Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Modeling drug efficacy in the tumour microenvironment with Saccharomyces cerevisiae genome-wide screens… Tran, Grant 2017

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2017_february_tran_grant.pdf [ 1.77MB ]
Metadata
JSON: 24-1.0340600.json
JSON-LD: 24-1.0340600-ld.json
RDF/XML (Pretty): 24-1.0340600-rdf.xml
RDF/JSON: 24-1.0340600-rdf.json
Turtle: 24-1.0340600-turtle.txt
N-Triples: 24-1.0340600-rdf-ntriples.txt
Original Record: 24-1.0340600-source.json
Full Text
24-1.0340600-fulltext.txt
Citation
24-1.0340600.ris

Full Text

MODELING DRUG EFFICACY IN THE TUMOUR MICROENVIRONMENT WITH SACCHAROMYCES CEREVISIAE GENOME-WIDE SCREENS IN HYPOXIC CONDITIONS by  Grant Tran  B.Sc., The University of British Columbia, 2013  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Pharmaceutical Sciences)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   January 2017  © Grant Tran, 2017 ii  Abstract  Hypoxia, the state of reduced oxygen, is a microenvironment found in many solid tumours and is correlated with an increased risk in patient mortality. This is due to an increase in resistance to radiotherapy and chemotherapy as well as a decrease in drug efficacy. The mechanisms and cellular factors (gene products) associated with this reduced chemotherapeutic efficacy in hypoxia remains unclear. This research looks to identify cellular processes and pathways that cancerous cells are able to exploit in order to survive and thrive in this microenvironment. The eukaryotic model baker’s yeast Saccharomyces cerevisiae combined with a genome-wide approach was used to screen the yeast knockout collection for specific genotypes that are sensitive to the hypoxic environment alone, and in combination with commonly used chemotherapeutics. Pathways and processes identified in these screens include transcriptional regulation, cytoskeleton maintenance, ribosomal biogenesis, macromolecular complex assembly and the heat shock response. The combination of heat and hypoxia was found to result in a synergistic effect that drastically affected cell fitness. DNA-damaging chemotherapeutics screened in hypoxic conditions showed reduced efficacy. Genotypes most sensitive to drugs in the hypoxic environment fall into Gene Ontology (GO) terms categorized in the response to the specific mechanism of the drug. This includes DNA repair processes such as homologous repair, post-replicative repair and mismatch repair. The mechanistic specificity uncovered in these screens suggests that the hypoxic environment exacerbates drug-specific stresses, and the identified genotypes highlight gene products and pathways critical for these responses. Cell survival and success in this microenvironment therefore requires adaptations to these exacerbated stresses, a iii  phenomenon successfully accomplished by resistant tumour cells. This research contributes to our understanding of cellular biology under this cancer microenvironment, and provides data to highlight the challenges in using chemotherapeutics to treat tumours.     iv  Preface All of the work presented was conducted in the laboratory of Dr. Corey Nislow and Dr. Guri Giaever in the Faculty of Pharmaceutical Sciences at the University of British Columbia, Point Grey campus.   Chapter 2 and 4 are based on the design of Dr. Corey Nislow, Dr. Guri Giaever, and Grant Tran. I was responsible for the experimental setup of generating a hypoxic environment, performing homozygous pool growth assays, microarray hybridization with assistance from Jennifer Chiang, and microarray data analysis with assistance from Dr. Guri Giaever. The enrichment analyses and figures generated are my original work.   Chapter 3 is based on the experimental design of Dr. Corey Nislow, Dr. Guri Giaever, and Grant Tran. I was responsible for the collection and quality control of RNA. Dr. Sunita Sinha performed the library preparation and RNA sequencing. Dr. Stephane Flibotte performed the sequencing alignment to generate read counts. I performed all data analysis for the differential gene expression analysis with input from Dr. Stephane Flibotte. For the hypoxia and heat experiments, I performed the temperature growth assays and the colony formation assay with input from Dr. Corey Nislow and Dr. Sunita Sinha.  v  Table of Contents  Abstract .................................................................................................................................... ii Preface ..................................................................................................................................... iv Table of Contents .....................................................................................................................v List of Tables .......................................................................................................................... ix List of Figures ...........................................................................................................................x List of Abbreviations ............................................................................................................ xii List of Symbols .......................................................................................................................xv Acknowledgements .............................................................................................................. xvi Chapter 1: Introduction ..........................................................................................................1 1.1 Introduction to the tumour microenvironment and hypoxia ..................................... 1 1.1.1 The tumour microenvironment ............................................................................. 2 1.1.2 Hypoxia within the tumour microenvironment..................................................... 3 1.1.3 Clinical significance of hypoxia ........................................................................... 4 1.2 Using chemogenomic screens with Saccharomyces cerevisiae ................................ 4 1.2.1 Saccharomyces cerevisiae as a model organism for functional genomics ........... 4 1.2.2 Chemogenomic screens using the yeast knockout collection ............................... 5 1.2.3 Modeling hypoxia in S. cerevisiae ........................................................................ 6 1.3 DNA-damaging agents as chemotherapeutics .......................................................... 8 1.3.1 Different mechanisms for DNA-damaging agents ............................................... 8 1.3.2 DNA-damaging agents in hypoxia........................................................................ 9 1.4 Thesis motivation .................................................................................................... 10 vi  Chapter 2: Characterization of the hypoxic environment .................................................11 2.1 Chemogenomic screens in hypoxic conditions ....................................................... 12 2.1.1 Homozygous pool grown under hypoxia compared to normoxia ....................... 12 2.1.2 Classification of sensitive strains using gene ontology and response signatures 14 2.1.3 Single-mutants strain validations from genome-wide screens under hypoxic stress… ............................................................................................................................ 18 2.2 Conclusions and summary ...................................................................................... 21 2.3 Materials and methods ............................................................................................ 22 2.3.1 Homozygous profiling ........................................................................................ 22 2.3.2 Generating a hypoxic environment ..................................................................... 23 2.3.3 Gene ontology analysis using Cytoscape and ClueGO....................................... 23 2.3.4 Single-strain growth profile and validations ....................................................... 24 Chapter 3: Validation of hypoxia processes, pathways and gene targets .........................26 3.1 Transcriptional profiling of wild-type cells under hypoxic conditions .................. 26 3.1.1 Experimental design of mRNA sequencing ........................................................ 28 3.1.2 Genes up and down regulated under hypoxia and/or raffinose conditions ......... 29 3.1.3 Gene expression analysis .................................................................................... 30 3.1.3.1 Hypoxia and glucose ................................................................................... 30 3.1.3.2 Normoxia and raffinose .............................................................................. 33 3.1.3.3 Hypoxia and raffinose ................................................................................. 38 3.1.3.4 Hypoxia raffinose compared to normoxia raffinose ................................... 41 3.2 Wild-type growth in heat stress and hypoxic stress ................................................ 44 3.2.1 Wild-type cell growth under hypoxia and heat stress ......................................... 45 vii  3.3 Conclusion and summary ........................................................................................ 48 3.4 Material and methods .............................................................................................. 51 3.4.1 RNA-Sequencing of wild-type cells in hypoxia and/or raffinose ....................... 51 3.4.1.1 RNA-Sequencing ........................................................................................ 51 3.4.1.2 RNA analysis .............................................................................................. 52 3.4.1.3 Gene enrichment analysis using ClueGO ................................................... 53 3.4.2 Wild-type growth in hypoxia and temperature stress ......................................... 53 3.4.3 Cell viability in hypoxia and temperature stress ................................................. 53 Chapter 4: Genome-wide screens with chemotherapeutics ...............................................55 4.1 More drug is required to generate the same level of inhibition under hypoxia compared to normoxia ........................................................................................................ 55 4.2 Genome-wide screens with DNA-damaging agents ............................................... 56 4.2.1 Genome-wide screen in hypoxia and cisplatin ................................................... 58 4.2.2 Genome-wide screen in hypoxia and doxorubicin.............................................. 64 4.2.3 Genome-wide screen in hypoxia and hydroxyurea ............................................. 67 4.2.4 Genome-wide screen in hypoxia and tirapazamine ............................................ 69 4.3 Conclusion and summary ........................................................................................ 72 4.4 Material and methods .............................................................................................. 73 4.4.1 Homozygous profiling with chemotherapeutics ................................................. 73 4.4.2 Gene ontology analysis using Cytoscape and ClueGO....................................... 74 Chapter 5: Summary and future directions ........................................................................75 5.1 Summary ................................................................................................................. 75 5.2 Future directions ..................................................................................................... 78 viii  Bibliography ...........................................................................................................................80 Appendices ..............................................................................................................................93 Appendix A :Data from chapter 2 ....................................................................................... 93 Appendix B :Data from chapter 3 ....................................................................................... 95  ix  List of Tables  Table 1 - Strains sensitive to hypoxia and their log2 ratios..................................................... 14 Table 2 - Single-mutants validated for sensitivity to the hypoxic environment. .................... 18 Table 3 –Doubling time of prefoldin subunits tested in hypoxia compared to normoxia. ..... 19 Table 4 - Strains used in this study. ........................................................................................ 24 Table 5- Experimental design of mRNA sequencing and the four conditions. ...................... 29 Table 6- Number of up-regulated genes in the three experimental conditions compared to the control condition - normoxia in glucose. ................................................................................ 29 Table 7 - Number of down-regulated genes in the three experimental conditions compared to the control condition - normoxia in glucose. .......................................................................... 29 Table 8 - List of drugs and concentrations required for 15-25% inhibition of homozygous pool growth. ............................................................................................................................ 56 Table 9 - Grouped categories for sensitive strains in chemogenomic screens. ...................... 57 Table 10 – Cisplatin chemogenomic profile. .......................................................................... 60 Table 11 - Doxorubicin chemogenomic profile. ..................................................................... 65 Table 12 - Hydroxyurea chemogenomic profile. .................................................................... 68 Table 13 - Tirapazamine chemogenomic profile. ................................................................... 70 Table 14 - GO enrichment of the 49 hypoxia-sensitive strains. ............................................. 93 Table 15 - Colonies counted for CFU viability assay. ............................................................ 95  x  List of Figures  Figure 1 - Scatterplot highlighting non-essential strains sensitive to hypoxia (<0.2% O2). .. 13 Figure 2 - Gene ontology network for 49 hypoxia-sensitive genes. ....................................... 16 Figure 3 - Growth curves of prefoldin subunits single-mutants. ............................................ 20 Figure 4 – GO Terms from 49 up-regulated genes in hypoxia and glucose. .......................... 30 Figure 5 – GO Terms from the list of 23 down-regulated genes in hypoxia and glucose. ..... 33 Figure 6 – GO Terms from the list of 704 up-regulated genes in hypoxia and raffinose. ...... 34 Figure 7 – GO Terms from the list of 533 down-regulated genes in hypoxia and raffinose. . 36 Figure 8 – GO Terms from the list of 979 up-regulated genes in hypoxia and raffinose. ...... 38 Figure 9 – GO Terms from the list of 660 down-regulated genes in hypoxia and glucose. ... 40 Figure 10 - GO Terms from the list of 212 up-regulated genes in hypoxia and raffinose compared to normoxia and raffinose. ..................................................................................... 42 Figure 11 - GO Terms from the list of 43 down-regulated genes in hypoxia and raffinose compared to normoxia and raffinose. ..................................................................................... 43 Figure 12 - Wild-type BY4743 cells grown in normoxia (18% O2) and varying temperatures.................................................................................................................................................. 46 Figure 13 – Wild-type BY4743 cells grown in hypoxia (<0.2% O2) and varying temperatures.................................................................................................................................................. 47 Figure 14 - Viability of wild-type BY4743 cells. ................................................................... 48 Figure 15 - Venn diagram describing the gene universe of all strains detected in the homozygous pool. ................................................................................................................... 58 Figure 16 – Leading GO Terms from 133 sensitive strains in hypoxia and cisplatin. ........... 62 xi  Figure 17 – Sub-network of 133 sensitive strains to cisplatin and hypoxia. .......................... 63 Figure 18 - Gene ontology network analysis of 126 hypoxia and doxorubicin specific sensitive strains. ...................................................................................................................... 66 Figure 19 – Partial Gene ontology network analysis of 98 hypoxia and TPZ specific sensitive strains. ..................................................................................................................................... 71   xii  List of Abbreviations ADP Adenosine diphosphate AIM Altered inheritance rate of mitochondria ATG Autophagy ATP Adenosine triphosphate BCL B-cell lymphoma CCT Chaperonin containing TCP-1 CHS Chitin synthase ChIP Chromatin immunoprecipitation  COG Conserved oligomeric golgi complex Co-IP Co-Immunoprecipitation COX Cytochrome c oxidase CSC Cancer stem cell CWI Cell wall integrity DGE Differential gene expression DNA Deoxyribonucleic acid dNTP Nucleotide triphosphate ER Endoplasmic reticulum  ESCRT Endosomal sorting complexes required for transport FDR False discovery rate GLM Generalized linear model GO Gene ontology xiii  HAP Heme activator protein HIF Hypoxia-inducible factor HIP Haploinsufficiency profiling  HMG 3-hydroxy-3-methyl-glutaryl  HOP Homozygous profiling IR Ionizing radiation MAP Mitogen-activated protein MDR Multi drug resistance mRNA Messenger ribonucleic acid MRPL Mitochondrial ribosomal protein, large subunit MVB Multivesicular bodies NEF Nuclear excision repair factor OD Optical density OMIM Online Mendelian Inheritance in Man ORF Open reading frame PCR Polymerase chain reaction PDAC Pancreatic ductal adenocarcinoma P-GP P-glycoprotein P-TEFb Positive transcription elongation factor b\ QCR ubiQuinol-cytochrome C oxidoReductase qPCR Quantitative polymerase chain reaction RNR Ribonucleotide reductase xiv  ROS Reactive oxygen species SGD Saccharomyces genome database TCA Tricarboxylic acid TCP T-complex protein TOP Topoisomerase TPZ Tirapazamine UFA Unsaturated fatty acid UV Ultraviolet VEGF Vascular endothelial growth factor VPS Vacuolar protein sorting YG Yeast grower YKO Yeast knockout YPD Yeast extract peptone dextrose   xv  List of Symbols °C Degrees Celsius  Δ Deletion  O2 Oxygen  xvi  Acknowledgements I would like to express my gratitude and sincerely thank Dr. Corey Nislow for the opportunity to pursue science and his endless support in my scientific endeavor. He always had his door open for me, and encouraged all of my work. I would like to thank you Dr. Guri Giaever for all of her support and countless suggestions in my research.  I would like to thank my committee, Dr. Brian Cairns, Dr. Peter Stirling, Dr. Judy Wong and Dr. David Grierson, for all of their support and expertise. Thank you to our funding support, the United States Department of Agriculture (USDA), Canadian Cancer Society Research Institute (CCSRI) and the National Aeronautics and Space Administration (NASA).  