L. María Sierra Isabel Gaivão Editors

Home , Ames test

Methods in Pharmacology and Toxicology

Genotoxicity and DNA Repair A Practical Approach M ETHODS IN PHARMACOLOGY AND TOXICOLOGY

Series Editor Y. James Kang University of Louisville School of Medicine Prospect, Kentucky, USA

For further volumes: http://www.springer.com/series/7653

Genotoxicity and DNA Repair

A Practical Approach

Edited by L. María Sierra

Universidad de Oviedo, Oviedo, Spain Isabel Gaivão

Universidade de Trás-os-Montes e Alto Douro, Vila Real, Portugal Editors L. María Sierra Isabel Gaivão Universidad de Oviedo Universidade de Trás-os-Montes e Alto Douro Oviedo, Spain Vila Real, Portugal

ISSN 1557-2153 ISSN 1940-6053 (electronic) ISBN 978-1-4939-1067-0 ISBN 978-1-4939-1068-7 (eBook) DOI 10.1007/978-1-4939-1068-7 Springer New York Heidelberg Dordrecht London

Library of Congress Control Number: 2014940854

© Springer Science+Business Media New York 2014 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifi cally the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfi lms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifi cally for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s location, in its current version, and permission for use must always be obtained from Springer. Permissions for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecution under the respective Copyright Law. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specifi c statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. While the advice and information in this book are believed to be true and accurate at the date of publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein.

Printed on acid-free paper

Humana Press is a brand of Springer Springer is part of Springer Science+Business Media (www.springer.com) Preface

When invited to work on, design, and edit a book on “Genotoxicity and DNA Repair,” our goal was to compile a volume that would provide a reference for determining how to analyze the genotoxic activity of molecules or materials and, at the same time, serve as a practical tool for researchers in the Environmental Mutagenesis and DNA Repair fi elds. Because of this, we have focused on genotoxicity assays recommended by the “OECD guidelines for the testing of chemicals,” presenting both theoretical information and updated standard protocols, as well as modifi ed protocols that could be of use in specifi c situations. In addition, we have also covered other assays not included in the OECD guidelines but of proven usefulness in the fi eld, such as gene mutation assays, the comet assay—in different species and applications—and the SMART assays in Drosophila. Throughout the book, special emphasis is placed on the analysis of nanoparticles and nanomaterials. With respect to DNA repair, we have included several assays that give information on repair activity in vitro and recent applications to study repair in humans. This part does not set out to be exhaustive, but aims to be of help when the analysis of DNA repair is necessary. We were fortunate enough not only to obtain the approval of the publisher, but also and especially to secure the interest and commitment of relevant scientists in the fi eld who agreed to write the different chapters. We thank all of them heartily for their support, their patience with our questions and requests, and their excellent work. We hope that you will enjoy the result and fi nd it as useful as we intended it to be.

Oviedo, Spain L. María Sierra Vila Real, Portugal Isabel Gaivão

Contents

Preface...... v Contributors ...... ix

PART I GENOTOXICITY ASSAYS

1 Ames Test (Bacterial Reverse Mutation Test): Why, When, and How to Use ...... 3 Araceli Pillco and Eduardo de la Peña 2 The Ames II and Ames MPF Penta I Assay: A Liquid Microplate Format Modification of the Classic Ames Test ...... 23 Sini Flückiger-Isler and Markus Kamber 3 Revised Procedure of the Bacterial Reverse Mutation Test for Genotoxic Evaluation of Nanoparticles ...... 43 Christophe Pagnout, Stéphane Jomini, and Pascale Bauda 4 The Control of Hydrophobic Compound Exposure in In Vitro Tests for Genotoxicity ...... 59 Kilian E. C. Smith 5 The In Vitro Micronucleus Assay and FISH Analysis ...... 73 Lucia Migliore, Sebastiano Di Bucchianico, and Chiara Uboldi 6 The In Vivo Rodent Micronucleus Test...... 103 Edson Luis Maistro 7 Chromosomal Aberration Test Utilities In Vitro and In Vivo...... 115 Ana Paula A. Guimarães, Adriana C. Guimarães, Diego Á. Alcântara, Luiz Raimundo Cunha, Patrícia L. Lima, Marne C. Vasconcellos, Raquel C. Montenegro, Bruno M. Soares, Marucia M. Amorim, and Rommel R. Burbano 8 Chromosome Painting of Mouse Peripheral Blood and Spleen Tissues...... 141 Dayton M. Petibone, James D. Tucker, and Suzanne M. Morris 9 T-Cell Receptor Mutation Assay for Monitoring Human Genotoxic Exposure ...... 159 Seishi Kyoizumi 10 The Human RBC PIG-A Gene Mutation Assay...... 169 Vasily N. Dobrovolsky and Robert H. Heflich 11 The Applicable Use of the HPRT Gene Mutation Assay as a Practical Tool in Mutagenesis and DNA Repair Studies...... 185 Zoulikha M. Zaïr and George E. Johnson 12 The Comet Assay: High Throughput Use of FPG ...... 199 Amaya Azqueta and Andrew R. Collins

vii viii Contents

13 The Comet Assay In Vivo in Humans ...... 219 Carla Costa and João Paulo Teixeira 14 Analysis of Nanoparticle-Induced DNA Damage by the Comet Assay ...... 241 Julia Catalán, Satu Suhonen, Anna Huk, and Maria Dusinska 15 The Comet Assay in Drosophila: Neuroblast and Hemocyte Cells ...... 269 L. María Sierra, Erico R. Carmona, Leticia Aguado, and Ricard Marcos 16 The SMART Assays of Drosophila: Wings and Eyes as Target Tissues ...... 283 Ricard Marcos, L. María Sierra, and Isabel Gaivão 17 Testing the Genotoxic Potential of Nanomaterials Using Drosophila ...... 297 Mohamed A. Abdalaziz, Balasubramanyam Annangi, and Ricard Marcos 18 Transgenic Rodent Gene Mutation Assay in Somatic Tissues ...... 305 John D. Gingerich, Lynda Soper, Christine L. Lemieux, Francesco Marchetti, and George R. Douglas 19 The Mouse Lymphoma Assay ...... 323 Tao Chen, Xiaoqing Guo, and Martha M. Moore 20 Bhas 42 Cell Transformation Assay for Genotoxic and Non- Genotoxic Carcinogens ...... 343 Kiyoshi Sasaki, Anna Huk, Naouale El Yamani, Noriho Tanaka, and Maria Dusinska

PART II DNA REPAIR ASSAYS

21 Methods for Measuring DNA Repair: Introduction and Cellular Repair...... 365 Andrew R. Collins and Amaya Azqueta 22 A Standardized Protocol for the In Vitro Comet-Based DNA Repair Assay ...... 377 Jana Slyskova, Sabine A.S. Langie, Isabel Gaivão, Andrew R. Collins, and Amaya Azqueta 23 Use of the Comet Assay to Study DNA Repair in Drosophila melanogaster . . . . 397 Isabel Gaivão, Rubén Rodríguez, and L. María Sierra 24 Use of RNA Interference to Study DNA Repair ...... 413 Elise Fouquerel, Jianfeng Li, Andrea Braganza, Zhongxun Yu, Ashley R. Brown, Xiao-Hong Wang, Sandy Schamus, David Svilar, Qingming Fang, and Robert W. Sobol

PART III OTHERS

25 The LacZ Plasmid-Based Transgenic Mouse Model: An Integrative Approach to Study the Genotoxicity of Nanomaterials ...... 451 Henriqueta Louro, Miguel Pinto, Nádia Vital, Ana M. Tavares, Pedro M. Costa, and Maria João Silva

Index ...... 479 Contributors

MOHAMED A. ABDALAZIZ • Group of Mutagenesis, Department of Genetics and Microbiology , Universitat Autònoma de Barcelona , Barcelona , Spain LETICIA AGUADO • Genetics Area, Department of Functional Biology , University of Oviedo , Oviedo , Spain ; University Oncology Institute from the Principado de Asturias , University of Oviedo , Oviedo , Spain DIEGO Á. ALCÂNTARA • Laboratório de Citogenética Humana, Instituto de Ciências Biológicas, Universidade Federal do Pará , Pará, Brazil MARUCIA M. AMORIM • Laboratório de Citogenética Humana, Instituto de Ciências Biológicas, Universidade Federal do Pará , Pará, Brazil ; Universidade da Amazônia , Belém , Brazil BALASUBRAMANYAM ANNANGI • Group of Mutagenesis, Department of Genetics and Microbiology , Universitat Autònoma de Barcelona , Barcelona , Spain AMAYA AZQUETA • Department of Pharmacology and Toxicology , University of Navarra , Spain PASCALE BAUDA • Laboratoire Interdisciplinaire des Environnements Continentaux (LIEC), CNRS 7360, Université de Lorraine , Metz , France ANDREA BRAGANZA • Department of Pharmacology & Chemical Biology , University of Pittsburgh School of Medicine , Pittsburgh, PA, USA ; Hillman Cancer Center, University of Pittsburgh Cancer Institute , Pittsburgh, PA, USA ASHLEY R. BROWN • Hillman Cancer Center , University of Pittsburgh Cancer Institute , Pittsburgh, PA , USA ROMMEL R. BURBANO • Laboratório de Citogenética Humana, Instituto de Ciências Biológicas, Universidade Federal do Pará , Pará , Brazil ERICO R. CARMONA • Núcleo de Investigación en Estudios Ambientales, Grupo de Genotoxicología, Escuela de Ciencias Ambientales, Facultad de Recursos Naturales, Universidad Católica de Temuco , Temuco, Chile JULIA CATALÁN • Nanosafety Research Center and Systems Toxicology, Health and Work Ability , Finnish Institute of Occupational Health , Helsinki , Finland TAO CHEN • Division of Genetic and Molecular Toxicology, National Center for Toxicological Research, US Food and Drug Administration , Jefferson , AR , USA ANDREW R. COLLINS • Department of Nutrition, University of Oslo , Oslo, Norway CARLA COSTA • Environmental Health Department, Portuguese National Institute of Health , Porto , Portugal PEDRO M. COSTA • Departmento de Genética Humana, Instituto Nacional de Saúde Dr. Ricardo Jorge (INSA) , Lisboa, Portugal; IMAR - Instituto do Mar, Departamento de Ciências e Engenharia do Ambiente, Faculdade de Ciências e Tecnologia, Universidade Nova de Lisboa, Caparica, Portugal LUIZ RAIMUNDO CUNHA • Laboratório de Citogenética Humana, Instituto de Ciências Biológicas, Universidade Federal do Pará , Pará , Brazil SEBASTIANO DI BUCCHIANICO • Department of Translational Research and New Technologies in Medicine and Surgery, Medical Genetics Unit, University of Pisa , Pisa, Italy

ix x Contributors

VASILY N. DOBROVOLSKY • Division of Genetic and Molecular Toxicology, National Center for Toxicological Research, US Food and Drug Administration , Jefferson , AR , USA GEORGE R. DOUGLAS • Environmental Health Science and Research Bureau, Health Canada , Ottawa, ON , Canada MARIA DUSINSKA • Health Effects Laboratory, Department of Environmental Chemistry , NILU (Norwegian Institute for Air Research) , Kjeller , Norway NAOUALE EL YAMANI • Health Effects Laboratory, Department of Environmental Chemistry , NILU (Norwegian Institute for Air Research) , Kjeller , Norway QINGMING FANG • Department of Pharmacology & Chemical Biology , University of Pittsburgh School of Medicine , Pittsburgh , PA, USA ; Hillman Cancer Center, University of Pittsburgh Cancer Institute , Pittsburgh , PA, USA SINI FLÜCKIGER-ISLER • Xenometrix AG , Allschwil , Switzerland ELISE FOUQUEREL • Department of Pharmacology & Chemical Biology , University of Pittsburgh School of Medicine , Pittsburgh , PA, USA ; Hillman Cancer Center, University of Pittsburgh Cancer Institute , Pittsburgh , PA, USA ISABEL GAIVÃO • Department of Genetics and Biotechnology, CECAV , University of Trás-os-Montes and Alto Douro , Vila Real, Portugal JOHN D. GINGERICH • Environmental Health Science and Research Bureau, Health Canada , Ottawa , ON , Canada ADRIANA C. GUIMARÃES • Laboratório de Citogenética Humana, Instituto de Ciências Biológicas, Universidade Federal do Pará , Pará , Brazil ANA PAULA A. GUIMARÃES • Centro de Ciências Biológicas e da Saúde, Universidade do Estado do Pará , Pará , Brazil ; Laboratório de Citogenética Humana, Instituto de Ciências Biológicas, Universidade Federal do Pará , Pará , Brazil XIAOQING GUO • Division of Genetic and Molecular Toxicology, National Center for Toxicological Research, US Food and Drug Administration , Jefferson , AR , USA ROBERT H. HEFLICH • Division of Genetic and Molecular Toxicology, National Center for Toxicological Research, US Food and Drug Administration , Jefferson , AR , USA ANNA HUK • Health Effects Laboratory, Department of Environmental Chemistry , NILU (Norwegian Institute for Air Research) , Kjeller , Norway GEORGE E. JOHNSON • College of Medicine, Institute of Life Science, Swansea University , Swansea, UK STÉPHANE JOMINI • Laboratoire Interdisciplinaire des Environnements Continentaux (LIEC), CNRS 7360, Université de Lorraine , Metz , France MARKUS KAMBER • Xenometrix AG , Allschwil, Switzerland SEISHI KYOIZUMI • Laboratory of Immunology, Department of Radiobiology/Molecular Epidemiology , Radiation Effects Research Foundation , Hiroshima , Japan SABINE A.S. LANGIE • Environmental Risk and Health Unit, Flemish Institute of Technological Research (VITO) , Mol, Belgium CHRISTINE L. LEMIEUX • Environmental Health Science and Research Bureau, Health Canada , Ottawa , ON , Canada ; Water and Air Quality Bureau, Health Canada , Ottawa, ON, Canada JIANFENG LI • Department of Pharmacology & Chemical Biology, University of Pittsburgh School of Medicine , Pittsburgh , PA , USA ; Hillman Cancer Center, University of Pittsburgh Cancer Institute , Pittsburgh , PA , USA PATRÍCIA L. LIMA • Centro de Ciências Biológicas e da Saúde, Universidade do Estado do Pará , Pará , Brazil ; Laboratório de Citogenética Humana, Instituto de Ciências Biológicas, Universidade Federal do Pará , Pará , Brazil Contributors xi

HENRIQUETA LOURO • Departmento de Genética Humana, Instituto Nacional de Saúde Dr. Ricardo Jorge (INSA) , Lisboa, Portugal EDSON LUIS MAISTRO • Faculdade de Filosoﬁ a e Ciências, Departamento de Fonoaudiologia, Universidade Estadual Paulista, UNESP , Marilia , SP , Brazil FRANCESCO MARCHETTI • Environmental Health Science and Research Bureau, Health Canada , Ottawa , ON , Canada RICARD MARCOS • Department of Microbiology and Genetics, Group of Mutagenesis, Faculty of Biosciences, Autonomous University of Barcelona , Barcelona , Spain LUCIA MIGLIORE • Department of Translational Research and New Technologies in Medicine and Surgery, Medical Genetics Unit, University of Pisa , Pisa, Italy RAQUEL C. MONTENEGRO • Laboratório de Citogenética Humana, Instituto de Ciências Biológicas, Universidade Federal do Pará , Pará, Brazil MARTHA M. MOORE • ENVIRON International Corporation, Little Rock, AR , USA SUZANNE M. MORRIS • Division of Genetic and Molecular Toxicology, National Center for Toxicological Research, US Food and Drug Administration , Jefferson, AR, USA CHRISTOPHE PAGNOUT • Laboratoire Interdisciplinaire des Environnements Continentaux (LIEC), CNRS 7360, Université de Lorraine , Metz , France EDUARDO DE LA PEÑA • Laboratory of Environmental Mutagenesis , Spanish National Research Council (CSIC) , Madrid, Spain DAYTON M. PETIBONE • Division of Genetic and Molecular Toxicology, National Center for Toxicological Research, US Food and Drug Administration , Jefferson, AR, USA ARACELI PILLCO • Laboratory of Environmental Mutagenesis , Spanish National Research Council (CSIC) , Madrid , Spain MIGUEL PINTO • Departmento de Genética Humana, Instituto Nacional de Saúde Dr. Ricardo Jorge (INSA) , Lisboa, Portugal RUBÉN RODRÍGUEZ • Genetics Area, Department of Functional Biology , University of Oviedo , Oviedo , Spain ; Universitary Oncology Institute from the Principado de Asturias , University of Oviedo , Oviedo, Spain ; Institute of Functional Biology and Genomics (IBFG), University of Salamanca-CSIC, IBSAL , Salamanca , Spain KIYOSHI SASAKI • Laboratory of Cell Carcinogenesis, Division of Alternative Toxicology Test , Hatano Research Institute, Food and Drug Safety Center , Kanagawa, Japan SANDY SCHAMUS • Hillman Cancer Center , University of Pittsburgh Cancer Institute , Pittsburgh, PA , USA L. MARÍA SIERRA • Genetics Area, Department of Functional Biology , University of Oviedo , Oviedo , Spain ; University Oncology Institute from the Principado de Asturias , University of Oviedo , Oviedo , Spain MARIA JOÃO SILVA • Departmento de Genética Humana, Instituto Nacional de Saúde Dr. Ricardo Jorge (INSA) , Lisboa, Portugal JANA SLYSKOVA • Department of Molecular Biology of Cancer , Institute of Experimental Medicine ASCR , Prague, Czech Republic KILIAN E. C. SMITH • Convergence Environment Team, Korean Institute of Science and Technology Europe , Saarbrücken , Germany BRUNO M. SOARES • Laboratório de Citogenética Humana, Instituto de Ciências Biológicas, Universidade Federal do Pará , Pará , Brazil ROBERT W. SOBOL • Department of Pharmacology & Chemical Biology , University of Pittsburgh School of Medicine , Pittsburgh , PA, USA ; Hillman Cancer Center, University of Pittsburgh Cancer Institute , Pittsburgh , PA , USA ; Department of Human Genetics, University of Pittsburgh Graduate School of Public Health , Pittsburgh , PA , USA xii Contributors

LYNDA SOPER • Environmental Health Science and Research Bureau, Health Canada , Ottawa, ON , Canada SATU SUHONEN • Nanosafety Research Center and Systems Toxicology, Health and Work Ability , Finnish Institute of Occupational Health , Helsinki, Finland DAVID SVILAR • Department of Pharmacology & Chemical Biology , University of Pittsburgh School of Medicine , Pittsburgh , PA , USA ; Hillman Cancer Center, University of Pittsburgh Cancer Institute , Pittsburgh , PA , USA NORIHO TANAKA • Laboratory of Cell Carcinogenesis, Division of Alternative Toxicology Test, Hatano Research Institute, Food and Drug Safety Center , Kanagawa , Japan ANA M. TAVARES • Departmento de Genética Humana, Instituto Nacional de Saúde Dr. Ricardo Jorge (INSA) , Lisboa , Portugal JOÃO PAULO TEIXEIRA • Environmental Health Department, Portuguese National Institute of Health , Porto , Portugal JAMES D. TUCKER • Department of Biological Sciences, Wayne State University , Detroit , MI , USA CHIARA UBOLDI • Department of Translational Research and New Technologies in Medicine and Surgery, Medical Genetics Unit , University of Pisa , Pisa, Italy MARNE C. VASCONCELLOS • Faculdade de Ciências Farmacêuticas , Universidade Federal do Amazonas , Manaus , Brazil NÁDIA VITAL • Departmento de Genética Humana, Instituto Nacional de Saúde Dr. Ricardo Jorge (INSA) , Lisboa, Portugal XIAO-HONG WANG • Hillman Cancer Center , University of Pittsburgh Cancer Institute , Pittsburgh, PA , USA ZHONGXUN YU • Hillman Cancer Center , University of Pittsburgh Cancer Institute , Pittsburgh, PA , USA ; School of Medicine, Tsinghua University , Beijing , China ZOULIKHA M. ZAÏR • Institute of Life Science, College of Medicine, Swansea University , Swansea, UK Part I

Genotoxicity Assays Chapter 1

Ames Test (Bacterial Reverse Mutation Test): Why, When, and How to Use

Araceli Pillco and Eduardo de la Peña

Abstract

The Salmonella typhimurium /mammalian microsome assay is the most widely used short-term test to identify genetic damage. This is used to assess the mutagenic and antimutagenic potential of compounds and mixtures. This assay uses histidine-dependent strains to detect mutations, e.g., substitutions, addi- tions, or deletions of one or several DNA nucleotides reverting originally changed gene sequence of the tester strains. The addition of a mutagenic chemical agent to a plate of cultured cells results in the growth of mutant colonies; the number of such colonies is an indicator of the mutagenic potency of the agent. The Ames test has many advantages, it is a very versatile assay, its different modiﬁ cations have been developed to determine mutagenic potencies, and it is recommended by several regulatory agencies. This chapter provides a detailed description of how the standard plate incorporation method should be performed, including the experimental design and interpretation of results.

Key words Ames test , Mutagenicity , Toxicity , Spontaneous revertant , Salmonella , Standard incorporation plate

1 Introduction

The bacterial reverse mutation test developed by Bruce Ames and his colleagues [ 1 , 2 ] is perhaps the most widely used short-term bioassay to identify genetic damage that leads to gene mutation. This is a simple tool that can be used to detect the mutagenic and antimutagenic potential of environmental chemicals, environmental mixtures, body ﬂ uids, foods, drugs, and physical agents [3 – 7 ]. This is a reverse mutation assay that employs histidine- dependent Salmonella strains with mutations at various genes in their histidine operon, that render them incapable of synthesizing the amino acid histidine. The strains restore their functional capability to synthesize histidine and to grow in the absence of the amino acid required by the parent strain [ 8 ]. This event occurs with low frequency. When the Salmonella tester strains are grown on a minimal media agar plate containing a trace of histidine, only

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_1, © Springer Science+Business Media New York 2014 3 4 Araceli Pillco and Eduardo de la Peña

Fig. 1 Mutagenic dose response with strain TA100, (a , b) control, spontaneous revertants, (c ) positive control, methyl methane sulfonate 1 μM/plate

those bacteria that revert to histidine independence are able to form colonies. The number of spontaneously induced revertant colonies per plate is relatively constant. However, when a mutagen is added to the plate, the number of revertant colonies is increased, usually in a dose-related manner [ 9 ] (see Fig. 1 ). The Ames test has many advantages to identify compounds or mixtures that cause gene mutations. For instance, it allows making replicates and obtaining results in a relatively short time, studying a large number of test materials inexpensively, and identifying the molecular mechanism effect of test materials. Also, although Salmonella is a prokaryotic organism, the combination of the cytochrome-based P450 metabolic oxidation system with the Ames test system allows determining some mutagenic agents, which are biologically inactive unless they are metabolized to active forms [ 4 , 6 , 10 ]. Apart from all the mentioned advantages, the Ames test system is very versatile, and many modiﬁ cations have been developed to determine mutagenic potencies, e.g., preincubation method, the desiccator assay, and microsuspension assay [ 6 , 11 – 15 ]. The most common assay procedures are the spot test and the standard plate incorporation method. The latter is described in detail in this chapter. Besides, the Ames test is required or recommended by regulatory agencies prior to registration or acceptance of many chemicals, drugs, and biocides [16 – 22 ]. Also the Ames test is recommended when it is necessary to perform a general screening, and it can be used as a backup in equivocal or negative results that are obtained in other experiments. According to Claxton et al. [ 7 ], the Ames test is commonly used to evaluate environmental samples, as well as agents associated with metabolism, or personal exposure, its use peaked in the 1980s but is still widespread. Relatively few publications have been associated with soil and sediment samples; papers of air samples Ames Test Protocol 5

follow the overall declining trend seen since 1983; and publications dealing with water reached a plateau starting in 1980 and have remained stable. However, reports dealing with natural substances have increased since the mid-1990s. This increase is due largely to a search for and analysis of antimutagens, mainly from plant extracts. The pharmaceutical industry has also tested thousands of substances in the bacterial reverse mutation assay [ 7 , 23 ].

1.1 Safety General laboratory safety items (biohazard waste bags, goggles or Considerations protective eye wear, gloves, lab coats, biohazard bags, protective covers for work space) should be applied. Also, it is important that basic bacteriological laboratory procedures are used to minimize exposure to the Salmonella tester strains. Though wild-type S. typhimurium can cause diarrhea and food poisoning, the gal and rfa (deep rough) mutations that are present in all the Salmonella tester strains described here eliminate, to different levels, the polysaccharide side chain of the lipopolysaccharide (LPS) layer that coats the bacterial surface, which makes the bacteria nonpathogenic [ 3 , 6 , 24 ]. It is nevertheless recommended to use caution at all times and to practice standard laboratory safety procedures such as using plugged pipettes and autoclaving all contaminated materials. Contamination by bacteria or fungi will interfere with the test. Contaminated frozen or working cultures of the Salmonella strains will render them unusable, as so will do contaminated overnight cultures. Surface areas must be properly disinfected before and after use. All cultures and labware used to handle the cultures must be autoclaved before being discarded. As a general rule, it is advised to consider all chemicals as if they were mutagens and carcinogens. All handling of chemicals, as well as the test itself, should be performed in a chemical safety cabinet. Workers should protect themselves from chemical exposure by wearing gowns, eyeglasses, and gloves. Wearers of contact lenses should wear regular eyeglasses since some volatile chemicals might react with the contact lens. All contaminated materials (e.g., test tubes, pipettes and pipette tips, gowns, and gloves) should be properly disposed of as well as the unused chemical dilutions and stock solutions of the test chemical, and positive control chemicals.

2 Materials

2.1 Supplies Autoclave and Equipment Automatic micropipettes (adjustable volumes up to 200 and 500 μL) Balances Boiling water bath or microwave oven Colony counter 6 Araceli Pillco and Eduardo de la Peña

Disposable spectrophotometer cuvettes Disposable tips General laboratory glassware Glass pipettes (1, 2, 5, and 10 mL) Laminar fl ow hood equipped with gas line Liquid and solid waste disposal Magnetic stir bars Magnetic stirrers Refrigerated centrifuge (9,000 × g ) Refrigerator (4 °C) and freezer (−20 °C) Shaking incubator set at 2 × g and 37 ºC Spectrophotometer for monitoring cell density Stationary incubator Sterile cryogenic storage vials Sterile glass test tubes (100 × 16 mm) and racks Sterile microbiological loops Sterile petri dishes (100 × 15 mm) Sterile syringes (5, 10, and 50 mL) Sterilizing membrane fi lters (0.22 and 0.45 μm) Ultralow-temperature freezer set at −80 °C or liquid nitrogen tank Vortexer Water bath set at 43–48 °C to maintain temperature of top agar Water purifi cation system

2.2 Reagents Use: to grow the tester strains overnight and Media: Recipes Ingredients per liter: 2.2.1 Nutrient Broth Difco bacto nutrient brotha 8 g Sodium chloride 5 g Distilled water 1,000 mL

Add the nutrient broth and sodium chloride to distilled water in a 2-L ﬂ ask containing a magnetic stir bar. Autoclave, loosely capped, for 20 min to 120 ºC. When cooled, store in the dark at room temperature. a If using Oxoid nutrient broth #2: To 1,000 mL of deionized water, add 25 g of nutrient broth powder (Oxoid) and the sodium chloride is omitted. Ames Test Protocol 7

2.2.2 Glucose Minimal Use: bottom agar for mutagenicity assay (GM) Agar Plates Ingredients per liter:

Agar 15 g Glucose solution (40 %) 50 mL Vogel-Bonner salts (50×) 20 mL Distilled water 930 mL

Add the agar to the distilled water in a 2-L ﬂ ask. Autoclave for 20 min at 120 ºC. When the solution has cooled slightly (about 65 ºC), add 20 mL of sterile 50× Vogel-Bonner salts and 50 mL of sterile 40 % glucose. After all the ingredients have been added, the solution should be stirred thoroughly. Dispense the agar medium in 100- × 15-mm petri dishes (approximately 25 mL/plate). When solidiﬁ ed, the plates can be stored at 4 ºC, but we recommend preparing new plates in every experiment. Note : The 50× Vogel-Bonner salts and 40 % glucose should be autoclaved separately.

2.2.3 Glucose Solution Use: as carbon source for the GM agar plates (40 % w/v) Ingredients per 100 mL:

Glucose 40 g Distilled water 100 mL

Weigh glucose and add cold distilled water. Stir immediately until completely dissolved. Autoclave at 120 ºC for 20 min, making sure the caps are on loosely. When cooled, tighten the caps and store at 4 ºC.

2.2.4 Vogel-Bonner Use: salts for the GM agar plates Salts (50×) Ingredients per liter:

Magnesium sulfate (MgSO4 ⋅ 7H2 0) 10 g

Citric acid monohydrate (C6 H8 O7 ⋅ H2 O) 100 g

Potassium phosphate, dibasic, anhydrous (K2 HPO4 ) 500 g

Sodium ammonium phosphate (NaHNH4 (PO4 ⋅ 4 H2 0)) 175 g Warm distilled water (about 45 ºC) 670 mL

Add the above ingredients in the order indicated to warm water in a 2-L beaker or ﬂ ask, making sure that each salt is dissolved thoroughly by stirring on a magnetic stirrer before adding the next salt. Adjust the volume to 1 L. Distribute in 250-mL aliquots. Autoclave, loosely capped, for 20 min at 120 ºC. When the solutions have cooled, tighten the caps and store at room temperature in the dark. 8 Araceli Pillco and Eduardo de la Peña

2.2.5 Top Agar Use: mutagenicity Supplemented Ingredients per liter: with Histidine/Biotin Agar 6 g Sodium chloride 5 g Histidine/biotin (0.5 mM) 100 mL Distilled water 900 mL

Add the agar and sodium chloride to distilled water in a ﬂ ask. Autoclave, loosely capped, for 20 min to 120 ºC. Then, in a laminar ﬂ ow hood, add 100 mL of limited histidine and biotin solution (0.5 mM). Dispense 200-mL aliquots in 500-mL screw-cap bottles. When ready to use, melt the top agar in a microwave oven or in boiling water.

2.2.6 Histidine/Biotin Use: to supplement top agar with excess biotin and a trace amount Solution (0.5 mM) of histidine Ingredients per 250 mL:

D -Biotin (F.W. 247.3) 0.0309 g

L -Histidine (F.W. 191.7) 0.0244 g Distilled water 250 mL

Dissolve the biotin and histidine by heating the water to the boiling point. Sterilize by ﬁ ltration through a 0.22-μm membrane ﬁ lter or autoclave 20 min at 120 ºC. Store at 4 ºC in a glass bottle.

2.2.7 Enriched GM Use: to provide medium supplemented with essential nutrients and Agar Plates antibiotics for the strain checking or propagation of the strains and preparation of stock culture master plates Prior to preparing the 1-L GM agar plates, add the following item(s):

● Biotin solution: 8 mL of 0.01 % solution. ● Histidine solution: 8 mL of 0.5 % solution. ● Histidine/biotin solution: 8 mL of 0.01 % biotin and 0.5 % histidine solution. ● Histidine/biotin/ampicillin solution: same as histidine/biotin solution, but add 3 mL of ampicillin solution (8 mg/mL) to give a ﬁ nal concentration of ampicillin of 24 μg/mL. ● Histidine/biotin/tetracycline solution: same as histidine/ biotin solution, but add 0.25 mL of tetracycline solution (8 mg/mL), which will give a ﬁ nal concentration of 2 μg/mL. Note: Mix well before dispensing. The solutions could be stored protected from direct light up to 1 year at 4 ºC. Ames Test Protocol 9

2.2.8 Crystal Violet Use: to conﬁ rm the presence of the rfa mutation in all the tester Solution (0.1 % w/v) strains Ingredients per 100 mL:

Crystal violet 0.1 g Distilled water 100 mL

Dissolve the crystal violet in the 100 mL of water. Mix well and store at 4 ºC in a lightproof glass bottle with screw cap.

2.2.9 Metabolic Use: metabolic activation Activation System Ingredients per 50 mL: (S-9 Mix) Standard S-9 mix High S-9 mix

Phosphate buffer (0.2 M, pH 7.4) 25 mL 25 mL NADPa 0.1 M 2 mL 2 mL

D -Glucose-6-phosphate (1 M) 0.25 mL 0.25 mL

KCl (1.65 M)–MgCl2 (0.4 M) salts 1 mL 1 mL Liver S-9 fraction 2 mL (4 %) 5 mL (10 %) Distilled water 19.75 mL 16.75 mL a Nicotinamide adenine dinucleotide phosphate This mixture must be prepared in a laminar fl ow hood and all components must be kept in an ice bath. Add components aseptically into screw-cap tubes, or fl asks, under volume intended for preparation, take care of adding the components in the order presented above (S-9 to last), and maintain fi nal solution in an ice bath or refrigerator at 4 ± 2 °C until used in the test. Any leftover of S-9 or S-9 mix should be discarded. It is recommended not to refreeze.

2.2.10 Phosphate Buffer Use: S-9 mix (0.2 M, pH 7.4) Ingredients per 500 mL:

Sodium dihydrogen phosphate (0.2 M) 60 mLa

(NaH2 PO ⋅ 4H2 O) (13.8 g/500 mL) Disodium hydrogen phosphate (0.2 M) 440 mLa

(Na 2 HPO4 ) (14.2 g/500 mL) a These are approximate values. Adjust pH, as needed, using dibasic sodium phosphate. Sterilize by autoclaving for 20 min at 121 ºC. Store up to 1 year at 4 ºC.

2.2.11 Nicotinamide Use: S-9 mix Adenine Dinucleotide Ingredients per 10 mL: Phosphate (NADP) (0.1 M) NADP (F.W. 765.4) 0.07654 ga Sterile distilled water 10 mL 10 Araceli Pillco and Eduardo de la Peña

Dissolve the NADP in sterile distilled water. Sterilize membrane ﬁ lter (0.45 μm) in a screw-cap vial previously sterilized. This solution must be prepared the same day and in the volume needed for the experiment. a This amount of NADP applies to a formula weight of 765.4. Check the corrected formula weight indicated for each lot of NADP.

2.2.12 D -Glucose-6- Use: S-9 mix Phosphate (1 M) Ingredients per 10 mL:

Glucose-6-phosphate 2.82 g Sterile distilled water 10 mL

Dissolve the glucose-6-phosphate in sterile distilled water. Sterilized membrane ﬁ lter (0.45 μm) in a presterilized disposable vial. Tubes containing a solution of glucose-6-phosphate must be stored in freezer (−20 ± 2 °C) for a maximum period of 60 days.

2.2.13 Salt Solution (KCl Use: S-9 mix (1.65 M)–MgCl2 (0.4 M)) Ingredients per 500 mL:

Potassium chloride (KCl) 61.5 g

Magnesium chloride (MgCl2 ⋅ 6H2 0) 40.7 g Distilled water To ﬁ nal volume of 500 mL

Dissolve the potassium chloride and magnesium chloride in distilled water. Autoclave for 20 min at 120 °C. The salt solution must be stored under refrigeration (2–8 °C) for a maximum period of 60 days.

2.2.14 Liver S-9 Fraction see support protocol 4.1 .

3 Methods

3.1 Reception If the new strain is received on a small sterile ﬁ lter disk embedded of New Strains in nutrient agar: 1. Transfer the disk to 5 mL of nutrient broth. 2. Incubate the culture overnight at 37 °C with shaking. Note : The overnight culture should contain 1–2 × 10 9 colony

forming units (cfu)/mL (O.D. 540 between 0.1 and 0.2). 3. Check the broth cultures for bacterial growth. If the new strain comes as a lyophilized culture: 1. Add 1 mL of sterile nutrient broth to rehydrate the culture. 2. Transfer the rehydrated culture to 4 mL of nutrient broth. Ames Test Protocol 11

3. Incubate the culture overnight at 37 °C with shaking. 4. Check the broth culture for bacterial growth.

3.2 Prior Steps It is recommended to perform the following steps: to the Standard (a) Select the bacterial strains. Incorporation Plate A set of histidine-requiring strains is used for mutagenicity Procedure testing. The most common tester strains are TA97, TA98, TA100, TA102, TA104, TA1535, TA1537, and TA1538. The strains also have additional mutations and genetic alterations that greatly increase their ability to detect mutagens. These are uvrB and rfa mutations, and introduction of pKM101 and pAQ1 plasmids. The uvrB mutation, which is present in all strains except TA102, arises from a deletion-type mutation through the uvrB-bio genes that eliminates the accurate DNA repair and makes the cells biotin dependent. All strains have the rfa mutation that affects the bacterial cell wall, resulting in a defective LPS layer that provides more permeability to bulky chemicals. Existence of pKM101 plasmid in TA97, TA98, TA100, and TA102 provides ampicillin resistance and sensitivity for chemical and induced mutagenesis associated with error-prone DNA repair pathway. TA102 strain also has a pAQ1 plasmid that carries a tetracycline resistance gene and a copy of hisG428 in the context of a multicopy plasmid. The mutation hisG428 was introduced in strainTA102 with the aim of amplifying the number of target sites. To enhance the ability of this strain to detect DNA crosslinking agents, the uvrB gene was retained, making the bacterium DNA repair proﬁ cient [6 , 10 , 25 ] (Table 1 ).

Table 1 Genotypic properties of the most common Salmonella tester strains

Strain Histidine mutation DNA target uvrB rfa Plasmid Reversion event

TA97 hisD6610 CCCCCC + + pKM101 Frameshifts TA98 hisD3052 CGCGCGCG + + pKM101 Frameshifts TA100 hisG46 GGG + + pKM101 Base-pair substitutions TA102 hisG428 TAA – + pKM101 Base-pair substitutions pAQ1 TA104 hisG428 TAA + + – Base-pair substitutions TA1535 hisG46 GGG + + – Base-pair substitutions TA1537 hisC3076 CCCCC + + – Frameshifts TA1538 hisD3052 CGCGCGCG + + – Frameshifts 12 Araceli Pillco and Eduardo de la Peña

In addition to the Salmonella tester strains, many laboratories employ Escherichia coli strain WP2 uvrA pKM101 as a bacterial tester strain in the bacterial reverse mutation test. This E. coli strain is similar in mutagen detection to S. typhimurium strain TA102. The E. coli strain carries a mutant gene that prevents synthesis of the essential amino acid tryptophan. Therefore, this strain can only survive and grow on medium that contains excess tryptophan. However, in the presence of a mutagenic chemical, the defective genes may be mutated back to the functional state, allowing the bacterium to grow on standard medium that does not contain supplemental tryptophan. The E. coli WP2 assay procedures are the same as those described for the Ames Salmonella assay with the exception that limited tryptophan instead of limited histidine is used [5 ]. (b) Select the positive controls. Each experiment should include solvent controls and diagnostic positive control chemicals speciﬁ c for each strain and for the metabolic activation system. Table 2 lists the most representative positive controls without metabolic activation and Table 3 those with metabolic activation. (c) Perform a genetic analysis of the bacterial strains. The analysis should be performed immediately after receiving the cultures, when a new set of frozen permanent or lyophilized cultures is prepared, when the number of spontaneous revertants per plate falls out of the normal range, or when there is a loss of sensitivity to standard mutagens. 1. Inoculate 5 mL of nutrient broth with a single colony and incubate this culture overnight at 37 °C. 2. Check the histidine dependence of all tester strains. 2.1 Streak a loop of the overnight culture on the surface of a GM agar plate enriched with excess of biotin. 2.2 Incubate at 37 °C for 48 h. 2.3 Check the agar plates. Note : No growth on plate is expected because all the Salmonella strains are histidine dependent. 3. Check the biotin dependence of all tester strains except TA102 strain. 3.1 Streak a loop of the overnight culture on the surface of a GM agar plate enriched with excess of histidine. 3.2 Incubate at 37 °C for 48 h. 3.3 Check the agar plates. Note: No growth on plate is expected, except for TA102 which is a biotin-nondependent strain. Ames Test Protocol 13

Table 2 Recommended positive co ntrol chemicals without metabolic activation

Strain Control chemicala Dosesb (μg/per plate) References

TA97 9-AA 50–100 [ 29 , 30 ] MMS 1 [ 31 ] 4-NPD 200 [ 32 ] TA98 4-NQO 0.5 [ 29 , 33 ] MMS 1* [ 34 ] 2-NF 10 [ 35 ] 4-NPD 200 [ 32 ] TA100 MMS 1* [ 33 , 34 ]

NaN3 1.5–10 [29 , 31 , 32 , 35 , 36 ] MNG 10 [ 36 ] TA102 MMS 1* [ 33 ] MMC 0.5–1 [ 32 , 36 ] TA104 MG 50 [ 29 , 33 ] TA1535 MMS 250 [ 30 ]

NaN3 1.5–5 [29 , 36 ] TA1537 9-AA 50 [ 30 ] TNF 0.1 [ 36 ] TA1538 4-NPD 2.5 [ 30 ] TNF 0.1 [ 36 ] a 9AA 9-aminoacridine, MMS methyl methane sulfonate, 4-NPD 4-nitro-o- phenilenediamine, 4-NQO 4-nitroquinoline-1-oxide, 2-NF 2-nitroﬂ uorene, NaN 3 sodium azide, MNG methyl nitroso guanidine, MG methylglyoxal, TNF 2,4,7-trinitro-9-ﬂ uorenone b Concentration based on 100- × 15-mm petri plate containing 20–25 mL of GM agar * 1 microliter/per plate

4. Check the biotin and histidine dependence of all tester strains. 4.1 Streak a loop of the overnight culture on the surface of a GM agar plate enriched with excess of biotin and histidine. 4.2 Incubate at 37 °C for 48 h. 4.3 Check the agar plates. Note : Growth on all plates is expected. 5. Check rfa mutation ( defective LPS) of all tester strains. 5.1 Streak a loop of the overnight culture on the surface of a GM agar plate enriched with excess of biotin and histidine. 5.2 Place a sterile ﬁ lter paper disk in the middle of the plate and apply 10-μL crystal violet solution (0.1 %, w/v) onto the disk. 14 Araceli Pillco and Eduardo de la Peña

Table 3 Recommended positive control chemicals with metabolic activation

Strain Control chemicala Dosesb (μg/plate) References

TA97 2-AA 1–15 [ 29 , 30 ] BaP 1 [ 31 ] TA98 2-AA 1–15 [ 29 , 30 ] BaP 1–10 [ 31 , 34 , 36 ] 2-AF 10 [ 35 , 36 ] TA100 2-AA 1–15 [ 29 , 30 ] BaP 1–10 [ 31 , 34 , 36 ] 2-AF 10 [ 35 , 36 ]

NaN 3 10 [ 32 ] TA102 2-AA 5–10 [ 30 , 32 ] TA104 2-AA 5–15 [ 29 , 30 ] TA1535 2-AA 2–15 [ 29 , 30 ] TA1537 2-AA 2–10 [ 30 ] 2-AF 10 [ 36 ] BaP 10 [ 36 ] TA1538 2-AA 2–10 [ 30 ] 2-AF 10 [ 36 ] BaP 10 [ 36 ]

a 2-AA 2-aminoanthracene, BaP benzo[a]pyrene, 2-AF 2-aminoﬂ uorene, NaN 3 sodium azide b Concentration based on 100- × 15-mm petri plate containing 20–25 mL of GM agar

5.3 Incubate at 37 °C for 48 h. 5.4 Check the agar plates. Note: An inhibition halo surrounding the disk is expected. 6. Check uvrB mutation of all tester strains. 6.1 Streak a loop of the overnight culture on the surface of a GM agar plate enriched with excess of biotin and histidine. 6.2 Unseal the top and cover half of the plate with sterile aluminum foil. Expose the plate to a low level of UV irradiation (15-W lamp at a distance of 35 cm) for a short time (approx. 8–10 s) that kills the uvrB strain but not its isogenic DNA repair proﬁ cient strain. 6.3 Incubate at 37 °C for 48 h. 6.4 Check the agar plates. Note : Normal growth on the nonexposed part of the plate is expected, but not on the exposed part. Ames Test Protocol 15

7. Check the presence of plasmid pKM101 (ampicillin resistance). 7.1 Streak a loop of the overnight culture on the surface of a GM agar plate enriched with excess of biotin and histidine and 24 μg/mL of ampicillin. Another way is to use a sterile fi lter paper disk containing 10-μg ampicillin in the middle. It can be placed on a streak of the strain on a GM agar plate supplemented with an excess of histidine and biotin. 7.2 Incubate at 37 °C for 48 h. 7.3 Check the agar plates. Note: Growth only in the ampicillin-resistant pKM101 strains is expected (TA97, TA98, TA100, TA102, and TA104). 8. Check the presence of plasmid pAQ1 (tetracycline resistance). 8.1 Streak a loop of a TA102 overnight culture on the surface of a GM agar plate enriched with excess of biotin and histidine and 2 μg/mL of tetracycline. Another way is to use a sterile fi lter paper disk containing 2-μg tetracycline in the middle. It can be placed on a streak of the strain on a GM agar plate supplemented with an excess of histidine and biotin. 8.2 Incubate at 37 °C for 48 h. 8.3 Check the agar plates. Note : Growth on plates is expected, which demonstrates the presence of pAQ1 plasmid. (d) Determine the viability assay and test concentrations. 1. Establish a total of eight concentrations, spaced in half-log intervals, with the highest dose limited by solubility, or by an arbitrary value (usually 5,000 or 10,000 μg/plate). 2. Employ the standard incorporation plate procedure with and without metabolic activation system. Positive and solvent control chemicals should be included. 3. Analyze toxic characteristics: a substantial decrease in the number of revertant colonies on the test plates; absence of the background bacterial lawn growth; and presence of a sparse bacterial lawn with pinpoint-size visible colonies. 4. A minimum of fi ve dose levels covering a range of at least three logs should be selected for the defi nitive test. For toxic chemicals, only the highest dose used should exhibit toxicity. For nontoxic chemicals, a high dose of 5,000 or 10,000 μg/ plate is acceptable. 16 Araceli Pillco and Eduardo de la Peña

Non-soluble chemicals may be tested as a suspension up to a dose level that does not interfere with handling of the suspension (e.g., risk of losing some of the sample into pipettes or pipette tips). Note: It is recommended to determine an appropriate dose range to evaluate the mutagenicity assay. The strain TA100 with and without metabolic activation is enough to perform the toxicity assay when more than one tester strain will be used. When multiple chemicals with similar properties are to be tested, it may be sufﬁ cient to perform a toxicity test on one representative chemical and use these results to estimate the dose range to be used with the other chemicals. Each test should be performed using a single batch of reagents, media, and agar to avoid variation in the results. (e) Verify the spontaneous mutation rate. Each tester strain reverts spontaneously at a frequency that is characteristic of the strain. Each laboratory has a characteristic range of revertant colonies, which is referred to as “historical control values.” Table 4 shows a sample of acceptable control values per plate with and without metabolic activation for the most common Salmonella tester strains. If spontaneous control values fall outside an acceptable historical range, the genetic integrity of the strain(s) is considered compromised, and a new culture should be isolated [6 ].

Table 4 Spontaneous control mutation rates for the Salmonella tester strains

Number of revertants

Maron and Ames [2 ] Mortelmans and Zeiger [30 ] Pillco [33 ]

Strain −S9 +S9 −S9 +S9 −S9 +S9

TA97 90–180 – 75–200 100–200 – – TA98 30–50 – 20–50 20–50 22–80 22–80 TA100 120–200 – 75–200 75–200 60–220 60–220 TA102 240–320 – 100–300 200–400 240–320 288–350 TA104 – – 200–300 300–400 275–425 275–425 TA1535 – – 5–20 5–20 – – TA1537 – – 5–20 5–20 – – TA1538 – – 5–20 5–20 – – Ames Test Protocol 17

3.3 Protocol The standard plate incorporation test consists of exposing the for Standard Plate Salmonella strains to the tested chemical directly on a glucose Incorporation Test minimal agar plate, usually in the presence and absence of an exogenous metabolic activation system. The different components are ﬁ rst added to sterile test tubes containing 2 mL of molten top agar supplemented with limited histidine and biotin. The contents of the tubes are mixed and poured on GM agar plates. After the top agar has hardened, the plates are incubated in an inverted orienta- tion for 48 h. After incubation, the revertant colonies are counted on all plates and compared to the number of spontaneous revertant colonies on solvent control plates. Figure 2 depicts the setup for performing the standard plate incorporation test. 1. Inoculate 0.05 mL of tester strain(s) cultures from frozen per- manents in 50 mL of nutrient broth and incubate for 12–16 h at 37 ºC. This is achieved by gently shaking (2 × g ). Care should be taken that the size of the culture bottle be at least three to ﬁ ve times the volume of the culture medium to ensure adequate aeration. 2. Prepare an appropriate number of GM agar plates and sterile glass test tubes for each test chemical, taking into consideration that each experiment should contain a series of duplicate or triplicate plates for: ● Negative control (solvent) ● Five or more concentrations of the test substance (diluted to at least half-log intervals), with the highest dose limited by toxicity or solubility ● Positive controls 3. Prepare the exogenous metabolic activation system (S-9) and keep it on ice until use. 4. Prepare the dilution of the test compounds. 5. Melt top agar supplemented with 0.05 mM histidine and biotin and maintain at 43–48 ºC. 6. Add the following items respectively into sterile glass tubes maintained at 43 ºC: ● 2 mL of molten top agar ● 0.10 or 0.50 mL of metabolic activation (S-9) mix or buffer ● 0.05 mL of the test compound dilution or control ● 0.10 mL of overnight culture of the Salmonella strain 7. Pour contents of each test tube onto the surface of the corresponding GM plate and gently swirl to evenly distribute the molten top agar. 18 Araceli Pillco and Eduardo de la Peña

Fig. 2 Diagram of the steps involved in the standard incorporation plate procedure

8. Invert plates when the top agar has hardened and place in an incubator at 37 ºC for 48 h. If after 48 h growth retardation is observed (smaller than anticipated colony sizes), the plates could be incubated for an additional 12–24 h. 9. Count colonies either manually or by an electronic colony counter. 10. Express the test results as number of revertant colonies per plate. Note: Usually the results are reported as mean revertant colonies per plate ± the S.D. Ames Test Protocol 19

3.4 Data Evaluation The evaluation data criterion that has been widely used is to set a minimum fold increase (usually 2–3-fold, over the solvent/negative control) as the cutoff between mutagenic and non-mutagenic response [26 ]. However, according to Zeiger et al. [13 ], nonstatis- tical procedure also has been established to evaluate the results. Using this procedure, the following criteria are used to interpret results. Positive: A compound is considered a mutagen if it produces a reproducible, dose-related increase in the number of revertant colonies in one or more strains. Note : It is considered a weak mutagen if it produces a reproducible, dose-related increase in the number of revertant colonies in one or more strains, but the number of revertants does not double the background number of colonies. Negative: A compound is considered a non-mutagen if no dose-related increase in the number of revertant colonies is observed in at least two independent experiments. Inconclusive: If a compound cannot be identiﬁ ed clearly as a mutagen or a non-mutagen, the results are classiﬁ ed as inconclusive.

4 Support Protocols

4.1 Preparation The strains used in the bacterial reverse mutation test are not of Metabolic capable of performing the cytochrome-based metabolic oxidations Activation that occur in mammalian systems. For this reason, before perform- System (S-9) ing an experiment, the fraction of a rodent liver homogenate (referred to as S-9 fraction) is added to the enzymic cofactors, NADP and NADPH, to make up the S-9 mix. Both should be stored frozen. To increase the enzyme activity, the animals used to obtain the S-9 fraction are pretreated with Aroclor or a combined injection of phenobarbital and B-naphthofl avone (5,6-benzo- fl avone) [27 – 29 ]. The metabolism of different chemicals to their mutagenic forms requires different optimum concentrations of S-9. It is recommended to use between 4 % and 30 % (v/v in the cofactor mix) of the S-9 fraction [ 1 , 2 ]. Some laboratories use a low S-9 concentration (e.g., 5 % or 10 %) in the fi rst mutagenicity assay. If negative results are obtained, a subsequent test with a high S-9 level (30 %) is performed. Other laboratories use the higher level of S-9 fi rst and use the low level in the second experiment if negative results were obtained in the initial experiment. The procedure described in this protocol can also be followed when animals other than rats, e.g., hamsters, are used for preparing the metabolic activation system from livers. The S-9 can be prepared by the testing laboratory and aliquots frozen at −80 ºC. When more than one animal is used, the tissues should be pooled in order to minimize animal-to-animal variation among batches. All steps of the procedure should be carried out at 0–4 ºC. 20 Araceli Pillco and Eduardo de la Peña

For inducing liver enzymes: 1. Dilute Aroclor 1254 in corn oil to a concentration of 200 mg/ mL. Using a 27-G needle, give each animal a single intraperitoneal injection of 500 mg/kg (0.5 mL of this mixture), 5 days before kill them [ 6 ]. 2. Give the rats unrestricted access to drinking water and food until 12 h prior to kill them. At that point, remove the food. 3. Kill the animal by cervical dislocation and place it on its back on an autopsy board. Note : All subsequent steps should be performed at 0 –4 º C, using cold, sterile solutions and glassware. For preparing the S-9 fraction: 4. Remove livers and place in preweighed beaker containing 1 mL of 0.15-M KCl/g wet liver. 5. Wash livers several times in fresh 0.15-M KCl. 6. Transfer livers to sterile beaker containing 3 mL of 0.15-M KCl/g wet liver. 7. Mince the livers with sterile scissors and homogenize using a Potter-Elvehjem homogenizer with a loose Teﬂ on pestle. 8. Centrifuge homogenate 10 min at 9,000 × g , at 0 –4 ºC. 9. Decant and save the supernatant (containing the S-9 fraction) into a sterile beaker that is kept on ice. 10. Dispense 1-mL aliquots of the S-9 fraction into sterile 1- or 5-mL cryogenic vials. 11. Store the S-9 homogenate in a −80 ºC freezer or in liquid nitrogen in appropriately labeled boxes. 12. Before being used for routine screening, test each batch of S-9 for sterility, as well as for effectiveness against the laboratory’s standard positive control mutagens, or against the chemical class of interest. Note: A number of commercial suppliers are now providing S-9 preparations. This has the advantage that information is provided about the enzyme activities and the effectiveness against standard mutagens. It may also be more cost-effective than purchasing and maintaining animals for this purpose.

4.2 Long-Term For long-term preservation, the Salmonella tester strains should be Storage of the Tester kept frozen at −80 ºC or liquid nitrogen. Healthy-looking single Strains colonies should be chosen to prepare the frozen stock cultures. Dimethylsulfoxide (DMSO) or glycerol is suggested as cryoprotective agent. The ﬁ nal concentration of the cryoprotective should be at least 10 % (v/v) [6 ]. Ames Test Protocol 21

Acknowledgments

The authors express their thanks to Dr. Oscar Herrero, Dr. Eduardo de la Peña Jr., Ms. Antonia Martinez, and the Institute of Agricultural Sciences (CSIC).

References

1. Ames BN, McCann J, Yamasaki E (1975) incubation assay: increased sensitivity for Methods for detecting carcinogens and muta- detecting mutagens in human urine. Mutat Res gens with the Salmonella /mammalian-micro- 112:25–32 some mutagenicity test. Mutat Res 31:347–364 12. Hughes TJ, Simmons DM, Monteith LG et al 2. Maron D, Ames BN (1983) Revised methods (1987) Vaporization technique to measure for the Salmonella mutagenicity test. Mutat mutagenic activity of volatile organic chemicals Res 113:173–215 in the Ames/Salmonella assay. Environ 3. Ames BN, Lee FD, Durston WE (1973) An Mutagen 9:421–441 improved bacterial test system for the detection 13. Zeiger E, Anderson B, Haworth S et al (1992) and classifi cation of mutagens and carcinogens. Salmonella mutagenicity tests: V. Results from Proc Natl Acad Sci U S A 70(3):782–786 the testing of 311 chemicals. Environ Mol 4. Ames BN, Durston WE, Yamasaki E et al Mutagen 19(21):1–141 (1973) Carcinogens are mutagens: a simple 14. Araki A, Noguchi T, Kato F et al (1994) test system combining liver homogenates for Improved method for mutagenicity testing of activation and bacteria for detection. Proc Natl gaseous compounds by using a gas sampling Acad Sci U S A 70(8):2281–2285 bag. Mutat Res 307:335–344 5. Mortelmans K, Riccio ES (2000) The bacterial 15. Tejs S (2008) The Ames test: a methodological tryptophan reverse mutation assay with short review. Environ Biotechnol 4(1):7–14 Escherichia coli WP2. Mutat Res 455(1–2): 16. Dearfi eld KL, Auletta AE, Cimino MC et al 61–69 (1991) Considerations in the U.S. Environ- 6. Mortelmans K, Zeiger E (2000) The Ames mental Protection Agency’s testing approach Salmonella /microsome mutagenicity assay. for mutagenicity. Mutat Res 258:259–283 Mutat Res 455(1–2):29–60 17. Kramers PGN, Knaap AGAC, van der Heijden 7. Claxton LD, Umbuzeiro G, DeMarini DM CA et al (1991) Role of genotoxicity assays in (2010) The Salmonella mutagenicity assay: the the regulation of chemicals in The Netherlands: stethoscope of genetic toxicology for the 21st considerations and experiences. Mutagenesis 6: century. Environ Health Perspect 118(11): 487–493 1515–1522 18. FDA (US Food and DruGg Administration) 8. Eaton DL, Gilbert SG (2008) Principles of (1993) Toxicological principles for the safety toxicology. In: Klaassen CD (ed) Casarett and assessment of direct food additives and color Doull’s toxicology, the basic science of poisons, additives used in food “Redbook II” [Draft]. 7th edn. McGraw-Hill, New York, pp 11–45 FDA, Center for Food Safety and Applied 9. McCann J, Choi E, Yamasaki E et al (1975) Nutrition, Washington, DC Detection of carcinogens in the Salmonella / 19. HPBGC (Health Protection Branch Geno- microsome test: assay of 300 chemicals. Proc toxicity Committee Guidelines; Canada) Natl Acad Sci U S A 72:5135–5139 (1993) The assessment of mutagenicity. Health 10. Mehmet K, Özlem B, Medine G (2012) Protection Branch mutagenicity guidelines. Salmonella as a unique tool for genetic toxicol- Environ Mol Mutagen 21:15–37 ogy, Salmonella—a diversifi ed superbug. In: 20. Kirkland DJ (1993) Genetic toxicology testing Kumar Y (ed), ISBN: 978-953-307-781-9. requirements: offi cial and unoffi cial views from InTech. http://www.intechopen.com/books/ Europe. Environ Mol Mutagen 21: 8–14 salmonella-a-diversifi ed-superbug/salmonella- 21. OECD 471 (1997) Guideline for testing of as-aunique-tool-for-genetic-toxicology . chemicals. Bacterial reverse mutation test. Accessed 15 Nov 2013 Organization for Economic Cooperation and 11. Kado NY, Langley D, Eisenstadt E (1983) Development, Paris, Adopted 26 May 1983, A simple modifi cation of the Salmonella liquid last updated 21st July 22 Araceli Pillco and Eduardo de la Peña

22. Regulation (EC) No 1907/2006 of the 30. Zeiger E, Mortelmans K (1999) The Salmonella European Parliament and of the Council of 18 (Ames) test for mutagenicity. Curr Protoc December 2006 concerning the Registration, Toxicol Unit 3.1:3.1.1–3.1.29. Doi: 10.1002/ Evaluation, Authorization and Restriction of 0471140856.tx0301s00 Chemicals (REACH) (2006) 31. Ansari MI, Malik A (2009) Genotoxicity of 23. Houk VS (1992) The genotoxicity of industrial agricultural soils in the vicinity of industrial wastes and effl uents. Mutat Res 277:91–138 area. Mutat Res 673:124–132 24. Wilkinson RG, Gemski P Jr, Stocker BAD 32. Líman R, Eren Y, Akyil D, Konuk M (2012) (1972) Non-smooth mutants of Salmonella Determination of mutagenic potencies of typhimurium : differentiation by phage sensitiv- aqueous extracts of Thermopsis turcica by Ames ity and genetic mapping. J Gen Microbiol test. Turk J Biol 36:85–92 70:527–554 33. Pillco A (2011) Evaluación toxicológica de 25. Walker GC, Dobson PP (1979) Mutagenesis mezclas complejas mediante ensayos alternativos and repair defi ciencies of Escherichia coli umuC a la experimentación animal de genotoxicidad, mutants are suppressed by the plasmid mutagenicidad y ecotoxicidad (Toxicological pKM101. Mol Gen Genet 172:17–24 evaluation of complex mixtures employing 26. Mahon G, Green M, Middleton B et al (1989) genotoxicity, mutagenicity and ecotoxicity Analysis of data from microbial colony assays. assays as an alternative to animal testing) In: Kirkland DJ (ed) UKEMS Sub-committee Doctoral thesis, Autonomous University of on Guidelines for Mutagenicity Testing. Madrid, p 173 Report: part III. Statistical evaluation of muta- 34. Rank J, Nielsen MH (1998) Genotoxicity test- genicity test data. Cambridge University Press, ing of wastewater sludge using the Allium cepa Cambridge, pp 28–65 anaphase-telophase chromosome aberration 27. Ong TM, Mukhtar CR, Wolf CR et al (1980) assay. Mutat Res 418:113–119 Differential effects of cytochrome P450- 35. Monarca S, Feretti D, Zerbini I et al (2002) inducers on promutagen activation capabilities Soil contamination detected using bacterial and enzymatic activities of S9 from rat liver. and plant mutagenicity test and chemical analy- J Environ Pathol Toxicol 4:55–65 ses. Environ Res Sect A 88:64–69 28. Barrueco C, de la Peña E (1988) Mutagenic 36. Sandoval AM (2008) Ensayo de mutagenic- evaluation of pesticides captan, folpet, captafol, idad con la bacteria Salmonella typhimurium . dichlofl uanid and related compounds with the Prueba de Ames (Mutagenicity assay with mutants TA102 and TA104 of Salmonella Salmonella typhimurium. Ames test). In: typhimurium . Mutagenesis 3(6):467–480 Ramirez P, Mendoza A (eds) Ensayos toxi- 29. Barrueco C, Herrera A, de la Peña E (1991) cológicos para la evaluación de sustancias Mutagenic evaluation of trichlorfon using químicas en agua y suelo, 1st edn. La experien- different assay methods with Salmonella cia en México, México typhimurium . Mutagenesis 6(1):71–76 Chapter 2

The Ames II and Ames MPF Penta I Assay: A Liquid Microplate Format Modiﬁ cation of the Classic Ames Test

Sini Flückiger-Isler and Markus Kamber

Abstract

This chapter describes a protocol modiﬁ cation of the bacterial reverse mutation test (Ames test). It is based on the same principle as the Ames test but uses a liquid, low-volume microplate version of the ﬂ uctuation method. Clear strengths are the low compound requirement and the increased throughput as compared to the standard format. The liquid system allows for processing several replicates at once with the possibility of using pipetting robots and has an easy colorimetric readout. The Ames II/Ames MPF also uses less S9 and produces less hazardous waste due to the low-volume multiwell format.

Key words Ames II , Ames MPF , Liquid microﬂ uctuation method , Test kits , High throughput

1 Introduction

The bacterial reverse mutation test using Salmonella typhimurium [1 , 2 ] or Escherichia coli [ 3 ] is the most widely used in vitro test to evaluate mutagenicity of chemical substances and environmental samples. The principle of this test is described in detail in Chap. 1 of this book. With at least two strains detecting base-pair and frameshift mutations, the bacterial reverse mutation test is generally used for genotoxicity screening of drugs, chemicals, environmental samples, and food additives. The bacterial reverse mutation test is also required for regulatory genotoxicity testing and performed according to the guidelines of various regulatory agencies such as OECD (Organisation for Economic Co-operation and Development) and ICH (International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use). The traditional format of a bacterial reverse mutation test described in OECD guideline 471 [ 4 ] relies on scoring bacterial growth (colonies) on selective agar plates after exposure of the bacterial cells to a test chemical, either by incorporation of the test chemical into the agar plates or by preincubation prior to plating.

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_2, © Springer Science+Business Media New York 2014 23 24 Sini Flückiger-Isler and Markus Kamber

This format uses a lot of agar plates and consumes large amounts of test sample and S9. The need for relatively large amounts of test chemicals and the low-throughput format limit the use of the standard Ames plate incorporation or preincubation assay during early phase of drug development. Similar constraints apply for the testing of surface waters or drinking water which requires large amounts of raw material for extraction and concentration steps. Scientists’ efforts have therefore led to the development of scaled- down and high-throughput genotoxicity assays which allow screening of a greater number of samples. One of such tests is the Ames II/Ames MPF assay, a liquid microplate version of the bacterial reverse mutation test which has a very good conformity to the standard testing procedure [5 ], decreases the amount of test sample required for a study and allows for automation.

2 Ames II Assay

2.1 Principle The Ames II assay is a 384-well microplate format modiﬁ cation of the classic bacterial reverse mutation test. It uses a small volume, liquid format for exposure of the bacterial tester strains to test samples in the absence and presence of S9, and a colorimetric readout for ease of scoring the revertant bacteria. In this assay, the frameshift mutations are detected by the traditional strain TA98, and base-pair substitutions by a mixture of six strains named TAMix (Table 1 ). The TAMix strains and the microﬂ uctuation test procedure that is used in the Ames II assay were developed by Gee et al. [5 , 6 ]

Table 1 Genotypes of TA98 and TAMix Salmonella typhimurium strains

Strain Mutation Type Target Cell walla Repairb pKM101 c

TA98 hisD3052 Frameshift GC rfa uvrB Yes TAMix contains TA7001 hisG1 775 b.p. subst. T:A>C:G rfa uvrB Yes TA7002 hisC9138 b.p. subst. T:A > A:T rfa uvrB Yes TA7003 hisG9074 b.p. subst. T:A > G:C rfa uvrB Yes TA7004 hisG9133 b.p. subst. C:G>T:A rfa uvrB Yes TA7005 hisG9130 b.p. subst. C:G > A:T rfa uvrB Yes TA7006 hisC9070 b.p. subst. C:G > G:C rfa uvrB Yes a (rfa ): This mutation leads to a defective lipopolysaccharide (LPS) layer that coats the cell surface, making the bacteria more permeable to bulky chemicals b (uvrB ): The uvrB deletion mutation eliminates the accurate excision repair mechanism, thereby allowing more DNA lesions to be repaired by error-prone DNA repair mechanisms. The deletion through the biotin gene makes the bacteria biotin dependent c (pKM101): This R factor plasmid enhances chemical and UV-induced mutagenesis via an error-prone recombinational DNA repair pathway. The plasmid also confers ampicillin resistance Ames II and Ames MPF Penta I Assay 25

and originally designed to serve both as a screen for mutagenic substances and, by using the TAMix strains individually, to allow the identifi cation of the specifi c base-pair substitution mutations produced. The TAMix comprises six his − Salmonella tester strains, TA7001–TA7006, each with a different base-pair substitution. Each of these mutants can be reverted only by one specifi c transition or transversion such that all possible base-pair changes can be detected and identifi ed. Like the traditional strains, the genetic background of the TA700x series of strains has been modifi ed to improve the sensitivity of their reversion ( uvrB , rfa , plasmid pKM101).

2.2 General Freshly prepared overnight cultures of TA98 and TAMix are Assay Description exposed to six concentrations of a test agent, as well as to a positive and negative control for 90 min in medium containing sufﬁ cient histidine to support a few cell divisions. After 90 min the exposure cultures are diluted in pH indicator medium lacking histidine and aliquoted into 48 wells of a 384-well plate. Within 2 days, cells that have undergone the reversion to histidine prototrophy—either spontaneously, or as a result of the exposure to a mutagen—will grow into colonies. Bacterial metabolism reduces the pH of the medium, changing the color of that well from purple to yellow. The number of wells containing revertant colonies is counted for each dose and compared to a solvent (negative) control. Each dose is tested in triplicate to allow for statistical analysis of the data. An increase in the number of revertant wells relative to the solvent controls indicates that the chemical is mutagenic in the Ames II assay. The mutagenic potential of substances is assessed directly and in the presence of liver S9 (Fig. 1 ).

2.3 Advantages Important advantages of this test system are that it consumes much less test sample than the standard plate or preincubation tests. It requires less hands-on time since the low-volume liquid format allows working with multichannel pipettes and processing several replicates at once. It needs considerably less S9 and plasticware, and it can be partly processed by pipetting stations. The assay has an easy colorimetric readout in 48-well sections of a 384-well plate. Typically, triplicate plates with six sample dilutions, negative and positive controls are scored by eye in 1–2 min. The Ames II assay is available as a kit including all necessary ingredients ready to use and with detailed instruction for use (see Sect. 2.5.1 ). The Ames II procedure has found its use as an early mutagenicity screening procedure for pharmaceutical and chemical companies, as well as for environmental samples.

2.4 Validation There are several validation studies comparing the Ames II with Studies the traditional agar plate assay using several strains. In all these studies, many chemical classes were evaluated. The ﬁ rst Ames II 26 Sini Flückiger-Isler and Markus Kamber

Fig. 1 Ames II/Ames MPF test procedure

validation study with 25 compounds was published in 1998 and gave a concordance of 79 % for TA98 and TAMix with traditional data obtained with up to fi ve strains [ 5 ]. The authors concluded that the high concordance with the traditional Salmonella test and the reproducibility among cultures and replicates demonstrate that the Ames II test procedure is an effective screen for identifying Salmonella mutagens. An international Ames II round robin study performed with 19 coded compounds [ 7 ] gave an 84 % concordance with published traditional results and an almost 90 % interlaboratory consistency. Another validation study was performed by Gervais et al. [ 8 ]. They tested 42 proprietary compounds and obtained a concordance of 83 %. The disagreement in the test results was obtained mostly with compounds that specifi cally revert E. coli or TA1535. In a study released in 2009 [9 ], 71 compounds tested with the Ames II assay were compared with published data for the traditional test, resulting in an agreement of 84 %, which is comparable to the intra- and interlaboratory reproducibility of 87 % for the traditional test itself. The Ames II assay is routinely used for the screening of early drug candidates by several pharmaceutical companies such as BASF, Janssen Pharmaceutica NV, Sanofi -Aventis GmbH, and Servier, or for the investigation of genotoxic impurities. Sanofi - Aventis reports a predictability of ~92 % when comparing the results of the Ames II with the full-scale traditional test, and a throughput of 40 compounds weekly [ 10 ]. The Servier Group uses the Ames II test for early evaluation of in-house compounds [11 ]. Ames II and Ames MPF Penta I Assay 27

All compounds that resulted positive in the Servier Ames II screening assay were conﬁ rmed positive in the regulatory bacterial reverse mutation test. The Ames II test is also regularly used for investigating drinking water quality [12 ].

2.5 Detailed Assay The TA98 strain which is also used in the traditional bacterial Description reverse mutation assay can be re-isolated and stored as described earlier [1 , 2 ]. The individual TA700x strains of the TAMix are pat- 2.5.1 Materials ent protected. The TAMix is an equimolar mixture of the six individually grown strains. Strains have to be stored at ≤−70 °C until use. TAMix, as well as growth, exposure, and indicator media, which are stored at room temperature, can be purchased ready to use [ 13 , 14 ]. S9 can be purchased from different vendors, either in liquid (storage −80 °C) or lyophilized form (storage −20 °C), and has a shelf life of ≥2 years when stored correctly. Positive control chemicals and ampicillin are available from different suppliers. Reagents for preparing the S9 mix (buffer–salts, G-6-P, and NADP) can be prepared individually or are available as ready-to-use solutions [13 ]. The necessary plasticware for performing the assay (culture tubes, multiwell plates, reagent reservoirs) can be bought individually. Complete test kits for performing the Ames II assay are also commercially available [13 , 14 ], and the procedure is described in the Ames II Instructions for Use [ 15 ].

2.5.2 Equipment Apart from the abovementioned items, a laboratory should feature the following equipment to perform the Ames II assay: an environmental shaker with approximately 2.5–3 cm amplitude capable of 37 °C, 250 rpm incubations, a 37 °C dry incubator, an autoclave, a spectrophotometer for measuring optical density at 600 nm, 1.5 mL spectrophotometer cuvettes, single and multichannel adjustable pipettes and sterile tips, a programmable stepper for dispensing 2.6 mL (recommended), an 8-channel repeating pipette for the transfer of the exposed cultures to the 384-well plates (highly recommended), 5 and 10 mL serological pipettes and solvents for sample dilution and solvent control. A light box facilitates the scoring of the 384-well plates. Note: Shakers with different characteristics may be used but might require different incubation ﬂ asks for optimal bacterial growth .

2.5.3 Overnight Culture Using sterile technique, overnight cultures of TA98 and TAMix Preparation are prepared by performing the following steps: 10 μL of freshly thawed and carefully mixed bacterial stocks are added to 50 mL culture tubes containing 10 mL growth medium and 10 μL of 50 mg/mL ampicillin. The culture tubes are capped loosely (or ﬁ lter caps are used) to allow for sufﬁ cient aeration. The cultures are grown for 12–15 h at 37 °C, 250 rpm in an environmental shaker to the late exponential phase. Incubation time varies depending on the shaker type and amplitude. 28 Sini Flückiger-Isler and Markus Kamber

2.5.4 Determination One hundred microliter aliquots of the overnight cultures are of Optical Density (OD600 ) diluted 1:10 with growth medium and the OD 600 is measured in semi-micro-cuvettes (light path: 1 cm). The OD600 reading is multiplied by 10 to obtain the actual optical density and should be at least 2.0. Note: The OD measurement of bacterial cultures is actually based on the amount of light scattered by the culture rather than the amount of light absorbed. In their standard conﬁ guration, spectrophotometers are not optimized for light scattering measurements, thus resulting in differences in measured absorbance between instruments.

2.5.5 Preparation of Test Test samples and positive controls are prepared as 25-fold concen- Sample Dilutions, Positive trated stock solutions because they will be diluted 1:25 in the assay Controls (10 μL in a fi nal volume of 250 μL). Sample dilutions are performed in a 96-well plate. The dilution scheme is confi gured for six concentrations, plus a negative (solvent) and positive control; every other well is used to allow for the transfer to the 24-well exposure plate with a multichannel pipette (Fig. 2 ). The dilution scheme is performed by fi rst adding the appropriate amount of solvent to wells #0–5 of the plate. The 25X stock is transferred to well #6 of the dilution plate. The serial dilutions are performed by transferring a defi ned volume of the test sample from well #6 to #1. The “0” well is the solvent control and the “+” well is the positive control for the assay. For the testing of solid chemicals, a top dose of 5 mg/mL (corresponding to a 25-fold stock solution of 125 mg/mL) is recommended. If there are solubility limitations, the lowest workable suspension may be used as the highest concentration. For liquid test samples, a top dose of 5 μL/mL is recommended. Environmental samples (e.g., surface waters, wastewater, sediment) can be tested in the Ames II assay after appropriate extraction/concentration.

Fig. 2 Preparation of the 96-well dilution plate Ames II and Ames MPF Penta I Assay 29

Table 2 Preparation of 30 % S9 mix using self-made reagents (one test compound in TA98 and TAMix)

Stock (M) Reagent Volumes (mL)

1.00 KCl 0.076

0.25 MgCl 2 × 6H2 O 0.074 0.20 Glucose-6-phosphate 0.058 0.04 NADP 0.230

0.20 NaH 2 PO4 buffer, pH 7.4 1.173 S9 fraction 0.690 Final volume 2.301

2.5.6 Amount of Test The required volumes are dependent on the number of replicates Sample Needed (usually 3) and the dilution factor used. 60 μL of each stock in the Ames II Assay concentration will be required for dosing each strain when testing triplicates with and without S9 (6× 10 μL). Therefore, enough of the 25X stock should be added to well #6 such that there will be sufﬁ cient volume for the serial dilutions. It is recommended to calculate also a dead volume (pipetting reserve) of approximately 20 μL. The required test sample amount and volume can be automatically calculated by using the “Ames MPF dilutions calculator” which is included in the “Ames calculation sheet” and can be downloaded from the Internet [13 ]. For example, an Ames II assay (two strains) performed in triplicates in the presence and absence of S9, and using a 5 mg/mL top dose and ½ log dilution steps (10 0.5 = 3.16-fold dilutions), requires 205 μL of a 125 mg/mL stock solution, resulting in a total test sample amount of 25.6 mg. The dilution scheme for the abovementioned example corresponds to the following volumes for Fig. 2 : 205 μL of a 125 mg/mL stock is added to well #6 and 140 μL solvent is added to wells #0–5. 65 μL is transferred from well #6 to #5 and mixed. The serial dilutions are completed from well #5 to #4, #4 to #3, #3 to #2, and #2 to #1; the 0 well is the solvent control for the assay.

2.5.7 Preparation The Ames II assay is performed in the absence and presence of of 30 % S9-Mix exogenous metabolic activation. Aroclor-1254- or phenobarbital/ ß- naphthoﬂ avone-induced rat liver S9 is usually used. The instructions for preparing the S9 mix are given in Table 2 . Immediately before use, a 30 % S9 mix is prepared by combining the volumes of reagents listed in Table 2 in a sterile tube. All (thawed) reagents are kept on ice.

2.5.8 Preparation Note: This dosing protocol applies to one 24 -well plate corresponding of the Exposure Cultures to an assay with one test sample at six concentrations in triplicates in one strain without or with S9 . 30 Sini Flückiger-Isler and Markus Kamber

1 2345 1 234 5 6 A 0 A 0A 4A 0B 4B0C 4C B 4

C 1 B 1A 5A 1B 5B 1C 5C D 5

E 2 C 2A 6A 2B 6B 2C 6C F 6 3 G D 3A +A 3B +B 3C +C H +

Fig. 3 Transfer of test sample from 96-well dilution plate to 24-well exposure plate

Using an 8-channel pipette with four evenly spaced tips (every other channel), 10 μL from the ﬁ rst column of the dilution plate (wells #0–3) is transferred to columns 1, 3, and 5 of the 24-well plate. Then, 10 μL from the second column of the dilution plate (wells #4–6 and +) is transferred to columns 2, 4, and 6 of the 24-well plate (Fig. 3 ). The exposure culture is prepared by adding 6.3 mL exposure medium and 0.7 mL bacterial culture to a sterile reagent reservoir (assay without S9), or by adding 5.25 mL exposure medium, 0.7 mL bacterial culture, and 1.05 mL of freshly prepared 30 % S9 mix to a sterile reagent reservoir (assay with S9). Using an 8- channel pipette, 240 μL is transferred to all wells of the 24-well exposure plate. (The pipette is set to 120 μL since two tips dispense into each well.) Each well of a 24-well plate now contains 10 μL of test sample diluted to the appropriate concentration, 25 μL of bacterial culture, and 215 μL of exposure medium (or 177.5 μL of exposure medium and 37.5 μL of 30 % S9 mix), giving a total volume of 250 μL. This mixture is incubated for 90 min at 37 °C with agitation at 250 rpm. Precipitation of the sample in the exposure mixture should be recorded.

2.5.9 Transfer After the 90 min incubation, 2.6 mL indicator medium is added to of the Exposed Cultures each well of the exposure plates. Each replicate (8 wells) is trans- to 384-Well Plates ferred to a separate 384-well plate. Using an 8-channel repeating pipette, 50 μL aliquots of column 1 are dispensed into columns 1–12 of a ﬁ rst 384-well plate. Column 2 of the 24-well exposure plate is dispensed accordingly into columns 13–24 of the ﬁ rst 384- well plate. This procedure is repeated for the remaining columns of the Ames II and Ames MPF Penta I Assay 31

24-well plates. Columns 3 and 4 of the 24-well plate are dispensed into a second 384-well plate, and columns 5 and 6 are dispensed into a third 384-well plate. This procedure is repeated for each 24-well plate. The transfer can also be processed by a pipetting workstation. After completion of the transfer, the 384-well plates are placed into a sealable plastic bag to prevent evaporation during incubation. The plastic bag is placed into a 37 °C dry incubator for 2 days.

2.5.10 Scoring, After 2 days, the 384-well plates are scored ideally by placing them Calculation, and Evaluation on top of a light box. The use of a transparent template that divides Criteria the 384-well plate in eight 48-well sections can be helpful. The number of positive wells in each 48-well section is scored. Positive wells are those that have turned yellow or have a bacterial colony visible on the bottom of the well (Fig. 4 ). Any indication of a color change from purple to yellow should be included in the positive count. Note: The plates are easily scored by eye, but they can also be measured with a microplate reader. An Ames II assay is considered valid if the solvent control and the positive control values (mean of positive wells of all replicates) are within the acceptable range (Table 3 ). Toxicity can be assessed

SC C4

C1 C5

C2 C6

C3 PC

Fig. 4 Plate scoring. SC solvent control, C1 –C6 sample concentrations, PC positive control

Table 3 Positive controls and acceptable range for solvent and positive controls in the Ames II assay

Acceptable range Strain S9 Chemical (positive wells/48)

TA98 and − None ≤8 TAMix − 2-NF/4-NQO >25 (2 μg/mL/0.5 μg/mL) + None ≤8 + 2-AA (5 μg/mL) >25 2 - NF 2-nitroﬂ uorene, 4 - NQO 4-nitroquinoline-N -oxide, 2 - AA 2-aminoanthracene 32 Sini Flückiger-Isler and Markus Kamber

by a signifi cant decrease of positive wells relative to the solvent control (generally at high test sample concentrations) and by an increase of the brilliance of the purple medium as compared to the solvent control (cell lysis, absence of bacterial cells). The fold increase of revertants relative to the solvent control is determined by dividing the mean number of positive wells at each dose by that of the solvent control baseline. The solvent control baseline is derived from the mean number of positive wells in the solvent control plus 1 standard deviation. If the baseline is less than 1, the value is set to 1 for calculation. A fold increase greater than two times the baseline level is generally considered as an alert. Multiple alerts with a dose–response will lead to the test sample being classifi ed as a clear positive. A test sample is classifi ed negative when no response greater than two times the baseline is recorded. Fold inductions are calculated from the baseline of the actual overnight culture and not from historical data. Although the acceptable range for most cultures is between 0 and 8, the overnight cultures of different experiments have a distinct narrower range (e.g., 2-2-1 or 4-5-3). The use of the baseline will accom- modate for larger standard deviations. If more than one compound is tested with the same overnight culture, the negative (solvent) control wells can be pooled. For example, when three samples were tested with the same culture on the same day (e.g., TA98-S9), the three corresponding triplicate negative control scores are pooled to a mean of nine replicates. A large number of negative control counts give a more reliable range of the spontaneous revertants of a given culture.

3 The Ames MPF Penta I Assay

The Ames II test has found its use as an early mutagenicity screening procedure with pharmaceutical and chemical companies [8 , 10 ], as well as in the fi eld of water testing [12 ]. Recently, the liquid microplate procedure used in the Ames II screen has been adapted to the strains listed in the OECD guideline 471. The regulatory full- scale bacterial reverse mutation assay uses several strains of S. typhimurium and E. coli , and has been used extensively in genetic toxicology testing. The OECD guideline 471 recommends using at least fi ve tester strains that carry different target sites in order to be sensitive to a broad range of chemicals. They include four strains of S. typhimurium , TA98, TA100, TA1535, and TA1537 or TA97a or TA97. These strains all have GC base pairs at the primary reversion site and may therefore not detect certain classes of chemicals. Such chemicals can be detected by either E. coli WP2 strains or S. typhimurium TA102 which have an AT base pair at the primary reversion site. The Ames microplate fl uctuation format (Ames MPF) assay can be performed with S. typhimurium TA98, TA100, TA1535, Ames II and Ames MPF Penta I Assay 33

Table 4 Genotypes of S. typhimurium and E. coli strains

Strain Mutation Type Cell wall Repair pKM101

S. typhimurium (histidine auxotrophic) TA98 hisD3052 Frameshift rfa uvrB Yes TA100 hisG46 Base-pair subst. rfa uvrB Yes TA1535 hisG46 Base-pair subst. rfa uvrB No TA1537 hisC3076 Frameshift rfa uvrB No E. coli wp2 (tryptophan auxotrophic) uvrA trpE65 Base-pair subst. – uvrA No [pKM101] trpE65 Base-pair subst. – – Yes rfa : This mutation leads to a defective lipopolysaccharide (LPS) layer that coats the cell surface, making the bacteria more permeable to bulky chemicals uvrB / uvrA : The uvrB / uvrA deletion mutations eliminate the accurate excision repair mechanism, thereby allowing more DNA lesions to be repaired by error-prone DNA repair mechanisms pKM101: This R factor plasmid enhances chemical and UV-induced mutagenesis via an error-prone recombinational DNA repair pathway. The plasmid also confers ampicillin resistance

TA1537, and E. coli WP2 uvrA and E. coli WP2 [pKM101] (Table 5 ). All reagents and the strains can be purchased separately or as a kit (Ames MPF Penta I, [ 13 ]). The Salmonella strains are histidine auxotrophs and are therefore tested in media with limiting histidine (and biotin) supplementation (exposure medium) or lacking histidine (indicator medium). The E. coli WP2 strains are tested in the same way as the Salmonella strains with the exception that the exposure medium is supplemented with limiting tryptophan and the indicator medium lacks tryptophan. The genotypes of the strains used in the Ames MPF Penta I assay, including additional mutations which enhance sensitivity to certain mutagens, are shown in Table 4 .

3.1 Assay The procedure for the Ames MPF Penta I assay is basically the same Description as described for the Ames II assay (Fig. 1 , paragraphs 2.5.3 – 2.5.10 ), except that different strains are used (some of them being differently diluted in the exposure medium), and that the E. coli strains with their tryptophan requirement need a different exposure and indicator medium than the histidine-requiring Salmonella strains. The Ames MPF Penta I assay, in brief, is as follows: 1. Growth of tester strains overnight. 2. Exposure to test sample and S9 if employed. 3. Distribution and plating of cells in a medium which selects for revertants. This medium contains a pH indicator dye that turns from purple to yellow upon bacterial growth. 34 Sini Flückiger-Isler and Markus Kamber

4. Incubation in 384-well plates for 48 h to allow growth of revertant bacteria. 5. Scoring of plates for positive (yellow) wells, data entry, and evaluation of mutational potential.

3.2 Advantages A main advantage of the Ames MPF over the Ames II assay is that strains recommended in the OECD guideline 471 [ 4 ] are used, thus making the microﬂ uctuation procedure a valid alternative to the regulatory plate incorporation/preincubation assay. The technical advantages of the Ames MPF test system are the same as those for the Ames II test: – Almost four times less test chemical is needed than in the standard plate or preincubation test. – It requires less hands-on time since the low-volume liquid format allows working with multichannel pipettes and robotic systems. – It needs less plasticware and 10 times less S9 mix. – The assay has an easy colorimetric readout in 48-well sections of a 384-well plate. – The Ames MPF assay is available as a kit including all necessary ingredients ready to use and constantly quality controlled, and with detailed instruction for use. – Easy and quick colorimetric readout.

3.3 Validation In the fi rst validation study of 1998 with base-specifi c tester strains, Studies Gee et al. [ 5 ] included also the traditional frameshift strains TA98 and TA1537 for comparing the liquid microplate format with published data of the preincubation method. There was an overall agreement of 84 % (21/25) and of 94 % (18/20) in the TA98 and TA1537 results, respectively. The liquid format seemed to be more sensitive in this comparison. Several posters [ 16 – 18 ] presented at different congresses showed the excellent concordance of the Ames MPF test results obtained with a combination of the standard set of tester strains with those of published data for the traditional test. The Ames MPF assay using the four Salmonella strains TA98, TA100, TA1535, and TA1537 showed the best correlation with a miniaturized plate incorporation assay (“Mini Ames”) when compared to the Vitotox and the Ames II at UCB Pharma [ 19 ]. In 2010, Umbuzeiro et al. [ 20 ] compared the Ames MPF method and the microsuspension assay developed by Kado et al. [ 21 ] by concurrently testing environmental samples (air, surface water, effl uent water) with strain TA98. The results of both assays correlated very well, and the authors considered the Ames MPF method a valid alternative to the microsuspension assay due to its easy handling. In their 2012 publication [ 22 ], Flückiger-Isler and Kamber tested 15 equivocal to weakly positive chemicals selected Ames II and Ames MPF Penta I Assay 35

2 Plate (FI over NC) MPF (FI over BL) 1 Fold induction over NC or BL

0 3.125 6.25 12.5 25 50 Glutaraldehyde (µg per ml or plate) Fig. 5 Dose–response of glutaraldehyde in TA100 without S9. NC negative control, BL baseline (mean negative control + 1 SD), FI fold induction, Plate preincubation method, MPF microplate ﬂ uctuation method

from the National Toxicology Program (NTP) database concurrently in the Ames MPF and the standard Ames preincubation method. Thirteen of the 15 chemicals showed concordant results in both tests despite the challenging choice of chemicals. The Ames MPF method appeared to be more sensitive, since in half of the positive results, the mutagenic effect or cytotoxicity, if present, was seen at lower doses in the Ames MPF than in the Ames plate incorporation method (Fig. 5 ).

3.4 Deviations The assay procedure is described in detail in the “Ames MPF Penta from the Ames II Test I Instructions for Use” [ 23 ]. Furthermore, a “visual guide for Protocol Ames MPF” illustrates all steps of the test procedure and the evaluation of the results [ 13 ].

3.4.1 Preparation The Salmonella strains were developed in the laboratory of Dr. of Strains and Media Bruce Ames, University of California, Berkeley, and have been established in many laboratories. The E. coli strains are available at the National Collections of Industrial and Marine Bacteria Limited, Aberdeen, Scotland, UK. Ready-to-use kits for performing the Ames microﬂ uctuation assay with the standard tester strains are commercially available [ 13 , 14 ]. The tester strains are cryopreserved and delivered in a semisolid form which allows for shipment at ambient temperatures for up to 10 days [13 ]. The kits also include ampicillin, positive controls, and S9. Reagents for preparing the S9 mix (buffer–salts, G-6-P, and NADP) can be prepared individually or are available as ready-to-use solutions [ 13 ]. The necessary plasticware (culture tubes, multiwall plates, reagent reservoirs) can be bought individually. 36 Sini Flückiger-Isler and Markus Kamber

3.4.2 Equipment see Ames II (Sect. 2.5.2 ).

3.4.3 Overnight Culture The strains of the Ames MPF Penta I assay [23 ] are cryopreserved Preparation in a different medium (semisolid) than those of the Ames II assay (liquid). This necessitates a small additional step in the preparation of the overnight cultures. After thawing and the addition of 200 μL growth medium, the semisolid pellet is disrupted mechanically by pipetting up and down until a uniform suspension is obtained which can be pipetted repeatedly without clogging the tip and which shows visually a homogeneous distribution of the dark fragments. 25 μL of the dispersion is added to 50 mL culture tubes containing 10 mL growth medium with ampicillin (strains TA98, TA100, E. coli [pKM101]) or without ampicillin (strains TA1535, TA1537, and E. coli WP2 uvrA ). The culture tubes are capped loosely (or ﬁ lter caps are used) in order to allow for sufﬁ cient aeration. The cultures are grown for 12–15 h at 37 °C, 250 rpm in an environmental shaker to the late exponential phase.

3.4.4 Determination see Ames II (Sect. 2.5.4 ). of Optical Density (OD600 )

3.4.5 Preparation of Test The procedure is the same as described for the Ames II assay Sample Dilutions, Positive (Sect. 2.5.5 ), except that the different positive controls should be Controls added directly to the corresponding wells of the 24-well exposure plates. The recommended sample top dose for the Ames MPF assay is the same as described for the Ames II assay.

3.4.6 Amount of Test As in the Ames II protocol, 60 μL of each stock concentration will Sample Needed be required for dosing each strain when testing in triplicates with in the Ames MPF Assay and without S9 (6× 10 μL), and there should be sufﬁ cient volume (See also Sect. 2.5.6 ) (including a dead volume) of the 25X stock added to well #6 to perform the serial dilutions. When the calculated volume of the sample in well #6 is >360 μL, it is recommended to use a 24-well plate as the dilution plate and to increase the dead volume to 40–50 μL. The required test sample amount and volume can be automatically calculated by using the “Ames MPF dilutions calculator” which is included in the Ames calculation sheet and can be downloaded [13 ]. For example, a full-scale Ames MPF Penta I assay with ﬁ ve tester cultures performed in the presence and absence of S9, and using triplicates, a 5 mg/mL top dose and ½ log dilution steps, requires 497 μL of a 125 mg/mL stock solution, resulting in a total test sample amount of 62.2 mg. The dilution scheme for the abovementioned example corresponds to the following volumes for Fig. 2 : 497 μL of a 125 mg/mL stock is added to well #6 and 340 μL solvent is added to wells #0–5. 157 μL is transferred from well #6 to #5 and mixed. The serial dilutions are completed from well #5 to #4, #4 to #3, #3 to #2, and #2 to #1; the 0 well is the solvent control for the assay. Ames II and Ames MPF Penta I Assay 37

3.4.7 Preparation The S9 mix is prepared as described for the Ames II assay of 30 % S9-Mix (Sect. 2.5.7 ). Instructions for preparing appropriate volumes of the S9 mix for one to ﬁ ve strains are given in Table 5 .

3.4.8 S9 Booster Some batches of S9 can lead to signs of toxicity when tested with Solution (Optional) the positive control chemical 2-AA, especially in strains TA100 and TA1537. Xenometrix AG therefore provides the S9 together with a “S9 booster solution” to protect strains TA100 and TA1537 from possible toxic S9 effects. This solution will be mixed with the exposure medium when using S9 in the Ames MPF assay. Procedure for assays with S9 : The S9 booster solution is mixed with the Salmonella exposure medium at a ratio 1:667 (e.g., 10 mL exposure medium + 15 μL booster solution). The required volume of exposure medium/S9 booster solution mixture is prepared at the day of the assay.

3.4.9 Preparation The transfer from the dilution plate to the 24-well exposure plates of Exposure Cultures is performed as described for the Ames II (Sect. 2.5.8 ). The preparation of the TA98, TA1535, and TA1537 exposure cultures is the same as described for the Ames II, whereas TA100 and the E. coli are more diluted in the exposure medium. The two E. coli strains are grown individually overnight and then exposed together to a test sample. This “EC Combo” has the following advantages over an individual exposure: The two strains are not equally sensitive to different chemicals and it is always the more sensitive strain that dominates in the EC Combo assay which allows to detect more mutagens than with the single strains alone [ 17 ]. Table 6 lists the volumes of the exposure culture components. The 24-well plates are incubated at 37 °C for 90 min ( Salmonella , E. coli without S9) or 20 min ( E. coli with S9) at 37 °C, 250 rpm. It is strongly recommended to use a shortened

Table 5 Preparation of 30 % S9 mix using self-made reagents (for one test compound)

Volume for Volume for Volume for Volume for Volume for Stock one strain two strains three strains four strains ﬁ ve strains (M) Reagent (mL) (mL) (mL) (mL) (mL)

1.00 KCl 0.043 0.076 0.110 0.146 0.179

0.25 MgCl 2 × 6H2 O 0.042 0.074 0.106 0.141 0.172 0.20 G-6-P 0.033 0.058 0.083 0.111 0.136 0.04 NADP 0.130 0.230 0.330 0.440 0.540

0.20 NaH 2 PO4 buffer 0.663 1.173 1.683 2.244 2.754 S9 fraction 0.390 0.690 0.990 1.320 1.620 Final volume 1.301 2.301 3.302 4.402 5.401 38 Sini Flückiger-Isler and Markus Kamber

Table 6 Preparation of exposure cultures

Exposure Bacterial Strain medium culture Strain dilution (mL) (mL) S9

TA98 1:10 6.3 0.7 – 1:10 5.25 0.7 1.05 mL TA100 1:20 6.65 0.35 – 1:20 5.6 0.35 1.05 mL TA1535 1:10 6.3 0.7 – 1:10 5.25 0.7 1.05 mL TA1537 1:10 6.3 0.7 – 1:10 5.25 0.7 1.05 mL E. coli Combo uvrA 1:14.3 6.3 0.49 – pKM101 1:33.3 0.21 uvrA 1:14.3 5.25 0.49 1.05 mL pKM101 1:33.3 0.21

exposure time of 20 min for E. coli in the presence of S9. This clearly improves the relatively weak response after 90 min to the positive control 2-aminoanthracene (2-AA), as well as to benzo(a) pyrene (B(a)P), another indirect mutagen (Fig. 6 ). For other tested chemicals that are positive in the absence and presence of S9 (4-nitroquinoline-N -oxide, methyl methanesulfonate, streptoni- grin, N 4 -aminocytidine, cumene hydroperoxide, and glutaraldehyde), the reduction of the exposure time in assays with metabolic activation results in an equal or slightly better sensitivity. For all Salmonella tester strains and for E. coli without S9, the exposure time is 90 min.

3.4.10 Transfer The procedure is the same as for the Ames II (Sect. 2.5.9 ), except of the Exposed Cultures that 2.6 mL E. coli indicator medium is added to the E. coli Combo to 384-Well Plates plates, and that the E. coli Combo with S9 should be processed and Incubation after 20 min of exposure (see above).

3.4.11 Scoring, Plate scoring, toxicity assessment, and evaluation criteria for the Calculation, and Evaluation Ames MPF are the same as described in detail for the Ames II Criteria under Sect. 2.5.10 . Preliminary results for the TA100, TA1537, and TA1535 cultures can be obtained already after 1 day of incubation. Revertant growth of TA100 and TA1537 cultures is almost completed after 28 h, whereas the number of positive wells of the TA1535 plates will further increase overnight. TA98 and E. coli plates can be scored reliably only after 2 days. Ames II and Ames MPF Penta I Assay 39

50 50

40 40

30 30

20 min 20 20 90 min

Revertants/48 wells 10 10

0 0 0 6.25 12.5 25 50 0 5 10 20 40 2-aminoanthracene (µg/ml) Benzo(a)pyrene (µg/ml) Fig. 6 Effect of exposure time on the 2-AA and B(a)P induction proﬁ le with E. coli strains in the presence of Aroclor 1254-induced S9

Table 7 Positive controls and acceptable range for solvent and positive controls in the Ames MPF Penta I assay mean of positive wells

TA98 TA100 TA1535 TA1537 E. coli Combo

Without S9 Solvent control ≤ 8 ≤ 12 ≤ 8 ≤ 8 ≤ 12 2-NF (2.0 μg/mL) ≥25 4-NQO (0.1 μg/mL) ≥ 25 N4-ACT (100 μg/mL) ≥25 9-AAc (15 μg/mL) ≥ 25 4-NQO (2 μg/mL) ≥ 25 With S9 Solvent control ≤ 8 ≤12 ≤ 8 ≤ 8 ≤ 12 2-AA (0. 5–5.0 μg/mL) a ≥ 25 ≥ 25 ≥ 25 ≥25 2-AA (50 μg/mL) >2-fold baselineb 2 -NF 2-nitrofl uorene, 4 - NQO 4-nitroquinoline-N -oxide, N4 - ACT N4 -aminocytidine, 9 - AAC 9 aminoacridine, 2 - AA 2-aminoanthracene a Different concentrations of 2-AA may be used depending on the S9-inducing agent (Aroclor-1254 or phenobarbital/ ß- naphthofl avone), on the different Salmonella strains, and on the actual S9 batch. When using Aroclor-1254-induced S9, a 2-AA concentration of 2.5 μg/mL may be used for all Salmonella strains if the preparation of different concentrations seems to be too laborious. Since strain TA98 is considerably more sensitive to 2-AA than strains TA1535 and TA1537, the use of strain-specifi c 2-AA concentrations will better indicate eventual problems with the condition of a culture. The 2-AA dose–responses of all tester strains in the presence of S9 are given in the Xenometrix S9 Certifi cates of Analysis for each individual batch b 20 min exposure

An Ames MPF assay is considered valid if the solvent control and the positive control values are within the acceptable range (Table 7 ). An alternative way to evaluate the Ames MPF results is to use a detection limit based on the binomial formula employed by Smith et al. [ 24 ], since Ames test responses are not normally distributed, but follow a binomial distribution. 40 Sini Flückiger-Isler and Markus Kamber

4 Testing of Non-concentrated Water Samples

The Ames microplate format has also been adapted to test native non-concentrated water samples [ 25 ]. The use of a tenfold concentrated exposure medium, which will be diluted to single strength by the aqueous sample, allows to decrease the dilution factor in the assay 18.5-fold.

5 Conclusion

The Ames II/Ames MPF procedure has several signifi cant advantages over the standard Ames test: less test sample, less hands-on time, less S9 consumption, less hazardous waste, suitable for being automated, and easy colorimetric readout. Based on the many comparative studies, there is no evidence that the Ames II/Ames MPF is less capable than the standard Ames at detecting mutagens. The Ames II test can be used for the screening of a large number of samples or samples available in limited quantities in early drug discovery. It contains the frameshift strain TA98 and covers all six possible base-pair mutations detecting strains in TAMix. However, it will not detect frameshift mutations that specifi cally revert TA1537. Furthermore, the six base-specifi c strains are combined to a single culture—the TAMix—and thus diluted sixfold. Alternatively, TAMix can be replaced by the traditional strain TA100 in the Ames MPF 98/100 assay, hence using a classic base- pair strain with comparable spectrum of responsiveness [ 16 ]. The liquid Ames MPF procedure (e.g., Ames MPF Penta I) with its substantial advantages can be used with the traditional Salmonella and E. coli tester strains and covers all the genetic endpoints recommended in the OECD 471. It is therefore a valid alternative to the standard Ames plate test for testing of chemicals, pharmaceuticals, cosmetics, and—with an appropriate combination of strains—environmental samples such as surface water, wastewater, air, or sediment.

References

1. Maron DM, Ames B (1983) Revised methods 4. OECD (1997) Guideline for testing of chemi- for the Salmonella mutagenicity test. Mutat cals. Test guideline no. 471: bacterial reverse Res 113:173–215 mutation test. OECD, Paris 2. Mortelmans K, Zeiger E (2000) The Ames 5. Gee P, Sommers CH, Melick AS et al (1998) Salmonella /microsome mutagenicity assay. Comparison of responses of base-speciﬁ c Mutat Res 455:29–60 Salmonella tester strains with the traditional 3. Mortelmans K, Riccio ES (2000) The bacte- strains for identifying mutagens: the results rial tryptophan reverse mutation assay with of a validation study. Mutat Res 412: Escherichia coli WP2. Mutat Res 455:61–69 115–130 Ames II and Ames MPF Penta I Assay 41

6. Gee P, Maron D, Ames BN (1994) Detection ing in liquid microplate format using S. and classifi cation of mutagens: a set of base- typhimurium TA98 and TA100. EEMS 2006, specifi c Salmonella tester strains. Proc Natl Prague Acad Sci U S A 91:11606–11610 17. Flückiger-Isler S, Kamber M (2007) The Ames 7. Flückiger-Isler S, Baumeister M, Braun K et al MPF™ assays: novel mutagenicity testing in (2004) Assessment of the performance of the liquid microplate format using S. typhimurium Ames II assay: a collaborative study with 19 TA98, TA100, TA1535 and TA1537. SOT coded compounds. Mutat Res 558:181–197 2007, Charlotte 8. Gervais V, Bijot D, Claude N (2003) Assessment 18. Flückiger-Isler S, Kamber M (2009) The Ames of a screening experience with the Ames II™ test MPF™ Penta I assay: mutagenicity testing in and future prospects. In: European liquid microplate format using OECD Environmental Society 33th Annual Meeting: Guideline 471 compliant strains S. From Hazard to Risk 2003, Aberdeen, p 120 typhimurium TA98, TA100, TA1535, TA1537 9. Kamber M, Flückiger-Isler S, Engelhardt G and E. coli WP2 uvrA plus E. coli WP2 et al (2009) Comparison of the Ames II and [pKM101]. ICEM 2009, Florence traditional Ames test responses with respect to 19. Atienzar F (2009) Evaluation of a battery of early mutagenicity, strain specifi cities, need for genotoxicity assays to predict regulatory testing. metabolism and correlation with rodent carci- ADMET Meeting, January 22–23, Brussels nogenicity. Mutagenesis 24(4):359–366 20. de Aragão Umbuzeiro G, Rech CM, Bergamasco 10. Braun K (2001) Automation of the Ames II AM et al (2010) Comparison of the Salmonella / Assay: high through-put screening of muta- microsome microsuspension assay with the new genic substances; Aventis Pharma Deutschland microplate fl uctuation (MPF) protocol for test- GmbH, DI & A, Lead Optimization, Drug ing the mutagenicity of environmental samples. Safety Evaluation. MipTec ICAR, Basel Environ Mol Mutagen 51:31–38 11. Lorge E, Gervais V, Becourt-Lhote N et al 21. Kado NY, Langley D, Eisenstadt E (1983) A (2007) Genetic toxicity assessment: employing simple modifi cation of the Salmonella liquid the best science for human safety evaluation incubation assay. Mutat Res 121:25–32 part IV: a strategy in genotoxicity testing in 22. Flückiger-Isler S, Kamber M (2012) Direct drug development: some examples. Toxicol Sci comparison of the Ames microplate format 98(1):39–42 (MPF) test in liquid medium with the standard 12. Heringa MB, Stang A, van Vugt MATM et al Ames pre-incubation assay on agar plates by (2009) Ames II and high throughput Comet use of equivocal to weakly positive test com- assay for effi cient screening of drinking water pounds. Mutat Res 747:36–45 (sources) for genotoxic contaminants. Firenze, 23. Ames MPF™ Penta I instructions for use ICEM 2009, August 20–25, p 217 (2012) Xenometrix AG version 4.5_S 13. Xenometrix AG, Allschwil, Switzerland ( www. 24. Smith KEC, Heringa MB, Uytewaal M et al xenometrix.ch ) (2013) The dosing determines mutagenicity of 14. Molecular Toxicology Inc., Boone, USA hydrophobic compounds in the Ames II assay 15. Ames II instructions for use (2012) Xenometrix with metabolic transformation: passive dosing AG version 4.5_L versus solvent spiking. Mutat Res 750:12–18 16. Flückiger-Isler S, Kamber M (2006) The Ames 25. Ames MPF™ 98/100 AQUA instructions for MPF™98/100 assay: novel mutagenicity test- use (2012) Xenometrix AG version 4.51_S Chapter 3

Revised Procedure of the Bacterial Reverse Mutation Test for Genotoxic Evaluation of Nanoparticles

Christophe Pagnout , Stéphane Jomini , and Pascale Bauda

Abstract

A recent review of in vitro genotoxicity testing strategies for nanoparticles (NPs) revealed that the conventional Bacterial Reverse Mutation Test (Ames test) primarily yielded negative results, whereas the other in vitro genotoxicity assays were usually positive. Accordingly, the effi ciency of the test for NP evaluation was questioned, as was the NP entrance in bacterial cells. Indeed, prokaryotes are unable to perform endocytosis and NPs are too large to be transported through the pores of the cell wall. However, regardless of whether they have been internalized, the mutagenic potential of free radicals, produced intrinsically or indirectly by NPs adsorbed onto the bacterial cell walls, should be detected by the Bacterial Reverse Mutation Test . Another phenomenon that can explain the low test effi ciency is the lack of interactions between NPs and cells due to strong NP aggregation in the exposure medium of the assay and the presence of electrostatic repulsive forces between NPs and bacteria, which both carry overall negative charge. This hypothesis is supported by our recent study that revealed the mutagenic potential of NP-TiO2 using a revised procedure of the Bacterial Reverse Mutation Test , which improves the NP-cell interactions and the sensitivity of the test. In this chapter, we provide several recommendations for the genotoxic evaluation of NPs and propose a revised version of the Bacterial Reverse Mutation Test more in line with the specifi c properties of NPs.

Key words Bacterial reverse mutation assay , Ames test , Fluctuation , Nanoparticle , Mutagenicity , Genotoxicity , In vitro , Oxidative stress , Salmonella typhimurium , Escherichia coli

1 Introduction

The defi nition of nanoparticles (NPs), or ultrafi ne particles, continues to be the subject of debate. However, there is widespread agreement that NPs are particles with sizes between 1 and 100 nm that are manufactured specifi cally for their particular physicochemical properties, compared to bulk material. Engineered nanoparticles are increasingly being produced for use in a wide range of industrial and consumer products, raising concerns regarding their impacts on human and environmental health.

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_3, © Springer Science+Business Media New York 2014 43 44 Christophe Pagnout et al.

A large number of scientiﬁ c papers have recently been published to highlight these issues, but given the large number of variables such as the characteristics of NPs tested (different chemical composition, size, shape, surface recovery), suspension/dispersion protocols (sonication, stirring, mixing, use of a dispersing agent or not), toxicological assays used (endpoints, tested cells/species, methods of administration, dose range, exposure periods), and the relevance of these assays for evaluation of NPs, conclusively establishing safety or toxicity of NPs is still a major challenge. Additionally, it is becoming evident that NPs cannot be treated as chemical compounds with regard to their safety as their unique physicochemical properties introduce an additional level of complexity. According to the scientiﬁ c community, more work should be conducted to standardize (1) physicochemical characterization procedures; (2) dispersion protocols; (3) methods for (eco)toxicity evaluation, the existing assays having been designed for the evaluation of chemical substances; and (4) to identify standard reference NPs to correlate test results across laboratories to improve the reliability and reproducibility of data. The Bacterial Reverse Mutation Test is recommended by several regulation agencies (Organisation for Economic Co-operation and Development, Food and Drug Administration, International Conference on Harmonisation) for substance evaluation and, due to its simplicity and relatively low cost, is commonly employed as an initial screening method for genotoxic activity. In this chapter, we present a revised version of this test, taking into account in a more precise way particular properties of NPs, and we provide recommendations for genotoxic evaluation of NPs.

1.1 The Bacterial The Bacterial Reverse Mutation Test was initially developed by Reverse Mutation Bruce Ames in the 1970s [1 – 3 ]. This test evaluates the genotoxic- Test and Its Different ity of compounds by measuring their ability to induce reverse Procedures mutations (frameshift mutations or base-pair substitutions) in auxotrophic bacterial strains of Salmonella typhimurium (histidine –) or Escherichia coli (tryptophan –). The revertant bacteria are then detected by their ability to grow in absence of the amino acid required by the auxotroph strain. The strains used in the Bacterial Reverse Mutation Test present speciﬁ c features that make them more sensitive for the detection of mutagens such as (1) increased cell wall permeability with regard to large molecules due to mutation, which causes partial loss of the lipopolysaccharide barrier ( rfa mutation); (2) mutation in the bacterial cell system to excise and repair defects in the DNA resulting in the inability to repair damaged/mutated sections ( uvrB mutation); and (3), for speciﬁ c strains, a R-factor plasmid pKM101 that increases chemical and spontaneous mutagenesis by enhancing an error-prone DNA repair system [4 ]. Strain TA102 also contains the hisG428 mutation on the multicopy plasmid pAQ1 to enlarge the number of target sites (Table 1 ) [ 4 , 5 ]. The OECD guideline Revised Procedure of the Bacterial Reverse Mutation Test for Genotoxic Evaluation… 45

b g/mL) g/mL) g/mL) uorene g/mL) μ μ μ μ (10 (0.4 (25 ng/mL) (3.5 (0.1 Sodium azide 9-Aminoacridine 2-Nitroﬂ Sodium azide Cumene hydroperoxide 4-Nitroquinoline-1-oxide Positive controls a Spontaneous revertants 3–10 3–10 10–30 60–70 3–15 3–15 ) ) ) ) + ) r r r r r ) r pAQ1(Tet pAQ1(Tet pKM101(Amp Plasmid (resistance) − − pKM101(Amp pKM101(Amp pKM101(Amp pKM101(Amp

uvrA uvrA Repair defect uvrB uvrB uvrB uvrB uvrB uvrB

Additional mutations LPS defect rfa rfa rfa rfa rfa rfa uctuation procedure tester strains recommended by the OECD guidelines tester strains recommended by the OECD guidelines Base-pair substitutions Frameshifts Frameshifts Frameshifts Base-pair substitutions Transitions/transversions Transitions/transversions Transitions/transversions

Escherichia mutation Reversion event Reversion Histidine mutation hisG46 hisD3076 hisD6610 hisD3052 hisG46 hisG428 trpE trpE and

Salmonella

city of Bacterial strains S. typhimurium TA1535 TA1537 TA97 TA98 TA98 TA100 TA102 E. coli WP2 uvrA WP2 uvrA (pKM101) Positive controls used without S9 mix Spontaneous revertants expected in the 96-well control plate with the ﬂ a b Table Table 1 Speciﬁ 46 Christophe Pagnout et al.

for testing of chemicals (OECD No. 471) recommends the use of at least the fi ve following strains in its testing strategy: S. typhimurium TA1535, TA1537 (or TA97 or TA97a), TA98, TA100, and TA102 (or E. coli WP2 uvrA or E. coli WP2 uvrA (pKM101)) [6 ]. The conventional method for performing the Bacterial Reverse Mutation Test , also called the plate incorporation method / procedure or more generically the Ames test , consists of exposing the bacterial suspensions (100 μL of ~10 9 viable cells/mL) to grade concentrations of the test substance (50 or 100 μL) in the presence or absence of an exogenous metabolic activation system (500 μL of S9 mix containing an adequate amount of post-mitochondrial fraction of rat liver homogenate). This metabolic system is optionally added to simulate the effects of mammalian metabolism as some compounds are not mutagenic themselves, unlike their metabolic products. These mixtures are immediately supplemented with a soft overlay agar (2.0 mL) and poured into Petri dishes onto minimal medium. After 2 or 3 days of incubation at 37 °C, revertant colonies are counted and compared to the background rate of reverse mutation to establish the genotoxicity of the substance. The signifi cance is then determined by several statistical approaches [ 7 , 8 ]. A detailed description and recommendations for performing this test are given in [4 , 9 – 13 ]. Several other procedures for performing the Bacterial Reverse Mutation Test have been developed in the past decades, with the main objectives of making the test more sensitive or more suitable for the evaluation of substances with specifi c properties (volatile chemicals or gases, small volume of sample to test, presence of histidine in the test substance). The most commonly used variant is the preincubation method / procedure , in which an incubation step of 20–90 min at 30–37 °C in phosphate buffer or S9 mix (500 μL) is introduced prior to adding the soft agar and pouring into Petri dishes as previously described. During the preincubation period, tubes are aer- ated by shaking. The preincubation procedure is often reported to be more sensitive than the conventional plate incorporation method as short-lived mutagenic metabolites have a better chance of reacting with the tester strains than when the incubation mixture is immediately poured into Petri dishes. The preincubation procedure , fully described in [10 ], was used in the 1980s and 1990s for large-scale chemical testing [14 – 17 ]. Another popular alternative to the plate incorporation method is the fl uctuation method / procedure , in which the bacterial growth is entirely performed in liquid medium and not on solid agar medium [ 18 , 19 ]. The medium contains a purple pH indicator (purple bromocresol), which changes color and turns yellow when the pH decreases due to the metabolic process of growing bacteria. After bacterial cell incubation with the test substance for up to 5 days Revised Procedure of the Bacterial Reverse Mutation Test for Genotoxic Evaluation… 47

in 96-well microplates, the number of revertants is scored by counting the number of yellow wells. Genotoxicity is then established by comparing the number of yellow wells in treated plates and in the control and signifi cance is established by Chi-square analysis [ 20 ]. The fl uctuation method has not been reported to be much more sensitive than the plate incorporation method , but it allows testing of higher concentrations of samples (>80 % against <2 % v/v for the plate incorporation), and is less time consuming in terms of protocol execution (no agar plate to prepare) and reading results (counting 96 wells faster than counting colonies on agar plates). Detailed protocols for the fl uctuation procedure have been described in [19 , 21 , 22 ]. Other alternative procedures have been developed and used less frequently, such as the suspension method [ 23 ] and procedures using desiccators or other sealed vessels to test volatile substances or gases [10 , 24 , 25 ].

1.2 NP Genotoxicity NPs can induce genotoxic effects via physical interactions with and Evaluation Using genomic DNA or components that determine its integrity (direct the Bacterial Reverse primary mechanism), or indirectly by the production of reactive Mutation Test oxygen species (ROS) via active redox cycling on their surfaces, especially on metal-based NPs (indirect primary mechanism), or also through activation of neutrophils and macrophages mediating an inﬂ ammatory reaction (secondary mechanism). For review on NP genotoxicity, refer to [ 26 – 30 ]. Over the past decade, the traditional Bacterial Reverse Mutation Test (Ames test) has been used by several authors for NP evaluation. In about 20 published studies, the test was always negative or weakly positive, whereas other in vitro genotoxicity assays (e.g., comet assay, micronucleus assay), sometimes performed in parallel with the same NPs, gave mostly positive results [ 29 ]. Therefore, the efﬁ ciency of the Bacterial Reverse Mutation Test , and more particularly the degree of NP uptake by cells, was questioned. Indeed, prokaryotes are unable to perform endocytosis and NPs are too large to be transported through the pores of the cell wall [ 27 , 29 , 31 , 32 , 34 ]. Even if some authors are able to locate NPs in the bacterial cytoplasm [ 33 , 34 ], the ability of NPs to penetrate the double-lipid bilayer of Gram-negative bacteria is still the subject of debate. The presence of these NPs inside the cells may be attributed to the precipitation of metal ions released from water-soluble NPs or a forced entry following multiple preparation steps for transmission electron microscopy (centrifugation, dehydration, embedding, slicing). However, regardless of internalization, the mutagenic potential of free radicals produced intrinsically by NPs adsorbed onto the bacterial cell walls should be detected by the Bacterial Reverse Mutation Test , particularly for strains TA102 and TA104, which are especially sensitive to oxidative mutagens [5 , 10 ]. 48 Christophe Pagnout et al.

If we closely examine the way in which the traditional Bacterial Reverse Mutation Test is carried out, bacterial cells and NPs are mixed together in an aqueous medium containing high salt concentrations and various organic compounds, originating from the nutritive medium in which the cells were grown overnight. This mixture is then supplemented with a soft agar that will quickly solidify and that also contains salts and organic substances. Considering theories on the adhesion of colloidal particles such as the DLVO [ 35 , 36 ], the probability of NPs entering into contact with bacteria is strongly minimized by their own aggregation and their possible interactions with material other than bacterial cells present in the exposure medium.

1.3 Revised We previously highlighted the paramount role that electrostatic Procedure forces play in the interactions between NPs and bacteria, and the of the Bacterial importance of medium characteristics such as pH, ionic strength, Reverse Mutation Test and electrolyte composition in controlling these forces [ 37 ]. We for NP Evaluation also examined the behavior of TiO2 -NPs in complex medium, such as the test medium of the Bacterial Reverse Mutation assay, and revealed their immediate and rapid aggregation as soon as they are

introduced [ 38 ]. It was also shown that TiO2 -NPs, which usually display an isoelectric point (IEP) around pH 7, were also shown to be negatively charged in this medium. Similarly, the S. typhimurium strains carry an overall negative charge due to the presence of car- boxylic and phosphate groups on their surfaces [ 39 ]. Accordingly, interaction and subsequent contact between NPs and bacteria are limited by repulsive electrostatic forces between each other, which may explain the negative results commonly obtained by the Bacterial Reverse Mutation Test when evaluating the genotoxicity of NPs. This hypothesis is supported by the fact that a simple preexposure step in a medium favoring NP-cell interactions before conducting the test leads to the outcome of a positive Bacterial

Reverse Mutation assay on NP-TiO 2 [ 38 ]. S. typhimurium TA102 was found to be more sensitive than TA97, TA98, and TA100 for the detection of mutagenic events. To make the Bacterial Reverse Mutation assay more sensitive and accurate for the detection of NP genotoxicity, we suggest that several aspects of the test should be considered. Preexposure Medium : It has been shown that in weakly acid solution mainly composed of monovalent ions at low concentration, such as a 10 mM NaCl solution prepared in ultrapure water (pH 5.5–6), NPs with an IEP ≥ 6.5 are positively charged and quickly adsorb onto the bacterial cell surface [ 37 , 38 ]. This saline solution can then be used as preexposure medium for the assessment of NPs having IEPs ≥ 6.5 and may be slightly acidiﬁ ed with HCl for NPs having IEPs below this limit. It should be noted that the phosphate buffer commonly used in the preexposure procedure is not appropriate for the NP evaluation. Indeed, because this Revised Procedure of the Bacterial Reverse Mutation Test for Genotoxic Evaluation… 49

buffer has a pH close to 7.4 and a high concentration of mono-/ trivalent ions, it promotes rapid and strong NP aggregation. Fluctuation vs. Plate Incorporation Procedure : The fl uctuation procedure was found to be more sensitive than the plate incorporation method for evaluation of NPs. A possible explanation could be the low diffusion effi ciency of NPs into the agar gel when compared to the liquid medium. Several studies demonstrated that particle properties (size, distribution, shape, surface charges) along with the relaxation time of the viscoelastic material each play a role in determining the extent to which NP mobility/transport is reduced within a gel matrix [ 40 , 41 ]. Mid -log vs. Early Stationary Phase of Growth : Bacterial cells in the mid-log phase instead of those in the early stationary phase (as recommended in the conventional Bacterial Reverse Mutation procedure ) are more sensitive for the evaluation of NPs [38 ]. Several factors can explain this phenomenon, such as the intracellular level of the alternative sigma factor RpoS, which accumulates in stationary phase and is responsible for the transcriptional activation of genes involved in stress resistance and DNA repair activity [ 42 , 43 ], or the cell membrane composition and fl uidity, known to change during different bacterial growth phases [ 44 ]. Use of S9 Mix : Adding the S9-mix fraction to mimic the mammalian metabolic conditions has no reason to be performed during the Bacterial Reverse Mutation Test since NPs do not produce metabolites. Moreover, this fraction contains a high protein concentration that may increase the aggregation of NPs and limit NP-cell interactions. As a consequence, we do not recommend performing the test using the S9-mix fraction.

1.4 Other Important Characterization : To correlate the properties of the NPs to their Considerations Prior (geno)toxic potential and ensure that results are reproducible and to Testing NPs accurate, characterization of NPs is essential. However, such determination of every possible characteristic of NPs is complex, time 1.4.1 Nanoparticles consuming, and expensive. Indeed, in addition to the information to Be Tested generally given by the manufacturer (CAS no., structural formulae, molecular structure, composition, purity, morphology, surface chemistry, major commercial uses, method of production, etc.), a minimum set of important properties must be addressed in (geno) toxicological studies (Table 2 ). These include primary size, shape, surface area, chemical composition (purity or additives), crystalline phase, surface chemistry (coating/modiﬁ cation), and, for nanoparticles in suspension (e.g., stock suspension), size distribution (aggregation/agglomeration state) and surface charge (zeta potential) [ 45 ]. As recommended by the OECD Working Party on Manufactured Nanomaterials [ 46 ], other relevant information, characteristics, or properties can be given such as dustiness, water

solubility, porosity, K ow , crystallite size, redox potential, radical formation potential, photocatalytic activity, etc. 50 Christophe Pagnout et al. On a surface On a surface Dispersed in liquid Dry powder Dispersed in liquid Aerosol On a surface On a surface On a surface Dispersed in liquid NP state during analysis Embedded Dispersed in liquid m m μ μ m m m m μ μ μ μ ~2 nm–1 ~0.2 nm–1 ~1 nm–7 ~5 nm–10 ~5 nm–1 ~5 nm–2 ppt range ppm range ppb range Variable ppm Size Shape Agglomeration/aggregation Size Shape Agglomeration/aggregation Size Agglomeration/aggregation Dispersion stability Surface area Surface charge Agglomeration/aggregation Dispersion stability Size Surface chemistry/contamination Surface chemistry/contamination Surface chemistry/contamination Photocatalytic activity Surface chemistry/contamination Photocatalytic activity Solubility/dissolution + + + + + Liquid dispersant Measured parameters Sensitivity NP suspension + + + + + + + + + + NP powder + + + + + + (EDX) Techniques Techniques Scanning electron microscopy (SEM) electron microscopy (TEM) Transmission Dynamic light scattering (DLS) (BET) Emmett, and Teller Brunauer, Zeta potential Scanning mobility particle sizer (SMPS) X-ray photoelectron spectroscopy (XPS) dispersive X-ray spectroscopy Energy Secondary ion mass spectrometry (SIMS) Fluorescence spectroscopy Redox potential UV–Vis spectroscopy Ion exchange chromatography pH Table Table 2 for (geno)toxicity studies Common techniques for measuring NP characteristic relevant Revised Procedure of the Bacterial Reverse Mutation Test for Genotoxic Evaluation… 51

Stock Suspension Preparation : The method used to prepare the stock suspension directly impacts the NP degree of dispersion and, consequently, their effects on organisms. Several procedures have been described in the literature, but we recommend the use of the Protocol for Nanoparticle Dispersion developed by PROSPEcT in support of the OECD Sponsorship Programme [47 ], which is applicable to a wide range of NPs. This protocol is derived from the BS ISO 14887 [ 48 ] sample preparation — dispersing procedures for powders in liquids . Briefl y, a mass of dry powder (200 mg for a fi nal NP stock suspension at 10 g/L) was weighed into a 20–25 mL glass container. The contents of the container were then mixed into a thick paste with a few drops of deionized (DI) water using a pre-cleaned metal spatula with suffi cient energy to remove visible aggregates. Next, the paste was suspended in 15 mL DI water and sonicated (ultrasonic probe tip half way down the small vial) with an ultrasonic probe for 20 s at 90 % amplitude to disperse the NP agglomerates. Where required, this sonication time can be extended with an appropriate cooling system. Finally, the sonicated suspension was mixed into the end volume (20 mL for a fi nal NP stock suspension at 10 g/L) of DI water using a glass stirring rod to make the stock. For NPs coated with hydrophobic organic molecules, a similar protocol can be used but with ethanol, DMSO, or another appropriate solvent as a dispersant instead of DI water. Storage and Stability : When stored, NP stock suspension must be kept at 4–6 °C away from light in a glass/polyethylene sealed container. Additionally, the stability of the suspension (agglomeration, metal ion/capping agent released) must be investigated before beginning a new experiment. It is recommended that dynamic light scattering (DLS) or UV/visible spectrometric measurements in solution be conducted before the test to ensure that the NP characteristics meet those obtained with the freshly prepared suspension. Metal ions and capping agent released in the solution can be measured in the supernatant by ICP-MS after NP removal by (ultra)centrifugation. The speed and time of (ultra)centrifugation are determined by the size of the NPs. We recommend 2 h at 40,000 × g for NPs 10–30 nm, 1 h at 20,000 × g for NPs 30–60 nm, and 30 min at 14,000 × g for NPs 60–100 nm. A second (ultra) centrifugation or a fi ltration can be performed to ensure that all NPs are removed. Sterility : The presence of microorganisms in the NP suspension may affect the accuracy of the test. Careful attention should be paid when attempting to fi nd a suitable method of sterilization for each NP suspension, and the effects of sterilization on the integrity of the physicochemical properties of NPs need to be investigated.

1.4.2 Controls In addition to the conventional controls performed during the to Be Performed genotoxicity assays (dispersant/solvent, sterility, and positive controls), parallel controls must be included when assessing NPs. 52 Christophe Pagnout et al.

Supernatant Control : Components other than NPs in the stock suspension, such as impurities remaining from the synthesis process, stabilizers or surfactants added during the dispersion procedure (SDS, PVP, Tween, etc.), or even products from the NP alteration (metal ions or capping agents), may adversely affect test results. Accordingly, we recommend an additional control consisting of the supernatant of the NP stock suspension after NP removal by (ultra)centrifugation at the speed/time mentioned above. Metal Ions / Coating Control : Throughout the assay, metal ions or capping agents may be released from NPs. To ensure that the results obtained from the test are NP dependant and not related to an artifact of the NP alteration, we recommend additional controls for water-soluble or embedded NPs consisting of the correspond-

ing metal salt in solution (AgNO 3 , CuCl2 , ZnCl2 , etc.) and the free capping agent (citrate, PVP, PEG, etc.). NP Positive Controls : Suitable reference NP suspensions with good physical and chemical stability and mutagenic effects detectable by the Bacterial Reverse Mutation Test must be found or designed. A small number of reference materials already exist in the fi eld of manufactured NPs, but most of them are certifi ed for size and are used mainly to calibrate instruments which measure particle size. Currently, in nano-genotoxicology no NP positive control has been established or universally accepted for the tests. These reference NPs are necessary in the validation of tests, assessment of test method performances, and laboratory profi ciency.

2 Materials

2.1 Equipment Note 1 : This section does not consider the preparation of the NP stock and Supplies suspension and the NP characterization. For these purposes refer to [47 , 48 ] and [45 ], respectively. Moreover , this section does not consider the preparation of frozen working cultures and their long -term storage , or confi rmation of tester strain genotypes. For these purposes, refer to [4 , 10 , 13 ]. Large equipment : Autoclave, laminar fl ow hood, centrifuge, ultracentrifuge for supernatant control preparation, gyratory and stationary incubators at 37 °C, and a gyratory shaker at room temperature for preincubation, ultrapure, and deionized water source, 4 °C fridge, −80 and −20 °C freezer. Small equipment : Sterile 5 or 10 mL polypropylene culture tubes, 1.5 mL Eppendorf tubes, 50 mL Falcon tubes, micropipettors (0–1,000 μL) and supply of sterile disposal tips, multichannel dis- penser for distribution into wells of the microtiter plates, sealed fi ltration units (0.22 μm pore size) for fi lter sterilization of stock solutions, balance, vortex, and 96-well microtiter plates. Revised Procedure of the Bacterial Reverse Mutation Test for Genotoxic Evaluation… 53

2.2 Stock Solutions Nutrient broth medium : 2.5 g Oxoid nutrient broth No. 2 is dissolved and Media in 100 mL of deionized water and autoclaved at 121 °C for 15 min. Medium stored at room temperature in the dark can be used for at least 1 month. NaCl preexposure solution ( 10 mM ): 58.44 mg of NaCl is dissolved in 100 mL of ultrapure water (18.2 MΩ⋅cm at 25 °C) and autoclaved at 121 °C for 15 min. This preexposure solution should be slightly acidiﬁ ed with HCl for NPs having IEPs below 6.5. Sterile water : Ultrapure water (18.2 MΩ⋅cm at 25 °C) and deionized water are sterilized by autoclaving at 121 °C for 15 min and used for the preparation of NP dilutions and for the test procedure, respectively.

Biotin solution (0.01 % w / v ): 10 mg of D -biotin is dissolved in 100 mL of deionized water, and the solution is then sterilized by ﬁ ltration through a 0.22 μm ﬁ lter. Stock solutions stored at 4 °C can be used for at least 6 months.

Histidine solution ( 0.1 % w / v ): 100 mg of L -histidine is dissolved in 100 mL of deionized water, after which the solution is sterilized by ﬁ ltration through a 0.22 μm ﬁ lter. Stock solutions stored at 4 °C can be used for at least 6 months.

Tryptophan solution (0.1 % w / v ): 100 mg of L -tryptophane is dissolved in 100 mL of deionized water, and the solution is then sterilized by fi ltration through a 0.22 μm fi lter. Stock solutions stored at 4 °C can be used for at least 6 months. Ampicillin solution (5 % w / v ): 500 mg of ampicillin is dissolved in 10 mL of deionized water, after which the solution is sterilized by fi ltration through a 0.22 μm fi lter. Stock solution stored at −20 °C can be used for at least 4 months. Tetracycline solution (0.5 % w / v ): 50 mg of tetracycline is dissolved in 10 mL of ethanol. The solution need not be sterilized. Stock solution stored at −20 °C can be used at least 4 months.

Glucose solution ( 40 % v / v ): 40 mg of D -glucose is dissolved in deionized water to a fi nal volume of 100 mL. The solution is then sterilized by fi ltration through a 0.22 μm fi lter. Stock solution stored at 4 °C can be used at least 6 months.

Davis -Mingioli salts (5.5 ×): 38.5 g of K2 HPO4 , 11.0 g of KH2 PO4 , 5.5 g of (NH4 )2 SO4 , 1.375 g of trisodium citrate, and 550 mg of MgSO4⋅7H2 O are added and dissolved in the order listed in 600 mL of deionized water. The ﬁ nal volume is then adjusted to 1 L with deionized water, after which the solution is autoclaved at 121 °C for 15 min. Stock solution stored at room temperature can be used for at least 6 months. Purple bromocresol solution ( 0.2 % w / v ): 300 mg of purple bromocresol is dissolved in 150 mL of deionized water, and the solution 54 Christophe Pagnout et al.

is then autoclaved at 121 °C for 15 min. Stock solutions stored at 4 °C can be used for at least 6 months. Test reagent mixture : A volume of 108 mL of Davis-Mingioli salts (5.5×) is supplemented with 24 mL of D -glucose (40 % w / v ), 6 mL of D -biotin (0.01 % w / v ), 0.3 mL of L -histidine (0.1 % w / v ) (for tester strains S. salmonella ) or tryptophan (0.1 % w /v ) (for tester strains E. coli ), and 12 mL of bromocresol purple solution (0.2 % w / v ). This mixture must be prepared the day of the assay and immediately before the test and should not be stored. Positive controls : Positive controls are dissolved under sterile conditions in dimethyl sulfoxide at the following concentrations, 9-aminoacridine (6 mg/mL), 2-nitroﬂ uorene (160 μg/mL), sodium azide (50 μg/mL), cumene hydroperoxide (2 mg/mL), and 4-nitroquinolone-1-oxide (100 μg/mL). These stock solutions stored at −20 °C can be used at least 1 year.

3 Methods

Note 2: idem Note 1

3.1 Preparation Bacterial overnight cultures are started from frozen glycerol stocks of Bacterial Strains by scraping a small portion of frozen cells and inoculating into liquid Oxoid nutrient broth No. 2 supplemented with appropriate antibiotics (Table 1 ). Ampicillin was added to a ﬁ nal concentration of 25 μg/mL and tetracycline to 2 μg/mL. Cultures are then grown overnight in the dark at 37 °C in a rotary shaker and used the next morning to inoculate 50 mL of the same fresh medium in 250 mL Erlenmeyer ﬂ asks. These new cultures are incubated as

described above until reaching an absorbance density (OD 600nm ) between 0.4 and 0.6 (mid-log phase of growth), at which point the Erlenmeyer ﬂ asks are removed and the bacterial suspensions transferred into individual 50 mL Falcon tubes. Cells are then washed twice by centrifugation (10 min at 7,000 × g at room temperature) and resuspension in 50 mL of NaCl preexposure solution. Cells are ﬁ nally redispersed in an adequate volume of NaCl preexposure

solution to obtain a ﬁ nal absorbance density (OD600nm ) of 5.0.

3.2 Preparation The concentration range is prepared from the NP stock suspen- of the NP sions in sterile 1.5 mL Eppendorf tubes by serial dilutions (1:1, Concentration Range 1:10, 1:100, and 1:1,000) in sterile ultrapure water. These ratios can be modiﬁ ed depending on the concentration of the initial NP stock suspension (usually 10 g/L) and potential toxicity of NP evaluated (refer to the toxicity assessment of NP section). Dilutions must be prepared immediately before use in the preincubation procedure and cannot be stored for future assays. Revised Procedure of the Bacterial Reverse Mutation Test for Genotoxic Evaluation… 55

3.3 Preincubation A total of 200 μL of the bacterial suspension prepared as described Procedure above was transferred to 100 mL ﬂ asks containing 20 mL of the NaCl preexposure solution. Next, 200 μL of NPs at various concentrations were added and the ﬂ ask was incubated for 30 min at 20 °C in the dark under orbital agitation (150 rpm). The time of incubation could be extended or reduced depending of the NP toxicity.

3.4 Mutagenicity After the preexposure step, fl asks are removed and aliquots of Test 500 μL are transferred into 50 mL Falcon tubes containing 2.5 mL of the test reagent mixture. A volume of 17 mL of sterile distilled water is then added to the tubes, giving a fi nal NP dilution factor of 4,000 (based on the concentration described above). A total of 200 μL from these mixtures is subsequently dispensed into 96-well microtiter plates, after which the plates are sealed in plastic bags and incubated at 37 °C for 5 days in the dark. Various controls are run concurrently, such as (1) negative controls corresponding to sterile distilled water or other disper- sants used in the preparation of the NP stock suspension (DMSO, ethanol); (2) positive control substances to confi rm the reversion properties of the strains (Table 1 ); (3) supernatant controls prepared as described in the Controls to Be Performed section; (4) in case of genotoxic evaluation of water-soluble or coated NPs, the corresponding metallic salt or the capping agent; and (5) all appropriate sterility controls: medium, NPs, reference mutagens used as positive controls, etc.

3.5 Interpretation A yellow well is considered positive and a purple well negative. The of Results test is considered valid if after 5 days of incubation, (1) all wells in the sterility control plates are still purple, (2) the results from the negative control plates are as expected (spontaneous revertants), and (3) between 90 and 96 wells in the positive control plates are yellow. NPs tested are considered mutagenic if the number of positive wells is signifi cantly greater in treated plates than in both negative control plates, the supernatant control, and when performed, in the metallic salt/capping agent treatments. Results obtained were tested for signifi cance by the chi-square analysis using Table 3 [20 ]. Based on the number of positive wells obtained with the negative control (column control), this table gives the minimum number of positive wells that must be obtained from the treated plates (column exposed) to be statistically signifi cant at α = 0.05.

3.6 Toxicity Toxicity is generally indicated by a reduction of positive wells in Assessment of the the treated plates relative to the negative control plates. It is also NPs possible to carry out an initial toxicity test to determine the most appropriate concentration range to assay. In the case of NP toxicity, only the higher tested dose should exhibit some toxicity. 56 Christophe Pagnout et al.

Table 3 Signiﬁ cance thresholds with α = 0.05 [20 ]

Control Exposed Control Exposed Control Exposed Control Exposed

0 3 25 36 50 62 75 84 1 5 26 37 51 63 76 85 2 7 27 39 52 64 77 86 3 9 28 40 53 65 78 87 4 10 29 41 54 66 79 87 5 12 30 42 55 67 80 88 6 13 31 43 56 68 81 89 7 15 32 44 57 68 82 90 8 16 33 45 58 69 83 90 9 17 34 46 59 70 84 91 10 19 35 47 60 71 85 92 11 20 36 48 61 72 86 93 12 21 37 49 62 73 87 93 13 22 38 50 63 74 88 94 14 24 39 51 64 75 89 94 15 25 40 52 65 76 90 95 16 26 41 53 66 77 91 96 17 27 42 54 67 78 92 96 18 28 43 55 68 78 93 96 19 30 44 56 69 79 94 20 31 45 57 70 80 95 21 32 46 58 71 81 96 22 33 47 59 72 82 23 34 48 60 73 83 24 35 49 61 74 83

3.7 Safety In addition to the safety precautions that must be taken in labora- Precautions tories when using potentially mutagen/carcinogen substances and microorganisms, even with the nonpathogenic tester strains used in the Bacterial Reverse Mutation assay, speciﬁ c recommendations have to be taken into account when handling NPs. All manipulation of NP powders has to be carried out in a glove box, glove bag, chemical fume hood, or other airborne con- taminant control system. NP in suspensions may be handled on the Revised Procedure of the Bacterial Reverse Mutation Test for Genotoxic Evaluation… 57

lab bench with appropriate personal protective equipment such as goggles, antistatic gloves, and a lab coat. The Bacterial Reverse Mutation assay has to be performed under sterile conditions in a well-ventilated area.

3.8 Disposal All materials contaminated with microorganisms have to be auto- of Wastes claved. Wastes containing NPs and/or mutagen compounds, such as positive controls, must be collected and managed as hazardous waste.

