Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Synthetic lethality and synthetic cytotoxicity strategies for selective killing of ATM deficient cells 2012

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
ubc_2012_spring_li_xuesong.pdf [ 3.23MB ]
Metadata
JSON: 1.0072731.json
JSON-LD: 1.0072731+ld.json
RDF/XML (Pretty): 1.0072731.xml
RDF/JSON: 1.0072731+rdf.json
Turtle: 1.0072731+rdf-turtle.txt
N-Triples: 1.0072731+rdf-ntriples.txt
Citation
1.0072731.ris

Full Text

 SYNTHETIC LETHALITY AND SYNTHETIC CYTOTOXICITY STRATEGIES FOR SELECTIVE KILLING OF ATM DEFICIENT CELLS  by  Xuesong Li  B. Sc., Fudan University, 2005  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE  in  The Faculty of Graduate Studies  (Genome Science and Technology)  THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)     April 2012  © Xuesong Li, 2012   ii Abstract Chromosome instability (CIN) is a hallmark of cancer cells and could, in theory, be exploited in the design of cancer therapeutics. Tumor cells harboring CIN mutations may be dependent on certain DNA repair pathways for viability. Thus, inhibition of specific DNA repair enzymes may enhance the CIN phenotype to an intolerable level, or may sensitize cells to DNA damage stress. To test this hypothesis, I focused on the CIN gene ATM, which is often mutated in human tumors. I hypothesized that knockdown of certain second site DNA repair genes would selectively kill ATM-deficient cells resulting in synthetic lethality (SL), or sensitize ATM-deficient cells to a sub-lethal dose of DNA damaging agent resulting in synthetic cytotoxicity (SC). The goal of this research is to use budding yeast as a model system to identify candidate SL or SC interaction partner genes for ATM with/without sub- lethal doses of DNA damaging agents, using mutations in the yeast ATM homologues, TEL1 and MEC1. I tested for interactions with TEL1 and MEC1 in a small matrix of three DNA repair genes (RAD27, TDP1 and TPP1) and four DNA damaging agents (hydroxyurea, 5- fluorouracil, bleomycin, and camptothecin). I also performed a genome-wide screen for interactions between TEL1 and ~5000 non-essential genes, both in the presence and absence of low doses of camptothecin. I discovered one SL interaction with MEC1 and fourteen SC interactions with TEL1.  Most of the SC interaction partner genes are involved in DNA repair and show sensitivity specifically to camptothecin. These data provide a rationale for testing specific combination therapies for selective killing of cancer cells bearing ATM mutations. Specifically, the Shu complex, Ku complex, Rrm3, Rad27 and CK2β subunits can be further tested as potential combination therapeutic targets, together with a sub-lethal dose of camptothecin, to kill ATM-deficient cancer cells.   iii Table of Contents  Abstract .............................................................................................................................................. ii Table of Contents ........................................................................................................................... iii List of Tables ..................................................................................................................................... v List of Figures .................................................................................................................................. vi List of Abbreviations ................................................................................................................... vii Acknowledgements ..................................................................................................................... viii Chapter 1 : Introduction ............................................................................................................... 1 1.1 Accumulation of mutations underlies cancer ................................................................... 1 1.2 Chromosomal instability (CIN) as a hallmark of cancer................................................ 3 1.3 Synthetic Lethality (SL) and Synthetic Cytotoxicity (SC) strategies to target CIN 4 1.4 CIN gene ATM mutations in cancer cell ............................................................................... 7 1.5 Conserved SL interactions and the ATM homologue gene in yeast ........................ 10 1.6 Overview of thesis .................................................................................................................... 11 Chapter 2 : Direct testing for SL and SC with ATM mutation in combination with three DNA repair enzymes ........................................................................................................ 13 2.1 Introduction ............................................................................................................................... 13 2.2 Material and methods ............................................................................................................. 16 2.2.1 Yeast strains and media .................................................................................................................... 16 2.2.2 DNA damaging agent sensitivity assay ....................................................................................... 18 2.2.3 Quantitative growth curve analysis ............................................................................................. 19 2.3 Results .......................................................................................................................................... 20 2.3.1 Most double mutant strains are viable without growth defects ...................................... 20 2.3.2 SL interactions between rad27 and mec1 mutations ........................................................... 21 2.3.3 Most double mutants show the same DNA damaging agent sensitivity as single mutants ................................................................................................................................................................... 22 2.3.4 SC interactions between rad27 and tel1 mutations .............................................................. 23 2.4 Discussion ................................................................................................................................... 25 Chapter 3 : Genome wide screen for synthetic interactions with TEL1 .................... 27   iv 3.1 Introduction ............................................................................................................................... 27 3.2 Materials and methods ........................................................................................................... 30 3.2.1 Yeast strains and media .................................................................................................................... 30 3.2.2 Genome-wide screens ....................................................................................................................... 33 3.2.2.1 tel1 deletion query strain construction ............................................................................................. 33 3.2.2.2 SC screen by synthetic genetic array (SGA) ..................................................................................... 34 3.2.3 Bioinformatic analysis ....................................................................................................................... 36 3.2.3.1 Fitness score and normalization .......................................................................................................... 36 3.2.3.2 SGA fitness data comparison by R ....................................................................................................... 36 3.2.4 Hit validation ......................................................................................................................................... 37 3.2.4.1 Random spore assay (RSA) ..................................................................................................................... 37 3.2.4.2 Validation by tetrad dissection followed by spot assay and growth curve analysis ..... 38 3.3 Results and discussion ........................................................................................................... 38 3.3.1 Camptothecin (CPT) concentration for SGA ............................................................................. 38 3.3.2 SC/SL candidate genes list obtained through two-step analysis ..................................... 42 3.3.3 No SL and 13 SC interactions with TEL1 were uncovered ................................................. 45 3.3.3.1 SHU genes ....................................................................................................................................................... 47 3.3.3.2 YKU genes ....................................................................................................................................................... 51 3.3.3.3 RRM3 gene ..................................................................................................................................................... 55 3.3.3.4 Casein kinase 2 beta subunit genes ..................................................................................................... 56 3.3.3.5 Spindle checkpoint genes and ERG5 gene ........................................................................................ 58 Chapter 4 : Conclusion, implications and future directions .......................................... 61 4.1 Summary ..................................................................................................................................... 61 4.2 Improvement of methodology ............................................................................................. 62 4.3 Clinical implications ................................................................................................................ 64 4.4 Conclusion ................................................................................................................................... 67 References ....................................................................................................................................... 68 Appendices ...................................................................................................................................... 77 Appendix A Other positive SC interaction hits validated by RSA .......................................... 77 Appendix B SHU1 SHU2 SC interactions with TEL1 by growth curve assay ...................... 83    v List of Tables  Table 2.1 DNA damaging agents ............................................................................................ 15 Table 2.2 Strains used in the study in Chapter 2..................................................................... 17 Table 2.3 Concentrations of DNA damaging agents for the sensitivity test ........................... 18 Table 3.1 Yeast strains used in the study in Chapter 3 ........................................................... 32 Table 3.2 Comparison of most sensitive genes to CPT .......................................................... 41 Table 3.3 Twenty-seven candidate genes (include linkage genes) list after SGA data analysis ................................................................................................................................................. 44 Table 3.4 SL/SC candidate validation .................................................................................... 46    vi List of Figures  Figure 1.1 SL strategy to target tumor cells with a CIN mutation ............................................ 5 Figure 1.2 SC strategy to target tumor cells with a CIN mutation ........................................... 7 Figure 2.1 Screen in 2X3X5 matrix ........................................................................................ 16 Figure 2.2 Inviable spores due to SL after tetrad dissection ................................................... 21 Figure 2.3 Example of no SC interaction (TDP1, TEL1/MEC1 and CPT) by spot assay ...... 22 Figure 2.4 SC interaction between RAD27 and TEL1 by spot assay ..................................... 23 Figure 2.5 SC interaction between RAD27 and TEL1 by growth curve assay ...................... 24 Figure 2.6 Quantitative SC analyses of fitness data from growth curve assay ....................... 24 Figure 3.1 Synthetic Genetic Array (SGA) ............................................................................. 29 Figure 3.2 SC screen using synthetic genetic array (SGA) .................................................... 35 Figure 3.3 Two steps to filter for SC and SL interaction hits ................................................. 43 Figure 3.4 SC interactions between SHU genes and TEL1 by spot assay .............................. 47 Figure 3.5 SC interaction between PSY3 and TEL1 by growth curve assay ......................... 48 Figure 3.6 SC interactions between SHU genes and TEL1 are specific to CPT .................... 49 Figure 3.7 SC interactions between SHU genes and TEL1 are SRS2 independent ............... 51 Figure 3.8 SC interactions between YKU genes and TEL1 ................................................... 52 Figure 3.9 Interactions between YKU genes and TEL1 are CPT specific ............................. 54 Figure 3.10 SC interactions between RRM3 and TEL1 ......................................................... 54 Figure 3.11 SC interactions between RRM3 and TEL1 are not specific to CPT ................... 56 Figure 3.12 SC interactions between CKB1 and TEL1 by RSA ............................................ 57 Figure 3.13 Only CK2 beta subunits show SC interactions with TEL1 ................................. 58 Figure 3.14 SC interactions between spindle checkpoint genes and TEL1 ............................ 59 Figure 3.15 SC interactions between ERG5 genes and TEL1 ................................................ 59    vii List of Abbreviations  AT: ataxia telangiectasia ATM: Ataxia Telangiectasia Mutated ATR: ATM- and Rad3-related AUC: area under the curve BER: base excision repair CIN: chromosome instability DDR: DNA damage response DMA: deletion mutant array DSB: double strand break dSLAM: diploid-based synthetic lethality analysis with microarrays GCR: gross chromosomal rearrangements IR: ionizing radiation HR: homologous recombination MIN: microsatellite instability MMR: mismatch repair MMS: methyl methanesulfonate NBD: nuclear binding domain OD: optical density OFP: Okazaki fragment processing PARP: poly (ADP ribose) polymerase PI3K-related: phosphoinositide 3-kinase-related RNAi: RNA interference SC: Synthetic Cytotoxicity SGA: Synthetic Genetic Array SL: Synthetic Lethality WT: Wild Type YKO: Yeast Knock-out deletion collection YPD: yeast peptone dextrose   viii Acknowledgements  I am especially grateful to my supervisor, Dr. Phil Hieter, who gave me a chance to pursue my master degree in his lab. His scientific guidance and experienced vision as well as continuous encouragement were of great help during my study. I am very fortunate to have trained with such a nice, patient, and knowledgeable mentor.  I also would like to thank all members of the Hieter lab for a supportive and learning environment. Derek van Pel introduced me to the basics of yeast biology and genetics, along with Peter Stirling whose almost limitless knowledge helped me get started early on. Jessica McLellan helped me plan experiments and trouble-shoot experimental difficulties. Jan Stoepel provided much helpful advice, especially with the statistical analysis of this study. I thank Nigel O'Neil for all his help, patience and advice throughout the course of this thesis. I am thankful to Irene Barrett and Megan Kofoed, who never hesitated to offer the technical guidance and assistance in the laboratory. I also thank Noushin Moshgabadi, Alina Chan, Sean Minaker, Melanie Bailey and Supipi Duffy for being very supportive and friendly towards me.  Finally, my thanks and utmost appreciation goes out to my beautiful and brilliant wife Chengcheng, without whose love, support, and patience none of this work would have been accomplished.       1 Chapter 1 : Introduction  1.1 Accumulation of mutations underlies cancer Although most spontaneous mutations are neutral, they can often be harmful rather than beneficial (Taddei et al., 1997). The survival of a species requires the protection of genetic material and the passage of intact accurately copied genetic information. This can be quite a challenge as DNA is constantly under attack by both endogenous and exogenous DNA damaging factors (Hoeijmakers, 2009). DNA damage can result from endogenous processes such as cellular metabolism or spontaneous errors in DNA replication and recombination (De Bont and van Larebeke, 2004). In addition, exogenous genotoxic agents, such as ultraviolet light, ionizing radiation (IR) and chemical mutagens, can cause nucleotide alterations and DNA breaks (Wogan et al., 2004, Friedberg et al., 2006). Unlike damaged protein, which can be replaced by synthesis, any modification of DNA has to be accurately repaired. Unrepaired DNA damage can lead to cell cycle arrest, cell death or mutations that may contribute to oncogenesis or other disease phenotypes (Vogelstein and Kinzler, 2001). Cells have evolved a response system to maintain and repair the organized state of DNA structure and the sequence of the nucleotide code. The DNA damage response (DDR) system activates the appropriate DNA repair pathway and arrests or slows the cell cycle to allow time for repair, or, in the case of irreparable damage, induces apoptosis (Jackson and Bartek, 2009). There are multiple pathways that must collaborate for the efficient repair of DNA damage and because of the critical role of genome maintenance mechanisms in all cells, DDR genes tend to be highly conserved throughout eukaryotes (Friedberg et al., 1995). Approximately 230 human DNA repair genes have been identified to date and most have homologues in budding yeast (Wood et al., 2005, Milanowska et al., 2011).   