Modulation of Pharmacologic Effects of 5-Azacytidine by Reductase Antisense GTI-2040

Dissertation

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By

Josephine Eki Aimiuwu, B.S.P.S

Graduate Program in Pharmacy

The Ohio State University

2011

Dissertation Committee:

Kenneth K. Chan, Ph.D., Advisor

Guido Marcucci, M.D., Co-advisor

Robert Lee, Ph.D.

Robert M. Snapka, Ph.D.

ABSTRACT

Cancer, like most human diseases is complex and the therapeutic approaches available to treat this disease have limited efficacy. Therefore, combination studies of two or more anticancer drugs is expected to be essential in achieving a better therapeutic response in patients and may also provide a cure for drug-resistant cancers. Leukemia is a cancer of the blood, which results from an uncontrolled proliferation of white blood cells, thereby inhibiting its functions. Important insights into the pathogenesis of this disease have led to the development of a number of anti-leukemia drugs, including analogs and new antisense compounds, that intervene at the level of disease progression. 5-

Azacytidine and decitabine are hypomethylating agents recently approved by the U.S Food and Drug Administration for the treatment of Myelodysplastic Syndrome (MDS) and are also in clinical trials for the treatment of hematological malignancies, such as acute myeloid leukemia (AML). These drugs are approved based on their ability to induce DNA demethylation, resulting in reactivation of hypermethylation-associated silencing of tumor suppressor genes. Aracytidine (Ara-C) is another nucleoside drug, widely used as antimetabolite for the treatment of acute myelogenous leukemia. Antisense GTI-2040 is a

20-mer oligonucleotide inhibiting the expression of ribonucleotide reductase subunit M2

(RRM2) mRNA, an that has been found to be over-expressed in most cancers. In this dissertation, investigations on pharmacodynamic studies of nucleoside analogs in combination with GTI-2040 were carried out. It has been demonstrated that inhibition of

ii cellular RR, which subsequently decreases triphosphate (dNTP) pools, could enhance the anti-tumor activity of subsequently administered nucleoside analogs. We have in our studies, provided both in vitro and in vivo evidences to support the novel combination treatment of antisense GTI-2040 and 5-azaC, leading to a synergistic effect. In addition, GTI-2040 decreases RRM2 levels, and most notably, we discovered that 5-azaC modulates RRM2 for the first time and this result makes RR a novel target for 5-azaC. In addition, the biomarkers involved in the development of 5-azaC and decitabine (DAC) resistances were assessed in order to elucidate the potential mechanisms that contribute to the induction of resistance in cancer cells. In a phase II evaluation of GTI-2040 in combination with Ara-C in patients with AML at The James Cancer Hospital and Research

Institute at The Ohio State University, clinical pharmacokinetic of GTI-2040 and the in vitro-in vivo pharmacodynamic analysis with Ara-C was established to assist in the exploration of their pharmacokinetic-pharmacodynamic (PK-PD) correlations in relation to clinical response. Finally, our studies of GTI-2040, 5-AzaC, DAC and Ara-C provide valuable insights into their clinical development as a single agent or in combination with other drugs.

iii

DEDICATION

Dedicated to my family

iv

ACKNOWLEDGEMENTS

I would like to thank my adviser, Dr. Kenneth K. Chan, for his intellectual guidance, encouragement and motivation, continuing support during my graduate studies. I sincerely appreciate his help in every aspect of this research and for his invaluable suggestions. I also would like to thank my co-adviser, Dr. Guido Marcucci for his support in the clinical study and his invaluable suggestions and comments on this project. I also thank Dr. Rebecca Klisovic, for the opportunity to work in the clinical trial. Appreciation also goes to my committee members, Dr. Robert Lee and Dr. Robert M. Snapka for their time, constructive suggestions and comments on this project. I would like to thank Dr.

Duxin Sun’s support, sound advice and friendship during my graduate studies.

Appreciation also goes to our collaborators in Dr. Marcucci’s laboratory. Special thanks go to Dr. Shujun Liu for the valuable scientific discussions on biological studies, and stimulating discussions on this project. Ms. Jiuxia Pang assistance on western blotting experiments and without both of their help and effort, my work would have undoubtedly been more difficult. I would also like to express my thanks to Ms.

LeNguyen Huynh for the time she spent on collecting patient samples and help with real time PCR experiments.

I would also like to thank Dr. Ping Chen for her valuable scientific discussions and the development of most of the assays used in our biochemical modulation experiments. I am

v thankful for Dr. Zhongfa Liu’s suggestions and assistance earlier on in my studies. I benefited from his immense experience in analytical chemistry and help using mass spectrometers. My sincere thanks also go to Dr. Zhiliang Xie for the great help and assistance on sample preparation and analysis and Dr. Ming Chiu for his help in animal work. I also thank Dr. Hongyan Wang for stimulating discussion and assistance on this project and her immense friendship, Dr. Yonghua Ling for her support and warm friendship. I would like to thank all of my labmates and friends who directly or indirectly have allowed me to be where I am today. Thanks are also to Dr. Tracey Wei, Dr. Hao

Cheng, Dr. Vijay Sarathi, Dr. Jiang Wang and Dr. Xianhua Cao for their assistance in various aspects.

I am also very grateful to Ms. Joy Scott for her administrative help, encouragement and warm friendship and Ms. Kathy Brooks. I would also like to express my thanks to all of my friends and colleagues in the College of Pharmacy, especially

Casey Hoerig for his help in computer related issues.

Finally, my special thanks go to my family. I am deeply and forever indebted to my husband, Osagie Christopher for his love, support and encouragement throughout my studies and to my children, Izoduwa Maria and Osagie Kingsley for making my time in school so much easier than I could have imagined. I also like to thank my parents for their everlasting and unconditional love and for believing in me at all times.

vi

VITA

June 2005 ...... B.S. in Pharmaceutical Science,

The Ohio State University

2005 to present ...... Graduate Research Associate,

The Ohio State University

PUBLICATIONS Research Publications

1. Eric H. Kraut, Christopher Rhoades, Yilong Zhang, Hao Cheng, Josephine Aimiuwu, Ping Chen, James Lang, Donn Young, Amit Agrawal, Janet Dancey, Kenneth Chan, Michael R Grever. Phase I and pharmacokinetic study of erlotinib (OSI-774) in combination with docetaxel in squamous cell carcinoma of the head and neck (SSCHN). Cancer Chemotherapy and Pharmacology, 2010, 1-8.

2. Guisheng Zhang, Lanyan Fang, Lizhi Zhu, Josephine Aimiuwu, Jie Shen, Hao Cheng, Mark Muller, Gun Lee, Duxin Sun and Peng Wang. Synthesis and Biological Activities of Disaccharide Daunorubics. J.Med.Chem. 2005, 48, 5269-5278.

3. Ping Chen, Zhongfa Liu, Shujun Liu, Zhiliang Xie, Josephine Aimiuwu, Jiuxia Pang, Rebecca Klisovic, William Blum, Michael Grever, Guido Marcucci and Kenneth K. Chan. A LC-MS/MS Method for the Analysis of Intracellular levels. Pharmaceutical Research. 2009, 26(6), 1504-1515.

4. Zhongfa Liu, Shujun Liu, Zhiliang Xie, Ryan E. Pavlovicz, Jiejun Wu, Ping Chen, Josephine Aimiuwu, Jiuxia Pang, Deepak Bhasin, Paolo Neviani, James R. Fuchs, Christoph Plass, Pui-Kai Li, Chenglong Li, Tim H-M Huang, Lai-Chu Wu, Laura Rush, Guido Marcucci, and Kenneth K. Chan. Modulation of DNA Methylation by a Sesquiterpene Lactone Parthenolide. Journal of Experimental Therapeutics. 2009, 329(2), 505-514.

5. Ping Chen, Josephine Aimiuwu, Zhiliang Xie, Xiaohui Wei, Shujun Liu, Rebecca Klisovic, Guido Marcucci, and Kenneth, K. Chan. Biochemical Modulation of Aracytidine (Ara-C) Effects by GTI-2040, a Ribonucleotide Reductase Inhibitor, in K562

vii Human Leukemia Cells. American Association Pharmaceutical Scientists Journal (In press).

Research Abstracts:

1. Mitch A. Phelps, Danxin Wang, Josephine E. Aimiuwu, Sherri L. Starrett, Audrey C. Papp, Ahmed A. Ghany, William J. Hicks, Kavitha V. Kosuri, Gregory A. Otterson, Mark A. Socinski, Thomas E. Stinchcombe, Weiqiang Zhao, Larry J. Schaaf, Sanford H. Barsky, Kenneth K. Chan, Wolfgang Sadee, Miguel A. Villalona-Calero. Erlotinib in African Americans with lung cancer: a prospective, clinical, molecular, pharmacokinetic and pharmacogenetic characterization. American Association for Cancer Research, April 2009.

2. Josephine Aimiuwu, Ping Chen, Zhiliang Xie, Zhongfa Liu, Vijay Sarathi, Shujun Liu, Rebecca Klisovic, Guido Marcucci, Kenneth K. Chan. In vitro-in vivo Pharmacodynamic Analysis of GTI-2040 Combined with Ara-C in Acute Myeloid Leukemia. American Association for Cancer Research, April 2009.

3. Ping Chen, Ph.D., William Blum, MD, Josephine Aimiuwu, Vijayasarat Upadhyayula,Ph.D., Zhongfa Liu, Ph.D., Shujun Liu, Ph.D, Jiuxia Pang, Alison Walker, Rebecca B.Klisovic, MD, Ramiro Garzon, MD, Michael R. Grever, MD, Miguel Villalona-Calero, MD, John C. Byrd, MD, Kenneth K Chan, PhD and Guido Marcucci, MD. Quantification of Intracellular Decitabine-Triphosphate with A Novel, Highly Sensitive and Specific LC-MS/MS Assay in Acute Myeloid Leukemia Patients Treated with Low Dose Decitabine. The American Society of Hematology, December 2009.

4. Josephine Aimiuwu, Ping Chen, Zhiliang Xie, Zhongfa Liu Shujun Liu, Rebecca Klisovic, William Blum, Guido Marcucci, Kenneth K. Chan. Development and Characterization of 5-Azacytidine Resistance in Human Cancer Cell Lines. American Association of Pharmaceutical Scientist, November 2009.

5. M. J. Karpenko, Z. Liu, J. Aimiuwu, L. Wang, X. Wu, M. A. Villalona-Calero, D. Young, K. Chan, M. R. Grever, G. A. Otterson. Phase I study of 5-aza-2’- in Combination with Valproic acid in Patients with NSCLC. American Society of Clinical Oncology, March 2008.

viii 6. Zhongfa Liu, Zhiliang Xie, Josephine Aimiuwu, Yonghua Ling, Joseph M. Covey, Kenneth K. Chan. Cytotoxicity and Hypomethylation Activity of 5-Fluoro-2’- deoxycytidine and decitabine on Human Cancer Cell Lines. American Association for Cancer Research, April 2009.

7. Josephine Aimiuwu, Ping Chen et al. 5-Azacytidine, a Possible New Inhibitor of Ribonucleotide Reductase (R2) and its Combined Biological Effect with GTI-2040 in Chronic and Acute Myeloid Leukemia Cells. American Association for Cancer Research, April 2008.

8. Liu, Z, Liu, S, Xie Z, Li C, Aimiuwu J, Chen P, Marcucci G and Chan K. NF- kappaB inhibitor, Parthenolide mediates DNA Demethylation and Histone Acetylation in Leukemia Cells. American Association for Cancer Research, April 2008.

9. Chen P, Liu Z, Aimiuwu J, Xie Z, Liu S, Marcucci G and Chan K.K. A LC- MS/MS Method for the Analysis of Intracellular Nucleoside Triphosphate Levels. American Association for Cancer Research, April 2007.

10. Ping Chen, Zhongfa Liu, Xie Z, Aimiuwu J. Biochemical Modulation of Intracellular Nucleoside Triphosphate Levels by GTI-2040, An Inhibitor of Ribonucleotide Reductase in K562 Human Leukemia Cells. American Association of Pharmaceutical Scientist, November 2007.

11. Liu Z, Xie Z, Wu J, Aimiuwu J, et al. Specific and Sensitive LC-MS/MS Method for Accurate Quantification of Gene Specific Methylation (GSM). American Association of Pharmaceutical Scientist, November 2007.

FIELDS OF STUDY

Major Field: Pharmacy

-with studies in preclinical, clinical pharmacokinetics/pharmacodynamics, and drug development and mechanism studies.

ix TABLE OF CONTENTS

ABSTRACT ...... ii

DEDICATION ...... iv

ACKNOWLEDGEMENTS ...... v

VITA ...... vii

LIST OF TABLES ...... xvi

LIST OF FIGURES ...... xvii

CHAPTER 1 BACKGROUND AND INTRODUCTION ...... 1 1.1 Background ...... 1 1.2 Hypothesis...... 2 1.3 Introduction ...... 5 1.3.1 Ribonucleotide reductase, a strategic anticancer therapeutic target ...... 5 1.3.2 Pertubation of DNA and RNA in cancers ...... 7 1.3.3 GTI-2040- a specific inhibitor of RRM2 ...... 9 1.3.4 Nucleoside drug, Aracytidine (Ara-C) ...... 11 1.3.5 Molecular profiling of Acute Myeloid Leukemia (AML) ...... 12 1.3.6 Epigenetic remodeling in cancer ...... 14 1.3.7 Hypomethylating drugs and resistance ...... 16 1.3.7.1 5-Azacytidine (5-azaC) ...... 16 1.3.7.2 Decitabine (DAC) ...... 17 1.4 Specific Aims ...... 18 1.5 Rationale of the project ...... 19 1.5.1 Combined biological effects of 5-azacytidine (5-azaC) and GTI-2040 ...... 19 1.5.2 Development and characterization of 5-azacytidine and decitabine resistant cell lines ...... 20 1.5.3 Combination therapy of GTI-2040 with Aracytidine (Ara-C) ...... 21

CHAPTER 2 x IN VITRO MODULATION OF THE PHARMACOLOGIC EFFECTS OF 5-AZACYTIDINE BY ANTISENSE GTI-2040 IN MV4-11 AND K562 HUMAN LEUKEMIA CELLS ...... 31 2.1 Abstract ...... 31 2.2 Introduction ...... 32 2.3 Materials and Methods ...... 35 2.3.1 Cell culture and drug treatments ...... 35 2.3.2 Determination of intracellular GTI-2040 concentrations by a Hybridization-based ELISA...... 36 2.3.3 Growth inhibition assay ...... 37 2.3.4 Determination of drug synergy ...... 37 2.3.5 Quantification of RRM2 and DNMT1 mRNA levels by real-time RT- PCR ...... 38 2.3.6 Measurement of RRM2 and DNMT1 protein expression by western blot ...... 39 2.3.7 Determination of intracellular dNTP/NTP and DAC triphosphate (DAC-TP) levels ...... 40 2.4 Results ...... 41 2.4.1 Determination of GTI-2040 in human leukemia MV4-11 cells ...... 41 2.4.2 GTI-2040 enhances cytotoxic effects of 5-azaC in combination treatment ...... 42 2.4.3 Synergy between 5-azaC and GTI-2040 ...... 42 2.4.4 GTI-2040 and 5-azaC reduce RRM2 mRNA levels in MV4-11 and K562 cell lines...... 42 2.4.5 GTI-2040 and 5-azaC decrease RRM2 protein expression ...... 43 2.4.6 GTI-2040 and 5-azaC perturb the intracellular ribonucleotide (NTPs) and deoxyribonucleotide (dNTPs) pools in MV4-11 and K562 cells ...... 44 2.4.7 5-AzaC reduces DNMT1 mRNA and protein levels in MV4-11 cells ...... 44 2.4.8 GTI-2040 when combined with 5-azaC reduces DAC-TP levels in MV4-11 cells ...... 45 2.4.9 GTI-2040 in combination with 5-azaC inhibits the reduction of DNMT1 protein level in MV4-11 cells ...... 46 2.5 Discussion ...... 46 2.6 Conclusion ...... 51

CHAPTER 3 IN VIVO MODULATION OF THE PHARMACOLOGIC EFFECTS OF 5- AZACYTIDINE BY ANTISENSE GTI-2040 ...... 64 xi 3.1 Abstract ...... 64 3.2 Introduction ...... 65 3.3 Materials and methods ...... 69 3.3.1 Chemicals ...... 69 3.3.2 Cell culture conditions ...... 69 3.3.3 Xenograft tumor model ...... 70 3.3.4 Treatment of mice with GTI-2040 and 5-azaC ...... 70 3.3.5 Quantification of RRM2 mRNA levels by real-time RT-PCR ...... 71 3.3.6 Measurement of RRM2 protein expression by western blot ...... 72 3.3.7 Tumor growth pharmacodynamic modeling (control group) ...... 73 3.3.8 Tumor growth model for treated groups ...... 74 3.3.9 Computer Software ...... 75 3.3.10 Statistical analysis ...... 75 3.4 Results ...... 75 3.4.1 Anti-tumor effect of GTI-2040 and 5-azaC treatments ...... 75 3.4.2 Reduction in tumor weight after treatment ...... 76 3.4.3 Measurement of mouse body weight ...... 77 3.4.4 GTI-2040 and 5-azaC decrease RRM2 mRNA and protein levels in xenograft tumor tissues ...... 77 3.4.5 Tumor growth model in control animals ...... 78 3.4.6 Molecular pharmacodynamics and tumor growth model in the combination treatment of GTI-2040 with 5-azaC ...... 79 3.5 Discussion ...... 79 3.6 Conclusion ...... 82

CHAPTER 4 RIBONUCLEOTIDE REDUCTASE IS A NOVEL TARGET OF 5-AZACYTIDINE IN VITRO AND IN VIVO ...... 93 4.1 Abstract ...... 93 4.2 Introduction ...... 94 4.3 Materials and Methods ...... 99 4.3.1 Chemicals ...... 99 4.3.2 Instrumentation ...... 99 4.3.3 HPLC and MS conditions ...... 100 4.3.4 Cell culture and treatments ...... 100 4.3.5 Western blot analysis ...... 101 4.3.6 Determination of intracellular dNTP/NTP pools, DAC and DAC-TP levels ...... 102 4.3.7 RNA Isolation and RT-qPCR ...... 103 4.3.8 mRNA stability assay ...... 104 xii 4.3.9 RNA reduction and hydrolysis...... 105 4.3.10 RNA reduction and hydrolysis...... 105 4.4 Results ...... 106 4.4.1 5-AzaC decreases RRM2 protein level in leukemia cells ...... 106 4.4.2 5-AzaC decreases RRM2 mRNA levels in leukemia cells ...... 107 4.4.3 5-azaC perturbs the intracellular ribonucleotide (NTPs) and deoxyribonucleotide (dNTPs) pools in leukemia cells ...... 108 4.4.4 5-AzaC destabilizes RRM2 mRNA ...... 108 4.4.5 5-AzaC is incorporated into RNA...... 109 4.4.6 Inhibition of RR expression decreases the levels of DAC and DAC-TP formed following 5-azaC treatment ...... 111 4.4.7 RRM2 expression knockout abolishes 5-azaC mediated down-regulation of DNMTs ...... 112 4.4.8 5-AzaC decreases RRM2 mRNA and protein in MV4-11 xenografts ...... 112 4.4.9 5-AzaC reduces RRM2 mRNA and protein in bone marrow cells ex-vivo ...... 113 4.4.10 5-AzaC treatment induces up-regulation of endogenous miR29b and miR181a ...... 113 4.5 Discussion ...... 114 4.6 Conclusion ...... 120

CHAPTER 5 DEVELOPMENT AND CHARACTERIZATION OF 5-AZACYTIDINE AND DECITABINE RESISTANCE IN HUMAN CANCER CELL LINES ...... 137 5.1 Abstract ...... 137 5.2 Introduction ...... 138 5.3 Materials and methods ...... 140 5.3.1 Cell culture and treatment methods ...... 140 5.3.2 Growth inhibition assay and IC50 determination ...... 141 5.3.3 RNA isolation and RT-qPCR...... 142 5.3.4 Western blot analysis ...... 143 5.3.5 Determination of intracellular dNTP/NTP pools and DAC triphosphate (DAC-TP) levels ...... 143 5.3.6 Global DNA methylation (GDM) analysis ...... 144 5.3.7 Characterization of resistance ...... 145 5.3.8 Cross-resistance of DAC with 5-azaC resistant cell lines ...... 145 5.3.9 Statistical analysis ...... 146 5.4 Results ...... 146 5.4.1 Establishment of 5-azaC and DAC resistant cell lines ...... 146 xiii 5.4.2 Reduced drug sensitivity in resistant cell lines ...... 146 5.4.3 mRNA levels of DNMTs in 5-azaC and DAC resistant cell lines ...... 147 5.4.4 DNMTs’ protein expression levels in resistant cell lines relative to parental cell lines ...... 148 5.4.5 Global DNA methylation (GDM) pattern in resistant cell lines ...... 149 5.4.6 Ribonucleotide (NTP) and Deoxyribonucleotide (dNTP) triphosphate level measurement ...... 150 5.4.7 Reduced DAC-TP accumulation in DAC resistant relative to the parental cell lines ...... 151 5.4.8 Relationship of the DAC and DAC-TP in MV4-11 cells ...... 152 5.4.9 Cross-resistance of DAC with 5-azaC resistant cell lines ...... 152 5.4.10 Cross-resistance of DAC to 5-azaC resistant cell lines reduces DAC-TP accumulation ...... 153 5.5 Discussion ...... 153 5.6 Conclusion ...... 158

CHAPTER 6 BIOCHEMICAL MODULATION OF ARACYTIDINE (ARA-C) EFFECTS BY GTI- 2040, A RIBONUCLEOTIDE REDUCTASE INHIBITOR, IN K562 HUMAN LEUKEMIA CELLS ...... 174 6.1 Abstract ...... 174 6.2 Introduction ...... 175 6.3 Materials and methods ...... 178 6.3.1 Chemicals ...... 178 6.3.2 Cell culture ...... 178 6.3.3 Determination of intracellular GTI-2040 concentrations by a hybridization-based ELISA ...... 179 6.3.4 RNA isolation and RT-qPCR...... 179 6.3.5 Western blot analysis ...... 180 6.3.6 Determination of intracellular dNTP and NTP Pools ...... 181 6.3.7 Growth inhibition assay ...... 182 6.3.8 Determination of intracellular Ara-CTP level in K562 human leukemia cells after treatment with Ara-C alone ...... 183 6.3.9 Statistical analysis ...... 183 6.4 Results ...... 184 6.4.1 Determination of GTI-2040 concentrations in K562human leukemia cells ...... 184 6.4.2 GTI-2040 reduces RRM2 mRNA levels in K562 cells ...... 184 6.4.3 GTI-2040 decreases RRM2 protein expression ...... 184 6.4.4 GTI-2040 perturbs the intracellular ribonucleotide xiv (NTPs) and deoxyribonucleotide (dNTPs) pools in K562 cells ...... 185 6.4.5 Inhibition growth assay ...... 186 6.4.6 Dose-dependent accumulation of Ara-C in K562 cells ...... 186 6.5 Discussion ...... 187 6.6 Conclusion ...... 190

CHAPTER 7 A PHASE II PHARMACOKINETICS AND PHARMACODYNAMICS ANALYSIS OF GTI-2040 AND ARACYTIDINE (ARA-C) IN PATIENTS WITH ACUTE MYELOID LEUKEMIA (AML) ...... 199 7.1 Abstract ...... 199 7.2 Introduction ...... 201 7.3 Materials and methods ...... 204 7.3.1 Drugs administration ...... 204 7.3.2 Clinical trial design ...... 205 7.3.3 Pharmacokinetic sampling and assay...... 205 7.3.4 Pharmacodynamic assessment ...... 206 7.3.5 Correlation of dNTP reduction and Ara-CTP accumulation ...... 206 7.3.6 PK/PD modeling and simulations of GTI-2040 combined with Ara-C on the reduction of dCTP and accumulation of Ara-CTP based on indirect response model ...... 207 7.3.7 Data analysis ...... 209 7.4 Results ...... 210 7.4.1 Plasma Pharmacokinetics of GTI-2040 ...... 210 7.4.2 Intracellular uptake of GTI-2040 in bone marrow mononuclear cells ...... 210 7.4.3 Pharmacodynamic results ...... 211 7.4.4 Disease response ...... 211 7.4.5 Correlations among pharmacokinetics, pharmacodynamics and response ...... 212 7.4.6 PK/PD modeling and simulations of GTI-2040 combined with Ara-C ...... 212 7.5 Discussion ...... 213 7.6 Conclusion ...... 216

CHAPTER 8 CONCLUSIONS AND PERSPECTIVES ...... 232 BIBLOGRAPHY ...... 239

xv LIST OF TABLES

Table Page

Table 2.1 IC50 values, the concentration of drug that inhibit 50% of cells, combination index (CI) and fold of sensitization (inverse of CI) in MV4-11 and K562 cell lines ...... 52 Table 5.1 IC50 values of the parental and resistant cell lines and their resistant Indices ...... 159 Table 5.2 Cross-resistance of 5-azaC resistant cell lines to DAC...... 160 Table 7.1 Dose schedule treatment plan and demographic characteristics for GTI-2040 combined with high dose cytarabine in AML patients ...... 217 Table 7.2 Relevant PK parameters of GTI-2040 in AML patients in pilot PD and phase II groups giving GTI-2040 with 5 mg/kg/day ...... 218 Table 7.3 Pooled PK parameters of GTI-2040 in AML patients...... 219 Table 7.4 Clinical response in patients treated with GTI-2040 in combination with high dose Ara-C ...... 220 Table 7.5 An example of observed and the PK-PD model predicted (A) GTI-2040 plasma concentrations (B) dCTP depletion (C) Ara-CTP accumulation with time in patient 07028-10 ...... 221 Table 7.6 Simulated PK/PD model predicted GTI-2040 plasma concentrations, dCTP reductions and Ara-CTP accumulation with time in patient 07028-10 ...... 222

xvi LIST OF FIGURES

Figure Page

Figure 1.1 Reduction of ribonucleotide to deoxyribonucleotide catalyzed by ribonucleotide reductase (RR) (1) ...... 23 Figure 1.2 Main mechanisms of action of antisense oligonucleotides (2) and the sequence and coding region of GTI-2040 ...... 24 Figure 1.3 Structure and mechanism of action of Aracytidine (Ara-C) ...... 25 Figure 1.4 Schematic representation of acute myeloid leukemia (AML) ...... 26 Figure 1.5 Schematic representations of the various epigenetic mechanisms. A: Methylation of DNA leading to gene silencing. B: Acetylation of histones in chromatin leading to activation of gene transcription. C: RNA-mediated transcriptional and post-transcriptional gene silencing. D: Interrelationship of the three epigenetic mechanisms (3) ...... 27 Figure 1.6 Representation of a typical CpG island of a tumor suppressor gene in a normal and tumor cell. The molecular environment of the cell is changed completely due to dense hypermethylation (4) ...... 28 Figure 1.7 Structures of some nucleoside analogs as examples of hypomethylating agents ...... 29 Figure 1.8 Intracellular mechanisms of 5-azacytidine and decitabine and their hypomethylating effects at low doses (5, 6) ...... 30 Figure 2.1 Diagrammatic rationale for the combination of 5-azaC with GTI-2040 in leukemia cells...... 53 Figure 2.2 Schematic representation of the principles of electroporation ...... 54 Figure 2.3 Intracellular accumulation of GTI-2040 in MV4-11 cells following introduction of GTI-2040 at the indicated concentrations for 24 hour by electroporation...... 55 Figure 2.4 Effect of GTI-2040 on the cytotoxicity of 5-azaC. (A) Pretreatment of MV4-11 cells with 1 and 5 µM GTI-2040 decreased the IC50 of 5-azaC (B) Pretreatment of K562 cells with 0.5 µM GTI-2040 also decreased the IC50 of 5-azaC...... 56 Figure 2.5 (A) Plot of combination index vs GTI-2040 concentration at 50% drug effect level...... 57 Figure 2.6 GTI-2040 reduced RRM2 mRNA expression levels in MV4-11 and K562 cells and no further reduction in RRM2 mRNA expression caused by addition of 5 µM 5-azaC. 5-azaC alone also decreased RRM2 mRNA levels by 40-50%...... 58 Figure 2.7 GTI-2040 decreased RRM2 protein expression (50-60%) in MV4-11 and K562 cells and addition of 5-azaC did not contribute to further reduction xvii in RRM2 protein levels...... 59 Figure 2.8 Perturbation of intracellular dNTP/NTP pools by GTI-2040 and 5-azaC. MV4-11 cells were treated with GTI-2040 at 10 µM in the presence and absence of 5-azaC (A) GTI-2040 reduced the dTTP, dATP and dCTP by 30-60%; and 5-azaC alone decreased dATP and dCTP by 20-30%. (B) GTI-2040 did not change GTP, CTP and UTP pools, but dGTP/ATP increased...... 60 Figure 2.9 Perturbation of intracellular dNTP/NTP pools by GTI and 5-azaC. K562 cells were treated with 10 µM GTI-2040 in the presence and absence of 5-azaC (A) GTI-2040 reduced the dTTP, dATP and dCTP by 30-60%; and 5-azaC alone by 20-40%. (B) GTI-2040 did not change GTP, CTP, UTP and dGTP/ATP pools...... 61 Figure 2.10 (A) 5-AzaC at the indicated concentrations decreased DNMT1 mRNA level in MV411 cells in a dose-dependent manner following 48 hour treatment (B) 5-AzaC at 5 µM depletes DNMT1 protein levels...... 62 Figure 2.11 (A) 5 µM 5-azaC alone in MV4-11 cells resulted in the formation of DAC-TP, but when the cells were pre-treated with 5 or 10 µM GTI-2040, DAC-TP levels significantly decreased. (B) Pretreatment of MV4-11 cells with 5 or 10 µM GTI-2040 followed by 1 and 5 µM 5-azaC did not reduce the DNMT1 protein level, but 1 and 5 µM 5-azaC alone for 4 hour decreased DNMT1 levels by 10-50%...... 63 Figure 3.1 Photograph of athymic nu/nu mice bearing tumor after inoculation with MV4-11 cells 16 days post-treatment ...... 84 Figure 3.2 Tumor volume change as a function of time following various drug treatments...... 85 Figure 3.3 Tumor weight change following various drug treatments...... 86 Figure 3.4 Mouse body weight change as a function of time following various drug treatments...... 87 Figure 3.5 Changes in RRM2 mRNA (A) and protein (B) levels in tumor tissue obtained from MV4-11 engrafted mice following various drug treatments. GTI-2040 combined with 5-azaC reduces RRM2 mRNA level in tumor tissues by about 50%, relative to the saline treated control...... 88 Figure 3.6 Control tumor growth model ...... 89 Figure 3.7 Tumor growth in MV4-11 engrafted mice as a function of time as analyzed by the control tumor growth model. (A) Tumor size changes as a function of time as fitted to the tumor growth model and (B) Tumor weight change as a function of time...... 90 Figure 3.8 Tumor growth model in treated animals ...... 91 Figure 3.9 Simulated tumor growth-time profiles following various drug treatment...... 92 xviii Figure 4.1 Structure of 5-azacytidine...... 122 Figure 4.2 Function of ribonucleotide reductase (RR) ...... 123 Figure 4.3 5-AzaC decreases RRM2 protein expression in MV4-11 cells. (A) Western blot analysis of RRM2 protein following 5-azaC treatment at the indicated concentrations for 24, 48 and 72 hours, respectively. (B) Representative densitometry plots of time- and dose dependent decrease of RRM2 protein expression following 5-azaC treatment in MV4-11 cells...... 124 Figure 4.4 5-AzaC decreases RRM2 protein expression in K562 cells. (A) Western blot analysis of RRM2 protein following 5-azaC treatment at the indicated concentrations for 24, 48 and 72 hours, respectively. (B) Representative densitometry plots of time- and dose dependent decrease of RRM2 protein expression following 5-azaC treatment in K562 cells...... 125 Figure 4.5 (A) Effects of DAC, triapine and GTI-2040 on RRM2 protein expression at the indicated concentrations and times in MV4-11 and K562 cells. (B) Comparative plot of RRM2 protein level in MV4-11 cell line following 5-azaC, triapine and GTI-2040 treatment, respectively...... 126 Figure 4.6 5-AzaC down-regulates RRM2 mRNA. (A) 5-AzaC causes a dose-and time-dependent decline in total RNA level in MV4-11 and K562 cells. (B) 5-AzaC decreases RRM2 mRNA levels at 24 hour in MV4-11 and K562 cells in a dose dependent manner. (C) Changes in RRM2 mRNA levels as a function of time in MV4-11 and K562 cell lines following treatment with 5µM 5-azaC ...... 127 Figure 4.7 5-AzaC decreases intracellular dNTP and NTP pools. 5-azaC dose-dependently reduces levels of dTTP, dATP and dCTP in MV4-11 (A) and K562 cells (B) in a dose dependent manner. 5-AzaC decreases GTP, CTP, UTP and dGTP/ATP in MV4-11 (C) and K562 cells (D) in a dose-dependent manner ...... 128 Figure 4.8 5-AzaC reduces mRNA stability of RRM2. (A) 5-AzaC shortens the half-life (t1/2) of RRM2 mRNA, but not abl mRNA or GAPDH mRNA (B) Blockade of protein synthesis facilitates 5-azaC-induced destabilization of RRM2 ...... 129 Figure 4.9 5-AzaC incorporates into RNA. The extracted ion chromatograms (XICs) for (A) RNA hydrolysate from blank cell sample treated with NaBH4. (B) RNA hydrolysate from 5-azaC treated cell sample followed by treatment with NaBH4. (C) RNA hydrolysate from blank cell sample treated with NaBD4. (D) RNA hydrolysate from 5-azaC treated cell sample followed by treatment with NaBD4 ...... 130

xix Figure 4.10 Knock-down of RRM2 expression prevents (A) DAC and (B) DAC-TP conversion from 5-azaC in MV4-11 cells (C) Stability of DAC over time. MV4-11 cells in 1 mL PBS were heat deactivated for 6 minutes at 100 °C and exposed to 10 µM 5-azaC or DAC at 37°C. DACremained stable during the time course of the experiment and DAC was not formed from 5-azaC due to the inactivation of the by heat. The arrows indicate fresh addition of 5-azaC ...... 131 Figure 4.11 Knock-down of RRM2 expression inhibits 5-azaC-induced down-regulation of DNMT 1 and 3a ...... 132 Figure 4.12 5-AzaC reduces RRM2 in MV4-11 tumor xenografts in mice (A) RRM2 protein was dramatically reduced by 5-azaC in engrafted tumor tissues (B) 5-AzaC reduces RRM2 mRNA level in tumor tissues by about 50% ...... 133 Figure 4.13 5-AzaC treatment reduces RRM2 mRNA and protein expression in bone marrow samples from AML patients...... 134 Figure 4.14 5-AzaC treatment increases the endogenous levels of miR29b and miR181a following treatment with 5-azaC in MV4-11 cells...... 135 Figure 4.15 5-AzaC treatment increases the endogenous levels of miRNA29b and miRNA 181a in primary bone marrow mononuclear cells from patients with AML...... 136 Figure 5.1 Dose-response curves of several parental and DAC and 5-azaC resistant leukemia and solid tumor cell lines treated with DAC and 5-azaC...... 161 Figure 5.2 (A) mRNA levels of DNMTs in 5-azaC resistant cell lines relative to their parental counterparts. (B) DNMTs mRNA levels in DAC resistant relative to their parental cell lines...... 162 Figure 5.3 DNMTs protein expression in parental and 5-azaC resistant cell lines. (A) DNMT1 increased (10-60%) in all of the 5-azaC resistant cell lines relative to the parental counterparts. (B) DNMT 3a increased by 30% in MV4-11/5-azaC but decreased (40-70%) in K562/5-azaC, IGROV1/5-azaC and HCT-15/5-azaC cell lines...... 163 Figure 5.4 DNMTs protein expression in parental and DAC resistant cell lines. (A) DNMT1 increased (20-90%) in both the DAC resistant cell lines relative to the parental cell lines. (B) DNMT3a decreased by 50% (C) DNMT3b increased by 3-fold for MV4-11/DAC but K562/DAC decreased by 30% ...... 164 Figure 5.5 Global DNA methylation (GDM) pattern in parental and resistant cell lines (A) 5-AzaC resistant cell lines: K562/5-azaC showed a 15% increase, with no significant change in the other cell lines. (B) In DAC resistant cell lines, MV4-11/DAC and K562/DAC showed a decrease xx (1.5-2 folds ...... 165 Figure 5.6 Baseline levels of deoxyribonucleotide and ribonucleotide pools (dNTP/NTP) in K562/5-azaC and MV4-11/5-azaC and their parental cells. (A) K562/5-azaC showed a 40% increase in dATP with no change in dTTP and dCTP levels. All NTPs decreased by 2-4 folds (B) MV4-11/5-azaC showed a 30% increase in dATP level, and a 40% decrease in dTTP level, with no change in dCTP levels ...... 166 Figure 5.7 Baseline levels of deoxyribonucleotide and ribonucleotide pools (dNTP/NTP) in IGROV1/5-azaC and HCT-15/5-azaC and their parental cell lines. (A) IGROV1/5-azaC showed a 63% increase in dTTP with no change in the dATP and dCTP levels. NTPs increased by 2-5 folds. (B) HCT-15/5-azaC showed a 40% increase in dCTP levels, with no change in the dTTP and dATP levels. NTPs increased (50%) or remained unchanged, all compared with those of the corresponding parental cell lines. Vertical bars are mean ± of SD in triplicate. **p<0.01, * p<0.05 ...... 167 Figure 5.8 Baseline levels of deoxyribonucleotide and ribonucleotide pools (dNTP/NTP) in MV4-11/DAC and K562/DAC relative to their parental cell lines. (A) In MV4-11/DAC, all dNTPs levels increased by 4-30 fold. GTP and dGTP/ATP increased 2-2.5 fold, while the CTP and UTP remained unchanged (B) K562/DAC showed a 60% decrease in dTTP levels, while the dATP levels increased by 40% and dCTP levels was unchanged. dGTP/ATP levels decreased by 30% and GTP, CTP and UTP levels remained unchanged, all relative to the corresponding parental cell lines. Vertical bars are mean ± of SD in triplicate. **p<0.01, * p<0.05 ...... 168 Figure 5.9 DAC-TP and dNTPs levels in K562 and K562/DAC cell lines following DAC treatment. (A) A 2-3 folds decrease in DAC-TP accumulations in the K562/DAC resistant cell lines relative to their parental cells in all time points assessed were seen. (B) dATP and dCTP levels increased by 30-60% and dTTP levels decreased by 50-70%. Vertical bars are mean ± of SD in triplicate **p<0.01 *p<0.05 ...... 169 Figure 5.10 DAC-TP and dNTPs levels in MV4-11 and MV4-11/DAC cell lines following DAC treatment (A) A 50% decrease in DAC-TP accumulation in the MV4-11/DAC resistant cell lines relative to their parental cells at 4 hours was seen. (B) All dNTPs increased dramatically (2-30 folds). Vertical bars are mean ± of SD in triplicate **p<0.01 *p<0.05 ...... 170 Figure 5.11 DAC levels and DAC-TP accumulation in MV4-11...... 171 Figure 5.12 Dose-response curves for 5-azaC resistant cell lines for cross-resistance to DAC ...... 172 Figure 5.13 DAC-TP levels in parental and 5-azaC resistant cell lines following xxi treatment with DAC...... 173 Figure 6.1 Structures of GTI-2040 and cytarabine (Ara-C) ...... 191 Figure 6.2 Diagrammatic rationales for the combination of GTI-2040 with Ara-C in leukemia cells...... 192 Figure 6.3 Intracellular accumulation of GTI-2040 in K562 cells following introduction of GTI-2040 at the indicated concentrations for 24 hours by electroporation...... 193 Figure 6.4 GTI-2040 reduces RRM2 mRNA expression in K562 cells alone...... 194 Figure 6.5 GTI-2040 decreases RRM2 protein expression in K562 cells. RRM2 protein levels were decreased by about 50% following GTI-2040 treatment at 10 µM ...... 195 Figure 6.6 Perturbation of intracellular dNTP/NTP pools by GTI-2040 and Ara-C. (A) GTI-2040 at 10 µM reduced levels of dTTP, dATP and dCTP by 40%; whereas Ara-C showed no effect; (B) GTI-2040 did not change the NTP pools; however, Ara-C at 10 and 20 µM decreased GTP, CTP, UTP and dGTP/ATP by 40-60%...... 196 Figure 6.7 Effect of GTI-2040 on cytotoxicity of Ara-C in K562 cells. (A) Pretreatment of 5 or 10 μM GTI-2040 via electroporation, decreased the IC50 of Ara-C; (B) Pretreatment of 1 μM Ara-C followed by the GTI-2040 treatment via neophectine also decreased the IC50 of Ara-C (*p<0.05, versus control) ...... 197 Figure 6.8 Dose-dependent accumulation of Ara-CTP following 4 hours treatments in K562 cells. Vertical bars represent mean ± SD from triplicate experiments ...... 198 Figure 7.1 Illustration of a two-step ELISA assay in determination of GTI-2040 in plasma and bone marrow cells (7) ...... 223 Figure 7.2 A two compartment model for pharmacokinetic analysis of GTI-2040 and a simplified PK/PD model of GTI-2040 and Ara-C in AML patients ...... 224 Figure 7.3 Representative WinNonLin data plots for one patient each for pilot and phase II PD concentration-time profile of GTI-2040 in patients with AML after treatment with 5 mg/kg/day ...... 225 Figure 7.4 (A) Pilot PD and phase II PD group concentration-time profiles (B) of GTI-2040 in patients with AML after treatment with 5 mg/kg/day drug as a 96 and 144 h, respectively, continuous infusion using a two compartment infusion model in WinNonLin ...... 226 Figure 7.5 Examples of GTI-2040 intracellular levels in bone marrow cells of AML patients for day 1 and 2 following treatment with GTI-2040...... 227 xxii Figure 7.6 Mean intracellular dNTPs/NTPs levels (A) and mean Ara-CTP (B) levels in bone marrow cells of AML patients in pilot and phase II PD groups (p>0.05) ...... 228 Figure 7.7 The fitted and observed GTI-2040 plasma concentration-time data from Patient 07028-10 using a two-compartment intravenous infusion model ...... 229 Figure 7.8 Composite plots of fitted and observed GTI-2040 concentration-time profiles with % changes in PK-PD model simulated and observed dCTP and Ara-CTP levels for Patient 07028-10 ...... 230 Figure 7.9 Composite plots of simulated GTI-2040 concentration-time profiles with % changes in PK-PD model simulated dCTP depletion and Ara-CTP accumulation...... 231

xxiii CHAPTER 1

BACKGROUND AND INTRODUCTION

1.1 Background

Human cells have the capacity to salvage and for the synthesis of used for DNA synthesis (1). Structurally modified nucleoside analogs can also be utilized but would interfere with normal DNA function and metabolism. Thus, nucleoside analogs of these precursors are a major group of antimetabolite cytotoxic drugs currently used in clinical practice. These antimetabolites are still one of the most important drugs for cancer treatment and their mechanisms of action are primarily due to their inhibition of DNA and RNA synthesis in tumor cells (2). Since 1991, about nine nucleoside analogs have been approved by the

US Food and Drug Administration (FDA) for the treatment of various malignances (1).

Forty-five percent of these new agents were approved since 2004 and several others are still currently being evaluated in clinical trials either alone or in combination with other drugs (1). These recent FDA approvals are indications that the pharmacological efficacies of these drugs are still a dynamic area for the treatment of various diseases, especially cancers.

1 Nucleoside analogs have been primarily used for the treatment of hematologic diseases such as leukemia. Leukemia is a group of blood cancers, which results from the uncontrolled proliferation of white blood cells, thereby inhibiting its functions (3).

Common blood cancers include multiple myeloma, chronic myeloid leukemia (CML), acute lymphocytic leukemia (ALL) and acute myeloid leukemia (AML). AML is still a group of heterogeneous leukemia which is widespread in adults and constitutes about

40% of all leukemia in the western countries (4). Nucleoside analogs are the first class of agents that had noticeable activity and provided better prognosis (5). Recently, other biologics like antisense therapy has emerged as a promising gene-targeting strategy to specifically inhibit gene expressions. Antisense oligonucleotides are short, stretch of

DNA molecules with sequence complementary to their target mRNAs (6).

Oligonucleotides hybridize with its mRNA complements and interfere with transcription and eventually translation of the target proteins (7). Antisense represents a promising genetic intervention strategy for chemotherapy and when compared to the conventional anticancer treatment produces less side effects and higher selectivity (8).

1.2 Hypothesis

5-Azacytidine (5-AzaC) is a analog that has been widely used for the treatment of hematological malignancies, such as AML. It was recently (2004) approved by the FDA for the treatment of Myelodysplastic Syndrome (MDS) (9). 5-AzaC is converted to 5’-azaC monophosphate by the enzyme - kinase, then to its diphosphate (5-Aza-CDP) by monophosphate kinase and finally to its active triphosphate (5-Aza-CTP) by diphosphate kinase (9). Cytotoxicity of 5-azaC is associated

2 with its incorporation into cellular DNA and/or RNA (10). Conversion of 5-azaC deoxynucleotide diphosphates then to decitabine (DAC) diphosphate, and subsequently to

DAC triphosphate occurs through a reduction via ribonucleotide reductase (RR) (9, 11).

The reduction allows 10-20% of 5-azaC to be incorporated into DNA as DAC (12). DAC at lower doses (5-20 mg/m2/day) induces hypomethylating effects resulting in reactivation of silenced tumor suppressor genes or inhibition of oncogenes (9, 13, 14). At higher doses (50-100 mg/m2/day), DAC causes DNA damages and cell deaths (10, 15,

16). About 80-90 % of 5-azaC is incorporated into RNA resulting in alteration in RNA processing, thereby causing inhibition of protein synthesis (9, 17). Therefore, incorporation of 5-azaC into RNA may result in RNA instability and potential up/down regulation of RNA, possibly including certain microRNAs, a class of non-coding RNA discovered in recent years (18).

Acute myeloid leukemia (AML) is a heterogeneous disease that could be characterized at the molecular and cytogenetic level. Because of this, there are various clinical responses to standard chemotherapy (19). In order to treat this malignancy, continuing drug development, especially in the area of combination therapy, is necessary.

One of the widely used antimetabolites for the treatment of AML is aracytidine (Ara-C)

(20, 21). Ara-C needs to be phosphorylated to Ara-C triphosphate (Ara-CTP) by deoxycytidine kinase in cells to compete with dCTP for DNA incorporation. This incorporation causes DNA synthesis inhibition and apoptosis (22). Therefore, if intracellular dNTP levels, especially dCTP, are reduced, an increase in Ara-CTP accumulation in cells is expected.

