Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Whole genome mutational profiling of the chemotherapeutic agent mitomycin C Tam, Annie 2014

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

Item Metadata

Download

Media
24-ubc_2014_september_tam_annie.pdf [ 966.57kB ]
Metadata
JSON: 24-1.0167493.json
JSON-LD: 24-1.0167493-ld.json
RDF/XML (Pretty): 24-1.0167493-rdf.xml
RDF/JSON: 24-1.0167493-rdf.json
Turtle: 24-1.0167493-turtle.txt
N-Triples: 24-1.0167493-rdf-ntriples.txt
Original Record: 24-1.0167493-source.json
Full Text
24-1.0167493-fulltext.txt
Citation
24-1.0167493.ris

Full Text

WHOLE GENOME MUTATIONAL PROFILING OF THE  CHEMOTHERAPEUTIC AGENT MITOMYCIN C  by Annie Tam  B.Sc., Simon Fraser University, 2011 (Honours)  A THESIS SUBMITTED IN PARTIAL FULFILLMENT  OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in The Faculty of Graduate and Postdoctoral Studies (Medical Genetics)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  May 2014  © Annie Tam, 2014        ii  Abstract  Cancer therapy largely depends on chemotherapeutic agents that generate DNA lesions. However, our understanding of the nature of the resulting lesions, as well as the mutational profiles of these chemotherapeutic agents, is limited. Among these lesions, DNA interstrand crosslinks are among the more toxic types of DNA damage. Here I have characterized the mutational spectrum of a commonly used DNA interstrand crosslinking agent mitomycin C (MMC). Using a combination of genetic mapping, whole-genome sequencing, and bioinformatics tools, I have identified and confirmed several genomic lesions linked to MMC-induced DNA damage. These results indicate that MMC predominantly causes deletions, with a 5´-CpG-3´ sequence context prevalent in the deleted regions of DNA. Furthermore, I identified microhomology flanking the deletion junctions, indicative of DNA repair via non-homologous end joining. Based on these results, I propose a general repair mechanism that is likely to be involved in the biological response to this highly toxic agent.  In conclusion, the systematic study I have described provides insight into potential sequence specificity of MMC with DNA.                  iii  Preface  The forward genetic screens and dosage experiments with mitomycin C were performed by A.M. Rose. I maintained the Caenorhabditis elegans strains, and performed all of the three-factor mapping and complementation testing. Selection of bioinformatics software and writing of scripts to parse whole genome sequencing data were performed by J.S.C. Chu. Parameters of Pindel and VarScan were mainly determined by J.S.C. Chu. Genomic DNA was prepared by J.S.C. Chu, M. Tian, K. Wong, and F. Muir. Library preparation and whole genome sequencing was performed by the Michael Smith Genome Sciences Centre.  I used Integrative Genomics Viewer (IGV) for visible inspection and curation of sequence data, and I used Pindel and VarScan to identify mutations caused by mitomycin C.                          iv     Table of Contents Table of Contents Abstract ......................................................................................................................................................... ii Preface ......................................................................................................................................................... iii Table of Contents ......................................................................................................................................... iv List of Tables ............................................................................................................................................... vi List of Figures ............................................................................................................................................. vii Acknowledgements .................................................................................................................................... viii Chapter 1: Introduction ................................................................................................................................. 1 1.1 Overview ................................................................................................................................................. 1 1.2 DNA crosslinking: interstrand crosslinks ............................................................................................... 2 1.2.1 General mechanism of DNA interstrand crosslinking agents .......................................................... 2 1.2.2 Basic model of interstrand crosslink repair based on bacterial and yeast studies ............................ 3 1.2.3 Current model of interstrand crosslink repair in mammalian cells .................................................. 5 1.2.4 The Fanconi anemia pathway suppresses interstrand crosslink sensitivity by promoting homologous recombination ....................................................................................................................... 6 1.3 Mitomycin C as a representative of interstrand crosslinking agents ....................................................... 8 1.3.1 Chemistry of DNA crosslinking and alkylation by mitomycin C .................................................... 8 1.3.2 DNA sequence specificity of mitomycin C ..................................................................................... 9 1.3.3 Genotoxic effects of mitomycin C ................................................................................................. 10 1.4 Caenorhabditis elegans as a model to study genomic lesions .............................................................. 11 1.4.1 Genetic balancers used to capture mutations and determine potency of mutagens ....................... 11 1.4.2 Caenorhabditis elegans provides a systematic method to characterize mutagens ......................... 13 1.5 Thesis objective .................................................................................................................................... 14 Chapter 2: Methods ..................................................................................................................................... 16 2.1 Characterization of mitomycin C as a DNA mutagen in Caenorhabditis elegans ............................... 16 2.1.1 Caenorhabditis elegans strains and culture conditions .................................................................. 16 2.1.2 Mutagenesis with varying doses of freshly prepared and stored mitomycin C ............................. 16 2.1.3 Forward genetic screen .................................................................................................................. 17 v  2.1.4 Three-factor mapping of mitomycin C-induced lethal mutations .................................................. 18 2.1.5 Genetic complementation analysis ................................................................................................. 20 2.2 Mutational consequences of mitomycin C treatment ............................................................................ 20 2.2.1 Genomic DNA extraction .............................................................................................................. 20 2.2.2 Illumina HiSeq sequencing platform ............................................................................................. 21 2.2.3 Data processing and bioinformatic analysis ................................................................................... 22 2.2.4 Physical identification of putative lethal mutation sites ................................................................ 23 2.2.5 Verification of lethal mutation sites by complementation testing and Sanger sequencing ............ 23 Chapter 3: Results ....................................................................................................................................... 26 3.1 Forward genetic screen and genetic mapping of mitomycin C-induced lethal mutations .................... 26 3.1.1 Forward mutation frequency of mitomycin C-induced lethals ...................................................... 26 3.1.2 Genetic mapping of mitomycin C-induced lethal mutations ......................................................... 27 3.1.3 Complementation tests to optimize genetic location of mitomycin C-induced lethal mutations ... 30 3.2 Characterizing mitomycin C-induced lesions in the hT2 balanced chromosomes ................................ 31 3.2.1 Identifying mitomycin C-induced lethal mutations using three-factor mapping data in conjunction with bioinformatics ................................................................................................................................. 31 3.2.2 Validation of the putative mitomycin C-induced lethal mutations ................................................ 32 3.3 Global view of mitomycin C-induced damage to the C. elegans genome ............................................ 34 3.3.1 Extending analysis of mitomycin C-induced deletions to the rest of the genome ......................... 34 3.3.2 Genome-wide distribution of mitomycin C-induced single nucleotide variants ............................ 37 3.3.3 Putative insertions induced by mitomycin C ................................................................................. 38 Chapter 4: Discussion ................................................................................................................................. 40 4.1 Overview ............................................................................................................................................... 40 4.2 Research conclusions ............................................................................................................................ 40 4.2.1 Summary of findings ...................................................................................................................... 40 4.2.2 Significance of findings ................................................................................................................. 46 4.2.3 Future directions ............................................................................................................................ 48 References ................................................................................................................................................... 49 Appendices .................................................................................................................................................. 55 Appendix A: Supplementary material for chapter 2 ............................................................................... 55 Appendix B: Supplementary material for chapter 3 ............................................................................... 58   vi  List of Tables  Table 1.1: Commonly used DNA cross-linking agents and their target sequences in an in vitro system…………………………………………………………………………………..... 3 Table 2.1: Summary of varying doses and ages of mitomycin C used in mutagenesis……….... 17 Table 2.2: List of known alleles used to verify the putative lethal mutation sites…………….... 24 Table 3.1: Forward mutation frequencies for recessive lethal mutations…………………….… 27 Table 3.2: Complementation testing using the deficiency hDf10…………………………….… 30 Table 3.3: Mitomycin C induced deletions of varying sizes…………………………………… 31 Table 3.4: Validation of mitomycin C-induced lethal mutation sites using complementation testing and Sanger sequencing………………………………………………………….. 33 Table 3.5: Genome-wide sequence context of mitomycin C-induced deletions……………...... 35 Table 3.6: Putative mitomycin C-induced insertions identified by Pindel……………………... 39                                vii  List of Figures  Figure 1.1: Basic model of interstrand crosslink repair in bacteria and yeast…………………… 4 Figure 1.2: Current model for mammalian replication-independent and replication-independent interstrand crosslink repair……………………………………………………………….. 6 Figure 1.3: Mitomycin C-DNA interactions are affected by sequence context………………… 10 Figure 1.4: Structure of the specialized chromosome hT2(I;III)...……………………………... 13 Figure 2.1: Schematic of the forward genetic screen used to isolate lethal mutation-bearing strains……………………………………………………………………………...……. 18 Figure 3.1: Genetic map of mitomycin C-induced lethal mutations in the hT2-balanced region of chromosome I…………………………………………………………………………… 29 Figure 3.2: Putative mitomycin C-induced deleted regions of DNA were enriched for 5´-CpG-3´ dinucleotides……………………………………………………………………………. 36 Figure 3.3: Genome-wide profile of mitomycin C-induced single nucleotide variants, compared to N2 and EMS-treated animals……………………...…………………………………. 38  Figure 4.1: Possible mechanism for the generation of deletions with flanking sequence microhomologies………………………………………………………………...……… 47                             viii  Acknowledgements I would like to thank everyone who has helped me with this project. In particular, I would like to thank my supervisor, Dr. Ann M. Rose, for entrusting me with this great project, and for providing me with the utmost support and guidance throughout my program. Thank you to my committee members Dr. Donald G. Moerman and Dr. Steven J.M. Jones for their time, and invaluable advice. I would also like to thank Dr. David L. Baillie for insightful and thought-provoking conversations. Special thanks to past, present, and honorary Rose lab members: George Chung, Dr. Jeffrey S.C. Chu, Dr. Martin R. Jones, Dr. Ann Marie Davison, Shu-Yi Chua, Jason Luce, Meng Tian, Kathy Wong, Fraser Muir, and Jessica Wang. Thanks, and lots of love to my family.                         1  Chapter 1: Introduction 1.1 Overview  Tumourigenesis is driven by the accumulation of somatic mutations. These mutations may arise during development, from environmental or exogenous factors, or as a result of defective DNA repair pathways. Cancer-initiating mutations can be caused by exposure to exogenous mutagens, such as ultraviolet light resulting in skin cancer or tobacco smoking in lung cancer [1]. Abnormalities in DNA maintenance pathways may lead to colorectal cancer or the cancer predisposition syndrome Fanconi anemia [2, 3]. Chemotherapeutic agents, though widely used to treat cancers, are exogenous mutagens themselves causing various types of mutations in order to limit or halt tumour growth. Chemotherapeutic agents can target DNA to generate lesions, such as DNA adducts, interstrand and intrastrand crosslinks, and DNA double strand breaks. Among chemotherapeutic agents, DNA crosslinking agents are considered the more toxic types of DNA damaging drugs [4, 5]. However, our understanding of the lesions, and the mutational profile generated by these agents, is limited. With modern technological advances, it is possible to precisely and rapidly sequence entire genomes, and determine the mutational spectra induced by chemotherapeutic agents [6, 7]. For this reason, I am analyzing genomes exposed to a DNA crosslinking agent mitomycin C to systematically characterize the spectrum of mutagenic lesions. Identification of the nature and frequency of mutagenic lesions caused by DNA crosslinking agents will not only further our understanding of this widely used drug, but may also provide insight into the molecular mechanisms required for DNA crosslink repair.   