2 The integrity of the genome is crucial for tumor suppression in somatically proliferating cells (Hoeijmakers, 2001). To ensure faithful genome transmission eukaryotic cells have a remarkable DNA repair capacity (Vilenchik and Knudson, 2000). However, errors do happen, some of which escape the repair mechanisms and apoptosis, thus becoming permanent mutations. These mutations can be propagated to the next cellular generation and will accumulate with time. Accumulation of multiple mutations underlies many genetic diseases including cancer. During early cancer development, it is hypothesized that tumor cell precursors undergo a multistep mutational process to acquire a set of genetic changes that leads to oncogenesis. These oncogenic mutations in tumor cells result in a set of specific phenotypes or “hallmarks” that define tumor cells. These traits include uncontrolled proliferation, escape from apoptotic signals, and metastasis (Hanahan and Weinberg, 2000, Michor et al., 2004). Therefore, anticancer therapeutics are designed to specifically exploit one or more of these hallmarks to kill tumor cells.  The effectiveness of the most common chemotherapy agents is based on increased proliferation of cancer cells. Cytotoxic drugs, which are more toxic to proliferating cells, can kill the highly proliferative tumor cells. However, they also harm normal proliferating cells, such as cells in the bone marrow, digestive tract, and hair follicles. This results in many side effects, such as myelosuppression, mucositis and alopecia (Chabner and Roberts, 2005, Weinberg, 2007). Therapeutic effects could be improved by developing therapies that are highly selective for cancer cells, and do not rely on cell proliferation alone for their toxicity.       3 1.2 Chromosomal instability (CIN) as a hallmark of cancer The phenotypic hallmarks of cancer cells are caused by an accumulation of two main kinds of mutations: activating mutations in oncogenes and inactivating mutations in tumor suppressor genes (Weinberg, 2007). Usually, when cancer is detected, a large number of mutations have accumulated in tumor cells (Tomlinson et al., 2002). However, given the extremely high fidelity of replication, repair, and segregation of chromosomal DNA during normal cell division, the multiple cancer mutations are accumulating at a much higher rate than the spontaneous mutation rate in normal cells. So there is a third class of mutations that lead to genome instability that occur in the early development stage of tumorigenesis (Gatenby and Vincent, 2003, Hahn, 2004).  Many tumor cells acquire a mutator or chromosome mis-segregation phenotype early in their development, which increases the frequencies of errors that occur during replication, repair, and segregation of chromosomes in dividing cells, and therefore increases the likelihood of accumulating the other two categories of cancer gene mutations (Lengauer et al., 1997, Negrini et al., 2010).  Multiple pathways function to maintain genetic integrity in normal cells. Accordingly, genetic instability in cancer can be caused by different classes of mutations. There are at least two general types that are observed in tumor cells: 1) chromosome instability (CIN) involving changes of chromosome structure and number as a result of an abnormal mitosis; 2) microsatellite instability (MIN) involving deletions or amplifications at the nucleotide level, especially at short tandem repeats (microsatellites) due to loss of mismatch repair (MMR) (Storchova and Pellman, 2004). The majority of solid colorectal tumors (>80%) exhibit CIN (Rajagopalan et al., 2003), and it is believed that the emergence of CIN provides   4 a fundamental source of mutation to drive tumorigenic progression rather than being simply a passenger phenotype (Jallepalli and Lengauer, 2001, Vogelstein and Kinzler, 2004). A large amount of data from human tumors suggests that CIN correlates with tumor grade and prognosis (Kronenwett et al., 2004, Carter et al., 2006). In addition, the high mutation rates and heterogeneity of tumor cells caused by CIN can provide growth advantages particularly in cells challenged by chemotherapeutic agents. So CIN contributes to both tumor evolution and therapeutic resistance (Rajagopalan and Lengauer, 2004). However, because CIN is specifically present in most types of cancer cells and not normal cells, it could be exploited for selective killing of tumor cells based on that phenotypic difference.  1.3 Synthetic Lethality (SL) and Synthetic Cytotoxicity (SC) strategies to target CIN One strategy to target CIN is to identify gene mutations that are synthetic lethal with mutations that cause CIN. Synthetic lethality (SL) is a phenomenon that is well-characterized and studied in yeast and the fruit fly: two gene mutations that have little or no effect on cell viability individually but cause lethality when combined in a double mutant are said to be synthetic lethal (Kaelin, 2005). Early in 1997, Hartwell and colleagues hypothesized that cancer cells harboring known mutations or deletions might represent genetically sensitized cells relative to the normal surrounding cells that may be susceptible to drug therapies selectively targeting a second gene product (Hartwell et al., 1997). Recently this approach has proven effective in targeting cancer cells with loss of function of the homologous recombination (HR) pathway for DNA double-strand break repair (BRCA1/BRCA2 mutation), using small molecule inhibitors of PARP, a protein involved in DNA single-strand break repair (Bryant et al., 2005, Farmer et al., 2005). The CIN as a “driver” mutation in   5 cancer cells may therefore create opportunities for selective killing of cancer cells by mimicking the effect of the second genetic mutation with a targeted therapy (McManus et al., 2009, McLellan et al., 2009, McLellan et al., 2012) (Figure1.1).   Figure 1.1 SL strategy to target tumor cells with a CIN mutation Tumor cells harboring a CIN mutation may require certain DNA repair pathways for viability. The inhibition or mutation of the DNA repair enzyme in this pathway may selectively kill the tumor cell as a monotherapy.  DNA repair enzymes might be good candidates for targeting tumors for SL (Shaheen et al., 2011). In normal proliferating cells, DNA lesions are detected and repaired by the DNA damage response machinery. Many CIN proteins are involved in damage-induced signaling cascades and various DNA repair pathways (Paulsen et al., 2009, Schvartzman et al., 2010). The tumor cells harboring CIN mutations may exhibit a high level of endogenous DNA lesions that result in genetic instability. While being the origin of DNA mutation, CIN may also be a weakness: for example, causing reduced ability to replicate DNA and proliferate   6 (Negrini et al., 2010). Successful CIN cells have acquired additional mutations that adapt its genome maintenance system to cope with these genome imbalances. It could be said that some tumor cells become addicted to certain DNA repair pathways and would be sensitive to the inhibition of these pathways. Hence, in tumors with defined somatic CIN gene mutations, certain DNA repair enzymes may be excellent drug targets as monotherapies with high efficacy.  The level of DNA damage occurring spontaneously in cancer cells might be not enough to convert to fatal lesions and trigger cell death in the absence of certain DNA repair functions. Therefore, not all tumors will be susceptible to DNA repair inhibitor monotherapies. It is possible that these tumors will be sensitive to a combination therapy of DNA repair enzyme inhibitor and DNA damaging agent. Inhibiting the function of DNA repair enzymes could synergize with somatic mutations in the tumor cell to sensitize the tumors to DNA-damaging therapeutic agents resulting in Synthetic Cytotoxicity (SC) (Figure 1.2). In addition to the potential to improve efficacy and selectivity of treatment, another benefit of using the SC strategy to target cancer cells is to minimize the cytotoxic side effects because SC would permit significant reductions in the dose of cytotoxic agents from those used in standard chemotherapy. Furthermore, the toxicity of chemotherapy can be reduced by the increased activities of several DNA repair pathways (Helleday et al., 2008).  So by identifying and targeting DNA repair enzymes as damage agent-enhancers, it may be possible to overcome natural and acquired tumor cell chemoresistance. Currently, clinical trials are assessing the efficacy of combination therapy with PARP inhibitors coupled with DNA damaging agents such as carboplatin, gemcitabine, and temozolomide (Chan and Giaccia, 2011). Although the   7 molecular mechanism behind the increased sensitivity to these therapeutics remains to be elucidated, these findings demonstrate the power of DNA repair enzyme inhibitors as sensitizers of tumor cells to chemotherapy. The same SC strategy could be used to target tumors with a CIN gene mutation.   Figure 1.2 SC strategy to target tumor cells with a CIN mutation The inhibition or mutation of certain DNA repair enzymes may sensitize the tumor cell harboring a CIN mutation to low doses of a DNA damaging agent resulting in Synthetic Cytotoxicity (SC).   1.4 CIN gene ATM mutations in cancer cell To test SL and SC strategies to target tumor cells with CIN mutations, I chose the ATM (Ataxia Telangiectasia Mutated) gene, a CIN gene frequently mutated in many types of cancer and which represents the major CIN mutational spectrum of a tumor cell.    8 The ATM gene is a large gene encoding a 13 kb transcript. It has 66 exons extending over 160kb of genomic DNA (Shiloh, 2003). This gene was first discovered in the hereditary cancer-prone human disease ataxia telangiectasia (AT) (Savitsky et al., 1995). People with this disorder have mutations in both copies of the ATM gene in each cell. It is an autosomal recessive disorder which is caused by absent or inactivated ATM protein (Broeks et al., 2000). Researchers have identified several hundred mutations in the ATM gene in AT patients. The majority of these mutations are nonsense changes that lead to the generation of a premature stop codons and give rise to truncated proteins (Stankovic et al., 1998). People with AT have a 40% risk of developing cancer. The most common types of cancer seen are leukemia and lymphoma and often appear in childhood (Boultwood, 2001).  As AT patients live longer, there appears to be an increased risk of other cancer types, including breast cancer, ovarian cancer, stomach cancer, melanoma, and sarcoma (Ahmed and Rahman, 2006, Kastan and Bartek, 2004, Renwick et al., 2006).  AT cells exhibit radiosensitivity, altered DNA repair, cell cycle checkpoint defects and chromosomal instability (Lavin, 2008).  AT is rare disease with a frequency of ∼1/40,000 (Savitsky et al., 1995). However, about 1% of the western populations are ATM heterozygotes (Renwick et al., 2006). It has been estimated that ATM gene mutation carriers have an increased hereditary pancreatic cancer risk (Fillon, 2012). They also have a 2.2-fold higher risk of developing breast cancer, and this number increases to nearly 5-fold for women under 50 (Thompson et al., 2005). The genomic instability present in the primary cells from AT patients and the high propensity of ATM mutation carriers to develop cancer strongly support the idea that CIN gene mutations in ATM drive tumor development by increasing the spontaneous mutation rate.   9 In addition to hereditary cancer, CIN gene ATM mutations were often detected in diverse sporadic cancers as well. Several studies have reported a high prevalence of ATM gene mutations in sporadic lymphoid cancer (Stankovic et al., 2002, Gumy-Pause et al., 2004). Recently, research showed that ATM is the 7th most frequently mutated gene in lung adenocarcinoma, detected in nearly 10% of cases (Kang et al., 2008). Mutation or reduced expression of ATM protein was found in 15% of gastric cancers. Low ATM protein levels were also detected in 10-20% of breast cancer cases and a strong correlation between ATM disruption and breast cancer was suggested (Hall, 2005).  The phenotype of AT patients and of cells carrying ATM mutations provided many clues to understand the function of ATM protein. ATM is a member of the phosphoinositide 3- kinase-related (PI3K-related) kinase family and is mainly involved in initial response to double-stranded DNA breaks (DSB) as a principal sensor. The activation of ATM in response to DSBs leads to phosphorylation and activation of multiple downstream protein effectors, which in turn results in the delay of cell cycle progression to facilitate DNA repair or the induction of apoptosis to eliminate the damaged cells. Many of these integrated cellular response processes are also required for subsequent activation of another master checkpoint kinase, the ATM- and Rad3-related (ATR) protein which also belongs to the PI3K-related kinase family (Shiloh, 2003).  Because of its central role in the DNA damage response, ATM is an important protein for preventing malignant transformation through induction of cell cycle checkpoints, senescence and apoptosis. However, in cancer cells, the response to genome alterations is commonly   10 impaired owing to accumulation of mutations in the ATM checkpoint regulator. Such alterations disrupt the ability of the cell to maintain genomic stability, generating the CIN and mutator phenotype, and making the acquisition of additional genomic alterations in the cancer cell more likely. Due to well-established problems in the means of gene delivery, therapeutic restoration of ATM is currently not a viable option for cancer therapy. However, an ATM mutation inducing CIN may represent a weakness relative to normal cells that can be exploited for tumor therapy. Therefore, the study to exploit SL and SC strategies-based target therapy through investigating the synthetic genetic interactions with ATM mutation is of interest in cancer research.  1.5 Conserved SL interactions and the ATM homologue gene in yeast Given the large number of DNA repair enzymes and various types of DNA damaging agents, it is necessary to develop a robust method to screen for combinations of gene pair and DNA damaging agents for synthetic lethal and synthetic cytotoxic interactions with ATM mutations. Since genome stability and DNA repair are basic and fundamental cellular processes, the mechanisms and genes involved are highly conserved among eukaryotes (Kitagawa and Hieter, 2001). Many insights into genetic interactions have been obtained from the model organism Saccharomyces cerevisiae, in which a genome-wide deletion collection has been generated that can be used with high-throughput screening techniques (Scherens and Goffeau, 2004). Instead of searching for SL and SC genetic interactions with ATM mutation directly in human cells, I use a budding yeast with a dysfunctional ATM homologue gene to mimic ATM mutations in cancer cells. I then knocked out second site DNA repair genes to mimic the inhibition of the DNA repair enzyme. Yeast is a genetically   11 amenable model organism with a short life cycle that greatly facilitates my research and allows the rapid translation of data from the genetic networks with ATM to Human. In this way, I can narrow down the number of synthetic lethal and cytotoxic candidate gene pairs and then directly test these candidates in human cells. This approach is based upon the observation that many synthetic interactions are conserved across species (McManus et al., 2009, McLellan et al., 2009).  The yeast homologues of ATM and the similar PI3K-related kinase ATR are TEL1 and MEC1, respectively.  TEL1 and MEC1 have overlapping functions in mediating the cellular response to DNA damage.  While ATM/TEL1 is not essential for growth in mammals and yeast, loss of function of the related kinase ATR/MEC1 is lethal in both human and budding yeast (Brown and Baltimore, 2000, de Klein et al., 2000).  However, MEC1 deletion mutants are viable when an inhibitor of ribonucleotide reductase, SML1, is also deleted (Zhao et al., 1998).  Therefore, I can create a set of clean deletion strains for my assays bearing tel1 or mec1 deletions or tel1 and mec1 double deletions. sml1 deletion strains will serve as an additional “wild type” control for any strain deleted in MEC1 to ensure that SML1 is not having an effect.  1.6 Overview of thesis I propose to use the model organism Saccharomyces cerevisiae to screen for synthetic lethal interactions between DNA repair genes and the CIN gene ATM, and for synthetic cytotoxic interactions in combination with sub-lethal doses of DNA damaging agents. I hypothesize that knockdown of certain second site DNA repair genes will selectively kill TEL1 deficient   12 cells resulting in synthetic lethality, or sensitize TEL1 deficient cells to DNA damaging agents resulting in synthetic cytotoxicity.  