Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A cross species approach to identify potential therapeutic targets through synthetic lethal interactions McLellan, Jessica 2011

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

Item Metadata

Download

Media
24-ubc_2011_fall_mclellan_jessica.pdf [ 16.24MB ]
Metadata
JSON: 24-1.0072371.json
JSON-LD: 24-1.0072371-ld.json
RDF/XML (Pretty): 24-1.0072371-rdf.xml
RDF/JSON: 24-1.0072371-rdf.json
Turtle: 24-1.0072371-turtle.txt
N-Triples: 24-1.0072371-rdf-ntriples.txt
Original Record: 24-1.0072371-source.json
Full Text
24-1.0072371-fulltext.txt
Citation
24-1.0072371.ris

Full Text

A CROSS SPECIES APPROACH TO IDENTIFY POTENTIAL THERAPEUTIC TARGETS THROUGH SYNTHETIC LETHAL INTERACTIONS by Jessica McLellan B. Sc., McGill University, 2006 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Medical Genetics) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  October 2011 © Jessica McLellan, 2011  Abstract Chromosome instability (CIN) is characterized by the loss or gain of large portions of DNA and is characteristic of ~85% of solid tumours. Sequencing of the human orthologues of ~200 genes that cause CIN in yeast identified mutations in approximately 25% of tumours tested. Mutations in cohesin genes and CDC4 represented the two major mutational categories identified. Large scale genetic interaction networks in model organisms can provide insight into the biology underlying tumour mutations and can identify potential therapeutic targets. This approach is based on the concept of synthetic lethality (SL); cell death resulting from the combination of two sub-lethal mutations. Therapies that take advantage of SL distinguish a cancer cell from a normal cell based on their genetic background. This thesis investigated genetic interactions of three cohesin genes, SMC1, SCC1, and SCC2, using high throughput synthetic genetic array (SGA) methods in S. cerevisiae. The overlay of these three genome wide SGA screens and validation using growth curve analysis found that sub-optimal cohesin requires the presence of proteins that mediate replication fork progression. The protruding vulva assay was developed to identify genetic interactions in the somatic cells of C. elegans. It was used to test whether the cohesin interactions were conserved in a multicellular animal. 80% of the validated interactions identified with cohesin in yeast are conserved in C. elegans. Additional fork mediators, namely the pme/PARP family of genes was found to interact with him-1/SMC1 in both C. elegans and human cells. Human cells depleted of SMC1 by siRNA are selectively sensitive to the PARP inhibitor olaparib, currently being evaluated in phase II clinical trials. Additional genetic interaction testing found that CDC4 has a distinct genetic interaction profile from that of cohesin, suggesting different mutational consequences. Work in C. elegans with the lin-23 mutant suggested LIN-23 is involved in controlling CYE-1/Cyclin E levels. lin-23 mutants and human CDC4-/- cells are unable to properly respond to alkylating DNA damage, suggesting CDC4 is important for the DNA damage response. Keywords: colon tumours, chromosome instability, cohesin, CDC4, genetic interactions, SGA, replication fork, PARP.  ii  Preface A version of chapter 2 has been published. 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 C. elegans identify the alternative RFCCTF18 as a candidate cancer drug target. Molecular Biology of the Cell Dec; 20(24): 5306-13. PMID: 19846659. Copyright the American Society for Cell Biology. http://www.molbiolcell.org/content/20/24.toc. I conducted all of the experiments but had help with training and trouble shooting from S. Tarailo. N. O’Neil was instrumental in experimental design and P. Hieter, A. Rose, and N. O’Neil, helped with overall project direction. J. Bryan performed the statistical analysis. J. McLellan, N. O’Neil, and P. Hieter wrote the manuscript upon which this section was based. A version of chapter 3 will be submitted for publication. McLellan J*, O’Neil N*, Barrett I, Ferree E, van Pel DM, Ushey K, Sipahimalani P, Bryan J, Rose A & Hieter P (2011). A cross species candidate approach to identify cohesin genetic interactions reveals that cohesin and PARP genetically interact in human cells. *These authors contributed equally to this work. I performed most of the experiments with significant input from N. O’Neil. N. O’Neil performed some of the C. elegans double mutant construction and brood analysis and I had help in the initial stages of the human tissue culture experiments from I. Barrett and D. van Pel. E. Ferree was an HHMI summer student who helped with the yeast genetic interaction validation stage of this project. P. Sipahimalani performed SGA screens for the essential alleles for SCC2 and SCC1. The statistical analysis was performed by K. Ushey and J. Bryan. J. McLellan, N. O’Neil, and P. Hieter wrote the manuscript upon which this section was based. The mammalian cell culture experiments presented in chapter 4 were performed with help from I. Barrett and the C. elegans experiments were performed with input from N. O’Neil. These experiments involve cell lines, not human tissue and do not require ethics approval.  iii  Table of Contents Abstract	
  .............................................................................................................................................	
  ii	
   Preface	
  ...............................................................................................................................................	
  iii	
   Table of Contents	
  ............................................................................................................................	
  iv	
   List of Tables	
  .................................................................................................................................	
  viii	
   List of Figures	
  ..................................................................................................................................	
  ix	
   List of Abbreviations	
  .....................................................................................................................	
  xi	
   Acknowledgements	
  .......................................................................................................................	
  xiv	
   Dedication	
  ........................................................................................................................................	
  xv	
   Chapter	
  1:	
  Introduction	
  ...............................................................................................................	
  1	
   1.1	
   The genetic basis of cancer	
  .............................................................................................................	
  1	
   1.2	
   Major classes of genes mutated in cancer	
  ...................................................................................	
  2	
   1.3	
   Cohesin genes are mutated in tumours	
  .......................................................................................	
  5	
   1.4	
   Cohesin	
  ...............................................................................................................................................	
  6	
   1.4.1	
   The sister chromatid cohesion cycle	
  ....................................................................................................	
  6	
   1.4.2	
   Cohesin, the DDR, and recombinational repair	
  ...............................................................................	
  8	
   1.4.3	
   Cohesin and regulation of gene expression	
  ....................................................................................	
  10	
   1.5  The role of genetic interactions in interpreting the biology underlying cancer mutations  	
  .....................................................................................................................................................................	
  12	
   1.5.1  Genetic interactions	
  .................................................................................................................................	
  12	
    1.5.2  Synthetic sickness/ lethality	
  .................................................................................................................	
  13	
    1.5.3  Genetic interaction networks	
  ...............................................................................................................	
  14	
    1.5.4  SL and cancer treatment	
  ........................................................................................................................	
  16	
    1.6  Hypothesis and Research Goals	
  ................................................................................................	
  17	
    Chapter	
  2:	
  Development	
  of	
  a	
  cross	
  species	
  approach	
  to	
  identify	
  and	
  quantify	
   genetic	
  interactions	
  ....................................................................................................................	
  18	
   2.1	
   Introduction	
  ....................................................................................................................................	
  18	
   2.1.1	
   C. elegans vulval development	
  ...........................................................................................................	
  20	
    iv  2.2	
   Materials and methods	
  .................................................................................................................	
  22	
   2.2.1	
   S. cerevisiae random spore analysis	
  ..................................................................................................	
  22	
   2.2.2	
   S. cerevisiae growth curves	
  ..................................................................................................................	
  24	
   2.2.3	
   S. cerevisiae growth curve analysis	
  ...................................................................................................	
  27	
   2.2.3.1	
   An alternate definition of genetic neutrality	
  ...........................................................................................	
  32	
   2.2.3.2	
   The special case of smc1-259, ctf8Δ	
  ..........................................................................................................	
  34	
    2.2.4	
   C. elegans general methods and strains	
  ...........................................................................................	
  38	
   2.2.5	
   C. elegans RNAi	
  ......................................................................................................................................	
  38	
   2.2.6	
   C. elegans Pvl assay	
  ...............................................................................................................................	
  39	
   2.3	
   Results	
  ..............................................................................................................................................	
  39	
   2.3.1	
   Genes mutated in colon tumours share genetic interactions with a common set of genes in yeast	
  39	
   2.3.2	
   Random spore and growth curve analysis	
  .......................................................................................	
  40	
   2.3.3	
   Using C. elegans to identify genetic interactions in proliferating somatic cells	
  ................	
  43	
   2.3.4	
   The vulval lineage can be used as a read out for defects in cell division	
  .............................	
  45	
   2.3.5	
   Defects in division are seen in the post-embryonic seam cell lineage	
  ...................................	
  48	
   2.3.6	
   CIN SL genetic interactions in yeast are conserved in C. elegans	
  .........................................	
  50	
   2.3.7	
   Genetic interactions are specific	
  .........................................................................................................	
  53	
   2.3.8	
   Pvl pilot screen	
  .........................................................................................................................................	
  56	
   2.4	
   Discussion	
  ........................................................................................................................................	
  58	
    Chapter 3: A cross-species candidate approach to identify cohesin genetic interactions reveals that cohesin and PARP genetically interact in human cells	
  ...................................	
  63	
   3.1	
   Introduction	
  ....................................................................................................................................	
  63	
   3.2	
   Materials and methods	
  .................................................................................................................	
  66	
   3.2.1	
   S. cerevisiae strain construction and SGA screens	
  ......................................................................	
  66	
   3.2.2	
   Double mutant reconstruction & random spore analysis	
  ...........................................................	
  67	
   3.2.3	
   S. cerevisiae growth curves	
  ..................................................................................................................	
  71	
   3.2.4	
   Growth curve data analysis	
  ..................................................................................................................	
  71	
   3.2.4.1	
   Estimation of strain fitness	
  ...........................................................................................................................	
  72	
   3.2.4.2	
   Estimation of interaction effects	
  .................................................................................................................	
  73	
   3.2.4.3	
   Determination of statistical significance	
  ..................................................................................................	
  74	
    3.2.5	
   C. elegans genetic interactions & SYTO12 staining	
  ...................................................................	
  74	
   3.2.6	
   C. elegans double mutant strain construction & brood analysis	
  .............................................	
  75	
    v  3.2.7	
   HCT116 cell culture & siRNA	
  ...........................................................................................................	
  75	
   3.3	
   Results	
  ..............................................................................................................................................	
  76	
   3.3.1	
   S. cerevisiae cohesin genetic interactions	
  .......................................................................................	
  76	
   3.3.2	
   Negative genetic interactions with cohesin are conserved in C. elegans	
  .............................	
  95	
   3.3.3	
   Analysis of cohesin network sub-groups	
  .........................................................................................	
  97	
   3.3.4	
   SMC1/him-1 mutants have increased apoptosis in the C. elegans germline when fork mediators are disrupted	
  ......................................................................................................................................	
  100	
   3.3.5	
   SMC1/him-1 genetically interacts with the PARP pathway in C. elegans	
  ........................	
  102	
   3.3.6	
   SMC1 genetically interacts with PARP in human cells	
  ...........................................................	
  107	
   3.4	
   Discussion	
  ......................................................................................................................................	
  110	
   3.4.1	
   Cohesin and cancer	
  ..............................................................................................................................	
  110	
   3.4.2	
   Cohesin network	
  ...................................................................................................................................	
  111	
   3.4.3	
   Model organisms for the discovery of therapeutic targets	
  ......................................................	
  114	
    Chapter 4: CDC4 plays a role in the DNA damage response	
  ............................................	
  117	
   4.1	
   Introduction	
  ..................................................................................................................................	
  117	
   4.1.1	
   CDC4 in S. cerevisiae and C. elegans	
  ...........................................................................................	
  119	
   4.2	
   Materials and methods	
  ...............................................................................................................	
  121	
   4.2.1	
   cdc4-10 SGA	
  ..........................................................................................................................................	
  121	
   4.2.2	
   C. elegans general strains and methods	
  ........................................................................................	
  121	
   4.2.3	
   C. elegans MMS treatment	
  ...............................................................................................................	
  122	
   4.2.4	
   Mammalian cell culture	
  ......................................................................................................................	
  122	
   4.3	
   Results	
  ............................................................................................................................................	
  122	
   4.3.1	
   S. cerevisiae CDC4 SGA genetic interactions had little overlap with cohesin interactions 	
    122	
    4.3.2	
   C. elegans has two CDC4 like genes	
  .............................................................................................	
  125	
   4.3.3	
   cye-1/Cyclin E RNAi improves viability of lin-23 mutants	
  ....................................................	
  126	
   4.3.4	
   cye-1(RNAi) reduces seam over proliferation in lin-23 mutants	
  ..........................................	
  129	
   4.3.5	
   lin-23 worms respond inappropriately to MMS induced DNA damage	
  ............................	
  130	
   4.3.6	
   hCDC4-/- human cells show genomic instability in the presence of DNA damage	
  ........	
  131	
   4.4	
   Discussion	
  ......................................................................................................................................	
  134	
   4.4.1	
   The spectrum of CDC4 genetic interactions is distinct from that of cohesin	
  ...................	
  134	
   4.4.2	
   C. elegans as a model to study CDC4 and Cyclin E regulation	
  ...........................................	
  136	
    vi  4.4.3	
   CDC4 loss of function and the DDR	
  .............................................................................................	
  137	
    Chapter 5: Conclusion	
  ...............................................................................................................	
  139	
   5.1	
   C. elegans as a model to identify cancer relevant SL interactions	
  ...................................	
  140	
   5.2	
   Challenges in interpreting data from large scale genetic interaction screens	
  ................	
  142	
   5.3	
   Conservation of genetic interactions	
  .......................................................................................	
  144	
   5.4	
   Fork mediators in the presence of sub-optimal cohesin	
  .....................................................	
  146	
   5.5	
   CDC4 mutations in colon tumours	
  ..........................................................................................	
  148	
   5.6	
   Clinical implications	
  ...................................................................................................................	
  150	
    References	
  ....................................................................................................................................	
  153	
   Appendix	
  .......................................................................................................................................	
  169	
   List of 200 RNAi clones used in the Pvl pilot screen	
  ......................................................................	
  169	
    vii  List of Tables Table 2.1  S. cerevisiae strains used for random spore analysis ........................................... 23	
    Table 2.2  S. cerevisiae strains used for growth curve analysis ............................................ 25	
    Table 2.3  Growth curve ensemble ....................................................................................... 26	
    Table 2.4  Evidence for genetic interactions, based on a minimum or additive definition of  neutrality ................................................................................................................................. 33	
   Table 2.5  Statistical evidence for smc1-258, ctf8D interaction ........................................... 37	
    Table 2.6  C. elegans strains................................................................................................. 38	
    Table 2.7  RNAi clones enhanced in the rrf-3 background .................................................. 52	
    Table 3.1  Diploid S. cerevisiae strains used in this study .................................................... 68	
    Table 3.2  Haploid S. cerevisiae strains used in this study ................................................... 70	
    Table 3.3  C. elegans strains ................................................................................................. 75	
    Table 3.4  Mutations in cohesin genes seen in tumours ....................................................... 77	
    Table 3.5  S. cerevisiae, C. elegans and human orthologs.................................................... 80	
    Table 3.6  Double mutant interactions ranked by T-statistic at 26oC ................................... 91	
    Table 3.7  Double mutant interactions ranked by T-statistic at 30oC ................................... 92	
    Table 3.8  Summary of interactions identified by Growth Curve Analysis.......................... 93	
    Table 3.9  Reported CIN phenotypes of cohesin interacting genes ...................................... 99	
    Table 3.10  SGA scores for the major genes involved in homologous recombination ...... 103	
    Table 4.1  C. elegans strains ............................................................................................... 121	
    Table 4.2  Enriched MIPS categories in the CDC4 SGA screen ........................................ 124	
    Table 4.3  BLAST results for CDC4-like genes in C. elegans and H. sapiens .................. 126	
    viii  List of Figures Figure 2.1  Growth curves analyzed ..................................................................................... 29  Figure 2.2  scal values for the 21 double mutants tested by growth curve analysis ............. 30  Figure 2.3  Strength of the genetic interactions (Sqd) as assayed by growth curve analysis 31  Figure 2.4  Comparison of the additive and minimum definition of genetic neutrality. ...... 34  Figure 2.5  Growth curves for the smc1-259,ctf8Δ double mutant. ...................................... 35  Figure 2.6  Evidence for an smc1-259,ctf8Δ genetic interaction. ......................................... 36  Figure 2.7  Synthetic lethal and synthetic sick interactions in yeast. ................................... 42  Figure 2.8  Paradigms for genetic interaction screening in C. elegans................................ 44  Figure 2.9  him-1(RNAi) causes Pvl...................................................................................... 47  Figure 2.10  him-1(RNAi) causes defects in cell division in an independent lineage ........... 49  Figure 2.11  Genetic interactions of cohesin can be recapitulated in C. elegans ................. 51  Figure 2.12  Genetic interactions are specific rather than general. ....................................... 55  Figure 2.13  C. elegans Pvl pilot screen ............................................................................... 57  Figure 3.1  Cohesin mutations found in colon tumours ........................................................ 78  Figure 3.2  S. cerevisiae SGA network ................................................................................. 82  Figure 3.3  S. cerevisiae growth curves ................................................................................ 85  Figure 3.4  Strain fitness at 26oC ranked by interaction magnitude. .................................... 87  Figure 3.5  Strain fitness at 30oC ranked by interaction magnitude. .................................... 88  Figure 3.6  Interaction T-statistics (26oC) ............................................................................. 89  Figure 3.7  Interaction T-statistics (30oC) ............................................................................. 90  Figure 3.8  Validated S. cerevisiae network ......................................................................... 94  Figure 3.9  C. elegans genetic interactions ........................................................................... 96  Figure 3.10  Cohesin mutations are SL with replication fork mediators ............................ 100  Figure 3.11  Apoptosis in him-1 mutants treated with fork mediator RNAi ...................... 101  Figure 3.12  Knockout of RAD51 does not rescue the lethality of cohesin, fork mediator  double mutants ...................................................................................................................... 104 Figure 3.13  him-1 interacts with pme/PARP mutants in C. elegans .................................. 106  Figure 3.14  Cohesin deficient cells are sensitive to the Parp inhibitor benzamide ........... 108  ix  Figure 3.15  SMC1 siRNA treated human cells are sensitive to the PARP inhibitor olaparib  ............................................................................................................................................... 109 Figure 4.1  cdc4-10 SGA data only partially overlaps with cohesin SGA data.................. 125  Figure 4.2  Phenotypes of lin-23(e1883) and sel-10 mutant worms ................................... 128  Figure 4.3  lin-23 seam cell hyperplasia is partially rescued in lin-23, cye-1(RNAi) ......... 129  Figure 4.4  lin-23 mutants continue to proliferate in the presence of DNA damage .......... 131  Figure 4.5  CDC4-/- cells have elevated Cyclin E ............................................................... 132  Figure 4.6  Human CDC4 prevents aneuploidy after genotoxic stress ............................... 133  Figure 4.7  Micronulei in CDC4-/- cells treated with MMS ................................................ 134  x  List of Abbreviations AC: anchor cell APC: anaphase promoting complex ATP: adenine nucleotide triphosphate AUC: area under the curve BER: base excision repair CDK: cyclin dependent kinase CIN: chromosome instability COSMIC: catalogue of somatic mutations in cancer Damp: decreased abundance by mRNA perturbation DAPI: 4',6-diamidino-2-phenylindole DDR: DNA damage response DMA: deletion mutant array DMSO: dimethyl sulfoxide DSB: double strand break dSLAM: diploid-based synthetic lethality analysis with microarrays Egl: egg laying defective E-MAP: epistatic mini array profiling FACS: fluorescent activated cell sorting FWER: family wise error rate G1: cell cycle growth phase 1 G2: cell cycle growth phase 2 GFP: green fluorescent protein GOF: gain of function Gro: growth defective GTP: guanine nucleotide triphosphate HC: hierarchical clustering HC-DIM: high content digital imaging microscopy Hh: hedgehog proteins HR: homologous recombination xi  HygB: hygromycin B IPTG: isopropyl-beta-D-1-thiogalactopyranoside IR: ionizing radiation Let: lethal LOF: loss of function LOH: loss of heterozygosity LOI: loss of imprinting M: cell cycle mitotic phase MIPS: Munich information center for protein sequences MMR: mismatch repair MMS: methyl methanesulfonate NBD: nuclear binding domain NER: nucleotide excision repair NGM: nematode growth medium OD: optical density PAR: poly (ADP ribosylation) PARP: poly (ADP ribose) polymerase PBS: phosphate buffered saline PCG: polycomb group of proteins PFA: paraformaldehyde PSCS: precocious sister chromatid separation Pvl: protruding vulva RFC: replication factor C RNAi: RNA interference S: cell cycle synthetis phase SAC: spindle assembly checkpoint SCF: SKP1/CDC53-Cullin-F-box SCM: seam cell marker SD: standard deviation SEM: standard error of the mean SGA: synthetic genetic array xii  shRNA: small hairpin RNA siRNA: small interfering RNA SL: synthetic lethality Sma: small body size SMC: structural maintenance of chromosomes SS: synthetic sickness SSC: sister chromatid cohesion SS/L: synthetic sickness/ lethality Ste: sterile Stu: sterile & uncoordinated ts: temperature sensitive VPCs: vulval precursor cells VU: ventral uterine precursor Vul: vulvaless WT: wild type YJM: yeast Jessica McLellan; identifier referring to the yeast database of JMcLellan YPD: yeast peptone dextrose  xiii  Acknowledgements I thank my supervisor, Dr. P Hieter, and my colleague and mentor, Dr. N, O’Neil, for their guidance over the past five years. My committee members, Dr. D. Moerman, Dr. J Kronstad, and Dr. M Kobor, including a recently retired committee member, Dr. D Riddle, have been helpful in keeping this thesis work on track. Their questions and suggestions were very much appreciated. I also thank my colleagues, Dr. P Stirling and Dr. M Bailey for their daily companionship and scientific input. I thank I. Barrett, the Hieter lab Manager, for her constant help in day to day matters, both big and small. I thank my family. My mother and father, Chris and Peter, my brother Thomas, and my sister Emily, for unwavering support. I also thank Todd for putting up with all the ups and downs of working towards a PhD and for always being there to listen and to care. I acknowledge financial support from the University of British Columbia, the Michael Smith Foundation for Health Research, the Natural Sciences and Engineering Research Council, and the Canadian Institutes for Health Research.  xiv  Dedication This thesis is dedicated to my family whose love and support made this work possible.  xv  Chapter 1: Introduction Cancer is a multigenic disease characterized by uncontrolled and inappropriate cell proliferation. Cancerous cells are distinct from normal cells and show genetic and epigenetic changes that lead to phenotypic differences including alterations in gene expression, the accumulation of mutations, inappropriate response to cellular signaling, and resistance to cell death. Much work has been done over the last 50 years to identify the changes that occur in cancer cells and to understand their significance and outcome. This thesis describes work that was done to explore the biology underlying mutations in cohesin genes, which are found in a wide variety of tumours (Barber et al. 2008, Solomon et al. 2011, reviewed in Xu, Tomaszewski & McKay 2011). Genetic analysis of these mutated genes has identified potential cancer therapeutic targets and has shed light on the cellular pathways that are critical in the context of these mutations.  1.1  The genetic basis of cancer  Genetic changes can result in phenotypic alterations. DNA sequencing, first of candidate genes and then of entire genomes, has confirmed that indeed cancerous cells harbour a variety of mutations (reviewed in Stratton 2011). Expression analysis, such as microarrays, have confirmed that some genes are either over expressed or under A note on nomenclature used throughout this thesis: S. cerevisiae gene names are upper case italics, eg. SMC1. S. cerevisiae alleles are lower case italics with an allele identifier, eg. smc1-259. S. cerevisiae proteins are capitalized first letter, eg. Smc1. C. elegans gene names are lower case italics with a dash. eg. him-1. C. elegans alleles are the same as gene names with an added allele identifier in brackets, eg. him-1(e879). C. elegans proteins are upper case with a dash, eg. HIM-1. C. elegans phenotypes are capitalized first letter and abbreviated, eg. Pvl. Human gene names are upper case italics, eg. SMC1. Human proteins are upper case, eg. SMC1. Non-abbreviated human protein names, such as Securin, are only captalized on the first letter.  1  expressed in patient tumour samples (reviewed in Meyerson, Gabriel & Getz 2010). One major challenge now is to identify which mutations have a functional role in tumour formation and or progression and which mutations are phenotypically silent, or act as ‘passenger’ mutations. Cancer is considered a multigenic disease where several gene alterations are required for the cancer phenotype to be manifest (reviewed in Loeb 2011). Which combinations of mutations can result in a cancerous phenotype and how these mutations interact is not clearly understood. As cancer therapy becomes more sophisticated knowing which mutations are relevant for disease state and progression may inform a patient’s treatment strategy (reviewed in Gonzalez-Angulo, Hennessy & Mills 2010). Indeed, an elevated mutational frequency, where mutations in a particular gene occur more often than by chance, suggests that a mutation is causative and this finding is often a starting point for further investigation.  1.2  Major classes of genes mutated in cancer  Mutated genes that have a role in tumour progression can be classified into three main categories: oncogenes, tumour-suppressor genes (TSGs), and genome stability genes (reviewed in Vogelstein & Kinzler 2004). These groupings are not mutually exclusive as mutations in some genes result in phenotypes that can be classified under multiple categories.  Proto-oncogenes, the normal cellular counterpart of oncogenes, typically promote cell growth and proliferation. Expression or protein activity of these positive cell cycle regulators is normally tightly controlled so that cell cycle progression and cell  2  proliferation only occur under the correct circumstances and at the appropriate time. Oncogenes have gain of function (GOF) mutations that uncouple regulation and growth. Activating mutations in a single allele are often sufficient to confer a cellular growth advantage. Treatment strategies are based on interfering with protein function by down regulating activity or inducing degradation. Well studied oncogenes include HER2 and RAS. The tyrosine kinase HER2 is an epidermal growth factor receptor family member involved in growth and differentiation. 20-30% of human breast cancers have increased levels of HER2 and this is associated with poor prognosis (reviewed in Gutierrez & Schiff 2011). Herceptin is a monoclonal antibody that functions to down regulate HER2 signaling (reviewed in Callahan, Hurvitz 2011). RAS is the most commonly mutated oncogene and is a membrane bound guanosine nucleotide binding protein that is controlled by a GTP hydrolysis cycle (reviewed in Vigil et al. 2010). It is key to cell growth control as it stimulates several serine/threonine signal transduction cascades that initiate transcription of genes involved in cell proliferation and differentiation, cell cycle regulation, cell survival and angiogenesis. Oncogenic, activating mutations in RAS are seen in approximately 30% of cancers and inhibition of function by antisense oligonucleotides have been investigated as a cancer therapy (Adjei 2001).  Tumour suppressor genes (TSGs) act to restrict cell proliferation and promote cell death. These genes often impinge on cell cycle checkpoints to oversee critical events, such as DNA replication and chromosome segregation. They initiate and enforce cell cycle delays to allow for completion of these events or correction of errors. Inactivation of tumour suppressors can therefore allow inappropriate cell proliferation. Since loss of  3  function (LOF) is required to confer a cancerous phenotype, mutation of both alleles is typically needed to confer a growth advantage. TP53 is the most commonly mutated TSG and functions as a phosphoprotein involved in the G1 and G2 DNA damage checkpoints (reviewed in Zilfou & Lowe 2009). It controls the transcription of genes that inhibit cyclin dependent kinase (CDK) activity and induces cell cycle arrest. P53 promotes apoptosis if DNA damage cannot be repaired. Upstream proteins that promote P53 activation, such as RB, also function as tumour suppressors while proteins that inhibit its function, such as MDM2, act as proto-oncogenes. Treatment strategies for TSG are fundamentally different from oncogenic mutations because disease often results from loss of function.  Genome stability genes protect the integrity of the genome and alteration of these genes typically results in genomic instability. It is believed that genomic instability can accelerate or facilitate tumour evolution by increasing the probability that protooncogenes and tumour suppressor genes acquire mutations (reviewed in Vogelstein & Kinzler 2004). These mutations are often LOF although GOF or overexpression of some genes can cause instability (Oikawa et al. 2004). Genetic instability can affect large portions of DNA such as entire chromosomes (chromosome instability (CIN)) or small sections of nucleotides (microsatellite instability (MIN)). Stability genes can be divided into three main functional classes: replication, chromosome segregation, and repair. Different pathways predominantly contribute to either large or small scale changes. For example, genes whose encoded proteins participate in the DNA repair pathways mismatch repair (MMR), base excision repair (BER), and nucleotide excision repair  4  (NER) prevent MIN by proof reading newly synthesized DNA and removing aberrant or damaged bases and nucleotides (reviewed in Charames & Bapat 2003). On the other hand, loss of homologous recombination (HR) mediated replication fork restart and DNA repair, such as in BRCA1 and ATM mutants, leads to CIN (reviewed in Florl & Schulz 2008).  1.3  Cohesin genes are mutated in tumours  CIN is seen in approximately 85% of solid tumours, which suggests that large-scale genomic instability plays a pivotal role in tumourgenesis. The link between CIN and tumourigenesis led the Hieter and Vogelstein groups to sequence the human orthologs of approximately 200 yeast CIN genes in a panel of CIN colon tumours. They found that over 20% of samples harboured mutations in genes involved in sister chromatid cohesion (SCC) (Barber et al. 2008, Wang et al. 2004, Rajagopalan et al. 2004, Lyons & Morgan 2011). Mutations in cohesin-associated genes have also been reported in various cancer types in the Catalogue of Somatic mutations in Cancer (COSMIC) database (Xu, Tomaszewski & McKay 2011). The prevalence of cohesin-associated mutations in a variety of tumours argues that the cohesin complex is important in tumour biology. Cohesin-associated genes are best characterized in terms of their role in tethering sister chromatids together until anaphase onset but also have roles in the DNA damage response (DDR), recombinational repair, and control of gene expression (reviewed in Xu, Tomaszewski & McKay 2011). Correct execution and regulation of all of these processes is important to maintain genome integrity and it is currently unclear which aspect of cohesin function is important in cancer development.  