2  1.2 DNA crosslinking: interstrand crosslinks  DNA crosslinking agents are highly potent chemotherapeutic drugs. In particular, DNA interstrand crosslinks (ICLs) are among the more toxic forms of DNA damage [4, 5]. ICLs are lesions that covalently link two bases on different strands of DNA. This prevents DNA strand separation, and critical cellular processes such as replication and transcription are blocked [8]. Since both DNA strands are involved, ICLs present a challenge for DNA repair, and it is clear that ICL formation activates many cellular responses such as cell cycle arrest, DNA damage repair, and apoptosis [9]. A single ICL can be lethal to a repair-deficient bacteria or yeast cell, whereas 20-40 ICLs can kill repair-deficient mammalian cells [10, 11]. Due to the potent cytotoxic nature of crosslinking agents, this class of drug is extensively used in cancer therapy. However, ICLs can induce mutations and rearrangements of DNA that not only affect microevolution of cancer cells, but also the genetic integrity of normal non-cancerous cells [12]. Therefore, there is direct clinical relevance in understanding the nature of lesions induced by interstrand crosslinking agents.  1.2.1 General mechanism of DNA interstrand crosslinking agents  A variety of agents are capable of inducing DNA crosslinks in the genome. Commonly used agents include platinum compounds, nitrogen mustards, and other chemicals such as mitomycin C, psoralen, and nitrosoureas [13]. While the clinical applications of these agents vary, the mechanism of crosslinking drugs is broadly similar. Interstrand crosslinks (ICLs) form when a bifunctional crosslinking agent reacts with two nucleotides on opposite strands of DNA. This reaction requires a geometric alignment of the bifunctional crosslinking agent with DNA. In addition to this geometric specificity, in vitro studies using synthesized oligonucleotide duplexes 3  containing ICLs indicate these crosslinks are also likely to be sequence specific as well (Table 1.1) [14-18]. However, the nature of these lesions in vivo is not well understood, and the frequency and distribution of mutations is not known.   ICL-inducing agent Target sequence References Platinum compounds 5´-GC-3´ [14] Nitrogen mustards 5´-GNC-3´ [15] Mitomycin C 5´-CG-3´ [16] Psoralen 5´-TA-3´ [17] Nitrosoureas G-C base pair [18] Table 1.1: Commonly used DNA cross-linking agents and their target sequences in an in vitro system.   1.2.2 Basic model of interstrand crosslink repair based on bacterial and yeast studies  The mechanisms involved in interstrand crosslink (ICL) repair are highly conserved in most prokaryotes. The major ICL repair pathway in Escherichia coli has been well characterized both genetically and biochemically [19, 20]. A similar repair mechanism is also seen in budding yeast [21]. Genetic studies of epistasis with bacterial and yeast mutants that were sensitive to a range of DNA damaging agents identified three major groups of genes involved in ICL repair [22-24]. These groups of genes are involved in nucleotide excision repair (NER), homologous recombination (HR), and translesion synthesis (TLS). Models have been proposed to explain recombination-dependent, error-free repair, and recombination-independent, error-prone repair in bacteria and yeast (Figure 1.1) [25, 26]. To initiate these parallel pathways, the NER machinery is required to initiate strand-unhooking by introducing incisions flanking the ICL lesion. The 4  ICL-intermediate is processed by two parallel pathways involving either error-free HR or error-prone TLS. Both pathways result in DNA synthesis across the ICL, thereby filling the gap created by the initial NER incisions. In both pathways, NER is co-opted to remove the ICL adduct left behind after incision.   Figure 1.1: Basic model of interstrand crosslink repair in bacteria and yeast. This basic model of interstrand crosslink repair is based on bacterial and yeast studies. Two parallel pathways involve nucleotide excision repair (NER), homologous recombination (HR), and translesion synthesis (TLS). The NER pathway recognizes the ICL, and initiates repair by introducing incisions around the lesion. The unhooked repair intermediate is then processed by the error-prone TLS pathway or the error-free HR pathway. Both independent pathways end with a subsequent round of NER to fully remove the crosslink. Yellow crooked line indicates an interstrand crosslink. Scissors indicate incisions by nucleases. Adapted from Cole, R.S 1973, and McHugh et al, 2001 [25,26].  ICL unhookingHRDNA templateNERTLSNERNERGap fillingGap filling5  These are the basic elements of a repair pathway found in all organisms: recognition of the ICL, followed by unhooking, gap formation and repair by recombination, or lesion bypass (See Muniandy et al, 2011 for review) [27]. Although this process would appear to provide a general solution to ICL repair, there are major variations in yeast and mammalian cells.  1.2.3 Current model of interstrand crosslink repair in mammalian cells  The basic model of interstrand crosslink (ICL) repair in bacteria and yeast is also relevant to mechanisms in higher eukaryotes. However, current understanding of ICL repair in mammalian cells is complicated by the presence of multiple overlapping DNA repair pathways. Different ICL repair pathways are thought to be activated depending on the phase of the cell cycle [28, 29]. Replication-dependent and replication-independent ICL repair is highlighted in Figure 1.2. The replication-dependent model is based on biochemical experiments using plasmids with site-specific ICLs and a defined replication system in Xenopus egg extracts [30-32]. The replication-independent model is based on experiments observing ICL repair using plasmids with site-specific ICLs that have no replication origins [33-37]. Like ICL repair in bacteria and yeast, repair in mammalian cells involves sequential excision of the crosslink from one strand of DNA, followed by the next strand [38]. This prevents the creation of multiple double-strand breaks (DSBs), which may drive the cell towards an end-joining pathway. During replication, the ICL is thought to be unhooked from the lagging strand template by introduction of two sequential incisions 5´, then 3´ of the incision (Figure 1.2) [39]. Once the crosslink is unhooked, translesion synthesis (TLS) extends the leading strand past the unhooked ICL, creating a duplex molecule that can now undergo homologous recombination. It is not known at what point the unhooked crosslink is removed, and it is not known if the crosslink is removed via 6  nucleotide excision repair (NER), hydrolysis, or another method. In non-replicating cells, ICLs are thought to be repaired via unhooking of the ICL from one strand of DNA, possibly by nucleotide excision repair, followed by TLS to fill the resulting gap [33-37, 40].    Figure 1.2: Current model for mammalian replication-independent and replication-independent interstrand crosslink repair. See text for details. Yellow lines represent the lesion caused by an interstrand crosslink. Scissors indicate incisions by nucleases. Adapted from Clauson et al, 2013 [41]. 1.2.4 The Fanconi anemia pathway repairs interstrand crosslinks by promoting homologous recombination  The current model of the predominant interstrand crosslink (ICL) repair pathway in mammalian cells is highlighted in Chapter 1.2.3. Homologous recombination (HR) occurs in Replication fork approaches the ICLSequential incisions flanking the crosslink1 2Translesion synthesis across the lesionNucleotide excision repair?Hydrolysis?5’3’Homologous recombinationReplication dependentReplication independentRecognition:During transcription or Helix distortionNucleotide excision repairTranslesion synthesisNucleotide excision repairGap filling7  S/G2 phases of the cell cycle, utilizes an undamaged homologous sequence as a repair template, and is thus a more precise method of repairing double-strand breaks (DSBs). Throughout the cell cycle, non-homologous end joining (NHEJ) can compete with HR to fix DSBs. NHEJ is an efficient pathway that functions to fix DSBs by ligation of DNA ends, which can lead to error-prone repair. NHEJ involves the joining of DNA ends with no or little homology (also called microhomology). Small sequence microhomologies are thought to help guide NHEJ-mediated repair [42]. NHEJ involves recruitment of Ku proteins, which bind to exposed DNA ends following a DSB, preventing end resection, a prerequisite of HR [43]. A number of processing factors are recruited to the site of damage, joining end structures together [44]. Experiments have demonstrated that factors involved in HR and NHEJ most likely compete to fix lesions. It has been shown that loss of NHEJ factors promotes end resection, and thus HR [45]. It has also been demonstrated that when Fanconi anemia (FA) pathway components are mutated, there is an increase in NHEJ-mediated events [46-48]. The FA pathway promotes HR-dependent stabilization of the replication fork and DNA repair, and FA mutants have been shown to be defective in HR in the repair of DSBs [49]. In addition, cells deficient for the FA repair pathway components are extremely sensitive to crosslinking agents, and show increased formation of radial chromosomes, possibly as a result of aberrant NHEJ  [49, 50]. Due to the one ended nature of the DSBs that arise as a consequence of ICL repair, the NHEJ components do not have a natural substrate available to rejoin the breaks. Consequently, breaks can remain unrepaired, leading to deletions resulting from direct ligation of the broken ends of DNA, or the formation of radial chromosomes through ligation between DSBs in different chromosomes [46-48]. To combat the ongoing threat of DNA damage, eukaryotic cells have evolved overlapping DNA 8  repair pathways that work parallel to, or in concert with one another. The coordinated steps of these different DNA repair pathways play important roles in maintaining genomic stability.  1.3 Mitomycin C causes interstrand crosslinks  The mitomycins are a family of antibiotics isolated from the microorganism Streptomyces caespitosus [51]. One member of the family, mitomycin C (MMC), finds use as a cancer chemotherapy by virtue of its broad spectrum activity against solid tumours [52-56]. In the clinic, MMC is widely used as a chemotherapeutic agent to combat breast, lung, cervical, colorectal, and bladder cancers, amongst others [52-56]. In addition to its potent antitumour activity, MMC has a variety of biological effects including induction of DNA repair, inhibition of DNA synthesis, chromosome breakage, and sister chromatid exchange [57, 58]. These biological effects are attributed to the potent DNA crosslinking ability of MMC [59]. Outside of its use in cancer chemotherapy, MMC is also clinically used as a diagnostic criterion for Fanconi anemia, a condition marked by early onset cancer and genomic instability. Other clinical uses of MMC include usage in laser eye surgery and treatment of glaucoma [60, 61]. MMC is a widely used DNA crosslinking agent with various clinical applications. Therefore, there is clinical relevance in characterizing the mutational signature of MMC as a starting point for studying DNA interstrand crosslinking agents.     1.3.1 Chemistry of DNA crosslinking and alkylation by mitomycin C  A unique feature of mitomycin C (MMC) is that the molecule must first undergo reduction in the cell to become reactive with DNA [57]. Upon reduction, the reduced MMC 9  substrate generates DNA-reactive species that can crosslink DNA due to its bifunctional alkylating property [57]. Although MMC can also undergo monofunctional alkylation with DNA (attachment of the drug molecule to one strand of DNA), it has been shown that the antitumour activity of MMC is predominantly due to inhibition of DNA replication caused by DNA crosslinking [57, 62]. It has been demonstrated that on exposure to MMC and a reducing agent in cell-free systems, DNA interstrand crosslinks (ICLs) readily form [57]. In vitro methods of ICL detection include gel electrophoretic separation of crosslinked and non-crosslinked DNA or oligonucleotides, high-performance liquid chromatography, and other chromatography methods [63-65]. Molecular modeling studies have shown that the MMC molecule binds to the minor groove of DNA, both in the monoadduct and crosslinked forms [66, 67]. This binding of MMC with DNA causes very little perturbation to the structure of the duplex DNA [68]. While there is significant understanding of the chemical nature of MMC-DNA interactions, the efficiency of MMC activation in a cellular environment is poorly understood.  1.3.2 DNA sequence specificity of mitomycin C  Multiple studies have explored the use of in vitro assays to determine possible methods in which mitomycin C (MMC) may preferentially interact with DNA. The current understanding indicates MMC preferentially interacts with guanines, and that this reactivity is affected by the sequence context [63]. Studies that have investigated MMC-DNA monoalkylation formation indicate that the preferred sites are predominantly in the 5´-CpG-3´ sequence context as opposed to the other 5´-NpG-3´ dinucleotide sequence, though 5´-GpC-3´ may also be observed [69]. Furthermore, it has been shown that the 3´-flanking base only has a small effect on the CpG specificity [69]. Formation of the DNA interstrand crosslink – the second step following monoalkylation formation – is specific for the same 5´-CpG-3´ sequence (Figure 1.3) [70]. This 10  sequence specificity of DNA crosslinking has been attributed to the chemistry of MMC [69]. It has been proposed that in the first step of the reaction, the active monoalkylated MMC molecule forms a specific hydrogen bond with the 2-amino group of the opposite-strand guanine, stabilizing the structure [69]. In support of this hypothesis, NMR-derived structures of the DNA monoadduct show the presence of this hydrogen bond [71].   Figure 1.3: Mitomycin C-DNA interactions are affected by sequence context. In vitro studies indicate that mitomycin C (MMC) preferentially interacts with guanines in the 5´-CpG-3´ sequence context due to its chemical properties. In step 1, a monoadduct can form with any guanine, independent of sequence context. In step 2, an interstrand crosslink can only form in a 5´-CpG-3´ sequence context due to the availability of the 2-amino group of the opposite-strand guanine. Yellow rectangle denotes the MMC adduct.  