To all the members, past and present, of the Nislow/Giaever team: Sunita Sinha, Elisa Wong, Stephane Flibotte, Jennifer Chiang, Erica Acton, Amy Lee, Seth Tigchelaar, and Neira Mauricio, thank you for all of your support. Thank you Elisa, for showing me all I know about working with yeast and for pushing me to become a tough cookie. Thank you to Sunita, for all of the funny moments and countless life and science advice. Thank you Erica, for being the best office mate anyone could ask for, and pushing my limits on coffee intake.  I would like to thank all my friends for their support whenever I needed an ear for my frustration. Thank you to Jeff Yeo for sharing the graduate school experience with me, and for being a great coffee buddy.   Lastly, I would like to thank my family for their limitless love and support throughout my academic career.   1  Chapter 1: Introduction 1.1 Introduction to the tumour microenvironment and hypoxia  Solid tumours are composed of a diverse population of cells that grow within a particular environmental niche. This environment influences which cells are most fit and imposes specific selection such that the population is constantly undergoing Darwinian evolution1. Two models have been proposed to describe the development of tumours, 1) the clonal evolution model and 2) the cancer stem cells (CSCs) model1. In the former, cancer development is seen as an evolutionary process based on Darwinian natural selection2,3. This selection, combined with the inherent genetic instability of cancer cells, contributes to the development of malignancies and also to treatment resistance. The cancer stem cells (CSCs) model poses that a subpopulation of self-renewing cells generates the diverse cell population found within a tumour, and that this subpopulation continually feeds tumour progression1,4,5. Many different CSCs have been characterized, and like non-cancer stem cells, they have the ability to self-renew and differentiate. In contrast to non-cancer stem cells, CSCs fail to the down-regulate in these self-renewing and differentiation processes, which are normally strictly regulated1,4. These two models are not mutually exclusive, as is the case in the clonal evolution model, CSCs are subjected to environmental pressure as well which contributes to their tumourigenic potential1. Certain environments are beneficial to the neoplastic population, and/or possibly damaging the healthy cell population. Effects from the microenvironment on cancerous cells can come from different sources, and their interactions are important aspects to consider when developing cancer therapies.    2  My research looks to understand how the microenvironment surrounding neoplastic cells contributes to their properties, and how these challenges can affect treatments in the clinic. In particular, I examine the low-oxygen environment known as hypoxia, and how hypoxia affects the growth of cells of different genotypes in the presence and absence of chemotherapeutics.  I use the baker’s yeast Saccharomyces cerevisiae as a model organism, owing to its high degree of homology to Homo sapiens and the plethora of genetics tools that have been developed over the years. My research comprises three aims, 1) describing effects of hypoxic stress alone, identifying pathways and processes that are most sensitive to this environment, 2) validating these pathways and processes, and 3) combining the information gained from the first two aims to examine the changes in response to chemotherapeutics in hypoxia.  1.1.1 The tumour microenvironment Multiple variables make up the tumour microenvironment. For instance, an “inflammatory microenvironment” can be created within the neoplastic cell population by mutations that increase production of inflammatory mediators or by external sources such as nearby macrophages1,6. Tumour-associated macrophages have been shown to protect breast cancer cells by expressing cathepsin, a lysosomal peptidase after treatment with chemotherapies including Taxol and doxorubicin7. The importance of the nearby-stroma cell populations and their impact on tumour development has led to research to “re-educate” stroma cells in an attempt to alter their effects on neighboring neoplastic cells8. Another factor in the tumour environment is the low pH of solid tumours. This is a result of the increase in glucose metabolism within tumours,  generating more H+, and cause the tumour microenvironment to become acidic9. One model postulates that the acidic extracellular environment can induce the release of cathepsin, which promotes extracellular degradation. Furthermore, the overall acidity of tumour cells can be toxic 3  to nearby normal cells9. This neighbor effect creates an “acid-mediated invasion” at the site of the primary tumour, which promotes metastasis into nearby tissue, followed by the process of angiogenesis mediated by vascular endothelial growth factor (VEGF). The role of increased vascularization tumour progression is well-documented1,9.   My research examines hypoxia, another important microenvironment variable. The state of low oxygen is commonly observed within many solid tumours. While the inflammatory response, tumour acidity, and hypoxia are often studied in isolation, in vivo they combine to contribute to tumour biology and should therefore be considered together in cancer therapy. Hypoxia and other microenvironment factors combine to influence overlapping phenotypes, and understanding the contribution of each factor will aid in the understanding of how the environment affect therapeutic response.   1.1.2 Hypoxia within the tumour microenvironment Hypoxia describes the state of reduced oxygen levels, ranging from 0-3%1. It is caused by a combination of processes, including poorly developed vasculature, poor cell differentiation, and arterio-venous malformations4,10. Under normal physiological conditions, the level of oxygen within normal tissues ranges from 50-80mmHg (7-10% oxygen), in contrast hypoxic tumour masses oxygen ranges from 0-30mmHg (0-3% oxygen)4. Levels can vary within a tumour mass as well, creating a heterogeneous oxygen environment with pockets of high oxygen10. For example, in pancreatic cancer patients the hypoxia indicator pimonidazole estimated that the oxygen levels can vary even between adjacent surgical sections or within 100 microns or less10. This environmental heterogeneity compounds the challenges of treating tumour masses and their heterogeneous cell populations.  4  1.1.3 Clinical significance of hypoxia Patients with hypoxic tumours have a significantly increased risk of metastasis and mortality11,12. This is due in part, to an increase in hypoxic tumour cell resistance to radiotherapy and chemotherapy1,13,14. In radiotherapy, ionizing radiation (IR) produces free radicals from oxygen, which in turn induces DNA strand breaks in irradiated cancerous and adjacent normal cells. Radiotherapy and treatment efficacy are dampened in hypoxic conditions4. Drug resistance and reduced drug efficacy for a variety of drugs have been noted including cisplatin, doxorubicin, etoposide, 5-fluoruracil and docetaxel10,14. Chemotherapies that affect DNA and rely on altered DNA repair are affected by hypoxia15. While some of these drugs, such as cisplatin and doxorubicin, are routinely used in the clinic, their specific activity in hypoxia is not known. To uncover specific processes and gene products that these cells rely on to survive in hypoxia, functional genomics can be a useful experiment approach.    1.2 Using chemogenomic screens with Saccharomyces cerevisiae  1.2.1 Saccharomyces cerevisiae as a model organism for functional genomics  Saccharomyces cerevisiae (S. cerevisiae) is a single-cell eukaryotic organism that has been used for genetic studies for many years. As the first eukaryotic organism to have its whole genome sequenced, its ~6000 genes and their functional annotation provides geneticists with powerful tools to understand eukaryotic genetics. Its genetic and functional homology to humans, fast generation times and simple maintenance makes yeast particularly suitable for genetic research. Unique to yeast is the availability of the knockout collection, where single-mutants are available for every open-reading frame (ORF), each with a unique molecular barcode16. These molecular barcodes are 20 base-pair DNA sequences at the gene knockout locus that identify that strain. By combining individual single-mutants into pools, genome-wide phenotypic screens can be 5  performed in diverse biological conditions17. Early studies using the genome-wide approach characterized bud site selection18 and genes required for resistance to the K1 killer toxin19. These studies demonstrated the power in genome-wide screens to assign functional annotation to both previously known genes as well as unknown genes (for full review, see Giaever & Nislow, 201420).  1.2.2 Chemogenomic screens using the yeast knockout collection  One powerful application of the yeast knockout collection is for drug screening. Two powerful complementary assays known as haploinsufficency profiling (HIP) and homozygous profiling (HOP) have been used on hundreds of compounds to date21,22,23. In HIP, the ~1100 essential heterozygous mutants are interrogated with a specific perturbation, commonly drugs or other small molecules21. This approach can identify direct targets of the drug, by virtue of a gene dosage model, where a drug will target a gene product, and the heterozygous mutant of the gene target which grows normally in the absence of drug, will be sensitized in the presence of the drug that targets that gene product17,21. In HOP, ~4800 homozygous non-essential single gene mutants are interrogated. Here, the drug perturbation will identify processes and pathways involved in responding to the specific drug’s cellular stress, i.e. mutants deleted for members of these buffering pathways will show increased drug sensitivity17,21. When used together, HIP-HOP allows the identification of strains deleted for genes encoding the drug target (through HIP) and associated pathways and processes important in buffering drug-induced stress (through HOP). Examples of studies that have successfully identified gene drug targets include methotrexate (targeting DFR1), atorvastatin and lovastatin (targeting HMG1), benomyl ( targeting alpha- and beta tubulin), and latrunculin A targeting actin affecting actin polymerization21,24,25,26. Other 6  environmental perturbations have been tested, including UV-induced DNA damage to identify members of the DNA repair pathways27. 1.2.3 Modeling hypoxia in S. cerevisiae The hypoxic response in yeast is regulated primarily through heme biosynthesis via the heme activator protein Hap128. In the presence of oxygen, heme biosynthesis operates through a multi-step process that includes HEM13, a coproporphyrinogen III oxidase that requires oxygen in order to function29. The heme-Hap1p interaction induces Hap1p to homodimerize and then binds to specific DNA promoters, among them the downstream regulator ROX1. Rox1p mediates its regulation through transcriptional repression of specific hypoxic and anaerobic growth genes28. In anaerobic conditions when heme biosynthesis is compromised, Rox1p levels are reduced and expression of hypoxic genes occurs. The Hap family including HAP2, HAP3, HAP4, and HAP5, are also involved in heme/oxygen regulation, along with survival in non-fermentable carbon sources28,29.   Oxygen levels in yeast directly impact fatty acid metabolism, notably in the synthesis of unsaturated fatty acids (UFAs). The essential gene OLE1encodes the Δ-9 fatty acid desaturase which is required for monounsaturated fatty acid synthesis and for the normal distribution of mitochondria30. OLE1 transcription is repressed by the presence of UFAs, and is activated by low oxygen. Repression is also regulated by the action of Mga2p,  Spt23p, and endoplasmic reticulum (ER) membrane proteins.  Mga2p and Spt23p are paralogue DNA-binding proteins that are activated by ubiquitination in the ER. Once they are ubiquintinated they up-regulate a variety of hypoxic genes including OLE1.  7  The relationship between lipid metabolism and cancer cell physiology has been previously studied using yeast31. Yeast is known to sense oxygen concentration via the sterol biosynthesis pathway which requires 12 molecules of oxygen32, and therefore reduction of oxygen limits ergosterol biosynthesis. Reduction in available sterols induces expression of the regulators UCP2 and ECM22, which control expression of hypoxic genes33. Another consequence of hypoxia in S. cerevisiae is a perturbation in the uptake of sterols from the environment, and the up-regulation of ERGosterol biosynthesis (ERG) genes including ERG2 and ERG334.   In metazoans, contrast to yeast, hypoxia is regulated by hypoxia-induced transcriptional factors (HIFs)1,10,11. The roles of Hif-1 transcription regulation in the response has been well-characterized and is recognized as a major contributor to cancer progression14,35. Multiple therapeutics have been developed to target Hif-1 inhibition36,37. Although the direct oxygen sensing mechanisms differs between yeast and human, the downstream effects are similar and modeling hypoxia in yeast is intriguing given the similarities in cellular metabolism between yeast and cancer cells. For example, the Warburg effect was described in 1956 as the cellular phenomenon where, even in the presence of oxygen, cancer cells undergo glycolysis rather than oxidative phosphorylation38. While early studies postulated that cancer cells had defective mitochondria, which required them to use glycolysis to generate ATP, more recent studies have shown that cancer cells actually have functional mitochondria but that the Warburg effect reflects a metabolic re-programming to alter their energy metabolism to support tumourigenesis39,40. The classification of the “hallmarks of cancer” was recently expanded to reflect the importance of energy metabolism in cancer biology41,42.  8  S. cerevisiae is a facultative aerobe that uses glycolysis as the primary energy generation pathway even when oxygen is present. It also however, exhibits a Pasteur Effect43, where oxygen is used for respiration if it is available. These similarities between yeast and cancer metabolism suggests there may be mechanistic similarities between the hypoxic responses in cancer cells and yeast. Given the homology between the yeast and human genes (~50%), yeast can be used as a model to understand the hypoxic response. Previous work with S. cerevisiae identified ~330 mutants that demonstrated a hypoxia-specific growth defect 44  when grown on solid agar media as single colonies. These sensitive strains included transcriptional regulators such as MGA2, mRNA stability genes, and members of the CCR4-NOT complex which plays a role in transcriptional regulation44. This large-scale approach validated previous small-scale studies of the hypoxic response and identified new genes important for survival in hypoxia. In my work I use a genome-wide approach, combining hypoxia with other cellular perturbation (such as drug exposure) to expand our understanding of drug action in hypoxia in general and drug activity in hypoxia specifically.  1.3 DNA-damaging agents as chemotherapeutics 1.3.1 Different mechanisms for DNA-damaging agents  DNA damage can arise from a variety of sources. It occurs as a consequence of internal cellular processes, during replication, and as a response to a variety of environmental factors. DNA replication can introduce DNA damage that if left unrepaired, can result in mutations. Such alteration can provide the raw material for evolution45. Chronic exposure to DNA damaging conditions can lead to apoptosis46.  In fact, the induction of cell death by DNA damage is used therapeutically. This approach to combat cancer cells dates back to the discovery of nitrogen mustard compounds during the world wars47, and remains in use today. DNA-damaging agents 9  include a number of mechanistically different compounds, such as cisplatin, an alkylating-like compound that forms intra-strand crosslinks by binding primarily to guanine residues. Topoisomerase poisons, including doxorubicin and tirapazamine, induce apoptosis by inhibiting the release of DNA from topoisomerase II48,49. Others agents such as hydroxyurea affect DNA replication by altering nucleotide levels by inhibiting ribonucleotide reductase 50. While DNA-damaging agents have been important in treating cancer, their use is limited by acquired resistance and other mechanism that decrease their efficacy51. For example, these drugs are subjected to efflux by transporters such as P-glycoprotein and other multi-drug resistance transporters37. Another consideration of drug efficacy is the environmental factors affecting drug action within a tumour.  1.3.2 DNA-damaging agents in hypoxia  The environment surrounding tumour cells can provide specificity for drug therapy. Similarly, cellular pathways are important in phenotypes such as drug resistance or reduced drug efficacy. In vitro studies demonstrate that in hypoxia, different cancer cell types are more resistant to chemotherapy, including those involved in neuroblastoma, rhabdomyosarcoma, osteosarcoma, and head & neck squamous cell carcinoma14,48. Drug resistance can be gained through a variety of mechanisms, including those that reduce drug efficacy rather than affecting drug mechanism intracellularly. For example, Hif-1 targets MDR1, or P-glycoprotein, to increase drug efflux and help cancer cells remove drugs such as doxorubicin and cisplatin from of the cell52. Drugs that induce apoptosis (such as cisplatin and etoposide) have reduced efficacy due to an increase in expression of anti-apoptotic factors IAP3 and Bcl-2 proteins53. Cisplatin-specific resistance is also observed by the increase in autophagy54. An example of a direct drug mechanistic resistance gained for cells in hypoxia is seen with doxorubicin. Doxorubicin uses molecular oxygen to 10  increase reactive oxygen species (ROS) to damage DNA, proteins and membranes55, and ROS formation has been shown to be compromised in hypoxia56.  1.4 Thesis motivation Hypoxia is recognized as a tumour parameter that has a strong influence on clinical outcomes. It is evident that this microenvironment plays an important role in radiotherapy and chemotherapy resistance, and contributes to a negative prognosis and an increased risk of mortality. Therapies tailored to the hypoxic microenvironment are being developed to address the need of treating hypoxic tumours. Specifically, for chemo-resistance, what remains unclear is which cellular processes and pathways contribute to this detrimental drug phenotype.  This research project proposes using Saccharomyces cerevisiae, a single-cell eukaryotic organism with functional homology to humans, as a model organism to identify the specific genes and gene products most sensitive to the hypoxia environment. Previous work using yeast in combination with small molecules successfully identified many drug targets, and relevant conserved biological pathways. What is less well understood is the relationship between drugs, cells and the environment. The Yeast Knockout collection allows for parallel genome-wide screens to rapidly assess the differences observed between normoxic and hypoxic conditions, and of clinically relevant chemotherapies. My research takes a functional genomics approach to elucidate the biological factors that tumour cells overcome to thrive in this microenvironment.  