References

1. Ames BN, Gurney EG, Miller JA et al (1972) 13. Pillco A, de la Peña E (2014) Bacterial reverse Carcinogens as frameshift mutagens: metabo- mutation test: why, when and how to use. In: lites and derivatives of 2-acetylaminofl uorene Sierra LM, Gaivao I (eds) Genotoxicity and and other aromatic amine carcinogens. Proc DNA repair: a practical approach. Springer Natl Acad Sci U S A 69:3128–3132 Science, Heidelberg 2. Ames BN, Lee FD, Durston WE (1973) An 14. Mortelmans K, Haworth S, Lawlor T et al improved bacterial test system for the detec- (1986) Salmonella mutagenicity tests: tion and classifi cation of mutagens and carcin- II. Results from the testing of 270 chemicals. ogens. Proc Natl Acad Sci U S A 70:782–786 Environ Mutagen 8(Suppl 7):1–119 3. Ames BN, Durston WE, Yamasaki E et al 15. Zeiger E, Anderson B, Haworth S et al (1987) (1973) Carcinogens are mutagens: a simple Salmonella mutagenicity tests: III. Results test system combining liver homogenates for from the testing of 255 chemicals. Environ activation and bacteria for detection. Proc Natl Mutagen 9(9):1–110 Acad Sci U S A 70:2281–2285 16. Zeiger E, Anderson B, Haworth S et al (1988) 4. Maron D, Ames BN (1984) Revised methods Salmonella mutagenicity tests: IV. Results for the Salmonella mutagenicity test. Mutat from the testing of 300 chemicals. Environ Res 113:173–215 Mutagen 9(Suppl 12):1–158 5. Levin DE, Hollstein M, Christman MF et al 17. Zeiger E, Anderson B, Haworth S et al (1992) (1982) A new Salmonella tester strain (TA102) Salmonella mutagenicity tests: V. Results from with AT base pairs at the site of mutation the testing of 311 chemicals. Environ Mol detects oxidative mutagens. Proc Natl Acad Sci Mutagen 19(Suppl 21):1–141 U S A 79:7445–7449 18. Green MHL, Muriel WJ (1976) Mutagen test- 6. OECD No. 471 (1997) Organization for eco- ing using Trp+ reversion in E. coli . Mutat Res nomic development and cooperation guide- 38:3–32 lines for testing of chemicals, Paris, France 19. Hubbard SA, Green MHL, Gatehouse D et al 7. Margolin B, Kaplan N, Zeiger E (1981) (1984) The Fluctuation test in bacteria. In: Statistical analysis of the Ames Salmonella / Kilbey BJ, Legator M, Nichols W, Ramel C microcosm test. Proc Natl Acad Sci U S A (eds) Handbook of mutagenicity test proce- 78:3779–3783 dures, 2nd edn. Elsevier, Amsterdam 8. Stead A, Hasselblad J, Claxton L (1981) 20. Gilbert RI (1980) The analysis of fl uctuation Modeling the Ames test. Mutat Res 85:13–27 tests. Mutat Res 74:283–289 9. Gatehouse D, Haworth S, Cebula T et al (1994) 21. Flückiger-Isler S, Kamber M (2014) The Ames Recommendations for the performance of bacte- II and Ames MPF Penta I assay: a liquid micro- rial mutation assays. Mutat Res 312:217–233 plate format modifi cation of the classic Ames 10. Mortelmans K, Zeiger E (2000) The Ames test. In: Sierra LM, Gaivao I (eds) Genotoxicity Salmonella /microcosm mutagenicity assay. and DNA repair: a practical approach. Springer Mutat Res 455:29–60 Science, Heidelberg 11. Tejs S (2008) The Ames test: a methodological 22. Environment Canada (1993) Protocole-Test short review. Environ Biotechnol 4:7–14 de fl uctuation. Laboratoire C&P (CSL). 12. Gatehouse D (2012) Bacterial mutagenicity Montreal, Canada assays: test methods. In: Parry JM, Parry EM 23. Thompson ED, Melampy PJ (1981) An exam- (eds) Genetic toxicology: principles and meth- ination of the quantitative suspension assay for ods, methods in molecular biology, vol 817. mutagenesis with strains of Salmonella Springer Science, Heidelberg typhimurium . Environ Mutagen 3:453–465 58 Christophe Pagnout et al.

24. Hugues TJ, Simmons DM, Monteith LG et al 36. Verwey EJW, Overbeek JTG (1948) Theory of (1987) Vaporization technique to measure the stability of lyophobic colloids. Elsevier, mutagenic activity of volatile organic chemicals Amsterdam in the Ames/ Salmonella assay. Environ 37. Pagnout C, Jomini S, Dadhwal M et al (2012) Mutagen 9:421–441 Role of electrostatic interactions in the toxicity 25. Araki A, Noquchi T, Kato F et al (1994) of titanium dioxide nanoparticles toward Improved method for mutagenicity testing of Escherichia coli . Colloids Surf B Biointerfaces gaseous compounds by using gas sampling 92:315–321 bag. Mutat Res 307:335–344 38. Jomini S, Labille J, Bauda P et al (2012) 26. Gonzalez L, Lison D, Kirsch-Volders M Modifi cations of the bacterial reverse mutation (2008) Genotoxicity of engineered nanomate- test reveals mutagenicity of TiO2 nanoparticles rials: critical review. Nanotoxicology and byproducts from a sunscreen TiO2 -based 2:252–273 nanocomposite. Toxicol Lett 215:54–61 27. Singh N, Manshian B, Jenkins GJS et al (2009) 39. Jiang W, Saxena A, Song B et al (2004) NanoGenotoxicology: the DNA damaging Elucidation of functional groups on gram- potential of engineered nanomaterial. positive and gram-negative bacterial surfaces Biomaterials 30:3891–3914 using infrared spectroscopy. Langmuir 28. Botta A, Benameur L (2011) Nanoparticle 20:11433–11442 toxicity mechanisms: genotoxicity. In: Marano 40. Fatin-Rouge N, Starchev K, Buffl e J (2004) F, Lahmani M, Houdy P (eds) Nanoethics and Size effects on diffusion processes within aga- nanotoxicology. Springer, Heidelberg rose gels. Biophys J 86:2710–2719 29. Doak SH, Manshian B, Jenkins GJS et al (2012) 41. Chen-Ming L (2008) Nanoparticle mobility In vitro genotoxicity testing strategy for nano- in viscoelastic media. Ph.D thesis, University materials and the adaptation of current OECD of Iowa guidelines. Mutat Res 745:104–111 42. Hengge-Aronis R (1993) Survival of hunger and 30. Magdolenova Z, Collins A, Kumar A et al stress: the role of RpoS in early stationary phase (2013) Mechanisms of genotoxicity. A review gene regulation in E. coli . Cell 72:165–168 of in vitro and in vivo studies with engineered 43. Lange R, Hengge-Aronis R (1991) nanoparticles. Nanotoxicology. doi: Identifi cation of a central regulator of 10.3109/17435390.2013.773464 stationary- phase gene expression in Escherichia 31. Landsiedel R, Kapp MD, Schulz M et al coli . Mol Microbiol 5:49–59 (2009) Genotoxicity investigations on nano- 44. Yatvin MB, Gipp JJ, Klessig DR et al (1986) materials: methods, preparation and character- Hyperthermic sensitivity and growth stage in ization of test material, potential artifacts and Escherichia coli . Radiat Res 106:78–88 limitations—many questions, some answers. 45. PROSPECT (2009) Evaluation and assign- Mutat Res 681:241–258 ment of nanoparticle dispersion/characterisa- 32. Woodruff RS, Li Y, Yan J et al (2012) tion methodologies. In: Ecotoxicology Test Genotoxicity evaluation of titanium dioxide Protocols for representative Nanomaterials in nanoparticles using the Ames test and Comet support of the OECD Sponsorship assay. J Appl Toxicol 32:934–943 programme 33. Kumar A, Pandey AK, Singh SS et al (2011) 46. OECD (2008) List of manufactured nanoma- Cellular uptake and mutagenic potential of terials and list of endpoints for phase one of metal oxide nanoparticles in bacterial cells. the OECD testing programme, ENV/JM/ Chemosphere 83:1124–1132 MONO (2008) 13/REV, No. 6 34. Clift MJD, Raemy DO, Endes C et al (2012) 47. PROSPEcT (2010) Protocol for nanoparticle Can the Ames test provide an insight into nano- dispersion. In: Ecotoxicology Test Protocols object mutagenicity? Investigating the interac- for representative Nanomaterials in support of tion between nano-objects and bacteria. Nano- the OECD Sponsorship programme toxicology. doi: 10.3109/17435390.2012. 48. BS ISO 14887 (2000) Sample preparation— 741725 dispersing procedures for powders in liquids. 35. Derjaguin BV, Landau LD (1941) Acta International organization for standardization, Physicochim. URSS 14:733–742 Geneva, Switzerland Chapter 4

The Control of Hydrophobic Compound Exposure in In Vitro Tests for Genotoxicity

Kilian E. C. Smith

Abstract

Several of the OECD recommended tests for genotoxicity are in vitro tests, involving the use of open plastic vessels and culture plates, rich culture media, or metabolic activation. These result in volatile, sorptive, and biotransformation losses of the test compound, and thus to poorly defi ned exposure. This has implications for the apparent sensitivity of the assay as well as for establishing a reliable relationship between concentration and response. These issues are particularly relevant for hydrophobic organic compounds (HOCs) which are diffi cult to initially dissolve in aqueous media and particularly susceptible to sorptive losses. Therefore, the in vitro genotoxicity testing of HOCs requires techniques that compensate for losses and lead to defi ned and constant dissolved exposure concentrations. Here, passive dosing can play a useful role. Passive dosing involves a dominating reservoir of polymer-sorbed HOC acting as a partitioning source to the test medium, maintaining defi ned and constant dissolved concentrations. This chapter intro- duces passive dosing for the control of HOC dissolved concentrations during in vitro genotoxicity tests, covering considerations for selecting a suitable passive dosing polymer and its dimensioning, as well as procedures for cleaning and loading the dosing polymer with the test HOC(s).

Key words In vitro test , Genotoxicity , Exposure , Hydrophobic organic compound , Passive dosing

1 Introduction

A number of the OECD recommended tests for genotoxicity are in vitro tests, for example, the bacterial reverse mutation test, mammalian chromosome aberration test, mammalian cell gene mutation test, and mammalian cell micronucleus test [ 1 – 4 ]. Such in vitro tests represent important progress in the development of assays for elucidating the underlying mechanisms behind genotoxicity, minimizing material consumption, facilitating automation for enhanced throughput, but also reducing animal testing which is important for their regulatory and public acceptance. Many such in vitro tests involve the use of plastic culture ﬂ asks or microwell plates. Their manufacture out of plastic makes them biologically inert and cheap to produce, and their open nature allows for efﬁ cient gas

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_4, © Springer Science+Business Media New York 2014 59 60 Kilian E.C. Smith

exchange to support the biology. Unfortunately, these properties can also lead to unacceptable losses of some test compounds via sorption to the plastic material or volatilization [5 – 7 ]. This is com- pounded by experimental conditions that are often not optimal for minimizing losses. For example, in vitro assays with mammalian cells are normally conducted at elevated temperatures of 37 °C which can lead to high volatile losses. Furthermore, test chemicals can bind to the protein and lipid which are an integral part of the cell culture medium [ 8 , 9 ], further reducing their availability. Therefore, there is an incompatibility between the conditions typically used for in vitro testing, and those required for the optimal testing of chemicals. These problems are especially acute for hydrophobic organic compounds (HOCs). Due to their “water- hating” nature, it is challenging to initially dissolve them in aqueous media, and they are also particularly prone to sorption to the plastic culture vessel surfaces and medium constituents [10 , 11 ]. This can lead to quite considerable differences between the amount of test HOC added (i.e., the nominal concentration), and that which is actually available to cause the genotoxicity (i.e., the effective concentration) [8 , 9 ]. For an HOC that requires metabolic activation to exert its mutagenicity, the control of exposure is even more challenging due to continuous consumption by the enzymes [ 12 ]. Indeed, the issue of low availability, and the resulting limited effectiveness of metabolic activation, is specifi cally mentioned in the OECD bacterial reverse mutation test guideline [1 ]. The most serious consequence would be when a hydrophobic mutagen gives a false-negative result in the in vitro test due to low availability rather than because of any absence of genotoxic activity. However, poor defi nition and control of the exposure during in vitro testing also present additional challenges. It becomes diffi cult to compare test results for compounds with markedly different physical-chemical properties as these have very different exposure scenarios. Furthermore, for a given compound, varying exposure regimes between different genotoxicity tests complicate comparison of the results. Particularly relevant in this regard are efforts to elucidate the relationship between in vitro and in vivo observations [ 13 – 15 ]. These issues apply equally to the reference chemicals used as controls for the in vitro genotoxicity assays. If the aim is to simply establish whether a chemical is genotoxic or not, a suitable solution might be to add excess compound to the test system (provided of course this does not lead to cytotoxicity). In fact, a number of the existing OECD protocols for in vitro genotoxicity tests specifi cally advocate the inclusion of a test concentration lying above the test compound’s solubility in the medium [ 1 – 4 ]. However, when the goal is to describe the underlying relationship between exposure and response, more attention needs to be given to the actual exposure concentrations during the test. Note also that evidence of a consistent concentration-response is often used as a criterion for a positive response in in vitro Control of Hydrophobic Compound Exposure 61

genotoxicity assays [1 – 4 ]. Here, the incorrect control of exposure can mask such a relationship and thus remove this important assessment criterion. Several possibilities exist for the proper control of HOC exposure in in vitro genotoxicity tests. It is possible to better defi ne the exposure by analytical measurements [8 ] or modeling [5 ], but this does not solve the critical issue of limited compound availability leading to a negative test result. In such cases it is then necessary to take active steps to control HOC exposure, with the simplest strategy being reducing losses as much as possible [ 11 ]. For example, efforts might be directed towards using vessels made from more suitable materials such as glass. However, particularly for cell culture plates, this has the downside of increased costs and a less optimal cell adherence. Moreover, the need for gas exchange and thus of volatile losses remains, as does sorption of the HOC to medium constituents. Therefore, a more comprehensive approach is to maintain constant exposure concentrations via a resupply of test substance during the test. Flow-through systems might be used [ 16 ], but these are technically diffi cult to apply routinely in cell culture plates. Simpler approaches for maintaining constant and defi ned levels of test HOCs in in vitro set-ups are thus needed, and here passive dosing can play an important role [ 17 , 18 ].

2 Principles of Passive Dosing

Different metrics can be used to describe exposure, with one of the most ubiquitous and experimentally accessible being the dissolved concentration in the test medium. This is considered to be the effective concentration driving uptake and toxicity, or in the case of metabolic activation the enzymatic activity [19 , 20 ]. Passive dosing involves introducing a dominating reservoir of sorbed HOC into the test system where it functions a partitioning source such that equilibrium partitioning is established with the dissolved HOC (Fig. 1 ). Typically this partitioning source comprises the HOC dissolved in a chemically inert and biocompatible polymer such as silicone [ 21 , 22 ]. Any losses perturbing the dissolved HOC concentration are then compensated for by further partitioning from this reservoir of sorbed HOC. −1 When the polymer-sorbed (C Sorbed , μg L ) and dissolved (C Free , −1 μg L ) HOC are at equilibrium, C Free can be calculated from C Sorbed −1 via the equilibrium partitioning ratio ( KSorbed/Free , L L ), which depends on the properties of both the HOC and sorbing polymer: C C = Sorbed (1 ) Free K Sorbed/ Free 62 Kilian E.C. Smith

Volatilization Biomass dilution

Dissolved

Sorption to medium and vessel Partitioning Biotransformation

Reservoir of polymer sorbed HOC Fig. 1 Passive dosing for controlling the dissolved concentrations of hydrophobic organic compounds in in vitro tests for genotoxicity

The size of the passive dosing reservoir is dimensioned to be sufﬁ ciently large that it dominates any loss processes. In this

way, CSorbed can be assumed to remain constant, and thus also C Free (see (1 )), and passive dosing permits a precise deﬁ nition of the

test CFree . For an HOC requiring metabolic activation for its genotoxic activity, passive dosing brings additional beneﬁ ts [23 ]. The high solubility of the HOC in the dosing polymer means a large mass of test compound can be introduced into the assay. This buffers any high compound turnover in the aqueous phase, and thus leads to an increase in the apparent assay sensitivity. Due to their hydrophobic nature, HOCs preferentially partition to the dosing phase, and

this high turnover can be achieved even when CFree is low (see ( 1 )). Thus, passive dosing becomes a practical tool for introducing sufﬁ cient HOC mass into the in vitro set up at low noncytotoxic concentrations [ 12 ]. This is not possible when using conventional approaches such as spiking, where simply adding excess compound

leads to values of C Free that are close the compound’s aqueous solubility. In summary, for controlling the dissolved exposure concentration of an HOC in in vitro genotoxicity tests, passive dosing has the following desirable features:

1. C Free is determined by partitioning, and by varying C Sorbed can be controlled to any value between zero and aqueous solubility.

2. With proper dimensioning of the passive dosing format, C Free is determined by equilibrium partitioning and can be precisely

deﬁ ned by applying K Sorbed/Free values (see below for where to obtain KSorbed/Free values when using silicone as the passive dosing phase). Control of Hydrophobic Compound Exposure 63

3. When the polymer-sorbed reservoir of HOC is sufﬁ ciently large to dominate any losses, signiﬁ cant depletion is avoided

and C Free remains constant. 4. When the above features apply to all compounds in an HOC mixture, the mixture proﬁ le is both deﬁ ned and remains constant. 5. The HOC is introduced associated with a biocompatible and chemically inert polymer and artifacts associated with spiking cosolvents are avoided.

6. CFree can be analytically conﬁ rmed at experiment completion. This can be done either by direct measurement or by measuring the polymer-sorbed HOC levels and applying partitioning ratios. Passive dosing has been applied in a range of in vitro tests targeting different toxicity endpoints. These include general toxicity in the Microtox bacterial assay [ 18 ], immunotoxic effects in human cell lines [ 9 ], cytotoxicity and EROD activity in rainbow trout cell lines [ 24 ], dioxin-like activity in the DR-Calux® assay [25 ], as well as mutagenicity in the Ames II assay [ 23 , 26 ]. These studies have all shown the utility of passive dosing for providing rigorously deﬁ ned and stable C Free exposure concentrations, leading to an increase in the apparent test sensitivity when compared to conventional solvent spiking [23 – 26 ]. For example, a passive dosing format based on silicone O-rings was applied in the Ames II assay for determining the mutagenicity of amino- and nitro-PAHs in the absence of metabolic activation

[ 26 ]. The CFree -based effect concentrations when passive dosing were 3–33 times lower than those based on the nominal spiked concentrations, with the differences increasing for the more hydrophobic compounds due to their enhanced sorption and thus lower availability. The same silicone format was applied in the Ames II assay to determine the concentration-response of benzo(a)pyrene after metabolic activation by S9 [ 23 ]. When spiking, nominal concentrations had to exceed aqueous solubility before any mutagenicity was observed. In contrast, with passive dosing the concentration-response curves were more reproducible and shifted to

C Free levels that were lower than the nominal concentrations by several orders of magnitude. These differences were attributed to the sorptive losses and slow dissolution kinetics of the precipitated benzo(a)pyrene when spiking. This underlines the above observation that although adding excess compound can be used to determine a compound’s genotoxicity, to establish reliable concentration-response relationships, a more rigorous control of exposure is necessary. Such increases in the apparent test sensitivity directly result from having deﬁ ned and stable C Free values. In contrast, using nominal concentrations does not take into account the reductions in test compound availability due to test losses. 64 Kilian E.C. Smith

This chapter will cover passive dosing as a technique for the control of HOC dissolved concentrations during in vitro genotoxicity tests, including the considerations behind selecting a suitable passive dosing polymer, proper dimensioning of the format and procedures for cleaning, and then loading the dosing polymer with the test HOC(s).

2.1 Step 1: Selection For its proper application, the passive dosing polymer should of a Passive Dosing possess a number of characteristics: Phase 1. Linear sorption isotherms. HOC partitioning between the polymer and aqueous phases should be characterized by a single partition coeffi cient, and competitive sorption between test chemicals, as well as with medium constituents, avoided. 2. High capacity for the test HOC. This ensures that the polymer- sorbed compound can adequately compensate for any losses. 3. High permeability of the polymer towards the test HOC. This permits a rapid release of the test substance into the aqueous medium to effi ciently compensate for losses. 4. Biocompatibility. The polymer introduced into the in vitro assay should not result in any kind of toxicity. 5. Chemical inertness. This widens the selection of the solvents that can be used for pre-cleaning and loading the polymer with HOC while avoiding issues of swelling or cracking. The criteria of linear partitioning and a high capacity are normally met for absorbents but not for adsorbents. Silicone, specifi cally polydimethylsiloxane (PDMS) silicone, has proved to be a suitable choice of absorbent polymer for the passive dosing of HOCs in a number of toxicity test systems [ 21 , 22 ]. In view of its widespread application, the following discussion therefore focuses on silicone. Both single HOCs and their mixtures dissolve in the silicone matrix, and the sorption isotherms are linear over the full range of concentrations up to the upper limit as set by solubility [ 27 – 29 ]. Silicone has a high affi nity for HOCs, and the amounts of silicone needed to ensure it dominates as a partitioning source are practical to work with. The silicone matrix has a low mass transfer resistance towards HOCs, mediating their rapid release [30 , 31 ]. Under well-mixed conditions, mass transfer into the aqueous medium is typically rate limited by the HOC diffusion through the unstirred water layer immediately adjacent to the silicone surface [ 31 , 32 ]. Concentration gradients do not develop within the silicone matrix and it can be considered as a homogenous phase, which is necessary for using equilibrium partitioning ratios to

calculate C Free . Although a number of other absorbent polymers such as low-density polyethylene or polyoxymethylene are commonly used for the analytical handling of HOCs [ 33 , 34 ], these have lower permeabilities which under certain circumstances can Control of Hydrophobic Compound Exposure 65

lead to the formation of undesirable HOC concentration gradients within the polymer matrix. Silicone can also be obtained at reasonable cost and in a wide range of formats, which simpliﬁ es selection of formats suited for different experimental set-ups. It is chemically inert, can be easily sterilized and being available in highly pure forms, including medical-grade silicone, has excellent biocompatibility. Finally, a large database of silicone to water equilibrium partitioning ratios for HOCs exists (e.g., [35 ]), complemented by predictive relationships for compounds lacking measurements [36 ].

2.2 Step 2: For the correct dimensioning of the passive dosing format, the Dimensioning mass of silicone is important for ensuring negligible depletion, of the Passive Dosing whereas the silicone surface area in contact with the aqueous Phase medium ensures exchange is sufﬁ ciently rapid such that equilibrium partitioning is maintained in the face of sorptive and other losses.

2.2.1 Step 2.1: Passive Extensive depletion of the silicone should be avoided since this is

Dosing Phase Depletion directly reﬂ ected in decreasing C Free values (see (1 )). The mass distribution of an HOC between the silicone and the water compart- ments is given by [37 ]: 1 f = (2 ) Silicone V 1 + Water KV× Silicone/ Free Silicone

where ƒ Silicone is the fraction of the total compound mass in the silicone at equilibrium and V Silicone (L) and V Water (L) are the volumes of silicone and water, respectively. Provided that KSilicone/ Free >> V Water / VSilicone , then f Silicone tends to a value of 1 and the ﬁ rst criterion of negligible depletion is met. Thus when passively dosing

HOCs with log K Silicone/Free values of around 1,000 and above, V Water / VSilicone phase ratios less than around 10–20 are typically used [9 , 23 – 26 ]. For example, Bougeard et al. [ 26 ] and Smith et al. [23 ] applied silicone O-rings with a volume of 0.171 mL to passively dose 0.5 and 1 mL culture medium, respectively. Kramer et al. [24 ] used 0.177 mL silicone disks to passively dose 1.7 mL medium, and Booij et al. [ 25 ] cast 0.01–0.1 mL PDMS silicone into the base of culture wells for dosing 2 mL medium. For com-

pounds with KSilicone/Free values below 1,000, the selection of a suitable ratio is more critical since the HOC mass distribution is

shifted towards the aqueous phase, and reduced VWater / VSilicone ratios are needed to avoid depletion (see (2 )). An obvious solution for these compounds would be to apply polymers with a higher afﬁ nity, but these still remain to be identiﬁ ed and characterized. The above scenario pertains to pure water and without any

losses, and as such sets the upper limit of the V Water / VSilicone ratio required for negligible depletion. However, for in vitro genotoxicity 66 Kilian E.C. Smith

testing, passive dosing is applied to control C Free in solutions that contain medium constituents and cells that sorb the HOC. These increase the capacity of the aqueous phase for the HOC, and depletion of the silicone dosing phase is determined by the respective capacities of the silicone, water, and sorbing material for the compound. Experience has shown this not to be a signiﬁ cant issue when

working with more hydrophobic compounds (K Silicone/Free > 1,000), solutions containing moderate amounts of sorbing materials, and with typical volume ratios of less than 10–20 [9 , 23 – 26 ]. Nevertheless, when working with particularly rich media and dense suspensions of cells, additional depletion of the silicone by this sorbing material should be considered (see, e.g., [38 ]). In summary, silicone is a robust choice for the passive dosing

of HOCs with K Silicone/Free values greater than around 1,000. K Silicone/ Free can be used in (2 ) to ensure that the selected V Water / VSilicone ratio does not lead to an unacceptable decrease in C Silicone . In addition, it is a good strategy to use a conservatively low V Aqueous / VSilicone ratio, and a pilot study measuring depletion of the silicone under the experimental conditions can be a worthwhile investment.

2.2.2 Step 2.2: Passive The silicone surface area relative to the aqueous volume to be

Dosing Phase Exchange dosed plays a determining role in how quickly the equilibrium C Free Kinetics level is reached in the medium, and whether this is maintained in the face of losses. In most cases a simple pre-equilibration step can be incorporated into the in vitro test protocol to ensure equilibrium partitioning prevails. For example, in Smith et al. [ 23 ] the benzo(a)pyrene loaded silicone was left to equilibrate overnight with the medium prior to starting the exposure of the bacterial cells in the presence of S9 activation. When passive dosing is used in treatment regimens with metabolic activation, HOC loss kinetics due to the (bio)transformation process require consideration. The existence of (bio)transformation losses per se implies that the silicone dosing phase will become progressively depleted. However, provided that this is kept within acceptable limits, it is valid to

assume that CSilicone and thus also C Free remain constant (see (1 )). The existence of a signiﬁ cant (bio)transformation loss process also

means that C Free is deﬁ ned by steady state rather than equilibrium partitioning considerations, and passive dosing formats with faster

exchange kinetics will result in steady-state CFree levels that are closer to the equilibrium partitioning concentrations.

2.3 Step 3: Initial The value of the genotoxicity data relies on the controls showing Cleaning limited toxicity. Problems related to impurities in the silicone can of the Silicone be easily minimized by initially selecting high-quality silicones and incorporating a cleaning step. A rigorous cleaning process of washing the silicone with analytical grade and water-miscible solvents Control of Hydrophobic Compound Exposure 67

such as ethanol or methanol, followed by sequential rinsing of the silicone with distilled water to remove solvent traces has been shown to be effective [ 22 , 23 ]. For example, application of the above cleaning procedure to silicone used for passive dosing in the Ames II assay resulted in the silicone blanks displaying no cytotoxicity, and having assay responses that were the same as those of the DMSO-only negative controls.

2.4 Step 4: Loading Correct loading of the silicone with the test HOC is an important of the Silicone step since the silicone concentrations directly determine CFree (see with the Test HOCs (1 )). The silicone can be loaded to any level up to saturation,

directly translating into CFree levels in the medium up to the upper limit as set by aqueous solubility. Two approaches exist for loading the silicone: partitioning from a solution containing the test HOC or spiking using a concentrated solution. Irrespective of the loading approach used, during each step sufﬁ cient time should be factored in to ensure equilibrium partitioning between the loading solution and silicone. This is usually relatively fast, under quiescent conditions equilibration of an HOC between pure methanol or methanol:water and silicone is completed within hours [9 , 22 , 24 ]. Using a water-miscible solvent to make up the HOC loading solution simpliﬁ es later removal of the solvent via washing using water (see below). Loading by partitioning . Partitioning-based loading with a solution containing excess HOC saturates the silicone to provide

equilibrium C Free levels in the test that are at aqueous solubility. For solid HOCs, the crystals present in the loading suspension dissolve to replenish any depletion due to partitioning into the silicone [9 , 21 , 22 ]. Their continued presence at the end of the loading step indicates that the loading solution, and thus also the silicone, is saturated. This approach should also work for liquid HOCs, although most experience has so far been with solid HOCs. An alternative approach to saturate the silicone is to embed pure solid in the silicone matrix. For example, Kwon and Kwon [ 39 ] ﬁ rst deposited a layer of solid HOC on the bottom of a glass vial and then covered this with silicone. In this way direct contact of the solid HOC with the aqueous phase was avoided, and the embedded solid HOC continually saturated the silicone for passive dosing at aqueous solubility.

Often the aim is to investigate genotoxicity at increasing C Free levels up to the limit set by aqueous solubility. This requires loading the silicone with increasing concentrations of test HOC up to and including the saturation level. This can be done by either non- depletive or depletive partitioning. For non-depletive partitioning, the volume of loading solution used is sufﬁ ciently large such that depletion due to partitioning 68 Kilian E.C. Smith

into the silicone can be neglected [ 22 , 23 ]. An analogous equation to the above ( 2 ) is used to calculate the volume of loading solution needed for negligible depletion: 1 f = (3 ) Loading_ solution 1 V 1 +×Silicone K V Loading_/ solution Silicone Loading_ solution

where f Loading_solution (unitless) is the fraction of HOC remaining in −1 the loading solution after completion, KLoading_solution/Silicone (L L ) is the equilibrium partitioning ratio between the loading solution and −1 silicone, V Silicone (L ) is the volume of silicone, and V Loading_solution −1 (L ) is the volume of loading solution. If ƒ Loading_solution remains acceptably high, it can be assumed that the loading solution concentration is unchanged and thus the same as the nominal con-

centration. C Silicone can then be approximated by: C C = Loading_ solution ( 4 ) Silicone K Loading_/ solution Silicone

−1 where C Loading_solution (μg L ) is the nominal concentration of HOC in the loading solution. The above approach can be incorporated into a simple proto-

col for passive dosing at deﬁ ned CFree levels from aqueous solubility and below: 1. Make up a saturated loading solution. 2. Remove excess compound by ﬁ ltration, centrifugation, etc. 3. Make stepwise dilutions of the saturated solution. 4. Use the saturated solution and dilutions thereof to load the silicone in negligible-depletion mode as described above. In this way, a series of silicone dosing phases are prepared with

C Silicone ranging from saturation down to known dilutions of this. These translate into passive dosing C Free values that decrease from aqueous solubility in known dilutions steps. When analytical con- ﬁ rmation is not possible, literature values of the test compound’s

aqueous solubility can be used to calculate the CFree dilutions in the test. An alternative approach to above is loading by depletive partitioning. A deﬁ ned volume of loading solution is added to the silicone, the HOC simply allowed to partition into the silicone and

ﬁ nally CSilicone measured for calculating C Free using (1 ). Analytical measurements are used to conﬁ rm the loading of the silicone, and thus it does not matter that the loading solution becomes depleted. Note that with this approach it is not possible to saturate the silicone for testing at aqueous solubility, since the partitioning into the silicone will reduce the loading solution concentrations to below saturation levels. Control of Hydrophobic Compound Exposure 69

Loading by “spiking” from a concentrated solution . Here, a deﬁ ned mass of test compound is quantitatively transferred into the silicone from a small volume of concentrated spike solution. This is useful when working with HOCs that are available in small

quantities or are expensive. With this approach, CSilicone is known and C Free then calculated using the appropriate partitioning ratio or measured. Two variations are possible: (1) partitioning from a water-miscible solvent [ 12 , 40 ] and (2) partitioning from a volatile solvent followed by its evaporation [ 25 , 41 ]. In the fi rst variation, a concentrated spiking solution of the test HOC is made up in a water-miscible solvent, and the appropriate volume (and thus the required mass of compound) added to the silicone. The HOC is allowed to partition into the silicone, and small volumes of water then added incrementally, allowing suffi cient time between each addition for a new partitioning equilibrium to be reached. By gradually increasing the water to spiking solvent ratio, the HOC increasingly partitions into the silicone such that it eventually is quantitatively transferred (e.g., [ 12 , 40 ]). The water increments should be suffi ciently small to avoid the formation of crystals in the spiking solvent:water mixture. In the second variation, a volatile solvent is used to make up the loading spike, and the compound then forced into the silicone by slow evaporation of the solvent (e.g., [ 25 , 41 ]). Removal of loading solution traces . After loading is completed, remaining solvent traces need to be completely removed prior to the genotoxicity test. When a water-miscible solvent is used for the loading solution, this can be easily achieved by repeated washing of the silicone with small volumes of water (e.g., [ 9 , 21 , 22 ]). The HOCs preferentially remain in the silicone, while the solvent is quantitatively removed with the water washes. In the case of spiking with a volatile solvent, the evaporation step should be optimized to completely remove the solvent while minimizing losses of the target compound [25 , 41 ].

2.5 Step 5: Exposure At experiment completion, exposure can be conﬁ rmed to show Conﬁ rmation that the passive dosing silicone has been (1) loaded to the correct level and (2) negligibly depleted. For example, the silicone concentrations can be measured by simple solvent extraction and analysis, and compared to the initial values [ 9 ]. In this case the analysis is straightforward since the HOC concentration in the silicone is relatively high. Alternatively, the silicone can be equilibrated with a small volume of pure water at the test temperature

[ 21 , 22 ]. Measuring these equilibrium C Free values provides a direct measure of the dissolved exposure concentrations in the experimental setup. However, for some HOCs this approach is limited by their low water concentrations precluding their direct analysis. 70 Kilian E.C. Smith

3 Future Outlook of Passive Dosing for In Vitro Genotoxicity Testing

As testing efforts move on from the simple conﬁ rmation of genotoxic activity to better understanding of the underlying exposure-

response relationships, a better control of the C Free levels throughout the in vitro assay will become increasingly important. Techniques to achieve this need to be practical to implement, reliable, and easily interfaced with the existing test protocols for in vitro genotoxicity assays. One promising technique is passive dosing, which

has already shown its utility for the a priori control of CFree in various in vitro test systems. Although so far mainly used for single HOCs, the properties of passive dosing also make it useful for the genotoxicity testing of HOC mixtures since it can maintain constant mixture levels and profi les in the test. Despite the progress, a number of challenges still remain for passive dosing. The approach could be further simplifi ed, particularly with regard to the loading of the silicone. This will facilitate its application in high-throughput systems. In this context, smaller formats will also lead to reduced consumption of test substance. The exchange kinetics between the passive dosing and aqueous phases determine the speed of equilibration, and thus how effectively losses are compensated. The fastest passive dosing formats have equilibration times in the order of tens of minutes (e.g., [12 ]), which is suffi cient for longer in vitro genotoxicity assays, as well as those where a pre-equilibration step can be incorporated. However, faster passive dosing formats would still be benefi cial for assays with very short response times. The application domain of silicone as a

passive dosing polymer covers HOCs with log K Silicone/Free values greater than around 1,000. Such HOCs are particularly challenging with regard to the proper control of their exposure in in vitro assays. Nevertheless, another area for further developing passive dosing will be extending its application domain to include more polar and volatile organic compounds. Many of these face similar challenges to HOCs when it comes to test losses, albeit via different mechanisms. This will require ﬁ nding passive dosing polymers with higher afﬁ nities for these compounds and which still exhibit the desirable characteristics of inertness, linear partitioning, and low internal mass transfer resistance.

References

1. OECD (1997) Test no. 471: bacterial reverse section 4. OECD Publishing, Paris (doi: mutation test, OECD guidelines for the testing 10.1787/9789264071261-en ) of chemicals, section 4. OECD Publishing, 3. OECD (1997) Test no. 476: in vitro mam- Paris (doi: 10.1787/9789264071247-en ) malian cell gene mutation test, OECD 2. OECD (1997) Test no. 473: in vitro mam- guidelines for the testing of chemicals, malian chromosome aberration test, OECD section 4. OECD Publishing, Paris (doi: guidelines for the testing of chemicals, 10.1787/9789264071322-en ) Control of Hydrophobic Compound Exposure 71

4. OECD (2010) Test no. 487: in vitro mam- 16. Lammer E, Kamp HG, Hisgen V et al (2009) malian cell micronucleus test, OECD guide- Development of a fl ow-through system for the lines for the testing of chemicals, section 4. fi sh embryo toxicity test (FET) with the zebrafi sh OECD Publishing, Paris (doi: 10.1787/ ( Danio rerio ). Toxicol In Vitro 23(7): 9789264091016-en ) 1436–1442 5. Schreiber R, Altenburger R, Paschke A et al 17. Mayer P, Wernsing J, Tolls J et al (1999) (2008) How to deal with lipophilic and volatile Establishing and controlling dissolved concen- organic substances in microtiter plate assays. trations of hydrophobic organics by partition- Environ Toxicol Chem 27(8):1676–1682 ing from a solid phase. Environ Sci Technol 6. Tanneberger K, Rico-Rico A, Kramer NI et al 33(13):2284–2290 (2010) Effects of solvents and dosing pro- 18. Brown RS, Akhtar P, Akerman J et al (2001) cedure on chemical toxicity in cell-based Partition controlled delivery of hydrophobic in vitro assays. Environ Sci Technol 44(12): substances in toxicity tests using poly(dimeth- 4775–4781 ylsiloxane) (PDMS) fi lms. Environ Sci Technol 7. Riedl J, Altenburger R (2007) Physicochemical 35(20):4097–4102 substance properties as indicators for unreliable 19. Escher BI, Hermens JLM (2004) Internal exposure in microplate-based bioassays. exposure: linking bioavailability to effects. Chemosphere 67(11):2210–2220 Environ Sci Technol 38(23):455A–462A 8. Heringa MB, Schreurs R, Busser F et al (2004) 20. Groothius FA, Heringa MB, Nicol B et al Toward more useful in vitro toxicity data with (2013) Dose metric considerations in in vitro measured free concentrations. Environ Sci assays to improve quantitative in vitro – in vivo Technol 38(23):6263–6270 dose extrapolations. Toxicology. doi: 9. Smith KEC, Oostingh GJ, Mayer P (2010) 10.1016/j.tox.2013.08.012 Passive dosing for producing defi ned and 21. Mayer P, Holmstrup M (2008) Passive dosing constant exposure of hydrophobic organic of soil invertebrates with polycyclic aromatic compounds during in vitro toxicity tests. Chem hydrocarbons: limited chemical activity explains Res Toxicol 23(1):55–65 toxicity cutoff. Environ Sci Technol 10. ECETOC (1996) Monograph no. 26. Aquatic 42(19):7516–7521 toxicity testing of sparingly soluble, volatile 22. Smith KEC, Dom N, Blust R et al (2010) and unstable substances. European Centre for Controlling and maintaining exposure of Ecotoxicology and Toxicology of Chemicals hydrophobic organic compounds in aquatic (ECETOC), Brussels toxicity tests by passive dosing. Aquat Toxicol 11. OECD (2000) Guidance document on aquatic 98(1):15–24 toxicity testing of diffi cult substances and 23. Smith KEC, Heringa MB, Uytewaal M et al mixtures; ENV/JM/MONO(2000)6. Organi- (2013) The dosing determines mutagenicity sation for Economic Co-operation and Develo- of hydrophobic compounds in the Ames II pment (OECD), Paris assay with metabolic transformation: passive 12. Smith KEC, Rein A, Trapp S et al (2012) dosing versus solvent spiking. Mutat Res 750 Dynamic passive dosing for studying the (1–2):12–18 biotransformation of hydrophobic organic 24. Kramer NI, Busser FJM, Oosterwijk MTT et al chemicals: microbial degradation as an exam- (2010) Development of a partition-controlled ple. Environ Sci Technol 46(9):4852–4860 dosing system for cell assays. Chem Res Toxicol 13. Gulden M, Seibert H (2003) In vitro - in vivo 23(11):1806–1814 extrapolation: estimation of human serum 25. Booij P, Lamoree MH, Leonards PE et al concentrations of chemicals equivalent to cyto- (2011) Development of a polydimethylsilox- toxic concentrations in vitro . Toxicology ane fi lm-based passive dosing method in the 189(3):211–222 in vitro DR-Calux® assay. Environ Toxicol 14. Gulden M, Seibert H (2005) Impact of bio- Chem 30(4):898–904 availability on the correlation between in vitro 26. Bougeard C, Gallampois C, Brack W et al cytotoxic and in vivo acute fi sh toxic concentra- (2011) Passive dosing: an approach to control tions of chemicals. Aquat Toxicol 72(4): mutagen exposure in the Ames fl uctuation test. 327–337 Chemosphere 83(4):409–414 15. Kramer NI, Hermens JLM, Schirmer K et al 27. Mayer P, Vaes WHJ, Hermens JLM et al (2009) The infl uence of modes of action (2000) Absorption of hydrophobic compounds and physicochemical properties of chemicals into the poly(dimethylsiloxane) coating of on the correlation between in vitro and acute solid-phase microextraction fi bers: high parti- fi sh toxicity data. Toxicol In Vitro 23(7): tion coeffi cients and fl uorescence microscopy 1372–1379 images. Anal Chem 72(3):459–464 72 Kilian E.C. Smith

28. Poerschmann J, Gorecki T, Kopinke FD et al passive sampling devices. Environ Sci Technol (2000) Sorption of very hydrophobic organic 45(23):10124–10132 compounds onto poly(dimethylsiloxane) and 35. Difi lippo EL, Eganhouse ER (2010) dissolved humic organic matter. 1. Adsorption Assessment of PDMS-water partition coeffi - or partitioning of VHOC on PDMS-coated cients: implications for passive environmental solid-phase microextraction fi bers—a never- sampling of hydrophobic organic compounds. ending story? Environ Sci Technol 34(17): Environ Sci Technol 44:6917–6925 3824–3830 36. Poole SK, Poole CF (2001) Chromatographic 29. Seethapathy S, Gorecki T (2012) Applications models for the sorption of neutral organic of polydimethylsiloxane in analytical chemistry: compounds by soil from water and air. J a review. Anal Chim Acta 750:48–62 Chromatogr A 845:381–400 30. Rusina TP, Smedes F, Klanova J et al (2007) 37. Schwarzenbach RP, Gschwend PM, Imboden Polymer selection for passive sampling: a com- DM et al (2003) Environmental organic chem- parison of critical properties. Chemosphere istry. Wiley, New York 68(7):1344–1351 38. Gouilarmou V, Smith KEC, de Jonge LW et al 31. Kwon JH, Wuethrich T, Mayer P et al (2007) (2012) Measuring binding and speciation Dynamic permeation method to determine of hydrophobic organic chemicals at controlled partition coeffi cients of highly hydrophobic freely dissolved concentrations and without chemicals between poly(dimethylsiloxane) and phase separation. Anal Chem 84:1601–1608 water. Anal Chem 79(17):6816–6822 39. Kwon HC, Kwon JH (2012) Measuring aque- 32. ter Laak TL, Busser FJM, Hermens JL et al ous solubility in the presence of small cosolvent (2008) Poly(dimethylsiloxane) as passive sam- volume fractions by passive dosing. Environ Sci pler material for hydrophobic chemicals: effect Technol 46(22):12550–12556 of chemical properties and sampler characteris- 40. Birch H, Gouliarmou V, Lützhøft HCH et al tics on partitioning and equilibration times. (2010) Passive dosing to determine the Anal Chem 80(10):3859–3866 speciation of hydrophobic organic chemicals 33. Lohmann R (2012) Critical review of low- in aqueous samples. Anal Chem 82(3): density polyethylene’s partitioning and diffu- 1142–1146 sion coeffi cients for trace organic contaminants 41. Bandow N, Altenburger R, Lübcke-Von Varel and implications for its use as a passive sampler. U et al (2009) Partitioning-based dosing: Environ Sci Technol 46(2):606–618 an approach to include bioavailability in the 34. Endo S, Hale SE et al (2011) Equilibrium par- effect-directed analysis of contaminated sedi- tition coeffi cients of diverse polar and nonpolar ment samples. Environ Sci Technol 43(10): organic compounds to polyoxymethylene (POM) 3891–3896 Chapter 5

The In Vitro Micronucleus Assay and FISH Analysis

Lucia Migliore , Sebastiano Di Bucchianico , and Chiara Uboldi

Abstract

The cytokinesis-block micronucleus cytome (CBMN-cyt) assay was originally established as an ideal system for evaluating chromosomal damage in terms of micronuclei formation. Throughout the years, the micronucleus assay evolved in a comprehensive system for assessing cytogenetic damage, cytostasis and cytotoxicity. The CBMN-cyt assay in peripheral blood lymphocytes and in other cultured mammalian cells is the most common approach to evaluate chromosomal damage induced by environmental agents, including emerging compounds as nanomaterials, and it is the most frequent test system in biomonitoring human populations. When coupled with ﬂ uorescence in situ hybridization (FISH), CBMN-cyt assay is able to reveal the capability to induce structural chromosome aberrations (clastogenic activity) and/or numerical chromosome changes (aneuploidogenic activity). The methods for CBMN-cyt assay and FISH described here refer to the use of separate lymphocytes and whole blood cultures involving the block of cytokinesis with cytochalasin B (cyt-B) but other cell systems of different origin can be successfully used. This chapter describes in details well-established protocols for sample processing, slide preparation and scoring criteria.

Key words Micronuclei , Genotoxicity , Aneuploidy , Clastogenicity , Nanomaterials , DNA damage , Lymphocytes , Whole blood , Cytome assay

1 Introduction

If we search in the electronic PubMed database using “in vitro micronucleus assay” as key words, we fi nd about 2,000 scientifi c papers, starting since the mid-1970s of the last century. Indeed it is used to date back the fi rst observations on those chromatin particles that we now call micronuclei (MN), visible in the cytoplasm of cells of many different species so far investigated, ranging from plants to fi shes, to mammalians, humans included, to some papers published in the middle of the last century. Among them there is that of Hutchinson and Ferguson-Smith who observed the presence of nuclear fragments in red cell precursors, commonly known as Howell–Jolly bodies, in haematological disease, usually associated with clear evidence of vitamin B12 or folic acid defi ciency [ 1 ].

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_5, © Springer Science+Business Media New York 2014 73 74 Lucia Migliore et al.

The fi rst application of the scoring of micronuclei (MN) as an index of chromosomal damage to evaluate induced effects of in vitro treatments in human lymphocytes is offi cially recognized to Countryman and Heddle, who studied the production of MN by in vitro treatments of ionizing radiations [2 ]. The test was proposed as a rapid, simple and not expensive assay for the assessment of chromosomal damage induced in vitro by environmental agents. Then Michael Fenech, under the supervision of Morley [3 ], developed the fi rst version of the cytokinesis-block micronucleus (CBMN) assay by using the cytochalasin-B (cyt-B) to block cell cytokinesis, obtaining as binucleated cells lymphocytes undergone to one cell division. Subsequently M. Fenech evolved and refi ned the assay into a comprehensive CBMN cytome assay (CBMN-cyt) of DNA damage, cell death and cytostasis [ 4 ], which is the topic of this chapter. The main milestones in the evolution of the assay have been linked to the possibility: (a) to score cytoplasmic chromatin bodies in interphase cells, in alternative to the classical chromosomal aberration analysis, that is time-consuming and requiring rather expert operators; (b) to identify cells surely undergone a cell division in a lymphocyte population, not always synchronous; only those cells are indeed able to show micronuclei as the result of chromatin fragmentation or mis-segregation events; (c) to distinguish between agents able to induce mainly clastogenic events from those mainly aneuploidogenic; (d) to identify, at the same time, several induced cytotoxic/genotoxic effects, as in the last version of the assay, the CBMN cytome assay. The four milestones are epitomized in the present version of the CBMN cytome assay which is a comprehensive system for measuring DNA damage, cytostasis and cytotoxicity. DNA damage events are scored specifi cally in once-divided binucleated cells and include (1) micronuclei (MN), a biomarker of chromosome breakage and/ or whole chromosome loss, (2) nucleoplasmic bridges (NPB), a biomarker of DNA misrepair and/or telomere end-fusion and (3) nuclear buds (NBUD), a biomarker of elimination of amplifi ed DNA and/or DNA repair complexes (Fig. 1 ). Cytostatic effects are measured via the proportion of mono-, bi- and multinucleated cells and cytotoxicity via necrotic and/or apoptotic cell ratios. To determine whether the micronuclei formed are the result of a clastogenic (chromosome fragmentation) or aneuploidogenic (whole chromosome loss) mode of action, the micronucleus assay is coupled to fl uorescence in situ hybridization (FISH) with pancentromeric probes (Fig. 2 ). This combination allows to discriminate between micronuclei containing a whole chromosome (centromere positive micronucleus) and an acentric chromosome fragment (centromere negative micronucleus). Over the nearly 30 years after its proposal, the in vitro CBMN assay has achieved considerable popularity, because it is suitable to CBMN-cyt Assay and FISH 75

Fig. 1 Schematic representation of the possible scenarios following cytotoxic and/or genotoxic events. The CBMN-cyt assay offers information on the cytotoxic potential by the evaluation of the apoptotic and necrotic indices; simultaneously, the genotoxicity is typifi ed by MN (chromosome loss or breakage), NPB (dicentric chromosome) and NBUD (DNA amplifi cation) (Colour fi gure online)

be used both for basic studies, to explore the mechanisms of action of various physical, mainly ionizing radiation, and chemical agents, but also for screening programs involving many sort of substances such as drugs, pesticides, dyes and environmental contaminants. The main reasons are due to its ease of implementation, applicability to different cell types, but also to the constant efforts of many researchers to improve its performance. An important contribution to the diffusion and improvement of the assay was given in 1997 with the HUMN (HUman MicroNucleus) project [ 5 ], run mainly by Michael Fenech and Stefano Bonassi. It was aimed at collecting data from laboratories worldwide for a comparison of the human lymphocyte culture methods used to deﬁ ne the main variables affecting lymphocyte MN basal frequency; moreover, by evaluating the performance of an inter- and intra-laboratory slide scoring exercise, criteria of 76 Lucia Migliore et al.

Fig. 2 FISH analysis with pancentromeric probes (in yellow ) allows to distinguish between MN originating from chromosome loss (MN C+) and those generating from chromosome breakage (MN C−) in binucleated cells obtained following cytochalasin-B treatment (Colour ﬁ gure online)

scoring objective and shared were established for the assay [ 4 – 10 ]. Finally, a prospective study to test the hypothesis that MN frequency in lymphocytes can predict cancer risk was successfully carried out [ 11 ]. Another contribution that allowed to consider the MN assay a valid alternative to the chromosome aberration test, even for regulatory purposes, was developed in 2008 at the European Centre for the Validation of Alternative Methods (ECVAM). A retrospective validation of existing data was performed, to evaluate the validity of the in vitro MN test on the basis of the modular validation approach. The ECVAM Validation Management Team concluded that the in vitro MN test is reliable and relevant and can therefore be used as an alternative method to the in vitro chromosome aberration test [ 12 ]. Following peer review, these conclusions were formally endorsed by the ECVAM Scientiﬁ c Advisory Committee. It allowed to introduce the MN test in OECD (Organisation for Economic Co-operation and Development) Guidelines [ 13 ] that include the most relevant internationally agreed testing methods used by government, industry and independent laboratories. CBMN-cyt Assay and FISH 77

The OECD Guidelines are used primarily in regulatory safety testing and subsequent chemical and chemical product notiﬁ cation and registration; they can also be used for the selection and ranking of candidate chemicals during the development of new chemicals and products, and in toxicology research.

1.1 Cytological The molecular mechanisms that induce the formation of micronu- and Molecular Basis clei and of other nuclear anomalies, such as NPB and NBUD, are multiple [ 14 ]. Displaced acentric chromosomes or fragments that do not attach to the mitotic spindle during anaphase and that, consequently, are erroneously excluded from the nuclei of the daughter cells are believed to generate tiny extranuclear bodies called MN [ 4 , 15 ]. The presence of acentric chromosome fragments can depend on misrepair of DNA double strand breaks, on the excision repair of damaged or inappropriate DNA bases or on fragmented chromosomes that originated during formation, stretching and breakage of NPB [16 ]. MN can also originate from whole chromosome mis-segregation in anaphase or from chromosome loss, and it can depend on a defective assembly of the mitotic spindle and of mitosis checkpoint proteins [ 17 – 20 ]. MN formation via chromosome mis-segregation can be caused also by the hypomethylation of cytosine residues in the satellite DNA of centromeric regions [21 , 22 ]. Under normal conditions, the cytosine residues of the satellite DNA are methylated and contribute to the condensation of the heterochromatin structure of chromosomes 1, 9 and 16. When cytosines in the centromeric DNA are demethyl- ated, in contrast, the pericentromeric chromatin of the chromosomes 1, 9 and 16 is elongated, the condensed structure is lost and these chromosomes result mis-segregated causing MN formation [23 – 26 ]. This mis-segregation or loss of chromosomes can also be ascribed to a defective kinetochore assembly [ 27 ]. The correct assembly of the kinetocore proteins CENP-A and CENP-B, in fact, depends on the methylation of cytosines in the centrosomes and on the correct methylation of histones [27 , 28 ]. NPB represent dicentric chromosome bridges and can be used to evaluate chromosomal rearrangements [29 ]. Moreover, compared to the sole counting of MN, NPB scoring provides a direct indication of genome damage resulting either from misrepaired DNA breaks either from telomere end-fusions [4 ]. NPB are usually formed during mitosis, at anaphase, when the centromers of dicentric chromosomes are moved to the opposite poles of the cell/mitotic spindles: if the anaphase bridge is not broken during cytokinesis, then the nuclear membrane encircles the daughter nucleus and the newly formed NPB. For this reason, when performing CBMN-cyt assay, that requires the use of cyt-B to block cytokinesis, NPB are present in binucleated cells offering therefore an indication of chromosome rearrangements [4 , 30 ]. The dicentric chromosomes from which NPB generate can form in different 78 Lucia Migliore et al.

ways: misrepair of chromosome breaks or of telomere end-fusion caused by telomere shortening, loss of telomere capping proteins or defects in telomere cohesion [14 , 31 –34 ]. Another cytogenetic anomaly that can be scored via CBMN- cyt assay is the nuclear budding. NBUD are biomarkers of the elimination of ampliﬁ ed DNA extruded from the nuclei [ 35 – 37 ], are morphologically similar to MN and are identiﬁ ed by the presence of a thin tail of nucleoplasmic material connecting the NBUD with the nucleus [ 38 , 39 ]. NBUD can, in particular, be formed following exposure to gamma-irradiation that induces the formation and then the extrusion of Rad51-recombination proteins from the nuclei [40 ].

1.2 Post-mitotic The MN formed during mitosis have been studied to clarify their Fate of Micronuclei post-mitotic fate and understand what happens to them. Although the MN fate has been largely investigated [41 – 45 ], it is still unclear; it seems to depend on different mechanisms driving the MN to:

● Re -incorporation in the main nucleus . When this happens, the re-incorporated genomic material results indistinguishable from the main nucleus and resumes its normal biological activity. ● Expulsion from the cell . When a cell recognizes that the DNA contained in the MN is not functional or it cannot be replicated due to the absence of the necessary replication machinery, MN is physically extruded from the cell. ● Retention in the cytoplasm and DNA / chromosome replication . If the MN remains in the cytoplasm, it can replicate (once or more cycles) at the same time of the main nucleus. As consequence, multiple copies of the chromatine mass contained in the MN can be found. For example, in human lymphocytes the replication of the genetic material contained in MN induced the formation of multiple copies (up to four to ten copies) of the sexual X-chromosome, as shown by FISH [41 ]. As reviewed by Terradas and colleagues, the transcriptional activity of the MN depends on its genetic content and on the functionality of the nuclear membrane, the nuclear pores and the nuclear lamina [44 ]. In fact it has been demonstrated that MN derived from chromosomal fragmentation contain a combination of transcriptionally active and inactive forms, being the inactive forms much more frequent than the active ones. This lack of replication seems to depend on the low density of the nuclear pore complexes on the MN envelop [16 ]. ● Elimination by apoptosis . MN can be erased from cells by apoptosis of the MN directly, or the BNMN cell itself undergoes programmed cell death. While apoptosis of MN depends on the presence or on the absence of DNA repair and DNA CBMN-cyt Assay and FISH 79

replication checkpoint proteins within the MN, micronucleated cells seem to represent a signal triggering apoptosis. This was proven by supplementing the micronucleated cells with inhibitors of caspase-8 and caspase-9: the presence of the inhibitors increased the number of cells with MN, conﬁ rming that micronucleated cells can undergo apoptosis [46 ]. Moreover, in HeLa cells an altered gene expression in MN was associated to the ability of micronucleated cells to die via programmed cell death [37 ], and MN-bearing cells were shown to suffer from apoptosis more frequently than cells without MN [47 ].

1.3 Cell Lines: The CBMN-cyt assay can be successfully performed in primary Methodological cells and cell lines of different origin (human and rodent), in the Aspects presence or in the absence of the actin polymerization inhibitor cyt-B. Nevertheless, several methodological aspects must be taken into consideration to obtain robust, reproducible and reliable results, as stated in the OECD test guideline 487 [13 ]. The criteria for the selection of cell lines for the MN test require that the chosen cell line has a low and stable background frequency of MN. To date, the MN assay has been validated for some cell lines such as the rodent-derived ones V79 [47 ], CHL/ IU [48 ], CHO and L5178Y [49 ], but also other cell lines like TK6 [ 50 ], HepG2 [51 ], A549 [52 ] and Syrian Hamster Embryo (SHE) cells [53 ] are reported in the literature. Of fundamental importance when using established or primary cells is the knowledge on the duration of their cycle: this parameter drives the experimental design in terms of culture and treatment duration. CBMN-cyt assay, in fact, requires that the cells have divided at least once during the treatment or in the post-treatment incubation period, and therefore the culture duration should vary depending on the doubling time of each speciﬁ c cell line. Additionally, the treatment period is crucial, since it should last long enough to cover all the stages of the cell cycle and ensure a proper interaction of the cells with the test compound [13 ]. Moreover, when performing the CBMN-cyt assay with cell lines, it is important to remind that freshly cultured cells (early passages after thawing) normally display a lower MN frequency than the same cells kept in culture for several passages. Another important methodological aspect for the correct running of the CBMN-cyt assay is the use of cyt-B, which affects the experimental design. A requirement of the CBMN-cyt assay is that the cells being scored had completed mitosis during the treatment or the post-treatment incubation period. Therefore, when CBMN- cyt is performed in the absence of cyt-B, to prevent the formation of false positives it is necessary to demonstrate that the cells scored in the culture have divided during or following treatment with the test substance. Bromodeoxyuridine (BrdU), proliferation index (PI), relative increase in cell count (RICC) or cytotoxicity/cytostasis 80 Lucia Migliore et al.

markers (conﬂ uency, cell number, apoptosis, necrosis, metaphase counting) can be used to ensure that cells underwent mitotic division. In contrast, when CBMN-cyt assay is performed in the presence of cyt-B, cyt-B itself ensures the inhibition of the assembly of the actin ﬁ laments, preventing thus the separation of the daughter cells after mitosis and allowing the formation of BN cells.

1.4 Application A complete evaluation of the genotoxic potential of chemicals of the FISH Analysis requires the assessment of the capability to induce structural to the MN Assay chromosome aberrations (clastogenic activity) and/or numerical chromosome changes due to chromosome non-disjunction or to chromosome loss (aneuploidogenic activity). Aneuploidy is a major cause of human reproductive failure and an important contributor to cancer, and it is therefore important that any increase in its frequency due to chemical exposures should be recognized and controlled [54 ]. The rapid development of repetitive chromosome specifi c probes for centromeric, pericentromeric or telomeric regions of different mammalian chromosomes provided the opportunity to study the basic mechanisms underlying the origin of micronuclei. Several approaches have sought to improve the study of aneuploidogenic phenomena either using a classifi cation of the micronuclei diameter [55 ] either displaying centromeres in micronuclei by using anti-kinetochore antibodies from serum of CREST- patients (Calcinosis, Raynaud phenomenon, Oesophageal Dysmotility, Sclerodactyly, Telangiectasia) [56 ]. However, the fi rst approach did not lead to reliable results (because sometimes MN with high diameter can contain more chromosome fragments, instead of a whole chromosome) and the latter suffers from the fact that the kinetochore associate proteins and not the centromeric DNA are visualized, making undetectable the loss of chromosomes due to the inactivation or inhibition of kinetochore proteins formation. FISH technique was then successfully combined with CBMN- cyt assay to evaluate simultaneously the overall level of chromosome damage and to localize a defi nite genome domain within an individual cell. In the past decades, the CBMN-cyt assay has been widely applied for the detection of aneuploidogenic compounds in vitro [ 57 – 59 ]. Recently, the OECD adopted a test guideline (TG487) for conducting the micronucleus assay recognizing that hybridization with centromeric and/or telomeric probes can provide information on the mechanisms of chromosome damage and micronucleus formation [ 13 ]. By coupling the FISH technique with an alphoid DNA probe specifi c for centromere of all chromosomes, it is possible to discriminate MN generated by acentric fragments (clastogenic damage) from those containing whole chromosomes (mis-segregation CBMN-cyt Assay and FISH 81

Fig. 3 Micrograph of FISH analysis with pancentromeric probes applied to CBMN-cyt assay. Centromere negative micronuclei (a ) are easily distinguishable from centromere positive micronuclei (b ) by the absence of fl uorescent spots (in red ). The insert shows the specifi city of the centromeric hybridization on human chromosomes (Colour fi gure online)

events, i.e. chromosome loss) (Fig. 2 ). The use of centromeric probes for specifi c chromosomes can be useful in special cases. For example, epigenetic drugs that induce DNA hypomethylation, such as 5-azacytidine, induce a specifi c increase in the frequency of MN containing whole chromosomes 1, 9, 15, 16 and Y [ 60 ] associated with the induction of a signifi cant reorganization of constitutive pericentromeric heterochromatin, chromosomal rearrangements and centromere under-condensation [ 61 ]. The use of telomeric probes can solve and differentiate the mechanism of NPB formation: when an NPB originates from a telomere end-fusion, the telomere signal will be present within the bridge, while when an NPB originates from a DNA misrepair event, the telomeric signal will attend in MN arising from acentric fragments. The labeling and hybridization methods can be used when there is an increase in MN and/or NPB formation. Aneuploidy can result not just from mis-segregation of chromosomes, but also from spindle disruption, microtubule dissociation, centriole inactivation and epigenetic modifi cation; therefore, to observe aneuploidy the test substance should be present during all phases of mitosis. Separate slides from CBMN cytome assay can be used to achieve FISH with centromeric and/or telomeric probes. Slides of treatments, positive and negative control can be examined for the presence of centromeric spots in MN (Fig. 3 ) and classifi ed as centromere positive (MN C+) and centromere negative (MN C−).

1.5 Testing The steadily development, production and release of engineered Nanomaterials nanomaterials (NM) and NM-based products in many everyday with MN Assay life applications (e.g. cosmetics, paints, plastics, papers, food and feed additives) poses questions on their effects to the human health. Due to their extremely small size (at least one dimension 82 Lucia Migliore et al.

smaller than 100 nm), NM are characterized by an elevated surface- area-to-volume-ratio which makes them highly bioreactive, and their bioreactivity can lead to toxicity due to harmful interactions of NM with biological systems [ 62 ]. The exposure of biological systems to NM can induce either primary genotoxicity, if the contact occurs directly, or secondary genotoxicity, with the mediation of proinfl ammatory cytokines and oxidative factors. To investigate the genotoxic potential of NM, an intelligent testing approach should take advantage, for an easier comparison of the results, of the use of validated assays. Mechanistic considerations and existing data indicate how, compared to other genotoxicity validated tests, the CBMN-cyt assay is suitable for the study of the genotoxicity exerted by NM [ 63 ]. The in vitro CBMN-cyt assay, in fact, allows to differentiate between clastogenic chromosomal damages mainly induced by reactive oxidative species, and aneuploidogenic effects caused by the alteration of the mitotic spindles [ 64 ] or, when coupled with FISH analysis, by chromosome loss [ 65 ]. In addition, CBMN-cyt assay detects apoptotic and necrotic fi gures, chromosomal breakage and loss or nondisjunction via the evaluation of MN, NPB and NBUD. Nevertheless, as for the other genotoxicity tests, CBMN-cyt assay has been developed to investigate conventional chemical products and its experimental design might not be directly applicable to NM. In fact, testing NM genotoxicity poses several issues, and data gaps and confounding factors have been identifi ed. The lack of a positive particulate control is, for example, an important issue questioning the suitability of CBMN-cyt assay in testing the genotoxicity of NM [ 66 ]. Although CBMN-cyt assay compares the results of the test NM with its respective negative or solvent/ vehicle control, the lack of a positive particulate control affects the appropriateness of the experimental result [ 67 ]. Additionally, the interaction of NM with the cell culture medium and with serum proteins are very important factors to take into account in order to improve the quality of the CBMN-cyt assay and to obtain reliable and reproducible data [ 68 , 69 ]. Moreover, the use of cyt-B is a crucial step when the genotoxicity of NM is investigated by performing in vitro CBMN-cyt assay. In fact, as described by Doak and colleagues cyt-B, which is used to block cytokinesis and generate binucleated cells, might inhibit the endocytotic machinery and interfere with the internalization of NM by the cells in culture [ 68 ]. This is particularly evident in the case of a co-treatment: when NM and cyt-B are simultaneously incubated with cultured cells, the uptake of NM into the cells might be inhibited by the presence of cyt-B, leading to an underestimation of the genotoxic damage. To overcome this problem, adaptations on the CBMN-cyt assay protocol were proposed [ 70 ]. In the case of cyt-B post- treatment, cells are fi rst incubated in the presence of NM, CBMN-cyt Assay and FISH 83

then NM are removed and culture medium is refreshed before cyt-B is supplemented. This option has been shown inducing a signifi cant dose-dependent increased MN formation in MCL-5 cells exposed to dextran-coated ultrafi ne superparamagnetic iron oxide nanoparticles [ 68 ]. Alternatively, to ensure that cells have suffi cient time to internalize the test compound and that they have already completed one cell cycle, a delayed co-treatment can be performed and cyt-B is added 20 h after the beginning of the exposure (44 h after cell seeding). A delayed co-treatment, in addition, guarantees that cells are exposed to NM during mitosis, which is the step at which NM can directly enter in contact with DNA, as the integrity of the nuclear envelop is lost and the genetic material is most accessible [ 69 –71 ]. The delayed co-treatment seems to be the most effi cient because it restricts the scoring of MN to binucleated cells and prevents from false results caused by suboptimal or altered cell division kinetics [ 4 ]. In fact, if cells are exposed to the sole treatment without the addition of cyt-B, an underestimation of MN formation can occur due to the absence of cyt-B that inhibits the nuclear division, as demonstrated in experiments with human lymphocytes [72 ]. In conclusion, CBMN-cyt assay is suitable for testing the genotoxicity of NM and it has the advantage of offering not just information on chromosomal damage, but also on cellular proliferation and cytostasis. Nevertheless, the presence of NM requires some precautions and adjustments to obtain more reliable and reproducible data. Additionally, the presence/absence of serum proteins and the composition of the culture medium play a pivotal role in the assessment of the genotoxic potential of NM. It has been shown, in fact, that the absence of serum increases the MN formation [ 73 , 74 ], and the cell culture media composition impacts signifi cantly on the induction of chromosomal aberrations [ 75 ]. Moreover, when the solubility of the NM is not optimal, the highest concentration tested should be the one causing minimal precipitate in cultures and not interfering with scoring. The evaluation of NM physico-chemical properties should be performed before performing the CBMN-cyt assay and NM that precipitate or that change the pH of the culture medium should not be tested.

2 Materials

2.1 Materials – Sodium heparin Vacutainer or a syringe with 25 U/mL preservative-free sodium heparin – Class-II biosafety cabinet – Test tubes for culturing cells – Conical 15 mL and 50 mL polypropylene test tubes 84 Lucia Migliore et al.

– Microscope slides (76 × 26 mm and 1 mm thick) wiped with alcohol, rinsed with distilled water and allowed to dry. The slides are stored at −20 °C before use – Glass coverslips 22 × 22 mm and 22 × 50 mm – Plastic dropper – Pasteur pipette – Fixogum or Rubber Cement – Eukitt or DePex mounting medium – Hot plate – Fluorescence microscope – Cytocentrifuge – 6-well/plates

2.2 Reagents – Dimethylsulfoxide (DMSO) for CBMN-cyt Assay – 0.9 % isotonic saline: 0.9 g NaCl in 100 mL water, sterile – RPMI-1640 culture medium with 2 mM glutamine and 25 mM Hepes, sterile, liquid. Store at 4 °C. Use at 37 °C – Foetal bovine serum (FBS) heat-inactivated, sterile. Store frozen at −20 °C. Thaw in a 37 °C water bath before adding to the culture medium Note : Heat-inactivation (56 °C for 30 min) is only required if it is considered necessary to destroy heat-labile complement proteins that may cause cell lysis. However, pre-warming of FBS to 37 °C is enough to inactivate heat-labile complement taking into account that certain vitamins (e.g. folic acid), growth factors, amino acids, etc. may be diminished by the heat treatment causing effects on genome stability [ 4 ]. The CBMN-cyt assay for lymphocytes and whole blood can be performed using not inactivated FBS – Penicillin–streptomycin: 10,000 U/mL Penicillin, 10,000 μg/ mL Streptomycin. The aliquots may be stored at −20 °C for up to 6 months – Hank’s balanced salt solution (HBSS) – Ficoll-paque – Phytohemagglutinin (PHA) M form: 1 mg/mL, liquid. PHA should be aliquoted in sterile tubes into a volume appropriate for use. The aliquots may be stored at −20 °C for up to 6 months. If lyophilized powder form is available, reconstitute to obtain 1 mg/mL solution – Cytochalasin-B (cyt-B) solution: take the 10 mg vial of cyt-B from −20 °C and allow it to reach RT. Sterilize the top of the rubber seal with ethanol, but do not remove the seal. Vent the vial seal with a sterile needle and add 1.25 mL of sterile DMSO using a syringe; mix gently. Remove the 1.25 mL from the vial CBMN-cyt Assay and FISH 85

and eject into a sterile conical 50 mL polypropylene test tube. Add 15.4 mL of 0.9 % isotonic saline to reach a fi nal volume of 16.65 mL. This procedure gives a fi nal concentration of 600 μg/mL cyt-B solution. Mix and dispense adequate volumes into sterile cryogenic capped vials to make multiple aliquots and store at −20 °C for up to 12 months – Hypotonic solution (75 mM KCl): 2.8 g of KCl in 500 mL of distilled water. Do not store. Use only freshly prepared solution – Methanol – Glacial acetic acid – Prefi xing solution: 3:5 ratio of methanol: glacial acetic acid. The prefi xative should be freshly prepared each time. This solution should be made in a well-ventilated fume hood with appropriate safety precaution – Fixing solution: 6:1 ratio of methanol: glacial acetic acid. The fi xative should be freshly prepared each time and used at −20 °C. This procedure should be performed in a well- ventilated fume hood with appropriate safety precaution – 2 % (v/v) Giemsa solution: fi lter the Giemsa solution with 0.2 μm pore size fi lters. Protect from light. Prepare 100 mL of 2 % (v/v) Giemsa staining solution by adding, at RT, 96 mL of distilled water with 2 mL of Sorensen’s buffer and 2 mL of fi ltered Giemsa – Sorensen’s buffer (0.2 M): prepare stock solutions A and B and, prior to use them, combine the solutions to reach pH 6.8. Solution A: 0.2 M potassium dihydrogen phosphate (27.2 g

KH 2 PO4 per litre of water); Solution B: 0.2 M dibasic sodium phosphate (34.8 g Na 2 HPO4 · 2H2 0 per litre of water). To obtain 100 mL of Sorensen’s buffer solution (pH 6.8), mix

51 mL of KH2 PO4 (solution A) with 49 mL of Na2 HPO4 (solution B). Utilize the Sorensen’s buffer to prepare the Giemsa staining solution

2.3 Reagents – SSC (sodium chloride/sodium citrate), 20×: 175 g NaCl for FISH (3 M), 88 g trisodium citrate dihydrate Na 3 C6 H5 O 7 · 2H2 O (0.3 M), H 2 O to 800 mL, adjust pH to 7.0 with 1 M HCl, add H2 O to 1 L – Igepal – Ethanol – Centromeric probes – Phosphate-buffered saline (PBS), pH 7.3: 8.0 g NaCl, 0.2 g

KCl, 2.16 g Na 2 HPO4 · 7H2 O, 0.2 g KH 2 PO4 , H 2 O to 1 L. Filter, sterilize and store at 4 °C 86 Lucia Migliore et al.

– DAPI solution: dissolve 1 mg 4′,6-diamidino-2-phenylindole

in 10 mL H2 O to obtain a 0.3 mM DAPI solution. Aliquot into aluminium foil-wrapped tubes and store at −20 °C. Working solution : dilute the stock solution 1/1000 in PBS – Antifade mounting media: Vectashield, purchase from Vector Laboratories. Follow their instruction for use

3 Methods

3.1 In Vitro MN Test Human peripheral whole blood and isolated lymphocytes should be from 18 to 35 years old, healthy and non-smoking donors, not recently exposed to genotoxic chemicals or radiation. The selection of the donors must consider that MN frequency increases with age, and that this aspect is more pronounced in females than in males. Blood in sodium heparin can be held for ≤1 day at RT and still be cultured successfully, but cultures are best initiated as soon as possible. Cultured cells can be chosen among those validated [48 , 49 , 76 ] or any other cell line can be selected for in vitro MN assay. Nevertheless, before starting the assay with cell lines or with primary cells, any information on the length of the cell cycle is of fundamental importance. The experimental design and the treatment duration, in fact, depend on this information, as well as the timing for cyt-B addition. All the cell culturing and treatment must be performed in a Class II biosafety cabinet.

3.1.1 Primary Cell (a) Lymphocytes Cultures 1. Collect peripheral blood by venipuncture into a sodium heparin Vacutainer or a syringe with 25 U/mL preservative-free sodium heparin. 2. Dilute whole blood with an equal volume of HBSS at RT and gently invert to mix. In a ﬁ nal volume of 4 mL usually one can expect to collect up to 2 × 10 6 leukocytes. 3. Carefully layer the diluted blood sample on Ficoll-paque using a 1:3 ratio (e.g. 1 mL Ficoll-paque: 3 mL diluted blood). Note: It is important layering the sample : do not mix Ficoll- paque and the diluted blood sample. 4. Weigh and balance before spinning the tubes at 400 g for 30 min at 18–20 °C. 5. Draw off the upper layer using a clean Pasteur pipette, leaving the leukocytes layer undisturbed at the interface. The upper layer of plasma, which is essentially free of cells, may be stored at −20 °C for other purposes. CBMN-cyt Assay and FISH 87

6. Using a clean Pasteur pipette, transfer the leukocyte layer located at the interface of Ficoll-paque to a clean centrifuge tube taking care not to remove too much Ficoll-paque. 7. Add at least 3 volumes (6 mL) of HBSS to the leukocytes in the test tube at RT, mix gently and then centrifuge at 180 g for 10 min. 8. Remove the supernatant and resuspend the cell pellet in 2× volume removed of HBSS, then centrifuge at 100 g for 10 min at RT. 9. Discard the supernatant and resuspend the leukocytes in the medium appropriate (1–2 mL) using a Pasteur pipette. 10. Count the cells and resuspend 1 × 106 cells into 750 μL of complete culture medium in a 6 mL round-bottomed culture tube. Set up duplicate cultures per subject and/or treatment. Culture of lymphocytes 11. Stimulate mitotic division of lymphocytes by adding 10 μL of the PHA solution to 750 μL culture to give a ﬁ nal concentration of 30 μg/mL. 12. Incubate the cell cultures at 37 °C with lids loose in a

humidifi ed atmosphere containing 5 % CO2 for exactly 44 h. Addition of cyt-B 13. Thaw out stock vial containing 100 μL solution of cyt-B at 600 μg/mL and aseptically add 900 μL of culture medium equilibrated to RT to the vial to obtain a 1,000 μL solution of 60 μg/mL cyt-B. 14. Remove 56 μL of medium from the top of the 750 μL culture and replace with 56 μL of the 60 μg/mL cyt-B solution to give a fi nal concentration of 4.5 μg/mL. 15. Return cultures to the incubator and incubate for a further 28 h. Harvesting of cells, slide preparation, fi xation and staining 16. Harvest cells onto slides by a cytocentrifuge 72 h after stimulation following manufacturer’s instructions. Note: Prepare a concentration of cells that is suffi cient to produce a monolayer of cells on each spot. 17. Place the slides horizontally on a slide tray and allow the cells to air-dry for 10 min at RT. 18. Place slides vertically in a dry staining rack and place in methanol for 10–15 min. 19. Air-dry the slides overnight. Avoid drying them under laminar fl ow or fume hood. 88 Lucia Migliore et al.

20. Immerse the slides for 10 min at RT in a staining jar containing 2 % Giemsa solution. Note : Staining time is determined by trial and error. The aim is to obtain an optimal contrast between the nuclear and the cytoplasmic staining, so that the different biomarkers in the CBMN-cyt assay are easily and clearly scored. 21. Rinse the slides for 1 min in deionized water. 22. Leave the slides to air-dry completely for at least 1 h before putting coverslips on. 23. Place the slides on paper and set out coverslip alongside each. 24. Put two large drops of Eukitt or DePex (use a plastic dropper) on each coverslip. 25. Invert the slide over the coverslip and allow the Eukitt or DePex to spread. 26. Dry the slides under the fume hood. 27. Store the slides in slide boxes at RT. (b) Whole blood 1. Collect peripheral blood as indicated above. 2. Inoculate 0.3 mL of the whole blood into a sterile 15 mL centrifuge tube containing 4.7 mL complete medium RPMI-1640 supplemented with 10 % FBS and 1 % penicillin–streptomycin. Set up duplicate cultures per subject and/or treatment. 3. Add 75 μL of PHA solution to each test tube. T lymphocytes in whole blood are stimulated with the mitogenic PHA within 12–24 h and continue to proliferate for 2–4 days. 4. Incubate the test tubes at 37 °C with unsealed lids in a

humidiﬁ ed atmosphere containing 5 % CO2 . 5. At 44 h, add 50 μL of cyt-B solution to block the cytokinesis process. 6. Return cultures to the incubator. 7. At 72 h after starting the culture, harvest the cells. Harvesting of cells 8. Centrifuge the cell suspension for 10 min at 180 × g (1000 rpm in a rotor with 16 .1 cm radius) at RT. Discard supernatant and keep the pellet in 100 μL supernatant. 9. Gently resuspend the cell pellets and add 5 mL of 75 mM KCl at RT. Let stand 3 min to allow red blood cell lysis. Note : Hypotonic treatment is a critical step for slide preparation. Mild treatment is suggested to avoid the loss of necrotic and apoptotic cells, and also to avoid cytoplasmic loss. Some laboratories may vary the length of hypotonic treatment. Note that decreasing the amount of hypotonic solution will impact more than decreasing the time of treatment. CBMN-cyt Assay and FISH 89

10. Add 400 μL of prefi xing solution and gently mix. 11. Centrifuge as in step 8. 12. Discard the supernatant, gently resuspend the cell pellets in 100 μL supernatant and replace with 10 mL of cold (−20 °C) methanol. Samples in methanol can be stored at −20 °C for months. 13. Centrifuge as in step 8. 14. Remove supernatant and replace with 5 mL of fi xing solution. 15. Repeat step 12 and 13 twice. 16. Remove the supernatant and resuspend the pellet in an appropriate volume of fi xative (300–500 μL). Slide preparation, fi xation and staining 17. Drop 80–100 μL of cell suspension directly onto clean iced slides. Make two slides for each sample. 18. Continue as described in (a) Lymphocytes , passage 19. (c) Cell lines 1. Determine the optimal number of cells to be plated into 6-well/plates. 6 × 104 cells/well (3 mL complete culture medium per well) is recommended. Proceed with metabolic activation if the cell line in use requires it. Note: The cell density should be enough to collect an adequate number of cells after treatment, but it should not allow 100 % confl uency. 2. 24 h after seeding, cells are exposed to the test compounds. 3. At 44 h add 30 μL cyt-B to each well. 4. At the end of the exposure period, cells are harvested and centrifuged for 10 min at 180 × g (1000 rpm in a rotor with 16.1 cm radius) at RT. 5. Discard supernatant and keep the pellet in 100 μL supernatant. 6. Gently resuspend the cell pellets and add 5 mL of 75 mM KCl at RT. Note: The length of hypotonic treatment can vary depending on the cell line used. Note that decreasing the amount of hypotonic solution will impact more than decreasing the time of treatment. 7. Continue as described in (b) Whole blood, passage 10.

3.1.2 Exposure When exposing cells to the compound of interest, it is important of the Cells to consider some parameters such as the toxic potential, the length of the exposure to the test substance and the timing for the addi- Test Compounds tion of cytochalasin-B to the cultures (Fig. 4 ). Prior to expose whole blood cells or lymphocytes, the cytotoxicity of the test compound should be evaluated. 90 Lucia Migliore et al.

Fig. 4 Proposed timeline of CBMN-cyt assay. The co-treatment with cytochalasin-B is delayed compared to the treatment with the compound of interest and starts at the 44th hour from the cell seeding. The micrograph shows the desired cell density obtained after slides preparation (Colour ﬁ gure online)

If no toxicity is detected:

● It is recommended that at least three doses of the test compound are assayed by CBMN-cyt assay ● The chosen concentrations should be separated by no more threefold spacing In case toxicity data sets are available for compound of interest:

● The concentrations to be used in CBMN-cyt assay are at least three, with no more that threefold spacing in between ● The dosage covers a range of concentrations ranging from low or no toxicity, to intermediate to high toxicity, with the highest tested dose equal to the concentration inducing 60 % cytotoxicity [77 ] The length of the exposure period to the test compound is equally important when cells are exposed to the test substance during CBMN-cyt assay. It is necessary to ensure that CBMN-cyt assay is performed on cells that have replicated at least once during exposure, and this is especially required in the case of aneuploidogenic compounds. Therefore, in the case of PHA- stimulated primary cells a treatment period of at least 24 h is recommended to be sure that all the different stages of the cell cycle are covered. If cells are not PHA stimulated, in contrast, a short exposure (3–6 h) is sufﬁ cient. CBMN-cyt Assay and FISH 91

Equally, critical is the addition of cyt-B to the cultures [68 – 70 ]. As explained in the Introduction section of this chapter, there are four different options to add cyt-B to the cells incubated with the test compound: (a) Co-treatment : the substance of interest and cyt-B are added to the cultured cells at the same time and, consequently, are harvested together. (b) Post-treatment : cells are ﬁ rst incubated with the test compound and then with cyt-B. Before cyt-B is added to the cultures, the test compound is removed and medium refreshed. (c) Delayed co-treatment : the treatment is started 24 h after cell seeding and cyt-B is added 20 h after the beginning of the exposure (44 h after cell seeding). (d) Treatment alone : cells are exposed solely to the test compound and not to cyt-B.

Solvent/Vehicle Solvent/vehicle controls should be included in each experiment. The solvent/vehicle should not chemically react with the test substance and should not be cytotoxic and/or genotoxic to cells. If solvent/vehicle other than water and cell culture medium are used, their use should be supported by data indicating their lack of cytotoxic and genotoxic effects. In the absence of these data, it is mandatory to include untreated controls to prove that no effects are induced by the chosen solvent. Generally, solvent controls should not exceed 1 % (v/v).

Positive Control Concurrent positive controls are needed to demonstrate the sensitivity of the test system to identify clastogens and aneuploidogens. The positive control should be a substance known to induce micronuclei formation at concentrations expected to give reproducible frequencies over the background control. Examples of positive control substances to use in experiments with and without metabolic activation are indicated in Table 1 . Other positive control substances may be used if justiﬁ ed.

3.1.3 Scoring Criteria The scoring criteria we refer are those proposed by Fenech [4 ]. Scoring CBMN-cyt slides offers information on cytostatic, cytotoxic and genotoxic effects of the test compounds. Cytostasis is determined by the frequency of mono-, bi- and multinucleated viable cells, cytotoxicity by the apoptotic and necrotic indices calculated on 500 cells, and genotoxicity, in contrast, is calculated on 1000 binucleated cells and is given by the frequency of the chromosomal aberrations (MN, NPB and NBUD) identiﬁ ed. 92 Lucia Migliore et al.

Table 1 List of controls to be used for CBMN-cyt assay and FISH

Category Chemical CAS number Solvent

Aneuploidogens Diazepam 439-14-5 DMSO Methyl-2-benzimidazole 10605-21-7 DMSO Thiabendazole 148-79-8 DMSO Vinblastine 143-67-9 DMSO Diethylstilbestrol 56-53-1 DMSO Colchicine 64-86-8 Water Clastogens requiring 2-Acetylaminoﬂ uorene 53-96-3 DMSO metabolic activation 7,12-Dimethylbenzanthracene 57-97-6 DMSO Clastogens not requiring Actinomycin D 50-76-0 Ethanol metabolic activation Cytosine arabinoside 147-94-4 Water Bleomycin sulphate 11056-06-7 Water Cadmium sulphate 10124-36-4 Water Ethylmethanesulphonate 62-50-0 Water 5-Fluorouracil 51-21-8 Water Mitomycin C 50-07-7 Water Negative substances Pyrene 129-00-0 DMSO Di(2-ethylhexyl)phthalate 117-81-7 DMSO Sodium chloride 7647-14-5 Water Nalidixic acid 389-08-02 DMSO

Cytostasis: Scoring To determine the cytostasis of the test compound, the number of Viable Mono-, Bi- and mono-, bi- and multinucleated cells is calculated (Fig. 5a, b ). To be Multinucleated Cells counted as mono-, bi- or multinucleated, cells must have speciﬁ c characteristics:

● Be viable ● Present an intact cytoplasm and normal nuclear morphology ● Mononucleated cells contain one nucleus, binucleated cells present two nuclei and multinucleated cells show three or more nuclei Mono-, bi- and multinucleated cells must be scored either if they contain one or more MN or NBUD or if they do not, and in the case of bi- and multinucleated cells NPB should be scored as well. The total number of these cell types, in fact, is necessary to determine parameters such as the proliferation index and the number of binucleated micronucleated (BNMN) cells. The cytokinesis-block proliferation index (CBPI) is calculated on the ﬁ rst 500 viable cells counted using the following formula:

12 number of mononucleated number of binucleated   3 nuumber of multinucleated  CBPI    500 viable cells counted

CBMN-cyt Assay and FISH 93

Fig. 5 Micrographs of the endpoints scored by CBMN-cyt assay. (a ) mono- and binucleated cells; (b ) bi- and multinucleated cells; (c , d ) necrosis; (e , f ) early apoptotic cells; (g , h ) late apoptotic cells; (i ) micronucleated binucleated cell; (j ) nucleoplasmic bridge; (k ) nucleoplasmic bridge and a micronucleus; (l ) nuclear bud (Colour ﬁ gure online) 94 Lucia Migliore et al.

Cytotoxicity: Scoring Necrotic cells, as in the case of apoptotic ones, should not be the Necrotic Index counted as viable cells. And can be divided into early or late necrotic cells. Early necrotic cells are characterized by:

● The presence of cytoplasmic vacuoles ● Whitish cytoplasm ● Damaged cell membrane ● Intact nuclei Late necrotic cells, in contrast, are characterized by:

● Loss of cytoplasm and cell membrane ● Loss of the integrity of the nuclear membrane ● Release of the nuclear content On the CBMN-cyt slides necrotic cells are also recognized by a pale intensity of the cytoplasmatic and nuclear staining compared to viable cells (Fig. 5c, d ). The necrotic index is calculated as the percentage of necrotic cells (early and late necrosis) in 500 viable cells scored. Cytotoxicity: Scoring the Apoptotic Index Apoptotic cells must not be considered viable cells. They are divided into early or late apoptotic cells depending on speciﬁ c features (Fig. 5e–h ). Early apoptotic stages:

● Intact cytoplasm and nuclear membrane ● The presence of condensed chromatin in the nuclei When the condensed chromatin is fragmented into nuclear bodies and the cytoplasm is still intact, then cells are in a late apoptotic stage. The apoptotic index is calculated as the percentage of apoptotic cells (early and late apoptosis) in 500 viable cells scored. The intensity of the nuclear and cytoplasmic staining in apoptotic cells is higher than in viable cells.

Genotoxicity: Scoring MN must be scored only in viable binucleated cells (BN) character- Micronuclei ized by the presence, in the same undamaged cytoplasm, of two nuclei with intact nuclear membranes in close proximity to each other, but not overlapping (Fig. 5i ). Furthermore, the nuclei must have equal size and staining intensity, and can be connected by an NPB. MN must be carefully identiﬁ ed and distinguished from staining residues or cellular debris that, sometimes, are visible on CBMN-cyt slides. To correctly identify them, MN must:

● Have same morphology, but smaller size (size ranging from 1/3rd to 1/16th of the diameter of the main nuclei), than the nuclei CBMN-cyt Assay and FISH 95

● Not be attached to the main nuclei (in this case they should be scored as NBUD if a connecting peduncle is present) or overlapping with them ● At the microscope MN appear non-refractive and on the same focal plane as the nuclei A BN cell can contain more than one MN. The BNMN frequency is evaluated as

total number of BNcellscontainingoneor moreMN 1000BNcells

and the MN frequency as

total number of MN 1000BNcells

Scoring Nucleoplasmic NPB are dicentric chromosomes linking the two nuclei in a BN Bridges cell. They

● Appear as a thin bridge of genomic material ● Show the same staining intensity and focal plane as the main nuclei ● Their size is one-fourth of the nuclei in the BN cell A BN cell can contain more than one NPB, and NPB can be detected in BN cells with one or more MN. BN cells with NPB and no MN are also detectable (Fig. 5j, k ). The BN-NPB frequency is calculated as

total number of BNcellscontainingoneor moreNPB 1000BNcells

and the NPB frequency as

total number of NPB 1000BNcells

Scoring Nuclear Buds NBUD are morphologically similar to MN, but characterized by

● The presence of a tail of nucleoplasmic material connecting the NBUD with the nucleus (Fig. 5l ) ● Same staining intensity and focal plane as the main nuclei ● Their size is equal to the MN 96 Lucia Migliore et al.

The BN-NBUD frequency is calculated as total number of BNcells containingoneNBUD 1000BNcells

and the NBUD frequency as total number of NBUD 1000BNcells

3.2 FISH (a) Pretreatment 1. Pretreat prepared slides in 2× SSC, 0.5 % Igepal, pH 7.0 at 37 °C for 15 min. Dehydrate, for 1 min each, in 70, 85 and 100 % ethanol. Air-dry at RT (b) Co-denaturation 2. Add 10 μL of centromeric probes each 22 × 22 mm fi eld 3. Cover with glass coverslip and seal with Fixogum or Rubber Cement 4. Denature sample and probe on a hot plate at 75 °C for 8–10 min (c) Hybridization, post-hybridization washing and counterstaining 5. Incubate overnight at 37 °C in a humidifi ed chamber 6. Remove fi xogum and slide off the coverslips 7. Wash the slides in 0.4× SSC/0.3 % Igepal for 2 min at 72 °C (±1) without agitation, and then in 2× SSC/0.1 % Igepal for 1 min at RT without agitation. For slides with highly cytoplasmic background, pretreat the dry sample in 2× SSC at 37 °C for 2 min and incubate the slides 5–10 min in 0.005 % Pepsin solution in 0.01 M HCl at 37 °C. Wash slide for 3 min in PBS at RT and dehydrate in 70, 85 and 100 % ethanol for 1 min each. Air- dry at RT and proceed with the co-denaturation step Note: Please remember that stringency of hybridization and washing (temperatures and buffer concentrations) are very important, as lower stringency can result in non-specifi c binding of the probe to other sequences, and higher stringency can result in lack of signal. Incomplete denaturation of target DNA can result in lack of signal 8. Dehydrate in 70, 85 and 100 % ethanol for 1 min each 9. Air-dry at room temperature 10. Add 50 μL DAPI solution on the slide, cover with coverslips and allow to stain 10 min at RT Note: Note that detection reagents and DAPI are light sensitive; avoid exposure to ambient light by wrapping containers in aluminium foil CBMN-cyt Assay and FISH 97

11. Add 7 μL antifade mounting medium to the stained slide and add a coverslip 12. Seal with fi xogum 13. Store at −20 °C in slide box with desiccant (d) Scoring criteria 14. Examine the slides using a fl uorescence microscope with epi-illumination and an appropriate fi lter set for the fl uorochrome used. Examine for the presence of one or more centromeric spots the micronuclei present in the binucleated cells with intact cytoplasm, and classify them as centromere positive (MN C+) or centromere negative (MN C−)

4 Future Perspectives

New features and new techniques to investigate micronuclei formation and genotoxic damage are constantly explored given the increased employment of MN assay and the challenging opportuni- ties of CBMN-cyt assay. For this reason, in addition to the use of whole blood and separated lymphocytes, other tissues and cell types are currently investigated for toxicological studies as well as for biomonitoring. However, although new knowledge has been added in the recent past, with many publications exploring the feasibility of CBMN-cyt assay to nasal epithelial, stomach and hair-root cells [78 –80 ], still validated and standardized protocols are required. To further improve the protocols and the scoring techniques currently used for the detection of MN induction, other innovative approaches such as fl ow cytometry, laser scanning cytometry and automated image analysis scoring have been developed. Each of these approaches has the advantage of reducing the time of analysis compared to the classical visual scoring at microscope, and of increasing the amount of analysed cells improving, therefore, the quality of the results. Nevertheless, at the same time, there are critical steps that should be taken into account. While image analysis does not seem to be easy and does not signifi cantly improve the scoring speed, fl ow cytometry requires some arrangements to separate reticulocytes from mature whole blood cells. With mouse samples this problem has been overcome by using a double- staining approach which allows the distinction between micronucleated reticulocytes and blood cells [81 ], and in human blood physical enrichments have been successfully adopted [82 – 85 ]. Overall, these automated scoring techniques present other disadvantages. There is still lack of agreement on the experimental design and on the statistical method to be used for data analysis, and the improvement of the technique constantly requires a re-evaluation of the acquired results. In addition, the disruption of the plasma membrane, needed to detect MN by fl ow cytometry protocol, can generate false positives because the technique does 98 Lucia Migliore et al.

not allow the recognition of particles, organelles or fragments from MN. After lysis, in fact, it is difﬁ cult to distinguish individual chromosomes or chromosome aggregates isolated from mitotic cells, fragments of nuclear chromatin or chromatin granules from apoptotic cells and apoptotic bodies from “real” MN [ 86 ]. Moreover, ﬂ ow cytometry does not solve the distribution of micronuclei scored in a cell population: cells with more than one MN cannot be recognized from cells without or with one single MN, leading thus to a mis-evaluation of the data. In conclusion, although many efforts have been made in the last years in order to improve the quality of the CBMN-cyt data, the proposed techniques need further development, making the manual scoring still up-to-date.

Acknowledgements

SDB and CU are granted by the FP7 project No 280716, SANOWORK ( www.sanowork.eu ). We would like to acknowledge Davide Tesoro for drawing ﬁ gures.

Glossary

BN Binucleated cells BNMN Binucleated micronucleated cells BrdU Bromodeoxyuridine CBMN Cytokinesis-block micronucleus CBMN-cyt Cytokinesis-block micronucleus cytome CBPI Cytokinesis-block proliferation index CENP-A Centromere protein A CENP-B Centromere protein B cyt-B Cytochalasin-B DAPI 4 ′,6-Diamidin-2-fenilindolo DMSO Dimethylsulfoxide DNA Deoxyribonucleic acid FBS Foetal bovine serum FISH Fluorescence in situ hybridization HBSS Hank’s balanced salt solution MN Micronucleus/i MN C Micronucleus/i centromere negative MN C+ Micronucleus/i centromere positive NBUD Nuclear bud/s NM Nanomaterial/s NPB Nucleoplasmic bridge/s PBS Phosphate-buffered solution PHA Phytohemagglutinin CBMN-cyt Assay and FISH 99

PI Proliferation index RICC Relative increase in cell counts R T Room temperature SSC Sodium chloride/sodium citrate

References

1. Hutchison HE, Ferguson-Smith MA (1959) 12. Corvi R, Albertini S, Hartung T et al (2008) The signifi cance of Howell-Jolly bodies in red ECVAM retrospective validation of in vitro cell precursors. J Clin Pathol 12:451–453 micronucleus test (MNT). Mutagenesis 2. Countryman PI, Heddle JA (1976) The pro- 23(4):271–283 duction of micronuclei from chromosome 13. OECD. In vitro Mammalian Cell Micronucleus aberrations in irradiated cultures of human Test (MNvit). OECD Guideline for Testing of lymphocytes. Mutat Res 41(2–3):321–332 Chemicals No. 487. OECD, Paris. http:// 3. Fenech M, Morley A (1985) Solutions to the www.oecd.org/env/testguidelines kinetic problem in the micronucleus assay. 14. Fenech M, Kirsch-Volders M, Natarajan AT Cytobios 43(172–173):233–246 et al (2011) Molecular mechanisms of micro- 4. Fenech M (2007) Cytokinesis-block micronu- nucleus, nucleoplasmic bridge and nuclear bud cleus cytome assay. Nat Protoc 2:1084–1104 formation in mammalian and human cells. 5. HUMN: HUman MicroNucleus Project. Mutagenesis 26(1):125–132 www.humn.org . Accessed 8 Oct 2013 15. Fenech M (2010) The lymphocyte cytokinesis- 6. Bonassi S, Fenech M, Lando C et al (2001) block micronucleus cytome assay and its appli- HUman MicroNucleus project: international cation in radiation biodosimetry. Health Phys database comparison for results with the 98:234–243 cytokinesis-block micronucleus assay in human 16. Hoffelder DR, Luo L, Burke NA et al (2004) lymphocytes: I. Effect of laboratory protocol, Resolution of anaphase bridges in cancer cells. scoring criteria, and host factors on the fre- Chromosoma 112:389–397 quency of micronuclei. Environ Mol Mutagen 17. Fenech M (2006) Cytokinesis-block micronu- 37(1):31–45 cleus assay evolves into a “cytome” assay of 7. Bonassi S, Neri M, Lando C et al (2003) Effect chromosomal instability, mitotic dysfunction of smoking habit on the frequency of micronu- and cell death. Mutat Res 600(1–2):58–66 clei in human lymphocytes: results from the 18. Gisselsson D (2008) Classifi cation of chromo- Human MicroNucleus project. Mutat Res some segregation errors in cancer. 543(2):155–166 Chromosoma 117:511–519 8. Fenech M, Chang WP, Kirsch-Volders M et al 19. Guerrero AA, Gamero MC, Trachana V et al (2003) HUMN project: detailed description (2010) Centromere-localized breaks indicate of the scoring criteria for the cytokinesis-block the generation of DNA damage by the mitotic micronucleus assay using isolated human lym- spindle. Proc Natl Acad Sci U S A 107(9): phocyte cultures. Mutat Res 534(1–2):65–75 4159–4164 9. Bonassi S, Biasotti B, Kirsch-Volders M et al 20. Payne CM, Crowley-Skillicorn C, Bernstein C (2009) State of the art survey of the buccal et al (2010) Hydrophobic bile acid-induced micronucleus assay—a fi rst stage in the micronuclei formation, mitotic perturbations, HUMN(XL) project initiative. Mutagenesis and decreases in spindle checkpoint proteins: 24(4):295–302 relevance to genomic instability in colon carci- 10. Fenech M, Holland N, Chang WP et al (1999) nogenesis. Nutr Cancer 62(6):825–840 The HUman MicroNucleus Project—An 21. Fenech M, Baghurst P, Luderer W et al (2005) international collaborative study on the use of Low intake of calcium, folate, nicotinic acid, the micronucleus technique for measuring vitamin E, retinol, beta-carotene and high DNA damage in humans. Mutat Res intake of pantothenic acid, biotin and ribofl a- 428(1–2):271–283 vin are signifi cantly associated with increased 11. Bonassi S, Znaor A, Ceppi M et al (2007) An genome instability—results from a dietary increased micronucleus frequency in periph- intake and micronucleus index survey in South eral blood lymphocytes predicts the risk of Australia. Carcinogenesis 26:991–999 cancer in humans. Carcinogenesis 28(3): 22. Schueler MG, Sullivan BA (2006) Structural 625–631 and functional dynamics of human centromeric 100 Lucia Migliore et al.

chromatin. Annu Rev Genomics Hum Genet 36. Dutra A, Pak E, Wincovitch S et al (2010) 7:301–313 Nuclear bud formation: a novel manifestation 23. Xu GL, Bestor TH, Bourc’his D et al (1999) of Zidovudine genotoxicity. Cytogenet Chromosome instability and immunodefi ciency Genome Res 128(1–3):105–110 syndrome caused by mutations in a DNA meth- 37. Utani K, Kohno Y, Okamoto A et al (2010) yltransferase gene. Nature 402(6758): Emergence of micronuclei and their effects on 187–191 the fate of cells under replication stress. PLoS 24. Suzuki T, Fujii M, Ayusawa D (2002) One 5(4):e10089 Demethylation of classical satellite 2 and 3 38. Montero R, Serrano L, Ostrosky P (1997) In DNA with chromosomal instability in senes- vitro induction of micronuclei in lymphocytes: cent human fi broblasts. Exp Gerontol 37(8–9): the use of bromodeoxyuridine as a prolifera- 1005–1014 tion marker. Mutat Res 391:135–141 25. Putiri EL, Robertson KD (2011) Epigenetic 39. Serrano-García L, Montero-Montoya R mechanisms and genome stability. Clin (2001) Micronuclei and chromatid buds are Epigenetics 2(2):299–314 the result of related genotoxic events. Environ 26. Hovhannisyan G, Aroutiounian R, Liehr T Mol Mutagen 38(1):38–45 (2012) Chromosomal composition of micronu- 40. Haaf T, Raderschall E, Reddy G et al (1999) clei in human leukocytes exposed to mitomycin Sequestration of mammalian Rad51- C. J Histochem Cytochem 60(4):316–322 recombination protein into micronuclei. J Cell 27. Gieni RS, Chan G, Hendzel MJ (2008) Biol 144(1):11–20 Epigenetics regulate centromere formation 41. Leach NT, Jackson-Cook C (2004) and kinetochore function. J Cell Biochem 104: Micronuclei with multiple copies of the X 2027–2039 chromosome: do chromosomes replicate in 28. Hayden KE, Strome ED, Merrett SL et al micronuclei? Mutat Res 554:89–94 (2013) Sequences associated with centromere 42. Mateuca R, Lombaert N, Aka PV et al (2006) competency in the human genome. Mol Cell Chromosomal changes: induction, detection Biol 33(4):763–772 methods and applicability in human biomoni- 29. Thomas P, Umegaki K, Fenech M (2003) toring. Biochimie 88(11):1515–1531 Nucleoplasmic bridges are a sensitive measure of 43. Decordier I, Dillen L, Cundari E et al (2002) chromosome rearrangement in the cytokinesis- Elimination of micronucleated cells by apopto- block micronucleus assay. Mutagenesis 18(2): sis after treatment with inhibitors of microtu- 187–194 bules. Mutagenesis 17:337–344 30. Thomas P, Fenech M (2011) Cytokinesis- 44. Terradas M, Martín M, Tusell L et al (2010) block micronucleus cytome assay in lympho- Genetic activities in micronuclei: is the DNA cytes. Methods Mol Biol 682:217–234 entrapped in micronuclei lost for the cell? 31. Acilan C, Potter DM, Saunders WS (2007) Mutat Res 705:60–67 DNA repair pathways involved in anaphase 45. Kirsch-Volders M, Plas G, Elhajouji A et al bridge formation. Genes Chromosomes (2011) The in vitro MN assay in 2011: origin Cancer 46(6):522–531 and fate, biological signifi cance, protocols, high 32. Sofueva S, Osman F, Lorenz A et al (2011) throughput methodologies and toxicological Ultrafi ne anaphase bridges, broken DNA and relevance. Arch Toxicol 85(8):873–899 illegitimate recombination induced by a repli- 46. Decordier I, Cundari E, Kirsch-Volders M cation fork barrier. Nucleic Acids Res 39(15): (2005) Infl uence of caspase activity on micro- 6568–6584 nuclei detection: a possible role for caspase-3 in 33. Murnane JP (2012) Telomere dysfunction and micronucleation. Mutagenesis 20:173–179 chromosome instability. Mutat Res 730(1–2): 47. Yasui M, Koyama N, Koizumi T et al (2010) 28–36 Live cell imaging of micronucleus formation 34. Pampalona J, Frías C, Genescà A et al (2012) and development. Mutat Res 692:12–18 Progressive telomere dysfunction causes cyto- 48. Matsushima T, Hayashi M, Matsuoka A et al kinesis failure and leads to the accumulation of (1999) Validation study of the in vitro micro- polyploid cells. PLoS Genet 8(4):e1002679 nucleus test in a Chinese hamster lung cell line 35. Shimizu N, Shimura T, Tanaka T (2000) (CHL/IU). Mutagenesis 14(6):569–580 Selective elimination of acentric double min- 49. Lorge E, Thybaud V, Aardema MJ et al (2006) utes from cancer cells through the extrusion of SFTG international collaborative study on micronuclei. Mutat Res 448(1):81–90 in vitro micronucleus test I. General conditions CBMN-cyt Assay and FISH 101

and overall conclusions of the study. Mutat Res pline evolving from studies of ultrafi ne parti- 607(1):13–36 cles. Environ. Health Perspect 113:823–839 50. Sobol Z, Homiski ML, Dickinson DA et al 63. Pfuhler S, Elespuru R, Aardema MJ et al (2012) Development and validation of an (2013) Genotoxicity of nanomaterials: refi n- in vitro micronucleus assay platform in TK6 ing strategies and tests for hazard identifi ca- cells. Mutat Res 746(1):29–34 tion. Environ Mol Mutagen 54(4):229–239 51. Westerink WM, Schirris TJ, Horbach GJ et al 64. Sargent LM, Shvedova AA, Hubbs AF et al (2011) Development and validation of a high- (2009) Induction of aneuploidy by single- content screening in vitro micronucleus assay walled carbon nanotubes. Environ Mol in CHO-k1 and HepG2 cells. Mutat Res Mutagen 50:708–717 724(1–2):7–21 65. Di Bucchianico S, Fabbrizi MR, Misra SK et al 52. Gonzalez L, Thomassen LC, Plas G et al (2013) Multiple cytotoxic and genotoxic (2010) Exploring the aneugenic and clasto- effects induced in vitro by differently shaped genic potential in the nanosize range: A549 copper oxide nanomaterials. Mutagenesis human lung carcinoma cells and amorphous 28(3):287–299 monodisperse silica nanoparticles as models. 66. Landsiedel R, Kapp MD, Schulz M et al Nanotoxicology 4:382–395 (2009) Genotoxicity investigations on nano- 53. Schmuck G, Lieb G, Wild D et al (1988) materials: methods, preparation and character- Characterization of an in vitro micronucleus ization of test material, potential artifacts and assay with Syrian hamster embryo fi broblasts. limitations—many questions, some answers. Mutat Res 203(6):397–404 Mutat Res 681(2–3):241–258 54. Parry EM, Parry JM, Corso C et al (2002) 67. Warheit DB, Donner EM (2010) Rationale of Detection and characterization of mechanisms genotoxicity testing of nanomaterials: regula- of action of aneugenic chemicals. Mutagenesis tory requirements and appropriateness of avail- 17(6):509–521 able OECD test guidelines. Nanotoxicology 55. Yamamoto KI, Kikuchi Y (1980) A compari- 4:409–413 son of diameters of micronuclei induced by 68. Doak SH, Griffi ths SM, Manshian B et al (2009) clastogens and by spindle poisons. Mutat Res Confounding experimental considerations in 71(1):127–131 nanogenotoxicology. Mutagenesis 24:285–293 56. Vig BK, Swearngin SE (1986) Sequence of 69. Doak SH, Manshian B, Jenkins GJS et al centromere separation: kinetochore formation (2012) In vitro genotoxicity testing strategy for in induced laggards and micronuclei. nanomaterials and the adaptation of current Mutagenesis 1(6):461–465 OECD guidelines. Mutat Res 745:104–111 57. Eastmond DA, Tucker JD (1989) Identifi cation 70. Gonzalez L, Sanderson BJ, Kirsch-Volders M of aneuploidy-inducing agents using (2011) Adaptations of the in vitro MN assay cytokinesis-blocked human lymphocytes and for the genotoxicity assessment of nanomateri- an antikinetochore antibody. Environ Mol als. Mutagenesis 26(1):185–191 Mutagen 13(1):34–43 71. Singh N, Manshian B, Jenkins GJS et al (2009) 58. Migliore L, Bocciardi R, Macrì C et al (1993) NanoGenotoxicology: the DNA damaging Cytogenetic damage induced in human lym- potential of nanomaterials. Biomaterials phocytes by four vanadium compounds and 30:3891–3914 micronucleus analysis by fl uorescence in situ 72. Fenech M (2000) The in vitro micronucleus hybridization with a centromeric probe. Mutat technique. Mutat Res 455(1–2):81–95 Res 319(3):205–213 73. Corradi S, Gonzalez L, Thomassen LC et al 59. Migliore L, Zotti-Martelli L, Scarpato R (1999) (2012) Infl uence of serum on in situ prolifera- Detection of chromosome loss and gain tion and genotoxicity in A549 human lung induced by griseofulvin, estramustine, and van- cells exposed to nanomaterials. Mutat Res adate in binucleated lymphocytes using FISH 745(1–2):21–27 analysis. Environ Mol Mutagen 34(1):64–68 74. Gonzalez L, Lukamowicz-Rajska M, 60. Guttenbach M, Schmid M (1994) Exclusion Thomassen LC et al (2014) Co-assessment of of specifi c human chromosomes into micronu- cell cycle and micronucleus frequencies dem- clei by 5-azacytidine treatment of lymphocyte onstrates the infl uence of serum on the in vitro cultures. Exp Cell Res 211(1):127–132 genotoxic response to amorphous monodis- 61. Herrera LA, Prada D, Andonegui MA et al perse silica nanoparticles of varying sizes. (2008) The epigenetic origin of aneuploidy. Nanotoxicology 8(8):876–884 Curr Genomics 9(1):43–50 75. Prasad RY, Wallace K, Daniel KM et al (2013) 62. Oberdorster G, Oberdorster E, Oberdorster J Effect of treatment media on the agglomera- (2005) Nanotoxicology: an emerging disci- tion of titanium dioxide nanoparticles: impact 102 Lucia Migliore et al.

on genotoxicity, cellular interaction, and cell 81. Grawé J, Zetterberg G, Amnéus H (1992) cycle. ACS Nano 7(3):1929–1942 Flow-cytometric enumeration of micronucle- 76. Von der Hude W, Kalweit S, Engelhardt G ated polychromatic erythrocytes in mouse et al (2000) In vitro micronucleus assay with peripheral blood. Cytometry 13(7):750–758 Chinese hamster V79 cells—results of a collab- 82. Grawé J, Biko J, Lorenz R et al (2005) orative study with in situ exposure to 26 chem- Evaluation of the reticulocyte micronucleus ical substances. Mutat Res 468(2):137–163 assay in patients treated with radioiodine for 77. Kirsch-Volders M, Sofuni T, Aardema M et al thyroid cancer. Mutat Res 583(1):12–25 (2003) Report from the in vitro micronucleus 83. Dertinger SD, Miller RK, Brewer K et al assay working group. Mutat Res 540(2): (2007) Automated human blood micronucle- 153–163 ated reticulocyte measurements for rapid 78. Fenech M, Holland N, Zeiger E et al (2011) assessment of chromosomal damage. Mutat The HUMN and HUMNxL international col- Res 626(1–2):111–119 laboration projects on human micronucleus 84. Flanagan JM, Howard TA, Mortier N et al assays in lymphocytes and buccal cells—past, (2010) Assessment of genotoxicity associated present and future. Mutagenesis 26(1):239–425 with hydroxyurea therapy in children with 79. Knasmueller S, Holland N, Wultsch G et al sickle cell anemia. Mutat Res 698(1–2):38–42 (2011) Use of nasal cells in micronucleus 85. Witt KL, Livanos E, Kissling GE et al (2008) assays and other genotoxicity studies. Comparison of ﬂ ow cytometry- and Mutagenesis 26(1):231–238 microscopy-based methods for measuring 80. Ohyama W, Okada E, Fujiishi Y et al (2013) In micronucleated reticulocyte frequencies in vivo rat glandular stomach and colon micro- rodents treated with nongenotoxic and geno- nucleus tests: kinetics of micronucleated cells, toxic chemicals. Mutat Res 649(1–2):101–113 apoptosis, and cell proliferation in the target 86. Darzynkiewicz Z, Smolewski P, Holden E et al tissues after a single oral administration of (2011) Laser scanning cytometry for automa- stomach- or colon-carcinogens. Mutat Res tion of the micronucleus assay. Mutagenesis 755(2):141–147 26(1):153–161 Chapter 6

The In Vivo Rodent Micronucleus Test

Edson Luis Maistro

Abstract

The in vivo mammalian micronucleus assay is the primary test in a battery of genotoxicity tests recommended by the regulatory agencies worldwide. The purpose of the assay is to identify substances that cause clastogenicity (chromosome breakage) and aneugenicity (chromosome lagging due to spindle dysfunction), and bone marrow toxicity by estimating the ratio of polychromatic erythrocytes to normochromatic erythrocytes. This chapter describes the mechanism of micronucleus formation, presents practical guidelines for designing studies, and gives the step-by-step protocols of the in vivo micronucleus test in bone marrow and peripheral blood cells of rodents.

Key words Micronucleus test , MN protocol , Mice , Rats , PCE/NCE ratio , Clastogenicity test , Aneugenicity test

1 Introduction

Genetic toxicology tests are in vitro or in vivo assays designed to detect compounds that induce genetic damage directly or indirectly. Fixation of DNA damage can result in gene mutations, loss of heterozygosity, chromosomal loss or gain, and chromosome aberrations. These events may play an important role in the neo- plastic development of certain tumors and may also induce heritable effects leading to birth defects [1 , 2 ]. Thus, identifying genotoxic effects of biological, physical, and chemical agents is important for the risk/beneﬁ t assessment of their potential use in humans. The in vivo rodent micronucleus (MN) assay is the primary test in a battery of genotoxicity tests and is recommended by the regulatory agencies around the globe to be conducted as part of product safety assessment [3 ]. The test procedure, developed by Schmid and coworkers [ 4 – 8 ] presented important advantages over the analysis of bone marrow metaphase chromosomes: it is simpler, faster, and at least as sensitive as chromosome analysis in the same material without expense of accuracy; in addition, it allows the

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_6, © Springer Science+Business Media New York 2014 103 104 Edson Luis Maistro

Fig. 1 Schema of micronucleus origin in vivo in bone marrow cells

detection of aneugenic effect of the test substance, as a result of disturbance of the chromosome segregation machinery. The micronucleus test affords a procedure for the detection of aberrations involving anaphase chromosome behavior utilizing a particularly useful cell type, the bone marrow erythroblast. The principle of the test is based on the formation of “micronuclei” (also known as Howell–Jolly bodies in hematological routine) from particles of chromatin material which, due to chromosome breakage or spindle dysfunction, do not migrate to the poles during anaphase and are not incorporated into the telophase nuclei of the dividing cell. Such chromatin fragments, or even whole chromosomes in the case of chromosome lag, result in the formation of one or more small satellite nuclei in the cytoplasm of the daughter cells (Fig. 1 ). The expulsion of the mammalian erythrocyte nucleus during the erythroblast development into a polychromatic erythrocyte (PCE) follows the ﬁ nal mitotic division by several hours, but the resulting enucleated erythrocyte still contains RNA, retaining its cytoplasmic basophilia for approximately 24 h after nuclear extrusion. Micronuclei in the cytoplasm of these cells are not expelled with the nucleus. If scoring of micronuclei is restricted to this cell type (PCE), then it is known that the anomalies were bound to arise mostly during the course of the immediately preceding mitosis. An increase in the frequency of micronucleated PCE (MNPCE) in test agent-treated animals is an indication of The In Vivo Rodent Micronucleus Test 105

induced chromosome damage [ 9 , 8 ]. In addition, estimating the ratio of PCE to normochromatic erythrocytes ((NCE) formed prior to test compound exposure) is useful to detect any perturbations in hematopoiesis as a result of treatment. It is considered that a decrease of the PCE/NCE ratio is an indicator of bone marrow toxicity induced by mutagens [3 , 10 ]. This mammalian in vivo test is especially relevant to assessing genotoxicity hazard because it allows the consideration of factors of in vivo metabolism, pharmacokinetics and DNA-repair process although these may vary among species, among tissues, and among genetic endpoints [ 11 ]. The assay is also useful in further investigation of a chromosome aberration detected by an in vitro test system.

2 Materials

– Mouse or rat – Polypropylene boxes with metal railings to keep the animals – Drinking fountains – Food for mouse or rat – Shavings (white pine sawdust) to line the boxes animals. It should be sterilized by autoclaving. – Picric acid for animal identiﬁ cation – Alcohol 70 % – Tweezers and scissors for animal dissection – Absorbent towel paper, to clean debris from tissues of the femur – 1 mL syringe with needle – Conical centrifuge tubes (15 mL) – Pasteur pipettes with rubber pipettor – Support for centrifuge tubes – Fetal bovine serum (FBS) (store in freezer) – 0.9 % NaCl (store in refrigerator) – Positive controls: Cyclophosphamide, Mitomicin C, N - nitroso- N -ethylurea, Doxorubicin – Slides with matte border – Support for the slides drying – Leishman eosin-methylene blue dye – Coverslips for the slides mounting – Permount ® or Entellan® for permanent slides mounting 106 Edson Luis Maistro

– Immersion oil for analysis of the slides – Xylene for mounting the slides – Boxes for storing slides

3 Assay Protocol

3.1 Animal Selection, Mice and rats are the species most commonly used to detect micro- Number, and Housing nucleus induction because this model has been utilized in other toxicological studies regarding the test agent under study. Any strain can be used. When the analysis involved peripheral blood, it is recommended the use of mice, since the spleen of rats eliminate most of the micronucleated erythrocytes from their peripheral blood [12 ]. Note : In the case of mice , the micronucleus assay can be used not only for detection of acute genetic damage , but also chronic damage , because of little or no selective removal of micronucleated cells from bone marrow and peripheral blood circulation . If rats are chosen, the use of fl ow cytometric analysis is recom- mendable, because the population of young peripheral PCE can be easily and reproducibly determined, allowing fundamentally the analysis of cells before MNPCE elimination [ 13 ]. Other appropriate mammalian species may be used if the spleen of these species does not remove micronucleated erythrocytes. Five animals of each sex per group were recommended for a standard study in the micronucleus test, with a single dosing regimen and two euthanasia times (24 and 48 h). The use of one sex, usually male, is suffi cient if the test agent does not cause differential toxicity between the sexes. If lethality is expected at the high dose of test agent, extra animals may be included at this group. Where human exposure to chemicals may be sex-specifi c, the test should be performed with animals of the appropriate sex. Note : If the availability of test substance is not a limiting factor , it can be administered to animals in a multiple dosing regimen , three or more consecutive times , with a 24 h interval between each administration. The bone marrow cells must be collected 24 h after the last dose applied. In this way , the number of animals of each experimental group is reduced to half , since there is no need for two euthanasia times. Thus , each experimental group could consist of fi ve animals of each sex , or only fi ve male animals , if the test substance does not cause differences in toxicity between sexes . Animals should be obtained from a recognized source of laboratory animals and should be acclimated to the laboratory environment for a minimum of 5 days. Strains of young healthy animals should be used. Animals may be housed individually, or caged in small groups of the same sex. The In Vivo Rodent Micronucleus Test 107

For any given experiment, the weight variation of animals should be minimal and not exceed ±20 % of the mean weight of each sex. The temperature in the experimental animal room should be 22 °C (±3 °C). The relative humidity should be at least 30 % and not exceed 70 % (preferably 50–60 %). Lighting should be artiﬁ cial, the sequence being 12 h light, 12 h dark. For feeding, conventional laboratory diets may be used, with food and water available ad libitum .

3.2 Test-Agent Note : It is recommended to obtain in the literature all information Dose Selection, available about the chemical to be tested , or other substances related Administration, to it , before conducting the micronucleus test. This information is and Controls important to properly identify the doses to be tested . The high dose selected for the rodent micronucleus assay is one that produces any compound-related signs of toxicity, or signifi cantly reduce survival. The analysis of fi nding the dose range should be performed in the same laboratory, using the same species, strain, sex, and treatment regimen to be used in the main study. It has also been recommended that the intermediate dose be one-half of the high dose and the low dose be one-half of the intermediate dose. If a test agent does not produce signs of toxicity until the limit dose ≥2,000 mg/kg/day [ 11 ], a full study using three dose levels may not be considered necessary and a limit dose test at 2,000 mg/kg may be suffi cient.

For the dose selection, also the Lethal Dose 50 % (LD50 ) can be utilized. In this case, the highest dose must be 80 % of the LD50 and the small ones, 50 and 25 % of the LD50 . For studies of a longer duration, the limit dose by day, for treatment up to 14 days is 2,000 mg/ kg, and 1,000 mg/kg/body weight/day if treatment longer than 14 days could be performed. Expected human exposure may indicate the need for a higher dose level to be used in the limit test. The route of administration of the test substance preferable has been that expected for human exposure. Usually, the compound is administered by gavage using a stomach tube or a suitable intubation cannula, or by intraperitoneal injection. Note : To administer intraperitoneally the chemical substance to mice , the animal should be placed with his head down , to prevent intestinal perforation. Using a ﬁ ne needle ( insulin ), the test substance is injected into the animal in the lower left quadrant of the abdomen. The volume to be administered should not exceed 10 mL /kg of body weight . Other routes, e.g., subcutaneous, intradermal, inhalation, intranasal, and intravenous, may be acceptable when they can be justiﬁ ed. 108 Edson Luis Maistro

Table 1 Designed groups for acute MN test

Treatment Dose (mg/kg) Animals/sex group number

Negative control (vehicle/solvent) M (5)a M (5)b F (5)a F (5)b Test substance Low M (5)a M (5)b F (5)a F (5)b Test substance Mid M (5)a M (5)b F (5)a F (5)b Test substance High M (5)a M (5)b F (5)a F (5)b Positive control (cyclophosphamide) 50 M (5)a F (5)a M male, F female a Bone marrow collected 24 h after treatment b Bone marrow collected 48 h after treatment

Note : For environmental chemicals or food additives studies , the test material is usually mixed in the diet or drinking water. The doses are commonly expressed in terms of concentration ( ppm ), or in terms of the quantity of material received by the animal ( mg / kg / day ), based on the amount of water or diet consumed. In order to ensure a dosage consistent with the increased weight of the animal , the investigator must check the weight of the animal as well as the consumption of water and food . Regarding the controls, it is important to determine the frequency of micronucleated erythrocytes in the group treated only with the vehicle or solvent control relative to the spontaneous frequency in the untreated animals group. This is to certify that solvent control is nontoxic at the dose volume used. After each laboratory veriﬁ es that the solvent control is nontoxic at the dose volume used and that it is not known to produce chemical reaction with the test agent, the use of the untreated animals group is generally not necessary. Frequently, water or methylcellulose aqueous solutions are used as solvent control. To provide an evaluation of the sensitivity of the assay, a positive control group must be performed. An agent known to induce micronuclei in the animal cells should be used. Cyclophosphamide i.p.—20–25 mg/kg for rats, 40–50 mg/kg for mice; mitomycin C i.p.—0.5 mg/kg for rats and mice, are examples. Table 1 exempli- ﬁ es a routine designed study.

3.3 Test Procedures: Animals are observed for clinical signs of toxicity as goosebumps, Bone Marrow/Blood reluctance to move, scream, ﬁ ght, panting behavior, seizures and Preparation tremors, at various intervals after treatment, and at 24 and 48 h. Animals are killed 24 and 48 h after administration of a single

dose of the test agent, by cervical dislocation, CO2 asphyxiation or other euthanasia method accepted internationally for laboratory animal use and care, and country law for animal protection and welfare. The In Vivo Rodent Micronucleus Test 109

3.3.1 Bone Marrow – Femurs are quickly removed and freed from muscle tissue. Preparations – After cutting a femur epiphysis, the bone marrow is removed by inserting a syringe needle in the canal of the bone marrow and, with soft movements, small amount of FBS is injecting, collecting the solution in a small Petri dish. – Using a Pasteur pipette, bone marrow cells are harvested and transferred into centrifuge tubes containing FBS. Saline solution (0.9 % NaCl) can also be used (about 3 mL/femur). – Cell suspensions are centrifuged at 1,000 rpm for 5–10 min and supernatants are discarded. – The pellets are suspended in a 0.5 mL of FBS and used for preparing bone marrow smears. Note: In this step , 3 – 4 drops of formaldehyde 4 % can be added for cell cytoplasm retaining . – To do the smears, a small drop of the viscous suspension is put on the end of a slide and spread by pulling the material behind a polished cover glass held at an angle of 45° (prepare at least two slides per animal). – After 24 h, air-dried smears are ﬁ xed in absolute methanol (10 min), stained with May-Grünwald Giemsa [ 8 ] or Giemsa alone [14 ], and all slides are coded for further blind evaluation.

3.3.2 Blood Preparations – Peripheral blood is obtained from the tail vein or other appropriate blood vessel. – Blood cells are immediately stained supravitally [15 – 17 ], or smear preparations are made, air-dried, fi xed in absolute methanol, and then stained. – The use of a DNA-specifi c stain (acridine orange [18 ] or Hoechst 33258 plus pyronin-Y [ 19 ]) can eliminate some of the artifacts associated with using a non-DNA-specifi c stain. Generally, micronuclei that arise from unrepaired chromosome breaks (clastogenic damage) lack a kinetochore, and those related to chromosome loss and nondisjunction of chromosomes contain a kinetochore. Commercially available antibodies can identify the kinetochore presence in the micronuclei [ 20 , 21 ].

3.4 Analysis Two hundred erythrocytes are examined for each animal to determine the ratio of polychromatic (PCE, immature) to NCE (mature), which is an indication of toxicity to bone marrow cells [10 ]. Note: The proportion of PCE among the bone marrow erythrocytes is usually 50 – 60 %. 110 Edson Luis Maistro

Table 2 Frequencies of micronucleated polychromatic erythrocytes (MNPCEs) and PCE/NCE ratio in bone marrow cells of male Swiss mice treated with three different doses of the test compound, and respective controls

MNPCEs

Treatments Number of PCEs analyzed Number Percentage PCE/NCE ratio (mean ± SD)

NaCl 0.9%a 10,000 3 0.03 1.11 ± 0.06 CPA (50 mg/kg)b 10,000 28 0.28 c 1.23 ± 0.15 Test agent dose 1 10,000 4 0.04 1.24 ± 0.12 Test agent dose 2 10,000 21 0.21 c 1.21 ± 0.12 Test agent dose 3 10,000 25 0.25 c 1.31 ± 0.13 a Negative control b Positive control c Signiﬁ cantly different from negative control (Student t -test, p < 0.001)

For bone marrow, 2,000 PCE per animal are scored for the incidence of micronucleated polychromatic erythrocytes (MNPCE), when fi ve animals are treated per concentration (two slides, 1,000 PCE scored per slide). All slides, including those of negative and positive controls, should be coded before microscopic analysis. For peripheral blood, 1,000 erythrocytes should be scored per animal (two slides, 500 PCE scored per slide). The unit of analysis is PCE and not the number of micronuclei per PCE, as a PCE may contain more than one micronucleus [ 3 ]. Systems for automated analysis (image analysis and cell suspensions fl ow cytometry) are acceptable alternatives to manual evaluation if appropriately justifi ed and validated. The number of cells evaluated could be much higher, up to 100,000 total erythrocytes, and the proportions of PCE, NCE, MNPCE, and MNNCE (micronucleated normocromatic erythrocytes) may be quantifi ed.

3.5 Data Individual animal data should be presented in tabular form. When Presentation the analyzed number of PCEs is 2,000/animal for at least 5 animals, or 1,000 PCEs for 8–10 animals, the animal–animal variability is lower than the sampling variability for the number of micronucleated cells. Therefore, presenting the data of all animals within the same treatment group has been a common practice [22 ]. When females and males are tested using the same treatment protocol and the same doses, the data of both sexes may be combined for statistical analysis, if there is no evidence of a difference in response between the sexes. The PCE number analyzed, the MNPCE number and percentage of MNPCE, as well as the PCE/ NCE ratio, should be presented, by gender, by sample time, and treatment group. Table 2 shows a suggestion for data presentation. The In Vivo Rodent Micronucleus Test 111

Table 3 Number of micronucleated polychromatic erythrocytes (MNPCE) observed in the bone marrow cells of male (M) Swiss mice treated with the test agent, and respective controls

Number of MNPCE per animal MNPCE PCE/NCE

Treatments M1 M2 M3 M4 M5 M6 mean ± SD mean ± SD

Control 7 9 8 6 7 8 7.50 ± 1.04 1.26 ± 0.08 Test agent ( dose 1 ) 9 6 7 10 11 11 9.0 ± 2.09 1.20 ± 0.08 Test agent ( dose 2 ) 17 16 19 12 21 17 17.0 ± 3.03 a 1.23 ± 0.13 Test agent ( dose 3 ) 19 17 16 17 16 19 17.33 ± 1.3 a 1.25 ± 0.08 Cyclophosphamide 27 29 26 29 27 16 25.66 ± 4.8 a 1.20 ± 0.06 (CPA) (50 mg/kg) Two thousand cells were analyzed. SDM standard deviation of the mean a Signiﬁ cantly different from negative control (ANOVA analysis of variance followed by the Tukey multiple comparison test, p < 0.001)

Other alternative for data presentation is presented in the Table 3 , when each animal data is presented and then the average and standard deviation for each group is calculated and compared with the control group.

3.6 Statistical In the data evaluation, both statistical and biological criteria are Analysis and Data considered. Interpretation Numerical data are analyzed using appropriate statistical tests, e.g., MNPCE data could be analyzed with a one-sided test for an increase and PCE data may also be analyzed with a one-sided test for a decrease [ 23 , 24 ]. If there is no evidence for a difference in response between sexes, the data from both sexes may be combined for statistical analysis. Indeed, a variety of tests have been used for statistical analysis and there is not one preferably indicated. The criteria to determine a result as positive should include a dose-related increase in the frequency of MNPCE, or a clear increase in the frequency of MNPCE at the highest dose, as a result of chromosomal damage or damage to the mitotic apparatus in the erythroblasts of the test species. However, if a statistically signiﬁ cant increase in the number of MNPCE is within the range of variation of historical control data from the laboratory, and if all animals show an incidence of MNPCE within the normal range of variation of the historical control, the test substance cannot be considered positive. A negative result indicates that, under the test conditions, the test substance does not produce a signiﬁ cant increase in MNPCE in the species tested. 112 Edson Luis Maistro

Often, when the data from animals of the same treatment group are put together, the frequency of MNPCE in treated groups can be analyzed using the Cochran–Armitage test to evaluate an increase in the response, depending on the treatment (this is only a suggestion). Furthermore, each test group treated is compared with the negative control group, using the Chi-Square test to identify a positive response.

Acknowledgements

I am grateful to Eduardo de Souza Marques for his help in preparing the micronucleus origin schema shown in Fig. 1 .

References

1. Yunis JJ (1983) The chromosomal basis of marrow micronucleus test. Mutat Res 347: human neoplasia. Science 221:227–236 97–99 2. Hagmar L, Bonassi S, Stromberg U et al 11. OECD (Organization for Economic (1998) Chromosomal aberration in lympho- Co-operation and Development) (1997) cytes predict human cancer: a report from the Mammalian Erythrocyte Micronucleus Test. European Study Group on Cytogenetic Bio- Guideline for the testing of chemicals 474. markers and Health (ESCH). Cancer Res 58: Paris, France 4117–4121 12. Schlegel R, MacGregor JT (1984) The persis- 3. Krishna G, Hayashi M (2000) In vivo rodent tence of micronucleated erythrocytes in the micronucleus assay: protocol, conduct and data peripheral circulation of normal and spenecto- interpretation. Mutat Res 455:155–166 mized Fischer 344 rats: implication for cytoge- 4. Boller K, Schmid W (1970) Chemische netic screening. Mutat Res 127:169–174 Mutagenese beim Säunger. Das Knochenmark 13. Abramsson-Zetterberg L, Grawé J, Zetterberg des Chinesischen Hamsters als in vivo-Testsystem. G (1999) The micronucleus test in rat erythro- Hämatologische Befunde nach Behandlung mit cytes from bone marrow, spleen and peripheral Trenimon. Humangenetik II:35–54 blood: the response to low doses of ionizing 5. Matter B, Schmid W (1971) Trenimon- radiation, cyclophosphsmide and vincristine induced chromosomal damage in bone marrow determined by ﬂ ow cytometry. Mutat Res 423: cells of six mammalian species, evaluated by the 113–124 micronucleus test. Mutat Res 12:417–425 14. Gollapudi B, Kamra OP (1979) Application of 6. Schmid W, Arakaki DT, Breslau NA et al a simple Giemsa-staining method in the micro- (1971) Chemical mutagenesis. The Chinese nucleus test. Mutat Res 64:45–46 hamster bone marrow as an in vivo test system. 15. Hayashi M, Morita T, Kodama Y et al (1990) I. Cytogenetic results on basic aspects of the The micronucleus assay with mouse peripheral methodology, obtained with alkylating agents. blood reticulocytes using acridine orange- Humangenetk II:103–118 coated slides. Mutat Res 245:245–249 7. Schmid W (1973) Chemical mutagen testing 16. CSGMT (Collaborative Study Group for the on in vivo somatic mammalian cells. Agents Micronucleus Test) (1992) Micronucleus test Actions 3:77–85 with mouse peripheral blood erythrocytes by 8. Schmid W (1975) The micronucleus test. acridine orange supravital staining: The sum- Mutat Res 31:9–15 mary report of the 5th collaborative study by 9. Miller RC (1973) The micronucleus test as an CSGMT/JEMS.MSM. Mutat Res 278:83–98 in vivo cytogenetic method. Environ Health 17. CSGMT (Collaborative Study Group for the Perspect 6:167–170 Micronucleus Test) (1995) Protocol recom- 10. Gollapudi BB, McFadden LG (1995) Sample mended for the short-term mouse peripheral size for the estimation of polychromatic to nor- blood micronucleus test. Mutagenesis 10: mochromatic erythrocyte ratio in the bone 153–159 The In Vivo Rodent Micronucleus Test 113

18. Hayashi M, Sofuni T, Ishidate-Jr M (1983) An marrow micronucleus assay. In: Ashby J, application of acridine orange fl uorescent staining deSerres FJ, Shelby MD, Margolin BH, to the micronucleus test. Mutat Res 120:241–247 Ishidate-Jr M, Becking (eds) Evaluation of 19. Macgregor JT, Wehr CM, Langlois RG (1983) short-term tests for carcinogens. Report of the A simple fl uorescent staining procedure for international programme on chemical safety’s micronuclei and RNA in erythrocytes using collaborative study on in vivo assay. Oxford Hoechst 33258 and Pyronin Y. Mutat Res University Press, Oxford, pp 29–44 120:269–275 23. Richold M, Ashby J, Bootman J et al (1990) In 20. Krishna G, Fiedler R, Theiss JC (1992) Simultaneous vivo cytogenetic assays. In: Kirkland DJ (ed) evaluation of clastogenicity, aneugenicity, and Basic mutagenicity tests, UKEMS recom- toxicity in the mouse micronucleus assay using mended procedures. UKEMS subcommittee immunofl uorescence. Mutat Res 282:159–304 on guidelines for mutagenicity testing. Report. 21. Miller BM, Zitzelsberger HF, Weier HU et al Part I revised. Cambridge University Press, (1991) Classifi cation of micronuclei in mouse Cambridge, pp 115–141 erythrocytes: immunofl uorescent staining using 24. Lovell DP, Anderson D, Albanese R et al CREST antibodies compared to in situ hybrid- (1989) Statistical analysis of in vivo cytogenetic ization with biotinylated gamma satellite DNA. assays. In: Kirkland DJ (ed) Statistical evalua- Mutagenesis 6:297–302 tion of mutagenicity test data. UKEMS sub- 22. Margolin BH, Risko KJ (1998) The statistical committee on guidelines for mutagenicity analysis of in vivo genotoxicity data. Case studies testing. Report. Part III. Cambridge University of the rat hepatocyte UDS and mouse bone Press, Cambridge, pp 184–232 Chapter 7

Chromosomal Aberration Test Utilities In Vitro and In Vivo

Ana Paula A. Guimarães , Adriana C. Guimarães , Diego Á. Alcântara , Luiz Raimundo Cunha , Patrícia L. Lima , Marne C. Vasconcellos , Raquel C. Montenegro , Bruno M. Soares , Marucia M. Amorim , and Rommel R. Burbano

Abstract

Human populations are frequently exposed to several mutagenic agents that have the potential to damage the DNA, and this, in many cases, may result in the formation of chromosomal aberrations (CAs). CAs are recognized as an important biomarker of human exposure, being a very important tool for environmental biomonitoring. Although there are several types, little is known about the mechanisms involved in the processing of induced lesions in DNA and how these could result in CAs. Thus, cytogenetics and molecular cytogenetics are tools of great importance for identifying these agents, the conditions that can exercise their mutagenic potential, and their action mechanism. This chapter discusses the history of CA formation and some cytogenetic protocols that may be used to perform the chromosomal aberration test in in vivo and in vitro studies.

Key words Biomonitoring , Chromosomal aberrations , DNA damage , Cytotoxicity , Mutagenesis

1 Introduction

1.1 Origin DNA damage, formed spontaneously or induced by various of Chromosomal chemical and physical agents, can be repaired or processed, but Aberrations: Uses many of them can lead to the formation of numerical and struc- In Vitro and In Vivo tural chromosomal aberrations (CAs). Although there are several studies and technological advances, little is known about the mechanisms involved in the processing of induced lesions in DNA and how these could result in chromosomal aberrations. However, it is known that such changes may indicate changes in cellular homeostasis, which are important in genomic instability and crucial to understanding the processes of mutagenesis and carcinogenesis [1 ]. In general, there are numerical and structural chromosomal aberrations, which are both associated with human health, such as congenital anomalies in newborns and cancer [ 2 ].

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_7, © Springer Science+Business Media New York 2014 115 116 Ana Paula A. Guimarães et al.