1.3.3 Genotoxic effects of mitomycin C  The DNA crosslinking ability of mitomycin C (MMC) was first discovered in bacteria [59]. Subsequently, the biological effects of MMC were studied in a variety of different organisms. In studies using Drosophila melanogaster as a model organism, it was found that MMC changed the pattern of recombination, and also induced X-Y chromosome and Y-Y chromosome interchanges [72, 73]. Fluorescence Plus Giemsa (FPG) experiments using Chinese hamster ovary (CHO) cells and mouse bone marrow cells demonstrated that MMC can induce Step 1: MonoadductSt p 2: Crosslink11  sister-chromatid exchanges (SCEs) [74, 75]. In both studies, there were differences in the persistence of the lesions caused by MMC, which has been suggested to be due to the intrinsic differences in the organism’s initial sensitivity to lesion induction, and the ability to remove and repair DNA lesions [76, 77]. In the mouse system, MMC was shown to induce deletions and tandem-base substitutions in the genome [78]. In this study, the genetic analysis of bone marrow from transgenic mice revealed the presence of isolated deletions and point mutations in lambda DNA that have been integrated into the genome. An analysis of this data revealed that larger deletions were frequently flanked by two short direct repeat sequences 2 to 6 base pairs long, suggesting a type of end-joining repair of double-strand breaks [78]. From the existing literature, it is therefore clear that MMC can induce a wide range of genetic lesions. However, the biological frequency of these lesions has never been studied with high resolution methods. To understand the biological consequences of clinically used mutagens, it is imperative to first characterize the spectrum of mutagenic changes caused by these agents in a biological context.  1.4 Caenorhabditis elegans as a model to study genomic lesions  The nematode Caenorhabditis elegans is a well characterized genetic model system with which to study genomic damage generated by exposure to crosslinking agents. In this section, I will describe some of the established methods that have been used to capture and assay mutational events in a biological context in this system.  1.4.1 Genetic balancers used to capture mutations and determine potency of mutagens  Caenorhabditis elegans presents an excellent genetic model due to the ease of capturing and maintaining specific mutations using specialized chromosomes, also known as genetic 12  balancers (Reviewed in Jones et al, 2011) [79]. These specialized chromosomes include deletions, chromosomal insertions, reciprocal translocations, extrachromosomal arrays, complex rearrangements, as well as transgenic constructs. The term “genetic balancer” was first used in Drosophila genetics to describe a method of using a chromosomal rearrangement to maintain a lethal mutation [80]. Lethal mutations are mutations that cannot be maintained as homozygotes, because the mutated locus is essential for the animal’s survival. A major class of genetic balancer is those that reduce or prevent recombination between the wild-type homologue and mutation-bearing chromosome. In this class, the reciprocal translocation balancers are used to balance two different chromosomes. A segment of DNA from one region of the chromosome is displaced, and can no longer pair and recombine with its homologous regions. Therefore, a lethal mutation captured in this balanced region can be stably maintained, since recombination is suppressed.  Characterization of the reciprocal translocation, eT1, demonstrated that the non-homologous balanced arms do not genetically pair or recombine in animals with the reciprocal translocation [81, 82]. In the case of the genetic balancer hT2, the right portion of chromosome III translocated to the right of chromosome I, disjoining from the normal chromosome I, while the left of chromosome I translocated to the left of chromosome III, disjoining from the normal chromosome III (Figure 1.4) [83]. The resulting reciprocal translocation balances the left portion of chromosome I from the left end through unc-101, and the right portion of chromosome III from the right end through dpy-17 [83]. To maintain the translocation as a heterozygote, the normal homologues are marked with mutations with visible phenotypes in the recombination-suppressed region. For example, the strain KR4949 used in this study carries the genotype dpy-5 unc-13 / hT2, where dpy-5(e61) and unc-13(e51) were used as visible markers. 13   Figure 1.4: Structure of the specialized chromosome hT2(I;III). The specialized chromosome hT2 results from a reciprocal translocation between the left half of chromosome I and right half of chromosome III. The dotted lines represent the homologue recognition regions (HRR) [83, 84], the regions of homologous chromosomes that pair. Recombination is suppressed where non-homologous chromosomes are aligned. Black: chromosome I. Blue: chromosome III. (Adapted from Jones et al, 2011 and McKim et al, 1993) [79, 83].  There are practical applications of using genetic balancers in C. elegans genotoxicity assays. Multiple experiments have shown its efficacy of determining the forward mutation frequency of mutagens, which is a measure of the total number of lethal events recovered by the balancer under defined conditions. These experiments include investigating the mutagens ethylmethane sulfonate (EMS), formaldehyde, ultraviolet irradiation, gamma irradiation, cosmic radiation, and also accumulated mutations from mutator strains [85-90]. 1.4.2 Caenorhabditis elegans provides a systematic method to characterize mutagens  C. elegans has been extensively used as a model to assay genotoxicity. Genetic studies indicate C. elegans is a very useful system to determine the dose-response to mutagens such as formaldehyde, a DNA crosslinking agent [85, 91]. Complementation testing with deficiencies demonstrated formaldehyde can induce putative point mutations, deletions, and complex rearrangements [85]. However, technology at the time prevented verification of these mutations. Large-scale studies employing whole genome sequencing have been shown to be feasible to 14  identify point mutations in ethyl methanesulfonate (EMS), N-ethyl-N-nitrosourea (ENU), and ultraviolet trimethylpsoralen (UV-TMP)-mutagenized C. elegans strains, and when coupled with comparative genomic hybridization, could also be used to find deletions [92, 93]. In addition, bioinformatic approaches have been developed for identifying lethal mutations in a heterozygous background [96].  In this study, I used whole genome sequencing and a suite of bioinformatics tools to identify MMC-induced mutations. This study provided a high resolution, global view of the genomic events induced by a mutagen. This process was simplified by the inherent biological properties of the model organism C. elegans. C. elegans has a low generational time, allowing for quick mutagenesis and mutation capture, well-defined genetics, as well as a smaller genome size to expedite whole genome sequencing and analysis. 1.5 Thesis objective  The aim of my project was to characterize the mutational spectrum of mitomycin C (MMC), a widely used DNA crosslinking agent. Specifically, my main goals were 1) to assess the mutagenic potency of MMC; 2) to use whole genome sequencing to identify the nature of lesions caused by MMC in terms of their frequency and distribution; and 3) to determine possible mechanisms by which MMC-induced lesions arise, which may lead to general insight into interstrand crosslink repair. Previously, the methods used to identify structural variants such as insertions, deletions, and copy number variations have been poorly understood. Here, I describe a two-part, overlapping method to identify variants that have been caused by MMC. In this study, I parsed two genomic regions in the same strains to determine the nature of MMC-induced lesions. In the first method, I generated and genetically mapped recessive lethal mutations using the genetic balancer hT2 to capture the state of the genome immediately after 15  mutagenesis. The chromosomes balanced by hT2 were maintained in a heterozygous state, and therefore these regions represent a catalogue of the extent of damage caused by MMC. In the second method, I extended my analysis to the rest of the genome, cataloguing mutations that were maintained in a homozygous state. Using these complementary methods, the total extent and consequence of MMC-induced DNA damage are described here.                   16   Chapter 2: Methods  2.1 Characterization of mitomycin C as a DNA mutagen in Caenorhabditis elegans  2.1.1 Caenorhabditis elegans strains and culture conditions  Wild-type and mutant C. elegans strains were cultured in Petri dishes on agar nematode growth medium (NGM) streaked with Escherichia coli OP50 [94]. C. elegans were maintained at 20ºC as previously described [94]. The nomenclature for genes and alleles follows the uniform system adopted for C. elegans in Horvitz et al (1979) [95]. Strains were obtained from the Caenorhabditis Genetics Center (CGC) unless otherwise indicated. The genetic balancer hT2(I;III) used in this study was induced by gamma irradiation [83]. The deficiency hDf10 was isolated in a screen for lethal mutations using hT2(I;III) as a balancer [83], and a dominant pharyngeal GFP marker from an insertion of a transgene that expresses Pmyo-2::GFP was subsequently incorporated. The region balanced by hT2(I;III) has previously been identified to be from the left of chromosome I to between unc-101 and unc-59, and right of chromosome III to dpy-13 [83, 84]. All mutations denoted with the h prefix originated from the Rose laboratory. 2.1.2 Mutagenesis with varying doses of freshly prepared and stored mitomycin C  The chosen mitomycin C (MMC) concentration range was based on Schewe, Suzuki and Erasmus’ (1971) findings in Drosophila melanogaster [72, 73]. They showed dosages of 374 µM MMC in females and 1496 µM MMC in males were sufficient to induce lethal mutations [72]. Two 2 mg vials of MMC were prepared (summarized in Table 2.1). In the first batch, freshly prepared, stored 1 week, and stored 2 weeks MMC were prepared by first dissolving in 200 µL 17  dH2O, followed by dilution with M9 buffer to give a final concentration of 750 µM. In the second batch, MMC was dissolved in dH2O, followed by dilution with M9 buffer to give the following concentrations: 375 µM, 750 µM, and 1400 µM.  Vial number Age of vial Dose 1 Fresh 750 µM 1 Stored 1 week 750 µM 1 Stored 2 weeks 750 µM 2 Stored 1 week 375 µM 2 Stored 1 week 750 µM 2 Stored 1 week 1400 µM Table 2.1: Summary of varying doses and ages of mitomycin C used in mutagenesis.  KR4949 nematodes of genotype hT2[bli-4(e937)] let-? (q782) qIs48] I; III  / unc-13 (e51); dpy-5(e61), subsequently referred to as dpy-5 unc-13 / hT2 were washed off plates with M9 buffer. The animals were collected by centrifugation and soaked with the concentrations of MMC summarized in Table 2.1 for 4 hours at 20ºC. The animals were removed from solution and placed onto fresh Petri plates. After 3 days, the F1 progeny were screened for the absence of Dpy-Uncs indicating lethal mutations captured in the hT2-balanced region.  2.1.3 Forward genetic screen  Single F1 animals from the P0 hermaphrodite plates were placed onto individual Petri plates. The progeny of the F1 animals were screened for recessive lethal mutations using the method summarized in Figure 2.1.  18    Figure 2.1: Schematic of the forward genetic screen used to isolate lethal mutation-bearing strains. dpy-5 unc-13 / hT2 P0 animals were mutagenized with mitomycin C (MMC), and lethal mutations were screened in the progeny of the F1 animals. Blue bars indicate the normal chromosomes. Purple bars indicate balanced regions. Red ticks indicate the visible markers dpy-5 and unc-13, as well as possible MMC-induced lethal mutations.  Due to linkage with the visible markers, the absence of mature Dpy-Unc animals indicated the presence of a lethal mutation in the balanced region of chromosome I. If a lethal mutation was induced in the balanced region of chromosome III, there would be an absence of mature Dpy-Unc animals due to pseudolinkage with the visible markers. Mitomycin C-induced lethal mutations in the unbalanced regions will die as homozygotes and be lost.  Mutant strains were established by picking and maintaining heterozygotes that did not segregate mature Dpy-Unc animals. Isolated lethal mutation-bearing strains were frozen for further analysis.  2.1.4 Three-factor mapping of mitomycin C-induced lethal mutations  Three-factor mapping was used to determine the genetic location of the lethal mutations. The dpy-5 unc-13 / hT2 lethal mutation-bearing strains were crossed to N2 males. Non-GFP L4 hT2(I) hT2(III)P0If lethal mutation occurred in the balanced regions:hT2(I)hT2(I) hT2(III)hT2(III)Mitomycin C treatmentlethal mutation lethal mutationlethal mutation lethal mutationIf lethal mutation occurred in the unbalanced regions:orSelf-fertilization Self-fertilizationMutation is lost when homozygousLethal mutation maintained as heterozygoteF1orPhenotype: wild-type animals Phenotype: wild-type or Dpy-Unc animals19  F1 hermaphrodites were picked from each strain and transferred daily over four days. In each brood, the wild-type, Dpy-Unc, Dpy, and Unc animals were scored. Genetic distance of the putative lethal mutations were calculated using the equation for recombination frequency  (p) = 1 - √(1-2R) [94], where R denotes the fraction of 2 x Dpy-Uncs (recombinants) over 4/3 wild type (to calculate total progeny). Using an example where the lethal mutation was induced in cis to the dpy-5 unc-13 markers, the derivation of this formula is as follows: The + symbol denotes wild type, and let denotes the lethal mutation. let dpy-5 unc-13 + + +  The predicted progeny genotypes are:  let d u + + + let + + + d u let d + + + u let d u Lethal Wild-type Lethal Dpy-Unc Lethal Unc + + + Wild-type Wild-type Wild-type Wild-type Wild-type Wild-type let + + Lethal Wild-type Lethal Wild-type Lethal Wild-type + d u Dpy-Unc Wild-type Wild-type Dpy-Unc Dpy Unc let d + Lethal Wild-type Lethal Dpy Lethal Wild-type + + u Unc Wild-type Wild-type Unc Wild-type Unc   R = % Recombinant animals =      Number of Dpy-Unc animals      _                                               Total number of progeny  But the total number of progeny could not be scored because lethal homozygotes resulted in arrested progeny. Only fertile Dpy-Uncs and wild types could be scored, therefore the adjusted calculation of % recombination (R) is:   R =      2 x Number of Dpy-Unc animals      _    Total number of wild types x 4/3  The distance between dpy-5 and unc-13 is approximately 2 map units (equal to a frequency of 0.02). Confidence limits of 95% were calculated using binomial statistics.  