11  Chapter 2: Characterization of the hypoxic environment  To understand the differences in drug efficacy in hypoxia versus normoxia, I first look to understand how different genotypes respond to the hypoxia environment alone. In this chapter, I identified specific strains that are sensitive to the hypoxic environment. I used a competitive growth approach to mimic the cellular environment of cells grown in hypoxic conditions. Cell growth is performed in liquid cultures within a controlled environment allowing for better control of anaerobic conditions. Testing of the hypoxic environment is done with a chemical and a biological indicator. The yeast species Kluyveromyces lactis (K.lactis) is a used as a biological control for ensuring the generation of a hypoxic environment because it cannot grow under anaerobic conditions and its growth is reduced in environments with reduced oxygen33,57.   Using the Yeast Knockout collection, specifically the ~4800 non-essential homozygous single-mutant strains, genome-wide screens were performed under both normoxic and hypoxic conditions. While genes such as HEM13 and OLE1 are essential genes and play a vital role in oxygen-sensing and response, given the robustness of the heterozygous essential gene mutants22, I predicted fewer perturbations can be detected under the subtle stress of hypoxia. Rather, I focused on the non-essential genes to capture as many different genotypes as possible with many different perturbations (such as drugs) in combination with hypoxia. Indeed, wild-type and mutant pool growth analysis demonstrated nearly identical growth profiles, suggesting almost no growth defect under hypoxia. However, the sensitivity of our homozygous deletion profiling assay allowed for the detection of subtle changes for a reproducible set of strains. The following results acts as the foundation to formulate further hypotheses on the cellular components important in cell survival in the hypoxic environment.   12  2.1 Chemogenomic screens in hypoxic conditions Using HOP assays, the non-essential homozygous pool (hereby referred to as the homozygous pool) was grown in both normoxic and hypoxic conditions. The molecular barcodes specific to each strain allow identification of the differences in growth of each individual strain between the two conditions, and strains with the greatest difference in abundance represent the genotypes most sensitive to the hypoxic environment. 2.1.1 Homozygous pool grown under hypoxia compared to normoxia  The homozygous pool was subjected to growth in both hypoxia (<0.2%) and normoxia (18%) conditions and sensitive strains were identified based on barcode abundance. The level of oxygen is detected by an OXY-SEN Oxygen Sensor built into a hypoxia chamber, monitoring oxygen levels in real time. Normoxia experiments are performed identically except in standard laboratory conditions, and the oxygen level is determined by allowing the sensor to equilibrate to ambient oxygen levels. Figure 1 depicts the 49 sensitive strains identified by comparing the growth of pooled strains in 13 hypoxia replicates versus 15 normoxia replicates. Table 1 lists the human homologs for the yeast genes as identified using YeastMine58 and the Online Mendelian Inheritance in Man59 (OMIM) database.   13   Figure 1 - Scatterplot highlighting non-essential strains sensitive to hypoxia (<0.2% O2). In red and with text are strains with a log2 ratio of 1 or higher. Log2 ratios are generated from 13 replicates of the homozygous pool grown in hypoxia and 15 replicates of the homozygous pool grown in normoxia. The blue line denotes the threshold cutoff at log2 ratio of 1 (2-fold change).        14  Table 1 - Strains sensitive to hypoxia and their log2 ratios. Human homologs were taken from the YeastMine database. Bold strains are part of the heat shock/prefoldin response signature described by Lee et al.60 Gene Log2 Ratio Human Homolog Gene Log2 Ratio Human Homolog CDC26 4.15  RSA1 1.45 NUFIP1 NPT1 3.92  TOM70 1.45 TOMM70A TOM1 3.82 UBR5 / HUWE1 VAC14 1.37 VAC14 MFT1 3.25 THOC7 BUD27 1.35 MED10 NKP2 3.00  END3 1.35  THP2 2.88  YPT6 1.26 RAB6 GIM5 2.83 PFDN5 BUL1 1.25  DBF2 2.67 DMPK LSM7 1.19 LSM7 YKE2 2.64 PFDN6 NPL6 1.17 SUZ12 RRP6 2.56 EXOSC10 POR1 1.14 VDAC 1-3 PAC10 2.53 PFDN3 SEC66 1.13  ECM2 2.47 RBM22 LRP1 1.12 C1D YML094C-A 2.27  PIH1 1.10 PIH1D1 HTL1 2.15  YOR309C 1.10  RIC1 2.03 RIC1 SEM1 1.09 SHFM1 PTC1 2.03 PPM1 YNL140C 1.07  LEA1 1.91 SNRPA1 YKU70 1.07 XRCC6 VPS61 1.87  ARC18 1.05 ARPC3 SLT2 1.84 MAPK7 CTK3 1.03  RGP1 1.83  VRP1 1.03 WIPF 1-3 MGA2 1.79  FMC1 1.01  SHE4 1.70  TMA23 1.00  VPS63 1.60  BCK1 1.00 MAP3K 1-3 BEM4 1.57 RAP1GDS1 PAT1 1.00 PATL 1-2 AAT2 1.57 GOT1     2.1.2 Classification of sensitive strains using gene ontology and response signatures I classified the sensitive strains using several analytical tools. Gene Ontology (GO) is a classification system that aggregates information and gene annotation from the literature. It includes research across many disciplines and organisms and uses a common vocabulary. GO has three over-arching categories: biological process, molecular function, and cellular component, each with specific “GO Terms”61. GO Terms in these categories represent an 15  annotation about a biological process, function or component, and each term is described by a number of genes. For example, “metabolic process” (a child term) is a downstream branch from the domain “biological processes” (a parent term). GO Terms branch from less specific to more specific as one derive more children terms, the process, function or component described is more specific, and the number of genes/term decreases. This provides a classification system to identify individual genes, and groups of genes that behave similarly, with respects to co-localization, genetic interactions, same complex membership or participation in the same biological processes or functions. GO annotation terms can group genes into networks which allows for the visualization of interactions in particular conditions. Bioinformatics tools can curate and visualize these interactions, such as GeneMania62 and Cytoscape63. I used Cytoscape in combination with the ClueGO64 to compare my 49 gene list to Gene Ontology as annotated by the Saccharomyces Genome Database (SGD)65 in Figure 2. To ensure I capture as many relevant pathways and processes as possible, I used a high P-value threshold (see Methods) and performed a right-sided hypergeometric test for gene enrichment. 16   Figure 2 - Gene ontology network for 49 hypoxia-sensitive genes. Darker coloured nodes represent lower P-values determined by a right-sided hypergeometric test (with Bonferroni correction). Network was generated from the ClueGo application in Cytoscape.  Gene Ontology analysis identified enriched GO Terms in the set of 49 genes, containing at least 5 genes from the list of 49. In Figure 2, the most significant (darker coloured) nodes are RNA-related processes such as mRNA processing, RNA localization, and RNA processing. Less significant (brighter coloured) nodes terms include protein assembly and large complex biogenesis. This network analysis suggested enrichment for transcriptional regulation, ribosomal biogenesis, cell cycle and protein folding and complex assembly.  I compared the 49 genes to previous systematic genome-wide chemogenomic screens. Lee et al. characterized 3250 compounds using HIP (haploinsufficiency profiling) and HOP to identify 17  specific chemogenomic fitness signatures60. This large-scale study classified signatures into 45 fundamental small-molecule cellular responses, and outlined the relationships between individual genes under similar perturbations. Each “response signature” is a list of single-mutant strains that behaved similarly when perturbed with that particular signature’s stresses (represented by compounds that are specific to each response). Such genes within a response class are described as “co-fit”. For example, the “heat shock/prefoldin” stress response includes 36 genes that are co-fit when perturbed by heat (or other stresses) on genes that encode components of the prefoldin protein complex. I cross-referenced the 49 hypoxia-sensitive genes across all response signatures, and found the hypoxia-specific list  had strains belonging to the “heat shock/prefoldin” response signature60. This result suggests that the hypoxia induced a stress similar to heat.  The processes identified by GO enrichment analysis are consistent with the literature on the hypoxia response; in both yeast and human cells the majority of cellular adaptation occurs at the level of transcription to hypoxic stress28,66, and therefore such mutants will be sensitive. Examples include MFT1, encoding a member of the THO complex which functions in transcription elongation, mRNA transport and mitotic recombination67, MGA2, encoding an endoplasmic reticulum -membrane protein and known transcription factor responsible for OLE1 regulation during hypoxic stress29, and SLT2, encoding a serine/threonine MAP kinase involved in cell wall integrity, cell cycle progression and nuclear mRNA retention during heat shock68. These results point to the processes important for cells to survive hypoxic stress.  18  2.1.3 Single-mutants strain validations from genome-wide screens under hypoxic stress To confirm strains identified in the genome-wide screens, I selected individual single-mutants from my list to validate sensitivity to hypoxia. The network analyses in Figure 2 allowed me to formulate hypotheses to test the relationship between cellular responses and the gene deletion strains identified. Table 2 lists the single-mutants that were tested. False positives and true positives were identified by growth curve analysis. I focused on the prefoldin complex, where multiple subunits of this hexameric complex were identified as sensitive to hypoxia. This complex is a cytosolic chaperone involved in protein folding, most notably tubulin and actin monomers in both yeast and humans68. While some evidence has suggested that protein folding is affected by low-oxygen conditions69, no specific association between the prefoldin complex and hypoxia has been reported. GIM5, aka PFD5, was identified from the hypoxia screen with a high log2 ratio (2.83). Other members of this complex identified included YKE2/PFD6 and PAC10/PFD3. To evaluate if other members of the prefoldin complex are sensitive to hypoxia, I tested all members of the prefoldin complex by growth curve.   Table 2 - Single-mutants validated for sensitivity to the hypoxic environment. Average generation ratio (Avg_G ratio) values are calculated using the YeastGrower software, controlled to wild-type BY4743 grown under the same environmental condition (normoxia or hypoxia). Avg_G Ratio values are averaged across 6 replicate wells.  Gene Normoxic Average Generation Ratio Hypoxic Average Generation Ratio AAT2 0.70 0.56 GIM5 0.80 0.75 TOM1 0.95 0.56 19  Gene Normoxic Average Generation Ratio Hypoxic Average Generation Ratio YKE2 0.92 0.92 RRP6 1.10 1.10 BCK1 0.83 0.80 SLT2 1.04 1.05 PAC10 0.87 0.33  Table 3 –Doubling time of prefoldin subunits tested in hypoxia compared to normoxia. Average doubling times are calculated across three replicate wells.  Relative doubling times are hypoxia doubling times/normoxia doubling times.  Subunit Normoxia Doubling Time (minutes) Hypoxia Doubling Time (minutes) PFD1 132 129 (0.98x) PFD2 117 143 (1.22x) PFD3 146 445 (3.05x) PFD4 175 129 (0.73x) PFD5 161 316 (1.96x) PFD6 146 401 (2.75x)  20   Figure 3 - Growth curves of prefoldin subunits single-mutants. In blue are mutants grown under normoxia and in red for growth under hypoxia.  Figure 3 shows the growth curves of prefoldin subunit single-mutants. As predicted from our genome-wide screen, pfd3Δ/Δ, pfd5Δ/Δ, and pfd6Δ/Δ had severe growth defects in hypoxic conditions. While PFD1, PFD2 and PFD4 did not show sensitivity in the genome-wide screen, their single-mutants demonstrated minor but reproducible growth defects.   No literature observations exist regarding an association between this protein-folding complex and hypoxia. Studies have however, identified different utilization of individual members of this complex in other cellular functions70. For example, pfd1, pfd5, and pfd6, have mRNA biogenesis defects for long transcripts, and these subunits are localized to the nucleus, as opposed to the canonical location in the cytoplasm70. The Pfd5 protein is recruited to actively transcribed genes and functions in transcriptional elongation. In my genome-wide screen, I identified the pfd5 and pfd6 as sensitive mutants. There may be a functional overlap between these non-canonical function and hypoxia, beyond protein folding. 21  2.2 Conclusions and summary In this chapter I identified biological processes and specific strains that are sensitive to the hypoxic environment. Using the homozygous deletion collection of ~4800 non-essential single-mutants, I identified pathways that demonstrated this sensitivity based on growth abundance. These include RNA metabolism, cell cycle, cytoskeleton, protein folding, cell wall integrity, transcription elongation, and a similar response to heat stress, determined by Gene Ontology. In my list49 hypoxia genes, 30 have a human homolog. Two examples that are associated with cancer is AAT2 and TOM1. AAT2 encodes  an aspartate aminotransferase involved in the formation of oxaloacetate and L-glutamate, and its human homolog GOT1 has been reported to be a part of a non-canonical pathway of glutamine use for tumour growth in human pancreatic ductal adenocarcinoma (PDAC)71. Overexpression of this glutamate oxaloacetate transaminase is part of PDAC metabolism and promotes tumour growth71. TOM1, an E3 ubiquitin ligase involved in histone degradation and mRNA export in yeast, is homologous to the human gene UBR5, where it shares similar functions. Amplification of the UBR5 loci has been found in different carcinomas and results in its overexpression in breast and ovarian cancers72. These two examples of yeast and human homologs are consistent with the existence of a conserved mechanism for cell survival in hypoxia. While these genes are non-essential in yeast (in rich media and normoxic conditions), deletion of these gene products causes an observable growth defect in hypoxia. In contrast, overexpression of these genes in mammalian cells can allow the cell to thrive in neoplastic populations. These observations suggest that in both cases, gene dose plays a role in the hypoxic response. To further explore the roles of hypoxia-related genes, I focused on the relationship between heat stress and hypoxia. The heat shock factor 1 (HSF-1) controls the majority of proteins responsible 22  for heat stress adaptation. Hsf-1 expression is regulated by Hif-1, where Hif-1 binds directly to the Hsf-1 promoter and up-regulates its transcription73,74. In mammalian cell culture, the chaperone Hsp90 stabilizes Hif-1, creating a positive feedback loop in hypoxic conditions74. Clearly there is a relationship between heat stress and hypoxic stress in mammalian cell, which prompted me to explore this phenomenon in yeast, which shares many well-conserved HSP genes. In previous chemogenomic screens, different strains are sensitized to heat shock and this includes the prefoldin complex. The 6 prefoldin subunits in addition to 30 strains make up the heat shock/prefoldin response signature. I examined the prefoldin subunits identified (PFD3, PFD5, PFD6) along with the subunits not identified (PFD1, PFD2, PFD4). I confirmed the subunits found in my screen were sensitive as individual mutants. Furthermore I confirmed growth defects for deletion strains corresponding to other (albeit of less severity) for the other three subunits of the prefoldin complex. These observations suggest the prefoldin complex is required to survive and provides data for hypothesis-driven experiments to further explore the role of this complex.  2.3 Materials and methods 2.3.1 Homozygous profiling The genotype of the deletion collection is listed in Table 4. The homozygous deletion pools were grown for 5 generations in YPD (1% yeast extract, 2% peptone, 2% glucose) media at 700uL volumes. YPD is kept in the hypoxic environment until used for yeast growth. Experimental pool assays are performed within a plate reader and monitored by the Yeast Grower (YG) software75. Optical density (OD) at 595nm is measured every 15 minutes throughout the growth assay. Cells are harvested after 5 generations and are processed immediately or centrifuged, media removed and frozen at -20°C. Genomic DNA extraction is performed using the extraction kit YeaStar 23  (ZymoResearch). Genomic barcode PCR amplification is performed using the universal primers that flank each barcode, and PCR products are then hybridized on to a Tag4 microarray. The GeneChip Operating Software (Affymetrix)17 was used to identify intensity values for each experimental condition, and log2 ratios between conditions were calculated based on these intensity values.  2.3.2 Generating a hypoxic environment A hypoxic environment is generated within a PLAS 856-HYPO (Plas-Labs) hypoxia chamber, with nitrogen gas displacing oxygen within the enclosed airtight space. Oxygen levels are detected via an OXY-SEN sensor, and allow control down to 0% oxygen. The chamber is also equipped with a temperature control unit, allowing up to 40oC. All hypoxic experiments were at oxygen levels <0.2%. To test that an adequate hypoxic environment is achieved within the chamber, multiple controls were used. A resazurin anaerobic chemical indicator (Oxoid/Thermo Fisher Scientific) was used to evaluate a colour change in the presence of oxygen. Kluyveromyces lactis, an obligate respiratory yeast, was grown in hypoxia and normoxia. A growth defect was observed when K.lactis was grown in hypoxia. Lastly, wild-type (BY4743) cells grown in the hypoxic environment did not show the diauxic shift seen when cells are grown with oxygen after 40 hours of growth. The chamber is also equipped with a Genios Tecan reader capable of temperature control and OD readings through the YG software75, which allows for monitored yeast cell growth required for genome-wide screens.  2.3.3 Gene ontology analysis using Cytoscape and ClueGO Analysis of the 49 hits from the HOP assay was analyzed using gene enrichment from curated Gene Ontology information as downloaded by SGD (downloaded April 25th 2016)65. One-way hypergeometric analysis is performed with the 49 identified strains, against a gene universe set 24  of all strains detected via microarray. Figure 2 was generated with statistical parameters of a p-value < 0.05, a minimum of 5 genes per GO Term and hypergeometric multiple-testing correction using the Bonferroni step-down correction method64.  2.3.4 Single-strain growth profile and validations Individual strains from the Yeast Deletion collection were grown on solid media and selected after streaking from a single colony. From each colony, an overnight culture is grown until saturation, diluted in fresh media and grown for 2-3 generations before growth in normoxia or hypoxia conditions. OD values were taken every 15 minute and plotted using RStudio with the package ggplot2. Strain genotypes are listed in Table 4.  