5  1.4 1.4.1  Cohesin The sister chromatid cohesion cycle  Smc1, Smc3, Scc1, and Scc3 form the cohesin ring that encircles newly replicated sister chromatids together until anaphase onset (Michaelis, Ciosk & Nasmyth 1997, reviewed in Koshland, Guacci 2000 and Nasmyth, Haering 2009). The SMC (structural maintenance of chromosomes) family of proteins are composed of an intramolecular antiparallel coiled coil that has a globular hinge domain at one end and an ABC type ATP nucleotide binding domain (NBD) at the other (Haering et al. 2002). In cohesin complexes the Smc1/3 NBD is bound by the alpha-kleisin subunit Scc1 to form a tripartite ring (Haering et al, 2002, Haering et al. 2004, Guacci, Koshland & Strunnikov 1997). Scc3 is a HEAT repeat containing protein that binds to Scc1. The loading of cohesin complexes onto chromatin depends on the Scc2/Scc4 complex and occurs starting after mitosis and continuing throughout the cell cycle (Ciosk et al. 2000). Loading of cohesin onto DNA and its dissociation is dynamic and a large fraction of cohesin is thought to cycle between soluble and bound fractions. A cohesin complex is associated with chromatin approximately every 10kb in S. cerevisiae and 20kb in higher organisms (Losada, Hirano & Hirano 1998, Blat, Kleckner 1999, Tanaka et al. 1999, Glynn et al. 2004), although their spatial and temporal organization differs among organisms (Lengronne et al. 2004, Misulovin et al. 2008). SCC is established during S phase and requires acetylation of Smc3 by Eco1/Ctf7, an essential acetyl transferase (Unal et al. 2008, Rolef Ben-Shahar et al. 2008, Skibbens et al. 1999, Toth et al. 1999). Proper cohesion requires the presence of several accessory proteins. Scc3 forms a subcomplex with Pds5 and Rad61/WAPL (Sutani et al. 2009). Specific mutations in  6  SMC3, SCC3, or PDS5 or deletion of RAD61 render the normally essential ECO1 dispensable for viability. These proteins thus have a role in counteracting Eco1’s cohesion establishment activity (Sutani et al. 2009, Rowland et al. 2009). Loss of Pds5 or Rad61 in yeast, however, compromises SCC and it has been suggested that this complex controls cohesin dynamics (Kueng et al. 2006). Cohesion establishment is restricted to S phase by the recent finding that Eco1 is targeted for degradation via the SCFCdc4 proteasome (Lyons & Morgan 2011). A number of accessory proteins such as Ctf4, the alternate replication factor C Ctf18, Rad27, Chl1, Csm3, and Tof1, are also required for efficient SCC (Mayer et al. 2001, Mayer et al. 2004, Skibbens et al. 2004, Hanna et al 2001). How most of these proteins contribute to SCC is currently unknown but one commonality is that they are found at or near the replication fork during S phase, the time when cohesion becomes established (Lengronne et al. 2006).  In most eukaryotic cells cohesion release and dissociation of cohesin complexes from chromosomes occurs in two phases. Most cohesin complexes that tether the chromosome arms dissociate during prophase and prometaphase in a process termed the prophase pathway. During this time centromeric cohesin is protected from dissociation by localization of Shugoshin (Sgo1) that is directed to centromeres by Bub1 (Kitajima et al. 2005, Waizenegger et al. 2000). Shortly before anaphase onset the cohesin that tethers the centromeres and the remaining arm cohesin are released by cleavage of the Scc1 alphakleisin subunit by Esp1/Separase (Uhlmann et al 1999). The timing of the latter cohesin release is regulated by holding separase inactive by binding to Pds1/Securin until chromosomes have achieved biorientation (Michaelis et al. 1997, Cohen-Fix et al. 1996,  7  Ciosk et al. 1998). The prophase pathway, which does not require Scc1 cleavage, is mostly (if not completely) absent in yeast (Sumara et al. 2000). Following cohesin dissociation from chromosomes, Smc3 is deacetylated by Hos1 to allow cohesin complexes to be re-used during the next phases of the cell cycle (Borges et al. 2010)  Aberrant levels, misregulation, and mutation of cohesin proteins have been shown to cause precocious sister chromatid separation (PSCS) (reviewed in Xu, Tomaszewski & McKay 2011), which in turn can lead to aneuploidy and CIN (Barber et al. 2008). CIN is implicated in carcinogenesis as 80% of colorectal tumours exhibit a CIN phenotype (Pino & Chung 2010). PSCS can also occur if the spindle assembly checkpoint (SAC) is compromised. This checkpoint delays anaphase onset until chromosomes are correctly bioriented on the metaphase plate, thereby ensuring proper chromosome segregation by inhibiting Scc1 cleavage by the Anaphase Promoting complex (APC) (reviewed in Yanagida 2000). Mutations in the MAD genes (MAD1, MAD2, MAD3) or the BUB genes (BUB1, BUB2, BUB3) abrogate the ability of the cell to activate the SAC and result in PSCS under microtubule depolymerizing conditions (Straight et al. 1996). Indeed, many human tumours also harbour mutations that compromise the SAC, providing support for the idea that PSCS is important in tumour progression (Chin & Yeong 2010).  1.4.2  Cohesin, the DDR, and recombinational repair  Cohesin has additional roles in the DNA damage response (DDR) and recombinational repair of double strand breaks (DSBs). In human cells cohesins are phosphorylated by Tel1/ATM in response to ionizing radiation (IR) (Yazdi et al. 2002) and by Mec1/ATR in  8  response to replication stress (Bolderson et al. 2004). In yeast, in response to DSBs that occur post S phase, Mec1 and Tel1 phosphorylate histone H2AX (γH2AX) in a region spanning approximately 60kb on either side of the break (Unal et al. 2004). The γH2AX modification is required for de novo cohesin complex loading by Scc2/Scc4 and this binding is enabled by Mre11 (Heidinger-Pauli, Unal & Koshland 2009, Unal et al. 2004, Strom et al. 2004). Newly loaded cohesin spans approximately 50kb on either side of the break and additional cohesin that is loaded genome-wide forms cohesive links between sister chromatids. A signaling cascade where Mec1 phosphorylates Chk1, which in turn phosphorylates Scc1, allows Eco1 to acetylate Scc1. Scc1 acetylation promotes cohesion establishment around the DSB and this establishment is critical to the repair of DSBs (Unal, Heidinger-Pauli & Koshland 2007, Heidinger-Pauli, Unal & Koshland 2009, Sjogren & Nasmyth 2001). Eco1 is the same acetyltransferase that modifies Smc3 in S phase to promote cohesion establishment and in both cases this modification counteracts the anti-establishment activity of Rad61/WAPL (Sutani et al. 2009, Rowland et al. 2009).  Cohesins are also critical during meiosis for the repair of Spo11 induced double strand breaks and for synaptonemal complex axial element formation (Klein et al. 1999). Rec8 replaces Scc1 during meiosis and, along with the other cohesin core components, is required for resolution of chiasmata that exist between chromatids of homologous chromosomes (Buonomo et al. 2000). In mammals meiosis specific versions of Smc1 (SMC1B) (Revenkova et al. 2004) and Scc3/STAG1, STAG2 (STAG3) also exist (Prieto et al. 2001)  9  Cohesin’s involvement in the repair of naturally occurring and exogenous DNA damage through HR may also be important for tumour development. Defective HR can lead to chromosomal aberrations such as loss of heterozygosity (LOH), translocations, inversions and deletions, all of which are characteristic of oncogenesis (reviewed in Heyer, Ehmsen & Liu 2010). The importance of HR proteins in protecting against cancer is exemplified by the increased risk of breast cancer in individuals harbouring germline mutations in BRCA1 and BRCA2 and high mutation rates of HR genes, such as MRE11, in other tumours (reviewed in Trainer et al. 2010). Cohesin’s role as an effector in the DDR sensing pathways and in execution of the S phase checkpoint may also play a role in tumour development. Many groups have documented mutation of central DDR genes, including the upstream kinases ATM, ATR, CHK2, CHK1 and TP53, in human cancers (reviewed in Smith et al. 2010). The inability of cells to repair DNA lesions and to activate the appropriate checkpoints that allow repair to occur are critical to maintenance of genome integrity. Indeed, many DDR and checkpoint genes also cause aneuploidy and CIN (Stirling et al. 2011, Yuen et al. 2007).  1.4.3  Cohesin and regulation of gene expression  The first indication that cohesin may have a role in controlling gene expression came from the observation that mutation of some cohesin genes is the cause of several hereditary disorders characterized by mental and developmental abnormalities. RobertsSC phocomelia syndrome results from defects in the acetytransferase ESCO2 (Vega et al. 2005, Schule et al. 2005), one of the human orthologs of ECO1, and Cornelia de Lange syndrome is due to mutation of NIPBL/SCC2 (Krantz et al. 2004, Tonkin et al. 2004) or  10  SMC1L1 (Musio et al. 2006). Much of the pathology of these disorders is thought to be due to changes in gene expression (reviewed in Dorsett 2007). It was later observed that Smc1 and Smc3 are required to limit heterochromatin spreading from the silent HMR mating locus in yeast (Donze et al. 1999) and that Nipped-B/Scc2 in Drosophila can influence the expression level of several genes by affecting long range enhancerpromoter interactions (Rollins, Morcillo & Dorsett 1999). Cohesin complexes localize to sites of convergently transcribed genes in yeast (Lengronne et al. 2004) and to introns and 5’ untranslated regions of transcription units in Drosophila (Misulovin et al. 2008), suggesting that although the spatial organization is not conserved, the use of transcription to relocate cohesin and the subsequent use of these complexes to control gene expression is common.  In mammals cohesin interacts with the CTCF insulator and higher order chromatin organizing protein (Parelho et al. 2008, Wendt et al. 2008). CTCF, a potential tumour suppressor gene encodes a protein involved in regulating the transcriptional level of several cancer-implicated genes including MYC and CSF2 (reviewed in Wendt, Peters 2009). Additionally, CTCF dependent loss of imprinting (LOI) resulting in aberrant expression of the insulin-like growth factor 2 (IGF2) has been observed (Nativio et al. 2009). LOI of IGF2 has been observed in several cancers and LOI in general is an early and common event in tumourgenesis (Cui 2007, Jelinic, Shaw 2007). A potential orthologue of SCC1 in Drosophila, vtd, is a member of the trithorax group of proteins that suppress polycomb (PcG) and hedgehog (Hh) proteins, and both PcG and Hh are  11  implicated in various tumour types (reviewed in Lin, Chen & Fang 2011, Takebe et al. 2011).  1.5  The role of genetic interactions in interpreting the biology underlying cancer  mutations One approach to identify which cellular processes are compromised in the context of cohesin gene mutations is to look at their pattern of genetic interactions. On a large scale, genetic interaction networks can identify sets of genes that are required in a particular mutant background. The pattern of genetic interactions gives a genome wide impression of processes that are required in for example the context of a cohesin mutation.  1.5.1  Genetic interactions  A genetic interaction exists when two combined genetic perturbations result in a phenotype that deviates from an additive effect. In other words, the resultant phenotype of the double mutant differs significantly from a simple combination of the phenotypes of the two single mutants. The measured phenotype can vary depending on the experimental context and the specific question being asked. For example, studies in yeast that focused on strain fitness have used growth rate or colony size as the phenotypic measure (Tong et al. 2001, Baetz et al. 2004), whereas studies in C. elegans have looked for enhanced embryonic lethality as a measure of a genetic interaction (Byrne et al. 2007, Lehner et al. 2006).  Genetic interactions can be positive or negative (reviewed in Boone, Bussey, Andrews 2007). Positive interactions are also described as ameliorating or suppressive, and occur 12  when the phenotype of the double mutant is less severe than that predicted by a null hypothesis using an additive model of genetic interactions. Negative interactions are also known as aggravating or synergistic, and describe a situation where the double mutant phenotype is more severe than that predicted by an additive model.  1.5.2  Synthetic sickness/ lethality  Synthetic lethality (SL) is a specific type of negative genetic interaction where two sublethal mutations result in cell or organismal lethality when in combination. SL was first documented in the early 1920s when it was observed that crosses between some nonallelic mutations in fruit flies were inviable (Nijman 2011). The term ‘synthetic lethality’ was first used to describe this phenomenon in the mid 1940s (Dobzhansky 1946). Inviability is the most severe form of a negative interaction. Synthetic sickness (SS) is the term used to describe less severe genetic interactions that result in slow growth or sublethal growth impairments.  Hartwell et al. suggested in 1997 that model organisms could be used to predict SL interactions and potential chemotherapeutic targets in humans. In this paradigm a mutation acts like an ideal drug by eliminating or reducing the function of a specific target. SS/L genetic interactions with genes known to be mutated in cancer thus define potential chemotherapeutic targets; since normal cells, by definition, do not harbor cancer associated mutations they would be relatively unaffected. This concept takes advantage of the fact that the mutational profile of a tumour distinguishes it from normal body cells resulting in selective elimination of cancerous cells (Hartwell et al. 1997, Kaelin 2005).  13  McManus et al. provided proof for this principle when they recapitulated an interaction between two yeast genes, RAD27 and RAD54, in mammalian cells. They demonstrated selective killing of a colon cancer cell line deficient for RAD54B when depleted of FEN1, the human homolog of yeast RAD27. RAD54B mutations are found at conserved residues in human colon tumours and primary lymphomas, and this work suggests that inhibitors of FEN1 may be able to selectively eliminate cancerous cells harbouring mutations in RAD54B (McManus et al. 2009).  1.5.3  Genetic interaction networks  The breadth of genetic interaction testing and the ability to perform genome wide screens has greatly increased the amount of interaction data available. Recently, a global profile of the yeast interaction map was published. This work interrogated genetic interactions between 1712 query genes and the entire non-essential yeast deletion mutant collection and identified over 170,000 interactions (Costanzo et al. 2010). Similarly, large screens in C. elegans have examined interactions between 37 query mutants and ~1,750 library genes by RNAi (Lehner et al. 2006). Compiling this information produces genetic interaction networks. As technology continues to advance these types of networks will also be available for human cells and the addition of drugs to yeast genetic interactions assays are adding complexity to these networks (Bandyopadhyay et al. 2010).  Genetic networks are informative because they present a cell wide view of processes and pathways that are required in the presence of specific mutations. Comparison of genetic interaction profiles of different mutants has been informative. Genes with unknown  14  functions have been assigned to pathways and previously unappreciated connections between pathways have been established (Tong et al. 2004, Pan et al. 2004, Schuldiner et al. 2005, Collins et al. 2007, Costanzo et al. 2010). It is widely accepted that genetic networks in different species share organizational characteristics. The extent of conservation was debated when interaction screens in more complex model systems, such as C. elegans, were first performed (Byrne et al. 2007, Lehner et al. 2006). Other studies in C. elegans have suggested that there is indeed a high degree of conservation, on the order of 43%, and that cross species comparisons yield valuable information (Tarailo, Tarailo & Rose 2007). Studies in S. pombe have found that the degree of conservation depends on the subset of genes examined. Genes involved in various aspects of chromosome function show conservation of negative genetic interactions between budding and fission yeast ranging between 17 and 33% (Roguev et al. 2008) whereas studies that interrogated the genetic interactions of transcription factors and kinases found conservation on the order of 4-8% (Beltrao et al. 2009).  Technology has recently developed to the point where genome wide genetic interaction screens are possible in human cells. These screens are typically performed using lentiviral small hairpin RNA (shRNA) libraries targeting the entire genome (Scholl et al. 2009, Barbie et al. 2009, Luo et al. 2009). While model organism screening is more time and cost effective and the data analysis is more straightforward, human screening technology does have more direct relevance to human disease. There are, however, a number of additional advantages that exist while screening in model organisms. The large collections of mutants that exist in yeast, which now contain at least one mutant for most  15  of the genes in the yeast genome (Giaver et al. 2002, Mnaimneh et al 2004, Schuldiner et al. 2005, Breslow et al. 2008, Ben-Aroya et al. 2008), allow screens to be performed without the use of gene knockdown. A large collection of C. elegans mutants has also been generated. Comparatively few matched human cell lines exist with engineered mutations. This breadth of mutants in yeast and C. elegans allows large numbers of screens to be performed and the construction of dense genetic interaction networks. Networks are invaluable if one goal is to identify potential therapeutic targets that will be effective in treating tumours with varying genetic backgrounds. As human screening technology continues to advance, both methods will contribute to our understanding of disease progression and lead to the identification of therapeutic targets in human cells.  1.5.4  SL and cancer treatment  SL can be effective in treating abnormal cancer cells for a variety of reasons. One situation occurs when two redundant pathways are compromised. This can be as straightforward as when two enzymes that are able to catalyze the same reaction are non functional (reviewed in Boone, Bussey & Andrews 2007). It can also be quite complicated as when two alternate pathways that accomplish the same end goal are inactivated. A classic example of this latter situation is the compensatory role of the single strand break (SSB) and double strand break (DSB) repair pathways. When both pathways are inactivated DNA lesions result in replication fork collapse and lethality (Bryant et al. 2005, Farmer et al. 2005). An alternate situation where SL can result in cancer cell death is observed when these abnormal cells become addicted to the presence or overexpression of a particular gene. This has been seen in cervical tumours that  16  overexpress WAPL. Loss of WAPL expression is correlated with cell death suggesting these cells require elevated WAPL for survival (Oikawa et al. 2004).  1.6  Hypothesis and Research Goals  Cohesin genes are mutated at an elevated frequency in colon tumours, suggesting these mutations are functional in promoting disease. The cohesin complex contributes to many processes that are important to maintain genome integrity and prevent cancer development. Genetic interactions of cohesin genes can be used to help understand which processes are important for cancer development, and to inform the biology underlying the disease state. The overlay of individual cohesin genetic interaction profiles to create genetic interaction networks can identify SL interactions that are common between multiple query genes. These common SL interactors are potential chemotherapeutic targets that could prove effective in the treatment of cancers that harbor cohesin mutations.  A) Develop an approach to identify and validate genetic interactions in the single celled yeast S. cerevisiae and the multicellular animal C. elegans. B) Perform unbiased genome-wide genetic interaction screens in S. cereviaise using cohesin gene mutations as queries. C) Validate these interactions in S. cerevisiae and C. elegans using the previously developed platform. D) Test whether these interactions are conserved in human cells.  17  Chapter 2: Development of a cross species approach to identify and quantify genetic interactions 2.1  Introduction  One major challenge in applying the principles of SL to understand tumour biology and for the identification of potential therapeutic targets is the identification of genetic interactions with cancer relevance. It is difficult to assay mutant pairs in human cells, which makes genetically tractable model organisms such as S. cerevisiae and C. elegans enticing for this application. The first goal of this work was thus to develop a platform where genetic interactions could be quantitatively interrogated in a multi-species approach. The aim was then to use individual genetic interactions to build larger interaction networks. The second goal was to identify potential therapeutic targets through SL interactions for cancers that harbor mutations in the subset of SCC defective genes of interest. The cross species approach made use of the single celled yeast S. cerevisiae and the animal model C. elegans. S. cerevisiae affords the ability to quickly screen for genetic interactions for a wide variety of double mutant pairs. On the other hand, C. elegans is multicellular with a more complex gene repertoire akin to that of humans (reviewed in Hodgkin, Plasterk & Waterston 1995). Many genes in yeast are conserved in the nematode but C. elegans also contains orthologs of human genes not found in yeast. These two model organisms thus have complementary properties to investigate genetic interactions. A version of this chapter has been published, see: 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 C. elegans identify the alternative RFCCTF18 as a candidate cancer drug target. Molecular Biology of the Cell Dec; 20(24):5306-13. PMID: 19846659. Copyright the American Society for Cell Biology. http://www.molbiolcell.org/content/20/24.toc 18  Systematic screening of genetic interactions causing SS/L has been performed in S. cerevisiae and to a lesser extent in C. elegans (Tong et al. 2001, Lehner et al. 2006, Tong et al. 2004, Pan et al. 2004, Pan et al. 2006, Davierwala et al. 2005, Tischler et al. 2006, Ceron et al. 2007). The model system where a genetic interaction is quantified will affect the choice of the measured phenotype. One major challenge in this field is that the selection of possible phenotypes to score (to measure a genetic interaction) has not been well defined. For example, in yeast, where strain fitness is only one measure of a genetic interaction, there exist many different techniques to measure strain fitness of single and double mutants. These include: colony spot assays, the ability to grow at elevated temperatures and growth curves performed in liquid media (Baetz et al. 2004). Many techniques are not amenable to quantitative analysis and those that are, such as growth curves, can be analyzed by looking at a number of different parameters. When the combination of two mutants results in a synthetic slow growth phenotype, it is often difficult to determine if a genetic interaction exists. For this reason it is important to be able to evaluate a phenotype quantitatively.  In C. elegans the readout for genetic interactions has been quite different from yeast because nematodes are multicellular organisms. Overall strain fitness is measured by assessing brood size and viability of single and double mutants. Brood analysis is a time consuming process that cannot be easily scaled, however automated microscopy has had some success (Pulak 2006). Currently, in C. elegans, there is no way to systematically create double mutants and therefore most of the high throughput genetic interaction mapping has been done using the RNA interference (RNAi) library, which targets 19  approximately 70-80% of the predicted genes in C. elegans (Kamath et al. 2003). The ability to screen for genetic interactions causing cell lethality is more challenging in multicellular animals because many single mutations that are tolerated in somatic cells or single celled yeast are not tolerated in the developing embryo and singly cause embryonic lethality. This then excludes the ability to use embryonic arrest as an indicator for a synthetic interaction in many cases, as has been done in previous studies (Lehner et al. 2006). To circumvent this problem, this work developed an assay system using the somatic vulval cells of the multicellular animal C. elegans, a cell lineage where knockdown of essential genes does not cause organismal lethality.  2.1.1  C. elegans vulval development  The C. elegans test system developed in this thesis takes advantage of the fact that cell division defects in the vulval cell lineage result in a visible post-embryonic phenotype (Weidhaas et al 2006). The formation of the vulva has been extensively studied and is mediated by an EGFR/RAS/MAPK pathway that is highly conserved from C. elegans to humans (Sundaram 2006). Three vulval precursor cells (VPCs) go through three mitotic cycles to form the wild-type vulva consisting of 22 cells (Sulston, Horvitz 1977). Weidhaas et al. (2006) have shown that loss of vulval cells by non-apoptotic cell death results in a protruding vulva phenotype (Pvl). Treatment with radiation can cause postmitotic vulval cells to die in a stochastic fashion resulting in fewer than 22 vulval cells and the generation of abnormal vulval phenotypes, including Pvl (Weidhaas et al. 2006). Along this same line of reasoning, reduced post-embryonic cell division in the vulval lineage would similarly generate a Pvl phenotype and could be used as an assay for genetic interactions that cause defects in somatic cell proliferation. The screen also 20  capitalizes on the use of post-embryonic RNAi to knock down genes that may be essential for embryonic development. One advantage of RNAi is that the timing of gene knockdown can be controlled and this is especially useful when the genes of interest are essential.  The genes of interest in this study were CDC4, BUB1, MRE11, SMC1, CSPG6/SMC3, STAG3/SCC3, NIPBL/SCC2, and DING/PDS1, which are mutated in CIN colorectal tumours (Barber et al. 2008, Wang et al. 2004, Kemp et al. 2005, Rajagopolan et al. 2004). These genes were initially sequenced in colon tumours because they are known to cause CIN in yeast when mutated. Interestingly, they all have a defect in SCC, suggesting that cohesin is important in tumour initiation and/or progression. Previous studies have shown that CTF4, RAD27, and members of the alternative replication factor C CTF18 (Alt RFCCTF18), CTF18, CTF8, and DCC1, share many common interactions with this particular set of colon mutated genes (Yuen et al. 2007).  This chapter describes a quantitative growth curve assay in yeast to examine interactions between genes mutated in colon tumours and five common interacting genes. These interactions were then assayed in C. elegans using a newly developed and validated Pvl assay for defects in somatic cell proliferation. A high degree of conservation was found between genes with SCC defects and RAD27, CTF4, CTF18, CTF8, and DCC1. The synthetic cell lethality of the colon cancer mutated genes with CTF4, RAD27, CTF8, CTF18 and DCC1 identifies these genes as potential therapeutic targets. A major goal of  21  this chapter was also to develop the methods required for a larger and unbiased study of cohesin genetic interactions described in Chapter Three.  2.2 2.2.1  Materials and methods S. cerevisiae random spore analysis  Heterozygous double mutant strains were first constructed by mating each of the corresponding single mutants. The haploid query gene deletion strains (rad27Δ, ctf4Δ, ctf81Δ, ctf8Δ, dcc1Δ) were disrupted with the HygB (HphMX) resistance gene in strains containing the MAT a specific marker can1Δ::MFA1pr-HIS3::LEU2 or can1Δ::STE2prHIS5+ (Giaever et al. 2002). Four of the colon tumour implicated genes were essential and therefore temperature sensitive (ts) alleles were used to analyze these strains (cdc4-10, smc1-259, scc2-4, smc3-42). Ts alleles were fused to a URA3 marker as described by Ben-Aroya et al. (2008). Knockout alleles for the two non-essential colon tumour implicated genes, MRE11 and BUB1, were created by targeted disruption with the G418 resistance gene. Random spore analysis was performed as described by Tong and Boone (2006) with slight modifications. Briefly, spores from each of the double heterozygous mutant strains were plated onto haploid, single, and double selection plates. Haploid plates were Sc-his-arg+canavanine (60mg/mL) +glutamic acid (0.9g/L) and single selection plates were additionally either +hygromycinB (300mg/L), +G418 (200mg/L) or –ura. Spores were diluted to an OD600 of .01-.1, and plated at volumes of 75µl, 150µl, and 300µl on haploid, both single selection, and double selection plates, respectively. Strains used can be found in Table 2.1.  22  Table 2.1  S. cerevisiae strains used for random spore analysis  Strain YJM358 YJM343 YJM374 YJM325 YJM327 YJM384 YJM96 YJM386 YJM421 YJM423 YJM353 YJM337 YJM366 YJM303 YJM313 YJM348  Relevant Genotype cdc4-10::URA3/CDC4, ctf8Δ::HygB/CTF8 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/CAN1 cdc4-10::URA3/CDC4-10, ctf18Δ::HygB/CTF18 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/CAN1 cdc4-10::URA3/CDC4-10, dcc1Δ::HygB/DCC1 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/CAN1 cdc4-10::URA3/CDC4-10, ctf4Δ::HygB/CTF4 can1Δ::LEU2-MFA1pr::HIS3/CAN1 cdc4-10::URA3/CDC4-10, rad27Δ::HygB/RAD27 can1Δ::LEU2-MFA1pr::HIS3/CAN1 pds1-1::URA3/PDS1, ctf8Δ::HygB/CTF8 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/CAN1 pds1-1::URA3/PDS1, ctf18Δ::HygB/CTF18 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/CAN1 pds1-1::URA3/PDS1, dcc1Δ::HygB/DCC1 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/CAN1 pds1-1::URA3/PDS1, ctf4Δ::HygB/CTF4 can1Δ::LEU2-MFA1pr::HIS3/CAN1 pds1-1::URA3/PDS1, rad27Δ::HygB/RAD27 can1Δ::LEU2-MFA1pr::HIS3/CAN1 smc1-259::URA3/SMC1, ctf8Δ::HygB/CTF8 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/CAN1 smc1-259::URA3/SMC1, ctf18Δ::HygB/CTF18 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/CAN1 smc1-259::URA3/SMC1, dcc1Δ::HygB/DCC1 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/CAN1 smc1-259::URA3/SMC1, ctf4Δ::HygB/CTF4 can1Δ::LEU2-MFA1pr::HIS3/CAN1 smc1-259::URA3/SMC1, rad27Δ::HygB/RAD27 can1Δ::LEU2-MFA1pr::HIS3/CAN1 scc2-4:URA3/SCC2, ctf8Δ::HygB/CTF8 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/ can1Δ::LEU2MFA1pr::HIS3 scc2-4:URA3/SCC2, ctf18Δ::HygB/CTF18 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/ can1Δ::LEU2MFA1pr::HIS3 scc2-4:URA3/SCC2, dcc1Δ::HygB/DCC1 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/ can1Δ::LEU2MFA1pr::HIS3 scc2-4:URA3/SCC2, ctf4Δ::HygB/CTF4 can1Δ::LEU2-MFA1pr::HIS3/CAN1 scc2-4:URA3/SCC2, rad27Δ::HygB/RAD27 can1Δ::LEU2-MFA1pr::HIS3/CAN1 smc3-42::URA3/SMC3, ctf8Δ::HygB/CTF8 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/ can1Δ::LEU2MFA1pr::HIS3 smc3-42::URA3/SMC3, ctf18Δ::HygB/CTF18 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/ can1Δ::LEU2MFA1pr::HIS3 smc3-42::URA3/SMC3, dcc1Δ::HygB/DCC1 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/ can1Δ::LEU2MFA1pr::HIS3 smc3-42::URA3/SMC3, ctf4Δ::HygB/CTF4 can1Δ::LEU2-MFA1pr::HIS3/CAN1 smc3-42::URA3/SMC3, rad27Δ::HygB/RAD27 can1Δ::LEU2-MFA1pr::HIS3/CAN1 scc1-73::URA3/SCC1, ctf8Δ::HygB/CTF8 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/CAN1 scc1-73::URA3/SCC1, ctf18Δ::HygB/CTF18 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/CAN1 scc1-73::URA3/SCC1, dcc1Δ::HygB/DCC1 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/CAN1 scc1-73::URA3/SCC1, ctf4Δ::HygB/CTF4 can1Δ::LEU2-MFA1pr::HIS3/CAN1 scc1-73::URA3/SCC1, rad27Δ::HygB/RAD27 can1Δ::LEU2-MFA1pr::HIS3/CAN1 bub1Δ::G418/BUB1, ctf8Δ::HygB/CTF8 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/ can1Δ::LEU2MFA1pr::HIS3 bub1Δ::G418/BUB1, ctf18Δ::HygB/CTF18 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/ can1Δ::LEU2MFA1pr::HIS3 bub1Δ::G418/BUB1, dcc1Δ::HygB/DCC1 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/ can1Δ::LEU2MFA1pr::HIS3 bub1Δ::G418/BUB1, ctf4Δ::HygB/CTF4 can1Δ::LEU2-MFA1pr::HIS3/CAN1 bub1Δ::G418/BUB1, rad27Δ::HygB/RAD27 can1Δ::LEU2-MFA1pr::HIS3/CAN1 mre11Δ::G418/MRE11, ctf8Δ::HygB/CTF8 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/ can1Δ::LEU2MFA1pr::HIS3 mre11Δ::G418/MRE11, ctf18Δ::HygB/CTF18 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/ can1Δ::LEU2MFA1pr::HIS3  YJM347 YJM362 YJM307 YJM315 YJM349 YJM339 YJM364 YJM305 YJM311 YJM354 YJM335 YJM370 YJM291 YJM297 YJM394 YJM407 YJM398 YJM413 YJM415 YJM392 YJM405  23  Strain YJM403 YJM409 YJM411  Relevant Genotype mre11Δ::G418/MRE11, dcc1Δ::HygB/DCC1 lyp1Δ/LYP1 can1Δ::STE2pr-HIS5+/ can1Δ::LEU2MFA1pr::HIS3 mre11Δ::G418/MRE11, ctf4Δ::HygB/CTF4 can1Δ::LEU2-MFA1pr::HIS3/CAN1 mre11Δ::G418/MRE11, rad27Δ::HygB/RAD27 can1Δ::LEU2-MFA1pr::HIS3/CAN1  All strains are MATa/MATα ura3Δ0/ura3Δ0 leu2Δ0/leu2Δ his3Δ1/his3Δ1 lys2Δ0 or LYS2 met15Δ0 or MET15 in addition to the genotype listed.  2.2.2  S. cerevisiae growth curves  Strains can be found in Table 2.2. MAT a strains containing each of the single and double mutants were isolated by sporulating the double mutant strains constructed for random spore analysis on Sc –his –arg +canavanine (60mg/L). All strains were grown overnight in YPD at 30oC except for strains for smc1-259 analysis, which were grown at 25oC. Cultures were diluted to an OD600 of 0.2 in the morning, allowed to grow for 4 hrs in YPD while shaking, and then diluted to an OD600 of .05 in fresh YPD in 96 well plates. OD620 measurements were taken after 5 minutes of shaking, every 30 minutes, for 24 hours in either a Multiskan Ascent or Tecan M1000 plate reader. At least three replicates were performed for each strain analyzed with higher levels of replication for wild type and the non-essential query gene deletions (rad27Δ, ctf18Δ ctf8Δ, ctf4Δ, and dcc1Δ). The smc1-259, ctf8Δ double mutant (and relevant single mutants and WT strain) was grown at 25oC due to the severe slow growth phenotype of smc1-259 at higher temperatures. smc3-42 double mutants were grown at 32oC due to the mild phenotype of smc3-42 at lower temperatures. All other double mutants were grown at 30oC. Due to this division based on growth temperature and the 9 ‘missing’ double mutants that were SL, the data was partitioned into 6 ensembles, induced by the 5 query genes and the special  24  case of smc1-259,ctf8Δ. These ensembles are summarized in Table 2.3 and growth curves are presented in Figure 2.1.  Table 2.2 Strain YJM425 YJM428 YJM431 YJM434 YJM438 YJM443 YJM444 YJM448 YJM543 YJM450 YJM452 YJM464 YJM466 YJM468 YJM470 YJM472 YJM474 YJM476 YJM479 YJM480 YJM483 YJM484 YJM487 YJM488 YJM490 YJM492 YJM494 YJM496B YJM498 YJM500 YJM502 YJM504 YJM506 YJM510  S. cerevisiae strains used for growth curve analysis Relevant Genotype ctf8Δ::HygB ctf18Δ::HygB dcc1Δ::HygB ctf4Δ::HygB rad27Δ::HygB cdc4-10:URA3 smc1-259::URA3 scc2-4:URA3 smc3-42::URA3 bub1Δ::G418 mre11Δ::G418 ctf8Δ::HygB, cdc4-10::URA3 ctf18Δ::HygB, cdc4-10::URA3 dcc1Δ::HygB, cdc4-10::URA3 ctf4Δ::HygB, cdc4-10::URA3 rad27Δ::HygB, cdc4-10::URA3 ctf8Δ::HygB, smc1-259::URA3 dcc1Δ::HygB, smc1-259::URA3 ctf8Δ::HygB, smc3-42::URA3 ctf18Δ::HygB, smc3-42::URA3 dcc1Δ::HygB, smc3-42::URA3 ctf8Δ::HygB, scc2-4::URA3 ctf18Δ::HygB, scc2-4::URA3 dcc1Δ::HygB, scc2-4::URA3 ctf4Δ::HygB, scc2-4::URA3 ctf8Δ::HygB, bub1Δ::G418 ctf18Δ::HygB, bub1Δ::G418 dcc1Δ::HygB, bub1Δ::G418 rad27Δ::HygB, bub1Δ::G418 ctf8Δ::HygB, mre11Δ::G418 ctf18Δ::HygB, mre11Δ::G418 dcc1Δ::HygB, mre11Δ::G418 ctf4Δ::HygB, mre11Δ::G418  All strains are MATa can1Δ::LEU2-MFA1pr::HIS3 ura3Δ0 leu2Δ0 his3Δ1 lys2Δ0 or LYS2 met15Δ0 or MET15 LYP1 in addition to the genotype listed.  25  Table 2.3  Growth curve ensemble  ctf4Δ  ctf8Δ  ctf18Δ  dcc1Δ  rad27Δ  bub1Δ  rad27Δ  cdc4-10  ensemble ctf4Δ  ctf8Δ  ctf18Δ  dcc1Δ  ensemble  ensemble  ensemble  ensemble  mre11Δ scc2-4 smc3-42 smc1-259  smc1-259 ensemble  Growth curves were performed in batches due to differences in temperature and space availability in the plate reader. Gray blocks denote growth curve ensembles that were analyzed together (6 in total). Blank spaces denote double mutants that were SL when tested by random spore analysis and were therefore not further tested by growth curve analysis  26  2.2.3  S. cerevisiae growth curve analysis  A four-parameter logistic growth model was fit to the growth curves (Pinheiro, Bates 2000): y(x) = A +  B! A 1 + exp[(xmid ! x) / scal]  [1]  where x is time and y is the OD reading, a proxy for cell density or population size. A is the starting point of growth or minimum OD reading and B is the carrying capacity or maximum OD reading. xmid is the time at which 50% of total growth is achieved and scal is the approximate time taken to move from 50 to 75% of growth.  