20  2.1.5 Genetic complementation analysis  Complementation testing was utilized to narrow down the genetic location of the lethal mutations balanced by chromosome I. Lethal-mutation bearing strains were crossed to the deficiency hDf10 using the strain, KR4633 (hDf10[unc-29 dpy-5] / hT2[bli-4(e937)] let-? (q782) qIs48]). hDf10 is balanced by hT2 and carries the markers dpy-5 and unc-29. Lethal-mutation bearing strains were crossed to N2 males, and the non-GFP (non-hT2) heterozygous males were crossed to hDf10. The mated hDf10 hermaphrodites were brooded, and the F1 progeny were scored for wild-type and Dpy hermaphrodites and males. The absence of Dpy male progeny indicated a failure to complement, since the strains carrying hDf10 also carry a dpy-5 mutation. 2.2 Mutational consequences of mitomycin C treatment 2.2.1 Genomic DNA extraction  Eleven lethal mutation-bearing strains KR4968, KR4969, KR4978, KR4984, KR4995, KR5006, KR5009, KR5034, KR5035, KR5037, and KR4992 were prepared for whole genome sequencing. These strains carried the lethal mutations h2717, h2718, h2727, h2733, h2744, h2755, h2758, h2784, h2785, and h2787 on chromosome I, and h2741 on chromosome III. The following pipeline was previously described in Chu et al, 2012 [96]. Briefly, each strain was grown on ten Petri dishes on agar NGM streaked with Escherichia coli OP50 until food depletion (approximately 5 days at 20ºC). Worms were collected and pelleted by washing plates with M9 buffer, followed by centrifugation for 1 minute at 1500 x g. The pellet was washed three times with M9 buffer, followed by 2-3 hrs of incubation at 20ºC. Worms were then pelleted and resuspended in 0.5 mL TE buffer. The worms were frozen at -20ºC and lysed in lysis solution (50 µL 5% SDS, 2.5 µL 20 mg/mL proteinase K) at 60ºC for 2 hr. Genomic DNA was 21  purified using phenol/chloroform extraction followed by ethanol precipitation. Each DNA sample was treated with 4 µL of 5 mg/mL RNase A for 1 hr at 37ºC. A second round of phenol/chloroform extraction and ethanol precipitation was used to ensure purity. Approximately 10 µg of DNA was sheared using Sonic Dismembrator 550 (Fisher Scientific) at power setting “7” for 30 sec pulse-cooling cycles for 10 min and subsequently analyzed on an 8% PAGE gel. A 180- to 220-bp DNA fraction was excised from the gel slice and eluted in 300 mL elution buffer [5:1, LoTE buffer (3 mM Tris–HCl, pH 7.5, 0.2 mM EDTA):7.5 M ammonium acetate] overnight at 4ºC. The DNA fraction was purified using a Spin-X Filter Tube (Fisher Scientific) and by ethanol precipitation. The whole genome shotgun sequencing library was prepared using an Illumina paired-end protocol. The modified protocol involved DNA end-repair, formation of overhangs using Klenow fragment, and ligation to Illumina adapters. The adapter-ligated products were purified on Qiaquick spin columns (Qiagen) and PCR-amplified using Phusion DNA polymerase for 10 cycles using the PE primer 1.0 and 2.0 (Illumina). PCR products of the desired size range were purified using an 8% PAGE gel. DNA quantity and quality was assayed using Nanodrop 7500 spectrophotometer (Nanodrop). DNA was diluted to 20 nM, and the final concentration was confirmed using Qubit fluorometer (Invitrogen). 2.2.2 Illumina HiSeq sequencing platform  The Illumina HiSeq 2000 platform was used for whole genome sequencing at the Michael Smith Genome Sciences Centre, Vancouver, B.C, Canada. Clusters were generated on the Illumina cluster station and paired-end reads were generated following the manufacturer’s instructions. The V1.0 Illumina Genome Analyzer analysis pipeline was used for image analysis, base calling, and error calibration. 22  2.2.3 Data processing and bioinformatic analysis  Each genomic sequence of the 11 strains was aligned to the annotated sequence of C. elegans available on WormBase (WS200) (http://www.wormbase.org) using BWA at the default setting [97]. Each genomic sequence was compared to the other 10 sequenced strains, as well as a composite parental strain to identify strain specific events. The composite parental strain was derived from combining the bam files of two strains used to make the parental strain, hDf10 dpy-5 unc-29 / hT2 and dpy-5 unc-13 / dpy-5 unc-13 (combined by Chu, J.S.C). Integrative Genomics Viewer (IGV) was used to visualize, browse, and analyze genomes [98]. Single nucleotide variants (SNVs) were called using VarScan with the following parameters: –min-coverage 3 –min-avg-qual 5 –p-value 0.1 –str-filter 0 –min-freq-homozy 0.9 [99]. A custom script was used to parse each strain for strain-specific SNVs that were only present in one strain and absent in the other sequenced strains and parental strain. The distribution and number of SNVs were compared to 10 spontaneously mutating N2 strains, using raw data published in Denver et al, 2009, and also 11 EMS-treated strains [100]. P-values of the distribution of SNVs were derived from chi-square statistics applied to the distribution of SNVs (p<0.05). Insertions and deletions were called using Pindel with default parameters, and set to filter insertions and deletions (indels) occurring in homopolymer regions [101]. A custom script was used to parse each strain for homozygous, strain-specific indels (absent in the composite parental strain and absent in all other sequenced MMC-treated strains). The strain specific indels were independently verified on IGV and the left and right 15 flanking nucleotides (greater than 1 helical turn of DNA) were extracted to provide regional context of insertion and deletion lesions. 23    2.2.4 Physical identification of putative lethal mutation sites  The physical location of each mutation was estimated from the genetic location determined by three-factor mapping. The range (confidence limits of 95% calculated using binomial distribution statistics applied to the map units) was determined for each strain, and the physical location of the closest known genes mapped to those regions was used to estimate the starting and ending point of analysis. The putative lethal mutation sites were identified through IGV, searching for any unique mutations in the mapped region (an example of a mutation identified with IGV is shown in Figure S2.1). This was used to test the sensitivity of Pindel, and to verify that Pindel could detect lesions induced by mitomycin C (MMC). The putative lethal mutation sites were identified with the following criteria: (1) allelic ratio was approximately 50%, indicating heterozygosity; (2) the identified mutations fell within the range determined by three-factor mapping; (3) the mutation was unique to the strain in question and absent in the composite parental strain, and absent in all the other sequenced MMC-treated strains.  2.2.5 Verification of lethal mutation sites by complementation testing or Sanger sequencing  After a putative lethal site was identified, the mutation was subsequently verified by complementation testing with a known allele. Ten of the sequenced strains had mutations that mapped to chromosome I, and seven out of ten mutations were identified using the methods described in Chapter 2.2.4. Five of the seven strains carrying h2718, h2727, h2733, h2755, and h2758 were identified as having one gene affected after validation. One strain carrying h2787 was identified as having more than one disrupted gene, and h2717 was identified as a large deletion that affected 56 genes. Complementation testing with known alleles verified the putative 24  essential gene in four of the seven strains (Table 2.2). Before complementation testing, the known alleles in Table 2.2 were ordered from the Caenorhabditis Genetics Center (CGC) and crossed with dpy-5 unc-13 / hT2 animals to balance the known allele with the genetic balancer hT2. The known alleles originated from multiple C. elegans labs and many were C. elegans Gene Knockout Consortium alleles. The lethal strains were outcrossed with N2 males and subsequently crossed to the modified strains carrying the known alleles. The P0 hermaphrodites were brooded in short broods, and egg viability was scored to test complementation. An example of a complementation cross is shown in Appendices (Figure S2.2). The candidate genes in strains carrying h2727 and h2733 were determined by examining RNAi phenotypes, existing alleles, and absence of other mutations in the mapped region.  Putative lethal mutation Strain used in complementation test Origin of strain (lab) Genotype of strain h2758 VC1118 Moerman D, C. elegans Reverse Genetics Core, Vancouver, BC, Canada npp-13(ok1534)/szT1[lon-2(e678)] I; +/szT1 X h2718 VC2112 Moerman D, C. elegans Reverse Genetics Core, Vancouver, BC, Canada Y71F9AL.17(ok2824) I/hT2[bli-4(e937) let-?(q782) qIs48](I;III) [also called copa-1] h2787 HY604 Aroian R, UCSD, La Jolla CA mat-1(ye121) I h2755 VC1344 Moerman D, C. elegans Reverse Genetics Core, Vancouver, BC, Canada T09B4.9(ok1792) I/hT2[bli-4(e937) let-?(q782) qIs48](I;III) Table 2.2: List of known alleles used to verify the putative lethal mutation sites.  The strain carrying h2717 was identified as having the putative lethal mutation that affected 56 genes. Complementation testing was not used to verify the mutation, but PCR was 25  used to verify that the mutation was on the normal chromosome since the PCR primers designed around the deletion would not amplify a wild-type sequence. In this method, h2717 dpy-5 unc-13 / hT2 animals were outcrossed to N2 males, and lysates made from h2717 dpy-5 unc-13 / + + +, hT2 / +, and N2 animals were PCR amplified. PCR primers designed to flank the deletion was used as a diagnostic. The deletion was only found in the h2717 dpy-5 unc-13 / + + + animals, and subsequent Sanger sequencing verified the presence of the lethal mutation on the normal chromosome (Figure S2.3a). BioEdit Sequence Alignment Editor v7.2.5 was used to visualize Sanger sequencing trace files [102].                26  Chapter 3: Results  3.1 Forward genetic screen and genetic mapping of mitomycin C-induced lethal mutations  3.1.1 Forward mutation frequency of recessive lethal mutations  The forward mutation frequency was measured by determining the number of lethal events recovered using a genetic balancer. The frequency was calculated as the number of lethal mutations recovered divided by the total number of F1 animals screened. The forward mutation frequency gives an estimate of the potency of a specific mutagen. In this study, the forward mutation frequency was also used as an indication of mutagenesis reasoning that if the concentration of mitomycin C (MMC) induced a lethal mutation it may have induced mutations elsewhere in the genome. To investigate the stability of MMC in solution and variability between batches as reported by others [103, 104], mutagenesis screens were set up using two different vials of MMC.  Using doses similar to those used previously  in Drosophila melanogaster [72], a range of concentrations (375 µM to 1400 µM) were tested in C. elegans [72]. The dose-response data are summarized in Table 3.1. The dose refers to the concentration of MMC that the worms were exposed to. The response refers to the forward mutation frequency. Three separate genetic screens using vial 1 at various storage times gave statistically similar forward mutation frequencies with an average of 5.2% (Table 3.1) (Figure S3.1a). Forward mutation frequencies generated from the mutagenesis using the first vial indicate that MMC retained its potency in frozen solution over a two week period. With regard to concentration, a second batch was dissolved and designated vial 2.  In this case MMC concentrations of 375 µM and 750 µM gave results similar to each other, 2.1% and 2.2%, respectively, but lower than those for vial 1 (Figure S3.1b). Batch variability of MMC has previously been observed, and has been attributed to 27  variations in purity between batches [104, 105]. A dose of 1400 µM MMC resulted in a large number of sterile F1 animals, and was thus excluded from further calculations. The results show that MMC was stable for at least two weeks when stored at the conditions of this experiment (frozen at -20ºC at a concentration of 2 mg in 0.2 ml dH2O prior to dilution). An optimal dosage for mutagenesis was in the range of 375 – 750 uM.  Vial Age of vial Dose Number of F1s tested Number of lethal mutations Forward mutation frequency (%)a 1 Fresh 750 µM 437 24 5.2% 1 Stored 1 week 750 µM 550 25 4.5% 1     Stored 2 weeks 750 µM 436 24 5.5% 2 Stored 1 week 375 µM 145 3 2.1% 2 Stored 1 week 750 µM 580 13 2.2% 2 Stored 1 week 1400 µM 145 1 0.7% a Forward mutation frequency was calculated as a percentage of lethal mutations  (isolated lethal mutations/F1s tested)x100%. Refer to text for details. Table 3.1: Forward mutation frequencies for recessive lethal mutations.  3.1.2 Genetic mapping of mitomycin C-induced lethal mutations  The forward genetic screen produced 90 mitomycin C-induced lethal mutation-bearing strains. Of these 90, 14 did not survive the freezing process and were not further analyzed. The lethal mutation was lost from a further 7 strains prior to mapping. In these 7 strains, the GFP marker was also lost, indicating a breakdown of the hT2 balancer rather than the crossing over of a closely linked mutation. Using three-factor mapping, 69 strains carrying MMC-induced lethal mutations were mapped to either chromosome I or III (data summarized in Table S3.1). 48 lethal mutations were mapped to chromosome III. 21 lethal mutations were mapped to chromosome I 28  (Figure 3.1). The presence of the visible markers dpy-5 and unc-13 allowed for recombination mapping of the lethal mutations relative to the visible markers. Ten (47.6%) of these lethal mutations were to the left of dpy-5, ten (47.6%) were to the right of unc-13, and one (4.8%) lethal mutation was positioned between dpy-5 and unc-13. The map distance for each of the mutations was determined by calculating the 95% confidence intervals as described in Methods.  29   a Map position of the hT2 breakpoint was determined by McKim et al, 1993 [83], which was used to identify the hT2 physical breakpoint at 11 Mbp (Chu, J.S.C., unpublished results). The physical location was inferred from the position of pes-2.2, the gene closest to the hT2 physical breakpoint. Figure 3.1: Genetic map of mitomycin C-induced lethal mutations in the hT2-balanced region of chromosome I. The h alleles depicted were mapped to chromosome I using three-factor mapping. The numbers displayed beside each h allele represents the mapped genetic location, and the dotted lines represent 95% confidence intervals of three-factor mapping. The genes bli-3, dpy-5, unc-13, and unc-101 are shown as references. Map positions of lethal mutations are displayed as map units. h2744 (-16.86)h2717 (-13.23)h2733 (-11.62)h2727 (-12.63)h2758 (-12.10)h2718 (-6.09) h2787 (-2.06)h2784 (-1.82)h2785 (-1.80)h2772 (-0.067)dpy-5 unc-13h2755 h2726 (+2.37)h2729 (+2.52)h2754 (+2.80)h2798 (+3.05)h2715 (+3.68)h2732 (+4.22)h2793 (+4.65)h2728 (+4.50)h2738 (+5.52)h2776 (+6.42)-20 -15 -10 -5 0 5 10bli-3 unc-101hT2 balanced regionhT2 breakpointa30  3.1.3 Complementation tests to optimize genetic location of mitomycin C-induced lethal mutations  The lethal mutations that mapped to the left of dpy-5 collectively span a large distance (approximately 5.4 Mbp), therefore deletion mapping was used to narrow down the physical location of the lesions.  