3 Ribonucleotide reductase (RR) is a highly regulated enzyme that catalyzes the reduction of to their corresponding , which is a rate-limiting step for DNA synthesis and repair (23). Over-expression of RR has been shown to play a role in the critical early events of tumor promotion common in malignant cells (24-26). This will increase the endogenous pool of triphosphate, therefore increasing the rate of DNA synthesis to serve the proliferative nature of malignant cells. Additionally, over-expression of RR is a potential mechanism of chemoresistance to nucleoside analogs (27). The important roles of RR in cell proliferation, malignancy, DNA repair and chemoresistance has made it a target of chemotherapy and cancer drug development (28, 29). Recently, RRM2 has been found to be inhibited by GTI-2040, a 20-mer phosphorothioate oligonucleotide (PS ODN) antisense to the mRNA coding region of RRM2 (27). GTI-2040 in combination with high dose Ara-C was evaluated in a phase I/II clinical trial at our Institution. Through this investigation, significant pharmacokinetic and pharmacodynamic knowledge has been gained on this antisense oligonucleotide.

Based on our preliminary result on the pharmacokinetic analysis and pharmacodynamic effects of GTI-2040 when combined with Ara-C, previous reports on the pharmacologic studies of 5-azaC on its incorporation into cellular DNA and/or RNA and the implication of nucleoside drugs in chemoresistance, we hypothesize that:

1. Inhibition of RRM2 expression by GTI-2040 and subsequent reduction in

deoxyribonucleotides may result in enhanced incorporation of 5-azaC into RNA,

thereby increasing mRNA instability and certain miRNA regulation.

4 2. Identification of potential biomarkers in 5-azaC and DAC induced resistant cell

lines may assist elucidation of mechanisms involved in acquired resistances of

these drugs.

3. Depletion of intracellular dCTP levels induced by GTI-2040 may result in an

increased accumulation of Ara-CTP, following sequential combination of these

two drugs, which in turn will increase the antitumor effect of Ara-C.

In this chapter, I will:

1. Discuss ribonucleotide reductase (RR) as a strategic anticancer therapeutic target and its association with DNA and RNA imbalance.

2. Discuss molecular profiling of AML.

3. Review mechanisms of antisense activity of GTI-2040 and its pharmacokinetic and pharmacodynamic (PK/PD) analysis.

4. Discuss assessment of the relationship between epigenetics and cancer.

5. Appraise mechanisms of action of antimetabolite drugs (5-azacytidine, ara-C and decitabine) and their implication in drug resistance. In addition, specific aims of these projects will be proposed and rationale for these projects will be provided.

1.3 Introduction

1.3.1 Ribonucleotide reductase, a strategic anticancer therapeutic target

Ribonucleotide reductase (RR) plays a unique role in nucleotide metabolism and is a highly regulated enzyme that provides the four deoxyribonucleotides required for

5 DNA synthesis (30, 31). RR catalyzes the reduction of all 5’-diphosphates

(ADP, GDP, UDP and CDP or NTPs) to their corresponding deoxyribonucleotides

(dADP, dGDP, dUDP and dCDP or dNTPs), and this is the rate-limiting step during

DNA synthesis and repair (23). RR is composed of two dimeric subunits RRM1 and

RRM2. They are encoded by different genes and their expression is required for the enzyme activity (31). RRM1 protein is constitutively active and stable throughout the cell cycle, whereas RRM2 is only expressed in the late G1 and early S phase. RRM1/RRM2 complex serves as a major provider of dNTPs for DNA replication during S phase (26,

32). A p53-inducible small RR subunit, p53M2, has also been discovered (33, 34). It complexes with RRM1 in non-proliferating cells and provides dNTPs for p53-dependent

DNA repair and mitochondrial DNA synthesis (35).

Over-expression of RR is commonly found in malignant cells (24-26). It increases the endogenous pool of dNTPs, therefore increasing the rate of DNA synthesis to serve the proliferative nature of malignant cells. Studies have demonstrated that inhibition of

RR leads to a depletion of nucleotide pools and arrests of the cell cycle in S phase, thus providing antiproliferative and antineoplastic benefits (26, 32). However, several established RR inhibitors including hydroxyurea, gemcitabine and fludarabine are associated with considerable limitations. For example, the use of hydroxyurea, one of the most commonly used RR inhibitors, is limited by its low affinity for RR, high hydrophilicity, short half-life and early development of resistance (23, 36). Other RRM2 inhibitors, such as antisense GTI-2040 and triapine are still in development (37, 38).

Consequently, there is still a need to develop new RR inhibitors or to employ

6 combination therapy strategies in the use of existing RRM2 inhibitors to provide better clinical outcomes.

1.3.2 Perturbation of DNA and RNA in cancers

Cancers, like most human diseases, result from inability of cells to maintain DNA levels or gene mutations (36, 39). Maintenance of accurate DNA levels as determined by concentrations of deoxyribonucleotides (dNTPs) at replication sites is critical, since the balance of the pools is a determinant of both the rate and fidelity of DNA replication

(39). RR is responsible for the production of deoxyribonucleotides from their corresponding . Enzyme activity of RR directly relates to the production of dNTPs in cells because it ensures efficient and yet accurate genome replication (40). A substantial amount of dNTPs are required by cells in the S phase of the cell cycle and so closely linked with the growth control mechanisms of cells. Consequently, perturbed dNTP pools by RR inhibitors can cause a disturbance in the DNA synthesis and replication processes, thereby resulting in mutagenesis or chromosomal instability (26).

Furthermore, over-expression of RRM2 closely correlates with total dNTP pool expansion and therefore drug resistance (41). Therefore, in the in vitro studies with GTI-

2040 or other RR inhibitors alone or in combination with other drugs, there is a need to monitor pre- and posttreatment intracellular levels of dNTPs. This information becomes critical in establishing status of cell damage effect. RR activity may also be mediated by

RNA interference (42). The activity of RR cooperates with a variety of oncogenes like v- fms, v-src, a-raf and c-myc to act as determinants in enhancing tumor progression (42).

For instance, c-myc, which promotes cell cycle progression, has notable inducible

7 expression caused by RR and this can directly be linked to RNA interference, acting as negative regulators of gene expression.

MicroRNAs (miRNAs) are endogenously expressed small regulatory non-protein coding that were discovered recently. They negatively regulate target mRNAs and can down-regulate various genes by translational repression, when in partial complementarities to the 3’ untranslated regions (3’UTR) of the target mRNA (43).

Clearly, mechanisms that lead to genomic destabilization are critical for cancer transformation and progression; therefore, there is a need to develop therapeutic agents that can disrupt normal DNA and/or RNA processes (44). It has been reported that down-regulation of genes could involve both on-target and off-target effect and the latter has been shown to be highly associated with microRNAs (45). Therefore, the mechanism of mRNA stability located in the untranslated region (UTR) and the modulation of RRM2 mRNA may also be impacted by miRNA-like targeting (46). MicroRNAs are also directly involved in tumorigenesis, such as leukemia and lymphomas (47). For example, miR-15a/miR-16 is frequently deleted or down-regulated in patients with B-cell chronic lymphocytic leukemia (CLL) and multiple myeloma (MM). This down-regulation contributes to malignant transformations through BCL-2 up-regulation hence inhibition of apoptosis (18, 47). Therefore, understanding the functions of miRNAs should provide new insights into the molecular basis of cancer and serve as new biomarkers for cancer therapy and diagnostics.

8 1.3.3 GTI-2040 – a specific inhibitor of RRM2

Antisense oligonucleotides are short, synthetic DNA molecules with sequences that are complementary to their target mRNAs. Once hybridized to their mRNA complement, it can interfere with transcription and hence translation of the mRNA to the target proteins (48, 49). Antisense inhibition involves physical blockade of the ribosomal machinery and/or recruitment of RNase H (ribonuclease) that cleaves mRNA at the hetero-duplex site (6). GTI-2040 is a 20-mer phosphorothioate oligonucleotide that is complementary to the coding region of RRM2 mRNA with sequence of 5’-

GGCTAAATCGCTCCACCAAG-3’. As with other antisense, GTI-2040 needs to penetrate into the targeted cells, and the uptake occurs through active transport that depends on concentration of antisense in the cells (50, 51). In vitro, GTI-2040 is introduced into cells by either transfecting agent or with an electroporation device (32) that permits charged or uncharged oligonucleotides to enter cells rapidly (52, 53). GTI-

2040 is designed to specifically bind to RRM2 mRNA sequence and has been shown to cause degradation of both the mRNA and protein levels in several human leukemia cells, human T4 bladder cancer cells and murine L cells (31, 32).

Antitumor activity of GTI-2040 has been demonstrated. In nude and SCID mice,

GTI-2040 was tested against a wide range of human xenograft tumors such as leukemia, colon, pancreatic, and breast cancers (7). GTI-2040 causes a decrease in tumor growth, metastasis and increase in animal survival in a dose-dependent manner (32). GTI-2040 also decreased the RRM2 levels in tumors isolated from mice treated with GTI-2040, which supports its function in-vivo. Rats and monkeys were used to characterize preclinical pharmacokinetic studies of GTI-2040. In monkeys dosed with GTI-2040 at 2,

9 10 or 50 mg/kg/day the plasma half-life appeared to reach steady state by 8 hours. For the 10 and 50 mg/kg GTI-2040 dose groups, the maximum plasma concentration (Cmax) was 2.94 and 23.6 µmol/mL, plasma steady-state concentration (Css) was 1.88 and 14.3

µmol/mL and clearance (CL) 250 and 160 mL/hr/kg, respectively, (54). Analysis of circulating GTI-2040 levels in rats and monkeys demonstrate that stability, metabolism and excretion were similar with other class of phosphorothioate oligonucleotides (32, 55).

Based on the preclinical studies and the phase I study using GTI-2040 alone for the treatment of advanced solid tumors with manageable toxicity, the therapeutic potential of GTI-2040 combined with aracytidine (Ara-C) on refractory or relapsed AML was evaluated in a phase I clinical trial at The Ohio State University. The study was based on the hypothesis that down-regulation of RRM2 by GTI-2040 will result in a decrease in the dCTP pool that may favor higher Ara-CTP accumulation and thereby induce more apoptosis. GTI-2040 was given at variable doses (3.5 to 7 mg/kg/day) as a long-term continuous IV infusion from days 1 to 7 and Ara-C (1.5 to 3 g/m2/dose) infused over days 2 to 7. The pharmacokinetic (PK) and pharmacodynamic (PD) analysis was performed. A specific picomolar hybridization-based ELISA assay developed in our laboratory was used to determine the concentrations of GTI-2040 in plasma and in bone marrow cells (56). The PK analysis showed that GTI-2040 reached a plasma steady-state concentration (Css) within 4 hour and remained stable till the end of infusion. The clearance (CL) was 13.9 L/hr, the half-lives, t½α and t½β were 0.71 and 31.1 hr, respectively (27). GTI-2040 also found to decrease RRM2 protein expression in about

40% of the patients treated and these were the patient population that achieved complete remission (CR). However, these patients that achieved CR had a higher pretreatment

10 levels of RRM2 expression. This suggest that baseline levels of RRM2 may serve as a possible predictor for clinical response in GTI-2040 acting as a marker for treatment stratification of GTI-2040 with Ara-C therapy in patients with relapsed/refractory disease

(27).

Based on the encouraging result of the phase I trial of GTI-2040 and Ara-C especially in younger patients with refractory/relapsed AML, a phase II clinical trial for this combination is currently being evaluated. Several biomarkers such as RRM2 mRNA and protein, intracellular GTI-2040 drug levels, dNTP/NTP pools, and the intracellular accumulation of Ara-CTP levels that are important biological endpoints are being assessed in addition to clinical outcomes. As RRM2 plays an important role in cell proliferation and its increased activity causes cancer metastasis (57), the development of better strategies to down-regulate RRM2 expression is necessary. This gives us the basis to exploit the combination of GTI-2040 with other nucleoside drugs.

1.3.4 Nucleoside drug, Aracytidine (Ara-C)

Aracytidine (Ara-C) is widely used for the first line treatment of AML. Ara-C is taken up into cells via nucleoside transporters, such as human equilibrative 1 (hENT1), which becomes saturated at >10 µM concentration and above this concentration the uptake is via passive diffusion (58, 59). Once inside the cell, Ara-C is sequentially phosphorylated to its active metabolite; Ara-CTP, via Ara-CMP and Ara-

CDP by deoxycytidine kinase, deoxycytidine monophosphate kinase and nucleotide- diphosphate kinase, respectively. The activated Ara-CTP competes with endogenous deoxycytidine triphosphate (dCTP) for incorporation into DNA by DNA polymerase (27,

11 60). Ara-CTP incorporation into DNA inhibits the DNA polymerases resulting in termination of DNA strand elongation, thus DNA synthesis and resulting in apoptosis

(60-62). A direct correlation between intracellular levels of Ara-CTP and antileukemic effect exists (3, 28) and different strategies to increase the levels of the active metabolite are being investigated. In the combination of RRM2 inhibitors with Ara-C, the assessment and correlation among important pharmacodynamic endpoints (RRM2 mRNA/protein, dNTPs and Ara-CTP levels) are critical. The assessment of these biomarkers will provide evidence of cooperative activity of the combination and ultimately provide insight into clinical outcome.

1.3.5 Molecular profiling of Acute Myeloid Leukemia (AML)

Human acute myeloid leukemia is a hematopoietic malignancy of the myeloid lineage characterized by clonal expansion of immature myeloid cells in the bone marrow

(BM) (19, 32). The uncontrolled accumulation of leukemic cells blocks the production of normal marrow cells leading to deficiency of red blood cells (anemia), platelets

(thrombocytopenia) and normal white blood cells (3). More than half of the de novo

AML cases exhibit cytogenetic abnormalities that result in chromosomal translocations and inversions that cause gene fusion (19). These non-random abnormalities occur in approximately 55% of adult AML patients and have been recognized as one of the most important independent prognostic indicator for achieving complete remission (CR) after intense chemotherapy (63). For instance, translocations in v(16) and t(8;21) results in a chimeric fusion proteins CBFB/MYH11 and AML/ETO, respectively, and these characterize AML with favorable prognosis (64-67).

12 Development of AML also indicates the contribution of aberrant DNA methylation that arises as a result of epigenetic alterations of the genome without changes in the DNA sequence. DNA methylation and histone methylation/acetylation are the common epigenetic changes that are contributing factors to cancer initiation and progression by causing inactivation of gene expression (68, 69). Progression of AML arises by the inactivation of gene expression like tumor suppressor genes and activation of oncogenes (70). In AML, genes, such as p15, MDR1, ER and HICI, are frequently inactivated by methylation (71, 72). Studies have shown that 15 out of 20 AML patients showed aberrant methylation in two or more cancer-related promoters, suggesting AML might be differentiated by a deregulation of CpG island methylation (19). Knowledge of these epigenetic changes provides new insight into the molecular biology of AML towards developing relevant epigenetic targets for anti-cancer drug design (19, 73).

The management of AML is complex with approximately one half of treated patients achieving a long-term remission, even when patients receive chemotherapy as soon as diagnosed (4). Current treatment of AML is even more challenging in adult patients, with about 75% of AML patients older than 60 years exhibiting significant drug toxicity with overall poor response and survival rates. Overall, about 20% to 30% of

AML patients never achieve CR, and among the patients who achieve CR, more than

50% eventually have disease relapse (4, 74). Although there has been significant progress made in the treatment of AML, improvement, especially in the area of combination therapy, is still much needed.

13 1.3.6 Epigenetic remodeling in cancer

Epigenetics deals with biochemical alterations in DNA and chromatin that do not involve changes in the DNA coding sequence itself (9). The interpretation of genetic information coded within the DNA is regulated by mechanisms that involve stable and heritable modifications of DNA and histones. The major modification that occurs in cells is DNA methylation, histone acetylation/methylation and RNA-associated silencing which are known to interact with each other (69). These modifications result in changes in gene expressions that involve the silencing of some tumor suppressor genes and the reactivation of certain oncogenes in cells (46, 69). Epigenetic changes are involved in almost all the different stages of cancer development and progression (75). However, these changes are almost completely reversible, making it a promising therapeutic targeting area.

DNA methylation involves the enzymatic addition of a methyl group to the carbon-5 position of cytosine residues that is followed by a (CpG dinucleotides or islands). CpG islands (0.5-5kb DNA fragments), often unmethylated during normal development, are needed for processes like genomic imprinting and X chromosome inactivation (68, 76). However, in cancer cells, hypermethylation in this region becomes aberrant and causes demethylation of the other sites (68). Hypermethylation is associated with leukemia and other hematologic malignancies that involve genes, such as p15, p21,

ER and MDR. The p15 gene has been demonstrated to be hypermethylated in about 65% of myelodysplastic syndromes (MDS) (72) and at diagnosis was associated with lower survival, transformation and progression into AML (77). This may suggest the role of

14 p15 as a marker of leukemic transformations (4, 78). In mammals, DNA methylation occurs via DNA methyltransferases (DNMTs 1, 3a and 3b) at the 5’ position of the cytosine residues through an enzymatic reaction which uses S-adenosyl-methionine as a methyl group donor (79). Epigenetic processes have been demonstrated to be reversible by treatment with DNA methylation inhibitors, such as 5-azacytidine and decitabine.

These have proven to be effective in restoring gene expression and normal patterns of differentiation and apoptosis in malignant cells (80, 81).

Histone acetylation of lysine residues carried out by histone aceyltransferases

(HATs) on histones 3 and 4 results in an active or open chromatin structure. This allows various transcription factors to access the promoters of target genes, hence occurrence of transcription (81). On the contrary, de-acetylation of lysine residues via histone deacetyltransferases (HDACs) removes acetyl groups from the histone tail giving a compact or closed chromatin structure that inactivates gene expression (46, 81). DNA methylation changes and histone modifications interact with each other and directly with

DNA or histone tails. In addition, RNA-associated gene silencing is another epigenetic mechanism that has recently attracted attention (82). RNA in forms, such as antisense transcripts, non-coding RNAs or RNA interference (RNAi), can also induce transcriptional silencing of gene expression by facilitating histone modification and DNA methylation (69, 82). These changes play an important role in the process of cellular differentiation and allow cells to maintain different characteristics, despite containing the same genomic material.

15 1.3.7 Hypomethylating drugs and resistance

Epigenetic changes to chromatin structure are crucial for cell development and differentiation (68). Exposure to DNA hypomethylating agent can alter the patterns of gene expression without changing the coding sequence of those genes or their transcriptional control elements. Hypomethylating agents at low doses (83) primarily involve the processing of RNA and initiation of transcriptional silencing by hypermethylation. Silencing induced by DNA methylation and/or histone deacetylation can be reversed via pharmacologic inhibition of DNA methyltransferases by 5-azaC and

DAC at low-doses (13). Epigenetic changes can also be a crucial factor in the acquisition of drug resistance, especially in most cytotoxic drugs like 5-azaC and DAC that have low therapeutic index (75). Studies have shown that drug resistant cell line models have multiple changes in methylation regulation following drug exposure (84). Defects in the methylation pattern may therefore contribute to genetic instability and constitute one of the mechanisms of 5-azaC and DAC resistance (85-87) and eventually clinical treatment failure (75). Consequently, the identification and evaluation of surrogate endpoints more specific to drug resistance may be more informative in determining response and overall survival of patients.

1.3.7.1 5-Azacytidine (5-AzaC)

5-Azacytidine is a hypomethylating drug that has recently been approved for the treatment of Myelodysplastic Syndrome (MDS) (88). 5-AzaC is a cytidine analog, in which the position five of the cytidine ring has been replaced by nitrogen and this site represents the position of cytosine methylation in genomic DNA (89). 5-AzaC is taken up

16 into cells via nucleoside transporters, especially human equilibrative nucleoside transporter 1 (hENT1), and is inactive in its original form but converted to its phosphorylated active metabolite 5-Aza-CTP by cytidine kinases. The biological activity of 5-azaC is associated with the incorporation into cellular DNA via prior conversion to decitabine, and/or RNA with subsequent sequestration of DNMTs via covalent bond formation (10). Therefore, 5-azaC can inhibit DNA, RNA and protein synthesis (13). 5-

AzaC was synthesized 45 years ago and has widely been used for the treatment of hematological malignancies, including acute myeloid leukemia (AML) (9). 5-AzaC causes hypomethylation via its DNA incorporation (10-20%) by inhibiting DNA methyltransferases at low-doses (1-10 µM in leukemia MV4-11 and K562 cells in-vitro or 10-75 mg/m2/day in the clinic) (9, 10, 90). However, about 80-90% of 5-azaC incorporates into RNA causing RNA destabilization that interferes with transfer RNA and ribosomal RNA processing (9). The cytotoxic effect of 5-azaC is considered to be primarily due to its incorporation into RNA, with resultant interference in the synthesis of nucleic acids and proteins. Despite decades of efforts made to delineate the mechanisms of action of 5-azaC in terms of its interference with RNA metabolism (91), the precise basis of its clinical efficacy still remains uncertain (10). Therefore, we decide to further investigate the effects of 5-azaC via its RNA incorporation.

1.3.7.2 Decitabine (DAC)

Decitabine is a deoxycytidine analog with the carbon-5 position in the pyrimidine ring replaced by nitrogen (92). Decitabine is phosphorylated by kinase to its active form

DAC-TP. Decitabine can reverse the epigenetic silencing of normal genes common in

17 solid tumor and hematologic malignances by the pharmacologic inhibition of DNMTs

(93). Previous reports show that DAC at low doses (0.3-2.5 µM in vitro or 5-20 mg/m2/day in the clinic) can cause a sustained DNMTs enzyme depletion by the covalent binding of the DNMTs with the DAC moieties and therefore restore normal pattern of gene expression. At higher doses (50-100 mg/m2/day), DAC is primarily a cytotoxic

(DNA damaging) agent (94). At these low-doses, complete remissions have been achieved with relatively low toxicity in patients with myeloid malignances, such as acute myeloid leukemia (AML), myelodysplastic syndrome (MDS), and chronic myelogenous leukemia (CML) (95). The pharmacological activity of DAC is associated with several factors. Firstly, the continuous trend for a decrease of DNMTs expression observed following DAC treatment supports the notion that DNMTs enzymes are primary targets and serve as useful biomarkers. Secondly, the determination of DAC-TP inside the cell following DAC treatment is critical for establishing drug-effect relationship.

1.4 Specific Aims

A.1. To evaluate the pharmacodynamic activities of low-dose 5-azacytidine when combined with ribonucleotide reductase (RRM2) antisense GTI-2040 under in vitro and vivo conditions.

1. To study the in-vitro pharmacologic effects of 5-azaC when combined with GTI-

2040 in leukemia cell lines (K562 and MV4-11) and this will allow identification

of synergistic/additive effect and therefore provide evidence for their potential

combined use in the clinic.

18 2. To evaluate the antitumor effect of the combination of 5-azaC with GTI-2040 in-

vivo and consequently determine the optimal dosing regimen for future clinical

application.

3. To evaluate the modulation of RRM2 expression by 5-azaC which has not been

previously been investigated. This aims at further clarification of mechanism of

action of 5-azaC and will serve as a novel target for 5-azaC.

A.2. To develop and characterize 5-azaC and DAC induced resistant leukemia and solid tumor cell lines. This will allow identification of biomarkers that may be responsible for mechanisms of nucleoside drug resistance.

A.3. To evaluate the in vitro-in vivo pharmacodynamic analysis of GTI-2040 combined with Ara-C and also characterize the clinical pharmacokinetic of GTI-2040 which will assist in the exploration of their PK/PD correlations and their relationship with clinical response.

1.5 Rationale of the project

1.5.1 Combined biological effects of 5-Azacytidine (5-AzaC) and GTI-2040

The proposed study is rationally designed based on the premise that the major mechanism of action of 5-azaC appears to be due to its RNA effect and that only 80-90% of the drug is incorporated into RNA. This RNA effect includes disruption of RNA elongation (9, 10) and protein synthesis inhibition. The remaining 10-20% of 5-azaC is converted to DAC via RR. To enhance its RNA effect, we seek to mechanistically enhance the incorporation of 5-azaC into RNA by inhibiting RR expression that converts

19 5-azaCDP to DAC with GTI-2040 (specific inhibitor of RRM2), thereby enhancing the cytotoxic effects of 5-azaC. Furthermore, in terms of cell cycle activity, 5-azaC has been shown to induce G2/M-phase arrest and GTI-2040 inhibits RRM2 protein expression during G1/early S phase (96). Therefore, this combination may produce synergy and offers the potential use of low dose 5-azaC still invoking markedly increased rates of apoptosis. In addition, the strategic increase in the incorporation of 5-azaC into RNA may be used to establish a link between RNA interference and certain miRNA perturbation.

Since RRM2 plays a critical role in cell proliferation and its increased activity causes cancer metastasis, this combination study will utilize the specific role of GTI-2040 to inhibit RRM2, and subsequent perturbation of dNTP/NTP pools will thereby simultaneously increase the cytotoxic effects of 5-azaC at low dose. This combination strategy will produce optimal dose balance of both GTI-2040 and 5-azaC for desired drug effects and ultimately enables the efficient and effective translation of preclinical findings into clinical evaluation.

1.5.2 Development and characterization of 5-Azacytidine and Decitabine resistance cell lines

Development of resistance to antitumor nucleosides analogs, such as 5-azaC and

DAC, remain a major problem during chemotherapy (89). Resistance to anticancer drugs results from a variety of factors that arises from drug induction mechanisms (97-99). 5-

AzaC and DAC are both hypomethylating agents, and defects in the DNA methylation pattern might contribute to the origin of resistance by causing genomic instability and therefore elevated mutations (85-87, 100). 5-AzaC and DAC are taken up into cells by

20 hENT1 and subsequently phosphorylated to their active triphosphate form by kinases.

Inefficient cellular uptake in cells and insufficient intracellular triphosphate levels of

DAC and 5-azaC may therefore contribute to the development of resistance (12, 101).

Drug resistance to DAC can also arise due to an enhanced level of dCTP, which will compete with DAC-TP for incorporation into DNA and which also reduces the phosphorylation of DAC by deoxycytidine kinase, since dCTP is a feedback inhibitor of this enzyme (15). Another potential mechanism of resistance for nucleoside analog is the over-expression of RRM2 that leads to an increase in the endogenous dNTP/NTP pools

(102-106). Therefore, we will examine in vitro models of induced 5-azaC and DAC resistant cancer cell lines to identify biomarkers that will facilitate the understanding of the mechanisms of acquired drug resistance. This information may help the understanding of the development of resistance of nucleoside drugs in ongoing clinical studies.

1.5.3 Combination therapy of GTI-2040 with Aracytidine (Ara-C)

In the clinic, combination therapy has proven to have a great advantage over the use of a single agent. We hypothesized that using a specific inhibitor of RRM2 expression, such as GTI-2040, will cause a reduction in intracellular dNTP pools, especially dCTP and when combining with Ara-C, the depletion of dCTP will thereby reduce the competition with Ara-CTP for DNA incorporation, hence enhancing apoptosis. Thus, based on this rationale, a phase I clinical trial was conducted using GTI-

2040 in combination with Ara-C in refractory/relapsed AML patients at The Ohio State

University. The results were encouraging and a phase II clinical trial for this combination

21 was initiated and is about to complete. In order to provide a full validation of this combination strategy, evaluation of several pharmacodynamic biomarkers, such as

RRM2 mRNA and protein expression levels was sought. In the preclinical evaluation, in vitro-in-vivo intracellular drug levels, dNTP/NTP pools, and intracellular accumulation of

Ara-CTP levels will be analyzed from leukemia cells treated with this combination and from bone marrow samples from AML patients. Clinical pharmacokinetics (PK) of GTI-

2040 will be characterized. Finally, correlation of the PK-PD endpoints will be explored.

Analysis of the endpoints will also provide initial parameters for PK/PD modeling and simulation that will eventually provide insight into clinical outcomes.

22

Ribonucleotides Deoxyribonucleotides

RR

DNA

Figure 1.1 Reduction of ribonucleotide to deoxyribonucleotide catalyzed by ribonucleotide reductase (RR) (107).

23

A Hum (R2) mRNA Hum (RRM2) mRNA 194 CODING 1364 2475 5’UTR 3’UTR Poly A

626 645 3’ GAACCACCTCGCTAAATCGG 5’ GTI-2040

Figure 1.2 Main mechanisms of action of antisense oligonucleotides (108) and sequence and coding region of GTI-2040.

24 NH2 NH2 NH2

N N HC N N N HC N O N O N O HO HO HO O O O H H H H H H H H H H H H OH OH OH OH OH cytidine 5-aza-cytidine decitabine N HNH2 2

N Hum (R2) mRNA

194 1364 2475 N O CODING

5’UTR 3’UTR Poly A H O

O

H O H 626 645

H H 3’ GAACCACCTCGCTAAATCGG 5’ GTI -2040 O H cytarabine

Ara-C CDA Deamination dCK

Ara-CMP dCMK Ara-CDP

NDP kinase NH2 O NH2

N HN N

O N O N O N O O O

H O O H O O HO P O P O P O O OH OH OH OH OH OH

OH OH OH Aracytidine (Ara-C ) Arauridine (Ara-U ) Aracytidine 5'-triphosphate (Ara-CTP)

NH NH 2 2 N N

O N O N O O O O O O

HO P O P O P O O HO P O P O P O O OH OH OH Figure 1.3 Structure and mechanismOH OofH actionOH of Aracytidine (Ara-C).

OH OH OH

Deoxycytidine 5'-triphosphate (dCTP) Cytidine 5'-triphosphate (CTP)

25

Normal bone marrow cells AML bone marrow cells

Figure 1.4 Schematic representation of acute myeloid leukemia (AML).

26

Figure 1.5 Schematic representations of the various epigenetic mechanisms. A: Methylation of DNA leading to gene silencing. B: Acetylation of histones in chromatin leading to activation of gene transcription. C: RNA-mediated transcriptional and post- transcriptional gene silencing. D: Interrelationship of the three epigenetic mechanisms (69).

27

Figure 1.6 Representation of a typical CpG island of a tumor suppressor gene in a normal and tumor cell. The molecular environment of the cell is changed completely due to dense hypermethylation (68).

28

NH2 NH2 NH2

N N HC N N N HC N O N O N O

HO HO HO O O O H H H H H H H H H H H H OH OH OH OH OH cytidine 5-aza-cytidine decitabine NH2

NH2 NH2 NH2 N Hum (R2) mRNA N HC N N 194 1364 2475 N N N O CODING

HC 5’UTR 3’UTR Poly A HO N O N O N O O HO HO HO H OH 626 645 O H O H 3’ GAACCACCTCGCTAAATCGG 5’ GTI -2040 O OH H H H H H H cytarabine H H H H H H OH OH OH OH OH cytidine 5-aza-cytidine decitabine NH2

N Figure 1.7 Structures of some nucleosideHum analogs (R2) mRNA as examples of hypomethylating

agents. 194 1364 2475 N O CODING

5’UTR 3’UTR Poly A HO

O

H OH 626 645

H H 3’ GAACCACCTCGCTAAATCGG 5’ GTI -2040 OH cytarabine

29

~80-90%

Figure 1.8 Intracellular mechanisms of 5-azacytidine and decitabine and their hypomethylating effects at low doses (9, 109).

30 CHAPTER 2

IN VITRO MODULATION OF THE PHARMACOLOGIC EFFECTS OF 5- AZACYTIDINE BY ANTISENSE GTI-2040 IN MV4-11 AND K562 HUMAN LEUKEMIA CELLS

2.1 Abstract

5-Azacytidine (5-AzaC) appears to exert its anticancer activity not only through

DNA mechanism but also by direct incorporation into RNA leading to disruption of RNA elongation. In order to mechanistically enhance the incorporation of 5-azaC into RNA that may inhibit protein synthesis and subsequently cause cytotoxicity, we combined 5- azaC with GTI-2040. GTI-2040 is an antisense to the RRM2 subunit of ribonucleotide reductase (RR), which catalyzes conversion of ribonucleotides into deoxyribonucleotides and which is commonly over-expressed in every type of cancer. GTI-2040 was introduced into the cells via electroporation and transfecting agent. MTS cytotoxicity assay was used to determine the synergistic/additive effect of the combination of GTI-

2040 and 5-azaC in K562 and MV4-11 cell lines. Quantitative RT-PCR was used to measure the transcriptional level of the RRM2 and DNA methyltransferase I (DNMT1) and western blot assay was used to evaluate their corresponding protein levels. GTI-2040 intracellular levels were measured by a hybridization-based ELISA assay and a non- radioactive LC-MS/MS method was used to quantify (NTP), deoxynucleotides (dNTP) pools and DAC-TP levels before and after drug treatments.

31 Combination of GTI-2040 (0.5, 1 or 5 μM ) with variable 5-azaC concentrations showed a synergistic effect based on the reduction in IC50 values as compared to 5-azaC or GTI-2040 alone. Treatment with GTI-2040 alone induced down-regulation of RRM2 mRNA and protein after 48 hour exposure and addition of 5-azaC to pretreated GTI-2040 cells did not result in further reduction of the enzyme; however, 5-azaC decreased RRM2 mRNA levels in both MV4-11 and K562 cell lines. GTI-2040 and 5-azaC alone and their combination decreased the dNTP pools with no significant change in the NTP pool, supporting functional inhibition of RRM2. Treatment of MV4-11 cells with 5 µM 5-azaC alone resulted in the formation of 1.04 pmol/106 cells of DAC-TP and decreased (10-

50%) DNMT1 protein expression after 4 hour; however, for the cells pretreated with 5 or

10 µM GTI-2040, DAC-TP levels were reduced (3 to 9-fold) and DNMT1 levels were unchanged. Our data support the combination of 5-azaC at low-dose with GTI-2040 through sensitization effects of GTI-2040. The combination use of 5-azaC with GTI-2040 appeared to potentiate the anti-leukemia activity of the two compounds in vitro and needs to be further explored in vivo.

2.2 Introduction

5-Azacytidine (5-AzaC) has been used effectively for the treatment of hematological malignancies, such as acute myeloid leukemia (110), and recently approved for the treatment of Myelodysplastic Syndrome (MDS) (9). 5-AzaC is converted to 5’-azaC monophosphate (5-Aza-CMP) by the enzyme uridine-cytidine kinase, then to its diphosphate (5-Aza-CDP) by cytosine monophosphate kinase and finally to its active triphosphate form (5-Aza-CTP) by diphosphate kinase (9).

32 Cytotoxicity of high dose 5-azaC is associated with its incorporation into cellular DNA and/or RNA (10). Formation of 5-azaC deoxynucleotide diphosphates to decitabine

(DAC) diphosphate through the reduction by ribonucleotide reductase (RR) and subsequently to DAC triphosphate by diphosphate kinase allows 10-20% of 5-azaC to be incorporated into DNA, resulting in hypomethylating effects (low dose) or DNA damaging effect (high dose). Major interest on 5-azaC has been focused on its purported ability to inhibit DNA methylation (111, 112) via its DNA incorporation (likely through conversion to DAC), athough the RNA incorporation remains very crucial for its cytotoxicity. About 80-90 % of 5-azaC is incorporated into RNA, which can result in increased protein synthesis inhibition and disruption in metabolism, hence an enhanced rate of apoptosis (9, 10). Inhibition of protein synthesis via mRNA destabilization is probably the major cause for the strong cytotoxic activity of the drug

(113). Since this effect can be suppressed by actinomycin D (transcription inhibitor), it may suggest that the incorporation of 5-azaC into RNA is crucial for its inhibition process (114, 115). Majority of the combination studies done so far has been focused on

5-azaC inhibition of DNMTs and HDAC inhibitors; however, these combination strategies have been primarily centered on the DNA pathway of 5-azaC given that the

RNA pathway of 5-azaC is less well understood. Our study will therefore, focus on the combination that exploits the advantages of RNA-mediating 5-azaC effects.

Ribonucleotide reductase (RR) is a highly regulated enzyme that catalyzes the reduction of ribonucleotides to their corresponding deoxyribonucleotides, which is the rate- limiting step for DNA synthesis and repair (23). RR is primarily composed of two

33 dimeric subunits RRM1 and RRM2, in addition to a newly discovered p53-inducible small analog of RRM2, p53M2 (33, 34). RRM1 and RRM2 are encoded by different genes and their expression is required for the enzyme activity (31). RRM1 protein is constitutively active throughout the cell cycle, whereas RRM2 is only expressed in the late G1 and early S phase. Over-expression of RR is commonly found in malignant cells

(24-26). It elevates the endogenous pool of deoxynucleoside triphosphate, therefore increasing the rate of DNA synthesis to serve the proliferative malignant cells. Studies have demonstrated that inhibition of RR leads to a depletion of nucleotide pools and arrest of the cell cycle in S phase, thus providing antiproliferative and antineoplastic benefits (26, 32). RRM2 protein has been found to be inhibited by GTI-2040, a 20-mer phosphorothioate oligonucleotide (PS ODN) with sequence of 5’-

GGCTAAATCGCTCCACCAAG-3’. GTI-2040 was designed to specifically bind to and target the mRNA coding region of the RRM2 (27). Given that RR enables the conversion of 5-azaC to DAC, clearly a combination of GTI-2040 with 5-azaC should allow GTI-

2040 to block RRM2 expression and prevent the formation of DAC and this may oppose the ability of 5-azaC to induce hypomethylation, while enhancing 5-azaC RNA mediating effects.

The purpose of this study was to mechanistically enhance the incorporation of 5- azaC into RNA, using a specific mRNA inhibitor (GTI-2040) to maximize the RNA effect of this nucleoside, allowing investigation of its biochemical mechanism. We therefore propose (Figure 2.1) that, by blocking the conversion of 5-azaC to DAC by targeting RRM2 expression with GTI-2040, it can lead to a depletion of nucleotides pools

34 and which simultaneously, enhance the cytotoxicity of 5-azaC through its RNA incorporation. Therefore, this combination strategy will be used to produce maximal synergy by invoking increased rates of apoptosis with low-dose 5-azaC that will decrease its untoward toxicity. The results from these experiments will provide important information on its involvement in DNA methylation. We believe that this combination will ultimately enable the efficient and effective translation of preclinical findings to clinical evaluation.

2.3 Materials and Methods

2.3.1 Cell culture and drug treatments

Human leukemia cell lines K562 and MV4-11 obtained from American Type

Culture Collection (ATCC, Manassas, VA) were used for these studies. All cells were cultured in RPMI 1640 medium supplemented with L-glutamine (Supplied by Tissue

Culture Shared Resource, Comprehensive Cancer Center, The Ohio State University,

Columbus, OH), 1% Penicillin-Streptomycin (Gibco, Rockville, MD) and 10% fetal bovine serum (FBS) (Invitrogen, Rockville, MD). The cell lines were maintained at 37

°C in a humidified environment with 5% CO2. Viability and cell counts were determined using trypan blue dye exclusion assay (54).

GTI-2040 was introduced into cells using either neophectin transfecting agent or with an electroporation device (BIO-RAD Lab, CA, USA) (Figure 2.2). However, transfecting agent can be toxic to cells and may have potential unknown biological

35 activity (54). Therefore, further studies with GTI-2040 were done using electroporation.

5-AzaC was obtained from The National Cancer Institute (Bethesda, MD).

2.3.2 Determination of intracellular GTI-2040 concentrations by a Hybridization-

based ELISA

3×106 MV4-11 cells were treated with GTI-2040 at 0.1, 0.3, 1.0, 3.0, 10, 30 μM for 24 hours using an electroporation delivery technique. Cells were then harvested and intracellular GTI-2040 levels were measured by a previously developed two-step hybridization-ligation ELISA assay (56). Briefly, GTI-2040 was first base paired to the capture probe in a polypropylene 96-well plate. 10% Triton X-100 was added to the mixture solution and incubated at 42C for 2.5 hours for hybridization. The resulting solution was transferred to a NeutrAvidin-coated 96-well plate (Pierce, IL) and further incubated at 37C for 30 minutes to ensure the attachment of labeled capture probe to NeutrAvidin-coated wells. After washing six times, the ligation solution containing T4 ligase and detection probe was added to each well followed by the addition of S1 nuclease solution. The reaction was blocked with Superblock buffer (Pierce, IL). Anti- digoxigenin-alkaline phosphatase (AP) (Fermentas, MD) was then added into each well.

Following addition of substrate solution (36 mg Attophos [Pierce, IL] in 60 ml diethanolamine buffer), fluorescence intensity was measured at Ex 430/Em 560

(filter=550nm) using a Gemini XS fluorescence microtiter plate reader (Molecular

Devices, Sunnyvale, CA).

36 2.3.3 Growth inhibition assay

To evaluate the growth inhibitory effect of 5-azaC and GTI-2040 on MV4-11 and

K562 leukemia cells, MTS cytotoxicity assay was performed. GTI-2040 was introduced into K562 and MV4-11 cells by using either neophectin transfecting agent or with an electroporation device. Briefly, cells were seeded into 96 well plates and the cells were pretreated with fixed concentrations 0.5, 1, and 5 µM GTI-2040 for 24 hours. Increasing concentrations of 5-azaC at 0.01, 0.1, 1, 10, 100 µM, were then added and the incubation was continued for 72 hours. Afterward, 20 µL of 3-(4,5-dimethylthisazol-2-yl)-5-(3- carboxymethoxyphenyl-2-(4-sulfophenyl)-2H-tetrazolium, inner salt (MTS) and phenazine methosulfate (PMS) (Promega, Madison, WI) were added to each well in a

20:1 ratio. After 2 hours incubation, the absorbance was read at 490 nm on a microplate reader Gemini XS (Molecular Devices, CA) to determine levels of formazan product as a measure of cell viability against a non-drug treated control. Determination of IC50 values was performed using WinNonLin software (Pharsight, Mountain View, CA).

2.3.4 Determination of synergy

The nature of the interaction between the combination of GTI-2040 and 5-azaC was carried out using the combination index (CI) values based on the Chou-Talalay method

(116). Concentration-effect curves for single agents and their combinations were used to determine the amount of each agent, either alone or in combination, needed to achieve a

50% level of effect. The combination index (CI) was calculated as:

IC50 IC CI  A,B  B IC50 A IC50 B

37 where IC50A and IC50B are drug concentrations of GTI-2040 and 5-azaC that produce

50% effect level of cytotoxicity when used alone, while IC50A,B is the concentration that produced the same effect when used in combination and ICB is the fixed concentration of

GTI-2040 used. Isobologram analysis was done using the combination index (CI) values to determine whether the combination was additive, synergistic or antagonistic (116).

2.3.5 Quantification of RRM2 and DNMT1 mRNA levels by real-time RT-PCR

To measure intracellular RRM2 mRNA level, 3×106 K562 and MV4-11cells were first treated with GTI-2040 alone at 1, 5, 10 µM for 24 hours followed by continuous treatment with 5-azaC at a concentration of 5 µM for 48 hours, while the GTI-2040 exposure was maintained. In another study, MV4-11 cells were treated with 5-azaC alone at 1, 5, 10, 20 µM for 48 hours to investigate its hypomethylating effects. Total RNA was isolated using Trizol reagent (Invitrogen, Carlsbad, CA). Briefly, cell lysate was treated with chloroform and the total RNA was precipitated with isopropyl , followed by a washing step with 75% . RNA was then dissolved in RNase free water and its concentration and purity was measured by a Nanodrop spectrophotometer (Nanodrop

Technologies, Wilmington, DE). cDNA was synthesized from 2 µg total RNA using

Moloney murine leukemia virus reverse transcriptase (Invitrogen). The cDNA templates and primers were then mixed with reagents from a SYBR Green PCR Master Mix

(Applied Biosystem, Foster City, CA). Reactions were carried out in triplicate in ABI prism 7700 sequence detector (Applied Biosystems, Foster City, CA), and data were analyzed by comparative CT method. Dissociation curves were also obtained to examine

38 the purity of the amplified products. The amount of RRM2 and DNMT1 mRNA in each sample was normalized to an internal control, abl. The relative changes in treatment groups were expressed as a percentage of untreated control (arbitrarily set at 1). The results were expressed as the mean ± SD from triplicate determinations.

2.3.6 Measurement of RRM2 and DNMT1 protein expression by western blot

5×106 K562 and MV4-11 cells were first treated with GTI-2040 alone at 1, 5, 10

µM for 24 hours followed by 5-azaC continuous exposure at 5 µM for 48 hours. MV4-11 cells were also treated with 5 µM 5-azaC alone for 48 hours to investigate its hypomethylating effects. In another design, 5×106 MV4-11 cells were either treated with

5-azaC alone at 1 and 5 µM for 4 hours or pre-treated with GTI-2040 at 5 and 10 µM for

48 hours, followed by 5-azaC continuous exposure at 1 and 5 µM for an additional 4 hours, while the GTI-2040 exposure was maintained. Cells were then harvested, washed with 1 mL ice-cold PBS, and centrifuged at 1000 g for 5 minutes at 4 ºC. The pellet was obtained and re-suspended in 100 μL lysis buffer (50 mM pH 7.6 Tris-HCl, 250 mM

NaCl, 5 mM EDTA, 2 mM Na3VO4, 50 mM NaF and 1% protease inhibitor cocktail)

(P8340, Sigma) for 30 minutes on ice. The lysate was sonicated for 10 seconds. Total protein concentration was determined using the BCA protein assay method (Pierce,

Rockford, IL). Equal amounts of protein for each sample were incubated with 6x SDS loading buffer (100 mM, pH 6.8 Tris, 200 mM DTT, 4% SDS, 20% glycerol, and 0.015% bromphenol blue) and boiled for 5 minutes. The proteins were then separated on 15%

SDS-polyacrylamide gels and transferred to nitrocellulose membranes (Amersham,

39 Piscataway, NJ). The RRM2 and DNMT1 proteins were recognized following treatment with a goat antihuman RRM2 polyclonal antibody (E-16) (Santa Cruz Biotechnology,

Santa Cruz, CA) and a rabbit antihuman DNMT1 polyclonal antibody (New England

Biolabs, MA), respectively, as the first antibody. This was followed by a HRP

(horseradish peroxidase) conjugated anti-goat IgG and anti-rabbit IgG secondary antibody, respectively. RRM2 (MWT 45,000 dalton) and DNMT1 (MWT 180,000 dalton), respectively, was detected by ECL (Amersham, Arlington Heights, IL) and

GAPDH was used as the internal loading control. RRM2 and DNMT1 protein expressions were quantified by densitometry and normalized to GAPDH.

2.3.7 Determination of intracellular dNTP/NTP and DAC triphosphate (DAC-TP)

levels

10×106 K562 and MV4-11 cells were treated either with 5-azaC alone at 5 µM for

24 hours or pre-treated with GTI-2040 at 10 µM, and 24 hours later treated with continuous exposure of 5-azaC at 5 µM for 48 hours. Cells were then lysed and dNTPs/NTPs were extracted and quantified by our previously published method (117).

