Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mapping, characterization and cosmid rescue of essential genes in the dpy-5 unc-13 (I) region of Caenorhabditis… McDowall, Jennifer Susan 1996

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

Item Metadata

Download

Media
831-ubc_1996-091309.pdf [ 23.19MB ]
Metadata
JSON: 831-1.0087269.json
JSON-LD: 831-1.0087269-ld.json
RDF/XML (Pretty): 831-1.0087269-rdf.xml
RDF/JSON: 831-1.0087269-rdf.json
Turtle: 831-1.0087269-turtle.txt
N-Triples: 831-1.0087269-rdf-ntriples.txt
Original Record: 831-1.0087269-source.json
Full Text
831-1.0087269-fulltext.txt
Citation
831-1.0087269.ris

Full Text

MAPPING, CHARACTERIZATION AND COSMID RESCUE OF ESSENTIAL GENES IN T H E dpy-5 unc-13 (I) REGION OF Caenorhabditis elegans by JENNIFER SUSAN M T O W A L L B.Sc, The University of Edinburgh, 1983 M.Sc, The University of British Columbia, 1990 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF T H E REQUIREMENTS FOR T H E DEGREE OF DOCTOR OF PHILOSOPHY in T H E F A C U L T Y OF G R A D U A T E STUDIES (Genetics Graduate Programme) We accept this thesis as conforming to the required standard T H E UNIVERSITY OF BRITISH COLUMBIA April 1996 © Jennifer Susan M cDowall, 1996 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. itapartrnont" 'of (j&Oe-rJC'S Pf QCjcarfrmP The University of British Columbia Vancouver, Canada Date ftfXto.Wk ABSTRACT This thesis describes the analysis of essential genes in the 1.5 map unit dpy-5 unc-13 region of chromosome I in the C. elegans genome. A l l the lethal mutations analyzed came from a set of 495 EMS-induced, sD/?2-rescued lethals described in Howell etal. (1987). The 15 map unit sDp2 region is 80% saturated for essential genes (McDowall 1990). This set of lethals comprises a substantial 'mutant library', which should play an important role in correlating the function of essential genes with the sequence currently being generated by the C. elegans genome project (Waterston and Sulston 1995). In order to fulfill this role, these lethal mutations must first be placed on both the genetic and physical maps, so as to render them accessible for cloning and functional analysis. The goal of this thesis is to place a subset of these lethal mutations on both maps, and to further define their mutant phenotypes. In this study, I report the identification, mapping, cosmid rescue and characterization of essential genes in the dpy-5 unc-13 region. Forty lethal mutations lying between the breakpoints of hDpl3 and hDpl6 in the dpy-5 unc-13 region have been mapped and complementation tested. Seventeen of these lethals identify six new essential genes, while the remaining 23 lethals were allelic to nine known genes. This analysis brings the number of essential genes in the dpy-5 unc-13 region to 64, as defined by the sDp2 rescue of lethal mutants. 61% of these essential genes are identified by more than one mutation. Positioning of mutations lying between hDpl3 and hDp!6 was done using the breakpoints of six duplications. These genes appear to be nonrandomly distributed throughout the region, based on their arrest phenotypes. This completes the genetic mapping of the entire sDp2 set of lethal mutations between dpy-5 and unc-13. The mutant phenotypes of the loci essential for fertility were further characterized by Nomarski microscopy and DAPI-staining of their nuclei. In total, 14 genes in the dpy-5 unc-13 region, plus one gene to the right of unc-13, were characterized. None of the mutants were Ill rescued by male sperm, indicating that none of the mutations caused defects in spermatogenesis alone. The cytological data showed that four genes produced mutants with defects in gonadogenesis, let-395, let-603, let-605 and let-610. Seven genes, let-355, let-367, let-384, let-513, let-544, let-545, and let-606, produced mutants that affected germ cell proliferation or gametogenesis. Mutants of the remaining four genes produced eggs that failed to develop or hatch, let-370, let-538, let-599 and let-604, thereby acting as maternal effect lethals. Finally, a subset of the essential loci in the dpy-5 unc-13 region were positioned on the physical map by phenotypically rescuing lethal mutations with transgenic strains. Germline transformation with cosmid DNA was used to create 30 heritable extrachromosomal array-bearing transgenic strains. These arrays were used as small duplications for fine scale mapping. Mutations in 13 essential genes have been phenotypically rescued by 10 different transgenic strains. These rescues provide an alignment of the genetic and physical maps in this region. The frequency of transmission of extrachromosomal arrays to hermaphrodite gametes was calculated. In 12/16 transgenic strains, the arrays were found to be transmitted 2- to 7-fold less frequently in oocytes compared to hermaphrodite sperm. Three strains showed a subsequent increase in array stability in oocytes. This phenomenon may be influenced by cosmid sequences. Early mitotic loss of arrays was observed in all 17 strains examined, suggesting that loss of the array can occur at any time during development when cell divisions are occurring. This research has helped to refine the genetic map of the dpy-5 unc-13 region of chromosome I, as well as integrate the genetic and physical maps over 0.5 map units. There are now 64 essential genes in the dpy-5 unc-13 region identified from the sDp2 set of lethal mutations, of which 22% have been characterized with respect to their mutant phenotypes. As a result of this work, 13 of these loci are anchored to the physical map, providing links between the genetic and physical maps on average every 85 kb. These links will facilitate the future cloning and functional analysis of these loci. IV TABLE OF CONTENTS ABSTRACT ii LIST OF TABLES viii LIST OF FIGURES ix ACKNOWLEDGMENTS • • xi GENERAL INTRODUCTION 1 Chapter 1: Genetic Analysis of Essential Genes in the hDp]3-hDpl6 Region INTRODUCTION 8 MATERIALS and METHODS 10 S tock Maintenance and Strains 10 Chromosomal Rearrangements 11 Complementation Analysis of Lethal Strains 12 Duplication Mapping 15 RESULTS 17 Complementation Analysis of Lethal Strains 17 Mapping Lethal Mutations 20 DISCUSSION 23 Identification of Essential Genes 23 s£)/?2-rescued Recessive Lethals . 24 Forward Mutation Rates and Lethal Saturation 25 Developmental Arrest of Lethal Mutations , 27 Nonrandom Distribution of C. elegans Essential Genes 29 Genomic Organization of Essential Genes 31 Chapter 2: Characterization of Sterile Mutants in the dpy-5 unc-13 (I) Region INTRODUCTION 34 MATERIALS and METHODS 42 Test for Sperm-defective Mutants 42 Microscopic Characterization of Sterile Adult Mutants 43 RESULTS 45 Test for Sperm-defective Mutants 45 Microscopic Characterization of Sterile Adult Mutants: 45 (i) Mutations Affecting Gonadogenesis 46 (ii) Mutations Affecting Proliferation and Gametogenesis 47 (iii) Maternal Effect Lethals 49 DISCUSSION 72 Characterization of Adult Mutant Phenotypes 72 Mutations Affecting Gonadogenesis: 73 (i) let-395, let-605, let-610 ' 73 (ii) let-603 74 Mutations Affecting Proliferation and Gametogenesis 74 (i) let-355, let-384 75 (ii) let-606 76 (iii) let-513 78 (iv) let-544, let-545 79 (v) let-367 80 Maternal Effect Lethals 80 (i) let-538, let-599, let-604 82 (ii) let-370 83 Distribution of Sterile Mutations 84 Chapter 3: Cosmid Rescue of Lethal Mutations in the bli-4 (I) Region INTRODUCTION 86 MATERIALS and METHODS 88 vi Preparation of Cosmid DNA for Germline Transformation 88 Germline Transformation 88 Segregation Stability 92 Complementation Mapping to the Transgenic Strains 94 Polymerase Chain Reaction Analysis of Transgenic Worms 96 RESULTS 97 Germline Transformation 97 Segregation Stability of the Extrachromosomal Arrays 101 Rescue of Lethal Mutations by Complementation with Arrays: 111 (i) Complete Rescue of Lethal Phenotypes 114 (ii) Partial Rescue of Lethal Phenotypes 117 (iii) Alignment of the Genetic and Physical maps 129 Confirmation of the Presence of Cosmid Sequences 133 DISCUSSION 134 Germline Transformation 134 Germline Transmission of Extrachromosomal Arrays 136 Transmission Frequencies of Extrachromosomal Arrays 138 (i) Transmission Frequencies in Oocytes and Sperm 138 (ii) Changes in Transmission Frequencies 141 (iii) Early Mitotic Loss 143 Phenotypic Rescue of Lethals using Extrachromosomal Arrays 144 (i) Partial Rescue of Lethal Phenotypes 144 (ii) Phenotypic Variability of Partially Rescued Animals 148 (ii) Artificial Positive Results 149 , Alignment of Genetic and Physical Maps 151 Extrachromosomal Arrays as Mapping Tools 153 CONCLUSIONS 156 vii BIBLIOGRAPHY 161 Appendix 1: Strains Used 178 Appendix 2: Strains of New Lethal Mutations Used 180 Appendix 3: Complementation Results for Essential Genes 182 Appendix 4: Deriving the Equations for the Calculation of Transmission Frequencies 186 Appendix 5: Primers Used for PCR Analysis 188 Vlll LIST OF TABLES Table 1 Newly Identified Essential Genes in the hDp!3-hDpl6 Region 18 Table 2 Essential Genes in the hDpl3-hDpl6 Region 19 Table 3 Duplication Mapping Results 21 Table 4 Essential Genes in the dpy-5 unc-13 Region 32 Table 5 Characterization of Adult Mutant Phenotypes 52 Table 6 Plasmid and Cosmid DNA Used for Transformation 91 Table 7 Transgenic Strains Constructed 99 Table 8 Extrachromosomal Array Stability I: Frequency of Rol-6 Progeny 105 Table 9 Extrachromosomal Array Stability II: Frequency of Hermaphrodite Array-bearing Gametes 107 Table 10 Extrachromosomal Array Stability III: Frequency of Early Mitotic Loss 109 Table 11 Changes in Transmission Frequency 110 Table 12 Transgenic Complementation Results 113 Table 13 Rescued Transgenic Strains 116 Table 14 Partial Cosmid Rescue of Essential Genes 121 Table 15 Alignment of Rearrangement Breakpoints with the Cosmid Contig 132 IX LIST OF FIGURES Figure 1 Genetic Map of Essential Genes in the hDpl3-hDpl6 Region 13 Figure 2 Protocol for Complementation Analysis 14 Figure 3 Protocol for Duplication Mapping of Lethal Mutations 16 Figure 4 New Genetic Map of Essential Genes in the hDpl3-hDpl6 Region 22 Figure 5 Genetic Map of Essential Genes in the dpy-5 unc-13 Region 33 Figure 6 Gonadogenesis and Anatomy in Wild-type Hermaphrodites • 41 Figure 7 Protocol for let-538 44 Figure 8 Germline of Wild-type Hermaphrodites 53 Figure 9 Germline of Unc-13 Hermaphrodites 54 Figure 10 Germline of Dpy-5 Unc-13 Hermaphrodites 55 Figure 11 Germline of let-395(h271) Hermaphrodites 56 Figure 12 Germline of let-603(h289) Hermaphrodites 57 Figure 13 Germline of let-605(h312) Hermaphrodites 58 Figure 14 Germline of let-610(h695) Hermaphrodites 59 Figure 15 Germline of let-355(h81) Hermaphrodites 60 Figure 16 Germline of let-384(h84) Hermaphrodites ' , 61 Fi gure 17 Germline of let-606(h292) Hermaphrodites 62 Figure 18 Germline of let-513(h752) Hermaphrodites 63 Fi gure 19 Germline of let-544(H692) Hermaphrodites 64 Figure 20 Germline of let-545(h842) Hermaphrodites 65 Figure 21 Germline of let-367(hi 19) Hermaphrodites 66 Figure 22 Germline of let-538(h990) Hermaphrodites 67 Figure 23 Germline of let-599(h290) Hermaphrodites 68 Figure 24 Germline of let-604(h293) Hermaphrodites 69 Figure 25 Geimline of let-604(h490) Hermaphrodites 70 Figure 26 Germline of let-370( hl28) Hermaphrodites 71 Fi gure 27 Distribution of Genes Essential for Fertility in the dpy-5 unc-13 Region 85 Figure 28 Physical Map of the Cosmid Contig used in this Study 90 Figure 29 Equations Used to Calculate Transmission Frequencies 93 Figure 30 Protocol for Complementation Analysis of Transgenic Strains 95 Figure 31 Partial Rescue of let-376(h!30) 122 Figure 32 Partial Rescue of let-530(h798) 123 Figure 33 Partial Rescue of let-395(h271) 124 Figure 34 Partial Rescue of let-603(h289) 125 Figure 35 Partial Rescue of let-602(h283) 126 Figure 36 Partial Rescue of let-61 l(h850) 127 Figure 37 Partial Rescue of let-393(h225) 128 Figure 38 Map Showing the Cosmid Rescue of Essential Genes 131 XI ACKNOWLEDGMENTS I would like to thank my research supervisor, Dr. Ann Rose, and my supervisory committee Dr. Don Moerman, Dr. David Baillie, Dr. Steve Wood and Dr. Peter Candido for their helpful advice. The transformation techniques used in this thesis were taught to me by Dr. John Sulston and I would like to thank him for his time and effort. Many thanks to my colleagues in the lab, Monique Zetka, Ken Peters, Colin Thacker, Bob Johnsen, and Mark Edgely, for their help and discussions. I would especially like to thank my family, Phil, Richard, Elizabeth, and my parents, for their encouragement and faith in my abilities. This work was supported in part by a Natural Sciences and Engineering Research Council of Canada Graduate Fellowship. GENERAL INTRODUCTION Currently, the genomes of many organisms are being mapped and sequenced, including those of vertebrates (Adams etal. 1995; Dietrich et al. 1995), invertebrates (Spradling et al. 1995; WaterstonandSulston 1995), plants (Goodman etal. 1995; Carels etal. 1995), yeast (Oliver et al. 1992; Dujon etal. 1994), and bacteria (Fleischmann et al. 1995; Fraser et al. 1995). The advantage of sequencing the genomes of many organisms is that it enables a comparison of different species with regard to genome organization, the number of genes involved in different cellular functions, and the identification of species-specific genes. Such a collation of data requires the elucidation of the biological function of the genes identified in each genome. Hence, with the current trend in providing physical maps and sequence data for entire genomes, it becomes increasingly important to provide genetic data to help correlate genes with their functions, if the full benefits of whole genome analysis are to be realized. This need becomes apparent by the fact that the functions of 50% of the 4255 sequenced genes in Caenorhabditis elegans are unknown (Hodgkin etal. 1995; Waterston and Sulston 1995). Even in the smallest genome of a free-living organism, Mycoplasmagenitalium, which carries the minimal set of genes required for survival, 22% of the open reading frames have no similarity to those in current databases (Fraser etal. 1995; Goffeau 1995). Genetic analysis provides a powerful approach to discern the functions of genes and to integrate the biological data. This will become increasingly important as the genome projects reach their conclusion, and the emphasis of research turns from a structural to a functional approach. A well characterized 'mutant library', consisting of a set of mutations within a defined genetic interval, can provide a resource for such genetic studies. To be valuable, these mutations must be placed on both the genetic and physical maps, because the integration of these maps enables the correlation of sequence with biological function. The goal of this study is to describe the distribution of essential loci in a region of the gene cluster on L G I of Caenorhabditis elegans. The approach involves the genetic mapping of a set of lethal mutations, and the systematic cosmid rescue of several of these mutations, in order to align the genetic and physical maps over a 0.5 map unit region. C. elegans provides a powerful model system for the study of many biological problems, including eukaryotic genome organization, and even human disease loci. Over 40% of the genes in C. elegans show similarity to those in other organisms (Hodgkin etal. 1995). In particular, 80% of the human disease genes so far identified have significant similarity to C. elegans genes. These conserved genes can be mutated in C. elegans and subjected to genetic analyses not possible in human studies. Such analyses include both forward and reverse genetics, whereby mutants can be linked to the physical map by transgenic analysis (this study), or sequences can be used to create mutants by transposon gene disruption (Plasterk and Groenen 1992; Zwaal etal. 1993). The short generation time and large brood sizes (300-1400) of C. elegans allows genetic analysis to progress rapidly. C. elegans normally exists as self-fertilizing hermaphrodites, which facilitates the genetic analysis of severe morphological and behavioural mutants, since they are often difficult to mate. Males are produced by spontaneous X chromosome nondisjunction, making cross-fertilization possible (Hodgkin 1987). The development of this organism has been described (Hirsh etal. 1976; Sulston and Horvitz 1977; Riddle 1982; Chalfie 1984; Waterston and Francis 1985; Hodgkin 1987), providing invaluable groundwork for experimentation. The developmental expression patterns of cloned genes can be determined by fusing them to reporter genes and examining their P-galactosidase expression in transgenic animals (Lynch et al. 1995), or by somatic mosaic analysis (Herman 1984; Bucher and Greenwald 1991). These approaches enable the function of a gene to be examined within the context of an intact organism. As such, C . elegans is a suitable organism for addressing questions relating to functional genomics. The relatively small size of the C. elegans genome facilitates genomic investigation. The haploid genome of C. elegans consists of 100 Mb of DNA apportioned among five autosomes and an X chromosome (Sulston and Brenner 1974). For this reason, C. elegans has been used to pioneer physical mapping and sequencing techniques (Sulston et al. 1992; Wilson et al. 1994; Waterston and Sulston 1995). This has led to the most complete physical map (14 contigs), the most genomic sequence (22.5 Mb), and the most sequenced genes (4255 of the predicted 13,500) being available for C. elegans compared to any other organism. Of the 13,500 predicted genes in C. elegans (Waterston and Sulston 1995), about 1,400 genes have been mapped genetically (Edgley and Riddle 1993; Hodgkin et al. 1995), 4255 genes have been identified through sequencing (Wilson et al. 1994), and 1,194 genes have been identified by cDNA tagging (Waterston 1992). In a comprehensive effort to identify mutants for functional analysis, many large scale hunts for mutants have been carried out. These have been directed at saturating for specific mutant classes (Kemphues etal. 1988; Ward et al. 1988), generating transposon-induced mutant banks (Plasterk and Groene 1992; Zwaal etal. 1993), and saturating for essential genes (Meneely and Herman 1979; Sigurdson etal. 1984; Howell etal. 1987; Clark et al. 1988; Johnsen and Baillie 1991). A study by Barnes etal. (1995) combined genetic recombination data with physical map data to address the question of the organization of visible genes in C. elegans. The most striking feature of the C. elegans genetic map is the clustering of loci in the central region of each of the six chromosomes (Edgley and Riddle 1993). This clustering is due to a higher density of genes in the center region of the chromosomes (Waterston et al. 1992), as compared to the chromosome arms, and is enhanced by recombination suppression in the autosome arms (Kim and Rose 1987). Barnes etal. (1995) showed that a striking similarity existed between the autosomes in terms of gene density, arrangement, and recombination rates across the chromosome. Based on data accumulated for visible mutations, they concluded that the chromosomes carried a similar number of genes, and had a similar number of total mutational hits. Lethal genes were not included in this analysis, because data were available for only discrete regions which have been genetically balanced. This study points out the necessity of gathering further information regarding the distribution of essential genes, especially since they represent a large fraction of the mutable targets in C. elegans. This thesis focuses on genes that are essential for the development and fertility of C. elegans. Lethal and sterile mutations are potentially important tools for understanding the genetic basis of eukaryotic development and genome organization. A comprehensive set of lethal mutations should be fairly representative of each aspect of biological function, such as metabolism, cell structure, cell division, protein synthesis, and communication. Many large scale screens have been carried out in C. elegans with the intention of identifying essential genes that are located in a defined region by generating lethal mutations. Extensive lethal analysis has been carried out for defined chromosomal regions on L G I (Rose etal. 1980; Howell etal. 1987; Howell and Rose 1990; McKim etal. 1992), L G II (Sigurdson etal. 1984), L G IV (Rogalski etal. 1982, 1985; Clark etal. 1988), L G V (Rosenbluth etal. 1988; Johnsen and Baillie 1991), and L G X (Meneely etal. 1979, 1981). Lethal screening is an effective and comprehensive method of identifying a wide variety of genes, which can then be analyzed to discern their developmental expression patterns, their functional roles in development, and their possible relationships with other loci. Extensive lethal analysis has been carried out in other eukaryotic organisms, most notably Drosophilamelanogaster (Hochman 1971; Shannon etal. 1972; Leicht and Bonner 1988). Mutational analysis provides valuable information on the organization of the genome by examining the distribution of essential genes over defined regions, as well as providing a valuable resource for the in vivo investigation of the biological function of these genes. The genetic, physical, and cell lineage maps, as well as the DNA sequence of the genome, provide a framework for the developmental analysis of a model organism, provided the information can be integrated. The integration of genetic and physical maps provides a powerful biological tool, because genetic markers (such as RFLPs) and genes become unified on a single physical framework. Genetic map data from various sources can be linked on a physical map, thereby positioning markers that were not genetically mapped against one another. Furthermore, the positional cloning of known genes will aid in correlating the biological function of a gene with its nucleotide sequence. This is especially important for elucidating the function of sequenced genes that have no significant similarity to those in current databases. The genetic and physical links provide convenient guideposts, to which new mutations can be quickly mapped and cloned. These links also enable a direct comparison of genetic and physical map data, yielding valuable information on the structural and functional organization of a genome with regards to gene density, sequence organization, and recombination. The importance of integrating the genetic and physical maps so as to increase the potential use of the forthcoming sequencing information has been stressed by every genome project initiated (Dietrich et al. 1995; Goodman et al. 1995; Guyer and Collins; Olson 1995; Spradling etal. 1995; Waterston and Sulston 1995). The integration of the genetic and physical maps has been approached by various means. A common approach has involved the use of various polymorphic markers, such as restriction fragment length polymorphisms (RFLPs). RFLPs result from DNA sequence variation between individuals. In mice, this has been enhanced by interspecific crosses, which can maximize polymorphisms in the progeny (Dietrich etal. 1995). The mouse genome project has also made use of simple sequence length polymorphisms (SSLPs), based on microsatellite repeats that can be amplified by PCR (Cornall etal. 1991). Yeast artificial chromosomes (YACs) corresponding to each SSLP on the genetic map can then be isolated to integrate the maps. Similarly, in Arabidopsis, RFLPs that have been genetically mapped can be fingerprinted and integrated into the physical cosmid map (Goodman etal. 1995; Schmidt etal. 1995). In the human genome project, DNA polymorphisms represent a major contribution to the integration of the genetic and physical maps. Both RFLPs and microsatellite repeats are used as a direct means of linking the genetic map with the physical contig map. In addition, sequence-tagged sites (STSs) developed from cosmids can be localized along a chromosome by PCR screening, thereby placing these markers simultaneously on both the genetic and physical maps (Chumakov etal. 1995; Collins et al. 1995; Doggetteta/. 