Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The meiotic pattern and chromatin structure in Caenorhabditis elegans Leger, Michelle Marie 2007

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

Item Metadata

Download

Media
831-ubc_2007-0475.pdf [ 3.92MB ]
Metadata
JSON: 831-1.0101000.json
JSON-LD: 831-1.0101000-ld.json
RDF/XML (Pretty): 831-1.0101000-rdf.xml
RDF/JSON: 831-1.0101000-rdf.json
Turtle: 831-1.0101000-turtle.txt
N-Triples: 831-1.0101000-rdf-ntriples.txt
Original Record: 831-1.0101000-source.json
Full Text
831-1.0101000-fulltext.txt
Citation
831-1.0101000.ris

Full Text

THE MEIOTIC PATTERN AND CHROMATIN STRUCTURE IN CAENORHABDITIS ELEGANS by  MICHELLE MARIE LEGER B.Sc, University of York, 2005  THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Genetics)  THE UNIVERSITY OF BRITISH COLUMBIA October 2007 © Michelle Marie Leger, 2007  ABSTRACT  In the nematode Caenorhabditis elegans, local variations in frequencies of crossing over have been observed along chromosomes. These variations give rise to a genetic map that differs from the physical map, with a recombinationally suppressed cluster of genes located in the centre of the autosomes and flanked by highly recombinogenic arms. This distribution of crossover events is known as the meiotic pattern. The pattern is modified in a mutant, rec-1. Mutational alteration of the meiotic pattern allows for investigation of the  mechanisms governing the generation of the meiotic pattern.  In other species, recombination hotspots have been linked to chromatin modifications. In this thesis, I have investigated whether or not chromatin modifiers are required to maintain the unique meiotic pattern found in C. elegans. Mutations in two chromatin modifiers, him17 and lin-35, were analyzed and found to eliminate the meiotic pattern, producing an  increase in the frequency of crossing over within a recombinationally suppressed region, and a decrease in the frequency of crossing over within a highly recombinogenic region. This is the first demonstration that meiotic patterning in C. elegans is established by chromatin structure.  ii  INDEX  ABSTRACT  ii  PWDEX  iii  LIST OF T A B L E S  vi  LIST OF F I G U R E S  vii  ACKNOWLEDGEMENTS  viii  DEDICATION  ix  I. INTRODUCTION  1  I . l . Introduction to the meiotic pattern  1  I. 2. Caenorhabditis elegans as a model organism  2  1.3. Meiotic prophase in Caenorhabditis elegans  3  I. 4. The meiotic pattern is altered in a mutant  8  I. 5. Chromatin modifiers and the meiotic pattern  10  I. 6. Why a meiotic pattern?  12  I. 7. Research goals  15  II. M A T E R I A L S A N D M E T H O D S  ,  18  II. 1. General methods  18  II. 2. Calculating recombination frequencies  21  II. 3.ccls4251  22  II. 4. rec-1 mapping using dpy-5, dpy-13, ccls4251, lin-11 and unc-75  23  II. 5. Sequencing  24  II. 6. Scoring males  26  III. R E S U L T S III. 1. Crossing over in a him-] 7 mutant III. 1.1. Mutant him-17 increased crossing over within the central cluster on LGI  27 27 27  in  III. 1. 2. Mutant him-17 decreased crossing over within a flanking region on LGI  29  III. 1. 3. Mutant him-17 decreased crossing over within a flanking region on LGIII... 29 III. 1. 4. Mutant him-17 increased crossing over in a hybrid region on LGIII  31  III. 1. 5. Crossing over in the cluster of LGI in a rec-l;him-l 7 double mutant was not significantly different from that in each single mutant  31  III. 1. 6. Crossing over in a flanking region on LGIII in a rec-1 ;him-17 double mutant was not significantly different from that in each single mutant  33  III. 1. 7. The effects of rec-1 and him-17 on crossing over in a hybrid region were additive III. 1. 8. Summary: mutant him-17 eliminates the meiotic pattern III. 2. Crossing over in a lin-35 mutant  33 36 41  III. 2. 1. Mutant lin-35 increased crossing over within the central cluster of LGV  41  III. 2. 2. Mutant lin-35 decreased crossing over within a hybrid region on LGV  43  III. 2. 3. Mutant lin-35 decreased crossing over within a flanking region on LGIII  43  III. 2. 4 Summary: mutant lin-35 eliminates the meiotic pattern  45  III. 3. Crossing over in the presence of a heterologous insertion  45  III. 3. 1. 7>ara-heterozygous ccls4251 decreased crossing over in the cluster and on the arm on the same chromosome III. 3. 2. The effects of ccls4251 on crossing over were localized  45 49  III. 3. 3. Homozygous mutant tam-1 reversed the local recombinational suppression of heterozygous ccls4251 over a flanking region  49  III. 3. 4. Summary: a multicopy transgene decreases the frequency of crossing over on the same chromosome  49  III. 4. Mutant him-17 does not suppress ccls4251 expression  51  III. 5. Mutant rec-1 does not suppress ccls4251 expression  51  iv  III. 6. Mapping rec-1 III. 6. 1. Gross mapping of rec-1 using LGI markers and a transgene III. 6. 2. Summary: the rec-1 gene maps between dpy-5 and the insertion site of ccls4251 III. 7. rec-1 \s not klp-16  III. 8. X-chromosome nondisjunction or loss IV. DISCUSSION IV. 1. General discussion  61  IV. 2. Future work  69  IV. 3. Conclusion  71  V. BIBLIOGRAPHY  72  APPENDIX 1  84  v  LIST OF T A B L E S  Table 1: Primers used to amplify klp-16  25  Table 2: Crossing over in a him-17 strain in a cluster region  28  Table 3: Crossing over in a him-17 strain in a flanking region on LG1  30  Table 4: Crossing over in a him-17 strain in a flanking region on LGIII  30  Table 5: Crossing over in a him-17 strain in a hybrid region  32  Table 6: Crossing over in a rec-1 ;him-17 strain in a cluster region  32  Table 7: Crossing over in a rec-1 ;him-17 strain in a flanking region  34  Table 8: Crossing over in a rec-1 ;him-17 strain in a hybrid region  35  Table 9: Crossing over in a //«-35 strain in a cluster region Table 10: Crossing over in a lin-35 strain in a hybrid region  42 44  Table 11: Crossing over in a lin-35 strain in a flanking region  44  Table 12: The local effects of rrara-heterozygous ccls4251 on crossing over  48  Table 13: The effects of heterozygous ccls4251 on a different chromosome  50  Table 14: The effects of mutant tam-1 on crossing over when heterozygous ccls4251 is present Table 15: ccls4251 mapping data  50 54  Table 16: Left-right mapping of rec-1 in dpy-5 ccls4251 recombinants  54  Table 17: Left-right mapping oi rec-1 in ccls4251 unc-75 recombinants  55  Table 18: Positioning rec-1 in dpy-5 unc-13 ccls4251 strains Table 19: Strains used  57 84  LIST OF FIGURES  Figure 1: Events in C. elegans meiotic prophase 1  7  Figure 2: Mutant rec-1 eliminates the meiotic pattern on LGI  9  Figure 3: Mutant him-17 eliminates the meiotic pattern on LGI  37  Figure 4: Mutant him-17 eliminates the meiotic pattern on LG1II  38  Figure 5: Double mutant rec-1 ;him-l 7 has the same phenotype as rec-1 and him-17 in the cluster on LGI  39  Figure 6: The effects of rec-1 and him-17 on the meiotic pattern are additive  40  Figure 7: Mutant lin-35 increases crossing over in the cluster of LGV  46  Figure 8: Mutant lin-35 decreases crossing over in a region on the arm of LGIII  47  Figure 9: Mutant him-17 does not suppress the expression of a multicopy tandem-array transgene  51  Figure 10: Mutant rec-1 does not suppress the expression of a multicopy tandem-array transgene  52  vii  ACKNOWLEDGEMENTS  I would like to thank my research supervisor, Ann Rose, who introduced me to this project and has supported me throughout. I am greatly indebted to her for her guidance, advice, patience, and valuable discussion. I would also like to thank the other two members of my supervisory committee, Louis Lefebvre and Angela Brooks-Wilson, for their thoughtful comments as my work progressed. 1 would further like to thank Maja Tarailo, Jillian Youds, Yang Zhao, Shir Hazhir and Sanja Tarailo, to whom I am grateful for their patience and encouragement as they introduced me to the quirks of C. elegans; Petr Pospisil, for his help in positioning rec-1; members of the Rose lab for helpful comments on this thesis; and Rachel's Brain for fruitful discussion. Finally, I would like to thank Dr. Nigel O'Neil, whose advice and discussion, particularly about rec-1, proved especially helpful.  viii  Dedicated to my parents, Lucien and Mary Leger  I. INTRODUCTION  1.1. Introduction to the meiotic pattern  Non-random patterns of crossing over have been observed in many species, including humans; most often, these are apparent in the form of recombination hotspots (reviewed in Lichten and Goldman, 1995, de Massy, 2003, and Mezard, 2006). However, the distribution of crossing over varies considerably between species (Jensen-Seaman et al., 2004), and the frequency of crossing over may vary widely within a species: sexual dimorphism in recombination rates is common (Morgan, 1914, Singer et al., 2002), and variations between individuals have been reported (reviewed in Lynn et al., 2004). The nematode Caenorhabditis elegans provides an excellent model system for the study of the mechanisms  governing crossover distribution, because of its striking meiotic pattern and strong crossover interference.  The first genetic maps of C. elegans were produced by Sydney Brenner in 1974. Although at the time no physical maps were available for comparison, it was apparent that the genetic maps of each chromosome displayed a strong clustered pattern of markers toward the centre of each chromosome. Brenner suggested that this might be attributed to local differences in recombination frequency: "One possibility is that the clusters are produced by a lower frequency of recombination in a defined region of a chromosome - for example, near the centromere" (Brenner, 1974). Brenner's hypothesis was confirmed when the physical and genetic maps were compared with one another (Greenwald et al., 1987). The centre of each chromosome is recombinationally suppressed, while the arms are recombinogenic,  1  undergoing relatively high rates of recombination, although on the X chromosome this pattern is less pronounced (Brenner, 1974).  In the hermaphrodite, strong crossover interference has been noted, producing a single recombination event per chromosome (Brenner, 1974), although double crossovers have been found on a male autosome (Hodgkin et al., 1978). An intriguing extension of crossover interference has been reported in whole chromosome fusions, in which crossover remains limited to a single event per fused pair (Hillers and Villeneuve, 2003). Further, crossover suppression in one region of the chromosome as a result of translocation results in a compensatory increase in crossing over in the region immediately adjacent (McKim et al., 1988). The lack of double crossovers in C. elegans eliminates the need to distinguish between a lack of crossing over, and two crossovers placed relatively close together. It also permits more precise determination of the site of each crossover event.  I. 2. Caenorhabditis  elegans as a model organism  C. elegans is an ideal model organism for geneticists on account of its small size, relatively compact genome, easy maintenance, and the possibility of either outcrossing or selffertilizing. Wild-type C. elegans hermaphrodites (5A, XX) are self-fertilizing, initially producing sperm, which is stored in a spermatheca until it can be used to fertilize the eggs that are subsequently produced. Males (5A, XO) arise spontaneously through Xchromosome loss or nondisjunction at a rate of approximately 1.07 per 1000 progeny at 20°C (Rose and Baillie, 1979a). Because outcrossed hermaphrodites use male sperm preferentially, males can be maintained by crossing to hermaphrodites, producing 50% hermaphrodite and 50% male progeny.  2  I. 3. Meiotic prophase in Caenorhabditis elegans  Meiosis is the mechanism by which genetic material is segregated into germ cells, in the process reducing the amount of genetic material in each cell from 2n chromosomes in somatic cells to n chromosomes in the germ cells. Correct segregation of DNA is vitally important: failure of a germ cell to receive its complement of DNA results in aneuploid progeny if that germ cell contributes to an embryo.  In C. elegans, the progression of germ nuclei through the stages of meiosis can be observed by 4',6-diamidino-2-phenylindole (DAPI) staining in each of the two gonads in an adult hermaphrodite (Dernburg et al., 1998), where oogenesis occurs. In contrast, germ nuclei undergoing spermatogenesis, which takes place in L4 stage larval hermaphrodites or in males, are not as easily observable. Each gonad can be divided into a number of distinct zones, each corresponding to a particular stage of meiosis (Dernburg et al., 1998). The distal tip of the gonad arm contains the mitotic zone, in which germ cells proliferate through mitotic division. Adjacent to this zone is the transition zone, in which germ nuclei begin to enter meiotic prophase I (Dernburg et al., 1998). Prophase I is subdivided into five stages, leptotene, zygotene, pachytene, diplotene and diakinesis. The transition from diakinesis to metaphase 1 begins in the gonad arm and is completed in the spermatheca, where the oocyte is fertilized (McCarter et al., 1999). Meiotic progression relies on three interdependent processes that take place in prophase I: pairing, recombination, and synapsis.  Pairing is first observed in leptotene and zygotene nuclei (Dernburg et al., 1998). Pairing involves recognition and stable alignment of the homologs, governed by homolog  3  recognition regions (HRRs) (McKim et al., 1993), also known as pairing centres (PCs) (Villeneuve, 1994). These short sequence elements, located at one end of each homolog, have been mapped by examining pairing in different translocation or duplication heterozygotes (McKim et al., 1988; Villeneuve, 1994). Secondary pairing sites do not confer the ability to pair, yet are necessary for recombination to occur over the entire chromosome (McKim et al., 1993). Pairing is mediated by four zinc-finger proteins, each limited to specific autosome pairs: ZIM-1 (LGII and LGIII), ZIM-2 (LGV), Z1M-3 (LGI and LGIV) and HIM-8 (LGX) (Phillips and Dernburg, 2006). Homologs are tethered to the nuclear envelope during pairing (Phillips et al., 2005; Penkner et al., 2007); however, while this appears to be necessary, it is not sufficient to produce stable pairing (Phillips et al., 2005).  Chromosomes then progress through synapsis, with the formation of the synaptonemal complex (SC), a proteinaceous scaffold that can be observed in pachytene (Dernburg et al., 1998). While stable pairing can occur in the absence of synapsis, synapsis does not normally progress in unpaired chromosomes (MacQueen et al., 2005). Chromatin is organized along chromosome axes, in a series of loops along which proteins align to form the axial elements (Colaiacovo, 2006). These axes have been implicated in crossover interference, as mutants lacking one of the protein components, HIM-3, undergo double crossovers (Nabeshima et al., 2004). When reinforced in the context of the mature SC, these are known as lateral elements, and are bridged by transverse filaments that form the central region of the SC (reviewed in Colaiacovo, 2006). Coordination between pairing and synapsis is maintained by the HTP-1 protein, which inhibits SC polymerization until mature chromosome axes have formed, preventing non-homologous synapsis, and recombination between sister chromatids (Couteau and Zetka, 2005; Martinez-Perez and Villeneuve, 2006).  