The main type of numerical CA in humans is represented by aneuploidy, which is usually a change in dosing of a single chromosome which has one extra copy (trisomy) or one copy less (mono- somy), being the latter less compatible with human survival. Most aneuploidy originates by a chromosome nondisjunction in one of the meiosis divisions or during mitosis. This failure can occur acci- dentally, but it can also be caused by advancing age, especially in females, and the exposure to aneugenic agents which interfere with the formation of the spindle fi bers [3 ]. The chromosomal aberrations are recognized as an important biomarker of the human exposure to ionizing radiation and genotoxic chemicals, being a very important tool for the environmental biomonitoring [ 2 ]. Several studies have a purpose to evaluate the environmental exposition of populations to several elements that are possibly genotoxic. The CA can be induced by chemical and physical agents, or may arise during the process of DNA repair, and this is highly infl u- enced by the chromatin structure and the transcriptional activity [4 ]. According to Palitti [ 5 ], the induction of damage in DNA and its processing by repair enzymes are affected by different levels of chromatin organization. The breaks in DNA produced by mutagens may involve single-strand and double-strand breaks (SSB and DSB, respectively). It is believed that DSB may be the main damage resulting directly in various types of CA, which can be detected in the fi rst mitotic division, subsequent to the exposure of the cells to the mutagenic compound [6 ]. The chemical and physical agents capable of inducing the formation of chromosomal aberrations are called clastogenic agents. The clastogenicity may result from the direct interaction of these agents with the DNA or the indirect mechanisms which interfere with the replication and repair of the molecule. The chemical agents that induce chromosomal damage include alkylating agents, intercalators, DNA repair inhibitors, and other substances, while the physical agents can be represented by radiations [ 7 ]. Clastogenic agents may have different mechanisms of action. They may be divided into S-phase-dependent agents, which induce lesions that require a period of DNA synthesis (S phase) for the aberrations to be produced, such as ultraviolet (UV) light, alkylating agents, and most chemicals, and the S-phase-independent agents, which induce aberrations in all phases of the cell cycle, such as for instance ionizing radiation and bleomycin [ 8 , 9 ]. Ionizing radiations, UV light, alkylating agents, and other chemical mutagenic agents are effi cient inducers of CA in vivo and in vitro, as they are capable of causing DNA damage responsible for the formation of structural chromosomal alterations [10 ]. According to Natarajan [ 11 ], these lesions can be of various types depending on the agent used: breaks in the DNA strands (X-ray, UV light, bleomycin); DNA crosslinks (polyfunctional alkylating agents), Chromosomal Aberration Test 117 pyrimidine dimers (UV light), apurinic and apyrimidinic sites (monofunctional alkylating agents); interpolations, among others. The various studies conducted with chemical agents present many evidences that these agents and/or their metabolites can induce CA by secondary mechanisms, such as the inhibition of the DNA synthesis [12 , 13 ]. Studies involving structural chromosomal aberrations have been developed from the beginning of the last century to try to elucidate their origin and their consequences in different organisms. The fi rst studies describing the process of CA formation were performed by Perthes [ 14 ], using irradiated oocytes of Ascaris , and in the same year Koernicke studied the chromosomal aberrations using irradiated root cells of Vicia faba and Pisum sativum [15 ]. De Vries [16 ], studying plant chromosomes, noted the presence of translocations, and in the chromosomes of the fl ies of the genus Drosophila , he noted several paracentric inversions, coming to the conclusion that the spontaneous chromosomal aberrations play an important role in organism evolution. Muller [17 ] was the fi rst to use fl ies of the genus Drosophila to describe the relationship between chromosomal aberrations and X-rays. The use of these fl ies, as a research model in this fi eld, is quite effective because in some cells they have polytene chromosomes, whose large size and well-defi ned banding patterns make it possible to determine the complexity of the rearrangements and the position of the breaks involved in the changes induced by X-rays. Scott [18 ] and Evans [19 ] demonstrated that X-rays are able to induce chromosomal aberrations in the root cells of Vicia faba , highlighting that this radiation is able to induce aberrations at different stages of the cell cycle. Most of these studies were conducted in an era when almost nothing was known about the different pathways to repair DNA damage. Later, with the advances in techniques, it was observed that cells irradiated with X-rays and post-treated with Neurospora endonuclease show a signifi cant increase in the frequencies of CA. The interpretation of this fact was that X-rays induce SSB and these would be converted into DSB by the endonuclease, resulting in increased chromosomal damage [ 11 ], reinforcing the idea that the DSBs in the DNA molecule are the most important lesions that lead to the formation of CA. Other studies have shown that mutant cell lines with repair defi ciency of double-strand breaks are more sensitive to ionizing radiation, with an increase of about three to fi ve times in the incidence of aberrations in these cells, compared to wild-type cells, after the X-irradiation [20 ]. These are some examples of how knowledge of chromosomal aberrations, both from in vitro and in vivo studies, is important to analyze exposures to various environmental conditions. Amorim et al. [21 ] evaluated the environmental exposure to mercury in an Amazonian region, where the activity of gold mining using mercury in the 118 Ana Paula A. Guimarães et al.

process is very widespread. Studies in the Tapajos region, a major affl uent of the Amazon River show that mercury is present in all environmental conditions (soil, water, and plant). For the biomonitoring of this population, a group of people was selected for the collection of blood to perform the chromosomal aberration test. Natarajan et al. [ 22 ] performed a study accompanying victims of the Goiania accident occurred with Cesium 137 in 1987 from translocation data obtained by the FISH technique. Guimarães et al. [ 23 ] studied the exposure of CHO-k1 and XRS-5 cell lines (mutant cells defi cient for double-strand break rejoining) to verify the interaction effects of the VP-16 and 5 azaC drugs in terms of induction of chromosomal aberrations. In K1 cells, the combined treatments induced a signifi cant reduction in the frequency of induced aberrations; however, in XRS-5 cells, the drug combination caused a signifi cant increase in the frequency of induced aberrations, but with a concomitant reduction in the randomly induced aberrations. The use of mutant cells with distinct cytoplasmic and nuclear organization and with impaired patterns of DNA repair is very useful to identify the different types of induced biological effects.

2 Chromosomal Aberrations In Vitro

2.1 Chromosomal Various regulatory agencies such as the FDA (U.S. Food and Drug Aberrations Administration), KFDA (Korea Food and Drug Administration) in Cultured Cells [ 24 ], and ANVISA (Agência Nacional de Vigilância Sanitária— National Health Surveillance Agency—Brazil) suggested a review to determine genetic risk through tests that have been developed, validated, and integrated in international policies such as the OECD (Organization for Economic Cooperation and Development) [25 ]. In this sense, the evaluation of chromosomal aberrations follows the OECD 473 and the recommendations of the reports of the International Workshops on Genotoxicity Testing (IWGT) [26 ]. Genotoxicity assays, performed in vitro and in vivo, are designed to detect compounds that induce DNA damage. The test for chromosomal aberrations in vitro, using mammalian cells, has the purpose of identifying causative agents of structural changes in chromosomes [ 27 –29 ]. These assays increase the sensitivity for detection of carcinogens. However, the increased incidence of positive results is not always correlated with the processes of carcinogenesis, since a single test is not able to detect all relevant mechanisms in the development of tumors. There is therefore a need to carry out further tests in parallel, such as the bacterial reverse mutation test (Ames test), the micronucleus test, and/or the test for gene mutation in L5178Y TK +/− cells from mouse lymphoma (MLA). Furthermore, in vivo testing must be included in this battery of tests, since some agents are mutagenic in vivo but not in vitro [24 ]. Chromosomal Aberration Test 119

Thus, the test for chromosomal aberrations in cultured mammalian cells is an important tool in safety assessment of compounds that could be used as drugs, cosmetics, and/or food additives.

2.1.1 Cells The selection of the appropriate cell line to be used in the in vitro chromosomal aberration test depends on the stability of the karyotype, the number and the diversity of the chromosomes, and the frequency of the spontaneous chromosomal aberrations. Furthermore, the status of the P53 gene must be considered, since the cell lines that have an impairment of this gene overestimate the genotoxic potential of a given agent [ 25 , 26 ]. The cell lines must be maintained in an exponential growth phase. For this, it is necessary to culture cells in a suitable medium and under appropri-

ate incubation conditions (culture ﬂ asks, CO 2 concentration of 5 %, at 37 °C and humidity of 85 %). The cultures must be routinely monitored so that the modal number of chromosomes is established and the absence of contamination by mycoplasma is guaranteed [25 , 30 ].

2.1.2 Study Design For the performance of chromosomal aberrations test in cell lines, it is necessary to be sure that the cells in culture suffered division during and after the treatment with the test substance. Cytotoxicity assays can be conducted to ensure that the treated cells passed through the cell division process during the genotoxicity testing. The cytotoxic potential of the test agent must be determined with and without metabolic activation, using an appropriate integrity and cell proliferation indicator, and may not exceed a reduction of about 50 % in cell growth. These assays can help determine the concentration of the test substance to be used in the genotoxicity evaluation [31 – 33 ]. The treatment with the test substance must be of 3–6 h (pulse treatment), with and without metabolic activation. Parallel to the treatment, a negative control, which consists only on the vehicle in the culture medium, must be made and must be handled in the same way as the treated cultures. Positive controls are also necessary because they demonstrate the ability of the cell to identify clastogenic agents. Some recommended positive controls, such as benzo(a)pyrene and cyclophosphamide, are used for testing with exogenous metabolic activation, while others, like methylmethane- sulphonate, mitomycin C, and 4-nitroquinoline- N -oxide, do not require such activation [24 , 25 ].

Procedure The number of concentrations used for the treatment of cultures must be enough to provide conﬁ dence in the evaluation. It is recommended that the cultures be made in duplicate, at least for the negative control [ 25 ]. 120 Ana Paula A. Guimarães et al.

To proceed with the in vitro chromosomal aberration test, using the method without metabolic activation, the cells (5,000 cells/mL) are seeded in culture plates (60 mm), in 5 mL of medium and incubated for a period of 24 and 48 h. Treatments with different concentrations of the test agent and the positive control (e.g., mitomycin C in concentrations of 0.00005 and 0.0001 mg/mL) are added to the plates. The cultures are exposed to the pulse treatment, for a period of 3–6 h, or to continuous treatment, for 24 and 48 h. To block the cell division in metaphase, 100 μL of colchicine 10 μg/mL are added to the culture 2 h prior to cell collection. After this period, the cells from each culture are harvested separately using 0.05 % of tripsin/0.53 mM EDTA to detach the cells from the culture ﬂ ask. The collected cells are immersed in a hypotonic solution, 75 mM KCl, for 30 min at 37 °C. After that, the cells are harvested by centrifugation (550 × g ), the supernatant is removed and the cells are ﬁ xed with cold Carnoy’s (Acetic acid + methanol, 3:1). The cell suspension is then spread on heated slides and, after drying, staining is performed using Giemsa dye. The karyotype can be analyzed using G-banding techniques [34 , 35 ]. For the metabolic activation, the enzyme mixture is added 3 days after the start of the culture, allowing a greater number of treated cells, in comparison with the experiments without metabolic activation. The cells are treated with the positive and negative controls and the test agent for a period of 3 or 6 h. After the exposure time to the treatment, the cells are washed in brine and then recultured with fresh culture medium for 18 h. This procedure allows the removal of the test agent and its metabolites and also the enzyme mixture. In general, the experiments with metabolic activation are wrist to resemble the in vivo experiments, where test agent is eliminated after metabolization. The preparations to obtain the metaphase chromosomes are performed as described previously [14 ], and as indicated below for human lymphocytes.

2.2 Chromosomal Cytogenetic studies are of great importance for the identiﬁ cation Aberrations in Human of the numerical and structural chromosomal aberrations. These Lymphocytes changes are associated with several genetic syndromes, and with the effects of genomic exposure to chemical and physical mutagenic agents, capable of producing diseases like cancer. The obtainment of the chromosomes from the advent of cytogenetic techniques allowed the establishment of the human chromosome set, considering the morphology and the number of chromosomes per diploid cell, and the examination of the karyotype, which consists in organizing the chromosome set of an individual, in groups (A, B, C, D, E, F, G) from the observation of the position of the centromere and the size of each chromosome, to investigate possible changes. The establishment of the technical culture of human lymphocytes for the analysis of chromosomal aberrations played a crucial role, being quite used to this day, to Chromosomal Aberration Test 121

analyze the human chromosomes; and it is well accepted as a fundamental tool to study populations exposed to environmental and occupational mutagenic agents (biomonitoring), and to test agents to check their mutagenic potential, by using cultured lymphocytes collected from healthy individuals and exposing them in vitro.

2.2.1 History The fi rst ideas on chromosomes arose in the late nineteenth century, when the fi rst studies on mitosis were performed. In the 1950s there was the advent of the techniques for chromosome preparations, such as the addition of colcemid and the hypotonic treatment, thus leading to the obtaining of metaphase chromosomes and the establishment of the diploid number of chromosomes in man as 46 [36 ]. However, it was in 1960 that the lymphocyte culture technique of human peripheral blood was developed, when Moorhead et al. [37 ] observed that these lymphocytes which normally do not divide, lying in the G0 phase of the cell cycle, could be stimulated to divide in culture when a protein substance, phytohemagglutinin (PHA), derived from the black beans ( Phaseolus vulgaris ), was added to the culture. This substance reacts with N -acetyl-d - galactosamine and stimulates a broad spectrum of T lymphocytes (responsible for cell immunity), transforming them into blast cells. In the period from 48 h after the addition of PHA the lymphocyte undergoes a nuclear and cytoplasmic cell volume transformation about fi ve times the initial volume. This technique has as one of its main advantages the easy obtaining of lymphocytes by making the collection of blood from a venipuncture. After the discoveries of Moorhead et al. [ 37 ], several techniques have emerged as complementary steps for the lymphocyte culture technique for obtaining the metaphases, in order to enhance the details of chromosome analysis. So in the late 1960s, Carpersson et al. [ 38 ] observed that when the metaphases were stained with quinacrine mustard and examined by fl uorescence microscopy, a series of fl uorescent bands (Q bands) were seen, being each particular pattern specifi c for each chromosome pair. After the Q-banding, some other banding techniques have been developed such as G, R, and Nor each with its own and specifi c applications. Besides the banding techniques, there are others that can be employed, as for example, the viewing of sister chromatid exchange (SCE) from the treatment of cell cultures with 5-BrdU for two cycles, allowing the differentiation between the subsequent staining of the sister chromatids [ 39 ] as a result of semiconservative replication of DNA; or the association with the FISH technique, thus performing molecular cytogenetics with the study of specifi c chromosome segments, or an entire chromosome, resulting in data of specifi c chromosomal and translocations losses [40 ]. 122 Ana Paula A. Guimarães et al.

Since at metaphase chromosomes are more condensed and individualized, the metaphase chromosomes are those used for their classifi cation, according to Denver [ 41 ]. The human genome thus comprises 46 chromosomes, 44 autosomes, and two sex chromosomes, morphologically classifi ed according to the position of the centromere as metacentric (centromere centrally located), acrocentric (centromere near the end), and submetacentric (centromere in intermediate position). Once the correct description of human chromosomes with well-established number was made, the chromosomal abnormalities associated with specifi c birth defects, or with environmental exposures, could be analyzed, specially because the development of the human lymphocytes culture technique made easy the obtaining of human material to perform cytogenetic analysis.

2.2.2 The Use Various tissues can be used for the preparation of chromosomes, of Lymphocytes to Obtain and the blood is the tissue most frequently used for the cytogenetic Human Chromosomes analysis, since it is obtained easily from individuals of all ages. Lymphocytes have a recirculation time that takes on average 12 h, when they pass through the spleen, lymph nodes, and other tissues, returning to the circulation via the lymphatic ducts. This means that the lymphocytes with mutations in any part of the body will be sometime in the peripheral blood, allowing the detection of the mutations, both in human peripheral blood lymphocytes, and in those distributed in different organs [ 42 ]. In this context, the culture of lymphocytes from peripheral blood is a critical step in obtaining quality chromosome preparations. Most human peripheral lymphocytes are in G0 of the cell cycle stage and can be stimulated in vitro to develop mitotic divisions by mitogenic substances (PHA) that induce the DNA synthesis and the cell division, by binding to glycoprotein receptors of the lymphocyte surface without requiring prior sensitization [43 ].

2.2.3 Protocol According to the Moohead et al. [37 ] technique, with modiﬁ cations and adjustments, the following protocol can be perform to analyze chromosomal aberrations in human lymphocytes: 1. In heparinized disposable syringes, 5–10 mL of blood is collected. 2. The blood is transferred into vials and the material is placed at room temperature for depositing. 3. Approximately 1 mL of plasma with lymphocytes is added to 5 mL of complete culture medium (supplemented with 20 % of FBS). Alternatively, the lymphocytes can be isolated with Ficoll-Pack and seeded in culture medium afterwards. Chromosomal Aberration Test 123

4. 200 μL of PHA is added to reach a 1 % ﬁ nal concentration in the total volume of 5 mL complete culture medium.

5. The culture(s) is incubated in the CO2 incubator for 72 h. 6. Approximately 90 min before the end of the incubation 20 μL of colchicine (1 μg/mL) is added. 7. At the end of 72 h, each culture is transferred to centrifuge tubes. 8. Centrifuge at 550 × g for 5 min. 9. Remove the supernatant and gently resuspend the pellet with a Pasteur pipette in 0.5 mL left supernatant. 10. Add 5 mL of hypotonic solution of KCl (0.075 M) heated at 37 ºC and gently resuspend. Note: The hypotonic treatment with KCL causes a swelling in the cells through the process of osmosis , providing the dispersion of the chromosomes . 11. Centrifuge at 550 × g for 5 min. 12. Remove the supernatant and gently resuspend the pellet in 0.5 mL left supernatant. 13. Add 5 mL of Carnoy’s fi xative and gently resuspend the material. Note: In the Carnoy ’ s fi xative , the methanol component acts denaturing and precipitating proteins by dehydration and under acidic conditions. The acetic acid coagulates the nucleo- protein and causes swelling in the cells , previously contracted by dehydration and shrinkage from the action of the methanol. This fi xative easily penetrates the cell and preserves the structure of the chromosomes . 14. Centrifuge at 550 × g for 5 min. 15. Repeat the steps 12–14 two more times. 16. Remove the supernatant leaving approximately 0.5 mL of the pellet for slide preparation. 17. Place a wet slide in tilted position and distribute three drops of the material leaving it spread on the slide. Note: The slides must be well washed with soap and water , even if they are new , and stored in ice -cold distilled water. With a forceps , each slide is removed and tilted to receive the drops of the suspension that , because of the position of the slides and the contact with the water , slip and spread . 18. Pass the slide quickly in an alcohol lamp fl ame and allow it to dry at room temperature. Note : For drying the slides , both the fl ame of an alcohol lamp , over where the slides are passed quickly , or the air drying are acceptable . 124 Ana Paula A. Guimarães et al.

19. Add 3 mL of Giemsa, diluted (1:30) in 0.06 M phosphate buffer pH 6.8, throughout the surface of each slide, place horizontally. Leave the stain to act for 5 min, then rinse it in running water. 20. Allow the slide to dry at room temperature and view under a light microscope using the 100× lens. Most laboratories prefer to perform the culture of 72 h for routine samples because of the high number of metaphases and better quality of the material obtained. However, there is great interest in optimizing the execution time of the procedure, to reduce the waiting time for the diagnosis, performing cultures of 48 h, already used in several laboratories [44 ].

2.2.4 Treatment The treatment protocol will vary with the objectives of the work, with Test Substance that is, when it is desired to perform the continuous treatment of the lymphocytes with the test substance. The chemical solutions (at different concentrations) are added to the 5 mL of culture medium with lymphocytes about 6 h after the incubation start, when the cells in G0 have been stimulated by PHA. To check the chemical effect in the G1, S, and G2 phases of the cell cycle, the chemical solution can be added in pulses of 30–60 min to the culture medium, after 10 h of incubation for the G1 phase, 24 h for the S phase and 48 or 69 h for the G2 phase. In the case of G1, S, and G2-48 h phase analysis, it is necessary centrifuging the cultures and replacing the medium two or three times, in order to totally removed the test chemical. In the case of treatment in G2-69 h the test chemical could be left in the culture until the 72 h are completed, whereupon the harvest [ 45 ] must be performed. As already mentioned, it is important that the study with the test chemical of interest contains, besides the cultures exposed at different times and/or concentrations of the agent, the cultures which correspond to the positive and negative controls.

3 Molecular Cytogenetic Techniques

The advances in molecular cytogenetic techniques improved the chromosomal diagnosis of diseases, both in the early stages of the disease, as in the embryonic stages of development in prenatal monitoring, and facilitated the experimental study in the investigation of chromosomal aberrations in vitro and in vivo, besides being currently very useful in the research of chromosomal changes in populations exposed to the ionizing radiation or other environmental agents (chemical, physical, or biological) . The FISH Method (Fluorescent in situ Hybridization ) is the result of a series of changes in the hybridization techniques, since the late 1970s to the early 1980s. Currently, it consists of using Chromosomal Aberration Test 125 short sequences of DNA, called probes, labeled with fl uorochromes (fl uorescent dyes) which have analogy to the genomic DNA, so that these probes hybridize to the target sequence in the metaphase or interphase chromosome or microarrays ( arrays ) DNA, emitting fl uorescence. The fl uorescence can be detected by fl uorescence microscope, except in the case of microarray hybridization, where digital image analysis is used [ 38 , 46 – 48 ]. The FISH Method is ideal for viewing and analyzing the genetic disorders that affect one or a few genes or chromosomal sub-regions, both in the metaphase nuclei as in the interphase nuclei. With this technique it is possible to detect chromosomal aberrations, which by classical cytogenetics is not possible to do accurately (or it depends on a person with much experience in the analysis), as small deletions and chromosomal rearrangements [49 ]. The samples to be analyzed may be from suspension cells (such as the peripheral blood lymphocytes or the bone marrow cells), or cells from solid tissues (such as the biopsies of healthy tissue or solid tumors), so that it is feasible to mount on slides for the hybridization. In the cases of suspension cells, the material is easily fi xed on slides and the procedures for the hybridization and analysis proceed thereon. On the other hand, when working with cells in solid tissue, embedded in paraffi n on the slide, it is also possible the development of the whole procedure for the hybridization if prior deparaffi nization [50 , 51 ] is performed. There are various types of DNA probes, that detect regions from less than 1 to 50 kb, which are used for the FISH Method, such as: the probe for the entire chromosome, known as “chromosome painting”; the probe for a chromosome arm called “partial chromosome painting”; the centromeric probe for chromosome centromeres; the telomeric probe for chromosome telomeres, and the locus-specifi c probe for an specifi c gene or chromosomal sub- region. Combined with the probes, a variety of binders or fl uorochromes, which allows the simultaneous analysis of more than one target sequence, can also be found on the market [40 , 49 ]. As previously mentioned, over the years there have been technological advances which allowed improvements and variations of the FISH technique, keeping the principle of the hybridization. The techniques available today are SKY, mFISH, COBRA-FISH, and aCGH [49 , 52 ]. The SKY (Spectral Karyotyping ) technique is a chromosome analysis tool based on the principles of the FISH technique and the difference is that it uses combinations in pairs of fi ve fl uorochromes, allowing the painting of each of the 24 different human chromosomes (22 autosomes, X and Y sex-chromosomes) with different colors. It is very useful for the diagnosis of a variety of congenital disorders such as trisomies, abnormalities associated with sex chromosomes, and chromosomal abnormalities associated with malignant tumors. It shows some advantages compared to the classical 126 Ana Paula A. Guimarães et al.

cytogenetics, such as the identifi cation of derivative chromosomes, complex translocations, and gene amplifi cation. On the other hand, it has few advantages or disadvantages involving the analysis of inversions, deletions, and numerical aberrations, when compared to the classic method analysis [53 , 54 ]. The mFISH (multicolor-FISH) was performed fi rst by Speicher et al. [ 55 ], which developed epifl uorescent types of fi lters and a computer program for detection and discrimination of up to 27 different DNA probes simultaneously hybridized. Similar to the SKY technique, it is also performed by the combination of fl uorochromes, and it is extremely useful to detect translocations and complex changes, as observed in the tumor cells [ 53 , 55 ]. The COBRA-FISH (Combined Binary Ratio -labeling FISH ) is a multicolored FISH methodology which allows the recognition of all the chromosomes of a genome based on the combination of colors. The principle of the technique is based on a probe composed of a mixture of different probes with different fl uorescent markings, usually well distinguishable fl uorescences like blue, red, and green. The number of separated fl uorophores is used for the proportional marking, so that two fl uorophores are used to produce a certain color grading. When this is applied to three fl uorophores and each pair of fl uorophores results in fi ve graduated colors (including the original colors), a total of twelve colors are generated. A second type of probe with twelve colors recognizes different targets and has the addition of a hapten (biotin or digoxi- genin), that binds to a fourth fl uorophore, then the type 12 is multiplied by two, resulting in 24 colors using four fl uorophores. Extra fl uorophores can be used to repeat this process by exploiting a second hapten, resulting in double viable colors (48 colors). However, in practice, the proportional marking is considered complex. The identifi cation of the chromosomes stained with proportional marking probes is not simply a matter of “yes or no,” but it requires accurate measurement of color [ 56 ]. In the technique of aCGH (array -Comparative Genomic Hybridization ) the equivalent amounts of a test sample containing DNA, marked with Cyanine 5 (Cy5), and a reference sample, marked with Cyanine 3 (Cy3), are co-hybridized on the array . This mix is transferred to a microarray where images will be scanned. The intense points are measured and the image fi les are quantifi ed using an extraction program, since the text fi les of the quantitative analyses are imported into a program of copy number analysis. The ratio of the resulting fl uorescence intensities is proportional to the ratio between the numbers of copies of DNA sequences in the test and reference genome. If the intensities of the fl uorescent colors on a probe are equal, this region of the genome of the patient is interpreted as having the equivalent amounts of DNA in the test and reference samples, but if there is a change in the ratio Cy3: Cy5 this indicates a loss or a gain of an individual’s Chromosomal Aberration Test 127

DNA in a specifi c region of the genome. This tool can be applied to studies where changes are below the resolution of the conventional cytogenetic methods, such as submicroscopic critical changes involving the number of copies (deletions, duplications, or both changes) of a locus or loci [48 , 49 ]. Fluorescence hybridization can also be used in association with other cytogenetic techniques, such as in assays of micronuclei (MN) in binucleated cells (see Chap. 5 of this book), as in the work of Guimarães et al. [57 ], which assessed clastogenic effects in exposed populations by using pan-centromeric probes in MN in binucleated cells, and two specifi c probes in the FISH Method to check the presence of the IGH/BCL2 translocation in interphase nuclei. Summary of the FISH Method, according to Min [58 ]: 1. Spread the slide At this early stage, the preparation of the slide is done in the same way as for the classical cytogenetics. The important thing is to have the least possible cytoplasmic waste or extracellular matrix, to avoid diffi culties in the hybridizations and hence interference in the analyses. The sample can be obtained from metaphases, but for many analyses it is not necessary to obtain cell division, only the use of hypotonic solution, to fi x and perform the washes (three times) with the Carnoy’s fi xative. Note : It is critical for molecular cytogenetics that the slides are cleaned for using ; therefore they can be washed in ethanol and stored at − 20 °C prior to use . In cells in suspension, around 15 μL or 3–4 drops can be used, so that a coverslip (22 × 22 mm) covers the area of the sample. A diamond marker can be used to defi ne the area. A phase contrast microscope is useful for ensuring that chromosomes or interphase nucleus are spread out, with good contrast and that there is little cytoplasm. If necessary, a pretreatment of the slide using RNase before the hybridization can be done. 1.1. Pretreatment of the slide Three methods can be performed depending on the situation. If the slides are discolored or stored for more than 1 month, the fi rst or second method may be used. The fi rst one is with SSC (saline sodium citrate) and the alcohol series. The second is with the acetic acid series followed by the alcohol series, both with decreasing dilution in water. If there is much residual cytoplasm, the third method may be more appropriate, using RNase diluted in SSC, and applying it on the slide. It is incubated in a humid chamber and then washed in SSC. 2. Denature the target DNA on the slide For the denaturation of DNA on the slide one can proceed following two methods, using in both of them 60 % formamide in 2 × SSC (fi nal concentration), for it is useful in reducing the 128 Ana Paula A. Guimarães et al.

need for high temperature to separate double-stranded DNA. One method is by immersing the slides in a hot solution (70 °C) of formamide and SSC. Then, pass them into series of cold ethanol (70 %, 80 %, 100 %), and fi nally allow them to dry at room temperature, awaiting the next step (hybridization). The second method consists of placing the slide on a hot plate (about 72 °C), dripping the same formamide solution on the hot slide, then pass it through the battery of cold ethanol (70 %, 80 %, 100 %), and allow it to dry at room temperature for the start of the next step (hybridization). 3. Denature the probe To denature the probe, it is necessary to separate an aliquot of 10 μL in a microtube, expose it to a brief microfuge, and denature it at about 70 °C in a water bath . Immediately afterwards, dip the microtube on ice to prevent renaturation. Proceed with the pre-annealing, incubating the probe at 37 °C for up to 60 min. 4. Hybridize the probe to the target DNA Continuing from the previous steps, add the probe to the sample on the slide, cover it with coverslip, seal around the coverslip with rubber cement, and allow it to incubate at 37 °C in a humid chamber. Note : The incubation time may vary depending on the type of probe , so it is essential to observe the manufacturer ’s instructions 5. Wash to remove any unhybridized probe After the hybridization, the unbound probes are removed by a series of washes. These washes serve to remove unbound probes and only allow the presence of the probes bound to the target DNA, using a stringent solution of 60 % formamide in 2 × SSC (fi nal concentration). However, if there is a lot of unwanted hybridization, the stringency can then be increased by increasing the concentration of formamide in the hybridization buffer. 6. Counterstaining Finally, proceed with the propidium iodide counterstain (PI) or DAPI. PI is most often used as a contrast to yellow fl uorochromes and DAPI to red or green. Add it on the slide and cover it with coverslip. These slides can be stored at 4 °C (the time varies according to the manufacturer’s recommendation). The molecular cytogenetic analysis is performed with a fl uorescence microscope, with triple fi lter DAPI/FITC/ TRICT. It should be equipped with a system to capture images and analyze them, which can be purchased from commercial enterprises. It is crucial that the analysis is carried out on slides of good quality and with in situ hybridization exceeding 70 % (78). For each sample 200 interphase nuclei are analyzed. Chromosomal Aberration Test 129

4 Chromosomal Aberrations In Vivo: Human Biomonitoring Using Lymphocytes

Human populations may present genetic damage by accidental, occupational, or environmental exposure to chemical and physical genotoxic agents. These agents may interfere with the proper cell development, causing damage to their genetic material, representing high risk for the development of neoplasms [59 ]. The induced CA may be stable or not, and the fi rst refer to the minor damage, reciprocal translocations and some aneuploidy, which do not prevent cell division and proliferation, while the non-stable CA such as dicentric and ring chromosomes, large deletions and fragments, sometimes are lethal to the cell. Different changes can accumulate in successive cell divisions and produce mutations in genes that have a key role in the process of carcinogenesis [60 ]. Since the induced chromosomal aberrations result from the interaction of a given mutagen with the DNA, one can use such anomalies as indicators of damage in this molecule. In other words, by means of the chromosomal damage one can assess the mutagenic activity of these agents [61 , 62 ]. Clastogenicity can be detected by various tests, such as the analysis of chromosomal aberrations in metaphase cells, the micronucleus test, the test of exchange between the sister chromatids and the comet assay, among others, and as already indicated it may result from the direct interaction of the clastogenic agents with the DNA, or from indirect mechanisms which interfere with the replication and the repair of the molecule [1 , 63 – 65 ]. The accumulation of changes in the genetic material of the cell depends on several endogenous factors such as gender, age, and genetic constitution, as well as exogenous factors such as consumption habits, food, medications, and exposure to chemicals. Thus, besides the identifi cation of these agents that can interfere with the cellular instability, one must consider the level of individual exposure and the susceptibility of each individual, emphasizing the existence of an interrelationship between environmental triggering factors and predisposing genetic factors [66 ]. Since most of the exposures to genotoxic agents occur in occupational environments, it is important to monitor these populations at risk, whose consequences can be seen in the short, medium, or long term. The monitoring techniques allow the early identifi cation of risks, enabling the intervention and, consequently, minimizing or eliminating their effects on health. The analysis of the combination of these factors aims to infer the individual risk of a person to develop diseases like cancer, compared to a given exposure [ 67 ]. 130 Ana Paula A. Guimarães et al.

No one reported in the literature the minimum size population to be studied, but in general samples of less than 50 individuals are considered small samples. It is important to undertake the study of an exposed population, selecting the individuals who will be part of the study (exposed population) and individuals who will be part of the control group. For the selection of both individuals that will be part of the sample, and those who will be part of the control, an interview with questions that address the lifestyle habits, such as drinking and smoking, should be carried out. It is ideal that the groups are matched so that the only difference between them is the exposure factor. The subjects should be matched with controls based on gender, age, and habits like smoking and/or drinking. If the study is ethnically heterogeneous, pairing of the population by ethnic group should also be considered [ 45 ]. The lymphocytes circulating in the human blood represent a great system for testing a substance for its ability to produce chromosomal aberrations [ 68 ], which can be assessed by the direct analysis of these changes, by the micronucleus test, or via molecular tests such as in situ ﬂ uorescent hybridization (FISH). These analyses can assess the genotoxicity of the exposure to the ionizing radiation, drugs, handling of pesticides, and other chemicals, either for professional, therapeutic, or accidental reasons. Assuming that there is a correlation between the damage induced in the blood cells and other somatic cells, the lymphocytes would serve as a sen- tinel system for the high-risk groups [ 69 , 70 ].

5 Analysis and Interpretations of Results

5.1 Analysis For the analysis of the metaphases, the prepared slides must be of the Results identifi ed independently of each other, and analyzed in a blind test. in Cell Culture The cells scored must contain a number of centromeres equal to the modal number ±2, for all types of cell. Generally four slides are made per treatment, and at least 100 well-spread metaphases must be analyzed per treatment. This number can be reduced when there is a large number of aberrations [25 ]. The chromosomal and chromatid aberrations must be recorded separately and classifi ed by sub-types: chromatid breaks (CTB), chromatid exchange (CTE), chromosome breaks (CSB), chromatid and chromosome gap (CTG), and chromosome exchanges (CSE), including the dicentric and ring chromosomes. Also, the total number of cells that have aberrant chromosomes, including CTG and no CTG is measured. The procedures performed in the laboratory must be done by trained professionals and, if necessary, be assessed in pairs [25 , 71 ]. The fi nal results are evaluated as follows: negative (−) if the frequency of aberrant cells is <5 %, inconclusive (±) if it is ≥5 % but <10 % and positive (+) if it is ≥10 %. This assessment needs to be Chromosomal Aberration Test 131

conﬁ rmed by a statistical analysis. The statistics is performed by analysis of variance (ANOVA), with post Dunnett or Tukey’s test, with p value <0.05 considered statistically signiﬁ cant [34 , 71 ].

5.2 Analysis For each individual (or treatment), the analysis of CAs must be of the Results performed in a blind test and 100 metaphases analyzed. The meta- in Cultured phase must be sought in a 10× objective, then moving to the lens Lymphocytes of 40×, checking the quality of the material. When a good metaphase to perform the analysis is found, one uses the immersion objective (100×), initially counting the number of the chromosomes present; if a change in the number (which in human somatic cells is 46 chromosomes) is found, one checks if there was no technical failure, as for instance, the complete destruction of the nuclear membrane (during the treatment with hypotonic solution) allowing the chromosomes to spread and/or get lost along the preparation. If the preparation is of good quality, the change found must be a type of numerical chromosomal aberration. In addition to the counting of chromosomes number, checking that the chromosome structures are intact must be performed. The structural chromosomal aberrations that may be found during the analysis are generally divided into breaks or gaps , which may be of chromosome (observed in the two chromatids) or chromatid (observed in only one of the chromatids) type. A region of discon- tinuity smaller than the width of the chromatid [ 72 ] is considered a gap . Such discontinuities are diffi cult to be counted due to their large variation between the experiments and the lack of insight into their true nature [15 ]. Complex rearrangements occur when there is an exchange of parts of chromosomes, homologous or not, resulting from breaks followed by the reunion of the injured parties. As an example of complex rearrangements ring chromosomes, dicentric ones and triradial and quatriradial chromosome formation may be cited (Fig. 1 ) [73 ]. The breaks or gaps of the chromosomal type generally arise as a result of the exposure to S-dependent agents during the G1 phase of the cell cycle and may be viewed on the fi rst metaphase. The cells that suffered damage during the G2/S result in gaps or breaks of the chromatid type. The structural chromosomal aberrations, above described, are often found in the studies of the populations exposed to mutagenic agents and are not necessarily fi xed in the cell population, since the cells do not survive these changes [ 15 ]. For the classifi cation and distribution of the chromosomes within the different groups (A, B, C, D, E, F, G), the following criteria must be met: – GROUP A: Comprising the six largest chromosomes, the metacentric pair 1, the submetacentric pair 2, and the metacentric pair 3 that are slightly smaller than the pair 1. – GROUP B: Includes the pairs 4 and 5, which are submetacentric. The size of the short arm of each chromosome is equal to 132 Ana Paula A. Guimarães et al.

Fig. 1 Chromosomal alterations in metaphasic lymphocytes: (a ) chromatid gap, (b ) chromatid break, ( c ) dicentric chromosome, (d ) the triradial chromosome formation, (e ) the quatriradial chromosome formation and (f ) ring chromosome (Micrograph courtesy of Doctor Andre Salim Khayat)

1/3 of their long arms, and the pairs are indistinguishable from each other. – GROUP C: Comprising 15 chromosomes in men and 16 in women because the X chromosome is included in this group. They are metacentric or submetacentric chromosomes smaller than those of the groups A and B; they are difﬁ cult to identify individually, and only the pairs 6 and 7 (because they are the largest ones of the group), and the X chromosome can be clearly identiﬁ ed. Chromosomal Aberration Test 133

– GROUP D: Consists of three pairs of acrocentric chromosomes (13, 14, and 15) of medium size that are indistinguishable from each other. – GROUP E: Includes three pairs of small chromosomes (but not the smallest), of which 16 are metacentric, and 17 and 18 are submetacentric. – GROUP F: Consists of the smallest metacentric chromosomes (pairs 19 and 20) that are indistinguishable from each other. – GROUP G: Comprises 4 chromosomes in women and 5 in men because the Y chromosome belongs to this group. They are the smallest acrocentric chromosomes. Besides the veriﬁ cation of the possible changes in the number and/or morphology of the chromosome set, one must also check the Mitotic Index (MI), which represents the number of the metaphases found in a total of 1,000 interphase cells. This data indicates the cytotoxicity of an agent to be studied [ 74 ].

5.3 Steps to Perform All slides, including those of positive and negative controls, should the Analysis be independently coded before microscopic analysis for chromo- of the Results somal aberrations. Since ﬁ xation procedures often result in the breakage of a proportion of metaphase cells with chromosomes losses, as indicated above, the cells scored should, therefore, contain a number of centromeres equal to the modal number ±2 for all cell types, and at least 100 cells should be scored per concentration and control, equally divided among the duplicates, if applicable. In case of single culture per dose, at least 100 well-spread metaphases should be scored in this single culture. This number can be reduced when high numbers of aberrations are observed. In general, four slides are analyzed per culture. For in vitro studies, at least three analyzable test concentrations from duplicate cultures should be evaluated. For substances demonstrating little or no toxicity, concentration intervals of approximately two to threefold will usually be appropriate. However, many substances exhibit steep concentration–response curves and, in order to obtain data at low and moderate toxicity, it will be necessary to use more closely spaced concentrations. When it is desirable to study the dose–response relationship in detail, more than three concentrations will be needed. In these cases, a larger number of concentrations (single cultures or duplicates) will be necessary. If single cultures are used, then the negative control should be in duplicate. Where cytotoxicity occurs, the test concentrations selected should cover a range in which the ﬁ rst concentration producing cytotoxicity is on the top. If the maximum concentration is based on cytotoxicity, the highest concentration should aim to achieve 55 ± 5 % cytotoxicity, using the recommended cytotoxicity parameters, i.e., RICC and RPD ( see below) for cell lines, and mitotic index for primary cultures of lymphocytes. 134 Ana Paula A. Guimarães et al.

Care should be taken at interpreting as positive results only those found at the higher end of this range and including concentrations at which there is moderate and little or no cytotoxicity.

5.4 Formulas – Mitotic index (MI): for Cytotoxicity Number of mitotic cells Assessment MI =´100 Total number of cells scored

– Relative Increase in Cell Counts (RICC) and Relative Population Doubling (RPD) are recommended, as both take into account the proportion of the cell population which has divided.

Increase in number of cells in treated cultures() final- starting RICC = ´100 Increase in number of cells in control cultures final- starting (() No. of Population doublings in treated cultures RPD = ´100 No. of Populatioon doublings in control cultures

Where

Population Doubling=¸élog () Post- treatment cell number Initial cell nnumber ù ¸ log 2 ë û In any case, the number of cells before treatment should be measured and the same for positive and negative control cultures. Either duplicate or single treated cultures may be used at each concentration tested. In either event, the number of concentrations used must be sufﬁ cient to provide conﬁ dence in the evaluation. Particularly, in situations where the chemical is negative or weakly positive, it maybe advisable to use single treated cultures, and increase the number of different concentrations evaluated in a single experiment Because of the importance of the negative controls, it is recommended to use duplicate negative control cultures. Cells with structural chromosomal aberration(s) including or excluding gaps should be scored. Procedures in use in the laboratory should ensure that analysis of chromosomal aberrations is performed by well-trained scorers and peer-reviewed if appropriate. Although the purpose of the test is to detect structural chromosomal aberrations, it is important to record polyploidy and endoreduplication frequencies when these events are seen.

5.5 Treatment The percentage of cells with structural chromosomal aberration(s) of Results should be evaluated. Chromatid and chromosome-type aberrations classiﬁ ed by sub-types (breaks, exchanges) should be listed separately with their numbers and frequencies for experimental and control cultures. Gaps are recorded and reported separately but not included in the total aberration frequency. Chromosomal Aberration Test 135

Concurrent measures of cytotoxicity for all treated, negative and positive control cultures in the main aberration experiment(s) are necessary and should be recorded. Individual culture data should be provided. Additionally, all data should be summarized in tabular form. The following eligibility criteria are considered for the evaluation of the results: 1. There is a dose-related increase in the CA frequency. 2. At least one of the test concentrations exhibits a statistically signifi cant increase compared to the concurrent negative control. 3. The positive result is reproducible (e.g., between duplicates or between experiments). 4. The positive result is outside the distribution of the historical negative control data (e.g., 95 % confi dence interval). A test substance that meets all the above criteria, in at least one experimental condition, is considered to induce chromosomal aberrations in cultured mammalian somatic cells in this system. A test substance that meets none of the above criteria under all experimental conditions is considered unable to induce chromosomal aberrations in cultured mammalian somatic cells in this system. There is no requirement for verifi cation of a clear positive or negative response. A test substance which meets only some of the above criteria should be evaluated by expert judgment and/or further investigations from the existing experiments (e.g., consider analyzing more cells, more cultures). If the results remain inconclusive, they should be clarifi ed by further testing, preferably using modifi cation of experimental conditions (e.g., other metabolic activation conditions (i.e., S9 concentration or S9 origin), length of treatment, sampling time, concentration spacing). In rare cases, even after further investigations, the data set will preclude making a conclusion of positive or negative results and will therefore be concluded as equivocal. Positive results from the in vitro chromosomal aberration test indicate that the test substance induces structural chromosomal aberrations in cultured mammalian somatic cells. Negative results indicate that, under the test conditions, the test substance does not induce chromosomal aberrations in cultured mammalian somatic cells. An increase in the number of polyploidy cells may indicate that the test substance may have the potential to inhibit mitotic processes and to induce numerical chromosomal aberrations under certain conditions. An increase in the number of cells with endoreduplicated chromosomes (Fig. 2 ) may indicate that the test substance has the potential to inhibit cell cycle progression and should therefore be cautiously interpreted as they could refl ect phenomena not resulting in aneuploidy. 136 Ana Paula A. Guimarães et al.

Fig. 2 Metaphase chromosome spread from human cells with endoreduplicated chromosomes (Micrograph courtesy of Doctor Eleonidas Moura Lima)

References

1. Galloway SM (1994) Chromosome aberrations 8. Preston RJ, San Sebastian JR, McFee AF induced in vitro : mechanisms, delayed expres- (1987) The in vitro human lymphocyte assay sion and intriguing questions. Environ Mol for assessing the clastogenicity of chemical Mutagen 23(24):44–53 agents. Mutat Res 189:175–183 2. Natarajan AT (2002) Chromosome aberration: 9. Natarajan AT, Zwanenburg TSB (1982) past, present and future. Mutat Res 504 Mechanisms for chromosomal aberrations in (1–2):3–16 mammalian cells. Mutat Res 95:1–6 3. Snustad. DP, Simmons MJ (2013) Variação 10. Natarajan AT, Balajee AS, Boli JJWA et al no número e estrutura dos cromossomos. In (1996) Mechanisms of induction of chromo- Snustad DP, Simmons MJ (eds) Fundamentos somal aberrations and their detection by in situ de genética, 6ª edn. Guanabara Koogan, Rio hybridization. Mutat Res 372:247–258 de Janeiro (RJ) 11. Natarajan AT (1984) Origin and signiﬁ cance 4. Surralés J, Perto S, Ramirez MJ et al (1998) of chromosomal alteration. In: Obe G (ed) Links between chromatin structure, CNA repair Mutations in man. Springer, Berlin, pp 156–176 and chromosome fragility. Mutat Res 404:39–44 12. Kilman BA, Nichols WW, Levan A (1963) 5. Palitti F (1998) Mechanisms of the origin The effects of deoxyadenosine and cytosine of chromosomal aberrations. Mutat Res 404: arabinoside on the chromosomes of human 133–137 leukocytes in vitro . Hereditas 50:139–143 6. Swierenga SHH, Heddle JA, Sigal EA et al 13. Nichols WW (1964) In vitro chromosome (1991) Recommended protocols based on a breakage induced by arabinosyladenine in survey of current practice in genotoxicity testing human leukocytes. Cancer Res 8:1502–1505 laboratories, IV. Chromosome aberration and 14. Perthes G (1904) Versucheüber den Einﬂ uss sister-chromatid exchange in Chinese hamster der Röntgenstrahlen und Radiumstrahlen auf ovary, V79 Chinese hamster lung and human die Zellteilung. Deut Med Wochenschr 30: lymphocyte cultures. Mutat Res 246: 301–322 632–634 7. Evans HJ (1977) Molecular mechanisms in the 15. Natarajan AT, Vyas RC, Darroudi F et al induction of chromosome aberrations. In: (1990) DNA lesions, DNA repair and chromo- Scott D, Bridges BA, Sobels FH (eds) Progress somal aberrations. In: Obe G, Natarajan AT genet. Toxicologic, vol 2. Elsevier/North (eds) Chromosomal aberrations: basic and Holland, Amsterdam, pp 57–74 applied aspects. Springer, Berlin, pp 31–40 Chromosomal Aberration Test 137

16. De Vries H (1918) Mass mutations and twin research, vol 5. Elsevier Science Publishers, hybrids in Oenotheragrandifl ora . Ait Bot Gaz Amsterdam, New York, Oxford, pp 427–432 65:377–422 29. Galloway SM, Armstrong MJ, Reuben C et al 17. Muller HJ (1928) The production of muta- (1987) Chromosome aberration and sister tions by X-rays. Proc Natl Acad Sci U S A chromatid exchanges in Chinese hamster ovary 14(9):714–726 cells: evaluation of 108 chemicals. Environ Mol 18. Scott D, Evans HJ (1967) X-ray-induced chro- Mutagen 10(10):1–175 mosomal aberrations in Viciafaba. Changes in 30. Peres CM, Curi R (2005) Como Cultivar response during the cell cycle. Mutat Res células, 1st edn. Guanabara Koogan, Rio de 4(5):579–599 Janeiro 19. Evans HJ, Scott D (1969) The induction of 31. Galloway S (2000) Cytotoxicity and chromo- chromosome aberrations by nitrogenmustard some aberrations in vitro : experience in indus- and its dependence on DNA synthesis. Proc R try and the case for an upper limit on toxicity in Soc Lond Ser B 173:491–512 the aberration assay. Environ Mol Mutagen 20. Natarajan AT, Obe G, Hayashi M (2008) 35:191–201 Chromosomal aberrations. Mutat Res 657(1): 32. Lorge E, Hayashi M, Albertini S et al (2008) 1–2 Comparison of different methods for an accu- 21. Amorim MIM, Mergler D, Bahia MO et al rate assessment of cytotoxicity in the in vitro (2000) Cytogenetic damage related to low micronucleus test. I. Theoretical aspects. Mutat levels of methyl mercury contamination in Res 655:1–3 the Brazilian Amazon. An Acad Bras Cienc 33. Galloway S, Lorge E, Aardema MJ et al (2011) 72(4):497–507 Workshop summary: top concentration for in 22. Natarajan AT, Santos SJ, Darroudi F et al vitro mammalian cell genotoxicity assays; (1998) 137Cesium-induced chromosome and report from working group on toxicity aberrations analyzed by fl uorescence in situ measures and top concentration for in vitro hybridization: eight years follow up of the cytogenetics assays (chromosome aberrations Goiânia radiation accident victms. Mutat Res and micronucleus). Mutat Res 723:77–83 400(1–2):299–312 34. Dusinska M, Kazimırova A, Barancokova M 23. Guimarães APA, Dias FL, Cardoso RS et al et al (2003) Nutritional supplementation with (2003) Cromosomal aberration induced by antioxidants decreases chromosomal damage in 5-azacytidine combined with VP-16 (etopo- humans. Mutagenesis 18:371–376 side) in CHO-K1 and XRS-5 cell lines. Teratog 35. Shaffer LG, Tommerup N (2005) ISCN 2005: Carcinog Mutag 23(51):171–186 an international system for human cytogenetic 24. FDA—Food and Drug Administration (2012) nomenclature. Cytogenetic & genome research. ICH S2 (R1), Genotoxicity Testing and Data S. Karger AG, Basel, Switzerland Interpretation for Pharmaceuticals Intended 36. Tijo JH, Levan A (1956) The chromosome for Human Use number in man. Hereditas 44:1–6 25. OECD—Organization for Economic Coopera- 37. Moorhead PS, Noweel PC, Mellman WJ et al tion and Development. Guideline for testing of (1960) Chromosome preparation of leuko- chemicals. Proposal for updating test guideline cytes cultured from Human peripheric Blood. 473 (2012) OECD 473— In vitro Mammalian Exp Cell Res 20:613–616 Chromosome Aberration Test. Adopted: 21st 38. Caspersson T, Farber S, Foley GE et al (1968) July 1997 Chemical differentiation along metaphase 26. Pfuhler S, Fellows M, Benthem JV et al (2011) chromosomes. Exp Cell Res 49(1):219–222 In vitro genotoxicity test approaches with 39. Latt AS, Schreck RR (1980) Sister chromatid better predictivity: summary of an IWGT exchange. Am J Hum Genet 32:297–313 workshop. Mutat Res 723(2):101–107 27. Evans HJ (1976) Cytological methods for 40. Carpenter NJ (2001) Molecular cytogenetics. detecting chemical mutagens. In: Hollaender Semin Pediatr Neurol 8(3):135–146 A (ed) Chemical mutagens, principles and 41. Denver Conference (1960) The identifi cation methods for their detection, vol 4. Plenum of individual chromosomes especially in man. Press, New York, London, pp 1–29 Am J Hum Genet 12:384–389 28. Ishidate M Jr, Sofuni T (1985) The in vitro 42. International Atomic Energy Agency (1986) chromosomal aberration test using Chinese Biological dosimetry chromosome aberrations hamster lung (CHL) fi broblast cells in culture. analysis for dose assessment. Technical Report In: Ashby J et al (eds) Progress in mutation series n 260, Vienna, IAEA, STI/PUB/10/260 138 Ana Paula A. Guimarães et al.

43. Obe G, Pfeiffer P, Savage JRK et al (2002) ants of a large uranium mineralization área: the Chromosomal aberration: formation, identifi - municipalities of Monte Alegre, Phainha and cation and distribuition. Mutat Res 504:17–36 Alenquer, in the State of Pará, Brazil. Cell Biol 44. Gomes JJ, Frade SPPJ, Figueiredo BJ et al Toxicol 26(5):403–419 (2011) Comparação de Métodos de cultura 58. Min T (2003) Fish techniques. In: Swansbury de Linfócitos em sangue periférico para análise J (ed) Cancer cytogenetics. Methods and pro- citogenética, 2ª edn. Reunião Brasileira de tocols. Humana Press, Totowa, NJ Citogenética. Águas de Lindóia (SP) 59. Natarajan AT (1993) Techniques for biomoni- 45. Ribeiro LR, Salvadori DMF, Marques EK toring of human populations for genetic (2003) Mutagênese Ambiental. Ed ULBRA, effects. Rev Brasil Genet 16(3):841–847 Canoas, pp 151–172 60. Little John B (2000) Ionizing radiation. In: 46. Pardue ML, Gall JG (1969) Molecular hybrid- Bast RC Jr, Kufe DW, Pollock RE et al (eds) ization of radioactive DNA to the DNA of Cancer medicine, 5th edn. BC Decker Inc., cytological preparations. Proc Natl Acad Sci 2400 pp, Hamilton (ON) USA 64(2):600–604 61. Therman E, Susman M (1996) Cromossomos 47. Kallioniemi A, Visakorpi T, Karhu R et al (1996) humanos: estructura, comportamiento y efec- Gene copy number analysis by fl uorescence in tos, 3ª edn. Revista Brasileira de Genética, situ hybridization and comparative genomic Ribeirão Preto, 404 p hybridization. Methods 9(1):113–121 62. Bahia MO (1997) Le potentielgénotoxique 48. Shinawi M, Cheung SW (2008) The array du mercure: mutations PRET et effects cyto- CGH and its clinical applications. Drug Discov génétiques. Dissertação de Mestrado em Today 13(17–18):760–770 Biologia. Montreal, Université du Quebéc á 49. Maluf SW, Riegel M (2011) Citogenética Montreal, 119 p humana. Ed Artmed, Porto Alegre, p 334 63. Amorim MI, Ferrari I, Bahia Mde O et al 50. Arnoldus EP, Dreef EJ, Noordermeer IA et al (2008) Genotoxic effects of white fl uorescent (1991) Feasibility of in situ hybridisation with light on human lymphocytes in vitro. Mutat chromosome specifi c DNA probes on paraffi n Res 652(2):204–207 wax embedded tissue. J Clin Pathol 44(11): 64. Movajagh A, Maleki F, Mohammadzadeh SG 900–904 et al (2005) Association of glutathione 51. Costa Raiol LC, Figueira Silva EC, Mendes da S-transferase and chromosomal aberrations as a Fonseca D et al (2007) Interrelationship means to determine occupational exposure. Int between MYC gene numerical aberrations and Cong Ser 1276:197–198 protein expression in individuals from northern 65. Castañeda AN, Téllez MGO, Rodriguez- Brazil with early gastric adenocarcinoma. Annaiz R (2006) Genotoxic profi le of inhibi- Cancer Genet Cytogenet 181(1):31–35 tors of topoisomerase I (camptotecin) and II 52. Das K, Tan P (2013) Molecular cytogenetics: (etoposide) in a mitotic recombination and recent developments and applications in cancer. sex-chromosome loss somatic eye assay of Clin Genet 84(4):315–325 Drosophila melanogaster. Mutat Res 604 53. Liehr T, Starke H, Weise A et al (2004) (1–2):83–90 Multicolor FISH probe sets and their appli- 66. Du Four VA, Janssen CR et al (2005) cations. Claussen Histol Histopathol 19: Genotoxic and mutagenic activity of environ- 229–237 mental air samples from different rural, urban 54. Imataka G, Arisaka O (2012) Chromosome and industrial sites in Flanders, Belgium. Mutat analysis using spectral karyotyping (SKY). Cell Res 588(2):106–117 Biochem Biophys 62:13–17 67. Gattás GJF, Segre M, WünschFilho V (2002) 55. Speicher MR, Gwyn Ballard S, Ward DC Genetics, molecular biology and ethics: work (1996) Karyotyping human chromosomes by and health connections. Ciência Saúde Coletivan combinatorial multi-fl uor FISH. Nat Genet 7(1):159–167 12(4):368–375 68. Natarajan AT, Obe G (1980) Screening of 56. Tanke HJ, Wiegant J, van Gijlswijk RP et al human populations for mutations induced (1999) New strategy for multi-colour fl uore- by environmental pollutants: use of human scence in situ hybridisation: COBRA: lymphocite system. Ecotoxicol Environ Saf 4: COmbined Binary RAtio labelling. Eur J Hum 468–481 Genet 7(1):2–11 69. Rabello-Gay MN, Rodrigues MA, La R et al 57. Guimarães AC, Antunes LM, Ribeiro HF et al (1991) Mutagênese, teratogênese e car- (2009) Cytogenetic biomonitoring of inhabit- cinogênese: métodos e critérios de avaliação. Chromosomal Aberration Test 139

Sociedade Brasileira de Genética, Revista 72. Preston JR, Au W, Bender MA et al (1981) Brasileira de Genética, Ribeirão Preto, 246 p Mammalian in vivo and in vitro cytogenetic 70. Brusick DJ, Young RR (1981) IERL-RTP assay: a report of the U.S.EPA’s Gene Tox procedure: manual level. In: Environmental Program. Mutat Res 87:143–188 assessment biological tests. (EPA-600/8-81- 73. Savage JRK (1976) Classiﬁ cation and relation- 024), NTISPB 82-228966. Kenson. Litton ships of induced chromosomal structural Bionetics changes. J Med Genet 13:102–122 71. Chaudhary M, Payasi A (2013) Evaluation of 74. Al Sabati K, Lloyd DC, Edwards AA et al (1992) genotoxicity of CVA1020 through Ames and A survey of lymphocyte chromosomal damage in vitro chromosomal aberration tests. Br J in Slovenian workers exposed to occupational Pharmacol Toxicol 4(3):95–100 clastogens. Mutat Res 280(3):215–223 Chapter 8

Chromosome Painting of Mouse Peripheral Blood and Spleen Tissues

Dayton M. Petibone , James D. Tucker , and Suzanne M. Morris

Abstract

Exposure to genotoxins may induce both structural and numerical chromosome aberrations. Whole chromosome ﬂ uorescence in situ hybridization (FISH) of metaphase cells provides a means for analyzing chromosome aberrations with single-cell level resolution. The laboratory mouse model is an essential tool for in vivo genotoxicity studies. However, few published resources are available that provide a comprehensive background and instructions on how to perform whole chromosome FISH on mouse metaphase cells. Here, we consolidate several techniques into a single, comprehensive protocol describing the preparation of mouse metaphase cells for whole chromosome FISH analysis. Also presented are basic recommendations on how to visualize the slides and how to organize the data for analysis and interpretation.

Key words Genotoxicity , Risk assessment , Regulatory science , Mouse , Peripheral blood culture , Spleen tissue culture , FISH painting , Cytogenetics , Microscopy

1 Introduction

Chromosome aberrations are a well-accepted biomarker for evaluating exposures to clastogenic compounds and ionizing radiation. Data from experimental and epidemiological studies suggest that chromosome aberrations are involved in carcinogenesis, and individuals with higher frequencies of chromosomal aberrations appear to be at greater risk for cancer [ 1 , 2 ]. Structural chromosomal aberrations that are stable and therefore persist through mitosis, particularly translocations, can result from exposure to

genotoxic events in the G 0 and G1 phases of the cell cycle and appear in the ﬁ rst mitotic metaphase division in cultured cells [ 3 ]. Chromosome aberrations can be induced through several mechanisms, including direct damage to DNA by double strand breakage, replication on a damaged DNA template, and/or damage to DNA synthesis and DNA repair proteins [ 4 , 5 ]. Whole chromosome painting by ﬂ uorescence in situ hybridization (FISH) enables rapid and accurate assessment of chromosome translocations, insertions,

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_8, © Springer Science+Business Media New York 2014 141 142 Dayton M. Petibone et al.

and dicentrics, in addition to detection of acentric fragments and hyperdiploid events. The cell type most frequently used for mouse FISH genotoxicity studies are T-lymphocytes from peripheral blood or from spleens. There are several advantages to using T-lymphocytes: cells are easily isolated from peripheral blood, and are grown in suspension so trypsinizing the cell monolayer is not necessary. In addition, T-cells are quiescent which allows damage to accumulate during pro- tracted exposure periods; the damage may be diluted in cells that are actively dividing. Mouse molecular cytogenetic studies evaluating genotoxic risks of chemicals can use FISH to analyze changes in telomeres, centromeres, specifi c regions of a chromosome or whole chromosomes. The objective of these molecular cytogenetic studies is to identify rare chromosome translocation events and related structural chromosome aberrations. To achieve this objective, scoring a large number of metaphase cells is necessary. To make translocation events readily identifi able, a two-color FISH painting with a counterstain produces an excellent 3-color scoring system. As is commonly done in human cells, simultaneously FISH painting several mouse chromosomes each in red and in green, and using a 4′,6-diamidino-2-phenylindole (DAPI) fl uorescent counter is one approach that has worked well [ 6 , 7 ]. The drawback with this 3-color scoring system is that painting only a few of the chromosomes will not identify all translocation events, i.e., detectable chromosome exchanges will only be those that involve chromosomes labeled in different colors, while exchanges involving chromosomes labeled the same color are not detectable. The more chromosomes that are painted and the more unique colors used, the greater the ratio of detectable exchange events. In these molecular cytogenetic studies, increasing the number of painted cells analyzed will increase the number of whole genome cell equivalents (CEs) and increase the statistical power of analysis. The advantage of using FISH in cytogenetic applications is the ability to detect translocations in genotoxicity risk assessments. Translocations are diffi cult to quantify with classical chromosome banding techniques, as highly trained personnel are necessary to perform the analysis and the methodology is time consuming. Identifi cation of translocations with FISH does not require extensive training for slide readers, and under optimal hybridization conditions an experienced slide reader can score 1,000 metaphase cells in one workday. However, the high cost of both slide reader labor and the FISH probes are a disadvantage and are limiting factors in many study designs. Any chemical that induces DNA breaks can theoretically result in chromosome aberrations. Conducting FISH analysis of animals treated at doses of the test article that are positive in the micronucleus (MN) and comet assays for DNA damage can be used as a follow-up assay. If MN and comet datum are not available, begin Chromosome Painting of Mouse Peripheral Blood and Spleen Tissues 143

by analyzing a small subset of the doses (e.g., negative control, medium dose, and high dose) to determine if there is a detectable response. The FISH analysis of chromosome aberration frequencies, when used in conjunction with MN and comet data, can inform investigators as to the persistence and frequency of chromosome aberrations that could result in disease. A minimum of fi ve animals per treatment dose (including negative and solvent controls) is generally recommended, although this number may be adjusted depending on the needs of the study, including tolerances for Type I and Type II statistical errors. In animal studies, the unit of statistical analysis is the animal , not the number of cells scored. The number of metaphase cells evaluated per animal depends on two primary considerations: (a) the desired level of accuracy in determining the frequency of events (e.g., translocations per cell) per animal and (b) the number of chromosomes painted and the number of colors used. The fi rst issue depends on the desired level of statistical certainty, but in general, 100–200 whole genome equivalents per animal is considered suffi cient in most studies. The number of metaphase cells that need to be scored to achieve the desired number of whole genome equivalents depends on several factors. With painting, the only rearrangements that can be seen are those between chromosomes labeled in different colors. Exchanges between chromosomes labeled in the same color will not be observed. This means that the more chromosomes that are painted, and the more colors that are used, the greater will be the fraction of structural chromosome aberrations observed. Furthermore, chromosome size is important, as larger chromosomes are more likely to be involved in rearrangements than smaller chromosomes. The mathematical relationship between the number of chromosomes painted, the size of those chromosomes, and the number of colors used has been described in detail [ 8 ]. The terms “cell equivalents” or “whole genome equivalents” are used to relate the number of metaphase cells scored to the corresponding number of whole genomes evaluated. For example, if a particular set of chromosome paints identifi es 0.333 of all translocations, then 1/0.333 = 3 metaphase cells must be scored to obtain the equivalent amount of information obtained in one whole genome (e.g., a Giemsa-banded cell). This means that if 200 whole genome equivalents are needed per animal, then 600 metaphase cells must be scored.

2 Materials

2.1 Blood Draw, 95 % ethanol (general purpose) diluted to 70 % with dH2 O in a Spleen Harvest, spray bottle and carboy and Sample 3 mL syringe with a 22 gauge × 1 ½″ needle (BD Bioscience, Preparation 14-826-85) 144 Dayton M. Petibone et al.

1 mg/mL heparin (10 mg, Fisher, BP2524-10) in 1× HBSS solution, ﬁ ltered with a 0.22 μm syringe ﬁ lter and stored at 4 °C. 15 mL polypropylene tubes with plug-seal cap (Corning, 430052) Dissecting corkboard (Fisher, 36-119) Dissecting T-pins (Fisher, S99385) Adsorbent dissection underpads (Fisher, 14-206-62) 4″ Curved forceps (Roboz, RS-5138), foil wrapped and autoclaved 4 ½″ scissors (Roboz, RS-5912), foil wrapped and autoclaved Disposable sterile scalpels (Fisher, 31-200-32) RPMI 1640 (Invitrogen, 11875-119), ice cold Iodine tincture (Fisher, S71952) 1× HBSS [-] calcium chloride [-] magnesium chloride [-] magnesium sulfate (Invitrogen, 14175-092) Ice bucket (general purpose) with water ice 60 × 15 mm cell culture dishes (Corning, 430166) 70 μm tissue strainers (Falcon, 352350) 10× RBC lysis buffer (BioLegend, 420301) Lymphocyte separation media (LSM), Accu-paque (Accurate Chemical, AN-3100) 5 ¾″ Pasteur pipettes (Fisher, 22042816) autoclaved in sterilization can (Fisher, 03-475-1)

2.2 Tissue Culturing RPMI 1640 supplemented with 2 mM L -glutamine and phenol red (Invitrogen, 11875-119) 200 mM L -glutamine (Invitrogen, 25030081) 5,000 U/mL Penicillin 5,000 μg/mL Streptomycin, (Invitrogen, 15070063) Fetal bovine serum (FBS, Atlanta Biologicals, S12650) Phytohemagglutinin (PHA), purifi ed (Remel, 30852801) 10 μg/mL Colcemid in PBS (Invitrogen, 15212-012) 25 cm2 tissue culture fl ask with 0.2 μm vented cap (Corning, 430639) 500 mL Disposable Sterile Filter Systems, 0.22 μm Polyethersulfone (PES) fi lter (Corning, 431097) 5 mL pipettes, paper wrapped (Corning, 4487) 10 mL pipettes, paper wrapped (Corning, 4488) 50 mL pipettes, paper wrapped (Corning, 4490) 1,000 μL aerosol resistant barrier pipette tips (Fisher, 2779) 1,000 μL pipette (e.g., Rainin, L-1000XLS) Chromosome Painting of Mouse Peripheral Blood and Spleen Tissues 145

2.3 Cell Fixation, Potassium chloride (KCl, Fisher, BP366-500) Metaphase Cell Sodium citrate (Fisher, BP327-500) Spreads, and FISH Methanol (Fisher, A412-1) Painting Glacial acetic acid (Fisher, BP2401500) Absolute ethanol (general purpose) Gold seal 3 × 1″ microscope slides (Fisher, 12-518-102A) Delicate task wipes (Kimberly-Clark, 34133) 25 mm square cover glass (Corning, 12-524C) Mouse single- or multicolor FISH probe cocktail: Applied Spectral Imaging, single- and multicolor probe ( http://www.spectral-imaging.com/ ) Cytocell Ltd., single color probes (http:// www.cytocell-us.com ) Cambio, single color probes ( http://www.cambio.co.uk/ ) 100 μL aerosol resistant barrier pipette tips (Fisher, 2779) 100 μL pipette LTS lite-touch (Rainin, L-1000XLS) Formamide (Fisher, 50-720-4034) Plastic tube rack (general purpose) placed in water bath as a support for slides Glass Coplin jars with lid (general purpose) Gold Antifade Reagent with DAPI (Molecular Probes, S36938) 20× SSC (Fisher, BP1325-1) Triton-X detergent (Fisher, BP151-100) Rubber cement (Elmer’s, E425) Hand counter (Fisher, 07-905) Slide boxes (Fisher, 03-448-1) Humidity sponge packets (Fisher, 07-580) Pouch heat sealer (Ampac, 101-6) Plastic pouches, heat-sealing (Ampac, 404-24)

2.4 Equipment CO2 euthanizing chamber (Euthanex, E-22000)

CO2 tank (general purpose) Heating blocks (Torrey Pines, IC20) 2 each

H 2 O dual bath (Fisher, 15-474-30) Slide moat (Boekel, 05-450-30), 2 each set at 70 and 37 °C Inverted tissue culture microscope with 20× phase contrast objective (e.g., TMS, Nikon) Upright ﬂ uorescence microscope (e.g., 80i microscope, Nikon) equipped with: 146 Dayton M. Petibone et al.

DAPI, Cy3, FITC single band pass fi lters and DAPI-FITC-Cy3 triple band pass fi lters 10× (dry), 20× (dry), 40× (oil immersion), 60× (oil immersion), 100× (oil immersion) objectives Computer with 24″ color monitor (Hewlett Packard) Microsoft Offi ce/Excel Elements Basic Research software (Nikon) DS-Qi1Mc monochrome camera (Nikon, DS-Qi1Mc)

3 Methods

3.1 Blood Draw The cardiac puncture technique will obtain a single, 0.1–1 mL vol- and Spleen Harvest umes of a quality blood sample from a mouse under CO 2 narcosis. Cardiac puncture is a skill that requires much practice to be successful, 3.1.1 Terminal Cardiac and should therefore be performed by highly trained personnel. Puncture

1. In an animal sacriﬁ ce room containing a CO2 euthanasia device consisting of a euthanizing chamber attached to a CO2 tank, place 2–3 mice into the euthanizing chamber, and turn

on the CO 2 gas for approximately 2–3 min. The mice will stop moving, but their hearts should still be beating. Remove a mouse from the chamber, hold it ﬁ rmly by its back skin, spray the chest area with 70 % ethanol and rub gently with an index ﬁ nger to sanitize the area. Note: An animal facility should be readily available to house the mouse strain used in your study under the appropriate conditions and in accordance with the guidelines from the Institutional Animal Care and Use Committee ( IACUC ). A suitable work area should be available that contains a euthanasia chamber

attached to a CO 2 tank. Perform blood draw through terminal cardiac puncture on mice under CO 2 narcosis , that is , while the heart is still beating. The terminal cardiac puncture euthanizes the mice for the subsequent spleen isolation step . 2. Insert a 3 mL syringe with a 22 gauge × 1 ½″ needle containing 0.1 mL of 1 mg/mL heparin solution into the left side of the chest, through the diaphragm into a ventricle and begin drawing blood. Draw the blood slowly to prevent the heart from collapsing. Note : All blood draws performed for molecular cytogenetics should use heparin as an anticoagulant. Other anticoagulants , such as EDTA , inhibit cell division needed to produce metaphase cells in culture [9 ]. 3. Once the blood draw is complete, dispense the blood from the syringe into a 15 mL disposable centrifuge tube, labeled with sample ID, containing 1 mL of RPMI 1640 and place on ice. Chromosome Painting of Mouse Peripheral Blood and Spleen Tissues 147

Dispose of the uncapped used syringe into a sharps container. Gently mix the blood with the heparin and the RPMI 1640. The RPMI 1640 can contain an additional 0.1 mL of 1 mg/ mL heparin solution to prevent clotting (optional). Repeat until all the euthanized mice have had their blood drawn. If

mice remaining in the CO 2 chamber come out from under anesthesia, turn on the gas until they again stopped moving. Continue drawing blood from the remaining mice and place the processed ones on a corkboard covered with a dissection pad for spleen removal. Note : Dispensing drawn blood into ice - cold media and storing on ice until processing will maintain healthy cells and prevent hemolysis. As quickly as possible , aseptically remove the spleen and place in ice -cold media as is done with the blood sample. Shipping blood or fully dissociated spleen samples on water ice overnight , from or to , an off - site laboratory is permissible. Extended periods on ice up to 24 h will not negatively affect the health of the cells ; in fact , our observations indicate that incubation on ice improves metaphase cell production as long as care is taken to make sure the cells do not freeze .

3.1.2 Spleen Harvest 1. Pin the legs of euthanized mice with T-pins to the corkboard covered with a dissecting pad, spray the mouse abdomen with 70 % ethanol, and gently rub it with ﬁ ngers to disinfect the area. 2. Holding the skin with forceps on the left side of the abdomen below the rib cage, begin by creating an incision in the skin with a disposable scalpel, and continue making the incision into the peritoneal cavity with dissecting scissors until the spleen is visible. Holding the spleen with forceps, trim away any connected tissues and blood vessels. 3. Place the spleen in a 15 mL disposable centrifuge tube, labeled with sample ID, containing 1 mL RPMI 1640 and 0.1 mL of 1 mg/mL heparin solution on ice. Continue removing spleens from the remaining mice, and dispose of the carcasses in accordance with your IACUC guidelines.

3.2 Single Cell Perform all the following sample manipulations in a laminar ﬂ ow Suspensions hood equipped with a vacuum trap. from Whole Blood Mix the LSM by shaking vigorously, allow it to stand until the and Spleen Tissues air bubbles disappear, and aliquot 3 mL into 15 mL disposable centrifuge tubes. Using a 5 mL pipette, mix the diluted blood sample by 3.2.1 LSM Separation pipetting up and down several times before drawing into the pipette. of T-Lymphocytes Note: LSM consists of a high mass, hydrophilic, highly branched polysaccharide used for separation of blood components (e.g., monocytes) based on their buoyancy properties. LSM with a density = 1.0875 ± 0.0001 g / cm 3 composed of sodium diatrizoate and Dextran , pH 6.9 , performs well for murine lymphocyte isolation. 148 Dayton M. Petibone et al.

abc Before After centrifugation centrifugation

Diluted blood

Blood serum

Monocytes

LSM LSM Granulocytes/ erythrocytes

Fig. 1 Isolation of mouse mononuclear cells using density gradient LSM. (a ) Carefully, layer diluted blood solution over LSM in a 15 mL tube. ( b ) The blood solution and LSM will result in a distinct interface between the two layers. (c ) Following centrifugation, the components should separate into blood plasma, monocyte, LSM, and granulocyte/erythrocyte layers

This LSM formulation induces erythrocyte aggregation while reducing platelet aggregation to optimize viable lymphocyte isolation. Diluted whole blood layered on top of LSM separates into distinct layers following centrifugation. From top to bottom , these are the blood plasma layer , mononuclear cell layer at the plasma – LSM interface , the LSM layer and the ﬁ nal layer containing platelets , granulocytes, and erythrocytes (Fig. 1 ). Layering whole blood over the LSM must be done slowly and carefully to prevent mixing the blood with the LSM , and is therefore a time -consuming step. However , if numerous samples are processed , specialized porous devices and ready -made tubes are available that prevent mixing of blood samples with LSM during the layering . 1. Hold the tip of the 5 mL pipette against the tube wall and slowly dispense the diluted blood on top of the LSM, being careful not to allow mixing of the blood with the LSM. 2. Centrifuge the blood and LSM at 400 × g in a swinging bucket rotor for 30 min at 22 °C. With a sterile Pasteur pipette attached to a vacuum trap, remove the serum level to ~3 to 5 mm above Chromosome Painting of Mouse Peripheral Blood and Spleen Tissues 149

the monocyte layer at the LSM–serum interface. Using a 5 mL pipette, remove approximately 2 mL of the monocyte layer and LSM taking care not to disturb the pellet containing erythrocytes, granulocytes, and platelets below. Dispense the monocytes into a 15 mL disposable centrifuge tube, labeled with sample ID, containing 2 mL 1× HBSS. Securely replace the caps, and mix by gently inverting the tubes. Centrifuge the samples for 5 min at 200 × g at 22 °C in a swinging bucket rotor. 3. Aspirate the media down to ~3 to 5 mm above the cell pellet (the cell pellet may not be visible), taking care not to get too close to the bottom that cells would be aspirated with the media. Wash the cells in 5 mL 1× HBSS. Centrifuge for 5 min at 200 × g at 22 °C in a swinging bucket rotor. Repeat the wash step twice more.

3.2.2 Prepare Single Cell Perform all the following sample manipulations in a laminar fl ow Suspensions from Spleen hood equipped with a vacuum trap. 1. Transfer a spleen sample in RPMI 1640 into a 60 mm cell culture dish and aspirate the RPMI with a Pasteur pipette attached to the vacuum trap. Using sterile forceps and scissors, remove approximately half of the spleen, place in a 60 mm cell culture dish containing 1 mL 1× HBSS, and macerate the tissue with a forceps and scissors. Note: If microbial contamination becomes an issue in spleen cultures , a 1 % iodine tincture wash and rinse will disinfect the spleen. In a 60 mm tissue culture dish , rinse the exterior of the spleen briefl y with 1 mL 1 % iodine tincture using a 1,000 μL pipette tip. Aspirate the iodine tincture and rinse the spleen three times with 1 mL 1 × HBSS , aspirating the 1 × HBSS each time . 2. Pour the spleen slurry into a 70 μm tissue strainer sitting in a 60 mm cell culture dish, and holding the strainer fi rmly with forceps, push the tissue through the strainer with the plunger from a 3 mL syringe to produce a suspension of single cells. Transfer the cell suspension into a fresh 15 mL disposable centrifuge tube. Add another 1 mL 1× HBSS to the 70 μm tissue strainer and push any remaining tissue through the tissue strainer. Rinse the outer surface of the tissue strainer with 1 mL 1× HBSS twice to rinse off any cells, collect the rinsed cells, and place in the same 15 mL disposable centrifuge tube. Centrifuge the single cell suspension at 200 × g for 5 min in a swinging bucket rotor at 22 °C, aspirate the supernatant, and suspend in 1 mL complete RPMI medium. Note: If T -lymphocytes from spleen divide poorly in culture , removing the erythrocytes may improve the mitotic index. Spleen samples are not as amenable to cell isolation by LSM as peripheral blood. Therefore , process dissociated spleen tissues using an RBC lysis buffer containing ammonium chloride to lyse RBCs by osmotic disruption while leaving leukocytes unaffected [ 10 ]. In contrast , for unclear 150 Dayton M. Petibone et al.

reasons , we have found that peripheral blood treated with RBC lysis buffer does not produce metaphase cells .

3.3 Mouse 1. Suspend the entire cell pellets from the peripheral blood T-Lymphocyte harvests in 10 mL of complete medium containing RPMI Culture and Harvest 1640, 10 % FBS, 50 U/mL penicillin 50 μg/mL streptomycin, of Mitosis- Blocked 0.25 mM L -glutamine , and 0.25 μg/mL HA 16 PHA, and Cells transfer to 25 cm 2 tissue culture fl asks using a 10 mL pipette. Dispense 250 μL of the 1,000 μL spleen suspension into (up to four) 25 cm2 tissue culture fl asks containing 10 mL of RPMI 1640 complete media. Note: PHA is a plant lectin that acts as a mitogen to stimulate mitogen activated protein kinase pathways and induces cell division in lymphocytes. There are two grades of commercial PHA , a reagent grade in crude protein extracts , and in a highly purifi ed , HA 16 form. Only the HA 16 PHA ( not the reagent grade ) will stimulate murine T -cell mitosis and its use is essential for generating good quality metaphase cells . 2. Incubate the 25 cm2 tissue fl asks in a 37 °C incubator with 5 %

CO2 , sitting upright to concentrate the cells at the bottom of the fl ask, which will promote cell division. After 36 h incubation add Colcemid to a fi nal concentration of 0.1 μg/mL, and incubate an additional 4 h to arrest cells in metaphase. Centrifuge the cell cultures at 200 × g in a swinging bucket rotor for 5 min and remove the supernatant. Note : Incubating the cultures for 36 h in the presence of PHA will produce mainly fi rst division metaphase cells. The addition of Colcemid will block mitosis through microtubule -depolymer- ization. If the mitotic indices are low , increasing the Colcemid block up to 6 h may improve the metaphase cell yield . 3. Gently suspend the cells by adding 8 mL of hypotonic solution containing 75 mM KCl and 300 mM sodium citrate and incubate for 10 min at 37 °C. Handle the cells gently to prevent rupture, as the cells are swollen and extremely fragile at this point. Fix the cells by slowly adding 2 mL of methanol:glacial acetic acid (3:1) drop-wise to the side of the tube and mix by gently fl icking the tube. Centrifuge the cells at 200 × g for 5 min at 22 °C and aspirate the fi xative and hypotonic solution. Repeat the methanol:glacial acetic acid cell fi xation three more times, carefully adding the fi xative drop-by-drop on the side of the tube and then gently mix the cell pellet. Storing the fi xed cells at 4 °C in tightly capped tubes for extended periods does not have any negative effect on FISH, provided the fi xative does not evaporate. Prior to slide preparation following long storage, the fi xative in which the cells are stored should be replaced, i.e., the cells should be fi xed one more time. Chromosome Painting of Mouse Peripheral Blood and Spleen Tissues 151

3.4 Slide Preparation To standardize slide preparation and increase the likelihood of of Metaphase Cell obtaining well-spread metaphase cells, drop the cell suspensions Spreads onto high quality glass slides sitting on a grated support (a plastic tube rack works well) protruding 2–3 cm above the water line in a water bath set to 70 °C (Fig. 2 ). These conditions will minimize the effect of ambient environmental conditions and improve metaphase cell spreading. Note : The selection of high quality microscope slides made of optical quality borosilicate glass is important for successful molecular cytogenetic studies. Many low quality slides have imperfections such as pits or grooves that trap FISH probes and fl uorescent counterstains. Poor - quality slides reduce FISH probes hybridization effi ciency , and create brightly fl uorescent blemishes that make analysis diffi cult. Microscope slides , even those labeled as pre - cleaned , should fi rst be soaked in 70 % ethanol for a minimum of 1 h , then wiped with a delicate task wipe ,

rinsed in dH 2 O and allowed to air dry. This will remove glass particles and residual oil used in slide machining and polishing that would interfere with metaphase cells spreading . Note: Humidity and fi xative evaporation rate are crucial to obtaining well - spread metaphase cells on microscope slides [11 –13 ]. To standardize slide preparation and ensure that ambient environmental conditions minimally affect metaphase cell spreading , the fi xed cell suspension should be dropped onto a slide over a water bath set to 70 °C (Fig. 2 ). This serves to create a localized high humidity environment in which the fi xative can evaporate slowly and allow better metaphase cell spreading. Chilling the slides at 4 ° C and

Fig. 2 Slide preparation under a localized high humidity environment to control for ambient conditions. (a ) Place a microscope slide (chilled to 4 °C) at an angle on a supporting grate that allows the free ﬂ ow of steam from the water bath to form a layer of condensate on the slide surface. (b ) Flip the slide over so that the side with the layer of condensate is facing up, and using a micropipette, dispense 15–20 μL of cell suspension onto the slide from a height of 2 to 3 cm. Allow the cell suspension to spread and the ﬁ xative to dry before removing slides from the supporting grate for overnight aging 152 Dayton M. Petibone et al.

briefl y inverting them over the water bath until a uniform layer of condensate is present on the slide also helps slow the rate of fi xative evaporation [ 14 ]. Churning during the mixing of the water condensate and organic fi xative is thought to facilitate metaphase cell spreading [11 ]. 1. Suspend cell pellets in methanol:glacial acetic acid fi xative, generally between 100 and 200 μL. How much fi xative to add to the cell pellet is determined intuitively, based on the size of the cell pellet and how densely the cells will appear on the slide. 2. Chill the microscope slides at 4 °C and place them face down over a grated support above the 70 °C bath until a uniform layer of condensate forms on the slides (Fig. 2a ). 3. Turn the slides on the grated support over, and then apply two drops per slide of 15–20 μL of cell suspension by dropping the cells from a height of 2–3 cm (Fig. 2b ). Leave the slides in place on the grated support until all the fi xative has completely evaporated. Do not cover with the water bath lid, as condensate may drip onto the slides and wash away the cells. 4. Peripheral blood samples do not provide a lot of cells. Thus, as many slides as possible should be made to provide suffi cient numbers of cells for scoring. Spleen samples provide more cells and typically have higher mitotic indices. If at all possible, for each animal make at least 6 slides, each with 2 drops of cells, and store any remaining cells in fi xative in tightly sealed tubes at 4 °C for additional slide preparation. 5. When dry, remove the slides from the water bath and age them overnight at room temperature. Slides containing metaphase cell spreads should be aged overnight at room temperature before FISH painting. Hybridizing freshly prepared slides (without being aged overnight) usually results in chromosomes with a poorly defi ned morphology. To store the slides following aging, place them in a slide box in a plastic heat- sealing pouch with sponge packs that absorb humidity. Blow

N2 gas into the pouch several times to replace the air completely. Fill the pouch with N2 gas before sealing with a heated pouch sealer. The slides sealed in the pouches should then be stored at −20 °C until used for FISH.

3.5 Fluorescence 1. To prepare for the FISH, at least 1 h before beginning hybrid- In Situ ization, prewarm 50 mL 70 % formamide in 2× SSC in a Coplin Hybridization (FISH) jar set in a 73 °C water bath; and check the temperature of the pre-denaturation solution with a thermometer to ensure that it is 73 °C. Unless otherwise stated, perform all washes in glass Coplin jars. In washes performed in a water bath, the water line should cover ¾ the height of the Coplin jar, in both the FISH hybridization and slide washing steps. Chromosome Painting of Mouse Peripheral Blood and Spleen Tissues 153

2. Remove the slide boxes from the freezer and allow them to reach room temperature before opening the sealed pouch. Select slides with suitable numbers of metaphase cells by inspecting them under phase contrast on an inverted microscope with a 20× objective. Mark the lateral edges of the slide with a pencil to identify the location of cell spreads to indicate where to apply the FISH probe. Gentle breathing on the slides will briefl y cover them with a layer of condensate and reveal the boundaries of the cell drop. 3. Dim the overhead lights, remove the FISH probe from the freezer, and allow it to reach room temperature. Minimize exposing the FISH paints to light while handling, as they will photo-degrade. Place an aliquot of the required volume of single- or multicolor FISH probe cocktail, corresponding to the number of hybridizations performed, into a 0.5 mL tube and return the FISH probe stock to −20 °C. Place the tube containing the FISH probe into a heating block set to 37 °C, cover with foil, and incubate for 10 min. Transfer the FISH probes to a heating block set to 65 °C, cover with foil, and incubate for 10 min, then back to the 37 °C heating block for 30 min. Note: For molecular cytogenetic applications , the probe vendor must formulate and optimize the multicolor FISH cocktail because combining single probes into an effective cocktail for producing multicolor metaphase cells may not provide the necessary results. Multicolor FISH probes are often a custom item. Thus , it is recommended that a probe vendor able to meet the needs of your project be identifi ed , and if possible , establish a contract to ensure uninterrupted delivery during the course of the study . 4. While the FISH probe cocktail is incubating, pre-denature the slides sequentially in 70 % formamide in 2× SSC at 73 °C for 5 min, and then dehydrate the slides by treating 5 min each in a 70 %, 85 %, and absolute ethanol series at room temperature. Place the slides on a 37 °C slide moat lined with aluminum foil and allow them to dry. Note: Over denaturation of DNA during pre -denaturation step often leads to distortion of chromosome morphology and inhibits DAPI intercalation . 5. Retrieve the FISH probe from the 37 °C heating block, aliquot 15 μL onto the cell spreads, and cover with a 25 mm square cover glass. Remove any air bubbles trapped under the cover glass by fi rmly but gently holding the cover glass in place on the slide and pushing air bubbles out from under the cover glass with a soft pencil eraser. Incubate the slides for 30 min at 37 °C on a slide moat covered with foil. 154 Dayton M. Petibone et al.

6. Transfer the slides to a 70 °C slide moat for 5 min to allow the DNA in both the cells and the FISH probe to denature. Return the slides to the 37 °C slide moat lined with aluminum foil. Using a 3 mL syringe (without a needle) ﬁ lled with rubber cement; seal the edges of the cover glass on the slides to prevent evaporation of the FISH probe cocktail from the slide. Cover the slide moat lid with aluminum foil to protect the slides from light and hybridize the slides at 37 °C for 48 h. Note : A rubber cement that dries ﬂ exible , instead of one that dries hard , allows for easy removal of the cover glass following hybridization .

3.6 FISH Slide 1. Following FISH probe hybridization, use forceps to remove the Washing and Antifade rubber cement and cover slips, being careful not to drag the cover Mounting slip across the surface of the slide, as this will scratch the metaphase and interphase cells off the slides. 2. Wash the slides three times for 5 min each in 50 % formamide in 2× SSC and once each in 0.1 % Triton-X in 2× SSC and 2× SSC. Perform all washes at 45 °C in a water bath with the water line covering ¾ of the Coplin jar, agitating the slides for 5 s every minute.

3. Remove the slides, rinsing brieﬂ y in a beaker containing dH 2 O and dry for approximately 1 h in a dark place such as a drawer. Mount the slides with 25 μL Slowfade/DAPI mounting solution and cover with a 25 mm 2 cover glass. The slides are now ready for coding and distribution to the slide readers.

3.7 Slide Scoring 1. The slides should be coded and scored in a blinded fashion to avoid observer bias. A person who is not involved in the slide reading, e.g., the supervisor or primary investigator, should allocate all coded slides to the scorers. The slide codes are written or printed on adhesive labels together with the study number; apply these labels to the slides as to thoroughly obscure the original label. The backside of the slide, under the label, may also be obscured with a black felt pen if desired. Note: The use of coded slides will blind the slide readers to the treatments used in this study and prevent subjective bias during analysis. During scoring , slide readers can scan slides at 40 × ( oil ) using a DAPI - FITC - Cy3 triple band pass ﬁ lter to rapidly locate FISH-painted metaphase cells. When the scorer locates a metaphase cell , switch the microscope turret to view the cell with the 100 × ( oil ) objective to check for chromosome aberrations. If an aberrant metaphase cell is located , use a monochrome camera to capture an image with each of the DAPI , FITC , and Cy3 single band pass ﬁ lters to create an image overlay . 2. Standard criteria for scoring FISH-painted metaphase cells have been established. These are: (a) the cell must be intact, Chromosome Painting of Mouse Peripheral Blood and Spleen Tissues 155

Fig. 3 Example of a mouse metaphase cell containing aberrant chromosomes. Chromosomes 1 and 3 are painted red and chromosomes 2 and 8 are simultaneously painted green. The cell contains a reciprocal translocation where a segment of a green chromosome translocated onto a red chromosome (t(Bc) [16 ], is shown with a yellow arrow ) and a segment of a red chromosome translocated onto a green chromosome (t(Cb), is shown with a white arrow )

(b) all of the centromeres must be visible under the DAPI fi lter, and (c) all the centromeres from the FISH-painted chromosomes must be present. The protocol for aberration identifi cation and nomenclature terminology (PAINT) system provides nomenclature and classifi cation criteria for identifi ed aberrations [ 15 , 16 ]. This scoring system gives one or more symbols that describe particular chromosome aberration events within a cell. For example, aberration types are categorized as t = translocation, dic = dicentric, and ace = acentric fragment. Categorization of aberrant events is according to color, where “A or a” represent an unpainted, DAPI counterstained chromosome, “B or b” represent a red painted chromosome, and “C or c” represents a green painted chromosome. The upper case letters (A, B, and C) each represent a section of chromosome possessing a centromere, while the lower case letters (a, b, and c) each represent a section of chromosome without a centromere. An example of a multicolor FISH-painted mouse metaphase cell containing a reciprocal translocation is shown in Fig. 3 . To determine what fraction of inter-chromosomal exchanges is detectable within a specifi c organism and with a particular combination of painting probes, use the formula 2 pq , where p represents the percentage of total chromosomes 156 Dayton M. Petibone et al.

painted (e.g., either green or red) and q represents the fraction of the genome that is counterstained with DAPI [ 17 ]. Tucker has described in detail the calculations for the fraction of the genome covered and how to determine the fraction of detectable exchanges using two color combinations of painted chromosomes (and the counterstain) [8 ]. 3. Scoring is accomplished using an appropriate fl uorescence microscope and recording the number of metaphase cells on hand counters. The scorers should record their name, date, slide code, number of metaphase cells scored, and any data pertaining to abnormal cells (i.e., aberration types and Vernier coordinates) onto the score sheets. Storing the slides in slide boxes at 4 °C will slow degradation of the FISH probe. Scoring the slides within 1–2 weeks following hybridization will ensure that the FISH probes have not degraded. The use of high- resolution cameras to document aberrant cells, or those that are suspected to be aberrant, is recommended given that FISH probes can quickly fade during analysis or degrade in storage following analysis. Photographic documentation allows review and confi rmation of any aberrant cells identifi ed in the study. 4. An Excel spreadsheet-based system can be used for entry and quality control (a “master calculation spreadsheet”) of structural chromosome aberration data obtained by FISH whole chromosome painting. This system includes data entry spreadsheets for use by slide readers, and may be used either in an electronic form or as a hard copy if a computer is not available at the microscope workstation. The quality control spreadsheet compiles data from individual specimen slides, and may also be used for amassing data for specimens within a treatment group. The master calculation spreadsheet contains formulas that tally the type and frequency of a wide range of structural chromosome aberrations. In addition, the spreadsheet contains calculations to check for and identify data entry errors. Once data collection for an individual sample is complete, inserting a SUMLINK of the totals pasted into a master calculation spreadsheet provides a fi nal con- solidation of the data for statistical analyses. In order to insert a SUMLINK, copy the totals row from an individual sample spreadsheet. In the master calculation spreadsheet, click the fi rst cell in the row that you wish the data to be pasted. Go the “edit” menu on the tool bar, and select “paste special…” to open the dialog box. Select SUMLINK in the lower left corner. Any subsequent data entered into the individual calculation sheets will automatically be updated in the SUMLINK upon opening the master calculation sheet. Note : Templates for the data entry spreadsheet and master calculation spreadsheet are available upon request from the author . Chromosome Painting of Mouse Peripheral Blood and Spleen Tissues 157

4 Molecular Cytogenetics in Regulatory Science and Genotoxicity Assays

Assessment of chromosome aberration frequencies, particularly stable rearrangements such as translocations, insertions, and dicentrics are of interest to regulatory science. These types of chromosome rearrangements are not easily evaluated by other (e.g., Giemsa) staining methods and the level of cytogenetic damage may be underestimated [18 , 19 ]. Therefore, molecular cytogenetics assessment of stable chromosome rearrangements provides for a more comprehensive and accurate estimation of cytogenetic damage for regulatory science and genotoxicity assay applications. However, cells with chromosome aberrations are a rare event, which necessitates screening large numbers of metaphase cells. To make analysis of FISH-painted metaphase cells more manageable, future studies may rely on automated microscopy systems equipped with software to scan slides and identify metaphase cells, capture images and possibly classify aberrations. Even with its limitations, FISH of mouse metaphase cells is a useful tool in risk assessment studies. When used in conjunction with other endpoints, such as micronucleus and the Comet assay, FISH molecular cytogenetics provide a comprehensive assessment of a test chemical’s genotoxicity.

Declaration

The views presented in this article are those of the authors and do not necessarily reﬂ ect those of the U.S. Food and Drug Administration.

References

1. Ceppi M, Gallo F, Bonassi S (2011) Study 6. Tucker JD, Marples B, Ramsey MJ et al (2004) design and statistical analysis of data in human Persistence of chromosome aberrations in mice population studies with the micronucleus assay. acutely exposed to 56Fe+26 ions. Radiat Res Mutagenesis 26(1):247–252 161(6):648–655 2. Feneh M, Bonassi S (2011) The effect of age, 7. Breneman JW, Swiger RR, Ramsey MJ et al gender, diet and lifestyle on DNA damage mea- (1995) The development of painting probes sured using micronucleus frequency in human for dual-color and multiple chromosome analy- peripheral blood lymphocytes. Mutagenesis sis in the mouse. Cytogenet Cell Genet 26(1):43–49 68(3–4):197–202 3. Mateuca R, Lombaert N, Aka PV et al (2006) 8. Tucker JD (2010) Chromosome translocations Chromosomal changes: induction, detection and assessing human exposure to adverse envi- methods and applicability in human biomoni- ronmental agents. Environ Mol Mutagen toring. Biochimie 88(11):1515–1531 51(8–9):815–824 4. Ballarini F, Ottolenghi A (2004) Models of 9. Tucker JD, Christensen ML (1987) Effects chromosome aberration induction: an example of anticoagulants upon sister-chromatid based on radiation track structure. Cytogenet exchanges, cell-cycle kinetics, and mitotic Genome Res 104(1–4):149–156 index in human peripheral lymphocytes. Mutat 5. Bailey SM, Bedford JS (2006) Studies on chro- Res 190(3):225–228 mosome aberration induction: what can they 10. Delano MD (1995) Simple physical cons- tell us about DNA repair? DNA Repair (Amst) traints in hemolysis. J Theor Biol 175(4): 5(9–10):1171–1181 517–524 158 Dayton M. Petibone et al.

11. Claussen U, Michel S, Mühlig P et al (2002) aberrations detected by whole chromosome Demystifying chromosome preparation and painting. Mutat Res 347(1):21–24 the implications for the concept of chromo- 16. Tucker JD, Morgan WF, Awa AA et al (1995) some condensation during mitosis. Cytogenet A proposed system for scoring structural aber- Genome Res 98(2–3):136–146 rations detected by chromosome painting. 12. Deng W, Tsao SW, Lucas JN et al (2003) Cytogenet Cell Genet 68(3–4):211–221 A new method for improving metaphase 17. Lucas JN, Tenjin T, Straume T et al (1989) chromosome spreading. Cytometry A 51(1): Rapid human chromosome aberration analysis 46–51 using fl uorescence in situ hybridization. Int J 13. Henegariu O, Heerema NA, Lowe Wright L Radiat Biol 56(1):35–44. Erratum. Int J Radiat et al (2001) Improvements in cytogenetic slide Biol 56:201(1989) preparation: controlled chromosome spread- 18. Zhang L, Rothman N, Wang Y et al (1999) ing, chemical aging and gradual denaturing. Benzene increases aneuploidy in the lympho- Cytometry 43(2):101–109 cytes of exposed workers: a comparison of data 14. Petibone DM, Morris SM, Hotchkiss CE et al obtained by fl uorescence in situ hybridization (2008) Technique for culturing Macaca mulatta in interphase and metaphase cells. Environ Mol peripheral blood lymphocytes for fl uorescence Mutagen 34(4):260–268 in situ hybridization of whole chromosome 19. Smith MT, Zhang L, Wang Y et al (1998) paints. Mutat Res 653(1–2):76–81 Increased translocations and aneusomy in chro- 15. Tucker JD, Morgan WF, Awa AA et al (1995) mosomes 8 and 21 among workers exposed to PAINT: a proposed nomenclature for structural benzene. Cancer Res 58(10): 2176–2181 Chapter 9

T-Cell Receptor Mutation Assay for Monitoring Human Genotoxic Exposure

Seishi Kyoizumi

Abstract

Mutant T-cells defective in T-cell receptor (TCR) α or β gene expression among CD4+ T-cells can be detected as CD3− CD4+ cells by two-color fl ow cytometry using anti-CD3 and anti-CD4 monoclonal antibodies labeled with different fl uorescent dyes. This is because incomplete TCR αβ/CD3 complexes are not transported to the surface membrane. TCR mutations are spontaneously generated at a frequency of about 2 × 10 −4 in peripheral CD4+ T-cells, and the mutant frequencies are dose-dependently increased by exposure to genotoxic substances such as ionizing radiation and chemicals. The TCR mutation assay was developed in 1990, and this method has since been applied to monitoring human genotoxic exposure. It has been proved that the assay has advantage in detection of chronic exposure in large-scale studies. Here, I will briefl y review past studies reporting the applications, and describe methods for both preparation of the target cells and detection of the mutant cells.

Key words Somatic mutation , Flow cytometry , T-cell receptor , CD4 T-cell , Human peripheral blood

1 Introduction

Monitoring of somatic mutations in vivo is useful in evaluating cancer risk from exposure to environmental genotoxic substances, including ionizing radiation and chemicals. Assays of in vivo somatic mutations have been established for various target genes [1 ]. Of these assays, the T-cell receptor (TCR) mutation assay enables reproducible measurement of mutant frequencies (Mfs) at the TCR α and β genes of peripheral mature CD4+ T-cells in humans [2 ]. TCR mutation assay can be applied to individuals without any restriction of genetic background, in contrast to erythrocyte glycophorin A [3 , 4 ] and T-cell HLA class I [5 , 6 ] mutation assays. Furthermore, because the TCR mutation assay does not require long-term culture, whereas HPRT mutation assay does [7 , 8 ], and uses the general ﬂ ow cytometric method for lymphocytes, measurement can be completed within 4 h for one blood sample.

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_9, © Springer Science+Business Media New York 2014 159 160 Seishi Kyoizumi

TCR αβ are expressed on the cell surface of normal peripheral CD4 and CD8 T-cells. In the majority of these T-cell populations, only one of the two alleles of each TCR chain gene is active in protein expression, although it has been reported that a minor T-cell subpopulation co-expresses dual α or β chains [9 , 10 ]. Inactivation of one allele is due to allelic exclusion mechanisms [ 11 ], so mutants that do not express TCR can be generated in the majority of a T-cell population by a single inactivation event, even though TCR genes are autosomally located. In addition, TCR α and β chains can be expressed on the cell surface only after formation of large molecular complexes with CD3-γ, δ, ε, ζ, and η chains. If the expression of TCR α or β chain gene is inactivated, the TCR αβ/CD3 complex cannot be transported to the cellular membrane, and defective complexes accumulate in the cytoplasm [ 2 , 12 ]. Thus, defective mutations in TCR α or β genes among CD4+ T-cells can be detected as CD3− CD4+ mutant cells by fl ow cytometry using monoclonal antibody against CD3 and CD4 molecules. Although CD4 + TCR γδ + T-cells have been found in peripheral blood [13 ], they are too rare to infl uence TCR Mf. Specifi cally, the fraction of CD3 − cells in a population of mature CD4+ T-cells is considered to be the total TCR-α and TCR-β gene Mf in CD4+ T-cells. The background Mf of CD3− cells in populations of human mature CD4 + T-cells increases signifi cantly with age [2 , 14 ], but is about 2 × 10 −4 [2 ]. TCR mutants were found to be dose-dependently induced in normal CD4 + T-cells and in a lymphoma cell line by in vitro exposure to ionizing radiation [ 15 – 17 ] or chemicals [16 ]. Although expression of a TCR mutant phenotype can require as long as several months in vivo, the assay has been improved to shorten the expression time of the TCR mutant phenotype by using growth stimulation of lymphocytes in culture [ 17 ]. The Mf of TCR has been found to be elevated in patients having autosomal recessive inherited diseases with defective DNA repair and premature aging, such as ataxia telangiectasia [ 2 , 18 ], Fanconi’s anemia [2 ], and Werner’s syndrome [14 ]. The TCR mutation assay can be used to monitor human exposure to radiation. It has, for example, been applied to lymphocytes from cancer patients who had recently received radiotherapy [17 , 19 , 20 ], from patients who had been treated during the 1930s and 1940s with Thorotrast-a colloidal preparation of radioactive thorium-232 used as a radiological contrast medium [19 , 21 ], from a person who was heavily exposed to radiation during the 1986 Chernobyl accident [19 ], from cleanup workers in the Chernobyl accident [22 ], from residents in a contaminated area near Chelyabinsk in Russia [23 , 24 ] and Semipalatinsk in Kazakhstan [25 ], and from nuclear workers [26 ]. The assay has also been used to monitor genotoxic chemical- exposed donors, including cancer patients who had received chemotherapy [27 –29 ] and workers occupationally exposed to cancer drugs [ 29 –31 ] and lead [32 –35 ]. T-Cell Receptor Mutation Assay 161

Notably, no signifi cant elevation has been detected in atomic bomb survivors who were exposed to high-dose radiation about 70 years ago [ 19 ]. This is consistent with the observation that elevated Mfs in radiotherapy patients decline gradually to background levels within about 10 years after exposure (half-life: about 2 years) [ 20 , 36 , 37 ]. In other words, if study subjects show signifi cant elevation of TCR Mfs, they have received recent exposure or are currently receiving chronic exposure. A mouse model was also established to demonstrate the in vivo kinetics and dose–response of radiation-induced TCR mutations [ 38 ]; precise methods for the mouse TCR mutation assay are described elsewhere [ 39 ]. In this system, expression of TCR mutation phenotype reached a peak about 2 weeks after whole-body irradiation and the Mf then decreased, with a half-life of about 2 weeks [38 ]. This chapter gives precise methods for the preparation of these target cells and the methods of fl ow cytometric detection used to quantify fractions of CD3 − cells among human CD4+ T-cell populations.

2 Materials

2.1 Preparation 1. Heparinized peripheral blood (3–5 mL). of Human Peripheral 2. Ficoll–Hypaque solution (speciﬁ c density; 1.077) (e.g., Lymphocyte Blood Separation Medium; ICN Biomedicals, Inc., Irvine, CA). Mononuclear Cells 3. 15 mL polypropylene centrifuge tube. 4. Phosphate-buffered saline (PBS) (e.g., Sigma Chemicals Co., St. Louis, MO). 5. PBS containing fetal bovine serum (FBS) (e.g., Thermo Scientiﬁ c HyClone, South Logan, UT) (2.5 %) (heat inactivated for 30 min, 56 °C) (PBS-S). 6. Hemocytometer (e.g., Becton Dickinson and Company [BD], Franklin Lakes, MD). 7. Turk’s solution (e.g., MERCK, Darmstadt, Germany). 8. Trypan blue stain (0.4 %) (e.g., Invitrogen).

2.2 Immuno- 1. Fluorescein isothiocyanate (FITC)-labeled anti-human CD4 ﬂ uorescence Staining monoclonal antibody (clone: SK3, e.g., BD). 2. Phycoerythrin (PE)-labeled anti-human CD3ε monoclonal antibody (clone: SK7, e.g., BD). 3. PBS containing NaN3 (0.01 %) and FCS (1 %) (PBS-NS). 4. PBS-NS containing propidium iodide (10 μg/mL). 5. 1.5 mL Eppendorf tube. 6. 5 mL polystyrene round-bottom tube (e.g., BD). 162 Seishi Kyoizumi

2.3 Flow Cytometry 1. Flow cytometer (e.g., CyAn™ ADP Analyzer, Beckman Coulter, Indianapolis, IN) with computer software installed for data acquisition and analysis (e.g., Summit, Beckman Coulter). [ Note : BD FACSCalibur installed with BD FACStation is also usable for this assay].

3 Methods

3.1 Preparation 1. Place heparinized blood (3–5 mL) into 15 mL centrifuge tube. of Human Peripheral [Note : Tubes for one-step mononuclear cell separation from Blood whole blood are commercially available (e.g., BD Vacutainer Mononuclear Cells CPT tube). These tubes contain anticoagulant (sodium heparin or sodium citrate) and a cell separation medium that is composed of a polyester gel and a density gradient liquid]. 2. Add an equal volume of PBS at room temperature and mix well. 3. Slowly layer the Ficoll–Hypaque solution underneath the blood/PBS mixture by placing the tip of the pipette containing the Ficoll–Hypaque at the bottom of the sample tube. Use 3 mL Ficoll–Hypaque per 10 mL blood/PBS mixture. 4. Centrifuge for a total of 30 min at 400 × g at room temperature with no brake (raise the centrifuge speed slowly to 400 × g ). 5. Using a pipette, remove the upper layer that contains the plasma and platelets. Using another pipette, transfer the mononuclear cell layer (interface between the upper and Ficoll– Hypaque layers) to a new 15 mL centrifuge tube. 6. Wash cells by adding excess PBS-S (about three times the volume of the mononuclear cell layer) at room temperature and centrifuging 10 min at 510 × g . 7. Discard supernatant and resuspend cells in 10 mL PBS-S and centrifuge for 10 min at 240 × g . 8. Repeat step 7. 9. Discard supernatant and resuspend cells in 1 mL PBS-S. 10. Count mononuclear cells with Turk’s solution using a hemocytometer, and calculate cell yield. Use trypan blue exclusion to determine cell viability. Average yield will be about 1 × 10 6 mononuclear cells from 1 mL blood.

3.2 Immuno- 1. Transfer 2 × 106 human peripheral blood mononuclear cells ﬂ uorescence Staining suspended in PBS-NS to 1.5 mL Eppendorf tube and centrifuge at 340 × g , 4 °C for 2 min. 2. Discard supernatant and add 2 μg each of FITC-labeled antihuman CD4 and PE-labeled anti-human CD3 antibodies to the cell pellet, mix well, and incubate for 30 min on ice. T-Cell Receptor Mutation Assay 163

3. Wash cells by adding 0.75 mL of PBS-NS and centrifuging at 340 × g , 4 °C for 2 min. 4. Discard supernatant and resuspend cells in 0.5 mL of PBS-NS containing propidium iodide to stain dead cells. [ Note : If the fl ow cytometer is installed with a violet laser, 4′, 6-diamidino- 2-phenylindole (DAPI) (Invitrogen) at the fi nal concentration of 1 μg/mL can be used for dead-cell staining]. 5. Transfer cell suspension to 5 mL polystyrene tube for fl ow cytometry.

3.3 Flow Cytometry 1. TCR mutant CD4+ T-cells (CD3- CD4+ ) can be measured using a fl ow cytometer installed with operation and analysis software. Set up the fl ow cytometer and the analysis software, and optimize setting according to manufacturer’s instructions. 2. First, apply a small number of the stained lymphocytes (about 1,000 events) to the fl ow cytometer. Set a gate in the region for a lymphocyte fraction on the forward and side light scatter (FSC and SSC) profi le (Fig. 1a ). 3. Acquire and store FL1 (CD4 FITC fl uorescence) and FL2 (CD3 PE fl uorescence) data for a minimum of 500,000 lymphocyte-gated events. [ Note : Four-parameter (FSC, SSC, FL1, and FL2) correlated data can be acquired and stored if disk storage space is large enough. Lymphocyte gate should be set on the light scatter profi le for the mutant analysis of stored four-parameter data]. 4. Display acquired data on the screen in histograms of FL-1 (CD4) and FL-2 (CD3) and in density plot of FL1 and FL2 correlated data (Fig. 1b ). Obtain the mode fl uorescence intensities (channel number) of the FL-1 (CD4) and FL-2 (CD3) of normal CD3 + CD4+ cell population in the histograms by gating this population in the density plot (gate out the propidium iodide-stained dead cells from the population) (Fig. 1b ). [ Note : In the case where CD4 low CD3− cell population overlaps the mutant window, this cell population may be eliminated by staining with anti-CD14 antibody (e.g., APC-Cy7-labeled antibody, Invitrogen), as shown in Fig. 1c, d ]. 5. Set a mutant window on the region for CD3 − CD4+ in the density plot as follows. Set the left and right limits of FL1 at half and double values, respectively, of the mode intensity of FL1 (CD4) for normal CD3 + CD4+ cells. Set the upper limit of the FL-2 for the mutant window at 1/25, the mode intensity of CD3 for normal CD3 + CD4+ cells as mentioned above, and the lower limit at 10 0 . [ Note : The mutant window may be set using other reasonable rules. For example, the upper limit of the mutant window can be set at the value of the mean plus 3 164 Seishi Kyoizumi

ab4 10

3 10 + Total CD4 ) (n = 23.3 104

102 SSC

1 - + 10 FL2(PE/CD3) CD3 CD4 Lymphocyte (n = 110)

gate -4 100 Mf = 4.7 10 0 20K 40K 60K FSC FL1(FITC/CD4) cd 104

103 Total CD4+ (n = 22.8 104)

102 SSC

1 - +

10 FL2(PE/CD3) CD3 CD4 (n = 105)

0 -4 10 Mf = 4.6 10 100 101 102 103 104 FL3(APC-Cy7/CD14) FL1(FITC/CD4)

Fig. 1 ( a , b ) Representative fl ow cytograms of human peripheral blood mononuclear cells stained with FITC- labeled anti-CD4 (FL1) and PE-labeled anti-CD3 (FL2) antibodies. (c , d ) Elimination of CD3− CD4low cells overlapping into the mutant window by staining with these antibodies in combination with allophycocyanin-Cy7 (APC-Cy7)-labeled anti-CD14 (FL3) antibody. (a ) A gate for lymphocytes on forward and side light scatter (FSC and SSC) profi le (dot plot ). ( b) Windows for total CD4+ and mutant CD3− CD4 + T-cells on fl uorescence profi les (density plot ). ( c ) A gate for CD14− cells on SSC and CD14 profi le. ( d) Windows for total CD4+ and mutant CD3− CD4+ T-cells on fl uorescence profi les after gating out of CD14+ cells in c . The number of events in each window is shown to the right of panels (b , d ). The Mf was calculated as the number of events in the mutant window divided by the number of events among all CD4+ T-cells. The majority of CD3− CD4low cells (open arrow ) are eliminated by gating out of CD14 + cells (b , d ). Events representing the highest FL2 fl uorescence ( closed arrow) are dead cells stained with propidium iodide (b , d ). The blood sample was obtained from a laboratory control (male, age 59)

standard deviations of PE ﬂ uorescence intensity (FL2) of CD3− CD4− cells [22 ]]. 6. Calculate the mutant frequency (Mf) as the number of events in the mutant window divided by the total number of events corresponding to CD4 + cells (Fig. 1b ). T-Cell Receptor Mutation Assay 165

3.4 Statistical To demonstrate signifi cant changes of TCR Mfs in given study Analysis subjects, selection of study subjects and statistical analysis should be conducted according to proper epidemiological methods. The study subjects should comprise exposed and control groups. Because TCR Mf has been found to signifi cantly increase with increase of age [ 2 , 14 ], and to differ between males and females [19 ], age and gender distributions should be matched between exposed and control groups or adjusted in the statistical analysis. The difference in TCR Mfs between control and exposed groups can be analyzed by the Mann–Whitney U test if exposed and control groups are matched. If the Mfs can be assumed to be normally distributed, the difference can be analyzed by a t -test. If the Mf values are normally distributed and exposure doses are defi ned, multiple linear regression analysis can be applied using age, gender, and exposure dose as variables. If the Mfs are not normally distributed, log transformation of the Mfs may be required [ 19 , 40 ]. Further analyzes, such as with case–control data, may benefi t from consultation with a biostatistician.

Acknowledgements

The authors would like to acknowledge Y. Kubo and M. Yamaoka for excellent technical help, and Y. Kusunoki and T. Hayashi for valuable suggestions.

References

1. Cole J, Skopek TR (1994) Somatic mutant Mutations in human lymphocytes studied by frequency, mutation rates and mutational spec- an HLA selection system. Mutat Res 198: tra in the human population in vivo. Mutat Res 221–226 304:33–105, International Commission for 6. Kushiro J, Hirai Y, Kusunoki Y, Kyoizumi S, Protection Against Environmental Mutagens Kodama Y, Wakisaka A, Jeffreys A, Cologne and Carcinogens. Working paper no. 3 JB, Dohi K, Nakamura N et al (1992) 2. Kyoizumi S, Akiyama M, Hirai Y, Kusunoki Y, Development of a fl ow-cytometric HLA-A Tanabe K, Umeki S (1990) Spontaneous loss and locus mutation assay for human peripheral alteration of antigen receptor expression in mature blood lymphocytes. Mutat Res 272:17–29 CD4+ T cells. J Exp Med 171:1981–1999 7. Hakoda M, Akiyama M, Kyoizumi S, Kobuke 3. Kyoizumi S, Nakamura N, Hakoda M, Awa K, Awa AA, Yamakido M (1988) Measurement AA, Bean MA, Jensen RH, Akiyama M (1989) of in vivo HGPRT-defi cient mutant cell fre- Detection of somatic mutations at the gly- quency using a modifi ed method for cloning cophorin A locus in erythrocytes of atomic human peripheral blood T-lymphocytes. bomb survivors using a single beam fl ow Mutat Res 197:161–169 sorter. Cancer Res 49:581–588 8. Albertini RJ, Nicklas JA, O’Neill JP, Robison 4. Langlois RG, Nisbet BA, Bigbee WL, Ridinger SH (1990) In vivo somatic mutations in DN, Jensen RH (1990) An improved fl ow humans: measurement and analysis. Annu Rev cytometric assay for somatic mutations at the Genet 24:305–326 glycophorin A locus in humans. Cytometry 9. Padovan E, Casorati G, Dellabona P, Meyer S, 11:513–521 Brockhaus M, Lanzavecchia A (1993) 5. Janatipour M, Trainor KJ, Kutlaca R, Bennett Expression of two T cell receptor alpha chains: G, Hay J, Turner DR, Morley AA (1988) dual receptor T cells. Science 262:422–424 166 Seishi Kyoizumi

10. Davodeau F, Peyrat MA, Romagne F, Necker and dicentric chromosome frequencies in lym- A, Hallet MM, Vie H, Bonneville M (1995) phocytes from radiotherapy patients. J Radiat Dual T cell receptor beta chain expression on Res 35:92–103 human T lymphocytes. J Exp Med 181: 21. Umeki S, Kyoizumi S, Kusunoki Y, Nakamura 1391–1398 N, Sasaki M, Mori T, Ishikawa Y, Cologne JB, 11. Brady BL, Steinel NC, Bassing CH (2010) Akiyama M (1991) Flow cytometric measure- Antigen receptor allelic exclusion: an update ments of somatic cell mutations in Thorotrast and reappraisal. J Immunol 185:3801–3808 patients. Jpn J Cancer Res 82:1349–1353 12. Clevers H, Alarcon B, Wileman T, Terhorst C 22. Saenko AS, Zamulaeva IA, Smirnova SG, (1988) The T cell receptor/CD3 complex: a Orlova NV, Selivanova EI, Saenko VA, dynamic protein ensemble. Annu Rev Matveeva NP, Kaplan MA, Nugis VY, Immunol 6:629–662 Nadezhina NM, Tsyb AF (1998) 13. Kyoizumi S, Akiyama M, Hirai Y, Kusunoki Y Determination of somatic mutant frequencies (1991) Isolation and characterization of at glycophorin A and T-cell receptor loci for human peripheral blood CD4 + T cell clones biodosimetry of prolonged irradiation. Int J expressing gamma delta T cell receptors. Radiat Biol 73:613–618 Immunol Lett 29:197–203 23. Blinova EA, Veremeyeva GA, Akleyev AV 14. Kyoizumi S, Kusunoki Y, Seyama T, Hatamochi (2012) Apoptosis of peripheral blood lympho- A, Goto M (1998) In vivo somatic mutations cytes and mutations in the gene of the T-cell in Werner’s syndrome. Hum Genet 103: receptor in survivors of chronic radiation 405–410 exposure. Health Phys 103:58–60 15. Iwamoto KS, Mizuno T, Ito T, Tsuyama N, 24. Veremeyeva G, Akushevich I, Pochukhailova Kyoizumi S, Seyama T (1996) Gain-of- T, Blinova E, Varfolomeyeva T, Ploshchanskaya function p53 mutations enhance alteration of O, Khudyakova O, Vozilova A, Kozionova O, the T-cell receptor following X-irradiation, Akleyev A (2010) Long-term cellular effects in independently of the cell cycle and cell sur- humans chronically exposed to ionizing radia- vival. Cancer Res 56:3862–3865 tion. Health Phys 99:337–346 16. Mei N, Kunugita N, Nomoto S, Norimura T 25. Taooka Y, Takeichi N, Noso Y, Kawano N, (1996) Comparison of the frequency of T-cell Apsalikov KN, Hoshi M (2006) Increased receptor mutants and thioguanine resistance T-cell receptor mutation frequency in induced by X-rays and ethylnitrosourea in cul- radiation-exposed residents living near the tured human blood T-lymphocytes. Mutat Res Semipalatinsk nuclear test site. J Radiat Res 357:191–197 47(Suppl A):A179–A181 17. Ishioka N, Umeki S, Hirai Y, Akiyama M, 26. Ivanov VK, Kashcheev VV, Zamulaeva IA, Kodama T, Ohama K, Kyoizumi S (1997) Saenko AS, Orlova NV, Smirnova SG, Korelo Stimulated rapid expression in vitro for early AM, Gorsky AI, Maksioutov MA (2012) detection of in vivo T-cell receptor mutations Formation of potential radiation risk groups to induced by radiation exposure. Mutat Res render timely targeted medical care: lessons of 390:269–282 Chernobyl. Radiat Prot Dosimetry 151: 18. Lantelme E, Palermo B, Granziero L, 666–670 Mantovani S, Campanelli R, Monafo V, 27. Kubota M, Lin YW, Hamahata K, Sawada M, Lanzavecchia A, Giachino C (2000) Cutting Koishi S, Hirota H, Wakazono Y (2000) edge: recombinase-activating gene expression Cancer chemotherapy and somatic cell muta- and V(D)J recombination in CD4+CD3 low tion. Mutat Res 470:93–102 mature T lymphocytes. J Immunol 164: 28. Lanza A, Robustelli della Cuna FS, Zibera C, 3455–3459 Pedrazzoli P, Robustelli della Cuna G (1999) 19. Kyoizumi S, Umeki S, Akiyama M, Hirai Y, Somatic mutations at the T-cell antigen recep- Kusunoki Y, Nakamura N, Endoh K, Konishi tor in antineoplastic drug-exposed popula- J, Sasaki MS, Mori T et al (1992) Frequency of tions: comparison with sister chromatid mutant T lymphocytes defective in the expres- exchange frequency. Int Arch Occup Environ sion of the T-cell antigen receptor gene among Health 72:315–322 radiation-exposed people. Mutat Res 265: 29. Hongping D, Jianlin L, Meibian Z, Wei W, 173–180 Lifen J, Shijie C, Wei Z, Baohong W, Jiliang H 20. Iwamoto KS, Hirai Y, Umeki S, Kusunoki Y, (2006) Detecting the cytogenetic effects in Kyoizumi S, Kodama T, Ohama K, Nakamura workers occupationally exposed to vincristine N, Akiyama M (1994) A positive correlation with four genetic tests. Mutat Res 599: between T-cell-receptor mutant frequencies 152–159 T-Cell Receptor Mutation Assay 167

30. Deng H, Zhang M, He J, Wu W, Jin L, Zheng genotoxic effects of workers exposed to lead W, Lou J, Wang B (2005) Investigating genetic using micronucleus assay, comet assay and damage in workers occupationally exposed to TCR gene mutation test. Toxicology 223: methotrexate using three genetic end-points. 219–226 Mutagenesis 20:351–357 36. Vershenya S, Biko J, Drozd V, Lorenz R, 31. Jiang W, Lu Y, Chen Z, Chen S, Zhang M, Jin L, Reiners C, Hempel K (2004) Dose response Lou J, He J (2008) Studying the genotoxicity of for T-cell receptor (TCR) mutants in patients vincristine on human lymphocytes using comet repeatedly treated with 131 I for thyroid cancer. assay, micronucleus assay and TCR gene muta- Mutat Res 548:27–33 tion test in vitro. Toxicology 252:113–117 37. Umeki S, Kusunoki Y, Cologne JB, Iwamoto 32. Garcia-Leston J, Mendez J, Pasaro E, Laffon B KS, Hirai Y, Seyama T, Ohama K, Kyoizumi S (2010) Genotoxic effects of lead: an updated (1998) Lifespan of human memory T-cells in review. Environ Int 36:623–636 the absence of T-cell receptor expression. 33. Garcia-Leston J, Roma-Torres J, Vilares M, Immunol Lett 62:99–104 Pinto R, Cunha LM, Prista J, Teixeira JP, Mayan 38. Umeki S, Suzuki T, Kusunoki Y, Seyama T, O, Pasaro E, Mendez J, Laffon B (2011) Fujita S, Kyoizumi S (1997) Development of a Biomonitoring of a population of Portuguese mouse model for studying in vivo T-cell recep- workers exposed to lead. Mutat Res 721:81–88 tor mutations. Mutat Res 393:37–46 34. Garcia-Leston J, Roma-Torres J, Vilares M, 39. Kyoizumi S, Kusunoki Y, Hayashi T (2014) Pinto R, Prista J, Teixeira JP, Mayan O, Conde Flow cytometric quantiﬁ cation of mutant T J, Pingarilho M, Gaspar JF, Pasaro E, Mendez cells with altered expression of TCR: detecting J, Laffon B (2012) Genotoxic effects of occu- somatic mutations in humans and mice. In: pational exposure to lead and inﬂ uence of Keohavong P, Grant SG (eds) Molecular toxi- polymorphisms in genes involved in lead toxicology protocols, 2nd edn. Springer, cokinetics and in DNA repair. Environ Int New York, pp 245–253 43:29–36 40. Watson JV (1992) Flow cytometry data analy- 35. Chen Z, Lou J, Chen S, Zheng W, Wu W, Jin sis: Basic concepts and statistics. Cambridge L, Deng H, He J (2006) Evaluating the University Press, Cambridge Chapter 10

The Human RBC PIG-A Gene Mutation Assay

Vasily N. Dobrovolsky and Robert H. Heflich

Abstract

Mutation affects the expression of genes and the function of the proteins that they encode. Of relevance to safety assessments, somatic cell mutation in key endogenous tumor suppressor genes and the genes for cell-cycle regulators drives the carcinogenesis process. Safety assessments, though, typically measure mutations in reporter genes that have easily identifiable phenotypes. Mutation in genes that contribute to the biosynthesis of the anchor molecule glycosylphosphatidylinositol (GPI) may dramatically alter the complement of cell-surface markers at the exterior of the cell. Such aberrant mutant cells, deficient in GPI anchors and specific GPI-anchored cell-surface markers, can be differentiated from wild-type cells and enumerated on high-throughput flow cytometers. In humans, one of the enzymes involved in the synthesis of GPI is coded by the endogenous X-linked PIG-A gene, which makes flow cytometry-based analysis of GPI- deficient cells particularly sensitive to the detection of mutation in this gene. Hematopoietic cells and the peripheral blood, in general, are the most amenable cell types for detecting mutation in the PIG-A gene: blood cells don’t require extraordinary steps for antibody labeling and flow analysis. Labeling reagents (antibodies and stains) for the PIG-A assay are readily available from commercial sources, which makes most research and clinical laboratories proficient in the use of flow cytometry capable of detecting PIG-A mutation in human blood.

Key words Genotoxicity, Glycosylphosphatidylinositol (GPI) anchor, CD59 cell-surface marker, Paroxysmal nocturnal hemoglobinuria (PNH), Glycophorin, Erythrocytes, Peripheral blood, Flow cytometry

1 Introduction

1.1 Mutation The seminal work by Bruce Ames and his colleagues, demonstrating and Safety Evaluations that many known human and animal carcinogens are mutagens, made it clear that genotoxicity in general, and gene mutation in particular, are intimately involved in carcinogenesis [1]. Various types of heritable changes to DNA, including the loss and rearrangement of chromosomes, large multi-loci deletions, as well as small deletions and insertions, frameshifts, transitions and transversions in single genes, all have been implicated in the susceptibility to, and the initiation and progression of cancer. Mutation can inactivate tumor suppressor function, can cause an

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_10, © Springer Science+Business Media New York 2014 169 170 Vasily N. Dobrovolsky and Robert H. Heflich

imbalance in the networks of the cell-cycle control and DNA damage repair systems, and can permanently activate endogenous proto-oncogenes. For these reasons, the identification of potential mutagens in many respects equates to the identification of poten tial carcinogens. The sensitivity and specificity of the bacterial systems that Dr. Ames developed to identify carcinogens were far less than 100 %, so additional models were developed for detecting genotoxicity in cultures of eukaryotic cells and in animals. Today, government regulatory agencies, depending upon the scope of their authority, require that the products that they regulate be evaluated for safety in a battery of genotoxicity tests. For novel drugs intended for human use, the battery and general approach to testing for genotoxicity were recently updated in the S2(R1) guidance document by the International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH, for short), an organization consisting of repre- sentatives from national regulatory authorities and from the increasingly globalized pharmaceutical industry [2]. ICH S2(R1) describes the standard in vitro (in bacterial and eukaryotic cells) and in vivo tests (in mouse and rat) for genotoxicity assessment and the acceptable combinations of tests necessary for a comprehensive safety evaluation. Historically, testing batteries, including those recommended in ICH S2(R1) are biased toward measuring gene mutation in bacteria and measuring cytogenetic/chromosomal damage in eukaryotic cells and in vivo. There is no absolute requirement for measuring gene mutation in eukaryotic cells (either in vitro or in vivo), relegating the few available nonclinical in vivo gene mutation tests as optional or follow-up assays. A major reason for the reliance on cytogenetic tests for measuring in vivo genotoxicity is that it is difficult to measure gene mutation in cancer-relevant genes, and the majority of tests for mutation in surrogate targets (reporter genes) are limited by prac- ticality issues. The methodology for measuring in vivo mutation is cumbersome (e.g., requiring long-term maintenance of primary cell cultures or calling for extensive manipulation of tissue samples) and expensive (due to the use of transgenic animals and subchronic treatment schedules). The International Workshop on Genotoxicity Testing (IWGT) acknowledges the value of in vivo gene mutation testing for safety evaluations and, in conjunction with the International Life Sciences Institute/Health and Environmental Sciences Institute (ILSI-HESI), is currently evaluating a novel flow cytometry-based methodology for measuring mutation in the endogenous phosphatidylinositol glycan, class-A, gene (Pig-a for the gene in rodents and PIG-A for the human gene) in the cells of peripheral blood [3]. The basics of the assay are described in the following section. The Human RBC PIG-A Gene Mutation Assay 171

1.2 The Genesis Eukaryotic cells have extracellular surface markers (usually clusters of a Human Somatic of differentiation (CD) markers), anchored to the cytoplasmic Cell Gene Mutation membrane by glycosylphosphatidylinositol (GPI). GPI is synthe Assay: GPI, PIG-A, sized within the cell at the surface of the endoplasmic reticulum and PNH (ER), with the participation of about two dozen genes. The synthesis begins at the cytoplasmic side of the ER membrane on a phosphatidylinositol imbedded with its fatty chain in the membrane; then the intermediate product flips over and the synthesis continues in the ER lumen. In the lumen, independently synthesized CD polypeptides are conjugated with the complete GPI anchors, and the whole assembly is exported to the exterior of the cytoplasmic membrane [4]. Only one gene involved in the synthesis of GPI anchors, the PIG-A gene, resides on the X-chromosome in most (if not all) mammals and thus is present in a single functional copy in male and female cells. In male cells the PIG-A gene is physically a single- copy gene; in female cells the second (nonfunctional) copy of the PIG-A gene is present on the transcriptionally silenced X-chromosome. PIG-A codes for a catalytic subunit of acetylglu- cosamine transferase, which is responsible for the initial step of GPI biosynthesis. An inactivating mutation in the single functional PIG-A gene will result in disruption of GPI synthesis, and, subsequently, will cause a deficiency in GPI-anchored markers at the surface of the cell [5]. In humans, there is a rare disorder, paroxysmal nocturnal hemoglobinuria (PNH), which is characterized by an abnormally high fraction (up to 90 %) of cells in the bone marrow and peri pheral blood that are deficient in GPI-anchored markers [5–7]. In most cases, the disease is accompanied by clonal expansion of PIG-A mutant cells (i.e., cells that have the same PIG-A mutation). PNH is therefore a genetic disorder, though not inherited, but acquired. The origins of this disease (the expansion of the PIG-A mutant fraction) remain a mystery, but the evidence suggests that PIG-A mutant stem cells may colonize bone marrow and produce PIG-A mutant cells of all lineages after an episode of bone marrow failure, in which the surviving fraction of stem cells contains a PIG- A mutant cell that either was present spontaneously or was induced through prior exposure to a noxious agent [8]. Healthy humans have a small fraction of T-cells in peripheral blood with a phenotype consistent with PIG-A mutation (i.e., deficient in GPI anchors and having PIG-A mutation); the frequency of PIG-A mutant lymphocytes in normal individuals is comparable to the frequency of cells containing mutations in another endogenous X-linked gene, the hypoxanthine-guanine phosphoribosyltransferase gene (HPRT) [9, 10]. In patients with PNH, not only the nucleated cells have the characteristic PIG-A mutant phenotype. Red blood cells (RBCs), which develop from nucleated precursors, also are deficient in 172 Vasily N. Dobrovolsky and Robert H. Heflich

GPI-anchored markers, although the genetic origin of the phenotype cannot be easily confirmed in these DNA- and mostly RNA-free cells. RBCs, however, are the preferred cell population for the diagnosis or confirmation of PNH, since RBCs are the most abundant cell fraction in peripheral blood and only microliter volumes of blood are sufficient to clearly observe an elevated fraction of GPI-anchored marker-deficient cells [11, 12]. A small volume of peripheral blood can be labeled with fluorescent antibodies against a ubiquitous GPI-anchored marker, e.g., CD59, and then the sample is analyzed by flow cytometry. Any substantial fraction of marker-deficient RBCs will be visualized on a cytogram or a histogram as a population with reduced fluorescence. The flow cytometric diagnostic procedure may be refined and focused at different subsets of cells for a more detailed characterization of the disease. This may require using additional antibodies (e.g., for the positive identification of the cells of interest, such as lymphoid or myeloid subpopulations), additional preprocessing of peripheral blood (e.g., enriching the population of mononuclear cells by gradient centrifugation), etc. PNH diagnostic procedures form the basis for a PIG-A somatic cell mutation assay—the mutants lack GPI-anchored cell surface markers, while wild-type cells have a full complement of surface constituents. The pathways for biosynthesis of GPI and the expression of GPI-anchored markers in hematopoietic cells also are conserva- tive [13], so that analogous models for the detection of Pig-a mutation can be and have been developed for laboratory animals. In fact there has been great progress recently in the development of advanced protocols for the detection of Pig-a mutation in RBCs and T-cells from the rat and, to a lesser extent, the mouse [14]. Much has been learned about the expression time and the magnitude of the response in rats treated with various genotoxins, including weak and powerful gene mutagens, clastogens, ionizing radiation, and structurally related non-mutagens.

1.3 An Assay Araten et al. extended the basic RBC- and granulocyte-based flow for Detecting PIG-A cytometric protocols for PNH diagnosis to a small set of non-PNH Mutation in Human volunteers and determined that the normal subjects have a low Peripheral Blood RBCs frequency of cells in the peripheral blood consistent with PNH [15]. We have built upon Araten’s approach for analysis of mutant RBCs, modified the labeling protocol and sample analysis strategy (presented in the following section of this chapter in greater detail), and determined the frequency of CD59-deficient, presumed PIG-A mutants, in the blood of a large number of untreated self-identified healthy individuals of both sexes and from various ethnic backgrounds. We also have measured RBC PIG-A mutant frequencies in newly diagnosed cancer patients before the start of prescribed chemotherapy with genotoxic antineoplastic drugs and at several time points while on or post-chemotherapy [16]. The Human RBC PIG-A Gene Mutation Assay 173

In people with an unknown history of exposure to mutagenic compounds (presumably normal healthy individuals), the average background RBC PIG-A mutant frequency is about 5.3 × 10−6, although there are outliers. In one patient receiving genotoxic cisplatin, we measured a gradual increase in PIG-A mutant frequency from 5.3 × 10−6 (before the start of chemotherapy) to 16.5 × 10-6 over 6 months into the treatments. So far the data for human populations are scarce, but they suggest that the values for RBC PIG-A MF are stable in unexposed healthy individuals and may be increased in individuals exposed to genotoxins. A follow-up study indicated that cisplatin induces Pig-a mutation in rats [17], which suggests that the increase in PIG-A MF that was measured in humans is real. Although there are similarities between approaches for using PIG-A mutation to diagnose PNH and for measuring frequencies of PIG-A mutant cells in non-PNH individuals, there are also significant differences. As the frequency of true PIG-A mutant cells is expected to be low in non-PNH cases (in the 1 × 10−6 to 1 × 10−4 range), much lower than what would be considered a “population” in conventional flow cytometry applications, special attention must be paid to technical details in performing the assay as well as to the quality of materials and instrumentals used for the assay. Successful implementation of the PIG-A mutation detection assay has several requirements: 1. The tissue sample should be available as a single-cell cell suspension suitable for later processing on a flow cytometer. For that reason, PIG-A mutation detection assays have been established only for the cells from the peripheral blood (human and animal) and bone marrow (animal). Dissociation protocols for making cell suspensions out of solid tissues usually require enzymatic treatments, which could negatively impact GPI- anchored proteins and artificially increase the fraction of marker-deficient cells (i.e., generating false mutants). Peripheral blood should be collected in anticoagulant-containing tubes to avoid the clumping of cells and clogging the flow cytometer.

Heparin and K2EDTA can be used as anticoagulants; evacuated anticoagulant-coated blood collection tubes (vacutainer- type) are suitable. Ideally, the blood should be processed as soon as possible after collection; if shipping to an off-site labo-

ratory is required, then K2EDTA-coated collection tubes are preferred. The shipping should be done via overnight carrier on wet ice to avoid freezing and hemolysis of the samples. Ethanol or formaldehyde fixing of blood samples for long- term storage is not permissible: fixing causes permanent denaturation of the GPI-anchored antigens used in this protocol, so that they are no longer sufficiently reactive with the antibodies. 174 Vasily N. Dobrovolsky and Robert H. Heflich

2. The GPI-anchored cell-surface marker chosen as an indicator of PIG-A mutation should be expressed on the target cells ubiquitously and at a high level. The expression of CD markers is variable in different hematopoietic cells [18]. CD59 is an excellent marker for the human RBC assay, but it would be inappropriate for a lymphocyte PIG-A assay. 3. The antibody chosen for detecting GPI-anchored markers should be conjugated with a bright fluorophore; the use of secondary antibodies for the detection of GPI-anchored marker deficiency is discouraged. The antibody should have low nonspecific binding and provide substantial separation of the fluorescence signals measured in the wild-type labeled cells and mutant non-labeled cells. The emission of the fluorophore should not be in the portion of the spectrum where there may be excessive autofluorescence of the target cells. FITC- conjugated mouse antihuman CD59 (clone p282) is acceptable for the human RBC PIG-A assay. 4. The additional fluorescent antibody (or antibodies) used for the positive identification of target cells (or for the exclusion of nontarget cells) should have fluorescence emission spectra substantially separated from the emission of the main anti-GPI- anchored marker antibody in order to avoid performing compensation adjustments that, if applied inappropriately, may distort the results. The best practice is to use fluorophores that are excited by different lasers. The cell-type-specific marker also should be expressed on the target population ubiquitously and at a high level. Glycophorins are specific for RBCs and are highly expressed on the surface of RBCs. APC-conjugated mouse antihuman CD235ab (clone HIR2) is very specific for human RBCs; APC is one of the brightest fluorophores, and it is excited by a red laser, which, on many flow cytometers, is spatially separated from the blue laser employed for excitation of FITC- conjugated anti-CD59 antibody. With this combination of FITC- and APC-conjugated antibodies, compensation is not required for the precise measurement of specific fluorescence signals. Anti-glycophorin antibodies cause aggregation and (at higher levels) lysis of the RBCs from many species, including humans, mouse, rat, and rhesus macaque; a preliminary titration should be performed in order to determine the optimal ratio of blood to anti-CD235ab that provides acceptable fluorescent labeling, yet minimal aggregation and hemolysis. 5. The frequency of CD59-deficientPIG-A mutant RBCs often is determined only with a single sample of blood; in theory, the probability to detect k number of mutant cells in repeated analyses of samples from the same source with the average number of λ mutants in the sample is described by the Poisson lk ×e-l distribution, Pr(k, λ) = . When the true average number k!-1 The Human RBC PIG-A Gene Mutation Assay 175

of mutants in the sample is low (e.g., λ = 1), the probability of detecting zero mutants (k = 0) in a single sample will be the same as the probability of detecting one mutant (k = λ = 1), 1/e ≈ 0.37. In other words, the estimate of mutant frequency based on a sample containing a low number of measured mutants may be inaccurate. In addition, “zero” mutant frequencies complicate further statistical analysis of data (e.g., comparing responses determined in control and exposed groups, or before and after treatment, especially for weak responses). In order to estimate mutant frequency with greater accuracy and to reduce the chance of finding zero mutants in analyzed samples, the size of the sample should be relatively large, that is, as many RBCs should be processed in each sample as possible, and as many mutants should be counted as possible. In practical terms, there is a limit to how many events can be processed and stored as a continuous sample/file by flow cytometer acquisition software. A good starting point is to use a sample size that contains an average of five mutants, which will have a <0.01 probability of detecting zero mutants. With the average RBC PIG-A MF in a diverse healthy population of 5 × 10−6, that would be equivalent to analyzing of approximately 1 × 106 RBCs per sample. The analysis of 1 × 106 RBCs should be the lowest starting point. On a modern clinical flow cytometer (e.g., BD FACSCanto II, FACSAria I, or equivalent), the analysis of 1 × 106 single-cell RBCs may take less than 3 min. Consequently, the analysis of 2 or 3 × 106 RBCs can be achieved in a reasonable time frame while providing much more reliable mutant frequency data and greater statistical power for distinguishing between samples having small differences in mutant frequency. 6. As the frequency of mutant cells is extremely low by conventional flow cytometry standards, and often comparisons are conducted on data collected at different time points, it is important that the instruments used for analysis are verified for drift-free performance using manufacturer-suggested procedures (e.g., using factory-standardized beads as in the Cytometer Setup and Tracking (CST) procedure for BD flow cytometers). Fluidics pressure and the dimensions of the flow cell (e.g., nozzle size on some sorter models) should be chosen to maximize throughput and cell survival. For determining mutant frequencies, an instrument capable of fluorescence-activated cell sorting is not required: instruments capable of just fluorescence analysis can be used. Human blood samples (especially clinical samples) should be considered biohazardous and should be handled using personal protective equipment (PPE). After labeling, blood samples can be fixed in formaldehyde to inactivate potential biohazards. In the protocol below, fixation preserves the fluorescent properties of the cells stored at 4 °C 176 Vasily N. Dobrovolsky and Robert H. Heflich

for at least several days so that the analysis can be performed at a later time. If fixation is undesirable, then appropriate aerosol control measures should be employed on sorters, and, upon the completion of analysis, a decontamination routine should be performed on both sorters and analyzers. Human RBC PIG-A Assay on FACSAria I Flow Cytometer/Sorter

2 Materials

1. Blood collection. Evacuated K2EDTA-coated tubes (e.g., 4-mL purple top Vacutainer tubes, BD 367861); optional—styrofoam shipping container with pre-chilled Cool Packs. 2. Labeling. FITC antihuman CD59 antibody (clone p282(H19); BD Pharmingen, 555763); APC antihuman CD235ab (clone HIR2; BioLegend, 306608); tissue culture grade sterile Ca-, Mg-free Dulbecco’s phosphate buffered saline (DPBS); methanol-free 16 % formaldehyde (Ted Pella, Inc., 18505), dilute fresh to 1 % formaldehyde with DPBS; 35 μm strainer cap 5 mL flow cytometry tubes (BD, 352235); 1.5 mL microcentrifuge tubes and a microcentrifuge capable of generating 300 × g; fluid removal apparatus (e.g., VACUSAFE Comfort, Integra Biosciences). 3. Flow cytometry. Two-laser (blue 488 nm and red 633 nm, spatially separated) BD FACSAria I flow cytometer with a PMT equipped with 530/30 nm emission filter for detection of FITC fluorescence on the octagon array and a PMT with 660/20 nm emission filter for detection of APC fluorescence on the trigon array.