As proof of principle for this hypothesis, I first directly tested genetic interactions in a small matrix of three DNA repair enzyme genes and then used high throughput genetic methods to screen a large matrix of the complete non-essential gene deletion collection in budding yeast to discover synthetic lethal and synthetic cytotoxic relationships between TEL1, DNA repair enzymes and sub-lethal doses of DNA damaging therapeutic agents. SL and SC interactions data derived in model organisms may or may not be conserved in humans and will need to be further validated.  This study would provide the basis for monotherapies and combination therapies for selective killing of ATM deficient cancer cells. The identification of negative genetic interactions (the extreme example: SL) with ATM mutation can be exploited directly as drug targets for monotherapies for ATM deficient cancer. Furthermore, the genes that upon knockdown sensitize cells to a specific DNA damaging agent can be used as targets in combination therapy.       13 Chapter 2 : Direct testing for SL and SC with ATM mutation in combination with three DNA repair enzymes  2.1 Introduction Studies show that the Saccharomyces cerevisiae homologue of ATM, TEL1 plays a similar role in maintaining genome stability as a central mediator of the DNA damage and replication checkpoints (Ritchie et al., 1999, Zhou and Elledge, 2000). Also, based on the observation of molecular structures, the homology between ATM and TEL1 is not restricted to the catalytic domain at the C terminus of the protein, but rather extends over the whole length (Mallory and Petes, 2000). Unlike its homologue ATM, TEL1 is also involved in telomere maintenance. TEL1 was originally isolated as a mutant that had stable shorter (about 50 bp) telomeres (Greenwell et al., 1995, Lustig and Petes, 1986). Although the phenotype of TEL1 mutated yeast cells are not identical to ATM mutated mammalian cells, it has been observed that the TEL1 gene, when transformed into an AT cell line, could complement many of the mutant phenotypes associated with the ATM mutation (Fritz et al., 2000).  In human cells, many proteins involved in cell cycle checkpoints and DNA repair are phosphorylated in an ATM-dependent manner in response to DNA damage, including many proteins involved in the homologous recombination (HR) repair pathway (Mre11, Nbs1, Rad50, Rad51, Rad52, Rad54, RPA, Exo1, CtIP and BRCA1/2) and several non-homologous end joining (NHEJ) repair pathway core proteins (DNA-PKcs, XLF and Artemis) (Bensimon et al., 2010). HR and NHEJ are the two main pathways for DNA DSB repair in human cells (Mao et al., 2008). Although ATM has many substrates, its contribution to the DNA damage   14 response is partially masked by ATR. Many IR-induced phosphorylation sites targeted by ATM can also be phosphorylated by ATR (Matsuoka et al., 2007). In yeast, TEL1 has been shown to partially restore the DNA damage checkpoint in cells with a MEC1 (ATR in humans) deficiency (Clerici et al., 2004). And cells lacking both TEL1 and MEC1 are more sensitive to DNA damaging agents than either single mutant. Also, because of this redundancy in function, the TEL1 or MEC1 mutation alone demonstrates low or moderate effects on genome stability, while loss of both genes leads to synergistically elevated levels of genome instability in budding yeast.  Genetic interactions of the CIN genes TEL1 and MEC1 have been widely explored in the model organism budding yeast. Only a few SL interactions with TEL1 have been found and most of the SL partners are essential genes, such as DNA2 and MEC1 (Budd et al., 2005, Chakhparonian et al., 2005). In S. cerevisiae, mec1 null cells are inviable, so most genetic interaction data are based on hypomorphic mec1 mutations or in combination with the sml1 deletion background. In my primary screen in budding yeast for SL and SC genetic interactions with TEL1 or MEC1 mutations, I focused on three DNA repair enzymes, tyrosyl-DNA-phosphodiesterase 1 (TDP1), DNA 3’phosphatase (TPP1/PNKP in humans) and RAD27 5’flap endonuclease (FEN1 in humans), for which small molecule inhibitors have been identified or are being developed (Huang et al., 2011, Freschauf et al., 2009).  In my study, the inhibition of the DNA repair enzyme is modeled through knockout of the corresponding yeast homologue. The combination of mutations in three repair enzyme genes (TDP1, TPP1, RAD27) and two CIN genes (TEL1, MEC1) results in a 2x3 matrix of double   15 gene deletion strains. Any inviable double mutant combinations of CIN gene and DNA repair gene will be considered SL and can be prioritized for validation in human cells as a potential single-agent target.  On the other hand, viable double mutants will be screened for synthetic cytotoxicity by exposing them to sub-lethal doses of various classes of DNA-damaging agents.  In order to test whether SC occurs with general DNA damage or a specific lesion type, four DNA damaging agents that cause different types of lesions, were chosen: bleomycin (BLEO), 5-fluorouracil (5FU), camptothecin (CPT) and hydroxyurea (HU) (Helleday et al., 2008) (Table 2.1). After the construction of all double gene deletion strains between CIN genes and DNA repair genes, their sensitivities to 4 types of DNA damage were tested by plate assays on YPD media with/without corresponding DNA damaging agents, resulting in 2X3X5 matrix screen (Figure 2.1). The SC interactions were quantified by growth curve analyses.  DNA damaging agent Compound Description Toxic Lesions 5-Fluorouracil (5FU) Antimetabolite Base damage; Replication lesions Bleomycin (BLEO) Radiomimetic DNA-cleaving agents DNA strand breaks; Base damage Camptothecin (CPT) Topoisomerase I inhibitor DNA strand breaks; Replication lesions Hydroxyurea (HU) Ribonucleotide reductase inhibitor Stalled/collapsed replication forks Table 2.1 DNA damaging agents Four DNA damaging agents that cause different types of lesions were chosen for DNA damage sensitivity test: bleomycin (BLEO), 5-fluorouracil (5FU), camptothecin (CPT) and hydroxyurea (HU) (Helleday et al., 2008).   16  Figure 2.1 Screen in 2X3X5 matrix The combination of double mutations of three repair enzyme genes (TDP1, TPP1 and RAD27) and two CIN genes (TEL1 and MEC1) results in a 2x3 matrix of double gene deletion strains. All double mutants were first tested for synthetic lethality by growing in media without DNA damaging agent and then screened for cytotoxicity by exposing to sub- lethal doses of four DNA-damaging agents: bleomycin (BLEO), 5-fluorouracil (5FU), camptothecin (CPT) and hydroxyurea (HU).  2.2 Material and methods 2.2.1 Yeast strains and media All strains are BY4743 background (Brachmann et al., 1998). The heterozygous double mutant strains were constructed by mating each of the corresponding single mutants. The haploid double mutant strains were generated by sporulation of diploid heterozygous double mutant strains followed by tetrad dissection. The genotypes of all the strains used in the experiment were checked by PCR for confirmation of the gene deletions (Table 2.2).  Yeast was grown in rich media at 30°C (Amberg et al., 2005). Plasmid bearing strains were grown in synthetic complete media lacking the appropriate nutrient.     17 Strain Genotype BY4741 MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 BY4742 MATα his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 PSY437 MATa his3 leu2 lys2 ura3 tel1::KANMX PSY438 MATα his3 leu2 lys2 ura3 tel1::KANMX PSY765 MATα his3 leu2 lys2 ura3 mec1::HYG sml1::KANMX PSY766 MATa his3 leu2 lys2 ura3 mec1::HYG sml1::KANMX YDV103 MATa his3 leu2 lys2 ura3 rad27::HYG YDV104 MATα his3 leu2 LYS2 ura3 rad27::HYG YNM1 MATa/MATα his3/his3 leu2/leu2 LYS2/lys2 MET15/ met15 TRP1/TRP1 ura3/ura3 CAN1/can1::LEU2-MFA1pr-HIS3 TDP/tdp1::KANMX YNM10 MATα his3 leu2 lys2 ura3 tel1::NAT YNM11 MATα his3 leu2 lys2 ura3 mec1::NAT sml1::KANMX YNM12 MATα his3 leu2 lys2 ura3 mec1::HYG sml1::NAT YNM14 MATa his3 leu2 LYS2 TRP1 ura3 tpp1::KANMX YNM15 MATa his3 leu2 lys2 ura3 rad27::HYG  sml1::KANMX YNM16 MATa his3 leu2 lys2 ura3 rad27::HYG sml1::KANMX YNM17 MATa his3 leu2 lys2 ura3 tdp1::KANMX tel1::NAT YNM18 MATa his3 leu2 lys2 ura3 tdp1::KANMX tel1::NAT YNM2 MATa/MATα his3/his3 leu2/leu2 LYS2/lys2 MET15/ met15 TRP1/TRP1 ura3/ura3 CAN1/can1::LEU2-MFA1pr-HIS3 TPP/tpp1::KANMX YNM20 MATα his3 leu2 lys2 ura3 tdp1::KANMX tel1::NAT YNM22 MATα his3 leu2 lys2 ura3 tdp1::KANMX mec1::HYG sml1::NAT YNM23 MATa his3 leu2 lys2 ura3 tpp1::KANMX tel1::NAT YNM24 MATα his3 leu2 lys2 ura3 tpp1::KANMX tel1::NAT YNM25 MATa his3 leu2 LYS2 ura3 tpp1::KANMX mec1::HYG sml1::NAT YNM26 MATα his3 leu2 lys2 ura3 tpp1::KANMX mec1::HYG sml1::NAT YNM27 MATa his3 leu2 lys2 ura3 rad27::HYG tel1::NAT YNM28 MATα his3 leu2 lys2 ura3 rad27::HYG tel1::NAT YNM3 MATa his3 leu2 LYS2 TRP1 ura3 tdp1::KANMX YNM30 MATα his3 leu2 LYS2 ura3 rad27::HYG sml1::KANMX YNM31 MATa his3 leu2 lys2 ura3 tel1::KANMX mec1::HYG sml1::NAT YNM32 MATα his3 leu2 lys2 ura3 tel1::KANMX mec1::HYG sml1::NAT YNM4 MATα his3 leu2 lys2 TRP1 ura3 tdp1::KANMX YNM6 MATα his3 leu2 lys2 TRP1 ura3 tpp1::KANMX YNM7 MATa his3 leu2 lys2 ura3 rad27::NAT YNM8 MATα his3 leu2 LYS2 ura3 rad27::NAT YNM9 MATa his3 leu2 lys2 ura3 tel1::NAT Table 2.2 Strains used in the study in Chapter 2    18 2.2.2 DNA damaging agent sensitivity assay Both plate assays and growth curve analyses were utilized to identify SC interactions by comparing the agent sensitivity of double mutant strains with the two single mutant parental strains.  All cells were grown in YPD media without DNA damaging agent until log phase before spotting on YPD plates or addition to YPD liquid media containing sub-lethal doses of the DNA damaging agents. OD600   values were checked in order to transfer the same number of single and double mutant cells. The sub-lethal doses of the DNA damaging agents were decided based on previous tests of the fitness of single mutants in serially diluted concentration of agents on plates or in liquid media. The concentration at which the growth of one parental single mutant was mildly affected was used as the inhibitory concentration. The double mutants and parental strains being tested were exposed to three concentrations (inhibitory concentration diluted 2X, 8X, and 16X) of each DNA-damaging agent during on plate assay and growth curve analyses (Table 2.3). DNA damaging agents Concentrations for on plate assay Concentrations for growth curve assay Camptothecin (CPT) 1μg/ml 2 μg/ml 8 μg/ml 0.5μg/ml 1 μg/ml 4 μg/ml Bleomycin (BLEO) 0.5μg/ml 1 μg/ml 4 μg/ml 0.1μg/ml 0.2μg/ml 1 μg/ml 5Fluorouracil (5FU) 2.5μg/ml 5 μg/ml 20 μg/ml 1 μg/ml 2 μg/ml 8 μg/ml Hydroxyurea (HU) 10 mM 20 mM 80 mM 2 mM 4 mM 16 mM Table 2.3 Concentrations of DNA damaging agents for the sensitivity test    19 For on plate spot assays, an identical optical density (OD600) of cells was serially diluted 5- fold and spotted on the indicated plate at the indicated temperature for 72 hours. For growth curve analysis logarithmic phase cultures were diluted to an OD600   value of 0.15±0.02 in 96- well plates and grew for 24 hours in a TECAN M200 plate reader at 30°C. Each strain was tested in three replicates. 2.2.3 Quantitative growth curve analysis The fitness of strains in growth curve analysis is often calculated by two ways: doubling time (the time required for a population to double in size during log phase); and specific growth rate (the maximum growth rate estimated through the slope) (Shah et al., 2007). However, both calculations only focus on the exponential growth region, a small area of the growth curve. They work well with most of the healthy strains, but when estimating very sick strains with slow growth, it is hard to define the exponential growth regions. In addition, the ability to adapt to growth conditions is different between mutants. And as population increases, the limited pool of resource and toxic waste buildup prevent growing beyond certain threshold. So the resulting lag and plateau phases are also very important factors when estimating fitness, especially for the mutants with growth defects. To avoid these issues, I used the area- base calculation method developed by Kevin Ushey (UBC Statistics MSc student). Strain fitness F was defined as the presentation of the logarithm of the area under the curve (AUC) of mutants relative to wild type (Ushey, 2011, McLellan et al., 2012).  The genetic interactions are quantified through the comparison of fitness of double mutants and expected phenotype base on two single mutants. Mutations in independent genes (two   20 genes with a neutral interaction) often combine in a multiplicative manner. The expected fitness of resulting double mutant is assumed to be Fab*=Fa x Fb. The interaction can be quantitatively measured by comparing double mutant Fab against expected Fab Mani et al., 2008 * ( ).  2.3 Results 2.3.1 Most double mutant strains are viable without growth defects All double heterozygous deletion strains for genetic analysis were generated by mating two haploid single deletion strains bearing either the CIN or DNA repair gene deletion. After sporulation, the viability of double mutants was first checked by the growth of double mutant offspring after tetrad dissection. Most spores from tel1 and DNA repair gene double heterozygous deletion diploids were viable. When selected for the antibiotic resistance maker, most tetrads exhibited 2:2 segregation patterns. Because MEC1 is an essential gene, we used a mec1 sml1 double mutant, which is viable. As expected, only half of the offspring spores generated from triple heterozygous deletion strains (mec1Δ sml1Δ tdp1Δ or mec1Δ sml1Δ tpp1Δ) were viable.  All 5 combination double mutants grew similar to wild type, and did not show sick/slow growth on plates. The viable double deletion strains were further tested for synthetic sickness by comparing to parental single deletion strains. The fitness relative to wild type was quantified using liquid growth curve analysis (Ushey, 2011, McLellan et al., 2012), which show no synthetic growth defects.    21 2.3.2 SL interactions between rad27 and mec1 mutations For the mec1 rad27 double heterozygote, only 13 in total 55 spores derived from asci separated on YPD plates were viable suggesting other reasons causing death beside the separation of mec1 deletion and sml1 deletion. mec1 sml1 rad27 triple deletion haploid strains could not be isolated after replication to selection media plates. The genotype of some inviable spores deduct from the rest spore on selection plate should be triple deletion (representative tetratype tetrad is shown in Figure 2.2). A rad27 mec1 double heterozygous deletion and sml1 homozygous diploid was constructed to eliminate the effect of SML1 and confirm the rad27 mec1 synthetic lethal interaction. Inviability of the rad27 mec1 double mutant synthetic lethality was confirmed in 22 tetrads, and no viable double deletion haploid was found.   Four haploid spores from one mec1 sml1 rad27 triple heterozygous deletion diploid cell were dissected onto an YPD plate and replica-plated onto marker selection plates with different antibiotics: YPD HYG and YPD NAT. Only rad27Δ::Hyg mutants can survive on YPD HYG plates while mec1Δ::NAT mutants can grow on YPD NAT plates. There were three viable spores that grew to colonies on the YPD plate. Marker segregation patterns were 2:2, as expected. The genotype of the inviable spores could be deducted from the growth of viable spores on selection plates and corresponded to rad27Δ::Hyg mec1Δ::NAT deletion mutants.   Figure 2.2 Inviable spores due to SL after tetrad dissection   22 2.3.3 Most double mutants show the same DNA damaging agent sensitivity as single mutants Comparing the concentrations of the DNA damaging agents used in the on plate spot assay and in liquid growth curve assay, it was clear that yeast cells are much more sensitive to DNA damaging agents in liquid than on the surface of solid media. Among all single deletion strains, mec1Δ and rad27Δ are more sensitive to DNA damaging agents especially camptothecin and bleomycin, tel1 null mutants are mildly sensitive to CPT. tpp1 and tdp1 single deletion strains grow similar to wild type and double mutants show similar sensitivity with their parental mec1 or tel1 single deletion strains. In addition, growth curve assays to quantify the interaction also show the same results.  Figure 2.3 Example of no SC interaction (TDP1, TEL1/MEC1 and CPT) by spot assay A serial 5x dilution of cells was spotted on YPD plates with/without CPT. On YPD plates all mutant strains grow similarly as compared to wild type. rad52Δ mutant which is hypersensitive to CPT, was used as positive control. tdp1Δ strain was not sensitive to CPT even at 16 µg/ml CPT. tdp1Δ mec1Δ sml1Δ mutants show slow growth on YPD plate with 8 µg/ml CPT, similar to mec1Δ sml1Δ mutants. tdp1Δ tel1Δ double mutants grow sick similarly with tel1Δ mutant on 16 µg/ml CPT plate.  Synthetic cytotoxicity of viable double mutants was directly tested by plating onto YPD plates with sub-lethal doses of DNA damaging agents and comparing sensitivity to the two single mutant parental strains. The double mutants and parental strains were tested along with the wild type as native control and rad52 single deletion as positive control (one example of spot assay for tdp1Δ CPT sensitivity test is shown in Figure 2.3). Tdp1 participates in the   23 repair of topoisomerase I inhibitor-induced DNA damage, but knock out of TDP1 in budding yeast is not sensitive to CPT (Pouliot et al., 2001). This is consistent with the observation on spot assays that the tdp1Δ mutant did not show a growth defect on CPT plates, suggesting the existence of other repair mechanisms in addition to Tdp1 to remove Top1-DNA cleavage complexes trapped by CPT (Liu et al., 2002).  