1995; Gemmill etal. 1995; Krauter etal. 1995). Transposable elements have been instrumental in creating RFLPs, as well as other markers for integrating the genetic and physical maps. In C. elegans, RFLPs can be created by the insertion of the transposable element Tel , and the number of such polymorphisms can be increased by crossing subspecies that carry different numbers of Te l elements (Emmons et al. 1983; Rose etal. 1982). Six such RFLPs have been positioned within the gene cluster on chromosome I (Starr etal. 1989), thereby providing valuable genetic and physical guideposts. Tcl-induced RFLPs have also been used to generate sequence-tagged sites for mapping (Williams et al. 1992). In Drosophila, researchers have made use of the P transposable element to create mutations for integrating the maps. P element-mediated mutation employs genetically engineered P elements to disrupt open reading frames. The DNA flanking the site of insertion is then sequenced so as to place a series of genetic markers on the physical map, as well as to correlate the biological function of a gene with its sequence (Spradling etal. 1995). In contrast to the approach of linking the maps by using transposons to induce mutations, the maps can also be linked by generating transgenic strains carrying cosmid/plasmid DNA to 'rescue'genetic mutations (Fire 1986; Mello etal. 1991). This technique involves the phenotypic rescue of a lethal or visible mutation by the activity of a wild-type gene on a cosmid/plasmid that has either been integrated into the genome, or has been incorporated into an extrachromosomal array. In a similar fashion, mouse geneticists use Y A C s or BACs (bacterial artificial chromosomes) for the positional cloning of genes by phenotypically rescuing mutations via transgenesis (Dietrich et al. 1995). Such phenotypic rescue experiments integrate the genetic and physical maps by associating the rescue of a genetic mutation with a particular DNA clone. In addition, the sequence of the transgenic DNA can be linked to the function of the rescued genes. A mutant library provides a valuable tool for discerning the biological roles of essential genes, and ultimately correlating this information with the nucleotide sequence currently being generated (Wilson et al. 1994; Waterston and Sulston 1995). This study involves a subset of lethal mutants from a 'mutant library' covering the left third of chromosome I in C. elegans, which were recovered using the duplication sDp2 (Rose etal. 1984; Howell etal. 1987). These lethal mutations lie in a defined region (1.5 map unit) between the visible markers dpy-5 and unc-13 in the middle of the gene cluster. Before the genetic and physical maps could be integrated, several of these lethals first had to be placed on the genetic map, so as to complete the mapping of all 495 lethal mutations with respect to the dpy-5 and unc-13 genetic markers. Genetic mapping involved the use of a series of duplications with breakpoints within the defined region. A subset of these lethals were characterized by Nomarski microscopy and DAPI-staining of their nuclei, so as to improve the defini t ion o f their mutant phenotypes. T h i s subset consisted o f a l l the lethal mutants l y i n g between dpy-5 and unc-13 that arrested i n development as sterile adults. F i n a l l y , the genetic and phys ica l maps were integrated over a 0.5 map unit region between dpy-5 and bli-4 by the serial c o s m i d rescue o f le thal mutat ions. C o s m i d and p l a s m i d D N A ca r ry ing a genetic marke r was introduced into non-lethal strains by microin jec t ion ( M e l l o etal. 1991). These transgenic animals were then genetically crossed to a series o f lethal mutations w i th in the region o f interest, p rov id ing a r ap id method o f sys temat ica l ly test ing several mutant strains against c o s m i d D N A wi thout repeatedly having to inject animals. C o s m i d rescue experiments integrate the genetic and phys ica l maps by p lac ing genetically mapped lethal mutations on the c o s m i d cont ig map. These mutations are then rendered accessible for c lon ing , w h i c h w i l l u l t imately lead to the correlat ion of b io log i ca l funct ion w i t h the u p c o m i n g sequence for ch romosome I ( H o d g k i n et al. 1995; Waters ton and Sulston 1995). In addi t ion, mutations i n new and k n o w n genes can be mapped by recombina t ion relative to the rescued lethals, a l l o w i n g the mutations to be p laced wi th in specific phys ica l intervals defined by these convenient guideposts. A s more lethals f rom the mutant l ibrary are placed on the genetic and phys ica l maps, the density o f anchored l o c i a long the phys ica l map w i l l increase, and w i t h it the power and p rec i s ion o f genetic and p h y s i c a l m a p p i n g methods. Consequent ly , the placement o f these lethals on both the genetic and phys ica l maps considerably increases the value of this mutant l ibrary , and expands their future role in a large scale funct ional analysis o f genes essential for the growth and development o f C . elegans. 8 CHAPTER 1: Genetic Analysis of Essential Genes in the hDpl3-hDpl6 Region INTRODUCTION Lethal mutants are important for investigating many aspects of Caenorhabditis elegans development. A collection of mutant alleles for each essential gene provides a resource for studying its biological function. The Rose lab has carried out a large-scale screen directed at identifying essential genes within the left third of chromosome 1 in C. elegans (Howell et al. 1987), a 15 map unit region defined by the breakpoints of the free duplication sDp2 (Rose et al. 1984). A region of this size should carry a wide variety of chromosomal loci, representing most classes of genes and regulatory sites. Within this region, visible genes are unevenly distributed along the genetic map, with most genes being clustered together near the center of the chromosome around unc-13, leaving the remainder of the chromosome sparsely populated. This clustering is due to both recombination suppression (Kim and Rose 1987), and to a higher gene density within the gene cluster (Waterston etal. 1992). Barnes etal. (1995) have derived distinct boundaries for each of the autosome gene clusters, based on recombination rates and physical map data. They determined the chromosome I gene cluster to be bounded by unc-73 on the left, and lin-11 on the r i g h t The dpy-5 unc-13 region analyzed in this study lies in the center of this gene cluster. In an effort to identify essential loci within the left third of chromosome I, Howell et al. (1987) induced lethal mutations using 12 mM ethyl methane sulphonate (EMS). In total, 495 recessive lethal mutations were recovered with the use of sDp2 as a genetic balancer. sDp2 has been useful for the maintenance of recessive lethal mutations, because crossing-over does not occur between the duplication and the normal homologs. The 15 map unit sDp2 region is estimated to have 225 essential genes using a Poisson calculation (McDowall 1990), of which a large fraction have been identified in this mutational analysis. As such, this 'mutant library' represents a substantial collection of mutant alleles that can be used in a large-scale functional analysis of C. elegans essential genes. The initial step in characterizing these lethal mutations was to divide them into complementation groups and to position them on the genetic map. Through a collaborative effort, including this study, at least half of the mutations were assigned to complementation groups, and were positioned on the genetic map using deficiencies, duplications, and recombination analysis (Howell etal. 1987; Howell and Rose 1990; McDowall 1990; Peters etal. 1991; M c K i m etal. 1992). The duplications consisted of a set of 62 gamma-induced derivatives of the duplication sDp2, generated by McKim and Rose (1990). The derivations behaved as free duplications, and have no influence on the recombination frequency between the normal homologs in the duplicated regions. These duplications were efficient tools for mapping lethal mutations. The duplication and deficiency breakpoints used in this study and in previous studies (McDowall 1990; McKim et al. 1992), divided the dpy-5 unc-13 region into 25 intervals. This chapter describes the mapping and complementation testing of 40 recessive lethal mutations lying between the breakpoints of the duplications hDpl3 and hDp!6 within the dpy-5 unc-13 region. Seventeen of the 40 lethal mutations identified six new essential genes. The characterization of lethal mutations with respect to their developmental arrest suggested that the distribution of arrest phenotypes in this region was nonrandom. The positioning of these lethal mutations on the genetic map represented the first step in integrating the genetic and physical maps with respect to these essential genes, thereby rendering them available for cloning and functional analysis. 10 Chapter 1: MATERIALS AND METHODS Stock Maintenance and Strains: Wild-type (N2) and mutant C. elegans strains were maintained and mated on petri plates containing nematode growth medium (NGM) streaked with Escherichia coli OP50 (Brenner 1974). Strains were usually maintained as self-fertilizing hermaphrodites (5AA;XX). Males (5AA;XO) arise spontaneously as a result of X chromosome nondisjunction (Hodgkin et al. 1979). Al l experiments were carried out at 20°C (Rose and Baillie 1979). Matings were done by placing eight to ten hermaphrodites on a petri plate with 15 to 20 males, which were then left at 20°C for 24 hours. After the mating, the hermaphrodites were transferred individually to fresh plates. In C. elegans, male sperm is used preferentially over the hermaphrodite's own sperm (Ward and Carrel 1979). The mutants studied in this thesis are listed by their strain name and genotype in Appendices 1 and 2. A genetic map of these genes prior to this study is shown in Figure 1. The lethal mutations on L G I used in this study were induced with 12, 15 or 17 mM ethyl methane sulphonate (EMS) and recovered using the .sDp2-lethal rescue system as described by Howell et al. (1987). The following visible mutations were used: dpy-5(e61)I, dpy-14(e188)1, unc-1 l(e47)I, unc-13(e51)I, bli-4(e937)I and lon-2(e678)X. The following sDp2-derived duplications were used: hDpl2, hDpl3, hDplS, hDpl6, hDpl7, hDp37, hDp41, hDp61 and hDp72 (McKim and Rose 1990). The translocation szTl(I;X) of Fodor and Deak (1985) was used in a strain with the genotype unc-11 dpy-14/szTl(I;X)[+ + ,ion-2] (McKim etal. 1988). The names of genetic loci are abbreviated using three italicized lower case letters followed by a number, and refer to a detectable phenotype. For example: let (/emal), dpy (dumpy), unc (uncoordinated), bli (Mstered) and Ion (long). Alleles are assigned one or two lower case italicized letters, to indicate the laboratory of origin, followed by a number, for example h293. The allele name can be used alone, or in parentheses after the gene name. Strain names consist of two 11 or three non-italicized uppercase letters, to indicate the lab of origin, followed by a number. The designations for Ann Rose's lab are 'KR' for strains and 'h' for alleles. Phenotypes of mutant worms are indicated by using the gene name without italics, with the first letter in upper case. Chromosomal rearrangements are abbreviated as follows: 'Df for deficiencies, 'Dp' for duplications, 'T for translocations, 'Ex' for extrachromosomal arrays and 'Is' for integrated sequences (extrachromosomal DNA that has integrated into the genome). Such rearrangements are named by listing the lab mutation name, followed by the abbreviation and a number, for example hDpl6. Chromosomal Rearrangements Duplications: All the duplications used in this study are free duplications of the left third of chromosome I. sDp2 was generated in D. Baillie's lab at Simon Fraser University (Rose et al. 1984). The duplications with the 'h' allele designation were generated and mapped with respect to visible markers by K. McKim (McKim etal. 1988, 1990). They were generated by breaking down sDp2 with gamma rays. These duplications carry wild-type alleles for the farthest left known markers on chromosome I. However, they vary in their right end breakpoints, thereby creating a series of endpoints within the dpy-5 bli-4 region. The breakpoint of sDp2 lies between dpy-14 and unc-13, therefore it carries the wild-type alleles of dpy-5 and dpy-14, but not of the unc-13 locus (Rose etal. 1984). The breakpoints of hDp!3, hDpl5, hDp37, hDp41 and hDp72 were mapped to lie between the visible markers dpy-5 and unc-40. These duplications carry wild-type alleles of dpy-5 and the markers to the left of dpy-5, but not of the unc-40 locus. The hDpl2, hDpl6, hDpl7 and hDp61 breakpoints lie between unc-40 and bli-4. The extent of all the duplications used are shown in Figure 1. Translocations: The translocation szTl (I;X) was induced on a lon-2(e678)(X) chromosome using 7000R of X-radiation (Fodor et al. 1985), and was isolated as a dominant X chromosome crossover suppressor. The szTl chromosome is composed of two abnormal 12 chromosomes derived from the normal chromosomes I and X. Animals homozygous for the szTl chromosome arrest as embryos. Fodor et al. showed that when heterozygous hermaphrodites are allowed to self-cross, they produce Lon-2 males at the elevated frequency of 0.08-0.12%, and that recombination was reduced in the dpy-7 unc-3(X) interval from 19.2 map units to 0.3 map units. The strain of szTl used in this study was obtained from K. McKim, who introduced the markers unc-11 and dpy-14 onto the chromosome I portion of the szTl chromosome (McKim etal. 1988). Complementation Analysis of Lethal Strains: Forty lethal mutations previously mapped between the duplication breakpoints of hDpl3 and hDpl6 (McDowall 1990) were tested for complementation to all the known essential genes in the same region. Complementation analysis was carried out in order to identify previously undescribed essential genes, new alleles of known essential genes, and deletions (failure to complement more than one gene). Tests were performed by crossing the sDp2-balanced lethal-bearing strains (dpy-5 let-x unc-13; sDp2) to N2 wild-type males to generate heterozygous males lacking sDp2 (dpy-5 let-x unc-13l+ + +), then crossing these heterozygous males to sDp2-balanced hermaphrodites bearing other lethal mutations (dpy-5 let-y unc-13; sDp2). The progeny were screened for Dpy-5 Unc-13 males and fertile Dpy-5 Unc-13 hermaphrodites (dpy-5 let-x unc-13ldpy-5 let-y unc-13), their presence indicating complementation (Figure 2). let-377, let-393, let-512, let-513, let-514, let-531, let-532 let-370 let-599 let-355 let-367 let-530 let-395 let-376 let-378 let-379 dpy-5 I I I I I I I let-388 let-603 let-384 let-391 unc-40 bli-4 I unc-13 I WplS Wpl3 hDp57 • . hDP48 hDp41 hDp72 hDp37 hDp!7 hDp6I hDpll hDp!4, Wpl6, HDpl8, HDp32 Figure 1: Genetic Map of Essential Genes lying between the breakpoints of hDp!3 and hDp!6 in the dpy-5 unc-13 Region. L O F igure 2 : Protocol for Complementation Analysis + + dpy-5 let-x unc-13 dpy-5 let-x unc-13 Unc-13 hermaphrodites (carrying sDp2) dpy-5 let-x unc-13 dpy-5 let-z unc-13 Dpy-5 Unc-13 + + + + Wild-type males Pick Wild-type males dpy-5 let-x unc-13 ± + + X + + dpy-5 let-z unc-13 dpy-5 let-z unc-13 Unc-13 hermaphrodites (carrying sDp2) dpy-5 let-x unc-13 ± + + 'Wild-type' + + + dpy-5 let-z unc-13 'Wild-type* + + dpy-5 let-x unc-13 dpy-5 let-z unc-13 Unc-13 + + dpy-5 let-x unc-13 + + + 'Wild-type' + + + + + dpy-5 let-z unc-13 'Wild-type' 15 Duplication mapping: Duplications can be used to map recessive mutations by determining whether or not a wild-type copy of a given gene exists in a duplication strain. Seven essential genes previously mapped between the breakpoints of hDpl3 and hDpl6 (McDowall 1990) were positioned with respect to the breakpoints of the duplications hDpl2, hDpl7, hDp37, hDp41, hDp61 and hDp72. Duplication mapping was done following the protocol in Figure 3. sDp2-balanced lethal-bearing hermaphrodites (dpy-5 let-x unc-13; sDp2(I;f)) were crossed to unc-11 dpy-14;szTl(I;X)[lon-2J males, where let-x represents the unmapped lethal mutation. The (I;X) translocation in szTl males was useful for producing males of the required genotype to cross to hermaphrodites carrying the new duplication. szTl(I;X) males carry a normal chromosome I, but the only X chromosome they carry is rearranged in a (I;X) translocation. All the progeny receiving the paternal chromosome I will be male, while those receiving the paternal (I;X) rearrangement will be hermaphrodites. As a result, all males derived from the above cross came from unc-11 dpy-14;0 sperm (McKim etal. 1992), and were wild-type in phenotype with a normal X chromosome and heterozygous for markers on L G I. These males (+ dpy-5 let-x + unc-13/unc-ll + + dpy-14 +) were crossed to dpy-5 dpy-14; hDpz(I;f) hermaphrodites, where hDpz represents the duplications used for mapping. All the duplications used were dpy-5(+) dpy-14(-), and were Dpy-14 in phenotype. From this second cross, only the wild-type outcross progeny would carry both the new duplication and the lethal mutation, and were of the genotype (dpy-5 let-x + unc-13/dpy-5 + dpy-14 +;hDpz(l;f)). These wild-type hermaphrodites were allowed to self-fertilize, and their progeny were screened for the presence of viable, fertile Unc-13 animals, indicating that the duplication carried the wild-type allele of the lethal being mapped. To ensure that the fertile Unc-13 worms had not lost the lethal mutation from recombination, their progeny were screened for the presence of fertile Dpy-5 Unc-13 worms. Since Dpy-5 Unc-13 worms are homozygous for the lethal mutation, but do not carry the duplication, they should arrest in their development. The presence of fertile Dpy-5 Unc-13 worms would therefore suggest that the lethal mutation had been lost. Figure 3: Protocol for Duplication Mapping of Lethal Mutations dpy-5 let-x unc-13 I dpy-5 let-x unc-13 ; sDp2 X unc-11 dpy-14 ; szTl Unc-13 hermaphrodites Lon-2 males i Pick Wild-type males dpy-5 let-x unc-131 unc-11 dpy-14 X dpy-5 dpy-141 dpy-5 dpy-14 ; hDpz Dpy-14 hermaphrodites (carrying new duplication) Pick Wild-type hermaphrodites dpy-5 dpy-141 dpy-5 let-x unc-13 ; hDpz I self-cross Check for presence of fertile Unc-13s Expected Ratio dpy-5 let-x unc-13 I dpy-5 let-x unc-13 Dead 1 dpy-5 let-x unc-131 dpy-5 let-x unc-13 ; hDpz Unc-13 1 dpy-5 let-x unc-131 dpy-5 dpy-14 Dpy-5 2 dpy-5 let-x unc-131 dpy-5 dpy-14 ; hDpz 'Wild-type 2 dpy-5 dpy-141 dpy-5 dpy-14 Dpy-5 Dpy-14 1 dpy-5 dpy-141 dpy-5 dpy-14 ; hDpz Dpy-14 1 17 Chapter 1: RESULTS Complementation Analysis of Lethal Strains: Complementation tests were carried out for the 40 lethal mutations positioned in the hDp!3-hDpl6 region (results in Appendix 3). Each lethal mutation was complementation tested to the 11 known essential genes identified between the same breakpoints. Complementation testing identified six new essential genes, represented by 17 lethal mutations (Table 1). The remaining 23 lethal mutations were alleles of nine previously identified essential genes. There are now a total of 17 essential genes mapping between the breakpoints of hDp!3 and hDpl6, 12 of which are represented by more than one allele (Table 2). One lethal mutation, h823, failed to complement two essential genes, let-376 and let-391, and was therefore considered to be a double hit. h823 was not considered to be a large deletion, since it complemented all the essential genes mapping between let-376 and let-391. h823 was assigned a second allele number, hi 175, to reflect the second mutation. The genotype of this mutant is dpy-5 let-376(h!175) let-391(h823) unc-13; sDp2. Reciprocal complementation tests for each allele within a complementation group were also carried out. In every case, all the alleles within a given complementation group failed to complement each other. 18 Table 1: Newly Identified Essential Genes in the hPpl3-hDpl6 Region Strain Allele New Gene Name KR1440 1x797 let-512 KR1393 h752 let-513 KR1394 h753 let-514 KR1441 h798 let-530 KR1374 h733 let-531 KR1339 H715 let-532 19 Table 2: Essential Genes in the hDpl3-hDpl6 Region G e n e A l l e l e Stage of Arres t G e n e A l l e l e Stage of Arres t let-367 M19 Sterile A d u l t let-393 h225 E a r l y larval h375 Ear ly larval let-376 hl30 E a r l y larval h416 E a r l y larval let-395 h271 Sterile A d u l t h764 Ear ly larval h804 Ear ly larval let-512 h797 M i d larval h805 . Ear ly larval h351 M i d larval h832 Ear ly larval h362 M i d larval h870 Ear ly larval h510 M i d larval hi 175 Ear ly larval h741 M i d larval h808 N o t determined let-377 hi 10 E a r l y larval h433 Ear ly larval let-513 h752 Sterile A d u l t h766 E a r l y larval let-514 h753 Late larval let-378 hl24 M i d larval h401 Ear ly larval let-530 h798 Ear ly larval h359 Ear ly larval let-379 M27 Ear ly larval h677 E a r l y larval K712 Ear ly larval h775 M i d larval h794 E a r l y larval let-384 h84 Sterile A d u l t h828 E a r l y larval h388 Sterile A d u l t h454 Sterile A d u l t let-531 h733 Late larval h865 Sterile A d u l t h767 Sterile A d u l t let-388 h88 Ear ly larval let-532 h715 E g g h729 Sterile A d u l t h843 Sterile A d u l t let-603 h289 Sterile A d u l t h408 Sterile A d u l t let-391 h91 Sterile A d u l t h437 Sterile A d u l t h475 E a r l y larval h812 Sterile A d u l t h736 Ear ly larval h823 M i d larval Stage o f arrest previously determined by J . M c D o w a l l (1990). 20 Mapping L e t h a l Mutations: A linear array of duplication breakpoints with one end in common provides an effective means of mapping genes, since their breakpoints allow one to subdivide the region of interest into very precisely defined intervals. Eight duplication breakpoints have been found to lie between the breakpoints of hDpl3 and hDpl6, dividing this region into nine intervals (J. McDowall 1990; S. McKay, unpublished results). Ten of the 17 essential genes in this region were previously mapped to these eight duplication breakpoints by S. McKay (unpublished results). In this study, seven essential genes mapping between the breakpoints of hDpl3 and hDpl6 were tested by complementation against the duplications hDpl2, hDpl7, hDp37, hDp41, hDp61 and hDp72. The results of the duplication mapping are summarized in Table 3 and Figure 4. This duplication mapping completes the genetic map of all the lethal mutations generated in the Rose lab that lie in the dpy-5 unc-13 region. 