Table 4 - Strains used in this study. Abbreviation Genotype & Source BY4743 (wt) MATa/α his3Δ1/his3Δ1 leu2Δ0/leu2Δ0 LYS2/lys2Δ0 met15Δ0/MET15 ura3Δ0/ura3Δ076 pfd1Δ/Δ Isogenic to BY4743, except for pfd1::KanMX/pfd1::KanMX16 pfd2Δ/Δ Isogenic to BY4743, except for pfd2::KanMX/pfd2::KanMX16 pfd3Δ/Δ Isogenic to BY4743, except for pfd3::KanMX/pfd3::KanMX16 pfd4Δ/Δ Isogenic to BY4743, except for pfd4::KanMX/pfd4::KanMX16 pfd5Δ/Δ Isogenic to BY4743, except for pfd5::KanMX/pfd5::KanMX16 pfd6Δ/Δ Isogenic to BY4743, except for pfd6::KanMX/pfd6::KanMX16 aat2Δ/Δ Isogenic to BY4743, except for aat2::KanMX/aat2::KanMX16 tom1Δ/Δ Isogenic to BY4743, except for tom1::KanMX/tom1::KanMX16 rrp6Δ/Δ Isogenic to BY4743, except for rrp6::KanMX/rrp6::KanMX16 bck1Δ/Δ Isogenic to BY4743, except for bck1::KanMX/bck1::KanMX16 25  Abbreviation Genotype & Source slt2Δ/Δ Isogenic to BY4743, except for slt2::KanMX/slt2::KanMX16 Array ORFΔ MATa geneX::KanMX4 LYS2 his3Δ1 leu2Δ0 met15Δ0 ura3Δ0  26  Chapter 3: Validation of hypoxia processes, pathways and gene targets To build on my genome-wide screen data in hypoxia, I sought to further interrogate the identified pathways. I first looked at the transcriptional profile of wild-type cells in hypoxia and normoxia to complement the fitness profile and to understand how cells adapt to the hypoxic environment. I then focused on pathways and processes identified in Chapter 1, specifically the overlap between the hypoxia response and the heat shock response. I interrogated this by looking at wild-type cell growth under hypoxia at varying temperatures and then compared cell viability under hypoxia and heat stress versus heat stress alone.  3.1 Transcriptional profiling of wild-type cells under hypoxic conditions  Different studies have used a variety of techniques to examine the transcriptome in both yeast32,77 and human cell culture78,79,80. In mammalian cells, recent studies have used 1) expression profiling in hypoxic pulmonary disease81, 2) polysome profiling to analyze the transcription-translation relationship in hypoxia78, 3) hypoxic cardiomyocytes82, and 4) hypoxic cancer cells to analyze microRNAs and mRNAs80,83. These studies used a variety of techniques including microarray, qPCR and RNA-sequencing. Not surprisingly, each study demonstrated a large transcriptional difference in each model. Some of the affected processes include the inflammatory responses in the brain79, translational differences through compartmentalization78 and the transcriptional machinery, particularly transcription elongation84.   In yeast, Millan-Zambrano et al. identified non-canonical prefoldin subunit functions in transcription elongation, with PFD5 playing a key role70. Safronova et al. examined transcription elongation with regards to the positive transcription elongation factor b (P-TEFb)-dependent phosphorylation of Ser2, where they observed repression of transcription elongation through P-27  TEFb inhibition in hypoxic conditions84. In my genome-wide screens in hypoxia, sensitive strains involved in transcription elongation include CTK3, HTL1, MFT1, PFD5, PFD6, and THP2 suggesting the effect of hypoxia on this process is conserved. Studies in S. cerevisiae have characterized the mRNA profile in hypoxia as well. One early genome-wide analysis in aerobic and anaerobic conditions (ter Linde et al.) used microarrays to identify genes with a large fold-change in hypoxic conditions (described as >10-fold difference by microarray detection)85. They identified genes known to be down-stream targets of Rox1p activation, and proteins involved in sterol uptake (SUT3), the ATP/ADP translocator AAC3, and FET3, involved in ferrous iron uptake. In a larger study (Becerra et al.), mutants involved in transcriptional regulation in hypoxia were grown under normoxic and hypoxic conditions, and referenced to wild-type to identify changes in gene expression77. They found that deletion of rox1, rox3, and srb10 does not significantly change gene expression under aerobic conditions. This is in contrast to the hap1 mutant that showed an up-regulation of a specific subset of genes in aerobic conditions. Examining these genes, a majority of them are related to hexose transport, and Becerra et al.77 attribute this to Hap1p’s repressor activity in aerobic conditions. As expected, deletion of any of these regulators alters gene expression under hypoxia.   My enrichment analysis of the hypoxic-specific genes suggested that RNA processing and metabolism are affected by hypoxia, and previous work in the literature supports this observation. Although the gene expression phenotype in any particular deletion mutant is not definitive86, an updated transcriptome analysis is useful to provide a high quality reference data set. In this section, I characterize the mRNA expression profile, or transcriptome, of wild-type cells in hypoxia using next generation sequencing (RNA-seq).  28  3.1.1 Experimental design of mRNA sequencing  The RNA sequencing experiment was designed to build upon the chemogenomic screens. For example, the growth curve data showed there is subtle difference when wild-type and homozygous pools grow under rich media (YPD). Under conditions with high (2%) glucose, yeast’s preferred carbon source, almost all energy generation is derived from glycolysis, even in the presence of oxygen (i.e. Crabtree effect)87.   In previous transcriptional studies, cell growth in both normoxic and hypoxic conditions were designed such that glucose was limiting (usually 0.5%)77,85. Conflicting results in transcript analysis were noted by M. ter Linde et al. and DeRisi et al., the former found respiratory sugar metabolism genes showed little to no repression under anaerobic conditions compared to aerobic conditions85, while the latter noticed respiratory genes are induced by switching from fermentative to respiratory growth88. They attributed this variation to the lack of glucose. The effect of S. cerevisiae growing in different sugars and subsequent changes in metabolism has been previously described, including raffinose, which forces cells to rely more on aerobic respiration than compared to glucose89. I hypothesized that by using the less efficient, fermentable carbon source raffinose, cells will change their metabolism to more respiration-dependent processes, therefore exacerbating the stress induced by hypoxia. The control condition for this experiment was yeast grown in normoxia and 2% glucose. To capture a steady-state transcriptome profile, cells were allowed to adapt to their respective conditions, and then harvested in mid-log phase. This corresponds to 4-6 hours of growth under the outlined conditions, on a similar time scale as used by Becerra et al.  Table 5 outlines the 4 condition experimental design. 29  Table 5- Experimental design of mRNA sequencing and the four conditions. Three biological replicates were analyzed for each condition. Carbon concentrations were 2% in Yeast-Peptone (YP) media.  Normoxia Glucose (Control) Hypoxia Glucose Normoxia Raffinose Hypoxia Raffinose  3.1.2 Genes up and down regulated under hypoxia and/or raffinose conditions  Using differential gene expression (DGE) analysis overlaps (see Methods), I identified genes that were up-regulated or down-regulated in all conditions as compared to the control condition (normoxia and glucose). Table 6- Number of up-regulated genes in the three experimental conditions compared to the control condition - normoxia in glucose. Up-Regulated Normoxia (Control Oxygen Levels) Hypoxia (Experimental Oxygen Levels) Glucose (Control) Control 49 Raffinose (Experimental) 704 979  Table 7 - Number of down-regulated genes in the three experimental conditions compared to the control condition - normoxia in glucose. Down-Regulated Normoxia (Control Oxygen Levels) Hypoxia (Experimental Oxygen Levels) Glucose (Control) Control 23 Raffinose (Experimental) 533 660   Table 6 and Table 7 display the number of genes that are either up or down-regulated under their respective conditions, using a threshold of a log2 ratio change of 1 or greater and a corrected adjusted P-value of 0.05 or less. As I hypothesized based on my genome-wide screen data and a 30  review of the literature, I found subtle changes in hypoxia when the preferred carbon source glucose was used. Many more differentially expressed genes were seen in raffinose, and this number is greater when hypoxia and raffinose treatments were combined. In all three conditions, more genes were up-regulated than down-regulated. This may reflect the short time frame before RNA collection, where cells are up-regulating required genes for adaptation for growth. Cluster analysis90 identified the carbon source as a greater driver for transcript variation than hypoxia, in agreement with our observation of these treatments on cell growth.    3.1.3 Gene expression analysis  3.1.3.1 Hypoxia and glucose To identify differentially expressed genes, I took a similar approach to the chemogenomic data. Gene enrichment analyses of genes in each condition produce a different enrichment profile, allowing the characterization under each condition.   Figure 4 – GO Terms from 49 up-regulated genes in hypoxia and glucose. GO Terms are listed by significance, with the lowest adjusted P-values listed on top.  Fractions displayed are detected genes / number of genes in each GO Term.  31  Coincidentally, 49 genes were up-regulated with glucose in hypoxic conditions. The most significant GO Term was “retrotransposon nucleocapsid”. A closer inspection at this GO Term identified a series of transcripts related to the complex of retrotransposons, and transposable elements, known as Ty (Transposon yeast) DNA sequences. These transcripts are highly transcribed and are also responsible for the categorization of related GO Terms: viral life (retrotransposon nuclecapsid; viral release from host cell; viral life cycle), “DNA integration”, “endonuclease activity”, and “DNA polymerase activity”. The GO term “DNA recombination” contains genes related to meiosis and DNA repair (MEI4, DMC1, and REC8). Two of these are genes specifically involved in repair of double-strand breaks (DSBs) during meiosis. Mei4p is meiosis-specific and is involved in DSB formation during meiotic-recombination, and Dmc1p is a meiotic recombinase responsible for DSB repair, conserved in both yeast and humans.  DMC1 is a functional homolog of RAD51 work together in meiotic recombination91. It is known that RAD51 in multiple cancer cells is down-regulated in hypoxia, contributing to the repression of DNA repair92, making it intriguing that DMC1 is found to be up-regulated in hypoxia.    Biologically relevant up-regulated genes can be identified by comparing our gene list with previously published data and examination of the hypoxic chemogenomic screen results from Chapter 2. Ranking the list of 49 genes by greatest fold-change shows ANB1 is the transcript with the greatest up-regulation in hypoxia. This gene codes for the transcription elongation factor eIF-5A and is known to be expressed in hypoxic conditions32, supporting the experimental design. Another predicted target for up-regulation is HEM13 (given that oxygen represses its expression in aerobic conditions), a transcript with a log2 ratio 2.48, equivalent to a more than 3 fold-changes. Hem13p is coproporphyrinogen III oxidase and is repressed under aerobic 32  conditions via Rox1p and Hap1p.  HSP12 is also up-regulated in hypoxia. This heat shock protein is responsible for membrane re-organization and is induced under many environmental stress conditions including heat shock, oxidative stress, glucose depletion and DNA replication stress. Becerra et al. found 54 genes up-regulated in their transcriptome analysis. Interestingly, there are only two overlaps between genes that are significant in their list and mine. This discrepancy can be due to experimental difference in carbon concentration, where 0.5% glucose was used for cell growth and could force a transcriptional response earlier on. Despite the lack of overlap, we both found OLE1 and HSP12 to be up-regulated in hypoxia.   Enrichment analyses of the down-regulated genes found fewer GO Terms, which is expected given that only 23 genes were identified. Figure 5 outlines the most significantly enriched terms. Thioredoxin peroxidase activity is important in protecting cells from oxidative stress by reducing peroxides such as H2O2 to harmless products93. Specific genes in this response include PRX1, a mitochondrial peroxiredoxin, AHP1, a thiol-specific peroxiredoxin, and TSA1, a thioredoxin peroxidase. A simple explanation for this reduction in thioredoxin peroxidase activity under hypoxia could be that under non-respiratory conditions there is a decreased need for protection against oxidative stress. This is supported in a study that determined ROS production decreases in hypoxic conditions94, although other studies have shown an increase in ROS formation in hypoxia95,96. Lastly, copper ion transport is down-regulated, including CTR1, a high-affinity copper transporter that mediates copper levels. This transporter is also responsible for the influx of cisplatin in both yeast and humans97. FRE1 and FRE7, ferric reductases that are expressed in response to low iron or copper levels, are also reduced.    33   Figure 5 – GO Terms from the list of 23 down-regulated genes in hypoxia and glucose. GO Terms are listed by significance, with the lowest adjusted P-values listed on top. Displayed fraction is detected genes / number of genes in GO Term. 3.1.3.2 Normoxia and raffinose Examining the larger gene lists obtained in raffinose involved a “control” or candidate gene to evaluate the experimental setup. I specifically looked at SUC2 transcript levels in raffinose conditions because SUC2 encodes an invertase that is responsible for hydrolyzing raffinose to fructose and melibose, and mutants lacking SUC2 cannot to use raffinose for growth89. Of 704 up-regulated genes, SUC2 is up-regulated with a log fold-change greater than 4,  consistent with a functional role in raffinose metabolism. Gene function analysis can be challenging with hundreds of genes showing changes. In these cases, enrichment analyses can be a useful statistical tool to categorize and identify biologically meaningful changes. I focus on the GO Terms not with the most genes in them, but rather on those with the smallest P-value from the one-sided hypergeometric test (these two characteristics are often but not always correlated).  34   Figure 6 – GO Terms from the list of 704 up-regulated genes in hypoxia and raffinose. (A) Leading GO Terms are displayed, with the lowest adjusted P-values listed on top. Displayed fraction is detected genes / number of genes in GO Term. (B) Network of GO Terms with the leading GO Terms bolded. Bolded terms are leading GO Terms with the highest significance in their related-processes. Related GO Terms are significant with a larger P-value than the leader term.   In Figure 6, statistically significant GO Terms meeting the threshold of corrected P-value < 0.0005 from the 704 up-regulated genes in normoxia and raffinose condition are shown. Genes 35  encoding components of the mitochondria are highly up-regulated; this enrichment profile supports our hypothesis to select a carbon source that would require yeast to shift their cellular metabolism to rely more on respiratory pathways, such as raffinose. Examining the child-terms from the lead term “single-organism metabolic processes” shows many of the ribonucleoside triphosphate metabolic processes are up-regulated and represent significant GO Terms. This may reflect a cellular response to adapt to the source of energy molecules from ribose sugars carrying tri-phosphates. The term “respiratory chain” has 28 genes in my list, representing over 90% of this GO Term’s total genes including those with molecular function in “cytochrome-c oxidase activity” and “electron carrier activity”. The “oxidoreductase complex” and “oxidoreductase activity” is also up-regulated. Examining the children terms for these shows my gene list contains members of cellular respiration, and in particular genes involved in the tricarboxylic acid (TCA) cycle such as CIT2, MDH3, KGD2, and ICL1. This reinforces the idea that cells in normoxia with raffinose as a carbon source switch more to respiratory metabolism in order to maintain energy levels, up-regulating the genes involved in the electron transport chain. I detected multiple members of the Heme Activator Protein (HAP) family, including HAP2, HAP4, and HAP5. Similar to HAP1, these members of the HAP family are involved in transcriptional regulation of genes involved in heme biosynthesis and respiration. The up-regulated genes suggest raffinose is affecting oxygen levels and therefore these processes, , and may explain why these hypoxia regulators are being up-regulated.  36   Figure 7 – GO Terms from the list of 533 down-regulated genes in hypoxia and raffinose. (A) Leading GO Terms are displayed, with the lowest adjusted P-values listed on top. Displayed fraction is detected genes / number of genes in GO Term. (B) Network of GO Terms with the leading GO Terms bolded. Bolded terms are leading GO Terms with the highest significance in their related-processes. Related GO Terms are significant with a larger P-value than the leader term.  37  Down-regulated genes in normoxic and raffinose conditions show enrichment for ribosomal biogenesis and rRNA activity, including genes involved in the nucleolus, the site for rRNA transcription and maturation. This may be due to the fact that cells in raffinose are not proliferating as quickly (as seen by the slower doubling time), which would require less translational activity. Ribosomal biogenesis and translation is a very energy demanding process and this observation may reflect the cells attempting to conserve energy. This observation supports previous work on hypoxia tolerance, where genes deleted in ribosome biogenesis show greater fitness in anaerobic conditions98. 38  3.1.3.3 Hypoxia and raffinose  Figure 8 – GO Terms from the list of 979 up-regulated genes in hypoxia and raffinose. (A) Leading GO Terms are displayed, with the lowest adjusted P-values listed on top. Displayed fraction is detected genes / number of genes in GO Term. (B) Network of GO Terms with the leading GO Terms bolded. Bolded terms are leading GO Terms with the highest significance in their related-processes. Related GO Terms are significant with a larger P-value than the leader term.  39  The GO enrichment analyses between raffinose in normoxic and hypoxic conditions are similar, for both the up and down-regulated gene lists. Genes for respiratory chain with cytochrome and electron carrier activity remains up-regulated, along with oxidoreductase activity. Looking at the GO Terms with the most significance, a term that appears in the hypoxia list but not the normoxia list is “peroxidase activity”. The normoxia/raffinose up-regulated list contains 8/17 members from this GO Term, while hypoxia increases this list to cover 12/17. All members of the peroxidase activity GO Term are involved in reducing reactive oxygen species. In raffinose with increased respiration, there is a functional requirement for these genes. In particular, three of the hypoxia-specific up-regulated genes from this GO Term handle hydrogen peroxide sensing and reduction (CTA1, CTT1, and HYR1). HYR1 is a paralogue of GPX1, and while HYR1 is specific to hypoxia, GPX1 is up-regulated in both normoxic and hypoxic conditions with raffinose. Gpx1p is a lipid peroxidase and Hyr1p is a thiol peroxidase with hydrogen peroxide signaling functionality. This up-regulation in peroxidase activity is contrary to the down-regulation of thioredoxin peroxidase activity in hypoxia/glucose conditions. Similar to multiple studies, the increase or decrease of ROS formation differs between different studies and models94,95,96. One explanation for this discrepancy in my experiment may be the usage of carbon source in hypoxia. Studies that found an increase in ROS formation attribute this to an active electron transport chain95,96. In hypoxia/glucose conditions, cells are not relying on respiration for energy and therefore ROS production should decrease. In raffinose and respiratory metabolism, requirement for peroxidase activity increases with an active ETC, supported by up-regulation of ETC genes.   