Inclusion or exclusion with hDf10 would give information about the physical location of the lesions.  Analysis of the genome sequence of a strain carrying hDf10 demonstrated that the right breakpoint is close to 1.1 Mbp on chromosome I (Chu, J.S.C., Baillie, D.L., and Rose, A.M, unpublished results). Complementation tests using the deficiency hDf10 were used with lethal mutations that were mapped to the far left of chromosome I. Male hDf10 dpy-5 unc-29 / hT2 animals were crossed to 6 strains (Figure 3.1, Table 3.2). The P0 hermaphrodites were brooded and the F1 progeny were scored for the absence of male Dpy animals.  One of the strains carrying h2717 dpy-5 unc-13 / hT2 failed to complement hDf10 dpy-5 unc-29 / hT2, as indicated by the lack of male Dpy animals (Table 3.2).        Genotype ( ) Total progeny Wild-type Dpy  Wild-type Dpy Allelic h2715 dpy-5 unc-13 / hT2 1051 605 92  284 70 No h2717 dpy-5 unc-13 / hT2 933 494 3  436 0 Yes h2718 dpy-5 unc-13 / hT2 1340 651 193  427 69 No h2727 dpy-5 unc-13 / hT2 331 177 36  100 18 No h2733 dpy-5 unc-13 / hT2 1014 561 103  314 36 No h2744 dpy-5 unc-13 / hT2 378 175 52  139 12 No h2758 dpy-5 unc-13 / hT2 643 385 54  198 6 No Table 3.2: Complementation testing using the deficiency hDf10.      31  3.2 Characterizing mitomycin C-induced lesions in the hT2 balanced chromosomes 3.2.1 Identifying mitomycin C-induced lethal mutations using three-factor mapping data in conjunction with bioinformatics  Ten of the 21 chromosome I lethal mutation-bearing strains and one of the chromosome III lethal mutation-bearing strains were sent to the Michael Smith Genome Sciences Centre for whole genome sequencing. These were strain carrying h2717, h2718, h2727, h2733, h2744, h2755, h2758, h2784, h2785, h2787, and h2741.The genome sequences were first analyzed for the presence of the balancer, indicated by heterozygosity for the dpy-5 and unc-13 markers. This was determined by the ratio of reference:variant reads visualized using IGV (Table S3.2) [98]. Using three-factor mapping data and manual curation with IGV of sequences within the genetically mapped region, putative lethal-causing mutations were found for 7 of the 10 strains (carrying h2717, h2718, h2727, h2733, h2755, h2758, and h2787) (Table 3.3).  All of these putative-lethal mutations were deletions and ranged from 8 bp to 318,826 bp (Table 3.3). Strain Allele Genetic mutation Mapped location  (map units) Physical Location (bp) KR4968 h2717 318,826 bp deletion -13.23 ± 2.0 I:753,745-1,072,570 KR4978 h2727 2382 bp deletion -12.63 ± 1.5 I:1,742,835-1,745,216 KR4984 h2733 8 bp deletion -11.62 ± 2.7 I:1,833,730-1,833,739 KR5009 h2758 4962 bp deletion -12.10 ± 2.1 I:2,100,906-2,105,867 KR4969 h2718 23 bp deletion -6.09 ± 2.0 I:2,897,024-2,897,044 KR5037 h2787 15,059 bp deletion -2.06 ± 1.6 I:5,112,626-5,127,686 KR5006 h2755 4877 bp deletion Between markers dpy-5 (0.00 ± 0.002) and unc-13 (2.07 ± 0.004) I:6,175,541-6,180,400 Table 3.3: Mitomycin C induced deletions of varying sizes. 32  The strain KR4995 carrying h2744 was mapped to -16.86 ± 1.9 cM, approximately at the physical location I:618,207 – 1,328,228 bp (determined from methods described in Chapter 2.2.4). The entire balanced region was scanned for unique mutations, but none were found. Complementation testing with hDf10 showed genetic complementation, therefore h2744 could be located beyond the hDf10 breakpoint, towards the end of chromosome I, which is difficult to sequence. h2784 and h2785 were mapped to -1.82 ± 0.54 cM and -1.80 ± 1.6 cM, approximately I:3,807,457 – 4,508,930 bp and I:3,745,589 – 4,702,866 bp, respectively. The putative lethal mutations were not found in the mapped regions, and upon extending analysis to the entire chromosome I balanced region, the putative lethal mutations were still not found. It is possible that the mutations were located in regions that had poor coverage, and therefore could not be visualized using the methods described in this thesis. Chapter 3.3 describes further analysis of h2744, h2784, and h2785 using bioinformatics methods. 3.2.2 Validation of the putative mitomycin C-induced lethal mutations  Initial analysis of the strain carrying h2718 revealed three candidate lethal mutations: a 5 bp deletion, 2 bp insertion, and a 23 bp deletion, disrupting the genes itx-1, W05F2.4, and Y71F9AL.17 (also called copa-1), respectively (using annotated reference sequence WS200). To narrow down the list of candidates, steps were taken to verify that the mutation occurred on the normal chromosome and not the balancer chromosome. h2718 dpy-5 unc-13 / hT2 animals were outcrossed to N2 males. h2718 dpy-5 unc-13 / + + +, hT2 / +, and N2 animals were isolated and Sanger sequenced with designed primers flanking the putative lethal mutations to verify the presence of the mutation on the balancer or normal chromosome. Sanger sequencing revealed that the 5 bp deletion disrupting itx-1 was on the hT2 balancer chromosome, and the mutations affecting W05F2.4 and Y71F9AL.17 were on the normal chromosome (Figure S2.3b). 33  Complementation testing with a known allele of Y71F9AL.17 (ok2824) revealed that h2718 failed to complement ok2824, indicating the 23 bp deletion affecting Y71F9AL.17 is the causative mutation in the strain h2718 (complementation test cross outlined in Figure S2.2). The putative lethal mutation in h2717 is a deletion affecting 56 annotated genes. Complementation testing was not used to verify the lesion, but Sanger sequencing was used to verify that the lesion was on the normal chromosome (Figure S2.3a). Complementation testing with a known allele was systematically used to validate the identified lethal mutation in the other 5 strains. More than one gene was found to be potentially affected in the strain h2787, subsequently candidates were narrowed down by examining RNAi phenotypes, available alleles, and protein function. The mutations causing the lethal phenotype, the predicted essential genes, as well as human orthologs of these genes, are summarized in Table 3.4.  Allele Genetic lesion Number of  genes affected Confirmed with Sanger sequencing Predicted  essential gene Human ortholog h2717 318,826 bp deletion 56 Yes N/A N/A h2727 2382 bp deletion 1 N/A mppa-1 PMPCA h2733 8 bp deletion 1 N/A C53H9.2 LSG1 h2758 4962 bp deletion 1 N/A npp-13 NUP93 h2718 23 bp deletion 1 Yes Y71F9AL.17 COPA h2787 15,059 bp deletion 4 N/A mat-1 CDC27 h2755 4877 bp deletion 1 N/A T09B4.9 TIMM44 Table 3.4: Validation of mitomycin C-induced lethal mutation sites using complementation testing and Sanger sequencing.   34  3.3 Global view of mitomycin C-induced damage to the C. elegans genome 3.3.1 Extending analysis of mitomycin C-induced deletions to the rest of the genome  The characterization of the lethal mutations provided an indication of the types of mutations that might be induced genome-wide. In addition, the results of the analysis confirmed that the bioinformatics software used in this study could robustly call variants. Pindel was used to identify insertions and deletions in the sequenced strains and called the same deletions described above. Characterizing the mitomycin C (MMC)-induced lethal mutations showed that MMC induced mainly deletions of varying sizes (Table 3.3). The analysis was extended to the rest of the genome, excluding the hT2-balanced region. Pindel was used to identify homozygous deletions in 11 sequenced genomes. A custom script comparing the read support of a deletion to the average depth across a deletion was used to parse the Pindel data for unique homozygous events. From this analysis, 25 unique deletions were found in the 11 strains, and all 25 deletions were manually validated using IGV (Table S3.3). Consistent with the results from analyzing the balanced regions of the MMC-treated animals, on extending analysis to the rest of the genome, MMC also induced deletions of varying sizes. These deletions range from 2 bp to 13,671 bp (Table 3.5).                35  Deletion Size (bp) Sequence 2 TTCTTTTCATTTCTCtaTGTTTGCCTATCACT 3 AAGAAATAATTCCAGctcATGTCAAATGCTCTT 4 TAATTTCACTGGAAtcatTGTCTTCCTTATCAC 4 ATCCAATTTTCCGCcgcgCAGTCGCCAAAAAGG 5 TAAATGACTACTgtaacGCTTGTGTCGATTTA 5 TTCTAGGCATATattgcGAAATATCTTTATAA 5 CTTGCGTGGGACCAgtcgtGTGGTCGAAACAGAC 5 CAAATTGGAATGCTGccgaaCTCTTGCCCATGTCT 6 TTTGGAGCTGTCGAcacgttCCGCGCCGCACTATA 7 CATAAATCGCAAActgcgttCTTCAGCAACAATAT 8 ATATTCGTGAAGAcattgttcCAACGCTGCACAATT 9 GTGACCTTAAGAAgaacgagatGAGATGGAGAGTTGA 10 TCCAACACAGAGAtgccgtagcgTGCAAATTAGGCTTT 11 GAAATTCTcccccccccccCCCCCCCCCCACTGA 13 ACTGCATCTGgaattccgatatcGATTTCACGATCGAA 19 CGGAGAAGTTGtagaatttaatattactttTTGAAGTAGAAGAAC 22 AAAaccacgcaggcgacgcctacatACCACGCAGGCAGCC 58 CTTTTATACGTAACCtttcc<48>ataaaTGAAAAATTGCTTCC 77 GACGACGACGAGGAAccccc<67>aaagcACCACGTGGAGATTA 104 TGTTTCAAAATATACatctg<94>ctaatATAAATATCCATGCC 165 AGtgccaacaacaatgtattc<137>caacaatcaTGCCAACAACAATGC 331 GAGGAAAAGGATAACacatt<321>aaaaaTCGCAAAAAACGCAT 394 TGGTTCGGCCAATGAatttt<384>ctcagTGTTTACGGTTTATA 1003 ATAGAGAATATACGGtatgc<993>gcaatTATCAGATTTCTTGT 13,671 ATCCCACTTTGTAGAagaac<13661>ggagtTCCAAAGAGTCATGC Underline: deletion. Bolded yellow highlight: 5´-CpG-3´ context. Blue highlight: microhomology flanking the deletion junctions. Table 3.5: Genome-wide sequence context of mitomycin C-induced deletions.   Analysis of the sequence context in the deleted regions of DNA revealed a prevalence of 5´-CpG-3´ dinucleotides versus the other dinucleotide combinations (Figure 3.2). Comparison of the deleted regions of DNA with the 15 nucleotides flanking both sides of the deletion revealed a higher proportion of 5´-CpG-3´ dinucleotides in the deleted regions of DNA versus the flanking regions (Figure S3.2). Further, on comparison with an N2 C. elegans strain, the 5´-CpG-3´ 36  dinucleotide content in the deleted regions of DNA was 3.4-fold higher than in N2. This 5´-CpG-3´ dinucleotide footprint is consistent with in vitro data [63, 69, 70].  Figure 3.2: Putative mitomycin C-induced deleted regions of DNA were enriched for 5´-CpG-3´ dinucleotides. Heatmap representation of the proportion of dinucleotide combinations in the putative mitomycin C-induced deleted regions of DNA, 15 nucleotides flanking left and right, and an N2 C. elegans strain. Red represents a low number of observed events, while green represents a high number of events. Blue triangle indicates the 5´-CpG-3´ dinucleotides.   The nucleotides flanking the deletions were inspected to determine sequence specificity of the deletions. Analysis of flanking regions around the deletion breakpoints revealed small microhomologies in 19 of the 25 deletions, ranging from 1 bp to 14 bp (Table 3.5). Analyzing the sequence context of the previously identified deletions causing lethality also revealed the same sequence specificity in terms of prevalence of 5´-CpG-3´ dinucleotides and flanking microhomologies. It is clear that MMC interacts with DNA in a process that may favour certain sequences. Both of the methods that were used to characterize the spectrum of mutations in the balanced and unbalanced regions of DNA indicate that MMC interacts with preferred sequences in the DNA, leaving behind an identifiable footprint associated with deletions of variable length.    37  3.3.2 Genome-wide distribution of mitomycin C-induced single nucleotide variants  The 11 sequenced mitomycin C (MMC)-treated strains were analyzed for unique, homozygous single nucleotide variants (SNVs) using VarScan. SNVs were called by aligning to the C. elegans reference genome (WS200). Based on an approach previously described in Chu et al, 2012, homozygosity was defined as a reference to variant read ratio that is greater than 90% [96]. For statistical purposes, only SNVs that had coverage greater than 7 overlapping reads were considered. Any SNVs that were present in the composite parental strain, or occurred more than once in the 11 sequenced MMC-treated strains were filtered out. VarScan and subsequent filtering identified 290 unique, homozygous SNVs in the 11 strains. These SNVs were pooled for further analysis since each SNV represents a unique event. Variants were grouped as complementary base pairs – G:C>C:G, A:T>C:G, A:T>G:C, G:C>T:A, A:T>T:A, G:C>A:T – where the colons signify the complementary base pairs and the arrow (>) signifies the change. The distribution and number of SNVs were compared to 10 spontaneously mutating N2 strains, using raw data published in Denver et al, 2009 [100], and 11 sequenced EMS-treated strains (Figure 3.2). EMS is a known monoalkylating agent with a distinct genetic profile [93]. The total number of SNVs identified in each strain is summarized in Table S3.4. Using chi-square statistics, the distribution of SNVs in the MMC-treated animals was statistically different from the untreated N2 strains (p<0.05) (Figure 3.2).  However, the distribution of SNVs in MMC-treated animals did not show a skew towards any one type of single nucleotide change. The distribution in EMS-treated animals differed significantly from both N2 and MMC-treated strains (Figure 3.2). The greatest proportion of SNVs in the EMS-treated strains were G:C>A:T events (Figure 3.2).  38  The total number of SNVs in the MMC-treated animals (290 unique SNVs in 11 strains) was about the same as the number of SNVs in untreated spontaneously mutating strains (391 unique SNVs in 10 strains), but 7-fold lower than the number in the EMS-treated strains (2130 unique SNVs in 11 strains). It is apparent that MMC treatment did not induce many, if any, single nucleotide changes, and the observed SNVs were most likely spontaneous events.     Figure 3.3: Genome-wide profile of mitomycin C-induced single nucleotide variants, compared to N2 and EMS-treated animals. Untreated N2, EMS-treated, and MMC-treated strains were compared to determine the distribution of SNVs in each strain relative to one other. Distribution of SNVs in the MMC-treated animals (column 3) differed somewhat from the untreated N2 strains (column 1) (p < 0.05). The distribution of SNVs in the EMS-treated strains (column 2) was dramatically different than either spontaneous or MMC-treated. Total number of events is shown above each column.  3.3.3 Putative insertions induced by mitomycin C  Characterization of the lethal mutations revealed that mitomycin C (MMC) predominantly induced deletions.  Three of the 11 lethal mutations (h2744, h2784, and h2785) were not found, therefore Pindel and custom scripts were used to further analyze the sequencing n = 391 n = 2130 n = 290 39  data for other structural variants that may have been missed, such as insertions. The entire balanced region was analyzed, but no candidates, neither single nucleotide variants, nor deletions nor insertions, were observed. Using these parameters, analysis was expanded to include the entire genomes of the 11 MMC-treated strains in order to verify if insertions can be observed elsewhere in the genome. Using Pindel and custom scripts, 4 insertions were identified, and subsequently verified using IGV (Table 3.6). 