To simultaneously account for the repeated measurement of individual wells and for the systematic effects of gene mutation on growth, we fit a mixed effects logistic growth model using the R package nlme (Pinheiro et al. 2008, R Development Core Team: R Foundation for Statistical Computing 2008) (R code available upon request). Each of the four growth parameters (A, B, xmid, scal) could therefore be modeled with a combination of fixed gene mutation effects and random well effects. To identify genetic interactions, we focused on the scal parameter, which is proportional to doubling time under conditions of unconstrained exponential growth. The logistic growth parameter scal is also inversely related to the underlying exponential growth rate or constant. The model for scal had the following form:  scalqd = Swt + Sq + Sd + Sqd  [2]  27  where q specified the colon tumour implicated genes (q: wild type, bub1Δ, cdc4-10, mre11Δ, scc2-4, smc1-259, and smc3-42) and d specified the query genes (d: wild type, ctf4Δ, ctf8Δ, ctf18Δ, dcc1Δ, and rad27Δ). The term Swt gave the typical value of scal for the wild type strain and the main effect terms Sq and Sd were the typical change in scal associated with a single mutation in a colon implicated gene q or query gene d, respectively. The interaction term Sqd captured the difference between the experimental value of scal for the double mutant (scalqd) and that predicted from a simple sum of Swt and the associated single mutant effects:  Sqd = scalqd – (Swt + Sq +Sd)  [2]  Since scal is related to doubling time, a gene mutation that results in reduced fitness will be associated with a positive main effect, e.g., Sq > 0. In this work we used an additive definition of genetic neutrality which states that a genetic interaction is present when the double mutant’s growth deviates significantly from the expected growth when the effect of both single mutants are combined. The fitness of the double mutant for non-interacting genes using this definition is scalqd = Swt + Sq + Sd  [where Sqd = 0 ]  In the case of a genetic interaction Sqd will be non-zero. Positive values of Sqd indicate synthetic interactions and negative values of Sqd represent alleviating interactions. scal (S) values for all interactions tested are graphically represented in Figure 2.2 and the magnitude of Sqd is graphically represented in Figure 2.3.  28  Growth curves analyzed  mut0  100 200 300 400 500  Lethal  ctf4  bub1  bub1 ctf18  Lethal ctf18 Slow growth  100 200 300 400 500  cdc4−10 mre11  cdc4−10  scc2−4  dcc1  scc2−4  scc2−4  ctf8 Slow growth  dcc1 Slow growth  mre11  cdc4−10 mre11  cdc4−10  scc2−4 ctf8 Slow growth  query0  query0  query0 bub1  bub1 ctf4  Lethal  dcc1  ctf18  smc3−42 smc1−259  mut0  ctf18 Slow growth  rad27  dcc1  ctf18  ctf8  smc3−42 smc1−259  scc2−4  ctf4 Slow growth  dcc1  ctf18  ctf8  ctf8 Slow growth  100 200 300 400 500  ctf18  ctf8  mre11  bub1 cdc4−10  ctf4 Slow growth  smc3−42 smc1−259  1.2 1.0 0.8 0.6 0.4 0.2  ctf4 Slow growth  mut0  scc2−4  1.2 1.0 0.8 0.6 0.4 0.2  ctf4  mre11  mut0  mre11  cdc4−10  mut0  smc3−42 smc1−259  Optical Density(620nm)  1.2 1.0 0.8 0.6 0.4 0.2  Lethal  100 200 300 400 500  ctf8  query0  query0  WT  mut0  bub1  1.2 1.0 0.8 0.6 0.4 0.2  ctf4  dcc1 Lethal dcc1 Slow growth  smc3−42 smc1−259  100 200 300 400 500  WT  smc3−42 smc1−259  Figure 2.1  rad27 Slow growth  rad27 Slow growth  rad27 Lethal  1.2 1.0 0.8 0.6 0.4 0.2  1.2 1.0 0.8 0.6 0.4 0.2  rad27 Lethal rad27 Lethal  1.2 1.0 0.8 0.6 0.4 0.2  rad27 Lethal  100 200 300 400 500  time (minutes)  Query genes are shown in columns and colon tumour implicated genes are shown in rows. Wild type growth curves are shown in the upper left corner. A minimum of three replicates was performed for each strain included in this analysis. The x-axis represents time in minutes and the y-axis represents optical density at 620 nm. A designation of ‘lethal’ (9 strains) or ‘slow growth’ (12 strains including the smc1-259,ctf8Δ special case) is indicated next to each set of growth curves that showed a genetic interaction between the two mutations tested.  29  Figure 2.2  scal values for the 21 double mutants tested by growth curve analysis  scc2−4,ctf4 bub1 ,rad27 mre11 ,ctf4 mre11 ,ctf8 smc3−42,ctf8 mre11 ,ctf18 mre11 ,dcc1 cdc4−10,ctf4 smc3−42,ctf18 cdc4−10,rad27 smc3−42,dcc1 smc1−259,ctf8 bub1 ,ctf8 bub1 ,dcc1 scc2−4,ctf8 scc2−4,ctf18 scc2−4,dcc1 cdc4−10,ctf8  WT qDel dDel doubMut neutPred  cdc4−10,ctf18 cdc4−10,dcc1 bub1 ,ctf18 −20  0  20  scal experimental - predicted scal neutral  The scal values for the wild type, both single mutants, and the double mutant are shown on a single horizontal line. scal values are aligned according to the predicted double mutant value of scal if no genetic interaction exists (vertical line and diamond shapes). Double mutants are ranked according to the strength of the observed interaction. Interactions that were found to be SS are the top 11 ranked double mutants and are separated by the rest of the strains tested by a dashed horizontal line. qDel denotes the colon implicated genes (CDC4, MRE11, BUB1, SMC1, SMC3, and SCC2). dDel denotes the query genes (RAD27, CTF18, CTF8, CTF4, and DCC1). The double mutant (doubMut) is indicated by a square and the wild type (WT) is denoted by a circle. The double mutant analyzed is indicated on the left of the figure, in line with the relevant scal values. Note that smc1-259,ctf8Δ interaction was not found to be SS according to the scal values presented in this figure however it was found to be significant with respect to another growth parameter, B, and was analyzed separately (Figure 2.5, 2.6).  30  Figure 2.3  Strength of the genetic interactions (Sqd) as assayed by growth curve analysis  scc2−4,ctf4 bub1 ,rad27 mre11 ,ctf4 mre11 ,ctf8 smc3−42,ctf8 mre11 ,ctf18 mre11 ,dcc1 cdc4−10,ctf4 smc3−42,ctf18 cdc4−10,rad27 smc3−42,dcc1  synergistic  smc1−259,ctf8  alleviating  bub1 ,ctf8 bub1 ,dcc1 scc2−4,ctf8 scc2−4,ctf18 scc2−4,dcc1 cdc4−10,ctf8 cdc4−10,ctf18 cdc4−10,dcc1 bub1 ,ctf18 −20  −10  0  10  20  30  scal experimental - predicted scal neutral  The magnitude of each bar corresponds to the magnitude of the genetic interaction (Sqd) for each double mutant. Synthetic slow growth interactions are shown with positive values and alleviating interactions are shown with negative values. The zero mark on the x-axis represents no genetic interaction and occurs when the experimental value of scal for the double mutant is equal to the predicted value of scal when calculated under a hypothesis of no genetic interaction or genetic neutrality. The strength of the interaction is the degree of deviation of the experimental value of scal from the predicted value (see supp figure 2). All 11 SS interactions shown in this figure were statistically significant after correction for multiple testing at a p-value of .05. Double mutants are ranked according to the magnitude of the interaction, with SS interactions shown at the top. The associated p-values are shown in Table 2.4.  31  2.2.3.1  An alternate definition of genetic neutrality  We used an alternate definition of neutrality, the minimum neutrality function (54), for comparison. Using this definition, non-interacting mutations yield the fitness of the least fit single mutant. As mentioned above, a reduced fitness will result in a positive S value. For each interaction (q, d), we defined Smore-fit = min(Sq, Sd) and Sless-fit = max(Sq, Sd), which leads to a restatement of equation [2]: [2A]  scalqd = Swt + Sless-fit + Smore-fit + Sqd  The difference between the observed double knockout fitness and that expected under neutrality is then given by:  scalqdfull = Swt + Sless-fit + Smore-fit + Sqd scalqdneut = Swt + Sless-fit scalqdfull ! scalqdneut = Smore-fit + Sqd [2A] where scalfull represents the experimental value of scal for the double mutant strain. scalneut represents the predicted value of scal for the double mutant under the assumption of genetic neutrality or the absence of a genetic interaction. A synthetic sick interaction is observed when scalfull is greater than the value of scalneut or when Smore-fit + Sqd > 0. Likewise an alleviating interaction is observed when scalfull < scalneut. A comparison of the magnitude of the interactions identified using both the additive and the minimum models of neutrality are given in Table 2.4 and Figure 2.4.  32  Table 2.4  Evidence for genetic interactions, based on a minimum or additive definition of neutrality  Double mutant scc2-4,ctf4Δ bub1Δ ,rad27Δ mre11Δ ,ctf4Δ mre11Δ ,ctf8Δ smc3-42,ctf8Δ mre11Δ ,ctf18Δ mre11Δ ,dcc1Δ cdc4-10,ctf4Δ smc3-42,ctf18Δ cdc4-10,rad27Δ smc3-42,dcc1Δ smc1-259,ctf8Δ bub1Δ ,ctf8Δ bub1Δ ,dcc1Δ scc2-4,ctf8Δ scc2-4,ctf18Δ scc2-4,dcc1Δ cdc4-10,ctf8Δ cdc4-10,ctf18Δ cdc4-10,dcc1Δ bub1Δ ,ctf18Δ  DF 1273 1023 1273 2052 2052 2055 2052 1273 2055 1023 2052 504 2052 2052 2052 2055 2052 2052 2055 2052 2055  Additive definition of neutrality tStat p Value 1.40E+01 1.08E-41 1.80E+01 3.47E-63 1.55E+01 1.08E-49 1.29E+01 1.87E-36 1.48E+01 5.21E-47 1.19E+01 1.65E-31 1.01E+01 2.53E-23 9.41E+00 2.18E-20 6.00E+00 2.39E-09 4.04E+00 5.83E-05 3.85E+00 1.21E-04 -3.34E-01 7.39E-01 1.87E+00 6.23E-02 4.88E+00 1.13E-06 6.76E+00 1.85E-11 9.78E+00 4.06E-22 9.57E+00 2.89E-21 1.42E+01 1.32E-43 1.59E+01 1.44E-53 1.68E+01 1.86E-59 2.14E+01 1.02E-91  Minimum definition of neutrality tStat p Value 1.77E+01 6.42E-63 2.48E+01 7.73E-107 2.30E+01 3.30E-98 1.68E+01 2.39E-59 1.98E+01 1.25E-79 2.20E+01 4.06E-96 1.57E+01 2.70E-52 1.59E+01 3.91E-52 1.27E+01 1.34E-35 1.18E+01 4.87E-30 1.07E+01 5.70E-26 2.05E+00 4.12E-02 2.57E+00 1.02E-02 1.56E+00 1.19E-01 2.76E+00 5.80E-03 4.28E+00 1.92E-05 3.34E+00 8.51E-04 1.12E+01 2.84E-28 4.10E+00 4.23E-05 1.17E+01 1.03E-30 8.72E+00 5.76E-18  The t-statistics and p-values for the scal parameter are shown using an ‘additive’ and a ‘minimum’ definition of neutrality. Degrees of freedom (DF) are also indicated. After correction for multiple testing of 21 possible interactions, 11 mutants were found to have a statistically significant p-value of < .05/21. Both models make largely the same predictions and are in accordance with respect to the eleven synthetic slow growth genetic interactions.  33  Figure 2.4  Comparison of the additive and minimum definition of genetic neutrality.  A.  neutral (additive model)  B.  neutral (minimum model)  scc2−4,ctf4 bub1 ,rad27 mre11 ,ctf4 mre11 ,ctf8 smc3−42,ctf8 mre11 ,ctf18 mre11 ,dcc1 cdc4−10,ctf4 smc3−42,ctf18 cdc4−10,rad27 smc3−42,dcc1 smc1−259,ctf8 bub1 ,ctf8  synergistic  bub1 ,dcc1  alleviating  scc2−4,ctf8 scc2−4,ctf18 scc2−4,dcc1 cdc4−10,ctf8 cdc4−10,ctf18 cdc4−10,dcc1 bub1 ,ctf18 −20  0  20  40  −20  0  20  40  Apparent interaction effect (scal)  Ranked interaction effects (scal) when A) an additive model or B) a minimum model of a genetic interaction was used. Note that the two models give the same results with respect to the top 11 SS interactions but some variations exist between the remainder of the interactions.  2.2.3.2  The special case of smc1-259, ctf8Δ  This data was collected at a substantially lower temperature (25oC) than the rest of the growth curves due to the severe phenotype of smc1-259 at higher temperatures. The lower temperature, separate from any gene mutation effects, had a noticeable effect on growth. Specifically, growth of wild type and the single ctf8Δ mutant was initially  34  slower, when compared to growth observed at 30oC or 32oC. Eventually, however, both wild type and ctf8Δ achieved a higher steady state OD reading at the cooler temperature, i.e. more growth was ultimately achieved. Therefore, it was concluded that the carrying capacity, which corresponded to the growth parameter B, seemed to have a different, and perhaps greater, biological meaning at this lower temperature. Due to this temperaturerelated effect, the data relevant to the smc1-259,ctf8Δ interaction was analyzed separately. With respect to the growth parameter scal, the evidence for a genetic interaction was equivocal. It was statistically significant given a minimum model of genetic neutrality but not with an additive model. However, with respect to the growth parameter B, there was striking evidence for a genetic interaction. The growth curves for this case are presented in Figure 2.5, the estimated growth parameters for scal and B are presented in Figure 2.6, and the associated p-values are presented in Table 2.5.  Figure 2.5  Growth curves for the smc1-259,ctf8Δ double mutant.  A.  B. 1.2  1.0  Optical Density (620nm)  Optical Density (620nm)  1.2  WT ctf8  0.8  smc1-259  0.6  0.4  1.0  WT ctf8  0.8  smc1-259  0.6  0.4  smc1-259,ctf8  0.2  100  200  300  Time (minutes)  400  smc1-259,ctf8  0.2  500  100  200  300  400  500  Time (minutes)  A) Raw growth curves and B) smoothed, averaged growth curves for the smc1-259, ctf8Δ double and single mutants.  35  Figure 2.6  Evidence for an smc1-259,ctf8Δ genetic interaction.  WT ●  smc1−259  double mutant  ctf8  A.  predicted  scal  neut (min) neut (add) experimental values 30  35  40  45  50  55  B  B. neut (min) neut (add) experimental values 0.6  0.8  1.0  1.2  Graphical depiction of the A) scal and B) B parameters for the smc1-259,ctf8Δ double mutant strain and the associated control strains. scal is the time required to move from 50 to 75% of maximum growth. It was the growth parameter used to analyze all other growth curves in this work. B is a measure of the carrying capacity of a culture as measured by the maximum OD620 obtained over a 24 hour growth period. The predicted values of scal and B under a hypothesis of a neutral genetic interaction are shown as ‘neut (add)’ and ‘neut (min)’ for the additive and minimum definitions of a genetic interaction, respectively. Both the additive and minimum models for genetic neutrality are shown for scal and B. With respect to the growth parameter scal the evidence for a genetic interaction was equivocal. It was statistically significant given a minimum model but not with an additive model. The carrying capacity, B strongly supported the presence of a statistically significant genetic interaction between smc1-259 and a deletion allele of CTF8 because the experimentally measured B was statistically less than the predicted value using both models of genetic neutrality. Note that a larger value of B correlates with improved growth whereas a smaller value of scal correlates with improved growth.  36  Table 2.5  Statistical evidence for smc1-258, ctf8Δ interaction  growth parameter B scal  DF 504 504  Additive definition of neutrality tStat pValue -33.43 2.73E-130 -0.33 0.63  Minimum definition of neutrality tStat pValue -38.8 8.77E-154 2.05 0.02  The tStat (t-statistic) and p-values are shown for the case of the smc1-259,ctf8Δ double mutant. Values were calculated using both an ‘additive’ and a ‘minimum’ definition of genetic neutrality. Two different growth parameters, B and scal, were considered. scal provides evidence for a genetic interaction using the minimum model but not with the additive model. B supports a statistically significant genetic interaction using either definition of neutrality.  37  2.2.4  C. elegans general methods and strains  Strains used can be found in Table 2.6. All nematodes were grown at 20oC on standard Nematode Growth Medium (NGM) seeded with the E. coli strain OP50 except in cases where worms were fed E. coli expressing RNAi constructs. Young adult worms were imaged on a Zeiss Axioplan 2 Imaging microscope.  Table 2.6  Strain VC2010 CB879 VC193 JR667 SU93 KR4543 KR4544 VC695 VC1245 PS3239  2.2.5  C. elegans strains  Genotype Wild Type him-1(e879) I him-6(ok412) IV unc-119(e2498::Tc1) III; wIs51[SCM::GFP, unc-119(+)] jcIs1[ajm-1::gfp; unc-29(+); rol-6(su1006)] IV him-1(e879) I;wIs5[SCM::GFP, unc-119(+)]1 him-1(e879) I;jcIs1[ajm-1::gfp; unc-29(+); rol-6(su1006)] IV scc-1(ok1017)/mIn1[mIs14 Δpy-10(e128)] II smc-3(ok1703) III/hT2[bli-4(e937) let-?(q782) qIs48](I;III) dpy-20(e1282) syIs49[pMH86(dpy-20(+) + pJB100(zmp-1::GFP)] IV  C. elegans RNAi  RNAi was administered by feeding as previously described by Fraser et al. (2000). Briefly, 5ml bacterial over night cultures containing ampicillin (50mg/ml) and tetracyclin (5mg/ml) were spun down and 4 ml of the supernatant removed. 60ul of concentrated bacterial culture was spotted onto nematode growth medium (NGM) plates containing tetracycline (5mg/ml), ampicillin (50mg/ml) and isopropyl-beta-D-1-thiogalactopyranoside (IPTG) (0.5mM). Plates were incubated at 37oC for 4 hours and then 20oC until the following day. All RNAi constructs, except CTF8, were obtained from the RNAi feeding library (Kamath et al. 2003). A bacterial RNAi clone for CTF8/T22C1.4 was not contained in the library and was kindly provided by K. McManus. unc-22(RNAi) was used as a positive control to ensure RNAi effects could be observed. C. elegans treated with unc-22(RNAi) exhibit a twitching 38  phenotype that is easily observable. If no twitching was observed in unc-22(RNAi) treated worms the entire experiment was discarded. It also served as an additional measure of the background Pvl frequency.  2.2.6  C. elegans Pvl assay  8-30 gravid hermaphrodites were alkaline lysed using 10µl of a 1:1 solution of 2N NaOH and 10-13% sodium hypochlorite onto plates seeded with E. coli expressing different RNAi constructs. L1 worms hatched and then began feeding on the RNAi, allowing for all essential embryonic divisions to occur before RNAi exposure. N2, him-1(e879), and him-6(ok416) adults were scored after 4 days for the presence of Pvl. scc-1(ok1017) and smc-3(ok1703) homozygotes were transferred to fresh RNAi plates after 3 days and scored for Pvl after 5 days due to slower development.  2.3 2.3.1  Results Genes mutated in colon tumours share genetic interactions with a common set  of genes in yeast In order to investigate genetic interactions with multiple genes mutated in colon tumours, a small network comprising the yeast homologs of genes mutated in colon tumours and common interacting partner genes was examined. The yeast orthologs of the colon cancer mutated genes were CDC4, BUB1, MRE11, SMC1, SMC3, SCC3, SCC2, and PDS1. The five potential common “SL gene partners” were RAD27, CTF4, CTF18, CTF8, and DCC1. The partners were selected because the collation of results from several high throughput genetic interaction screens (synthetic genetic array; SGA and heterozygote diploid-based synthetic lethality analysis with microarrays: dSLAM) and direct genetic tests showed that CTF18, 39  CTF8, CTF4, DCC1, and RAD27 genetically interact with many of the yeast orthologs of genes mutated in colon tumours (Tong et al. 2004, Pan et al. 2006, Mayer, Gygi & Aebersold 2001). This work aimed to complete and quantify the synthetic growth phenotypes of this network of genetic interactions by assessing all pairwise combinations of double mutants in yeast.  2.3.2  Random spore and growth curve analysis  The genetic interactions in S. cerevisiae were first assayed by random spore analysis for all pairwise combinations in a 6 x 5 gene matrix. Nine double mutant combinations were found to be lethal and the remaining 21 viable double mutants were further tested by growth curve analysis (Figure 2.1). Growth curve analysis was used to determine if viable double mutants showed a synergistic or an additive effect. The former is indicative of a genetic interaction while the later, in this case, is the combination of two deleterious effects on strain fitness and may indicate epistasis, or the absence of a genetic interaction. Twenty-one mutant combinations were tested for a genetic interaction by growth curve analysis. Smoothed growth curves for all strains can be found in Figure 2.1  Of the 21 strains analyzed 9 double mutants were SL by random spore analysis and 12 double mutant combinations were found to have a synthetic slow growth (SS) phenotype by growth curve analysis (Figure 2.7). Quantification of genetic interactions by growth curves or other quantitative method is an important step because random spore or tetrad analysis does not always clearly identify synthetic interactions. The analysis revealed that the yeast orthologs of several genes mutated in colon tumours share genetic interactions with a common set of genes, namely RAD27, CTF4, and members of the alternative RFCCTF18. This 40  analysis highlighted the importance of these five genes as potential therapeutic targets that may selectively target colon cancer cells that harbour mutations in genes such as SMC1/SMC1L1. SCC1 is the fourth component of the cohesin complex, and although mutations in this gene were not identified in colon tumours, the hypomorphic allele scc1-73 is also SL with deletions in RAD27, CTF4, CTF18, CTF8, and DCC1.  41  Figure 2.7  Synthetic lethal and synthetic sick interactions in yeast.  42  A) The human orthologs of the genes on the outside of the wheel are mutated in colon tumours. Genes on the inside were identified as being SL with a subset of the genes mutated in colon tumours. Solid lines and dashed lines represent SL and SS interactions, respectively. Circles shaded in grey represent essential genes in yeast. The temperature indicated next to the allele indicates the temperature at which genetic interactions were tested. B and C) Examples of growth curves performed that yielded a SS interaction and no interaction, respectively. D) The scal parameter extracted from the ensembles of growth curves in B and C are shown. The growth profile of each curve can be summarized by scal, a value inversely proportional to growth rate. Double mutant strains with a scal value statistically greater than the predicted scal of a neutral genetic interaction were considered SS. scal values for the double mutant, the associated single mutants, and the WT strain are shown on a horizontal line and more poorly growing strains have a greater scal value. PDS1 is not included because single mutants are too slow growing to be informative in growth curve analysis. SCC3 is not included because asuitable ts mutant was not available at the time this work was performed.  2.3.3  Using C. elegans to identify genetic interactions in proliferating somatic cells  This study aimed to test whether the genetic interactions identified in yeast were conserved in a higher organism, and therefore more likely to also be conserved in human cells. C. elegans was chosen because it is a well established metazoan model system with the ability to use RNA interference (RNAi) to knockdown gene function at various stages of development (Figure 2.8). This was important for this study because many of the genetic interactions of interest involved genes that are essential for embryogenesis and therefore these interactions had to be studied in somatic, post-embryonic cells.  43  Figure 2.8  Paradigms for genetic interaction screening in C. elegans  A.  score embryonic lethality  RNAi effect embryonic divisions  postembryonic divisions  B.  embryonic divisions  postembryonic divisions  score postembryonic phenotypes (eg. Pvl)  RNAi effect embryonic divisions  postembryonic divisions  C.  Zygote  embryonic divisions  Seam cell divisions  post-embryonic divisions Vulval cell divisions D.  E.  44  A) Typical method for screening SL genetic interactions using RNAi. RNAi treatment is initiated during the L1-L4 stages of development to deplete maternal, and hence embryonic, stores of mRNA. A frequent readout is enhanced embryonic lethality in the subsequent generation. B) The Pvl screening method where RNAi treatment begins at the L1 stage, once animals have completed the embryonic cell divisions. This experimental approach is designed to identify phenotypes caused by defects in the postembryonic cell divisions, such as Pvl. Black arrows denote when RNAi treatment is initiated. Grey arrows indicate when phenotype scoring takes place. Black wedges depict the phenotypes caused by RNAi knockdown and the associated phenotypic lag. C) C. elegans cell lineage diagram showing the embryonic and post-embryonic cell divisions and highlighting the vulval and seam cell divisions. Figure adapted from the C. elegans server www.umdnj.edu/mobioweb/moss/celegansonline.html. D and E) In the Pvl screening assay worms are visually scored on a Zeiss dissecting microscope at 160x for the presence of a small protrusion, a protruding vulva. D) Wild type worms treated with no RNAi. E) him1 worms treated post embryonically with him-1(RNAi). Black arrow heads denote a protruding vulva.  2.3.4  The vulval lineage can be used as a read out for defects in cell division  The first goal was to determine whether or not the C. elegans vulval cell lineage could be used to identify general defects in somatic cell proliferation. To do so, the frequency of individuals exhibiting a Pvl phenotype was scored after knockdown by RNAi of the essential cohesin subunit HIM-1/Smc1. A loss of function mutation in him-1 results in embryonic lethality (Broverman, Meneely 1994) presumably due to loss of SCC during embryonic cell proliferation. The canonical allele of him-1, e897, is a hypomorphic mutation that does not result in lethality but causes chromosome instability (CIN), characterized by a high frequency of X chromosome non-disjunction, resulting in a High Incidence of Males (Him) phenotype (Hodgkin, Horvitz & Brenner 1979). him-1(e879) also exhibits a low frequency of Pvl (3.9 +/-1.0%) (Figure 2.9A) suggesting that the hypomorphic mutation does have a minor effect on somatic cell proliferation. When RNAi targeting him-1 was administered to the wild-type strain by feeding, after completion of embryogenesis, the percentage of Pvl in the  45  RNAi treated animals was 36.7 +/-1.5%. When him-1(RNAi) was administered to the him1(e879) hypomorph 79.3 +/- 3.6% of animals were Pvl (Figure 2.9).  The zmp-1::GFP marker was next used to count a subset of cells present in the vulva of wild type worms and worms with Pvl to ascertain whether defects were the result of reduced cell numbers. zmp-1 is expressed in the two vulD and vulE cells of the vulva from L4 through adult stages. Expression is also seen in the four vulA cells in the adult stage (Inoue et al. 2002). Adults have a characteristic expression pattern (Figure 2.9E) where the four vulA and two vulD cells are highlighted as four distinct areas of fluorescence. The vulE cells appear to have a more diffuse fluorescent signal and are visible in different plane of focus. The number of GFP positive foci was counted in animals with and without Pvl. 90.4 +/-1.6% of animals with Pvl have fewer than 4 GFP foci, suggesting that these animals have fewer than the wild type number of cells (Figure 2.9D). One caveat of this approach is that a change in cell morphology or expression of GFP could also result in fewer than four GFP foci. Although on its own not conclusive, a reduced number of zmp-1 foci supports the notion that him-1(RNAi) results in fewer cell divisions in the vulval lineage. Collectively, these results reinforced the proposal that the Pvl phenotype could be used as an indicator of defective somatic cell proliferation.  46  Figure 2.9  him-1(RNAi) causes Pvl  A) N2 (wild type) worms and him-1(e879) worms were treated with both him-1(RNAi) and no RNAi and then scored for the presence of Pvl. N2 normally shows a low frequency of Pvl that is enhanced when treated with him-1(RNAi). The him-1 hypomorphic mutant has a low frequency of Pvl when treated with no RNAi. Treatment of him-1 worms with him-1(RNAi) produces a more penetrant phenotype as 79.3 +/-3.6% of animals are Pvl. Error bars represent SEM. n > 500 for each experiment. B) DIC image of N2; (no RNAi) C) DIC image of N2; him-1(RNAi). D) zmp-1::GFP was used to score the number of GFP positive foci in worms with no Pvl and worms with Pvl. Wild type typically have 4 GFP positive foci while 90.4 +/-1.6% of worms with a Pvl have fewer than 4 GFP positve foci. Error bars represent SEM n > 120 for each experiment. E) Image of the 4 GFP foci seen in worms with a wild type vulva. These foci are composed of the 4 VulA cells and the 2 VulD cells. F) Image of a Pvl worm with 2 GFP foci. 47  2.3.5  Defects in division are seen in the post-embryonic seam cell lineage  To confirm that somatic cell loss was responsible for the observed Pvl phenotype, proliferation of an independent post-embryonic cell lineage was assayed. The seam cells are a subset of cells that divide four times in the developing larva, the final divisions occurring at the same time as the vulval cell divisions (Fig 2.8C). These cells normally fuse during the L4 stage to form lateral syncytia along the A-P axis of the worm (Sulston, Horvitz 1977, Hedgecock, White Dev Biol. 1985 Jan;107(1):128-33., Nasmyth, Peters & Uhlmann 2001, Podbilewicz, White 1994). The ajm-1::GFP apical junction marker (Köppen et al. 2001) was used to visualize defects in the normally continuous lateral seam cell syncytium after postembryonic him-1(RNAi) treatment. It was found that him-1(RNAi) caused a statistically significant increase in seam defects in both N2 and him-1 hypomorphic worms (Figure 2.10A-C). To ensure that seam discontinuities were due to defects in cell division rather than defects in cell fusion the SCM::GFP marker (seam cell marker) (Terns et al. 1997) was used. This marker enables the number of seam nuclei in adult worms to be individually counted. Normally adult worms have two sets of 15 seam nuclei on either side of the body, not including H0 (Sulston, Horvitz 1977). When N2 worms were treated post-embryonically with him-1(RNAi) 73% had fewer than 15 seam nuclei and many had irregular nuclear bodies (Figure 2.10D-H). These data demonstrated that discontinuities in the seam syncytium are due to defects in cell division. In conclusion, the loss of him-1 gene function responsible for the Pvl phenotype is also responsible for seam cell loss, indicating that the Pvl phenotype can be used as a general, and easily scorable, readout for proliferation defects in somatic cells.  48  Figure 2.10  him-1(RNAi) causes defects in cell division in an independent lineage  A) ajm-1::GFP worms were treated with no RNAi or him-1(RNAi) and adults were scored for gaps in their lateral seam syncytia. Error bars represent SEM. n > 120 for each experiment. B) ajm-1::GFP;(no RNAi). C) ajm-1::GFP; him-1(RNAi). D) Worms carrying the SCM::GFP marker were treated with either no RNAi or him-1(RNAi). E) SCM::GFP;(no RNAi). 15 nuclei, not including H0, can be identified F) SCM::GFP; him-1(RNAi). Fewer than 15 nuclei are seen in 65% of worms including irregular nuclear bodies (arrow head). G and H) Examples of irregular nuclear bodies. n=30 for each experiment.  49  2.3.6  CIN SL genetic interactions in yeast are conserved in C. elegans  The protruding vulva assay was used to test for conservation of the synthetic genetic interactions identified in yeast, with a focus on the cohesin genes. Interactions with the C. elegans SMC1 ortholog, him-1 were first tested. him-1(e879), a viable hypomorphic mutation, was used to screen for interactions with the C. elegans orthologs of RAD27, CTF4, and three members of the alternative RFCCTF18. As previously mentioned, him-1 exhibits CIN and a low frequency of Pvl (3.9 +/-1.0%). Post-embryonic RNAi was administered by feeding newly hatched him-1(e879) mutant animals to test for an increased frequency of individuals exhibiting the Pvl phenotype. Post-embryonic treatment of him-1(e879) with RNAi against crn-1/RAD27, F17C11.10/CTF4, Y47G6A.8/ CTF18, T22C1.4/CTF8, and K09H9.2/DCC1, caused a statistically significant increase in the number of Pvl individuals (Figure 2.11A). This demonstrated that these yeast genetic interactions are conserved in C. elegans. It was formally possible that the increase in the Pvl phenotype observed in the him-1 mutant was due to enhanced RNAi sensitivity in this mutant background. As part of a separate, larger screen (O’Neil, N. & Tarailo, S.) the wild type and him-1 strains were screened against 17 RNAi clones known to be enhanced in the rrf-3 RNAi sensitive background (Table 2.7) (Simmer et al. 2003). No enhancement of the RNAi phenotype in the him-1 mutant was observed as compared to wild type in all 17 cases. This implied the increases in Pvl were due to bona fide genetic interactions.  50  Figure 2.11  Genetic interactions of cohesin can be recapitulated in C. elegans  A) An increase in Pvl frequency is seen when him-1 worms are treated with RNAi against the C. elegans orthologs of RAD27, CTF4, CTF18, CTF8, and DCC1 but not with control RNAi. B and C) smc-3(ok1703) and scc-1(ok1017) homozygotes show a similar profile of genetic interactions. Error bars represent SEM. n > 500 for each experiment.  51  Table 2.7  RNAi clones enhanced in the rrf-3 background  RNAi gene clone san-1 cdk-8 pfd-6 pd-3 F16D3.4 ben-1 rev-1 F32A11.4 ZK858.1 ubc-1 rfc-2 C54G10.2 kin-20 C08B11.2 F59E10.1 T21D12.9b K09B11.2  Phenotype in N2 WT WT WT WT WT WT WT WT WT WT WT WT WT WT WT WT WT  Phenotype in him-1 WT WT WT WT WT WT WT WT WT WT WT WT WT WT WT WT WT  Phenotype in rrf-3 Pvl Dpy Ste, Pvl Ste, Pvl Ste, Pvl Emb Emb Emb Ste Let Let Gro Gro Let Ste Sma Ste  WT: Wild type, Pvl: Protruding Vulva, Dpy: Dumpy, Ste: Sterile, Emb: Embryonic Lethal, Let: Lethal, Gro: Growth Defective, Sma: Small Body Size.  Colon tumour sequencing identified multiple mutations in cohesin components, highlighting cohesin as a major mutational target. One prediction is that the interactions observed with him-1/SMC1 would be conserved with mutations in the other cohesin subunits. Two additional C. elegans mutant strains defective for cohesin components were analyzed: smc3(ok1703) and scc-1(ok1017), both of which are null mutants. 90% of smc-3 mutant homozygotes reach adulthood and are sterile. The remaining 10% of progeny arrest at the L3 or L4 larval stage. 8.6 +/-2% of the sterile adults exhibited the Pvl phenotype. scc-1 mutant homozygotes had a similar phenotype to smc-3; 50% of larva arrest and the remaining 50% that reach the adult stage are sterile. 10.3 +/-2% of these adult worms are Pvl. Although these mutations are knockouts of essential genes, it was possible to test for genetic interactions using the Pvl assay because those animals reaching the adult stage often developed a wild-  52  type vulva, probably due to maternal contribution of wild type protein. The fact that null mutations and knockdown of essential genes can be assayed with this system is a major advantage over other multicellular genetic screening systems. Both smc-3(ok1703) and scc1(ok1017) mutants showed synthetic interactions with crn-1/RAD27, F17C11.10/CTF4, Y47G6A.8/ CTF18, T22C1.4/CTF8, and K09H9.2/DCC1 in the RNAi assay (Figure 2.11B and C). These results suggest that subunits of the multi-protein cohesin complex share at least a subset of genetic interactions in both yeast and worms, and further emphasize their importance as potential therapeutic targets.  2.3.7  Genetic interactions are specific  In order to determine if the observed Pvl enhancement was a general effect caused by RNAi treatment, we tested him-1(e879) with 15 different control RNAi clones, 10 of which were randomly chosen from chromosome I. None of the 15 RNAi experiments resulted in a significant increase in the frequency of Pvl with either N2 or him-1(e879) (Figure 2.12A). This demonstrated that the increase in Pvl frequency observed when the cohesin mutants are treated with RNAi against crn-1/RAD27, F17C11.10/CTF4, Y47G6A.8/ CTF18, T22C1.4/CTF8, and K09H9.2/DCC1 are specific. Furthermore, three of the control RNAi clones targeted DNA metabolism/repair genes, xpa-1/XPA, rad-51/RAD51, and cku70/CKU70. Notably, him-1(e879) did not show an interaction when treated with these DNA repair gene targeting RNAi clones, thereby demonstrating that the interaction is not due to a general interaction with DNA metabolism genes. Additionally, we tested whether him1(e879) interacted with fkh-10 and F31C3.2, two genes that were previously shown to interact with the checkpoint protein mdf-1/MAD1 in C. elegans (Tarailo, Tarailo & Rose 2007). We did not observe a synthetic increase in Pvl frequency after treatment with the 53  control RNAi constructs, indicating that the interactions identified for the cohesins are specific. In addition, T08B2.5 and C43E11.11 were two of the 10 randomly chosen clones. These clones have been shown to have a phenotype in the RNAi sensitive background, rrf-3, but not in WT (Table 2.7) (Simmer et al. 2003). In this work T08B2.5 and C43E11.11 did not produce a phenotype in either N2 or in the him-1 mutant background, providing further support that him-1 is not at RNAi sensitive strain.  To test whether RAD27, CTF4, and members of the alternative RFCCTF18 were specifically interacting with the cohesin mutants or with chromosome instability mutants in general, we tested another C. elegans chromosome instability mutant, him-6(ok416). him-6 is the RecQ helicase most similar to BLM in humans, which is mutated in a variety of tumours (Wicky et al. 2004, Grabowski, Svrzikapa & Tissenbaum 2005, Ouyang, Woo & Ellis 2008). him-6 homozygotes are viable and exhibit general chromosome non-disjunction that results in a high incidence of males and dead aneuploid embryos. him-6(ok416) have a low frequency of Pvl (5.9 +/-0.5%). A statistically significant increase in Pvl frequency was observed in him-6; crn-1/RAD27(RNAi) and him-6; F17C11.10/CTF4(RNAi) animals but not in him-6 mutants treated with RNAi against components of the alternative RFCCTF18 (Figure 2.12B). This demonstrated that different CIN mutants have specific interaction profiles.  54  Figure 2.12  Genetic interactions are specific rather than general.  A) The frequency of Pvl did not synergistically increase when him-1 was treated with randomly chosen RNAi clones. a. RNAi clones shown to interact with the MAD1/mdf-1 checkpoint component. b. Genes involved in different aspects of DNA metabolism. c. Randomly chosen genes from chromosome I. RNAi against RAD27/crn-1 was included as a positive control. Error bars represent SEM. n > 150 for each experiment. B) him-6 worms show a synthetic interaction when treated with RNAi against RAD27 and CTF4 whereas no interaction is detected with the components of the alternative RFCCTF18. Error bars represent SEM. n > 500 for each experiment.  55  2.3.8  Pvl pilot screen  To assess how the Pvl screen can be scaled to assay a larger number of interactions, 200 genes were screened against the him-1 hypomorph. These genes were hand picked to sample from repair, replication, and cell cycle control including the spindle assembly checkpoint, and chromatin modification pathways. Genes that were described as having a role in maintaining genome repeat stability by a screen performed by Tijsterman et al. were also included. In the Tijsterman screen an out of frame lacZ reporter can induce expression as a measure of repeat instability (Tijsterman, Pothof & Plasterk 2002). RNAi clones that induced expression were considered to have a role in DNA stability. The list of genes, including S. cerevisiae and human homologs, can be found in Appendix A. It is feasible for one person to screen on the order of 50 RNAi clones against a wild typre reference strain and one mutant per week making this approach scalable to a level of several hundred mutant-RNAi tests. This approach is at this point too labor intensive to screen genome wide.  56  s-2  5.3  0 pm  10  3A  20  A. 29  '!"  C5  30  7D 3  i-1 lig -1 hi m F1 1 4F 9 ZK .5 85 6.1 cd kF1 5 8A 1.5 sa n1 em b27 fk h1 CO 0 6H 2 F5 .3 3H 4.1 bi ra r-1 d5/ cl T0 k-2 4A 8 C3 .15 5A 5. pf 9 d2 m rt2 ph g1 F0 2E 9.4 hp r-2 1 rsr -1 ce p1 do g1 b Y5 ir-2 6A 3A Y1 .29 1B 2 F0 a.8 2D 10 . kb 7 p5 xr n1 hi m -6 K0 2F 3 F1 .1 6D 3.4  pr  % of adults with Pvl 40  Y4  B.  sir -2 Y5 .1 4E 2 F5 A.6 7C 1 R0 2.2 4D 3 F4 .3 5E 1.6 K1 0E K0 9.1 7F 5. ra 14 d54 Đƞ Ͳϰ C3 4G 6 R1 .5 0E 4. ct 5 f-1 8 ƌĨĐ ͲϮ tb a1 ƌĨĐ Ͳϰ bu b1 m sh -5 C2 4G F3 6.3 9H 1 C1 1.1 7E 4 m .6 re -1 1 ůŝŐ Ͳϰ R0 6C 7 m .7 df -1 rc q5 K0 7H 8 R1 .1 3F 6 R4 .10 1H 1 D1 0.4 08 F4 1.8 6G 1 F3 0.7 4D 1 C3 0.2 0G 7 lin .1 -2 3 M 18 .5  % of adults with Pvl  Figure 2.13 C. elegans Pvl pilot screen  A.  