3 Methods

1. Collect blood by venipuncture into an anticoagulant-containing vacutainer tube. Keep on ice in the dark until ready for use. 2. Label the blood with antibodies: in a 1.5 mL microcentrifuge tube combine 200 μL of PBS, 20 μL of FITC-conjugated anti- CD59 antibody, 0.1 μL of APC-conjugated anti-CD235ab antibody, and 3 μL of anticoagulant-preserved blood. In parallel, prepare an unstained control, 200 μL DPBS plus 3 μL of blood, and prepare a mutant-mimic (single antibody) control, 200 μL DPBS, 3 μL of blood, and 0.1 μL APC-conjugated anti-CD235ab antibody. The controls may be prepared from any blood samples, but for the purposes of the initial setup and designing a gating strategy, the blood from a known healthy The Human RBC PIG-A Gene Mutation Assay 177

donor is recommended. If more than one blood sample is processed, it is recommended to make an antibody and PBS master mix, dispense aliquots into 1.5 mL tubes, and add blood as the last component into the antibody mix. It is important to expose all blood to the antibody cocktail and not let even the smallest droplets of blood remain unstained on the walls or the lids of the tubes. Unlabeled or insufficiently labeled blood cells may appear on cytograms as CD59-deficient mutants. Incubate all experimental and control samples in the dark at room temperature for 30–60 min. 3. Centrifuge the blood labeling reactions for 3 min at 300 × g and remove as much supernatant as possible using a VACUSAFE without disturbing the pellet. Due to aggregation caused by anti-CD235ab antibody, the pellets formed by two-antibody- labeled samples and the mutant-mimic control should be slightly more compacted than the unstained control. Resuspend the pellets in 1 mL of 1 % formaldehyde in DPBS by trituration 7–10 times with a 1 mL pipette tip and filter directly through the strainer-caps into 5 mL flow tubes. Filtration through the strainer eliminates excessive cell aggregates and any particulate matter from the sample which may clog nozzles on sorters. Keep the filtered samples refrigerated until ready for flow cytometric analysis. 4. Flow cytometry setup. Install the 70 μm nozzle on the FACSAria I, and set the configuration for a sheath pressure of 70 psi. Next, perform the fluidics start-up, start and optimize the stream, and perform the CST calibration procedure according to instrument specifications. The optimal PMT voltages determined in the CST routine are usually a good starting point for subsequent adjustments. In the Browser window create new experiment. In the Cytometer settings, choose collecting data for forward (FSC) and side (SSC) light scatter, FITC and APC fluorescence in the peak area mode (A mode). Collect FSC in linear scale and all other parameters in log scale. Set the threshold for FSC to 5,000 to minimize acquiring data on subcellular debris. In the experiment folder, create specimens and tubes for the samples that will be analyzed. For example, the Control specimen may contain a tube for an unlabeled sample and a tube for a mutant-mimic sample. Each two-color-labeled sample may belong to individual specimens. In the Global Worksheet window, create an FSC (horizontal) vs. SSC (vertical) cytogram, an APC histogram, and an APC (horizontal) vs. FITC (vertical) cytogram; right-click on any cytogram window and open the population hierarchy box. For each cytogram and histogram, in the Inspector menu under the Title tab, check the boxes to display specimen name, tube 178 Vasily N. Dobrovolsky and Robert H. Heflich

name, and populations in the heading. In the Inspector menu under the Plot tab, choose bi-exponential scales for the APC histogram and the APC vs. FITC cytogram. 5. Load the unstained control sample. In the Acquisition dashboard adjust the flow rate to achieve a threshold of 15–20 × 103 events/s. Adjust voltages on the FSC and SSC detectors to place the main cell population approximately in the middle of the bivariate FSC vs. SSC cytogram (consult Fig. 1 for examples of placing cell populations within the cytograms and drawing gates). On the light scatter cytogram make a lose gate surrounding the main population and name it “Cells”; in the population hierarchy box, assign a solid color to the “Cells” gate, e.g., green. Right-click on the APC histogram and choose to display events only from the “Cells” gate. Monitor the appearance of events in the APC histogram—a single peak in the left part of the histogram (as in Fig. 1, top center); if the peak exceeds 2 × 102, then reduce the voltage on the APC PMT. In the Acquisition dashboard choose “All events” for storage and stopping gates. Collect data for 100,000 events in the stopping gate; stop acquisition and unload the unstained sample tube. For the unlabeled control, the events in the APC vs. FITC cytogram will be mostly in the lower left quadrant (Fig. 1, top right). 6. Load a mutant-mimic control sample and start acquisition. The main population on the light scatter cytogram should appear similar to that of the unstained control, with the main population of events falling into the “Cells” gate. On the APC histogram the events should appear as two peaks—a major peak somewhere in the middle of the histogram and a satellite peak (or several satellite peaks of decreasing size) on the right side of the major peak. Adjust the voltage on the APC PMT to place the maximum of the main (leftmost) peak at approximately 1–2 × 103. If the leftmost peak appears to be the same size or smaller than the peak(s) right of it, then the level of anti-CD232ab antibody in the labeling reaction was too high and there were too many RBC aggregates (of two and more RBCs) in the sample. In this case the labeling should be repeated with appropriate adjustments. Draw a tight interval gate covering the leftmost peak and call it “Singlet RBCs”; in the population hierarchy assign the red color to the gate. Right-click on the APC vs. FITC cytogram and choose to display only events from the “Singlet RBCs” gate. Adjust the voltage on the FITC PMT to place the cell population in the lower third of the cytogram (approximately within the same range of values for FITC fluorescence as in Fig. 1, second row, on the right). In the Acquisition board, choose “All events” as the storage gate and “Singlet RBCs” as the stopping gate; record data for 100,000 singlet RBCs. Stop the acquisition The Human RBC PIG-A Gene Mutation Assay 179

Fig. 1 Typical cytograms and gating strategy for detecting CD59-deficient PIG-A( mutant) RBCs in human peripheral blood. Once determined, the oval gate on light scatter cytograms (left column) and the rectangular gate on CD235ab vs. CD59 cytograms (right column) remain the same for all experimental samples (even for experiments performed on different days); the interval gate on the histograms in the center column can be manually repositioned within a small range from sample to sample. The satellite peaks outside the interval gate on the APC histograms in the middle of the figure represent RBC doublets and higher order aggregates in labeled samples. RBCs aggregate in stacks (in a “concertino” style) which are difficult to resolve with standard approaches for doublet discrimination (using the height and the width of the measured signal). Note linear scale for forward light scatter (FSC) and bi-exponential scales for CD235ab-APC and CD59-FITC fluorescence; bi-exponential displays help in visualizing “negative” populations in their entirety which is essential for properly adjusting PMT voltages and determining the gate for CD59-deficient PIG-A mutant RBCs. In healthy individuals, the fraction of events in the “CD59-negative RBCs” gate out of the total events displayed in the APC vs. FITC cytogram should be extremely low (~5 × 10−6, as in third row on the right). In this specific example, the total number of singlet RBCs is approximately 6 × 105, with only two RBCs being negative for CD59 (the dots representing these two deficient RBCs were enlarged for better visibility). The calculated RBCPIG-A mutant frequency for this blood sample is 3.3 × 10−6. The three panels in the bottom row were produced with blood from an individual having an unusually high fraction of PIG-A mutant RBCs (ca. 200 CD59-negative RBCs out of ~6 × 105 total singlet RBCs, with a resulting mutant frequency of ~300 × 10−6), but not as high as observed in PNH cases (where it can be in excess of 50 %). The interpretation of such results may be complicated without knowing the prior health history of the donor; the possibilities include a low-level clonal expansion of a PIG-A mutant bone marrow erythroid cell (without expressing characteristic PNH clinical signs of the disease) and exposure to a potent mutagen 180 Vasily N. Dobrovolsky and Robert H. Heflich

and unload the tube. Draw a rectangular gate comprising the tight population on APC vs. FITC cytogram with generous tolerance at the top of the population and extending it toward the bottom of the cytogram; name the gate “CD59-negative RBCs” and assign the blue color to the gate. 7. Load a two-antibody labeled sample from a healthy individual with an expected low (normal) background level of spontaneous CD59-deficientPIG-A mutant RBCs. On the light scatter cytogram, the main population should fall in the “Cells” gate; on the APC histogram the main peak and the satellites should be approximately in the same place as for the mutant-mimic control, so that the previously created “Singlet RBCs” gate tightly covers the main peak. If needed, without adjusting the APC PMT voltage, adjust the position of the “Singlet RBCs” gate on the histogram one tick left or right to precisely cover the main peak with the interval gate. The vast majority of singlet RBCs from a healthy person will be CD59 positive. Adjust the voltage on the FITC PMT to place cells above the “CD59- negative RBCs” gate with sufficient clearance (as in Fig. 1, third row, on the right). In the Acquisition dashboard choose “All events” as the storage gate and “Singlet RBCs” as the stopping gate; collect data for 1,000,000 (or more) singlet RBCs. Stop acquisition and unload the tube. After the software renders the image of all events collected for this sample, there should be very few or no events in the “CD59-negative RBCs” gate. Thus the optimal voltages for the light scatter and fluorescence detectors (set in steps 5–7) are the reference values for subsequent analysis of the experimental samples. With these optimized parameters, the unstained and mutant-mimic control samples may be reanalyzed to verify that the populations of properly stained cells remain within the intended coordinates and in the appropriate gates in the cytograms/ histograms. 8. The detector voltages should remain the same while performing analysis of multiple blood samples on any specific experimental day. The CST routine should be performed on each experimental day before processing the samples to assure that there is no drift in cytometer performance. The unstained control and single-color labeled mutant-mimic control also should be prepared on each experimental day together with the two- color labeled samples. If the two controls are processed prior to analysis of the experimental samples, it is recommended that a “blank” sample, containing only sheath fluid, be loaded on the flow cytometer and run for 20–30 s in order to flush the probe and the sampling line of any potential carryover control cells (especially the cells from the mutant-mimic control that may appear as mutants if contaminating the two-color-labeled The Human RBC PIG-A Gene Mutation Assay 181

experimental samples). Alternatively, if the cytometer shows no drift over time and the labeling procedure is consistent in the hands of the investigator, the controls can be processed at the end of the experimental day, after all two-color-labeled samples have been analyzed. The only acquisition change between the samples (that may or may not be required) is fine adjustment to the position of the “Singlet RBCs” interval gate covering the main peak on the APC histogram. Load an experimental 2-color-labeled sample; from the Acquisition dashboard choose “All events” as the storage gate and “Singlet RBCs” as the stopping gate; collect data for at least 1,000,000 singlet RBCs. Monitor the appearance of events on the cytograms in real time; any abrupt changes or gross deviations from the expected appearance of the populations may indicate problems with the instrument or sample preparation. Upon acquisition, note the number of events in the “CD59-negative RBCs” gate and the “Singlet RBCs” gate (in this example it should be 1,000,000). The frequency of CD59-deficient, presumedPIG- A mutant, RBCs is calculated as a ratio [CD59-negative RBCs]/[Singlet RBCs]. Collecting data for more than 1 × 106 RBCs (e.g., for 2 × 106 or 3 × 106) may require longer acquisition times and create larger, difficult-to-handle files within FACSDiva software, but with the benefit of much more accurately detecting RBC PIG-A mutant frequency (see discussion in Sect. 1.3).

4 The Potential Role of the Human PIG-A Assay in Regulatory Science

Currently, regulatory authorities do not require routine measurement of genotoxicity in humans for hazard identification. However, there are essentially no gene mutation assays used for regulatory safety assessments that have the ability for translation to humans as does the Pig-a assay. The rodent Pig-a assay is rapidly being developed for regulatory use (e.g., it is currently mentioned as a possible in vivo follow-up assay in the ICH M7 Draft Guideline [19]), and we have described here a practical method for measuring PIG-A mutant RBCs in humans. The modern paradigm for safety assessments is always evolving, and the concept of monitoring humans for genotoxicity in clinical trials of new drugs, for example, may now be reasonable. Arguments in favor of monitoring humans for genotoxicity include the possibility of human-specific reactions to products such as drugs, food supplements, and medical devices, and the fact that in vivo safety assessments are performed using healthy young rodents having uniform genetic background, while human exposures may occur to individuals with diverse and/or unusual characteristics. A human PIG-A gene mutation assay may 182 Vasily N. Dobrovolsky and Robert H. Heflich

also be used to “validate” the results of present safety assessment paradigms that are used to predict genotoxicity in humans. Finally, a practical human gene mutation assay, like the PIG-A assay, may be used for monitoring genotoxic risks in populations affected by industrial or environmental exposures, for example, as occurred at the Fukushima Daiichi nuclear power plant in Japan or in patients given a contaminated drug. For the present, additional longitudi- nal studies on heterogeneous human populations are required to identify whether RBCs, which have obvious practical advantages, or other cell types in peripheral blood (e.g., a younger cohort of RBCs—reticulocytes, or high turnover mononuclear cells— granulocytes [20], or longer lasting lymphocytes) are the best cell type for monitoring mutation in humans.

Declaration

The views presented in this chapter are not necessarily those of the US Food and Drug Administration.

References

1. Ames BN, Durston WE, Yamasaki E et al 7. Bessler M, Mason PJ, Hillmen P et al (1994) (1973) Carcinogens are mutagens: a simple Paroxysmal nocturnal haemoglobinuria (PNH) test system combining liver homogenates for is caused by somatic mutations in the PIG-A activation and bacteria for detection. Proc Natl gene. EMBO J 13:110–117 Acad Sci U S A 70:2281–2285 8. Young NS, Maciejewski JP, Sloand E et al 2. The International Conference on Harmoni (2002) The relationship of aplastic anemia and sation of Technical Requirements for Regis PNH. Int J Hematol 76(Suppl 2):168–172 tration of Pharmaceuticals for Human Use 9. Brodsky RA, Hu R (2006) PIG-A mutations (ICH) (2011) Guidance on genotoxicity test- in paroxysmal nocturnal hemoglobinuria and in ing and data interpretation for pharmaceuticals normal hematopoiesis. Leuk Lymphoma 47: intended for human use. http://www.ich.org/ 1215–1221 fileadmin/Public_Web_Site/ICH_Products/ 10. Ware RE, Pickens CV, DeCastro CM et al Guidelines/Safety/S2_R1/Step4/S2R1_ (2001) Circulating PIG-A mutant T lympho- Step4.pdf cytes in healthy adults and patients with bone 3. Schuler M, Gollapudi BB, Thybaud V et al marrow failure syndromes. Exp Hematol 29: (2011) Need and potential value of the Pig-ain 1403–1409 vivo mutation assay—a HESI perspective. 11. Hall SE, Rosse WF (1996) The use of mono- Environ Mol Mutagen 52:685–689 clonal antibodies and flow cytometry in the 4. Kinoshita T, Fujita M, Maeda Y (2008) diagnosis of paroxysmal nocturnal hemoglo- Biosynthesis, remodelling and functions of binuria. Blood 87:5332–5340 mammalian GPI-anchored proteins: recent 12. Richards SJ, Rawstron AC, Hillmen P (2000) progress. J Biochem 144:287–294 Application of flow cytometry to the diagnosis 5. Nishimura J, Murakami Y, Kinoshita T (1999) of paroxysmal nocturnal hemoglobinuria. Paroxysmal nocturnal hemoglobinuria: an Cytometry 42:223–233 acquired genetic disease. Am J Hematol 62: 13. Tiede A, Bastisch I, Schubert J et al (1999) 175–182 Biosynthesis of glycosylphosphatidylinositols in 6. Smith LJ (2004) Paroxysmal nocturnal hemo- mammals and unicellular microbes. Biol Chem globinuria. Clin Lab Sci 17:172–177 380:503–523 The Human RBC PIG-A Gene Mutation Assay 183

14. Dertinger SD, Phonethepswath S, Weller P et al glycosylphosphatidylinositol-anchored proteins (2011) International Pig-a gene mutation assay on different subsets of peripheral blood cells: a trial: evaluation of transferability across 14 labo- frame of reference for the diagnosis of paroxys- ratories. Environ Mol Mutagen 52:690–698 mal nocturnal hemoglobinuria. Cytometry B 15. Araten DJ, Nafa K, Pakdeesuwan K et al (1999) Clin Cytom 70:71–81 Clonal populations of hematopoietic cells with 19. The International Conference on Harmon paroxysmal nocturnal hemoglobinuria genotype isation of Technical Requirements for Regis and phenotype are present in normal individu- tration of Pharmaceuticals for Human Use als. Proc Natl Acad Sci U S A 96:5209–5214 (ICH) (2013) Assessment and control of DNA 16. Dobrovolsky VN, Elespuru RK, Bigger CA reactive (mutagenic) impurities in pharmaceu- et al (2011) Monitoring humans for somatic ticals to limit potential carcinogenic risk, M7. mutation in the endogenous PIG-a gene using Draft consensus guideline. http://www.ich. red blood cells. Environ Mol Mutagen 52: org/fileadmin/Public_Web_Site/ICH_ 784–794 Products/Guidelines/Multidisciplinary/M7/ 17. Bhalli JA, Shaddock JG, Pearce MG et al M7_Step_2.pdf (2013) Sensitivity of the Pig-a assay for detect- 20. Rondelli T, Berardi M, Peruzzi B et al (2013) ing gene mutation in rats exposed acutely to The frequency of granulocytes with spontane- strong clastogens. Mutagenesis 28:447–455 ous somatic mutations: a wide distribution in 18. Hernandez-Campo PM, Almeida J, Sanchez a normal human population. PLoS One 8: ML et al (2006) Normal patterns of expression of e54046 Chapter 11

The Applicable Use of the HPRT Gene Mutation Assay as a Practical Tool in Mutagenesis and DNA Repair Studies

Zoulikha M. Zaïr and George E. Johnson

Abstract

The HPRT gene mutation assay is a notable tool that detects for genotoxic substances and allows for the isolation and screening for inducible mutation types. As with the thymidine kinase (TK) mouse lymphoma assay (MLA), the HPRT gene mutation assay is considered a significant tool as set out in the guidelines for mammalian gene mutation tests [OECD Guideline for the Testing of Chemicals. In Vitro Mammalian Cell Gene Mutation Test: 476]. Since its refinement, the HPRT assay has been widely utilized in mechanistic studies using human knock-out cell lines for DNA repair, providing details of the mode of action (MOA) of the test substance. This chapter provides up-to-date methodology for carrying out the assay in different cell lines in the presence and absence of metabolism with relevant technical information and emerging advances in statistical analysis of data generated from the HPRT gene mutation assay.

Key words HPRT, Mutagenicity, DNA repair, Human lymphoblastoid cells

1 Introduction

1.1 Recommended There are three main features of the mammalian hypoxanthine- Use of the HPRT Assay guanine phosphoribosyltransferase (HPRT) gene mutation assay which have led to its wide use: (a) The target gene is encoded on the mammalian X-chromosome, and consequently it is easy to select for loss-of-function mutants in cells derived from males, which in mammals are heterogametic for sex chromosomes. (b) The biochemical selection systems for loss of function with cells that survive in the presence of 6-thioguanine (6-TG) and/or 8-azaguanine are simple and effective [1]. (c) Also an advantage of the HPRT gene is that mutations in the same gene can be compared between cell lines, experimental animals, and with humans [2].

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_11, © Springer Science+Business Media New York 2014 185 186 Zoulikha M. Zaïr and George E. Johnson

Human lymphoblastoid cell lines [3] were first used for HPRT assays. Chinese hamster ovary (CHO) cells then became the main cell line [4] and these adherent cells have been the most widely used cells to date. However, the test is relatively flexible and the human lymphoblastoid AHH-1 and MCL-5 suspension cell lines can also be used [5, 6]. The AHH-1 cell line is very suitable for these experiments, and because it is heterozygous at the TK+/− locus, we were able to validate our HPRT results by carrying out both the HPRT and TK gene mutation assays alongside one another. In vivo experiments can also be performed using this methodology, and mice or rats are commonly treated with the test substance by an appropriate route of exposure. After a fixation period, the rodent lymphocytes are removed and established. T-lymphocytes are particularly useful to examine because they circulate throughout many tissues, which allows them a greater chance to come into contact with an administered mutagen than cells that are permanently resident in a single tissue. T-lymphocytes are also long-lived in circulation, and they continue to undergo cell division, which makes the identification of mutant cells possible. In addition to T-lymphocytes, mutant frequency has also been determined in many cells, including those from the spleen, kidney, thy- mus, and lymph nodes.

1.2 Molecular HPRT is a key enzyme in the purine salvage pathway with more Principles of the than 300 disease-associated mutations reported [7, 8]. The main HPRT Assay being through partial enzyme deficiency leading to gouty arthritis and complete deficiency leading to Lesch–Nyhan disease [9]. The HPRT gene is located on the X-chromosome and encodes for HPRT protein, which plays a key role in the purine salvage pathway. HPRT catalyzes the transformation of purines (hypoxanthine, guanine, or 6-mercaptopurine) to monophosphates, which are therefore cytotoxic to HPRT-proficient cells. Cells with mutations in the HPRT gene have lost the HPRT enzyme and thus survive treatment with purine analogues. As large losses in the X-chromosome lead to cell lethality, only small changes, such as point mutations and exon deletions, are detected in the HPRT gene. The spontaneous MF is also lower at the HPRT locus, than the TK locus, as high proportions of mutations at HPRT are lethal (but not at TK). These events include nondisjunction and translocation, which are known to lead to viable TK mutants but not viable HPRT mutants. After a chemical insult, further multiple events are required to transform DNA changes (pro-mutagenic DNA lesions) into selectable phenotypes: (a) Fixation of the mutation (b) Reduction of the preexisting enzyme to a level with no biological activity The HPRT Gene Mutation Assay 187

Fixation of a mutation requires the initial lesion in the DNA (i.e., adduct, strand break, or damage to DNA-dependent proteins) being translated into a DNA sequence change, such as point mutation, deletion, or loss. For a point mutation, the mutant strand must be separated from the wild type strand by cell division; thus one of the progeny cells no longer produces active mRNA and/or protein. Point mutations can only occur after the cells have undergone cell division, as the lesion affecting the base may be removed or the base may be repaired by DNA repair. The mutation is therefore only relevant in these assays once it has been incorporated into both strands. For selection, the existing enzyme or mRNA must be reduced by either cell division or degradation to nonfunctional levels, so that the original phenotype can no longer be identified. Cells are incubated in microwell culture plates with the selective agent 6-TG, a purine analogue that is a substrate for HPRT and is toxic to nonmutant cells. A significant increase in mutant frequency in treatment cultures compared with controls indicates the test chemical has induced mutation at the HPRT locus. The average spontaneous mutant frequency at the HPRT locus is in the range of 10−6 [10]. Using standard techniques, further molecular analysis of HPRT mutations can be performed, if desired. Mutation spectra utilizing HPRT mutants from AHH-1 cells treated with the direct-acting genotoxicant methyl-nitrosourea (MNU) [11] and HPRT from harvested liver cells from transgenic mice treated with phenobarbital [12] have provided invaluable insight into mutation hot spots, mutation type, and frequency.

2 Materials

2.1 Cell Lines Our group in Swansea University has recently used the AHH-1 cell line to investigate the dose–response relationships of DNA reactive genotoxic agents in the low-dose region of exposure [13, 14]. The AHH-1 cell line was very suitable for these experiments, and because it is heterozygous at the TK+/− locus, we were able to validate our HPRT results by carrying out both the HPRT and TK gene mutation assays alongside one another. Other suitable cell lines for the HPRT assay include L5178Y mouse lymphoma cells; the CHO, AS52, and V79 lines from Chinese hamsters [15–17]; and AHH-1, MCL-5, and TK6 human lymphoblastoid cells [5, 6, 13, 18]. Note: L5178Y mouse lymphoma cells and CHO cells have a published spontaneous MF of 2–50 × 106 [19]. AHH-1 and MCL-5 human lymphoblastoid cells have a published spontaneous MF of 6–80 × 106 [20, 21]. The HPRT locus is on the X-chromosome and therefore primary male cell lines can also be used to study mutagenic effects in mice and rats [22, 23] and this methodology can also be used for 188 Zoulikha M. Zaïr and George E. Johnson

human biomonitoring [13, 24]. In this chapter we will focus on the in vitro HPRT assay mainly in human lymphoblastoid cell lines. Each cell line requires specific culture medium and this along with the cell culture conditions are stated in the batch details, provided upon purchase of the cell line.

2.2 Compounds HAT (hypoxanthine–aminopterin–thymidine) supplement is Required for the HPRT added to culture medium at the mutant cleansing stage. The ami- Gene Mutation Assay nopterin in HAT medium blocks the salvage pathway, leaving cells reliant on the endogenous pathway, i.e., HPRT and TK, and therefore HPRT− and TK−/− mutants are killed. This reduces the spontaneous (background) MF values and is a crucial step. Concentrated HAT medium may be purchased from Sigma or GIBCO and diluted in sterile water. Alternatively 1× HAT media may be made thus: 100 μM hypoxanthine, 1 μM aminopterin, 20 μM thymidine. For freezer stock, it is recommended to freeze aminopterin separately thus: dissolve 136 mg hypoxanthine and 48.4 mg thymidine in a total volume of 100 mL sterile water to give 100× stock. Filter sterilize and freeze 1 mL aliquots. Protect from light. Dissolve 4.4 mg aminopterin in a few mL of sterile 0.1N NaOH up to 100 mL sterile water to give 100× stock. Ensure pH is 7.0, adjust with HCl if above 7.0. Filter sterilize and store in 1 mL aliquots at −80 °C. Protect from light. Following HAT treatment, HT (hypoxanthine–thymidine) supplement is added to culture medium, and both the de novo nucleotide biosynthesis pathway and the salvage pathway are able to function from this stage onwards. After treatment with the test compound, the HPRT− mutant cells are selected for using 6-TG. HPRT+ cells incorporate 6-TG into the DNA and die, and HPRT− cells do not incorporate this toxic analogue into their DNA, and they survive. Ethyl methanesulphonate (EMS) and ethyl-nitrosourea (ENU) can be used as the positive controls in the absence of exogenous metabolic activation [17]. 3-Methylcholanthrene, N-nitrosodime- thylamine, or 7,12-dimethylbenzanthracene can be used as a positive control in the presence of exogenous metabolic activation [17]. Note: To avoid false-negative results from the effects of metabolic cooperation, in which wild type HPRT+ cells cross-feed killing off TG-resistant mutants, keep cell densities below the threshold at which this occurs. To determine the threshold, plate cells across a series of cell densities keeping 6-TG levels constant. The frequency should not be high enough to yield more than a few colonies. Suspension cells are not prone to this phenomenon and should be feasible at cell densities of 10E5 cells per 100 μL well. Note: For nanotoxicity studies, it is advisable to use fresh HAT whenever deciding to stagger plating (used to make large dose-series more manageable), as cells may exhibit delayed or no cell revival or proliferation with free-thaw HAT stocks [25]. The HPRT Gene Mutation Assay 189

3 Method

3.1 Metabolism Compounds that require metabolic activation require either an exogenous source such as S9 (treatment time of 3–6 h with the test compound), or a genetically modified cell line such that metabolic activation is endogenous. For example, MCL-5 cells are derived from L3 cells, a subpopulation of AHH-1 cells that express a particularly high level of CYP1A1 activity [6, 26]. The MCL-5 cell line has also been transfected with two plasmids: one containing two copies of CYP3A4 cDNA and one copy of CYP2E1 cDNA and the second containing one copy of each CYP1A2, CYP2A6, and microsomal epoxide hydrolase cDNA [6, 26]. Therefore MCL-5 cells stably express all five cDNAs and also have increased levels of CYP1A1 compared to AHH-1 cells, and test compounds that are known to be metabolized by these enzymes can be tested using MCL-5 cells. Genetically modified cell lines that stably express metabolic enzymes can promote more stable and reliable results as there are fewer variables such as pH, osmolality, or high levels of cytotoxicity than when adding crude cell extracts of liver, S9.

3.2 HPRT Mutant Mammalian gene mutation assays depend upon the ability to quan- Cleansing tify mutant cells using selective media. The mutation frequency (MF) at each test concentration is compared to the control MF, and the control is a measure of the spontaneous MF. Therefore, the spontaneous MF should be maintained at a low and stable level within each laboratory. To decrease the spontaneous MF of cultures, the number of HPRT− mutants is reduced using HAT medium, which inhibits the endogenous de novo nucleotide biosynthesis pathway so that the salvage pathway is required for dNTP synthesis for DNA replication. Therefore, cells that are incapable of using the salvage pathway (i.e., HPRT− mutants) can no longer divide and undergo cell death.

3.3 Treatment Following “mutant cleansing,” the cells are sub-cultured and Protocol grown for 24 h in HT medium. The cells are washed and grown in normal culture medium for 3–4 days to attain sufficient numbers for treatment. This can be a good time to cryogenically freeze down the cells for storage, if for any reason the experiment needs to be carried out at a later date. However, one should be careful that there are no extra days added to the assay as these can cause clonal expansion and the spontaneous MF can be increased. Ten milliliter treatment flasks or six-well plates are set up and the test chemical is added to the cultures at predefined concentrations, which should usually be separated by no more than a factor of between 2 and √10 and should also cover a range of high toxicity to little or no toxicity [17]. There are currently revisions taking 190 Zoulikha M. Zaïr and George E. Johnson

place to this guideline, which currently states that “the maximum concentration is based on cytotoxicity then it should result in approximately 10–20 % (but not less than 10 %) relative survival (relative cloning efficiency) or relative total growth.” A typical design of the HPRT assay would be testing up to 10 mM or 5,000 μg/mL [17] with a maximum of one insoluble concentration, because cells grow in suspension and precipitate cannot be removed. However, this has recently changed for pharmaceutical testing which follows the guidelines produced by the International Conference on Harmonisation, where the top dose has been “reduced from 10 to 1 mM in order to reduce the number of irrel- evant/false-positive in vitro findings” 27[ ]. Treated cultures are then incubated for 4 or 24 h, depending on the half-life of the compound and whether S9 is added, and each test compound should be prepared and dissolved in the correct solvent using sources of information such as the batch guidelines or the Merck index. A typical design of the HPRT assay would be with and without S9 (4 h treatments) plus a 24 h treatment without S9. Each experiment is then carried out in at least duplicate, and each duplicate treatment flask is treated using a stock solution that is prepared each time (i.e., two times for a duplicate experiment) from the purchased product, to allow for variation in preparation procedures. HPRT assays usually include four or more concentrations with duplicate treatments per concentration, or in triplicate if more advanced statistical analysis is required. Negative controls must be used, and the solvent should not have a significantly different MF than the spontaneous MF. Positive controls must also be used [17]. Following chemical exposure, the cells are centrifuged to remove the test medium, washed, and resuspended in fresh culture medium for mutant expression. For the HPRT mutation assay, this involves incubation for 13 days. This allows any mutations to become fixed and any existing HPRT proteins/RNA to become degraded. During the phenotypic expression period, the cells are sub-cultured every other day, on days 1, 3, 5, 7, 9, and 11, by centrifugation and re-suspension of the cells in fresh culture medium. Cells can be cryogenically frozen down at this stage which helps when carrying out large dose–response studies where large numbers of plates are required per dose. The day of freezing down should be recorded so that the number of expression period days is not altered due to this process. It is not advisable to freeze them down after day 9 as this is close to the time of plating, and we have previously shown this to increase the background mutant frequency. After the phenotypic expression period, the cells are added to 96-well microplates at 40 % confluence (4 × 104 cells/well for AHH-1 and MCL-5 cells) in culture medium The HPRT Gene Mutation Assay 191

with selection using the toxic analogue 6-thioguanine (6-TG from Sigma UK) at 0.6 μg/mL. It is difficult to treat a sufficiently large number of cells (>105 cells per petri dish) to produce statistically powerful assays. Therefore, we make use of a microplate protocol to improve the sensitivity of assay. Plates are scored for colony formation after 14 days of incubation at 37 °C in humidified incuba-

tor with 5 % CO2. The outside wells of the 96-well microplate can dry up due to the long growth period in the incubator, to give false-negative colony growth. Note: Special plates can be purchased that have a channel for water which keeps the cells in a humid environment, or other methods can be designed to keep the cells in a humid environment. For the cloning efficiency (P.E) calculation, 2–200 cells/well can be plated with no selection at each dose. Note: Scoring colonies of attached cell lines is different from scoring colonies of suspension cell lines, and this should be considered when deciding whether they are defined as viable colonies (e.g., >20 cells diameter) or not (e.g., <20 cells diameter or the cells are dead). Note: Adding 6-TG can kill the colonies at a late stage of growth, and therefore you should be careful that you are scoring viable colonies and not dead cells. This can be determined by observing the morphology of the cells. For example, dead AHH-1 lymphoblastoid cells are darker and their cell walls are also less circular than viable cells.

3.4 Scoring Method The criteria for colony counting include only scoring colonies of 20+ cells in diameter and ensuring separate colonies are clearly apart, thereby accounting for clonal expansion. This value was defined by us in Swansea and other laboratories may wish to define other scoring criteria. There will also be a large number of dead HPRT+ colonies due to 6-TG selection, and these must not be scored as viable colonies.

3.5 Statistical In the presence of metabolic activation and detoxification of a Method genotoxic compound, the kinetics of DNA-adduct formation is considered first order. A biologically relevant increase in mutation rate which deviates from a linear response may only be observed at dose levels that result in saturation of the DNA repair processes. In order to determine this point of departure (PoD), several statistical tests may be utilized. The Genetic Toxicology Technical Committee (GTTC) quantitative workgroup, formally known as the International Life Sciences Institute Health and Environmental Sciences Institute (ILSI-HESI), recommends the use of the 1-sided Dunnett’s test for data with normal distribution and homogeneous variance, 1-sided Dunnett’s T3 for normal distribution and 192 Zoulikha M. Zaïr and George E. Johnson

heterogeneous variance, and 1-sided Dunn’s test for nonparametric data. These tests are available in packages such as SPPS and SAS, but the ILSI-HESI group has recently submitted a package called “DRSMOOTH” to CRAN, which will allow these tests and more PoD tools to be used on the open-source statistical software R (http://cran.r-project.org/). No observed and lowest observed genotoxic effect levels (NOGEL and LOGEL) can be defined using these approaches for use as PoD metrics for HPRT data with suitable statistical power. Other PoD methods include bilinear modeling (two lines with different slopes) such as the hockey stick model, which may be used to determine the “line of best fit” [28], and the segmented package available in R [29, 30]. A novel package based on mgcv, which provides generalized additive modeling functions based on a penalized regression spline approach with automatic data smooth- ness function, can also be defined as a usable PoD metric called a slope transition dose. The most well-regarded approach for defining a PoD using HPRT data is the benchmark dose (BMD) approach [30, 31]. BMD analysis can be conducted using PROAST, the dose–response modeling software developed at the National Institute for Public Health and the Environment (RIVM) in the Netherlands (www. proast.nl; version 36.9 [32], 2002). BMD analysis can also be conducted using EPA’s Benchmark Dose Software [33]. For continu-

ous datasets such as HPRT mutant frequencies, the BMDL10 and the BMDL1SD are recommended [30].

3.6 Clonal Expansion Mutant HPRT− colonies are removed from the 96-well microplates of Mutants and DNA using Pasteur pipettes and re-suspended in fresh culture medium. Extraction Cultures are maintained and grown to confluence for sufficient numbers of cells for DNA/RNA extraction.

3.7 HPRT Mutation The HPRT gene has been characterized largely using the poly- Spectrum Analysis merase chain reaction (PCR), with multiplex PCR being used for exon deletion detection, i.e., many DNA fragments (i.e., different exons) can be amplified simultaneously, which alongside sequencing can detect smaller mutations [34]. The coding region of the human HPRT gene is distributed over 39.8 kb DNA and contains nine coding exons. The RNA transcript can also be sequenced using reverse transcriptase PCR, and the resulting mutation spectrum from this technique provides sequence information of mRNA (cDNA) and this is least costly, with regard to both time and money, than multiplex PCR [11]. Note: For molecular analysis, it is imperative to obtain independent mutational events. To ensure this plate, replicate cultures at the minimum cell density required for TG resistance. When harvesting cells, harvest a single mutant per replicate. The HPRT Gene Mutation Assay 193

4 Calculations

4.1 Mutation Note: Each replicate should have a minimum of 4 × 96-well Frequency and Cloning microplates for 6-TG selection and 2 × 96-well microplates for Efficiency Equations 3[ ] cloning efficiency (viability). This can be increased if the experiment is designed to detect smaller changes in MF increase. For example, 100 × 96-well microplates for 6-TG selection and 50 × 96-well microplates for cloning efficiency are required in total for threshold analysis. Note: Cloning efficiency is a crucial step in the HPRT gene mutation assay. It is important to optimize the number of cells plated for this step (2–200 cells/well) for the particular laboratory and particular cell line.

4.1.1 Cloning Efficiency Cloning Efficiency %/CE =−Ln XN×100 () ()oo

CE 4.1.2 Cell Viability Cell Viability % =×100 (Relative Cloning CE of control Efficiency) −Ln()XN/ 4.1.3 Mutant Fraction Mutation frequency ()MF = ss× DF −Ln XN/ ()oo

()No.tof initial cells per well Nonselective conditions Dilution factor()DF = ()No.of initial cells per well Selective connditions

X s = No.of wells without colonies Selecctive conditions N s = Total no.of wells 

X o = No.of wells without colonies Nonseelective conditions N o = Total no.of wells 

4.1.4 Worked Examples HPRT− mutation frequency (MF) calculation for replicate B at 0 μg/mL MMS

−Ln()1699 / 1800 20 MF = ×=0./0577468 0.455115××=0..0005 6 344×10−5 −Ln()1218 / 1920 40000

If cells have a low cloning efficiency, it can be due to [23]: 1. A bad batch of 96-well microplates (relatively rare) 2. A bad batch of horse serum (relatively rare). Make sure you use the same batch of horse serum throughout your experiment 194 Zoulikha M. Zaïr and George E. Johnson

3. High pH on the 96-well microplates (common) due to: (a) Opening the incubator too much during the first 4–5 days

(b) Low CO2 setting (c) Incorrect medium pH (should be 6.8–7.0 before adding serum) 4. Poorly growing cells 5. Colonies not well disaggregated before being plated If cells have a high negative control mutant fraction, it can be due to [23]: 1. An artifact due to low cloning efficiency 2. Improper HAT/HT treatment due to: (a) Thymidine starvation (b) Inadequate aminopterin 3. Exposure to a mutagen (sunlight) 4. Inadequate selective agent

5 Assay Benefits and Limitations

The HPRT assay is adapted for use in mammals as well as mammalian cultured cells offering pertinent advantages over microbial tests. These include the genomic organization being similar to that of humans (and is absent in bacteria) as well as mammalian specific cell metabolism cannot be replicated in bacteria. As an endogenous gene, HPRT is transcriptionally active and subject to transcription-coupled DNA repair, making it an ideal assay for DNA damage and DNA repair studies alike. Indeed HPRT gene mutation assay detects point mutations, small insertions, frameshifts, and small deletions and has been successfully used in mapping mutation hot spots specific to particular geno- toxicants. As with any assay there are, however, limitations. HPRT is an X-linked gene and thus hemizygous in function. Detection of large deletions, nondisjunction events, and chromosomal recombination is very poor due to such events being lethal to the cell and being hemizygous there is no compensation for loss of gene function resulting inevitably in cell death. Negative results in the HPRT assay, for example, were observed when using V79 cells treated with cyclopentenone [35]. The authors concluded that large chromosome deletions may have been the predominant form of induced mutations. This hypothesis is supported by investigations showing that oxidative stress is at most weakly mutagenic in terms of point mutations and small deletions, but mutagenic through a mechanism involving large rearrangements [36]. Where the intention is to look for nondisjunction and translocation events then use of thymidine kinase (TK) gene as the gene The HPRT Gene Mutation Assay 195

of target would be a suitable alternative and has been reported to have a higher mutation rate relative to HPRT [10]. In animal experiments it is worth remembering that dilution of mutant T-lymphocytes in circulation occurs as peripheral lymphocyte populations are renewed. T-lymphocyte number is also affected by the age of the animal [37]. As a result, sampling in the spleen (bone marrow may also be used) should be carefully timed to detect the maximum mutant frequency. It is worth noting that published studies typically focus on T-lymphocytes, which prevents identification of mutagenic effects that may arise preferentially in other target tissues. Different sensitivities of the cell lines used may also play a role in reported discrepancies observed for mutagenicity of several compounds of interest. These variations in cell line sensitivity may be due to the species of origin as well as the relative fidelity of the DNA repair capacity in the cells selected. If cells with reduced repair capacity are utilized for the HPRT assay, then false-positive datasets could potentially be the result. Recent examples include nanotoxicity studies whereby conflicting data exists in studies that have utilized the HPRT assay in determining the genotoxic poten-

tial of TiO2 and carbon nanotubes [25, 38–40]. Despite the HPRT being ideal in isolating and characterizing genotoxic mutants, it is not ideal for dose–response analysis given the number of replicates and thus number of cells and doses required. Latest advances have seen the introduction of HPRT activity assays that may be used as an alternative. Such assays, which include PRECICE® HPRT Assay Kit from NovoCIB, allow the investigator to monitor HPRT activity in a convenient 96-well plate format using spectrophotometric analysis, after treatment with the compound of interest. These authors would like to see flow cytometric approaches for novel in vitro mutation tests coming forward, and they hold out much promise for things such as the in vitro PigA assay where 10,000 cells per dose can be scored in a matter of minutes from a single tissue culture flask, compared to hours of scoring in ~100 × 96-well plates.

References

1. Caskey CT, Kruh GD (1979) The HPRT compounds evaluated at the hgprt locus in the locus. Cell 16(1):1–9 standard and suspension CHO assays. 2. Chen T, Harrington-Brock K, Moore MM Mutagenesis 6(1):77–85 (2002) Mutant frequency and mutational spec- 5. Crespi CL, Thilly WG (1984) Assay for gene tra in the Tk and Hprt genes of N-ethyl-N- mutation in a human lymphoblast line, AHH- nitrosourea-treated mouse lymphoma cells 1, competent for xenobiotic metabolism. dagger. Environ Mol Mutagen 39(4):296–305 Mutat Res 128(2):221–230 3. Furth EE, Thilly WG, Penman BW et al 6. Crespi CL, Gonzalez FJ, Steimel DT et al (1981) Quantitative assay for mutation in dip- (1991) A metabolically competent human loid human lymphoblasts using microtiter cell line expressing five cDNAs encoding plates. Anal Biochem 110(1):1–8 procarcinogen-activating enzymes: application 4. Moore MM, Parker L, Huston J et al (1991) to mutagenicity testing. Chem Res Toxicol 4(5): Comparison of mutagenicity results for nine 566–572 196 Zoulikha M. Zaïr and George E. Johnson

7. Torres RJ, Puig JG (2007) Hypoxanthine- reactive carcinogens. Cancer Res 67(8): guanine phosphoribosyltransferase (HPRT) 3904–3911 deficiency: Lesch-Nyhan syndrome. Orphanet 22. van Dam FJ, Natarajan AT, Tates AD (1992) J Rare Dis 2(48):1–10 Use of a T-lymphocyte clonal assay for deter- 8. Jinnah H, Ceballos-Picot I, Torres RJ et al mining HPRT mutant frequencies in individ- (2010) Attenuated variants of Lesch-Nyhan ual rats. Mutat Res 271(3):231–242 disease. Brain 133(3):671–689 23. Tates AD, van Dam FJ, de Zwart FA et al 9. Jolly DJ (1986) The role of the HPRT gene in (1994) Development of a cloning assay with human disease. Horiz Biochem Biophys 8: high cloning efficiency to detect induction of 123–168 6-thioguanine-resistant lymphocytes in spleen 10. Johnson GE (2012) Mammalian cell HPRT of adult mice following in vivo inhalation gene mutation assay: test methods. In: Parry exposure to 1,3-butadiene. Mutat Res EM, Parry JM (eds) Genetic toxicology. 309(2):299–306 Springer, New York, pp 55–67 24. Albertini RJ (2001) HPRT mutations in 11. Thomas AD, Jenkins GJ, Kaina B et al (2013) humans: biomarkers for mechanistic studies. Influence of DNA repair on nonlinear dose- Mutat Res 489(1):1–16 responses for mutation. Toxicol Sci 132(1): 25. Doak SH, Manshian B, Jenkins GJS et al (2012) 87–95 In vitro genotoxicity testing strategy for nano- 12. Shane BS, de Boer J, Watson DE et al (2000) materials and the adaptation of current OECD LacI mutation spectra following benzo[a] guidelines. Mutat Res 745(1–2):104–111 pyrene treatment of Big Blue® mice. 26. Woodruff NW, Durant JL, Donhoffner LL et al Carcinogenesis 21(4):715–725 (2001) Human cell mutagenicity of chlorinated 13. Parry JM, Parry EM, Johnson G et al (2005) and unchlorinated water and the disinfection by The detection of genotoxic activity and the product 3-chloro-4-(dichloromethyl)-5-hydroxy- quantitative and qualitative assessment of the 2(5H)-furanone (MX). Mutat Res 495(1–2): consequences of exposures. Exp Toxicol 157–168 Pathol 57(Suppl 1):205–212 27. ICH (2011) S2 (R1) Guidance on genotoxicity 14. Doak SH (2008) Aneuploidy in upper gastro- testing and data interpretation for pharmaceuti- intestinal tract cancers—a potential prognostic cals intended for human Use. International marker? Mutat Res 651(1–2):93–104 Committee on Harmonisation. pp 1–25. 15. Tindall KR, Stankowski LF Jr, Machanoff R http://www.ich.org/fileadmin/Public_Web_ et al (1986) Analyses of mutation in pSV2gpt- Site/ICH_Products/Guidelines/Safety/S2_ transformed CHO cells. Mutat Res 160(2): R1/Step4/S2R1_Step4.pdf 121–131 28. Lutz WK, Lutz RW (2009) Statistical model 16. Tindall KR, Stankowski LF Jr (1989) to estimate a threshold dose and its confidence Molecular analysis of spontaneous mutations limits for the analysis of sublinear dose- at the gpt locus in Chinese hamster ovary response relationships, exemplified for muta- (AS52) cells. Mutat Res 220(2–3):241–253 genicity data. Mutat Res 678(2):118–122 17. OECD (1997) OECD guideline for the test- 29. Muggeo V (2008) Segmented: an R package ing of chemicals. In vitro mammalian cell to fit regression models with broken-line rela- gene mutation test: 476. DOI: 10.1787/ tionships. R-News: http://cran.r-project.org/ 9789264071322-en doc/Rnews/. pp 20–25 18. Moore MM, DeMarini DM, DeSerres FJ et al 30. Johnson GE, Hernandez LG, Gollapudi BB, (1987) Banbury report 28: mammalian cell Pottenger LH, Bodger OG, Dearfield KL, mutagenesis. Cold Spring Harbor Laboratory, Lovell DP, Heflich RH, Hixon G, Macgregor New York JT, Thompson V, Thybaud V, Zeiger E, van Benthem J, and White PA (2014) Derivation 19. DeMarini DM, Brock KH, Doerr CL et al of points of departure (PoD) estimates in (1988) Mutagenicity and clastogenicity of genetic toxicology studies and their potential proflavin in L5178Y/TK +/− −3.7.2.C cells. application in risk assessment. Environ Mol Mutat Res 204(2):323–328 Mutagen, DOI: 10.1002/em.21870 20. Gentest (1994) AHH-1 TK+/- human lym- 31. Gollapudi BB, Johnson GE, Hernandez LG et al phoblastoid cells. Routine cell culture, metab- (2013) Quantitative approaches for assess- olite production and gene-locus mutation ing dose-response relationships in genetic assay. Procedures for use. Gentest Corporation, toxicology studies. Environ Mol Mutagen Woburn, MA 54(1):8–18 21. Doak SH, Jenkins GJ, Johnson GE et al 32. Slob W (2002) PROAST: software for dose- (2007) Mechanistic influences for mutation response modeling and benchmark dose analy- induction curves after exposure to DNA- sis. RIVM The HPRT Gene Mutation Assay 197

33. EPA (2013) BMDS v2.2 38. Wang JJ, Sanderson BJ, Wang H (2007) Cyto- 34. McPherson M, Taylor G, Quirke P (1991) and genotoxicity of ultrafine TiO2 particles in PCR 1: a practical approach, Practical approach cultured human lymphoblastoid cells. Mutat series. Oxford University, Oxford, p 274 Res 628(2):99–106 35. Solecki GM, Groh IAM, Kajzar J et al (2013) 39. Huang S, Chueh PJ, Lin Y-W et al (2009) Genotoxic properties of cyclopentenone Disturbed mitotic progression and genome prostaglandins and the onset of glutathione segregation are involved in cell transforma- depletion. Chem Res Toxicol 26(2):252–261 tion mediated by nano-TiO2 long-term 36. Gille JJP, van Berkel CGM (1994) exposure. Toxicol Appl Pharmacol 241(2): Mutagenicity of metabolic oxygen radicals in 182–194 mammalian cell cultures. Carcinogenesis 40. Asakura M, Sasaki T, Sugiyama T et al (2010) 15(12):2695–2699 Genotoxicity and cytotoxicity of multi-wall 37. OECD (2005) Chapter 2: existing genotoxicity carbon nanotubes in cultured Chinese hamster assays. In: OECD Environment HaSPS (ed). lung cells in comparison with chrysotile A pp. 21–36 fibers. J Occup Health 52(3):155–66 Chapter 12

The Comet Assay: High Throughput Use of FPG

Amaya Azqueta and Andrew R. Collins

Abstract

The alkaline comet assay in its standard form is well established as a genotoxicity testing assay, widely used in screening novel chemicals and pharmaceuticals for potentially carcinogenic effects. Incorporation of a digestion of DNA with lesion-specifi c enzymes is an accepted modifi cation which has allowed, for example, the quantitative assessment of levels of 8-oxoguanine in DNA as a measure of oxidative damage, using the enzyme formamidopyrimidine DNA glycosylase (FPG). However, FPG is not restricted to the measurement of oxidized bases, and we describe here its use in a wider context to detect various kinds of DNA damage. A limitation of the standard assay is the relatively low number of samples that can be run in one experiment (restricted by the number of microscope slides fi tting in the electrophoresis tank). Recent developments of high throughput versions of the comet assay have alleviated this problem, and we describe a modifi cation based on the use of 12 minigels on each slide. We provide a detailed protocol for running 12 minigels per slide, with the inclusion of FPG to obtain enhanced sensitivity. We emphasize the conditions of the comet assay that are most critical for reproducibility and accuracy.

Key words DNA damage , Comet assay , Lesion-speciﬁ c enzymes , High throughput , Genotoxicity testing

1 Introduction

The alkaline comet assay (single cell gel electrophoresis) is a sensitive and versatile genotoxicity assay that detects DNA breaks. It can be applied to virtually any eukaryotic cell or nuclear suspension applying very similar protocols. It has been applied to cultured cells, primary animal and human cells (e.g., blood cells, cells from different tissues and sperm), hemolymph cells from molluscs and insects, yeast, and nuclei from plant tissues. It has also been successfully applied to isolated chromosomes [ 1 ]. It is widely used in genotoxicity testing, biomonitoring, and ecogenotoxicology, in both basic and applied research. The comet assay detects between a few hundred and several thousand DNA breaks in a cell; an extent of damage that is within

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_12, © Springer Science+Business Media New York 2014 199 200 Amaya Azqueta and Andrew R. Collins

abStandard comet assay Comet assay in combination with FPG

Cells embedded in agarose Cells embedded in agarose on microscope slide on microscope slide

Lysis Lysis

Oxidized/alkylated purines

Nucleoids Nucleoids

Alkaline incubation Buffer/FPG digestion

Alkaline incubation

Electrophoresis Electrophoresis

Nucleoid without Nucleoid containing strand breaks strand breaks Buffer digestion FPG digestion

Neutralisation Neutralisation Stain and visualisation Stain and visualisation

Fig. 1 Schemes of the (a ) standard comet assay and (b ) comet assay in combination with FPG (Color ﬁ gure online)

the ability of the normal cell to repair. The protocol is quite simple to run, but it is crucial to understand every step to be able to interpret the results (Fig. 1a ). To start, cells in suspension are mixed with agarose and set as a gel on a microscope slide. They are lysed by immersion in a buffer containing a detergent and a high salt concentration. Membranes and soluble cytoplasmic and nuclear components are removed during lysis to obtain nucleoids (i.e., supercoiled DNA attached at intervals to a nuclear matrix, in effect forming loops). Then, DNA is denatured by an alkaline treatment and electrophoresis is carried out. If the DNA contains a break, the loop containing the break will be relaxed and able to migrate during the electrophoresis. In this case, and after ﬂ uorescent staining, the image that appears under the microscope resembles the stellar comets and this is the origin of the name of the assay (Fig. 2 ). The relative ﬂ uorescence in the tail of the comet represents the amount of DNA that has been able to migrate and is proportional to the number of DNA breaks present in that cell. Meanwhile, the New In Vitro Comet Assay 201

Fig. 2 Comets obtained after performing the standard comet assay with TK-6 cell treated with 70 μM H 2 O 2 during 5 min

fl uorescence in the head of the comet refl ects the intact DNA. Each individual cell gives a comet and so a measure of DNA damage. Results from all comets in a sample need to be pooled to produce the mean or median value for the % of DNA in the tail. The comet assay can be regarded as a quantitative method since calibration against the known strand-breaking effects of ionizing radiation can be done [2 ]. Based on alkaline sucrose sedimentation studies, 1 Gy of ionizing radiation produces 0.31 breaks per 10 9 Da [3 ]. Breaks per 109 Da, breaks per 106 bases or base pairs, or breaks per cell can be used as quantitative units. It is important to take into account that the alkaline treatment, before and during the electrophoresis, converts apurinic/apyrimidinic sites (AP sites) into breaks so the extension of the tail of the comet refl ects not only strand breaks but also AP sites. Though Singh et al. are considered to be the originators of the comet assay, important work was done before them [4 ]. The history of the comet assay starts with the theory of “nucleoids” elaborated by Cook et al. [5 ]. They demonstrated that the DNA of cells lysed in nonionic detergent and high concentration of salt is supercoiled and attached to a matrix at intervals, giving a series of loops. In 1978, Rydberg and Johanson described a method for checking the DNA damage in individual cells [6 ]. They used the same 202 Amaya Azqueta and Andrew R. Collins

system that is used today for immobilizing the cells, embedding them in agarose on a glass slide. Cells were lysed, unwound, and stained with acridine orange, a dye that gives different signals if bound to double- or single-stranded DNA. Some years later, in 1984, Östling and Johanson introduced the electrophoresis step and obtained what are today called comets [7 ]. They embedded γ-irradiated cells on a glass slide and lysed them in a weak alkaline solution, nearly neutral, before performing electrophoresis. The authors explained the formation of the comets by the presence of breaks inducing a relaxation of supercoiled DNA and so facilitating the migration—explicitly referring to nucleoids. Singh et al. developed the alkaline version of the assay [4 ]. They used a similar protocol to the one used nowadays including a lysis step, alkaline unwinding, and electrophoresis. The quality of the comets they obtained was very good. The authors described the alkaline version as more sensitive, being able to detect SSB, DSB, ALS, and phosphotriesters. A few years later, Olive et al. published some optimizations of the assay to detect SSBs [8 ] and DSB [ 9 ]. They proposed the tail moment as the best parameter to describe the damage of a comet. Since then, many different variations of the assays have been described. One of the most important is the inclusion of specifi c enzymes to detect specifi c DNA lesions (Fig. 1b ). Collins et al. were the fi rst to use enzymes in combination with the comet assay to increase its sensitivity [ 10 ]. The protocol included the incubation of lysed cells with endonuclease III (Endo III), a bacterial DNA repair enzyme. The enzyme removes oxidized pyrimidines and induces a break at the site of the resulting AP site. The break is detected following the rest of the standard protocol, i.e., alkaline unwinding and electrophoresis. Since then, different enzymes have been used successfully in combination with the comet assay: bacterial formamidopyrimidine DNA glycosylase (FPG) to detect oxidized purines—mainly 8-oxoGua and formamidopyrimidines (FaPy bases) that result from breakdown of alkylated or oxidized purines [ 11 – 13 ]; T4 endonuclease V (T4endoV) to detect cyclobu- tane pyrimidine dimmers [ 14 ]; bacterial AlkA to detect 3- methyladenine residues [15 ]; bacterial uracil DNA glycosylase to detect misincorporated uracil [ 16 ]; and human OGG1, the mammalian counterpart of the bacterial FPG, to detect 8-oxoGua [ 17 ]. The combination of the comet assay with enzymes gives the opportunity to detect not only breaks but specifi c lesions. Some other approaches have also been described which allow the detection of different types of lesions. The presence of DNA- DNA or protein-DNA cross-links inhibits DNA migration so they can be studied by their ability to block the migration in cells

subsequently treated with H2 O 2 or irradiated [18 ]. Collins et al. describe the use of DNA synthesis inhibitors to induce the accumulation of the DNA breaks formed during the process of New In Vitro Comet Assay 203 nucleotide excision repair [19 ]—a way of revealing damage such as bulky adducts (or UV-induced pyrimidine dimers) that do not normally present as detectable DNA breaks. Very recently the comet assay in combination with enzymes has been used to detect global methylation. Two groups have described the use of restriction endonucleases that recognize the sequence 5′CCGG-3′ but are inhibited by the presence of methylated cytosine [20 , 21 ]. The comet assay is sensitive, economical, and quick and has been employed for many years in genotoxicity testing for the in vitro screening of novel cosmetics, pharmaceuticals, and chemicals. Its sensitivity is similar to that of other genotoxicity assays (i.e., micronucleus, sister chromatid exchange, and chromosome aberration assays) [ 22 ], and it gives similar verdicts on genotoxicity as the chromosome aberration test [ 23 ]. It also showed good sensitivity and specifi city with 67 known carcinogens that gave equivocal or negative results in the micronucleus assay [24 ]. Nevertheless, due to the lack of a standard protocol and misunder- standing of critical steps, its reliability is not fully accepted among scientists, industry, and regulatory bodies. Several efforts have been made in this direction during recent years to improve the reputation of the assay as a genotoxicity tool. The in vitro comet assay is recommended under the Registration, Evaluation, Authorisation and Restriction of the Chemical Substances— REACH—programme of the European Commission. Nowadays, the in vivo comet assay is included in one of the strategies developed by the International Conference on Harmonisation for testing drugs for genotoxicity [25 ]. In the same way, the European Food Safety Authority (EFSA) has also included the in vivo version of the assay in the guidelines for testing chemicals found in foods or food-related products [ 26 ]. Very recently the Organisation for Economic Co-operation and Development (OECD) has published a draft guideline for performing the in vivo mammalian alkaline comet assay [ 27 ]. The Japanese Center for the Validation of Alternatives Methods (JaCVAM), which was in charge of the validation of the in vivo comet assay, is also taking care of the international validation trials for the in vitro version. Only the standard comet assay, without the use of enzymes, has been adopted by the regulatory bodies. (The enzyme-modifi ed assay is, however, seen by EFSA as a tool for assessing oxidative damage to DNA and establishing claims for antioxidant micronutrients [28 ].) The original, simple version of the assay detects DNA strand breaks and AP sites while many chemicals induce other types of DNA lesion, namely oxidized and alkylated bases, adducts, and cross-links. The use of the standard comet assay is sometimes justifi ed by the fact that breaks are produced during repair or spontaneous breakdown of these lesions, as intermediates. This is true but they have such a short existence that they are diffi cult to detect with the comet assay. The treatment of nucleoids from lysed cells 204 Amaya Azqueta and Andrew R. Collins

with lesion-specifi c enzymes can in part solve this issue. Enzymes such as FPG remove the altered bases and leave AP sites which are converted to breaks by an associated AP lyase activity or by the high pH used to unwind the DNA. We compared the ability of the in vitro alkaline comet assay in its standard form and in combination with FPG, to detect the genotoxicity of 11 compounds (three non-cytotoxic, two cytotoxic, and six genotoxic) in TK-6 cells [ 29 ]. We demonstrated that the use of the enzyme dramatically increases the sensitivity of the comet assay (ability to detect a genotoxic effect at non-cytotoxic concentrations) without reducing its specifi city. Therefore it would be advantageous to include the digestion with FPG in genotoxicity testing. Several strategies have been followed to increase the throughput of the comet assay. The number of samples that can be analyzed in each experiment depends on the number of slides that can be easily handled (without compromising reliability), the number of gels per slide (normally two gels per slide are used as technical replicates), and the size of the electrophoresis tank (which typically holds about 20 slides). Taking into account this picture, 20 samples can be analyzed including controls when the standard comet assay is used but only six samples can be analyzed when using one lesion- specifi c enzyme (three slides per sample; one without incubation, one incubated with buffer, one incubated with enzyme; if more than one enzymes are used the number of samples per run is further reduced). During the last few years, several high throughput approaches have been developed. Normally they imply the reduction of the volume of the agarose gel creating the so-called minigel, which contains enough number of cells to be analyzed, though other systems have also been used. GelBond fi lms [30 , 31 ], special 96-well plates [32 , 33 ], micro cell arrays [34 – 36 ], and glass microscope slides [37 – 39 ] have been used to increase the throughput. Stang and White validated their new method against the standard assay [ 32 ]. The same was done by Azqueta et al.; we obtained similar results using two conventional gels per slide, 12 minigels per slide, or 48 minigels on a GelBond fi lm [40 ]. An automated image analysis system is of great help when high throughput approaches are used. Stang et al. [41 ] and Azqueta et al. [42 ] demonstrate that automated image analysis system can be successfully used in combination with high throughput approaches. The use of high throughput methods in genotoxicity testing represents a great advantage for in vitro testing (different compounds and concentrations can be analyzed in the same run) and for in vivo studies (multiple tissue analysis). Moreover, it facilitates the use of lesion-specifi c enzymes without reducing the number of samples that can be analyzed in one experiment. The protocol to carry out the in vitro comet assay in combination with FPG using 12 minigel per slide [ 29 , 38 ] will be described in this chapter. This method gives the opportunity to incubate the New In Vitro Comet Assay 205

Fig. 3 The 12-gel incubator chamber from Severn Biotech (from the webpage www.severnbiotech.com )

nucleoids in each gel separately by using the 12-gel incubator chamber containing a silicone gasket with holes to isolate each gel (Severn Biotech: www.severnbiotech.com ) (Fig. 3 ). The slide is placed on a metal base, the gasket is placed on top of the slides without disturbing the minigels, and a plastic top plate with corresponding holes is clamped on, sealing the gasket and isolating the gels. The use of this device is not mandatory, but less enzyme is needed than if the whole slide is immersed. Both approaches, with and without using the device, are described here.

2 Materials

2.1 Equipment, General laboratory and cell culture equipment, consumables, and Consumables, reagents are needed. The speciﬁ c items for performing the assay and Reagents are listed below:

2.1.1 Equipment ● 12-gel chamber (Severn Biotech) ● Coplin jars (large capacity; takes eight slides, or 16 back-to-back) ● Electrophoresis tank (horizontal platform, e.g., from Thistle Scientifi c) ● Fluorescence microscope (plus fi lters appropriate for DAPI or other fl uorescent dyes) ● Image analysis software (not essential but recommended; several options) 206 Amaya Azqueta and Andrew R. Collins

● Incubator (simple, 37 °C, not CO2 ) ● Metal plates (large enough for about ten slides) ● Microwave oven ● Moisture chamber (plastic box with lid and rack to hold slides above water) ● Plastic trays (to contain ice for cooling down materials) ● Power supply, up to 50 V, 500 mA (to work at 4 °C) ● Water baths (37 °C, 55 °C)

2.1.2 Consumables ● Cover slips (20 × 20 mm, 22 × 22 mm) ● Glass microscope slides with frosted ends

2.1.3 Reagents ● Bovine serum albumin (BSA) ● Ethanol ● Ethylenediaminetetraacetic acid disodium salt (EDTA) ● FPG ● Low melting point agarose (LMP agarose) ● Normal melting point agarose (NMP agarose) ● Phosphate buffered saline (PBS) ● Potassium chloride (KCl) ● Potassium hydroxide (KOH) ● Sodium chloride (NaCl) ● Sodium hydroxide (NaOH) ● Tris base ● Triton X-100 ● 4-(2-Hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) ● 4′,6-Diamidino-2-phenylindole dihydrochloride (DAPI) or SYBR Gold

2.2 Solutions Composition, preparation, and storage conditions of the different and Material solutions used to perform the in vitro high throughput comet assay in combination with FPG are listed in this section. Instructions to handle potentially toxic solutions are also included.

2.2.1 Solution Composition: 1 % NMP agarose (electrophoresis grade). for Pre-coating Slides Preparation: The volume required depends on the number of slides to be pre-coated (about 100 mL is enough to pre-coat more than 100 slides). Dissolve the agarose in the correct amount of distilled water using a microwave oven. Storage: Pre-coating solution can be kept in a closed bottle at 4 °C for several months. After pre-coating some slides, the remaining solution can be stored and reused. New In Vitro Comet Assay 207

2.2.2 Agarose Composition: 1 % LMP agarose. for Embedding Cells Preparation: The volume to prepare depends on the number of gels that are going to be needed but generally only small volumes are required. Dissolve the agarose in the correct amount of PBS using a microwave oven. Prepare aliquots of about 5 mL. Storage: Aliquots can be kept in small closed glass test tubes (they can be closed by using paraﬁ lm) or small bottles at 4 °C for several months. After using an aliquot, the remaining solution can be stored and reused once (repeated heating in the microwave oven is not recommended since the solution will become more concentrated by evaporation).

2.2.3 Lysis Solution Composition: 2.5 M NaCl, 0.1 M EDTA, 10 mM Tris, 1 % Triton X-100, pH 10. Preparation: Prepare 1 or 2 L. Dissolve the correct amount of NaCl, EDTA, and Tris in distilled water (less than the fi nal volume), set pH to 10 with 10 M NaOH solution (add about 35 mL of NaOH to dissolve the EDTA, and then add dropwise to pH 10), and add the required volume of water to reach the fi nal volume. Just before use, add 1 mL Triton X-100 per 100 mL of solution. Storage: Lysis solution without Triton X-100 can be kept in a closed bottle at 4 °C for several months. Once the Triton X-100 is added the fi nal solution cannot be kept. Instructions to handle: Use gloves to handle NaOH.

2.2.4 Enzyme Reaction Composition: 40 mM HEPES, 0.1 M KCl, 0.5 mM EDTA, Buffer for FPG 0.2 mg/mL BSA, pH 8.0. Preparation: Prepare 0.5 or 1 L of a 10× concentrated stock. Dissolve the correct amount of HEPES, KCl, EDTA, and BSA in distilled water (less than the ﬁ nal volume), and set pH to 8 with 6 M KOH solution. Prepare 50 mL aliquots in plastic centrifuge tubes and store frozen. On the day of the assay, thaw an aliquot and add 450 mL of distilled water. Storage: 10× stock solution should be kept at −20 °C. Once the reaction buffer is prepared for use, it can be kept in a closed bottle at 4 °C for a week. Intructions to handle: Use gloves to handle KOH.

2.2.5 Electrophoresis Composition: 0.3 M NaOH, 1 mM EDTA (pH < 13.0). Solution Preparation: The volume to prepare depends on the size of the electrophoresis tank. Dissolve the correct amount of NaOH and EDTA in distilled water (the pH does not need to be adjusted). Storage: Electrophoresis solution should be kept in a closed bottle at 4 °C for a week. It can be reused once. Instructions to handle: Use gloves since the solution is strongly alkaline. 208 Amaya Azqueta and Andrew R. Collins

2.2.6 Neutralizing Composition: 2 mM KH2 PO4 , 10 mM Na2 HPO4 , 2.7 mM KCl, Buffer: PBS 137 mM NaCl, pH 7.4, or commercial PBS. Preparation: Prepare 1 or 2 L. Dissolve the correct amount of

KH 2 PO4 , Na2 HPO4 , KCl, and NaCl in distilled water (less than the ﬁ nal volume), set pH to 7.4 with 1 M HCl solution, and add the required volume of water to reach the ﬁ nal volume. A 20× stock can also be prepared. Storage: Neutralizing buffer or the 20× stock solution can be kept in a closed bottle at 4 °C for several months.

2.2.7 70 % Alcohol Preparation: Mix 30 mL absolute ethanol with 70 mL of distilled water. Storage: In a closed bottle at room temperature.

2.2.8 Staining Composition: 1 μg/mL of diamidino-2-phenylindole (DAPI) in Solution: DAPI distilled water. Preparation: Prepare a stock solution of 1 mg/mL in distilled water. Dilute 10 μL of the stock solution in 10 mL of water and make aliquots of 1 mL for use (working solution). Proportional volumes can be used. Storage: All aliquots and stock solution should be protected from light and stored at −20 °C. The working solution can be thawed and frozen several times. Instructions to handle: Use gloves since DAPI is possibly genotoxic. Avoid working with the powder; do not weigh it, but just add distilled water to the commercial bottle to make the stock solution or a more concentrated solution for subsequent dilution.

2.2.9 Enzymes: FPG Pure FPG can be purchased from different commercial sources, but it can also be isolated from bacteria containing overproducing plasmids. In the latter case, a crude extract, where other bacterial enzymes are present only in very small amounts, is used. The advantage of using an extract, instead of a pure enzyme, is the stability probably conferred by the presence of other proteins. If a commercial enzyme is used, the manufacturer’s instructions should be followed to store it and to prepare the working solution. In the case of a bacterial crude extract, titration experiments should be performed in order to avoid nonspeciﬁ c nuclease activity. A big batch of bacterial extract is normally produced so small aliquots should be prepared and stored at −80 °C to avoid repeated freezing and thawing. The ﬁ nal working solution varies from batch to batch.

2.2.10 Slide Preparation Ordinary clean microscope glass slides with frosted ends should be pre-coated with 1 % NMP agarose on one side:

● Dissolve or melt the agarose solution using a microwave oven (loosen the lid ﬁ rst). New In Vitro Comet Assay 209

● Keep the agarose solution in a water bath, in a glass beaker, until it reaches about 55 °C (the beaker should be big enough to introduce a slide; a staining (Coplin) jar is suitable). ● Dip a slide in the agarose solution, drain off the excess agarose, wipe the back clean, and leave it in a bench, at least overnight, to dry. ● Label the pre-coated side of the slide (mark it with a pencil on the frosted end). Pre-coated slides can be stored for years in the cardboard slide boxes they came in.

3 Methods: Step by Step

All solutions and pre-coated slides should be prepared before the day of the experiment. In what follows it is assumed that six samples (including reference standards) are to be assayed for strand breaks and FPG- sensitive sites. If the planned experiment involves treating cells, e.g., with a DNA-damaging chemical or radiation, this is normally done before embedding them in agarose. However, cells can also be treated after embedding, using the 12-gel incubator chamber. This prevents the loss of cells, but the agarose matrix can interfere with the treatment and the lack of contact between the cells can change their behavior. This approach can be suitable when very short treatments are performed (the suitability for long treatments needs to be explored) and is useful if it is desirable to minimize the ability of cells to repair damage before lysis. To start the protocol, cells must be in suspension at a concentration of 2.5 × 10 5 mL−1 in PBS (they should be trypsinized if adherent cells are used). The concentration of the cells in the gel is a critical point; there should be enough to score but not so many that they overlap to a signiﬁ cant extent as this makes scoring difﬁ - cult. Between 50 and 100 comets are scored in each minigel. Keep the cell suspension on ice throughout the process to avoid repair of the DNA damage. It is advisable to assess cell viability before performing the comet assay (or any other genotoxicity assay) to avoid misinter- preting the results. Though several assays can be used to assess the viability, it seems that the proliferation assay gives the greatest accuracy [29 , 43 ]. It is recommended that the viability of the cells should be more than 70 % [ 44 ] although some authors put the limit to 50 % [45 ]. 210 Amaya Azqueta and Andrew R. Collins

C0 C1 C2 C3 C4 C5 Lysis

C0 C1 C2 C3 C4 C5

Buffer C0 C1 C2 C3 C4 C5

C0 C1 C2 C3 C4 C5 c

C0 C1 C2 C3 C4 C5 FPG

C0 C1 C2 C3 C4 C5

Fig. 4 ( a ) Scheme of the three slides needed for an assay (C0, negative control; C1–C6, cell from ﬁ ve different concentrations). ( b) Base of the 12-gel incubator chamber. (c ) Twelve minigels placed on a pre-coated microscope slide

3.1 Embedding Cells Prepare three identical pre-coated slides. Label them as “Lysis,” in Agarose “Buffer,” and “FPG” (Fig. 4a ).

● Melt an aliquot of LMP agarose solution in a beaker of hot water or in the microwave oven (loosen the lid ﬁ rst) and place in a water bath or in a thermoblock so that it cools down to 37 °C. ● Mix 15 μL of cell suspension with 70 μL of LMP agarose solution (or equivalent proportions to maintain the ﬁ nal concentration of agarose at about 0.8 %) and pipette the mix up and down once. ● Place 2 × 5 μL drops of each agarose-cell suspension on each pre-coated slide placed on a cold metal plate, following the pattern of two rows of six minigels in each slide (or use the base of the 12-gel incubation chamber where the exact positions for two rows of six minigels is marked) (Fig. 4b, c ).

3.2 Lysis ● Place the slides in the lysis solution in a staining jar. ● Leave at 4 °C for at least 1 h.

3.3 Enzyme The concentration of FPG is a critical step. Follow the instruction Treatment (FPG) of the manufacturer in case of using a commercial enzyme or perform a titration experiment if using an extract from bacteria containing overproducing plasmid. The concentration of FPG should be enough to detect all the lesions present in the set incubation time while avoiding interference from nonspeciﬁ c nucleases. New In Vitro Comet Assay 211

Titration experiments should be performed by using cells treated with a compound that induces the speciﬁ c lesion. In this case the photosensitizer Ro 19-8022 (Hoffmann-La Roche) plus light, which induce predominantly 8-oxoguanine, can be used.

● Transfer the “Buffer” and “FPG” slides to a new jar containing enzyme reaction buffer and leave for 5 min at 4 °C. Repeat twice.

● (Meanwhile, prepare the working solution of FPG and keep it on ice.) ● Place the slide on the cold base of the 12-gel incubator chamber (which should have been in the fridge for at least 30 min) and assemble and clamp the gasket and top plate. ● Add 30 μL of cold buffer or FPG to each of the wells of the chamber of the “Buffer” or “FPG” slides, respectively, and seal the chambers (do it quickly to avoid warming the slides). ● Incubate for 30 min at 37 °C. Note: The time of incubation can vary depending on titration experiments; concentration and time of incubation are the parameters to play with to obtain the maximum conversion of damaged bases to strand breaks—without obtaining nonspeciﬁ c breaks.

● Place the chamber on ice and remove the slide carefully. If the 12-gel incubator chamber is not used, the buffer/FPG treatment should be done by immersing the slides in a bath containing either buffer or an appropriate concentration of enzymes (established by a titration experiment).

3.4 Alkaline The tank should be level. Treatment ● Place the slides in the electrophoresis tank and add cold electrophoresis solution (alkaline solution) to cover the gels. Do not forget to add the “Lysis” slides and ﬁ ll gaps on the platform with blank slides. ● Incubate for 40 min at 4 °C. It is advisable to always use the same amount of electrophoresis buffer. Note: The alkaline treatment can also be done in a Coplin jar, transferring slides to the electrophoresis tank just before performing the electrophoresis.

3.5 Electrophoresis Electrophoresis should be run at 4 °C in a cold room or a fridge.

● Connect the electrophoresis tank to the power supply. ● Set the voltage to reach approximately 1.3 V/cm across the platform carrying the slides. The current should not exceed the limit of the power supply. 212 Amaya Azqueta and Andrew R. Collins

Note: The total voltage depends on the size of the electrophoresis platform. A voltmeter with electrodes on each side of the platform gives the most accurate data.

● Leave for 20 min. Note: About 0.85 V/cm for 30 min gives very similar results.

3.6 Neutralization ● Transfer the slides to a Coplin jar containing PBS and leave for 10 min at 4 °C. ● Transfer the slides to a Coplin jar containing water and leave for 10 min at 4 °C.

3.7 Fixation This step is crucial to avoid the appearance of anomalous comets at the edge of the minigel, the so-called edge effect [ 42 ].

● Transfer the slides to a Coplin jar containing 70 % ethanol and leave for 15 min at room temperature. ● Transfer the slides to a Coplin jar containing absolute ethanol and leave for 15 min at room temperature. ● Leave the slides at room temperature.

3.8 Staining ● Place a drop of the working solution of DAPI onto each minigel and cover with a large coverslip (24 × 60 mm). Alternative stains can be used: SYBR Gold (following manufacturer’s instruction), propidium iodide (2.5 μg/mL), Hoechst 33258 (0.5 μg/mL), ethidium bromide (20 μg/mL), etc. Note: Depending on which stain is used, you will need an appropriate ﬁ lter set in the ﬂ uorescence microscope.

3.9 Quantitation The number of comets that should be analyzed is between 50 and 100 per minigel. Anomalous comets at the edge of the minigel, if present, should not be evaluated. There are two ways to quantify the DNA damage: visually or using computer image analysis software. Visual scoring involves the quantitative analysis of the comets by simple microscopic examination. Comets are classifi ed into fi ve categories depending on the intensity of tail fl uorescence: from class 0 (undamaged comets without a visible tail) to class 4 (highly damaged comets with nearly all the DNA in the tail) (Fig. 5 ). A total comet visual score from the minigel is calculated following this formula: (0 × % of comets in class 0) + (1 × % of comets in class 1) + (2 × % of comets in class 2) + (3 × % of comets in class 3) + (4 × % of comets in class 4). This value goes from 0 to 400 arbitrary units. There are different computer image analysis systems on the market. In all cases a digital camera is mounted on the microscope. The software analyzes the comet image and measures different parameters: % of total fl orescence in head and tail, tail length, and “tail moment”. It is recommended to use % of total fl orescence in New In Vitro Comet Assay 213

Fig. 5 Typical comet images of the different categories according to visual scoring

tail (% DNA in tail). There are two types of computer image analysis software: semiautomatic, where the operator selects the comets to be analyzed; and automatic, where the software selects the comets. In some cases, it may be useful or necessary to analyze individual comet data, for instance, looking at dispersion of data to check whether subpopulations of comets with different degrees of damage are present. However, for most purposes, the overall damage level of the whole population of comets is the important parameter. Therefore, to pool the results, the median % tail DNA of all the scored comets in a minigel should be calculated, since this prevents undue weight being given to a few highly damaged cells. In some cases, where a wide distribution of comets is seen, the mean % tail DNA may be more meaningful. As was explained in Sect. 1 , the actual DNA break frequency can also be calculated by using calibration curves. This can be useful in some cases; for instance, it was possible in the ESCODD project [46 ] to estimate the frequency of 8-oxoguanine relative to unaltered guanine and to compare the accuracy of chromatographic and FPG-based methods in making this measurement. Highly damaged comets containing almost all the DNA in tail, the so-called hedgehog comets, should always be analyzed; “highly damaged” is a relative term and reﬂ ects the sensitivity of the comet assay. Such levels of damage have been shown clearly to be repair- able and so are compatible with life. They may represent the very earliest stage of the process of apoptosis but certainly should not be regarded as diagnostic of cell death, as is often assumed [47 ]. Results obtained with visual scoring, semiautomatic image analysis, and fully automated image analysis systems are in agreement qualitatively, although all of them have some weak points [41 ].

3.10 Calculations Different lesions are obtained with the different slides. “Lysis” and Interpretation slides reveal the strand breaks presented in the samples; “Buffer” of the Results slides provide an estimation of the strand breaks plus any nonspe- ciﬁ c breaks induced during the incubation; and “FPG” slides measure the strand breaks plus any “buffer-induced” breaks plus the FPG-sensitive sites. Results obtained in “Lysis” and “Buffer” slides are usually very similar; if the value for the “Buffer” slide is high, the buffer should be discarded and the experiment repeated, if possible. Subtracting the % of DNA in tail (or arbitrary units) for 214 Amaya Azqueta and Andrew R. Collins

the “Buffer” slides from that for the “FPG” slides, for each of the samples, gives the “net FPG-sensitive sites”. This calculation is correct when the % DNA in tail values of both the “Buffer” and the “FPG” slides are on the linear part of the dose-effect curve of the comet assay (% DNA in tail plotted against DNA break frequency or X-ray dose) . If some of the values are out of the linear part of the curve, a calculation of the frequency of breaks should be done (using an X-ray calibration curve) before performing the subtrac- tion; otherwise there will be an underestimation of the FPG- sensitive sites. The % of DNA in tail obtained with cells treated with up to 10 Gy of X- or γ-irradiation shows a more or less linear response until about 75 % [2 ]. The mean of the technical replicate samples should be calculated; the biological or experimental unit is the cell culture and only one data point should be obtained per cell sample, condition, and experiment. In genotoxicity studies a negative control, of non-treated or solvent-treated cells, is always included. A positive control should also be included; in this case the best positive control is cells treated with photosensitizer Ro 19-8022 plus light to produce FPG- sensitive sites (predominantly 8-oxoguanine). Statistical approaches to analyze comet assay results are very well described in the papers written by Lovell and Omori [48 ] and Lovell [ 49 ]. Important issues are, for instance, the key concept of “experimental unit,” identifi cation of suitable endpoints, the choice between mean and median, and the appropriate statistical tests to be used in different circumstances. In the case of the in vitro assay, normally three independent experiments are enough to show the genotoxic effect. It is diffi - cult to properly apply a statistical analysis when the size of the sample is so low. Although the statistical signifi cance of the results has to be taken into account, the biological meaning is the key concept. It is important to test different concentrations of the test compound in order to detect a dose-response. Reproducibility (essentially, whether similar results are obtained when experiments are performed on different occasions) has also to be taken into consideration.

4 Discussion and Conclusion

Considerable efforts have been put into attempts to standardize the protocol for carrying out the comet assay. In 2011, two research groups identifi ed some of the critical points of the comet assay [ 50 , 51 ]. The % of DNA in tail varies with the agarose concentration, duration of alkaline incubation, electrophoresis voltage, and electrophoresis duration. The electrophoresis step is crucial, % DNA in tail increasing steeply with the increase of the voltage or New In Vitro Comet Assay 215 the duration (two parameters that can be varied reciprocally to a certain extent). It is important to measure the exact voltage gradient (volts per centimeter across the platform holding the slides) by using a voltmeter. This measurement can be done once and the same conditions reproduced in each experiment (i.e., voltage and run time, volume of electrophoresis solution, temperature, and rows and position of slides). Changes in the current have little effect on the outcome. Current is varied by changing the volume of the electrophoresis buffer so any effect observed on comet tail formation is due to consequent changes in the voltage (since voltage depends on resistance which depends on the depth of electrophoresis solution). High current can increase the temperature so electrophoresis should be run in a fridge or a cold room (at about 4 °C). The protocol described in this chapter incorporates the recommendations given in the papers of Ersson and Möller [ 50 ] and Azqueta et al. [ 51 ] in order to increase the sensitivity and reproducibility of the assay. Gutzkow et al. demonstrated that recirculation of the electrophoresis buffer during electrophoresis decreases the variability of the results between minigels on GelBond fi lms [31 ]. This surely also applies also to the 12 minigels per slide. The period spent by slides in lysis solution is generally regarded as unimportant, and times between 1 h and a week are common. However, Banáth et al. studied the effect of lysis duration and found that longer lysis increases the effi ciency for detecting radiation-induced strand breaks [ 52 ] . More studies are needed to reach a defi nitive conclusion. It was demonstrated by Azqueta et al. that the use of the comet assay in combination with FPG increases the sensitivity of the assay in two ways, identifying known genotoxic compounds such as benzo( a )pyrene that do not act via strand breakage (i.e., reducing the “false negatives”) and detecting the effects at considerably lower concentrations [ 29 ]. The specifi city of the comet assay in combination with FPG was no different from that of the standard comet assay, i.e., there was no increase in “false positives.” The regulatory authorities have yet to be persuaded of this advantage, and an OECD guideline for the use of the FPG-comet assay is a long way off. The use of the 12-gel system and the 12-gel incubator chamber increases the throughput of the assay since more compounds can be tested at once. This approach can be further simplifi ed by omitting lysis and buffer slides and relying on the results with FPG, since in genotoxicity testing the distinction of different lesions, SB and FPG-sensitive sites, is not needed. Some testing is needed in order to validate this new simplifi ed version. The application of this approach to in vivo studies has not been explored though it is reasonable to suppose that it will work. 216 Amaya Azqueta and Andrew R. Collins

References

1. Cortés-Gutiérrez EI, Dávila-Rodríguez MI, 14. Collins AR, Mitchell DL, Zunino A et al Fernández JL et al (2011) New application of (1997) UV-sensitive rodent mutant cell lines of the comet assay: chromosome—comet assay. J complementation groups 6 and 8 differ pheno- Histochem Cytochem 59:655–660 typically from their human counterparts. 2. Collins AR, Azqueta-Oscoz A, Brunborg G Environ Mol Mutagen 29:152–160 et al (2008) The comet assay: topical issues. 15. Collins AR, Dusinska M, Horska A (2001) Mutagenesis 23:143–151 Detection of alkylation damage in human lym- 3. Ahnström G, Erixon K (1981) Measurement phocyte DNA with the comet assay. Acta of strand breaks by alkaline denaturation and Biochim Pol 48:611–614 hydroxyapatite chromatography. In: Friedberg 16. Duthie SJ, McMillan P (1997) Uracil misin- EC, Hanawalt PC (eds) DNA repair. A labora- corporation in human DNA detected using tory manual of research procedures. Marcel single cell gel electrophoresis. Carcinogenesis Dekker, New York 18:1709–1714 4. Singh NP, McCoy MT, Tice RR et al (1988) A 17. Smith CC, O’Donovan MR, Martin EA (2006) simple technique for quantitation of low levels hOGG1 recognizes oxidative damage using of DNA damage in individual cells. Exp Cell the comet assay with greater specifi city than Res 175:184–191 FPG or ENDOIII. Mutagenesis 21:185–190 5. Cook PR, Brazell IA, Jost E (1976) 18. Spanswick V, Hartley J, Hartley J (2010) Characterization of nuclear structures contain- Measurement of DNA interstrand crosslinking ing superhelical DNA. J Cell Sci 22:303–324 in individual cells using the single cell gel elec- 6. Rydberg B, Johanson KJ (1978) Estimation of trophoresis (comet) assay. In: Fox KR (ed) single strand breaks in single mammalian cells. Drug-DNA interaction protocols, methods in In: Hanawalt PC, Friedberg EC, Fox CF (eds) molecular biology, vol 613. Humana, DNA repair mechanisms. Academic, New York New York, pp 267–282 7. Östling O, Johanson KJ (1987) Bleomycin, in 19. Collins AR, Dobson VL, Dusinská M et al contrast to gamma irradiation, induces extreme (1997) The comet assay: what can it really tell variation of DNA strand breakage from cell to us? Mutat Res 375:183–193 cell. Int J Radiat Biol 52:683–691 20. Wasson GR, McGlynn AP, McNulty H et al 8. Olive PL, Banath JP, Durand RE (1990) (2006) Global DNA and p53 region-specifi c Heterogeneity in radiation-induced DNA hypomethylation in human colonic cells is damage and repair in tumor and normal cells induced by folate depletion and reversed by measured using the “comet” assay. Radiat Res folate supplementation. J Nutr 136: 122:86–94 2748–2753 9. Olive PL, Wlodek D, Banath JP (1991) DNA 21. Wentzel JF, Gouws C, Huysamen C et al double-strand breaks measured in individual (2010) Assessing the DNA methylation status cells subjected to gel electrophoresis. Cancer of single cells with the comet assay. Anal Res 51:4671–4676 Biochem 400:190–194 10. Collins AR, Duthie SJ, Dobson VL (1993) 22. Uhl M, Helma C, Knasmüller S (1999) Single- Direct enzymic detection of endogenous oxi- cell gel electrophoresis assays with human- dative base damage in human lymphocyte derived hepatoma (Hep G2) cells. Mutat Res DNA. Carcinogenesis 14:1733–1735 441:215–224 11. Dusinska M, Collins AR (1996) Detection of 23. Hartmann A, Agurell E, Beevers C et al (2003) oxidised purines and UV-induced photoprod- Recommendations for conducting the in vivo ucts in DNA of single cells, by inclusion of alkaline Comet assay. Mutagenesis 18:45–51 lesion-specifi c enzymes in the comet assay. 24. Kirkland D, Speit G (2008) Evaluation of the Altern Lab Anim 24:405–411 ability of a battery of three in vitro genotoxicity 12. Li Q, Laval J, Ludlum DB (1997) FPG protein tests to discriminate rodent carcinogens and releases a ring-opened N-7 guanine adduct non-carcinogens: III. Appropriate follow-up from DNA that has been modifi ed by sulfur testing in vivo . Mutat Res 654:114–132 mustard. Carcinogenesis 18:1035–1038 25. ICH harmonised tripartite guideline S2 (R1) 13. Speit G, Schütz P, Bonzheim I et al (2004) (2011) Guidance on genotoxicity testing and Sensitivity of the FPG protein towards alkyla- data interpretation for pharmaceuticals intended tion damage in the comet assay. Toxicol Lett for human use. International conference on 146:151–158 harmonisation of technical requirements for New In Vitro Comet Assay 217

registration of pharmaceuticals for human use. processing of multiple samples. Mutat Res http://www.ich.org/products/guidelines/ 721:153–156 safety/article/safety-guidelines.html 40. Azqueta A, Gutzkow KB, Priestley CC et al 26. EFSA Scientifi c Committee (2011) Scientifi c (2013) A comparative performance test of opinion on genotoxicity testing strategies standard, medium- and high-throughput applicable to food and feed safety assessment. comet assays. Toxicol In Vitro 27:768–773 EFSA J 9:2379 41. Stang A, Brendamour M, Schunck C et al 27. Organisation for Economic Cooperation & (2010) Automated analysis of DNA damage in Development (OECD) (2013) Draft OECD the high-throughput version of the comet guideline in vivo mammalian alkaline comet assay. assay. Mutat Res 698:1–5 In: OECD guidelines for testing of chemicals. 42. Azqueta A, Meier S, Priestley C et al (2011) 28. EFSA Panel on Dietetic Products, Nutrition The infl uence of scoring method on variability and Allergies (2011) Guidance on the scientifi c in results obtained with the comet assay. requirements for health claims related to anti- Mutagenesis 26:393–399 oxidants, oxidative damage and cardiovascular 43. Kirkland D (2011) Improvements in the reli- health. EFSA J 9:2474 ability of in vitro genotoxicity testing. Expert 29. Azqueta A, Arbillaga L, Lopez de Cerain A Opin Drug Metab Toxicol 7:1513–1520 et al (2013) Enhancing the sensitivity of the 44. Tice RR, Agurell E, Anderson D et al (2000) comet assay as a genotoxicity test, by combin- Single cell gel/comet assay: guidelines for in ing it with bacterial repair enzyme vitro and in vivo genetic toxicology testing. FPG. Mutagenesis 28:271–277 Environ Mol Mutagen 35:206–221 30. McNamee JP, McLean JRN, Ferrarotto CL 45. Burlinson B (2012) The in vitro and in vivo et al (2000) Comet assay: rapid processing of comet assays. In: Parry JM, Parry EM (eds) multiple samples. Mutat Res 466:63–69 Genetic toxicology: principles and methods. 31. Gutzkow KB, Langleite TM, Meier S et al Springer, New York (2013) High-throughput comet assay using 96 46. Gedik CM, Collins A, European Standards minigels. Mutagenesis 28:333–340 Committee on Oxidative DNA Damage 32. Stang A, Witte I (2009) Performance of the ESCODD (2005) Establishing the background comet assay in a high-throughput version. level of base oxidation in human lymphocyte Mutat Res 675:5–10 DNA: results of an interlaboratory validation 33. Stang A, Witte I (2010) The ability of the study. FASEB J 19:82–84 high-throughput comet assay to measure the 47. Lorenzo Y, Costa S, Collins AR et al (2013) sensitivity of fi ve cell lines toward methyl meth- The comet assay, DNA damage, DNA repair anesulfonate, hydrogen peroxide, and penta- and cytotoxicity: hedgehogs are not always chlorophenol. Mutat Res 701:103–106 dead. Mutagenesis 28:427–432 34. Mercey E, Obeïd P, Glaise D et al (2010) The 48. Lovell DP, Omori T (2008) Statistical issues in application of 3D micropatterning of agarose the use of the comet assay. Mutagenesis 23: substrate for cell culture and in situ comet 171–182 assays. Biomaterials 31:3156–3165 49. Lovell DP (2009) Statistical analysis of comet 35. Wood DK, Weingeist DM, Bhatia SN et al assay data. In: Anderson D, Dhawan A (eds) (2010) Single cell trapping and DNA damage The comet assay in toxicology. The Royal analysis using microwell arrays. Proc Natl Acad Society of Chemistry, Cambridge Sci U S A 107:10008–10013 50. Ersson C, Möller L (2011) The effects on 36. Weingeist DM, Ge J, Wood DK et al (2013) DNA migration of altering parameters in the Single-cell microarray enables high- throughput comet assay protocol such as agarose density, evaluation of DNA double-strand breaks and electrophoresis conditions and durations of the DNA repair inhibitors. Cell Cycle 12:907–915 enzyme or the alkaline treatments. Mutagenesis 37. Ritter D, Knebel J (2009) Genotoxicity testing 26:689–695 in vitro —development of a higher throughput 51. Azqueta A, Gutzkow KB, Brunborg G et al analysis method based on the comet assay. (2011) Towards a more reliable comet assay; Toxicol In Vitro 23:1570–1575 optimising agarose concentration, unwinding 38. Shaposhnikov S, Azqueta A, Henriksson S et al time and electrophoresis conditions. Mutat Res (2010) Twelve-gel slide format optimised for 724:41–45 comet assay and fl uorescent in situ hybridisa- 52. Banáth JP, Kim A, Olive PL (2001) Overnight tion. Toxicol Lett 195:31–34 lysis improves the effi ciency of detection of 39. Zhang LJ, Jia JF, Hao JG et al (2011) A modi- DNA damage in the alkaline comet assay. fi ed protocol for the comet assay allowing the Radiat Res 155:564–571 Chapter 13

The Comet Assay In Vivo in Humans

Carla Costa and João Paulo Teixeira

Abstract

In the last decades, the comet assay has been used as a molecular biomarker to detect DNA damage in several human biomonitoring studies. Nevertheless, many of these studies are often inconclusive or offer poor quality data due to study design constraints. This chapter describes not only the experimental protocol to carry the alkaline version of the comet assay in human cells but also considers many of the obstacles that the researcher faces while carrying a human biomonitoring study, such as sample size, sample type, sample collection, statistics, and result communication. Understanding assay limitations, both experimental and biological, is essential to improve both data quality and data relevance obtained and to guarantee that comet assay continues to provide enhanced reliability as a biomarker in human biomonitoring studies.

Key words Comet assay , Human biomonitoring , Exposure , Lymphocytes , Whole blood

1 Comet Assay for Human Biomonitoring

Comet assay (also designated of single cell gel electrophoresis) is a rapid, simple, visual, and sensitive method to detect DNA breakage in eukaryotic cells [1 ]. Briefl y, cells are embedded in agarose on a microscope slide, lysed with detergent to remove cell and nuclear membranes, and treated with a high salt solution. Nucleoids are formed, containing non-nucleosomal but still supercoiled DNA. Any breaks present in the DNA cause the supercoiling to relax locally and loops of DNA are then free to extend toward the anode during electrophoresis. At the end of this procedure, damaged cells will resemble comets (hence the name comet assay ) while non-damaged cells maintain their round appearance. Comets are viewed by fl uorescence microscopy after staining with a suitable fl uorescent DNA-binding dye [ 2 ]. The pH of unwinding solution is responsible for differences in method sensitivity, as the conversion of alkali-labile sites to strand breaks is a process dependent on the pH and time of alkaline treatment. Under alkaline (pH > 13) conditions, the assay potentially

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_13, © Springer Science+Business Media New York 2014 219 220 Carla Costa and João Paulo Teixeira

detects single- and double-stranded breaks, incomplete repair sites, and alkali-labile sites in virtually any eukaryotic cell population that can be obtained as a single cell suspension [ 3 ]. Throughout the years, there were several modifi cations and innovations to the original protocol which led to an array of comet assay variants that detect different types of DNA alterations. Specifi c types of DNA lesions can be measured by using lesion-specifi c repair enzymes [4 ]; the comet assay can also analyze DNA repair capacity of cells [ 5 ], DNA cross-links [ 6 ], gene-specifi c DNA damage [7 ], and also DNA methylation [8 ]. This technique provides information on the genotoxic potential of chemicals (genotoxicity testing) as well as the chemoprotec- tive properties of others (antigenotoxicity testing); its application on suitable organisms can provide data on genetic damage due to environmental contamination (ecological monitoring or ecogenotoxicology); another main application concerns human biomonitoring that examines the effects of occupational and environmental exposure to different agents. In addition to all these uses, comet assay also offers the possibility to further understand DNA damage and DNA repair [9 ]. This chapter is particularly focused on the application of the alkaline version of the comet assay as a molecular biomarker to detect DNA damage in human biomonitoring studies. Biomarkers are biological indicators of the body’s response to exposure and indicate early subclinical changes, which if sustained, may yield pathological consequences [ 10 ]. Furthermore, biomarkers may provide information on human exposure and risk indication of ill-health [11 ]. The high sensitivity of the comet assay makes it a very useful biomarker particularly in cases of low-dose exposures. It is, in most cases, accepted that the range of damage detectable with the comet assay is roughly from one hundred to several thousand breaks per human. This interval covers physiologically relevant levels of damage, ranging from the normal background levels of damage and damage levels induced by nonlethal doses of cell- damaging agents [12 ]. The relevance of the endpoints measured with the alkaline comet assay is still not fully understood as they mirror a temporary strand break which, under normal circumstances, will be repaired in a short period before being fi xed as a mutation [ 13 ], and often constitutes an error-free DNA repair process [ 14 ]. Nevertheless, several studies have been showing a very good agreement between DNA damage assessed by alkaline version of the comet assay and other cytogenetic tests [ 13 ] that detect fi xed damage (chromosomal aberrations) [ 15 ]. Biological relevance of the assay can be signifi cantly increased if associated with lesion-specifi c endonucleases; these will allow the detection of altered bases (e.g., oxidized purines, oxidized The Comet Assay In Vivo in Humans 221

pyrimidines, and alkylated bases), which are biologically more signifi cant than DNA breaks, since they are more likely to lead to mutation [ 2 ]. For further reading on this matter, we recommend Chap. 12 of this book. Comet assay has been used in biomonitoring studies to assess (1) clinical status, (2) dietary effects, (3) occupational exposure, and (4) environmental exposure. There are two basic approaches to assess if exposure is associated with a particular outcome (in this case, genetic damage): experimental (or intervention) studies and observational studies. In intervention studies (usually limited to nutraceutical research), subjects are treated with a specifi c agent (suspected to be benefi cial), and the effects on DNA damage levels are monitored and compared with placebo (or no treatment) group [16 ]. In some cases, results obtained after treatment are compared with the ones attained before treatment [17 ]. This same type of result analysis has also been carried out in studies looking at genetic effects after chemotherapy in cancer patients [18 ]. Observational studies using comet assay often comprise case–control and cross- sectional study designs. Case-control studies compare diseased with disease-free individuals (e.g., [19 ]); cross-sectional are used to determine the levels of genetic damage associated to a specifi c event (either exposure or disease) in a defi ned population, at a particular point in time. Results are also usually compared with a referent population [20 ]. Each study design has its own limitations and strengths; the researcher must understand the specifi cities of each study design in order to design a good quality study and envisage all the potential obstacles; we suggest Silva [ 21 ] for further reading on this matter. In the next section, some basic issues related to study design are described.

2 Considerations on Study Design

Human biomonitoring studies are affected by both random and systematic errors (bias). Random error is related to individual biological variation, sampling error, and measurement errors and can be reduced with appropriate sample size and protocol standardization that will lead to an improvement in study precision. On the other hand, systematic errors are associated with loss of study validity and can be grouped in two main types: selection bias and measurement/classiﬁ cation bias. Regardless of the type of assay used, some questions must always be addressed in study design in order to obtain reliable results. Albertini et al. [22 ] provides guidelines for study planning and result interpretation in human monitoring studies. 222 Carla Costa and João Paulo Teixeira

Establishment of study populations, concretely subject selection, participant eligibility, and defi nition of referent populations are a main issue of study design. A list of criteria for inclusion and exclusion in each group (exposed and referent) must be clearly defi ned by the researcher. These criteria must be established before the subjects are recruited for the study to ensure that no biases enter into the selection process. All epidemiological studies must be reviewed and approved by appropriate ethical authorities, and all data must be kept confi den- tial. A prerequisite for participation is a signed informed consent from each subject based on an adequate and understandable explanation of the intent of the study; participants must also be informed in advance of the meaning of results both on an individual and at group level in order to avoid false expectations [22 ]. Participants must also be aware if it is planned to store the collected samples or perform additional analyses in the future (specimen banking). In consideration to sample size, it is necessary to ensure that the proposed number of subjects to be recruited will be appropriate to answer the main objective of the study. In addition, it is important to bear in mind that the numbers have to be large enough to allow for failure of compliance, unacceptable controls, or loss of samples due to a failure of transport or laboratory handling [ 22 ]. Statistical planning, namely the size and power of the study, must be among priorities to be addressed in order to improve the quality of comet assay biomonitoring studies [23 ]. Statistical power depends on sample size, signifi cance level, and effect size. In order to assure that the planned study has enough statistical power (ability to fi nd a difference when a real difference exists), a priori power analysis is useful to estimate adequate sample size; specifi c software that allows comparison of study power using different combinations of research parameters are available and may be very useful for researchers during study design (e.g., IBM SPSS SamplePower); generally, a power of 80 % or higher is regarded as acceptable. To address the issue of power, it might be necessary to run pilot studies to estimate intra- and interindividual variability; in addition, power estimates must rely on plausible assumptions (e.g., for one-tailed tests choice) and cost/ benefi t analysis. Confounding is a serious source of bias that can affect results obtained in human biomonitoring studies. Regarding comet assay, the most commonly evaluated confounding factors include age, gender, smoking habits, alcohol consumption, exercise, and drug intake [ 23 ] in accordance to what has been previously recommended by Møller et al. [ 24 ]. It is important to notice that some confounding may not be relevant for all the versions of the comet assay. For example, studies have reported seasonal variation of DNA damage, whereas enzyme-sensitive sites do not appear to show the same pattern [25 ]. Confounding can be controlled at the The Comet Assay In Vivo in Humans 223

design stage by randomization (ensuring that potential confounding variables are equally distributed in the established groups of experimental studies), restriction (limiting the study to populations that present the confounding characteristics), and matching (potential confounding variables are evenly distributed in the two compared groups) [26 ]. Nevertheless, while randomization is diffi cult to carry out, restriction and matching may limit the informativeness of the study and are incompatible with multiple confounders [27 ]. For this reason, it is usual to address confounding at result analysis stage by using stratifi cation and statistical modeling. This last method can be particularly useful if one is dealing with a great number of confounders. As mentioned later on this chapter, the usage of statistical models to control for confounding is usually associated to a loss of power that can only be compensated by an increase in sample size. All possible confounders must be identifi ed during study design in order to allow data collection. This information may be gathered by, for example, questionnaires, interviews (for ethnic group, age, sex, tobacco smoking, drug usage, exposure to X-rays, or chemical exposure at home), or even laboratorial analyses (in the case of genetic polymorphisms) [22 ]. DNA damage identifi ed by the comet assay is not chemical specifi c, and therefore the association with exposure to a particular xenobiotic must be established by an independent measure as, for example, the analysis of the chemical/metabolite in biological matrices [ 22 ]. Environmental concentrations and/or chemical- specifi c biomarkers of exposure constitute important and valuable information in environmental and occupational monitoring studies. This information improves study validity and overall interpretation of data as these measurements form a good basis to establish who is exposed and, more importantly, the degree of exposure. Studies relying only on questionnaire data for exposure status categorization may suffer of considerable misclassifi cation. Selection of human specimen type, timing for sample collection, and sample integrity are also key issues on study design, but these will be considered separately in the next section.

3 Sample Type, Collection, and Processing

Human biomonitoring requires a suitable biological matrix of easy access in sufﬁ cient amount for routine procedures but preferably with diminutive impact for the individual [ 28 ]. Blood and urine are the most commonly used matrices, but hair, exhaled air, teeth, nails, and saliva can also be used for toxicological quantiﬁ cations. Human biomonitoring studies use mainly peripheral blood cells (particularly lymphocytes) as the surrogate tissue, as these cells are easily obtained, in large numbers, and do not require cell culture, 224 Carla Costa and João Paulo Teixeira

are diploid, and are almost all in the same phase of the cell cycle [12 , 13 ]. In addition, it has been hypothesized that genetic damage observed in peripheral blood lymphocytes (PBLs) refl ects a similar damage of cells of target tissues [ 29 ] and therefore can be used as a surrogate for the target tissue in which alterations can occur [30 ]. Nevertheless, this assay can be performed in other types of cells that in some cases may be more appropriate considering the exposure route, exposure duration, and target tissue. In this context, aside from lymphocytes, the assay has been carried out using whole blood [ 31 ], buccal epithelial cells [32 ], tear duct epithelial cells [33 ], and sperm [34 ]. The selection of the cell population to use must be based on exposure conditions, persistence of DNA damage, and rates of cell turnover. Albertini et al. [ 22 ] recognize some additional key issues regarding sample collection and processing; these are timing of sample collection, concurrent and identical handling of samples from both exposed and referent subjects, and blinded analysis of coded samples. Sampling should be performed during a relevant time period with respect to exposure; for identifi cation of possible DNA damaging agents, the optimal timing for sample collection from any cell population is during long-term, chronic exposure when the induction and repair of DNA damage are presumed to be at steady- state equilibrium [20 ]. Below, one can fi nd the protocols for blood sample collection (as this is the most commonly analyzed tissue), lymphocyte isolation, sample preservation, and thawing. In addition, some considerations are made regarding sample quality assessment methods as this is still a particularly controversial and widely discussed issue.

3.1 Sample Blood collection must be assigned to a trained phlebotomist, famil- Collection iarized with good-work practices and different blood-sampling sys- (and Transportation) tems to perform venepuncture [ 35 ]. Samples are usually collected in a tube already containing an anticoagulant to prevent clotting; different anticoagulants have been used but EDTA is generally used as a small concentration of this substance may prevent endonuclease activity. The possible effect of different anticoagulants on the comet assay results remains unclear [ 23 ]. In alternative to venepuncture, blood can be collected by ﬁ n- ger prick; a much smaller volume will be available but enough to run the comet assay. To samples that will not be immediately processed, an adequate volume of anticoagulant must be added to prevent clotting. If samples are not collected at the laboratory (the most common situation), they must be transported under refrigeration (at 4 °C). The Comet Assay In Vivo in Humans 225

3.2 Lymphocyte If blood was drawn to cell separation tubes, e.g., BD CPT™, Isolation centrifuge in accordance to manufacturer’s instructions to obtain the lymphocytes/monocytes cell suspension and skip to point 4: 3.2.1 After Blood Collection by Venepuncture 1. Layer 3.5 mL of blood onto 5 mL of density gradient medium (e.g., Lymphoprep™) in a 15 mL conical plastic centrifuge tube. 2. Centrifuge for 30 min at 700 × g (room temperature). 3. Remove band containing lymphocytes, just above the density gradient medium using a pipette; transfer to a 15 mL plastic centrifuge tube. 4. To the lymphocytes/monocytes cell suspension, add 5 mL of ice-cold PBS pH 7.2 (or pH 7.4) and invert the tube. Centrifuge for 10 min at 200 × g (room temperature). 5. Discard the supernatant, resuspend the pellet in 1 mL of ice-cold PBS pH 7.2, and add another 4 mL of PBS pH 7.2. Invert the tube. Centrifuge for 10 min at 200 × g (room temperature). 6. Repeat the previous step. 7. Discard the supernatant, resuspend the pellet in 1 mL of ice- cold PBS pH 7.2, and transfer the suspension to a microtube. 8. Take a sample for cell count. 9. Aliquot your suspension to obtain optimal cell number. Note : This will depend on how the assay will be performed ; for two gels per slide , 10 4 cells / gel is recommended. 10. Centrifuge at 200 × g for 3 min. Remove as much supernatant as possible using micropipette and use the pellet to perform the comet assay (see Sect. 4.3 , step 5).

3.2.2 After Blood 1. Take ≈50 μL of blood from ﬁ nger prick and add 1 mL of cold Collection by Finger Prick PBS pH 7.2 in a 1.5 mL microtube. Mix and leave on ice. 2. Underlay with 100 μL of density gradient medium (e.g., Lymphoprep™), using a micropipette. Mark the boundary between PBS and density gradient medium with a line. 3. Centrifuge at 200 × g for 3 min (4 °C). 4. Retrieve lymphocytes in 100 μL from just above the marked boundary, using pipette (to another microtube). 5. Suspend the retrieved lymphocytes in 1 mL of cold PBS pH 7.2. Leave on ice. 6. Aliquot your suspension to obtain optimal cell number (this will depend how the assay will be performed; for two gels per slide, 10 4 cells/gel is recommended). 7. Centrifuge again. Remove as much supernatant as possible using pipette and use the pellet to perform the comet assay (see Sect. 4.3 , step 5). 226 Carla Costa and João Paulo Teixeira

3.3 Sample If time between sampling and processing is variable and/or Preservation excessive, it may be preferable to cryopreserve cells [22 ]. Although many studies perform comet assay in fresh samples, it is generally accepted that both cryopreserved lymphocytes and cryopreserved whole blood are suitable for the alkaline version of the comet assay [ 36 , 37 ]. Sample stability is affected by numerous factors [38 ], but timing before initial processing and temperature are of particular concern for comet assay. Whenever possible, sample processing must immediately follow sample collection as time between these two steps may both decrease or increase damage (by DNA repair processes or inadequate transport/storage conditions, respectively) and therefore increases bias [ 22 ]. Different studies have looked at sample stability at 4 °C for short-term storage, but confl icting results have been reported for whole blood samples; while some authors fi nd that these are not stable for 24 h [ 31 , 39 ], others suggest a stable 4-day period [40 ]. Another work suggests that isolated human lymphocytes are slightly more stable at 4 °C presenting stable levels of DNA damage up to 48 h [41 ]. Despite all this, if samples are not immediately analyzed, they are usually stored at −80 °C. At this temperature, whole blood samples shown to be stable for at least 4 months [ 36 ], while lymphocytes are expected to present stable levels of DNA damage for several months or even years [9 ]. Below, one can fi nd some overall guidelines for lymphocytes and whole blood freezing. In alternative to −80 °C storage, cells can also be stored in liquid nitrogen.

3.3.1 Lymphocytes 1. After cell count (step 8 of lymphocyte isolation), centrifuge at 200 × g for 3 min. 2. Remove supernatant, resuspend pellets in freezing medium (90 % heat-inactivated fetal bovine serum (FBS) and 10 % dimethyl sulfoxide (DMSO)) at 3 × 10 6 cells/mL, and transfer them to labeled cryotubes or microtubes. 3. Place the tubes in a controlled-rate freezing apparatus (decreasing temperature approximately 1 °C per minute) in the −80 °C freezer, and leave overnight. Note : Slow freezing is essential to preserve DNA without shearing; if there is no freezing box in your lab, you may use styrofoam racks to initiate the freezing process. 4. Transfer to labeled boxes and store at −80 °C or in liquid nitrogen.

3.3.2 Whole Blood 1. To the blood sample, add an equal amount of a 20:80 (v/v) mixture of DMSO and RPMI 1640 cell culture medium. The Comet Assay In Vivo in Humans 227

2. Place tubes in a controlled-rate freezing apparatus (decreasing temperature approximately 1 °C per minute) in the −80 °C freezer and leave overnight (see above). 3. Transfer to labeled boxes and store at −80 °C or in liquid nitrogen. Although this is the most frequent procedure for sample preservation, Al-Salmani et al. [42 ] showed that whole blood can be preserved in small aliquots (of approximately 250 μL) without cryopreservative for up to 1 month at −80 °C without artifactual formation of DNA damage.

3.4 Sample Thawing 1. Place the tube in a 37 °C water bath. Note: For whole blood thawing , this is the only necessary step ; for lymphocytes it is necessary to perform the rest of this protocol 2. As soon as the ice is melted, transfer the sample to a conical centrifuge tube containing 15 mL of prechilled thawing medium, consisting of 50 % FBS, 40 % RPMI, and 10 % dex- trose (one tube/vial). 3. Centrifuge for 10 min at 200 × g. 4. Resuspend the cell pellet in 1 mL of ice-cold PBS (pH 7.2). 5. Centrifuge again. Remove as much supernatant as possible using pipette, and use the pellet to perform the comet assay (see Sect. 4.3 , step 5).

3.5 Sample Quality Nonspecific damage may originate from cell processing or cell death (either apoptosis or necrosis) may originate from nonspecific DNA damage, identified as an assay artifact [22 ]. Thus, it is common to analyze cell viability before performing the assay, and trypan blue exclusion test is the most frequently chosen assay. Nevertheless, this assay does not measure cell death but cell integrity, and therefore its usage is considered as inadequate by many authors. In biomonitoring studies assessing damage in peripheral blood cells, a poor result in viability is more likely to be related with poor handling of cells [12 ] than possible cytotoxic effects of a certain exposure. In contrast, if the study is using terminally differentiated cell populations with a high renewal rate, such as buccal cells, the presence of apoptotic and necrotic cells is to be expected [ 22 ]. Nevertheless, it is yet not clear whether increased DNA fragmentation due to cell death can result in the generation of false-positive results in the comet assay [43 ]. Despite all this, most authors still assess cytotoxicity prior to comet assay; other than dye viability assays, histopathology (in tissues) and a low molecular weight (LMW) test [43 ] can be used to assess evidence of necrosis or apoptosis. LMW DNA diffusion assay can be easily conducted concurrently with the comet assay by preparing an extra comet slide that will be ﬁ xed just after the lysis step [ 44 ]. 228 Carla Costa and João Paulo Teixeira

After staining, 100 cells per slide can be scored visually for the percentage of diffused and condensed cells. The percentage of diffused cells is considered representative of cytotoxicity in the sampled cell population [ 45 ]. With this result, it is possible to detect DNA fragmentation that occurs in the earliest stages of cell death and that can confound the determination of genotoxicity using the comet assay [ 44 ]. If increased levels of DNA migration are present in samples of exposed individuals, the possibility that this increase is due to apoptosis or necrosis need to be properly evaluated [22 ].

4 Comet Assay Detailed Protocol

4.1 Critical Factors One of the reasons for the widespread use of comet assay in biomonitoring studies is related to the simplicity of the method. Nevertheless, throughout the years, independent studies [ 46 , 47 ] have shown that there are key steps of the protocol that may introduce high variability in the results and therefore must be properly controlled during experiments; furthermore, experimental conditions must be stated in scientifi c publications for critical evaluation of the results. These key steps include (1) agarose concentration (low melting point (LMP) agarose), (2) alkali unwinding period, (3) electrophoretic conditions, and (4) comet scoring (this last point will be addressed later on in a dedicated section of this chapter). In addition, temperature during alkaline treatment and electrophoresis cannot be disregarded; fi ndings of Speit et al. [48 ] show that increased temperature during these steps strongly enhances DNA migration, and therefore this parameter must be strictly controlled throughout any study in order to obtain reproducible results. Regarding LMP agarose concentration, it has been observed that higher concentrations may impair DNA migration. On the other hand, if the concentration is too low, gels will be too fragile for handling. Azqueta et al. [ 46 ] recommend a fi nal LMP agarose between 0.6 and 0.8 % (w/v). Alkali unwinding period that has also been identifi ed as a critical step [46 , 47 ] of the procedure is highly variable among laboratories hampering result comparison. The commonly used period of 20 min is considered acceptable for most purposes [49 ], but higher periods are recommended. Azqueta et al. [46 ] indicate a period of 40 min for optimal alkali-labile expression while Ersson and Möller [47 ] suggests protocol adjustment in accordance to the type of damage in study. Electrophoresis and, in particular, the voltage applied are critically important [50 ]. Describing the voltage and current used during the assay is not suffi cient to illustrate assay conditions as the critical The Comet Assay In Vivo in Humans 229

parameter is in fact voltage gradient expressed in V/cm [12 ]. Electrophoresis for 20 min at 1.15 V/cm, or 30 min at 0.83 V/cm offers acceptable results [46 ]. One of the steps of the assay that also varies signiﬁ cantly from one laboratory to another is the period of lysis (from 1 h to several days). Nonetheless, it has been shown that this is not a critical step of the protocol and that if no problems of cell adherence are encountered in longer periods of lysis, these will have no evident effect on the results.

4.2 Issues Biomonitoring studies usually involve a large number of samples on Quality Control that are analyzed in different days and that may introduce experimental variability in the results even if the assay, and particularly the key steps identified above, is performed under controlled conditions. One possible way to integrate all the sources of variation and compare measurements obtained in different work sessions is the incorporation of an internal standard into the assay [ 50 ]. Different publications use a control slide in each work session; the slide contains cells with a known level of damage that are either treated with a genotoxic compound (positive control) or untreated cells (negative control). Boeck et al. [ 50 ] states that standardization of measurements against a negative standard is most useful as this shows to be more stable across experiments and operators. Still, the inclusion of a positive control is suggested to confirm DNA migration in each experiment. In alternative to result normalization, acceptance criteria may be established according to laboratory’s historical data, and if DNA migration of control cells does not fall within reference values, data from that work session is rejected [51 ]. It is important to note that these types of standards do not account for inter-gel variability as they are parallel standards analyzed in separate slides. Zainol et al. [ 52 ] suggested the use of an internal standard in which reference cells were analyzed in the same gel that test cells. Reference cells had their DNA substituted by BrdU, and by using a fl uorescent anti-BrdU antibody and an additional barrier fi lter, it was possible to distinguish these cells from the remaining test cells. The use of this type of internal standard provided a decrease of twofold in the coeffi cient of variation for intra- and inter-experimental measures of DNA damage. Below, one can fi nd a detailed protocol to perform the alkaline version of the comet assay. It is essential to guarantee blinded analysis (processing and scoring) of coded samples to avoid measurement bias. This is a standard protocol that may be adjusted in some critical steps (Sect. 4.1 ) to improve assay sensitivity.

4.3 Slide Preparation 1. Prepare a solution of 1 % standard agarose (normal melting

point) in deionized H2 O in a vertical jar. Note : In order to melt the agarose, use the microwave or another heating method. 230 Carla Costa and João Paulo Teixeira

2. Dip the slides in a vertical jar of melted agarose, drain off the excess, and wipe the back of the slide. 3. Place the slide to dry horizontally. Note: Place them at room temperature overnight or in an oven at 50 °C for quicker drying. 4. Prepare a solution of 0.6 % LMP agarose in PBS pH 7.2 (see above); place it in a 37 °C water bath. Note : Allow low melting point agarose to reach 37 ° C before performing the next step. 5. Mix the cell pellet with 70 μL of LMP agarose and prepare duplicate drops in the same slide (using the previously prepared precoated slides). 6. Place 18 mm × 18 mm coverslips over the drops and place on icepacks until the agarose layer is solidiﬁ ed (this takes approximately 5 min). 7. After gel solidiﬁ cation remove the coverslips carefully.

4.4 Lysis 1. Immerse the slides in cold, freshly prepared lysing solution

(NaCl 2.5 M, Na 2 EDTA 100 mM, Tris–base 10 mM, NaOH 10 M, pH 10; just before use, add Triton X-100 to 1 %) (use Coplin jars). 2. Store at 4 °C for at least 1 h. 3. Remove slides from the lysing solution. All the following steps must be performed under dim yellow light to prevent additional DNA damage and at 4 °C.

4.5 Alkaline 1. Place the slides in a horizontal gel electrophoresis tank. Treatment Note: The tank must be cooled throughout the entire period of alkaline treatment and electrophoresis.

2. Add cold freshly made electrophoresis buffer (1 mM Na2 EDTA and 0.3 M NaOH) until the slides are completely covered without formation of air bubbles over the agarose gel and leave for 40 min to allow DNA unwinding and alkali-labile site expression.

4.6 Electrophoresis 1. Run the electrophoresis at 1.15 V/cm for 20 min; adjust the and Neutralization current by raising or lowering the buffer level. Note: To estimate V / cm , you may use the spreadsheet available at http ://comics.vitamib.com / electrophoresis -physics / . 2. Turn off the power and remove the slides from the buffer. 3. Immerse the slides with PBS pH 7.2 and keep it at 4 °C for 10 min. Note : Neutralization is classically performed using Tris buffer, but PBS usage is just as good and much cheaper. 4. Remove PBS and add cold deionized water; keep it at 4 °C for 10 min. The Comet Assay In Vivo in Humans 231

4.7 Fixation 1. Remove the water from the jar and ﬁ ll it with 70 % ethanol; keep it at room temperature for 15 min. 2. Remove the 70 % ethanol from the jar and ﬁ ll it with 96 % ethanol; keep it at room temperature for 15 min. 3. Remove slides from the jar and place the slides to dry horizontally (at room temperature overnight).

4.8 Staining Good staining of the comets is essential for accurate scoring; various fl uorochromes can be used to stain DNA, e.g., ethidium bromide, DAPI (4,6-diamino-2-phenylindole), SYBR® green/ gold, acridine orange, YOYO dye, and propidium iodide. To choose the best fl uorochrome, one has to consider possible fading during image capture, DNA content of analyzed cells, and background staining [51 ]. Below, you can fi nd two different staining options.

4.8.1 Option 1: 1. Place the slides in the staining container, such as a petri dish, Ethidium Bromide the lid of a pipette-tip box, or a polypropylene container. 2. To prepare the staining solution dilute 200 μL of ethidium

bromide (commercially available at 10 mg/mL in H2 O) in 99.8 mL of deionized water (ﬁ nal concentration 20 μg/mL). 3. Rehydrate the slides with ice-cold deionized water for 30 min. 4. Add enough staining solution to completely cover the slides; protect the staining solution from light by covering it with aluminum foil or by placing it in the dark. Leave it for 20 min. 5. At the end of this period, wash the slides with ice-cold deionized water for 20 min. 6. Place the slides to dry horizontally. Note: These slides can be kept for months if necessary. 7. Before reading/scoring, rehydrate by adding a few drops of deionized water and lay a coverslip on the slide.

4.8.2 Option 2: 1. Place the slides in the staining container, such as a petri dish, SyberGold the lid of a pipette-tip box, or a polypropylene container. 2. To prepare the staining solution dilute 20 μL of SyberGold (previously diluted in DMSO to 1,000×) in 20 mL of TE buffer (10 mM Tris–HCl and 1 mM EDTA). 3. Add enough staining solution to completely cover the slides; protect the staining solution from light by covering it with aluminum foil or by placing it in the dark. 4. Place the staining container to agitate gently at room temperature for 30 min. 5. At the end of this period, quickly wash the slides with deionized water at room temperature. 232 Carla Costa and João Paulo Teixeira

6. Place the slides to dry horizontally (see above). 7. Before reading/scoring, rehydrate by adding a few drops of deionized water and lay a coverslip on the slide.

4.9 Comet For visualization, use a fl uorescent microscope equipped with an Visualization adequate fi lter combination (for the stain you are using) and a and Scoring magnifi cation of approximately ×500 (magnifi cation may vary according to cell type). Comets must be selected without bias and should represent the whole gel. Comets seen in edges, air bubbles, and overlaps must be rejected. In total, at least 50 cells should be analyzed by replicate of sample (reaching a total of 100 cells per individual). The optimum number of cells to score has not been defi ned, but in case of small differences in the level of DNA damage, by increasing the number of cells scored, the coeffi cients of variance in each group will decrease and the sensitivity of the assay increase [53 ]. The scorer must not know which slides he/she is analyzing (blind score); if analysis is performed by two or more technicians, a comparison between both must be carried out in order to guarantee reproducibility. Technician training and experience must not be overlooked as comet scoring is a potential source of variability. The most common approaches for comet scoring are visual inspection and computer-based analysis; recent studies have demonstrated the validity of both methods [ 54 ]. The key point is that the scoring method must be consistent over the study [55 ]. Visual scoring can be performed either by using an eyepiece micrometer to measure tail length or by using a classifi cation scheme for DNA damage (Collins [ 9 ] suggests a scheme ranging from 0, no damage, to 4, corresponding to almost all the DNA in the tail). The great advantage of computer-based analysis is that it collects data on different endpoints simultaneously (further information on endpoints will be given in the next section). Most laboratories use semiautomatic systems that require cell selection making comet scoring by software analysis as time-consuming as comet scoring by visual analysis. At this time, fully automated systems that rely only on software are already available allowing the increase of number of cells scored without human effort and consequent increase in assay sensitivity. Nevertheless, this constitutes a signifi cant investment that may not always be worthy specially as evidence shows a good agreement between visual scoring, semiautomated, and fully automated systems [54 ]. One the main drawback of software analysis is their inability to evaluate cells that are usually designated as clouds or hedgehogs . These are characterized by a small or absent head and a highly diffused tail; software are often not able to assess this type of cells due to the great DNA dispersion in the tail. Researchers are advised to record their number alongside with image analysis of The Comet Assay In Vivo in Humans 233

the remaining cells. Although the percentage of hedgehogs is usually excluded from formal statistical analysis (how this data can be integrated with other comet measurements is still an unsolved issue), this data can be important to assess the quality of the study [55 ].

4.10 Endpoints Among the different parameters captured by image analysis software, the most frequently used are tail length (μm; this can also be evaluated by visual scoring as mentioned above), tail moment (there are several measures of tail moment; the most commonly described olive tail moment is the amount of DNA in the tail multiplied by the mean distance of migration in the tail) that aims to summarize comet shape and size, and tail intensity (% of DNA in the tail). Tail length, which was widely used in the ﬁ rst years of the comet assay, is now not recommended for biomonitoring as it is highly dependent of protocol variations and quickly reaches a plateau level being therefore useless in cases of high levels of damage [56 ]. Considering tail moment, its proper measurement requires the geometrical calibration of the image analysis software (i.e., number of pixels per micrometer) and should be expressed in μm/%. If publications do not report units, then interlaboratory comparison is impossible. In addition, tail moment is an indirect measurement that depends on tail length and thus the limitations mentioned above for that parameter also affect this one. Several advantages have been recognized to the use of percentage of DNA in the tail (% tDNA); (1) it immediately gives an idea of the shape of the comet, (2) uses a quantitative measure of damage (from 0 to 100 %), and (3) is less variable across studies [ 9 , 23 ]. In addition, this parameter is linearly related to DNA break frequency over most of the assay range [12 ]. One possible problem arising from % tDNA is related to the presence of zero values that may complicate statistical analysis [55 ].

5 Data Interpretation

Properly designed and executed studies enable researchers to access the existence (and strength) or absence of an association between the agent and the outcome. Nevertheless, the association may or may not be causal [ 57 ]. To estimate accurately the association between any agent and a health effect, we need reliable and valid measurements of exposure, covariates (potential confounders and effect modiﬁ ers), and outcomes. In addition, for result interpretation it is essential to consider how much of the observed association between the studied factor and the outcome may have been affected by bias, confounding, and chance [58 ]. Result interpretation is also dependent of the type of cell analyzed. Lymphocytes are relatively long-lived cells, and DNA 234 Carla Costa and João Paulo Teixeira

damage observed in these cells must be differently interpreted from DNA damage observed in whole blood (contains 60–75 % of short-lived cells neutrophils). In addition to cell life-span, researchers must also consider differences in cell sensitivity; Wojewódzka et al. [ 59 ] reported that T and B lymphocytes showed different sensitivity after exposure to ionizing radiation. In addition, sample collection timing in regard to exposure can also greatly infl uence results and must be adequately addressed in result interpretation (some considerations were already made in Sect. 3 ). Comparison of obtained results with basal DNA breakage may be useful to detect possible issues related to storage conditions, scoring, or protocol. Møller [ 25 ] has established a reference value of 8.6 % for tDNA in control populations based in a pooled analysis of previously published studies. In human biomonitoring studies, various statistical approaches are commonly used to evaluate the effect of a specifi c condition on the comet assay results. Whenever measurements are obtained using specifi c software, a large amount of data, usually with varied distributions curves, is obtained for each endpoint and for each individual. The choice of summary statistics for the experimental unit (in biomonitoring studies, the individual) is crucial and depends on the endpoint chosen to report results; there are a number of measures that can be used to describe a sample, namely mean, median, 90th and 95th percentile value. Despite its sensitivity to small numbers of extreme values, and the fact that summary statistics can greatly alter the results, mean value is the most commonly used measure; nevertheless, Wiklund and Agurell [ 60 ] suggests mean of log transformed data and 90th percentile for tail moment and tail length, respectively. The choice of statistical analyses should be determined, in a large extent, by the experimental design of the study [55 ]. Statistical tests used to analyze comet assay results in published studies include parametric, nonparametric, multivariate, and univariate testing. Wiklund and Agurell [ 60 ] recommends the use of linear models, in particular ANOVA that permits the inclusion of additional relevant factors. Nevertheless, ANOVA make assumptions about randomization which are unlikely to hold for observational studies. Univariate testing may be appropriate in some cases, but in order to fully control confounding factors and to identify interactions, statistical modeling is necessary. It is important to refer that confounding control with statistics is usually associated to a loss of power or precision of the study. Effects must be preferably reported using quantitative measures of effect accompanied of confi dence intervals; these will complement the negative or positive fi nding derived from the hypothesis testing by estimating the precision of the effects found and are better suited to bridge the gap between statistical and biological signifi cance [60 ]. In this context, the criteria for a positive effect The Comet Assay In Vivo in Humans 235

should be based on the size of effect produced that would be biologically important [55 ]. Even in a well-designed and well-analyzed study, lack of association may only mean that either the sample size was not large enough to detect a weak association or the outcome has multiple causes [57 ].

6 Result Communication

Researchers typically use the peer-reviewed scientifi c literature as the principal approach to communicate study results [ 61 ]. On the other hand, researchers are increasingly called upon to communicate with study subjects when designing, interpreting, and reporting their work [ 62 ]. Result communication must be seriously considered and planned to avoid result misinterpretation. As already pointed for result interpretation, in result communication the researcher must be able to clearly distinguish between statistically signifi cant effects and biologically signifi cant effects. In the case of the comet assay, special caution must be taken as this issue remains to be clarifi ed even within the scientifi c community. It is also essential to deliver to participants (preferably during recruitment) that results will be population-based and not valid at the individual level. Overall, in result communication there must be a great concern on the subject’s ability to understand study information, including the associated uncertainties and limitations; pamphlets, posters, graphic narratives, and videos may prove to be useful to reach a wider audience [61 ].

7 Future Prospects

The amount of samples processed in a single experiment is an important issue that has been recently addressed. Improving throughput has been attempted by combining multiple gels in one substrate, and/or by reducing the volume of agarose gel per sample [ 56 ]. Some examples are the 12 gelBond fi lm, 96 gelBond fi lm, the 96-well multichamber plate, and the 12 minigel system [ 63 ]. This last example constitutes a medium throughput method that permits the incubation of gels under different conditions (lesion- specifi c endonucleases, DNA probes, etc.). Although the implementation of these systems may present additional value to the comet assay, their validation is still critical alongside with the need of reliable automated scoring systems to actually make the entire assay less labor intensive. Validation of a biomarker assay should be seen as crucially important to their use in biomonitoring studies [ 64 ]. At this time, 236 Carla Costa and João Paulo Teixeira

comet assay has not yet been established as a reliable biomonitoring tool for human studies and the biological relevance of the assay is not completely defi ned. A network project has been recently launched, ComNet, that aims to investigate the reliability and validity of comet assay as a biomarker for human biomonitoring studies [65 ]. One of the main shortcomings of this assay as a tool for biomonitoring studies is the lack of a standardized methodology [ 13 ] that would signifi cantly reduce or minimize experimental variation. This concept is related to reliability of the assay and in simple words means that the assay is reproducible in different laboratories (the same concept can also be applied in a single laboratory but to multiple operators) and repeatable at different times. Regarding this matter, different trials have been performed to address this issue, as the ESCODD project [66 ] and ECVAG [67 ]. Both found considerable interlaboratory variability, even after implementation of a standard protocol [68 ], and thus much effort is still necessary to overcome this question. Validity, on the other hand is related with the ability of the assay to predict a certain outcome [69 ]. In the case of comet assay, researchers aim to associate the obtained endpoints with specifi c types of DNA damage after a specifi c exposure. Meta-analyses or pooled analyses constitute tools that can be used to address the validity of biomarkers [ 58 ] although they can be affected by publication bias. The link between DNA damage and risk of disease in healthy individuals may be a future step, but this can only be effi ciently evaluated by prospective cohort studies as previously accomplished for other biomarkers of early genomic damage [23 , 29 , 70 ]. In conclusion, comet assay is valuable tool in biomonitoring studies, but researchers must be aware of all assay limitations (both experimental and biological). This is essential to improve both data quality and data relevance obtained with this assay and to guarantee that comet assay continues to provide enhanced reliability as a biomarker in human biomonitoring studies.