2.3.4 SC interactions between rad27 and tel1 mutations The spot assays suggested rad27 and tel1 exhibited a clear SC interaction on CPT plates and weak negative interaction on HU plates (Figure 2.4). However, on BLEO and 5FU plates, double mutants grew similar to rad27Δ single mutants indicating the sensitivity of double mutants is specific to CPT and HU induced DNA lesions. The growth curve assays show the similar results (Figure 2.5). In 0.5 µg/ml CPT, the double mutant growth curve was much slower than either single mutant growth curve, about 7% worse than the prediction, indicating a SC interaction (Figure 2.6).  Figure 2.4 SC interaction between RAD27 and TEL1 by spot assay A serial 5x dilution of cells was spotted on YPD plates containing different DNA damaging agents. On YPD control plates, tel1Δ grows as well as wild type. Both rad27Δ single mutant and rad27Δ tel1Δ mutant grow a little slower than wild type. On CPT plate, rad52Δ (positive control) and rad27Δ tel1Δ mutants are hypersensitive to CPT. tel1Δ is slightly sick compared to wild type, while rad27Δ shows a mild growth defect. On HU plates, the double mutant is a little sicker than rad27Δ and relatively more sensitive compared to tel1Δ. On BLEO and 5FU plates growth of the double mutant is similar to rad27Δ single mutant.    24  Figure 2.5 SC interaction between RAD27 and TEL1 by growth curve assay In YPD control media, tel1Δ grows as same as wild type, while rad27Δ and rad27Δ tel1Δ mutants grow a little slower. A low dose of CPT (0.5µg/ml) shows no/little effect on all strains except double mutants. In YPD media with 1 µg/ml CPT, the growth of double mutants is much slower than single mutants. The growth curves shown are using average reads from 3 replicates data.   Figure 2.6 Quantitative SC analyses of fitness data from growth curve assay The fitness compared to wild type during the growth curve assay is calculated using curve area based methods (Ushey, 2011, McLellan et al., 2012). The percentages of each strain relative to wild type are shown above the line and average numbers are noted. The expected fitness values of double mutants with no interaction between rad27 and tel1 are obtained by multiplying two single mutant fitness values together. The actual fitness of rad27Δ tel1Δ double mutants in the presence of 1 µg/ml CPT is about 7% lower than expected, indicating a SC interaction between rad27 and tel1.   25 2.4 Discussion Similar to their homologues ATM and ATR in human cells, MEC1 and TEL1 initiate the DNA damage response in response to single-strand and double-strand DNA breaks. A cascade of protein kinases are activated to arrest or slow the cell cycle, allowing time to repair the damage (Clerici et al., 2004). FEN1/RAD27 is highly conserved through evolution. RAD27 encodes a nuclease involved in many aspects of DNA metabolism including processing Okazaki fragments during DNA replication, base excision repair (BER), and other pathways maintaining genome stability (Liu et al., 2004). Inactivation of Rad27 in budding yeast leads to the accumulation of double-strand breaks, chromosome instability, and sensitivity to methylmethane sulfonate (MMS) (Debrauwere et al., 2001). I have found that rad27 null mutants are also sensitive to CPT. Because of its multifunctional role, more than 30 SL interactions with RAD27 have been reported that include many DNA damage checkpoint genes including RAD9, RAD24, RAD17 and MEC1 (Pan et al., 2006), which is consistent with my experimental results. It suggests that double-strand breaks created in rad27 mutants may require the TEL1/MEC1 checkpoint to delay cell division to allow DNA damage repair to maintain viability. The three DNA enzymes tested during the first screening did not provide any hint of the mechanism behind the only SC interaction identified. Also because of the multiple functions of RAD27 and the indirect DNA damage caused by camptothecin, there are still many questions about this SC interaction between RAD27 and TEL1 in the presence of CPT. But the discovery of SC interaction in yeast proved my hypothesis and suggested SC interaction could be specific to certain types of DNA lesions.    26 Furthermore, the direct screening of the matrix (2 CIN genes X 3 DNA repair enzyme genes X with/without 4 damage agents) revealed FEN1/RAD27 SL with ATR/MEC1 and SC to camptothecin with ATM/TEL1 in budding yeast. This work suggests a new target for combination therapy: FEN1 inhibitor combined with low non-toxic dose of camptothecin may enhance cytotoxicity to ATM deficiency cancer cell.   27 Chapter 3 : Genome wide screen for synthetic interactions with TEL1  3.1 Introduction From directly testing a 2 gene X 3 gene X 4 DNA damaging agent matrix, I found that the rad27 tel1 double mutant shows a strong Synthetic Cytotoxicity (SC) to CPT treatment in yeast. Because Rad27 is a multifunctional protein with roles in DNA replication, base excision repair (BER) and double-strand break repair (DSBR) pathways (Liu et al., 2004), the mechanism behind the SC with TEL1 is not clear. Testing of more synthetic interactions with TEL1 (i.e. with more than one DNA repair gene in certain repair pathway) would provide a more comprehensive view and may lead to a better insight into the mechanism. In addition to DNA repair enzymes, genes involved in other process may lead to SC as well, such as other CIN genes or proteins that exclude DNA damaging agents from the cell. To get a better view of the TEL1 SC network and to identify more candidate genes that target tel1 deficient cells, I expanded my search to the collection of non-essential gene deletions in yeast using high throughput Synthetic Genetic Array (SGA) technology.  With the SGA technique, it is possible to systematically construct  double mutants by introducing the tel1 deletion allele in a query strain and mating to an ordered array of ~5000 non-essential gene deletion mutants (representing ~80% of all yeast genes) (Tong and Boone, 2006). (Figure 3.1) The generated single and double deletion haploid strains were transferred and grown on selective plates separately. Because the amount of cells of each strain transferred by robotic pinning are relatively similar, the fitness of haploid mutants can be estimated based on strain spot size on the final plate. By comparing the fitness of single mutants and their corresponding double mutants, a quantitative measurement of genetic   28 interactions can be determined. In this way, I could identify all non-essential SL partner genes with tel1. In parallel, each strain was replicated onto media containing sub-lethal doses of DNA damaging agent, and comparison of the growth on plates with and without DNA damaging agent demonstrates the differential sensitivity of every single and double deletion strain. The SC partner genes were identified when single gene deletion strains exhibited low sensitivity to DNA damaging agent and the corresponding double mutants exhibited hypersensitivity. All hits identified in high throughput fashion could be further validated by tetrad dissection or random spore assay.  Because the only SC interaction identified in the previous experiment in the 2X3X5 matrix shows specificity to CPT, I chose CPT as the DNA damaging agent in the SGA screen. All SC hits identified during SGA screening could be tested with other DNA damaging agents to test whether the SC was specific for a type of lesion or was for DNA damage in general.    29  Figure 3.1 Synthetic Genetic Array (SGA) MATα strain carrying the magic marker (can1Δ::STE2pr-Sp_HIS5) and a query gene TEL1 replaced by selectable marker (tel1Δ::NAT) is crossed to an ordered array of MATa deletion mutants (ykoΔ::kanMX). The resulting heterozygous diploids are induced to sporulate on medium with reduced carbon and nitrogen. The generated haploid meiotic spores are selected on medium lacking histidine and containing canavinine (a toxic analogue of arginine) for MATa cells which express S.pombe His5 protein and are deleted of CAN1 gene (encoding arginine permease). The selected MATa progeny are transferred to a medium that contains G418 / G418+NAT, which selects for single/double mutants.     30 3.2 Materials and methods 3.2.1 Yeast strains and media All strains are BY4743 background (Brachmann et al., 1998). The heterozygous double mutants were constructed by mating each of the corresponding single mutants. The haploid double mutants were generated from the sporulation of diploid heterozygous double mutants followed by tetrad dissection. The genotypes of all the strains used in the experiment were checked by PCR confirmation of the genomic DNA (Table 3.1).  MATa deletion mutant array (DMA) for SGA was obtained from Research Genetics Company. The MATa deletion collection was originally derived from genome-wide “bar- coded” yeast knockout (YKO) mutants constructed by the Saccharomyce Genome Deletion Project (http://www.sequence.stanford.edu/group/yeast_deletion_project/deletions3.html The overall genotype of the MATa deletion mutants from the DMA collection is: ). MATa ura3Δ leu2Δ his3Δ lys2Δ met15Δ ykoΔ::kanMX.  Heterozygous gene deletion collection provided by Jef Boeke was used to obtain single gene knockout alleles through tetrad dissection. It was originally constructed for the dSLAM technology (heterozygous diploid-based Synthetic Lethality Analysis on Microarrays) (Pan X et al., 2004). The overall genotype of strains from this Boeke collection is: MATa/α ura3Δ leu2Δ his3Δ lys2Δ/LYS2 met15Δ/MET15 can1Δ::LEU2-MFA1pr- HIS3/CAN1 xxx::kanMX/XXX  The media for SGA were the same as described in (Tong et al., 2004).   31 Strain Genotype DHL1 MATα his3 leu2 LYS2 TRP1 ura3 lyp1Δ can1Δ::STE2pr-Sp _HIS5 DHL10 MATa his3 leu2 LYS2 TRP1 ura3 yku70::KANMX DHL11 MATα his3 leu2 lys2 TRP1 ura3 yku70::KANMX DHL12 MATα his3 leu2 lys2 TRP1 ura3 yku70::KANMX tel1::NAT DHL13 MATa his3 leu2 lys2 TRP1 ura3 yku80::KANMX tel1::NAT DHL14 MATa his3 leu2 lys2 TRP1 ura3 yku80::KANMX DHL15 MATα his3 leu2 lys2 TRP1 ura3 yku80::KANMX DHL16 MATα his3 leu2 lys2 TRP1 ura3 yku80::KANMX tel1::NAT DHL17 MATa his3 leu2 lys2 TRP1 ura3 dnl4::KANMX tel1::NAT DHL18 MATa his3 leu2 lys2 TRP1 ura3 dnl4::KANMX DHL19 MATα his3 leu2 lys2 TRP1 ura3 dnl4::KANMX DHL20 MATa his3 leu2 LYS2 TRP1 ura3 psy3::KANMX DHL21 MATα his3 leu2 lys2 TRP1 ura3 psy3::KANMX DHL22 MATα his3 leu2 lys2 TRP1 ura3 dnl4::KANMX tel1::NAT DHL23 MATa his3 leu2 lys2 TRP1 ura3 shu1::KANMX tel1::NAT DHL24 MATa his3 leu2 lys2 TRP1 ura3 csm2::KANMX DHL25 MATα his3 leu2 lys2 TRP1 ura3 csm2::KANMX DHL26 MATα his3 leu2 lys2 TRP1 ura3 shu1::KANMX tel1::NAT DHL27 MATa his3 leu2 lys2 TRP1 ura3 shu2::KANMX tel1::NAT DHL28 MATa his3 leu2 lys2 TRP1 ura3 erg5::KANMX DHL29 MATα his3 leu2 lys2 TRP1 ura3 erg5::KANMX DHL30 MATa his3 leu2 LYS2 TRP1 ura3 pif1::KANMX DHL31 MATα his3 leu2 lys2 TRP1 ura3 pif1::KANMX DHL32 MATα his3 leu2 lys2 TRP1 ura3 shu2::KANMX tel1::NAT DHL33 MATa his3 leu2 LYS2 TRP1 ura3 psy3::KANMX tel1::NAT DHL34 MATα his3 leu2 lys2 TRP1 ura3 erg5::KANMX tel1::NAT DHL35 MATα his3 leu2 lys2 TRP1 ura3 rrm3::KANMX DHL36 MATα his3 leu2 lys2 TRP1 ura3 psy3::KANMX tel1::NAT DHL37 MATa his3 leu2 lys2 TRP1 ura3 csm2::KANMX tel1::NAT DHL38 MATa his3 leu2 LYS2 TRP1 ura3 yku70::KANMX tel1::NAT DHL4 MATa his3 leu2 lys2 TRP1 ura3 tel1::NAT DHL40 MATα his3 leu2 lys2 TRP1 ura3 lte1::KANMX DHL41 MATa his3 leu2 lys2 TRP1 ura3 lte1::KANMX DHL42 MATα his3 leu2 lys2 TRP1 ura3 csm2::KANMX tel1::NAT DHL43 MATa his3 leu2 lys2 TRP1 ura3 erg5::KANMX tel1::NAT DHL44 MATα his3 leu2 lys2 TRP1 ura3 bfa1::KANMX      32 Strain Genotype DHL45 MATa his3 leu2 lys2 TRP1 ura3 ckb2::KANMX DHL46 MATα his3 leu2 lys2 TRP1 ura3 cka2::KANMX DHL47 MATa his3 leu2 LYS2 TRP1 ura3 pif1::KANMX tel1::NAT DHL48 MATα his3 leu2 lys2 TRP1 ura3 ckb2::KANMX DHL49 MATa his3 leu2 lys2 TRP1 ura3 cka2::KANMX DHL5 MATα his3 leu2 lys2 TRP1 ura3 tel1::NAT DHL50 MATα his3 leu2 lys2 TRP1 ura3 rrm3::KANMX tel1::NAT DHL51 MATa his3 leu2 lys2 TRP1 ura3 bfa1::KANMX DHL54 MATα his3 leu2 lys2 TRP1 ura3 lte1::KANMX tel1::NAT DHL55 MATa his3 leu2 lys2 TRP1 ura3 bfa1::KANMX tel1::NAT DHL56 MATa his3 leu2 lys2 TRP1 ura3 srs2::KANMX DHL58 MATα his3 leu2 lys2 TRP1 ura3 srs2::KANMX DHL59 MATa his3 leu2 LYS2 TRP1 ura3 rrm3::KANMX DHL6 MATa his3 leu2 lys2 TRP1 ura3 shu1::KANMX DHL60 MATα his3 leu2 lys2 TRP1 ura3 ckb2::KANMX tel1::NAT DHL61 MATa his3 leu2 lys2 TRP1 ura3 lte1::KANMX tel1::NAT DHL62 MATα his3 leu2 lys2 TRP1 ura3 srs2::KANMX tel1::NAT DHL63 MATa his3 leu2 LYS2 TRP1 ura3 srs2::KANMX tel1::NAT DHL64 MATa his3 leu2 LYS2 TRP1 ura3 psy3::KANMX srs2::KANMX DHL65 MATα his3 leu2 lys2 TRP1 ura3 psy3::KANMX srs2::KANMX DHL66 MATα his3 leu2 lys2 TRP1 ura3 shu1::KANMX srs2::KANMX DHL67 MATa his3 leu2 lys2 TRP1 ura3 shu1::KANMX srs2::KANMX DHL68 MATα his3 leu2 lys2 TRP1 ura3 bfa1::KANMX tel1::NAT DHL69 MATa his3 leu2 lys2 TRP1 ura3 ckb2::KANMX tel1::NAT DHL7 MATα his3 leu2 lys2 TRP1 ura3 shu1::KANMX DHL71 MATa his3 leu2 lys2 TRP1 ura3 cka2::KANMX tel1::NAT DHL72 MATa his3 leu2 LYS2 TRP1 ura3 psy3::KANMX srs2::KANMX  tel1::NAT DHL73 MATα his3 leu2 lys2 TRP1 ura3 psy3::KANMX srs2::KANMX  tel1::NAT DHL76 MATα his3 leu2 lys2 TRP1 ura3 shu1::KANMX srs2::KANMX  tel1::NAT DHL77 MATa his3 leu2 lys2 TRP1 ura3 shu1::KANMX srs2::KANMX  tel1::NAT DHL8 MATa his3 leu2 lys2 TRP1 ura3 shu2::KANMX DHL9 MATα his3 leu2 lys2 TRP1 ura3 shu2::KANMX YJM164 MATα his3 leu2 lys2 TRP1 ura3 cka2::KANMX YNM9 MATa his3 leu2 lys2 ura3 tel1::NAT Table 3.1 Yeast strains used in the study in Chapter 3    33 3.2.2 Genome-wide screens 3.2.2.1 tel1 deletion query strain construction The SGA selection scheme was used to introduce the tel1 deletion into MATa deletion mutants. The MATα tel1 deletion query strain, DHL1, was created by a non-essential gene switching method (Tong et al., 2004). I crossed YNM9 bearing tel1Δ::NAT with YJM164 carrying the magic marker (LYS2 lyp1Δ can1Δ::STE2pr-Sp _HIS5) (Tong and Boone, 2006). The resulting heterozygous diploid was sporulated and followed by tetrad dissection to get the MATα tel1Δ strain containing the magic maker. The genotype of DHL1 (MATα tel1Δ::NAT his3Δ LYS2 lyp1Δ can1Δ::STE2pr-Sp_HIS5) was confirmed by PCR.  The introduction of the magic maker was to allow selection of haploids during SGA. HIS3, a S.cerevisiae gene required for histidine biosynthesis, was deleted from both the MATα query strain DHL1 and MATa deletion mutant array strains thus they cannot grow without the addition of histidine. S.pombe HIS5 gene (Sp_HIS5) can complement a S. cerevisiae HIS3 mutation, so his3Δ Sp_HIS5 cells can grow on minimal medium without histidine.  STE2 encodes the α-factor pheromone receptor which is only expressed in MATa cells. The STE2 promoter linked with the S.pombe HIS5 gene (STE2pr-Sp_HIS5) allows specific selection of MATa his3Δ haploids in media lacking histidine, because only MATa cells express the S.pombe HIS5 gene (Figure 3.1). The CAN1 gene and LYP1 gene encode an arginine permease and a lysine permease, respectively, which allows respective antibiotics, canavanine (a toxic analogue of arginine) and thialysine (a toxic analogue of lysine), to enter and kill cells. So only can1Δ lyp1Δ double mutants can survive in the presence of these toxic analogues. Sometimes mitotic crossover events between homologous chromosomes in   34 MATa/α diploids can result in MATa/a or MATα/α diploids. The STE2pr-Sp_HIS5 in MATa/a diploid can be also expressed like MATa haploid and live in the media without histidine. Introduction of can1Δ and lyp1Δ markers into the query strain enable MATa/a diploid cells to be killed by canavanine and thialysine because they carry a wild-type copy of the CAN1 and LYP1 genes (Tong and Boone, 2006). The magic maker (LYS2 lyp1Δ can1Δ::STE2pr-Sp_HIS5) in query strain DHL1 ensured the specific selection of MATa haploid progeny and substantially reduced the potential for false positives.  3.2.2.2 SC screen by synthetic genetic array (SGA) SGA analyses were performed using a Singer RoToR as described in (Tong et al., 2004) with the following modifications (Figure 3.2). The MATa yeast deletion mutant set (ykoΔ::kanMX) was arrayed at a density of 1536 colonies/plate by robotic pining. The query strain DHL1 was mated with yeast deletion mutants on YPD medium for 1 day. This mating step was done in triplicate to increase confidence. All plates were incubated at 30 degrees. Then diploids were selected twice on YPD plates contain 200µg/ml G418 and 200µg/ml clonNAT. Diploids were then pinned onto sporulation medium for 7 days at 25 degrees to induce sporulation. The resulting haploid spores were a combination of wild type, single and double mutants due to the independent assortment of chromosomes and recombination within the chromosomes. MATa spore progenies were selected twice on haploid selection medium (synthetic complete medium lacking histidine, lysine and arginine but containing 50µg/ml thialysine and 50µg/ml canavanine). Then MATa deletion mutants were selected twice on single deletion selection plate containing 200µg/ml G418 and double deletion selection plate containing 200µg/ml G418 and 200µg/ml clonNAT. The single deletion selection medium   35 allowed the germination of the MATa meiotic progeny carrying a ykoΔ::kanMX single gene deletion and carrying a ykoΔ::kanMX tel1Δ::NAT double gene deletion. The double deletion selection medium only allowed the ykoΔ::kanMX tel1Δ::NAT double gene deletion MATa progeny to grow. The resulting single/double gene deletion strains were also pinned onto single/double selection plate with a sub-lethal dose of camptothecin to test sensitivity. The final selection plates with/without camptothecin were replicated to three identical selection plates to increase the statistical power. All SGA plates were incubated at 30 degrees for 24 hours before scanning.  