21 Table 3: Duplication Mapping Results Gene(allele) thDpl3 hDp41 hDp72 hDp37\hDpl7\hDp61 hDpl2 1hDpl6 let-377(110) let-393(225) Iet-513(h752) *OUT OUT OUT OUT OUT [..OUT OUT OUT I N •w www www I N OUT J I N | I N I N 1 I N i \ ;vvvvvvvvvvvvv "^''"'"'""''','"""'"|WWW\W\'vW^ L°UL. I 2 I N 1 I N j I N 1 I N _ J N _ | J N _ OUT j I N I N let-514(h753) OUT OUT OUT | OUT I N I N | I N let-532 (h715) OUT j I N I N ^ J N _ | j N l j>L_ I N | I N | I N JUN[ l l N mmrnd I N t Results for the duplications hDpl3 and hDpl6 previously determined (McDowall 1990). *0UT signifies that the duplication does not complement the lethal mutation (mutation not covered by the duplication), and I N indicates complementation. let-370 let-599 dpy-5 let-531 , let-532 let-355 let-367 let-530 let-395 let-376 let-378 let-393 let-513 let-379 let-377 unc-40 I let-388 let-514 let-603 let-384 let-391 let-512 let-396 let-602 let-611 let-380 let-381 let-387 let-601 let-606 bli-4 I KDplS hDp57 , - Wp48 hDp41 HDp72 HDp37 hDp!7 hDp61 KDpl2 hDp!4, hDp!6, hDpl8, hDp32 HDp31 Figure 4: New Genetic Map of Essential Genes in the hDpl3-hDpl6 Region. N> 23 Chapter 1: DISCUSSION The goal of this study was to describe the distribution of essential loci in a region of the gene cluster on L G I. This chapter described the analysis of essential genes in a region of the C. elegans genome, as defined by the right breakpoints of the chromosomal duplications hDpl3 and hDpl6, lying between the visible markers dpy-5 and unc-13 on chromosome I. I examined 40 from a set of 495 EMS-indiiced lethal mutations described in Howell etal. (1987, 1990) the results defined six new essential loci, and 23 new alleles of nine previously described genes. Duplication mapping was used to position seven essential loci on the genetic map. This study completed the mapping and complementation testing of the set of sZ)/?2-rescued lethal mutations identified in this lab that lie in the dpy-5 unc-13 region. This work represented the initial step in using these lethal mutations to align the genetic and physical maps in this region. Identification of Essential Genes: From this analysis, six new essential genes were identified in the 0.5 map unit region between the right breakpoints of the duplications KDpl3 and hDpl6. There were 11 essential genes in this region identified in previous studies (Howell et al. 1987; McDowall 1990). This brings the total number of identified essential genes in the hDpl3-hDpl6 region to 17, as defined by the sDp2 rescue of lethal mutants. In addition, 23 new alleles were identified for nine known essential genes in this region. As a result, 12 essential genes are now identified by more than one mutant allele. Seven of these essential genes were positioned with respect to six duplication breakpoints. This analysis completed the mapping of all the sDp2-rescued, lethal mutations generated in the Rose lab that lie in the 1.5 map unit region between dpy-5 and unc-13 in the gene cluster on chromosome I. There are a total of 64 essential loci, including bli-4 and unc-37, identified 24 between dpy-5 and unc-13, six identified in this study and 58 identified in previous studies (Brenner 1974; Rose and Baillie 1980; Howell etal. 1987; McDowall 1990; Peters et al. 1991; McKim etal. 1992; D. Pilgrim, unpublished results), as shown in Table 4. 39 (61%) of these essential genes are identified by more than one mutant allele. Six duplication breakpoints were used to yield a high resolution genetic map. In combination with previous studies (McDowall 1990; McKim et al. 1992; S. McKay, unpublished results), there are now a total of 23 duplication breakpoints and two deficiency breakpoints positioned between dpy-5 and unc-13, which subdivide this region into 25 intervals. On average, they were found to have breakpoints every 1/17 of a map unit, giving a resolution of one breakpoint for every 2.8 genes identified in this region by mutational analysis (Figure 5). Mapping lethal mutations using an array of duplications was found to be an effective method, suitable for positioning large numbers of lethal mutations. sD/*2-rescued Recessive Lethals: Using sDp2 as a balancer (Howell etal. 1987; Howell and Rose 1990; McDowall 1990; McKim et al. 1992), the Rose lab recovered 495 lethal mutations on L G I from 31,600 chromosomes mutagenized with EMS (using dose in the range of 12 to 17 mM). The data set can be used to compare intervals since the mutations were isolated in similar screens using the same balancer. Although the dpy-5 unc-13 interval (1.5 map units) is genetically a small fraction of the sDp2 region, which is at least 17 map units in length (Howell etal. 1987), 38% (188) of the lethal mutations have been mapped here (Howell etal. 1987; McDowall 1990; Peters et al. 1991; McKim etal. 1992; D. Pilgrim, unpublished results). By comparison, 11% (54) of the lethal mutations have been positioned to the 1.5 map unit interval deleted by hDf6, to the left of dpy-5 (Howell and Rose 1990). The dpy-5 gene is located near the center of chromosome I in a recombinationally-suppressed, gene-rich region of the chromosome (Kim and Rose 1987; Starr et al. 1989; Waterston et al. 1992; Barnes etal. 1995). This density of genetic information is reflected in the number of mutations identified by the Rose lab screen. In the 1.5 map unit region to the left of dpy-5, 19 genes are represented by 54 mutations, whereas in the 1.5 map unit region to the right, 64 genes (including bli-4 and unc-37) are represented by 188 mutations. Thus, this analysis would indicate that there are three times as many essential genes in the same size genetic interval to the right of dpy-5 as to the left. This could reflect a difference in recombination rates and/or a difference in the physical gene density of the two regions. Kim and Rose (1987) demonstrated that the extent of recombination suppression varied across the chromosome I gene cluster, by measuring the increase in recombination frequencies of different intervals after treatment with gamma radiation. In the unc-11 dpy-5 interval (left of dpy-5) a 1.5-fold increase was observed, while in the dpy-5 dpy-14 and dpy-14 unc-13 intervals (right of dpy-5) two-fold and five-fold increases were observed, respectively. Similarly, Starr etal. (1989) showed that the variation in the amount of DNA per map unit varied across the chromosome I gene cluster, reaching a peak between dpy-14 and unc-13. These studies indicate that recombination suppression is greatest to the right of dpy-5, between dpy-14 and unc-13, which is consistent with the results obtained in this study. Forward Mutation Rates and Lethal Saturation: The average hit frequency for the sDp2 lethals was 2.9 mutations per gene. Since the screen was unbiased in regards to the position of the lethal mutations recovered, the frequency of mutational hits should have been the same across the entire sDp2 region. This was indeed what we found, since the same frequency was obtained for both the /?D/6-deleted region (Howell and Rose 1990), and for the dpy-5 unc-13 region. The most frequently hit genes in the sDp2 region were let-385 (18 mutations, or 5.7 x 10"4 mutations/gene) and let-354 (17 mutations, or 5.4 x 10~4 mutations/gene), the latter being identified independently by mutations causing dominant temperature-sensitive maternal-effect lethality (Mains etal. 1990; Howell and Rose 1990). bli-4, originally described because of a visible morphological defect resulting in blistering of the cuticle 26 (Brenner 1974), was a moderately large target in the sDp2 screens, having nine alleles (Peters et cd. 1991), or 2.8 x 10 4 mutations/gene. These genes may be favored targets for EMS mutagenesis because of the physical size of these loci, or because of increased sensitivity to EMS. Since EMS alkylates guanine residues causing GC to A T transitions, the GC content of a gene could affect its susceptibility to EMS mutagenesis. Considering the data from the /zD/5~-deleted region and the dpy-5 unc-13 region, 242 mutations representing 83 genes were recovered from 31,600 chromosomes, giving a mutation frequency of 9.2 x 10s mutations/chromosome. This frequency is similar to that of Clark and Baillie (1992), who recovered lethals on chromosome IV using the translocation balancer nTl (8.7 x 10 5 at 25 mM EMS), but lower than that of Brenner (1974) (50 x 10 5 at 50 mM EMS), even after adjusting for the difference in mutagen dose (Rosenbluth etal. 1983). Based on forward mutation rates and Poisson calculations (McDowall 1990), the sDp2 region was considered to be 80% saturated for mutations in essential genes. By extrapolation from the Poisson calculation, the 15 map unit sDp2 region was estimated to have 225 essential genes (McDowall 1990). The C. elegans sequencing project (Waterston and Sulston 1995) estimated there to be 13,500 genes in the C. elegans genome. Since each of the six chromosomes are believed to carry approximately equal numbers of genes (Barnes et al. 1995), the sDp2 region could contain up to 1000 genes. If the Poisson estimate of 225 essential genes was accurate, then only a quarter of the genes in this region were essential. It is possible that the Poisson calculation was too low. This calculation attempted to estimate the number of zero hit genes based on the number of genes hit once, twice, thrice, and so on. However, it assumed each gene was an equal target for mutagenesis, which is untrue. Furthermore, even though frequently hit genes were omitted from the calculation, moderately hit genes could still have caused an overestimate of the frequency of hits obtained. Alternatively, this could have been an accurate reflection of the number of essential genes in this region. In yeast, the majority of open reading frames identified by the sequencing of chromosome III did not appear to encode essential genes (Oliver et al. 1992). Only 6% of the 55 open reading frames analyzed by gene disruption encoded essential genes, compared 27 to 76% non-essential genes. When mutated, 33% of these non-essential genes gave rise to visible phenotypes and 67% were silent. The number of open reading frames appears to have exceed the number of mutable loci for yeast (Dujon et al. 1994), C. elegans (Sulston et al. 1992), and Drosophila (Bossy etal. 1984; Kozlova etal. 1994). Mutations in these silent genes may cause subtle phenotypes that are not easily detected during screening. Alternatively, their function could be carried out by the activity of related genes, thereby masking their mutant phenotypes. Park and Horvitz (1986) estimated that up to 50% of C. elegans genes may be wild-type when mutant. Some members of redundant multigene families can be identified by dominant gain-of-function mutations (Park and Horvitz 1986). Consequently, it seems reasonable to assume that essential genes do not account for the majority of loci in C. elegans. However, essential genes do account for the majority of mutable loci. In the dpy-5 unc-13 region, 75% (65/87) of the loci identified by mutation are essential genes. This includes bli-4 and unc-37, both of which have lethal alleles. Similarly, in the region to the left of the dpy-5 unc-13 interval, defined by the deficiency hDf6, 79% (19/24) of the loci identified by mutation are essential genes (Howell and Rose 1990). In contrast, gene disruption experiments on yeast resulted in more visible mutations (25%) than lethal mutations (6%) (Oliver et al. 1992). The difference between the two organisms could reflect a difference in the approaches taken to identify mutable loci, a difference in the gene content of the two genomes, or both. For instance, it is possible that the sDp2 region of chromosome I in C. elegans is more saturated for lethal mutations compared to visible ones, while the gene disruption approach carried out in yeast is unbiased with regards to the types of genes hit. Alternatively, these results could present an accurate image of the two genomes, and as such reflect the greater complexities of a multicellular organism. Developmental Arrest of Lethal Mutations: The lethal blocking stages were previously determined for the majority of essential genes mapping between dpy-5 and unc-13 (McDowall 1990), in order to examine if the time of 28 developmental arrest was a representative indicator of gene function. The completion of the genetic map in the region between the right breakpoints of hDpl3 and hDpl6 enabled a comparison of the arrest stages for the alleles of each complementation group in the dpy-5 unc-13 region. Most genes (26/37, or 70%) identified by more than one allele exhibited the same arrest phenotype for each allele. The similarity in arrest stages suggested that most of the lethal mutations could be nulls, although this assumption must be tested genetically. The consistency between the arrest stages of different alleles was in agreement with the findings described by others (Rose and Baillie 1980; Meneely and Herman 1981; Rogalski etal. 1982; Clark etal. 1988; Rosenbluth etal. 1988; Howell and Rose 1990; Johnsen and Baillie 1991). McKim etal. (1992) demonstrated that most of the lethal alleles in the region around dpy-14 had the same phenotype as a homozygote as they did when heterozygous over a deficiency, indicating that they were amorphs, or nulls (Muller 1932). If these results were representative of most lethal mutations in the sDp2 set, then the majority of lethal mutations may have caused null phenotypes. In these cases, the arrest stages may have coincided with the time in development when the gene products were required. The remaining 30% of essential loci (11/37) identified in the dpy-5 unc-13 region displayed more variability in arrest stages between mutant alleles. Leicht and Bonner (1988) suggested that variability between alleles could be due to linked second site mutations. However, linked second site mutations are likely to be scarce among the mutations in the sDp2 set, because they were generated with a relatively low EMS dose (Howell etal. 1987). To date, only one linked second site mutation has been identified in the sDf>2-balanced lethal mutations identified between dpy-5 and unc-13 (Chapter 1). Furthermore, all the genetic mutations were induced in the same genetic strain (KR235), therefore any phenotypic range is likely to result from intrinsic properties of the mutant alleles. Muller (1932) proposed that later blocking mutations could be hypomorphic alleles that produce a reduced amount of gene product, which might be less severe than null mutations. Both Clark etal. (1988) and McKim etal. (1992) found that later blocking alleles often arrested earlier in their development as hemizygotes compared to homozygotes, suggesting that they were hypomorphic alleles. Rather than a difference in the amount of gene product produced, the 29 differences between alleles could also be due to a difference in their ability to function. Shannon et al. (1972) proposed that the later arresting alleles might have a "leaky" phenotype, the leaky mutants producing a defective gene product that functions above a certain threshold, thereby enabling them to survive to a later developmental stage. Alternatively, genes with alleles that display variable mutant phenotypes may represent more complex loci. In at least one case, there is data attributing the range in phenotypes to tissue specific products produced by alternative splicing. Mutations in bli-4 have a complex pattern of intragenic complementation, that correlates with the variability in phenotype of the mutant alleles (Peters et al. 1991). Mutations in bli-4 affect the forms of at least four, structurally different, transcripts produced by alternative splicing (Thacker et al. 1995). The gene encodes a kex2/subtilisin-like endoprotease sharing sequence identity with mammalian prohormone convertases. Mutations affecting specific bli-4 gene products have been shown to cause different phenotypes. The phenotypes of the mutants range from visible defects, such as blistering of the adult cuticle (resulting from a deletion of the blisterase A isoform), to embryonic lethality. The different phenotypes most likely result from the loss of function of the different substrates that the bli-4 isoforms normally process. Nonrandom Distribution of C. elegans Essential Genes: During the course of this work, there appeared to be a relationship between mutant phenotypes and their map positions. I have examined whether there are regional differences in the stage of lethal arrest across the dpy-5 unc-13 region. In the dpy-5 half of the interval (dpy-5 to the hDpl6 right breakpoint), there are proportionally more mutants surviving to the late larval stage or beyond (50%) compared to early larval mutants (40%). By contrast, in the unc-13 half of the region (hDpl6 right breakpoint to unc-13), there are proportionally more mutants that fail to develop past the early larval stage (64%) compared to late larval mutants (25%). It appears that the distribution of arrest phenotypes across this interval is nonrandom. Such an assumption should be based upon a collection of null mutations. McKim etal. (1992) demonstrated that the majority of 30 sDp2-rescued lethal mutations in the dpy-14 appeared to be nulls based on the similarity in phenotypes between hemizygotes and homozygotes. However, it should be noted that the majority of the sZ)p2-rescued lethals in the dpy-5 unc-13 region have not been tested against a deficiency, and as such are not known to be nulls. Hence, the null phenotype of genes designated as late blocking could block at an earlier stage in development. This region can be compared to other regions in the C. elegans genome for which lethal analyses have been carried out. In the 1.5 map unit region defined by the deficiency HDf6 on L G I, Howell and Rose (1990) showed that there are proportionally more mutants surviving to the late larval stage (47%) compared to early larval mutants (16%). In a 2 map unit region on L G IV, Clark and Baillie (1992) showed that the number of late to early arresting mutants was equal (29% each). Finally, in the 23 map unit region on L G V, Johnsen and Baillie (1991) showed on average there were 56% late arresting mutants to 21% early arresting mutants. If this large chromosomal region is subdivided into smaller intervals of approximately 2.5 map units (similar in size to those above), regional differences in the stages of developmental arrest become apparent. The number of late arresting genes across this region remains fairly constant (56%), but the number of early arresting mutants varies from 9% to 36%. The comparison of these discrete regions suggests that the distribution of arrest phenotypes across the genome is nonrandom. This prediction brings forth a cautionary note, namely that the distribution of phenotypes across a small interval should only be used with circumspection when extrapolating to the entire genome. When the data from these regions are collated, there are a total of 209 essential genes characterized with respect to developmental arrest stages. From this large data set, 13% of essential genes are required for egg development, 37% are required to develop past the early larval stage, 21% are required to develop past the mid larva stage, 10% are required to develop past the late larval stage, and 19% are required for fertility. However, when you look at any individual region, these assumptions can change dramatically. A more accurate estimate of the proportion of essential genes required for development at each larval stage is dependent upon studies encompassing the entire genome. 31 Genomic Organization of Essential Genes: Analysis of the sDp2 set of lethal mutations show that the essential genes display the same clustering on the genetic map as do the nonessential loci (Edgley and Riddle 1993). This clustering is also seen in other chromosomal regions that have been examined with regard to the distribution of essential genes in C. elegans (Sigurdson etal. 1984; Rosenbluth etal. 1988; Howell and Rose 1990; Clark and Baillie 1992; Johnsen and Baillie 1991). The gene cluster is recombinationally suppressed (Kim and Rose 1987), and has a physically high density of genes (Waterston et al. 1992). The density of genes is further enhanced by the grouping of loci within operons, which comprise approximately 26% of the genome (Zorio etal. 1994). It is interesting to note that in Tea Mays, the majority of genes are also narrowly distributed within gene spaces comprising 10-20% of the genome (Carels etal. 1995). However, in contrast to C. elegans, most recombinational exchanges in maize have been found to occur in gene-rich regions (Civardi et al. 1994). Hence, the clustering of loci does not necessitate a corresponding decrease in recombination potential between the loci involved. Rather, C. elegans' chromosomes are unusual for their recombinationally inert gene clusters. In order to correlate the genetic and physical map data, it is important to place these genetically mapped loci on the physical map. 32 Table 4 : Essential Genes in the dpy-5 unc-13 Region Gene No of Stage of Gene No. of Stage of Gene No. of Stage of (allele) alleles arrest (allele) alleles arrest (allele) a lleles arrest IiDpl3-hDpl6 Region hDp!6-hDpl9 Region hDpl9-s Dp2 Region 3let-367(hU9) 1 Sterile Adult 3let-380(h80) 3 Early larval 2let-86(sl41) 1 Early larval 3let-376(hl30) 8 Early larval 3lel-38](IU07) 3 Early larval 3let-383(h82) 4 Early larval 3let-377(hlL0) 3 Early larval 3lel-382(/i82) 2 Mid larval 3let-385(h85) 18 Early larval 3let-378(hl24) 2 Early larval 3let-386(hll7) 5 Early larval 3let-389(hl06) 10 Early larval 3Iet-379(hl27) 2 Early larval 3let-387(h87) 1 Mid larval 3let-392(hl20) 9 Egg/early 3let-384(h84) 4 Sterile adult 3let-390(h44) i Early larval 6let-394(h262) 7 Early larval 3let-388(h88) 3 Early larval 3let-396(h217) 5 Early larval 6let-397(h228) 3 Early larval 3let-391(h91) 4 Early larval 7let-601(h281) 1 Egg 6let-398(/i257) 1 Early larval 3let-393(h225) 2 Early larval 7let-602(h283) 3 Early larval 6let-399(h273) 1 Late larval 3kt-395(h271) 1 Sterile adult 4let-604(h293) 2 Sterile adult 6let-400(h269) 4 Early larval let-512(h797) 6 Mid larval 4let-606(h292) 1 Sterile adult 6let-520(h690) 1 Mid larval let-513(h752) 1 Sterile adult 4let-607(h402) 1 Egg 6let-521(h704) 1 Late larval let-514(h753) 1 Late larval 4let-608(h706) 3 Early larval 6let-522(h735) 3 Egg/early let-530(h798) 6 Early larval 4let-6I0(h695) 1 Sterile adult 6Iet-523(h751) 2 Hatching let-531(h733) 2 Late larval 4kt-6U(h850) 3 Early larval Het-524(h442) 5 Early larval let-532(h715) 1 Egg l5Mi-4(e937) 11 Hatching 6let-525(h874) 3 Mid larval 4let-603(h289) 4 Sterile adult 6let-527(h207) 3 Early larval 6let-528(hl012) 1 Late larval 6let-529(h238) 4 Early larval 6lel-534(h260) 2 Mid larval 6let-543(h792) 1 Late larval Covered by hDp!3 sDp2 to unc-13 Het-544(h692) 1 Sterile adult 3let-355(li81) 1 Sterile adult 2let-75(sl01) 4 Early larval 6let-545(h842) I Sterile adult 3Iel-370(hl28) 1 Sterile adult 6let-542(h986) 1 Early larval 4lel-605(h312) 5 Sterile adult 4let-599(h290) 1 Sterile Adult Hel-539(h938) 1 Early larval ^2unc-37(e262) 3 Sterile adult The earliest blocking allele was taken for genes with more than one allele. Previously unknown essential genes identified in this study are underlined. 