Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Analysis of chromosome I rearrangements in Caenorhabditis elegans McKim, Kim Stewart 1990

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

Item Metadata


831-UBC_1991_A1 M35.pdf [ 12.77MB ]
JSON: 831-1.0100352.json
JSON-LD: 831-1.0100352-ld.json
RDF/XML (Pretty): 831-1.0100352-rdf.xml
RDF/JSON: 831-1.0100352-rdf.json
Turtle: 831-1.0100352-turtle.txt
N-Triples: 831-1.0100352-rdf-ntriples.txt
Original Record: 831-1.0100352-source.json
Full Text

Full Text

ANALYSIS OF CHROMOSOME I REARRANGEMENTS IN CAENORHABDITIS ELEGANS By KIM STEWART McKIM B.Sc, Simon Fraser University, 1986 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES GENETICS PROGRAMME We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December 1990 ©Kim Stewart McKim, 1990 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of VV<>&SL^. QyJ^ssS^^ The University of British Columbia Vancouver, Canada Date -"v^s^W^o^ V - l " ^ s S ^ DE-6 (2/88) ii ABSTRACT In this thesis, chromosome I rearrangements were used to study the organization of essential genes and regions important for chromosome behaviour in the nematode Caenorhabditis elegans. To facilitate the genetic mapping of mutations in essential genes, rearrangements were isolated using a procedure designed to recover derivative chromosome I duplications shortened by gamma radiation from existing duplications. Sixty-two duplications were isolated in this way. These duplications, along with three deletions isolated in this study and 9 existing deletions of the region, divided the left half of chromosome I into at least 24 regions. Protocols were developed and used to rapidly map mutations into the regions defined by the breakpoints. The techniques and results described demonstrate the feasibility of carrying out a similar analysis on the whole genome. The majority of duplications behaved as if they were free; that is they segregated independently of the euploid chromosome set. While size was an important determinant of mitotic stability, clear exceptions to a size - stability correlation were observed. For example, despite its larger size, hDp72 was lost during cell division more frequently than hDpl8, suggesting features of chromosome structure were important. Shortening of duplications in the unc-11 dpy-5 region caused greater reductions in mitotic stability than similar sized shortenings in the dpy-5 unc-13 region. Therefore, specific sequences appear to influence duplication stability. Some free duplications were also observed to break spontaneously. Breakage occurred at different frequencies for different duplications and correlated with mitotic instability. The meiotic properties of four translocations involving chromosome I were examined. No recombination was observed in any of the translocation heterozygotes along the left (let-362 - unc-13) portion of chromosome I. By isolating a half-translocation chromosome as a free duplication, I mapped the breakpoints of three of the translocations. The boundaries of cross-over suppression coincided with the physical breakpoints. These results agree with the proposal that DNA sequences at the right end of chromosome I are essential for homologue recognition followed by iii meiotic synapsis and recombination. The published data of other translocations and duplications indicates that each of the other five C. elegans chromosomes has DNA sequences localized to one end that are required for homologue recognition and recombination. iv TABLE OF CONTENTS Abstract ii List of Tables vi List of Figures vii Acknowledgements viii 1. Introduction 1 1.1 Organization of essential genes. 5 1.2 Chromosome behaviour. 8 2. Materials and Methods 14 2.1 General 14 2.2 Strains. 16 2.3 Isolation of new rearrangements. 20 2.4 Complementation analysis. 22 2.5 Genetically determining the size of translocation-derived duplication chromosomes. 27 2.6 Complementation with him strains. 30 2.7 Scoring and calculating recombination. 31 2.8 Linkage of unlinked markers to chromosome I duplications. 35 3. Results 37 3.1 Source of breakpoints on the left half of chromosome I. 37 3.1.1. Isolation of gamma radiation induced lethal mutations. 37 3.1.2. Induced deletion of duplication chromosomes 40 3.2 Mapping the gamma radiation lethal mutations and the new duplication breakpoints. 44 3.3 The left end of chromosome I. 57 3.4 The dpy-14 region. 61 3.4.1. Source of mutations. 61 3.4.2. Mapping mutations to rearrangements. 64 3.5 Lethal mutations isolated with the translocation hTl. 74 3.6 Translocations 78 3.6.1. szTl(I;X) 78 3.6.2. hTl(I;V) 91 3.6.3. hT2(I;III) 95 3.6.4. hT3(I;X) 101 V 3.7 Insertions 106 3.8 Meiotic behaviour of duplications. 113 3.8.1. Duplications with meiotic pairing behaviour: Chromosome I. 115 3.8.2. Duplications with meiotic pairing behavior: The X chromosome. 121 3.9 Non-homologous disjunction. 128 3.10 Effect of deletions on recombination. 133 3.10.1. Chromosome I deletions 134 3.10.2. X-chromosome deletions 143 3.11 Meiotic mutants him-3, him-6 and him-8. 146 3.11.1. Effects on recombination. 148 3.11.2. Autosomal nondisjunction. 151 3.12 Tests for an inter-chromosomal effect. 154 3.13 Segregational behavior of free duplications 156 3.13.1. Segregation analysis of the duplications. 156 3.13.2. Spontaneous shortening of duplications: 166 Spontaneous breakdown of hDpl4. 176 3.13.3. Genetic mosaics. 180 4. Discussion 185 5. Bibliography 222 vi LIST OF TABLES Table 1. Abbreviations 15 Table 2. Mutations isolated in other studies. 17 Table 3. Summary of gamma radiation induced lethal mutations. 38 Table 4. Isolation of duplications with breakpoints between unc-11 and dpy-5 42 Table 5. Isolation of duplications with breakpoints between dpy-5 and dpy-14 42 Table 6. Isolation of duplications with breakpoints between dpy-5 and unc-13 42 Table 7. Mapping of mutations at the left end of chromosome I. 59 Table 8. Mapping of mutations in the dpy-14 region. 65 Table 9. Duplication mapping of hTl lethal mutations. 75 Table 10. Deficiency mapping of hTl lethal mutations. 76 Table 11. Recombination in control crosses. 89 Table 12. Recombination in szTl(I;X) heterozygotes. 90 Table 13. Recombination in hTl(I;V) heterozygotes. 93 Table 14. Recombination in hT2(I;III) heterozygotes. 97 Table 15. Recombination in hT3(I;X) heterozygotes. 104 Table 16. Mapping and recombination in hDpl4 strains. 107 Table 17. Recombination in insertion strains. I l l Table 18. Effects of duplications on recombination. 114 Table 19. Recombination involving chromosome I duplications. 116 Table 20. Recombination in X-chromosome duplication bearing males. 126 Table 21. Duplication stability. I. Male sperm. 129 Table 22. Recombination in deficiency heterozygotes: The left end of chromosome I. 135 Table 23. Recombination in deficiency heterozygotes: The unc-29 region. 136 Table 24. Recombination in deficiency heterozygotes: The right end of chromosome I. 139 Table 25. Segregation and recombination in mnl64(I;X) strains. 141 Table 26. Recombination in X-chromosome deficiency heterozygotes. 145 Table 27. The effect of him mutations on recombination frequency. 149 Table 28. The effect of Him mutations on duplication stability. 153 Table 29. The effect of crossover suppressors on recombination of other chromosomes. 155 Table 30. Duplication stability: Hermaphrodite gametes. 157 Table 31. Duplication stability: Oocytes. 162 Table 32. Stability of spontaneous duplications: Hermaphrodite gametes. 173 Table 33. Recovery of spontaneously shortened duplication strains. 175 Table 34. Recovery of genetic mosaics from duplication strains. 182 Vll LIST OF FIGURES Figure 1. Simplified genetic map of the C. elegans genome. 19 Figure 2. Determining linkage of lethal mutations isolated in hT2(I;III) strains. 21 Figure 3. Mapping mutations to duplication breakpoints. 23 Figure 4. Complementation testing with hDf8. 28 Figure 5. Punnett square of dpy-5 unc-z; + lhTl(I;V) heterozygote. 32 Figure 6. Isolating shortened duplications. 41 Figure 7. Duplications and deficiencies on the left half of chromosome I. 45 Figure 8. Duplications which break in the unc-11 dpy-5 region. 46 Figure 9. Duplications which break in the dpy-5 unc-13 region. 47 Figure 10. Position of breakpoints of dpy-5(-) derivatives of ILXLszTl. 49 Figure 11. Genetic map of the left end of chromosome I. 58 Figure 12. Genetic map of the dpy-14 unc-29 region. 62 Figure 13. Diagram of the translocation chromosomes. 79 Figure 14. Punnett square of dpy-5; unc-3/szTl(I;X). heterozygote. 81 Figure 15. Punnett square of unc-29 unc-751 unc-29 unc- 75/IRVLhTl hermaphrodite. 118 Figure 16. Diagram of I^X^szTl and some derivatives. 122 Figure 17. Punnett square of dpy-5; unc-11 unc-llhDp56 hermaphrodite. 124 Figure 18. Non-homologous segregation of linked duplications. 131 Figure 19. Deletions at the right end of chromosome I. 138 Figure 20. Deletions at the right end of the X-chromosome. 144 Figure 21. Genetic map of the dpy-13 unc-31 region. 147 Figure 22. Segregational stability and duplication size. 159 Figure 23. Pedigree of spontaneous duplications. 167 Figure 24. Structure of spontaneous duplications 169 Figure 25. Proposed ring chromosomes hDp4 and KDp23. 172 Figure 26. Spontaneous breakdown products of hDpl4. 178 Figure 27. Summary of C. elegans cell lineage. 183 Figure 28. Distribution of lethal mutations in the dpy-14 region. , 190 Figure 29. Distribution of lethal mutations from other studies. 192 Figure 30. Chromosome fragments with and without meiotic pairing activity. 197 Figure 31. Homologue recognition sites on each C. elegans chromosome. 200 viii ACKNOWLEDGMENTS The last four years in the laboratory of my supervisor, Dr. Ann Rose, have been very stimulating and enjoyable. I was given all the freedom I wanted, but the advice, encouragement and stimulating discussion from Dr. Rose was always useful and available. Similarly, I am grateful for the advice and encouragement from the members of my supervisory committee, Dr. David Baillie, Dr. Pat Dennis, Dr. Fred Dill and Dr. David Holm. I also thank Dr. A. T. C. Carpenter for helpful comments on my research and an earlier manuscript, and Theresa Rogalski for comments on a draft of this thesis. I owe special thanks to Dr. David Baillie and Raja Rosenbluth for my initial, and ongoing, lessons in worm genetics. I have also benefitted from working along side many good student colleagues. I would like to thank Joe Babity, Dr. Denise Clark, Dr. Linda Harris, Dr. Robert Johnsen, Jennifer McDowall, Ken Peters, Dr. Shiv Prasad and the many undergraduates who provided most of the technical support. I would also like to thank Dr. Terry Starr for teaching me some molecular biology. I am especially grateful for the discussions and friendship of Dr. Ann Marie Howell and Monique Zetka. I have worked very hard on my thesis. So hard in fact I have tried the patience of a number of people. My Mother, Father, Cara Lee and some wonderful friends endured my many tardy performances. Without their support and encouragement, the last four years would not have been so enjoyable. This work was supported in part by a Studentship from the Medical Research Council of Canada. ix "Years later, Evelyn Wilkin-at the time a young bacterial geneticist on fellowship at Cold Spring Harbor-asked her how she could have worked for two years without knowing what was going to come out. 'It never occurred to me that there was going to be any stumbling block. Not that I had the answer, but [I had] the joy of going at it. When you have the joy, you do the right experiments. You let the material tell you where to go, and it tells you at every step what the next has to be because you're integrating with an overall brand new pattern in mind. You're not following an old one; you are convinced of a new one. And you let everything you do focus on that. You can't help it, because it all integrates. There were no difficulties.'" From "A Feeling for the Organism: The life and work of Barbara McClintock." By Evelyn Fox Keller. 1983 W. H. Freeman and Company. 1 1. Introduction The genetic material in the nucleus of a eukaryotic organism is organized into chromosomes. During the cell divisions of mitosis and meiosis, a series of tightly regulated steps ensures the accurate replication and segregation of genetic material to the two daughter cells. By grouping the genes into a limited set of chromosomes, the cell can divide the replicated genetic material and ensure that the daughter cells do not receive deficiencies or duplications of genetic material. The chromosomes of all organisms have the same basic structural components. The DNA of the mitotic chromosomes is tightly folded into chromatin fibers which are anchored by looping through a scaffold structure (reviewed in Earnshaw 1988). During meiosis, this scaffold is replaced by an axial core through which the chromatin loops. This core eventually forms part of the synaptonemal complex (Moens and Pearlman 1988, Moens and Earnshaw 1989), the structure in the context of which synapsis and recombination occur during meiosis I. In addition, all chromosomes require a kinetochore to which the spindle fibers attach. A distinct entity, called the centromere, holds the sister chromatids together prior to anaphase of mitosis or meiosis II. The kinetochore and centromere are usually located in the same region of the chromosome. Many chromosomes also contain a large amount of repetitive DNA, often associated with centromeric regions. Clark and Baum (1990) have shown that the centromeric regions of Schizosaccharomyces pombe require the associated repeat DNA for proper functioning. Thus repetitive DNA may be utilized by chromosomes to add structural stability during meiosis or mitosis. In contrast, the centromeres of Saccharomyces cerevisiae require less than 200 base pairs for proper function (reviewed in L. Clark 1990). In general, the contribution of repetitive DNA to chromosome structure and function is largely unknown. Telomeres are also an important part of chromosome function during mitosis and meiosis (reviewed in Zakian 1989; Zakian, Runge and Wang 1990). Muller and Herskowitz (1954) showed that terminal deletions in Drosophila were exceptionally rare, and proposed chromosomes required a specialized end, the telomere, for stability. McClintock (1941, 1942) observed the fate 2 of broken chromosomes in maize. The exposed ends were reactive and fused with other broken ends. This is an intolerable situation for an organism because every cell division would be plagued with dicentric bridges. In the sporophyte, however, McClintbck observed the broken ends could be stabilized and permanently healed. Despite rare examples of "stable" chromosomes lacking a telomere, it is generally excepted that chromosomes require a telomere cap in order to avoid shortening and degradation (Zakian 1989). Telomeres have been cloned from S. cerevisiae, a few protozoans, Arabidopsis thaliana and humans (listed in Zakian 1989). In all cases there is a repeat of a short sequence, GT rich on one strand and AC rich on the other. When two homologues are present, these sequences are believed to form four stranded configurations (Sen and Gilbert 1988, 1990; Sundquist and Klug 1989; Williamson, Raghuraman and Cech 1989). The possible role of telomeres in DNA replication was initially discussed by Watson (1972). DNA polymerase synthesizes only in the 5' to 3' direction, thus a telomere is required to prevent incomplete replication and shortening every generation. Two models have been proposed for how telomeres accomplish this task (Zakian 1989). One model states that sequences are added to telomeres de novo by an enzyme (King and Yao 1982). In support of this model, an enzyme (telomerase) has been isolated from ciliates (Greider and Blackburn 1985) and humans (Morin 1989) that adds telomere repeats to an existing telomere sequence without a template. The importance of telomerase in vivo was shown by Yu et al. (1990). They engineered a mutant telomerase in Tetrahymena that added the wrong repeats to the ends of the chromosomes. The result was senescence or cell death. The second model, with its many variations, invokes inter-telomere recombination as the mechanism for the maintenance of chromosome ends. There is experimental evidence in favour of some form of recombination between telomeres (Zakian 1989; Wang and Zakian 1990; Zakian, Runge and Wang 1990). In vivo, both models may be applicable. Chromosomes are often attached at their ends to the nuclear envelope. Although convincing evidence is lacking, telomere attachment to the envelope could mediate the three-dimensional organization of the nucleus. This is an old idea. Early in cytogenetic research it was 3 observed that telomeres often associate together at one end of the nucleus while the centromeres are associated together at the opposite end of the nucleus. This has been termed the "Rabl" orientation after its discoverer (reviewed in Fussell 1987). Telomeres may contribute to homologous pairing of chromosomes during meiosis. The Rabl orientation has been proposed to set the stage for meiosis by maintaining the chromosomes in discrete regions of the nucleus. This could eliminate the problem of homologues finding each other if the chromosomes end interphase in essentially random locations. Some authors have proposed the unique structural properties of telomeres makes them suitable loci for initiating meiotic pairing between homologous chromosomes (Sen and Gilbert 1988, 1990; Sundquist and Klug 1989). Extensive cytological observations (reviewed by Rasmussen and Holm 1980; Maguire 1984), however, do not support such a role. Even though evidence supporting this function for telomeres is lacking, their attachment to the nuclear membrane may be utilized by the cell to prepare the chromosomes for synapsis. It is clear that prophase I chromosomes of meiosis form a "bouquet" orientation with the telomeres clustered in one region of the nuclear envelope (von Wettstein, Rasmussen and Holm 1984). The bouquet could be a variation of the Rabl orientation with the telomeres clustered in one area of the nuclear envelope. This clustering may be required to bring the chromosomes close enough for synapsis. The feature of a chromosome receiving the most attention is the gene content. An approach to understanding the gene content of a chromosome is to sequence the genomes of a select group of organisms. This group includes Escherichia coli, Saccharomyces cerevisiae, Drosophila melanogaster, Caenorhabditis elegans, the mouse and human. While sequencing an entire genome is a monumental task, it has begun in E. coli and C. elegans (Roberts 1990). What questions to ask and how to investigate them is one of the reasons the sequencing effort includes a host of "genetic organisms", in addition to human. The challenge of these efforts is to extract the biological information from the DNA sequence. One of the best ways to elucidate the function of a gene is to study the effects on the organism of mutations in that gene. A genetically compiled list 4 of all the genes in an organism will not only be of basic interest, but practical for application to the genome sequencing projects. A feature of eukaryotic chromosomes that is more difficult to define is the chromosome's organization. For example, it is known that the expression of a gene can be affected by its chromosomal environment (in Drosophila: Spofford 1976; Daniels et. al. 1986; Clark and Chovnick 1986 and references therein). Whether genes are located close together as a result of selection for certain groups of alleles which segregate together at meiosis or as a remnant of the random shuffling of DNA may be tested as more data are collected regarding the organization of the genes on chromosomes. These questions are being investigated in Caenorhabditis elegans. This nematode was introduced for genetic study by Sydney Brenner (1974; see also Wood 1988) because of its hermaphrodite biology, short generation time and its ability to be stored cryogenically. The small genome size, 100Mb (A. Coulson and J. Sulston, personal communication), and low amount of repeated DNA, 15% (Sulston and Brenner 1974), have made C. elegans attractive for molecular studies. The biological descriptions of this nematode have been very thorough. The complete wiring diagram of the nervous system is known (White et al. 1986). On a similar level of complexity, the entire embryonic and postembryonic lineages are known (Sulston and Horvitz 1977; Sulston et al. 1983). The genetic map of C. elegans is extensive, with more than 800 genes mapped on six chromosomes (Edgley and Riddle 1987). Each autosome has a "cluster" of genes (Brenner 1974). It is believed these regions of the chromosomes have less recombination per base pair than the genome average of 330kB per map unit (Brenner 1974; Greenwald et al. 1987; Prasad and Baillie 1989; Starr et al. 1989). More recently, almost the entire genome has been physically mapped using overlapping cosmids and yeast artificial chromosomes (Coulson et al. 1986, 1988). The sequencing of these clones is underway. As the sequencing progresses, there will be a need to ascribe functions to various components of the C. elegans sequence. Both genes and structural elements of each chromosome are expected to have important sequence components. In order to help elucidate the function of 5 these sequences, it will be important to have genetic descriptions of the C. elegans chromosomes at two levels. The first level is the gene. Genetic analysis can be used to identify how many genes there are, where they are located in the genome, and can begin to give information as to the function of the genes. The second level is the structural organization of the chromosome. Genetic analysis can identify important structural elements of a chromosome such as a centromere. In addition, genetic analysis can determine the sequences required for certain aspects of chromosome behavior, such as segregation and meiotic recombination. My project used chromosome rearrangements to investigate these two levels of chromosome organization. On the first level, chromosome rearrangement breakpoints were used to map mutations. Because rearrangement breakpoints can be positioned on both the genetic and physical maps, they can be used to localize genes identified through mutation to small regions of the physical map. This facilitates using the mutations to understand the functions of the coding regions in the C. elegans genome. On the second level, rearrangements were used to study chromosome behaviour. Chromosome fragments were isolated and their meiotic and mitotic properties were studied. In this way, insights were obtained into components essential for the normal behavior of C. elegans chromosomes. 1.1 O r g a n i z a t i o n o f e s s e n t i a l genes . It is no longer sufficient to study the genetic basis of development and cell biology with a select group of genes identified with biased criteria. For example, extensive research has described the properties of the genes controlling segment identity in Drosophila melanogaster. These genes are believed to regulate another set of genes that carry out the processes of morphogenesis (effector genes). The identities of the genes in this effector group, however, are not known (Gehring and Hiromi 1986; Duncan 1987; Beachy 1990). It is likely that many of these downstream genes provide functions used by many different cell types for a variety of functions. Thus, they may not be identified in specialized mutant screens or through sequence homology mediated cloning (see Brenner et al. 1990). In the analysis of developmental mutants in C. 6 elegans, some of the genes identified with mutations disrupting vulval development are turning out to have lethal null phenotypes (M. Han and P. Sternberg, personal communication). Even the genes with so-called redundant functions (Ferguson and Horvitz 1989) appear to have additional unique functions essential for C. elegans viability or fertility. Another method of identifying genes important in development and cellular homeostasis is unbiased searching. This is possible on a molecular level if the entire nucleotide sequence of the genome is available. Relatively unbiased gene searches can also be carried out using genetic analysis. Essential genes can be identified using screens designed to recover recessive lethal mutations. These complementation groups represent the most diverse group of genes. Essential genes have been identified in all genetically studied organisms, including man. Intensive mutant hunts usually require an organism suitable for mass culturing and with rapid generation times. Two metazoans that fit this criteria are C. elegans and D. melanogaster. The rationale for choosing essential genes as targets of a mutant hunt is the hypothesis that 80-90% of all genes in the genome can be mutated to give a recessive lethal phenotype. The remaining classes of genes have non-lethal mutant phenotypes. These classes include the morphological mutants (e.g. Unc and Dpy) and those with no detectable phenotype because they are a member of a gene family or they are not essential under laboratory conditions. Evidence supporting this assumption comes from studies in D. melanogaster (Lefevre and Watkins 1986) and C. elegans (Meneely and Herman 1979, 1981; Rose and Baillie 1980; Moerman and Baillie 1982; Sigurdson, Spanier and Herman 1984; Rogalski and Baillie 1985; Howell et al. 1987; Howell and Rose 1990; Rosenbluth et al. 1988; Rogalski and Riddle 1988 and see Discussion). In yeast, the question has not been adequately investigated. The sDp2 system has facilitated the rapid isolation of over 500 lethal mutations on the left half of chromosome I. As described by Howell et al. (1987), free duplications as lethal balancing systems have a number of advantages over other systems. Unfortunately, there was no efficient system for mapping these mutations and placing them into complementation groups. Unlike some other C. elegans regions, few rearrangement breakpoints existed on chromosome I that could be 7 used to order the lethal mutations on the genetic map. One of the goals of my thesis was to generate a large number of chromosome breakpoints that could be used to map mutations in the region. The poor viability of strains carrying the one deletion that did exist in the left half of chromosome I (Rose 1980; Howell et. al. 1987) made duplication breakpoints the rational choice for dividing up the region. In this thesis, I describe a large number of new chromosome breakpoints that were isolated as gamma radiation-induced shorter derivatives of pre-existing duplications. Two duplications were used to isolate the derivatives. The first duplication, sDp2(I;f), was initially characterized by Rose, Baillie and Curran (1984). The second duplication is a chromosome segregated from a reciprocal translocation strain. This chromosome, ILXLszTl (see Results), is one half of the szTl(I;X) translocation (Fodor and Deak 1985) and is used here in addition to the euploid genetic background. ILXLszTl is larger than sDp2 and is attached at one end to a fragment of the X chromosome. Like sDp2, I^X^szTl does not recombine with the normal chromosome I. In this study, I used gamma radiation to isolate 62 new chromosome I duplications. Most of these duplications are free; that is they are not translocated to another chromosome. In addition, I isolated three new deficiencies among a set of 31 gamma radiation induced lethal mutations. Nine deficiencies of the region were isolated by other workers. These rearrangements were characterized with respect to the genetic position of their breakpoints and their meiotic/mitotic properties. The duplication breakpoints, in combination with 18 deficiency breakpoints, potentially divide the left half of chromosome I into 0.1 m.u. intervals. All 550 sDp2 - rescued lethal mutations isolated by the Rose laboratory have now been mapped to a subset of the duplications (Howell 1989; McDowall 1990; A. Rose unpublished results; this thesis). These mutations have identified approximately 80% of the essential genes in the sDp2 region. Most of the transcribed regions of the sDp2 region are thus represented in this set of lethal mutations. These duplications will facilitate the alignment of the physical and genetic maps so the mutations can be assigned to individual coding regions. 8 1.2 C h r o m o s o m e b e h a v i o u r . Mitosis and meiosis are the evolutionary conserved processes by which all eukaryotic cells reproduce themselves. I have used chromosome rearrangements to study mitosis and meiosis in C. elegans. In particular, I have studied the influence of chromosome structure on chromosome behaviour during meiosis and mitosis. Unique to meiosis is the process of homologue pairing and high rates of recombination. The pairing of homologous chromosomes has been extensively described using cytogenetic techniques (for review, Moens 1987; Giroux 1988). The replicated chromosomes are immediately attached to an axial core. The next step is the poorly understood process whereby the homologous chromosomes align side by side in the nucleus. As this occurs, synapsis of the chromosomes proceeds by parallel alignment of the axial cores. When this is complete, the homologous chromosomes are tightly held together along their entire length. The structure holding them together is called the synaptonemal complex and is composed of the axial core, then called the lateral elements, and a structure linking the two lateral elements, the central element. When first formed during zygotene, the synaptonemal complex is limited to homologous synapsis. Later in pachytene, however, this constraint is relaxed and the complex can form between non-homologous chromatin (reviewed in Moses, Dresser and Poorman 1984; von Wettstein, Rasmussen and Holm 1984). It is in the context of the synaptonemal complex that the high levels of recombination observed during meiosis occur. The first cytological manifestations of recombination are the "recombination nodules" (Beyers and Goetsch 1975; Carpenter 1975; reviewed in Carpenter 1988). Later in meiosis, when the synaptonemal complex dissolves, crossover events become visible as chiasmata. Crossing-over is a tightly regulated process (Jones 1987; Carpenter 1988). Every bivalent has at least one chiasma and the distribution of exchanges among the chromosomes is narrower than predicted by the Poisson distribution. In addition, the location of the crossover events are nonrandomly located on the chromosomes. Recombination is often reduced close to centromeres and telomeres (Beadle 1932; Hawley 1980; reviewed in Jones 1984). 9 Szauter (1984) mapped sites distributed throughout the D. melanogaster X-chromosome which influenced the distribution of recombination events. He also suggested the pattern of recombination events was a reflection of chromatin structure or events related to synapsis. In many systems, exchange events and the resulting chiasmata, facilitate the proper disjunction of the homologues (Darlington 1932; reviewed in Hawley 1988). When the synaptonemal complex dissolves at diplotene, the chiasmata probably serve to keep the bivalents attached until anaphase. It is interesting to note that recombination is not essential to this process. In some organisms one sex is achiasmatic (reviewed in White 1973). The usual solution to the problem of how to keep the homologues together is to maintain the synaptonemal complex until anaphase (von Wettstein, Rasmussen and Holm 1984). Another achiasmatic situation is found in the Drosphila male where there is no synaptonemal complex. In this case disjunction is mediated by specific chromosomal sites (reviewed in Hawley 1988). Finally, there are systems in D. melanogaster (Grell 1976) and S. cerevisiae (Dawson, Murray and Szostak 1986; Mann and Davis 1986) which are present in addition to recombination mediated systems but do not require homology or recombination for disjunction. Of substantially more controversy is the role that genetic recombination plays in the pairing and synapsis of chromosomes during prophase I. Of particular interest is the observation that gene conversion events occur at least as frequently, and sometimes more often, as reciprocal exchanges. One explanation for the excess of gene conversion events is that, because isomerisation in favour of reciprocal exchange is an inefficient process, the cell needs an excess of initiation events to insure that enough result in reciprocal exchanges (reviewed in Carpenter 1987). In contrast, the "excess" gene conversion events may play an essential role in the process by which homologous chromosomes correctly align in register (Smithies and Powers 1986; Powers and Smithies 1986; Carpenter 1987). The proposal is that close alignment of homologous chromosomes, culminating in synaptonemal complex formation and reciprocal exchange, is mediated by simple gene conversion events detecting DNA sequence identity. Recent data from S. 10 cerevisiae are consistent with this hypothesis (Alani, Padmore and Kleckner 1990; Engebrecht, Hirsch and Roeder 1990). The powerful genetics of C. elegans make it a good choice for studies on mitosis and meiosis. A notable difference from the other organisms mentioned is that C. elegans mitotic chromosomes are "holokinetic" (Albertson and Thomson 1982); at anaphase the microtubules of the spindle are attached to a kinetochore which extends the entire length of each chromatid. The C. elegans kinetochore has the same structure as in other organisms. When viewed under the electron microscope, the kinetochore has unique staining properties and is composed of a tripartite structure (Albertson and Thomson 1982). Organisms with holokinetic chromosomes are rare but can be found in a wide variety of species. Some organisms have "polykinetic" chromosomes; these have multiple but discrete spindle attachment sites. In addition to many Gout not all) nematodes (Triantaphyllou 1987), holokinetic chromosomes are found in insect orders Heteroptera, Homoptera and Lepidoptera and in some plants (White 1973). The existence of holokinetic chromosomes has been proposed as the reason why free duplications are so readily isolated in C. elegans. The fact that non-overlapping segments of the X-chromosome could be isolated as free duplications was also attributed to the holokinetic structure (Herman, Madl and Kari, 1979). There has been no explanation, however, as to how these duplications repair their exposed ends, if they do at all. I have used duplications to analyze mitotic properties of C. elegans chromosomes. For example, what are the features of a holokinetic chromosome which contribute to its segregation stability? How is a holokinetic chromosomes different from a monocentric chromosome. Is the kinetochore truly diffuse, or are there discrete sites influencing chromosome stability, and if so, do these sites correspond to sites of spindle attachment? I isolated chromosome fragments of various sizes and compared their segregation stabilities. The size of the duplications was determined using genetic mapping techniques. Knowing their structure, I could look for correlations between size and stability. While size is an important factor, it is clear from the data that other structural 11 factors are also important. In addition, I found that duplications spontaneously break, the frequency of which varies with the stability of the duplication. While a holokinetic structure poses no significant problems for mitotic chromosomes, a similar configuration does present problems for meiotic chromosomes. Microtubules attached along the entire length of the chromosomes at meiosis I could make resolving of chiasma very difficult. Theoretically, a meiotic chromosome must have a discrete centromere. This is the site where the sister chromatids are held together until meiosis II. Thus a holokinetic chromosome could have two sites; one for the centromere, and a second, perhaps diffuse site, for spindle attachment. Recent studies of the nematode Parascaris univalens have partially solved the problem. Goday and Pimpinelli (1989) have shown that the kinetochore structure is different between somatic and germ cells. As opposed to somatic cells, the microtubules attach to the ends of the meiotic chromosomes. A standard kinetochore structure is not visible, the microtubules appear to insert directly into the chromatin. This region is heterochromatic, is comprised of AT rich satellites and comprises 85% of the genome. It is this region which is disposed of during chromatin diminution. Similar observations have been made with C. elegans (D. Albertson, personal communication) with the notable exception that C. elegans does not have vast amounts of satellite DNA or heterochromatin. Chromosome pairing and synapsis can be studied using rearrangements and mutations which disrupt the normal patterns of recombination. Numerous authors have addressed the problem of how the homologous chromosomes find each other in order to undergo synapsis and recombination (for example Comings and Riggs 1971; Burnham 1972; Holliday 1977; Hawley 1980; Chandley 1986; Carpenter 1987; Giroux 1988). A similar and controversial problem concerns the mechanisms regulating the frequency of recombination along the chromosome and how this is related to the mechanisms of homologue pairing (Jones 1984; Szauter 1984; Giroux 1988). While the control of the number and distribution of recombination events is under genetic control, the time of action of these controls is not known. The frequency and selection of recombination sites could be decided as a part of homologue synapsis, or it could occur after and be 12 independent of synapsis. These problems can be studied by observing the effect that meiotic mutations and chromosome rearrangements have on homologous pairing and the distribution of recombination events. Crossover suppression associated with chromosome rearrangements in maize and Drosophila has been attributed to the failure of the chromosomes to pair. These rearrangements provide a means to study events leading to synapsis and recombination of chromosomes. In Drosophila, translocation heterozygosity usually results in a reduction in cross-over frequency (Dobzhansky 1931; Kossikov and Muller 1935; Roberts 1970; reviewed in Roberts 1976; Hawley 1980). Dobzhansky (1931) proposed that crossover reductions were the result of competitive pairing between the partial homologues. Roberts (1970) concluded that crossover reductions in translocation heterozygotes were due to disturbed pairing, and not due to elimination of crossover strands. Hawley (1980) found evidence for the existence of specific chromosomal sites required for pairing and recombination. In each of the above cases, some aspect of defective pairing has been implicated as the cause of the failure to recombine. Studies in maize (Burnham et al. 1972) and mouse (Moses, Dresser and Poorman 1984) have also found defects in pairing and crossover reductions associated with chromosome rearrangement heterozygotes. Studies on human meiosis have benefited greatly from improved cytogenetic techniques and new molecular genetic mapping techniques (Chandley 1988). The pairing of the X and Y chromosomes in the male is an ideal model system for studying meiotic processes because the region of homology and crossing over is restricted to a small portion of each chromosome. These studies show that meiosis in humans is strikingly similar to that in other organisms. The progress of the chromosomes through the stages of meiosis is a highly conserved process; it is virtually the same in organisms as diverse as plants, invertebrates and humans. V In C. elegans the reduction in crossing over in translocation heterozygotes is usually extensive, resulting in severe reductions of exchange along one end of a chromosome (Herman 1978; Rosenbluth and Baillie 1981; Herman, Kari and Hartman 1982; Ferguson and Horvitz 1985). How the recombination suppression corresponds to translocation breakpoints has been 13 examined in only a few cases because it is not possible to map breakpoints cytologically in C. elegans. For both translocation chromosomes in eTl(III;V) (Rosenbluth and Baillie 1981) and mnT2(II;X) (Herman, Kari and Hartman 1982) heterozygotes, recombination was suppressed on only one side of the translocation breakpoint. These authors showed that the crossover-suppressed regions lacked chromosomal features essential for meiotic homologue pairing. The other portion of the translocations recombined and segregated from its normal homologue and therefore possessed sequences enabling it to engage in meiotic pairing and recombination. Rosenbluth and Baillie (1981) pointed out these results demonstrated the nonequivalence of the two halves of each chromosome in C. elegans. Rose, Baillie and Curran (1984) extended these studies with an analysis of two duplications of chromosome I, sDpl(I;f) and sDp2(I;f). The duplications were of similar size but covered opposite halves of chromosome I. Their meiotic properties were, however, dramatically different. Only sDpl(I;f) paired and recombined with the normal homologues. These authors proposed there were sites on chromosome I which enabled the homologues to pair for recombination. It was proposed that sDpl, and not sDp2, carried enough of these sites to enable it to pair for recombination with a normal homologue. To further the understanding of the localized chromosome features essential for meiotic pairing and recombination in C. elegans, I studied the meiotic properties of four translocations and a variety of other rearrangements of chromosome I. In all four translocations, the boundary of crossover suppression corresponded to the position of the breakpoint. The chromosome rearrangements were used to genetically map one region of chromosome I which is required for homologue pairing and recombination. I also studied the effects that pairing disruptions had on recombination frequency in the genome. While I found no evidence for interchromosomal effects of crossover suppressors, there were intrachromosomal enhancements of recombination, suggesting there exists a mechanism ensuring at least one recombination event per bivalent. 14 2. Materials and Methods 2.1 General C. elegans is a self-fertilizing hermaphroditic (5AA;XX) species generating males (5AA;XO) spontaneously at a frequency of approximately 1/700 at 20°C (Rose and Baillie 1979). These males result from X-chromosome nondisjunction (Hodgkin, Horvitz and Brenner, 1979). Wild-type and mutant strains were maintained and mated on Petri plates containing nematode growth medium (NGM) streaked with Escherichia coii OP50 (Brenner 1974). Unless otherwise noted, experiments described in this thesis were conducted at 20° Celsius (Rose and Baillie 1979). The nomenclature follows the uniform system adopted for C. elegans (Horvitz et al. 1979). Names of genetic loci are abbreviated with a three letter code followed by a number. For example, dumpy genes, mutation of which give the worms a short and fat morphology, are abbreviated "dpy". There are more than 25 dpy genes which are distinguished by a number. The most commonly used dpy gene in this thesis, dpy-5, is the fifth dumpy identified. Mutations are given a number following a one or two letter designation of the laboratory in which it was isolated. The designation for Dr. A. M. Rose's laboratory is h. For example, h85 is the eighty-fifth mutation originating in the Rose laboratory. Finally, strains are given names based on their laboratory of origin. This designation is composed of a two letter code written in capital letters followed by a number. The Rose laboratory strain designation is KR. For example, KR900 is a strain constructed in the Rose laboratory. Nomenclature of rearrangements has been modified since the Horvitz et al. (1979) paper. Mutations on translocation chromosomes (T) are shown in square brackets. A strain with both unc-29 and lon-2 on the szTl(I;X) chromosomes is written dpy-5 unc-29; + l szTl(I;X)[+ unc-29; lon-2]. The order of genes is the same as they would be on the normal chromosomes; in this case chromosome I first, the X-chromosome second. The nomenclature provides no information on segregation patterns of the translocation chromosomes. When discussing both components of a reciprocal translocation, the formal name {e.g. szTl(I;X) or hTl(I;V)) is used. When discussing the individual component chromosomes of a translocation, the nomenclature describes the 15 Table 1 Abbreviations Abbreviation Description bli Mutant alleles of gene produce a blistering cuticle phenotype. dpy Mutant alleles of gene produce a short and fat (dumpy) phenotype. dpy-5(+) Wild type allele of dpy-5 gene. let Mutant alleles of gene produce a lethal or sterile phenotype. lev Mutant alleles of gene produce a levamisole resistant phenotype. spe Mutant alleles of gene produce a sperm defective phenotype. unc Mutant alleles of gene produce an uncoordinated phenotype. Wt Wild-type Dpy-5 Phenotype of dpy-5 homozygote (short and fat) Him Phenotype of High Incidence of Males m.u. map units C.I. 95% confidence interval T Translocation Df Deficiency Dp Duplication lL Left portion of chromosome I. IN Normal complete chromosome I. 16 structure of the new chromosomes. For example, szTl(I;X) = ILXLszTl + IRXRszTl and hTl(I;V) = IRVLhTl + ILVRhTl (as shown in Figure 13). The IRXR terminology implies that the new chromosome carries the right portion of the X chromosome joined to the right portion of chromosome I. The "N" superscript signifies the normal chromosome. Duplications are designated with "Dp". The chromosomes involved in the duplication are written in parentheses and if the duplication is free (that is not linked to an intact chromosome), an f is included. If a mutation is on the duplication chromosome, it is written in square brackets. sDp2(I;f)[dpy-5(h585)] is an sDp2 chromosome on which a dpy-5 mutation had been induced. Other abbreviations used in this thesis can be found in Table 1. 2.2 Strains. Many strains used in this study were generous gifts from other workers. The wild-type strain N2 and some mutant strains of C. elegans var. Bristol were obtained from D. L. Baillie, Simon Fraser University, Burnaby, Canada or from the Caenorhabditis Genetics Center at the University of Missouri, Columbia. The following mutant genes and alleles were used: I dpy-5(e61); dpy-5(h660); bli-3(e767); bli-4(e937); dpy-14(el88); egl-30(n686); egl-30(n715); fog-l(q253ts); spe-8(hc50); spe-ll(hc90); lev-ll(xl2); Un-6(el466); Un-17(n671); sup-ll(n403n682); unc-ll(e47); unc-13(e450); unc-14(e57); unc-15(e73); unc-29(e403); unc-35(e259); unc-37(e262); unc-38(e264); unc-40(e271); unc-54(el90); unc-55(e402); unc-57(e406); unc-63(e384); unc-73(e936); unc-74(e883); unc-74(xl9); unc-75(e950); unc-87(el216); unc-89(el460); unc-lOl(ml). II unc-52(e444). III dpy-17(el64); dpy-18(e364); unc-36(e251); unc-45(e286ts); unc-64(e246). IV dpy-9(el2); dpy-13(el84sd); unc-22(s7). V dpy-ll(e224); unc-42(e270); unc-51(e369); unc-60(m35). X dpy-3(e27); dpy-7(e88); dpy-7(sc27); dpy-8(el30); Un-15(n309); lon-2(e678); unc-l(e719); unc-3(el51); unc-7(e5); unc-9(el01); unc-20(ell2). Table 2 Mutations i s o l a t e d i n other studies. S t r a i n Genotype Method Origin BA00662 spe-11(hc90) dpy-5; sDp2(I;f) EMS L'Hernault et al. 1988 BC00070 eTl (III/V) P32 Rosenbluth and B a i l l i e 1981 BC00159 dpy-5 unc-13; sDpl 7000R gamma Rose, B a i l l i e and Curran 1984 BC03003 eDfl9 unc-22;+/nTl(IV;V) Clark 1990 BC03175 sDf2;+/nTl(IV;V)[let-(m435)] formaldehyde Moerman and B a i l l i e 1981 BC03176 dpy-13 mDI7; +/nTl(IV;V)[let-(m435)] 1500R gamma Rogalski and Riddle 1988 BC03314 sDf60;+/nTl(IV;V)[let-(m435)] formaldehyde Clark 1990 BC03355 unc-60 sDf50 unc-46; +/eTl(III;V) formaldehyde Johnsen and B a i l l i e 1988 BC03737 sDp8; eTl UV Stewart et al., per. comm. BW01102 dpy-5 mei-2(ctl02) unc-29; sDp2 P. Mains, per. comm. CB01256 him-3(el256) EMS Hodgkin et al. 1979 CB01479 him-6(el423) EMS Hodgkin et al. 1979 CB02769 eDf3/eDf24 DEO 2mM 1 Anderson and Brenner 1984 CB02770 eDf4/ eDf24 DEO 2mM Anderson and Brenner 1984 CB02772 eDf6/ eDf24 DEO 2mM Anderson and Brenner 1984 CB02773 eDf7/eDf24 DEO 2mM Anderson and Brenner 1984 DR00684 mDf9;+/nTl(IV;V) 1500R gamma Rogalski and Riddle 1988 Table 2 (con't). S t r a i n Genotype Method O r i g i n DR00799 mDf4;+/nTl(IV;V) 1500R gamma Rogalski and Riddle 1988 DR00814 dpy-13 ama-1 mDf8;+/ nTl(IV;V)[let(m435); +] 1500R gamma Rogalski and Riddle 1988 GE01385 tD£3 dpy-5; unc-3/ szTl(I;X) 40mM EMS R. Feichtinger (per. comm.) KR00016 sDp2 / unc-11 dpy-5 / unc-11 dpy-5 7000R gamma Rose, B a i l l i e and Curran 1984 KR01001 sDf6; +/ szTl(1;X) formaldehyde Rose and B a i l l i e (1980) KR01011 sDf5; +/ szTl(I;X) EMS Rose and B a i l l i e (1980) KR01053 lin-17(n671) dpy-5 EMS Ferguson and Horvitz 1985 KR01062 spe-8(hc50) dpy-5; unc-3/ szTl(I;X) EMS L'Hernault et al. 1988 KR01069 sDf4; +/ hTl(I;V)[+; unc-42] formaldehyde Rose (1980) KR01233 hDf8/ dpy-5 unc-13 formaldehyde T. Starr, A. Rose(per. comm.) KR01234 hT2(I;III)[bli-4(e937);+] 1500R gamma K. Peters (per. comm.) KR01255 hT2(I;III)[dpy-5(h660) bli-4(e973);+] EMS J. Babity (per. comm.) KR01275 hT2 (I;III) [bli-4 (e937) ; dpy-18 (h.662) ] EMS K. Peters (per. comm.) KR01458 unc-11 dpy-14/ szTl(I); +/ szTl(X) X-Ray 7500R Fodor and Deak 1985 KR01569 hT2(I;III)[bli-4 unc-29(hl011);+] EMS A. M. Howell (per. comm.) KR01695 dpy-5 nDf23; +/ hTl (I;V)[unc-29;+] gamma Ferguson and Horvitz 1985 KR01696 dpy-5 nD£25; +/ hTl(I;V)[unc-29;+] gamma Ferguson and Horvitz 1985 KR01713 dpy-5 nD£24; +/ hTl(I;V)[unc-29;+] gamma Ferguson and Horvitz 1985 Table 2 (con't). S t r a i n Genotype Method O r i g i n KR01828 fog-1 (q253ts) dpy-5; him-5 (el490) EMS Barton and Kimble 1990 KR01857 sup-11(n403n682) dpy-5 unc-29; sDp2 EMS Greenwald and Horvitz 1983 KR01912 qDf3 dpy-5;+/szTl(I;X) 7500R gamma Barton and Kimble 1990 KR01937 dpy-11; sDp30 1500R gamma Rosenbluth et al. 1988. KR01948 dpy-5 nDf(n2088);+/szTl(I;X) EMS M. Basson (per. comm.) MT01442 lin-6(el466) dpy-5;+/ szTl(I;X) Horvitz and Sulston 1980 MT02936 unc-13; nDp4 J . Thomas (per. comm.) MT? dpy-5(e61) sem-4(nl378) EMS M. Basson (per. comm.) SP00075 mnDp25; unc-3 X-Ray Herman, Madl and Kari 1979 SP00115 mnDp8; unc-3 X-Ray Herman, Madl and Kar i 1979 SP00268 mnDf7; mnDpl X-Ray Meneely and Herman 1979, 1981 SP00272 mnDfl1; mnDpl X-Ray Meneely and Herman 1979, 1981 SP00278 mnD£20; mnDpl X-Ray Meneely and Herman 1979, 1981 SP00423 let-2(mnl53) unc-3; mnDpl(X;V) EMS Meneely and Herman 1979, 1981 SP00511 mnDf43; mnDpl X-Rays Meneely and Herman 1979, 1981 SP00580 mnl64(I;X) X-rays Herman, Kari and Hartman 1982 SP00957 dpy-8 stDfl; mnDp30 R. Waterston, per. comm. Or i g i n r e f e r s to the study i n which the rearrangement or mutation was i s o l a t e d . In many cases I constructed new s t r a i n s from these mutations; these are designated with a KR number. 18 Chromosome I essential genes let-351 - let-392 were first reported by Howell et al. (1987). Chromosome I essential genes let-75(sl01); let-76(s80); let-80(s96), let-81(s88), let-82(s85), let-85(sl42), let-86(sl41), let-87(sl06), let-88(sl32), let-89(sl33) and let-90(sl40) were isolated by Rose and Baillie (1980). Essential genes at the right end of chromosome I, let-201(el716), let-202(el720), let-204(el719) and let.-208(el718), were isolated by Anderson and Brenner (1984). More strains from other laboratories are listed in Table 2. A genetic map of the C. elegans genome is shown in Figure 1. This map represents a summary of the rearrangements and regions of the genome which were studied. Greater detail of particular regions can be found in the pertinent section of the Results (see Table of Figures). In general, only one male stock, N2, was maintained. In most cases the required males for an experiment could be generated by mating a hermaphrodite stock to either N2 males or males acquired from an szTl(I;X) stock. szTl(I;X) is a reciprocal translocation between the chromosomes I and X. Strains containing szTl were very useful because they spontaneously generated males due to X-chromosome nondisjunction (section 3.6.1). These males were frequently used to generate duplication or translocation males. When szTl heterozygous males were crossed to non-translocation hermaphrodites, only two of the four types of sperm produced viable progeny. These two were I^;0 and I^X^szTl; IRXRszTl, the other two produced lethal zygotes due to aneuploidy. Thus all the wild-type males from crossing dpy-5;0/szTl males to dpy-5 dpy-14; hDpx hermaphrodites were dpy-5 + /dpy-5 dpy-14; hDpx, if hDpx was dpy-5(+). All the wild-type male progeny from crossing hT2(I;III)[dpy-18]/szTl(I;X) males to let-x dpy-5 unc-29/+;hT2(I;III)[dpy-18] hermaphrodites were let-x dpy-5 unc-291 + ;hT2(I;III)[dpy-l8] males. hT2(I;III) is a reciprocal translocation between chromosomes I and III (section 3.6.3). Generation of duplication males. The most common method to generate males was to cross duplication carrying hermaphrodites to szTl(I;X) males. For example, to generate hDpl2 males, dpy-5 dpy-14; hDp!2 hermaphrodites were crossed to dpy-5; 01 szTl males. Wild-type males from this cross carried the duplication and were used for subsequent crosses. All male •101 J .76 -.29 _ -13-.5 -•11 H T e D f 3 _L nDf25 sDf4 _L nDp4 sDp2 19 0 0 ° ' an0' •3 H l ) 0 C -9 dpv -7 J 2o J l i O c ' TT mnDpl T SfD/7 J _ -6* dpv -18 H dpv -36 H •17 H U 0 C -93 H - 4 6 H sDp8 31 i •22 -8 dP* -19 J dP*' J _ -61 dpv ,42 H 11 H 0 ^ ° -60 H TT sDp30 IL sDf50 1 Figure 1: Simplified map of the C. elegans genome showing markers and regions studied in this thesis. Duplications are drawn with double lines. Deficiencies are drawn with single lines. 