Figure 3.2 SC screen using synthetic genetic array (SGA) During SGA, tel1 deletion was introduced into non-essential gene mutants by mating MATα query strain (tel1Δ::NAT) to the MATa yeast deletion mutant set (ykoΔ::kanMX). Mating step was done in three biological replicates. The resultant double heterozygous diploid mutants were selected and induced to sporulate. The MATa haploid spores were selected by magic marker and then selected by single deletion (ykoΔ::kanMX) and double deletion (ykoΔ::kanMX tel1Δ::NAT) in parallel. Finally, the single/double mutants were also copied onto single/double selection plate with a sub-lethal dose of camptothecin to test sensitivity. The final selection plates with/without camptothecin were replicated to three identical selection plates before scanning.    36 3.2.3 Bioinformatic analysis 3.2.3.1 Fitness score and normalization The fitness of each strain was determined by Balony, a yeast colony plate scanning image analysis program developed by Barry Young in the Loewen Lab (http://code.google.com/p/balony/).   The area of each spot was first measured by pixel counts on the SGA plate scanning pictures. The program averaged the area values of all spots on the entire plate and normalized to 1. The fitness score resulted from comparing the area of every spot with average value. Edge effects, in which spots on the plate edge grow better than at plate centre due to more nutrition distribution, was also buffered and eliminated during normalization.  So the fitness score of most strains was around 1. A score below 1 means this strain grew worse than others and 0 means an empty spot or lack of growth. Because three biological replicates at the mating step and three technique replicates at the selection step were done during SGA, every strain had nine fitness scores. Also, four sets of fitness scores of each non-essential gene from both single/double selection media with or without CPT were collected. 3.2.3.2 SGA fitness data comparison by R R program was used to compare the fitness score between four sets of data for each gene. The difference of the average fitness score on single selection media and double selection media gives the EC score (experiment versus control) of certain gene. The more negative EC score that a certain gene was, the worse the double mutant grew compared with single mutants. Also because there were nine fitness scores for each data set, R program also gave the P value of this EC score.   37 3.2.4 Hit validation 3.2.4.1 Random spore assay (RSA) Double heterozygous deletion diploids were created by the same method in SGA by mating query strain DHL1 to single gene deletion strains derived from the DMA collection which was the same collection used in the SGA screen. After sporulation, a small amount of germinated spores was resuspended in 1ml sterile H2 1) 20 µL on SC-His/Arg/Lys + canavanine/thialysine select for the MATa spore progeny; O and plated out onto different medium: 2) 40 µL on SC-His/Arg/Lys + canavanine/ thialysine/clonNAT select for MATa tel1 mutants; 3) 40 µL on SC-His/Arg/Lys + canavanine/ thialysine/G418 select for MATa candidate gene mutants; 4) 80 µL on SC-His/Arg/Lys + canavanine/ thialysine/clonNAT/G418 select for MATa tel1 candidate gene double mutants.  In parallel, the same amount of spores was plated out onto corresponding selection media with 4µg/ml CPT to test the sensitivity. All plates were incubated at 30 degrees for 2 days. Colony growth under the 8 conditions was compared by eye and the double mutants were scored as SL, SC or no interaction.  The SC interaction between CSM2 and TEL1 had been validated by tetrad dissection followed by spot assay and growth curve assay. No SC interaction between TDP1 and TEL1 had been tested in the small matrix. CSM2 and TDP1 were used as positive control and negative control, respectively, to test RSA sensitivity.   38 3.2.4.2 Validation by tetrad dissection followed by spot assay and growth curve analysis Methods of spot assay and growth curve assay were the same as described in Chapter 2.  To increase confidence as well as strains with a clean genetic background, all single deletion mutants for construction of double deletion strains were derived by tetrad dissection from the Boeke collection (Pan et al., 2004), which was different from the DMA collection used in SGA screen. It was a heterozygous gene deletion collection with less contamination and more stable deletion alleles. Gene deletions in strains for spot assay and growth curve analysis were confirmed by PCR.  3.3 Results and discussion 3.3.1 Camptothecin (CPT) concentration for SGA From the previous small matrix screen, single deletion mutants themselves show different sensitivities to camptothecin (CPT). It was inevitable that some hypersensitive single deletion mutants were killed in the presence of a sub-lethal dose of CPT during the screen. In contrast, if the concentration was too low it would not elicit an effect on double deletion mutants. Unfortunately, there was little information from the literature about using SGA in the presence of CPT. It was therefore important to define an appropriate CPT concentration for the genome-wide screen that slightly inhibited tel1 mutants and allowed most of the other single mutants to grow without killing them.    39 In previous on plate spot experiments, the tel1 strain exhibited a mild sensitivity on YPD plates containing 16µg/ml CPT. But the SGA plates used synthetic complete media (SC- His/Arg/Lys + canavanine/thialysine) to select the MATa spore progeny and I found that the same strains grown on synthetic complete medium show higher sensitivities to CPT than on YPD rich medium with the same concentration of CPT (data not show). To mimic the final DNA damaging agent sensitivity assay during SGA, I ran CPT sensitivity tests on the third plate of 1536 density single deletion strains array from the MATa single selection step of the other SGA experiment (“G418 plate” step in Figure 3.1) because the genotype of strains on this plate was roughly the same as the SGA final single deletion array plate (“Expansion 4” plate in Figure 3.2) and it contains tel1 deletion strain. The strains were transferred by robot onto single haploid selection medium (SC-His/Arg/Lys + canavanine/thialysine + G418) with serial diluted concentrations of CPT (diluted from 16µg/ml to 1µg/ml) and grown at 30 degrees C for 24 hours. The severity of the phenotype was assessed qualitatively by eye. Because rad52 and rad54 are hypersensitive and tel1 is mildly sensitive to CPT, rad52Δ and rad54Δ spots exhibited no growth on plates with 4µg/ml CPT and there was ~20% reduction in growth in comparison to the plate with no CPT present at tel1Δ spot. On the plate with 8µg/ml CPT, most strains grew slower than on 4µg/ml CPT plate and tel1Δ was very sick. Comparing the growth on all plates, 4µg/ml was chosen as the concentration of CPT for SGA. After SGA was completed and fitness data of each strain obtained, to determine if an appropriate concentration of CPT was chosen for SGA plates, I compared my data of single mutant sensitivity to CPT with the published data of the chemical-genetic interaction profiles generated by screening CPT against the S. cerevisiae viable deletion set from two groups   40 (Deng et al., 2005, Parsons et al., 2003). The fitness of single mutants in the presence of CPT was compared to the fitness in the absence of CPT. The larger the difference of fitness value (EC value) there was between the CPT plate and the no CPT plate, the more sensitive the single mutant was. The comparison provided a genome-wide view of the CPT effect on each non-essential gene single deletion strain during my SGA screen. The Deng et al study used a barcode pool based assay by growing homozygous diploid deletion strains in liquid YPD medium containing 50 μM (~17μg/ml) CPT. The Parsons et al study used colony size on plate based assay by growing MATa hypoid deletion strains onto YPD solid YPD medium containing 15 μg/ml CPT. I compared my top 30 most sensitive genes (P value<0.05) with the top 30 genes from the Deng paper and the 27 genes that show strong sensitivity from Parsons paper (Table 3.2). I found 15 and 17 overlapping genes respectively. They included 13 of top 15 CPT sensitivity genes from Deng paper. These suggested that 4µg/ml CPT works well on SGA plates and was an appropriate concentration for the screen.    41  Top 30 sensitive mutant strains in SGA screen (Deng et al., 2005) (Parsons et al., 2003) Rank ORF GENE (a) EC value P value GENE (b) GENE (c) 1 YOR368W RAD17 -0.9285 6.09E-07 MMS4 CTF4 2 YGL175C SAE2 -0.89549 1.72E-05 (mms4) DDC1 3 YDR386W MUS81 -0.87715 8.61E-07 RAD59 MMS1 4 YBR094W PBY1 -0.83529 4.74E-05 SRS2 MMS22 5 YBR098W MMS4 -0.82795 2.69E-07 MUS81 MMS4 6 YGL163C RAD54 -0.82453 1.87E-05 SAE2 MRE11 7 YDR388W RVS167 -0.80862 0.000753 RAD54 MUS81 8 YLR288C MEC3 -0.80441 1.02E-05 RAD55 PAT1 9 YJL047C RTT101 -0.80415 9.69E-06 RAD51 RAD24 10 YDR076W RAD55 -0.8006 3.51E-06 RTT101 RAD50 11 YDL059C RAD59 -0.79696 5.72E-05 MMS1 RAD51 12 YJL092W SRS2 -0.77672 8.71E-05 RAD57 RAD52 13 YBR099C YBR099C -0.76826 1.09E-06 RAD50 RAD54 14 YJL124C LSM1 -0.75732 1.18E-05 VAC14 RAD55 15 YDR369C XRS2 -0.74981 1.27E-05 TOF1 RAD57 16 YCL016C DCC1 -0.74375 1.17E-05 CLB5 RAD59 17 YJL115W ASF1 -0.69261 3.50E-05 HMO1 RTT101 18 YML032C RAD52 -0.68893 1.32E-05 (mms4) RTT109 19 YOR233W KIN4 -0.67891 0.001929 PPH3 RVS161 20 YNL250W RAD50 -0.67503 3.37E-07 HTZ1 SPT10 21 YHR191C CTF8 -0.67347 6.84E-05 PAT1 SPT20 22 YHR154W RTT107 -0.66587 4.17E-05 PSY2 XRS2 23 YDR004W RAD57 -0.64499 0.000154 FUN30 PBY1 24 YMR258C ROY1 -0.63599 0.028126 LSM1 YBR099C 25 YER095W RAD51 -0.63033 0.000107 CSM3 YBR100W 26 YBR260C RGD1 -0.61873 0.0137 NUP60 YLR235C 27 YMR190C SGS1 -0.61561 0.004059 SPT21 ZUO1 28 YOL106W YOL106W -0.60491 0.005285 ILM1  29 YMR048W CSM3 -0.60411 0.000455 ULA1  30 YLL002W RTT109 -0.58245 5.80E-06 UBC4  Table 3.2 Comparison of most sensitive genes to CPT GENE (a) list shows the top 30 sensitive single mutant strains in my SGA screen ordered by EC value (P value < 0.05). More sensitive mutant has more negative EC value. GENE (b) list shows top 30 gene list from Deng paper ordered by rank of sensitivity. (mms4) indicates that deletion of the ORF deletes MMS4 rather than a new gene. GENE (c) list shows 27 genes with strong sensitivity from Parsons paper ordered by alphabetical. Red box indicates mutations common to my top 30 CPT sensitivity gene list and the Parsons et al 2003 and Deng et al 2005 studies.    42 3.3.2 SC/SL candidate genes list obtained through two-step analysis To ensure I did not miss any potential SL and SC interactions, I used a two-step strategy to analyze the SGA data. (Figure 3.3)  First, I compared the data from single selection plates without CPT against double selection plates with CPT to identify all double mutants that were sensitive to CPT which had largest fitness difference compared with single mutant without CPT. 173 candidates genes were deemed significant (EC<-0.4 P value<0.05). These hits included all potential SC/SL interaction as well as the ones in which both single and double mutants were sensitive to CPT.  Second, I sorted out all the single mutants sensitive to CPT by comparing the fitness of single and double mutants in the presence of CPT (EC<-0.4 P value<0.05). There were 27 candidates on the final gene list (Table 3.3).  After checking each gene profile and phenotype on SGA plates, I identified 2 SL candidate genes, 20 SC candidate genes and 5 linkage genes (including TEL1) whose ORF is “linked” to the query TEL1 gene. Because double mutants were created by meiotic recombination, the linkage genes tended to form double mutants at a reduced frequency, appearing as false positive synthetic lethal/sick with the query gene mutation.    43     Figure 3.3 Two steps to filter for SC and SL interaction hits All hits (value<0.05) shown on scatter chart are ordered by EC value. The comparison of the data from single selection plates without CPT against double selection plate with CPT filtered 173 candidates (EC<-0.4). The comparison of the data from single selection plates with CPT against double selection plate with CPT identified 27 candidates (EC<-0.4).      44 GENE Linkage EC value P value Function HBT1  -1.18115 1.31E-06 Cell morphogenesis SCS22 linked -1.1702 8.33E-06 Phospholipids metabolism RAS1  -1.1282 3.87E-06 Adenylate cyclase activating pathway MRP21 linked -1.00322 8.94E-05 Mitochondrial ribosomal protein YBL071C linked -0.87391 5.27E-06 Putative protein of unknown function CKI1  -0.84728 0.000266 Phosphatidylcholine synthesis YJR128W  -0.79801 0.03098 Dubious open reading frame SRO77 linked -0.7667 0.013652 Exocytosis and cation homeostasis RRM3  -0.55443 0.000345 DNA helicase YPR063C  -0.54455 0.013964 Protein of unknown function MVB12  -0.53027 0.006265 ESCRT-I subunit YPR1  -0.52772 0.047448 NADPH-dependent aldoketo reductase TEL1 linked -0.52577 0.000147 DNA damage cell cycle checkpoint RRT6  -0.50463 0.03739 Putative protein of unknown function YCL046W  -0.47278 0.008209 Dubious open reading frame RUB1  -0.47079 0.005797 Ubiquitin-like protein FPK1  -0.45977 0.005096 Ser/Thr protein kinase CSM2  -0.43987 0.003578 Subunit of SHU complex CKB2  -0.4374 0.002063 B Subunit of casein kinase 2 (CK2) TRE1  -0.43697 0.002569 Regulation of metal transporter YKU80  -0.43685 0.012981 Subunit of the telomeric Ku complex LTE1  -0.43321 0.028525 spindle orientation checkpoint YOR223W  -0.42895 0.044181 Protein of unknown function AGX1  -0.42142 0.000498 Glycine synthesis ERG5  -0.42098 0.013983 Ergosterol synthesis YBR134W  -0.41134 0.032558 Dubious open reading frame PSY3  -0.41089 0.000613 Subunit of SHU complex Table 3.3 Twenty-seven candidate genes (include linkage genes) list after SGA data analysis    45 3.3.3 No SL and 13 SC interactions with TEL1 were uncovered The random spore assay test demonstrated that the two SL candidate genes interactions were false positives. Double mutants grew normally and didn’t show sensitivity to CPT. Only two genes, MEC1 and DNA2, are reported synthetic lethal with TEL1 in yeast and both of them are essential genes and were not tested in this experiment (Budd et al., 2005, Chakhparonian et al., 2005).  Among 20 SC candidates, 4 genes are related to DNA repair and 2 of them are in the same complex. So I decided to validate these 4 hits directly by tetrad dissection, while the remaining 16 were confirmed by random spore assay. After removing false positive SC candidates through random spore assay and tetrad dissection, I identified 7 candidates that appear to have SC interactions with TEL1. Among them, CSM2 and PSY3 are in the same SHU complex which contains the other two subunits SHU1 and SHU2; YKU70 is a component of the yKu70-yKu80 heterodimer; CKB2 is the subunit of casein kinase 2 (CK2) which is also comprised of CKA1, CKA2 and CKB1; LTE1 is involved in spindle checkpoint and interacts with BFA1-BUB2 complex. All function related non-hit genes mentioned have negative EC value >-0.4 or P value > 0.05 (Table 3.4). So I also included them to test the SC interactions with TEL1 by random spore assay and tetrad dissection. 6 of 9 were also found to be SC to CPT with tel1. In total 13 SC interactions were uncovered.    46 Function Gene EC value P value Random spore assay Tetrad dissection Interaction Shu complex CSM2 -0.43987 0.003578 √ √ SC PSY3 -0.41089 0.000613  √ SC SHU1 -0.29365 0.135906  √ SC SHU2 -0.24395 0.054303  √ SC Ku complex YKU80 -0.43685 0.012981  √ SC YKU70 -0.22265 0.01981  √ SC DNA helicase RRM3 -0.55443 0.000345  √ SC PIF1 0.321353 0.65881 X X  Bfa1- Bub2 complex LTE1 -0.43321 0.028525 √ √ SC BFA1 -0.37092 0.059806 √ √ SC BUB2 -0.30669 0.018664 √ √ SC Casein kinase 2 CKB2 -0.4374 0.002063 √ √ SC CKA2 -0.31277 0.117162 X X  CKB1 -0.27383 0.051928 √ √ SC CKA1 -0.06364 0.77864 X  Other function YPR1 -0.52772 0.047448 X   CKI1 -0.84728 0.000266 X   MVB12 -0.53027 0.006265 X   RUB1 -0.47079 0.005797 X   FPK1 -0.45977 0.005096 X   ERG5 -0.42098 0.013983 √ √ SC TRE1 -0.43697 0.002569 X   AGX1 -0.42142 0.000498 X   RAS1 -1.1282 3.87E-06 X   HBT1 -1.18115 1.31E-06 X  Unknown function YJR128W -0.79801 0.03098 X   YBR134W -0.41134 0.032558 X   YCL046W -0.47278 0.008209 X   YPR063C -0.54455 0.013964 X   YOR223W -0.42895 0.044181 X   RRT6 -0.50463 0.03739 X   Table 3.4 SL/SC candidate validation The red box indicates P value >0.05 or EC value >-0.4. “√” and “X” indicates confirmed true and false positive respectively.     47 3.3.3.1 SHU genes SHU genes (CSM2, PSY3, SHU1, and SHU2) were first identified in the same epistasis group that suppresses severe sensitivity to HU and MMS in sgs1 and top3 mutants. All four SHU gene products interact in two-hybrid assays suggesting they are stably associated as a multi-protein complex in the cell. The single SHU gene deletion mutants all demonstrate similar phenotypes: a mutator phenotype, increased gross chromosomal rearrangement (GCR) and moderate sensitivity to methyl methane sulfonate (MMS).  Mutation of all four genes does not cause any additive effects.  It has been proposed that these proteins exist in a multimeric complex that fails to function when any one member is missing (Shor et al., 2005). In spot assays, compared with YPD control plates, both TEL1 and SHU gene single mutants exhibited weak sensitivity to CPT, while all 4 double mutants were almost completely inviable on CPT plate (Figure 3.4). The SC interactions between SHU genes and TEL1 were clearly demonstrated.  