TACTICS for Bioimaging Informatics and Analysis of T Cells Raz Shimoni Submitted in total fulfilment of the requirements of the degree of Doctor of Philosophy Centre for Micro-Photonics (CMP) Faculty of Science, Engineering and Technology SWINBURNE UNIVERSITY OF TECHNOLOGY 2014 i "Absence of evidence is not evidence of absence" ii Dedicated to my princess Liliana and to my lovely wife Olga You are my sunshine and my hope iii Abstract T cells are a highly specialized type of white blood cell, with a critical role in cell-mediated immunity. The normal function of T cells is imperative for an efficient immune response, providing adaptable defence against new foreign bodies. Unfortunately, in some cases this system fails, leaving our body exposed to disease. Immunologists are now trying to dissect how T cells participate in the immune response, which might lead to the development of new drugs and treatments. One of the latest approaches to studying T cells is to utilize fluorescence microscopy imaging of live cells over time. Using this technique it is possible, for instance, to monitor how the selective recruitment of molecules to polarized regions within the cells can affect functionality and fate. Despite the rapid developments in fluorescence microscopy imaging and the wide range of available analytical software, the analysis of time-lapse data is not yet perfected. The development of novel strategies to aid the interpretation of information collected by fluorescence time-lapse imaging, and how they can be used to improve our analysis of live cell imaging, are the main subjects of this thesis. Herein, a modular and adaptable high-throughput toolbox named TACTICS has been developed. The TACTICS pipeline is first described, offering features not seen in other standard Bioimaging informatics software. Taking advantage of TACTICS as a computational platform, several challenges in the analysis of fluorescence microscopy data are addressed. In particular, the dependency of fluorescence measurements on experimental setup, the difficulty of automatically processing large data sets, and the effects of imaging settings on the accuracy of the analysis were explored. Next, these new tools were employed to measure protein polarization in T cells during migration and cell-cell interactions. Finally, multi- parametric analysis, integrating image cytometry and interactive lineage informatics analysis, was demonstrated to be extremely useful to study the fate of T cells. Together these advances have been used for the analysis of Asymmetric Cell Division (ACD) and polarity studies, providing an innovative toolbox to help elucidate key insights in the biology of T cells. iv v Acknowledgements This work was conducted under the direct supervision of Prof. Sarah Russell, the head of the Immune Signalling Laboratory at Peter MacCallum Cancer Centre. Sarah, thank you for selecting me to carry out this project, thank you for your infinite patient, for the time you spent with me to look together into raw data, and for many ideas and suggestions to improve the analysis. It has been a great honour for me to be a part of the Centre for micro- Photonics (CMP), supervised by Prof. Min Gu, the director of the CMP. Thank you Min for your support and for encouragement, it means a great deal for me. I would like to express my sincere appreciation to my previous advisor, Dr. Zeev Bomzon, who initially suggested for the usability of image cytometry software dedicated to study polarity from time-lapse imaging. I would like to thank Dr. Daniel Day, who served as advisor in my first year of candidature, for our intriguing intellectual discussions, and for his valuable microfabrication training. Special thanks to my mentor Dr. Pavel Lobachevsky who represented me at the student committees, in addition to his research advices and support. I also would like to thank all members of my PhD committee, in particular the head of committee Dr. Ilia Voskoboinik and education officer Dr. Caroline Owens. The interaction between programming and biology has been vital to the success of establishment new tools throughout this project. For exchanging ideas and opportunities, setting requirements, and provided their data for software development, I wish to thank my collaborators: Mohammed Yassin, Dr. Kim Pham, Dr. Jane Oliaro, Kelly Ramsbottom, Mandy Ludford-Menting, Dr. Edwin Hawkins and Dr. Kerrie-Ann McMahon. I have been benefited to work with you and to learn more about the biology aspect of your research, and enjoyed your friendly company at the lab. Additionally, I wish to thank TACTICS beta-testers Amelia Poetter, Emily O’Connor and Adam Poetter for helping to identify bugs and useful suggestions. Your work was essential to improve and validate the quality of the software and analyzed data. It was awesome to work with the next generation of scientists. I would like to the thank Dr. Sarah Ellis and Cameron Nowell for microscopy training. I would like to the thank Mandy Ludford-Menting for her PC2 and tissue culture training. I wish to thank Pierrette