N2 (WT, VC2010)  him-1(e879)  20  10  0  C. 100  40 80  30 60  40  20  0  57  him-1(e879) was screened aginst 200 RNAi clones for a set of hand picked genes. N was between 80 and 160 for each data point, which is an average of two replicates. Error bars are SD. SD could not be calculated for bars indicated with * because of contamination of one of the two replicates. A) – C) are all from the same data set but have been split into 3 graphs to allow for more detail. Note the difference in scale in C). All plates were initially visually inspected. Only plates that had > 4 Pvl worms on a plate were scored in detail (70/200, shown above). Interactions were ranked according to the difference in magnitude between the frequency of Pvl in the wild tipe and him-1 strains. 2.4  Discussion  In this work genetic interactions that are conserved between S. cerevisiae and C. elegans were identified. RAD27, CTF4, CTF8, CTF18 and DCC1 are potential therapeutic targets because they each interact with multiple colon cancer mutated genes, giving them the potential to selectively eliminate cancerous cells with a variety of mutational backgrounds.  It was found that, as predicted (Yuen et al. 2007), a subset of genetic interactions are common for members of the same protein complex. Null mutations of smc-3 and scc-1 in C. elegans, two additional components of cohesin, share the same tested interactions as SMC1/him-1. The fact that members of the same protein complex share genetic interactions is reasonable from a functional standpoint and has been previously observed and used to identify new protein complexes or new members of known complexes (Collins et al. 2007). Additional work in this area has the potential to identify additional cohesin components and to further elucidate the individual roles of known components. It was also found that him-6, the C. elegans ortholog of the human BLM gene, genetically interacts with CTF4 and RAD27 but not with components of the alternative RFC. This finding suggests that therapeutic targeting of either CTF4 or RAD27 may also be effective in the treatment of cancers harbouring mutations in BLM, such as certain leukemias. These results collectively demonstrated that different CIN mutants have specific interaction profiles. This in turn 58  suggested that different types of cancer will respond uniquely to chemotherapeutic strategies, and that this difference will be in part dependent upon a cancer’s specific mutational profile. For this reason it is important to expand our current genetic networks both in yeast and in more complex multicellular animals, which have a greater gene repertoire and a genome more similar in complexity to a human genome. For the purpose of cancer therapeutic target identification, focus should be placed on mapping the genetic interactions of the genes most often mutated in a wide variety of tumour types.  The experimental approach presented in this work is well poised to comprehensively elucidate the genetic interactions, both essential and non-essential, of genes mutated in tumours. Interestingly, mutagenesis screens using the Pvl phenotype as a readout have been performed previously, however the goal of these screens was to identify genes involved in vulval development (Seydoux, Savage, Greenwald 1993, Eisenmann, Kim 2000). These types of screens identified a host of genes involved in diverse signaling pathways such as EGF, RAS, NOTCH, and WNT signaling, all of which are required for correct vulval development. Mutants in vulval signaling and/or development result in a range of phenotypes including Egg-laying defective (Egl), Vulvaless (Vul), and Pvl. Additional mutants with a role in cell division were also recovered from these banks of Pvl mutants, including evl14/PDS5 and scc-3/SCC3, (Wang et al. 2003). This latter finding demonstrated that the Pvl phenotype could also be used to identify cell division mutants. Additionally, several genes known to be involved in the cell cycle, such as air-1/aurora kinase, air-2, sep-1/separase, cdk-4/CDK4/6, and san-1/MAD3, exhibit a Pvl phenotype when mutated (Simmer et al. 2003, Park, Krause 1999, Woollard, Hodgkin 1999, Furuta, Baillie & Schumacher 2002). Similarly,  59  chromosome stability mutants, such as the catalytic subunit of telomerase, trt-1, are also Pvl (Meier et al. 2006). These mutant phenotypes are consistent with the view that mutations in genes required for the cell division cycle or to maintain genomic integrity result in defects in the vulval post-embryonic lineage and can give rise to the Pvl phenotype.  In theory, defects in any post-embryonic lineage could be used to identify cell division cycle mutants. O’Connell et al. (1998) used this concept to screen for temperature-sensitive cell division mutants by screening for animals with the Stu (sterile and uncoordinated) phenotype. Along this same reasoning, defects in the post-embryonic lineages could theoretically be exploited to identify combinations of mutants that result in cell proliferation defects and we chose to use the vulval lineage for genetic screening because defects are easily identified and quantified. Some of the advantages of this technique result from the single generation screening approach that has allowed us to interrogate the interactions of essential genes.  This study shed some light on the question of whether or not genetic interaction networks are conserved between organisms. Previous studies by Lehner et al. (2006) and Byrne et al. (2007) focused on genes involved in signaling pathways and suggested that a relatively low percentage of genetic interactions are conserved. On the other hand, work by Tarailo et al. (2007) that focused on the spindle assembly checkpoint suggested that a much higher fraction of interactions, on the order of 43%, are conserved between S. cereviaise and C. elegans. Work by Roguev et al. in S. pombe suggested that between 17 and 33% of negative genetic interactions are conserved between budding and fission yeast, with respect to genes  60  involved in chromosome biology (Roguev et al. 2008). The degree of conservation appears to be on the order of 4-8% for more rapidly evolving genes, such as kinases and transcription factors (Beltrao et al. 2009). Recent work by McManus et al. (2009) demonstrated the usefulness of a cross-species candidate approach for drug target identification. They showed that an interaction first identified in yeast between RAD27 and RAD54 (Symington 1998) was conserved in human HCT116 cells.  This data suggested that there is a high degree of conservation between species, at least with respect to essential genes involved in chromosome biology. One caveat of this cross-species candidate approach is that the selection pressures that shape genetic networks in a tumour may be quite different from those in a model organism. Further genetic interaction testing of essential genes in model organisms and tumour derived cell lines will help to answer these questions.  The interaction network in yeast and the conserved interactions identified in somatic cell divisions of C. elegans suggest that CTF4, RAD27 and members of the alternative RFCCTF18 are important therapeutic targets worthy of further investigation. The alternative RFC components are especially attractive as therapeutic targets because of their relative lack of phenotype when knocked down singly in C. elegans and mammalian cells. On the other hand, others have found that single knockdown of CTF4 or RAD27 has a detrimental effect on otherwise normal cells (Zhu et al. 2007, Saharia et al. 2008). The ideal cancer therapeutic would eliminate cancerous cells while causing no harm to normal cells. Potentially, many genes have a SL relationship with genes mutated in cancer but the best therapeutic targets are  61  those genes that cause the least amount of harm to normal body cells during treatment. For genes whose knockdown causes a detrimental effect on normal cells it may be a matter of titrating the therapeutic dose to induce selective killing of cancerous cells. The approach presented in this study identified DCC1, CTF8, and CTF18 as potential cancer therapeutic targets. The C. elegans Pvl assay is a powerful tool for many reasons. It is a convenient approach to test for conservation of genetic interactions identified in simpler organisms. This will help to prioritize testing of interactions in mammalian cells with the ultimate goal of therapeutic target identification and drug development. The Pvl assay also offers the possibility of detecting new SL genetic interactions not present in yeast. This type of testing in a multicellular animal (with a genome similar in complexity to the human genome) may greatly benefit our understanding of genes that cause slow growth or lethality in combination with genes mutated in cancers, thereby expanding our list of potential therapeutic targets.  62  Chapter 3: A cross-species candidate approach to identify cohesin genetic interactions reveals that cohesin and PARP genetically interact in human cells 3.1  Introduction  Defects in cohesin-associated genes are emerging as potential drivers of tumour progression. Sequencing of over 200 human orthologs of yeast chromosome instability (CIN) genes from 130 colon tumours found that cohesin genes are mutated in 8% of tumour samples (Barber et al. 2008). Mutations in cohesin genes have also been identified in other tumour types (Reviewed in Xu, Tomaszewski & McKay 2011). Furthermore, altered cohesin gene expression, either overexpression or underexpression is characteristic of many tumours (Oikawa et al. 2004, Xu et al. 2011, Ghiselli, Iozzo 2000, Zhang et al. 2008, Hagemann et al. 2011). These observations suggested that cohesin dysfunction can contribute to tumour development and progression. Cohesin genes are defined by their contribution to sister chromatid cohesion (SCC) and screens for defects in SCC have identified the core cohesin complex, composed of Smc1, Smc3, Scc1 and Scc3, and additional accessory and regulatory proteins (Michaelis, Ciosk & Nasmyth, reviewed in Koshland, Guacci 2000 and Nasmyth, Haering 2009). Cohesin proteins contribute to several cellular processes including chromosome segregation, DNA repair, and regulation of gene expression (reviewed in Xu, Tomaszewski & McKay 2011). Although much is known about the function of cohesin in regulating SCC and  A version of this work has been submitted for publication. See McLellan J, O’Neil N, Barrett I, Ferree E, van Pel DM, Ushey K, Sipahimalani P, Bryan J, Rose A & Hieter P. Cohesin are syhthetic lethal with PARPs and replication fork mediators, 2011. 63  DNA damage repair, it is unclear how cohesin mutations might contribute to tumour progression.  One approach to understanding the functional spectrum associated with a gene of interest relies on the identification of genetic interactions with other gene mutations. Comprehensive genetic interaction networks can lead to new functional insights (Collins et al. 2007, Costanzo et al. 2010). Synthetic genetic array (SGA) and epistatic mini array profiling (E-MAP) are two large-scale genetic interaction screening approaches in yeast that facilitate the collection and analysis of positive and negative genetic interaction data (Vogelstein & Kinzler 2004, Tong et al. 2001, Collins et al. 2007). Negative genetic interactions occur when the double mutant shows a synthetic growth defect, manifest as severe slow growth or lethality (synthetic sickness/lethality, SS/L) when compared to either single mutant (reviewed in Boone, Bussey, Andrews 2007). SS/L interactions with genes mutated in cancer can identify potential therapeutic targets (Hartwell et al. 1997). An excellent example of a synthetic lethal genetic interaction with relevance to cancer chemotherapy is the SL interaction between mutations in the breast cancer susceptibility genes BRCA1 or BRCA2 and loss of function of the Poly-ADP Ribose Polymerases (PARPs). Two groups have shown that both BRCA1- and BRCA2-defective cells are selectively sensitive to knockdown of PARP or chemical inhibition of PARP activity (Bryant et al. 2005, Farmer et al. 2005). PARP inhibitors are being evaluated in Phase II clinical trials for their use in breast and ovarian cancer therapy (Audeh et al. 2010, Fong et al. 2010, O'Shaughnessy et al. 2011, Khan et al. 2011).  64  One challenge is to screen large numbers of gene pairs for genetic interactions with genes found mutated in cancers. Genetically amenable model organisms can provide a platform in which to rapidly query large numbers of potential genetic interactions. The use of model organisms to identify genetic interactions with potential cancer therapeutic value has proven effective (McManus et al. 2009). The inclusion of the metazoan animal model, C. elegans, in the genetic interaction testing pipeline can also contribute new insights as nematodes have a gene complement more akin to humans and contain several cancer-relevant genes not found in yeast, such as BRCA1, BRCA2, TP53 and the family of poly(ADP)-ribose polymerases (PARPs) (Boulton et al. 2004, Martin et al. 2005, Derry et al. 2001, Schumacher et al. 2001, Gagnon et al. 2002). Furthermore, C. elegans mutants and double mutants also present informative phenotypes, such as apoptotic defects, cell cycle checkpoint dysfunction, and chromosome loss, which can lead to a better understanding of the biological processes affected by specific genetic interactions (Kirienko, Mani, Fay 2010).  In this work SGA screens were performed on three hypomorphic cohesin mutants to identify common processes required for survival when cohesin is mutated. Interactions were tested for conservation in C. elegans using an assay for defects in somatic cell proliferation detailed in chapter 2 (McLellan et al. 2009). Proteins that mediate replication fork progression and stability are required in both S. cerevisiae and C. elegans in the presence of cohesin mutations. The screen in C. elegans was expanded to include a fork stability regulator, PARP1, that is present in higher organisms but not yeast. It was found that the PARP family of mutants genetically interact with SMC1 in C. elegans. In  65  human cells a small molecule PARP inhibitor that is currently being evaluated in clinical trials was effective in inhibiting growth in cohesin depleted cells. This demonstrated that the dependence of cohesin mutated cells on replication fork mediators is conserved across distantly related species.  3.2 3.2.1  Materials and methods S. cerevisiae strain construction and SGA screens  Temperature sensitive (ts) cohesin alleles were marked with URA3 as described by Ben aroya et al. and were used as query genes in SGA screens. Query genes were screened against the Boeke non-essential deletion collection and a collection of Damp and ts alleles representing essential genes (Schuldiner et al. 2005, Ben-Aroya et al. 2008). SGA screens were performed in biological triplicate, each with three technical replicates using a Singer RoToR. SCC1 and SCC2 were screened at 30oC while SMC1 was screened at 25oC due to slow growth at higher temperatures. Screens were performed essentially as described (Tong et al. 2004) with slight modifications detailed in Stoepel et al (manuscript in preparation). Briefly, each query strain was mated to the array and split into three biological replicates. Haploids were selected using canavanine, thialysine and minus histidine selection as in Tong et al. 2004. For the final selection step single mutants (array mutants) were selected on plates containing G418 (Invitrogen) and double mutant selection was performed on G418 containing minus uracil plates. For final data collection the three biological replicates were expanded in triplicate for a total of nine replicates and plate images were acquired on a flat bed scanner. Colony size was normalized for each plate and converted to pixel area measurements using custom software developed by the Loewen lab at UBC. Statistical analysis was done in RTM (Stoepel et al. manuscript in 66  preparation) using a script that compares the colony size measurements for each gene pair. The program output is a measure of the difference in colony size, a proxy for strain fitness, of corresponding colonies on the single and double selection plates along with a measure of statistical significance.  3.2.2  Double mutant reconstruction & random spore analysis  Double heterozygous mutants were recreated by mating each of the single mutants and selecting on –ura, +G418 plates. The cohesin mutant parent strains were the same query genes used for SGA screens and the other parent was pulled directly from the array plates. Heterozygotes were confirmed by PCR analysis to confirm the identity of the array strain. Strains used can be found in table 3.1 and 3.2. Random spore was performed at 25oC as described in Chapter 2. Briefly, spores were plated onto haploid selection plates containing canavanine and thialysine. Single selection plates were additionally either –ura or +G418. Double mutants that did not grow on the double selection plates (ura, +G418) were considered SL. Several isolates of viable double mutants were isolated from random spore analysis for growth curve analysis.  67  Table 3.1  Diploid S. cerevisiae strains used in this study  Genotype smc1-259,lpd1Δ scc1-73, lpd1Δ scc2-4, lpΔ1Δ smc1-259, trm112-Damp scc1-73, trm112-Damp scc2-4, trm112-Damp smc1-259, gim4Δ scc1-73, gim4Δ scc2-4, gim4Δ smc1-259, clb2Δ scc1-73, clb2Δ scc2-4, clb2Δ smc1-259, stu2-12 scc1-73, stu2-12 scc2-4, stu2-12 smc1-259, stu2-13 scc1-73, stu2-13 scc2-4, stu2-13 smc1-259, tub2-443 scc1-73, tub2-443 scc2-4, tub2-443 smc1-259, cdc20-2 scc1-73, cdc20-2 scc2-4, cdc20-2 smc1-259, tub4-ΔDSY scc1-73, tub4-ΔDSY scc2-4, tub4-ΔDSY smc1-259, lst8-15 scc1-73, lst8-15 scc2-4, lst8-15 smc1-259, hos1Δ scc1-73, hos1Δ scc2-4, hos1Δ smc1-259, rrp4-1 scc1-73, rrp4-1 scc2-4, rrp4-1 smc1-259, rna15-58 scc1-73, rna15-58 scc2-4, rna15-58 smc1-259, rpn11-14 scc1-73, rpn11-14 scc2-4, rpn11-14 smc1-259, bub3Δ scc1-73, bub3Δ scc2-4, bub3Δ smc1-259, ypr1Δ scc1-73, ypr1Δ scc2-4, ypr1Δ smc1-259, pac10Δ scc1-73, pac10Δ scc2-4, pac10Δ  YJM# 591 592 635 636 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 709 710 711 712 713 714 715 716 719 720 727 728 749 750 751 752 754 755 756 807 808 813 814 815 816 817 818 819 820 825 826  Genotype smc1-259, eaf3Δ scc1-73, eaf3Δ scc2-4, eaf3Δ smc1-259, arc1Δ scc1-73, arc1Δ scc2-4, arc1Δ smc1-259, sac3Δ scc1-73, sac3Δ scc2-4, sac3Δ smc1-259, pcf11-ts9 scc1-73, pcf11-ts9 scc2-4, pcf11-ts9 smc1-259, pcf11-1 scc1-73, pcf11-1 scc2-4, pcf11-1 smc1-259, rad61Δ scc1-73, rad61Δ scc2-4, rad61Δ smc1-259, bim1Δ scc1-73, bim1Δ scc2-4, bim1Δ smc1-259, rps31-Damp scc1-73, rps31-Damp scc2-4, rps31-Damp smc1-259, gim3Δ scc1-73, gim3Δ scc2-4, gim3Δ smc1-259, irc15Δ scc1-73, irc15Δ scc2-4, irc15Δ smc1-259, kar3Δ scc1-73, kar3Δ scc2-4, kar3Δ smc1-259, doc1Δ scc1-73, doc1Δ scc2-4, doc1Δ smc1-259, csm3Δ scc1-73, csm3Δ scc2-4, csm3Δ smc1-259, mdm20Δ scc1-73, mdm20Δ scc2-4, mdm20Δ smc1-259, chl1Δ scc1-73, chl1Δ scc2-4, chl1Δ smc1-259, ctf4Δ scc1-73, ctf4Δ scc2-4, ctf4Δ smc1-259, rad27Δ scc1-73, rad27Δ scc2-4, rad27Δ  YJM# 893 894 895 896 897 898 899 900 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 935 936 937 938 939 940 941 924 943 944 945 946 947 947 949 950 951 952 953 954 955 956 957 958 959 960 961  68  Genotype smc1-259, tof1Δ scc1-73, tof1Δ scc2-4, tof1Δ smc1-259, rps16BΔ scc1-73, rps16BΔ scc2-4, rps16BΔ  YJM# 855 856 863 864 879 880  Genotype smc1-259, dcc1Δ scc1-73, dcc1Δ scc2-4, dcc1Δ smc1-259, ctf8Δ scc1-73, ctf8Δ scc2-4, ctf8Δ  YJM# 962 963 964 965 966 967  All cohesin alleles (smc1-259, scc1-73, scc2-4) are marked with URA3. All other alleles listed in the table are marked with KanMX. In addition to the genotype listed strains are Mat a/alpha ura3Δ0/ura3Δ0 leu2Δ0/leu2Δ0 his3Δ1/his3Δ1 met15Δ0/ MET15 or met15Δ0 LYS2/LYS2 can1Δ::STE2pr_pombeHIS5/CAN1 LYP1/lyp1Δ. YJM: Yeast Jessica McLellan; an identifier corresponding to the yeast database of JMcLellan.  69  Table 3.2  Haploid S. cerevisiae strains used in this study  Genotype lpd1Δ scc1-73, lpd1Δ scc2-4, lpΔ1Δ trm112-Damp smc1-259, trm112-Damp scc1-73, trm112-Damp scc2-4, trm112-Damp gim4Δ smc1-259, gim4Δ scc1-73, gim4Δ scc2-4, gim4Δ clb2Δ smc1-259, clb2Δ scc1-73, clb2Δ scc2-4, clb2Δ stu2-12 smc1-259, stu2-12 scc1-73, stu2-12 scc2-4, stu2-12 stu2-13 smc1-259, stu2-13 scc1-73, stu2-13 tub2-443 smc1-259, tub2-443 scc1-73, tub2-443 scc2-4, tub2-443 cdc20-2 smc1-259, cdc20-2 scc1-73, cdc20-2 scc2-4, cdc20-2 tub4-ΔDSY smc1-259, tub4-ΔDSY scc1-73, tub4-ΔDSY scc2-4, tub4-ΔDSY lst8-15 smc1-259, lst8-15 scc1-73, lst8-15 scc2-4, lst8-15 hos1Δ smc1-259, hos1Δ scc1-73, hos1Δ scc2-4, hos1Δ rrp4-1 smc1-259, rrp4-1 scc1-73, rrp4-1 scc2-4, rrp4-1 rna15-58 smc1-259, rna15-58 scc1-73, rna15-58 scc2-4, rna15-58 rpn11-14  YJM# 589 593 595 597 599 601 603 605 607 609 611 613 615 617 619 621 623 625 627 629 631 633 637 639 641 643 645 647 649 651 653 655 657 659 661 663 665 667 669 671 673 675 693 695 697 699 701 703 705 707 717  Genotype smc1-259, pac10Δ scc1-73, pac10Δ scc2-4, pac10Δ rps16BΔ smc1-259, rps16BΔ scc1-73, rps16BΔ scc2-4, rps16BΔ eaf3Δ smc1-259, eaf3Δ scc1-73, eaf3Δ scc2-4, eaf3Δ arc1Δ smc1-259, arc1Δ scc1-73, arc1Δ scc2-4, arc1Δ sac3Δ smc1-259, sac3Δ scc1-73, sac3Δ scc2-4, sac3Δ pcf11-ts9 smc1-259, pcf11-ts9 scc1-73, pcf11-ts9 scc2-4, pcf11-ts9 pcf11-1 smc1-259, pcf11-1 scc1-73, pcf11-1 scc2-4, pcf11-1 rad61Δ scc2-4, rad61Δ bim1Δ smc1-259, bim1Δ scc2-4, bim1Δ rps31-Δamp smc1-259, rps31-Damp scc1-73, rps31-Damp scc2-4, rps31-Damp gim3Δ smc1-259, gim3Δ scc1-73, gim3Δ scc2-4, gim3Δ irc15Δ smc1-259, irc15Δ scc1-73, irc15Δ scc2-4, irc15Δ kar3Δ scc1-73, kar3Δ scc2-4, kar3Δ doc1Δ scc1-73, doc1Δ scc2-4, doc1Δ mdm20Δ  YJM# 743 745 747 757 759 761 763 766 767 770 772 773 775 777 779 781 783 785 787 789 791 793 795 797 799 801 803 805 811 821 823 827 829 831 833 835 837 839 841 843 845 847 849 851 853 857 859 861 865 867 877  70  Genotype scc1-73, rpn11-14 scc2-4, rpn11-14 bub3Δ scc1-73, bub3Δ scc2-4, bub3Δ ypr1Δ smc1-259, ypr1Δ scc1-73, ypr1Δ scc2-4, ypr1Δ pac10Δ *Mat alpha smc1-259  YJM# 721 723 725 729 731 733 735 737 739 741 561  Genotype scc1-73, mdm20Δ rad27Δ smc1-259, rad27Δ scc1-73, rad27Δ scc2-4, rad27Δ smc1-259 scc1-73 scc2-4 *Mat alpha scc1-73 *Mat alpha scc2-4  YJM# 881 901 903 905 907 925 927 929 931 219 212  All cohesin alleles (smc1-259, scc1-73, scc2-4) are marked with URA3. All other alleles listed in the table are marked with KanMX. In addition to the genotype listed strains are Mat a ura3Δ0 leu2Δ0 his3Δ1 MET15 LYS2 can1Δ::STE2pr_pombeHIS5 lyp1Δ *SGA query strains. These strains are identical in genotype to the other strains in this table except that they are Mat alpha met15Δ0 or MET15 3.2.3  S. cerevisiae growth curves  All viable double mutants were analyzed by growth curve analysis as described in Chapter 2 with slight modifications. Strains were grown overnight in YPD, diluted to an optical density600nm (OD) of 0.2 and allowed to grow for 4 hours while shaking at 25oC. Strains were then diluted to an OD of 0.3 and 100ul was added to each well of a 96 well plate. 100ul of fresh YPD was added to each well for a final OD of 0.15. For each plate, fifteen replicates were performed for wild type, while three replicates were performed for each of the other strains analyzed. OD readings were made every 30min, after 3 min of shaking, over a period of 24 hours at 26oC (5 plates) or 30oC (6 plates) on a Tecan M1000.  3.2.4  Growth curve data analysis  The goal of the growth curve analysis of the viable S. cerevisiae double mutants was to identify synthetic sick (SS) interactions between three cohesin query genes (scc1-73, smc1-259, scc2-4), and 28 of the 33 conserved genes listed in Table 3.3 (the remaining 5 71  were SL with all three cohesin query genes). Yeast strains were grown for 24 hours on 11 different 96 well plates, each run on a separate day. Of these 11 plates, five were grown at 26oC, and six at 30oC. On each plate there were fifteen replicate wells for the wild type strain and three replicate wells for each of the other strains analyzed. 3.2.4.1  Estimation of strain fitness  Strain fitness is often defined with implicit reference to exponential growth:  y (t ) = y0 e rt where y(t) is the population size at time t, y0 is the initial population at time t=0, and r is the exponential growth constant. In reality, unconstrained growth does not exist and it is common to employ a related model, the S-shaped function known as logistic growth:  y (t ) = A +  B− A 1 + exp[−r (t − tmid )]  where y(t) and r are as defined before. B is the carrying capacity or upper asymptote, and A is the lower asymptote. tmid is the time at which population size y is halfway between A and B. Successfully employed approaches based on these models have been used in the past (Baetz et al. 2004, McLellan et al. 2009); however, in this work, the heterogeneity of the growth curve shapes prohibited the use of either model. Therefore a non-parametric method in which the strain fitness was designated as the logarithm of the total area under the curve (AUC) was used. This measure is still well-justified in either of the above models and thus represents an attractive strategy that is achievable in practice and compatible with relevant growth models. High values of log(AUC) are associated with large exponential growth constants and hence high strain fitness, and vice versa.  72  The AUC was calculated for each growth curve through use of Simpson’s rule in RTM (Pinheiro et al. 2008, R Development Core Team: R Foundation for Statistical Computing 2008). After fitness estimates for each growth curve were obtained and prior to identifying SS interactions, it was necessary to address the effect of the difference in temperatures and to normalize for plate (or, equivalently, day) effects. Because the temperature difference had profound effects on the shape of the growth curves, separate analyses were done for plates grown at 26oC and 30oC. Within each temperature-specific analysis, the high level of replication for wild type was exploited to estimate plate effects. All fitness estimates were then adjusted relative to a reference plate (10/07/13 for 26oC, 10/06/29 for 30oC).  3.2.4.2  Estimation of interaction effects  A derived variable analysis was performed on normalized fitness estimates with a linear model of the form:  Fq , g = FWT + τ q + τ g + τ q , g Fq , g  is the fitness for a double-knockout of query gene q and non-essential gene g, FWT  is the fitness estimate for wild type growth curves,  τq  is the single-knockout effect for  each of the three query genes (scc1-73, smc1-259, scc2-4),  τg  effect for each of the other non-essential genes assessed, and  is the single-knockout  τ q,g  is the double-knockout  interaction effect associated with genes q and g. After fitting this model, the significance  73  of the interaction effects  τ q,g  was assessed by comparing these estimates to those  obtained under the assumption of additive neutrality:  Fqneut , g = FWT + τ q + τ g Hence, values of  τ q,g > 0  τ q,g = 0  indicate SS interactions, while values of  τ q,g < 0  indicate  alleviating interactions.  3.2.4.3  Determination of statistical significance  To test for the statistical significance of each interaction, T-statistics were calculated as:  Tq , g = τ q , g / SE (τ q , g ) The T-statistics are the interaction effects normalized according to the standard error of that particular estimate. Bonferroni-corrected p-values were used to control for the family wise error rate (FWER), set at 0.05.  3.2.5  C. elegans genetic interactions & SYTO12 staining  The Pvl somatic cell proliferation assay was performed as described in chapter 2. SYTO12 staining was performed on young adult worms (that had been pre-treated with RNAi by feeding) by bleaching gravid adults onto RNAi plates. Worms were stained and protected from light for 3 hours, then destained on fresh RNAi plates. Images were captured immediately after destaining on a Zeiss Axioplan 2 microscope using a 40x lens.  74  3.2.6  C. elegans double mutant strain construction & brood analysis  See Table 3.3 for strains used. him-1(e879) males were mated to either pme-1(ok988), pme-2(ok344), pme-2(tm3401), pme-3(gk120), pme-4(ok980) or pme-5(ok446) hermaphrodites. Homozygous him-1 mutants were followed by the high frequency of males (~10%) and all other mutations were followed by PCR. pme-1; him-1 animals were balanced by hT2 and homozygous him-1; pme-2(ok344) animals were balanced by mIn1.  Table 3.3  C. elegans strains  Strain Name CB879 RB1042 VC1171 FX3401 VC130 VC641 RB684 PH001 PH002 PH003 PH004 PH005 PH006  3.2.7  Genotype him-1(e879) I pme-1(ok988) I pme-2(ok344) II pme-2(tm3401) II pme-3(gk120) IV pme-4(ok980) IV pme-5(ok446) V pme-1(ok988) I ; him-1(e879) I him-1(e879) I ; pme-2(ok344) II him-1(e879) I ; pme-2(tm3401) II him-1(e879) I ; pme-3(gk120) IV him-1(e879) I ; pme-4(ok980) IV him-1(e879) I; pme-5(ok446)  HCT116 cell culture & siRNA  Cells were cultured as in Mcmanus et al. 2009. siRNAs and transfection reagent were purchased from Dharmacon, olaparib was purchased from Selleck and benzamide was purchased from Aldrich. Cells were transfected 24 hours after seeding and were transferred after an additional 24 hours and expanded for three to 10 days depending on the assay. For the clonogenic survival assays 1000 and 500 cells were seeded per well in 24 well dishes and 10 cm plates, respectively. Cells were allowed to attach for 10 hours and then olaparib was added at the indicated concentration. Cells were fixed with 4% paraformaldehyde (PFA) in PBS and stained using 0.1% crystal violet in 95% ethanol  75  after seven and 10 days for the 24 well plat and 10 cm dishes, respectively. Drug exposure was continuous during this time and the 10cm dishes were supplemented with fresh media containing 0.6uM olaparib at day five. High content digital imaging microscopy (HC-DIM) was used to count cell numbers after plating 1000 siRNA transfected cells per well in 96 well plates. Cells were allowed to settle for 10 hours and then either Benzamide or Olaparib was added at the indicated concentration. Cells were fixed with PFA after 3 days of drug exposure and counter stained with Hoechst. Plates were subjected to HC-DIM using a Cellomics ArrayScan with a 20x (81 images/well) or 10x (16 images/well) dry lens. The total number of Hoescht-positive nuclei was determined using a Cellomics Target Activation algorithm and normalized to each siRNA treatment (GAPDH, BRCA1 or SMC1). Western blots were performed on asynchronous, sub-confluent cells harvested 3 days post-transfection. Antibodies were purchased from Millipore (BRCA1 07-434) and Abcam (SMC1 ab9262, alpha tubulin ab56476).  3.3 3.3.1  Results S. cerevisiae cohesin genetic interactions  Synthetic genetic array (SGA) technology was used to screen temperature sensitive alleles of two cohesin components (smc1-259, scc1-73) and one cohesin loader (scc2-4) against ~95% of genes in S. cerevisiae. All three alleles have mutations in similar regions as those identified in colon tumours (SMC1, SCC2) or in COSMIC (SCC1) (Table 3.4, Figure 3.1).  76  Table 3.4  Mutations in cohesin genes seen in tumours  Human Gene (Yeast Gene) SMC1L1 (SMC1)  CSPG6 (SMC3) RAD21 (SCC1) STAG3 (SCC3) NIPBL (SCC2)  Mutations reported in CIN colon tumours F369L (1186C>T) R434W (1300C>T) I560M (1680 C>G) I1186V (3556G>A) R879X (2635C>T)  I795T (24117T>C) R479STOP (1435C>T) Frameshift after aa 992 Q554STOP (1660C>T) M1793K (5378T>A)  Cancer mutations reported in COSMIC*; cancer type  A44V (131C>T); CNS Q474stop (1420C>T); lung E498K (1492G>A); skin F660L (1980C>A); lung E1674K (4939G>A); breast S349C (1045A>T); lung Deletion, Frameshift after aa 2241; lung  *Silent mutations not reported in this table. Amino acid: aa  77  Figure 3.1  Cohesin mutations found in colon tumours  Comparison of SMC1 orthologs in H. sapiens, C. elegans, and S. cerevisiae. Mutations identified in colon tumours are indicated on the human gene and protein domains are shown on the S. cerevisiae gene. The number of amino acids (aa) are shown on the right hand side. A) SMC1 mutations B) SCC2 mutations. C) SCC1 mutations. D) Schematic of cohesin and loaders (adapted from Nasmyth & Haering, 2009).  The SGA interaction data was sequentially filtered using several criteria to increase its quality and focus. Only negative genetic interactions were considered that had  78  statistically significant p-values less than 0.05 and that had relatively large interaction magnitudes (E-C value less than -0.3, see Methods). Filtering based on magnitude enriched for interactions that caused a severe fitness defect when compared to the single mutant. Because one goal of this work was to identify second site mutations that cause sickness or lethality in the context of a cohesin mutation that might be suitable for cancer treatment, the fitness differential between single and double mutants was a useful metric. In addition, because mutations are found in different cohesin genes in tumours, it was important to identify genes that interacted with several of these cohesion genes in order to identify broader spectrum targets. This would increase the mutational scope of cancers that could be treated with a single therapeutic rather than identifying targets that are specific for a relatively small subset of cancers with somatic mutations in a particular gene. For this reason, and to reduce false positives, genes that interacted with only a single cohesion query gene were eliminated from further analysis.  Finally, genes that were not conserved in humans were eliminated to focus on interactions that may have implications in the biology of cohesins and cancer. Using these criteria, 55 genes interacted with more than one cohesin query gene, defining 90 putative negative genetic interactions. Of these, 39 (71%) had an identifiable ortholog in humans (Table 3.5). 6 genes were removed from further analysis for technical reasons and this left 33 genes comprising 78 genetic interactions. This filtering process is graphically depicted in Figure 3.2 A and the SGA filtered network is found in Figure 3.2 B.  79  Table 3.5  S. cerevisiae, C. elegans and human orthologs  S. cerevisiae Gene (ORF name)  C. elegans gene (ORF name, e-value)  H. sapiens gene (e-value)  ARC1 (YGL105W) BIM1 (YER016W)¶  mars-1 (F58B3.5, 4E-36) ebp-1 (Y59A8B.7, 1.00E-26) ebp-2 (VW02B12L.3, 3.00E21) No human, C. elegans match bub-3 (Y54G9A.6, 2.00E-28) fzy-1 (ZK177.6, 8.00E-31) M03C11.2 (9.00E-92) No human, C. elegans match No human, C. elegans match cyb-3 (T06E6.2, 4.00E-26) F23C8.9 (5.00E-06) No human, C. elegans match No human, C. elegans match F17C11.10 (1E-53) T22C1.4 (3E-09) K09H9.2 (2E-24) apc-10 (F15H10.3, 3.00E-08) mrg-1 (Y37D8A.9, 1.00E-08) pfd-4 (B0035.4, 2.00E-05) pfd-2 (H20J04.5, 2.00E-05) No human, C. elegans match hda-1 (C53A5.3, 2.00E-43) hsf-1 (Y53C10A.12, 3.00E-19) No human, C. elegans match LLC1.3 (9.00E-73) klp-17 (W02B12.7, 9.00E-53) xrn-1 (Y39G8C.1, 4.60E-174) LLC1.3 (9.00E-160) C10H11.8 (3.00E-30) No human, C. elegans match No human, C. elegans match No human, C. elegans match  Aminoacyl tRNA synthase interacting protein (2.00E-39) APC-binding protein EB1 (7.00E-37)  RNAi construct available? Y Y  BUB3 (9.00E-30) CDC20 (3.00E-69) DDX11 (2.00E-121)  Y Y Y  Cyclin B2 (5.00E-55) TIMELESS interacting protein (8.00E-07)  Y Y  WDHD1 (4.00E-18) CHTF8 (5E-10) DCC1 (6.00E-08) APC10 (5.00E-22) Mortality factor 4-like protein 1 (7.00E-33) PFD4 (8.00E-14) PFD2 (4.00E-14)  Y Y Y Y Y Y Y  HD1 (3.00E-44) HSF1 (3.00E-30)  Y Y  Dihydrolipoyl dehydrogenase (2.00E-71) Kinesin family member C1 (2.00E-84) 5'-3' exoribonuclease 1 (0.00E+00) Dihydrolipoyl dehydrogenase (4.00E-157) LST8 (4.00E-82)  Y Y Y Y Y  BNA2 (YJR078W) BUB3 (YOR026W) CDC20 (YGL116W) CHL1 (YPL008W) CHL4 (YDR254W) CIK1 (YMR198W) CLB2 (YPR119W) CSM3 (YMR048W) CTF19 (YPL018W) CTF3 (YLR381W) CTF4 (YPR135W)§ CTF8 (YHR191C)§ DCC1 (YCL016C)§ DOC1 (YGL240W) EAF3 (YPR023C) GIM3 (YNL153C) GIM4 (YEL003W) GPI15 (YNL038W) HOS1 (YPR068C) HSF1 (YGL073W) IML3 (YBR107C) IRC15 (YPL017C) KAR3 (YPR141C) KEM1 (YGL173C) LPD1 (YFL018C) LST8 (YNL006W) MCM16 (YPR046W) MCM21 (YDR318W) MCM22 (YJR135C)  80  S. cerevisiae Gene (ORF name)  C. elegans gene (ORF name, e-value)  H. sapiens gene (e-value)  MDM20 (YOL076W)* MRC1 (YCL061C) PAC10 (YGR078C) PCF11 (YDR228C) PML39 (YML107C) RAD27 (YKL113C) RAD61 (YDR014W)§ RNA15 (YGL044C) RPN11 (YFR004W)  cra-1 (R13F6.10, 0.14) No human, C. elegans match pfd-3 (T06G6.9, 4.00E-22) pcf-11 (R144.2, 3.00E-10) No human, C. elegans match crn-1 (Y47G6A.8, 1.00E-112) wpl-1 (R08C7.10, 1.00E-53) cpf-2 (F56A8.6, 3.00E-17) rpn-11 (K07D4.3, 2.00E-110)  N(alpha)-acetyltransferase 25 (1.00E-51)  RNAi construct available? Y  PFD3 (4.00E-29) PCF11 (5.00E-16)  Y Y Y Y Y Y  RPS16B (YDL083C) RPS20 (YHL015W) RPS31 9YLR167W) RRP4 (YHR069C) SAC3 (YDR159W) SPT10 (YJL127C) STU1 (YBL034C) STU2 (YLR045C) SWI6 (YLR182W) TOF1 (YNL273W)§ TRM112 (YNR046W) TUB2 (YFL037W) TUB4 (YLR212C) YNL171C YPR1 (YDR368W)  rps-16 (T01C3.6, 2.00E-55) rps-20 (Y105E8A.16, 3.00E-31) ubl-1 (H06I04.4, 7.00E-39) exos-2 (Y73B6BL.3, 3.00E-43) F20D12.2 (1.00E-26) No human, C. elegans match No human, C. elegans match zyg-9 (F22B5.7, 4.00E-13) No human, C. elegans match tim-1 (Y75B8A.22, 4.00E-66) C04H5.1 (6.00E-10) tbb-4 (B0272.1, 0.00E+00) tbg-1 (F58A4.8, 9.00E-55) dubious ORF in S. cerevisiae Y39G8B.1 (4.00E-49)  FEN1 (4.00E-118) Wings apart-like protein homolog (7.9) cleavage stimulation factor subunit 2 (6.00E-17) 26S proteasome non-ATPase regulatory subunit 14 (1.00E-115) 40S ribosomal protein S16 (9.00E-51) 40S ribosomal protein S20 (4.00E-32) ubiquitin-40S ribosomal protein S27a precursor (7.00E-59) RRP4 (2.00E-56) MCM3-associated protein (7.00E-33)  cytoskeleton-associated protein 5 (3.00E-32)  Y  timeless homolog (2.00E-09) tRNA methyltransferase 112 homolog (4.00E-12) Beta tubulin (0.00E+00) Gamma tubulin (7.00E-98)  Y Y Y Y  aldo-keto reductase family 1, member A1 (9.00E-56)  Y  Y Y N N Y  ¶  No RNAi constrauct was available for ebp-1. Experiments were performed with RNAi against the second best BLAST match, ebp-2. *No direct S. cerevisiae to human match. Used C. elegans gene sequence to find a human homolog. Reported e-value for the human gene is from a BLAST search from C. elegans against the human database. § No direct S. cerevisiae to C. elegans match. Used the human gene sequence to find a C. elegans homolog. Reported e-value for the C. elegans gene is from a BLAST search from human against the C. elegans database.  81  Figure 3.2  A.  