4  Homologous recombination ensures that the two homologs remain together in preparation for segregation, in the process acting as an engine of genetic diversity by generating new combinations of alleles. Double-strand break (DSB) formation is not a prerequisite for the initiation of synapsis (Dernburg et al., 1998), nor is synapsis required for the initiation of DSB formation (Colaiacovo et al., 2003), but DSB formation initiates prior to synapsis. Recombination initiates with cleavage of one of the homologs by the endonuclease SPO-11 to create a DSB. In the absence of SPO-11, DSBs can be formed by irradiation, rescuing the Spo-11 phenotype (Dernburg et al., 1998). Following the removal of SPO-11, strand resection occurs: the 5' end of each cut double-strand is degraded, possibly by MRE-11 and/or RAD-50, leaving a short section of single-stranded 3' DNA. This single strand then undergoes strand invasion, in which the single strand locates its equivalent sequence on the homolog and invades it to create a displacement-loop (D-loop). The recombinase RAD-51 is a homolog of RecA, which has been shown to bind ssDNA and mediate strand invasion in yeast; RAD-51 is believed to carry out the same function in C. elegans (reviewed in GarciaMuse and Boulton, 2007). The C. elegans BRCA2 homolog CeBRC-2 is believed to act as a co-mediator by loading RAD-51 onto ssDNA (Martin et al., 2005) and stimulating D-loop formation (Petalcorin et al., 2006); two further proteins, HIM-14 and MSH-5, have also been implicated in this step (Carlton et al., 2006). RAD-51 forms discrete foci, and antibody staining of these foci has provided a means of measuring the timing of strand exchange, which starts in late leptotene/early zygotene and peaks in pachytene (Alpi et al., 2003). Information on events following strand invasion is relatively sparse in C. elegans, but abundant in yeast. In yeast pachytene, extension of the D-loop and capture of the second 3' single strand from the first homolog leads to the formation of a double Holliday junction (dHJ), a four-way connection between both strands of each homolog that can be resolved by cleavage of two strands (reviewed in Villeneuve and Hillers, 2001; Garcia-Muse and  5  Boulton, 2007). Resolution in C. elegans is mediated by MUS-81 (Garcia-Muse and Boulton, 2007). DSB formation and RAD-51 loading do not appear to be dependent on pairing, as RAD-51 foci appear on schedule in the pairing-deficient him-3 mutant. However, in these mutants, RAD-51 levels persist beyond wild-type and the DNA damage checkpoint is activated, suggesting that the DSBs undergo non-crossover repair (Couteau et al., 2004). Synapsis is required for the completion of recombination, and failure of even one pair of homologs to synapse leads to widespread appearance of recombination intermediates and delay of meiotic progression (Colaiacovo et al., 2003; Carlton et al., 2006). However, although crossing over is almost completely eliminated in synapsis-deficient syp-3 mutants, DSBs are repaired through sister-chromatid mediated repair and non-homologous endjoining (Smolikov et al., 2007).  SYP-1 and SYP-2 are two structural proteins that are found distributed evenly throughout the central region of the SC. In late pachytene and early diakinesis, these proteins become asymmetrically localized, and asymmetric disassembly of the SC follows (Nabeshima et al., 2005). At this point, chiasmata become evident at the site of crossovers (Nabeshima et al., 2005); these physical connections ensure that the homologs remain stably associated in preparation for segregation (Dernburg et al., 1998). In crossover-deficient mutants, asymmetric localization of SYP-1 and SYP-2 is disrupted, but can be rescued by the formation of DSBs as a result of gamma irradiation (Nabeshima et al., 2005).  6  HRR • Secondary pairing site | Sislcr chromatid cohesion complex • H  IZIM protein or HIM-X  s i s t e r  One homolog  Transition zone  c n r o m a t i d s  DSB formation  \  ^  SPO-11  EMRE-U?RAD-5O?  Early pachytene Strand resection  Strand exchange m  •  HTP-1 allows SC polymerization  •  RAD-51 CcBRC-2  •  Mature SC  IIIM-14  Mid pachytene  MSH-5  DNA synthesis and double Holliday junction formation?? XXXX Synaptonemal complex - lateral elements and transverse filaments Chiasina  Asymmetric SC dissociation  Resolution  .  Homologous recombination  xxxxxxxx-  Diplotene  Synapsis  Diakinesis  Figure 1: Events in C. elegans meiotic prophase I. Question marks indicate incomplete data; arrows between the Homologous recombination and Synapsis categories indicate steps in one category necessary for steps in the next.  7  During mitosis, spindle fibres attach along the entire length of the C. elegans chromosomes (Albertson and Thomson, 1982). However, during meiosis, genetic analysis demonstrated that a single, localized meiotic centromere must be present. Early studies of recombination in duplication (Rose et al., 1984), translocation (Rosenbluth and Baillie, 1981; McKim et al., 1988) and inversion (Zetka and Rose, 1992) heterozygotes found that recombination could still occur at one end of the affected chromosome, consistent with a single, localized meiotic centromere. In addition, cytological studies indicate that microtubules attach to a localized point along the chromosome (Albertson and Thompson, 1993).  I. 4. The meiotic pattern is altered in a mutant  The pattern of crossing over can be altered in C. elegans. The frequency of crossing over in a cluster region is dependent on external factors, decreasing with maternal age and increasing with temperature (Rose and Baillie, 1979a). The frequency of crossing over is higher in hermaphrodite sperm than in male sperm, leading to a higher overall frequency of crossing over in the hermaphrodite (Zetka and Rose, 1990); the reasons for this are not yet understood. The meiotic pattern can be eliminated completely by gamma irradiation, which introduces DSBs that may become substrates for meiotic crossing over (Kim and Rose, 1987).  A mutation in rec-1 that alters the meiotic pattern by increasing recombination over the meiotic cluster was discovered by Rose and Baillie (1979b). This provided the first evidence that the meiotic pattern was genetically controlled, rather than being solely a result of sequence features such as GC content. Later it was found that rather than producing a general increase in recombination, the rec-1 mutation eliminates the meiotic pattern by  8  producing a multifold increase in recombination over the recombinationally suppressed meiotic cluster, and a similar decrease in recombination over the recombinogenic regions on the arms (Fig. 2; Zetka and Rose, 1995).  dpy-5 Genetic map in N2 hermaphrodite  unc-1l\ unc-13  bli-3  unc-11 dpy-5 unc-13  bli-3  unc-54  unc-101  unc  .101  unc-54  Physical map  G e n e t i c m a p bli-3  unc-11  dpy-5  unc-13  unc-101 unc-54  in rec-1 hermaphrodite  Genetic maps: Physical map: •  5 map units 1 Mb  Figure 2: Mutant rec-1 eliminates the meiotic pattern on LGI. Figure based on Zetka and Rose, 1995; genetic map data taken from Zetka and Rose, 1995, physical map data taken from WormBase. Dashed lines are used to contrast the relative positions of each gene along each genetic map with their position along the physical map.  Several factors have confounded mapping efforts. In order to determine the Rec-1 phenotype, the progeny must be counted, limiting the number of individuals that can be scored (Rose and Baillie, 1979b). The fact that mutant rec-1 produces a decrease in crossing over in some regions (Zetka and Rose, 1995) does not solve the labour-intensive problem. 9  The mutation arose early, and exists in a number of strains not scored for recombination (A. Rose, personal communication). As a result, rec-1 has been difficult to map. It was positioned on the right arm of LGI (A. Rose, unpublished); this position was confirmed when it was shown to be in the region encompassed by the duplication sDpl, which covers the right arm of LGI (Zetka and Rose, 1995). Subsequent positioning has placed it between dpy-5 and unc-75 (M. Tarailo, personal communication). To date, rec-1 has not been cloned,  and its molecular basis has not been determined.  I. 5. Chromatin modifiers and the meiotic pattern  The effect of chromatin modifiers and meiotic recombination has been studied most extensively in yeast. In 1994, Wu and Lichten discovered a correlation between the site of recombinational hotspots and promoter regions. In 1997, a single base pair mutation was discovered in Schizosaccharomyces pombe that created a recombinational hot spot. Mutants displayed new micrococcal nuclease-hypersensitivity and altered nucleosome phasing at the site of the mutation (Mizuno et al., 1997). Mizuno et al. have suggested that this mutation produced local alterations in chromatin structure, promoting access of double-strand breakcreating enzymes to the DNA. Following this discovery, Yamada et al. (2004) reported hyperacetylation of H3 and H4 residues of histones surrounding the new hotspot. They also identified two cellular components implicated in hyperacetylation at this site, a histone acetyltransferase, and a SWI/SNF-ATP-dependent chromatin remodelling factor.  In other species, a role of chromatin remodelling in homologous recombination is being studied. A mutation in INO80, a SWI/SNF ortholog in Arabidopsis thaliana, decreases homologous recombination to 15% of wild-type, apparently without affecting DNA damage  10  repair pathways. The authors suggest that the fact that transcription of genes in recombination and repair pathways in this mutant were not altered points to INO80 having a direct effect on recombination, rather than affecting transcription of a distinct recombination factor (Fritsch et al., 2004). In mammals, recombination in the assembly of immunoglobulin genes in B lymphocyte development provides an interesting opportunity to study recombination independent of meiosis. In mouse, this process has been found to depend on nucleosome remodelling and histone H3 hyperacetylation (McMurray, 2000, Maes et al., 2006).  Recently, a paper by Reddy and Villeneuve (2004) reported a link between meiosis and chromatin structure in C. elegans. Germ cells in him-17 mutants undergo generalized meiotic nondisjunction, which can be rescued by irradiation, consistent with the absence of double-strand break formation, him-17 animals also exhibit altered patterns of the Lys9 residue of the histone H3 tail, a chromatin modification associated with a closed chromatin configuration. They proposed that the localization of crossovers is regulated by alterations in chromatin structure; these allow or deny access of SPO-11 to chromatin, affecting the placement of double-strand breaks (see Discussion). Reddy and Villeneuve (2004) note that the altered H3-Lys9 methylation pattern and meiotic phenotype observed in him-17 mutants may be ascribed to independent functions of HIM-17. However, they also note that the effect oi lin-35 (RNAi) on the meiotic phenotype supports the idea that the two functions are linked. They acknowledge the possibility of HIM-17 acting indirectly in meiosis by affecting gene expression, but cite normal expression of meiotic machinery genes as evidence against this.  11  Other work on meiosis and chromatin structure in C. elegans is sparse, but some noteworthy observations have been made in the context of unpaired DNA. Unpaired DNA is transcriptionally silenced in meiosis in C. elegans (Kelly et al., 1997). Reporter transgene arrays that are well expressed in somatic tissues are nevertheless silenced in the germline. The male X chromosome - which is unpaired - accumulates H3-Lys9 methylation at pachytene. In Him-8 mutants, X chromosome pairing is defective; in these mutants a similar pattern of H3-Lys9 methylation has been observed on both X chromosomes (Bean et al., 2004).  Bean et al. (2004) also found that three autosomal duplications at pachytene are enriched for H3-Lys9 methylation. One of these, sDp2, is unpaired and does not undergo recombination (Rose et al., 1984, McKim and Rose, 1990). However, a second duplication, sDpl, which is paired and undergoes recombination (Rose et al., 1984, McKim and Rose, 1990), also shows high levels of H3-Lys4 methylation - a modification that is found in transcriptionally competent chromatin (Bean et al., 2004).  I. 6. Why a meiotic pattern?  While the possibility of chromatin structure as the mechanism underlying the meiotic pattern is interesting, it does not present an evolutionary explanation for the meiotic pattern. Any non-random pattern of chromatin modifications begs the question of why there should be such a pattern at all.  In Saccharomyces cerevisiae, an elegant answer has already been proposed. Wu and Lichten (1994) showed that the sites of DSBs in three regions of the genome correspond closely to  12  predicted promoter sites. They suggest that the location of DSBs can be accounted for by pre-existing chromatin modifications involved in transcriptional regulation, which make these regions accessible to transcriptional machinery, and as a side effect, to yeast Spol 1. However, this does not explain the pattern in C. elegans, where the meiotic pattern does not correlate positively with gene density (and, therefore, with promoter sites). Even a pattern correlating with the promoter regions of genes during meiosis would require an explanation as to why these genes might be found preferentially on the arms, especially on all chromosomes.  A number of alternative explanations have been suggested to explain the meiotic pattern in C. elegans. That crossover location should be determined by location along a chromosome -  i.e. that crossover factors might more easily initiate recombination a certain distance from each telomere - seems unlikely, given that cluster size does not appear to be a function of overall chromosome length (Barnes et al., 1995).  One possible explanation for the meiotic pattern is the need to protect the genome from transposable elements (TEs) (Kelly et al., 1997). This is particularly important in the germline, at a time when the mutagenic effects of TE excision and insertion could damage members of the next generation. This hypothesis is supported by the observation that the C. elegans transposon Tel has been found to excise at a high frequency only in somatic cells  (Emmons and Yessner, 1984). Moreover, the distribution of (DNA-based) transposons and non-autonomous miniature inverted-repeat transposable elements (MITEs), though not of (RNA-based) retrotransposons, has been found to correlate positively with recombination rate. This relationship holds true in both coding and noncoding regions, suggesting that the mechanism responsible is independent of selection against transposons inserted into coding  13  regions (Duret et al., 2000). This might be seen as evidence that chromatin modifications are indeed involved in suppressing the spread of TEs, by preventing reinsertion rather than excision. However, the distribution of retrotransposons appears to be a counterargument. Surely these elements, which require transcription in order to transpose, should be preferentially silenced; instead, their density does not correlate either positively or negatively with recombination rates (Duret et al., 2000). In addition, the findings in C. elegans contrast with those in other species, such as Arabidopsis thaliana (Wright et al.,  2003), Drosophila melanogaster (Bartolome et al., 2002) and human (Boissinot et al., 2001), where high density of TEs occurs in regions of low recombination.  