40   Figure 9 – GO Terms from the list of 660 down-regulated genes in hypoxia and glucose. (A) Leading GO Terms are displayed, with the lowest adjusted P-values listed on top. Displayed fraction is detected genes / number of genes in GO Term. (B) Network of GO Terms with the leading GO Terms bolded. Bolded terms are leading GO Terms with the highest significance in their related-processes. Related GO Terms are significant with a larger P-value than the leader term.  The enrichment profiles for down-regulated genes between the two environmental variables in raffinose are nearly identical with respect to significant GO Terms. The cellular component 41  nucleolus, along with Ribosomal RNA and ribosomal biogenesis remains to be highly down-regulated.  3.1.3.4 Hypoxia raffinose compared to normoxia raffinose To examine the hypoxia-specific gene regulation in raffinose, I use the normoxia/raffinose condition as my control and compare hypoxia/raffinose. As anticipated, this reduced the number of regulated genes in hypoxia/raffinose, and kept in trend with more up-regulated (212) genes than down-regulated (43) genes. Figure 10 shows the enrichment profile for up-regulated genes. In the GO Term “monocarboxylic acid metabolic process” are genes involved in fatty acid metabolism and import, including FAA2, YAT2, POX1, POT1, and FOX2. The GO Term “sexual sporulation” contains genes related to sporulation and meiosis function or has been annotated to be expressed during sporulation. These include SPO23, SPO24, SPO74, SPR28, SPR3, SPS1, and SPS100. The genes up-regulated here are specific to hypoxia rather than raffinose metabolism, and complements the few genes found up-regulated in glucose conditions that are related to sporulation (MEI4, DMC1).  Figure 11 shows the enrichment profile for the 43 down-regulated genes in hypoxia/raffinose compared to normoxia/raffinose. Similar to glucose conditions, ion transport is down-regulated including the same membrane proteins, Ctr1p, Fre1p, and Fre7p. These proteins are involved in ion regulation including copper and iron. Iron regulation is further down-regulated in raffinose conditions with FET4 and MRS4.  42   Figure 10 - GO Terms from the list of 212 up-regulated genes in hypoxia and raffinose compared to normoxia and raffinose. (A) Leading GO Terms are displayed, with the lowest adjusted P-values listed on top. Displayed fraction is detected genes / number of genes in GO Term. (B) Network of GO Terms with the leading GO Terms bolded. Bolded terms are leading GO Terms with the highest significance in their related-processes. Related GO Terms are significant with a larger P-value than the leader term.   43   Figure 11 - GO Terms from the list of 43 down-regulated genes in hypoxia and raffinose compared to normoxia and raffinose. (A) Leading GO Terms are displayed, with the lowest adjusted P-values listed on top. Displayed fraction is detected genes / number of genes in GO Term. (B) Network of GO Terms with the leading GO Terms bolded. Bolded terms are leading GO Terms with the highest significance in their related-processes. Related GO Terms are significant with a larger P-value than the leader term.  44  3.2 Wild-type growth in heat stress and hypoxic stress In the next section, I focus on the overlap seen in hypoxia and heat stress. This observation is of interest because it is known that certain heat shock response proteins play a role in cancer biology. As previously noted, up-regulation of Hsf-1 is regulated through Hif-1 binding at promoters in human genes74. Hsf-1 has pro-oncogenic functionality in cancer cells, perhaps enabling them to avoid proteomic instability99.   In yeast, the heat shock response is well characterized68, but there are no published links between heat and hypoxic stresses. Many of the genes in the heat shock response pathway are conserved between human and yeast. For example, human Hsp90 is encoded by yeast HSC82 and HSP82. Studies in yeast have helped develop Hsp90 inhibitors47. The heat shock response in yeast has specific phenotypes that include, metabolic reprogramming, altered cell wall dynamics and protein aggregation68. Certain genes identified in my initial genome-wide survey related to these processes. For example, the mutant slt2Δ/Δ undergoes autolysis at temperatures around 37°C100. This mutant lacks the mitogen-activated protein (MAP) kinase activity that is responsible for cell wall integrity and signaling, and demonstrated sensitivity under hypoxic conditions. PIH1is a component of the R2TP complex, which works with prefoldin-like proteins to interact with HSP90 to promote RNA polymerase II assembly in human cells101. Interestingly, PIH1 and the prefoldin-like BUD27 were both strains identified in my hypoxia chemogenomic screen. This relationship between genes involved in the transcriptional response and HSPs could connect the relationship between heat and hypoxic stress. Another cellular response that connects our gene list with the heat shock response is P-body formation. P-bodies are cytoplasmic foci that form during stress and which are responsible for mRNA degradation, particularly histone mRNA and 45  under heat stress102. P-bodies have been observed in DNA replication stress, for example, induced by hydroxyurea103. Relevant genes in my list of hypoxia-sensitive strains include LSM7 and PAT1.  3.2.1 Wild-type cell growth under hypoxia and heat stress To begin interrogation of heat stress and hypoxia, I grew cells at 37°C to observe any synergistic growth defects. Yeast grow optimally at 30°C but will grow well at 37°C, even strains deleted for the chaperone HSP90 show almost no growth defect at 37°C 104. In fact, yeast that are shifted to 37°C are capable of mounting a transcriptional stress response105, and maintain growth up to 42°C106. It was therefore a surprise to observe a complete lack of cell growth at 37°C in hypoxia. To further explore this phenotype, I titrated the temperature down to see at which point growth recovered. Figure 13 shows that the wild-type growth defect depends on the temperature in hypoxia. For comparison, normoxic controls show that the temperature defects are minimal under ambient oxygen conditions (Figure 12).  46   Figure 12 - Wild-type BY4743 cells grown in normoxia (18% O2) and varying temperatures. Cells are grown until saturation. Average OD values are calculated by taking an average across 12 replicate wells. 47   Figure 13 – Wild-type BY4743 cells grown in hypoxia (<0.2% O2) and varying temperatures. Cells are grown until saturation or for 25 hours. Average OD values are calculated by taking an average across 12 replicate wells.   I next asked if the cells stressed in hypoxia and heat are still viable i.e. is the combination of heat and hypoxic stress to a cell lethal or does it simply avert cell growth. To evaluate this, I performed a viability CFU count assay with cells following hypoxia and stress at the lethal temperature of 37°C. Results from Error! Reference source not found. suggests that while ecovery is slower in cells stressed in hypoxia and at 37°C, there is still an appreciable amount of growth post-stress.  An explanation for the lack of growth in hypoxia at elevated temperatures may be a growth arrest mechanism, rather than cell death.  48   Figure 14 - Viability of wild-type BY4743 cells. Cells are stressed for 1 hour or 4 hours and assessed for viability through colony formation units. Cells were stressed either 37°C in normoxia, 37°C in hypoxia or unstressed at 30°C, then plated on to 2% YPD agar plates and recovered for 2 days at 30°C. Colonies are counted and log(CFU/mL) values are calculated for each condition. Error bars are standard deviation of n=3. ANOVA analysis indicates no significant changes between conditions.    3.3 Conclusion and summary In this chapter, I explored the overall mRNA transcriptome of wild-type cells in hypoxia using RNA sequencing. I complement my findings in Chapter 2, reinforcing the idea that with the preferred carbon source (glucose) the hypoxic environment represents a specific cellular stress and does not require cells to alter their metabolism. This is reflected in the small number of transcripts that are up or down-regulated in hypoxia/glucose conditions (49 and 23 respectively). In the processes that do change, I observe genes involved in double-strand break repair activity 01000000020000000300000004000000050000000600000007000000080000000900000001 Hour Stress 4 Hours StressCFU/mL Colony Formation Units of Hypoxic & Heat Stressed Wild-Type Cells 30°C + Normoxia37°C + Normoxia37°C + Hypoxia49  being up-regulated and thioredoxin peroxidase activity being down-regulated. DMC1 is a meiotic recombination protein involved in DSB repair that is expressed in early meiosis107. Its up-regulation in hypoxia suggests a requirement for DSB repair, or some cells are entering meiosis. Down-regulation of thioredoxin peroxidase activity likely reflects the reduced level of reactive oxygen species in hypoxia in conditions where cells do not rely on respiratory metabolism. Another set of down-regulated genes included ion transporter, specifically CTR1, FRE1 and FRE7. These Ferric REductase family genes are involved in iron uptake and their expression is decreased in low copper environments. Ctr1p is a high-affinity copper transporter that is induced at low copper levels and during DNA replication stress. The human homolog SLC31A1 can complement the yeast version, suggesting its functions are conserved. Interestingly, in yeast and human this copper channel mediates influx of cisplatin, and deletion of CTR1 confers cisplatin resistance in yeast97. Down-regulation of CTR1 is observed in hypoxia/raffinose, but not under normoxic conditions. This down-regulation of a drug import channel provides a distinct mechanism for reduced drug efficacy in hypoxia.  I hypothesized that in raffinose, which is a less-efficient fermentable carbon source, cells will shift their metabolism to respiration pathways. By RNA-sequencing, the data shows a greater number of genes regulated in the presence of raffinose, in both normoxia and hypoxia. The enrichment profiles for the up-regulated genes in both environments and (+ raffinose) suggests more mitochondria activity, purine nucleoside metabolism and cellular respiration. In hypoxic conditions with raffinose, I observe an increase in peroxidase activity, reflecting the increase ROS formation under respiration. This is in contrast to down-regulation of peroxidase activity in the presence of glucose, reflecting the difference between carbon source and overall cell 50  metabolism. Overall down-regulation of ribosomal RNA and ribosomal biogenesis is observed with raffinose present in both normoxia and hypoxia. This reflects the conservation of energy for energy-demanding processes such as ribosomal biogenesis and translation, a response observed in the environmental stress response108. Similar to many othe r stresses (heat, osmolarity, drugs), growth in raffinose induces this reduction in ribosomal activity. This is reinforced in comparing gene regulation in hypoxia/raffinose conditions to normoxia/raffinose, where the apparent down-regulation of ribosomal activity is not observed. This is because in both conditions of raffinose, ribosomal activity must be similar and is not detected in DGE analysis.  Comparing hypoxia gene regulation with normoxia in raffinose showed an up-regulation in fatty acid metabolism and sporulation activity. An explanation for this up-regulation in fatty acid metabolism is the cell attempting to maximize efficiency of respiration and the production of subtrates for respiration metabolism. This effect is not detected in conditions with glucose, where cells use glucose through glycolysis for energy production. Sporulation activity and related genes were minimal in conditions with glucose, with MEI4 and DMC1 being up-regulation. These genes increase in raffinose conditions, suggesting that hypoxia may induce cells to sporulate in conditions with hypoxia and raffinose.     I demonstrated a heat-dependent growth defect for cells grown in hypoxia; at 37°C cells cannot grow in hypoxia while reducing the temperature towards 30°C restores growth. A colony formation assay shows that cells stressed under hypoxia + 37°C for four hours are still viable, 51  showing that the lack of cell growth in hypoxia and 37°C is not due to cell death, but due to growth arrest.   3.4 Material and methods 3.4.1 RNA-Sequencing of wild-type cells in hypoxia and/or raffinose 3.4.1.1 RNA-Sequencing RNA-sequencing was performed with wild-type (BY4743) cells. Each biological replicate was streaked to obtain individual colonies from a 25% glycerol stock on to 2% agar YPD (2% glucose) plates. Single colonies were grown for 2 days and stored at 4°C for up to 1 week for RNA experiments. A single colony is picked into a 4mL overnight culture (2% YPD) and grown to saturation. Experiments are performed by back-diluting this saturated culture to an OD of 0.1 in 700uL in their respective carbon source (2% glucose or 2% raffinose), and grown in their respective environments (normoxia or hypoxia). Cell growth is monitored until mid-log phase (between 4-6 hours). 700uL of cells are collected by centrifugation and RNA is extracted immediately using the RNeasy Mini Kit for yields roughly between 100-500ng/μL. RNA quality was assessed on a 2% agarose gel to check for 28S and 16S band intensities and for minimal DNA contamination. A NanoDrop spectrophotometer is used to determine 260/230 and 280/260 ratios. For both, a ratio of ~2.0 is acceptable for RNA. The RNA samples are then quantified using Qubit RNA HS Assay Kit (Thermofisher), which uses a fluorescent intensity dye, to measure RNA concentration without interference of DNA, protein, or free nucleotides. 500ng of RNA was used as input for library preparation using the Illumina TruSeq mRNA Stranded library preparation kit. With this method, mRNA is first captured using a poly-A tail capture and then reverse transcribed. After second strand synthesis, adapters are ligated which contain the Illumina flow-cell binding sequences and unique indices to enable sample multiplexing. 52  Libraries were pooled and sequenced on a HiSeq2500 (Illumina), to generate paired-end 100 bp reads. 3.4.1.2 RNA analysis 12 samples were processed in two batches for three biological replicates in each condition. After quality control, reads were aligned using the STAR aligner109 to the S288c reference genome from SGD (downloaded September 22nd 2016)65 to produce a raw count matrix of 7126 transcripts. Counts were used for two separate differential gene expression (DGE) analysis: DESeq290 and edgeR110. Raw counts were filtered using edgeR’s “count per million” (cpm) function that tally total reads and removes transcripts with less than five cpm. This reduces our total transcript count to 5519 transcripts.  Analysis with DESeq2 was performed using the “DESeq” function, which estimates the size factors of each variable relative to all samples, estimates the dispersion and then fit a negative binomial generalized linear model (GLM) and tested for significance90. A model design matrix for the meta-data is designed with each condition: carbon source (glucose or raffinose), oxygen (normoxia or hypoxia), and batch (first batch or second batch processed). Each condition is then compared to the control group (normoxia / glucose) for a log2 fold change, standard error and adjusted P-values. Analysis with edgeR was performed with the raw counts and running the “DGEList” function on the count matrix with meta data outlining carbon source, oxygen and batch110. Size-corrected normalized factor are calculated, then dispersion estimation. Direct comparisons between groups were made relative to the control group (normoxia / glucose) after fitting a GLM. Results are reported with a log fold-change and false discovery rate (FDR).   53  Lists of up-regulated and down-regulated transcript lists were then cross-referenced, filtering in both packages transcripts with a log2 fold change of 1 or greater, and an adjusted P-value or FDR less than 0.05. A transcript is classified as up or down-regulated in my final list if it passes the criteria in both packages.  3.4.1.3 Gene enrichment analysis using ClueGO Similar to previous gene enrichment analysis, up and down-regulated gene lists are analyzed in Cytoscape + ClueGO against the GO from SGD. The gene universe of detection was defined as the 5519 transcripts that passed filter before the DGE. Filtering criterion was determined by a minimum count of 5cpm in at least one sample. Statistical thresholds were set at Bonferroni-corrected P-value < 0.0005 and GO Terms were required to contain at least 5 genes. GO Term enrichment bar plots were created in RStudio with the package ggplot2111.  3.4.2 Wild-type growth in hypoxia and temperature stress  BY4743 cells were grown in an overnight culture until saturation. Cells are then back-diluted to OD 0.1 and grown for ~6 hours for 3-4 generations, and then placed in their respective environments: hypoxia + heat or normoxia + heat. Cell growth is tracked by YG and OD readings are taken every 15 minutes. Growth curves are plotted in RStudio using the package ggplot2.  3.4.3 Cell viability in hypoxia and temperature stress BY4743 cells were grown in an overnight culture until saturation. Cells were then diluted to an OD of 0.1 in 700uL with multiple replicate wells and placed in a Genios Tecan reader in their respective conditions: normoxia/30°C, normoxia/37°C, or hypoxia/37°C. Cells were stressed for 1hour or 4 hours. At each time point, 700uL of cells were harvested and diluted to OD 0.1, a 10-1 dilution, in ambient laboratory conditions. Cells were then diluted 10-fold 3 times. For the four 54  dilutions, 100uL were plated on agar 2% YPD plates, and spread using glass beads. Cells were recovered for 40 hours at normoxia/30°C, and then counted for colonies. Only plates with 25-300 colonies were used for counting, as variation increases beyond this range. Colony formation units were calculated by:                                         The dilution used was either 10-4 or 10-5, depending on which plate had 25-300 colonies. The volume used was 100μL per plate.   55  Chapter 4: Genome-wide screens with chemotherapeutics  In this chapter, I combine the genome-wide screens in hypoxia with clinically relevant chemotherapeutics. The goal is to identify processes and individual strains that demonstrate sensitivity to either the hypoxic environment alone or when combined with a drug. Each drug is screened against the homozygous deletion collection in both normoxia and hypoxia, and then compared back to the control (normoxia and no drug). I take advantage of the previous work that screened over 3200 compounds using the YKO collection, providing a profile for each drug in normoxia and acts as a reference for my own screens60. I also compare and contrast the similarities and differences between the profiles for each drug in relation to their specific mechanisms. 