References

1. McKelvey-Martin V, Green M, Schomezer P Comet assay workgroup. Mutat Res 627: et al (1993) The single cell gel electrophoresis 31–35 assay (comet assay): a European review. Mutat 4. Collins A, Duthie S, Dobson V (1993) Direct Res 288:47–63 enzymic detection of endogenous oxidative 2. Collins AR, Dobson VL, Dusinska M et al base damage in human lymphocyte (1997) The Comet assay: what can it really tell DNA. Carcinogenesis 14:1733–1735 us? Mutat Res 375:183–193 5. Gedik C, Ewen S, Collins A (1992) Single-cell 3. Burlinson B, Tice RR, Speit G et al (2007) gel electrophoresis applied to the analysis of Fourth International Workgroup on UV-C damage and its repair in human cells. Genotoxicity testing: results of the in vivo Int J Radiat Biol 62:313–320 The Comet Assay In Vivo in Humans 237

6. Pfuhler S, Wolf HU (1996) Detection of DNA- 22. Albertini RJ, Anderson D, Douglas GR et al crosslinking agents with the alkaline comet (2000) IPCS guidelines for the monitoring of assay. Environ Mol Mutagen 27(1):196–201 genotoxic effects of carcinogens in humans. 7. Santos J, Singh N, Natajaran A (1997) Mutat Res 463:111–172 Fluorescence in situ hybridization with com- 23. Collins A, Koppen G, Valdiglesias V et al ets. Exp Cell Res 232(2):407–411 (2014) The comet assay as a tool for human 8. Wentzel JF, Gouws C, Huysamen C et al biomonitoring studies: the ComNet project. (2010) Assessing the DNA methylation status Mutat Res 759:27–39 of single cells with the comet assay. Anal 24. Møller P, Knudsen LE, Loft S et al (2000) The Biochem 400:190–194 comet assay as a rapid test in biomonitoring 9. Collins AR (2004) The comet assay for DNA occupational exposure to DNA-damaging agents damage and repair. Principles, applications, and effect of counfounding factors. Cancer and limitations. Mol Biotechnol 26:249–261 Epidemiol Biomarkers Prev 9:1005–1015 10. Links JM, Kensler TW, Groopman JD (1995) 25. Møller P (2006) Assessment of reference val- Biomarkers and mechanistic approaches in ues for DNA damage detected by the comet environmental epidemiology. Annu Rev Public assay in human blood cell DNA. Mutat Res Health 16:83–103 612:84–104 11. IEH (1996) In: The use of biomarkers in envi- 26. Bonita R (2006) Basic epidemiology. World ronmental exposure assessment, Leicester, UK Health Organization, Geneva. http://nla.gov. 12. Collins AR, Oscoz AA, Brunborg G et al au/nla.cat-vn4201434 (2008) The comet assay: topical issues. 27. Pearce N, Checkoway H, Kriebel D (2007) Mutagenesis 23(3):143–151 Bias in occupational epidemiology studies. 13. Faust F, Kassie F, Knasmüller S et al (2004) Occup Environ Med 64:562–568 The use of the alkaline comet assay with lym- 28. Angerer J, Ewers U, Wilhelm M (2007) phocytes in human biomonitoring studies. Human biomonitoring: state of the art. Int J Mutat Res 566:209–229 Hyg Environ Health 210(3–4):201–228 14. Speit G, Hanelt S, Helbig R et al (1996) 29. Hagmar L, Bonassi S, Strömberg U et al Detection of DNA effects in human cells with (1998) Chromosomal aberrations in lympho- the comet assay and their relevance for muta- cytes predict human cancer: a report from the genesis. Toxicol Lett 88:91–98 European Study Group on Cytogenetic 15. Hartmann A, Plappert U, Poetter F et al Biomarkers and Health (ESCH). Cancer Res (2003) Comparative study with the alkaline 58:4117–4121 Comet assay and the chromosome aberration 30. Van Delft JHM, Baan RA, Roza L (1998) test. Mutat Res 536(1–2):27–38 Biological effect markers for exposure to carci- 16. Wong VW, Szeto Y, Collins AR et al (2005) nogenic compound and their relevance for risk The Comet assay: a biomonitoring tool for assessment. Crit Rev Toxicol 28(5):477–510 nutraceutical research. Curr Top Nutraceut 31. Chuang C-H, Hu M-L (2004) Use of whole Res 3(1):1–14 blood directly for single-cell gel electrophore- 17. Astley S, Langrish-Smith A, Southon S et al sis (comet) assay in vivo and white blood cells (1999) Vitamin E supplementation and oxida- for in vitro assay. Mutat Res 564:75–82 tive damage to DNA and plasma LDL in type 32. Eren K, Özmeriç N, Sardas S (2002) 1 diabetes. Diabetes Care 22(10):1626–1631 Monitoring of buccal epithelial cells by alkaline 18. Uriol E, Sierra M, Comendador MA et al comet assay (single cell gel electrophoresis (2013) Long-term biomonitoring of breast technique) in cytogenetic evaluation of cancer patients under adjuvant chemotherapy: chlorhexidine. Clin Oral Investig 6:150–154 the comet assay as a possible predictive factor. 33. Rojas E, Valverde M, Lopez MC et al (2000) Mutagenesis 28(1):39–48 Evaluation of DNA damage in exfoliated tear 19. Schabath MB, Spitz MR, Grossman HB et al duct epithelial cells from individuals exposed (2003) Genetic instability in bladder cancer to air pollution assessed by single cell gel elec- assessed by the comet assay. J Natl Cancer Inst trophoresis assay. Mutat Res 468:11–17 95(7):540–547 34. Morris I, Ilott S, Dixon L et al (2002) The 20. Valverde M, Rojas E (2009) Environmental spectrum of DNA damage in human sperm and occupational biomonitoring using the assessed by single cell gel electrophoresis Comet assay. Mutat Res 681:93–109 (Comet assay) and its relationship to fertiliza- 21. Silva I (1999) Cancer epidemiology: principles tion and embryo development. Hum Reprod and methods. IARC, Lyon 17(4):990–998 238 Carla Costa and João Paulo Teixeira

35. WHO (2010) WHO guidelines on drawing and electrophoresis on results obtained with the blood: best practices in phlebotomy. Agency comet assay. Toxicol Lett 110:73–78 for Healthcare Research and Quality (AHRQ). 49. Tice RR, Agurell E, Anderson D et al (2000) http://www.guideline.gov/content.aspx?id= Single cell gel/comet assay: guidelines for 37621 . Accessed 20 Nov 2013 in vitro and in vivo genetic toxicology testing. 36. Hininger I, Chollat-Namy A, Sauvaigo S et al Environ Mol Mutagen 35:206–221 (2004) Assessment of DNA damage by comet 50. Boeck MD, Touil N, Visscher GD et al (2000) assay on frozen total blood: method and evalu- Validation and implementation of an internal ation in smokers and non-smokers. Mutat Res standard in comet assay analysis. Mutat Res 558:75–80 469:181–197 37. Duthie S, Pirie L, Jenkinson A et al (2002) 51. Kumaravel T, Vilhar B, Faux SP et al (2009) Cryopreserved versus freshly isolated lympho- Comet assay measurements: a perspective. Cell cytes in human biomonitoring: endogenous and Biol Toxicol 25:53–64 induced DNA damage, antioxidant status and 52. Zainol M, Stoute J, Almeida GM et al (2009) repair capability. Mutagenesis 17(3):211–214 Introducing a true internal standard for the 38. Holland NT, Smith MT, Eskenazi B et al Comet assay to minimize intra- and inter- (2003) Biological sample collection and pro- experiment viability in measures of DNA dam- cessing for molecular epidemiological studies. age and repair. Nucleic Acids Res 37(22):e150 Mutat Res 543:217–234 53. Sharma AK, Soussaline F, Sallette J et al (2012) 39. Narayanan S, Donovan M, Duthie S (2001) The infl uence of the number of cells scored on Lysis of whole blood in vitro causes DNA the sensitivity in the comet assay. Mutat Res strand breaks in human lymphocytes. 749:70–75 Mutagenesis 16(6):455–459 54. Azqueta A, Meier S, Priestley C et al (2011) 40. Anderson D, Yu T-W, Dobrzynska MM et al The infl uence of scoring method on variability (1997) Effects in the Comet assay of storage in results obtained with the comet assay. conditions on human blood. Teratog Carcinog Mutagenesis 26(3):393–399 Mutagen 17:115–125 55. Lovell DP, Omori T (2008) Statistical issues in 41. Villavicencio D (2006) Evaluation of storage the use of the Comet assay. Mutagenesis 23(3): conditions for assessing DNA damage using 171–182 the comet assay. Indiana University 56. Azqueta A, Collins AR (2013) The essential 42. Al-Salmani K, Abbas HHK, Schulpen S et al comet assay: a comprehensive guide to mea- (2011) Simplifi ed method for the collection, suring DNA damage and repair. Arch Toxicol storage, and comet assay analysis of DNA dam- 87(6):949–968 age in whole blood. Free Radic Biol Med 51: 57. Bailey LA, Gordi L, Green M et al (1996) 719–725 Reference guide on epidemiology. Jurimetrics 43. Hartmann A, Agurell E, Beevers C et al (2003) 36(2):159–168 Recommendations for conducting the in vivo 58. Vineis P, Gallo V (2007) The epidemiology alkaline Comet assay. Mutagenesis 18(1):45–51 theory: principles of biomarker validation. In: 44. Vasquez MZ (2010) Combining the in vivo Vineis P, Gallo V (eds) Epidemiology concepts comet and micronucleus assays: a practical of validation of biomarkers for the identifi ca- approach to genotoxicity testing and data tion/quantifi cation of environmental carcino- interpretation. Mutagenesis 25(2):187–199 genic exposures. ECNIS, London 45. Vasquez M (2007) Cytotoxicity and its impact 59. Wojewódzka M, Machaj EK, Goździk A et al on the in vivo comet assay. Environ Mol (2008) DNA damage in subpopulations of Mutagen 48:624 human lymphocytes irradiated with doses in 46. Azqueta A, Gutzkow KB, Brunborg G et al the range of 0–1 Gy of X-radiation. Nukleonika (2011) Towards a more reliable comet assay: 53(4):145–149 optimising agarose concentration, unwinding 60. Wiklund SJ, Agurell E (2003) Aspects of time and electrophoresis conditions. Mutat design and statistical analysis in the Comet Res 724:41–45 assay. Mutagenesis 18(2):167–175 47. Ersson C, Möller L (2011) The effects on 61. Bates MN, Hamilton JW, LaKind JS et al DNA migration of altering parameters in the (2005) Workgroup report: biomonitoring comet assay protocol such as agarose density, study design, interpretation, and communica- electrophoresis conditions and durations of tion—lessons learned and path forward. the enzyme or the alkaline treatments. Environ Health Perspect 113(11):1615–1621 Mutagenesis 26(6):689–695 62. Sandman PM (1991) Emerging communica- 48. Speit G, Trenz K, Schütz P et al (1999) The tion, responsibilities of epidemiologists. J Clin infl uence of temperature during alkaline treatment Epidemiol 44(Suppl I):41S–50S The Comet Assay In Vivo in Humans 239

63. Azqueta A, Gutzkow KB, Priestley CC et al 67. Møller P, Möller L, Godschalk RW et al (2010) (2013) A comparative performance test of Assessment and reduction of comet assay varia- standard, medium- and high-throughput tion in relation to DNA damage: studies from comet assays. Toxicol In Vitro 2:768–773 the European Comet Assay Validation Group. 64. Dusinska M, Collins AR (2008) The comet assay Mutagenesis 25(2):109–111 in human biomonitoring: gene–environment 68. Forchhammer L, Ersson C, Loft S et al (2012) interactions. Mutagenesis 23(3):191–205 Inter-laboratory variation in DNA damage 65. Collins A, Anderson D, Coskun E et al (2012) using a standard comet assay protocol. Launch of the ComNet (comet network) project Mutagenesis 27(6):665–672 on the comet assay in human population studies 69. Bennett DA, Waters MD (2000) Applying during the International Comet Assay Workshop biomarker research. Environ Health Perspect meeting in Kuasadasi, Turkey (September 108(9):907–910 13–16, 2011). Mutagenesis 27(4):385–386 70. Bonassi S, Znaor A, Ceppi M et al (2007) An 66. Gedik CM, Collins A (2005) Establishing the increased micronucleus frequency in periph- background level of base oxidation in human eral blood lymphocytes predicts the risk of lymphocyte DNA: results of an interlaboratory cancer in humans. Carcinogenesis 28(3): validation study. FASEB J 19(1):82–84 625–631 Chapter 14

Analysis of Nanoparticle-Induced DNA Damage by the Comet Assay

Julia Catalán , Satu Suhonen , Anna Huk , and Maria Dusinska

Abstract

The rapid and enormous development of nanotechnology has been accompanied by a deep concern for the effects that nanoparticles may have on human health and the environment. In this context, it is essential to assess the ability of nanoparticles to cause DNA damage. Single cell gel electrophoresis assay (comet assay) is widely used for evaluating nanoparticle-induced DNA damage in cells and is the most used assay for genotoxicity testing of nanomaterial. Here, we describe the standard alkaline version of the comet assay, both in vitro and in vivo, as well as the lesion-speciﬁ c enzyme-modiﬁ ed comet assay (for detection of oxidized DNA lesions) to test nanoparticles. We also highlight critical points that need to be taken into consideration when assessing nanomaterial genotoxicity.

Key words Nanoparticles , Comet assay , Genotoxicity , DNA damage , DNA oxidation

1 Introduction

Nanotechnology has led to a huge development of new products with unique properties, allowing new commercial applications. However, the widespread industrial use of nanoparticles (NP) has raised concern about their potential health effects. Thus, evaluation of the genotoxic effects of nanoparticles is essential. Compared to known genotoxic compounds, NPs are unique because of their behavior and physicochemical characteristics, which are completely different from those of ﬁ ne-sized particles of similar chemical composition [1 ]. NP-induced genotoxicity may primarily result from direct interaction of particles with the genomic DNA or indirectly through the enhanced production of reactive oxygen species (ROS) by cellular constituents in response to their interaction with the particles [ 2 , 3 ]. Both pathways may relate to surface properties, the presence of transition metals, intracellular iron mobili- zation, or lipid peroxidation processes. Other aspects relevant to primary genotoxicity are particle size, shape, particle uptake, and the presence of mutagens carried with the particles [4 ].

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_14, © Springer Science+Business Media New York 2014 241 242 Julia Catalán et al.

Secondary genotoxicity is also an oxidative stress-driven response, but in this case mediated by infl ammatory cells [2 , 3 ]. In fact, oxidative stress has often been described as a key mechanism underlying the ability of NP to cause DNA damage [5 ]. The DNA lesions caused by NP can still be repaired by cellular DNA repair systems. If the lesions are misrepaired or if unrepaired lesions cause replication errors, gene mutations and chromosomal mutations can be formed [ 6 ]. One of the most common DNA damage assays is the single cell gel electrophoresis assay, commonly known as the comet assay. Depending on the pH used, the comet assay is able to detect a wide variety of DNA damage such as single- and double-strand breaks, incomplete excision repair sites, cross-links (by decreased comet tail), alkali-labile sites (e.g., abasic sites) and, by using lesion- specifi c enzymes, oxidized DNA lesions [3 , 7 , 8 ]. Additionally, photogenotoxic effects of NP can be measured in combination with ultraviolet radiation [ 9 ]. A more detailed description of this method has already been provided in previous chapters of this book, and its applications in studies of nanomaterial genotoxicity can be found in recent reviews [ 3 , 5 , 10 ]. The assay has the advantage that it can be carried out in both proliferating as well as nonproliferating cells and single cells from different tissues can be evaluated in large numbers [ 11 ]. Medium- and high-throughput versions are already in use to allow analysis of large numbers of NP in a relatively short time [12 ]. Scoring can be semiautomated using image analysis software, and improvements toward full automation are already in progress. In addition, due to its simple and inexpensive setup, it can be used in conditions where more complex assays are not available [4 ]. Virtually any cell type with a nucleus can be used, and thus the comet assay can assess both cell- and tissue-specifi c DNA damage induced by NP. An advantage as well as a main disadvantage is its extreme sensitivity that allows detection of weak genotoxic agents. Therefore, to ensure the reproducibility of results, reference standards must be included, and quality control exercised at crucial steps. Additionally, there has been concern that tested nanomaterials could interfere with the comet assay, giving rise to false-positive/false-negative results. All crucial aspects for genotoxicity testing of nanomaterials using the comet are discussed below.

2 Methodological Considerations

Variability in the results of nanogenotoxicity studies can be attributed to the source and type of NP, the method of preparation or synthesis, stabilizers used, surface properties, the dispersion protocol, and variables in experimental conditions such as physicochemical speci- ﬁ cations (pH, temperature, presence of impurities, or sonication), Comet Assay as a Useful Tool for Testing Nanoparticles 243

treatment regime, cell type used, exposure time, dose, possible interference with the detection system or the method itself, and other factors [13 –15 ]. Some of the most critical points applying to the testing of NP with the comet assay are discussed below.

2.1 Dispersion Primary properties of nanomaterials change depending on the sur- of Nanomaterial rounding environment. The critical aspect of nanomaterial testing, and Secondary especially if applied in liquid form, is the vehicle, as secondary Characterization properties of NP and their genotoxic effects can be infl uenced by the solvents used for dispersing or dissolving them. Solvents differ in their physical and chemical properties. Factors such as pH, salin- ity, water hardness, temperature, and the presence of dissolved or natural organic particles could infl uence the biological response or toxicity. NP thus may behave differently in water, culture medium, PBS, and other solvents [ 16 ]. This could have pronounced effects on their uptake, cellular localization, and hence the observed toxic response. NPs tend to precipitate, agglomerate and aggregate and they can become surrounded by biomolecules such as proteins, forming a corona—all of which could affect their genotoxic potential. It is important to note that the treatment medium should mimic real in vivo conditions as much as possible (composition and proportion of proteins and other components should be similar to real exposure) as the effect depends largely on secondary NP properties. Also note that NP should be characterized before the treatment, and secondary properties such as state of agglomeration, surface properties, and stability of the dispersion stock solution as well as of the NP dispersed in the fi nal treatment medium should

be measured. Our studies with TiO2 NP showed that the same NP gave a positive genotoxic response when dispersed following one protocol but not when dispersed by another procedure [17 ]. As the stability of dispersion is in most cases limited, the dispersion solution always needs to be freshly prepared, i.e., immediately before the experiment. Most common dispersions use bovine serum albumin (BSA) or fetal bovine/calf serum (FBS, FCS), as the presence of proteins prevents agglomeration. Nevertheless, one must be aware that the protein corona formed around NP affects their toxicological properties. Sonication of the dispersion also protects against agglomeration and is widely used. However, severe sonication can affect the properties of nanomaterials, e.g., by breaking down carbon nanotubes.

2.2 In Vitro The choice of biological models (in vivo/in vitro or human) and and In Vivo Models cell type (human/mammalian, primary cells or cell lines) depends on the type of NP, the aim of the study, route of exposure, and target organ and tissue. For in vivo studies, both mice and rats, and cells from different organs, most commonly from the lungs, 244 Julia Catalán et al.

blood, or liver, have been used. The in vitro comet assay for nanomaterial testing utilizes primary cells as well as numerous cell lines of human or mammalian origin most commonly derived from skin, blood, brain, lung, gastrointestinal tract, kidney, or liver [14 , 15 ]. Primary blood cells from human volunteers were also used to study DNA-damaging effects of several NP [18 ].

2.3 Exposure Time The exposure time is crucial for both in vitro and in vivo studies. and Concentrations For testing chemicals in vitro, 3–6 h exposure time is usually recommended. However, NP may need more time to enter the cells. Thus, to study genotoxicity of NP, both short and 24 h exposure should be used [8 ]. Time of exposure is important also in vivo to detect effects of NP. For poorly soluble NPs, at least 24 h treatment is recommended for both mice and rats to detect the effect and produce a signiﬁ cant increase of DNA damage in target organs and tissue. Partly soluble NP could exert their effects at shorter times. In addition, particle deposition could give rise to a continuous insult to DNA (either by direct ROS production or indirectly by elicita- tion of an inﬂ ammatory response) that could be even detected after several months [6 ]. For both in vitro as well as in vivo tests, concentrations used should be realistic, i.e., relevant to possible exposures. Recommended concentrations for in vitro studies should correspond to at least 60 % viability, since—especially when detecting DNA strand breaks by the standard alkaline comet assay—DNA breaks can be secondary effects of cytotoxicity and so could give false-positive results. Thus for in vitro studies cytotoxicity should always be tested with the same cells (and ideally in the same experiment). The most reliable assays are relative growth activity and colony-forming ability, since MTT, WST1, and other colorimetric assays are prone to interference by the NP [19 ]. NPs have a tendency to agglomerate, and it is suggested that the concentration/dose of NP should not exceed the level at which agglomeration is enhanced. The agglomeration of NP affects their bioavailability to the cell and thus might lead to false-positive/ false-negative results. The concentration of NP is commonly expressed as μg/mL or μg/cm2 . The relationship between both metric units varies depending on the type of culture plates that are used. Other metrics include number of particles per mL or cm 2 as well as surface area per mL or cm 2 . It is advised to use at least two metric units to express dose.

2.4 Positive For reproducibility and quality assurance, positive and negative and Negative Controls controls should always be used in each experimental setup to and Reference ensure the correct performance of the comet assay. Ideally, a so- Standards called reference standard should be used to indicate whether the Comet Assay as a Useful Tool for Testing Nanoparticles 245

level of damage as detected by the comet assay is constant with time—or, if there is variation over time, to permit normalization of data. Thus a reference standard helps to minimize variation. This could be any type of cell (either untreated or treated with genotoxic substance), in aliquots frozen slowly, in freezing medium with serum and DMSO, and stored at −80 °C [ 20 ]. Negative controls consist of dispersion solutions that do not contain NP but, otherwise, have been processed in a similar way as NP dispersions (e.g., same sonication schedule). Several positive controls can be used depending on the aim of the study, the material being tested, and whether a metabolic activation system is present. Virtually any DNA strand break-inducing agent could be used. The most common is hydrogen peroxide

(H2 O2 ). The treatment of cells can differ. In our laboratories we use 5–10 min treatment with 20–100 μM H2 O2 . As DNA breaks are quickly repaired, optionally the cells can be treated in the gel to minimize time of processing of cells before lysis. For metal NP and some carbon nanotubes, it is advised to use metal ions as an additional control as metal ions can be released from NP and cause production of reactive oxygen species via Fenton-like reactions, and so it is important to test whether their presence does not induce genotoxicity, rather than the NP themselves. Coating materials or NP stabilizers can also cause genotoxicity and thus should also be tested and included in the experimental setup as additional reference material. It is also important to note that NPs are good carriers, and if a stabilizer or coating is toxic, even lower, normally nontoxic concentrations can cause damage due to their enhanced internalization into cells. The challenge for nanogenotoxicity studies is choosing a nano- specifi c positive control. There are several initiatives currently focusing on selection of nanomaterials with appropriate properties to be recommended as reference standards [21 , 22 ]. ZnO NPs were suggested as a positive control for the comet assay at the EU NanoGenotox project [ 23 ]. The results were positive but not reproducible, being particularly affected by the type of cell used. Unfortunately, there is still a lack of certifi ed reference standards to be used as positive controls for genotoxicity. Our recommended positive control for in vivo experiments with mice is 1 mg/mouse of tungsten carbide–cobalt mixture (WC–Co), applied by pharyngeal aspiration 24 h before sacrifi cing the animals. This treatment provides a signifi cant increase of DNA damage in both BAL and lung cells.

2.5 Possible Properties of NP such as adsorption capacity, optical properties, Interference hydrophobicity, chemical composition, surface charge and surface of the Comet Assay properties, catalytic activities as well as agglomeration can result in with Nanomaterial interference with standard toxicity tests [ 19 , 24 ]. The interference of NP was demonstrated with a range of in vitro cell viability assays 246 Julia Catalán et al.

(MTT, LDH, WST-1, Annexin V/PI, neutral red, caspase propidium iodide, 3 H-thymidine incorporation, and automated cell counting), infl ammatory responses (ELISA for GM-CSF, IL-6, and IL-8), and oxidative stress detection (monoBromoBimane, dichlorofl uorescein, NO assays) [19 , 24 ]. Interferences found were assay- as well as NP-specifi c. It is clear that for nanotoxicity testing, most of the assays need to be adapted and modifi ed to avoid measuring artifacts. Additional standards need to be included as controls for the interference. For genotoxicity, interference was reported so far with the comet assay as well as the micronucleus test. The protocol for the micronucleus assay needed modifi cation as cytochalasin B (used in this assay) inhibits endocytosis and may prevent uptake of NP [25 ]. The comet assay has been criticized because it often gives a positive result with nanomaterials, which resulted in speculation about false-positive results [ 5 , 26 ]. Landsiedel et al. [27 ] compared the comet assay results (14 of 19 studies positive) with the micronucleus assay (12 of 14 positive) which corresponds to 73.6 % of positive with the comet assay and 85 % positive with micronucleus assay, and Rajapakse et al. [ 26 ] surprisingly regarded this comparison as indicating inconsistency between two assays (rather than a broad agreement) and speculated that the higher frequency of positives with the comet assay could be due to measuring artifacts. Their study relies on an experimental setting which includes exposure of acellular systems (i.e., gel-embedded nucle-

oids from lysed cells) to TiO2 NP. As they measured a high level of DNA breaks with NP in the gel, they assume that the apparent damage seen on exposure of cells before embedding could be due to NP being carried over in the gel and present after lysis—thus

giving false-positive results. In our study, we also included TiO2 NP into gels and found no interference [ 25 ]. Interestingly, TiO2 NP genotoxicity depended on the dispersion protocol used [ 17 ] and not on the presence of NP in gel. The possibility of residual NP present in the gel interacting directly with non-nucleosomal DNA and potentially causing additional DNA damage, raised by Stone et al. [28 ], Karlsson [5 ], as well as Rajapakse et al. [26 ], should certainly be taken into consideration, and tests should be carried out with each NP before genotox-

icity testing. In our study, we tested fi ve NPs (TiO2 , uncoated and sodium oleate coated Fe3 O4 , and silica NP) present at high concentration with the lysed cells and found no increase in strand breaks. A modifi cation of the comet assay, involving digestion of DNA with lesion-specifi c endonucleases—such as formamidopyrimidine DNA glycosylase (FPG) which detects oxidized purines—raised a different concern about artifacts resulting from the presence of residual NP with the DNA of lysed cells. In this case, there is a risk of false-negative results if NP interferes with the FPG reaction. Comet Assay as a Useful Tool for Testing Nanoparticles 247

An interaction of NP with FPG has been reported by Kain et al. [29 ] for two tested NPs; they suggested that this interaction could lead to underestimation of DNA oxidation caused by NP. These authors mixed NP directly with FPG at high concentrations, which do not correspond to experimental conditions in the comet assay; even so, FPG was efﬁ cient after mixing with 3 out of the 5 NPs. The study of Magdolenova et al. [ 25 ] showed that incubation of nuclei (treated with a known 8-oxoguanine-inducing agent) with NP does not lead to any decrease in the yield of FPG-sensitive sites. No interference with repair enzyme FPG was found with all 5 NP tested when present in the gel. To conclude, both positive and negative results have been reported with the comet assay so far [3 ]. The assay is sensitive to detect DNA damage induced by NP, but it does seem that there is a possibility of artifactual damage occurring if NP remains in the gel. Thus, thorough rinsing and testing for possible interference of NP in the gel, using both untreated cells and cells exposed to a known genotoxic compound (causing DNA strand breaks as well as oxidized DNA lesions), as suggested by Magdolenova et al. [25 ], would seem to be a sensible precaution to be sure that no overestimation or underestimation of damage is occurring.

3 Materials

3.1 Speciﬁ c ● Dimethyl sulfoxide (DMSO) for in Vivo Approach ● Hank’s Balanced Salt Solution (HBSS) ● Beakers (100 mL) ● Cell strainer (40 μm, Nylon, ~27 mm Ø) ● Centrifuge tubes (ﬂ ip top, 15 mL) ● Petri dishes (60 mm Ø)

3.2 Speciﬁ c ● dPBS (Ca2+ , Mg2+ free) for In Vitro Approach ● Fetal bovine serum (FBS) ● Penicillin/streptomycin ● Culture media RPMI 1640, DMEM, BEGM ● Trypsin 0.05 %—EDTA 1× ● Cell culture ﬂ asks ● Multidishes (Nunclon surface, sterile, 6 or 24 wells)

3.3 Common ● Bottle top ﬁ lter units 0.2 μM PES (for sterilization of reagent for Both Approaches solutions) ● Bovine serum albumin (BSA) 248 Julia Catalán et al.

● Disodium EDTA (Titriplex III)

● dH2 O (distilled/deionized) ● Ethanol 100 % ● Ethidium bromide ● Hydrochloric acid (HCl concentrated >10 M)

● Hydrogen peroxide (H2 O 2 ) 30 % ● Low melting point agarose (LMPA) ● Normal melting point agarose (NMPA) ● PBS (Ca2+ , Mg2+ free) ● Sodium chloride (NaCl) ● Sodium hydroxide (NaOH) ● Triton X-100 ● Trizma base ● Centrifuge ● Centrifuge tubes (50 mL) ● Centrifuge tubes (conical, 14 mL) ● Centrifuge tubes (conical, sterile, 11 mL)

● CO2 incubator +37 °C, 5 % CO, 80 % humidity ● Coplin jars ● Coverslips (no. 1, 24 mm × 40 mm) ● Culture tubes (round bottom, sterile, 14 mL) (for dispersion of NP) ● Digital microscope or inverted phase contrast microscope ● Fluorescence microscope ● Frozen ice packs ● Glass tubes (sterile, round bottom, 10, 30 mL) (for dispersion of NPs stock solutions) ● Hemacytometer ● Horizontal electrophoresis tank(s) ● Ice ● Komet 5,5 (imaging program, Andor Bioimaging) or Komet VI (Perceptive), Metafer ● Micropipettor and tips ● Microscope slide tray (aluminum/plastic/cardboard) ● Microscope slides, conventional ● Microwave oven ● Permanent marker for labeling slides ● Pincers Comet Assay as a Useful Tool for Testing Nanoparticles 249

● Power unit ● Sonicator ● Syringe ﬁ lter 0.2 μm ● Waterbath +37 °C

4 Methods

Here we describe the standard alkaline version of the comet assay for detection of strand breaks and alkali-labile sites (adapted to NP testing from the protocol of Tice and Vasquez [30 ] and Jonas Nygren’s advice) and the modiﬁ ed version of the comet assay for detection of strand breaks, alkali-labile sites, and oxidized purines and pyrimidines (NILU standard operating procedure, adapted to NP testing from Collins [31 ]; Dusinska and Collins [7 ]; Harris et al. [12 ]).

4.1 In Vitro HBSS (Hank’s Balanced Salt Solution) containing 20 mM and In Vivo Standard EDTA. Alkaline Comet Assay Add 3.72 g EDTA to 400 mL of 1× HBSS (Ca2+ , Mg2+ free) and for Detection adjust pH to 7.0–7.5 with 10 M NaOH or concentrated HCl. of Strand Breaks Adjust volume to 500 mL with 1× HBSS (Ca2+ , Mg2+ free). 4.1.1 Preparation Store at room temperature. Sterilize by ﬁ ltering. of Reagents Mincing solution (only for in vivo approach).

For Processing of Samples

For Embedding of Cells 0.7 % LMPA LMPA 0.7 g PBS 100 mL

Microwave or heat until near boiling and the agarose dissolves. Aliquot 25 mL samples into 50 mL centrifuge tubes and refrigerate until needed. When needed, brieﬂ y melt agarose in microwave or by another appropriate method. Place LMPA vial in a 37 ºC water bath to cool and stabilize the temperature. Preparation of Comet Slides 1.5 % NMPA (1.5 g NMPA in 100 mL PBS). Microwave or heat until the agarose dissolves. While NMA agarose is hot, dip conventional slides up to one-half of their frosted area and gently remove. Wipe underside of the slide to remove agarose and place it on a tray on a ﬂ at surface to dry. 250 Julia Catalán et al.

The slides may be air-dried or warmed for quicker drying. Store the slides at room temperature until needed; avoid high-humidity conditions. Prepare slides at least 1 week before use. Lysing solution

NaCl 141 g EDTA 37.2 g Trizma base 1.2 g

Add dH2 O to ~700 mL Begin stirring Add NaOH ~8 g

Allow the mixture to dissolve. Adjust the pH to 10.0 using concentrated HCl or NaOH.

Add dH2 O to 890 mL. Sterilize by ﬁ ltering. Store at room temperature. Final Lysing Solution for In Vitro Approach (1,000 mL)

Add 100 mL of sterile dH2 O and 1 % Triton X-100; then refrigerate for at least 30 min prior to slide addition. Final Lysing Solution for In Vivo Approach (1,000 mL) Add 10 % DMSO and 1 % Triton X-100; then refrigerate for at least 30 min prior to slide addition. Note: The purpose of the DMSO in the lysing solution is to scavenge radicals generated by the iron released from hemoglobin when blood or animal tissues are used. It is not needed for other situations or where the slides will be kept in lysing for a brief time only. Unwinding Solution for In Vivo Approach .

For Electrophoresis 0.3 M NaOH 12 g and Analysis of Slides 1 M NaCl 58.4 g (to enhance DNA unwinding) 2 mM EDTA 0.744 g (to prevent activity of Ca/Mg-dependent nucleases)

add dH2 O to 1,000 mL

Sterilize by ﬁ ltering. Store at room temperature. Comet Assay as a Useful Tool for Testing Nanoparticles 251

Electrophoresis Buffer for In Vitro Approach (300 mM NaOH/ 1 mM EDTA)

Stock solutions

10 M NaOH NaOH 200 g

dH2 O 500 mL 200 mM EDTA EDTA 14.89 g

dH 2 O 150 mL Adjust pH to 10.0 with 10 M NaOH or concentrated HCl

Store both at room temperature. Note : Prepare the NaOH and EDTA stock solutions every ~ 2 week . 1× buffer—rpepare from stock solutions Prepare fresh buffer before each electrophoresis run.

10 M NaOH 30 mL 200 mM EDTA 5 mL

Add dH2 O to 1,000 mL

Mix well. Measure the pH prior use to ensure that it is >13. Note: The total volume to be prepared depends on the gel box capacity. Electrophoresis Buffer for In Vivo Approach (0.03 M NaOH, 2 mM EDTA)

Stock solutions ( 0.6 M NaOH / 40 mM EDTA ) NaOH 12 g EDTA 7.44 g

dH 2 O to 500 mL

Final electrophoresis buffer (1×)

Add 50 mL of stock solution to 950 mL of deionized H2 O. Note : In vivo electrophoresis buffer is still alkaline, but pH is lower than in vitro (~10 instead of > 13 ) Neutralization buffer

Trizma base 48.5 g

Add dH2 O to ~800 mL 252 Julia Catalán et al.

Adjust to pH 7.5 with concentrated (>10 M) HCl.

Add dH2 O to 1,000 mL. Store at room temperature. Staining Solution

Add 40 μl of ethidium bromide to 12.5 mL dH2 O, ﬁ lter sterilize, and store at room temperature. Caution: Handle dyes appropriately .

4.1.2 Dispersion of NP Note : The below protocols are applied to test NP in human bronchial epithelial cells (in vitro procedure ) and in mice treated by respiratory exposure (in vivo procedure ).

● Add 10 mg of NPs to 5–10 mL of sterile water in 10 mL sterile glass tube to prepare a stock solution of 1–2 mg/mL. Note : For in vitro approach, NP can be also dispersed in exposure medium ( BEGM + BSA 0.6 mg / mL ). For in vivo approach, NP can be also dispersed in PBS ( alone or with 0.6 mg /mL BSA ).

● Sonicate the stock solution for 20 min. at 37 °C using a 37 kHz Elmasonic Ultrasound Cleaner. Note : Bath or probe sonicators can both be used but be aware that the applied energy will be different depending on the model. Severe sonication can affect nanomaterials properties , e.g., breaking down carbon nanotubes .

● Make serial dispersions of the NP with water, exposure medium (BEGM + BSA 0.6 mg/mL), or PBS (alone or with 0.6 mg/ mL BSA), to cover a range of concentrations. We usually test 5–250 μg/mL range concentration for cell treatment in vitro (sterile, round bottomed, 14 mL, culture tubes) and 20 μg/ mL to 4 mg/mL range concentration for treating mice in vivo. ● Sonicate NP dispersions for 20 min at 37 °C using a 37 kHz Elmasonic Ultrasound Cleaner. ● Vortex dispersions efﬁ ciently and immediately applied to the cells or animals. Note : Dispersions of NP should be freshly prepared on the day of treatment . Note : For in vivo experiments , NP needs to be dispersed in a liquid when applying by intratracheal instillation or pharyngeal aspiration. However , dry NP can be also used by aerosolizing . Note : In addition to the above dispersion protocol , we have also used the protocol developed at the EU -project NanoGenotox ( http://www.nanogenotox.eu/files/PDF/web%20nanogeno- tox%20dispersion%20protocol.pdf ). Comet Assay as a Useful Tool for Testing Nanoparticles 253

4.1.3 Treatment ● Subculture 1.2 × 105 BEAS 2B cells/well in 24-well plates in 1 mL of BEGM medium (BEBM + supplements) and incubate In Vitro Approach: Exposure them at 37 °C for 24 h. and Harvest of Cells ● Remove media from the cells by aspiration. ● Replace medium with desired concentration of NP (0.5 mL of dispersion/well) and incubate at 37 °C for 0.5–24 h (according to the chosen exposure times). Note: Positive and negative control cultures should be included in each experimental set .

● Remove the medium and brieﬂ y rinse the cells with 1 mL of 1× PBS. ● Add 0.25 mL/well 0.05 % Trypsin–EDTA solution and incubate cells at 37 °C for 6–20 min to detach the cells. ● Add 0.5 mL/well RPMI 1640 medium containing 10 % FBS and 1 % penicillin–streptomycin to stop trypsin effect. ● Transfer cell suspension to a cold centrifuge tube (conical, 14 mL centrifuge tube), take an aliquot, and count the cells by using a hemocytometer. ● Centrifuge at 210 × g for 5 min. ● Remove supernatant. Keep the cell pellet cold on ice until embedding on slides.

In Vivo Approach: Exposure Human exposure to nanoparticles occurs mostly through the and Processing of Samples respiratory route, although there are also other exposure pathways (oral, dermal or—in nanomedicine applications—intravenous). There are several procedures to mimic experimentally respiratory exposure: inhalation (whole body or nose-only exposure), intranasal instillation, intratracheal instillation, or pharyngeal aspiration. All these procedures can be performed with different exposure frequencies (single vs. multiple exposures), exposure times, and sampling times after exposure. A short-term inhalation approach for testing of poorly soluble nanomaterials has been recently proposed [ 32 ]. Whatever the chosen exposure procedure is, the processing of samples for the Comet assay is done in a similar way. Bronchoalveolar lavage

● Collect samples into sterile 15 mL centrifuge tubes with ﬂ ip top, take an aliquot, and count the cells by using a hemocytometer. ● Centrifuge the cells down (210 × g , 6 min, 4 °C). ● Remove supernatant. 254 Julia Catalán et al.

Lung Tissue

● Collect samples into Eppendorf tubes containing 1 mL of cold mincing solution and kept on ice. ● Cut the lung lobules to pieces. ● Add 1 mL of chilled mincing solution to a cell strainer (40 μm, ~27 mm Ø) fi tted to a small Petri dish on ice. ● Place the lung pieces into the cell strainer. ● Mash the tissue through with the piston (with rubber head) of a small syringe. ● Rinse the cell strainer with more chilled mincing solution whereas processing the sample (up to a fi nal volume of 6 mL). ● Collect the cell suspension into a 15 mL centrifuge tube (sterile, fl ip top). Note : Usually ~¼ of the suspension provides enough cells for the assay . Note : Divide one sample — make a positive control from the other

half ( 20 mM H 2 O 2 , 10 min , RT = > 900 μL PBS + 100 μL H 2 O 2 → 10 μL of dilution to 1 mL of cell suspension ).

● Wash the cells three times with cold PBS by centrifugation (210 × g , 6 min, 4 °C). ● Resuspend the pellet in 2 mL of cold PBS and count the cells by using a hemocytometer. Note: This procedure can also be used with other solid tissues , e.g., liver . ● Centrifuge the cells down and remove supernatant. Keep cell pellets on ice until embedding cells on comet slides.

4.1.4 Embedding of Cells ● Mix cells with warm (37 °C) 0.7 % LMPA (one sample at a time; 500,000 cells/1 mL LMPA for in vitro approach and ~200,000 cells/300 μL LMPA for in vivo approach) and immediately pipette 60 μL mixture on a Comet slide. Add coverslip and place the slide on a slide tray resting on ice packs (4 °C). Let the slide lay there until the agarose layer hardens (~5 to 10 min). ● Remove coverslip and slowly immerse slide in prechilled, freshly made lysing solution. Leave at 4 °C in the dark for a minimum of 1 h (usually overnight).

4.1.5 Electrophoresis ● After at least 1 h at 4 °C, gently remove slides from the lysing and Analysis of Slides solution. Comet Assay as a Useful Tool for Testing Nanoparticles 255

● For the in vitro approach, place slides in Coplin jars containing freshly made cold pH > 13 electrophoresis buffer and let them sit there for 20 min. Then, remove slides from Coplin jars and place them side by side on the horizontal gel box, near one end, sliding them as close together as possible. Fill buffer reservoirs with freshly prepared cold pH > 13 electrophoresis buffer until the liquid covers completely the slides (avoid bubbles over the agarose). ● For the in vivo approach, place slides in Coplin jars containing unwinding solution (room temperature) and let them sit there in dark for 45 min to allow for unwinding of the DNA and the expression of alkali-labile damage. Then, remove slides from Coplin jars and place them side by side on the horizontal gel box, near one end, sliding them as close together as possible. Fill buffer reservoirs with freshly prepared cold pH ~10 electrophoresis buffer until the liquid covers completely the slides (avoid bubbles over the agarose). ● Set power supply to ~0.74 V/cm for the in vitro approach and a bit gentler treatment (~0.67 V/cm) for in vivo approach and apply voltage for 15 min. Note: Perform all these steps under dimmed or yellow light to prevent damage from UV radiation .

● Gently put slides out from the buffer and place them on a drain tray. Dropwise coat slides with neutralization buffer and let sit for at least 5 min. Drain slides and repeat two more times. ● Drain excess neutralization buffer, rinse by dipping in distilled water, and air-dry slides. At this stage, samples may be stored at room temperature with desiccant prior to scoring. ● Place 40 μL of 1× EtBr onto each slide before placing coverslip to ensure even hydration and staining and cover with a fresh coverslip. Before viewing slides, drain excess liquid from the slide edges. Note: SYBR Green or SYBR Safe can be used instead of ethidium bromide. Ethidium bromide provides a more stable staining although it is very toxic. There are also commercially available kits .

● View slides under ﬂ uorescence microscope using image analysis program for measurements. ● After scoring, remove coverslip, rinse in 100 % alcohol to remove stain, let dry, and store at room temperature. 256 Julia Catalán et al.

4.2 In Vitro Alkaline Stock solution Comet Assay for Detection of Strand 10 M NaOH (200 g in 500 mL distilled H2 O), keep at room temperature Breaks and Oxidized DNA Lesions 0.2 M EDTA (37.2 g in 500 mL distilled H2 O), keep at 4–8 °C

4.2.1 Preparation 4 M Tris (242.2 g in 500 mL distilled H2 O), keep in incubator at 37 °C of Reagents, Positive Controls, and Enzymes Lysis solution

2.5 M NaCl (146 g/L distilled H2 O)

0.1 M EDTA (37.2 g/L distilled H2 O)

10 mM Tris (2.5 mL of 4 M stock/L distilled H2 O)

Prepare 1 L. Adjust to pH 10 with either solid NaOH or preferably concentrated (10 M) NaOH solution. (Add 35 mL of NaOH straight away to ensure that EDTA dissolves, and then add dropwise to pH 10.) Add 1 mL Triton X-100 per 100 mL immediately before use. Electrophoresis Solution

0.3 M NaOH (30 mL/L distilled H2 O from 10 M NaOH stock solution)

1 mM EDTA (5 mL/L distilled H2 O from 0.2 M EDTA stock solution) PBS Solution 8 g/L NaCl 0.2 g/L KCl

1.15 g/L Na2 HPO4

0.2 g/L KH2 PO4 Set pH to 7.2 with NaOH

Preparation of H2 O2 Dilution Solution A: 11 μL stock solution (30 % w/v, i.e., 30 g/100 mL; 9.82 M) in 1 mL of PBS solution = 100 mM Solution B: 10 μL solution A in 1 mL of PBS solution = 1 mM

The working solution with selected concentration of H2 O 2 is prepared from solution B, e.g., 50 μM H 2 O2 solution is prepared as follows: 50 μL solution B + 950 μL PBS. Ro 19-8022 (Photosensitizer) Obtained from Hoffmann La Roche (contact angela.perrin @ roche.com ) Comet Assay as a Useful Tool for Testing Nanoparticles 257

Dissolve in 70 % ethanol at 1 mM and store in small aliquots in microtubes at –20 ºC. Avoid excessive light during preparation and wrap tubes in Al foil. Working solution: 1–2 μM in PBS. Enzyme Reaction Buffer for Endonuclease III and FPG (Buffer F) 40 mM HEPES (9.53 g/L) 0.1 M KCl (7.45 g/L) 0.5 mM EDTA (0.18 g/L) 0.2 mg/mL BSA (0.02 g/100 mL) Adjust to pH 8.0 with KOH (can be made as 10× stock, adjusted to pH 8.0 and frozen at −20 °C) Slide Preparation We recommend the use of ordinary clear glass slides precoated with agarose:

● The slides for precoating should be grease-free; clean if necessary (soak the slides in alcohol for about 24 h and then wipe dry with a clean tissue).

● Dip slides in melted 0.5 % standard agarose in H2 O. ● Drain off excess agarose, wipe the back clean and dry by leaving on a ﬁ lter paper overnight. ● Mark the coated side with a pencil mark in one corner (e.g., top left). ● Dry, precoated slides can be stored indeﬁ nitely, packed in slide boxes.

Enzymes Endonuclease III (ENDO III) and FPG are isolated from bacteria containing overproducing plasmids. Because such a high proportion of protein is the enzyme, a crude extract is perfectly satisfactory; in our experience, there is no nonspeciﬁ c nuclease activity at the concentrations employed. The enzyme extracts are best obtained from a laboratory producing them. On receipt, the enzyme (which should have been refrigerated in transit) should be dispensed into small aliquots (say, 5 μL) and stored at −80 °C. This minimizes repeated freezing and thawing. The ﬁ nal dilution of the working solution will vary from batch to batch.

FPG

● Dispense the stock solution into 5 μL aliquots and refreeze at −80 °C. ● Take one of these aliquots and dilute 30× using the regular buffer F with the addition of 10 % glycerol (5 μL aliquot of 258 Julia Catalán et al.

enzyme +145 μL buffer F with BSA and 10 % glycerol). Dispense this into 10 μL (2 μL, 5 μL, etc.) aliquots in 1.5 mL micro centrifuge tubes (mark the volume on the top of tube and label the box in which enzymes are stored as “TO BE DILUTED 1:100”) and freeze at −80 °C. These aliquots are 30× dilutions of enzyme, and they need to be further diluted 100× for the ﬁ nal working concentration (which is 1:3,000). ● For use, follow description on box, DILUTE 1:100, i.e., dilute one of these 10 μL enzyme aliquots to 1,000 μL with buffer F with BSA (no glycerol) and keep on ice until you add it to the gels: Do not refreeze this working solution.

Endo III (Is More Stable)

● Dispense the stock solution into 5 μL aliquots and refreeze at −80 °C. ● Take one of these aliquots and dilute to 0.5 mL using the regular buffer F (no need to add glycerol). Dispense this into 10 μL aliquots (label as “100× diluted”) and freeze at −80 °C. ● For use normally, dilute one of these 10 μL aliquots to 300 μL with buffer (no glycerol) and keep on ice until you add it to the gels.

4.2.2 Procedure Cells are cultured in complete culture medium and incubated in culture dishes or fl asks in humidifi ed atmosphere at 37 °C, 5 % Cell Culture Conditions CO 2 . Both suspension cells as well as cells growing attached to the surface can be used for testing nanoparticles for genotoxicity. Ideally they should be within 4 weeks of establishing the culture from a frozen stock and subcultured or given fresh medium 1–2 days before use to ensure that they are in a proliferating state. Monolayer cells should be used when approximately 75 % confl uent.

Exposure Conditions Negative and positive controls are always used. The concentration and Treatment With NPs range should be established with regard to expected genototoxic- and Controls ity, solubility in the test system, and changes in pH or osmolarity. A negative control and at least three concentrations should be used. For relatively noncytotoxic nanomaterial, the maximum concentration is recommended to be 100 μg/mL. For dispersion of NPs, follow recommendations in Sect. 2 and use dispersion protocol recommended for your NP. Be sure that NPs are characterized in dispersion media. Cells are exposed to the test substance for a deﬁ ned period of time, most commonly for 24 h. It is recommended to perform at least one experiment and two repeats. Comet Assay as a Useful Tool for Testing Nanoparticles 259

Positive Controls As reference standard for the comet assay, these positive controls

should be used: H 2 O 2 (to detect strand breaks) and Ro 19 -8022 ( photosensitizer ) for detection of oxidized purines.

Treatment of Cells with H2 O2

● The cells are treated with 50 μM H2 O 2 in PBS (1 mL), for 5 min at 4 °C (on ice). ● After treating, spin the cells at 200 × g , 4 °C, 3 min, wash with 1 mL of cold PBS. Note : It is necessary to work quickly so as not to allow repair of the induced damage .

Alternative: Treatment of Cells with H2 O2 on Gel For 2 gels on slide:

● Treat the cells after embedding in gel on glass slide with 50 μM

H2 O2 PBS, for 5 min at 4 °C in staining jar. For 12 gels on slide (see Fig. 1 ): ● Mix 10,000 cells with 200 μL LMP agarose. ● Put 5–10 μL drops of mixture on the slide in parallel for each

sample (place drop at the end of slides—less volume of H 2 O2 will be needed). ● Put slides into fridge for 5 min.

● Put slides into jar with cold 50 μM H2 O2 , for 5 min at 4 °C (in fridge). ● Put slides into jar with cold PBS (in fridge) for 5 min. ● Take slides out of jar and place in fridge for 5 min. Note : It is necessary to work quickly so as not to allow repair of the induced damage . Treatment of Cells with RO 19- 8022 in Suspension

● Spin cells, pour off medium, wash cells with PBS, spin, and add 5 mL of cold PBS containing Ro 19-8022 (1 μM) on Petri dish. Note : Avoid excessive light during preparation .

Fig. 1 Scheme of slides for 12-gel and 2-gel format 260 Julia Catalán et al.

● Place Petri dish with cells (in suspension) and Ro on ice 30 cm from a 500 W halogen lamp and irradiate for 5 min. ● Spin, remove Ro solution, wash with PBS, spin and add medium with serum, and prepare slides for the comet assay. Alternative for RO Treatment and Later Use of Frozen Cells in Aliquots

● Cells exposed to RO, spin, remove medium, and resuspend in 1 mL media. ● Count cells and dilute them to ﬁ nal concentration 20,000 cells/mL in mixture medium and DMSO (9:1). ● Aliquot mixture to Eppendorf tubes and store in ultrafreezer. ● For experiments, thaw contents of one microcentrifuge tube, spin cells, remove supernatant, wash in 1 mL cold PBS, and spin again. ● Remove supernatant and resuspend pellet in 200 μL LMP agarose. ● Put two drops of mixture on precoated slides. ● Place slides into fridge for 5 min. Note: It is necessary to work quickly so as to avoid repair of the induced damage .

Measurement of Survival At the end of the exposure period, cells are washed, counted, and subcultured to determine survival rate 48 h after the treatment for determination of relative growth activity or seeded in small inocula (100–200 cells per dish) to establish colony-forming ability.

Embedding Cells After the exposure and washing, appropriate number of cells are in Agarose taken from the subcultures, placed in a microcentrifuge tube, and centrifuged (200 × g , 3 min, 4 °C) and used for comet assay. For 2 gels on Slide (see Fig. 1 ) Work quickly as the agarose sets quickly at room temperature!

● Tap tube (ﬂ ick with ﬁ nger) to disperse cells in the small volume of medium remaining. ● Quickly add 140 μL of 1 % LMP agarose in PBS at 37 °C and mix with pipette. ● Take 140 μL of mixture and transfer as two equal drops on each slide. 2 × 10 4 cells/70 μL LMP agarose gel. ● Cover each with a coverslip. Leave slides in fridge for 5 min. For 12 Gels on Slide (see Fig. 1 )

● Prepare 0.2 mL vials on a heat block setup to 37 °C in rows of 6. ● Put a glass slide on cold metal plate. ● Add 120 μL of 1 % LMP agarose in PBS to each vial. Comet Assay as a Useful Tool for Testing Nanoparticles 261

● Add 5 × 103 cells into the vial with agarose. ● Put two drops 5 μL of mixture on the slide in parallel for each sample. 200–300 cells/5 μL LMP agarose gel. ● Put the slides into fridge for 5 min. For 12 Gels on Slides for Higher Number of Samples/Medium-Throughput Method (see Figs. 1 and 3 )

● Cells are exposed to NPs in 96 well plate with ﬂ at bottom. After the exposure remove medium, wash cells with PBS, trypsinize, and resuspend in 50 μL medium. ● Count untreated cells (negative control). ● Take 10,000 cells (calculate from counting cells from negative control) from each well by using automatic multichannel pipette and transfer cells to 96 well plate with U-bottom. It is not necessary count cells for each well. ● Put plate with slides (see picture) into ice to avoid repair of induced damage. ● Put 200 μL LMP of agarose in to six wells by using multichannel pipettes, mix, take all mixture into the tips, wait when most bob- bles disappear, push small volume back into wells, and immediately place six small drops on precoated slides (smaller drops approximately 5 μL for FPG slides and bigger drop 10–15 μL for rest of slides—e.g., for DNA breaks). Repeat all procedure to place all cells from 96 well plate into precoated slides. ● Place all slides in fridge for 5 min.

Lysis Add 1 mL Triton X-100 to 100 mL of lysis solution (4 °C), mix at magnetic blender. Store at 4 °C until use.

● Remove coverslips from slides and place in this solution in a (horizontal) staining jar. ● Leave at 4 °C for at least 1 h up to 24 h.

Enzyme Treatment The enzyme incubation takes place after lysis. FPG or Endo III enzyme removes the damaged base leaving an apurinic/apyrimidinic (AP) site that is converted into a break by the AP lyase/endonuclease normally associated with the repair enzyme or by the alkaline conditions of the assay. For enzyme treatment, use a special plate either a commercially available chamber (see Fig. 2 ) or a plate similar as made in our lab (see Fig. 3 ). FPG, Endo III

● Thaw 300 mL FPG enzyme reaction buffer with bovine serum albumin (BSA) and put aside 1 mL for enzyme dilutions. ● Wash slides twice in FPG enzyme reaction buffer without BSA (stock stored in fridge) and once with complete FPG enzyme reaction buffer (4 °C) in staining jar, for 5 min each. 262 Julia Catalán et al.

Fig. 2 Slide chamber commercially available

● Meanwhile, prepare dilutions of enzyme. ● Remove slides from last wash, and dab off excess liquid with tissue. For 2 Gels Format (Fig. 1 )

● Place 50 μL of enzyme solution (or buffer alone, as control) onto gel and cover with a square cut from kitchenfoil, paraﬁ lm (or coverslip). ● Put slides into moist box (prevents desiccation) and incubate at 37 °C for 30 min—FPG (or ENDO III) enzyme. For the 12 Gels Format (with Commercial Chamber)

● For this format for incubation with enzyme, use commercial slide chamber (see Fig. 2 ). ● Place the slide on the chamber. ● Cover with silicon. ● Place 30 μL of enzyme solution (or buffer alone, as control) onto each gel. ● Cover with plastic cover, and then secure with screws. ● Put slides into moist box (prevents desiccation) and incubate at 37 °C for 30 min—FPG (or ENDO III) enzyme. Note : The chamber should be screwed in identical way from both sides to prevent move of enzyme solution from one part to another. This could cause higher level of DNA damage in gels from one row . For 12 Gels Format (Medium-Throughput Method) (see Figs. 1 and 3 ) Comet Assay as a Useful Tool for Testing Nanoparticles 263

Fig. 3 Metal plate holder for 12 slides and scheme of preparation of gels in 12-gel format for large number of samples using multichannel pipette. For incubation of gels with repair enzyme, kitchen foil (or paraﬁ lm) is used

● For this format for incubation with enzyme, we use special metal plate (slides holder) (see Fig. 3 ). ● Place 30 μL of enzyme solution (or buffer alone, as control) onto each gel and cover with square cuts from kitchen wrapping plastic foil or paraﬁ lm. ● Put slides into moist box (prevents desiccation) and incubate at 37 °C for 30 min—FPG enzyme (or ENDO III enzyme). ● To avoid losing gel drops after the incubation with enzymes, it is recommended to place slides to fridge for 5–10 min. 264 Julia Catalán et al.

● After incubation, slides should be placed in the fridge to avoid loss of gels. Carefully remove foil avoiding damage to gels. Note: Check that , during incubation , foil is well attached to surface and there is no bubble of air under the foil. This allows enzyme to reach all places in the gel .

Alkaline Treatment ● Electrophoresis solution should be cooled before use, e.g., (20 min) by pouring into the electrophoresis tank in the cold incubator at 4 °C (acceptable up to 12 °C) an hour or so before it is needed. ● Gently place slides (minus coverslips) on platform in tank, immersed in solution, forming complete rows (gaps ﬁ lled with blank slides). ● Make sure that tank is level and gels are just covered. Leave 20 min. ● It is recommended to place one slide at the end of the tank ( see picture) to avoid moving of slides when you ﬁ ll the tank with electrophoresis solution.

Electrophoresis (20 min) ● Run at 25 V (~1.25 V/cm) for 20 min. ● If there is too much electrolyte covering the slides, the current may be so high that it exceeds the maximum—so set this at a higher level than you expect to need. If necessary, i.e., if 25 V is not reached, remove some solution. Normally the current is around 300 mA, but this is not crucial.

Neutralization ● Wash the slides for 10 min with cold PBS in staining jar followed by 10 min in cold water. ● Dry (room temperature) for storage.

Staining Staining with DAPI For 2 gels: Place approximately 20 μL (10 μL per gel) of 1 μg/mL

DAPI solution in distilled H 2 O (stored at −20 °C) onto each slide and cover with a coverslip. For 12 gels: Use 50 μL and spread it evenly in small droplets to cover the whole slide and cover with preferably large coverslip. Staining with SYBR Gold For 2 gels: Place 20 μL (10 μL per gel) of SYBR Gold (0.1 μL/mL

in TE buffer—10 mM Tris–HCl, 1 mM Na2 EDTA, pH 7.5–8) onto each slide and cover with coverslip. For 12 gels: Use 50 μL and spread it evenly in small droplets to cover the whole slide and cover with preferably large coverslip. Comet Assay as a Useful Tool for Testing Nanoparticles 265

Scoring Slides are analyzed by ﬂ uorescent microscope using image analysis system Comet assay IV (Perceptive instruments) or Metasystems (Metafer) by scoring 50 comets per gel (2 gels per concentration). These software operate with a charge-coupled device camera mounted on the microscope, automatically analyzes individual comet images. The programs are designed to differentiate comet head from tail and to measure a variety of parameters including tail length, % of total ﬂ uorescence in head and tail, and “tail moment.” We use % DNA in tail as the most informative parameter. Alternatively comets could be evaluated by visual scoring (100 comets per gel). During one study, the same method of analysis should be used.

5 Statistical Analyses of the Data

The comet assay is a hierarchical or nested design with animals (in the in vivo design) and cultures (in the in vitro design) within doses, a number of slides from each animal or culture, and a number of cells measured per slide [11 ]. The statistical models underlying these designs are based upon a general linear model approach, such as analysis of variance (ANOVA). Currently, the widely preferred measurement for statistical analysis is the tail intensity or percent of DNA in tail (% tail DNA), since it can be “standardized” over studies while other measures of DNA (as tail length and tail moment), although consistent within a study, may not be comparable across studies [ 11 , 33 ]. At our laboratories, in vitro, we work with a minimum of 3 doses of nanomaterials plus the corresponding positive and negative controls. We use two replicates or cultures per dose and we score 50 cells per slide (having two slides per culture). Every in vitro experiment is repeated at least twice. In vivo, we perform only one experiment, with a minimum of 3 doses of nanomaterials plus corresponding controls, minimum of six mice (or rats) per dose, and two slides per animal, scoring 50 cells per slide. Measurements made on 50 cells per slide have appeared as a recommendation following extensive realistic simulations of both in vitro and in vivo studies that compared measurements on different number of cells per slide [33 ]. We analyze our data by ANOVA (factorial or hierarchical depending on the design), using the mean value of 50 cells (mean or median % tail DNA per slide). This test is completed by a posteriori multiple comparison test. In addition, linear regression analysis is also tested to ﬁ nd out a linear dose–response effect. As recommended by Bright et al. [33 ], the positive control group is 266 Julia Catalán et al.

omitted from the linear model (because the variability in the animal summary statistic is usually markedly smaller in this group than in the other groups) and compared separately versus the negative control group to ensure the reliability of the assay. Another recommendation from the same authors is that the focus should be on the conﬁ dence intervals for the treatment effects and not on the p-values because the former conveys the sizes of effects consistent with the study data. Where this range of effect sizes sits relative to what would be considered biologically important (usually an increase of about twofold to threefold from the negative control value) is the key to reach an appropriate conclusion.

6 Current Validation Status of the Comet Assay in Testing NP

The comet assay has been the most widely used assay in genotoxicity testing of nanomaterials, both in vitro and in vivo [ 3 , 10 ]. However, it is not included among internationally validated OECD genotoxicity test guidelines, nor was recommended by the OECD Working Party on Manufactured Nanomaterials [34 ] for assessing nanomaterials genotoxicity. However, scientifi cally the comet assay (both standard method for detection of DNA breaks as well as modifi ed version to detect oxidized DNA lesions) was validated within several European projects such as FP5 ESCCOD [35 ], FP6 ECVAG and for genotoxicity testing of NP in FP7 NanoTEST [ 8 , 36 ] and NanoGenotox. Because of its great potential for medium- and high-throughput analysis and automation [12 ], it is considered as one of the most promising tools for genotoxicity testing of nanomaterials. Recently, the European Chemical Agency, at its guidance for testing of nanomaterials [ 37 ], has listed the in vivo comet assay among applicable in vivo genotoxicity assays as a follow-up assay for positive in vitro fi ndings concerning clastogenicity or gene mutation induction.

Acknowledgments

We thank Dr. Jonas Nygren for providing useful advice when developing our in vivo approach. We thank Andrew R Collins, Zuzana Magdolenova, Lise M Fjellsbø, Elise Runden-Pran for adopting protocol and preparation of standard operating procedure. We also thank Mr. Leszek Huk for making plate for incubation slides and Mr. Michal Zagrodzki for helping with the graphic design. Comet Assay as a Useful Tool for Testing Nanoparticles 267

References

1. Knaapen AM, Borm PJA, Albrecht C et al 14. Dusinska M, Rundén-Pran E, Carreira SC et al (2004) Inhaled particles and lung cancer. Part (2012) In vitro and in vivo toxicity test meth- A: mechanisms. Int J Cancer 109:799–809 ods. Chapter 4. Critical evaluation of toxicity 2. Donaldson K, Poland CA, Schins RPF (2010) tests. In: Fadeel B, Pietroiusti A, Shvedova A Possible genotoxic mechanisms of nanoparti- (eds) Adverse effects of engineered nanomate- cles: criteria for improved test strategies. rials: exposure, toxicology and impact on Nanotoxicology 4:414–420 human health. Elsevier, New York 3. Magdolenova Z, Collins AR, Kumar A et al 15. Dusinska M, Magdolenova Z, Fjellsbø LM (2013) Mechanisms of genotoxicity. Review of (2013) Toxicological aspects for nanomaterial recent in vitro and in vivo studies with engine- in humans. In: Ogris M, Oupicky D (eds) ered nanoparticles. Nanotoxicology 8(3):233– Nanotechnology, methods in molecular biol- 278. doi: 10.3109/17435390.2013.773464 ogy. Humana, Totowa 4. Bowman L, Castranova V, Ding M (2012) 16. Handy RD, van den Brink N, Chappell M et al Single cell gel electrophoresis assay (comet (2012) Practical considerations for conducting assay) for evaluating nanoparticles-induced ecotoxicity test methods with manufactured DNA damage in cells. Methods Mol Biol 906: nanomaterials: what have we learnt so far? 415–422 Ecotoxicology 21:933–72 5. Karlsson HL (2010) The comet assay in nano- 17. Magdolenova Z, Bilaničová D, Pojana G et al toxicology research. Anal Bioanal Chem 398: (2012) Impact of agglomeration and different 651–666 dispersions of titanium dioxide nanoparticles 6. Alenius H, Catalán J, Lindberg H et al (2014) on the human related in vitro cytotoxicity and Nanomaterials and human health. In: Vogel U, genotoxicity. J Environ Monit 14:455–64 Savolainen K, Wu Q, van Tongeren M, 18. Tulinska J, Kazimirova A, Kuricova M, et al. Brouwer D, Berges M (eds) Handbook of (2014) Immunotoxicity and genotoxicity test- nanosafety. Measurement, exposure and toxi- ing of PLGA-PEO nanoparticles in human cology. Academic, San Diego blood cell model. Nanotoxicology (in press) 7. Dusinska M, Collins AR (1996) Detection of 19. Guadagnini R, Halamoda Kenzaoui B, et al. oxidised purines and UV-induced photoprod- (2014) Toxicity screenings of nanomaterials: ucts in DNA, by inclusion of lesion-specifi c challenges due to interference with assay enzymes in the comet assay (single cell gel processes and components of classic in vitro electrophoresis). Alternatives Lab Anim 24: tests. Nanotoxicology (in press) 405–411 20. Dusinska M, Collins A (2008) The Comet 8. Magdolenova Z, Drlickova M, Henjum K, assay in human biomonitoring: gene–environ- et al. (2014) Coating-dependent induction of ment interactions. Mutagenesis 23:191–205 cytotoxicity and genotoxicity of iron oxide 21. OECD (2008) OECD work on the safety of nanoparticles. Nanotoxicology (in press) manufactured nanomaterials. Environment, 9. Jha AN (2008) Ecotoxicological applications health and safety division. Environment Direc- and signifi cance of the comet assay. Mutagenesis torate. OECD. www.oecd.org/dataoecd/54/ 23:207–21 27/41567645.ppt 10. Doak SH, Manshian B, Jenkins GJS et al 22. Stone V, Nowack B, Baun A et al (2010) (2012) In vitro genotoxicity testing strategy Nanomaterials for environmental studies: clas- for nanomaterials and the adaptation of current sifi cation, reference material issues, and strate- OECD guidelines. Mutat Res 745:104–111 gies for physico-chemical characterization. Sci 11. Lovell DP, Omori T (2008) Statistical issues in Total Environ 408:1745–54 the use of the comet assay. Mutagenesis 23: 23. NanoGenotox project report. http://www. 171–182 nanogenotox.eu/files/PDF/nanogenotox_ 12. Harris G, Palosaari T, Magdolenova Z, et al. web.pdf (2014) Iron oxide nanoparticle toxicity testing 24. Kroll A, Pillukat MH, Hahn D et al (2012) using high throughput analysis and high content Interference of engineered nanoparticles imaging. Nanotoxicology (in press) with in vitro toxicity assays. Arch Toxicol 86: 13. Dusinska M, Fjellsbø LM, Magdolenova Z 1123–1136 et al (2011) Safety of nanomaterial in nano- 25. Magdolenova Z, Lorenzo Y, Collins A et al medicine. In: Hunter RJ, Preedy VR (eds) (2012) Can standard genotoxicity tests be Nanomedicine in health and disease. CRC, applied to nanoparticles? J Toxicol Environ Boca Raton Health A 75:800–806 268 Julia Catalán et al.

26. Rajapakse K, Drobne D, Kastelec D et al REV1. In: 10th meeting of the Working party on (2013) Experimental evidence of false-positive manufactured nanomaterials, OECD; 27–29 Comet test results due to TiO2 particle-assay June 2012; Paris, France interactions. Nanotoxicology 7:1043–1051 33. Bright J, Aylott M, Bate S et al (2011) 27. Landsiedel R, Kapp MD, Schulz M et al (2009) Recommendations on the statistical analysis of Genotoxicity investigations on nanomaterials: the Comet assay. Pharm Stat 10:485–493 methods, preparation and characterization of 34. OECD (2009). Preliminary review of OECD test material, potential artifacts and limita- test guidelines for their applicability to manufac- tions—many questions, some answers. Mutat tured nanomaterials. Environment Directorate, Res 681:241–458 Organisation for Economic Co-operation and 28. Stone V, Johnston H, Schins RPF (2009) Development, Paris. OECD Environment, Development of in vitro systems for nanotoxi- Health and Safety Publications Series on the cology: methodological considerations. Crit Safety of Manufactured Nanomaterials, No. 15 Rev Toxicol 39:613–626 35. Collins AR, Dušinská M, Mišľanová C, among 29. Kain J, Karlsson HL, Möller L (2012) DNA other ESCODD members (2002) Inter- damage induced by micro- and nanoparticle— laboratory validation of procedures for measuring interaction with FPG infl uences the detection 8-oxo-7,8-dihydroguanine/8-oxo-7,8-digydro- of DNA oxidation in the comet assay. 2′-deoxyguanosine in DNA. Free Radic Res Mutagenesis 27:491–500 36:239–245 30. Tice R, Vasquez M (1999) Protocol for the 36. Juillerat L, Dusinska M, Fjellesbo LM, et al. application of the pH > 13 alkaline single cell (2014) Biological impact assessment of nano- gel (SCG) assay to the detection of DNA dam- material used in nanomedicine. Introduction age in mammalian cells. http://cometassay. to the NanoTEST project. Nanotoxicology com/index_fi les/Page290.htm (in press) 31. Collins AR (2004) The comet assay for DNA 37. ECHA (2012) Guidance on information require- damage and repair: principles, applications, and ments and chemical safety assessment. Appendix limitations. Mol Biotechnol 26:249–261 R7-1 Recommendations for nanomaterials 32. OECD (2011) A steering group 7 case-study applicable to Chapter R7a Endpoint specifi c for hazard identifi cation of inhaled nanomateri- guidance. European Chemicals Agency. http:// als: an integrated approach with short-term inha- echa.europa.eu/documents/10162/13632/ lation studies. ENV/CHEM/NANO(2011) 6/ appendix_r7a_nanomaterials_en.pdf Chapter 15

The Comet Assay in Drosophila : Neuroblast and Hemocyte Cells

L. María Sierra , Erico R. Carmona , Leticia Aguado , and Ricard Marcos

Abstract

At present, the comet assay has been applied to Drosophila melanogaster for the in vivo analysis of genotoxicity, using three different larvae cells: hemocyte, midgut, and neuroblast cells. Due to the advantages of this higher eukaryotic organism, in terms of its similarities to mammals in DNA damage response, the comet assay in Drosophila has been successfully used in several studies to analyze the in vivo genotoxicity of chemicals, including chemotherapeutic drugs, environmental contaminants, and metals. The obtained results clearly conﬁ rm the usefulness of this combination (Drosophila and comet assay), and open its possibilities for a more widely use, selecting new cell targets and exposure scenarios. In this context, we present here detailed protocols to perform this assay using neuroblast and hemocyte cells.

Key words Drosophila , Comet assay , Third-instar larvae , Neuroblast cells , Hemocytes

1 Introduction

Since its design by Östling and Johanson [1 ], and its posterior modiﬁ cation by Singh et al. [2 ], the comet assay has become a very useful genotoxicity assay [ 3 – 6 ]. Its usefulness has been widely applied in different ﬁ elds such as the biomonitoring of exposed populations ([ 7 – 15 ]; see Chap. 13 of this book), the analysis of DNA repair ([16 – 20 ]; see Chaps. 21 and 22 of this book), and the analysis of DNA damage in a variety of organisms and cell types [ 21 – 24 ], as well as in studies of ecogenotoxicology [25 –28 ]. Its demonstrated utility in in vivo and in vitro mammalian model systems, as well as in other organism models, was determinant for its development in Drosophila melanogaster [ 29 ]. In this way, the advantages of a model organism like D. melanogaster , referred to homologies to mammals in key processes in the DNA damage response, such as the xenobiotic metabolism [30 , 31 ] and DNA repair activity [ 32 , 33 ], are in support of the usefulness of the comet assay applied to Drosophila , mainly as in vivo model.

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_15, © Springer Science+Business Media New York 2014 269 270 L. María Sierra et al.

It is important to remark that Drosophila was the pioneering organism used to detect the DNA-damaging effect of ionizing radiation [34 ] and chemicals [35 ]. Since the fi rst work by Bilbao et al., in which the comet assay was performed in neutral and alkaline conditions, in different defi - cient repair mutant strains, and using neuroblast cells from brain ganglia [ 29 ], several more works were published, using these same cells as well as other types of cells. In some of them, the comet assay was performed with midgut cells from wild-type strains, to check the genotoxicity of chemicals, metals, and contaminants [ 36 – 42 ], to detect oxidative DNA damage [43 ], to analyze nano- compounds using transgenic strains [ 44 ], and to study DNA repair [39 , 40 , 42 ]. The comet assay was also performed with Drosophila S2 cultured cells to check DNA damage induction [ 45 ] and to study DNA repair enzymes [ 46 ]. Spermatocytes were used to perform the comet assay in D. simulans to analyze oxidative DNA damage in Wolbachia -infected fl ies [ 47 ]. In other studies the comet assay was performed with hemocytes, analyzing metal genotoxicity [ 48 – 50 ], DNA damage induced by gold nanoparticles [51 ], genotoxicity of pesticides [ 52 ], and genotoxic effects of fl avor ingredients [ 53 ]. The comet assay in Drosophila , with neuroblast cells, was also used to analyze the relationship between chemically induced adducts and DNA strand breaks, considering the functional status of the nucleotide excision repair system [ 54 ]. In all these works, the comet assay in Drosophila has demonstrated its usefulness and its potential as an in vivo genotoxicity assay in a higher eukaryotic organism. In summary, the comet assay in Drosophila can be carried out in vivo with at least three different cell types: neuroblast, midgut, and hemocytes. It must be remembered that, in addition to this three well-established protocols, the comet assay can be carried out in any kind of cells, once a suitable protocol to get individualized cells must be obtained. In this chapter, we are presenting the protocols to perform the comet assay in Drosophila , using two of these cells: neuroblast cells from brain ganglia and hemocyte cells from hemolinfa [55 ].

2 Materials and Reagents

2.1 Materials ● All the reagents necessary to prepare the described solutions and buffers ● Coplin jars ● Coverslips 22 × 22 mm and 24 × 60 mm ● Electrophoresis power supply ● Eppendorf/microcentrifuge tubes 1.5 mL Drosophila Comet Assay 271

● Fine forceps ● Fluorescence microscope with CDD camera ● Freezer (−20 ºC) ● Horizontal electrophoresis tank ● Microscopic slides ● Microwave ● Refrigerator ● Small scalpel ● Tungsten wires ● Water bath

2.2 Media ● Baker yeast, 100 g for Growing Flies ● Sugar, 100 g and Maintaining ● Agar-agar, 9 g Strains ● Salt, small spoon 2.2.1 Standard Yeast–Sugar Medium: Mix everything, heat it, and keep it boiling for 30 min. Cool it For 1 L of Water down to 55 ºC, add 5 mL of propionic acid, mix well, and start to pour around 20 mL of medium in glass bottles of 250 mL.

2.2.2 Standard Corn To maintain the strains and obtain larvae, the following standard Flour Medium medium can be used:

● Corn ﬂ our, 170 g ● Yeast, 120 g ● NaCl, 2 g ● Agar-agar, 10 g Mix the components, add 1,200 mL of cold water, mix well the mix, and heat to boiling. Remove from heat, cool, and add to the mixture 4 mL of Nipagin fungicide, dissolved in 10 % ethanol and 4 mL of propionic acid. Once prepared, the culture medium is distributed in culture glass bottles of 125 mL. In each bottle, approximately 25 mL of media are placed and allowed to stand for a few hours before use. Place a small piece of paper soaked with the insecticide tetradifon into each bottle in order to control mite populations, while maintaining the humidity inside the culture bottles.

2.3 Medium Carolina Drosophila Medium Formula 4-24® (Carolina Biological for Treatments Supply Company, USA) for neuroblast cell analysis: 3 mL of medium (approx. 0.76 g) hydrated with 3 mL of distilled water (or solvent) or chemical solutions. 272 L. María Sierra et al.

2.4 Ringer Solution: ● NaCl 130 mM, 1.899 g For 250 mL ● KCl 35 mM, 0.652 g

● CaCl2 2 mM, 0.055 g Adjust the pH to 6.5 with NaOH, and sterilize by autoclaving. If not contaminated, it can last until 3 months, at 4 ºC.

2.5 Solution ● PBS, 9.3 mL to Collect Hemocytes ● Phenylthiourea (PTU), 7 mg As PTU pose some difﬁ culties to dissolve to reach the working concentration (0.07 %), long stirring is recommended.

2.6 Lysis Buffer: ● NaCl 2.5 M, 146.10 g

For 1 L ● Na2 EDTA 100 mM, 37.22 g ● Tris-(hydroxymethyl)-aminomethane 10 mM, 1.21 g ● NaOH 0.25 M, 10.00 g ● N -Lauroylsarcosine sodium salt 0.77 %, 23.10 mL ● (30 % aqueous solution) Mix and dissolve the salts, add the N -lauroylsarcosine sodium salt (keep it in darkness from here), adjust the pH to 10 with HCl, and sterilize by autoclaving. Maintain it in darkness at room temperature not longer than 1 month.

2.7 Lysis Solution: ● Lysis buffer 89 %, 178 mL For 200 mL ● Dimethyl sulfoxide (DMSO) 10 %, 20 mL ● Triton X-100 1 %, 2 mL Mix well and keep it stirring until use.

● 2.8 Denaturation Na2 EDTA 1 mM 20 mL from a 100 mM Na2 EDTA stock and Electrophoresis ● NaOH 300 mM 120 mL from a 5 M NaOH stock Buffer: For 2 L Adjust the pH to 12.6 with HCl. Prepare it fresh and keep it at

4 ºC. The Na2 EDTA stock is sterilized by autoclaving. Both stocks are maintained at room temperature for at least 1 month.

2.9 Neutralization ● Tris-(hydroxymethyl)-aminomethane 0.4 M, 48.46 g Buffer: For 1 L Adjust the pH to 7.5 with HCl and sterilize by autoclaving. Maintain it at room temperature for at least 1 month.

2.10 Staining ● Ethidium bromide 0.2 mg/mL 1.8 mg in 9 mL of distilled water Solutions Mix and keep it at 4 º C and darkness. 2.10.1 Ethidium Bromide (Stock Solution) Drosophila Comet Assay 273

2.10.2 DAPI ● DAPI (4′,6-diamidine-2-phenylindole), 10 mg ● Water, 50 mL To prepare the stock solution, add 10 mg DAPI in 50 mL of

d-H 2 O (200 μg/mL). Mix 1 mL of stock solution with 19 mL of d-H2 O (20 μg/mL). Finally, for 1X stock, mix 1 mL with 9 mL of d-H2 O.

3 Methods

3.1 Comet Assay ● Place 100 females and 60 males per bottle, with 20 mL stan- with Brain Ganglia dard yeast–sugar medium. This proportion is recommended Cells: Specifi c when wild-type strains or mutant strains with standard viability Procedure are used. For mutant strains with reduced viability, place 120 females and 80 males, to ensure enough number of larvae. 3.1.1 Larvae Collection ● Bottles are kept at 24 ± 1 ºC for 24 h. ● Remove the fl ies and place the bottles at 21 ± 1 ºC for fi ve additional days. In this way, after 144 ± 12 h of total development, we will be handling third-instar larvae of rather large size that will not move to the vial wall for at least another 12 h.

3.1.2 Treatment ● Prepare treatment vials: For each vial, add 3 mL of Carolina instant medium (around 0.76 g) plus 3 mL of the chemical solution to be used or distilled water. Mix by shaking the vial and let it stand. Prepare 2 vials for each concentration of the chemical analyzed, 2 vials for the negative control (distilled water or solvent buffer), and 2 vials for the positive control (we use 1 mM methyl methanesulfonate [MMS]). ● For the ﬁ rst concentration, take several larvae from the collecting bottles with a spoon. Place them in a Petri dish and wash thoroughly with tap water to remove standard medium debris. Place 10–15 larvae in each of the two treatment vials. ● Repeat this process for the rest of concentrations at intervals of at least 30 min, because this is the shortest necessary time to collect the brain ganglia corresponding to each analyzed concentration. ● Keep the larvae in these vials for 12 h at 24 ± 1 °C. Note: Be sure that the time interval for treatments is the time you need to collect all the brain ganglia. Otherwise, you would accumulate delay, and therefore, treatment would be longer than 12 h.

● Let the negative control as the last treatment. 274 L. María Sierra et al.

3.1.3 Isolation ● Cover part of one microscopic slide with paraﬁ lm, and place of Single Cells there one drop of 15 μL of Ringer solution. ● Add a new drop (15 μL) of Ringer solution in another slide without paraﬁ lm, and place it under the stereomicroscope with diascopic illumination, with the lowest possible light intensity. ● With the help of tweezers, take one larvae from the treatment medium and place it on the Ringer drop. Hold the larvae there with the tweezers and cut it in two parts with a scalpel as shown in Fig. 1a . ● Remove the posterior part of the larvae, and with the help of two tungsten wires, start looking for the brain ganglia in the anterior part (Fig. 1b ).

Fig. 1 Process of brain ganglia isolation. (a ) Anterior and posterior half of a larva after cutting it with a small scalpel. (b ) Identiﬁ cation of brain ganglia in the anterior half of the cut larva, signed with arrows. (c) Brain ganglia already cleaned Drosophila Comet Assay 275

● Clean it of tissue debris without breaking the lobes (Fig. 1c ). ● Take both lobes with the tungsten wires, place them on the parafi lm Ringer drop, and break/tear/shred them with the wires. ● Repeat the process with three more larvae for each parafi lm slide. ● Add 15 μL of Ringer solution to the drop with the brains of four larvae, and pipette up and down several times to individualize cells. Take everything to one Eppendorf tube and keep it on ice and darkness until all the work with the remaining concentrations is fi nished. ● Two slides, and therefore two Eppendorf tubes, with four larvae brains each one, must be prepared per concentration.

3.1.4 Agarose Slide ● Smooth slides with one frosted end are kept in ethanol, at Preparation −20 ºC, for at least 24 h. Before use, they are removed from the ethanol and let air-dry. ● First layer is prepared between 1 and 3 days before the comet assay is performed. Prepare 5 mL of normal melting point agarose (NMA; Invitrogen, Life Technologies Cat. No. 15510- 019), at 0.5 %, in the microwave and keep it warm. Add 150 μL in one dry slide and spread it with the fi nger, covering every part except the frosted end. Repeat with the other slides. Place the slides at 60–65 ºC for 15–20 min in a heat cabinet, until agarose became solid and transparent. Keep the slides at room temperature. ● Second layer: Prepare 5 mL of low melting point agarose (LMA; Gibco BRL, Life Technologies Cat. No. 15517-014), at 0.73 %, in the microwave and keep it at 45–50 ºC. From now on, work under red light. Add 65 μL of this agarose to the 30 μL with the brain cells in one Eppendorf tube. Mix by pipetting but avoid bubbles. Add the resultant 95 mL to one slide with the fi rst agarose layer in random drops. Place one coverslip (24 × 60 mm) checking that the agarose spreads on the slide. Repeat the process with the other slides. Place the slides at 4 ºC and darkness, for at least 15 min. Remove the coverslip carefully sliding it along the slide. ● Be aware that the fi nal concentration of agarose in this second layer should be 0.5 %. ● Third layer: It is optional. Prepare 5 mL of LMA at 0.5 % in the microwave and keep it warm. Add 75 μL of this agarose over the second layer, in random drops, and place a new coverslip, checking again that the agarose spreads on the slide. Repeat the process for the other slides. Place them at 4 ºC and darkness for at least 15 min. 276 L. María Sierra et al.

3.1.5 Cell Lysis ● Remove the coverslip carefully sliding it along the slide in all the slides. ● Place the slides in a slide support and submerged it in 200 mL of lysis solution, freshly prepared, in a Coplin jar. Keep them at 4 ºC and darkness for 2 h. ● In these 2 h, prepare the denaturing and electrophoresis buffer, and keep it at 4 ºC.

3.1.6 Denaturation ● Place the electrophoresis tank in a cold chamber, or in a box and Electrophoresis with ice, to keep the process at 4 ºC. ● Remove the slides from the lysis solution and place them in the electrophoresis tank, with their frozen ends to the cathode and touching each other. Avoid empty spaces among slides. ● Cover them with the cold denaturing and electrophoresis buffer, and keep them there for 20 min. Do not add much buffer or you would have to remove it. ● Connect the tank to a power supply. Fix the voltage to 0.9 V/ cm (across the platform holding the slides), and adjust the intensity to 300 mA adding more denaturing and electrophoresis buffer. ● Run the electrophoresis in cold and darkness for 20 min.

3.1.7 Neutralization ● Under red light, remove one slide, wash it two times with and Fixation 2 mL of neutralization buffer, and place it vertical in a tray. Repeat for the other slides. ● Let them air-dry for a few minutes. ● Place the slides ﬂ at in a tray and cover them with absolute ethanol. Keep them there for 3 min. ● Remove the slides from the ethanol, one by one, place them in a new tray over ﬁ lter paper, and let them dry overnight at room temperature and darkness.

3.2 Comet Assay ● In bottles containing standard corn fl our medium, about 300 with Hemocyte Cells: females and the corresponding males are placed. Collect eggs Specifi c Procedure from these females during 8-h periods. To do that, all males and females are moved to new bottles containing standard 3.2.1 Larvae Collection medium each 8 h. ● After 3 days, 72 ± 4 h larvae (third instar) are collected by washing the culture bottles softly with tap water. Then, fi lter the larvae through a sieve of 500 and 400 μm to separate the larvae from the standard culture media rests.

3.2.2 Treatment ● Place the resulting 3-day-old larvae in disposable plastic vials containing 4.5 g of lyophilized instant Drosophila medium hydrated with 10 mL of the test solution (or water or solvent as controls). Drosophila Comet Assay 277

● The treatment vials are covered with a cotton plug and kept in an incubator at 25 °C and 60 % humidity for 24 h. ● Negative control larvae receive instant Drosophila medium hydrated with distilled water (or solvent), while positive control larvae usually receive 1 mM of ethyl methanesulfonate (EMS).

3.2.3 Hemocyte Isolation ● Third-instar larvae of 96 ± 2 h are extracted from the culture medium, washed, sterilized with ~5 % sodium hypochlorite, and dried with fi lter paper. ● To collect the hemolymph and circulating hemocytes, the cuti- cle of each larva is disrupted using two fi ne forceps, avoiding damage to internal organs. The procedure is performed into a drop (40 μL) of cold PBS containing ~0.07 % of phenylthiourea deposited in a well (6 mm ø) of a Tefl on-printed diagnostic slide. ● After collecting the hemolymph from 10 larvae, the drop of PBS/hemocytes is removed from the diagnostic slide with a micropipette and placed within a microcentrifuge tube (1.5 mL). Complete up to 40–60 larvae per treatment. ● The tubes obtained from each treatment are centrifuged at 300 g for 10 min at 4 °C. ● Finally, the supernatant is removed and the cell pellet is resuspended in 20 μL of cold PBS containing ~0.07 % PTU.

3.2.4 Agarose Slides ● First layer: Prepare 1 % of normal melting point agarose (NMA,

Preparation 500 mg/50 mL d-H2 O). ● Heat the NMA in a microwave until it is diluted and avoid boiling. ● Immerse slides gently in hot NMA. Then, drain and clean the back of the slide with paper and dry at room temperature, avoiding high humidity conditions. ● Second layer: Prepare 0.75 % of low melting point agarose (LMA, 375 mg/50 mL PBS). ● Heat the LMA in a microwave until it is diluted and avoid boiling. ● Cell samples from each treatment are carefully resuspended in 75 μL of 0.75 % LMA at 37 °C and mixed by pipetting. ● Place two drops, each with a volume of 20 μL of the mixture, in the slide precoated with 1 % NMA (one replica per treatment). ● Cover each drop with a coverslip of 22 × 22 mm. Store at 4 ºC during 10 min. 278 L. María Sierra et al.

3.2.5 Cell Lysis ● Immediately after agarose solidiﬁ cation, the coverslips are removed. ● The slides are immersed in a cold fresh lysis solution for 2 h at 4 °C in a dark chamber.

3.2.6 Denaturation ● The slides are removed from the lysis solution and transferred and Electrophoresis into an electrophoresis chamber. ● Fill the electrophoresis chamber with fresh and cold electrophoresis buffer until the microgel slides are covered. ● Keep the slides in this alkaline buffer for 25 min at 4 °C to allow DNA unwinding and expression of alkali-labile sites. ● Turn the power supply to 25 V (0.7 V/cm) and adjust the current to 300 mA by placing or removing alkaline buffer. Electrophoresis is performed for 20 min at 4 °C.

3.2.7 Neutralization ● Slides are neutralized with two washes of 5 min (0.4 M Tris, and Fixation pH 7.5). ● After that, slides are air-dried for a few minutes. ● Place the slides ﬂ at in a tray and cover them with absolute ethanol. Keep them there for 3 min. ● Remove the slides from the ethanol, one by one, place them in a new tray over ﬁ lter paper, and let them dry overnight at room temperature and darkness.

3.3 Common Steps ● Code the slides, for blind scoring. for Both Methods ● Prepare the working 1x solution. Dilute the ethidium bromide 3.3.1 Staining or the DAPI stocks with distilled water. ● In the case of ethidium bromide, each slide is stained with 40 μL of a 1:4 dilution of the working solution, that is, with 40 μL of a 0.4 μg/mL fi nal concentration. In the case of DAPI, each slide is stained with 20 μL of a concentration of 1 μg/mL. ● When staining with ethidium bromide, add 1 μL of fl uorescence protector VECTASHIELD® (Vector Laboratories, Inc., Burlingame, CA 94010, USA) to the staining solution, for each slide. ● Spread the 41 μL on one slide, in random drops, and place a coverslip. Repeat for the other slides. ● In the microscope, use an excitation fi lters adequate for the ethidium bromide (530–560 nm) and for DAPI (360–370) staining. ● Place the slide on the stage of the microscope, with the frozen end to the right. ● Take pictures of at least 50 cells per slide. Drosophila Comet Assay 279

3.3.2 Image Analysis We use the software Komet 5 (Kinetic Imaging Ltd., UK) for image analysis, but there are other software programs available in the market. We collect the information provided for four Comet parameters: tail DNA, tail length, tail extent moment, and Olive tail moment.

● Tail DNA is the percentage of DNA that is in the tail. ● Tail length is the length of the tail, measured in μm, from the border of the head. ● Tail extent moment is the product of the tail DNA and tail length divided by 100. ● Olive tail moment is the product of the tail DNA and the difference between tail mean and head mean (proﬁ le centers of gravity), divided by 100.

3.4 Statistical This analysis is performed independently of the origin (neuroblasts Analysis or hemocytes) of the comet. At least three different chemical concentrations should be analyzed in each experiment, in addition to the negative and positive controls. As none of the parameters used (tail DNA, tail length, tail extend moment, and Olive tail moment) follows a normal distribution, the comparison between the result of each chemical concentration and the negative control should be performed with a nonparametric statistical test, such as the Mann–Whitney U test. We use the STATISTICA software for Windows (StatSoft, Inc., 1995, STATISTICA for Windows, Computer Program Manual, Tulsa, OK). To improve the statistical analysis of results, three different independent experiments can be performed for each analyzed chemical. Then comparisons between each chemical concentration and the negative control are carried out comparing the arithmetic means of the average values of the three experiments with a Student’s t -test. This test can be performed with any statistical program, even with the Excel software. In addition to this analysis, sometimes it is necessary, or only interesting, to perform dose–response regression analysis, to determine a possible response increase with the increasing concentration, when working on the linear part of the dose– response curve. This analysis can be performed with any statistical program. However, in this analysis, the most important part is to know whether the regression slope is statistically different from zero. For this analysis, we normally used the LightStat3 program, developed by our colleague Dr. P. Casares, and freely distributed. Nevertheless, other programs can be also used. 280 L. María Sierra et al.

Acknowledgments

The authors thank the ﬁ nancial support of MEC Spain (project CT2004-03005) and FICYT (PCTI Asturias, project PC07-018) to LMS and CIRIT (project 2009SGR-725) to RM. ERC thanks the support of Dirección General de Investigación y Postgrado, UC Temuco, DGIP UCT CD 2010-01 project, and MECESUP UCT 0804 project. LA thanks the support of Instituto Universitario de Oncología del Principado de Asturias, Obra Social Cajastur.

References

1. Östling O, Johanson KJ (1984) Microelectro- 10. Møller P, Knudsen LE, Loft S et al (2000) The phoretic study of radiation-induced DNA dam- Comet Assay as a rapid test in biomonitoring ages in individual mammalian cells. Biochem occupational exposure to DNA-damaging Biophys Res Commun 123:291–298 agents and effect of confounding factors. 2. Singh NP, McCoy MT, Tice RR et al (1988) A Cancer Epidemiol Biomarkers 9:1005–1015 simple technique for quantitation of low levels 11. Faust F, Kassie F, Knasmüller S et al (2004) of DNA damage in individual cells. Exp Cell The use of the alkaline comet assay with lym- Res 175:184–191 phocytes in human biomonitoring studies. 3. Speit G, Hartmann A (1999) The comet assay Mutat Res 566:209–229 (single cell gel test). A sensitive genotoxicity 12. Hoffmann H, Högel J, Speit G (2005) The test for the detection of DNA damage and effect of smoking on DNA effects in the comet repair. In: Henderson DS (ed) Methods in assay: a meta-analysis. Mutagenesis 20:455–466 molecular biology. DNA repair protocols: 13. Burlinson B, Tice RR, Speit G et al (2007) In eukaryotic systems, vol 113. Humana Press, vivo Comet Assay workgroup, part of the Fourth Totowa, pp 203–212 International Workgroup on Genotoxicity 4. Tice RR, Agurell E, Anderson D et al (2000) Testing: results of the in vivo Comet Assay work- The single cell gel/comet assay: guidelines for group. Mutat Res 627:31–35 in vitro and in vivo genetic toxicology testing. 14. Dušinská M, Collins AR (2008) The comet assay Environ Mol Mutagen 35:206–221 in human biomonitoring: gene-environment 5. Hartmann A, Agurell E, Beevers C et al interactions. Mutagenesis 23:191–205 (2003) Recommendations for conducting the 15. Uriol E, Sierra M, Comendador MA et al in vivo alkaline comet assay. Mutagenesis 18: (2013) Long-term biomonitoring of breast 45–51 cancer patients under adjuvant chemotherapy: 6. Collins AR (2004) The Comet assay for DNA the Comet assay as a possible predictive factor. damage and repair. Principles, applications, and Mutagenesis 28:39–48 limitations. Mol Biotechnol 26:249–261 16. Collins AR, Horváthová E (2001) Oxidative 7. Collins AR, Gedik CM, Olmedilla B et al DNA damage, antioxidants and DNA repair: (1998) Oxidative DNA damage measured in applications of the comet assay. Biochem Soc human lymphocytes; large differences between Trans 29:337–341 sexes and between countries, and correlations 17. Collins AR, Dusinska M, Horvathova E et al with heart disease mortality rates. FASEB J (2001) Inter-individual differences in DNA 12:1397–1400 base excision repair activity measured in vitro 8. Somorovská M, Szabová E, Vodicka P et al with the comet assay. Mutagenesis 16:297–301 (1999) Biomonitoring of genotoxic risk in 18. Collins AR, Gaivão I (2007) DNA base excision workers in a rubber factory: comparison of the repair as a biomarker in molecular epidemiology Comet assay with cytogenetic methods and studies. Mol Aspects Med 28:307–322 immunology. Mutat Res 445:181–192 19. Gaivão I, Piasek A, Brevik A et al (2009) 9. Kassie F, Parzefall W, Knasmüller S (2000) Comet assay-based methods for measuring Single cell gel electrophoresis assay: a new DNA repair in vitro; estimates of inter- and technique for human biomonitoring studies. intra-individual variation. Cell Biol Toxicol Mutat Res 463:13–31 25:45–52 Drosophila Comet Assay 281

20. Dušinská M, Collins AR (2010) DNA oxida- 34. Müller HJ (1927) Artifi cial transmutation of tion, antioxidant effects, and DNA repair mea- the gene. Science 143:581–583 sured with the comet assay. In: Aldini G, Yeum 35. Auerbach C, Robson JM (1942) Experiments KJ, Niki E, Russell R (eds) Biomarkers for anti- on the action of mustard gas in Drosophila. oxidant defense and oxidative damage: princi- Production of sterility and mutation. Report to ples and practical applications. Blackwell Ministry of Supply, w. 3979 Publishing Ltd., Oxford, pp 261–282 36. Mukhopadhyay I, Chowdhuri DK, Bajpayee M 21. Navarrete MH, Carrera P, de Miguel M et al et al (2004) Evaluation of in vivo genotoxicity (1997) A fast comet assay variant for solid tis- of cypermethrin in Drosophila melanogaster sue cells. The assessment of DNA damage in using the alkaline Comet assay. Mutagenesis higher plants. Mutat Res 389:271–277 19:85–90 22. Menke M, I-Peng Chen I-P, Angelis KJ et al 37. Siddique HR, Chowdhuri DK, Saxena DK et al (2001) DNA damage and repair in Arabidopsis (2005) Validation of Drosophila melanogaster thaliana as measured by the comet assay after as an in vivo model for genotoxicity assessment treatment with different classes of genotoxins. using modifi ed alkaline Comet assay. Mutat Res 493:87–93 Mutagenesis 20:285–290 23. Lee RF, Steinert S (2003) Use of the single cell 38. Siddique HR, Gupta SC, Dhawan A et al gel electrophoresis/comet assay for detecting (2005) Genotoxicity of industrial solid waste DNA damage in aquatic (marine and freshwater) leachates in Drosophila melanogaster. Environ animals. Mutat Res 544:43–64 Mol Mutagen 46:189–197 24. Dhawan A, Bajpayee M, Parmar D (2009) 39. Siddique HR, Sharma A, Gupta SC et al (2008) Comet assay: a reliable tool for the assessment DNA damage induced by industrial solid waste of DNA damage in different models. Cell Biol leachates in Drosophila melanogaster: a mech- Toxicol 25:5–32 anistic approach. Environ Mol Mutagen 49: 25. Verschaeve L, Gilles J (1995) Single-cell gel 206–216 electrophoresis assay in the earthworm for the 40. Mishra M, Sharma A, Negi MP et al (2011) detection of genotoxic compounds in soils. Tracing the tracks of genotoxicity by trivalent Bull Env Contam Toxicol 54:112–119 and hexavalent chromium in Drosophila mela- 26. Cotelle S, Férard JF (1999) Comet Assay in nogaster. Mutat Res 722:44–51 genetic ecotoxicology: a review. Environ Mol 41. Sharma A, Mishra M, Shukla AK et al (2012) Mutagen 34:246–255 Organochlorine pesticide, endosulfan induced 27. Dixon DR, Pruski AM, Dixon LRJ et al (2002) cellular and organismal response in Drosophila Marine invertebrate eco-genotoxicology: a melanogaster. J Hazard Mater 221–222: methodological overview. Mutagenesis 17: 275–287 495–507 42. Mishra M, Sharma A, Shukla AK et al (2013) 28. Jha AN (2008) Ecotoxicological applications Transcriptomic analysis provides insights on and signifi cance of the comet assay. Mutagenesis hexavalent chromium induced DNA strand 23:207–221 breaks and their possible repair in midgut cells 29. Bilbao C, Ferreiro JA, Comendador MA et al of Drosophila melanogaster larvae. Mutat Res (2002) Infl uence of mus201 and mus308 muta- 747–748:28–39 tions of Drosophila melanogaster on the geno- 43. Shukla AK, Pragya P, Chowdhuri DK (2011) A toxicity of model chemicals in somatic cells modifi ed alkaline Comet assay for in vivo detec- in vivo measured with the comet assay. Mutat tion of oxidative DNA damage in Drosophila Res 503:11–19 melanogaster. Mutat Res 726:222–226 30. Hallström I, Blank A, Atuma S (1984) Genetic 44. Siddique YH, Fatima A, Jyoti S et al (2013) variation in cytochrome P450 and xenobiotic Evaluation of the toxic potential of Graphene metabolism in Drosophila melanogaster. copper nanocomposite (GCNC) in the third Biochem Pharmacol 33:13–20 instar larvae of transgenic Drosophila melano- 31. Søndergaard L (1993) Homology between the gaster (hsp70-lacZ)Bg(9.). PLoS One 8:e80944. mammalian liver and the Drosophila fat body. doi: 10.1371/journal.pone.0080944 Trends Genet 9:193 45. Guanggang X, Diqiu L, Jianzhong Y et al 32. Henderson DS (1999) DNA repair defects and (2013) Carbamate insecticide methomyl con- other (mus)takes in Drosophila melanogaster. fers cytotoxicity through DNA damage induc- Methods 18:377–400 tion. Food Chem Toxicol 53:352–358 33. Sekelsky JJ, Brodsky MH, Burtis KC (2000) 46. Radyuk SN, Michalak K, Rebrin I et al (2006) DNA repair in Drosophila: insights from the Effects of ectopic expression of Drosophila Drosophila genome sequence. J Cell Biol DNA glycosylases dOgg1 and RpS3 in mito- 150:F31–F36 chondria. Free Radic Biol Med 41:757–764 282 L. María Sierra et al.

47. Brennan LJ, Haukedal JA, Earle JC et al (2012) 52. Demir E (2012) In vivo genotoxicity assess- Disruption of redox homeostasis leads to oxi- ment of diﬂ ubenzuron and spinosad in dative DNA damage in spermatocytes of Drosophila melanogaster with the comet assay Wolbachia-infected Drosophila simulans. using haemocytes and the SMART assay. Insect Mol Biol 21:510–520 Fresenius Environ Bullet 21:3894–3900 48. Carmona ER, Guecheva TN, Creus A et al 53. Demir E, Kaya B (2013) Studies on the geno- (2011) Proposal of an in vivo comet assay using toxic properties of four benzyl derivatives in haemocytes of Drosophila melanogaster. the in vivo comet assay using haemocytes of Environ Mol Mutagen 52:165–169 Drosophila melanogaster . Fresenius Environ 49. Carmona ER, Creus A, Marcos R (2011) Bullet 22:1590–1596 Genotoxic effects of two nickel-compounds in 54. García Sar D, Aguado L, Montes Bayón M et al somatic cells of Drosophila melanogaster. (2012) Relationships between cisplatin- Mutat Res 718:33–37 induced adducts and DNA strand-breaks, 50. Carmona ER, Creus A, Marcos R (2011) mutation and recombination in vivo in somatic Genotoxicity testing of two lead-compounds in cells of Drosophila melanogaster, under differ- somatic cells of Drosophila melanogaster. ent conditions of nucleotide excision repair. Mutat Res 724:35–40 Mutat Res 741:81–88 51. Sabella S, Brunetti V, Vecchio G et al (2011) 55. Marcos R, Carmona ER (2013) The wing-spot Toxicity of citrate-capped AuNPs: an in vitro and the comet tests as useful assays detecting and in vivo assessment. J Nanoparticle Res genotoxicity in Drosophila. Methods Mol Biol 13:6821–6835 1044:417–427 Chapter 16

The SMART Assays of Drosophila: Wings and Eyes as Target Tissues

Ricard Marcos , L. María Sierra , and Isabel Gaivão

Abstract

Drosophila melanogaster is a useful model for genetic studies, including those aiming to detect genotoxicity. The SMARTs (somatic mutation and recombination tests) in Drosophila are in vivo short-term assays that assess genetic damage induction in somatic cells of adult fl ies after larval exposure. They are less oner- ous than other Drosophila tests (the germinal ones) and are very sensitive, specifi c, and accurate. All of them are based in obtaining heterozygous offspring with particular phenotypes, in which the occurrence of genotoxic events in proliferating (imaginal disk) cells leads to an alteration of this phenotype, which is manifested as mutant spots easily detectable in a normal phenotypic background. Among the various SMARTs available, in this chapter, we will discuss the wing-spot assay ( mwh/fl r 3 ) and the eye-spot white/ white+ ( w/w+ ) assay, provide a comprehensive overview, introduce the principles of the assays, and provide the details to properly conduct both of them.

Key words Drosophila melanogaster , SMART assays , Larvae instar , white/white+ , mwh/ﬂ r 3 , Eye-spot , Wing-spot

1 Introduction

The somatic mutation and recombination (SMART) assays are in vivo genotoxicity assays designed for and carried out in Drosophila melanogaster [1 ]. Using genetic markers phenotypically observed in adult tissues, the assays detect in heterozygous or transheterozygous individuals the loss of heterozygosity for the selected genetic markers [ 2 , 3 ]. Although this loss of heterozygosity may be theoretically due to different events such as point mutations/deletions, nondisjunction, and homologous mitotic recombination, nondisjunction processes are generally not especially relevant for most of the tested chemicals [ 4 , 5 ]. There are two different SMART assays that, although with the same genetic base, targeted two different adult tissues: the wings in the case of the wing-spot test [ 4 ] and the eyes in the case of the eye-spot test [ 5 ]. In both assays, the detection of mutation and recombination in somatic cells is carried out

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_16, © Springer Science+Business Media New York 2014 283 284 Ricard Marcos et al.

Fig. 1 Electronic microscopy view of mwh and ﬂ r morphologies. (a ) and (b ) morphology of mwh phenotype. (c ) and (d ) morphologies of ﬂ r phenotype

through the scoring of adult tissues looking for phenotypically mutant spots. Since they detect mainly the induction of gene mutation and homologous recombination, they are very useful tools to analyze in vivo the potential genotoxicity of chemicals in the somatic cells of a higher eukaryotic organism [ 4 , 5 ]. It should be noted that the quantifi cation of mitotic recombination in somatic cells is relevant for genotoxicity screening, because aberrant recombination activity is commonly associated with carcinogenesis [6 ]. In fact, these assays have been used in the analysis of the genotoxicity of different types of agents, such as radiation [7 – 9 ], metals [10 – 13 ], chemicals [14 – 20 ], plant extracts/products [21 , 22 ], therapy drugs [ 23 –26 ], food products [27 – 29 ], pesticides [ 30 – 32 ], different types of pollutants [ 33 – 35 ], and nanoparticles (see Chap. 17 ), among others. They have also been used in the analysis of possible antimutagens [36 – 42 ]. The wing-spot test uses the recessive genetic markers multiple wing hair ( mwh ) and fl are-3 (fl r 3 ) (Fig. 1 ), localized on chromosome 3 [ 43 ], to check the induction of mutant spots that refl ect the loss of heterozygosity due to point mutation, deletion, nondisjunction, or mitotic recombination (Fig. 2 ) [4 ]. The eye w/w + SMART assay uses the X-chromosome white ( w ) gene as recessive marker to monitor, on wild-type eyes, the presence of white clones that indicate the occurrence of loss of The Drosophila Wing and Eye SMART Assays 285

a centromere mwh flr Bd S

0.3 38.8 47.7 91.9

b + flr3 + mwh + +

X ++BdS mwh + +

mwh ++ mwh + +

+ flr3 + ++BdS

Transheterozigous Balanced heterozigous

Fig. 2 ( a) Relative position of the mwh , ﬂ r , and Bd S markers on chromosome 3. ( b) Cross between mwh females and ﬂ r males with the resulting offspring: transheterozygous and balanced heterozygous individuals

heterozygosity, by point mutations and/or deletions, at the white locus, and nondisjunction and homologous mitotic recombination in w/w + somatic cells of Drosophila in vivo (Fig. 3 ), but without discriminating among these genetic endpoints [5 ]. A more elaborated version of this assay, also developed by Vogel and Nivard [44 ], allows the detection of chromosomal aberrations induced in late larval stages. However, the high sensitivity of the necessary strains makes this version not necessarily worth the effort, and because of that, the protocol presented here corresponds to the general version and not the improved one. Nevertheless, due to the signiﬁ cance of the male eyes analysis, indicated in this last version [44 ], we have incorporated it in the protocol presented in this chapter. This allows the quantiﬁ cation of mitotic homologous recombination, because the loss of the w + marker in hemizygous cells also leads to white clones in males; in this case, white clones can be the result of intrachromosomal recombination and/or point mutations and deletions at the white locus [44 ]. It is important to remark that, for both assays, the obtained results can be modulated by the ability of the used strain to properly metabolize xenobiotic compounds [15 , 45 , 46 ]. In this chapter, protocols for both SMART tests are presented with enough detail to allow their performance in any laboratory. 286 Ricard Marcos et al.

Heterozygous female cell w

w + +

MITOTIC DIVISION

Daughter cells Gene Homologous mutation Non recombination disjunction

w w w

+ w m m +

m + + + Deletion w

+ m

Fig. 3 Scheme showing the different genetic processes that can lead to white mutant spots in daughter cells

2 Materials and Reagents

● Drosophila strains ● Culture chambers (18, 21, and 24 ºC) ● Sieve (500 and 400 μm pores) ● Fine tweezers ● Microscope ● Microscopic slides The Drosophila Wing and Eye SMART Assays 287

● Coverslips (24 × 60 mm) ● Freezer (−20 ºC) ● Refrigerator ● Eppendorf/microcentrifuge tubes (1.5 mL) ● Small brushes to handle ﬂ ies ● Stereoscopic microscope ● Black plaque ● Cold light lamp

2.1 Drosophila Wing-spot test —two D. melanogaster strains are used for the wing- Strains spot test:

● The multiple wing hair strain with genetic constitution y ; mwh j ● The fl are-3 strain with genetic constitution fl r 3 /Ln (3LR ) TM3, Bd s Both strains carry mutations that visibly affect the phenotype of the trichomes (hairs) of the wing cells [2 ]. The mwh mutation is recessive and viable in homozygosis, and it is located on chromosome 3. Its phenotypic expression is characterized by the appearance of three or more hairs in each cell, instead of one per cell, which is the normal phenotype. The fl r 3 mutation is also located on the same arm of chromosome 3. It is recessive and produces lethality in homozygosis in the germ line, but not in somatic cells. The fl are phenotype is quite variable, ranging from short, thick, and deformed to amorphous-like globular hairs. The phenotype of the markers is visible in Fig. 1 . More details of the other genetic markers and phenotype descriptions of the strains employed are given by Lindsley and Zimm [43 ]. Both strains are grown up in a culture chamber at 25 ± 1 °C, with a relative humidity of ~60 %. Eye-spot test —two strains are used also for the eye w/w + assay:

● A wild-type strain homozygous for the wild-type allele of the white gene ( w+ /w+ ). ● A mutant strain for the white gene, and therefore with white eyes ( w/w ), if possible with the same genetic background as the wild-type strain. These eyes are white because the pigmented cells on the ommatidia do not have pigment, that is, they are transparent. The white gene is located at position 1.5 of the X chromosome [43 ]. 288 Ricard Marcos et al.

2.2 Medium To maintain the strains and to obtain larvae, the following stan- for Growing dard medium can be used: and Maintaining ● Corn ﬂ our, 170 g the Strains ● Yeast, 120 g 2.2.1 Standard Corn ● NaCl, 2 g Flour Medium (for the Wing-Spot Test) ● Agar-agar, 10 g Mix the components, add 1,200 mL of cold water, mix well the mix, and heat to boiling. Remove from heat, cool, and add to the mixture 4 mL of Nipagin fungicide, dissolved in 10 % ethanol and 4 mL of propionic acid. Once prepared, the culture medium is distributed in culture glass bottles of 125 mL. In each bottle, approximately 25 mL of media are placed and allowed to stand for a few hours before use. Place a small piece of paper soaked with the insecticide tetradifon into each bottle in order to control mite populations, while maintaining the humidity inside the culture bottles.

2.2.2 Standard ● Baker yeast, 100 g Yeast-Sugar Medium ● Sugar, 100 g (for the Eye-Spot Test): ● Agar-agar, 9 g for 1 L of Water ● Salt approx. 2 g Mix everything, heat it, and keep it boiling for 30 min. Cool it down to 55 ºC, add 5 mL of propionic acid, mix well, and pour 20 mL of medium in glass bottles of 250 mL volume. Let the bottles dry overnight before using them.

2.3 Medium ● Carolina Drosophila Medium Formula 4-24® (Carolina Biological for Treatments Supply Company, USA) (wing-spot test): 4.5 g of medium hydrated with 10 mL distilled water or chemical solutions ● Carolina Drosophila Medium Formula 4-24® (Carolina Biological Supply Company, USA) (eye-spot test): 20 mL of medium (approx. 6 g) hydrated with 20 mL of distilled water/ solvent or chemical solutions, per bottle

2.4 Faure’s Solution ● Gum arabic, 30 g ● Choral hydrate, 50 g

● d-H2 O, 50 mL ● Glycerol, 20 mL Mix and dissolve well the three ﬁ rst components. Heating— without boiling—is recommended to facilitate dilution. Remove from heat and cool for 20 min. Then, add the 20 mL of glycerol and mix well. Once prepared, transfer the solution into an amber glass bottle with dropper. The Drosophila Wing and Eye SMART Assays 289

2.5 Eye-Scoring ● Solution 1: for 100 mL, mix 90 mL of absolute ethanol, 9 mL Solutions of distilled water, and 1 mL of Triton X-100. ● Solution 2: for 100 mL, mix 10 mL of Solution 1 with 90 mL of distilled water.

3 Methods

3.1 SMART Assay ● The assay use transheterozygous individuals, with (mwh +/ Using Wing Markers + ﬂ r 3 ) genotype, resulting from the cross between ﬂ r 3 females (the Wing-Spot Assay) and mwh males (see Fig. 2 ).

● 3.1.1 Obtaining To get enough number of individuals to perform the cross, the Transheterozygous Larvae number of individuals (culture bottles) in the parental strains for Treatments must be previously increased. ● Once a suffi cient number of virgin females from the fl r 3 strain (~300 individuals) have been obtained, they are mated with males of the mwh strain. For the mates, 125 mL bottles with standard growing medium are used. To be sure that females are virgin, they must be collected no more than 6 h after the pupa hatching. ● Collect eggs from this cross during 8-h periods in culture glass bottles containing standard medium. ● After 3 days of culture, collect the resulting 72 ± 4 h larvae (third instar), by washing the culture bottles softly with tap water. Then, fi lter the larvae through a sieve of 500 and 400 μm to separate the larvae from the standard culture media rests. ● Place the resulting 3-day-old larvae in disposable plastic vials containing 4.5 g of lyophilized instant Drosophila medium hydrated with 10 mL of the test solution (or water or solvent as controls). The treatment vials are covered with a cotton plug and kept in an incubator at 25 °C and 60 % humidity until emergence of adult individuals.

3.1.2 Preparation ● Collect the emerging adults. and Mounting ● Classify the adult according to their phenotype. of Drosophila Wing Slides Note: from the cross, two different types of offspring are obtained: transheterozygous individuals ( mwh +/ + fl r ) and balanced heterozygous ( mwh + / + Bd S ). These last fl ies are easily visualized because they have the border of the wing sawed (due to the mutation Bd S Beadle serrate). ● Remove the wings. If wings are not prepared immediately after adult collection, emerged adult individuals can be stored in plastic vials with 70 % ethanol at cold temperature (4 ºC) until their use. ● To remove possible rests of culture media deposited on the wings, the fl ies should be washed with a mix of ethanol and

d-H2 O into a watch glass. 290 Ricard Marcos et al.

● Each fl y is placed on a well slide, with a drop of Faure’s solution. To proceed with the extraction of its wings, fi ne forceps under a binocular stereomicroscope are used. ● Once the two wings are separated from the body of the fl y, they are carefully taken with fi ne tweezers and placed in pairs and aligned on a clean microscopic slide. ● For each slide, 48 wings (24 individuals) should be placed, with the same proportion between males and females. The preparation is maintained for 24 h in a dust-free place (i.e., inside a Petri dish) for drying. ● Since both transheterozygous and balanced heterozygous individuals are obtained in the crosses, both sets of individuals must be used (normal and sawed wings). ● Wing preparations can be permanent. To proceed, a few drops of Faure’s solution are placed in the center of the slide and then covered with a coverslip of 24 × 60 mm, making a light pressure for an even distribution of the solution. To prevent formation of air bubbles and to spread the wings, a weight (between 250 and 500 g) must be placed on the slide preparation. After 24 h, the preparation is completely dry and the weights can be removed.

3.1.3 Microscopic ● The wings are scored at 400 times magnifi cation for the pres- Analysis of the Wings ence of mutant spots. ● Wing from transheterozygous individuals should be scored fi rst. If results are positive, the serrate set of wings should be scored to provide further information about the mechanisms producing genotoxicity. Since balanced heterozygous (mwh +/ + Bd S ) fl ies do not recombine, due to the presence of the TM3 inversion, the observed differences between genotypes must be attributed to somatic recombination. ● In each series, between 40 and 80 wings are scored (20–40 individuals). ● Mutant spots are recorded according to their position in each area of the wing (i.e., A, B, C, C′, D, D′, and E) (see Fig. 4 ), category, and size. ● Mutant spots can be classifi ed as: ● Small single mwh spots (1–2 cells) ● Large single mwh spots (3 or more cells) ● Twin (adjacent mwh and fl are ) spots ● Single fl r 3 spots

3.1.4 Data Analysis ● To determine the efﬁ ciency of the treatments, the frequency of each type of mutant spot observed in the treatments must be compared with the values observed in the negative control. The Drosophila Wing and Eye SMART Assays 291

Fig. 4 Scheme showing the different sectors from a wing

This comparison is carried out by using the conditional binomial test of Kastenbaum and Bowman [ 47 ] with a signiﬁ cance level of α = β = 0.05. ● To determine the overall response of an agent as positive, weakly positive, negative, or inconclusive, the multiple-decision procedure described by Frei and Würgler [48 ] is used. ● The frequency of clone formation is calculated, without size correction, by dividing the number of mwh clones per wings by 24,400, which is the approximate number of cells analyzed per wing [49 ].

3.2 SMART Assay ● Collect virgins, females and males separately, of both strains for Using Eye Markers some days. We recommend collecting them from Monday to (the w/w+ Friday morning. Eye-Spot Assay) ● Mix all the females isolated over these days, and also the males, 3.2.1 Obtaining from each strain, to avoid age effect. Heterozygous w/w + Larvae ● Mate 50 w+ /w+ females with at least 30 w/ Y males in vials for 2 days (from Friday afternoon to Monday morning). In this way,

both F1 females and males could be scored. If w/w females + were mated with w /Y males, only the F1 females should be scored. ● Move the fl ies in each vial to bottles with instant treatment medium. Let them lay eggs for the chosen time. At this point, it is necessary to mention that there are two different types of treatment: chronic and surface. In the chronic treatment, the mated fl ies are allowed to lay eggs in medium hydrated with the chemical solutions or distilled water/solvent for the negative control. The fl ies lay eggs for 24 or 48 h. When removed, the adults should be disposed off. After hatching, the larvae spend all their time exposed to the chemical. In the surface treatment, the mated fl ies are allowed to lay eggs in medium hydrated with distilled water for 24 h. After removing the adults, the larvae may be treated at different times, to treat different larval stages. The removed adults can be used for 292 Ricard Marcos et al.

egg-laying in other bottles. The treatment is performed spreading 1.5 mL of the chemical solution or distilled water/ solvent (for the negative control) over the medium of each bottle. ● Bottles are kept at 24 ± 1 ºC until adult eclosion. ● Collect and count the adults, separate males and females, and keep them in vials with fresh standard medium until the eye scoring. ● Since it is impossible to count eggs in instant Carolina medium, the estimation of toxicity in these experiments is semiquantita- tive, comparing the number of eclosed adults in each chemically treated bottles with those of the negative control. Note: The chronic treatment needs more chemical product, and because the larvae will spend all their life exposed to the chemical, the analyzed concentrations cannot be very high. The surface treatment needs lesser product and the analyzed concentration can be at least 10 times higher than in the chronic one. Note: For surface treatments, the same adults may be used to lay eggs in four different bottles, changing them to new bottles on Tuesday, Wednesday, and Thursday and eliminating them on Friday. On this last day, the age of the larvae would be between 0 and 24 h in the last bottles (12 ± 12 h), between 24 and 48 h in the previous ones (36 ± 12 h), between 48 and 72 h in the second bottles (60 ± 12 h), and between 72 and 96 h in the ﬁ rst ones (84 ± 12 h), which are prepared on Monday.

3.2.2 Eye Scoring ● Place the fl ies, slept, in a black plaque and cover them with Solution 1. Place the plaque under the stereoscopic microscope and adjust the cold light (maximum intensity) to the observed fl y, using optic fi ber with two foci. Observe the fl ies at least at ×40 magnifi cation. ● Check the eyes for dark spots that distort the precise order of ommatidia, like those presented in Fig. 5 , moving the eyes under the light and playing with the microscope focus.

Fig. 5 Drosophila eyes with white mutant spots ( dark spots that interfere with the ommatidia lines), small and medium size, and large white spots in the border of the eyes The Drosophila Wing and Eye SMART Assays 293

Note: Be aware that the spots are not dark because of a dark pigment in the pigmented cells, but because these cells in the spots are transparent and surrounded by pigmented cells. Because of this, when the spots are in the border of the eye, they are not dark but uncolored, transparent (Fig. 5 ).

● When Solution 1 evaporates, cover the fl ies with Solution 2. Keep the fl ies always covered with liquid. ● Two independent experiments must be carried out for each tested chemical, and at least 300 eyes should be scored per concentration, between both experiments. ● The spots are classifi ed, considering the number of white ommatidia, into small (S), when 1–2 ommatidia are affected; medium (M), when the size of the spots goes from 3 to 8 ommatidia; and large (L), for spots larger than 8 ommatidia. ● Since the spots found in males could not be induced by homologous recombination, the differences in spot frequencies between males and females represent the recombination effect.

3.2.3 Data Analysis ● Two different parameters may be determined in this test: (i) the number of spots per 100 eyes, as proposed by Vogel and Nivard [ 5 ], considering that two spots are independent when they are separated by at least four rows of wild-type ommatidia and (ii) the number of mosaic eyes, or eyes with at least one spot, as proposed by Ferreiro et al. [50 ]. The analysis of results is performed comparing the results of each analyzed concentration to that of the negative control, with a χ2 test. ● In addition, the frequency of mutant clones per 104 cells can be estimated, as described by Vogel and Nivard [ 5 ], as 2 nm/ NC, where n is the number of spots, m is the average clone size, N is the number of analyzed eyes, and C is 800 (the number of ommatidia per eye). ● As described for the wing-spot test, to determine the overall response of an agent as positive, weakly positive, inconclusive, or negative, the multiple-decision procedure described by Frei and Würgler [ 48 ] can be used.

Acknowledgments

The authors thank the ﬁ nancial support of their research activity: RM to CIRIT (project 2009SGR-725); LMS to MEC Spain (project CT2004-03005), FICYT (PCTI Asturias, project PC07-018), and Instituto Universitario de Oncología del Principado de Asturias, Obra Social Cajastur; and IG to the Portuguese Science and Technology Foundation (FCT) under the Project PEst-OE/ AGR/UI0772/2014. 294 Ricard Marcos et al.

References

1. Würgler FE, Vogel EW (1986) In vivo muta- 14. Sierra LM, Barros AR, García M et al (1991) genicity testing using somatic cells of Drosophila Acrolein genotoxicity in Drosophila melanogaster . In: de Serres FJ (ed) Chemical melanogaster.I. Somatic and germinal muta- Mutagens: Principles and Methods for their genesis under proﬁ cient repair conditions. Detection, vol 10. Plenum Press, New York, Mutat Res 260:247–256 pp 1–59 15. Rodriguez-Arnaiz R, Vogel EW, Szakmary A 2. Graf U, Würgler FE, Katz AJ et al (1984) (1993) Strong intra-species variability in the Somatic mutation and recombination test in metabolic conversion of six procarcinogens to Drosophila melanogaster . Environ Mutagen somatic cell recombinagens in Drosophila. 6:153–188 Mutagenesis 8:543–551 3. Vogel EW, Zijlstra JA (1987) Mechanistic and 16. Aguirrezabalaga I, Santamaría I, Comendador methodological aspects of chemically-induced MA (1994) The w/w+ SMART is a useful tool somatic mutation and recombination in for the evaluation of pesticides. Mutagenesis Drosophila melanogaster . Mutat Res 182: 9(4):341–346 243–264 17. Cunha KS, Campesato VR, Reguly ML et al 4. Graf U, Frei H, Kägi A et al (1989) Thirty (1995) Tannic acid is not mutagenic in germ cells compounds tested in the Drosophila wing spot but weakly genotoxic in somatic cells of Drosophila test. Mutat Res 222:359–373 melanogaster . Mutagenesis 10:291–295 5. Vogel EW, Nivard MJM (1993) Performance 18. Gaivão I, Sierra LM, Comendador MA (1999) of 181 chemicals in a Drosophila assay pre- The w/w+ SMART assay of Drosophila mela- dominantly monitoring interchromosomal nogaster detects the genotoxic effects of reac- mitotic recombination. Mutagenesis 8:57–81 tive oxygen species inducing compounds. 6. Sengstag C (1994) The role of mitotic recom- Mutat Res 440:139–145 bination in carcinogenesis. Crit Rev Toxicol 19. Kaya B, Yanikoglu A, Marcos R (1999) 24:323–353 Genotoxicity studies on the phenoxyacetates 7. Guzmán-Rincón J, Delfín-Loya A, Ureña- 2-4-D and 4-CPA in the Drosophila wing spot Núñez F et al (2005) Genotoxicity of neutrons test. Teratog Carcinog Mutagen 19:305–312 in Drosophila melanogaster. Somatic mutation 20. Costa WF, de Oliveira AB, Nepomuceno JC and recombination induced by reactor neu- (2010) Genotoxicity of lapachol evaluated by trons. Radiat Res 164:157–162 wing spot test of Drosophila melanogaster. 8. Kaya B, Kocaoğlu S, Demir E (2006) Analysis Genet Mol Biol 33:558–563 of UV-stimulated recombination in the 21. Téllez MG, Rodríguez HB, Olivares GQ et al Drosophila SMART assay. Environ Mol (2007) A phytotherapeutic extract of Mutagen 47:639, Erratum Equisetum myriochaetum is not genotoxic 9. Demir E, Kocaoğlu S, Kaya B et al (2010) either in the in vivo wing somatic test of Induction of adaptive response in Drosophila Drosophila or in the in vitro human micronu- after exposure to low doses of UVB. Int J cleus test. J Ethnopharmacol 111:182–189 Radiat Biol 86:957–963 22. Toledo VM, Tellez MG, Sortibrán AN et al 10. Graf U, Heo OS, Ramirez OO (1992) The (2008) Genotoxicity testing of Cecropia obtu- genotoxicity of chromium (IV) oxide in the sifolia extracts in two in vivo assays: the wing wing spot test of Drosophila melanogaster is somatic mutation and recombination test of over 90 % due to mitotic recombination. Mutat Drosophila and the human cytokinesis-block Res 266:197–203 micronucleus test. J Ethnopharmacol 116: 11. Ramos-Morales P, Rodríguez-Arnaiz R (1995) 58–63 Genotoxicity of two arsenic compounds in 23. Cunha KS, Reguly ML, Graf U et al (2001) germ cells and somatic cells of Drosophila Taxanes: the genetic toxicity of paclitaxel and melanogaster. Environ Mol Mutagen 25: docetaxel in somatic cells of Drosophila mela- 288–299 nogaster. Mutagenesis 16:79–84 12. Rizki M, Kossatz E, Creus A et al (2004) 24. Lehmann M, Vilar Kde S, Franco A et al (2004) Genotoxicity modulation by cadmium treat- Activity of topoisomerase inhibitors daunoru- ment: studies in the Drosophila wing spot test. bicin, idarubicin, and aclarubicin in the Environ Mol Mutagen 43:196–203 Drosophila Somatic Mutation and 13. Carmona ER, Creus A, Marcos R (2011) Recombination Test. Environ Mol Mutagen Genotoxic effects of two nickel-compounds in 43:250–257 somatic cells of Drosophila melanogaster. 25. Yüksel M, Sarıkaya R, Bostanci N (2010) Mutat Res 718:33–37 Genotoxic evaluation of antiepileptic drugs by The Drosophila Wing and Eye SMART Assays 295

Drosophila somatic mutation and recombination 38. Kaya B, Creus A, Velázquez A et al (2002) test. Food Chem Toxicol 48:2682–2687 Genotoxicity is modulated by ascorbic acid. 26. García Sar D, Aguado L, Montes Bayón M et al Studies using the wing spot test in Drosophila. (2012) Relationships between cisplatin- Mutat Res 520:93–101 induced adducts and DNA strand-breaks, 39. Taira K, Miyashita Y, Okamoto K et al (2005) mutation and recombination in vivo in somatic Novel antimutagenic factors derived from the cells of Drosophila melanogaster, under differ- edible mushroom Agrocybe cylindracea. Mutat ent conditions of nucleotide excision repair. Res 586:115–123 Mutat Res 741(1–2):81–88 40. Anter J, Campos-Sánchez J, Hamss RE et al 27. Kounatidis I, Papoti VT, Nenadis N et al (2009) (2010) Modulation of genotoxicity by extra- Evaluation of potential genotoxicity of virgin virgin olive oil and some of its distinctive com- olive oil (VOO) using the Drosophila wing-spot ponents assessed by use of the Drosophila test. J Agric Food Chem 57:7785–7789 wing-spot test. Mutat Res 703:137–142 28. Sotibrán AN, Ordaz-Téllez MG, Rodríguez- 41. Felício LP, Silva EM, Ribeiro V et al (2011) Arnaiz R (2011) Flavonoids and oxidative Mutagenic potential and modulatory effects of stress in Drosophila melanogaster. Mutat Res the medicinal plant Luehea divaricata 726:60–65 (Malvaceae) in somatic cells of Drosophila 29. Demir E, Marcos R, Kaya B (2012) Genotoxicity melanogaster: SMART/wing. Genet Mol Res studies in the ST cross of the Drosophila wing 10:16–24 spot test of sunﬂ ower and soybean oils before 42. Vlastos D, Mademtzoglou D, Drosopoulou E and after frying and boiling procedures. Food et al (2013) Evaluation of the genotoxic and Chem Toxicol 50:3619–3624 antigenotoxic effects of Chios mastic water by 30. Torres C, Ribas G, Xamena N et al (1992) the in vitro micronucleus test on human lym- Genotoxicity of four herbicides in the Drosophila phocytes and the in vivo wing somatic test of wing spot test. Mutat Res 280:291–295 Drosophila. PLoS One 8:e69494 31. Kaya B, Marcos R, Yanikoğlu A et al (2004) 43. Lindsley DL, Zimm GG (1992) The Genome Evaluation of the genotoxicity of four herbi- of Drosophila melanogaster . Academic, San cides in the wing spot test of Drosophila mela- Diego, CA nogaster using two different strains. Mutat Res 44. Vogel EW, Nivard MJM (2000) Parallel moni- 557:53–62 toring of mitotic recombination, clastogenicity 32. Fragiorge EJ, Rezende AA, Graf U et al (2008) and teratogenic effects in eye tissue of Comparative genotoxicity evaluation of imid- Drosophila. Mutat Res 455:141–153 azolinone herbicides in somatic cells of 45. Frölich A, Würgler FE (1989) New tester Drosophila melanogaster. Food Chem Toxicol strains with improved bioactivation capacity for 46:393–401 the Drosophila wing-spot test. Mutat Res 33. Rodrigues F, Lehmann M, do Amaral VS 216:179–187 (2007) Genotoxicity of three mouthwash 46. Gaivão I, Comendador MA (1996) The w/w+ products, Cepacol, Periogard, and Plax, in the somatic mutation and recombination test Drosophila wing-spot test. Environ Mol (SMART) of Drosophila melanogaster for Mutagen 48:644–649 detecting reactive oxygen species: characteriza- 34. Dihl RR, da Silva CG, do Amaral VS et al tion of 6 strains. Mutat Res 360:145–151 (2008) Mutagenic and recombinagenic activity 47. Kastenbaum MA, Bowman KO (1970) Tables of airborne particulates, PM10 and TSP, for determining the statistical signiﬁ cance of organic extracts in the Drosophila wing-spot mutation frequencies. Mutat Res 9:527–549 test. Environ Pollut 151:47–52 48. Frei H, Würgler FE (1988) Statistical methods 35. García-Quispes WA, Carmona ER, Creus A to decide whether mutagenicity test data from et al (2009) Genotoxic evaluation of two halo- Drosophila assays indicate a positive, negative, nitromethane disinfection by-products in the or inconclusive result. Mutat Res 203:297–308 Drosophila wing-spot test. Chemosphere 75: 49. Alonso-Moraga A, Graf U (1989) Genotoxicity 906–909 testing of antiparasitic nitrofurans in the 36. Graf U, Abraham SK, Guzmán-Rincón J et al Drosophila wing somatic mutation and recom- (1998) Antigenotoxicity studies in Drosophila bination test. Mutagenesis 4:105–110 melanogaster. Mutat Res 402:203–209 50. Ferreiro JA, Sierra LM, Comendador MA 37. Rizki M, Amrani S, Creus A et al (2001) (1995) Methodological aspects of the white- Antigenotoxic properties of selenium: studies ivory system assay of Drosophila melanogaster in the wing spot test in Drosophila. Environ in relation with genotoxicity testing. Mutat Res Mol Mutagen 37:70–75 335:151–161 Chapter 17

Testing the Genotoxic Potential of Nanomaterials Using Drosophila

Mohamed A. Abdalaziz , Balasubramanyam Annangi , and Ricard Marcos

Abstract

Nanogenotoxicology is an emergent area of research aiming to determine the potential risk of nanomaterials. Since most of the established studies use in vitro approaches, neglecting the repair and metabolic properties of the whole organism, some doubts about the accuracy of the obtained results exist. To overcome this gap more in vivo approaches testing the potential genotoxic risk of nanomaterials are required. In this context we propose to use Drosophila melanogaster as a useful model to study the possible genotoxic risk associated to nanoparticles exposure. Until now, only few studies have been carried out and they all use the wing-spot assay that detects the induction of somatic mutation and recombination events in the wing imaginal disks. This test is based on the principle that the loss of heterozygosis and the corresponding expression of the suitable recessive markers, multiple wing hairs and ﬂ are-3 , can lead to the formation of mutant clone cells in growing up larvae, which are expressed as mutant spots on the wings of adult ﬂ ies. The protocol to perform the wing-spot assay is presented.

Key words Drosophila , Wing-spot test , Nanomaterials , Genotoxicity

1 Introduction

Nanoparticles (NPs) are engineered structures or materials characterized by its size that must be less than 100 nm in at least one dimension. They present unique physicochemical features such as small size, large surface area, unique shape, and high mechanical, thermal, and electrical properties [ 1 ]. There is a rising debate concerning the possible harmful effects derived from the use of NPs, and due to their growing use, the risks associated with their exposure, routes of entry, and molecular mechanisms of cytotoxicity ought to be well deﬁ ned [2 ]. So far, there are considerable toxicological studies that have addressed the effects of NPs in different systems and environments, which raise concerns about the adverse effects on biological systems [ 3 – 6 ]. These studies suggest that NPs are not inherently benign and they

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_17, © Springer Science+Business Media New York 2014 297 298 Mohamed A. Abdalaziz et al.

affect biological processes at cellular, subcellular, and molecular platforms [ 7 ]. It led to the birth of one of the most important subdisciplines of nanotechnology known as nanotoxicology that is used for the research requirements to identify the real risks that comes with exposure to the engineered NPs [ 8 , 9 ]. Several attempts have been made to study the toxicity/genotoxicity of nanomaterials with the already known toxicological methods, since these traditional methods may be applicable to study the toxicity of NPs as well. Nevertheless, due to the unique properties of NPs, some of the methods may fail to measure its toxicity. Although there are many in vitro studies dealing with the toxic/genotoxic properties of nanomaterials, more in vivo studies are required due to their relevance in terms of health risk. In this context the use of Drosophila as a model can be a very useful alternative in front of the mammalian models, due to its objective advantages. Thus, Drosophila presents a short lifespan (10–14 days from embryo to adulthood), it is easy to manipulate, and it has a good cost-effective ratio. In addition, Drosophila shares some 60–70 % homology to human genes including those that encode molecules essential for carcinogenesis, pigmentation, and the ner- vous system [10 ]. Moreover, counterparts of several genes responsible for more than 700 different human genetic diseases including neurological, immunological, cardiovascular, auditory, visual, devel- opmental, and metabolic disorders are found in Drosophila [11 – 14 ]. Till recently, most of the studies carried out with this model organism indicate that Drosophila is amenable for studying a wide range of biological effects of NPs, particularly sensitive to detect their size effects. Many studies have been carried out with Drosophila showing the toxic effects of different nanomaterials using different biological end points. Thus, silica NPs are internalized in the midgut cells increasing the production of reactive oxygen species (ROS) and producing apoptosis [15 ]. It was shown that gold NPs could enter the fat body increasing lipid levels without triggering stress responses [16 ]; in addition it increased the expression of genes related to stress response, DNA damage recognition, and apoptosis pathway [ 17 ]. The effects of silver NPs on the development of Drosophila from eggs into pupae and adult are size dependent since 20–30 nm size effects are less harmful than 100 nm and 500– 1,200 nm sizes of silver NPs [ 18 ]. Titanium dioxide NPs induced oxidative stress mediated by the expression of superoxide dismutase and glutathione genes [ 19 ]. Finally, with respect to the effects of carbon nanotubes, its exposure does not induce detectable larval effects but in adults it impairs locomotor activity [20 ]. The studies evaluating genotoxicity of nanomaterials in Drosophila are scarce and, with the exception of one work dealing with the effects of CdSe-ZnS quantum dots inducing ROS and apoptosis in larval hemocytes [ 21 ], the other studies have been Testing Nanomaterials Using Drosophila 299

carried out using the wing SMART assay. The wing-spot assay is a simple test that measures the signiﬁ cant increase in the frequency of mutant spots ( mwh or ﬂ r ) on the wings to detect the possible genotoxic potential risk of NPs in Drosophila (see chapter 16 in this book). Until now this assay has been used to demonstrate the genotoxicity of silver NPs [ 22 ], cobalt NPs [23 ], and titanium, zirconium, and aluminum NPs [ 24 ] after larval feeding. No genotoxic effects have been detected when multiwalled carbon nanotubes were evaluated [25 ]. Due to the demonstrated usefulness of the wing SMART assay in determining the genotoxic potential of nanomaterials, we reported here the development of this assay.

2 Materials and Reagents

● Drosophila strains ● Culture chamber ● Glass bottles (125 mL) ● Disposable plastic vials (50 mL) ● Culture medium ● Faure’ solution ● Eppendorf tubes 1.5 mL ● Fine tweezers ● Sieve (of about 500 μm pore) ● Microscope slides ● Coverslips (24 × 60 mm) ● Microscope ● Freezer (−20 ºC) ● Refrigerator (4 ºC)

2.1 Drosophila For the development of the wing-spot assay, two D. melanogaster Strains strains are needed: the mwh and the fl r strains. Both strains carry mutations that affect the phenotype of the trichomes (hairs) of the wings cells [ 26 ]. Thus, the phenotypic expression of the mwh mutation is characterized by the appearance of three or more hairs in each cell, instead of one per cell, which is the normal phenotype. The mwh mutation is recessive and viable in homozygosis, and it is located on chromosome 3. On the other hand, the fl are phenotype is quite variable, ranging from short, thick, and deformed to amorphous-like globular hairs. The fl r 3 mutation is located on the same arm of chromosome 3. It is recessive and produces lethality in homozygosis in the germ line, but not in somatic cells. For this reason, to maintain the strain it is necessary to balance the 300 Mohamed A. Abdalaziz et al.

Fig. 1 Phenotype morphology of mwh and ﬂ r wing-spot markers. (a , b ) show electronic and light microscopy view of a mwh clone, respectively. (c , d ) show electronic and light microscopy view of a ﬂ r clone, respectively

chromosome carrying the ﬂ r mutation with a chromosome carrying overlapped inversions that avoid recombination ( 3LR and TM3 ). In addition, a dominant mutation with a phenotype given a swayed border wing ( Bd S ) is included in the balanced chromosome to differentiate both chromosomes. The phenotype of the wing markers is visible in Fig. 1 To get more details of the other genetic markers and on the phenotype descriptions of the employed strains, see Lindsley and Zimm [27 ]. These strains can be easily obtained by request to any of the groups working with this assay, such as the group presenting this chapter.

2.2 Culture For maintaining both strains, as well for growing treated larvae, Conditions they are grown in a culture chamber at 25 ± 1 °C, with a relative humidity of ~60 % and a light/dark cycle of 12/12 h.

2.2.1 Culture Medium To 1,200 mL of cold water add 170 g of corn ﬂ our, 120 g of for Growing yeast, 2 g of NaCl, and 10 g of agar-agar. Apply heat to the mix and Maintaining until boliling point, remove from heat and cool the mix, and after the Strains that, a mixture of 4 mL of Nipagin fungicide, dissolved in 10 % ethanol and 4 mL of propionic acid, is added. At this point the medium is ready to be used. Testing Nanomaterials Using Drosophila 301

2.2.2 Medium To carry out the different larval treatments with NPs, a dehy- for Treatments drated growth medium designed as Carolina Drosophila Medium Formula 4-24 ® (Carolina Biological Supply Company, USA) is used. Thus, 3.5 g is rehydrated with 10 mL of distilled water (or NP dispersions).

2.2.3 Dispersion of NPs To be sure that the treatments with NPs are applied at nanosized and Treatments level, avoiding large agglomeration formation, a correct dispersion must be obtained. Thus, NPs are weighted to obtain the amount necessary for the stock concentration, pre-wetted in 0.5 % absolute ethanol, and mixed with a concentration of 10 mM in 0.05 % bovine serum albumin (BSA) in double distilled water. The dispersion solution is submitted to 30 min of sonication. At this point the stock dispersion is ready to be used. It is recommended to carry out the different concentrations no more than 2 h after sonication.

2.2.4 Faure’s Solution In 50 mL of distilled water, 30 mL of gum arabic and 50 g of chloral hydrate are mixed, heated without boiling and chilled for 20 min. Add 20 mL of glycerol and mix well. Once the Faure’ solution is prepared, it must be transferred into an amber glass bottle with dropper.

3 Speciﬁ c Procedures

3.1 Obtaining The assay use transheterozygous individuals, with (mwh +/+ ﬂ r 3 ) Transheterozygous genotype, to detect both somatic mutation and recombination. Larvae for Treatments These individuals result from the cross between ﬂ r 3 virgin females and mwh males (see Fig. 2 ), although the reciprocal cross can also be used. The cross is carried out in 125 mL bottles containing standard food medium, and the resulting eggs are collected during

+flr3 + mwh + +

X ++BdS mwh + +

mwh + + mwh + +

+ flr3 + ++BdS

Transheterozigous Balanced heterozigous

Fig. 2 Offspring resulting from a cross between mwh females and ﬂ r males: transheterozygous and balanced heterozygous individuals 302 Mohamed A. Abdalaziz et al.

8-h periods. The resulting 72 ± 4 h larvae (third instar) are collected, by washing the culture bottles softly with tap water, ﬁ ltered through a sieve of about 500 μm to separate larvae from the standard culture media rests, and placed in disposable plastic vials containing 3.5 g of lyophilized instant Drosophila medium hydrated with 10 mL of the NP dispersion. The treatment vials are kept in an incubator at 25 °C and 60 % humidity until the emergence of adult individuals.

3.2 Preparation The emerging adults from the NP treatment vials are collected the and Mounting day of its emergence, classiﬁ ed according their phenotype, and of Drosophila Wing stored in 1.5 mL plastic vials (Eppendorf) with 70 % ethanol at Slides cold temperature (4 ºC) until their use. Since two different types of offspring are obtained (transheterozygous and balanced heterozygous), both types of individuals must be kept separately. Before removing the wings the ﬂ ies are washed with a mix of

ethanol and d-H 2 O into a watch glass. After that, individuals are placed on a well slide, with a drop of Faure’s solution, to proceed with the extraction of its wings using ﬁ ne forceps under a binocular stereo microscope. Wings are carefully taken with ﬁ ne tweezers and placed in pairs and aligned on a clean microscopic slide. A few drops of Faure’s solution are placed in the center of the slide that is covered with a coverslip of 24 × 60 mm, to obtain a permanent preparation.

3.3 Microscopic The presence of mutant clones on each of the two wing blades is Analysis of the Wings determined at 400 times magnifi cation under a light microscope. Slides with wings from transheterozygous individuals are scored fi rst. Only if results are positive, the serrate set of wings is scored to provide further information on the underlying mechanisms producing genotoxicity. Thus, since balanced heterozygous (mwh +/+ Bd S ) individuals do not recombine, due to the presence of the TM3 inversion, the observed differences between genotypes must be attributed to somatic recombination. Between 40 to 80 wings are scored per dose of the NP treatment and the observed mutant spots are recorded according to their position in each area of the wing (see Fig. 3 ). Spots are classifi ed as (a) small single mwh spots (1–2 cells), (b) large single mwh spots (3 or more cells), (c) twin (adjacent mwh and fl are ) spots, and (d) single fl r 3 spots.

3.4 Statistical The recorded data from the scoring sheets must be translated into Analysis a tabular form and properly analyzed from a statistical point of of the Recorded Data view. To carry out the statistical analysis, the frequency of each type of mutant spot observed in the different treatments with the selected NPs is compared with the values observed in the negative control. This comparison is carried out by using the conditional binomial test of Kastenbaum and Bowman [ 28 ] with a signiﬁ cance level of α = β = 0.05. To determine the overall response of an agent Testing Nanomaterials Using Drosophila 303

Fig. 3 Scheme showing the different sectors from a Drosophila wing used for the scoring of mutant clones

as positive, weakly positive, negative, or inconclusive, the multiple- decision procedure described by Frei and Würgler [ 29 ] is used. The frequency of clone formation is calculated, without size correction, by dividing the number of mwh clones per wings by 24,400, which is the approximate number of cells analyzed per wing [30 ].