A different model, put forward by Barnes et al. (1995), held that meiotic recombination is promoted by short sequence elements located in noncoding DNA. The meiotic pattern can then be seen as a consequence of gene distribution along the chromosome, with the frequency of crossing over being greater in regions of noncoding DNA. However, later physical mapping appears to show a more uniform distribution of genes along the chromosome, in contrast to the meiotic pattern (Brenner, 1974; C. elegans Sequencing Consortium, 1998).  A third, promising model involves the need for asymmetry inherent in meiosis I. Disassembly of the synaptonemal complex is asymmetrical, and dependent on crossing over for completion; it has been suggested that crossovers act as symmetry-breaking events. Asymmetric disassembly in turn is hypothesized to determine the point of attachment of the spindle fibres at metaphase I (Nabeshima et al., 2005). Crossovers would therefore need to occur on either side of the chromosome, but not at the centre. The Rec-1 phenotype argues against this, as rec-1 mutants do not display severe meiotic phenotypes consistent with  14  major disruption of either SC disassembly or failure to form proper spindle attachment. One could argue that crossing over in only a tiny portion of the centre would result in failure of the SC to disassemble correctly, resulting in a very slight decrease in viability that would not be detected as significant in most cases. However, this raises the question of why recombination should then be suppressed over the entire meiotic cluster. Furthermore, spindle attachment cannot be solely dependent on crossover location, as the unpaired male X chromosome is able to attach despite not undergoing crossing over (Wicky and Rose, 1996).  A non-random pattern of crossing over appears to be a universal theme in eukaryotes (Lichten and Goldman, 1995, de Massy, 2003, and Mezard, 2006), although the precise nature of the pattern differs between organisms. The evolutionary basis of the pattern in C. elegans has not yet been satisfactorily explained. Clarification of the mechanism underlying  the pattern may elucidate the evolutionary basis and importance of the meiotic pattern.  I. 7. Research goals  To examine whether the meiotic pattern might be established by altering chromatin structure, I examined the effects of known chromatin modifiers on the meiotic pattern.  LIN-35 is a chromatin modifier of the retinoblastoma family. It has recently been identified as a member of the DRM complex (Harrison et al., 2006), which is believed to function in transcriptional repression, and interacts with the SWI/SNF complex during development (Cui et al., 2004). lin-35 mutants are synthetic multivulva class B (SynMuv B) mutants (Lu and Horvitz, 1998). Double mutants with loss-of-function mutations in a SynMuv B gene and any gene in another SynMuv class (A or C) exhibit a multivulva phenotype that is not  15  found in either single mutant (Lu and Horvitz, 1998, Ceol and Horvitz, 2004). This synthetic phenotype can be attributed to the redundant action of SynMuv genes to antagonize the Ras signalling pathway; however, SynMuv genes have been implicated in generalized transcriptional repression (reviewed in Fay and Yochem, 2007). lin-35 mutants also exhibit enhanced RNAi (Lehner et al., 2006) and silencing of tandem-array transgenes (Hsieh et al., 1999). The silencing effects of mutant LIN-35 are context-dependent: when the transgene consists of simple repeats, transgene expression is suppressed. However, when the repeats are interspersed with random fragments of genomic DNA, expression is not affected. Thus LIN-35 would appear to act on the repetitive nature of the transgene, rather than being recruited by individual signals within the sequence (Hsieh et al., 1999).  HIM-17 is a chromatin modifier that is required for double-strand break formation and meiotic progression. The Him (high incidence of males) phenotype of him-17 mutants is an indicator of X chromosome nondisjunction or X chromosome loss, him-17 mutants display massive meiotic nondisjunction and altered patterns of histone H3-Lys9 methylation (Reddy and Villeneuve, 2004), and him-17 interacts genetically with ego-1, another gene implicated in H3-Lys9 methylation of unpaired DNA (Maine et al., 2005). Recombinational suppression has also been reported in these mutants (Reddy and Villeneuve, 2004); however, recombination was examined only in the null mutant, in which chiasmata do not form, lin-35 (RNAi) was found to enhance the severity of the non-null phenotype of him-17 animals (Reddy and Villeneuve, 2004). Additionally, a function of HIM-17 in regulating the meiotic entry vs. germline proliferation decision has been described (Bessler et al., 2007).  TAM-1 is a putative chromatin modifier discovered by Hsieh et al. (1999). It contains a RING-finger motif, which has been associated with members of the polycomb family of  16  transcriptional repressors. Like lin-35 mutants, tam-1 mutants are SynMuvB mutants and exhibit context-dependent gene silencing, tam-1 has recently been linked to developmental regulation in foregut cells (Kiefer et al., 2007).  In this thesis, the meiotic phenotypes of mutant lin-35, him-17 and tam-1 were scored and compared to the Rec-1 phenotype, to test whether chromatin modifiers are required to maintain the meiotic pattern in C. elegans. The effects on the meiotic pattern of a large multicopy tandem-array transgene, ccls4251, were also studied. In addition, mutants were used to refine the map position of rec-1.  Understanding the molecular basis of the meiotic pattern in C. elegans will permit a more thorough understanding of a basic biological process, and will facilitate the use of C. elegans as a model organism. It may help to elucidate the factors governing the distribution  of crossing over events in other, less easily studied species and explain the wider importance of this distribution.  17  II. MATERIALS AND METHODS  II. 1. General methods  Most strains used in this thesis were originally obtained from David Baillie at Simon Fraser University, or from the Caenorhabditis Genetics Centre at the University of Minnesota. The CB4523 strain, along with a second non-lethal him-17 strain, CB6066, were generously  provided by Jonathan Hodgkin at the University of Oxford. The following alleles were used  (all alleles are shown left to right): LGI: dpy-5(e61), unc-13(e51), Un-35(n745), rec-l(sl80) ccls4251, lin-ll(n389), unc-75(e950), unc-lOl(ml), unc-54(el90). LGIII: unc-32(el89), dpy-18(e364), unc-64(e246). LGV: unc-60(m35), tam-1 (cc567), dpy-1 l(e224), unc42(e270), him-17(e2707), him-5(el467). For strains used, see Appendix 1.  All strains were maintained and mated at 20° on Nematode Growth Medium (NGM) streaked with Escherichia coli strain OP50. All experiments were carried out at 20°. The frequency of X-chromosome nondisjunction rises with temperature. Where only hermaphrodites of a desired genotype were available, males for mating were obtained by incubating L4 hermaphrodites at 30° for 5 - 6 hours. For him-17 animals, this step was unnecessary, as a much higher proportion of males are produced at 20°.  At the L4 larval stage, hermaphrodites do not yet produce oocytes (Schedl, 1997), and the vulva has not yet fully developed (Greenwald, 1997). By picking hermaphrodites at this stage, one can be sure that they have neither been mated by a male nor self-fertilized. Mating was achieved by placing males together with L4 hermaphrodites; large numbers of males in the Fl generation were used as an indication that successful mating had occurred.  18  L4 hermaphrodites from this generation were picked to individual plates and allowed to selffertilize to produce the F2 generation, which were then scored for most crosses.  The genetic distance between markers (a, left and b, right) in a mutant him-17 background was measured by scoring recombinants from an a b /+  him-17/him-17 hermaphrodite.  Animals bearing the him-17 deletion allele ok424 were not used, as viability in these animals is only 5% (Reddy and Villeneuve, 2004), making recombination very difficult to score in these mutants. Instead, the CB4523 strain, bearing the non-null allele e2707, was used. Viability in this strain is 63% (Reddy and Villeneuve, 2004). Visible markers were chosen that had already been used to characterize the Rec-1 phenotype. The dpy-5 unc-13 interval is located in the meiotic cluster of LGI; it is one of the original intervals used to characterize the Rec-1 phenotype (Rose and Baillie, 1979b). Recombination over this interval undergoes a multifold increase in a rec-1 background (Rose and Baillie, 1979b, Zetka and Rose, 1995, see Introduction), and is frequently used to test strains for the presence of mutant rec-1. The unc-101 unc-54 interval, located on LGI(right), was shown to undergo a multifold decrease  in recombination in a mutant rec-1 background (Zetka and Rose, 1995).  him-17 e2707 males were first crossed to double mutant hermaphrodites. The progeny were  picked at L4 and allowed to self-fertilize. L4 hermaphrodites of the F2 generation were picked to individual plates, and progeny of animals giving both double mutant and male progeny were scored. Progeny of animals giving double mutant progeny but no males were scored as controls. All recombinants were plated individually and progeny-tested (where hermaphrodite) or inspected 24 hours later to ensure that the phenotype had been correctly identified (where male).  19  rec-1;him-17 double mutant strains were constructed by crossing him-17 males to dpy-5 unc-13 rec-1 hermaphrodites. L4 hermaphrodites from the F l generation were allowed to self-fertilize, and Dpy Uncs giving male progeny were picked from the F2 generation. A dpy-5 unc-13 rec-1(1);him-J 7(V) strain was bred from these animals, rec-1 males were then mated to dpy-5 unc-13 rec-1 (I);him-17(V) hermaphrodites, and L4 hermaphrodites from the F l generation were allowed to self-fertilize. L4 wild-type hermaphrodites from the F2 generation were then picked individually, and allowed to self-fertilize. The progeny of animals giving Dpy Unc and male progeny were scored; the progeny o f animals giving Dpy Unc but no male progeny were scored as controls. A rec-1 ;him-17 strain was isolated from these crosses. Experiments measuring crossing over on LG11I in the double mutant were achieved by mating rec-1 ;him-17 males to strains bearing the rec-1 mutation on LGI, and the desired markers on LGIII, and scoring the progeny of F2 animals giving Dpy Uncs and male progeny.  lin-35 is on LGI, and has been mapped to +0.46 c M (WormBase). lin-35 mutants do not have a visible phenotype at 20°; the null mutation used is a SNP that does not introduce a restriction site, and thus cannot be followed using PCR. Building a strain bearing both mutant lin-35 and visible markers on LGI such as dpy-5 (map position 0.0) would thus be unduly cumbersome, particularly in light of the absence of a visible Lin-35 phenotype or PCR markers. For this reason, markers on other chromosomes were used to test for effects of mutant lin-35 on crossing over. The presence of homozygous mutant lin-35 in the desired crosses was assured by crossing lin-35 males to a previously constructed ccls4251(I);dpy-l 1 unc-42(V) or cch4251(I);unc-60 dpy-ll(V) strain, and picking L4 hermaphrodites of the F2 generation individually. ccls4251 animals express nuclear-targeted and mitochondriallytargeted G F P under the control of the myo-3 promoter. The progeny of animals giving Dpy  20  Uncs and not expressing GFP were scored. All recombinants were plated individually and progeny-tested. Where progeny testing could not be accomplished due to sterility or premature death of the recombinants, initial recombination frequencies were calculated based first on the assumption that the recombinant phenotype had been correctly identified, then on the assumption that it had not, and that the animals were in fact wild-type. The most conservatively estimated recombination rate - i.e. that which differed least from the wildtype recombination rate - was then retained and included here.  In order to measure the dpy-18 unc-64 distance in rec-1 and lin-35 backgrounds, blind crosses were first carried out. rec-1 or lin-35 males were crossed to dpy-18 unc-64 hermaphrodites, hermaphrodites from both the Fl and F2 generation were picked at L4 and allowed to self-fertilize, and the progeny of F2 hermaphrodites giving Dpy Unc progeny were scored. A quarter of the F2 were expected to be mutant for rec-1 or lin-35 respectively; the rest could act as negative controls. Although the frequency of crossing over in both mixed populations was lower than wild-type, it was difficult in some cases to make a clear distinction between mutant and wild-type plates that would have allowed for accurate calculations of map distances in each type of background. For this reason, Dpy Uncs from plates with recombination rates clearly lower than wild-type were picked and mated to both mutant males and N2 males, and the F2 generations of these crosses were scored.  II. 2. Calculating recombination frequencies  Recombination frequency was measured by scoring the number of recombinant progeny of animals that were c«-heterozygous for the visible markers used, under the conditions  21  described by Rose and Baillie (1979a). The recombination frequency (p) between two markers was calculated according to the method described by Zetka and Rose (1995), as p = 1 - (1 - 2R) , where R is calculated as the total number of recombinant progeny scored U2  divided by the total progeny. In cases where one mutant class exhibited lower viability than the other or was indistinguishable from the double mutant, R was calculated as 2 x (the number of recombinants of the more viable or visible mutant class). To correct for possible low viability of the double mutant progeny, total progeny number was calculated as 4/3 x (wild-type progeny + one recombinant class).  95% confidence intervals were calculated as described in Zetka and Rose (1995), using the statistics of Crow and Gardner (1959) where fewer than 300 recombinants were scored. Where more than 300 recombinants were scored, confidence intervals were calculated from the expected maximum and minimum numbers of recombinants, using n±\.96(nxy) , m  when  n is the number of recombinants, x is the number of recombinants divided by the number of recombinant plus wild-type progeny, andy is 1 -x. Both classes of recombinants were used except in the cases of the unc-101 to unc-54, dpy-5 to ccls425l, and ccls4251 to unc-75 intervals.  