20 progeny sired by szTl heterozygotes would be expected to have a Dpy-5 phenotype unless they received the duplication, which carries the wild-type allele of dpy-5. 2.3 Isolation of new rearrangements. Three mutagenesis procedures were adopted using gamma radiation as the mutagen. The first was a general screen for lethal mutations using the translocation hT2 as the balancer. The second was a screen designed to recover deletion derivatives of pre-existing duplications. The third was a screen designed to detect translocations based on pseudolinkage. The protocol for the screens using pre-existing duplications are described in the Results section. In all screens, gravid worms were treated with either 1500 or 3000 rad of gamma radiation from a Cobalt 60 source (see Rosenbluth, Cuddeford and Baillie 1985). The dose rate varied from 7.5 to 9.0 rad/sec. The screen for lethal mutations was done using strain the KR1261 (dpy-5 + unc-29;-!- lhT2(I;III)[+ bli-4 +; dpy-18]. Radiated wild-type L4 and gravid hermaphrodites were set individually on plates. Their FI wild-type progeny were then set individually on plates and their progeny, the F2, examined for the absence of fertile Dpy-5 Unc-29 hermaphrodites. If this class were absent, it was assumed that a lethal mutation had been induced on one of the two normal chromosomes in the balanced region of hT2 and the strain was retained for further analysis. The new lethal mutations were analyzed to determine which linkage group the lethal mutation was on. This was accomplished through one of two methods. First, some strains were crossed to N2 males and the progeny of the lethal heterozygotes were scored. If linkage was shown to dpy-5 and unc-29, then the mutation was retained; otherwise it was discarded. The second method of testing linkage involved replacing the normal chromosome III in the lethal bearing strains with an untreated chromosome and then determining if the lethal mutation was still present. The protocol for this method is described in Figure 2. Isolation of hTl(I;V): The translocation hTl(I;V) was isolated in a screen for mutations which caused the unlinked genes unc-13(1) and dpy-ll(V) to behave as if they were linked. This is referred to as pseudolinkage. unc-23 +1 + unc-42 (V) males were treated with 1500 rads gamma radiation and mated to unc-13; dpy-11 hermaphrodites. Wild type FI hermaphrodites 21 Viable? Chrom. I m m m m Chrom. Ill . X-Chrom. Figure 2: Protocol for determining linkage of lethal mutations isolated in hT2(I;III) strains. The "*" refers to the mutagenized chromosomes. In the first cross, hermaphrodites from the lethal bearing strain, dpy-5 unc-29 *;+ */ hT2(I;III)[bli-4; dpy-18] were crossed to hT2(I;III)[bli-4 unc-29; dpy-18]/ szTl(I;X)[lon-2] males. The wild-type male progeny were dpy-5 unc-29 *; + / hT2(I;III)[bli-4; dpy-18]. The mutagenized chromosome I was still present but the mutagenized chromosome III was replaced. The mutagenized chromosome I was then placed over a new hT2 chromosome in the next cross. If the wild-type progeny from this cross segregated viable Dpy-5 Unc-29 progeny, the lethal mutation mapped to chromosome III. If no Dpy-5 Unc-29 progeny were produced, then the lethal mutation mapped to chromosome I. The wild-types were retained for further analysis of the lethal mutations. 22 were plated individually and their progeny were examined for lack of independent assortment of dpy-11 and unc-13. One isolate among 2019 FI progeny behaved as a translocation between chromosomes I and V and was named hTl(I;V). Isolation of spontaneous duplications and levamisole selection: Some spontaneously shortened duplications were observed during routine crosses and strain maintenance because of their exceptional phenotypes. More spontaneously shortened duplications were isolated in screens using resistance to 1 mM levamisole as a selective agent to screen a large number of worms. Levamisole is a potent acetylcholine agonist that kills sensitive worms. Lewis et al. (1980) showed Unc-63 and Unc-74 worms are resistant to ImM levamisole. In order to select for spontaneously shortened duplications, two or three unc-74 dpy-5; Dpx or unc-63 dpy-5; Dpx wild-type hermaphrodites were placed on each small Petri plate. When the F2 progeny were larvae, they were washed off in M9 buffer (Brenner 1974) and transferred onto new large (100 X 15mm) plates containing 1 mM levamisole. The plates were prepared as described in Lewis et al. (1980). Two or three days later the plates were screened for levamisole resistant non-Dpy worms. Survivors were placed onto normal plates to observe their progeny. 2.4 Complementation analysis. The novel aspect of duplication mapping protocol was the use of szTl(I;X) males to generate trans -heterozygous lethal bearing males. In a cross of a duplication hermaphrodite to a male of the correct genotype, all the wild-type progeny carried the duplication. At least half of these also carried the mutation to be tested. The pseudolinkage between chromosomes I and X in szTl(I;X) heterozygotes made this possible. Males heterozygous for szTl(I;X) have a Lon-2 phenotype (Fodor and Deak 1985) and carry only the rearranged X-chromosome. These males cannot sire hermaphrodite progeny with a paternal normal chromosome I. For the same reason these males sire only male progeny which carry the paternal normal chromosome I. The Lon males were used to generate males frans-heterozygous for the mutation a to be tested and another mutation b in the duplication strain. The mutation to be tested was linked to a common marker c. 23 Figure 3: Protocol for mapping genes to duplication breakpoints. Lon males from szTl were used to generate phenotypically wild-type males carrying the lethal mutation being mapped. These males were mated to hermaphrodites carrying the duplication being tested;. Wild-type hermaphrodites from this cross were selected and their progeny examined for the presence of absence of Unc-13 hermaphrodites. Viable Unc-13 hermaphrodites indicated that the duplication covered the lethal mutations. 23a Po Unc-13 6 sDp2 d p y - 5 l e t - x unc-13 | d p y - 5 l e t - x unc-13 | X X Wild-type ^ / d p y - 5 l e t - x unc-13 unc-11 dpy-14 unc-11 dpy-14 L O n ( 5 d p y - 5 dpy-14 d p y - 5 dpy-14 szT1 szT1 hDpx Dpy-14 <jT X X F1 d p y - 5 Wild-type hDpx • dpy-14 d p y - 5 l e t - x unc-13 or d p y - 5 hDpx dpy-14 d p y - 5 l e t - x unc-13 F2 hDpx d p y - 5 l e t - x unc-13 d p y - 5 l e t - x unc-13 Arrested Unc-13 hDpx d p y - 5 l e t - x unc-13 d p y - 5 l e t - x unc-13 Unc-13 24 The trans-heterozygous males {a c +1 + + b) were then crossed to duplication strains (c b; Dp[c(+)]). With the exception of rare recombinants, at least half the wild-type hermaphrodite progeny carried both the duplication and the mutation to be tested (ac +1 + cb; Dp[c(+)]). If the Dp carried b(+) the other half would be + c bl + + b; Dp. These two possibilities could be distinguished by the phenotypes of the F2 progeny. If the duplication complemented the mutation a, then a non-mutant a c; Dp[a(+)c(+)] strain could be established. 1) Visible mutation with duplication [Dp(I;f)]. Markers linked to either dpy-5, dpy-14 or unc-13 were used to map the duplication breakpoints (Figure 3). The exception to this was the mapping of Dp[unc-ll(+) dpy-5(-)] duplications. This mapping used visible mutations linked to unc-11. To facilitate mapping, most duplications isolated with non-dpy-5 dpy-14 chromosomes were resegregated into dpy-5 dpy-14; Dp strains. - To map a duplication breakpoint with respect to a visible mutation (e.g. unc-x), the progeny of unc-xl+; Dp(I;f) hermaphrodites were examined to see if their unc-x; Dp progeny were Unc [i.e. Dp is unc-x(-)] or wild type [Dp is unc-x(+)]. Most parental hermaphrodites were constructed by crossing unc-x bearing males to a duplication stock. For example, unc-x dpy-5 hermaphrodites were crossed to unc-11 dpy-14; 01 szTl(l;X)[+ ;lon-2] males. The only resulting male progeny (+ unc-x dpy-5 +1 unc-11 + + dpy-14) were crossed to duplication strains. In crosses to dpy-5 dpy-14; hDpz(I;f) strains, most wild-type hermaphrodite progeny were unc-x dpy-5 + / + dpy-5 dpy-14; hDpz, if the duplication did not cover dpy-14. If the duplication was [dpy-14(+)], then + dpy-5 dpy-14 I unc-11 + dpy-14; hDpz hermaphrodites were also recovered. These possibilities were distinguished by examining the resulting self fertilization progeny. Upon examining the progeny of the desired (unc-x dpy-5 +1 + dpy-5 dpy-14 ; hDpz) hermaphrodites, Dpy-5 and Unc-x progeny were observed if the duplication did not carry unc-x(+). No Unc-x progeny were observed if the duplication carried unc-x(+). The only exceptions to the above procedure resulted from recombinant chromosomes. 25 In crosses of + unc-x dpy-5 + 1 unc-11 + + dpy-14 males to unc-11 dpy-5 ; hDpz(I;f) strains (where hDpz was unc-11(-) dpy-5(+)), most wild-type hermaphrodite progeny were unc-11 + dpy-5/ + unc-x dpy-5 ; hDpz(I;f). For crosses to unc-11 dpy-5 ; hDpz(I;f) strains (where hDpz was unc-11(+) dpy-5(-)), the mutation to be tested was linked to unc-11 and brought in through the male of genotype unc-11 unc-x +1 + + dpy-5. Using similar methodology, dpy-14 unc-13; hDpz[dpy-5(-) unc-13(+)] duplications were crossed to unc-x dpy-14 +1 + + unc-13 males when mapping a visible mutation. In some of the complementation experiments unc-x dpy-5/ + dpy-5; hDpz hermaphrodites were generated by crossing duplication bearing males to homozygous mutant hermaphrodites. For example, hDpl8 males were generated by crossing dpy-5;0/szTl(I;X)[+;lon-2] males to dpy-5 dpy-14; hDpl8 hermaphrodites. Wild-type males from this cross {dpy-5 +/dpy-5 dpy-14; KDplS) were crossed to unc-x dpy-5 hermaphrodites to generate unc-x dpy-5/+ dpy-5; hDpl8 hermaphrodites. 2) sDp2-balanced lethal with duplication /Dp(I;f)7. Lethal bearing hermaphrodites [i.e. dpy-5 let-x unc-13; sDp2(I;f)] were crossed to unc-11 dpy-14; 01 szTl(I;X)[+ ;lon-2] males (Figure 3). The resulting male progeny {+ dpy-5 let-x + unc-131 unc-11 + + dpy-14 +) were then crossed to dpy-5 dpy-14; hDpz hermaphrodites. Male progeny carrying sDp2 had a distinctive phenotype and did not mate. Wild-type hermaphrodite progeny from this cross were dpy-5 let-x + unc-13/ dpy-5 + dpy-14 +; hDpz or + dpy-5 dpy-14 /unc-11 + dpy-14; hDpz if hDpz was dpy-14(+). These two classes were distinguished by examining their progeny. If the duplication carried let-x(+), Dpy-5 and viable Unc-13 progeny were observed. hDp[dpy-5(-) unc-13(+)] strains were tested by generating dpy-5 + let-x unc-13/ + dpy-14 + unc-13; hDpz hermaphrodites and looking for viable Dpy-5 progeny. These hermaphrodites were recovered from crossing dpy-14 + unc-13; hDpz hermaphrodites to + dpy-14 + +/ dpy-5 + let-x unc-13 males. In addition to the unc-13 linked lethals, mei-2, spe-11, fog-1, and sup-11 (Table 2) were complementation tested mapped to the duplications. Mei-2 homozygotes are sterile adults due to defective meiosis. Spe-11 and Fog-1 worms are sterile adults owing to a lack of functional sperm 26 and sup-ll(n403n682) homozygotes arrest development in embryogenesis. In the strain BW1102, mei-2 was linked to dpy-5 and unc-29. The unc-29 marker served the purpose of the unc-13 in the complementation testing of lethal mutations (Figure 3). spe-ll(hc90) and sup-Il(n403n682) were already linked to dpy-5. To facilitate their mapping, unc-29 was crossed onto both chromosomes. For example, Unc-29 worms were picked from spe-11 dpy-5 + + /+ + dpy-14 unc-29 hermaphrodites and the spe-11 dpy-5 unc-29 chromosome was then maintained with sDp2. To map fog-1, fog-1 dpy-5 +1+ dpy-5 dpy-14; hDpz worms were constructed and 10 to 20 wild-type hermaphrodites were set individually on plates. If any of worms were sterile and laid oocytes, the duplication failed to complement fog-1. If any of the wild-types were fertile and segregated only sterile Dpy-5 progeny, then the duplication complemented fog-1. 3) Lethal with lethal or deficiency. These complementation tests required that the two lethal mutations be tightly linked to a common visible marker. Most of the lethal mutations or deficiencies isolated in this lab were linked to dpy-5. Any that were not had a dpy-5 marker introduced onto the chromosome by recombination. For example, dpy-5 was crossed onto the qDf3 chromosome by picking Dpy-5 recombinants from unc-38 dpy-5/ qDf3 heterozygotes. Similarly, dpy-5 was crossed onto the nDf23, nDf24 and nDf25 chromosomes by picking Dpy-5 recombinants from dpy-5 unc-13/ nDfx strains. The dpy-5 Dfx chromosomes were then crossed into balancer strains. The qDf3 dpy-5 chromosome was balanced by szTl(I;X). The dpy-5 nDfx. chromosomes were balanced with hTl(I;V)[unc-29]. Complementation tests were done by crossing one lethal heterozygote to another lethal or deficiency heterozygote and looking for Dpy-5 males and fertile Dpy-5 hermaphrodites in the F l (or Dpy-5 Unc-13 if both lethals were also linked to unc-13). In most cases the heterozygous male, dpy-5 let-x unc-13/ + + +, could be generated by crossing a duplication strain, dpy-5 let-x unc-13; sDp2, to N2 males. If the lethal mutation was loosely linked to dpy-5, a balancer was used. For example, hTl(I;V)/szTl(I;X) males crossed to let-x dpy-5 unc-13; sDp2 hermaphrodites produced balanced let-x dpy-5 unc-13/hTl males. Males heterozygous for the lethal mutation were 27 crossed to hermaphrodites from duplication balanced strains, let-x dpy-5 unc-13; sDp2, or translocation strains such as tDf3 dpy-5;unc-3/szTl. The deficiency hDf8 was isolated by Starr (1990) using formaldehyde in a precomplementation screen for dpy-14 mutations, dpy-5 could not be crossed onto the hDf8 deficiency chromosome because it is associated with a suppressor of chromosome I recombination (section 3.10.1). To allow this deficiency to be used in the general complementation analysis of mutations around dpy-14, a special protocol was developed. This is diagrammed in Figure 4. The heterozygous lethal males were the same genotype as those used in the duplication crosses described above. Therefore, the inclusion of hDf8 into the complementation analysis was simple. When these males were crossed to dpy-5 dpy-141 hDf8 hermaphrodites, the only wild-type progeny would be let-xlhDf8. These wild-types were viable and fertile only if the lethal was outside the deletion. Rare wild-types could also result from recombination in the male, but these were easily identified by progeny testing. 2.5 G e n e t i c a l l y d e t e r m i n i n g the s i z e o f t r a n s l o c a t i o n - d e r i v e d d u p l i c a t i o n c h r o m o s o m e s . ILXLszTl: To map the duplication I^X^szTl with respect to visible mutations, dpy-14; 01 szTl(I;X) males were crossed to dpy-5 unc-z hermaphrodites. All the wild-type male progeny were + dpy-14 + / dpy-5 + unc-z and these were mated to dpy-5 dpy-14; I^X^szTl hermaphrodites. With the exception of rare recombinants, wild-type progeny have to carry the ILXLszTl chromosome in addition to the dpy-5 dpy-14 chromosome. From this cross dpy-5 unc-zl dpy-5 dpy-14; fi-'X^-'szTl hermaphrodites were isolated (along with + dpy-141 dpy-5 dpy-14; I^X^szTl hermaphrodites) and their progeny scored to see if the I^X^szTl chromosome carried unc-z(+). Among the progeny of these hermaphrodites, a large number of Unc-z worms indicated the I^X^szTl chromosome did not carry unc-z(+). To map I^X^szTl with respect to lethal mutations, unc-13 let-x; unc-3/ szTl(I;X) strains were constructed and their progeny observed for the presence of Unc-3 progeny. The presence of the Unc-3 progeny indicated the duplication 28 dpy-5 let-x unc-13; sDp2 C3 X dpy-14;0/szT1 Q wild-type <^  dpy-5 let-x unc-13/ dpy-14 $ X dpy-5 dpy-14/hDf8 (jj I dpy-5 let-x unc-13/ hDf8 wild type Figure 4: Protocol for complementation testing with hDf8. A l l progeny from the second cross were Dpy-5 or Dpy-14, with the exception of dpy-5 let-x unc-13lhDf8, which were wild-type. Fertile wild-types from this cross indicated that the deficiency complemented the lethal mutation. Rare recombinants could have been wild-type, but these were distinguished by progeny testing. 29 carried the wild-type allele for the lethal mutation. If found, the Unc-3 hermaphrodites were confirmed to be unc-13 let-x; unc-31 unc- 31 ILXLszTl by progeny testing. IRVLhTl: To map visibles against the duplication, dpy-5 unc-x/ + +; dpy-11 unc-421 + +; IR V^hTl strains were constructed and their progeny observed for rescue of the unc-x mutation (i.e. Dpy-5 progeny produced). The dpy-11 unc-42 (V) chromosome was present to monitor the presence of the duplication {IRV^hTl is unc-42(-)). Lethal mutations were mapped in a similar fashion except the strain used was unc-13 let-x/ + +; dpy-11 unc- 421 + +; IRVLhTl. hDplOl was isolated from a dpy-5 unc-29; IRV^hTl strain. To show it carried chromosome V markers, +; dpy-11 unc-42/hTl males were crossed to dpy-5 unc-29; hDplOl and dpy-5 unc-29/ + +/ hDplOl; dpy-11 unc-421+ + strains were isolated. Unc-42 worms were then picked {hDplOl; dpy-11 unc-42). The Unc-42 hermaphrodites were crossed to +; unc-60 dpy-11/hTl males and unc-60 dpy-11 +1 + dpy-11 unc-42; hDplOl hermaphrodites were isolated. No Unc-60 progeny were seen in the progeny of these worms. Furthermore, wild-type progeny of the genotype unc-60 dpy-11; hDplOl were produced. Thus hDplOl, like IRVLhTl, complemented unc-60 and dpy-11 but not unc-42. IRIIIRhT2: Strains of the genotype dpy-5; unc-36; IRIIIRhT2 were Dpy but not Unc, indicating IRIIIRhT2 carried unc-36(+). The Dpy hermaphrodites were crossed to +; dpy-17 unc-36/hT2 males, and the dpy-5/+; -f unc-36/ dpy-17 unc- 36; IRIIIRhT2 progeny were scored. These did not segregate any Dpy-17 progeny. In addition, wild-types of the genotype dpy-17 unc-36; IRIIIRhT2 were recovered, showing IRIIIRhT2 carried dpy-17(+). To test iURIIIRhT2 carried unc-64(+), the dpy-17 unc- 36; IRIIIRhT2 hermaphrodites were crossed to +1 dpy-17 unc-641 hT2 males, and the dpy-17 + unc-64/dpy-l 7 unc-36 +; IRIIIRhT2 were scored. These did not segregate Unc-64 progeny, and wild-type dpy-17 unc-64; IRIIlRhT2 hermaphrodites were recovered. These data showed IRIIIRhT2 carried sequences from dpy-17 to the right end {unc-64) of chromosome III. ILXLhT3: The left end of ILXLhT3 was determined by crossing unc-11 dpy-5; ILXLhT3 hermaphrodites to bli-3 unc-11 +; +1 hT2(I;III)[+ + dpy-5; + ] males. The wild type 30 hermaphrodite progeny were bli-3 unc-11 + 1 + unc-11 dpy-5; I^X^hT3. These did not segregate non-Unc-11 Bli-3 progeny, indicating I^X^hTS carried bli-3(+). To show the duplication carried material from the left end of the X-chromosome, unc-11 dpy-14/ unc-11 dpy-14; ILXLhT3 hermaphrodites were crossed to unc-11 dpy-5 +/ + + dpy-14; unc-11 hDp31(I;X)[+; + ] males. Some of the wild-type progeny from this cross were unc-11 dpy-5 + / unc-11 + dpy-14; unc-1 I +; ILXLhT3. No Non-Dpy Unc-1 progeny were observed segregating from these wild types as would be expected if I^X^hTS did not carry the unc-1 ( + ) region. In a similar cross, + dpy-5 +1 unc-11 + dpy-14; unc-3/O; mnDp25 males were crossed to unc-11 dpy-5; I^XRhT3 hermaphrodites and unc-11 + dpy-141 unc-11 dpy-5 +; I^XRhT3; unc-31+ progeny were recovered. If I^XRhT3 carried unc-3(+), then these hermaphrodites would not segregate Unc-3 non-Dpy progeny. Unc-3 progeny would have to be of the genotype unc-11 + . dpy-141 unc-11 dpy-5 +; I^XRhT3; unc-3/ unc-3. Unc-3 worms were observed, however, indicating that I^XRhT3 did not carry unc-3(+). The same experiment was done using dpy-7. unc-11 dpy-5; /Ax^/iT3 hermaphrodites were crossed to + dpy-5/ unc-11 +; dpy-710 males. Wild-type progeny from this cross of the genotype unc-11 dpy-51 unc-11 +; I^XRhT3/dpy-7l'+ were observed. Non-Unc-11 Dpy-7 progeny were not observed. 2.6 Complementation with him strains. Complementation tests were done by crossing deficiency strains of the genotype Df/nTl(IV;V) (Table 2) to him-3 or him-6 males. In most cases the him/Df progeny could be selected because of the dominant expression of unc-22 mutations or dpy-13(el84). This did not apply to mDf9 because this strain did not have either mutation. When him males were crossed to sD2, sDf60, eDfl9 or mDf7/nTl, the him/Df heterozygotes were selected by picking the worms in a 1% nicotine solution. unc-22(-) heterozygotes, and therefore the him/Df heterozygotes, twitched in nicotine (Brenner 1974). In crosses of him males to mnDf4, mnDf7 or mnDfS/nTl heterozygotes, semi-dominant expression of the dpy-13(el84) mutation was used to select the deficiency progeny. 31 2.7 S c o r i n g a n d c a l c u l a t i n g r e c o m b i n a t i o n . S t a n d a r d r e c o m b i n a t i o n m a p p i n g : Recombination frequencies between pairs of markers were determined by scoring the self progeny of cis-heterozygous hermaphrodites under the conditions described by Rose and Baillie (1979) at 20° C. The recombination frequency, p, was calculated using the formula p = 1 - (1 - 2R)1/2, where R is the fraction of marked (non-Wt) recombinant individuals over total progeny (Brenner 1974). Map distances are reported as map units (m.u. = lOOp). For standard crosses, the total number of progeny (the denominator) was calculated as 4/3 X (the number of wild type plus one recombinant class). When the viability of one recombinant class reduced its recovery, the number of recombinant individuals (the numerator) was calculated as twice the number of the more viable recombinant class (Rose and Baillie 1979). Some of the control cross data are in Table 11. T r a n s l o c a t i o n s : When the interval being examined was in cis to a translocation breakpoint, one of the markers crossed onto the translocation chromosome. Assuming alternate and adjacent-a segregation frequencies were equal (see Results) and p was small, examination of the Punnett square (Figure 5) for the heterozygote dpy-5 unc-zl T(I; ?), where dpy-5 was to the left and unc-z was to the right of the translocation breakpoint, shows that recovery of the Unc recombinant was expected at a frequency twice that of the reciprocal Dpy recombinant. As p increased, the ratio of Dpy to Unc recombinants was expected to increase. In these situations, recombination frequency was calculated as; 1) p = (l-(l-[4B/W+B])1/2)/2; where B was the number of recombinant Unc progeny and W was the number of wild types. When the calculation was based on the number of Dpy progeny, then the recombination frequency was; 2) p = [(4A + 2W)-(4W2-48A2)1/2]/2(4A+W); where A was the number of recombinant Dpy progeny. When recombination could occur between the interval being examined and the translocation breakpoint, for example in the unc-101 unc-54lszTl(I;X) experiment, formula 2) had •32 •f o s <i5uZ V N d5uZ j L y R , R V L V N j R v L , L V R dS + V N d5 + , L V R I R V L u Z X N I R V L o Z j L v R d5uZ V N Dpy-5 Unc-Z Wt Dpy-5 Unc-Z d5uZ j L v R Wt Unc-Z jRyL V N Wt Wt , R V L , L V R Wt Wt d5 + V N Dpy-5 Wt Dpy-5 Wt d5 + j L y R Wt Wt I R V L u Z V N Unc-Z Wt I R V L « Z j L y R Unc-Z Wt Figure 5: Punnett square showing the predicted zygotes from a dpy-5 unc-zl hTl(I;V) heterozygote in which recombination has occurred between the two recessive markers. The dpy is on the left and the unc is on the right side of the translocation breakpoint. The first four squares along the top and left-most columns are the non-recombinant gametes while the second four are the recombinant gametes. The normal chromosome I is indicated by the markers it carries as follows: dpy-5 (d5) and unc-z {uz). This diagram can be applied to any reciprocal translocation. Empty boxes represent zygotes presumed lethal due to severe aneuploidy. 33 to be modified to compensate for the loss of Wt progeny caused by recombination in this interval. In this case, the recombination frequency was calculated as; 3) p = t(4A-4Ai + 2W + 2Wi)-((4A-4Ai + 2W + 2Wi)2-4(4A-2Ai + 3 Ai2)(4A+W)) 1 / 2]/2(4A + W) where A was the number of Unc-101 recombinants. The i value was the recombination frequency between the breakpoint and the interval being scored (for unc-101 unc-54; 15 m.u., see Results). One arrives at equation 2) if i = 0 is inserted into equation 3). Confidence limits of 95% were calculated using the Poisson statistics according to Crow and Gardner (1959). Free Duplications: When scoring recombination from the progeny of dpy-x unc-yl + +; Dp(I;f) hermaphrodites, the calculation was modified for the fraction of gametes which carried the duplication. In most cases this was 41-42% (see section 3.13.1). The recombinant frequency was calculated as 4) p = t(.17W - .17D) - ((.17D - .17W)2 - 4(.09W + .03D)(.74D))1/2]/2(.09W + .03D) where W was the number of wild types (or Wt+Unc-13 with sDp2). Depending on the cross, D was the number of Dpy or Dpy Unc progeny. The frequency at which a duplication was found in the germ line was calculated by scoring the progeny of ml ml Dpx hermaphrodites; where m was dpy-5, unc-11 or unc-13. The fraction of duplication bearing gametes was x/(2 + x); where x = Wt/Mutant. In these calculations the duplication homozygotes were not considered as they were not scored. These worms were usually small, clear and slow growing. sDp2 homozygotes were smaller than other Dp homozygotes and only rarely produced progeny. Homozygotes of sDp2 derivatives deleted at the left end also had this phenotype. Dp[dpy-5( +) dpy-14(-)] strains, however, produced homozygotes which were fertile but still small and clear. 34 sDpl: I derived two formulas for calculating recombination in dpy-5 unc-xldpy-5 unc-x/sDpl hermaphrodites. In the first formula, I assumed sDpl - I;I segregations did not occur. If there was no duplication loss, such that half of the gametes carried the duplication, R was 2(D)/2(W + D) or D+TJ/(2(W+D)). As duplication stability decreased, the number of Dpy recombinant progeny (D) increased relative to wild-types (W), artificially increasing R. To compensate, a recombination formula was derived which took into account duplication loss. 5) p = [(.58D + -62W) - (.02DW + .38W2 - .86D2)1 /2]/(1.2D +.74W) The next formula assumed that when sDpl paired and recombined with a normal chromosome I, sDpl -lN;lN and sDpl;!^ - 1^ segregations occurred with a equal probability. In this calculation, no compensation was made for duplication loss. 6) p = [(.75D + .5W) - (.25W2 - .81DW - ,44D2)1 /2]/2(.5D + .3W) jRyLhTl: Formula 6) also applied to recombination in unc-29 unc-ylunc-29 unc-y/IRV^hTl hermaphrodites. In formula 6), D was the number of Unc-29 progeny. Deletions: In most cases calculating recombination from deletion heterozygote strains was straight forward. Total progeny was calculated as 2(wild-types + one recombinant class). The only problem would arise if there was frequent recombination between the interval being scored and the deletion breakpoints. This was the case in the studies of h655. In the h655 + +1 + unc-101 unc-54 heterozygote, I deduced from the dpy-14 unc-101 Ih655 + + experiments there was 18.6 m.u. between unc-101 and h655. To compensate for this recombination, the following formula was derived: 7) p = [(1.9A + 2W) - (4W2 - 12.42A2 - .41AW)1/2]/2(2A +W) where A was the number of Unc-101 recombinants. 35 Heterozygotes for deficiencies of the left end of chromosome I were constructed by crossing Df bearing males to strains homozygous for the markers to be scored. Typical was the hG5 51 dpy-14 unc-101 experiment. By crossing h655/dpy-14 males to dpy-14 unc-101 hermaphrodites, all wild-type progeny were of the correct genotype to be scored. In experiments with deletions from the right end of chromosome I, the markers to be scored were usually introduced through the male. eDf/dpy-5 unc-101 was constructed by crossing dpy-5 unc-1011+ + males to unc-54/eDf hermaphrodites and picking wild-type progeny. The correct hermaphrodites segregated Dpy-5 Unc-101 worms but no Unc-54 worms. To make the dpy-5 eDflhT2 hermaphrodites, the dpy-5 marker had to be crossed onto the eDf chromosome before crossing to hT2 bearing males. First, dpy-5 unc-101/ eDf hermaphrodites were constructed by crossing dpy-5 unc-101/+ + males to unc-54/eDf hermaphrodites. Dpy-5 progeny were picked from the dpy-5 unc-101/ eDf hermaphrodite progeny. These worms, dpy-5 unc-101/ dpy-5 eDf, were crossed to dpy-5/hT2 males. Wild-type progeny from this cross were of two types. The ones segregating Dpy-5 Unc-101 worms were not scored. Those segregating only Dpy-5 worms, and therefore of the genotype dpy-5 eDf/hT2, were scored. Hermaphrodites heterozygous for X-linked deletions and markers were constructed by crossing mnDpl/Df males to hermaphrodites homozygous for the markers. All hermaphrodite progeny from this cross were Df/ dpy unc. The presence of mnDpl was ignored when the interval being scored was outside the duplication. When the interval was covered by mnDpl, such as with unc-7 lin-15, only non-duplication worms were scored. These were easily distinguished from duplication worms because approximately 25% of their progeny were Unc Lin. In contrast, less than 10% of the progeny from duplication worms were Unc Lin. 2.8 Linkage of unlinked markers to chromosome I duplications. To determine if a chromosome I duplication carried material from another chromosome, the following cross was done, dpy-5; hDpx males were crossed to dpy-5; unc-y hermaphrodites and wild-type hermaphrodite progeny were picked. The unc-y mutation was on a chromosome 36 other than I or X. If the duplication was free, then the wild-types were of the genotype dpy-5ldpy-5; unc-y/+; hDpx. If the duplication covered unc-y, then the wild-types would not segregate any Unc-y non-Dpy-5 progeny. If the duplication was linked to another chromosome, then the wild-types were either dpy-5/dpy-5; unc-ylhDpx or dpy-5ldpy-5; unc-y I+; hDpx/ +. In the former genotype, the duplication was linked to unc-x, and the wild-type would segregate few Dpy-5 or Unc-y progeny. In the latter case, the duplication was not linked to unc-x, and the wild-type would segregate many Dpy-5 and Unc-x progeny. To determine if a duplication which carried dpy-5(+) had X-chromosome material, dpy-5; hDpx males were crossed to dpy-5; unc-y hermaphrodites, where unc-y was an X-linked mutation. If wild-type males were observed, then the duplication did carry unc-y(+). If non-Dpy Unc males were observed, the duplication did not carry unc-y(+). The presence of X-chromosome material on the ILXLszTl[dpy-5(-) unc-13(+ )] duplication derivatives had to be detected using a different method. For unknown reasons, crosses to ILXLszTl[dpy-5(-) unc-13( + )] strains produced males carrying these duplications at very low frequency. The rare males recovered were usually sterile. To test for rescue of unc-1, dpy-14 unc-13; hDp hermaphrodites were crossed to + dpy-14 +/dpy-5 + unc-13; hDp31/unc-1 wild-type males. Most of the wild-type hermaphrodites from this cross were dpy-14 +ldpy-14 unc-13; hDp; unc-l/+ or dpy-5 + unc-13l+ dpy-14 unc-13; hDp; unc-ll+. If these segregated Unc-1 progeny, then the duplication did not carry unc-1 (+). If the hermaphrodites did not segregate Unc-1 progeny, then the duplication did carry unc-1 (+). If no Unc-1 hermaphrodites were observed, the results were confirmed by setting more wild-type progeny on plates to isolate the hDp; unc-l/unc-1 hermaphrodites. 37 3. Results 3.1 Source of breakpoints on the left half of chromosome I. Attempts to isolate breakpoints in the left half of chromosome I were of two types. The first involved screening for gamma radiation induced recessive lethal mutations which could then be analyzed to determine whether they were associated with a chromosomal rearrangement. Since this method did not prove too fruitful in terms of producing chromosome I deletions or duplications (see below), the second method was more specific and consisted of screening for deletions in pre-existing duplications treated with gamma radiation. 3.1.1. Isolation of gamma radiation induced lethal mutations. There were three sources of gamma ray induced lethal mutations. Twenty mutations, induced with 1500 rad of gamma radiation on chromosome I marked with dpy-5 unc-13, were available prior to this study. Six of these mutations, h53, h58, h61, hl67, hl70 and hl71, had been recovered using the sDp2 system (Howell et al. 1987). The remaining 14 mutations, h545, h546, h549, h.550, h556, h559, h563, h564, h565, h590, h601, h654, h655, and h657, had been recovered using szTl(I;X) as a balancer (McKim, Howell and Rose 1988; Howell 1989). As part of this thesis, I isolated 27 lethal mutations in a screen using hT2(I:III) as a balancer. I chose a translocation balancer because the screens with sDp2 had not produced any deficiencies. The mutagenized strain was treated either with 1500 or 3000 rads of gamma radiation and recessive lethal mutations were recovered (section 2.3). In one set of experiments (the A set), 659 chromosomes were tested after treatment with 1500 rads and 59 mutations were recovered. An additional 497 chromosomes treated with 1500 rads and 416 chromosomes treated with 3000 rads were tested but only a few randomly chosen mutations from these screens were kept (the B set). Two methods were used to map the mutations to either chromosome I or III (section 2.3). From the 59 mutations of the A set, 19 mutations were kept which mapped exclusively to chromosome I. From the B set, eight mutations mapping exclusively to chromosome I were kept. The rest either mapped to chromosome HI or were more complicated and did not act like a simple chromosome I lethal mutation. All of those were discarded. 38 Table 3 Summary of gamma r a d i a t i o n induced l e t h a l mutations A l l e l e Balancer m. u. from dpy-5 a Progeny*3 Mapper c Region d F i n a l name (h00053) (h00058) (h00061) (h00075) (h00167) (hOOHO) (h00171) (h00331) (h00545) (h00546) (h00549) (h00550) (h00556) (h00559) (h00563) (h00564) (h00565) (h00590) (h00601) (h00654) (h00655) (h00657) sDp2 sDp2 sDp2 sDp2 sDp2 sDp2 sDp2 sDp2 szTl(IfX) szTl(IfX) szTl ( I f X ) szTl(IfX) szTl(IfX) szTl(IfX) szTl(IfX) szTl(IfX) szTl(IfX) szTl(IfX) szTl(IfX) szTl(IfX) szTl(IfX) szTl(IfX) 0.0 643 (25) 0.2R 1208 (302) 0.0 584 (25) 16.6L 2293 (208) 0.2L 869 (145) 0.2L 1006 (210) 2.0L 1797 (200) 0.2R 1057 (352) 2.1L 1049 (350) 0.7L 602 (201) 0.8R 1261 (315) 9.0L 348 (174) 0.2R 837 (279) 0.6R 624 (312) 0.4L 568 (284) 2.2R 1204 (301) 0.0 1055 (264) 0.0 947 (237) 0.0 950 (238) 10.4L 605 (302) RM RM RM AMH RM RM AMH KMC AMH AMH AMH AMH AMH AMH AMH AMH KMC KMC KMC KMC KMC Dp58/Dp50 Dpl9/Dp62 h61 dpy-5 h75 dpy-5 h331 dpy-5 h545 dpy-5 Dpl3/Dp37 Dp3/Df6 Df6/dpy-5 Dpi 2/Dpi 6 hDflO Dpl3/Dp37 Dpi 2/Dpi 6 Df6/dpy-5 unc-29 h.601 dpy5 h654 unc29 hDflO hDflO l e t - 5 2 4 let-529 hT?e l e t - f l e t -l e t -let-362 l e t -hDf6 l e t -let-354 l e t -l e t -l e t -let-378 l e t -l e t -let-400 hDf9 l e t -l e t -l e t -39 A l l e l e Balancer m. u. from dpy-5 Progeny Mapper Region F i n a l name (h00893) hT2(I;III) 10.IR 187 (93) KMc unc-29 h893 l e t (h00896) hT2(I;III) 2.3L 341 (114) KMc h896 dpy-5 hDflO (h00901) hT2(I;III) KMc unc-29 h901 l e t (h00904) hT2(I;III) 0.0 1062 (265) KMc h904 dpy-5 l e t (h00910) hT2(I;III) KMC unc-29 h910 l e t (h00912) hT2(I;III) KMC dpy5 l e t unc29 l e t (h00913) hT2(I;III) KMC h913 dpy-5 l e t (h00915) hT2(I;III) KMC h915 dpy-5 hDpl02 (h00916) hT2(I;III) KMC h916 dpy-5 hT3(I;X) a Distance from dpy-5. R = l e t h a l maps to the ri g h t of dpy-5, L = l e t h a l maps to the l e f t of dpy-5. b Progeny: t o t a l number of progeny scored. The average progeny per l e t h a l heterozygote i s indicated i n parentheses. c Mapper (two and three f a c t o r data): RM = Ricardo Mancebo; KMc = Kim McKim; AMH = Ann Marie Howell. d Region:" map p o s i t i o n based on 3-factor or Dp/Df mapping. e Translocation; other chromosome(s) involved are not known. f E s s e n t i a l genes were not defined on the basis of gamma r a d i a t i o n induced mutations. 40 Nine of the 27 chromosome I mutations isolated with hT2; h893, h896, h901, h904, h910, h912, h913, h915 and h916, were tested to determine if they were covered by the 1^ portion of ILXLszTl. The remaining 18 mutations have not been tested. The dpy-5 unc-29 (let-x); + /hT2(I;IH)[bli-4; dpy-18] hermaphrodites were crossed to bli-4 dpy-14;OlszTl(I;X) males. The non-Bli Wt male progeny from this cross, dpy-5 + + unc-29 (let-x)/'+ bli-4 dpy-14 + + were crossed to dpy-5 dpy-14; I^X^szTl hermaphrodites. If the wild-type progeny from this cross, dpy-5 + + unc-29 (let-x)/ dpy-5 dpy-14; ILXLszTl segregated Unc-29 F2 progeny, then the lethal mutation was inside I^X^szTl and the mutation was analyzed further (see below). Five of the mutations, h896, K904, h913, h915 and h916, mapped inside the ILXLszTl region. A summary of all the gamma induced mutations is shown in Table 3. Available two and three factor map data are also shown. 3.1.2. Induced deletion of duplication chromosomes Duplications of different sizes with one end in common would provide a linear array of breakpoints for rapid mapping of mutations. To generate such an array, gamma radiation was used to shorten pre-existing duplications. Shortened duplications were recovered by screening for the exposure of mutant phenotypes previously rescued by the duplication. The protocol for one such experiment is diagrammed in Figure 6. sDp2 derivatives: In the first screen (Figure 6), derivatives of sDp2 were recovered which had breaks in the unc-11 - dpy-5 region. The Po had a wild-type phenotype and segregated wild-type and Unc-11 Dpy-5 progeny. The FI progeny of gamma irradiated unc-11 dpy-5/ unc-11 dpy-5; sDp2 worms were examined for rare Unc-11 and Dpy-5 individuals. Because no recombination has been observed between the normal chromosomes and sDp2 (Rose, Baillie and Curran 1984), the rare Unc-11 and Dpy-5 hermaphrodites could have resulted from; 1) shorter duplications derived from a break between unc-11(+) and dpy-5(+) on sDp2, 2) a point mutation within unc-11 or dpy-5 on sDp2, 3) somatic loss of sDp2 resulting in a mosaic worm, or 4) intragenic revertants. Thirteen exceptional events were recovered. Genetic mapping of the duplication 41 unc-11 __L_ unc-11 I d p y - 5 dpy-14 unc-13 d p y - 5 dpy-14 unc-13 I I • Treat with gamma radiation Unc-11 Dpy-5 unc-11 unc-11 I dpy -5 unc-11 l _ dpy-5 • dpy-5 unc-11 L_ dpy-5 I Figure 6: Protocol for isolating shortened duplications. sD/j2-bearing hermaphrodites were treated with gamma radiation and the progeny examined for rare Dpy-5 or Unc-11 individuals. The results of these experiments are shown in Table 5. Similar experiments are shown in Tables 6 and 7. 42 Table 4 Isolation of duplications with breakpoints between unc-11 and dpy-5 Dose Dp[ull + d5~~\ Dp[ull'd5+] Chromosomes Rate 1500R 1 4 16000 3.1X10'4 3000R 3 4 7900 8.9X10'4 Table 5 Isolation of duplications with breakpoints between dpy-5 and dpy-14 Dose Dp[d5 + dl4'] Dp[d5-dl4 + ] Chromosomes Rate 1500R 5 0 4500 1.1X10"3 3000R 3 0 2500 1.2X10-3 Table 6 Isolation of duplications with breakpoints between dpy-5 and unc-13 Dose Dp[d5+ul3-] Dp[d5-ul3 + ] Chromosomes Rate 1500R a 6 4 4600 2.2X10'3 b 6 3 4600 1.9X10"3 3000R a 9 8 5000 3.4X10"3 b 3 3 4700 1.3X10-3 43 breakpoints described below revealed 12 class (1) mutations and one class (2) mutation (sDp2 [dpy-5(h585)]). Four duplications with the right end and eight duplications with the left end of sDp2 deleted were recovered (Table 4). Two pairs of duplications, hDp3; hDp7 and hDp5; hDp21, were isolated from the same mutagenized hermaphrodite. The breakpoints of hDp3 and hDp7 are the same (section 3.2), while the latter two have different breakpoints. Both of these situations could have been caused by a premeiotic mutation which was, in the case of hDp21, accompanied by a spontaneous post-mutation shortening (section 3.13.2). A similar approach was used to obtain breaks in the dpy-5 dpy-14 region. In this screen, dpy-5 dpy-141 dpy-5 dpy-14; sDp2 worms were treated with gamma radiation and their FI progeny examined for rare Dpy-5 and Dpy-14 worms. The same possibilities as outlined above apply to this screen. Eight Dpy-14 worms were recovered (Table 5). Genetic mapping of the breakpoints (below) has shown all of these to be shortened derivatives of sDp2. In addition, one unlinked semi-dominant Dpy mutation (h630) was recovered. This was shown by complementation testing to be an allele of dpy-13 (IV). Unlike the previous screen, no duplications with the left end (dpy-5) deleted were recovered. I^X^szTl derivatives: A third screen used a different duplication, I^X^szTl. This duplication was one of the translocation chromosomes from szTl(I;X) that had been recovered in a strain where it existed in addition to the euploid complement of chromosomes (section 3.6.1). ILXLszTl carries the portion of chromosome I left of unc-29 (I^1) joined to a fragment of the X-chromosome (X^1) (Figure 13). This X-chromosome portion carries unc-1 (+) and dpy-3(+) but not unc-20( + ). dpy-5 unc- 13; ILXLszTl worms were phenotypically wild-type and Him due to X-nondisjunction. The dpy-5 unc- 13; ILXLszTl worms were treated with gamma radiation and their FI progeny were examined for rare Dpy-5 or Unc-13 individuals. Twenty-four Unc-13 and 18 Dpy-5 hermaphrodites were recovered (Table 6). Genetic mapping of the breakpoints (below) showed 23 of the Unc and all of the Dpy were shorter derivatives of ILXLszTl. One Unc-13 strain had no dpy-5 marker and was not analyzed further. Two dpy-5(+) dpy-14(-) duplications, hDp40 and hDp63, caused the hermaphrodites to be sick and slow growing, dpy-5 dpy-14; Dp 44 strains could not be constructed for these two duplications and therefore no extensive mapping of their breakpoints was done. Unlike the sDp2 screen of this region, duplications with the left end (dpy-5) deleted were recovered. Included in Table 6 as entries in columns two and three is a worm which segregated Unc-13, Dpy-5, Wt and Dpy-5 Unc-13 progeny. The wild-types were picked and these segregated more wild-types, Unc-13, Dpy-5, and Dpy-5 Unc-13 worms. Two duplications were isolated from the wild-type strain, the Unc-13 worms carried an unc-13(-) duplication (hDp64) while the Dpy-5 worms carried a dpy-5(-) duplication (hDp73). Since hDp64 and hDp73 have similar breakpoints (section 3.4.2) and complementary parts of the I^X^szTl chromosome, they could be products of the same mutational event. The recovery frequency for the I^X^szTl derivative duplications (Table 6) was approximately the same as that for the sDp2 derivative duplications in the dpy-5 dpy-14 screen (Table 5). The duplications with breaks in the unc-11 dpy-5 region were recovered four times less frequently (Table 4) than duplications with breaks in the dpy-5 unc-13 region. 3.2 Mapping the gamma radiation lethal mutations and the new duplication breakpoints. Mapping was done primarily on the basis of complementation tests (section 2.4). In addition, some of the lethal mutations were mapped by two and three factor crosses with respect to their cz's-linked visible markers (Table 3). For the complementation tests, use was made of visible and lethal mutations previously located on the left half of chromosome I. In addition, there were five deletions available in this region (Figure 7). Most of the visible mutations were originally mapped by Brenner (1974). Modifications of these data are in Edgley and Riddle (1987). Some of the lethal mutations were from the work of" Rose and Baillie (1980). The rest of the lethal mutations [except sup-11 (n403n682~} were from the set isolated with sDp2 (Howell et al. 1987; Howell 1989). These lethal mutations had already been positioned relative to dpy-5, unc-13 and sDf4. The mapping results are shown in Figures 7, 8, 9, and 10. 45 Figure 7: Duplications and deficiencies of the left half of chromosome I. Not all the duplications are shown. Markers on the map, except bli-3, egl-30 and unc-35, were tested for complementation with each duplication. When it was not known if a duplication covered a marker, the duplication was drawn with a broken double line. Gamma induced lethal mutations h61, hl67, and hi 71 were mapped to the left of hDp 18 but not with respect to the other duplications. h 6 5 5 b l i - 3 e g l - 3 0 u n c - 3 5 l e t - 3 6 2 I h61,hl67,hI71 l i n - S lin-17 sup-11 I dpy-14 h 6 S 4 f o g - 1 unc-11 u n c - 7 3 d p y - 5 u n c - 4 0 b l i - 4 unc-13 u n c - 2 9 I I I I I I I I I I  s D £ 4 hOf6 nDt 2 3 nDp4 hDp102 hDp70 s D p 2 hDp3 hDp 13 hDpl4,16 1 h D p l 9 hDplB I hDp62 hDp72 — I h D p 6 9 H hDp9 h D p S l Lr 0) 46 Figure 8: Position of breakpoints of sDp2 derived duplications which break in the unc-11 dpy-5 region. Genes and gamma ray-induced mutations of essential genes (designated by allele number) are indicated above the line. Mapping of essential genes with respect to sDf4, hDf6, hDp3 and hDp20 was the work of Howell (1989). When it is not known if a duplication complements a marker, the duplication is drawn with a broken double line. All duplications were tested for complementation with all the other markers on the map. unc-11 let-353 let-356 let-366 let-373 let-354 let-374 let-351 let-359 let-364 unc-73 unc-38 unc-89 unc-74 unc-57 let-375 h550 h565 h913 let-352 let-361 let-363 let-371 unc-63 spe-11 dpy-5 unc-40 bli-4 unc-37 dpy-14 unc-13 I I I sDf4 hDf6 sDp2 hDplO,ll hDp8,9 hDp2,3,5,7 hDp4 hDp6 hDp20,21 H 1 m . u . 47 Figure 9: Position of breakpoints of sDp2 and ILXLszTl derived duplications which break in the dpy-5 unc-13 region. Most of the duplications extend to the left end of the chromosome which was shown in Figure 7. With the exception of hDpl8, the gamma ray induced sDp2[dpy-5(+) dpy-14(-)\ derivatives (hDps 12-19) carried let-362(+) at the left end of the chromosome. Nine ILXLszTl[dpy-5(+) unc-13(-J] derivatives (hDp31, 32, 33, 34, 35, 36, 37, 38, 58) were tested and rescued let-362. hDp62 and hDp72 have left end deletions (see Figure 7). With regard to X-chromosome sequences, seven I'-'X^szTl[dpy-5(+) unc-13(-)] derivatives (hDp33, 35, 37, 39, 41, 42, 54) were tested and failed to rescue unc-1. Two duplications, hDp31 and hDp56, rescued unc-1. hDp!3 hDP15 hDp31 hDpl2 hDp41 hDp51 hDp 17 hDpU hDp55 hDp 16 hDp39 hDp61 hDpl8 hDp54 hDp 19 hDp72 hDpSl hDp56 hDp33 hDp58 hDp 3 2 hDp 3 4 hDp 3 5 hDp 6 2 sDp2 hDp38 hDp36 hDp42 hDp43 hDp 60 hDp 64 ILXLszTl 0.3 m.u. dpy-5 let-378 hi 70, h546 unc-40 hS56, h564 bli-4 let-6 02 unc-37 dpy-14 unc-14 let-75 unc-13 49 Figure 10: Position of breakpoints of dpy-5(-) derivatives of I^X^szTl. Most of these duplications do not carry any left end sequences; they are terminal deletions. Six I^X^szTl [dpy-5(-) unc-13(+)] duplications (hDp44, 45, 48, 49, 53, 65) failed to complement let-362. As shown in Figure 6, hDp51 and hDp69 are internal deletions of ILXLszTl. Seven ILXLszTl[dpy-5(-) unc-13(+ )] duplications were tested (hDp45, 46, 50, 51, 65, 66, 73) and all rescued unc-1. When it is not known if a duplication complements a marker, the duplication is drawn with a broken double line. dpy-5 3 m.u. ILXLszTl let-75 unc-13 51 Three deletions were found in this analysis of the 29 gamma radiation induced lethal mutations. The remaining 26 mutations failed to complement no more than one complementation group (see below). Further analysis of some mutations in this group of 26 is described below. The mutation h545, which has been renamed hDf6, failed to complement unc-73, unc-74, unc-57 and unc-38. Mapping of essential genes with respect to the deficiencies sDf4 and hDf6 is published elsewhere (Howell et al. 1987; Howell 1989). Mutation h601, which has been renamed hDf9, failed to complement nDf23, nDf24, nDf25 and let-82. let-82 complemented the three nDf mutations, positioning it to the right of unc-29 (Figure 12). Mutation h896 is a deletion at the left end of chromosome I and has been renamed hDflO. It is described in section 2.3. Mapping strategy: The use of autosomal duplications to map genes requires observation of the F2 generation to determine if the duplication carries the wild-type copy of the gene being tested. The objective of the protocol was to selectively recover duplication-bearing strains from a cross between a duplication strain and a strain carrying the mutation to be tested. The F2 progeny from these crosses were observed to see if the duplication could rescue the tested mutation. I developed the protocol described in section 2.4 in order to map rapidly a large numbers of mutations (Figure 3). Mapping Results: The duplications in combination with sDf4 and hDfS have subdivided the 4 m.u. region between unc-11 and unc-13 into at least 18 intervals. None of the seven unc-11 (-) dpy-5(+) duplications had breakpoints in the interval between the right breakpoint of hDf6 and dpy-5. The breakpoints were in or close to hDfB. Similarly, none of the duplications tested had a break in the unc-3 7 - dpy-14 interval (Figures 9 and 10), even though this interval is recombinationally 1/5 of the dpy-5 unc-13 region (Edgley and Riddle 1987). Twelve of 41 ILXLszTl duplications had breakpoints in the dpy-14 unc-13 region; an interval genetically 1/5 the size of the dpy-5 unc-13 region (Rose and Baillie 1979). Further mapping of these duplications with respect to essential genes is described in section 2.4. These data suggest some regions are refractory to duplication breakpoints. 