Acknowledgments

The authors thank the ﬁ nancial support of their research activity by CIRIT (project 2009SGR-725). BA was supported by a postdoctoral fellowship from the Universitat Autònoma de Barcelona (UAB). MAA was supported by a predoctoral fellowship from the Egyptian Government.

References

1. Fubini B, Ghiazza M, Fenoglio I (2010) Physico- stress in the mouse brain after exposure to silver chemical features of engineered nanoparticles nanoparticles. Toxicol Lett 187:15–21 relevant to their toxicity. Nanotoxicology 6. Pan X, Redding JE, Wiley PA et al (2010) 4:347–363 Mutagenicity evaluation of metal oxide 2. Medina C, Santos-Martinez MJ, Radomski A nanoparticles by the bacterial reverse mutation et al (2007) Nanoparticles: pharmacological assay. Chemosphere 79:113–116 and toxicological signiﬁ cance. Br J Pharmacol 7. Nel A, Xia T, Madler L et al (2006) Toxic 150:552–558 potential of materials at the nanolevel. Science 3. Oberdorster G, Oberdorster E, Oberdorster J 311:622–627 (2005) Nanotoxicology: an emerging disci- 8. Donaldson K, Stone V, Tran CL et al (2004) pline evolving from studies of ultraﬁ ne parti- Nanotoxicology. Occup Environ Med 61: cles. Environ Health Perspect 113:823–839 727–728 4. Casey A, Herzog E, Lyng FM et al (2008) 9. Colvin VL (2003) The potential environmen- Single walled carbon nanotubes induce indirect tal impact of engineered nanomaterials. Nat cytotoxicity by medium depletion in A549 Biotechnol 21:1166–1170 lung cells. Toxicol Lett 179:78–84 10. Rand MD (2010) Drosophotoxicology: the 5. Rahman MF, Wang J, Patterson TA et al growing potential for Drosophila in neurotoxi- (2009) Expression of genes related to oxidative cology. Neurotoxicol Teratol 32:74–83 304 Mohamed A. Abdalaziz et al.

11. Reiter LT, Potocki L, Chien S et al (2001) A larval dietary uptake is benign, but adult systematic analysis of human disease-associated exposure causes locomotor impairment and gene sequences in Drosophila melanogaster . mortality. Environ Sci Technol 43:6357–6363 Genome Res 11:1114–1125 21. Galeone A, Vecchio G, Malvindi MA et al 12. Koh K, Evans JM, Hendricks JC et al (2006) (2012) In vivo assessment of CdSe-ZnS quan- A Drosophila model for age associated changes tum dots: coating dependent bioaccumulation in sleep: wake cycles. Proc Natl Acad Sci U S A and genotoxicity. Nanoscale 4:6401–6407 103:13843–13847 22. Demir E, Vales G, Kaya B et al (2011) 13. Wolf MJ, Amrein H, Izatt JA et al (2006) Genotoxic analysis of silver nanoparticles in Drosophila as a model for the identiﬁ cation of Drosophila . Nanotoxicology 5:417–424 genes causing adult human heart disease. Proc 23. Vales G, Demir E, Kaya B et al (2013) Natl Acad Sci U S A 103:1394–1399 Genotoxicity of cobalt nanoparticles and ions 14. Khurana V, Lu Y, Steinhilb ML et al (2006) in Drosophila . Nanotoxicology 7:462–468 TOR-mediated cell-cycle activation causes 24. Demir E, Turna F, Vales G et al (2013) In vivo neurodegeneration in a Drosophila tauopathy genotoxicity assessment of titanium, zirco- model. Curr Biol 16:230–241 nium and aluminium nanoparticles, and their 15. Pandey A, Chandra S, Chauhan LK et al (2013) microparticulated forms, in Drosophila . Cellular internalization and stress response of Chemosphere 93:2304–2310 ingested amorphous silica nanoparticles in the 25. de Andrade LR, Brito AS, Melero AM et al midgut of Drosophila melanogaster . Biochim (2014) Absence of mutagenic and recombina- Biophys Acta 1830:2256–2266 genic activity of multi-walled carbon nanotubes 16. Wang B, Chen N, Wei Y et al (2012) Akt in the Drosophila wing-spot test and Allium signaling-associated metabolic effects of dietary cepa test. Ecotoxicol Environ Saf 99:92–97 gold nanoparticles in Drosophila. Sci Rep 26. Graf U, Würgler FE, Katz AJ et al (1984) 2:563 Somatic mutation and recombination test in 17. Vecchio G, Galeone A, Brunetti V et al (2012) Drosophila melanogaster . Environ Mutagen 6: Concentration-dependent, size-independent 153–188 toxicity of citrate capped AuNPs in Drosophila 27. Lindsley DL, Zimm GG (1992) The Genome of melanogaster . PLoS One 7:e29980 Drosophila melanogaster . Academic, San Diego 18. Gorth DJ, Rand DM, Webster TJ (2011) Silver 28. Kastenbaum MA, Bowman KO (1970) Tables nanoparticle toxicity in Drosophila : size does for determining the statistical signiﬁ cance of matter. Int J Nanomedicine 6:343–350 mutation frequencies. Mutat Res 9:527–549 19. Posgai R, Cipolla-McCulloch CB, Murphy KR 29. Frei H, Würgler FE (1988) Statistical methods et al (2011) Differential toxicity of silver and to decide whether mutagenicity test data from titanium dioxide nanoparticles on Drosophila Drosophila assays indicate a positive, negative, or melanogaster development, reproductive effort, inconclusive result. Mutat Res 203:297–308 and viability: size, coatings and antioxidants 30. Alonso-Moraga A, Graf U (1989) Genotoxicity matter. Chemosphere 85:34–42 testing of antiparasitic nitrofurans in the 20. Liu X, Vinson D, Abt D et al (2009) Differential Drosophila wing somatic mutation and recom- toxicity of carbon nanomaterials in Drosophila : bination test. Mutagenesis 4:105–110 Chapter 18

Transgenic Rodent Gene Mutation Assay in Somatic Tissues

John D. Gingerich , Lynda Soper , Christine L. Lemieux , Francesco Marchetti , and George R. Douglas

Abstract

Some chemicals found in the environment may cause DNA damage. The mutations which may result from this damage can cause various diseases including cancer. It is important to have methods available to test chemicals to which humans may be exposed for their mutagenic potential. In 2011 the Organization for Economic Cooperation and Development adopted a test guideline on the use of transgenic rodent (TGR) assays for investigating the mutagenic potential of chemical agents. The TGRs used in these assays carry a transgene consisting of multiple copies of a bacterial gene which is incorporated into the genome and thus resides in every cell of every tissue. These transgenes have no effect on the animal but are easily recovered and tested for DNA mutations. These TGR assays have an advantage over bacterial and in vitro assays in that the exposure to the test agent occurs within a live animal with all of the various metabolic and DNA repair processes in place, thus more closely modeling actual human exposure. The possibility to investigate tissue speciﬁ city by examining DNA from various tissues in the same animal adds to the value of the assay. Herein we describe the use of the Muta™Mouse transgenic mouse model for determining the mutagenic potential of chemical agents.

Key words Transgenic , Rodent , Gene mutation , lacZ , DNA isolation

1 Introduction

DNA mutations may be caused by exposure to chemicals in the environment and can result in various human diseases including cancer. This means that it is important to have available methods for measuring the potential DNA-damaging properties of chemicals to which we may be exposed. Many in vitro tests such as the Ames assay [ 1 ] are commonly used for this purpose. These tests provide a rapid method of screening chemicals for their ability to cause genetic damage. However, it is frequently necessary to add mammalian enzymes to these test systems to more accurately reﬂ ect the metabolic processes that occur in mammalian cells, and this can greatly affect the outcome of exposure to various chemicals.

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_18, © Springer Science+Business Media New York 2014 305 306 John D. Gingerich et al.

In vivo mammalian assays have the obvious advantage of possessing the intrinsic mammalian metabolic repair capacity needed for the effective determination of the result of exposure to mutagens. However, it has been diffi cult to measure the induction of mutations in vivo in endogenous genes from mammalian cells and tissues since there is no easy way to recover such genes and measure their mutation frequencies. This diffi culty has been overcome with the development of animal models harboring transgenic reporter genes, which are carried on vectors designed to be easily recovered, allowing for rapid assessment of mutant frequency. Current evidence suggests that these reporter genes respond to mutagens in the same way as endogenous genes [2 , 3 ]. This technology makes it possible to screen for changes in mutant frequency in virtually any tissue following in vivo exposure to agents of interest. In 2011 the Organization for Economic Cooperation and Development (OECD) adopted a test guideline on transgenic rodent somatic and germ cell gene mutation assays (updated in 2013) [ 4 ] which includes the following model systems: Muta™Mouse [5 , 6 ], lacZ plasmid mouse [7 ], Big Blue® mouse and rat [6 , 8 ], and gpt delta mouse and rat [ 9 ]. Here we describe the use of Muta™Mouse to measure changes in mutant frequency in various somatic tissues following in vivo exposure to mutagens. The application of transgenic rodent (TGR) models to investigate the induction of mutations in germ cells involves many considerations both in terms of the timing for sample collection and in the methods used to isolate DNA. This topic has been covered elsewhere [10 ]. A typical study to assess mutagenic effects in somatic tissues involves the steps illustrated in Fig. 1 . Animals are treated with the agent of interest via an appropriate route. After an appropriate expression time, during which mutations are allowed to become fi xed, tissues are collected and frozen for later DNA isolation and analysis. The reporter vector is recovered and the mutant frequency is determined by counting the number of mutant plaques recovered under selection conditions as compared to the number recovered in the absence of selection. Although the methods described here for the recovery of genomic DNA from tissues should be broadly applicable to other TGR assay models, the testing details will necessarily be different [ 11 ]. While not covered here, it is possible to pick mutant plaques for DNA sequencing in order to study the spectrum of mutations induced by the study agent [6 , 12 ].

2 Materials

Aqueous solutions and buffers stored for more than 24 h should be kept sterile. Stock solutions for buffers are autoclaved and then stored at room temperature for up to 6 months, unless otherwise speciﬁ ed. Gene Mutation Determination Using Muta™Mouse 307

Fig. 1 Overview of lacZ selective assay using Muta™Mouse and E. coli C, lacZ −, galE − host. (a ) Animals are treated daily for 28 days in accordance with the OECD test guideline. (b ) Tissues are recovered from the animals 3 days following the last exposure and frozen for storage. (c ) Genomic DNA is isolated from the frozen tissues. The transgene locus in Muta™Mouse consists of mostly concatenated copies of lambda bacteriophage carrying a lacZ gene [21 ]. (d ) Viable phages are recovered from the genomic DNA preparation using Transpack packaging extracts (Agilent Technologies). The extracts provide the enzymes which cut the transgene at intact lambda cos sites and assemble viable phage using capsules, also supplied by the extracts, in the same way as would normally occur during phage propagation. (e ) The phage produced by packaging are used to infect the E. coli C, lacZ − , galE − host cells. (f ) Following adsorption, the infected cells are mixed with top agar and plated on minimal agar plates with and without the selective agent, phenyl-β- D-galactopyranoside (p-gal). In the absence of p-gal, there is no source of galactose. Accordingly, all phage infections, whether wild type or mutant, result in phage propagation with eventual lysis of infected cells and plaque formation. However, when the host is infected with wild-type lacZ + phage in the presence of p-gal, the cells cleave p-gal with the aid of the phage-borne lacZ gene producing galactose. Because the bacterial host cells lack a functional galE gene, they are unable to complete the conversion of galactose to UDP-glucose resulting in the accumulation of toxic UDP-galactose. This causes early cell death in which case there is no phage propagation and thus no plaque formation. Infection of the E. coli host cells with a phage carrying a mutant (nonfunctional) lacZ gene has no effect on the host’s inability to metabolize p-gal. Thus, no toxic product is produced and phage propagation continues resulting in cell lysis and plaque formation. (g ) After overnight incubation at 37 °C, plaques are scored and the data used to determine the mutant frequency. Illustration by Sheena Gingerich, used with permission (color ﬁ gure online)

2.1 Exposure Vehicle The choice of vehicle is dictated by the chemical characteristics of the compound being tested. Generally, in our laboratory, olive oil is used for non-water-soluble compounds and phosphate buffer is used otherwise.

● Olive oil (Sigma)

● 0.067 M Na2 HPO4

● 0.067 M KH2 PO4 308 John D. Gingerich et al.

Note : Solutions of various pH can be prepared by combining the two phosphate solutions in different proportions. For example, 1.3 mL of

the Na 2 HPO 4 solution plus 8.7 mL of the KH 2 PO 4 solution give 10 mL of phosphate buffer at a pH of 6.0 which is used for dissolving ENU.

2.2 Isolation ● PBS: phosphate buffered saline without magnesium (Invitrogen) Reagents ● 1 M Tris–HCl, pH 7.6 ● 0.5 M EDTA ● 5 M NaCl ● TMST buffer (per L): 50 mL 1 M Tris–HCl, pH 7.6 (50 mM ﬁ nal), 0.64 g Mg acetate (3 mM), 20 g sucrose (250 mM), 2 mL Triton X-100 (0.2 %). Autoclave to sterilize and store at 4 °C for up to 1 month ● Lysis buffer (per 100 mL): 1 mL 1 M Tris, pH 7.6 (10 mM), 2 mL 0.5 M EDTA (10 mM), 2 mL 5 M NaCl (100 mM),

100 mg proteinase K (1 mg/mL), H2 O to 100 mL (prepared just before use) ● 10 % SDS ● Phenol (Invitrogen) Note: Phenol is caustic. During organic extraction wear gloves and eye protection and work in a fume hood to avoid skin contact and breathing vapors.

● 1 M Tris–HCl, pH 8 ● Chloroform (Caledon Labs) Note : Chloroform is volatile and is a suspected human carcinogen. During organic extraction, wear gloves and eye protection and work in a fume hood to avoid skin contact and breathing vapors.

● Isoamyl alcohol (Caledon Labs) ● Proteinase K (Invitrogen) ● 1 M KCl ● RNase A (10 mg/mL) (Sigma) stored at −20 °C.

● − 4 TE7.6 buffer: 100 μL 1 M Tris–HCl, pH 7.6 (10 mM), 2 μL −4 0.5 M EDTA (10 M), H2 O to 10 mL

● 2.3 Test Reagents 1 M MgSO4 ● 20 % maltose ● Packaging extract: Transpack (Agilent Technologies Inc.) ● 2 % gelatin (autoclaved and then stored at 4 °C) ● SM buffer: 500 μL 1 M Tris–HCl, pH 7.6 (50 mM), 200 μL

5 M NaCl (100 mM), 0.16 mL 1 M MgSO4 (16 mM), 50 μL 2 % gelatin (0.01 %), H2 O to 10 mL Gene Mutation Determination Using Muta™Mouse 309

● LB powder: Lennox L Broth Base (Invitrogen) ● p-gal: phenyl-β-D -galactopyranoside (Sigma) ● Minimal agar recipe (/L): – LB powder 5.0 g – NaCl 6.4 g – Agar (Difco) 7.5 g ● Dimethylformamide (Caledon Labs) ● Host cells: E. coli C, ΔlacZ − , galE − , recA − , pAA119 [5 , 13 ]

3 Methods

3.1 Animal Exposure Muta™Mouse animals are available commercially. They are housed using a 12-h light/dark schedule, temperature at 22–25 °C, and 3.1.1 Animals humidity at 50–60 %. Food and water are supplied ad libitum. Animals are allowed to acclimatize for at least a week after arrival and before the start of a study. Male animals, 8–12 weeks of age at the start of the study, are assigned randomly to four groups: one vehicle control group and three dose groups. A fi fth group, used as a positive control treatment group, may be included, and a sixth untreated group may be added in cases where there are no data on the mutagenicity/toxicity of the vehicle used. Note : ENU is commonly used as a positive control in our laboratory. It is prepared in a pH 6 phosphate buffer no more than 2 h before it is used. The half-life of ENU is measured in minutes at pH 7 and in hours at pH 6 [ 14 ]. At least fi ve animals per group are used but this number may increase depending on the combination of the size of the change in mutant frequency that is to be detected and the desired power of the result. Note : A study employing fi ve animals per group would detect a doubling of mutant frequency with a power of 90 % for the bone marrow, 70 % for the liver, or 85 % for germ cells from the testis based on historical data in this lab. Unless toxicity data on the test agent administered via the chosen route are available, a preliminary study is done where animals are treated over a wide range of doses to determine the maximum tolerated dose (MTD). Note : Generally the highest dose tested is 1,000 mg/kg/day or the highest dose that can be dissolved or suspended in the vehicle if less than this. Usually groups of two animals per dose are suffi cient for MTD determination. The treatment duration should be the same as that proposed for the main study, which will be 28 daily doses when following the OECD test guideline [ 4 ]. These animals are closely 310 John D. Gingerich et al.

monitored for signs of toxicity and are removed from the study when toxicity becomes evident. Note : Signs of toxicity are determined in consultation with animal care personnel. Here the signs to watch for include marked lethargy, marked hypothermia, persistent dehydration (no response to rehydration after 72 h), self-mutilation, over 20 % loss in body weight, ulcers or open wounds, herniation, marked abdominal distention, severe diarrhea, and impaired mobility [15 ]. The MTD is the highest dose at which no overt signs of toxicity are observed. The OECD test guideline recommends that the three doses used for screening be the MTD, 2/3 of MTD, and 1/3 of MTD [ 4 ].

3.1.2 Exposures The test agent is prepared by dissolving or suspending it in the appropriate vehicle daily for labile compounds or as needed for agents known to be stable in the vehicle. The dose is administered at a rate of 0.005 mL/g of body weight, which determines the concentration of the test agent in the vehicle. A stock concentration of 1 mg/mL provides a dose of 5 mg/kg when administered at this rate. Note : The rate may be changed in consultation with animal care personnel. It may be possible to administer up to 0.01 mL/g body weight if necessary, depending on the route of administration. The route of exposure should be chosen with regard to the potential environmental exposure being studied. For example, oral gavage is often used to mimic the ingestion route of exposure and is preferred as the default route. Other possible routes of exposure would be inhalation (requires specialized facilities) or by addition to food or drinking water. The latter two suffer from the disadvantage that it is complicated to determine the dose delivered. Intraperitoneal injection (i.p.) is generally not recommended since it does not mimic most intended routes of exposure, but can be used for administering single exposure positive controls such as ethylnitrosourea.

3.2 DNA Preparation Following exposure and a 3-day ﬁ xation period, the animals are euthanized and the tissues of interest are collected and frozen until 3.2.1 Tissue Collection analysis. The method of euthanasia should be chosen in consultation with the local animal care committee and should be appropriate considering the tissues to be collected. For example, if blood is to be collected for a micronucleus or Pig-a assay (see, e.g., www.litronlabs. com ), euthanasia could be by anesthetic overdose to allow for the collection of blood via cardiac puncture. If only organs are being collected, cervical dislocation may be appropriate. Following euthanasia, tissues are excised and divided as appropriate (see below). Aliquots are then placed into cryovials and ﬂ ash frozen in liquid nitrogen followed by long-term storage at −80 °C. Gene Mutation Determination Using Muta™Mouse 311

The bone marrow is ﬁ rst ﬂ ushed from the femurs into a microtube using a 1-mL syringe with a 26-G needle and 1 mL of PBS. The sample is centrifuged in a microcentrifuge at 15,000g for 10 s. The supernatant is discarded and the pellet is resuspended in 100 μL of fresh PBS. The sample is then divided into two cryovials, 50 μL in each, and frozen as above. The liver is divided into two to four vials before freezing and the left and right lungs are usually divided between two vials. Paired organs such as the kidney, testes, etc. are stored individually or at least separated in the freezing vial so they can be recovered individually for DNA isolation.

3.2.2 Tissue Digestion Tissues are defrosted on ice for 1 h following their removal from −80 °C storage, after which they are digested, and DNA is puriﬁ ed and precipitated. Note : Care should be taken to use polypropylene centrifuge tubes with caps that seal tightly. Chloroform will dissolve polystyrene tubes!

Bone Marrow 1. The defrosted tissue, equal to the contents of one femur, is resuspended in 0.5 mL of lysis buffer in the cryovial then transferred to 4.0 mL of lysis buffer in a 15 mL centrifuge tube, and the suspension is mixed by inverting the tube. 2. 0.5 mL of 10 % SDS is added to give a ﬁ nal concentration of 1 % SDS. 3. The tube contents are mixed by gentle inversion and then left at 37 °C overnight with gentle agitation, if possible (e.g., 50 oscillations per minute (opm) on a reciprocal shaker) to digest.

Solid Tissues (Kidney, 1. Defrosted tissues are ﬁ rst minced in 1 mL of PBS using curved Spleen, Heart, Forestomach) scissors or crossed scalpel blades to reduce the tissue to 1–2 mm3 pieces. 2. The tissue is transferred to a 15 mL centrifuge tube and centrifuged at 800g , and the supernatant is decanted. 3. The tissue pellet is resuspended and digestion is initiated in the same way as for the bone marrow, Steps 1–3.

Lung 1. The defrosted tissue is minced as above (see Sect. 3.2.2.2 ) and then transferred to a 15 mL tube in 2–3 mL PBS. 2. The sample is subjected to a vacuum for 2–5 min to remove air from the tissue. This allows the tissue to sink in the buffer. 3. The sample is centrifuged, the supernatant decanted and the pellet resuspended in 100 μL of PBS. 4. Digestion is initiated in the same way as for the bone marrow, Steps 1–3. 312 John D. Gingerich et al.

Liver The liver is fi rst processed to collect cell nuclei before digestion is initiated. Approximately one-quarter of the liver from an animal is defrosted on ice. 1. The tissue is homogenized in 5 mL of ice-cold TMST buffer using a Dounce conical glass homogenizer with a Tefl on pestle. The pestle is rotated at about 60 rpm and the glass tube is raised and lowered six to ten times. Note : Care should be taken to avoid cavitation which would risk rupturing cell nuclei resulting in the loss of DNA during centrifugation and washing. 2. The sample is allowed to sit on ice for a few minutes to allow larger tissue clumps to settle. 3. Approximately 3 mL of the supernatant is slowly poured into a 15 mL centrifuge tube while taking care not to disturb the tissue clumps settled at the bottom of the tube. 4. An additional 2 mL of TMST is added to the tissue remaining in the Dounce tube, mixed, and again placed on ice to allow larger tissue pieces to settle before another 2 mL is transferred to the centrifuge tube containing the fi rst 3 mL. 5. The samples are centrifuged at 700g at 4 °C for 6 min. 6. The supernatant is decanted and the pellet resuspended in 0.5 mL TMST with the aid of a vortex mixer. An additional 4.5 mL TMST is added and the sample mixed using a vortex mixer. 7. The samples are centrifuged at 700g at 4 °C for 6 min and decanted. 8. Steps 7 and 8 are repeated once more, following which the pellets are resuspended, using a vortex mixer, in approximately 50 μL of supernatant which has been left in the tube. 9. The tissue pellet is resuspended and digestion is initiated in the same way as for the bone marrow, Steps 1–3.

Small Intestine and Colon The epithelial cells from the small intestine or colon are separated from the supporting tissue using a technique we have called “cracking.” Note : The term “crack” comes from the sound of cavitation which is unavoidable here. Try to keep the number of “cracks” to a minimum to reduce the amount of free DNA generated from ruptured cells. 1. Cracking buffer is prepared as follows. To 100 mL of water, add 0.1 mL of 1 M KCl (1 mM) and 2 mL of 0.5 M EDTA (10 mM) and cool on ice. 2. The organ is either inverted using two forceps or slit open lengthwise with scissors and placed in 1–2 mL cold cracking buffer in a petri dish on ice. Gene Mutation Determination Using Muta™Mouse 313

3. The tissue is sucked into and expelled from a 1 mL tuberculin syringe without a needle two to three times. 4. The buffer and released cells are discarded. This wash will remove residual intestinal contents and some of the surface cells. 5. An additional 2 mL of cracking buffer is added to the tissue which is then “cracked” four to eight times with the syringe. 6. The remaining intact tissue is discarded and the released cells are collected in a 15 mL centrifuge tube in a fi nal volume of about 4 mL of cracking buffer. 7. The cell suspension is centrifuged at 1,800g for 10 min after which the supernatant is discarded. Note : Some samples may contain a gelatinous-looking material which prevents the cells from forming a pellet. As much of the buffer as is possible is removed but the gelatinous material is retained. 8. The cell pellet is resuspended in 100 μL of cracking buffer. 9. Digestion is then initiated in the same way as for the bone marrow, Steps 1–3. Note : The following four steps are normally applied only to small intestine and colon tissues. They could also be used for other tissues if there are problems getting clean DNA preparations, but this is not normally necessary. 10. Following overnight digestion, RNase A is added to a fi nal concentration of 100 μg/mL (50 μL of a stock solution of 10 mg/mL for a 5 mL digest). 11. After 1 h at 37 °C, 0.4 volume of 5 M NaCl is added to give a fi nal NaCl concentration of about 1.5 M. 12. The sample is mixed and centrifuged at 2,000 g for 20 min to “salt out” and pellet proteins. 13. The aqueous sample (including any gelatinous “globs”) is transferred to a new 15 mL tube to begin DNA purifi cation using organic solvent extraction as described below (Sect. 3.2.3 ).

Glandular Stomach 1. The defrosted organ is cut open and rinsed clean with cold PBS. 2. In a petri dish on ice, the stomach mucosa is scraped off into 2 mL lysis buffer (without proteinase K) using a scalpel blade. 3. The released cells are transferred to a Dounce homogenizer and scraping of the stomach is repeated with an additional 3 mL of lysis buffer. 4. The cells released by the second scraping are added to those in the homogenizer. The resulting 5 mL of cell suspension is homogenized with three to four passes of the Teﬂ on pestle. 314 John D. Gingerich et al.

5. The sample is transferred to a 15 mL centrifuge tube and 0.5 mL of 10 % SDS along with 0.1 mL Rnase A stock solution at 5 mg/mL (0.1 mg/mL ﬁ nal) is added. 6. After incubating for 1 h at 37 °C and 70 opm in a reciprocal shaker, proteinase K is added to a ﬁ nal concentration of 1 mg/mL (0.25 mL of a 20 mg/mL stock solution). 7. Incubation is continued overnight at 37 °C with tubes in a horizontal position and gentle agitation at about 50 opm.

3.2.3 DNA Puriﬁ cation There are several methods available to purify genomic DNA including extraction with organic solvents or using commercially available DNA isolation kits. Here, we describe the use of phenol/ chloroform extraction, which is the method in use in our laboratory. 1. Previously hydrated phenol is equilibrated in polypropylene centrifuge tubes with an equal volume of 1 M Tris–HCl, pH 8, by shaking vigorously for 15 s. Note: Remember, phenol is caustic. During organic extraction wear gloves and eye protection and work in a fume hood to avoid skin contact and breathing vapors. Note: Remember, care should be taken to use polypropylene centrifuge tubes with caps that seal tightly. Chloroform will dissolve polystyrene tubes! 2. The phases are allowed to separate for 5 min and the upper aqueous phase is discarded. 3. Equilibration is repeated as above using 0.1 M Tris–HCl, pH 8.0, this time. Note : Equilibrated phenol can be stored for a week or 2 when protected from light at 4 °C. We prefer to equilibrate and use it on the same day. 4. An equal volume of chloroform: isoamyl alcohol 24:1, is added, mixed well, and then centrifuged for 10 min at 1,500 g to separate the phases. 5. The upper aqueous phase is discarded, and a volume of the lower organic phase, consisting of the equilibrated phenol/ chloroform, approximately equal to the sample volume, is added to each sample. 6. The samples are slowly inverted end-over-end at approximately 22 rpm for 20–30 min. Note: We use a rotator that allows a rack of tubes to be rotated end-over-end at 22 rpm, its slowest setting. 7. The samples are centrifuged at approximately 1,500g for 10 min. 8. The aqueous upper fraction is transferred to a new 15 mL tube taking care to leave behind any white precipitate at the interface. Gene Mutation Determination Using Muta™Mouse 315

9. Sufﬁ cient 5 M NaCl is added to each sample to bring the NaCl concentration to 200 mM (1/50 volume) and mixed with two to three inversions of the sample tube. 10. A volume of chloroform/isoamyl alcohol equal to the sample volume is added. 11. The sample is mixed on the rotator and then centrifuged as described above. 12. The aqueous fraction is transferred to a new 15 mL tube leaving behind any visible precipitate at the interface of the organic and aqueous layers. 13. DNA precipitation is initiated by the gentle addition of two volumes of ethanol. 14. The tube is gently rotated about its long axis. After several rota- tions it is slowly inverted end-over-end. These two motions are alternated until precipitation is complete. Note: With practice, the combination of the two motions will allow the DNA to precipitate into a single “glob” or “string.” Care should be taken to avoid too vigorous mixing which results in small, individual “strings” of DNA which cannot be spooled and which could cause shearing of the genomic DNA. Should this happen, the DNA may be recovered by centrifugation; see below. 15. Once the precipitation is complete as evidenced by the appearance of a stringy, white DNA precipitate, it is spooled onto a heat-sealed glass Pasteur pipet. 16. The DNA is washed by swirling the pipet tip in 70 % ethanol and left to air-dry, tip upward, for 5–10 min. 17. The pipet tip with the DNA is then placed into an appropriate − 4 volume of TE7.6 buffer in a microtube, to dissolve. 18. After 5–10 min the DNA will have dissolved free of the pipet which can then be removed. Note: Experience will determine the appropriate volume of buffer used to dissolve the DNA. This will depend on the amount and type of starting tissue and the quality of the isolation. The DNA concentration may be determined using a spectrophotometer after allowing a day or 2 for the DNA to dissolve and form a homogenous solution. The ideal concentration is about 1 μg/μL. The suggested starting volumes for dissolving DNA recovered from various tissues are as follows :

● Bone marrow (1 femur) 50 μL ● Kidney 100 μL ● Spleen 75 μL ● Heart 50 μL ● Brain (1/4) 50 µL ● Forestomach 25 µL 316 John D. Gingerich et al.

● Lung (1/2) 100 µL ● Liver (1/4) 100 µL ● Small intestine (1/3) 100 µL ● Colon (1/3) 75 μL ● Glandular stomach 50 μL

Note : If the DNA does not coalesce into a single mass, it can be recovered by centrifugation at 1,500 g for 10 min. After centrifugation, the ethanol is poured off, and 1 mL of 70 % ethanol is added to wash the sample which is then recentrifuged. The wash is poured off and the tube is left inverted overnight to allow the − 4 remaining ethanol to evaporate. In the morning, TE7.6 buffer is added to the tube to dissolve the DNA which is then transferred to an appropriate storage tube. 19. The DNA is allowed to dissolve at least overnight at 4 °C or preferably for several days. The concentration and quality of the DNA can be estimated using a spectrophotometer to measure optical density at 260 and 280 nm. Clean DNA will have a 260/280 ratio of 1.8–2.0. A concentration of about 1 μg/ μL is considered ideal for packaging. Note : In the past we have not routinely measured DNA concentration. Using a spectrophotometer with a 1 cm light path requires that the DNA be diluted. Due to the diffi culty of accurately diluting a small sample of a viscous DNA preparation, we have found that there is no consistent relationship between the measured DNA concentration and packaging effi ciency. Having established usual starting volumes for the DNA recovered from particular tissues, we fi nd it preferable to try to maximize the number of plaques recovered with each packaging reaction rather than maximizing packaging effi ciency. If the dissolved DNA is very viscous, the total number of plaques recovered may be increased by diluting the sample. While it may be possible to further improve packaging effi ciency in terms of plaques recovered per unit of DNA by diluting the sample even more, this may lead to a reduction in the overall number of plaques recovered. Additional packaging extracts may therefore be required to achieve the desired number of plaques per sample which for us is 125,000 plaques. More recently, using a spectrophotometer with a very short light path such as a NanoDrop , the determination of concentration can be accomplished without diluting the sample. This seems to improve the accuracy of the measurement, and as a result plaque recovery appears to be more consistent when the buffer volume is adjusted to give a DNA concentration of 1–2 µg /µL.

3.3 Mutant Detection Eight plates are required per sample, four to determine titer and four for mutant selection. Plates are poured with 8-mL minimal 3.3.1 Day 1: Host Cell agar each—usually the day before an assay is performed. This is and Agar Plate Preparation referred to as the “bottom” agar. Agar Plates Gene Mutation Determination Using Muta™Mouse 317

Each sample requires 32 mL of “top” agar for titer determination as well as 32 mL for selection of mutants. The recipe for both

“top” and “bottom” agar is the same except that 1 M MgSO 4 is added to the top agar after it has been autoclaved and cooled to

60 °C to give a ﬁ nal concentration of 10 mM MgSO4 . For selection plates only, p-gal is dissolved in dimethylformamide at 1 g/

mL and this is added to the top agar along with the MgSO 4 at a rate of 3 mL/L to give a ﬁ nal concentration of 0.3 % p-gal. This is done after the agar has been autoclaved and then cooled to 50 °C.

galE − Host Cells A 50-mL tube is prepared containing 10 mL of LB, 0.1 mL of

20 % maltose, 0.1 mL of 1 M MgSO4 , 50 μg/mL ampicillin, and 10 μg/mL kanamycin. This is inoculated with a scraping of the gal E − host cells from a frozen stock (or a colony on an agar plate) and incubated overnight in a shaking bath at 37 °C at about 200 rpm. The next day, the cells are subcultured by diluting 1/100 in LB

with maltose and MgSO 4 but without antibiotics. The amount needed is about 8 mL per sample to be tested. This culture is incu-

bated in a 37 °C shaking bath for 3.5 h at which time the OD 600 should be about 1.0 Note: In the event that cell growth is slower than expected, the culture time may be extended or the volume used to resuspend the cells may be

reduced to achieve a ﬁ nal cell concentration which gives an OD 600 of approximately 2. The culture is centrifuged and the cell pellet is resuspended in

1/2 the original volume of LB containing 10 mM MgSO 4 (OD600 ≈ 2). These are the host cells used below.

3.3.2 Day 2: Packaging/ 1. Four to ﬁ ve microliters of genomic DNA are dispensed into a Plating 1.5 mL microtube. Packaging Note : The genomic DNA may be very viscous. Cut off the pipet tip to widen the bore, and after taking up the desired volume of DNA, push the tip against the bottom of the tube and twist to cut off the sample in the tip from the rest of the DNA in the tube. Otherwise the DNA in the tip may be pulled back out when attempting to remove it from the sample tube. 2. 4.8 μL of packaging extract from a Transpack red tube is added, and the preparation is mixed by stirring. 3. When all of the samples have been prepared, they are pulse-centrifuged in a benchtop centrifuge and placed in a 30 °C water bath. 4. After 1.5 h the samples are removed from the bath, and 4.8 μL of extract from a Transpack blue tube is added with stirring and a pulse spin when all samples have been prepared. 5. After a further 1.5 h incubation at 30 °C, 500 μL of SM buffer is added to each sample. 318 John D. Gingerich et al.

6. Samples are rotated end-over-end on a rotator for approximately 30 min after which they are vortexed forcefully for 5–10 s to break up remaining genomic DNA. Note : After packaging, it is important that the packaged phage be distributed homogenously throughout the preparation. Because a small volume will be used to determine titer, this small volume must be representative of the concentration of phage in the whole preparation in order to get an accurate estimate of total phage numbers. This is the reason for the rotation and vortexing to break up remaining genomic DNA.

Plating 1. Prepared host cells are dispensed into 50 mL tubes at 2 mL per tube. Two tubes are required for each sample. 2. To one tube of host cells is added 500 μL of packaged phage. This is mixed briefl y using a vortex mixer and then allowed to stand at room temperature to allow the phage to adsorb to the host cells. This is the “test” tube. 3. After 25 min, the sample is mixed briefl y with a vortex mixer to ensure that the cells are evenly distributed, and 5 μL of the phage/host cell mix is transferred to a second tube containing 2 mL of host cells and again mixed briefl y. This tube is the “titer” tube (i.e., no selection). Note: Exercise care to ensure that this volume is as accurate as possible. Ensure that no excess medium is carried on the outside of the tip. Small inaccuracies will have a large effect on the estimate of titer and thus on the calculation of mutant frequency. 4. Thirty milliliters of top agar is added to the cells in the titer tube, and this is immediately distributed equally into four “titer” dishes (8 mL per dish). 5. Thirty milliliters of top agar, to which p-gal has been added to a fi nal concentration of 0.3 %, is added to the cells in the original “test” tube, and this is immediately distributed equally onto four “test” dishes (for selection of mutants). 6. The agar is allowed to cool for 10 min, following which the plates are inverted and incubated overnight at 37 °C.

3.3.3 Day 3: Plaque After overnight incubation, plaques are counted and the totals Scoring/Mutant Frequency recorded. There will probably be many plaques on some of the titer Determination plates. In these cases, in order to facilitate counting, it may be necessary to count a known fraction of the plate and estimate the total count from that. Note : We use a template that allows us to mark two opposing one- eighth sectors of a plate spaced 180° apart from each other. This allows us to score plaques on one-quarter of the plate and helps to compensate for uneven distribution of plaques across the plate. Gene Mutation Determination Using Muta™Mouse 319

The mutant frequency is determined by dividing the number of mutant plaques by the number of potential plaque-forming units (PFU) on the test plates. The number of mutant plaques is simply the sum of the number of plaques observed on the four “test” plates. The number of potential PFU on these plates (in the absence of selection) is calculated as follows: The number of PFU countedon4 titer plates (d volume of cells with adsorbed phage transferred from“” test tube to “ titre ” tube )[( voluume of cells in“” test tube volume of phage preparation added)–( voolume of cells with adsorbed phage transferred to“” titre tube )]

As described above, the volume of cells in the “test” tube is 2,000 μL, the phage preparation volume is 500 μL, and the volume of cells with phage transferred from “test” tube to “titer” tube to determine titer is 5 μL. Therefore the number of potential PFU in this case is determined as follows: #PFU on 4 titre dishes 5 PFU/, L2 000 500 5  L  # potential PFU oon4“ test ” dishes It is usual to multiply the number obtained for mutant frequency by 100,000 and express it as mutants per 100,000 plaques or simply as mutant frequency (×10 5 ). The results obtained are plotted as mutant frequency vs. dose to check for a dose-response. A doubling of mutant frequency in response to treatment would be considered positive providing that the power criteria have been met. In general, nonparametric statistics should be used for any analysis. Pair-wise comparisons can be used to determine the signiﬁ cance of individual doses, e.g., Wilcoxon signed-ranks test [16 ]. The Cochran-Armitage test is useful for determination of a dose-response relationship [17 ]. Also, see the following for a more in-depth discussion of statistical analyses such as Poisson regression [ 18 ] or logistical models [19 ]. While rare, it is possible that results for one animal will differ signiﬁ cantly from the values for the rest of the group. A statistical test for outliers such as Grubbs’ test may be used to determine if such a result is indeed an outlier and may be disregarded [ 20 ]. This situation may occur as a result of the clonal expansion of a spontaneous mutation in the transgene during the development of the animal. This can occur in a single tissue or the whole animal depending on when the mutation occurred. In the latter case, the mutation may have occurred in one of the germ cells before con- ception. If this were the case, a mutant frequency in excess of 1,000 × 10−5 would be expected, as one copy of lacZ (out of ~60 that constitute the transgene in each cell [ 21 ]) will carry the same mutation. 320 John D. Gingerich et al.

3.4 Mutant It may be of interest to determine the molecular characteristics of Characterization mutations resulting from treatment with a particular agent in order to elucidate the mechanism by which mutations are induced. Originally, sequencing the lacZ gene was a laborious process involving the use of complementation strains of E. coli to narrow the region of interest. This was followed by the use of multiple primers to cover the region of the gene where the mutation occurred [ 22 ]. More recently it has become common practice to isolate cII mutants for sequencing, cII being a much shorter gene (~300 bp) contained in the lambda vector, which can be sequenced with one primer pair [11 ]. The advent of next-generation sequencing techniques has removed the constraints formerly imposed by gene size meaning it is now possible to easily sequence large numbers of lacZ mutants to study and compare the spectra of the mutations induced by various chemical agents [23 ].

References

1. Ames BN, Mccann J, Yamasaki E (1975) 10. O’Brien JM, Beal MA, Gingerich JD, Soper L, Methods for detecting carcinogens and mutagens Douglas GR, Yauk CL et al (2014) Transgenic with the Salmonella/mammalian-microsome rodent assay for quantifying male germ cell mutagenicity test. Mutat Res 31(6):347–364 mutant frequency. J Vis Exp e51576. 2. Lambert IB, Singer TM, Boucher SE et al doi: 10.3791/51576 (2005) Detailed review of transgenic rodent 11. Manjanatha MG, Cao X, Shelton SD et al mutation assays. Mutat Res 590(1–3):1–280 (2013) In vivo cII , gpt , and Spi− gene mutation 3. OECD (2009) Detailed review paper on trans- assays in transgenic mice and rats. In: Dhawan genic rodent mutation assays. OECD, Paris Alok D, Bajpayee M (eds) Genotoxicity assess- 4. OECD (2013) Test no. 488: transgenic rodent ment: methods and protocols, methods in somatic and germ cell gene mutation assays. molecular biology, vol 1044. Springer OECD guideline for the testing of chemicals, Science + Business Media, New York Section 4: health effects. OECD, Paris 12. Douglas GR, Jiao J, Gingerich JD et al (1996) 5. Gossen JA, Molijn AC, Douglas GR et al Temporal and molecular characteristics of lacZ (1992) Application of galactose-sensitive E. coli mutations in somatic tissues of transgenic mice. strains as selective hosts for LacZ plasmids. Environ Mol Mutagen 28:317 Nucleic Acids Res 20:3254 13. Mientjes EJ, van Delft JHM, op’t Hof BM et al (1994) An improved selection method of 6. Jakubczak JL, Merlino G, French JE et al − (1996) Analysis of genetic instability during lambda lac phages based on galactose sensitiv- mammary tumor progression using a novel ity. Transgenic Res 3:67–69 selection-based assay for in vivo mutations in a 14. Tosato ML, Terlizzese M, Dogliotti E (1987) bacteriophage λ transgene target. Proc Natl Effects of buffer composition on water stability Acad Sci U S A 93(17):9073–9078 of alkylating agent, the example of N-ethyl-N- 7. Boerrigter ME, Dollé ME, Martus H-J et al nitrosourea. Mutat Res 179:123–133 (1995) Plasmid-based transgenic mouse model 15. OECD (2000) Guidance document on the for studying in vivo mutations. Nature recognition, assessment and use of clinical 377(6550):657–659 signs as humane endpoints for experimental 8. Kohler SW, Provost GS, Fieck A et al (1991) animals used in safety evaluation: series on test- Analysis of spontaneous and induced muta- ing and assessment. OECD, Paris tions in transgenic mice using a lambda 16. Wilcoxon F (1945) Individual comparisons by ZAP/ lacI shuttle vector. Environ Mol ranking methods. Biometrics 1:80–83 Mutagen 18(4):316–321 17. Armitage P (1955) Tests for linear trends in 9. Nohmi T, Katoh M, Suzuki H et al (1996) A proportions and frequencies. Biometrics 11(3): new transgenic mouse mutagenesis test system 375–386 using Spi- and 6-thioguanine selections. 18. Lemieux CL, Douglas GR, Gingerich J et al Environ Mol Mutagen 28(4):465–470 (2011) Simultaneous measurement of benzo[a] Gene Mutation Determination Using Muta™Mouse 321

pyrene-induced Pig-a and lacZ mutations, λgt10-lacZ transgene: evidence for in vivo micronuclei, and DNA adducts in Muta™Mouse. rearrangements. Mutagenesis 25(6):609–616 Environ Mol Mutagen 52:756–765 22. Vijg J, Douglas GR (1996) Bacteriophage 19. Fung KY, Xihong L, Krewski D (1998) Use of lambda and plasmid lacZ transgenic mice for generalized linear models in analyzing mutant studying mutations in vivo . In: Pfeifer G (ed) frequency data from transgenic mouse assay. Technologies for detection of DNA damage Environ Mol Mutagen 31:48–54 and mutations, part II. Plenum, New York, NY 20. Grubbs FE (1969) Procedures for detecting 23. Besaratinia A, Li H, Yoon J-I et al (2012) A high- out-lying observations in samples. throughput next-generation sequencing- based Technometrics 11:1–21 method for detecting the mutational ﬁ nger- 21. Shwed PS, Crosthwait J, Douglas GR et al print of carcinogens. Nucleic Acids Res (2010) Characterisation of Muta™Mouse 40:e116 Chapter 19

The Mouse Lymphoma Assay

Tao Chen, Xiaoqing Guo, and Martha M. Moore

Abstract

The mouse lymphoma assay (MLA) is the most widely used mammalian cell gene mutation assay for regulatory purposes and is included in the core battery of genotoxicity tests for the registration of pharmaceuticals, pesticides and for other regulatory decision-making. The assay detects mutations in the thymidine kinase (Tk) gene. The Tk mutants recovered in the MLA can be classified as large and small colonies, indicating the potential for the test chemical to induce gene mutations and/or chromosomal mutations, respectively, and thus revealing the mutagenic mode of action of test agents. There are two methods for enumerating mutants in the assay, the agar and microwell versions. In this chapter, we introduce the principle of the assay and provide details for conducting both versions of the assay. We include our strategy for test chemical concentration selection, the appropriate way to measure cytotoxicity, and the treatment time. We also include the current internationally harmonized approach for data interpretation and acceptance criteria for valid assays. As an example assay, we provide data for analysis of mutagenicity of 5-nm silver nanoparticles.

Key words Mouse lymphoma assay, Thymidine kinase, Mutagens, Genotoxicity test, Point mutations, Large-colony mutants, Small-colony mutants, Nanomaterials

1 Introduction

The mouse lymphoma assay (MLA) using the Thymidine kinase gene (Tk1) as the target for mutation induction detects a broad spectrum of genetic damage due to the nature and autosomal location of the Tk gene. The assay has been clearly demonstrated to detect both point mutations and chromosomal mutations [1–7]. The Tk gene codes for a cytosolic protein, a phosphotransferase enzyme, involved in the pyrimidine nucleotide salvage pathway. The TK enzyme phosphorylates deoxythymidine to deoxythymidine 5′-phosphate so that deoxythymidine can be incorporated into DNA. Several features of the Tk gene allow the assay to detect different types of mutations. First, a functional Tk gene is not necessary for cells in culture so that a Tk mutant can still grow and develop into a colony. Second, Tk-deficient cells can be selected with the pyrimidine analog trifluorothymidine (TFT). While normal

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_19, © Springer Science+Business Media New York 2014 323 324 Tao Chen et al.

cells cannot grow in the TFT selective growth medium because the toxic TFT is incorporated into their DNA resulting in cytocidal effects, the mutant cells can grow in the medium and develop into colonies due to their nonfunctional pyrimidine salvage pathway. Third, the Tk locus has an autosomal location and is present in two copies, one of which is mutant and does not produce functional TK enzyme. Unlike the X-linked functionally hemizygous Hprt locus that primarily recovers mutations within the gene, the Tk locus is sensitive to mutations involving not only intragenic events but also chromosome alterations associated with the Tk locus, such as chromosome deletion and rearrangements due to mitotic recombination. Hence, the MLA detects both gene and chromosome mutations. The L5178Y/Tk+/− −3.7.2C mouse lymphoma cell line is used for the MLA. This Tk heterozygous cell line was developed specifi- cally for this assay because the frequency at which Tk functional homozygous cells can be mutated is extremely low, too low to be practical for a mutation assay. The L5178Y/Tk+/− −3.7.2C mouse lymphoma cell line, with only one functional copy (Tk1b) located on mouse chromosome 11 can be mutated with frequencies compatible with easy mutation detection. A distinctive feature of the Tk mutant colonies recovered from this cell line is that mutants grow to different size colonies. Large-colony mutants grow at a normal rate similar to the Tk heterozygous cells and small-colony mutants grow at a slower rate. The relative frequency of the two colony classes is mutagen dependent [8]. Although the relationship is not absolute, clastogenic chemicals, that is, chemicals that break chromosomes and induce few point mutations tend to produce more small-colony mutants, whereas chemicals inducing point mutations and do not cause chromosome breakage tend to induce more large-colony mutants [8, 9]. Individual Tk mutant clones can be further characterized using a combination of molecular and cytogenetic analysis to determine the exact type of mutation that caused the loss of Tk function. Thus, colony sizing together with molecular analysis of the mutants can reveal the mutagenic mechanism of action of the test agents [3–5, 10, 11]. There are currently two equally acceptable methods for enumerating mutants in the MLA. The assay was originally developed using the formation of cell clones immobilized in soft agar to enumerate mutants [12, 13]. In 1983, Jane Cole and her coworkers published a method using liquid medium and 96-well microwell plates for mutant frequency determination [14]. Many protocol improvements have been made since the MLA was originally developed by Don Clive and his coworkers more than 40 years ago, and currently the assay is for measuring the mutagenicity of chemicals and other test agents [13–22]. Because of its ability to detect a broad range of mutational events, including both point and chromosomal mutations, the MLA is generally the assay recommended by regulatory agencies as the The Mouse Lymphoma Assay 325

preferred in vitro mammalian gene mutation assay for the core battery of genotoxicity tests [23–25]. A specific guidance for the conduct of the MLA can be found in the Redbook located on the website of the US Food and Drug Administration, Center for Food Safety and Applied Nutrition [26]. In this chapter, we update our previous article [12] and describe the MLA, including materials, cell culture, chemical treatment, cloning, and data calculation and interpretation. We also provide our strategy for test chemical concentration selection, the appropriate way to measure cytotoxicity, and the treatment time. We also provide the internationally harmonized (IWGT reference) approach for data interpretation and acceptance criteria for valid assays. Because we have recently been applying the MLA to the evaluation of nanomaterials, we provide data here, as an example, for analysis of 5-nm silver nanoparticles (AgNPs).

2 Materials

We routinely use the following materials for conducting the MLA. The sources and use of these materials are also given: 1. BBL agar (Baltimore Biological Laboratories, Baltimore, MD) for the soft agar assay 2. Benzo(a)pyrene (BP) (Sigma, St. Louis, MO) as a positive control when S9 mix is applied 3. Cyclophosphamide (CP) (Sigma) as a positive control when S9 mix is applied 4. Dimethyl sulfoxide (DMSO) (Sigma) as a solvent for chemicals not soluble in water and as a component of the cell-freezing medium for liquid nitrogen storage 5. Glycine (Sigma) for THMG and THG media 6. Horse serum (Invitrogen, Carlsbad, CA) for preparing different culture media 7. Hypoxanthine (Sigma) for preparing THMG and THG media 8. Fischer’s medium for leukemic cells of mice with glutamine (Quality Biologicals, Inc., Gaithersburg, MD) as the basic medium for different culture and freezing media 9. Methotrexate (Sigma) as a component for THMG medium 10. Methymethanesulfonate (MMS) (Sigma) as a positive control for testing without S9 mix 11. Nicotinamide adenine dinucleotide phosphate (NADP) (Sigma) as a component of S9 mix 12. 4-Nitroquinoline-1-oxide (NQO) (Sigma) as a positive control for testing without S9 mix 326 Tao Chen et al.

13. Penicillin-streptomycin (Invitrogen) as an antibiotic in cell culture media 14. Pluronic F68 (Invitrogen), used to prevent mechanical disruption of cells during shaking 15. Thymidine (Sigma), a component for THMG and THG media 16. Trifluorothymindine (TFT) (Sigma) forTk mutant selection 17. S9 (In vitro Technologies, Baltimore, MD), used for metabolic activation of chemicals that are not mutagenic without metabolism 18. Sodium pyruvate (Sigma), a component of culture media

3 Methods

3.1 Cell Maintenance The L5178Y/Tk+/− −3.7.2C mouse lymphoma cell line is the only appropriate cell line for conducting the MLA. This cell line, which was originally obtained from Dr. Donald Clive, is available from our laboratory at the National Center for Toxicological Research/ FDA, Jefferson, AR. A current effort, coordinated by ILSI/HESI, is establishing a cell repository for all of the cell lines commonly used for genetic toxicology evaluations. Once this repository is fully functional, it should be considered to be the preferred source of this cell line. Fischer’s medium for leukemic cells of mice supplemented with horse serum is routinely used for this assay in our laboratory. Other laboratories prefer to use RPMI 1640 medium, for all or part of the assay, and this is acceptable. Horse serum is inactivated by heating for use with both Fischer’s and RPMI media, but it is particularly important when RPMI 1640 medium is used for mutant selection [16, 27]. In addition, a threefold higher concentration of TFT (3 μg/mL) is required for mutant colony selection when using RPMI 1640 although the stringency of the mutant selection conditions may be different in different laboratories [9, 15]. It is important that every laboratory verify their selection conditions, both medium used and concentration of TFT, by isolating a large number of colonies isolated as presumed mutants and testing them to assure that they actually are Tk mutants.

The basic medium (F0P) consists of Fischer’s medium supplemented with 100 unit/mL penicillin and 100 μg/mL streptomycin, 200 μg/mL sodium pyruvate, and 0.05 % (v/v) Pluronic F68. Note: Pluronic is important for use with cultures that are either rolled on a roller drum or incubated in a rotating shaker platform. It is less important and generally felt to be unnecessary if the cell cultures are incubated in a stationary setting. Several different media are made from the basic medium and different concentrations of horse serum. Freshly thawed horse The Mouse Lymphoma Assay 327

serum should be heat inactivated at 56 °C for 30 min before using. It is important that both the temperature and time be carefully monitored if RPMI 1640 medium is used for TFT mutant selection. All media should be properly filter sterilized and stored in a lighttight refrigerator at 4 °C. Media and heat-inactivated horse serum should be warmed to room temperature before use. Growth

medium (F10P) is made by adding 10 % (v/v) heat-inactivated horse serum into F0P that is utilized for cell maintenance and growth. Cloning medium (F20P) is made by adding 20 % (v/v) heat- inactivated horse serum to F0P. The cells should be periodically checked for mycoplasma contamination. Only cells with normal cell line karyotype and population doubling times and free of mycoplasma contamination should

be used. Cells can be cryopreserved for storage in liquid N2 using F20P (adding 20 % (v/v) serum to F0P) containing 5 % DMSO. In the past, we would grow our stock cultures for several weeks, although not more than 3 months. Current discussions related to the ILSI/ HESI cell repository exercise and the development of improved guidance for cell growth lead us to change this recommendation. We now cleanse a large stock of cells and freeze them in liquid nitrogen. Individual vials are thawed for single use or for only a very small number of uses. The cultures are grown in polypropylene

tissue culture flasks that are gassed with 5 % CO2 in air and placed on a shaker incubator at 37 °C with constant slow mixing. Note: Stationary cultures incubated at 37 °C in a humidified incu-

bator gassed with 5 % CO2 in air are also acceptable. The cells should be maintained in log phase, with a doubling time of 9–10 h. Cell density is determined by using a Coulter counter or a hemocytometer, and the cultures are routinely diluted with 5 fresh F10P medium each day to 2 × 10 cells/mL. Each Friday, the cells are diluted to 7 × 103 cells/mL for a weekend. Doubling times should be carefully monitored, and cultures showing doubling times in excess of 10 h should not be used for experiments.

3.2 Cleansing To eliminate preexisting Tk mutants in cultures, the mouse lym- Cultures of Preexisting phoma cells should be treated with THMG and THG media to Tk Mutants cleanse the culture within the week preceding each assay or prior to freezing a large number of vials for storage. A 100× THMG stock medium is made with F0P containing 300 μg/mL thymidine, 500 μg/mL hypoxanthine, 10 μg/mL methotrexate, and 750 μg/ mL glycine. A 100× THG stock medium contains the same components as THMG stock medium but without methotrexate. These stock media can be made and stored at –20 °C for future use. The cleansing procedure is performed by adding 0.5 mL of THMG stock to 50 mL of the stock culture at 2 × 105 cells/mL

F10P, mixing and gassing the culture with 5 % CO2 in air, and placing it in an incubator at 37 °C for 24 h. After performing a cell 328 Tao Chen et al.

count (the cell count should not be less than 1.0 × 106 cells/mL), centrifuge the cells at 200 × g, 10 min, and resuspend the cell pellet at a concentration of 2 × 105 cells/mL in THG medium

(F10P medium containing 1 % THG stock). The culture is gassed with 5 % CO2 in air and placed in an incubator at 37 °C for 24 h. The cells can be expected to grow at longer doubling times during cleansing than they do normally. Normal growth should resume after the first 24 h in THG medium. The cells should not be exposed to test chemicals until they have completely recovered from cleansing. Cleansed cells may be grown and cryopreserved at a density of 5 × 106 cells/mL/tube in freezing medium. New cultures for assays may be started directly from the cryopreserved cleansed stocks.

3.3 Chemical S9 Mix. S9 mix is used to provide metabolic activation of pre- Treatment mutagens. A chemical must be adequately tested both with and without metabolic activation before declaring it to be negative in the assay. Aroclor-induced rat liver S9 is routinely utilized for this purpose. S9 is available from commercial sources or made according to the published methods [28]. In the treatment, the S9 mix for

each culture contains 3 mL of cofactor mix (F0P supplemented with NADP (8 mg/mL) and sodium isocitrate (15 mg/mL), neutralized to an orange-red color with 1N NaOH and filter sterilized) and 1 mL of S9 (equivalent of 25 mg of protein per mL). It should be freshly prepared and kept on ice prior to use. Positive and Negative Controls. To know whether each assay is successful, positive and negative controls should be always included with each experiment. The positive control chemicals commonly used include MMS (10–20 μg/mL) and NQO (0.05–0.1 μg/mL) in the absence of S9 and BP (2–3 μg/mL) and CP (3–5 μg/mL) for testing with S9. Although MMS is widely used as a positive control chemical, there are some reservations over its use because it is volatile and hydrolytic. In addition, some commercial supplies of MMS are not as mutagenic as others. NQO and BP can be prepared with DMSO as 100-fold concentrated stock solutions and stored as frozen aliquots at –80 °C in the dark [21]. However, MMS and CP 100× stocks should be freshly prepared with physi-

ological saline. If a solvent other than saline or F0P is used, the solvent control should receive a dose of the solvent equivalent to the highest amount used for a treated culture but should not exceed a final volume of 1 %. Test Compound Exposure Concentrations. Test chemicals should be

dissolved in a suitable solvent such as saline, DMSO, or F0P to make appropriate concentrations of the stock solutions. Treatment is initiated by adding these stock solutions into treatment medium containing cells in suspension. The final volume should not exceed 1 % when DMSO or other nonaqueous solvents are used; the The Mouse Lymphoma Assay 329 amount of the solvent should be the same in all cultures. This may require that additional solvent be added to cultures, depending upon the strategy used to dilute the test chemical, that is, how much of the stock solution is added to each test culture. The exposure concentrations of a test chemical are selected according to its toxicity spanning from ~100 to 10–20 % relative total growth (RTG; see Sect. 3.5 for calculation). RTG is a combination of the relative suspension growth (RSG) during the expression period and the relative cloning efficiency determined at the time of mutant selection. Different laboratories utilize various strategies to identify the appropriate concentrations to be used. Generally, in the absence of cytotoxicity information of the test compound, we perform a preliminary experiment using half-log dilutions (around eight serial concentrations between 5000 and 0.1 μg/mL) of the chemical. Generally, a second (and perhaps more) preliminary experiment is required to fine-tune the dose range to obtain a sufficient number of cultures that adequately cover the dose range. A full experiment will usually be conducted with five or more concentrations. It is important to have more than one data point that can be used to establish whether the chemical is positive or negative. For chemicals that induce high mutant frequencies, it may not be necessary to use doses covering the entire 100–10 % RTG range. For chemicals that are negative or weakly positive, more concentrations are generally necessary with a focus on using doses that are near the 10–20 % RTG cytotoxicity level. The current OECD and Redbook guidelines specify that at least four analyzable concentrations are required with duplicate test cultures and eight analyzable concentrations are required for single cultures. Many laboratories prefer to use duplicate (or triplicate) cultures, but the use of single cultures and more different doses increases the probability of having the appropriate concentrations upon which a decision can be made. It is advisable to use duplicate (or triplicate) cultures for the negative/solvent control. For chemicals deemed to be noncytotoxic or weakly cytotoxic, the maximum concentration is 2 mg/mL, 2 μL/mL, or 10 mM, whichever is the lowest. This is a new recommendation that will be incorporated into the OECD revisions of all of the in vitro mammalian genetic toxicology assays, and it represents a lowering of the top concentration from the historically recommended levels. It should be noted that the ICH (for pharmaceuticals intended for human use) recommends a top level of 1 mM, unless the chemical has a low molecular weight (defined as less than 250). Compounds that have limited solubility should be tested at a concentration up to or beyond their limit of solubility under the culture conditions. The new OECD test guidelines will recommend that only one precipitating dose be used. Compounds that are not soluble in any acceptable solvent cannot be appropriately evaluated for their mutagenicity in the MLA. The reader is referred to the publications of the IWGT MLA Workgroup for more recommendations on concentration selection [21]. 330 Tao Chen et al.

Treatment of Cell Cultures. Generally, it is recommended that the test be conducted with and without S9 metabolic activation. However, there are some circumstances, particularly when the test chemical is positive without activation, in which the S9 treatment is either unnecessary or not recommended. Table 1 provides an example of an experiment for the evaluation of a chemical both with and without S9. The cells should always be maintained in logarithmic growth including for cell treatment. The serum level in the cultures used for

treatment should be reduced to 5 % (v/v) with F0P. This can be done by centrifuging cells and resuspending them in fresh medium containing 5 % serum. Generally for the individual test cultures, we use 50-mL sterile disposable centrifuge tubes containing 6 × 106 cells in

6 mL of F5P. To each tube, we add either 4 mL of F0P (without metabolic activation) or 4 mL of S9 mix (with metabolic activation). The test chemical is added to each tube with gentle mixing. After the

addition of the test chemical, all cultures are gassed with 5 % CO2 in air (or placed in a CO2 incubator for stationary cultures) and incubated in a roller drum at 37 °C for 4 h. While we have generally used a 4-h treatment period, others, particularly those conducting the microwell version of the assay, normally use a 3-h treatment time. There is also a recommendation for pharmaceuticals intended for human use that there be a 24-h treatment (without S9) for those chemicals that are negative in the short treatment. After the incubation period, the cells are centrifuged at 200 × g for 10 min, and

the supernatant is discarded. Each culture is then washed with F0P twice by resuspending the cells in fresh medium and centrifugation. After the final centrifugation, the cell pellet is resuspended in 5 20 mL of fresh F10P at a concentration of 3 × 10 cells/mL. After these steps, the treatments are incubated as noted above for an expression period of 2 days. Cell densities are determined approximately 24 h following treatments and adjusted to 5 2 × 10 cells/mL with fresh F10P. On completion of the 2-day expression period, cell densities are determined. The cell densities from day 1 and day 2 are utilized for calculating the RSG and the RTG (see Sect. 3.5. Calculations and examples in Tables 1 and 2). Cultures with cell densities less than 2 × 105 mL−1 will not be considered for cloning. If a test chemical yields negative responses in the 3- or 4-h treatment assay with and without S9 mix, a 24-h treatment trial may be conducted in the absence of S9 mix [20]. For the 24-h treatment incubations, cultures of 50 mL at 2 × 105 cells/mL culture medium are treated in flasks with a series of diluted test chemi-

cals for 24 h in a 37 °C, 5 % CO2 humidified incubator. The cells are then centrifuged and washed twice. They are transferred to new flasks and adjusted to 50 mL at 2 × 105 cells/mL with fresh medium for growth through the 2-day expression period. For the 24-h treatment, the RSG and the RTG should include the cytotoxicity that occurs during the 24-h treatment. The Mouse Lymphoma Assay 331

Table 1 Sample data for demonstrating the treatment condition, main parameters, and calculation of cell cytotoxicity and mutant frequency in the mouse lymphoma assay

Conc. RSG RPEV RTG PEM MF −6 −6 Culture no. (μg/mL) S9 SG1 SG2 (%) PEV (%) (%) (×10 ) (×10 ) %SC Sol. Con.-100 0 − 4.60 6.02 100 0.92 103 103 40 44 49 Sol. Con.-101 0 − 4.64 5.98 100 0.87 97 97 44 51 47 Dose 1-102 10 − 4.61 6.03 100 0.94 105 105 45 48 48 Dose 2-103 20 − 4.49 5.82 94 0.80 89 84 50 62 52 Dose 3-104 40 − 4.22 5.83 89 0.78 87 77 120 154 60 Dose 4-105 80 − 3.88 5.74 80 0.74 83 66 164 222 62 Dose 5-106 160 − 3.53 5.62 72 0.68 76 55 271 399 68 Dose 6-107 320 − 2.92 5.02 53 0.60 67 36 419 698 71 Dose 7-108 640 − 2.31 4.77 40 0.51 57 23 441 865 79 Dose 8-109 1,280 − 1.82 3.89 26 0.46 51 13 412 896 81 Pos. Con. 1-150 20 − 2.24 4.11 33 0.47 53 18 636 1,353 51 Sol. Con.-200 0 + 4.51 5.83 96 0.94 103 99 51 54 51 Sol. Con.-201 0 + 4.72 6.01 104 0.89 97 101 54 61 53 Dose 1-202 10 + 3.89 5.88 84 0.90 98 82 65 72 50 Dose 2-203 20 + 3.63 5.62 75 0.78 85 64 180 231 62 Dose 3-204 40 + 3.46 5.71 72 0.73 80 58 290 397 70 Dose 4-205 80 + 3.15 5.79 67 0.70 77 52 340 486 67 Dose 5-206 160 + 2.77 5.19 52 0.65 71 37 440 677 76 Dose 6-207 320 + 2.04 4.52 34 0.55 60 20 520 946 74 Dose 7-208 640 + 1.10 4.29 19 0.53 58 11 602 1,136 85 Dose 8-209 1,280 + 0.20 Discard culture Pos. Con. 2-250 10 + 1.81 4.86 32 0.45 49 16 550 1,222 66

Abbreviations: SG1 suspension growth rate between day 0 and day 1 of the expression time, SG2 suspension growth rate between day 1 and day 2 of the expression time, RSG relative suspension growth, PEV plating efficiency for viability,

RPEV relative plating efficiency for viability, RTG relative total growth rate, PEM plating efficiency for mutants, MF mutant frequency, SC small colony, Conc. concentration, Sol. Con. solvent control, Pos. Con. positive control

3.4 Cloning Before the cloning, TFT stock solution and cloning medium have to be prepared. TFT stock solution is made by mixing 10-mg TFT with 100-mL physiological saline in a foil-wrapped bottle. The stock solution should be filter sterilized, dispensed in 15-mL aliquots into sterile tubes, and stored at −20 °C for up to 3 months.

F20P is used for the cloning medium. For the agar method, pre- warmed F20P is measured into an Erlenmeyer flask, and autoclaved 332 Tao Chen et al.

Table 2 Suspension growth of mouse lymphoma cells treated with different concentrations of 5-nm silver nanoparticles

Day 1 cell Day 2 cell Dose concentration concentration 5 5 Culture I.D. (μg/mL) (×10 ) SG1 (×10 ) SG2 SG RSG Control-1 0.00 17.00 5.70 11.00 5.50 31.35 1.08 Control-2 0.00 16.00 5.30 10.00 5.00 26.50 0.92 AgNP-1 3.00 15.00 5.00 9.00 4.50 22.50 0.78 AgNP-2 3.50 14.00 4.70 9.50 4.80 22.56 0.78 AgNP-3 4.00 13.50 4.50 9.00 4.50 20.25 0.70 AgNP-4 4.50 12.00 4.00 9.00 4.50 18.00 0.62 AgNP-5 5.00 10.00 3.30 9.00 4.50 14.85 0.51 AgNP-6 5.50 9.00 3.00 8.00 4.00 12.00 0.41 AgNP-7 6.00 6.00 2.00 7.00 3.50 7.00 0.24 NQO 0.10 14.50 4.80 9.00 4.50 21.60 0.75

BBL agar (at 95 °C) at a final concentration of 0.28 % is added with thorough mixing. The soft agar cloning medium is freshly made and kept at 37 °C prior to use [29]. For mutant enumeration, each culture is centrifuged and the 5 cell pellet resuspended into F20P at a density of 2 × 10 cells/mL. It is important that the cell pellet be resuspended in such a way as to assure that the cells are in a single cell suspension so that individual cells are plated and that the colonies that form come from single cells. The cultures should be mixed and incubated for at least 30 min to minimize trauma and adapt to the medium. The cells then are diluted to the appropriate densities to plate for TFT resistance and cell viability. Cloning for Mutant Selection and Plating Efficiency. For the soft agar version of the assay, 3 × 106 cells from each sample are centrifuged, and the cell pellet is resuspended in 100-mL soft agar cloning medium and mixed thoroughly. Prior to the addition of 3 μg/ mL TFT, 0.5 mL of the soft agar medium cell mixture is taken and placed into a flask containing 50 mL of soft agar cloning medium. After thorough mixing, a 2-mL sample of the cells in medium is taken and placed into a flask containing 98 mL of cloning medium, thus giving the 600 cells in 100-mL cloning medium needed for determining cloning efficiency in the absence of TFT see( below). The Mouse Lymphoma Assay 333

Both the 100-mL cultures for TFT selection and for cloning efficiency are distributed into three 100-mm tissue culture petri dishes. The plates are chilled at –20 °C for 12 min to solidify the

agar. The plates are then placed in a 37 °C, 5 % CO2 incubator. For the microwell version of the assay, the cells are agitated to form a single cell suspension, and the cell concentrations are 4 −1 adjusted to 1 × 10 mL F20P. Each culture is then sampled and diluted to give a culture with 8 cells/mL using a two-step dilution. TFT (3 μg/mL) is then added to the selection flask. Using a multichannel pipette, place 200 μL of each TFT containing suspension into each well of four flat-bottom 96-well plates and 200 μL of the 8 cells/mL culture from each sample into each well of two flat- bottom 96-well plates. Incubation and Colony Counting. The plates with seeded cells are

incubated at 37 °C in a humidified incubator gassed with 5 % CO2 in air for 11–14 days. For the soft agar version of the assay, colony counting and sizing from selection and viability plates are performed using an automatic colony counter fitted with the capability to evaluate the size of the colonies. Mutant colonies approximately <0.6 mm in diameter are considered to be small- colony mutants and those larger to be large-colony mutants. For the microwell version of the assay, colonies are identified by low- power microscope or eye, and small colonies are defined as less than a quarter of the diameter of the well, while large colonies are more than a quarter of the diameter of the well. The morphology is generally compact for small colonies and may be diffuse for large colonies.

3.5 Data Calculation Mutant Frequency. The mutant frequency (MF) is determined by and Interpretation the plating efficiencies of mutant colonies (PEM) and adjusted with plating efficiencies of viable cells (PEV) from the same culture. The calculation is MF = PE / PE MV Table 1 shows a set of sample data and calculation of MF and

RTG. For negative Control 1, Culture 100 in Table 1, the PEM −6 and PEV are 40 × 10 and 0.92, respectively. Therefore, this spon- −6 −6 taneous mutant frequency is PEM/PEV = 40 × 10 /0.92 = 44 × 10 . For the positive Control 1, Culture 150, the MF is 636 × 10−6/0.47 = 1,353 × 10−6.

PEM and PEV are calculated by using the number of colonies and the total number of cells used for the cloning: PE = CT/ MMM PE = CT/ VVV 334 Tao Chen et al.

where CM is the number of colonies on the selective plates, TM is the total number of cells used for selection, CV is the number of colonies on the viability plates, and TV is the total number of cells used for viability.

In soft agar version of the assay, CM and CV are obtained by directly counting the clones. When 600 cells are plated for cloning efficiency and 3 × 106 cells are used for mutant selection, PE =×CC//310366= ()×10− MM()M PE = C / 600 VV For example, if we count 150 mutants from TFT selection plates and 480 colonies from non-selection plates, PE = 150 / 31×=05−−6601× 0 M () PE =480 /.600 = 080 V

MF ==PE //PE 50×10−−6608..06=×25 10 MV()

In the microwell version of the assay, however, CM and CV are determined as the product of total number of microwells (TW) and the probable number of colonies per well (P) in the microwell plates: CP=×TW MM M CP=×TW VV V From the zero term of the Poisson distribution, P is given by P =−ln EW / TW () where EW is empty wells and TW is total wells. Therefore, PE ==CT//PT× TW MMMM()MM PE ==CT//PT× TW VVVV()VV With the knowledge that 768,000 cells are plated in four 96-well plates (2 × 103 cells/well) for mutant selection and 307 cells are plated in two 96-well plates (1.6 cells/well) for viability, PE =×()PP384 //768000 =×2103 MM M () PE =×PP192 //307 = 16. VV() V The Mouse Lymphoma Assay 335

If we find 284 empty wells from 384 total wells in the four TFT selection plates and 50 empty wells from 192 total wells in the two viability plates, PE =−ln()284 //384 21× 036=×150 10− M () PE =−ln 50 //192 16..= 084 V ()

MF =×150 10−−66/.084 =×179 10 () Relative Total Growth. RTG is the measure for cytotoxicity of the test chemical. It is calculated as RTGR=×SG RPE V

where RSG is the relative suspension growth and RPEV is the relative plating efficiency for viability:

=× × RSGS()GS12()test GS()test / ()GS12()controlcG ()ontrol

RPEP= EP/ E V Vt()estV()control

SG1 is the growth rate between day 0 and day 1 (cell concentration at day 1/cell concentration at day 0), and SG2 is the growth rate between day 1 and day 2 (cell concentration at day 2/cell concentration at day 1). Here is an example for calculating the RTG of Culture 107 in

Table 1. First, we determine the means of SG1 × SG2 and PEV from the duplicate solvent controls: SG ×=SG ()46..06×+02 46..45× 98 /.22= 77 12()controlc()ontrol PE =+()09..2087 /.20= 90 Vc()ontrol

Then, we calculate the RSG and RPEV of Culture 107: RSG =×29..2502 /.27 70= .53 () RPE = 06./00..90 = 067 V Finally, we get the RTG as =×== RTG 05..3067 03.%636 Percentage of Small- and Large-Colony Mutants. The percentage of small-colony mutants is simply calculated as below: %/SC = ()()smallcolonyMF ()totalMF ×100

=− %%LC 100 % SC 336 Tao Chen et al.

where SC and LC represent small-colony mutants and large-colony mutants, respectively. Data Interpretation. The IWGT MLA Workgroup reached consensus on the criteria for defining an acceptable assay and for positive and negative results [30, 31]. A summary of the defined criteria is presented below: 1. Assay acceptability: An assay is considered acceptable and valid only if the negative/solvent control meets the following criteria: Soft Agar Method Mutant frequency: 35–140 × 10−6 Cloning efficiency: 65–120 % Suspension growth: 8–32-fold Microwell Method Mutant frequency: 50–170 × 10−6 Cloning efficiency: 65–120 % Suspension growth (corrected for cytotoxicity during treatment and during expression): 8–32-fold The positive control must also adequately demonstrate that the assay was properly conducted and that small-colony mutants were optimally detected. The IWGT MLA Workgroup recommends that either there be an induced small-colony mutant frequency of at least 150 × 10–6, or that there be an induced total mutant frequency of 300 × 10–6, with 40 % of the total mutant frequency consisting of small- colony mutants. 2. Criteria for positive and negative results: The harmonized IWGT MLA Workgroup criteria for a positive test chemical response require that there be an induced MF of at least 90 × 10−6 for the soft agar version and 126 × 10−6 for the microwell version of the assay. The attainment of these values, termed the global evaluation factor or GEF and a concentration- related increase evaluated by an appropriate statistical method has been determined to represent a biologically relevant increase in the assay. A compound is negative if it does not meet the criteria for a positive response. Before a negative response can be concluded, however, the treatment concentrations must include doses that exhibit sufficient cytotoxicity (10–20 % RTG). It is recognized, however, that positive responses seen only between 10 and 20 % RTG should be interpreted with caution, and positive responses seen only below 10 % RTG should not be considered to be an indication that the test chemical is mutagenic. The Mouse Lymphoma Assay 337

4 An Example for Conducting the MLA for Genotoxicity Assessment of Nanoparticles

We have evaluated the genotoxicity of different nanomaterials using several different assays including the Ames test, the MLA, the in vivo and in vitro Comet assay, the in vivo and in vitro micronucleus assays, and the in vivo Pig-a assay [32–35]. For the details, the readers are encouraged to read the original articles. Here, we focus on the method for conducting the MLA to evaluate the genotoxicity of 5-nm uncoated AgNPs [33]. Characterization of AgNPs. The primary sizes of the AgNPs were determined using transmission electron microscopy. More than 100 nanoparticles were measured, and the size distribution of the particles and aggregates were calculated. The particles measured in diameters for sizes of less than 4, 4–8, 8–12 nm, and above 12 nm were 4 %, 66 %, 24 %, and 6 %, respectively. Dynamic light scattering (DLS) technique was used to characterize the behavior of the AgNPs, and the hydrodynamic sizes and surface charge were measured using Zetasizer. The mean agglomerate sizes ranged from 61.2 nm (in water) to 1,608.7 nm (in medium), and the surface charge ranged from −9.37 mV (in water) to −8.20 mV (in medium). Cell uptake of the particles was confirmed with a confocal microscopy. Preparation of the Working Solution for Treatment. A 5-nm AgNP stock solution (1 mg/mL) was suspended in sterilized water. Prior to use, it was vortexed for 5–10 min and then sonicated for 10 min in an ultrasonic water bath to ensure a uniform suspension. The solution was centrifuged at 78 × g for 5 min to remove possible aggregates. The supernatant was filtered (0.20 μm) to sterilize. A preliminary experiment for defining the cytotoxicity dose range showed that doses lower than 3 μg/mL were associated with little cytotoxicity while doses higher than 6 μg/mL resulted in less than 10 % RTG. Accordingly, six working solutions of 300, 350, 400, 450, 500, 550, and 600 μg/mL AgNPs were prepared by a series of dilutions using sterilized water. Cell Treatment and Mutation Expression. Since AgNPs are pure metal and therefore do not require metabolic activation, only a treatment without S9 was conducted. For the treatment, mouse lymphoma cells were suspended in 50-mL centrifuge tubes con- 6 taining 6 × 10 cells in 10 mL of treatment medium (F5P). One hundred microliters of the AgNP working solutions were added to the cell cultures to make the final concentrations of 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, and 6.0 μg/mL AgNPs. The cells were gassed with 5 % (v/v) CO2 in air and placed on a roller drum (15 rpm) in a 37 °C incubator for 4 h. Two untreated controls and one positive control (0.1 μg/mL NQO) were included in this experiment. After the treatment, the cells were centrifuged at 78 × g for 5 min 338 Tao Chen et al.

and washed twice with fresh growth medium (F10P) and were then resuspended in growth medium at a density of 3 × 105 cells/ mL. The culture tubes were placed on a roller drum in a 37 °C incubator to begin the 2-day phenotypic expression. The cell con-

centrations were counted 24 h after the incubation, and SG1 was calculated accordingly. The cells were then adjusted to 2 × 105 cells/ mL and were cultured for another 24 h. Following the 2-day expression period, the cell concentration for each culture was

counted. The SG2 and SG were determined (Table 2). Cloning and MF Calculation. The treated cells were collected and divided into two parts to perform mutant selection experiments using both the microwell and soft agar versions of the MLA. In the soft agar version, 1 μg/mL of TFT was added in the cloning medium to enumerate the mutants. Cells were suspended in 100 mL of cloning medium with 0.28 % agar and poured onto three 100-mm-diameter tissue culture dishes to give a final density of 30,000 cells/mL. Six hundred cells were suspended in 100 mL of 0.28 % soft agar cloning medium for determining plating efficiency. In the microwell version, 3 μg/mL of TFT was added to the cells in the cloning medium (F20P) for mutant enumeration, and the cells were seeded into four 96-well flat-bottom microtiter plates using 200 μL per well and a final density of 2,000 cells/well. For the determination of plating efficiency, the cultures were adjusted to 8 cells/mL medium and aliquoted in 200 μL per well into two 96-well flat-bottom microtiter plates. All 96-well plates and 100-mm tissue culture dishes were

incubated at 37 °C in a humidified incubator with 5 % CO2 in air. After 11 days of incubation, colonies were counted, and mutant colonies were categorized as small or large. For the microwell version, counting was performed visually, and the small colonies were defined as those smaller than 25 % of the diameter of the well. MFs were calculated using the Poisson distribution (see Sect. 3.5). For the soft agar version, colony counting and sizing were performed using an automatic colony counter fitted with the capability to evaluate the size of the colonies. Mutant colonies approximately <0.6 mm in diameter were considered to be small-colony mutants. Cytotoxicity was measured as RTG (Tables 3 and 4). Data Interpretation. For the soft agar version, the average MF was 49 × 10–6; the average cloning efficiency was 93 %; and the average suspension growth was 28.9 for the two negative controls (Tables 2 and 3). The results for the microwell version were also within the normal range. The positive control (NQO) also significantly increased the MFs, and the small-colony mutants were optimally detected for both of the versions. These values indicate that the assays are acceptable. The induced MFs were 506.3 × 10−6 for the soft agar version and 992.5 × 10−6 for the microwell version in the highest dose at which the RTGs were larger than The Mouse Lymphoma Assay 339 % SC 49 66 67 57 63 76 81 81 84 50 ) −6 version 15.9 22.8 11.7 15.7 22.8 16.2 35.4 40.2 81 95 relative total growth rate, LC MF (×10 ) −6 RTG 15.2 44.3 23.7 20.8 38.9 51.4 95 150.9 171.4 425.3 SC MF (×10 45 93 71 117 125 154 362 442 676 239 SC # ) 4-nitroquinoline-1-oxide as the positive control −6 NQO 31.1 67 35.4 36.6 61.7 67.7 186.3 211.6 506.3 190.1 MF (×10 relative plating efficiency for viability, relative plating efficiency for viability, ) V −6 RPE (×10 M 30.33 59 46.33 41.33 66.33 67.67 149 182 268.33 159.33 PE silver nanoparticles, AgNP 91 177 139 124 199 203 447 546 805 478 Colony # in S.P. large colony, colony, large 87 93 81 67 45 38 14 68 plating efficiency for viability, plating efficiency for viability, 113 110 RTG (%) RTG V LC PE (%) V 95 86 93 57 90 105 141 122 116 108 RPE small colony, small colony, V SC selection plates, PE 0.98 0.88 1.31 1.13 1.08 1 0.8 0.86 0.53 0.84 S.P. Colony # in non-S.P. 585 528 785 678 645 600 480 516 318 503 mutant frequency, mutant frequency, MF 92 78 77 70 62 52 41 24 75 RSG (%) 108 g/mL) relative suspension growth rate, μ Dose ( 0 0 3 3.5 4 4.5 5 5.5 6 0.1 RSG : 1 2 3 4 5 6 7 plating efficiency for mutants, M Culture I.D. Control-1 Control-2 AgNP- AgNP- AgNP- AgNP- AgNP- AgNP- AgNP- NQO PE Abbreviations Table 3 Table Mutant frequencies and relative total growth rate in mouse lymphoma cells treated with 5-nm silver NPs measured the soft agar 340 Tao Chen et al.

Table 4 Mutant frequencies and relative total growth rate in mouse lymphoma cells treated with 5-nm silver NPs measured with the microwell version

Well #

Culture Dose RSG with RPEV RTG Empty PEM MF SC SC MF LC MF % −6 −6 −6 −6 I.D. (μg/mL) (%) colony PEV (%) (%) well # (×10 ) (×10 ) # (×10 ) (×10 ) SC Control-1 0 108 157 1.06 102 110 358 35.05 33 15 19 13.9 58 Control-2 0 92 154 1.01 98 90 333 71.25 70.4 28 38.6 31.7 55 AgNP-1 3 78 173 1.45 139 108 336 66.77 46.2 32 30.8 15.4 67 AgNP-2 3.5 77 162 1.16 112 86 323 86.5 74.6 40 48.9 25.7 66 AgNP-3 4 70 173 1.45 139 98 311 105.42 72.9 47 47 26 64 AgNP-4 4.5 62 156 1.05 101 63 293 135.23 129.3 64 90.9 38.4 70 AgNP-5 5 52 157 1.06 102 53 220 278.51 261.8 122 194.7 67 74 AgNP-6 5.5 41 143 0.85 82 34 197 333.72 391 122 255.1 135.9 65 AgNP-7 6 24 102 0.47 46 11 150 470 992.5 170 721 271.5 73 NQO 0.1 75 148 0.92 89 67 216 287.68 312.4 79 146.9 165.5 47

Abbreviations: RSG relative suspension growth rate, PEV plating efficiency for viability,RPE V relative plating efficiency for viability, RTG relative total growth, PEM plating efficiency for mutants, MF mutant frequency, SC small colony, LC large colony, AgNP silver nanoparticles, NQO 4-nitroquinoline-1-oxide as the positive control

10 % (14 % and 11 %). This response meets the requirement of exceeding the GEF to define a positive response. In addition, the MFs were clearly increased dose-dependently for the both versions of the assay (Tables 3 and 4). Therefore, the 5-nm AgNP is clearly mutagenic in the MLA.

Declaration

The views presented in this article do not necessarily reflect those of the Food and Drug Administration.

References

1. Hozier J, Sawyer J, Clive D et al (1985) tk-/- mouse lymphoma mutants by loss of het- Chromosome 11 aberrations in small colony erozygosity (LOH) and by whole chromosome L5178Y TK-/- mutants early in their clonal 11 painting: detection of recombination. history. Mutat Res 147:237–242 Mutagenesis 13:461–474 2. Applegate ML, Moore MM, Broder CB et al 4. Chen T, Harrington-Brock K, Moore MM (1990) Molecular dissection of mutations at (2002) Mutant frequencies and loss of heterozy- the heterozygous thymidine kinase locus in gosity induced by N-ethyl-N-nitrosourea (ENU) mouse lymphoma cells. Proc Natl Acad Sci U S in the thymidine kinase (TK) gene of L5178Y/ A 87:51–55 Tk+/–3.7.2C mouse. Mutagenesis 17:105–109 3. Liechty MC, Scalzi JM, Sims KR et al (1998) 5. Chen T, Harrington-Brock K, Moore MM Analysis of large and small colony L5178Y (2002) Mutant frequency and mutational The Mouse Lymphoma Assay 341

spectra in the Tk and Hprt genes of 18. Majeska JB, Matheson DW (1990) N-ethyl-N-nitrosourea-treated mouse lym- Development of an optimal S9 activation mix- phoma cellsdagger. Environ Mol Mutagen ture for the L5178Y TK+/- mouse lymphoma 39:296–305 mutation assay. Environ Mol Mutagen 6. Mei N, Guo X, Moore M (2014) Methods for 16:311–319 using the mouse lymphoma assay to screen for 19. Honma M, Hayashi M, Shimada H et al (1999) chemical mutagenicity and photo-mutagenicity. Evaluation of the mouse lymphoma tk assay In: Caldwell GW, Yan Z (eds) Optimization in (microwell method) as an alternative to the drug discovery, vol 18. Humana, Totowa, NJ, in vitro chromosomal aberration test. pp 561–592 Mutagenesis 14:5–22 7. Schisler MR, Moore MM, Gollapudi BB 20. Honma M, Zhang LZ, Sakamoto H et al (2013) In vitro mouse lymphoma (L5178Y (1999) The need for long-term treatment in Tk(+)/(-)-3.7.2C) forward mutation assay. the mouse lymphoma assay. Mutagenesis Methods Mol Biol 1044:27–50 14:23–29 8. Moore MM, Clive D, Howard BE et al (1985) 21. Clements J (2000) The mouse lymphoma In situ analysis of trifluorothymidine-resistant assay. Mutat Res 455:97–110 (TFTr) mutants of L5178Y/TK+/- mouse 22. Moore MM, Honma M, Clements J et al lymphoma cells. Mutat Res 151:147–159 (2002) Mouse lymphoma thymidine kinase 9. Moore MM, Clive D, Hozier JC et al (1985) gene mutation assay: follow-up international Analysis of trifluorothymidine-resistant (TFTr) workshop on genotoxicity test procedures, mutants of L5178Y/TK+/- mouse lymphoma New Orleans, Louisiana, April 2000. Environ cells. Mutat Res 151:161–174 Mol Mutagen 40:292–299 10. Liechty MC, Rauchfuss HS, Lugo MH et al 23. Dearfield KL, Auletta AE, Cimino MC et al (1993) Sequence analysis of tka(-)-1 and (1991) Considerations in the U.S. Environmental tkb(+)-1 alleles in L5178Y tk+/- mouse- Protection Agency’s testing approach for muta- lymphoma cells and spontaneous tk-/- genicity. Mutat Res 258:259–283 mutants. Mutat Res 286:299–307 24. ICH (1995) Topic S2A genotoxicity: guidance 11. Liechty MC, Hassanpour Z, Hozier JC et al on specific aspects of regulatory genotoxicity (1994) Use of microsatellite DNA polymor- tests for pharmaceuticals. International confer- phisms on mouse chromosome 11 for in vitro ence on harmonisation of technical require- analysis of thymidine kinase gene mutations. ments for registration of pharmaceuticals for Mutagenesis 9:423–427 human use, harmonised tripartite guideline 12. Chen T, Moore MM (eds) (2004) Screening CPMP/ICH/141/95, approved September for chemical mutagens using the mouse lym- 1995 (see website http://www.ifpma.org/ phoma assay. Humana, Totowa, NJ ich1.html) 13. Clive D, Flamm WG, Machesko MR et al 25. HHS (1997) Genotoxicity: a standard battery (1972) A mutational assay system using the for genotoxicity testing of pharmaceuticals. thymidine kinase locus in mouse lymphoma International conference on harmonization of cells. Mutat Res 16:77–87 technical requirements for registration of phar- 14. Cole J, Arlett CF, Green MH et al (1983) A maceuticals for human use. Food and Drug comparison of the agar cloning and microtitra- Administration, Rockville, MD tion techniques for assaying cell survival and 26. FDA (2001) Redbook 2000, toxicological mutation frequency in L5178Y mouse lym- principles for the safety of food ingredients: phoma cells. Mutat Res 111:371–386 IV.C.1.c. Mouse lymphoma thymidine kinase 15. Clive D, Spector JF (1975) Laboratory proce- gene mutation assay. http://www.fda.gov/ dure for assessing specific locus mutations at Food/GuidanceRegulation/ the TK locus in cultured L5178Y mouse lym- GuidanceDocumentsRegulatoryInformation/ phoma cells. Mutat Res 31:17–29 IngredientsAdditivesGRASPackaging/ ucm078336.htm 16. Moore-Brown MM, Clive D, Howard BE et al (1981) The utilization of trifluorothymidine 27. Moore MM, Howard BE (1982) Quantitation (TFT) to select for thymidine kinase-deficient of small colony trifluorothymidine-resistant (TK-/-) mutants from L5178Y/TK+/- mouse mutants of L5178Y/TK+/- mouse lymphoma lymphoma cells. Mutat Res 85:363–378 cells in RPMI-1640 medium. Mutat Res 104:287–294 17. Moore MM, Clive D (1982) The quantitation of TK-/- and HGPRT- mutants of L5178Y/ 28. Amacher DE, Turner GN (1980) Promutagen TK+/- mouse lymphoma cells at varying activation by rodent-liver postmitochondrial times post-treatment. Environ Mutagen 4: fractions in the L5178Y/TK cell mutation 499–519 assay. Mutat Res 74:485–501 342 Tao Chen et al.

29. Meyer M, Brock K, Lawrence K et al (1986) 32. Sadiq R, Bhalli JA, Yan J et al (2012) Evaluation of the effect of agar on the results Genotoxicity of TiO(2) anatase nanoparticles obtained in the L5178Y mouse lymphoma in B6C3F1 male mice evaluated using Pig-a assay. Environ Mutagen 8:727–740 and flow cytometric micronucleus assays. 30. Moore MM, Honma M, Clements J et al Mutat Res 745:65–72 (2006) Mouse lymphoma thymidine kinase 33. Mei N, Zhang YB, Chen Y et al (2012) Silver gene mutation assay: follow-up meeting of nanoparticle-induced mutations and oxidative the international workshop on genotoxicity stress in mouse lymphoma cells. Environ Mol testing—Aberdeen, Scotland, 2003—assay Mutagen 53:409–419 acceptance criteria, positive controls, and 34. Li Y, Chen DH, Yan J et al (2012) Genotoxicity data evaluation. Environ Mol Mutagen of silver nanoparticles evaluated using the 47:1–5 Ames test and in vitro micronucleus assay. 31. Moore MM, Honma M, Clements J et al (2003) Mutat Res 745:4–10 Mouse lymphoma thymidine kinase gene muta- 35. Woodruff RS, Li Y, Yan J et al (2012) tion assay: international workshop on genotox- Genotoxicity evaluation of titanium dioxide icity tests workgroup report—Plymouth, UK nanoparticles using the Ames test and Comet 2002. Mutat Res 540:127–140 assay. J Appl Toxicol 32:934–943 Chapter 20

Bhas 42 Cell Transformation Assay for Genotoxic and Non-Genotoxic Carcinogens

Kiyoshi Sasaki , Anna Huk , Naouale El Yamani , Noriho Tanaka , and Maria Dusinska

Abstract

Cell transformation assays (CTAs) are in vitro carcinogenicity tests measuring morphological transformation of cells either as transformed colonies (SHE cells) or foci (C3H/10T1/2 and BALB/c 3T3 including Bhas 42 cells) derived from a single cell. CTAs such as Bhas 42 CTA can detect both genotoxic and non- genotoxic carcinogens. When used as an initiation assay to test tumor-initiating activity, cells at low density are treated with a test chemical for 3 days, whereas a promotion assay to test for tumor-promoting activity, near-conﬂ uent cells are treated with a test chemical for a period of 10 days. The Bhas 42 CTA has advantages compared with BALB/c 3T3 and other CTAs due to its simplicity, higher sensitivity, less time needed for assay performance, and robustness (exempliﬁ ed by its adaptation to a high-throughput method). The Bhas 42 CTA has been validated together with other CTAs and recommended for development of an OECD guideline. The assay has already been applied in testing various chemical and physical agents including particles and nanomaterials. Protocols for both 6 and 96-well plate formats of initiation and promotion assays are described in detail.

Key words Cell transformation assay , Bhas 42 , Carcinogenicity in vitro , Initiation , Promotion , Genotoxic carcinogens , Non-genotoxic carcinogens

1 Introduction

Long-term carcinogenicity assays have played a central role in the risk assessment of chemicals, but for ethical, economical, and practical reasons their use is diminishing. Consequently, short-term genotoxicity tests are used in pre-screening for carcinogenicity. These tests, however, are not sensitive enough to detect all genotoxic carcinogens and cannot detect non-genotoxic carcinogens. In contrast, in vitro cell transformation assays (CTAs) have the potential to detect both non-genotoxic as well as genotoxic carcinogens.

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_20, © Springer Science+Business Media New York 2014 343 344 Kiyoshi Sasaki et al.

1.1 Principle of CTA Two types of cells can be used for detecting morphological transformation after exposure to test chemicals: primary Syrian hamster embryo (SHE) cells [ 1 ], or mouse cell lines such as BALB/c 3T3 [ 2 ] or C3H/10T1/2 [3 ]. A CTA of SHE cells measures the conversion of normal mammalian cells to morphologically transformed cells [ 4 ]. The morphologically transformed cells in colonies are characterized by the loss of density-dependent regulation of growth and the formation of colonies with criss-crossed and piled-up cells that are not observed in untreated controls. Subsequent passage of the cells from the aberrant colonies allows the expression of other characteristic phenotypes of malignant transformation, which include growth in soft agar and tumor formation in syngeneic hosts. On the other hand, CTA of a cell line measures the conversion of morphologically non-transformed cells to morphologically transformed cells [ 5 , 6 ]. The morphologically transformed cells in foci are characterized by criss-crossed, piled- up, and spindle-shaped cells that are observed in contact-inhibited non-transformed cells which show a monolayer. Unlike SHE cells, the transformed cells from the foci show growth in soft agar and tumor formation in hosts without subsequent passage.

1.2 Bhas 42 CTA Bhas 42 cells were established by Sasaki et al. [7 ] from the BALB/c 3T3 cells (derived from BALB/c mouse whole embryo), transfected with v-Ha- ras gene. The cells are derived from a morphologically non-transformed, TPA (12- O -tetradecanoylphorbol 13-acetate)-sensitive clone [ 8 ]. The end point of Bhas 42 CTA is the formation of morphologically transformed foci derived from a single cell. While the Bhas 42 cells are regarded as already initiated (according to the two-stage paradigm of carcinogenesis), the assay can predict both tumor-initiating and tumor-promoting activities of chemical carcinogens [ 9 , 10 ]. To detect promoting the activity of chemicals, the BALB/c 3T3 CTA needs pre-treatment with an initiator but the Bhas 42 CTA does not need such pre-treatment. To distinguish between initiation and promotion, the initiation assay involves a 3-day treatment of low density cells, obtained 1 day after plating, with a test chemical, whereas the promotion assay involves treatment of near-conﬂ uent cells with a test chemical for a period of 10 days (Day 4–14) (Fig. 1 ) [11 ]. The ability of the Bhas 42 assay to detect genotoxic and non-genotoxic carcinogens was evaluated on 98 chemicals [ 10 ]. The assay was compared with the BALB/c 3T3 transformation assay [12 , 13 ], and pre-validated and validated in interlaboratory studies using two protocols, for the 6- and 96-well plates [ 14 – 17 ]. The procedures are in principle the same for both formats. The cells are apportioned relative to the size of wells, and according to whether initiation or promotion is of interest; they are cultured and treated with a test chemical and the production of transformed foci is observed. The Bhas 42 CTA Bhas 42 Cell Transformation Assay 345

100

30 cells/Well) 4 10

No. of cells (x 10 0.3 0246810 Days after plating

Treatment period

Initiation assay

Promotion assay Continue on Day 14 Fig. 1 Growth curve of Bhas 42 cells in 6-well plates in the initiation and promotion assays. Bhas 42 cells are plated at 0.4 × 104 cells (fi lled circle) and 1.4 × 104 cells (circle ) per well in the initiation and promotion assays, respectively. Chemicals are treated during the growth phase (fi lled square: on Day 1–4 after plating) in the initiation assay and during the stationary phase ( square : on Day 4–14 after plating) in the promotion assay. Inverted fi lled triangle : medium change

is reproducible and reliable and compared with the BALB/c 3T3, CTA is simpler, more economical, shorter and more sensitive with high speciﬁ city [18 ].

1.3 Validation of CTA The international validation study considered both SHE and BALB/c 3T3 including Bhas 42 models as scientiﬁ cally valid for assessing the carcinogenic potential of hazardous compounds [16 , 19 ], and to be suitable alternatives to the in vivo 2-year carcinogenicity test [ 19 , 20 ]. Based on the validation studies, the CTAs were recommended for development of OECD Test Guidelines [21 –24 ]. It was recently proposed that they be included in integrating testing strategies as efﬁ cient tools to identify genotoxic carcinogens and as stand-alone tests to detect non-genotoxic carcinogens [25 ].

1.4 Application The CTAs have been applied to screen various chemical and physical of CTAs as In Vitro agents for potential in vitro carcinogenicity [ 12 , 26 – 28 ] including Carcinogenicity Tests metals [29 , 30 ], and inorganic arsenic compounds [13 ]. The CTAs also appeared to be useful for testing particles and especially nanoparticles (NPs) [31 , 32 ]. As no interference of NPs with the assay has been reported, the assay does not need modiﬁ cation for testing nanomaterial. The Bhas 42 model is of particular interest also for its adaptation to high-throughput testing and robustness. The Bhas 42 CTA was used to test the particles present in cigarette smoke that are known to induce tumors in vivo and have 346 Kiyoshi Sasaki et al.

been shown to induce morphological transformation in vitro. A dose-dependent increase in type III foci, and a signiﬁ cant increase in focus formation at moderately toxic concentrations of these particles were found [ 33 ]. Ponti et al. [31 ] studied the cytotoxicity, genotoxicity and morphological transforming activity of cobalt NPs (CoNPs) and cobalt ions (Co 2+ ) in BALB/c 3T3 cells. Dose- dependent cytotoxicity for both compounds was found, though the toxicity was higher for CoNPs than for Co 2+ and a signiﬁ cant increase in the formation of morphologically transformed foci (type III) was detected only for CoNPs. Amorphous silica NPs

(aSiO2 NPs) have been internalized by BALB/c 3T3 mouse ﬁ broblasts without inducing cytotoxicity, genotoxicity, or morphological transformation [34 ]. Ponti et al. [35 ] investigated the toxicological

effects of nude and chemically functionalized (–NH2 , –OH, and – COOH groups) multiwall carbon nanotubes (mwCNTs) using BALB/c 3T3 CTA. Additionally, they assessed the basal cytotoxicity, genotoxicity, and interaction of cells with NPs. The authors found no cytotoxicity and genotoxicity but signiﬁ cant carcinogenic potential of mwCNTs. These results emphasized that different toxicological endpoints have to be considered when studying NPs: acute (immediate) effects such as cytotoxicity and genotoxicity, as well as long-term effects, such as carcinogenicity. Thus, the Bhas 42 CTA is of particular value for its ability to measure several endpoints in addition to morphological transformation.

2 Materials

2.1 Cell Line Bhas 42 cells, a clone of v-Ha-ras -transfected BALB/c 3T3 A31- 1- 1 cells [36 ]: JCRB0149, JCRB cell bank (Japanese Collection of Research Bioresources Cell Bank, Osaka, Japan), conﬁ rmed to be free of bacteria, fungi, and mycoplasma.

2.2 Reagents ● Eagle’s minimum essential medium (MEM): e.g., 11095-080, Life Technologies, Grand Island, NY. Store at 4 °C. ● Dulbecco’s modiﬁ ed Eagle’s medium/Ham’s F12 (DMEM/ F12): e.g., 11330-032, Life Technologies. Store at 4 °C. ● Fetal bovine serum (FBS): e.g., Life Technologies. Selected by showing a low spontaneous focus formation and a high focus formation in the positive control. Store at –80 °C. ● Penicillin–streptomycin 100× solution (PS): e.g., 10,000 units/ mL Penicillin G sodium and 10 mg/mL streptomycin sulfate, 15140-122, Life Technologies. Store at less than –20 °C. ● 0.25 % Trypsin: e.g., 15050-065, Life Technologies. Store at less than –20 °C. ● 0.02 % EDTA-PBS(−): e.g., 15040-066, Life Technologies. Store at room temperature. Bhas 42 Cell Transformation Assay 347

● WST-8: CK04, Dojindo Laboratories, Kumamoto Techno Research Park, Japan. Store at 4 °C. ● Giemsa solution: e.g., 1.09204, Merk, Darmstadt, Germany. Store at room temperature. ● Dimethyl sulfoxide (DMSO): e.g., D8418-50ML, Sigma- Aldrich. Store at room temperature. ● 3-Methylcholanthrene (MCA): e.g., 213942-100MG, Sigma- Aldrich. Store at less than –20 °C. ● 12-O -tetradecanoylphorbol 13-acetate (phorbor 12-myristate 13-acetate, TPA): e.g., P-1585-1MG, Sigma-Aldrich. Store at less than –20 °C.

2.3 Equipment ● Microplate reader: e.g., Vmax-S, Molecular Devices, Sunnyvale, CA. ● Inverted microscope: e.g., CKX31, Molecular Devices, Tokyo, Japan.

2.4 Reagent Set Up ● M10F (MEM supplemented with 10 % FBS): 500 mL MEM + 56 mL FBS + 5 mL PS. Used for the expansion of provided cells, cell storage, and the fi rst culture after thawing. Store at 4 °C. ● DF10F (DMEM/F12 supplemented with 10 % FBS): 500 mL DMEM/F12 + 56 mL FBS + 5 mL PS. Used for the expansion of provided cells, cell storage, and the fi rst culture after thawing. Store at 4 °C. ● DF5F (DMEM/F12 supplemented with 5 % FBS): 500 mL DMEM/F12 + 26.5 mL FBS + 5 mL PS. Used for routine passages, cell growth assay and transformation assay. Store at 4 °C. ● 5 % Giemsa solution: 95 mL pure water + 5 mL Giemsa solution. Needed 13–15 mL for one plate. Store at room temperature. ● Test chemicals : Dissolve or suspend in an appropriate solvent (usually distilled water/ultra pure water or DMSO) at a high concentration as a stock solution and store in aliquots at less than –20 °C. Before treatment, dilute to each individual concentration for 20× the fi nal concentration with distilled water/ ultra pure water and 1,000× the fi nal concentration with DMSO. Never refreeze the aliquots. ● Negative controls : The solvent for test chemicals. The fi nal concentration in the medium: 5 % for distilled water/ultra pure water and 0.1 % for DMSO (permissible up to 0.5 % when test chemicals do not dissolve). It is possible to use acetone but ethanol is not recommended because it easily affects cell functions [ 37 ]. ● Positive controls : MCA (fi nal concentration: 1 μg/mL) for the initiation assay and TPA (fi nal concentration: 50 ng/mL) for 348 Kiyoshi Sasaki et al.

the promotion assay. Prepare 1,000× the ﬁ nal concentration with DMSO as stock solutions (MCA: 1 mg/mL and TPA: 50 μg/mL) and store in aliquots at less than –20 °C. Never refreeze the aliquots.

3 Procedures

The protocol of the transformation assay is illustrated at Fig. 2 . The Bhas 42 CTA consists of the initiation assay and promotion assay [10 , 15 , 16 ]. The both assays do not necessarily apply to a chemical and either one or the other can be selectively used depending on the purposes. Between the initiation and promotion assays, only two points, the number of cells plated and the treatment procedure of the test chemicals, are different. In the initiation assay, a less number of cells are plated into the wells and the cells are treated with the test chemicals in the beginning of growth phase for a shorter period (3 days: from Day 1 to 4 after plating). On the other hand, in the promotion assay, a greater number of cells are plated, then the treatment is started at sub-conﬂ uence and continued during the stationary phase for a longer period (10 days: from Day 4 to 14 after plating). The entire process of the Bhas 42 CTA is similar to the other in vitro short-term assays such as chromosome aberration and mouse lymphoma assays, i.e., optimum doses are set by the cell

Days 0 4 7 11 14 21 Medium change

Initiation assay

Cell growth assay

Transformation assay

Treat at growth phase

Promotion assay

Cell growth assay

Transformation assay

Treat at stationary phase Fig. 2 Schedule of the Bhas 42 CTA. In the 6-well format, 0.4 × 104 and 1.4 × 104 cells are plated into wells in the initiation and promotion assays, respectively. In the 96-well format, 200 and 400 cells are plated into wells in the initiation and promotion assays, respectively. Square : medium containing chemicals, −: fresh medium Bhas 42 Cell Transformation Assay 349

growth assay as the fi rst step, and then, the transformation assay and the concurrent cell growth assay are performed as the second step. The concurrent cell growth assay is needed to confi rm the reproducibility of the cytotoxicity of the chemicals and check whether the transformation assay is properly performed at the optimum concentrations. In the cell growth assay, the cells are treated with the test chemicals in the same manner as in the transformation assay, except the shorter culture period. i.e., the cells are fi xed on Day 7 after plating (Fig. 2 ). Negative control (water or DMSO) and positive control (MCA for the initiation assay and TPA for the promotion assay) must be included in the transformation assay and the concurrent cell growth assay in all cases to confi rm induction of transformation by MCA and/or TPA. When the solvent of a test chemical is water, DMSO control is needed for MCA and/or TPA in addition to water control for the chemical. In the cell growth assay to set doses, MCA and/or TPA are not needed. Usually, M10F or DF10F is used for cell storage and the culture from thawing frozen cells to the fi rst passage, and DF5F is used in all other procedures. Because 10 % FBS allows the cells to attach and grow well and 5 % FBS reduces spontaneous transformation [ 12 ]. Although growth is a little worse, DF5F can be used for the fi rst culture after thawing. DF5F is never used for cell storage because 10 % or more concentration of FBS is essential for keeping high viability of frozen cells. 6-well and 96-well plates are available in the Bhas 42 CTA. Only two points, the number of cells plated and expression of transformation frequency, differ between the protocols of using 6-well and 96-well plates. The 6-well plate format is almost the same as other standard transformation assays (using 60-mm or 100-mm dishes) with only smaller changes of culture area, therefore the process of cell culture is not especial. On the other, the 96-well plate format requires fi ne manipulation since tiny 96 wells are used for cell culture with a small volume of medium. However, the 96-well plate format can be applied for a high-throughput assay using an automated system [ 38 ]. The following protocol focuses on the 96-well plate format and describes subsequently 6-well plate format.

3.1 Solubility Test A solubility test is needed to determine a solvent (distilled water/ ultra pure water or DMSO) and the maximum concentration of test chemicals in medium. Based on the maximum concentration of the stock solution, the maximum ﬁ nal concentration is calculated. 1. A small amount of the test chemical is weighted into two tubes (e.g., 10 mg/tube) for distilled water/ultra pure water and DMSO. 350 Kiyoshi Sasaki et al.

Table 1 An example of ﬁ nal concentrations of chemicals when chemical solutions are prepared at 10 mg/mL

Solvents Distilled water/ultra pure water DMSO

Maximum concentration of the stock 10 mg/mL (20× the fi nal 10 mg/mL (1,000× the solution concentration) fi nal concentration) Maximum fi nal concentration in medium 0.5 mg/mL 0.01 mg/mL Solvent concentration in medium 5 % 0.1 %

Make many tubes Test 1 Make many tubes

Test 2

Test 3

Master stock Working stock Fig. 3 Preparation of cell stocks for CTAs. To obtain reproducible results as many as possible, master stocks are prepared followed by working stocks, and then a tube from working stocks is thawed one by one for each transformation assay

2. Distilled water/ultra pure water or DMSO is added into its respective tube little by little until the chemical is dissolved or equally suspended. 3. The solvent with the better solubility is chosen and the maximum concentration of the stock solution is obtained. 4. The maximum ﬁ nal concentration is calculated as shown in Table 1 .

3.2 Preparation Using the same passaged cells is most important, not only for Bhas of Cell Stock 42 cells but also for other cell lines (BALB/c 3T3 and C3H10T1/2 cells), to obtain reproducible results of CTAs [26 , 39 ]. To prevent spontaneous transformation, Bhas 42 cells should be preserved at as low a passage as possible after obtaining the cells. In addition, a master stock and a working stock should be prepared (Fig. 3 ), and then, one frozen tube of the working stock should be thawed in each transformation assay (Fig. 4 ). 1. Frozen cells are thawed quickly in a 37 °C water bath. 2. After centrifugation (200–400 × g , 3–5 min) to remove DMSO, the cell pellet is suspended in 50 mL of M10F or DF10F. Bhas 42 Cell Transformation Assay 351

Thaw Passage Passage No

Test 1 Test 2 Test 3 Never use for transformation assay

Thaw Test 1

OK Test 2

Test 3

Fig. 4 How to use working stocks for CTAs. To obtain reproducible results, cells are used at the same and lower passage (one passage or without passage from thawing to plating for each transformation assay). Never use different or higher passaged cells. However, these cells can be used for the cell growth assay or biochemical experiments

3. The cells are plated into fi ve 100-mm plates (10 mL/plate) and cultured. 4. When one tube (2.5 × 105 cells) is thawed, usually the cultures become 50–70 % confl uent (2–2.5 × 106 cells/plate) in 5–6 days. At 50–70 % confl uence, the cells are trypsinized and suspended at a cell density of 5 × 10 5 cells/mL in M10F or DF10F containing 5 % DMSO. 5. The cells are frozen in 0.5 mL aliquots (2.5 × 105 cells/tube) at –80 ˚C for more than 2 h and stored in liquid nitrogen as a master stock (30–50 tubes). 6. One tube of the master stock is thawed to fi ve to ten 100-mm plates and a working stock (50–100 tubes) is prepared the same as the master stock. 7. Every transformation assay is started from the working stock. 8. After using up all the working stock, new one is made again from the master stock.

3.3 Cell Culture It is known that long-term culture at conﬂ uence, passages with and Passage high density plating and many passages causes spontaneous transformation in cell lines used in the focus formation method [26 , 39 ]. Also Bhas 42 cells should be passaged before conﬂ uency and used at low passage for the transformation assay. However, to save frozen cell stocks, the cells at higher passages can be used other than 352 Kiyoshi Sasaki et al.

the transformation assay, such as the cell growth assay and biochemical experiments. 1. Bhas 42 cells are cultured in 100-mm plates with 10 mL DF5F

in a humidiﬁ ed 5 % CO2 incubator at 37 °C. 2. At about 50–70 % conﬂ uence, medium is removed and the cells are washed once with about 10 mL of 0.02 % EDTA-PBS(-). 3. After adding 0.5–1 mL of 0.25 % trypsin into the plate, the plates are incubated in the incubator until the cells are detached. 4. The cells are suspended with 5–10 mL DF5F and the cell suspension is centrifuged (200–400 × g , 3–5 min) to remove trypsin. 5. The cells are re-suspended with appropriate volumes of DF5F and the cell concentration is adjusted. 6. In a convenient method, the cells are passaged once a week at 500–1,000-fold dilution without cell counting. Medium change is not needed for a week.

3.4 Preparation The medium containing test chemicals should be prepared before of Medium Containing treatment of the cells. All chemical treatment media contain an Test Chemicals equal concentration of the solvent. 1. Bottles (for 6-well plate format) or multi-channel pipette reservoirs (for 96-well plate format) are placed and DF5F is dispensed into them. 2. Chemical solutions are added to the medium.

3.5 Medium Change In 96-well format, draining off by inverting plates is recommended (96-Well Format) as a method of removing medium, rather than aspiration using Pasteur pipettes. The draining off method has the following advantages: (1) quick removal, (2) complete removal, and (3) removal without detaching cells by touching with Pasteur pipettes. 1. Diapers or paper towels are placed to catch the medium in a laminar ﬂ ow cabinet. In a clean room, no microbial contamination is observed even when medium is changed outside the cabinet. 2. The medium is removed by inverting the plates, i.e., the plates are swung and shaken three times to drain off the medium completely. 3. The medium is dispensed gently from tips touched to the upper walls of the wells.

3.6 Preliminary A preliminary cell growth assay is strongly recommended before Cell Growth Assay the cell growth assay for dose setting as the ﬁ rst step. The cell (96-Well Format) growth is inhibited at wide range concentrations (ng/mL–mg/mL) depending on test chemicals. Furthermore, some chemicals show Bhas 42 Cell Transformation Assay 353

steep dose response curves in the cell growth assay. Therefore, fi nding optimum concentrations is diffi cult by a single cell growth assay. The preliminary cell growth assay can cut down on repeated cell growth assays for dose setting. In the preliminary cell growth assay, the cell growth is judged visually for a few days. For example, many detergents lyse the cells immediately after treatment at high concentrations, and apoptosis- inducing chemicals change the morphology of cells drastically, such as cell shrinkage and pyknosis. The treatment at such concentrations is not needed in the growth assay for dose setting. Five to nine concentrations with a wide range are set (e.g., 0.03, 0.1, 0.3, 1, and 3 mg/mL). One well is used for each concentration. Day 0 1. The cells are plated (400 cells/well) in the same manner as the initiation assay (described below), except dispensing 0.05 mL of 8,000 cells/mL suspensions. Day 1 2. 0.05 mL of DF5F containing the test chemical (2× the fi nal concentration) is added. Day 1–4 3. The cells are observed everyday under an inverted microscope.

3.7 Initiation Assay Similar to other in vitro short-term assays, genotoxic chemicals (96-Well Format) have a tendency to induce the highest response at concentrations showing severe cytotoxicity in the Bhas 42 CTA. Therefore, a fi nd- 3.7.1 Cell Growth Assay ing of a concentration near IC90 is important. Five to nine con- for Dose Setting centrations are set. Eight wells are used for each group. Day –3 1. At about 50–70 % confl uence, 0.7–1 × 105 cells are plated into 100-mm plates with 10 mL of DF5F. Day 0 2. At about 50–70 % confl uence, the cells are plated (200 cells/ well) by dispensing 0.05 mL of 4,000 cells/mL suspensions into 96-well plates. 3. Because of the small volume of medium, the plates are tapped lightly to spread the cells. 4. After keeping the plates at room temperature for 15 min until cell attachment, the plates are placed into an incubator. Day 1 5. 0.05 mL of DF5F containing the test chemical (2× the fi nal concentration) is added. 354 Kiyoshi Sasaki et al.

Day 4 6. The medium is changed with 0.1 mL fresh DF5F. Day 7 7. 0.1 mL of 10 % WST-8 medium (ﬁ nal concentration: 5 %) is added and the cells are incubated for 3 h. 8. The plates are shaken for several seconds (shaking function of a microplate reader is recommended) and the absorbance at 450 nm is measured. 9. The absorbance of each well is corrected by subtracting that of blank wells. The growth rates relative to the solvent control culture are calculated. Usually wells with medium only are used as blank wells, however, wells with medium containing a chemical are used as blank when the color of WST-8 is changed by the chemical.

3.7.2 Transformation Five to nine concentrations are set based on the results of the cell Assay and Concurrent Cell growth assay. These concentrations cover a range from little or no Growth Assay cytotoxicity to the highest cytotoxicity (less than 20 % survival compared to the control). Ideally, one concentration below NOEL (no observable effect level, 20 % above or below the control), two concentrations between NOEL and IC50 and two concentrations between IC50 and IC90 are assessed. Dilution factor should not be greater than square root 10, e.g., acceptable: 0.1, 0.3, 1, and 3 mg/mL, not acceptable: 0.003, 0.03, 0.3, and 3 mg/mL. Eight wells and 96 wells are used for each group in the cell growth assay and the transformation assay, respectively. Day –6 or –7 1. Frozen cells (2.5 × 105 cells/tube) of the working stock are thawed into two to ﬁ ve 100-mm plates with M10F or DF10F. Day –3 2. The cells are passaged in the same manner as the cell growth assay. Day 0 3. The cells are replated for the transformation assay and the concurrent cell growth assay in the same manner as the cell growth assay. Day 1 4. The cells are treated with a chemical in the same manner as the cell growth assay. Day 4 5. The medium is changed in the same manner as the cell growth assay. Bhas 42 Cell Transformation Assay 355

Day 7 6. Transformation assay: The medium is changed with 0.1 mL fresh DF5F. 7. Concurrent cell growth assay: The absorbance is measured in the same manner as the cell growth assay. Day 10 or 11, and 14 8. The medium is changed with 0.1 mL fresh DF5F. Day 21 9. After removing the medium, the cells are ﬁ xed with 0.1– 0.2 mL of methanol for at least 10 min. 10. After removing the methanol, the cells are stained with 0.1– 0.2 mL of 5 % Giemsa solution for at least 30 min, then the cells are washed with water. It is not necessary to rinse the cells after treatment with methanol.

3.8 Promotion Assay Tumor promoters affect different responses to the cell growth (96-Well Format) depending on chemical properties, i.e., enhance the cell growth dose-dependently (e.g., TPA), inhibit the cell growth dose- 3.8.1 Cell Growth Assay dependently (e.g., okadaic acid) and enhance at low concentrations for Dose Setting but inhibit at high concentrations (e.g., lithocholic acid). Because optimum concentrations for the transformation assay vary depending on the growth curve, accurate ranges of enhancement and/or inhibition should be obtained. Five to nine concentrations are set. Day –3 1. The cells are passaged in the same manner as the initiation assay. Day 0 2. The cells are plated (400 cells/well) in the same manner as the initiation assay except dispensing 0.1 mL of 4,000 cells/mL suspension. Tapping the plates is not needed because of a sufﬁ cient volume of medium. Day 4 3. The medium is changed with 0.1 mL DF5F containing the test chemical (1× the ﬁ nal concentration). Day 7 4. The absorbance is measured in the same manner as the initiation assay.

3.8.2 Transformation Five to nine concentrations are set based on the results of the cell Assay and Concurrent Cell growth assay. The concentrations determined in the cell growth Growth Assay assay are sometimes not suitable for the transformation assay 356 Kiyoshi Sasaki et al.

because of the different treatment period between the cell growth assay for dose setting (3 days) and the promotion assay (10 days). Therefore, concentrations should be set with particular care [ 40 ]. For test chemicals that exhibit enhancement and inhibition of cell growth, optimum concentrations are selected to cover a wide range from little or no effect to inhibition interposing enhancement. Ideally, one concentration below NOEL, three concentrations in the range of growth enhancement, and one concentration in the range of weak growth inhibition are assessed. For test chemicals that inhibit cell growth dose-dependently, optimum concentrations are selected from little or no cytotoxicity to moderate cytotoxicity (less than 50 % survival compared to the control). Ideally, two concentrations below NOEL, two concentrations between NOEL and IC50 and one concentration above IC50 are assessed. Similar to the initiation assay, the dilution factor should not be greater than square root 10. Day –6 or –7 1. The cells are thawed in the same manner as the initiation assay. Day –3 2. The cells are passaged as described in the initiation assay. Day 0 3. The cells are replated (400 cells/well) in the same manner as the initiation assay except dispensing 0.1 mL of 4,000 cells/mL suspension. Tapping the plates is not needed because of a suffi cient volume of medium. Day 4 4. The medium is changed with 0.1 mL DF5F containing the test chemical (1× the fi nal concentration). Day 7 5. Transformation assay: The medium is changed with 0.1 mL DF5F containing the test chemical. 6. Concurrent cell growth assay: The absorbance is in the same manner as the initiation assay. Day 10 or 11 7. The medium is changed with 0.1 mL DF5F containing the test chemical. Day 14 8. The medium is changed with 0.1 mL fresh DF5F. Day 21 9. The cells are fi xed and stained as described in the initiation assay. Bhas 42 Cell Transformation Assay 357

Fig. 5 Various morphological types of Bhas 42 transformed foci. Each transformed focus has the following distinctive characteristics, (a ) piled-up, spindle-shaped, and criss-crossed, (b ) piled-up, (c ) spindle-shaped, and (d) only high density (not counted as a focus)

3.9 Focus Count The transformed foci are counted visually under an inverted microscope. Transformed foci have the following morphological characteristics: (a) more than 100 cells, (b) spindle-shaped cells, (c) deep basophilic staining, (d) criss-cross, (e) piling up, and (f) invasive growth into the monolayer of surrounding contact-inhibited cells. All transformed foci do not have all these characteristics (Fig. 5 ) and a certain level of subjective judgments cannot be avoided (Fig. 6 ). Because the morphology of transformed foci of Bhas 42 cells is the same as that of BALB/c 3T3 cells, it is preferable to refer to the photo catalogue of BALB/c 3T3 transformed foci [ 41 ]. The number of wells having transformed foci is recorded, i.e., a well having one focus is counted as one and a well having two or more foci is also counted as one. The transformation frequency is expressed as the number of wells having transformed foci per 96 wells (Fig. 6 ).

3.10 6-Well Format In the 6-well format, all procedures are the same as the 96-well format except the number of cells plated and the expression of transformation frequency as follows: 1. The cells, 4,000 and 14,000 cells/well, are plated for the initiation assay and the promotion assay in 2 mL medium, respectively. 358 Kiyoshi Sasaki et al.

Fig. 6 Difference of expression of transformation frequency between 6-well and 96-well formats. In the 6-well format, transformation frequency is expressed as the number of transformed foci per well. On the other hand, in the 96-well format, transformation frequency is expressed as the number of wells having transformed foci per 96 wells. Transformed foci with various sizes and morphological types are observed. Even if difﬁ cult to judge transformed foci ( B is not known whether it is an original focus or a daughter focus derived from A , a larger foci showing the same morphology as B. C is lower piled-up.), the well is counted as one well having foci quickly if there is one typical focus (D )

2. On Day 1 after plating in the initiation assay, chemical solutions are added into wells. When using distilled water/ultra pure water as a solvent, 0.1 mL of chemical solutions is added. When DMSO is used, 2 μL of chemical solutions is added. However, to lower the error, adding of 20 μL of the chemical solutions diluted 10-fold with medium is preferable. It is also possible to replace the medium dispensed on Day 0 with the medium containing test chemicals. 3. In the cell growth assay, after adding WST-8 100 μL (ﬁ nal concentration: 5 %) into wells with 2 mL medium, 0.1 mL of the medium is transferred to 96-well plates then the absorbance is measured. 4. The medium is aspirated by Pasteur pipettes. When the cells are not conﬂ uent (e.g., on Day 4 of the initiation assay), Pasteur pipettes are contacted, the bottom of wells and medium is removed completely. Never contact the bottom with Pasteur pipettes (e.g., on Day 7 of the initiation assay) to prevent scratching the cells. No problem even if a small volume of the medium is remaining. 5. Transformation frequency is expressed as the number of transformed foci per well (Fig. 6 ). Bhas 42 Cell Transformation Assay 359

3.11 Test The initiation or promotion assay is repeated independently, as Acceptance of Criteria needed (e.g., microbial contamination, inappropriate concentrations, and technical problems). If the cells are not conﬂ uent due to cytotoxicity at the end of transformation assay, the concentration is invalid.

3.12 Statistical After scoring foci, averages are calculated and results are assessed Analysis by an appropriate statistical analysis (e.g., 96-well format: chi- square test with Bonferroni adjustment, 6-well format: Dunnett's test).

3.13 Notes 1. Other plates (12-well plate, 24-well plate, 60-mm plate, etc.) and Troubleshooting can be used for the Bhas 42 CTA [42 ]. Of any plate format, it Advice should be examined when the cells become confl uent and the number of cells plated is determined (Fig. 1 ). 2. In the 96-well format, transformation frequency can be expressed as the number of transformed foci per well [ 43 ]. However, expression as the number of wells having transformed foci per 96 wells is easier because the well can be counted as positive without hesitation as long as one focus showing typical transformed cells is observed in some foci with various morphological types (Fig. 6 ) [38 ]. 3. In the cell growth assay, other dyes monitoring metabolic activity of living cells (MTT, neutral red, alamarBlue, etc .) or staining dead cells (crystal violet) can be used in place of WST- 8. Although almost the same results are obtained no matter the chosen dye, it is preferable to use suitable dyes based on features of some chemicals, e.g., WST-8: the cells are dead but not detached, crystal violet: the color of WST-8 is changed by the chemical. 4. In the case of crystal violet, after fi xing the cells with methanol, the cells are stained with 0.1 % crystal violet solution for 15 min, the stained dye was extracted with extraction solution (containing 0.02 mol/L HCl and 50 % ethanol) for 10 min, then the absorbance is measured at 540–570 nm [38 ]. 5. Recently, hydrogen peroxide method, a spectrophotometric quantitative transformation assay using a cell viability chromo- genic reagent, has been developed [ 41 ]. The method is based on that normal cells are selectively killed by treatment with hydrogen peroxide. Briefl y, after induction of transformation, the cells are treated with 0.0015 % hydrogen peroxide for 24 h (Fig. 7 ) followed by a dye monitoring metabolically active cells, such as WST-8 and alamarBlue, and then the absorbance or fl uorescence is measured. 6. Possible problems and their solutions are referred in Table 2 . 360 Kiyoshi Sasaki et al.

Fig. 7 Selection of Bhas 42 transformed foci by treatment with hydrogen peroxide. When Bhas 42 cells with a transformed focus (a ) are treated with 0.0015 % hydrogen peroxide for 24 h, normal cells are selectively killed (b ). Both the pictures show the same position

Table 2 Troubleshooting and advices

Problems Possible causes Solutions

Growth is slow Serum is not suitable Change the lot of serum Medium contains something Remove it (e.g., fungicide) Foci are detached Dispensing of medium is strong Dispense medium gently Dispense medium to the walls of wells Use wide-bore pipette Spontaneous Serum is not suitable Change the lot of serum transformation frequency Culture period (3 weeks) is long Shorten the culture period (2.5 weeks) is high Cultures contain transformed Prepare a working stock again, obtain cells another lot or obtain cells from another source Transformation frequency Serum is not suitable Change the lot of serum induced by positive Culture period (3 weeks) is Extend the culture period (3.5–4 weeks) controls is low short Chemicals are inactivated Prepare them freshly Do not increase the concentrations to conﬁ rm the sensitivity v-Ha-ras transfected cells are lost A low possibility, because 100 % Bhas 42 cells contain v-Ha-ras at high passages A microplate reader records Incubation period (3 h) is long Shorten the incubation period (2–2.5 h) a maximum value Bhas 42 Cell Transformation Assay 361

Acknowledgements

This work was supported by the EC FP7 QualityNano [INFRA-2010-1.131], contract no: 214547-2, EC FP7 NANoREG, [NMP.2012.1.3-3] contract no: 310584, and by the NEDO grant (New Energy and Industrial Technology Development Organization).

References

1. Berwald Y, Sachs L (1963) In vitro cell trans- 10. Sakai A, Sasaki K, Muramatsu D et al (2010) formation with chemical carcinogens. Nature A Bhas 42 cell transformation assay on 98 200:1182–1184 chemicals: the characteristics and performance 2. Kakunaga T (1973) A quantitative system for for the prediction of chemical carcinogenicity. assay of malignant transformation by chemical Mutat Res 702:100–122 carcinogens using a clone derived from 11. Sakai A (2008) BALB/c 3T3 cell transforma- BALB/3T3. Int J Cancer 12:463–473 tion assays for the assessment of chemical carci- 3. Reznikoff CA, Bertram JS, Brankow DW et al nogenicity. AATEX 14(Special Issue):367–373 (1973) Quantitative and qualitative studies of 12. Ohmori K, Sasaki K, Asada S et al (2004) An chemical transformation of cloned C3H mouse assay method for the prediction of tumor pro- embryo cells sensitive to postconﬂ uence inhibi- moting potential of chemicals by the use of tion of cell division. Cancer Res 33:3239–3249 Bhas 42 cells. Mutat Res 557:191–202 4. Evans CH, DiPaolo JA (1981) In vitro mam- 13. Muramatsu D, Sasaki K, Kuroda S et al (2009) malian cell transformation for identiﬁ cation of Comparison of sensitivity to arsenic com- carcinogens, cocarcinogens, and anticarcino- pounds between a Bhas 42 cell transformation gens. In: Stich HF, San RHC (eds) Short-term assay and a BALB/c 3T3 cell transformation tests for chemical carcinogens. Springer, assay. Mutat Res 675:66–70 New York, pp 306–322 14. Ohmori K, Umeda M, Tanaka N et al (2005) 5. Sakai A, Iwase Y, Nakamura Y et al (2002) Use An inter-laboratory collaborative study by the of a cell transformation assay with established Non-Genotoxic Carcinogen Study Group in cell lines, and a metabolic cooperation assay Japan, on a cell transformation assay for tumour with V79 cells for the detection of tumor pro- promoters using Bhas 42 cells. Altern Lab moters: a review. Altern Lab Anim 30:33–59 Anim 33:619–639 6. Schechtman LM (2012) Rodent cell transfor- 15. Tanaka N, Sasaki K, Hayashi K et al (2009) mation assays—a brief historical perspective. An interlaboratory collaborative study on a cell Mutat Res 744:3–7 transformation assay using Bhas 42 cells. 7. Sasaki K, Mizusawa H, Ishidate M (1988) AATEX 14:831–848 Isolation and characterization of ras-transfected 16. Sakai A, Sasaki K, Hayashi K et al (2011) An BALB/3T3 clone showing morphological trans- international validation study of a Bhas 42 cell formation by 12-O-tetradecanoyl-phorbol-13- transformation assay for the prediction of acetate. Jpn J Cancer Res 79:921–930 chemical carcinogenicity. Mutat Res 725: 8. Sasaki K, Mizusawa H, Ishidate M (1990) 57–77 Establishment of a highly reproducible trans- 17. Hayashi M, Kojima H, Corvi R et al (2012) formation assay of a ras-transfected BALB 3T3 Bhas 42 cell transformation assay validation clone by treatment with promoters. In: Kuroda study report. EURL ECVAM Y, Shankel DM, Waters MD et al (eds) Recommendations, Ispra, Italy Antimutagenesis and anticarcinogenesis mech- 18. Ohmori K (2009) In vitro assays for the predic- anisms II, vol 52, Basic life sciences. Plenum, tion of tumorigenic potential of non-genotoxic New York, pp 411–416 carcinogens. J Health Sci 55:20–30 9. Asada S, Sasaki K, Tanaka N et al (2005) 19. Vanparys P, Corvi R, Aardema MJ et al (2012) Detection of initiating as well as promoting Application of in vitro cell transformation activity of chemicals by a novel cell transforma- assays in regulatory toxicology for pharmaceu- tion assay using v-Ha-ras-transfected BALB/c ticals, chemicals, food products and cosmetics. 3T3 cells (Bhas 42 cells). Mutat Res 588:7–21 Mutat Res 744:111–116 362 Kiyoshi Sasaki et al.

20. Balls M, Clothier R (2010) A FRAME response chloride: an in vitro study in Balb/3T3 mouse to the Draft Report on Alternative (Non- fi broblasts. Mutagenesis 24:439–445 animal) Methods for Cosmetics Testing: 32. Ohmori K, Sato Y, Nakajima D et al (2013) Current Status and Future Prospects—2010. Characteristics of the transformation frequency Altern Lab Anim 38:345–353 at the tumor promotion stage of airborne par- 21. Corvi R, Aardema MJ, Gribaldo L et al (2012) ticulate and gaseous matter at ten sites in Japan. ECVAM prevalidation study on in vitro cell Environ Sci Process Impacts 15:1031–1040 transformation assays: general outline and con- 33. Weisensee D, Poth A, Roemer E et al (2013) clusions of the study. Mutat Res 744:12–19 Cigarette smoke-induced morphological trans- 22. Maire MA, Pant K, Poth A et al (2012) formation of Bhas 42 cells in vitro . Altern Lab Prevalidation study of the Syrian hamster Anim 41:181–189 embryo (SHE) cell transformation assay at pH 34. Uboldi C, Giudetti G, Broggi F et al (2012) 7.0 for assessment of carcinogenic potential of Amorphous silica nanoparticles do not induce chemicals. Mutat Res 744:64–75 cytotoxicity, cell transformation or genotoxic- 23. Pant K, Bruce SW, Sly JE et al (2012) ity in Balb/3T3 mouse fi broblasts. Mutat Res Prevalidation study of the Syrian hamster 745:11–20 embryo (SHE) cell transformation assay at pH 35. Ponti J, Broggi F, Mariani V et al (2013) 6.7 for assessment of carcinogenic potential of Morphological transformation induced by chemicals. Mutat Res 744:54–63 multiwall carbon nanotubes on Balb/3T3 cell 24. Tanaka N, Bohnenberger S, Kunkelmann T model as an in vitro end point of carcinogenic et al (2012) Prevalidation study of the BALB/c potential. Nanotoxicology 7:221–233 3T3 cell transformation assay for assessment of 36. Kakunaga T, Crow JD (1980) Cell variants carcinogenic potential of chemicals. Mutat Res showing differential susceptibility to ultraviolet 744:20–29 light-induced transformation. Science 209: 25. Benigni R, Bossa C, Tcheremenskaia O (2013) 505–507 In vitro cell transformation assays for an inte- 37. Pontes H, Carvalho M, de Pinho PG et al grated, alternative assessment of carcinogenicity: (2008) Ethanol, the forgotten artifact in cell a data-based analysis. Mutagenesis 28:107–116 culture. Arch Toxicol 82:197–198 26. IARC/NCI/EPA Working Group (1985) 38. Arai S, Sakai A, Hayashi K et al (2013) A high- Cellular and molecular mechanisms of cell throughput cell transformation assay applicable transformation and standardization of transfor- to automation for detecting potential chemical mation assays of established cell lines for the carcinogens using Bhas 42 cells. AATEX prediction of carcinogenic chemicals: overview 18:1–19 and recommended protocols. Cancer Res 39. Heidelberger C, Freeman AE, Pienta RJ et al 45:2395–2399 (1983) Cell transformation by chemical agents— 27. Kerckaert GA, Brauninger R, LeBoeuf RA et al a review and analysis of the literature. A report of (1996) Use of the Syrian hamster embryo cell the U.S. Environmental Protection Agency transformation assay for carcinogenicity predic- Gene-Tox Program. Mutat Res 114:283–385 tion of chemicals currently being tested by the 40. Arai S, Tanaka N, Sasaki K et al (2010) A study National Toxicology Program in rodent bioas- on the dose setting of test chemicals for the says. Environ Health Perspect 104:1075–1084 promotion assay in Bhas 42 cell transformation 28. Hasegawa G, Shimonaka M, Ishihara Y (2013) assay. AATEX 15:6–13 Differential genotoxicity of chemical proper- 41. Sasaki K, Sakai A, Tanaka N (2013) High- ties and particle size of rare metal and metal throughput quantifi cation of morphologically oxide nanoparticles. J Appl Toxicol 32:72–80 transformed foci in Bhas 42 cells (v-Ha- ras 29. Mazzotti F, Sabbioni E, Ponti J et al (2002) transfected Balb/c 3T3) using spectrophotom- In vitro setting of dose–effect relationships of etry. In: Steinberg P (ed) High-throughput 32 metal compounds in the Balb/3T3 cell line, screening methods in toxicity testing. Wiley, as a basis for predicting their carcinogenic New York, pp 317–339 potential. Altern Lab Anim 30:209–217 42. Suganuma M, Kurusu M, Suzuki K et al (2005) 30. Ponti J, Munaro B, Fischbach M et al (2007) New tumor necrosis factor-α-inducing protein An optimised data analysis for the BALB/c released from Helicobacter pylori for gastric 3T3 cell transformation assay and its applica- cancer progression. J Cancer Res Clin Oncol tion to metal compounds. Int J Immunopathol 131:305–313 Pharmacol 20:673–684 43. Suganuma M, Okabe S, Marino MW et al (1999) 31. Ponti J, Sabbioni E, Munaro B et al (2009) Essential role of tumor necrosis factor α (TNF-α) Genotoxicity and morphological transforma- in tumor promotion as revealed by TNF-α- tion induced by cobalt nanoparticles and cobalt defi cient mice. Cancer Res 59:4516–4518 Part II

DNA Repair Assays Chapter 21

Methods for Measuring DNA Repair: Introduction and Cellular Repair

Andrew R. Collins and Amaya Azqueta

Abstract

DNA repair pathways provide a critically important cellular defence system, effectively protecting us from mutations and cancer. Distinct pathways deal with various classes of damage: single- and double-strand breaks, oxidized and alkylated bases, bulky adducts, intra- and inter-strand cross-links. A simple approach to measuring the DNA repair capacity of cell lines, or samples of blood cells, for instance, is the cellular repair assay, or challenge assay, in which cells are treated with a speciﬁ c DNA-damaging agent, and incubated; at intervals, samples are taken and the residual damage is measured. The comet assay is well suited for measuring strand break rejoining, excision repair of oxidized or alkylated bases (with a lesion-speciﬁ c endonuclease to convert the altered bases into breaks), and nucleotide excision repair of UV-induced lesions (again, using an appropriate enzyme to detect the damage). An alternative way to assess nucleotide excision repair capability is the incision assay: after UV irradiation, cells are incubated with inhibitors of repair DNA synthesis, so that incomplete repair sites accumulate as DNA breaks. We provide protocols for these DNA repair assays, and discuss their applications—and their limitations. We also raise some important, so far unanswered questions concerning the regulation of repair, and the factors that might account for the wide variations seen in individual repair capacities.