Briefly, cells were counted and monitored for viability using trypan blue exclusion test before extraction. Following centrifugation at 1000 g for 5 minutes, cell pellets were washed with phosphate buffered saline (PBS) and deproteinized with an addition of 1 mL

60% methanol. The resulting solution was vortex-mixed for 20 seconds, incubated in -20

°C for 30 minutes and sonicated for 15 minutes in an ice bath. Cell extracts were centrifuged at 1000 g for 5 minutes at 4 °C and the supernatant was separated and dried under a stream of nitrogen. The residues were reconstituted with 300 μL of water, vortex-

40 mixed for 20 seconds and the cell extracts were centrifuged at 1000 g for 5 minutes at 4

°C. A 50 μL aliquot of the resulting supernatants was injected into an ion-trap LC-

MS/MS system (LCQ, Thermo Scientific, San Jose, CA) for dNTP and NTP measurements.

For DAC-TP levels, MV4-11 cells were either treated with 5-azaC alone at 5 µM for 4 hour, or were pre-treated with GTI-2040 at 5 and 10 µM for 48 hours, followed by

5-azaC continuous exposure at 5 µM for additional 4 hours. 2.5 µM DAC at 4 hours was included as positive control. The cells (10×106) were harvested, washed with phosphate buffered saline (PBS), and the triphosphates were extracted and quantified using the same procedure as described above.

2.4 Results

2.4.1 Determination of GTI-2040 in human leukemia MV4-11 cells

In order to confirm the uptake of GTI-2040 into cells, intracellular GTI-2040 levels were determined. Following the exposure of MV4-11 cells to GTI-2040 at 0.1, 0.3,

1, 3, 10 and 30 µM for 24 hours, GTI-2040 cellular concentrations were found to be

13.77 ± 0.43, 26.87 ± 1.79, 130.79 ± 18.16, 457.93 ± 23.62, 2872.92 ± 265.52 and

8647.32 ± 1150.47 pmol/106 cells, (mean ± SD, n=3), respectively (Figure 2.3). Thus, intracellular GTI-2040 content in MV4-11 cells was found to increase with increase in extracellular exposure concentrations.

41 2.4.2 GTI-2040 enhances the cytotoxic effects of 5-azaC in combination treatment

Effects of GTI-2040 on 5-azaC cytotoxicity in K562 (0.5 µM GTI-2040) and

MV4-11 (1 and 5 µM GTI-2040) cells were examined. As shown in Figure 2.4, we observed a shift in the dose-response curves to the left for the combination (increase in cytotoxicity), compared to the single drugs alone. Following pre-incubation with GTI-

2040 at 0.5, 1 and 5 µM for 24 hours, the IC50 values of 5-azaC were reduced in both

K562 and MV4-11 cell lines (Table 2.1) This data suggest that there was a synergistic effect between GTI-2040 and 5-azaC.

2.4.3 Synergy between 5-azaC and GTI-2040

To verify the synergistic effect of GTI-2040 and 5-azaC, their combination index

(CI) was calculated and isobologram constructed to evaluate their interaction at the 50% effect level (Figure 2.5). The fold of sensitization was determined as the inverse of the

CI. Following pretreatment with GTI-2040 at 0.5, 1 and 5 µM, the extents of synergy were found to be 3.3, 5.5 and 1.6 fold, respectively. Our data, therefore, showed a synergistic effect between GTI-2040 and 5-azaC, since the CI values calculated for all the combination were less than 1 (116).

2.4.4 GTI-2040 and 5-azaC reduces RRM2 mRNA levels in MV4-11 and K562

cells

Following treatment of MV4-11 cells with GTI-2040 (1, 5 and 10 µM) and K562 cells (10 µM) for 24 hours, RRM2 mRNA levels were decreased by 40-50%, relative to the untreated control and the small variation due to different GTI-2040 exposure levels

42 was within the statistical errors. Addition of 5-azaC at 5 µM to the 24 hours pre-treated cells with 1, 5 and 10 µM GTI-2040 for 48 hours did not result in further reduction of

RRM2 mRNA. Unexpectedly, treatment of these cells with 5 µM 5-azaC alone caused a

50% reduction in RRM2 mRNA (Figure 2.6). These data confirmed that GTI-2040 regulates RRM2 at the transcriptional level and indicated that RRM2 is a target of 5-azaC as well. However, the regulation of RRM2 mRNA by both GTI-2040 and 5-azaC alone may be through separate mechanisms.

2.4.5 GTI-2040 and 5-azaC decreases RRM2 protein expression

To investigate the effects of GTI-2040 on RRM2 protein expression, MV4-11 cells were pre-treated with GTI-2040 (1, 5 and 10 µM) and K562 (10 µM) for 24 hours.

We found a significant reduction in RRM2 protein expression by GTI-2040 (50-60%) relative to the untreated control. Subsequent combination treatment with 5 µM 5-azaC for additional 48 hours did not cause further decrease in RRM2 protein expression. However, treatment of these cells with 5 µM 5-azaC alone at 24 hours did not cause a significant reduction in RRM2 protein (Figure 2.7). These data confirmed that GTI-2040 regulates

RRM2 protein at both the transcriptional and translational level, while 5-azaC at 5 µM regulates RRM2 protein expression at the translational level at a longer exposure time

(>24 hours).

43 2.4.6 GTI-2040 and 5-azaC perturb the intracellular ribonucleotide (NTPs) and

deoxyribonucleotide (dNTPs) pools in MV4-11 and K562 cells

Since RRM2 is required for catalyzing the direct reduction of diphosphates to the corresponding deoxyribonucleotides (118, 119), we probed the intracellular dNTP/NTP pool changes caused by GTI-2040 or by 5-azaC alone and their combination. MV4-11 and K562 cells were first exposed to 10 µM GTI-2040 for 24 hours, followed by treatment with 5 µM 5-azaC for 48 hours. dNTPs and NTPs were extracted and measured by LC-MS/MS method as described above (117). We found a significant decrease (30-60%) in intracellular dTTP, dATP and dCTP levels for both cell lines following 10 µM GTI-2040 exposure alone when compared with the untreated control, but with no effect on the NTP pool (Figures 2.8A and 2.9A). No further decrease was apparent with co-treatment with 5 µM 5-azaC; however, 5 µM 5-azaC alone also decreased the dNTPs especially dATP and dCTP by 20-40%. There was no change in the

GTP, UTP and CTP levels by GTI-2040 alone, 5-azaC alone or their combination for both MV4-11 and K562 cells (Figures 2.8B and 2.9B). dGTP/ATP levels for MV4-11 cells were significantly increased, which may be due to feedback regulations to compensate for decrease in dNTPs.

2.4.7 5-AzaC reduces DNMT1 mRNA and protein levels in MV4-11 cells

Based on the mechanism of action of 5-azaC, 10-20 % of the drug is converted to

DAC, which targets DNA methyltransferase enzymes (DNMTs) leading to DNA hypomethylation (31, 120). Therefore, we sought to demonstrate the effects of 5-azaC treatments on DNMT1 mRNA and protein levels. As shown in Figure 2.10A, following

44 treatment of MV4-11 cells with 5-azaC at 5, 10 and 20 µM, DNMT1 mRNA levels decreased by 30, 35 and 50%, respectively, relative to the untreated control. This indicates that down-regulation of DNMT1 mRNA levels by 5-azaC is exposure- concentration dependent.

Following exposure of MV4-11 cells to 5 µM 5-azaC, DNMT1 protein level was totally depleted (Figure 2.10B). This confirms the hypomethylating effect of 5-azaC via a decrease in the DNMT1 protein expression. However, the decrease in DNMT1 mRNA level at 5 µM of 5-azaC did not change as significantly, when compared to the reduction in the protein level.

2.4.8 GTI-2040 when combined with 5-azaC reduces DAC-TP levels in MV4-11

cells

To investigate the effects of RRM2 inhibition by GTI-2040 on DAC-TP formation following exposure of cells to 5-azaC, MV4-11 cells were pre-treated with 5 and 10 µM GTI-2040 for 48 hours, followed by incubation with 5-azaC for an additional

4 hours. Our data showed that DAC-TP formed following 5 µM 5-azaC alone was 1.04 pmol/106 cells; however, when 5 µM 5-azaC was added to cells pre-treated with 5 and 10

µM GTI-2040, the DAC-TP levels were significantly reduced to 0.12 and 0.36 pmol/106 cells, respectively, and the small variation due to different GTI-2040 dose was within the statistical errors (Figure 2.11A). Treatment of cells with 2.5 µM DAC, yielded 1.92 pmol/106 cells of DAC-TP. This data further confirmed that GTI-2040 inhibits RRM2 expression and thereby preventing the formation of DAC from 5-azaC, hence less formation of DAC-TP when combined with GTI-2040 in cells.

45 2.4.9 GTI-2040 in combination with 5-azaC inhibits the reduction of DNMT1

protein level in MV4-11 cells

In order to further substantiate that pretreatment of cells with GTI-2040 prevents the formation of DAC-TP through the reduction of 5-azaC, which can result in abolishing the down-regulation of DNMT1 expression by 5-azaC, MV4-11 cells were first exposed to 5 and 10 µM GTI-2040 for 48 hours, followed by exposure with 1 and 5 µM 5-azaC for an additional 4 hours. The DNMT1 protein level was then measured. As shown in

Figure 2.11B, GTI-2040 alone did not alter the DNMT1 protein level as expected; instead it increased by ~30-50%, showing that DNMT1 is not regulated by GTI-2040.

Subsequent combination treatment with 1 and 5 µM 5-azaC for 4 hours did not cause any decrease in DNMT1 protein expression, which may be due to the opposing effects of

GTI-2040 and 5-azaC on DNMT1. Treatment with 1 and 5 µM 5-azaC alone decreased

DNMT1 protein levels by 10 and 50%, respectively, relative to the untreated control and this effect could presumably be due to its conversion to DAC. Taken together, these data confirmed that 5-azaC alone but not GTI-2040 regulates DNMT1 protein level and the combination of GTI-2040 with 5-azaC abrogates 5-azaC inhibitory activity on DNMT1 protein.

2.5 Discussion

The management of cancer is complex. While single drug therapy has contributed significant clinical response, there are still limitations because of multiple targets beyond single agent and development of drug resistance. Efforts are being made in the development of novel combination therapy to improve patients’ treatment. 5-AzaC has

46 activity in almost all leukemia and is currently being investigated in AML; however, a comprehensive understanding and exploration of 5-azaC in terms of its RNA incorporation is still critical in defining its exact mode of action to further advance its clinical use alone and in combination with other drugs. In this chapter, we demonstrate for the first time, that combination treatment of GTI-2040 with 5-azaC in K562 and

MV4-11 leukemia cells in vitro is synergistic, presumably through enhancing the RNA mediated effect of 5-azaC. Direct incorporation of 5-azaC into RNA can lead to the disruption of RNA elongations. Therefore, we have attempted to mechanistically enhance the potential 5-azaC RNA incorporation by inhibiting its reduction to DAC. Based on this rationale, GTI-2040 was used as a chemo-sensitizer to specifically down-regulate the expression of RRM2 and thereby enhance the cytotoxicity of 5-azaC.

GTI-2040 is a large and hydrophilic molecule and traditionally delivered by transfecting agents; however, the transfecting agents may be toxic and have biological activity of their own. Therefore, we delivered the antisense compound with an electroporation device to avoid this complication. In order to evaluate the delivery effectiveness of electroporation, GTI-2040 intracellular concentration was measured. As shown, about 60-70% of GTI-2040 was introduced into the cell with electroporation, which is comparable to that when using a transfection reagent (54). We have shown that

GTI-2040 sufficiently accumulated in MV4-11 cells at a dose as low as 0.1 µM. Growth inhibition assay was used to evaluate sequential treatment of GTI-2040 with 5-azaC. As indicated by the reduction in the IC50 values when the drugs were used in combination, a synergistic effect from the combined use of 5-azaC and GTI-2040 was seen. Using

47 combination index analysis (CI), we evaluated whether the cytotoxicity of both drugs was augmented in leukemia cells. At fixed concentrations 0.5-5 µM GTI-2040 combined with

0.01-100 µM 5-azaC, all CI values were less than 1.0, suggesting that the combination is synergistic. Our data support that the use of a combination of GTI-2040 with a low-dose

5-azaC generates a better therapeutic effect.

GTI-2040 alone and in combination with 5-azaC was found to cause a down- regulation of RRM2 mRNA and protein levels in both K562 and MV4-11 cells. The results are consistent with the proposed mechanism of GTI-2040 in leukemia cells (32) and provide an experimental verification of the theoretical basis for its use in the treatment of leukemia. Additionally, 5 µM 5-azaC alone caused ~ 50% reduction in the

RRM2 mRNA levels observed for the first time, but not the protein. A simple explanation is that down-regulation of RRM2 protein expression at 5 µM 5-azaC treatment is time dependent (>24 hours), which may be due to the turnover rate of RRM2 mRNA (~4 hours), therefore requiring longer drug exposure time in cells. Thus, it seems that 5-azaC is more potent than GTI-2040 in RRM2 mRNA regulation in both MV411 and K562 cells. Nevertheless, the reduction in the RRM2 mRNA levels by GTI-2040 alone and 5- azaC alone may be through different processes. Furthermore, following GTI-2040 treatment in both cell lines, a significant decrease in dNTPs levels were observed, which confirmed the inhibitory effect on ribonucleotide reductase by GTI-2040. However, no significant changes were observed among NTP levels except a slight increase in dGTP/ATP level in MV4-11 cells. This effect may also be due to the salvage pathways of intracellular NTP , which compensates for perturbation in NTPs. We also

48 showed that 5-azaC alone can modulate the dNTP/NTP levels but did not cause further reduction in combination. This may be explained by the fact that both GTI-2040 and 5- azaC can independently decrease dNTP levels (~50%); hence, combination of these two drugs results in an additive effect, at least in nucleotides modulation.

In our study, we seek to demonstrate the demethylating effect of 5-azaC, since about 10-20 % of the 5-azaC is converted to DAC targeting DNA methyltransferase

(DNMTs). Treatment with 5-azaC has been shown to reverse DNA hypermethylation via incorporation of its reduced analog DAC-TP into DNA by covalently trapping of DNA methytransferases enzyme (DNMTs) (13). Our data showed that DNMT1 mRNA and protein levels were decreased following exposure of MV4-11 cells to 5-azaC alone.

DNMT1 protein expression was totally depleted at 5 µM and the mRNA level decreased in a dose-dependent manner. We noticed that, the expression level of DNMT1 mRNA level at 5 µM of 5-azaC did not change as significantly, when compared to the depletion of the protein at the same dose. A simple explanation may be the degradation of DNMT1 protein level upon 5-azaC treatment might not necessarily be due merely to a decline in mRNA level but that 5-azaC induced degradation of DNMT1 protein may be a post- translational event (121).

GTI-2040 is a specific target of RRM2 enzyme and currently used in a phase I clinical trial in AML patients. Thus, in our effort to improve the RNA mediated effect of

5-azaC, we sought to investigate the ability of GTI-2040 to inhibit RRM2 expression levels and temporarily blocks DNA methylation pathway of 5-azaC. We showed that

49 sequential treatment of GTI-2040 with 5-azaC demonstrates a synergistic effect in terms of their cytotoxicity; however, the ability of 5-azaC to cause demethylation was abolished. This is consistent with the inhibition of reduction of 5-azaC to DAC via

RRM2 by GTI-2040. Previous pharmacodynamic studies of GTI-2040 have shown that longer exposure of cells to GTI-2040 improves its cellular uptake and hence RRM2 target down-regulation (54). Therefore, in our studies, MV4-11 cells exposed to GTI-2040 for a longer period (48 hours), followed by treatment with 5-azaC resulted in a significant inhibition of RR expression and a very low level of DAC-TP formed compared to 5-azaC alone. Inability to form DAC-TP from 5-azaC resulted in the failure of down-regulation of DNMT1 at 5 µM of 5-azaC, when cells were pre-treated with GTI-2040 for 48 hours.

We have found that in MV4-11 cells, shorter exposure time to GTI-2040 (24 hours), followed by continuous treatment with 5-azaC did not affect the ability of 5-azaC to down-regulate DNMT1 (data not shown), also demonstrating the increased RRM2 target down-regulation of GTI-2040 at longer exposure time. Thus far, we have demonstrated successful inhibition of the formation of DAC-TP by RRM2 inhibition with GTI-2040, thus likely increasing the incorporation of 5-azaC into RNA. Because of this, we would like to further extend our efforts to establish a link between RNA mediated effect of 5- azaC and microRNA (miRNA). In our initial microarray analysis and validation following treatment of MV4-11 cells with 5-azaC, our data demonstrate that 5-azaC can modulate certain miRNA such as miR29b and miR181a. Details of this study will be discussed in Chapter 4.

50 2.6 Conclusion

Cytotoxity of 5-azaC in combination with 5-azaC was found to be synergistic. GTI-

2040 alone and in combination with 5-azaC was found to inhibit ribonuclotide reductase by down-regulation of RRM2 expression levels. The inhibition of RRM2 by GTI-2040 results in a reduction in dATP, dTTP and dCTP levels. The down-regulation of RRM2 expression levels and reduction in the dNTPs by 5-azaC was observed for the first time.

Reduction in RRM2 expression levels by GTI-2040 was also found to result in an opposing effect to the hypomethylating effect of 5-azaC via its conversion to DAC-TP and subsequent reduction in DNMT1 protein expression due to the blockade of reduction of 5-azaC to DAC. These results will provide a laboratory and mechanistic justification for clinical evaluation of this novel combination.

51

GTI-2040 (µM) IC50 (µM) of Combination Fold of 5-Azacytidine alone and/or Index (CI) Sensitization GTI-2040 combined 0.0 (K562/MV4-11) 7.50 ± 4.90/7.23 ± 1.29 N/A N/A 0.5 (K562) 1.85 ± 1.90 0.30 3.3 1.0 (MV4-11) 0.49 ± 3.44 0.20 5.5 5.0 (MV4-11) 0.55 ± 2.11 0.63 1.6

IC50 values of GTI-2040 alone is approximately 9.03-11.50 ± 2.90 µM in both cell lines.

Table 2.1 IC50 values, the concentration of drug that inhibit 50% of cells, combination index (CI) and sensitization fold (inverse of CI) were calculated in MV4-11 and K562 cells.

52

Inhibition by GTI-2040

~80-90%

-Increased incorporation into RNA -Inhibition of protein synthesis -Hence, increased cytotoxicity

Figure 2.1 Diagrammatic rationale for the combination of 5-azaC with GTI-2040 in leukemia cells. GTI-2040 inhibits RR, and causes down-stream depletion of dNTP and NTP pools, while enhancing the incorporation of 5-azaC into RNA resulting in increased cytotoxic effects.

53

Figure 2.2 Schematic representation of the principles of electroporation.

54 ** 10000

8000

6000

4000 **

** 500

400

(pmol/10^6 cells) (pmol/10^6 300 ** Intracellular level of GTI-2040 level Intracellular 200

100 **

0

GTI -2040 (µM) 0.0 0.1 0.3 1.0 3.0 10.0 30.0

Figure 2.3 Intracellular accumulation of GTI-2040 in MV4-11 cells following introduction of GTI-2040 at the indicated concentrations for 24 hours by electroporation. Vertical bars represent mean ± SD from triplicate experiments and compared with GTI- 2040 at 0.1 µM with asterisks showing p<0.01.

55

A

5µM GTI + variable 5-azacytidine (*p<0.05) 1µM GTI + variable 5-azacytidine (*p<0.05) 5-azacytidine alone 120.0 MV4-11 cells 100.0

80.0 60.0

40.0 control) 20.0

Cell Survival (% 0.0 0.001 0.01 0.1 1 10 100

Concentration of 5-azacytidine (µM)

B 5-azacytidine alone 0.5µM GTI + variable 5-azaC (*p<0.05) GTI-2040 alone K562 cells 140 120

100

80

60 40

20 Cell Survival (% control) (% Survival Cell 0 0.0001 0.001 0.01 0.1 1 10 100 Concentration of 5-azacytidine (µM)

Figure 2.4 Effect of GTI-2040 on the cytotoxicity of 5-azaC. (A) Pretreatment of MV4- 11 cells with 1 and 5 µM GTI-2040 decreased the IC50 of 5-azaC (B) Pretreatment of K562 cells with 0.5 µM GTI-2040 also decreased the IC50 of 5-azaC. We observed a significant decrease (p<0.05) in survival for the combination relative to the 5-azaC or GTI-2040 alone (a left shift in the curves). Data are means ± SD of three replicates.

56

B A

2

8

7

6

5 1 4

3 5-azaC alone 5-azaC

2 Combination Index Combination

1

0 0 0.0 1.0 2.0 3.0 4.0 5.0 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 GTI-2040 (µM) GTI-2040 alone

Figure 2.5 (A) Plot of combination index vs GTI-2040 concentration at 50% drug effect level. Combination index < 1, >1 and 1 are synergistic, antagonistic and additive, respectively (B) Isolobogram graph showing line of additivity and the intercepts are IC50 of GTI-2040 and 5-azaC when present alone.

57

1.4 MV4-11 cells 1.2

1

0.8 * * * * 0.6 *

0.4

RRM2/ablmRNA (Relativecontrol) to 0.2

0 GTI-2040 (µM) - 1 5 10 1 5 10 - 5-azaC (µM) - - - - 5 5 5 5

1.20

K562 cells 1.00

0.80 * * * 0.60

RRM2/ablmRNA 0.40 (Relativecontrol) to

0.20

0.00 GTI-2040 (µM) Control - 10 µM10 GTI 5 µM 5-azaC- 10 µM GTI 10 + 5 µM 5- 5-azaC (µM) - - 5 azaC 5

Figure 2.6 GTI-2040 reduced RRM2 mRNA expression levels in MV4-11 and K562 cells and no further reduction in RRM2 mRNA expression caused by addition of 5 µM 5- azaC. 5-azaC alone also decreased RRM2 mRNA levels by 40-50%. RRM2 mRNA levels were normalized by abl and vertical bars are mean ± SD from triplicate experiments and compared with the control with asterisks showing p<0.05.

58

MV4-11 cells

GTI-2040 (µM) - 1 5 10 1 5 10 - 5-azaC (µM) - - - - 5 5 5 5 RRM2

GAPDH RRM2/GAPDH 1.0 0.8 0.8 0.4 0.4 0.4 0.5 0.8 Ratio

K562 cells

GTI-2040 (µM) - 10 10 - 5-azaC (µM) - - 5 5

RRM2

GAPDH

RRM2/GAPDH 1.0 0.4 0.4 0.9 Ratio

Figure 2.7 GTI-2040 decreased RRM2 protein expression (50-60%) in MV4-11 and K562 cells and addition of 5-azaC did not contribute to further reduction in RRM2 protein levels. 5-azaC at 5 µM alone for 24 hours had little or no effect on RRM2 protein. RRM2 protein levels were quantified by densitometry and expressed as the ratios over the loading control GAPDH.

59 A MV4-11 cells

B

Figure 2.8 Perturbation of intracellular dNTP/NTP pools by GTI-2040 and 5-azaC. MV4-11 cells were treated with GTI-2040 at 10 µM in the presence and absence of 5- azaC (A) GTI-2040 reduced the dTTP, dATP and dCTP by 30-60%; and 5-azaC alone decreased dATP and dCTP by 20-30%. (B) GTI-2040 did not change GTP, CTP and UTP pools, but dGTP/ATP increased. Vertical bars are mean ± of SD of triplicate experiments with asterisks showing p<0.05 versus control.

60

A K562 cells

B

Figure 2.9. Perturbation of intracellular dNTP/NTP pools by GTI and 5-azaC. K562 cells were treated with 10 µM GTI-2040 in the presence and absence of 5-azaC (A) GTI-2040 reduced the dTTP, dATP and dCTP by 30-60%; and 5-azaC alone by 20-40%. (B) GTI- 2040 did not change GTP, CTP, UTP and dGTP/ATP pools. Vertical bars are mean ± SD of triplicate experiments with asterisks showing p<0.05 versus control.

61

A

1.2

1 * 0.8 *

0.6 *

0.4

DNMT1/ablmRNA

0.2

to (relative untreated control) 0 5-azaC (µM) - 5 10 20

B 5-AzaC (5µM)

- +

DNMT1

GAPDH

Figure 2.10 (A) 5-AzaC at the indicated concentrations decreased DNMT1 mRNA level in MV411 cells in a dose-dependent manner following 48 hours treatment (B) 5-azaC at 5 µM depletes DNMT1 protein levels. DNMT1 mRNA levels were normalized by abl and vertical bars are mean ± SD of triplicate experiments with asterisks showing p<0.05 versus control.

62

A 2.50

2.00 cells) 6 1.50 * 1.00 ** 0.50 DAC-TP level (pmol/10 DAC-TPlevel **

0.00

GTI-2040 (µM) - - 5 10 5-azaC (µM) - 5 5 5 DAC (µM) 2.5 - - -

B

GTI-2040 (µM) - 5 10 5 5 10 10 - - 5-azaC (µM) - - - 1 5 1 5 1 5 DNMT1

GAPDH

DNMT1/GAPDH 1.0 1.5 1.5 1.5 1.3 1.5 1.5 0.9 0.5 Ratio

Figure 2.11 (A) 5 µM 5-azaC alone in MV4-11 cells resulted in the formation of DAC- TP, but when the cells were pre-treated with 5 or 10 µM GTI-2040, DAC-TP levels significantly decreased. The vertical bars represent mean ± SD from triplicate experiments and compared with DAC at 2.5 µM. *p<0.05, ** p<0.01 (B) Pretreatment of MV4-11 cells with 5 or 10 µM GTI-2040 followed by 1 and 5 µM 5-azaC did not reduce the DNMT1 protein level, but 1 and 5 µM 5-azaC alone for 4 hour decreased DNMT1 levels by 10-50%. DNMT1 protein levels were quantified by densitometry and expressed as the ratios over the loading control GAPDH.

63

CHAPTER 3

IN VIVO MODULATION OF THE PHARMACOLOGIC EFFECTS OF 5- AZACYTIDINE BY ANTISENSE GTI-2040

3.1 Abstract

GTI-2040 is an antisense to the M2 subunit of ribonucleotide reductase (RR), which catalyzes conversion of ribonucleotides to deoxyribonucleotides and is commonly over-expressed in most type of cancers. 5-Azacytidine (5-azaC) exerts its anticancer activity mainly through direct incorporation into RNA (80-90%), thereby leading to disruption of RNA stability and elongation. A substantial portion of 5-azaC is channeled to DNA incorporation through its conversion to decitabine via the enzyme RRM2. Thus,

5-azaC’s antitumor effect consists of both DNA and RNA mechanisms. We hypothesized that by using GTI-2040 in combination with 5-azaC, the RNA incorporation of 5-azaC, thus RNA mechanism will be enhanced. In Chapter two, we have demonstrated that in vitro, we enhanced the incorporation of 5-azaC into RNA that resulted in increased cytotoxicity and protein synthesis inhibition in MV4-11 and K562 human leukemia cell lines. We further showed that the combination of 5-azaC and GTI-2040 was synergistic.

In this chapter, in vivo studies were conducted by inoculating nu/nu mice with MV4-11 acute myeloid leukemia cells. Tumor volume and weight of the xenografted implants were measured following treatments with the combination of GTI-2040 and 5-azaC and compared with those of animals treated with 5-azaC, GTI-2040 and vehicle alone. Mouse

64 body weights were also measured. Quantitative RT-PCR was used to measure the transcriptional level of RRM2 expression levels and Western blot assay was used to evaluate the corresponding protein levels. A pharmacodynamic (PD) model was developed to evaluate the tumor growth rate of xenograft MV4-11 tumor models in control and treatment groups (GTI-2040 and 5-azaC). All drugs treatment groups but not the untreated control showed decreased tumor size and weight after 32 days and this suggests antitumor effect in the drug combination group. All treatment groups showed a minimal level of toxicity as evidence by negligible change in body weights (<2%).

RRM2 mRNA and protein expression was decreased by about 50% in tumor tissues in the GTI-2040 treated group alone, 5-azaC treated group alone and their combination treated group. The pharmacodynamic model developed fitted the control tumor growth very well and demonstrated the tumor growth rate and tumor death rate with 0.007 mm3/h and 0.000001 mm3/h, respectively. A synergistic antitumor effect was further simulated in the combination treatment of GTI-2040 and 5-azaC and demonstrated inhibition of more than 70% tumor growth through this modeling. The combination of low doses of

GTI-2040 and 5-azaC seems to potentiate the anti-leukemia activity of the two compounds in vivo and will be explored in the clinic. This model has utility for optimization of the dosing schedules of the combined drugs.

3.2 Introduction

5-AzaC is a cytidine analog that belongs to the family of pyrimidines. Being a nucleoside, incorporation of 5-azaC into RNA causes direct cytotoxicity and anti- proliferative effects in tumor cells. The biological activity of 5-azaC is associated with

65 the incorporation into cellular DNA with subsequent sequestration of DNMTs via covalent bond formation (10), presumably through its reduced product decitabine and/or through direct incorporation into RNA. Therefore, 5-azaC can inhibit DNA, RNA and protein synthesis (13). About 10-20% of 5-azaC was found to be incorporated into the

DNA fraction through indirect formation of 5-azaC deoxynucleotide diphosphates catalyzed by ribonucleotide reductase (RR). This DNA incorporation of 5-AzaC causes hypomethylation via inhibition of DNA methyltransferases at low-doses (3-5 µM in vitro or 10-75 mg/m2/day) (9, 10, 90). DNA hypomethylation may affect transcription of messenger RNA and differential gene expression, which subsequently triggers terminal differentiation and eventually apoptotic processes. About 80-90% of 5-azaC is incorporated into RNA causing RNA destabilization that may interfere with transfer

RNA and ribosomal RNA processing (9). It has been reported that 5-azaC is active against leukemia cells in mice and with the most effective dose starting at 10 mg/kg when in combination with other drugs (122). For example, when combining 5-azaC at a daily intraperitoneal dose of 20 mg/kg with arsenic trioxide in BALB/c nude mice inoculated with myeloma cells, significant delayed growth in the treated groups compared to the control group was found without increased toxicity, as evidenced by no weight loss or sudden death (123). In the same token, we hypothesize that combination treatment of 5- azaC and GTI-2040 might be effective in reducing tumor growth without causing significant toxicity in vivo. Our previous in vitro studies (Chapter 2) have demonstrated that 5-azaC treatments at clinically achievable concentrations induce synergistic effects in combination with GTI-2040. Therefore, it is important to pursue this combination in vivo in order to assess the optimum dosing regimen potentially useful in the clinic.

66 Additionally, this combination can be used for exploiting the mechanism of 5-azaC via its

RNA incorporation in a mouse model.

Ribonucleotide reductase (RR) is a highly regulated enzyme that catalyzes the reduction of ribonucleotides to their corresponding deoxyribonucleotides, which is a rate- limiting step for DNA synthesis and repair (23) and also plays a role in the critical early events of tumor promotion common in malignant cells (24-26). Over-expression of RR has been found in almost every type of cancer studied (32) and has been linked to malignant status of tumor cells via the cooperation with a number of activated oncogenes.

For example, enhanced expressions of oncogenes, such as, v-fms, v-src, A-raf, v-fes, c- myc in tumor cells are associated with over-expressed RR (124). In addition, over- expression of RRM2 mRNA and its enzymatic activity contributed to a 3-fold increase in tumor’s invasiveness in RRM2 gene transfected cells compared to the KB wild-type cells

(26). Recently, RRM2 has been found to be inhibited by GTI-2040, a 20-mer phosphorothioate oligonucleotide (PS ODN) with sequence of 5’-

GGCTAAATCGCTCCACCAAG-3’. GTI-2040 is designed to specifically bind to and target the mRNA coding region of the RRM2 (27). The antitumor activity of GTI-2040 has previously been evaluated in-vitro and in-vivo. In nude and SCID mice, activity of

GTI-2040 was tested against a wide range of human xenograft tumors, including colon, pancreatic, liver, lung, breast, renal, and prostate cancers. In these animal models, GTI-

2040 acted in a dose-dependent manner to down-regulate RRM2 expression with a concomitant decrease in tumor growth, metastasis and an increase in animal survival

(32). GTI-2040 also increased survival in animals with lymphoma and leukemia. In vivo

67 antisense effects are best evaluated by the characterization of the pharmacodynamics of the target mRNA and its correlation to the drug level and the therapeutic effects (49,

125). To date, little is known about the tumor growth pattern after antisense therapy through pharmacodynamic modeling of the target mRNA, compared to its pharmacokinetics (125). Therefore, effort will be made to understand the pharmacodynamics of the combination of GTI-2040 and 5-azaC through pharmacodynamic modeling and simulation.

The antitumor activity of GTI-2040 has been investigated in combination with standard chemotherapeutic agents such as 5-FU, vinblastine and IL-2 against the growth of tumor-bearing nude/SCID mice and has shown to have promising activity (126).

Preclinical studies also support targeting RRM2 in combination with cytotoxic agents. In

Chapter two, I described the synergistic effect of GTI-2040 in combination with 5-azaC in K562 and MV4-11 human leukemia cell lines. In this chapter, we will evaluate the synergistic effect of 5-azaC and GTI-2040 alone and in combination in tumor-bearing mice inoculated with MV4-11 cells in vivo. In addition to monitoring the tumor size and weight in the mice, we also examined the pharmacodynamic modulation of the RRM2 mRNA and protein expression by GTI-2040 and 5-azaC in xenograft tumor tissues at the end of the chapter. At the same time, as part of the preclinical evaluation in preparation for further clinical studies, a mechanism-based pharmacodynamic model was developed to monitor the tumor growth rate and evaluate the synergistic antitumor effect addressing the combination treatment of 5-azaC and GTI-2040. The goal of the present study was to

68 characterize and evaluate the combination effect of GTI-2040 ad 5-azaC on tumor growth and provides PD profiles for preclinical assessment to assist future clinical study design.

3.3 Materials and methods

3.3.1 Chemicals

5-AzaC was obtained from The National Cancer Institute (Bethesda, MD) and a stock solution was prepared in dimethyl sulphoxide (DMSO) and further dilution was performed in phosphate buffered saline (PBS, Invitrogen, Carlsbad, CA). The 20-mer phosphorothioate oligonucleotide, GTI-2040, with sequence of 5’-

GGCTAAATCGCTCCACCAAG-3’ was provided by the The National Cancer Institute

(Bethesda, MD) and used without further purification. Lipofectamine 2000TM (lipo)

(Invitrogen, Rockville, MD) was used as transfecting agent for GTI-2040.

3.3.2 Cell culture conditions

Human leukemia cell line MV4-11 was obtained from American Type Culture

Collection (ATCC, Manassas, VA). The cells were cultured in RPMI 1640 medium supplemented with L-glutamine (Supplied by Tissue Culture Shared Resource,

Comprehensive Cancer Center, The Ohio State University, Columbus, OH), 1%

Penicillin-Streptomycin (Gibco, Rockville, MD) and 10% fetal bovine serum (FBS)

(Invitrogen, Rockville, MD). The cell line was maintained at 37 °C in a humidified environment with 5% CO2. The culture media was refreshed every 3-4 days to maintain a

69 cell density of 0.4 million cells/ml. Viability and cell counts were determined using trypan blue dye exclusion.

3.3.3 Xenograft tumor model

Female athymic nu/nu mice, 4-6 weeks old and weighing 18-23 g, were obtained from Charles River Laboratory (Wilmington, MA). Animals were housed in sterile cages and experiments carried out in accordance with the guidelines of the Association for

Assessment and Accreditation of Laboratory Animal Care International (AAALAC).

MV4-11 Cells (10x106 cells per mouse/side) were suspended with Matrigel (Becton

Dickinson) and phosphate buffer saline in a 1:1 ratio and implanted subcutaneously into the right and left flanks, respectively, of the athymic nu/nu mice (Figure 3.1). After inoculation, mice were randomly assigned into six different cohorts to receive saline, lipo alone, 5-azaC/lipo, 5-azaC alone, GTI-2040/lipo and GTI-2040/lipo plus 5-azaC. Each experiment group contains six mice. When the tumor size reached a mean tumor volume of 100 to 150 mm3, treatments were initiated.

3.3.4 Treatment of mice with GTI-2040 and 5-azaC

GTI-2040 and 5-azaC alone and/or in combination were dissolved in saline. The drugs were administered via intra-peritoneal (i.p) injection and dose at 5 mg/kg for GTI-

2040 alone (GTI-2040-lipo complex, ratio 1:2.5 formulated in 100 µL saline solution), 5- azaC at 20 mg/kg alone and co-injected with GTI-2040 twice a week for the combination treatment. The treatment was continued to 32 days period for the anti-tumor growth studies. Antitumor activity was estimated by the measurement of tumor size with a

70 caliper and the body weight of animals at the start of experiment and at four days intervals. The tumor volume was then calculated by the formula L x W x H/2, where L is the length, W is the width, and H indicates the height of tumor. After 32 days treatment, the animals were sacrificed, the tumors removed and the final weight measured. The tumor tissues were snapped frozen in liquid nitrogen and stored in -80°C until use.

3.3.5 Quantification of RRM2 mRNA levels by real-time RT-PCR

Frozen tumor tissues excised from the untreated and treated mice were thawed on ice, homogenized using Trizol reagent (Invitrogen, Carlsbad, CA) and total RNA was isolated according to the manufacturer’s protocol. Briefly, tumor tissues in Trizol reagent was treated with chloroform and the total RNA was precipitated with isopropyl alcohol, followed by a washing step with 75% ethanol. RNA was then dissolved in RNase free water and its concentration and purity was measured by a Nanodrop spectrophotometer

(Nanodrop Technologies, Wilmington, DE). cDNA was synthesized from 2 µg total RNA using Moloney murine leukemia virus reverse transcriptase (Invitrogen). The cDNA templates and primers were then mixed with reagents from a SYBR Green PCR Master

Mix (Applied Biosystem, Foster City, CA). Reactions were carried out in triplicate in

ABI prism 7700 sequence detector (Applied Biosystems), and data were analyzed by comparative CT method. Dissociation curves were also obtained to examine the purity of the amplified products. The amount of RRM2 mRNA in each sample was normalized with respect to an internal control, abl. The relative changes in treatment groups were

71 expressed as percentages of untreated control (arbitrarily set at 1). The results were expressed as the mean ± SD from triplicate determinations.

3.3.6 Measurement of RRM2 protein expression by western blot

Frozen tumor tissue samples obtained from mice of the six cohorts were thawed on ice and homogenized in 100 μL lysis buffer (50 mM PH 7.6 Tris-HCl, 250 mM NaCl,

5 mM EDTA, 2 mM Na3VO4, 50 mM NaF and 1% protease inhibitor cocktail) (P8340,

Sigma) for 30 minutes on ice. The lysate was sonicated for 10 seconds. Total protein concentration was determined using the BCA protein assay method (Pierce, Rockford,

IL). Equal amounts of protein for each sample were incubated with 6x SDS loading buffer (100 mM, pH 6.8 Tris, 200 mM DTT, 4% SDS, 20% glycerol, and 0.015% bromphenol blue) and boiled for 5 minutes. The proteins were then separated on 15%

SDS-polyacrylamide gels and transferred to nitrocellulose membranes (Amersham,

Piscataway, NJ). The RRM2 proteins were recognized with a goat antihuman RRM2 polyclonal antibody (E-16) (Santa Cruz Biotechnology, Santa Cruz, CA) as the first antibody, followed by a HRP (horseradish peroxidase) conjugated anti-goat IgG. RRM2

(MWT 45,000 dalton) was detected by ECL (Amersham, Arlington Heights, IL) and

GAPDH was used as the internal loading control. RRM2 protein expressions were quantified by densitometry and normalized to GAPDH.

72 3.3.7 Tumor growth pharmacodynamic modeling (control group)

In vivo tumor growth in xenograft models is known to follow exponential growth

(127) up to a threshold weight (128), at least in its early phases of development, and subsequently follows a linear growth which eventually reaches a plateau as described by the Gompertz model (129-131). The Gompertz model of growth has been widely used as a simple and adequate discriptor of tumor growth curves (132) and has been used repeatedly to study tumor growth dynamics in many in vitro studies of tumor growth kinetics both in animals and in cell lines (133). It incorporates all biologically essential phenomena of cell population dynamics such as cell proliferation and cell loss (134, 135), and allows for discrimination between untreated tumor growth and drug dependent model parameters. In our approach, using Gompertz-like tumor growth model, we assumed that there is a maximum tumor size, Ninf, at which tumor cannot grow further (136). When tumor size approaches Ninf, tumor will grow slower and slower, as described by the following equation.

dN N  K0 *(1 )* N  N * K E dt Ninf

3 -1 where N (cells x 10 ) represents the tumor size at time t (h). K0 (h ) is first order rate

-1 constant characterizing the rate of exponential tumor growth and KE (h ) is first order parameter for cell death. This model was used to fit the control tumor growth rate in the

MV4-11 xenograft model, when the initial tumor sizes reached 100-150 mm3. The estimates for K0 and KE obtained from modeling the observed tumor size with the control growth tumor model were then fixed (used as an initial estimate values) to simulate the tumor growth in the combination treatment of GTI-2040 and 5-azaC.

73 3.3.8 Tumor growth model for treated groups

In the control model, all the tumor cells were assumed to be proliferating naturally, whereas the treated tumor model assumes that our anticancer treatments (GTI-

2040 w/o 5-azaC) will sensitize tumor cells to become non-proliferating or reduce tumor growth rate and eventually leading to tumor cell death. In the mechanism studies with

GTI-2040 (27) alone and/or 5-azaC (Chapter 2), it has been shown that both drugs can inhibit RRM2 mRNA expression levels during cell proliferation and contributes to the degradation of the enzyme, in addition to the cytotoxic effects of 5-azaC. In our approach, we assume that the RRM2 expression level is directly related to the tumor growth rate. In the untreated tumors, over-expression of RRM2 will drive the full tumor growth by increasing DNA synthesis and repair, while the reduction in RRM2 expression levels by GTI-2040 w/o 5-azaC treatments, will contribute significantly to the inhibition of tumor growth. The tumor size differential equations are presented as follows. dN1 RRM 2GTI 20405azaCcomb N1  K0 * 0 *(1 )* N  N * KE dt RRM 2 Ninf

dN2 RRM 2GTI 2040 N2  K0 * 0 *(1 )* N  N * KE dt RRM 2 Ninf

dN3 RRM 2GTI 2040 N3  K0 * 0 *(1 )* N  N * KE dt RRM 2 Ninf where, RRM2GTI-2040+5azaCcomb, RRM2GTI2040 and RRM25azaC are the RRM2 levels in the tumor cells with GTI-2040 plus 5-azaC, GTI-2040 alone and 5-azaC alone, respectively.

RRM20 is the initial RRM2 levels in tumor cells, which is assumed to be similar to

RRM2 levels in the control tumor cells. The ratio RRM2 /RRM20 is the rate of change of

74 0 RRM2 (initial RRM2 levels) to RRM2 levels following treatment for each drug. N1, N2 and N3 represent the tumor size reached in the treatment groups for GTI-2040 and 5-azaC combination, GTI-2040 alone and 5-azaC alone, respectively. The ratio N/Ninf implies the rate of change when N (tumor size) approaches Ninf (maximum tumor size). The model simulation was done using these equations put together but added in a step-wise manner.

3.3.9 Computer Software

WinNonLin (Pharsight, Mountain View, CA) was used in the PD model development and simulation. Excel (Microsoft Office Word 2007, WA) was used in the data plot.

3.3.10 Statistical analysis

Statistical significance of differences observed in treated mouse groups compared with the control group was determined using a Student’s t-test using Minitab statistical software (Minitab, State College, PA). The minimal level of significance was a p-value below 0.05.

3.4 Results

3.4.1 Anti-tumor effect of GTI-2040 and 5-azaC treatments

In order to ascertain the anti-tumor activities in vivo using the in vitro data of the combined treatment of GTI-2040 and 5-azaC, human MV4-11 cells were inoculated into athymic nu/nu mice. The tumors were allowed to grow until they were palpable and

75 could be measured accurately with calipers (25cm long). Subsequently, mice were treated with GTI-2040 and 5-azaC alone separately and in combination along with the appropriate controls (saline and lipo treated groups). Tumor growth was assessed by measurement of tumor size over time. Tumor regression becomes prominent at about 16 days for all agents used and was clearly distinguishable from the untreated control.

Tumor growth rate of lipofectamine control group alone was the same as that of the saline treated group group and had no effect, demonstrating that lipofectamine did not contribute to the toxicity of the drugs. Tumor growth rate of GTI-2040 alone decreased

(10%) at later time (day 24) compared to the untreated control, which may be due to longer residence time or higher dose needed for sufficient GTI-2040 intracellular concentration needed in this tumor model. Single agent 5-azaC alone or 5-azaC/lipo reduced tumor growth by 20-24%; however, the combination of GTI-2040 with 5-azaC reduced the tumor growth by 61% as compared to the saline treated control, (p<0.05)

(Figure 3.2).

3.4.2 Reduction in tumor weight after treatment

Following 32 days treatment of mice, the animals were sacrificed and the tumors were removed. The tumor weight at the end of the experiment was measured. The tumor weight of the lipofectamine treatment group alone was the same as saline treated control.

GTI-2040 alone treated group showed decreased tumor weight, however, with no statistical significance relative to the saline treated control. 5-AzaC alone or 5-azaC/lipo treated group showed decreased tumor weight by 1.4-1.8 fold and the small variation was within the statistical errors. However, GTI-2040 when combined with 5-azaC reduced the

76 total tumor weight by 3.1 and 2.7-fold relative to the untreated control and to the GTI-

2040 alone treatment group, respectively. When compared to 5-azaC alone, GTI-2040 combined with 5-azaC decreased the tumor weight by 2.7-fold (p<0.05) (Figure 3.3).

3.4.3 Measurement of mouse body weight

The mouse body weight measurement was taken at the start of experiment and every four days thereafter. It was found that the mouse body weights for the treatment groups showed no difference from those of the control groups. As shown in Figure 3.4,

GTI-2040 and 5-azaC treatment alone and in combination produced minimal toxicity as evidenced by <2% body weight loss. This finding suggested that the combination of GTI-

2040 with 5-azaC may reduce the toxicity associated with 5-azaC treatment alone when used at high dose and for longer period of time (122). Due to the limited number of mice per group (n = 6) for this experiment, our conclusion may not be explicit, however, our data may suggest that the combination therapy of GTI-2040 and 5-azaC at this dose level may be well-tolerated in vivo furthering addition to further toxicology studies.

3.4.4 GTI-2040 and 5-azaC decreases RRM2 mRNA and protein levels in xenografted tumor tissues

Since GTI-2040 and 5-azaC treatments significantly lowered RRM2 mRNA and protein levels in leukemia cell lines in vitro, we next examined the same in MV4-11 engrafted nu/nu mice tumor tissues. Tumor tissues collected from mice following end of treatment were homogenized, lysed and analyzed for RRM2 mRNA and protein expression levels. As shown in Figure 3.5A, lipofectamine alone did not affect the RRM2

77 mRNA level as expected. GTI-2040 alone decreased the RRM2 mRNA level, albeit not statistically different relative to the untreated control (p>0.05). Notably, RRM2 mRNA expression was reduced (~40%) by 5-azaC alone or 5-azaC/lipo relative to the untreated control; however, no significant difference was found when compared to that of GTI-

2040 alone. Combination of GTI-2040 with 5-azaC reduced the RRM2 mRNA levels by

~50%, relative to the untreated control, although it was found not to be significantly different compared with that of GTI-2040 alone.