S. cerevisiae SGA network  i  828 SCC1 72  44  ii  148  84  110  SCC2  SCC1  13  14  SMC1 246  20  13  9  BUB3  Conserved in humans  SWI6  12  SCC2  434  7  HSF1  RAD27  PAC10  28  2  SMC1 0  94 Validated by RS, GC 33  RPN11  TOF1  CHL4  YPR1  YNL171C PML39  GIM3  DOC1  IML3  DCC1 CTF8  CTF19  SCC1  RPS20  SCC2  CTF4  CTF3  CSM3  MCM21 MCM22  CHL1  PCF11  SPT10  RRP4  LPD1  TUB2  BIM1 MDM20  IRC15  BNA2 RPS16B EAF3  SAC3  GIM4 CLB2  RAD61  KAR3  HOS1  TRM112  GPI15  CIK1 MRC1  TUB4  KEM1  RPS31  MCM16 LST8  CDC20  SCC2  STU1  Essential in S. cerevisiae  STU2  0  2  0  90 (78) 2/3 query genes, Conserved 39 (33)  Not conserved  RNA15  1 SCC1  SMC1  E-C < -.3 364  B.  iv  7  SCC1  SCC2  SMC1 191  1692 Negative interaction, p value < .05 1435  iii  13  12  ARC1  SMC1  A) Venn diagrams depicting how SGA data was filtered. i. Interactions that had a negative interaction value and were statistically significant (p-value < 0.05). ii. Interactions that had a relatively weak interaction score (Experimental value – Control value > -.3) were eliminated to enrich for biologically significant interactions. iii. Genes were eliminated if they failed to interact with > 2 of the cohesin query genes and if they did not have a human ortholog (Table 3.5). iv. Summary of the final network after 82 Supplementary Figure 2  random spore and growth curve retesting and validation. The total number of interactions in a given Venn diagram is underlined. The total number of genes is shown in red. Numbers in brackets indicate remaining interactions and genes after removal of 6 genes for technical reasons. B) The network represented in A iii. Green circles indicate genes conserved in humans and purple circles represent genes with no identifiable sequence orthologs. Circles outlined in black represent essential S. cerevisiae genes.  99 double mutants (33 by 3 cohesin query genes) were reconstructed and retested for a genetic interaction by random spore analysis. Fourteen double mutants were inviable and had a SL interaction. To assess whether growth defects were greater than additive, growth curve analysis was performed on all viable double mutants. Growth curves can be found in Figure 3.3. Gene interactions were analyzed by defining a mutant’s fitness as the area under the growth curve (AUC). Each single and double mutant was assigned an interaction estimate. Double mutants that had a fitness defect that deviated from that predicted under a model of interaction neutrality were considered genetic interactions. See methods for additional details. The fitness of each strain analyzed is graphically presented in Figures 3.4 and 3.5 and the t-statistics are graphically represented in Figures 3.6 and 3.7. Tables 3.6 and 3.7 list the interaction magnitudes, p-values, and t statistics. A summary of interactions identified by growth curve analysis is presented in Table 3.8. Random spore and growth curve analyses achieved several goals: 1) Reduced the false positive rate by removing genes with an incorrect well address in the high throughput arrays (8%); 2) Eliminated condition artifacts by ensuring that genetic interactions were reproducible under the same drug selective conditions as SGA (random spore) and in rich medium (growth curve); 3) Yielded an additional quantitative measure of each SS genetic interaction; and 4) Identified additional true positives not identified under SGA conditions. 83  During random spore and growth curve analysis of double mutants 4 (5%) out of 78 interactions identified by SGA did not result in a negative genetic interaction and an additional 20 interactions not observed by SGA were identified for a total of 94 negative genetic interactions with the three cohesin query genes. The final, validated network can be found in Figure 3.6. 31% of these interactions involve essential genes represented by either temperature sensitive (ts) or decreased abundance by mRNA perturbation (Damp) alleles, highlighting the importance of screening against essential gene collections.  84  Figure 3.3  S. cerevisiae growth curves  Growth curves (26oC)  A) Phenotype (OD600nm)  smc1  smc1 bim16  1.5  smc1 bub36  1.0  smc1 cdc20−2  smc1 doc16  SL  0.5 scc1  scc1 bim16  1.5 1.0  smc1 eaf36  smc1 gim36  smc1 gim46  smc1 hos16  smc1 irc156  smc1 kar36  SL  scc1 bub36  scc1 cdc20−2  scc1 doc16  SL scc1 eaf36  scc1 gim36  scc1 gim46  scc1 hos16  scc1 irc156  smc1 lpd16  smc1 lst8−15  smc1 mdm206  SL  scc1 kar36  smc1 pac106  smc1 pcf11−1  smc1 pcf11−ts9  smc1 rad276  smc1 rad616  SL  scc1 lpd16  scc1 lst8−15  scc1 mdm206  scc1 pac106  scc1 pcf11−1  scc1 pcf11−ts9  scc1 rad276  SL  scc1 rad616  scc1 rna15−58  scc1 rpn11−1  smc1 rpn11−14  smc1 rps16B6  smc1 rps316  smc1 rrp4−1  smc1 sac36  smc1 stu2−12  smc1 stu2−13  smc1 trm112−Damp  smc1 smc1 tub2−443 tub4−6DSY  smc1 ypr16  bub36  cdc20−2  doc16  eaf36  gim36  gim46  hos16  irc156  kar36  lpd16  lst8−15  mdm206  pac106  pcf11−1  pcf11−ts9  rad276  1.5 1.0  SL  0.5  scc1 rpn11−14  scc1 rps16B6  scc1 rps316  scc1 rrp4−1  scc1 sac36  scc1 stu2−12  scc1 stu2−13  scc1 trm112−Damp  scc1 scc1 tub2−443 tub4−6DSY  scc1 ypr16  1.5 1.0  SL  bim16  1.5  smc1 rpn11−1  SL  SL  0.5  smc1 rna15−58  0.5  rad616  rna15−58  rpn11−1  rpn11−14  rps16B6  rps316  rrp4−1  sac36  stu2−12  stu2−13  trm112−Damp  tub2−443 tub4−6DSY  ypr16  1.5 1.0  1.0  0.5 00 00 0 80  40 0  00 0  00 0  40  0  80  00  00  40 0  0  80 0  0  40 00 0 80 00 0  0  00 0  00 0  40  40 00 0 80 00 0  0  80  0  40 00 0 80 00 0  40 00  80  80  0  0 00 0  0  40 00 0 80 00 0  0  40 00 0 00 0  00 0 80 00 0  0  40  40 0  80  0  00  0  40 00  40 00 0 80 00 0  40 00 0 80 00 0  0  80 0  0  0 00 0  0  40 00 0 80 00 0  0  00  0.5  Time (seconds)  Growth curves (30oC)  B) scc2 bim16  scc2 bub36  scc2 cdc20−2  scc2 doc16  scc2 eaf36  scc2 gim36  scc2 gim46  scc2 hos16  scc2 irc156  scc2 kar36  scc2 lpd16  scc2 lst8−15  1.0  scc2 mdm206  scc2 pac106  scc2 pcf11−1  scc2 pcf11−ts9  scc2 rad276  scc2 rad616  scc2 rna15−58  scc2 rpn11−1  scc2 rpn11−14  scc2 rps16B6  scc2 rps316  scc2 rrp4−1  scc2 sac36  scc2 stu2−12  SL  0.5 scc1  scc1 bim16  1.5  scc1 bub36  scc1 cdc20−2  scc1 doc16  scc1 eaf36  scc1 gim36  scc1 gim46  scc1 hos16  scc1 irc156  scc1 kar36  scc1 lpd16  scc1 lst8−15  scc1 mdm206  scc1 pac106  scc1 pcf11−1  scc1 pcf11−ts9  scc1 rad276  scc1 rad616  scc2 trm112−Damp  scc2 scc2 tub2−443 tub4−6DSY  scc2 ypr16  scc1 rna15−58  scc1 rpn11−1  scc1 rpn11−14  scc1 rps16B6  scc1 rps316  scc1 rrp4−1  scc1 sac36  scc1 stu2−12  scc1 stu2−13  bim16  bub36  cdc20−2  doc16  eaf36  gim36  gim46  hos16  irc156  kar36  lpd16  lst8−15  mdm206  pac106  pcf11−1  pcf11−ts9  rad276  1.5 1.0 0.5  scc1 trm112−Damp  scc1 scc1 tub2−443 tub4−6DSY  scc1 ypr16  1.5 1.0  SL  0.5  1.5  scc2 stu2−13  SL  SL  1.0  0.5  rad616  rna15−58  rpn11−1  rpn11−14  rps16B6  rps316  rrp4−1  sac36  stu2−12  stu2−13  trm112−Damp  tub2−443 tub4−6DSY  ypr16  1.5  0  0  00  00  40  0  80  0  0  00  00  40  0  80  0  0  00  00  40  0  80  0  0  00  00  40  0  80  0  40 00 0 80 00 0  0  40 00 0 80 00 0  0  0  00  00  80  0  40  0  0  00  00  80  0  40  0  0  00  00  80  0  40  0  40 00 0 80 00 0  0  0  00  00  80  0  40  0  0  00  00  40  0  80  0  0  00  00  40  0  80  40 00 0 00 0  0  80  0 00  00  80  0  40  00  00  40  0  0  0.5 0  1.0  0.5 0  1.0  80  Phenotype (OD600nm)  scc2  1.5  Time (seconds) ‘Gene A’ Mutation  ‘Gene B’ Mutation  SL: Lethal by Random spore. No Growth curve analysis performed  Interaction found with alternate allele. Second allele not analyzed Strain could not be isolated due to high frequency of suppressors on random spore plates Both scc1 and smc1 are SL. No growth curve performed at 26oC  85  Growth curve replicates at A) 26oC and B) 30oC. scc2-4 growth curves were only run at 30oC and smc1-259 curves were only assayed at 26oC. scc1-73 curves were run at both temperatures because unlike the other two alleles, scc1-73 shows a phenotype at both temperatures. In most cases if an interaction with scc1-73 was present, it was more pronounced at 30oC. Some interactions were tested with multiple alleles of the same gene. If an interaction was identified the second allele was not always assayed (denoted by black circles). Double mutants that were SL according to random spore could not be analyzed by growth curve analysis and are marked with ‘SL’. Gene A mutations refer to cohesin alleles and gene B mutations refer to genes identified in the SGA screens.  86  Figure 3.4  Strain fitness at 26oC ranked by interaction magnitude.  rSQï , scc1 LUF 6, scc1 WXEï6DSY, scc1 mdm20 6, scc1 pac10 6, scc1 VWXï , scc1 bub3 6, scc1 OVWï , smc1 SFIïWV , scc1 SFIï , scc1 LUF 6, smc1 gim3 6, smc1 rrSï , scc1 WXEï , smc1 rrSï , smc1 hos1 6, smc1 sac3 6, smc1 trPï'DPS, smc1 sac3 6, scc1 hos1 6, scc1 VWXï , scc1 SFIïWV , smc1 rQDï , smc1 pac10 6, smc1 VWXï , smc1 gim4 6, smc1 eaf3 6, scc1 WXEï , scc1 FGFï , scc1 rQDï , scc1 rad27 6, scc1 rad27 6, smc1 trPï'DPS, scc1 rps31 6, scc1 rps16B 6, scc1 rps16B 6, smc1 VWXï , smc1 FGFï , smc1 rps31 6, smc1 ypr1 6, scc1 gim4 6, scc1 kar3 6, scc1 OVWï , scc1 eaf3 6, smc1 lpd1 6, scc1 SFIï , smc1 doc1 6, scc1 ypr1 6, smc1 WXEï6DSY, smc1 gim3 6, scc1 bim1 6, smc1  −1.0  −0.5  0.0  Strain Fitness  0.5  WildType GeneA GeneB Neutral Interaction  Each growth curve was assigned an individual estimate of strain fitness reflecting the area under the curve (AUC) and these were averaged for each strain. A neutral strain fitness estimate was computed for each interaction and represented the theoretical strain fitness of the double mutant under conditions of an additive, non-synergistic genetic interaction. Supplementary Figure 4 Synergistic interactions occur when the experimental double mutant strain fitness deviates from the neutral estimate. Interactions were ranked according to the difference between the experimental and neutral strain fitness estimates with stronger negative interactions occurring at the top of the figure. 87  Figure 3.5  Strain fitness at 30oC ranked by interaction magnitude.  FGFï , scc1 rSQï , scc2 kar3 6, scc2 LUF 6, scc2 doc1 6, scc2 rSQï , scc1 LUF 6, scc1 VWXï , scc1 VWXï , scc2 WXEï6'6Y, scc1 mdm20 6, scc1 bub3 6, scc2 bim1 6, scc2 sac3 6, scc1 SFIïWV , scc2 WXEï6'6Y, scc2 gim3 6, scc2 SFIïWV , scc1 SFIï , scc1 trPï'DPS, scc1 VWXï , scc1 rrSï , scc1 pac10 6, scc2 pac10 6, scc1 rQDï , scc1 rad27 6, scc2 hos1 6, scc1 sac3 6, scc2 rQDï , scc2 rad27 6, scc1 gim4 6, scc1 eaf3 6, scc1 lpd1 6, scc2 rps16B 6, scc1 rps31 6, scc1 OVWï , scc2 bub3 6, scc1 rrSï , scc2 FGFï , scc2 hos1 6, scc2 rad61 6, scc2 WXEï , scc1 rps16B 6, scc2 WXEï , scc2 gim4 6, scc2 rps31 6, scc2 ypr1 6, scc1 trPï'DPS, scc2 lpd1 6, scc1 eaf3 6, scc2 SFIï , scc2 doc1 6, scc1 kar3 6, scc1 OVWï , scc1 gim3 6, scc1  ï  ï  ï  ï  Strain Fitness (30oC)    WildType GeneA GeneB Neutral Interaction  Each growth curve was assigned an individual estimate of strain fitness reflecting the area under the curve (AUC) and these were averaged for each strain. A neutral strain fitness estimate was computed for each interaction and represented the theoretical strain fitness Supplementary Figure 5 of the double mutant under conditions of an additive, non-synergistic genetic interaction. Synergistic interactions occur when the experimental double mutant strain fitness deviates from the neutral estimate. Interactions were ranked according to the difference between the experimental and neutral strain fitness estimates with stronger negative interactions occurring at the top of the figure.  88  Figure 3.6  Interaction T-statistics (26oC)  bim1 6, smc1 gim3 6, scc1 WXEï6DSY, smc1 ypr1 6, smc1 doc1 6, scc1 SFIï , smc1 lpd1 6, scc1 eaf3 6, smc1 OVWï , scc1 kar3 6, scc1 JLP 6, scc1 ypr1 6, scc1 FGFï , smc1 rps31 6, smc1 VWXï , smc1 rps16B 6, smc1 trPï'DPS, scc1 rad27 6, smc1 rps16B 6, scc1 rad27 6, scc1 rps31 6, scc1 FGFï , scc1 rQDï , scc1 JLP 6, smc1 VWXï , smc1 WXEï , scc1 eaf3 6, scc1 pac10 6, smc1 rQDï , smc1 hos1 6, scc1 SFIïWV , smc1 VWXï , scc1 sac3 6, scc1 trPï'DPS, smc1 sac3 6, smc1 hos1 6, smc1 rrSï , smc1 rrSï , scc1 gim3 6, smc1 irc15 6, smc1 WXEï , smc1 SFIï , scc1 OVWï , smc1 SFIïWV , scc1 bub3 6, scc1 VWXï , scc1 mdm20 6, scc1 pac10 6, scc1 irc15 6, scc1 WXEï6DSY, scc1 rSQï , scc1  Bonferroni Corrected p-value = .05 p-value = .05  −15  −10  −5  0  5  Interaction T-Statistic (26oC) Supplementary Figure 6  T-statistics are ranked according to magnitude. Dotted lines indicate a p-value cut off of .05, and dashed lines indicate a Bonferroni corrected p-value of .05. 89  Figure 3.7  Interaction T-statistics (30oC)  gim3 6, scc1 OVWï , scc1 kar3 6, scc1 doc1 6, scc1 SFIï , scc2 eaf3 6, scc2 ypr1 6, scc1 lpd1 6, scc1 trPï'DPS, scc2 rps31 6, scc2 JLP 6, scc2 WXEï , scc2 rps16B 6, scc2 rad61 6, scc2 WXEï , scc1 hos1 6, scc2 rrSï , scc2 FGFï , scc2 rps31 6, scc1 bub3 6, scc1 OVWï , scc2 rps16B 6, scc1 rad27 6, scc1 lpd1 6, scc2 eaf3 6, scc1 sac3 6, scc2 JLP 6, scc1 rQDï , scc2 rad27 6, scc2 hos1 6, scc1 rQDï , scc1 rrSï , scc1 pac10 6, scc1 pac10 6, scc2 SFIïWV , scc1 VWXï , scc1 trPï'DPS, scc1 SFIïWV , scc2 SFIï , scc1 gim3 6, scc2 sac3 6, scc1 WXEï6DSY, scc2 bim1 6, scc2 mdm20 6, scc1 bub3 6, scc2 WXEï6DSY, scc1 VWXï , scc2 rSQï , scc1 VWXï , scc1 irc15 6, scc1 rSQï , scc2 irc15 6, scc2 doc1 6, scc2 kar3 6, scc2 FGFï , scc1  Bonferroni Corrected p-value = .05 p-value = .05  −40  −30  −20  −10  0  Interaction T-Statistic (30oC) Supplementary Figure 7  T-statistics are ranked according to magnitude. Dotted lines indicate a p-value cut off of .05, and dashed lines indicate a Bonferroni corrected p-value of .05 90  Table 3.6  Double mutant interactions ranked by T-statistic at 26oC  Rank  Double Mutant$  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51  scc1, rpn11-14 scc1, tub4-Δ - DSY scc1, irc15Δ scc1, pac10Δ scc1, mdm20Δ scc1, stu2-12 scc1, bub3Δ scc1, pcf11-ts9 smc1, lst8-15 scc1, pcf11-1 smc1, tub2-443 smc1, irc15Δ smc1, gim3Δ scc1, rrp4-1 smc1, rrp4-1 smc1, hos1Δ smc1, sac3Δ smc1, trm112-Damp scc1, sac3Δ scc1, stu2-13 smc1, pcf11-ts9 scc1, hos1Δ smc1, rna15-58 smc1, pac10Δ scc1, eaf3Δ scc1, tub2-443 smc1, stu2-12 smc1, gim4Δ scc1, rna15-58 scc1, cdc20-2 scc1, rps31Δ scc1, rad27Δ scc1, rps16BΔ smc1, rad27Δ scc1, trm112-Damp smc1, rps16BΔ smc1, stu2-13 smc1, rps31Δ smc1, cdc20-2 scc1, ypr1Δ scc1, gim4Δ scc1, kar3Δ scc1, lst8-15 smc1, eaf3Δ scc1, lpd1Δ smc1, pcf11-1 scc1, doc1Δ smc1, ypr1Δ smc1, tub4-Δ DSY scc1, gim3Δ smc1, bim1Δ  Interaction Estimate -1.09 -0.74 -0.82 -0.68 -0.71 -0.63 -0.63 -0.52 -0.55 -0.51 -0.45 -0.49 -0.49 -0.48 -0.43 -0.42 -0.39 -0.38 -0.34 -0.29 -0.28 -0.32 -0.27 -0.26 -0.25 -0.23 -0.26 -0.26 -0.21 -0.22 -0.19 -0.21 -0.18 -0.20 -0.19 -0.16 -0.15 -0.12 -0.13 -0.11 -0.09 -0.05 -0.03 -0.02 0.03 0.07 0.07 0.08 0.09 0.13 0.27  Standard Error 0.058 0.051 0.058 0.051 0.058 0.058 0.058 0.051 0.058 0.058 0.051 0.058 0.058 0.058 0.058 0.058 0.058 0.058 0.058 0.051 0.051 0.058 0.051 0.051 0.051 0.051 0.058 0.058 0.051 0.058 0.051 0.058 0.051 0.058 0.058 0.051 0.051 0.051 0.058 0.058 0.058 0.058 0.058 0.051 0.058 0.058 0.058 0.058 0.051 0.058 0.058  P Value  T Statistic  7.45E-52 5.57E-36 1.55E-34 6.06E-32 3.14E-28 2.16E-23 4.81E-23 7.75E-21 9.98E-19 9.46E-17 1.05E-16 1.61E-15 2.45E-15 7.96E-15 1.32E-12 3.60E-12 1.47E-10 3.18E-10 1.34E-08 3.79E-08 8.30E-08 1.06E-07 3.17E-07 4.54E-07 2.53E-06 8.24E-06 8.28E-06 1.27E-05 4.35E-05 1.89E-04 3.07E-04 4.32E-04 6.64E-04 7.27E-04 1.14E-03 1.42E-03 3.36E-03 2.01E-02 2.81E-02 5.42E-02 1.16E-01 4.26E-01 6.14E-01 7.38E-01 5.61E-01 2.10E-01 1.99E-01 1.73E-01 8.94E-02 2.86E-02 5.95E-06  -18.74 -14.45 -14.05 -13.34 -12.30 -10.90 -10.79 -10.13 -9.48 -8.84 -8.83 -8.43 -8.37 -8.20 -7.42 -7.26 -6.65 -6.52 -5.85 -5.65 -5.50 -5.45 -5.24 -5.16 -4.80 -4.54 -4.54 -4.44 -4.15 -3.78 -3.65 -3.56 -3.44 -3.42 -3.29 -3.22 -2.96 -2.34 -2.21 -1.93 -1.58 -0.80 -0.51 -0.34 0.58 1.26 1.29 1.37 1.70 2.20 4.61  91  $  Alleles used for SMC1, SCC1, and SCC2 were smc1-259, scc1-73 and scc2-4, respectively.  Table 3.7  Double mutant interactions ranked by T-statistic at 30oC  Rank*  Double Mutant$  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45  scc1, cdc20-2 scc2, kar3Δ scc2, doc1Δ scc2, irc15Δ scc2, rpn11-1 scc1, irc15Δ scc1, stu2-12 scc1, rpn11-14 scc2, stu2-12 scc1, tub4-Δ DSY scc2, bub1Δ scc1, mdm20Δ scc2, bim1Δ scc2, tub4-Δ DSY scc1, sac3Δ scc2, gim3 scc1, pcf11-1 scc2, pcf11-ts9 scc1, trm112-damp scc1, stu2-13 scc1, pcf11-ts9 scc2, pac10Δ scc1, pac10Δ scc1, rrp4-1 scc1, rna15-58 scc1, hos1Δ scc2, rad27Δ scc2, rna15-58 scc1, gim4Δ scc2, sac3Δ scc1, eaf3Δ scc2, lpd1Δ scc1, rad27Δ scc1, rps16BΔ scc2, lst8-15 scc1, bub3Δ scc1, rps31Δ scc2, cdc20-2 scc2, rrp4-1 scc2, hos1Δ scc1, tub2-443 scc2, rad61Δ scc2, rps16BΔ scc2, tub2-443 scc2, gim4Δ  Interaction Estimate -1.97 -1.55 -1.41 -1.48 -1.60 -1.18 -1.08 -1.22 -1.07 -1.01 -0.90 -0.91 -0.90 -0.74 -0.87 -0.73 -0.68 -0.76 -0.66 -0.65 -0.73 -0.56 -0.54 -0.58 -0.48 -0.43 -0.48 -0.36 -0.36 -0.40 -0.34 -0.32 -0.36 -0.32 -0.28 -0.27 -0.31 -0.24 -0.27 -0.23 -0.20 -0.21 -0.16 -0.14 -0.12  Standard Error 0.045 0.045 0.043 0.045 0.051 0.045 0.045 0.051 0.045 0.043 0.045 0.051 0.051 0.043 0.051 0.045 0.045 0.051 0.045 0.045 0.051 0.045 0.045 0.051 0.045 0.045 0.051 0.045 0.045 0.051 0.045 0.045 0.051 0.045 0.045 0.045 0.051 0.045 0.051 0.045 0.045 0.051 0.045 0.045 0.045  P Value  T Statistic  5.51E-145 1.06E-114 5.41E-110 7.03E-110 6.73E-104 5.79E-85 4.76E-76 1.48E-75 1.49E-75 1.67E-74 4.31E-60 1.15E-50 4.08E-50 6.98E-49 1.18E-47 1.88E-44 2.38E-40 4.08E-39 4.97E-38 5.54E-37 1.82E-36 5.71E-30 2.56E-28 2.26E-25 2.07E-23 1.15E-19 1.05E-18 1.83E-14 2.63E-14 4.19E-14 2.00E-13 3.79E-12 1.00E-11 1.04E-11 1.92E-09 4.00E-09 5.47E-09 2.09E-07 3.11E-07 4.69E-07 9.00E-06 7.01E-05 4.41E-04 1.55E-03 1.09E-02  -43.72 -34.29 -32.94 -32.91 -31.24 -26.20 -23.93 -23.81 -23.81 -23.55 -19.99 -17.69 -17.56 -17.26 -16.96 -16.17 -15.15 -14.84 -14.57 -14.31 -14.18 -12.50 -12.06 -11.26 -10.72 -9.63 -9.34 -8.00 -7.94 -7.88 -7.64 -7.19 -7.04 -7.03 -6.16 -6.03 -5.98 -5.29 -5.22 -5.13 -4.51 -4.02 -3.55 -3.19 -2.56  92  Rank*  Double Mutant$  46 47 48 49 50 51 52 53 54 55  scc2, rps31Δ scc2, trm112-Damp scc1, lpd1Δ scc1, ypr1Δ scc2, eaf3Δ scc2, pcf11-1 scc1, doc1Δ scc1, kar3Δ scc1, lst8-15 scc1, gim3Δ  $  Interaction Estimate -0.11 -0.09 -0.09 -0.10 -0.07 0.00 0.01 0.07 0.11 0.14  Standard Error 0.051 0.045 0.045 0.051 0.045 0.045 0.043 0.045 0.045 0.045  P Value  T Statistic  3.85E-02 5.64E-02 5.81E-02 6.41E-02 1.22E-01 9.22E-01 8.55E-01 1.42E-01 1.92E-02 2.32E-03  -2.08 -1.91 -1.90 -1.86 -1.55 0.10 0.18 1.47 2.35 3.07  Alleles used for SMC1, SCC1, and SCC2 were smc1-259, scc1-73 and scc2-4, respectively.  Table 3.8  Summary of interactions identified by Growth Curve Analysis  Temperature 26oC 30oC  # of interactions analyzed 51 55  # Alleviating interactions  # SS interactions  7 5  44 50  This table does not include SL interactions identified by random spore analysis  93  84  SCC1 72  44  SCC2  12  SCC1  13  14  SMC1 246  20  SCC2 9  SMC1 191  1692 434 Negative interaction, E-C < -.3 364 p value < .05 Figure 3.8 Validated S. cerevisiae network 1435  A.  7  SCC1 13  12  SCC2 7  SMC1  90 (78) 2/3 query genes, Conserved 39 (33)  SCC1 0  2 28  SCC2 2  SMC1 0  94 Validated by RS, GC 33  B. scc2-4, irc15630 C irc156  OD600nm  1.5  scc2-4  1.0  0.5  WT  scc2-4, irc156 -32.9  scc1-73, irc15630 C  OD600nm  irc156 scc1-73  WT  scc1-73, irc156 -26.2  smc1-259, irc15626 C  OD600nm  irc156  smc1-259  WT smc1-259, irc156 -8.43 0  5  10  15  Time (hrs)  20  Spindle/Kinetochore Cohesin  Replication  Protein/Cellular Metabolism  Chromatin  Microtubule/Prefoldin  mRNA Processing  Figure 1  A) Representative subset of growth curve data. The T-statistic, which takes into account the interaction magnitude and statistical significance, is shown in blue. B) Expanded view of the final network summarized in Figure 3.2 A iv. Red lines indicate SL interactions and black lines represent SS interactions. The black line thickness increases with increases interaction strength.  94  3.3.2  Negative genetic interactions with cohesin are conserved in C. elegans  One of the main interests of this work was to identify interactions that are conserved in mammalian cells and thus relevant to the development of therapeutics. Interactions that are conserved among eukaryotes are more likely to be conserved in higher animals and therefore we tested our set of validated S. cerveisiae interactions in the model metazoan, C. elegans. The visual screen developed in chapter 2 was used to monitor defects in development of the C. elegans vulva. This assay can identify synthetic genetic interactions that result in defects in somatic cell proliferation (McLellan et al. 2009). The vulval cell divisions occur late in nematode development (Sulston & Horvitz 1977), so perturbations do not affect organismal viability. This allowed screening for interactions using gene mutations or RNAi knockdowns that are potentially organismal lethal. The C. elegans orthologs of all genes that had a verified genetic interaction with one of the cohesin genes in S. cerevisiae and for which there was an RNAi construct available were tested (Table 3.5). For the cohesin mutant, a viable hypomorphic mutation in the C. elegans SMC1 orthologue, him-1(e879), was used. These worms were treated with RNAi by feeding after embryos hatched. Defects in the vulva apparent as a protruding vulva (Pvl) were scored once worms had matured to adults. An increased frequency of defects in the mutant treated with RNAi, compared to the predicted additive effect for the mutant and RNAi separately, is indicative of a genetic interaction. A clear increase over the predicted additive frequency of Pvl was observed in 23 out of 28 (82%) interactions tested suggesting that many interactions are conserved between S. cerevisiae and C. elegans (Figure 3.9).  95  Figure 3.9 35  VC2010 (WT) SMC1/him-1(e879)  30  % of adults with Pvl  C. elegans genetic interactions  25  20  15  10 Randomly chosen RNAi clones  5  no  RN un Ai cF4 22 4F F3 1.3 5 E0 E2.9 3H F1 4.4 2 W B6.1 03 D T 8 EA 23 .3 YP F3 B3.4 R1 / m / Y rg 3 -1 TU 9G B4 8B. BI / tb 1 M g CD 1/ e -1 C2 bp TR GI 0/ f -2 M M4 zy1 / 1 CH 12/ pfd L1 C0 -2 / M 4H 5 BU 03C .1 B3 11 / . G b 2 LS IM ubT8 3/ 3 / C pfd PA 10H -4 C 1 HO 10/ 1.8 S p DO 1/ fd-3 C1 hda RN / a -1 CT A15 pc-1 F4 / 0 / F cpf RA 17C -2 SA D2 11 C3 7/ .10 DC / F2 crn C1 0D -1 CS / K0 12.2 M 9H 3 M / F2 9.2 DM 3 2 C8 KA 0/ c .9 r R RA 3/ a-1 D6 klp 1 -1 ST /w 7 U2 ap CT / z l-1 F8 yg PC / T2 -9 F1 2C 1 1 LP TU /pc .4 D1 B2 f-1 ,IR / t 1 C1 bb 5/ -4 LL C1 .3  0  RNAi  B. depicting Image of the a WTfrequency worm Graph of worms with a protruding vulva (Pvl) when VC2010 (WT) and SMC1/him-1(e879) strains are treated with various RNAi constructs. The RNAis tested targeted the C. elegans orthologs of genes that interact with one of the cohesin query C. Image of a worm with Pvl genes in the S. cerevisiae validated network (Figure 1C). Orthologs in S. cerevisiae and humans can be found in Table 3.5. Interactions are ranked by the difference between the frequency of Pvl in VC2010 and him-1 strains. The predicted value is the sum of both the Figure 2 him-1 and WT background Pvl frequencies and the effect of the RNAi on WT. unc-22 and 6 randomly chosen RNAi clones from chromosome 1 are included as negative controls. Error bars represent SEM. n > 150 for each condition.  96  3.3.3  Analysis of cohesin network sub-groups  Each gene in the S. cerevisiae network was assigned to one of several broad functional groups including the spindle, microtubules and kinetochore (39%), mRNA processing (12%), replication factors (24%), and a group that contains diverse genes involved in cellular metabolism (24%). The absence of genes involved in the DNA damage response is notable given the role of cohesin in maintaining genome stability.  For the phenotypic analysis, information collected from the Saccharomyces genome database (SGD) and a recently performed screen for CIN (Stirling et al. 2011) in essential genes was used to ask how many genes in this network were also yeast CIN genes (Table 3.9). Cohesin genes are themselves CIN genes. One hypothesis for a genetic interaction is that two mutations that cause the same phenotype, for example CIN, when combined cause a cumulative phenotype that overwhelms the tolerance level of the cell. 85% of genes, excluding those that were grouped into the metabolism category, cause CIN, as measured by a variety of assays such as chromosome transmission fidelity (CTF) and gross chromosomal rearrangements (GCR) in yeast (Stirling et al. 2011).  To profile the identified interactions based on strength, the network was filtered to include only SL interactions, based on the hypothesis that these would indicate the most sensitized processes in the presence of sub-optimal cohesin (Figure 3.10). Genes involved in replication were central to the SL network as the five genes that were SL with the three cohesin query genes were members of this group. The replication gene mutations that are SL with cohesin mutants are specifically involved in replication fork progression and stability. Ctf8 and Dcc1 are components of the alternative replication factor C Ctf18 97  clamp loader (altRFCCTF18) that controls the speed and restart activity of the replication fork (Terret et al. 2009). Rad61 acts with Pds5 to bind cohesin and regulates its association with chromatin (Sutani et al. 2009, Kueng et al. 2006). This complex acts in opposition to Smc3 acetylation by Eco1, a modification required for processive DNA synthesis and SCC (Rolef Ben-Shahar et al. 2008, Rowland et al. 2009, Beckouet et al. 2010). Chl1 is a DNA helicase that interacts genetically with CTF18 and physically with Eco1 (Skibbens 2004). Ctf4/AND1 has a role in coupling the Mcm2-7 helicase replication progression complex to DNA polymerase alpha (Tanaka et al. 2009). Csm3 functions in a complex with Tof1 and Mrc1 to control stable pausing of the replication fork (Bando et al. 2009, Komata et al. 2009). These genes were collectively called replication fork mediators. The SL interaction of these replication fork mediators with multiple cohesin mutants suggested that the cell is sensitized to perturbations of the replication fork in the presence of cohesin mutations.  98  Table 3.9  Reported CIN phenotypes of cohesin interacting genes  S. cerevisiae Gene BIM1 BUB3 CDC20 CHL1 CSM3 CTF4 CTF8 DCC1 DOC1 EAF3 GIM3 GIM4 HOS1 IRC15 KAR3 LPD1 LST8 MDM20 PAC10 PCF11 RAD27 RAD61 RNA15 RPN11 RPS16B RPS31 RRP4 SAC3 STU2 TRM112 TUB2 TUB4 YPR1  CIN phenotypes CTF, BiM, ALF CTF, BiM, ALF CTF CTF, BiM, ALF CTF, BiM, ALF, LOH CTF, BiM, ALF CTF, BiM, ALF, GCR CTF, BiM, ALF, LOH CTF, ALF N/A N/A ALF, BiM CTF (overexpression) CTF, BiM, ALF CTF, ALF N/A N/A N/A ALF, BiM CTF CTF, BiM, ALF, LOH, GCR CTF, BiM, ALF CTF CTF N/A N/A CTF, GCR ALF N/A N/A Chromosome missegregation CTF N/A  CTF: chromosome transmission fidelity; BiM: bi-mater; ALF: A-like faker; LOH: loss of heterozygosity; GCR: gross chromosomal rearrangement  99  Figure 3.10  Cohesin mutations are SL with replication fork mediators  STU2  CHL1  DCC1  Red lines indicate SL interactions and genes are color coded according to Figure 3.8.  3.3.4  SMC1/him-1 mutants have increased apoptosis in the C. elegans germline  when fork mediators are disrupted To further investigate the biology underlying these interactions the frequency of apoptosis in the C. elegans germline was analyzed. In the C. elegans germ line, nuclei with DNA damage are removed by apoptosis and increased germline apoptosis is indicative of increased DNA damage (Gartner, MacQueen & Villeneuve 2004). Apoptosis levels were quantified in the germline of the him-1(e879) mutant and were found to be elevated above that of wild type. When him-1 mutant animals were treated with RNAi against CSM3, RAD61, CTF8, CTF4 and DCC1 a further increase in apoptosis was detected. Apoptotic bodies in these cases were typically found in large clusters, rather than being distributed throughout the pachytene region, reminiscent of irradiated animals (Figure 3.11). These results suggested that compromised HIM-1 and fork mediator function leads to irreparable DNA damage events resulting in apoptosis. 100  plementary Figure 8  Figure 3.11  A.  Apoptosis in him-1 mutants treated with fork mediator RNAi  Average number of apoptotic corpses per gonad arm  18 16 14 12 10 8 6 4 2  N2  (V C hi 20 m 10 -1 (e ) 87 9) N2 ,R hi AD m 6 -1 , R 1(RN AD A 61 i) (R NA i) N2 ,C SM hi m 3 -1 , C (RN SM Ai) 3( RN Ai ) N2 , hi CTF m -1 4(RN ,C TF Ai) 4( RN Ai ) N2 ,C TF hi m -1 8(RN ,C TF Ai) 8( RN Ai ) N2 ,D hi CC m 1( -1 , D RNA CC i 1( ) RN Ai )  0  no RNAi  B.  RAD61  CTF4  CTF8  N2  him-1  A) Graph showing the average number of apoptotic corpses per gonad arm. B) Representative images showing apoptotic corpses in untreated WT and him-1 worms and worms treated with RAD61, CSM3, and CTF8 RNAi. Error bars represent SEM. n= 30-35 gonad arms examined for each condition.  101  3.3.5  SMC1/him-1 genetically interacts with the PARP pathway in C. elegans  Cohesin mutants are SL with mutations in genes that mediate replication fork progression but are not SL with any DNA repair mutants (Table 3.10). These results suggested that the cohesin mutations lead to replication fork progression defects but not directly to DNA damage. This was further supported by the observation that the cohesin mutations do not result in RAD52 foci accumulation, which are indicative of HR repair intermediates (Data not shown, personal communication P. Stirling, S. Minaker, P. Sipahimalani). Additionally, genetic elimination of the HR repair pathway by deletion of RAD51 in mediator gene-cohesin gene double mutants did not rescue lethality, suggesting that ineffective HR is also not the cause for lethality in a cohesin mutant background (Figure 3.12). This evidence collectively suggests that the interaction between cohesin mutants and fork mediators is intimately tied to the regulation of replication progression and led us to further investigate this relationship.  102  Table 3.10  HR Gene RAD51 RAD52 RAD54 RDH54 RAD55 RAD57 RAD59 RAD50 MRE11 XRS2 RAD1 RAD10 RAD9 SGS1 MUS81  SGA scores for the major genes involved in homologous recombination  SMC1 screen E-C (p value) -.027 (.82) -.13 (.70) -.26 (.48) -.14 (.05) +.39 (.10) -.02 (.96) +.10 (.30) -.36 (.13) -.39 (.20) -.004 (.98) +.27 (.59) +.15 (.05) +.13 (.79) -.29 (.42) +.06 (.26)  SCC1 screen E-C (p value) -.006 (.82) -.004 (.96) +.05 (.19) -.02 (.45) -.09 (.01) -.77 (2.5e-14) -.29 (7.2e-5) -.04 (.36) -.16 (.07) -.03 (.72) -.14 (.0002) +.08 (.005) -.005 (.82) +.03 (.42) +.07 (.005)  SCC2 screen E-C (p value) +.13 (.37) -.004 (.97) +.07 (.62) -.01 (.66) +.04 (.30) -.02 (.83) -.09 (.08) -.05 (.52) -.12 (.23) +.01 (.81) -.01 (.72) -.03 (.96) +.06 (.26) -.056 (.087) +.07 (.055)  The grey box highlights the one interaction that would have been considered significant in terms of p value and interaction magnitude in the data filtering process used in this study.  103  Figure 3.12  Knockout of RAD51 does not rescue the lethality of cohesin, fork mediator double  mutants  A)  haploid selection  haploid +G418  B)  haploid selection  haploid +G418  haploid -ura  haploid +Hygromycin  haploid -ura  haploid +Hygromycin  haploid -ura +G418  haploid -ura +G418 +Hyg  haploid -ura +G418  haploid -ura +G418 +Hyg  scc1-73::URA3 csm36::KanMX rad516::Hyg  smc1-259::URA3 ctf86::KanMX rad516::Hyg  RAD51 was replaced with the resistance gene for hygromycin in double heterozygous smc1-259 and fork mediator (CSM3, CTF4, CTF8, DCC1, RAD61) mutants. The same was done for scc1-73, fork mediator double mutants. Random spore was performed on all 10 triple mutants and no rescue of lethality was seen in any cases. Random spore results are shown for A) scc1-73, csm3Δ, rad51Δ and B) smc1-259, ctf8Δ, rad51Δ triple heterozygotes.  In higher eukaryotes there are additional factors that protect and regulate the replication fork. A major early mediator of the replication fork is the family of Poly (ADP-ribose) polymerases (PARPs). PARPs have been shown to localize to stalled forks and mediate restart (Bryant et al. 2009). PARPs are not present in yeast but are present in C. elegans. Supplementary Figure 9  To assess whether PARP metabolism plays a role in maintaining viability in a cohesin  104  mutant, double mutants with him-1(e879) and the five PARP metabolism (pme) genes in C. elegan were constructed. pme-1 and pme-2 are the C. elegans orthologues of PARP1 and PARP2, respectively (Gagnon, Hengartner & Desnoyers 2002). pme-3 and pme-4 are homologs of Poly (ADP-ribose) glycohydrolase (PARG), which degrades the ADP-ribose polymers into monomeric ADP-ribose units (St-Laurent et al. 2007). pme-5 is the C. elegans orthologue of PARP5, which is also known as TANKYRASE (Gravel et al. 2004). All him-1; pme double mutants exhibit decreased brood sizes and an increase in the frequency of arrested embryos (Figure 3.13). Strikingly, him-1; pme-2 double mutants had a very high frequency of protruding vulva phenotype (Figure 3.13) indicative of somatic cell proliferation defects.  105  m pm -1 e ; p -5 m e5  hi  m pm -1 e ; p -5 m e5  hi  m pme -1 ; p -4 m e4  hi  Embryonic	
  lethality  m pme -1 ; p -4 m e4  m pm -1 e;p 1 m e1 hi pm m -1 e-2 ;p m (ok e- ) 2( ok ) hi p m m -1 e;p 2( m tm e- ) 2( tm ) hi p m m -1 e;p 3 m e3  hi  1 hi 0 m -1  20  VC  Brood	
  size  hi  m pm -1 e ;p 1 m e1 hi pm m -1 e-2 ;p m (ok e- ) 2( ok ) hi p m m -1 e;p 2( m tm e- ) 2( tm ) hi p m m -1 e ;p 3 m e3  hi  20 1 hi 0 m -1  VC  Brood	
  size/	
  Embryonic	
  lethality  A) 350 %	
  males  300  250  200 15  150  100 10  50 5  0 0  B) 50  40  30  20  10  0  106  %	
  of	
  males  %	
  of	
  hermaphrodites	
  with	