'Brenner 1974 5Peters et al. 1991 2Rose and Baillie 1980 6McKim et al. 1992 3Hovvell etal. 1987 7 D. Pilgrim, unpublished results 4McDowall 1990 1,1-531 , 1,1-532 l,t-37S 1,1-393 lti-380 1,1-381 1,1-384 1,1-396 ltl-387 1,1-391 1,1-602 ltl-601 1,1-608 Ul-386 lrt-3S2 1,1-397, 1,1-398. 1,1-528. 1,1-529. 1,1-534, lrt-544, lrl-545. lrl-605 1,1-385 1,1-394 1,1-399 1,1-521 1,1-599 1,1-355 lrl-367 1,1-530 1,1-395 let-376 lcl-513 lrl-379 lel-377 1,1-314 1,1-603 1,1-312 Ir 1-611 1,1-606 Ut-610 1,1-604 1,1-390 1,1-607 1,1-392 1,1-383 1,1-343 1,1-96 unc-11 dpy-5 unc-40 bb-4 unc-37 unc-87 dpy-14 lcl-389 lri-322 1,1-400 1,1-323 1,1-324 1,1-520 1,1-525 l,t-527 1,1-339 1,1-75 ln-542 1,1-53! srm-4 unc-lS unc-13 hDJX hPpIS hPp!3 hPpSl hDp48 hDp-tl hDp72 hPp!7 hPp!7 hPp6\ hPpIl hPpl4, hDp!6, hPplS. hPp32 Figure 5: hPplI hDp56 hPp39 Genetic M a p of Essential Genes in the dpy-5 unc-13 Region. Underlined genes were identified in this study. Visible genes are in bold typeface. hPp}4 hPp!9 hDp33. hDp35 hPpS8 "I I' hPpSO ;Pp2 hPp60 hPp36, HDp43 hPp64 34 CHAPTER 2: Characterization of Sterile Mutants in the dpy-5 unc-13 (I) Region INTRODUCTION Lethal mutants are important for investigating many aspects of Caenorhabditis elegans development. In particular, sterile mutants facilitate the genetic investigation of germline and gamete development. A collection of alleles for essential genes provides the resource for studying how these genes are organized at both the genetic and molecular levels. In C. elegans, the investigation of germline development and fertilization has demonstrated the existence of mutations affecting meiosis (Mournier and Brun 1980; Clandinin and Mains 1993), spermatogenesis (Edgar and Hirsh 1985; Doniach 1986; L'Hernault et al. 1988; Schedl and Kimble 1988; Ward etal. 1988; Barton and Kimble 1990), oogenesis (Graham and Kimble 1993), germ cell proliferation (Kimble and White 1981; Austin and Kimble 1987; Beanan and Strome 1992), and embryogenesis (Isnenghi etal. 1983; Priess etal. 1987; Wood etal. 1980). In many cases, information about the functions that the genes provide has resulted from cytological analysis of gonads from sterile mutants (Mournier and Brun 1980). Gonad development in wild-type hermaphrodites has been characterized with regards to the four larval stages (L1-L4) (Hirsh etal. 1976; Kimble and Hirsh 1979; Kimble and White 1981), as shown in Figure 6. At hatching, the gonadal primordium consists of two somatic (Zl and Z4) and two germline (Z2 and Z3) precursor cells. Z l and Z4 descendants give rise to the anterior and posterior somatic gonad, respectively. Two descendants (Zl.aa and Z4.pp) become the distal tip cells (DTCs). During the second larval stage, the DTCs begin to migrate, guiding the proliferating 35 germline cells. Their migration creates two reflexed arms by the L4 stage, each with a distal-proximal axis with respect to vulval position. The majority of the somatic cells are located in the proximal region. The gonad reflexions may serve to increase the length of the gonad arms for proper gametogenesis to occur. Mutations affecting DTC migration, such as mig-6 (migration) which results in bulb-shaped ovotestes (Hedgecock etal. 1987), can support normal germline proliferation, but gametes often fail to mature. The majority of the somatic gonad cells remain in the proximal region. During the second larval stage, uterine precursor cells draw together to form the uterus, forcing the germ cells away from the center of the gonad. Two cells in the developing uterus have an equal potential to become the anchor cell (AC), Zl.ppp or Z4.aaa. The other cell becomes a ventral uterine precursor cell (VU). Kimble (1981) used laser ablation experiments to show that the V U cell fate is determined by cell-cell interactions, presumably by the V U cell receiving a signal from the A C via a lin-12 (lineage) receptor (Seydoux and Greenwald 1989). During L2 these cells are in direct contact with germline cells, however by L3 a morphologically distinct A C emerges, completely surrounded by somatic tissue. A diffusible signal from the A C induces the vulval precursor cells to produce vulval cells. During the L3 stage, the A C fuses the uterus to the newly formed vulval cells. By the L4 stage, the vulval cells fuse to form an invagination, with the A C at the top. The A C later disappears to create an open passageway between the uterus and the vulva. Ablation of the A C disrupts vulval invagination and vulval-uterus attachment. Each arm of the adult hermaphrodite gonad consists of six regions: the ovary, the loop, the oviduct, the spermatheca, the spermatheca valve, and the uterus (Figure 6). Germ cell maturation occurs along a distal-proximal axis, with mitotically dividing germ cells at the distal end, and gametogenesis at the proximal end. Germ cells are descended from two precursor cells, Z2 and Z3, located in the two somatic gonad arms. These cells divide mitotically until the L3 stage, when the most proximally located germ cells enter meiosis. During the L4 stage, the most proximal germ cells develop as sperm, while subsequent germ cells differentiate as oocytes. Somatic and germline sex determination result from separate, interconnected pathways that involve many genes 36 in common (Hodgkin 1990; Ellis and Kimble 1995). However, the differentiation of gametes as sperm or oocytes is not influenced by the somatic gonad, but is germline autonomous. The terminal regulator in somatic sex differentiation is tra-1 (Zftmsformed), which encodes a putative zinc-finger protein, and thus may act at the transcriptional level (Zarkower and Hodgkin 1992). The presence of the wild-type tra-1 product enables female development of the soma, while suppression of tra-1 enables male development of the soma. The independence of soma and germline cell fates is apparent in the following mutants: fog-1 (feminization of germline) X O males have a male soma that produce gametes resembling oocytes (Barton and Kimble 1990), while fem-3 (/feminization) gain-of-function X X hermaphrodites have a female soma that produce only sperm (Barton etal. 1987). The fog-1 and fog-3 wild-type gene products are required for germ cells to differentiate as sperm (Barton and Kimble 1990; Ellis and Kimble 1995). Likewise, the wild-type gene product of gld-1 (defective in germ/ine development) is required for germ cells to differentiate as oocytes (Francis etal. 1995a, 1995b). Ellis and Kimble (1995) suggested that the genes of the sex-determining pathway may act on fog-1 and fog-3, instead of gld-1, thereby specifying the spermatogenic cell fate prior to the oogenic cell fate. This regulation involves inhibition of tra-2 activity to enable spermatogenesis (Donaich 1986). Temperature-sensitive mutants of fog-1 indicate that the determination of germ cells as sperm or oocytes occurs as germ cells switch from mitosis to meiosis (Barton and Kimble 1990). The decision between spermatogenesis and oogenesis, as well as between mitosis and meiosis appear to be linked by gld-1, which selects between the oogenic and mitotic cell fates, thereby linking the decisions controlling sexual identity and mitosis (Ellis and Kimble 1995; Francis etal. 1995a, 1995b). The sperm-oocyte switch is controlled by at least 14 genes in the germline sex determination pathway (Hodgkin 1990; Kuwabara and Kimble 1992). Five genes, fem-1, fem-2, fem-3, fog-1, and fog-3, act in concert to repress oogenesis and promote spermatogenesis, with fog-1 and fog-3 acting as the terminal regulators (Ahringer and Kimble 1991; Ellis and Kimble 1995). The subsequent switch to oogenesis involves the post-translational, negative regulation of fem-3, and possibly other fern or 37 fog genes, by tra-2, tra-3, and mog-1 (masculinization of germline), positioned upstream in the pathway (Donaich 1986; Graham and Kimble 1993; Francis etal. 1995a). The three genes tra-2, tra-3, and mog-1 repress spermatogenesis, but they do not specify the oogenic cell fate. Instead, this role is assumed by gld-1, and possibly by other genes (Francis etal. 1995a, 1995b). Ellis and Kimble (1995) have proposed a model, whereby the determination of germ cell fate involves a network of inhibitory interactions between genes specifying different cell fates, so as to ensure that only a single fate is adopted: mitosis, spermatogenesis, or oogenesis. The sperm-oocyte switch is germline autonomous, however germline development involves soma-germline interactions as well. The somatic DTCs maintain mitotic proliferation of adjacent germ cells (Kimble and White 1981) by inhibiting meiosis through cell-cell interactions involving the wild-type glp-1 (germ/ine proliferation) receptor (Austin and Kimble 1987; Yochem and Greenwald 1989). The wild-type glp-1 protein is associated with the membranes of mitotic germ cells, and is spatially restricted to the distal mitotic region (Crittenden etal. 1994). Germ cell proliferation occurs within the distal 25% of each ovary under the influence of a signal emitted from the DTCs. The extracellular domain of the glp-1 protein is believed to receive the DTC signal, while the intracellular domain directs germline mitosis (Roehl and Kimble 1993). This signal is believed to be the transmembrane protein lag-2 (Henderson et al. 1994). Consequently, the terminal regulators of spermatogenesis and oogenesis are both inhibited in this region. As germ cells move progressively further from the DTC, they are no longer influenced by glp-1 activity and can enter meiosis and begin to differentiate (Ellis and Kimble 1995). Ablation of the DTCs, or mutations in glp-1, cause all the germ cells to prematurely enter meiosis. The glp-1 gene product is also required for a cell-cell interaction in the developing embryo for induction of the anterior pharynx (Priess etal. 1987). The maternal requirement for glp-1 protein is believed to be responsible for the different mechanisms of glp-1 protein localization in hermaphrodite and male germlines(Crittenden etal. 1994). Soma-germline interactions are also involved in some aspects of Drosophila (Schupbach 1987) and mammalian (McLaren 1987) germline development. 38 It is equally important for an organism to be able to block inappropriate cell-cell interactions. An interaction between the somatic A C and the germline is blocked by the surrounding somatic gonad cells that express the wild-type lin-12 receptor (Seydoux etal. 1990). The prevention of an AC-germline interaction aids in the determination of germ cell fates. In mutants lacking the lin-12 gene product, the A C can interact with the wild-type glp-1 receptor in the germline causing mitotic proliferation in the uterus. The lin-12 and glp-1 genes are structurally related (Yochem and Greenwald 1989), both sharing homology with the Drosophila neurogenic gene Notch (Wharton et al. 1985). These three genes encode putative transmembrane proteins with extracellular domains that contain EGF-like repeats and cysteine-rich L N G {lin-12/Notch! glp-1) repeats, and cytoplasmic domains that contain cdclO/SW16 motifs. The cdclO/SW16 repeats in the cytoplasmic domain are thought to be involved in signal transduction through protein interactions (Yochem and Greenwald 1989). Both glp-1 and lin-12 are believed to share the same ligand, lag-2, which shares homology with the putative Notch ligand, Delta (Henderson et al. 1994) The cdclO/SW16 motifs are found in two yeast genes, cdclO and SW16, that function during the cell cycle (Breeden and Nasmyth 1987; Nurse and Bissett 1981). The C. elegans sex-determining gene, fem-1, also contains cdclO/SW16 repeats, which forms a structural domain that mediates specific protein-protein interactions (Spence et al. 1990). Notch, lin-12, and glp-1 may be members of a gene family involved in the control of cell fates during development. The wild-type glp-1 and lin-12 gene products may be biochemically interchangeable (Lambie and Kimble 1991). Their different roles in development may be due to their different patterns of gene expression, glp-1 being expressed in germline tissue and lin-12 being expressed in somatic tissue. Gametogenesis begins in the proximal gonad arms during the L4 stage and continues throughout adulthood (Hirsh etal. 1976). Germline nuclei enter meiotic prophase in the ovary of the distal arm, where they undergo synapsis followed by the pachytene stage of meiosis I. Germline tissue in the ovary is syncytial, with the nuclei sharing a common cytoplasmic core, or rachis. The rachis is present at the beginning of the pachytene stage, and communicates with the germ cells through large gaps in the plasma membranes. Gibert etal. (1984) demonstrated that a 39 metabolite transfer occurs from the pachytene germ cells to the rachis, suggesting a trophic role for the rachis as a common nutritive pool during oogenesis. The rachis is necessary for the formation of normal oocytes, since mutants with a reduced or absent rachis show no oocyte production. In the loop region, the germ cells enter the diplotene stage and begin to cellularize. The rachis assumes a central position towards the gonadal loop, with germ cell nuclei arranged peripherally around the gonad. Programmed cell death occurs in the loop region, as evidenced by the accumulation of refractile bodies (M. Hengertner and B. Horvitz, unpublished results). At least a third of the germ cells are believed to undergo cell death. These programmed cell deaths serve to increase the cytoplasmic to nuclear ratio, enabling the growing oocytes to incorporate the large number of ribosomes necessary to sustain later embryonic development. After the oocytes reach the oviduct, they keep their connection to the rachis so that they can continue to grow until they reach the spermatheca. The sheath cells of the oviduct contain muscle filaments that act to push the oocytes towards the spermatheca. Yolk proteins are first expressed during the L4 molt to adulthood (Kimble and Sharrock 1983). Yolk proteins are synthesized in the intestine and transported to the gonad, where they are taken up by mature oocytes in the oviduct. The mature oocytes in the oviduct are arrested at diakinesis of meiosis I, and only complete the two meiotic divisions after fertilization. Fertilization occurs as the oocytes pass through the spermatheca. C. elegans sperm lack flagella. Instead, they use a single pseudopod for movement (Roberts and Ward 1982). Ward and Carrel (1979) showed that male sperm outcompete hermaphrodite sperm by displacing the hermaphrodite sperm from their position in the spermatheca. However, sperm from two different males will compete equally within the spermatheca. Mating can stimulate oogenesis in a hermaphrodite. After fertilization, the zygote must resume meiosis, extrude the polar bodies, then switch to mitotic cell division. Mutations in 22% (14) of the 64, sDp2-rescued, essential genes mapping between dpy-5 and unc-13 result in sterility. This chapter describes the phenotypic characterization of 15 essential genes that mutate to sterility, by the use of Nomarski microscopy and DAPI-staining techniques. Fourteen of these loci map between dpy-5 and unc-13 , and one locus maps to the right of unc-13. 40 The effect of sterile mutations on the development of the gonad and its gametes is described for these essential genes. However, it is not the intention of this study to fully investigate the role of each essential gene in germline development, but rather to further characterize their mutant phenotypes so as to make them more accessible to researchers studying specific classes of mutants. The identification and preliminary characterization of loci involved in different aspects of fertility will facilitate their further investigation, thereby ultimately contributing to our knowledge of C. elegans germline and embryonic development. 41 Proximal ro o toooooooo /oN o o o o o o Adult .Pachytene .Synapsis r / o o o o m u y Ovary Loop I 0 O O 0 O 0 0 0 0 0 0 0 0 0 0 0) i 0 0 0 o o o o 'o\ c 0 s s s s s s s s o 0 o Diplotene Diakinesis Oocyte Egg Uterus Vulva Oviduct Spermatheca Figure 6: Gonadogenesis and Anatomy in Wild-type Hermaphrodites. Morphology of the gonad given for the four larval stages and for adulthood, based on drawings by Kimble and White, 1981. ( p )^ Somatic gonad region; (•) Mitotic germ cells; (o) Meiotic germ cells; (s) Sperm; (D) Distal tip cell; (A) Anchor cell. 42 C h a p t e r 2 : MATERIALS AND METHODS The source of mutations and chromosomal rearrangements is given in Chapter 1: Materials and Methods. Test for Sperm-defective Mutants: To test for sperm-defective mutants, sterile adult hermaphrodites were crossed to wild-type males, thereby introducing wild-type sperm into the hermaphrodite. If the hermaphrodites are producing viable oocytes but are defective in spermatogenesis, then the wild-type sperm should rescue the sterile phenotype, producing viable progeny. However, this test would not detect mutants defective in both sperm and oocyte production. The wild-type male rescue experiment is also useful to distinguish between strict maternal effect lethals (cannot be rescued by wild-type male sperm) and partial maternal effects (can be rescued). If maternal expression of the gene product is strictly required for the viability of the embryo, then the embryos of homozygous mothers will not survive, even if they carry a wild-type allele from their father (Wood etal. 1980; Kemphues etal. 1988). For two lethals, let-544 and let-545, sterile Unc-13 hermaphrodites homozygous for dpy-5 let-x unc-13 and carrying a duplication that does not rescue the lethal were used (dpy-5 let-x unc-13; hDpl3), since mating is more easily accomplished in Unc-13 animals than in Dpy-5 Unc-13 animals. These hermaphrodites were crossed to N2 males and plates were checked for the presence of progeny over a period of one week. For the remaining two lethals, let-370 and let-599, sterile Dpy-5 Unc-13 hermaphrodites homozygous for dpy-5 let-x unc-13 were tested for sperm defects. These mutants did not carry a duplication, because none of the duplications will separate the lethal mutations from dpy-5. These hermaphrodites were mated to N2 males, and the plates were checked for the presence of progeny over a period of one week. 43 Microscopic Characterization of Sterile Adult Mutants: Sterile Unc-13 adult animals, generated as described below, were observed using Nomarski differential interference contrast (DIC) microscopy, and using fluorescence microscopy after DAPI-staining. Steriles were outcrossed before analysis to expose the lethal mutations, as well as to remove any linked mutations that could affect the sterile phenotypes. All the lethal mutations analyzed were linked to dpy-5 and unc-13. However, the gonads in Dpy-5 Unc-13 animals are difficult to examine, since these animals are internally more disorganized. Therefore, where possible, each mutation was placed over a duplication that carried a wild-type allele for dpy-5, but did not cover the lethal mutation, nor unc-73. sDp2-balanced, sterile strains (dpy-5 let-x unc-13; sDpT) were crossed to the duplication strain dpy-5 dpy-14; hDp!3, except for let-355. This strain was crossed to a strain bearing hDpl5, since hDp!3 carries a wild-type allele of let-355. Sterile Unc-13 worms (dpy-5 let-x.unc-13; hDpz) were obtained following the protocol for duplication mapping in Figure 3. The protocol in Figure 7 was followed for crossing hDp!3 into the sterile strain carrying let-538, since this strain was balanced by szTl[lon-2 unc-29] instead of sDp2. Homozygous lethal Dpy-5 Unc-13 hermaphrodites carrying no duplication were examined for let-370 and let-599, because these lethal mutations could not be separated from dpy-5 by any of the duplications used. Sterile Unc-13 (dpy-5 let-x unc-13; hDpz), or Dpy-5 Unc-13 (dpy-5 let-x unc-13) worms were selected and allowed to grow for one week to reach adulthood, and were then subjected to microscopic examination. One week old N2, Unc-13 (dpy-5 unc-13; hDpl3) and Dpy-5 Unc-13 (dpy-5 unc-13) animals were used for comparison. DAPI-staining followed the protocol of Ellis and Horvitz (1986). 