II. 3. ccls4251  ccls4251 is a large tandem-array transgene, expressing multiple copies of myo-3::GFP in somatic nuclei and mitochondria. It is integrated into LGI (Fire et al., 1998). The expression of ccls4251 is suppressed (although some GFP expression is still visible) by a homozygous mutation in tam-1 (Hsieh et al., 1999). As GFP is expressed dominantly, ccls4251 provided a means of following the Tam-1 phenotype while allowing scoring of visible mutants on the  22  same chromosome (LGI). To take into account possible effects of ccls4251 itself on crossing over, crosses were also carried out in animals heterozygous for ccls4251 using visible markers on the same and on a different chromosome.  ccls4251 males undergo successful mating at low rates; to ensure that successful mating could occur, N2 males were first mated to ccls4251 hermaphrodites. ccls4251/+ males from the progeny were then crossed to double mutants, GFP hermaphrodites from the Fl were picked individually at L4, and their progeny scored. Where crossing over in tam-l animals was scored, N2 males were mated to ccls4251/ccls4251 (I); tam-l/tam-1 (V) hermaphrodites, and GFP hermaphrodite progeny were picked at L4. L4 wild-type hermaphrodites of the F2 generation were then examined under UV light with the aid of a Zeiss Stemi SV11 microscope; those expressing GFP at a low intensity were picked individually, while those expressing GFP at wild-type levels were picked as controls. The progeny of animals giving Dpy Uncs were scored.  II. 4. rec-1 mapping using dpy-5, dpy-13, ccls4251, lin-11 and unc-75  ccls4251/+ males were crossed to dpy-5 rec-1 unc-75 hermaphrodites. One Dpy GFP recombinant and one GFP Unc recombinant were picked from the F2 generation and were allowed to self-fertilize until homozygous dpy-5 ccls4251 and ccls4251 unc-75 strains were produced. N2 males were crossed to dpy-5 rec-1 unc-75 hermaphrodites to produce dpy-5 rec-1 unc-75/+ + + males that were crossed to the ccls4251-bearing recombinant strains. Four Dpy Unc mutants expressing GFP were then picked from the F2 generation and allowed to self-fertilize to produce homozygous Dpy Unc GFP strains. These strains were tested for the presence of the rec-1 mutation by crossing to rec-1 males and scoring the  23  progeny. Two strains were found to bear mutant rec-1, while the other two were found to bear wild-type rec-1. To map the relative positions of ccls4251 and rec-1, lin-11 males were mated to a dpy-5 rec-1 ccls4251 unc-75 strain, and the F2 generation was scored for Dpy and Une animals. Recombinants were scored for mutant lin-11. GFP Dpy and GFP Unc recombinants were also tested for the presence of the rec-1 mutation by crossing to rec-1 males and scoring the distance between the ccls4251 insertion and dpy-5 or unc-75. This was done by picking all Dpys or Uncs from plates and scoring the frequency of these animals giving some non-GFP progeny.  In a separate experiment, ccls4251/+ males were crossed to dpy-5 unc-13 rec-1 animals, and four Dpy Uncs expressing GFP were isolated from the F2 generation. Four homozygous Dpy Unc GFP strains were obtained from these animals. These strains were tested for the presence of mutant rec-1.  This mapping work was carried out in part by a work study student, Petr Pospisil.  II. 5. Sequencing DNA from Rec-1 animals was obtained by lysing individual adult rec-1 hermaphrodites for one hour at 56°C in worm lysis buffer (lx New England Biolabs PCR buffer, 60ng/ul Proteinase K, 1.5mM MgCh), then for 15 minutes at 15°C. klp-16 was then amplified using PCR (PCR mix: lx New England Biolabs PCR buffer minus MgCl , 0.25 mM dNTP, 1.6 2  mM MgCb, 0.5U Taq polymerase, 0.6mM each of primers, 5u.l template DNA, dE^O to 25 |il).  24  The following program was used: 1. Incubation  30 s at  96°C  2. Melting  45 s at  94°C  3. Annealing  45 s at  56°C  4. Elongation  180 s at  72°C  5. Final elongation  300 s at  72°C  6. Final hold  4°C  (steps 2. - 4.:35 cycles)  Table 1: Primers used to amplify klp-16  Primer name Primer sequence NON101  5' atg aat gtc get cgt aga ag 3'  NON102  5' cag tga ccc aat caa cag c 3'  NON103  5' gec ata ttc tga aag tta gaa gac 3'  NON104  5' caa tea ctg cag gaa ttc ttc 3'  ML1  5' cca gtc tea cga tat cca tg 3'  ML2  5' gat act caa teg cag aac tc 3'  ML3  5' gtt gca caa cga tgt cgt c 3'  ML4  5' cat etc ctg aac tgg teg 3'  ML5  5' age aca tea gec ate aac tg 3'  PCR products were tested by running 5 pi on a 1% agarose gel stained with ethidium bromide. PCR products were purified using a QIAGEN QIAquick® PCR purification  according to the protocol provided with the kit, and sent for sequencing at the U B C Nucleic Acid Protein Service (NAPS) Unit.  II. 6. S c o r i n g males  In C. elegans, the percentage of male progeny is a useful tool for measuring the frequency of X chromosome loss or nondisjunction. The percentage of males rises with temperature and age (Rose and Baillie, 1979a). Although males were recorded regardless of phenotype, it might have been difficult to identify males among some of the Dpy or Dpy Unc progeny. For this reason, the frequency of males was calculated by using only the number of wildtype male progeny, divided by the total number of wild-type progeny. This value was multiplied by 100 to obtain the percentage of males.  26  III. RESULTS  III. 1. Crossing over in a him-17 mutant  At 20°C, the standard temperature for scoring experiments (Rose and Baillie, 1979a), the low viability of him-17 null mutants (Reddy and Villeneuve, 2004) makes it hard to obtain the numbers necessary for scoring data. Crossing over was therefore measured in a him-17 e2707 non-null mutant. Two well-characterized intervals were chosen on LGI: one in the recombinationally-suppressed cluster, and one in a highly recombinogenic region on the arm. Two further intervals were chosen on a different chromosome, LGIII. One of these intervals is in a highly recombinogenic region on the arm; the other is a large, hybrid region that covers part of the cluster and part of a flanking region.  III. 1. 1. Mutant him-17 increased crossing over within the central cluster on LGI  Crossing over in the dpy-5 unc-13 interval in a him-17 e27'07 background produced a map distance of 7.7 m. u. (95% C.I. 5.6 - 10.3 m. u.), compared to 2.4 m. u. (95% C.I. 1.9-3.1) in the wild-type N2 strain. The genetic distance between dpy-5 and unc-13 in a mutant rec-1 background was 6.8 m. u. (95% C.I. 5.2 - 8.6 m. u.). The more than threefold increase in a him-17 e27'07 background is of a similar magnitude to that found in a mutant rec-1 background. Thus, like mutant rec-1, mutant him-17 increased crossing over across a recombinationally-suppressed region (Table 2).  27  Table 2: Crossing over in a him-17 strain in a cluster region  Genotype  Wild-type  Recombinants  progeny '  Dpy  1 2  dpy-5 unc-13 rec-17+ + + dpy-5 unc-13 rec-17++ rec-1  1967 (0) 713 (0)  dpy-5 unc-13 rec-17++ +; him-17/him-17 427 (110)  2  3  1  Unc'  /?xl00 Total  95% C . I /  1  31  33  64  2.4  34  30  64  6.8  5.2-8.6  25  18  43  7.7  5 . 6 - 10.3  Male progeny included Total wild-type progeny are shown, with the number of male progeny given in parentheses 95% Confidence Interval, see Materials and Methods  to  00  1.9-3.1  III. 1. 2. Mutant him-17 decreased crossing over within a flanking region on LGI  unc-101 mutants display a coiled body phenotype, while unc-54 mutants are limp and nearly completely paralysed. Because of this, unc-54 mutants could not be distinguished from the double mutants; for this reason, only unc-101 animals were scored.  In a him-17 background, the map distance was 5.1 m. u. (95% C. I. 3.1 - 7.7 m. u.). The map distance of this interval in the wild-type background was 9.9 m. u. (95% C.I. 8.0 - 11.9 m. u.). Thus a loss of function of him-17 produces a twofold decrease in recombination over this region compared to wild-type (Table 3).  III. 1. 3. Mutant him-17 decreased crossing over within a flanking region on LGIII  In order to test whether the effect of him-17 on recombination is general, crossing over was measured on LGIII. The dpy-18 unc-64 interval, on the right chromosome arm, was chosen. In the wild-type background, the distance between these two markers was 14.4 m. u. (95% C. I. 14.0 - 14.7 m. u.). In him-17 animals, the map distance between dpy-18 and unc-64 was 7.9 m. u. (95% C. I. 6.0 - 10.3). This demonstrates that the effects of mutant him-17 on recombination are not limited to LGI. In rec-1 animals, this region underwent a comparable decrease, to 9.1 m. u. (95% C. I. 8.7 - 9.4) (Table 4).  29  Table 3: Crossing over in a him-17 strain in a flanking region on LGI  Wild-type progeny '  Genotype  pxlOO 95%C.I/  Recombinants Unc-101 Unc-54' 1  1728(1) 618(169) unc-101 unc-54/+ +;him-17/him-17  116 21  unc-101 unc-5 4/+ +  N/S* N/S  9.9 5.1  4  8.0-11.9 3.1-7.7  'Male progeny included Total wild-type progeny are shown, with the number of male progeny given in parentheses 95% Confidence Interval, see Materials and Methods N/S: Not Scored. Only unc-101 recombinants were scored, as unc-54 animals could not be distinguished from unc-101 unc-54 animals  2  3 4  Table 4: Crossing over in a him-17 strain in a flanking region on LGIII  Genotype  1  >sRecombinants  Wild-type progeny '  pxlOO 95% C.I/  Dpy Unc' Total 1  1  dpy-18 unc-64/+ +  6526 (13)  458  856  dpy-18unc-64/+ +; rec-1/rec-1  3190 (15)  105  297  dpy-18 unc-64/+ +;him-17/him-17 537 (176) 1 2 3  24  34  1314 402 58  14.4 9.1  14.0- 14.7 8.7-9.4  7.9  Male progeny included Total wild-type progeny are shown, with the number of male progeny given in parentheses 95% Confidence Interval, see Materials and Methods  o  6.0- 10.3  III. 1. 4. Mutant him-17 increased crossing over in a hybrid region on LGIII  Mutant rec-1 and mutant him-17 were further compared to assess whether or not these mutants affect crossing over in identical ways across different intervals. Crossing over was measured in the unc-32 dpy-18 interval on LGIII, a hybrid region that covers part of the meiotic cluster and part of the flanking region. In rec-1 animals, this region was 6.1 m. u. (95% C. I. 5.9 - 6.3), a decrease from the wild-type distance of 7.4 m. u. (95% C. I. 7.2 7.6). In contrast, an increase in the frequency of crossing over was noted in a him-17 background (map distance 9.9 m. u., 95% C. I. 8.4 - 11.5) compared to wild-type (Table 5).  III. 1. 5. Crossing over in the cluster of LGI in a  rec-l;him-17  double mutant was not  significantly different from that in each single mutant  Crossing over was measured in a rec-1 ;him-l7 double mutant, to determine the nature of the interaction (or its absence) between rec-1 and him-17. Crossing over was measured in an interval in the cluster of LGI that had previously been measured in each single mutant. In the double mutant, this region was 8.5 m. u. (95% C. I. 6.0 - 11.3) - not significantly different from the distance in either the rec-1 mutant (6.9 m. u., 95% C. I. 5.9 - 7.9) or the him-17  mutant (7.7 m. u., 95% C. I. 5.6 - 10.3) (Table 6). The progeny of two animals was  excluded from the analysis, as the very low brood size in these animals (12 and 15) yielded no recombinants in one case, and a single Dpy with no Dpy Uncs in the other did not give a clear indication as to whether the genotype of the parent might have been dpy-5 rec-1/+ rec-1 (I.); him-17/him-17(V).  Excluding these results does not significantly change the  overall frequency of crossing over in the double mutant. Taking into account these low  31  Table 5: Crossing over in a him-17 strain in a hybrid region Genotype  Wild-type progeny '  Unc  unc-32 dpy-18/+ + unc-32 dpy-18/++; rec-l/rec-1  6067(4) 4842 (26)  pxlOO  Recombinants Dpy  1  Total  1  1  302 306 608 188 214 402  unc-32 dpy-18/+ +; him-17/him-17 1370 (394)  87  98  95% C . I /  7.4 6.1  185  7.2-7.6 5.9-6.3  9.9  8.4- 11.5  Male progeny included Total wild-type progeny are shown, with the number of male progeny given in parentheses 95% Confidence Interval, see Materials and Methods  2 3  Table 6: Crossing over in a rec-1;him-17 strain in a cluster region Genotype  pxlOO 95% C.I/  Recombinants  Wild-type  1,2  Dpy Unc Total' 1  1  dpy-5 unc-13 rec-17+ + +  1967 (0)  31  33  64  2.4  1.9-3.1  dpy-5 unc-13 rec-l/+ + rec-1  2301(9)  113  101  214  6.9  5.9-7.9  dpy-5 unc-13 rec-l/+++;him-17/him-17  427 (110)  dpy-5 unci3 rec-l/+ + rec-1;him-17/him-17 391(116) 1 2 3  25  18 28  43 17  7.7 45  5.6- 10.3 8.5  Male progeny included Total wild-type progeny are shown, with the number of male progeny given in parentheses 95% Confidence Interval, see Materials and Methods  1>J  6.0- 11.3  brood sizes, average total brood size in the double mutant (52.5) did not differ from that in the him-17 mutant alone (50.1).  III. 1. 6. Crossing over in a flanking region on LGIII in a rec-l;him-17 double mutant was not significantly different from that in each single mutant  To confirm that the rec-1 ;him-l 7 double mutant randomizes the meiotic pattern, rather than producing a general increase in the frequency of crossing over, crossing over was measured in the dpy-18 unc-64 interval on the arm of LGIII. In the double mutant, this region was 6.7 m. u. (95% C. I. 4.7 - 9.4) - a decrease compared with the wild-type distance (14.4 m. u., 95% C. I. 14.0 - 14.7), but not significantly different from the distance in either the rec-1 mutant (9.1 m. u., 95% C. I. 8.7 - 9.4) or the him-17 mutant (7.9 m. u., 95% C. I. 6.0 - 10.3) (Table 7).  III. 1. 7. The effects of rec-1 and him-17 on crossing over in a hybrid region were additive  To further examine the relationship between rec-1 and him-17, the frequency of crossing over was measured between unc-32 and dpy-18. Crossing over in this interval is suppressed compared to wild-type in the rec-1 mutant, and increased in the him-17 mutant. The map distance in wild-type animals was 7.4 m. u. (95% C. I. 7.2 - 7.6), and the map distance in the double mutant was 8.0 m. u. (95% C. I. 6.6 - 9.7), not significantly different from wildtype. The distance in the double mutant was higher than that in rec-1 animals (6.1 m. u., 95% C. I. 5.9 - 6.3), and lower than the distance in him-17 animals (map distance 9.9 m. u., 95% C. I. 8.4-11.5) (Table 8).  33  Table 7: Crossing over in a rec-l;him-l 7 strain in a flanking region Genotype  Wild-type progeny'  12  pxlOO 95%C.I.  Recombinants  J  Dpy Unc Total 1  1  1  dpy-18 unc-647+ +  6526(13)  458 856 1314 14.4  14.0- 14.7  dpy-18 unc-64/+ +;rec-l/rec-1  3190 (15)  105  dpy-18 unc-64/+ +;him-17/him-17  537(176)  dpy-18 unc-64/+ +;rec-l/rec-l;him-17/him-17 375 (102)  1 2 3  24  297  402  9.1  8.7-9.4  34  58  7.9  6.0-10.3  17  17  34  Male progeny included Total wild-type progeny are shown, with the number of male progeny given in parentheses 95% Confidence Interval, see Materials and Methods  4^  6.7  4.7-9.4  Table 8: Crossing over in a rec-1,-him-17 strain in a hybrid region Genotype  Wild-type progeny '  pxlOO 95% C.I/  Recombinants Dpy Unc Total 1  unc-32 dpy-18/+ + unc-32 dpy-18/+ +;rec-l/rec-l unc-32 dpy-187++;him-17/him-17  6067 (4) 4842(26) 1370(394)  unc-32 dpy-18/+ +;rec-l/rec-l;him-17/him-17 991 (316)  1 2 3  1  1  302 306 608 7.4 188 214 402 6.1 87 98 185 9.9 50  58  108  8.0  Male progeny included Total wild-type progeny are shown, with the number of male progeny given in parentheses 95% Confidence Interval, see Materials and Methods  <J1  7.2 - 7.6 5.9-6.3 8.4- 11.5 6.6-9.7  III.  1. 