3 of the 4 insertions were 4 bp insertions of the sequence TACC or TAGG. One insertion was a 1 bp cytosine insertion. Browsing WormBase for transposon insert sites, low complexity regions, and repetitive DNA did not reveal additional information about the 3 regions with TACC or TAGG insertions. However, this analysis showed the 1 bp cytosine insertion was located in a repetitive region. The data indicate insertions were not identified in the balanced region, and very few were identified in the unbalanced regions of DNA.  Those that were observed may be due to spontaneous events.  Thus, there is little evidence for MMC generating insertions.  Allele Physical location Disruption Sequence h2718 III:6090831 Intron               TACC                ˅ TTTGAAAAAAAAACA TACCTACCTTTTCGA h2733 II:8244559 Exon               TAGG                ˅ TGGCACAAACGAGCA TAATCGGACCAATGC h2741 V:14356033 Intergenic               TACC                ˅ TGCACGTTTTTGAAT TAGTTTTGTTTCTTT h2787 V:18314060 Intergenic                C                ˅ TTTTTGTAAGGGGGA CCTTTGGATTTTTTT Table 3.6: Putative mitomycin C-induced insertions identified by Pindel.   40  Chapter 4: Discussion 4.1 Overview The aim of this thesis was to characterize the nature of lesions induced by the DNA crosslinking agent mitomycin C (MMC). Specifically, the main goals were 1) to assess the mutagenic potency of MMC; 2) to use whole genome sequencing to identify the type of lesions caused by MMC; and 3) to determine possible mechanisms by which MMC-induced lesions may arise. Two complementary methods were employed. The first method involved characterizing lethal mutation-causing lesions that were captured and maintained by the genetic balancer hT2. In this way, the state of the genome immediately after mutagenesis could be captured and maintained for analysis. The second method involved analysis of the DNA sequence in the rest of the genome. The unbalanced regions of the genome represent a method of cataloguing mutations that were maintained in a homozygous state. The first method recovered mutations resulting in a phenotypic change, while the second method identified mutations that occurred globally across the entire genome.  4.2 Research conclusions 4.2.1 Summary of findings  In this study, the mutagenic potency of mitomycin C (MMC) was assessed. Prior to this study, an optimal dose of MMC had not been established in Caenorhabditis elegans. The concentrations used in this study maximized mutagenesis without causing sterility of the animals. The forward mutation frequency – the number of lethal mutations that were captured in 100 haploid copies of the region balanced by hT2 – was determined for each condition examined. This number can be calculated by the percentage of lethal mutations recovered divided by the 41  total number of F1 animals isolated in the forward genetic screen. The forward mutation frequencies were the same for both 375 µM and 750 µM of MMC; whereas a concentration of 1400 µM MMC resulted in a large number of sterile and low-fecundity animals, as indicated by the low number of F1 animals isolated in the genetic screens. Therefore, a dose in the range of 375 µM to 750 µM MMC is recommended.  Both anecdotal and experimental findings point to variability between different batches of MMC, possibly due to chemical impurities [103, 104]. Mutagenesis experiments using two different vials of MMC revealed batch variability as much as two-fold in these experiments. Although batch variability was observed, lethal mutations were recovered and mapped and the variability did not appear to affect subsequent analysis.  Storage of MMC dissolved in water and frozen at -20ºC did not appear to affect its effectiveness over a two week period.  Thus, it would seem best, when possible, to compare results from experiments using one batch of MMC over a few weeks rather than comparing results between different batches.  The highest forward mutation frequency of MMC was calculated at 5%, which is approximately 5 lethal mutations per 100 haploid copies of the region balanced by hT2, and covers approximately one-fifth of the genome (18.2 Mbp). The forward mutation frequency for a number of mutagens has been determined in C. elegans using the reciprocal translocation, eT1 [81-87].  Using eT1, the forward mutation frequency for MMC was observed to be 2% with an equal number of lethal mutations recovered in the first and second broods (A.M. Rose, unpublished results). This compares favorably with gamma irradiation (4.0%) [86]), UV-TMP (3.3%, A. M. Rose, unpublished results), UV irradiation (3.1% [87]), and formaldehyde (1.6% [85, 91]), but MMC is not as potent as the monoalkylating agent EMS (10.3% [90]). 42  Two methods were used to identify mutations in the balanced and unbalanced regions of the genome, with each region providing different but overlapping information. The first method involved characterizing lethal mutation-causing lesions that were captured and maintained by the genetic balancer hT2. These mutations were generated by the forward genetic screen with MMC, and subsequent three-factor mapping that placed 69 lethal mutations to the hT2-balanced regions of chromosome I or III. 21 lethal mutations were mapped to chromosome I, and 10 of the strains carrying these lethal mutations were analyzed to find the causative mutation. The balancer hT2 results from a reciprocal translocation between the left half of chromosome I and right half of chromosome III (Figure 1.4) [83]. Chromosome I in C. elegans is approximately15 Mbp, while chromosome III is approximately 13 Mbp. Analysis of major changes in sequencing coverage revealed the breakpoints of the translocations were at 13.18 Mbp on chromosome I and 4.98 Mbp on chromosome III (Chu, J.S.C., unpublished results), in agreement with genetic determination of the regions of crossover suppression [83]. Since hT2 balances approximately 13 Mbp of chromosome I and approximately 5 Mbp of chromosome III, one might expect more lethal mutations to occur on chromosome I due to the larger size. However, approximately 70% of the induced lethal mutations mapped to chromosome III. The reason for this result is unknown. It presumably reflects a difference in susceptibility of the DNA to the generation of deletions or the nature of the genes (size, number, essentiality) in the two regions.    In this project, the hT2 balancer strain was designed to recover and map lethal mutations on chromosome I using the visible markers dpy-5 and unc-13.  Using the approach outlined in this thesis, the lethal mutations in 7 of the 10 sequenced MMC-treated strains were successfully identified. The data from this analysis revealed the mutations were deletions of varying sizes, ranging from 8 bp – 318,826 bp. It was also noted that with the exception of one strain (h2718), 43  no other type of structural variant was identified in the lethal-bearing strains analyzed. The data demonstrates that MMC predominantly causes deletions of DNA in an animal model.  In this study, two methods were used to identify the effect of MMC on the genome. In the first method, identifying the molecular basis of lethal mutations captured by a genetic balancer provided evidence that DNA variants caused by MMC could be accurately identified. In this case, it was known that MMC had caused a variation in the DNA resulting in a phenotypic change.  Using genetic approaches, such as complementation testing, the causative change in the DNA could be confirmed. The second method of analysis involved extending variant analysis to the rest of the genome. In the remainder of the genome without a mechanism to maintain balanced heterozygosity, crossing over and segregation will establish homozygous chromosomes. Here, analysis with Pindel revealed that MMC predominantly induced deletions of varying sizes (2 bp to 13,671 bp), along with a very low number of insertions. This result is in agreement with the results of the lethal analysis.  In both methods, deletions of a range of sizes were observed after treatment with MMC.   However, I also wanted to know if MMC caused single nucleotide changes.  Since MMC had been characterized in vitro to both monoalkylate and bialkylate DNA [69, 70], the analysis was extended to include the distribution and frequency of single nucleotide variants (SNVs).  MMC induced a much lower number of SNVs when compared to the alkylating agents ethyl methanesulfonate (EMS), N-ethyl-N-nitrosourea (ENU), and ultraviolet trimethylpsoralen (UV-TMP) [93].  MMC induced nearly 1/10th the number of SNVs when compared to EMS.   Thus, the data presented here indicate that the SNVs observed in the MMC-treated genomes could have occurred spontaneously, as they did not differ as greatly from N2 strains left to propagate on 44  plates for thirty generations without chemical treatment [100].  It is clear from the data that MMC does not act like the monofunctional alkylating agent EMS, nor other alkylating agents such as ENU and UV-TMP.    Analysis of whole genome sequences showed that MMC predominantly induced deletions, and not other variants such as insertions or base substitutions.  This finding agrees with studies using a mouse model, which described deletions of varying sizes (1 bp to 8326 bp) induced by MMC [78]. The mouse study assayed bone marrow from transgenic mice, and isolated mutations in lambda DNA integrated into the mouse chromosomes. In other organisms, it has been reported that MMC induced chromosome interchanges and sister-chromatid exchanges (SCEs) [72-75].   To determine possible mechanisms by which MMC-induced lesions may arise, the deletions were analyzed for recurring sequences.  Examining the sequence context of the deletions induced by MMC revealed a 5´-CpG-3´ bias of the deleted regions of DNA. This finding is consistent with in vitro studies that indicate MMC preferentially interacts with guanines [63]. Analysis of the proportion of SNVs induced by MMC did not show any nucleotide bias, but this could be due to the low number of total events that occurred. In addition, on comparison with untreated N2 animals, the total number of SNVs was similar, implicating the distribution of SNVs were mainly affected by spontaneous mutations. Others have also observed the 5´-CpG-3´ sequence bias in the MMC-treated strains [63, 69, 70].  They have proposed that due to the chemical properties of MMC, interstrand crosslink (ICL) formation specifically occurs in a 5´-CpG-3´ sequence context [63, 69, 70].  45  The analysis done in this thesis examines the end product of mutagenesis and repair, and therefore would not reveal whether or not the lesion resulted from an interstrand crosslink, instrastrand crosslink, or monoadduct. Nevertheless, the results in this thesis indicate that MMC-DNA interactions most likely involve interstrand crosslinks, as indicated by the potential bias towards a 5´-CpG-3´ sequence context. Analyzing the sequence context of the deletions revealed small microhomologies flanking the deletions, consistent with previous findings by Takeiri et al, 2003 [78]. As highlighted by Takeiri et al, 2003, these results point to some form of end-joining repair [78]. It has been noted that small sequence microhomologies could help guide non-homologous end joining-mediated repair [42]. Analysis of the sequence context flanking the deletions revealed that deletions were also generated with no microhomology, implicating NHEJ where blunt ends of DNA were joined together [44, 106].  The experiments outlined in this thesis indicate MMC predominantly induced deletions. It has previously been noted that on induction of double-strand breaks, deletion events are 2- to 8-fold more efficient than inversion events [106]. The study also reported that when the NHEJ component KU80 is absent, mutagenic rejoining, and in particular microhomology-mediated repair occurs more efficiently [106]. Analysis of the sequence context of the MMC-induced mutations indicated that MMC preferentially interacts with a 5´-CpG-3´ sequence context, most likely as a result of interstrand crosslinking. Deletions can be observed as a footprint of NHEJ-mediated repair, with the NHEJ machinery using sequence microhomology to ligate the DSB. Taken together, it is possible that overloading the genome with ICLs mimics the effects of having the NHEJ component KU80 absent, leading to microhomology-mediated repair. 46   4.2.2 Significance of findings   Cancer therapy largely depends on chemotherapeutic agents that generate DNA lesions, which may lead to micro-evolution of cancer, as well as the accumulation of mutations in normal cells. The biological nature of these lesions, as well as the mutational profile of these drugs is not fully understood. In this study, I determined the mutational spectrum of the DNA crosslinking agent mitomycin C (MMC), and characterized the mutagenic potency of this commonly used agent.  The results presented here indicate that MMC is a strongly mutagenic drug. A dose of 750 µM of MMC was sufficient to induce as many as 5 lethal mutations per 100 haploid copies of the region balanced by hT2, which is as mutagenic as commonly used mutagens such as UV-TMP, UV-radiation, gamma-radiation, and formaldehyde, but less potent than EMS [85-87, 90].  This demonstrates the potent cytotoxic nature of MMC – it can induce a very few mutations that are lethal to cells.  Subsequent bioinformatics analysis revealed a low mutational load, as indicated by the low number of variants other than the handful of deletions that were identified.   This result may be relevant for selection of a mutagen that can induce a high number of deletions accompanied by few background mutations.   A mechanism of how these deletions may arise is shown in Figure 4.1. When an interstrand crosslink (ICL) forms, nucleases cleave the regions around the lesion, leading to a double-strand break (DSB) [39]. Due to the one ended nature of the DSBs that arise as a consequence of ICL repair, the non-homologous end joining (NHEJ) components do not have a natural substrate available to rejoin the breaks. Consequently, breaks can remain unrepaired, 47  leading to deletions resulting from direct ligation of the broken ends of DNA [46-48]. This ligation may be guided by small sequence microhomologies around the deletion junctions (Figure 4.1) [42]. Subsequently, the deletions can be observed from the sequencing data as a fingerprint of NHEJ-mediated repair, marked by flanking microhomologies.   Figure 4.1: Possible mechanism for the generation of deletions with flanking sequence microhomologies. When an interstrand crosslink forms, nucleases cause incisions around the lesion, leading to a double-strand break. NHEJ is activated, and NHEJ components degrade and resection the DNA, forming overhangs for re-annealing. This ligation may be guided by small sequence microhomologies. Subsequently, the deletions can be observed from the sequencing data as a fingerprint of NHEJ-mediated repair, marked by flanking microhomologies.  