Michaux for clean room training. Additionally, I vi wish to thank Ricardas Buividas and Dr. David MacDonald for their assistance with the laser setup for microfabrication experiments. I wish to thank all members of the CMP for their support and intriguing environment. The contribution of the MATLAB open-code community has been imperative to the development of TACTICS. I wish to thank many developers who contributed their source code or gave free advice. I hope that my work will inspired others as you inspired me. I would like to thank my supervisor Prof. Sarah Russell for her kind help with editing and correcting the chapters of my thesis. Selected sections were edited and corrected with additional assistance from Dr. Olga Shimoni, Dr. Jane Oliaro, Dr. Simon Partridge, and Cameron Nowell, Mohammed Yassin, and Mandy Ludford- Menting. I am deeply grateful for the generous financial support I received for the last 3.5 years through Swinburne University Postgraduate Research Award (SUPRA). I also thank my Supervisor Dr. Sarah Russell who generously chose to further top up this scholarship. Further modest student budgets were kindly received from Faculty of Engineering and Industrial Science (FEIS) and Peter MacCallum Cancer Centre education department. The submission of this thesis closes a long chapter in my life. First of all, I wish to thank my father, who always kept me in the frontier of science many years ago: my first computer Atari XL, life science books and kits, and even a first simple light microscope to explore new worlds beyond our sight. I would like to thank all my family and friends in Australia and in Israel, who supported and encouraged me to achieve my dream. In particularly, this thesis is dedicated to my beautiful daughter Liliana, who was born during the first year of my candidature. Special thanks to the wonderful staff of Goodstart Early Learning Centre in Altona, who provided me the security that my daughter is in their devoted care, allowing me spending the time required for this research. Last but not least, I dedicate this thesis to my wise wife Olga, who is a scientist herself and gave many useful advices during my research. Olga, thank you for your support, and that you always cheer me up. I have been truly privileged to have such a family. vii Declaration This is to certify that: 1. The thesis contains no material which has been accepted for the award of any other degree or diploma, except where due reference is made. 2. To my best knowledge this thesis contains no material previously published or written by another person except where due reference is made in the text of the examinable outcome. 3. Where the work is based on joint research or publications, I disclose the relative contributions of the respective workers or authors. Raz Shimoni viii Preface In Chapter 3, the development of TACTICS version 2.0 was done mainly in collaboration with Dr Kim Pham, that her project required the development of TACTICS-ACD and TACTICS-Polarization Modules. The source code of TACTICS version 2.2 was published in co-authorship with Dr Kim Pham in [1]. The development of TACTICS version 3.0 was done mainly in collaboration with Mohammed Yassin who provided data images of CD8+ T cells and that his project required the development of TACTICS-Lineage and further improvements in the TACTICS-Tracking Modules. Particularly, the interactiveness of the TACTICS- Lineage Module, and novel ideas in the analysis of lineage trees reconstructions, gating, improved tools to the TACTICS-Tracking Module for interactive user corrections, and the revolutionary idea of selective tracking operator. Manual corrections were performed by Amelia Poetter, Emily O’Connor, and Mohammed Yassin. In Chapter 4, MLA cell line was provided by Mandy Ludford-Menting-. Mohammed Yassin assisted in establishing the time-lapse microscopy setup and validation of microwells. In addition, the development of the ACD module and new approaches to investigate ACD were done in collaboration with Dr Kim Pham, who investigated the hypothesis of the existence of asymmetric cell division in immune cells. In Chapter 5, time-lapse microscopy images of MLA and thymocytes undergoing migration and division were provided by Dr Kim Pham, who also conducted all supporting biological assays. Plasmid constructs of eGFP-Numb and the mutant variation eGFP-Numb-2a were fused by Mandy Menting-Ludford, using plasmid generated by C.J. McGlade. Microfabricated microwells were supplied by Daniel Day (Microsurfaces Pty Ltd). Cameron Nowell supplied the lif2tif journal to convert .lif files to categorized .tif files. Pavel Lubachevski supplied splitting algorithm in C code. Adam Poetter assisted me with beta-testing, manual corrections of segmentation, and tracking. My innovation of the major-minor normalization approach would be impossible without statistical assistance from Prof Terry Speed, who initially suggested looking on polarization ratios across other angles of polarity, and is based on previous normalization method formed by Dr Zeev Bomzon.