52 Most duplications appeared to result from a single break in the selected region, producing a terminal deletion (Figures 7, 8, 9, 10). For example, hDp8, hDp9, hDplO and hDpll have a single breakpoint in the unc- 73 dpy-5 region. Their left endpoint is probably the same as sDp2 because all four complement bli-3. Some of the duplications, however, resulted from more than one break (Figure 7). hDp51 and hDp69, two I^X^szTl[dpy-5(-) unc-13(+)] duplications, carry an internal deletion involving dpy-5. hDp69 was dpy-5(-) but rescued let-362, unc-11 and unc-74. hDp51 rescued let-362 and lin-6 but not unc-11. It was not tested against lin-17 or sup-11 (see also section 2.3). In addition, hDp71, a third I^'X^'szTl[dpy-5(-) unc-13(+)] duplication, failed to rescue dpy-5, but did rescue all other markers tested. Thus hDp71 may have a small deletion involving dpy-5 or an intragenic mutation. hDpl8, in addition to being deleted for the dpy-14 region, was also found to be deficient for the left end of the chromosome. hDp 18 rescued unc-11 and fog-1 but not sup-11, lin-17, lin-6 or let-362. Thus, at least one unselected break occurred in the region between sup-11 and fog-1. hDp62 and hDp72, two of the l^X^szTl[dpy-5( + ) unc-13(-)] derivatives, failed to rescue egl-30, let-362 and lin-6. In addition, hDp72 rescued sup-11 but hDp62 did not. Like hDpl8, these duplications probably have a terminal deletion caused by a second break to the left of unc-11. The X-chromosome sequences were expected to be lost in the terminal deletions of the ILXLszTl[dpy-5(+) unc-13(-J] type if the X-chromosome sequences were attached to the right end oULXLszTl. Nine duplications were tested and, as predicted, seven did not rescue unc-1 (Figure 9). Two duplications, hDp31 and hDp56, did rescue unc-1 and therefore contained this portion of the X-chromosome. These two duplications resulted from internal deletions of ILXLszTl. Conversely, terminal deletions of the lLxLszTl[dpy-5(-) unc-13(+)] type were expected to retain the X-chromosome material because the X-chromosome sequences were attached to the other end of I^X^szTl. Consistent with this prediction, all seven of the ILXLszTl[dpy-5(-) unc-13(+)] duplications tested rescued unc-l(+) (Figure 10). Two breaks were also involved in the formation of hDp20. In addition to the chromosome I material (Figure 8), this duplication carries sequences from the unc-60 region of chromosome V. 53 In addition to unc-60, a deletion from this region of chromosome V, sDf50 (Figure 1), was rescued by hDp20. An additional duplication of the region was nDp4. nDp4 was isolated in H. R. Horvitz's laboratory. Its breakpoints were mapped using the same techniques as the other duplications. nDp4 is linked to another chromosome (J. Thomas and S. Kim, personal communication). I have been able to isolate slow, thin wild-types that did not segregate non-nDp4 progeny (Unc-13) from the unc-13; nDp4l+ strain. This is the behaviour expected of a linked, homozygous duplication. Most free duplications were not stable as homozygotes (section 3.9). These hermaphrodites probably had the genotype unc-13; nDp4/nDp4. Data from the MT lab (Edgley and Riddle 1987) showed that nDp4 includes the portion of chromosome I between dpy-5 and unc-75 inclusive. I have confirmed this by showing that nDp4 rescues bli-4, dpy-14, unc-29 and unc-75. In contrast to the original data, however, my copy of nDp4 failed to rescue dpy-5 or unc-40. Therefore, the left breakpoint of nDp4 is between unc-40 and bli-4. nDp4 failed to rescue unc-101. Therefore, the right breakpoint of nDp4 is between unc-75 and unc-101. Furthermore, these data establish that unc-75 is to the left of unc-101. Figure 8 shows that KDp4 rescued let-351(h43). In fact, h43lh43;hDp4 worms grew to be adults but were sterile and, therefore, by definition hDp4 failed to rescue h43. In Figure 8 hDp4 is shown to include let-351 for two reasons. First, the h43/h43 phenotype was a mid-larval lethal (Howell 1989), not a sterile adult as observed in the hDp4 bearing homozygotes. Second, h43/h43; hDp78 worms were viable. As described in section 3.13.2, hDp78 is a spontaneous deletion derivative of hDp4 linked to chromosome I. Since hDp78 had the wild-type allele of let-351, so must hDp4. These data suggest that hDp4 carries the wild-type allele of leU351 but is unable to rescue h43 due to properties unique to this duplication. This is one of only three examples in this study of a duplication failing to rescue a mutation for reasons other than map position (see sections 3.4.2 and 3.13.2). Further analysis of gamma radiation induced lethal mutations. Some of the gamma ray induced lethal mutations mapped in this study were not separable from dpy-5 or unc-13 in 54 recombination experiments (Table 3). Duplication mapping of these mutations positioned some of them into unexpected places on the map. Five gamma ray induced lethal mutations, h61, hi67, hi 71, h655 and h904, recombinationally inseparable from dpy-5 were mapped to the left end of the chromosome because they were rescued by hDpl2 but not by hDp!8. The h904 chromosome was peculiar in that the dpy-5 h904 unc-29; Dp worms were not Unc-29 but a weak coiler Unc. Either this chromosome carried a mutation modifying or epistatic to unc-29, or the unc-29(e403) mutation reverted and a new unc mutation was induced. Two other mutations {h601, h654) recombinationally inseparable from dpy-5 mapped to the right of unc-13 based on the failure of I^X^szTl to rescue them (Figure 12). These two mutations were complementation tested against the deficiencies in the unc-29 region. h654 and h601 fully complemented each other and unc-29. h654 was mapped inside nDf24 and nDf25 but outside nDf23. h601 failed to complement all three deficiencies as well as let-82. let-82, however, complemented nDf23, nDf24 and nDf25 and, therefore, mapped outside these deletions. Because h601 failed to complement two separable loci, the nDfs and let-82, I concluded it was a deficiency and renamed it hDf9. As described in section 3.4.2, these two mutations complement all other lethal mutations in the region. All seven of these gamma ray induced mutations should have been recombinationally separable from dpy-5. h61, h655 and h904 suppressed recombination in trans to a bli-3 unc-11 chromosome, while h601 and h654 suppressed recombination in trans to an unc-11 dpy-14 chromosome (section 3.10.1). These results suggest these five mutations disrupt sequences required for normal recombination, hi67 and hi 71 did not suppress recombination in trans to bli-3 unc-11 and were probably not crossover suppressors. These chromosomes may carry two lethal mutations, one near the left end and the other in the dpy-5 unc-13 region. This was shown to be true for the hi 71 chromosome because it failed to complement sDf4 h61: Unlike the other crossover suppressors mapping to the left end of chromosome I, h61 did not behave like a simple intrachromosomal rearrangement. The ratio of wild-type to Unc-11 progeny from unc-11/ h61 dpy-5 unc-13 heterozygotes was abnormally high, 3.6 (117 wild-types 55 and 42 Unc-11), and the average progeny size (25; Table 14) was abnormally low. The "homozygotes" (h.61 dpy-5 unc-13; sDp2) were noticeably more fertile. Furthermore, in an egg count from unc-11/ h61 dpy-5 unc-13 heterozygotes, only 20% (87/438) of the eggs went on to produce viable worms. This data suggests h61 is associated with a rearrangement that generates aneuploid progeny when heterozygous. It is possible h61 is associated with a chromosome I translocation, although no pseudolinkage with other markers has been found. h915: The strain containing the chromosome I linked mutation h915 also carries an X-linked duplication. It was possible to separate the duplication from the chromosome I linked h915 mutation, but the reciprocal was not possible. This duplication was named hDpl02. The original dpy-5 h.915 unc-29lhT2 stock retained the duplication as did a dpy-5 K915 unc- 29; ILXLszTl stock, dpy-5 h915 unc-29/+ strains may require the duplication for viability suggesting the h915 mutation disrupts or deletes a haploinsufficient site. In the mapping experiments, dpy-5 h.915 unc-29/ unc-11 dpy-14 males were constructed (see section 2.2), and crossed to dpy-5 dpy-14; hDpx (where the hDpx was dpy-5(+) but dpy-14(-)) strains. The F I wild-type hermaphrodites were of two types. The first type had the h915 chromosome as judged by the presence of Dpy-5 and Unc-29 progeny. The second type of F I wild-type segregated the Unc-11 Dpy-14 chromosome and therefore carried a novel dpy-14(+) duplication, dpy-5 h915 unc- 29; ILXLszTl hermaphrodites were used to generate dpy-5 h915 unc-29ldpy-14 males. These males, when crossed to dpy-14 hermaphrodites, produced Dpy-14 male progeny but no Dpy-14 hermaphrodites. An X-linked dpy-14(+) duplication would have produced these results. Duplication homozygotes were healthy and viable. In the progeny of unc-13; hDpl02 hermaphrodites were 223 Unc-13 hermaphrodites, 643 wild-type hermaphrodites and one Unc-13 male. The ratio of wild types to Unc-13 hermaphrodites, 2.88 to 1, and the fact that many strains were isolated which failed to segregate non-duplication progeny suggested that hDpl02 homozygotes were viable. hDpl02 failed to rescue all markers tested to the left of dpy-5, including let-352(h45), the closest marker to the left (section 3.13.2). hDpl02 rescued let-370(hl28), a marker which had not 56 been separated from dpy-5 by recombination experiments (Howell et al. 1987) or duplication mapping experiments (McDowall 1990). Because h915 was induced on a dpy-5(e61) containing chromosome, the left duplication breakpoint could not be positioned with respect to dpy-5 (Figure 7). At the right end, hDpl02 failed to rescue hDf9 and unc-101 but did rescue let-88 and h654. These data placed the right hDpl 02 breakpoint in the unc-29 region. hDpl 02 was induced on an unc-29(e403) containing chromosome so I could not position the breakpoint with respect to unc-29. h916: The h916 lethal mutation is associated with a I;X translocation. During initial mapping of h916, it was found to be inside I^X^szTl and hDpl5. The experiments were irregular in that when h916 dpy-5 unc-29l+ '+ dpy-14 + males were crossed to dpy-5 dpy-14; hDp!5 hermaphrodites, no FI Wt males or Dpy hermaphrodites were observed. One explanation for this observation was that h916 involved a I;X rearrangement. As described in section 3.6.4., this was shown to be true and h916 was renamed hT3(I;X). Allelism of mutations: Some of the gamma ray induced lethal mutations were complementation tested against known EMS mutations of essential genes (Table 3). h549 was shown by Howell (1989) to be an allele of let-354. Howell (1989) also tested h550 and h565 for allelism with known genes in the hDfS to dpy-5 interval. No allelism was found, indicating h550 and h565 define new loci. Mutation h913 was mapped to the same region but has not been complementation tested against the known mutations. I found h563 to be an allele of let-378. hi 70 and h546 failed to complement each other but were not tested against other lethal mutations in the region. Among the mutations mapping near the left end of the chromosome (Figure 7), hi 71 was an allele of let-362(h86). Mutations h53, h58 and h590 were mapped into the hDpl9 -sDp2 region and are described in section 3.4.2. Mutations in genes not previously identified are indicated in the figures by their allele designation. While these mutations define new essential loci, it is still possible they delete more than one complementation group. 57 3.3 The left end of chromosome I. The duplication breakpoints were intended to be used in the detailed analysis of small regions. Towards this end, two regions of the left half of chromosome I were studied. The first region was the left end of chromosome I (Figure 11). In section 3.10.1, I describe four gamma radiation induced mutations, h61, hDflO, h904 and h655, which are also crossover suppressors. To characterize these further, a genetic map of deletions and genes at the left end was needed. I arbitrarily defined the region as being to the left of sDf4 (Figure 11). The source of mutations was a) those already mapped to this region by other workers and b) lethal mutations generated in this lab and recombination mapped to this region. The collection of lethal mutations were selected to provide a baseline with which to map the rearrangement breakpoints (Table 7). They do not represent all the mutations from a particular screen (as in the dpy-14 region analysis; see below) and do not represent a randomly chosen group of mutations. As described above, four duplications had breakpoints in the region (Figure 11). In addition, the deletions tDf3 and qDf3 were available from other laboratories (Table 2). One of the four crossover suppressors, hDflO, was shown to be a deletion by this analysis. hDflO failed to complement at least eight genetic loci. The loci represented by let-362, let-548. let-and h75 were not positioned with respect to hDflO. Positioning these genes with respect to hDflO might resolve if hDflO is an internal deletion. Two of the crossover suppressors, h655 and h904, failed to complement each other and hDfl 0, but no other loci. The fourth crossover suppressor, h61, complemented h655 and h904, but was not tested against hDflO. Many of the lethal mutations caused the worms to arrest development as late larvae or as adults (Table 7). This usually did not change when the mutations were made heterozygous with a deletion. There were three exceptions. These three mutations may be hypomorphs because their phenotypes are more severe over a deletion. First, let-549(h!93) dpy-5 unc-13/ tDf3 dpy-5 + worms arrested development early in development. The hi 93 homozygotes, however, arrested development as adults and some produced a few progeny that never progressed beyond an early larval stage. Second, Un-17(n671) dpy-51 dpy-5 tDf3 hermaphrodites were sterile adults or 58 hDflO bli-3 unc-35 egl-30 h655 k904 spe-8 let-546, let-547 let-552 h559, h651 hDp51 iDfS A let-362, let-548 let-551, h75 hn-6 let-549 let-553 lin-17 sup-11, let-550 hDp 72 qDf3 hDp 18 let-360 let-365 fog-1 let-357 hDp 6 2 sDf4 unc-11 Figure 11: Genetic map of the left end of chromosome I. hDp51 was not tested with lin-17. 59 Table 7 Mapping of mutations at the l e f t end of chromosome I. Gene and a l l e l e Arrest D i s t a Map b hDpl8 hDp62 hDp72 hDp51 hDflO tDf3 qDf3 bli-003(e00767) viable -egl-030(n00686) viable -fog - 0 0 1 ( q 0 0 2 5 3 ) adult + - + - c let-357(h00089) late 3.0 AMH + + -let-360(h00096) late 4.6 AMH + - + let-362(h00086) mid 15.4 AMH - + + + let-365(h00129) adult 5.5 AMH + - + let-365{h00295) late 7.0L JM + - + let-546(h00227) mid 18.OL LJH + -let-547(h00277) adult 12.6L RLK + -let-548(h00356) early- - - + + + let-549(h00193) adult 11.2L LJH - + + -let-550(h00294) late 4.0L JM + - + let-551(h00347) early + + -let-552(h00220) early 17. 9L LJH -let-553(h00247) early 8.6L KMC - + + -lin-006(e01466) mid lin-006(h00092) mid 10.2 AMH - + lin-006(h00099) mid 11.3 AMH -lin-017(n00671) viable - - -spe-008(hc0050) adult -sup-ll(n403n682) early - + -unc-035(e00259) viable -(h00075) early 16.6 AMH + + (h00559) early 9.0 AMH -(h00657) early 10.4 KMC - + -a Distance from dpy-5. b Map = Mapper; KMc = Kim McKim; AMH = Ann Marie Howell; LJH = Linda J . Harris ; RLK = Rohinish Kisun; JM = Jennifer McDowall. c Barton and Kimble (1990) 60 produced very few progeny. This suggests the null phenotype of lin-17 may be a sterile adult or perhaps arrest at an early stage of development. The third putative hypomorph was bli-3(e767). Two crosses were carried out with e767. In the first, hDflO dpy-5 unc-131 unc-11 dpy-14 males were crossed to bli-3 unc-11 hermaphrodites. With the exception of rare recombinants (hDflO is a crossover suppressor; section 3.10.1), the wild-type progeny from this cross were bli-3 unc-11/hDflO. When the cross was done, no wild-types were observed, suggesting that bli-3(e767) over a deletion was an early larval lethal. In a second cross, bli-3 dpy-5/hTl males were crossed to hDfl 0 dpy-5 unc- 13: ILXLszTl hermaphrodites. In this cross small Dpy progeny (similar to the double Dpy phenotype of Dpy-5 Dpy-14) were observed, a minority of which were fertile. This result did not change when the reciprocal cross was done. Perhaps the early larval lethality of bli-3/hDfl0 observed in the first cross was partially suppressed in the second cross by the Dpy-5 phenotype. It is known that some dpy mutations suppress bli mutations (K. Peters, personal communication). In contrast, the spe-8(hc50) (L'HernauIt, Shakes and Ward 1988) phenotype was the same in the hDflO heterozygote as in the homozygote. spe-8 homozygotes are sterile because of defective sperm. These results suggest the sperm defect is the null phenotype. One mutation, let-547(h277), had a maternal effect phenotype. Most of the adult homozygotes produced eggs which hatched. The larvae usually did not progress beyond an early larval stage. To test if the zygotes could be rescued by male sperm, the adult lethal homozygotes were crossed to N2 males. Iet-547(h277) homozygotes were rescued by male sperm, suggesting that zygotic gene expression was necessary and sufficient for viability. Homozygous h277 hermaphrodites had the same maternal effect lethal phenotype as h277lhDflO hermaphrodites, suggesting this was the null phenotype of let-54 7. The other adult arrest lethal mutations were crossed to N2 males to see if they had a sperm defect or a zygotically rescuable maternal defect. None of the mutations were rescued by this cross. Alleles of lin-6 are dominant suppressors of the Lon-2 phenotype. The males from MT1442 [Un-6(el466) dpy-5; + l szTl(I;X)] and GE1385 [tDf3 dpy-5; unc-3/ szTl(I;X)] were not 61 Lon-2 but looked wild-type. These males usually failed to mate. To show these strains still carried and expressed the lon-2(e678) mutation, they were crossed to N2 males. As expected, many Lon-2 progeny were produced. The suppression was not allele specific since a null allele of lin-6, tDf3, dominantly suppressed lon-2(e678). An examination of Figure 11 shows the genes may not be evenly distributed along the left end of the chromosome. There may be a concentration of genes at the left end with a corresponding low density region between lin-6 and unc-11. The unc-11 region marks the fringe of the gene cluster centered around dpy-14. Perhaps the left end of chromosome I is a mini gene cluster with respect to the neighbouring lin-6 unc-11 interval. Similar observations on chromosome V were made by Johnsen (1990). 3.4 The dpy-14 region. The region between the right-end breakpoints of hDpl9 and sDp2 was chosen for more extensive analysis. This region, shown in Figure 12, will be referred to as "the dpy-14 region". 3.4.1. Source of mutations. Using the procedure of Howell et al. (1987), the Rose laboratory has tested more than 31 000 chromosomes mutagenized with ethyl methane sulfonate (dose varied from 12-17 mM) and recovered 550 lethal mutations. All of the strains were of the genotype (let-(hx) dpy-5) unc-13; sDp2. The availability of my duplications plus hDf6 and sDf4 has been used effectively to partition most of the 550 mutations into smaller groups for analysis. Howell (1989) mapped 114 mutations to the region deleted by sDf4. Most of the remaining lethal mutations were mapped with duplications. The duplication hDpl3 was used by J. McDowall and A. M. Rose to divide the lethal mutations into two groups. Those that mapped to the right of hDpl3 were further mapped with respect to hDp 16 and hDpl 9 by McDowall (1990). Forty-one lethal mutations were not mapped in these experiments. Thirty-eight of these mutations were tested with hDpl 9 by myself and A. M. Rose (Table 8). The remaining three mutations were lost. In total 90 lethal mutations have been mapped to the dpy-14 region; 72 by 62 Figure 12: Genetic map of the dpy-14 - unc-29 region. This map is not drawn to scale but represents approximately 1 m.u. of chromosome I. All genes on the map with the exception of bli-4 and let-602 are to the right of hDpl9 (Figure 9). Essential genes let-81, let-85 and let-90 were not tested against hDp50, 65 or hDp73. The essential genes to the right of the line in the unc-3 7 dpy-14 region have not been positioned relative to each other or the genes on the line, let-540 has not been positioned relative to unc-13. If sDf6 extends further to the right than sDf5, let-540 could be to the right of unc-13. 63 kDp33 hDp35 1 hDp62 I hDp58 sDp2 hDp38 hDp42 hDp60 hDp36 hDP43 hDp64 hD}8 nDf(n2088) hDp65 hDp50 hDp73 sDf6 sDf5 nDf24 nDf25 nDf23 T hDf9 ± bli-4 let-602 let-392 unc-37 let-383, let-385, let-394, let-397, let-398, lei-399, let-521, let-528, let-534, unc-87 dpy-14 let-543, lei-544, let-544, let-605 let-86, let-529 lei-389, let-400, let-520, sem-4 unc-14, let-522, let-523, let-525 let-524, let-527 let-539 let-75 unc-15, let-542 let-540 unc-13, let-87 let-88, let-81, let-85, let-90 let-80, lei-89, let-535 let-537, let-538, h654 unc-29, let-536, let-541 let-82 64 McDowall (1990) and 18 by A. M. Rose and myself. Many of the 20 mutations inside hDpl9 have little else known about them and should eventually be mapped against the other duplications for inclusion in those regions. With the exception of mutant strains which could not be analyzed because they did not grow upon thawing or did not produce fertile males, this group of mutations was derived from the entire set of 552 lethal mutations isolated with sDp2. The results are summarized in Table 8. 3.4.2. Mapping mutations to rearrangements. The lethal mutations were mapped against rearrangement breakpoints in the hDpl9 -sDp2 interval. An examination of Figures 9 and 10 shows there were 17 breakpoints in the region. This includes hDf8, the hDpl9 group: hDps 33, 35, 39, 54, 56; the hDp45 group: hDp45, hDp46, hDp66; hDp62, hDp58, hDp50, hDp65 and the sDp2 group: hDps 38, 42, 60. Unfortunately, most of these were found to break outside the hDpl9 - sDp2 region. From the hDP19 group (Figure 9), McDowall (1990) found hDps 39, 54, 56 had breakpoints to the left of hDpl9. I have shown that all members of the hDp45 group complement let-602(h283) (Figure 10) and therefore also break to the left of hDpl9. The breakpoints of the four members of the sDp2 group were not differentiated with the mutations available. The six remaining breakpoints were informative in the analysis of this region (Figure 12). The dpy-14 region was initially divided into three intervals with hDp62 and hDp58. Mutations mapped to one of these intervals were then placed into complementation groups. The hDpl9 -hDp62 region had the largest number of complementation groups with 15 identified genes. The largest mutational target in the dpy-14 region was also in this region. Eighteen alleles of let-385 were isolated in this analysis. In addition, one spontaneous allele, h578, was isolated. Seventeen of the EMS alleles and h578 caused worm development to arrest at an early larval stage. Homozygotes for let-385(h748), however, had a late larval arrest phenotype. This allele was classified as a hypomorph because its phenotype was more severe over a null allele or deficiency. When h748 was heterozygous to a hf8 or h85, the worms arrested development at an early larval stage. Table 8 Mapping of mutations i n the dpy-14 region. Gene a A l l e l e A r r e s t 3 hDp62 hDp58 hDp65 hDp50 hDf8 hDp83 dpy-014* (e00188) viable + - - - + l e t - 8 6 * (s00141) early + -let-383* (h00115) early + + + l e t - 3 8 3 #(h00389) early l e t - 3 8 3 (h00527) early + + l e t - 3 8 3 #(h00727) early l e t - 3 8 5 * (h00085) early + + + let-385 (h00109) let-385 (h00135) l e t - 3 8 5 (h00202) l e t - 3 8 5 #(h00431) early let-385 (h00435) early let-385 #(h00453) early -l e t - 3 8 5 #(h00470) early l e t - 3 8 5 #(h00492) early l e t - 3 8 5 #(h00514) early l e t - 3 8 5 #(h00523) early l e t - 3 8 5 (h00524) + + l e t - 3 8 5 (h00578)b early l e t - 3 8 5 #(h00686) early l e t - 3 8 5 #(h00687) early l e t - 3 8 5 #(h00717) early l e t - 3 8 5 #(h00742) early l e t - 3 8 5 (h00748) l a t e l e t - 3 8 5 (h00789) early l e t - 3 8 9 (h00106) early -l e t - 3 8 9 (h00190) early - + l e t - 3 8 9 * (h00428) early - + l e t - 3 8 9 (h00429) early + l e t - 3 8 9 (h00680) early - + + let-389 (h00698) l a t e + + Table 8 (con 1 t ) . Gene A l l e l e Arres t hDp62 hDp58 hDp65 hDp50 hDf8 hDp83 l e t - 3 8 9 (h00801) ear ly - + l e t - 3 8 9 (h00807) ear ly + l e t - 3 8 9 (h00825) ear ly + l e t - 3 8 9 (h00836) ear ly + let-392* (h00120) ear ly + + -l e t - 3 9 2 (h00122) ear ly l e t - 3 9 2 #(h00367) ear ly l e t - 3 9 2 #{h00489) ear ly l e t - 3 9 2 (h00575) ear ly + + -l e t - 3 9 2 (h00576) ear ly + + l e t - 3 9 2 #(h00762) ear ly l e t - 3 9 2 #(h00780) ear ly l e t - 3 9 2 #(h00800) egg l e t - 3 9 4 (h00235) ear ly + + let-394* (h00262) ear ly + + -l e t - 3 9 4 #(h00350) ear ly + + l e t - 3 9 4 (h00361) ear ly + + l e t - 3 9 4 #(h00472) ear ly l e t - 3 9 4 (h00770) l e t - 3 9 4 (h00806) ear ly l e t - 3 9 1 (h00221) ear ly + + let-397* (h00228) + -let-397 (h00445) ear ly + + -l e t - 3 9 8 (h00257) mid + + -l e t - 3 9 9 (h00273) l a te + -l e t - 4 0 0 (h00243) ear ly - + let-400* (h00269) ear ly - + + + l e t - 4 0 0 #(h00793) ear ly l e t - 4 0 0 #(h00795) ear ly l e t - 5 2 0 * (h00690) la te - + + + Table 8 (con't). Gene A l l e l e Arrest hDp62 hDp58 hDp65 hDp50 hDf8 hDp83 let-521* (h00704) late + + - + l e t - 5 2 2 (h00240) early - -l e t - 5 2 2 (h00519) egg - - + + l e t - 5 2 2 * (h00735) early - + - + l e t - 5 2 3 (h00479) egg -l e t - 5 2 3 * (h00751) early - - + -l e t - 5 2 4 (h00360) early - -l e t - 5 2 4 (h00406) early - -l e t - 5 2 4 * (h00442) early - - + + l e t - 5 2 4 (h00444) early - -l e t - 5 2 5 (h00410) mid - -let-525* (h00874) mid - + -let-527* (h00207) early - - + + let-527 (h00357) adult - - + + let-528* (h01012) late + + - -let-529* (h00238) early + + - + l e t - 5 2 9 (h00249) early + + l e t - 5 2 9 (h00342) early + + l e t - 5 2 9 (h00516) early + + + l e t - 5 3 4 (h00196) mid + + l e t - 5 3 4 * (h00260) mid + + - + l e t - 5 4 3 * (h00792) l a t e + + - - + l e t - 5 4 4 * (h00692) adult + + - - + let-545* (h00842) adult + - - + let-605* (h00312) adult + + - - + l e t - 6 0 5 (h00486) adult -l e t - 6 0 5 (h00533) adult + + -l e t - 6 0 5 (h00537) adult + + l e t - 6 0 5 (h00707) adult + -68 Table 8 (con't). Gene A l l e l e Arrest hDp62 hDp58 hDp65 hDp50 hDf8 sem-004 (h00769) raid - + + + sem-004* (n01378) viable unc-014* (e00057) viable - - + unc-037* (e00262) viable + + -unc-037 (S00080) early- + -unc-037 (h00763) adult + + -unc-087* (e01216) viable + + -* The canonical a l l e l e , the a l l e l e which defines the gene and upon which a l l e l i s m was based, # Mutations assigned to complementation group by A. M. Rose (personal communication). a Most of the arrest data from McDowall (1990) with additions from t h i s study. Lethal dpy-5 l e t - x unc-13; hDp worms (phenotypically Unc-13) were measured. Sizes used to determine arr e s t stage were as follows: Early l a r v a l = L1/L2 = 0.15-0.3mm Mid-larval = L2/L3 = 0.3-0.4mm Late l a r v a l = L3/L4 = 0.4-0.6mm Adult stage = 0.6-0.8mm b Spontaneous a l l e l e . The following a l l e l e s were tested with hDpl9 and found to be rescued by i t : hl89, h203, h248, h261, h272, h275, h346, h347, h348, h473, h485, h496, h513, h517, h531, h535, h536, h540, h577, h872. 69 The hDpl 9 - hDp62 interval has no useful breakpoints from other rearrangements. In these experiments, the right breakpoint of the deficiency hDf8 was not differentiated from the right breakpoint of hDp62. Each mutation rescued by hDp62 failed to complement hDf8. The left breakpoint of hDf8 has been shown by McDowall (1990) to be in the hDpl 6 - hDpl9 interval. Only the breakpoints of hDp33 and hDp35 fall in this region but are separated from hDpl 9 by only one gene, let-392. It was possible, however, that hDp!9, hDp33 and hDp35 all had the same breakpoints and hDpl 9 carried a point mutation in the let-392 gene. At the right end of this region, the hDp83 is breakpoint to the right of dpy-14 but to the left of let-86 (Figure 24). hDp83 is a spontaneous derivative of hDp26 (section 3.13.2). I confirmed the let-86 mapping with a three factor experiment. From the progeny of dpy-14 let-86 unc-13/ + + + hermaphrodites I recovered 3 non-lethal Dpy-14 and 3 non-lethal Unc-13 recombinants. These data placed let-86 between dpy-14 and unc-13. let-529 was tentatively mapped to the same region as let-86 because hDp83 partially rescued let-529(h238). While dpy-5 let-529(h238) uncl3 homozygotes were early larval lethals, dpy-5 let-529(h238) unc-13; hDp83 hermaphrodites were sick, slow growing and either sterile or produced few progeny. Two other let-529 alleles, h342 and h516, were also poorly rescued by hDp83. Hermaphrodites with the h516 allele rescued by hDp83, however, were noticeably more fertile than strains with either of the other two alleles. Why hDp83 partially rescues let-529 mutants is not clear. If let-529 were outside the region duplicated by hDp83, perhaps the duplication of nearby genes partially suppressed the lethal phenotype. Alternatively, let-529 could be in the region duplicated by hDp83, but expressed at an abnormally low level. Reduced expression of the let-529(+) product could have resulted from the hDp83 breakpoint being in the let-529 gene, but only partially inactivating it. Alternatively, the low level of let-529 expression may be a position effect. The visible genes dpy-14, unc-37 and unc-87 were also mapped to the hDpl9 - hDp62 region. Mutations of these genes were complementation tested against representative alleles of the essential genes. Only unc-3 7 was found to have lethal alleles. It was originally defined by the viable allele e262 (Brenner 1974) which causes an Unc phenotype when homozygous. I have 70 found two lethal mutations, s80 and h763, which failed to complement the Unc phenotype of unc-37(e262). The h763le262 and s80/e262 heterozygotes were fertile Unc worms. One allele, s80, was the only allele of let-76 (Rose and Baillie 1980). The unc-37 designation will be retained and the let-76 designation will no longer be used. Both as a homozygote and heterozygote to hDf8, s80 had an early larval arrest phenotype. These data suggested the null phenotype of unc-37 was early larval lethality. In contrast, h763 homozygotes arrested development as adults. The h763/hDf8 hermaphrodites had uncooridinated movement and arrested development at a late larval stage. In addition, the viable allele e262 was also viable when heterozygous to hDf8. These data suggested the null phenotype was uncoordinated movement and late larval lethality. The reason s80 arrested development at an early larval stage may have been because it was induced on a dpy-14(el88) chromosome. Perhaps the Dpy-14 phenotype increased the severity of the lethal phenotype in s80/s80 and s80/hDf8 worms. The lethal mutations could have other phenotypes in addition to lethality. Morphological and behavioral phenotypes of the lethal mutations would not have been observed, however, because the lethal mutations were induced on chromosomes marked with dpy-5 and unc-13. The Dpy and Unc phenotypes of the lethal homozygotes masked other phenotypes. Another ramification of inducing lethal mutations on marked chromosomes was the potential for of synthetic lethals; mutations which were lethal only in the presence of Dpy-5 and/or Unc-13 phenotypes. The complementation testing protocol with hDf8 addressed these two possibilities with lethal mutations in the hDp 19 - hDp62 interval. The crosses with hDf8 were different from the other complementation tests because the let-x/hDf8 progeny were wild-type instead of being marked with an Unc-13 or Dpy-5 phenotype (section 2.4). All lethal mutations in the hDpl9 -hDp62 region, failed to complement hDf8. In this region, therefore, there was no evidence for synthetic lethals. In most of the cases when the lethal mutation failed to complement AD/8, the time of arrest of the wild-type letlDf was no earlier than in the let homozygote. These lethal mutations could be amorphs. As already described, let-385(h748) was classified as a hypomorph based on the deficiency test. 71 Because the let-x/hDfS worms were wild-type, behavioral phenotypes could be observed. Hermaphrodites of the genotype let-605(h312)/ hDf8 were sterile adults, had a distinctive coiling Unc phenotype and had a vulval "blip". Homozygous h312 worms were sterile adults in addition to the Dpy-5 Unc-13 phenotype. Three more alleles of let-605, h486, h533, and h.707, also had the sterile adult Unc phenotype in trans to hDf8. The Fifth allele, h537, was not tested. Because four alleles had the same phenotype, the Unc phenotype was probably not due to a second unc mutation in the hDf8 region but was a property of the let-605 locus, let-605 may be important for neurological function as well as viability. In the hDp62 - hDp58 interval four genes have been identified. hDp65 covers this region, but its breakpoint was not been distinguished from hDp62. hDp65 was useful, however, for finding and analyzing a double hit in the strain KR440. let-389 was originally defined by KR440 and the mutation was designated hi06 (Howell et al. 1987). Mapping of hl06, however, was complex. hl06lhl06; hDp65 worms arrested as late larvae, while hl06lhl06; hDp62 worms arrested as early larvae. In addition, hl06/hDf8 worms arrested as late larvae. These data could be explained with a double hit chromosome: one mutation in the hDp 19 - hDp62 interval causing late larval lethality, and the second mutation in the hDp62 - hDp58 interval causing early larval lethality. This proposal was confirmed as ten other lethal mutations failed to complement the lethal chromosome in KR440. All of these mutations mapped to the hDp62 - hDp58 interval. In future, the late larval mutation in the hDpl9 - hDp62 region will be referred as mutation hl012 of the gene let-528. let-389 has been redefined with the canonical allele h428. The let-389 mutation in KR440 retains the designation hi06. An eleventh let-389 allele, h997, was isolated in a lethal screen using a translocation (section 3.5). Homozygotes of ten let-389 alleles caused development to arrest at an early larval stage. The eleventh allele, h698, caused worms to arrest development at a late larval stage. This allele also had the interesting characteristic of partially complementing at least one let-389 allele. h698/hl06, h698/h680 and h698/h997 heterozygotes arrested development at an early larval stage. In contrast, h698/hl90 hermaphrodites were sterile L4/adult hermaphrodites and 72 h698/h428 heterozygotes were fertile, albeit less so than if full complementation was observed. Other allele combinations with h698 were not tested. The deletion nDf(n2088) (see below) was used as a null allele of let-389 to test if h698 was a hypomorph. The nDf(n2088)lh698 heterozygotes arrested development as early larvae, supporting the view that h698 was a hypomorph. The mutations hi06, h680 and h997 which, when heterozygous to h698, produced an early larval arrest phenotype, may be similar to nDf(n2088) and thus null mutations of let-389. Using this criteria, h428 and hl90 are not null mutations. Because the h428lh698 worms were healthier than either homozygote, the products from let-389(h428) and let-389(h698) must have complementary functions. Another gene in the hDp62 - hDp58 region was sem-4. Mutations in this gene produce an egg laying defective (Egl) phenotype (M. Basson, personal communication). M. Basson (personal communication) mapped sem-4(nl378) to the hDp62 - hDp58 interval. He also isolated a lethal mutation, n2088, that failed to complement nl378. I have done complementation testing with nl378 and n2088 against the lethal mutations in the hDp62 - hDp58 region. Mutations in three essential genes in the region failed to complement n2088. Heterozygotes of n2088 with let-389(h428), let-400(h269) and let-520(h690) produced Dpy progeny that arrested development at about the same time as the let homozygotes. In contrast, these three lethal mutations fully complemented the Egl defect of nl378. For these reasons, I believe n2088 is a deletion and sem-4 is a different gene than these three essential genes. nDf(n2088) complemented let-86 and dpy-14 on the left and let-522(h735), let-523(h751) and let-524(h874) on the right. h769l n.1378 and h769l nDf(n2088) heterozygotes were viable but Egl. Thus, h769 appeared to be an allele of sem-4 because of the failure to complement nl378 for the Egl phenotype. The viability of the h769/n2088 heterozygotes was similar, if not better, than h769lh769 homozygotes. The dpy-5 h769 unc-13 homozygotes were poorly fertile and their progeny did not grow past a late larval stage. It was possible the Unc-13 phenotype decreased the viability of h769, making them sicker than the h769/n2088 heterozygotes which had a Dpy-5 phenotype. 73 In the hDp62 - sDp2 interval six essential genes were identified, unc-14(e57) was also mapped to this interval but complemented representative alleles from the essential genes. This region was divided into two groups by the hDp50 breakpoint (Figure 12). The homozygous phenotype of one mutation in this region acted differently depending on the genetic background of the hermaphrodite parent, dpy-5 let-527(h357) unc-13 homozygotes segregating from the dpy-5 let-527(h357) unc-13; sDp2 strain arrested development as early larvae. The same Dpy-5 Unc-13 hermaphrodites segregating from dpy-5 let-527(h357) unc-131+ + + heterozygotes were sperm defective adults (Spe) and produced viable oocytes. That is, the Dpy-5 Let-527 Unc-13 homozygotes were fertile if crossed to wild-type males. The presence of a duplication per se was not responsible for increasing the severity of the h357 phenotype; Unc-13 progeny segregating from dpy-5 let-527(h357) unc-13/ dpy-5 dpy-14; hDp58 hermaphrodites were also sterile and Spe. These data are consistent with a maternal effect of the let-527 locus. For example, the wild-type let-527 product may not be equivalent in either quality or quantity in h357/h357/Dp[+'] hermaphrodites compared to h357l+ hermaphrodites. Alternatively, the h357 allele could have an antimorphic quality and maternal expression of two copies of this allele increases the severity of the lethal phenotype from sperm defect to early larval lethality. Another allele of let-527, h207, was identified. The h357/h207 heterozygotes were sperm-defective adults. While the phenotype of h207 homozygotes was like K357, early larval arrest, h207 did not exhibit any maternal rescue. Perhaps h207 is a more severe allele of let-527 than K357. Unfortunately there were no deficiencies in the region to test if h357 was a hypomorph. A third allele of let-527, h932, was identified (section 3.5) and had the same properties as h207. Three gamma radiation induced lethal mutations were mapped to the hDpl 9 - sDp2 interval (section 3.2). In each case the lethal mutation failed to complement alleles of only one gene. None of the three were found to be deletions. h53 failed to complement let-524(h422), h58 was an allele of let-529(h238) and h590 was found to be an allele of let-400(h269). 74 3.5 Lethal mutations isolated with the translocation hTl. As described in the Introduction, the duplication approach to essential gene analysis has many attractive features. One potentially serious drawback of this approach, however, is that duplications may not rescue all recessive lethal mutations. Whereas let-x/ + is viable, let-x/let-x; Dp[let-x(+J] may be lethal. This possibility affects the analysis of chromosome I in two ways. First, if sDp2 did not rescue the majority of lethal mutations in the region it duplicates, then attempting to identify most of the genes in the region would be impossible. Second, using duplications to map such mutations would be misleading. Mutations which cannot be rescued by a duplication would be mapped to the wrong location. To address the possibility that a class of lethal mutation was not being recovered with the sDp2 system, a screen for lethal mutations was done using the translocation hTl(I;V) as a balancer (Howell and Rose 1990). Lethal mutations were recovered in the region between the left end of chromosome I and unc-29, the recombination suppressed region in hTl heterozygotes (section 3.6.2). Forty-four mutations from this screen that were also balanced by szTl(I;X) were kept for further analysis. Howell (1989) divided 41 of these mutations into two groups based on coverage by hDpl3. Howell also tested the 41 mutations for complementation with hDfS. She found four mutations in hDfS, and all four were rescued by hDpl 3. Thus there was no evidence for a large class of lethal mutation in hDf6 that could not be rescued by a duplication. Unfortunately hDfS is small relative to the sDp2 region and may have restricted the chances of detecting a new class of lethal mutation. For a more complete analysis, I needed to know the genetic position of all 44 lethal mutations. Then the possibility of rescue by a duplication could be determined. The 19 mutations outside of hDpl 3 plus the three not mapped by Howell were tested with sDp2 and hDpl 9 by myself and A. M. Rose (personal communication) (Table 9). Nine mutations were rescued by hDpl9 and five more were rescued by sDp2 but not by hDpl9. Obviously this group of 14 mutations could be recovered in a duplication based screen for lethal mutations. The remaining Table 9 Duplications mapping of hTl balanced l e t h a l mutations. Gene 3 and A l l e l e Arrest hDpl9 hDp62 hDp58 sDp2 hDp70 nDp4 hDp38 hDp42 hDp60 hDp36 hDp43 hDp64 hDp73 let-400(h00878) early - + + let- (h00879) +*b let-536(h00882) mid —* _* - + let-540(h00884) l a t e _* - + + -let- (h00885) + + + let-541(h00886) early - - - + -let- (h00931) + * let-527(h00932) early — * - +* let- (h00933) + let-539(h00938) mid - - - + + + + let-385(h00940) early + + + + let- (h00984) + let-542(h00986) e a r l y — * + + - - + let-538(h00990) adult - - - + let- (h00991) + + + + let-535(h00993) mid -* - + let- (h00994) + + Table 9 (con't) Gene 3 and A l l e l e Arrest hDpl9 hDp62 hDp58 sDp2 hDp70 nDp4 hDp38 hDp42 hDp60 hDp36 hDp43 hDp64 hDp73 let-389(h00997) e a r l y — * — + +* let-537(h00999) e a r l y - - + let-520(h01003) l a t e - + + + + let- (h01005) + + let- (h01007) + + a Gene names assigned to genes i n hDpl9 - sDp2 region. b * =• data from A. M. Rose (personal communication). 76 Table 10 Def ic iency mapping of n i l balanced l e t h a l mutations. Gene and A l l e l e sDf6 hDf8 hDf9 nDf23 nDf24 nDf25 sDf5 let-536 (h00882) + let-540 (h00884) - + + + + let-541 (h00886) + let-385 (h00940) let-542 (h00986) + + let-538 Sh00990) + Let-535 (hOQ993) + let-537 (h00999) + let-520 (h01003) + 77 eight mutations were either outside sDp2 or were within the sDp2 region but not rescuable in a let-x/let-x; Dp[let-x(+)] strain. To test if these eight mutations mapped within the sDp2 region, they were mapped with respect to other duplications and deficiencies in the region (Figure 12) between the sDp2 breakpoint and the hTl crossover suppression boundary. Two duplications known to cover part (hDp70; Figure 7) or all of the region (nDp4; Figure 7) were used to determine if the lethal mutations could be complemented by a duplication. Three were rescued by hDp70 and all eight were rescued by nDp4. Therefore, no mutations were inside the sDp2 region and lethal in let-x/let-xlDp[let-x(+)] genotypes. More precise mapping of these eight mutations (Figure 12 and Tables 9 and 10) has shown they all map to the right of the sDp2 breakpoint. All eight mutations defined new genes (Table 9). This was not surprising since there have been no large scale lethal mutation screens in the region to the right of sDp2. The group of lethal mutations mapping between hDpl 9 and sDp2 were mapped and complementation tested to the rearrangements and complementation groups identified in the dpy-14 region. Three of the mutations mapped to the hDp62 - hDp58 region; all three failed to complement a preexisting lethal mutation (Table 9). One mutation, h932, mapped to the hDp58 -sDp2 interval and was found to be an allele of let-527(h207). Only two of the five mutations were in the hDpl 9 - hDp62 interval. Two mutations were not efficiently rescued by duplications. H940 failed to complement KDf8 but h940/h940;Dp strains with hDp62, hDp58 or hDp64 were rarely fertile. For example, 1/10 Unc-13 set from dpy-5 h940 unc-13ldpy-5 dpy-14;hDp64 hermaphrodites were fertile, the other nine were mid-larval lethal. h940 was complementation tested against lethal mutations in the hDf8 region and was found to be an allele of let-385. With this information, hDp64;h940lh940 hermaphrodites should have been fertile. hl003 was similar to h940 although the effect was not as severe. Only one of five Unc-13 hermaphrodites from dpy-5 hi 003 unc-13/dpy-5 dpy-14;sDp2 hermaphrodites were fertile. The rest were sterile adults. When seven F2 Unc-13 progeny from the single fertile FI were tested, 78 three were sterile. Interestingly, hi 003 failed to complement h690, the only other allele of let-520. In contrast to these results, both h940 and hi003 were efficiently rescued by nDp4 and hi 003 was efficiently rescued by hDp58. These effects were probably not due to second site semi-sterile mutations because when either h940 or hi 003 were derived from a rare fertile let;Dp strain, the same effect was observed. 3.6 Translocations In the previous section, rearrangemented chromosomes, primarily duplications, were used to investigate the organization of essential genes. Now I will describe the structure and behaviour of a variety of chromosome rearrangements. This analysis was aimed at defining the localized sites and structures which influence chromosome behaviours such as recombination and segregation during meiosis and mitosis. The first group of rearrangements analyzed were four translocations involving chromosome I. 3.6.1. szTl(I;X) The translocation szTl(I-;X) was initially isolated by Fodor and Deak (1985) as a dominant X-chromosome crossover suppressor. It was induced on a lon-2(e678)(X) chromosome using 7000 rad of X-radiation. Their analysis showed that szTl consists of two abnormal chromosomes derived from the normal chromosomes I and X, that the homozygotes arrest as embryos, that heterozygous hermaphrodites produce Lon male self-progeny at a frequency of 0.08 - 0.12 and recombination was reduced in the dpy-7 - unc-3 (X) interval from 19.2 m.u. to 0.3 m.u. To characterize this translocation further, the physical breakpoints were genetically mapped and the relationships of these breaks to segregation and recombination in the translocation heterozygote were determined. To summarize the findings, szTl(I;X) is composed of two chromosomes, ILXLszTl and IRXRszTl. Figure 13a shows the arrangement of these chromosomes. The IRXRszTl chromosome recombines with and segregates from normal chromosome I and I^X^szTl recombines with and segregates from the X-chromosome. The experiments leading to these conclusions are described below. 79 i i _ J I J I I a) szTtihX) R R I X szT1 L L I X szTl b) hTKhV) R L I V hT1 ZD ILATI c) hT2(l;lll) R R I III hT2 L L I III hT2 d) hT3(l;X) R R I X hT3 L L I X hT3 Figure 13: Diagram of the chromosomes comprising (a) szTl(I;X), (b) hTl(I;V), (c) hT2(I;III) and (d) hT3(I;X). The orientation of the translocation arms is not known; they are drawn in a manner requiring the least number of breaks. For example, the X R piece of szTl(I;X) could be inverted, but this would require at least two more breaks, one at the tip of the X chromosome and another to cap the unc-20 end. 80 Characterization of the I^X^szTl chromosome. Hermaphrodites of the genotype dpy-5 unc-13; unc-31 szTl(I;X)[+; lon-2] (strain KR900) segregated wild-type hermaphrodites (1009), Lon males (71), Dpy-5 Unc-13 Unc-3 (222 hermaphrodites and 19 males) and Unc-3 (50 hermaphrodites and 17 males) worms. Among the progeny of dpy-5; unc-31 szTl(I;X) hermaphrodites, the absence of Dpy-5 (non-Unc-3) progeny and the low frequency of Unc-3 progeny confirmed Fodor and Deak's (1985) findings that the progeny of szTl heterozygotes show a high degree of pseudolinkage between dpy-5 and unc-3. The presence of a few Unc-3 progeny showed that the pseudolinkage was not 100%. Since the pseudolinkage phenomenon of translocation heterozygotes is the result of both (a) the inviability of aneuploid zygotes and 0D) the suppression of recombination between the test genes and their translocation breakpoints, the Unc-3 progeny could have been either viable aneuploid progeny or the result of recombination. The latter was unlikely since crossing-over would have resulted in both Unc-3 and Dpy-5 progeny from the dpy-5; unc-31 szTl(I;X) heterozygotes. As described below, the genotype of the Unc-3 progeny was found to be dpy-5 (unc-13); unc-3 ; ILXLszTl (i.e. hyperploid for I^1 and X^-1). As one of the translocation chromosomes in szTl(I;X), I^X^szTl was used to map the translocation breakpoints (see below). In the szTl(I;X) segregation (Figure 14), some of the Unc-3 males could have resulted from normal segregation, but the Unc-3 hermaphrodites required a nondisjunction event (e.g. 3:1 segregation). The experiments leading to these results are described below. As shown in Figure 14, the Unc-3 males were expected to equal half the number of wild types. This was not observed probably because of the slow development of these hyperploid males, even when compared to their hyperploid hermaphrodite siblings. Investigation of the presumed aneuploids was done with a strain similar to KR900, but carrying dpy-7 instead of unc-3 (i.e. KR1118; dpy-5 unc-13; dpy-7/ szTl(I;X) [ + ;lon-2]). This strain segregated Dpy-7 progeny. If the Dpy-7 hermaphrodites were dpy-5 unc-13ldpy-5 unc-13; dpy-7/ dpy-7/ ILXLszTl, then in crosses to wild-type males, all the F I wild-type hermaphrodites should have been heterozygous for the dpy-5 unc-13 and dpy-7 chromosomes and segregated Dpy-5 Unc-13 and Dpy : 7 F2 progeny; this was in fact observed. Moreover, among the F l progeny, 81 3 ? I N X N 1 N j L x L j R x R X N j R x R j L x L I N X N j L x L I N j R x R X N I L X L j R x R I N x N Dpy-5 Unc-3 Unc-3 6" D p I L X L Wt Unc-3 Dpy-5 Unc-3 6* Dpy X X X i N Unc-3 o" D p I L X L Wt Lon cT Wt Dplhc1-Lono' j R x R + X N Wt Dpy X X X j R x R I L X L Wt Lon d* D p I L X L Wt DpI^C1-Lon 6* ,N X N j L x L Unc-3 D p I L X L Dpy X X X Wt D p I L X L Unc-3 6* DpI L X L Wt I N Dpy-5 Unc-3 6* Lon<? Unc-3 c? DpI L X L Wt j R x R X N j L x L Dpy X X X Wt D p I L X L Wt j R x R Lon o" Wt Figure 14: Punnett square showing the predicted segregation from a dpy-5; unc-3/ szTl(I;X) heterozygote. In addition to the alternate and adjacent-a segregations (see Results), 3:1 segregations from nondisjunction of the X chromosome and ^ X^szTl are shown. Empty boxes represent zygotes presumed lethal due to severe aneuploidy. Viable classes are indicated by phenotype and any aneuploidy they may carry. All viable zygotes are hermaphrodites unless indicated as male. The triplo-X worms have a weak Dpy phenotype. 82 1/2 were expected to receive ILXLszTl; these were expected to segregate wild-type and Dpy-7 to Dpy-5 Unc-13 progeny in an 11:1 ratio (self fertilization of dpy-5 unc-13/ + + ; +1 dpy-7/ ILXLszTl hermaphrodites assuming the Dpi Dp worms died). From 25 tested hermaphrodites, 12 were observed to give high wild-type+ Dpy-7:Dpy-5 Unc-13 ratios, and the numbers observed (1853:180) were consistent with an 11:1 ratio. All 12 ILXLszTl carrying wild-type hermaphrodites segregated Dpy-7 hermaphrodites, and were therefore dpy-5 unc-13/ + +; dpy-71 + IILXLszTl. The dpy-5 unc-131 + + ; +IILXLszTl sibling worms were probably males. In summary, the Dpy-7 progeny segregating from KR1118 or the Unc-3 progeny from KR900 carried one half of szTl(I;X) (I^X^szTl) in addition to the normal diploid chromosome complement. This analysis also shows that ILXLszTl carries dpy-5(+) and unc-13(+) but not dpy-7(+) or unc-3(+), The I^X^szTl duplication chromosome induces X-nondisjunction. The Dpy-7 aneuploid hermaphrodites (dpy-5 unc- 13; dpy-7; ILXLszTl) produced Dpy-7 non-Dpy-5 non-Unc-13 males (dpy-5 unc-13; dpy-7/O; I^X^szTl) as well as hermaphrodites (772 Dpy-7 hermaphrodites, 59 Dpy-7 males and 558 Dpy-5 Dpy-7 Unc-13 hermaphrodites and males). Duplication carrying wild-type males in the progeny of dpy-5 unc-13; + lszTl(l;X)[+ + ;lon-2] hermaphrodites did not mate, a result comparable to that reported by Rose, Baillie and Curran (1984) for sDp2(I;f) males. sDp2 is also a duplication of the left end of chromosome I (Figure 1). Mapping the szTl breakpoints. 1) Chromosome X: The presence of the aneuploid Dpy-7 and Unc-3 segregants from the respective szTl heterozygotes showed that ILXLszTl carries neither dpy-7 nor unc-3. Finer mapping of the X-chromosome breakpoint was done by crossing dpy-5 unc-13; OlszTl(l;X)[+ +; lon-2] males to hermaphrodites homozygous for a X-chromosome marker. If wild-type male progeny were observed, ILXLszTl must have carried a wild-type copy of that gene. When these Lon males were crossed to unc-20 hermaphrodites, only Unc-20 male progeny were observed. When the same test was done using dpy-3 or unc-1 hermaphrodites, both wild-type males (i.e. dpy-310; I^X^szTl) and Dpy-3 or Unc-1 males were observed. The X-83 chromosome breakpoint is therefore between dpy-3 and unc-20, and the duplication carries XL (Figure 13a). 2) Chromosome I: Strains were constructed with markers on a normal chromosome I and tested for complementation with wild type alleles on I^X^szTl (section 2.5). The essential genes let-88 and let-80 map to the right of sDf6 (section 3.4.2) but to the left of unc-29 (Figure 12). I^X^szTl includes the wild type alleles of let-362, dpy-5, let-88 and genes deleted by sDf6, but not of let-80 or unc-29. The chromosome I breakpoint of szTl(I;X) maps between let-88 and let-80 (Figure 13). Although the translocation homozygotes die, Lon males (e.g. dpy-5 unc-13; 01 szTl(I;X)) survived. Therefore, the szTl -associated lethal mutation does not involve an X-chromosome gene (Fodor and Deak 1985). The duplication sDpl(I;f), which covers the unc-74 - unc-54 interval (Figure 1), rescued the szTl-associated lethal mutation, placing the lethal mutation to the right of unc-74. The szTl -associated lethal mutation was complemented by both nDf24 and nDf25. Assuming that the breakpoint generated the lethal site and knowing that unc-29 is within nDf24 and nDf25 (Figure 12) and to the right of the translocation breakpoint, then the szTl(I;X) breakpoint is to the left of nDf24 and nDf25. The following experiment shows and X ^ segregate together and are therefore probably linked. The IRXRszTl chromosome was isolated from szTl (I;X) by creating a segmental aneuploid with sDp2(I;f). dpy-5 unc-29; 01 szTl(I;X) males were crossed to dpy-5 unc-29; sDp2 hermaphrodites. Some of the wild-type hermaphrodites (2/15) did not segregate Lon males but did segregate Unc-29 hermaphrodites, indicating sDp2 was present; presumably in the genotype dpy-5 unc-291 IRXRszTl; sDp2; +10, and IRXRszTl carries unc-29(+). Among the progeny of the dpy-5 unc-29IIRXRszTl; sDp2; +10 hermaphrodites, no wild-type males were observed even though Unc-29 and Dpy-5 Unc-29 males were observed. When 1^ was present resulting in a wild-type phenotype, the X ^ piece was present resulting in a hermaphrodite phenotype. 