In Figure 3.5B, we showed that lipofectamine alone did not affect the RRM2 protein expression level, but GTI-2040 decreased RRM2 protein level by 40%, while, 5- azaC or 5-azaC/lipo reduced the RRM2 level by 60-65% relative to the untreated control.

More so, when GTI-2040 was combined with 5-azaC, the RRM2 protein levels was decreased by 80% in these tumor tissues, when compared to the saline treated group. Our data therefore confirmed the use of GTI-2040 with 5-azaC in combination will enhance the therapeutic potential of their individual drugs in terms of RRM2 down-regulation.

3.4.5 Tumor growth model in control animals

When two parameters, K0 and KE, were examined via the control tumor growth model, it was found that the average tumor growth values obtained from the control animals inoculated with MV4-11 leukemia cell line as a function of time fitted well to the model, as shown in Figures 3.6 & 3.7. The evaluated tumor growth rate (K0) and tumor

3 3 death rate (KE) were 0.007 mm /h and 0.000001 mm /h, respectively, through modeling the observed tumor size with the control growth model.

78 3.4.6 Molecular pharmacodynamics and tumor growth model in the combination treatment of GTI-2040 with 5-azaC

In order to better evaluate the synergistic antitumor effect of the combination of

GTI-2040 with 5-azaC, the initial PD parameters obtained from our in vitro and in vivo experiments were used as our initial estimates. We have shown that both GTI-2040 or 5- azaC alone and their combination reduced the RRM2 expression level in MV4-11 cell line and tumor tissues by >50%. Using our PD model, we further simulated the synergistic antitumor growth effect in the combination treatment as shown in Figure 3.8.

We show an inhibition of >70% tumor growth at the end of drug treatment for the combination (Figure 3.9) and the RRM2 expression levels were synergistically reduced to

50%. Our model simulation of GTI-2040 and 5-azaC alone also reduced RRM2 expression to 45% and 50%, respectively, confirming our in vitro data.

3.5 Discussion

Chemotherapy that utilizes methods of increasing the cytotoxic effects of nucleoside analogs is especially desirable. In our previous chapter, we mechanistically enhanced protein synthesis inhibition and therefore increased the cytotoxic effects of 5- azaC via its RNA incorporation by combining with GTI-2040. We demonstrated for the first time that treatment of GTI-2040 in combination with 5-azaC is synergistic, enhancing the down-regulation of RRM2 expression by GTI-2040 alone or in combination with 5-azaC. GTI-2040 was used as a chemo-sensitizer to enhance the overall tumor cell death by 5-azaC and this effect is investigated in vivo in our studies. In our nude mice model, we demonstrated a significant decrease in tumor growth following

79 treatment with GTI-2040 and 5-azaC compared to that of the untreated control mice or the two drugs when used alone. At the end of treatment period, the tumor weight was measured and was found to be reduced by more than 3-fold for the combination treatment of GTI-2040 and 5-azaC compared to the untreated control group, GTI-2040 alone or 5- azaC alone groups. Although, lipofectamine used as a vehicle did not show significant contribution to the antitumor effects of the drugs (tumor volume and final weight), it may have potential effects on RRM2 expression level as shown in Figure 3.5. These data support the use of these two drugs in the treatment of leukemia, with minimal or no modification in the dosing regimen. GTI-2040 alone did not show a significant decrease in tumor growth reduction in this tumor model, which may be due to insufficient GTI-

2040 intracellular concentration in this MV4-11 tumor model. Treatment of tumor- bearing mice with 5-azaC at higher dose and extended period of time may result in severe loss in body weight (122). In our study, we show that the use of 20 mg/kg of 5- azaC in combination with GTI-2040, did not show any significant loss in weight for the mice. This data may show a marked ability to reduce the toxicity associated with high dose of 5-azaC while increasing the sensitization effect of GTI-2040 at a 5 mg/kg dose level.

Ribonucleotide reductase (RR) is a very important and highly regulated enzyme during DNA synthesis and repair (23) and has cooperative activity with a variety of oncogenes like c-myc (cell cycle promoter) to act as determinants in enhancing tumor progression (42). Studies in our laboratory, have found that GTI-2040 can induce down- regulation of RRM2 mRNA and protein expression in vitro which supports the fact that

80 notable inducible expression of RR may be linked directly to RNA interference that acts as negative regulators of gene expression (42). Several phase I clinical studies also demonstrated that GTI-2040 can reduce RRM2 mRNA and protein levels (31) and combinations with other agents like cytarabine in AML patients also confirm this reduction (27). Therefore, to further substantiate the modulation of RRM2 expression by either GTI-2040 alone or with 5-azaC in our developed tumor model, real time PCR and western blot was used to measure the levels of the RRM2 expression in tumor tissues after excision from mice. We showed that both GTI-2040 and 5-azaC alone and in combination can reduce the level of RRM2 expression in these tumor tissue models relative to the control. These data confirmed for the first time a reduction of RRM2 mRNA and protein levels by 5-azaC in mice tumor tissues. Our studies therefore, generate a major advancement in the pharmacology of 5-azaC and will enable effective translation of preclinical studies to clinical effectiveness.

The in vivo evaluation of the antitumor activity using xenografted mice is a fundamental step in the preclinical development process of anticancer drugs and the most usual metric for evaluation of tumor growth inhibition (130). Although, this process aids in selecting the most appropriate dosing regimen for future clinical trial, the pharmacodynamic time course of the tumor growth is often neglected. As part of the preclinical evaluation in preparation for further studies and future clinical trial, a pharmacodynamic model was developed to assess the tumor control group based on two physiologically relevant parameters (K0 and KE). In our approach, we assumed that there is a maximum tumor size, Ninf, at which tumor could not grow any more (136). When

81 tumor size approaches Ninf, tumor will grow slower, where N represents the tumor size at time t, and K0 and KE are first order parameters characterizing the rate of exponential tumor growth and cell death, respectively. This approach was flexible enough to describe accurately the growth patterns of MV4-11 cells in untreated nu/nu mice. In the implementation of the control tumor model, the growth curve was successfully modeled and the K0 and KE were estimated. In the model for the treated group, we were able to estimate the decrease of about ~50% in the RRM2 mRNA levels, following treatment with GTI-2040 and 5-azaC. The initial estimates for RRM2 expression was derived from our previous in vitro data and the turn-over rate of this protein. Further simulation of the synergistic antitumor growth effect in the combination treatment of GTI-2040 and 5- azaC, through our PD inhibition model, show an inhibition of >70% tumor growth at the end of drug treatment.

3.6 Conclusion

In summary, GTI-2040 combined with 5-azaC reduced tumor growth rate and a subsequent reduction in final tumor weight when compared to the untreated control or the two drugs alone, without loss in body weights for the mice. GTI-2040 and 5-azaC alone and their combination decreased RRM2 expression level significantly in this tumor model. A tumor growth profile and pharmacodynamic based modeling was established to evaluate the antitumor effect of the combination treatment of GTI-2040 and 5-azaC. This information will form the basis for further characterization of the mechanism for the synergistic effects of the combination treatment of other antisense with nucleoside analog in leukemia. Our model may be used prospectively to make an educated design of in vivo

82 experiments and provide valuable tools in estimating dose regimen design. A more realistic safety margin can be obtained for future preclinical and clinical studies about other combination treatment in this tumor model design.

83

Figure 3.1 Photographs of athymic nu/nu mice bearing tumor after inoculation with MV4-11 cells 16 days post-treatment.

84 800 Control GTI-Lipo + 5 AzaC 700 GTI-Lipo 600 5-azaC only 5-AzaC-Lipo 500 Lipo only

400

300 Tumorvolume (mm3)

200

100

0 0 4 8 12 16 20 24 28 32 Days after tumor injection

Figure 3.2. Tumor volume change as a function of time following various drug treatments. GTI-2040 combined with 5-azaC reduces tumor volume by 61%, relative to the untreated group and causes a >2-fold reduction when compared with 5-azaC or GTI- 2040 alone. The results are mean ± SD from six mice for each group and compared with the control.

85 16.00

14.00 # 12.00 # 10.00 * 8.00 * 6.00 *

Tumor Weight (grams) Weight Tumor 4.00

2.00

0.00 GTI-2040 (µM) - - + - - + 5-azaC (µM) - - - + + + Vehicle - + + - + +

Figure 3.3. Tumor weight change following various drug treatments. Tumor weight was reduced by 3-fold in the combination treatment of GTI-2040 with 5-azaC and >2-fold reduction, when compared with GTI-2040 or 5-azaC alone. The Vertical bars are mean ± SD from six tumors from each group.*p<0.05 compared with the control, #p<0.05, compared to GTI-2040 or 5-azaC alone.

86

Control GTI-lipo + 5-AzaC 5-AzaC only 5-AzaC-Lipo GTI-Lipo Lipo-only 140 120

100

80

60 (grams)

40 Mice body Mice weight 20 0 0 4 8 12 16 20 24 28 32 Days after tumor injection

Figure 3.4. Mouse body weight change as a function of time following various drug treatments. Mouse body weights showed a minimal toxicity as manifested by a minimum weight loss (<2%). The results are mean ± SD of six mice for each group.

87 A

1.5

1 * *

0.5 RRM2 mRNA ratio

0 GTI-2040 - - + - + - 5-azaC - - - + + + Vehicle - + + - + +

B

GTI-2040 - - + + - - - - + + - - 5-AzaC - - + + + + + + - - - - Lipo - - + + - - + + + + + + RRM2

GAPDH

RRM2/GAPDH 1.0 0.2 0.4 0.4 0.6 1.1 Ratio

Figure 3.5. Changes in RRM2 mRNA (A) and protein (B) levels in tumor tissue obtained from MV4-11 engrafted mice following various drug treatments. GTI-2040 combined with 5-azaC reduces RRM2 mRNA level in tumor tissues by about 50%, relative to the saline treated control. Remarkably, 5-azaC alone without vehicle also decreased RRM2 mRNA levels by 50%. RRM2 mRNA levels were normalized by abl and vertical bars represent mean ± SD from three mice tissues. *p<0.05, as compared with the untreated control. RRM2 protein expression decreased the protein level by 80% in the combination; GTI-2040 alone decreased by ~40% and 5-azaC also reduced the protein expression levels about 60%. RRM2 protein levels were quantified by densitometry and expressed as the ratios over the loading control GAPDH.

88 K0 Tumor Size (N) KE

dN N  K0 *(1 )* N  N * K E dt Ninf

Figure 3.6 Control tumor growth model.

89

Figure 3.7. Tumor growth in MV4-11 engrafted mice as a function of time as analyzed by the control tumor growth model. (A) Tumor size changes as a function of time as fitted to the tumor growth model and (B) Tumor weight change as a function of time.

90 RRM2 K 0 K Tumor Size E (N)

dN1 RRM 2GTI 20405azaCcomb N1  K0 * 0 *(1 )* N  N * KE dt RRM 2 Ninf

dN2 RRM 2GTI 2040 N2  K0 * 0 *(1 )* N  N * KE dt RRM 2 Ninf

dN3 RRM 2GTI 2040 N3  K0 * 0 *(1 )* N  N * KE dt RRM 2 Ninf

Figure 3.8 Tumor growth model in treated animals.

91 101.4

101.2 Control 101 GTI+5-azaC GTI alone 100.8 5-azaC alone 100.6

100.4

100.2 Relative tumor Relative volume (%)

100

99.8 0 5 10 15 20 25 30 35 Days after tumorhr injection

Figure 3.9. Simulated tumor growth-time profiles following various drug treatment. Synergistic inhibition of tumor growth was seen in the combination treatment of GTI- 2040 with 5-azaC in MV4-11 leukemia tumor xenografts.

92 CHAPTER 4

RIBONUCLEOTIDE REDUCTASE IS A NOVEL TARGET OF 5-AZACYTIDINE IN VITRO AND IN VIVO

4.1 Abstract

5-Azacytidine, an azanucleoside synthesized over five decades ago, is now approved to be used for the treatment of myelodysplastic syndrome and other hematological malignancies. It inhibits DNA synthesis through conversion to decitabine triphosphate and subsequent DNA incorporation; however, its precise mechanism of action remains unclear. Ribonucleotide reductase (RR) is a highly regulated enzyme comprising two subunits RRM1 and RRM2 that provides the deoxyribonucleotides required for DNA synthesis/repair. RR appears to play a role in the critical early events of tumor promotion. In this study we report the effects of 5-azaC on RR, which serves as a novel post-transcriptional target of 5-azaC both in-vitro and in-vivo. Leukemia cells,

MV4-11 and K562 were continuously exposed to 5-azaC in a dose– and time dependent manner. Western blot analysis was used to determine RRM2 and DNMTs protein expression. A LC-MS/MS method published in our laboratory was used to quantify endogenous NTPs and dNTPs levels. Quantitative RT-PCR was used to measure RRM2 mRNA levels. In vivo studies were performed on tumor tissues excised from experiments conducted by inoculating Nu/Nu mice with MV4-11 leukemia. Ex-vivo studies were performed on mononuclear cells (MNCs) from bone marrow samples of 6 untreated acute 93 myeloid leukemia (AML) patients obtained from The Ohio State University Leukemia

Tissue Bank. Evidence of 5-azaC RNA incorporation was performed on RNA hydrolysate by chemical reduction and LC-MS/MS. In MV4-11 cell line, 5-AzaC was found to down-regulate RRM2 mRNA levels and protein expression significantly

(p<0.05) in a dose- and time manner, while the same in K562 cells was not as significant.

5-AzaC treatment causes a major perturbation of dNTP/NTP pools. 5-AzaC was found to destabilize RRM2 mRNA, which is influenced by protein synthesis inhibition. Evidence of 5-azaC RNA incorporation was found. Knockout of RRM2 expression by siRRM2 blocks the conversion of 5-azaC to DAC and DAC-TP, and abolishes the effects of 5- azaC-mediated decrease of DNMT 1 and 3a. In tumor tissues, 5-azaC decreased RRM2 mRNA level and protein expression by 50-75%. In ex vivo studies, samples from five patients (83%) showed a reduction in RRM2 mRNA levels and consistent with the reduction in RRM2 mRNA, there was a proportionate decrease (>90%) in the protein expression. We herein report the down-regulation of 5-azaC on RRM2 for the first time via its RNA mechanism, and such serves as a novel post-transcriptional target of 5-azaC.

Our findings will bear clinical relevance and serve as additional important surrogate biomarker for 5-azaC.

4.2 Introduction

5-Azacytidine (5-AzaC; Vidaza) (Figure 4.1), a pyrimidine analog that was synthesized 45 years ago, has been widely used for the treatment of hematological malignancies, including acute myeloid leukemia (AML) (9). Currently, 5-azaC was approved by the United States Food and Drug Administration (FDA) for the treatment of

94 myelodysplastic syndrome (MDS) as a hypomethylating agent (9). 5-AzaC acts as a

DNA hypomethylating agent, presumably through first conversion to its deoxy analog decitabine, at low doses (10-75 mg/m2/day). 5-AzaC can reactivate previously silenced genes, including tumor suppressor genes, restore apoptosis and inhibit proliferation of cancer cells (13). Despite decades of efforts made to delineate the mechanisms of action for 5-azaC in terms of its interference with RNA and DNA metabolism (91), the precise basis of its clinical efficacy remains uncertain (10). A comprehensive understanding of the pharmacology of 5-azaC is therefore critical in defining its exact mode of action and will further advance its clinical use alone and in combination with other drugs.

5-AzaC is transported into cells by human equilibrative nucleoside transporters

(hENT) (137-139). Once inside the cell, 5-azaC is first anabolized to its nucleoside monophosphate by uridine-cytidine kinase enzyme and eventually phosphorylated by diphosphate kinase to, 5-aza-CTP, which is thought to be incorporated into RNA (140).

About 80-90 % of 5-azaC was found to be incorporated into RNA, which can increase the rate of apoptosis resulting in protein and nucleic acid metabolism disruption (9, 10).

Inhibition of protein synthesis via mRNA destabilization is probably the main target accounting for the strong cytotoxic activity and protein inhibition effect of the drug (113,

115). Since this effect can be suppressed by actinomycin D (transcription inhibitor), the incorporation of 5-azaC into RNA is crucial for its protein inhibition process (114, 115).

A fraction (10-20%) of 5-azaC intermediate product 5-aza-CDP is indirectly and reversibly converted to decitabine (DAC) via 5-aza-dCDP by ribonucleotide reductase

(RR), followed by further phosphorylation to 5-azaC-dCTP (DAC-TP) (9). Decitabine

95 (DAC) is another pyrimidine analog, which differs structurally from 5-azaC by only one hydroxyl group at the 2’ position of the ring, that allows for DNA incorporation.

DAC has significant activity against a variety of solid tumors and hematologic malignancies (141-145) and also has recently been approved by the FDA for the treatment of myelodysplastic syndrome (MDS). DAC was found to have two different pharmacologic actions. At higher doses (i.e., 50-100 mg/m2/day), it acts primarily as a cytotoxic agent due to its incorporation into DNA, leading to inhibition of DNA synthesis and cell death (15, 16), while at lower doses (i.e., 5-20 mg/m2/day) it induces DNA demethylation, resulting in reactivation of hypermethylation-associated silencing of tumor suppressor genes (13, 14). DAC undergoes a three-step phosphorylation to its active anabolite, DAC-TP in the cell. Therefore, incorporation of DAC-TP into DNA results in covalent trapping of DNA methyltransferases (DNMT1, 3a/b), leading to the depletion of the enzyme activity and causing demethylation of the DNA (146). While recent attention has focused on the hypomethylating effects of 5-azaC via its DNA pathway, the RNA pathway is less understood. Therefore, we pursued further characterization of its RNA pathway and discovered a previously uncharacterized role of

5-azaC in the regulation of ribonucleotide reductase.

Ribonucleotide reductase (RR) (Figure 4.2) is a highly regulated enzyme that catalyzes the reduction of ribonucleotides to their corresponding deoxyribonucleotides, which is the rate- limiting step for DNA synthesis and repair (23), and also plays a role in the critical early events of tumor promotion (24-26). RR is composed of two non- identical dimeric subunits RRM1 and RRM2. They are encoded by different genes and

96 their expression is required for the enzyme activity (31). RRM1 protein is constitutively active throughout the cell cycle, whereas RRM2 is only expressed in the late G1 and early S phase. RRM1/RRM2 complex serves as a major provider of dNTPs for DNA replication during S phase (26, 32). Relatively recently, Tanaka et al., (34) discovered a p53-inducible analog of RRM2 subunit, p53R2 (33, 34), which complexes with RRM1 in non-proliferating cells and provides dNTPs for p53-dependent DNA repair and mitochondrial DNA synthesis (35). It is likely that there exist two independent pathways involving RRM2 and p53R2 to supply human cells with dNTPs for DNA synthesis and repair. However, RRM2 is mainly involved in DNA replication at S phase, while p53R2 is involved in DNA repair at arrested G1 and G2 phases (34, 147).

Over-expression of RR is commonly found in malignant cells (24-26) and increases the endogenous pool of deoxynucleoside triphosphate (dNTP), therefore increasing the rate of DNA synthesis to serve the proliferative nature of malignant cells.

Additionally, RR over-expression is also a potential mechanism of chemoresistance to nucleoside analogs competing for DNA incorporation (27). These important roles in cell proliferation and DNA repair and its over-expression in malignant cells has made RRM2 a target for chemotherapeutic strategies (28, 29). The mechanisms that regulate

RRM1/RRM2 complexes are very critical for the cell, because imbalanced dNTP supply can lead to genetic abnormalities and cell death (148). Studies have demonstrated that inhibition of RRM2 with agents leads to a depletion of nucleotide pools and arrest of the cell cycle in S phase, thus providing antiproliferative and antineoplastic benefits (26, 32).

However, several established RR inhibitors are associated with considerable limitations.

For example, hydroxyurea, one of the most commonly used RRM2 inhibitors, is limited

97 by its low affinity for RR, high hydrophilicity, short half-life and early development of resistance (23, 32, 36). A new class of RRM2 inhibitors, such as antisense GTI-2040

(LOR-2040) is still in development. Although, antisense therapy represents a promising genetic intervention strategy for chemotherapy, while producing less side effects and high selectivity (8), it faces challenges in their hydrophilic nature, low achievable target drug concentration, and lack of efficient delivery method. Another class of RR inhibitors is the iron chelators, such as triapine, which is also under development; however, their effective

RR inhibitory properties are still being evaluated.

Major interest has been focused on the ability of 5-azaC to inhibit cytosine methylation, which affects gene expression and differentiation via its DNA incorporation; however, RNA incorporation of 5-azaC is also very crucial for its cytotoxic inhibition processes. Chapter 2 of this dissertation discussed the combination of 5-azaC with GTI-

2040 and the involvement of 5-azaC on RRM2 expression was implied for the first time.

In an effort to study the 5-azaC RNA effect, herein we report the discovery of RRM2 as a novel post-transcriptional target of 5-azaC and also to increase our understanding of the mechanism involved in this 5-azaC-RNA mediated outcome. Furthermore, the mechanisms of action of 5-azaC and DAC, that might explain their equivalence and/or clinical differences, are being examined. In the present studies, 5-AzaC was found to dramatically decrease RRM2 expression in cell lines and tumor xenografts model. Most notably, RRM2 down-regulation also occurs ex vivo in human bone marrow cells treated with 5-azaC. The mechanism might be through direct incorporation of 5-azaC into RNA, which results in mRNA destabilization and shortened half-life. In addition, the

98 involvement of 5-azaC in RNA regulation will be used to establish a link between RNA interference and certain miRNA perturbations and provide evidence for endogenous miRNA involvement in transcriptional gene silencing. Using an HPLC-MS/MS method and chemical reduction, we confirmed the presence of 5-azaC in total RNA. Our findings will bear clinical relevance and serve as additional surrogate biomarker for 5-azaC.

4.3 Materials and Methods

4.3.1 Chemicals

5-Azacytine and decitabine were obtained from The National Cancer Institute

(Bethesda, MD) and a stock solution was prepared in dimethyl sulphoxide (DMSO) and further dilution was made with phosphate buffered saline (Invitrogen, Rockville, MD).

Triapine and GTI-2040 (LOR-2040) were also obtained from The National Cancer

Institute. The siRRM2 now, LOR-1284 was obtained from Invitrogen, Rockville, MD.

4.3.2 Instrumentation

For qualification and DAC measurement, the LC-MS system used consisted of a

Finnigan TSQ Quantum EMR Triple Quadruple mass spectrometer (ThermoFinnigan) coupled to Shimadzu HPLC system (Shimadzu), which was equipped with a CBM-20A system controller, an LC-20 AD pump, a SIL-20AC autosampler, CTO-20A column oven, DGU-20A5 degasser and FCV-11AL valve unit. Analysis of dNTP/NTP and DAC-

TP were done on an ion-trap LC-MS/MS system (LCQ, Thermo Scientific, San Jose,

CA).

99 4.3.3 HPLC and MS conditions

5-Azacytidine, uridine, and dihydro-5-azaC, HD-5-azaC or dideuterio-5-azaC were separated on a 250×2.1mm Hypersil Aquasil C18 5mm stainless steel column

(Thermo Hypersil-Keystone), which was coupled to a 2mm Aquasil precolumn (Thermo

Hypersil-Keystone). Isocratic program was used to qualify 5-azacytidine. The mobile phase consisted of 10mM ammonium formate containing 1% methanol and flow rate was set at 0.2 mL/min. The mass spectrometer was operated in the positive ESI mode with a helium pressure of 20 psi, a typical electrospray needle voltage of 4988 V, a sheath nitrogen gas flow of 27 (arbitrary unit) and a heated capillary temperature of 300 °C. 5-

Azacytidine, uridine, dihydro-5-azaC, HD-5-azaC and dideuterio-5-azaC were analyzed by multiple reaction monitor (MRM) mode using ion transitions at m/z 245.10>113.16,

245.10>113.16, 247.10>115.1, 248.3>116.2 and 249.2>117.3 respectively. The relative collision energy of 5-azacytidine, uridine, dihydro-5-azaC, HD-5-azaC and dideuterio-5- azaC were 15%, 15%, 16%, 16% and 15%, respectively. All operations were controlled by Finnigan Xcalibur software on a Windows NT 4.0 system.

4.3.4 Cell culture and treatments

The human leukemia cell lines MV4-11 and K562 were obtained from American

Type Culture Collection (ATCC). Cells were cultured in RPMI 1640 media (Supplied by

Tissue Culture Shared Resource, The Ohio State University Comprehensive Cancer

Center) supplemented with 10% fetal bovine serum (FBS) (Invitrogen, Rockville, MD),

L-glutamine (Invitrogen) and Penicillin-Streptomycin antibiotics (Gibco, Rockville,

MD). The cell lines were maintained at 37 °C in a humidified environment with 5% CO2.

100 Viability and cell counts were determined using trypan blue dye exclusion assay. GTI-

2040 was transfected into cells by electroporation device according to manufacturer’s instruction. Daily treatment with 5-azaC was necessitated to ensure continued exposure due to the short half-life (~30 to 60 min) of the drug (17, 149). Mononuclear cells

(MNCs) from bone marrow samples of 6 untreated AML patients were obtained from

The Ohio State University (OSU) Leukemia Tissue Bank. MNCs were prepared and cultured as previously described (150). All experiments with these cells were performed in accordance with the protocols approved by The Ohio State University institutional review board (IRB).

4.3.5 Western blot analysis

5×106 MV4-11 and K562 cells were serially treated with 5-azaC at doses of 0, 1,

5, 10, 20 and 30 µM for 24, 48 and 72 hour. Both cell lines were also treated independently with GTI-2040, triapine and decitabine at 0, 1, 5, 10 and 30 µM for 72 hour. Mononuclear cells (MNCs) bone marrow samples from AML patients were exposed to 5 and 10 µM 5-azaC at 24 and 48 hour. In another design, MV4-11 cells were also pretreated w/o 10 µM siRRM2 for 24 hour, followed by treatment with 5-azaC at 10

µM for 2, 4 and 24 hour, and then after 24 hour, to a continuous exposure of a second dose of 10 µM 5-azaC for an additional 26 and 28 hour. Exposure to DAC at 10 µM was used as a positive control. Cells were harvested and washed with 1 mL ice-cold PBS, and centrifuged at 1000 g for 5 minutes at 4 ºC. The pellet was obtained and re-suspended in

100 μL lysis buffer (50 mM PH 7.6 Tris-HCl, 250 mM NaCl, 5 mM EDTA, 2 mM

Na3VO4, 50 mM NaF and 1% protease inhibitor cocktail) (P8340, Sigma) for 30 minutes

101 on ice. The lysate was sonicated for 10 seconds. Total protein concentration was determined using the BCA protein assay method (Pierce, Rockford, IL). Equal amounts of protein for each sample were incubated with 6x SDS loading buffer (100 mM, pH 6.8

Tris, 200 mM DTT, 4% SDS, 20% glycerol, and 0.015% bromphenol blue) and boiled for

5 minutes. The proteins were then separated on 4-15% SDS-polyacrylamide gels (Biorad,

Herculeus, CA) and transferred to nitrocellulose membranes (Amersham, Piscataway,

NJ). The RRM2 and DNMT1/3a proteins were recognized following treatment with a goat antihuman RRM2 polyclonal antibody (E-16) (Santa Cruz Biotechnology, Santa

Cruz, CA) and a rabbit antihuman DNMT1/3a polyclonal antibody (New England

Biolabs, MA), respectively as the first antibody. This was followed by a horseradish peroxidase (HRP) conjugated anti-goat IgG and anti-rabbit IgG secondary antibody, respectively. RRM2 (MWT 45,000 dalton) and DNMT1/3a (MWT 180,000/120,000 dalton), respectively, were detected by ECL (Amersham, Arlington Heights, IL) and

GAPDH was used as the internal loading control. RRM2 and DNMT1/3a protein expressions were quantified by densitometry and normalized to GAPDH.

4.3.6 Determination of intracellular dNTP/NTP pools, DAC and DAC-TP levels

10×106 MV4-11 and K562 cells were treated with 5-azaC at doses of 0, 1, 5, and

10 µM for 72 hours for dNTP and NTP analysis. Cells were then lysed and the dNTPs/NTPs, were extracted and quantified by our previously described method (117).

Briefly, cells were counted and monitored for viability using trypan blue exclusion test.

Following centrifugation at 1000 g for 5 minutes, cell pellets were washed with phosphate buffered saline (PBS) and deproteinized with an addition of 1 mL 60%

102 methanol. The resulting solution was vortex-mixed for 20 seconds, incubated in -20 °C for 30 minutes and sonicated for 15 minutes in an ice bath. Cell extracts were centrifuged at 1000 g for 5 minutes at 4 °C and the supernatant was separated and dried under a stream of nitrogen. The residues were reconstituted with 300 μL of water, vortex-mixed for 20 seconds and the cell extracts were centrifuged at 1000 g for 5 minutes at 4 °C. A

50 μL aliquot of the resulting supernatants was injected into an ion-trap LC-MS/MS system (LCQ, Thermo Scientific, San Jose, CA) for dNTP and NTP measurements.

For the measurement of DAC and DAC-TP levels, 10×106 MV4-11 cells were pre-treated in the presence or absence of 10 µM siRRM2 for 24 hours, followed by treatment with 5-azaC at 10 µM for 2, 4 and 24 hour, then after 24 hours, to a continuous exposure of a second dose of 10 µM 5-azaC for an additional 26 and 28 hours.

Furthermore, 10×106 MV4-11 cells in 1 mL PBS were heat deactivated for 6 minutes at

100 °C and exposed to 10 µM 5-azaC or DAC at 37°C, to ascertain their stability. Cells were harvested, washed with phosphate buffered saline (PBS), and the triphosphates were extracted and quantified using the same procedure as described above. DAC was measured by Triple Quadruple LC/MS-MS system (ThermoFinnigan, San Jose, CA) and

DAC-TP by ion-trap LC-MS/MS system (LCQ, Thermo Scientific, San Jose, CA).

4.3.7 RNA Isolation and RT-qPCR

To measure intracellular RRM2 mRNA level, 5×106 MV4-11 and K562 cells were treated with 5-azaC at doses of 0, 1, 5, 10, 20, 30 µM for 6, 12, 24, 24, 48 and 72 hours and analyzed for RRM2 mRNA expression levels. Total RNA was isolated using

103 Trizol reagent (Invitrogen). Briefly, cell lysate was treated with chloroform and the total

RNA was precipitated with isopropyl alcohol, followed by a washing step with 75% ethanol. RNA was then dissolved in RNase free water, its concentration and purity measured by a Nanodrop spectrophotometer (Nanodrop Technologies, Wilmington,

Delaware). cDNA was synthesized from 2 µg total RNA using Moloney murine leukemia virus reverse transcriptase (Invitrogen). The cDNA templates and primers were then mixed with reagents from a SYBR Green PCR Master Mix. Reactions were carried out in triplicate in ABI prism 7700 sequence detector (Applied Biosystems), and data were analyzed by comparative CT method. The amount of RRM2 mRNA in each sample was normalized with respect to an internal control, abl. The relative changes in treated groups were expressed as a percentage of untreated control (arbitrarily set at 1). The results were expressed as the mean ± SD from triplicate determinations.

4.3.8 mRNA stability assay

3×106 MV4-11 cells were treated with 1µM 5-azaC for 4 hours. At the same time, 1µM actinomycin D (Sigma) was used to block new transcription. Cells were harvested at various time points (0, 0.5, 1, 2, 4 hours) after actinomycin D treatment.

Total RNA was isolated and RRM2 and abl mRNA levels were then quantified by qRT-

PCR as previously described. Their half-lives were calculated from the plots of their mRNA levels as a function of time. In order to determine whether proteins were involved in the effect of 5-azaC on RRM2 mRNA stability, we also performed the experiment with a potent protein synthesis inhibitor, cycloheximide (Sigma). MV4-11 cells were pretreated with 10µg/mL cycloheximide for 2 hours, 1µM 5-azaC for additional 2 hours,

104 followed by treatment with 1µM actinomycin for 4 hours. RRM2 and abl mRNAs were then measured by qRT-PCR for each condition.

4.3.9 RNA reduction and hydrolysis

MV4-11 cells were treated with 300 µM 5-azaC. Cells were harvested after 6 hours and the total RNA was isolated as previously described. 10 μL of 10 mg/ml NaBH4 or NaBD4 was added into 1 μg of total RNA. The mixture was incubated at room temperature for 15 minutes and then neutralized to pH 7.0 by the addition of HCl. RNA hydrolysis was performed as previously described (151). Briefly, a 1/10 volume of 0.1 M ammonium acetate (pH 5.3) and two units of nuclease P1 were added to 1 μg of total

RNA. The mixture was incubated at 45°C for 2 hours followed by the addition of 1/10 volume of 1 M ammonium bicarbonate and 0.002 unit of venom phosphodiesterase I. The resulting mixture was incubated at 37°C for 1 hour. After adding 0.5 unit of alkaline phosphatase, the mixture was incubated at 37°C for another 1 hour. A 50 μL of resulting mixture was injected into LC-MS for further qualification.

4.3.10 Xenograft mouse model

Xenograft athymic Nu/Nu mice (4-6 weeks old, 18-22 g) were obtained from

Charles River Laboratory (Wilmington, MA). All experiments were conducted in accordance with the guidelines of the Association for Assessment and Accreditation of

Laboratory Animal Care International (AAALAC). Briefly, MV4-11 Cells (10x106 cells) was suspended with Matrigel (Becton Dickinson) and PBS in a 1:1 ratio and implanted subcutaneously into the right and left flank of mice, respectively. Tumor diameters were

105 measured and body weights were monitored weekly after implantation. Treatments were initiated when the average tumor size reached 100 to 150 mm3. 5-AzaC was injected intraperitoneally (i.p) at the dose of 20 mg/kg, twice a week. After 32-day treatment, the mice were sacrificed and the tumor tissues were snapped frozen in liquid nitrogen and stored in -80°C for further analysis.

4.4 Results

4.4.1 5-AzaC decreases RRM2 protein level in leukemia cells

To investigate the effects of 5-azaC on RRM2, we first examined RRM2 protein expression in two leukemia cell lines following 5-azaC treatments, replenished at 24 hours intervals, at a dose range of 1 to 30 µM, which are in the clinically relevant concentration range based on previous pharmacokinetics studies (90). 5-AzaC was found to cause a robust decrease of RRM2 protein in a dose- and time-dependent manner in both MV4-11 (Figure 4.3) and K562 (Figure 4.4) cell lines, with MV4-11 cells being more sensitive. In MV4-11 cell line, RRM2 protein was reduced by 60% after 48 hours treatment with 5 µM 5-azaC and was completely depleted after 72 hours. However, K562 cells showed only minimal reduction in RRM2 protein expression even at high concentrations (20 or 30µM) of 5-azaC for 48 hour. Incubation for 72 hours resulted in further decrease of RRM2 protein level by 50% in these cells.

We then asked whether the down-regulation effect on RRM2 protein expression was specific to 5-azaC or it also applies to other nucleoside analogs, such as decitabine

(DAC, 5-aza-2’-deoxycytidine). In contrast to 5-azaC, DAC was found not to induce any

106 reduction in RRM2 protein level in MV4-11 and K562 cell lines, not even at high concentration of 30 µM and at exposure time of 72 hours. Next, we evaluated the potency of 5-azaC in comparison with two established RRM2 inhibitors, triapine and

GTI-2040. Consistent with the previous report of triapine in leukemia cells (152, 153), we indeed showed that triapine decreased RRM2 protein level in MV4-11 cells but with no decrease in K562 cells. We also confirmed that the antisense of RRM2, GTI-2040, down-regulated RRM2 protein expression in both of these two leukemia cell lines (Figure

4.5A) (32). Interestingly, we found that 5-azaC was significantly more potent than triapine and GTI-2040, at least in MV4-11 cells (Figure 4.5B). Collectively, these data demonstrate that 5-azaC down-regulates RRM2 protein in leukemia cells.

4.4.2 5-AzaC decreases RRM2 mRNA levels in leukemia cells

We next investigated the mechanism through which 5-azaC down-regulates

RRM2 protein. First, we observed that 5-azaC induces linearly a dose- and time- dependent decline in total RNA levels in both MV4-11 and K562 cell lines (Figure 4.6A).

By quantitative RT-PCR, we showed that 5-azaC decreased RRM2 mRNA by ~40, 65,

85 and 92% at 5, 10, 20, and 30 µM, respectively, at 24 hours with no detectable effect at

1 µM in MV4-11 cells (Figure 4.6B). In K562 leukemia cells, 5-azaC, starting at 10 µM, was found to induce a 30-60% down-regulation of RRM2 mRNA. Additionally, we demonstrated a time-dependent decrease of RRM2 mRNA level at 5 µM for both MV4-

11 and K562 cell lines (Figure 4.6C). Taken together, these data suggest that 5-azaC regulates RRM2 at transcriptional level.

107 4.4.3 5-AzaC perturbs the intracellular ribonucleotide (NTPs) and deoxyribonucleotide (dNTPs) pools in leukemia cells

Since RRM2 is required for catalyzing the reduction of ribonucleoside diphosphates to the corresponding deoxyribonucleoside then to deoxyribonucleotides

(118, 119), we subsequently reasoned that 5-azaC should also alter the intracellular dNTPs/NTP pools. Thus, leukemia cells were exposed to different doses of 5-azaC for

72 hours and the dNTP and NTPs were extracted and measured by a non-radioactive LC-

MS/MS method developed in our laboratory. As expected, a significant dose-dependent decrease in the intracellular dTTP, dATP and dCTP levels was observed in both MV4-11 and K562 cells (Figures 4.7A and 4.7B). Interestingly, 5-azaC also reduced the UTP and dGTP/ATP in a dose-dependent manner in MV4-11 cells, whereas GTP and CTP remained unperturbed (Figure 4.7C). K562 cells showed similar results, except that the

GTP level showed a dose-dependent decrease (Figure 4.7D). It has previously been reported that perturbation of the dNTP/NTP pools could enhance the accumulation of nucleoside analogs into DNA, thereby increasing their cytotoxic effects (154). Our data therefore suggest that 5-azaC’s cytotoxic effects might be at least in part due to its ability to reduce RRM2 and exert its effect on the downstream dNTP/NTP pools.

4.4.4 5-AzaC destabilizes RRM2 mRNA

To address whether 5-azaC-induced RRM2 mRNA down-regulation is due to its effects on mRNA stability, we analyzed the half-life of RRM2 mRNA (t1/2) following treatment with actinomycin D, which blocks the de novo transcription. The levels of

RRM2 mRNA and the housekeeping control abl mRNA was monitored by quantitative

108 RT-PCR. The RRM2 mRNA was found to be remarkably destabilized by 5-azaC with an estimated t1/2 = 2.5 hour, as compared to that in untreated group (t1/2 = 4.2 hour) (Figure

4.8A). The half-life of the housekeeping control abl, GAPDH mRNA (t1/2 = 2.5, 4.2 hour for both groups), however, was not changed as expected. This suggests that RRM2 mRNA, but not abl or GAPDH mRNA, is a specific target for 5-azaC. We also determined whether the 5-azaC-induced destabilization of RRM2 mRNA depends on protein synthesis. Surprisingly, pretreatment with specific protein synthesis inhibitor, cycloheximide, was found to facilitate 5-azaC’s destabilizing effects on RRM2 mRNA

(Figure 4.8B). RRM2 mRNA level was further decreased by 70%, when protein synthesis is blocked. Therefore, protein synthesis is essential for the 5-azaC-mediated reduction in

RRM2 mRNA stability.

4.4.5 5-AzaC is incorporated into RNA

It has been previously shown that 5-azaC is incorporated into RNA by measuring the incorporation of radiolabeled 5-azaC (11, 155). In order to further confirm the incorporation of 5-azaC into RNA, we employed a chemical method coupled to LC-

MS/MS assay. Since 5-azaC is chemically rather labile (109, 113, 156), it is not expected to be detectable in the total RNA acid hydrolysate. However, it can be readily converted to stable dihydro-5-azaC (157) and dideuterio-5-azaC with NaBH4 and NaBD4, respectively, in aqueous solution under mild condition and these products can be monitored by LC-MS/MS as a means to probe 5-azaC RNA incorporation. Therefore, we first examined the LC-MS retention times and monitored the ion transitions of standard dihydro-5-azaC, HD-5-azaC and dideuterio-5-azaC in the mobile phase and found them

109 to be at about 4.5 min. Then the total RNA obtained from the 5-azaC treated and untreated (control) MV411 cells were treated with NaBH4 followed by acid hydrolysis.

No peak with significant intensity was observed at the retention time at about 4.5 min at the ion transitions for 5-azaC or dihydro-5-azaC in the control RNA samples. However, a peak with significant intensity was observed at 4.56 min corresponding to dihydro-5- azaC with the correct ion transition in the treated samples, which indicated the presence of 5-azaC in the total RNA. In order to further substantiate these data, we also used deuterium labeled sodium borohydride NaBD4 to reduce the blank and 5-azaC treated total RNA. Since each of the deuterium atom increases the mass by one relative to the hydrogen atom, we would expect that the observed new peaks, if they relate to 5-azaC, would have a proportional increase in the mass unit corresponding to dihydro-5-azaC,

HD-5-azaC and dideuterio-5-azaC Additionally, since NaBD4 could undergo partial H-D exchange during reduction, HD-5-azaC was also monitored. Following NaBD4 treatment, similar to the NaBH4 experiment no significant peak was observed at 4.7 min for any of the reduced products in the blank. However, a significant peak was observed at

4.7 min with ion transitions corresponding to dihydro-5-azaC, HD-5-azaC and dideuterio-

5-azaC. These data indicated the presence of 5-azaC in the total RNA of the 5-azaC treated samples (Figure 4.9) and confirmed the expected 5-azaC incorporation into total

RNA in a human AML cell line.

110 4.4.6 Inhibition of RR expression decreases the levels of DAC and DAC-TP formed following 5-azaC treatment

Ribonucleotide reductase converts a fraction (10-20%) of 5-azaC intermediate product 5-aza-CDP to DAC through reversible indirect conversion via 5-aza-dCDP (9,

11). Therefore, to confirm that RRM2 is indeed a target of 5-azaC, we used siRRM2 to block RRM2 expression and obstruct the formation of DAC from 5-azaC. Following treatment of MV4-11 cells with siRRM2 for 24 hours followed by 5-azaC treatment, cells were harvested and analyzed for DAC and DAC-TP levels. Inhibition of RRM2 expression levels significantly hindered the conversion of 5-azaC to DAC or DAC-TP with the amount either much reduced or not detectable. Figure 4.10A shows 0.55 and

0.07 pmol/106 cells of DAC formed at 2 and 4 hours, respectively, for the samples without RRM2 expression inhibition, as compared to 0.11 and 0.01 pmol/106 cells formed for samples with inhibition, respectively. DAC-TP was detected at 2 hours without RRM2 inhibition but not detectable following RRM2 inhibition (Figure 4.10B).

Interestingly, exposure of the cells to a second dose of 5-azaC did not result in further formation of DAC or DAC-TP. Therefore, these data supplement the strong inhibition of

RRM2 expression by 5-azaC in a time-dependent manner (>24 hours) as shown in Figure

4.3 and further confirms RRM2 as a valid target for 5-azaC. In addition, after incubation with DAC or 5-azaC in heat inactivated cell lysate, it was found that these nucleosides were stable during the time course of the experiment (Figure 4.10C), alleviating a potential problem that the failure of detection of DAC was due to degradation of DAC or

5-azaC.

111 4.4.7 RRM2 expression knockout abolishes 5-azaC mediated down-regulation of

DNMTs

It has been shown that down-regulation of DNMTs by DAC requires DNA incorporation and formation of DNA-drug-enzyme complex (158). We further postulate that blockage of RRM2 expression will result in abolishing the down-regulation DNMTs effect of 5-azaC, since its formation of DAC or DAC-TP is blocked. MV4-11 cell samples were first exposed to siRRM2 for 24 hours, and then 5-azaC at 10 µM was added. Cells were harvested at 0, 2 and 4 hours for DNMTs protein expression levels measurement. There was no down-regulation of DNMTs (1 and 3a) protein expression levels in cells with RRM2 inhibition compared to those without RRM2 inhibition (Figure

4.11). Thus, inhibition of DNMTs of 5-azaC requires its conversion to DAC.

4.4.8 5-AzaC decreases RRM2 mRNA and protein in MV4-11 xenografts

Since 5-azaC was shown to significantly lower RRM2 mRNA and protein levels in leukemia cell lines, we next examined this activity in a MV4-11 engrafted nude mice model. Consistent with our in vitro data, twice a day intraperitoneal administration of 20 mg/kg 5-azaC resulted in a significant decrease of RRM2 protein in tumor xenografts, when compared to the untreated controls (~75%, p=0.05, Figure 4.12A). Furthermore, the

RRM2 mRNA level was dramatically reduced in 5-azaC-treated tissues, relative to the untreated group (~47%, p=0.01, Figure 4.12B).

112 4.4.9 5-AzaC reduces RRM2 mRNA and protein in bone marrow cells ex-vivo

The consistent results between the in vitro cell data and tumor xenografts prompted us to examine ex-vivo human bone marrow samples obtained from AML patients. The samples were then cultured, followed by treatment with 5-azaC. As shown in Figure 4.13, notably five patients (83%) showed a reduction (ranging from 60-90%) in

RRM2 mRNA levels. Therefore, overall there was a significant decrease in RRM2 mRNA levels within individual samples in a time- and dose dependent manner.

Consistent with the reduction in RRM2 mRNA, there was a corresponding decrease

(>90%) in the protein expression. These data further supports our in vitro and in vivo animal data and provides clinical relevance of our study.

4.4.10 5-AzaC treatment induces up-regulation of endogenous miR29b and miR181a

To further investigate the mechanism of 5-azaC regulation of RNA, miRNA analysis and validation was performed. Following treatment of MV4-11 cells with 5-azaC at serial dose and time, miR29b levels was found to be up-regulated by 30-50% at 4 and

24 hours of 10 µM 5-azaC with no change at 5 µM 5-azaC, while miR181a levels increased by 50% at 24 hours at both 5 and 10 µM 5-azaC but with no change at 4 hours

(Figure 4.14). Mononuclear bone marrow cells from AML patients were also exposed to

5-azaC, as shown in Figure 4.15, relative to the control, 75% of the samples (n = 4) showed an increase in miR29b levels, ranging from 20 to 60%, but one sample show a

50% decrease, while miR181a levels increased from 20 to >90%. The up-regulation of the miRNA in the mononuclear bone marrow cells was more significant at a shorter

113 exposure time (24 hours) compared to longer time (48 hours). Therefore, 5-azaC can induce up-regulation of miR29b and miR181a in both in vitro and in vivo models.