8. Summary: mutant  him-17 eliminates the meiotic pattern  From these data, I conclude that mutant him-17 alters the meiotic pattern by increasing crossing over within regions that are normally recombinationally suppressed, and decreasing crossing over within regions that are normally highly recombinogenic regions. This effect produces a genetic map in him-17 mutants that is similar to the physical map (see Introduction; Fig. 3, Fig. 4). Its effects appear to be general, affecting autosomes in trans. A rec-1 ;him-l 7 double mutant exhibits crossing over in a cluster region and in a highly recombinogenic flanking region that is not significantly different from that found in either single mutant (Fig. 5, Fig. 6). However, the frequency of crossing over in a hybrid region is not significantly different from wild-type, differing from the frequency of crossing over found in both single mutants (Fig. 6).  36  dpy-5 Genetic map in N2 hermaphrodite  unc-54  unc-101  unc-11 unc-13  bli-3  |1  \ bli-3  unc-11 dpy-5 unc-13  unc-101  unc-54  Physical map  G e n e t i c m a p bli-3  unc-11  dpy-5  unc-13  unc-101 unc-54  in rec-1 hermaphrodite  dpy-5  Genetic map  unc-13  unc-101  unc-54  in him-17 hermaphrodite Genetic maps: Physical map:  5 m a p units 1 Mb  Figure 3: Mutant him-17 eliminates the meiotic pattern on LGI. Figure based on Zetka and Rose, 1995; N2 and rec-1 genetic map data taken from Zetka and Rose, 1995, physical map data taken from WormBase, him-17 data presented in this thesis. Dashed or dotted lines are used to contrast the relative positions of each gene along each genetic map with their position along the physical map. Large dashes on the him-17 map indicate regions for which no data has been gathered.  37  unc-32 JJ  Genetic map inN2 hermaphrodite  Physical map  *** unc-32 -|  Geneticmap in rec-1 hermaphrodite Genetic map in him-17 hermaphrodite Genetic maps: Physical map:  \ ;  dpy-18 | ;  unc-64 |_  ! dpy-18 -J  unc-32 |  unc-32 I  dpv-18 | dpy-18 _|  unc-64 1-  unc-64 ) unc-64 j  5 map units 1 Mb  Figure 4: Mutant him-17 eliminates the meiotic pattern on LGIII. Figure based on Zetka and Rose, 1995; physical map data taken from WormBase, genetic map data presented in this thesis. Dashed or dotted lines are used to contrast the relative positions of each gene along each genetic map with their position along the physical map.  38  Genetic map in N2 hermaphrodite  dpy-5 unc-11 unc-13  bli-3  |i  \ bli-3  unc-54  unc-101  unc-11 dpy-5 unc-13  unc-101  unc-54  Physical map  Genetic map bli-3 in rec-1 hermaphrodite  unc-11  Genetic map  dpy-5  unc-13  dpy-5  unc-13  dpy-5  unc-13  unc-101 unc-54  unc-101  unc-54  in him-17  hermaphrodite  Genetic map  in rec-1,him-17 double mutant hermaphrodite  Genetic maps: Physical map:  •5 map units  1 Mb  Figure 5: Double mutant rec-l;him-17 has the same phenotype as rec-1 and him-17 i the cluster on LGI. Figure based on Zetka and Rose, 1995; N2 and rec-1 genetic map data taken from Zetka and Rose, 1995, physical map data taken from WormBase, him-17 and rec1;him-17 data presented in this thesis. Dashed or dotted lines are used to contrast the relative positions of each gene along each genetic map with their position along the physical map.  39  Genetic map in N2 hermaphrodite  unc-32  unc-64  dpy-18  dpy-18  unc-32  unc-64  Physical map  unc-32  Genetic map in rec-1 hermaphrodite Genetic map  unc-32  dpy-18  dpy-18  unc-64  unc-64  in him-17  hermaphrodite Genetic map  unc-32  dpy-18  unc-64  in rec-1;him-17  hermaphrodite Genetic maps: Physical map:  5 map units  1 Mb  Figure 6: The effects of rec-1 and him-17 on the meiotic pattern are additive. Figure  based on Zetka and Rose, 1995; physical map data taken from WormBase, genetic map data presented in this thesis. Dashed or dotted lines are used to contrast the relative positions of each gene along each genetic map with their position along the physical map.  40  III. 2. Crossing over in a lin-35 mutant  Crossing over was measured in a lin-35 null mutant. To overcome difficulties arising from the lack of a convenient way of following mutant lin-35, and the use of markers closely linked to it, crossing over was measured on a different chromosome, LGV. Recombination was measured over an interval in the cluster for which mutant rec-1 data was readily available for comparison. Crossing over across a different interval on the same chromosome was chosen to determine whether mutant lin-35 produces an overall increase in recombination; this interval covers both cluster and flanking regions. In the absence of suitable flanking intervals on LGV, crossing over was measured across a flanking region on the arm of LGIII which had already been measured in rec-1 and him-17 mutants.  III. 2. 1. Mutant lin-35 increased crossing over within the central cluster of LGV  To look for an effect of mutant lin-35 on recombination, the dpy-11 unc-42 interval, a wellcharacterized interval in the cluster on LGV, was chosen. In a lin-35 background, the genetic distance increases to 5.2 m. u. (95% C. I. 4.0 - 6.6 m. u.), a more than twofold increase from that in a wild-type background, which was 2.0 m. u. (95% C. I. 1.4 - 2.7 m. u.). This increase is likely to be an underestimate, as Dpy or Unc animals that died before producing offspring would not have been scored as recombinants (see Materials and Methods). Sterility and/or premature death were more prevalent in animals originally identified as Unc, explaining the disparity between Dpy and Unc animals shown in the table (Table 9). As rec1 and lin-35 mutants have similar phenotypes, and as rec-1 maps close to lin-35 (see III.7), a  complementation test was carried out between rec-1 and lin-35. lin-35 males were mated to  41  Table 9: Crossing over in a lin-35 strain in a cluster region  Genotype  Wild-type progeny ' Recombinants  pxlOO 95%C.I.  J  Dpy Unc Total' 1  dpy-11 unc-42/+ +  1523 (1)  dpy-11 unc-42/+ +; lin-35/lin-35 956 (0)  1  21 55  1  19 14  40  2.0  1.4-2.7  69  5.2  4.0-6.6  Male progeny included Total wild-type progeny are shown, with the number of male progeny given in parentheses 95% Confidence Interval, see Materials and Methods  to  dpy-5 unc-13 rec-1 hermaphrodites, and crossing over was scored in the dpy-5 unc-13 interval, lin-35 was found to complement rec-1 (data not shown). Although the lin-35 null mutation gives rise to a protein product lacking both binding pocket domains (Lu and Horvitz, 1998), it was nevertheless conceivable that complementation between different protein domains might be occurring. However, later rec-1 mapping data that puts rec-1 to the right of unc-13 instead of to the left (see III. 6.) eliminated this possibility.  III. 2. 2. Mutant lin-35 decreased crossing over within a hybrid region on LGV  To test whether mutant lin-35 produces an overall increase in the frequency of crossing over, recombinants were scored over the unc-60 dpy-11 interval. This interval is a large hybrid region that encompasses mainly flanking region, but also covers part of the central cluster. Crossing over in a rec-1 background was also measured. Crossing over in the lin-35 (map distance 11.5 m. u., 95% C. I. 9.7- 13.4), but nottherec-7 (map distance 13.6 m. u., 95% C. I. 11.5 - 15.7), was significantly decreased from wild-type (map distance 15.0 m. u., 95% C. I. 14.5-15.6) (Table 10).  III. 2. 3. Mutant lin-35 decreased crossing over within a flanking region on LGIII  To test whether mutant lin-35, like mutant him-17 and mutant rec-1, decreases the frequency of crossing over in flanking regions, crossing over was measured over the dpy-18 unc-64 interval on LGIII. Crossing over in this region was also measured in the rec-1 background. The frequency of crossing over was significantly lower in both the rec-1 (9.1 m. u., 95% C. I. 8.7 - 9.4) and lin-35 (10.5 m. u., 95% C. I. 7.9 - 13.6) backgrounds than in the wild-type background (14.4 m. u., 95% C. I. 14.0 - 14.7) (Table 11).  43  Table 10: Crossing over in a lin-35 strain in a hybrid region  Genotype  Wild-type progeny ' Recombinants  pxlOO 95%C.I/  Unc Dpy Total 1  1  1  unc-60 dpy-ll/++  2347 (0)  257  223  480  15.0  14.5- 15.6  unc-60 dpy-11/++; rec-1/rec-1  1113 (0)  99  106  205  13.6  11.5- 15.7  unc-60 dpy-ll/+ +; lin-35/lin-35 1057 (0) 1 2 3  85  80  165  11.5  9.7- 13.4  Male progeny included Total wild-type progeny are shown, with the number of male progeny given in parentheses 95% Confidence Interval, see Materials and Methods  Table 11: Crossing over in a lin-35 strain in a flanking region  Genotype  Wild-type progeny ' Recombinants  />xl00 95% C.I.  3  Dpy Unc Total 1  dpy-18 unc-64/+ +  6526(13)  458  1  856  1  1314  14.4  14.0-14.7  dpy-18 unc-64/+ +; rec-l/rec-1 3190 (15)  105  297 402  9.1  8.7-9.4  dpy-18 unc-64/+ +; lin-35/lin-35 431 (0)  22  41  10.5  7.9-13.6  2 3  63  Male progeny included Total wild-type progeny are shown, with the number of male progeny given in parentheses 95% Confidence Interval, see Materials and Methods  III. 2. 4 Summary: mutant lin-35 eliminates the meiotic pattern  In lin-35 null mutants, the frequency of crossing over increases within the meiotic cluster, and decreases within flanking regions on the arm. Thus, it appears that the meiotic pattern is eliminated in lin-35 mutants (Fig. 7, Fig. 8). Despite its proximity to the mapped location of rec-1 (Zetka and Rose, 1995; M. Tarailo, personal communication), lin-35 complemented rec-1.  III. 3. Crossing over in the presence of a heterologous insertion  In order to test for the suitability of the multicopy myo-3::GFP insertion ccls4251 on LGI as a visible marker when measuring crossover frequencies, crossover frequency was scored over one recombinationally suppressed interval (dpy-5 unc-13) and one highly recombinogenic interval (unc-54 unc-101) on LGI with cch4251 in trans. Crossing over was then measured in tam-1 mutants in the presence of the heterologous insertion.  III. 3. 1. 7>fl«.s-heterozygous ccls4251 decreased crossing over in the cluster and on the arm on the same chromosome  Both the cluster and a flanking region on LGI were found to be recombinationally suppressed in ccls4251 heterozygotes compared to N2. The map distance between dpy-5 and unc-13 was 0.5 m. u. (95% C. I. 0.3 - 0.7) when ccls4251/+ was present heterozygously,  and 1.1 m. u. (95% C. I. 0.8 - 1.6) in wild-type. The unc-101 unc-54 interval was 10.7 m. u. (95% C. I. 8.0 - 13.9) in the ccls4251/+ heterozygote, compared with 19.2 m. u. (95% C. I. 16.1 -22.45) in wild-type (Table 12).  45  Genetic map in N2 hermaphrodite  unc-60  dpy-11 unc-42  4  1 1—  • unc-60 _|  Physical map  Genetic map in lin-35 hermaphrodite  unc-60 _j  Genetic maps: Physical map:  dpy-11 1  dpy-11 1  unc-42 1  unc-42 1  5 map units 1 Mb  Figure 7: Mutant lin-35 increases crossing over in the cluster of LGV. Figure based on Zetka and Rose, 1995; physical map data taken from WormBase. Dashed or dotted lines are used to contrast the relative positions of each gene along each genetic map with their position along the physical map.  46  unc-32  Genetic map in A/2 hermaphrodite  dpy-18  unc-64  dpy-18  unc-32  unc-64  Physical map  Genetic map in rec-1 hermaphrodite  Genetic map in him-17 hermaphrodite  unc-32  dpy-18  '.  •  unc-32 I '  dpy-18 _J !  Genetic map  dpy-18  in lin-35  1  unc-'64 •  J' • •  unc-64 '. I :  unc-64 L  hermaphrodite  Genetic maps: Physical map:  5 m.u. 1 Mb  Figure 8: Mutant lin-35 decreases crossing over in a region on the arm of LGIII. Figure based on Zetka and Rose, 1995; physical map data taken from WormBase. Dashed or dotted lines are used to contrast the relative positions of each gene along each genetic map with their position along the physical map.  47  Table 12: The local effects of frans-heterozygous ccls4251 on crossing over Genotype  Wild-type progeny  Recombinants  1  pxlOO 95% C l /  Left Right marker Total marker 1  1  1  dpy-5 unc-13 rec-1/ + + +  2115  dpy-5 unc-13 rec-1/ + + + ccls4251 4725 unc-101 unc-54/+ + unc-101 unc-54/+ + ccls4251 1 2 3  1409 787  18  9 184 57  14  32  20 N/S N/S 3  3  29  1.1  0.8- 1.6  0.5 0.3 - 0.7 19.2 16.1-22.45 10.7 8.0-13.9  Male progeny included 95% Confidence Interval, see Materials and Methods N/S: Not Scored. Only unc-101 recombinants were scored, as unc-54 animals could not be distinguished from unc-101 unc-54 animals  oo  III. 3. 2.  The effects of ccls4251 on crossing over were localized  To test whether the effects of ccls4251 on the meiotic pattern were restricted to LGI, crossover frequency was measured over a cluster interval on LGV (dpy-11 unc-42). This interval did not undergo significantly different rates of crossing over in ccls4251/+ (map distance 2.9 m. u., 95% C. I. 2.0 - 3.9) and N2 (map distance 2.6 m. u., 95% C. I. 2.0 - 3.3) backgrounds (Table 13).  III. 3. 3. Homozygous mutant tam-1 reversed the local recombinational suppression of heterozygous ccls4251 over a flanking region  The effects of mutant tam-1 on crossing over in a flanking region were measured. Crossing over in the well-characterized unc-101 unc-54 interval was scored in animals transheterozygous for ccls4251, using ccls4251 /rara'-heterozygotes not homozygous for tam-1 as controls. This interval was found to undergo higher rates of crossing over in ccls4251/+;tam-1 animals (map distance 16.2 m. u., 95% C. I. 13.1 - 19.5) compared with  heterozygous ccls4251 alone (map distance 10.7 m. u., 95% C. I. 8.0 - 13.9); indeed, the map distance in a ccls4251/+;tam-1 background was restored to wild-type levels (19.2 m. u., 95% C. I. 16.1 -22.45) (Table 14). III. 3. 4. Summary: a multicopy transgene decreases the frequency of crossing over on the same chromosome  7ra«.s-heterozygous ccls4251 was found to suppress crossing over along the same chromosome, both in the cluster and in a flanking region. In tam-1 mutants, this effect was eliminated for an interval on the arm. 49  Table 13: The effects of heterozygous ccls4251 on a different chromosome Genotype  Wild-type progeny Recombinants  pxlOO 95% CI/  1  Dpy Unc' Total' 1  dpy-11 unc-42/+ +  1941  dpy-11 unc-42/+ +; ccls4251/+ 979  35  32  67  19  19  38  2.6 2.9  2.0-3.3 2.0-3.9  Male progeny included 95% Confidence Interval, see Materials and Methods  Table 14: The effects of mutant tam-1 on crossing over when heterozygous ccls4251 is present Genotype  Wild-type progeny' Recombinants  pxlOO 95% C.I/  Unc-101 Unc-54' 1  unc-101 unc-54/+ +  1409  184  N/S  19.2  16.1-22.45  unc-101 unc-5 4/+ + ccls4251  787  57  N/S  10.7  8.0-13.9  122  N/S  16.2  13.1-19.5  unc-101 unc-54/+ + ccls4251; tam-l/tam-11105 1 2 3  3  3  3  Male progeny included 95% Confidence Interval, see Materials and Methods N/S: Not Scored. Only unc-101 recombinants were scored, as unc-45 animals could not be distinguished from unc-101 unc-54 animals  o  III.  4. Mutant him-17 does not suppress ccls4251 expression  M u t a t i o n s in  tam-1 a n d lin-35  h a v e b o t h b e e n s h o w n to s u p p r e s s the e x p r e s s i o n o f a t a n d e m  a r r a y t r a n s g e n e ( H s i e h et a l . , 1999). T o test f o r a s i m i l a r e f f e c t o f m u t a n t  transgene expression,  him-17 e2707  him-17 o n  m a l e s w e r e c r o s s e d to a n i m a l s e x p r e s s i n g the  t r a n s g e n e . L 4 h e r m a p h r o d i t e s f r o m the F 2 g e n e r a t i o n w e r e p i c k e d i n d i v i d u a l l y , a n d  ccls4251 him  a n i m a l s w e r e e x a m i n e d u n d e r u l t r a v i o l e t light f o r the i n t e n s i t y o f G F P e x p r e s s i o n , u s i n g a  Zeiss Stemi S V 1 1 microscope,  l e v e l s than  him-17 a n i m a l s  non-him-17 a n i m a l s  ( F i g . 