84 Segregational properties of szTl. jRx^szTl segregates from chromosome I. An unc-29 marked IRXRszTl was used to determine which translocation chromosome segregates from the normal chromosome I. dpy-14 unc-29; +1 szTl(I;X) [ + unc-29; lon-2] hermaphrodites were crossed to dpy-5 unc-29; 01 szTl(I;X)[+ +; lon-2] males. The ratio of FI wild-type hermaphrodites segregating Dpy-14 Unc-29 F2 progeny to those segregating Dpy-5 Unc-29 F2 progeny depended on which chromosome unc-29 segregates from. If the half-translocation chromosome marked with unc-29(e403) (IRXRszTl) segregated independently of chromosome I in males, sperm containing both dpy-5 unc-29 and IRXRszTl would be produced. If this sperm fertilized an oocyte containing a normal X and the other half-translocation chromosome I^X^szTl, hermaphrodites segregating Dpy-5 Unc-29 progeny would be produced. Thirty-two FI wild-types were tested and all of them segregated Dpy-14 Unc-29 progeny. Therefore, the IRXRszTl component usually segregates from chromosome I because no Dpy-5 Unc-29 segregating FI progeny were recovered. Since the chromosome I breakpoint lies between unc-13 and unc-29 and alternate and adjacent-a segregation are equal in hermaphrodites (see below), the other half translocation, ILXLszTl, must segregate independently of chromosome I in hermaphrodites. Since I^X^szTl does not carry dpy-7 or unc-3, these genes must be on IRXRszTl (see below), dpy-5 unc-13; dpy-710; I^X^szTl are not only Dpy-7 in phenotype but also male, therefore I^X^szTl does not carry enough of the X chromosome to produce a hermaphrodite phenotype. Alternate and adjacent-a are equal in hermaphrodites. The traditional terms adjacent-1 and adjacent-2 (see Rickards 1983 for review) cannot be used with C. elegans because the location of a centromere on each chromosome has not been mapped. Therefore, I refer to I R segregating from 1^ (and 1^  segregating from X^) as adjacent-a and I R segregating from X ^ (and I^ 1 segregating from 1^ ) as adjacent-b. Alternate segregation is when the two translocation chromosomes segregate together. I^X^szTl recombines with the X-chromosome (see below), but there is no information concerning its segregation with respect to the normal chromosome I. Several lines of evidence 85 suggest that alternate (I^X^szTl and IRXRszTl segregate to the same pole) and adjacent-a ([LXLSZTI a n a jN segregate to the same pole) occur at equal frequencies during hermaphrodite meiosis. First, when recombination was scored on chromosome I in a small interval including the szTl(I;X) breakpoint, one recombinant class was expected to be recovered at twice the frequency of its reciprocal if alternate and adjacent-a segregations were equal and no aneuploids survived (see Figure 5 and section 2.7). In the experiments reported in Table 12 and in McKim, Howell and Rose (1988), the observed ratio of reciprocal recombinant progeny was 176/393. Compared to the expected ratio of 190/379, the results of these experiments agree with the 2:1 prediction. The egg survival frequency and progeny ratios also support the proposal that alternate and adjacent-a segregations occur at equal frequency. Figure 14 shows the expected egg survival frequency is between 5/16 (31%) and 9/16 (56%) depending on the survival of aneuploid progeny. I observed 34% (37 adult hermaphrodites from 110 eggs), which is similar to the egg survival frequency (31%) observed by Fodor and Deak (1985). In addition, a dpy-5 unc-13; unc-3/ szTl(l;X) hermaphrodite should have segregated wild-type and Dpy-5 Unc-13 Unc-3 hermaphrodites in a 4:1 ratio (Figure 5), if there were no 3:1 segregations. I observed 1009:222 or 4.5:1. If alternate and adjacent-a were not equal, the egg survival frequency would have been higher and the wild-type:Dpy Unc ratio would have been smaller. If adjacent-b segregation also occurred with equal probability, the expected results would have been 22% for egg survival and a 6:1 ratio for wild-type to Dpy-5 Unc-13 Unc-3 progeny. As noted above, there was a low frequency (about 10%) of aberrant 3:1 segregations from X-nondisjunction. That is, sperm or oocytes were formed which carried a normal X and the I^X^szTl chromosome. This would inflate the number of wild-type progeny in the above ratios and could partly explain the deviations from a 4:1 ratio. Segregation in szTl(I;X) males. To test segregation patterns in the male, 40 wild-type hermaphrodite progeny from crossing dpy-14; 01 szTl(I;X)[+;lon-2] males to dpy-5 unc-29; +1 szTl(I;X)[+ unc-29;lon-2] hermaphrodites were tested by observing their progeny (F2). If the ILXLszTl chromosome segregated randomly as a univalent with respect to IRXRszTl and 86 chromosome I, then FI wild types segregating either Dpy-5 Unc-29 or Dpy-14 progeny were expected in equal numbers. If I^X^szTl segregated from 1 ,^ then only FI wild-types segregating Dpy-5 Unc-29 progeny were expected. Twenty-seven FI hermaphrodites were dpy-14; + lszTl(I;X)[+ unc-29;lon-2] and 13 were dpy-5 unc-29; +1 szTl(I;X)[+ + ;lon-2]. Assuming alternate and adjacent-a segregations were equal in the hermaphrodite, it appears that I N ; I^X^szTl gametes were produced twice as often as 1 ;^ O gametes. This relative excess 0flN; I^X^szTl gametes formally implies that I^X^szTl segregates from IRXRszTl. This is surprising considering the presumed absence of homology between the two szTl chromosomes. This happens only in males, however, where there is no X homologue. A similar result was obtained in a different experiment in which dpy-5 unc- 291 IRXRszTl; sDp2; + 10 hermaphrodites (see above for their construction) were crossed to dpy-5 unc-13; 01 szTl(I;X) males. Thirteen wild-type progeny were tested (by allowing them to self-fertilize) and eleven were dpy-5 unc- 29; +1 szTl(I;X). The IRXRszTl chromosome in these hermaphrodites must have come from the maternal parent. The other two wild-types were dpy-5 unc-13; +1 szTl (I;X); these hermaphrodites got their IRXRszTl from the paternal parent. As observed above, the bias in recovering one of the classes of wild-type could be explained by proposing chromosome I and I^X^szTl tend to segregate together in the male germ line. S p o n t a n e o u s a t t achmen t o f I^X^szTl a n d the X - c h r o m o s o m e . Two more unusual segregant classes were detected when the progeny of 1104 untreated wild type hermaphrodites from the strain KR900 [dpy-5 unc-13; unc-3/ szTl(I;X)] we examined. Thirty-four strains were recovered which failed to segregate Dpy-5 Unc-13 Unc-3 progeny, as if a lethal mutation had been induced in the balanced region of the szTl heterozygotes. These strains did not, however, behave like a standard szTl strain. Of the 34 strains, 28 segregated approximately one-third of their progeny as Lon males ("class 1") and 6 segregated no Lon males ("class 2"). This segregation pattern is not characteristic of normal lethal bearing szTl strains (—10% Lon males). In addition, the class 1 strains often reverted to produce strains similar to the original KR900. 87 Both class 1 and class 2 strains can be explained with one chromosome structure derived from I^X^szTl. The evidence for these conclusions are described below. The new chromosome is hypothesized to carry at least a full complement of X material, for example a fusion ILXLszTl and the normal X (X^) to produce a I^X^X^szTl chromosome. Class 1 strains would then be dpy-5 unc-13/ IRXRszTl; lLXLszTll ILXLXNszTl, and the class 2 strains would be dpy-5 unc-131 IRXRszTl; ILXLXNszTl. In the original isolation of these strains, the class 1 strains were isolated more frequently (28:6). This is consistent with the proposal that the class 2 zygotes were derived from the fertilization of a nullo-X gamete by one carrying the ILXLXNszTl chromosome. This is rare because the nullo-X gamete results from a nondisjunction event in the szTl heterozygote. Class 2 worms were healthy looking wild types while the class 1 worms were thin and clear. This phenotype for the class 1 hermaphrodites was consistent with them having a lL duplication. There are, of course, other possible explanations for these results. For example, an expected segregation product from KR900 is dpy-5 unc- 13/IRXRszTl; unc- 31 ILXLszTll I^X^szTl. With the added assumption that the I^X^szTl chromosomes regularly segregate from each other, this hermaphrodite would have a class 1 phenotype. The derivation of a class 1 strain from a class 2 strain described below, however, does not support this hypothesis because a class 2 hermaphrodite cannot be derived from a dpy-5 unc-13/ IRXRszTl; unc-3/ ILXLszTll ILXLszTl hermaphrodite. Analysis of the class 2 strains suggested the normal X-chromosome had become physically linked to one of the translocation chromosomes, forming a compound chromosome. When the class 2 strains were crossed to dpy-5/ + ; +10 males, all Dpy-5 progeny were males. The origin of the X-chromosome in these males was shown by crossing the class 2 strains to dpy-710 males. All the male progeny from this cross were Dpy-7. The abundant patroclinous Dpy-7 males most likely resulted from 100% nondisjunction of the maternal X-chromosome centromeres. The failure to recover Dpy-5 hermaphrodites in the initial cross indicated the presumed compound chromosomes segregated away from the normal chromosome I. j 88 The abnormal chromosome in the class 2 strains was the same chromosome which caused the class 1 phenotype. This was shown by crossing unc-11 dpy-14; 01 szTl(I;X) males to a class 2 hermaphrodite strain and observing the F l progeny. Twenty-five FI progeny were individually set, eight of these segregated the class 1 phenotype and 17 segregated the class 2 phenotype. Thus a class 1-type strain could be derived from a class 2 strain. The failure to recover a strain which segregated Dpy Unc progeny indicated there was no normal X-chromosome in either the male or the hermaphrodite of the cross. The abnormal chromosome was shown not to be an altered form of IRXRszTl by .testing five of the eight class 1 strains for the presence of a dpy-5 unc-13 chromosome. Since three of the five carried this chromosome of maternal origin, their IRXRszTl chromosome must have been of paternal origin and not from the original class 1 strain. This test was done by crossing the new class 1 strains to dpy-5 dpy-141 + + males. In these crosses, the class 1 strains segregated rare Dpy males in their progeny, presumably the result of nondisjunction of the two translocation chromosomes containing portions of the X chromosome. Recombination suppression. Fodor and Deak (1985) observed in szTl(I;X) heterozygotes that recombination is reduced from 19.2 to 0.3 m.u. in the dpy-7 - unc-3 (X) interval. I have found that this suppression occurs in the dpy-3 - unc-20 and unc-20 - dpy-8 intervals (Table 12). Recombination was also measured in strains that were heterozygous for szTl and chromosome I markers. No recombination was observed from let-362 (the left end) (Howell 1989) to let-88 (0.4 map units to the right of unc-13) (Table 12). Recombination was observed between unc-13 and let-80, indicating that the boundary of crossover suppression lies between let-88 and let-80. This boundary appears to coincide with the physical breakpoint of the translocation (Figure 13). Recombination enhancement. Recombination frequencies were assayed in chromosome I intervals to the right of the szTl(I;X) breakpoint. Recombination in the 1 m.u. interval adjacent to the breakpoint was increased three-fold in szTl heterozygotes (McKim, Howell and Rose 1988; Howell 1989). This map expansion extended into the unc-29 unc-75 interval which increased two-fold from 5.9 m.u. [dpy-5 unc-75 distance (Table 11); 8.9 - dpy-5 unc-29 distance (McKim, Howell 89 Table 11 Recombination frequency in control crosses. Genotpye Wt. Recombinants m.u.(C.L) let-362 dpy-5 unc-13/+ + +a 1406 149 Dpy-5 Unc-13 15.4 (13.0-17.9) 20 Unc-13 1.9 (1.2-2.9) bli-3 unc-11/ + + 723 86 Unc-11 17.5(13.5-21.9) dpy-5 unc-1011 + + b 889 66 Unc-101 12.0(10.1-14.0) 79 Dpy-5 dpy-14 unc-101/ + + 529 47 Unc-101 13.1(9.3-17.1) dpy-5 unc-75/ + + 1767 108 Unc-75 8.9(7.6-11.2) unc-101 unc-541 + + 874 84 Unc-101 14.2(11.3-17.3) dpy-17 unc-64/ + + 859 125 Dpy-17 22.3(19.2-25.4) 136 Unc-64 unc-45 dpy-171 + + 727 117 Dpy-17 23.7(20.2-27.5) 119 Unc-45 unc-60 dpy-111 + + c 2820 349 Dpy-11 17.9(16.1-20.3) dpy-11 unc-51l+ + 1362 254 Dpy-11 27.3(23.4-31.4) unc-1 dpy-7i + + 1996 275 Dpy-7 19.7(18.1-21.4) 262 Unc-1 dpy-7 unc-31 + + 1186 149 Dpy-7 19.1(16.8-21.5) 160 Unc-3 + unc-7 Un-15 + /unc-3 + + let-2 858 14 Unc-7 1.6(0.9-2.6) a Data from Howell et al. (1987). b Data from M.C. Zetka (personal communication). c Data from McKim et al. (1987). Table 12 Recombination in szTl (I;X) heterozygotes Genotype Wts Recombinants pXlOO(C.L) X-chromosome: unc-20 dpy-8l+ + + ;unc-20 dpy-8/szTl dpy-3 unc-20/+ + + ;dpy-3 unc-20/ szTl + ;unc-1 dpy-7IszTl 778 1887 882 470 487 38 Dpy-8 37 Unc-20 0 30 Dpy-3 29 Unc-20 14 Dpy-3 6 Unc-20 117 Unc-1 51 Dpy-7 7.2 (5.6-8.9) 0.0 (0-0.2.0) 5.0 (3.7-6.3) 3.0 (1.7-4.9) 26.3 (20.8-32.9) 19.6 (14.7-25.5) Chromosome I: let-362 dpy-5 unc-13; +IszTl 817 unc-13 let-88/+ + 1173 unc-13 let-88; + IszTl 2812 unc-13 let-80l+ + 1897 unc-13 let-80; + lszTl 1639 dpy-5 unc-75; + IszTl 581 dpy-5 unc-101; + IszTl 562 unc-101 unc-54; + IszTl 640 0 3 Unc-13 0 10 Unc-13 7 Unc-13 45 Dpy-5 91 Unc-75 45 Dpy-5 105 Unc-101 101 Unc-101 0.0 (0-0.2) 0.4 (0.1-1.0) 0.0 (0.0-1.4) 0.8 (0.4-1.4) 0.8 (0.4-1.7) 14.6 (10.5-19.0) 15.1 (10.9-19.6) 24.4 (19.8-29.3) 91 and Rose 1988); 3.0] to 14.6 m.u. (Table 12). The adjacent interval, unc-101 unc-54 (Table 12), increased less than two-fold in szTl heterozygotes from 14.2 m.u. (Table 11) to 24.4 m.u. As described in section 2.7, this calculation is based in part on the dpy-5 unc-101 distance in szTl(I;X) heterozygotes (Table 12). On the X-chromosome a 50% increase was observed in the unc-1 - dpy-7 interval from 19.7 m.u. to 26.3 m.u. (Table 12). The increase was actually much larger than 50%, however, because the crossover suppression boundary was near unc-20. unc-20 is approximately 5 m.u. from unc-1 (Edgley and Riddle 1987). Recombination immediately adjacent to the szTl breakpoint on the X chromosome was not increased compared to the controls. An apparent decrease was observed from 5.0 m.u. to 3.0 m.u. in the dpy-3 unc-20 interval in szTl(I;X) heterozygotes (Table 12). Unfortunately, the location of the breakpoint and crossover suppression boundary in the dpy-3 unc-20 interval is not known to the same resolution as on chromosome I. 3.6.2. hTl(I,V) Isolation of hTl(I;V): The translocation hTl(I;V) was isolated in a screen for mutations causing pseudolinkage between the unlinked markers unc-13(1) and dpy-11(V) (section 2.3). Psueodolinkage was not detected in hTl heterozygotes between dpy-5(I) and unc-3(X), unc-4(II), dpy-10(H), unc-22(IV) nor between unc-13(1) and dpy-18(111). The translocation homozygotes are lethal and arrest at a mid-larval stage. The experiments below describe how the structure of hTl(I;V) was deduced. The conclusions are that hTl(I;V) is composed of two chromosomes IRVLhTl and ILVRhTl (Figure 13b). IRVLhTl segregates from and recombines with chromosome I while I^VRhTl segregates from and recombines with chromosome V. Structure of IRVLhTl. The half-translocation IRVLhTl was isolated in a free duplication (hyperploid) strain from rare nondisjunction events in the progeny of unc-13; dpy-11/ hTl(I;V). The original duplication strain had an Unc-13 phenotype and was found to be of the genotype unc-13; dpy-11; IRV^hTl. When this strain was crossed to wild-type males, all the FI progeny segregated Dpy-11 and Unc-13 progeny. Thus, like the I^X^szTl strain (section 3.6.1), 92 the IR V^hTl strain carried a normal diploid complement in addition to the duplicated region of chromosomes I and V. As with szTl, the duplication strain was used to map the chromosome I breakpoint of the translocation (section 2.5). Because the duplication strain was not Dpy-11, IRVLhTl must carry dpy-ll(+). In addition, IRVLhTl carries unc-60(+). The chromosome I breakpoint of hTl(I;V) is between let-80 and unc-29. This places the hTl(I;V) breakpoint to the right of the szTl(I;X) breakpoint. IRV^hTl was shown to carry wild-type alleles of unc-29 and unc-75, but not of let-80 or dpy-5 (Figure 13b). Knowing that let-80 and unc-29 are in nDf24 and nDf25 places the hTl breakpoint inside the regions deleted by these two deficiencies. hTl(I;V)[unc-29;-h] I nDf24; + and hTl(I;V)[unc-29; + ] / nDf25; + strains were constructed and were viable, suggesting the hTl associated lethal mutation and the chromosome I breakpoint did not coincide. The chromosome V breakpoint may disrupt an essential gene. Recombination in hTl(I;V) heterozygotes. Similar to the observations for szTl, the boundary of crossover suppression correlated closely with the physical breakpoint. On chromosome I, recombination was not observed in let-362 dpy-5 unc-13(I)/hTl heterozygotes. In addition, crossover suppression was observed in the unc-13 - let-80 interval (Table 13). Recombination on chromosome V between unc-60 and the breakpoint would have been observed as Dpy-5 or Unc-60 progeny segregating from a dpy-5; unc-60/ hTl heterozygote (Table 13). Dpy-5 progeny were recovered, but they all resulted from nondisjunction events (see below). Thus hTl(I;V) suppresses recombination on the left end of chromosome V. Recombination suppression ends before unc-42 as Unc-42 recombinants were recovered from unc-13; dpy-11 +1 hTl(I;V)[+; + unc-42] hermaphrodites. I measured 0.3 m.u. between the boundary of crossover suppression and unc-29 (Table 13). The expected distance between let-80 and unc-29 is 0.4 m.u. (McKim, Howell and Rose 1988). If there was an increase in this interval as was observed with szTl (McKim, Howell and Rose 1988; Howell 1989), it would place the hTl breakpoint very close to, and to the left of unc-29. Since the breakpoint has not been mapped relative to other markers between let-80 and unc-29, I cannot be certain if the breakpoint is close to unc-29 and there is an enhancement in this 93 Table 13 Recombination in hTl(I;V) heterozygotes Genotype Wts Recombinants pXlOO(C.L) unc-13 let-80; + lhTl 3215 0 (0-0.2) dpy-5 unc-29; + lhTl 1631 1 Dpy-5 0.1 (0-0.6) 5 Unc-29 0.3 (0.1-0.7) dpy-5 unc-75; + lhTl 612 23 Dpy-5 7.3 (4.8-10.6) 41 Unc-75 6.7 (4.7-9.1) dpy-5 unc-101 ; + lhTl 500 34 Dpy-5 12.9 (9.2-17.6) 76 Unc-101 15.6 (12.0-19.7) unc-101 unc-541 hTl 282 49 Unc-101 27.0 (19.9-35.6) unc-13; dpy-lllhTl 401 dpy-5; unc-601 hTl 1420 14Dpy-5a a Thirteen of the Dpy-5 hermaphrodites were recovered from only two Po individuals. A single wild-type hermaphrodite segregated 11 Dpy-5s while a second produced two Dpy-5 worms. The Dpy-5 progeny were not recombinants but were the result of non-disjunction events (see text). 94 region, or if the breakpoint is close to let-80 and there is no enhancement. Recombination in the adjacent unc-29 - unc-75 interval (as inferred from the dpy-5 unc-75 experiment since the crossover suppression boundary is very close to unc-29) did not increase as it did for szTl(I;X) (7 m.u. vs. 14.6 m.u.; Tables 12 and 13). The hTl(I;V) chromosome I breakpoint, which falls in the same general region (unc-13 to unc-29) as the szTl breakpoint, did not enhance recombination in the adjacent DNA. In the unc-101 unc-54 region, recombination was increased almost two-fold from 14.2 m.u. (Table 11) to 27.0 m.u. (Table 13). As with szTl(I;X), this calculation was based partly on the dpy-5 unc-101 distance in hTl(I;V) heterozygotes (section 2.7). In contrast to the region adjacent to the breakpoint, this region behaves similarly in hTl and szTl heterozygotes. Segregation in hTl(I;V) heterozygotes. Segregation of IRVLhTl was assayed in the following cross, dpy-5 unc-13; + / hTl(I;V) males were crossed to dpy-5 unc-29; +1 hTl(I;V):[+ unc-29;+] hermaphrodites. If IRVLhTl segregated independently of chromosome I, the dpy-5 unc-29 and IRV^hTl unc-29 chromosomes could be in the same oocyte. If fertilized by a V-W; ILVRhTl sperm, an Unc-29 male would result. From this cross, 247 wild-type males were observed but no Unc-29 males. Thus, as with szTl(I;X), it is the portion of the translocation with the right half of chromosome I (IRV^hTl) that usually segregates from chromosome I. Alternate and adjacent-a segregations appear to occur with equal frequency in hTl(I;V) heterozygotes. The same arguments stated for szTl apply to hTl. From 302 eggs scored from dpy-5 unc-13; +1 hTl(I;V) parents, 76 reached adulthood, or 25.2%. This value was quite low (expected 31%), suggesting either there was a viability problem or an appreciable level of adjacent-b segregation. The adjacent-b possibility is not likely considering segregation experiments described above. The results of two experiments suggest alternate and adjacent-a segregation are equal in hTl heterozygotes. First, the wild-type:Dpy Unc ratio from a dpy-5 unc-29/ hTl(I;V) heterozygote (Table 13) was close to the expected 4:1 (1636/352 = 4.6:1). Second, in recombination experiments of intervals spanning the chromosome I breakpoint, the ratio of 95 reciprocal recombinants was close to the 2:1 ratio (58 Dpy-5 to 122 Unc-x or 1:2.1) expected if the alternate and adjacent-a segregation frequencies were equal. Nondisjunction in hTl heterozygotes was assayed by scoring the progeny of either dpy-5; unc-60/hTl or unc-13; dpy-lllhTl hermaphrodites (Table 13). Unc or Dpy progeny resulted either from crossovers in the recombination suppressed portion of hTl or from nondisjunction. In the experiment with dpy-5 and unc-60, 14 Dpy-5 hermaphrodites were recovered. Eleven of these, however, came from a single hermaphrodite. Two more came from a second hermaphrodite. Four of these Dpy-5 strains (including two from the group of eleven and one from the group of two) were crossed to N2 males, and 10 wild-type progeny were kept to observe their progeny. If the Dpy-5 progeny were recombinants, I expected half of the wild-types to segregate noDpy-5 progeny if the I^VRhTl chromosome had picked up dpy-5 (i.e. dpy-5; unc-60/hTl(I;V)[dpy-5; + ] and hTl is homozygous lethal) or half to segregate no unc-60 progeny if the normal chromosome V had lost the unc-60 marker (i.e. dpy-5; unc-60/+). In fact, all of the wild-types segregated both Dpy-5 and Unc-60 progeny. Furthermore, none of the Dpy-5 strains bred true homozygotes. These data are consistent with the Dpy-5 strains being of the genotype dpy-5; unc-60; IRV^hTl; they resulted from nondisjunction events of chromosome I and I^V^hTl. Thirteen of the exceptional Dpy-5 progeny appeared to arise from pre-meiotic events because they appeared in clusters. The occurrence of pre-meiotic non-disjunction events makes it difficult to accurately calculate the nondisjunction frequency. I estimate the frequency of I N - IRVLhTl nondisjunction to be (2X3/(1420/4)) 1.7%. 3.6.3. hT2(I;Ul) hT2(I;III) is a reciprocal translocation between chromosomes I and III. It was isolated by K. Peters (personal communication) as a mutation causing pseudolinkage between unc-13(1) and dpy-18(III). hT2 was induced on a bli-4(e937) marked chromosome and was homozygous viable. Based on experiments described below, I conclude that hT2 was made up of two novel chromosomes (Figure 13c). One chromosome, designated I^JIII^'hT2, carries the part of 96 chromosome I between the left end and unc-59 attached to a fragment from the left end of chromosome III. The other chromosome, designated IRIIIRhT2, is the unc-54 region of chromosome I attached to the right end of chromosome III. The unc-54(+) containing chromosome of hT2, IRIIIRhT2, recombined with chromosome I (Table 14), therefore it most likely segregated from it as described for I R bearing chromosomes of szTl(I;X) and hTl(I;V). The linkage of I R to IIIR was demonstrated in the following two experiments. The First showed that I R is linked to a segment of chromosome III that carries dpy-18. Recombination between dpy-18 and unc-54 was measured in homozygous KT2 hermaphrodites: hT2(I;III)[+ unc-54; dpy-18]/hT2[dpy-5 +; + ]. The progeny included 120 Dpy-18 recombinants, 663 wild-types and 244 Dpy-5, giving a recombination frequency of 0.265. Thus, on the hT2 chromosomes unc-54 and dpy-18 show linkage. The second experiment, in which the progeny of dpy-5; unc-361 hT2(I;III) were scored (Table 14), showed that the dpy-18 bearing segment is IIIR. Most of the Dpy-5 progeny were found to be carrying an unc-36(+) duplication. This duplication was further analyzed (section 2.5) and shown to carry the dpy-17(+), dpy-18(+) and the unc-64(+) genes of chromosome III (Figure 13c). The duplication chromosome was therefore IRIIIRhT2. Approximately 20% (2/10 tested) of the Dpy-5 hermaphrodites, however, did not carry a duplication. The non-duplication exceptions are discussed below. Because the duplication IRIIIRhT2 did not carry dpy-5(+), the chromosome I breakpoint must be between dpy-5 and unc-54. To show the breakpoint was between unc-29 and unc-54, the absence of linkage between unc-29 and dpy-18 in hT2 homozygotes was shown by scoring the progeny of hT2(I;III)[unc-29; + ]/ hT2(I;III)[+; dpy-18] hermaphrodites. The data for the segregation of unc-29 and dpy-18 (538 wild-types, 166 Dpy-18, 192 Unc-29 and 76 Dpy-18 Unc-29) are close to a 9:3:3:1 ratio, indicating the chromosome of hT2 containing unc-29 does not have dpy-18. This data shows that dpy-5 and unc-29 are on ILIIILhT2. In addition, K. Peters (personal communication) has shown that unc-29 on the translocation chromosome segregates independently of the normal chromosome I. As described below in HT2 heterozygotes, 97 Table 14 Recombination in hT2(I;III) heterozygotes. Genotpye0 Wt. Wt/Db Recombinants m.u.(C.L) Chromosome I: dpy-5; unc-36lhT2[unc-29;dpy-18] 1274 bli-3 unc-11; +/hT2[dpy-5] 1035 dpy-5 unc-29; + IKT2 1082 dpy-5 unc-75; +lhT2 716 dpy-5 unc-101; + /KT2 1575 dpy-5 unc-59; + lhT2[dpy-18] 479 dpy-5 unc-54; +/hT2[dpy-18] 444 4.3C 4.1d 3.7C 4.5C 3.5C 4.5d 22 Dpy-5 4 Unc-36 10 Unc-59 100 Dpy-5 see text 2.1 (1.0-3.8) 43.4 (33.7-57.1) Chromosome III: +; dpy-17 unc-36/hT2 854 +; unc-45 dpy-17IKT2 375 +; dpy-17 unc-64/hT2[dpy-5] 679 4.4 96 Unc-45 65 Dpy-17 28.3 (21.8-38.3) 32.1 (24.8-41.8) a All hT2 chromosomes carried bli-4(e937). b Ratio of Wild-type to homozygote Dpy or DpyUnc progeny. c Ratio calculated from normal chromosome homozygote. ^ Ratio calculated from translocation homozygote. 98 recombination was suppressed from the left end of chromosome I to unc-101 but not suppressed as far as unc-59. By analogy to the previous two translocations, the chromosome I breakpoint may be between unc-101 and unc-59. I have placed the unc-45(+) end of chromosome ILT linked to I*-1, although there is no direct supporting evidence for this. The Dpy-5 duplication bearing progeny from dpy-5; unc-361 hT2(l;III) hermaphrodites must have been the product of nondisjunction of chromosome I and IRIIIRhT2 during meiosis. Assuming no viability problems with the duplication worms, the frequency of nondisjunction was (2(.8X22))/(1274/4)) 11.6%. Nondisjunction of chromosome I in hTl heterozygous hermaphrodites (1.7%; section 3.6.2) was found to occur at a lower frequency than in hT2 heterozygotes. In addition, the hT2 nondisjunction events did not appear in clusters. More nondisjunction events were detected in the recombination experiments of Table 14 where hT2 was marked with both unc-29 and dpy-18. In these experiments I observed both Dpy-18 and Unc-29 Bli-4 progeny. The Dpy-18 progeny were expected from chromosome I nondisjunction events. These worms were predicted to be trisomic for chromosome I and of the genotype hT2[dpy-18; unc-29]lhT2[dpy-18; unc-29]!dpy-5. Similarly, the Unc-29 Bli-4 progeny were predicted to be trisomic for chromosome III and of the genotype hT2[dpy-18; bli-4 unc-29]/hT2[dpy-l8; bli-4 unc-29]luhc-36. Trisomy for chromosome I is a viable condition, as described in section 3.10.1. Recombination in hT2(I;III) heterozygotes. To calculate the recombination frequency in hT2 heterozygotes, a couple of assumptions have been made (section 2.7). Most important is the assumption that the two types of segregation at meiosis I, alternate and adjacent-a occur at equal frequency. That is, the four types of gamete IRIIIRhT2; ILIIILhT2, IRIIIRhT2; i n N , I N ; ILIIILhT2 and IN; U l N are produced at equal frequency. Two predictions can be made from this assumption. The first is the ratio of translocation heterozygote to homozygote (either for the translocation or normal chromosomes) progeny produced from a self-fertilizing heterozygous hermaphrodite should be 4:1. In the recombination experiments of Table 14, ratios close to 4:1 were observed. The second prediction is that when recombination is scored in an interval of a translocation chromosome, one recombinant (the translocation heterozygote) should be 99 approximately twice as frequent as the other recombinant (the homozygote) (Figure 5). Using formulas derived from these assumptions and independently calculating from each of the two recombinant classes in the unc-45 dpy-17 experiment, the same recombination frequency was determined. Therefore, the formulas for calculating recombination frequency in hT2 heterozygotes are probably accurate. Recombination was suppressed over a considerable part of chromosome I in hT2(I;III) heterozygotes. Between bli-3 and unc-101 recombination was almost eliminated (Table 14). Recombination suppression ended between unc-101 and unc-59. The observation of 43.4 m.u. between dpy-5 and unc-54 in hT2 heterozygotes indicated that recombination frequency was enhanced at least three-fold near unc-54. The distance between unc-54 and the hT2 breakpoint can not be more than 14 m.u. of the normal map (distance between unc-101 and unc-54; Table 11). Such an increase in recombination frequency in the region of a translocation heterozygote which still recombined was also observed with szTl and hTl. That this enhancement is a property of the translocation heterozygote and not the breakpoint is shown by the scoring of recombination in hT2 homozygotes. This was done for the dpy-18 unc-54 interval. The results (see above) show there can be no more than 26.5 m.u. between unc-54 and the hT2 breakpoint. Surprisingly, an enhancement of similar magnitude was not observed on chromosome III. The unc-45 dpy-17 interval, which includes most of the recombining part of chromosome HI in hT2 heterozygotes, was approximately 30 m.u. While this was an enhancement when compared to the control value of 23.7 m.u., it failed to compensate fully for the recombination suppression to the right of dpy-17 (Table 14). Alternatively, high level of double crossovers could mask the magnitude of the enhancement. Rare recombination events have been observed in the crossover suppressed region of hT2 heterozygotes. In scoring the dpy-5; unc-36/ hT2(I;III) hermaphrodites, four Unc-36 progeny were recovered (Table 14). Two of these were tested and found be of the genotype dpy-5 +1 + bli-4; unc-36lunc-36. Recombination involving ILIIILhT2 had occurred between bli-4 and the hT2 breakpoint. The unc-29(hl011) marker on ILIIILhT2 in this experiment and if the crossover was 100 between bli-4 and unc-29 the resulting chromosome would have been dpy-5(+) bli-4(-) unc-29(+). The two Unc-36 exceptions were tested by crossing them to unc-29l+ males. In one case unc-29(hlOU) was present, indicating the crossover occurred between unc-29 and the hT2 breakpoint. In the other strain unc-29(+) was present, indicating the crossover was between bli-4 and unc-29. The exchange of flanking markers shows these recombination events were true crossovers and not simply conversion events. They were also more frequent than expected for spontaneous mutation. The reciprocal product of these recombination events would generate a Dpy-5 worm. Of ten Dpy-5 strains tested, eight were duplication strains (see above) and two were not. These two were probably recombinant hT2 heterozygotes with a dpy-5 marker on I^'III^'hT2. Although the two Dpy-5 strains were not extensively tested, they did not carry an unc-36(+) duplication and true breeding Dpy-5 strains could not be obtained. These results could be explained if the genotype of the Dpy-5 worms was dpy-5; unc-36/hT2(I;III)[dpy-5 unc-29; dpy-18] . These would have a dpy-5 mutation on ILIIILhT2 and also unc-29 if the recombination event occurred between dpy-5 and unc-29. If there is recombination in the crossover suppressed region of hT2 heterozygotes, then it should be detectable by scoring the progeny of dpy-5 unc-101; + lhT2 hermaphrodites. Surprisingly, I recovered no recombinant progeny in 1575 wild-type progeny (Table 14). Based on the data from the dpy-5; unc-36/ hT2 experiment, I expected 4 Dpy-5 and 4 Unc-101 recombinants in this experiment. Additional experiments with dpy-5 unc-29 and dpy-5 unc-75 also failed to detect recombinants in the normally crossover suppressed region. I have no explanation for the discrepancy in these experiments. Crossovers in the recombination suppressed region of hT2 have been observed in other experiments. For example, an Unc-13 worm isolated from unc-13; dpy-18lhT2(l;III) heterozygotes by K. Peters (personal communication) was likely unc-131 unc-13; dpy-18/+. In this case, the recombination event was inside the crossover suppressed region on chromosome LTI between dpy-18 and the breakpoint. K. Peters also isolated apparently new bli-4 alleles in a precomplementation screen using the bli-4(e937) allele on hT2. Molecular analysis of one new 101 allele, however, showed that it was probably e937 (K. Peters, personal communication). This could have resulted from the normal chromosome I picking up the e93 7 mutation by a recombination event. Rare recombination events involving normally suppressed regions of translocation heterozygotes have also been detected by R. Rosenbluth (personal communication) with other translocations. Not all the exceptional progeny from hT2(I;III) heterozygotes can be explained so simply. A typical example is illustrated by a wild-type strain isolated from dpy-5 unc-101 ;+lhT2(l;III)[bli-4; dpy-18] heterozygotes. This strain was identified because the wild-types did not segregate any Dpy-5 Unc-101 progeny. In addition, 7% of the wild-type progeny from these worms had blisters which, unlike normal Bli-4 worms, covered less than 50% of the body and were often only small bubbles. These worms also segregated a larger fraction of Dpy-18 progeny (46/118 = 39%) than the expected 25%. To determine the genotype of this strain, it was crossed to unc-29; + lhT2(I;III)[dpy-5; + ] males and FI wild-type progeny were picked and classified based on the types of progeny produced. This experiment showed that the exceptional strain was still heterozygous for hT2(I;III)[bli-4; dpy-18]. In addition, the abnormal chromosome I was homozygous lethal and the dpy-5 mutation was not being expressed. This lead to the proposal that the dpy-5 unc-101 chromosome had been modified by deletion of bli-4(+) and other essential genes and by addition of more chromosome I material. The added material had dpy-5(+) and therefore originated from I^'III^'hT2. The low level of blister expression was probably due to the dominant suppressing effects of dpy-5 mutations (K. Peters, personal communication). Perhaps a recombination event, possibly involving non-homologous sequences, was not properly resolved and produced the abnormal chromosome I. 3.6.4. hT3(I;X) hT3(I;X) was isolated as a gamma radiation induced lethal mutation on chromosome I. The isolation and mapping of hT3 is described in section 3.2. The lethal mutation, designated h916, was duplication mapped close to the left of dpy-5 between the breakpoints of hDp24 and 102 hDp76 (Figure 24). h916 failed to complement h60, a gamma-ray induced allele of let-363, which was previously mapped to the same interval. The chromosome I breakpoint (see below) and lethal site may be at the same site. Subsequent analysis (see below) demonstrated h916 to be a reciprocal interchange between chromosomes I and X. The translocation breakpoint did not disrupt an X-linked essential gene since hT3/+ males were viable and fertile and the lethal phenotype of hT3lhT3 was rescued by chromosome I duplications. Structure of hT3(I;X). hT3 is made up of two parts (Figure 13); ILXLhT3, which recombines with the X-chromosome, and IRXRhT3 which recombines with chromosome I. The structure of hT3 was deduced by isolating one of the translocation chromosomes as a duplication and then mapping the breakpoints. When the progeny of unc-11 dpy-14; + / hT3(I;X)[dpy-5 unc-29;+] hermaphrodites were scored, two types of Dpy-14 were recovered. The First type, the result of recombination, had the genotype unc-11 dpy-14; + / hT3(I;X)[dpy-5 dpy-14; -f ]. This was concluded because these Dpy-14 recombinants segregated larval Dpy-5 Dpy-14 progeny that were probably translocation homozygotes. The second type did not segregate Dpy-5 Dpy-14 and did not produce Dpy-5 progeny in crosses to dpy-5/+ males. While these Dpy-14 worms could have resulted from a recombination event between the translocation breakpoint and dpy-5, further analysis showed this was not the case and that the genotype for these Dpy-14 worms was unc-11 dpy-14/ unc-11 dpy-14; ILXLhT3. That is, these worms carried an unc-11(+) duplication. This was shown by crossing these Dpy-14 worms to unc-11 dpy-5 +/ + + dpy-14 males and observing the wild-type progeny. The wild type progeny segregated more wild-types, Unc-11, Unc-11 Dpy-5, Unc-11 Dpy-14, Dpy-14 and Dpy-5 progeny. The wild-types, therefore, carried three copies of the unc-11 region but only two copies of the dpy-14 region. It could not be determined had the dpy-5 region because the translocation was induced on a dpy-5(e61) unc-29(e403) chromosome. The duplication breakpoints were mapped as described in section 2.5. In addition to unc-1 l(+),ILXLhT3 covers bli-3(+) at the left end of chromosome I. The breakpoint and the lethal site [kt-363(h916)] map to similar locations and may be the same site. On the X-ar chromosome, ILXLhT3 includes unc-1 (+), dpy-7(+) but not unc- 3(+). Therefore, the ILXLhT3 103 chromosome includes the bli-3 unc-63 region of chromosome I attached to a large portion of the X chromosome extending from unc-1 to dpy-7. The other translocation chromosome presumably carries the reciprocal components of chromosomes I and X (IRXRhT3). Segregation was assayed in hT3 heterozygous males. I crossed bli-4 dpy-14;0/hT3[+; + ] males to dpy-14 unc-29; + lhT3[unc-29; + ] hermaphrodites and 46 wild-type progeny were picked onto individual plates. Thirty-two wild-types segregated Bli-4 Dpy-14 progeny and fourteen segregated Dpy-14 Unc-29 progeny. If segregation of the two translocation chromosomes was random in hT3 males, then I should have recovered these two classes in approximately equal numbers. The excess of wild-types that segregated Bli-4 Dpy-14 progeny indicated the two translocation chromosomes had a tendency to segregate from each other during male meiosis. This conclusion is the same made for szTl heterozygous males and may be a form of non-homologous disjunction (section 3.9). Recombination in hT3 heterozygotes. The assumptions on which the calculations of recombination frequency in translocation heterozygotes are based has been described (section 2.7). These assumptions make some predictions about the segregation products from translocation heterozygotes. The data with hT3 heterozygotes, however, did not agree with these predictions (Table 15). The ratio of translocation heterozygotes to normal chromosome homozygotes was less than 4:1. The data from scoring the progeny of unc-11 ; + lhT3 hermaphrodites were 763 wild-types to 221 Unc-11. The wild-type/Unc-11 ratio was 3.45. In addition, the expected 2:1 ratio of recombinant classes was not observed for X-chromosome intervals. For example, when the progeny of unc-20 dpy-8lhT3 hermaphrodites were scored, I expected more Unc-20 than Dpy-8 recombinants. As shown in Table 15, this was not observed. In the chromosome I experiments, the data were confounded by the fact that in all the experiments one of the recombinant phenotypes was not scored (Unc-54) or had reduced viability (Dpy-14). These deviations from the expectations may have been due to reduced viability of the translocation heterozygotes or to increased frequency of alternate segregation relative to adjacent-a segregation. These two possibilities were resolved by doing an egg survival experiment. A lower frequency of egg survival 104 Table 15 Recombination in hT3(I;X) heterozygotes.0 Genotypek Wt. Recombinants m.u.(C.I.)c Chromosome I: bli-3 unc-11; +1 hT3[unc-29; + ] 431(7) bli-3 unc-63; +1 hT3 707(19) unc-40 dpy-14; +lhT3 668 + ; +/hT3[unc-29; + ] 294 unc-11 dpy-14; + lhT3[unc-29; + ] 449 dpy-14 unc-101; +/hT3[unc-29; + ] 408 unc-101 unc-54; +/hT3 336 31 Unc-29 3 Dpy-5 6 Unc-40d 3 Dpy-5 13 Unc-29 5 Unc-11 14 Dpy-14 24 Unc-29 103 Unc-101c 29 Unc-29 42 Unc-101d 1 Dpy-5 1.8 (0.9-3.8) see text 25.6 (20.5-33.0) 17.5 (12.6-23.4) X-chromosome: + ; dpy-7 unc-3lhT3 + ; unc-20 dpy-8lhT3 + unc-1 dpy-7lhT3[unc-29; + ] 536 338 271 5 Dpy-76 7 Unc-3rf 25 Unc-20c 34 Dpy-8rf 1 Dpy-5 63 Dpy-7d 75 Unc-l c 6 Unc-29 0.9 (0.4-2.1) 3.0 (0.7-5.1) 7.4 (5.0-10.8) 18.8 (13.4-25.5) 43.9 (33.0->50) 30.8 (23.3-50.0) ° See the text for the map distance between the hT3 breakpoint and dpy-5 or unc-29. b All hT3 chromosomes carried the dpy-5(e61) mutation. c Recombinant a translocation heterozygote (see text). ^ Recombinant a normal chromosome homozygote (see text). 105 than the predicted 5/16 (Figure 6) was expected if there was reduced viability of the translocation heterozygotes. If there was an increased frequency of alternate segregation relative to adjacent-a segregation, an increase in egg survival was expected. Out of 660 eggs from unc-11; + lhT3 heterozygotes, 129 wild-type and 34 Unc-11 progeny were produced. The egg survival frequency (25%) was less than the expected 31% (5/16). I conclude the data are best explained by reduced viability of hT3 heterozygotes. In fact, in all the X-chromosome recombination experiments, the recombinants which were heterozygous for the translocation were often slow growing and sick. This viability problem resulted in an over-representation of the recombinant which was not a translocation heterozygote. The calculation of recombination frequency derived from the recombinant which was heterozygous for the translocation should be more accurate because the wild-types were also translocation heterozygotes. Unfortunately, the mutant phenotype of the recombinant may have exacerbated the viability problem, resulting in a reduced recovery of recombinant progeny and an artificially low recombination frequency. For this reason, the true value is probably somewhere between the values calculated from the two types of recombinant. If the data were available, the results from both types of recombinant are given Table 15. Recombination on chromosome I in hT3 heterozygotes was suppressed to the left of unc-63 (Table 15). Recombination was observed in intervals to the right of dpy-5. Recombination also occurred between the hT3 breakpoint and dpy-5. hT3 was induced on a dpy-5 marked chromosome and Dpy-5 progeny resulted from recombination between the breakpoint and dpy-5. In the recombination experiments of Table 15, 8 Dpy-5 progeny were recovered in 2049 wild-types, giving a distance of 0.4 m.u. Howell et al. (1987) found 0.3 m.u. between let-363 and dpy-5. Thus like szTl and hTl, the crossover suppression boundary and the breakpoint are very close. Some of the recombination experiments were also done with an unc-29 marked chromosome (Table 15); 103 Unc-29 hermaphrodites were scored in 1853 wild-types for a calculated distance of 5.6 m.u. between the breakpoint and Unc-29. McKim, Howell and Rose (1988) found 3.1 m.u. between dpy-5 unc-29 in normal chromosomes. Thus there is a small enhancement of recombination in this region. This enhancement was also observed in other intervals further to 106 the right. The distance in the dpy-14 unc-101 interval was 25.6 m.u. In the adjacent unc-101 unc-54 interval the distance was 17.5 m.u.. The total amount of recombination on chromosome I in hT3 heterozygotes was approximately 45 m.u. This amount of recombination was high compared to the 25 m.u. normally observed in this region of chromosome I (Table 11). The loss of recombination in the left end of chromosome I was compensated for in the portion of the chromosome that was still recombining. On the X-chromosome, recombination was reduced in the dpy-7 unc-3 interval from 19.1 m.u. to 0.9 m.u. Recombination was enhanced in the unc-1 dpy-7 interval from 19.7 m.u. to at least 30.8 m.u. The low level of nondisjunction observed in hT3 heterozygotes is probably due to the increased recombination frequency in the unc-1 dpy-7 interval. This is in contrast to szTl(I;X) where X-chromosome nondisjunction was high and the total level of X-chromosome recombination was low. The breakpoint of szTl is between dpy-3 and unc-20; much closer to the left end of the X-chromosome than the hT3 breakpoint. 3.7 Insertions The effects of some insertions and duplications on recombination are now described. hDpl4(I;X): The duplication hDpl4 (Figure 7 and 9), is a derivative of sDp2 inserted into the X chromosome. The insertion event did not disrupt an essential gene since hDpl4 homozygotes and males were viable. Hermaphrodites of the genotype dpy-5ldpy-5; hDpl4/hDpl4 were very slow growing; they had a generation time at 2 0 ° of 6 days compared to 4 for hDp!4 heterozygotes. These homozygous duplication strains also had a Him phenotype, and were never observed to segregate Dpy-5 progeny, dpy-5/dpy-5; hDpl 410 males were fertile and contributed the duplication to all their hermaphrodite progeny. The insertion event was associated with loss of the sequences at the right end of sDp2 {bli-4 - dpy-14) but no detectable loss occurred at the left end of sDp2. let-362, egl-30 and unc-35 were complemented by hDpl4 (Figure 7). In hDpl4 heterozygous hermaphrodites, males were produced at a frequency of 10% (Table 30). 107 Table 16 Mapping and recombination in hDpl4(I;X) strains. Genotypea Wtb Recombinants 0 Fraction Him m.u.(C.I.)c dpy-3 unc-20/hDpl4 389 19 Dpy-3 18/19 6.3 (4.6-8.9) 12 Unc-20rf 0/12 7Dpy-5 e 0/7 10.6 (4.4-22.4/ unc-20 dpy-8lhDpl4 273 32 Dpy-8d 3/27 14.6 (9.4-21.3) 16 Unc-20 13/16 dpy-7 unc-31hDpl4 248 3 Unc-3 3/3 1.8 (0.9-3.9) 2 Dpy-7d 0/2 dpy-5;dpy-7 unc-31+ +;hDp16296 36 Unc-3 not calculated hDpl4/ unc-7 lin-15 932 1 Unc-7 0/1 0.1 (0-0.5) 9 Dpy-5e 3.9 (1.7-7.3/ dpy-5; unc-3/ hDpl4 1189 25 Unc-3 17/17 2.7 (2.0-3.1/ 18 Dpy-5 0/18 dpy-5; unc-1/ hDp 14 207 49 Dpy-5 1/19 34.7(24.0-48.8/ 38 Unc-l£ 42/56 a All strains were heterozygous for dpy-5 on chromosome I except where indicated that dpy-5 was homozygous. b Male progeny not included. c C.I. = 95% confidence interval. ^ 1/4 of this class was not observed because of dpy-5 segregating. e The Dpy-5 recombinants were multiplied by 8 in order to calculate the insertion to unc-20 or unc-7 distance. f Distance between marker and hDpl4. In case of unc-20 dpy-8, the marker was unc-20. § see text (sections 4.8 and 108 The insertion site was positioned by three factor mapping the Him phenotype. The Him mutation must have been tightly linked to the duplication since the duplication and the Him phenotype always segregated together. The Him mutation was mapped to the right of dpy-3 and unc-20 but to the left of dpy-8 (Table 16). The single non-Him Dpy-3 recombinant could have been the result of a double crossover event, one between dpy-3 and unc-20, and the second between unc-20 and the insertion. The unc-20 dpy-8 experiment showed the insertion was between these two markers. The fraction of Unc-20 recombinants with the duplication to Dpy-8 recombinants with the duplication did not constitute evidence that the insertion mapped closer to dpy-8 because, as described next, the insertion disrupted the normal distribution of recombination on the X-chromosome. To the left of the insertion, recombination was slightly higher than normal. The dpy-3 unc-20 interval was 6.3 m.u. in duplication heterozygotes (Table 16) compared to 5.0 m.u. on the normal map (Table 12). In the dpy-3 unc-20 experiment, the Dpy-5 recombinants were the result of recombination between unc-20 and the insertion. Because dpy-5(e61) was heterozygous on chromosome I, the number of Dpy-5 recombinants was multiplied by eight in the calculation of the insertion - unc-20 distance. The 10.6 m.u. observed in the unc-20 hDpl4 interval was only slightly higher than the 7.2 m.u. normally observed in the unc-20 dpy-8 region (Table 11). Similarly, 14.6 m.u. was observed in the unc-20 dpy-8 interval of hDpl4 heterozygotes. With an assumption that the Unc-20 duplication bearing worms had a viability problem, this calculation was based on the number of Dpy-8 progeny. In contrast, to the right of the insertion recombination was greatly reduced. In hDpl4 heterozygotes, recombination in the dpy-7 unc-3 interval was reduced ten fold from 19.1 m.u. to 1.8 m.u. As a control, recombination in dpy-5; hDpl6(I;f); dpy-7 unc-3/ + + hermaphrodites was scored (Table 16). hDpl6 is of comparable size to hDpl4 but is unlinked. This result indicated the hyperploidy in hDpl 4 strains did not cause the reduction in X-chromosome recombination frequency. The hDpl 4 induced crossover suppression was observed over the entire portion of the X-chromosome to the right of the insertion. Even in the unc-7 lin-15 interval, which is located at 109 the far right end of the chromosome (Figures 1 and 20), recombination was reduced in hDpl4 heterozygotes from 1.6 m.u. (Table 11) to 0.1 m.u. (Table 16). A direct measure of the genetic distance between the insertion site and a marker was found by scoring the progeny of hermaphrodites homozygous for dpy-5 but heterozygous for hDpl4 and an X-linked marker. For example, in dpy-5ldpy-5; hDpl4/unc-3 hermaphrodites, the frequency of Dpy-5 and Unc-3 recombinants was a function of the map distance between the insertion site and unc-3. The data for the distance from the insertion to unc-1 or unc-3 are in Table 16. Many of the Unc-1 progeny were not true recombinants but instead carried broken hDpl4 chromosomes. They are discussed in section Despite the large number of these unusual events, the number of Dpy-5 and Unc-1 progeny from these crosses was close to equality. Because the reciprocal products of recombination were recovered in equal numbers, it is possible the breakage of the hDp 14 chromosome was initiated by a recombination event. The calculated recombination frequency between unc-1 and the insertion site, 34.7%, was based only on the Dpy-5 recombinants. The level of recombination in the unc-1 hDpl 4 interval was an enhancement compared to normal chromosomes. With approximately 20 m.u. between dpy-3 and the insertion (the dpy-3 unc-20 plus the unc-20 dpy-8 distance in hDpl4 heterozygotes), there was 16 m.u. between unc-1 and dpy-3 in hDpl4 heterozygotes; a level substantially greater than normally observed (5 m.u.; Edgley and Riddle 1987). Combining the data, the total size of the X-chromosome in hDp 14 heterozygotes was not more than 38 m.u. hDpl02(I;X): hDpl 02 (section 3.2 and Figure 7) is a second insertion of chromosome I material into the X-chromosome. In contrast to hDpl 4, this insertion did not cause high levels of X-chromosome nondisjunction (Table 17). The insertion site was mapped by three-factor analysis. unc-13; hDpl02 males were crossed to unc-13; dpy-7 unc-3 hermaphrodites to construct the unc-13; hDpl02(1;X)[+; + +]/ dpy-7 unc-3 hermaphrodites whose progeny were scored (Table 17). Dpy-7 or Unc-3 recombinants were recovered if the recombinant chromosome carried the insertion. As both types of recombinant were recovered, these data are consistent with the insertion being located between dpy-7 and unc-3. Recombination was reduced in this interval to 110 5.2 m.u from 19 m.u. Similar recombination suppression was not observed in the unc-7 Unl5 region of hDpl02 heterozygotes. When the progeny of hDpl021 unc-7 lin-15 hermaphrodites were scored (Table 17), the distance measured for the unc-7 lin-15 interval, 2.1 m.u., was not significantly different from the control (Table 11). The progeny of unc-13; hDpl 02(1 ;X)[+; + + ]/ unc-1 dpy-7 hermaphrodites were also scored (Table 17). The data was consistent with the insertion being inserted to the right of dpy-7. The single Dpy-7 recombinant could have been the product of two recombination events, one in the unc-1 dpy-7 interval, and the second in the dpy-7 -hDpl02 interval. Overall, the genetic length of the X-chromosome in the hDpl 02 heterozygote was 24 m.u. sDp30(V;X): A third X-chromosome insertion available was sDp30 (Figure 1). Rosenbluth et al. (1988) showed sDp30 is fragment of chromosome V attached to the X-chromosome. The insertion homozygotes were healthy and fertile. I was interested in the location of the insertion site and the effects on X-chromosome recombination. First, the X-linkage of the duplication was confirmed, dpy-1 IIdpy-11; sDp30 males were crossed to unc-60 dpy-11 hermaphrodites. The FI progeny consisted of only Dpy-11 males and wild-type hermaphrodites. These results confirmed the X-linkage of sDp30 because no males received the duplication while all the hermaphrodites did. sDp30 is inserted between dpy-7 and unc-3. dpy-11; sDp30/O males were crossed to dpy-11; dpy-7 unc-3 hermaphrodites. The progeny of the wild-type F l progeny were scored (Table 17). Because the Dpy-11 Dpy-7 progeny looked Dpy-11, recombinant Dpy-7 or Unc-3 progeny were recovered only if the recombinant chromosome picked up the insertion. As both Dpy-7 and Unc-3 progeny were found, sDp30 was the inserted between these two markers. Almost 3/4 of the recombinants were Unc-3, therefore the insertion site was probably closer to dpy-7. The insertion also caused a small but significant level of X-chromosome nondisjunction (Table 17). Unlike the previous hDpl4 and hDpl 02, however, sDp30 had little effect on X-chromosome recombination. The observed recombination fraction of 0.16 in the dpy-7 unc-3 interval (Table 17) was only slightly lower than the control (Table 11). The unc-1 dpy-7 distance I l l Table 17 Recombination in insertion strains. Genotype Wt. a Recombinants m.u.(CJ.) unc-13; hDpl02l dpy-7 unc-3 852 5 Dpy-7 0.9(0.3-1.9)^ 26 Unc-3 5.2(3.5-7.4)c 4 Unc-13 4.5(2.9-6.7)rf unc-13; hDpl02/ unc-1 dpy-7 802 (2) 83 Unc-1 15.0(12.1-19.1)e 1 Dpy-7 5 Unc-13 unc-13l+; hDpl02l unc-7 lin-15 2114 (3) 31 Unc-7 2.1(1.5-3.0/ 26 Unc-13 7.5(4.8-11.1)S dpy-11; dpy-7 unc-31 sDp30 1710 54 Dpy-7 16.2(13.8-18.9) 135 Unc-3 unc-1 dpy-7/ sDp30 879 128 Unc-1 21.3(17.3-25.6) unc-60 dpy-11/ mnDpl; unc-3 1340 8 Unc-60 0.7(0.4-1.3) 10 Dpy-11 dpy-11 unc-421 mnDpl; unc-3 1419 7 Unc-42 0.5(0.2-1.0) 9 Unc-3 0.6(0.3-1.2)A dpy-5 unc-54 +1 dpy-5 + hDp78 1182(7) 1 Dpy-5 0.08 (0-0.5) 1 Unc-54 dpy-5 unc-101 unc-54 +/dpy-5 + + hDp78 774 50 Unc-101 6.2(4.7-8.2) 2 Dpy-5 0.2(0-0.8) Males indicated in parentheses. 0 Insertion - dpy-7 distance. R = 2XDpy-7/Total progeny = 10/1176. c dpy-7 - unc-3 distance. R = 2(Unc-3 + Uncl3)/Total progeny = 60/1176. d Insertion - unc-3 distance. R = 2(Unc-3)/Total progeny = 52/1176. e unc-1 - dpy-7 distance. R = 2(Unc-l)/Total progeny = 166/1180. f unc-7 lin-15 distance. R = 2(Unc-7)/Total progeny = 62/2895. 8 Insertion - unc-7 distance. R = 8(Unc-13)/ Total progeny = 208/2895. h Insertion - unc-42 distance. R = 2(Unc-3)/ Total progeny = 18/2870. 