Key words DNA repair , Strand break rejoining , Base excision repair , Nucleotide excision repair , Incision assay

1 Introduction

DNA-damaging agents abound, in the environment and in the body. Arguably, the most (potentially) dangerous damage we suffer is endogenous, in the form of free radicals (reactive oxygen species), simply because we respire. A small fraction of the oxygen passing down the respiratory chain is released from the mitochon- dria in the form of superoxide radicals [1 ]. Through evolution, we have very effective antioxidant defences, in the form of enzymes

such as superoxide dismutase that converts superoxide to H2 O2 and catalase that degrades the H2 O 2 to water, and also molecules such as glutathione that scavenge free radicals, becoming (reversibly)

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_21, © Springer Science+Business Media New York 2014 365 366 Andrew R. Collins and Amaya Azqueta

oxidized in the process. Some reactive oxygen does, however, reach the DNA, causing strand breaks (SBs) and oxidizing bases. In addition to endogenous agents, we are exposed to exogenous genotoxic chemicals and radiation, capable of causing a wide range of lesions in the DNA—single- and double-strand breaks (SSBs and DSBs), oxidized and alkylated bases, bulky adducts (large chemical groups covalently bound to bases), intra- and inter-strand cross-links, and DNA–protein cross-links [2 ]. Chemicals, once in the body, are likely to be intercepted by antioxidant defences and/or by the xenobiotic metabolizing system, mainly in the liver, consisting of “phase I enzymes” (cytochromes P450) which tend to oxidize compounds to a more reactive form, and “phase II enzymes” that conjugate these reactive intermediates to carrier molecules (including glutathione) prior to excretion [3 ]. Again, the defences are not 100 % effective, and some damage does occur. DNA repair is the body’s second line of defence, dealing with almost all the DNA damage that occurs in spite of the primary defences. Any damage that remains in the DNA when the cell enters S-phase is likely to result in blocked replication or mis- incorporation (for instance, 8-oxoguanine tends to pair with adenine rather than cytosine) [4 ]. The resulting mutation, if in a proto-oncogene or tumor suppressor gene, a gene related to cell proliferation control, or a gene involved in the cell’s defences (such as DNA repair itself), can lead to cell transformation and cancer. Thus DNA repair is of the utmost importance in protecting us from cancer. Distinct repair pathways deal with the various classes of damage. The simplest, and quickest, repair process is the rejoining of SSBs— essentially little more than ligation. Typically, SSBs induced experi-

mentally by ionizing radiation or H2 O2 have a half-life of several minutes [5 ]. Radiation and H2 O2 induce far more SSBs than DSBs, but the latter are potentially more serious, as they can disrupt chromosome structure. Two pathways are known for repair of DSBs; homologous recombination, limited to S-phase, in which the homologous chromatid provides a template which ensures correct pairing and sequence conservation; and nonhomologous end-joining, which ligates broken ends with a risk of (limited) loss of sequence; it is therefore error-prone and can lead to mutations [6 ]. DSB repair has a much slower time course than does SSB repair. The pathway known as base excision repair (BER) deals with small base changes, i.e., oxidation or alkylation [ 2 ]. One of a range of more or less speciﬁ c glycosylases removes the damaged base, leaving a baseless sugar (an apurinic/apyrimidinic or AP site). This is converted to a strand break by an AP lyase activity associated with the glycosylase, or by a separate AP endonuclease. The broken ends are cleaned up, and thus a small gap is created, which is Cellular DNA Repair Assays 367

fi lled by a DNA polymerase and ligated to the preexisting DNA strand. (In fact there are two variants of BER; so-called short-patch and long-patch BER) [7 ]. Bulky adducts, and dimerised pyrimidines (“intra-strand cross-links,” caused by short wavelength UV radiation), which cause helix distortion, are the substrates for nucleotide excision repair (NER), carried out by a complex of proteins including endonucleases that cut on each side of the lesion, removing the damage in an oligonucleotide and leaving a gap of about 28 nucleotides to be fi lled by DNA polymerase and ligase [2 ]. Repair of inter-strand cross-links is not completely understood, but they are probably removed by NER in two stages; excision of one terminal of the cross-link creates a gap in that strand, which must be fi lled before excision of the other terminal (otherwise a DSB will be created) [8 ]. Covalent links between DNA and protein are probably repaired by a combination of proteolysis and nucleotide excision [9 ]. In a few cases, damage is directly reversed, rather than repaired. A good example is the repair of O 6 -methylguanine, which involves a so-called suicide enzyme, methylguanine methyltransferase (MGMT), which transfers the methyl group to itself, becoming inactivated in the process [ 10 ]. Little is known about the regulation of DNA repair. It might be expected that BER, since it deals with ubiquitous, unavoidable oxidative damage, would be constitutively active, while NER, dealing with sporadic, relatively low frequency damage, is inducible via the activation of p53 tumor suppressor protein by DNA damage [11 ]. Regulation might occur at the level of transcription, or by post-translational enzyme modifi cation. The availability of deoxy- ribonucleotide precursors (dNTPs) can be a factor: peripheral blood mononuclear (PBMN) cells are able to incise at sites of UV-induced dimers, but incision events accumulate as DNA breaks, since the exiguous pool of dNTPs in these nonproliferating cells limits synthesis of the repair patch [ 12 ]. DNA damage (especially SBs and oxidized bases) is routinely measured in samples of PBMN cells (or sometimes leukocytes in whole blood) collected in human biomonitoring trials. What is measured is a dynamic steady state, since the input of damage is normally balanced by the output, i.e. repair. This underlines the importance of being able to measure individual DNA repair capacity and relating it to the level of DNA damage. Here, and in the following chapter, we describe two general approaches to the measurement of DNA repair in mammalian cells; assays involving the incubation of cells containing damage (cellular repair assays, this chapter), and assays in which a cell extract is incubated in vitro with a substrate of DNA containing specifi c lesions (Chap. 22 ). 368 Andrew R. Collins and Amaya Azqueta

2 Measuring DNA Repair by Monitoring the Removal of Lesions

The simplest DNA repair assay, at least conceptually, is the cellular repair assay, or challenge assay. Cells are treated with an appropriate DNA-damaging agent, and incubated under normal cell culture conditions. At intervals, samples are taken and the residual damage is measured. Repair of SSB, induced experimentally by X- or γ-irradiation,

or by treating cells with H 2 O2 , is readily followed using the comet assay (a quick, sensitive method based on the fact that strand breaks relax supercoiled loops of DNA, freeing them to extend towards the anode during electrophoresis, forming a comet-like image under ﬂ uorescence microscopy; the relative intensity of comet tail DNA indicates the break frequency) ( see Chap. 12 of this book). The yield of SSBs is many times higher than that of DSBs, after

irradiation or H2 O2 treatment, and so it is predominantly the rejoining of SSBs that is followed. The half-life of SSB in the cell is a matter of minutes (Fig. 1 ). A slow phase of break joining, after disappearance of almost all breaks, may be interpreted as the repair of DSBs. Note: Contrary to what is often stated, it is not possible to measure DSB repair speciﬁ cally by running the comet assay at neutral pH , since SSBs can relax DNA supercoiling even at neutral pH .

Fig. 1 Kinetics of rejoining of strand breaks induced by H2 O 2 , and repair of FPG-sensitive sites (8- oxoguanine) induced by light in the presence of photosensitizer. Bars represent standard deviation of results from three experiments. From [13 ] with permission from Elsevier Cellular DNA Repair Assays 369

8-Oxoguanine, and other oxidized bases, can be measured directly, using chromatographic techniques, but in exhaustive studies under the ESCODD project [ 14 ] it was shown that these methods were subject to an artifact of oxidation during sample preparation. In contrast, a modifi ed version of the comet assay seemed to be more accurate. The modifi cation is an additional step, in which the DNA is digested with a lesion-specifi c enzyme— formamidopyrimidine DNA glycosylase (FPG) from Escherichia coli , or the mammalian analogue 8-oxoguanine DNA glycosylase (OGG)—to convert the oxidized base to a SSB ( see Chap. 12 of this book). This approach, then, can be applied to studying BER, by following the removal of 8-oxoguanine from cells’ DNA using FPG or OGG. BER is slower than SSB repair, with a half-life for 8- oxoguanine (induced by treatment of cells with photosensitizer Ro 19-8022 plus light) measured in hours [ 5 ] (Fig. 1 ). Similarly, alkylated bases and their removal can be measured with an alkylation-specifi c enzyme (e.g., AlkA) [15 ]. Repair of UV-induced pyrimidine dimer repair is monitored with T4endoV [ 16 ]. In the outline protocols that follow, the comet assay is performed using either the standard format (1 or 2 gels per slide) or a high throughput version (12 gels per slide) as described in Chap. 12 of this book. Note: Doses of damaging agent are suggested , but should be checked with the particular cell type to be used ; the initial level of damage should ideally give around 50 – 60 % DNA in tail . Note : Numbers of dishes of cultured cells will depend on doses and incubation times chosen , and on the method , and so are not specifi ed .

2.1 Repair of SBs; 1. Treat cells (as monolayer or suspended in PBS) on ice with X-

Practical Details or γ-irradiation (suggested dose; 7 Gy), or with H 2 O2 in PBS (suggested dose; 100 μM) for 5 min. If monolayer cells are used, sufﬁ cient dishes should be prepared for all time points. Include untreated cells as controls.

(a) Alternatively, irradiation or incubation with H2 O2 can be performed after embedding the cells in agarose; cells remain viable after embedding, at least as far as their DNA repair capability is concerned [5 ]. This greatly reduces the possibility that cells might repair SSBs before the start of the “ofﬁ cial” repair incubation. One slide should be prepared for each time point. Place the whole slide in the irradiation chamber or in a Coplin staining jar containing cold

H2 O2 in PBS. 2. Rinse attached monolayer cells and add fresh medium. In the case of cells treated in suspension, centrifuge (700 × g , 5 min), and resuspend in culture medium. Place cells in a 37 °C incu-

bator and incubate for up to 1 h, taking samples at t0 and at 370 Andrew R. Collins and Amaya Azqueta

intervals of 5–10 min. Spin cells (after trypsinisation in the case of a monolayer), wash in PBS, spin again, suspend pellet in agarose, and prepare gels.

(a) If cells were irradiated or treated with H2 O2 after embedding, transfer the slides to medium in a Coplin jar in a 37 °C incubator, remove slides at intervals during the repair incubation period and transfer to lysis solution. 3. Perform lysis, alkaline incubation, and electrophoresis as usual [17 and Chap. 12 of this book].

2.2 Repair 1. Add to monolayer or suspended cells the photosensitizer Ro of Oxidized Bases 19-8022 (obtained from Hoffmann-La Roche) at 1 μM (from (BER): Practical stock at 1 mM in 70 % ethanol). Place cells on ice below a Details 500 W source of visible light at a distance of 300 mm for 5 min. This induces mainly 8-oxoguanine with very few SBs. If monolayer cells are used, sufﬁ cient dishes should be prepared for all time points. Include untreated cells as controls. 2. Rinse attached monolayer cells and add fresh medium. In the case of cells treated in suspension, centrifuge (700 × g, 5 min), and resuspend in culture medium. Place cells in a 37 °C incubator

and incubate for several hours, taking samples at t 0 and at 30 min intervals. Spin cells (after trypsinisation if necessary), wash in PBS, spin again, suspend pellet in agarose, and prepare gels. (a) Alternatively, embed cells in agarose after treatment with Ro 19-8022 + light (preparing one slide for each time point), place slides in medium in a Coplin jar in a 37 °C incubator and remove slides at intervals during the repair incubation period. 3. Lyse cells, wash, and incubate with FPG (or OGG) as described [17 ]. 4. Perform alkaline incubation and electrophoresis as usual.

2.3 Repair 1. Place dish (without lid) containing PBS-rinsed cell monolayer of UV-Induced or cells suspended in PBS below a germicidal (short wave- Pyrimidine Dimers length) UV lamp (e.g., General Electric BIAX) mounted in a (NER): Practical simple cabinet, with a shutter to allow timed exposure. The Details cabinet should be light-tight to protect the operator from UV. The dose rate (measured with a UV radiometer) can be adjusted by adjusting the shelf position, or by placing metal gauze or plastic sheets between the lamp and the cells. A suitable dose for studying cellular repair is 0.1 J m–2 . If monolayer cells are used, sufﬁ cient dishes should be prepared for all time points. Include untreated cells as controls. 2. Spin cells (700 × g , 5 min), and resuspend in culture medium (or simply rinse attached monolayer cells and add fresh medium). Place cells in a 37 °C incubator and incubate for

several hours, taking samples at t0 and at 30 min intervals. Spin Cellular DNA Repair Assays 371

cells (after trypsinisation if necessary), wash in PBS, spin again, suspend pellet in agarose, and prepare gels. (a) Alternatively, embed cells in agarose after treatment with UV (preparing one slide for each time point), place slides in medium in a Coplin jar in a 37 °C incubator and remove slides at intervals during the repair incubation period. 3. Lyse cells, wash, and incubate with T4endoV (adapt method described in [13 ]). 4. Perform alkaline incubation and electrophoresis as usual.

3 Measuring DNA Repair by Blocking Incision

Over 30 years ago, it was established that repair DNA synthesis occurring after UV irradiation of mammalian cells could be inhibited by hydroxyurea, and/or 1-β-D -arabinofuranosylcytosine (araC); as a result, incomplete repair sites accumulated as DNA breaks (measured at that time using techniques such as alkaline unwinding) [18 , 19 ]. Note : Hydroxyurea acts by blocking DNA precursor production , while araC inhibits polymerisation — as does aphidicolin , more recently available . This allows the estimation of incision rates; since incision is the rate-limiting step in NER, this approach can give information about repair capabilities in different species, or in mutant cell lines with UV sensitivity [20 ], or in connection with human disease [21 ] (Fig. 2 ).

Fig. 2 Kinetics of repair of UV(C) induced DNA lesions, in HeLa cells and in ﬁ broblasts from patient with xeroderma pigmentosum complementation group A. Lesions were measured using T4endoV with the alkaline unwinding assay. Drawn from data in [22 ]; bars represent standard errors of means from replicate experiments 372 Andrew R. Collins and Amaya Azqueta

Fig. 3 Time course of repair of UV(C)-induced lesions, measured in HeLa cells by the accumulation of DNA breaks (repair intermediates) during a 30 min pulse of hydroxyurea, cytosine arabinoside, and aphidicolin. Drawn from data in [22 ]; bars represent the range of duplicates

3.1 Practical Details 1. Place dish (without lid) containing PBS-rinsed cell monolayer or cells suspended in PBS below a UV lamp ( see step (1) in Sect. 2.3 ). Irradiate with 1 J m–2 (suggested dose). If monolayer cells are used, sufﬁ cient dishes should be prepared for all time points. Include untreated cells as controls. 2. Spin cells (700 × g , 5 min), and resuspend in culture medium (or simply rinse attached monolayer cells and add fresh medium). Add hydroxyurea, araC, and aphidicolin to ﬁ nal concentrations of 2 mM, 0.1 mM, and 15 μM, respectively. Note : The addition of both araC and aphidicolin should maximize the inhibition , but may not be necessary . 3. Place cells in a 37 °C incubator and take samples at 0, 5, 10, 20, and 30 min. Spin cells (after trypsinisation if necessary), wash in PBS, spin again, suspend pellet in agarose, and prepare gels. 4. Perform lysis, alkaline incubation, and electrophoresis as usual. 5. Plotting the time course of break accumulation over 30 min allows accurate estimation of the initial incision rate. Note : If inhibitors are added at intervals after irradiation , for a 30 min pulse , the kinetics of repair —i.e., the falling incision rate as the amount of damage declines — can be followed (Fig . 3 ). Cellular DNA Repair Assays 373

4 Discussion

Applications of the cellular approach to measuring DNA repair include:

● Characterization of the repair capacity of artifi cially produced or naturally occurring mutants in repair pathways [16 , 20 , 21 ]. ● Following the various repair pathways by use of enzymes with specifi cities appropriate for the induced lesion, such as AlkA for alkylated bases, FPG for oxidized bases, or T4endoV for pyrimidine dimers. Bulky adducts have so far been recalcitrant to this approach; one possibility would be to incubate with an extract prepared from eukaryotic cells, since this will contain the appropriate enzymes (as well as nonspecifi c nucleases that might increase the background of DNA breaks unless inhibited). ● Detecting inhibitory effects of chemical compounds on repair pathways (an aspect of genotoxicity), by preincubating the cells with the compound and then treating with an appropriate damaging agent (depending on the pathway targeted). However, if the compound is also capable of damaging DNA, disentangling the direct genotoxic effect from the repair inhibitory effect will be diffi cult. ● Investigating the possible stimulation or induction of repair, for example by micronutrients [23 ]: a similar scheme of preincubation of the cells with the putative enhancer is followed. Realistic concentrations (such as could be achieved in the blood by consuming reasonable amounts of the food containing the micronutrient) should be used. In many reports of DNA repair capacity, comparing different cell types, or cancer patients and controls, or different treatments (of cells or subjects), a full time course is not followed; levels of initial damage and residual damage at the end of the repair incubation are measured. There are problems with this approach, which are illustrated in Fig. 4 . This represents a comparison of the ability of lymphocytes from two individuals to repair oxidized DNA bases induced by photosensitizer plus light. They show signifi cantly different levels of induced damage (oxidized purines measured with

FPG) at t0 . Residual damage is measured during a 4 h incubation. At the end of this incubation, each subject has repaired three quar- ters of the initial damage. This could be taken as meaning that they are equally proﬁ cient at repair. But there are other possible interpretations. The subject shown by circle symbols has fewer lesions at the end of the 4 h incubation. However, the total number of lesions repaired in 4 h is greater for the “square” individual, who 374 Andrew R. Collins and Amaya Azqueta

Fig. 4 Schematic representation of repair kinetics for two individuals; one (squares ) incurred 50 % more damage than the other (circles ) from the same dose of damaging agent. Triangles represent the background level of damage

started with a higher level of damage. Ideally, individuals being compared should start with the same level of damage, but this cannot be assured, since individuals differ in antioxidant status. The best measure of repair would then be the initial rate—which requires sampling at closely spaced times just after inducing the damage (starting within minutes), in order to obtain an accurate

value from the regression slope. The t 1/2 for damage removal is an alternative valid measure, also requiring sampling at intervals during the repair period. Cellular repair assays, measuring removal of lesions or accumulation of blocked repair sites, involve complicated cell handling procedures and in some cases lengthy cell incubations. They are not convenient for samples from human trials when typically many samples need to be assayed—and there is a report that cryopreserved lymphocytes do not repair DNA damage as effectively as fresh cells [24 ]. For this application, we recommend an alternative, in vitro assay, which is described in full in Chap. 22 of this book. Different approaches to measurement of DNA repair would beneﬁ t from a validation exercise, involving analysis of identical samples of cells with different assays. There are many unanswered questions concerning DNA repair ability in humans. (This discussion is relevant to the use of both kinds of repair assay, i.e. cellular and in vitro.)

● Does poor DNA repair result in cancer? In extreme cases the answer is certainly yes. Xeroderma pigmentosum is the classic case of a mutated DNA repair gene leading to a deﬁ ciency in NER and, as a consequence, an enormously elevated risk of Cellular DNA Repair Assays 375

skin cancer caused by exposure to UV in sunlight [ 25 ]. DNA repair is frequently reported to be depressed in cancer patients compared with controls. The problem with all simple case– control studies is that we do not know whether the difference observed contributes to the etiology of the disease, or whether it is an effect. There are indications that repair pathways become deregulated in colorectal cancer, via an elevation of oxidative stress [26 ]. Prospective studies are needed, in which repair is estimated in a population that is then followed to see whether subjects with low repair capacity have an increased risk of cancer. Such studies do not exist, and so meanwhile we cannot deduce from case–control studies whether poor intrinsic repair capacity is a risk factor for cancer. ● Is a high measured repair capacity necessarily benefi cial? It is generally assumed so; however, there is patchy evidence from environmental and occupational mutagenesis studies [ 27 ] that repair may be induced by exposure to DNA-damaging agents— in which case a high measured repair capacity might simply refl ect the adverse exposure rather than an intrinsic survival advantage. ● What accounts for the wide variation observed in individual repair activities (strand break repair, BER and NER) measured in human cohort studies [ 27 ]? Effects of SNPs in repair genes on phenotypic activity of the corresponding enzymes are very limited, and cannot account for the variation. Environmental factors are presumably responsible for at least some of the variations, and in addition to the likely effects of exposure to DNA-damaging agents, it is clear that dietary factors can also infl uence repair [23 ]. ● Does declining repair capacity contribute to aging, as is often assumed? The evidence here is mixed; some studies reported a negative association between repair capacity and age, some claimed a positive association, and others found no effect of age [27 ]. Various assays, for different repair pathways, have been used, which makes comparisons and conclusions diffi cult. Furthermore, at least in human studies, repair is almost always measured in white blood cells (usually PBMN cells), and these might not be the most appropriate for looking at age-related effects. ● Do studies of expression of DNA repair genes provide useful information? To date, there seems to be very little evidence of correlation between phenotypic DNA repair activity (as measured with the comet assay or other techniques) and the level of expression of genes in the corresponding repair pathway. 376 Andrew R. Collins and Amaya Azqueta

References

1. Brand MD, Affourtit C, Esteves TC et al 15. Collins AR, Dusinska M, Horska A (2001) (2004) Mitochondrial superoxide: production, Detection of alkylation damage in human biological effects, and activation of uncoupling lymphocyte DNA with the comet assay. Acta proteins. Free Radic Biol Med 37:755–767 Biochim Pol 48:611–614 2. Friedberg EC, Walker GC, Siede W et al (2006) 16. Collins AR, Mitchell DL, Zunino A et al DNA repair and mutagenesis. ASM Press, (1997) UV-sensitive rodent mutant cell lines of Washington, DC complementation groups 6 and 8 differ pheno- 3. Iyanagi T (2007) Molecular mechanism of typically from their human counterparts. phase I and phase II drug-metabolizing Environ Mol Mutagen 29:152–160 enzymes: implications for detoxifi cation. In: 17. Collins AR, Azqueta A (2012) Single-cell gel Kwang WJ (ed) International review of cytol- electrophoresis combined with lesion-specifi c ogy, vol 260. Academic, Waltham, pp 35–112 enzymes to measure oxidative damage to 4. Grollman AP, Moriya M (1993) Mutagenesis DNA. Methods Cell Biol 112:69–92 by 8-oxoguanine: an enemy within. Trends 18. Collins AR (1977) DNA damage in ultraviolet- Genet 9:246–249 irradiated HeLa and CHO-K1 cells examined 5. Collins AR, Horvathova E (2001) Oxidative by alkaline lysis and hydroxyapatite chromatog- DNA damage, antioxidants and DNA repair; raphy. Biochim Biophys Acta 478:461–473 applications of the comet assay. Biochem Soc 19. Hiss ES, Preston RJ (1997) The effect of cyto- Trans 29:337–341 sine arabinoside on the frequency of single- 6. Goodarzi AA, Jeggo PA (2013) The repair and strand breaks in DNA of mammalian cells signaling responses to DNA double-strand following irradiation or chemical treatment. breaks. In: Theodore Friedmann JCD (ed) Biochim Biophys Acta 478:1–8 Advances in genetics, vol 82. Academic, 20. Busch DB, Zdzienicka MZ, Natarajan AT et al New York, pp 1–45 (1996) A CHO mutant, UV40, that is sensitive 7. Fortini P, Dogliotti E (2007) Base damage and to diverse mutagens and represents a new com- single-strand break repair: mechanisms and plementation group of mitomycin C sensitivity. functional signifi cance of short- and long-patch Mutat Res 363:209–221 repair subpathways. DNA Repair 6:398–409 21. Squires S, Johnson RT, Collins AR (1982) 8. Muniandy R, Liu J, Majumdar A et al (2010) Initial rates of DNA incision in UV-irradiated DNA interstrand crosslink repair in mamma- human cells; differences between normal, xero- lian cells: step by step. Crit Rev Biochem Mol derma pigmentosum and tumour cells. Mutat Biol 45:23–49 Res 95:389–404 9. Reardon JT, Cheng Y, Sancar A (2006) Repair 22. Klaude M, Gedik CM, Collins AR (1995) DNA of DNA–protein cross-links in mammalian damage and repair after low doses of UV-C cells. Cell Cycle 5:1366–1370 radiation; comparable rates of repair in rodent 10. Pegg AE (2000) Repair of O6-alkylguanine by and human cells. Int J Radiat Biol 67:501–508 alkyltransferases. Mutat Res 462:83–100 23. Collins AR, Azqueta A, Langie SAS (2012) 11. Ford JM (2005) Regulation of DNA damage Effects of micronutrients on DNA repair. Eur J recognition and nucleotide excision repair: Nutr 51:261–279 another role for p53. Mutat Res 577:195–202 24. Duthie SJ, Pirie L, Jenkinson AM et al (2002) 12. Collins AR, Ma A, Duthie SJ (1995) The Cryopreserved versus freshly isolated lympho- kinetics of repair of oxidative DNA damage cytes in human biomonitoring: endogenous (strand breaks and oxidised pyrimidines) in and induced DNA damage, antioxidant status human cells. Mutat Res 336:69–77 and repair capability. Mutagenesis 17:211–214 13. Collins AR (2014) Measuring oxidative dam- 25. DiGiovanna JJ, Kraemer KH (2012) Shining a age to DNA and its repair with the comet assay. light on xeroderma pigmentosum. J Invest Biochim Biophys Acta. 1840(2):794–800. Dermatol 132:785–796 http://dx.doi.org/10.1016/j.bbagen. 2013. 26. Tudek B, Speina E (2012) Oxidatively damaged 04.022 DNA and its repair in colon carcinogenesis. 14. ESCODD, Gedik CM, Collins A (2005) Mutat Res 736:82–92 Establishing the background level of base 27. Collins AR, Azqueta A (2012) DNA repair as a oxidation in human lymphocyte DNA: results biomarker in human biomonitoring studies; of an interlaboratory validation study. FASEB J further applications of the comet assay. Mutat 19:82–84 Res 736:122–129 Chapter 22

A Standardized Protocol for the In Vitro Comet-Based DNA Repair Assay

Jana Slyskova, Sabine A.S. Langie, Isabel Gaivão, Andrew R. Collins, and Amaya Azqueta

Abstract

DNA repair is regarded as an important biomarker to be measured alongside DNA damage when considering the risk of cancer from environmental or genetic causes. Efficient repair deals with DNA lesions before they can disrupt replication and create mutations. Repair capacity can be readily assessed using an in vitro comet-based DNA repair assay, which is particularly useful in human biomonitoring studies where many samples are collected over an extended period, stored frozen, and analyzed at a later date. In this assay, a protein lysate is extracted from studied cells or tissues and is incubated with damage-containing substrate DNA. Repair proteins in extract are able to recognize and incise DNA lesions and cumulate DNA breaks, which are quantified with the comet assay. Here we provide detailed protocols for the in vitro estimation of base excision repair (on a substrate containing 8-oxoguanine induced by visible light in the presence of a photosensitizer) and nucleotide excision repair (with UV-induced pyrimidine dimers and 6-4 photoproducts as substrate). We describe the preparation of extracts from different kinds of source material (cultured cells, peripheral blood mononuclear cells, animal tissues, human biopsies) and emphasize the need for careful control of the extract concentration. Furthermore, we discuss not only conventional comet assay format (2 gels on microscope slide), but also a medium-throughput version (12 minigels in microscope slide), which is recommended for reduction of experimental variability.

Key words DNA repair, Base excision repair, Nucleotide excision repair, Comet assay, Biomonitoring

1 Introduction

It is becoming ever more apparent that the complexity of DNA repair processes and their contribution to individual susceptibility to malignant diseases cannot be easily accounted for by single- nucleotide polymorphisms or other genome-level differences, or by gene expression analysis. Individual susceptibility is assumed to be formed by multiple additive effects of many medium- and low- penetrance gene variants, combined with the influence of environmental components, which are very variable at the individual level. A common quest for phenotyping approaches, able to directly quantify the activity of DNA repair enzymes or functioning of

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_22, © Springer Science+Business Media New York 2014 377 378 Jana Slyskova et al.

repair pathways, is gaining momentum. Alongside other assays (e.g. plasmid- or oligonucleotide-based assays), the in vitro comet- based DNA repair assay is clearly becoming a practical tool with a high potential for wide applicability. The method has been adapted for different kinds of biological specimen, including blood, cultured cells, and solid tissues. The method can be performed on fresh as well as deep-frozen samples. It is relatively high throughput (as with the 12-minigel format system optimized by us) and not excessively laborious. The in vitro comet-based repair assay measures the enzymatic activity of a cell extract, incubated with nucleoids (i.e. protein- depleted nuclei of supercoiled DNA that remain after cell lysis) containing artificially induced DNA lesions, embedded in agarose on a microscopic slide. The assay is able to measure activity of two particular repair pathways—base and nucleotide excision repair (BER and NER), taking advantage of their common feature of generating intermediate DNA strand breaks at the sites of eliminated DNA lesions. Quantification of these breaks reflects the activity of the repair process. Although the commonly accepted terminology for what is measured by this assay is “DNA repair capacity/activity,” it actually represents only the initial incision step of the whole excision repair mechanism. Incision is generally regarded as being the rate-limiting step in the intact cell. Only if deoxyribonucleotides and ATP are provided will the extract be capable of carrying out the synthetic stages of repair culminating in ligation, as was demonstrated by Collins et al. [1].

1.1 Historical The fact that repair proteins in a cellular extract can penetrate the Background agarose, recognize DNA lesions, and disrupt phosphodiester binding in agarose-embedded nucleoids was first exploited in 2001 by Collins and colleagues [2]. This in vitro DNA repair assay was designed to measure BER in extracts from human lymphocytes, using a substrate containing 8-oxoguanine induced by treating cells with a photosensitizer (Ro 19-8022) plus visible light [2]. From then on, the in vitro approach became increasingly popular due to its several positive aspects. Above all, it opens up the possibility to study DNA repair activity in diverse biological material, unlike “cellular repair” assays in which the removal of damage over time is monitored in cells in culture. The cell extract to be used in the in vitro assay can be prepared from virtually any tissue. Furthermore, this method is far less laborious and time-consuming compared to time-course experiments and can therefore be performed on numerous samples on a large scale. The BER-specific in vitro assay was modified in 2005 for evaluation of NER activity in human lymphocytes, using benzo(a) pyrene-diolepoxide (BPDE)-treated cells to provide the substrate nucleoids [3]. In 2009, the alternative UV(C)-treated substrate was introduced [4]. Just recently, methods for assessing BER and In Vitro DNA Repair Assay 379

NER activities in solid tissues were developed and optimized. While Langie et al. modified both BER and NER assays to measure DNA repair activities from solid tissues of animal origin [5, 6], human samples were assayed for both repair pathways by Slyskova et al. [7]. In this chapter, we provide a comprehensive overview with detailed protocols for measuring both BER and NER activities in extracts from different kinds of biological material, including isolated cells and solid tissues. We describe the NER assay based on a UV(C)-treated substrate, but do not describe the BPDE-NER assay, since this assay requires further standardization.

1.2 Principle A protein lysate is extracted from cells or tissues and incubated with of the Assay damage-containing substrate DNA. The DNA substrates consist of gel-embedded nucleoids from cells that were pretreated with the photosensitizer Ro 19-8022 plus light for the measurement of BER or UV(C) for the measurement of NER. Incubation of these substrate nucleoids with cell or tissue extracts allows the initial steps of BER or NER to occur. This will result in single-strand breaks that can then be determined by single cell electrophoresis (also known as comet assay). Thus, the increase in %DNA in the tail is proportional to the DNA repair incision activity of the extracts. A schematic overview of the assay and the principle is shown in Fig. 1.

1.3 Workflow To obtain reliable results by the assay, emphasis should be given to of the Experiment optimizing and standardizing experimental procedures. Below we mention some critical points to consider when setting up the in vitro repair assay in your own laboratory. (1) An optimal design includes preparing more than enough frozen aliquots from a single batch of substrate cells (treated, or not, with DNA-damaging agent) to provide positive and negative standards for the whole experimental run, especially critical when there are many samples from a biomonitoring study, for example. (2) The level of induced specific lesions in contrast to nonspecific damage induced during the substrate preparation should be determined in preliminary experiments using specific enzymes; ideally, the substrate should contain enough lesions to saturate the in vitro DNA repair assay while still having few background strand breaks. (3) Extracts can be prepared just before an experiment, or they can be stored at −80 °C until use. It is recommended to measure the protein concentration of fresh extract and then store it undiluted in aliquots (each aliquot enough for one experiment). (4) With samples of the biological material under test (e.g. tissues, lymphocytes, or cultured cells), preliminary experiments should be set up to determine the optimal concentration of extract, i.e. the extract concentration giving maximal discrimination between lesion-specific incision and nonspecific nuclease activity in the extract. (5) Another parameter that needs to be tested is the time of extract-substrate incubation. 380 Jana Slyskova et al.

a SUBSTRATE CELLS EXTRACT MATERIAL

Ro 19-8022 + light or UV or or

Cell culture PBMC Tissue or

Protein/enzyme Cell culture PBMC extraction

Substrate cells embedded in gel Extract

Immersion of the REACTION cells in lysis solution

ALKALINE TREATMENT ELECTROPHORESIS NEUTRALIZTION WASHING STAINING Substrate: nucleoids COMET ANALYSIS containing specific lesions Cell lysis b Recognition

T=T 3’ G A CAAT 5’ DNAincision

T=T 3’ G A CAAT 5’ Activity of the enzymes

containing in the EXTRACT Alkaline treatment: denaturation

P T=T OH OH G A P 3’ CAAT 5’

Electrophoresis

Fig. 1 Overview of the in vitro comet-based repair assay (adapted from [5, 11]). (a) Workflow: substrate cells are exposed to the photosensitizer Ro 19-8022 plus visible light, or to UV(C), to induce specific lesions. After lysis, gel-embedded nucleoids are incubated with cell/tissue extracts. Subsequent standard single cell gel electrophoresis reveals the incisions (detected as single-strand breaks) introduced by the DNA repair enzymes. (b) Principle of the comet-based repair assay (NER taken as an example for this scheme). Substrate DNA contains UV-induced thymidine dimers (T = T). DNA repair enzymes present in the cell extract will recognize and incise the lesions. Subsequent standard alkaline treatment reveals the single-strand breaks introduced by the NER enzymes In Vitro DNA Repair Assay 381

The optimal time should be on the linear part of the incubation curve, avoiding saturation. (6) The number of samples processed per experiment can be varied, within the limit imposed by the number of slides that can be placed together in the electrophoresis unit. After setting up the reaction conditions, the measurement of samples can be started. Briefly, an aliquot of substrate cells—both treated and untreated—is defrosted, and cells are embedded in agarose gel on slides. During lysis of cells, extracts can be diluted (with addition of ATP, aphidicolin, etc. if required), and positive-control enzyme(s) and negative-control buffer(s) are prepared. After lysis, substrate nucleoids are washed with buffer, and incubated for the required time with extracts/control solutions followed by alkaline incubation and electrophoresis. Slides can be stained and scored on the day of the experiment or stored unstained. It is, however, recommended that the whole cohort of samples should be scored by one person, to minimize inter-operator variation. The protocol below gives a detailed description of comet- based in vitro repair assays and their use in DNA repair incision activity quantification. See also supplementary material in Azqueta et al. [8] for more details and additional notes.

2 Materials

2.1 Solutions 1. Standard agarose (for pre-coating slides): 1 % in distilled water. 2. Low-melting-point agarose (for embedding cells): 0.8 % in PBS. 3. Cell freezing medium (for freezing cells): For example, MEM or RPMI, 10 % serum, 10 % DMSO. Store frozen. 4. Buffer A (extraction buffer): 45 mM HEPES, 0.4 M KCl, 1 mM EDTA, 0.1 mM dithiothreitol, 10 % (v/v) glycerol, adjusted to pH 7.8 with KOH. Prepare 100 mL. Store frozen. 5. Buffer A/1 % Triton X-100: Prepare 1 % Triton X-100 in buffer A. Store frozen in 1 mL aliquots (for use in one experiment). 6. Buffer A/0.25 % Triton X-100 (for background control incubation): Prepare 0.25 % Triton X-100 in extraction buffer A. Store frozen in 0.5 mL aliquots (for use in one experiment). 7. Lysis solution: 2.5 M NaCl, 0.1 M EDTA, 10.0 mM Tris, adjusted to pH 10 with 10 M NaOH. Prepare 1 L. Store at 4 °C. Before use, add 1 mL of Triton X-100 per 100 mL. 8. Buffer B (washing buffer after lysis and reaction buffer for BER): 40 mM HEPES, 0.5 mM EDTA, 0.2 mg/mL BSA, 0.1 M KCl, adjusted to pH 8 with KOH. Prepare 500 mL of 10x concentrated stock and freeze in 50 mL tubes (to use for washing slides after lysis) and in 1 mL aliquots (to use as reaction buffer). Dilute 10x in distilled water on the day of use. 382 Jana Slyskova et al.

9. Buffer N (washing buffer after lysis and reaction buffer for NER): 45 mM HEPES, 0.25 mM EDTA, 0.3 mg/mL BSA, 2 % (v/v) glycerol, adjusted to pH 7.8 with KOH. Prepare 500 mL of 10x concentrated stock and freeze in 50 mL tubes (to use for washing slides after lysis) and in 1 mL aliquots (to use as reaction buffer). Dilute 10x in distilled water on the day of use. 10. Electrophoresis solution: 0.3 M NaOH, 1 mM EDTA. Store at 4 °C (for up to 1 week). 11. Neutralizing solution: 1xPBS. Store at 4 °C.

2.2 Reagents 1. Photosensitizer Ro 19-8022 (for preparing BER substrate): and Enzymes Store frozen at 1 mM in 70 % ethanol. Avoid excessive light and wrap tubes in aluminum foil. 2. Formamidopyrimidine DNA glycosylase (FPG): Positive control for BER. 3. T4 endonuclease V (T4 endoV): Positive control for NER. 4. ATP (for extract activation in NER assay): Store frozen at 25 mM in distilled water. Working concentration is 2.5 mM.

2.3 Equipment 1. GelBond films can be purchased for mounting of the gels instead of using microscope slides. The advantage is that gels are guaranteed to stick to the GelBond film. However, a low- cost alternative is to pre-coat slides with agarose (see Sect. 2.4). When left to dry overnight, the risk of losing gels by using pre- coated slides is negligible. 2. 500 W tungsten halogen lamp (for activation of the photosensitizer Ro 19-8022). 3. Germicidal UV lamp (for induction of UV-induced damage in substrate DNA). 4. Microtube pestles (for homogenization of tissue). 5. Staining (Coplin) jars (for cell lysis and slide washing). 6. Humidity chamber (for extract-substrate incubation). 7. General comet assay equipment: See, e.g. [9] for a complete list.

2.4 Pre-coating Standard microscope slides with frosted end are used. Optional Microscope Slides step: degrease the slides with ethanol and leave them to dry.

1. Prepare 1 % agarose solution in H2O and keep it at 60 °C in water bath. 2. Dip the slides into the gel until the frosted part. 3. Wipe one side of the dipped slide clean and put the slide flat to dry. Remember to indicate with a mark on the frosted part, which side of the slide is the coated one. 4. Leave to dry overnight and then store in the original slide boxes. In Vitro DNA Repair Assay 383

3 Methods

3.1 Preparation Various types of cells can be used to prepare substrate DNA; of Substrate Cells cultured cells (e.g. HeLa, TK6) or human peripheral blood mononuclear cells (PBMC) are convenient. The dose of DNA-damaging agent should be tested in a titration experiment for the optimal dose introducing a high level of specific DNA damage but a relatively low level of background strand breaks. The ratio between specific and nonspecific DNA damage should be tested using FPG (for 19-8022-induced lesions) or T4 endoV (for UV-induced lesions). Substrate cells can be stored in frozen aliquots (−80 °C) for a long period, so that all extracts in an experiment or trial will be tested on substrate nucleoids from the same batch.

3.1.1 Induction 1. Prepare enough cells (cultured cells or PBMC) to provide both of DNA Damage exposed and non-exposed substrates (i.e. treatment control) to be used in all assays within a project. 2. For cells in suspension, wash cells in PBS, count cells, and centrifuge at 700x g, 5 min at 4 °C. In the case of using adherent cells, just wash them with PBS.

8-Oxoguanines 1. After centrifugation of cells, add cold PBS containing photo- to Study BER sensitizer Ro 19-8022 (1–2 μM) and suspend pellet. Avoid excessive light. In the case of using adherent cells, simply add cold PBS containing photosensitizer Ro 19-8022. 2. Place cells on ice, 33 cm from a 500 W tungsten halogen lamp, and irradiate for 5 min. 3. Prepare control cells in PBS, with no photosensitizer, and expose them to light for 5 min in the same way as you expose the cells to Ro 19-8022. 4. After exposure, collect and centrifuge cells at 700x g, 5 min at 4 °C, wash with PBS, and spin again. In the case of using adherent cells, wash them with PBS and trypsinize to collect them. 5. Suspend cells at ~1 × 106/mL in cold freezing medium, and prepare 0.3 mL aliquots in 1.5 mL microtubes. Each aliquot will have enough cells for 20 gels in the 2 gels/slide format. 6. Freeze slowly to −80 °C using a “Mr. Frosty”® (Nalgene) freezing container or in a thick-walled box of expanded polystyrene.

UV Photoproducts 1. After centrifugation, add cold PBS and place cells in a petri to Study NER dish. In case of using adherent cells, wash them with PBS. 2. Place the dish under a UV(C) source and irradiate with 1–2 Jm−2. Irradiate without the lid, as plastic reduces UV exposure. 384 Jana Slyskova et al.

3. Prepare control, non-exposed cells, from the same batch. 4. After exposure, collect and centrifuge cells at 700x g, 5 min at 4 °C, wash with PBS, and spin again. In the case of using adherent cells, wash them with PBS and trypsinize to collect them. 5. Suspend cells at ~1 × 106/mL in cold freezing medium and prepare 0.3 mL aliquots in 1.5 mL microtubes. Each aliquot will have enough cells for 20 gels in the 2 gels/slide format. 6. Freeze slowly to −80 °C using a “Mr. Frosty”® (Nalgene) freezing container or in a thick-walled box of expanded polystyrene.

3.1.2 Checking The ratio of specific damage (8-oxoguanines or UV photoproducts) of Substrate DNA to nonspecific damage (strand breaks and alkali-labile sites) should be controlled in preliminary experiments by incubating: 1. Nucleoids from Ro 19-8022-exposed cells with FPG (recognizing 8-oxoguanine) for the BER assay and nucleoids from UV-irradiated cells with T4 endoV (detecting UV-induced pyrimidine dimers) for the NER assay. Extracts must be provided with DNA substrates containing an excess of damage to work on (ideally, the lesions in the substrate should saturate the assay). 2. Nucleoids from non-exposed cells with FPG or T4 endoV to show the background level of endogenous lesions (normally insignificant in the case of T4 endoV, but PBMC and cell lines contain variable levels of oxidized bases recognizable by FPG). If the background lesions are excessive, another batch of cells should be prepared.

3.2 Preparation Extract can be prepared from viable cells stored in freezing medium of Extract or from cells stored as dry cell pellets (both can be prepared from either PBMCs or cultured cells) or from deep-frozen solid tissues. 3.2.1 Source Material Peripheral Blood Approximately 5–10 mL of blood is needed for preparing extract. Mononuclear Cells Isolate PBMC from whole blood according to standard procedure: or Cultured Cells 1. Wash cells with PBS and count a sample (adherent cultured cells need to be trypsinized). Centrifuge at 700x g for 10 min at 4 °C. Either 2. Suspend at 5 × 106/mL in cold freezing medium and prepare 1 mL aliquots in 1.5 mL microtubes. 3. Freeze slowly to −80 °C using a “Mr. Frosty”® (Nalgene) freezing container or in a thick-walled box of expanded polystyrene. Or 1. Suspend cells at 5 × 106 cells/mL in cold 3x diluted buffer A and prepare 1 mL aliquots in microtubes. In Vitro DNA Repair Assay 385

2. Centrifuge at ~2,000x g for 5 min at 4 °C. 3. Carefully remove as much supernatant as possible, avoiding disturbance to the pellet. 4. Drop the tubes into liquid nitrogen. They can then be stored at −80 °C for use at a later date or processed immediately.

Solid Tissues Solid tissues sampled into the cryotubes should be snap frozen in liquid nitrogen after sampling and then stored at −80 °C: 1. Grind the frozen tissue under liquid nitrogen using a pestle and mortar prechilled in liquid nitrogen, weigh, and divide into aliquots of 30–50 mg. 2. Ground tissue aliquots can be stored at −80 °C for use at a later date or processed immediately.

3.2.2 Extract Preparation 1. Thaw the frozen cells, and centrifuge at 700x g for 5 min at 4 °C.

From Frozen Cells 2. Suspend pellet in cold PBS, count cells, and centrifuge again. 3. Suspend cells at 5 × 106 cells/mL in cold 3x diluted buffer A in 1 mL aliquots in microtubes. 4. Centrifuge at ~2,000x g for 5 min at 4 °C. 5. Discard supernatant. The pellet should be almost dry. 6. Continue as described for frozen cell pellet.

From Frozen 1. Add 50 μL of buffer A to each pellet of 5 × 106 cells. Cell Pellets 2. Snap freeze by dropping into liquid nitrogen and immediately thaw again. 3. To each 50 μL aliquot, add 15 μL of buffer A/1 % Triton X-100. 4. Vortex for 5 s and leave for 10 min on ice. 5. Centrifuge at ~15,000x g for 5 min at 4 °C to remove cell debris. 6. Collect the supernatant.

From Frozen Ground 1. Thaw tissue aliquots. Tissues 2. Add 100 μL of buffer A per 30 mg. 3. Vortex, snap freeze by dropping into liquid nitrogen, and immediately thaw again. 4. Add 30 μL of 1 % Triton X-100 in buffer A to each 100 μL aliquot. 5. Vortex vigorously and incubate for 10 min on ice. 6. If necessary, homogenize larger particles of tissue with a microtube pestle. 7. Centrifuge at ~15,000x g for 5 min at 4 °C to remove cell debris. 386 Jana Slyskova et al.

8. Collect the supernatant. Note: Keep extracts on ice all the time.

3.2.3 Concentration Protein concentration should ideally be measured prior to the reac- of Proteins in the Extract tion so that all the extracts can be used at the same concentration in the experiment. Retrospective normalization of the activity according to protein concentration (when extracts are used at different concentrations in experiments and results are adjusted for protein concentration afterward) is not recommended because protein concentration and activity as measured in the assay are not proportionally related. When storing the extracts for later use, it is advisable to measure protein concentration after the extract preparation but to store it undiluted. Dilution with buffer should be performed just before the reaction.

3.3 Embedding Cells 1. Thaw frozen substrate cells by adding 1 mL of ice-cold PBS. in Agarose and Lysis 2. Centrifuge at 700x g for 5 min at 4 °C. for Substrate Nucleoid 3. Wash in cold PBS and spin again. Preparation 4. Remove the supernatant, disperse the pelleted cells by tapping vigorously, and add 1.5 mL of 0.8 % low-melting-point agarose at 37 °C (to reach the concentration of 2 × 105 cells/mL). 5. Add two 70 μL drops to each pre-coated microscope slide (the final number of cells per gel is ~14,000). 6. Cover gels with coverslips and keep for a few minutes at 4 °C. 7. Remove coverslips and place slides in lysis solution at 4 °C (with 1 % Triton X-100 added just before use). They can be left in lysis solution for between 1 and 48 h. 8. Wash slides in three changes of buffer B or N (depending on whether BER or NER is being assayed), 5 min each, at 4 °C.

3.4 Reaction The optimal concentration of the extract as well as time of incubation of the extract with substrate DNA should be tested in titration experiments. Use randomly selected extracts to establish the optimal conditions for the reaction. Several experimental controls should be included in the assay for the correct interpretation of the results.

3.4.1 Extract-Substrate Ideally, protein concentrations should be identical, so if possible Incubation measure the protein concentration of a sample of the extract and dilute all extracts to a standard concentration with buffer A/1 % Triton X-100. Always keep protein extracts on ice: 1. Keep slides on ice to prevent premature enzyme activity when the extract is added. 2. Place humidity chamber in a 37 °C incubator. In Vitro DNA Repair Assay 387

Fig. 2 Overview of experimental design and different types of experimental controls; (a) non-exposed substrate incubated with buffer represents any background DNA damages that may result from rough handling of the substrate cells b (background control); (b) exposed substrate incubated with buffer represents DNA breaks resulting from treatment with DNA-damaging agent (treatment control); (c) non-exposed substrate incubated with extract reflects nonspecific endonuclease activity of the extract (specificity control); (d) exposed substrate incubated with extract represents breaks introduced by the incision of total DNA damage (both endogenous and experimentally induced lesions); (e) non-exposed substrate incubated with specific enzyme (FPG or T4 endoV) represents the (normally low) level of endogenous damages recognized by the enzyme; (f) exposed substrate incubated with specific enzyme (FPG or T4 endoV) indicates the total of experimentally induced and endogenous enzyme-specific lesions (positive control). Data are shown as mean TI of two independent incubations within one experiment. Bars indicate SD. Reprinted from [11] with permission from Elsevier

BER 1. Add to the extract 4 volumes of reaction buffer B. 2. Prepare a buffer control: buffer A/0.25 % of Triton X-100, mixed with buffer B in a ratio 1:4. Continue as for NER 3. step.

NER 1. Add to the extract 4 volumes of buffer N (buffer N contains ATP in a ratio 7:1, so the final working concentration of ATP in the extract is 2.5 mM). 2. Prepare a control buffer: buffer A/0.25 % of Triton X-100, mixed with buffer N + ATP in a ratio 1:4. 3. Add 50 μL of diluted extract or buffer control to each of two gels (containing nucleoids of either non-exposed or exposed cells; see Fig. 2 bars C and D). Each sample then has technical replicates. 4. Cover with coverslips and incubate at 37 °C in the humidity chamber. The incubation time is normally around 10–30 min. 388 Jana Slyskova et al.

3.4.2 Experimental Experimental controls should be used in each experiment in order Controls (i) to verify the reliability of the experiment, (ii) to calculate the net DNA repair activity, and (iii) to perform inter-experimental calibration:

●● Background control: Non-exposed substrate incubated with reaction buffer represents any background DNA damage (Fig. 2, bar A). ●● Treatment control: Exposed substrate incubated with reaction buffer reveals the presence of any nonspecific damage, i.e. strand breaks or abasic sites, resulting from the treatment with damaging agent (Fig. 2, bar B). ●● Specificity control: Non-exposed substrate incubated with extract will determine the nonspecific nuclease activity of the extract not related to specific BER or NER activity (Fig. 2, bar C). The calculation of final incision activity is explained in Sect. 3.6. Nonspecific activity of the extract can be suppressed by applying inhibitors of nucleases into the reaction. Aphidicolin at a concentration of 1.5 μM has been tested in the BER assay to increase the specificity of the reaction [5].

3.5 Alkaline The remaining steps are as in the standard comet assay protocol [10]: Treatment and Single 1. Incubation in cold electrophoresis solution for 40 min. Cell Gel Electrophoresis 2. Electrophoresis at ~1 V/cm for 20–30 min. (Voltage gradient should be measured across the platform carrying the slides.) 3. Neutralization by washing slides for 10 min in ice-cold PBS and 10 min in ice-cold water. 4. Drying of gels overnight. 5. The day before scoring comets, slides can be stained with ethidium bromide (0.01 μg/mL in water) or DAPI (1 μg/mL in water), adding 20 μL to each gel and covering with a coverslip. If using SYBR Gold® (Invitrogen), which gives intense fluorescence, it is recommended to immerse slides in a bath of the stain at a dilution of 1:10,000 in water for 20 min, followed by two 10 min washes with water. Other DNA fluorescence stains are also suitable. 6. Slides are left to dry, and for viewing, 20 μL of water is added to each gel and covered with a coverslip.

3.6 Scoring Comets Comets can be scored by visual scoring, but computer-assisted and Calculation image analysis using commercially available software gives the most of Repair Rates reproducible results. Tail intensity (TI) that represents % of DNA in the tail of the comet is the recommended and most precise indica- 3.6.1 Image Analysis tor of damage. Usually, 50 comets are scored per gel, i.e. 100 In Vitro DNA Repair Assay 389

comets per sample when working in duplicates. To obtain the net TI value, calculate first the median TI per each gel and then the mean TI of replicate gels.

3.6.2 Calculation The procedure for calculating the incision activity is demonstrated of Repair Rates here, by taking the measurement of BER on a substrate with 8-oxoguanine as an example. For net DNA incision activity calculation, four values of % tail DNA are needed: 1. Reaction buffer incubated with untreated substrate, i.e. background control (TI noRo/buffer) 2. Reaction buffer incubated with exposed substrate, i.e. treatment control (TI Ro/buffer) 3. Extract incubated with non-exposed substrate, i.e. specificity control (TI noRo/extract) 4. Extract incubated with exposed substrate (TI Ro/extract) To calculate the DNA repair-related incision activity, first subtract the background control value (TInoRo/buffer) from all data to obtain “net” breaks, which can be used in the following formula:

DNA incision activityn=−et TIRo//Extract netTInnoRoExtract − et TIRo/Buffer Results are normally reported as change in TI in a given time.

4 Notes

4.1 Optimization There are three different ways for standardizing the concentration of Extract Density of the extracts; they can be (i) prepared from equal numbers of cells, (ii) diluted to reach the same protein concentration in all samples, or (iii) diluted according to the DNA content. Clearly, the choice of procedure is dependent on the source material used for extract isolation. In case of cell suspensions, all three approaches can be applied to prepare standardized extracts. When working with frozen solid tissues, the density of extracts can only be standardized according to the protein or DNA concentration. It is recommended to run titration experiments to identify the preferred protein concentration (i.e. the concentration that gives the highest specificity). We have observed that the protein concentration to be used in the assay is tissue specific, and we advise to test the optimal concentration range for each particular tissue type [5, 6]. This fact, however, poses a difficulty in a comparison of DNA repair activity among different tissue types. Another possibility for optimizing an extract density is to use the amount of DNA (measured at an early stage, before spinning down the cell and nuclear debris). This, in case of solid tissues especially, might eliminate the inconsistencies in protein ratios caused by histological heterogeneity of tissues. However, this option remains to be tested. Alternatively, if the 390 Jana Slyskova et al.

extract density cannot be standardized prior to the experiment, results can be adjusted according to the cell, protein, or DNA concentration, respectively. This option, however, is not ideal and is not recommended, as explained in the next section.

4.2 How to Express The incubation of extract with substrate should be performed for Results? a standard length of time, so that results can be expressed as a rate of accumulation of DNA breaks which are optimally expressed as % of DNA in tail. Comet assay results can be further converted to an actual DNA break frequency, using a calibration curve based on irradiated cells, so that the results can be expressed in terms of breaks per 109 Da [9, 10]. Another way of expressing results is to calculate DNA strand breaks relative to protein concentration, e.g. breaks per mg/mL protein. It is, however, important to keep in mind that activity as assayed with this method is not linearly proportional to protein concentration but increases less than twofold for a doubling of concentration [5, 8].

4.3 Which Interindividual variations in NER incision activity measured by the Interindividual in vitro BPDE-NER assay using lymphocyte extracts were in the Variations in DNA range of ~8-fold [11]. Similar variations of ~10-fold were observed Incision Activity in human lymphocytes by Gaivao et al., conducting in vitro to Expect? UV-NER assay [4]. These are relatively wide variations in NER incision activity compared with interindividual variations for BER that were reported to be approximately ~4-fold [4]. Notably, some individuals even appear to have negligible DNA repair incision activity as measured by comet-based in vitro repair assays, which might lead to an accumulation of DNA lesions after exposure to genotoxic agents and a subsequent increased risk of mutations and cancer development. However, on the other hand, a lower DNA repair incision activity might also reflect the absence of lesions, because DNA damage can activate DNA repair proteins directly via feedback control mechanisms. It is therefore necessary to assess DNA repair activity in combination with exposure (e.g. by using biomarkers of exposure) in prospective cohort studies before we can make any reliable predictions on cancer susceptibility and risk. However, practical differences in the performance of the comet assay or quantitation by image analysis could also contribute to the variation. In the comparison of different laboratories performing in vitro repair assay by the European Comet Assay Validation Group (ECVAG), it was shown that the variation between laboratories is considerable [12].

4.4 An Update A medium-throughput format for the comet assay was introduced for Large-Scale in 2009, comprising 12 minigels (each of 5 μL of agarose) per slide Studies [13]. Each minigel can be incubated separately with different extract/enzyme/buffer when placed into the 12-gel comet assay unit produced by Severn Biotech (www.severnbiotech.com, Fig. 3). This unit allows the analysis of many more extracts in one In Vitro DNA Repair Assay 391

Fig. 3 12-gel comet assay unit (Severn Biotech) increases throughput and allows to incubate the gels separately; metal positioning guide (1), silicone gasket (2), and top plate (3) tightened by metal clamps (4)

experiment and uses fewer substrate cells and less extract volume per reaction. A guide to performing an experiment using this format is briefly described here.

4.4.1 Substrate Cells Minigels require approximately 10 times fewer substrate cells per reaction than gels of conventional format. Therefore, the substrate cells are stored in aliquots of 2.5 × 105 cells in 0.5 mL of freezing medium. This is more than enough for 20 slides, i.e. 240 minigels.

4.4.2 Setting Substrate cells are mixed with 0.8 % low-melting-point agarose to the Minigels a final concentration of 0.5 × 105/mL. Using a multi-dispensing pipettor, place 12 gels of 5 μL (~250 cells per gel) on each microscope slide at the positions defined by the metal positioning guide of the 12-gel comet assay unit (Fig. 3). Precool the metal positioning guide to 4 °C so the gels will set very quickly. Coverslips are not used, and so the gels are dome shaped.

4.4.3 Extract-Substrate Cover microscope slide with the silicone gasket, place the top plate Incubation on top, and tighten everything together with metal clamps (Fig. 3). Pipette 30 μL of extract (or buffer or specific enzyme) onto each gel in the wells formed by the gasket and incubate for the required time at 37 °C. At the end of incubation, put chambers on ice. Remove slides from chambers and immediately transfer to the electrophoresis tank (already containing cold alkaline solution). Speed is important, to stop the reactions and to avoid cross-contamination between different extracts when the slides are removed from the chamber.

4.4.4 Dehydration To avoid “disorientated” comets [14], gels need to be dehydrated of the Minigels after the neutralization. Dehydrate gels by immersing slides in 70 % ethanol for 15 min and in absolute ethanol for another 15 min. 392 Jana Slyskova et al.

4.5 In Vitro DNA The molecular epidemiological approach, measuring molecular or Repair Assay cellular indicators of disease risk or exposure to causative factors, in Molecular is a valuable tool in addition to conventional epidemiology. There Epidemiological is an increasing demand for phenotyping assays in the field of Studies human functional genetics. DNA repair activity is representative of this functional approach. It is a complex marker that defines the multifactorial process of DNA repair, involving various genetic and nongenetic factors that modulate the final phenotype. The in vitro DNA repair assay has been widely used in the field of molecular epidemiological studies. There are many studies focused on the relation between DNA repair activity and age, following the hypothesis that aging is driven by accumulation of errors in macro- molecules, including DNA, which might be accompanied by a reduced DNA repair activity. Nevertheless, an age-dependent decrease in DNA repair remains still a hypothesis to be confirmed, due to an inconsistency of observations, possibly caused by the lack of a standardized protocol. Some groups have reported a negative correlation of repair with age, some a positive association, and some did not observe any relationship, as reviewed in [15]. So far, the in vitro DNA repair assay is most often applied in trials investigating whether DNA repair activity is modulated by dietary supplementation with phytochemicals, antioxidant-rich food, or other dietary supplements. In this respect, supplementation by a mixture of antioxidant-rich plant-based food [16, 17] or kiwifruit only [18] showed enhancement of repair and/or reduction of the level of DNA damage. In some other cases, supplementation with selenium plus vitamins A, C, and E [19], fruit juice [11], or broccoli [20] did not produce any significant influence in the level of DNA repair activity. In a retrospective study based on data from food frequency questionnaires and plasma levels of antioxidants, DNA repair activity of NER pathway was increased with higher plasma levels of ascorbic acid and alpha- carotene [21]. Recently, Langie et al. observed maternal folate depletion during pregnancy and lactation to result in a significant induction of the BER-related incision activity in the brain of offspring at weaning, which in the long term leads to a significantly decreased BER-related incision activity in the brain of the adult offspring (6-month-old mice), though mothers were unaffected [22]. The review by Collins et al. gives a further overview of human and animal studies looking at the modulation of DNA repair by micronutrients [23]. Another branch of research is represented by studies that evaluate genotoxic effects of various chemicals, usually potential carcinogens that are used in industrial production. For environmentally induced diseases, molecular biomarkers play a key role in understanding the relationship between exposure to toxic chemicals and the development of chronic diseases and in identifying individuals at increased risk. They can be used to monitor levels of exposure to some disease-causing agents, or they may inform about In Vitro DNA Repair Assay 393

interindividual variation in response to these factors. In this respect, biomarkers of DNA damage have been used to assess genotoxic exposure to industrially produced chemicals in numerous epidemiological studies. DNA repair, as a marker of individual response to DNA damage, has been less used to date. However, there are few studies that applied this approach, in relation to exposure to asbestos [24], mineral fibers [25], and styrene [26–30] and in a rubber tire factory [31]. No consistent overall picture arises from these studies; some agents seem to enhance repair, other seems to reduce its level, and sometimes no effect of exposure is observed. The lack of a standardized protocol has in all probability contrib- uted to this situation. Finally, in vitro DNA repair techniques have been utilized in relation to cancer, although not very frequently. There are a few reports of modulation of DNA repair activity by anticancer drugs as measured by the in vitro repair assay [32]. A fundamental feature of cancer is genomic and chromosomal instability. Furthermore, most agents recognized to be potential carcinogens operate via generating DNA damage and causing mutations. It is also known that inherited or acquired defects in DNA damage-response mechanisms contribute significantly to the onset of cancer [33]. Studies analyzing NER activity in association with cancer by (not only) in vitro assays were reviewed by Slyskova et al. [34]. NER was shown to be suboptimal in PBMC of patients with cancers of various types. However, in most cases, it is impossible to say whether this is a cause or an effect of the disease. Evaluation of BER incision activity in human solid tissues—colon tumors and adjacent mucosa in particular—was reported by Herrera et al. [35] and by Slyskova et al. for not only BER but also NER activity in colorectal tissues [7]. In both reports, DNA repair activity of the tumor was seen to be the same or just moderately higher than that of adjacent healthy tissue.

4.6 Open Questions The interindividual variability of DNA repair seems to be substantial. Nevertheless, at present we still cannot derive any absolute reference values characterizing the physiological range of DNA repair activity in healthy individuals, to be used for further comparisons with other populations with different characteristics (affected by disease, occupationally or environmentally exposed to toxicants or simply of different ethnicity or age/sex composition). The reason is that the studies so far available have used slightly different protocols to measure DNA repair, with different DNA damage-inducing agent, concentrations, time intervals, etc. Recently, however, there has been an effort toward the optimization and interlaboratory validation of the methods to measure DNA repair. For instance, ECVAG (European Comet Assay Validation Group) and the ComNet group (Comet Assay Network group) have been established in part to develop commonly accepted protocols for human biomonitoring of DNA 394 Jana Slyskova et al.

damage and DNA repair with comet assay-based methods [36, 12]. The current major challenges are:

●● To validate all indeterminate steps in the assay (i.e. optimization of the assay to be used on BPDE-treated substrates; the use of aphidicolin not only in BER but also in the NER assay; the normalization of the extract according to DNA content) ●● To validate the in vitro comet-based assays by comparison with other in vitro repair assays ●● To organize a “ring study” to involve several laboratories in different countries in measuring DNA repair in standard cell extracts, distributed by the lead laboratory, using standard protocols (based on the protocols described in this paper)

Acknowledgments

Ro 19-8022 was kindly provided by Hoffman la Roche. SL was supported by a postdoctoral grant from the AXA Research Fund. AA thanks the Ministerio de Educación y Ciencia (“Juan de la Cierva” programme, 2009) of the Spanish Government for personal support. IG thanks the Portuguese Science and Technology Foundation (FCT) under the Project PEst-OE/AGR/UI0772/2014.

References

1. Collins AR, Fleming IM, Gedik CM (1994) In 7. Slyskova J, Korenkova V, Collins AR et al vitro repair of oxidative and ultraviolet-induced (2012) Functional, genetic, and epigenetic DNA damage in supercoiled nucleoid DNA by aspects of base and nucleotide excision repair in human cell extract. Biochim Biophys Acta colorectal carcinomas. Clin Cancer Res 1219(3):724–727 18(21):5878–5887 2. Collins AR, Dusinska M, Horvathova E et al 8. Azqueta A, Langie SA, Slyskova J et al (2013) (2001) Inter-individual differences in repair of Measurement of DNA base and nucleotide DNA base oxidation, measured in vitro with excision repair activities in mammalian cells and the comet assay. Mutagenesis 16(4):297–301 tissues using the comet assay: a methodological 3. Langie SA, Knaapen AM, Brauers KJ et al (2006) overview. DNA Repair 12(11):1007–1010 Development and validation of a modified comet 9. Collins AR, Oscoz AA, Brunborg G et al assay to phenotypically assess nucleotide excision (2008) The comet assay: topical issues. repair. Mutagenesis 21(2):153–158 Mutagenesis 23(3):143–151 4. Gaivao I, Piasek A, Brevik A et al (2009) Comet 10. Forchhammer L, Johansson C, Loft S et al assay-based methods for measuring DNA repair (2009) Variation in the measurement of DNA in vitro; estimates of inter- and intra-individual damage by comet assay measured by the variation. Cell Biol Toxicol 25(1):45–52 ECVAG inter-laboratory validation trial. 5. Langie SA, Cameron KM, Waldron KJ et al Mutagenesis 25(2):113–123 (2011) Measuring DNA repair incision activity 11. Langie SA, Wilms LC, Hamalainen S et al of mouse tissue extracts towards singlet oxygen- (2010) Modulation of nucleotide excision induced DNA damage: a comet-based in vitro repair in human lymphocytes by genetic and repair assay. Mutagenesis 26(3):461–471 dietary factors. Br J Nutr 103(4):490–501 6. Langie SA, Kowalczyk P, Tudek B et al (2010) 12. Godschalk RW, Ersson C, Riso P et al (2013) The effect of oxidative stress on nucleotide- DNA-repair measurements by use of the modi- excision repair in colon tissue of newborn pig- fied comet assay: an inter-laboratory compari- lets. Mutat Res 695(1–2):75–80 son within the European Comet Assay In Vitro DNA Repair Assay 395

Validation Group (ECVAG). Mutat Res 26. Vodicka P, Koskinen M, Stetina R et al (2003) 757(1):60–67 The role of various biomarkers in the evalua- 13. Shaposhnikov S, Azqueta A, Henriksson S et al tion of styrene genotoxicity. Cancer Detect (2010) Twelve-gel slide format optimised for Prev 27(4):275–284 comet assay and fluorescent in situ hybridisa- 27. Hanova M, Vodickova L, Vaclavikova R et al tion. Toxicol Lett 195(1):31–34 (2011) DNA damage, DNA repair rates and 14. Azqueta A, Gutzkov KB, Priestley CC et al mRNA expression levels of cell cycle genes (2013) A comparative performance test of (TP53, p21(CDKN1A), BCL2 and BAX) with standard, medium- and high-throughput respect to occupational exposure to styrene. comet assays. Toxicol In Vitro 27(2):768–773 Carcinogenesis 32(1):74–79 15. Collins AR, Azqueta A (2012) DNA repair as a 28. Hanova M, Stetina R, Vodickova L et al (2011) biomarker in human biomonitoring studies; Modulation of DNA repair capacity and mRNA further applications of the comet assay. Mutat expression levels of XRCC1, hOGG1 and XPC Res 736(1–2):122–129 genes in styrene-exposed workers. Toxicol 16. Guarnieri S, Loft S, Riso P et al (2008) DNA Appl Pharmacol 248(3):194–200 repair phenotype and dietary antioxidant sup- 29. Vodicka P, Koskinen M, Naccarati A et al plementation. Br J Nutr 99(5):1018–1024 (2006) Styrene metabolism, genotoxicity, and 17. Brevik A, Karlsen A, Azqueta A et al (2011) potential carcinogenicity. Drug Metab Rev Both base excision repair and nucleotide excision 38(4):805–853 repair in humans are influenced by nutritional 30. Slyskova J, Dusinska M, Kuricova M et al (2007) factors. Cell Biochem Funct 29(1):36–42 Relationship between the capacity to repair 18. Collins AR, Harrington V, Drew J et al (2003) 8-oxoguanine, biomarkers of genotoxicity and Nutritional modulation of DNA repair in a individual susceptibility in styrene-exposed human intervention study. Carcinogenesis workers. Mutat Res 634(1–2):101–111 24(3):511–515 31. Vodicka P, Kumar R, Stetina R et al (2004) 19. Caple F, Williams EA, Spiers A et al (2010) Markers of individual susceptibility and DNA Inter-individual variation in DNA damage and repair rate in workers exposed to xenobiotics in base excision repair in young, healthy non- a tire plant. Environ Mol Mutagen 44(4): smokers: effects of dietary supplementation 283–292 and genotype. Br J Nutr 103(11):1585–1593 32. Sliwinski T, Czechowska A, Szemraj J et al 20. Riso P, Martini D, Moller P et al (2010) DNA (2008) STI571 reduces NER activity in BCR/ damage and repair activity after broccoli intake ABL-expressing cells. Mutat Res 654(2): in young healthy smokers. Mutagenesis 162–167 25(6):595–602 33. Hoeijmakers JH (2001) Genome maintenance 21. Slyskova J, Lorenzo Y, Karlsen A et al (2014) mechanisms for preventing cancer. Nature Both genetic and dietary factors underlie dif- 411(6835):366–374 ferences in DNA damage levels and DNA 34. Slyskova J, Naccarati A, Pardini B et al (2012) repair capacity. DNA repair 16:66–73 Differences in nucleotide excision repair capac- 22. Langie SA, Achterfeldt S, Gorniak JP et al ity between newly diagnosed colorectal cancer (2013) Maternal folate depletion and high-fat patients and healthy controls. Mutagenesis feeding from weaning affects DNA methyla- 27(2):225–232 tion and DNA repair in brain of adult offspring. 35. Herrera M, Dominguez G, Garcia JM et al FASEB J 27(8):3323–3334 (2009) Differences in repair of DNA cross- 23. Collins AR, Azqueta A, Langie SA (2012) links between lymphocytes and epithelial Effects of micronutrients on DNA repair. Eur J tumor cells from colon cancer patients mea- Nutr 51(3):261–279 sured in vitro with the comet assay. Clin Cancer Res 15(17):5466–5472 24. Dusinska M, Collins A, Kazimirova A et al (2004) Genotoxic effects of asbestos in 36. Collins A, Anderson D, Coskun E et al (2012) humans. Mutat Res 553(1–2):91–102 Launch of the ComNet (comet network) project on the comet assay in human population studies 25. Dusinska M, Barancokova M, Kazimirova A during the International Comet Assay Workshop et al (2004) Does occupational exposure to meeting in Kusadasi, Turkey (September 13–16, mineral fibres cause DNA or chromosome 2011). Mutagenesis 27(4):2 damage? Mutat Res 553(1–2):103–110 Chapter 23

Use of the Comet Assay to Study DNA Repair in Drosophila melanogaster

Isabel Gaivão , Rubén Rodríguez , and L. María Sierra

Abstract

Drosophila melanogaster is a useful model for genetic studies, including DNA repair. In addition to all the advantages of this model organism, there are several strains available which are effi cient and defi cient in different DNA repair pathways, and the mechanisms of action of many compounds are very well known in vivo. The analysis of DNA repair capacity in Drosophila can be achieved by following two different approaches with comet assay: in vivo, DNA repair assay , by performing the assay with different repair condition strains, direct analysis of cells/individuals, and comparison between/among them through statistical regression analysis, and in vitro, DNA repair assay , by incubating a substrate nucleoid DNA from cells pretreated with a specifi c damaging agent with a cell-free protein extract obtained from mashed adult fl ies. In this chapter, we provide a comprehensive overview, introduce the principal of the assays, and provide the details for the conduct of both versions of the assay.

Key words DNA repair , Drosophila melanogaster , Nucleotide excision repair , Base excision repair , In vivo and in vitro assays

1 Introduction

Drosophila melanogaster is a useful model organism for genetic studies, including DNA repair. It is easy to manipulate and to maintain; it has a short life cycle (approx. 14 days at 24 ºC); its xenobiotic compound metabolism system, antioxidant enzymes, and repair DNA systems are similar/equivalent to those in mammals [ 1 , 2 ]; it is very well characterized and cataloged [3 ]. Nowadays, due to restrictions on the use of experimental higher animals in research, Drosophila can play an important role. It is an established insect model for human diseases and toxicological research, recommended by the European Centre for the Validation of Alternative Methods (ECVAM) [ 4 ], and has the important advantage of allowing for the reduction in the number of animals necessary for extensive genotoxicity and histological analysis, in compliance with the 3Rs agenda of the EU to Replace, Reduce, and Reﬁ ne animal experimentation for scientiﬁ c purposes [5 ].

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_23, © Springer Science+Business Media New York 2014 397 398 Isabel Gaivão et al.

For DNA repair studies, there are several different mutant repair strains available, defi cient (to various degrees) for most of the DNA repair systems and the processes involved in DNA damage response (see Table 1 ). Additionally, the availability of methods to induce mutations in targeted genes [6 ] makes the obtaining of mutant strains for specifi c genes a not impossible task. This fact allows the analysis of DNA repair in vivo in germ cells of males [7 ] and females [8 ], in response to the exposure to genotoxic/mutagenic chemicals. However, although this type of analysis would be desirable and of great value in somatic cells, it is apparently not possible with the SMART tests [ 9 ] that are the genotoxicity assays most widely performed in these cells (see Chap. 16 of this book). Fortunately, the application of the comet assay in Drosophila opens the possibility of analyzing DNA repair in vivo in somatic cells, as indicated before [10 –14 ]. In fact, at least for cisplatin, a clear correlation between the levels of DNA adducts and the number of DNA strand breaks (measured as tail moment) was found when analyzing the infl uence of the nucleotide excision repair (NER) system [11 ]. In these works, the repair analysis was carried out with a direct treatment and analysis of wild-type and mutant repair individuals, comparing their responses to DNA damage with statistical regression analysis. This approach is what we call in vivo repair assay . The comet assay in mammals and humans has been recently applied to the analysis of DNA repair in vitro, using cell extracts from cells and tissue samples, and measuring its incision capacity in damaged DNA in vitro using repair enzymes as positive controls (see Chap. 22 of this book). This methodology consists of an incubation of a substrate nucleoid DNA (i.e., protein-depleted nuclei of supercoiled DNA that remain after cell lysis) from cells pretreated with a specifi c damaging agent with a cell-free protein extract. The repair enzymes present in the extracts recognize the damage and the DNA is incised (fi rst step of DNA repair), causing single-strand breaks. Since the comet assay detects these breaks, increased values of comet parameters (e.g., DNA in tail) indicate the DNA repair capacity of the cell extract. This repair activity is called in vitro, because the reaction is carried out with cell extracts on nucleoid DNA. Breaks refl ect the activity of the repair process. Although the commonly accepted terminology for what is measured by this assay is “DNA repair capacity/activity,” it actually represents only the initial incision step of the whole excision repair mechanism. Only if the cells have conditions, deoxyribonucleotides and ATP in the medium would the repair process be complete [15 ]. In this way, the activities of both nucleotide excision repair (NER) and base excision repair (BER) systems have been analyzed in humans [ 16 , 17 ] and in mammals [18 , 19 ], using blood and/or tissue for extract preparation. This approach has been applied to Drosophila as an alternative to study DNA repair activity and has been called the in vitro repair assay . Table 1 List of some of the Drosophila strains, mutant for different DNA damage response (DDR) systems, including DNA repair systems, cell cycle control, and apoptosis control

Mutant straina DDR systemb Referencesc

mus101 CCC [ 20 – 22 ] grp/dmChk1 CCC [ 21 , 23 ] mus304/ATR-IP CCC [ 24 – 26 ] mei41/mus103/mus104/dmATR CCC, HRR [ 21 , 27 – 29 ] dmp53 AP [ 30 – 32 ] mus201/dmXPG NER [ 26 , 33 , 34 ] mus210/dmXPC NER [ 21 , 26 , 33 , 35 ] mus312 NER [ 26 , 36 ] haywire/dmXPB NER [ 21 , 37 ] mei-9/dmXPF NER, MMR, HRR [ 21 , 24 , 26 , 38 ] mus209/dmPCNA MMR, NER, replication [ 21 , 26 , 39 ] spelI/dmMSH2 MMR [ 40 ] dmMsh6 MMR [ 41 ] mus205/dmREV3 TLS [ 21 , 24 , 26 , 42 ] mus309/dmBLM/Ku70 HRR [ 21 , 26 , 43 – 45 ] dmBrca2 HRR [ 46 ] spnA/dmRAD51 HRR [ 47 , 48 ] okr/dmRAD54 HRR, [ 21 , 48 – 50 ] rad50 HRR [ 51 ] mus308/dmPOLQ Alt-EJ, PDD-R [ 52 – 56 ] mus301/spnC/dmHEL308 ICL-R [ 26 ] mus322/snm1 ICL-R [ 26 ] lig4 NHEJ [ 48 , 57 , 58 ] mus207 NHEJ [ 26 , 48 ] mus306 NHEJ [ 26 , 48 ] mus206 SSA, NHEJ, ER [ 26 , 48 , 59 ] ssar SSA [ 48 ] phr Photo-repair [ 21 , 60 ] mus302 ER, PRR [ 24 , 26 , 61 ] mus310 ER [ 24 , 26 , 61 ] a There are more mutant strains possibly related to DDR, because they are sensitive to chemicals, but since their deﬁ ciencies are not known, they are not included here b CCC cell cycle control, HRR homologous recombination repair, AP apoptosis control, NER nucleotide excision repair, MMR mismatch repair, TLS translesion synthesis, SSA single-strand annealing, Alt-EJ alternative end-joining, PDD-R persistent DNA damage repair, NHEJ nonhomologous end-joining, ICL-R interstrand cross-link repair, ER excision repair, PRR post-replication repair c See References on the text 400 Isabel Gaivão et al.

The assay measures the enzymatic activity of cell extracts prepared from adult fl ies, incubated with nucleoids. The extracts can be obtained from any type of strain, such as effi cient repair strains (wild type) or mutant for different repair systems. The assay can be performed using as substrate Drosophila cells, either larva cells treated in vivo or culture cells (not covered in this book), or even mammalian cells from cell lines, incubated with different cell extracts obtained from smashed adult fl ies. As indicated in Chap. 15 of this book, the comet assay in Drosophila can be performed in several cell types of third instar larvae: neuroblast, hemocyte, and anterior mid gut cells. All the studies of DNA repair activity with the comet assay, until now, have been performed with neuroblast and midgut cells, in the in vivo repair assay approach, which uses the standard protocol with different mutant repair strains. Although a detailed protocol for the standard comet assay in neuroblasts is presented in Chap. 15 of this book, detailed protocols to perform the in vitro repair assay with neuroblasts treated in vivo, and with the reference for other different substrates, are presented in this chapter.

2 Materials and Reagents

● All the reagents necessary to prepare the described solutions and buffers (see below). ● All the materials necessary for working with cell cultures (for details see Chap. 22 of this book) ● Coplin jars ● Coverslips 18 × 18, 22 × 22, and 24 × 60 mm ● Electrophoresis power supply, low voltage ● Electrophoresis tank (horizontal) ● Eppendorf/microcentrifuge tubes (0.5 and 1.5 mL) ● Fluorescence microscope with CCD camera ● Freezer (−20 ºC) ● Low melting point agarose (LMP) (e.g., Invitrogen Life Technologies, UK) ● Microscope slides ● Microwave oven ● Normal melting point agarose (NMP) (e.g., Invitrogen Life Technologies, UK) ● Positive controls (repair enzymes), for example: – Formamidopyrimidine DNA glycosilase (FPG), for BER – T4 Endonuclease V (T4 endoV), for NER Comet Assay to Study DNA Repair in Drosophila melanogaster 401

Note: They are available from commercial companies, such as Sigma- Aldrich. In addition, Professor Andrew Collins prepares FPG in his laboratory (University of Oslo, Norway), and Gunnar Brunborg (Norwegian Institute of Public Health, Oslo, Norway) prepares T4 endonuclease V, and usually they sell and send with instructions.

● Refrigerator (4 ºC) ● Small scalpel ● Tungsten wires ● Water bath (37–65 ºC)

2.1 Standard To a 1 L of water, add 100 g of baker yeast, 100 g of sugar, 9 g of Yeast–Sugar Medium agar-agar, and approx. 2 g of NaCl. Mix well, heat, and keep it for Growing Flies boiling for 30 min. Cool it down to 55 ºC, add 5 mL of propionic and Maintaining acid, and mix well. At this moment, it is ready for serving. Strains Note: When the medium cools down, it becomes solid. In this process, water is condensed inside the bottles/vials. Let them dry 24 h before placing the ﬂ ies inside.

2.2 Medium for Fly Carolina Drosophila instant medium Formula 4-24® (Carolina Treatments Biological Supply Company, USA): 3 mL of medium (approx. 0.76 g) hydrated with 3 mL of distilled water (or solvent) or chemical solution.

2.3 Ringer Solution 130 mM NaCl, 35 mM KCl, and 2 mM CaCl 2 . Adjust the pH to 6.5 with NaOH, and sterilize by autoclaving. If not contaminated, it can last until 3 months, at 4 ºC. Autoclave 50 mL aliquots.

2.4 Extraction Buffer 45 mM HEPES, 0.4 M KCl, 1 mM EDTA, 0.1 mM DTT, and 10 % glycerol. Adjust pH to 8 with 6 M KOH, and sterilize by autoclaving. Maintain it at 4 ºC.

2.5 Lysis Buffer 2.5 mM NaCl, 100 mM Na2 EDTA, 10 mM Tris-(hydroxymethyl)- aminomethane, and 0.25 M NaOH. Then add 0.77 % N-lauroylsarcosine sodium salt (30 % aqueous solution), adjust the pH to 10 with HCl, and sterilize by autoclaving. Maintain it in darkness, at 4 ºC, not longer than 1 month. Note : For 1 L of solution, mix and dissolve 146.10 g NaCl, 37.22 g

Na2EDTA, 1.21 g Tris-(hydroxymethyl)-aminomethane, 10.00 g NaOH. When all this is dissolved, add 23.10 mL N-lauroylsarcosine sodium salt (30 % aqueous solution). Keep it in darkness from this moment.

2.6 Lysis Work On the assay day, mix 89 % lysis buffer, 10 % dimethyl sulfoxide Solution (DMSO), and 1 % Triton X-100. Mix well and keep it stirring in darkness until use. 402 Isabel Gaivão et al.

Note: For 200 mL add 178 mL of lysis buffer, 20 mL of DMSO, and 2 mL of Triton X-100.

2.7 Reaction Buffer 40 mM HEPES, 0.1 M KCl, 0.5 mM EDTA, 0.2 mg/mL BSA, adjust pH to 8 with 6 M KOH, and sterilize by autoclaving. Maintain it at 4 ºC. Note: It is also possible to prepare 10x concentrate buffer and frozen at −20 ºC, in 50 mL aliquots.

2.8 Denaturation 1 mM Na 2 EDTA and 300 mM NaOH. Adjust the pH to 12.6 with and Electrophoresis HCl, or check that it is over 13. Prepare it fresh from stock solu-

Buffer tions, with cold water, and keep it at 4 ºC. The Na 2 EDTA stock is sterilized by autoclaving. Both stocks are maintained at room temperature for at least 1 month.

Note: The stocks are 100 mM Na2EDTA and 5 M NaOH. The solution volume prepared would depend on the size of the electrophoresis tank. Note: For Drosophila in vivo treated cell substrate, the pH is 12.6. For mammalian cell substrates, the pH is over 13.

2.9 Neutralization 0.4 M Tris-(hydroxymethyl)-aminomethane, adjust the pH to 7.5 Buffer with HCl and sterilize by autoclaving. Maintain it at room temperature for at least 1 month. Note: PBS 1X can also be used for neutralization and it is cheaper.

2.10 Ethidium Dissolve 1.8 mg in 9 mL of distilled water to reach a 0.2 mg/mL Bromide (Stock concentration. Mix and keep it at 4 ºC and darkness. Solution) Note: Depending on the available microscope ﬁ lter, it is also possible to use other ﬂ uorescent dyes, e.g., DAPI. For cell culture solutions and procedures, see Chap. 22 of this book.

3 Methods

3.1 The In Vitro This assay consists of the analysis of the repair activity displayed by Repair Assay cellular extracts, obtained from adult individuals of Drosophila . These fl ies might be from strains, effi cient and defi cient for most of the DNA repair systems, to evaluate the difference between strains, or from only one Drosophila strain, treated with different compounds/conditions, to evaluate the effect of treatment on the DNA repair systems. The extracts are incubated on substrate damaged DNA. This substrate DNA is nucleoid DNA embedded in agarose gels obtained from cells treated with specifi c damaging agents. The analysis is performed comparing the repair activity of each mutant strain with that of an effi cient repair wild-type strain in the fi rst case and comparing the effect of treatments in the second one. Comet Assay to Study DNA Repair in Drosophila melanogaster 403

3.1.1 Cell Extracts For each of the repair efﬁ cient and deﬁ cient strains:

● Sleep adult ﬂ ies (with ethylic ether or CO2 ) and place 75 females and 75 males in cold glass mortars on ice. Males and females are used to avoid sex-effects. Note: Do not use very old ﬂ ies (between 1 and 2 weeks after adult eclosion).

● Smash the ﬂ ies in 500 μL of cold extraction buffer with the glass pestles keeping the mortars on ice. ● Divide the solution in ten 50 μL aliquots, in 1.5 mL cryotubes/microcentrifuge tubes, and immediately place them in liquid nitrogen and store at −80 ºC until use in the repair assay. Note: Different numbers of ﬂ ies can be used, but with these numbers differences among strains can be detected. For use in the assay:

● Thaw on ice the number of aliquots necessary, knowing that extracts for six slides are obtained from one of them. ● Add 12 μL of Triton X-100 at 1 % to every aliquot, vortex for 5 s, and place on ice for 5 min. Centrifuge at high speed for 10 min and take 50 μL supernatant to a new tube. ● Add 200 μL of cold reaction buffer to each tube and keep them at −20 ºC until use. Note: The protein concentration at this point should be around 2.5 μg/μL, for all the strains.

3.1.2 Cell Substrates The aim is to produce cells with a high level of speciﬁ c DNA damage so that, after embedding in agarose and lysis, the nucleoids would contain excess damage for the enzymes in the extract to work on (lesions are the enzyme substrate) while keeping the level of background strand breaks low. Repair enzymes present in the extracts would recognize the damage present on the substrate, converting them in the DNA strand breaks detected by the assay. Different cell types can be used.

Drosophila Cell Substrate The DNA substrates consist of gel-embedded neuroblasts nucleoids from wild-type larvae that were treated in vivo with a known genotoxic agent. Incubation of these substrate nucleoids with cell or tissue extracts allows the initial steps of BER or NER to occur. This will result in single-strand breaks that can be determined by subsequent single cell electrophoresis (comet assay). Thus, the increase in comet parameter values (e.g., % of DNA in the tail) is proportional to the DNA repair incision activity of the extracts. A schematic overview of the assay and the principle is shown in Fig. 1 . The use of Drosophila cells as substrate has the advantage that larvae are treated in vivo, and then the whole experiment will 404 Isabel Gaivão et al.

subtracts extract

Treated with specific agent adult flies

3th instar larvae brain Cell culture

Protein enzyme Cell embeded in LMP agarose extraction

lysis incubation extract

Nucleoide with specific damage electrophoresis

Breaks due to enzyme repair activity

Fig. 1 Steps and overview of the in vitro comet repair assay. Substrate cells ( Drosophila brain cells or human/ mammalian culture cells) are exposed and are treated with genotoxic agents to induce speciﬁ c DNA lesions (e.g., Drosophila in vivo treatment with chemicals with well-known mechanisms of action; cultured cells exposed to photosensitizer Ro-19-8022 plus visible light or to UV(C)). After lysis, gel-embedded nucleoids are incubated with cell/tissue extracts. Subsequent standard single cell gel electrophoresis reveals the incisions (detected as single-strand breaks) introduced by the DNA repair enzymes

provide information about the capability of the analyzed chemical to induce DNA strand breaks in vivo. Additionally, another important advantage is that the endogenous xenobiotic metabolism, equivalent to that present in mammals, is working. The main disadvantages are the laborious process (compared to cell cultures) and that there could be some inter individual genetic variation, not expected when working with cell cultures. For the analysis of the repair activity of different mutant repair strains, chemicals inducing DNA damages that would be repaired by the studied repair systems should be chosen.

Protocol Since a very detailed protocol for larva treatment, brain ganglia isolation, and neuroblast cell handling is presented in Chap. 15 , detailed information for the speciﬁ c steps of this repair assay and a summary of the common processes are present in this chapter:

● In vivo treatment : Larvae from an efﬁ cient repair wild-type strain are treated in Carolina instant medium hydrated with the chemical solution(s) or distilled water/solvent (as negative control), for 12 h. 10–15 larvae should be treated per vial as described. Only one chemical concentration is generally used in addition to the negative control. This concentration should be used to induce DNA damage but not a high level of DNA strand breaks. Comet Assay to Study DNA Repair in Drosophila melanogaster 405

For both chemical and negative controls, the number of vials should equal the number of slides/half slides. And each extract/ buffer should be analyzed in duplicates. One of the extracts should always be that of the efﬁ cient repair wild-type strain.

● Brain ganglia isolation : As described in Chap. 15 of this book, each larva is placed on a Ringer solution drop (15 μL) under the stereomicroscope with diascopic illumination, and, with the help of tweezers and a small scalpel, it is cut it in two parts. Brain ganglia should remain in the anterior part of the larva, and they are isolated, cleaned, and taken with two tungsten wires into another Ringer solution drop (15 μL), where they should be torn apart with the help of the wires, to help with the freeing of neuroblast cells. Each slide should contain the brain ganglia of four larvae. Add 15 μL of Ringer solution to the drop with the brains of four larvae, and pipet up and down several times to individualize cells. Take everything to one Eppendorf tube and keep it on ice and darkness until all the work with the remaining slides is ﬁ nished.

● Agarose slide preparation : Use one end frosted slides cleaned in ethanol for 24 h, at −20 ºC. For each slide, prepare the fi rst agarose layer with 150 μL of normal melting point (NMP) agarose, at 0.5 % in water, and spread it with the fi nger. After the agarose dries, by heating the slides at 60–65 ºC for 15–20 min, keep the slides at room temperature at least for 1 day. The second layer should be prepared with low melting point (LMP) agarose, at 0.73 % in water, adding 65 μL of this agarose to the 30 μL Ringer solution containing the brain cells (the fi nal LMP agarose concentration should be 0.5 %). Mix by pipetting but avoid bubbles. Spread the resultant 95 μL on one slide with the help of one coverslip (24 × 60 mm). Alternatively, you can use the two-gel format, adding two times 70 μL in one slide, spreading each one with one 18 × 18 mm coverslip. Place the slides at 4 ºC and darkness, for at least 15 min, and carefully remove the coverslips. A third layer is optional; if you choose it, add 75 μL of 0.5 % LMP agarose over the second layer, and spread it with a new coverslip. Again, place the slides at 4 ºC and darkness for at least 15 min, and remove the coverslips. ● Cell lysis should be carried out in Coplin jars with lysis solution freshly prepared, at 4 ºC and darkness for 2 h. Note: Lysis time can be raised until 12 h without problem, but nothing is gained with respect to 2 h.

● After lysis, wash the slides with reaction buffer two times, for 10 min (in Coplin jars), and placed them in a cold surface waiting for the extract/buffer incubation. 406 Isabel Gaivão et al.

Culture Cell Substrate The use of cell cultures treated with a speciﬁ c damaging agent as subtract has the advantages that is much less laborious; it is more homogeneous and allows division in aliquots with the right number of cells for one assay; store at −80 ºC until use. It is possible to use mammal/human cell culture. The protocol for the use of mammal/human cell culture is presented in Chap. 22 of this book. Brieﬂ y as example: HeLa cells can be used, cultured in dishes with D-MEM (Sigma-Aldrich ) medium supplemented with 10 % fetal calf serum (Sigma-Aldrich) and penicillin/streptomycin, at

37 ºC in a 5 % CO 2 atm. Near-conﬂ uent cell cultures are incubated with 1 μM Ro 19-8022 (F. Hoffmann-La Roche) and irradiated on ice with visible light (5 min at 30 cm from a 500 W tungsten halogen source) to induce 8-oxoguanine, as the substrate for BER. Alternatively, the cells can be irradiated on ice with 1 Jm−2 ultraviolet light (UVC), which creates pyrimidine dimers and 6-4 photoproducts, repaired by NER. Cells are then suspended in freezing medium (D-MEM with 10 % serum and 10 % dimethyl sulfoxide) at 3 × 10 6 cell/mL, frozen slowly to −80 ºC in 400 μL aliquots, and kept at this temperature until needed for repair assays [16 ].

3.1.3 Extract Incubation With the substrate nucleoid DNA embedded in agarose on a microscopic slide, the incubation with the extract would allow to measure repair activity from different repair pathways, depending of the damage induced, and for different mutant repair strains comparing their activity with that of the efﬁ cient repair one or with the activity of speciﬁ c enzymes.

● To the slides placed on a cold surface, add 60 μL of reaction buffer/cell extracts, cover them with coverslips, and place on a humidity chamber for 30 min. Place this chamber at room temperature (24 ºC) in the case of Drosophila cells and at 37 ºC in the case of mammalian/human cells. Note: Since for positive controls it is advisable to use repair enzymes, speciﬁ c for the damage induced, and they are usually from bacteria, these slides should be incubated at 37 ºC. Therefore, the origin of the extract (or control enzyme) will determine the incubation temperature.

3.1.4 Comet Assay After the incubation with the extracts, the comet assay is resumed with the denaturation, electrophoresis, neutralization, ﬁ xation, and staining steps, as follows:

● Denaturation and electrophoresis : Avoiding empty spaces among them, place the slides in an electrophoresis tank, located in a cold chamber or in a box with ice, and cover them with cold denaturing buffer, in darkness. Comet Assay to Study DNA Repair in Drosophila melanogaster 407

Note: As indicated above, when working with Drosophila cells, the buffer pH should be 12.6; for the rest of cell types, the pH should be over 13. After 20 min, connect the tank to a power supply, at 0.9 V/cm (across the platform holding the slides) and 300 mA, during another 20 min.

● Neutralization and fi xation : Wash each slide two times with 2 mL of neutralization buffer and place them vertically in a tray. Alternatively, place the slides in Coplin jars and wash them with the buffer for 5 min, three times. Place the slides fl at in a tray and cover them with absolute ethanol for 3–5 min. After that, let them dry overnight at room temperature and darkness. Code the slides, for blind scoring. ● Staining and microscope scoring : Prepare the working 1x solution ethidium bromide from the stock, and dilute it 1:4. Use 40 μL of this dilution (0.4 μg/mL) and 1 μL of fl uorescence protector Vectashield® (Vector Laboratories, Inc., Burlingame, CA 94010, USA) for staining each slide. Spread with a coverslip. In the microscope, use an excitation fi lters adequate for the ethidium bromide (530–560 nm) staining. Analyze at least 50 cells per slide; depending on the system analysis, save images or analyze them in real time . For image analysis, we use the software Komet 5 (Kinetic Imaging Ltd, UK), but there are other software programs available in the market. We collect the information provided for four comet parameters: tail DNA, tail length, tail extent moment, and olive tail moment. Tail DNA is the percentage of DNA that is in the tail. Tail length is the length of the tail, measured in μm, from the border of the head. Tail extent moment is the product of the tail DNA and tail length divided by 100. Olive tail moment is the product of the tail DNA and the difference between tail mean and head mean (profi le centers of gravity) divided by 100.

3.1.5 Statistical Analysis As none of the parameters used (tail DNA, tail length, tail extend moment, and olive tail moment) follows a normal distribution, the comparison between the results of buffer and different extracts and between extracts in each experiment should be performed with a nonparametric statistical test, such as the Mann–Whitney U -test. We use the STATISTICA software for Windows (StatSoft, Inc., 1995. STATISTICA for Windows. Computer Program Manual. Tulsa, OK, USA). To improve the statistical analysis of results, three different independent experiments can be performed for each extract and type of induced DNA damage. Then comparisons between them are carried out comparing the arithmetic means of the average 408 Isabel Gaivão et al.

values of the three experiments with a Student t -test. This test can be performed with any statistical program, even with the Excel software.

Statistical Procedures When using Drosophila cells treated in vivo as substrate, especially Using Drosophila In Vivo if the treatment induces DNA strand breaks, the analysis of extract Treated Cells as Substrate repair activities should take into account the fact that quite important differences in the level of DNA damage/adducts can be present between untreated and treated cells. Therefore, comparison between extracts should be limited to their effects on equivalent amounts of DNA damage, because the effect of the extracts on spontaneous DNA damage may be proportionally larger than that on induced DNA damage, for instance, because of saturation. Moreover, in this case, the goal is the determination of the repair activity of mutant repair strains, and this is achieved by comparing the effects of their extracts with those of an effi cient repair one. For every analysis, then, six values for each of the comet parameters should be necessary: (1) untreated cells exposed to buffer (background control, BC), which indicates the level of spontaneous DNA damage inducing strand breaks; (2) treated cells exposed to buffer (treatment control, TC), which allows the estimation of direct chemically induced DNA strand breaks (TC-BC); (3) untreated cells exposed to extracts from effi cient repair strain (effi cient repair activity on background control, ERBC), which gives information on the DNA strand breaks generated by the effi cient repair extracts on spontaneous DNA damage (ERBC-BC); (4) treated cells exposed to extracts from effi cient repair strain (effi cient repair on treatment control, ERTC), which indicates the levels of DNA strand breaks generated by the effi cient repair extracts on specifi c plus spontaneous DNA damage (ERTC-TC); (5) untreated cells exposed to extracts from mutant repair strains (mutant repair activity on background control, MRBC), which allows the estimation of the DNA strand breaks generated by the mutant extracts on spontaneous DNA damage (MRBC-BC); and (6) treated cells exposed to extracts from mutant repair strains (mutant repair activity on treatment control, MRTC), which gives information about the DNA strand breaks generated by the mutant repair extracts on specifi c plus spontaneous DNA damage (MRTC-TC). In these analyses, contrary to the analysis performed when working with mammalian cells as substrate, the effects of the extracts on spontaneous DNA damage cannot be subtracted from their effects on the treated cells, because we can end up with negative values. To start the analysis, treatment control should be compared with background control to check the induction of DNA strand breaks by the treatment (TC versus BC ). After this, the results of the extracts should be compared with the results of the buffer, to check for the possible incision activity of the extracts (ERBC versus Comet Assay to Study DNA Repair in Drosophila melanogaster 409

BC , MRBC versus BC , ERTC versus TC , and MRTC versus TC ). Finally, the effects of the extracts of the mutant repair strain are compared to those of the effi cient repair one, both for spontaneous and induced DNA damage, fi rst to check for statistically signifi cant differences ( ERBC versus MRBC and ERTC versus MRTC ) and second to estimate the relative repair activity of the mutant strain on both types of damage: Relative activity on spontaneous DNA damage%/ MRBC BC ERBC  BC 100   Relative activity on induced DNA damage%/ MRTC TC ERTC  TC 1000  

Statistical Procedures There is no one way to analyze these data. An independent-sample Using Cultured Cells t- test can be used to assess the differences between groups. To as Substrate determine the possible association between the different variables in the whole sample, a Pearson correlation coeffi cient can be used. An intraclass correlation coeffi cient can be used to observe comet assay test–retest reliability. Coeffi cients of variation (CV) can be useful to analyze repair rate variability. Data analysis can be performed using the software “Statistical Program for Social Sciences (SPSS),” for example.

3.2 The In Vivo This assay consists of the performance of the standard comet assay Repair Assay in vivo, with larvae from different Drosophila effi cient and defi cient repair strains. The analysis is carried out comparing the DNA damage response of each mutant strain to that of the effi cient one, with regression analyses. To be able to do it, the chemical concentrations analyzed should be within the linear part of the dose–response relationship and should be the same for all the analyzed strains. However, this need might represent a problem because mutant repair strains are in general more sensitive than wild-type strains, and therefore the chosen concentrations might not be within the linear part of the dose–response curve for all the strains. All the methodology to perform the standard comet assay is presented in detail in Chap. 15 of this book and also above, if the extract incubation step is removed.

3.2.1 Statistical Analysis As indicated above, the comparison between the result of each chemical concentration and the corresponding negative control should be performed with a nonparametric statistical test, such as the Mann–Whitney U -test. As well, three different independent experiments should be performed for each analyzed chemical in each Drosophila repair strain. Then, as already described, comparisons between each chemical concentration and their negative control are carried with a Student t -test. The comparisons between the efﬁ cient repair wild-type strain and the deﬁ cient ones are carried out with dose–response 410 Isabel Gaivão et al.

regression analysis, checking and comparing the slopes of the respective regression lines: checking whether the regression slopes are statistically different from zero and different between the compared strains. For this analysis we normally used the LightStat3 program, developed by our colleague Dr. P. Casares, and freely distributed. Nevertheless, other programs can be also used.

Acknowledgments

The authors thank the ﬁ nancial support of their research activity: LMS to MEC Spain (project CT2004-03005), FICYT (PCTI Asturias, project PC07-018), and Instituto Universitario de Oncología del Principado de Asturias, Obra Social Cajastur. IG to the Portuguese Science and Technology Foundation (FCT) under the Project PEst-OE/AGR/UI0772/2014.

References

1. Søndergaard L (1993) Homology between the to chemically-induced DNA modifi cations. mammalian liver and the Drosophila fat body. Mutat Res 476:149–165 Trends Genet 9:193 10. Bilbao C, Ferreiro JA, Comendador MA et al 2. Sekelsky JJ, Brodsky MH, Burtis KC (2000) (2002) Infl uence of mus201 and mus308 DNA repair in Drosophila: insights from the mutations of Drosophila melanogaster on the Drosophila genome sequence. J Cell Biol genotoxicity of model chemicals in somatic 150:F31–F36 cells in vivo measured with the comet assay. 3. Adams MD, Celniker SE, Holt RA et al (2000) Mutat Res 503:11–19 The genome sequence of Drosophila melano- 11. García Sar D, Aguado L, Montes Bayón M et al gaster. Science 287:2185–2195 (2012) Relationships between cisplatin- 4. ECVAM http://ihcp.jrc.ec.europa.eu/our_ induced adducts and DNA strand-breaks, activities/alt-animal-testing-safety-assessment- mutation and recombination in vivo in somatic chemicals cells of Drosophila melanogaster, under different conditions of nucleotide excision repair. 5. European Union (2010) Directive 2010/63/ Mutat Res 741:81–88 EU of the European parliament and of the Council of 22 September 2010 on the protec- 12. Siddique HR, Sharma A, Gupta SC et al (2008) tion of animals used for scientifi c purposes DNA damage induced by industrial solid waste NCIL. Offi cial J European Union leachates in Drosophila melanogaster: a mechanistic approach. Environ Mol Mutagen 49(3): 6. Thibault ST, Singer MA, Miyazaki WY et al 206–216 (2004) A complementary transposon tool kit for Drosophila melanogaster using P and pig- 13. Mishra M, Sharma A, Negi MP et al (2011) gyBac. Nat Genet 36:283–287 Tracing the tracks of genotoxicity by trivalent and hexavalent chromium in Drosophila mela- 7. Vogel EW, Nivard MJM, Ballering LAB et al nogaster. Mutat Res 722:44–51 (1996) DNA damage and repair in mutagene- 14. Mishra M, Sharma A, Shukla AK et al (2013) sis and carcinogenesis: implications of Transcriptomic analysis provides insights on structure-activity relationships for cross-species hexavalent chromium induced DNA strand extrapolation. Mutat Res 353:177–218 breaks and their possible repair in midgut cells 8. Hernando J, Alvarez L, Ferreiro JA et al (2004) of Drosophila melanogaster larvae. Mutat Res Female germ cell mutagenicity of model chem- 747–748:28–39 icals in Drosophila melanogaster: mechanistic 15. Collins AR, Fleming IM, Gedik CM (1994) In information and analysis of repair systems. vitro repair of oxidative and ultraviolet-induced Mutat Res 545:59–72 DNA damage in supercoiled nucleoid DNA by 9. Vogel EW, Nivard MJM (2001) Phenotypes of human cell extract. Biochim Biophys Acta Drosophila homologs of human XPF and XPG 1219:724–727 Comet Assay to Study DNA Repair in Drosophila melanogaster 411

16. Gaivao I, Piasek A, Brevik A et al (2009) 29. Laurençon A, Purdy A, Sekelsky J et al (2003) Comet assay-based methods for measuring Phenotypic analysis of separation-of-function DNA repair in vitro: estimates of inter- and alleles of MEI-41, Drosophila ATM/ intra-individual variation. Cell Biol Toxicol ATR. Genetics 164:589–601 25:45–52 30. Brodsky MH, Nordstrom W, Tsang G et al 17. Slyskova J, Korenkova V, Collins AR et al (2000) Drosophila p53 binds a damage (2012) Functional, genetic, and epigenetic response element at the reaper locus. Cell aspects of base and nucleotide excision repair in 101:103–113 colorectal carcinomas. Clin Cancer Res 31. Jin S, Martinek S, Joo WS et al (2000) 18:5878–5887 Identifi cation and characterization of a p53 18. Langie SA, Kowalczyk P, Tudek B et al (2010) homologue in Drosophila melanogaster. Proc The effect of oxidative stress on nucleotide- Natl Acad Sci U S A 97:7301–7306 excision repair in colon tissue of newborn pig- 32. Ollmann M, Young LM, Di Como CJ et al lets. Mutat Res 695:75–80 (2000) Drosophila p53 is a structural and func- 19. Langie SA, Cameron KM, Waldron KJ et al tional homolog of the tumor suppressor p53. (2011) Measuring DNA repair incision activity Cell 101:91–101 of mouse tissue extracts towards singlet 33. Sekelsky JJ, Hollis KJ, Eimerl AI et al (2000) oxygen-induced DNA damage: a comet-based Nucleotide excision repair endonuclease genes in vitro repair assay. Mutagenesis 26:461–471 in Drosophila melanogaster . Mutat Res 20. Fujikawa K, Kondo S (1986) DNA repair 459:219–228 dependence of somatic mutagenesis of 34. Calleja FM, Nivard MN, Eeken JC (2001) transposon-caused white alleles in Drosophila Induced mutagenic effects in the nucleotide melanogaster after treatment with alkylating excision repair defi cient Drosophila mutant agents. Genetics 112:505–522 mus201(DI) , expressing a truncated XPG pro- 21. Henderson DS (1999) DNA repair defects and tein. Mutat Res 461:279–288 other mustakes in Drosophila melanogaster. 35. Henning KA, Peterson C, Legerski R et al Methods 18:377–400 (1994) Cloning the Drosophila homolog of 22. Yamamoto RR, Axton JM, Yamamoto Y et al the xeroderma pigmentosum complementa- (2000) The Drosophila mus101 gene, which tion group C gene reveals homology between links DNA repair, replication and condensation the predicted human and Drosophila polypep- of heterochromatin in mitosis, encodes a pro- tides and that encoded by the yeast RAD4 tein with seven BRCA1 C-terminus domains. gene. Nucleic Acids Res 22:257–261 Genetics 156:711–721 36. Yildiz O, Majumder S, Kramer B et al (2002) 23. Fogarty P, Campbell SD, Abu-Shumays R et al Drosophila MUS312 interacts with the nucleo- (1997) The Drosophila grapes gene is related to tide excision repair endonuclease MEI-9 to checkpoint gene chk1/rad27 and is required generate meiotic crossovers. Mol Cell for late syncytial division fi delity. Curr Biol 10:1503–1509 7:418–426 37. Mounkes LC, Jones RS, Bee-Choo L et al 24. Boyd JB, Mason JM, Yamamoto AH et al (1992) A Drosophila model for xeroderma pig- (1987) A genetic and molecular analysis of mentosum and Cockayne’s syndrome: haywire DNA repair in Drosophila. J Cell Sci Suppl encodes the fl y homologue of ERCC3 , a 6:39–60 human excision repair gene. Cell 71:925–937 25. Brodsky MH, Sekelsky JJ, Tsang G et al (2000) 38. Sekelsky JJ, McKim KS, Chin GM et al (1995) mus304 encodes a novel DNA damage check- The Drosophila meiotic recombination gene mei- point protein required during Drosophila 9 encodes a homologue of the yeast excision development. Genes Dev 14:666–678 repair protein R ad1. Genetics 141:619–627 26. Laurençon A, Orme CM, Peters H et al (2004) 39. Henderson DS, Banga SS, Grigliatti TA et al A large-scale screen for mutagen-sensitive loci (1994) Mutagen sensitivity and suppression of in Drosophila. Genetics 167:217–231 position-effect variegation result from muta- 27. Hari KL, Santerre A, Sekelsky JJ et al (1995) tions in mus209 , the Drosophila gene encoding The mei-41 gene of D. melanogaster is a struc- PCNA. EMBO J 13:1450–1459 tural and functional homolog of the human 40. Flores C, Engels W (1999) Microsatellite insta- ataxia telangiectasia gene. Cell 82:815–821 bility in Drosophila spellchecker1 (MutS homo- 28. Garner M, van Kreeveld S, Su TT (2001) mei- log) mutants. Proc Natl Acad Sci U S A 41 and bub1 block mitosis at two distinct steps in 96:2964–2969 response to incomplete DNA replication in 41. Radford SJ, Sabourin MM, McMahan S et al Drosophila embryos. Curr Biol 11:1595–1599 (2007) Meiotic recombination in Drosophila 412 Isabel Gaivão et al.

Msh6 mutants yields discontinuous gene con- 52. Boyd JB, Sakaguchi K, Harris PV (1990) version tracts. Genetics 176:53–62 mus308 mutants of Drosophila exhibit hyper- 42. Eeken JC, Romeijn RJ, de Jong AW et al sensitivity to DNA cross-linking agents and are (2001) Isolation and genetic characterisation defective in a deoxyribonuclease. Genetics of the Drosophila homologue of (SCE)REV3, 125:813–819 encoding the catalytic subunit of DNA poly- 53. Aguirrezabalaga I, Sierra LM, Comendador merase zeta. Mutat Res 485:237–253 MA (1995) The hypermutability produced by 43. Beall EL, Rio DC (1996) Drosophila IRBP/Ku the mus308 mutation of Drosophila is not spe- p70 corresponds to the mutagen-sensitive cifi c for cross-linking agents. Mutat Res mus309 gene and is involved in P-element exci- 336:243–250 sion in vivo. Genes Dev 10:921–933 54. Harris PV, Mazina OM, Leonhardt EA et al 44. Kusano K, Berres ME, Engels WR (1999) (1996) Molecular cloning of Drosophila Evolution of the RECQ family of helicases: a mus308 , a gene involved in DNA cross-link Drosophila homolog, Dmblm , is similar to the repair with homology to prokaryotic DNA human bloom syndrome gene. Genetics polymerase I genes. Mol Cell Biol 16: 151:1027–1039 5764–5771 45. Adams MD, McVey M, Sekelsky JJ (2003) 55. Chan SH, Yu AM, McVey M (2010) Dual roles Drosophila BLM in double-strand break repair for DNA polymerase theta in alternative end- by synthesis-dependent strand annealing. joining repair of double-strand breaks in Science 299:265–267 Drosophila. PLoS Genet 6:e1001005 46. Klovstad M, Abdu U, Schüpbach T (2008) 56. Díaz-Valdés N, Comendador MA, Sierra LM Drosophila brca2 is required for mitotic and (2010) Mus308 processes oxygen and nitrogen meiotic DNA repair and effi cient activation of ethylation DNA damage in germ cells of the meiotic recombination checkpoint. PLoS Drosophila. J Nucleic Acids ID 416364, Genet 4:e31 doi: 10.4061/2010/416364 47. Staeva-Vieira E, Yoo S, Lehmann R (2003) An 57. Gorski MM, Eeken JC, de Jong AW et al essential role of DmRad51/SpnA in DNA (2003) The Drosophila melanogaster DNA repair and meiotic checkpoint control. EMBO Ligase IV gene plays a crucial role in the repair J 22:5863–5874 of radiation-induced DNA double-strand 48. Wei DS, Rong YS (2007) A genetic screen for breaks and acts synergistically with Rad54 . DNA double-strand break repair mutations in Genetics 165:1929–1941 Drosophila. Genetics 177:63–77 58. Romeijn RJ, Gorski MM, van Schie MA et al 49. Ghabrial A, Rayand RP, Schüpbach T (1998) (2005) Lig4 and rad54 are required for repair okra and spindle-B encode components of the of DNA double-strand breaks induced by RAD52 DNA repair pathway and affect meio- P-element excision in Drosophila. Genetics sis and patterning in Drosophila oogenesis. 169:795–806 Genes Dev 12:2711–2723 59. Huang W, Smith PD (1997) The mus206 gene 50. Kooistra R, Pastink A, Zonneveld JB et al of Drosophila melanogaster is required in the (1999) The Drosophila melanogaster excision repair of alkylation-induced DNA DmRAD54 gene plays a crucial role in double- lesions. Mutat Res 384:81–88 strand break repair after P-element excision and 60. Boyd JB, Harris PV (1987) Isolation and acts synergistically with Ku70 in the repair on characterization of a photorepair-defi cient X-ray damage. Mol Cell Biol 19:6269–6275 mutant in Drosophila melanogaster . Genetics 51. Gorski MM, Romeijn RJ, Eeken JC et al (2004) 116:233–239 Disruption of Drosophila Rad50 causes pupal 61. Harris PV, Boyd JB (1993) Reevaluation of lethality, the accumulation of DNA double- excision repair in the mus304 , mus306 and strand breaks and the induction of apoptosis in mus308 mutants of Drosophila. Mutat Res third instar larvae. DNA Repair 3:603–615 301:51–55 Chapter 24

Use of RNA Interference to Study DNA Repair

Elise Fouquerel , Jianfeng Li , Andrea Braganza , Zhongxun Yu , Ashley R. Brown , Xiao-Hong Wang , Sandy Schamus , David Svilar , Qingming Fang , and Robert W. Sobol

Abstract

DNA repair pathways maintain the integrity of the genome and thereby help prevent the onset of cancer, disease, and aging phenotypes [DNA Repair and Mutagenesis, ASM, Washington, DC]. As such, the critical requirement for DNA repair proteins and pathways in response to radiation and genotoxic chemotherapeu- tics implicates DNA repair proteins as prime targets for improving response to currently available anticancer regimens. Although defects in critical DNA repair pathways or proteins can predispose to cancer onset [FASEB J 26:2094–2104, 2012], such cancer-specifi c DNA repair defects offer novel approaches for tumor-selective therapy. To effectively evaluate the functional role of a specifi c DNA repair protein with regard to cell survival, response to genotoxins, and genome stability, it has become standard practice to employ select genetic tools to alter expression of the gene of interest and/or reexpress a mutant transgene [Cancer Res 71:2308–2317, 2011; Mol Cancer Res 8:67–79, 2010; Neuro Oncol 13:471–486, 2011]. A useful approach to reduce or suppress a specifi c gene of interest in cells is RNA interference. Briefl y, RNA interference is a posttranscriptional gene-silencing biological mechanism whereby RNA molecules inhibit gene expression either by translational suppression or by the targeted degradation of specifi c mRNA molecules. Once the gene of interest is suppressed in this manner and validated for gene expression loss, the resulting knockdown (KD) cells can be used for functional analysis to defi ne the cellular impact of gene loss and provide a resource for evaluating mutants of the gene, such as somatic or germ-line mutations, for impact on function [PLoS Genet 8:e1003086, 2012]. Herein, we describe methods to modify human cells via RNA interference as well as methods to validate gene KD and some measures of cellular response to genotoxins to uncover functional DNA repair defects in the absence of the gene.

Key words RNA interference , siRNA , shRNA , Lentivirus , Knockdown , qRT-PCR , PARP , DNA damage response , Cell survival

1 Introduction

In this chapter, we describe several methods we have employed for regulating gene expression in human cells via RNA interference. We present approaches for both transient (siRNA) and stable

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_24, © Springer Science+Business Media New York 2014 413 414 Elise Fouquerel et al.

Fig. 1 Simpliﬁ ed diagram of siRNA and shRNA processing to facilitate gene knockdown. The transduced shRNA is exported from the nucleus to the cytoplasm and processed by Dicer to yield a siRNA-like molecule similar to the transfected siRNA. These are bound with the RISC complex and the Ago proteins to trigger translational suppression or target mRNA degradation

(shRNA) gene knockdown (KD) as well as the details for several procedures routinely used for validation of gene expression loss (Fig. 1 ). Finally, we describe numerous functional endpoints useful in characterizing the impact of the targeted gene depletion with regard to DNA repair and the cellular response to genotoxins.

1.1 RNA Interference RNA interference (RNAi) is a highly speciﬁ c posttranscriptional in Human Cells Using gene-silencing mechanism that is mediated by the introduction of siRNA double-stranded (ds) RNA molecules 21 bp in length, referred to as short interfering RNA or siRNA (Fig. 1 ). One strand of the siRNA is then incorporated into an RNA-silencing complex (RISC) to target the complementary RNA for destruction by a second nuclease, Ago2 [ 1 – 3 ]. Note that there are multiple Ago family proteins (Ago1–Ago4) involved in siRNA and microRNA (miRNA) processing [4 ]. Molecules of siRNA are now routinely used to transiently silence gene expression in a gene- and sequence-speciﬁ c manner [ 5 – 7 ]. Molecules such as siRNAs are important tools to RNAi and DNA Repair 415

investigate DNA repair pathways because the expression of select DNA repair factors can be transiently suppressed (knocked down; KD) with high specifi city and selectivity using these complementary oligoribonucleotide molecules. This method of gene regulation has become invaluable to identify gene function, validate anticancer gene targets, and is currently being evaluated for clinical potential [ 8 ]. The utility of RNAi as a genetic tool for the study of biological pathways and stress responses in mammalian cells has made signifi cant strides since the fi rst demonstration of RNAi in mammalian cells in 2001 [6 ]. The siRNA can be delivered directly to cells using multiple approaches, including microinjection, electroporation, and transfection [ 9 ]. Currently there are two variations of siRNA transfection procedures being used: forward transfection and reverse transfection. They differ in the order and timing of the addition of the three necessary components for transfection: (1) the siRNA molecules, (2) the lipid-based transfection reagent, and (3) the cells to be targeted. In the forward (standard) transfection protocol, the siRNA and lipid are complexed together and then added to pre-plated cells (e.g., normally the cells are seeded to the wells of a plate 24 h before addition of the siRNA/lipid complex). In the reverse transfection protocol, the siRNA/lipid mixture is added to the wells of the plate fi rst and then the cells are added subsequently. Reverse transfection offers the fl exibility of testing siRNA reagents at various concentrations or utilizing existing siRNA library resources to perform large-scale screens [10 ]. To demonstrate the utility of siRNA, we will focus on the reverse transfection procedure optimized in our lab and will provide detailed information on the procedures and expected outcomes using siRNA to knock down the DNA repair gene UNG [ 10 ].

1.2 RNA Interference RNAi-mediated knockdown can also be achieved via the expression in Human Cells Using of double-stranded RNA (dsRNA) hairpins of 21–23 bp in length shRNA (Fig. 1 ), referred to as small hairpin RNA or short hairpin RNA (shRNA). The expressed shRNA is transported from the nuclease to the cytoplasm and cleaved by a nuclease (Dicer) into dsRNAs of 21–23 bp in length to yield the desired siRNA and is processed as indicated above. Stable expression of shRNAs can mediate effective gene knockdown [ 11 – 13 ]. Most recently, expression of shRNA in cells is typically accomplished using retroviral or lentiviral vectors although lentiviral vectors have become the most routinely used and are the most commonly available resource. To prepare HIV-based lentivirus particles, the shuttle vector (encoded to express the shRNA and a selection cassette such as puromycin, G418, or EGFP) is co-transfected with the required lentiviral packaging vectors into a packaging cell line such as 293FT [ 14 ], a clonal derivative of the human kidney 293T cell line (Invitrogen). Complete libraries of shRNA shuttle vectors speciﬁ c to any human 416 Elise Fouquerel et al.

or mouse gene are available from a number of commercial sources, including Sigma, Thermo Scientiﬁ c, System Biosciences, among many others. Numerous packaging systems have been reported and most are commercially available or available from Addgene ( http://www.addgene.org ). We routinely use the so-called “third- generation” packaging system in which the gene-speciﬁ c shRNA plasmids are co-transfected into 293FT cells together with the packaging plasmids pMD2.g(VSVG), pRSV-REV, and pMDLg/ pRRE to generate replication-defective viral particles capable of transducing numerous cell types [ 15 , 16 ]. Virus particles are isolated from the cell culture supernatant and may be stored at −80 °C or used immediately for transduction of the target cell. With the proper selection, a stable gene silenced or KD mammalian or human cell line can be readily established [ 10 , 15 , 17 , 18 ].

1.3 Validating Gene A critical and essential aspect of any RNAi experiment is validation Knockdown by of gene knockdown. This is most readily accomplished by a quan- qRT-PCR titative reverse transcription polymerase chain reaction (qRT-PCR) analysis using a real-time qRT-PCR protocol [ 19 ]. Although the specifi city of siRNA is normally very precise [20 ], it is essential to also test the expression of other relevant genes to most effectively conduct your study. For example, when studying the role of DNA polymerase ß (Polß) in response to chemotherapeutic agents [21 ], we routinely test the expression of several of the relevant base excision repair (BER) genes such as MPG, APE1, and PARP1 [22 ]. An advantage of analysis by qRT-PCR is the ready availability of specifi c reagents. Material for RNA isolation or qRT-PCR-ready cell lysates are readily available, and validated primers and probes for real-time qRT-PCR are available for any gene. In general, we routinely analyze a transfected (siRNA) or transduced (shRNA) cell line for gene knockdown using RNA from three separate samples, comparing expression to a control such as the same cell line transfected with a control siRNA or transduced with a GFP-expressing or scrambled shRNA-expressing lentivirus. Analysis across samples is normalized to the expression of ß-actin, measured for each sample. As expected, the goal is to achieve complete gene expression knockdown. In practice, knockdown greater than 75 % is likely to yield a signifi cant loss of steady-state protein levels and result in a functional impact from the loss of the expressed mRNA [15 , 17 , 22 ].

1.4 Validating Gene The ultimate and most stringent validation to demonstrate that you Knockdown by have succeeded in “knocking down” your gene is to show a loss or Immunoblot suppression of protein expression [23 ]. We routinely use immunoblotting to validate the loss of gene expression and in some cases, complement this analysis with a functional test [22 ]. However, the loss of protein, as determined by immunoblotting, is a highly reliable method for validating gene expression loss. This standard laboratory analytical procedure is only limited by the availability of the antibody. RNAi and DNA Repair 417

1.5 PARP Activation Genomic DNA damage from genotoxins must be repaired to in Response prevent gene mutations, aneuploidy, senescence, or cell death that to DNA- can contribute to cancer, early-onset aging, and/or degenerative Damaging Agents diseases [ 24 ]. There are multiple overlapping mechanisms and processes in human cells that govern and orchestrate the response to genotoxins, collectively referred to as the DNA damage response (DDR) [ 25 ]. The initial DDR of a cell involves the recognition of the DNA damage followed by the propagation of a series of signals ranging from alterations in RNA or protein expression and modifi cation of protein function or stability through posttranslational modifi cation, among other signals. The cell’s defense to genotoxic lesions is triggered and accomplished by a series of events that mediate and regulate proliferation, cell death, or DNA repair crucial to its survival [26 , 27 ]. PARP1 is a major response protein in the DDR pathway and binds to and is activated by damaged DNA (strand breaks) to initiate a signal cascade by utilization of NAD + to modify numerous proteins with poly-ADP-ribose (PAR), including histones, poly- merases, topoisomerases, and PARP1 itself [28 ]. PARP1 (as well as PARP2 and PARP3) [29 , 30 ] acts as sensors of DNA damage such as DNA single-strand or double-strand breaks and becomes activated, consuming NAD + to synthesize PAR [15 ]. It is clear that activated PARP1 (together with PARP2 and PARG) facilitates DNA repair via roles in BER [31 ] and nonhomologous end joining (NHEJ) [30 , 32 – 39 ], the latter also involving PARP3 [30 ]. In this context, being able to detect PAR levels in cells is an effective method to defi ne the sensitivity of a cell to a DNA-damaging agent and its capacity to repair such damage. In this chapter, we describe two standard approaches to detect and quantify PAR, either by immunoblot or immunofl uorescence.

1.6 Complementing One of the advantages of developing a stable KD cell line, deﬁ cient Gene KD Cells in the expression of the endogenous gene, is the capacity to then and Evaluating Cell complement the cell line by reexpression of the cDNA, for exam- Survival in Response ple, with the wild-type (WT) cDNA or encoding a mutation to DNA- (Fig. 2 ). By analysis of normal and cancer tissues, it has been found Damaging Agents that most of the 150–200 genes among the DNA repair pathways harbor germ-line or somatic mutations. For example, there are numerous mutations for the BER gene Polß found to have either protein-altering germ-line single-nucleotide polymorphisms (SNPs) or functionally altering somatic mutations when isolated from tumor tissue [40 ]. In fact, it was recently described that “… we can expect all cancer cells to be defective in some aspect of DNA repair … There are at least 150 different proteins that cata- lyze DNA repair … To seed new therapies, geneticists and molecular biologists are needed to explore the detailed consequences of an alteration in each of these repair pathways…” [ 41 ]. To facilitate effective complementation and expression of the transgene while reducing endogenous expression by RNAi, one 418 Elise Fouquerel et al.

Fig. 2 Scheme for transgene complementation after gene knockdown. The target for RNA interference can be within the open reading frame (ORF) of the target mRNA or speciﬁ c to the 3′untranslated region (3′UTR). As indicated, effective transgene design for shRNA-resistant expression of your target gene requires either deletion of the shRNA target site within the ORF, mutation of the shRNA target site within the ORF, or expression of the transgene in cells targeted with a 3′UTR-speciﬁ c shRNA

must either mutate the cDNA to eliminate the RNAi target sequence or use an shRNA specifi c for the 3′untranslated region of the target gene (Fig. 2 ). Once accomplished, the utilization of an RNAi-resistant expression system is a robust method to reexpress or complement a KD cell line. Following gene knockdown, we routinely develop lentiviral-based cDNA expression vectors to complement the KD cells. With such a system, the investigator is open to effectively study KD phenotype rescue, the functional role of specifi c amino acid residues such as those in the enzyme active site, those implicated in protein-protein interactions, or those that may be a target for posttranslational modifi cation. Once in hand, these modifi ed cell lines are then valuable resources to evaluate the role of these proteins (and mutants) in response to genotoxins. Cells treated with cytotoxic compounds will respond differently depending on the DNA repair capacity of the cell. For example, cells expressing the protein MGMT are highly resistant to the cytotoxic effects of the alkylating agent-induced DNA lesion O 6 - methyl- dG, but in the absence of MGMT expression, the lesion is highly toxic [ 22 ]. There are many methods to measure cytotoxicity. Here, we will focus on the MTS assay and the CyQUANT assay that are used to monitor the cytotoxicity of a genotoxin after short-term or long-term exposure to the agent, respectively. The MTS assay measures the reducing potential of the cell using a colorimetric reaction. Viable cells can reduce the agent 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2- (4-sulfophenyl)-2 H -tetrazolium (MTS) to produce a formazan RNAi and DNA Repair 419

product that has an absorbance maximum at 490–500 nm in phosphate-buffered saline (PBS), in the presence of phenazine methosulfate [ 42 ]. The MTS assay is more convenient than the MTT assay, since the reagent can be added to the cells directly without the intermittent steps required in the MTT assay. However, this convenience makes the MTS assay susceptible to colorimetric interference [ 43 ]. In general, one needs to take caution when using this assay, depending on the cell type, the DNA- damaging agent, and the experiment as a whole. Since the assay measures cellular reducing potential (mitochondrial function), a resting cell may have a lower overall level of reducing potential, whereas a highly proliferative cell is likely to have a high reducing potential. In that regard, treatments that induce senescence but not cell death may appear in this assay to reduce cell number. To avoid such potential problems, we routinely test cells using the MTS assay as well as a second assay, as described below. The CyQUANT assay is based on a cell-permeable DNA- binding dye in combination with a background suppression reagent to estimate cell number [ 44 ]. We routinely use this assay to evaluate genotoxic response 8–12 days after exposure to the agent. As DNA content is highly regulated, cell number estimates are very accurate. The masking dye blocks staining of dead cells and cells with compromised cell membranes, causing only healthy cells to be stained. Recent studies using diverse sets of cytotoxic compounds and assay types have shown that cell number (proliferation) is among the most sensitive indicators of cytotoxicity [ 45 ]. In our hands, this assay yields cell survival or proliferation outcomes similar to that of a clonogenic assay if control cells are allowed to undergo seven or more doublings before analysis. Therefore, the CyQUANT assay provides a highly accurate measure of cell number, allowing for the evaluation of proliferation as well as cytotoxicity in response to genotoxin exposure.

2 Materials

2.1 RNA Interference ● siRNA speciﬁ c for human UNG (Ambion; cat# 4390824; in Human Cells Using siRNA ID: s14679) siRNA ● Silencer® Negative Control #2 siRNA (50 μM) (Ambion; AM4613) ● siPORT™ NeoFX™ Transfection Agent (Applied Biosystems; AM4511) ● Opti-MEM® 1 Reduced-Serum Medium (1×), liquid (Invitrogen; 31985-070) ● Trypsin, 0.25 % (1×) with EDTA, liquid (Invitrogen; 25200-114) ● PBS (Invitrogen; 14190-250) 420 Elise Fouquerel et al.

All the listed reagents are stored according to the manufacturers’ instructions.

2.2 RNA Interference ● The three lentiviral packaging plasmids (pMD2.g(VSVG), in Human Cells Using pRSV-REV, and pMDLg/pRRE), available from Addgene shRNA (cat# 12259, #12253, and cat #12251, respectively) ● The 293FT cell line (cat# R700-07), available from Life Technologies ● TransIT-293 (cat# MIR-2704, 2700, 2705, or 2706), available from Mirus ● Millipore Sterifl ip vacuum fi lters (cat# SE1M003M00), available from Fisher Scientifi c (Fisher Scientifi c cat# SC50 FL0 25) ● Lenti-X Concentrators (cat# 631231), available from Clontech ● Ultracentrifuge tubes (Beckman Ultra-Clear tubes, cat# 344058), available from Fisher Scientifi c (Fisher Scientifi c cat# NC9146666)

2.3 Validating Gene ● TaqMan® Gene Expression Cells-to-CT™ Kit (cat# AM1728), Knockdown by available from Life Technologies. qRT-PCR ● PBS (cat# 14190-250), available from Invitrogen.

● DNase and RNase-free H2 O. ● Two heat blocks set to 37 and 95 °C or a PCR thermocycler. ● Applied Biosystems StepOnePlus™ Real-Time PCR System. ● Applied Biosystems StepOne™ Software. ● TaqMan® Gene Expression Assays (the speciﬁ c assay will depend on the gene you are targeting) are available from Life Technologies. ● TaqMan® Gene Expression Master Mix is available from Life Technologies.

2.4 Validating Gene ● Cell extracts diluted twice in the sample buffer (2×). For cell Knockdown by extracts prepared for detection of PAR, see Sect. 2.5. Immunoblot Note: Alternatively and depending on the molecular weight of the protein of interest, a higher or lower concentration of Acryl/ Bis 30 % can be used . ● Sample buffer (2×) (LaemmLi blue 2×: 2 % SDS, 20 % glycerol, 62.5 mM Tris-HCl, pH 6.8, 0.01 % Bromophenol blue). ● Stacking gel buffer (Tris-HCl 0.5 M, pH 6.8, SDS 0.4 %)— also sold as a ready-to-use solution by Bio-Rad (cat# 161- 0799) (stable at room temperature). ● Separating gel buffer (Tris-HCl 1.5 M, pH 8.8, SDS 0.4 %)— also sold as a ready-to-use solution by Bio-Rad (cat# 161- 0798) (stable at room temperature). RNAi and DNA Repair 421

● Isopropanol (100 %). ● Acryl/Bis 30 %; ratio 37.5/1 (Bio-Rad cat#) (keep at 4 °C). ● Ammonium persulfate (APS). ● N , N ,N ′, N ′-Tetramethylethylenediamine (TEMED) (Bio-Rad cat# 161-0800) kept at 4 °C. ● Running buffer, 5× (30 g/L Tris-base, 144 g/L glycine, 0.5 % SDS) stored at room temperature. A 1× solution is prepared by mixing 200 mL of a 5× stock solution and 800 mL of

dH2 O. This solution can be used up to ﬁ ve times and is kept at room temperature. ● Vertical electrophoresis system, 1.0 mm gel thickness (e.g., Mini-PROTEAN® Tetra Cell, Bio-Rad). ● Prestained protein molecular weight markers—for example, Precision Plus Protein Kaleidoscope™ from Bio-Rad (cat# 161-0375). ● Blotting transfer buffer, 10× (30 g/L Tris-base, 144 g/L glycine, 1 % SDS). The stock solution can be stored at room temperature. Before the transfer, prepare a 1× solution by

diluting 100 mL of the 10× stock solution in 615 mL of dH 2 O and adding 285 mL of 70 % ethanol. This solution can be reused up to ﬁ ve times if stored at 4 °C. ● Blotting transfer system (Mini Trans-Blot Electrophoretic Transfer Cell, Bio-Rad). ● Nitrocellulose transfer membrane (Bio-Rad Nitrocellulose Membranes 0.45 μm; cat# 162-0115). ● TBS-Tween 0.3 % buffer (TBT): prepared by diluting Tris- buffered saline, pH 7.4 (10×; Boston BioProducts; cat#

BM-300), in 900 mL dH 2 O and adding 3 mL of 100 % Tween 20. The TBS is also available in powder format (Sigma-Aldrich; cat# D5773). ● Blocking powder (Blotting-Grade Blocker; cat# 170-6404) or low-fat dry milk. Note: BSA can be used for special antibodies if speciﬁ ed by the company . ● Primary antibody against the protein of interest. ● Secondary antibody (we use goat anti-mouse or goat anti- rabbit horseradish peroxidase (HRP) conjugate from Bio-Rad; an alternate secondary antibody can be used depending on the primary antibody species). ● Detection reagents : For expected normal-to-strong signals, we use Immun-Star™ HRP from Bio-Rad (cat# 170-5040). If the signal is expected to be weak or low, a more sensitive product can be used such as SuperSignal ® West Femto Maximum Sensitivity Substrate from Thermo Fischer (cat# 34095). 422 Elise Fouquerel et al.

● Detection system : The HRP signal can be detected using X-ray ﬁ lm. We use Premium X-ray Film from Phenix (#EBA45) or a Bio-Rad Molecular Imager® Chemidoc™ XRS+ system.

2.5 PARP Activation ● DNA-damaging agents speciﬁ c to your experiment (MNNG, in Response methyl methanesulfonate (MMS), H 2 O2 , ionizing radiation, to DNA- etc). In the example shown here, we use MNNG, N ′-methyl- Damaging Agents N ′-nitro- N -nitrosoguanidine (prepared as a 10 mM stock solution in DMSO), available from TCI America (# N0527; 2.5.1 Poly-ADP-ribose Portland, OR). Detection by Immunoblot ● Plated cells. ● Ice-cold PBS (cat# 14190-250), available from Invitrogen. ● Sample buffer 2× (see Sect. 2.4). ● Materials for SDS-PAGE and immunoblot as described in Sect. 2.4. ● Mouse monoclonal PAR antibody 10H [15 , 17 ]. ● Secondary antibody; goat anti-mouse Ab conjugated with HRP.

2.5.2 Poly-ADP-ribose ● Coverslips washed in 5 mL 70 % ethanol Detection by ● Microscope slides Immunoﬂ uorescence ● DNA-damaging agents speciﬁ c to your experiment (MNNG,

MMS, H2 O2 , ionizing radiation, etc) ● Methanol and acetone (100 %), stored at −20 °C ● PBS-Triton X-100, 0.1 % (alternatively, PBS-Tween 0.3 % can be used) ● PBS-BSA, 0.5 % ● PBS-BSA, 2 % ● Primary antibody mouse; monoclonal anti-PAR, clone 10H [15 , 17 ] ● Secondary antibody; goat anti-mouse Ab conjugated with Alexa Fluor 488 ● Mounting solution containing DAPI ( ProLong® Gold Antifade Reagent with DAPI; Life technologies cat# P-36931)

2.6 Complementing Complementation or reexpression of the protein of interest after Gene KD Cells gene KD can be accomplished using any plasmid- or viral-based and Evaluating Cell transgene expression system. We have found success with lentiviral Survival in Response vectors (e.g., from Clontech), allowing expression of DNA poly- to DNA- Damaging merase ß (Polß) with numerous selection agents such as hygromy- Agents cin, G418, or EGFP. Most of the shRNA lentiviral vectors we have used employ a puromycin cassette for selection. Therefore, using 2.6.1 Complementation separate selection agents for complementation is critical. We ﬁ nd Tools no difﬁ culty with transducing a cell multiple times, for example, to RNAi and DNA Repair 423

express shRNA for gene KD followed by transduction with a cDNA expression vector to reexpress the gene of interest or to express mutants of the cDNA [ 42 ]. Once stable cell lines are developed, they yield a valuable resource for analysis of cell survival in response to DNA-damaging agents [10 , 22 , 46 ].

2.6.2 MTS Assay ● Cell culture medium (depends on which cell used). ● PBS (cat# 14190-250), available from Invitrogen. ● MNNG, N ′-methyl-N ′-nitro-N -nitrosoguanidine (prepared as a 10 mM stock solution in DMSO), available from TCI America (# N0527; Portland, OR). ● Promega CellTiter 96 AQueous One, available from Promega (Catalogue No. G356B), aliquot and store at −30 °C. ● Hemocytometer, CASY counter, or similar cell counting device. ● Multi-well aspirator. ● Multichannel pipettor.

2.6.3 CyQUANT Assay ● Cell culture medium (depends on which cell used). ● PBS (cat# 14190-250), available from Invitrogen. ● MNNG, N ′-methyl-N ′-nitro-N -nitrosoguanidine (prepared as a 10 mM stock solution in DMSO), available from TCI America (# N0527; Portland, OR). ● CyQUANT GR reagent (Catalogue No. C7026, Life technologies), store at −30 °C. ● Lysis buffer, store at 4 °C. ● Paraﬁ lm. ● Hemocytometer, CASY counter, or similar cell counting device. ● Multi-well aspirator. ● Multichannel pipettor.

3 Methods

3.1 RNA Interference Validated siRNAs are available from a number of sources and are in Human Cells Using generally available as individual siRNAs or a gene-specifi c pool. siRNA Further, libraries can be obtained for the entire genome or those that are limited to specifi c functional biological pathways or other subgroups such as the druggable genome [ 10 ]. The siRNA provided by Ambion is supplied as a lyophilized powder. The user is advised to centrifuge briefl y at the highest speed and resuspend the powder with nuclease-free water to a fi nal concentration of 2 μM. All the pipette tips should also be 424 Elise Fouquerel et al.

nuclease-free. The siRNA stock solutions can be stored at −20 °C for up to 3 months. For each transfection condition, include three replicate transfections. Remember to include a Negative Control (scrambled) siRNA and a non-transfected control (cells that are mock- transfected with Opti-MEM 1 Medium but no transfection reagent and no siRNA). When possible, prepare master mixes to minimize variability. For this procedure, cells are seeded in each well of a 6-well plate at 3 × 10 5 cells for each well. Note: The amount of cells for transfection could range from 2 × 10 5 to 3 × 10 5 , depending on the cell type. For this specifi c gene (UNG), we used 135 nM siRNA (fi nal concentration after transfection) for transfection. Note: The fi nal concentration should be modifi ed according to the specifi c siRNA. The recommended concentration ranges from 10 to 150 nM. These conditions generally result in effective gene knockdown, with loss of UNG mRNA as much as 65 % in this example (Fig. 3 ), suffi cient to reduce UNG protein levels below detection and to suppress UNG enzymatic activity to background levels [ 10 ]. 1. On the day of transfection, aspirate the growth media from the cells. 2. Wash cells twice with 1× PBS. 3. Aspirate the PBS. 4. Add enough trypsin to cover the cells (0.5 mL) and place the plate in the incubator for 5 min to allow the cells to detach. 5. After 5 min, gently tap the plate/dish to detach any unbound cells from the surface and suspend the cells in Opti-MEM 1 Media (2.5 mL).