9).  I ccls4251  ccls4251;tam-1  A J M H H H ^ M ^ H  w e r e not f o u n d to e x p r e s s G F P at l o w e r  ccls4251;him-17  nMm^m^m^m^m^M  c  ^  ^  ^  ^  ^  H  H  ^  H  Figure 9: Mutant him-17 does not suppress the expression of a multicopy tandem-array transgene. m\o-3::GFP expression viewed under ultraviolet light in (A) ccls4251;tam-1 (B) ccls4251 and (C) ccls4251;him-17 animals. ccls4251 control animals are placed in the centre figure for easier comparison with the two mutants. A marked decrease in GFP expression was observed in ccls4251;tam-1, but not in ccls4251;him-17, animals.  III.  5.  Mutant rec-1 does not suppress ccls4251 expression  T o test f o r t a n d e m a r r a y t r a n s g e n e e x p r e s s i o n s u p p r e s s i o n i n  l(sI80) ccls4251 unc-75(e950)  and  rec-1 m u t a n t s , dpy-5(e61) rec-  dpy-5(e6l) ccls4251 unc-75(e950)  were constructed and  51  examined by eye under ultraviolet light, using a Zeiss Stemi SV11 microscope, rec-1 animals were not found to express GFP at lower levels than animals bearing wild-type rec-1 (Fig. 10).  dpy-S unc-13 ccls42S1  I dpy-5 unc-13 rec-1 ccls4251  BJ Figure 10: Mutant rec-1 does not suppress the expression of a multicopy tandem-array transgene. m\o-3::GFP  expression viewed under ultraviolet light in (A) dpx-5 unc-13  ccls425l and (B) dpv-5 unc-13 rec-1 ccls4251 animals, dpv-5 unc-13 animals were used to test for the presence of rec-1. No decrease in GFP expression was observed in rec-1 animals compared to controls.  III. 6. Mapping rec-1  To narrow down the position of rec-1 with a view to facilitating strain construction, dpy-5, ccls425l, lin-11 and unc-75 were used to left-right map rec-1.  III. 6. 1. Gross mapping of rec-1 using LGI markers and a transgene  The ccls425l insertion was found to map to the left of lin-11, 1.3 m. u. (95% C. I. 0.8 - 1.9) from dpy-5, and 1.2 m. u. (95% C. I. 0.8 - 1.8) to lin-11 (Table 15).  52  Map distances in the four Dpy ccls4251 recombinant strains between dpy-5 and ccls4251 were 3.8 m. u. (95% C. I. 2.2 - 6.0), 5.9 m. u. (95% C. I. 3.8 - 8.4), 7.6 m. u. (95% C. I. 5.4 - 10.1), and 8.3 m. u. (95% C. I. 6.4 - 10.7), all significantly higher than the distance in wild-type control (1.3 m. u., 95% C. I. 0.6 - 2.3), consistent with the presence of mutant rec1 in all four recombinant strains (Table 16).  Map distances in the four Unc ccls4251 recombinant strains between unc-75 and the ccls4251 insertion were 3.0 m. u. (95% C. I. 1.8 - 4.8), 3.2 m. u. (95% C. I. 1.8 - 5.0), 3.5  m. u. (95% C. I. 1.8 - 5.9) and 5.0 m. u. (95% C. 1. 3.4 - 7.2). These are not significantly different from a control distance of 3.4 m. u. (95% C. I. 2.4 - 3.3). This is consistent with the absence of mutant rec-1 from three of the ccls4251 unc-75 strains, and its presence in the fourth (Table 17).  Taken together, the results given in Tables 16 and 17 put rec-1 to the left of the ccls4251 insertion. In one strain, the map distance was elevated compared to wild-type, but not significantly so, raising the possibility that this strain might in fact contain mutant rec-1. Although this precludes more accurate mapping of the position of rec-1 relative to ccls4251, it does not alter the conclusion that rec-1 maps to the left of ccls4251. Caution should in any case be exercised when mapping rec-1 more closely in a eels4251 -bearing strain, as the degree of local suppression of ccls4251 at different intervals along LGI has not been fully studied.  53  Table 15: ccls4251 mapping data Interval measured Wild-type Recombinants 1  pxlOO 95% C.I.  2  3  Left marker Right marker Total  2  dpy-5 to ccls4251  1210  11  11  22  1.3  0.8-1.9  ccls4251 tolin-U  1210  7  13  20  1.2  0.8- 1.8  lin-ll to unc-75  1210  15  36  36  2.2  1.5-3.0  Although all classes of recombinants were obtained from a single experiment, they are presented here separately in the interests of clarity Male progeny included 95% Confidence Interval, see Materials and Methods  2 3  Table 16: Left-right mapping of rec-1 in dpy-5 ccls4251 recombinants Genotype  Wild-type progeny Recombinants  pxlOO 95% C.I.  1  2  Dpy ccls4251/ccls4251 Total 1  dpy-5 ccls4251 + unc-75/ + + lin-ll + (control)1210  1  l  dpy-5 rec-1 ccls4251/+ rec-1 +  634  11 16  11 N/S  dpy-5 rec-1 ccls4251/+ rec-1 +  686  27  N/S  3  5.9  0.6-2.3 2.2-6.0 3.8-8.4  dpy-5 rec-1 ccls4251/+ rec-1 +  935  48  3  N/S  7.6  5.4-10.1  dpy-5 rec-1 ccls4251/+ rec-1 +  1083  61  N/S  8.3  6.4-10.7  3  3  22  1.3 3.8  Male progeny included 95% Confidence Interval, see Materials and Methods Only Dpy progeny were picked to test for homozygous ccls4251\ any ccls4251 non-Dpy progeny would not have been distinguished from dpy-  2  3  5 ccls4251/+ + animals  -fe.  Table 17: Left-right mapping of rec-1 in ccls4251 unc-75 recombinants Genotype  pxlOO 95% C.I/  Wild-type progeny' Recombinants ccls4251/ccls4251  Unc Total'  22  34  l  dpy-5 ccls4251 + unc-75/ + + lin-11 + (control) 1210 + ccls4251 unc-75/rec-1 + + + ccls4251 unc-75/rec-1 + + + ccls4251 unc-75/rec-l + + rec-1 ccls4251 unc-75/rec-1 + +  804 782 547 829  1  N/S  3  16  N/S  3  17  N/S  3  13  N/S  3  28  56  3.4 3.0 3.2 3.5 5.0  2.4-3.3 1.8-4.8 1.8-5.0 1.8-5.9 3.4-7.2  Male progeny included 95% Confidence Interval, see Materials and Methods Only Unc progeny were picked to test for homozygous ccls4251; any ccls4251 non-Unc progeny would not have been distinguished from  1 2  3  ccls4251 unc-75/+ + animals  Four strains were constructed by crossing ccls4251 onto LGI bearing dpy-5 and unc-13, to confirm the position of rec-1 to the left of the ccls4251 insertion, and to determine whether  rec-1 is to the left or right of unc-13. In two strains, the distance between dpy-5 and unc-13  (3.0 m. u., 95% C. I. 2.3 - 3.8; 3.0 m. u., 95% C. I. 2.4 - 3.8) underwent a fivefold increase in crossing over when compared with that in the other two strains (0.4 m. u., 95% C. I. 0.2 0.8; and 0.6, 95% C. I. 0.2 - 1.4) and in the negative control (0.5 m. u., 95% C. I. 0.2 - 1.1) (Table 18). This result is consistent with the presence of mutant rec-1 in these strains. That the map distance in these strains is not equal to that in a rec-1 strain lacking cch4251 can be attributed to the local suppression of crossing over in strains containing ccls4251 (see Table 12).  III.  6. 2.  Summary: the rec-1 gene maps between dpy-5 and the insertion site of  ccls4251  The ccls4251 transgene was found to map to the left of lin-11. It was subsequently used in gross mapping of rec-1, which was positioned to the left of the insertion site of ccls4251 and to the right of dpy-5. This position is consistent with that found by Tarailo (personal communication).  56  Table 18: Positioning rec-1 in dpy-5 unc-13 ccls4251 strains Genotype  Wild-type progeny Recombinants  p x 100 95% C.l.  1  z  Dpy Unc Total 1  dpy-5 unc-13 rec-l/+ + + ccls4251 (control) 865  1186 dpy-5 unc-13 + ccls4251/+ + rec-1 + 600 dpy-5 unc-13 rec-1 ccls4251/+ + rec-1 + 1568 dpy-5 unc-13 rec-1 ccls4251/+ + rec-1 + 1625 dpy-5 unc-13 + ccls4251/+ + rec-1 +  2  Male progeny included 95% Confidence Interval, see Materials and Methods  5 4 4 32 43  1  1 2 1 31 23  1  6 6 5 63 66  0.5 0.4 0.6 3.0 3.0  0.2- 1.1 0.2-0.8 0.2-1.4 2.3-3.8 2.4-3.8  III. 7. rec-1 is not klp-16  A paper by Page and Hawley (2005) had reported a Rec-1-like phenotype in Drosophila melanogaster. The mutation responsible was found to be in the kinesin-like protein gene,  Klp-3A. KLP has two recently-duplicated homologs in C. elegans, klp-15 and klp-16, the latter of which is on LGI. This raised the possibility that a mutation in klp-16 might be responsible for the Rec-1 phenotype; it was believed that subfunctionalization between klp15 and klp-16 might account for the non-lethality of the Rec-1 phenotype in C. elegans. klp-  16 was amplified from rec-1 animals and sent for sequencing (see Materials and Methods). The entirety of the coding region of klp-16, along with 400bp of the upstream region and 600bp of the downstream region, were sequenced, and found not to differ from wild-type.  III. 8. X-chromosome nondisjunction or loss  Males were scored as a measure of X chromosome loss or nondisjunction.  The percentage of males amongst the progeny of dpy-5 unc-13 rec-l/+ + +(I); him-17/him17(V) animals was found to be 25.8% (Table 2). No males were found among the progeny of dpy-5 unc-13 rec-l/+ + +(I) or dpy-5 unc-13 rec-l/+ + rec-l(l) animals. The percentage of  males was 27.35% amongst the progeny of unc-101 unc-54/+ +(I); him-17/him-l 7 animals, compared with a percentage of 0.06% in the wild-type background (Table 3). The percentage of male progeny from the dpy-18 unc-64 cross in a him-17 background was 32.8%; the percentage of male progeny in the wild-type background was 0.19% (Table 4). Male progeny in a him-17 background with no other markers present were scored. 154 males  58  were scored out of a total of 471 animals, a percentage of 32.7% (95% C. I. 27.7 - 38.3) (data not shown).  None of the crosses in lin-35 background yielded any male progeny (Tables 9 - 11).  A slight Him phenotype was observed in the progeny of some rec-1 mutant crosses. The progeny of rec-l/rec-1 (I); dpy-18 unc-32/+ + (111) hermaphrodites yielded 0.53% males (95% C. I. 0.35 - 0.77) compared with 0.07% males in the wild-type controls (95% C. I. 0.02 - 0.16) (Table 5). That a Him phenotype was not observed for all crosses can be attributed to the low increase in the number of males, and consequently the high number of progeny needed to obtain a significant result. A Him phenotype has previously been reported in rec-1 animals (Rattray and Rose, 1988).  Some variation was observed between different him-17 and rec-1 ;him-l 7 experiments. The percentage of male progeny from the dpy-5 unc-13 cross in a rec-1; him-17 double mutant background was 29.7% (95% C. I. 25.6 - 33.4). This is not significantly different from the percentage of males observed in the equivalent cross in a him-17 background (25.5%; 95% C. I. 22 - 29.4) (Table 6). When crossing over was measured over the dpy-18 unc-64 interval, rec-1;him-17 animals were found to give a lower percentage of male progeny (27,2, 95% C. I. 23,3 - 31,0) compared with him-17 alone (32,8% male progeny, 95% C. 1. 29,4 36,0) (Table 7), whereas when crossing over was measured in the unc-32 dpy-18 region, rec1;him-17 animals were found to give a higher percentage of male progeny (31,9, 95% C. I.  30,5 - 33,2) compared with him-17 alone (28,8% male progeny, 95% C. I. 27,7 - 29,8) (Table 8). It is likely that the high proportion of males is in these cases giving rise to  59  artificially narrow confidence intervals that do not reflect the large variation in the proportion of males between individual broods, arising from small brood sizes.  IV. DISCUSSION  In Caenorhabditis elegans, the distribution of crossing over events is non-random, producing disparities between the physical and genetic maps; the distribution of crossing over is known as the meiotic pattern. A mutation in rec-1 eliminates the meiotic pattern by increasing the frequency of crossing over in the meiotic cluster and decreasing it in flanking regions on the arms, producing a genetic map in rec-1 animals that is similar to the physical map. The effect of chromatin modifiers on the meiotic pattern was tested by measuring the frequency of recombination in animals bearing mutations in chromatin-modifying genes, him-17  and lin-35. The frequency of crossing over was measured both in intervals that are  recombinationally suppressed in N2, and in intervals that are highly recombinogenic in N2.  IV. 1. General discussion  Mutant him-17 was found to increase recombination in the meiotic cluster of LGI more than threefold from wild-type, an increase comparable to that measured in rec-1 mutants. The increase measured in rec-1 mutants agrees with that described in the literature (Zetka and Rose, 1995). In a flanking region that is highly recombinogenic in N2, mutant him-17 produced a two-fold decrease in the frequency of crossing over from wild-type. The frequency of crossing over in him-17 was, comparable to that described in a rec-1 background, in which the map distance between these two markers is 4.6 m. u. (95% C. I. 3.6 - 5.8, Zetka and Rose, 1995). Although the frequency of crossing over has not been measured across the entirety of LGI in him-17 mutants, it is clear that the pattern in these mutants is drastically altered. When the frequency of crossing over in him-17 animals is  61  compared with the physical and N2 genetic maps, it appears that the genetic map in him-17 animals is more similar to the physical map over the regions measured.  The frequency of crossing over was also measured over two intervals on LGIII. Comparisons of the physical and genetic maps in N2 and rec-1 animals have not yet been done for this chromosome. However, based on Brenner's genetic maps, which show a central cluster of genes across all of the autosomes (Brenner, 1974), and on data showing that mutant rec-1 alters crossover frequencies on LGV as well as on LGI (Zetka and Rose, 1995), one would expect that such a comparison would yield a figure similar to Fig. 2. Crossing over was measured across a flanking region on the arm. Based on its position, this region was expected to be highly recombinogenic in N2, and thus it was expected to undergo a decrease in recombination in rec-1 animals. In both him-17 and rec-1 animals, crossing over across this interval was found to decrease significantly. Crossing over was also measured across an interval covering part of the meiotic cluster and a flanking region. Crossing over in this region decreased in rec-1 animals compared with wild-type, but increased in him-17 animals, a disparity that can be attributed to a differential in the contribution of an increase in crossing over towards the centre of the chromosome and a decrease in crossing over on the arm for the two genes. Although crossing over was only measured along part of the chromosome, the genetic map in him-17 animals again appears more similar to the physical map than to the genetic map, suggesting that mutant him-17 eliminates the meiotic pattern on all autosomes.  Crossing over was measured in lin-35 null mutants. Like mutant rec-1 (Zetka and Rose, 1995), mutant lin-35 increased recombination over the recombinationally-suppressed dpy-11 unc-42 interval more than two-fold. Mutant lin-35 did not produce a similar increase in  62  recombination in a flanking region on LGV, eliminating the possibility that it simply increases recombination over all intervals. In lin-35 mutants, a highly recombinogenic region on the arm of LGIII undergoes a significant decrease in recombination to a degree comparable to that found in rec-1 and him-17 mutants. Thus, the lin-35 null mutant appears to eliminate the meiotic pattern.  