112 in sDp30 heterozygotes was 21.3 m.u., also similar to the controls. These results raise the question of why the other two insertions, hDpl4 and hDpl 02, have recombination reduction phenotypes but sDp30 does not? Autosomal insertions: A similar type of analysis of autosomal insertions would be useful. Do they have effects on recombination frequency? Herman, Albertson and Brenner (1976) reported that the X-chromosome duplication mnDpl was linked to chromosome V. They also reported that recombination on the left end of chromosome V was reduced in mnDpl heterozygotes. To determine the genetic location of the insertion, I scored the progeny of unc-60 dpy-lll mnDpl(I;V)[+ +; + ]; unc-3lunc-3 hermaphrodites (Table 17). Dpy-11 progeny would be recovered only if the + dpy-11 recombinant chromosome also picked up mnDpl. Dpy-11 progeny were recovered, positioning the insertion site of mnDpl to the left of dpy-11. It was not possible to tell if the unc-60 + recombinant chromosomes also had mnDpl because Unc-60 is epistatic to Unc-3. The insertion site is probably close to unc-60 or even to the left because the total unc-60 dpy-11 distance, as calculated by the number of Unc-60 progeny, was the same as the mnDpl dpy-11 distance as calculated with the number of Dpy-11 progeny. To confirm that mnDpl was to the left of dpy-11, the progeny of dpy-11 unc-42 I mnDpl(I;V)[+ + ; + ]; unc-3/unc-3 hermaphrodites were scored. The only recombinants recovered were Unc-42 and Unc-3. The Unc-42 progeny were the result of recombination between dpy-11 and unc-42, producing a mnDpl + unc-42 chromosome. The Unc-3 progeny resulted from recombination outside the dpy-11 unc-42 interval producing a dpy-11(+) unc-42(+) chromosome V. In both these experiments, the calculated recombination frequencies showed that mnDpl reduced recombination in both regions of chromosome V tested, a result previously reported by Herman, Albertson and Brenner (1976). The normal unc-60 dpy-11 distance is approximately 18.0 m.u. (Table 11) and the dpy-11 unc-42 distance is approximately 3 m.u. (Edgley and Riddle 1987). Another autosomally inserted duplication is hDp78. This duplication was isolated as a spontaneous derivative of hDp4, a free duplication of chromosome I (section 3.13.2). The insertion 113 site was mapped 0.08 m.u. from unc-54 (Table 17) at the right end of chromosome I. Recombination was reduced in hDp78 heterozygotes between unc-101 and unc-54 (Table 17). The two Dpy-5 recombinants (putative genotype dpy-5 + +1 dpy-5 unc-101 unc-54) from the dpy-5 + + hDp78l dpy-5 unc-101 unc-54 + experiment positioned the insertion to the right of unc-54. Finally, I have shown that a duplication induced with ultraviolet radiation by H. Stewart, R. E. Rosenbluth and D. L. Baillie (personal communication) is linked to dpy-5. This duplication, sDp8(III;I), covers a small region around unc-36 on chromosome III (Figure 1). Among the progeny of dpy-5lsDp8; unc-36 hermaphrodites were 429 wild-types, 127 Dpy-5 Unc-36, 4 Dpy-5 and 5 Unc-36 hermaphrodites. Further analysis is required to determine the location of the insertion and it effects on recombination frequency. 3.8 M e i o t i c b e h a v i o u r o f d u p l i c a t i o n s . sDp2, hDp79, a n d I^X^szTl h a v e no effect o n r e c o m b i n a t i o n f r equency : Duplications of the left end of chromosome I do not recombine with the normal homologues (Rose, Baillie and Curran 1984 and section 3.13.2). They could still, however, pair with the normal homologues, thereby reducing recombination in the duplicated region. To test this possibility, three duplications which did not normally recombine with the normal chromosome I were examined to see if they reduced the frequency of recombination between the normal homologues. Strains carrying a duplication in addition to normal chromosomes heterozygous for markers were scored and the recombination frequencies were calculated as described in section 2.7 (Table 18). sDp2 had no effect on recombination in either the dpy-5 unc-13 interval or the let-362 dpy-5 interval. In the dpy-5 unc-13 interval, ILXLszTl had no detectable effect on recombination frequency. In the let-357 dpy-5 interval hDp79 had no effect on recombination frequency. The first two duplications are free, whereas hDp79 is linked to an autosome (section 3.13.2). It appears that duplications of the left end of chromosome I have no influence on the meiotic pairing of chromosome I homologues. 114 Table 18 Effects of duplications on recombination Genotype Wt Recombinants m.u.(C.I.)° dpy-5 unc-13 unc-291 + + +; sDp2b 1279 4 Dpy-5 1.3 (0.3-3.1) 10 Unc-29 1.4 (0.7-2.4) dpy-5 unc-13 unc-29/ + + +c 2678 34 Dpy-5 1.9(1.3-2.6) 21 Unc-29 1.2(0,8-2.6) let-362 dpy-5 unc-13l+ + +; sDp2 652 20 Dpy-5 Unc-13 15.0 (9.3-24.3) let-362 dpy-5 unc-13l+ ++d 1406 149 Dpy-5 Unc-13 15.4(13.0-17.9) dpy-5 unc-131'+ +; ILXLszTl 1769 6 Dpy-5 1.5 (0.7-3.3) dpy-5 unc-131 ++d 4775 52 Dpy-5 1.6 (1.3-1.9) 51 Unc-13 let-357 dpy-5 unc-13/+ +. +; hDp79 685 13 Dpy-5 Unc-13 6.3 (4.4-8.6)e 24 Unc-13 let-357 dpy-5 unc-13/+ + +d 1261 25 Unc-13 3.0 (2.0-4.4) a C.I. = 95% confidence interval. 0 This data was collected by A. M. Howell. c Data from McKim, Howell and Rose (1988). d Data from Howell et al (1987). e let-357 - dpy-5 distance. R = 2(Dpy-5 Unc-13 + Unc-13)/ Total progeny = 74/1211. 115 3.8.1. Duplications with meiotic pairing behaviour: Chromosome I. Previous work (Rose, Baillie and Curran 1984) showed the duplication sDpl(I;f) had the ability to recombine with the normal homologues. With rare exceptions, {e.g. mnDpl; Rogalski and Riddle 1988) other comparably sized duplications in C. elegans are not recombination proficient. I have extended this analysis by studying the pattern of recombination in sDpl and comparing it to strains carrying the IRV^hTl chromosome (section 3.6.2) as a duplication. Recombination was assayed between sDpl and the normal homologues by scoring recombination in dpy-5 unc-x/ dpy-5 unc-x/ sDpl strains. The data are shown in Table 19. Two recombination formulas (section 2.7) were derived. The first assumed the conclusion from experiments described below, that is the two normal homologues segregated at meiosis I despite the pairing activity of sDpl. The second assumed the segregation of one normal homologue was random when sDpl recombined with the other normal homologue. As it turns out, both formulas give similar answers (Table 19). Recombination over most of the length of sDpl could be measured since dpy-5 and unc-54 are at opposite ends of the duplication. The fraction of meiocytes with a duplication mediated recombination event (2 times the recombination frequency) was 12-14%. This was quite a bit less than the 66% expected if sDpl competed in pairing and recombination as effectively as the normal homologues. Furthermore, sDpl had little effect on recombination between the normal homologues in an interval not covered by the duplication. Recombination was scored in let-362 dpy-5 unc-13/ + + + / sDpl hermaphrodites. The results (Table 19) showed that recombination was either unaffected or slightly increased by the presence of the duplication. Analysis of the regional pattern of sDpl recombination showed there was relatively more recombination in the unc-101 unc-54 interval than in the dpy-5 unc-101 interval (Table 19). Recombination measured in the dpy-5 unc-101 interval of sDpl was 10 fold lower than between normal chromosomes (12 m.u. to 1 m.u.) and recombination between dpy-5 and unc-29 was 30 times lower (3 m.u. to 0.09 m.u.). The effect was even more severe further towards the left end of sDpl. Rose, Baillie and Curran (1984) have shown that recombination in the dpy-5 unc-13 116 Table 19 Recombination involving chromosome I duplications. Genotype Wt. Recombinants m.u.(C.I.) dpy-5 unc-54ldpy-5 unc-54/sDpl 114 dpy-5 unc-75/dpy-5 unc-75/sDpl 823(27) dpy-5 unc-101/dpy-5 unc-101/sDpl 580(16) dpy-5 unc-291 dpy-5 unc-29/sDpl 836(19) let-362 dpy-5 + unc-13l+ + +/sDpl 731(8) unc-29 unc-75/ + + 877 unc-29 unc-75/unc-29 unc- 75/IRVLhTl 660 unc-29 unc-54/+ + 551 unc-29 unc-54lunc-29 unc- 54IIRVLhTl 359 unc-29 unc-54lunc-29 unc-54/hDpl01 728 63 Dpy-5 13 Dpy-5 13 Unc-75 8 Dpy-5 8 Unc-101 1 Dpy-5 2 Unc-29 27 Dpy-5Unc-21 Unc-29 18 Unc-75 23 Unc-29 3 Unc-75 92 Unc-29 86 Unc-29 15 Unc-29 6.1°(4.7-7.7) 7.4fc(5.9-9.5) 1.2a(0.7-2.0) 1.0a(0.4-2.0) 0.09°(0-0.9) 13 18.6c(11.7-28.5) 3.3(2.4-4.4) 3.36(2.2-4.9) 24.0(19.1-30.1) 20.2fe(16.6-23.9) 1.6a(0.8-2.5) a Formula 5 (Materials and Methods). b Formula 6. c Formula 4. 117 region of sDpl occurred at a frequency of 10"4. In contrast, when I measured in the dpy-5 unc-54 interval, recombination was only 4 times lower (25 m.u. to 6.1 m.u.). Recombination in the unc-101 unc-54 interval of sDpl was estimated by subtracting the dpy-5 unc-101 distance from the dpy-5 unc-54 distance (6-1 = 5 m.u.). This level of recombination was only three times lower than normal (14.2 m.u. to 5 m.u.)). In summary, sDpl does recombine with the normal homologues, but an altered distribution of events results in 80% of recombination occurring in the unc-101 unc-54 interval. When sDpl recombines with one of the normal chromosomes, what is the segregation behavior of the second normal homologue? The results of the following two experiments support the hypothesis that the two normal chromosomes continue to segregate from each other, regardless of the pairing activity of the sDpl chromosome. In the first experiment, the progeny of dpy-5ldpy-5lsDpl hermaphrodites were scored. From examination of Figure 15, some predictions can be made concerning the outcome of this experiment. If the duplication does not cause nondisjunction of the normal homologues, then 80% of the zygotes are expected to survive assuming the Dp homozygotes die. Higher zygote mortality was expected if the pairing activity of the duplication caused nondisjunction of the normal homologues (e.g. Figure 15; 18/36 dead zygotes or 50%). In the experiment, 81% of the eggs produced by these hermaphrodites eventually developed to adults (495 adults/610 eggs). The ratio of wild-type to Dpy progeny was 1.51 (495 wild-type and 328 Dpy) which was equivalent to a Dp gamete frequency of 43% (section 2.7). Another prediction in Figure 15 is that the frequency of recombinant phenotypes depends on whether the two normal chromosomes segregate independently. If the two normal homologues segregate independently, and assuming the trisomic I progeny are scored, one recombinant class will be recovered in excess over the other. Only if the two normal homologues always segregate from each other when there is also duplication recombination, will there be an equal number of the two recombinant classes (Figure 15). As the data in Table 19 shows, the reciprocal recombinant classes from sDpl strains were recovered in equal numbers. 118 o o u29u75 u29u75• l R V L u29 + u29 + u29u75 * I R V L M75 * u29u75 I R V L t t75 u29u75 Unc-29 Unc-75 Wt Unc-29 Unc-29 I I I * * Unc-75 u29u75 l R v L Wt Wt * * u29 + Unc-29 Wt Unc-29 Unc-29 I I I * * Wt u29 + u29u75 * Unc-29 I I I * * Unc-29 I I I * * Wt * * I R V L u 7 5 * * * * Wt * * u29u75 I R V L « 7 5 Unc-75 Wt * * Figure 15: Punnett square showing the predicted segregation products from unc-29 unc-75lunc-29 unc-75; IRV^hTl hermaphrodites. This figure can be applied to other duplication hermaphrodites. For example, to apply this figure to the dpy-5 unc-54ldpy-5 unc-54;sDpl, replace IRVLhTl with sDpl and unc-29 unc-75 with dpy-5 unc-54. The trisomy I progeny are shown with "111". Empty boxes represent presumed lethal zygotes due to severe aneuploidy. The gamete and zygote boxes marked with "*" occurred only if the the presence of the duplication caused nondisjunction of the normal homologues. As described in Results, these boxes should be ignored when applying the Punnett to sDpl. 119 The second experiment attempted to directly recover sDpl induced nondisjunction events. If sDpl does cause one normal homologue to segregate randomly, then gametes with two dpy-5 marked normal chromosomes should be produced. If these gametes are fertilized by a gamete nullo for all or most of chromosome I, then the resulting progeny should be detectable. As described below, 6-7% of gametes from an sDpl carrying hermaphrodite are recombinant for the duplication. Assuming random segregation of the non-recombinant chromosome I, approximately 3% of the gametes from sDpl worms would be disomic for chromosome I. To test this prediction, dpy-5/dpy-5/sDpl; unc-60lunc-60 hermaphrodites were crossed to +IKT2 males. One quarter of the gametes from + lhT2 males would be the correct genotype (IRIIIRhT2; IIIN) to produce a viable worm upon fertilizing a disomic I oocyte. Normally only the alternate segregation products (IN; n i N o r iRniRhT2; ILIIILhT2 = 50% of total) could fertilize an egg from a duplication hermaphrodite and produce a viable zygote. But a IRIIIRhT2; III N gamete could fertilize a disomic dpy-5/dpy-5 gamete to produce a viable Dpy-5 worm. The viability of this hyperploid has already been shown in section 3.6.3. When +lhT2 males were crossed to dpy-5/dpy-5/sDpl; unc-60lunc-60 hermaphrodites no Dpy-5 worms were observed in 960 progeny. If disomic I gametes were produced, they were at a frequency of less than 1% (3X4/960). From these experiments, I conclude that in sDpl hermaphrodites, the two normal chromosome I homologues continue to segregate from each other despite the pairing activity of sDpl. The meiotic properties of the IRV^hTl chromosome when present as a duplication were determined for comparison to sDpl. IRVLhTl is one of the chromosomes from translocation hTl(I;V) and has the unc-29 unc-54 end of chromosome I attached to the unc-60 dpy-11 region of chromosome V (Figure 13b). There are two significant structural differences between IRV^hTl and sDpl (Figure 1). One is that IRV^hTl is shorter; sDpl includes most of the gene cluster on chromosome I while IRV^hTl does not. The second difference is that a large piece from the left end of chromosome V is attached to the chromosome I sequences of IRV^hTl (Figure 13). IRV^hTl does not recombine with the normal chromosome V in either a euploid translocation strain (section 3.6.2) or as a duplication. 120 Recombination was frequent between IRVLhTl and the normal chromosome I homologues (Table 19). The recombination frequency in the unc-29 unc-54 interval was 20%, a region covering most of the chromosome I portion of IRVLhTl. Thus, IRVLhTl engaged in recombination three times more frequently with the normal homologues than sDpl did. In addition, recombination was more frequent near the left end of/^V^/iTi than that observed in a similar region of sDpl (Table 19). In the unc-29 unc-75 interval, I observed 3.3% recombination, the same region in sDpl had 1% recombination. This data also showed that the distribution of recombination events involving IRV^hTl was more like the normal chromosome. In contrast to sDpl, the relative amounts of recombination in the unc-29 unc-75 and unc-29 unc-54 intervals was approximately the same as in the controls. In addition, in the unc-29 unc-75 experiment I recovered many more Unc-29 than Unc-75 progeny. As shown in Figure 15, this is consistent with the proposal that IRV^hTl pairing activity leads to nondisjunction of the normal homologues. Further evidence that IRV^hTl induced nondisjunction of the normal homologues was collected. Only 68% (329/482) of the zygotes from unc-29lunc-29IIRVLhTl hermaphrodites developed to adults. From these same hermaphrodites, 35% of the gametes had the duplication (464 wild-types and 427 Unc-29). In a similar experiment with sDpl (see above), 81% of the zygotes survived and 43% of the gametes had the duplication. These differences can be explained if IRV^hTl induced nondisjunction of the normal homologues (Figure 15). A recombination defective derivative of IRV^hTl was fortuitously discovered. This duplication, called hDplOl, recombined much less frequently with chromosome I than IRV^hTl (Table 19). Recombination in the unc-29 unc-54 interval was reduced from 20.2% to 1.6%. With such a low level of recombination, is was not surprising that hDpl 01 strains had characteristics more like sDpl strains than IRVLhTl strains. For example, unc-29lunc-29/hDpl 01 hermaphrodites produced a higher proportion of viable zygotes (647/782 = 83%) and more gametes with the duplication (41%; 599 wild-types and 433 Unc-29) than similar IRVLhTl hermaphrodites. These differences were expected if hDpl 01 is was less efficient at meiotic pairing resulting in nondisjunction of the normal homologues (Figure 15). 121 I attempted to map the locus responsible for the reduced recombination activity found with hDpl 01. Unc-29 recombinants from unc-29 unc-54/ unc-29 unc-54/ hDpl 01 were picked. The unc-29 + recombinant chromosome carried the region of the hDpl01 chromosome around unc-54. Recombination was measured in the unc-101 unc-54 interval in the presence of the unc-29 + chromosome by scoring recombination in unc-29 + +1 + unc-101 unc-54 hermaphrodites. Recombination in these strains was normal, indicating the site responsible for the recombination defective phenotype in IRV^hTl did not map close to unc-54. Furthermore, the chromosome V portion of hDplOl carried the unc-60 dpy-11 region (section 2.5). Thus, the recombination defective phenotype did not result from a large deletion of the chromosome V sequences in IRVLhTl. 3.8.2. Duplications with meiotic pairing behavior: The X chromosome. Duplications that meiotically pair with the normal X-chromosome were isolated by modifying the structure of I^X^szTl (section 3.1.2). This chromosome carries the left end of the X chromosome attached to the left end of chromosome I (Figure 16). Strains with ILXLszTl have a Him phenotype (Table 30), presumably because the ILXLszTl chromosome occasionally pairs with one X-chromosome, causing the other to be a univalent and segregate randomly at anaphase. Derivatives of I^X^szTl were isolated that had lost the material around dpy-5 or unc-13. Most of these events were terminal deletions of the I^X^szTl chromosome (section 3.2). For example, most of the dpy-5(-) duplications were unc-1 (+) but let-362(-). The Him phenotype of the parental I^X^szTl results from X-nondisjunction induced by the pairing behavior of the sequences in the unc-1 region (section 3.8.2). Since most of the dpy-5(-) duplications still contained the X-chromosome material (section 3.2 and Figure 10), it was expected they would also cause X-chromosome nondisjunction. It was thus surprising that most of the dpy-5(-) duplications lacked the Him phenotype (Table 30 and data not shown). A resolution to this paradox was suggested by the behavior of hDp69 and hDp71, which were both dpy-5(-) but let-362(+). hDp69 is an internal deletion while hDp71 is either a small internal deletion or a point mutation within 122 bli-4 unc-40 let-362 unc-1 dpy-3 unc-371 | dpy-5 unc-11 bli-3 lLXLszT1 unc-1 dpy-3 dpy-7 unc-11 bli-3 lLXLhT3 unc-1 dpy-3 bli-4 dpy-5 bli-3 hDp56 unc-1 dpy-3unc-40 dpy-5 bli-3 hDp31 unc-1 dpy-3 unc-37 bli-3 hDp51 Figure 16: Diagram oULXLszTl and its derivatives hDP31, hDP56 and hDpSl. ILXLhT3 i shown for comparison. 123 the dpy-5 locus. Unlike the other dpy-5(-) duplications, hDp69 and hDp71 had a Him phenotype, suggesting that the presence of the left end of chromosome I was a critical component in the meiotic pairing activity of I^X^szTl and its derivatives. A third dpy-5(-) deletion derivative of ILXLszTl, hDp51, was deleted in the sup-11 to bli-4 region but had sequences from the left end (let-362) of chromosome I (Figures 11 and 16). hDp51 did not have a Him phenotype but, based on evidence described below, did have weak meiotic pairing activity. The weaker ability to pair might be related to the lesser amount of chromosome I sequences in hDp51 compared to hDp69. Most of the unc-13(-) derivatives of ILXLszTl were also unc-1 (-) but let-362(+) (section 3.2 and Figure 9). These unc-13(-) derivatives also lacked a Him phenotype, as expected from a derivative of I^X^szTl which lacked the X-chromosome material. Two exceptions to this, hDp31 and hDp56, did have meiotic pairing behavior. Both duplications have X-chromosome material from the unc-1 dpy-3 region (Figure 16) and therefore represent internal deletions of I^X^szTl. The meiotic pairing behavior was demonstrated with three types of data. First, hDp31 and hDp56 hermaphrodite strains had a mild Him phenotype (Table 30), indicating these two duplications could successfully compete with the normal X-chromosomes for meiotic pairing causing X-chromosome nondisjunction. Recombination between the duplication and the normal X-chromosome was assayed to test if the nondisjunction resulted from pairing activity of the duplications. To do this, the progeny of dpy-5; unc-1 lunc-1 lhDp56 hermaphrodites were scored. Recombination between the unc-1 bearing X-chromosome and the unc-1 (+) bearing hDp56 resulted in Dpy-5 or Unc-1 progeny. Among 886 wild-types (not including 29 males) and 648 Dpy-5 Unc-1 were 18 Dpy-5 and 19 Unc-1 hermaphrodites. The presence of Dpy-5 and Unc-1 progeny indicated recombination was frequent between hDp56 and the X-chromosome. The Punnett square for this experiments is shown in Figure 17. Examination of this figure reveals that the number of wild-type males observed in the experiment was much more than expected. According to the Punnett, the most common event to produce a wild-type male was a Dp[unc-2]/nullo-X gamete fertilizing a unc-l(+) gamete. The frequency of these zygotes is a function of p^ and thus should have been infrequent relative to the recombinant classes such as 124 0 d5 d5 d5 d5 d5 d5 ul ul + ul + o Dp ulDp ul ulDp d5 Dpy-5 Wt Dpy-5 Unc-1 XXX Unc-1 ul Unc-1 o d5 Wt Wt XXX ul Dp d5 Unc-1 Wt XXX ul ulDp d5 Dpy-5 Wt Dpy-5 Wt XXX Wt + o d5 Unc-1 Wt wt ulDp o o d5 X X X XXX X X X XXX wt + ul Figure 17: Punnett square of predicted segregation from dpy-5; unc-1 /unc-1 lhDp56 hermaphrodites. The trisomic X progeny are indicated with X X X ; these have reduced viability and are Dpy. All zygotes are hermaphrodite except those indicated as male. The normal chromosome I and chromosome X are indicated by abbreviations for the gene they carry; dpy-5 (d5) and unc-1 (ul). 125 Dpy-5 and Unc-1. This was not the case as 29 wild-type males were recovered with the 37 recombinant progeny. Thus another source of wild-type males must be invoked. The most frequent event would be a Dp[unc-1(+^/nullo-X gamete fertilizing an unc-1 gamete. In this situation, the Dp[unc-1 (+j]/nullo-X gamete was generated from the hDp56 causing non-disjunction of the X-chromosomes without itself demonstrating evidence of recombination. This could happen if many recombination events occurred to the left of unc-1, or double crossing over was frequent, or there was hDp56 - X pairing and segregation with no recombination. There is no reason to believe there would be much recombination to the left of unc-1 because it does not occur between normal X-chromosomes. Although Hodgkin, Horvitz and Brenner (1979) observed high levels of interference on the X-chromosomes of hermaphrodites, double crossovers could be more frequent in the situation with the duplication. In the second demonstration of meiotic pairing activity, hDp31 and hDp56 segregated from the X-chromosomes at a very high frequency in males. To score recombination between the duplication and the X-chromosome in males, dpy-5; hDpx[unc-l(+)]/unc-l males were crossed to dpy-5; unc-1 hermaphrodites (Table 20). The wild-type males resulted from non-recombinant duplication/nullo-X sperm while the Dpy-5 hermaphrodites and Unc-1 males resulted from nullo-Dplunc-1(+) and Dp [unc-l]/nullo-X sperm respectively. The nullo-Dp/unc-1 sperm produced Dpy-5 Unc-1 progeny which were not distinguishable from self-fertilization progeny and thus not scored. The wild-type hermaphrodites resulted from nondisjunction of the Dp and the X-chromosome while the reciprocal, Dpy-5 Unc-1 males, resulted from nullo-Dp/nullo-X sperm. The data for the two duplications were similar. The duplications regularly segregated from the X-chromosome. The segregation was more regular than observed with males in which the duplication did not have meiotic pairing activity but segregated from the X-chromosome by a "non-homologous" mechanism 70% of the time (section 3.9). Greater than 90% of the time KDp31 or hDp56 segregated from the X-chromosome in males. All non-disjoining chromosomes in this experiment were non-recombinant; there were no Unc-1 hermaphrodite progeny. Since 30% of the disjoining duplications were recombinant in the interval between unc-1 and the left end of the Table 20 Recombination in X-chromosome duplications bearing males. Duplication Wt Wto* Dpy-5 Dpy-SUnc-ltf Unc-1 Unc-ltf hDp31 2 159 79 10 0 87 hDp56 4 63 29 6 0 31 All crosses were dpy-5; hDp/unc-ltf X dpy-5; unc-1. 127 duplication, approximately 30% of the Dp - X segregation events were not associated with a recombination event. In addition to non-recombinants, these could be products of double crossover events or recombination to the left of unc-1. Similar observations were described above for duplication behaviour in the hermaphrodite. The amount of recombination in the region of the duplication was very high compared to normal chromosomes. High levels of recombination involving a duplication in males was also observed by Herman and Kari (1989). In the third demonstration of meiotic pairing activity, hDp31 homozygous strains were stable, producing very few nullo-Dp gametes. This suggests the two duplications in a homozygote regularly segregate from each other at meiosis. In contrast, duplications without meiotic pairing activity failed to segregate from each other in the homozygous state (section 3.9). Duplication homozygous worms usually had a distinctive phenotype; they are slow growing, thin and clear and produced few progeny. hDp31 homozygous strains were easily maintained as few of the progeny were Dp hemizygotes. No Dpy-5 progeny were observed in 226 wild-type progeny segregating from dpy-5/dpy-5; hDp31/hDp31 hermaphrodites. Dpy-5 progeny were observed in the next generation. hDp56 homozygotes were very sick, produced few progeny and similar numbers could not be generated. To determine the fraction of nullo-Dp gametes produced by a duplication homozygote, dpy-5l+ males were crossed to dpy-5; hDp31lhDp31 hermaphrodites. Only one Dpy-5 was observed in 400 progeny for a nullo-Dp frequency of 0.5%. Based on this type of analysis, hDpSl has meiotic pairing activity. hDp51 homozygous strains were fairly stable; no Unc-13 progeny were recovered in 425 wild-types segregating from unc-13; hDp51lhDp51 hermaphrodites. Many Unc-13 hermaphrodites were observed in the next generation and the hDp51 homozygous strains were difficult to maintain. In summary, hDp31 and hDp56 had meiotic pairing activity based on two observations, the Him phenotype and recombination with the X-chromosome. A third observation suggesting meiotic pairing activity in hDp31 and hDp51 was the segregation of homozygous duplications. These three duplications competed poorly for pairing with the two normal X chromosomes in 128 hermaphrodites, but in males and in Dp homozygous strains they had very proficient pairing activity. A chromosome with similar structure to hDp31 and hDp56 but containing more X-chromosome sequences is I^X^hT3 (Figure 16). Strains carrying I^X^hT3 had a much stronger Him phenotype than hDp31 or hDp56 (Table 30). I^X^hT3 may be more effective at competing with the normal X-chromosomes for pairing because it contains more X-chromosome sequences than hDp31 or hDp56. 3.9 Non-homologous disjunction. In the analysis of the translocations and meiotically pairing duplications, it was concluded that homology and recombination were driving segregation at the reductional division. A second disjunctional system, however, exists in C. elegans. This system is like that observed in Drosophila (Grell 1976), it can drive the segregation of chromosomes independently of homology and recombination. Disjunction of non-homologous chromosomes in C. elegans was first observed by Herman, Madl and Kari (1979). They found that free duplications of chromosome II segregated from the univalent X chromosome in males. I observed the same phenomenon with chromosome I duplications (Table 21). An excess of Dp; nullo-X and nullo-Dp; X gametes over Dp; X and nullo-Dp; nullo-X gametes indicated segregation of the duplication from the X-chromosome. The ratio of (wild-type males + Dpy-5 hermaphrodites) to (wild-type hermaphrodites + Dpy-5 males) indicated the amount of Dp - X segregation. The greater this value (termed the segregation value), the more Dp - X segregation. A segregation value of one indicated random segregation of the duplication with respect to the X-chromosome. Values greater than one indicated some X - Dp segregation. In these experiments, as with those of Herman, Madl and Kari (1979), approximately 70% of the Dp sperm were nullo-X. Smaller duplications {e.g. hDp3, hDp6, hDp22) segregated from the X-chromosome with less efficiency than the larger duplications (e.g. hJDpl2, hDp20). By comparison to homologous segregation, however, non-homologous segregation was not as proficient. In the case of KDp31 and hDp56, which participate in homologous segregation, 90% of the Dp sperm were nullo-X. 129 Table 21 Duplication stability. I. Male Sperm. Duplication Wt. male Wt. herm. Dpy male Dpy herm. Frequency of Dp sperm hDp2 46 36 598 583 0.07 hDp3 168 105 278 294 0.32 hDp4 17 9 219 249 0.05 hDp6 58 37 192 182 0.20 hDpl 2 486 178 241 517 0.47 hDp20 175 73 50 163 0.54 KDp22 165 122 693 674 0.17 hDp31 101 13 16 84 0.53 hDp56 293 16 16 296 0.50 hDP78 1155 886 802 1051 0.52 hDP79 391 248 231 350 0.52 hDp83 911 780 757 904 0.50 All crosses were dpy-5; hDpx male X dpy-5; unc-36 hermaphrodite. 130 Non-homologous disjunction was not limited to free duplications. Herman, Madl and Kari (1979) observed linked duplications also segregated from the X-chromosome in males. As described in the analysis of szTl(I;X) translocation heterozygote males (section 3.6.1), the univalent chromosome ILXLszTl segregated from the other translocation chromosome, IRXRszTl, 67% of the time even though they apparently shared no homologous D N A and the IRXRszTl chromosome regularly recombined with and segregated from the normal chromosome I. Similarly, in hT3(I;X) heterozygous males the two translocation chromosomes tended to segregate from each other (section 3.6.4). Three chromosome I duplications that were linked to an autosome, hDp78, hDp79 and hDp83 (section 3.13.2), also had a tendency to segregate from the X-chromosome in males. As shown in Table 21, these duplications were recovered in half the total sperm but in significantly greater than 50% of nullo-X sperm. Thus, these duplications still segregated from the X-chromosome even though they were linked to another chromosome. These experiments were carried out by mass mating duplication males to dpy-5; unc-36 hermaphrodites followed by transfer of individual hermaphrodites to separate plates. There was a wide range of segregation values for the progeny of each of the individual hermaphrodites from the dpy-5; hDpx X dpy-5; unc-36 mass matings. In the progeny of some hermaphrodites there was no evidence of Dp - X segregation while in others there was strong Dp - X segregation. Since the male X-chromosome had no pairing partner, variation most likely occurred within the population of males. This hypothesis was confirmed by crossing individual dpy-5; hDp79 males to dpy-5; unc-36 hermaphrodites. Some males had segregation values of approximately one, indicating random segregation of hDp79 and the X-chromosome, while others had segregation values of approximately two, indicating non-random segregation such that approximately 70% of Dp sperm were nullo-X. In the initial mass mating experiments, this variation within the duplication bearing male population was hidden because the individual hermaphrodites may have mated with more than one male. The data for hDp78, hDp79 and hDp83 are shown graphically in Figure 18. Most of the data comes from mass mating experiments and thus the X-axis represents individual 131 Figure 18: Non-homologous segregation of linked duplications. Segregation values for the linked duplications a) hDp78, b) hDp79, c) hDp83 and d) the free duplications hDpl2 and hDp20. The cross was dpy-5; hDp males crossed to Dpy-5 Unc-36 hermaphrodites. The strength of the Dp - X segregation is indicated by the" segregation ratio; (Wt. males + Dpy-5 hermaphrodites)/ Wt. hermaphrodites + Dpy-5 males). Each point on the abscissa represents an individual hermaphrodite. Each hermaphrodite has been given a number and they are ordered by segregation ratio. a) Seg. Ratio 2.5 i 2 1.5 0 ™ ™ ™ ^ » ^ — — — — — — — — — — — — — — — — — — — 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Hermaphrodite b) Seg. Ratio 3.5 | 1 2 3 4 5 6 7 8 9 10 11 12 13 Hermaphrodite S e g . ra t io • Wt male*D5 herm/ Wt herm*D5 m a l e 131b Seg. ratio • Wt m a l e * D 5 herm/ Wt h e r m * D 5 male 132 hermaphrodites, not individual males. For comparison, data from similar experiments with free duplications hDpl2 and hDp20 are also shown. Figure 18 shows the variation of segregation values within each experiment with a linked duplication. When a free duplication was used, less variation was observed. It is also evident from Figure 18 that hDp79 was more proficient at segregating from the X-chromosome than hDp78 or hDp83. This may be related to the fact that hDp79 is larger than the other two duplications (see Figure 24). Qualitatively, however, there may be two types of male; one with high, the other with low segregation values. To explain this data, I suggest non-homologous segregation is partly dependent on premeiotic events. Perhaps the frequency of non-homologous segregation depends on events in the stem cell nuclei, such as the relative positions of the two chromosomes. Does non-homologous segregation occur in the hermaphrodite germ lines? Previously, Rose, Baillie and Curran (1984) observed that sDpl strains had a Him phenotype and they suggested this could result from the non-homologous pairing of the X-chromosome and sDpl. Similarly, hermaphrodite strains carrying the linked duplications hDp78, hDp79 or hDp83 had a Him phenotype (Table 32). This phenotype could have resulted from non-homologous pairing of the X-chromosome with the linked duplication fragments. In an attempt to directly observe non-homologous segregation in hermaphrodites, the progeny of free duplication homozygotes of the genotype dpy-5/ dpy-5; hDpx/hDpx were scored. These duplications were not able to recombine with the normal homologues and therefore lacked the chromosome features facilitating homologous pairing. Duplication homozygotes were easily recognized by their slender and clear appearance. Few Dpy-5 progeny would be observed if the two duplications were to segregate from each other. When the free duplication homozygotes were scored, however, there was no evidence for nonrandom disjunction. From a dpy-5; hDp39/hDp39 strain, 30 Dpy-5 hermaphrodites were recovered in 132 wild types. From a dpy-5; hDp!2JhDpl2, strain, 129 Dpy-5 hermaphrodites were recovered in 458 wild types. The wild types included both duplication homozygotes and hemizygotes in a 1 to 2 ratio. The large number of Dpy-5 and duplication hemizygote progeny indicated the homozygous duplications did not segregate from each other during hermaphrodite 133 meiosis. By comparison, homozygotes of hDp31, which did have meiotic pairing activity, rarely produced nullo-Dp zygotes (section 3.8.2). The inability of identical duplications to segregate did not rule out the possibility that non-homologous segregation could occur in hermaphrodites. It may have been significant that in the these experiments, the duplications lacked meiotic pairing activity. When evidence for non-homologous segregation was observed in hermaphrodites, the chromosomes involved, sDpl and the linked duplications, were linked to sequences with meiotic pairing activity. It is possible that non-homologous segregation a priori requires meiotic pairing activity. The analysis of hDpl 02 suggested non-homologous segregation could occur in hermaphrodites. The genetic length of the X-chromosome in hDpl02 heterozygotes was only 24 m.u. (section 3.7). This was substantially less than the 40-45 m.u. for the normal chromosome and implied that half of the hDpl 02 bivalents were non-recombinant. With a genetic map length of 24 m.u., tetrads without an exchange would be expected half the time and would result in approximately 12% male progeny. In fact, the male frequency observed in hDpl 02 heterozygotes was only 0.1% (Table 17), suggesting the non-recombinant X-chromosomes in hDpl 02 heterozygotes efficiently segregated from each other. 3.10 Effect of deletions on recombination. If a deletion removed a site required for homologue pairing, recombination in heterozygotes would be reduced in regions adjacent to the deletion. Rosenbluth, Johnsen and Baillie (1990) recently described the effects that deletions of the left end of chromosome V have on recombination. Terminal deletions were found to have less effect on recombination in adjacent regions than internal deletions. In addition, the effect of the internal deletions was polar; the recombination reduction was only observed on the internal side of the deletion. To determine if the effects of deletions were region specific, I assayed the influence that deletions in different regions of the genome have on recombination frequency. 134 3.10.1. Chromosome I deletions I have compared the effects of deletions on recombination in three regions of chromosome I. Deletions located at both ends of the chromosome and some near the middle have been tested for their effects on recombination in adjacent regions. Mutations mapping near the left end of chromosome I reduced recombination in adjacent regions. To characterized these mutations, a map of the left end of chromosome I was generated (section 3.3 and Figure 11). Of the four crossover suppressors, only one, hDflO, was shown to be a deletion. Two other mutations, h655 and h904, failed to complement each other and hDflO, but no other loci. I did not determine if any of the four mutations were internal deletions. h655 had a severe effect on recombination; no recombination was observed in the bli-3 unc-11 region (Table 22). h61 also had a severe effect on recombination, but it is also associated with a complex rearrangement (section 3.2). h904 and hDflO caused lesser but significant reductions of recombination frequency in adjacent regions (Table 22). For example, recombination was reduced in the bli-3 unc-11 interval from 17.5 m.u. to 2.0 m.u. in the presence of h904. Also tested for recombination suppression properties was tDf3. This deletion has internal breakpoints, does not overlap with hDflO (Figure 11), but does reduce the recombination frequency in the region between it and dpy-5 (Table 22). The distance between the right breakpoint of tDf3 and dpy-5 must be at least the let-357 dpy-5 distance (Table 18; 3.0 m.u.). Thus the observed map distance of 0.4 m.u. was lower than normal. Recombination was measured in other regions of chromosome I in h655 heterozygotes. Recombination suppression continued from the left end to the unc-29 region. In contrast, the recombination frequency in the unc-101 unc-54 interval was elevated compared to controls. This enhancement may be mechanistically similar to recombination enhancements which accompany recombination suppression in translocation heterozygotes. Deficiencies in the unc-29 region (Figure 12), located near the middle of chromosome I, were been tested for recombination suppression. Two gamma induced lethal mutations isolated with szTl as a balancer, hDf9 and h654, were mapped near unc-29 (section 3.2). Both of these 135 Table 22 Recombination in deficiency heterozygotes: The Left end of chromosome I. Genotpye Wt. Recombinants m.u.(C.I.) h655 + +1+ bli-3 unc-11 h655 + +1+ dpy-14 unc-101 h655 + +1+ unc-101 unc-54 h904 + +1 + bli-3 unc-11 hDflO dpy-5 unc-29/ + + + hGl + +1 + bli-3 unc-11 tDf3 dpy-51 + + 568 602 635 651 250 212 1613 0 Unc-11 122 Unc-101 155 Unc-101 13 Unc-11 4 Dpy-5 Unc-29 2 Unc-29 0 Unc-11 4 Dpy-5 0.0(0.0-0.5) 18.6(14.9-22.4) 23.1(19.6-26.7) 2.0(1.1-3.2) 2.3(0.3-6.0) 1.2(0-4.2) 0.0(0.0-1.4) 0.4(0.1-0.9) 136 Table 23 Recombination in deficiency heterozygotes: The unc-29 region of chromosome I. Genotpye Wt. Recombinants m.u.(C.I.) + + nDf23l dpy-5 unc-13 + + + + nDf23l let-362 dpy-5 unc-13 + 1225 18 Dpy-5 1.5(1.0-2.0) 18 Unc-13 376 22 Dpy-5 Unc-13 5.7(3.6-8.4) + + nDf24l dpy-5 unc-13 + 2210 12 Dpy-5 14 Unc-13 0.6(0.4-0.9) + + nDf25l dpy-5 unc-13 + + nDf25 +1 dpy-14 + unc-101 + + + nDf25l let-362 dpy-5 unc-13 + 1687 13 Dpy-5 6 Unc-13 485 39 Unc-101 40 Dpy-14 257 36 Dpy-5 Unc-13 0.6(0.3-0.9) 8.0(6.2-9.8) 13.2(8.9-18.4) dpy-5 unc-13 hDf9l + + + + dpy-5 + unc-13 hDf9/unc-ll + dpy-14 + + + + dpy-5 unc-13 KDf9lbli-3 unc-11 + + + dpy-5 + unc-13 hDf9 +/+ dpy-14 + + unc-101 dpy-5 unc-13 h654l + + + 791 0 1281 4 Unc-11 712 159 Unc-11 676 79 Unc-101 710 0 0.0(0-0.6) 0.3(0.1-0.8) 20.3(16.8-24) 11.8(8.8-13.9) 0.0(0-0.6) 137 mutations reduced the recombination frequency in the dpy-5 unc-13 region (Table 23). hDf9 reduced recombination across the gene cluster and over to unc-11. To the left of unc-11, as measured with bli-3 unc-11, recombination was not reduced and possibly enhanced. To the right of hDf9, recombination was not affected in the dpy-14 unc-101 interval. Both of these mutations were mapped inside the deletions nDf24 and nDf25 (Figure 12). Like hDf9, nDf24 and nDf25 reduced the recombination frequency for a short distance to the left but not to the right of the deletions. A smaller deletion of the unc-29 region, nDf23, did not reduce recombination in the dpy-5 unc-13 region like the other mutations. nDf23 was the only mutation, however, to reduce recombination in the left end of chromosome I. These deletions were unusual in that they had an interchromosomal effect. These data are described in section 3.12. I also tested the effect on recombination of deletions in the vicinity of the right end of chromosome I. Do internal deletions of the right end of the chromosome behave like those at the left end, or can a different pattern of recombination suppression be found? Sequences essential for meiotic homologue pairing appear to be located in this region of chromosome I (section 3.6 and Discussion). For this reason, it was expected that of this region it would reduce or eliminate recombination in adjacent intervals. A genetic map of the right end of chromosome I, with the markers and deletions used, is shown in Figure 19. The deletions of this region were isolated by Anderson and Brenner (1984) using eDf24 as a balancer. eDf24 is a partial deletion of the ribosomal gene cluster, the most distal genetic marker known on chromosome I. Therefore, none these deletions must have internal breakpoints and not delete significant amounts of the ribosomal cluster. I found no effects on recombination frequency with any of the deletions tested (Table 24). It was clear that internal deletions had a different effect on recombination at the right end of chromosome I than at the left end. For example, in eDf4 heterozygotes, there was no effect on recombination in the adjacent unc-101 lev-11 region. It was interesting to observe that recombination could occur between two large heterologies. In hT2leDf3 and hT2leDf6 worms there was high frequencies of recombination 138 nDp4 1 m.u. Limit of hT2 crossover suppression eDf3 eDf6 eDJ7 eDf4 unc-75 unc-101 let-201 unc-59, let-202 let-204 lev-10 lev-11 unc-54 lei-208 Figure 19: Deletions at the right end of chromosome I. Deletion breakpoints were mapped by Anderson and Brenner (1984). The extent of hT2 crossover suppression, which includes the left two-thirds of chromosome I is shown. 139 Table 24 Recombination in deficiency heterozygotes: The right end of chromosome I. Genotpye Wt. Recombinants m.u.(C.I.) unc-101 unc-54 +1+ + eDf24 344 52 Unc-101 14.1(10.1-18.7) let-202/+ 2384 50 Unc-54 3.1(2.3-4.1) let-202 +1+ eDf24 996 20 Unc-54 2.0(1.3-3.0) let-201l+ 2741 56 Unc-54 3.0(2.3-3.9) let-201 +1+ eDf24 220 5 Unc-54 2.2(0.9-4.9) + + eDf3/dpy-5 unc-101 + 404 39 Unc-101 12.1(8.5-16.5) + + eDf6/dpy-5 unc-101 + 408 71 Unc-101 16.1(12.2-20.7) + + eDpldpy-5 unc-101 + 187 31 Unc-101 15.4(10.1-22.2) unc-101 lev-Ill + + 680 32 Lev-11 6.2(4.6-7.6) 25 Unc-101 unc-101 lev-11 +1+ + eDf4 218 20 Unc-101 8.1(5.6-11.2) 17 Lev-11 dpy-5 eDf3; +IKT2 122 9 Dpy-5 13.9(6.4-25.9) dpy-5 eDf4; +lhT2 244 76 Dpy-5 > 50.0(47.6-) dpy-5 eDfS; +IHT2 144 26 Dpy-5 33.7(22.0-54.6) dpy-5 eDf7; +IKT2 44 10 Dpy-5 43.9(21.2->50J 140 (Table 24) between the translocation breakpoint and the deletion breakpoint, suggesting synapsis could initiate at subterminal locations or that synapsis could "jump" heterologies. The low level of recombination in eDf3 heterozygotes may have resulted from the very small region available for recombination in these worms. Another deletion of the right end of chromosome I was mnl64. Originally isolated by Herman, Kari and Hartman (1982), mn.164 was described as a homozygous viable, dominant X-chromosome nondisjunction mutation. Subsequent cytogenetic analysis (D. Albertson, personal communication) has shown it to be a fusion between chromosomes I and X. The mnl 64 chromosome is composed of chromosome I joined at its right end to the left end of the X-chromosome. The attachment was associated with a partial deletion of the ribosomal gene cluster. A simple interpretation of mnl 64 is that a complete X-chromosome attached to a chromosome I which is missing the material distal to the ribosomal cluster. If the sequences essential for for homologue pairing and recombination were located to the right of the ribosomal gene cluster, then mnl 64 would not be expected to recombine with the other chromosome I homologue. Pseudolinkage was observed between unc-54 and unc-1 in mnl 64 heterozygotes. In the progeny of mnl64l unc-54; unc-1 hermaphrodites, only wild-type and Unc-54 hermaphrodites and males were observed (Unc-54 is epistatic to Unc-1). D. Albertson (personal communication) has shown using cytogenetic analysis that mnl 64 homozygotes are trisomic for chromosome I; they are mnl64lmnl 64IIN. This explains how mnl 64 homozygotes could produce 20% of their progeny as males (Herman, Kari and Hartman 1982). If the homozygotes had two copies of chromosome I, genotype mnl64lmnl64, the spontaneous males, genotype mnl64/0, would be lethal due to the monosomy of chromosome I. During meiosis of mnl 641 mnl 6411^ hermaphrodites, however, nondisjunction of the mnl 64 chromosomes could produce a mnl 64/1^ male. Chromosome I trisomy was confirmed in mnl 64 homozygotes with a genetic test, mnl 64/ unc-54 hermaphrodites segregated wild-type and Unc-54 progeny. In the next generation, some of the wild-type progeny segregated no Unc-54 progeny. These were mnl64 homozygotes. When males from one of these strains were crossed to dpy-5 unc-54 hermaphrodites, nine Unc-54 non-Dpy-5 progeny were Table 25 Segregation and recombination in mnl64(I;X) strains. Genotype Wts Wto- Other phenotypes mnl64lunc-101 1214 367 127 Unc-101 ; 218 Unc-lOlo" mnl64lunc-101 unc-54 1274 315 32 Unc-101 ; 9 Unc-101^ mnl64l dpy-5 unc-101 543 190 16 Dpy-5 a ; 13 Dpy-5o-17 Unc-101 ; 21 Unc-lOlo -14/14 Dpy-5 hermaphrodites tested were Him. 142 observed among 218 wild types. The presence of Unc-54 progeny indicated that the mnl 64 homozygotes had a third copy of chromosome I marked with unc-54. Recovering nine Unc-54 progeny among a total of 218 was far less than the expected 50%. While the Unc-54 worms had severely reduced viability, this could not account for their low recovery in this experiment Segregation in mnl 64 heterozygotes did not produce equal amounts of mnl 64 and 1^ gametes. The mnl 64 heterozygotes had two copies of chromosome I. Herman, Kari and Hartman (1982) found one-third of the progeny from mnl64/dpy-3 heterozygotes were homozygotes. In a couple of different mnl 64 heterozygotes I observed a similar frequency of homozygote production (14/35). The production of trisomy I mnl64 homozygous progeny from mnl64 heterozygotes was frequent, indicating high frequency chromosome I nondisjunction at meiosis I or II. The frequency of mnl 64 homozygotes from the heterozygote was so frequent, in fact, that chromosome I segregation must have been almost random. In the data of Herman, Kari and Hartman (1982), equational nondisjunction of mnl 64, but not its homologue, occurred in approximately 4% of ova. Such ova could be fertilized by a I N ; nullo-X sperm (frequency 25% in ova; from Herman, Kari and Hartman (1982)) to produce a mnl 64 homozygote. This could not, however, account for the frequency of mnl 64 homozygotes actually produced. Gametes from a mnl 641+ hermaphrodite carrying both an mnl 64 chromosome and a normal chromosome I must have been frequent. Calculation of recombination frequencies in mnl 64 heterozygotes was complicated by the high frequency of chromosome I nondisjunction and the production of trisomy I progeny. As discussed already, chromosome I and rnril 64 were not recovered equally from the meiosis of mnl 64 heterozygotes. In both the heterozygous male experiment described above and in heterozygous hermaphrodites, the normal chromosome I homozygotes were recovered at a low frequency. When the mnl64lunc-54 males were crossed to dpy-5 unc-54 hermaphrodites, half of the progeny were expected to be Unc-54 if chromosome I and mnl64 sperm were produced at equal frequency. When the progeny of mnl64/unc-101 hermaphrodites were scored (Table 25), only 9% of hermaphrodite progeny were Unc-101. In contrast, 37% of male progeny were Unc-101. The results of two recombination experiments are shown in Table 25. It is clear that 143 recombination occurs between mnl 64 and the normal chromosome I. The mnl 64 chromosome still retains sequences allowing it to pair for recombination during meiosis. hDf8: Using the mutagen formaldehyde, Starr (1990) isolated hDf8 in a precomplementation screen for dpy-14 mutations. This deficiency was used in the analysis of lethal mutations in the bli-4 - dpy-14 region (section 3.4, Figure 12). The hDf8 chromosome is also a crossover suppressor. It prevented recombination in the let-362 dpy-5 interval (A. Rose, personal communication). I assayed recombination to the right of dpy-5. I scored the progeny of dpy-5 unc-75/ hDf8 and found 3.8% recombination (406 wild-types, 15 Dpy-5 and 16 Unc-75). This result was lower than the normal frequency of 8.9% (Table 11), suggesting the crossover suppression boundary was close to unc-75. The hDfS chromosome did not behave like a translocation, the most common type of crossover suppressor in C. elegans. Translocation heterozygotes produce a large number of lethal zygotes due to aneuploidy. I did an egg count from dpy-5lhDf8 hermaphrodites to determine if there was excessive zygote lethality. From 433 eggs there were 219 wild-type and 108 Dpy-5 hermaphrodites. The egg survival frequency of 75.5% agrees with the predictions for an intrachromosomal rearrangement tightly linked to a recessive zygotic lethal mutation. The ratio of wild-types to Dpy-5 progeny, 2.02:1, was also consistent with the hDf8 chromosome rearrangement(s) being limited to chromosome I. 3.10.2. X-chromosome deletions Some deletions behave as crossover suppressors on autosomes I and V (section 3.10 and Rosenbluth, Johnsen and Baillie 1990). Can similar deletions be found on the X-chromosome? Meneely and Herman (1979, 1981) isolated deletions at the right end of the X-chromosome. A map of these deletions is shown in Figure 20. As can be seen from an examination of Table 26, the X-chromosome deletions did not effect recombination like the autosomal deletions. In fact, only in the case of mnDfl 1 was any effect seen at all. mnDfl 1 is a putative terminal deletion. The known internal deletions, mnDf20 and mnDf7, had no effect on recombination. This was tested for an interval to the left (dpy-7 unc-9) and an interval to the right (unc-7 lin-15) of the deletions. 144 1 m.u. mnDpl mnDfl mnDf4S mnDfl 1 mnD}20 mnDf41 unc-9 unc-84 let-4 unc-3 m- lei-l unc-7 Unl5 let-2 Figure 20: Genetic map of the right end of the X-chromosome. The mapping of deletion and mDpl breakpoints is from Meneely and Herman (1979, 1981) and Egdley and Riddle (1987). 145 Table 26 Recombination in X chromosome deficiency heterozygotes Genotype Wt. Recombinants m.u.(C.I.) dpy-7 unc-9 + +1+ + unc-3 let-2 979 100 Unc-9 9.7(7.7-11.8) dpy-7 unc-9 +/+ + mnDf20 930 108 Unc-9 11.0(9.0-13.3) dpy-7 unc-9 +/+ + mnDfl 1 962 44 Unc-9 4.5(3.2-6.0) dpy-7 unc-9 +/+ + mnDf41 790 70 Unc-9 8.5(6.6-10.7) dpy-7 unc-9 +1+ + mnDf7 635 76 Unc-9 11.3(8.7-14.2) dpy-7 unc-9 +1+ + mnDf43 1227 129 Unc-9 10.0(8.2-11.9) + dpy-7 unc-9IstDfl + + 812 93 Unc-9 11.0(8.7-13.4) + unc-7 lin-15 -h/unc-3 + + let-2 858 14 Unc-7 1.6(0.9-2.6) + unc-7 Un-15lmnDf7 + + 1078 19 Unc-7 1.7(1.0-2.6) + unc-7 Un-15lmnDf20 + + 1631 32 Unc-7 1.9(1.3-2.6) 146 As discussed already (section 3.7), two large insertions into the X-chromosome, hDpl4 and hDpl 02, caused significant, but polar, decreases in recombination. In contrast, none of the deletions tested had a similar polar effect. It might have been expected that mnDf7 would have reduced recombination in the unc-7 lin-15 interval like hDpl4 since both would produce a heterology in the synapsed X-chromosome. The deletion stDfl (Figure 1), which is located in the same region of the X-chromosome as hDpl4, had no effect on recombination in the dpy-7 unc-9 region (Table 26). 