deletionRe-annealTGCCGTAGCGTGCACGGCATCGCACGDouble-strand break5’3’ 5’3’TGCCGT  AGCGTGCACGGCA  TCGCACG5’3’ 5’3’TG        CGTGCACGG        ACG5’3’ 5’3’Degradation and resectionTGCACG5’3’ 5’3’TGCACG5’3’ 5’3’Deletion mutantWildtypesequencetgccgtagcgTGCacggcatcgcACG5’3’ 5’3’48   4.2.3 Future directions  In this study, I determined the mutagenic spectrum of mitomycin C (MMC) in a non-DNA repair-deficient strain of C. elegans. The pipeline outlined in this thesis described a method of systematically identifying mutations such as single nucleotide variants, deletions, and insertions. To gain better insight into the mechanisms required for interstrand crosslink (ICL) repair, it is important to extend this analysis to strains that are deficient in DNA repair components. Components of the Fanconi anemia (FA) pathway have been shown to suppress ICL sensitivity by promoting homologous recombination (HR) [49]. Experiments have demonstrated that factors involved in HR and non-homologous end joining (NHEJ) most likely compete to fix lesions. It has been shown that loss of NHEJ factors promotes end resection, and thus HR [45]. Experiments also indicate that when FA pathway components are mutated, there is an increase in NHEJ-mediated events [46-48]. Therefore, the analysis in this thesis can be extended to C. elegans strains deficient in components of HR and NHEJ to assess how the mutagenic spectrum changes. This analysis can also be extended to characterize other chemotherapeutic agents. These future studies would provide greater insight into the DNA repair pathways required for ICL repair, and also a better understanding of the biological consequences of treating patients with chemotherapeutic agents.      49  References  1. Pfeifer, G.P., Environmental exposures and mutational patterns of cancer genomes. Genome Med, 2010. 2(8): p. 54. 2. Pena-Diaz, J., et al., Noncanonical mismatch repair as a source of genomic instability in human cells. Mol Cell, 2012. 47(5): p. 669-80. 3. Soulier, J., et al., Detection of somatic mosaicism and classification of Fanconi anemia patients by analysis of the FA/BRCA pathway. Blood, 2005. 105(3): p. 1329-36. 4. Clingen, P.H., et al., The XPF-ERCC1 endonuclease and homologous recombination contribute to the repair of minor groove DNA interstrand crosslinks in mammalian cells produced by the pyrrolo[2,1-c][1,4]benzodiazepine dimer SJG-136. Nucleic Acids Res, 2005. 33(10): p. 3283-91. 5. O'Connor, P.M. and K.W. Kohn, Comparative pharmacokinetics of DNA lesion formation and removal following treatment of L1210 cells with nitrogen mustards. Cancer Commun, 1990. 2(12): p. 387-94. 6. Alexandrov, L.B., et al., Signatures of mutational processes in human cancer. Nature, 2013. 500(7463): p. 415-21. 7. Nik-Zainal, S., et al., Mutational processes molding the genomes of 21 breast cancers. Cell, 2012. 149(5): p. 979-93. 8. Piette, J., et al., Mutagenesis induced by site specifically placed 4'-hydroxymethyl-4,5',8-trimethylpsoralen adducts. Nucleic Acids Res, 1988. 16(21): p. 9961-77. 9. Osawa, T., D. Davies, and J.A. Hartley, Mechanism of cell death resulting from DNA interstrand cross-linking in mammalian cells. Cell Death Dis, 2011. 2: p. e187. 10. Magana-Schwencke, N., et al., The fate of 8-methoxypsoralen photoinduced crosslinks in nuclear and mitochondrial yeast DNA: comparison of wild-type and repair-deficient strains. Proc Natl Acad Sci U S A, 1982. 79(6): p. 1722-6. 11. Phillips, D.H., DNA adducts in human tissues: biomarkers of exposure to carcinogens in tobacco smoke. Environ Health Perspect, 1996. 104 Suppl 3: p. 453-8. 12. Spanswick, V.J., et al., Repair of DNA interstrand crosslinks as a mechanism of clinical resistance to melphalan in multiple myeloma. Blood, 2002. 100(1): p. 224-9. 13. Deans, A.J. and S.C. West, DNA interstrand crosslink repair and cancer. Nat Rev Cancer, 2011. 11(7): p. 467-80. 14. Jamieson, E.R. and S.J. Lippard, Structure, Recognition, and Processing of Cisplatin-DNA Adducts. Chem Rev, 1999. 99(9): p. 2467-98. 15. Ojwang, J.O., D.A. Grueneberg, and E.L. Loechler, Synthesis of a duplex oligonucleotide containing a nitrogen mustard interstrand DNA-DNA cross-link. Cancer Res, 1989. 49(23): p. 6529-37. 16. Tomasz, M., Mitomycin C: small, fast and deadly (but very selective). Chem Biol, 1995. 2(9): p. 575-9. 17. Cimino, G.D., et al., Psoralens as photoactive probes of nucleic acid structure and function: organic chemistry, photochemistry, and biochemistry. Annu Rev Biochem, 1985. 54: p. 1151-93. 18. Fischhaber, P.L., et al., Direct demonstration in synthetic oligonucleotides that N,N'-bis(2-chloroethyl)-nitrosourea cross links N1 of deoxyguanosine to N3 of deoxycytidine on opposite strands of duplex DNA. Cancer Res, 1999. 59(17): p. 4363-8. 50  19. Cole, R.S., D. Levitan, and R.R. Sinden, Removal of psoralen interstrand cross-links from DNA of Escherichia coli: mechanism and genetic control. J Mol Biol, 1976. 103(1): p. 39-59. 20. Sladek, F.M., et al., In vitro repair of psoralen-DNA cross-links by RecA, UvrABC, and the 5'-exonuclease of DNA polymerase I. J Biol Chem, 1989. 264(12): p. 6755-65. 21. Simon, J.A., et al., Differential toxicities of anticancer agents among DNA repair and checkpoint mutants of Saccharomyces cerevisiae. Cancer Res, 2000. 60(2): p. 328-33. 22. Berardini, M., W. Mackay, and E.L. Loechler, Evidence for a recombination-independent pathway for the repair of DNA interstrand cross-links based on a site-specific study with nitrogen mustard. Biochemistry, 1997. 36(12): p. 3506-13. 23. Henriques, J.A. and E. Moustacchi, Isolation and characterization of pso mutants sensitive to photo-addition of psoralen derivatives in Saccharomyces cerevisiae. Genetics, 1980. 95(2): p. 273-88. 24. Lawley, P.D. and P. Brookes, Molecular mechanism of the cytotoxic action of difunctional alkylating agents and of resistance to this action. Nature, 1965. 206(983): p. 480-3. 25. Cole, R.S., Repair of DNA containing interstrand crosslinks in Escherichia coli: sequential excision and recombination. Proc Natl Acad Sci U S A, 1973. 70(4): p. 1064-8. 26. McHugh, P.J., V.J. Spanswick, and J.A. Hartley, Repair of DNA interstrand crosslinks: molecular mechanisms and clinical relevance. Lancet Oncol, 2001. 2(8): p. 483-90. 27. Muniandy, P.A., et al., DNA interstrand crosslink repair in mammalian cells: step by step. Crit Rev Biochem Mol Biol, 2010. 45(1): p. 23-49. 28. Barber, L.J., et al., DNA interstrand cross-link repair in the Saccharomyces cerevisiae cell cycle: overlapping roles for PSO2 (SNM1) with MutS factors and EXO1 during S phase. Mol Cell Biol, 2005. 25(6): p. 2297-309. 29. McHugh, P.J. and S. Sarkar, DNA interstrand cross-link repair in the cell cycle: a critical role for polymerase zeta in G1 phase. Cell Cycle, 2006. 5(10): p. 1044-7. 30. Raschle, M., et al., Mechanism of replication-coupled DNA interstrand crosslink repair. Cell, 2008. 134(6): p. 969-80. 31. Knipscheer, P., et al., The Fanconi anemia pathway promotes replication-dependent DNA interstrand cross-link repair. Science, 2009. 326(5960): p. 1698-701. 32. Long, D.T., et al., Mechanism of RAD51-dependent DNA interstrand cross-link repair. Science, 2011. 333(6038): p. 84-7. 33. Wang, X., et al., Involvement of nucleotide excision repair in a recombination-independent and error-prone pathway of DNA interstrand cross-link repair. Mol Cell Biol, 2001. 21(3): p. 713-20. 34. Zheng, H., et al., Nucleotide excision repair- and polymerase eta-mediated error-prone removal of mitomycin C interstrand cross-links. Mol Cell Biol, 2003. 23(2): p. 754-61. 35. Shen, X., et al., REV3 and REV1 play major roles in recombination-independent repair of DNA interstrand cross-links mediated by monoubiquitinated proliferating cell nuclear antigen (PCNA). J Biol Chem, 2006. 281(20): p. 13869-72. 36. Hlavin, E.M., et al., Cross-link structure affects replication-independent DNA interstrand cross-link repair in mammalian cells. Biochemistry, 2010. 49(18): p. 3977-88. 51  37. Enoiu, M., J. Jiricny, and O.D. Scharer, Repair of cisplatin-induced DNA interstrand crosslinks by a replication-independent pathway involving transcription-coupled repair and translesion synthesis. Nucleic Acids Res, 2012. 40(18): p. 8953-64. 38. Kuraoka, I., et al., Repair of an interstrand DNA cross-link initiated by ERCC1-XPF repair/recombination nuclease. J Biol Chem, 2000. 275(34): p. 26632-6. 39. Hanada, K., et al., The structure-specific endonuclease Mus81-Eme1 promotes conversion of interstrand DNA crosslinks into double-strands breaks. EMBO J, 2006. 25(20): p. 4921-32. 40. Scrima, A., et al., Structural basis of UV DNA-damage recognition by the DDB1-DDB2 complex. Cell, 2008. 135(7): p. 1213-23. 41. Clauson, C., O.D. Scharer, and L. Niedernhofer, Advances in understanding the complex mechanisms of DNA interstrand cross-link repair. Cold Spring Harb Perspect Med, 2013. 3(10): p. a012732. 42. Simsek, D. and M. Jasin, Alternative end-joining is suppressed by the canonical NHEJ component Xrcc4-ligase IV during chromosomal translocation formation. Nat Struct Mol Biol, 2010. 17(4): p. 410-6. 43. DeFazio, L.G., et al., Synapsis of DNA ends by DNA-dependent protein kinase. EMBO J, 2002. 21(12): p. 3192-200. 44. Critchlow, S.E. and S.P. Jackson, DNA end-joining: from yeast to man. Trends Biochem Sci, 1998. 23(10): p. 394-8. 45. Lee, S.E., et al., Saccharomyces Ku70, mre11/rad50 and RPA proteins regulate adaptation to G2/M arrest after DNA damage. Cell, 1998. 94(3): p. 399-409. 46. Pichierri, P., D. Averbeck, and F. Rosselli, DNA cross-link-dependent RAD50/MRE11/NBS1 subnuclear assembly requires the Fanconi anemia C protein. Hum Mol Genet, 2002. 11(21): p. 2531-46. 47. Bennardo, N., et al., Alternative-NHEJ is a mechanistically distinct pathway of mammalian chromosome break repair. PLoS Genet, 2008. 4(6): p. e1000110. 48. Yun, M.H. and K. Hiom, CtIP-BRCA1 modulates the choice of DNA double-strand-break repair pathway throughout the cell cycle. Nature, 2009. 459(7245): p. 460-3. 49. Nakanishi, K., et al., Human Fanconi anemia monoubiquitination pathway promotes homologous DNA repair. Proc Natl Acad Sci U S A, 2005. 102(4): p. 1110-5. 50. Wang, W., Emergence of a DNA-damage response network consisting of Fanconi anaemia and BRCA proteins. Nat Rev Genet, 2007. 8(10): p. 735-48. 51. Hata, T., et al., Mitomycin, a new antibiotic from Streptomyces. I. J Antibiot (Tokyo), 1956. 9(4): p. 141-6. 52. Kato, T., et al., Chemosensitivity of human bladder cancer cells in long-term culture and clinical responses to the selected anticancer drug. Cancer, 1979. 44(1): p. 58-63. 53. Spain, R.C., The case for mitomycin in non-small cell lung cancer. Oncology, 1993. 50 Suppl 1: p. 35-50. 54. Michalaki, V., S. Gennatas, and C. Gennatas, Mitomycin C and UFT/leucovorin as salvage treatment in patients with advanced colorectal cancer. J BUON, 2010. 15(2): p. 270-3. 55. Greenberg, B.R., et al., Sequential combination of bleomycin and mitomycin in advanced cervical cancer--an American experience: a Northern California Oncology Group study. Cancer Treat Rep, 1982. 66(1): p. 163-5. 52  56. Doll, D.C., R.B. Weiss, and B.F. Issell, Mitomycin: ten years after approval for marketing. J Clin Oncol, 1985. 3(2): p. 276-86. 57. Iyer, V.N. and W. Szybalski, Mitomycins and Porfiromycin: Chemical Mechanism of Activation and Cross-Linking of DNA. Science, 1964. 145(3627): p. 55-8. 58. Carrano, A.V., et al., DNA crosslinking, sister-chromatid exchange and specific-locus mutations. Mutat Res, 1979. 63(1): p. 175-88. 59. Iyer, V.N. and W. Szybalski, A Molecular Mechanism of Mitomycin Action: Linking of Complementary DNA Strands. Proc Natl Acad Sci U S A, 1963. 50: p. 355-62. 60. Srinivasan, S., A. Drake, and S. Herzig, Photorefractive keratectomy with 0.02% mitomycin C for treatment of residual refractive errors after LASIK. J Refract Surg, 2008. 24(1): p. S64-7. 61. Mandal, A.K., et al., Mitomycin C-augmented trabeculectomy in refractory congenital glaucoma. Ophthalmology, 1997. 104(6): p. 996-1001; discussion 1002-3. 62. Tomasz, M., et al., The mode of interaction of mitomycin C with deoxyribonucleic acid and other polynucleotides in vitro. Biochemistry, 1974. 13(24): p. 4878-87. 63. Gargiulo, D., et al., Structural and function modification of DNA by mitomycin C. Mechanism of the DNA sequence specificity of mitomycins. Nucleic Acids Symp Ser, 1995(34): p. 169-70. 64. Borowy-Borowski, H., et al., Duplex oligodeoxyribonucleotides cross-linked by mitomycin C at a single site: synthesis, properties, and cross-link reversibility. Biochemistry, 1990. 29(12): p. 2992-9. 65. Warren, A.J. and J.W. Hamilton, Synthesis and structural characterization of the N2G-mitomycin C-N2G interstrand cross-link in a model synthetic 23 base pair oligonucleotide DNA duplex. Chem Res Toxicol, 1996. 9(7): p. 1063-71. 66. Tomasz, M., et al., Reaction of DNA with chemically or enzymatically activated mitomycin C: isolation and structure of the major covalent adduct. Proc Natl Acad Sci U S A, 1986. 83(18): p. 6702-6. 67. Tomasz, M., et al., Isolation and structure of a covalent cross-link adduct between mitomycin C and DNA. Science, 1987. 235(4793): p. 1204-8. 68. Norman, D., et al., NMR and computational characterization of mitomycin cross-linked to adjacent deoxyguanosines in the minor groove of the d(T-A-C-G-T-A).d(T-A-C-G-T-A) duplex. Biochemistry, 1990. 29(11): p. 2861-75. 69. Kumar, S., R. Lipman, and M. Tomasz, Recognition of specific DNA sequences by mitomycin C for alkylation. Biochemistry, 1992. 31(5): p. 1399-407. 70. Borowy-Borowski, H., R. Lipman, and M. Tomasz, Recognition between mitomycin C and specific DNA sequences for cross-link formation. Biochemistry, 1990. 29(12): p. 2999-3006. 71. Sastry, M., et al., Solution structure of the monoalkylated mitomycin C-DNA complex. J Mol Biol, 1995. 247(2): p. 338-59. 72. Schewe, M.J., D.T. Suzuki, and U. Erasmus, The genetic effects of mitomycin C in Drosophila melanogaster. I. Induced mutations and X-Y chromosomal interchanges. Mutat Res, 1971. 12(3): p. 255-67. 73. Schewe, M.J., D.T. Suzuki, and U. Erasmus, The genetic effects of mitomycin C in Drosophila melanogaster. II. Induced meiotic recombination. Mutat Res, 1971. 12(3): p. 269-79. 53  74. Natarajan, A.T., et al., Induction of sister-chromatid exchanges (SCEs) and chromosomal aberrations by mitomycin C and methyl methanesulfonate in Chinese hamster ovary cells. An evaluation of methodology for detection of SCEs and of persistent DNA lesions towards the frequencies of observed SCEs. Mutat Res, 1983. 121(3-4): p. 211-23. 75. Russo, A., et al., Persistence of chromosomal lesions induced in mouse bone marrow cells by mitomycin C, as evaluated by SCE analysis. Mutat Res, 1993. 287(2): p. 275-82. 76. Littlefield, L.G., S.P. Colyer, and R.J. DuFrain, SCE evaluations in human lymphocytes after G0 exposure to mitomycin C. Lack of expression of MMC-induced SCEs in cells that have undergone greater than two in vitro divisions. Mutat Res, 1983. 107(1): p. 119-30. 77. Daza, P., et al., Mitomycin C, 4-nitroquinoline-1-oxide and ethyl methanesulfonate induce long-lived lesions in DNA which result in SCEs during successive cell cycles in human lymphocytes. Mutat Res, 1992. 270(2): p. 177-83. 