  Pvl  Figure 3.13 him-1 interacts with pme/PARP mutants in C. elegans  30  25  20  A) Brood size, embryonic lethality, and % males in him-1 and pme single and double mutants. B) % of hermaphrodites with Pvl. Error bars represent SEM. n=6-9 broods analyzed for each mutant. 3.3.6  SMC1 genetically interacts with PARP in human cells  The strong negative genetic interaction in C. elegans between a hypomorphic SMC1 mutant and PARP mutants prompted testing in human cells to see if this interaction was conserved. An early generation PARP inhibitor, benzamide (Li & Zhang 2001), was used to inhibit PARP function in a panel of HCT116 cells treated with siRNA against a variety of cohesin genes (SMC1, SMC3, SCC2/NIPBL, SCC1/RAD21, SCC3/STAG1 and SCC3/STAG3) in a high throughput format using high content digital imaging microscopy (HC-DIM) to count Hoescht stained nuclei (Figure 3.14). This preliminary assay suggested that siRNA against several cohesin genes induces sensitivity to PARP inhibition. The interaction with SMC1 and PARP was further explored using both clonagenic survival assays and additional screening with HC-DIM. Olaparib, a PARP inhibitor that has undergone phase II clinical trials was used in these subsequent assays (Audeh et al. 2010, Tutt et al. 2010). This assay found that BRCA1 siRNA treated cells are highly sensitive to low doses of olaparib, as has been previously reported (Figure 3.14) (Bryant et al. 2005, Farmer et al. 2005). SMC1 siRNA treated cells were also qualitatively sensitive to a range of olaparib concentrations using a 24 well plate survival assay while untreated and GAPDH siRNA treated cells were only mildly affected (Figure 3.15 A). This sensitivity was quantified using an expanded 10cm dish survival assay where cells treated with the various siRNAs were continuously exposed to 0.6uM olaparib and colonies were stained and counted after a 10 day period (Figure 3.15 B). Proliferation defects were also evident in HCT116 SMC1 knockdown cells exposed to  107  olaparib using HC-DIM. A significant dose dependent decrease in cell number in the BRCA1 and SMC1 treated cells was observed as compared to GAPDH (Figure 3.15 C). Knockdown of SMC1 and BRCA1 was evidenced in these cells by western blotting with the appropriate antibody (Figure 3.15 D).  Figure 3.14  Cohesin deficient cells are sensitive to the PARP inhibitor benzamide  1.4  EtOH  5mM Benzamide  Cell number normalized to no drug control  1.2 1.0 0.8 0.6 0.4 0.2  ST AG 3  ST AG 1  1 D2 RA  C3 SM  C1 SM  2 BR  CA  1 CA BR  GA  PD H  0  siRNA  HCT116 cells were transfected with the siRNA indicated and exposed to 5mM benzamide for 3 days before fixing, staining with Hoescht, and counting cell number using HC-DIM. Error bars represent SEM. n=8 technical replicates for each condition.  Supplementary Figure 11  108  Figure 3.15  SMC1 siRNA treated human cells are sensitive to the PARP inhibitor olaparib  109  In all panels HCT116 cells were treated with siRNA targeting GAPDH, BRCA1 or SMC1 or untreated. A) 24 well plate clonagenic survival assay with HCT116 siRNA treated cells exposed to olaparib concentrations up to 1.5uM. B) 10cm dish clonagenic survival assay looking at the number of colonies after 10 days in the presence of 0.6uM olaparib. C) HC-DIM counting Hoescht positive nuclei of siRNA treated cells after 3 days of olaparib exposure D) Western blot of protein preparation from asynchronously growing cells collected from the 10cm survival assay in B showing knockdown of SMC1 and BRCA1 only in the respectively treated siRNA samples. Protein samples were not treated with olaparib. Error bars represent SEM.  3.4 3.4.1  Discussion Cohesin and cancer  Cohesin dysfunction is emerging as a potential driver of tumour formation and progression (Reviewed in Xu, Tomaszewski & McKay 2011). Mutations in cohesin components and altered expression of cohesin and cohesin-associated genes have been found in a number of human cancers (Oikawa et al. 2004, Barber et al. 2008, Xu, Tomaszewski & McKay 2011, Xu et al. 2011, Ghiselli, Iozzo 2000, Zhang et al. 2008, Hagemann et al. 2011). Given the critical roles of cohesin in the prevention of aneuploidy through the regulation of sister chromatid cohesion (SCC) (Michaelis, Ciosk & Nasmyth 1997), and in DNA repair (Sjogren, Nasmyth 2001, Unal, Heidinger-Pauli & Koshland 2007, Heidinger-Pauli, Unal & Koshland 2009), it is not surprising that disruption of cohesin function might contribute to tumour formation and/or progression. This work should contribute to the understanding of the cellular consequences of cohesin gene mutation and may lead to identification of potential chemotherapeutic targets.  SL genetic interactions in model organisms identify candidate genes or pathways that can be targeted for inhibition leading to specific killing of tumour cells with specific mutations (Hartwell et al. 1997). SGA technology was used in yeast to generate a 110  network of negative genetic interactions using hypomorphic mutations in cohesin as query genes, with the aim of elucidating the genetic pathways needed for survival when cohesin function is compromised. These SGA screens queried both non-essential gene deletions and conditional alleles of essential genes. The inclusion of ts and Damp essential alleles (Schuldiner et al. 2005, Breslow et al. 2008, Ben-Aroya et al. 2008) allowed these screens to be comprehensive and the value of this expanded set was demonstrated by the finding that 30% of validated interactions occurred with essential genes.  3.4.2  Cohesin network  By overlaying the results from three separate SGA screens, two cohesin core components (SMC1 and SCC1) and one cohesin loader (SCC2), common interactions with compromised cohesin were identified rather than interactions that were allele or component-specific. While these screens found scores of single negative interactions with each of the three query mutants (Figure 3.2), the goal of this study was to identify interactions that are more likely to be SL with the wide-range of cohesin mutations observed in human tumours. Negative genetic interactions specific to only one of the three cohesin mutants may reveal specific aspects of cohesin components and these interactions warrant further investigation.  The filtered interaction set identified many processes that would be predicted to interact with cohesin dysfunction. For example, mutations affecting the mitotic spindle, kinetochore and microtubules, including prefoldin, were found to result in synthetic growth defects with at least two of the three query mutants. Given the role of cohesin in 111  regulating chromosome segregation leading to the anaphase transition (Michaelis, Ciosk & Nasmyth 1997), mutations affecting the spindle or kinetochore would be predicted to synergize with cohesin mutations. Other interactions that do not appear to be related to cohesin function may represent more general effects on cell viability; for example ten interactions were identified with components of mRNA processing and the translation machinery. In these cases, further work will be needed to ascertain whether this is a general synergistic effect on viability or whether the interactions result in a specific defects.  When interactions were filtered based on strength, keeping only those interactions that were SL, the predominant genes were those involved in mediating replication fork (RF) progression. These included two of the three genes in the alternative RFCCTF18, the replisome components Csm3 and Ctf4, the replication and cohesion-associated helicase Chl1, and the cohesin regulator Rad61. All of these interactors are known to mediate replication fork stability and progression (Rolef Ben-Shahar et al. 2008, Sutani et al. 2009, Rowland et al. 2009, Kueng et al. 2006, Terret et al. 2009, Beckouet et al. 2010, Skibbens 2004, Tanaka et al. 2009, Bando et al. 2009, Komata et al. 2009). Furthermore, mutations affecting these genes also result in chromosome cohesion defects (Mayer et al. 2004), thereby linking replication fork stability and SCC.  The SL interactions are not specific to mutations that accelerate or impair replication fork progression. In fact, the Alt-RFCCTF18, which promotes RF progression, does so by regulating the acetylation of Smc3 (Terret et al. 2009), which in turn inhibits the  112  association of Smc3 and Rad61 (Rowland et al. 2009, Beckouet et al. 2010). In contrast to the Alt-RFCCTF18, Rad61 binding to Smc3 slows replication fork progression (Terret et al. 2009). Rad61/WAPL is known to destabilize the association between cohesin and chromosomes (Sutani et al. 2009). The requirement for Rad61 in the cohesin mutants is particularly interesting given the correlation of elevated expression of the human Rad61 orthologue, WAPL, with certain cancers. Its importance in maintaining the viability of tumour cells was demonstrated by the fact that knockdown of WAPL in cervical cancer cell lines resulted in cell death (Oikawa et al. 2004). Similarly, SL was observed when RAD61 was knocked out in the cohesin mutants, thereby demonstrating that Rad61 function is needed when cohesin function is compromised.  The SL of the cohesin mutations with RF mediator mutations suggested that the replisome must be stabilized when cohesin is mutated to ensure proper progression. Interestingly, cohesin mutants did not show negative genetic interactions with DNA repair genes such as the RAD52-complementation group of HR genes, the RecQ helicase SGS1, and the structure-specific endonuclease MUS81. In addition, the cohesin mutants do not accumulate RAD52 foci (Data not shown P. Stirling, S. Minaker, P. Sipahimalani), which is indicative of increased DNA damage or DNA repair intermediates. These data suggested that cohesin mutations on their own do not lead to DNA double strand breaks or fork collapse, both of which require HR for resolution. Cohesin dysfunction may impact replication fork dynamics, but these events are tolerable in the presence of fork stabilizing proteins.  113  Cohesin may also have a role in modulating fork progression. As human replicons span between 60 and 140kb (Blumenthal, Kriegstein, Hogness 1974, Walter, Newport 1997) and cohesin is associated, on average with DNA every 10-20kb (Laloraya et al. 2000, Parelho et al. 2008, Wendt et al. 2008, Wendt et al. 2009), each fork must theoretically pass several cohesin complexes (Terret et al. 2009). One mechanism to prevent fork collapse is to keep the distance between the leading strand helicase to the polymerase to a minimum. If the polymerase stalls and the helicase continues to advance, single stranded DNA is exposed and the fork becomes more fragile. Rad61 association with cohesin is thought to induce a closed cohesin conformation that limits fork progression (Terret et al. 2009). Tightening of cohesin may be one mechanism to prevent helicase and polymerase separation. Support for cohesin moderating fork progression comes from the finding that SMC1 is phosphorylated by ATR in response to S phase stress (Kim, Xu & Kastan 2002). This modification is required for activation of the replication checkpoint.  3.4.3  Model organisms for the discovery of therapeutic targets  In this work the strong negative interactions between cohesin mutants and replication fork mediators were first observed in yeast and were found to be conserved in C. elegans and human cells. In higher eukaryotes, additional factors regulate replication fork stability and progression, one example being the family of poly(ADP-ribose) polymerases (PARPs). PARPs play a major role in DNA metabolism including aspects of repair and replication (reviewed in Helleday 2011). PARP was recently found to be activated by stalled replication forks to enable restart pathways (Bryant et al. 2009). The work in this thesis demonstrated that loss of PARP, PARG, and TANKYRASE orthologs synergized with a mutation in him-1/SMC1 in C. elegans resulting in a decrease in cell 114  and organismal viability. In human cells depletion of SMC1 by siRNA caused sensitivity to the PARP-inhibitors olaparib and benzamide. Interestingly, this sensitivity was comparable to that observed when we depleted BRCA1 by siRNA and then treated cells with olaparib. Recent emerging hypotheses have proposed alternate fork restart pathways where one branch is HR mediated and the other requires PARP (Helleday 2011). It has been suggested that BRCA1/2 are SL with PARP1/2 because of an inability to perform fork restart. This hypothesis is interesting given the connection we observe between cohesin and fork progression mediators and is suggestive that cohesin may have additional roles that relate to replication fork dynamics.  Given the findings regarding the importance of RF mediators in the presence of cohesin mutations it is possible that PARP activity is also needed for replication fork stability in the cohesin mutants or knockdown lines. However, it is also possible that the SL of cohesin knockdown with PARP inhibition is related to the role of cohesin in HR (Sjogren, Nasmyth 2001, Unal, Heidinger-Pauli & Koshland 2007, Heidinger-Pauli, Unal & Koshland 2009), although in yeast this requirement has only been clearly shown post S phase. PARP inhibition is known to be effective in killing cells with defective HR, such as BRCA1, BRCA2 and ATM-deficient cells (Bryant et al. 2005, Farmer et al. 2005). In light of this discovery, PARP inhibitors are currently being developed for the treatment of tumours with defective HR (Audeh et al. 2010, Khan et al. 2011, Tutt et al. 2010). Given the broad range of biological roles of both cohesin and PARP, it is possible that the SL of cohesin knockdown and PARP inhibition is due to either or both mechanisms. Although further investigation is needed to identify the specific mechanism of lethality, this data  115  indicated PARP inhibition is a potential treatment for tumours with cohesin-related mutations and SCC defects. This work should have a significant impact on therapy for a variety of tumours as overall 20% of tumours harbor cohesin-associated mutations (Xu, Tomaszewski & McKay 2011, Solomon et al. 2011, Barber et al. 2008).  116  Chapter 4: CDC4 plays a role in the DNA damage response 4.1  Introduction  Sequencing of yeast CIN genes in human colon tumours has identified a set of eight genes that are mutated at an elevated frequency, and collectively account for CIN gene mutations in approximately 25% of cancers examined (Barber et al. 2008, Wang et al. 2004, Yuen et al. 2007). CDC4 has the highest mutation frequency in this set (9%) (Kemp et al. 2005, Rajagopalan et al. 2004). Other groups have reported mutations in CDC4 in cancers of the bile duct (35%), blood (T-cell acute lymphocytic leukemia 31%), endometrium (9%), pancreas (9%), stomach (15%), lung (3%), bone (2%), and ovary (2%) (Akhoondi et al. 2007, Calhoun et al. 2003, Cassia et al. 2003, Kwak et al. 2005, Lee et al. 2006, Malyukova et al. 2007, Spruck et al. 2002). The overall mutation rate of CDC4 across all cancers examined is 6% (Akhoondi et al. 2007). The high mutation rate of CDC4 in colon cancer and a wide variety of other tumour types prompted further investigation into the biology underlying these mutations.  CDC4/FBXW7 is a member of the F-box family of proteins that act as substrate recognition components for Skp1/Cdc53-Cullin-F-box (SCF) complexes (reviewed in Crusio et al. 2010). SCFs ubiquitinate specific substrates that have been primed by phosphorylation (Skowyra et al. 1997). SCF targets are tagged with covalently bound ubiqiuitin chains that mark them for degradation by the 26S proteasome. CDC4 is the substrate recognition component of the SCFCDC4 (reviewed in DeSalle, Pagano 2001). Most missense CDC4 mutations found in cancers cluster in the substrate binding domain, suggesting aberrant substrate recognition and degradation as a phenotype relevant in tumours (Akhoondi et al. 2007). 117  Known CDC4 degradation targets include Cyclin E, MYC, JUN, Aurora A kinase, Notch, Sp/Kruppel-like factor 5 (KLF5), peroxisome proliferator-activating receptor γ coactivator-1α (PGC-1α), and the sterol regulatory element binding protein (SREBP) (Wei et al. 2005, Tsunematsu et al. 2004, Wu et al. 1998, Olson et al. 2008, Sundqvist et al. 2005, Liu et al. 2010, Mao et al. 2004). Cyclin E, a positive regulator of the Cyclin dependent kinase 2 (CDK2), was one of the first described SCFCDC4 targets. CDK2-Cyclin E is involved in promoting entry into S phase and CDK2 activity levels peak at the G1/S transition (reviewed in Stamatakos et al. 2010). All of the SCFCDC4 targets have roles in promoting cell cycle progression and their degradation by the proteasome removes their signals (reviewed in Tan, Sangfelt & Spruck 2008). It is critical for CDC4 degradation targets to be properly controlled as many have oncogenic potential. For example, elevated Cyclin E levels have been observed in multiple tumour types and this is associated with poor prognosis (Cassia et al. 2003, Donnellan & Chetty 1999). The oncogenic potential of the many SCFCDC4 targets may in part explain why mutations in the CDC4 are so common in a wide variety of tumours.  CDC4-/- human cells have a SCC defect and CDC4 has recently been implicated in controlling SCC establishment (Barber et al. 2008, Lyons, Morgan 2011). Establishment requires the function of Eco1 in yeast and is normally restricted to S phase. Lyons et al. demonstrated that Eco1 is phosphorylated and degraded at the end of S phase by the proteasome and that degradation is dependent on Cdc4/CDC4. Based on this finding and the fact that mutations in cohesin genes (8%) and CDC4 (9%) are prevalent in colon  118  tumours (Barber et al. 2008, Wang et al. 2004, Kemp et al. 2005), it is possible that the cohesin-associated role of CDC4 is relevant to colon tumour progression. To assess this, the spectrum of S. cerevisiae CDC4 genetic interactions and whether they overlapped with the validated cohesin genetic interactions was investigated. Additionally, since the SCFCDC4 is involved in the degradation of many cellular targets the consequences of CDC4 mutations in cancer progression are complex and therefore poorly understood. As shown in chapter 3, genetic interactions can uncover important biological insights regarding the cellular consequences of altered gene function. To this end the CDC4 genetic interactions in yeast and C. elegans were investigated.  4.1.1  CDC4 in S. cerevisiae and C. elegans  S. cerevisiae Cdc4 has two yeast specific degradation targets, Sic1 and Far1 (Henchoz et al. 1997, Feldman et al. 1997). Sic1 is an inhibitor of the Cdc28 kinase that controls progression from G1 into S phase (Schwob et al. 1994) and therefore most alleles of S. cerevisiae CDC4 result in a predominantly G1/S phase arrest at the restrictive temperature (Goh, Surana 1999). This arrest phenotype makes it difficult to study the effects of defective degradation of other Cdc4 targets in S. cerevisiae and complicates conclusions drawn for higher organisms where Sic1 is not a target of Cdc4. Far1 functions as part of the yeast pheromone response (Butty et al. 1998) and is not conserved in higher organisms.  In C. elegans, sel-10 has the best sequence match to human CDC4 (Kipreos, Gohel & Hedgecock 2000) and was first identified in a screen for suppressors of partial loss of function lin-12/NOTCH alleles (Sundaram, Greenwald 1993). SEL-10 has subsequently 119  been shown to bind and negatively regulate LIN-12 activity (Wu et al. 1998, Hubbard et al. 1997). sel-10 mutants have a masculinizing phenotype because proteins that promote male development, such as FEM-1 and FEM-3, are not properly degraded (Jager et al. 2004). Notch signaling is known to be involved in cell fate decisions and work in C. elegans has detailed the contribution of LIN-12 levels on the anchor cell (AC)/ Ventral uterine precursor (VU) cell fate decision (Kimble, Hirsh 1979). Another closely related CDC4-like gene, lin-23, is known to have roles in restricting cell proliferation and is most similar S. cerevisiae MET30 and human β-TRCP based on amino acid sequence (Kipreos, Gohel & Hedgecock 2000).  A CDC4 genetic interaction profile was elucidated for comparison with the cohesin genetic interaction profiles. There was little overlap between the cohesin allele genetic interactions and the cdc4-10 allele interactions suggesting that this CDC4 mutation has other major phenotypic consequences that are not the result of defective cohesin. Phenotypic analysis of C. elegans lin-23 supported a role for LIN-23 in regulating Cyclin E levels. The importance of LIN-23 was demonstrated as lin-23 mutants were found to have a defective DNA damage checkpoint. In addition, CDC4-/- human cells were found to have an abnormal response to DNA damage and became polyploid when exposed to a DNA alkylating agent, suggesting that the loss of cell cycle control is conserved. This work sheds light on the consequences of altered CDC4 function and the implications for maintaining genome integrity.  120  4.2  Materials and methods  4.2.1  cdc4-10 SGA  The SGA screen was performed as detailed in Chapter 2. The cdc4-10 query gene was screened against the DMA, ts, and Damp collections. The data was filtered using the same criteria as the cohesin query mutant screens where hits were considered significant if they had a p-value less than .05 and an E-C value less than -0.3. Enriched MIPS functional categories were identified using the online web based cluster interpreter FunSpec (http://funspec.med.utoronto.ca/cgi-bin/funspec).  4.2.2  C. elegans general strains and methods  Strains used can be found in table 4.1. Seam cell analysis was done as detailed in chapter 2. For DAPI (4',6-diamidino-2-phenylindole) staining young adults were stained for one hour at room temperature, protected from light in DAPI (100ng/ml) in 95% ethanol. Worms were destained in M9 buffer overnight at 4oC.  Table 4.1  C. elegans strains  Strain Name VC2010 CB3514 MT2244 GS307 JR667 KR4670 KR4690  Genotype Wild Type lin-23(e1883)/dpy-10(e128) II sel-10(n1077) V dpy-5(e61), cye-1(ar95)/unc-13(e51) I unc-119(e2498::Tc1) III; wIs51[SCM::GFP, unc-119(+)] lin-23(e1883)/dpy-10(e128) II; unc-119(e2498::Tc1) III; wIs51[SCM::GFP, unc-119(+)] dpy-5(e61), cye-1(ar95)/unc-13(e51) I; unc-119(e2498::Tc1) III; wIs51[SCM::GFP, unc-119(+)]  Mutation W450>stop G565>E W231>stop  121  4.2.3  C. elegans MMS treatment  MMS plates were made by overlaying 200ul of a 1M MMS stock onto 8ml solid agar plates pre-seeded with E. coli OP50 for a final plate concentration of 0.2mM MMS. Plates were left to dry over night. lin-23(e1883)/dpy-10(e128) heterozygote worms were bleached onto MMS containing plates and DAPI stained after 3-5 days of chronic exposure.  4.2.4  Mammalian cell culture  Cells were cultured as detailed in Chapter 3. Cells were prepared for FACS analysis by fixing in ethanol overnight at 4oC. Cells were resuspended in PBS containing propidium iodide (.04mg/ml) and Rnase (1mg/ml) for 30 min at 37oC and resuspended in PBS. FACS profiles were analyzed using FlowJoTM software. Micronuclei were counted from interphase cell spreads stained with DAPI. The Cyclin E and alpha tubulin antibody were purchased from Abcam.  4.3 4.3.1  Results S. cerevisiae CDC4 SGA genetic interactions had little overlap with cohesin  interactions Overlapping cohesin SGA genetic interaction profiles identified 33 genes that interacted with at least two of the three cohesin mutant query genes. Once this set had been validated the strongest SL interactions were with replication mediators. To determine if CDC4 had a similar genetic interaction profile as cohesin mutants, an SGA screen using a hypomorphic allele of CDC4, cdc4-10, was performed. CDC4 is an essential gene in yeast and most hypomorphic alleles, such as cdc4-1, have a G1 arrest phenotype. cdc4122  10, however, has a mixed G1/S and G2/M arrest (Goh, Surana 1999). Since human CDC4-/- cells have a sister chromatid cohesion (SCC) defect (Barber et al. 2008) and the degradation of the SCC establishment factor Eco1 was shown to be dependent on S. cerevisiae Cdc4 (Lyons, Morgan 2011), it was hypothesized that the cdc4-10 mutant allele might have a similar genetic interaction profile to the cohesin mutant alleles.  Analysis of the cdc4-10 SGA data showed that, as expected, many genes associated with ubiquitin and proteasome degradation were identified. Table 4.2 shows the top categories of MIPS functional classification that were enriched in this screen. Cell cycle control and spindle function were also among the top enriched categories. When the 33 genes that were validated in the cohesin screen were compared to the CDC4 interactions, which comprised over 1500 genes, only eleven hits were found in common. These eleven genes were mainly comprised of genes whose encoded proteins function as components of the microtubule or prefoldin apparatus or general cellular metabolism genes, such as ribosomal components. Of note, none of the replication mediators, which were the strongest interactors with the cohesin mutants, were found in this set. From this data it appeared that in budding yeast the cdc4-10 genetic interaction profile had a different signature than that of cohesins.  123  Table 4.2  Enriched MIPS categories in the CDC4 SGA screen  MIPS Functional Classification Category  p-value*  Proteasomal degradation (ubiquitin/proteasomal pathway) [14.13.01.01] Mitotic cell cycle and cell cycle control [10.03.01]  3.314e-11  Spindle pole body/centrosome and microtubule cycle [10.03.05.01] Cell growth / morphogenesis [40.01]  3.427e-07  Protein processing (proteolytic) [14.07.11]  6.248e-06  rRNA synthesis [11.02.01]  7.763e-06  rRNA processing [11.04.01]  4.732e-05  Organization of chromosome structur [42.10.03]  7.969e-05  2.908e-07  2.863e-06  *P-value cut off 1e-05 MIPS: Munich information center for protein sequences  124  Figure 4.1  cdc4-10 SGA data only partially overlaps with cohesin SGA data  TUB4  RPS31  Replication Mediators  KAR3 RPS16B  Chromatin BIM1  TRM112  CDC4  STU2  mRNA Processing Spindle/Kinetochore Cohesin (subunit or loader)  YPR1  Protein/Cellular Metabolism BUB3 RNA15  Microtubule/Prefoldin  PCF11  The eleven hits that were common between the CDC4 SGA screen and the set of cohesin validated interactions are shown. Significant cdc4-10 SGA interactions were considered as those that had a p-value < .05 and an interaction magnitude (E-C) value < -0.3, the same criteria as was used to filter the cohesion SGA data in chapter 3. Common interactors are colored according to function.  4.3.2  C. elegans has two CDC4 like genes  In yeast Cdc4 has additional degradation targets, namely Sic1 and Far1, that are not conserved in more complex eukaryotes. To evaluate the potential role of CDC4 mutations in tumour development in this thesis, C. elegans, a eukaryote with CDC4 activities more akin to humans was used. sel-10 is the most similar gene to human and yeast CDC4 based on sequence analysis. There is, however, another closely related CDC4-like gene, lin-23 (Kipreos, Gohel & Hedgecock 2000). lin-23 is the second closest match to S. cerevisiae CDC4 and human FBXW7 but has a reciprocal best blast match with S. cerevisiae MET30 and human β-TRCP (Table 4.1). Met30 is also an F-box protein but has a role in sulfur metabolism and methionine biosynthesis. sel-10 mutants appear  125  grossly wild type although they are characterized by a masculinizing phenotype where some male specific neurons are not removed in hermaphrodites (Jager et al. 2004). lin-23 mutants show hyperplasia of the embryonic intestinal lineage and over-proliferation of a number of post embryonic tissues (Kipreos, Gohel & Hedgecock 2000). lin-23 null mutant analysis found additional seam cells, neuroblasts and ventral hypodermoblasts, suggesting a general loss of cell proliferation control. Additional gonad arms were also observed due to additional distal tip and gonadoblast cells (Kipreos, Gohel & Hedgecock 2000). lin-23; sel-10 double mutant analysis suggested that these two genes have no significant functional redundancy (Kipreos, Gohel & Hedgecock 2000). Aside from phenotypic characterization of lin-23 mutant alleles, relatively little is known about lin-23 function in C. elegans. Given the interesting over-proliferation phenotype of lin-23 mutants, the cell cycle defect in lin-23 defective animals was further investigated.  Table 4.3  BLAST results for CDC4-like genes in C. elegans and H. sapiens  Gene (organism) sel-10 (C. elegans) lin-23 (C. elegans) FBXW7 (H. sapiens) β-TRCP2 (H. sapiens)  4.3.3  Best Blast Match FBXW7 β-TRCP2 sel-10 lin-23  Expect value 9e-130 0.0 9e-130 0  cye-1/Cyclin E RNAi improves viability of lin-23 mutants  Cyclin E is a key driver of the cell cycle transition from G1 to S phase. Cyclin E levels are also elevated in certain tumour types (Cassia et al. 2003, Ekholm-Reed et al. 2004). Although sel-10 is more similar to CDC4 in terms of sequence, it was possible that LIN23, and not SEL-10, controls Cyclin E levels. This is supported by the fact that sel-10  126  mutants show no cell cycle defects, which may be expected in a mutant where cyclin E levels are improperly regulated. lin-23 null mutants, on the other hand, have a variety of cell cycle defects, are homozygous lethal and have a high percentage of Pvl. Previous work in this thesis demonstrated that Pvl is characteristic of mutants with defects in somatic cell proliferation.  If LIN-23 is involved in degrading CYE-1/Cyclin E then down regulating cye-1/Cyclin E by RNAi should reverse some of the defects seen in lin-23 null mutants. lin-23(e1883) worms are normally thin and uncoordinated and their gonads contain an abundance of mitotic stage cells that have failed to enter meiosis (Figure 4.2). lin-23;cye-1(RNAi) worms appeared healthier and even produced embryos (20/20 worms) some of which hatched and arrested as larvae (1/20). sel-10 mutants appeared grossly WT and sel10;cye-1(RNAi) worms appeared more sickly as seen by gonadal defects and an increased incidence of Pvl but were indistinguishable from cye-1(RNAi) animals (Figure 4.2).  127  Figure 4.2  Phenotypes of lin-23(e1883) and sel-10 mutant worms  lin-23  cye-1(RNAi)  untreated  sel-10  *  Representative images of DAPI stained, young adult worms. sel-10 mutants worms produce embryos and viable progeny and have a low incidence of Pvl. sel-10; cye-1(RNAi) worms have a variety of defects including Pvl (denoted with an asterisk) and brightly staining DAPI spots (denoted with an arrow head). lin-23 mutants have germ lines devoid of embryos that are packed with mitotic cells (no meiotic marks, such as crescent nuclei were visible) along its length. lin-23; cye-1(RNAi) worms have gonads that appeared more similar to wild type.  128  4.3.4  cye-1(RNAi) reduces seam over proliferation in lin-23 mutants  lin-23 mutants have hyperplasia of a number of embryonic and post embryonic lineages (Kipreos, Gohel & Hedgecock 2000). In order to quantitatively address whether reducing CYE-1 levels can alter lin-23 over-proliferation, seam cell numbers in young adults were analyzed. lin-23(e1883) homozygous worms have excess seam cells (average of 23 in lin-23 vs 15 in WT worms) as was previously reported (Kipreos, Gohel & Hedgecock 2000) (Figure 4.3). cye-1(ar95) truncation mutants are sterile and had under-proliferation of the seam cells (average of 12 in cye-1). This suggested that altered CYE-1 levels, as expected, affect cell proliferation. lin-23; cye-1(RNAi) animals had a reduced number of seam cells as compared to the lin-23 mutant alone. This demonstrated that the lin-23 phenotype is dependent on the levels of CYE-1. These results raised the possibility that LIN-23 is the F-box protein involved in Cyclin E degradation in C. elegans.  Figure 4.3  A)  lin-23 seam cell hyperplasia is partially rescued in lin-23, cye-1(RNAi)  25  B) cye-1(ar95);dpy-5(e61)/unc-13(e51)  12  # of seam cells  20 15 C) lin-23(e1883)/dpy-10(e128) 10 22  5  ) li cy n-­‐2 e-­‐ 3; 1( RN Ai  3 -­‐2 lin  e-­‐ 1 cy  N2  0  A) Seam cell counts per lateral syncytium. n ranges from 30 to 35 in each trial. Error bars represent SEM. B and C) Images of worms of the indicated genotype carrying the SCM::GFP marker. The number on the left of each image is the number of seam cells. 129  4.3.5  lin-23 worms respond inappropriately to MMS induced DNA damage  Cyclin E levels drive progression into S phase. Given that CYE-1 levels may be elevated in lin-23 mutants, these animals may be unable to impose cell cycle checkpoints appropriately. To investigate this, question lin-23 mutants were chronically exposed to the alkylating agent methyl methanesulfonate (MMS). MMS lesions, if not removed, can impair replication progression and this leads to checkpoint activation (reviewed in Branzei, Foiani 2009). Worms that were heterozygous for wild type lin-23 showed arrested germline development and restricted proliferation in other post embryonic lineages, such as the seam (Figure 4.4 A). lin-23 homozygotes appeared unresponsive to MMS and had numerous germ nuclei that had continued to divide even in the presence of MMS (Figure 4.4 B).  130  Figure 4.4  lin-23 mutants continue to proliferate in the presence of DNA damage  A.)  lin-23 heterozygous  B.)  lin-23 homozygous  Representative images of A) lin-23(e1883)/dpy-10(e128) heterozygous and B) homozygous worms were chronically treated with .2mM MMS and stained with DAPI. 4.3.6  hCDC4-/- human cells show genomic instability in the presence of DNA damage  To determine whether human cells lacking CDC4 also responded inappropriately to DNA damage induced by MMS exposure, heterozygous and homozygous CDC4 deletion cell lines were treated with MMS. In the HCT116 CDC4 +/- and -/- lines, Cyclin E regulation is affected as both lines have elevated levels of Cyclin E relative to the CDC4 proficient line (Figure 4.5). Cells were chronically treated with .15mM MMS for a period of 48 hrs and fluoresence activated cell sorting (FACS) was used to monitor the cell cycle profile.  131  CDC4-/- cells have elevated Cyclin E  CDC4-/-  CDC4+/-  CDC4+/+  CDC4-/-  CDC4+/-  .15mM MMS .25mM MMS  CDC4-/-  CDC4+/-  CDC4+/+  asynch  CDC4+/+  Figure 4.5  Cyclin E  α Tubulin  Protein samples were collected from asynchronously growing cells, normalized and then probed with antibodies against Cyclin E and alpha Tubulin. Asynchronous CDC4-/- cells had a slight accumulation of cells with a 4N DNA content, suggestive of a cell cycle defect (Figure 4.6). Cells lacking CDC4 accumulated a large aneuploid population of cells (24.9% of total population) by 48 hours when exposed to .15mM MMS, whereas CDC4 proficient cells cycled between 2N and 4N DNA content and accumulated a smaller (6.6%) population of aneuploid cells. This demonstrated that CDC4 is required to maintain genomic integrity in the continual presence of a DNA damaging agent.  132  Figure 4.6  Human CDC4 prevents aneuploidy after genotoxic stress  .15mM MMS  .25mM MMS  6.61  86.1  13.9  95.9  4.06  87.8  12.2  75.1  24.9  80.0  20.0  # Cells  93.4  CDC4-/-  CDC4+/-  CDC4+/+  asynchronous  FL3-H  Cells were treated with 0.15mM MMS for 48 hours and analyzed by FACS. Profiles are gated to identify the percentage of cells with a greater than 4N DNA content.  To look more closely at the phenotype of CDC4-/- cells, the frequency of micronuclei was examined. The percentage of CDC4-/- cells with micronuclei was increased as compared to CDC4 wild type cells, a phenotype previously observed by Rajagopalan et al. 2004.  133  Treatment with MMS increased the frequency of cells with micronuclei in all cell lines examined. The fold increase was similar in all cases (Figure 4.7). This suggested that loss of CDC4 does not dramatically affect small-scale changes in nuclear DNA content.  Incidence of Micronuclei (%)  Figure 4.7  Micronulei in CDC4-/- cells treated with MMS  18 16  untreated  14  .15mM MMS  12 10 8  2.5x 4.17x  3.08x  6 4 2 0  CDC4+/+  CDC4+/-  CDC4-/-  n ranges from 210 to 350 interphase nuclei for each sample.  4.4 4.4.1  Discussion The spectrum of CDC4 genetic interactions is distinct from that of cohesin  Sequencing of CIN genes in colon tumours found that mutations in genes involved in SCC are relatively common. The recent finding that Cdc4 also has a role in regulating cohesion establishment in yeast raised the possibility that CDC4 and cohesin mutations in colon tumours may share a common disease progression mechanism. In other words, the major defect found in Cdc4/CDC4 mutant cells may be akin to the defects seen in cohesin mutant cells. To investigate this possibility, an SGA screen was performed on a temperature sensitive allele of CDC4. The genetic interactions that overlapped between the CDC4 and 134  cohesin data identified eleven genes. A common finding in many SGA screens is that they identify non-specific interactions with house keeping metabolism genes. These genetic interactions are generally attributed to an inability to deal with general cell stress, leading to general fitness defects. SGA screens done in the Hieter lab that have used query genes with a variety of functions have identified between 15 and 25% of genetic interactions with genes involved in cellular metabolism. This suggested that, for example, mutations in ribosomal genes are not a specific interaction and may interact with mutations in many other genes. Indeed, four of the eleven were classified as general cell metabolism genes and included a subunit of tRNA methyltransferase (TRM112), an aldo-keto reductase (YPR1), and two small ribosomal subunits (RPS16B and RPS31). Interestingly, Rps31 is a fusion protein that is cleaved to generate a ribosomal component and ubiquitin (Ozkaynak et al. 1987). Many genes involved in the ubiquitin cycle were identified in this screen, as expected. Of note, none of the replication mediators, the strongest class of genetic interactors with cohesin as a whole, interacted with CDC4.  Lyons and Morgan found that the degradation of Eco1 depended on Cdc4. They also showed that when Eco1 is over expressed, cells are viable and there is no gross defect in chromosome segregation. A slight timing defect in the ability of sister chromatids to separate synchronously was observed but the consequence of this defect, at this time, appears minimal. The accumulation of other Cdc4 targets, such as Cyclin E, appear to have greater phenotypic consequences than accumulation of Eco1.  135  4.4.2  C. elegans as a model to study CDC4 and Cyclin E regulation  This thesis work has shown that the over proliferative phenotype of lin-23 mutants in C. elegans depends on the levels of CYE-1/Cyclin E. This raises the possibility that LIN-23 is the F-box protein that is involved in degrading Cyclin E. If this is the case, then C. elegans has at least two CDC4 orthologs, each of which takes on different aspects of CDC4’s function in higher organisms. Further work will be required to determine whether CYE-1 levels are elevated in lin-23 mutants and whether CYE-1 ubiquitination directly depends on LIN-23.  The apparent splitting of CDC4 function makes C. elegans an ideal model to study the loss of degradation control of specific targets. In terms of CYE-1 control, the lin-23 mutant may enable specific phenotypes of over abundant cyclin E to be identified. It was previously identified that Cyclin E is a positive cell cycle regulator and the phenotype of the lin-23 mutant suggested that increased levels of Cyclin E result in increased and inappropriate cell divisions. C. elegans will be an interesting model to identify additional regulators of Cyclin E as reduced CYE-1 levels in the lin-23 mutant dramatically increase viability. Screens based on the presence of embryos, which are normally lacking in lin-23 mutants, could be used as a readout to identify suppressor mutations. A reduced number of seam cells may also be a convenient readout for reduced proliferation. Both of these phenotypes are theoretically amenable to high throughput screening using an embryonic or seam cell GFP marker and a COPASTM Flow-Sorting system that can detect fluorescence along the length of the worm (Pulak 2006).  136  4.4.3  CDC4 loss of function and the DDR  Since lin-23 mutants have excess cell divisions, one question that was raised was whether cells can arrest the cell cycle in response to DNA damage. In C. elegans it was found that when worms were chronically treated with MMS, gonad and somatic cells continued to proliferate. In heterozygous lin-23 mutants, on the other hand, cell division is arrested and gonads do not develop. This suggested that loss of LIN-23 impairs the normal DNA damage response and that, at least with respect to this phenotype, this mutant is recessive. It will be interesting to know whether this defect can be rescued by down regulating CYE-1 in lin-23 mutants, as was done when seam cells were counted in the absence of any DNA damaging agaent.  An impaired response to MMS is also evident in human CDC4 deficient cells. FACS analysis demonstrated that CDC4-/- cell populations accumulated a large fraction of aneuploid cells after chronic MMS exposure. It is currently not known whether these aneuploid cells are viable or whether they have additional chromosomal defects. It will be interesting to know whether over-expression of Cyclin E phenocopies loss of CDC4 in this respect. This data suggests that CDC4 has a role in enabling the cell to respond appropriately to DNA damage in C. elegans and human cells. Additional questions remain regarding this defective DDR and which checkpoint components are compromised.  This thesis work has implications with respect to cancers that harbour mutations in CDC4 and potentially those that over express Cyclin E. Many cancer therapies include agents that induce DNA damage. If these therapies are administered to patients with CDC4 defects, this  137  may induce rapid tumour evolution and the emergence of resistant clones. A more effective therapeutic intervention for patients with CDC4 mutations may be the administration of therapies to down regulate CDC4 degradation targets as compared to more traditional therapies  138  Chapter 5: Conclusion Cancer is a polygenic disease caused by dysfunction in a number of different processes, including cell cycle control, DNA repair and execution of apoptosis, chromosome segregation, and maintenance of genome stability. As precursor cancerous cells acquire functional genetic lesions they develop more of the phenotypes that are characteristic of cancer cells, a theory known as the multi-hit hypothesis (reviewed in Stratton 2011). Mutations in a definable set of genes account for these cancer phenotypes. On-going tumour sequencing and work in model systems are striving to identify the spectrum of mutations that lead to tumour development and their associated phenotypes. These data will provide a starting point to understand the role of causative mutations in disease progression. How these mutations impact normal cellular processes and also how they interact with other disease mutations is of utmost importance.  In this thesis a cross species screening and validation approach to identify genetic interactions relevant to cancer was developed. This work found that the genetic interaction profiles of cohesin and CDC4, the two main groups of mutations found in colon tumours to date (Barber et al. 2008, Wang et al. 2004, Rajagopalan et al. 2004), were distinct in S. cerevisiae. This supports the hypothesis that cohesin and CDC4 mutations lead to disease via two different mechanisms. Cohesin mutations were found to sensitize the cell to the absence of proteins that mediate replication fork progression while CDC4 appears to be important for the DNA damage response.  