Figure 3.4 SC interactions between SHU genes and TEL1 by spot assay In spot assays, single and double mutants were grown on the same plate, wild type and rad52Δ mutants were used as negative and positive controls, respectively. All strains grew normally on YPD plates except the rad52Δ mutant grows a little slower. On CPT plates, SHU gene single mutants show a minor growth defect and tel1 null cells grew slightly less, while the double mutants were almost dead.   48 To quantify the PSY3, SHU1, SHU2 SC interactions with TEL1, I used growth curve assays in presence of CPT and compared the fitness ratio of wild type and double mutants to the expected ratio based on each single mutant. The calculation methods of the fitness ratio and expected ratio are described in Chapter 2. An example of a PSY3 growth curve is shown in (Figure 3.5). The growth curve of psy3Δ tel1Δ double deletion strain in media with 2µg/ml CPT is much slower than both single mutants. The quantitative analysis results indicate psy3Δ tel1Δ, shu1Δ tel1Δ and shu2Δ tel1Δ double mutants are 15%, 10% and 10% respectively lower than expected when exposed to 2µg/ml CPT.    Figure 3.5 SC interaction between PSY3 and TEL1 by growth curve assay Growth curve assays were done in YPD media containing different concentrations of CPT. All single and double mutants grew similarly to wild type in YPD media without CPT. When treated with 2µg/ml CPT, the growth of psy3Δ tel1Δ double deletion is much slower than either single mutants. The quantitative analysis shows the fitness of double mutants is around 15% lower than expected when exposed to 2µg/ml CPT.    49 Next I asked if these SC interactions with TEL1 are specific to CPT-induced DNA damage. I tested sensitivity to high concentrations of hydroxyurea (HU, blocks DNA replication), and cycloheximide (CHX, inhibits protein synthesis). The results show that the SHU gene SC interactions with TEL1 are specific to CPT and do not extend to HU or CHX (Figure 3.6).  Figure 3.6 SC interactions between SHU genes and TEL1 are specific to CPT Spot assays were done on YPD plates with different chemicals: 20 µg/ml CPT, 100 mM HU (DNA replication inhibitor) and 0.1 µg/ml cycloheximide (CHX; protein synthesis inhibitor). The double mutants are completely inviable on CPT plate, but exhibit no/little growth defect on HU and CHX plates compared with single mutants.  As an inhibitor of protein synthesis, CHX induces defects in cell growth which is more general compared to the two DNA damaging agents: CPT and HU (del Pozo et al., 1991). The topoisomerase I inhibitor CPT induces formation of DNA Topoisomerase I (Top1) cleavage complexes (Top1cs) that block the leading strand during DNA synthesis, which often leads to replication fork collapse. Also, the accumulation of single-stranded nicks after CPT treatment can result in double-strand breaks (DSBs) upon passage of the fork (Redon et al., 2003). In contrast, HU is an inhibitor of ribonucleotide reductase that slows down fork progression and induces stalling of the replication fork by reducing dNTP pools (Lopes et al.,   50 2001, Alvino et al., 2007). So both HU and CPT can increase the DNA replication stress during S phase. Compared to HU, CPT induces more DSBs and DNA replication fork collapse rather than arresting or slowing down the replication process. On spot assays, the resistance to HU of double mutants indicates cells are able to resume DNA replication after HU-dependent fork arrest. The sensitivity to CPT suggests double mutants are defective either in recovery after CPT-dependent fork breakage or in DSB repair pathways.  While the formation of DSBs triggers activation of the DNA damage checkpoint TEL1 and MEC1, DNA repair enzymes in the cell also are activated to repair DSBs by either non- homologous end joining (NHEJ) or homologous recombination (HR) (Mao et al., 2008). Recently, it has been proposed that the Shu complex stabilizes Rad51 filaments to promote HR by inhibiting the disassembly reaction of Srs2 (Bernstein et al., 2011). To investigate if SRS2 is involved in the SC interaction with TEL1, I knocked out the SRS2 gene in double deletion mutants and tested the sensitivity to CPT by spot assay (Figure 3.7). The deletion of SRS2 gene doesn’t alleviate the CPT sensitivity of psy3Δ tel1Δ or shu1Δ tel1Δ double mutant. Hypersensitivities to CPT of triple mutants suggest that the SC interactions are SRS2 independent. The SHU complex is also suggested to be involved in the error-free repair pathway in DNA post-replication repair (PRR) (Ball et al., 2009). It functions to bypass replication-blocking lesions and prevent damage-induced cell death, but little is known about the detailed mechanism. A recent study shows the Shu complex has DNA-binding activity in vivo (Tao et al., 2012). It is possible that SHU facilitates efficient repair of broken replication forks and the activation of this repair is independent of TEL1 checkpoint, further investigation is needed to address this possibility.   51  Figure 3.7 SC interactions between SHU genes and TEL1 are SRS2 independent SHU gene, TEL1 and SRS2 triple deletion strains are constructed and test by spot assay. The triple deletion was still much more sensitive to CPT than tel1Δ single mutant and srs2Δ tel1Δ, srs2Δ psy3Δ, srs2Δ shu1Δ double mutants.  3.3.3.2 YKU genes The highly conserved Ku genes are involved in double strand break repair by NHEJ and telomere maintenance. In S. cerevisiae, the core components of the NHEJ machinery include the Yku (Yku70–Yku80), MRX (Mre11–Rad50–Xrs2) and Dnl4–Lif1 complexes, which are recruited rapidly to DSBs, where the two Yku subunits form a ring-like structure that binds to DNA ends and initiates NHEJ (Daley et al., 2005). In humans, the essential NHEJ components also include the catalytic subunit of the DNA-dependent protein kinase (DNA- PKcs), Artemis, and XRCC4. It has been reported that siRNA targeting the Ku70 protein enhances the response to topoisomerase II inhibition in human cancer cells (Ayene et al., 2005). Upon both the spot assay and growth curve assay, cells with mutations in a YKU gene alone show no growth defect compared with wild type. yku70Δ tel1Δ double mutants are very sensitive to CPT, similar to the SHU complex. The fitness of yku70Δ tel1Δ double mutant is around 13% lower than the expected fitness if there was no interaction (Figure3.8). Both subunits of the yKu70-yKu80 heterodimer are confirmed as SC partner genes of TEL1, and   52 these SC interactions are also specific to CPT-induced DNA lesions and do not extend to HU or CHX (Figure3.9).     Figure 3.8 SC interactions between YKU genes and TEL1 The SC interactions between YKU70/YKU80 and TEL1 are demonstrated by spot assays and growth curve assays. On the plate spot assay, yku70Δ and yku80Δ exhibit no sensitivity to CPT as compared with wild type, while double mutants of TEL1 and either YKU gene are more sensitive than tel1Δ single mutant, which is mildly sensitive to CPT. The quantitative analysis from growth curve data shows the fitness of yku70Δ tel1Δ double mutant is around 13% lower than the expected fitness if there was no interaction.     53 In S. cerevisiae, the deletion of YKU70 or YKU80 causes defective NHEJ and can suppress the lethality of mec1 mutants, and this suppression is dependent on wild type TEL1 (Corda et al., 2005).  It  has been reported that chemosensitization of ATM-deficient tumor cells occurs when DNA-PK, which is also involved in NHEJ but has no yeast homologue, is knocked down by RNAi or small molecule inhibitors in vitro and in vivo (Jiang et al., 2009). These findings suggest that there is a connection between the DNA damage checkpoint ATM/TEL1 and NHEJ pathway. However, another study reported that CPT shows a stronger inhibition of DNA replication in Ku80-/- cells than Ku80+/+ Wang et al., 2002  cells. The inhibition was correlated with the activities of ATR and CHK1 but not DNA-PK which suggests that Ku may function in the DNA damage checkpoint in a DNA-PK independent fashion ( ). In budding yeast, yku null mutants exhibit increased gross chromosomal rearrangements (GCR), and overexpression of the YKU complex can suppress GCRs after MMS treatment. Furthermore, this suppression cannot be attenuated by the deletion of DNL4 gene (DNA ligase involved in NHEJ, Lig4 in humans), suggesting another function of the Yku complex in maintaining genome stability besides NHEJ (Banerjee et al., 2006). DNL4/Lig4 is exclusively required for the Ku-dependent NHEJ pathway of DSB repair and no other DNA ligases have been found that can substitute for this function both in human cells and budding yeast (Adachi et al., 2001). To test whether the SC interaction with TEL1 is with NHEJ in general or just the Yku complex, I knocked out the DNL4 gene in a tel1Δ mutant to check if it too was SC to CPT (Figure 3.9). The dnl4Δ tel1Δ double mutant shows a similar sensitivity with tel1Δ single mutants suggesting that the SC interactions between YKU70/YKU80 and TEL1 is due to an NHEJ independent mechanism.   54  Figure 3.9 Interactions between YKU genes and TEL1 are CPT specific and due to an NHEJ independent mechanism On spot assays, YKU70 doesn’t exhibit SC interaction with TEL1 on HU or CHX plate. dnl4Δ tel1Δ double mutant shows a mild sensitivity similar to tel1Δ single mutants suggesting no SC interaction between DNL4 and TEL1.    Figure 3.10 SC interactions between RRM3 and TEL1 The SC interactions between RRM3 and TEL1 are demonstrated by spot assays and growth curve assays. The quantitative analysis from growth curve data shows the fitness of rrm3Δ tel1Δ double mutant is around 17% lower than expected.    55 3.3.3.3 RRM3 gene RRM3 shows the strongest SC interaction with TEL1. Like YKU genes, in medium containing 2µg/ml CPT, the rrm3Δ single mutant is not sensitive and rrm3Δ tel1Δ double mutant shows slow growth, about 17% slower than expected (Figure 3.10). In addition, on spot assays compared with single mutants, the slower growth of the rrm3Δ tel1Δ double mutant on all CPT, HU and CHX plates suggests the SC interaction between RRM3 and TEL1 is not specific to CPT (Figure 3.11).  Rrm3 protein in yeast is a 5’ to 3’ DNA helicase involved in rDNA replication and relieves replication fork pauses at telomeric regions (Ivessa et al., 2000, Makovets et al., 2004). It has a structurally and functionally related protein Pif1. Both of them belong to the conserved Pif1 subfamily, of which humans have only one identified Pif1 family member, Pif1. Cells mutated in either helicase are repair DNA proficient (Bochman et al., 2010). Therefore, I also tested PIF1 gene for SC interaction with TEL1. PIF1 also functions in mitochondrial DNA recombination and genome maintenance. The observation that the pif1 deletion strain loses functional mitochondria rapidly could explain the slow growth phenotype of pif1Δ mutants on spot assays (Foury and Kolodynski, 1983)(Figure 3.11).  Interestingly, the phenotypes of rrm3Δ and pif1Δ mutants on CPT plates are opposite: RRM3 show a negative interaction with TEL1 while PIF1 acts as a phenotypic suppressor. RRM3 and PIF1 appear to have different roles in several areas other than mitochondria in yeast. Rrm3 is required for telomere-proximal replication forks, while Pif1 is thought to be an inhibitor of telomerase (Schulz and Zakian, 1994). RRM3 and PIF1 also have different functions during Okazaki fragment maturation and HR DNA repair. The rrm3Δ mutant is synthetically lethal with dna2   56 mutants, a helicase/nuclease required for Okazaki fragment processing (OFP) and HR repair. In contrast, pif1Δ suppresses both the DNA replication and repair defects of dna2 mutants and even the lethality of deletion of DNA2 (Ivessa et al., 2000).   Figure 3.11 SC interactions between RRM3 and TEL1 are not specific to CPT Both the pif1Δ single mutant and pif1Δ tel1Δ double mutant show growth defects on YPD plates and HU plates. pif1Δ tel1Δ double mutant is even sicker. On CPT plate, the growth of rrm3Δ tel1Δ double mutant was entirely inhibited by CPT, whereas pif1 appears to partially suppress the tel1 growth defect. This positive interaction between PIF1 and TEL1 only occurs on CPT plate but not on HU or CHX plate.  3.3.3.4 Casein kinase 2 beta subunit genes Casein kinase 2 is a serine/threonine protein kinase, which has been implicated in cell cycle control, DNA repair, regulation of the circadian rhythm and other cellular processes (Guerra et al., 1999). In budding yeast, it is a tetramer of two alpha subunits (CKA1 and CKA2) and two beta subunits (CKB1 and CKB2). The alpha subunits have the catalytic kinase domain. The beta subunits function as the regulator of the two catalytic subunits, enhancing their stability, activity and specificity (Pinna, 1990). Random spore assay (RSA) of all 4 subunits shows only CKB1 and CKB2, the beta subunits of CK2 kinase are SC with TEL1. One example of RSA plates with CKB1 is shown (Figure 3.12)    57  Figure 3.12 SC interactions between CKB1 and TEL1 by RSA On the control plate (right), both the amount and size of colonies grown on double selection plate are similar with single selection plates. On experimental plate with 4µg/ml CPT (left), only a few small colonies can be found on the double selection plate.  The spot assay shows that knockout of CKA2 in a tel1Δ mutant doesn’t cause a more severe phenotype in the presence of CPT, however CKB2 displays a clear SC interaction with TEL1 (Figure 3.13).  These observations suggest only CK2 beta subunits exhibit SC interaction with TEL1. The same results from both random spore analysis and tetrad dissection followed   58 by spot assay indicate a CKA-independent role of CKB1/CKB2 in addition to regulating the CK2 tetramers leads to the SC interaction with TEL1. In S. pombe, only CK2β mutants show adaptation defectives in response to DNA damage and this defect is independent of the catalytic CK2 subunits (Toczyski et al., 1997). Study in mammalian cells demonstrates that CK2β interacts with Chk1 (G2 checkpoint) and this interaction does not extend to the CK2α subunit (Guerra et al., 2003).  It has been reported that high level expression of CK2β relative to CK2α was observed in a variety of tumors (Vilk et al., 2001). The distinct cellular functions of CK2β from CK2 including the SC interaction with TEL1, may offer insights into the multiple roles of the regulatory subunit of CK2 kinase.  Figure 3.13 Only CK2 beta subunits show SC interactions with TEL1 On spot assays, cka2Δ, ckb2Δ and tel1Δ single mutation strains show a mild sensitivity to CPT. ckb2Δ tel1Δ double mutant is dead on CPT plate, while cka2Δ tel1Δ grows similar to cka2Δ single mutant.  3.3.3.5 Spindle checkpoint genes and ERG5 gene The other SC partner genes confirmed by random spore analysis and further validated by tetrad dissection include three spindle checkpoint genes (BFA1, BUB2 and LTE1) and the ERG5 gene involved in ergosterol biosynthesis. On spot assay, lte1 null mutants are hypersensitive to CPT; both lte1Δ and lte1Δ tel1Δ   59 mutants are inviable on CPT plates. The double mutants also show strong sensitivity on HU plates but not on CHX plate, as compared with single mutants, suggesting that the SC interaction with the spindle checkpoint genes is specific to CPT and HU induced DNA damage (Figure 3.14).   Figure 3.14 SC interactions between spindle checkpoint genes and TEL1 On CPT plates, the bfa1Δ single mutant is not sensitive while the bfa1Δ tel1Δ mutant is very sick. Both lte1Δ and lte1Δ tel1Δ mutants are inviable in the spot assay to CPT. Both TEL1 SC interactions with LTE1 and BFA1 are clearly shown on HU plate but not CHX plate.   Figure 3.15 SC interactions between ERG5 genes and TEL1 Erg5 shows SC interaction with TEL1 on the CPT plate. Whereas both the single and double mutants are sensitive to HU and CHX.  Erg5 shows a clear SC interaction with TEL1 on CPT plates by spot assay. However, both erg5Δ and erg5Δ tel1Δ mutants display a similar sensitivity to HU. (Figure 3.15) One possible explanation is that membrane lipid composition affects CPT susceptibility in yeast. It has been observed that cells deleted for genes encoding ergosterol biosynthetic enzymes   60 such as ERG3, ERG4 and ERG6 were hypersensitive to various DNA damaging agents include CPT (Parsons et al., 2003). The defective membrane in erg5Δ cells may allow more CPT to enter cell and cause more DNA damage and eventually lead to cell death.  All four candidate genes have no detectable homologues in human.    