44 Figure 7: Protocol for let-538 dpy-5 let-538 unc-13; szTl(I;X)[unc-29; lon-2] X dpy-5 dpy-14; hDp!3 Lon-2 males Dpy-14 hermaphrodites I Pick Wild-type males .. dpy-5 let-538 unc-13; dpy-5 dpy-14; hDpl3 X (backcross) dpy-5 let-538-unc-13; szTl(I;X)[unc-29; lon-2]; X 'Wild-type' hermaphrodites i Pick Unc-13s dpy-5 let-538 unc-13 I dpy-5 let-538 unc-13 ; hDpl3 C h a p t e r 2 : RESULTS Test for Sperm-defective Mutants: All sterile adult mutants had a normal life-span of two to three weeks, and grew to the expected size for Unc-13 adults (0.6 - 0.8 mm), except let-604(h490), which was slightly dumpy and shorter (0.45 - 0.5 mm) than usual. The other allele of let-604, h293, grew to the full length expected for an adult. Since hermaphrodites produce both egg and sperm, a mutation affecting either of these two processes would cause sterility. A simple test to identify sperm-defective mutants is to mate an adult hermaphrodite homozygous for a sterile mutation to wild-type males. In this way, if the hermaphrodites are producing normal oocytes, the sperm from the wild-type males should rescue the sterile phenotype. Five lethal mutations, representing 5 essential genes, that resulted in sterile adult phenotypes were crossed to wild-type males, in an attempt to rescue the sterile phenotype, let-370(hl28), let-538(h990), let-544(h692), let-545(h842), and let-599(h290) . None were rescued by wild-type sperm. The remaining ten essential genes characterized in this study were previously tested for defects in spermatogenesis (McDowall 1990). Therefore, none of the lethal mutations could have affected the process of spermatogenesis alone. Furthermore, since none of the four mutants displaying a maternal effect, let-370, let-538, let-599, and let-604, were rescued by wild-type sperm, they must all be strict maternal effect lethals. Microscopic Characterization of Sterile Adult Mutants: To determine if the sterile mutants were defective in oogenesis, or in the structure of the gonad itself, adult sterile animals were observed under Nomarski microscopy, and using fluorescence microscopy after DAPI-staining of their nuclei. Sixteen lethal mutations, representing 15 essential genes, that resulted in sterile adult phenotypes were characterized: let-355(h81), 46 let-367(h!19), let-370(hl28), let-384(h84), let-395(h271), let-513(h752), let-538(h990), let-544(h692), let-545(h842), let-599(h290), let-603(h289), let-604(h293), let-604(h490), let-605(h312), let-606(h292) and let-610(h695). Microscopic examination revealed many visible defects in the gonads of these animals (Table 5). Fertile N2 (Figure 8), Unc-13 (Figure 9) and Dpy-5 Unc-13 (Figure 10) animals carrying no lethal mutation were used for comparison. There were no detectable differences in the gonads of N2 and Unc-13 animals. In the distal gonad arms of N2 and Unc-13 animals, the germ cells occurred as a syncytium in meiotic prophase, while mitotic proliferation occurred near the distal tip cell (Figures 8a-1, 9a-1). Differentiation occurred in the loop region, producing a single row of oocytes in the proximal arm (Figures 8a-2, 8b-2, 9a-2, 9b-2), which were fertilized by sperm in the spermatheca (Figures 8a-3, 8b-3, 9a-3, 9b-3). The DAPI-stained N2 and Unc-13 animals portrayed in Figures 8b and 9b, respectively, display dense staining of their mitotic and meiotic nuclei in their distal arms. Dpy-5 Unc-13 animals (Figure 10) were more disorganized internally. The gonads occupied a larger portion of the body cavity, but the same oocyte development was seen in their gonads. The sterile strains were placed into three phenotypic classes: gonadogenesis mutants, proliferation and gametogenesis mutants, and maternal effect lethals. (i) Mutations Affecting Gonadogenesis: Four of the mutants were defective in gonadogenesis, let-395, let-603, let-605 and let-610 (Figures 11 - 14). All of these arrested development as adult hermaphrodites, and either did not develop any recognizable gonad, or developed a malformed gonad. Homozygotes for let-395 (Figure 11) arrested as young adults. DAPI-staining revealed a group of nuclei clustered at the vulva (Figure 1 lb-1), but no structure that could be recognized as a gonad was observed. Homozygotes for let-603 (Figure 12) lacked both gonads and vulva. Where the gonads should have been located, there was a mass of undifferentiated, tumour-like tissue (Figures 12a-l, 12b-l, 12d-.l). This tissue mass caused these animals to bulge in the center, where the girth of the animal was larger than it was anteriorly or posteriorly. DAPI-staining reveals a group of nuclei 47 (Figure 12c-l), which could be undifferentiated germ cells. The tissue mass subsequently became necrotic, degenerating into a localized group of nuclei (Figure 12c-1). In let-605 homozygotes (Figure 13), rudimentary gonads could be detected. The gonads were extremely atrophied, being approximately one sixth of their normal size. There was no visible evidence of any germ cells. The small size of the gonads could be due to a lack of germ cells, but the gonad appeared malformed as well. In the proximal region of the gonad, the uterus appeared connected to the vulva (Figure 13a-4). However, the distal region did not extend much beyond the loop, widely separating the distal ends of the two gonad arms (Figure 13a-1, 13a-2). As such, the gonad arms were only partially reflexed in these animals. The DAPI-stained animals showed a clustering of nuclei around the vulva (Figure 13b-l). In let-610 homozygotes (Figure 14), the proximal region of the gonad appeared morphologically normal, with the uterus attached to the vulva (Figure 14a-4). However, the gonad was approximately a third of its normal size. There appeared to be greatly reduced number of germ cells (if any), which would account for the small size of the gonad. The distal region did not extend much beyond the loop, widely separating the distal ends of the two gonad arms (Figures 14a-l, 14a-2). As such, the gonad arms were only partially reflexed in these animals. In the DAPI-stained animals, none of the characteristic staining of germ cell differentiation could be seen, and no gametogenesis was detectable. (ii) Mutations Affecting Germ Cell Proliferation and Gametogenesis: The majority of the mutants examined were defective in germ cell proliferation and gametogenesis (Figures 15 - 21). There were seven mutants in this category, let-355, let-367, let-384, let-513, let-544, let-545, and let-606. These mutants arrested as adults and formed gonads, but had defects in germ cell proliferation, or in the formation of mature gametes. Homozygotes for let-355 (Figure 15) failed to undergo either spermatogenesis or oogenesis. The germ cells underwent proliferation, but not gamete differentiation. It is uncertain whether these germ cells remained in mitosis, or entered meiosis and failed to progress. However, 48 the nuclei in the loop region have the appearance of thread-like chromatin, characteristic of the pachytene stage of meiotic prophase, when the chromosomes become more conspicuous as they thicken. There was no overproliferation of germ cells. In the case of let-384 (Figure 16), the germ cells proliferated, but failed to differentiate into gametes, as seen by the lack of sperm or oocytes after DAPI-staining (Figure 16b). It is uncertain whether these germ cells remained in mitosis, or entered meiosis and failed to progress. However, none of the DAPI-stained nuclei exhibited the thread-like appearance of germ cells in the pachytene stage of meiotic prophase. There was no overproliferation of germ cells. The gonadal cells became necrotic soon after the animal reached adulthood. let-606 homozygotes (Figure 17) had enlarged proximal arms from the ectopic proliferation of germ cells in this region. The germ cells from the distal arms underwent limited differentiation in the loop region (Figure 17a-2), yet gametogenesis did not occur. The germ cells at the most proximal region of the gonad resembled the distal mitotic germline nuclei, whereas the germ cells in the loop region were characteristic of meiotic nuclei. The result was swollen proximal arms filled with the proliferating excess of germline nuclei that had a compact morphology (Figures 17a-1, 17b-1, 17c-1). The vulva appeared normal (Figure 17a-4). In let-513 homozygotes (Figure 18), there appeared to be a reduced number of germ cells in the distal arms. Spermatogenesis failed to occur, but the germ cells did undergo oogenesis. The presence of oocytes but not sperm is characteristic of a Fog (/feminization of germline) phenotype, whereby the germline assumes a female mode of expression. Oogenesis occurred earlier than usual, with large oocytes produced ectopically midway in the distal arms (Figures 18a-l, 18b-l), and a reduced number of syncytial nuclei occurring in the other half of the distal arms. The oocytes then degenerated midway along the proximal arms (Figure 18a-2). Oocytes do not normally form until after the loop region, in the proximal arms of the gonad. The oocytes did not form embryos when fertilized by wild-type sperm, however the proximal gonad near the vulva may not be properly formed (Figure 18a). The failure to develop embryos suggests that either the oocytes were abnormal, or they could not be reached by male sperm. In addition, dpy-5 let-513 49 unc-13;sDp2 animals segregated male progeny at the elevated frequency of 6.2% (77/1252), making this strain Him-like (Mgh incidence males). The Him phenotype is most likely due to an unlinked second site mutation, since it was lost upon outcrossing. let-544 homozygotes (Figure 19) underwent spermatogenesis, as revealed by DAPI-staining (Figure 19b-2). However, they failed to produce oocytes, characteristic of a Mog (masculinization of germline) phenotype. The germ cells in the distal region have the appearance of thread-like chromatin, characteristic of the pachytene stage of meiotic prophase (Figure 19b-3). The germ cells appeared to undergo limited germ cell differentiation in the most proximal region (Figure 19a-2), indicating that they may switch to oogenesis, but fail to progress. This is supported by no overproliferation of sperm occurring. In let-545 homozygotes (Figure 20), spermatogenesis occurred, as revealed by DAPI-staining (Figure 20b-2). However, they failed to produce oocytes, characteristic of a Mog phenotype. There are a reduced number of germ cells in the gonads, and they show less differentiation than those in let-544. However, the germ cell nuclei in the proximal arms have the characteristic thread-like appearance of meiotic nuclei (20b-3). The gonadal cells became necrotic and began to break down soon after the animal reached adulthood. Homozygotes for let-367 (Figure 21) produced sperm nuclei that were visible in the DAPI-stained animals (Figure 21b-l). Oogenesis occurred, but generated abnormal oocytes. Germline nuclei could be observed in the loop region, but they failed to localize in the periphery of the gonad. Differentiation was abnormal, and the oocytes in the proximal arms varied in size and were malformed (Figure 21a-l). These oocytes subsequently degenerated. These oocytes failed to form embryos either after self-fertilization, or after fertilization by a wild-type male. The vulva appeared normal (Figure 21a-2). (iii) Maternal Effect Lethals: The third category of mutants reached adulthood, accomplished gametogenesis, but produced fertilized eggs that were arrested (Figures 22 - 26). All four mutants, let-370, let-538, 50 let-599, and let-604, were strict maternal effect lethals (Kemphues etal. 1988). In each case, homozygotes produced from heterozygous hermaphrodites reached adulthood. However, when these homozygotes were crossed to wild-type males, heterozygous embryos could not be rescued by wild-type sperm and were inviable. Homozygotes for let-538 (Figure 22) produced sperm nuclei that were visible in the DAPI-stained animals (Figure 22d-l), sometimes in copious amounts (Figure 22c-l). Oogenesis occurred, but the oocytes in the proximal arms were malformed and varied in size, most being much smaller than normal (Figure 22b-l). In some cases large numbers of oocytes were produced (Figure 22a-1), however they often degenerated in the proximal arms. Very rarely did eggs form, and these showed no sign of morphogenesis (Figure 22b-2). In the case of let-599 (Figure 23), the proximal gonad arms appeared enlarged, even though only a few oocytes and eggs were present. Oocytes formed in the proximal arms. However, these appeared malformed, and smaller than normal (Figures 23a-3, 23b-l). Only rarely were eggs produced. These eggs were disorganized and more spherical than normal, and they failed to develop or be laid (Figure 23a-l). The DAPI-stained animal in Figure 23d shows one egg in the center with stained nuclei. The structure of the gonad also appeared disorganized, since the enlarged proximal arms appeared to lack the normal ordered pattern of oocyte development. ( Homozygotes for let-604(h293) (Figure 24) showed gamete differentiation in the loop region, with oocytes (Figure 24a-1) and fertilized eggs in the proximal region. However, the eggs did not develop past gastrulation, since no morphogenesis was visible (Figure 24a-2). None of the eggs were laid, and they became necrotic within the gonad (Figure 24b-l). The let-604(h490) homozygotes (Figure 25) were more severe. Sperm were present (Figure 25b-3), but the oocytes did not undergo differentiation until they reached the proximal arm (Figure 25a-2). The nuclei assume a peripheral arrangement around the gonad in the proximal arms (Figure 25a-l), just after the loop region. Normally, nuclei at pachytene form this arrangement in the distal arms near the loop region. Oocytes formed midway along the proximal arms, but then degenerated (Figure 51 25a-3). Very rarely did eggs form, and these showed no sign of morphogenesis (Figure 25b-l). The eggs became necrotic within the hermaphrodite without being laid. let-370 homozygotes (Figure 26) were the least severe. They produced oocytes (Figures 26b-1, 26c-1) which sometimes degenerated (Figure 26b-2, 26c-2), and sperm (Figure 26f-2). They laid eggs, but the eggs usually failed to hatch. A few of the eggs developed to the final stage of elongation (Figure 26a-l), and some occasionally hatched, but the larvae died soon thereafter during the first larval stage. 52 Table 5 : Characterization of Adu l t Mutant Phenotypes Oocytes Sperm Eggs Gonad Gene (allele) Formed Formed Formed Structure Comments Mutants Affecting Gonadogenesis: let-39S(h271) no no let-603(h289) let-60S(h312) let-610(h695) no no no no no no no no no no gonad No recogni zable gonad no gonad Tumour-like mass atrophied No germ cells; soma malformed; uncoordinated reduced Few/no germ cells; arms partially reflexed Mutants Affecting Germ Cell Proliferation and Gametogenesis: let-35S(h81) let-367(h!19) let-384(h84) let-513(h752) let-544(h692) Iet-S45(h842) let-606(h292) no yes no yes no no no no yes no no yes yes no no no no no no no no normal normal normal normal normal normal proximally enlarged Germ cell proliferation, but no gametogenesis Spermatogenesis; oocytes malformed, degenerate Germ cell proliferation, but no gametogenesis Reduced number of germ cells; no sperm; ectopic entry into oogenesis in mid-distal arm Spermatogenesis, but no oogenesis Spermatogenesis, but no oogenesis Tumour: ectopic proliferation of germ cells Maternal Effect Lethals: let-370(hl28) yes yes yes normal let-538(h990) yes yes very rare normal let-599(h290) yes yes rare. Iet-604(h293) yes yes yes normal let-604(h490) yes yes rare normal Morphogenesis, but eggs usually fail to hatch Eggs show no morphogenesis proximally enlarged E g g s s n o w n o m o r P h ° g e n e s ' s Eggs show no morphogenesis & become necrotic Eggs show no morphogenesis F i g u r e 8: Germline of Wild-type Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a) and under fluorescence after DAPI-staining (b). Mitotic germ cells occur in the most distal region of the gonad (a-1), these syncytial nuclei appearing heavily DAPI-stained (b-1). Oocytes begin to form in the loop region, and by the proximal gonad arm the oocytes are well-formed and clearly visible by Nomarski (a-2). DAPI-stained oocyte nuclei appear as a cluster of dots (b-2), representing individual chromosomes at diakinesis of meiosis I. The spermatheca occurs as a constriction midway along the proximal arm (a-3), many sperm nuclei being visible after DAPI-staining (b-3). Eggs are visible by Nomarski in many different stages of development (a-4), before they exit through the vulva (a-5, b-4). F i g u r e 9: Germl ine of U n c - 1 3 Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a) and under fluorescence after D A P I -staining (b). The gonads of Unc-13 animals display the same pattern of germ cell development as that found in wild-type animals (Figure 8). Mitotic germ cells (a-1, b-1), oocytes (a-2, b-2), sperm (b-3), and eggs (a-3) occur in the same relative positions as those found in wild-type animals. The vulva is also clearly visible (a-4). Figure 10: Germline of Dpy-5 Unc-13 Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a) and under fluorescence after D A P I -staining (b). Gonads occupy more of the pseudocoelom, but display the same pattern of germ cell development as in wild-type animals (Figure 8). The relatively small size of the pseudocoelom causes the oocytes to press upon the pharynx (a-1). The mitotic germ cells (a-2, b-1), oocytes (a-3, b-2), sperm (b-3), eggs (a-4), and vulva (a-5) are shown. Figure 11: Germline of let-395(h271) Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a) and under fluorescence after D A P I -staining (b). let-395 has an undeveloped gonad, in which gametogenesis does not occur. D A P I -staining reveals a group of nuclei at the vulva (b-1). 57 Figure 12: Germline of let-603(h289) Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a, b, c) and under fluorescence after D A P I -staining (d). (a, b) represent the same animal, let-603 has no recognizable gonad, and appears vulvaless. The tumour-like cell mass is marked (a-1, b-1). This cell mass appears as a group of DAPI-stained nuclei (d-1), that subsequently becomes necrotic (c-1). 58 Figure 13: Germline of let-605(li3I2) Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a) and under fluorescence after D A P I -staining (b). The distal ends of the atrophied gonad arms are widely separated (a-1, a-2). The distal arms are shorter than the proximal arms (a-3). Gametogenesis does not appear to occur. The vulva appears normal (a-4). DAPI-staining reveals a group of nuclei around the vulva (b-1) F i g u r e 14: Germline of let-61()(li695) Hermaphrodites A d u l t hermaphrodites v i e w e d under N o m a r s k i opt ics (a) and under f luorescence after D A P I -staining (b). let-610 shows many features in c o m m o n wi th let-605 (Figure 13). T h e distal ends o f the atrophied gonads are w i d e l y separated (a-1, a-2), and the distal arms are shorter than the proximal arms (a-3). Furthermore, gametogenesis does not occur , and the v u l v a appears normal (a-4). Howeve r , the gonad is larger than that found in let-605, and the group o f D A P I - s t a i n e d nuclei around the v u l v a in let-605 are l ack ing in let-610 (b-1). Figure 15: Germline of /et-JSS/fiS/JHermaphrod'ites A d u l t hermaphrodites v i e w e d under N o m a r s k i opt ics (a) and under f luorescence after D A P I -s ta in ing (b). D A P I - s t a i n i n g reveals the presence o f germ ce l l s (b-1) , h o w e v e r they f a i l to differentiate. The distal gonad arms (a-1, b-2), and the loop region (a-2, b-3) are shown. Figure 16: Germline of let-3H4(h84) Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a) and under fluorescence after D A P I -staining (b). The germ cells in the distal arms (a-1, b-1) do not for sperm or oocytes, as revealed by DAPl-staining. The DAPI-stained nuclei in the bend of the gonad arm (b-2) seem to exhibit the thread-like appearance characteristic of meiotic prophase. The vulva appears normal. Figure 17: Germline of let-606(h292) Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a) and under fluorescence after D A P I -staining (b, c). The proximal arms near the vulva are swollen with excess germ cell proliferation (a-1, b-1, c-1). These germ cells are distinct from those produced in the distal arms (a-2, b-2, c-2), since the later appear to undergo limited differentiation in the proximal arms (a-3). The vulva appears normal (a-4). 63 Figure 18: Germline of let-513(h752) Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a) and under fluorescence after D A P I -staining (b). Oocytes are produced early and are visible midway along the distal arm (a-1, b-1). Note the absence of sperm, and the degeneration of oocytes in the proximal arm. The spermatheca is marked (a-2). F i g u r e 19: Germline of let-544(h692) Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a) and under fluorescence after D A P I -staining (b). The germ cells in the distal arms (a-1, b-1) appear to undergo limited differentiation in the proximal arms (a-2). DAPI-staining reveals the presence of sperm nuclei (b-2). Nuclei in the distal arm have the characteristic thread-like appearance of meiotic nuclei (b-3). The vulva appears normal. F i g u r e 20 : Germline of let-545(h842) Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a) and under fluorescence after D A P I -staining (b). DAPI-staining reveals the presence of germ cell nuclei in the distal arm (b-1), and sperm nuclei in the proximal arm (b-2), however oocytes are not produced. Nuceli in the proximal arm have the characteristic thread-like appearance of meiotic nuclei (b-3). The distal arm (a-1), and the proximal arm (a-2) are shown. The vulva appears normal (a-3). 66 F igure 21: Germline of let-367(hl 19) Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a) and under fluorescence after D A P I -staining (b). Small, malformed oocytes appear in the proximal arms (a-1), which then degenerate. Sperm nuclei appear to be present, as revealed by DAPI-staining (b-1). The vulva appears to be normal (a-2). F i g u r e 2 2 : Germl ine of let-538(h990) Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a, b) and under fluorescence after DAPI-staining (c, d). Many oocytes were produced (a-1), however they varied in size , most being much smaller than normal (b-1). DAPI-staining revealed the presence of sperm nuclei (d-1), often in copious amounts (c-1). Eggs rarely formed, and showed no sign of morphogenesis (b-2). DAPI-stained germ cell nuclei in the distal arm (d-2), and oocyte nuclei in the proximal arms (c-2, d-3) are marked. The vulva appears normal (a-2). F i g u r e 23: Germline of lety-599(h290) Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a, b) and under fluorescence after D A P I -staining (c, d). The proximal gonad appears enlarged, even though only a few oocytes and eggs are present. The eggs contain several nuclei (d-1), however, they fail to undergo morphogenesis (a-1). Sperm nuclei are visible by DAPI-staining (c-1). Germ cells in the distal arm (a-2), and oocytes in the proximal arm (a-3, b-1) are marked. Figure 24: Germline of let-6()4(h293) Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a, b) and under fluorescence after D A P I -staining (c). The oocytes form just after the loop region in the proximal arm (a-1), as in wild-type animals (Figure 8). Many eggs are produced. These eggs contain several nuclei (c-1), however they fail to undergo morphogenesis (a-2), and subsequently become necrotic (b-1). The vulva appears normal (a-3). The distal germ cells (a-4, c-2), and the sperm nuclei (c-3) are shown. Figure 25: Germline of let-6()4(h490) Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a, b) and under fluorescence after D A P I -staining (c, d). Iet-6()4(h490) mutations were more severe than let-604(h293) mutations (Figure 24). The nuclei in the proximal arm shows a peripheral arrangement (a-1). Oocytes form midway along the proximal arm (a-2). Few of these oocytes become fertilized, while most degenerate (a-3). The eggs contain several nuclei (c-1), however they fail to undergo morphogenesis (b-1). The distal germ cells (a-4, c-2), and the sperm nuclei (c-3) are marked. 71 Figure 26: Germline of let-370(hl28) Hermaphrodites Adult hermaphrodites viewed under Nomarski optics (a, b, c) and under fluorescence after DAPI-staining (d, e, 0- Oocytes were produced (b-1, c-1, e-1), however they sometimes degenerated (b-2, c-2). Eggs (d-1) develop to the elongation stage (a-1). The germ cell nuclei in the distal arm (b-3, e-2), and the sperm nuclei in the proximal arm (f-1) are marked. 