3.11 Meiotic mutants him-3, him-6 and him-8. Meiotic mutations in the genes him-3(IV), him-6(IV) and him-8(IV) were originally isolated by Hodgkin et al. (1979) because they caused a high level of X-chromosome nondisjunction. Him-6, and possibly Him-3 mutants, were found to have elevated levels of autosomal nondisjunction. The him-3 and him-6 alleles used in the experiments reported here, el256 and el423, were the more severe of the two alleles known for each gene. Him-8 hermaphrodites produced many male progeny due to random disjunction of their X-chromosomes. Unlike the first two mutants, him-8 did not appear to affect autosomal disjunction. I have tested the effects of him-3 and him-6 on recombination frequency in different regions of chromosomes I, IH, V and X and tested the effects of all three mutants on the segregation patterns of free duplications. Previous to this study, the genetic map positions of him-3 and him-6 was not well defined. These genes were known to map on the right end of chromosome IV (Figure 1), between dpy-13 and unc-31 (Figure 21). In order to position them more accurately, him males were crossed to a series of deficiencies from the right end of chromosome IV (section 2.6). him-3 was complemented by all the deficiencies tested. This placed him-3 in the same region as unc-5 and unc-24 (Figure 21). Unfortunately, there are no deficiencies in this region. him-6 failed to complement sDf2 and sDf60 but complemented mDf7. This placed him-6 close to unc-22 (Figure 21). Using these deficiencies, it was possible to test if the him-6(1423) mutation was a hypomorph or an amorph. The frequency of males from him-6/sDf60 (18/83 = 0.22) and him-6lsDf2 (49/264 = 0.19) hermaphrodites was slightly higher than in him-6lhim-6 147 mDf9 mnDf4 T mDf8 1 m.u. eDf 19 mDfl j sDf2 sDf60 unc-17 lei-275 dpy-13 let-277 unc-5 him-3 unc-24 - daf-15 - daf-14 him-8 unc-22 him-6 unc-31 Figure 21: Genetic map of the dpy-13 unc-31 region. The mapping of deletion breakpoints is from D. V. Clark (1990) and Rogalski and Riddle (1988,) him-3 and him-6 were mapped to this interval by Hodgkin, Horvitz and Brenner (1979). The revised position based on experiments in this thesis are shown in the figure. 148 hermaphrodites (average of 13.7% in recombination experiments; Table 27). In addition, the number of progeny produced by him-6lsDfs hermaphrodites (279/7 = 40) was less than from him-6lhim-6 hermaphrodites (average of 110 in recombination experiments). Because the phenotype of el 423 was more severe when heterozygous to a deficiency, the el423 mutation was classified as a hypomorph. 3.11.1. Effects on recombination. Compared to the controls (Table 11), him-3 mutants had a reduced recombination frequency on chromosome I (Table 27). The reductions were more severe in the bli-3 unc-11 interval (25% of control) than in the dpy-5 unc-101 interval (68% of control). The severity of the reduction in the bli-3 unc-11 interval of him-3 hermaphrodites was confirmed when different markers (let-362 dpy-5) on the left end of chromosome I were used. On chromosomes III of him-3 mutants, recombination was reduced in the dpy-17 unc-64 interval (65% of control) and to a lesser degree in the unc-45 dpy-17 interval (80% of control). On chromosome V, recombination in the unc-60 dpy-11 interval was reduced to 50% of the control frequency. On the X-chromosome, him-3 had virtually no effect on recombination frequencies. On the left end of the X-chromosome, the recombination frequency in the unc-1 dpy-7 interval was 82% of controls and in the dpy-7 unc-3 interval on the right end of the X-chromosome the recombination frequency 97% of controls. There was a deficiency of Dpy-7 recombinants compared to Unc-3 recombinants in the latter experiment. This may have been a viability problem since the Dpy-7 Him-3 hermaphrodites were often sicker and slower growing than normal Dpy-7 worms. In the former experiment, a different dpy-7 allele was used and I did not observe any viability problem in this case. Considering the high level of X-chromosome nondisjunction in Him-3 hermaphrodites, the lack of an effect on X-chromosome recombination was surprising. Approximately 15% of the 149 Table 27 The effect of him mutations on recombination frequency. Genotype Wt. a Recombinants0 m.u.(C.L) Exp./Control him-3; bli-3 unc-11/ + + 666 19 Unc-11 4.3 (2.4-6.6) 0.25 let-362 dpy-5 unc-13/ + + 198 6 Dpy-5 Unc-13 4.4(2.2-9.9) 0.29 dpy-5 unc-101/+ + 502 24 Dpy-5 8.1(6.0-10.6) 0.68 31 Unc-101 unc-45 dpy-171 + + 535 64 Dpy-17 18.0((14.8-21.5) 0.80 67 Unc-45 dpy-17 unc-641 + + 403 40 Unc-64 14.6(9.9-16.2) 0.65 30 Dpy-17 unc-60 dpy-11/ + + 308 20 Dpy-11 9.1(6.2-12.5) 0.50 18 Unc-60 unc-1 dpy-7/ + + 360 40 Unc-1 16.1(12.6-20.3) 0.82 39 Dpy-7 dpy-7 unc-31 + + 1050 132 Unc-3 18.5(15.1-22.3) 0.97 65 Dpy-7 150 Table 27 (con't) Genotype Wt. c Recombinants'1 m.u.(C.I.) Exp./Control him-6; bli-3 unc-11/ + + 926 82 Unc-11 13.1(10.2-16.2) 0.75 dpy-5 unc-75/ + + 368 19 Dpy-5 8.6(5.6-10.6) 0.95 20 Unc-75 dpy-5 unc-54/ + + 877 152 Dpy-5 25.4(20.8-30.2) unc-45 dpy-171 + + 611 67 Unc-45 16.4(13.6-19.5) 0.69 69 Dpy-17 dpy-17 unc-64/ + + 425 33 Dpy-17 12.5(9.6-15.7) 0.56 39 Unc-64 unc-60 dpy-111 + + 1180 78 Dpy-11 10.0(8.3-11.6) 0.55 81 Unc-60 dpy-11 unc-511 + + 505 70 Dpy-11 20.3(15.6-26.1) 0.80 dpy-7 unc-31 + + 764 81 Unc-3 15.6(11.9-19.5) 0.82 unc-1 dpy-71 + + 560 56 Dpy-7 13.7(11.0-16.7) 0.82 48 Unc-1 Male progeny not included. 151 progeny from him-3 hermaphrodites were males. In addition, the lack of an effect of him-3 on X-chromosome recombination differed from the observations made on the autosomes. Results with him-6 indicated it may be more general in its effects on recombination frequency, but like him-3, affects some regions more than others (Table 27). In the bli-3 unc-11 region, recombination frequency was reduced in him-6 (75% of control) but not to the degree of reduction in him-3 mutants. There appeared to be no effect of him-6 in the dpy-5 unc-54 region. In contrast, recombination was reduced over the entire lengths of chromosome III and chromosome V. Recombination was also generally reduced on the X-chromosome. Unlike the situation with him-3, the X-chromosome non-disjunction phenotype of him-6 correlated with the reduction in recombination frequency. Overall, him-3 affected recombination in some regions more than others, whereas him-6 mutants, with the notable exception of the right end of chromosome I, had more generalized reductions in recombination. 3.11.2. Autosomal nondisjunction. The level of chromosome I nondisjunction in him-3 and him-6 hermaphrodites was assayed in a similar fashion to the experiments with sDpl (section 3.8.1). KT2I+ males were crossed to dpy-5; unc-64; him-3 or dpy-5; unc-36; him-6 hermaphrodites. If a disomy I gamete was produced by the Him strain, then there was a 25% chance it would be fertilized by a IRIIIRhT2; III^ sperm to produce a Dpy-5 worm. In the control, hT2l+ males were crossed to dpy-5; unc-36 hermaphrodites. No Dpy or Unc progeny were observed in 1882 progeny. Thus, the normal chromosome I nondisjunction frequency was less than 0.002. From the him-3 cross 14 Dpy-5 hermaphrodites and 4 Dpy-5 males were recovered in 114 wild-type hermaphrodites and 106 wild-type males. The excess of Dpy-5 hermaphrodites may have resulted from viability differences between hermaphrodites and males or from a segregation pattern such that disomic I gametes were often disomic X as well. The frequency of chromosome I nondisjunction in the Him-3 oocytes was (18/238) 0.076. I also recovered 4 Unc-64 hermaphrodites and 5 Unc-64 males for a chromosome III nondisjunction frequency of 0.039. 152 In the him-6 cross, seven Dpy-5 hermaphrodites and five Dpy-5 males were recovered in 471 wild-type hermaphrodites and 451 wild-type males. In this case there was no excess of Dpy-5 hermaphrodites. The frequency of chromosome I nondisjunction in the Him-6 oocytes was (12/934) 0.013. In the same experiment, four Unc-36 hermaphrodites and three Unc-36 males were recovered, giving a chromosome III nondisjunction frequency of 0.007. Effects on duplication segregation: I tested the effects of these two mutants and him-8 on the segregation of chromosomal duplications. In the experiments described here, the self-fertilization progeny of dpy-5; hDpx; him-y hermaphrodites or the progeny of dpy-5; hDpx; him-y males crossed to dpy-5; unc-36 hermaphrodites were scored (Table 28). With the possible exception of hDp31, the him mutants had little influence on duplication stability in the self-fertilization experiments. In these experiments, a variety of duplications were used; both stable and unstable, one with meiotic pairing activity (hDp31) and the rest without (section 3.9). In the male experiments, the effects of the him mutants on the segregation of duplications from the X-chromosome was tested. In normal males, hDp 12 segregated from the X-chromosome by a non-homologous process (section 3.9). him-3 disrupted the non-homologous segregation of hDpl2 resulting in equal numbers of male and hermaphrodite wild-types and Dpy progeny. In contrast, him-6 and him-8 did not effect non-homologous segregation of hDp!2 males. None of the him mutants had an effect on the frequency of duplication recovery in the male gametes. Normally in the male, hDp31 homologously pairs with the X-chromosome (section 3.8.2). hDp31 continued to segregate from the X-chromosome in both him-3 and him-8 males although the efficiency of segregation was reduced compared to controls. 153 Table 28 The effect of Him mutations on duplication stability. Genotype Wt. male Wt. herm. Dpy male Dpy herm. pa him-(+); dpy-5; hDp31 cfb 101 13 16 84 0.55 dpy-5; hDpl2 o"° 486 148 241 517 0.47 him-3; dpy-5; sDp2 24 102 10 72 0.43 dpy-5; hDp3 25 100 13 112 0.33 dpy-5; hDpl2 24 182 24 124 0.41 dpy-5; hDpl2 o"° 88 91 101 114 0.45 dpy-5; hDp31 103 227 72 225 0.36 dpy-5; hDp31 cfb 126 80 103 134 0.47 him-6; dpy-5; hDpl2 26 235 13 183 0.39 dpy-5; hDpl2 dh 62 26 27 48 0.53 him-8; dpy-5; hDpl2 375 557 252 486 0.39 dpy-5; hDpl2 o"° 102 70 91 126 0.44 dpy-5; hDp31 160 291 85 224 0.42 dpy-5; hDp31 o"6 188 53 51 191 0.49 a p = the fraction of duplication gametes; calculated from the sum of male and hermaphrodite progeny. b Progeny from crossing indicated males to dpy-5; unc-36 hermaphrodites. 154 3.12 Tests for an inter-chromosomal effect. Enhancements of recombination frequency in the portion of a chromosome still recombining in a variety of crossover suppressor heterozygotes has been described. Is this enhancement related to a more global inter-chromosomal effect or is it restricted to the crossover suppressed chromosomes, an intrachromosomal effect? This question has been addressed by McKim, Howell and Rose (1988) for szTl(I;X) heterozygotes. Four intervals on other chromosomes were tested in szTl(I;X) heterozygotes and in no case was an enhancement observed. The four intervals tested were located in gene cluster regions of three different autosomes. These intervals were chosen because in Drosophila, inter-chromosomal effects are observed in regions of low intrinsic exchange (Luchessi 1976). The recombination enhancements reported in this thesis have been observed in non-cluster regions. For example, in chromosome I crossover suppressors such as translocation heterozygotes, the unc-101 unc-54 region was expanded. To test for an interchromosomal effect, recombination was measured in a non-cluster interval in the presence of a heterozygous crossover suppressor of another chromosome (Table 29). No recombination enhancement in the unc-101 unc-54 interval (non-cluster) was observed in the presence of eTl(III;V)/+ (non-chromosome I crossover suppressor). Two chromosome I crossover suppressors, h655 and szTl, had no effect on recombination in non-cluster intervals of chromosome HI (Table 29). In conclusion, there is no evidence supporting the existence of an interchromosomal effect of crossover suppressors in C. elegans. The nDf deletions of the unc-29 region (Figure 12) caused small reductions in recombination on other chromosomes. For example, recombination was reduced in the dpy-11 unc-42 and unc-36 dpy-18 regions in the presence of the nDf deletions. hDf9 did not exhibit any of these interchromosomal effects. Because hDf9 reduced recombination in adjacent regions of chromosome I but not on other chromosomes, the interchromosomal effects of the nDf deletions may be caused by the haploidy of a gene(s) they delete and not be related to the effects they have on chromosome I recombination. 155 Table 29 Effect of crossover suppressors on recombination of other chromosomes. Genotpye Wt. Recombinants m.u.(C.I.) unc-101 unc-54/ + +; eTl(III;V)/+ 302 33 Unc-101 16.1(10.9-22.7)° szTl(I;X)l +; unc-45 dpy-171 + + 530 64 Dpy-17 18.8(15.6-22.5) 72 Unc-45 hDfdl +; unc-36 dpy-181 + + 1127 71 Dpy-18 9.1(7.5-10.8) 67 Unc-36 hDfdl +; dpy-11 unc-42/ + + 630 15 Dpy-11 4.3(2.9-5.7) 21 Unc-42 nDf23l +; dpy-11 unc-421 + + 970 23 Dpy-11 3.1(2.2-4.1) 17 Unc-42 nDf24l +; dpy-11 unc-42/ + + 558 4 Dpy-11 1.1(0.4-2.0) 4 Unc-42 nDf25l +; dpy-11 unc-421 + + 489 2 Dpy-11 0.7(0.3-1.7) 3 Unc-42 nDf23l +; unc-36 dpy-18/ + + 538 16 Dpy-18 5.9(4.4-7.9) 27 Unc-36 nDf25l +; unc-36 dpy-181 + + 310 11 Dpy-18 5.5(3.5-8.5) 13 Unc-13 ° The eTl homozygotes were ignored because their phenotype (Unc-36) was epistatic to Unc-101. 156 3.13 Segregational behavior of free duplications The remainder of the thesis deals with the segregational stability of free chromosome duplications; that is the fidelity at which the duplications are segregated at cell division into the daughter cells, and how well their structural integrity is maintained. The isolation of these duplications was described in section 3.1.2. With these duplications, the effect of chromosome size on segregational stability could be tested. As described in the Discussion, many of the experiments did not distinguish between mitotic and meiotic stability, although there was evidence that the conclusions applied to both mitotic and meiotic cell divisions. 3.13.1. Segregation analysis of the duplications. Duplication stability in the hermaphrodite: All of the duplications from the screens described in Tables 4 and 5 and some of those from Table 6 were tested for segregation stability. Strains of the genotype dpy-5/ dpy-5/ hDpx[dpy-5(+) unc-13(-J] or unc-11/ unc-11; hDpx[unc-11(+) dpy-5(-)] or unc-13/ unc-13; hDpx[dpy-5(-) unc-13(+)] were constructed. The ratio of wild type to mutant progeny was used to calculate the gametic frequency of each duplication (section 2.7). These results are summarized in Table 30 and Figure 22. The results from dpy-5; sDp2, dpy-5; I^X^szTl, unc-11; sDp2 and unc-13; I^X^szTl were not significantly different and thus experiments with different markers and duplications were comparable. The data in Table 30 does not include duplication homozygotes, which were slow growing, small and clear. Every duplication tested is not reported because those with similar breakpoints usually gave similar results. The calculations for gamete frequency assume the contributions from the sperm and oocyte were equal. As shown below, this was not always the case. Most duplications derived from deletion of sequences to the right of dpy-5 in either sDp2 or ILXLszTl were recovered in approximately 40% of the gametes. This included liDpl8 which is deleted at both ends of sDp2. Significant differences in duplication stability compared to sDp2 or ILXLszTl were found with hDpl4, hDp62, hDp72 and hDp73. hDp 14 was expected to be very stable since it is inserted into the X-chromosome (section 3.7). hDp73 was also very stable. Strains containing this duplication segregated worms which grew to adults but were sterile. Table 30 Duplication Stability. II. Hermaphrodite gametes. Duplication Wt a Dpy or Frequency of Unc a Dp gametes dpy-5; sDp2 563 405 0.41 dpy-5; hDp2 1067 1782 0.23 dpy-5; hDp3 374 328 0.36 dpy-5; KDp4 339 529 0.24 dpy-5; hDp6 351 693 0.20 dpy-5; hDp20 183 122 0.43 dpy-5; hDp 12 613 434 0.41 dpy-5; hDp 14 231 (27) 82 (32) 0.58 dpy-5; hDpl5 834 591 0.41 dpy-5; hDpl 8 464 352 0.40 dpy-5; ILXLszTl 401 (30) 273 (20) 0.42 dpy-5; hDp31 1332 (6) 936 0.42 dpy-5; hDp34 268 206 0.39 dpy-5; hDp36 381 283 0.40 dpy-5; hDp37 730 668 0.35 dpy-5; hDp54 751 488 0.43 dpy-5; hDp56 500 (22) 410 (4) 0.37 dpy-5; hDp57 209 128 0.45 dpy-5; hDp62 151 814 0.08 dpy-5; hDp72 134 303 0,18 158 Table 30 (con't) Duplication Wta Dpy or Frequency of U n c ° Dp gametes unc-11; sDp2 828 570 0.42 unc-11; hDplO 574 628 0.31 unc-11; hDpll 525 934 0.22 unc-11; ILXLhT3 515 (77) 341 (4) 0.43 unc-13; ILXLszTl 438 (16) 305 (6) 0.42 unc-13; hDp49 597 451 0.40 unc-13; JiDpSO 500 480 0.34 unc-13; hDp66 348 590 0.23 unc-13; hDp69 277 (15) 276 (3) 0.33 unc-13; hDp73 553 248 0.53 Male progeny scored are in parentheses. 159 Figure 22: The relationship between the extent of duplicated material and the stability of the chromosome. Within the brackets, the percentages indicate the fraction of gametes carrying a duplication. If only a single percentage is shown, this refers to the average recovery of hermaphrodite sperm and oocyte as measured from self fertilization experiments. When more than one percentage is shown, each number is derived from a different assay and presented in the form: [hermaphrodite self-fertilization; oocyte; male sperm]. See the text and Tables 21, 30 and 31 for details. b l i - 3 lec-362 l i n - 6 I lin-17 sup-11 _ I fog-1 I unc-11 unc-74 unc-57 I I I dpy-5 unc-40 b l i - 4 dpy-14 unc-13 I I I I I 3Dp2 [ 4 1 % ; 41%] hDp!2 ( 4 1 % ; 4 2%; 43%] hDp72 [ 1 8 % ] \ hDplB [ 4 0 % ; 32%;] hDp2 ( 2 3 % ; v a r i a b l e ; 7%] DDp3 [ 3 6 % ; 2 1 % ; 32%] hDp€ [ 2 0 % ; -; 20%] HDp4 [ 2 4 % ; 1.2%; 5%] hDp20 [ 4 3 % ; -; 54%] hDp22 [ 2 2 % ; -; 17%] hDplO [ 3 1 % ; 32%] hDpll [ 2 2 % ; 2.4%] hDp8 [ 2 2 % ] hDp62 [8%] hDp69 (33%) - _ | Ul 0> 160 Like hDpl4, hDp73 may have been an insertion into another chromosome, but unlike hDpl4, the insertion disrupted an essential gene. hDp72 and hDp62 were very unstable. These two duplications were also deleted at the left end of chromosome I, but this alone could not explain their instability because other duplications missing the left end, such as hDpl8, were much more stable. While most duplications with deletions to the right of dpy-5 had similar stabilities, duplications with deletions to the left of dpy-5 varied with respect to their recovery frequency in gametes. Most duplications with breaks in this region had reduced stability compared to sDp2 and other duplications with deletions to the right of dpy-5. This could have been caused by the larger size of the deletions relative to sDp2 and the duplications deleted to the right of dpy-5 (Figure 22). In addition, duplications carrying the unc-74 unc-57 region were often more stable than those deleting it. Thus unlike the dpy-5 unc-13 region where deletion of chromosomal material had only small effects on duplication stability, deletions in the unc-11 dpy-5 region had significant effects on stability. hDp20 was much more stable than expected for its size; its breakpoint was between unc-38 and unc-63 but was found in 43% of the hermaphrodite gametes. Other duplications of similar size {e.g. hDp6) were recovered in fewer (20%) gametes. The hDp20 chromosome also contained the left half of chromosome V (section 3.2). These sequences may have conferred the added stability to hDp20. hDp2, another duplication of anomalous behaviour, covered unc-74 but was found to be unstable like unc-74(-) duplications (Table 30 and below). Perhaps a unique structural feature of hDp2 caused these differences compared to hDp3, 5, and hDp7. Four duplications were isolated with deletions spanning from the left of dpy-5 through the unc-13 region (section 3.2). These duplications fell into two classes. hDp9 and hDplO were found in approximately 31% of the gametes while hDp8 and hDpll were found in approximately 22% of the gametes. The stabilities of these four duplications did not correlate with their genetic size (see Figure 8). For example, hDplO was more stable than hDp8 despite its smaller size. Like hDp2, features besides size, such as a structural feature, may account for these differences. 161 The stabilities of I^X^szTl derivatives which had lost the left end [dpy-5(-) unc-13(+)] of chromosome I were less stable than sDp2 but as or more stable than those sDp2 derivatives deleted to the left of dpy-5. The X-chromosome sequences attached to the right end of I^X^szTl might have conferred the added stability to the shorter duplications (Figure 10). With regard to comparing the effects of deleting DNA in the unc-11 dpy-5 region to the dpy-5 unc-13 regions, hDp69 was of interest because it was a small deletion of I^X^szTl in the unc-11 dpy-5 region with flanking regions intact (Figures 10 and 22). Its stability was similar to hDp3 and hDplO despite the fact it contained more DNA (Figure 22). The extra DNA may not have been as significant in determining the stability of hDp69 as the deletion of sequences in the unc-11 dpy-5 region. The following experiments show that most of the reduction in duplications stability could be accounted for by loss during oogenesis. The duplications were usually recovered at a higher frequency in hermaphrodite sperm than oocyte. The fraction of oocytes carrying a duplication chromosome was assayed by crossing heterozygous males to duplication hermaphrodites and scoring the cross progeny. In the crosses described below, the frequency of wild-type progeny (which carried the duplication) was equivalent to the frequency of oocytes which carried the duplication (Table 31). In Table 31, if the male was heterozygous for a crossover suppressor [hT2(I;III)] then all the wild-type progeny carried the duplication. If the male did not carry a crossover suppressor, the wild types had to be progeny tested to see if the duplication or a recombinant chromosome was present. When hT2(I;III)[+ dpy-5 +; + ]/ unc-11 + dpy-14; + males were crossed to dpy-5 dpy-14; sDp2 hermaphrodites, 41% of the cross progeny carried the duplication. Thus the contribution from sperm and oocyte in the sDp2 hermaphrodite was equal. Similar experiments with hDp!2 showed it was also recovered in similar frequencies in the two hermaphrodite germ lines. The smaller hDpl5 was recovered in slightly fewer oocytes. In a cross with hDpl8 hermaphrodites, 32% of the cross progeny carried the duplication. This reduced recovery frequency of hDpl8 compared to hDp!2 (42%) could be attributed to the left end deletion of hDpl 8. Whereas hDpl5 and hDpl8 had similar stabilities to sDp2 in the experiments 162 Table 31 Duplication Stability. III. Oocytes. Wt. herm. Wt male Dpy-5 herm. Dpy-5 male aFrequency of Dp oocytes bdpy-5 hT2l unc-11 dpy-14 X dpy-5 dpy-14; sDp2 118 112 97 66 0.41 dpy-5 hT2l unc-lldpy-14 X dpy-5 dpy-14; hDpl2 23 32 38 39 0.42 dpy-5 hT2l unc-lldpy-14 X dpy-5 dpy-14; KDpl5 58 51 97 92 0.37 dpy-5 hT2l unc-lldpy-14 X dpy-5 dpy-14; hDpl8 62 56 132 117 0.32 dpy-5 unc-13/ unc-11 dpy-14 X unc-11 dpy-5; hDp3 53 c N . D . C 261 248 0.21 dpy-5 unc-13/ unc-11 dpy-14 X unc-11 dpy-5; hDp4 3C N . D . C 243 258 0.012 bli-3 unc-11/ dpy-5 hT2 X unc-11 dpy-5; KDplO 16 17 52 50 0.32 bli-3 unc-11/ dpy-5 hT2 X unc-11 dpy-5; KDpll 3 0 71 53 0.024 163 Table 31 (con't) Wt. herm. Wt male Dpy-5 herm. Dpy-5 male aFrequency of Dp oocytes dpy-5 hT2l unc-11 dpy-14 X dpy-5 dpy-14; hDp59 12 6 100 82 0.032 dpy-5 hT2l unc-11 dpy-14 X dpy-5 dpy-14; hDp74 43 37 80 59 0.21 dpy-5 hT2l unc-11 dpy-14 X dpy-5 dpy-14; hDp22 94 59 212 180 0.13 a Fraction of duplication carrying oocytes calculated on the basis of Wt and Dpy progeny. Dp oocytes = Dp worms/ Total progeny. Dp worms = 2 X Wt except in the sDp2 experiment where Dp worms = Wt. Total progeny = Dp worms + (2 X Dpy). b The crosses are indicated above the data. The male genotype is on the left and the hermaphrodite genotype in on the right. c These experiments required progeny testing of the Wt progeny because no balancer was used in the male. Males were not progeny tested and thus not included in the data. 164 combining hermaphrodite sperm and ooctyes (Table 30), small differences were observed in the oocyte experiments. Thus the oocyte experiments may have been more sensitive for detecting duplication loss in the hermaphrodite. For hDp[unc-ll(-) dpy-5(+)] strains, unc-11 dpy-14/ dpy-5 unc-13 males were crossed to unc-11 dpy-5; hDpx hermaphrodites. KDp4 was found in only 1.2% of the oocytes (Table 31). When the same cross was done with hDp3, 21% of the oocytes carried the duplication. In self fertilization experiments, hDp3 and hDp4 were found in 36% and 24% of the gametes respectively (Table 30). Knowing the hermaphrodite oocyte (Table 31) and male sperm duplication frequencies (Table 21), it was calculated that both hDp3 and hDp4 were found in approximately 41% of the hermaphrodite sperm. Therefore, the low frequency of oocytes carrying hDp3 or hDp4 accounted for the reduction from 41% in sDp2 self fertilization experiments to 36% in hDp3 and 24% in the hDp4 self fertilization experiments. hDp2 was lost frequently during oogenesis, but the results were inconsistent. When dpy-5 unc-13/ unc-11 dpy-14 males were crossed to seven dpy-5 dpy-14; hDp2 hermaphrodites, the frequency of /iDp2-bearing oocytes varied from 1 to 25% (data not shown). This was probably because KDp2 was significantly less stable in the somatic cells than hDp3 and could be lost at various stages of germ cell proliferation (section 3.13.3). The frequency of hDplO and hDpll in the oocytes was tested by crossing bli-3 unc-11 +; + / hT2(I;III)[+ + dpy-5; + ] males to unc-11 dpy-5; hDpx hermaphrodites. hDplO was recovered in 32% of the oocytes (Table 31); a rate similar to the self fertilization frequency (Table 30). This indicated that duplication loss, while greater than that for sDp2, occurred equally in both germ lines. In contrast, hDpll was recovered in only 2.4% of the oocytes; a rate ten-fold lower than the self-fertilization frequency. A duplication of spontaneous origin, hDp22 (section 3.13.2; Figure 24) also had similar loss frequencies in the two germ lines (Table 31). These results suggest there are features of a duplication determining if it is hypersensitive to loss during oogenesis. 165 The instability of these duplications resulted in part from somatic loss. Mosaic worms were frequently recovered with strains carrying the unstable duplications but rarely with strains carrying the stable duplications (such as sDp2 and hDpl 5). Mosaic worms (having a Dpy or Unc phenotype) were recovered from strains carrying an unstable duplication and unc-11 dpy-5, unc-57 dpy-5 or dpy-5 dpy-14 chromosomes. Isolation of mosaics with unc-74 dpy-5 or unc-63 dpy-5 chromosomes is described later. If a duplication was lost early in germ line development, an unc dpy; Dp[+ +] hermaphrodite would lack the duplication in its germ line and produce no wild-type progeny. hDp2, hDp4, hDp59 (see below), hDp62 and hDp72 strains frequently produced germ line mosaics that segregated no duplication progeny. These results indicated a higher level of somatic loss compared to most other duplications, even those with similar stabilities when measured by duplication recovery in the gametes. Duplication stability in the male: Duplication stability in the male was tested by crossing dpy-5 +1 dpy-5 dpy-14; hDpx males to dpy-5; unc-36 hermaphrodites and scoring the progeny. The Dpy (non-Unc) progeny did not receive the duplication while the wild-type progeny resulted from fertilization with a duplication carrying sperm. The results are shown in Table 21 and Figure 22. The duplication recovery in the male sperm was similar to that observed in hermaphrodite oocytes and considerably lower than for hermaphrodite sperm. For example, hDp3 was found in 32% of the male sperm compared to 41% of the hermaphrodite gametes. hDp4 was found in 5% of the male sperm, 2% in the hermaphrodite oocytes and 40% of the hermaphrodite sperm. In general, more loss was detected in male sperm and hermaphrodite oocytes than in hermaphrodite sperm. An excess of wild-type males (Dp; XO) and Dpy hermaphrodites (nullo-Dp; XX) over wild-type hermaphrodites (Dp; XX) and Dpy males (nullo-Dp; XO) indicated preferential segregation of the Dp from the X-chromosome. This is more thoroughly discussed in section 3.9. 166 3.13.2. S p o n t a n e o u s s h o r t e n i n g o f d u p l i c a t i o n s : Several duplications have been observed to spontaneously shorten. These duplication derivatives were isolated on the basis of an exceptional phenotype, either by chance or in selective screens (see below). For example, unc dpy; Dp strains rarely segregated Unc worms that produced either no Unc worms in their progeny (and thus were probably a genetic mosaic; section 3.13.3) or produced both Unc and Dpy Unc progeny. This latter class may have been the result of a deletion in the duplication chromosome to uncover the unc mutation, or a new mutation either on the duplication or elsewhere in the genome. Fourteen of these Unc strains were analyzed by complementation testing (see below) and all of them were found to result from a loss of material from one end of the duplication and not through mutation or recombination. The shortening events were not limited to the left end of sDp2. Some spontaneous duplications, for example hDp30 and hDp77, were dpy-14(-)\ the result of sequences being lost from the right end of sDp2. In three cases, a duplication which had previously shortened continued to shorten at a high frequency. For example, an hDp3 strain segregated hDp23 which subsequently segregated even shorter duplication chromosomes. The relationship between the spontaneous duplications and their progenitors are shown in Figure 23. The first step in the analysis of the spontaneous duplications was to genetically analyze their structure. This was done by mapping their breakpoints with respect to mutations in the region. Extensive complementation experiments were carried out in the unc-11 dpy-5 region (Figure 24) because most of the new breakpoints were in this interval. R. Rosenbluth (personal communication) provided valuable contributions to the interpretation of this data. Two considerations in this analysis were: 1) did the breakpoints show any clustering and 2) did the complementation patterns for each duplication correspond to a linear map. Breakage of a ring chromosome, for example, could have appeared in the mapping experiments as a deletion of material from both "ends" of a linear chromosome. In the analysis of the gamma radiation induced duplications (section 3.2), the data were consistent with a rod structure of sDp2 and ILXLszTl. 167 SDOZ gamma hDp3 gamma hDps gamma hDp7 hQp74 gamma nDp23 hDp77 hDp93 hDp9i h0p9S nDp28 hDp29 hDp22 n0p24 hDpWO hDp26 UDP27 hDp30 nDpS9 S0p2 gamma hDp75 hDp84 n0p2 gamnu nDp25 KDpBS nOpBS n0p87 nDp88 hDp89 hDp90 \t\Dp9l\ \hDp92\ \hDp76\ \nOp78\ \nOp7 f>Dp20 hDp6 gamma gamma gamma t\ T6 r>0078 nOp 79 Figure 23: Pedigree of spontaneous duplications. All the spontaneous shortened duplications their progenitors are shown. If gamma radiation was used, this is indicated, otherwise the duplications were spontaneous. 168 The mapping of the duplications provided no evidence for clustering of the spontaneous breaks. In fact, the spontaneous duplications have provided at least five new breaks in the unc-74 dpy-5 region. They have been useful for the ordering of lethal mutations in the region. In addition, spontaneous duplications derived from deletion of sDp2 in the region from unc-40 to the left end of the chromosome were isolated {e.g. hDpl 00). Duplications of this type were not recovered in the gamma radiation experiments. Complementation testing of him-1 to the spontaneous duplications using both the viable allele, e879, and the lethal allele, hl34, resulted in some inconsistencies. Hodgkin, Horvitz and Brenner (1979) mapped him-l(e879) to the interval between unc-38 and dpy-5. Duplication mapping with most duplications, and in particular hDp76 and hDp78, was agreed with this result. hDp76 was the duplication with the right-most breakpoint that still rescued him-1. Given these data, all duplications with breakpoints to the left of hDp76 should have rescued him-1. This was not the case. Some of the unc-63(+ ) spontaneous duplications, (e.g. hDp23 and hDp79) and hDp21, which was probably spontaneous, failed to rescue the lethal phenotype of him-1 (hi34). In addition, hDp23 failed to rescue the Him phenotype of the viable allele e879. e879/e879; hDp20 hermaphrodites produced 4.3% male progeny while e879/e879; hDp23 hermaphrodites produced 22% male progeny. The male frequency of 0.043 from e879le879; hDp20 was higher than normal (0.1%; Rose and Baillie 1979). Howell (1989) has shown that e879 is semi-dominant when in a 2:1 ratio to him-1 (+). To explain these results, I propose that in the formation of hDp21, hDp23, hDp29 and hDp79, a second him-1 mutation was induced on the duplication chromosome. The total number of spontaneous duplications (including hDp21) which covered a larger region than hDp76, the smallest duplication to complement him-1, was eight. Of these eight, four (the three above plus hDp29) were associated with a him-1 mutation. All of the progenitor duplications were tested and found to be him-1 (+). This proposal requires a very high mutation rate at the him-1 locus when a duplication spontaneously shortens. 169 unc-38 let-363 unc-11 unc-13 unc-14 unc-57 unc-63 let-359 dpy-5 unc-40 bli-4 dpy-14 let-86 J I hDp2,3,1,5 hDp92*,100* hDp!8 hDp23* ,74* ,79* hDp85*, 86*, 88*, 90*, 91 *, 93*, 94*, 95* hDp81*, 89* hDp59 hDp30 hDp83 let-361 let-363 unc-38 let-364 let-311 him-1 unc-13 unc-14 unc-51 let-351 let-315 unc-63 let-359 spe-11 let-352 dpy-5 I I I I I I I I I I I I hDp4,25*,28* hDp!8 hDp20,23*, 14* hDp29*,59* hDp30 hDp26 hDp21 hDp!6* hDp24 hDp22*, 15* Figure 24: Structure of spontaneous duplications. Spontaneous duplications are indicated by an asterisk, a) General map of duplications. hDp92 has not been tested with let-86. b) More detailed map of unc-74 dpy-5 region. 170 The data could also be explained if him-1 was between unc-38 and unc-63 (Figure 24) and an internal deletion derivative of hDp20 formed hDp76. This internal deletion would be to the right of him-1 (+) and include the unc-63 - let-361 region. Two observations, however, argue against this proposal. First, hDp76 no longer carried the chromosome V material that was present on its progenitor, hDp20 (section 3.2). This result supports a terminal deletion mechanism for the formation of hDp76. Second, a spontaneous variant of hDp79 {hDp79 is linked to chromosome II, see below) was isolated which could rescue him-l(h!34). The isolation of a revertant indicated hDp79 had the ability to acquire a him-1 (+) allele. This would be most likely if the original hDp79 had a him-l(-) allele which could revert to wild-type. In fact, the ability to rescue him-1 (hi34) may have been a consequence of hDp79 becoming homozygous. The fact that in a cross of the original hDp79; him-l(hl34) dpy-5 unc-13 strain to dpy-5;0/szTl(I;X) males produced no Dpy-5 progeny indicated the duplication was homozygous. Furthermore, when hDp79; dpy-5 dpy-141 him-l(hl34) dpy-5 unc-13 was constructed from the original rescued strain, the frequency of him-1; hDp79 rescued hermaphrodites was rare. This was consistent with a requirement for hDp79 homozygosity since the double homozygote him-1; hDp79 would have constituted only 1/16 of the total progeny. The requirement for two copies of the him-1 allele on hDp79 to achieve rescue suggests this mutation was a hypomorph. Overall, these two observations support the hypotheses that him-1 is between let-361 and dpy-5 and that some spontaneous duplications which cover this region have a second site mutation at the him-1 locus. Figure 24 shows that some of the duplications had complex structures involving multiple breaks. The assigned positions of let-359 and spe-11 were based in part on these complex duplications and allowed for the least number of breakpoints. Other maps could be drawn for the data, however, so the map positions of let-359 and spe-11 should be considered tentative until confirmed by three-factor experiments or complementation to other rearrangements induced with radiation. If the complex duplications were not considered, let-359 would be placed with the unc-63 group and spe-11 would be placed with the let-361 group. 171 The complex duplications came from only three progenitors. The h duplications 26, 27, 30 and 59 were spontaneous unc-63(-) derivatives of hDp23. hDp78 was a spontaneous derivative of hDp4. A dpy-5(-) derivative of hDp74, hDpl00, was isolated by Shiv Prasad (personal communication). Two of the three progenitors, hDp23 and hDp74, were also of spontaneous origin. The remainder of the spontaneous deletion events studied here involved duplications induced with gamma radiation. Of these, only hDp4 produced a duplication with a complex structure. The complex duplications were characterized by double and sometimes triple deletion events relative to the genetic map. The deletions often involved both ends of the progenitor duplications. For example, hDp30 was produced by a deletion of hDp23 involving both the dpy-14, unc-63 and let-363 regions. A structure has been proposed for hDp4 based on the structure of its derivative, hDp78 (Figure 25). The proposed structure for hDp4 involves a ring chromosome with a superimposed inversion relative to the linear map. A similar structure for hDp23, and possibly hDp74, was derived based on the structure of their derivatives. If hDp23, like the proposal for hDp4, was a ring chromosome with a superimposed inversion, than the multiple deletions with respect to the linear map could be explained by a single deletion on the ring chromosome-inversion map. Figure 25 shows the proposed structure for hDp23 and the deletions which produced the four derivatives. hDp74 may also have been a ring chromosome because material from both the dpy-14 and dpy-5 regions was lost in forming hDplOO. The stabilities of some of the spontaneous duplications in the hermaphrodite germ line were determined (Table 32). The derivatives from a given duplication had a variety of stabilities. One derivative of hDp2, hDp88, was extremely unstable. In contrast, a second derivative, hDp91, was possibly more stable than the parental hDp2. To determine the frequency at which duplications shortened, a selective system was used to screen a large number of worms for exceptional individuals. Using the procedure described in section 2.3, the progeny of unc-74 dpy-5; hDpx or unc-63 dpy-5; hDpx hermaphrodites were screened for levamisole resistant Unc progeny. The recovery of these duplications is summarized 172 Figure 25: a) Proposed ring chromosome structure for hDp4 and hDp23. The inversion breakpoints are indicated by parentheses. The deletion that occurred in hDp4 to form hDp78 is shown by brackets, b) Normal gene order in this region, c) The proposed deletion breakpoints (shown in brackets) which occurred to form the hDp23 derivatives. A ) dpy-5 dpy-5 let-351 unc-63 spe-11 let-359 spe dpy-14 let-363 hDp23 dpy-14 unc-57 let-352 unc-63 B ) unc-63 spe-11 let-352 unc-57 let-351 let-359 let-363 dpy-5 4-f-dpy-14 C ) dpy-5 let-359 spe let-363 .„ hDp26 dpy-14 let-352 dpy-5 let-359 spe-11 dpy-14 let-352 unc-63 dpy-5 let-359 spe-11 dpy-14 let-352 unc-63 dpy-5 let-359 spe let-363 hDp59 dpy-14 let-352 unc-63 173 Table 32 Stability of spontaneous duplications: Hermaphrodite gametes. Duplication W t a Dpy or Frequency of U n c c Dp gametes dpy-5; hDp22 302 520 0.22 dpy-5; hDp23 451 415 0.35 dpy-5; hDp24 156 260 0.23 dpy-5; hDp25 368 470 0.28 dpy-5; hDp29 306 335 0.31 dpy-5; hDp30 104 189 0.22 dpy-5; hDp76 509 361 0.41 dpy-5; hDp78 462 (3) 221 0.51 dpy-5; hDp79 726 (6) 255 (1) 0.49 dpy-5; hDp83 496 (5) 164 0.50 dpy-5; hDp88 50 537 0.04 dpy-5; hDp90 332 445 0.27 dpy-5; hDp91 246 231 0.35 dpy-5; hDp95 111 231 0.19 unc-74; KDp92 150 1974 0.04 ° Male progeny scored are in parentheses. With the exception of hDp79 and hDp83, the duplication homozygotes were not scored. 174 in Table 33. Because the worms were tested in the second generation, it could not be ensured that each event was independent. Some Po hermaphrodites segregated more than one Unc F2. It was possible that some of the shortening events occurred premeiotically. For this reason the numbers in Table 33 may not reflect the per chromosome shortening frequency. Different chromosomes shortened at different rates and there was a correlation between spontaneous shortening rate and mitotic stability. The more stable duplications shortened at a lower rate. For example, sDp2, which is mitotically very stable, spontaneously shortened at a low frequency of 1.6X10-^. In a levamisole selection experiment, unc-74 dpy-5; sDp2 hermaphrodites segregated two Dp(unc-74(-)) derivatives among 127,000 chromosomes. hDp2, which was mitotically unstable as judged by the production of mosaics and reduced recovery in the gametes, shortened at a much higher frequency (4.2X10"4). Two experiments were done without levamisole selection. The progeny of either unc-74 dpy-5; hDp2 or unc-74 dpy-5; hDp3 worms were screened. Because there was no levamisole, I could screen for both Unc and Dpy progeny. In addition, because I was screening in the FI generation, there was no uncertainty with independent events. If shortening events occurred premeiotically, I would have expected to recover some clusters of exceptional progeny. In the hDp2 screen, 3600 worms were screened. Seven true breeding Unc hermaphrodites and one true breeding Dpy were obtained. One additional Unc strain was recovered but it was exceptionally unstable and could not be maintained. All nine worms were found as single exceptions among the progeny of a parent. Such data would be expected if shortening events were restricted to meiosis. Two of the seven Unc strains carried new duplications, hDp87 and hDp89, which had two deletions with respect to hDp2. One deletion was in the unc-74 and the other was in the dpy-14 regions. As described above for hDp4 and hDp23, these double deletions suggest hDp2 was a ring chromosome. In the hDp3 screen, 2800 progeny were screened and 4 true breeding Unc hermaphrodites were recovered. In this case two of these worms came from the same Po. If this does represent a Table 33 Recovery of spontaneously shortened duplication strains Duplication Total Shortened Frequency Mosaics Progeny Duplications unc-74 dpy-5;hDp2 unc-74 dpy-5;hDp5 unc-63 dpy-5;hDp20 unc-63 dpy-5;hDp23 unc-74 dpy-5;sDp2 72 000 30 116 000 3 134 000 1 41 000 18 127 000 2 4.2X10' 4 54 2.6X10"5 3 7.4X10"6 1 2.2X10"4 18 1.6X10' 5 2 176 single premeiotic event, and not two meiotic events, then recoverable shortening events could occur in mitotic cells, but at a lower frequency than in meiotic cells. A minority of the spontaneous shortening events were associated with fusions to another chromosome. hDp78, hDp79 and hDp83 were recovered from hDp4, hDp6 and hDp23 strains respectively. For each duplication the attachment to another chromosome resulted in a dramatic increase in mitotic stability (Table 32). This was the basis on which they were identified. In addition, stable homozygous strains of each duplication could be isolated and maintained. Free duplication strains were not stable as homozygotes (see section 3.9). hDp79 and hDp83 homozygotes were similar in phenotype to their hemizygous siblings while the hDp78 homozygotes were noticeably smaller, thinner and slower growing. hDp78 was shown to be linked to unc-54 on chromosome I. Among the progeny of dpy-5 unc-54 -hi dpy-5 + hDp78, recombinant Dpy-5 progeny were rare (Table 17). I have not yet detected strong linkage between hDp79 and another marker. hDp83 was closely linked to dpy-9 on the left end of chromosome IV. No recombinant progeny were observed in 1119 wild types from dpy-5; + dpy-9/ hDp79 + hermaphrodites. The three fusions were associated with the loss of material from the original duplications (Figure 24). As just described, hDp 78 was a deletion with respect to hDp4. hDp 79, a derivative of hDp6, lost material in the unc-38 - let-3 75 region. hDp83 had the same left end breakpoint as its parent, hDp26, but was deleted at the other end (the let-86 - unc-14 region; Figure 24 see also section 3.4.2). As all three duplications were detected due to increased stability, the loss of material was not selected for and may have been an integral part of the fusion event. Spontaneous breakdown of hDpl4. Another example of spontaneous chromosome breakage has been observed with the hDp 14 chromosome. hDpl4 is a large part of chromosome I inserted into the X-chromosome between unc-20 and dpy-8. In an attempt to determine the amount of recombination between the insertion site and unc-1, I constructed and scored the progeny of dpy-5; unc-11 hDpl 4 hermaphrodites (section 3.7; Table 16). Unc-1 recombinants from this experiment should have been Him because the recombination event would link the insertion to unc-1. Surprisingly, 14/56 Unc-1 177 "recombinants" were not Him. Two of these Unc-1 hermaphrodites were analyzed further. From this analysis it appeared the exceptional Unc-1 hermaphrodites carried, in addition to two unc-1 marked X-chromosomes, a derivative of the original hDpl 4 chromosome in which there was a deletion in the unc-1 region. These worms had the genotype dpy-5; unc-1 lunc-1 ;hDp[dpy-5(+); unc-1 (-)]. The duplications in the two Unc-1 strains were called hDp81 and hDp82. The conclusion that spontaneous breakage of the hDpl4 chromosome created hDp81 and hDp82 was established from complementation testing of the duplications to chromosome I and X markers (Figure 26). In the formation of hDp81 and hDp82, no material was lost from the chromosome I portion of hDpl4 (Figures 7, 9 and 26). Both duplications carried wild-type alleles of bli-3, unc-11 and h564 but not of dpy-14. Both dpy-5 dpy-14; hDp81 and dpy-5 dpy-14; hDp81 hermaphrodites were Dpy-14 but usually sterile. Rare fertile Dpy-14 hermaphrodites had very low fecundity and could not be maintained. In addition to unc-1, the duplications did not cover unc-20. These two markers are to the left of the hDpl4 insertion site. hDp81 and hDp82 failed to completely rescue lon-2. The data were ambiguous, however, because the lon-2; Dp worms were longer than wild-type, but did not appear as long as euploid Lon-2 worms. Both duplications did complement dpy-7 and unc-3, markers to the right of the insertion site. These results showed the left breakpoints of hDp81 and hDp82 were in a similar location as the hDp 14 insertion site (Figure 26). For this reason, and because the chromosome I sequences were unaffected in these events, the event in the dpy-5; unc-llhDpl4 hermaphrodites to create hDp81 and hDp82 may have involved an almost precise break at the point of insertion. The result was loss of material (unc-1(+)) to the left of this point. The characteristic loss of the Him phenotype probably resulted from losing the information for meiotic pairing and recombination located in the unc-1 region (section 3.6, 3.10 and Discussion). The two duplications were not identical. hDp82 strains were usually healthier than hDp81 strains. In Dp; dpy-5 hermaphrodites hDp82 was recovered in more progeny than hDp81. From a hDp81; dpy-5 strain I scored 204 wild-types and 369 Dpy-5 (wild-type/Dpy = 0.55) and from a hDp82; dpy-5 strains I scored 255 wild-types and 278 Dpy-5 (wild-type/Dpy = 0.92). 178 bli-3 unc-11 dpy-5 unc-40 hDpU bli-3 unc-11 dpy-5 unc-40 hDp81 hDp82 dpy-8 dpy-7 unc-3 bli-3 unc-11 dpy-5 unc-40 hDp80 dpy-8 dpy-7 Figure 26: Diagram of spontaneous breakdown products oihDpl4(I;X). The duplications were not tested for rescue of dpy-8. 179 While hDp82 may be mitotically more stable than hDp81, viability differences may also be a significant factor. It appears that both duplications were not linked to another chromosome. Two additional spontaneous changes to the hDpl4 chromosome occurred during routine stock maintenance. In the first case, a dpy-5; HDpl4 stock was observed to have lost its Him phenotype. The duplication in this strain, called hDp80 (Figure 26), was analyzed by genetic mapping of its breakpoints and it was found to have the same structure as hDpl 4 with respect to chromosome I material. hDp80 rescued let-362, unc-40 and h564 and dpy-5. Like the previous two duplications, however, hDp80 had deletions of the X-chromosome. On the X-chromosome, hDp80 failed to rescue unc-1, unc-20 and lon-2 but did rescue dpy-7. These results were similar to those with hDp81 and hDp82. Unlike these two duplications, however, hDp80 did not rescue unc-3. Finally, it was noticed that a dpy-5 dpy-14; hDpl4 strain had lost its Him phenotype. In this case, however, there was no evidence for a free duplication. Instead, there had been a modification of the normal dpy-5 dpy-14 chromosome I in this strain. The abnormal strain was crossed to N2 males and twenty F I wild types were set individually on plates. Eleven F I worms segregated only Dpy-5 Dpy-14 while the other nine segregated only Dpy-14 worms. Nine of these Dpy-14 hermaphrodites were placed onto new plates but eight were sterile. Sterile Dpy-14 worms were also present in the original stock. Ten Dpy-14 worms were set from this stock and six were sterile or very sick. The other four were fertile and segregated Dpy-5 Dpy-14 progeny. Thus there appears to be a new chromosome I in this strain with a dpy-5( +) dpy-14(-) genotype. There is an adult sterile mutation on this chromosome and it has been designated hi 013. The genotype of the original Dpy-14 strain was dpy-5(+) hl013 dpy-14(-)/dpy-5 dpy-14. This explains the loss of the Him phenotype since the original hDpl 4 X-chromosome was no longer present. dpy-5 + +1 + hi 013 dpy-14 hermaphrodites were constructed and their progeny scored. The ratio of wild-type to Dpy-5 (1.8:1; 303 wild-types, 166 Dpy-5 and 106 Dpy-14) is consistent with hi 013 being a simple intrachromosomal mutation. The hi 013 mutation must be closely 180 linked to dpy-14 because the sterile mutation was retained in the stock for many generations without selection. The hl013 dpy-14 chromosome was derived from a dpy-5 dpy-14 chromosome. Either the new chromosome retained the original dpy-5 allele and carried a duplication which included dpy-5(+), or it had only the dpy-5(+) allele. The progeny of + +1 hl013 dpy-14 hermaphrodites were scored but no Dpy-5 progeny were observed in 1589 wild types. Either the original dpy-5 mutation was no longer present or a dpy-5(+) duplication was present and tightly linked to dpy-14. To test if this chromosome had unusual meiotic properties, I scored hi 013 dpy-14/let-362 dpy-5 unc-13 hermaphrodites. The results, 482 wild-types, 60 Dpy-5 Unc-13 and 6 Unc-13, gave a p of 0.12. This was slightly but significantly lower than normal (Table 11). I conclude hl013 was derived from a dpy-5 dpy-14 chromosome. The hl013 chromosome retained its original dpy-14 allele, possibly the original dpy-5 allele, and gained a dpy-5(+) allele. Although it has not been proven, it is possible that duplication material, and possibly associated X-chromosome material, was inserted into chromosome I closely linked to dpy-14. To detect the presence of X-chromosome material on the hi 013 chromosome, dpy-5lhl013 males were crossed to hermaphrodites homozygous for an X-linked marker. Wild-type male progeny would have been observed if the hi 013 chromosome carried an X-linked marker. In crosses to a variety of X-linked markers, no wild-type males were observed. Therefore there was none or little X-chromosome material associated with the hi 013 chromosome. 3.13.3. Genetic mosaics. Genetically mosaic worms result from the loss of the duplication at one of the embryonic cell divisions, producing a worm in which some cells have the duplication and some do not (Herman 1984). The type of genetic mosaics which are recovered can be used to determine where in the cell lineage the expression of a gene is required. For example, in experiments with unc-74 dpy-5; hDp2 and unc-74 dpy-5; hDp3 hermaphrodites in which no levamisole selection was used, many mosaic worms were recovered (Table 34). These worms had a mutant phenotype but segregated only Dpy Unc progeny or wild types and Dpy Unc progeny. The types of progeny from 181 a mosaic worm can be used to determine the approximate cell division at which the duplication was lost. The unc-74 gene probably codes for a subunit of the acetylcholine receptor located in muscle cells (Lewis et al. 1980a, 1980b). Thus mosaic worms with an Unc phenotype probably lost the duplication in one of the P cell divisions which produce most of the muscle cells (Figure 27). This prediction was confirmed by the data. Most of the Unc mosaics segregated only DpyUnc progeny. Since the duplication was lost early in the P lineage, it was absent from the germ line. The semi-Unc worms were more mobile than the Unc worms but had noticeable movement impairment compared to wild-type. Some of the semi-Unc worms segregated wild-type progeny. Since the semi-Unc worms probably had unc-74(+) present in only a subset of their muscle cells, the duplication was probably lost later in development and could still be retained in the germ line. Many Dpy non-Unc mosaic worms were recovered. Some of these were not fully Dpy but were intermediate between wild-type and Dpy-5. To recover a Dpy mosaic worm, the duplication must have been absent in the cells requiring dpy-5(+) product for proper morphology but present in the muscle cells. While 14 Dpy-5 mosaics segregated wild-type progeny, indicating the duplication was present in the P lineage, six others segregated no wild-type progeny. Either the duplication was lost twice in these six worms, once in the AB lineage and again in the P lineage, or Dpy-5 mosaics were produced from duplication loss in the P lineage. The former was not likely because the second loss event was more frequent (6/20) than expected. All semi-Dpy mosaics produced wild-type progeny. Either the duplication was lost in one of the EMS divisions, or in an AB division. Semi-DpyUnc progeny were intermediate in length between wild-type and Dpy but were also Unc. This was an interesting class because the duplication had been lost from a lineage which utilizes both gene products. An intriguing recovery was mosaics that were Dpy, but only in the anterior half of their bodies. In the hDp2 experiment 2 of these were recovered* one had wild-type and DpyUnc progeny in the F2 and the other had only DpyUnc progeny. In the hDp3 experiment two more of Table 34 Recovery of genetic mosaics from d u p l i c a t i o n s t r a i n s 3 . Duplication Phenotypes Total FI F2 Unc Unc DU Wt + DU W-Unc W-Unc DU Wt + DU W-DU W-DU DU Wt + DU W-Dpy Wt + DU Dpy Dpy Wt + DU DU hDp2 hDp3 21 1 13 2 6 6 17 0 11 0 19 6 3586 2784 a Hermaphrodite parents were unc-74 dpy-5; hDp. b Abbreviations: Wt=wild-type, DU=DpyUnc, W-Unc=weak Unc, W-DU=weak Dpy but also Unc, W-Dpy=weak Dpy. 183 Figure 27: Diagram summarizing the founder cells and the cell type they form during C. elegans development (from Sulston et al. 1983). 