61 Chapter 4 : Conclusion, implications and future directions  4.1 Summary Chromosome instability (CIN) is characterized by aberrant chromosomal states including changes in ploidy, alterations in the number of chromosomes (aneuploidy), or gross chromosomal rearrangements (GCR), all of which are consistently observed in most cancers in contrast to normal cells (Storchova and Pellman, 2004). As a hallmark of cancer cells, CIN can be exploited as a tumor-specific target for selective killing of cancer cells. The emerging data from research of using PARP inhibitors to target tumor cells with BRCA1/BRCA2 mutation with/without DNA damaging agents suggests that synthetic lethality (SL) and synthetic cytotoxicity (SC) can be exploited to preferentially kill tumor cells with specific mutations relative to adjacent normal cells (Bryant et al., 2005, Farmer et al., 2005, Chan and Giaccia, 2011). Using the same SL/SC strategies, cancer cells with CIN gene mutations may be inviable with/without the presence of a sub-lethal dose of DNA damaging agents when combined with another non-essential gene mutation. While CIN mutations may contribute to cancer development, they may also cause cells to have difficulty in replicating their DNA and proliferating. In such cases, it is possible that tumor cells can proliferate by overcoming the negative consequence of CIN through adaptive DNA repair mechanisms. In other words, tumor cells harboring CIN mutations may become heavily dependent on certain DNA repair pathways for viability. Thus use of SL/SC strategies by inhibition of these DNA repair enzymes may enhance CIN to an intolerable level or sensitize cells to DNA damage stress and eventually lead to cell death. To test this theory, I focused on the CIN gene ATM, which is often mutated in human lymphoid and epithelial tumors (Ahmed and Rahman, 2006). I hypothesized that knockdown of certain second site DNA repair genes would selectively kill   62 ATM deficient cells resulting in SL, or sensitize ATM deficient cells to a specific DNA damaging agent resulting in SC. The goal of this thesis is to use Saccharomyces cerevisiae as a model to identify SL interaction partner genes with ATM, and SC interaction partner genes together with a sub-lethal dose of DNA damaging agents. After directly testing a small matrix of the ATM and ATR homologues, TEL1 and MEC1 with three DNA repair enzyme genes and four DNA damaging agents, I focused on TEL1 and camptothecin (CPT) and screened a large matrix of the non-essential gene collection using high throughput genetic methods. In total, fourteen SC interactions with TEL1 (ATM in humans) were uncovered in budding yeast. Most of the SC genes found are involved in DNA repair and show a specificity of sensitivity to CPT which proved my hypothesis.  4.2 Improvement of methodology To determine which combinations of the DNA repair enzymes inhibition and sub-lethal doses of various types of DNA damaging agent might result in effective killing of tumor cells compared to healthy cells, I adapted the SGA methods in yeast, which enabled rapid screening of the thousands of possible combinations on one plate and determining SL/SC interactions through bioinformatic analysis. Compared with the direct test in the small matrix with of three DNA repair enzymes, the SGA approach was much faster and more scalable. However, the high throughput approach yielded a high proportion of false positives in the initial screen data that need to be resolved by retesting.  After validation by random spore analysis and tetrad dissection, only 7 out of 22 candidate genes from SGA screen turned out to be true positive hits. The biological processes involved   63 and the single mutant phenotype of the 15 false positive hit genes are various. Also there were 6 false negative hits confirmed by tetrad dissection and random spore assay. The similarity of principles between random spore assay and SGA implied technical problems during SGA, such as the contamination or mis-location of single mutants in the DMA collection or growth variation caused by differences in cell transfer by pinning. Also, many DNA repair gene mutants are very sensitive to DNA damaging agents because of DNA damage response defects. These genes were filtered out during the second comparison analysis because the single mutant was itself inviable or sick when exposed to the sub-lethal dose of CPT. In addition, there was a maximum number of times that replicate plates could be performed which limited the number of DNA damaging agents that could be tested during SGA. Because the pins controlled by the robot picked cells always at the same position, one mother plate of 1536 density array can be used to replicate to ~5-6 daughter plates, at most.  One way to fix these problems is to make customized array plates with lower density (for example, 384 colonies/plate). The contamination can be minimized by barcode sequence checking. Low density would also prevent the occurrence of contamination and mis-location during SGA. The larger pin head allows robot transfer of more cells when replicate-plating and reduces the noise inherent in colony growth. In addition, the larger colonies area would show greater differences of the fitness when compared to different plates and result in a more confident EC value. Finally, greater replicate-plating enables multiple chemical sensitivity tests needed to provide the information about the specificity of the SC interaction during SGA.    64 Other model organisms and high throughput methods also can be employed for the SC screen. For example, using barcode analysis by deep sequencing in yeast could allow the growth of mutants pooled in liquid media with different DNA damaging agents and quantitatively assessing the complex pools following outgrowth (Smith et al., 2009). An RNAi based screen in Caenorhabditis elegans would allow testing of the human genes which are not conserved in yeast but are conserved in worm, such as BRCA1 and BRCA2 (Kirienko et al., 2010).  4.3 Clinical implications Fourteen SC interactions with TEL1 provide the basis to investigate the mechanism behind the SC interaction and can be used to propose potential drug targets for ATM deficient cancer cell-specific therapy. Among them, at least 8 proteins have known homologues or functional orthologues in humans.  Human Fen1 and yeast RAD27 show a high degree of sequence similarity (73%) with a BLAST value of e-104 Liu et al., 2004 . Similar to RAD27 in yeast, Fen1 is also a flap endonuclease and exhibits nick exonuclease activities which are required for DNA synthesis and DNA repair including BER and HR. However, Fen1 is an essential gene required for DNA replication in human cells ( ). FEN1 homozygous knockouts lead to embryonic lethality in mice (Larsen et al., 2003). Together, these findings suggest its critical role in DNA replication and repair. It has been reported that FEN1 silencing can specifically kill RAD54B-deficient tumor cell (McManus et al., 2009). FEN1 inhibitors are currently in pre-   65 clinical development. It will be of great interest to treat ATM deficient cells with a combination of FEN1 inhibitors and DNA damaging agents such as CPT.  The Shu complex subunits, PSY3 and SHU2, appear to have homologues in human cells (RAD51D and SWI1 respectively) (Martin et al., 2006) and a human Shu-like complex was been found recently, which consists of SWS1 and SWSAP1 proteins (Liu et al., 2011). Their functions are reminiscent of SHU with respect to promoting HR and suppressing RecQ family mutant phenotypes. Like the SHU complex in budding yeast, the depletion of SWS1 and/or SWSAP1 leads to increased MMS sensitivity but not appreciable sensitivity to CPT. The function in HR of SHU genes in both organisms is not yet clarified and additional studies are needed.  As an essential component in NHEJ pathway, the Ku heterodimer is composed of Ku70 and Ku80 homologues in all eukaryotes (Daley et al., 2005). There are some variations between yeast and human. In humans, Ku interacts with DNA-PK, recruiting DNA-PKcs to DNA ends (Gottlieb and Jackson, 1993); while in S. cerevisiae, direct homologues of DNA-PKcs do not exist, and Ku lacks the DNA-PKcs-recruitment motif. The lack of DNA-PKcs suggests a more fundamental role of Ku genes in budding yeast. Strikingly, the study from Weller et al. shows Ku is also conserved in prokaryotes, which reinforces the idea of conservation of DNA repair throughout evolution (Weller et al., 2002). On the other hand, unlike NHEJ as a dominant repair pathway for DSB in higher organisms with larger genomes, prokaryotes might not need for NHEJ. This could be explained by their small genomes size making homology searches during HR quite easy. Alternatively, the prevalence   66 of Ku might reflect its action in other processes beside NHEJ. My study shows that the YKU gene SC interaction with tel1 is NHEJ-independent. It will be of great interest to determine whether this interaction is conserved in human.  While S. cerevisiae has two Pif1 family helicases: Rrm3 and Pif1, higher eukaryotes contain only one: Pif1. Human Pif1 shows 24% identity with ScRrm3 and ScPif1 over the helicase domain (Mateyak and Zakian, 2006). Only RRM3 shows SC interaction with TEL1 in my study, which could be explained by the very different roles PIF1 and RRM3 play in yeast in several cellular processes including DNA replication, DNA repair and telomere maintenance (Ivessa et al., 2000). The reason that such similar proteins have very divergent roles is unknown. Also the functions and mechanisms of human Pif1 are still not very clear due to the difficulties in expressing and purifying full-length protein. A recent study shows human Pif1 helicase has activity akin to Rrm3 in the nucleus and Pif1 in the mitochondria (Bochman et al., 2010). It is therefore possible that the SC interaction between RRM3 and TEL1 is conserved in human cells.  CK2 beta subunit genes were unexpected hits from my screen. Protein kinase CK2 is composed of a catalytic and regulatory subunit and is ubiquitously present in eukaryotic organisms (Pinna, 1990). It is involved in a myriad of cellular processes including cell growth and proliferation with more than 300 substrates (Meggio and Pinna, 2003).  In S. cerevisiae, there are two regulatory isoforms: CKB1 and CKB2; while there is only one regulatory subunit, CK2 β, in humans. It has been shown that CK2 protein levels and activity are increased in many cancers (Ahmad et al., 2005). Several CK2 inhibitors have been   67 developed, but most of them target the ATP-binding pocket on the catalytic subunit (Mazzorana et al., 2008). A two-hybrid screening of a combinatorial library of peptide aptamers to target CK2 beta subunit was carried out (Martel et al., 2006). This group found one potent CK2β-interacting peptide (KD=0.4 µM) and it can induce apoptosis in cells with wild-type p53. The CK2β subunit will be particular interesting to test in combination therapy.  4.4 Conclusion In conclusion, I used Saccharomyces cerevisiae as a model to discover SL and SC relationships between DNA repair enzyme inhibitors, TEL1/ATM mutations and sub-lethal doses of DNA damaging therapeutic agents. After directly testing in a small matrix with three DNA repair genes and genome-wide screening in a large matrix with ~5000 non- essential genes, I identified 1 SL interaction with MEC1 and 14 SC interactions with TEL1 and CPT. SL and SC interaction data derived in model organisms may or may not be conserved in humans. The next step will be to validate these results in ATM deficient cultured mammalian cells. The data derived in model organism budding yeast provided the basis for testing specific combination therapies for selective killing of cultured cancer cells bearing ATM mutations. Specifically, the Shu complex, Ku complex, Rrm3, Rad27 and CK2 β subunits can be further tested as potential combination therapy targets with sub-lethal doses of camptothecin to kill ATM-deficient cells.     68 References  ADACHI, N., ISHINO, T., ISHII, Y., TAKEDA, S. & KOYAMA, H. 2001. DNA ligase IV- deficient cells are more resistant to ionizing radiation in the absence of Ku70: Implications for DNA double-strand break repair. Proc Natl Acad Sci U S A, 98, 12109-13. AHMAD, K. A., WANG, G., SLATON, J., UNGER, G. & AHMED, K. 2005. Targeting CK2 for cancer therapy. Anticancer Drugs, 16, 1037-43. AHMED, M. & RAHMAN, N. 2006. ATM and breast cancer susceptibility. Oncogene, 25, 5906-11. ALVINO, G. M., COLLINGWOOD, D., MURPHY, J. M., DELROW, J., BREWER, B. J. & RAGHURAMAN, M. K. 2007. Replication in hydroxyurea: it's a matter of time. Mol Cell Biol, 27, 6396-406. AMBERG, D. C., BURKE, D., STRATHERN, J. N. & COLD SPRING HARBOR LABORATORY. 2005. Methods in yeast genetics : a Cold Spring Harbor Laboratory course manual, Cold Spring Harbor, N.Y., Cold Spring Harbor Laboratory Press. AYENE, I. S., FORD, L. P. & KOCH, C. J. 2005. Ku protein targeting by Ku70 small interfering RNA enhances human cancer cell response to topoisomerase II inhibitor and gamma radiation. Mol Cancer Ther, 4, 529-36. BALL, L. G., ZHANG, K., COBB, J. A., BOONE, C. & XIAO, W. 2009. The yeast Shu complex couples error-free post-replication repair to homologous recombination. Mol Microbiol, 73, 89-102. BANERJEE, S., SMITH, S. & MYUNG, K. 2006. Suppression of gross chromosomal rearrangements by yKu70-yKu80 heterodimer through DNA damage checkpoints. Proc Natl Acad Sci U S A, 103, 1816-21. BENSIMON, A., SCHMIDT, A., ZIV, Y., ELKON, R., WANG, S. Y., CHEN, D. J., AEBERSOLD, R. & SHILOH, Y. 2010. ATM-dependent and -independent dynamics of the nuclear phosphoproteome after DNA damage. Sci Signal, 3, rs3. BERNSTEIN, K. A., REID, R. J., SUNJEVARIC, I., DEMUTH, K., BURGESS, R. C. & ROTHSTEIN, R. 2011. The Shu complex, which contains Rad51 paralogues, promotes DNA repair through inhibition of the Srs2 anti-recombinase. Mol Biol Cell, 22, 1599-607. BOCHMAN, M. L., SABOURI, N. & ZAKIAN, V. A. 2010. Unwinding the functions of the Pif1 family helicases. DNA Repair (Amst), 9, 237-49. BOULTWOOD, J. 2001. Ataxia telangiectasia gene mutations in leukaemia and lymphoma. J Clin Pathol, 54, 512-6. BRACHMANN, C. B., DAVIES, A., COST, G. J., CAPUTO, E., LI, J., HIETER, P. & BOEKE, J. D. 1998. Designer deletion strains derived from Saccharomyces cerevisiae S288C: a useful set of strains and plasmids for PCR-mediated gene disruption and other applications. Yeast, 14, 115-32. BROEKS, A., URBANUS, J. H., FLOORE, A. N., DAHLER, E. C., KLIJN, J. G., RUTGERS, E. J., DEVILEE, P., RUSSELL, N. S., VAN LEEUWEN, F. E. & VAN 'T VEER, L. J. 2000. ATM-heterozygous germline mutations contribute to breast cancer-susceptibility. Am J Hum Genet, 66, 494-500.   69 BROWN, E. J. & BALTIMORE, D. 2000. ATR disruption leads to chromosomal fragmentation and early embryonic lethality. Genes Dev, 14, 397-402. BRYANT, H. E., SCHULTZ, N., THOMAS, H. D., PARKER, K. M., FLOWER, D., LOPEZ, E., KYLE, S., MEUTH, M., CURTIN, N. J. & HELLEDAY, T. 2005. Specific killing of BRCA2-deficient tumours with inhibitors of poly(ADP-ribose) polymerase. Nature, 434, 913-7. BUDD, M. E., TONG, A. H., POLACZEK, P., PENG, X., BOONE, C. & CAMPBELL, J. L. 2005. A network of multi-tasking proteins at the DNA replication fork preserves genome stability. PLoS Genet, 1, e61. CARTER, S. L., EKLUND, A. C., KOHANE, I. S., HARRIS, L. N. & SZALLASI, Z. 2006. A signature of chromosomal instability inferred from gene expression profiles predicts clinical outcome in multiple human cancers. Nat Genet, 38, 1043-8. CHABNER, B. A. & ROBERTS, T. G., JR. 2005. Timeline: Chemotherapy and the war on cancer. Nat Rev Cancer, 5, 65-72. CHAKHPARONIAN, M., FAUCHER, D. & WELLINGER, R. J. 2005. A mutation in yeast Tel1p that causes differential effects on the DNA damage checkpoint and telomere maintenance. Curr Genet, 48, 310-22. CHAN, D. A. & GIACCIA, A. J. 2011. Harnessing synthetic lethal interactions in anticancer drug discovery. Nat Rev Drug Discov, 10, 351-64. CLERICI, M., BALDO, V., MANTIERO, D., LOTTERSBERGER, F., LUCCHINI, G. & LONGHESE, M. P. 2004. A Tel1/MRX-dependent checkpoint inhibits the metaphase-to-anaphase transition after UV irradiation in the absence of Mec1. Mol Cell Biol, 24, 10126-44. CORDA, Y., LEE, S. E., GUILLOT, S., WALTHER, A., SOLLIER, J., ARBEL-EDEN, A., HABER, J. E. & GELI, V. 2005. Inactivation of Ku-mediated end joining suppresses mec1Delta lethality by depleting the ribonucleotide reductase inhibitor Sml1 through a pathway controlled by Tel1 kinase and the Mre11 complex. Mol Cell Biol, 25, 10652-64. DALEY, J. M., PALMBOS, P. L., WU, D. & WILSON, T. E. 2005. Nonhomologous end joining in yeast. Annu Rev Genet, 39, 431-51. DE BONT, R. & VAN LAREBEKE, N. 2004. Endogenous DNA damage in humans: a review of quantitative data. Mutagenesis, 19, 169-85. DE KLEIN, A., MUIJTJENS, M., VAN OS, R., VERHOEVEN, Y., SMIT, B., CARR, A. M., LEHMANN, A. R. & HOEIJMAKERS, J. H. 2000. Targeted disruption of the cell-cycle checkpoint gene ATR leads to early embryonic lethality in mice. Curr Biol, 10, 479-82. DEBRAUWERE, H., LOEILLET, S., LIN, W., LOPES, J. & NICOLAS, A. 2001. Links between replication and recombination in Saccharomyces cerevisiae: a hypersensitive requirement for homologous recombination in the absence of Rad27 activity. Proc Natl Acad Sci U S A, 98, 8263-9. DEL POZO, L., ABARCA, D., CLAROS, M. G. & JIMENEZ, A. 1991. Cycloheximide resistance as a yeast cloning marker. Curr Genet, 19, 353-8. DENG, C., BROWN, J. A., YOU, D. & BROWN, J. M. 2005. Multiple endonucleases function to repair covalent topoisomerase I complexes in Saccharomyces cerevisiae. Genetics, 170, 591-600.   