4.5 Discussion

5-AzaC and DAC have long been considered as similar antitumor agents that possess both DNA synthesis inhibition and hypomethylation activities, despite showing different toxicity profiles (9) and rather different chemical stability (159-161). Clinically, they both are mainly used for treatment of hematopoietic malignancies and in fact both are recently approved by the US FDA for the treatment of MDS. Many have disregarded that DAC has its primary target on inhibition of DNA synthesis, DNA damaging effect

(high dose) and hypomethylation activity (low dose), whereas 5-AzaC could act on both

DNA and RNA pathways. Many have also overlooked that the reason for 5-azaC’s DNA effect was due its conversion to DAC. More importantly, only 20% of the drug has been shown to be converted to DAC and the major portion of 5-azaC (~80%) acts on the RNA pathway. Therefore, the difference in mechanisms of action of 5-azaC and DAC, that might explain their clinical differences, has not been fully investigated. Additionally, despite this distinct parity between 5-azaC and DAC, little work has been done on the mechanism(s) of 5-azaC’s RNA effect.

It has previously been demonstrated that the consequences of 5-azaC incorporation into RNA can include alterations in tRNA and rRNAs, that result in inhibition of protein synthesis (155), and these RNA effects are avoided with DAC (13).

In this study, we have identified and characterized the subunit of RR (RRM2) as a novel

114 target for 5-azaC, direct comparisons between 5-azaC and DAC activities are presented and our data support the distinction between these two compounds. We demonstrate for the first time 5-azaC-induced decrease in RRM2 protein and mRNA levels in a dose-and time dependent manner and this effect was specific to 5-azaC and not to its reduced analog, DAC. In fact, traditionally, DAC is considered as a more potent demethylating agent, since it is already a deoxyriboside that is readily phosphorylated (35) and incorporated fully into DNA and not RNA. In contrast, only <20% of 5-azaC integrates into DNA, while the remaining >80% was believed to be incorporated into RNA, so that

5-azaC incorporation into RNA could cause changes in RNA structure and may account for the inhibitory effects of 5-azaC during protein synthesis (113). In our studies, we further discovered RRM2 as a novel target of 5-azaC and evaluated it against antisense

GTI-2040 and Triapine. Our data showed that 5-azaC is more potent than these known

RRM2 inhibitors.

Since formation of the deoxyribonucleotides (dNTPs) from the corresponding ribonucleotides (NTPs) is controlled by RR (162, 163), enzyme activity of RR directly relates to the production of dNTPs in cells because it ensures efficient genome replication

(40). A substantial amount of dNTPs are required by cells in S phase of the cell cycle and so closely linked with the growth control mechanisms of cells. Consequently, perturbed dNTP pools by RR inhibitors can cause a disturbance in the DNA synthesis and replication processes thereby resulting in chromosomal instability (26). Since we know that the integrity of RR function maintains a balance between the intracellular concentrations of each pool, we sought to establish and support our data that the observed

115 decrease in the RRM2 levels by 5-azaC treatment resulted in the perturbation of the dNTP. It has been demonstrated that many nucleoside analogs cause perturbation of dNTP pools (44) and our data show that following 5-azaC treatment of leukemia cells, the dTTP, dATP and dCTP pools were reduced in a dose dependent manner. Unlike DAC with a deoxyribose base, 5-azaCTP is not expected to link directly with dNTP pools, so that this observed perturbation is probably be due to its effect on RRM2. However, we also observed a decrease, specifically the CTP and UTP pools, which may be due to uridine-cytidine kinase. During 5-azaC metabolism, the rate limiting step is catalyzed by uridine-cytidine kinase, which is prone to potent feedback inhibition by uridine and cytidine triphosphate (UTP and CTP) (17). Therefore, since 5-azaC can reduce the intracellular UTP and CTP pools, it may enhance its phosphorylation to the active metabolite due to reduced feedback inhibition (17). The decrease in UTP and CTP may also be attributed to inhibition of the enzyme, CTP synthetase, which was found to be modulated by pyrimidine antimetabolites (164), such as 5-azaC. This enzyme catalyzes the formation of CTP from UTP and plays a role in modulation of cellular CTP and dCTP pools (164, 165). Possibly, 5-azaC inhibits CTP synthetase and concomitantly, disrupts both the intracellular CTP and dCTP. Therefore the imbalanced supply of dNTP/NTP pools caused by 5-azaC treatment confirms the link between the reduction in RRM2 expression and the decline in the dNTP pools.

This study also investigated the mechanisms involved in 5-azaC-induced down- regulation of the mRNA and protein expression levels of RRM2. Our data demonstrate that the reduction in RRM2 mRNA levels was due to changes (2-fold reduction in RRM2

116 mRNA half-life). Interestingly, our data further showed that mRNA of RRM2 was further destabilized by cycloheximide, suggesting that protein synthesis/translation is involved in

RRM2 mRNA stabilization. The relationship between mRNA half-life and translation is not yet completely understood. However, most mRNAs are stabilized in cells exposed to translation inhibitors (e.g. cycloheximide), but different mRNAs might be stabilized by a distinct pathway under certain circumstances (166, 167). It has been known that some

RNA-binding proteins play essential roles in protecting mRNA from degradation, such as polyA-binding protein (168), AU-rich region-binding protein (169), iron regulatory protein (170), etc. Two proteins are currently known to influence mRNA stability of

RRM2 by direct binding to its 3’-UTR sequence: p75 was suggested to protect mRNA from RNase attack, whereas R2BP was proposed to function as a chaperone for nucleases thus mediating cleavage of RRM2 mRNA upon binding (30, 171). Previous studies have also implied that p75 might function as a dominant stabilizer when both p75 and R2BP bind to 3’-UTR simultaneously (30, 171). Continued translation is required to maintain the level or activity of a regulatory protein, therefore one simple explanation of our results is that complete blockage of translation by cycloheximide might reduce the level of dominant regulatory factor p75, thus further destabilize RRM2 mRNA.

Previous studies indicated that the incorporation of 5-azaC into RNA can lead to the inhibition of RNA and protein synthesis (11, 172, 173). We reported, for the first time, by chemical and LC-MS/MS methods to prove the presence of 5-azaC in RNA. In our initial attempt, we did not observe 5-azaC directly. Since 5-azaC is unstable in neutral or alkaline solutions with a half-life of approximately 4 hours (156), it was likely

117 degraded during the hydrolysis process. However, by the use of NaBH4 or/and NaBD4 to reduce the labile 5-6 bond of the ring in 5-azaC before RNA hydrolysis, the reduced dihydro-5-azacytidine can be detected. We observed peaks corresponding to the dihydro-

5-azaC, HD-5-azaC and dideuterio-5-azaC, following NaBH4 or NaBD4 treatment in total

RNA. These data confirmed the presence of 5-azaC in RNA. Previous methods that investigate the incorporation of 5-azaC into RNA were done by radioactivity (174). Our new developed chemical reduction and LC-MS/MS approach, although only qualitative, provide an equivocal structural evidence for the incorporation of 5-azaC into RNA. This will allow for the rapid detection of 5-azaC incorporation into RNA for future studies.

We have shown that RRM2 expression knockout abolishes 5-azaC mediated down-regulation of DNMTs (1 and 3a). These data strongly supports the DNA hypomethylation effect of 5-azaC is in fact through first conversion to DAC, which subsequently, through further activation then DNA incorporation of DAC-TP, inactivates

DNMTs via covalent complex formation. More interestingly, following the induction with a second dose of 5-azaC in cells 24 hours after the first exposure, there was still no

DAC or DAC-TP detected, and the DNMT1 and 3a protein expression remained down regulated. This suggests that DNA-DNMT-drug covalent complexes following the initial formation with the formed DAC-TP are rather stable (half-life of DNMT1 ~6 hours).

Furthermore, it is possible that, DAC-TP as the active metabolite is retained in the cell after the drug is removed, which leads to continued use by DNA polymerases leading to sustained inhibition of DNA replication (1).

118 In addition to the in vitro results, we also showed that 5-azaC treatment reduces

RRM2 mRNA level and protein expression in-vivo in tumor tissues collected from athymic nude mice inoculated with MV4-11 cells. Our data confirm for the first time that treatment with 5-azaC can decrease both the RRM2 mRNA and protein levels, which generates a major advancement in the pharmacology of 5-azaC and will enable efficient translation of preclinical studies to clinical effectiveness. Furthermore, we have shown that 5-azaC treatment of mononuclear bone marrow cells collected from AML patients also showed a dramatic down-regulation of RRM2 mRNA levels (range of 0.01 – 0.74 relative to 1) and protein expression in five out of six patients. Thus, it is possible that clinically 5-azaC is a strong RRM2 inhibitor at both transcriptional and translational levels. Such inhibition should enhance the antileukemia activity of ara-C or other nucleoside analogs and this hypothesis will be evaluated in a future Phase I protocol study being planned.

Since a significant fraction of 5-azaC works on the RNA pathway and we have shown its RNA destabilizing effects, it would be of interest if 5-azaC RNA effect extends to microRNAs, which were not discovered until relatively recently. MicroRNAs play an important role in cell proliferation, apoptosis and differentiation during mammalian development (45). It has been reported that miRNAs can regulate gene expression by base pairing imperfectly to the 3’ UTR of target mRNAs and therefore inhibit protein synthesis or mRNA degradation (175). Although RNA modifications including methylation carried out by DNMT2 (176) has been identified in tRNA/rRNA (176, 177) and now microRNA (177), their relationship is rather complex and only limited knowledge is available. In one example showing a relationship between the function of

119 RNA modification and microRNA, a degradation of RNA after the injection of miR-221 and 222, that acts as potential target of Kit mRNA was reported (178, 179). In addition, the authors stipulated that exposure of microRNA to early embryonic genome is able to induce permanent and heritable epigenetic change in gene expression (179). Therefore, since we confirmed total RNA incorporation of 5-azaC and demonstrated its RNA destabilization effect, it is possible that 5-azaC could influence specific RNA such as microRNA thereby leading to up or down-regulation of certain microRNAs. To this end, using first microarray analysis, followed by quantitative RT-PCR, we found an increase

(20-90%) in miR29b and miR181a compared to the untreated control in both MV4-11 cell line and human primary bone marrow cell samples.

4.6 Conclusion

Our data provide an in depth understanding of the molecular pharmacology of 5- azaC and open a newly expanded area of its potential clinical use in combinations involving RRM2 inhibitors with other drugs. For example, previous studies have established the feasibility and the synergistic interaction of biochemical modulation of nucleoside analogs by combination with other RR inhibitors (180-182). Other reports

(27) demonstrated in a phase 1 trial that, GTI-2040 induces down regulation of RR and, when combined with high-dose aracytidine (ara-C) and the accumulation ara-CTP in

AML blasts results in enhanced ara-C cytotoxic. As such, RR as a novel target of 5-azaC will serve to develop the feasibility of more effective combination chemotherapy regimens for treatment of cancer. For example, we have designed a protocol combining azacytidine (5-azaC) with mitoxantrone, etoposide, and cytarabine (MEC) as salvage

120 chemotherapy in patients with relapsed or refractory acute myeloid leukemia (AML).

This protocol is aiming at to enhance the antileukemia effect of cytarabine and has been approved by the Center of Therapy Evaluation Program (CTEP) at the National Cancer

Institute. In summary, we report RRM2 as a robust gene target of 5-azaC therapy in vitro and in vivo.

121

NH2 NH2 NH2

HC N N N N N HC N O N O N O HO HO HO O O O H H H H H H H H H H H H OH OH OH OH OH

cytidine 5-aza-cytidine decitabine NH2

Figure 4.1 StructureN of 5-azacytidine. Hum (R2) mRNA

194 1364 2475 N O CODING

5’UTR 3’UTR Poly A HO O H OH 626 645

H H 3’ GAACCACCTCGCTAAATCGG 5’ GTI -2040 OH cytarabine

122

Figure 4.2 Function of ribonucleotide reductase (RR).

123

A

B

Figure 4.3 5-AzaC decreases RRM2 protein expression in MV4-11 cells. (A) Western blot analysis of RRM2 protein following 5-azaC treatment at the indicated concentrations for 24, 48 and 72 hours, respectively. (B) Representative densitometry plots of time- and dose dependent decrease of RRM2 protein expression following 5-azaC treatment in MV4-11 cells. RRM2 protein levels were quantified by densitometry and expressed as the ratios over the loading control GAPDH.

124

A

B

Figure 4.4 5-AzaC decreases RRM2 protein expression in K562 cells. (A) Western blot analysis of RRM2 protein following 5-azaC treatment at the indicated concentrations for 24, 48 and 72 hours, respectively. (B) Representative densitometry plots of time- and dose dependent decrease of RRM2 protein expression following 5-azaC treatment in K562 cells. RRM2 protein levels were quantified by densitometry and expressed as the ratios over the loading control GAPDH.

125 A

B

Figure 4.5 (A) Effects of DAC, triapine and GTI-2040 on RRM2 protein expression at the indicated concentrations and times in MV4-11 and K562 cells. (B) Comparative plot of RRM2 protein level in MV4-11 cell line following 5-azaC, triapine and GTI-2040 treatment, respectively. Densitometry was performed to quantify each lane and the ratio of RRM2 over the loading control GAPDH is presented under each blot.

126

Figure 4.6 5-AzaC down-regulates RRM2 mRNA. (A) 5-AzaC causes a dose-and time- dependent decline in total RNA level in MV4-11 and K562 cells. (B) 5-AzaC decreases RRM2 mRNA levels at 24 hours in MV4-11 and K562 cells in a dose-dependent manner. (C) Changes in RRM2 mRNA levels as a function of time in MV4-11 and K562 cell lines following treatment with 5µM 5-azaC. The RRM2 mRNA levels were normalized by abl. Data are presented as the percentage of the untreated control and the. *p < 0.05, **p < 0.01, as compared to control.

127

Figure 4.7 5-AzaC decreases intracellular dNTP and NTP pools. 5-azaC dose- dependently reduces levels of dTTP, dATP and dCTP in MV4-11 (A) and K562 cells (B) in a dose dependent manner. 5-AzaC decreases GTP, CTP, UTP and dGTP/ATP in MV4-11 (C) and K562 cells (D) in a dose-dependent manner. Data are presented as the percentage of untreated control. The results are mean ± SD of an experiment performed in triplicate. * p < 0.05, ** p < 0.01, versus control.

128

Figure 4.8. 5-AzaC reduces mRNA stability of RRM2. (A) 5-AzaC shortens the half-life (t1/2) of RRM2 mRNA, but not abl mRNA or GAPDH mRNA. MV4-11 cells were treated with 1µM actinomycin D in the presence or absence of 1µM 5-azaC. mRNA levels of RRM2 and two internal controls, abl and GAPDH, were measured by qRT-PCR at the indicated time points after actinomycin D treatment. The data are presented as the percentage of the mRNA level measured at time 0 (without adding actinomycin D). Since transcription is blocked by actinomycin D, mRNA levels of RRM2, abl and GAPDH decline in a time-dependent manner. Their mRNA half-lives were then calculated and indicated in the plots, respectively. 5-AzaC decreases t1/2 of RRM2 mRNA from 4.2 hours to 2.5 hours, whereas it does not significantly change the t1/2 of abl and GAPDH mRNA. (B) Blockade of protein synthesis facilitates 5-azaC-induced destabilization of RRM2 mRNA. MV4-11 cells were pretreated with 10µg/ml cycleohexmide for 2 hours to block protein synthesis, then treated with 1µM 5-azaC for additional 2 hours and subjected to 4 hours treatment of 1µM actinomycin D to inhibit transcription. mRNA levels of RRM2 and abl were also quantified by qRT-PCR. Data are presented as the percentage of control (actinomycin D-only) and shown as mean ± SD from triplicate experiments. *p < 0.05, as compared to control (actinomycin D-only); #p < 0.05, as compared to the treatment with both 5-azaC and actinomycin D.

129

A. B.

RT: 5.88 NL: 4.58E5 RT: 5.82 5-azaC and Uridine 5-azaC and Uridine NL: 4.39E5

RT: 5.94 NL: 3.61E3 RT: 4.56 2H-5-azaC 2H-5-azaC NL: 3.68E4

NL: 4.12E1 HD-5-azaC RT: 4.54 HD-5-azaC NL: 3.92E2

2D-5-azaC NL: 3.85 2D-5-azaC NL: 3.69

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 Time (min) Time (min) C. D.

RT: 5.80 RT: 5.77 NL: 4.20E5 5-azaC and Uridine NL: 3.89E5 5-azaC and Uridine

RT: 5.77 RT: 5.77 RT: 4.75 2H-5-azaC NL: 5.45E3 2H-5-azaC NL: 6.41E3

RT: 5.89 RT: 4.73 HD-5-azaC NL: 6.44E1 HD-5-azaC NL: 1.85E4

RT: 4.76 2D-5-azaC NL: 3.29 2D-5-azaC NL: 3.78E2

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 Time (min) Time (min)

Figure 4.9 5-AzaC incorporates into RNA. The extracted ion chromatograms (XICs) for (A) RNA hydrolysate from blank cell sample treated with NaBH4. (B) RNA hydrolysate from 5-azaC treated cell sample followed by treatment with NaBH4. (C) RNA hydrolysate from blank cell sample treated with NaBD4. (D) RNA hydrolysate from 5- azaC treated cell sample followed by treatment with NaBD4.

130

= Dosing event

Figure 4.10 Knock-down of RRM2 expression prevents (A) DAC and (B) DAC-TP conversion from 5-azaC in MV4-11 cells. MV4-11 cells was transfected with 10 µM siRRM2 or mock (PBS) for 24 hours, followed by treatment with 10 µM 5-azaC for the indicated time. 5-AzaC converts to DAC and is further phosphorylated to DAC-TP at early time point (2 hours) after initial 5-azaC treatment. However, cells pre-treated with siRRM2 and long incubation (>24 hours) with 5-azaC results in reduced DAC or undetectable DAC-TP levels. Even repeated exposure to 10 µM 5-azaC could not restore the formation of DAC or DAC-TP. Cells treated with DAC itself served as a positive control. (C) Stability of DAC over time. MV4-11 cells in 1 mL PBS were heat deactivated for 6 minutes at 100 °C and exposed to 10 µM 5-azaC or DAC at 37°C. DAC remained stable during the time course of the experiment and DAC was not formed from 5-azaC due to the inactivation of the enzymes by heat. The arrows indicate fresh addition of 5-azaC. 131

Figure 4.11 Knock-down of RRM2 expression inhibits 5-azaC-induced down-regulation of DNMT 1 and 3a. MV4-11 cells was transfected with siRRM2 or PBS for 24 hours and subjected to treatment with 10 µM 5-azaC for the indicated time. Western blot analysis shows that 5-azaC alone time-dependently decreases the protein levels of DNMT1 and 3a, whereas RRM2 depletion by siRRM2 abolishes its effects.

132

Figure 4.12 5-AzaC reduces RRM2 in MV4-11 tumor xenografts in mice. Athymic nude mice inoculated with MV4-11 cells were treated with 5-azaC for 32 days and the tumor samples were collected for western blot and RT-qPCR analysis. (A) RRM2 protein was dramatically reduced by 5-azaC in engrafted tumor tissues. Densitometry measurement was also performed and the ratio of RRM2 over the loading control GAPDH was presented under the blot. (B) 5-AzaC reduces RRM2 mRNA level in tumor tissues by about 50%. The results are mean ± SD from tissue of three mice per each group. The data are presented as percentage of control. *p < 0.05, as compared to the untreated control.

133

Figure 4.13 5-AzaC treatment reduces RRM2 mRNA and protein expression in bone marrow samples from AML patients. Untreated bone marrow cells from AML patients were treated at the indicated concentrations for 24 and 48 hours and the samples were collected for RT-qPCR and western blot analysis. A 60-90% reduction of RRM2 mRNA levels and their corresponding protein expression was found in samples from five patients. The data are presented as the percentage of the untreated controls. Densitometry measurement was performed to quantify each lane and the ratio of RRM2 over the loading control GAPDH is presented under each blot. The results are mean ± of SD of triplicate experiments. *p < 0.05, **p < 0.01, versus control.

134

160 * 140 miR29b

120

100 4 h 24 h 80

miR29b/U6 60

40

20

0 0 5 10 Concentration (µM)

miR181a

180

160 ** ** 140

120

100 4h 80 24h

60 miR181a/RNU44 40 20 0 0 5 10

concentration (uM)

Figure 4.14 5-AzaC treatment increases the endogenous levels of miR29b and miR181a following treatment with 5-azaC in MV4-11 cells. The miRNA levels were normalized by U6; RNU44 and presented as a percentage of untreated control. Vertical bars represent mean ± SD from triplicate experiments *p<0.05, ** p<0.01, versus untreated control.

135

miR29b

180 ** 160 *** 140 ** 120

100 0 24 h 80 48 h

miR29b/RNU44 60

40 **

20 0 Pt 135 Pt 624 Pt 396 Pt 534

250 miR181a *** 200

p=0.06 150 0 * ** 24 h 48 h

100 miR181a/RNU44

50

0 Pt 135 Pt 624 Pt 396 Pt 534

Figure 4.15 5-AzaC treatment increases the endogenous levels of miRNA29b and miRNA 181a in primary bone marrow mononuclear cells from patients with AML. The miRNA levels were normalized by RNU44 and presented as a percentage of control, *p<0.05, ** p<0.01, *** p<0.001, versus untreated control.

136 CHAPTER 5

DEVELOPMENT AND CHARACTERIZATION OF 5-AZACYTIDINE AND DECITABINE RESISTANCE IN HUMAN CANCER CELL LINES

5.1 Abstract

5-Azacytidine (5-azaC) and decitabine (DAC) are two nucleosides widely used for the treatment of hematologic malignancies, now approved by the US FDA for the treatment of myelodysplastic syndrome. However, acquired drug resistance remains a major problem. To overcome this problem, cellular mechanisms involved in the acquired drug resistance need to be elucidated. Herein, we developed 5-azaC and DAC resistance cell lines for identification of biomarkers potentially involved in their acquired resistances. Leukemia cell lines K562 and MV4-11 and solid tumor cell lines IGROV1 and HCT-15 were exposed to increasing 5-azaC or DAC concentrations qd10-15 over a period of 6 months, starting with an initial concentration of 1 µM. The cytotoxicity of the resistant cell lines, MV4-11/5-azaC, K562/5-azaC, IGROV1/5-azaC, HCT-15/5-azaC,

K562/DAC and MV4-11/DAC was examined using MTS assay and the IC50 values were determined. Additionally, parental K562, MV4-11 and IGROV1 and their resistant cell lines were treated with 2 µM of DAC at 0, 1, 4 and 24 hours and DAC-TP levels were determined by an LC-MS/MS method developed in our laboratory. Quantitative RT-PCR was used to measure the DNA methyltransferase (DNMT1, 3a and 3b) mRNA levels, 137 while western blot analysis was used to evaluate the protein level. Global DNA methylation (GDM) level was determined using a published LC-MS/MS method. Cross- resistance of DAC with 5-azaC resistant cell lines was also determined. The IC50 values of K562/5-azaC, K562/DAC, MV4-11/5-azaC, MV4-11/DAC, IGROV1/5-azaC and

HCT-15/5-azaC were found to be 50, 220, 54, 508, 3, and 5-fold, respectively, higher than their respective parental cell lines. DNMT1, 3a and 3b mRNA and protein expression levels were found to be rather variable in these resistant cell lines. GDM was found to increase by 15% for K562/5-azaC but no significant change in the other cell lines. The baseline dNTP/NTP pools were found to vary significantly in these cell lines with no consistent trend of perturbation, when cells were rendered resistant. Following treatment with 2 µM DAC for 0, 1, 4 and 24 hours, DAC-TP levels for 5-azaC (MV4-11,

K562 and IGROV1) and DAC (MV4-11 and K562) resistant cell lines were lower than the corresponding parental cell lines. DAC was found to be cross-resistant to 5-azaC resistant cell lines. This study provides important in vitro models for understanding the major mechanisms of resistance to DAC and 5-azaC.

5.2 Introduction

5-Azacytidine (5-azaC) and Decitabine (DAC) are pyrimidine analogs that were synthesized over 45 years ago and have been widely used for the treatment of hematological malignancies, including acute myeloid leukemia (AML). 5-AzaC and

DAC are used as hypomethylating drugs and have recently been approved by the U.S.

Food and Drug Administration for the treatment of myelodysplastic syndrome (MDS)

(95, 183). However, a major problem with both drugs is the development of resistance

138 during treatment. There appear to be a number of potential intracellular mechanisms underlying drug resistance at the molecular levels. In order to provide better therapeutic strategies, in the current studies we investigate the relationship between several biochemical mechanisms with their acquired resistance of these drugs.

Methylation are epigenetic changes to chromatin structure that are crucial for cell development and differentiation (68). DNA methylation is a normal occurrence via DNA methyltransferases (DNMT 3a, 3b and 1) (68, 92). Aberrant epigenetic changes are involved in almost all the different stages of cancer development including progression.

For example, up-regulation of DNA methyltransferases leads to hypermethylation and silencing of some tumor suppressor genes and the reactivation of certain oncogenes (46,

69). Since epigenetic processes are almost completely reversible, hypermethylation can be overcome via pharmacologic inhibition of DNA methyltransferases (DNMTs). The reversal of epigenetic changes is usually achievable with low dose 5-azaC and DAC (4).

In humans, cellular influx of 5-azaC and DAC involves the membrane human equilibrative nucleoside transporters (hENT1 and hENT2) (41). Once inside the cell, 5- azaC is metabolized to its monophosphate by uridine-cytidine kinase enzyme, then to its diphosphate, and finally to its active triphosphate, 5-aza-CTP, by diphosphate kinase.

Resistance to 5-azaC and DAC may occur as a result of low accumulations of the active anabolites 5-azaCTP and DAC-TP due to deficiency in drug uptake, activation and/or degradation (12). Ribonucleotide reductase (RR) is responsible for the reduction of all nucleoside 5’-diphosphates (ADP, GDP, UDP and CDP or NTPs) to their corresponding

139 2’deoxyribonucleotides (dADP, dGDP, dUDP and dCDP or dNTPs) (30, 32) and this is an important step during DNA synthesis and repair. Over-expression of RR causes imbalance of endogenous dNTP/NTP pools and serves as a potential mechanism of resistance for nucleoside analogs, which is common in malignant cells (102-106).

In this study, we have developed and characterized several stable, 5-azaC and

DAC resistant leukemia and solid tumor cell lines, by culturing these cell lines in the presence of graded concentrations of the drugs. Additionally, we sought predictive potential biomarkers involved in induced 5-azaC and DAC resistances. Consequently, we identified several biomarkers that may be involved in the acquired and cross-resistance to these drugs. These findings will provide understanding of the resistance to these drugs and facilitate the finding of strategies that will overcome this problem under clinical situations. In addition, our studies will help identification of patients who are likely or less likely to be benefitted from clinical trials involving the use of these epigenetic agents.

5.3 Materials and methods

5.3.1 Cell culture and treatment methods

Human leukemia cell lines K562 and MV4-11, ovarian cancer cell line IGROV1, and colon cancer cell line HCT-15 were obtained from American Type Culture

Collection (Manassas, VA). The cells were maintained in RPMI 1640 medium (Gibco,

Rockville, MA), supplemented with 10% fetal bovine serum cultured with 5% CO2

140 atmosphere at 370C. To develop drug resistance, the cells were exposed to increasing concentrations of 5-azaC or DAC (Both obtained from the National Cancer Institute

Bethesda, MD) separately over a period of 6 months, starting with an initial concentration of 1 µM to a maximum of 30 µM. Cell lines resistant to 5-azaC were designated as

K562/5-azaC, MV4-11/5-azaC, IGROV1/5-azaC and HCT-15/5-azaC, while DAC resistant cell lines were designated as K562/DAC and MV4-11/DAC. Cells were maintained resistant in the culture medium containing 10 µM of the drug; however, cells were grown in a drug free media three days prior to experiment to avoid possible drug accumulation.

5.3.2 Growth inhibition assay and IC50 determination

For growth inhibition assay, parental and resistant cell lines in logarithmic growth phase were plated at a density of 5000 cells/well in 200 µL media in a 96-well plate.

Graded concentrations of 5-azaC or DAC were added to the medium. Afterward, 20 µL of 3-(4,5-dimethylthisazol-2-yl)-5-(3-carboxymethoxyphenyl-2- (4-sulfophenyl)-2H- tetrazolium, inner salt (MTS) and phenazine methosulfate (PMS) (Promega, Madison,

WI) were added to each well in a 20:1 ratio. After 2 hours incubation, the absorbance was read at 490 nm on a microplate reader Gemini XS (Molecular Devices, CA) to determine levels of formazan product as a measure of cell viability against a non-drug treated control. Cell viability at various drug exposure concentrations was measured and the drug concentration that inhibits 50% proliferation (IC50) was determined from these data using

WinNonLin computer Software (Pharsight, Mountain View, California). The resistance index was calculated as the ratio of the IC50 value of the resistant cell line over that of the

141 corresponding parental cell line. Statistical significance was determined using Student’s t test.

5.3.3 RNA isolation and RT-qPCR

To measure intracellular baseline levels of DNMTs mRNA, 5×106 cells from parental and their drug resistant cell lines were harvested and total RNA was isolated using Trizol reagent (Invitrogen,). Briefly, cell lysate was treated with chloroform and the total RNA was precipitated with isopropyl alcohol, followed by a washing step with 75% ethanol. RNA was then dissolved in RNase free water, and its concentration and purity was measured by a Nanodrop spectrophotometer (Nanodrop Technologies, Wilmington,

Delaware). cDNA was synthesized from 2 µg total RNA using Moloney murine leukemia virus reverse transcriptase (Invitrogen). The cDNA templates and primers were then mixed with reagents from a SYBR Green PCR Master Mix (Applied Biosystem, Foster

City, CA). Reactions were carried out in triplicates in ABI prism 7700 sequence detector

(Applied Biosystems) and data were analyzed by comparative CT method. Dissociation curves were also obtained to examine the purity of the amplified products. The amount of

DNMTs mRNA in each sample was normalized with respect to an internal control,

GAPDH. The relative changes in the resistant groups were expressed as a percentage of their corresponding parental cell lines (arbitrarily set at 1). The results were expressed as the mean ± of SD from triplicate determinations.

142 5.3.4 Western blot analysis

To determine the baseline levels of DNMTs’ protein expression, 5×106 cells of the parental and their resistant lines were harvested. K562 and MV4-11 cells were also treated with 2 µM of DAC for 24 hours, cell harvested, and used as a positive control.

These harvested cell were then washed with 1 mL ice-cold PBS, and centrifuged at 1000 g for 5 minutes at 4 ºC. The pellet was obtained and resuspended in 100 μL lysis buffer

(50 mM PH 7.6 Tris-HCl, 250 mM NaCl, 5 mM EDTA, 2 mM Na3VO4, 50 mM NaF and

1% protease inhibitor cocktail) (P8340, Sigma) for 30 minutes on ice. The lysate was sonicated for 10 seconds. Total protein concentration was determined using the BCA protein assay (Pierce, Rockford, IL). Equal amounts of protein for each sample were incubated with 6x SDS loading buffer (100 mM, pH 6.8 Tris, 200 mM DTT, 4% SDS,

20% glycerol, and 0.015% bromphenol blue) and boiled for 5 minutes. The proteins were then separated on 4-15% SDS-polyacrylamide gels (Biorad, Hercules, CA) and transferred to nitrocellulose membrane (Amersham, Piscataway, NJ). The membrane was then probed with anti-DNMT1 antibody (1:500, New England BioLabs, MA),

DNMT3a/b (1:500, Santa Cruz, California) and GAPDH (1:1000, Santa Cruz, CA) served as a loading control. Densitometry was also performed to quantify each lane by

Image J (Scion Image, Scion, Frederick, MD).

5.3.5 Determination of intracellular dNTP/NTP pools and DAC triphosphate

(DAC-TP) levels

10×106 parental cells and their resistant counterparts were monitored for viability using trypan blue exclusion test. The cells were washed with phosphate buffered saline

143 (PBS) and dNTPs/NTPs were extracted and quantified as previously described (117).

Briefly, cells were first deproteinized, then lysed, followed by separation and drying under a stream of nitrogen. The residues were reconstituted in mobile phase, centrifuged at 4 °C for 2 minutes and a 50 µL aliquot of the supernatant was then injected into an ion- trap LC-MS/MS system (LCQ, Thermo Scientific, San Jose, CA), for dNTP and NTP measurements.

For measurement of DAC-TP and dNTP/NTP levels, parental and resistant cell lines (20×106 cells each) were treated with 2 µM of DAC for a 0, 1, 4 and 24 hours. The cells were harvested, washed with phosphate buffered saline (PBS), and the triphosphates were extracted and quantified using the same procedure as described above.

5.3.6 Global DNA methylation (GDM) analysis

The genomic DNA was isolated from 2×106 untreated parental/resistant MV4-11 cells treated with 1 µM of DAC for 72 hours using DNeasy Tissue kit (Qiagen,

Minneapolis, MN) and the DNA concentrations were measured on a Nanodrop spectrophotometer (Nanodrop Technologies, Wilmington, DE). DNA hydrolysis was performed according to the manufacturer’s instruction and the resulting hydrolysate was analyzed for DNA methylation by LC-MS/MS under conditions as previously described

(92). Briefly, 1 µg of genomic DNA was first denatured by heating at 100 °C for 3 min and then chilled on ice. Following the addition of a 1/10 volume of 0.1M ammonium acetate (pH 5.3) and two units of nuclease P1, the mixture was incubated at 45 °C for 2 hours. One-tenth volume of 1M ammonium bicarbonate and 0.002 unit of snake venom phosphodiesterase I were added, and the mixture was incubated at 37 °C for 1 hour. Next,

144 0.5 unit of alkaline phosphatase was added, and the mixture was incubated at 37 °C for 1 hour.

5.3.7 Characterization of resistance

Following the exposure of the aforementioned cell lines to increasing concentrations of 5-azaC or DAC separately over a period of 6 months, the development of resistance in the cells was assessed based on an increase in IC50 values of the resistant cells relative to its parental cell line. The resistance index values were calculated, which indicated and confirmed the extent of resistance of the cells to 5-azaC or DAC.

Perturbations of different biomarkers (DNMTs mRNA and protein expression, dNTPs/NTPs, GDM and DAC-TP levels) in the parent and resistant cell lines were further used to characterize the induction of resistance in cells.

5.3.8 Cross-resistance of DAC with 5-azaC resistant cell lines

It is hypothesized that 5-azaC resistance cell lines may be cross-resistant with

DAC. In order to evaluate their cross-resistance, K562/5-azaC, MV4-11/5-azaC and

IGROV1/5-azaC cell lines and their parental counterparts were exposed to graded concentrations of DAC. Cytotoxicity assay was performed and the IC50 and the resistance index were calculated. On a separate experiment to determine how the treatment of K562,

MV4-11 and IGROV1 and their 5-azaC resistant cell lines with DAC will affect their

DAC-TP levels, the cells were treated with 2 µM DAC for 0, 1, 4 and 24 hours.

Decitabine triphosphate (DAP-TP) was then extracted and measured as previously described.

145 5.3.9 Statistical analysis

Statistical evaluation of differences observed between relevant experiments was determined using a Student’s t-test. The minimal level of significance was a p-value below 0.05 and accomplished using Minitab statistical software (Minitab, State College,

PA). Excel (Microsoft Office Word 2007, WA) was used in the data plot.

5.4 Results

5.4.1 Establishment of 5-azaC and DAC resistant cell lines

Resistance to 5-azaC or DAC was induced by stepwise exposure of the cells to increasing concentrations of the drugs. Two leukemia cell lines responsive to 5-azaC and

DAC, K562 and MV4-11, and two solid tumor cell lines, colon and ovarian cell lines,

HCT-15 and IGROV1, respectively, were selected. After development of resistance, these cell lines were found to maintain stable resistance to the respective drug for more than 30 cell passages.

5.4.2 Reduced drug sensitivity in resistant cell lines

To examine the sensitivity of resistant cell lines to 5-azaC or DAC relative to their parental counterparts, inhibition growth assay was performed (Figure 5.1). We observed a significant shift of the dose-response curves of the resistant cell lines to the right

(decrease in activity) as compared to their corresponding parental cell lines. IC50 values of the cell lines and the resistance index of the resistant cell lines are listed in Table 5.1.

As shown, K562/5-azaC and MV4-11/5-azaC were 49 and 54-fold, respectively, less

146 sensitive to 5-azaC than their parental cell lines. The solid tumor cell lines IGROV1/5- azaC and HCT-15/5-azaC showed a 3 and 5-fold, respectively, less sensitive to 5-azaC.

The IC50 values of K562/DAC and MV4-11/DAC were 180.6 and 86.4 µM, respectively, which were a 220 and 500-fold increase, respectively, as compared to their parental cell lines.

5.4.3 mRNA levels of DNMTs in 5-azaC and DAC resistant cell lines

In order to understand how induction of resistance to 5-azaC or DAC affects hypermethylation, mRNA levels of DNMTs were quantified. In MV4-11/5-azaC, we found that DNMT1, DNMT3a and 3b mRNA levels were increased about 2-5 fold. In

K562/5-azaC and IGROV1/5-azaC, no significant change in DNMT1 and DNMT3a mRNA levels was found; however, DNMT3b was decreased by 50-60% (Figure 5.2A).

In HCT-15/5-azaC, DNMT1 and 3b mRNA levels were increased by 1-2 fold, while

DNMT3a remained unchanged. All changes were estimated relative to their corresponding parental cell lines. Overall, DNMT1 levels in all resistant cell lines appeared to be either increased or no change, whereas DNMT3b levels appeared to have no consistent change between leukemia and solid tumor resistant cell lines

In MV4-11/DAC cell line, DNMT1 mRNA level increased about 3-fold and

DNMT3b increased dramatically (14 fold), but DNMT3a mRNA level decreased slightly by 40%. In K562/DAC cell line, all mRNA levels of DNMTs showed slight increase or no change (Figure 5.2B). Thus, in the 5-azaC and DAC resistant cell lines DNMT1 and

DMNT3a,b mRNA levels showed consistent increase consistent with their functions to

147 maintain a global and CpG island methylation in human cancer cells, although the role of

DNMT3b may be less important (4, 184).

5.4.4 DNMT protein expression levels in resistant cell lines relative to parental cell lines

Next, to examine how DNMTs protein expression is modulated by the induction of resistance, DNMTs protein expression levels in the cell lines were determined using western blot. Our data showed a 10-60% increase in the DNMT1 levels for K562/5-azaC,

MV4-11/5-azaC, IGROV1/5-azaC and HCT-15/5-azaC relative to their parental cell lines

(Figure 5.3A). In K562/5-azaC, IGROV1/5-azaC and HCT1-/5-azaC, DNMT3a expression decreased by 50-70%, except for MV4-11/5-azaC, which showed an increase of 30% (Figure 5.3B) relative to their parental cell lines. We further found that the increase in the DNMT1 mRNA levels in 5-azaC resistant cell lines correlated with their increased protein expression levels.

K562/DAC and MV4-11/DAC demonstrate an increase (20-90%) in the DNMT1 protein level (Figure 5.4A) but their DNMT3a level decreased by 50% (Figure 5.4B). In

K562/DAC, the DNMT3b levels decreased by 30% but in MV4-11/DAC it increased >3- fold (Figure 5.4C). Our data suggest that DNMT1 protein expression level in 5-azaC or

DAC showed a more consistent increase, which was also reflected by an in increase in mRNA, when these cell lines were rendered resistant. Treatment of K562 and MV4-11 with 2 µM DAC decreased DNMT1, 3a and 3b levels by >90% as expected as positive controls. Thus, the quantitative analysis of DNMTs mRNA and protein levels may suggest that 5-azaC and DAC induction of resistance in cells via DNMTs differ in the

148 expression of the three isoforms. Therefore, our data suggest that the regulation and distribution of DNA methylation by DNMT1, 3a and 3b in cancer cell occurs independently of each other. Overall, cell lines resistant to DAC appear to show a more consistent increase in methylation machinery consistent with its DNA methylation machinery, whereas 5-azaC showed either less significant and consistent change in methylation machinery, suggesting that DNA methylation is not a major effect of 5-azaC and may be through its partial conversion to DAC.

5.4.5 Global DNA methylation (GDM) pattern in resistant cell lines

Since it has been shown that DAC and 5-azaC induce hypomethylation at low dose, we sought to measure the global DNA methylation levels in the parental and resistant cell lines. Our data show that GDM level for K562/5-azaC increased by about

15%, while MV4-11/5-azaC showed no change relative to the parental cell lines.

IGROV1/5-azaC and HCT-15/5-azaC cell lines showed a decrease in their GDM levels

(Figure 5.5A). K562/DAC and MV4-11/DAC also revealed a decrease in the GDM levels

(Figure 5.5B). Treatment of MV4-11 cells with 2 µM DAC decreased GDM levels by 30-

70%. These data may suggest that GDM alone may not be a good index for induced resistance of these leukemia and solid tumor cell lines. Alternatively, the resistance may not cause a significant change in DNA methylation status.

149 5.4.6 Ribonucleotide (NTP) and Deoxyribonucleotide (dNTP) triphosphate level

measurement

Since perturbation of dNTP/NTP pools by nucleoside drugs has been reported

(185), we compared the changes in dNTP and NTP pools in the resistant and parent cell lines. Previously, we have shown that the basal levels of the pools varied in different cell lines (117). For the resistant cell lines, we found that dATP increased by ~40% in

K562/5-azaC, while dTTP and dCTP levels remained unchanged. The GTP, UTP and dGTP/ATP levels all decreased by 2-4 fold with no change observed in the CTP level

(Figure 5.6A). dATP level in MV4-11/5-azaC cells increased by ~30%, while the dTTP level decreased by 40% and there was no change in the dCTP level. The CTP levels increased by 20%, while the dGTP/ATP level decreased by 20%, but the GTP and UTP levels remained unchanged (Figure 5.6B). In the solid tumor cell lines, dTTP, dCTP and dATP levels increased by 10-63% and the GTP, CTP, UTP and dGTP/ATP levels increased by 2-5 fold in IGROV1/5-azaC cell line (Figure 5.7A). In HCT-15/5-azaC cells, dCTP level increased by 40% with no change in the dTTP and dATP levels.

However, the GTP, CTP and UTP levels remained unchanged, while the dGTP/ATP increased by 50%, (Figure 5.7B). Therefore, our data imply that the perturbation of dNTP/NTP pools in 5-azaC resistant cell differ vastly from their parental cell lines and further confirms the induction of resistance in these cells.

For the DAC resistant cell lines, MV4-11/DAC showed an increase (4-30 fold) in the dTTP, dATP and dCTP levels. The GTP and dGTP/ATP also increased by 2-2.5 fold, while CTP and UTP levels remained unchanged (Figure 5.8A). In K562/DAC, the dTTP levels decreased by 60%, dATP increased by 40%, while the dCTP levels remained

150 unchanged. The GTP, CTP and UTP remained unchanged, but the dGTP/ATP decreased by 30% (Figure 5.8B). All data are relative to their parental cell lines. Our data showed that status of the dNTP/NTP pools in MV4-11/DAC and K562/DAC by DAC may significantly contribute to resistance in these cell lines. Overall, in 5-azaC resistant cell lines, dNTPs levels seem to be the same or decreased except dATP, which increased.

This differed from DAC resistant cell lines, which showed much expanded dNTP pool in

MV4-11 cell line, although in K562 cell line, dTTP level decreased significantly. Overall

NTP pool increased in MV4-11 cell line, although there were not much changes in the

NTP pool in K562 cell line except a significant decrease in dGTP/ATP.

5.4.7 Reduced DAC-TP accumulation in DAC resistant relative to the parental cell

lines

Using our newly developed novel specific HPLC-MS/MS method for measuring

DAC-TP, we quantified the intracellular DAC-TP levels in parental and DAC resistant cell lines. In K562 cells, following treatment with 2 µM of DAC for 1, 4 and 24 hours, the DAC-TP levels were 2.85 ± 0.74, 3.28 ± 0.31 and 0.87 ± 0.23 pmol/106 cells at 1, 4 and 24 hours, respectively, but were decreased to 0.48 ± 0.08, 0.45 ± 0.21 and 0.48 ±

0.08 pmol/106 cells in K562/DAC, respectively (Figure 5.9A). The dATP and dCTP levels increased (30-60%), while the dTTP level decreased by 50-70%. (Figure 5.9B).

In MV4-11/DAC cell line, following exposure to 2 µM of DAC, DAC-TP levels decreased significantly (50%), especially at the 4 hour time point (Figure 5.10A) relative to the parental counterpart. These data demonstrates that induction of resistance in these cells may be attributed in part to the decreased accumulation of DAC-TP. Additionally,

151 all of the dNTP levels were found to dramatically increase (2-30 folds) relative to the parental cell line (Figure 5.10B). Therefore, perturbation of the basal levels of dNTP/NTP pools in cells following development of resistance and the dramatic increase in the dNTP pools after treatment with DAC further competes with DAC-TP formation, contributing to the overall cell resistance.

5.4.8 Relationship of the DAC and DAC-TP levels in MV4-11 cells

In order to demonstrate the transformation of DAC to its active form DAC-TP in cells, both compounds were measured. Figure 5.11 shows that the DAC levels decreased with time in MV4-11 cells at levels of 5.75, 4.13 and 0.33 pmol/106 at 1, 4, and 24 hours, respectively, while DAC-TP levels appeared rapidly and sustained over time with values of 6.64, 6.31 and 6.61 pmol/106 at the corresponding time points. Thus, our data showed that DAC was converted to DAC-TP over time, which peaked at 24 hours and sustained thereafter.

5.4.9 Cross-resistance of DAC with 5-azaC resistant cell lines

Since both DAC and 5-azaC contain the same azapyrimidine, we next evaluated whether their resistant cell lines were cross-resistant to each other. Following treatment of MV4-11, K562 and IGROV1 cell lines and their 5-azaC resistant complements with graded concentrations of DAC, the cell viability was measured. We found 5-azaC resistant cell lines (Figure 5.12) were cross-resistant to DAC. IC50 values of K562/5-azaC and MV4-11/5-azaC showed a 400 and 2400-fold increase respectively, relative to those

152 of their parental cell lines. IGROV1/5-azaC only showed a minimal cross-resistance (1.5- fold in its IC50 values) (Table 5.2) to DAC.