72 Chapter 2: DISCUSSION This chapter described the characterization of lethal mutations in 14 essential genes mapping between dpy-5 and unc-13, plus one gene to the right of unc-13, by wild-type male rescue experiments, Nomarski microscopy and DAPI-staining. None of the lethal mutations were rescued by wild-type sperm, indicating that in each case sterility could not have been due to defects in spermatogenesis alone (five were tested in this study and the remaining ten were tested in a previous study, McDowall 1990). These rescue experiments also classify any maternal effect lethals as strict (Wood etal. 1980; Kemphues etal. 1988). The cytological data showed that four genes produced mutants with defects in gonadogenesis, seven genes produced mutants that disrupted germ cell proliferation or gametogenesis, and mutants in the remaining four genes produced eggs that failed to hatch or develop, thereby acting as maternal effect lethals. Characterization of Adult Mutant Phenotypes: The 15 essential genes described in this analysis that resulted only in sterile adult mutant phenotypes were characterized by microscopy. Essential genes with alleles that had a range of lethal phenotypes, including sterility, were not examined, because these genes are more likely to be required outside of embryogenesis. A rare sterile allele could be hypomorphic, producing sufficient product to enable the survival of the animal, but not of its embryos, because of a stringent requirement for maternally produced gene products (Perrimon et al. 1986; Kemphues et at. 1988). Essential genes identified by a single allele were included in this characterization. Eleven of the characterized essential genes are identified by a single allele, while the remaining four essential genes are identified by at least two alleles (Table 4, Chapter 1). In each case, only one allele was characterized, with the exception of let-604, for which two alleles were examined. Microscopic examination of adult sterile hermaphrodites revealed a variety of visible defects 73 affecting their gonads, germ c e l l p ro l i fe ra t ion , gametogenesis, or the development o f embryos . However , it should be noted that different alleles o f a s ingle locus can often display very different characteristics. F o r instance, mutations i n gld-1 can result i n at least four different phenotypes depending upon the type of mutat ion i n v o l v e d (Francis etal. 1995a). These phenotypes inc lude tumour format ion (nul l ) , abnormal oocyte (partial loss o f funct ion) , mascu l in iza t ion o f germl ine (ga in o f func t ion ) , and f e m i n i z a t i o n o f g e r m l i n e (ga in o f func t ion ) . Th e re fo r e , further characterizations are required to complete ly define the functions o f these sterile mutations. Mutations Affecting Gonadogenesis: T h e gonadal p r imord ium in a wi ld- type L I w o r m consists o f two somatic ( Z l and Z 4 ) and two germline ( Z 2 and Z 3 ) precursor cel ls ( K i m b l e and H i r sh 1979). T h e Z l , Z 4 somatic ce l l s are essential for the prol iferat ion and maturation o f germ cel ls . Dis rup t ions i n their migra t ions m a y result in steril i ty. In addi t ion, mutations affecting the migrat ion o f the distal t ip ce l l s can disrupt both gonadogenesis and gametogenesis (Hedgecock etal. 1987). (i) let-395, let-605, let-610: Mutat ions in three genes affected the early development of the soma and germline , let-395, let-605, and let-610. let-605 mutations resulted in an uncoordinated ( M c K i m etal. 1991) adult that was sterile due to the l ack o f germ ce l l s . T h e atrophied gonads in these an ima l s c o u l d o n l y part ial ly be attributed to the lack o f germ cel ls . T h e smal l size o f the gonad cou ld also be due to the malformat ion of the soma, w h i c h was on ly part ial ly reflexed. Muta t ions in let-610 were s imi l a r to, but less severe than, those i n let-605. let-610 mutations resulted in a s m a l l , immature gonad, w h i c h cou ld be attributed to the lack of germ cells . T h e p rox ima l gonad appeared properly fo rmed , but the arms were o n l y pa r t i a l ly re f l exed . let-395 muta t ions resul ted i n a v i s i b l e structure resembling a t iny, undeveloped gonad. A l l three mutations, let-395, let-605, and let-610, resulted i n v i s ib le defects i n the development o f the soma, and i n the prol i ferat ion o f the germl ine . Soma-germline interactions are important for the identi ty, prol i ferat ion, and differentiation o f germ cel ls 74 (Kimble and White 1981; Austin and Kimble 1987; Seydoux etal. 1990). Several loci have been identified in C. elegans, whose products are required maternally for normal germline development (Capowski et al. 1991). Mutations in these loci cause mes (maternal effect sterile), or grandchildless phenotypes, whereby homozygosity in the mother prevents normal germline development in the progeny. Two mutations, mes-3 and mes-4, resulted in unreflexed or partially reflexed gonad arms that carried reduced germ cell numbers (Capowski etal. 1991). These mes mutations were similar in appearance to let-605 and let-610 mutations, except the let mutations had fewer germ cells. In a more severe mutation, mes-1, progeny fail to generate germline progenitor cells. The mes mutations revealed a role for maternal products in the development of the embryo's germline. However, zygotic products could also be involved, and mutations in such loci would cause aberrations in the mutant's germline, rather than in the germline of the mutants' progeny. Such zygotic genes could act upon maternal cues, let-395, let-605 and let-610 could each play a role in the early development of the germline. (ii) let-603: The most extreme developmental defect was seen with let-603 mutations, which caused a mass of undifferentiated tissue to form in lieu of a gonad. This mutation appears to have affected cell differentiation causing a tumorous overgrowth. The gonad appears formless, but may contain undifferentiated germ cells. This phenotype is unique among the description of mutants in C. elegans. A n y involvement of let-603 in cell fate determination would need to be determined at the lineage level. Mutations Affecting Germ Cell Proliferation and Gametogenesis: Germ cell fate is controlled by a network of regulatory pathways, that interact to decide between three cell fates: mitosis, spermatogenesis, or oogenesis (Ellis and Kimble 1995). Many of the genes involved in these pathways have been identified, and their roles in sex determination 75 elucidated (reviewed in Kuwabara and Kimble 1992). Disruptions in these genes can cause a variety of germline defects, involving proliferation, sexual identity, or viability of germ cells. (i) let-355, let-384: Mutations in two genes, let-355 and let-384, disrupted gametogenesis, resulting in agametic phenotypes. There was no gross reduction, or overproliferation of germ cells, they simply failed to differentiate. It is uncertain whether the germ cells in these mutants remained in mitosis, or i f they entered meiosis, then failed to progress, let-355 germ cells appeared to be mitotic, as their nuclei lacked the characteristic appearance of meiotic nuclei, let-384 germ cells may have entered meiosis, as they seemed to exhibit the thread-like appearance of pachytene chromatin, but this remains to be tested. If these assumptions are correct, then let-355 mutants may arrest earlier than let-384 mutants. Several genes are involved in the maintenance and progression through mitosis, glp-1 (germ/ine proliferation) and glp-4 are both involved in maintaining the mitotic proliferation of germ cells (Austin and Kimble 1987; Beanan and Strome 1991). glp-1 promotes mitosis under the regulation of the distal tip cell . Mutations in this gene cause a severe reduction in the number of germ cells generated, by causing them to prematurely enter meiosis and differentiate as gametes (Austin and Kimble 1987). Neither let-355, nor let-384 resemble glp-1, because there was no apparent reduction in the number of germ cells, nor did the germ cells differentiate as gametes. glp-4 is required for the progression of germ cells through mitosis (Beanan and Strome 1991). Mutations in this gene caused germ cells to arrest in mitotic prophase, thereby preventing them from either dividing mitotically, or entering meiosis and differentiating. Neither let-355, nor let-384 resembled glp-4, because there was no apparent reduction in the number of germ cells. It is unknown whether the germ cells in these two lethal mutations were able to exit mitosis, but there was no evidence for mitotic arrest, which would have severely limited the number of germ cells generated. Hence, these lethal mutations did not appear to interfere with mitotic proliferation. Instead, they may have interfered with the mitotic-meiotic switch, or progression through meiosis. gld-1 (germ/ine defective) is believed to have a nonessential role in inhibiting premeiotic 76 proliferation in the hermaphrodite germline, in addition to its essential role in oogenesis (Francis et al. 1995a, 1995b). Mutations in gld-1 that disrupted its ability to inhibit premeiotic proliferation resulted in a tumorous phenotype, whereby germ cells exited meiosis and proliferated mitotically. Neither let-355 nor let-384 mutants resembled the tumorous phenotype of gld-1 mutations, as there was no overproliferation of germ cells. Instead, these lethal mutations may be involved in another aspect of the meiotic-mitotic switch, or in the normal progression through meiosis. (ii) let-606: Mutations in let-606 caused a tumorous phenotype, in addition to disrupting gametogenesis. Agametic let-606 mutants displayed ectopic proliferation of germline nuclei in the proximal gonad arms. The germ cells entered meiosis in the loop region, then appeared to revert to mitotic cell division, causing overproliferation. This gene was independently identified as gld-2 (complements let-606) (Lisa Kadyk, personal communication). The gld-2 allele isolated by L. Kadyk was more severe then let-606, as the germ cells did not show any signs of differentiation that were apparent in let-606 (Figure 17a-3). Instead, gld-2 mutations re-entered mitosis shortly after entering meiosis. If the gld-2 allele isolated by L . Kadyk was a true null, then let-606(h292) may be a partial loss of function mutation, let-606 and gld-2 closely resembled the tumourous phenotype of gld-1, that mapped to the right of unc-13 on L G I (Francis et al. 1995a). Neither gld-1, nor let-606 (gld-2) produced oocytes or sperm. Both displayed a similar tumour morphology, in which germ cells proliferated ectopically in the proximal gonad, gld-1 was shown to have at least two nonessential functions in suppressing premeiotic proliferation and promoting spermatogenesis, as well as an essential function in directing oogenesis (Francis et al. 1995a, 1995b). gld-1 promoted meiotic development, but it was not required for entry into meiosis. In gld-1 mutants, germ cells entered meiosis, but could not progress through meiosis. Instead, they inappropriately exited meiosis and proliferated mitotically by circumventing the normal controls (does not require wild-type glp-1 activity), let-606 may have a similar function, in the decision between meiotic and mitotic cell fates, gld-1 also played a role in spermatogenesis, gld-1 gain of function mutants had a Mog (masculinization of germline) phenotype by disrupting the switch from 77 spermatogenesis to oogenesis. However, since gld-1 mutants can make sperm, they have a nonessential role in spermatogenesis. A similar function has not been demonstrated for let-606. Finally, gld-1 was essential for oogenesis, gld-1 mutants will only form tumours if the fern or fog spermatogenesis genes are inactive, thereby setting the female mode. In a background where the fern or fog genes are active, creating a male mode (i.e. in males or hermaphrodite mutants), germ cells differentiate as sperm instead of reverting to mitosis. It is unknown how the sex differentiation of the germline affected tumour formation in let-606 mutants. Therefore, the similarity in tumour morphology with gld-1 mutants suggested a similar role for the let-606 gene product in promoting meiotic development, but no roles in oogenesis or spermatogenesis were demonstrated. Several genes have been identified in Drosophila that are required for germline proliferation and germline sex determination. Mutations in ovo (Pauli etal. 1995) and otu (ovarian tumor) (Rodesch etal. 1995) can both give rise to a range of phenotypes that bear striking similarity to each other, and to gld-1 mutations in C. elegans. In particular, ovo and otu mutations can produce a tumorous phenotype, whereby egg chambers are fdled with mitotic, undifferentiated germ cells. These ovo and otu phenotypes are only produced in females, as males bearing null mutations in these genes are wild-type. Similarly, gld-1 null mutations are only expressed in hermaphrodites, males being unaffected (Francis etal. 1995a). Both Drosophila genes are required in the germline during oogenesis. The otu gene has two roles in germline development; it is required for oocyte differentiation, as well as for germ cell proliferation. Similarly, the Drosophila gene, enc (encore) (Hawkins etal. 1996), is required for oocyte differentiation, as well as for regulating the number of germline mitoses. Mutations in enc cause one extra round of mitosis in the germline, in addition to oocyte nucleus defects. These two effects are independent of each other in their requirement for enc product. Therefore, the model proposed by Ellis and Kimble (1995) that links the control of germline sex determination to the regulation of germline proliferation may be a common phenomenon. 78 (iii) let-513: Mutations in let-513 enabled the production of oocytes, but not of sperm, thereby conferring a Fog (feminization of germline) phenotype. This does not imply that let-513 acted in an analogous fashion to the fog loci, but that the germline of let-513 hermaphrodites could only enter the female mode of germline sex determination. Oogenesis in let-513 mutants did not progress normally. Oocytes were produced ectopically in the distal arms of the gonad, and subsequently degenerated in the proximal arms. The ectopic generation of oocytes combined with the lack of sperm, and the reduction in number of germ cells in the distal arms, could be due to defects in germ cell proliferation, identity or differentiation, let-513 mutations could cause the early entry of germ cells into meiosis, thereby accounting for the reduced number of germ cells near the distal tip region. Mutations in let-513 could also affect germ cell identity and differentiation, because let-513 mutants were unable to generate sperm. This suggested that either spermatogenesis was suppressed, or that germ cells could not progress through spermatogenesis. It seems most likely that germ cells could not enter spermatogenesis, as there was no accumulation of arrested germ cells apparent in the proximal arms by DAPI-staining. Instead, germ cells adopted a female mode, and developed as oocytes. These oocytes did not develop into embryos when fertilized by male sperm, suggesting that either the oocytes were abnormal, or that the gonad was obstructed, thereby preventing the sperm from reaching the oocytes, let-513 mutations could have pleiotropic effects, especially as the pathways of germline proliferation and sex determination are interconnected, such that the sexual identity of a germ cell is determined around the time it enters meiosis, and mutations affecting one process, often affect the other as well (Barton and Kimble 1990; Ellis and Kimble 1995). In particular, it is believed that inhibitory interactions occur between genes involved in determining the three germ cell fates of mitosis, spermatogenesis, and oogenesis, so that only one fate can be expressed in a cell (Ellis and Kimble 1995). As such, the let-513 gene product could be required for promoting mitosis, or suppressing meiosis, as well as being involved in the specification of the male germ cell fate. 7 9 Several mutations that alter the sexual identity of the hermaphrodite germline to a female mode have been identif ied and described, fem-1 (Spence etal. 1990), fem-2 ( K i m b l e et al. 1984), fem-3 (Barton etal. 1987) , fog-1 (Bar ton and K i m b l e 1990), and fog-3 ( E l l i s and K i m b l e 1995) mutations a l l resulted i n transforming ce l l s that w o u l d n o r m a l l y develop as sperm into oocytes . These f ive genes function as terminal regulators o f the sex determination pathway. T h e act ivat ion of these genes specifies the male mode of germline development, wh i l e their inact iva t ion specifies the female mode ( E l l i s and K i m b l e 1995). A s there was on ly one mutat a l le le ident i f ied for let-513, and double mutants were not made, it was diff icul t to speculate on its involvement in the germl ine . One must be cautious when interpreting the phenotype o f one al le le , because different al leles o f a gene can produce different mutant phenotypes. F o r instance, gld-1 ga in o f funct ion mutations can result either i n a F o g or a M o g (masculinizat ion of germl ine) phenotype, depending on the role o f the mutant gene product, ind ica t ing a role for the gld-1 gene product i n both the spec i f ica t ion o f spermatogenesis and oogenesis, respect ively. A s such, let-513 c o u l d have other roles not evident by this muta t ion . Fur thermore , let-513 has not been charac ter ized as to the type o f muta t ion i n v o l v e d . Genes invo lved i n germline sex determination can have opposite phenotypes, depending upon the type o f mutations i n v o l v e d . F o r instance, fem-3 loss o f funct ion mutations results i n a F o g phenotype, characterized by an overproduct ion o f oocytes and a l ack of sperm, w h i l e fem-3 ga in o f funct ion mutations result i n a M o g phenotype, characterized by an overproduct ion o f sperm and a l ack o f oocytes (Bar ton et al. 1987) . let-513 is u n l i k e l y to be a ga in o f func t ion muta t ion , as ga in o f funct ion mutations often results i n dominance (Park and H o r v i t z 1986). A s a recessive le tha l , let-513 is more l i k e l y to be a nu l l or a hypomorphic a l le le . ( iv) let-544, let-545: Muta t ions i n two genes, let-544 and let-545, enabled the product ion o f sperm, but not o f oocytes, thereby confer r ing a M o g phenotype. A s such, the germl ine o f these hermaphrodites cou ld only enter the male mode o f germline sex determination. In both o f these mutants, the germ cells appeared to have meiot ic nucle i . T h e germ cel ls may make the sperm-oocyte swi tch , but fa i l 80 to progress through oogenesis. Alternatively, the germ cells could fail to exit spermatogenesis. Since there is no overproduction of sperm in the proximal arms in either lethal mutant, it seems unlikely that the germ cells continually make sperm. Furthermore, in let-545 mutants there appeared to be a reduced number of germ cells present in the gonad. Therefore, germ cells may switch to oogenesis, then become arrested in their development. Further analysis, and in particular double mutant studies, are needed to define the roles of these loci in oogenesis, or the sperm-oocyte switch. Mutations that alter the sexual identity of the hermaphrodite germline to a male mode have been identified and described. mog-I and gld-1 mutations can result in transforming cells that would normally develop as oocytes into sperm, mog-1 is believed to be involved in the sperm-oocyte switch by negatively regulating/em-3 (Graham and Kimble 1993), while gld-1 is involved in either specifying or executing the oocyte fate (Francis etal. 1995a, 1995b). (v) let-367: Mutations in the remaining gene, let-367, enabled both spermatogenesis and oogenesis to occur, but resulted in the generation of abnormal oocytes, that varied in size and subsequently degenerated. Since these oocytes failed to develop after being fertilized by wild-type male sperm, the let-367 mutation could disrupt one of many stages required for oocyte development. It is unknown whether the abnormal oocytes progressed through to diakinesis of meiotic prophase, which is the stage of arrest in wild-type oocytes prior to fertilization. Maternal Effect Lethals: Mutations in four essential genes can be classified as strict maternal effect lethals, because they cannot be rescued by fertilization with wild-type male sperm (Wood etal. 1980; Kemphues et d. 1988). The progeny from a cross between a homozygous hermaphrodite and a wild-type male 81 w o u l d have been heterozygous for a recessive muta t ion , but they s t i l l f a i l ed to deve lop . T h i s indicated that these gene products are required maternal ly , and c o u l d not be corrected by zygo t ic transcription. Mate rna l proteins and R N A that are produced dur ing oogenesis p rov ide the basic machinery for cel lular events in the embryo. E m b r y o n i c development i n C. elegans is rapid , taking only about 14 hours f rom fer t i l iza t ion to hatching ( W o o d et al. 1980), and there i s no increase i n size before hatching f rom the egg. In o rgan isms that deve lop r ap id ly , it is advantageous to m a x i m i z e maternal contributions, because the gonad provides a more efficient synthetic machinery than the egg, and can meet a l l o f the requirements o f the d e v e l o p i n g embryo . In contrast, organisms that d isplay an increase i n the s ize o f the embryo , such as i n m a m m a l s , cannot re ly purely upon maternal products, because maternal products do not meet the greater demands o f the embryo (Wieschaus 1996). Mate rna l ly required gene products have been shown to p lay a prominent role in C. elegans embryogenesis . F o r instance, o f the 24 mutants iden t i f i ed by W o o d et al. (1980) that were required for embryogenesis , 21 had a maternal effect, such that maternal expression was sufficient for e m b r y o n i c s u r v i v a l , regardless o f the genotype o f the e m b r y o . T h e s t rong ma te rna l requirement of C. elegans embryos was i l lustrated by experiments i n v o l v i n g the gene ama-1, that encoded the enzyme R N A polymerase II used in m R N A synthesis (Bul le r j ahn and R i d d l e 1988; R o g a l s k i etal. 1988). Mutants homozygous for a recessive al lele o f ama-1 cou ld survive through embryogenesis and arrested after ha tch ing , dur ing the first l a rva l stage. These embryos w o u l d have diff icul ty transcribing their o w n genomes, therefore must have survived o n maternal products unti l after hatching, when death ensued. Recent ly , B r o w n i n g and Strome (1996) demonstrated the presence o f a paternal effect le thal , spe-11, that was suppl ied on ly by the sperm and participated directly in C. elegans embryogenesis. C. elegans embryogenesis is characterized by cytoplasmic streaming and reorganizat ion, as w e l l as an asymmetr ic first cleavage, g i v i n g rise to blastomeres w i t h h igh ly determined c e l l fates ( K e m p h u e s et al. 1985) . M a t e r n a l effect genes are d i rec t ly i n v o l v e d i n c o n t r o l l i n g pattern formation and ce l l fate determination, decisions that occur very early i n C. elegans embryogenesis 82 (Sulston etal. 1983). Many screens for maternal effect lethals and zygotic genes have been carried out in C. elegans, to identify genes involved in the control of these early embryonic processes. Many developmentally important loci have been identified, including several zyg (zygotic), mel (maternal effect /ethal), ooc (oocyte defective), him (Mgh incidence males), and let loci (Wood et al. 1980; Hirsh etal. 1985; Kemphues etal. 1988; Mains etal. 1990). For several of these loci , their roles in embryonic development have been elucidated. Based on the identification of alleles that can mutate to strict maternal effect lethals, the number of purely maternally required genes in C. elegans has been estimated to be as low as 12 (Kemphues et al. 1988). However, other estimates have been as high as 60 essential genes required maternally for embryogenesis (Johnsen and Baill ie 1991). (i) let-538, let-599, let-604: Mutants that were homozygous for any one of these three loci rarely produced eggs, either by self- or cross-fertilization. These mutants produced defective embryos, that seemed incapable of developing past the first few stages of embryogenesis, as indicated by the failure of the fertilized eggs to show any sign of morphogenesis. T w o mutants of let-604 were examined, with let-604(h490) producing more severe defects than let-604(h293). Iet-604(h293) mutants often produced eggs, but they showed no sign of morphogenesis. Iet-604(h293) could be a partial loss of function mutation, that enabled more oocytes to survive until they became fertilized, but did not meet the more stringent requirements of embryogenesis. Embryogenesis can be divided into three stages: the generation of founder cells with specific fates, gastrulation (beginning of cell differentiation), and morphogenesis (Wood etal. 1980). It is uncertain how far the embryos from these mutants progressed, but it was certainly not beyond gastrulation. Several maternal effect sterile mutations identified in C. elegans could produce eggs that failed to show any sign of morphogenesis. These include zyg-14 (Hirsh etal. 1985), which is required for asymmetric positioning of the first cleavage, mel-23 (Mains etal. 1990), which causes an abnormal orientation of cell types, and mel-26 (Mains etal. 1990), which causes an abnormal orientation of the first cleavage. 