4.1 More drug is required to generate the same level of inhibition under hypoxia compared to normoxia I begin by selecting a drug concentration that generates an inhibition level between 15-25% (IC15-25) relative to the homozygous pool grown without the drug in the same oxygen environment (for instance, a drug concentration that inhibits 15-25% of pool growth under hypoxia, relative to pool growth without the drug under hypoxia). To achieve the same level of inhibition required 1.1–6 times the drug dose in hypoxic conditions, except for tirapazamine, a hypoxic-specific compound. This is consistent with literature reports of increased drug resistance and decreased drug efficacy in hypoxic conditions, particularly DNA-damaging agents14. These compounds take advantage of rapidly proliferating cells and are less effective when replication is slowed in hypoxia. Table 8 describes the drugs evaluated and their respective inhibitory concentrations. 56  Table 8 - List of drugs and concentrations required for 15-25% inhibition of homozygous pool growth. Fold-change comparison of concentrations between hypoxia and normoxia are also listed. In each drug except the hypoxia-specific Tirapazamine, a higher concentration of drug is required to achieve the same inhibition in hypoxia than in normoxia.  Drug  Normoxia IC15-25 Hypoxia IC15-25 Concentration Fold Increase Hydroxyurea 30 mM 32.5 mM 1.1x Cisplatin  200 μM 300 μM 1.5x Doxorubicin  8.5 μM 50 μM 6.25x Tirapazamine  15 μM 7.5 μM 0.5x Benomyl 30 μM 50 μM 1.66x  4.2 Genome-wide screens with DNA-damaging agents I focused on four drugs: cisplatin, doxorubicin, hydroxyurea, and tirapazamine. For each drug, I identified sensitive strains specific to each condition, generating condition-specific gene lists. I sorted and binned the strains based on which condition they were sensitive to as outlined in Table 9. Figure 15 shows the overlaps between conditions: hypoxia, normoxia and drug, or hypoxia and drug. Certain strains are shared between two or more groups. For example, in some cases a strain will demonstrate sensitivity in both environments in a particular drug. Such strains are likely involved in buffering drug-induced stress. Another group of strains are sensitive exclusively to the hypoxic environment with drug. These are strains that may be “nearly-sensitive” to drug in normoxia, and the additional environmental stress of hypoxia potentiates the drug effect.  57  Table 9 - Grouped categories for sensitive strains in chemogenomic screens. Every strain belongs to one specific category. A "+" symbol denotes strains sensitive in that group's comparison to the normoxia no-drug control. Each group represents a unique gene list encapsulating all strains detected in the homozygous pool.  Hypoxia Alone  Normoxia Drug  Hypoxia Drug  Group Characteristic + + - Sensitive to Hypoxia & Normoxia Drug + - + Sensitive to Hypoxia & Hypoxia Drug + - - Sensitive to Only Hypoxia - + - Sensitive to Only Normoxia Drug - + + Sensitive to Drug Exclusively  - - + Sensitive to Drug with Hypoxia + + + Sensitive to All Conditions - - - Background   For each chemotherapeutic I screened against the homozygous pool, I create a unique drug-table with sensitive strains. I focused on the hypoxic-drug exclusive list, where there is the greatest number of sensitive strains, and performed enrichment analysis for these lists to compare the specific strains and biological processes affected.       58   Figure 15 - Venn diagram describing the gene universe of all strains detected in the homozygous pool. Strains sensitive to any of the conditions will fall within a category. Certain strains will demonstrate overlap between two or all three conditions. Background strains exist outside the group circles.  4.2.1 Genome-wide screen in hypoxia and cisplatin  Cisplatin is a common chemotherapeutic used to treat a variety of cancers including head, neck, bladder, and lung. Its mechanism of action involves cross-linking DNA and disrupting DNA replication, and eliciting a DNA repair response112. If drug-induced DNA damage cannot be repaired, apoptosis is induced. Cisplatin efficacy is reduced in hypoxia, presumably due to 59  reduced cell proliferation. Another process that cells engage when treated with cisplatin is autophagy, a “self-eating” mechanism that recycles molecules such as amino acids and fatty acids113. In a cell culture model, autophagy-related genes are up-regulated in hypoxia and contribute to the reduced effectiveness of cisplatin therapy through suppression of the pro-apoptotic protein BCL2 Interacting Protein 3 (BNIP3)54.   Cisplatin’s broad use and efficiency has motivated the search for other platinum-based drugs, and such related therapeutics are routinely used, for instance, carboplatin for neck and head cancers 112.  Some of the goals for these derivative compounds include better bioavailability and lowered toxicity. S. cerevisiae and the yeast deletion collection has been used to identify similar compounds23. The cisplatin chemogenomic screen profile has a unique signature of the RADiation sensitive family deletion strains, important in DNA repair processes such as nucleotide excision repair, base excision repair, homologous recombination and double-strand break repair. Strains deficient in these pathways cannot effectively repair this damage and thus do not survive the cytotoxicity. I screened cisplatin in both normoxia and hypoxia and compared the respective strains sensitive in either condition.      60  Table 10 – Cisplatin chemogenomic profile. In red are strains a part of the DNA repair response. Normoxia Cisplatin (26) Hypoxia Only (10) Hypoxia AND Hypoxia Cisplatin (36) Normoxia AND Hypoxia Cisplatin (21) Hypoxia Only AND Normoxia Cisplatin (1) Hypoxia Cisplatin Only (131) All Conditions (1) ADE2, CHS5, CIR2, CKA1, EMI5, ERG4, HAP5, LDB19, LOS1, MIP6, OAC1, OCA1, PAC1, PHO86, PRO2, RIM101, RRT6, SAS3, SOD1, TCB1, YEL028W, YER066C-A, YER186C, YIP4, YLR217W BUL1, LRP1, LSM7, PIH1, POR1, RSA1, SEM1, TMA24, YNL140C, YOR309C AAT2, ARC18, BCK1, BEM4, BUD27, CDC26, CTK3, DBF2, ECM2, FMC1, GIM5, LEA1, MFT1, NKP2, NPL6, NPT1, PAC10, PAT1, PTC1, RGP1, RIC1, RRP6, SHE4, SLT2, THP2, TOM1, TOM70, VAC14, VPS61, VPS63, VRP1, YKE2, YKU70, YML094-C, YPT6 FAR11, EOS1, HPR5, IMP2, MMS4, MUS81, RAD1, RAD2, RAD4, RAD5, RAD10, RAD14, RAD18, RAD51, RAD54, RAD55, RAD57, RAD59, RVS167, XRS1  MGA2  HTL1  Table 10 lists all strains that demonstrated sensitivity in at least one of the experimental conditions: hypoxia alone, normoxia and cisplatin, or hypoxia and cisplatin. The 49 strains identified in our hypoxia-alone experiment are present, with 37/49 strains continuing to demonstrate sensitivity in the presence of cisplatin and hypoxia. 10/49 hypoxia sensitive strains are not sensitive to any drug conditions, and the last two were also sensitive in either normoxic/cisplatin (MGA2) or both drug conditions (HTL1). This interesting relationship for 61  MGA2, which, according to my threshold for sensitivity of log2 > 1 relative to the normoxia/no drug control, demonstrates sensitivity in hypoxia and normoxia/cisplatin, but is not sensitive in hypoxia/cisplatin. This is a unique interaction between cisplatin and the environment. However, the log2 ratio value for MGA2 in hypoxia/cisplatin is 0.83, which would be eliminated by my filtering criterion. Nonetheless, it does show a relative decrease in abundance under hypoxia/cisplatin. Indeed, in an individual strain growth assay of the mga2Δ/Δ mutant dosed with 300μM of cisplatin (the same concentration as in the hypoxia HOP assay) I observed sensitivity in hypoxia relative to the mutant in normoxia/no drug conditions. This suggests in my chemogenomic screen, the MGA2 is a false negative and is in fact sensitive.  Strains sensitive in both normoxia/cisplatin and hypoxia/cisplatin represent strains specific to cisplatin. The RAD family of genes, for example, is in this list. These DNA repair strains show growth defects in any environment if cisplatin is present. Other cisplatin-specific genes include MMS4 and MUS81. These observations complement previous studies that used cisplatin (in addition to other platinum-based drugs) using the HOP chemogenomic approach23. The largest list from Table 10 is those strains (131) that are sensitive in hypoxia/cisplatin. The enrichment of GO Terms displayed in Figure 16 is generated from analyzing strains that appear sensitive only in hypoxia and cisplatin, and do not demonstrate sensitivity under cisplatin and normoxia. This criterion filters out genes typically seen in cisplatin, such as the RAD genes. One of the top GO Term from this set of 131 strains is “cellular response to DNA damage stimulus”. Looking at the 29 genes listed under the GO Term “DNA Repair” in this hypoxia/cisplatin list, I identify additional RAD genes such as RAD23 and RAD27, UBC13 and MMS1/MMS2. Rad23p interacts with Rad4p and is a part of the Nuclear Excision Repair Factor 2 (NEF2), and is also involved in 62  ubiquitylated protein turnover. Ubc13p interacts with Mms2p in post-replicative repair and together forms an active ubiquitin-conjugating enzyme. Mms1p is a subunit of the E3 ubiquitin ligase complex involved in replication repair. Other DNA repair genes include RTT101, RTT107. This suggests that hypoxia potentiates the requirement for the DNA repair, revealing sensitive strains not observed in normoxia/cisplatin.   Figure 16 – Leading GO Terms from 133 sensitive strains in hypoxia and cisplatin. GO Terms are listed by significance, with the lowest adjusted P-values listed on top. Displayed fraction is detected genes / number of genes in GO Term. Each leading GO Term showed here are branched to related significant GO Terms sharing similar function and processes.   63   Figure 17 – Sub-network of 133 sensitive strains to cisplatin and hypoxia. GO Terms displayed are related to the DNA repair response. The bolded GO Term “cellular response to DNA damage stimulus” represent the most significant GO Term in this sub-network.  Another sensitive strain is LDB7, a member of the RSC chromatin remodeling complex. It is not, however, sensitized in normoxic environments. The log2 ratio for this strain is -0.52 in normoxia/cisplatin (well below my cut-off) and it is not a hypoxic gene in our control experiments. However, in hypoxia and cisplatin conditions, LDB7 is sensitive with a log2 ratio of 1.95. NPL3 encodes a member of the RSC remodeling complex and its deletion strain is sensitive to cisplatin and hypoxia. Additional strains deleted for chromatin remodeling genes include NGG1, ASF1, SPT8, and EAF1. While it is known that cisplatin perturbs chromatin remodeling114, this interaction in hypoxia has not been well-studied and this result suggests that hypoxia can interfere with chromatin remodeling.  64  4.2.2 Genome-wide screen in hypoxia and doxorubicin Doxorubicin is an anthracycline-based drug used to treat many different types of cancers, including breast, lung, gastric, ovarian, and pediatric115. It primarily works as a topoisomerase II poison by stabilizing the TOP2-DNA complex and inhibiting the re-sealing of the DNA helix which causes improper DNA replication and double-strand DNA breaks, which triggers the DNA repair response. Intracellular oxidation/reduction of doxorubicin generates free radicals, which damages membranes, proteins, and DNA directly to cause cytotoxicity 48,55. Reduced efficacy of doxorubicin in hypoxic conditions is attributed to increased drug efflux (through P-gp), and the lack of oxygen molecules for generation of ROS14. Doxorubicin, and other anthracylines, have been screened in both HIP and HOP60 and sensitive strains included those required for double-strand repair (RAD55, RAD57). The HIP profile identified different members involved in chromatin remodeling and the RSC complex (RSC8, RSC58). A recent study described the role of doxorubicin and other anthracyclines on histone eviction, and other effects on chromatin biology116. They propose an alternative consequence of doxorubicin-DNA interaction by affecting the transcriptional response, in particular the DNA repair response, and other processes including epigenetic regulation and transcription.       65  Table 11 - Doxorubicin chemogenomic profile. In red are strains a part of the DNA repair response.  Normoxia Doxorubicin (68) Hypoxia Only (13) Hypoxia AND Hypoxia Doxorubicin (31) Normoxia AND Hypoxia Doxorubicin (15) Hypoxia Only AND Normoxia Doxorubicin (1) Hypoxia Doxorubicin Only (126) All Conditions (2)  ARC18, BUL1, LRP1, LSM7, PIH1, POR1, RSA1, SEM1, THP2, TMA23, VAC14, YNL140C, YOR309C AAT2, BUD27, CDC26, CTK3, DBF2, ECM2, END3, FMC1, GIM5, HTL1, LEA1, MFT1, MGA2, NKP2, NPL6, PAC10, RGP1, RIC1, RRP6, SEC66, SHE4, SLT2, TOM1, TOM70, VPS61, VPS63, VRP1, YKE2, YKU70, YML094C-A YPT6 DAL81, HPR5, IMP2, LDB7, MMS4, MUS81, PDR1, RAD54, RAD55, RAD57, RAD59, REG1, RPA49, RPL20A, THR1 BEM4  NPT1, PTC1  The doxorubicin profile is coherent with its known DNA-damaging mechanism. Strains deficient in the DNA repair pathways are sensitive to doxorubicin regardless of environment. Although there are a substantial number of strains are sensitive to doxorubicin only in the hypoxic environment (126), enrichment analyses do not identify many GO Terms (Figure 18). The most significant GO Term corresponds to “cellular response to DNA damage stimulus”, which is consistent given doxorubicin’s mechanism. The hypoxic environment appears to exacerbate this requirement for the DNA damage response. 66   Figure 18 - Gene ontology network analysis of 126 hypoxia and doxorubicin specific sensitive strains. Bolded terms are leading GO Terms with the highest significance in their related-processes. Network analysis is generated from ClueGo application in Cytoscape. Doxorubicin and cisplatin share many similarities with respect to the distribution of strains between conditions. Of the 49 hypoxia strains, 29 are sensitive to both doxorubicin and cisplatin. Both drugs induced a DNA repair response as seen by the RAD genes, and MMS4 and MUS81. In both drugs, strains potentiated by the hypoxic environment are involved in the global DNA damage response, and the hypoxic environment is driving this sensitivity. Hypoxia and doxorubicin, similar to hypoxia and cisplatin, sensitizes strains involved in DNA repair processes that rely on the ubiquitination protein system, with overlapping strains in RAD23, RAD27, MMS1, RTT101, and RTT107. The genes and gene lists differ between the two drugs, and to each drug’s mechanism is reflected in these differences. For example, the aforementioned LDB7 of the RSC chromatin remodeling complex is specifically sensitive to doxorubicin. This agrees with the previous identification of essential RSC complex members as being sensitive to doxorubicin under normoxic conditions116.    67  4.2.3 Genome-wide screen in hypoxia and hydroxyurea The third drug investigated was hydroxyurea, a compound used to treat sickle cell disease and cancers of the head, neck and brain117,118. In contrast to cisplatin and doxorubicin, hydroxyurea does not induce DNA damage by direct interaction with DNA. It reduces deoxyribonucleotides (dNTPs) available for DNA replication, stalling DNA polymerase at replication forks inducing DNA lesions including double strand breaks15,50. In yeast, dNTP biosynthesis relies on the ribonucleotide reductase (RNR) complex, in particular the essential gene RNR2. Hydroxyurea is used to arrest cells in the S-phase of the cell cycle, and strains involved in cell cycle regulation (DUN1, SWI4) were found to be sensitive to hydroxyurea. Similarly, the cell cycle can also be affected by hypoxia. The 49 hypoxia-specific strains suggest that defects in cell cycle regulation can result in a growth defect under hypoxic conditions, from strains related to cell cycle such as CDC26, and DBF2.           68  Table 12 - Hydroxyurea chemogenomic profile. In red are strains a part of the DNA repair response. Normoxia Hydroxyurea (62) Hypoxia Only (15) Hypoxia AND Hypoxia Hydroxyurea (30) Normoxia AND Hypoxia Hydroxyurea (19) Hypoxia Only AND Normoxia Hydroxyurea (0) Hypoxia Hydroxyurea Only (77) All Conditions (2)  BCK1, BEM4, BUL1, LRP1, LSM7, PIH1, POR1, RGP1, RIC1, RSA1, SEM1, SLT2, TMA23, TOM70, VAC14, YOR309C, YPT6 AAT2, ARC18, BUD27, CDC26, CTK3, DBF2, ECM2, END3, FMC1, GIM5, LEA1, MFT1, MGA2, NKP2, NPL6, NPT1, PAC10, PAT1, RGP1, RRP6, SEC66, SHE4, THP2, TOM1, VPS61, VPS63, VRP1, YKE2, YKU70, YML094C-A YNL140C ARP8, CHS3, DUN1, ERG4, GET2, LSM1, MMS4, MUS81, RAD5, RAD18, RAD24, RAD54, RAD55, RAD59, RVS167, SWI4, TDA1, YJL027C, YMR031W-A   NPT1, PTC1  The hydroxyurea HOP profile from previous experiments identified the ORF YJL027C as the most sensitive strain. This ORF is a dubious ORF (unlikely to code for a functional protein). It overlaps the upstream 5’ region of the essential gene RNR2 on the opposite strand, so that deletion of yjl027c disrupts the RNR2 promoter, generating a loss-of-function RNR2 allele that is sensitive to hydroxyurea. I find YJL027C sensitive to hydroxyurea in both normoxia and hypoxia, with log2 ratio values of 1.93 and 1.36, respectively. Similar to cisplatin and doxorubicin, the RAD genes related to DNA repair are observed in all hydroxyurea conditions.  69  Network analyses of the 77 genes sensitive only to hydroxyurea/hypoxia and 66 normoxia-specific strains in hydroxyurea yield no specific enrichments. One explanation may be that hydroxyurea-induced stress is a mild perturbation, and that cells are adapted to this drug. This explains the usefulness of this drug in cell studies. Although there are phenotypic overlaps between hydroxyurea and hypoxia in cell cycle arrest, there are few sensitive strains that suggest specific cellular stress on complexes or processes. Green et al. evaluated whether there are similar mechanisms leading to cell cycle arrest in hypoxia and hydroxyurea, and showed that different cellular pathways are responsible for this phenotype119.  It may be that hypoxia has a greater effect on cells stressed with more direct acting DNA damaging agents such as cisplatin and doxorubicin. This is consistent with that there are no clinical reports of resistance to hydroxyurea in hypoxic tumours.  4.2.4 Genome-wide screen in hypoxia and tirapazamine  Tirapazamine (TPZ) is a hypoxic-specific DNA-damaging compound. Its primary mechanism of action is through the reduction of the parent compound to a free radical, damaging DNA directly to produce complex lesions49. Its hypoxia specificity is through the aerobic “back-oxidation” that occurs in the presence of free oxygen, nullifying the toxicity generated from the reduction of the parent compound120. In humans, TPZ reduction is facilitated by cytochrome P450 reductase, and expression of this oxidoreductase in breast cancer cell lines increases TPZ sensitivity121. In yeast, overexpression of the yeast P450 oxidoreductase homolog NCP1 can induce the same increase in sensitivity122. TPZ was also shown to induce double-strand DNA breaks, and further investigation determined this was due to a topoisomerase II poison mechanism, similar to doxorubicin49. TPZ showed promising activity in pre-clinical studies in lung cancer models120. In clinical trials, when combined with cisplatin and radiotherapy it did not show any survival 70  benefit in lung, head, neck and cervical cancer123,124. Examining these genes important in the response to TPZ in hypoxia may help guide further development of TPZ and how its pharmacogenomics may inform current efforts to develop hypoxia prodrugs125.  