184 these worms were recovered and both produced wild-type and DpyUnc progeny. This observation was surprising because many dpy genes, dpy-5 included, were believed to be required for proper cuticle formation. The cuticle is secreted from the hypodermal cells, many of which are multinucleate. The largest of these, hyp-7, extends most of the length of the body and is formed from the fusion of 110 cells (White 1988). The observation that a worm can have localized Dpyness might imply the contents of the syncytium are not free to diffuse. Alternatively, dpy-5 expression may be required in a tissue type distinct from hyp-7. 185 4. D i s c u s s i o n The original goals of this thesis were to isolate rearrangement breakpoints on chromosome I and use these to map mutations and study chromosome behaviour. To provide the rearrangement breakpoints for mapping mutations, I used gamma radiation to shorten pre-existing duplications. The array of breakpoints thus created were used to determine the order of mutations along the chromosome. A n a r r a y o f d u p l i c a t i o n s fac i l i t a t es the a n a l y s i s o f a l a r g e c h r o m o s o m a l r e g i o n : In the process of identifying essential genes in the sDp2 region (Howell et al. 1987), over 500 EMS induced mutations have been isolated and are being mapped by recombination and complementation testing (Howell 1989; McDowall 1990; A. Rose, personal communication). Dividing the mutations into complementation groups would be an enormous task without rearrangement breakpoints to position the mutations into smaller intervals. The sixty-two duplications described in this thesis and twelve deficiencies from various sources have been used to divide the left 20 m.u. of chromosome I into at least 24 regions. This number will increase as the rearrangements are complementation tested with more mutations. While the spontaneous shortening of duplications could produce false negative results, this problem is avoided through frequent testing of strains with appropriate markers. The techniques developed in this thesis could be applied to the genetic mapping and analysis of essential genes in any region of the C. elegans genome. A large number of breakpoints can be easily generated in any region having an appropriate pre-existing duplication. As a tool for the genetic dissection of a region, duplication mapping may be an attractive alternative to the conventional deficiency mapping (Sigurdson, Spanier and Herman 1984; Rosenbluth et al. 1988). The isolation of duplication breakpoints is fast and easy. Furthermore, unlike deficiencies whose viability decreases with increasing size, larger duplications are more stable with increasing size (see section 3.13). Having such large aberrations is a useful tool to quickly localize unmapped mutations to a large region; that being either the region covered or not covered by the duplication. Subsequent experiments can map the mutation more precisely using additional breakpoints. Since 186 deficiencies are generally smaller, more experiments are required to map a mutation to a region of comparable size. In addition, most of the duplications used in this study have one end in common with sDp2. Because only one breakpoint varies, mapping is a linear procedure. In theory the same could be said of nested deficiencies, but such a collection is not easily isolated and can only be effective for a small region due to the smaller size of viable deficiencies. The nested duplications are a by-product of the method used to generate them and there is no restriction on the size of the region to be analyzed. The duplication breakpoints can be used as genetic landmarks that can be placed on the physical map of the C. elegans genome (Coulson et al. 1986, 1988). Within the sDp2 region approximately 2000 kB of DNA has been placed on the physical map. Work in this laboratory has accomplished alignment of the DNA map to the genetic map using DNA strain polymorphisms (Starr et al. 1989; Starr 1990; J . Babity and K. Peters, personal communication). Mapping the duplication breakpoints with respect to polymorphic sites limits the region of the physical map where the duplication breakpoint could be located. For example, the hDp62 and hDp83 breakpoints have been located between the polymorphisms sPl and hP9 (Starr 1990, K. McKim and T. Starr unpublished results). In addition, the hDp83 breakpoint is to the left of the right end of AD/8. The AD/8 breakpoint was found on the physical map by Starr (1990). This information limits the position of the hDp83 breakpoint to two or three cosmids. Using the duplications in a small region. I have used these techniques in the analysis of essential genes in the dpy-14 unc-13 region of chromosome I. This region had two attractive features. The first was that T. Starr has done extensive molecular characterization in the region (Starr et al 1989, Starr 1990). He mapped four DNA polymorphisms in the region. Close to dpy-14, Starr (1990) restriction mapped cosmids and characterized coding regions in 173 kb surrounding the sPl polymorphism (Rose et al. 1982). The second feature was that many of the lethal mutations from the sDp2 lethal set had already been mapped to this region by McDowall (1990) and A. M. Rose (unpublished results). Ninety mutations from the original 550 lethal mutations isolated with sDp2 were mapped to this region by myself, A. M. Rose and J . McDowall. 187 These mutations, in combination with pre-existing mutations in the region, have defined 29 genes. These genes have been mapped to four distinct regions based on mapping with duplications and two deletions. One of these regions, between hDpl9 and hDp62, contained notably more genes than the other three regions. I mapped 18 genes to this region but only four, four and two to the other three regions. The lack of breakpoints in this region may result from a peculiar chromosome structure. It is also possible that in comparison to the other regions, the hDpl9 hDp62 region contains many genes, and a correspondingly large number of mutable loci, but little DNA. If genetics is to be utilized in the molecular analysis of this region, some breakpoints in the region must be generated. A second set of lethal mutations in this region was generated in a lethal screen using the translocation hTl(I;V) as the balancing system (Howell and Rose 1990). This screen was carried out to test whether duplication screens were recovering all possible recessive lethal mutations. Rosenbluth et al. (1988) in their analysis of essential genes on chromosome V, reported the duplication sDp30(V;X) failed to complement some lethal mutations which mapped in the region it duplicated. Where let/+ was viable, let/let/sDp30 was not. One possible explanation of these results is that there is a class of recessive lethal mutations which are dominant when present in two doses. The analysis of the hTl lethal mutations indicated there is not a large class of dosage dependent, conditionally lethal mutations in the left end of chromosome I. Of the 44 lethal mutations from the hTl screen that were balanced by szTl, only two showed any evidence of dominant lethality when present in two doses. Both hi 003 and h940 caused some dominant lethality when present in two doses. This lethality was not, however, fully penetrant. Some of the duplication hermaphrodites with two doses of either of these two mutations inevitably survived. Despite the dominant lethality, both of these mutations were found to be alleles of previously identified genes. These data show that genes which cannot be identified in duplication lethal screen are not a large class. There are some genes, however, which may be poorly 1S8 represented in the set of mutations because some of their alleles are variably penetrant dominant lethals when present in two doses of a duplication strain. As opposed to a class of mutation which is a conditional dominant lethal, another interpretation is that the duplications themselves vary with respect to their ability to rescue lethal mutations. As an example, Rosenbluth et al. (1988) proposed the dominant lethality in sDp30 strains may have resulted from the duplication being X-linked and dosage compensated. I have evidence that properties of each duplication may influence the penetrance of dominant lethality. In contrast to the result described above, h940 showed no dominant lethality in nDp4 strains and hi 003 had no dominant lethality in hDp58 strains. In addition, hDp50 and hDp65 failed to complement let-88(s!32); despite the fact both duplicate the let-88 region and other duplications of the region rescued si 32 homozygotes. Most of the lethal mutations in the hDp62 - sDp2 region exhibited dominant variably penetrant lethality in hDp65 strains. The behaviour of hDp65 and hDp50 could be related to their structure; the chromosome I material is attached to a fragment from the X-chromosome. Another example was the failure of hDp4 to rescue let-351(h43). The fact that hDp4 was unusually unstable during mitosis (section 3.13.1) may explain why h43lh43;hDp4 worms were sterile adults. The quantity of let-351 gene product may be low in h43; hDp4 hermaphrodites because the only source of let-351 (+) product, the duplication, is frequently lost. If the hermaphrodite is extremely sensitive to let-351 dosage, the lower level in h43; hDp4 worms may cause sterility. In general, structural or other characteristics appear to influence the level of gene expression from a duplication and thus affect its ability to rescue certain lethal mutations. E s t i m a t e s o f gene n u m b e r s a n d m u t a t i o n i n d u c t i o n ra tes . Two other regions of chromosome I have benefited from intense analysis of essential gene organization. Howell and Rose (1990) defined 19 genes with 54 mutations mapping in the hDfS region. McDowall (1990) defined 16 genes with 44 mutations in the hDp 16 - hDpl9 region; immediately left of dpy-14. The chromosome rearrangements isolated and initially characterized in this thesis formed the base line 189 for the mapping in these two studies. Similar studies of essential genes in the unc-22 region of chromosome IV (D. V. Clark 1990) and in a region covering about 15% of the X-chromosome (Meneely and Herman 1979, 1981) are also relevant to this discussion. These authors attempted to estimate total gene numbers and forward mutation rates from allele frequency data. Using Poisson statistics with the number of alleles for each gene, one can estimate the number of genes in the "zero" class; the number of essential genes with no lethal alleles in this set of mutations. As noted by the authors of the four previous studies, the Poisson analysis carries two assumptions: 1) The probability of mutating all genes are equal, and 2) the ability to detect mutants is the same for all genes. Despite violations to both of these assumptions, the Poisson analysis is informative. The failure to adhere to the assumptions results in an underestimate of the zero class and therefore the estimate of gene numbers is a minimum one. Also, the deviations from the assumptions reveal interesting characteristics and variability amongst the sample of genes. As an alternative to the Poisson analysis, Lefevre and Watkins (1986) have proposed the negative binomial is a more accurate estimator of the zero class. Figure 28 shows the allele distribution in the dpy-14 region. It is obvious that genes are not equally mutable, let-385 has twice the number of alleles as the next most mutated gene. All previous studies of essential genes have observed inequality in gene mutability. The results of Howell and Rose (1990) are shown in Figure 29a. One gene in her set (let-354) was similar in mutability to let-385. The rest of her genes were close to that predicted from the Poisson distribution. To reduce the overemphasis the Poisson analysis would place on the highly mutable genes, Meneely and Herman (1979) used the formula f = (l-e"m-me'm)/(l-e'm), where f = the fraction of mutated genes represented by more than one allele and m = the average number of alleles. This formula underestimates the number of genes with no alleles if there are many genes with lower than average mutability. For the calculation in the dpy-14 region, /"was calculated from either the sDp2 set of lethal mutations or from the combining both the hTl and sDp2 sets of lethal mutations. With just the 190 a) 12 -f 0 1 2 3 4 5 6 7 -8 9 10 11 12 13 14 15 16 17 18 19 20 Mutations per gene b ) 12-1 10 -# g e n e s 6 7 8 9 10 11 Mutations per gene i i i i i i i i i 12 13 14 15 16 17 18 19 20 observed expected modified expected Figure 28: Histogram of mutations in essential genes of dpy-14 region. The observed distribution for the a) sDp2 lethal mutations and for b) the sDp2 and hTl lethal mutations combined are shown. Also shown is the distribution predicted by the Poisson analysis. Also plotted in b) are the expected values calculated when the four largest mutational targets are excluded from the data set ("modified expected"). 191 sDp2 lethal set, f = 15/25 = 0.6 and m= 1.6. With both sets oflethal mutations, f = 16/25 = 0.64 and m = 1.7. These numbers are similar to the value of m calculated by Howell (1989; 1.5) and McDowall (1990; 1.65). From this calculation, I estimate at least four or five essential genes in the dpy-14 region have not been identified by the existing mutations. Added to the 25 identified essential genes, and the three genes represented only by viable alleles, the minimum estimate of the total number of genes in the dpy-14 region is 33. This a minimum estimate because of the limitations of the Poisson analysis and because genes with maternal effect lethal phenotypes and genes with subtle mutant phenotypes are not included. Based oh mutability, there may be four classes of genes in C. elegans. In Figures 28 and 29, there are genes with one to five alleles, genes with seven to eleven alleles, and one gene with 19 alleles. Interestingly, the genes in the 1-5 allele class follow closely the expectations of the Poisson distribution. The fourth is the class of genes with exceptionally low mutability that are under-represented in any mutation hunt. Howell (1989) in her analysis of hDf6 and Meneely and Herman (1981) in their analysis on the X-chromosome, found two of these classes, the first class and the highly mutable class (Figure 29a and 29b). D. V. Clark's (1990) data for the unc-22 region also has two classes of gene, the first and middle classes (Figure 29c). After consideration for different EMS dosages, the studies of D. V. Clark (1990) and Meneely and Herman (1981) tested two-thirds the number of chromosomes for lethal mutations as the studies on chromosome I. This difference does not alter these conclusions. If there are different classes of gene mutability, then the large mutation targets should be excluded from the Poisson analysis because the region has been saturated for the large targets. Repeating the Poisson calculation without the data from the four heaviest hit genes, let-385, let-389, let-392, and let-394, gives a similar estimate for the total number of genes in the region. Using the data from both the sDp2 and hTl lethal sets, f = 12/21 = 0.57 and m = 1.5. The number of genes estimated to be in the zero class is five. When the total number of genes in a region is known, extrapolations can be made to estimate the total number of essential genes in the genome. These estimates have ranged 192 a) I M J L 2 3 4 5 8 7 S 9 1 0 1 1 1 2 1 3 1 4 Mutations/Gene 0 1 2 3 < 5 e 7 8 8 K) 11 12 13 H IS 16 Number of alleles c) 1 1 i 1 1 M W . . . . 0 1 2 3 4 6 8 7 8 9 10 Number of alleles M Oburvad f5S3 ExptcMd Figure 29: Histogram of mutations of essential genes for other regions of the genome. The number of mutations per essential gene are shown for a) the hDf6 region of chromosome I (Howell 1989) b) the mnDpl region of the X-chromosome (Meneely and Herman 1981) and c) the unc-22 region (Clark 1990) for comparison with Figure 28. 193 from 3500 (D. V. Clark 1990) to 3300 (Howell and Rose 1990) to 4500 genes (McDowall 1990). These estimates are underestimates; the magnitude of the underestimate depends on the difference between the size of the zero class of genes as estimated from Poisson analysis and in reality. Another approach to estimating total gene numbers has been to use molecular methods to determine the number of coding regions in a given length of DNA. These studies have come up with similar results; one gene for every 15-20 kB of genomic DNA (Heine and Blumenthal 1986; Prasad and Baillie 1989; Starr 1990). With 85 Mb of single copy DNA in the C. elegans genome [80% of C. elegans DNA is single copy (Sulston and Brenner 1974) and the genome size is 100Mb (A. Coulson and J . Sulston, personal communication)], this corresponds to 4000 to 5300 genes. The estimates from both genetic and molecular data are in very close agreement, indicating that the zero class was not grossly underestimated by the Poisson statistics. The total gene estimates as approximated by essential gene analysis appear to be accurate to a factor of 1.5. The corollary is that most of the genes in C. elegans are mutable to a lethal phenotype. Genetic "saturation" of a chromosome or a genome. The analysis of three regions of chromosome I are concordant in the prediction that 80% of the essential genes in the sDp2 region of chromosome I have been identified. In order to increase this number to 90% or 99%, a doubling of the number of chromosomes tested (from 30 000 to 60 000) is required. This may not be practical nor an efficient use of resources. Perhaps identification of 80% of the essential genes is good enough "saturation" and other approaches should be developed to identify the rest. It is more practical to move on to new regions of the genome. To move on to other regions, one needs crossover suppressors (translocations, inversions) or duplications. Because of the ease of manipulation, the duplication approach must be preferred because it provides the most rapid method for isolating and mapping lethal mutations. The essential gene analysis could be extended to the rest of chromosome I. To study the region from the end of sDp2 to unc-75, there is the duplication nDp4. nDp4 will recover lethal mutations in the bli-4 dpy-14 region and in adjacent unanalyzed regions. Thus in a screen with nDp4, some of the complementation groups will overlap with those indentified in the sDp2 set. 194 Being a duplication, nDp4 has many of the advantages of sDp2. nDp4 has been tested as a balancer for new lethal mutations (A. Rose, S. McKay and K. McKim, unpublished results). From the right end of nDp4 to the right end of chromosome I there are no suitable duplications. M. Zetka (personal communication) has recently isolated an intrachromosomal rearrangement which acts as a dominant crossover suppressor in the unc-101 unc-54 region. This rearrangement should allow for essential gene analysis to be carried out in the region to the right of nDp4. This region is also of interest because it contains the homologue recognition site (see below). Mutations in the region of the homologue recognition site would help in its analysis. With the tools and techniques described here, identification of 80% of the essential genes on all of chromosome I is feasible. Stud i e s o n c h r o m o s o m e b e h a v i o u r : The study of the types and organization of essential genes on a chromosome is important in understanding the elements controlling the development of an organism. In addition, chromosomes have structural and organization features which are important for their own behaviour during meiosis and mitosis. Some of these features can be revealed by studying the effects of chromosome rearrangements on chromosome behaviour. I will first discuss the meiotic properties and then the mitotic properties of the rearrangements. I n t r a n s l o c a t i o n he te rozygo tes , the c r o s s o v e r s u p p r e s s i o n b o u n d a r y a n d the b r e a k p o i n t c o i n c i d e . The study of translocations in C. elegans has revealed three distinct effects on meiotic recombination. The first is the suppression of crossing over in translocation heterozygotes, an effect well documented in this thesis and by others (Herman 1978; Rosenbluth and Baillie 1981; Herman, Kari and Hartman 1982; Ferguson and Horvitz 1985, Rosenbluth, Cuddeford and Baillie 1985). A second effect documented in this thesis is an increase in recombination frequency in the homologously paired regions of translocation heterozygotes. This effect is probably ubiquitous to crossover suppressors. The third alteration is the increase in recombination frequency in homologously paired regions adjacent to the szTl breakpoint (McKim, Howell and Rose 1988). 195 The four chromosome I translocations studied in this paper failed to recombine in heterozygotes along their I^ 1 segments. From each translocation, one of the translocation chromosomes was used as a duplication to genetically map the breakpoints of szTl, hTl and hT3. In all three cases, the boundary of crossover suppression coincides closely with the breakpoint of the translocation. The result is that recombination is only observed to the right of the breakpoint. Although the chromosome I breakpoint of hT2 has not been as precisely mapped as the other three, the data are consistent with the coincident location of the breakpoint and the crossover suppression boundary. In the case of szTl, crossover suppression begins in the interval between let-88 and let-80, where the breakpoint occurs, and continues to let-362 at the left end of the chromosome. Similarly, hTl which has a breakpoint between let-80 and unc-29, also suppresses recombination from this point to the left end of the chromosome. In hT3 heterozygotes, recombination is suppressed to the left of dpy-5. The breakpoint is also close to the left of dpy-5. In all four translocations, the cross-over suppressed region corresponds to the portion of the translocation that segregates independently of the normal chromosome I. These findings agree with previous studies where the translocation breakpoints were known. In eTl (Rosenbluth and Baillie 1981) and mnT2 (Herman, Kari and Hartman 1982) heterozygotes, crossover suppression begins at or near a breakpoint and continues to one end of the chromosome. Thus, in C. elegans translocation heterozygotes, recombination is suppressed on only one side of the breakpoint. Failure to pair as the cause of recombination suppression in translocation heterozygotes has been proposed for D. melanogaster (Dobzhansky 1931; Sandler 1956; Roberts 1970; reviewed in Roberts 1976; Hawley 1980), and maize (Burnham et al. 1972). In D. melanogaster, crossover suppression does not extend along the length of translocation arms (reviewed in Roberts, 1976). In Drosophila and many other organisms, tetravalent structures and abundant recombination have been observed in all arms of a translocation heterozygote (reviewed in Rickards 1983). Thus, there is nothing intrinsic about the configuration of a translocation that would prevent pairing for recombination. As originally proposed by Rosenbluth and Baillie (1981), the 196 observations of translocation behavior in C. elegans are also most easily explained by the failure of the crossover suppressed region of the chromosome to pair. For example, the portion of chromosome V which in eTl(III;V) is recombinationally suppressed segregates independently of chromosome V (Rosenbluth and Baillie 1981). Similarly, the P-1 portion which is recombinationally suppressed in szTl heterozygotes segregates independently of chromosome I. carries the sequences required for pairing and recombination. The question of cross-over suppression in translocation heterozygotes may be a question of homologue pairing. What are the chromosomal features that determine homologue pairing and recombination at meiosis? Two hypotheses to explain homologue pairing resulting in recombination are 1) pairing by DNA sequence identity (see Smithies and Powers 1986) or 2) pairing initiating at a specific site(s) on a chromosome and spreading from this site to the rest of the homologous sequence. It is clear in C. elegans that homologue identity is not determined solely by DNA sequence identity. In translocation heterozygotes, two separate sections of the chromosome each share extensive DNA sequence identity with their homologues, yet only one section pairs and recombines (Figure 30). This recombination can occur up to the translocation breakpoint. Translocation homozygotes, however, exhibit recombination in the regions where suppression is observed in heterozygotes (Rosenbluth and Baillie 1981; K. Peters, personal communication). Thus in C. elegans, chromosome pairing requires information localized to one region of chromosome I. The location of homologue pairing sequences (resulting in recombination) in chromosome I , which herein will be referred to as the "the homologue recognition site", can be assigned to I R . This deduction is made because only fragments carrying sequences from the right end of chromosome I can pair and recombine with a normal chromosome I homologue. These sequences could be one site on jR or a number of interspersed sites. In hT2 heterozygotes, recombination is suppressed from let-362 to at least unc-101; a region covering about two thirds of the genetic map of chromosome I. This localizes the homologue recognition site to the right of unc-101 (Figure 30). Consistent with the translocation results are the results from two free duplications of chromosome I which have been studied in C. elegans (Rose, Baillie and Curran 1984). Only one of 197 bli-3 I — unc-11 dpy-5 unc-29 unc-101 | _ | 1 1 hT2 szT1 sDpl hT3 unc-54 hT3 sDp2 szT1 hT2 Figure 30: Diagram summarizing the chromosome I fragments with and without meiotic pairing activity (data on sDpl and sDp2 from Rose, Baillie and Curran 1984). Those fragments shown above the genetic map have meiotic pairing activity. Those shown below the line do not have meiotic pairing activity. The homologue recognition region around unc-54 is shown by the thick line. 198 these duplications, sDpl, is capable of pairing and recombining with the normal homologue. This duplication includes the unc-54 end of chromosome I which pairs and recombines with chromosome I in the four translocations. The duplication which does not recombine, sDp2, is derived from the left (let-362) end of chromosome I. This is the end which is crossover suppressed in the translocation heterozygotes. Although sDp2 may pair with its homologue, it lacks the information to pair for recombination. Therefore, with regard to the localization of sequences involved in homologue pairing for recombination, both the duplication and translocation studies agree. Similarly for other chromosomes for which two or more translocations or duplications have been studied, one end of the chromosome can be identified as containing the homologue recognition site(s) necessary for homologue pairing and recombination (Figure 31). The homologue recognition sites must lie outside the crossover suppressed region. In two translocation heterozygotes of chromosome II, mnT2(II;X) (Herman, Kari and Hartman 1982) and sT5(II;III) (H. I. Stewart, R. E. Rosenbluth and D. L. Baillie, personal communication), it is the right end which is crossover suppressed, leaving the left end as the location for the homologue recognition site. Three translocations of chromosome IH which have been studied are suppressed across intervals at the right (dpy-18) end: eTl(III;V) by Rosenbluth and Baillie (1981); sTl(III;X) by Rosenbluth, Cuddeford and Baillie (1985) and hT2(I;III). Thus the left (dpy-1) end has the sequences necessary for pairing and recombination. For chromosome IV it is the right (unc-22) end which is crossover suppressed in nTl(I;IV) (Ferguson and Horvitz 1985) and sT2(IV;V) (Rosenbluth, Cuddeford and Baillie 1985). In addition, the duplication mDpl(IV;f) (Rogalski and Riddle 1988) recombines with the normal chromosome IV. This duplication includes the left end of chromosome IV, the same end which pairs and recombines in the translocations described above. For chromosome V, it is the left (unc-60) end which is crossover suppressed in eTl(III;V) (Rosenbluth and Baillie 1981), nTl(IV;V) (Ferguson and Horvitz 1985), sT2(IV;V) (Rosenbluth, Cuddeford and Baillie 1985) and hTl(I;V). The right (unc-51) end carries the homologue recognition site. In the case of the X-chromosome, the right (unc-3) end is suppressed in szTl(I;X) (Fodor and Deak 1985) and mnT2(II;X) (Herman, Kari and Hartman 1982) and hT3. Studies on szTl (Fodor and Deak 199 1985; this paper) shows the dpy-3 region pairs and recombines with the normal X-chromosome while the portion of the chromosome to the right of dpy-3 is translocated to chromosome I and cross-over suppressed. Herman and Kari (1988) have come to the same conclusion from the study of X duplications which pair and recombine with the normal X chromosome. The homologue recognition site is located at the left (unc-1) end of the X-chromosome. Interestingly, the results with mnT10(V;X) (Herman, Kari and Hartman 1982) suggest the homologue recognition site may not be a localized entity, but may consist of extensive sequences. Even though mnTIO has a break in a similar location on the X-chromosome as mnT2, unlike mriT2 it suppresses recombination locally around the breakpoint but not in regions further removed on either side of the breakpoint. In mnT2, both translocation chromosomes seem to have homologue recognition activity, implying the site can be divided and still retain some activity. A peculiar feature of C. elegans translocations is that the breakage and joining of the translocation arms does not appear to be random. In all four of the translocations studied here, the end of the chromosome carrying the homologue recognition information is attached to the fragment of the other chromosome which does not carry this information. Of the seven C. elegans translocations where the structure has been determined, six have this pattern of rearrangement (Rosenbluth and Baillie 1981; Herman, Kari and Hartman 1982; this thesis). Perhaps this structure results from organizational features of the nucleus. Alternatively, chromosomes with either two homologue recognition sites or none may be particularly unstable. What is the function of the homologue recognition site? Two possibilities are suggested by cytological observations in a variety of organisms (reviewed in von Wettstein, Rasmussen and Holm 1984; Fussell 1987; Giroux 1988). Early in prophase, the chromosomes are usually positioned randomly about the nucleus. A long distance (>300 nm) attraction between the two homologues is the first stage of the pairing process. The result is a migration of the chromosomes, via the attachment of the telomeres to the inner nuclear membrane, until they congregate into a small area of the nucleus. This long range process will be referred to as the "pairing" phase. The 200 IV V X Figure 31- The locations of the homologue recognition sites for each C. elegans chromosome. The sites are indicated by the black thickened lines. The length of the black line does not indicate the size of the homologue recognition site, but instead depends on the available chromosome rearrangements used to define it (adopted from Rose and McKim 1990). 201 distance over which this process occurs precludes DNA sequence identity playing an important role. Once the homologues are in proximity, closer associations (lOOnm or less) follow resulting in synaptonemal complex formation, tight alignment of the chromosomes and recombination. This latter process is will be referred to as the "synapsis" phase and is most likely based on DNA sequence identity. Gene conversion type events have been proposed to be the mediator by which the two chromosomes seek perfect alignment through an homology search (Smithies and Powers 1986, Albini and Jones 1987, Carpenter 1987). Alani, Padmore and Kleckner (1990) have proposed that RAD50 protein is required for the homology search in S. cerevisiae. The homologue recognition site could function in either or both of these phases. Maguire (1984) has proposed there are sites which mediate the long range interaction, perhaps with the aid of chromosome specific proteins. Maguire (1984) has ruled out the telomeres as having a direct role in homologue recognition. Sites required for synapsis and recombination have also been proposed in studies on other organisms (reviewed in Maguire 1984; Giroux 1988). For example, in D. melanogaster, Sandler (1956) proposed discrete pairing sites and Hawley (1980) provided evidence for the existence and location of four of these sites distributed along the X-chromosome. Nonetheless, the issue remains controversial and it is not always clear if these sites are required for the early pairing phase or the later synaptic phase. Cytological observations show that the number and pattern of synaptic initiations appear to be species specific, but it can vary within a species (reviewed in Fussell 1987; Giroux 1988). In maize, for example, synapsis has been shown to begin at interstitial sites at both ends (not centromeric or telomeric) of chromosomes 1 and 5 (Burnham et al. 1972). The precise position of the initial and secondary site of synaptic initiation was not random but could vary. Similar results have been obtained using genetic analysis. Surosky and Tye (1988) in S. cerevisiae and Roberts (1970, 1976) in D. melanogaster found the sequences required for normal pairing and recombination were located in the arms of chromosomes. 202 The evidence described here leaves not doubt there is a site on each C. elegans chromosome required for either pairing or synapsis. The question is whether the homologue recognition site functions in the long range pairing phase or the synapsis phase. The position of the translocation breakpoint does not influence the fidelity of synapsis. Synapsis and recombination occurred right up to the breakpoint of all four chromosome I translocations studied. Thus, synapsis may occur after and be distinct from homologue pairing. There appear to be other sites on each C. elegans chromosome which control the normal distribution of recombination events and therefore probably function in the synapsis phase (see below). These sites appear to act secondarily to the homologue recognition site and are required for the normal initiation of synapsis. Fragments of the right end of chromosome I have both the homologue recognition site and the sites for regulating synapsis and therefore they recombine with the normal homologue. Fragments from the left end of chromosome I have the sites for regulating synapsis (see below) but do not recombine with the normal homologue because they lack the homologue recognition site. The homologue recognition site appears to be required early in meiosis, perhaps for the long range pairing of chromosomes. A similar conclusion was reached by Rosenbluth, Johnsen and Baillie (1990). The homologue recognition site could directly lead to the homologues finding each other, or it could simply provide the means for the chromosomes to aggregate non-specifically in a region of the nucleus but close enough to allow DNA-DNA interactions to detect homology. Homologue recognition could even involve premeiotic events. The observation that hTl heterozygotes produce nondisjunction events in clusters, suggests that some premeiotic event produces a clone of meiotic cells in which nondisjunction has occurred or will occur. The proposal is that each C. elegans chromosome requires only a single site for homologue recognition. Proposals in other organisms, however, call for a number of pairing sites on each chromosome. On the X-chromosome of D. melanogaster, for example, Hawley (1980) proposed there were at least four pairing sites. Since C. elegans chromosomes are considerably smaller than those of most other organisms, such as those of Drosophila or maize, it is possible that fewer recognition sequences for pairing are required. 203 Recombination frequencies are altered in translocation heterozygotes. The two or three-fold enhancement of recombination found in a region (unc-101 - unc-54) removed from the breakpoints of szTl, hTl, hT2 and hT3 may reflect a general response of chromosomes to recombination suppression. In this situation it is reasonable to propose the enhancement is a compensation for the suppression existing elsewhere on the chromosome. The total recombination on I R in all four chromosome I translocation heterozygotes was at least 40 m.u., compared to 40-45 m.u. in the entire normal chromosome. These "compensatory" increases are strictly an intrachromosomal effect in that they occur on the same chromosome as does the suppression. Previously reported increases of recombination in Drosophila translocation heterozygotes (Hawley 1980; Barabanova, Mamon and Vatti 1985) were attributed by Barabanova, Manon and Vatti (1985) to the interchromosomal effect. The interchromosomal effect is a phenomenon where chromosome rearrangements (inversions, translocations) on one chromosome increase and sometimes decrease the general cellular levels of recombination on the other chromosomes (reviewed in Luchessi 1976). The increases are mostly observed in regions of low intrinsic exchange, such as centromeric regions. McKim, Howell and Rose (1988) tested the effects of szTl heterozygosity on recombination in three intervals of low intrinsic exchange (i.e. autosomal gene clusters) on other chromosomes and found no change in the recombination frequency. In this study, I tested the effects of translocation heterozygosity on recombination in unlinked non-cluster regions and found no effects. Thus, the enhancement of recombination observed in translocation heterozygotes cannot be attributed to general increases in recombination frequency. These increases indicate there are genetic controls on the frequency of recombination along any one C. elegans chromosome. Meiotic mutants have been isolated in C. elegans which only effect recombination and segregation of the X-chromosome. These mutants do cause interchromosomal increases in recombination frequency (Hodgkin, Horvitz and Brenner 1979; Herman and Kari 1989). Under a model proposed by Luchessi and Suzuki (1968) to explain interchromosomal effects in Drosophila, the observations in C. elegans can be explained. Luchessi and Suzuki proposed that pairing 204 difficulties caused by chromosomes rearrangements lengthened the time for prophase I to be completed. In this extended time frame, more exchange events could be initiated between chromosomes correctly synapsed. In C. elegans translocation heterozygotes, no pairing problems would be expected because the crossover suppressed regions have no pairing activity. If the him mutants have a pairing defect, they might prolong prophase I and increase the number of opportunities for exchange to be initiated on the autosomes. The observed amount of compensation is chromosome specific, not translocation specific. In a variety of chromosome I crossover suppressors with a variety of breakpoints, the frequency of recombination in the region still able to meiotically pair was always 45%. In the two translocation crossover suppressors of chromosome III examined in this study, eTl and hT2, recombination was enhanced in the homologously paired regions to a level of only 30%. The recombination enhancement on the X chromosome in szTl heterozygotes may also be compensatory. The results, however, showed only 26 m.u. in the recombining portion of the X -chromosome. If this is the extent of the "compensation", then the 10% nondisjunction in hermaphrodites is expected from non-crossover bivalents. It is not known if there is more crossing over to the left of unc-1 or if double crossovers occur at an increased frequency in the translocation heterozygotes. Segregation: While mutants defective in recombination are usually disjunction defective as well (Baker, et al. 1976), this is not to say that recombination is sufficient for proper disjunction (Hawley 1988). As described below, meiotic mutants in C. elegans and other organisms can be disjunction defective but recombination proficient. In addition, Merriam and Frost (1964) observed greater than 50% of the non-disjoining X-chromosomes in wild-type Drosophila were recombinant chromosomes. Similarly in humans, crossing over accompanies a significant number of chromosome 21 nondisjunction events (Stewart et al. 1989). I observed high levels of chromosome I nondisjunction in hT2 heterozygotes despite the fact that recombination was at wild-type levels. The high level of nondisjunction in hT2 heterozygotes may be caused by there being only a small 205 region available for pairing and recombination which produces problems for segregation. As described below, the mnl 64 mutation may also cause a defect in segregation. In contrast, recombination is not absolutely required for segregation. The observation of non-homologous segregation of duplications from the X-chromosome in males is an example (Herman, Madl and Kari 1979; this thesis). Segregation of non-homologous chromosomes has been known for a long time in Drosophila as "Distributive pairing" (reviewed in Grell 1976; Hawley 1989). A similar phenomenon has been recently described in yeast (Dawson, Murray and Szostak 1986). In both Drosophila and C. elegans, recombination per se does not exclude chromosomes from non-homologous segregation. Holm and Chovnick (1975) observed that compound chromosomes entered into nonhomologous pairing, even when there was recombination between the attached arms. Novitski (1975, 1978) observed that both exchange and non-exchange chromosomes could participate in the distributive pairing process. In C. elegans, linked chromosome fragments also have the ability to segregate from the X-chromosome in males. Also like Drosophila (reviewed in Grell 1976), the efficiency of non-homologous segregation in C. elegans appears to be influenced by chromosome size. While segregation does not require recombination, it does require some form of meiotic pairing. In D. melanogaster it has been proposed that the segregation of recombinant and non-recombinant (distributive) bivalents is dependent on a single pairing phase (Novitski 1964, 1975, 1978, Hawley 1989). The forces resulting in the segregation of non-homologues may not, hoewever, be active until after exchange pairing has occurred (Harger and Holm 1980; Hawley 1989). It is not known if the two types of segregation in C. elegans also result from a single pairing phase. The results did indicate, however, that premeiotic events were involved in determining the nonhomologous segregation of linked duplications. As suggested above, premeiotic events may also be involved in homologous pairing. This similarity would allow for the two types of segregation to have a similar or the same pairing phase. The cytogenetic observations that C. elegans chromosomes are holokinetic (Albertson and Thomson 1982) raises a problem for segregation. During mitosis and meiosis, the spindle is 206 observed to attach at many sites along the C. elegans chromosome. The problem is that during meiosis, dicentric bridges would result if spindles were attached to either side of a crossover. The results and conclusions of this paper, however, can be understood assuming that C. elegans chromosomes behave as if they are meiotically monocentric. No unusual mechanisms are needed to explain the behavior of the chromosomes reported in this paper. The observations with Parascaris nematodes (Goday and Pimpinelli 1989) are in agreement with the proposal that C. elegans chromosomes are meiotically monocentric. During mitosis, Parascaris chromosomes are holokinetic. During meiosis, however, the spindles attach to only one end of the chromosomes. Similar observations have been made in C. elegans (D. Albertson, personal communication). Investigating pairing and synapsis with other rearrangements. The proposed two stages of homologue pairing in C. elegans may be individually disrupted with a variety of chromosome rearrangements. The crossover suppression in translocation heterozygotes may have resulted from the inability of the suppressed regions to carry out long range alignment. Whatever the mechanism of the close alignment, some of the alterations in recombination frequency observed in the rearrangements reported here may result from disrupting this D N A dependent phase in C. elegans. While deficiencies suppress recombination in the region they delete, there are few reports of them having an effect on recombination in adjacent regions (reviewed in Rosenbluth, Johnsen and Baillie 1990). In C. elegans, Rosenbluth, Johnsen and Baillie (1990) discovered that some chromosome V deficiencies dominantly reduced recombination for a considerable distance beyond the deleted regions. Most of these deletions were internal deletions. They proposed that subsequent to homologue recognition, synapsis was initiated at both ends of chromosome V and then alignment was subsequently ensured via the action of "alignment points". Internal deletions, by allowing non-homologous alignment points to synapse, disrupted recombination in adjacent regions. In this study, I characterized deletions at three locations on chromosome I; at the left end, in the middle and at the right end of the chromosome. The effects of the deletions on recombination frequency was position dependent. 207 The deletions I characterized at the left end of chromosome I could be analogous to those on chromosome V described by Rosenbluth, Johnsen and Baillie (1990). The right end of chromosome I has the homologue recognition site. Thus, like chromosome V, I have isolated mutations which map to the end of chromosome I opposite to the homologue recognition site which dominantly reduce recombination. Since non-overlapping mutations suppress recombination, the data rules out the possibility that a single site on chromosome I left is required for effective synapsis. It is possible that internal deletions at the ends of chromosome V and chromosome I opposite the homologue recognition site are revealing a process operating at this end of each chromosome to insure normal levels of recombination in these regions. Furthermore, as proposed by Rosenbluth, Johnsen and Baillie (1990), the mutations are probably disrupting close synapsis and not long range pairing. Evidence supporting this conclusion comes from analysis of the crossover suppressor mutation h655. This mutation maps to the left end of chromosome I and suppresses recombination over the entire left half of chromosome I. On the right end of chromosome I of h655 heterozygotes, recombination was enhanced over the normal frequencies. Thus h655 alters the distribution of exchange events but not the fidelity of pairing. Deletions in the middle of the chromosome reduced recombination in the region close to the left but no where else on the chromosome. These deletions may be causing a localized disruption in synapsis. At the right end of chromosome I, the region containing the homologue recognition site, I was unable to detect an effect on recombination with deletions. The lack of an effect on recombination in this regions suggests synapsis may differ mechanistically between the two ends of the chromosome. These results also indicate that the deletions had no effect on homologue recognition. This may indicate it is impossible to delete the recognition site; perhaps because it is a haploinsufficient site or it covers a large area and deleting part of it has no observable effect. The region would benefit, from a more thorough analysis, particularly if more deletions were available. Complete coverage of the region by the existing deletions was not possible because they were isolated as heterozygotes with a subterminally located lethal mutation (Anderson and Brenner 1984). 