70 FARMER, H., MCCABE, N., LORD, C. J., TUTT, A. N., JOHNSON, D. A., RICHARDSON, T. B., SANTAROSA, M., DILLON, K. J., HICKSON, I., KNIGHTS, C., MARTIN, N. M., JACKSON, S. P., SMITH, G. C. & ASHWORTH, A. 2005. Targeting the DNA repair defect in BRCA mutant cells as a therapeutic strategy. Nature, 434, 917-21. FILLON, M. 2012. Gene linked to pancreatic cancer. J Natl Cancer Inst, 104, 438-9. FOURY, F. & KOLODYNSKI, J. 1983. pif mutation blocks recombination between mitochondrial rho+ and rho- genomes having tandemly arrayed repeat units in Saccharomyces cerevisiae. Proc Natl Acad Sci U S A, 80, 5345-9. FRESCHAUF, G. K., KARIMI-BUSHERI, F., ULACZYK-LESANKO, A., MERENIUK, T. R., AHRENS, A., KOSHY, J. M., RASOULI-NIA, A., PASARJ, P., HOLMES, C. F., RININSLAND, F., HALL, D. G. & WEINFELD, M. 2009. Identification of a small molecule inhibitor of the human DNA repair enzyme polynucleotide kinase/phosphatase. Cancer Res, 69, 7739-46. FRIEDBERG, E. C., AGUILERA, A., GELLERT, M., HANAWALT, P. C., HAYS, J. B., LEHMANN, A. R., LINDAHL, T., LOWNDES, N., SARASIN, A. & WOOD, R. D. 2006. DNA repair: from molecular mechanism to human disease. DNA Repair (Amst), 5, 986-96. FRIEDBERG, E. C., WALKER, G. C. & SIEDE, W. 1995. DNA repair and mutagenesis, Washington, D.C., ASM Press. FRITZ, E., FRIEDL, A. A., ZWACKA, R. M., ECKARDT-SCHUPP, F. & MEYN, M. S. 2000. The yeast TEL1 gene partially substitutes for human ATM in suppressing hyperrecombination, radiation-induced apoptosis and telomere shortening in A-T cells. Mol Biol Cell, 11, 2605-16. GATENBY, R. A. & VINCENT, T. L. 2003. An evolutionary model of carcinogenesis. Cancer Res, 63, 6212-20. GOTTLIEB, T. M. & JACKSON, S. P. 1993. The DNA-dependent protein kinase: requirement for DNA ends and association with Ku antigen. Cell, 72, 131-42. GREENWELL, P. W., KRONMAL, S. L., PORTER, S. E., GASSENHUBER, J., OBERMAIER, B. & PETES, T. D. 1995. TEL1, a gene involved in controlling telomere length in S. cerevisiae, is homologous to the human ataxia telangiectasia gene. Cell, 82, 823-9. GUERRA, B., BOLDYREFF, B., SARNO, S., CESARO, L., ISSINGER, O. G. & PINNA, L. A. 1999. CK2: a protein kinase in need of control. Pharmacol Ther, 82, 303-13. GUERRA, B., ISSINGER, O. G. & WANG, J. Y. 2003. Modulation of human checkpoint kinase Chk1 by the regulatory beta-subunit of protein kinase CK2. Oncogene, 22, 4933-42. GUMY-PAUSE, F., WACKER, P. & SAPPINO, A. P. 2004. ATM gene and lymphoid malignancies. Leukemia, 18, 238-42. HAHN, W. C. 2004. Cancer: surviving on the edge. Cancer Cell, 6, 215-22. HALL, J. 2005. The Ataxia-telangiectasia mutated gene and breast cancer: gene expression profiles and sequence variants. Cancer Lett, 227, 105-14. HANAHAN, D. & WEINBERG, R. A. 2000. The hallmarks of cancer. Cell, 100, 57-70. HARTWELL, L. H., SZANKASI, P., ROBERTS, C. J., MURRAY, A. W. & FRIEND, S. H. 1997. Integrating genetic approaches into the discovery of anticancer drugs. Science, 278, 1064-8.   71 HELLEDAY, T., PETERMANN, E., LUNDIN, C., HODGSON, B. & SHARMA, R. A. 2008. DNA repair pathways as targets for cancer therapy. Nat Rev Cancer, 8, 193- 204. HOEIJMAKERS, J. H. 2001. Genome maintenance mechanisms for preventing cancer. Nature, 411, 366-74. HOEIJMAKERS, J. H. 2009. DNA damage, aging, and cancer. N Engl J Med, 361, 1475-85. HUANG, S. N., POMMIER, Y. & MARCHAND, C. 2011. Tyrosyl-DNA Phosphodiesterase 1 (Tdp1) inhibitors. Expert Opin Ther Pat, 21, 1285-92. IVESSA, A. S., ZHOU, J. Q. & ZAKIAN, V. A. 2000. The Saccharomyces Pif1p DNA helicase and the highly related Rrm3p have opposite effects on replication fork progression in ribosomal DNA. Cell, 100, 479-89. JACKSON, S. P. & BARTEK, J. 2009. The DNA-damage response in human biology and disease. Nature, 461, 1071-8. JALLEPALLI, P. V. & LENGAUER, C. 2001. Chromosome segregation and cancer: cutting through the mystery. Nat Rev Cancer, 1, 109-17. JIANG, H., REINHARDT, H. C., BARTKOVA, J., TOMMISKA, J., BLOMQVIST, C., NEVANLINNA, H., BARTEK, J., YAFFE, M. B. & HEMANN, M. T. 2009. The combined status of ATM and p53 link tumor development with therapeutic response. Genes Dev, 23, 1895-909. KAELIN, W. G., JR. 2005. The concept of synthetic lethality in the context of anticancer therapy. Nat Rev Cancer, 5, 689-98. KANG, B., GUO, R. F., TAN, X. H., ZHAO, M., TANG, Z. B. & LU, Y. Y. 2008. Expression status of ataxia-telangiectasia-mutated gene correlated with prognosis in advanced gastric cancer. Mutat Res, 638, 17-25. KASTAN, M. B. & BARTEK, J. 2004. Cell-cycle checkpoints and cancer. Nature, 432, 316- 23. KIRIENKO, N. V., MANI, K. & FAY, D. S. 2010. Cancer models in Caenorhabditis elegans. Dev Dyn, 239, 1413-48. KITAGAWA, K. & HIETER, P. 2001. Evolutionary conservation between budding yeast and human kinetochores. Nat Rev Mol Cell Biol, 2, 678-87. KRONENWETT, U., HUWENDIEK, S., OSTRING, C., PORTWOOD, N., ROBLICK, U. J., PAWITAN, Y., ALAIYA, A., SENNERSTAM, R., ZETTERBERG, A. & AUER, G. 2004. Improved grading of breast adenocarcinomas based on genomic instability. Cancer Res, 64, 904-9. LARSEN, E., GRAN, C., SAETHER, B. E., SEEBERG, E. & KLUNGLAND, A. 2003. Proliferation failure and gamma radiation sensitivity of Fen1 null mutant mice at the blastocyst stage. Mol Cell Biol, 23, 5346-53. LAVIN, M. F. 2008. Ataxia-telangiectasia: from a rare disorder to a paradigm for cell signalling and cancer. Nat Rev Mol Cell Biol, 9, 759-69. LENGAUER, C., KINZLER, K. W. & VOGELSTEIN, B. 1997. Genetic instability in colorectal cancers. Nature, 386, 623-7. LIU, C., POULIOT, J. J. & NASH, H. A. 2002. Repair of topoisomerase I covalent complexes in the absence of the tyrosyl-DNA phosphodiesterase Tdp1. Proc Natl Acad Sci U S A, 99, 14970-5.   72 LIU, T., WAN, L., WU, Y., CHEN, J. & HUANG, J. 2011. hSWS1.SWSAP1 is an evolutionarily conserved complex required for efficient homologous recombination repair. J Biol Chem, 286, 41758-66. LIU, Y., ZHANG, H., VEERARAGHAVAN, J., BAMBARA, R. A. & FREUDENREICH, C. H. 2004. Saccharomyces cerevisiae flap endonuclease 1 uses flap equilibration to maintain triplet repeat stability. Mol Cell Biol, 24, 4049-64. LOPES, M., COTTA-RAMUSINO, C., PELLICIOLI, A., LIBERI, G., PLEVANI, P., MUZI-FALCONI, M., NEWLON, C. S. & FOIANI, M. 2001. The DNA replication checkpoint response stabilizes stalled replication forks. Nature, 412, 557-61. LUSTIG, A. J. & PETES, T. D. 1986. Identification of yeast mutants with altered telomere structure. Proc Natl Acad Sci U S A, 83, 1398-402. MAKOVETS, S., HERSKOWITZ, I. & BLACKBURN, E. H. 2004. Anatomy and dynamics of DNA replication fork movement in yeast telomeric regions. Mol Cell Biol, 24, 4019-31. MALLORY, J. C. & PETES, T. D. 2000. Protein kinase activity of Tel1p and Mec1p, two Saccharomyces cerevisiae proteins related to the human ATM protein kinase. Proc Natl Acad Sci U S A, 97, 13749-54. MANI, R., ST ONGE, R. P., HARTMAN, J. L. T., GIAEVER, G. & ROTH, F. P. 2008. Defining genetic interaction. Proc Natl Acad Sci U S A, 105, 3461-6. MAO, Z., BOZZELLA, M., SELUANOV, A. & GORBUNOVA, V. 2008. DNA repair by nonhomologous end joining and homologous recombination during cell cycle in human cells. Cell Cycle, 7, 2902-6. MARTEL, V., FILHOL, O., COLAS, P. & COCHET, C. 2006. p53-dependent inhibition of mammalian cell survival by a genetically selected peptide aptamer that targets the regulatory subunit of protein kinase CK2. Oncogene, 25, 7343-53. MARTIN, V., CHAHWAN, C., GAO, H., BLAIS, V., WOHLSCHLEGEL, J., YATES, J. R., 3RD, MCGOWAN, C. H. & RUSSELL, P. 2006. Sws1 is a conserved regulator of homologous recombination in eukaryotic cells. EMBO J, 25, 2564-74. MATEYAK, M. K. & ZAKIAN, V. A. 2006. Human PIF helicase is cell cycle regulated and associates with telomerase. Cell Cycle, 5, 2796-804. MATSUOKA, S., BALLIF, B. A., SMOGORZEWSKA, A., MCDONALD, E. R., 3RD, HUROV, K. E., LUO, J., BAKALARSKI, C. E., ZHAO, Z., SOLIMINI, N., LERENTHAL, Y., SHILOH, Y., GYGI, S. P. & ELLEDGE, S. J. 2007. ATM and ATR substrate analysis reveals extensive protein networks responsive to DNA damage. Science, 316, 1160-6. MAZZORANA, M., PINNA, L. A. & BATTISTUTTA, R. 2008. A structural insight into CK2 inhibition. Mol Cell Biochem, 316, 57-62. MCLELLAN, J., O'NEIL, N., TARAILO, S., STOEPEL, J., BRYAN, J., ROSE, A. & HIETER, P. 2009. Synthetic lethal genetic interactions that decrease somatic cell proliferation in Caenorhabditis elegans identify the alternative RFC CTF18 as a candidate cancer drug target. Mol Biol Cell, 20, 5306-13. MCLELLAN, J. L., O'NEIL, N. J., BARRETT, I., FERREE, E., VAN PEL, D. M., USHEY, K., SIPAHIMALANI, P., BRYAN, J., ROSE, A. M. & HIETER, P. 2012. Synthetic Lethality of Cohesins with PARPs and Replication Fork Mediators. PLoS Genet, 8, e1002574.   73 MCMANUS, K. J., BARRETT, I. J., NOUHI, Y. & HIETER, P. 2009. Specific synthetic lethal killing of RAD54B-deficient human colorectal cancer cells by FEN1 silencing. Proc Natl Acad Sci U S A, 106, 3276-81. MEGGIO, F. & PINNA, L. A. 2003. One-thousand-and-one substrates of protein kinase CK2? FASEB J, 17, 349-68. MICHOR, F., IWASA, Y. & NOWAK, M. A. 2004. Dynamics of cancer progression. Nat Rev Cancer, 4, 197-205. MILANOWSKA, K., KRWAWICZ, J., PAPAJ, G., KOSINSKI, J., POLESZAK, K., LESIAK, J., OSINSKA, E., ROTHER, K. & BUJNICKI, J. M. 2011. REPAIRtoire--a database of DNA repair pathways. Nucleic Acids Res, 39, D788-92. NEGRINI, S., GORGOULIS, V. G. & HALAZONETIS, T. D. 2010. Genomic instability--an evolving hallmark of cancer. Nat Rev Mol Cell Biol, 11, 220-8. PAN, X., YE, P., YUAN, D. S., WANG, X., BADER, J. S. & BOEKE, J. D. 2006. A DNA integrity network in the yeast Saccharomyces cerevisiae. Cell, 124, 1069-81. PAN, X., YUAN, D. S., XIANG, D., WANG, X., SOOKHAI-MAHADEO, S., BADER, J. S., HIETER, P., SPENCER, F. & BOEKE, J. D. 2004. A robust toolkit for functional profiling of the yeast genome. Mol Cell, 16, 487-96. PARSONS, A. B., GEYER, R., HUGHES, T. R. & BOONE, C. 2003. Yeast genomics and proteomics in drug discovery and target validation. Prog Cell Cycle Res, 5, 159-66. PAULSEN, R. D., SONI, D. V., WOLLMAN, R., HAHN, A. T., YEE, M. C., GUAN, A., HESLEY, J. A., MILLER, S. C., CROMWELL, E. F., SOLOW-CORDERO, D. E., MEYER, T. & CIMPRICH, K. A. 2009. A genome-wide siRNA screen reveals diverse cellular processes and pathways that mediate genome stability. Mol Cell, 35, 228-39. PINNA, L. A. 1990. Casein kinase 2: an 'eminence grise' in cellular regulation? Biochim Biophys Acta, 1054, 267-84. POULIOT, J. J., ROBERTSON, C. A. & NASH, H. A. 2001. Pathways for repair of topoisomerase I covalent complexes in Saccharomyces cerevisiae. Genes Cells, 6, 677-87. RAJAGOPALAN, H. & LENGAUER, C. 2004. Aneuploidy and cancer. Nature, 432, 338- 41. RAJAGOPALAN, H., NOWAK, M. A., VOGELSTEIN, B. & LENGAUER, C. 2003. The significance of unstable chromosomes in colorectal cancer. Nat Rev Cancer, 3, 695- 701. REDON, C., PILCH, D. R., ROGAKOU, E. P., ORR, A. H., LOWNDES, N. F. & BONNER, W. M. 2003. Yeast histone 2A serine 129 is essential for the efficient repair of checkpoint-blind DNA damage. EMBO Rep, 4, 678-84. RENWICK, A., THOMPSON, D., SEAL, S., KELLY, P., CHAGTAI, T., AHMED, M., NORTH, B., JAYATILAKE, H., BARFOOT, R., SPANOVA, K., MCGUFFOG, L., EVANS, D. G., ECCLES, D., EASTON, D. F., STRATTON, M. R. & RAHMAN, N. 2006. ATM mutations that cause ataxia-telangiectasia are breast cancer susceptibility alleles. Nat Genet, 38, 873-5. RITCHIE, K. B., MALLORY, J. C. & PETES, T. D. 1999. Interactions of TLC1 (which encodes the RNA subunit of telomerase), TEL1, and MEC1 in regulating telomere length in the yeast Saccharomyces cerevisiae. Mol Cell Biol, 19, 6065-75.   74 SAVITSKY, K., BAR-SHIRA, A., GILAD, S., ROTMAN, G., ZIV, Y., VANAGAITE, L., TAGLE, D. A., SMITH, S., UZIEL, T., SFEZ, S., ASHKENAZI, M., PECKER, I., FRYDMAN, M., HARNIK, R., PATANJALI, S. R., SIMMONS, A., CLINES, G. A., SARTIEL, A., GATTI, R. A., CHESSA, L., SANAL, O., LAVIN, M. F., JASPERS, N. G., TAYLOR, A. M., ARLETT, C. F., MIKI, T., WEISSMAN, S. M., LOVETT, M., COLLINS, F. S. & SHILOH, Y. 1995. A single ataxia telangiectasia gene with a product similar to PI-3 kinase. Science, 268, 1749-53. SCHERENS, B. & GOFFEAU, A. 2004. The uses of genome-wide yeast mutant collections. Genome Biol, 5, 229. SCHULZ, V. P. & ZAKIAN, V. A. 1994. The saccharomyces PIF1 DNA helicase inhibits telomere elongation and de novo telomere formation. Cell, 76, 145-55. SCHVARTZMAN, J. M., SOTILLO, R. & BENEZRA, R. 2010. Mitotic chromosomal instability and cancer: mouse modelling of the human disease. Nat Rev Cancer, 10, 102-15. SHAH, N. A., LAWS, R. J., WARDMAN, B., ZHAO, L. P. & HARTMAN, J. L. T. 2007. Accurate, precise modeling of cell proliferation kinetics from time-lapse imaging and automated image analysis of agar yeast culture arrays. BMC Syst Biol, 1, 3. SHAHEEN, M., ALLEN, C., NICKOLOFF, J. A. & HROMAS, R. 2011. Synthetic lethality: exploiting the addiction of cancer to DNA repair. Blood, 117, 6074-82. SHILOH, Y. 2003. ATM and related protein kinases: safeguarding genome integrity. Nat Rev Cancer, 3, 155-68. SHOR, E., WEINSTEIN, J. & ROTHSTEIN, R. 2005. A genetic screen for top3 suppressors in Saccharomyces cerevisiae identifies SHU1, SHU2, PSY3 and CSM2: four genes involved in error-free DNA repair. Genetics, 169, 1275-89. SMITH, A. M., HEISLER, L. E., MELLOR, J., KAPER, F., THOMPSON, M. J., CHEE, M., ROTH, F. P., GIAEVER, G. & NISLOW, C. 2009. Quantitative phenotyping via deep barcode sequencing. Genome Res, 19, 1836-42. STANKOVIC, T., KIDD, A. M., SUTCLIFFE, A., MCGUIRE, G. M., ROBINSON, P., WEBER, P., BEDENHAM, T., BRADWELL, A. R., EASTON, D. F., LENNOX, G. G., HAITES, N., BYRD, P. J. & TAYLOR, A. M. 1998. ATM mutations and phenotypes in ataxia-telangiectasia families in the British Isles: expression of mutant ATM and the risk of leukemia, lymphoma, and breast cancer. Am J Hum Genet, 62, 334-45. STANKOVIC, T., STEWART, G. S., BYRD, P., FEGAN, C., MOSS, P. A. & TAYLOR, A. M. 2002. ATM mutations in sporadic lymphoid tumours. Leuk Lymphoma, 43, 1563- 71. STORCHOVA, Z. & PELLMAN, D. 2004. From polyploidy to aneuploidy, genome instability and cancer. Nat Rev Mol Cell Biol, 5, 45-54. TADDEI, F., RADMAN, M., MAYNARD-SMITH, J., TOUPANCE, B., GOUYON, P. H. & GODELLE, B. 1997. Role of mutator alleles in adaptive evolution. Nature, 387, 700-2. TAO, Y., LI, X., LIU, Y., RUAN, J., QI, S., NIU, L. & TENG, M. 2012. Structural analysis of Shu proteins reveals a DNA-binding role essential for resisting damage. J Biol Chem.   75 THOMPSON, D., DUEDAL, S., KIRNER, J., MCGUFFOG, L., LAST, J., REIMAN, A., BYRD, P., TAYLOR, M. & EASTON, D. F. 2005. Cancer risks and mortality in heterozygous ATM mutation carriers. J Natl Cancer Inst, 97, 813-22. TOCZYSKI, D. P., GALGOCZY, D. J. & HARTWELL, L. H. 1997. CDC5 and CKII control adaptation to the yeast DNA damage checkpoint. Cell, 90, 1097-106. TOMLINSON, I., SASIENI, P. & BODMER, W. 2002. How many mutations in a cancer? Am J Pathol, 160, 755-8. TONG, A. H. & BOONE, C. 2006. Synthetic genetic array analysis in Saccharomyces cerevisiae. Methods Mol Biol, 313, 171-92. TONG, A. H., LESAGE, G., BADER, G. D., DING, H., XU, H., XIN, X., YOUNG, J., BERRIZ, G. F., BROST, R. L., CHANG, M., CHEN, Y., CHENG, X., CHUA, G., FRIESEN, H., GOLDBERG, D. S., HAYNES, J., HUMPHRIES, C., HE, G., HUSSEIN, S., KE, L., KROGAN, N., LI, Z., LEVINSON, J. N., LU, H., MENARD, P., MUNYANA, C., PARSONS, A. B., RYAN, O., TONIKIAN, R., ROBERTS, T., SDICU, A. M., SHAPIRO, J., SHEIKH, B., SUTER, B., WONG, S. L., ZHANG, L. V., ZHU, H., BURD, C. G., MUNRO, S., SANDER, C., RINE, J., GREENBLATT, J., PETER, M., BRETSCHER, A., BELL, G., ROTH, F. P., BROWN, G. W., ANDREWS, B., BUSSEY, H. & BOONE, C. 2004. Global mapping of the yeast genetic interaction network. Science, 303, 808-13. USHEY, K. 2011. Identifying interaction effects in high-throughput studies of growth. University of British Columbia. VILENCHIK, M. M. & KNUDSON, A. G., JR. 2000. Inverse radiation dose-rate effects on somatic and germ-line mutations and DNA damage rates. Proc Natl Acad Sci U S A, 97, 5381-6. VILK, G., DERKSEN, D. R. & LITCHFIELD, D. W. 2001. Inducible expression of the regulatory protein kinase CK2beta subunit: incorporation into complexes with catalytic CK2 subunits and re-examination of the effects of CK2beta on cell proliferation. J Cell Biochem, 84, 84-99. VOGELSTEIN, B. & KINZLER, K. W. 2001. Achilles' heel of cancer? Nature, 412, 865-6. VOGELSTEIN, B. & KINZLER, K. W. 2004. Cancer genes and the pathways they control. Nat Med, 10, 789-99. WANG, H., WANG, X., ZHOU, X. Y., CHEN, D. J., LI, G. C., ILIAKIS, G. & WANG, Y. 2002. Ku affects the ataxia and Rad 3-related/CHK1-dependent S phase checkpoint response after camptothecin treatment. Cancer Res, 62, 2483-7. WEINBERG, R. A. 2007. The biology of cancer, New York, Garland Science. WELLER, G. R., KYSELA, B., ROY, R., TONKIN, L. M., SCANLAN, E., DELLA, M., DEVINE, S. K., DAY, J. P., WILKINSON, A., D'ADDA DI FAGAGNA, F., DEVINE, K. M., BOWATER, R. P., JEGGO, P. A., JACKSON, S. P. & DOHERTY, A. J. 2002. Identification of a DNA nonhomologous end-joining complex in bacteria. Science, 297, 1686-9. WOGAN, G. N., HECHT, S. S., FELTON, J. S., CONNEY, A. H. & LOEB, L. A. 2004. Environmental and chemical carcinogenesis. Semin Cancer Biol, 14, 473-86. WOOD, R. D., MITCHELL, M. & LINDAHL, T. 2005. Human DNA repair genes, 2005. Mutat Res, 577, 275-83.   76 ZHAO, X., MULLER, E. G. & ROTHSTEIN, R. 1998. A suppressor of two essential checkpoint genes identifies a novel protein that negatively affects dNTP pools. Mol Cell, 2, 329-40. ZHOU, B. B. & ELLEDGE, S. J. 2000. The DNA damage response: putting checkpoints in perspective. Nature, 408, 433-9.      77 Appendices Appendix A  Other positive SC interaction hits validated by RSA   a CSM2 ERG5    78  b TDP1 YBR224W   79  c LTE1 ERG5   80  d BFA1 BUB2   81  e CKA1 CKA2   82  f CKB1 CKB2  The SC interaction between CSM2 and TEL1 had been validated by tetrad dissection followed by spot assay and growth curve assay. No SC interaction between TDP1 and TEL1 had been tested in small matrix. So CSM2 and TDP1 were used as positive control and negative control respectively during RSA to test other screen hits. Other uncovered real SC interaction hits shown here include BFA1, BUB2, LTE1, CKB1, CKB2 and ERG5. One example of false positive hits excluded by RSA (YBR224W) is also shown here.    83 Appendix B  SHU1 SHU2 SC interactions with TEL1 by growth curve assay      

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
United States 3 0
France 3 0
China 1 0
City Views Downloads
Unknown 3 2
Beijing 1 0
Chicago 1 0
Boston 1 0
Ashburn 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}

Share

Share to:

Comment

Related Items