5.4.10 Cross-resistance of DAC to 5-azaC resistant cell lines reduces DAC-TP accumulation

In addition, we sought to evaluate if 5-azaC resistant lines that are cross-resistant to DAC also affect the DAC-TP levels. Following the exposure of 5-azaC resistant lines to 2 µM DAC for 1, 4 and 24 hours, we found that the DAC-TP levels for K562 cell line did not show any statistically significant difference between the parental and resistant cell lines, except at 4 hours. However, in MV4-11/5-azaC cell line, the DAC-TP levels were

6.64, 6.31, 6.61 pmol/106 at 1, 4 and 24 hours as compared with 0.67, 1.55, 3.65 pmol/106 (a 9.9, 4 and 1.8-fold decrease) at 1, 4 and 24 hours, respectively, in the parental cell line. The DAC-TP levels in the IGROV1/5-azaC showed a significant decrease at 24 hours (by 2.8-fold) as compared to the parental cell line, with no significant change at 1 and 4 hours (Figure 5.13). Therefore, our data demonstrated 5- azaC resistant cell lines that are cross resistant to DAC also have decreased DAC-TP levels in most part after exposure to DAC, when compared to their parental counterparts.

5.5 Discussion

5-AzaC and DAC have been used as anticancer agents for the treatment of a number of hematopoietic neoplasms. Despite their recent approval for the treatment of myelodysplastic syndrome (MDS), resistance to these drugs is still a major problem. In

153 order to better understand the cellular mechanisms involved in the development of resistance to these nucleoside analogs, we developed a cellular model of 5-azaC and

DAC resistance in a panel of six leukemia and solid tumor cell lines. We now sought to identify the relationship between resistance and several biomarkers. The resistance index as estimated from the IC50s values of the both the 5-azaC and DAC resistant cell lines to those of their respective parental counterparts showed a wide range between 5-500 fold.

CpG island methylation is mainly carried out by DNMTs and their over- expression has led to tumor development and currently is treated with nucleoside hypomethylating agent such as low dose DAC or 5-azaC. Since, DAC can trap DNMT enzymes via a covalent protein-DNA adduct (95) causing inhibition and hypomethylation, defect in the complex formation could result in phenotypic resistance.

It is also possible that repeated drug treatment over time could link to development of drug resistance. Thus, we consider that DNMTs genes could be molecular markers that can be profiled for development of drug induced resistance. Based on our data, we demonstrated that altered DNMTs gene expression levels contributed partially to DAC/5- azaC-induced resistance in leukemia and solid tumor cell lines. In our 5-azaC/DAC resistant cell lines, DNMT1 expression but not DNMT3a/b was consistently increased with a more profound change in MV4-11/5-azaC cell. This may be due to the maintenance function of DNMT1 (186, 187), which serves as the major and ubiquitously expressed DNA methyltransferases (188). Another possible explanation may be that

DNMT1, but not DNMT3a/b, is directly coupled to the DNA replication machinery, which increases the precision in maintenance methylation. This suggests a fundamental

154 difference in the regulation and enzymatic mechanism between DNMT1, 3a and 3b

(188). Furthermore, increased DNMTs gene expression may also be associated with increased resistance as a result of exposure to cytotoxic dose of nucleoside drugs. This may play a role in how tumor cells evade the damaging and cytotoxic effects of anticancer agents (189).

DNMT3a and 3b are de novo methyltransferases (190) that are usually required for DNA methylation and worked cooperatively with DNMT1; however, their DNA methylating functions becomes amplified during cancer development (191). Our data showed varied expressions of DNMT3a in 5-azaC and DNMT3a/b in DAC resistant cell lines relative to the consistent increase in the DNMT1 expression. This selectivity of

DAC for DNMT1 may arise from differential nuclear localization of the DNMTs enzymes (184). DNMT1 localizes at the replication fork during S-phase, whereas

DNMT3a/b remains diffuse in the nucleus (184, 192). Therefore, when DNMT1 moves with the replication complex, it encounters DAC or 5-azaC molecules incorporated into the genome more frequently than the diffusely localized DNMT3a and 3b enzymes (184).

Hence, it is possible that following exposure of the cells to 5-azaC/DAC, more DNMT1, but not DNMT3a/b enzymes, is first exposed to DAC. Taken together, our data provide some evidence that drug-induced DNA hypermethylation may be capable of inducing drug-resistant phenotype by inactivating genes or gene products which are required for drug cytotoxicity (193).

It has been shown that cancer cells express high DNA hypermethylation patterns, and GDM may serve as a surrogate endpoint to determine the hypomethylating effect of

155 5-azaC and DAC (92, 94). Therefore, GDM levels were measured in our model resistant cell lines and their parental counterparts. Our data did not show any correlation between increase in DNMTs expression and GDM levels in the resistant cell line, even though the link between chemo-resistance and global hypermethylations has been established (193),

This might suggest that the exact changes in DNMTs expression and global DNA methylation pattern after drug exposure, leading to resistances still need to be further investigated.

Resistance to nucleoside antimetabolite antitumor agents, such as 5-azaC and

DAC, has been linked to increased ribonucleotide reductase (RR) expression and closely correlated with expansion of dNTP/NTP pools (194). Additionally, if the dNTP pools are not sufficiently down-regulated during nucleoside antimetabolite therapy, dNTP pools could compete with these drugs for DNA incorporation (185). Our data showed that the distribution of basal levels of dNTP and NTP pools is variable in different cell lines, similar to the reported (117, 195, 196). In three of the investigated resistant cell lines a significant expansion of the dATP levels was observed but with a modest decrease or unchanged levels of the dCTP in several resistant cell lines. The reason for a decrease in dCTP level and an increase level of other dNTPs is not clear. However, an explanation for a similar observation in HL60/Fara resistant cell line was provided (197). These authors reported that continuous exposure to fludarabine may select RR mutants with reduced drug sensitivity to allosteric inhibitors that may result in dominant mutation in the RR gene, which disables the allosteric feedback control that was an important process in the parental cell line. Therefore, the alteration of dNTPs by 5-azaC/DAC may be due

156 to a selection of RR enzymes. Taken together, our data suggest that sustained exposure of

5-azaC or DAC in cells results in the perturbation or altered dNTP pools, which may partly contribute to the induction of resistance.

When DAC and 5-azaC resistant cell lines were exposed to DAC in a time dependent manner, a significant increase in the dNTP pools was observed, except dTTP levels in K562/DAC. This may be due to the decreased formation of DAC-TP in the 5- azaC/DAC resistant cell lines, hence stimulating an increase in the dNTP levels compared to the basal level. The decreased DAC-TP levels might have also resulted from deficiency in uptake of DAC, following treatment with DAC (12). In the solid tumor resistant cell lines (IGROV1 and HCT-15), dNTP pools increased and subsequently, NTP pools also increased, the reason for which is unclear (Figure 5.7). On the other hand, the

NTPs decrease in the leukemia cell lines (K562 and MV4-11) may be due to salvage or feedback regulations (triphosphate metabolites of 5-azaC and DAC conceived as NTPs).

At the molecular level, there appear to be a number of mechanisms through which resistances to nucleoside drugs could be acquired. It is commonly known that 5-azaC antileukemia effects require sufficient intracellular concentration of 5-azaCTP or DAC-

TP in tumor cells. Our data showed that there was a significant time-dependent decrease in DAC-TP levels (decrease in anabolism) in the DAC/5-azaC resistant cell lines relative to their parental counterparts, following treatment with DAC. Therefore, it is possible that decreased DNA incorporation, therefore increased DNA synthesis, may account for in part the acquired resistances for 5azaC and DAC (58). We have directly demonstrated this by using measurement of DAC-TP with our newly developed LC-MS/MS method.

157 The ability to measure the extent of accumulation of DAC-TP in cells serves as a useful surrogate endpoint during laboratory and clinical investigation of DAC/5-azaC. In addition, reports have also verified a significant correlation between the expression levels of nucleoside transporters and sensitivity to 5-azaC or DAC in mediating uptake of the drugs (109). Therefore, inadequate accumulation of the triphosphate forms of the drugs could be due to defects in the nucleoside transporters expression patterns. For future studies, nucleoside transporters expression patterns in relation to DAC-TP accumulation could potentially be used as predictive biomarkers for therapy response. In our study,

DAC was found to be cross-resistant to 5-azaC resistant cell lines, when exposed to DAC in a time dependent manner as shown by the change in the IC50 values (higher values) and the reduced accumulation of DAC-TP, when compared with the corresponding parental cell lines. These data may provide utility to probe resistance and/or lack or response to DAC during clinical treatment with DAC.

5.6 Conclusion

In conclusion, the aim of this project is to identify potential mechanisms that contribute and lead to induction of resistances of nucleoside analogs, 5-azaC and DAC, in human leukemia and solid tumor cell lines. Our data will contribute to the understanding of their therapeutic efficacy by introducing a new era of biomarker driven method of searching for compounds to yield synergistic outcome with these drugs. Although, in appreciating the capabilities of these drugs, further comprehensive mechanistic and translational research will be vital.

158

IC50 (µM) Resistance index Cell lines Parental 5-AzaC DAC resistant resistant K562 1.54 ± 0.25 0.82 ± 0.45 76.17 ± 10.81 180.61 ± 60.29 49.5* 220* MV4-11 2.38 ± 1.09 0.17 ± 0.04 128.65 ± 47.84 86.37 ± 47.24 54.1* 508* IGROV1 76.07 ± 18.52 - 234.99 ± 152.7 - 3.1 - HCT-15 8.26 ± 1.44 - 40.51 ± 4.37 - 5.1* - Resistant index: Defined as IC50 resistant cell line/IC50 parental cell line. *p<0.05

Table 5.1 IC50 values of the parental and resistant cell lines and their resistant index. We show an increase (5 – 500 fold) in the resistant cells relative to the parental cells. All experiments were done in triplicate.

159

IC50 (µM) Resistance index Cell lines Parental 5-AzaC resistant K562 0.85 ± 0.57 373.04 ± 84.43 439* MV4-11 0.135 ± 0.016 325.31 ± 54.99 2408* IGROV1 301.15 ± 157.17 422.49 ± 167.03 1.5

Table 5.2 Cross-resistance of 5-azaC resistant cell lines to DAC. 5-AzaC resistant and parental cell lines were treated with variable concentrations of DAC and the IC50 values and resistant index determined. We show an increase (500 – >2000 fold) of resistant indices in the resistant cells relative to the parental cells. All experiments were performed in triplicate. *p<0.05

160

Figure 5.1 Dose-response curves of several parental and DAC and 5-azaC resistant leukemia and solid tumor cell lines treated with DAC and 5-azaC. Cells were incubated for 72 hours with variable concentrations of 5-azaC or DAC and cell survival was measured by MTS assay. A significant increase in survival for the resistant cells relative to the parental cell lines (a right shift in the curves) was seen. All experiments were performed in triplicate.

161

Figure 5.2 (A) mRNA levels of DNMTs in 5-azaC resistant cell lines relative to their parental counterparts. DNMT1, 3a and 3b mRNA levels were found to increase by 2-5 folds in MV4-11/5-azaC, while no change in DNMT1 and 3a levels was seen in K562/5- azaC and IGROV1/5-azaC; however DNMT3b decreased by 50-60%. DNMT1 and 3b levels increased slightly in HCT-15/5azaC with no change in the DNMT3a levels. (B) DNMTs mRNA levels in DAC resistant relative to their parental cell lines. DNMT1 and 3b mRNA levels increased by 3-14 fold in MV4-11/DAC but 3a decreased by 40%. DNMT1 level increased by 20% in K562/DAC in DNMT1 level but DNMT3a and 3b levels remained unchanged. Vertical bars are mean ± of SD in triplicate experiments. **p<0.01, * p<0.05

162

Figure 5.3 DNMTs protein expression in parental and 5-azaC resistant cell lines. (A) DNMT1 increased (10-60%) in all of the 5-azaC resistant cell lines relative to the parental counterparts. (B) DNMT 3a increased by 30% in MV4-11/5-azaC but decreased (40-70%) in K562/5-azaC, IGROV1/5-azaC and HCT-15/5-azaC cell lines.

163

Figure 5.4 DNMTs protein expression in parental and DAC resistant cell lines. (A) DNMT1 increased (20-90%) in both the DAC resistant cell lines relative to the parental cell lines. (B) DNMT3a decreased by 50% (C) DNMT3b increased by 3-fold for MV4- 11/DAC but K562/DAC decreased by 30%.

164

Figure 5.5 Global DNA methylation (GDM) pattern in parental and resistant cell lines (A) 5-AzaC resistant cell lines: K562/5-azaC showed a 15% increase, with no significant change in the other cell lines. (B) In DAC resistant cell lines, MV4-11/DAC and K562/DAC showed a decrease (1.5-2 fold). MV4-11 cells treated with 1 µM DAC was used as a positive control. Vertical bars are mean ± of SD in triplicate. **p<0.01, *p<0.05

165

Figure 5.6 Baseline levels of deoxyribonucleotide and ribonucleotide pools (dNTP/NTP) in K562/5-azaC and MV4-11/5-azaC and their parental cells. (A) K562/5-azaC showed a 40% increase in dATP with no change in dTTP and dCTP levels. All NTPs decreased by 2-4 fold (B) MV4-11/5-azaC showed a 30% increase in dATP level, and a 40% decrease in dTTP level, with no change in dCTP levels. NTPs levels decreased or remained unchanged. All changes were relative to the corresponding parental cell lines. Vertical bars are mean ± of SD in triplicate. **p<0.01, * p<0.05

166

Figure 5.7 Baseline levels of deoxyribonucleotide and ribonucleotide pools (dNTP/NTP) in IGROV1/5-azaC and HCT-15/5-azaC and their parental cell lines. (A) IGROV1/5- azaC showed a 63% increase in dTTP with no change in the dATP and dCTP levels. NTPs increased by 2-5 fold. (B) HCT-15/5-azaC showed a 40% increase in dCTP levels, with no change in the dTTP and dATP levels. NTPs increased (50%) or remained unchanged, all compared with those of the corresponding parental cell lines. Vertical bars are mean ± of SD in triplicate. **p<0.01, * p<0.05

167

Figure 5.8 Baseline levels of deoxyribonucleotide and ribonucleotide pools (dNTP/NTP) in MV4-11/DAC and K562/DAC relative to their parental cell lines. (A) In MV4- 11/DAC, all dNTPs levels increased by 4-30 fold. GTP and dGTP/ATP increased 2-2.5 fold, while the CTP and UTP remained unchanged (B) K562/DAC showed a 60% decrease in dTTP levels, while the dATP levels increased by 40% and dCTP levels was unchanged. dGTP/ATP levels decreased by 30% and GTP, CTP and UTP levels remained unchanged, all relative to the corresponding parental cell lines. Vertical bars are mean ± of SD in triplicate. **p<0.01, * p<0.05

168

Figure 5.9 DAC-TP and dNTPs levels in K562 and K562/DAC cell lines following DAC treatment. Cells were treated with 2 µM of DAC for 0, 1, 4 and 24 hours. DAC-TP and dNTPs levels were analyzed simultaneously by the LC-MS/MS method. (A) A 2-3 fold decrease in DAC-TP accumulations in the K562/DAC resistant cell lines relative to their parental cells in all time points assessed were seen. (B) dATP and dCTP levels increased by 30-60% and dTTP levels decreased by 50-70%. Vertical bars are mean ± of SD in triplicate **p<0.01 *p<0.05 169

Figure 5.10 DAC-TP and dNTPs levels in MV4-11 and MV4-11/DAC cell lines following DAC treatment. Cells were treated with 2 µM of DAC at 0, 1, 4 and 24 hours. DAC-TP and dNTPs levels were simultaneously analyzed by the LC-MS/MS method. (A) A 50% decrease in DAC-TP accumulation in the MV4-11/DAC resistant cell lines relative to their parental cells at 4 hours was seen. (B) All dNTPs increased dramatically (2-30 folds). Vertical bars are mean ± of SD in triplicate **p<0.01 *p<0.05

170

MV4-11 cells DAC DAC-TP 10 * ** 8

6 cells)

6

4

(pmol/10 2

DAC-TPAmount ofDAC and 0 1 4 24 Time (h)

Figure 5.11 DAC levels and DAC-TP accumulation in MV4-11. The cells were treated with 2 µM of DAC for 1, 4 and 24 h. DAC and DAC-TP levels were quantified by LC- MS/MS method. Vertical bars are mean ± SD of experiments performed in triplicate. **p<0.01,*p<0.05

171

Figure 5.12 Dose-response curves for 5-azaC resistant cell lines for cross-resistance to DAC. A significant increase in survival for the resistant cell relative to the parental cell lines was seen (A right shift in the curves). All experiments were performed in triplicate.

172

Figure 5.13 DAC-TP levels in parental and 5-azaC resistant cell lines following treatment with DAC. Cells were treated with 2 µM of DAC at 1, 4 and 24 hours. DAC- TP levels were analyzed by the LC-MS/MS method. A 30-80% decrease in DAC-TP accumulation was seen in 5-azaC-resistant cell lines relative to their parental counterparts. Vertical bars are mean ± of SD of experiments performed in triplicate. **p<0.01 *p<0.05

173 CHAPTER 6

BIOCHEMICAL MODULATION OF ARACYTIDINE (ARA-C) EFFECTS BY GTI-2040, A RIBONUCLEOTIDE REDUCTASE INHIBITOR, IN K562 HUMAN LEUKEMIA CELLS

6.1 Abstract

GTI-2040 is a potent antisense to the RRM2 subunit of the ribonucleotide reductase (RR), an enzyme involved in the de novo synthesis of nucleoside triphosphates.

We hypothesized that combination of GTI-2040 with aracytidine (Ara-C) could result in an enhanced cytotoxic effect with perturbed intracellular deoxynucleotide/nucleotide

(dNTP/NTP) pools including Ara-CTP. This study aims to provide a direct experimental support of this hypothesis by monitoring the biochemical modulation effects, intracellular levels of Ara-CTP, dNTPs/NTPs following the combination treatment of Ara-C and GTI-

2040 in K562 human leukemia cells. GTI-2040 was introduced into cells via electroporation. A hybridization-ligation ELISA was used to quantify intracellular GTI-

2040 concentrations. Real-time PCR and Western blot methods were used to measure the

RRM2 mRNA and protein levels, respectively. MTS assay was used to measure the cytotoxicity following various drug treatments. A non-radioactive HPLC-UV method was used for measuring the intracellular Ara-CTP, while a LC-MS/MS method was used to quantify intracellular dNTP/NTP pools. Intracellular GTI-2040 content was found to

174 increase with an increase in extracellular exposure concentrations. GTI-2040 was found to down-regulate RRM2 mRNA and protein levels and did not result in further reduction when combined with Ara-C. There was no change in RRM2 mRNA or protein level caused by Ara-C. GTI-2040 showed a significant decrease in dNTPs, but no further decrease in dNTPs was apparent with co-treatment with Ara-C. Interestingly, the GTP,

UTP, CTP and dGTP/ATP were significantly reduced by Ara-C alone, but not by GTI-

2040 alone or its combination with Ara-C. When combining GTI-2040 with Ara-C, a synergistic cytotoxicity was observed with no further change in dNTP/NTP pools.

Importantly, pretreatment of K562 cells with GTI-2040 was found to increase Ara-CTP level for the first time, and this effect may be due to inhibition of RR by GTI-2040. This finding provides a laboratory justification for the current phase I/II evaluation of GTI-

2040 in combination with Ara-C in patients with acute myeloid leukemia.

6.2 Introduction

Ribonucleotide reductase (RR), a highly regulated enzyme involved in the de novo synthesis of 2’-deoxyribonucleotides, plays a critical role in nucleoside metabolism

(198, 199). RR catalyzes the reduction of ribonucleotides (ADP, GDP, UDP, and CDP) to their corresponding deoxyribonucleotides (dADP, dGDP, dUDP, and dCDP) and this process is the rate-limiting step required for DNA replication (200). Human RR consists of a RRM1 subunit, which contains a substrate binding site, an allosteric site and a redox active disulfide, and a RRM2 subunit, which contains an oxygen-linked non- iron center and a tyrosine residue, all of which are essential for catalytic activity (201, 202).

RRM2 protein is only expressed during the late G1/early S phase essential for DNA

175 synthesis and repair, while RRM1 protein level remains relatively stable throughout the cell cycle (202). Another analog of RRM2 subunit has been discovered in recent times, which is p53-inducible known as p53R2 (33, 34). It complexes with RRM1 in non- proliferating cells and provides dNTPs for p53-dependent DNA repair and mitochondrial

DNA synthesis (35). It has been found that over-expression of RRM2 protein is associated with malignant and metastatic status of tumor cells. Inhibition of RR induces imbalance of ribonucleotide and deoxyribonucleotide levels, resulting in inhibition of

DNA synthesis and repair, cell cycle arrest and apoptosis (203). For this reason, RRM2 is considered a good target for anticancer drugs development (32, 38).

A number of RR inhibitors, such as hydroxyurea, gemcitabine, and antisense GTI

2040 have been developed (37, 38). GTI-2040, a 20-mer oligonucleotide complementary to the coding region of RRM2 mRNA with the sequence of 5’-

GGCTAAATCGCTCCACCAAG-3’ is designed to bind to RRM2 mRNA, resulting in the recruitment of RNase H which in turn leads to cleavage of the drug-mRNA complex and degradation of the target mRNA. In vitro studies have demonstrated that treatment of a variety of tumor cell lines, such as human H460 lung carcinoma, human T24 bladder cancer and murine L cell lines, with GTI-2040 led to a sequence- and target- specific down-regulations of RRM2 mRNA and protein levels (32). In mice bearing Burkitt's lymphoma, GTI-2040 treatment greatly increased their survival rate (32). A phase I clinical evaluation of GTI-2040 was conducted (31), which has established its clinically safe dose. Since RR mediates reduction of ribonucleotides, it is expected that its inhibition by GTI-2040 should result in alteration of intracellular dNTP levels and such

176 could provide potential combination treatment strategies with antimetabolite drugs that modulate DNA synthesis.

Ara-C is a widely used antimetabolite for the treatment of acute leukemia (20,

21). Ara-C needs to be phosphorylated to Ara-C triphosphate (Ara-CTP) by deoxycytidine kinase in the cell to compete with dCTP for incorporation into DNA. This incorporation causes DNA synthesis inhibition and cell death (22). If intracellular dNTP levels, especially dCTP, are reduced, an increase in Ara-CTP level is expected, leading to increased antitumor activity of Ara-C (Figure 6.2). Based on this rationale, a phase I studies of GTI-2040 in combination with Ara-C for the treatment of acute myeloid leukemia was carried out at our institution (27). This study has demonstrated that GTI-

2040 and Ara-C can be safely given to AML patients (31). However, the experimental verification of their biochemical modulation following combination therapy remains to be illustrated. Herein, we demonstrate the use of a newly developed sensitive LC-MS/MS method (117) to probe the perturbation of dNTP pools in the cells following pretreatment with GTI-2040 to provide a direct experimental support for the above rationale. This was coupled to the use of a newly developed non-radioactive HPLC-UV assay to quantify cellular Ara-CTP levels. Additionally, intracellular levels of GTI-2040, RRM2 mRNA and protein levels were determined to further explore the mechanism of action of GTI-

2040. This information will be critical in optimizing dosing regimen for future clinical trials. The effective correlation analysis of these important biomarkers will further substantiate our justification for the utilization of these two drugs in combination therapy in leukemia patients.

177 6.3 Materials and methods

6.3.1 Chemicals

GTI-2040 and Ara-C were provided by The National Cancer Institute (Bethesda,

MD) and used without further purification. Stock solutions of both drugs were prepared and further diluted in phosphate buffered saline (PBS, Invitrogen, Carlsbad, CA).

6.3.2 Cell culture

The human leukemia cell line K562 was obtained from American Type Culture

Collection (ATCC, Manassas, VA). Cells were cultured in RPMI 1640 media (supplied by Tissue Culture Shared Resource, Comprehensive Cancer Center, The Ohio State

University, Columbus, Ohio) supplemented with 10% fetal bovine serum (FBS)

(Invitrogen, Rockville, MD), L-Glutamine (Invitrogen) and penicillin-streptomycin antibiotics (Gibco, Rockville, MD). The cell line was maintained at 37 °C in a humidified environment with 5% CO2. Viability and cell counts were determined using trypan blue dye exclusion assay (204). GTI-2040 was introduced into cells using either neophectin transfecting agent (NeoPharm, IL) or with an electroporation device (BIO-RAD Lab, CA,

USA). However, transfecting agent can be toxic to cells and may have potential unknown biological activity. Therefore, further studies with GTI-2040 were done using electroporation.

178 6.3.3 Determination of intracellular GTI-2040 concentrations by a hybridization-

based ELISA

5×106 K562 cells were treated with GTI-2040 at 0, 1, 5, 10, 20 μM for 24 hours using an electroporation delivery technique. After 24 hours, cells were harvested and intracellular GTI-2040 levels were measured using a previously developed two-step hybridization-ligation ELISA assay (56). Briefly, GTI-2040 was first base-paired to the capture probe in a polypropylene 96-well plate. 10% Triton X-100 was added to the mixture solution and incubated at 42C for 2.5 hours for hybridization. The resulting solution was transferred to a NeutrAvidin-coated 96-well plate (Fermentas, MD), which was incubated at 37C for 30 minutes to ensure the attachment of biotin labeled capture probe to NeutrAvidin-coated wells. After washing six times, the ligation solution containing T4 ligase and detection probe was added to each well followed by the addition of S1 nuclease solution. The reaction was blocked with Superblock buffer (Pierce, IL).

Anti-digoxigenin-alkaline phosphatase (AP) was then added into each well. Following addition of substrate solution (36 mg Attophos [Promega, WI] in 60 ml diethanolamine buffer), fluorescence intensity was measured at Ex 430/Em 560 (filter=550nm) using a

Gemini XS fluorescence microtiter plate reader (Molecular Devices, Sunnyvale, CA).

6.3.4 RNA isolation and RT-qPCR

To measure intracellular RRM2 mRNA level, 5×106 K562 cells were first treated with GTI-2040 alone at 10 μM for 24 hours followed by continuous treatment with Ara-C at concentrations of 5, 10 and 20 µM for and additional 48 hours, while the GTI-2040 exposure was maintained. Total RNA was isolated using Trizol reagent (Invitrogen). 179 Briefly, cell lysate was treated with chloroform and the total RNA was precipitated with isopropyl alcohol, followed by a washing step with 75% ethanol. RNA was then dissolved in RNase free water and its concentration and purity was measured by a

Nanodrop spectrophotometer (Nanodrop Technologies). cDNA was synthesized from 2

µg total RNA using Moloney murine leukemia virus reverse transcriptase (Invitrogen).

The cDNA templates and primers were then mixed with reagents from a SYBR Green

PCR Master Mix (Applied Biosystem, Foster City, CA). Reactions were carried out in triplicate in ABI prism 7700 sequence detector (Applied Biosystems), and data were analyzed by comparative CT method. Dissociation curves were also obtained to examine the purity of the amplified products. The amount of RRM2 mRNA in each sample was normalized with respect to an internal control, abl. The relative changes in treated groups were expressed as a percentage of untreated control (arbitrarily set at 1). The results were expressed as the mean ± SD from triplicate determinations.

6.3.5 Western blot analysis

5×106 K562 cells were first treated with GTI-2040 alone at 10 µM for 24 hours followed by Ara-C continuous exposure at concentrations of 5, 10 and 20 µM for 48 hours. Cells were then harvested, washed with 1 mL ice-cold PBS, and centrifuged at

1000 g for 5 minutes at 4 ºC. The pellet was obtained and re-suspended in 100 μL lysis buffer (50 mM pH 7.6 Tris-HCl, 250 mM NaCl, 5 mM EDTA, 2 mM Na3VO4, 50 mM

NaF and 1% protease inhibitor cocktail) (P8340, Sigma) for 30 minutes on ice. The lysate was sonicated for 10 seconds. Total protein concentration was determined using the BCA protein assay method (Pierce, Rockford, IL). Equal amounts of protein for each sample

180 were incubated with 6x SDS loading buffer (100 mM, pH 6.8 Tris, 200 mM DTT, 4%

SDS, 20% glycerol, and 0.015% bromphenol blue) and boiled for 5 minutes. The proteins were then separated on 15% SDS-polyacrylamide gels and transferred to nitrocellulose membranes (Amersham, Piscataway, NJ). The RRM2 protein was recognized following treatment with a goat antihuman RRM2 polyclonal antibody (E-16) (Santa Cruz

Biotechnology, Santa Cruz, CA) as the first antibody, followed by a HRP (horseradish peroxidase) conjugated anti-goat IgG secondary antibody. RRM2 protein (MWT 45,000 dalton) was detected by ECL (Amersham, Arlington Heights, IL) and GAPDH was used as the internal loading control. RRM2 protein expressions were quantified by densitometry and normalized to GAPDH.

6.3.6 Determination of intracellular dNTP and NTP Pools

10×106 K562 cells each were treated either with Ara-C alone at concentrations of

5, 10, and 20 µM for 24 hours, or pre-treated with GTI-2040 at 10 µM and 24 hours later treated with continuous exposure of Ara-C at concentrations of 5, 10 and 20 µM for 48 hours. Cells were then lysed and dNTPs/NTPs were extracted and quantified by our previously published method (117). Briefly, cells were counted and monitored for viability using trypan blue exclusion test before extraction. Following centrifugation at

1000 g for 5 minutes, cell pellets were washed with phosphate buffered saline (PBS) and deprotenized with an addition of 1 mL 60% methanol. The resulting solution was vortex- mixed for 20 seconds, incubated in -20 °C for 30 minutes and sonicated for 15 minutes in an ice bath. Cell extracts were centrifuged at 1000 g for 5 minutes at 4 °C and the supernatant was separated and dried under a stream of nitrogen. The residues were

181 reconstituted with 300 μL of water, vortex-mixed for 20 seconds and the cell extracts were centrifuged at 1000 g for 5 minutes at 4 °C. A 50 μL aliquot of the resulting supernatants was injected into an ion-trap LC-MS/MS system (LCQ, Thermo Scientific,

San Jose, CA) for dNTP and NTP measurements.

6.3.7 Growth inhibition assay

To evaluate the growth inhibitory effect of GTI-2040 and Ara-C on K562 cells,

MTS cytotoxicity assay was performed. GTI-2040 was introduced into K562 cells by using either neophectin transfecting agent or with an electroporation device. Briefly, cells were seeded into 96 well plates and the cells were pretreated with fixed concentrations 5 or 10 µM of GTI-2040 for 24 hours. Increasing concentrations of Ara-C at 0.01, 0.1, 1,

10, 100 µM, were then added and the incubation was continued for 72 hours. In a reverse sequence, K562 cells were pretreated with a fixed concentration 1 µM of Ara-C for 24 hours, then graded concentrations of GTI-2040 at 1, 3, 10, 30, 100 µM, were added and the incubation was continued to 72 hours. Afterward, 20 µL of 3-(4,5-dimethylthisazol-

2-yl)-5-(3-carboxymethoxyphenyl-2-(4-sulfophenyl)-2H-tetrazolium, inner salt (MTS) and phenazine methosulfate (PMS) (Promega, Madison, WI) were added to each well in a

20:1 ratio. After 2 hours incubation, the absorbance was read at 490 nm on a microplate reader Gemini XS (Molecular Devices, CA) to determine levels of formazan product as a measure of cell viability against a non-drug treated control. Determination of IC50 values was performed using WinNonLin computer software (Pharsight, Mountain View, CA).

Combination index (CI) values were determined based on the Chou-Talalay method

(116).

182 6.3.8 Determination of intracellular Ara-CTP level in human K562 leukemia cells after treatment with Ara-C alone

10×106 K562 cells were treated with Ara-C for 4 hours at 1, 5, 10 and 20 μM. The 4 hours time point was chosen based on previous studies done in our laboratory that demonstrates the maximum accumulation of Ara-CTP occurred at that time point. After treating with trypan blue dye and counting with a hemocytometer, cells were centrifuged and the harvested pellets were washed three times with cold PBS. Ara-CTP was extracted and determined as described in the Method Section for dNTP/NTP extraction. K562 cells were appropriately used for these study since previous report demonstrated K562 cells incubated with Ara-C, were able to accumulate up to 100 µM of the drug (205, 206).

6.3.9 Statistical analysis

Statistical significance of differences observed was determined using a Student’s t-test. The minimal level of significance was a p-value below 0.05. Pearson correlation was used to determine the coefficient of correlation. Both analyses were done using

Minitab statistical software (Minitab, State College, PA). Excel (Microsoft Office Word

2007, WA) was used in the data plot.

183 6.4 Results

6.4.1 Determination of GTI-2040 concentrations in human leukemia K562 cells

In order to confirm the uptake of GTI-2040 into cells, intracellular GTI-2040 levels were determined. Following the exposure of K562 cells to GTI-2040 at 1, 5, 10, and 20 µM for 24 hours, GTI-2040 cellular concentrations were found to be 168.4 ± 66.8,

373.1 ± 59.5, 1073.9 ± 66.2 and 1842.8 ± 130.9 pmol/106 cells, (mean ± SD, n=3), respectively (Figure 6.3). Thus, intracellular GTI-2040 content was found to increase with an increase in extracellular exposure concentrations.

6.4.2 GTI-2040 reduces RRM2 mRNA levels in K562 cells

Following pretreatment of K562 cells with 10 µM GTI-2040 for 24 hours, RRM2 mRNA level was significantly decreased by about 40-50% relative to the untreated control. Treatment of these cells with 5, 10 and 20 µM of Ara-C caused no change in

RRM2 mRNA level and addition of Ara-C (5, 10 and 20 µM) to the 24 hour pre-treated cells with 10 µM GTI-2040 for an additional 48 hours also did not result in further reduction of RRM2 mRNA level (Figure 6.4). These data confirmed that GTI-2040 regulates RRM2 at the transcriptional level and indicated that RRM2 is not a target of

Ara-C.

6.4.3 GTI-2040 decreases RRM2 protein expression

To investigate the effects of GTI-2040 on RRM2 protein expression, K562 cells were exposed to either GTI-2040 alone or in combination with Ara-C. Pretreatment with

10 µM GTI-2040 for 24 hours caused a significant reduction in RRM2 protein (~50%)

184 and subsequent combination treatment with 5, 10 and 20 µM Ara-C did not cause further decrease in RRM2 protein expression (Figure 6.5), consistent with the expected mechanism of RRM2 mRNA down-regulation. Taken together, these data confirmed that

GTI-2040 but not Ara-C regulates RRM2 expression at both the transcriptional and translational level.

6.4.4 GTI-2040 perturbs the intracellular ribonucleotide (NTPs) and

deoxyribonucleotide (dNTPs) pools in K562 cells

Since RRM2 is required for catalyzing the reduction of ribonucleoside diphosphates to the corresponding deoxyribonucleotides (199), we wanted like to probe the intracellular dNTP/NTP pool changes caused by GTI-2040 (117) or Ara-C alone and by their combination. K562 cells were first exposed to 10 µM GTI-2040 for 24 hours, followed by treatment with 10 or 20 µM Ara-C for 48 hours. dNTPs and NTPs were extracted and measured by the LC-MS/MS method as described above. We have found a significant decrease (~40%) in intracellular dTTP, dATP and dCTP levels (Figure 6.6A) following 10 µM GTI-2040 exposure alone when compared to the untreated control, but with no effect on the NTP pool. No further decrease was apparent with co-treatment with

10 or 20 µM Ara-C; the small variation due to different Ara-C doses was within the statistical errors. Interestingly, the GTP, UTP, CTP and dGTP/ATP were significantly reduced by Ara-C alone (40-60%), but not by GTI-2040 alone or its combination with

Ara-C (Figure 6.6B). The lack of change for the latter may be due to the RRM2 inhibition of GTI-2040.

185 6.4.5 Inhibition growth assay

Effect of GTI-2040 on Ara-C cytotoxicity in K562 cells was examined following

5 or 10 μM GTI-2040 pretreatment. As shown in Figure 6.7A, cytotoxicity of Ara-C alone was observed with an estimated IC50 of 0.13 μM. Following pre-incubation with

GTI-2040 at 5 μM and 10 μM for 24 hours, IC50 values of Ara-C were reduced to 0.014

μM (a 9.3-fold increase in cytotoxicity) and 0.012 μM (a 10.8-fold increase in cytotoxicity), respectively. Cytotoxicity of their combination in the reverse sequence was also evaluated in K562 cells. As shown in Figure 6.7B, co-treatment of K562 cells, with

Ara-C and GTI-2040 with pre-incubation of Ara-C at 1 μM, reduced the IC50 values of

Ara-C to 0.00076 μM (a 171-fold increase in cytotoxicity compared to Ara-C alone).

These data suggest that there were synergistic effects between GTI-2040 and Ara-C, which may be more effective when Ara-C treatment precedes GTI-2040.

6.4.6 Dose-dependent accumulation of Ara-C in K562 cells

In order to monitor the accumulation of Ara-C, the concentration-time course of

Ara-CTP in K562 cells following its treatment was followed and the Ara-CTP level peaked at 4 hours and sustained to 24 hours as reported earlier from our laboratory.

Furthermore, the accumulation level of Ara-CTP occurred in a dose-dependent manner following Ara-C treatment for 4 hours (Figure 6.8).

186 6.5 Discussion

In this study, the biochemical modulation effects of Ara-C by GTI-2040 in K562 human leukemia cells were investigated. There are several human leukemia cell lines available and only K562 cell line was selected to mimic the clinical situation, since this cell line was found to have higher baseline levels of RRM2 expression (data not shown), which is similar to those responders of AML patients treated with GTI-2040 and high dose Ara-C (27). Ara-C is one of the most effective anticancer agents for the treatment of

AML and its therapeutic effect at low or high doses has been extensively studied (207-

214). It has been reported that the metabolism of Ara-C in vitro was greatly enhanced by

RR inhibitors, such as amidox and trimidox (215-217). Clinical combination studies of

Ara-C with RR inhibitors such as fludarabine and chlorodeoxyadenosine demonstrated significant accumulation of Ara-CTP in lymphocytes or sustained inhibition of DNA synthesis in circulating leukemia blasts from patients with leukemia (218-220). Thus, it appears that inhibition of RR resulted in inhibition of de novo dNTP synthesis and enhances Ara-C sensitivity and even overcomes Ara-C resistance (221, 222). However, these RR inhibitors are primarily small molecules and possess other pharmacological effects. As an antisense RR inhibitor, GTI-2040 may provide more advantages in combination treatment with Ara-C. First, GTI-2040 is less toxic with its IC50 > 100 μM in human leukemia K562 cells. Phase I studies of GTI-2040 alone for the treatment of advanced solid tumors or lymphoma indicated that GTI-2040 was generally well tolerated as a single agent with only two patients experiencing a dose limiting reversible hepatic toxicity (31). Therefore, GTI-2040 will unlikely result in significant adverse

187 effect in combination with Ara-C. Second, GTI-2040 is stable and efficiently inhibits RR expression through direct base-pairing to RRM2 mRNA, leading to RR down regulation.

On the contrary, conventional analogs such as fludarabine and clofarabine are susceptible to enzymatic phosphorolysis by Escherichia coli nucleoside phosphorylase and are acid labile. Taken together, GTI-2040 appears to be a good candidate in combination treatment with Ara-C for human leukemia.

Our results showed significant down-regulation effect of RRM2 mRNA and protein in K562 cells following GTI-2040 treatment. In this study, GTI-2040 was introduced into cells using electroporation method, aiming to avoid potential additional biological effects contributed from transfecting agents that are widely used for oligonucleotides. Our results showed that about 60-70% of GTI-2040 was found to be introduced into the cell with electroporation, which is comparable to that using a transfection agent (54). Following GTI-2040 delivery, there was a significant down- regulation of RRM2 mRNA and protein levels. These effects provide an additional experimental verification of the proposed mechanism of GTI-2040 in leukemia cells (32).

Our data also confirmed that when the intracellular RRM2 expression was reduced by

GTI-2040, down-stream dNTP pool was perturbed, which may enhance Ara-C’s incorporation into DNA (223, 224).

Our data showed a significant decrease in intracellular dTTP, dATP and dCTP levels following treatment with GTI-2040 alone or in combination with Ara-C as expected with no change in NTP levels. Interestingly, the GTP, UTP, CTP and

188 dGTP/ATP pools were significantly reduced by treatment with Ara-C alone (40-60%) and this may be due to salvage or feedback regulations (Ara-CTP conceived as CTP).

Alternatively, this could be due to its modulation of synthetases (e.g. CTP synthetase), as it was shown that some pyrimidine antimetabolites could inhibit CTP synthetase causing decrease of CTP level (225). This enzyme catalyzes the formation of CTP from UTP and plays a role in the determination of cellular CTP and dCTP pools (165, 225). Possibly,

Ara-C inhibits CTP synthetase and concomitantly, disrupts both the intracellular CTP and dCTP.

Our study also demonstrated a synergistic effect in cytotoxicity in K562 cells with combination of GTI-2040 and Ara-C. At pretreatment concentrations of 5 and 10 μM of

GTI-2040, followed by Ara-C, the combination indices (CI) (116) were estimated to be

0.5 and 0.9, respectively, (all <1) suggesting a synergistic effect of this combination according to the literature (116). Interestingly, the extent of the synergistic effects was more significant when Ara-C treatment preceded the treatment with GTI-2040 (a 171- fold increase) with a combination index of 0.1. Possibly, this further enhancement could be due to the first incorporation of Ara-CTP into DNA to compete with dCTP causing immediate chain termination and blockade of DNA synthesis (58). Further addition of

GTI-2040 resulted in an enhanced depletion of dCTP pool, a sustained higher cellular

Ara-CTP levels and thus potentiation of cytotoxicity. Despite these differences, in both treatment orders, we have demonstrated that GTI-2040 enhanced the cytotoxicity of Ara-

C in leukemia cells. Based on our cytotoxicity results, alternative sequencing schedules for the combination should be considered in future clinical trial design with careful

189 attention for relevant pharmacodynamic endpoints. A non-radioactive HPLC-UV method was optimized in our study to quantify the Ara-CTP level in the cell. This method was then applied to analyze intracellular Ara-CTP level following Ara-C treatment, which showed time and dose dependent Ara-CTP accumulation. Our laboratory also reported that pretreatment with GTI-2040 at 10 or 20 μM, followed by Ara-C exposure led to significant enhancement of intracellular Ara-CTP level supporting the role of GTI-2040 potentiation of Ara-C.

6.6 Conclusion

GTI-2040 was found to inhibit ribonucleotide reductase by down-regulation of

RRM2 mRNA and protein levels and its inhibition results in a decrease of dNTP levels.

Reduction in RRM2 expression levels correlated well with the depletion of dNTP levels following GTI-2040 treatment. Furthermore, the combination of GTI-2040 with Ara-C produced an increase in cytotoxicity. These provide a laboratory and mechanistic justification for the current phase I/II evaluation of GTI-2040 in combination with Ara-C in patients with acute myeloid leukemia.

190 NH2 NH2 NH2

N N HC N N N HC N O N O N O

HO HO HO O O O H H H H H H H H H H H H OH OH OH NH2 OH OH NH2 NH2 cytidine 5-aza-cytidine decitabine NH2 N N HC N N N N Hum (RRM2) mRNA HCHum (R2) mRNA 194 1364 2475 N O CODING N O N O N O 5’UTR 3’UTR Poly A HO HO HO O HO H OH 626 645 O O O H H 3’ GAACCACCTCGCTAAATCGG 5’ GTI -2040 OH H H H H H H cytarabine H H H H H H OH OH OH OH OH

cytidine 5-aza-cytidine decitabine N H 2

N Hum (R2) mRNA

194 1364 2475 N O CODING

5’UTR 3’UTR Poly A H O

O

H O H 626 645

H H 3’ GAACCACCTCGCTAAATCGG 5’ GTI -2040 O H cytarabine Aracytidine (Ara-C)

Figure 6.1 Structures of GTI-2040 and Aracytidine (Ara-C).

191 Ara - C

hENT1 Cell membrane CDA Deamination Ara - C

dCK 5’ -NT CDP Ara-C GTI-2040 P RR UMP- CMP Kinase dCDP Ara - C P P NDP Kinase

Ara - C P P P dCTP Nuclear membrane dCTP Ara - CTCTPP DNA incorporation Inhibit DNA synthesis

DNA

Apoptosis

Figure 6.2 Diagrammatic rationales for the combination of GTI-2040 with Ara-C in leukemia cells. GTI-2040 inhibits ribonucleotide reductase (RR), which causes depletion of dNTP and NTP pools, hence allowing phosphorylated Ara-C (Ara-CTP) to be incorporated into DNA resulting in DNA synthesis inhibition and apoptosis.

192

Figure 6.3 Intracellular accumulation of GTI-2040 in K562 cells following introduction of GTI-2040 at the indicated concentrations for 24 hours by electroporation. Vertical bars represent mean ± SD from triplicate experiments. * p<0.05, ** p<0.01, versus concentration at 1 µM.

193

Figure 6.4 GTI-2040 reduces RRM2 mRNA expression in K562 cells alone. Ara-C had no effect on RRM2 mRNA expression alone or no further reduction in RRM2 mRNA expression caused by 10 µM GTI-2040. RRM2 mRNA levels were normalized by abl and vertical bars represent mean ± SD from triplicate experiments and compared with the control with asterisks showing p<0.05.

194

Figure 6.5 GTI-2040 decreases RRM2 protein expression in K562 cells. RRM2 protein levels were decreased by about 50% following GTI-2040 treatment at 10 µM. Ara-C alone had no effect on RRM2 protein levels and addition of various concentrations of Ara-C did not contribute to further reduction of RRM2 protein levels caused by pre- incubation with GTI-2040 at 10 µM. RRM2 protein levels were quantified by densitometry and expressed as the ratios over the loading control GAPDH.

195

Figure 6.6 Perturbation of intracellular dNTP/NTP pools by GTI-2040 and Ara-C. K562 cells were treated with GTI-2040 at the indicated concentrations in the absence and presence of GTI-2040. (A) GTI-2040 at 10 µM reduced levels of dTTP, dATP and dCTP by 40%; whereas Ara-C showed no effect; (B) GTI-2040 did not change the NTP pools; however, Ara-C at 10 and 20 µM decreased GTP, CTP, UTP and dGTP/ATP by 40-60%. dNTP/NTP levels are presented as the percentage of untreated control. Vertical bars are mean ± SD of triplicate experiments with asterisks showing p<0.05 versus control.

196

A Ara-C alone 5µM GTI-2040 + variable ara-C 10µM GTI-2040 + variable ara-C 140 120 100 80 60 40 20

0 Cell survival (% of survival control)Cell 0.001 0.01 0.1 1 10 100 Concentration of Ara-C (µM)

B

GTI-2040 alone 140 Pre-treatment with 1 µM Ara-C + 120 variable GTI-2040 Ara-C alone 100

80

60

40

Cell Survival (% control) (% Survival Cell 20

0 0 0.01 0.03 0.1 1 3 5 10 30 Concentration (µM)

Figure 6.7 Effect of GTI-2040 on cytotoxicity of Ara-C in K562 cells. (A) Pretreatment of 5 or 10 μM GTI-2040 via electroporation, decreased the IC50 of Ara-C; (B) Pretreatment of 1 μM Ara-C followed by the GTI-2040 treatment via neophectine also decreased the IC50 of Ara-C (*p<0.05, versus control).