Fig. 3 Relative mRNA expression as measured by qRT-PCR after siRNA to knockdown UNG. Quantiﬁ cation of siRNA-mediated UNG knockdown as determined by qRT-PCR. TaqMan probes were used to quantify mRNA levels on an Applied Biosystems StepOnePlus machine. The qRT-PCR data was analyzed using the ΔΔCt method and was normalized to a mock-transfected control. Expression of UNG was normalized to the expression of human ß-actin. The mean of three independent experiments is plotted ± SEM RNAi and DNA Repair 425

Note: In practice, if you cannot prepare the cell solution as quickly as possible, we recommend using normal growth media (with serum) to neutralize the trypsin. Then prepare the cell solution with Opti-MEM 1 Medium. This helps to prevent cell death caused by trypsin . 6. Count cells using either a hemocytometer or an automated cell counter. 7. Make the required dilution of cells to yield a ﬁ nal suspension of 3 × 105 cells per well in a volume of 2 mL per well. Store the cell suspension on ice while you quickly prepare the transfection reagent and siRNAs. Note: Here, the cell solution volume is 1.4 mL and the siRNA transfection reagent complex is 0.6 mL. The total volume for each well is 2 mL. 8. For each well, dilute the 5 μL of siPORT NeoFX in 295 μL Opti-MEM 1 Medium and incubate the diluted siPORT NeoFX for 10 min at room temperature. Note: The 10 min incubation time should be controlled precisely and you can dilute your siRNA with Opti-MEM 1 Medium during this period of time. In addition, you can prepare the master mixes in a 15 mL falcon tube. 9. For each well, dilute 135 μL of 2 μM siRNA in 165 μL Opti- MEM 1 Medium. 10. After the 10 min incubation, mix the diluted siRNA with the diluted siPORT NeoFX in a 1:1 ratio. 11. Incubate at room temperature for 10 min. 12. Dispense 600 μL of the siRNA/siPORT NeoFX complex into the wells of a 6-well culture plate and set up the non-transfected controls (i.e., add only OPTI-MEM 1 Medium to the negative control wells). 13. Overlay 1.4 mL of the cells (3 × 105 cells) per well into each well and mix gently by tilting back and forth to distribute evenly. DO NOT swirl the plate. 14. Incubate plate at 37 °C for 24 h. Note: If you observe cytotoxicity caused by transfection, we recommend changing the media back to normal growth media 8–10 h after transfection instead of 24 h as indicated . 15. After 24 h, wash cells with fresh growth media. 16. Gene expression can be measured by qRT-PCR 24 h after transfection or by immunoblot 48 h after transfection.

3.2 RNA Interference Lentivirus production and cell transduction are essentially as we in Human Cells Using have described previously [10 , 15 , 17 , 22 ]. The method is a classical shRNA procedure for virus production. The shuttle or transfer vector (the 426 Elise Fouquerel et al.

Fig. 4 Demonstration of lentivirus production and transduction. Panels a –c are white light images of 293FT cells before transfection (Panel a ) or 48 h after transfection in the absence of plasmid (Panel b) or in the presence of the viral vectors (Panel c ). Panel d is a ﬂ uorescent image of cells after transduction with an EGFP- expressing lentivirus (color ﬁ gure online)

shRNA-expressing plasmid) is co-transfected into 293FT cells together with the packaging plasmids pMD2.g(VSVG), pRSV-REV, and pMDLg/pRRE to generate replication-defective viral particles capable of transducing numerous cell types [15 , 16 ]. Importantly, the cells should be sub-confl uent when transfected (Fig. 4 , Panel a ). Virus particles can be isolated after 48 h. If the transfection effi ciency is low or if not transfected, the cells will reach confl uence by 48 h (Fig. 4 , Panel b ). However, if transfection is successful, virus production leads to a suppression of cell growth (Fig. 4 , Panel c ). Effective virus preparation should allow for a high level of transduction, approaching 100 %, as shown here for cells transduced with an EGFP-expressing lentivirus (Fig. 4 , Panel d ).

3.2.1 Lentivirus We utilize this procedure for the preparation of ﬁ ve different Preparation lentiviral preparations: 1. Small-scale, low titer : Seed 1.7 × 106 293FT cells (less than 2 weeks old) in a 60 mm dish with 5.5 mL complete growth

media and culture the cells at 37 °C, 5 % CO 2 overnight. RNAi and DNA Repair 427

Transfect as indicated below (Table 1 ). The ﬁ nal yield of the lentiviral particle suspension is 4 mL and the virus titer is approximately 1 × 10 6 transforming units per mL (TU/mL). 2. Medium-scale, low titer : Seed 1.2 × 107 293FT cells (less than 2 weeks old) in a 150 mm dish with 27 mL complete growth

media and culture the cells at 37 °C, 5 % CO 2 overnight. Transfect as indicated below (Table 2 ). The ﬁ nal yield of the lentiviral particle suspension is 25 mL and the virus titer is approximately 1 × 10 6 transforming units per mL (TU/mL). 3. Large-scale, low titer : Seed 1.2 × 107 293FT cells (less than 2 weeks old) in each of three [ 47 ] 150 mm dishes with 27 mL complete growth media each and culture the cells at 37 °C,

5 % CO2 overnight. Transfect as indicated below (Table 2 ). The ﬁ nal volume of the lentiviral particle suspension is 75 mL and the virus titer is approximately 1 × 10 6 transforming units per mL (TU/mL). 4. Medium-scale, medium titer : Seed 1.2 × 107 293FT cells (less than 2 weeks old) in a 150 mm dish with 27 mL complete

growth media and culture the cells at 37 °C, 5 % CO2 overnight. Transfect and concentrate as indicated below (Table 2 ). The virus titer is approximately 1 × 10 8 transforming units per milliliter (TU/mL). An aliquot of 10 μL would normally yield ~100 % transduction efﬁ ciency when infecting a 100 mm dish of cells. 5. Large-scale, high titer : Seed 1.2 × 107 293FT cells (less than 2 weeks old) in three [47 ] 150 mm dishes with 27 mL complete

growth media and culture the cells at 37 °C, 5 % CO2 overnight. Transfect and concentrate as indicated below (Table 2 ). The ﬁ nal volume of the lentiviral particle suspension is 30 μL and the virus titer is approximately 1 × 10 9 TU/mL.

Protocol 1. Seed 293FT cells as indicated above. 2. On the second day, prepare the transfection mix according to the materials listed in Table 1 for transfection of cells in a 60 mm dish or as in Table 2 for transfection of cells in a 150 mm dish. 3. Add the plasmids, media, and transfection reagent to a tube in this order: (1) media, (2) TransIT-293, and (3) plasmids. 4. Tap gently to mix (do not vortex) and incubate at room temperature for 15–30 min. 5. Add the transfection mix to each dish (do not remove the old media), dropwise. 6. After incubation at 37 °C for 48 h, collect culture media containing viral particles from transfected cells and fi lter the collection through a 0.45 μm Sterifl ip vacuum fi lter. 428 Elise Fouquerel et al.

Table 1 Transfection mix recipe for a 60 mm dish

Serum- and Plate Plasmid antibiotic-free TransIT-293 (60 mm) Cell line Plasmid (0.5 μg/μL) media (Mirus)

For each 293FT Transfer vector such as pLLox3.7, 8 μL 0.375 mL 10 μL plate pLVX-IRES, or pLKO.1-puro pMD2.g(VSVG) 4 μL pRSV-REV 4 μL pMDLg/pRRE 4 μL

Table 2 Transfection mix recipe for a 150 mm dish

Serum- and Plate Plasmid antibiotic-free TransIT-293 (150 mm) Cell line Plasmid (0.5 μg/μL) media (Mirus)

For each 293FT Transfer vector such as pLLox3.7, 56 μL 2.7 mL 70 μL plate pLVX-IRES, or pLKO.1-puro pMD2.g(VSVG) 28 μL pRSV-REV 28 μL pMDLg/pRRE 28 μL

7. For the low titer preps, aliquot (1.0 mL per tube) into sterile vials and store at −80 °C. 8. For the medium titer and high titer concentration procedures, see below.

Medium Titer Lentivirus 1. After ﬁ ltering the viral particles, mix 1 vol of Lenti-X Concentration Concentrator with 3 volumes of the ﬁ ltered viral particles (simply measure the amount of viral supernatant to be concentrated, divide by 3, and add the resulting amount of Lenti-X Concentrator to your viral supernatant). 2. Mix by gentle inversion. Larger volumes may be accommo- dated through the use of larger (i.e., 250 or 500 mL) centrifuge tubes. 3. Incubate mixture at 4 °C for 30 min (can be up to overnight and larger volumes (>100 mL) may require longer incubation times). 4. Centrifuge sample at 1,500 × g for 45 min at 4 °C. After centrifugation, an off-white pellet will be visible. 5. Carefully remove the supernatant, taking care not to disturb the pellet. Residual supernatant can be removed with either a pipette tip or by brief centrifugation at 1500 × g . RNAi and DNA Repair 429

6. Gently resuspend the pellet in 25 μL 1× PBS at 4 °C and keep it at 4 °C overnight. The pellet can be somewhat sticky at fi rst, but will go into suspension quickly. The fi nal volume after the overnight incubation will be just over 100 μL. 7. Aliquot (10 μL per tube) into sterile vials. Determine titer of sample immediately or store at −80 °C in single-use aliquots (ten aliquots). 8. An aliquot of 10 μL would normally yield ~100 % transduction effi ciency when infecting a 100 mm dish of cells (50 % confl uent).

High Titer Lentivirus 1. After fi ltering the viral particles, add 36 mL of fi ltered super- Concentration natant to each of two ultracentrifuge tubes and balance the two tubes with additional media or supernatant. It may be useful to titer some of the leftover supernatant to determine if there is a loss of virus during concentration. 2. Cover tubes with a small piece of Parafi lm and spin tubes using an SW-28 rotor at 25,000 rpm for 90 min at 4 °C. 3. Decant liquid and leave tube upside down on KimWipe for 10 min in the tissue culture hood. Aspirate remaining media being careful not to touch the bottom of the tube. 4. Add 12.5 μL cold, sterile tissue culture grade PBS to each tube and cover the tubes with Parafi lm and then keep those tubes at 4 °C overnight without shaking. 5. To resuspend the pelleted virus, hold the tube at an angle and pipette fl uid (the PBS that was added last night) over the pellet 20 times, being careful not to touch the pellet with the pipette tip. It is expected that the pellet will not be resuspended after this is completed. This pellet does not contain virus and can be discarded. 6. Combine the two virus suspensions and aliquot at 5 μL per tube (six tubes), fl ash-freeze in liquid nitrogen, and store at −80 °C. There should be no change in titer with freezing concentrated virus. Avoid multiple freeze-thaws.

3.2.2 Viral Titer Analysis 1. For a rapid titer determination, the Lenti-X qRT-PCR Titration Kit (Clontech Cat. No. 632165) directly quantiﬁ es the viral genomes in your virus stock, which is much faster and often more useful than antibiotic selection since it exploits conserved regions contained in most lentiviral preps. 2. Lenti-X GoStix (Clontech Cat. No. 631244) can also be used for a quick viral titer determination if the viral titer measurement does not need to be highly accurate. It only takes about 30 min.

3.2.3 Lentiviral This protocol is based on a procedure optimized for transduction Transduction Protocol of the human glioma cell line LN428 [ 10 , 15 , 17 , 22 ] but should be applicable to most other cell lines with minimal alteration. 430 Elise Fouquerel et al.

Table 3 Virus dilution recipe for lentiviral transduction

Vol. polybrene (μL) Well Virus Vol. media (mL) Vol. virus (mL) (stock = 8 mg/mL)

1 Negative control (no virus) 2.0 0 2 2 Target shRNA virus 1.0 1.0 2 3 Nontarget shRNA control or GFP control 1.0 1.0 2

Cell Seeding 1. Thaw a vial of cells and culture the cells until they are 80 % conﬂ uent. 2. Passage the cells and culture the cells until they are 50 % conﬂ uent. 3. Trypsinize, count, and then seed 1 × 105 cells per well in a 6-well plate with 2 mL growth medium.

4. Incubate at 37 °C, 5 % CO2 overnight.

Transduction 5. On the second day, prepare the virus dilutions on ice in sterile tubes, according to Table 3 . 6. Remove the growth media from each well. 7. Add the lentivirus preparation to each well.

8. Incubate at 32 °C, 5 % CO2 overnight (16–18 h).

Second Transduction 9. In the morning, replace the lentivirus preparation with 2 mL

(Optional) fresh growth media and incubate at 37 °C, 5 % CO2 for 6–7 h. 10. In the late afternoon, remove the growth media from each well and add the lentivirus preparation (new prep but identical to the ﬁ rst transduction) to each well, according to Table 3 .

11. Incubate at 32 °C, 5 % CO2 overnight (16–18 h). 12. The next morning, replace the lentivirus preparation with 2 mL

fresh growth media and incubate at 37 °C, 5 % CO2 for 24 h.

Selection 13. Prepare selection medium by adding the proper concentration of Transduced Cells of selection antibiotic to the growth medium. The concentration of the antibiotic (puromycin, G418, hygromycin) is usually determined empirically. It is recommended that a selection antibiotic titration be performed to determine the lowest concentration of the selection antibiotic needed to efﬁ ciently select transduced cells and kill all non-transduced cells. 14. For LN428 cells, the ﬁ nal puromycin concentration is 1 μg/mL. 15. In the morning, remove the growth media and wash the cells in each well with 1 mL pre-warmed PBS. RNAi and DNA Repair 431

16. Add 300 μL trypsin and keep the plate in a 37 °C incubator for 5 min. 17. Add 1 mL selection media with the proper concentration of selection antibiotic to each well and pipette up and down to prepare a single cell suspension. 18. Transfer the cell suspension of each well to a freshly prepared 100 mm cell culture dish containing 9 mL of selection medium.

19. Incubate at 37 °C, 5 % CO2 until the cells reach 90 % conﬂ uence. Change medium with the selection medium every 3–4 days. 20. When the cells reach approximately 90 % conﬂ uent, remove the growth media and wash the cells in each dish with 3 mL pre-warmed PBS. 21. Add 1 mL trypsin and keep the plate in 37 °C incubator for 5 min and then add 4 mL selection medium and pipette up and down to make a single cell suspension. 22. Transfer 1 mL single cell suspension from each dish to a freshly prepared 100 mm dish with 9 mL selection medium. 23. Keep culturing the cells until the cells in the negative dish (mock transduction) without virus are 100 % dead. 24. Check the knockdown level via qRT-PCR (see below). 25. Freeze three vials of cells and passage the remaining cells for the designed experiment.

3.3 Validating Gene 1. This protocol is designed for use with the Applied Biosystems Knockdown by TaqMan® Gene Expression Cells-to-CT™ Kit (cat #AM1728). qRT-PCR Note: There are many different kits available to prepare cDNA. We 3.3.1 Cell Lysis prefer the Cells-to-C t Kit due to convenience. It is the fastest method and Reverse Transcription as it only takes about 1.5–2 h to go from cells to cDNA. However, you can choose any RNA extraction and RT kit . 2. Prepare cells for lysis (two options). 3. Option #1 : Prepare a cell pellet of 800,000 cells (pellet can be frozen at −80 °C or used immediately); resuspend cell pellet in 50 μL of PBS making a cell suspension of 16,000 cells/μL; distribute 5 μL of the cell suspension to a 1.5 mL tube and place on ice until continuing with the lysis. Note: You can use less than 800,000 cells if you do not have a lot of cells to work with. This number was chosen so you can see the cell pellet in the tube. If you use less, adjust the amount of PBS to resuspend the cells. 4. Option #2 : At least 24 h prior to lysis, seed enough cells into a well of a 96-well plate to yield around 80,000 cells/well at harvest (this will depend on the growth rate of your cell line); the following day, aspirate the media from each well, add 100 μL of PBS to wash, aspirate PBS and repeat to wash a second 432 Elise Fouquerel et al.

time, remove as much PBS as possible, and place on ice for immediate use or freeze at −80 °C to use at a later date. Note: The 96-well plate method is going to be easier if you are working with a lot of different cell lines. You will lyse directly in the plate so you do n’t have to worry about opening and closing a lot of tubes. If you use this method, DO NOT seed too many cell lines in the 96-well plate at one time. They will all have to be lysed at the same time and there are time points you have to hit. I usually do not seed more than 18–20 wells on each plate. Also, we usually prepare two or three replicates for each sample .

3.3.2 Cell Lysis 1. Prepare a master mix for the lysis consisting of 49.5 μL of lysis solution and 0.5 μL of DNase I (both supplied in the kit) for each sample. Add 0.5× excess to make sure you will not run out of the mix. Note: For example , if you have ten samples (10.5 × ) , prepare a master mix of 519.75 μL of lysis solution and 5.5 μL of DNase I. 2. Add 50 μL of the lysis master mix (prepared above) to each sample. 3. Mix the lysis reaction by pipetting up and down ﬁ ve times (avoid bubble formation). 4. Incubate the lysis reaction for 5 min at room temperature. Note: IMPORTANT! Do not allow reaction to exceed 5 min. 5. Add 5 μL of stop solution (supplied in the kit) to each sample. 6. Mix the reaction by pipetting up and down ﬁ ve times (avoid bubble formation). 7. Incubate at least 2 min at room temperature. Note: IMPORTANT! DO NOT allow reaction to remain at room temperature longer than 20 min; lysates can be stored on ice for up to 2 h or at − 20 °C for up to 5 months.

3.3.3 Reverse 1. Prepare a master mix for the RT reaction consisting of 25 μL Transcription of 2× RT buffer, 2.5 μL of 20× RT enzyme mix, and 12.5 μL of nuclease-free water (RT buffer and enzyme mix supplied in the kit) for each sample. Add 0.5× excess to make sure you will not run out of the mix. Mix the components gently, but thoroughly. Note: For example , if you have ten samples , prepare a 10.5 × master mix of 262.5 μL of 2 × RT buffer, 26.25 μL of 20 × RT enzyme

mix, and 131.25 μL of H 2 O. 2. Distribute 40 μL of the master mix prepared above to a nuclease-free tube for each reaction. 3. Add 10 μL of the sample lysate (or nuclease-free water as a control) to each tube for a ﬁ nal volume of 50 μL. RNAi and DNA Repair 433

4. Mix reactions gently, and then centrifuge brieﬂ y (assembled reactions can be stored at 4 °C for up to 4 h). 5. Run RT reaction (two options). 6. Option #1 : Distribute reactions into a PCR tube or multi-well plate and run on a thermocycler—step 1, 37 °C for 60 min; step 2, 95 °C for 5 min; and step 3, 4 °C for forever. 7. Option #2 : Distribute reactions into a 1.5 mL tube and run reaction in a 37 °C heat block for 60 min, immediately transferring to a 95 °C heat block for 5 min and then placing tubes on ice for at least 5 min. 8. Store reactions at −20 °C or immediately proceed with qRT-PCR analysis.

3.3.4 Quantitative 1. This protocol is designed for use with the Applied Biosystems RT-PCR StepOnePlus™ Real-Time PCR System, StepOne™ Software, and TaqMan® Gene Expression Reagents. It is important to thoroughly plan out your experiment before beginning. Note: The advantage to using the TaqMan ® Assays and Reagents vs. SYBR Green is time. There are hundreds of predesigned TaqMan ® Assays and each assay is designed for the same PCR conditions. This means there is no need for optimization on your part . 2. For each sample, you will run your gene(s) of interest plus one or more endogenous controls. You do not need to measure the concentrations of the cDNA before running the PCR. The endogenous controls will help account for differences in concentration between samples. When choosing your endogenous control, be sure to choose one that is not modifi ed by your experiments. 3. For each plate you run, you will need a “wild-type” control sample to compare to the rest of your samples. This is important. Your fi nal value is relative, based on your “wild-type” control. Your control will automatically be set to 1 (or 100 % expression), and all other samples will have an RQ (relative quantity) value compared to the control (e.g., sample 2 has an RQ value of 0.5; this means sample 2 has 50 %, or half, the expression of your control sample). 4. For each sample, it is recommended to run three reactions/ assay. The software will average these replicates together in your fi nal calculations. This will help eliminate pipetting errors across the wells. 5. The StepOnePlus™ system is designed to read up to 4 dyes/ well. This gives you the ability to add more than one TaqMan® assay/well. However, there is some optimization required with multiplexing. It is a great option if you have limited sample available and will save on reagents, but you have to optimize to ensure you are not getting false results. 434 Elise Fouquerel et al.

6. It is recommended to have an NTC (no template control) for each assay. We usually use water or you can use the water control from the Cells-to-CT reaction. This will allow the software to account for any background signal in the calculations. 7. Plan your plate layout and fi gure out how many wells you will be using for each assay. Then, prepare your master mixes as follows: (a) A “1×” master mix for each assay. (b) 10 μL of 2× TaqMan Universal PCR Mix. Note: Here , you can use either the Fast or Standard Mix. It is important to pay attention to which one you are using. The Fast Mix will allow you to run your entire reaction in about 40 min. The Standard takes about 2 h. They both will provide the same results; the Fast Mix is just more convenient . (c) 1 μL of 20× TaqMan Gene Expression Assay (some assays will come as a 40× stock; you must dilute these to 20× before use). (d) 5 μL of nuclease-free water. 8. When preparing your mix, add 0.5× excess so you will not run out of your master mix. 9. Keep master mixes on ice until ready for use. 10. Distribute 4 μL of cDNA from the Cells-to-CT™ reaction to each well. 11. Add 16 μL of your master mix to each well. Note: Between each of the two steps above, inspect the wells by eye to make sure each well has approximately the same amount of either cDNA or master mix added. It works best to add the cDNA fi rst, followed by the master mix. This way you can tell the difference in amounts of liquid in each well. It is diffi cult to see by eye the difference between 16 μL and 20 μL, but a lot easier to see the difference between 4 μL and 20 μL . 12. Cover the plate or tubes, mix gently, and centrifuge briefl y (mixed reactions can be stored at 4 °C for 1 h). 13. Place plate or tubes in the StepOne™ PCR machine, follow the prompts on the StepOne software to set up your experiment, and run the PCR.

3.3.5 Analysis 1. When analyzing data, ensure the three replicate wells have a similar CT (cycle time) values. 2. The AB software will calculate the RQ (relative quantity) value for you, taking into consideration the endogenous controls and sample replicates and comparing all your samples to the “wild-type” control. 3. Make sure the correct sample is selected as the “wild-type” control and the correct assay is selected as the endogenous RNAi and DNA Repair 435

Fig. 5 Relative mRNA expression as measured by qRT-PCR after shRNA to knock down several different genes involved in apoptosis signaling. Quantiﬁ cation of shRNA-mediated knockdown of the apoptosis genes CASP3, CASP9, BBC3, CASP8, BID, and DR4 as determined by qRT-PCR following lentiviral transduction and stable cell line development. TaqMan probes were used to quantify mRNA levels on an Applied Biosystems StepOnePlus machine. The qRT-PCR data was analyzed using the ΔΔCt method and was normalized to GFP-transduced control cells. Expression for each was normalized to the expression of human ß-actin. The mean of three independent experiments is plotted ± SEM

control in the experimental setup stage (these can be changed as many times as needed after the PCR has run). 4. Multiply the RQ values by 100 to ﬁ nd the percentage of mRNA expression in relation to your “wild-type” control (e.g., wild type has an RQ value of 1.0 = 100 % expression; sample 1 has an RQ value of 0.56 = 56 % expression). see Fig. 5 for an example.

3.4 Validating Gene The goal of any RNAi experiment is to reduce expression of the Knockdown by protein, preferably below detection or at least 75–90 % below the Immunoblot level of the control, such as we have described [10 , 15 , 17 , 21 ]. In this example, we show an example of loss of expression of the protein pRb, suggested to play a role in the cellular response to DNA damage [ 48 ]. As shown, protein lysates were prepared from cells transduced with a control shRNA (lane 1) and an shRNA speciﬁ c to pRb (lane 2), resulting in greater than 90 % loss of the expression of the protein (Fig. 6 ). The following describes the procedures required to validate gene knockdown by immunoblot although it is expected that most labs will have some general knowledge of the immunoblot 436 Elise Fouquerel et al.

Fig. 6 Demonstration of shRNA-mediated gene (pRb) knockdown by immunoblot. Depletion of pRb protein expression following lentiviral transduction and stable cell line development. Cell extracts from control and pRb-KD cell lines were resolved on a 4–20 % SDS-PAGE gel and immunoblotted for pRb. The blot was stripped and re-probed for tubulin, which was used as a loading control

procedure [49 ]. While precast SDS-PAGE gels are available (e.g., Bio- Rad, Life Technologies), SDS-PAGE gels prepared in the lab are inexpensive and easy to prepare. However, the reliability of the assay is of course dependent on the speciﬁ city and effectiveness of the antibody. 1. Assemble the spacer plate with a short plate in the casting frame and squeeze the system into the casting stand.

2. Prepare the 10 % separating gel mixing using 4.5 mL H2 O, 3 mL Acryl/Bis (30 %), and 2.5 mL of the separating buffer. Mix by inverting the tube two to three times and then add 68 μL APS and 14 μL TEMED. Mix and pour approximately 8 mL, leaving enough space for the stacking gel. Note: Alternatively and depending on the molecular weight of the protein of interest, a higher or lower concentration of Acryl/ Bis 30 % can be used. 3. Pour 200 μL of isopropanol carefully onto the gel solution. The polymerization can take greater than 10 min. 4. Remove the isopropanol by inverting the plates and wash with

now solidiﬁ ed gel with dH2 O.

5. Prepare the stacking gel by mixing 3.04 mL H2 O, 660 μL Acryl/Bis 30 %, 1.25 mL of the stacking buffer, 40 μL APS, and 8 μL of TEMED. Mix, pour, and insert the comb. Leave to polymerize approximately 10 min.

6. Gently remove the comb and wash the wells with dH2 O before inserting the gel into the electrophoresis unit. 7. Prepare the 1× running buffer and pour into the unit, avoiding bubbles under the plates to allow better protein migration. 8. Mix the cell extract and the 2× sample buffer at a 1:1 ratio and heat 10 min at 95 °C and load into the wells. Note: A 10-well gel can only contain up to 40 μL of sample per well . RNAi and DNA Repair 437

Fig. 7 Cartoon diagram demonstrating the preparation of the gel-membrane sandwich for transfer from SDS-PAGE to nitrocellulose in preparation for an immunoblot

9. Load 5 μL of the prestained protein molecular weight marker in one well. 10. Close the unit with the lead containing the power cables and connect to a power supply. 11. Run the gel at 90 V for 30 min, time enough for the samples to get through the stacking gel and then increase the voltage to 130 V for approximately 1.5 h or until the dye front reaches the bottom of the gel. 12. Carefully disassemble the plates using the gel releaser and soak the gel in a tray containing the 1× transfer buffer prechilled at 4 °C, two foam pads, two pieces of Whatman™ 3MM paper, and the nitrocellulose membrane cut to the size of the gel. 13. Set up the transfer sandwich in a gel holder cassette as shown in Fig. 7 : one foam pad, one piece of Whatman paper, the sheet of nitrocellulose membrane, and the gel. Remove any bubbles that might be trapped between the gel and the membrane by gently rolling a pipette on the gel. Then add the second piece of Whatman paper and the foam pad. Place the sandwich in the tray ensuring that the membrane is close to the anode (+) and the gel on the cathode side (−). Fill the tray with the cold 438 Elise Fouquerel et al.

transfer buffer and connect it to the power supply. The transfer is performed at 4 °C (or at RT with a cooling unit inserted in the tank) at 230 mA for 2 h. Note: Here , you can use either the Fast or Standard Mix. It is important to pay attention to which one you are using. The Fast Mix will allow you to run your entire reaction in about 40 min. The Standard takes about 2 h. They both will provide the same results; the Fast Mix is just more convenient . 14. At the end of the transfer, remove the membrane from the sandwich and verify the presence of the prestained molecular weight marker proteins. 15. Wash the membrane brieﬂ y in PBT buffer and incubate (30 min to 1 h) in TBT containing 5 % (w/v) Blotting-Grade Blocker, at room temperature, on a rocking platform. 16. Replace the blocking buffer with 10 mL blocking buffer containing the primary antibody diluted at the desired concentration. Incubate for 2 h at room temperature or overnight at 4 °C on a rocking platform. 17. Wash the membrane three times with PBT (10 min) at room temperature. 18. Incubate the membrane for 1 h in 10 mL blocking buffer containing the HRP-conjugated secondary antibody corresponding to the species of the primary antibody. 19. Wash the membrane three times with PBT (10 min each) at room temperature. 20. Place the membrane on a glass plate (proteins facing up) and add 1 mL of detection reagent freshly prepared by mixing an equal amount of each solution. Incubate in the dark for 2 min and detect with ﬁ lm or with the Bio-Rad Molecular Imager® Chemidoc™ XRS+.

3.5 PARP Activation PAR can be detected by immunoblot using one of several poly- in Response clonal or monoclonal antibodies. We routinely use the monoclo- to DNA- Damaging nal Ab 10H [ 15 , 17 , 22 ]. Care must be taken not only for the Agents analysis but also in the lysate preparation, as detailed below. No signal should be observed in samples isolated from an untreated 3.5.1 PAR Detection by cell, unless the cell is defective in homologous recombination [ 50 ] Immunoblot or defective in expression of the PAR-degrading enzyme PARG [ 22 ]. As shown (Fig. 8 ), PAR production is followed by loss of signal due to PAR hydrolysis mediated by PARG, as we have shown [ 22 ]. Note also that a more quantitative analysis using these anti-PAR antibodies is available by ELISA, as we have described elsewhere [ 17 ]. 1. Seed 105 to 5 × 10 5 cells in a 60 mm dish and incubate for 24 h. 2. When cells reach approximately 80 % conﬂ uence, treat the cells to trigger PAR production. The time of treatment may have to RNAi and DNA Repair 439

Fig. 8 PARP1 activation as measured by immunoblot. PAR production in LN428/ MPG cells or when depleted for PARG by shRNA (LN428/MPG/PARG-KD) detected by immunoblot with the 10H anti-PAR Ab after 1 h treatment with media or MMS (0.01 %)

be adjusted depending on the damaging agent, its concentration, and the cell sensitivity. 3. After treatment (note above that the treatment time varies depending on the genotoxin), wash cells two times with ice- cold PBS. 4. Remove excess PBS and add 60–100 μL of sample buffer (2×) directly to the plate. 5. Scrape the cells, collect the sample, and transfer to a 1.5 mL microcentrifuge tube. 6. Incubate for 10 min at 95 °C (boiling water bath). 7. Load 30 μL of the sample on an SDS-PAGE gel and transfer to the nitrocellulose membrane as described in Sect. 3.4. We have found that an overnight transfer at 0.08 mA (16 h) gives a better signal when probing for PAR. 8. Discard the blocking buffer and add TBT containing the monoclonal mouse anti-PAR antibody (clone 10H) diluted 1–1,000. Incubate 2 h at RT or overnight at 4 °C on a rocking platform. 9. Wash the membrane three times (10 min each) with PBT and incubate 1 h at RT with TBT containing the HRP-conjugated goat anti-mouse Ab diluted 1–5,000. 10. Detect signal as described in Sect. 3.4. An example of the expected result is shown in Fig. 8 . Here we show the accumulation of PAR after MMS treatment of LN428/MPG cells 440 Elise Fouquerel et al.

Fig. 9 PARP1 activation as measured by immunoﬂ uorescence. PAR production in cells after 0 or 15 min of MNNG treatment (5 μM). PAR is shown in cells exposed to MNNG (5 μM) using the anti-PAR 10H antibody (left ) and the nucleus is visualized using DAPI (right ) (color ﬁ gure online)

[15 ] that have been depleted of the PAR-degrading enzyme PARG, essentially as we have shown [22 ].

3.5.2 PAR Detection by The monoclonal Ab 10H can also be used to detect PAR in cells by Immunofl uorescence immunofl uorescence [51 ]. To perform this assay, cells must be cultured and grown on glass coverslips. This is readily accomplished by introducing a coverslip into the wells of a 6- or 12-well plate, depending on the number of cells needed. The diameter of the coverslip is chosen to fi t the wells of the plate selected. An example is shown in Fig. 9 , in which LN428/MPG cells [ 15 ] were treated with media or with media supplemented with MNNG (5 μM, 15 min). As indicated, no PAR is detectable in the untreated cells, but after treatment with the alkylating agent MNNG, the PAR signal is evident and coincides with nuclear staining by DAPI. 1. Sterilize coverslips in 70 % ethanol and let them dry on the edge of the wells. 2. Seed cells (around 5 × 104 cells/well for a 6-well plate) and incubate for 24 h at 37 °C. 3. Wash cells once with pre-warmed PBS and replace with fresh media containing the DNA-damaging agent. 4. Place the plate on ice and wash cells twice with ice-cold PBS. 5. Fix the cells by adding 3 mL of methanol-acetone (1:1) freshly prepared and stored at −20 °C. Incubate for 20 min on ice. RNAi and DNA Repair 441

6. Wash three times with PBS. 7. Permeabilize the cells by treating with PBS-Triton X-100 (0.1 %) for 15 min. 8. Wash three times with PBS followed by ﬁ ve washes with PBS- BSA (0.5 %). 9. Block with PBS-BSA (2 %) for 30–45 min. 10. Wash three times with PBS-BSA (0.5 %). 11. Dilute the monoclonal mouse anti-PAR antibody (clone 10H) to 1/500 in PBS-BSA (0.5 %) and gently drop 50–100 μL on each coverslip. Incubate for 1.5 h at room temperature or overnight at 4 °C in a box containing a wet tissue to keep the slides humid and to avoid dehydration. 12. Wash three times in PBS-BSA (0.5 %) and incubate for 1 h with 50–100 μL of secondary antibody (goat anti-mouse Alexa Fluor 488) at room temperature in the dark. 13. Wash two times in PBS-BSA (0.5 %) and three times in PBS. 14. Drop 25 μL of ProLong® Gold Antifade Reagent with DAPI on a microscope slide. Carefully pick up each coverslip with tweezers and dip in water to remove excess of salts from the PBS. Place the coverslip on the SuperGold with the face containing the cells down. Note: We use a ready-to-use mounting reagent containing DAPI to stain the nucleus. But it is also possible, prior to mounting, to incubate the cells for 20 min with 0.05 ng/μL DAPI (Sigma- Aldrich) in PBS or for 30 s with Hoechst (Sigma-Aldrich cat#

B-2883) diluted in dH 2 O at 1 mg/100 mL. The cells are then washed with PBS, rinsed in dH 2 O , and mounted with a drop of anti-fading reagent (Mowiol/Dabco or Gelvatol) . 15. Let dry overnight at room temperature in the dark (Fig. 9 ).

3.6 Complementing As described in Sect. 1 for this section, we routinely complement Gene KD Cells cells after gene KD with either the WT transgene or a mutant and Evaluating Cell transgene, depending on the focus of the investigation [ 15 , 17 , 21 , Survival in Response 22 , 52 ]. We use standard molecular biology procedures for the to DNA- development of these expression vectors and so we will not detail Damaging Agents these here. Importantly, as pointed out in Fig. 2 , care must be taken to ensure that the transgene is not a target of the transfected siRNA or the transduced shRNA. Evaluating cell survival in response to DNA-damaging agents after gene KD and/or after transgene complementation can be performed using transiently transfected cells, but unless the transfection efﬁ ciency is high or the measureable effect is strong, a stable gene KD is preferred for such assays. Examples of the results from either an MTS assay (Panel a ) or a CyQUANT assay (Panel b ) are shown in Fig. 10 . Here, we show the cellular response to the DNA alkylating agent MNNG. In the MTS assay, the cells are 442 Elise Fouquerel et al.

Fig. 10 Cell survival analysis in response to MNNG. (a ) Cell viability of LN428 cells expressing Flag-Polß(WT), Flag-Polß(K72A), or EGFP (as indicated) after MNNG treatment (3 μM), as measured by the MTS assay 48 h after exposure. Plots show the relative surviving fraction as compared to untreated (control) cells. Means are calculated from triplicate values in each experiment. Results indicate the mean ± S.D. of three independent experiments. (b ) Cell viability of T98G cells expressing Flag-Polß(WT) or EGFP (as indicated) after MNNG treatment (2 μM), as measured by the CyQUANT assay 10 days after exposure. T98G cells or those modiﬁ ed to express Flag-Polß(WT) or EGFP (control) were treated with MNNG and evaluated for cell viability. The plot shows the % viable cells as compared to untreated (control) cells. Results indicate the mean ± S.D. of three independent experiments

exposed to MNNG for 48 h and analyzed as detailed below. The LN428 cell line used here has reduced Polß levels, and as expected, complementation with the WT Flag-Polß transgene confers MNNG resistance (Fig. 10 , Panel a ). Note that complementation with the mutant (K72A) Flag-Polß transgene has reduced MNNG resistance, owing to its defect in the dRP lyase activity of Polß [15 ]. Similarly, the T98G cell line is sensitive to MNNG in a long- term assay as shown in Panel b (Fig. 10 ) in a Polß-dependant manner, as complementation with the WT transgene also confers resistance. The following provides details on each of these assays.

3.6.1 MTS Assay The MTS assay is performed as we have described previously [21 ]. Before starting the MTS assay, seed 1,000–10,000 cells/well into a 96-well plate and let the cells grow for 48 h. Then add MTS dye to the wells, incubate 3 h at 37 °C, and then measure absorbance at 490 nm. This step is to determine how many cells are needed to obtain the best result. In general, the seeded cell number depends on the duplication time and the size of the cells. You need to determine the number of cells that yield a robust signal but also yield a ½ signal strength when ½ of the cells were seeded (to ensure linearity). After you know how many cells are needed, then you RNAi and DNA Repair 443

can begin to prepare cells for the MTS assay. For LN428 cells, we suggest seeding 2,000 cells/well. 1. After your desired cells reach 70–80 % conﬂ uence, wash the cells with PBS, then trypsinize cells, and resuspend cells with media. 2. Take 100 μL of the cell suspension of the cells and count using a hemocytometer or an automated cell counter. 3. Dilute the cells and seed 200 μL of the required number of cells/well in 96-well plates from column 2 to 11 (eight rows per drug dose) and add media in column 1 and column 12. 4. Incubate seeded cells for 24 h at 37 °C. 5. Twenty-four (24) hours after seeding the cells, remove media from each well of the plate with a multi-well aspirator but leave medium in column 1. Note: This step of the MTS assay should be performed very carefully. Otherwise , you may remove cells and that will lead to unreliable data . 6. After mixing the medium with compound thoroughly, add 200 μL mixture with different concentrations of compound to each 8-well row using a multichannel pipettor. We prepare MNNG as a 10 mM stock solution in DMSO and store at −80 °C. Here, we dilute the 10 mM MNNG (10 mM) to 10 μM with media. In this example, we treated cells with 0 or 3 μM MNNG. 7. Incubate the treated cells for 48 h at 37 °C. 8. Remove the media very carefully and rinse each well with 200 μL fresh media. Note: This step of the MTS assay should be performed very carefully. Otherwise , you may remove cells and that will lead to unreliable data . 9. Prepare CellTiter Solution (Promega CellTiter 96 AQueous One, #G356B) as suggested by the manufacturer. 10. Place 120 μL of fresh medium mixed with CellTiter in each well including column 12 (control well) and incubate for 3 h at 37 °C. 11. Place microplate in the microplate reader to measure absorbance at 490 nm. 12. Calculate the result and plot the result as % control vs. compound concentration (Fig. 10 , Panel a ).

3.6.2 CyQUANT Assay The CyQUANT assay is performed as we have described previously [10 ]. Similar to the MTS assay, the seeding cell number depends on the duplication time and the size of the cells. For the 444 Elise Fouquerel et al.

CyQUANT assay, the cells need to be cultured for a minimum of six duplication cycles. As such, different cells may require different culture times. For example, for HCT116 cells, we incubate for 5 days after drug treatment, whereas for LN428 or T98G cells, we incubate for 10–12 days after drug treatment. Determining the proper cell number for seeding is very important for this assay. Usually we seed 50–400 cells/well and let the cells grow for 5–10 days to determine how many cells we need to seed per well to obtain the optimal result. For T98G or LN428 cells, we suggest to seed 50–70 cells/well; for HCT116 cells, we suggest to seed 150–200 cells/well. 1. After your cells reach 70–80 % confl uence, wash the cells with PBS, then trypsinize cells, and resuspend cells with 5–10 mL media and pass the cell suspension through 0.7 μm fi lter to obtain a single cell suspension. 2. Take 100 μL of the cell suspension of the cells and count using a hemocytometer or an automated cell counter. 3. Dilute the cells and seed 100 μL of the proper number of cells/well in a 96-well plate to columns 3–11 but seed half the number of cells in column 2. Add media only in column 1 and column 12. 4. Incubate seeded cells for 24 h at 37 °C. 5. Twenty-four (24) hours after seeding cells, remove media from the plate with a multi-well aspirator and leave media in column 1 and column 12. 6. After mixing the media with compound thoroughly, add 200 μL of the mixture with different concentrations of compound to the appropriate column using a multichannel pipettor. Before this step we need to calculate the volume of genotoxin, in this case MNNG, needed. We suggest to calculate the volume of the stock compound used and the volume of fi nal mixture needed. 7. Incubate the treated cells for six duplication cycles at 37 °C. 8. After six duplication cycles, remove the media and rinse each well with 200 μL PBS. Remove the PBS completely. Note: This step should be performed very carefully. Otherwise , you may remove cells and that will lead to unreliable data . 9. Seal edges of the plate with Parafi lm to prevent water leaving or entering plate. 10. Freeze plates at −80 °C overnight. 11. Prepare the lysis/dye solution. Remember this dye is light sensitive, so keep it in the dark and add to the plate in a dark environment. RNAi and DNA Repair 445

Table 4 Dilutions needed for the CyQUANT assay

Milli-Q Cell lysis CyQUANT Total cell No. wells/ No. water/ buffer/ dye/well Total Milli-Q lysis buffer Total CyQUANT plate plates well (mL) well (mL) (μL) water (mL) (mL) dye (μL)

80 4 0.1895 0.01 0.5 60.64 3.2 160 15 % overhead Milli-Q Total cell lysis Total added on at Water/well buffer for CyQUANT this step (mL) all cells dye for all (μL) cells (μL) 69.736 3.68 184

Note: CyQUANT reagent is light sensitive. Avoid light exposure for CyQUANT reagent. After lysis buffer with CyQUANT, dye is added to the plates ; remember to cover plates with aluminum foil . 12. Follow dilution shown in Table 4 to prepare lysis solution. 13. Remove the plates from the freezer and add 200 μL of cell lysis solution with dye to each well being tested. 14. Seal the edges of the plate with Paraﬁ lm and cover the plates with aluminum foil to keep the plate protected from the light. 15. Shake the plates for 1 h at room temperature and then incubate overnight at −30 °C. 16. Remove the plates from the freezer and thaw the plates slowly at room temperature. Remember to keep them from the light at all times. This process takes approximately 2 h. Note: Remember to wait for the solution to be thawed completely . 17. Read the plates on a ﬂ uorescent plate reader. 18. Calculate the result and plot result as % control vs. compound concentration (Fig. 10 , Panel b ).

Acknowledgments

This work was supported by grants from the National Institute of Health (NIH) (CA148629, GM087798, and GM099213) to R.W.S. Support for the UPCI Lentiviral Facility was provided by the Cancer Center Support Grant from the National Institutes of Health (CA047904). 446 Elise Fouquerel et al.

References

1. Dykxhoorn DM, Novina CD, Sharp PA (2003) 15. Tang J, Goellner EM, Wang XW et al (2010) Killing the messenger: short RNAs that silence Bioenergetic metabolites regulate base exci- gene expression. Nat Rev Mol Cell Biol sion repair-dependent cell death in response to 4:457–467 DNA damage. Mol Cancer Res 8:67–79 2. Mcmanus MT, Sharp PA (2002) Gene silenc- 16. Rubinson DA, Dillon CP, Kwiatkowski AV ing in mammals by small interfering RNAs. et al (2003) A lentivirus-based system to Nat Rev Genet 3:737–747 functionally silence genes in primary mam- 3. Paddison PJ, Hannon GJ (2002) RNA inter- malian cells, stem cells and transgenic mice ference: the new somatic cell genetics? Cancer by RNA interference. Nat Genet Cell 2:17–23 33:401–406 4. Schurmann N, Trabuco LG, Bender C et al 17. Goellner EM, Grimme B, Brown AR et al (2013) Molecular dissection of human (2011) Overcoming temozolomide resistance Argonaute proteins by DNA shuffl ing. Nat in glioblastoma via dual inhibition of NAD+ Struct Mol Biol 20:818–826 biosynthesis and base excision repair. Cancer 5. Caplen NJ, Parrish S, Imani F et al (2001) Res 71:2308–2317 Specifi c inhibition of gene expression by small 18. Mutamba JT, Svilar D, Prasongtanakij S et al double-stranded RNAs in invertebrate and ver- (2011) XRCC1 and base excision repair bal- tebrate systems. Proc Natl Acad Sci U S A ance in response to nitric oxide. DNA Repair 98:9742–9747 (Amst) 10:1282–1293 6. Elbashir SM, Harborth J, Lendeckel W et al 19. Heid CA, Stevens J, Livak KJ et al (1996) (2001) Duplexes of 21-nucleotide RNAs medi- Real time quantitative PCR. Genome Res ate RNA interference in cultured mammalian 6:986–994 cells. Nature 411:494–498 20. Semizarov D, Frost L, Sarthy A et al (2003) 7. Fire A, Xu S, Montgomery MK et al (1998) Specifi city of short interfering RNA deter- Potent and specifi c genetic interference by mined through gene expression signatures. double-stranded RNA in Caenorhabditis ele- Proc Natl Acad Sci U S A 100:6347–6352 gans. Nature 391:806–811 21. Trivedi RN, Wang XH, Jelezcova E et al (2008) 8. Izquierdo M (2005) Short interfering RNAs as Human methyl purine DNA glycosylase and a tool for cancer gene therapy. Cancer Gene DNA polymerase ß expression collectively pre- Ther 12:217–227 dict sensitivity to temozolomide. Mol 9. Ullu E, Djikeng A, Shi H et al (2002) RNA Pharmacol 74:505–516 interference: advances and questions. Philos 22. Tang JB, Svilar D, Trivedi RN et al (2011) Trans R Soc Lond B Biol Sci 357:65–70 N-methylpurine DNA glycosylase and DNA 10. Svilar D, Dyavaiah M, Brown AR et al (2012) polymerase beta modulate BER inhibitor Alkylation sensitivity screens reveal a conserved potentiation of glioma cells to temozolomide. cross-species functionome. Mol Cancer Res Neuro Oncol 13:471–486 10:1580–1596 23. Wang S, Sim TB, Kim YS et al (2004) Tools for 11. Brummelkamp TR, Bernards R, Agami R target identifi cation and validation. Curr Opin (2002) A system for stable expression of short Chem Biol 8:371–377 interfering RNAs in mammalian cells. Science 24. Friedberg EC, Walker GC, Siede W et al (2006) 296:550–553 DNA repair and mutagenesis, 2nd edn. ASM, 12. Paddison PJ, Caudy AA, Hannon GJ (2002) Washington, DC Stable suppression of gene expression by RNAi 25. Sirbu BM, Cortez D (2013) DNA damage in mammalian cells. Proc Natl Acad Sci U S A response: three levels of DNA repair regulation. 99:1443–1448 Cold Spring Harb Perspect Biol 5:a012724 13. Yu JY, Deruiter SL, Turner DL (2002) RNA 26. Harper JW, Elledge SJ (2007) The DNA interference by expression of short-interfering damage response: ten years after. Mol Cell RNAs and hairpin RNAs in mammalian cells. 28:739–745 Proc Natl Acad Sci U S A 99:6047–6052 27. Jackson SP, Bartek J (2009) The DNA-damage 14. Javanbakht H, Halwani R, Cen S et al (2003) response in human biology and disease. Nature The interaction between HIV-1 Gag and 461:1071–1078 human lysyl-tRNA synthetase during viral 28. Hassa PO, Haenni SS, Elser M et al (2006) assembly. J Biol Chem 278:27644–27651 Nuclear ADP-ribosylation reactions in RNAi and DNA Repair 447

mammalian cells: where are we today and 40. Sobol RW (2012) Genome instability caused where are we going? Microbiol Mol Biol Rev by a germline mutation in the human DNA 70:789–829 repair gene POLB. PLoS Genet 8:e1003086 29. Hottiger MO, Hassa PO, Luscher B et al 41. Alberts B (2009) Redefi ning cancer research. (2010) Toward a unifi ed nomenclature for Science 325:1319 mammalian ADP-ribosyltransferases. Trends 42. Berridge MV, Tan AS (1993) Characterization Biochem Sci 35:208–219 of the cellular reduction of 3-(4,5-dimethylthiazol- 30. Rulten SL, Fisher AE, Robert I et al (2011) 2-yl)-2,5-diphenyltetrazolium bromide (MTT): PARP-3 and APLF function together to accel- subcellular localization, substrate dependence, erate nonhomologous end-joining. Mol Cell and involvement of mitochondrial electron 41:33–45 transport in MTT reduction. Arch Biochem 31. Almeida KH, Sobol RW (2007) A unifi ed view Biophys 303:474–482 of base excision repair: lesion-dependent pro- 43. Kooistra SM, Helin K (2012) Molecular mecha- tein complexes regulated by post-translational nisms and potential functions of histone demeth- modifi cation. DNA Repair 6:695–711 ylases. Nat Rev Mol Cell Biol 13:297–311 32. Audebert M, Salles B, Calsou P (2004) 44. Jones LJ, Gray M, Yue ST et al (2001) Sensitive Involvement of poly(ADP-ribose) polymerase-1 determination of cell number using the and XRCC1/DNA ligase III in an alternative CyQUANT cell proliferation assay. J Immunol route for DNA double-strand breaks rejoining. Methods 254:85–98 J Biol Chem 279:55117–55126 45. Chi P, Allis CD, Wang GG (2010) Covalent 33. Audebert M, Salles B, Weinfeld M et al (2006) histone modifi cations—miswritten, misinter- Involvement of polynucleotide kinase in a preted and mis-erased in human cancers. Nat poly(ADP-ribose) polymerase-1-dependent Rev Cancer 10:457–469 DNA double-strand breaks rejoining pathway. 46. Goellner EM, Svilar D, Almeida KH et al (2012) J Mol Biol 356:257–265 Targeting DNA polymerase ß for therapeutic 34. Couto CA, Wang HY, Green JC et al (2011) intervention. Curr Mol Pharmacol 5:68–87 PARP regulates nonhomologous end joining 47. KeimLing M, Deniz M, Varga D et al (2012) through retention of Ku at double-strand The power of DNA double-strand break (DSB) breaks. J Cell Biol 194:367–375 repair testing to predict breast cancer suscepti- 35. Mansour WY, Rhein T, Dahm-Daphi J (2010) bility. FASEB J 26:2094–2104 The alternative end-joining pathway for repair 48. Shamma A, Takegami Y, Miki T et al (2009) of DNA double-strand breaks requires PARP1 Rb Regulates DNA damage response and cel- but is not dependent upon microhomologies. lular senescence through E2F-dependent sup- Nucleic Acids Res 38:6065–6077 pression of N-ras isoprenylation. Cancer Cell 36. Mitchell J, Smith GC, Curtin NJ (2009) 15:255–269 Poly(ADP-Ribose) polymerase-1 and DNA- 49. Towbin H, Staehelin T, Gordon J (1979) dependent protein kinase have equivalent roles Electrophoretic transfer of proteins from poly- in double strand break repair following ioniz- acrylamide gels to nitrocellulose sheets: proce- ing radiation. Int J Radiat Oncol Biol Phys dure and some applications. Proc Natl Acad Sci 75:1520–1527 U S A 76:4350–4354 37. Pears CJ, Couto CA, Wang HY et al (2012) 50. Mccabe N, Turner NC, Lord CJ et al (2006) The role of ADP-ribosylation in regulating Defi ciency in the repair of DNA damage by DNA double-strand break repair. Cell Cycle homologous recombination and sensitivity to 11:48–56 Poly(ADP-Ribose) polymerase inhibition. 38. Robert I, Dantzer F, Reina-San-Martin B Cancer Res 66:8109–8115 (2009) Parp1 facilitates alternative NHEJ, 51. Ame JC, Fouquerel E, Gauthier LR et al whereas Parp2 suppresses IgH/c-myc trans- (2009) Radiation-induced mitotic catastrophe locations during immunoglobulin class in PARG-defi cient cells. J Cell Sci 122: switch recombination. J Exp Med 206: 1990–2002 1047–1056 52. Jelezcova E, Trivedi RN, Wang XH et al (2010) 39. Wang M, Wu W, Wu W et al (2006) PARP-1 Parp1 activation in mouse embryonic fi bro- and Ku compete for repair of DNA double blasts promotes Pol beta-dependent cellular strand breaks by distinct NHEJ pathways. hypersensitivity to alkylation damage. Mutat Nucleic Acids Res 34:6170–6182 Res 686:57–67 Part III

Others Chapter 25

The LacZ Plasmid-Based Transgenic Mouse Model: An Integrative Approach to Study the Genotoxicity of Nanomaterials

Henriqueta Louro , Miguel Pinto , Nádia Vital , Ana M. Tavares , Pedro M. Costa , and Maria João Silva

Abstract

Numerous in vitro studies have been performed to address the potential genotoxicity of chemicals and of emerging products, e.g., nanomaterials. Although valuable for hazard assessment, the in vitro assays do not refl ect the complexity of an organism, including, bioavailability, toxicokinetics, and immune responses. Moreover, the biological effects at the target organs are known to be greatly infl uenced by factors as cell proliferative rate, metabolic, and DNA repair capacities. In this sense, data from suitable in vivo assays are useful to evaluate the performance of in vitro assays and to strengthen the knowledge about the genotoxicity of chemicals and nanomaterials, using several routes of exposure, at a whole organism level. This chapter provides an overview of an integrated experimental design, based on the use of a LacZ plasmid-based transgenic mouse model to investigate multiple genotoxicity endpoints in several organs, towards the safety evaluation of nanomaterials. This approach includes the analysis of chromosome instability, assessed by the micronucleus assay in blood or bone marrow cells and by sister chromatid exchanges in splenocytes, the analysis of DNA breaks and oxidative DNA damage by the comet assay, and the quantifi cation of gene mutations in multiple organs. Furthermore, histological markers e.g., of infl ammation and apoptosis, can add information about other relevant cell responses. A key point is that all assays are performed on the same animal, therefore increasing the effi ciency while reducing the cost and the number of animals under experimentation, in compliance with the EU recommendations. Overall, gathering the data from the several endpoints and organs of the same animal depicts the complex response of a whole-organism to nanomaterials exposure, thereby providing a better prediction of the effects on humans.

Key words In vivo genotoxicity testing , Comet assay , Micronucleus assay , LacZ mutations , Histopathology , Nanomaterials

1 Introduction

In vivo genotoxicity assays have been recognized as important tools to evaluate the carcinogenic potential of pharmaceuticals and environmental stressors, while accounting for the complexity of an

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7_25, © Springer Science+Business Media New York 2014 451 452 Henriqueta Louro et al.

organism. This comprises the variability among tissues and organs in physiological responses, e.g., infl ammatory response, immune response, metabolic and DNA repair capacities that can greatly infl uence the genotoxic effects in somatic and germinal cells [1 ]. On the other hand, in recent years there has been a considerable effort in order to replace and reduce animal experimentation for safety evaluation of chemicals and thus, the combination of genotoxicity assays (e.g., micronucleus and comet assays) in the same animal has been proposed as a valuable practical approach for in vivo genotoxicity testing [ 2 – 4 ] and will be further addressed in this chapter. Amongst in vivo assays, the transgenic rodent (TGR) mutation assays have been especially useful for hazard identifi cation and characterization, contributing to risk assessment of chemicals, physical agents, and nanomaterials (NMs; [5 , 6 ]). According to the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH), in vivo gene mutation assays can be used as follow-up tests to develop weight of evidence in assessing results of in vitro or in vivo assays [6 , 7 ]. This assay allows mutation analysis in any tissue during the lifetime of an animal, which constitutes an advantage over the use of in vivo selectable endogenous genes such as hypoxanthine-guanine phosphoribosyltransferase (Hprt ) or adenine phosphoribosyltransferase (Aprt ) genes, that are restricted to mutation analysis in a few cell types. An additional advantage of TGR models is that the transgenic loci are genetically neutral, which avoids the infl uence of selective pressures on in vivo mutations, allowing their accumulation and persistence. Consequently, the manifestation of mutations is less biased by the selection system and mainly dependent of the cell turnover and DNA repair mechanisms [1 ]. The LacZ plasmid-based transgenic mouse line derives from the C57Bl/6 mouse and harbors the pUR288 plasmid shuttle vector (containing the LacZ reporter gene) inserted in head-to-tail sequences in homozygosity on both chromosomes 3 and 4 [ 8 , 9 ]. The LacZ transgene has the double role of being a target for mutations and a reporter for their detection [ 1 ]. In this specifi c TGR model, the plasmid contains also the lacO sequence (to allow affi n- ity capture of the plasmid from genomic DNA using the lacI repressor protein) and the colE1 and amp R genes to allow propagation in Escherichia coli . Mutants are quantifi ed by the selective growth of plasmid-transformed E. coli and the mutant frequency is determined as the ratio between the number of mutant colonies (i.e., red colonies in selective medium) and the total number of colonies (i.e., blue colonies in nonselective medium). A further value of this system is the possibility of establishing the pattern of mutations for a given chemical, allowing the categorization of mutants into those carrying size-change (presumably deletions/ insertions) and no-size-change (presumably point mutations) An Integrated Approach for In Vivo Genotoxicity Testing 453 mutations after enzymatic digestion, and permitting the sequencing of point mutations to establish a complete mutation spectrum. Compared to other TGR models based on bacteriophage lambda shuttle vectors, this plasmid-based model is the only one that allows the identifi cation of large deletions, ranging in size from about 100 bp up to more than 5,000 bp, provided that the antibiotic resistance gene and the origin of replication are recovered [10 ]. Therefore, data from this in vivo TGR mutation assay can help to clarify specifi c pre-mutagenic adducts and mutational events that may be directly involved in the carcinogenic activity of chemicals and NMs [1 ]. The in vivo mammalian erythrocyte micronucleus (MN) test, which is included in the standard battery of genotoxicity assays after in vitro testing [ 11 ], is a reliable test for the analysis of chromosome damage, i.e., for the detection of clastogenic and aneugenic events [ 12 ]. In mice this test can be performed using either bone marrow or peripheral blood samples to score micronuclei in immature erythrocytes, i.e., polychromatic erythrocytes (PCE). The use of peripheral blood cells avoids the animals sacrifi ce and enables repeated sampling from the same animal [ 13 ], facilitating its integration in multi-endpoint toxicity studies [11 , 14 ]. Furthermore, the in vivo alkaline comet assay allows the identifi cation of a broad spectrum of DNA lesions and has been recommended for regulatory purposes [15 ], although its inclusion has not been formally accepted, yet. As a rapid, sensitive, and cost- effective method, this in vivo assay has the advantage of requiring a very low number of cells for analysis, from virtually any tissue, and being amenable for performance on blood cells at different timepoints along the exposure period. This assay is equally applicable to cells isolated from a frozen tissue [ 16 ], which facilitates the workfl ow when several endpoints have to be performed simultaneously, following animals sacrifi ce. With the recent introduction of enzymatic treatment steps, the comet assay can also be modifi ed in order to analyze specifi c types of DNA damage, particularly oxidative damage, using DNA repair endonucleases [17 , 18 ]. Recent works pointed to the combination of the acute in vivo MN and comet assays as representing a technically feasible and scientifi cally acceptable alternative to conducting independent assays, and indicated that the liver comet assay can complement the MN assay in blood or bone marrow in detecting in vivo genotoxins [2 – 4 ]. Another genotoxicity endpoint that can be relevant for some chemicals is mitotic recombination that can be analyzed through the sister chromatid exchanges (SCE) test in mouse splenocytes. The SCE assay detects reciprocal exchanges of DNA between two sister chromatids of a duplicating chromosome, which can be visualized after incorporation of 5-bromodeoxyuridine (BrdU) into chromosomal DNA for two cell cycles [ 19 ]. The use of the fl uorescence-plus-Giemsa staining procedure allows to 454 Henriqueta Louro et al.

differentiate between the chromatid having a single DNA strand substituted with BrdU and the other one having both strands substituted [20 ]. In addition to in vivo genotoxicity studies, histological analyses may provide critical and complementary information to differentiate cytotoxic/genotoxic events or to detect cellular responses such as infl ammation that may result in secondary genotoxicity. The main advantages lie in cost-effectiveness of standard protocols (most of which over a century-old) and the (micro)anatomical similarity between vertebrates, especially mammals, which render rodent models as appealing surrogates for human subjects. Detecting cell death in histological sections may also provide an important quality check if genotoxicity assessment, specifi cally through the Comet assay, is to be applied to solid tissue, especially from the liver, spleen or renal cortex, the latter two rich in haema- topoietic tissue [ 21 ]. This also gives an insight on secondary mechanisms that may underlie the genotoxicity of stressors, namely infl ammation and apoptosis/necrosis. In this chapter, an integrative approach that assembles a battery of genotoxicity tests, covering a range of endpoints in the LacZ plasmid-based transgenic mouse, is described. Such an approach has the important advantage of allowing a reduction of the number of animals necessary for an extensive genotoxicity and histological analysis, in compliance with the 3Rs agenda of the EU to Replace, Reduce, and Refi ne animal experimentation for scientifi c purposes [22 ]. The proposed experimental design can successfully be applied to assess the genotoxicity of chemicals and NMs and to provide additional information about their mode of action, while reducing costs and the quantity of test compound needed, and expediting data generation and interpretation. In the last years, the characterization of the genotoxic potential of NMs has been a challenging task, due to their specifi c physicochemical properties and dynamic behavior, which is refl ected by the numerous discrepancies in the data produced in in vitro and in vivo studies. To overcome these drawbacks, a careful characterization of the NMs physicochemical properties (e.g., dimension, surface area, reactivity, coating, presence of contaminants, etc.) previous to biological assays, as well as the use of standardized procedures for NMs dispersion and genotoxicity testing have been recommended [23 , 24 ]. Moreover, as pointed by Landsiedel et al. [ 25 ], given the complex nature of NMs dispersion in air or liquids (aerosols or suspensions) and the complex process of their uptake, deposition, and distribution in the body, in vivo genotoxicity studies have obvious advantages over in vitro studies. In a recent study we showed that a single concentration of a titanium dioxide NM was able to induce micronuclei in human lymphocytes [ 24 ]. As a follow-up of this in vitro equivocal result, the integrative approach herein described was applied, providing valuable information for safety assessment. An Integrated Approach for In Vivo Genotoxicity Testing 455

In the following sections the experimental strategy and the methods for the integrative study of the in vivo genotoxic effects of several classes of stressors are described, emphasizing the speciﬁ cities and some critical steps associated to NMs testing. Whenever possible, The Organisation for Economic Co-operation and Development (OECD) guidelines used for regulatory purposes are followed to warrant standardized procedures, although the present approach is targeted for investigation purposes only.

2 Experimental Design

A ﬂ ow chart of the integrative approach proposed for in vivo genotoxicity testing is presented in Fig. 1 , including the micronucleus assay in immature erythrocytes, comet assay, sister chromatid exchange analysis, gene mutation quantiﬁ cation, and histology. Although feasible, the combination of multiple endpoints in several tissues of the same animal is somewhat challenging with respect to the experimental design and considering the requirements for each test. A compromise is necessary for the most adequate experimental set up, namely, for selection of doses, concurrent

Fig. 1 Integrated approach proposed for assessing in vivo genotoxic effects of stressors and its application to the case of nanomaterials. Several samples derived from the same mouse can be used: (A ) Femur and spleen can be collected for steps (3) and (4), respectively, and should be processed immediately; (B ) for step (5), target organs are immediately ﬁ xed; (C ) for step (6), organs can be processed immediately or snap-frozen; and (D ) for step (7), target organs are snap-frozen in liquid nitrogen 456 Henriqueta Louro et al.

positive controls, sampling or euthanasia timepoints, target organs and tissues to be analyzed as well as their collection and processing for the different assays. Dose selection and route of administration : A minimum of three dose levels plus the negative and positive control groups are recommended [ 6 , 11 , 15 ]. The highest dose should be the maximum tolerated dose (MTD) and the other dose levels should cover a range from the MTD to no toxicity [ 6 ]. If no background information exists, a pilot study should be performed, using the same intended species, strain, gender, and treatment regimen [ 6 , 11 ] and, ideally, the pharmacokinetics and the cytotoxicity of the test compound should be determined prior to genotoxicity testing. The maximum volume of liquid that can be administered by injection (or gavage) can be a limiting factor for the top dose and depends mainly on the body weight of the test animal and on the characteristic of the test product; it should not exceed 2 mL/100 g body weight [ 6 ]. The exposure route should mimic the human exposure, whenever possible. Treatment schedule : The number of administrations and the duration of exposure should be scientifi cally justifi ed according to the aim of the study. Nevertheless, a repeated-dose administration is usually recommended; for an administration period of 28 days (i.e., 28 daily treatments), the top limit dose is 1,000 mg/kg body weight/day [6 ]. There are at least two study designs that have been successfully employed in a number of laboratories for the combination of the micronucleus and comet assays [2 ]. In the 3-day treatment protocol, animals are exposed at 0, 24, and 45 h followed by sacrifi ce at 48 h for collection of MN and comet assay tissues [3 , 4 ]. The other study design consists of four daily treatments of rats or mice, with sacrifi ce at 4 h after the last dose [ 26 ]. To produce a scenario of complete bioavailability to all organs studied, repeated dosing by the intravenous (iv) route can be acceptable. Nevertheless, depending on the tested chemicals or NMs, other administration routes are also appropriate, including oral administration, through food or drinking water, instillation or gavage. Note : To calculate the exact dose to be administered to each animal , the weight of each individually housed mouse should be fi rst registered. With respect to NMs administration by iv route , it has to be referred that the insoluble nanoparticles dispersed in an aqueous medium result in a thick dispersion that may be diffi cult to inject into the mouse tail veins. Thereby , it is advisable to inject the smallest possible volumes ( not exceeding 200 μL ), which might be a limiting factor for determining the top dose . Number and gender of experimental animals and controls : A minimum of fi ve healthy young adult animals per sex are randomly assigned to the control and treatment groups [6 , 11 ]; An Integrated Approach for In Vivo Genotoxicity Testing 457 concurrent positive and negative controls should be included in each assay, preferably using the same route of administration as the treated groups [6 , 11 , 15 ]. The negative control is the vehicle, (e.g., the NMs dispersion medium), whereas the positive control should ideally produce a positive result for all the endpoints analyzed. The selection of an adequate positive control for this integrative approach remains a challenge. A possible positive control is N -ethyl-N -nitrosourea (ENU), 120 mg/kg body weight, administered by intraperitoneal (ip) injection that proved to be an adequate positive control for the mutant frequency determination in mouse liver and testis [27 ], as well as for peripheral blood micronucleus assay and for comet assay in liver, even 28 days after injection. However, to date, the results in other organs cannot be predicted. Since administration of ENU at lower doses to male CD-1 mice resulted in a dose-dependent increase in sister chromatid exchanges (SCE) in bone marrow cells [ 28 ], it is expected that such treatment will be a good positive control for SCEs, as well. Another possibility is to use EMS at 150 mg/kg/day as reported by Bowen et al. [ 3 ] for the combination of the MN and comet assays. An adequate positive nanosized control for in vivo assays has yet to be found. Selection of sampling time and target organs for the genotoxicity endpoints : The selection of sampling times is critical and is generally driven by the time needed for the fi xation of DNA lesions into mutations that depends on the tissues’ proliferative rate. Thus, a compromise must be found to obtain the maximum information from all assays. Furthermore, it is desirable that the toxicokinetic data concerning the test compound or NM are also taken into account. A suitable sampling time for the measurement of mutant frequencies in both rapidly and slowly proliferating tissues is 3 days after the fi nal treatment, when using a 28 consecutive daily treatment [ 6 ] or 28 days after the last injection, when using iv or ip routes [10 ]. For the standard MN assay and if two or more treatments are used, it is recommended that peripheral blood samples are obtained at 36–48 h and bone marrow at 18–24 h after the last administration [ 11 ]. On the other hand, for the comet assay, sampling is recommended at 2–6 h or 22–26 h after the last administration [ 15 ]. Experience from our laboratory shows that, following a repeated iv administration of NMs, sampling at 28 days allows the analysis of gene mutations in several organs but might hinder the detection of primary DNA lesions by the comet assay, in the same organs. SCE analysis can be performed in splenocytes sampled at this time point. Another practical aspect is that analysis of two or more endpoints in the same tissue/organ may be hampered by the amount of tissue that is required for each one. For example, the Comet assay and the gene mutation assay can be performed in the same tissue because the former requires a very small sample. 458 Henriqueta Louro et al.

On the other hand, the whole spleen is necessary to obtain a good number of cells in metaphase for SCEs analysis, preventing its use for other assays. Tissue sampling and processing : The logistics of tissue sampling and processing is another practical aspect that should be carefully planned, according to each timepoint, each assay, and the tissue characteristics and subsequent utilization (e.g., for microscopy or for cell culture). Peripheral blood samples are collected from the tail vein to proceed for the comet and/or MN assay but the number of bleedings must be limited in order to minimize animals stress from repeated manipulations and potential disturbances to erythropoiesis. Following sacriﬁ ce, tissues are immediately removed and femur and spleen should be immediately processed for MN and SCEs, respectively. Tissues for histology must be instantly ﬁ xed, and afterward can be preserved for a long period of time until processing. Organs for mutation analysis are washed in cold saline, snap-frozen in liquid nitrogen, and stored at –80 °C. Finally, tissues for Comet assay can be immediately processed or snap- frozen. Further details of sampling and processing are given in the next sections that describe each of the genotoxicity assays integrated in this multi-endpoint approach.

3 Materials and Methods

3.1 Preparation The procedure described below is based on a method developed to of NMs obtain a stable dispersion of NMs within the EU Joint Action and Administration Nanogenotox [29 ]. Note: In addition to this dispersion protocol that was used in our Laboratory , other dispersion methods are also acceptable (e.g., method described in Chap. 14 of this book ).

3.1.1 Materials – Test NM – Bovine serum albumin (BSA), 1 % (w/v), sterile-ﬁ ltered; dilute with sterile water to obtain a BSA solution, 0.05 % (thereafter designated as dispersion medium) – Ethanol (≥96 %) – Concentrated phosphate buffer saline, sterile (PBS, 10×) – Glass scintillation vials (sterile) – Branson Soniﬁ er 400 Watt, S-450D (Branson Ultrasonics Corp., Danbury, CT, USA) equipped with a standard 13 mm disruptor horn (Model number: 101-147-037)

3.1.2 Methods 1. Weight the NM powder in a glass scintillation vial inside a ventilated weighing box, a glove box, or a fume hood, using a precision balance with an accuracy of at least 0.1 mg. For insolu ble An Integrated Approach for In Vivo Genotoxicity Testing 459

NMs (e.g., titanium dioxide, or multi-walled carbon nanotubes), prepare a stock dispersion of the NM by pre- wetting the weighted powder with ethanol (fi nal concentration 0.5 %) followed by the addition of an appropriate volume of sterile dispersion medium. 2. Place the vial in a container with ice-water (85–90 % ice plus 10–15 % cold water) and insert the sonicator’s probe between the upper quarter and upper half of the solution. Run the sonication for 16 min, at 400 W and 10 % amplitude. Note: Confi rm NMs behavior in solution / dispersion when using liquid medium and check the agglomerates formed (e.g., using DLS ). Try to set up the best procedure for the dispersion of the test NM. Use standardized methods , whenever possible . 3. Prepare the adequate dilutions of the stock dispersion (depending on treatment doses) in dispersion medium. Add concentrated PBS (10×) to obtain a fi nal dispersion containing 10 % PBS and use immediately for injection. 4. Prepare concurrent positive and vehicle control (dispersion medium). 5. Proceed to NMs administration through the selected exposure route according to the selected study design. 6. Following exposure, mice are maintained in adequate conditions until sacrifi ce, while weekly registering the body weight and any observations of discomfort or disease.

3.2 LacZ Mutation The procedure described below is based on methods initially devel- Analysis in Mouse oped by Gossen et al. [8 , 9 , 30 – 32 ] and adapted in previous Organs published work [10 , 27 , 33 ]. 3.2.1 Materials – Lysis buffer: Tris (10 mM), NaCl (150 mM), and EDTA (10 mM), pH 8.0 – TE-buffer: Tris (10 mM), EDTA (1 mM), pH 8.0

– Binding buffer 5×: Tris–HCl (50 mM), pH 7.5, MgCl2 (50 mM) and glycerol (25 %), pH 6.8; ﬁ lter sterilized – IPTG-elution buffer: Tris–HCl (10 mM), pH 7.5, EDTA (1 mM) and NaCl (125 mM); ﬁ lter sterilized – Potassium acetate (8 M) – Sodium dodecilsulfate (SDS) 10 % – Proteinase K (25 mg/mL in TE-buffer) – Kanamycin (50 mg/mL) – Phenol:chloroform:isoamyl alcohol (25:24:1) – Magnetic beads (Dynabeads M450 sheep antimouse IgG; Invitrogen) – Anti-β-galactosidase (2 mg/mL; Promega). 460 Henriqueta Louro et al.

– LacZ / LacI fusion protein (kindly provided by RIVM, The Netherlands, prepared in house) – PBS

– Glycerol (10 % in ddH2 O) – Isopropyl-β-d -galactopyranoside (IPTG) stock solution:

25 mg/mL in ddH2 O; ﬁ lter sterilized and stored at –20 °C

– ATP stock solution: 10 mM in ddH2 O; ﬁ lter sterilized solution and stored at –80 °C – HindIII enzyme – T4 DNA ligase (Gibco-Invitrogen) – Sodium acetate (3 M, pH 4.9); autoclaved – Glycogen (20 μg/μL, Gibco-Invitrogen) – Phenyl- β-d -galactoside (P-gal; Sigma-Aldrich). – 2,3,5-Triphenyl-2H -tetrazolium chloride – 5-Bromo-4-chloro-3-indolyl-β-d -galactoside (X-gal; Stratagene) at 50 mg X-gal/mL of dimethylformamide – Ampicilin (50 mg/mL) – Super Optimal Broth (SOB) medium – LB-topagar medium (6.125 g/L LB Broth Base; 6.125 g/L Antibiotic Medium #2) – Luria-Broth (LB) medium – E.coli C (Δ LacZ /galE - ) strain – Brinkman homogenizer – Magnetic particle concentrator (Dynal MPC-S) – Electroporation apparatus (Gen Pulser, Biorad, Hercules, CA) – Gallenkamp incubator shaker

3.2.2 Isolation 1. Place each organ sample in a 50 mL Falcon tube (e.g., 1/2 of of Genomic DNA the spleen, 1/3 of liver, 1 kidney, or 1 testicle). Add 9 mL from Mouse Organs of cold lysis buffer and homogenize rapidly with a Brinkman homogenizer. Add Proteinase K (0.5 mg/mL, ﬁ nal concentration), SDS (1 %), and RNAse A (120 μg/mL). Incubate overnight at 56 °C while rotating gently. Note: The LacZ mutation assay is a technically demanding assay and some sources of problems have been identiﬁ ed , especially associated with the transgene rescue procedure. Considering these challenges , it is recommended that , whenever possible , at least 2 / 3 of the target organs are snap -frozen to allow repetition , if needed . 2. Add an equal volume of phenol:chloroform:isoamyl alcohol (25:24:1), mix gently horizontally for 25 min and centrifuge An Integrated Approach for In Vivo Genotoxicity Testing 461

for 20 min at 3,200 × g in a bench or Beckman centrifuge. Remove the upper aqueous layer, containing the DNA, and transfer to a new 50 mL tube. Repeat the last step if the aqueous layer is still turbid. Add 1/5 of the total volume of potassium acetate (8 M) and mix gently. Add one volume of chloroform, mix horizontally for 25 min, and centrifuge for 20 min at 3,200 × g . Remove the upper aqueous layer, containing the DNA, and transfer to a new 50 mL tube. Repeat this step once more. 3. Precipitate the DNA by adding two volumes of ethanol (96 %). Mix gently and transfer the DNA to a microtube containing 1 mL of cold ethanol (70 %). Transfer the DNA to a clean microtube containing 0.5 mL of ethanol (70 %) and centrifuge shortly (1 min). Remove the ethanol from the DNA pellet. Add 400 μL of ethanol (70 %) and wash the DNA by gently inverting the microtube. Centrifuge shortly, remove the ethanol, and let it dry. 4. Dissolve in an appropriate volume (400–800 μL) of TE-buffer overnight. Check the quality and concentration of the DNA in

a 1 % agarose gel and by measuring the OD260 /OD280 ratio.

3.2.3 Pre-coating 1. Wash 1 mL of magnetic beads coated with sheep anti-mouse of Magnetic Beads IgG once with 1 mL PBS; precipitate the beads using the magnetic particle concentrator and discard the supernatant. 2. Resuspend the magnetic beads in 950 μL PBS, add 300 μg of anti-β-galactosidase, and incubate at 37 °C for 1 h in the hybridization oven. Pellet the beads on magnet and discard the supernatant. Wash the beads 3× in 1 mL PBS and resuspend the beads in 995 μL PBS. 3. Add 5 μL of LacZ / LacI fusion protein and incubate at 37 °C for 2 h in the hybridization oven. Wash the beads 3× in 1 mL PBS, resuspend in 1 mL PBS, and store at 4 °C (for a few days).

3.2.4 Preparation 1. Add 50 μL of an E.coli C (Δ LacZ /galE- ) glycerol stock [30 ] of Electrocompetent Cells and 5 μL of kanamycin (50 mg/mL) to 10 mL LB medium. Incubate overnight at 37 °C at 240 rpm in a Gallenkamp incubator shaker. 2. Add 1.5 mL of the overnight culture to each of the erlen- meyers with 500 mL LB (no kanamycin). Grow the cells to

an OD600 of 0.45 at 37 °C and 240 rpm. Place the erlen- meyers immediately on ice and shake horizontally for 30 min. Divide the cell cultures in 50 mL centrifuge bottles. Centrifuge (15 min at 4,000 × g , 4 °C) and discard the supernatant.

3. Resuspend each pellet in 25 mL ice-cold ddH 2 O, by gently shaking horizontally on ice. Centrifuge (15 min at 4,000 × g , 4 °C) and discard the supernatant. Repeat this step after pool- ing the tubes in pairs. 462 Henriqueta Louro et al.

4. Resuspend each pellet in 9 mL 10 % ice-cold glycerol, by gently shaking horizontally on ice. Pool the tubes together in pairs (two tubes remaining). Centrifuge (15 min at 4,000 × g , 4 °C) and discard the supernatant. 5. Resuspend each pellet in 2 mL 10 % ice-cold glycerol, by gently shaking horizontally on ice. Pool the tubes together (one ﬁ nal tube remaining).

6. Check the OD600 of cells (10 μL in 3 mL LB), which should be OD 600 = 0.19; if necessary, adjust the suspension density by diluting with 10 % glycerol. Aliquot the suspension (255 μL per cryovial) and freeze in a dry ice/ethanol bath. Store the competent bacteria at –80 °C.

3.2.5 Recovery 1. Incubate 50 μg of genomic DNA with 40 U of the restriction of the LacZ Transgene enzyme Hind III, 15 μL of binding buffer 5× and 60 μL of from Genomic DNA magnetic beads pre-coated with LacZ / LacI fusion protein for and Mutant Frequency 1 h at 37 °C, while rotating. Determination [30 ] 2. Use the magnetic particle concentrator to separate the DNA bounded to beads. Discard the supernatant ﬂ uid. Wash the beads using 250 μL of 1× binding buffer for three times. Resuspend in 75 μL IPTG-elution buffer, 5 μL IPTG stock

solution, 100 μL ddH2 O, and 20 μL NE Buffer #2 and incubate for 30 min at 37 °C, while rotating. After incubation, inactivate Hind III at 65 °C for 20 min. Spin to cool to room temperature. 3. Circularize the plasmid by adding ATP to a ﬁ nal concentration of 0.1 mM and 0.1 U of T4 DNA ligase. Vortex and incubate for 1 h at room temperature. 4. Pellet the beads in the magnetic particle concentrator, transfer the supernatant with the plasmid DNA to a clean microtube and precipitate by adding 30 μg of glycogen, 0.1 volume of 3 M sodium acetate, and 2.5 volumes of 95 % ethanol. Incubate for 1 h at –80 °C. 5. Spin down the DNA for 30 min at 14 000 × g and wash once with 250 μL of 70 % ethanol. Remove all the remaining ethanol using a ﬁ ne-tip and allow the DNA pellet to dry using a speed-vacuum concentrator for 5–10 min. Resuspend the

DNA in 5 μL ddH2 O. 6. Add 60 μL of the previously prepared highly competent E.coli C (Δ LacZ , galE- ) host cells and proceed with electroporation as described by the manufacturer: Gene Pulser: 25 mF and 1.8 kV; Pulse Controller: 200 W. Add 1 mL of SOB immediately after electroporation. Incubate cells in a Gallenkamp shaker for 30 min at 37 °C, 240 rpm. Meanwhile, prepare two An Integrated Approach for In Vivo Genotoxicity Testing 463

fl asks of Top Agar medium with 50 μg/mL 2,3,5-triphenyl- 2 H -tetrazolium chloride, 25 μg/mL kanamycin, and 150 μg/ mL ampicillin. 7. Plate 99.9 % of the transformed cells on a selective top agar plate containing 0.3 % of the lactose analogue P-gal. The remainder (0.1 %) is plated on nonselective medium containing X-gal (37.5 μg/mL) to determine the total number of colonies. Incubate plates 16 h, at 37 °C. Note: Use restriction analysis to check for recurrent deletion type mutants [34 ]. This may be due to unspecifi c HindIII star activity , low effi ciency in the recovery of the transgene , plasmid contami- nations , etc. [35 ]. Be sure to ascertain enzyme integrity , careful handling, and storing . To solve the problem of low rescue effi ciency leading to low colony number in X-gal plates : prepare new fresh electrocompetent cells ; check genomic DNA quality and concentration on a 1 % agarose gel .

3.2.6 Data Analysis Cell colonies are counted in X-gal (blue colonies, i.e., non-mutated and Interpretation plasmid with LacZ gene) and P-gal media (red colonies, i.e., mutated plasmid with LacZ gene). The mutant frequency (MF) for each organ is calculated as the ratio between the number of mutant colonies and the total number of colonies times the dilution factor, 1,000 [ 30 ]. MFs for the exposed groups can be compared with the negative control group using the Mann–Whitney U-test. The existence of a dose–response relationship based on the mutant frequency can be explored by regression analysis, in order to derive the mathematical function of the curve. From our experience, the negative (or vehicle) control group displays a mean MF of 7.6 × 10 –5 up to 11 × 10–5 in liver [27 ]. An appropriate positive control is ENU, 120 mg/kg body weight, which can cause an approximately 4-fold increase in the MF [ 27 ] while an ip injection of a dose of 30 mg/kg body weight of N -methyl-N -Nitrosourea induces a twofold increase in the MF in the same organ [ 33 ].

3.3 Comet Assay The outlined procedure is based on the method developed by in Mouse Tissues Azqueta and Collins [ 17 ] and following the previous recommen- and Organs dations of Hartmann et al. [ 36 ].

3.3.1 Materials – Mincing solution: EDTA (20 mM) and DMSO (10 %) in HBSS (Ca 2+ , Mg2+ free); adjust pH to 7.0–7.5 and store at room temperature (r.t.).

– Lysis buffer: NaCl (2.5 M), Na 2 EDTA·2H2 O (100 mM), Tris– HCl (10 mM) in dH 2 O, and adjust pH to 10 with NaOH 464 Henriqueta Louro et al.

(10 M). Store at r.t. Just before use, add Triton X-100 (1 %) and DMSO (10 %), and store at 4 °C in Coplin jars. – Neutralization buffer: Tris base (0.4 M) and HCl (0.4 M) in

dH 2 O; adjust pH to 7.5 and store at 4 °C.

– Electrophoresis buffer: NaOH (300 mM) and Na2 EDTA·2H2 O (1 mM) in dH2 O; store at 4 °C. – Buffer F for enzymatic reaction: HEPES (40 mM), KCl (100 mM), EDTA (0.5 mM), and BSA (0.2 mg/mL) in MiliQ water; adjust pH to 8 with KOH. Store at –20 °C. – Agarose-coated slides: immerse clean microscope slides in

a solution of 1 % normal melting point agarose in dH 2 O at 65 °C. Remove the excess of agarose from one side of the slide; allow solidifying horizontally and storing at r.t. – Agarose for cell incorporation: 0.8 % low melting point agarose (LMPA) in PBS. Store in 75 μL aliquots, at 4 °C. – Formamidopyrimidine-DNA-glycosylase (FPG) and endonuclease III (ENDO III) – Ethidium bromide (0.125 mg/mL)

Note: Further details on solutions preparation can be found in Chap. 14 .

3.3.2 Peripheral Blood 1. For blood sampling, immobilize the animal and pierce the tail and Organs Sampling blood vessel with a syringe needle (0.9 mm × 25 mm). Collect and Processing blood (approximately 5 μL) with a heparinized micropipette tip and add to 10 μL of mincing solution. Mix thoroughly and keep on ice until processing. 2. For tissues sampling, carefully remove the organ immediately after sacriﬁ ce and wash with sterile PBS. Add a small piece (e.g., 1/10 of the liver, 1/4 of the spleen, and 1/3 of the kidney) to 500 μL of mincing solution in a Petri dish and mince the tissue with the aid of sterile scissors, in order to mechanically release cells; enzymatic digestion can also be done using collagenase or trypsin [ 37 ]. Pipette up and down repeatedly to obtain a single cell suspension and keep on ice until processing for the comet assay. 3. To prepare two gels from each mouse, add 5 μL of cells suspension (or blood) to 75 μL of LMPA (at 37 °C), and homogenize. Spread two equally distributed 40 μL drops onto two separate microscope slides, and cover with a 20 mm × 20 mm coverslip to make the gels. Let slides to rest at 4 °C for at least 10 min, until solidiﬁ cation (two gels can be prepared per slide and the number of gels per organ depends on the repair enzymes that will be additionally used). An Integrated Approach for In Vivo Genotoxicity Testing 465

4. It is desirable to prepare extra slides from tissues sampled from an untreated animal to be employed as positive control slides

after immersion in a cold solution of H2 O2 (5 mM) for 5 min, followed by 2 min in cold PBS.

3.3.3 Comet 1. After gel solidiﬁ cation, carefully remove coverslips and immerse Assay Method slides into lysis solution for at least 1 h up to 16 h, at 4 °C (protected from light). 2. Wash slides three times, for 5 min, in Buffer F at 4 °C. For enzymatic treatment, place slides onto a metal plate on ice, in order to avoid premature enzymatic activity. Add 50 μL of enzymatic solution (FPG or ENDO III diluted in Buffer F) or Buffer F only to each slide, and cover with a 24 mm × 60 mm coverslip. Place slides into a humid chamber and incubate (30 min for FPG and 45 min for ENDO III, respectively) at 37 °C. 3. Remove coverslips, place slides into an electrophoresis tank (protect from light), and incubate in electrophoresis buffer for 40 min at 4 °C, to allow DNA unwinding. Proceed with electrophoresis for 30 min, 0.7 V/cm, at 4 °C. 4. Finally, neutralize the slides by washing with neutralization

buffer for 10 min at 4 °C, and with dH 2 O, for 10 min at 4 °C. Allow gels to dry overnight protected from light. 5. Before analysis, stain each gel with 18 μL of ethidium bromide (0.125 mg/mL) or another appropriate ﬂ uorescent dye, cover with a coverslip, and keep the slides in a humidiﬁ ed chamber at 4 °C, protected from light.

3.3.4 Data Analysis Slides are coded and blind scored under a fl uorescence micro- and Interpretation scope equipped with the appropriate fi lter, at a ×200 magnifi cation. Software is available for rapid measurement of several comet parameters (e.g., Comet Imager, MetaSystems, GmbH). The quantifi cation of the percentage of DNA in tail, tail length, and olive tail moment of 100 randomly selected nucleoids per animal (50 scored per gel), and per enzymatic treatment is generally recommended [ 17 , 21 ]. The percentage of the tail DNA is the preferred measurement and the mean (or median) of the exposed groups can be compared to that of the vehicle control group using a general linear model approach, such as analysis of variance (ANOVA) followed by pairwise comparisons test. In case the data do not follow a normal distribution and/or the homogeneity of variances is not verifi ed, nonparametric tests can be alternatively applied. In addition, regression analysis is also used to fi nd the mathematical model that best describes the dose–response relation ship. The same analyses also apply to the determination of 466 Henriqueta Louro et al.

oxidative DNA damage, or FPG and ENDO III sensitive-sites. A signifi cant increase in tail DNA is indicative of primary DNA damage, and an increase in FPG or ENDO III sensitive-sites is indicative of oxidative DNA damage. Note : When high levels of DNA damage are detected in the negative control, ensure the protection of slides from direct exposure to light and check Buffer F preparation. If tissue samples are maintained on ice for an extended period after collection , elevated damage can be observed. A low cell number can result when relying on a mild mechanic disaggregation of tissues and that can be overcome by using a more drastic procedure or proteolytic enzymes. In data interpretation confounding factors (e.g., cytotoxicity , long period between exposure and sampling ) that may have infl uenced the level of DNA damage should be taken into account. For example , if mice are acutely exposed to the genotoxin by iv administration and sacrifi ced 28 days after exposure , a negative result in the comet assay is most probably due to the repair of the primary lesions ( or to cell death ) that prevents its late detection .

3.4 Micronucleus The protocols described below are adapted from published meth- Assay in Mouse Blood ods [13 , 14 , 38 ] and follow the general recommendations of and Bone Marrow OECD guideline 474 [11 ].

3.4.1 Materials – May-Grünwald solution, M-G (Merck, Cat.# 1.01424.0500) – Giemsa (Merck, Cat.# 1.09204.0500) – Phosphate buffer, pH 6.8 – Acridine Orange (AO)-coated slides. Slides are prepared by preheating clean glass slides at 70 °C and spreading 10 μL of 1 mg/mL AO solution using a glass rod [ 38 ]. The AO-coated slides are air-dried and can be stored at r.t., protected from light.

3.4.2 Micronucleus 1. Follow the procedure described above (see Sect. 5.2 ) to collect Assay in Peripheral Blood two blood samples (3–5 μL); spread each drop onto an AO-coated slide and cover with a coverslip. Note : Spread blood on slide with a uniform movement. Also , gently pressing coverslip permits a better spreading of blood cells . 2. Store the slides into a humidifi ed chamber protected from light, at 4 °C, until analysis within 5 days. Slides are analyzed under a fl uorescence microscope equipped with appropriate fi lters, at ×1,000 magnifi cation (Fig. 2a ). Note: If the intensity of the fl uorescence is excessive , allow a longer settling period at 4 °C , in order to reduce fl uorescence . An Integrated Approach for In Vivo Genotoxicity Testing 467

Fig. 2 Microphotographs illustrating the micronucleus assay: (a ) mouse peripheral blood erythrocytes stained with AO, allowing the distinction of polychromatic (PCE) from normochromatic (NCE) erythrocytes [14 , 29 ]; micronuclei are easily viewed as green ﬂ uorescent dots within the cytoplasm of PCE; (b ) Mouse bone marrow preparation stained with May-Grünwald/Giemsa; the analysis is restricted to PCE [11 ] that are typically bluish colored and larger than NCE (orange–pink stained)

3.4.3 Micronucleus 1. After mouse sacrifi ce, remove the femur, cut off both ends, and Assay in Mouse fl ush out the bone marrow with 1 mL of fetal bovine serum Bone Marrow (FBS). 2. Centrifuge the cell suspension at 136 × g for 5 min and discard the supernatant. Resuspend the cell pellet and make smears on clean microscope slides. It is advisable to prepare at least three slides per animal, two to be immediately stained and analyzed and the third one to be kept as a backup. 3. Slides are immediately fi xed by immersion in cold methanol, during 5 min. After air-dried, they are stained with the M-G solution for 2.5 min followed by a diluted M-G (1:2 in phosphate buffer) for 2 min. 4. Wash generously with phosphate buffer (pH 6.8) and stain with Giemsa (4 % in phosphate buffer) for 8 min. 5. Finally, wash slides with tap water, allow them to air-dry, and mount with Entellan. 6. Slides are analyzed under a bright fi eld microscope, at a ×1,000 magnifi cation (Fig. 2b ). Note: The glass slides may have a special coating that cause the formation of dye micelles or dyes may have particles that form artifacts over the preparation. Change slide brand , pre -clean the slides with ethanol – ether ( 3 :1 ), and fi lter dyes before use. Staining should be fi rst tested on two or three slides to optimize its quality , which is essential for a reliable scoring of MN. A good staining method must ensure a good differentiation of PCEs and NCEs. If staining is too dark , a quick rinse with ethanol 70 % after staining followed by tap water immersion can improve the differentiation ; the duration of Giemsa staining should be also optimized . 468 Henriqueta Louro et al.

3.4.4 Data Analysis For each animal, micronucleated PCEs are blind score in at least and Interpretation 2,000 PCEs [11 ] from two previously coded slides. In addition, the number of PCEs and NCEs is determined by scoring 1,000 erythrocytes to calculate the ratio between PCEs/ NCEs as a measure of bone marrow toxicity. For each dose group, the frequency of micronucleated PCEs can be compared with that of the vehicle control group using the Chi-Square test (or Fisher’s exact test). A signifi cant increase in the MN frequency in the group of exposed mice as compared to control animals indicates a genotoxic effect. The existence of a dose–response relationship can be investigated using regression analysis. A signifi cant decrease of the PCEs/NCEs ratio in the exposed group is indicative of hematho- poiesis reduction, i.e., mielotoxicity. Our historical control data has shown that the basal frequency of micronucleated PCE in the LacZ transgenic mouse ranges from 7.4 up to 8.4 ± 3.1 [27 , 39 ], which are higher than those usually reported for nontransgenic mouse strains [40 ]. Ethylnitrosourea (120 mg/kg) is able to induce a signifi cant 6-fold increase in the frequency of MN, 48 h after ip administration, being an adequate positive control [ 27 , 39 ].

3.5 SCE Assay – 5-Bromodeoxyuridine (5 μM) in Splenocytes – Colcemid (0.2 μg/mL) 3.5.1 Materials – Culture medium: RPMI 1640 medium (Invitrogen) supplemented with 20 % heat inactivated fetal calf serum (FCSi, Invitrogen) and 2 μg/mL of concanavalin A. – Washing medium: HBSS with 2 % FCSi. – Hypotonic solution: KCl (0.075 M) plus 1 % sodium citrate (3:1) – Fixative: acetic acid:methanol (1:3)

3.5.2 Splenocyte 1. Remove the spleen and place it in a Petri dish containing 5 mL Isolation and Chromosome of sterile washing medium. Release splenocytes to the medium Preparation [19 ] by gently pressing the spleen after cutting off the two extremi- ties of the spleen; resuspend the cells by repeated pipetting. 2. Centrifuge the cells (400 × g for 10 min) and resuspend the pellet in 10 mL of culture medium. For each spleen, initiate two 5 mL cultures in complete medium and incubate for 24 h, 37 °C. 3. Add BrdU (5 μM) and re-incubate the cultures (protected from light) for 30 h more, at 37 °C. Add colcemid (0.2 μg/mL) 2 h before harvesting, to block cells’ division in metaphase. 4. For chromosome preparations, harvest cells by centrifugation (400 × g for 10 min) and resuspend the pellet in 5 mL of the pre-warmed hypotonic solution. After 18 min at 37 °C, centrifuge and recover the pellet in 5 mL of cold ﬁ xative; store at 4 °C for at least 30 min, and spread cells onto microscope slides. An Integrated Approach for In Vivo Genotoxicity Testing 469

5. The differential staining of the sister chromatids is obtained through the ﬂ uorescence-plus-Giemsa standard method that consists of sequential steps of staining with the ﬂ uorescent dye bisbenzymide H33342, UV irradiation and Giemsa staining [20 ].

3.5.3 Data Analysis Following two rounds of replication in the presence of BrdU and Interpretation (second division metaphases) all chromosomes are constituted by a BrdU-monosubstituted chromatid that stains dark blue and a chromatid with two substituted strands that appears lightly stained (Fig. 3 ). SCEs are scored exclusively in second division metaphases that are recognized as those displaying all chromosomes with dif- ferentially stained chromatids. Each chromatid exchange, seen as a reciprocal exchange between the sister (dark and light) chromatids at apparently homologous loci , is counted as a single event. For each animal, 30–50 s division metaphases are blind scored for SCEs in two different replicate cultures. To characterize cell cycle kinetics, at least 200 metaphases per animal are classiﬁ ed into 1st (M1), 2nd (M2), and 3rd (M3) cell cycle divisions, in the same slides. In ﬁ rst division metaphases sister chromatids are undistinguishable whereas in third division metaphases only some chromosomes present differentiated sister chromatids. The mean SCE counts per cell are calculated for each animal and for each group of treated or untreated animals. The comparison of the SCE frequencies between the treatment and the vehicle control group is performed by nonparametric tests, e.g., the Kruskal–Wallis or the Mann– Whitney U-test. Chi-square test is applied to analyze the data from high frequency cells (HFCs), estimated as the proportion of cells

Fig. 3 Metaphase from LacZ transgenic mouse splenocytes stained by the ﬂ uorescence-plus-Giemsa method for sister chromatid exchange analysis 470 Henriqueta Louro et al.

lying above the 95th percentile of negative controls [ 41 ]. For cell cycle kinetics analysis, the replication index (RI) is calculated as [(M1 + 2(M2) + 3(M3))/n ], where n is the number of cells scored.

3.6 Histological – Alcoholic eosin Y (Sigma) Analysis of Mouse – Haematoxylin, Harris (Sigma) Tissues – Bouin’s Fixative (10 % formalin, 7 % acetic acid, in distilled or 3.6.1 Materials ultrapure water, with picric acid added to saturation) – Parafﬁ n, histological grade (58–60 °C) – Microscopy slides, pre-coated (e.g., Polysine, Thermo Scientiﬁ c), and coverslips (24 × 50 mm are preferred) – DPX mounting medium

3.6.2 Preparation 1. A detailed histological protocol is provided in Table 1 . Harvest of Histological Samples fresh tissue samples (5 mm-wide samples are adequate for most histological procedures) and immerse them immediately (without washing) into the fi xative solution for at least 24 h, up to 48 h. The choice of the most adequate fi xative for histological analysis may be tricky and dependent of the objective of the study, organ, and organism. Formalin–based fi xatives such as buffered formalin, Davidson’s and Bouin’s are good choices for general-purpose histology. Cross-linking fi xatives such as 1–2 % glutaraldehyde and/or paraformaldehyde (buffered) are preferred for electron microscopy and may be used for histology as well. Following fi xation, the samples are washed in water to remove the excessive fi xative and dehydrated in a progressive series of ethanol. The samples may be safely archived in 70 % ethanol. The samples are intermediately infi ltrated with several immersions in xylene or toluene and embedded in molten paraffi n, overnight or up to a few days. The choice of the best embedding agent relies mostly on the type of tissue: specimens that are either too soft or too hard may yield better slides if embedded in resins (such as epoxies). Alternatively, cryosections may be obtained from chemically protected samples of fresh tissue which implies considerable differences in protocol and equipment. Note: All samples should be absolutely fresh and undamaged by micro- dissection and handling. Fixation should be immediately and adequately performed. Careful dehydration is important to maintain tissue integrity. If tissue appears hard and brittle , the fi xation time might have exceeded. Also , the concentration of formalin should not exceed 10 % of the fi nal solution . 2. The paraffi n blocks are obtained by pouring molten paraffi n in commercial molds or Leuckart bars, then placing the sample (sectioning surface downwards) into the mold and allowing An Integrated Approach for In Vivo Genotoxicity Testing 471

Table 1 Basic protocol for histology

Step Substance/solution Duration Temperature

Pre - sectioning Fixation Bouin’s 24–48 h 4 °C Washing Distilled water 3 × 2 h Room Dehydration 70 % ethanol 1 × 30 min Room 95 % ethanol 2 × 15 min Room 100 % ethanol 3 × 30 min Room Intermediate infi ltration Xylene or toluene 3 × 15 min Room Embedding Molten paraffi n Overnight 58–60 °C Post-sectioning Deparaffi nation Xylene or toluene 2 × 30 s Room Rehydration 100 % ethanol 2 × 30 s Room 95 % ethanol 2 × 30 s Room 70 % ethanol 1 × 30 s Room Distilled water 1 × 6 min Room Staining Stain Harris’ haematoxylin 2–4 min Room Blueing Tap water or substitute 2 × 2 min Room Counterstain 70 % ethanol 1 × 3 min Room Alcoholic eosin Y 1–2 min Room Dehydrating 70 % ethanol Brief rinse Room 95 % ethanol Brief rinse Room 100 % ethanol 2 × 30 s Room Clearing Xylene or toluene 1 × 30 s Room Mounting DPX or similar

the paraffi n to solidify, fi rst at room temperature, then at 4 °C if fast solidifi cation is desired. Work fast when adjusting the sample to its proper position in the mold with molten paraffi n. 3. Trim the block to obtain a trapezoidal sectioning surface, then section the paraffi n blocks in a rotary microtome (manual or automatic). Sections ≈5 μm thick are ideal for most histological procedures. Pick up the strips with the sections and allow them to stretch in a warm water bath. Collect the strips on pre- coated slides. Every slide should contain at least eight serial sections, depending on specimen size (typically, two strips with 472 Henriqueta Louro et al.

Fig. 4 Parafﬁ n-embedded histological section of the liver of a LacZ transgenic mouse

exposed to TiO2 NMs revealing an infl ammatory focus consisting of macrophages intruding into the hepatic parenchyma (arrow ), adjacently to a hepatic venule (vn). Nanoparticle aggregates (npa) are clearly visible inside the defence cells and remaining hepatocytes within the focus. Inset : dense nanoparticle aggregates phagocytized by macrophages infi ltrating between hepatocytes. Sample fi xed in 10 % buffered formalin, embedded in paraffi n, sectioned at 5 μm, and stained with H&E

6–8 sections should be easily obtained). At least two slides per sample should be prepared. 4. Allow the slides to dry completely (a few hours in a fume hood at room temperature should suffi ce). Before staining, the sections are deparaffi nated with xylene or toluene and brought to water with a regressive series of ethanol, fi nalized in distilled or ultrapure water. Stain the sections with Harris’ haematoxylin (an alum haematoxylin) and counterstain with alcoholic eosin Y (H&E staining). The slides are then dehydrated, cleared with xylene or toluene, allowed to dry, and mounted in a resinous medium, such as DPX. DPX requires several days to solidify completely, in a fume hood, but the slides may be workable in a few hours or less. Protect the stained slides from continuous direct light to preserve the stain.

3.6.3 Microscopy The primary histopathological alterations to look for should be Analysis NMs bioaccumulation and infl ammation, which is a nonspecifi c immune response that has been shown to occur in several tissues, as a result of exposure to NMs [ 42 ]. This relatively complex response can be chiefl y identifi ed histologically by abnormal blood affl ux to the affected area (hyperaemia) and defense cell infi ltration (Fig. 4 ). Localized hemorrhage may occur in hyperaemic foci, especially in highly damaged tissues, as well as oedemas. Still, diffuse postmortem blood stasis is a common artifact that may be confused with hyperhaemia, which compromises diagnostics. An Integrated Approach for In Vivo Genotoxicity Testing 473

Adequate identifi cation of histopathological lesions and alterations is paramount. Good histopathological appraisals involve careful preparation of samples and, critically important, an expert interpretation of the fi ndings. Terms are often misemployed or artifacts confounded with actual changes. Terms such as tissue “disarrangement” should be avoided for being ambiguous and potentially revealing of ill-prepared samples. As examples, the following terms are often confused, even though they stand for different concepts: “atrophy,” which stands for loss-of-form; “regression,” which means loss-of-function and “degeneration,” which implies change in function. These examples state the importance of correct identifi cation of histopathological traits, adequate preparation of slides, and the presentation of high-quality micrographs, since these are the primary quality checks of the work. Identifi cation of cell death is an important endpoint. While focal necrosis and apoptosis can be identifi ed by nuclear pleomor- phisms (such as pyknosis) and abnormal staining of the cells, during early stages of cell death, necrosis is commonly accompanied by infl ammation, unlike apoptosis. Apoptosis may be identifi ed by other features, such as the “blebbing” of the cell membrane. Tissue damage during harvesting and sample preparation may result in false positives for cell death.

4 An Integrated Analysis and Interpretation of Data from Multiple Endpoints

Standard statistical analysis methods with predetermined criteria (e.g., 95 % confi dence levels) and the appropriate adjustments for data distribution and variability should be used to determine the statistical signifi cance of the response from all endpoints analyzed [4 ]. In addition, defi ned positive response criteria should be also clearly assigned to each endpoint, taking into account the specifi c- ity of each tissue and test compound. Criteria for a positive result generally includes a dose-related increase in the genotoxicity endpoint and/or a statistically signifi cant increase observed in treated groups comparatively to the vehicle control group [6 , 11 ]; data from this control group should be within the historical values of the laboratory [ 15 ]. However, it is advisable to consider the possibility that a nonlinear or nonmonotonic dose–response relationship may exist, for example, when testing NMs. To assemble all data obtained in each treatment group, a table can be created. An example of a hypothetical table for the assessment of the global effects of a NM is presented (Table 2 ). The proposed table provides an overview of the data per endpoint that, taking into account its biological relevance, is aimed at facilitating an integrative interpretation of the in vivo data. In addition, the Pearson or Spearman correlation coeffi cients can 474 Henriqueta Louro et al.

Table 2 Example of a model table to use for an integrative view of the results

Organ/tissue Assay/endpoint Outcomea

Peripheral blood Micronucleus Comet assay Modifi ed comet assay Bone marrow Micronucleus Comet assay Modifi ed comet assay Liver Presence of NMs Comet assay Modifi ed comet assay Mutation Infl ammation Spleen SCEs Mutation Infl ammation a Positive, negative, or equivocal, according to the criteria previously defi ned for each endpoint

be determined to analyze the association between the different endpoints analyzed in the same mouse, providing further clues on possible mechanisms underlying the observations. Thus, using a reduced number of animals and combining standard procedures for genotoxicity characterization, the integrative approach herein presented is a feasible, sensitive, and valuable tool contributing to the identiﬁ cation of genotoxic agents, including nanomaterials, across a wide range of target organs within the same animals.

5 Future Prospects

New technologies, improvement of classical assays, and new applications are emerging in the ﬁ eld of genotoxicity testing. In this context, suggestions are made in order to potentially improve future studies using the described integrative approach in this mouse model. Namely, if a change in the mutant frequency is noted after exposure to one nanomaterial, elucidating the induced mutation spectra by sequencing may help clarifying the mechanisms of action of the NM, similarly to what has been carried out for chemicals [ 10 ]. However, a complete sequencing of the LacZ gene requires ﬁ ve sequencing reactions when using the classical automated sequencing. Noteworthy, the introduction of Next Generation Sequencing may contribute to a faster and more cost- effective establishment of a mutation spectrum. An Integrated Approach for In Vivo Genotoxicity Testing 475

Additionally, with recent modifi cations and improvements of the Comet assay technique, other endpoints could also be considered in in vivo studies of nanomaterials. The introduction and optimization of the DNA repair Comet assay, for example, reveals other potential endpoints that can be explored in an integrated approach, since it allows the analysis of the interference of a test compound with the DNA repair capacities of the exposed animals, either by measuring base excision repair or nucleotide excision repair activities [43 ]. Given that it enables the use of frozen tissue samples and that it can be potentially performed in multiple organs, it could potentially be used to analyze new endpoints in a single set of exposed animals [ 43 ]. Moreover, the recently developed high- throughput comet assay [ 44 , 45 ] can probably be used for in vivo experiments, which will signifi cantly reduce slide number, blood and tissue sample size, and the reagents needed. In addition, it would allow the simultaneous processing of a larger number of individuals and could permit the application of a higher number of enzymatic treatments in a single experiment, which reduces results variability. For the MN assay, with the introduction of the automated fl ow cytometry scoring of peripheral blood it is possible to perform a quick and reliable determination of genotoxic effects, reducing inherent intra- and interlaboratory variability of scoring [46 , 47 ]. Moreover with the possibility of scoring tenfold more cells in the same period of time, compared to the standard value of 2,000, the statistical power is increased relative to the conventional microscopic evaluation. In the future, in light of the emerging developments of the omics technologies, transcriptomic tools should be incorporated in the integrated analysis herein proposed. Furthermore, proteomic- based tools may help disclosing the complex mechanisms underlying the relevant biological effects of nanomaterials.

Acknowledgments

The authors thank Keld Jensen (NRCWE, Denmark), Conny van Oostrom, Edwin Zwart, Harry van Steeg, Jan van Benthem and Wim de Jong (RIVM, the Netherlands), Lisa Giovannelly and coworkers (University of Florence, Italy), and Andrew Collins (University of Oslo, Norway) for their contribution to the development of the described methodologies at INSA.

References

1. Louro H, Silva MJ (2010) In vivo mutagenic 2. Rothfuss A, Honma M, Czich A et al (2011) effects of alkylating agents eliciting different Improvement of in vivo genotoxicity assess- DNA-adducts. In: Alvarez E, Cunha R (eds) ment: combination of acute tests and integra- DNA adducts: formation, detection and mutation into standard toxicity testing. Mutat Res genesis. Nova Science, New York 723:108–120 476 Henriqueta Louro et al.

3. Bowen DE, Whitwell JH, Lillford L et al 15. EFSA (2012) Minimum criteria for the (2011) Evaluation of a multi-endpoint assay in acceptance of in vivo alkaline Comet Assay rats, combining the bone-marrow micronu- reports cleus test, the Comet assay and the fl ow- 16. Jackson P, Pedersen LM, Kyjovska ZO et al cytometric peripheral blood micronucleus test. (2013) Validation of freezing tissues and cells Mutat Res 722:7–19 for analysis of DNA strand break levels by 4. Vasquez MZ (2010) Combining the in vivo comet assay. Mutagenesis 28(6):699–707 comet and micronucleus assays: a practical 17. Azqueta A, Collins AR (2013) The essential approach to genotoxicity testing and data comet assay: a comprehensive guide to measur- interpretation. Mutagenesis 25:187–199 ing DNA damage and repair. Arch Toxicol 5. Boverhof DR, Chamberlain MP, Elcombe CR 87:949–968 et al (2011) Transgenic animal models in toxi- 18. Collins AR (2014) Measuring oxidative dam- cology: historical perspectives and future out- age to DNA and its repair with the comet assay. look. Toxicol Sci 121:207–233 Biochim Biophys Acta 1840(2):794–800 6. OECD (2011) OECD guideline for the testing 19. Silva MJ, Dias A, Barreta A et al (2002) Low of chemicals—transgenic rodent somatic and frequency noise and whole-body vibration germ cell gene mutation assays. OECD cause increased levels of sister chromatid 7. ICH—International Conference on Harmoni- exchange in splenocytes of exposed mice. sation of Technical Requirements for Regis- Teratog Carcinog Mutagen 22:195–203 tration of Pharmaceuticals for Human Use 20. Perry P, Wolff S (1974) New Giemsa method (2012) S2(R1) Genotoxicity Testing and Data for the differential staining of sister chromatids. Interpretation for Pharmaceuticals Intended Nature 251:156–158 for Human Use. US Department of Health 21. Tice RR, Agurell E, Anderson D et al (2000) and Human Services, Food and Drug Single cell gel/comet assay: Guidelines for Administration in vitro and in vivo genetic toxicology testing. 8. Boerrigter ME, Dollé ME, Martus HJ et al Environ Mol Mutagen 35:206–221 (1995) Plasmid-based transgenic mouse model 22. European Union 2010 Directive 2010/63/ for studying in vivo mutations. Nature 377: EU of the European parliament and of the 657–659 Council of 22 September 2010 on the protec- 9. Gossen JA, Martus HJ, Wei JY et al (1995) tion of animals used for scientifi c purposes Spontaneous and X-ray-induced deletion NCIL. Offi cial Journal of the European Union mutations in a LacZ plasmid-based transgenic 23. OECD (2010) Guidance manual for the test- mouse model. Mutat Res 331:89–97 ing of manufactured nanomaterials. ENV/ 10. Louro H, Silva MJ, Boavida MG (2002) JM/MONO(2009)20/REV Mutagenic activity of cisplatin in the lacZ 24. Tavares A, Louro H, Antunes S et al (2014) plasmid-based transgenic mouse model. Genotoxicity evaluation of nanosized titanium Environ Mol Mutagen 40:283–291 dioxide, synthetic amorphous silica and multi- 11. OECD (1997) Test No. 474: Mammalian walled carbon nanotubes in human lympho- erythrocyte micronucleus test, OECD guide- cytes. Toxicol In Vitro 28(1):60–69 lines for the testing of chemicals, Section 4. 25. Landsiedel R, Kapp MD, Schulz M et al (2009) OECD Genotoxicity investigations on nanomaterials: 12. OECD (2010) Test No. 487: In vitro mamma- methods, preparation and characterization of lian cell micronucleus test, OECD guidelines test material, potential artifacts and limita- for the testing of chemicals, Section 4: Health tions–many questions, some answers. Mutat effects. OECD Res 681:241–258 13. Sato S, Taketomi M, Nakajima M et al (1995) 26. Recio L, Hobbs C, Caspary W et al (2010) Effect of aging on spontaneous micronucleus Dose–response assessment of four genotoxic frequencies in peripheral blood of nine mouse chemicals in a combined mouse and rat strains: the results of the 7th collaborative micronucleus (MN) and Comet assay protocol. study organized by CSGMT/JEMS.MMS. J Toxicol Sci 35:149–162 Collaborative Study Group for the Micro- 27. Louro H, Faustino I, Dias A et al (2010) Poly nucleus Test. Environmental Mutagen Society (ADP-ribose) polymerase-1 defi ciency does of Japan. Mammalian Mutagenesis Study not affect ethylnitrosourea mutagenicity in Group. Mutat Res 338:51–57 liver and testis of lacZ transgenic mice. Environ 14. Heddle JA, Fenech M, Hayashi M et al (2011) Mol Mutagen 51:322–329 Refl ections on the development of micronu- 28. Charles JL, Jacobson-Kram D, Condie LW Jr cleus assays. Mutagenesis 26:3–10 et al (1986) The kinetics of in vivo sister An Integrated Approach for In Vivo Genotoxicity Testing 477

chromatid exchange induction in mouse bone blood reticulocytes using acridine orange-coated marrow cells by ethylnitrosourea and methylni- slides. Mutat Res 245:245–249 trosourea. Toxicol Appl Pharmacol 84:56–65 39. Louro H, Silva MJ (2011) Cost/benefi t of 29. Jensen KA, Kembouche Y, Christiansen E et al mutation induction under PARP1 defi ciency: (2011) The generic NANOGENOTOX disper- from genomic instability to therapy. In: Urbano sion protocol—Standard Operation Procedure KV (ed) Advances in genetics research. Nova (SOP). http://www.nanogenotox.eu Science, New York, pp 109–134 30. Gossen JA, de Leeuw WJ, Molijn AC et al 40. Dass SB, Ali SF, Hefl ich RH et al (1997) (1993) Plasmid rescue from transgenic mouse Frequency of spontaneous and induced micro- DNA using LacI repressor protein conjugated nuclei in the peripheral blood of aging mice. to magnetic beads. Biotechniques 14:624–629 Mutat Res 381:105–110 31. Gossen JA, Vijg J (1988) E. coli C : a conve- 41. Bender MA, Preston RJ, Leonard RC et al nient host strain for rescue of highly methyl- (1992) On the distribution of spontaneous ated DNA. Nucleic Acids Res 16:9343 SCE in human peripheral blood lymphocytes. 32. Gossen JA, de Leeuw WJ, Tan CH et al (1989) Mutat Res 281:227–232 Effi cient rescue of integrated shuttle vectors 42. Bonner JC, Silva RM, Taylor AJ et al (2013) from transgenic mice: a model for studying Interlaboratory evaluation of rodent pulmo- mutations in vivo. Proc Natl Acad Sci U S A nary responses to engineered nanomaterials: 86:7971–7975 The NIEHS Nano GO Consortium. Environ 33. Louro H, Pinheiro I, Costa P et al (2008) Health Perspect 121:676–682 Mutagenic effects of poly (ADP-ribose) poly- 43. Azqueta A, Langie SA, Slyskova J et al (2013) merase-1 defi ciency in transgenic mice. Mutat Measurement of DNA base and nucleotide Res 640:82–88 excision repair activities in mammalian cells and 34. Dolle ME, Snyder WK, van Orsouw NJ et al tissues using the comet assay—a methodologi- (1999) Background mutations and polymor- cal overview. DNA Repair 12:1007–1010 phisms in lacZ-plasmid transgenic mice. 44. Azqueta A, Gutzkow KB, Priestley CC et al Environ Mol Mutagen 34:112–120 (2013) A comparative performance test of 35. Boerrigter ME, Vijg J (1997) Sources of vari- standard, medium- and high-throughput comet ability in mutant frequency determinations in assays. Toxicol In Vitro 27:768–773 different organs of lacZ plasmid-based trans- 45. Shaposhnikov S, Azqueta A, Henriksson S et al genic mice: experimental features and statistical (2010) Twelve-gel slide format optimised for analysis. Environ Mol Mutagen 29:221–229 comet assay and fl uorescent in situ hybridisa- 36. Hartmann A, Agurell E, Beevers C et al (2003) tion. Toxicol Lett 195:31–34 Recommendations for conducting the in vivo 46. MacGregor JT, Bishop ME, McNamee JP et al alkaline Comet assay. 4th International Comet (2006) Flow cytometric analysis of micronuclei Assay Workshop. Mutagenesis 18:45–51 in peripheral blood reticulocytes: II. An effi - 37. Brendler-Schwaab SY, Schmezer P, Liegibel U cient method of monitoring chromosomal et al (1994) Cells of different tissues for in vitro damage in the rat. Toxicol Sci 94:92–107 and in vivo studies in toxicology: compilation 47. Dertinger SD, Bishop ME, McNamee JP et al of isolation methods. Toxicol In Vitro 8: (2006) Flow cytometric analysis of micronuclei 1285–1302 in peripheral blood reticulocytes: I. Intra- and 38. Hayashi M, Morita T, Kodama Y et al (1990) interlaboratory comparison with microscopic The micronucleus assay with mouse peripheral scoring. Toxicol Sci 94:83–91 INDEX

A Bhas 42 cells ...... 343–360 Biomonitoring ...... 7 3, 97, 116, 118, 121, Ames test ...... 3–20, 23–40, 43, 46, 47, 118, 337 129, 188, 199, 221, 219, 220, 222, 223, 227–229, Ames II ...... 23–40 233–236, 269, 367, 377, 393, 379 Ames MPF ...... 23–40 Bromodeoxyuridine (BrdU) ...... 79, 98, 121, 9-Aminoacridine (9-AAC) ...... 13, 39 229, 453, 454, 468, 469 2-Aminoanthracene (2-AA)...... 14, 31, 37–39 2-Aminofluorene (2-AF) ...... 14 C Aneugenicity test ...... 104 Carcinogenicity in vitro ...... 345–346 Aneuploidogenic activity ...... 80 Carcinogens Aneuploidy ...... 80, 81, 116, 129, 135, 417 genotoxic ...... 141, 170, 203, 284, Apurinic/apyrimidinic sites (AP-sites) ...... 117, 201–204, 343–360, 392, 451 261, 366 non genotoxic ...... 343–360 B CBMN-cyt assay. See Cytokinesis-block micronucleus cytome (CBMN-cyt) assay Bacterial reverse mutation assay ...... 5, 27, 32, 48, 56, 57 C57B1/6 mouse strain ...... 452 gene mutations pUR288 plasmid shuttle vector ...... 452 base-pair changes/substitutions ...... 1 1, 24, 25, CD59 cell surface marker ...... 1 72, 174, 177, 179 44, 45 CD4+ T cell ...... 159–164 frameshifts ...... 11, 23, 24, 33, 40, 44, 45 Cell storage ...... 260, 347, 349, 385 liquid microfluctuation method ...... 2 4 Cell survival ...... 1 75, 260, 354, 356, nanoparticle analysis ...... 43–57 417–419, 422–423, 441–445 Penta I assay ...... 23, 24, 27, 32, 33, 36, 39 Cell transformation assay plasmid pAQ1 ...... 11, 15, 44 cell growth assay ...... 347–349, 351–359 plasmid pKM101 ...... 11, 15, 24, 25, 44 cells pre-incubation method ...... 4 , 24, 46, 54, 55 BALB/c 3T3 ...... 343–346, 350, 357 reverse mutations ...... 3–20, 23, 24, 27, 32, Bhas 42 ...... 343–360 43–57, 59, 60, 118 solubility test ...... 349–350 rfa mutation ...... 5, 9, 11, 13, 24, 33, 44 transformation assay ...... 343–360 Salmonella tester strains ...... 3, 5, 11, 12, 16, tumor initiation assay...... 344, 345, 348, 20, 25, 33, 38, 54 349, 353, 355–359 Salmonella typhimurium ...... 2 3, 24, 44 tumor promotion assay ...... 344, 345, 348, 349, standard plate incorporation method ...... 4, 17–18, 356, 357, 359 24, 34 Chromosomal aberrations ...... 59, 74, 76, 80, 83, TAMix strains...... 2 4, 25, 26, 27, 29, 31, 40 91, 103, 105, 115–125, 129–131, 133–135, TA strains ...... 4 , 11, 12, 16, 24, 25, 27, 141–143, 154–157, 203, 220, 285, 348 32, 33, 44, 46–48 Chromosomal aberration test ...... 59, 76, 115–136, 203 uvrB mutation ...... 11, 14, 24, 25, 33, 44, 45 Chromosome banding ...... 142 BALB/c 3T3 cells ...... 343–346, 350, 357 G-bands ...... 120, 121 Base excision repair (BER) ...... 366, 367, 369, Chromosome painting ...... 125, 141–157 370, 375, 377, 378, 379, 383, 390, 392, 393, 397, Clastogenic activity ...... 80 398, 406, 416, 417 Clastogenicity ...... 74, 80, 82, 91, 109, 116, Benzo(a)pyrene (B(a)P) ...... 14, 38, 39, 63, 66, 119, 215, 325 127, 129, 141, 172, 266, 324

L. María Sierra and Isabel Gaivão (eds.), Genotoxicity and DNA Repair: A Practical Approach, Methods in Pharmacology and Toxicology, DOI 10.1007/978-1-4939-1068-7, © Springer Science+Business Media New York 2014 479 GENOTOXICITY AND DNA REPAIR: A PRACTICAL APPROACH 480 Index Clastogenicity test ...... 129, 453 D Colcemid ...... 121, 144, 150, 468 Comet assay DDR. See DNA damage response (DDR) high throughput assay version ...... 199–215, 242, DNA damage response (DDR) ...... 116, 269, 266, 369, 475 393, 398, 399, 409, 417 human/mammalian cell extracts, in vitro DNA isolation ...... 3 06, 307, 311, 314, 460–461 cultured cells ...... 384–385 DNA repair peripheral blood mononuclear cells ...... 367 measure solid tissues ...... 377–394, 454 Drosophila , comet assay ...... 269–280, in vitro, cultured cells ...... 247, 265 397–410 in vivo humans/mammals, comet assay ...... 199, 202, Drosophila ...... 2 69–280, 397, 398, 219–236, 253, 375, 390, 393, 397, 398, 404, 406 400, 403, 404, 409–410 RNA interference ...... 413–445 humans (biomonitoring) ...... 219–236, 243, systems 244, 252, 398 BER ...... 366, 367, 369, 370, 375, mouse...... 2 45, 247 377, 378, 379, 383, 390, 392, 393, 397, 398, 406, nanoparticles ...... 241–266, 270, 456 416, 417, 475 repair analysis, in vitro ...... 398 homologous recombination ...... 284, 285, repair analysis, in vivo 293, 366, 399, 438 Drosophila ...... 2 69, 270, 283, 285, 398, MGMT ...... 367, 418 400, 403, 404, 408, 409 NER ...... 367, 370–371, 374, humans ...... 219–236, 243–244 375, 378–380, 382–384, 386–388, 390, 392–394, incision assay ...... 378–380, 398 398–400, 403, 406 Controls NHEJ ...... 399, 417 negative ...... 17, 19, 25, 32, 35, 55, 67, rejoining of SSB...... 3 66, 368 81, 112, 119, 120, 124, 133–135, 143, 190, 194, Drosophila extracts ...... 397–410 210, 214, 229, 244–245, 253, 258, 261, 265, 266, Drosophila melanogaster 273, 277, 279, 290–293, 302, 328, 333, 338, 347, comet assay 349, 381, 404, 405, 409, 419, 424, 425, 457, 463, haemocytes isolation ...... 277, 291 466, 470 in vitro repair assay ...... 398, 400, 402–409 in vitro assays ...... 214, 374, 381 in vivo repair assay ...... 3 98, 400, 409–410 in vivo assays ...... 4 04, 405 neuroblast cells isolation ...... 274, 277, 404, 405 positive...... 4, 5, 12, 13, 14, 17, 20, 25, culture media ...... 271, 276, 277, 288, 27–29, 31, 35–39, 51, 52, 54, 55, 57, 91, 105, 108, 289, 299–302 110, 119, 120, 135, 188, 190, 214, 229, 245, 254, SMART assays 256–260, 265, 273, 277, 279, 309, 310, 325, 328, eye-spot test ...... 284–288, 287, 333, 336–338, 346, 347, 381, 382, 387, 398, 400, 288, 291–293 406, 456, 457, 463, 465, 468 wing-spot test ...... 283–291 in vitro assays ...... 60–61, 214, 398 E in vivo assays ...... 455, 457 CP. See Cyclophosphamide (CP) Endoreduplicated chromosomes ...... 134–136 Crystal violet ...... 9, 13, 359 Erythrocytes Cumene hydroperoxide ...... 38, 45, 54 polychromatic erythrocytes (PCE) ...... 104–106, Cyclophosphamide (CP) ...... 105, 108, 111, 109–111, 453, 467, 468 119, 325, 328 Escherichia coli ...... 1 2, 23, 44, 369, 452 Cytogenetics ...... 78, 120, 122, 126–128, Ethidium bromide 142, 157, 170, 220, 324 solutions...... 2 12, 231, 272, 278, Cytokinesis-block micronucleus cytome 388, 402, 465 (CBMN-cyt) assay ...... 74, 75, 77–85, staining ...... 212, 231, 252, 255, 272, 88, 90–94, 97, 98 278, 388, 407, 465 Cytome assay ...... 74, 81 Ethyl methanesulfonate (EMS)...... 188, 277, 457 Cytotoxicity analysis Ethylnitrosourea (ENU) ...... 188, 308–310, CyQUANT assay ...... 418, 419, 443–445 457, 463, 468 MTS assay ...... 418, 419, 442–443 Exposure control ...... 59–70, 80 GENOTOXICITY AND DNA REPAIR: A PRACTICAL APPROACH 481 Index F I

FISH chromosome painting ...... 141–143, Immunofluorescence staining ...... 161–163 145, 150–157 Ionizing radiation ...... 74, 75, 116, 117, 124, Flow cytometry 130, 141, 159, 160, 172, 201, 234, 270, 366, 422 PIG-A gene mutation assay ...... 1 70, 172–177, 180 TCR gene mutation assay ...... 159–164 K Fluorescence in situ hybridization (FISH) ...... 73–98, Kinetocore proteins ...... 77 118, 121, 124–127, 130, 141–143, 145, Knockdown gene expression ...... 414–416, 418, 150–154, 156 420–422, 431–438 Formamidopyrimidine DNA glycosylase (FPG) ...... 199–215, 246, 247, L 257, 261–263, 368–370, 373, 382–384, 387, 400, LacZ 401, 464–466 gene ...... 307, 320, 452, 453, 455, 463, 474 G mutations ...... 459–463 LacZ plasmid mouse ...... 306, 452, 453, Gene mutation ...... 3, 4, 103, 118, 169, 455, 459–463 170, 242, 266, 284, 286, 323, 417, 451, 455, 457 Lentivirus ...... 4 15, 416, 418, 420, 422, Gene mutation assay 425–430, 435, 436, 445 HPRT ...... 171, 185–195, 324, 452 Lesion-specific enzymes ...... 204, 242, 369 mouse lymphoma...... 118, 187, 323–342 Lymphocytes ...... 7 4–76, 78, 83, 86–89, 97, RBC PIG-A ...... 1 69–182 120–125, 129–133, 144, 160, 161, 163–165, 174, TCR ...... 159–165 182, 186, 223–227, 378, 379, 390, 454 transgenic rodent ...... 305–320, 452, 453, L5178Y/ Tk +/- –3.7.2C cell line ...... 79, 118, 187, 324, 326 455, 459–463 Genotoxicity testing ...... 2 3, 70, 83, 118, 119, M 170, 199, 203, 204, 215, 220, 242, 266, 454, 3-Methylcholanthrene (MCA)...... 188, 347–349 456, 474 Methylglyoxal (MG) ...... 13 Giemsa staining ...... 8 5, 88, 109, 120–124, Methylmethanesulfonate (MMS) ...... 13, 273, 325, 328, 143, 157, 347, 355, 453, 466, 467, 469 422, 439 Glycophorin ...... 159, 174 Methyl-nitrosourea (MNU) ...... 187 Glycosyl phosphatidyl inositol (GPI) anchor ...... 171–174 Mice ...... 97, 105–108, 110, 118, H 141–157, 161, 170, 172, 174, 186, 187, 243–245, 252, 265, 306, 309, 323–340, 344, 346, 392, 416, Histology 421, 422, 452–457, 459–464, 466–474 lesions ...... 453, 457 Microfluctuation test ...... 24, 34, 35 markers ...... 454, 455, 458, 470–473 Micronuclei ...... 47, 59, 73–99, 103–112, samples ...... 389, 456, 458, 470–473 118, 127, 129, 130, 157, 203, 246, 310, 337, Histopathology ...... 227, 472, 473 452–457, 466–468, 474 HOCs. See Hydrophobic organic chemicals (HOCs) Micronucleus test Human biomonitoring FISH ...... 73–99, 121, 124–127, 130, chromosomal aberration test ...... 115–136 141–145, 151–157, 420, 468 comet assay ...... 1 29–130, 219–224, 393 in vitro confounding factors ...... 222, 234 cultured cells ...... 118–120 gene mutation assays human lymphocytes ...... 74, 120–124, 188, PIG-A ...... 1 69–182, 310 378, 390, 454 TCR ...... 159–165 in vivo Human lymphoblastoid cells ...... 186–188, 191 blood ...... 106, 108–110, 159–162, 170, HUman MicroNucleus (HUMN) project ...... 75 223–227, 234, 244, 250, 398, 453, 466, 467, 475 Human peripheral blood ...... 86, 121, 122, bone marrow ...... 1 03–106, 108–111, 161, 162, 164, 171–176, 179, 223, 224, 383 453, 457, 467, 468 Hydrophobic mutagen ...... 60 nanoparticles ...... 246 Hydrophobic organic chemicals (HOCs) ...... 60–70 Microscopy ...... 4 7, 121, 157, 219, 300, Hypoxanthine-guanine phosphoribosyl transferase (HPRT) 337, 368, 458, 470, 472–473 gene ...... 171, 185–195, 452 Mitosis check-point proteins ...... 77 GENOTOXICITY AND DNA REPAIR: A PRACTICAL APPROACH 482 Index MLA. See Mouse lymphoma assay (MLA) mouse lymphoma assay ...... 325, 332, 337–340 MMS. See Methylmethanesulfonate (MMS) multiple genotoxicity endpoints ...... 451–475 MNU. See Methyl-nitrosourea (MNU) preparation/disaggregation ...... 242, 249–252, Molecular cytogenetics 256–259, 263, 302, 458–459, 461, 464, 466 regulatory science...... 157, 181–182 SMART assay ...... 284, 299 techniques Nanoparticles (NPs) ...... 297–303 aCGH (array-Comparative Genomic NER. See Nucleotide excision repair (NER) Hybridization) ...... 125, 126 2-Nitrofluorene (2-NF) ...... 13, 31, 39, 54 COBRA-FISH (Combined Binary Ratio-labeling 4-Nitro-o -phenilenediamine (4-NPD)...... 13 FISH) ...... 125, 126 4-Nitroquinoline-1-oxide (NQO) ...... 13, 31, 325, fluorescence in situ hybridization 328, 332, 337–340 (FISH) ...... 125, 126 N' -methyl-N' -nitro-N -nitrosoguanidine mFISH (multicolor-FISH) ...... 125, 126 (MNNG) ...... 422, 423, 440–444 Monoclonal antibody ...... 160, 161, 422, 438 NPs. See Nanoparticles (NPs) Mouse FISH probe cocktails Nucleotide excision repair (NER) ...... 203, 270, multicolor ...... 1 26, 145, 153, 155 367, 370–371, 374, 375, 378–380, 382–384, single...... 145, 153, 154 386–388, 390, 392–394, 398–400, 403, 406, 475 Mouse lymphoma assay (MLA) agar version ...... 3 32–334, 336, 338–339 O colonies Oxidative stress...... 1 94, 242, 246, 298, 375 large ...... 324, 333, 335, 336, 339, 340 Oxidative DNA damage ...... 2 47, 270, 451, 466 small ...... 323, 324, 331, 333, 8-Oxoguanine adducts 335, 336, 338–340 induction ...... 383–384

L5178Y/ Tk+/- –3.7.2C cell line ...... 118, 187, repair ...... 366, 368–370, 378, 324, 326 383, 384, 389, 406 microwell version ...... 330, 333, 334, 336, Oxygen peroxide ...... 241 338, 340 nanoparticles assessment ...... 3 25, 332, 337–340 P thymidine kinase gene ...... 194, 323 Packaging plasmids trifluorothymidine ...... 323 pMD2.g(VSVG) ...... 416, 420, 426 Multiple genotoxicity endpoints ...... 204, 284, 455, pMDLg/pRRE...... 416, 420, 426 473–474 pRSV-REV...... 416, 420, 426 Muta™ mouse ...... 305–320 Paroxysmal nocturnal hemoglobinuria Mutation type (PNH) ...... 171–173, 179 chromosomal aberrations ...... 59, 47, 76, 80, 83, PARP activation ...... 4 16, 417, 422, 438–440 91, 103, 105, 115–125, 129–131, 133–135, PAR detection ...... 417, 422, 438–441 141–143, 154–157, 203, 220, 285, 348 Passive dosing numerical ...... 115, 116, 120, 126, 131, 135 biocompatibility ...... 61, 63, 64 structural ...... 115, 117, 120, 131, phase ...... 64–66 134, 135, 141 PCE/NCE ratio. See Polychromatic erythrocyte (PCE)/ gene mutation ...... 3 , 4, 103, 118, 169, 170, normochromatic erythrocytes (NCE) ratio 242, 266, 284, 286, 323, 417, 451, 455, 457 Peripheral blood ...... 86, 88, 106, 109, 110, base pair changes ...... 25, 32, 40 122, 125, 141–157, 161, 170–173, 182, 223, 224, frameshifts ...... 11, 23, 24, 33, 34, 227, 367, 384, 453, 457, 458, 464–467, 475 40, 44, 45, 169, 194 Peripheral blood culture ...... 384–385 N Photogenotoxic effects...... 2 42 PIG-A gene mutation assay NADP solution ...... 9 –10, 19, 27, 35, 325, 328 CD59 cell surface marker ...... 172, 174, 176, N4 -Aminocytidine (N4 -ACT) ...... 38, 39 177, 179–181 Nanomaterials/nanoparticles conjugated antibodies bacterial reverse mutation assay ...... 43–57 APC ...... 174, 176 comet assay ...... 241–266, 270, 337, FITC ...... 174, 176 452–458, 463–466, 475 flow cytometry ...... 170, 172–177, 180 MN assay ...... 81–83, 475 glycophorin ...... 174 GENOTOXICITY AND DNA REPAIR: A PRACTICAL APPROACH 483 Index glycosyl phosphotidyl inositol (GPI) S9 mix anchor ...... 171–174 preparation ...... 27, 29, 30, 35, 37, 325, 328 paroxysmal nocturnal hemoglobinuria (PNH) ...... 171 use ...... 27, 29, 34, 35, 37, 46, 49, 328

PIG-A gene ...... 169–182 Sodium azide (NaN3 ) ...... 13, 14, 54, 161 Plasmid pAQ1 ...... 11, 15, 44 Somatic mutation ...... 159, 171–172, 283, Plasmid pKM101 ...... 11, 15, 24, 25, 33, 44 301, 305–320, 417 PNH. See Paroxysmal nocturnal hemoglobinuria (PNH) Spleen tissue culture poly-ADP-ribose (PAR) detection spleen harvest ...... 143–144, 146–147, 150 immunoblot ...... 417, 420–421, 438–440 Spontaneous revertant ...... 4 , 12, 17, 32, 45, 55 immunofluorescence ...... 417, 440–441 Stable gene knockdown (shRNA). See RNA interference Polychromatic erythrocyte (PCE)/normochromatic Strain storage ...... 20 erythrocytes (NCE) ratio ...... 110, 468 Strand break rejoining ...... 118, 368 Poly (ADP-ribose) polymerase (PARP) SyberGold staining ...... 231–232 activation ...... 417, 422, 438–441 T Q TAMix strains ...... 24, 25 Quantitative reverse transcription polymerase chain reaction TA strains ...... 4, 11–16, 24–29, 31–40, (qRT-PCR) ...... 416, 420, 424, 425, 44, 46–48 429, 431–435 T-cell receptor gene ...... 160 TCR gene mutation assay R CD4 T cells ...... 159, 160, 163, 164 Rats ...... 19, 20, 29, 46, 105, 106, 108, 170, flow cytometry ...... 162–164 172–174, 186, 187, 243, 244, 265, 306, 328, 456 immunofluorescence staining ...... 1 61–163 Regulatory science ...... 157, 181–182 monoclonal antibodies ...... 160 Reverse mutations ...... 3, 44, 46 T-cell receptor gene ...... 160 rfa mutation ...... 5 , 9, 11, 13, 24, 33, 44 12-O -Tetradecanoyl phorbol-13-acetate (TPA) ...... 344, Risk assessment ...... 103, 142, 157, 343, 452 347–349, 355 RNA interference TGR gene mutation assay. See Transgenic rodent (TGR) 293 cell line ...... 415, 416, 420, 426, 427 gene mutation assay gene knockdown analysis Thymidine kinase gene ...... 1 86, 187, 194, 323 immunoblot ...... 416, 435–438 Transgenic rodent (TGR) gene mutation assay qRT-PCR ...... 416, 420, 424, 431–435 lacZ plasmid mouse ...... 306 lentiviral transduction ...... 4 29–431, 435, 436 Muta™ mouse lentivirus preparation ...... 426–430 animal exposure ...... 309–310 packaging plasmids DNA isolation ...... 306, 307, 311, 315 pMD2.g(VSVG) ...... 416, 420, 426 mutant detection ...... 316–319 pMDLg/pRRE ...... 416, 420, 426 tissue collection ...... 310–311 pRSV-REV ...... 416, 420, 426 tissue digestion...... 3 11–314 PARP...... 4 16, 417, 422, 438–441 2,4,7-Trinitro-9-fluorenone (TNF) ...... 13 RISC ...... 414 U stable gene knockdown (shRNA) ...... 413–416, 418, 420, 422, 423, 425–431, 435, 436, 439, 441 UV light transient gene knockdown (siRNA) ...... 413–416, photoproducts 419–420, 423–425, 441 induction ...... 383, 384 Ro 19-8022 ...... 211, 214, 256, 259, 369, repair ...... 383, 384, 406 370, 378–384, 394, 404–406 uvrB mutation ...... 11, 14, 24, 25, 33, 44, 45

S V

Salmonella tester strains ...... 3, 5, 11, 12, 16, 20, 25, 38 Visible light ...... 3 70, 377, 378, 380, 404–406, 426 Salmonella typhimurium ...... 5 , 12, 23, 24, 32, 33, 44, 46, 48 W Small interfering RNA (siRNA). See RNA interference SMART assays White light. See Visible light eye-spot ...... 283, 287, 288, 291–293 Whole blood ...... 8 4, 86, 88, 89, 97, wing-spot ...... 283, 284, 287–291, 299 147–150, 162, 224, 226–227, 234, 367, 384