How might the meiotic pattern be determined in C. elegans! An early model hinged on the resemblance of the Rec-1 phenotype to that of gamma irradiation, which introduces random double-strand breaks in a non-SPO-11-dependent pattern (Kim and Rose, 1987). This raised the possibility that mutant rec-1 introduces extra non-SPO-11-dependent DSBs. Furthermore, irradiation does not rescue the Rec-1 phenotype (A. Rose, personal communication). Creating more DSBs cannot apparently make the pattern more random than it already is in rec-1 mutants; this is consistent with the aforementioned extra DSB model. Meanwhile, crossover interference would be expected to ensure that the vast majority of these DSBs are repaired without crossovers occurring. However, mutant rec-1 did not rescue spo-11 mutants (Vijayaratnam, 2000). Although this does not exclude the possibility of extra DSBs occurring in rec-1 mutants, it does confirm that any extra DSBs do not result in crossovers. Extra DSBs can act as substrates for crossing over in spo-11 mutants, rescuing the Spo-11 phenotype. Thus the absence of crossing over in rec-1 ;spo-11 mutants suggests that any extra DSBs that might be created in rec-1 mutants are SPO-11dependent, refuting this model but not the possibility of DSBs being created.  Work in other organisms raised the possibility of chromatin modifications as a mechanism underlying the meiotic pattern. Two alternative models might explain how such modifications might alter the pattern.  63  The first model I considered was based on findings linking recombinational hotspots to an open chromatin configuration in yeast (Wu and Lichten, 1994; Mizuno et al., 1997; Yamada et al., 2004). According to this model, it is the access of SPO-11 to chromatin that determines the meiotic pattern; and this access is determined by chromatin modifications. A looser chromatin configuration allows SPO-11 easy access, resulting in a high frequency of crossing over; a repressive chromatin configuration blocks access by SPO-11, resulting in regions where crossing over is suppressed. According to this model, an overall loosening of chromatin compaction would grant SPO-11 access to all parts of the chromosome. This would result in an increase in SPO-11-mediated DSB formation, and a randomized pattern of DSB formation and crossing over. Conversely, an overall tightening of chromatin compaction should limit the access of SPO-11 to all parts of the chromosome. This would result in an overall decrease in the frequency of DSB formation, and a random pattern of those DSBs that did occur. Although this model is not the only one to include SPO-11 access as important in DSB formation, emphasis here is solely on SPO-11 's access, rather than its action. I therefore refer to this model as the SPO-11 access model to differentiate it from Reddy and Villeneuve's model, below.  The second model I considered was Reddy and Villeneuve's model (2004). This model, similarly to that put forward by Mizuno et al. (1997), posits that chromatin state modulates DSB formation by allowing or preventing DSB formation by SPO-11. However, the Reddy and Villeneuve model holds that a degree of chromatin compaction is critical for the action of SPO-11. Chromatin compaction in one region of the chromosome might be required to produce a compensatory decrease in compaction in a different region, promoting access of SPO-11 to chromatin in loosely compacted regions. Alternatively, compaction on one part  64  of a chromatin loop might produce mechanical stresses on other parts of the loop sufficient to trigger conversion of a reversible SPO-11 cleavage complex to an irreversible SPO-11DNA complex (Reddy and Villeneuve, 2004). According to this model, a general loosening of chromatin compaction along the chromosome should result in greater access of SPO-11 to chromatin, but less efficient cutting, producing an overall decrease in the frequency of DSB formation. Any DSBs that are created should follow a random distribution. Conversely, tightened chromatin compaction along the chromosome should result in a greater number of chromatin loops suitable as substrates for the action of SPO-11, distributed randomly along the chromosome, resulting in an increase in the number of DSBs with a random distribution.  In him-17 null mutants, DSBs do not occur (Reddy and Villeneuve, 2004); it therefore seems likely that the altered meiotic pattern in the non-null mutant can be attributed to a decrease in the number of DSBs. Mutant lin-35 is an enhancer of the Him-17 non-null phenotype (Reddy and Villeneuve, 2004) in a manner consistent with a decrease in DSB formation. Thus the altered meiotic pattern in lin-35 mutants can similarly be attributed to a decrease in DSB formation, him-17 mutants display delayed timing of chromatin modifications associated with a closed chromatin configuration (Reddy and Villeneuve, 2004), and LIN-35 acts in transcriptional repression (Harrison, 2006). Thus the decrease in DSB formation in these mutants coincides with an overall loosening in chromatin compaction; this supports Reddy and Villeneuve's model, rather than the earlier SPO-11 access model.  The Rec-1 phenotype might be explained in two ways, each consistent with Reddy and Villeneuve's model. According to a first hypothesis, rec-1 mutants might display an overall increase in chromatin compaction. This would result in the creation of an increased number  65  of SPO-11-dependent DSBs, with a random distribution. Irradiation would not be expected to rescue these mutants, as introducing yet more DSBs could not further randomize an already random pattern of crossing over. According to a second hypothesis, rec-l mutant chromosomes might undergo loosened chromatin compaction, similar to that found in lin-35 and him-17 mutants. The Rec-1 phenotype could then be attributed to a decrease in DSB formation, with any DSBs created being randomly distributed. A small decrease in DSB formation might be compensated for by a greater proportion of DSBs acting as substrates for crossing over. Irradiation would not be expected to rescue the meiotic pattern phenotype, as the distribution of crossing over is already random in these mutants.  To attempt to differentiate between these hypotheses, crossing over was scored in a rec1;him-17 double mutant in the dpy-5 unc-13 interval. Several phenotypes might be observed  in such a mutant.  If the first hypothesis were correct, mutant rec-1 and mutant him-17 should rescue one another, as the two genes would have antagonistic functions. A decrease in crossing over should therefore be observed when compared to either single mutant, restoring the frequency of crossing over to wild-type levels, or at least partly reversing the increase seen in either mutant. This effect was not observed, arguing against the hypothesis that the elimination of the meiotic pattern in rec-1 mutants can be attributed to increased chromatin compaction. The double mutant instead displayed the same phenotype as either single mutant in the dpy5 unc-13 region; a further experiment found that the double mutant decreased crossing over  in a highly recombinogenic region on the arm of LGIII, again displaying the same phenotype as either single mutant. The possibility of synthetic lethality was also eliminated,  66  demonstrating that mutant rec-1 does not enhance the Him-17 phenotype in the same way as lin-35 (RNAi).  Several interpretations of these data were possible. Either rec-1 or him-17 might be epistatic. Despite the double mutant displaying the same Him and reduced viability phenotypes as the him-17 mutant, these phenotypes were not shared by rec-1, compatible with him-17 epistasis. rec-1 and him-17 might be functionally redundant, acting in independent branches of a non-linear pathway, with a common target. In this case, the Him and reduced viability phenotypes of him-17 mutants must result from a function not shared by rec-1. Alternatively, additive or synergistic effects of rec-1 and him-17 might be masked in regions in the cluster or on the arm if these regions attained a maximum degree of randomization in the single mutants.  To differentiate between epistasis, synergy and additive effects, crossing over was measured over the unc-32 dpy-18 interval on LGI, an interval that undergoes a decrease in crossing over in a rec-1 background, and an increase in crossing over in a him-17 background. In hybrid regions, an increase in crossing over in the cluster may counter a decrease in crossing over in the flanking region, or vice versa. Consequently, synergistic or additive effects might be more apparent in such a region than in a region in the cluster or the arm. Conversely, in the case of epistasis, the different meiotic phenotypes of rec-1 and him-17 in this region should clarify which mutant was epistatic to the other.  In the double mutant, the frequency of crossing over was not significantly different from wild-type, consistent with additive effects of him-17 and rec-1 in this region, and eliminating the possibilities of synergistic effects or epistasis between the two mutants.  67  The expression of multicopy tandem-array transgenes is suppressed in lin-35 mutants. In lin35 mutants, the meiotic pattern is altered; following on from this discovery, strains were constructed to determine whether mutant rec-1 and him-17 might do the same. No difference in transgene expression was observed in these mutants when compared with wild-type. LIN35 has been implicated in a wide range of processes (Lu and Horvitz, 1998, Lehner et al., 2006, Ouellet, 2007); it is possible that its role in transgene suppression is unrelated to its meiotic function. The complete lack of a Him phenotype in lin-35 mutants appears to support different meiotic functions for LIN-35 from REC-1 and HIM-17. Antagonistic functions between him-17 and lin-35 were initially considered as an explanation for the lack of transgene expression suppression in him-17 mutants. If this were the case, one might expect to see an increase, rather than a decrease, in transgene expresion in him-17 and rec-1 mutants. ccls4251 is strongly expressed, and an increase in transgene expression in mutants would not be visible with the naked eye. However, in light of the enhancement of the Him17 phenotype by lin-35 (RNAi), this explanation can be dismissed. It is more plausible that any effects of HIM-17 and REC-1 on chromatin structure, and therefore potential transgene suppression, are limited to meiosis, and thus not visible in a somatically-expressed transgene.  Crossing over was measured in the presence of an integrated multicopy tandem array transgene, ccls4251. A heterozygous copy of this transgene was found to suppress crossing over in a region of LGI. This crossover suppression, limited to the same chromosome as the site of insertion of the transgene, is similar to that reported for other large, integrated tandem-array GFP transgenes (Hammarlund et al., 2005). A mutation in a chromatin modifier, tam-1, antagonizes this suppression in a flanking region. Mutant tam-1 also  68  suppresses GFP expression of ccls4251. It might be that mutant tam-1 simply increases the frequency of crossing over across all intervals. However, it is intriguing that tam-1 should affect both ccIs425Ts expression and ccls425Vs suppression of crossing over. Transgenes are silenced in the germline; ccls4251 might suppress crossing-over as a result of changes to chromatin structure covering both the insertion and a region on either side. The reversal of this crossover suppression might then result from the same changes to chromatin structure that account for transcriptional repression in tam-1 mutants.  IV. 2. Future work  There are several ways in which work on the meiotic pattern might be expanded in the future.  The function of rec-1 might be narrowed down. Now that similarities have been found between it and two chromatin modifiers, the possibility that it, too, is a chromatin modifier should be explored. Antibody staining might be used to determine whether rec-1 mutants display similar abnormal histone modifications to him-17 mutants. That any phenotype might be limited to germ cells complicates the use of micrococcal nuclease, as has been done in yeast. Nevertheless, by dissecting the gonad and comparing against the wild-type gonad, this might be possible.  Chromatin-modifying effects of mutant rec-1 and mutant him-17 could be further tested by examining their effects on transgene expression in the germline. This is not trivial, as the majority of transgenes are silenced in the germline. Some transgenes have been reported exempt from this effect; however, somatic expression of these transgenes is not affected by  69  mutanttam-7(Hsiehet al., 1999). Furthermore, their non-repetitive nature does not make them ideal candidates for testing the effects of mutants of context-dependent silencing.  In wild-type, the meiotic map provides an excellent tool for comparison with data resulting from high-throughput ChlP-on-chip methods: nucleosome distribution, histone acetylation or methylation patterns.  The effects of other chromatin modifiers on the meiotic pattern should also be examined. Ideally, mutants should each affect a known chromatin modification, allowing the effect of each modification on the meiotic pattern to be examined separately. This is easier said than done: mutations in many known chromatin modifiers are lethal, or have a phenotype (e. g. very low brood size, transcriptional defects) that hampers the scoring of recombinants in the numbers required. Nevertheless, as more of these modifiers are better characterized, it may be that new modifiers or non-null mutations in already characterized modifiers will be of use. Further analysis could start with the other members of the DRM complex, to narrow down the mechanism by which LIN-35 affects the meiotic pattern. Three SynMuv genes, lin-9, lin-53 and lin-54, are needed to stabilize this complex, and although three null mutants  are sterile (Harrison et al., 2006), non-null mutants of these genes might be a good place to start. Mutant rec-1 has not been found to have a SynMuv phenotype (N. O'Neil, personal communication), but this does not exclude the possibility of a role relating to the DRM complex.  70  IV. 3. Conclusion  Caenorhabditis elegans displays a unique pattern of crossing over, with a recombinationally suppressed central region flanked by highly recombinogenic arms. Chromosomes display strong crossover interference, dispensing with the need to consider double crossovers in C. elegans in most cases.  