78. Takeiri, A., et al., Molecular characterization of mitomycin C-induced large deletions and tandem-base substitutions in the bone marrow of gpt delta transgenic mice. Chem Res Toxicol, 2003. 16(2): p. 171-9. 79. Jones, M.R., Z. Lohn, and A.M. Rose, Specialized chromosomes and their uses in Caenorhabditis elegans. Methods Cell Biol, 2011. 106: p. 23-64. 80. Muller, H.J., Genetic Variability, Twin Hybrids and Constant Hybrids, in a Case of Balanced Lethal Factors. Genetics, 1918. 3(5): p. 422-99. 81. Rosenbluth, R.E. and D.L. Baillie, The genetic analysis of a reciprocal translocation, eT1(III; V), in Caenorhabditis elegans. Genetics, 1981. 99(3-4): p. 415-28. 82. Adames, K.A., et al., Mapping a telomere using the translocation eT1(III;V) in Caenorhabditis elegans. Genetics, 1998. 150(3): p. 1059-66. 83. McKim, K.S., K. Peters, and A.M. Rose, Two types of sites required for meiotic chromosome pairing in Caenorhabditis elegans. Genetics, 1993. 134(3): p. 749-68. 84. McKim, K.S., A.M. Howell, and A.M. Rose, The effects of translocations on recombination frequency in Caenorhabditis elegans. Genetics, 1988. 120(4): p. 987-1001. 85. Johnsen, R.C. and D.L. Baillie, Formaldehyde mutagenesis of the eT1 balanced region in Caenorhabditis elegans: dose-response curve and the analysis of mutational events. Mutat Res, 1988. 201(1): p. 137-47. 86. Rosenbluth, R.E., C. Cuddeford, and D.L. Baillie, Mutagenesis in Caenorhabditis elegans. II. A spectrum of mutational events induced with 1500 r of gamma-radiation. Genetics, 1985. 109(3): p. 493-511. 87. Stewart, H.I., R.E. Rosenbluth, and D.L. Baillie, Most ultraviolet irradiation induced mutations in the nematode Caenorhabditis elegans are chromosomal rearrangements. Mutat Res, 1991. 249(1): p. 37-54. 88. Zhao, Y., et al., A mutational analysis of Caenorhabditis elegans in space. Mutat Res, 2006. 601(1-2): p. 19-29. 89. Zhao, Y., et al., Spectrum of mutational events in the absence of DOG-1/FANCJ in Caenorhabditis elegans. DNA Repair (Amst), 2008. 7(11): p. 1846-54. 90. Rosenbluth, R.E., Cuddeford, C., and Baillie, D. L., Mutagenesis in Caenorhabditis elegans: I. A rapid eukaryotic mutagen test system using the reciprocal translocation, eTI(III;V). Mut. Res./Fundam. Molec. Mech. Mutagenesis, 1983. 110: p. 39-48. 54  91. Moerman, D.G. and D.L. Baillie, Formaldehyde mutagenesis in the nematode Caenorhabditis elegans. Mutat Res, 1981. 80(2): p. 273-9. 92. Sarin, S., et al., Analysis of multiple ethyl methanesulfonate-mutagenized Caenorhabditis elegans strains by whole-genome sequencing. Genetics, 2010. 185(2): p. 417-30. 93. Flibotte, S., et al., Whole-genome profiling of mutagenesis in Caenorhabditis elegans. Genetics, 2010. 185(2): p. 431-41. 94. Brenner, S., The genetics of Caenorhabditis elegans. Genetics, 1974. 77(1): p. 71-94. 95. Horvitz, H.R., et al., A uniform genetic nomenclature for the nematode Caenorhabditis elegans. Mol Gen Genet, 1979. 175(2): p. 129-33. 96. Chu, J.S., et al., Allelic ratios and the mutational landscape reveal biologically significant heterozygous SNVs. Genetics, 2012. 190(4): p. 1225-33. 97. Li, H. and R. Durbin, Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics, 2009. 25(14): p. 1754-60. 98. Robinson, J.T., et al., Integrative genomics viewer. Nat Biotechnol, 2011. 29(1): p. 24-6. 99. Koboldt, D.C., et al., VarScan: variant detection in massively parallel sequencing of individual and pooled samples. Bioinformatics, 2009. 25(17): p. 2283-5. 100. Denver, D.R., et al., A genome-wide view of Caenorhabditis elegans base-substitution mutation processes. Proc Natl Acad Sci U S A, 2009. 106(38): p. 16310-4. 101. Ye, K., et al., Pindel: a pattern growth approach to detect break points of large deletions and medium sized insertions from paired-end short reads. Bioinformatics, 2009. 25(21): p. 2865-71. 102. Hall, T.A., BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symp Ser, 1999. 41: p. 95-98. 103. Woo, E., et al., Stability of the antiproliferative effect of mitomycin-C after reconstitution. J Glaucoma, 1997. 6(1): p. 33-6. 104. Cervenka, J. and B.A. Hirsch, Cytogenetic differentiation of Fanconi anemia, "idiopathic" aplastic anemia, and Fanconi anemia heterozygotes. Am J Med Genet, 1983. 15(2): p. 211-23. 105. Proctor, B.L. and M.E. Gaulden, Chemical stability of mitomycin C in culture medium with and without fetal calf serum as determined by high pressure liquid chromatography and mass spectrometry. Arch Environ Contam Toxicol, 1986. 15(2): p. 235-40. 106. Guirouilh-Barbat, J., et al., Impact of the KU80 pathway on NHEJ-induced genome rearrangements in mammalian cells. Mol Cell, 2004. 14(5): p. 611-23.      55  Appendices Appendix A: Supplementary material for chapter 2   Figure S2.1: IGV was used to identify the putative mitomycin C-induced lethal mutations. The balanced regions on chromosome I were visually analyzed for any mutation that only occurred in one strain (unique) and were heterozygous (mutation occurred after the mutagenesis). In this example (5 kb deletion in the strain carrying h2755), the evidence of the deletion was the global drop in coverage in the deleted region, split reads indicating the deletion breakpoint, and red colour coded reads indicating the insert size is larger than expected.    Split reads Split readsDeleted regionDrop in coverageh2755: 5 kb deletion IGV uses colour coding to flag anomalous insert sizes. Red reads indicate the inferred insert size is greater than the expected insert size. 56   Figure S2.2: An example of a complementation test with a known allele to verify the putative mitomycin C-induced lethal mutation. Complementation testing with a known allele of Y71F9AL.17 (also known as copa-1 (ok2824)) revealed that h2718 failed to complement ok2824, indicating the 23 bp deletion affecting Y71F9AL.17 was the lethal mutation. Details in text.Y71F9AL.17hT2♂h2718 dpy-5 unc-13+xPick non-GFP hermaphroditesY71F9AL.17      .h2718 dpy-5 unc-13Y71F9AL.17+orShort broodsY71F9AL.17+Y71F9AL.17Y71F9AL.17++Score eggs; expect ¼ dead eggsY71F9AL.17      .h2718 dpy-5 unc-13Y71F9AL.17Y71F9AL.17h2718 dpy-5 unc-13.h2718 dpy-5 unc-13Score eggs; if complement, expect ½ dead eggs. If the two are allelic, this animal is lethal and therefore would not produce eggsComplement Fail to complementNo eggs that can be scored57     Figure S2.3: Sanger sequencing was used to verify that the candidate lethal mutations were not located on the hT2 balancer chromosome. a) Validation of the 300kb deletion found in the strain carrying h2717. The black arrow indicates the deletion junction. b) Validating the candidate lethal mutations in the strain carrying h2718. Candidate lethal mutations: itx-1, Y71F9AL.17, W05F2.4 (blue boxes). Primers were designed around the mutation in each gene and subsequently Sanger sequenced. The 5 bp deletion disrupting itx-1 was identified on the hT2 balancer chromosome, and the mutations affecting W05F2.4 and Y71F9AL.17 were identified on the normal chromosome. Black arrows indicate wild-type sequence. Red arrows indicate the mutant sequence that can be visualized. Orange boxes indicate the genotype of the strain sequenced. Trace files were visualized using BioEdit [102].Deletion breakpointh2717/+h2718/+N2hT2/+Mutant sequenceWildtype sequenceWildtype sequenceWildtype sequenceWildtype sequenceMutant sequenceY71F9AL.17Wildtype sequenceWildtype sequenceMutant sequenceA)B)itx-1 W05F2.458  Appendix B: Supplementary material for chapter 3       F1 progeny  Genotype  (P0 hermaphrodite) Total Recombinants (Dpy-Unc) Recombinants (Dpy or Unc) Lethal Mutation (Distance in map unitsa) h2715 dpy-5 unc-13 / +++ 1113 12 15 Dpy +3.7 (2.7-4.6) h2717 dpy-5 unc-13 / +++ 1669 139 17 Unc -13.2 (11.3-15.2) h2718 dpy-5 unc-13 / +++ 1512 61 9 Unc -6.1 (4.1-8.1) h2720 dpy-5 unc-13 / +++ 1742 356 18 Dpy 9 Unc Chromosome III h2721 dpy-5 unc-13 / +++ 1885 380 10 Dpy 7 Unc Chromosome III h2722 dpy-5 unc-13 / +++ 1754 352 20 Dpy 14 Unc Chromosome III h2723 dpy-5 unc-13 / +++ 1517 324 18 Dpy 17 Unc Chromosome III h2724 dpy-5 unc-13 / +++ 2579 293 3 Dpy 11 Unc Chromosome III h2726 dpy-5 unc-13 / +++ 1931 4 10 Dpy +2.4 (1.9-2.8) h2727 dpy-5 unc-13 / +++ 2147 167 27 Unc -12.6 (11.1-14.2) h2728 dpy-5 unc-13 / +++ 1489 21 12 Dpy +4.5 (3.2-5.8) h2729 dpy-5 unc-13 / +++ 1581 4 10 Dpy +2.5 (2.1-2.9) h2730 dpy-5 unc-13 / +++ 1418 235 11 Dpy 1 Unc Chromosome III h2731 dpy-5 unc-13 / +++ 924 193 7 Dpy 0 Unc Chromosome III h2732 dpy-5 unc-13 / +++ 2053 27 23 Dpy +4.2 (3.2-5.2) h2733 dpy-5 unc-13 / +++ 1277 92 3 Unc -11.6 (8.9-14.3) h2734 dpy-5 unc-13 / +++ 2304 463 20 Dpy 16 Unc Chromosome III h2736 dpy-5 unc-13 / +++ 1779 392 20 Dpy 18 Unc Chromosome III h2737 dpy-5 unc-13 / +++ 2201 452 16 Dpy 17 Unc Chromosome III h2738 dpy-5 unc-13 / +++ 2082 47 9 Dpy +5.5 (4.5-6.5) h2739 dpy-5 unc-13 / +++ 1704 212 19 Dpy 15 Unc Chromosome III h2741 dpy-5 unc-13 / +++ 2194 207 20 Dpy 24 Unc Chromosome III h2744 dpy-5 unc-13 / +++ 2545 262 21 Unc -16.9 (15.0-18.8) h2745 dpy-5 unc-13 / +++ 1515 144 3 Dpy 17 Unc Chromosome III h2746 dpy-5 unc-13 / +++ 2017 407 14 Dpy 11 Unc Chromosome III h2747 dpy-5 unc-13 / +++ 2526 567 11 Dpy 22 Unc Chromosome III h2748 dpy-5 unc-13 / +++ 1529 343 8 Dpy 10 Unc Chromosome III h2749 dpy-5 unc-13 / +++ 2199 471 20 Dpy 19 Unc Chromosome III h2750 dpy-5 unc-13 / +++ 2108 448 12 Dpy 12 Unc Chromosome III h2751 dpy-5 unc-13 / +++ 1828 364 13 Dpy 8 Unc Chromosome III h2752 dpy-5 unc-13 / +++ 2160 427 12 Dpy 8 Unc Chromosome III h2753 dpy-5 unc-13 / +++ 1578 339 13 Dpy 9 Unc Chromosome III h2754 dpy-5 unc-13 / +++ 1649 8 12 Dpy +2.8 (2.3-3.3) h2755 dpy-5 unc-13 / +++ 1951 0 9 Dpy 11 Unc Not applicableb h2756 dpy-5 unc-13 / +++ 1906 227 7 Dpy 13 Unc Chromosome III h2758 dpy-5 unc-13 / +++ 1978 153 11 Unc -12.1 (10.0-14.2) h2759 dpy-5 unc-13 / +++ 1961 354 13 Dpy 7 Unc Chromosome III h2760 dpy-5 unc-13 / +++ 2319 414 15 Dpy 10 Unc Chromosome III 59       F1 progeny  Genotype  (P0 hermaphrodite) Total Recombinants (Dpy-Unc) Recombinants (Dpy or Unc) Lethal Mutation (Distance in map unitsa) h2761 dpy-5 unc-13 / +++ 2039 393 13 Dpy 9 Unc Chromosome III h2762 dpy-5 unc-13 / +++ 1415 288 9 Dpy 12 Unc Chromosome III h2763 dpy-5 unc-13 / +++ 1926 428 9 Dpy 10 Unc Chromosome III h2764 dpy-5 unc-13 / +++ 2174 452 8 Dpy 13 Unc Chromosome III h2766 dpy-5 unc-13 / +++ 2153 471 14 Dpy 14 Unc Chromosome III h2767 dpy-5 unc-13 / +++ 2061 413 13 Dpy 9 Unc Chromosome III h2768 dpy-5 unc-13 / +++ 1908 389 21 Dpy 8 Unc Chromosome III h2769 dpy-5 unc-13 / +++ 2190 454 23 Dpy 6 Unc Chromosome III h2770 dpy-5 unc-13 / +++ 1906 423 7 Dpy 3 Unc Chromosome III h2772 dpy-5 unc-13 / +++ 1876 1 8 Unc -0.067 (0.0-0.20) h2773 dpy-5 unc-13 / +++ 2020 286 20 Dpy 4 Unc Chromosome III h2774 dpy-5 unc-13 / +++ 2016 422 14 Dpy 14 Unc Chromosome III h2775 dpy-5 unc-13 / +++ 1591 386 8 Dpy 7 Unc Chromosome III h2776 dpy-5 unc-13 / +++ 1517 45 9 Dpy +6.4 (4.4-8.4) h2777 dpy-5 unc-13 / +++ 2080 448 10 Dpy 8 Unc Chromosome III h2778 dpy-5 unc-13 / +++ 2175 524 9 Dpy 11 Unc Chromosome III h2779 dpy-5 unc-13 / +++ 1470 298 8 Dpy 10 Unc Chromosome III h2780 dpy-5 unc-13 / +++ 1588 344 8 Dpy 19 Unc Chromosome III h2781 dpy-5 unc-13 / +++ 1467 311 9 Dpy 10 Unc Chromosome III h2782 dpy-5 unc-13 / +++ 1927 387 17 Dpy 8 Unc Chromosome III h2783 dpy-5 unc-13 / +++ 2140 445 19 Dpy 8 Unc Chromosome III h2784 dpy-5 unc-13 / +++ 2383 28 11 Unc -1.8 (1.3-2.4) h2785 dpy-5 unc-13 / +++ 2276 29 15 Unc -1.8 (0.87-2.7) h2786 dpy-5 unc-13 / +++ 1716 360 15 Dpy 11 Unc Chromosome III h2787 dpy-5 unc-13 / +++ 1877 28 24 Unc -2.1 (0.44-3.7) h2791 dpy-5 unc-13 / +++ 2201 465 16 Dpy 10 Unc Chromosome III h2792 dpy-5 unc-13 / +++ 2272 499 15 Dpy 14 Unc Chromosome III h2793 dpy-5 unc-13 / +++ 2678 45 35 Dpy +4.7 (3.8-5.5) h2795 dpy-5 unc-13 / +++ 1806 279 9 Dpy 16 Unc Chromosome III h2796 dpy-5 unc-13 / +++ 1990 94 5 Dpy 4 Unc Chromosome III h2798 dpy-5 unc-13 / +++ 1123 6 4 Dpy +3.1 (2.2-3.9) Table S3.1: Three-factor mapping data of 69 mitomycin C-treated strains with lethal mutations mapped to chromosomes I or III.         60     Reference : Variant read ratio  dpy-5 marker unc-13 marker Allele C>T I:5,432,433 bp C>A I:5,432,448 bp C>T I:7,434,404 bp h2717 50% : 50% 56% : 44% 35% : 65% h2718 58% : 42% 55% : 45% 61% : 39% h2727 45% : 55% 46% : 54% 44% : 56% h2733 59% : 41% 56% : 44% 58% : 42% h2741 43% : 57% 45% : 55% 48% : 52% h2744 50% : 50% 38% : 62% 38% : 62% h2755 53% : 47% 50% : 50% 25% : 75% h2758 48% : 52% 33% : 67% 44% : 56% h2784 43% : 57% 50% : 50% 41% : 59% h2785 49% : 51% 47% : 53% 63% : 37% h2787 38% : 62% 33% : 67% 50% : 50% Table S3.2: Heterozygosity determined by the ratio of reference:variant reads of dpy-5 and unc-13 markers         Chromosome  Allele I II III IV V X Total h2717   1 1   2 h2718 3 2 1    6 h2727   1  1 2 4 h2733 1    3  4 h2741   1 1 1  3 h2744      1 1 h2755     2  2 h2758   1    1 h2784       0 h2785       0 h2787    1  1 2 Table S3.3: 25 mitomycin C-induced deletions identified using Pindel.       61    Base-difference 10 spontaneously mutating N2 strains 11 EMS-treated strains 11 MMC-treated strains G:C to A:T 80 1832 75 A:T to T:A 86 130 68 G:C to T:A 110 45 51 A:T to G:C 41 71 45 A:T to C:G 44 34 24 G:C to C:G 30 18 27 Total 391 2130 290 Table S3.4: Total single nucleotide variants identified in untreated spontaneously mutating N2, EMS-treated, and MMC-treated strains.            62      Figure S3.1: Forward mutation frequency of mitomycin C, measured as the induction of lethal mutations recovered by the genetic balancer hT2. a) Three separate genetic screens using vial 1 at various storage ages (fresh, 1 week old, 2 week old) gave statistically similar forward mutation frequencies averaging 5.2% (95% confidence intervals) b) Three separate screens using vial 2 at various concentrations (375 µM, 750 µM, and 1400 µM). 375 µM and 750 µM had similar results, averaging 2.2%, while 1400 µM possibly induced sterility. c) The forward mutation frequencies of vial 1 and 2 were compared. The first vial had a forward mutation frequency averaging 5.2%, while the second vial had a forward mutation frequency averaging 2.2%, indicating these two batches of MMC were statistically different in terms of potency.   00.010.020.030.040.050.060.070.080.090-week old 1-week old 2-week oldForward mutation frequencyMitomycin C storageVial 1: forward mutation frequency of 750 µM of fresh, 1 week old and 2 week old mitomycin C00.0050.010.0150.020.0250.03375 uM 750 uM 1400 uMForward mutation frequencyMitomycin C dosageVial 2: forward mutation frequency of 1 week old mitomycin C at varying doses00.010.020.0340.050.060.07Vial 1 Vial 2Forward mutation frequencyMitomycin C batchForward mutation frequency of different batches of mitomycin Ca) b) c) 63   Figure S3.2: The proportion of 5´-CpG-3´ dinucleotides in the deleted regions of DNA was higher than the 15 flanking nucleotides on both sides of the deletion. Standard error bars are shown. n = 238n = 118n = 23802468101214161820Left flanking region Deleted region Right flanking regionProportion of CpG dinucleotides

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.24.1-0167493/manifest

Comment

Related Items