83 The degeneration of the majority of oocytes in each of theses three mutants, suggested that mutations at these loci may have interfered with processes that occurred before fertilization. As such, these loci may be required for oogenesis, rather than for the control of specific embryonic processes. Any mutations that affect meiosis or the meiotic-mitotic switch, such as mei-1 (Clandinin and Mains 1993), could disrupt both oocyte development and embryogenesis. In addition, many genes have pleiotropic effects, such as glp-1, which affects both germline proliferation (Austin and Kimble 1987) and embryogenesis (Priess etal. 1987). (ii) let-370: Mutants that were homozygous for let-370 produced embryos that develop to the final stages of elongation, regardless of whether these embryos were homozygous (self-fertilization) or heterozygous (fertilized by wild-type male sperm). Occasionally these embryos hatched, giving rise to dead LI larvae. This locus may be involved in the later stages of embryogenesis, such as late gastrulation or morphogenesis. The C. elegans gene, skn-1, is required late in embryogenesis for development of pharyngeal and intestinal cells (Bowerman etal. 1992). Because the pharynx differentiates late in embryogenesis, mutations at this locus are not discernible until gastrulation. In Drosophila, mutations affecting gastrulation, such as abo (abnormal oocyte), resulted in maternal effect lethality. Lethal embryos produced by abo females die during late embryogenesis, or during the early larval stages, and display a range of cuticle defects (Tomkiel etal. 1995). A second method of identifying loci involved in the early developmental control of C. elegans is by screening for embryonic lethals, whereby homozygous progeny cannot be rescued by the maternal products from a heterozygous mother. Bucher and Greenwald (1991) used a genetic mosaic screen to differentiate between zygotic genes required for general metabolic functions (present in every embryonic cell), and those required for embryonic control (present in a subset of embryonic cells). They concluded that the majority of zygotic genes appeared to have specialized functions. This hypothesis is consistent with the dominant role played by maternal gene products, many of which are required for the basic cellular functions of the embryo, leaving 84 zygotic genes to carry out more specialized functions. Wieschaus (1996) suggested that zygotic transcription could enable a more precise localization of transcripts, which would be useful for spatial patterning. He suggested that because evolutionary selection for efficiency is likely to favour maternal products, only a small number of embryonic genes may be required for controlling embryogenesis in organisms displaying rapid development. This appears to be true for C. elegans. Of the 64 vDp2-rescued essential loci identified in the dpy-5 unc-13 region, only four (6%) mutate to egg lethality. One of these, bli-4, encodes a kex2/subtilisin-like endoprotease that may be required for elongation of the nematode embryo (Thacker etal. 1995). The remaining three egg lethal mutants are good candidates for loci involved in C. elegans embryogenesis. The phenotypic characterization of sterile mutants has demonstrated the involvement of 15 loci in germline development, germline sex determination and embryogenesis. The elucidation of the roles that these loci play in the different developmental pathways, and their relationships to other loci, will require further functional studies. The phenotypic characterizations of the mutants in this study expands our knowledge of these loci, which aids in their classification with regards to particular developmental processes, as well as giving direction to future work. Distribution of Sterile Mutations Ten of the 14 essential genes involved in normal fertility of the hermaphrodite germline mapping to the dpy-5 unc-13 region appeared to be randomly distributed. However, four of these essential genes appeared to be clustered near dpy-5 (Figure 27). let-370, let-599, let-355 and let-367 were inseparable from dpy-5 in recombination experiments, and were later separated using duplication mapping experiments (McDowall 1990). These four essential genes appeared to be involved in normal germline development and embryogenesis. About 25% of C. elegans genes are believed to be grouped into operons, where they are coordinately expressed (Spieth et al. 1993; Zorio etal. 1994). The clustering of these four loci makes them good candidates for co-transcribed genes within an operon. M 1,1-370 M P 1,1-599 l«-355 .-11 dpy-5 P G P | 1,1-531 . 1,1-532 | 1,1-345, 1,1-605.1,1-544. 1,1-529, 1,1-534. 1,1-528, P 1,1-397, l,l-39S 1,1-606 1,1-385 P P 1,1-381 1,1-394 I,i513 1,1-384 1,1-396 1,1-387 G M 1,1-399 P G Ui-393 1,1-388 G 1,1-391 1,1-602 1,1-601 1,1-610 ltt-604 1,1-382 1,1-521 1,1-367 1,1-530 1,1-395 ltl-376 lti-378 1,1-379 1,1-377 1,1-514 1,1-603 1,1-512 lrt-611 1,1-380 1,1-608 1,1-386 1,1-390 1,1-607 1,1-392 1,1-383 1,1-543 1,1-86 unc~tO bli-4 unc-37 unc-87 dpy-14 | wtc-14 | 1,1-389 1,1-100 1,1-520 ltm-4 l,i-522 ta-5U 1,1-524 1,1-525 1,1-527 1,1-539 M 1,1-75 1,1-542 1,1-538 unc-lS unc-13 sD/4 hD/8 WpIS hDpU hl>p57 "I I' hDp48 hDp-ll hDp72 hPp37 hDpl7 hl>p61 hDpll hDpU. hDpl6. hDpia. hPp32 hPP31 hPpS6 Figure 27: Distribution of Genes Essential for Fertility in the dpy-5 unc-13 Region The sterile mutants are classified as G (affecting gonadogenesis), P (affecting germ cell proliferation and gametogenesis), or M (maternal effect lethal). hDp39 hDp54 hPpl» hPp33, hPp3S hDpSS "I I" hDpSO sDp2 hPp60 hDp36, hPp-13 hDp64 86 CHAPTER 3 : Cosmid Rescue of Lethal Mutations in the bli-4 (I) Region of Caenorhabditis elegans INTRODUCTION The ability to introduce functional genes into an organism can provide a valuable tool to relate DNA sequences to mutant phenotypes, as well as to investigate their biological function. Germline transformation has been used to characterize a wide variety of genes in different organisms, including Drosophila (Rubin and Spradling 1983), Xenopus (Harland and Laskey 1980; Forbes etal. 1983), sea urchin (McMahon etal. 1985), mice (Palmiter and Brinster 1986), and zebrafish (Stuart et al. 1990). In C. elegans, the microinjection of DNA into germ cells was pioneered by Kimble etal. (1982). Different methods have been developed for introducing DNA into C. elegans germ cells; DNA has been injected into the cytoplasm in the distal gonad arms (Kimble etal. 1982; Stinchcomb etal. 1985; Mello etal. 1991), as well as directly into oocyte nuclei (Fire 1986). The transgenic DNA can either be integrated into the genome (Fire 1986; Mello etal. 1991), or be assembled into an extrachromosomal array (Stinchcomb et al. 1985; Mello et al. 1991). Evidence suggests that both integrated and extrachromosomal transgenes can function close to the level of endogenous loci, and are capable of being expressed in the correct tissues (Fire 1986; Way and Chalfie 1988; Fire and Waterston 1989; Kim and Horvitz 1990; Mello et al. 1991). Furthermore, the offspring of transgenic animals can continue to express the integrated or extrachromosomal transgenes (Fire 1985; Mello etal. 1991), making them suitable for genetic analysis. 87 Transgenes can be used to phenotypically rescue genetic mutations in C. elegans by supplying a wild-type copy of a mutant gene. These cosmid rescue experiments correlate mutant phenotypes with transforming DNA, thereby facilitating access to the sequencing information currently being generated (Waterston and Sulston 1995). Consequently, loci become anchored on the physical map, and as such are convenient guideposts for positioning other genetic markers, in addition to facilitating their cloning. As such, transgenic strains provide a potent means of correlating the genetic and physical maps. The sDp2 'mutant library' is a rich source for identifying developmentally important genes. The systematic rescue of mutants within this library greatly increases its potential, by correlating genetic data with physical data. This chapter describes the use of extrachromosomal cosmid arrays as mapping tools, for increasing genetic mapping resolution, as well as for the correlation of the genetic and physical maps. The technique of germline transformation by cosmid clones was used to rescue lethal mutations in 13 essential genes in the dpy-5 unc-13 region. One of the mutants which was phenotypically rescued was bli-4. bli-4 was first identified by a 32P-induced visible allele e937 (Brenner 1974). e937 is recessive and results in the formation of fluid-filled blisters resulting from the separation of the basal and cortical layers in the cuticle, bli-4 was shown to have an essential function in the development of C. elegans embryos by the recovery of eleven lethal alleles induced by EMS or Tel mutagenesis (Peters etal. 1991). The bli-4 gene is structurally related to the Saccharomyces cerevisiae gene KEX2 and to the mammalian prohormone convertases (Bresnahan etal. 1990; Thacker etal. 1995). Transgenic animals were also used to study the heritability of extrachromosomal arrays. The data presented here suggests that in the hermaphrodite, extrachromosomal arrays are predominantly transmitted through sperm. 88 Chapter 3: MATERIALS AND METHODS The source of mutations and chromosomal rearrangements is given in Chapter 1: Materials and Methods. Preparation of Cosmid DNA for Germline Transformation: All the cosmids used in this study were kindly provided by A. Coulson and J. Sulston of MRC, Cambridge, England (Figure 28). All the cosmid clones contained C. elegans genomic DNA. Cosmid and plasmid amplification and purification by cesium chloride gradient followed the standard protocol of Sambrook et al. (1989), except that the phenol extraction was omitted, because the residue could be toxic to the injected worms. After cesium chloride gradient centrifugation, the ethidium bromide was removed from the DNA solution (Sambrook et cd. 1989), followed by ethanol precipitation of the DNA, and resuspension in lOOul of lOmM Tris ImM EDTA, pH 7.5. DNA concentrations were determined by UV spectrometry. The appropriate amount of DNA was mixed with injection solution (Fire 1986) prior to injection to a final concentration of approximately lOOug/ml (actual concentrations injected are given in Table 6). Germline Transformation: The syncytial injection technique was used for introducing the DNA into the cytoplasmic core in the bend in both gonad arms of the hermaphrodite following the protocols of Fire (1986) and Mello et al. (1991). Wild-type hermaphrodites and hermaphrodites homozygous for unc-11 (e47) were injected with cosmid/plasmid DNA mixtures, containing one to four different cosmids with overlapping sequences, and the plasmid pRF4 (Table 6). The plasmid pRF4 carries the dominant cuticle mutation rol-6(su!006), and was used to identify successful injections. pRF4 also carries an ampicillin-resistance gene. The cosmids used in this study employed two types of 89 vectors: the Lorist vector with kanomycin-resistance, and the pJB8 vector with ampicillin-resistance. The pRF4 plasmid shares sequence identity with the pJB8 vector, but not with the Lorist vector. It is important to have a stretch of sequence identity between the cosmids and plasmids to help ensure their co-transformation (Mello et al. 1991). Therefore, Lorist vector cosmids were always co-injected with pJB8 vector cosmids that had overlapping insert sequences to help ensure their co-transformation with the plasmid pRF4. In this way, the pJB8 and Lorist vector cosmids would share insert sequence identity, while the pJB8 cosmid and the pRF4 plasmid would share vector sequence identity. The only exception was with the Lorist vector cosmid T23H2, for which there was no pJB8 vector cosmid with an overlapping insert sequence. T23H2 was injected with a modified form of the plasmid pRF4 that was both ampicillin-resistant and kanomycin-resistant (kindly supplied by B. Barbazuk, Simon Fraser University). The kanomycin-resistance gene in the modified pRF4 plasmid provided sequence identity with the kanomycin-resistant T23H2 cosmid. The progeny from the injected Unc-11 hermaphrodites were screened for the presence of Rol-6 Unc-11 hermaphrodites. Often the Rol-6 Unc-11 transformants were transients, failing to segregate any progeny carrying the transforming DNA. Therefore, the Rol-6 Unc-11 worms were picked and allowed to self-fertilize, to determine if they segregated Rol-6 Unc-11 progeny, indicating that the injected DNA was stably inherited (either integrated into the genome, or as an extrachromosomal array, the former displaying Mendelian inheritance). Each transgenic strain was assigned either a hEx or a his number to represent an extrachromosomal array or integrated sequence, respectively. Figure 28: Physical Map of the Cosmid Contig used in this Study, showing the extent of the extrachromosomal arrays I hEx9 1 I- hE.x33 1 . I hEx26 1 \hExl6\ I -hEx36 1 I— hEx39 1 \-hEx32—\ C28A7 F40D12 M01A12 C04F1 C03E6 C30A11 F53D11 ZK411 C27D2 C40A4 C38F7 K04F10 C24H1 V101A10 C12H4 C07F10 C53A11 C30B6 C34G6 C32E7 C06A5 C27A12 C46H8 C32G12 C48E7 F26B1 ZC338 B0342 C30A6 ' M01E11 T02D7 C10F11 C29F10 C03F7 C44D11 T23H2 dpy-5 unc-40 bli-4 j hEx28- I I hExlO 1 I hEx44 1 I hEx27 1 I hEx34 1 \-hEx4-\ \hEx35\ s Table 6: Plasmid and Cosmid DNA Used for Transformation DNA Cone. Genotype DNA Cone. Genotype injected (ug/ml) injected injected (ug/ml) injected pRF4 100 wild-type C06A5 24 unc-11 C10F11 8 pRF4 100 unc-11 F53D11 8 __ C27A12 24 ~~C24H1 unc-11 pRF4 17 C28A7 26.7 C30A6 26.7 C46H8 25 unc-11 pRF4 20 C27D2 25 C32G12 25 C28A7 30 unc-11 pRF4 15 M01A10 30 F40D12 30 C32G12 25 unc-11 pRF4 20 C40A4 25 C48E7 25 F40D12 30 unc-11 pRF4 15 M01E11 30 C12H4 30 C48E7 25 unc-11 pRF4 20 C38F7 25 F26B1 25 30 unc-11 pRF4 15 M01A12 30 C07F10 30 K04F10 40 unc-11 pRF4 20 C29F10 40 pRF4 20 C07F10 30 unc-11 C04F1 30 ZC338 40 unc-11 C53A11 30 C03F7 40 pRF4 20 pRF4 20 _ unc-11 T23H2 50 unc-11 C03E6 26 pRF4 40 C30B6 26 pRF4 18 T23H2 60 unc-11 pRF4 30 _3l>E7 unc-11 (AmpR KanR) pRF4 40 92 Segregation Stability: The frequency at which the extrachromosomal arrays were transmitted to hermaphrodite gametes was calculated by scoring the progeny of selfed transgenic Rol-6 hermaphrodites. The ratio of array-bearing (Ex) gametes to non-£x gametes was calculated using the ratio of Rol-6 (transgenic) to non-Rol-6 (non-transgenic) progeny, as outlined in Figure 29. Dpy-5 Unc-13 progeny were not scored, because they were lethal and failed to express Rol-6 (Dpy-5 is epistatic to Rol-6). This formula provided an average estimate for the frequency of array-bearing gametes, but did not distinguish between the transmission frequencies for hermaphrodite sperm and oocytes, which may differ. The frequency of extrachromosomal array-bearing oocytes was calculated by scoring the progeny arising from a cross between transgenic hermaphrodites and non-transgenic males (Figure 29). Only cross progeny were scored, omitting the Dpy-5 Unc-13 lethal class. All array-bearing cross progeny must have received the array from the oocyte, since sperm were derived from non-transgenic males. The frequency of hermaphrodite sperm that carried the extrachromosomal arrays was calculated by comparing the frequencies of array transmission from self- and cross-progeny (Figure 29). Since self-progeny received the array from both the oocyte and the sperm, while cross-progeny received the array from only the oocyte, the difference between the two provides an estimate of the frequency of array-bearing sperm. Figure 29: Equations Used to Calculate Transmission Frequencies Calculation of the frequency of array-bearing gametes: p - frequency of array-bearing (Ex ) gametes 1 - p = frequency of non-Ex gametes Ratio of Ex : non-Ex = x Rol-6 non-Rol-6 2p (1 - p) + p2 (1 -P)2 This equation was derived to: Calculation of the frequency of array-bearing (Ex) oocytes: „ Rol-6 (cross-progeny) E x oocytes = . , , / c — 2 — r -J total (cross-progeny) Calculation of the frequency of array-bearing sperm: E x sperm = 1 - ( non-Rol-6 (self-progeny) ^ total (self-progeny) non-Rol-6 (cross-progeny) \ total (cross-progeny) / 94 Complementation Mapping to the Transgenic Strains: Extrachromosomal arrays can be used like duplications to map genes, by determining whether or not a wild-type copy of a given gene exists in the extrachromosomal array. Complementation would result in the rescue of a lethal phenotype. Thirty-seven lethal mutations and three visible mutations, representing 37 genes, were complementation tested to 15 different transgenic strains, following the protocol in Figure 30. sDp2-balanced, lethal-bearing hermaphrodites were crossed to unc-11 dpy-14; szTl(I;X)[lon-2]males. The male progeny lacking sDp2 (dpy-5 let-x unc-13/unc-l 1 dpy-14) were crossed to the Unc-11 Rol-6 transgenic hermaphrodites (unc-11 ;hExz), where hExz represents the extrachromosomal arrays. All the Rol-6 progeny (i.e. not Unc-11 Rol-6) had the genotype dpy-5 let-x unc-13/unc-l 1 ;hExz. The Rol-6 hermaphrodites were allowed to self-fertilize, and their progeny screened for the presence of viable, fertile Dpy-5 Unc-13s (dpy-5 let-x unc-13;hExz). Only if the cosmids carried a wild-type allele of the lethal mutation being mapped would these Dpy-5 Unc-13 individuals survive. Because dpy-5 is epistatic to rol-6, these Dpy-5 Unc-13 animals will not express the Rol-6 phenotype. Therefore, viable Dpy-5 Unc-13 hermaphrodites were crossed to wild-type males in order to determine if they segregated Rol-6 progeny, indicating that the pRF4 plasmid was still present. 95 Figure 30: Protocol for Complementation Analysis of Transgenic Strains dpy-5 let-x unc-131 dpy-5 let-x unc-13 ; sDp2 X unc-11 dpy-14 ; szTl Unc-13 hermaphrodites Lon-2 males 1 Pick Wild-type males dpy-5 let-x unc-13 I unc-11 dpy-14 X unc-11 ; hExz Unc-11 Rol-6 hermaphrodites (carrying extrachromosomal array) 1 unc-111 unc-11 dpy-14 Unc-11 unc-111 unc-11 dpy-14 ; hExz Unc-11 Rol-6 unc-111 dpy-5 let-x unc-13 'Wild-type' unc-111 dpy-5 let-x unc-13 ; hExz Rol-6 Pick Rol-6 hermaphrodites | s e . f - c r o s s Check for presence of fertile Dpy-5 Unc-13s unc-111 unc-11 Unc-11 unc-111 unc-111 hExz Unc-11 Rol-6 unc-111 dpy-5 let-x unc-13 'Wild-type' unc-111 dpy-5 let-x unc-13 ; hExz Rol-6 dpy-5 let-x unc-131 dpy-5 let-x unc-13 Dead dpy-5 let-x unc-131 dpy-5 let-x unc-13 ; hExz Dpy-5 Unc-13 or Dead (rescue) (no rescue) 96 Polymerase Chain Reaction (PCR) Analysis of Transgenic Worms: The presence of the Lorist and/or pJB8 vector in transgenic worms was confirmed using PCR of whole worm lysate. Two or three Unc-11 Rol-6 or Dpy-5 Unc-13 worms were lysed using proteinase K (Pharmacia), followed by multiplex PCR, as described by Barstead et al. (1991). Two sets of primers were used for PCR, and were obtained from S. McKay (this lab) (Appendix 5). RL12 and RL14 are left and right primers that gave a 580 base pair amplification product from the C. elegans adenosyl homocysteine hydrolase gene (AHH) gene (Prasad et al. 1993). A H H is present in the C. elegans genome, but not in the cosmids used in this study. This set of primers was use as an internal positive control. The second set of primers were RL17 and RL18. These primers gave an 880 base pair amplification product of sequences specific to the Lorist cosmid vectors. PCR using wild-type and no worms were performed as negative controls. 