197

K562 cells cells) 6 160 ** 140 120 100 ** 80

60 ** 40 20

Ara-CTP accumulation (pmol/10 accumulation Ara-CTP 0

Ara-C (µM) 1 5 10 20

Figure 6.8 Dose-dependent accumulation of Ara-CTP following 4 hours treatments in K562 cells. Vertical bars represent mean ± SD from triplicate experiments. ** p<0.01, versus concentration at 1 µM.

198 CHAPTER 7

PHARMACOKINETICS AND PHARMACODYNAMICS ANALYSIS OF GTI- 2040 AND ARACYTIDINE (ARA-C) IN PATIENTS WITH ACUTE MYELOID LEUKEMIA (AML) IN A PHASE II CLINICAL TRIAL

7.1 Abstract

Aracytidine (Ara-C) is a nucleoside analog that is sequentially activated to its triphosphate, Ara-CTP metabolite, which competes with endogenous deoxynucleotides triphosphates (dNTPs) for incorporation into DNA. Ara-C has been a mainstay for the treatment of acute myeloid leukemia (AML). GTI-2040 is a 20-mer antisense oligonucleotide that targets and down-regulates the RRM2 subunit of ribonucleotide reductase (RR), an enzyme that reduces endogenous nucleotides (NTP) to deoxyribonucleotides (dNTPs). Therefore, we hypothesize that GTI-2040 enhances the anti-leukemia activity of Ara-C by decreasing RRM2 expression and consequently down- regulates the endogenous dNTP pools, especially dCTP levels, allowing Ara-C to effectively compete for DNA incorporation. The purpose of this study is to characterize the pharmacokinetic and pharmacodynamic properties of GTI-2040 and Ara-C in patients with refractory/relapsed AML enrolled in a phase II clinical trial that tested the feasibility of this combination. GTI-2040 at 5 mg/kg/day was continuously administered to AML

199 patient by continuous IV infusion (CIVI) for 6 days and Ara-C at 3 g/m2/dose was infused over 2 hours every 12 hours. Pharmacokinetic (PK) and pharmacodynamic (PD) analysis was performed in 25 patients enrolled in this phase II clinical trial (Protocol

OSU #07028). To perform plasma pharmacokinetic analysis, GTI-2040 plasma concentration was determined by a previously developed ultra-sensitive and highly selective hybridization ELISA assay. Pharmacokinetic analysis was accomplished using

WinNonLin computer software using a two-compartment continuous IV infusion model.

A LC-MS/MS method developed in our laboratory was used to quantify endogenous

NTPs and dNTPs levels in bone marrow cells (BM) obtained from patients. A non- radioactive HPLC-UV assay was used to quantify the Ara-CTP levels in BM cell samples. PK-PD modeling and simulation was also performed. Our data showed that there was no difference in PK parameters of GTI-2040 in patients pre-treated with GTI-

2040 followed by Ara-C, when compared with those of co-treatment of GTI-2040 with

Ara-C. dNTPs/NTPs and Ara-CTP were detectable in most of the patient samples analyzed. However, we were unable to establish any correlation between the dNTP/NTP pools and Ara-CTP accumulation, due to the large variability of the data among samples before and after treatment. However, the PK-PD simulation suggests the reduction in dCTP levels can enhance the incorporation of Ara-CTP in cells. These data validate our hypothesis relative to pharmacodynamics. The correlation of the pharmacokinetic and pharmacodynamic measures of GTI-2040 and Ara-C may be useful in assessing the overall response rate of AML patients’ future clinical studies involving RRM2 inhibitors and nucleoside analogs.

200 7.2 Introduction

Acute myeloid leukemia (AML) remained a critical disease lacking effective therapy until the mid 1960s (5) and is also one of the most common types of leukemia in the Western Hemisphere. AML arises from the hematopoietic system and consists of accumulation and expansion of immature myeloid leukemia cells in the bone marrow

(BM) (226). A number of clinical factors such as age, prior chemotherapy and CD34+ immunophenotype have been found to be important in predicting prognosis. None of these factors can, however, predict reliably clinical outcome, because each of the aforementioned features must be correlated with each other (226). Non-random chromosomal abnormalities are one of the most important independent prognostic indicators for achievement of complete remission (CR), duration of the first CR and survival following intensive chemotherapy (227, 228). These abnormalities are identified at the cytogenetic level in about 55% of all adult AML patients and used to direct initial treatment strategies (227, 228). However, the management of AML is still complex with only approximately 40-50% of treated patients achieving a long-term remission (4), even when patients receive chemotherapy as soon as diagnosed. Current treatment of AML is even more challenging in older adult patients, with about 75% of AML patients older than 60 years exhibiting significant toxicity with overall poor response and survival rates.

Overall, about 20% to 30% of AML patients never achieve CR, and among the patients who achieve CR, about 50% eventually have disease relapse (20, 227, 228). The lack of significant expansion of AML treatments in patients, therefore calls for development of novel therapeutic strategies.

201 Nucleoside analogs constitute the backbone of a variety of primary and salvage chemotherapy regimens for AML patients (20, 74). These compounds mimic physiological nucleosides in terms of uptake and metabolism and therefore incorporate into newly synthesized DNA, resulting in synthesis inhibition and chain termination and eventually, apoptosis (5). Ara-C is one of the widely used nucleosides for the treatment of hematologic malignancy and serves as the backbone drug for initial or salvage treatment of AML (74, 83). Ara-C is carried into cells by a nucleoside transporter; human equilibrative nucleoside transporter (hENT1) (58, 59) and following intracellular uptake,

Ara-C undergoes rapid phosphorylation to its triphosphate form (Ara-CTP). Ara-CTP then competes with the natural deoxycytidine triphosphate for incorporation into DNA by

DNA polymerase, and results in termination of the strand elongation important for DNA synthesis/repair (229, 230). Due to a relationship between the intracellular levels of Ara-

CTP and antileukemia effect (181, 206, 231, 232), strategies to enhance their DNA incorporation is pursued. Inhibition of ribonucleotide reductase (RR) is one of the potential strategies to increase the cytotoxicity derived from nucleoside analogs. RR catalyzes the reduction of ribonucleotides (ADP, GDP, UDP and CDP or NTPs) to their corresponding deoxyribonucleotides (dADP, dGDP, dUDP and dCDP or dNTPs), which is a rate- limiting step for DNA synthesis and repair (23). Exposure of proliferating cells to RR inhibitors produces an imbalance in deoxyribonucleoside triphosphates, thereby resulting in irreversible cell damage and apoptosis (26, 32).

GTI-2040 is a 20-mer oligonucleotide that is complementary to the RRM2 mRNA, subunit of RR. RR is an important enzyme for the metabolism of nucleoside analogs such as Ara-C commonly used as upfront chemotherapy in treatments in AML

202 patients (59). Reduction in RRM2 expression by treatment with GTI-2040, will allow

Ara-C to incorporate into newly synthesized DNA resulting in synthesis inhibition and chain termination (59). The antitumor activity of GTI-2040 has been evaluated in a variety of studies in vitro and in vivo, in addition to assessment of its down-regulation of the target RRM2 mRNA and protein levels. Based on these preclinical studies, a phase I study of GTI-2040 in combination with Ara-C in patients with refractory or relapsed

AML has been completed at our institution. At the end of the study, the maximal tolerated dose (MTD) of GTI-2040 was established to be at 5 mg/kg/day continuous IV infusion (CIVI) for a total of 144 hours on days 1-7 combined with Ara-C at 3 g/m2/dose administered every 12 hours on days 2-7 for younger patients (ages < 60 years). Down- regulation of the target RRM2 by GTI-2040 was demonstrated and appeared to correlate well with clinical response. To improve the understanding of pharmacodynamic activities of both GTI-2040 and Ara-C, their surrogate endpoints need to be assessed in future clinical trial design of this combination.

Recently, a phase II clinical trial of the combination of GTI-2040 with high dose

Ara-C in chemotherapy patients (18-59 yrs) with refractory or relapsed AML has been implemented and is ongoing. This study would further justify our hypothesis that the depletion of deoxyribonucleotides pools by GTI-2040 will enhance the incorporation of

Ara-C into DNA, hence increase in apoptosis. The primary goal of this trial is to evaluate the clinical efficacy and the secondary goal is to evaluate the clinical pharmacokinetic of

GTI-2040 and pharmacodynamic activities of both drugs, which is the focus of this chapter. The study was strategized to assess the pharmacokinetic of GTI-2040 before and

203 after treatment with Ara-C. In addition, in order to further validate RRM2 target and to understand other relevant biological endpoints, pharmacodynamics of GTI-2040 before and after treatment with Ara-C was assessed. To effectively study the pharmacodynamic endpoints of both drugs, the trial was designed as a pilot PD and a phase II PD groups. In the pilot PD group, addition of GTI-2040 is delayed until 24 hours after initiation of Ara-

C to evaluate the pretreatment effect of GTI-2040. In the phase II PD group, in order to observe pretreatment effect of GTI-2040 then combined with Ara-C, GTI-2040 is given

24 hours prior to addition of Ara-C, Finally, the correlation of PK-PD endpoints with clinical response will be evaluated and in addition to model simulation. The evaluation of the pharmacokinetic and pharmacodynamic of this combination will eventually be used to determine the overall response rate complete remission (CR) and CR with incomplete blood count recovery (CRi).

7.3 Materials and methods

7.3.1 Drugs administration

GTI-2040 (5’-GGCTAAATCGCTCCACCAAG-3’) was supplied by Lorus

Therapeutics Inc. in a glass vial as a concentrated sterile solution (100 mg/ml) and was further diluted with sterile normal saline (0.9% sodium chloride, USP). Aracytidine (Ara-

C) is commercially available in 100 mg, 500 mg, 1g, 2g, multi-dose vials in 0.68% NaCl solutions, and was diluted with sterile water containing no preservatives. GTI-2040 was administered by continuous IV infusion at 5 mg/kg/day and Ara-C infused at 3 g/m2/dose.

204 7.3.2 Clinical trial design

Patients with relapsed or refractory AML were treated with GTI-2040 in combination with Ara-C in a phase II trial at the James Cancer Hospital at The Ohio State

University under Protocol OSU #07028 approved by the Institutional Review Board

(IRB) and by Center of Therapy and Evaluation Program (CTEP) of the National Cancer

Institute. GTI-2040 was administered by continuous IV infusion at 5 mg/kg/day and Ara-

C infused at 3 g/m2/dose. Plasma samples were collected and the pharmacokinetic analysis before and after GTI-2040 treatment was done. The pharmacodynamic assessment was performed by stratification of the trial into pilot PD and phase II PD groups (delayed GTI-2040 and early GTI-2040 treatments), respectively. In the pilot PD group, GTI-2040 (5 mg/kg/day) was administered by CIVI for a total of 96 hours (days

+1 through +4) and for a total of 144 hours (days -1 through +4) in the phase II PD study group (Table 7.1). Prior to any study procedure, all patients gave written informed consent approved by IRB. Clinical and safety status of all patients will be subject to periodic review by all study sites and the sponsor for the purpose of ensuring that continuing accrual did not represent an unacceptable risk to the population under study.

All toxicities were graded by The National Cancer Institute Common Toxicity Criteria

(version 2.0) of the Cancer Therapy Evaluation Program.

7.3.3 Pharmacokinetic sampling and assay

Up to 45 patients were planned for this study. Blood samples were drawn from patients at the following time points; pretreatment, 2, 4, 6, 24, 48 and 72 hours on the first day during the infusion and post infusion at 96 hours (pilot PD group) and 144 hours

205 (phase II PD group) at 0.25, 0.5, 1, 2, 4, 24, 48 and 72 hours. At each time point, approximately 6 ml of peripheral blood was collected in heparized tubes and labeled with the unique patient registration number, collection time, date and source. Samples were shipped to the Leukemia Tissue Bank Core Facility. Drug concentration of GTI-2040 in plasma and in bone marrow (BM) cells was measured by a previously validated ELISA assay (56) (Figure 7.1).

7.3.4 Pharmacodynamic assessment

Fifteen patients each are planned to enroll in the pilot PD and phase II PD groups, however, we were able to obtain eight and four patients samples, respectively, for this analysis due to limited availability of BM cells. Approximately, 2 ml of Bone marrow

(BM) aspirates was collected into heparinized tubes. In the pilot PD group, samples were collected at pretreatment and 24 hours after Ara-C initiation (before GTI-2040). Samples at pretreatment and 48 hours after Ara-C dose (GTI-2040 pretreatment) were collected in the phase II PD group. 20×106 bone marrow cells at the indicated time points were procured. Intracellular dNTP/NTP (117) and Ara-CTP levels were extracted and quantified by the method developed in our laboratory.

7.3.5 Correlation of dNTP reduction and Ara-CTP accumulation

Following measurement and analysis of the data of the dNTP/NTP pools and Ara-

CTP accumulation from bone marrow samples obtained from patients enrolled in this phase II clinical trial, correlation studies was performed. Pearson Correlation was used to

206 analyze the pretreatment and post-treatment data with GTI-2040 or Ara-C alone and in combination with Ara-C in AML patients.

7.3.6 PK-PD modeling and simulations of GTI-2040 combined with Ara-C on reduction of dCTP and accumulation of Ara-CTP based on an indirect response model

The mechanism-based rationale for the combination treatment of GTI-2040 with

Aracytidine (Ara-C) indicated that the decrease of intracellular dNTP levels, especially dCTP level due to down-regulation of RRM2 mRNA and protein expression by GTI-

2040, could increase deoxycytidine kinase activities, resulting in an increase in intracellular Ara-CTP level. Therefore, we will use PK-PD modeling and simulation to establish a link between the decrease in dCTP levels and increase accumulation of Ara-

CTP in the cell. The phase II clinical pharmacokinetics profile of GTI-2040 in human plasma was characterized by a two-compartment model and pharmacodynamic (PD) endpoints; dCTP and Ara-CTP accumulation, (Pre- and Post GTI-2040 and Ara-C treatments) were measured in bone marrow (BM) cells. GTI-2040 needs to diffuse into the cells for its pharmacological activity, so a hypothetical effect compartment was used to link plasma PK profile of GTI-2040 and the PD endpoints in BM, which was evaluated by indirect response model (Figure 7.2).

Due to the bulky size, negative charge, and hydrophilic properties of GTI-2040, it is difficult for it to directly permeate through the cell membrane by passive diffusion.

However, in vitro experiments indicated that an active endocytosis mechanism may be involved in the uptake of phosphorothioate oligonucleotide in cultured cells (50, 51, 233-

207 235). Therefore, we assume that the cellular uptake of GTI-2040 is through an active transport system. The distribution of GTI-2040 from the central compartment to BM will be described by Michaelis-Menten enzyme kinetics with certain assumptions: (1)

Equilibrium between central compartment and peripheral compartment is quickly achieved. (2) Efflux of GTI-2040 from the effect compartment (BM) to the central compartment is not significant. Below is the differential equation:

dA1 Vm  R  K21* A2  (K12  K10  )* A1 Eq.1 dt 0 A1 Km  V1 dA2  K12* A1 K21* A2 Eq.2 dt dA3 Vm * A1   K30* A3 Eq.3 dt A1 Km  V1 Eq.3 E n d(dCTP) Emax C  Rin  Kdeg40  (1 n n ) dCTP Eq.4 dt EC 50  C dAra  CTP A(dCTP)  K syn 05*(1 ) Kdeg 50* Ara  CTP Eq.5 dt IC50  A(dCTP)

where, equations 1, 2 and 3 are used for the PK model fitting. A1, A2, A3 represent the amounts of GTI-2040 in the central, peripheral and effect compartments, respectively.

K12 and K21 are the intercompartmental transfer rate constants for the central and peripheral compartments, respectively. K10 and K30 are the elimination rate constants from the central and the effect compartments, respectively. V1 is the volume of central compartment, Vm the maximum uptake rate of GTI-2040 into cells and Km the substrate concentration that reaches the half maximal uptake rate. Equations 4 and 5 were used for

208 the PD model fitting, where dCTP is the normalized level at time t, Rin represents a zero- order rate constant of synthesis of dCTP, Kdeg40 is a first-order degradation rate constant of dCTP, Emax is the maximum reduced level of dCTP by GTI-2040, EC50 is the concentration of GTI-2040 required for half maximal reduction of dCTP, C is the effective concentration of GTI-2040 needed to deplete dCTP, and n is the sigmoidicity factor. Ara-CTP is normalized level at time t, Ksyn05 and Kdeg50 are the synthesis and degradation rate constant of Ara-CTP, respectively and IC50 is the concentration of dCTP change required to reach half maximal accumulation of Ara-CTP. In order to perform the simulation, all parameters were obtained from recent phase II pharmacokinetic and pharmacodynamic analysis and from previously published report of clinical pharmacokinetic evaluation of GTI-2040 and Ara-CTP in bone marrow samples from patients (54).

7.3.7 Data analysis

WinNonLin (Pharsight, Mountain View, CA) was used to fit the plasma concentration-time profiles, calculate the PK parameters and used for the PK/PD modeling. Data plots were performed by SigmaPlot computer software (version 11.0) and

Excel (Microsoft). Mann Whitney test of significance and correlation coefficient calculated using Minitab Statistical Software (version Minitab® 15.1.1.0) were used for treatment of the data.

209 7.4 Results

7.4.1 Plasma Pharmacokinetics of GTI-2040

Representative plasma concentration-time profiles of GTI-2040 for one patient each of the pilot and phase II PD groups receiving 5 mg/kg/day in are shown in Figure

7.3. In both of these two groups, the steady-state concentrations (Css) were achieved within 4 hours and remained so until the end of infusion. The plasma concentrations of

GTI-2040 in all the patients analyzed were measurable up to 48 hours post infusion followed by a bi-exponential decay (Figures 7.4). Therefore, the profiles were fitted with a two-compartment infusion model and the relevant PK parameters were computed as shown in Table 7.2 for the pilot and the phase II PD groups. There were no significant differences between the PK parameters of the pilot and the phase II PD groups, including

CL and Vss (Mann-Whitney test, p>0.05). Therefore, the parameters from these two groups were pooled (Table 7.3). When comparing PK parameters in patients treated before and after GTI-2040 for 24 hours, no difference was also found. Therefore, co- treatment of GTI-2040 with Ara-C did not seem to affect the PK parameters and thus co- administration of Ara-C may not affect the PK behavior of GTI-2040.

7.4.2 Intracellular uptake of GTI-2040 in bone marrow mononuclear cells

Intracellular concentrations (ICs) of GTI-2040 were measured in BM samples collected at pretreatment, 24 hours (D+1) and 48 hours (D+2) and (D+29), after GTI-

2040 treatment in samples of a total of 6 out of 12 patients and were undetected in the remaining 6 samples due to insufficient amount of cell samples available. The 24 hours

210 mean value of GTI-2040 level in the BM cells was 0.76 nM (n = 4) with a range of 0.01-

1.58 nM and the 48 hours mean value was 2.10 nM (n = 6) with a range of 0.12-7.78 nM.

ICs of GTI-2040 were not detectable in samples from day 29. Therefore, it appeared that

GTI-2040 accumulated in BM cells with time following drug infusion (Figure 7.5) and was eventually eliminated. The observed IC levels of GTI-2040 in BM between the pilot and the phase II PD groups did not show any statistically significant difference.

7.4.3 Pharmacodynamic results

To evaluate the effects of GTI-2040 on perturbation of the nucleotide pools, dNTP/NTP pools were measured in BM samples collected at pretreatment, 24 and 48 hours from the start of GTI-2040 infusion for both the pilot and the phase II PD groups.

There was a large variability in levels of intracellular dNTP and NTP pools among the samples (Figure 7.6A); however, there was a tendency towards decrease in the dCTP pool in the phase PD II group relative to the pilot PD group. To also determine the effect of Ara-C following pretreatment with GTI-2040, the Ara-CTP accumulation in BM samples was measured. There was no statistical difference between the Ara-CTP accumulation obtained from the two groups of patients pre-treated with GTI-2040 and after the initiation of Ara-C (Figure 7.6B).

7.4.4 Disease response

Clinical responses in patients in the two groups after receiving the treatments of

GTI-2040 combined with high dose of Ara-C are summarized in Table 7.4. In the pilot

211 PD group, complete remission (CR) was noted in 3 out of 10 patients. In the phase II PD group, 3 patients had CR and 1 patient had incomplete blood count with recovery (Cri)

7.4.5 Correlations among pharmacokinetics, pharmacodynamics and response

No pharmacokinetic parameters of GTI-2040, including AUC and Css values, were found to correlate with pharmacodynamic endpoints or disease response.

Intracellular GTI-2040 levels were found not to correlate with PD endpoints and disease response. Perturbation of dNTPs and NTPs levels were not found to correlate with accumulation of Ara-CTP.

7.4.6 PK-PD modeling and simulations of GTI-2040 combined with Ara-C

Previously, our laboratory reported the influence of the key parameter of GTI-

2040 targeting RRM2 mRNA by model simulation in the combination of GTI-2040 with

Ara-C (54). In this study, using one patient as an example, we demonstrate the model fitting of our data in order to simulate changes in dCTP and Ara-CTP levels as a function of time (Figures 7.7-7.9). The parameter estimates on patient 07028-10 are listed on

Table 7.5. Using the PK-PD model, the observed plasma PK for GTI-2040 and the PD values for GTI-2040 and Ara-C obtained from patients’ BM were initially used to fit and perform our model analysis. Based on our model, the observed PK and PD data and the

WinNonLin predicted values fell closely to each other. We observed the curves showing the decrease in dCTP levels as the rise in Ara-CTP accumulation occurred as a function of time (Figures 7.7 & 7.8). Their effect occurred fairly rapidly following drug treatment and remained stable for sometime before returning to baseline, when the infusion

212 stopped. Furthermore, as shown in Figure 7.9, PK-PD model simulation independent of the observed values (Table 7.6) was performed to further support our rationale. However, given that our model is based on PK and PD data points from one patient, the results from the PK-PD analysis should be interpreted with caution.

7.5 Discussion

In these clinical studies of patients enrolled in phase II trials of the combined therapy of GTI-2040 and high dose of Ara-C in refractory and relapsed AML patients, our data provide detailed information on the PK profile of GTI-2040 (5 mg/kg/day) administered by CIVI in AML patients with younger population (18-59 yr) of age. With the use of a highly sensitive ELISA-based hybridization method, plasma concentrations of GTI-2040 were measured and found to remain detectable up to 48 hours following termination of the infusion. Our ELISA allows for the effective capture of the true elimination phase with the t½β > 30 hour, which is consistent with the previous reported half-life in the phase I trial of this combination (27). This number differs from the previously reported half-life < 3 hr (236) and the discrepancy may be due to the limited assay sensitivities in the previous studies. The long terminal half-life of GTI-2040 may be due in part to its high protein binding (54). This high protein binding may also be responsible to a low renal clearance, which only accounts for 0.02% of the total clearance

(54). Therefore, the major clearance pathway of GTI-2040 was thought to be due to metabolism (54). Pharmacokinetics of GTI-2040 before and after treatment with Ara-C did not seem to affect its pharmacokinetic behavior and their PK parameters showed no statistical difference. Plasma pharmacokinetics of GTI-2040 was also found not to

213 correlate with disease response. This demonstrates the importance of monitoring the effective drug concentration at the sites of drug action, such as in the bone marrow cells where GTI-2040 was found to be taken up. Using this highly sensitive method, it was also possible to monitor the intracellular concentrations of GTI-2040 in samples of bone marrow from AML patients, which is valuable in correlation with the drug effect. The longer residence time of GTI-2040 in vivo also favors its cellular uptake, which can be demonstrated by the higher ICs of GTI-2040 in patients with extended IV infusion.

However, the ICs of GTI-2040 in BM cells did not appear to correlate well with the PD endpoints and disease response, at least not in this patient population.

Using our newly developed method for the quantification of dNTP/NTP pools, we were able to measure for the first time the levels of these surrogate endpoints in bone marrow cells of patients with AML in this trial to assess the effect of pretreatment with

GTI-2040 or Ara-C alone then in combination with Ara-C. The dNTP/NTPs levels were detectable in all the patients’ samples measured, and up to day 29 in a few of the patients.

There was an enormous variability in levels of dNTPs and NTPs among patients’ samples and this might be due to dissimilar disease conditions of the patients, initial baseline levels of RRM2 expression and the uptake amount of GTI-2040 following infusion. The average of the dCTP levels showed a tendency towards decrease after GTI-2040 treatment (phase II PD group), as compared to the pilot group (delayed GTI-2040).

Furthermore, we were successfully able to measure the Ara-CTP levels in patients’ samples acquired in the pilot and phase II PD groups. In comparing the Ara-CTP accumulation between the pilot and the phase II PD groups, we did not, however, detect a

214 significant increase in the Ara-CTP accumulation in the phase II PD group as compared the pilot PD group. This might be attributed to the lack of optimal sustained effect from

GTI-2040 treatment in the phase II group. In addition, perturbation of dNTPs did not appear to correlate well with Ara-CTP accumulation contrary to the in vitro experiments.

This might be attributed to the lack of controlled environment associated with clinical trial sample collection, variable disease conditions of patients, time and the technique of storage before analysis.

Because the number of patients and tissue samples assessed thus far in this trial was small, all of the correlation results are preliminary in nature. This also underscores the complex nature of the clinical evaluation of the antisense activity of GTI-2040. The manifestation of an antisense effect may rely on many factors, including cellular drug uptake and distribution, endogenous levels of endonucleases (RNase H), and positive response of the leukemia cells to target down regulation. An integrated PK-PD model with indirect response was built and applied to evaluate the dynamics of dCTP reduction and Ara-CTP accumulation following treatment with GTI-2040 and Ara-C. The purpose of this PK-PD model simulation study was to evaluate the effects of perturbation of dCTP by GTI-2040 and the enhanced accumulation of Ara-CTP. In our preliminary PK-PD model, we observed a sustained reduction in the dCTP levels following the initiation of

GTI-2040; Ara-CTP was subsequently accumulated and remained steady until the end of infusion. Although, our model simulation demonstrates perturbation of dCTP in bone marrow cells resulting in enhanced accumulation of Ara-CTP, it should be interpreted with discretion.

215 7.6 Conclusion

In summary, evaluation of the pharmacokinetics of GTI-2040 before and after treatment with Ara-C is needed to eliminate the effect of drug-drug interaction. The intracellular concentrations of GTI-2040 and its subcellular distribution patterns are important in evaluation of antisense effect and disease response in patients. Examination of the surrogate endpoints of GTI-2040 is necessary for target validation of decrease in

RRM2 level and subsequent dNTP pools reduction by GTI-2040 and provides proof of

Ara-C enhanced DNA incorporation for irreversible cell death to occur. PK-PD analysis demonstrated target alterations after drug perturbation and their correlations with dNTP/NTPs, Ara-CTP and response.

216

Study design GTI-2040 dose Aracytidine dose CIVI 2h IV q12h Pilot 5 mg/kg/day 3 g/m2/dose (96 h CIVI) Days +1 to +4 Days 0, +1, +2 and +4 Delayed GTI-2040 for 24 h

Phase II 5 mg/kg/day 3 g/m2/dose (144 h CIVI) Days -1 to +4 Days 0, +1, +2 and +4 Early GTI-2040 for 24 h

Demographic Characteristics of Patients

Gender M F N

25 19 44

Age 18 – 59 yrs

Body weight Median 83.18 ± 19.9 kg (range) (55.4 to 121.9 kg)

Table 7.1 Dose schedule treatment plan and demographic characteristics for GTI-2040 combined with high dose Aracytidine in AML patients.

217

Parameters Pilot Phase II n = 6 n = 10 Css (nmol/L) 228.4 ± 137.5 207.4 ± 118.4 AUC0-∞(µmol/L x h) 23.3 ± 13.0 28.5 ± 15.9 CL (L/h) 4.5 ± 3.5 3.2 ± 2.4 Vss (L) 5.3 ± 3.9 4.3 ± 3.1 T½α (h) 0.57 ± 0.38 0.59 ± 0.25 T½β (h) 33.8 ± 17.1 39.8 ± 21.0

Table 7.2 Relevant PK parameters of GTI-2040 in AML patients in pilot PD and phase II PD groups giving GTI-2040 with 5 mg/kg/day.

218

Parameters Pilot and Phase II groups n = 16

Css (nmol/L) 215.3 ± 121.8 AUC0-∞(µmol/L x h) 26.58 ± 14.65 CL (L/h) 3.96 ± 3.12 Vss (L) 3.73 ± 3.43

T½α (h) 0.58 ± 0.29

T½β (h) 37.6 ± 19.3

Table 7.3 Pooled PK parameters. There was no statistical difference in the pharmacokinetic parameters between the pilot and the phase II PD groups (p>0.05, Mann Whitney test).

219

Response Pilot and Phase II groups n = 25

CRs 6 Cri 1 PD 18

Table 7.4 Clinical response in patients treated with GTI-2040 combined with high dose Ara-C.

220 Kdeg50 A B

Time Concentration Model Time Obs. dCTP Model (h) (nM) Estimate (h) level Estimate 0 0 0 0 1.54 100 2 563.2 358.6 2 47.78 4 837.4 418.1 4 38.71 6 1173.7 428.0 6 34.32 24 780.6 430.3 24 0.48 27.78 48 593.5 430.6 48 27.71 72 637.8 430.9 72 27.71 144 438.5 431.4 144 27.71 144.25 189.9 345.0 144.25 27.77 144.5 312.1 275.9 144.5 28.10 145 175.3 176.6 145 29.50 146 61.15 72.87 146 33.48 13.39 148 28.77 148 40.36 168 1.50 1.305 168 51.39 192 0.85 1.044 51.47 192 216 1.02 0.835 216 1.54 51.47

C Time Obs Ara-CTP Model (h) level Estimate 0 0 0 2 41.02 4 42.78 6 44.78 24 49.85 50.12 48 50.21 72 50.21 144 50.21 144.25 50.21 50.15 144.5 145 49.84 146 48.66

148 45.79 168 39.23 192 39.17 216 39.17

Table 7.5 Example of observed and the PK-PD model predicted (A) GTI-2040 plasma concentrations (B) dCTP depletion (C) Ara-CTP accumulation with time in patient 07028-10.

221

Time GTI-2040 plasma dCTP (h) Simulation Simulation Ara-CTP Simulation 0 0 50 0 2 356.7 21.52 19.85 4 415.2 19.98 21.11 6 424.7 15.83 26.02 24 427.0 4.850 69.39 48 427.5 4.693 71.08 72 427.8 4.692 71.09 144 428.4 4.690 71.11 144.25 342.0 4.691 71.11 144.5 273.2 4.693 71.09 145 174.4 4.710 70.92 146 71.67 4.863 69.40 148 13.26 6.355 57.07 168 1.481 31.61 13.82 192 1.138 32.63 13.41 216 0.876 33.14 13.21

Table 7.6 Simulated PK-PD model predicted GTI-2040 plasma concentrations, dCTP reduction and Ara-CTP accumulation with time in patient 07028-10.

222 5’ 3’ Analyte antisense 3’ GTG ATC AAT 5’ B 5’ 3’ Capture probe with 5’ overhang 3’ chain-shorten metabolites

1) hybridization 2) Apply to avidin-coated plate

3’ GTG ATC AAT 5’ D B P 5’CAC TAG TTA3’ 3’ GTG ATC AAT 5’ B 3)T4 ligase, ATP GTG ATC AAT 5’ B 3’ Ap D E-Ab 3’ Ap P CAC TAG TTA D 4)S1 nuclease B GTG ATC AAT 5’ D 5) wash B 3’ GTG ATC AAT 5’ P 5’CAC TAG TTA3’ B

Figure 7.1 Illustration of a two-step ELISA assay in determination of GTI-2040 in plasma and bone marrow cells (56).

223

Effect CPT 3

n n (1+Emax x C /EC50 + C ) Rin Inhibition Emax model Kdeg40 dCTP

(1-A(dCTP)/(IC50 + A(dCTP)) Stimulation

Ara-CTP K K syn05 deg50

Figure 7.2 A two compartment model for pharmacokinetic analysis of GTI-2040 and a simplified PK-PD model of GTI-2040 and Ara-C in AML patients.

224

Figure 7.3 Representative WinNonLin data plots for one patient each for pilot and phase II PD concentration-time profile of GTI-2040 in patients with AML after treatment with 5 mg/kg/day.

225

A

10000

1000

100

10

Plasma GTI-2040 (nmol/L) GTI-2040 Plasma 1

0.1 0 20 40 60 80 100 120 140 160 180 Time (h)

B

10000

1000

100

10

Plasma GTI-2040 (nmol/L) GTI-2040 Plasma 1

0.1 0 50 100 150 200 250 Time (h)

Figure 7.4 (A) Pilot PD and phase II PD group concentration-time profiles (B) of GTI- 2040 in patients with AML after treatment with 5 mg/kg/day drug as a 96 and 144 hours, respectively, continuous infusion using a two compartment infusion model in WinNonLin.

226 9 8 ** Pretreatment Day 1 7 Day 2 6

5

4

3

2 ICs of ICs GTI-2040 (nM)in BM ** 1

0 11 22 10 14 15 21

Pilot PD Phase II PD

Figure 7.5 Examples of GTI-2040 intracellular levels in bone marrow cells of AML patients for day 1 and 2 following treatment with GTI-2040. *p<0.05, **p<0.01 compared to day 1 for phase II PD group.

227

Figure 7.6 Mean intracellular dNTPs/NTPs levels (A) and mean Ara-CTP (B) levels in bone marrow cells of AML patients in pilot and phase II PD groups. (p>0.05)

228 10000

Observed plasma conc. Predicted plasma conc. 1000

100

10

1

GTI-2040 concentration in plasma (nM)

0.1 0 50 100 150 200 250

Time (hr)

Figure 7.7 The fitted and observed GTI-2040 plasma concentration-time data from Patient 07028-10 using a two-compartment intravenous infusion model.

229

Figure 7.8 Composite plots of fitted and observed GTI-2040 concentration-time profiles with % changes in PK-PD model simulated and observed dCTP and Ara-CTP levels for Patient 07028-10.

230

Figure 7.9 Composite plots of simulated GTI-2040 concentration-time profiles with % changes in PK-PD model simulated dCTP depletion and Ara-CTP accumulation.

231 CHAPTER 8

CONCLUSIONS AND PERSPECTIVES

In this dissertation, pharmacodynamics of nucleoside analogs in combination therapy with ribonucleotide reductase (RR) antisense GTI-2040 was investigated. Both in vitro and in vivo experiments were performed to study the novel combination treatment of antisense GTI-2040 with 5-azaC. Most notably, we discovered for the first time the modulation effect of 5-azaC on ribonucleotide reductase, which serves as a novel target for this azanucleoside. In addition, the biomarkers involved in the development of resistance to 5-azaC and DAC in cancer cell lines were assessed which help to elucidate the potential mechanisms that contribute to the induction of resistance. Finally, we conducted in-vitro and in-vivo pharmacodynamic analysis of GTI-2040 in combination with Ara-C to provide experimental supports of the already completed phase I clinical combination study, and further evaluate their clinical pharmacokinetics in a phase II clinical trial which is about to complete. All these studies will provide a in depth understanding of this combination which will greatly benefit future planning of combination treatment of patients with refractory or relapsed AML.

5-Azacytidine (5-AzaC), an azanucleoside synthesized over 45 years ago, has shown a wide range of antitumor activities and has recently been approved by the US

232 FDA for the treatment of myelodysplastic syndrome (MDS). 5-AzaC was approved as a hypomethylating agent via its DNA incorporation pathway, which only represents 10-

20% of its total utilization. Nevertheless, it has continually been utilized in clinical trials for the treatment of myeloid leukemia and other forms of neoplasia. Although, 5-azaC has produced remission or clinical improvements in more than half of the patients treated at low doses (237, 238), its pharmacologic effects has yet to be fully exploited. Inside the cell, 5-azaC is phosphorylated to its active anabolite, 5-Aza-CTP, before incorporation into RNA, which represents 80% of its total utilization, presumably leading to the inhibition of protein synthesis. However, the detailed mechanism of its RNA effect is much less clear. In order to better understand its RNA effect, we explored the use of a pharmacologic agent to block its DNA pathway in order to evaluate and maximize the

RNA-mediating effect of 5-azaC.

In Chapter 2, we investigated the use of a specific RR mRNA antisense GTI-2040 to maximize the RNA effect of 5-azaC, allowing investigation of its biochemical mechanism. Our data showed that the combination of GTI-2040 with 5-azaC showed synergistic cytotoxicity in human leukemia cell lines K562 and MV4-11.

Mechanistically, GTI-2040 alone inhibits RRM2 expression levels and subsequently resulted in a reduction in the dATP, dTTP and dCTP levels. Unexpectedly, the down- regulation of RRM2 expression levels by 5-azaC was also observed for the first time which resulted in a series of experiments that was described in Chapter 4. Furthermore, the reduction in RRM2 expression levels by GTI-2040 was found to result in an opposing effect to the hypomethylating effect of 5-azaC, since its conversion to DAC-TP is blocked, subsequently there was no reduction in DNMT1 protein expression, not

233 counting the RRM2 inhibition effect of 5-azaC itself (Chapter 4). Finally, the combination of 5-azaC and GTI-2040 may enhance the RNA mediated effects of 5-azaC that will result in mRNA destabilization and protein synthesis inhibition. Therefore, the novel combination of 5-azaC and GTI-2040 is feasible and may potentially be exploited in the clinic. In order to better understand and effectively utilize the combination of GTI-

2040 and 5-azaC in the clinic, it is imperative that in addition to the preclinical studies,

PK-PD modeling and simulation be extensively explored to provide more insight into the use of these drugs.

In Chapter 3, we established an in-vivo tumor growth profile and pharmacodynamic based model to further evaluate the antitumor effect of the combination treatment of GTI-2040 with 5-azaC. The combination of 5-azaC and GTI-

2040 reduced tumor growth volume and weight in MV4-11 xenograft tumor models and our result also confirmed the down-regulation of RRM2 mRNA and protein levels in vivo from the excised tumor tissues. Our established pharmacodynamic tumor growth model may be used prospectively for an educated design of in vivo experiments, while providing valuable tools to evaluate dose regimen design for preclinical evaluation. A more realistic safety margin may be obtained for future preclinical and clinical studies with other combination treatments using this tumor model design.

Following our observation of the reduction in RRM2 expression by 5-azaC, a succession of experiments was performed to determine the mechanism involved in this process. Based on our data, we confirmed RRM2 as a novel target for 5-azaC. In this study we reported the discovery and affirmation of RRM2 as a novel post-transcriptional

234 target of 5-azaC. 5-AzaC dramatically decreases RRM2 expression, resulting in negative regulation of the deoxyribonucleotide and nucleotide (dNTP/NTP) pools in cell lines and tumor xenografts model. Most notably, RRM2 down-regulation also occurred ex vivo in human bone marrow cells treated with 5-azaC, which might be through the direct incorporation of 5-azaC into RNA, resulting in mRNA destabilization and shortened half- life. Evidence of 5-azaC incorporation into RNA was also found. In addition, the involvement of 5-azaC into RNA was used to establish a link between RNA interference and certain miRNA perturbation which provide evidence for endogenous miRNA involvement in transcriptional gene silencing. Our data may create a new area of target investigation for 5-azaC and will provide an in-depth understanding of the molecular pharmacology of the drug, while opening up further biomarker studies. Our study may also represent an important expansion for combination therapy involving RRM2 inhibitors and other drugs. Previous studies have established the feasibility of biochemical modulation of aracytidine triphosphate (Ara-CTP) accumulation in AML blasts by combination with nucleoside analog (180-182). Other reports (27) and our laboratory also demonstrated in a phase I/II trial that, down-regulation of RRM2 by GTI-

2040 when combined with high-dose Ara-C, resulted in increased accumulation of Ara-

CTP. As such, RRM2 as a novel target of 5-azaC will serve to develop the feasibility of more effective combination therapy for the treatment of cancer and contribute to the improved clinical utility of this drug.

5-AzaC and DAC have been used as potent anticancer agents for the treatment of hematopoietic neoplasms and solid tumors (141-145). Despite their recent approval to be

235 used for the treatment of myelodysplastic syndrome (MDS) due to their hypomethylating effects, resistance to these drugs remains a major problem. To better understand the cellular mechanisms involved in the resistance, we developed cellular models of 5-azaC and DAC resistances in a panel of six cell lines and attempted to identify biomarkers involved in the resistance. Our study investigated the induction of resistance to 5-azaC and DAC in vitro in several leukemia and solid tumor cell lines by continuous exposure to 5-azaC or DAC. Our data showed that various processes, such as hypermethylation, low accumulation of the intracellular triphosphate forms of the drugs and cellular perturbation of the dNTP and NTP pools, contribute to the induction of resistance in cells following graded exposures to 5-azaC or DAC. Our findings will further facilitate the understanding of the therapeutic efficacy of 5-azaC and DAC by introducing a new era of biomarker driven research for seeking appropriate combination with other compounds.

Ultimately, the development of these molecular biomarkers for 5-azaC and DAC induced resistances will allow for a better understanding of 5-azaC and DAC modes of action and allow for rational and efficient evaluation of 5-azaC and DAC for clinical implication.

In Chapter 6, we discussed the biochemical modulation effects of Ara-C by GTI-

2040 in K562 human leukemia cells. In general, antisense oligonucleotides are introduced into cells by transfecting agents, which themselves have potentially unknown biological effects. In order to alleviate this complication, GTI-2040 was delivered into the cell through electroporation. Following GTI-2040 delivery, down-regulation of RRM2 mRNA and protein were observed, which subsequently lead to a decrease of dNTP pool.

These effects provided an additional experimental verification of the proposed mechanism of GTI-2040 in leukemia cells (32). As expected, down-regulation of RRM2

236 mRNA and protein levels by GTI-2040 was found to have a positive correlation with the reduction in intracellular dNTP pools following treatment with GTI-2040 and Ara-C. We determined the Ara-CTP level following exposure to Ara-C and our data showed a dose dependent accumulation of Ara-CTP. As previously reported in our laboratory, pretreatment of K562 cells with GTI-2040, followed by Ara-C increased the accumulation of Ara-CTP in a dose dependent manner, hence increasing in the rate of apoptosis. Our studies, therefore provides a laboratory and mechanistic justification for the phase I/II evaluation of GTI-2040 in combination with Ara-C in patients with acute myeloid leukemia.

Finally, Chapter 7 discussed the clinical evaluation of patients enrolled in a phase

II trial of the combined therapy of GTI-2040 and high dose of Ara-C in refractory and relapsed AML patients. Our study provides detailed information on the pharmacokinetic profile of GTI-2040 (5 mg/kg/day) administered by CIVI in AML patients (18-59 yr).

With the assistance of a highly sensitive ELISA-based hybridization method, cellular

GTI-2040 levels and plasma concentrations of GTI-2040 were measured and remained detectable up to 48 hours following termination of the infusion. Consequently, the pharmacokinetics of GTI-2040 before and after treatment with Ara-C is needed to eliminate the effect of drug-drug interaction. To understand the pharmacodynamic effects of GTI-2040, the intracellular concentrations of GTI-2040 was assessed, which are important in the evaluation of the antisense effect and disease response in patients.

Examination of surrogate endpoints of GTI-2040 (dNTP/NTP pools) and Ara-C (Ara-

CTP) accumulation, when the two drugs were used in combination, is necessary for target

237 validation of RRM2 and can provide proof of increased DNA incorporation of Ara-C for irreversible cell death to occur in cells.

Our PK-PD analysis demonstrated target alterations after drug perturbation.

Through correlation analysis, the intracellular dNTP/NTP levels and Ara-CTP accumulation in bone marrow cells was evaluated in relation to clinical response and this has provided valuable insight that will improve the therapeutic effect of antisense and nucleoside drugs in combination therapy. In our PK-PD model we observed a sustained reduction in the dCTP levels following the initiation of GTI-2040 followed by subsequent accumulation of Ara-CTP, which remained steady until the end of infusion. Our model simulation, therefore, demonstrated efficient perturbation of dCTP in bone marrow cells, resulting in enhanced accumulation of Ara-CTP levels, which further support our rationale for the combination of GTI-2040 and Ara-C for the treatment of AML patients.

In summary, investigation into the modulation of pharmacologic effects of 5-

AzaC, DAC and Ara-CTP by ribonucleotide reductase GTI-2040 alone and in combination provided valuable insights into their present and future clinical application.

The biomarker involved in the mechanism of nucleoside analog resistances was also determined. PK-PD modeling and simulation of combination therapies of GTI-2040 and

Ara-C were investigated. More importantly, our studies have launched a translational research platform to investigate the potential of using 5-azaC as a RRM2 inhibitor in the clinic alone or in combination with other drugs.

238

BIBLIOGRAPHY

1. Parker WB. Enzymology of purine and pyrimidine antimetabolites used in the treatment of cancer. Chemical reviews 2009;109(7):2880-93. 2. Tabata S, Tanaka M, Endo Y, Obata T, Matsuda A, Sasaki T. Anti-tumor mechanisms of 3'-ethynyluridine and 3'-ethynylcytidine as RNA synthesis inhibitors: development and characterization of 3'-ethynyluridine-resistant cells. Cancer letters 1997;116(2):225-31. 3. Lowenberg B, Griffin JD, Tallman MS. Acute myeloid leukemia and acute promyelocytic leukemia. Hematology / the Education Program of the American Society of Hematology American Society of Hematology 2003:82-101. 4. Das PM, Singal R. DNA methylation and cancer. J Clin Oncol 2004;22(22):4632- 42. 5. Faderl S, Gandhi V, Kantarjian HM. Potential role of novel nucleoside analogs in the treatment of acute myeloid leukemia. Current opinion in hematology 2008;15(2):101- 7. 6. Agrawal S, Kandimalla ER. Antisense and/or immunostimulatory oligonucleotide therapeutics. Current cancer drug targets 2001;1(3):197-209. 7. Dias N, Stein CA. Antisense oligonucleotides: basic concepts and mechanisms. Molecular cancer therapeutics 2002;1(5):347-55. 8. Bennett CF. Efficiency of antisense oligonucleotide drug discovery. Antisense & nucleic acid drug development 2002;12(3):215-24. 9. Kuykendall JR. 5-azacytidine and decitabine monotherapies of myelodysplastic disorders. The Annals of pharmacotherapy 2005;39(10):1700-9. 10. Kiziltepe T, Hideshima T, Catley L, et al. 5-Azacytidine, a DNA methyltransferase inhibitor, induces ATR-mediated DNA double-strand break responses, apoptosis, and synergistic cytotoxicity with doxorubicin and bortezomib against multiple myeloma cells. Molecular cancer therapeutics 2007;6(6):1718-27. 11. Li LH, Olin EJ, Buskirk HH, Reineke LM. Cytotoxicity and mode of action of 5- azacytidine on L1210 leukemia. Cancer research 1970;30(11):2760-9. 12. Michalowsky LA, Jones PA. Differential nuclear protein binding to 5- azacytosine-containing DNA as a potential mechanism for 5-aza-2'-deoxycytidine resistance. Molecular and cellular biology 1987;7(9):3076-83. 13. Christman JK. 5-Azacytidine and 5-aza-2'-deoxycytidine as inhibitors of DNA methylation: mechanistic studies and their implications for cancer therapy. Oncogene 2002;21(35):5483-95. 14. Schrump DS, Fischette MR, Nguyen DM, et al. Phase I study of decitabine- mediated gene expression in patients with cancers involving the lungs, esophagus, or pleura. Clin Cancer Res 2006;12(19):5777-85. 15. Momparler RL. Molecular, cellular and animal pharmacology of 5-aza-2'- deoxycytidine. Pharmacology & therapeutics 1985;30(3):287-99.