139  5.1  C. elegans as a model to identify cancer relevant SL interactions  Most of the genetic interaction data generated by the research community thus far has come from studies in yeast (Tong et al. 2001, Schuldiner et al. 2005, Collins et al. 2007, Costanzo et al. 2010). This is a consequence of the existence of comprehensive sets of yeast mutants and the fact that yeast is very amenable to high throughput genetic screening. Although screening for genetic interactions in yeast has many advantages, there are a number of drawbacks when investigating cancer associated genes and tumour development. The animal model C. elegans can overcome some of these limitations and has two main advantages. The first is that, given their greater complexity by comparison to yeast, nematodes have a much larger genome, in fact akin in terms of gene number to that of humans (reviewed in Hodgkin, Plasterk & Waterston 1995). Gene families found in humans that are not conserved in yeast are often found in C. elegans. For example, the PARP family of genes is not found in the yeast genome but is represented by five genes in C. elegans; pme-1 through 5 (Gagnon, Hengartner & Desnoyers 2002, St-Laurent et al. 2007, Gravel et al. 2004). The breast cancer susceptibility genes, BRCA1 and BRCA2, often found mutated in cases of familial breast cancer, also have orthologues in C. elegans, brc-1 and brc-2, respectively (Boulton et al. 2004), but not in yeast. C. elegans offers an animal model to study genetic interactions in a more complicated system that is still genetically tractable and amenable to phenotypic screening. The second major advantage in using C. elegans for genetic studies is the existence of a broad range of phenotypes that are characteristic of defects in different processes. Checkpoint and DNA damage response defects are often characterized by abnormal gonad architecture and proliferation. Enlarged nuclei in the mitotic tip are indicative of replicative stress and increased apoptosis is indicative of DNA damage  140  (reviewed in Kirienko et al. 2010). Aberrant cell lineage numbers that give rise to phenotypes such as a protruding vulva and seam cell defects suggest problems arising in cell proliferation control (Seydoux, Savage & Greenwald 1993). These phenotypes can be exploited as end point readouts for both primary and secondary genetic interaction screens. Chapter 2 described the development of an assay to query genetic interactions affecting cell proliferation in the somatic cells of C. elegans.  Previous groups have developed C. elegans genetic interaction screens that used animal viability as the phenotypic read out (Byrne et al. 2007, Lehner et al. 2006, Tischler et al. 2006). As mentioned in chapter 2, this experimental design makes it challenging to study genes that are essential for embryonic viability. Many genes that incur mutations in cancers, such as CDC4, are essential for proper embryonic viability and development. Some cancer mutations occur in the germline, as in familial cancers, but a vast number occur somatically. The processes these somatic mutations compromise are therefore intact during embryogenesis. Defects that are not tolerated in embryonic cells are tolerated in somatic cells and do not compromise organismal viability. By screening in the somatic cell lineages of C. elegans, the developmental and viability constraints that occur in embryonic cells can be bypassed. The Pvl assay for genetic interactions allows screening in a cell environment where the cellular pressures that exist in tumour cells can be more closely mimicked.  The Pvl screen, while effective, is at this point too labour intensive for extremely large genetic interaction screens. It is amenable to screening in numbers that are in the ballpark of hundreds, as opposed to thousands as would be needed to query the entire C. elegans  141  genome. It may be possible to use cell type specific GFP reporters, such as seam or vulval lineage GFP markers, and automated or semi-automated screening to scale the vulval development assay to high throughput levels.  5.2  Challenges in interpreting data from large scale genetic interaction screens  In chapter 3 the Pvl assay was used to identify genetic interactions with SMC1 that were conserved between yeast and worm. The motivation for this work was to better understand the effects of cohesin mutations on cell viability and to leverage this information to identify potential chemotherapeutic targets. The starting point for this genetic interaction data was three independent SGA screens that used cohesin mutants as query genes. One problem with large scale genetic interaction screens such as SGA is the “signal” to “noise” ratio. How can bona fide interactions be extracted from a screen with a large number of hits, many of which are false positives? This work illustrated the usefulness of using the overlapping interaction space of three large preliminary data sets to enrich for robust interactions. Each screen, on its own, identified hundreds of genetic interactions in a wide variety of cellular pathways and processes. These datasets are difficult to manage and it is challenging to extract biologically meaningful conclusions if they are not somehow filtered.  One approach to filtering large scale interaction data is to use a variety of secondary assays. These additional screens can help identify true interactions and provide a more accurate sense of the strength of the interaction. Depending on the assay design, they can also sometimes help to categorize interactions by highlighting those that result in a specific phenotype, such as increased chromosome missegregation. In this work, by validating the SGA data using independent measures of fitness, random spore and growth curve analysis, 142  the quality of the data set was increased by enriching for true positives. Partitioning data based on whether interactions were SL or SS extracted an additional laer of information. This screening and analysis uncovered the fact that cohesin mutants are SL with replication fork mediators. Secondary screens of the interaction data in C. elegans were, on the other hand, designed to answer questions regarding the mechanisms of cell death in double mutants. This screen queried the levels of apoptosis in the germline and suggested that cohesin mutants depleted of fork mediators had increased DNA damage. This finding reinforced the interaction between cohesin and fork mediators and provided additional insight into the underlying biology.  A different approach to processing large scale genetic interaction data that was not applied in this thesis is to look at the pattern of genetic interactions through hierarchical clustering (HC). In HC the interaction profiles of many different query genes are compared to each other. Queries that have similar profiles are clustered in relatively close proximity to each other and this can suggest similar function (Collins et al. 2007, Costanzo et al. 2010). This approach can be used to assign gene function to un-annotated genes or to uncover new functions of known genes. In the context of identifying strong SL interactions that may be useful in treating cancers, an overlay of genetic interaction data has major advantages over other types of secondary screens or profiling by genetic interaction signatures. Interactions found by this filtering method are more likely to be effective in treating cancers with a defect in cohesin in human cells, for example, as opposed to being specific to the query gene or even the query mutation used. Genes that are SL with multiple components of the same process will identify more general (non-specific) interactions with cohesin defects, which are  143  likely to be conserved in other organisms, even given the sequence differences that exist between orthologous genes. Filtering for interactions with mutations in different cancers also translates into therapeutics with a broader spectrum of usefulness. Ideally, a cancer therapeutic will be effective in treating many different types of cancers with varying genetic backgrounds.  5.3  Conservation of genetic interactions  The work described in this thesis also serves as a model for how yeast genetic interaction screens can be used to inform genetic interaction testing in more complex model systems, including human cells. Predictions are not restricted to genes that are conserved, as demonstrated by the forecasted interactions identified here between the pme/PARP and cohesin genes in both C. elegans and human cells. Although many genes in humans and C. elegans are not found in yeast, a multitude of processes, such as mediation of replication fork progression, are conserved. Identification of the cellular processes that are compromised and those that are essential in the context of a query mutation is often sufficient to inform testing in more complex organisms.  Whether genetic interactions are conserved between distantly related species has been contentious. The first large scale screens in C. elegans suggested that a very small percentage of negative interactions, ranging from 0.5-1% are conserved (Byrne et al. 2007, Lehner et al. 2006). Conversely, smaller more focused interaction screens have found that a large proportion of negative interactions, between 17 and 43% depending on the organism, are conserved (Tarailo, Tarailo, Rose 2007, Roguev et al. 2008). Indeed, work presented in this  144  thesis suggested that an even higher percentage of conservation exists, on the order of 80%, at least for cohesin interactions.  The discrepancy in the degree of conservation between experimental data may be partially attributed to experimental design. Large scale C. elegans screens have previously relied on gene knockdown by RNAi in liquid and there is some controversy as to the effectiveness of this gene knockdown method on a large scale. The presence or absence of a genetic interaction also depends on the phenotype observed. First generation C. elegans screens have relied on viability as a readout (Byrne et al. 2007, Lehner et al. 2006, Tischler et al. 2006), which may not be appropriate for all gene pair interactions. Other more subtle phenotypes may be required to observe some of these interactions. Additionally, many genes that exist singly in yeast exist as gene families or have paralogs in the nematode (reviewed in Copley et al. 1999). Paralogs often have subtly different functions but can act redundantly in some situations. Uncovering of these genetic interactions may require small molecules, such as PARP inhibitors that target multiple, related genes, or the creation of double and triple mutants to compromise all paralogous gene functions. Studies that have examined genetic interactions using chemicals in place of gene knockdown, as has been done in yeast (Magtanong et al. 2011, Tamble et al. 2011, Kemmer et al. 2009), would be informative if done systematically in C. elegans. The findings of these types of chemical-genetic screens can often be more quickly translated into therapeutic terms because the initial screening required to identify inhibitors can be bypassed.  145  5.4  Fork mediators in the presence of sub-optimal cohesin  The goal of the genetic interaction networks has been to discover new insights into the biology of cohesin mutations that may be relevant to tumours. By overlaying the data from three SGA screens using three different cohesin mutations as query genes, and filtering on the strongest negative interactions, replication fork mediators were found to be essential in cohesin mutated cells. In S. cerevisiae, hypomorphic mutations in cohesin are SL with the loss of proteins that increase fork velocity and also with proteins that induce fork slowing and stabilization. This implies that the ability of the cell to regulate fork dynamics is critical in the context of sub-optimal cohesin. This dependence is conserved in C. elegans and in human cells, suggesting that general defects in cohesin, as represented by hypomorphic mutations or knockdown by siRNA, exist across species. In C. elegans this interaction was seen by making double mutants and interestingly, defects were seen when him-1/SMC1 mutants were combined with any of the pme/PARP family of mutants. pme-1 and pme-2 are orthologues of PARP1 and PARP2, respectively, which add PAR moieties to a variety of proteins (Gagnon, Hengartner & Desnoyers 2002). pme-3 and pme-4 reverse this modification by removal of these PAR moieties (St-Laurent et al. 2007). pme-5, the orthologue of TANKYRASE, has roles in modulating cohesin at telomeres (Gravel et al. 2004). Although these mutants had slightly different specific defects when combined with him-1(e879), in general, viability defects, as represented by reduced brood sizes or increased embryonic lethality, was seen in all double mutants. This implied that a defect in any aspect of the PARP pathway is not well tolerated when cohesin is sub-optimal.  146  How cohesin defects impact fork progression is not well understood. The findings that PAR modifications occur at stalled forks to signal these events and are required for recruitment of proteins that function in fork restart, such as MRE11, are relatively recent (Bryant et al. 2009). End point measurements of fork velocity and restart capacity using labeled nucleotides and visualization by stretching DNA fibers on glass slides will help to answer how defects in cohesin and PARP affect fork dynamics. It has recently been shown in human cells that a defect in DCC1, a component of the altRFCCTF18, slows fork velocity, decreases replicon size and diminishes restart capacity as measured by this DNA fiber assay (Terret et al. 2009). The same group showed that WAPL does not affect fork progression under normal circumstances but depletion relieves the slowed fork progression that occurs when Eco1/ESCO1 and ESCO2 are co-depleted. Immunofluoresence studies using antibodies for other markers of stalled forks, such as the single strand binding protein RPA, will help researchers to understand the initial and downstream fork associated events that may occur in cohesin mutants. It will also be interesting to know whether modification of SMC1 by ATR, which is required for the intra S phase checkpoint (Bolderson et al. 2004), is somehow linked to cohesin’s role or impact on fork progression. Because of cohesin’s close proximity to replication events it may play a role in sensing and/or signaling the S phase checkpoint response. The finding that sister chromatid cohesion (SCC) is established in conjunction with replication, a process that is not clearly understood, implies that cohesin complexes undergo a change that is catalyzed by the process of replication. Most of the proteins that mediate fork progression, such as Csm3, Dcc1 and Ctf4, have SCC defects, the reason for which is unknown (Mayer et al. 2004). It is possible that the cohesion defects of these proteins stem  147  from their role in replication. Perhaps altered fork progression dynamics alter the efficiency of establishment of SCC.  The interaction between cohesin and PARP genes was also found in human cells and this has direct cancer therapeutic relevance. In this system, the enhanced defect in cohesin-depleted cells was seen with the PARP inhibitor olaparib, a drug undergoing evaluation in phase II clinical trials (Fong et al. 2009, Audeh et al. 2010, Tutt et al. 2010, Gelmon et al. 2011, Khan et al. 2011). The creation of cell lines that are genetically defective in cohesin genes will be useful tools to study not only the interaction with PARP but also the effect of cohesin dysfunction on its own in human cells. Mouse studies will be useful to address whether PARP inhibitors can be effective in altering the course of progression of tumours with genetic lesions in cohesin. If these studies are promising, an inhibitor that has already gone through several optimization stages will be poised to begin clinical trials benefitting a different patient group.  5.5  CDC4 mutations in colon tumours  The research conducted in this thesis was not limited to cohesin mutations but inclusive of many of the CIN genes found mutated in colon tumours (Barber et al. 2008, Wang et al. 2004, Kemp et al. 2005, Rajagopalan et al. 2004). The largest single target for mutations in colon tumours was CDC4. Given that a significant percentage of the panel of colon tumours examined harboured mutations in cohesin genes, it was possible that these two groups of mutations resulted in the same underlying biology. This hypothesis seemed more probable given the recent finding that Cdc4 has a role in regulating the protein levels of the cohesion  148  establishment factor Eco1 (Lyons & Morgan 2011). Comparison of the CDC4 and cohesin genetic networks in S. cerevisiae did not reveal significant overlap in negative interactors and suggested that CDC4 and cohesin mutations may induce distinct mechanisms of tumour development. There are two main caveats to this approach. The first is that the interactions identified in this work may be specific to the cdc4-10 allele and may not reflect all aspects of Cdc4 disruption. The second is that CDC4 function may be divergent in higher organisms that have a greater gene repertoire.  CDC4 dysfunction results in increased levels of its degradation targets, many of which have potential oncogenic properties. For these reasons it has been considered to act as a tumour suppressor and indeed tumours have been identified that have elevated levels of Cyclin E (Cassia et al. 2003, Ekholm-Reed et al. 2004), one of CDC4’s degradation targets. In terms of therapeutic approaches for cancers with CDC4 mutations, the identification of CDC4 SL targets may prove effective. The SGA performed as part of this thesis identified a number of strong negative genetic interactions that should be further investigated. It should be recognized that CDC4 mutations, because of the cellular consequences, are distinct from mutations in other genes, such as cohesins, BUB1 and MRE11, all of which are mutated in colon tumours (Barber et al. 2008, Wang et al. 2004, Kemp et al. 2005). Mutations in these latter genes do not, as far as is currently known, directly and rapidly impact the levels of other proteins. Alternate therapeutic discovery approaches to down regulate CDC4 targets should also be considered. Work in this thesis showed that in C. elegans down regulating Cyclin E in lin-23 mutants rescued several cell over proliferation phenotypes and dramatically increased viability. Promoter control studies of Cyclin E have shown that over  149  expression can induce tumourigenesis in mice. These tumours appear addicted to Cyclin E over expression because promoter shut off caused tumour regression (reviewed in Freemantle & Dmitrovsky 2010).  Preliminary work performed in this thesis in C. elegans and human cells suggested that cells are unable to properly execute the DDR in coordination with cell cycle control. The germ cells of C. elegans lin-23 mutants continued to divide in the presence of MMS, presumably accumulating massive levels of DNA damage. Human cells deficient in CDC4 accumulate large aneuploid populations under MMS treatment over relatively few cell cycles. In order for DNA lesions to be corrected the DDR activates a number of pathways, some of which cause the cell cycle to slow or be delayed (Branzei, Foiani 2009). This alteration in cell cycle timing allows time for lesions to be repaired. It is possible that when CDC4 degradation targets fluctuate outside of a normal range, the cell is not able to effectively down regulate these positive cell cycle regulators when their signals need to be removed. In both C. elegans and human cells, elevated Cyclin E is correlated with inappropriate cell division when genome integrity is compromised. It will be interesting to know whether elevated Cyclin E levels affect the DDR, and in particular whether the timing and length of the cell cycle can be shifted to allow for lesion repair.  5.6  Clinical implications  The era of full genome sequencing is relatively recent and researchers are still developing approaches to understand the massive amount of data that comes from this technology (reviewed in Stratton 2011). As genome sequencing technology advances and our  150  understanding and analysis methods of genome data improve, it will be increasingly common for health care practitioners to sequence patient tumours and to use this information to aid in directing therapy. From a research standpoint, many more tumour samples will need to be sequenced to have a better understanding of the types of mutational signatures that can exist in various tumour types and at various different stages of tumour development. This data will need to be interfaced with the types of analysis performed in this thesis to inform therapeutic target selection for new drug development or to identify new uses for existing therapeutics. If a more informed approach for choosing a particular therapeutic option can be implemented that is based on a better understanding of a patient’s individual disease, patient outcome and survival should improve significantly.  The work described in this thesis has demonstrated how model organism genetics, and in particular the yeast/worm/human tissue culture axis, can be used to investigate tumour sequence information. In theory, this approach can be used to investigate any number of genes that are mutated in cancer to uncover the biological consequences of these mutations and to identify potential therapeutic targets. Many strong cohesin SL interactions were identified that could be further investigated in more complex model systems such as human cells or mouse models. The finding that PARP inhibitors may be effective in treating a much larger sub group of patients, those with mutations in cohesin genes, has direct clinical therapeutic implications. PARP inhibitors have already been heavily developed by a number of pharmaceutical companies for the treatment of breast and ovarian cancer (reviewed in Calvert & Azariti 2011). Therefore, the time frame to complete mouse model studies to establish whether it is reasonable to attempt to repurpose PARP inhibitors for a different  151  cancer treatment application could occur relatively quick. This in turn should translate into a relatively short time frame before PARP inhibitors could be evaluated in clinical trials for the treatment of patients with tumours harbouring mutations in cohesin.  152  References Adjei, A.A. 2001, "Blocking oncogenic Ras signaling for cancer therapy", Journal of the National Cancer Institute, vol. 93, no. 14, pp. 1062-1074. Akhoondi, S., Sun, D., von der Lehr, N., Apostolidou, S., Klotz, K., Maljukova, A., Cepeda, D., Fiegl, H., Dafou, D., Marth, C., Mueller-Holzner, E., Corcoran, M., Dagnell, M., Nejad, S.Z., Nayer, B.N., Zali, M.R., Hansson, J., Egyhazi, S., Petersson, F., Sangfelt, P., Nordgren, H., Grander, D., Reed, S.I., Widschwendter, M., Sangfelt, O. & Spruck, C. 2007, "FBXW7/hCDC4 is a general tumor suppressor in human cancer", Cancer research, vol. 67, no. 19, pp. 9006-9012. Arnaudeau, C., Lundin, C. & Helleday, T. 2001, "DNA double-strand breaks associated with replication forks are predominantly repaired by homologous recombination involving an exchange mechanism in mammalian cells", Journal of Molecular Biology, vol. 307, no. 5, pp. 1235-1245. Audeh, M.W., Carmichael, J., Penson, R.T., Friedlander, M., Powell, B., Bell-McGuinn, K.M., Scott, C., Weitzel, J.N., Oaknin, A., Loman, N., Lu, K., Schmutzler, R.K., Matulonis, U., Wickens, M. & Tutt, A. 2010, "Oral poly(ADP-ribose) polymerase inhibitor olaparib in patients with BRCA1 or BRCA2 mutations and recurrent ovarian cancer: a proof-of-concept trial", Lancet, vol. 376, no. 9737, pp. 245-251. Baetz, K., McHardy, L., Gable, K., Tarling, T., Reberioux, D., Bryan, J., Andersen, R.J., Dunn, T., Hieter, P. & Roberge, M. 2004, "Yeast genome-wide drug-induced haploinsufficiency screen to determine drug mode of action", Proceedings of the National Academy of Sciences of the United States of America, vol. 101, no. 13, pp. 4525-4530. Bando, M., Katou, Y., Komata, M., Tanaka, H., Itoh, T., Sutani, T. & Shirahige, K. 2009, "Csm3, Tof1, and Mrc1 form a heterotrimeric mediator complex that associates with DNA replication forks", The Journal of biological chemistry, vol. 284, no. 49, pp. 34355-34365. Bandyopadhyay, S., Mehta, M., Kuo, D., Sung, M.K., Chuang, R., Jaehnig, E.J., Bodenmiller, B., Licon, K., Copeland, W., Shales, M., Fiedler, D., Dutkowski, J., Guenole, A., van Attikum, H., Shokat, K.M., Kolodner, R.D., Huh, W.K., Aebersold, R., Keogh, M.C., Krogan, N.J. & Ideker, T. 2010, "Rewiring of genetic networks in response to DNA damage", Science (New York, N.Y.), vol. 330, no. 6009, pp. 1385-1389. Barber, T.D., McManus, K., Yuen, K.W., Reis, M., Parmigiani, G., Shen, D., Barrett, I., Nouhi, Y., Spencer, F., Markowitz, S., Velculescu, V.E., Kinzler, K.W., Vogelstein, B., Lengauer, C. & Hieter, P. 2008, "Chromatid cohesion defects may underlie chromosome instability in human colorectal cancers. ", Proc. Natl. Acad. Sci., vol. 105, no. 9, pp. 3443-8. Barbie, D.A., Tamayo, P., Boehm, J.S., Kim, S.Y., Moody, S.E., Dunn, I.F., Schinzel, A.C., Sandy, P., Meylan, E., Scholl, C., Frohling, S., Chan, E.M., Sos, M.L., Michel, K., Mermel, C., Silver, S.J., Weir, B.A., Reiling, J.H., Sheng, Q., Gupta, P.B., Wadlow, R.C., Le, H., Hoersch, S., Wittner, B.S., Ramaswamy, S., Livingston, D.M., Sabatini, D.M., Meyerson, M., Thomas, R.K., Lander, E.S., Mesirov, J.P., Root, D.E., Gilliland, D.G., Jacks, T. & Hahn, W.C. 2009, "Systematic RNA interference reveals that oncogenic KRAS-driven cancers require TBK1", Nature, vol. 462, no. 7269, pp. 108112.  153  Beckouet, F., Hu, B., Roig, M.B., Sutani, T., Komata, M., Uluocak, P., Katis, V.L., Shirahige, K. & Nasmyth, K. 2010, "An Smc3 acetylation cycle is essential for establishment of sister chromatid cohesion", Molecular cell, vol. 39, no. 5, pp. 689-699. Beltrao, P., Trinidad, J.C., Fiedler, D., Roguev, A., Lim, W.A., Shokat, K.M., Burlingame, A.l. & Krogan, N.J. 2009, "Evolution of phosphoregulation: comparison of phosphorylation patterns across yeast species”, PLoS biology, vol 7, no. 6. Blat, Y. & Kleckner, N. 1999, "Cohesins bind to preferential sites along yeast chromosome III, with differential regulation along arms versus the centric region", Cell, vol. 98, no. 2, pp. 249-259. Blumenthal, A.B., Kriegstein, H.J. & Hogness, D.S. 1974, “The units of DNA replication in Drosophila melanogaster chromosomes, Cold Spring Harbor symposia on quantitative biology, vol. 38, pp. 205-223. Bolderson, E., Scorah, J., Helleday, T., Smythe, C. & Meuth, M. 2004, "ATM is required for the cellular response to thymidine induced replication fork stress", Human molecular genetics, vol. 13, no. 23, pp. 2937-2945. Boone, C., Bussey, H. & Andrews, B.J. 2007, "Exploring genetic interactions and networks with yeast", Nature reviews.Genetics, vol. 8, no. 6, pp. 437-449. Borges, V., Lehane, C., Lopez-Serra, L., Flynn, H., Skehel, M., Rolef Ben-Shahar, T. & Uhlmann, F. 2010, "Hos1 deacetylates Smc3 to close the cohesin acetylation cycle", Molecular cell, vol. 39, no. 5, pp. 677-688. Boulton, S.J., Martin, J.S., Polanowska, J., Hill, D.E., Gartner, A. & Vidal, M. 2004, "BRCA1/BARD1 orthologs required for DNA repair in Caenorhabditis elegans", Current biology : CB, vol. 14, no. 1, pp. 33-39. Branzei, D. & Foiani, M. 2009, "The checkpoint response to replication stress", DNA repair, vol. 8, no. 9, pp. 1038-1046. Breslow, D.K., Cameron, D.M., Collins, S.R., Schuldiner, M., Stewart-Ornstein, J., Newman, H.W., Braun, S., Madhani, H.D., Krogan, N.J. & Weissman, J.S. 2008, "A comprehensive strategy enabling high-resolution functional analysis of the yeast genome", Nature methods, vol. 5, no. 8, pp. 711-718. Broverman, S.A. & Meneely, P.M. 1994, "Meiotic mutants that cause a polar decrease in recombination on the X chromosome in Caenorhabditis elegans", Genetics, vol. 136, no. 1, pp. 119-127. Bryant, H.E., Petermann, E., Schultz, N., Jemth, A.S., Loseva, O., Issaeva, N., Johansson, F., Fernandez, S., McGlynn, P. & Helleday, T. 2009, "PARP is activated at stalled forks to mediate Mre11-dependent replication restart and recombination", The EMBO journal, vol. 28, no. 17, pp. 2601-2615. 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, vol. 434, no. 7035, pp. 913-917. Buonomo, S.B., Clyne, R.K., Fuchs, J., Loidl, J., Uhlmann, F. & Nasmyth, K. 2000, “Disjunction of homologous chromosomes in meiosis 1 depends on proteolytic cleavage of the meiotic cohesins Rec8 by separin”, Cell, vol 103, no. 3, pp. 387-398. Butty, A.C., Pryciak, P.M., Huang, L.S., Herskowitz, I. & Peter, M. 1998, "The role of Far1p in linking the heterotrimeric G protein to polarity establishment proteins during yeast mating", Science (New York, N.Y.), vol. 282, no. 5393, pp. 1511-1516. 154  Byrne, A.B., Weirauch, M.T., Wong, V., Koeva, M., Dixon, S.J., Stuart, J.M. & Roy, P.J. 2007, "A global analysis of genetic interactions in Caenorhabditis elegans", Journal of Biology, vol. 6, no. 3, pp. 8. Calhoun, E.S., Jones, J.B., Ashfaq, R., Adsay, V., Baker, S.J., Valentine, V., Hempen, P.M., Hilgers, W., Yeo, C.J., Hruban, R.H. & Kern, S.E. 2003, "BRAF and FBXW7 (CDC4, FBW7, AGO, SEL10) mutations in distinct subsets of pancreatic cancer: potential therapeutic targets", The American journal of pathology, vol. 163, no. 4, pp. 1255-1260. Callahan, R. & Hurvitz, S. 2011, "Human epidermal growth factor receptor-2-positive breast cancer: Current management of early, advanced, and recurrent disease", Current opinion in obstetrics & gynecology, vol. 23, no. 1, pp. 37-43. Cassia, R., Moreno-Bueno, G., Rodriguez-Perales, S., Hardisson, D., Cigudosa, J.C. & Palacios, J. 2003, "Cyclin E gene (CCNE) amplification and hCDC4 mutations in endometrial carcinoma", The Journal of pathology, vol. 201, no. 4, pp. 589-595. Ceron, J., Rual, J., Chandra, A., Dupuy, D., Vidal, M. & van den Heuvel, S. 2007, "Largescale RNAi screens identify novel genes that interact with the C. elegans retinoblastoma pathway as well as splicing-related components with synMuv B activity", Biomed Central, vol. 7, pp. 30. Charames, G.S. & Bapat, B. 2003, "Genomic instability and cancer", Current Molecular Medicine, vol. 3, no. 7, pp. 589-596. Chin, C.F. & Yeong, F.M. 2010, "Safeguarding entry into mitosis: the antephase checkpoint", Molecular and cellular biology, vol. 30, no. 1, pp. 22-32. Ciosk, R., Shirayama, M., Shevchenko, T., Tanaka, T., Toth, A., Shevchenko, A. & Nasmyth, K. 2000, “Cohesin’s binding to chromosomes depends on a separate comple consisting of Scc2 and Scc4 proteins”, Molecular cell, vol 5, no. 2, pp. 1-20. Ciosk, R., Zachariae, W., Michaelis, C., Shevchenko, A., Mann, M. & Nasmyth, K. 1998, “An ESP1/PDS1 complex regulates loss of sister chromatid cohesion at the metaphase to anaphase transition in yeast”, Cell, vol. 93, no.6, pp. 1067-1076. Cohen-Fix, O., Peters, J.M., Kirschner, M.W. & Koshland, D.E. 1996, “Anaphase initiation in Saccharomyces cerevisiae is controlled by the APC-dependent degradation of the anaphase inhibitor Pds1p”, Genes & development, vol 10, no. 24, pp. 3081-3093. Collins, S.R., Miller, K.M., Maas, N.L., Roguev, A., Fillingham, J., Chu, C.S., Schuldiner, M., Gebbia, M., Recht, J., Shales, M., Ding, H., Xu, H., Han, J., Ingvarsdottir, K., Cheng, B., Andrews, B., Boone, C., Berger, S., Hieter, P., Zhang, Z., Brown, G.W., Ingles, C.J., Emili, A., Allis, C.D., Toczyski, D.P., Weissman, J.S., Greenblatt, J.F. & Krogan, N.J. 2007, "Functional dissection of protein complexes involved in yeast chromosome biology using a genetic interaction map", Nature, vol. 446, no. 7137, pp. 806-10. Copley, R.R., Schultz, J., Ponting, C.P. & Bork, P. 1999, "Protein families in multicellular organisms", Current opinion in structural biology, vol. 9, no. 3, pp. 408-415. Costanzo, M., Baryshnikova, A., Bellay, J., Kim, Y., Spear, E.D., Sevier, C.S., Ding, H., Koh, J.L., Toufighi, K., Mostafavi, S., Prinz, J., St Onge, R.P., VanderSluis, B., Makhnevych, T., Vizeacoumar, F.J., Alizadeh, S., Bahr, S., Brost, R.L., Chen, Y., Cokol, M., Deshpande, R., Li, Z., Lin, Z.Y., Liang, W., Marback, M., Paw, J., San Luis, B.J., Shuteriqi, E., Tong, A.H., van Dyk, N., Wallace, I.M., Whitney, J.A., Weirauch, M.T., Zhong, G., Zhu, H., Houry, W.A., Brudno, M., Ragibizadeh, S., Papp, B., Pal, C., Roth, F.P., Giaever, G., Nislow, C., Troyanskaya, O.G., Bussey, H., Bader, G.D., 155  Gingras, A.C., Morris, Q.D., Kim, P.M., Kaiser, C.A., Myers, C.L., Andrews, B.J. & Boone, C. 2010, "The genetic landscape of a cell", Science (New York, N.Y.), vol. 327, no. 5964, pp. 425-431. Crusio, K.M., King, B., Reavie, L.B. & Aifantis, I. 2010, "The ubiquitous nature of cancer: the role of the SCF(Fbw7) complex in development and transformation", Oncogene, vol. 29, no. 35, pp. 4865-4873. Cui, H. 2007, "Loss of imprinting of IGF2 as an epigenetic marker for the risk of human cancer", Disease markers, vol. 23, no. 1-2, pp. 105-112. Davierwala, A., Haynes, J., Li, Z., Brost, R., Robinson, M., Yu, L., Mnaimneh, S., Ding, H., Zhu, H., Chen, Y., Cheng, X., Brown, G.W., Boone, C., Andrews, B.J. & Hughes, T.R. 2005, "The synthetic genetic interaction spectrum of essential genes", Nature Genetics, vol. 10, no. 1147, pp. 52. DeSalle, L.M. & Pagano, M. 2001, "Regulation of the G1 to S transition by the ubiquitin pathway", FEBS letters, vol. 490, no. 3, pp. 179-189. Dobzhansky, T. 1946, "Genetics of natural populations; recombination and variability in populations of Drosophila pseudoobscura", Genetics, vol. 31, pp. 269-290. Donnellan, R. & Chetty, R. 1999, "Cyclin E in human cancers", The FASEB journal : official publication of the Federation of American Societies for Experimental Biology, vol. 13, no. 8, pp. 773-780. Donze, D., Adams, C.R., Rine, J. & Kamakaka, R.T. 1999, "The boundaries of the silenced HMR domain in Saccharomyces cerevisiae", Genes & development, vol. 13, no. 6, pp. 698-708. Dorsett, D. 2007, "Roles of the sister chromatid cohesion apparatus in gene expression, development, and human syndromes", Chromosoma, vol. 116, no. 1, pp. 1-13. Eisenmann, D.M. & Kim, S.K. 2000, "Protruding vulva mutants identify novel loci and Wnt signaling factors that function during Caenorhabditis elegans vulva development", Genetics, vol. 156, no. 3, pp. 1097-116. Ekholm-Reed, S., Spruck, C.H., Sangfelt, O., van Drogen, F., Mueller-Holzner, E., Widschwendter, M., Zetterberg, A. & Reed, S.I. 2004, "Mutation of hCDC4 leads to cell cycle deregulation of cyclin E in cancer", Cancer Research, vol. 64, no. 3, pp. 795-800. 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, vol. 434, no. 7035, pp. 917-21. Feldman, R.M., Correll, C.C., Kaplan, K.B. & Deshaies, R.J. 1997, "A complex of Cdc4p, Skp1p, and Cdc53p/cullin catalyzes ubiquitination of the phosphorylated CDK inhibitor Sic1p", Cell, vol. 91, no. 2, pp. 221-230. Florl, A.R. & Schulz, W.A. 2008, "Chromosomal instability in bladder cancer", Archives of Toxicology, vol. 82, no. 3, pp. 173-182. Fong, P.C., Yap, T.A., Boss, D.S., Carden, C.P., Mergui-Roelvink, M., Gourley, C., De Greve, J., Lubinski, J., Shanley, S., Messiou, C., A'Hern, R., Tutt, A., Ashworth, A., Stone, J., Carmichael, J., Schellens, J.H., de Bono, J.S., Kaye, S.B. 2010, “Poly(ADP)ribose polymerase inhibition: frequent durable responses in BRCA carrier ovarian cancer correlating with platinum-free interval”, Journal of clinical oncology, vol 28, no. 15, pp. 2512-2519.  156  Freemantle, S.J. & Dmitrovsky, E. 2010, "Cyclin E transgenic mice: discovery tools for lung cancer biology, therapy, and prevention", Cancer prevention research (Philadelphia, Pa.), vol. 3, no. 12, pp. 1513-1518. Furuta, T., Baillie, D.L. & Schumacher, J.M. 2002, "Caenorhabditis elegans Aurora A kinase AIR-1 is required for postembryonic cell divisions and germline development", Genesis, vol. 34, no. 4, pp. 244-50. Gagnon, S.N., Hengartner, M.O. & Desnoyers, S. 2002, "The genes pme-1 and pme-2 encode two poly(ADP-ribose) polymerases in Caenorhabditis elegans", The Biochemical journal, vol. 368, no. Pt 1, pp. 263-271. Gartner, A., MacQueen, A.J. & Villeneuve, A.M. 2004, "Methods for analyzing checkpoint responses in Caenorhabditis elegans", Methods in molecular biology (Clifton, N.J.), vol. 280, pp. 257-274. Ghiselli, G. & Iozzo, R.V. 2000, "Overexpression of bamacan/SMC3 causes transformation", The Journal of biological chemistry, vol. 275, no. 27, pp. 20235-20238. Giaever, G., Chu, A.M., Ni, L., Connelly, C., Riles, L., Véronneau, S., Dow, S., LucauDanila, A., Anderson, K., André, B., Arkin, A.P., Astromoff, A., El-Bakkoury, M., Bangham, R., Benito, R., Brachat, S., Campanaro, S., Curtiss, M., Davis, K., Deutschbauer, A., Entian, K., Flaherty, P., Foury, F., Garfinkel, D.J., Gerstein, M., Gotte, D., Güldener, U., Hegemann, J.H., Hempel, S., Herman, Z., Jaramillo, D.F., Kelly, D.E., Kelly, S.L., Kötter, P., LaBonte, D., Lamb, D.C., Lan, N., Liang, H., Liao, H., Liu, L., Luo, C., Lussier, M., Mao, R., Menard, P., Ooi, S.L., Revuelta, J.L., Roberts, C.J., Rose, M., Ross-Macdonald, P., Scherens, B., Schimmack, G., Shafer, B., Shoemaker, D.D., Sookhai-Mahadeo, S., Storms, R.K., Strathern, J.N., Valle, G., Voet, M., Volckaert, G., Wang, C.Y., Ward, T.R., Wilhelmy, J., Winzeler, E.A., Yang, Y., Yen, G., Youngman, E., Yu, K., Bussey, H., Boeke, J.D., Snyder, M., Philippsen, P., Davis, R.W. & Johnston, M. 2002, "Functional profiling of the Saccharomyces cerevisiae genome", Nature, vol. 418, pp. 387-91. Glynn, E.F., Megee, P.C., Yu, H.G., Mistrot, C., Unal, D.E., DeRisi, J.L. & Gerton, J.L. 2004, “Genome-wide mapping of the cohesin complex in the yeast Saccharomyces cerevisiae”, PLoS biology, vol 2, no. 9, pp. 1325-1339. Goh, P.Y. & Surana, U. 1999, "Cdc4, a protein required for the onset of S phase, serves an essential function during G(2)/M transition in Saccharomyces cerevisiae", Molecular and cellular biology, vol. 19, no. 8, pp. 5512-5522. Gonzalez-Angulo, A.M., Hennessy, B.T. & Mills, G.B. 2010, "Future of personalized medicine in oncology: a systems biology approach", Journal of clinical oncology : official journal of the American Society of Clinical Oncology, vol. 28, no. 16, pp. 27772783. Grabowski, M.M., Svrzikapa, N. & Tissenbaum, H.A. 2005, "Bloom syndrome ortholog HIM-6 maintains genomic stability in C. elegans", Mechanisms of Ageing and Development, vol. 126, no. 12, pp. 1314-21. Gravel, C., Stergiou, L., Gagnon, S.N. & Desnoyers, S. 2004, "The C. elegans gene pme-5: molecular cloning and role in the DNA-damage response of a tankyrase orthologue", DNA repair, vol. 3, no. 2, pp. 171-182. Guacci, V., Koshland, D.E., & Strunnikov, A. 1997, “A direct link between sister chromatid cohesion and chromosome condensation revealed through the analysis of MCD1 in S. cerevisiae”, Cell, vol 91, no. 1, pp. 47-57. 157  Gutierrez, C. & Schiff, R. 2011, "HER2: biology, detection, and clinical implications", Archives of Pathology & Laboratory Medicine, vol. 135, no. 1, pp. 55-62. Haering, C.H., Lowe, J., Hochwagen, A., Nasmyth, K. 2002, “Molecular architecture of SMC proteins and the yeast cohesin complex”, Molecular cell, vol 9, no. 4, pp. 773-788. Haering, C.H., Schoffnegger, D., Nishino, T., Helmhart, W., Nasmyth, K., Lowe, J. 2004, “Structure and stability of cohesin’s Smc1-kleisin interaction”, Molecular cell, vol 15, no. 6, pp. 951-964. Hagemann, C., Weigelin, B., Schommer, S., Schulze, M., Al-Jomah, N., Anacker, J., Gerngras, S., Kuhnel, S., Kessler, A.F., Polat, B., Ernestus, R.I., Patel, R. & Vince, G.H. 2011, "The cohesin-interacting protein, precocious dissociation of sisters 5A/sister chromatid cohesion protein 112, is up-regulated in human astrocytic tumors", International journal of molecular medicine, vol. 27, no. 1, pp. 39-51. Hanna, J.S., Kroll, E.S., Lundblad, V. & Spencer, F.A. 2001, “Saccharomyces cerevisiae CTF18 and CTF4 are required for sister chromatid cohesion”, Molecular cell biology, vol 21, no. 9, pp. 3144-3158. 