Table 13 - Tirapazamine chemogenomic profile. In red are strains a part of the DNA repair response. Normoxia Tirapazamine (43) Hypoxia Only (12) Hypoxia AND Hypoxia Tirapazamine (27) Normoxia AND Hypoxia Tirapazamine (18) Hypoxia Only AND Normoxia Tirapazamine (2) Hypoxia Tirapazamine Only (98) All Conditions (7)  AAT2, FMC1, LEA1, LSM7, PIH1, RSA1, SEM1, TMA23, VAC14, YNL140C, YOR309C BUD27, BUL1, CDC26, CTK3, DBF2, ECM2, GIM5, MFT1, MGA2, NPT1, PAC10, PAT1, POR1, RIC1, RRP6, SEC66, SHE4, THP2, TOM1, TOM70, VPS61, VPS63, VRP1, YKE2, YKU70, YML094C-A, YPT6 CHS3, CHS5, CHS6, CHS7, COG7, EDE1, HOM6, HPR5, IMP2, RAD18, RAD51, RAD54, RAD55, RAD59, SSD1, YDR433W, YPR123C ARC18, BEM4  BCK1, END3, HTL1, NPL6, PTC1, RGP1, SLT2  The chemogenomic profile of TPZ is unique. As a DNA-damaging agent, TPZ induces the expected sensitivity in the RAD genes. Other TPZ-specific genes are the CHS (CHitin Synthase-related) genes, which are responsible for chitin biosynthesis. Chitin is a polysaccharide made up of the glucose derivative N-Acetylglucosamine (GlcNAc), a component of the cell wall that helps maintain wall shape and rigidity. CHS3 is required for the elongation of the polysaccharide by adding on GlcNAc, while CHS5/CHS6 both make up the complex responsible for transporting Chs3p from the Golgi to the plasma membrane. One explanation for this sensitivity is that the 71  damage to the plasma membrane and cell wall through free radical formation by TPZ requires that the cell wall integrity (CWI) pathway compensate for this damage. This hypothesis is supported by the fact that other members of the CWI signaling pathway are sensitive to TPZ, including BCK1 and SLT2. Hypoxia-induced stress on the CWI pathway compounded, with membrane damage from TPZ, may explain the synergistic effect of the environment and TPZ.   Figure 19 – Partial Gene ontology network analysis of 98 hypoxia and TPZ specific sensitive strains. In teal are child terms of the biological process “Localization”. “Establishment of protein localization” is the most significant all “Localization” child terms. In green are child terms of “ATP Transport”, where “ATP Export” is the most significant term. Non-overlapping GO Terms exclusive to “ATP Transport” is excluded from this network image.  Network analysis is generated from ClueGo application in Cytoscape.  The enrichment profile of strains sensitive specifically to TPZ and hypoxia shows GO Terms not seen in the other drugs. For example, protein localization and overall transport are heavily represented in this gene list (Figure 19). Protein localization and genes involved include COG5, 72  COG6, and COG8. These genes encode members of the Conserved Oligomeric Golgi (COG) complex and function in protein trafficking to mediate transport vesicles to Gogli compartments. The COG complex is conserved in both humans and yeast. It acts as a central hub for protein and lipid sorting and transport126. Deletion of different COG genes in yeast results in varied phenotypes including vesicle accumulation, growth defects and reduced glycosylation of proteins126. Another process associated with the COG genes is autophagy; COG mutants were found to have defective autophagy and the related pathway of cytoplasm to vacuole targeting (Cvt)127. In addition, the ESCRT complex is represented by the strains VPS20, VPS25, VPS28, and VPS36.  These genes encode members ofthe endosomal sorting complexes required for transport (ESCRT) complex. This complex is highly conserved between yeast and humans, and play rolls in protein degradation through the multivesicular body (MVB) pathway and ubiquitylation, lipid sorting, and membrane cleavage128. Examining the gene enrichment of strains sensitive to TPZ in normoxic conditions yield only one GO Term: “cellular response to DNA damage stimulus”, suggesting TPZ sensitivity is not observed in strains defective for protein localization.  4.3 Conclusion and summary The genome-wide approach used provides an overview of how genotype influences the cells response to drug stress in hypoxia. Each drug affects cells differently, in part dependent on their particular mechanism of action. Direct DNA-damaging agents such as cisplatin and doxorubicin induced a stringent response, revealing a large cohort of strains sensitive to the drug and environment. This is in contrast to hydroxyurea, where cells appear to adapt more readily, perhaps reflecting its mechanism to reducing available dNTPs. Tirapazamine demonstrates a distinct profile where protein localization is required for survival in hypoxic conditions. This 73  could reflect TPZ’s activity to damage other molecules as well as DNA. Cellular damage to membrane and proteins could impair cells in a more global manner.  A survey of the literature shows that hypoxia down-regulates the DNA repair response across different models. Here, I demonstrate that in hypoxia, the requirement for the DNA repair response is increased and strains that can handle DNA-damage stress induced by these drugs in normoxic conditions are now sensitive in hypoxia. Certain mechanisms, aside from DNA repair, are potentiated in hypoxia. For example, cisplatin’s effect on chromatin remodeling is detectable in hypoxia. For the hypoxia-specific compound tirapazamine, hypoxia potentiates the mechanism of action and strains defective in protein localization and transport become sensitive.  4.4 Material and methods 4.4.1 Homozygous profiling with chemotherapeutics  HOP chemogenomic screens were performed as described in 2.1.1 in combination with drugs dosed at concentrations shown in Table 8 to achieve an IC 15-25 in normoxia and hypoxia. Each drug was dissolved in a solvent to a stock concentration, and diluted to their respective doses with the homozygous pool in 700uL per well. Cisplatin (Toronto Research Canada) and hydroxyurea (Sigma Aldrich) were dissolved in water. Doxorubicin (Sigma Aldrich) and tirapazamine (Sigma Aldrich) were dissolved in DMSO. All experimental (no-drug) controls had an equal volume of solvent added to each well as a vehicle control. For drugs in DMSO, drug and vehicle volume concentrations were kept <2%. IC values were determined from the “Average Generation Ratio” as determined by YG relative to the no-drug control in the same environment (normoxia or hypoxia).  74  4.4.2 Gene ontology analysis using Cytoscape and ClueGO Gene enrichment analyses for chemogenomics screen were performed using Cytoscape with the ClueGO application. The gene universe was considered as those strains detectable above background in 15 normoxia control arrays (4552 strains). One-sided hypergeometric tests were performed on each unique gene list with the following thresholds: P-value < 0.0005 and a minimum of 5 genes in each GO Term. The ClueGO application was used to visualize enrichment networks, showing “level four” GO Terms. Each GO domain (biological process, molecular function, and cellular component) represent a level zero GO. Each child term is an increase of one level.  75  Chapter 5: Summary and future directions 5.1 Summary In Chapter 2:, I identified biological processes and pathways that are important for yeast to survive in the hypoxic environment. In Chapter 2:, I used a genome-wide approach to interrogate ~4800 non-essential homozygous single-mutant strains grown together in hypoxia (<0.2% O2), and compared each mutant’s growth to its growth in standard laboratory (normoxia) conditions. No overall growth defect of the homozygous deletion pool grown in hypoxic conditions was observed However; assessing growth differences between all mutants identified 49 strains consistently under-represented in hypoxic growth. Using Gene Ontology and network analysis, these 49 strains demonstrated enrichment in specific biological processes. These included cell cycle, cytoskeleton / cell wall integrity, RNA metabolism, transcription elongation, and protein-related processes including ribosomal biogenesis and protein folding. I compared these 49 strains to previous work describing cellular response to small molecules and other perturbations, and found 10 of the hypoxia-sensitive strains belong to the heat shock/prefoldin response signature,  representing 27% of this response. These strains are co-fit in heat shock and perturbation to the prefoldin complex, suggesting that hypoxic stress induces a similar cellular response as heat stress.   In Chapter 2, I explored the biological processes affected by hypoxia and looked to validate the observed similarity between hypoxia and heat shock. I took a global approach and examined the mRNA transcriptome of wild-type cells in normoxia/hypoxia, as well as cells grown in raffinose. Complementary to my chemogenomic screens, yeast grown in hypoxia/glucose exhibit few changes in gene expression. Enrichment analyses on the transcripts being up-regulated suggested 76  different DNA repair genes related to sporulation function (DMC1, MEI4) are up-regulated, while enrichment shows thioredoxin peroxidase and ion transport activity being down-regulated.   A greater phenotypic change is observed in transcriptional regulation (with many transcripts up and down-regulated) in growth with raffinose, a less-efficient carbon source that promotes respiratory metabolism for growth. This is supported by genes up-regulated belonging to the GO domain of cellular component in the mitochondria and related mitochondria membranes. Highly significant GO Terms from biological processes include cellular respiration, respiratory chain, ATP metabolism and multiple nucleoside biosynthesis pathways. Genes that were down-regulated in raffinose growth included ribosomal biogenesis and rRNA metabolism. This down-regulation can be attributed to an overall cellular response to reduced ribosomal-related processes and translation, energetically costly processes. This is consistent with previously described studies that demonstrated strains deleted for gene products in ribosomal biogenesis have greater fitness in hypoxia98, suggesting this is an adaptive response to hypoxic stress. However, previous work on the environmental stress response also identified ribosomal processes as a general stress response. In the comparison of cells grown in hypoxia/raffinose to normoxia/raffinose, there is no change in ribosomal activity, suggesting in raffinose expression of these genes are similar, regardless of the environment. Further studies are required to elucidate whether this regulation is hypoxia-sepcific. The greatest number of transcript that changes their expression relative to the control condition (normoxia/glucose) is when cells are grown in raffinose and hypoxia. Enrichment analysis of this condition shows many similarities with normoxia/raffinose. One of the biggest differences is an increase in genes in the GO Term of oxidoreductase activity, specifically peroxidase activity. This may be reflective of an increase of 77  ROS in hypoxia with an active electron transport chain using respiratory metabolism. Up-regulation specific to hypoxia when comparing raffinose conditions (hypoxia/raffinose to normoxia/raffinose) show fatty acid metabolism is up-regulated, along with genes with sporulation function, complementing genes found up-regulated in conditions of glucose.   To further explore the findings in Chapter 1, I examined the heat shock response and its relation to hypoxic stress. Many of my hypoxia-specific strains overlapped with the strains in the heat shock response signature, I therefore combined hypoxia and heat stress for wild-type cell growth. Surprisingly, cells did not grow in hypoxia at 37°C. By titrating the temperature down and holding the oxygen level constant (<0.2%) growth was restored. I next asked if the combination of hypoxia and heat stress induced cell death or simply halted cell proliferation. A colony formation unit assay showed that wild-type cells stressed for 4 hours in hypoxia/37°C were able to recover with no significant difference with cells stressed at only 37°C, suggesting the lack of growth in hypoxia/37°C was due to cell arrest.   In Chapter 3, I combined my hypoxic genome-wide approach with clinically relevant chemotherapeutics to produce unique chemogenomic profiles for four compounds (cisplatin, doxorubicin, hydroxyurea and tirapazamine) against the yeast homozygous collection. Examining drugs with direct and indirect DNA-damaging mechanisms, I found hypoxia exacerbated the need for DNA-repair, particularly in cisplatin and doxorubicin. Deletion strains pertaining to DNA-repair pathways that were not sensitive to these drugs under normoxic conditions became sensitive under hypoxic conditions. In addition, processes that may not be as greatly affected by these compounds become increasingly important under hypoxic conditions. 78  For example, cisplatin has only modest effects on chromatin function in standard conditions, whereas these strains become more sensitive in hypoxia and cisplatin. In the compound tirapazamine, the hypoxia-specific mechanism of action is known. I show that strains with genotypic backgrounds deficient in protein localization, transport and sorting are sensitive to tirapazamine in hypoxia. I conclude that hypoxic stress with DNA-damaging chemotherapeutics potentiates the requirement for DNA repair and exacerbate additional drug mechanisms.  5.2 Future directions My work has uncovered different directions for exploration regarding hypoxia on cellular responses with chemotherapeutics. Here, I propose different future directions and experiments to continue the interrogation of hypoxia.   1. Hypoxia and heat stress. In Chapter 2:, I observed that the cellular response to the two stresses is similar; however, I did not detect a change in the major heat shock response genes such as HSP82, HSC81 and HSP104 genome wide screens. Identifying the gene overlaps that mediates these stresses can explain the synergistic effect I observed.  2. Transcription machinery and transcription elongation. RNA metabolism and transcription regulation are evidently affected in hypoxia, but particularly the process of transcription elongation. The THO complex and partially the TREX complex are both observed in my hypoxia-only data, and again in the presence of cisplatin and doxorubicin. In vitro experiments such as co-immunoprecipitation (Co-IP) and chromatin immunoprecipitation (ChIP) in hypoxia can identify proteins in the transcriptional machinery being affected and specifically which genes are being affected with improper transcription regulation.  3. Peroxidase activity and the formation of reactive oxygen species. The data showed contradiction in the regulation of peroxidase activity, similar to the literature where 79  evidently there are two different perspectives on ROS formation in hypoxia. The experimental design using raffinose presents a suitable model to test hypotheses on ROS formation. A ROS staining microscopy experiment could elucidate if ROS formation is different in the two carbon sources in hypoxia.  4. Prefoldin subunits with canonical protein folding functionality and independent nuclear function. My data suggests that the cytoskeleton is affected by hypoxia, evidently with sensitive strains in the CWI pathway. The prefoldin complex along with the CCT complex is important in maintaining cytoskeleton and cell morphology, and their deletion can explain hypoxia sensitivity. However, the degree of sensitivity of the prefoldin subunits differs between each one. This suggests different utilization of each subunit outside of protein folding, and given previous work has described their localization and function in the nucleus, investigation of this additional functionality may be worthwhile. A possible experiment could be synthetic genetic array, to query a prefoldin subunit into the deletion collection to create a pool of double mutants. PFD5 would be an outstanding candidate given its great sensitivity to hypoxia and non-canonical function in the nucleus.         80  Bibliography 1. Carnero, A. & Lleonart, M. The hypoxic microenvironment: A determinant of cancer stem cell evolution. BioEssays 38, S65–S74 (2016). 2. Greaves, M. & Maley, C. C. Clonal evolution in cancer. Nature 481, 306–13 (2012). 3. Nowell, P. C. The clonal evolution of tumor cell populations. Science 194, 23–8 (1976). 4. Leszczynska, K. B. et al. Targeting Tumour Hypoxia with ParP Inhibitors: Contextual Synthetic Lethality. Cancer Drug Discov. Dev. 83, 5. Park, C. H., Bergsagel, D. E. & McCulloch, E. A. Mouse myeloma tumor stem cells: a primary cell culture assay. J. Natl. Cancer Inst. 46, 411–22 (1971). 6. Castells, M., Thibault, B., Delord, J.-P. & Couderc, B. Implication of tumor microenvironment in chemoresistance: tumor-associated stromal cells protect tumor cells from cell death. Int. J. Mol. Sci. 13, 9545–71 (2012). 7. Shree, T. et al. Macrophages and cathepsin proteases blunt chemotherapeutic response in breast cancer. Genes Dev. 25, 2465–79 (2011). 8. Quail, D. F. & Joyce, J. A. Microenvironmental regulation of tumor progression and metastasis. Nat. Med. 19, 1423–37 (2013). 9. Estrella, V. et al. Acidity generated by the tumor microenvironment drives local invasion. Cancer Res. 73, 1524–35 (2013). 10. Dhani, N., Fyles, A., Hedley, D. & Milosevic, M. The Clinical Significance of Hypoxia in Human Cancers. Semin. Nucl. Med. 45, 110–121 (2015). 11. Semenza, G. L. Cancer-stromal cell interactions mediated by hypoxia-inducible factors promote angiogenesis, lymphangiogenesis, and metastasis. Oncogene 32, 4057–63 (2013). 12. Gilkes, D. M., Semenza, G. L. & Wirtz, D. Hypoxia and the extracellular matrix: drivers 81  of tumour metastasis. Nat. Rev. Cancer 14, 430–9 (2014). 13. velaei, K., Samadi, N., Barazvan, B. & Soleimani Rad, J. Tumor microenvironment-mediated chemoresistance in breast cancer. The Breast 30, 92–100 (2016). 14. Doktorova, H., Hrabeta, J., Khalil, M. A. & Eckschlager, T. Hypoxia-induced chemoresistance in cancer cells: The role of not only HIF-1 INTRODUCTION – HYPOXIA-INDUCED CHEMORESISTANCE. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub 159, 166–177 (2015). 15. Helleday, T., Petermann, E., Lundin, C., Hodgson, B. & Sharma, R. A. DNA repair pathways as targets for cancer therapy. Nat. Rev. Cancer 8, 193–204 (2008). 16. Giaever, G. et al. Functional profiling of the Saccharomyces cerevisiae genome. Nature 418, 387–391 (2002). 17. Pierce, S. E., Davis, R. W., Nislow, C. & Giaever, G. Genome-wide analysis of barcoded Saccharomyces cerevisiae gene-deletion mutants in pooled cultures. Nat. Protoc. 2, 2958–2974 (2007). 18. Ni, L. & Snyder, M. A genomic study of the bipolar bud site selection pattern in Saccharomyces cerevisiae. Mol. Biol. Cell 12, 2147–70 (2001). 19. Pagé, N. et al. A Saccharomyces cerevisiae Genome-Wide Mutant Screen for Altered Sensitivity to K1 Killer Toxin Such a collection promotes the discovery of cellular. 20. Giaever, G. & Nislow, C. The yeast deletion collection: a decade of functional genomics. Genetics 197, 451–65 (2014). 21. Giaever, G. et al. Chemogenomic profiling: identifying the functional interactions of small molecules in yeast. Proc. Natl. Acad. Sci. U. S. A. 101, 793–8 (2004). 22. Deutschbauer, A. M. et al. Mechanisms of haploinsufficiency revealed by genome-wide 82  profiling in yeast. Genetics 169, 1915–25 (2005). 23. Cheung-Ong, K. et al. Comparative Chemogenomics To Examine the Mechanism of Action of DNA-Targeted Platinum-Acridine Anticancer Agents. ACS Chem. Biol. 7, 1892–1901 (2012). 24. Hillenmeyer, M. E. et al. The chemical genomic portrait of yeast: uncovering a phenotype for all genes. Science 320, 362–5 (2008). 25. Smith, A. M., Ammar, R., Nislow, C. & Giaever, G. A survey of yeast genomic assays for drug and target discovery. Pharmacol. Ther. 127, 156–64 (2010). 26. Hoepfner, D. et al. High-resolution chemical dissection of a model eukaryote reveals targets, pathways and gene functions. Microbiol. Res. 169, 107–120 (2014). 