239 16. Rivard GE, Momparler RL, Demers J, et al. Phase I study on 5-aza-2'- deoxycytidine in children with acute leukemia. Leukemia research 1981;5(6):453-62. 17. Glover AB, Leyland-Jones B. Biochemistry of azacitidine: a review. Cancer treatment reports 1987;71(10):959-64. 18. Nicoloso MS, Kipps TJ, Croce CM, Calin GA. MicroRNAs in the pathogeny of chronic lymphocytic leukaemia. British journal of haematology 2007;139(5):709-16. 19. Rush LJ, Dai Z, Smiraglia DJ, et al. Novel methylation targets in de novo acute myeloid leukemia with prevalence of chromosome 11 loci. Blood 2001;97(10):3226-33. 20. Estey EH. Therapeutic options for acute myelogenous leukemia. Cancer 2001;92(5):1059-73. 21. Ravindranath Y. Recent advances in pediatric acute lymphoblastic and myeloid leukemia. Current opinion in oncology 2003;15(1):23-35. 22. Creasey WA, Papac RJ, Markiw ME, Calabresi P, Welch AD. Biochemical and pharmacological studies with 1-beta-D-arabinofuranosylcytosine in man. Biochemical pharmacology 1966;15(10):1417-28. 23. Shao J, Zhou B, Chu B, Yen Y. Ribonucleotide reductase inhibitors and future drug design. Current cancer drug targets 2006;6(5):409-31. 24. Huang A, Fan H, Taylor WR, Wright JA. Ribonucleotide reductase R2 gene expression and changes in drug sensitivity and genome stability. Cancer research 1997;57(21):4876-81. 25. Madlener S, Illmer C, Horvath Z, et al. Gallic acid inhibits ribonucleotide reductase and cyclooxygenases in human HL-60 promyelocytic leukemia cells. Cancer letters 2007;245(1-2):156-62. 26. Zhou BS, Tsai P, Ker R, et al. Overexpression of transfected human ribonucleotide reductase M2 subunit in human cancer cells enhances their invasive potential. Clinical & experimental metastasis 1998;16(1):43-9. 27. Klisovic RB, Blum W, Wei X, et al. Phase I study of GTI-2040, an antisense to ribonucleotide reductase, in combination with high-dose cytarabine in patients with acute myeloid leukemia. Clin Cancer Res 2008;14(12):3889-95. 28. Tsimberidou AM, Alvarado Y, Giles FJ. Evolving role of ribonucleoside reductase inhibitors in hematologic malignancies. Expert review of anticancer therapy 2002;2(4):437-48. 29. Weber G. Biochemical strategy of cancer cells and the design of chemotherapy: G. H. A. Clowes Memorial Lecture. Cancer research 1983;43(8):3466-92. 30. Amara FM, Chen FY, Wright JA. A novel transforming growth factor-beta 1 responsive cytoplasmic trans-acting factor binds selectively to the 3'-untranslated region of mammalian ribonucleotide reductase R2 mRNA: role in message stability. Nucleic acids research 1993;21(20):4803-9. 31. Desai AA, Schilsky RL, Young A, et al. A phase I study of antisense oligonucleotide GTI-2040 given by continuous intravenous infusion in patients with advanced solid tumors. Ann Oncol 2005;16(6):958-65. 32. Lee Y, Vassilakos A, Feng N, et al. GTI-2040, an antisense agent targeting the small subunit component (R2) of human ribonucleotide reductase, shows potent antitumor activity against a variety of tumors. Cancer Res 2003;63(11):2802-11.

240 33. Nakano K, Balint E, Ashcroft M, Vousden KH. A ribonucleotide reductase gene is a transcriptional target of p53 and p73. Oncogene 2000;19(37):4283-9. 34. Tanaka H, Arakawa H, Yamaguchi T, et al. A ribonucleotide reductase gene involved in a p53-dependent cell-cycle checkpoint for DNA damage. Nature 2000;404(6773):42-9. 35. Link PA, Baer MR, James SR, Jones DA, Karpf AR. p53-inducible ribonucleotide reductase (p53R2/RRM2B) is a DNA hypomethylation-independent decitabine gene target that correlates with clinical response in myelodysplastic syndrome/acute myelogenous leukemia. Cancer research 2008;68(22):9358-66. 36. Wallace DC. A mitochondrial paradigm of metabolic and degenerative diseases, aging, and cancer: a dawn for evolutionary medicine. Annual review of genetics 2005;39:359-407. 37. Bianchi V, Borella S, Calderazzo F, Ferraro P, Chieco Bianchi L, Reichard P. Inhibition of ribonucleotide reductase by 2'-substituted deoxycytidine analogs: possible application in AIDS treatment. Proceedings of the National Academy of Sciences of the United States of America 1994;91(18):8403-7. 38. Smith BD, Karp JE. Ribonucleotide reductase: an old target with new potential. Leukemia research 2003;27(12):1075-6. 39. Chimploy K, Tassotto ML, Mathews CK. Ribonucleotide reductase, a possible agent in deoxyribonucleotide pool asymmetries induced by hypoxia. The Journal of biological chemistry 2000;275(50):39267-71. 40. Koc A, Mathews CK, Wheeler LJ, Gross MK, Merrill GF. Thioredoxin is required for deoxyribonucleotide pool maintenance during S phase. The Journal of biological chemistry 2006;281(22):15058-63. 41. Jordheim LP, Cros E, Gouy MH, et al. Characterization of a gemcitabine-resistant murine leukemic cell line: reversion of in vitro resistance by a mononucleotide prodrug. Clin Cancer Res 2004;10(16):5614-21. 42. Avolio TM, Lee Y, Feng N, et al. RNA interference targeting the R2 subunit of ribonucleotide reductase inhibits growth of tumor cells in vitro and in vivo. Anti-cancer drugs 2007;18(4):377-88. 43. Davis S, Lollo B, Freier S, Esau C. Improved targeting of miRNA with antisense oligonucleotides. Nucleic acids research 2006;34(8):2294-304. 44. Ewald B, Sampath D, Plunkett W. Nucleoside analogs: molecular mechanisms signaling cell death. Oncogene 2008;27(50):6522-37. 45. Saito Y, Liang G, Egger G, et al. Specific activation of microRNA-127 with downregulation of the proto-oncogene BCL6 by chromatin-modifying drugs in human cancer cells. Cancer cell 2006;9(6):435-43. 46. Saetrom P, Snove O, Jr., Rossi JJ. Epigenetics and microRNAs. Pediatric research 2007;61(5 Pt 2):17R-23R. 47. Venturini L, Eder M, Scherr M. RNA-Mediated Gene Silencing in Hematopoietic Cells. Journal of biomedicine & biotechnology 2006;2006(4):87340. 48. Dean NM, Bennett CF. Antisense oligonucleotide-based therapeutics for cancer. Oncogene 2003;22(56):9087-96. 49. Kamiya H, Akita H, Harashima H. Pharmacokinetic and pharmacodynamic considerations in gene therapy. Drug discovery today 2003;8(21):990-6.

241 50. Loke SL, Stein CA, Zhang XH, et al. Characterization of oligonucleotide transport into living cells. Proceedings of the National Academy of Sciences of the United States of America 1989;86(10):3474-8. 51. Yakubov LA, Deeva EA, Zarytova VF, et al. Mechanism of oligonucleotide uptake by cells: involvement of specific receptors? Proceedings of the National Academy of Sciences of the United States of America 1989;86(17):6454-8. 52. Bergan R, Connell Y, Fahmy B, Neckers L. Electroporation enhances c-myc antisense oligodeoxynucleotide efficacy. Nucleic acids research 1993;21(15):3567-73. 53. Liu Y, Bergan R. Improved intracellular delivery of oligonucleotides by square wave electroporation. Antisense & nucleic acid drug development 2001;11(1):7-14. 54. Wei X. Pharmacokinetics, pharmacodynamics and metabolism of GTI-2040, a phosphorothioate oligonucleotide targeting R2 subunit of ribonucleotide reductase. Columbus, Ohio: Ohio State University; 2006. p. xxiii, 308 p. 55. Chabes A, Thelander L. Controlled protein degradation regulates ribonucleotide reductase activity in proliferating mammalian cells during the normal cell cycle and in response to DNA damage and replication blocks. The Journal of biological chemistry 2000;275(23):17747-53. 56. Wei X, Dai G, Marcucci G, et al. A specific picomolar hybridization-based ELISA assay for the determination of phosphorothioate oligonucleotides in plasma and cellular matrices. Pharmaceutical research 2006;23(6):1251-64. 57. Rubin EH, Cory JG. Differential turnover of the subunits of ribonucleotide reductase in synchronized leukemia L1210 cells. Cancer research 1986;46(12 Pt 1):6165- 8. 58. Cros E, Jordheim L, Dumontet C, Galmarini CM. Problems related to resistance to cytarabine in acute myeloid leukemia. Leukemia & lymphoma 2004;45(6):1123-32. 59. Galmarini CM, Mackey JR, Dumontet C. Nucleoside analogues: mechanisms of drug resistance and reversal strategies. Leukemia 2001;15(6):875-90. 60. Hubeek I, Peters GJ, Broekhuizen AJ, et al. Potentiation of in vitro ara-C cytotoxicity by ribonucleotide reductase inhibitors, cyclin-dependent kinase modulators and the DNA repair inhibitor aphidicolin in paediatric acute myeloid leukaemia. British journal of haematology 2005;131(2):219-22. 61. Grant S. Ara-C: cellular and molecular pharmacology. Advances in cancer research 1998;72:197-233. 62. Hamada A, Kawaguchi T, Nakano M. Clinical pharmacokinetics of cytarabine formulations. Clinical pharmacokinetics 2002;41(10):705-18. 63. Stone RM. Treatment of acute myeloid leukemia: state-of-the-art and future directions. Seminars in hematology 2002;39(3 Suppl 2):4-10. 64. Liu S, Shen T, Huynh L, et al. Interplay of RUNX1/MTG8 and DNA methyltransferase 1 in acute myeloid leukemia. Cancer research 2005;65(4):1277-84. 65. Meyers S, Downing JR, Hiebert SW. Identification of AML-1 and the (8;21) translocation protein (AML-1/ETO) as sequence-specific DNA-binding proteins: the runt homology domain is required for DNA binding and protein-protein interactions. Molecular and cellular biology 1993;13(10):6336-45.

242 66. Glassman AB, Hayes KJ. Translocation (11;16)(q23;p13) acute myelogenous leukemia and myelodysplastic syndrome. Annals of clinical and laboratory science 2003;33(3):285-8. 67. Stam RW, den Boer ML, Passier MM, et al. Silencing of the tumor suppressor gene FHIT is highly characteristic for MLL gene rearranged infant acute lymphoblastic leukemia. Leukemia 2006;20(2):264-71. 68. Esteller M. CpG island hypermethylation and tumor suppressor genes: a booming present, a brighter future. Oncogene 2002;21(35):5427-40. 69. Peedicayil J. Epigenetic therapy--a new development in pharmacology. The Indian journal of medical research 2006;123(1):17-24. 70. Toyota M, Kopecky KJ, Toyota MO, Jair KW, Willman CL, Issa JP. Methylation profiling in acute myeloid leukemia. Blood 2001;97(9):2823-9. 71. Esteller M, Sanchez-Cespedes M, Rosell R, Sidransky D, Baylin SB, Herman JG. Detection of aberrant promoter hypermethylation of tumor suppressor genes in serum DNA from non-small cell lung cancer patients. Cancer research 1999;59(1):67-70. 72. Sakashita K, Koike K, Kinoshita T, et al. Dynamic DNA methylation change in the CpG island region of p15 during human myeloid development. The Journal of clinical investigation 2001;108(8):1195-204. 73. Wolffe AP, Matzke MA. Epigenetics: regulation through repression. Science (New York, NY 1999;286(5439):481-6. 74. Kimby E, Nygren P, Glimelius B. A systematic overview of chemotherapy effects in acute myeloid leukaemia. Acta oncologica (Stockholm, Sweden) 2001;40(2-3):231-52. 75. Glasspool RM, Teodoridis JM, Brown R. Epigenetics as a mechanism driving polygenic clinical drug resistance. British journal of cancer 2006;94(8):1087-92. 76. Hendrich B, Bird A. Mammalian methyltransferases and methyl-CpG-binding domains: proteins involved in DNA methylation. Current topics in microbiology and immunology 2000;249:55-74. 77. Leone G, Teofili L, Voso MT, Lubbert M. DNA methylation and demethylating drugs in myelodysplastic syndromes and secondary leukemias. Haematologica 2002;87(12):1324-41. 78. Uchida T, Kinoshita T, Nagai H, et al. Hypermethylation of the p15INK4B gene in myelodysplastic syndromes. Blood 1997;90(4):1403-9. 79. Bird A. DNA methylation patterns and epigenetic memory. Genes & development 2002;16(1):6-21. 80. Daskalakis M, Nguyen TT, Nguyen C, et al. Demethylation of a hypermethylated P15/INK4B gene in patients with myelodysplastic syndrome by 5-Aza-2'-deoxycytidine (decitabine) treatment. Blood 2002;100(8):2957-64. 81. Yoo CB, Jones PA. Epigenetic therapy of cancer: past, present and future. Nature reviews 2006;5(1):37-50. 82. Egger G, Liang G, Aparicio A, Jones PA. Epigenetics in human disease and prospects for epigenetic therapy. Nature 2004;429(6990):457-63. 83. Ravandi F, Kantarjian H, Giles F, Cortes J. New agents in acute myeloid leukemia and other myeloid disorders. Cancer 2004;100(3):441-54.

243 84. Wei SH, Brown R, Huang TH. Aberrant DNA methylation in ovarian cancer: is there an epigenetic predisposition to drug response? Annals of the New York Academy of Sciences 2003;983:243-50. 85. Chen RZ, Pettersson U, Beard C, Jackson-Grusby L, Jaenisch R. DNA hypomethylation leads to elevated mutation rates. Nature 1998;395(6697):89-93. 86. Lengauer C, Kinzler KW, Vogelstein B. Genetic instabilities in human cancers. Nature 1998;396(6712):643-9. 87. Raderschall E, Stout K, Freier S, Suckow V, Schweiger S, Haaf T. Elevated levels of Rad51 recombination protein in tumor cells. Cancer research 2002;62(1):219-25. 88. Rudek MA, Zhao M, He P, et al. Pharmacokinetics of 5-azacitidine administered with phenylbutyrate in patients with refractory solid tumors or hematologic malignancies. J Clin Oncol 2005;23(17):3906-11. 89. Lu LJ, Randerath K. Long term instability and molecular mechanism of 5- azacytidine-induced DNA hypomethylation in normal and neoplastic tissues in vivo. Molecular pharmacology 1984;26(3):594-603. 90. Khan R, Aggerholm A, Hokland P, Hassan M, Hellstrom-Lindberg E. A pharmacodynamic study of 5-azacytidine in the P39 cell line. Experimental hematology 2006;34(1):35-43. 91. Murakami T, Li X, Gong J, Bhatia U, Traganos F, Darzynkiewicz Z. Induction of apoptosis by 5-azacytidine: drug concentration-dependent differences in cell cycle specificity. Cancer research 1995;55(14):3093-8. 92. Liu Z, Liu S, Xie Z, et al. Characterization of in vitro and in vivo hypomethylating effects of decitabine in acute myeloid leukemia by a rapid, specific and sensitive LC-MS/MS method. Nucleic acids research 2007;35(5):e31. 93. Yang AS, Doshi KD, Choi SW, et al. DNA methylation changes after 5-aza-2'- deoxycytidine therapy in patients with leukemia. Cancer research 2006;66(10):5495-503. 94. Issa JP, Gharibyan V, Cortes J, et al. Phase II study of low-dose decitabine in patients with chronic myelogenous leukemia resistant to imatinib mesylate. J Clin Oncol 2005;23(17):3948-56. 95. Qin T, Youssef EM, Jelinek J, et al. Effect of cytarabine and decitabine in combination in human leukemic cell lines. Clin Cancer Res 2007;13(14):4225-32. 96. Ueno M, Katayama K, Yamauchi H, Nakayama H, Doi K. Cell cycle progression is required for nuclear migration of neural progenitor cells. Brain research 2006;1088(1):57-67. 97. Goldie JH, Coldman AJ. Genetic instability in the development of drug resistance. Seminars in oncology 1985;12(3):222-30. 98. Jaffrezou JP, Chen G, Duran GE, Kuhl JS, Sikic BI. Mutation rates and mechanisms of resistance to etoposide determined from fluctuation analysis. Journal of the National Cancer Institute 1994;86(15):1152-8. 99. Wang G, Wilson TJ, Jiang Q, Taylor DE. Spontaneous mutations that confer antibiotic resistance in Helicobacter pylori. Antimicrobial agents and chemotherapy 2001;45(3):727-33. 100. Ji W, Hernandez R, Zhang XY, et al. DNA demethylation and pericentromeric rearrangements of chromosome 1. Mutation research 1997;379(1):33-41.

244 101. Mackey JR, Baldwin SA, Young JD, Cass CE. Nucleoside transport and its significance for anticancer drug resistance. Drug Resist Updat 1998;1(5):310-24. 102. Fukushima M, Fujioka A, Uchida J, Nakagawa F, Takechi T. Thymidylate synthase (TS) and ribonucleotide reductase (RNR) may be involved in acquired resistance to 5-fluorouracil (5-FU) in human cancer xenografts in vivo. Eur J Cancer 2001;37(13):1681-7. 103. Iwasaki H, Huang P, Keating MJ, Plunkett W. Differential incorporation of ara-C, gemcitabine, and fludarabine into replicating and repairing DNA in proliferating human leukemia cells. Blood 1997;90(1):270-8. 104. Kubota T, Watanabe M, Otani Y, Kitajima M, Fukushiuma M. Different pathways of 5-fluorouracil metabolism after continuous venous or bolus injection in patients with colon carcinoma: possible predictive value of thymidylate synthetase mRNA and ribonucleotide reductase for 5-fluorouracil sensitivity. Anticancer research 2002;22(6B):3537-40. 105. Matsusaka S, Yamasaki H, Fukushima M, Wakabayashi I. Upregulation of enzymes metabolizing 5-fluorouracil in colorectal cancer. Chemotherapy 2007;53(1):36- 41. 106. Okamura H, Kamei T, Sakuma N, Hanai N, Ishihara T. Ribonucleotide reductase immunoreactivity in adenocarcinoma cells and malignant or reactive mesothelial cells in serous effusions. Acta cytologica 2003;47(2):209-15. 107. Cerqueira NM, Fernandes PA, Ramos MJ. Enzyme ribonucleotide reductase: unraveling an enigmatic paradigm of enzyme inhibition by furanone derivatives. The journal of physical chemistry 2006;110(42):21272-81. 108. Fattal E, Couvreur P, Dubernet C. "Smart" delivery of antisense oligonucleotides by anionic pH-sensitive liposomes. Advanced drug delivery reviews 2004;56(7):931-46. 109. Stresemann C, Lyko F. Modes of action of the DNA methyltransferase inhibitors azacytidine and decitabine. International journal of cancer 2008;123(1):8-13. 110. Taylor EM, McFarlane RJ, Price C. 5-Azacytidine treatment of the fission yeast leads to cytotoxicity and cell cycle arrest. Mol Gen Genet 1996;253(1-2):128-37. 111. Jones PA. Altering gene expression with 5-azacytidine. Cell 1985;40(3):485-6. 112. Jones PA, Taylor SM. Cellular differentiation, cytidine analogs and DNA methylation. Cell 1980;20(1):85-93. 113. Zielinski WS, Sprinzl M. Chemical synthesis of 5-azacytidine nucleotides and preparation of tRNAs containing 5-azacytidine in its 3'-terminus. Nucleic acids research 1984;12(12):5025-36. 114. Lee TT, Karon MR. Inhibition of protein synthesis in 5-azacytidine-treated HeLa cells. Biochemical pharmacology 1976;25(15):1737-42. 115. Reichman M, Penman S. The mechanism of inhibition of protein synthesis by 5- azacytidine in HeLa cells. Biochimica et biophysica acta 1973;324(2):282-9. 116. Chou TC, Talalay P. Quantitative analysis of dose-effect relationships: the combined effects of multiple drugs or enzyme inhibitors. Advances in enzyme regulation 1984;22:27-55. 117. Chen P, Liu Z, Liu S, et al. A LC-MS/MS method for the analysis of intracellular nucleoside triphosphate levels. Pharmaceutical research 2009;26(6):1504-15.

245 118. Reichard P. Interactions between deoxyribonucleotide and DNA synthesis. Annual review of biochemistry 1988;57:349-74. 119. Wright DG, LaRussa VF, Knight RD, Bednarek JM, Cutting MA. Guanine ribonucleotide metabolism and the regulation of myelopoiesis. Advances in experimental medicine and biology 1991;309B:301-4. 120. Grusch M, Fritzer-Szekeres M, Fuhrmann G, et al. Activation of caspases and induction of apoptosis by novel ribonucleotide reductase inhibitors amidox and didox. Experimental hematology 2001;29(5):623-32. 121. Ghoshal K, Bai S. DNA methyltransferases as targets for cancer therapy. Drugs Today (Barc) 2007;43(6):395-422. 122. Evans JS, Hanka LJ. The in vivo activity of combinations of 5-azacytidine and cytidine on leukemia L-1210. Experientia 1968;24(9):922-3. 123. Chen G, Wang Y, Huang H, et al. Combination of DNA methylation inhibitor 5- azacytidine and arsenic trioxide has synergistic activity in myeloma. European journal of haematology 2009;82(3):176-83. 124. Fan H, Villegas C, Huang A, Wright JA. The mammalian ribonucleotide reductase R2 component cooperates with a variety of oncogenes in mechanisms of cellular transformation. Cancer research 1998;58(8):1650-3. 125. Juliano RL, Alahari S, Yoo H, Kole R, Cho M. Antisense pharmacodynamics: critical issues in the transport and delivery of antisense oligonucleotides. Pharmaceutical research 1999;16(4):494-502. 126. Vassilakos A, Lee Y, Viau S, et al. GTI-2040 displays cooperative anti-tumor activity when combined with interferon alpha against human renal carcinoma xenografts. International journal of oncology 2009;34(1):33-42. 127. Simeoni M, Magni P, Cammia C, et al. Predictive pharmacokinetic- pharmacodynamic modeling of tumor growth kinetics in xenograft models after administration of anticancer agents. Cancer research 2004;64(3):1094-101. 128. Rocchetti M, Poggesi I, Germani M, et al. A pharmacokinetic-pharmacodynamic model for predicting tumour growth inhibition in mice: a useful tool in oncology drug development. Basic & clinical pharmacology & toxicology 2005;96(3):265-8. 129. Xu XL. The biological foundation of the Gompertz model. International journal of bio-medical computing 1987;20(1-2):35-9. 130. Bissery MC, Vrignaud P, Lavelle F, Chabot GG. Experimental antitumor activity and pharmacokinetics of the camptothecin analog irinotecan (CPT-11) in mice. Anti- cancer drugs 1996;7(4):437-60. 131. Norton L, Simon R. Growth curve of an experimental solid tumor following radiotherapy. Journal of the National Cancer Institute 1977;58(6):1735-41. 132. Kozusko F, Bajzer Z. Combining Gompertzian growth and cell population dynamics. Mathematical biosciences 2003;185(2):153-67. 133. Tham LS, Wang L, Soo RA, et al. A pharmacodynamic model for the time course of tumor shrinkage by gemcitabine + carboplatin in non-small cell lung cancer patients. Clin Cancer Res 2008;14(13):4213-8. 134. Moreno D, Troconiz IF, Enguita M, Bandres E, Garcia-Foncillas J, Garrido MJ. Semi-mechanistic description of the in-vitro antiproliferative effect of different antitumour agents. The Journal of pharmacy and pharmacology 2008;60(1):77-82.

246 135. Wen-tao Q, Ying Z, Juan M, et al. Optimization of the cell seeding density and modeling of cell growth and metabolism using the modified Gompertz model for microencapsulated animal cell culture. Biotechnology and bioengineering 2006;93(5):887-95. 136. Mehrara E, Forssell-Aronsson E, Ahlman H, Bernhardt P. Quantitative analysis of tumor growth rate and changes in tumor marker level: specific growth rate versus doubling time. Acta oncologica (Stockholm, Sweden) 2009;48(4):591-7. 137. Huang Y, Anderle P, Bussey KJ, et al. Membrane transporters and channels: role of the transportome in cancer chemosensitivity and chemoresistance. Cancer research 2004;64(12):4294-301. 138. Rius M, Stresemann C, Keller D, et al. Human concentrative nucleoside transporter 1-mediated uptake of 5-azacytidine enhances DNA demethylation. Molecular cancer therapeutics 2009;8(1):225-31. 139. Zhang J, Visser F, King KM, Baldwin SA, Young JD, Cass CE. The role of nucleoside transporters in cancer chemotherapy with nucleoside drugs. Cancer metastasis reviews 2007;26(1):85-110. 140. Issa JP, Kantarjian HM. Targeting DNA methylation. Clin Cancer Res 2009;15(12):3938-46. 141. Alleman WG, Tabios RL, Chandramouli GV, et al. The in vitro and in vivo effects of re-expressing methylated von Hippel-Lindau tumor suppressor gene in clear cell renal carcinoma with 5-aza-2'-deoxycytidine. Clin Cancer Res 2004;10(20):7011-21. 142. Claus R, Lubbert M. Epigenetic targets in hematopoietic malignancies. Oncogene 2003;22(42):6489-96. 143. Digel W, Lubbert M. DNA methylation disturbances as novel therapeutic target in lung cancer: preclinical and clinical results. Critical reviews in oncology/hematology 2005;55(1):1-11. 144. Hsi LC, Xi X, Wu Y, Lippman SM. The methyltransferase inhibitor 5-aza-2- deoxycytidine induces apoptosis via induction of 15-lipoxygenase-1 in colorectal cancer cells. Molecular cancer therapeutics 2005;4(11):1740-6. 145. Sorm F, Vesely J. Effect of 5-aza-2'-deoxycytidine against leukemic and hemopoietic tissues in AKR mice. Neoplasma 1968;15(4):339-43. 146. Schneider-Stock R, Diab-Assef M, Rohrbeck A, et al. 5-Aza-cytidine is a potent inhibitor of DNA methyltransferase 3a and induces apoptosis in HCT-116 colon cancer cells via Gadd45- and p53-dependent mechanisms. The Journal of pharmacology and experimental therapeutics 2005;312(2):525-36. 147. Yamaguchi T, Matsuda K, Sagiya Y, et al. p53R2-dependent pathway for DNA synthesis in a p53-regulated cell cycle checkpoint. Cancer research 2001;61(22):8256-62. 148. Chang L, Zhou B, Hu S, et al. ATM-mediated serine 72 phosphorylation stabilizes ribonucleotide reductase small subunit p53R2 protein against MDM2 to DNA damage. Proceedings of the National Academy of Sciences of the United States of America 2008;105(47):18519-24. 149. Kaminskas E, Farrell AT, Wang YC, Sridhara R, Pazdur R. FDA drug approval summary: azacitidine (5-azacytidine, Vidaza) for injectable suspension. The oncologist 2005;10(3):176-82.

247 150. Garzon R, Liu S, Fabbri M, et al. MicroRNA-29b induces global DNA hypomethylation and tumor suppressor gene reexpression in acute myeloid leukemia by targeting directly DNMT3A and 3B and indirectly DNMT1. Blood 2009;113(25):6411-8. 151. Song L, James SR, Kazim L, Karpf AR. Specific method for the determination of genomic DNA methylation by liquid chromatography-electrospray ionization tandem mass spectrometry. Analytical chemistry 2005;77(2):504-10. 152. Giles FJ. A new era for ribonucleoside reductase inhibition. Leukemia research 2007;31(9):1163-4. 153. Giles FJ, Fracasso PM, Kantarjian HM, et al. Phase I and pharmacodynamic study of Triapine, a novel ribonucleotide reductase inhibitor, in patients with advanced leukemia. Leukemia research 2003;27(12):1077-83. 154. Karp JE, Giles FJ, Gojo I, et al. A phase I study of the novel ribonucleotide reductase inhibitor 3-aminopyridine-2-carboxaldehyde thiosemicarbazone (3-AP, Triapine) in combination with the nucleoside analog fludarabine for patients with refractory acute leukemias and aggressive myeloproliferative disorders. Leukemia research 2008;32(1):71-7. 155. Hollenbach PW, Nguyen AN, Brady H, et al. A comparison of azacitidine and decitabine activities in acute myeloid leukemia cell lines. PloS one;5(2):e9001. 156. Chan KK, Giannini DD, Staroscik JA, Sadee W. 5-Azacytidine hydrolysis kinetics measured by high-pressure liquid chromatography and 13C-NMR spectroscopy. Journal of pharmaceutical sciences 1979;68(7):807-12. 157. Avramis VI, Biener R, Krailo M, et al. Biochemical pharmacology of high dose 1-beta-D-arabinofuranosylcytosine in childhood acute leukemia. Cancer research 1987;47(24 Pt 1):6786-92. 158. Kelecsenyi Z, Spencer DL, Caspary WJ. Molecular analysis of 5-azacytidine- induced variants in mammalian cells. Mutagenesis 2000;15(1):25-31. 159. Beisler JA. Isolation, characterization, and properties of a labile hydrolysis product of the antitumor nucleoside, 5-azacytidine. Journal of medicinal chemistry 1978;21(2):204-8. 160. Lin KT, Momparler RL, Rivard GE. High-performance liquid chromatographic analysis of chemical stability of 5-aza-2'-deoxycytidine. Journal of pharmaceutical sciences 1981;70(11):1228-32. 161. Rogstad DK, Herring JL, Theruvathu JA, et al. Chemical decomposition of 5-aza- 2'-deoxycytidine (Decitabine): kinetic analyses and identification of products by NMR, HPLC, and mass spectrometry. Chemical research in toxicology 2009;22(6):1194-204. 162. Hatse S, De Clercq E, Balzarini J. Role of antimetabolites of purine and pyrimidine nucleotide metabolism in tumor cell differentiation. Biochemical pharmacology 1999;58(4):539-55. 163. Wang J, Lohman GJ, Stubbe J. Enhanced subunit interactions with gemcitabine- 5'-diphosphate inhibit ribonucleotide reductases. Proceedings of the National Academy of Sciences of the United States of America 2007;104(36):14324-9. 164. Ellims PH, Gan TE, Medley G. Cytidine triphosphate synthetase activity in lymphoproliferative disorders. Cancer research 1983;43(3):1432-5. 165. Williams JC, Kizaki H, Weber G, Morris HP. Increased CTP synthetase activity in cancer cells. Nature 1978;271(5640):71-3.

248 166. Ernest MJ. Regulation of tyrosine aminotransferase messenger ribonucleic acid in rat liver. effect of cycloheximide on messenger ribonucleic acid turnover. Biochemistry 1982;21(26):6761-7. 167. Ross J. mRNA stability in mammalian cells. Microbiological reviews 1995;59(3):423-50. 168. Bernstein P, Peltz SW, Ross J. The poly(A)-poly(A)-binding protein complex is a major determinant of mRNA stability in vitro. Molecular and cellular biology 1989;9(2):659-70. 169. Burd CG, Dreyfuss G. Conserved structures and diversity of functions of RNA- binding proteins. Science (New York, NY 1994;265(5172):615-21. 170. Klausner RD, Rouault TA, Harford JB. Regulating the fate of mRNA: the control of cellular iron metabolism. Cell 1993;72(1):19-28. 171. Amara FM, Chen FY, Wright JA. Phorbol ester modulation of a novel cytoplasmic protein binding activity at the 3'-untranslated region of mammalian ribonucleotide reductase R2 mRNA and role in message stability. The Journal of biological chemistry 1994;269(9):6709-15. 172. Cihak A. Biological effects of 5-azacytidine in eukaryotes. Oncology 1974;30(5):405-22. 173. Lin HL, Glazer RI. The comparative effects of 5-azacytidine and dihydro-5- azacytidine on polysomal RNA in Ehrlich ascites cells in vitro. Molecular pharmacology 1981;20(3):644-8. 174. Zain BS, Adams RL, Imrie RC. Effect of 5-azacytidine on phytohemagglutinin- stimulated horse lymphocytes and cultured mouse L929 cells. Cancer research 1973;33(1):40-6. 175. Pillai RS, Bhattacharyya SN, Filipowicz W. Repression of protein synthesis by miRNAs: how many mechanisms? Trends in cell biology 2007;17(3):118-26. 176. Goll MG, Kirpekar F, Maggert KA, et al. Methylation of tRNAAsp by the DNA methyltransferase homolog Dnmt2. Science (New York, NY 2006;311(5759):395-8. 177. Schaefer M, Hagemann S, Hanna K, Lyko F. Azacytidine inhibits RNA methylation at DNMT2 target sites in human cancer cell lines. Cancer research 2009;69(20):8127-32. 178. Rassoulzadegan M, Grandjean V, Gounon P, Vincent S, Gillot I, Cuzin F. RNA- mediated non-mendelian inheritance of an epigenetic change in the mouse. Nature 2006;441(7092):469-74. 179. Lewis BP, Shih IH, Jones-Rhoades MW, Bartel DP, Burge CB. Prediction of mammalian microRNA targets. Cell 2003;115(7):787-98. 180. Estey E, Plunkett W, Gandhi V, Rios MB, Kantarjian H, Keating MJ. Fludarabine and arabinosylcytosine therapy of refractory and relapsed acute myelogenous leukemia. Leukemia & lymphoma 1993;9(4-5):343-50. 181. Gandhi V, Estey E, Keating MJ, Plunkett W. Fludarabine potentiates metabolism of cytarabine in patients with acute myelogenous leukemia during therapy. J Clin Oncol 1993;11(1):116-24. 182. Gandhi V, Estey E, Keating MJ, Plunkett W. Biochemical modulation of arabinosylcytosine for therapy of leukemias. Leukemia & lymphoma 1993;10 Suppl:109- 14.

249 183. Lyko F, Brown R. DNA methyltransferase inhibitors and the development of epigenetic cancer therapies. Journal of the National Cancer Institute 2005;97(20):1498- 506. 184. Robert MF, Morin S, Beaulieu N, et al. DNMT1 is required to maintain CpG methylation and aberrant gene silencing in human cancer cells. Nature genetics 2003;33(1):61-5. 185. Goan YG, Zhou B, Hu E, Mi S, Yen Y. Overexpression of ribonucleotide reductase as a mechanism of resistance to 2,2-difluorodeoxycytidine in the human KB cancer cell line. Cancer research 1999;59(17):4204-7. 186. Fuks F, Burgers WA, Godin N, Kasai M, Kouzarides T. Dnmt3a binds deacetylases and is recruited by a sequence-specific repressor to silence transcription. The EMBO journal 2001;20(10):2536-44. 187. Mizuno S, Chijiwa T, Okamura T, et al. Expression of DNA methyltransferases DNMT1, 3A, and 3B in normal hematopoiesis and in acute and chronic myelogenous leukemia. Blood 2001;97(5):1172-9. 188. Margot JB, Cardoso MC, Leonhardt H. Mammalian DNA methyltransferases show different subnuclear distributions. Journal of cellular biochemistry 2001;83(3):373- 9. 189. Mishra MV, Bisht KS, Sun L, et al. DNMT1 as a molecular target in a multimodality-resistant phenotype in tumor cells. Mol Cancer Res 2008;6(2):243-9. 190. Bestor TH. The DNA methyltransferases of mammals. Human molecular genetics 2000;9(16):2395-402. 191. Kim GD, Ni J, Kelesoglu N, Roberts RJ, Pradhan S. Co-operation and communication between the human maintenance and de novo DNA (cytosine-5) methyltransferases. The EMBO journal 2002;21(15):4183-95. 192. Bachman KE, Rountree MR, Baylin SB. Dnmt3a and Dnmt3b are transcriptional repressors that exhibit unique localization properties to heterochromatin. The Journal of biological chemistry 2001;276(34):32282-7. 193. Segura-Pacheco B, Perez-Cardenas E, Taja-Chayeb L, et al. Global DNA hypermethylation-associated cancer chemotherapy resistance and its reversion with the demethylating agent hydralazine. Journal of translational medicine 2006;4:32. 194. Zhou BS, Ker R, Ho R, et al. Determination of deoxyribonucleoside triphosphate pool sizes in ribonucleotide reductase cDNA transfected human KB cells. Biochemical pharmacology 1998;55(10):1657-65. 195. Bianchi V, Pontis E, Reichard P. Changes of deoxyribonucleoside triphosphate pools induced by hydroxyurea and their relation to DNA synthesis. The Journal of biological chemistry 1986;261(34):16037-42. 196. Hurta RA, Wright JA. Mammalian drug resistant mutants with multiple gene amplifications: genes encoding the M1 component of ribonucleotide reductase, the M2 component of ribonucleotide reductase, ornithine decarboxylase, p5-8, the H-subunit of ferritin and the L-subunit of ferritin. Biochimica et biophysica acta 1990;1087(2):165-72. 197. Mansson E, Spasokoukotskaja T, Sallstrom J, Eriksson S, Albertioni F. Molecular and biochemical mechanisms of fludarabine and cladribine resistance in a human promyelocytic cell line. Cancer research 1999;59(23):5956-63.

250 198. Fontecave M. Ribonucleotide reductases and radical reactions. Cell Mol Life Sci 1998;54(7):684-95. 199. Reichard P, Ehrenberg A. Ribonucleotide reductase--a radical enzyme. Science (New York, NY 1983;221(4610):514-9. 200. Jordan A, Reichard P. Ribonucleotide reductases. Annual review of biochemistry 1998;67:71-98. 201. Kolberg M, Strand KR, Graff P, Andersson KK. Structure, function, and mechanism of ribonucleotide reductases. Biochimica et biophysica acta 2004;1699(1- 2):1-34. 202. Nocentini G. Ribonucleotide reductase inhibitors: new strategies for cancer chemotherapy. Crit Rev Oncol Hematol 1996;22(2):89-126. 203. Cory JG. Ribonucleotide reductase as a chemotherapeutic target. Advances in enzyme regulation 1988;27:437-55. 204. Wei X. Pharmacokinetics, pharmacodynamics and metabolism of GTI-2040, a phosphorothioate oligonucleotide targeting R2 subunit of ribonucleotide reductase [dissertation]; 2006. 205. Gandhi V, Kemena A, Keating MJ, Plunkett W. Fludarabine infusion potentiates arabinosylcytosine metabolism in lymphocytes of patients with chronic lymphocytic leukemia. Cancer research 1992;52(4):897-903. 206. Gandhi V, Plunkett W. Modulation of arabinosylnucleoside metabolism by arabinosylnucleotides in human leukemia cells. Cancer research 1988;48(2):329-34. 207. Breithaupt H, Pralle H, Eckhardt T, von Hattingberg M, Schick J, Loffler H. Clinical results and pharmacokinetics of high-dose cytosine arabinoside (HD ARA-C). Cancer 1982;50(7):1248-57. 208. Capizzi RL, Poole M, Cooper MR, et al. Treatment of poor risk acute leukemia with sequential high-dose ARA-C and asparaginase. Blood 1984;63(3):694-700. 209. Capizzi RL, White JC, Powell BL, Perrino F. Effect of dose on the pharmacokinetic and pharmacodynamic effects of cytarabine. Seminars in hematology 1991;28(3 Suppl 4):54-69. 210. Estey E, Plunkett W, Dixon D, Keating M, McCredie K, Freireich EJ. Variables predicting response to high dose cytosine arabinoside therapy in patients with refractory acute leukemia. Leukemia 1987;1(8):580-3. 211. Qian SX, Li JY, Tian T, et al. Effect of low-dose cytarabine and aclarubicin in combination with granulocyte colony-stimulating factor priming (CAG regimen) on the outcome of elderly patients with acute myeloid leukemia. Leukemia research 2007;31(10):1383-8. 212. Willemze R, Zwaan FE, Colpin G, Keuning JJ. High dose cytosine arabinoside in the management of refractory acute leukaemia. Scandinavian journal of haematology 1982;29(2):141-6. 213. Yamauchi T, Negoro E, Arai H, et al. Combined low-dose cytarabine, melphalan and mitoxantrone for older patients with acute myeloid leukemia or high-risk myelodysplastic syndrome. Anticancer research 2007;27(4C):2635-9. 214. Zhang WG, Wang FX, Chen YX, et al. Combination chemotherapy with low- dose cytarabine, homoharringtonine, and granulocyte colony-stimulating factor priming

251 in patients with relapsed or refractory acute myeloid leukemia. American journal of hematology 2008;83(3):185-8. 215. Bauer W, Horvath Z, Hochtl T, et al. Amidox, an inhibitor of ribonucleotide reductase, potentiates the action of Ara-C in HL-60 human promyelocytic leukemia cells. Nucleosides, nucleotides & nucleic acids 2004;23(8-9):1541-4. 216. Fritzer-Szekeres M, Salamon A, Grusch M, et al. Trimidox, an inhibitor of ribonucleotide reductase, synergistically enhances the inhibition of colony formation by Ara-C in HL-60 human promyelocytic leukemia cells. Biochemical pharmacology 2002;64(3):481-5. 217. Plagemann PG, Behrens M, Abraham D. Metabolism and cytotoxicity of 5- azacytidine in cultured Novikoff rat hepatoma and P388 mouse leukemia cells and their enhancement by preincubation with pyrazofurin. Cancer research 1978;38(8):2458-66. 218. Gandhi V, Estey E, Keating MJ, Chucrallah A, Plunkett W. Chlorodeoxyadenosine and arabinosylcytosine in patients with acute myelogenous leukemia: pharmacokinetic, pharmacodynamic, and molecular interactions. Blood 1996;87(1):256-64. 219. Gandhi V, Estey E, Plunkett W. Modulation of arabinosylcytosine metabolism during leukemia therapy. Advances in experimental medicine and biology 1994;370:119- 24. 220. Gandhi V, Nowak B, Keating MJ, Plunkett W. Modulation of arabinosylcytosine metabolism by arabinosyl-2-fluoroadenine in lymphocytes from patients with chronic lymphocytic leukemia: implications for combination therapy. Blood 1989;74(6):2070-5. 221. Bhalla K, Swerdlow P, Grant S. Effects of and hydroxyurea on the metabolism and cytotoxicity of 1-B-D arabinofuranosylcytosine in highly resistant human leukemia cells. Blood 1991;78(11):2937-44. 222. Colly LP, Richel DJ, Arentsen-Honders MW, Kester MG, ter Riet PM, Willemze R. Increase in Ara-C sensitivity in Ara-C sensitive and -resistant leukemia by stimulation of the salvage and inhibition of the de novo pathway. Annals of hematology 1992;65(1):26-32. 223. Cooper T, Ayres M, Nowak B, Gandhi V. Biochemical modulation of cytarabine triphosphate by clofarabine. Cancer chemotherapy and pharmacology 2005;55(4):361-8. 224. Wang LM, White JC, Capizzi RL. The effect of ara-C-induced inhibition of DNA synthesis on its cellular pharmacology. Cancer chemotherapy and pharmacology 1990;25(6):418-24. 225. Verschuur AC, Van Gennip AH, Leen R, Voute PA, Brinkman J, Van Kuilenburg AB. Cyclopentenyl cytosine increases the phosphorylation and incorporation into DNA of 1-beta-D-arabinofuranosyl cytosine in a human T-lymphoblastic cell line. International journal of cancer 2002;98(4):616-23. 226. Lowenberg B, Downing JR, Burnett A. Acute myeloid leukemia. The New England journal of medicine 1999;341(14):1051-62. 227. Byrd JC, Mrozek K, Dodge RK, et al. Pretreatment cytogenetic abnormalities are predictive of induction success, cumulative incidence of relapse, and overall survival in adult patients with de novo acute myeloid leukemia: results from Cancer and Leukemia Group B (CALGB 8461). Blood 2002;100(13):4325-36.

252 228. Mrozek K, Heinonen K, Bloomfield CD. Prognostic value of cytogenetic findings in adults with acute myeloid leukemia. International journal of hematology 2000;72(3):261-71. 229. Grem JL, Takimoto CH, Multani P, et al. Antimetabolites. Cancer chemotherapy and biological response modifiers 1999;18:1-38. 230. Johnson SA. Clinical pharmacokinetics of nucleoside analogues: focus on haematological malignancies. Clinical pharmacokinetics 2000;39(1):5-26. 231. Estey E, Thall P, Andreeff M, et al. Use of granulocyte colony-stimulating factor before, during, and after fludarabine plus cytarabine induction therapy of newly diagnosed acute myelogenous leukemia or myelodysplastic syndromes: comparison with fludarabine plus cytarabine without granulocyte colony-stimulating factor. J Clin Oncol 1994;12(4):671-8. 232. Gandhi V, Estey E, Du M, Nowak B, Keating MJ, Plunkett W. Modulation of the cellular metabolism of cytarabine and fludarabine by granulocyte-colony-stimulating factor during therapy of acute myelogenous leukemia. Clin Cancer Res 1995;1(2):169-78. 233. Zhang Z, Weinschenk T, Schluesener HJ. Uptake, cellular distribution and novel cellular binding proteins of immunostimulatory CpG oligodeoxynucleotides in glioblastoma cells. Molecular and cellular biochemistry 2005;272(1-2):35-46. 234. Nakai D, Seita T, Terasaki T, et al. Cellular uptake mechanism for oligonucleotides: involvement of endocytosis in the uptake of phosphodiester oligonucleotides by a human colorectal adenocarcinoma cell line, HCT-15. J Pharmacol Exp Ther 1996;278(3):1362-72. 235. Wu-Pong S, Weiss TL, Hunt CA. Antisense c-myc oligonucleotide cellular uptake and activity. Antisense research and development 1994;4(3):155-63. 236. Helene C, Toulme JJ. Specific regulation of gene expression by antisense, sense and antigene nucleic acids. Biochimica et biophysica acta 1990;1049(2):99-125. 237. Silverman LR, Demakos EP, Peterson BL, et al. Randomized controlled trial of azacitidine in patients with the myelodysplastic syndrome: a study of the cancer and leukemia group B. J Clin Oncol 2002;20(10):2429-40. 238. Chitambar CR, Libnoch JA, Matthaeus WG, Ash RC, Ritch PS, Anderson T. Evaluation of continuous infusion low-dose 5-azacytidine in the treatment of myelodysplastic syndromes. American journal of hematology 1991;37(2):100-4.

253