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, vol. 278, pp. 10648. Hedgecock, E.M. & White, J.G. Dev Biol. 1985 Jan;107(1):128-33., "Polyploid tissues in the nematode Caenorhabditis elegans", Dev Biol, vol. 107, no. 1, pp. 128-33. Heidinger-Pauli, J.M., Unal, E. & Koshland, D. 2009, "Distinct targets of the Eco1 acetyltransferase modulate cohesion in S phase and in response to DNA damage", Molecular cell, vol. 34, no. 3, pp. 311-321. Helleday, T., 2011, “The underlying mechanism for the PARP and BRCA synthetic lethality: clearing up the misunderstandings”, Molecular oncology, vol. 5, no. 4, pp.387-383. Henchoz, S., Chi, Y., Catarin, B., Herskowitz, I., Deshaies, R.J. & Peter, M. 1997, "Phosphorylation- and ubiquitin-dependent degradation of the cyclin-dependent kinase inhibitor Far1p in budding yeast", Genes & development, vol. 11, no. 22, pp. 3046-3060. Heyer, W.D., Ehmsen, K.T. & Liu, J. 2010, "Regulation of homologous recombination in eukaryotes", Annual Review of Genetics, vol. 44, pp. 113-139. Hodgkin, J., Horvitz, H.R. & Brenner, S. 1979, "Nondisjunction Mutants of the Nematode CAENORHABDITIS ELEGANS", Genetics, vol. 91, no. 1, pp. 67-94. Hodgkin, J., Plasterk, R.H. & Waterston, R.H. 1995, "The nematode Caenorhabditis elegans and its genome", Science (New York, N.Y.), vol. 270, no. 5235, pp. 410-414. Hubbard, E.J., Wu, G., Kitajewski, J. & Greenwald, I. 1997, "sel-10, a negative regulator of lin-12 activity in Caenorhabditis elegans, encodes a member of the CDC4 family of proteins", Genes & development, vol. 11, no. 23, pp. 3182-3193. Inoue, T., Sherwood, D.R., Aspöck, G., Butler, J.A., Gupta, B.P., Kirouac, M., Wang, M., Lee, P.Y., Kramer, J.M., Hope, I., Bürglin, T.R. & Sternberg, P.W. 2002, "Gene expression markers for Caenorhabditis elegans vulval cells", Mechanisms of Development, vol. 119, pp. S203-9. Jager, S., Schwartz, H.T., Horvitz, H.R. & Conradt, B. 2004, "The Caenorhabditis elegans Fbox protein SEL-10 promotes female development and may target FEM-1 and FEM-3 for degradation by the proteasome", Proceedings of the National Academy of Sciences of the United States of America, vol. 101, no. 34, pp. 12549-12554.  158  Jelinic, P. & Shaw, P. 2007, "Loss of imprinting and cancer", The Journal of pathology, vol. 211, no. 3, pp. 261-268. Kaelin, W.G.J. 2005, "The concept of synthetic lethality in the context of anticancer therapy", Nature Reviews Cancer, vol. 5, pp. 689-698. Kamath, R.S., Fraser, A.G., Dong, Y., Poulin, G., Durbin, R., Gotta, M., Kanapin, A., Le Bot, N., Moreno, S., Sohrmann, M., Welchman, D.P., Zipperlen, P. & Ahringer, J. 2003, "Systematic functional analysis of the Caenorhabditis elegans genome using RNAi", Nature, vol. 421, no. 6920, pp. 231-7. Kemp, Z., Rowan, A., Chambers, W., Wortham, N., Halford, S., Sieber, O., Mortensen, N., von Herbay, A., Gunther, T., Ilyas, M. & Tomlinson, I. 2005a, "CDC4 mutations occur in a subset of colorectal cancers but are not predicted to cause loss of function and are not associated with chromosomal instability", Cancer Research, vol. 65, pp. 11361-6. Kemp, Z., Rowan, A., Chambers, W., Wortham, N., Halford, S., Sieber, O., Mortensen, N., von Herbay, A., Gunther, T., Ilyas, M. & Tomlinson, I. 2005b, "CDC4 mutations occur in a subset of colorectal cancers but are not predicted to cause loss of function and are not associated with chromosomal instability", Cancer research, vol. 65, no. 24, pp. 11361-11366. Khan, O.A., Gore, M., Lorigan, P., Stone, J., Greystoke, A., Burke, W., Carmichael, J., Watson, A.J., McGown, G., Thorncroft, M., Margison, G.P., Califano, R., Larkin, J., Wellman, S. & Middleton, M.R. 2011, "A phase I study of the safety and tolerability of olaparib (AZD2281, KU0059436) and dacarbazine in patients with advanced solid tumours", British journal of cancer, vol. 104, no. 5, pp. 750-755. Kim, S., Xu, B., Kastan, M.B., 2002. “Involvement of the cohesion protein, Smc1, in Atmdependent and independent responses to DNA damage”. Genes & Development, vol. 16, pp. 560-570. Kimble, J. & Hirsh, D. 1979, "The postembryonic cell lineages of the hermaphrodite and male gonads in Caenorhabditis elegans", Developmental biology, vol. 70, no. 2, pp. 396417. Kipreos, E.T., Gohel, S.P. & Hedgecock, E.M. 2000, "The C. elegans F-box/WD-repeat protein LIN-23 functions to limit cell division during development", Development (Cambridge, England), vol. 127, no. 23, pp. 5071-5082. Kirienko, N.V., Mani, K. & Fay, D.S. 2010, "Cancer models in Caenorhabditis elegans", Developmental dynamics : an official publication of the American Association of Anatomists, vol. 239, no. 5, pp. 1413-1448. Kitajima, T.S., Hauf, S., Ohsugi, M., Yamamoto, T. & Watanabe, Y. 2005, "Human Bub1 defines the persistent cohesion site along the mitotic chromosome by affecting Shugoshin localization", Current biology : CB, vol. 15, no. 4, pp. 353-359. Klein, F., Mahr, P., Galova, M., Buonomo, S.B., Michaelis, C., Nairz, K. & Nasmyth, K. 1999, "A central role for cohesins in sister chromatid cohesion, formation of axial elements, and recombination during yeast meiosis", Cell, vol. 98, no. 1, pp. 91-103. Komata, M., Bando, M., Araki, H. & Shirahige, K. 2009, "The direct binding of Mrc1, a checkpoint mediator, to Mcm6, a replication helicase, is essential for the replication checkpoint against methyl methanesulfonate-induced stress", Molecular and cellular biology, vol. 29, no. 18, pp. 5008-5019.  159  Köppen, M., Simske, J.S., Sims, P.A., Firestein, B.L., Hall, D.H., Radice, A.D., Rongo, C. & Hardin, J.D. 2001, "Cooperative regulation of AJM-1 controls junctional integrity in Caenorhabditis elegans epithelia", Nature Cell Biology, vol. 3, no. 11, pp. 983-91. Koshland, D.E & Guacci, V. 2000, “Sister chromatid cohesion: the beginning of a long and beautiful relationship”, Current opinion cell biology ; vol 12, no 3, pp. 297-301. Krantz, I.D., McCallum, J., DeScipio, C., Kaur, M., Gillis, L.A., Yaeger, D., Jukofsky, L., Wasserman, N., Bottani, A., Morris, C.A., Nowaczyk, M.J., Toriello, H., Bamshad, M.J., Carey, J.C., Rappaport, E., Kawauchi, S., Lander, A.D., Calof, A.L., Li, H.H., Devoto, M. & Jackson, L.G. 2004, "Cornelia de Lange syndrome is caused by mutations in NIPBL, the human homolog of Drosophila melanogaster Nipped-B", Nature genetics, vol. 36, no. 6, pp. 631-635. Kueng, S., Hegemann, B., Peters, B.H., Lipp, J.J., Schleiffer, A., Mechtler, K. & Peters, J.M. 2006, "Wapl controls the dynamic association of cohesin with chromatin", Cell, vol. 127, no. 5, pp. 955-967. Kwak, E.L., Moberg, K.H., Wahrer, D.C., Quinn, J.E., Gilmore, P.M., Graham, C.A., Hariharan, I.K., Harkin, D.P., Haber, D.A. & Bell, D.W. 2005, "Infrequent mutations of Archipelago (hAGO, hCDC4, Fbw7) in primary ovarian cancer", Gynecologic oncology, vol. 98, no. 1, pp. 124-128. Laloraya, S., Guacci, V. & Koshland, D.E. 2000, “Chromosomal addresses of the cohesin component Mcd1p”, Journal of cell biology, vol 151, no. 5, pp. 1047-1056. Lee, W.J., Soung, H.Y., Kim, Y.S., Nam, W.S., Park, S.W., Lee, Y.J., Yoo, J.N. & Lee, H.S. 2006, "Somatic mutation of hCDC4 gene is rare in lung adenocarcinomas", Acta Oncologica, vol. 45, no. 4, pp. 487-488. Lehner, B., Crombie, C., Tischler, J., Fortunato, A. & Fraser, A. 2006, "Systematic mappng of genetic interactions in Caenothabditis elegans identifies common modifiers of diverse signaling pathways", Nature Genetics, vol. 38, pp. 896-903. Lengronne, A., Katou, Y., Mori, S., Yokobayashi, S., Kelly, G.P., Itoh, T., Watanabe, Y., Shirahige, K. & Uhlmann, F. 2004, "Cohesin relocation from sites of chromosomal loading to places of convergent transcription", Nature, vol. 430, no. 6999, pp. 573-578. Lengronne, A., McIntyre, J., Katou, Y., Kanoh, Y., Hopfner, K.P., Shirahige, K. & Uhlmann, F. 2006, "Establishment of sister chromatid cohesion at the S. cerevisiae replication fork", Molecular cell, vol. 23, no. 6, pp. 787-799. Li, J.H. & Zhang, J. 2001, "PARP inhibitors", IDrugs : the investigational drugs journal, vol. 4, no. 7, pp. 804-812. Lin, Y.W., Chen, H.M. & Fang, J.Y. 2011, "Gene silencing by the Polycomb group proteins and associations with cancer", Cancer investigation, vol. 29, no. 3, pp. 187-195. Liu, N., Li, H., Li, S., Shen, M., Xiao, N., Chen, Y., Wang, Y., Wang, W., Wang, R., Wang, Q., Sun, J. & Wang, P. 2010, "The Fbw7/human CDC4 tumor suppressor targets proproliferative factor KLF5 for ubiquitination and degradation through multiple phosphodegron motifs", The Journal of biological chemistry, vol. 285, no. 24, pp. 18858-18867. Loeb, L.A. 2011, "Human cancers express mutator phenotypes: origin, consequences and targeting", Nature reviews.Cancer, vol. 11, no. 6, pp. 450-457. 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, vol. 412, no. 6846, pp. 557-561. 160  Losada, A., Hirano, M. & Hirano, T. 1998, "Identification of Xenopus SMC protein complexes required for sister chromatid cohesion", Genes & development, vol. 12, no. 13, pp. 1986-1997. Lundin, C., Erixon, K., Arnaudeau, C., Schultz, N., Jenssen, D., Meuth, M. & Helleday, T. 2002, "Different roles for nonhomologous end joining and homologous recombination following replication arrest in mammalian cells", Molecular and cellular biology, vol. 22, no. 16, pp. 5869-5878. Luo, J., Emanuele, M.J., Li, D., Creighton, C.J., Schlabach, M.R., Westbrook, T.F., Wong, K.K. & Elledge, S.J. 2009, "A genome-wide RNAi screen identifies multiple synthetic lethal interactions with the Ras oncogene", Cell, vol. 137, no. 5, pp. 835-848. Lyons, N. & Morgan, O. 2011, "Cdk1-dependent destruction of eco1 prevents cohesion establishment after s phase", Molecular Cell, vol. 42, no. 3, pp. 378-389. Malyukova, A., Dohda, T., von der Lehr, N., Akhoondi, S., Corcoran, M., Heyman, M., Spruck, C., Grander, D., Lendahl, U. & Sangfelt, O. 2007, "The tumor suppressor gene hCDC4 is frequently mutated in human T-cell acute lymphoblastic leukemia with functional consequences for Notch signaling", Cancer research, vol. 67, no. 12, pp. 5611-5616. Mao, J.H., Perez-Losada, J., Wu, D., Delrosario, R., Tsunematsu, R., Nakayama, K.I., Brown, K., Bryson, S. & Balmain, A. 2004, "Fbxw7/Cdc4 is a p53-dependent, haploinsufficient tumour suppressor gene", Nature, vol. 432, no. 7018, pp. 775-779. Mayer, M.L., Gygi, S.P., Aebersold, R.H., & Hieter, P. 2001, "Identification of RFC(Ctf18p, Ctf8p, Dcc1p): an alternative RFC complex required for sister chromatid cohesion in S. cerevisiae", Molecular Cell, vol. 7, pp. 959-70. Mayer, M.L., Pot, I., Chang, M., Xu, H., Aneliunas, V., Kwok, T., Newitt, R., Aebersold, R., Boone, C., Brown, G.W. & Hieter, P. 2004, "Identification of protein complexes required for efficient sister chromatid cohesion", Molecular biology of the cell, vol. 15, no. 4, pp. 1736-1745. 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", Molecular biology of the cell, vol. 20, no. 24, pp. 5306-5313. McManus, K., Barrett, I., Nouhi, Y. & Hieter, P. 2009, "Specific synthetic lethal killing of RAD54B-deficient human colorectal cancer cells by FEN1 silencing", Proc. Natl. Acad. Sci., vol. 106, pp. 3276-81. Meier, B., Clejan, I., Liu, Y., Lowden, M., Gartner, A., Hodgkin, J. & Ahmed, S. 2006, "trt-1 is the Caenorhabditis elegans catalytic subunit of telomerase", Public Library of Science Genetics, vol. 2, no. 2, pp. E18. Meister, P., Taddei, A., Vernis, L., Poidevin, M., Gasser, S.M. & Baldacci, G. 2005, "Temporal separation of replication and recombination requires the intra-S checkpoint", The Journal of cell biology, vol. 168, no. 4, pp. 537-544. Meyerson, M., Gabriel, S. & Getz, G. 2010, "Advances in understanding cancer genomes through second-generation sequencing", Nature reviews.Genetics, vol. 11, no. 10, pp. 685-696. Michaelis, C., Ciosk, R. & Nasmyth, K. 1997, "Cohesins: chromosomal proteins that prevent premature separation of sister chromatids", Cell, vol. 91, no. 1, pp. 35-45.  161  Misulovin, Z., Schwartz, Y.B., Li, X.Y., Kahn, T.G., Gause, M., MacArthur, S., Fay, J.C., Eisen, M.B., Pirrotta, V., Biggin, M.D. & Dorsett, D. 2008, "Association of cohesin and Nipped-B with transcriptionally active regions of the Drosophila melanogaster genome", Chromosoma, vol. 117, no. 1, pp. 89-102. Mnaimneh, S., Davierwala, A.P., Haynes, J., Moffat, J., Peng, W.T., Zhang, W., Yang, X., Pootoolal, J., Chua, G., Lopez, A., Trochesset, M., Morse, D., Krogan, N.J., Hiley, S.L., Li, Z., Morris, Q., Grigull, J., Mitsakakis, N., Roberts, C.J., Greenblatt, J.F., Boone, C., Kaiser, C.A., Andrews, B.J. & Hughes, T.R. 2004, “Exploration of essential gene functions via titratable promoter alleles”, Cell, vol 118, no. 1, pp. 31-44. Musio, A., Selicorni, A., Focarelli, M.L., Gervasini, C., Milani, D., Russo, S., Vezzoni, P. & Larizza, L. 2006, "X-linked Cornelia de Lange syndrome owing to SMC1L1 mutations", Nature genetics, vol. 38, no. 5, pp. 528-530. Nasmyth, K. & Haering, C.H. 2009, "Cohesin: Its Roles and Mechanisms", Annual Review of Genetics, vol. 43, pp. 525-558. Nasmyth, K., Peters, J.M. & Uhlmann, F. 2001, "Splitting the chromosome: cutting the ties that bind sister chromatids", Novartis Foundation symposium, vol. 237, pp. 113-33; discussion 133-8, 158-63. Nativio, R., Wendt, K.S., Ito, Y., Huddleston, J.E., Uribe-Lewis, S., Woodfine, K., Krueger, C., Reik, W., Peters, J.M. & Murrell, A. 2009, "Cohesin is required for higher-order chromatin conformation at the imprinted IGF2-H19 locus", PLoS genetics, vol. 5, no. 11, pp. e1000739. Nijman, S.M. 2011, "Synthetic lethality: general principles, utility and detection using genetic screens in human cells", FEBS letters, vol. 585, no. 1, pp. 1-6. Oikawa, K., Ohbayashi, T., Kiyono, T., Nishi, H., Isaka, K., Umezawa, A., Kuroda, M. & Mukai, K. 2004, "Expression of a novel human gene, human wings apart-like (hWAPL), is associated with cervical carcinogenesis and tumor progression", Cancer Research, vol. 64, no. 10, pp. 3545-3549. Olson, B.L., Hock, M.B., Ekholm-Reed, S., Wohlschlegel, J.A., Dev, K.K., Kralli, A. & Reed, S.I. 2008, "SCFCdc4 acts antagonistically to the PGC-1alpha transcriptional coactivator by targeting it for ubiquitin-mediated proteolysis", Genes & development, vol. 22, no. 2, pp. 252-264. O'Shaughnessy, J., Osborne, C., Pippen, J.E., Yoffe, M., Patt, D., Rocha, C., Koo, I.C., Sherman, B.M. & Bradley, C. 2011, "Iniparib plus chemotherapy in metastatic triplenegative breast cancer", The New England journal of medicine, vol. 364, no. 3, pp. 205214. Ouyang, K.J., Woo, L.L. & Ellis, N.A. 2008, "Homologous recombination and maintenance of genome integrity: cancer and aging through the prism of human RecQ helicases", Mechanisms of Ageing and Development, vol. 129, no. 7-8, pp. 425-40. Ozkaynak, E., Finley, D., Solomon, M.J. & Varshavsky, A. 1987, "The yeast ubiquitin genes: a family of natural gene fusions", The EMBO journal, vol. 6, no. 5, pp. 1429-1439. 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, vol. 124, no. 5, pp. 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", Molecular Cell, vol. 16, pp. 487-96.  162  Parelho, V., Hadjur, S., Spivakov, M., Leleu, M., Sauer, S., Gregson, H.C., Jarmuz, A., Canzonetta, C., Webster, Z., Nesterova, T., Cobb, B.S., Yokomori, K., Dillon, N., Aragon, L., Fisher, A.G. & Merkenschlager, M. 2008, "Cohesins functionally associate with CTCF on mammalian chromosome arms", Cell, vol. 132, no. 3, pp. 422-433. Park, M. & Krause, M.W. 1999, "Regulation of postembryonic G(1) cell cycle progression in Caenorhabditis elegans by a cyclin D/CDK-like complex", Developmental Biology, vol. 126, no. 21, pp. 4849-60. Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D. & team R Computing 2008, nlme: Linear and Nonlinear Mixed Effects Models. R package version 3.1-90 ed. Pinheiro, J.C. & Bates, D.M. 2000, Mixed Effects Models in S and S-Plus, 1st ed edn, Springer. Pino, M.S. & Chung, D.C. 2010, "The chromosomal instability pathway in colon cancer", Gastroenterology, vol. 138, no. 6, pp. 2059-2072. Podbilewicz, B. & White, J.G. 1994, "Cell fusions in the developing epithelia of C. elegans", Dev Biol, vol. 161, pp. 408-24. Prieto, I., Suja, J.A., Pezzi, N., Kremer, L., Martinez-A, C., Rufas, J.S. & Barbero, J.L. 2001, "Mammalian STAG3 is a cohesin specific to sister chromatid arms in meiosis I", Nature cell biology, vol. 3, no. 8, pp. 761-766. Pulak, R. 2006, "Techniques for analysis, sorting, and dispensing of C. elegans on the COPAS flow-sorting system", Methods in molecular biology (Clifton, N.J.), vol. 351, pp. 275-286. R Development Core Team: R Foundation for Statistical Computing 2008, R: A Language and Environment for Statistical Computing. Rajagopalan, H., Jallepalli, P.V., Rago, C., Velculescu, V.E., Kinzler, K.W., Vogelstein, B. & Lengauer, C. 2004, "Inactivation of hCDC4 can cause chromosomal instability", Nature, vol. 428, no. 6978, pp. 77-81. Revenkova, E., Eijpe, M., Heyting, C., Hodges, C.A., Hunt, P.A., Liebe, B., Scherthan, H. & Jessberger, R. 2004, "Cohesin SMC1 beta is required for meiotic chromosome dynamics, sister chromatid cohesion and DNA recombination", Nature cell biology, vol. 6, no. 6, pp. 555-562. Roguev, A., Bandyopadhyay, S., Zofall, M., Zhang, K., Fischer, T., Collins, S.R., Qu, H., Shales, M., Park, H.O., Hayles, J., Hoe, K.L., Kim, D.U., Ideker, T., Grewal, S.I., Weissman, J.S. Krogan, N.J. 2008, “Conservation and rewiring of functional modules revealed by an epistasis map in fission yeast”, Science, vol 322, no. 5900, pp. 405-410. Rolef Ben-Shahar, T., Heeger, S., Lehane, C., East, P., Flynn, H., Skehel, M. & Uhlmann, F. 2008, "Eco1-dependent cohesin acetylation during establishment of sister chromatid cohesion", Science (New York, N.Y.), vol. 321, no. 5888, pp. 563-566. Rollins, R.A., Morcillo, P. & Dorsett, D. 1999, "Nipped-B, a Drosophila homologue of chromosomal adherins, participates in activation by remote enhancers in the cut and Ultrabithorax genes", Genetics, vol. 152, no. 2, pp. 577-593. Rowland, B.D., Roig, M.B., Nishino, T., Kurze, A., Uluocak, P., Mishra, A., Beckouët, F., Underwood, P., Metson, J., Imre, R., Mechtler, K., Katis, V.L. & Nasmyth, K. 2009, "Building sister chromatid cohesion: smc3 acetylation counteracts an antiestablishment activity", Molecular Cell, vol. 33, no. 6, pp. 763-74.  163  Saharia, A., Guittat, L., Crocker, S., Lim, A., Steffen, M., K., S. & Stewart, S.A. 2008, "Flap endonuclease 1 contributes to telomere stability", Current Biology, vol. 18, no. 7, pp. 496-500. Saintigny, Y., Dumay, A., Lambert, S. & Lopez, B.S. 2001, "A novel role for the Bcl-2 protein family: specific suppression of the RAD51 recombination pathway", The EMBO journal, vol. 20, no. 10, pp. 2596-2607. Scholl, C., Frohling, S., Dunn, I.F., Schinzel, A.C., Barbie, D.A., Kim, S.Y., Silver, S.J., Tamayo, P., Wadlow, R.C., Ramaswamy, S., Dohner, K., Bullinger, L., Sandy, P., Boehm, J.S., Root, D.E., Jacks, T., Hahn, W.C. & Gilliland, D.G. 2009, "Synthetic lethal interaction between oncogenic KRAS dependency and STK33 suppression in human cancer cells", Cell, vol. 137, no. 5, pp. 821-834. Schuldiner, M., Collins, S.R., Thompson, N.J., Denic, V., Bhamidipati, A., Punna, T., Ihmels, J., Andrews, B., Boone, C., Greenblatt, J.F., Weissman, J.S. & Krogan, N.J. 2005, “Exploration of the function and organization of the yeast early secretory pathway through an epistatic miniarray profile, Cell, vol 123, no 3, pp. 507-519. Schule, B., Oviedo, A., Johnston, K., Pai, S. & Francke, U. 2005, "Inactivating mutations in ESCO2 cause SC phocomelia and Roberts syndrome: no phenotype-genotype correlation", American Journal of Human Genetics, vol. 77, no. 6, pp. 1117-1128. Schwob, E., Bohm, T., Mendenhall, M.D. & Nasmyth, K. 1994, "The B-type cyclin kinase inhibitor p40SIC1 controls the G1 to S transition in S. cerevisiae", Cell, vol. 79, no. 2, pp. 233-244. Seydoux, G., Savage, C. & Greenwald, I. Dev Biol. 1993 Jun;157(2):423-36., "Isolation and characterization of mutations causing abnormal eversion of the vulva in Caenorhabditis elegans", Developmental Biology, vol. 157, no. 2, pp. 423-36. Simmer, F., Moorman, C., van der Linden, A.M., Kuijk, E., van den Berghe, P.V., Kamath, R.S., Fraser, A.G., Ahringer, J. & Plasterk, R.H. 2003, "Genome-wide RNAi of C. elegans using the hypersensitive rrf-3 strain reveals novel gene functions", Public Library of Science, vol. 1, pp. E12. Sjogren, C. & Nasmyth, K. 2001, "Sister chromatid cohesion is required for postreplicative double-strand break repair in Saccharomyces cerevisiae", Current biology : CB, vol. 11, no. 12, pp. 991-995. Skibbens, R.V. 2004, "Chl1p, a DNA helicase-like protein in budding yeast, functions in sister-chromatid cohesion", Genetics, vol. 166, no. 1, pp. 33-42. Skibbens, R.V., Corson, L.B., Koshland, D.E., Hieter, P. 1999, “Ctf7p is essential for sister chromatid cohesion and links mitotic chromosome structure to the DNA replication machinery”, Genes & development, vol 13, no. 3, pp. 307-319. Skowyra, D., Craig, K.L., Tyers, M., Elledge, S.J. & Harper, J.W. 1997, "F-box proteins are receptors that recruit phosphorylated substrates to the SCF ubiquitin-ligase complex", Cell, vol. 91, no. 2, pp. 209-219. Smith, J., Tho, L.M., Xu, N. & Gillespie, D.A. 2010, "The ATM-Chk2 and ATR-Chk1 pathways in DNA damage signaling and cancer", Advances in Cancer Research, vol. 108, pp. 73-112. Solomon, D.A., Kim, T., Diaz-Martinez, L.A., Fair, J., Elkahloun, A.G., Harris, B., Toretsky, J.A., Rosenberg, S.A., Shukla, N., Ladanyi, M., Samuels, Y., James, C.D., Yu, H., Kim, J.S. & Waldman, T. 2011, “Mutational inactivation of STAG2 causes aneuploidy in human cancer”, Science, vol 333, no. 6045, pp. 1039-1043. 164  Spruck, C.H., Strohmaier, H., Sangfelt, O., Müller, H.M., Hubalek, M., Müller-Holzner, E., Marth, C., Widschwendter, M. & Reed, S.I. 2002, "hCDC4 gene mutations in endometrial cancer", Cancer Research, vol. 62, no. 16, pp. 4535-9. Stamatakos, M., Palla, V., Karaiskos, I., Xiromeritis, K., Alexiou, I., Pateras, I. & Kontzoglou, K. 2010, "Cell cyclins: triggering elements of cancer or not?", World journal of surgical oncology, vol. 8, pp. 111. Stirling, P.C., Bloom, M.S., Solanki-Patil, T., Smith, S., Sipahimalani, P., Li, Z., Kofoed, M., Ben-Aroya, S., Myung, K. & Hieter, P. 2011, "The Complete Spectrum of Yeast Chromosome Instability Genes Identifies Candidate CIN Cancer Genes and Functional Roles for ASTRA Complex Components", PLoS genetics, vol. 7, no. 4, pp. e1002057. St-Laurent, J.F., Gagnon, S.N., Dequen, F., Hardy, I. & Desnoyers, S. 2007, "Altered DNA damage response in Caenorhabditis elegans with impaired poly(ADP-ribose) glycohydrolases genes expression", DNA repair, vol. 6, no. 3, pp. 329-343. Straight, A.F., Belmont, A.S., Robinett, C.C. & Murray, A.W. 1996, "GFP tagging of budding yeast chromosomes reveals that protein-protein interactions can mediate sister chromatid cohesion", Current biology : CB, vol. 6, no. 12, pp. 1599-1608. Stratton, M.R. 2011, "Exploring the genomes of cancer cells: progress and promise", Science (New York, N.Y.), vol. 331, no. 6024, pp. 1553-1558. Strom, L., Lindroos, H.B., Shirahige, K. & Sjogren, C. 2004, “ Postreplicative recruitment of cohesin to double-strand breaks is required for DNA repair”, Molecular cell, vol 16, no. 6, pp. 1003-1015. Sulston, J.E. & Horvitz, H.R. 1977, "Post-embryonic cell lineages of the nematode, Caenorhabditis elegans", Dev Biol, vol. 56, no. 1, pp. 110-56. Sumara, I., Vorlaufer, E., Gieffers, C., Peters, B.H. & Peters, J.-M. 2000, “Characterization of vertebrate cohesin complexes and their regulation in prophase”, Journal of cell biology. Vol 151, no. 4, pp. 749-762. Sundaram, M.V. 2006, "RTK/Ras/MAPK signaling." in WormBook , pp. 1-19. Sundaram, M. & Greenwald, I. 1993, "Suppressors of a lin-12 hypomorph define genes that interact with both lin-12 and glp-1 in Caenorhabditis elegans", Genetics, vol. 135, no. 3, pp. 765-783. Sundqvist, A., Bengoechea-Alonso, M.T., Ye, X., Lukiyanchuk, V., Jin, J., Harper, J.W. & Ericsson, J. 2005, "Control of lipid metabolism by phosphorylation-dependent degradation of the SREBP family of transcription factors by SCF(Fbw7)", Cell metabolism, vol. 1, no. 6, pp. 379-391. Sutani, T., Kawaguchi, T., Kanno, R., Itoh, T. & Shirahige, K. 2009, "Budding yeast Wpl1(Rad61)-Pds5 complex counteracts sister chromatid cohesion-establishing reaction", Current biology : CB, vol. 19, no. 6, pp. 492-497. Symington, L. 1998, "Homologous recombination is required for the viability of rad27 mutants", Nucleic Acids Research, vol. 26, no. 5589, pp. 95. Takebe, N., Harris, P.J., Warren, R.Q. & Ivy, S.P. 2011, "Targeting cancer stem cells by inhibiting Wnt, Notch, and Hedgehog pathways", Nature reviews.Clinical oncology, vol. 8, no. 2, pp. 97-106. Tan, Y., Sangfelt, O. & Spruck, C. 2008, "The Fbxw7/hCdc4 tumor suppressor in human cancer", Cancer letters, vol. 271, no. 1, pp. 1-12.  165  Tanaka, H., Katou, Y., Yagura, M., Saitoh, K., Itoh, T., Araki, H., Bando, M. & Shirahige, K. 2009, "Ctf4 coordinates the progression of helicase and DNA polymerase alpha", Genes to cells : devoted to molecular & cellular mechanisms, vol. 14, no. 7, pp. 807-820. Tanaka, T., Cosma, M.P., Wirth, K. & Nasmyth, K. 1999, "Identification of cohesin association sites at centromeres and along chromosome arms", Cell, vol. 98, no. 6, pp. 847-858. Tarailo, M., Tarailo, S. & Rose, A.M. 2007, "Synthetic lethal interactions identify phenotypic "interologs" of the spindle assembly checkpoint components", Genetics, vol. 177, pp. 2525-30. Terns, R.M.,., Kroll-Conner, P., Zhu, J., Chung, S. & Rothman, J.H. 1997, "A deficiency screen for zygotic loci required for establishment and patterning of the epidermis in Caenorhabditis elegans", Genetics, vol. 146, no. 1, pp. 185-206. Terret, M.E., Sherwood, R., Rahman, S., Qin, J. & Jallepalli, P.V. 2009, "Cohesin acetylation speeds the replication fork", Nature, vol. 462, no. 7270, pp. 231-234. Tijsterman, M., Pothof, J. & Plasterk, R.H. 2002, "Frequent germline mutations and somatic repeat instability in DNA mismatch-repair-deficient Caenorhabditis elegans", Genetics, vol. 161, no. 2, pp. 651-660. Tischler, J., Lehner, B., Chen, N. & Fraser, A. 2006, "Combinatorial RNA interference in Caenorhabditis elegans reveals that redundancy between gene duplicates can be maintained for more than 80 million years of evolution", Genome Biology, vol. 8. Tong, A.H., Evangelista, M., Parsons, A.B., Xu, H., Bader, G.D., Pagé, N., Robinson, M., Raghibizadeh, S., Hogue, C.W., Bussey, H., Andrews, B., Tyers, M. & Boone, C. 2001, " Systematic genetic analysis with ordered arrays of yeast deletion mutants", Science, vol. 294, pp. 2364-8. 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., Ménard, 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, vol. 303, no. 5659, pp. 80813. Tonkin, E.T., Wang, T.J., Lisgo, S., Bamshad, M.J. & Strachan, T. 2004, "NIPBL, encoding a homolog of fungal Scc2-type sister chromatid cohesion proteins and fly Nipped-B, is mutated in Cornelia de Lange syndrome", Nature genetics, vol. 36, no. 6, pp. 636-641. Toth, A., Ciosk, R., Uhlmann, F., Galova, M., Schleiffer, A., & Nasmyth, K. 1999, “Yeast cohesin complex requires a conserved protein, Eco1(Ctf7), to establish cohesion between sister chromatids during DNA replication”, Genes & development, vol 13, no. 3, pp. 320-333. Trainer, A.H., Lewis, C.R., Tucker, K., Meiser, B., Friedlander, M. & Ward, R.L. 2010, "The role of BRCA mutation testing in determining breast cancer therapy", Nature reviews.Clinical oncology, vol. 7, no. 12, pp. 708-717. Tsunematsu, R., Nakayama, K., Oike, Y., Nishiyama, M., Ishida, N., Hatakeyama, S., Bessho, Y., Kageyama, R., Suda, T. & Nakayama, K.I. 2004, "Mouse Fbw7/Sel-  166  10/Cdc4 is required for notch degradation during vascular development", The Journal of biological chemistry, vol. 279, no. 10, pp. 9417-9423. Uhlmann, F., Lottspeich, F. & Nasmyth, K. 1999, “Sister chromatid separation at anaphase onset is triggered by cleavage of the cohesin subunit Scc1p”, Nature, vol 400, no. 6739, pp. 37-42. Unal, E., Arbel-Eden, A., Sattler, U., Shroff, R., Lichten, M., Haber, J.E. & Koshland, D. 2004, "DNA damage response pathway uses histone modification to assemble a doublestrand break-specific cohesin domain", Molecular cell, vol. 16, no. 6, pp. 991-1002. Unal, E., Heidinger-Pauli, J.M. & Koshland, D.E. 2007, “DN double-strand breaks trigger genome-wide sister-chromatd cohesion through Eco1 (Ctf7)”, Science, vol 317, no. 5838, pp. 245-248. Unal, E., Heidinger-Pauli, J.M., Kim, W., Guacci, V., Onn, I., Gygi, S.P. & Koshland, D.E. 2008, "A molecular determinant for the establishment of sister chromatid cohesion", Science (New York, N.Y.), vol. 321, no. 5888, pp. 566-569. Vega, H., Waisfisz, Q., Gordillo, M., Sakai, N., Yanagihara, I., Yamada, M., van Gosliga, D., Kayserili, H., Xu, C., Ozono, K., Jabs, E.W., Inui, K. & Joenje, H. 2005, "Roberts syndrome is caused by mutations in ESCO2, a human homolog of yeast ECO1 that is essential for the establishment of sister chromatid cohesion", Nature genetics, vol. 37, no. 5, pp. 468-470. Vigil, D., Cherfils, J., Rossman, K.L. & Der, C.J. 2010, "Ras superfamily GEFs and GAPs: validated and tractable targets for cancer therapy?", Nature reviews.Cancer, vol. 10, no. 12, pp. 842-857. Vogelstein, B. & and Kinzler, K. 2004, "Cancer genes and the pathways they control", Nature Medicine, vol. 10, pp. 789-799. Waizenegger, I.C., Hauf, S., Meinke, A. & Peters, J.M. 2000, “Two distinct pathways remove mammalian cohesin from chromosome arms in prophase and from centromeres in anaphase”, Cell, vol. 103, no. 3, pp. 399-410. Walter, J. & Newport, J.W. 1997, “Regulation of replicon size in Xenopus egg extracts”, Science, vol 275, no. 5302, pp. 993-995. Wang, F., Yoder, J., Antoshechkin, I. & Han, M. 2003, "Caenorhabditis elegans EVL14/PDS-5 and SCC-3 are essential for sister chromatid cohesion in meiosis and mitosis", Molecular Cell Biology, vol. 23, no. 21, pp. 7698-707. Wang, Z., Cummins, J.M., Shen, D., Cahill, D.P., Jallepalli, P.V., Wang, T.L., Parsons, D.W., Traverso, G., Awad, M., Silliman, N., Ptak, J., Szabo, S., Willson, J.K., Markowitz, S.D., Goldberg, M.L., Karess, R., Kinzler, K.W., Vogelstein, B., Velculescu, V.E. & Lengauer, C. 2004, "Three classes of genes mutated in colorectal cancers with chromosomal instability", Cancer Research, vol. 64, pp. 2998-3001. Wei, W., Jin, J., Schlisio, S., Harper, J.W. & Kaelin, W.G.,Jr 2005, "The v-Jun point mutation allows c-Jun to escape GSK3-dependent recognition and destruction by the Fbw7 ubiquitin ligase", Cancer cell, vol. 8, no. 1, pp. 25-33. Weidhaas, J.B., Eisenmann, D.M., Holub, J.M. & Nallur, S.V. 2006, "A Caenorhabditis elegans tissue model of radiation-induced reproductive cell death", Proc. Natl. Acad. Sci., vol. 103, pp. 9946-51. Wendt, K.S. & Peters, J.M. 2009, "How cohesin and CTCF cooperate in regulating gene expression", Chromosome research : an international journal on the molecular,  167  supramolecular and evolutionary aspects of chromosome biology, vol. 17, no. 2, pp. 201-214. Wendt, K.S., Yoshida, K., Itoh, T., Bando, M., Koch, B., Schirghuber, E., Tsutsumi, S., Nagae, G., Ishihara, K., Mishiro, T., Yahata, K., Imamoto, F., Aburatani, H., Nakao, M., Imamoto, N., Maeshima, K., Shirahige, K. & Peters, J.M. 2008, "Cohesin mediates transcriptional insulation by CCCTC-binding factor", Nature, vol. 451, no. 7180, pp. 796-801. Wicky, C., Alpi, A., Passannante, M., Rose, A., Gartner, A. & Müller, F. 2004, "Multiple genetic pathways involving the Caenorhabditis elegans Bloom's syndrome genes him-6, rad-51, and top-3 are needed to maintain genome stability in the germ line", Molecular Cell Bology, vol. 24, no. 11, pp. 5016-27. Woollard, A. & Hodgkin, J. 1999, "Stu-7/air-2 is a C. elegans aurora homologue essential for chromosome segregation during embryonic and post-embryonic development", Mechanisms of Development, vol. 82, pp. 95-108. Wu, G., Hubbard, E.J., Kitajewski, J.K. & Greenwald, I. 1998, "Evidence for functional and physical association between Caenorhabditis elegans SEL-10, a Cdc4p-related protein, and SEL-12 presenilin", Proceedings of the National Academy of Sciences of the United States of America, vol. 95, no. 26, pp. 15787-15791. Xu, H., Tomaszewski, J.M. & McKay, M.J. 2011, "Can corruption of chromosome cohesion create a conduit to cancer?", Nature Reviews Cancer, vol. 11, no. 3, pp. 199-210. Xu, H., Yan, M., Patra, J., Natrajan, R., Yan, Y., Swagemakers, S., Tomaszewski, J.M., Verschoor, S., Millar, E.K., van der Spek, P., Reis-Filho, J.S., Ramsay, R.G., O'Toole, S.A., McNeil, C.M., Sutherland, R.L., McKay, M.J. & Fox, S.B. 2011, "Enhanced RAD21 cohesin expression confers poor prognosis and resistance to chemotherapy in high grade luminal, basal and HER2 breast cancers", Breast cancer research : BCR, vol. 13, no. 1, pp. R9. Yanagida, M. 2000, "Cell cycle mechanisms of sister chromatid separation; roles of Cut1/separin and Cut2/securin", Genes to cells : devoted to molecular & cellular mechanisms, vol. 5, no. 1, pp. 1-8. Yazdi, P.T., Wang, Y., Zhao, S., Patel, N., Lee, E.Y. & Qin, J. 2002, "SMC1 is a downstream effector in the ATM/NBS1 branch of the human S-phase checkpoint", Genes & development, vol. 16, no. 5, pp. 571-582. Yuen, K.W., Warren, C.D., Chen, O., Kwok, T., Hieter, P. & and Spencer, F.A. 2007, "Systematic genome instability screens in yeast and their potential relevance to cancer. ", Proc. Natl. Acad. Sci., vol. 104, pp. 3925-30. Zhang, N., Ge, G., Meyer, R., Sethi, S., Basu, D., Pradhan, S., Zhao, Y.J., Li, X.N., Cai, W.W., El-Naggar, A.K., Baladandayuthapani, V., Kittrell, F.S., Rao, P.H., Medina, D. & Pati, D. 2008, "Overexpression of Separase induces aneuploidy and mammary tumorigenesis", Proceedings of the National Academy of Sciences of the United States of America, vol. 105, no. 35, pp. 13033-13038. Zhu, W., Ukomadu, C., Jha, S., Senga, T., Dhar, S.K., Wohlschlegel, J.A., Nutt, L.K., Kornbluth, S. & Dutta, A. 2007, "Mcm10 and And-1/CTF4 recruit DNA polymerase alpha to chromatin for initiation of DNA replication", Genes and Development, vol. 21, no. 18, pp. 2288-99. Zilfou, J.T. & Lowe, S.W. 2009, "Tumor suppressive functions of p53", Cold Spring Harbor perspectives in biology, vol. 1, no. 5, pp. a001883. 168  Appendix List of 200 RNAi clones used in the Pvl pilot screen C. elegans  S. cerevisiae  Function  GFI-3/APC-5 C33F10.2 san-1 apc-10 mdf-1 mdf-2 emb-30 F54D10.1 F29B9.6 bir-2 bir-1 F49E12.6 C17G10.4 abl-1 lin-23 phg-1 T10F2.3 cdk-5 prp-17 F34d10.2 K08E7.7 ZK856.1 C34G6.5 C02F5.7 K08D10.2 R09E12.3 cep-1 ape-1 hus-1 RAD-5/CLK-2 Y69A2A_2991.A pch-2 atl-1 atm-1 K10E9.1 F57C12.2 R13F6.10  APC5 BUB2 CDC20 DOC1 MAD1 MAD2 no homolog SKP1 UBC9 no homolog no homolog no homolog CDC14 CDC15 CDC4 no homolog ULP1 CDC28 CDC40 CDC45 CDC53 CDC53 CDC7 GRR1 JAC1 HSP90 no homolog no homolog no homolog no homolog PTP1 PCH2 MEC1 MEC1 ATM GTR1 no homolog  apc apc apc apc apc apc apc apc apc apoptosis apoptosis apoptosis cell cycle exit cell cycle exit cell cycle exit cell cycle exit cell cycle exit cell cycle start cell cycle start cell cycle start cell cycle start cell cycle start cell cycle start cell cycle start chaperone chaperone checkpoint checkpoint checkpoint checkpoint checkpoint checkpoint checkpoint checkpoint checkpoint chromatin chromatin  Mutator in C.elegans (Y/N)  Y  Y Y  Y Y  Y  Y  169  C. elegans  S. cerevisiae  Function  scc-1 him-1 mix-1 F54D5.14 C23H4.6 C08b11.6 C30G7.1 F45E1.6 C35A5.9 T28D6.4 F46G10.7 T26A5.5 lin-35 R09B3.4 C08B11.2 Y43F4B.3 F02E9.4 sir-2.1 F39H11.1 C06H2.3 F02D10.7 C17E4.6 hcp-1 sgo-1 hcp-2 msh-4 him-14 Y111B2a.8 kin-18 ubc-1 Y76A2B.5 R05D3.6 ebp-2 F16D3.4 klp-15 bmk-1 kbp-5 K07H8.1 C06A8.5 ben-1 klp-3  SCC1 SMC1 SMC2 SMC6 SMC6 ARP6 HHO1 HHT1 HOS2 HOS4 HST1 JHD1 no homolog HDA12 RPD3 SET1 SIN3 SIR2 SWC5 no homolog no homolog no homolog no homolog SGO1 no homolog MSH4 MSH4 SNF4 SPS1 UBC1 no homolog no homolog BIM1 CIN1 KAR3 KIP1 NDL1 PAC2 SPC110 TUB2 KAR3  cohesion cohesion cohesion cohesion cohesion histone histone histone histone histone histone histone histone histone histone histone histone histone histone histone histone histone kinetochore kinetochore kinetochore meiosis meiosis metabolism metabolism metabolism metabolism metabolism microtubule microtubule microtubule microtubule microtubule microtubule microtubule microtubule microtubule  Mutator in C.elegans (Y/N)  Y Y Y Y Y Y Y Y Y Y Y Y Y  Y  170  C. elegans  S. cerevisiae  Function  lfi-1 B0035.4 pfd-5 pfd-3 pfd-6 apn-1 ctf-18 ctf-8 K09H9.2 lig-4 kin-20 H05L14.1 T05A12.4 C24g6.3 mlh-1 F56A6.4 D2005.5 F15B10.2 mre-11 H26D21.2 Y47G6A.11 mus-81 hsr-9 msh-5 brc-1 brd-1 dog-1 rtel-1 C30A5.2 Y71F9AL.2 C32D5.6 R10E4.5 Y18H1A.6 pms-2 mrt-2 F12F6.7 T01G1.1c ZK20.3 crn-1 him-9 xpa-1  MLP1 GIM3 GIM5 PAC10 YKE2 APN1 CTF18 CTF8 DCC1 DNL4 HRR25 HRR25 IRC20 MET18 MLH1 MMS4 MPH1 MPH1 MRE11 MSH2 MSH6 MUS81 no homolog no homolog no homolog no homolog no homolog no homolog no homolog HRR25 no homolog NTH1 PIF1 PMS1 POL30 POL31 POL32 RAD23 RAD27 RAD1 RAD14  nuclear pore prefoldin prefoldin prefoldin prefoldin repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair  Mutator in C.elegans (Y/N)  Y  Y Y  Y Y  Y  Y Y  Y Y  171  C. elegans  S. cerevisiae  Function  hpr-17 F53H4.1 rad-50 rad-51 rad-54 rev-1 him-6 rcq-5 wrn-1 K02F3.12 Y56A3A.29 T21D12.9b cku-70 cku-80 AC8.1 F14F9.5 T04A8.15 F56a3.2 M03C11.2 F23C8.9 ctf-4 R04D3.3 F43G6.1 C48D5.1b M04F3.1 C43E11.10 F59E10.1 C54G10.2 rfc-2 rfc-4 F19G12.2 F32A11.4 F49E12.10 W08F4.3 T23G5.6 F41G4.2b R06F6.2 SRXA-6/W07A8.3 F41H10.4 ZK520.3 C37H5.1  RAD24 RAD26 RAD50 RAD51 RAD54 REV1 SGS1 SGS1 SGS1 SGS1 UNG1 XRS2 YKU70 YKU80 no homolog no homolog SLX4 SLX1 CHL1 CSM3 CTF4 DIA2 DNA2 MRC1 no homolog ORC1 ORC2 RFC1 RFC2 RFC4 RNR2 TOP2 ERG25 no homolog no homolog SRV2 PEP5 SWA2 USO1 no homolog no homolog  repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair repair replication replication replication replication replication replication replication replication replication replication replication replication replication replication sterol metabolism sterol metabolism sterol metabolism trafficking trafficking trafficking trafficking trafficking trafficking  Mutator in C.elegans (Y/N) Y  Y  Y  Y Y  Y  Y  Y  Y  172  C. elegans  S. cerevisiae  Function  Mutator in C.elegans (Y/N) Y Y  F15E6.1 Y113G7A.9 fkh-10 F29C4.6 F31C3.2  CTI6 DCS1 HCM1 NCS6 PAP2  transcription transcription transcription transcription transcription  Y56A3A.33  REX3  transcription  Y  R03D7.2  SEN1  transcription  Y  Y111B2A.1  SKY1  transcription  Y  spt-4  SPT4  transcription  cdk-8  SSN3  transcription  F43D2.1  SSN8  transcription  ZK858.1  TRF5  transcription  K09B11.2 rsr-1 B0495.2 Y54E2a.6 xrn-1 R06F6.8 R05D11.4 rpl-2 Y39A3CL.3 B0412.3 H14A12.3  no homolog MUC1 CTK1 SLH1 KEM1 no homolog ROK1 RPL2a no homolog no homolog no homolog  transcription transcription translation translation translation translation translation translation Transmembrane ptn. unknown unknown  Y  Y Y  Y  Y Y Y  173  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0072371/manifest

Comment

Related Items