208 The mnl 64 mutation may be a terminal deletion of the right end of chromosome I. mnl 64 is a I-X fusion chromosome (D. Albertson personal communication) and apparently causes high levels of chromosome I nondisjunction, mnl 64 frequently recombines with the normal chromosome I, but the disjunction of the homologues was almost random. Either exchange tetrads fail to nondisjoin in mnl 64 heterozygotes, or there are high levels of meiosis II nondisjunction. These results are similar to those of Herman, Kari and Hartman (1982) on the X-chromosome of mnl 64 heterozygotes. Despite high levels of recombination, the X-chromosomes in mnl 64 heterozygotes segregate almost randomly (25-30% male production). Herman, Kari and Hartman (1982) showed nondisjunction occurred at both meiosis I and meiosis II leading them to propose mnl 64 caused a segregation defect on the X-chromosome and did not prevent synapsis and recombination. I propose a similar explanation for the chromosome I mnl 64 defect. The high level of chromosome I nondisjunction could then result from precocious separation at metaphase I or equational nondisjunction at meiosis II. This would make the chromosome I defect similar to the X-chromosome defect. The location of the mnl 64 mutation is at the homologue recognition end of both chromosomes, mnl 64 may disrupt centromere activity localized at these sites. This effect could be due to the fusion per se or could result from the deletion of specific sequences. Further analysis of deletions located at the homologue recognition ends of C. elegans chromosomes may resolve this issue. In contrast to the results with autosomes I and V, internal deficiencies of the right end of the X chromosome (opposite to the homologue recognition site) had no effect on recombination frequency in adjacent regions. This observation suggests the mechanism of synapsis on the X-chromosome has significant differences compared to the autosomes. The existence of X-chromosome specific meiotic mutations also suggests the X-chromosome has unique genetic controls for meiotic pairing and recombination (Hodgkin, Horvitz and Brenner (1979). In addition, the X is the only chromosome in C. elegans that does not have a clustering of genes due to meiotic recombination suppression (Brenner 1974). It is possible that the differences in synapsis 209 mechanisms between the X-chromosome and the autosomes is partly responsible for the different patterns of recombination events. S o m e i n s e r t i o n he t e rozygo tes h a v e r e d u c e d r e c o m b i n a t i o n . The effects of several insertions on recombination frequency were analyzed. The duplication hDpl4, an insertion of chromosome I material into the X-chromosome, alters X-chromosome recombination frequencies. To the left of the insertion site recombination is either unaffected or slightly increased. To the right, however, recombination is decreased tenfold. The region for homologue recognition on the X-chromosome was proposed to be at the left (unc-1) region (see above and Herman and Kari 1989). If the progression of synapsis was polar from the homologue recognition end of the X-chromosome, the chromosome I sequences might inhibit recombination by blocking the progression of synapsis down the chromosome. Recombination is normal (or enhanced) on the side of the insertion nearest the homologue recognition site because pairing can proceed efficiently up to the insertion site. The insertion apparently cannot efficiently be "jumped" by the pairing process resulting in the disruption of synapsis. The disruption of pairing and recombination on one side of a heterology is analogous to that observed for translocations. Pairing and recombination occur on one half of a reciprocal translocation but cannot initiate on the other half because there is no homologue recognition information there. Smaller insertions had less severe effects on X-chromosome recombination. hDpl 02 decreased X-chromosome recombination but it was not as severe as hDpl4 since the reductions in recombination did not spread to the right end of the chromosome. In sDp30 heterozygotes there was little, if any, effect on X-chromosome recombination although it was inserted into a similar location as hDpl 02. These data suggest the ability and strength of the recombination suppression caused by X-chromosome insertions may be dependent on their size or the type of sequence inserted. hDpl4 is certainly the biggest insertion and hDpl 02 is likely larger than sDp30. Perhaps sDp30 is small enough to allow the progression of synapsis to "jump" the heterology. Similarly, Herman, Albertson and Brenner (1976) observed a polar effect on recombination with the duplication mnDpl(X;V). This duplication is inserted into chromosome V and lowers 210 recombination on the left half of this chromosome. I three-factor mapped the site of mnDpl on chromosome V and found it was located near the left end of the chromosome. The pattern of recombination suppression in mnDpl heterozygotes thus more closely resembles the internal deficiencies of Rosenbluth, Johnsen and Baillie (1990; and as described above) than the insertions on the X-chromosome. The duplication hDp78 is located close to the right of unc-54 on chromosome I. It has a localized recombination reduction phenotype at the right end of chromosome I but the rest of the chromosome recombines at a high frequency. The defect in hDp78 is probably a local disruption in synapsis because recombination is normal in the rest of the chromosome, indicating homologue recognition is functioning. The number of known autosomal insertions is limited in C. elegans. It would be informative to have more with corresponding information on their effects on recombination and segregation. hDp73 and sDp8 are an examples of insertions that would benefit from such a study because they are linked to an autosome (section 3.13.1). Two other insertions were reported by Herman, Madl and Kari (1979). mnDp27(X;II) caused an increase in recombination frequency on the left half of chromosome II, despite being located to the right of that region. This enhancement may have resulted from recombination suppression on the left end of chromosome II. The second insertion, mnDp33(X;IV), had little effect on chromosome IV recombination; a situation similar to that observed with sDp30. From the data presented here, it is most likely these insertions disrupt the second stage of chromosome pairing, the formation of tight synapsis, and not the first, long distance phase. The main reason for this conclusion is their limited or polar effects on recombination. If these insertions disrupted homologue recognition and impaired the ability to form any synapsis, I would expect recombination to be reduced in all regions of the chromosome. Duplications used to study recombination and segregation of chromosomes. Two observations support the assertion that duplications of the left half of chromosome I have no pairing activity during meiosis. The first is the failure of sDp2 to recombine with the normal homologue (Rose, Baillie and Curran 1984). The second is the observation that these duplications 211 have no influence on the recombination frequency between the normal homologues in the duplicated regions. These observations are in contrast to the two duplications of chromosome I, sDpl and IRVLhTl, which carry sequences on the right end of chromosome I and do compete for pairing and recombination with the normal homologues (sDpl reported by Rose, Baillie and Curran 1984). Despite the fact both sDpl and IRV^hTl seem to have an intact chromosome I homologue recognition site, they have different meiotic properties. IRV^hTl has less chromosome I sequences than sDpl but is more proficient at recombining with the normal homologues. IRVLhTl was involved in recombination more often with the normal homologues and also caused nondisjunction of the normal homologues while sDpl did not. The reason for the difference is not known. IRV^hTl has a large part of chromosome V attached to it. These sequences may influence the ability of the chromosome I sequences to engage in successful pairing. Another possibility that cannot be ruled out is that sDpl carries additional sequences which inhibit its ability to engage in successful pairing. Rhoades (1936) has described a chromosome fragment in maize with properties similar to sDpl. Despite the meiotic pairing activity of this fragment, the two normal homologues usually segregated from each other. A derivative of IRVLhTl, hDpl 01, was isolated which recombines less frequently with the normal homologues and causes less homologue nondisjunction. This derivative makes IRVLhTl behave more like sDpl. There were no large differences in structure between IRVLhTl and hDpl 01. For example, the chromosome V sequences of IRVLhTl are unchanged in hDplOl. More study of hDpl 01 is required to understand its defect in meiotic pairing. I have also used X-chromosome duplications to study chromosome behavior during meiosis. All four of the duplications studied have a portion from the left end of the X-chromosome attached to the left end of chromosome I, allowing me to use chromosome I markers to monitor the behavior of the X-chromosome. These duplications exhibit meiotic pairing activity, the strength of which depends on the size of duplicated material and whether or not there is a competitor. In hermaphrodites, these duplications cause low levels of nondisjunction and infrequently recombine 212 with one of the two normal X-chromosomes. It is likely that the two normal X-chromosomes form a more stable association than a Dp-X association. On the X-chromosome the homologue recognitions site is located at the left (unc-1) end (this thesis ; Herman and Kari 1989). Thus if the homologue recognition site specifies information for long range pairing, the difference between the duplications and the normal X-chromosomes must be in a later step, perhaps during synapsis. The ILXLhT3 chromosome as a duplication caused more X-chromosome nondisjunction than hDp31 or hDp56. I^X^hTS has a larger amount of X-chromosome material than either of the other duplications, which may explain how it can be a better competitor for pairing with the X-chromosomes. Without competition from another X-chromosome, hDp31 and hDp56 were very good pairing partners. In male meiosis, these duplications disjoined from the X-chromosome greater than 90% of the time. This is a greater segregation efficiency than observed with non-homologous segregation and probably reflects true meiotic pairing and synapsis. While recombination had a significant role in the segregation of the duplications and the X-chromosome, only 30% of the segregating chromosomes were recombinant. Consistent with these data are either a high frequency of double crossover events or high frequency segregation of non-recombinant chromosomes. Almost identical results were found by Herman and Kari (1989) for some of their unc-1 duplications. Recombination was enhanced relative to the hermaphrodite; a region of less than 10 m.u. in the hermaphrodite was increased to greater than 30 m.u. in the male. The high level of recombination observed between the duplication and the X-chromosome may be mechanistically similar to the enhancements observed in crossover suppressor heterozygotes. In addition, it may be similar to the high level of crossing over observed in the pseudoautosomal region of male mice (Soriano et al. 1987) and humans (reviewed in Ellis and Goodfellow 1989). The pseudoautosomal region is the relatively small region of homology where pairing and recombination occur between the X and Y chromosomes in males. Despite its small size, there is at least one crossover in the region every meiosis. In contrast to humans, there may be no crossover interference in the pseudoautosomal region of male mice. My data are consistent with 213 recombination between the duplications and the X-chromosome having little associated interference. When hDp31 was homozygous, the two duplication chromosomes segregated from each other at a high frequency. The frequency of hDp31 nondisjunction in the homozygotes was on the order of that seen in hermaphrodites for normal X-chromosomes. Given a homologue, hDp31 behaves like a normal C. elegans chromosome. In general there is a correlation between the ability of hDp31, hDp56 and I^X^hTS to compete for, pair with and recombine with a homologue, and the amount of sequence identity. This suggests the ability to synapse is dependent on the amount of homologous sequence. This type of correlation would be predicted by the models for synapsis based on a sequence homology search (Smithies and Powers 1986; Albini and Jones 1987; Carpenter 1987; Alani, Padmore and Kleckner 1990). I also believe hDp51 has meiotic pairing behavior, but not as strong as the other duplications. The main difference between hDp51 and the other duplications is the amount of chromosome I material. As with the comparison of I^V^hTl to sDpl, nonhomologous sequences may play a role in meiotic pairing behavior. This may simply be a dependency on chromosome size. T h e effects o f m e i o t i c m u t a n t s o n r e c o m b i n a t i o n a n d s e g r e g a t i o n . I have studied the effects that two meiotic mutations have on recombination frequencies in a variety of regions in the C. elegans genome, him-3 and him-6 were isolated by Hodgkin, Horvitz and Brenner (1979) because they caused X-chromosome nondisjunction. These authors also provided evidence that these mutations had increased amounts of autosomal nondisjunction. I have confirmed this result. Autosomal nondisjunction was measured using translocation heterozygotes to produce gametes which could rescue disomic I and disomic III gametes from him strains. Reduction in crossing over is the primary defect in many nondisjunction mutants from other organisms (reviewed in Baker et al. 1976). In this study, him-3 demonstrated high levels of nondisjunction, but had little effect on X-chromosome recombination. The effects of him-3 on autosomal recombination were regional. The most severe reductions in recombination frequency 214 on the autosomes were at the ends of the chromosomes opposite that containing the homologue recognition site. Because him-3 reduces crossing over in the same regions that are affected by deletions (Rosenbluth, Johnsen and Baillie 1990; this thesis), it is tempting to speculate that him-3 is involved in the same synapsis process proposed to be disrupted by the deletions. On the X-chromosome, however, recombinant tetrads must non-disjoin because recombination was not grossly affected in him-3 hermaphrodites. Therefore, on the X-chromosome, him-3 might disrupt a post synapsis stage such as chiasma maintenance or segregation, while on the autosomes it is also required for steps in the initiation or resolution of recombination. A variety of mutations from other organisms cause non-disjunction of recombinant tetrads [ncd ;the chromosome function locus of claret nondisjunctional, cand (Sequeira, Nelson and Szauter 1989; Yamamoto et al. 1989), aid in Drosophila, desynaptic (dy) in maize and red-1 in S. cerevisiae (all reviewed in Hawley 1988), DISl in S. cerevisiae (Rockmill and Fogel 1988)]. It is now known that ncd is a microtubule motor protein (Endow, Henikoff and Soler-Neidziela 1990). him-3 differs from these in its effects on recombination frequency. him-3 is similar to the class of Drosophila mutation termed precondition mutations by Carpenter and Sandler (1974) because they disrupt the normal distribution of crossover events. In particular, him-3 is similar to the ord mutant in Drosophila. ord is a precondition mutant, as crossing over is reduced more in some regions than others, and nondisjunction is increased (Mason 1976). Unlike the other recombination defective mutants, however, nondisjunction occurs independently of recombination. It remains to be determined if recombinant chromosomes non-disjoin in him-3 mutants. The effects of him-6 on autosomal recombination were similar to the effects on the X-chromosome. Since the effects of him-6 on recombination appear to be less regionally constrained than the effects of him-3 in hermaphrodites, the him-6 gene may encode a recombination function. A example of such a mutation from D. melanogaster is mei-9, which reduces crossing over to 8-16% of controls (Carpenter and Sandler 1974). mei-9 flies are defective in excision repair (Boyd, Golino and Setlow 1976) and the gene presumably encodes a product involved in DNA 215 metabolism. The effects of him-6(el423) on recombination are not as drastic as mei-9, but el423 does not appear to be a null mutation. Using segregation of hDpl2 and hDp31 as an assay, the effects of him-3, him-6 and him-8 mutants on segregation were determined. In males, hDp!2 segregates from the X-chromosomes by the non-homologous process, him-3 eliminated this non-homologous segregation in males. In contrast, him-6 and him-8 did not effect non-homologous segregation of hDpl2 males. The fact that him-6 has no effect on non-homologous segregation is consistent with the proposal that it encodes a recombination function. hDp31 segregation was not affected by either him-3 or him-8. In the absence of recombination in him-8 males (see Herman and Kari 1989), segregation could still continue using the non-homologous system, him-8 had no effect on non-homologous segregation in hDpl2 males. Assuming an absence of non-homologous disjunction in him-3 males, segregation may result from normal pairing and recombination activity directing some segregation of hDp31 and the X-chromosome. Genetic analysis of duplication stability: In the following section I discuss the results bearing on chromosome stability during mitosis. I analyzed the segregational and structural stability of a variety of duplications to gain insights into the features which influence chromosome behaviour in C. elegans. Shortening duplications decreases mitotic stability: Considerable variation in the mitotic stability of the chromosome I duplications was observed; some were lost more frequently than others. Instability was indicated by a decrease in the number of Dp carrying gametes and an increase in the frequency of genetically mosaic worms. The isolation of genetically mosaic worms results from mitotic loss of the duplications. It is not know how much loss occurrs during meiosis. Herman, Albertson and Brenner (1976) observed considerable amounts of premeiotic duplication loss in cytogenetic observations of duplication strains. This behavior has been interpreted as a consequence of the holokinetic structure of C. elegans mitotic chromosomes (Albertson and 216 Thomson 1982). Albertson and Thomson (1982) proposed the efficiency of spindle attachment was proportional to kinetochore length. Thus smaller chromosome fragments, with their smaller kinetochores, would sometimes have insufficient spindle attachment for proper segregation. Consistent with this hypothesis, size is an important determinant of chromosome I duplication stability. The results, however, show there are significant exceptions to a strict relationship between duplication size and stability. hDpl4, hDp20 and hDp73 are more stable than expected, but this is easily explained by their attachment to other chromosomes or fragments. Less easily explained are hDp2, hDp4, hDp62 and hDp72. The stabilities found in this group are surprisingly low for their size. hDp2, for example, is 30% less stable than three other duplications of the same size. One possibility is that different structures of the duplications result in different stabilities. For example, circular chromosomes are observed to have more anomalies in mitotic and meiotic behaviour than linear chromosomes (McClintock 1938; Leigh 1976; for a review in humans see Matalon et al. 1990). Another exception to a strict size - stability correlation was found by determining the effects of deleting certain regions from the duplications. Deletion of the dpy-5 unc-13 region had minor effects on duplication stability whereas a deletion of the unc-57 unc-74 region significantly reduced duplication stability. For example, there was little difference between sDp2, hDpl2 and hDp 15 whereas there were significant differences between the three groups hDp 18 and hDp3, 5, 10 and hDp4, 6, 11. The extreme instability that accompanied deletion in the unc-11 dpy-5 region when compared to similar sized deletions in the dpy-5 unc-13 region was not the result of deleting larger amounts of DNA. The dpy-5 unc-13 region contains at least as many genes and as much or more DNA than the unc-11 dpy-5 region (Howell et al. 1987; Kim and Rose 1987; Howell 1989; Howell and Rose 1990; Starr et al. 1989; Starr 1990). Since size is not the only determinant of stability, what other factors could be involved? The first possibility, differences in structure, has already been discussed. Secondly, a threshold level of chromosome size may be necessary for stability. If a size threshold is important, then deletions of chromosome material do not have significant effects above the size threshold. 217 Duplications like hDp3 would be smaller than the threshold and be in the range where the probability of spindle attachment is sensitive to size changes. A third possibility is the DNA sequences responsible for stability may not be evenly distributed along the chromosome. These sequences need not be kinetochore attachment sites. If DNA sequences are an important factor, then hDp6 is less stable than hDp3 because sequences conferring mitotic stability have been deleted. The results do not unambiguously resolve these possibilities. In fact, any combination of these may be influencing the stability of a given duplication. Some free X-chromosome duplications have higher levels of loss in the oocyte line than in sperm (Herman, Albertson and Brenner 1976). I observed a lower frequency of duplication recovery in the hermaphrodite oocyte and the male sperm than in the hermaphrodite sperm. Hermaphrodite sperm are made before oocytes and undergo fewer cell divisions (Kimble and Hirsh 1979). This could explain the different rates of loss. The more frequent and rapid cell divisions of oogenesis could provide more opportunity for duplication loss. Similarly, male sperm undergo more cell divisions than hermaphrodite sperm. hDpl 0 and hDp22 were recovered at equal frequencies in both hermaphrodite germ lines. Some duplications may be less sensitive to loss than others during the cell divisions of oogenesis and male spermatogenesis. Mosaic analysis can be carried out on chromosome I genes using the more unstable duplications. Herman (1984) has used unstable duplications of the X chromosome to create genetically mosaic worms. This allows one to ask questions regarding the tissue specificity and cell autonomous expression of a gene. sDp2 is too stable for mosaic analysis but strains with shorter duplications frequently produce mosaic worms. Using hDp2 or hDp23 levamisole selection, the frequency of mosaics was approximately one in 2500 progeny when dpy-5 and a levamisole resistant unc were used. This is similar to the frequency of genetic mosaics recovered by Herman (1984). Without the use of levamisole selection, a greater variety of mosaics was recovered. The frequency of mosaics was 1 in 40 worms with hDp2 and 1 in 120 worms with hDp3. 218 Duplication chromosomes spontaneously break: A striking aspect of these results was the spontaneous shortening of the duplications. Spontaneous shortening was observed with ten duplications. In addition, I observed spontaneous rearrangement of non-duplication chromosomes. For example, the insertion of chromosome I material into the X-chromosome (hDpl4) made it prone to breakage. High frequency spontaneous rearrangement of chromosomes in szTl and hT2 strains was also observed. The spontaneous shortening events appear to have significant differences from those induced with gamma radiation. The spontaneous duplications had breakpoints in regions not hit by the gamma radiation. For example, no dpy-5(-) dpy-14(+) derivatives of sDp2 were obtained with gamma radiation, but they were common among the spontaneous duplications. Another difference between gamma radiation induced and spontaneous duplications was that the latter were often extremely unstable, recovered in only a few percent of the gametes. Such duplications were rarely recovered in the gamma radiation induced set. Previously, Herman (1984) observed spontaneous shortening of one X-chromosome duplication in C. elegans. Possibly all duplications will spontaneously shorten. Whether spontaneous shortening at the frequencies observed here occurs in normal chromosomes in a euploid genetic background is not known. Depending on their structure, however, different duplications shorten at different rates. Mitotically stable duplications (sDp2, hDp5, hDp20) shorten less frequently than unstable duplications (hDp2, hDp23). The correlation between mitotic stability and spontaneous shortening leads to the suggestion that mitotic/meiotic segregation problems cause both duplication loss and breakage. It is possible that the unstable duplications have difficulty aligning with the spindle during nuclear divisions (meiotic or mitotic). If the chromosome is not properly aligned, the chromatids may not be equally divided between the sister cells, lost altogether or undergo a breakage event (see below). Albertson and Thomson (1982) provided cytological evidence for this type of behavior during mitosis with the free X-chromosome duplication mnDp2. They observed the duplication often lying on the outside of the metaphase plate or lagging at anaphase. 219 The improper alignment of duplication chromosomes during nuclear division may provide an opportunity for breakage. A holokinetic chromosome could be pulled apart if the different spindles from the same pole do not attach to the same chromatid. Misalignment might allow spindles from opposite poles to attach to the same chromatid. If two adjacent spindles were pulling in opposite directions the duplication chromosome might break apart and the result would be a shorter duplication. Improper spindle attachment to normal chromosomes probably occurs but the rate at which this occurs or at which exceptional events are recovered is not known. The reason for the elevated frequency of spontaneous shortening for some of the duplications could be because they misalign on the metaphase plate more often than sDp2. During meiosis, being univalent poses problems for a chromosome. Most duplications are not engaged in exchange events and should be univalents during meiosis. Other authors have observed that univalents in corn (Miller 1963), wheat (Sears 1952) and Drosophila melanogaster (Sandler and Braver 1954; Carpenter 1973; Baker 1975) are unstable. The univalent may divide equationally at meiosis I, or either chromatid may be lost at meiosis I or II, or the univalent may undergo misdivision. Misdivision occurs when a metacentric univalent attempts to separate transversely at the centromere resulting in telocentric chromosomes. Because the duplications studied here do not pair and recombine with chromosome I, I have not addressed the influence of pairing and recombination on the spontaneous shortening frequency of a duplication. In addition, the frequency at which a normal chromosome breaks has not been determined. Furthermore, it is possible that shortening is not entirely the result of mitotic instability, but could also be a consequence of a particular chromosome structure. Chromosome breakage is observed with dicentric chromosomes produced by recombination events involving inversion or circular chromosomes (McClintock 1938, 1941; Haber, Thorburn and Rogers 1984). Although it is assumed genetic mosaics are produced through duplication loss (Herman 1984), it is also possible that a genetic mosaic could be generated by a breakage event in the lineage under study [as observed in maize (McClintock 1938, 1941)]. Mosaics produced through a breakage event may be a significant fraction of the total number based on my observations that 220 the recovery of mosaics and breakage events occurs at similar frequencies. A mosaic produced through a breakage event can give different results than a mosaic produced through complete loss. In a chromosome loss event, all alleles on the duplication are lost from that lineage. In a breakage event, some of the alleles would remain. Two examples of spontaneous chromosome shortening have been observed in D. melanogaster. Biessmann and Mason (1988) described terminal deficiencies of the X chromosome which were isolated in strain carrying the mu-2 mutation. This mutation caused a defect in the repair of double strand breaks. It has been shown that these chromosome ends have no telomere D N A (Beissmann, Carter and Mason 1990). Levis (1989) isolated terminal deletions of chromosome 3 when a subterminally located P element was the site of double strand cleavage. Both of these deletion chromosomes shortened by approximately 75 bp per generation. These experiments did not report attempts to recover large deletion events. It is possible that the duplications described here are shortening in small increments. I have not done experiments to detect these events. Breakage producing chromosomes lacking a proper telomere are mitotically unstable in maize (McClintock 1941) and yeast (Haber, Thorburn and Rogers 1984) and until recently (see above) have not been isolated in Drosophila (Muller and Herskowitz 1954; Roberts 1975). The structures of natural C. elegans telomeres or the ends of duplication chromosomes are unknown. Also unknown are the behavioral characteristics of the unrepaired ends of C. elegans chromosomes. By analogy with Drosophila, it is reasonable to postulate that in the process of shortening sDp2 or I^X^szTl with gamma radiation, shortened chromosomes were capped with telomere carrying fragments of other chromosomes originating from the same gamma radiation exposure (Muller and Herskowitz 1954; Williamson and Parker 1976). In the case of the spontaneous breakage events, however, capping in this way cannot be proposed. The free ends created by a breakage event may be repaired by the acquisition of telomere containing sequences from another genomic location; involving either intrachromosomal or interchromosomal rearrangements (McClintock 1941; Haber and Thorburn 1984; Rudin and 221 Haber 1988). Supporting this proposal is the observation that some shortening events are associated with attachment to another chromosome (Herman and Kari 1989, this thesis). Alternatively, a stable end could be created through the de novo addition of a telomere as proposed in Tetrahymena (King and Yao 1982). This event could be facilitated by a telomerase activity (Zakian, Runge and Wang 1990). In nematodes where chromatin diminution occurs, it may be required that the ends created by breakage of the holokinetic chromosomes be stable or have the ability to self heal (White 1973; Pimpinelli and Goday 1989). Spontaneous "healing" of chromosome ends was observed in maize (McClintock 1941), Drosophila (Traverse and Pardue 1988; Biessman et al. 1990) Plasmodium falciparum (Pologe and Ravetch 1988) and S. pombe (Matsumoto et al. 1987). McClintock (1941) observed that unstable chromosome ends were healed when passed into the sporophyte. My observation that most recoverable spontaneous breakage events may only occur at or after meiosis is reminiscent of this observation. Still another solution to the telomere problem is a ring chromosome. Based on deletion analysis of the type described here, evidence in favor of the existence of ring chromosomes has been collected by myself and others (C. P. Hunter and W. B. Wood, personal communication). Not all duplications, however, are rings. sDp2 appears to be linear because deletions of one end usually do not carry over to the other end. The most radical possibility is that there is no healing event at all and perhaps no standard telomere structure in holokinetic organisms. The solution to the structure of the duplication telomere awaits the cloning of a duplication end. 222 5. Bibliography Alani, E. , R. Padmore and N. Kleckner, 1990 Analysis of wild-type and rad50 mutants of yeast suggests an intimate relationship between meiotic chromosome synapsis and recombination. Cell 61: 419-436. Albertson, D. G. and J . N. Thomson, 1982 The kinetochores of Caenorhabditis elegans. Chromosoma 86: 409-428. Albini, S. M. and G. H. Jones, 1987 Synaptonemal complex spreading in Allium cepa and A. fistulosum. Chromosoma 95: 324-338. Anderson, P. and S. Brenner, 1984 A selection for myosin heavy-chain mutants in the nematode Caenorhabditis elegans. Proc. Natl. Acad. Sci. 81: 4470-4474. Baker, B. S., 1975 Paternal loss (pal): a meiotic mutant in D. melanogaster causing loss of paternal chromosomes. Genetics 80: 267-296. Baker, B. S., A. T. C. Carpenter, M. S. Esposito, R. E. Esposito and L. Sandler, 1976 The genetic control of meiosis. Ann. Rev. Gen. 10: 53-134. Barabanova, L. V., L. A. Mamon and K. V. Vatti, 1985 Crossing over in the translocation heterozygotes with break points mapping outside the pericentric region of chromosomes 2 and 3 of Drosophila melanogaster. Soviet Genetics 21: 1560-1565. Barton, M. K. and J . Kimble, 1990 fog-1, a regulatory gene required for specification of spermatogenesis in the germ line of Caenorhabditis elegans. Genetics 125: 29-39. Beachy, P. A., 1990 A molecular view of the Ultrabithorax homeotic gene of Drosophila. Trends Genet. 6: 46-51. Beadle, G. W., 1932 A possible influence of the spindle fiber on crossing-over in Drosophila. Proc. Natl. Acad. Sci. USA 18: 160-165. Beyers, B. and L. Goetsch, 1975 Electron microscopic observations on the meiotic karyotype of diploid and tetraploid Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 72: 5056-5060. Biessmann, H., S. B. Carter and J. M. Mason, 1990 Chromosome ends in Drosophila without telomeric DNA sequences. Proc. Natl. Acad. Sci. USA 87: 1758-1761. Biessmann, H. and J . M. Mason, 1988 Progressive loss of DNA sequences from terminal chromosome deficiencies in Drosophila melanogaster. EMBO J. 7: 1081-1086. Biessmann, H., J . M. Mason, K. Ferry, M. d'Hulst, K. Valgeirsdottir, K. L. Traverse and M. Pardue, 1990 Addition of telomere associated HeT DNA sequences "heals" broken chromosome ends in Drosophila. Cell 61: 663-673. Boyd, J. B., M. D. Golino and R. B. Setlow, 1976 The mei-9a mutant of Drosophila melanogaster increases mutagen sensitivity and decreases excision repair. Genetics 84: 527-544. Brenner, S., 1974 The genetics of Caenorhabditis elegans. Genetics 77: 71-94. 223 Brenner, S., W. Dove, I. Herskowitz and R. Thomas, 1990 Genes and development: molecular and logical themes. Genetics 126: 479-486. Burnham, C. R., J . T. Stout, W. H. Weinheimer, R. V. Kowles and R. L. Phillips, 1972 Chromosome pairing in maize. Genetics 71: 111-126. Carpenter, A. T. C , 1973 A mutant defective in distributive disjunction in Drosophila melanogaster. Genetics 73: 393-428. Carpenter, A. T. C , 1975 Electron microscopy of meiosis in Drosophila melanogaster females. II. The recombination nodule - a recombination-associated structure at pachytene? Proc. Natl. Acad. Sci. USA 72: 3186-3189. Carpenter, A. T. C , 1987 Gene conversion, recombination nodules, and the initiation of meiotic synapsis. Bioessays 6: 232-236. Carpenter, A. T. C , 1988 Thoughts on recombination nodules, meiotic recombination and chiasmata, pp 529-548, in Genetic Recombination, edited by R. Kucherlapati and G. R. Smith. American Society for Microbiology, Washington, D.C. Carpenter, A. T. C. and L Sandler, 1974 On recombination-defective meiotic mutants in Drosophila melanogaster. Genetics 76: 453-475. Chandley, A. C , 1986 A model for effective pairing and recombination at meiosis based on early replicating sites (R-bands) along chromosomes. Hum. Genet. 72: 50-57. Chandley, A. C , 1988 Meiosis in man. Trends Genet. 4: 79-84. Clark, D. V., 1990 The unc-22 region in Caenorhabditis elegans: genetic analysis and molecular mapping. Ph.D. thesis, Simon Fraser University, Burnaby, Canada. Clark, D. V., T. M. Rogalski, L. M. Donati and D. L. Baillie, 1988 The unc-22(IV) region of Caenorhabditis elegans: Genetic analysis of lethal mutations. Genetics 129: 345-353. Clark, L. , 1990 Centromeres of budding and fission yeasts. Trends. Genet. 6: 150-154. Clark, L. and M. P. Baum, 1990 Functional analysis of a centromere from fission yeast: A role for centromere-specific repeated DNA sequences. Mol. Cell. Biol. 10: 1863-1872. Clark, S. H. and A. Chovnick, 1986 Studies on normal and position-affected expression of rosy region genes in Drosophila melanogaster. Genetics 114: 819-840. Comings, D. E. and A. D. Riggs, 1971 Molecular mechanisms of chromosome pairing, folding, and function. Nature 233: 48-50. Coulson, A., J . Sulston, S. Brenner, and J. Karn, 1986 Towards a physical map of the genome of the nematode Caenorhabditis elegans. Proc. Natl. Acad. Sci. 83: 7821-7825. Coulson, A., R. Waterston, J. ruff, J . Sulston, and Y. Kohara, 1988 Genome linking with yeast artificial chromosomes. Nature 335: 184-186. Crow, E. L. and R. S. Gardner, 1959 Confidence intervals for the expectation of a poisson variable. Biometrika 46: 441-453. 224 Daniels, S. B., M. McCarron, C. Love and A. Chovnick, 1986 The underlying bases gene expression differences in stable transformants of the rosy locus in Drosophila melanogaster. Genetics 113: 265-286. Darlington, C. D., 1932 Recent advances in Cytology. The Blakiston Co., Philadelphia. Dawson, D. S., A. W. Murray and J . W. Szostak, 1986 An alternative pathway for meiotic chromosome segregation in yeast. Science 234: 713-717. DeLong, L. , L. P. Casson and B. J . Meyer, 1987 Assessment of X chromosome dosage compensation in Caenorhabditis elegans by phenotypic analysis of lin-14. Genetics 117: 657-670. Dobzhansky, T., 1931 The decrease of crossing over observed in translocations, and its probable explanation. Am. Nat. 65: 214-232. Duncan, I., 1987 The Bithorax complex. Ann. Rev. Genet. 21: 285-319. Earnshaw, W. C , 1988 Mitotic chromosome structure. Bioessays 9: 147-150. Edgley, M. L. and D. L. Riddle, 1987 Caenorhabditis elegans. In Genetic Maps 1987, Vol. 4, Edited by S. J . O'Brien. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Endow, S. A., Henikoff, S. and L. Soler-Niedziela, 1990 Mediation of meiotic and early mitotic chromosome segregation in Drosophila by a protein related to kinesin. Nature 345: 81-83. Engebrecht, J . , J . Hirsch and G. S. Roeder, 1990 Meiotic gene conversion and crossing over: Their relationship to each other and to chromosome synapsis and segregation. Cell 62: 927-937. Ferguson, E. L. and H. R. Horvitz, 1985 Identification and characterization of 22 genes that affect the vulval cell lineages of the nematode Caenorhabditis elegans. Genetics 110: 17-72. Ferguson, E. L. and H. R. Horvitz, 1989 The multivulva phenotype of certain Caenorhabditis elegans mutants results from defects in two functionally redundant pathways. Genetics 123: 109-121. Fodor, A. and P. Deak, 1985 The isolation and genetic analysis of a C. elegans translocation (szTl) strain bearing an X-chromosome balancer. J . of Genet. 64: 143-157. Fussell, C. P., 1987 The Rable orientation: a prelude to synapsis, pp 275-299. In: Meiosis, Edited by P. B. Moens. Academic Press, London. Gehring, W. J . and Y Hiromi, 1986 Homeotic genes and the homeobox. Ann. Rev. Genet. 20: 147-173. Giroux, C. N., 1988 Chromosome synapsis and meiotic recombination, pp 465-496, in Genetic Recombination, edited by R. Kucherlapati and G. R. Smith. American Society for Microbiology, Washington, D.C. Goday, C. and S. Pimpinelli, 1989 Centromere organization in meiotic chromosomes of Parascaris univalens. Chromosoma 98: 160-166. 225 Greenwald, I., A. Coulson, J. Sulston and J . Priess, 1987 Correlation of the physical and genetic maps in the lin-12 region of Caenorhabditis elegans. Nucleic Acids Res. 15: 2295-2307. Greider, C. W. and E. H. Blackburn, 1985 Identification of a specific telomere terminal transferase activity in Tetrahymena extracts. Cell 43: 405-413. Grell, R. F., 1976 Distributive pairing, pp. 435-486. In: The Genetics and Biology of Drosophila, Vol. la, Edited by M. Ashburner and E. Novitski. Academic Press, New York. Haber, J. E. and P. C. Thorburn, 1984 Healing of broken linear dicentric chromosomes in yeast. Genetics 106: 207-226. Haber, J. E. , P. C. Thorburn and D. Rogers, 1984 Meiotic and mitotic behavior of dicentric chromosomes in Saccharomyces cerevisiae. Genetics 106: 185-205. Harger, H. and D. G. Holm, 1980 Meiotic behavior of compound autosomes in females of Drosophila melanogaster: Interchromosomal effects and the source of spontaneous nonsegregation. Genetics 96: 455-470. Hawley, R. S., 1980 Chromosomal sites necessary for normal levels of meiotic recombination in Drosophila melanogaster. I. Evidence for and mapping of the sites. Genetics 94: 625-646. Hawley, R. S., 1988 Exchange and Chromosomal Segregation in Eucaryotes, pp 497-527 in Genetic Recombination, edited by R. Kucherlapati and G. R. Smith. American Society for Microbiology, Washington, D.C. Hawley, R. S., 1989 Genetic and molecular analysis of a simple disjunctional system in Drosophila melanogaster, pp. 277-302 in Molecular and Cytogenetic studies of non-disjunction. Alan. R. Liss, Inc. Heine, U, and T. Blumenthal, 1986 Characterization of regions of the Caenorhabditis elegans X chromosome containing vitellogenin genes. J . Mol. Biol. 188: 301-312. Herman, R.K., 1978 Crossover suppressors and balanced recessive lethals in Caenorhabditis elegans. Genetics 88: 49-65. Herman, R. K., 1984 Analysis of genetic mosaics of the nematode Caenorhabditis elegans. Genetics 108: 165-180. Herman, R. K. and C. K. Kari, 1989 Recombination between small X-chromosome duplications and the X-chromosome in Caenorhabditis elegans. Genetics 121: 723-737. Herman, R. K., D. G. Albertson and S. Brenner, 1976 Chromosome rearrangements in Caenorhabditis elegans. Genetics 83: 91-105. Herman, R. K., C. K. Kari and P. S. Hartman, 1982 Dominant X-chromosome nondisjunction mutants of Caenorhabditis elegans. Genetics 102: 379-400. Herman. R. K., J . E. Madl and C. K. Kari, 1979 Duplications in Caenorhabditis elegans. Genetics 92: 419-435. Hodgkin, J . A., H. R. Horvitz and S. Brenner, 1979 Nondisjunction mutants of the nematode C. elegans. Genetics 91: 67-94. 226 Holliday, R., 1977 Recombination and meiosis. Phil. Trans. R. Soc. Lond. B. 277: 359-370 Holm, D. G. and A. Chovnick, 1975 Compound autosomes in Drosophila melanogaster: The meiotic behavior of compound thirds. Genetics 81: 293-311. Horvitz, H. R. and J . E. Sulston, 1980 Isolation and genetic characterization of cell-lineage mutants of the nematode Caenorhabditis elegans. Genetics 96: 435-454. Horvitz, H. R., S. Brenner, J. Hodgkin and R. Herman, 1979 A uniform genetic nomenclature for the nematode Caenorhabditis elegans. Mol. Gen. Genet. 175: 129-133. Howell, A. M., 1989 Essential genes in a region of chromosome I in Caenorhabditis elegans. Ph.D. thesis, University of British Columbia, Vancoucer, Canada. Howell, A. M. and A. M. Rose, 1990 Essential genes in the hDf6 region of chromosome I in Caenorhabditis elegans. Genetics 126: 583-592. Howell, A.M., S.G. Gilmour, R.A. Mancebo and A.M. Rose, 1987 Genetic analysis of a large autosomal region in Caenorhabditis elegans by the use of a free duplication. Genet. Res., Camb. 49: 207-213. Johnsen, R. C. and D. L. Baillie, 1988 Formaldehyde mutagenesis of the eTl balanced region in Caenorhabditis elegans. Mutat. Res. 201: 137-147. Jones, G. H., 1984 The Control of Chiasma Distribution, pp. 293-320 in Controlling Events in Meiosis, edited by C. W. Evans and H. G. Dickinson. The Society for Experimental Biology, Cambridge. Jones, G. H., 1987 Chiasma, pp 213-244. In: Meiosis, Edited by P. B. Moens. Academic Press, London. Kim, J. S. and A. M. Rose, 1987 The effect of gamma radiation on recombination in Caenorhabditis elegans. Genome 29: 457-462. Kimble, J. and D. Hirsh, 1979 Postembryonic cell lineages of the hermaphrodite and male gonads in Caenorhabditis elegans. Develop. Biol. 70: 396-417. King, B. O. and M.-C. Yao, 1982 Tandemly repeated hexanucleotide at Tetrahymena rDNA free end is generated from a single copy during development. Cell 31: 177-182. Kossikov, K. V. and H. J . Muller, 1935 In validation of the genetic evidence for branched chromomeres in the case of the Pale translocation in Drosophila. J . Hered. 26: 305-317. L'Hernault, S. W., D. C. Shakes and S. Ward, 1988 Developmental genetics of chromosome I spermatogenesis-defective mutants in the nematode Caenorhabditis elegans. Genetics 120: 435-452. Lambie, E. J . and G. S. Roeder, 1986 Repression of meiotic crossing over by a centromere (CEN3) in Saccharomyces cerevisiae. Genetics 114: 769-789. Lefevre, G. and W. Watkins, 1986 The question of total gene number in Drosophila melanogaster. Genetics 113: 869-895. 227 Leigh, B., 1976 Ring chromosomes and radiation induced chromosome loss, in The Genetics and Biology of Drosophila, Vol. lb. Edited by M. Ashburner and E. Novitski. Academic Press, London. Levis, R. W., 1987 Viable deletions of a telomere from a Drosophila chromosome. Cell 58: 791-801. Lewis, J. A., C. H. Wu, H. Berg and J.H. Levine, 1980a The genetics of levamisole resistance in the nematode Caenorhabditis elegans. Genetics 95: 905-928. Lewis, J. A., C. H. Wu, J.H. Levine and H. Berg, 1980b Levamisole resistant mutants of the nematode Caenorhabditis elegans appear to lack pharmacological acetylcholine receptors. Neuroscience 5: 967-989. Luchessi, J . C , 1976 Inter-chromosomal effects, pp. 315-329. In: The Genetics and Biology of Drosophila, Vol. la, Edited by M. Ashburner and E. Novitski. Academic Press, New York. Luchessi, J . C. and D. T. Suzuki, 1968 The interchromosomal control of recombination. Ann. Rec. Genet. 2: 53-86. McClintock B., 1938 The production of homozygous deficient tissue by means of aberrant mitotic behavior of ring-shaped chromosomes. Genetics 23: 315-376. McClintock B., 1941 The stability of broken ends of chromosomes in Zea mays. Genetics 26: 234-282. McClintock B., 1942 The fusion of broken ends of chromosomes following nuclear fusion. Proc. Natl. Acad. Sci. USA 28: 458-463. McDowall, J . S., 1990 Essential genes in the hDpl6/hDpl9 region of chromosome I in Caenorhabditis elegans. M.Sc. thesis, University of British Columbia, Vancouver, B.C., Canada. 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. Maguire, M. P., 1984 The mechanism of meiotic homologue pairing. J. theor. Biol. 106: 605-615. Mann, C. and R. W. Davis, 1986 Meiotic disjunction of circular minichromosomes in yeast does not require DNA homology. Proc. Natl. Acad. Sci. USA. 83: 6017-6019. Mason, J . M., 1976 Orientation disrupter (ord): a recombination-defective and disjunction-defective meiotic mutant in Drosophila melanogaster. Genetics 84: 545-572. Matalon, R., P. Supple, H. Wyandt and I. M. Rosenthal, 1990 Transmission of ring 14 chromosome from mother to two sons. Am. J . Med. Genet. 36: 381-385. Matsumoto, T., K. Fukui, O. Niwa, N. Sugawara, J . W. Szostak and M. Yanagida, 1987 Identification of healed terminal DNA fragments in linear minichromosomes of Schizosaccharomyces pombe. Mol. Cell. Biol. 7: 4424-4430. Meneely, P. M. and R. K. Herman, 1979 Lethals, steriles and deficiencies in a region of the X chromosome of Caenorhabditis elegans. Genetics 92: 99-115. 228 Meneely, P.M. and R. K. Herman, 1981 Suppression and function of X-linked lethal and sterile mutations in Caenorhabditis elegans. Genetics 97: 65-84. Merriam J. R. and J . N. Frost, 1964 Exchange and nondisjunction of the X chromosome in female Drosophila melanogaster. Genetics 49: 109-122. Miller, O., 1963 Cytological studies of asynaptic in maize. Genetics 48: 1445-1466. Moens, P. B. Ed., 1987 Meiosis. Academic Press, London. Moens, P. B. and Earnshaw, W. C., 1989 Anti-topoisomerase II recognized meiotic chromosome cores. Chromosoma 98: 317-322. Moens, P. B. and R. E. Pearlman, 1988 Chromatin organization at meiosis. Bioessays 9: 151-153. Moerman, D. G. and D. L. Baillie, 1981 Formaldehyde mutagenesis in Caenorhabditis elegans. Mut. Res. 80: 273-279. Morin, G. B., 1989 The human telomere terminal transferase enzyme is a ribonucleoprotein that synthesizes TTAGGG repeats. Cell 59: 521-529. Moses, M. J . , M. E. Dresser and P. A. Poorman 1984 Composition and role of the synaptonemal complex, pp. 245-270 in Controlling Events in Meiosis, edited by C. W. Evans and H. G. Dickinson. The Society for Experimental Biology, Cambridge. Muller, H. J. and I. H. Herskowitz, 1954 Concerning the healing of chromosome ends produced by breakage in Drosophila melanogaster. Am. Naturalist 88: 177-208. Novitski, E. 1964 An alternative to the distributive pairing hypothesis in Drosophila. Genetics 50: 1449-1451. Novitski, E. 1975 Evidence for the single phase pairing theory of meiosis. Genetics 79: 63-71. Novitksi, E. 1978 The relation of exchange to nondisjunction in heterologous chromosome pairs in the Drosophila female. Genetics 88: 499-503. Pimpinelli, S. and C. Goday, 1989 Unusual kinetochores and chromatin diminution in Parascaris. Trends Genet. 5: 310-315. Pologe. L. G. and Ravetch, J . V., 1988 Large deletions from breakage and healing of P. falciparum chromosomes. Cell 55: 869-874. Powers, P. A. and O. Smithies, 1986 Short gene conversions in the human fetal globin region: A by-product of chromosome pairing during meiosis? Genetics 112: 343-358. Prasad, S. S. and D. L. Baillie, 1989 Evolutionarily conserved coding sequences in the dpy-20 -unc-22 region of Caenorhabditis elegans. Genomics 5: 185-198. Rasmussen S. W. amd P. B. Holm, 1980 Mechanics of meiosis. Hereditas 93: 187-216. 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. 229 Rhoades, M. M., 1936 A cytogenetic study of a chromosome fragment in maize. Genetics 21: 491-501. Rickards, G. K., 1983 Orientation behavior of chromosome multiples of interchange (reciprocal translocation) heterozygotes. Ann. Rev. Genet. 17: 443-498. Roberts, L. , 1990 The worm project. Science 248: 1310-1313. Roberts, P. A., 1970 Screening for X-ray-induced crossover suppressors in Drosophila melanogaster: Prevalence and effectiveness of translocations. Genetics 65: 429-448. Roberts, P. A., 1975 In support of the telomere concept. Genetics 80: 135-142. Roberts, P. A., 1976 The genetics of chromosome aberration, pp. 68-184. In: The Genetics and Biology of Drosophila, Vol. la, Edited by M. Ashburner and E. Novitski. Academic Press, New York. Rockmill, B. and S. Fogel, 1988 DIS1: a yeast gene required for proper meiotic chromosome disjunction. Genetics 119: 261-272. Rockmill, B. and G. S. Roeder, 1988 RED1: a yeast gene required for the segregation of chromosomes during during the reductional division of meiosis. Proc. Natl. Acad. Sci. USA 85: 6057-6061. Rockmill, B. and G. S. Roeder, 1990 Meiosis in asynpatic yeast. Genetics 126: 563-574. Rogalski, T. M. and D. L. Riddle, 1988 A Caenorhabditis elegans RNA polymerase II gene, ama-1 IV, and nearby essential genes. Genetics 118: 61-74 Rose, A. M., 1980 Genetic studies on the gene coding for paramyosin in Caenorhabditis elegans: unc-15 and the adjacent region. Ph.D. thesis, Simon Fraser University, Burnaby, B.C., Canada. Rose, A. M. and D. L. Baillie, 1979 Effect of temperature and parental age on recombination and nondisjunction in Caenorhabditis elegans. Genetics 92: 409-418. Rose, A. M. and D. L. Baillie, 1980 Genetic organization of the region around unc-15 (I), a gene affecting paramyosin in Caenorhabditis elegans. Genetics 96: 639-648. 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. Rose, A. M. and K. S. McKim, 1990 Rose, A. M., D. L. Baillie, E. P. M. Candido, K. A. Beckenbach and D. Nelson, 1982 The linkage mapping of cloned restriction fragment length differences in Caenorhabditis elegans. Mol. Gen. Genet. 188: 286-291. Rosenbluth, R. E. and D. L. Baillie, 1981 Analysis of a reciprocal translocation, eTl(III;V), in Caenorhabditis elegans. Genetics 99: 415-428. Rosenbluth, R. E. , C. Cuddeford and D. L. Baillie, 1985 Mutagenesis in Caenorhabditis elegans. II. A spectrum of mutational events induced with 1500R of gamma-radiation. Genetics 109: 493-511. 230 Rosenbluth, R. E. , R. C. Johnsen, and D. L. Baillie, 1990 Pairing for recombination in LGV of Caenorhabditis elegans: A model based on recombination in deficiency heterozygotes. Genetics 124: 615-625. Rosenbluth, R. E. , T. M. Rogalski, R. C. Johnsen, L. M. Addison and D. L. Baillie, 1988 Genomic organization in Caenorhabditis elegans: deficiency mapping on linkage group V(left). Genet. Res. 52: 105-118. Rudin, N. and J. E. Haber, 1988 Efficient repair of HO-induced chromosomal breaks in Saccharomyces cerevisiae by recombination between flanking homologous sequences. Mol. Cell. Biol. 8: 3918-3928. Sandler, I., 1956 Studies in T(1;4)BS in Drospohila melanogaster. Master's thesis. University of Missouri. Sandler, L., and G. Braver, 1954 The meiotic loss of unpaired chromosomes in Drosophila melanogaster. Genetics 39: 365-377 Sears, E. R., 1952 Misdivision of univalents in common wheat. Chromosoma (Berlin) 4:S. 535-550. Sen, D. and W. Gilbert, 1988 Formation of parallel four-stranded complexes by guanine-rich motifs in DNA and its implications for meiosis. Nature 334: 364-366. Sen, D. and W. Gilbert, 1990 A sodium-potassium switch in the formation of four stranded G4-DNA. Nature 344: 410-414. Sequeira, W., C. R. Nelson and P. Szauter, 1989 Genetic analysis of the claret locus of Drosophila melanogaster. Genetics 123: 511-524. Sigurdson, D. C , G. J . Spanier and R. K. Herman, 1984 Caenorhabditis elegans deficiency mapping. Genetics 108: 331-345. Smithies, O. and P. A. Powers, 1986 Gene conversion and their relationship to homologous chromosome pairing. Phil. Trans. Roy. Soc. (Lond.) B. 312: 291-302. Soriano, P., E. A. Keitges, D. F. Schorderet, K. Harbers, S. M. Gartler and R. Jaenisch, 1987 High rate of recombination and double crossovers in the mouse psueodoautosomal region during male meiosis. Proc. Natl. Acad. Sci. USA 84: 7218-7220. Spofford, J. B., 1976 Position effect variegation in Drosophila. pp. 955-1018. In: Genetics and Biology of Drosophila, Vol. IC, Edited by M Ashburner and E. Novitski. Academic Press, New York. Starr, T., 1989 Molecular analysis of the dpy-14 region of chromosome I in Caenorhabditis elegans. Ph.D. thesis, University of British Columbia, Vancouver, Canada. Starr, T., A. M. Howell, J . McDowall, K. Peters and A. M. Rose, 1989 Isolation and mapping of DNA probes within the linkage group I gene cluster of Caenorhabditis elegans. Genome 32: 365-372. Stewart, G. D., M. L. Van Keuren, J . Gait, S. Kurachi, M. J . Buraczynska and D. M. Kurnit, 1989 Molecular structure of human chromosome 21. Ann. Rev. Genet. 23: 409-423. 231 Sulston, J. E. and S. Brenner, 1984 The DNA of Caenorhabditis elegans. Genetics 77: 95-104. Sulston, J . E. and H. R. Horvitz, 1977 Post-embryonic cell lineages of the nematode Caenorhabditis elegans. Dev. Bio. 56: 110-156. Sulston, J. E . and H. R. Horvitz, 1981 Abnormal cell lineages in mutants of the nematode Caenorhabditis elegans. Dev. Biol. 82: 41-55. Sulston, J. E. , E. Schierenberg, J. G. White and J. N. Thomson, 1983 The embryonic cell lineage of the nematode Caenorhabditis elegans. Dev. Biol. 100: 64-119. Sundquist, W. I. and A. Klug, 1989 Telomeric DNA dimerizes by formation of guanine tetrads between hairpin loops. Nature 342: 825-829. Surosky, R. T. and B. Tye, 1988 Meiotic disjunction of homologs in Saccharomyces cerevisiae is directed by pairing and recombination of the chromosome arms but not by pairing of the centromeres. Genetics 119: 273-287. Szauter, P., 1984 An analysis of regional constraints on exchange in Drosophila melanogaster using recombination-defective meiotic mutants. Genetics 106: 45-71. Traverse, K. L. and M. L. Pardue, 1988 A spontaneously opened ring chromosome of Drosophila melanogaster has aquired He-T DNA sequences at both new telomeres. Proc. Natl. Acad. Sci. USA. 85: 8116-8120. Triantaphyllou, A. C., 1987 Chromosomes in Evolution of Nematodes, pp. 77-102 in "Chromosomes in Evolution of Eukaryotic Groups, edited by A. K. Sharma and A. Sharma. CRC press, Boca Raton, Florida. von Wettstein, D., S. W. Rasmussen and P. B. Holm, 1984 The synaptonemal complex in genetic segregation. Ann. Rev. Genet. 18: 331-413. Wang, S. and V. A. Zakian, 1990 Telomere-telomere recombination provides an express pathway for telomere acquisition. Nature 345: 456-458. Ward, S. and J. S. Carrel, 1979 Fertilization and sperm competition in the nematode Caenorhabditis elegans. Dev. Biol. 73: 304-321. Watson, J. D., 1972 Origin of concatameric T7 DNA. Nature 239: 197-201. Wood, W. B. Ed., 1988 The nematode Caenorhabditis elegans. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. White, J . G., E. Southgate, J. N. Thomson and S. Brenner, 1986 The structure of the nervous system of Caenorhabditis elegans. Philos. Trans. R. Soc. Lond. Biol. Sci. 314: 1-340. White, M. J . D., 1973 Animal cytology and evolution. Cambridge University Press, Cambridge. Williamson, J . H. and D. R. Parker, 1976 Recombination between the X and Y chromosomes, in The Genetics and Biology of Drosophila, Vol. lb. Edited by M. Ashburner and E. Novitski. Academic Press, London. 232 Williamson, J . R., M. K. Raghuraman and T. R. Cech, 1989 Monovalent cation-induced structure of telomeric DNA: The G-quartet model. Cell 59: 871-880. Yamamoto, A. H., D. J . Komma, C. D. Shaffer, V. Pirrotta and S. A. Endow, 1989 The claret locus in Drosophila encodes products required for eyecolor and for meiotic chromosome segregation. EMBO J. 8: 3543-3552. Yu, G., J. D. Bradley, L. D. Attardi and E. H. Blackburn, 1990 In vivo alteration of telomere sequences and senescence caused by mutated Tetrahymena telomerase RNAs. Nature 344: 126-132. Zakian, V. A., 1989 Structure and function of telomeres. Ann. Rev. Genet. 23: 579-604. Zakian, V. A., K. Runge and S. Wang, 1990 How does the end begin? Formation and maintenance of telomeres in ciliates and yeast. Trends Genet. 6: 12-16. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items