In this thesis, I describe three genes that alter the distribution of the single crossover that occurs autosomally. This phenotype was previously unknown in these mutants. Two mutations eliminate the meiotic pattern. This evidence of a link between chromatin modifications and sites of crossing over is consistent with a similar link hinted at in other species. Further work with these mutants, and further clarification of the processes governing the meiotic pattern, will provide powerful tools for the study of meiotic doublestrand break formation, crossing over and crossover interference.  71  V. BIBLIOGRAPHY  ALBERTSON, D. G. and J. N. THOMSON, 1982 The kinetochores of Caenorhabditis elegans. Chromosoma 86: 409-428  ALBERTSON, D. G. and J. N. THOMPSON, 1993 Segregation of holocentric chromosomes at meiosis in the nematode, Caenorhabditis elegans. Chromosome Res. 1:1526  ALPI, A., P. PASIERBEK, A. GARTNER and J. LOIDL, 2003 Genetic and cytological characterization of the recombination protein RAD-51 in Caenorhabditis elegans. Chromosoma 112: 6-16  BARNES, T. M., Y. KOHARA, A. COULSON and S. HEKIMI, 1995 Meiotic recombination, noncoding DNA and genomic organization in Caenorhabditis elegans. Genetics 141: 159-179  BARTOLOME, C, X. MASIDE and B. CHARLESWORTH, 2002 On the abundance and distribution of transposable elements in the genome of Drosophila melanogaster. Mol. Biol. Evol. 19: 926-937  BEAN, C. J., C. E. SCHANER and W. G. KELLY, 2004 Meiotic pairing and imprinted X chromatin assembly in Caenorhabditis elegans. Nat. Genet. 36: 100-105  72  BESSLER, J. B., K. C. REDDY, M. HAYASHI, J. HODGKIN and A. M. VILLENEUVE, 2007 A role for Caenorhabditis elegans chromatin-associated protein HIM-17 in the proliferation vs. meiotic entry decision. Genetics 175: 2029-2037  BOISS1NOT, S., A. ENTEZAM and A. V. FURANO, 2001 Selection against deleterious LINE-1-containing loci in the human lineage. Mol. Biol. Evol. 18: 926-935  BRENNER, S., 1974 The genetics of Caenorhabditis elegans. Genetics 77: 71-94  C. ELEGANS SEQUENCING CONSORTIUM, 1998 Genome sequence of the nematode C. elegans: a platform for investigating biology. Science 282: 2012-8  CARLTON, P. M., A. P. FARRUGGION and A. F. DERNBURG, 2006 A link between meiotic prophase progression and crossover control. PLOS Genetics 2: el2  CEOL, C. J. and H. R. HORVITZ, 2004 A new class of C. elegans synMuv genes implicates a Tip60/NuA4-like HAT complex as a negative regulator of Ras signaling. Dev Cell 6: 563576  COLAIACOVO, M. P., 2006 The many facets of SC function during C. elegans meiosis. Chromosoma 115: 195-211  COLAIACOVO, M. P., A. J. MACQUEEN, E. MARTINEZ-PEREZ, K. MCDONALD, A. ADAMO, A. LA VOLPE and A. M. VILLENEUVE, 2003 Synaptonemal Complex  73  Assembly in C. elegans Is Dispensable for Loading Strand-Exchange Proteins but Critical for Proper Completion of Recombination. Dev. Cell 5: 463-474  COUTEAU, F. and M. ZETKA, 2005 HTP-1 coordinates synaptonemal complex assembly with homolog alignment during meiosis in C. elegans. Genes & Dev. 19: 2744-1756  COUTEAU, F., K. NABESHIMA, A. VILLENEUVE and M. ZETKA, 2004 A component of C. elegans meiotic chromosome axes at the interface of homolog alignment, synapsis, nuclear reorganization, and recombination. Curr. Biol. 14: 585-592  CROW, E. L. and R. S. GARDNER, 1959 Confidence intervals for the expectation of a Poisson variable. Biometrika 46: 441-453  CUI, M., D. S. FAY and M. HAN, 2004 lin-35/Rb Cooperates With the SWI/SNF Complex to Control Caenorhabditis elegans Larval Development. Genetics 167: 1177-1185  DE MASSY, B., 2003 Distribution of meiotic recombination sites. Trends Genet. 19: 514522  DERNBURG, A. F., K. MCDONALD, G. MOULDER, R. BARSTEAD, M. DRESSER and A. M, VILLENEUVE, 1998 Meiotic recombination in C. elegans initiates by a conserved mechanism and is dispensable for homologous chromosome synapsis. Cell 94: 387-398  74  DURET, L., G. MARAIS and C BIEMONT, 2000 Transposons but not retrotransposons are located preferentially in regions of high recombination rate in Caenorhabditis elegans. Genetics 156: 1661-1669  EMMONS, S. W. and L. YESNER 1984 High-frequency excision of transposable element Tc 1 in the nematode Caenorhabditis elegans is limited to somatic cells. Cell 36: 599-605  FAY, D. S. and J. YOCHEM, 2007 The SynMuv genes of Caenorhabditis elegans in vulval development and beyond. Dev. Biol. 306: 1-9  FIRE, A., S. XU, M. K. MONTGOMERY, S. A. KOSTAS, S. E. DRIVER and C. C. MELLO, 1998 Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391: 806 - 811  FRITSCH, O, G. BENVENUTO, C. BOWLER, J. MOLINIER and B. HOHN, 2004 The INO80 protein controls homologous recombination in Arabidopsis thaliana. Mol. Cell 16: 479-485  GARCIA-MUSE, T. and S. J. BOULTON, 2007 Meiotic recombination in Caenorhabditis elegans. Chromosome Res. 15: 607-621  GREEN WALD, I., 1997 Development of the Vulva, pp. 519-542 in C. elegans II, edited by D. L. RIDDLE, T. BLUMENTHAL, B. J. MEYER and J. R. PRIESS. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.  75  GREENWALD, I., A. COULSON, J. SULSTON, J. PRIESS, 1987 Correlation of the physical and genetic maps in the lin-12 region of Caenorhabditis elegans. Nucleic Acids Res. 15: 2295-307  HAMMARLUND, M., M. W. DAVIS, H. NGUYEN, D. DAYTON and E. M. JORGENSON, 2005 Heterozygous insertions alter crossover distribution but allow crossover interference in Caenorhabditis elegans. Genetics 171: 1047-1056  HARRISON, M. M., C. J. CEOL, X. LU, and H. R. HORVITZ, 2006 Some C. elegans class B synthetic multivulva proteins encode a conserved LIN-35 Rb-containing complex distinct from aNuRD-like complex. PNAS 103: 16782-16787  HILLERS, K. J. and A. M. VILLENEUVE, 2003 Chromosome-wide control of meiotic crossing over in C. elegans. Current Biol. 13: 1641-1647  HODGKIN, J., H. R. HORVITZ and S. BRENNER, 1978 Nondisjunction mutants of the nematode Caenorhabditis elegans. Genetics 91: 67-94  HSIEH, J., J. LIU, S. A. KOSTAS, C. CHANG, P. W. STERNBERG and A. FIRE, 1999 The RING finger/B-Box factor TAM-1 and a retinoblastoma-like protein LIN-35 modulate context-dependent gene silencing in Caenorhabditis elegans. Genes Dev. 13: 2958-2970  JENSEN-SEAMAN, M. I., T. S. FUREY, B. E. PAYSEUR, Y. LU, K. M. ROSKIN, C.-F. CHEN, M. A. Thomas, D. HAUSSLER and H. J. JACOB, 2004 Comparative recombination rates in the rat, mouse and human genomes. Genome Res. 14: 528 - 538  76  KELLY, W. G., S. XU, M. K. MONTGOMERY and A. FIRE, 1997 Distinct requirements for somatic and germline expression of a generally expressed Caenorhabditis elegans gene. Genetics 146: 227 - 238  KIEFER, J. C, P. A. SMITH and S. E. MANGO, 2007 PHA-4/FoxA cooperates with TAM1/TRIM to regulate cell fate restriction in the C. elegans foregut. Dev. Biol. 303(2): 611-624  KIM, J. S. and A. M. ROSE, 1987 The effect of gamma radiation on recombination frequency in Caenorhabditis elegans. Genome 29: 457-62  LEHNER, B., A. CALIXTO, C. CROMBIE, J. TISCHLER, A. FORTUNATO, M. CHALFIE and A. G. FRASER, 2006 Loss of LIN-35, the Caenorhabditis elegans ortholog of the tumor suppressor pl05Rb, results in enhanced RNA interference. Genome Biol. 7: R4  LICHTEN, M. and A. S. H. GOLDMAN, 1995 Meiotic recombination hotspots. Annu. Rev. Genetics 29: 423-44  LU, X. and H. R. HORVITZ, 1998 lin-35 and lin-53, two genes that antagonize a C. elegans  Ras pathway, encode proteins similar to Rb and its binding protein RbAp48. Cell 95: 981991  LYNN, A., T. ASHLEY and T. HASSOLD, 2004 Variation in human meiotic recombination. Annu. Rev. Genomics Hum. Genet. 5: 317-349  77  MACQUEEN, A. J., C. M. PHILLIPS, N. BHALLA, P. WEISER, A. M. VILLENEUVE and A. F. DERNBURG, 2005 Chromosome sites play dual roles to establish homologous synapsis during meiosis in C. elegans. Cell 123: 1037-1050  MAES, J., S. CHAPPAZ, P. CAVALIER, L. O'NEILL, B. TURNER, F. ROUGEON and M. GOODHARDT, 2006 Activation of V(D)J recombination at the IgH chain JH locus occurs within a 6-kilobase chromatin domain and is associated with nucleosomal remodeling. J. Immunol. 176: 5409-5417  MARTIN, J. S., N. WINKELMANN, M. I. R. PETALCORIN, M. J. MCILWRAITH and S. J. BOULTON, 2005 RAD-51-dependent and -independent roles of a Caenorhabditis elegans BRCA2-related protein during DNA double-strand break repair. Mol. Cell. Biol. 25: 3127-3139  MARTINEZ-PEREZ, E. and A. M. VILLENEUVE, 2006. HTP-1-dependent constraints coordinate homolog pairing and synapsis and promote chiasma formation during C. elegans meiosis. Genes & Dev. 19: 2727-2743  MAINE, E. M., J. HAUTH, T. RATLIFF, V. E. VOUGHT, X. SHE and W. G. KELLY, 2005 EGO-1, a putative RNA-dependent RNA polymerase, is required for heterochromatin assembly on unpaired DNA during C. elegans meiosis. Curr. Biol. 15: 1972-1978  MCCARTER, J., B. BARTLETT, T. DANG and T. SCHEDL, 1999 On the control of oocyte meiotic maturation and ovulation in Caenorhabditis elegans. Dev. Biol. 205: 111128  78  MCKIM, K.S., A. M. HOWELL and A. M. ROSE, 1988 The effects of translocations on recombination frequency in Caenorhabditis elegans. Genetics 120: 987-1001  MCKIM, K. S. and A. M. ROSE, 1990 Chromosome I duplications in Caenorhabditis elegans. Genetics 124: 115-132  MCKIM, K. S., K. PETERS and A. M. ROSE, 1993 Two types of sites required for meiotic chromosome pairing in Caenorhabditis elegans. Genetics 134: 749-768  MCMURRAY, M. T. and M. S. KRANGEL, 2000 A role for histone acetylation in the developmental regulation of VDJ recombination. Science 287: 495^198.  MEZARD, C, 2006 Meiotic recombination hotspots in plants. Biochem. Soc. Trans. 34: 531-534  MIZUNO, K., Y. EMURA, M. BAUR, J. KOHLI, K. OHTA and T. SHIBATA, 1997. The meiotic recombination hot spot created by the single base substitution ade6-M26 results in remodeling of chromatin structure in yeast. Genes & Dev. 11: 878-886  MORGAN, T. H., 1914 No crossing over the male Drosophila of genes in the second and third pair of chromosomes. Biol. Bull. 26: 195-204  79  NABESHIMA, K., A. M. VILLENEUVE and K. J. HILLERS, 2004 Chromosome-wide regulation of meiotic crossover formation in Caenorhabditis elegans requires properly assembled chromosome axes. Genetics 168: 1275-1292  NABESHIMA, K., A. M. VILLENEUVE and M. P. COLAIACOVO, 2005 Crossing over is coupled to late meiotic prophase bivalent differentiation through asymmetric disassembly of the SC. J. Cell. Biol. 168: 683-9  OUELLET, J. and R. ROY, 2007 The Iin-35/Rb and RNAi pathways cooperate to regulate a key cell cycle transition in C. elegans. BMC Dev Biol. 27: 7-38  PAGE, S. L. and R. S. HAWLEY, 2005 The meiotic mutant mei-352 is an allele ofklp3A and reveals a role for a kinesin-like protein in crossover distribution. Genetics 170: 1797— 1807.  PENKNER, A., L. TANG, M. NOVATCHKOVA, M. LADURNER, A. FRIDKIN, Y. GRUENBAUM, D. SCHWEIZER, J. LOIDL and V. JANTSCH, 2007 The nuclear envelope protein matefin/SUN-1 is required for homologous pairing in C. elegans meiosis. Dev. Cell 12: 873-885  PETALCORIN, M. I. R., J. SANDALL, D. B. WIGLEY and S. J. BOULTON, 2006 CeBRC-2 stimulates D-loop formation by RAD-51 and promotes DNA single-strand annealing. J. Mol. Biol. 361: 231-242  80  PHILLIPS, C. M., C. WONG, P. M. BHALLA, P. WEISER, P. M. MENEELY and A. F. DERNBURG, 2005 HIM-8 binds to the X chromosome pairing center and mediates chromosome-specific meiotic synapsis. Cell 123: 1051-1063  PHILLIPS, C. M . and A. F. DERNBURG, 2006 A family of zinc-finger proteins is required for chromosome-specific pairing and synapsis during meiosis in  C. elegans.  Dev. Cell  11:  817-829.  RATTRAY, B. and A. M. ROSE, 1988 Increased intragenic recombination and nondisjunction in the Rec-1 strain of Caenorhabditis elegans. Genet. Res. Camb. 51: 89-93  REDDY, K. C. and A. M. VILLENEUVE, 2004 C. elegans HIM-17 links chromatin modification and competence for initiation of meiotic recombination. Cell 118: 439-452  ROSE, A. M. and D. L. BAILLIE, 1979a Effect of temperature and age on recombination and nondisjunction in Caenorhabditis elegans. Genetics 92: 409-418.  ROSE, A. M. and D. L. BAILLIE, 1979b A mutation in Caenorhabditis elegans that increases recombination frequency more than threefold. Nature 281: 599-600.  ROSE, A. M., D. L. BAILLIE and J. CURRAN, 1984 Meiotic pairing behavior of two free duplications of linkage group I in Caenorhabditis elegans. Mol. Gen. Genet.  195:  52-56  ROSENBLUTH, R. E. and D. L. BAILLIE, 1981 The genetic analysis of a reciprocal translocation, eTl (III; V) in Caenorhabditis elegans. Genetics 99: 415-428  81  S C H E D L , T., 1997 Developmental Genetics of the Germline, pp. 241-270 in C. elegans II, edited by D. L . R I D D L E , T. B L U M E N T H A L , B . J. M E Y E R and J. R. PRIESS. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N Y .  SINGER, A . , H . P E R L M A N , Y . Y A N , C. W A L K E R , G . C O R L E Y - S M I T H , B . B R A N D H O R S T and J. POSTLETHWAIT, 2002 Sex-specific recombination rates in zebrafish (Danio rerio). Genetics 160: 649-657  S M O L I K O V , S., A . EIZINGER, A . H U R L B U R T , E. R O G E R S , A . M . V I L L E N E U V E and M . P. C O L A I A C O V O , 2007 Synapsis-defective mutants reveal a correlation between chromosome configuration and the mode of double-strand break repair during C. elegans meiosis. Dev. Biol. 308: 206-221  V I L L E N E U V E , A . M . , 1994 A cw-acting locus that promotes crossing over between X chromosomes in Caenorhabditis elegans. Genetics 136: 887 - 902  V I L L E N E U V E , A . M . and K . J. HILLERS, 2001 Whence meiosis? Cell 106: 647-650  V I J A Y A R A T N A M , V . , 2000 Studies of rec-1 function in Caenorhabditis elegans. U B C thesis  W I C K Y , C. and A . M . ROSE, 1996 The role of chromosome ends during meiosis in Caenorhabditis elegans. BioEssays 18: 447-452  82  WRIGHT, S. I., N. AGRAWAL and T. E. BUREAU, 2003 Effects of recombination rate and gene density on transposable element distributions in Arabidopsis thaliana. Genome Res. 13: 1897-1903  WU, T.-C. and LICHTEN, M., 1994 Meiosis-induced double-strand break sites determined by yeast chromatin structure. Science 263: 515-518  YAMADA, T., K.-I. MIZUNO, K. HIROTA, N. KON, W. P. WAHLS, E. HARTSUIKER, H. MUROFUSHI, T. SHIBATA and K. OHTA, 2004 Roles of histone acetylation and chromatin remodeling factor in a meiotic recombination hotspot. EMBO J 23: 1792-1803  ZETKA, M.-C. and A. M. ROSE, 1990 Sex-related differences in crossing over in Caenorhabditis elegans. Genetic 126: 355-363  ZETKA, M.-C. and A. M. ROSE, 1992 The meiotic behavior of an inversion in Caenorhabditis elegans. Genetics 131: 321-332  ZETKA, M.-C. and A. M. ROSE, 1995 Mutant rec-1 eliminates the meiotic pattern of crossing over in Caenorhabditis elegans. Genetics 141: 1339-1349.  WormBase web site, http://www.wormbase.org, release WS177, date July 31, 2007  83  APPENDIX 1 Table 19: Strains used  Strain name  Genotype  N2 BC62  Wild-type, variety Bristol dpy-5(e61) rec-1 (si80) unc-75(e950)I.  BC313  rec-l(sl80)l.  BC502  dpy-18(e364) unc-64(e246)III  1  BC657  unc-32(e!89) dpy-18(e364)IIl.  CB4523'  him-17 (e2707)V.  DR181  unc-60(m35) dpy-11 (e224)V.  KR177  dpy-5(e61) unc-13(e51) rec-1 (si80)1.  KR181  rec-1 (sl80)I; dpy-1 l(e224) unc-42(e270)V.  12  KR1473  unc-lOl(ml) unc-54(el90)I.  KR4340  ccls42511.  MT633  lin-ll(n389)I; him-5(el467)V.  MT10430'  Iin-35(n745)l.  PD6249  ccls42511.;tam-l(cc567) V.  3  1  1  1 2 3  Available from the Caenorhabditis Genetics Centre Provided by J. Hodgkin Reisolated from PD6249  84  

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.831.1-0101000/manifest

Comment

Related Items