27. Birrell, G. W., Giaever, G., Chu, A. M., Davis, R. W. & Brown, J. M. A genome-wide screen in Saccharomyces cerevisiae for genes affecting UV radiation sensitivity. Proc. Natl. Acad. Sci. U. S. A. 98, 12608–13 (2001). 28. Zhang, L. & Hach, A. Molecular mechanism of heme signaling in yeast: the transcriptional activator Hap1 serves as the key mediator. Cell. Mol. Life Sci. 56, 415–26 (1999). 29. Trendeleva, T. A., Aliverdieva, D. A. & Zvyagilskaya, R. A. Mechanisms of sensing and adaptive responses to low oxygen conditions in mammals and yeasts. Biochem. 79, 750–760 (2014). 30. Martin, C. E., Oh, C.-S. & Jiang, Y. Regulation of long chain unsaturated fatty acid synthesis in yeast. Biochim. Biophys. Acta - Mol. Cell Biol. Lipids 1771, 271–285 (2007). 31. Natter, K. & Kohlwein, S. D. Yeast and cancer cells - common principles in lipid metabolism. Biochim. Biophys. Acta 1831, 314–26 (2013). 83  32. Kwast, K. E., Burke, P. V, Staahl, B. T. & Poyton, R. O. Oxygen sensing in yeast: evidence for the involvement of the respiratory chain in regulating the transcription of a subset of hypoxic genes. Proc. Natl. Acad. Sci. U. S. A. 96, 5446–51 (1999). 33. González Siso, M. I. et al. Kluyveromyces lactis : A Suitable Yeast Model to Study Cellular Defense Mechanisms against Hypoxia-Induced Oxidative Stress. Oxid. Med. Cell. Longev. 2012, 1–14 (2012). 34. Davies, B. S. J., Wang, H. S. & Rine, J. Dual activators of the sterol biosynthetic pathway of Saccharomyces cerevisiae: similar activation/regulatory domains but different response mechanisms. Mol. Cell. Biol. 25, 7375–85 (2005). 35. Ratcliffe, P. J. Oxygen sensing and hypoxia signalling pathways in animals: the implications of physiology for cancer. J. Physiol. 591, 2027–42 (2013). 36. Semenza, G. L. Targeting HIF-1 for cancer therapy. Nat. Rev. Cancer 3, 721–732 (2003). 37. Yasuda, H. Solid tumor physiology and hypoxia-induced chemo/radio-resistance: Novel strategy for cancer therapy: Nitric oxide donor as a therapeutic enhancer. Nitric Oxide - Biol. Chem. 19, 205–216 (2008). 38. Warburg, O. On the Origin of Cancer Cells. Science (80-. ). 123, (1956). 39. Moreno-Sánchez, R., Rodríguez-Enríquez, S., Saavedra, E., Marín-Hernández, A. & Gallardo-Pérez, J. C. The bioenergetics of cancer: Is glycolysis the main ATP supplier in all tumor cells? BioFactors 35, 209–225 (2009). 40. Cairns, R. A., Harris, I. S. & Mak, T. W. Regulation of cancer cell metabolism. Nat. Rev. Cancer 11, 85–95 (2011). 41. Hanahan, D. & Weinberg, R. A. The Hallmarks of Cancer. Cell 100, 57–70 (2000). 42. Hanahan, D. & Weinberg, R. A. Hallmarks of Cancer: The Next Generation. Cell 144, 84  646–674 (2011). 43. Seagroves, T. N. et al. Transcription factor HIF-1 is a necessary mediator of the pasteur effect in mammalian cells. Mol. Cell. Biol. 21, 3436–44 (2001). 44. Samanfar, B. et al. Large-scale investigation of oxygen response mutants in Saccharomyces cerevisiae. Mol. Biosyst. 9, 1351–9 (2013). 45. Voordeckers, K. & Verstrepen, K. J. Experimental evolution of the model eukaryote Saccharomyces cerevisiae yields insight into the molecular mechanisms underlying adaptation. Curr. Opin. Microbiol. 28, 1–9 (2015). 46. Roos, W. P. & Kaina, B. DNA damage-induced cell death by apoptosis. Trends Mol. Med. 12, 440–450 (2006). 47. Chabner, B. A. & Roberts, T. G. Timeline: Chemotherapy and the war on cancer. Nat. Rev. Cancer 5, 65–72 (2005). 48. Nitiss, J. L. Targeting DNA topoisomerase II in cancer chemotherapy. Nat. Rev. Cancer 9, 338–50 (2009). 49. Peters, K. B. & Brown, J. M. Tirapazamine. Cancer Res. 62, (2002). 50. Koç, A., Wheeler, L. J., Mathews, C. K. & Merrill, G. F. Hydroxyurea arrests DNA replication by a mechanism that preserves basal dNTP pools. J. Biol. Chem. 279, 223–30 (2004). 51. Cheung-Ong, K., Giaever, G. & Nislow, C. Chemistry &amp; Biology Perspective DNA-Damaging Agents in Cancer Chemotherapy: Serendipity and Chemical Biology. Chem. Biol. 20, 648–659 (2013). 52. Parsons, A. B. et al. Exploring the mode-of-action of bioactive compounds by chemical-genetic profiling in yeast. Cell 126, 611–25 (2006). 85  53. Guaragnella, N. et al. The expanding role of yeast in cancer research and diagnosis: insights into the function of the oncosuppressors p53 and BRCA1/2. (2013). doi:10.1111/1567-1364.12094 54. Wu, H.-M. et al. Hypoxia-induced autophagy mediates cisplatin resistance in lung cancer cells. Sci. Rep. 5, 12291 (2015). 55. Thorn, C. F. et al. Doxorubicin pathways: pharmacodynamics and adverse effects. Pharmacogenet. Genomics 21, 440–6 (2011). 56. Qutub, A. A. & Popel, A. S. Reactive oxygen species regulate hypoxia-inducible factor 1alpha differentially in cancer and ischemia. Mol. Cell. Biol. 28, 5106–19 (2008). 57. Ishtar Snoek, I. S. & Yde Steensma, H. Why does Kluyveromyces lactis not grow under anaerobic conditions? Comparison of essential anaerobic genes of Saccharomyces cerevisiae with the Kluyveromyces lactis genome. FEMS Yeast Res. 6, 393–403 (2006). 58. Balakrishnan, R. et al. YeastMine--an integrated data warehouse for Saccharomyces cerevisiae data as a multipurpose tool-kit. Database (Oxford). 2012, bar062 (2012). 59. Hamosh, A., Scott, A. F., Amberger, J. S., Bocchini, C. A. & McKusick, V. A. Online Mendelian Inheritance in Man (OMIM), a knowledgebase of human genes and genetic disorders. Nucleic Acids Res. 33, D514-7 (2005). 60. Lee, A. Y. et al. Mapping the cellular response to small molecules using chemogenomic fitness signatures. Science 344, 208–11 (2014). 61. Ashburner, M. et al. Gene Ontology: tool for the unification of biology. Nat. Genet. 25, 25–29 (2000). 62. Warde-Farley, D. et al. The GeneMANIA prediction server: biological network integration for gene prioritization and predicting gene function. Nucleic Acids Res. 38, 86  W214-20 (2010). 63. Shannon, P. et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 13, 2498–504 (2003). 64. Bindea, G. et al. ClueGO: a Cytoscape plug-in to decipher functionally grouped gene ontology and pathway annotation networks. Bioinformatics 25, 1091–3 (2009). 65. Cherry, J. M. et al. Saccharomyces Genome Database: the genomics resource of budding yeast. Nucleic Acids Res. 40, D700–D705 (2012). 66. Marxsen, J. H. et al. Hypoxia-inducible factor-1 (HIF-1) promotes its degradation by induction of HIF-alpha-prolyl-4-hydroxylases. Biochem. J. 381, 761–7 (2004). 67. Sims, R. J., Belotserkovskaya, R. & Reinberg, D. Elongation by RNA polymerase II: the short and long of it. Genes Dev. 18, 2437–68 (2004). 68. Verghese, J., Abrams, J., Wang, Y. & Morano, K. A. Biology of the heat shock response and protein chaperones: budding yeast (Saccharomyces cerevisiae) as a model system. Microbiol. Mol. Biol. Rev. 76, 115–58 (2012). 69. Feldman, D. E., Chauhan, V. & Koong, A. C. The Unfolded Protein Response: A Novel Component of the Hypoxic Stress Response in Tumors. Mol. Cancer Res. 3, (2005). 70. Millán-Zambrano, G. et al. The Prefoldin Complex Regulates Chromatin Dynamics during Transcription Elongation. PLoS Genet. 9, e1003776 (2013). 71. Son, J. et al. Glutamine supports pancreatic cancer growth through a KRAS-regulated metabolic pathway. Nature 496, 101–105 (2013). 72. Clancy, J. L. et al. EDD, the human orthologue of the hyperplastic discs tumour suppressor gene, is amplified and overexpressed in cancer. Oncogene 22, 5070–5081 (2003). 87  73. Eickelberg, O. et al. Functional activation of heat shock factor and hypoxia-inducible factor in the kidney. J. Am. Soc. Nephrol. 13, 2094–101 (2002). 74. Baird, N. A., Turnbull, D. W. & Johnson, E. A. Induction of the heat shock pathway during hypoxia requires regulation of heat shock factor by hypoxia-inducible factor-1. J. Biol. Chem. 281, 38675–81 (2006). 75. Proctor, M. et al. The automated cell: compound and environment screening system (ACCESS) for chemogenomic screening. Methods Mol. Biol. 759, 239–69 (2011). 76. Baker Brachmann, C. et al. Designer deletion strains derived fromSaccharomyces cerevisiae S288C: A useful set of strains and plasmids for PCR-mediated gene disruption and other applications. Yeast 14, 115–132 (1998). 77. Becerra, M. et al. The yeast transcriptome in aerobic and hypoxic conditions: effects of hap1 , rox1 , rox3 and srb10 deletions. Mol. Microbiol. 43, 545–555 (2002). 78. Staudacher, J. J. et al. Hypoxia-induced gene expression results from selective mRNA partitioning to the endoplasmic reticulum. Nucleic Acids Res. 43, 3219–3236 (2015). 79. Agrawal, R. et al. Hypoxic signature of microRNAs in glioblastoma: insights from small RNA deep sequencing. BMC Genomics 15, 686 (2014). 80. Camps, C. et al. Integrated analysis of microRNA and mRNA expression and association with HIF binding reveals the complexity of microRNA expression regulation under hypoxia. Mol. Cancer 13, 28 (2014). 81. Zhou, L. et al. Expression profiling analysis of hypoxic pulmonary disease. Genet. Mol. Res. 12, 4162–4170 (2013). 82. Kim, D. K. et al. Differentially regulated functional gene clusters identified in early hypoxic cardiomyocytes. Mol. Biol. Rep. 39, 9549–9556 (2012). 88  83. Matouk, I. J. et al. The oncofetal H19 RNA connection: Hypoxia, p53 and cancer. Biochim. Biophys. Acta - Mol. Cell Res. 1803, 443–451 (2010). 84. Safronova, O. S., Nakahama, K.-I. & Morita, I. Acute hypoxia affects P-TEFb through HDAC3 and HEXIM1-dependent mechanism to promote gene-specific transcriptional repression. Nucleic Acids Res. 42, 8954–69 (2014). 85. ter Linde, J. J. et al. Genome-wide transcriptional analysis of aerobic and anaerobic chemostat cultures of Saccharomyces cerevisiae. J. Bacteriol. 181, 7409–13 (1999). 86. Birrell, G. W. et al. Transcriptional response of Saccharomyces cerevisiae to DNA-damaging agents does not identify the genes that protect against these agents. Proc. Natl. Acad. Sci. U. S. A. 99, 8778–83 (2002). 87. Crabtree, H. G. Observations on the carbohydrate metabolism of tumours. Biochem. J. 23, (1929). 88. DeRisi, J. L., Iyer, V. R. & Brown, P. O. Exploring the metabolic and genetic control of gene expression on a genomic scale. Science 278, 680–6 (1997). 89. Barnett, J. A. The Utilization of Sugars by Yeasts. Adv. Carbohydr. Chem. Biochem. 32, 125–234 (1976). 90. Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550 (2014). 91. Dresser, M. E. et al. DMC1 functions in a Saccharomyces cerevisiae meiotic pathway that is largely independent of the RAD51 pathway. Genetics 147, 533–44 (1997). 92. Bindra, R. S. et al. Down-regulation of Rad51 and decreased homologous recombination in hypoxic cancer cells. Mol. Cell. Biol. 24, 8504–18 (2004). 93. Ross, S. J., Findlay, V. J., Malakasi, P. & Morgan, B. A. Thioredoxin peroxidase is 89  required for the transcriptional response to oxidative stress in budding yeast. Mol. Biol. Cell 11, 2631–42 (2000). 94. Vaux, E. C., Metzen, E., Yeates, K. M. & Ratcliffe, P. J. Regulation of hypoxia-inducible factor is preserved in the absence of a functioning mitochondrial respiratory chain. Blood 98, 296–302 (2001). 95. Bell, E. L., Klimova, T. A., Eisenbart, J., Schumacker, P. T. & Chandel, N. S. Mitochondrial reactive oxygen species trigger hypoxia-inducible factor-dependent extension of the replicative life span during hypoxia. Mol. Cell. Biol. 27, 5737–45 (2007). 96. Guzy, R. D. et al. Mitochondrial complex III is required for hypoxia-induced ROS production and cellular oxygen sensing. Cell Metabolism 1, (2005). 97. Ishida, S., Lee, J., Thiele, D. J. & Herskowitz, I. Uptake of the anticancer drug cisplatin mediated by the copper transporter Ctr1 in yeast and mammals. Proc. Natl. Acad. Sci. U. S. A. 99, 14298–302 (2002). 98. Shah, A. N. et al. Deletion of a subgroup of ribosome-related genes minimizes hypoxia-induced changes and confers hypoxia tolerance. Physiol. Genomics 43, 855–72 (2011). 99. Dai, C. & Sampson, S. B. HSF1: Guardian of Proteostasis in Cancer. Trends Cell Biol. 26, 17–28 (2016). 100. Martín, H., Arroyo, J., Sánchez, M., Molina, M. & Nombela, C. Activity of the yeast MAP kinase homologue Slt2 is critically required for cell integrity at 37 degrees C. Mol. Gen. Genet. 241, 177–84 (1993). 101. Boulon, S. et al. HSP90 and its R2TP/Prefoldin-like cochaperone are involved in the cytoplasmic assembly of RNA polymerase II. Mol. Cell 39, 912–24 (2010). 102. Parker, R. & Sheth, U. P Bodies and the Control of mRNA Translation and Degradation. 90  Mol. Cell 25, 635–646 (2007). 103. Baryshnikova, A. et al. Quantitative analysis of fitness and genetic interactions in yeast on a genome scale. Nat. Methods 7, 1017–24 (2010). 104. Nathan, D. F., Vos, M. H. & Lindquist, S. In vivo functions of the Saccharomyces cerevisiae Hsp90 chaperone. Proc. Natl. Acad. Sci. U. S. A. 94, 12949–56 (1997). 105. Gasch, A. P. et al. Genomic expression programs in the response of yeast cells to environmental changes. Mol. Biol. Cell 11, 4241–57 (2000). 106. Morano, K. A., Grant, C. M. & Moye-Rowley, W. S. The response to heat shock and oxidative stress in Saccharomyces cerevisiae. Genetics 190, 1157–95 (2012). 107. Bishop, D. K., Park, D., Xu, L. & Kleckner, N. DMC1: a meiosis-specific yeast homolog of E. coli recA required for recombination, synaptonemal complex formation, and cell cycle progression. Cell 69, 439–56 (1992). 108. Gasch, A. P. et al. Genomic expression programs in the response of yeast cells to environmental changes. Mol. Biol. Cell 11, 4241–57 (2000). 109. Dobin, A. et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics 29, 15–21 (2013). 110. Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139–40 (2010). 111. Valero-Mora, P. M. & Wickham, H. ggplot2: Elegant Graphics for Data Analysis. JSS J. Stat. Softw. B. Rev. 35, (2010). 112. Dasari, S. & Tchounwou, P. B. Cisplatin in cancer therapy: molecular mechanisms of action. Eur. J. Pharmacol. 740, 364–78 (2014). 91  113. Mizushima, N. & Klionsky, D. J. Protein turnover via autophagy: implications for metabolism. Annu. Rev. Nutr. 27, 19–40 (2007). 114. Mymryk, J. S., Zaniewski, E. & Archer, T. K. Cisplatin inhibits chromatin remodeling, transcription factor binding, and transcription from the mouse mammary tumor virus promoter in vivo. Proc. Natl. Acad. Sci. U. S. A. 92, 2076–80 (1995). 115. Cortés-Funes, H. & Coronado, C. Role of anthracyclines in the era of targeted therapy. Cardiovasc. Toxicol. 7, 56–60 (2007). 116. Pang, B. et al. Drug-induced histone eviction from open chromatin contributes to the chemotherapeutic effects of doxorubicin. Nat. Commun. 4, 1908 (2013). 117. McGann, P. T. & Ware, R. E. Hydroxyurea therapy for sickle cell anemia. Expert Opin. Drug Saf. 14, 1749–1758 (2015). 118. Madaan, K., Kaushik, D. & Verma, T. Hydroxyurea: a key player in cancer chemotherapy. Expert Rev. Anticancer Ther. 12, 19–29 (2012). 119. Hammond, E. M., Green, S. L. & Giaccia, A. J. Comparison of hypoxia-induced replication arrest with hydroxyurea and aphidicolin-induced arrest. Mutat. Res. 532, 205–13 (2003). 120. Evans, J. W., Yudoh, K., Delahoussaye, Y. M. & Brown3, J. M. Tirapazamine Is Metabolized to Its DNA-damaging Radical by Intranuclear Enzymes1. CANCER Res. 58, 2098–2101 (1998). 121. Saunders, M. P., Patterson, A. V, Chinje, E. C., Harris, A. L. & Stratford, I. J. NADPH:cytochrome c (P450) reductase activates tirapazamine (SR4233) to restore hypoxic and oxic cytotoxicity in an aerobic resistant derivative of the A549 lung cancer cell line. Br. J. Cancer 82, 651–6 (2000). 92  122. S, □, Hellauer, K., Lesage, G., Sdicu, A.-M. & Turcotte, B. Large-Scale Analysis of Genes that Alter Sensitivity to the Anticancer Drug Tirapazamine in Saccharomyces cerevisiae. doi:10.1124/mol.105.012963 123. Reddy, S. B. & Williamson, S. K. Tirapazamine: a novel agent targeting hypoxic tumor cells. Expert Opin. Investig. Drugs 18, 77–87 (2009). 124. DiSilvestro, P. A. et al. Phase III Randomized Trial of Weekly Cisplatin and Irradiation Versus Cisplatin and Tirapazamine and Irradiation in Stages IB2, IIA, IIB, IIIB, and IVA Cervical Carcinoma Limited to the Pelvis: A Gynecologic Oncology Group Study. J. Clin. Oncol. 32, 458–464 (2014). 125. Guise, C. P. et al. Bioreductive prodrugs as cancer therapeutics: targeting tumor hypoxia. Chin. J. Cancer 33, 80–6 (2014). 126. Smith, R. D. & Lupashin, V. V. Role of the conserved oligomeric Golgi (COG) complex in protein glycosylation. Carbohydr. Res. 343, 2024–31 (2008). 127. Yen, W.-L. et al. The conserved oligomeric Golgi complex is involved in double-membrane vesicle formation during autophagy. J. Cell Biol. 188, 101–14 (2010). 128. Wollert, T. et al. The ESCRT machinery at a glance. J. Cell Sci. Cell Sci.J. Cell Sci 122, 2163–2166  93  Appendices Appendix A  :Data from chapter 2 Table 14 - GO enrichment of the 49 hypoxia-sensitive strains. GOID Ontology GO Term # of Genes % of Term Corrected P-Value Genes GO:0006397 Biological Process mRNA processing 8 10.25641 0.000202 ECM2, PAT1, SLT2, PIH1, MFT1, CTK3, LSM7, LEA1 GO:0006354 Biological Process DNA-templated transcription, elongation 7 12.5 0.000238 HTL1, THP2, YKE2, MFT1, GIM5, CTK3, NPL6 GO:0016071 Biological Process mRNA metabolic process 10 7.142857 0.000271 ECM2, PAT1, SLT2, PIH1, LRP1, MFT1, CTK3, LSM7, RRP6, LEA1 GO:0008092 Molecular Function cytoskeletal protein binding 7 11.86441 0.000332 PAC10, YKE2, VRP1, ARC18, GIM5, END3, SHE4 GO:0006396 Biological Process RNA processing 12 5.454546 0.000409 ECM2, PAT1, PTC1, TOM1, SLT2, PIH1, LRP1, MFT1, CTK3, LSM7, RRP6, LEA1 GO:0065003 Biological Process macromolecular complex assembly 13 4.659498 0.000858 ECM2, PAT1, SEM1, BUD27, CDC26, PAC10, PIH1, FMC1, YKE2, VRP1, ARC18, GIM5, RSA1 GO:0034622 Biological Process cellular macromolecular complex assembly 12 5.042017 0.000901 ECM2, PAT1, SEM1, BUD27, PAC10, PIH1, FMC1, YKE2, VRP1, ARC18, GIM5, RSA1 GO:0022613 Biological Process ribonucleoprotein complex biogenesis 10 6.060606 0.001115 ECM2, PAT1, TOM1, BUD27, PIH1, LRP1, TMA23, LSM7, RRP6, RSA1 GO:0016272 Cellular Component prefoldin complex 3 60 0.001129 PAC10, YKE2, GIM5 GO:0006368 Biological Process transcription elongation from RNA polymerase II promoter 6 12.5 0.001181 HTL1, THP2, YKE2, MFT1, GIM5, NPL6 94  GOID Ontology GO Term # of Genes % of Term Corrected P-Value Genes GO:0008380 Biological Process RNA splicing 6 11.32076 0.002087 ECM2, PTC1, SLT2, PIH1, LSM7, LEA1 GO:0006403 Biological Process RNA localization 7 8.75 0.002317 SEM1, TOM1, LRP1, THP2, MFT1, RRP6, SHE4 GO:0043623 Biological Process cellular protein complex assembly 8 6.060606 0.008488 SEM1, BUD27, PAC10, FMC1, YKE2, VRP1, ARC18, GIM5 GO:0007021 Biological Process tubulin complex assembly 3 33.33333 0.008569 PAC10, YKE2, GIM5 GO:0032535 Biological Process regulation of cellular component size 5 10.41667 0.014602 TOM1, SLT2, PIH1, VRP1, ARC18 GO:0090066 Biological Process regulation of anatomical structure size 5 10.41667 0.014602 TOM1, SLT2, PIH1, VRP1, ARC18 GO:0071822 Biological Process protein complex subunit organization 10 4.184101 0.023371 HTL1, SEM1, BUD27, PAC10, FMC1, YKE2, VRP1, ARC18, GIM5, NPL6 GO:0008023 Cellular Component transcription elongation factor complex 3 23.07692 0.027037 THP2, MFT1, CTK3 GO:0051301 Biological Process cell division 8 4.705883 0.044946 PAT1, RGP1, CDC26, DBF2, SLT2, NKP2, VRP1, END3 GO:0019787 Molecular Function ubiquitin-like protein transferase activity 3 4.285714 0.047849 TOM1, CDC26, BUL1    95  Appendix B  :Data from chapter 3 Table 15 - Colonies counted for CFU viability assay.  # of Colonies: 1 Hour Stress # of Colonies: 4 Hours Stress Replicate #1   30°C 239 72 37°C 57 34 37°C + Hypoxia 33 28 Replicate #2   30°C 51 86 37°C 40 60 37°C + Hypoxia 45 26 Replicate #3   30°C 52 61 37°C 44 62 37°C + Hypoxia 39 53    

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0340600/manifest

Comment

Related Items