97 Chapter 3: RESULTS Germline Transformation: Germline transformation of C. elegans oocytes with cosmid and plasmid DNA was used to generate heritable transgenic strains, suitable for genetic analysis. DNA injections were made into wild-type and Unc-11 hermaphrodite gonads, in order to establish stocks suitable for complementation for both lethal and visible mutations. Thirty overlapping cosmid clones containing C. elegans genomic DNA were injected in 14 groups of one to four cosmids, along with the plasmid pRF4. From a total of 1112 hermaphrodites injected with DNA, 143 (13%) produced transgenic progeny. The majority of transgenic progeny (238/279, or 85%) showed transient expression of the array, failing to segregate the extrachromosomal array to their offspring. This could be due to injected DNA which has not yet formed arrays. The remaining transgenic progeny (41/279, or 15%) were heritably transformed, producing from 1% to 74% transgenic progeny from self-fertilization. The concentration of DNA injected into the hermaphrodite gonads was found to influence the number and type of transgenic progeny produced. Injections using concentrations of DNA of 75 1 85 ug/ml resulted in a high frequency of transgenic progeny. However, the majority of these progeny showed transient expression of the array, heritable transformants rarely being produced. Injections using higher concentrations of DNA (90 - 110 ug/ml) resulted in a low frequency of transgenic progeny, but a higher proportion of these progeny were heritably transformed. Injections'using DNA concentrations in excess of 110 ug/ml resulted in sick, often sterile transgenic progeny that frequently displayed morphological abnormalities such as shortened length, dumpy appearance and increased girth in the center of the animal, regardless of the type of cosmids used. From the 41 heritable transgenic lines constructed, 15 strains produced less than 5% transgenic progeny from self-fertilization. The extrachromosomal arrays were subsequently lost from these strains. The remaining 26 strains carrying heritable extrachromosomal arrays were 98 maintained as stable stocks. These 26 strains represent the 14 different groups of cosmid/plasmid DNA used for injections (Figure 28; Table 7). These heritable lines segregated the dominant Rol-6 phenotype of pRF4 as extrachromosomal arrays, and displayed non-Mendelian transmission of the arrays. In addition, three heritable strains were generated by transformation with only pRF4 plasmid DNA. These strains have the genotype unc-1 l;pRF4 (hEx7, hEx8) and +;pRF4 (hEx6), and were created to use as negative controls. In one of these strains, KR2225 hEx7, the extrachromosomal DNA spontaneously integrated into the genome, giving rise to the strain KR24l4hIsl. For certain transgenic strains, a fraction of the transgenic hermaphrodites were completely sterile, failing to produce any fertilized eggs. The remaining transgenic hermaphrodites produced normal brood sizes. In the strain KR2362 hEx26, 13% (8/63) of the transgenic hermaphrodites were completely sterile. Two strains displayed the highest frequency of sterile hermaphrodites, KR2360 hEx5 and KR2365 hEx!6. In both strains approximately 50% of the transgenic hermaphrodites were sterile. However, the strain bearing hExl6 subsequently became very stable, with all the transgenic hermaphrodites being completely fertile and having normal brood sizes. Care was taken when maintaining semi-sterile strains to ensure that an adequate number of transgenic hermaphrodites were transferred to fresh plates. Table 7: Transgenic Strains Constructed Strain Array Genotype Transforming DNA Phenotype KR2224 hEx6 +;hEx6 pRF4 Rol -6 KR2225 hEx7 unc-1 l;hEx7 pRF4 Unc-11 Rol-6 KR2363 hEx8 unc-11;hEx8 pRF4 Unc-11 Rol-6 KR2414 hisl unc-11 ;hlsl pRF4 Unc-11 Rol-6 KR2263 hEx9 unc-11 ;hEx9 pRF4, C24H11, C28A7, C30A6 Unc-11 Rol-6 KR2268 hEx28 unc-11 ;hEx28 pRF4, C28A7, M01A10, F40D12 Unc-11 Rol-6 KR2401 hEx29 unc-11 ;hEx29 pRF4, C28A7, M01A10, F40D12 Unc-11 Rol-6 KR2502 hEx33 unc-1 l;hEx33 pRF4, F40D12, M01E11, C12H4 Unc-11 Rol-6 KR2657 hEx38 unc-11 ;hEx38 pRF4, F40D12, M01E11, C12H4 Unc-11 Rol-6 KR2288 hExlO unc-1 l;hExlO pRF4, C12H4, M01A12, C07F10 Unc-11 Rol-6 KR2409 hEx30 unc-1 l;hEx30 pRF4, C12H4, M01A12, C07F10 Unc-11 Rol-6 KR2361 hExlS unc-11 ;hEx15 pRF4, C07F10, C04F1, C53A11 Unc-11 Rol-6 KR2362 hEx26 unc-11 ;hEx26 pRF4, C07F10, C04F1, C53A11 Unc-11 Rol-6 KR2818 hEx44 unc-11 ;hEx44 pRF4, C53A11, C03E6, C30B6 Unc-11 Rol-6 KR2365 hExl6 unc-11 ;hExl 6 pRF4, C32E7 Unc-11 Rol-6 KR2400 hEx27 unc-1 l;hEx27 pRF4, C06A5, C10F11, F53D11, C27A12 Unc-11 Rol-6 KR2817 hEx36 unc-11 ;hEx36 pRF4, C46H8, C27D2, C32G12 Unc-11 Rol-6 KR2503 hEx34 unc-11 ;hEx34 pRF4, C32G12, C40A4, C48E7 Unc-11 Rol-6 KR2658 hEx39 unc-11 ;hEx39 pRF4, C48E7, C38F7, F26B1 Unc-11 Rol-6 KR2276 hEx4 unc-1 l;hEx4 pRF4, K04F10, C29F10 Unc-11 Rol-6 KR2360 hEx5 unc-1 l;hEx5 pRF4, K04F10, C29F10 Unc-11 Rol-6 100 Table 7: continued Strain Array Genotype Transforming D N A Phenotype KR2419 hEx31 unc-11 ;hEx31 pRF4, ZC338, C03F7 Unc-11 Rol-6 KR2420 hEx32 unc-11 ;hEx32 pRF4, ZC338, C03F7 Unc-11 Rol-6 KR2386 hExll unc-11; hExll pRF4, T23H2 Unc-11 Rol-6 KR2820 hEx41 unc-11 ;hEx41 pRF4, T23H2 Unc-11 Rol-6 KR2813 hEx42 unc-11 ;hEx42 pRF4, T23H2 Unc-11 Rol-6 KR2558 hEx37 unc-11 ;hEx37 pRF4 (AmpR KanR), T23H2 Unc-11 Rol-6 KR2814 hEx43 unc-1 l;hEx43 pRF4 (AmpR KanR), T23H2 Unc-11 Rol-6 KR2699 hEx40 unc-11 ;hEx40 pRF4 (AmpR KanR), T23H2 Unc-11 Rol-6 KR2504 hEx35 unc-11 ;hEx3 5 pRF4 (AmpR KanR), T23H2 Unc-11 Rol-6 101 Segregation Stability of the Extrachromosomal Arrays: The frequency of transmission of the extrachromosomal arrays to the hermaphrodite's gametes was calculated for 26 different array-bearing strains, and one integrated strain. The progeny from the self-fertilization of transgenic hermaphrodites were scored. For the strains tested, self-fertilized transgenic hermaphrodites produced between 7 and 74% Rol-6 transgenic progeny (Table 8). The ratio of transgenic to non-transgenic progeny was used to calculate the gametic frequency of extrachromosomal array transmission (see Materials and Methods). These results are summarized in Table 9. Each extrachromosomal array displayed a different transmission value, which remained characteristic for that array over several generations, fluctuating only slightly. Genetic background did not appear to affect the transmission frequencies of the arrays. The transmission of arrays to hermaphrodite gametes was found to vary between 4% {hEx42) and 32% (hExl6). The transgenic strain transformed with pRF4 alone (hhl) transmitted the dominant Rol-6 phenotype in a 3:1 ratio to its self-progeny, passing the transgene on to approximately 50% of gametes. The genetic behaviour of the marker in this strain was consistent with that of an allele linked to an endogenous chromosome, suggesting that the transforming sequences in this line have been integrated into the genome. The calculation for gametic frequency assumes an equal contribution from sperm and oocyte. However, the transmission frequency of the array to progeny was generally found to decrease substantially when transgenic hermaphrodites were crossed to non-transgenic males, as shown in Table 8. The decrease in transmission frequency suggests an unequal contribution from sperm and oocyte. The fraction of oocytes and sperm carrying the array was determined for 16 array-bearing strains and one integrated strain. The fraction of array-bearing oocytes was determined by scoring the cross-progeny arising from mating transgenic hermaphrodites to non-transgenic males. The frequency of extrachromosomal array-bearing cross-progeny was equivalent to the frequency of oocytes that carried the array, since sperm were derived from non-transgenic males. The results summarized in Table 9 show the frequency of extrachromosomal array-bearing oocytes to be lower than the average gametic frequency for 12 of the 16 array-102 bearing transgenic strains. The remaining four array-bearing strains, and the integrated strain, showed equivalent values for the frequency of Ex gametes and Ex oocytes. The fraction of hermaphrodite sperm carrying the extrachromosomal arrays was determined by comparing the progeny arising from transgenic hermaphrodites that were self-fertilized to those that were mated with males lacking arrays. As shown in Table 9, the extrachromosomal arrays were often recovered at a higher frequency in hermaphrodite sperm than in oocytes. For two of the 16 array-bearing strains, hEx5 and hEx32, the transmission frequency of Ex sperm was seven times higher than the frequency of Ex oocytes. For one strain hEx39, the transmission frequency of Ex sperm was four times higher than the frequency of Ex oocytes. For nine strains, hEx4, hEx9, hExlO, hExl6, hEx26, hEx28, hEx33, hEx35, and hEx36, the transmission frequency of Ex sperm was between two and three times higher than the frequency of Ex oocytes. The four remaining array-bearing strains, hExll, hEx31, hEx34 and hEx44, and the one integrated strain, hlsl, showed approximately equal frequencies of Ex sperm and Ex oocytes. These results suggested a more frequent loss of the array during oogenesis, or oocyte-specific meiotic processes, than during spermatogenesis in several of the extrachromosomal array-bearing strains examined. Thus, transmission of the array in these strains occurred predominantly through the sperm. Transgenic males were not examined for transmission frequencies, because Rol-6 males are very difficult to mate. Therefore, these results cannot be extended to male sperm. A comparison was made between transgenic strains that carried the same cosmid DNA to determine if this phenomenon was sequence-specific. Comparisons were made between hEx4 and hEx5 (K04F10, C29F10), between hExlO and hEx30 (C12H4, M01A12, C07F10), and between hEx31 and hEx32 (ZC338, C03F7). In each case, the discrepancy between the transmission frequency to oocytes and sperm varied between strains carrying the same cosmid DNA. hEx5 showed a similar frequency of Ex gametes (21%) to hEx4 (18%), which suggested that the two arrays were equally stable, yet the hEx5 array was less stable in oocytes (5%) than the hEx4 array (11%). This resulted in a greater disparity between the frequency of Ex oocytes and Ex sperm for hEx5 than for hEx4, with Ex oocytes being 7-fold and 2.3-fold lower, respectively. In another 103 case, hEx32 showed a higher frequency of Ex gametes (21%) than hEx3](\3%), but was less stable in oocytes (5%) than hEx31 (12%), resulting in a greater disparity between the frequency of Ex oocytes and Ex sperm for hEx32 than for hEx31, with Ex oocytes being 7-fold and 1.2-fold lower, respectively. In contrast, when comparing hExlO and hEx30, hEx30 had a higher frequency of Ex gametes (31%) than hExlO (21%), and was also found to be more stable in oocytes (27%), than hExlO (11%). Less disparity was found between the frequency of Ex oocytes and Ex sperm for hEx30 and hExlO, with Ex oocytes being 1.3-fold and 2.6-fold lower, respectively. These results argue against specific sequences within the cosmid DNA being responsible for the differences in array transmission frequencies in oocytes and sperm. However, it should be noted that two strains, hEx5 and HEx32, displayed the greatest difference between the frequency of Ex sperm and Ex oocytes, both producing seven times fewer Ex oocytes. These two strains should carry DNA sequences in common, since the cosmids K04F10 and C29F10 in hEx5 overlap the cosmid ZC338 in hEx32. A specific sequence contained in these two arrays might influence array transmission. For the 17 transgenic strains examined, a fraction of transgenic hermaphrodites failed to transmit the array to any of their progeny. The fraction of transgenic hermaphrodites failing to segregate any array-bearing progeny is summarized in Table 10. Loss of the array varied between 1% for hExl6 and hEx34, and 23% for hEx9. Presumably, the complete lack of array transmission occurred due to mitotic loss prior to gametogenesis. Pre-meiotic loss of the array did not occur more often in selfed hermaphrodites than in those mated to non-transgenic males (data not shown), so both classes were used to calculate the frequency of mitotic loss. In the strain hisl, pre-meiotic loss of the plasmid sequences was never observed among the hermaphrodites tested, consistent with the DNA being integrated into the genome. Many of the transgenic strains showed an increase in transmission frequency over a several generations. This increase usually occurred suddenly, resulting in a new, consistent transmission frequency for that transgenic strain. A decrease in transmission frequency was not observed. Tables 8,9 and 10 give an estimate of the average transmission frequencies. Table 11 summarizes 104 the changes in transmission frequencies recorded for five transgenic strains (sufficient data was not recorded to show the change in transmission frequencies for most of the strains examined). For strains hEx9 and hEx29, sufficient data was only available to compare the frequency of Ex gametes in the low- and high-transmission lines. However, both the low- and high-transmission lines of hEx9 were frozen in liquid nitrogen and are available for further studies. Four of the five strains (HEx4, hEx9, hEx29 and HEx39) examined showed an increase in the frequency of Ex gametes. For the strains hEx4 and hEx39, the increase in gametic frequency of transmission of the array is reflected by an increase in the frequency of Ex oocytes. The 1.5-fold increase in Ex gamete frequency in the strain hEx4 is reflected by a 4-fold increase in array transmission frequency to oocytes, but no increase to sperm. The 2.1-fold increase in the frequency of Ex gametes in the strain hEx39 is reflected by a 13-fold increase in the frequency of Ex oocytes, but only a 1.5-fold increase in Ex sperm. By contrast, the strain hEx32 shows no increase in the average gametic frequency of transmission of the array. Instead, this strain displays an increase in the frequency of Ex oocytes (12-fold) and a concomitant decrease in the frequency of Ex sperm (1.2-fold). These results indicate that increased array stability is primarily occurring in oocytes. However, it should be noted that in all three strains, hEx4, hEx32 and hEx39, the array transmission frequency to sperm is still higher than to oocytes. In general, the overall stability of extrachromosomal arrays was consistent within individual strains, but could change over time to give a new characteristic transmission frequency, reflected by increased array stability in oocytes. Twelve of the 16 extrachromosomal arrays examined were found to be transmitted 2- to 7-fold more frequently in hermaphrodite sperm than in oocytes. No difference was found between the frequency of transmission of Er-oocytes and Ex-sperm in the integrated strain. Early mitotic loss of the arrays was observed in all 16 extrachromosomal array-bearing strains examined, but not in the integrated strain, suggesting that loss of the array can occur at any time during development when cell divisions are occurring. 105 Table 8: Extrachromosomal Array Stability I: Frequency of Rol-6 Progeny A r r a y Self-progeny* 1 Cross - progeny* Frequency of Rol -6 R o l - 6 n o n - R o l - 6 R o l - 6 n o n - R o l - 6 se l f -progeny cross -progeny hlslc 3231 1135 281 343 0 .74 0 .45 hEx4 3269 6639 359 2932 0 .33 0 .11 hEx5 163 279 7 140 0 .37 0 .05 HEx9 1426 3508 114 981 0 .29 0 .10 hExlO 3575 6075 284 2324 0 .37 0 .11 hExll 7205 6340 437 1157 0 .53 0 .27 hEx!5 182 284 - - 0 .39 -hExl6 3442 2969 191 927 0 .54 0 .17 hEx26 2319 3953 210 1564 0 .37 0 .12 hEx27 12 151 0 .07 hEx28 1214 2542 169 1478 0.32 0 .10 hEx29 81 638 - - 0.11 -hEx30 72 462 - - 0 .13 -hEx31 377 1199 35 252 0 .24 0 .12 hEx32 2174 3604 116 2042 0 .38 0 .05 hEx33 3846 6167 287 1868 0 .38 0 . 1 3 hEx34 2076 7804 207 1757 0.21 0 .11 hEx35 6110 10306 419 2155 0 .37 0 .16 hEx36 1166 1894 94 561 0 .38 0 .14 hEx37 104 92 - - 0 .53 -hEx38 30 209 - - 0 . 1 3 -hEx39 1627 3362 208 2949 0 .33 0 .07 106 Table 8: continued Self-progeny" Cross-progeny* Frequency of Rol-6 Array Rol-6 non-Rol-6 Rol-6 non-Rol-6 self-progeny cross-progeny hEx40 5 39 - 0.11 hEx41 80 221 - •. 0.27 hEx42 43 462 - • 0.09 hEx43 11 39 - 0.22 hEx44 53 418 16 237 0.11 0.06 "Genetic background: unc-11 + ++/+ dpy-5 let-x unc-13;hExz or unc-1 l;hExz fc,Cross: unc-11 ;hExz X + dpy-5 let-x unc-13/unc-ll + + dpy-14 + cOnly heterozygous hlsl hermaphrodites were used in this experiment Table 9: Extrachromosomal Stability II: Frequency of Hermaphrodite Array-bearing Gametes A r r a y Frequency of Ex gametes" Frequency of Ex oocytes" Frequency of Ex sperm" hlsl 0.49 0.45 0.53 hEx4 0.18 0.11 0.25 hEx5 0.21 0.05 0.34 hEx9 0.16 0.10 0.21 hExlO 0.21 0.11 0.29 hExll 0.31 0.27 .0.36 hExl5 0.22 - -hExl6 0.32 0.17 0.45 hEx26 0.21 0.12 0.28 hEx27 0.04 hEx28 0.18 0.10 0.24 hEx29 0.06 - -hEx30 0.07 - -hEx31 0.13 0.12 0.14 hEx32 0.21 0.05 0.35 hEx33 0.21 0.13 0.29 hEx34 0.11 0.11 0.11 hEx35 0.21 0.16 0.25 hEx36 0.21 0.14 0.28 hEx37 0.31 - -hEx38 0.06 - -hEx39 0.18 0.07 0.28 108 Table 9: continued Array Frequency of Ex gametes" Frequency of Ex oocytes" Frequency of Ex sperm" hEx40 0.06 - -hEx41 0.14 - -hEx42 0.04 - -hEx43 0.12 - -hEx44 0.06 0.06 0.05 "Data from Table 8 used to calculate these values, using the equations in Figure 26. 109 Table 10: Extrachromosomal Array Stability III: Frequency of Early Mitotic Loss Array Hermaphrodites failing to segregate array*1 Total Hermaphrodites" Frequency of loss of array hlsl 0 37 0.00 hEx4 20 168 0.12 hEx5 1 9 0.11 hEx9 17 75 0.23 hExlO 9 141 0.06 hExll 5 127 0.04 hExl6 1 83 0.01 hEx26 4 121 0.03 hEx28 10 87 0.11 hEx31 2 12 0.17 hEx32 7 47 0.15 hEx33 5 • 180 0.03 hEx34 1 120 0.01 hEx35 5 199 0.03 hEx36 2 43 0.05 hEx39 5 49 0.10 hEx44 1 6 0.17 "Includes selfed and mated hermaphrodites, since the frequency of early loss of an array did not vary between mated and non-mated individuals. 110 Table 11: Changes in Transmission Frequency Array Self- progeny Cross -progeny Frequency of Ex Rol-6 non-Rol-6 Rol-6 non-Rol-6 gametes oocytes sperm hEx4 a1 39 118 57 1336 0.13 0.04 0.22 b2 3132 6051 302 1594 0.19 0.16 0.21 hEx9 a 176 497 - - 0.14 - -b 199 111 - - 0.40 - -hEx29 a 81 638 - - 0.06 - -b 59 5 - - 0.72 - -hEx32 a 342 551 18 1338 0.21 0.01 0.37 b 1688 2603 98 704 0.22 0.12 0.31 hEx39 a 53 261 22 1708 0.09 0.01 0.16 b 1525 3012 186 1241 0.19 0.13 0.24 1 Initial strain designated (a). 2 Strain after change in transmission frequency observed designated (b). I l l Rescue of Lethal Mutations by Complementation with Extrachromosomal Arrays: Germline transformation of non-lethal hermaphrodites was used to generate heritable array-bearing strains suitable for complementation with lethal and visible strains. The cosmids were injected in groups of one to four cosmids, so as to reduce the necessary number of complementation tests. The cosmids were grouped into overlapping sets, each group carrying one cosmid in common with the preceding group, and one cosmid in common with the following group of cosmids. For example, if cosmids A , B and C were co-injected as a group, then the next group of cosmids would be C, D and E, and the next group would be cosmids E, F and G, and so forth. The group containing cosmids C, D and E shares cosmid C in common with the preceding group, and shares cosmid E in common with the following group. The use of overlapping sets of cosmids enables the phenotypic rescue of mutations by array-bearing strains to be associated with single cosmids. For instance, if two array-bearing strains sharing one cosmid in common both rescue the lethal phenotype of a particular mutation, then the shared cosmid must be responsible for the rescued phenotype. It also enables the rescue of lethal mutations that are only partially represented in a cosmid, because a locus that has been split over two cosmids can often reform a wild-type locus when the two cosmids recombine to form a concatemer. However, it should be noted that the recombining of cosmids to form concatemers could also alter loci, thereby affecting their ability to rescue lethal mutations. The genetic map generated in Chapter 1 was used to help determine which array-bearing strains should be complementation tested to each mutation. When a transgenic strain was found to rescue a mutation, then the mutations mapping to the left of the rescued mutation would not be tested to strains carrying cosmids physically mapping to the right of the rescuing transgenic strain. Similarly, mutations mapping to the right of the rescued mutation would not be tested to strains carrying cosmids mapping to the left of the rescuing transgenic strain. Once a mutation was phenotypically rescued by complementation with a transgenic strain(s), the mutation would also be tested against strains carrying cosmids mapping to the left and 112 right of the rescuing transgenic strain, so as to obtain negative results. This was done to ensure that the rescued phenotype was not an artifact of the extrachromosomal arrays themselves. Fourteen different transgenic strains carrying cosmids as extrachromosomal arrays were used for complementation with 37 lethal mutations and two visible mutations, representing 37 genes. A total of 15 mutations were rescued, representing 13 genes. The lethal/visible phenotypes of eight mutations (6 genes) were completely rescued by complementation with array-bearing strains. The lethal phenotypes of the remaining seven mutations (7 genes) were only partially rescued by complementation to array-bearing strains. The results of the complementation tests are summarized in Table 12. 113 Table 12: Transgenic Complementation Results hEx9 hEx28 hEx33 hExlO hEx26 hEx44 hExl6 IIEX27 hEx36 hEx34 hEx39 hEx4 hEx32 hEx35 dpy-5 -let-599 - - -let-370 - - -let-355 - - - -let-367 - - + + -let-530 - - +P -lel-395 - - - +P - -let-376 - - - +P -lel-531 -let-532 + let-378 - + + - -let-393 - - -let-513 - - - - -let-379 - - -let-377 -lel-388 let-514 - - -let-603 - - +P let-384 - - -let-391 - - • -let-512 - - -let-396 - - - -let-602 +P - -let-611 - +P -let-381 - + - - -tbli-4 - + - -let-601 - + -let-380 - - - -let-387 - - - - -let-606 - - - - -let-608 - - -let-610 - - -let-386 -let-604 -let-382 -lel-390 -let-607 -Shaded areas represent complementation tests not required on the basis of the genetic map position oi" the mutations. Underlined genes are the adult stcriles characterized in Chapter 2. If complementation tests with bli-4 were done using the alleles Ii42, Ii754 and e937. ''Denotes partial rescue of the lethal penotype. (i) Complete Rescue of Lethal Phenotypes: Eight mutations, representing six genes, showed complete rescue of their lethal/visible phenotypes when complemented with transgenic strains. For the mutations linked to dpy-5 and unc-13, each rescued strain (dpy-5 let-x unc-13;hExz) produced both fertile Dpy-5 Unc-13 progeny and lethal progeny. The fertile Dpy-5 Unc-13 progeny did not display a Rol-6 phenotype, because dpy-5(e61) is epistatic to rol-6(sul006). However, each of the rescued strains segregated Rol-6 progeny when mated to N2 males. For the visible mutation bli-4(e937), the rescued strain (bli-4(e937);hExz) produced both Rol-6 and Bli-4 progeny. In each case, the proportion of lethal (or visible) to viable (or non-visible) progeny correlated to the array transmission frequency expected for the rescuing transgenic strain. The strain information for each of the rescued lines is summarized in Table 13. The transgenic strains bearing the extrachromosomal arrays hEx4 and hEx5 (K04F10, C29F10, pRF4) each complemented the three complementation groups of bli-4 (represented by the two lethal alleles h42, and h754, that arrest during hatching, and the visible blistering allele e937) (Peters etal. 1991). The cosmid K04F10 has been used for the cloning of bli-4 (Thacker et al. 1995). The sequence of the cosmid C29F10 is contained within the larger K04F10 cosmid, and was included with the injected DNA to provide vector sequence identity with the pRF4 plasmid. The complementation results provided evidence that the K04F10 cosmid contains the bli-4 locus. The lethal phenotypes of two mutations were each rescued by complementation to two transgenic strains. The adult sterile phenotype of the mutation let-367(hl 19) was rescued by the transgenic strains bearing hExlO or hEx33, which share a single cosmid in common, C12H4. The early larval lethal phenotype of the mutation let-378(hl24) was rescued by the transgenic strains bearing hExlO or hEx26, which share the cosmid C07F10 in common. These results suggested that in each case a single cosmid was responsible for the rescued sterile phenotype. The lethal phenotypes of three mutations were each rescued by complementation to one transgenic strain. The early larval lethal phenotype of the mutation let-381(hl07) was rescued by the transgenic strain bearing hEx39, but failed to be rescued by the strains bearing hEx4 or hEx34. 115 hEx34 shared one cosmid in common with hEx39, while hEx4 did not share any common cosmids with hEx39, but did overlap the F26B1 cosmid. These results suggested that either/both of the overlapping cosmids unique to hEx39, C38F7 and F26B1, could be responsible for the rescue of the sterile phenotype. The egg lethal phenotype of the mutation let-532(h715) was rescued by the transgenic strain bearing hEx26. This result suggested that one of the cosmids in the hEx26 array, C07F10, C04F1, or C53A11, was responsible for the rescued sterile phenotype. The egg lethal phenotype of the mutation let-601(h281) was rescued by the transgenic strain bearing hEx32, but failed to be rescued by the strains bearing hEx4 or hEx35. Neither hEx4, nor hEx35 shared cosmids in common with hEx32, however hEx4 did overlap the ZC338 cosmid in hEx32 (there was a gap in the cosmid contig between the cosmids in the array hEx32 and in the array hEx35). These results suggested that either/both of the cosmids in the hEx32 array, ZC338 and C03F7, was responsible for the rescued sterile phenotype. 116 Table 13: Rescued Transgenic Strains Strain Genotype Phenotype KR2356 dpy-5 bli-4(h754) unc-13 ; hEx5 Dpy-5 Unc-13 KR2366 dpy-5 bli-4(h42) unc-13; hEx4 Dpy-5 Unc-13 KR2692 bli-4(e937); hEx4 Rol-6 KR2841 dpy-5 let-367(hll9) unc-13 ;hExlO Dpy-5 Unc-13 KR2559 dpy-5 let-367(hll9) unc-13 ; hEx33 Dpy-5 Unc-13 KR2678 dpy-5 let-378(hl24) unc-13 ;hExlO Dpy-5 Unc-13 KR2679 dpy-5 let-378(h!24) unc-13 ; hEx26 Dpy-5 Unc-13 KR2815 dpy-5 let-381(h!07) unc-13 ; hEx39 Dpy-5 Unc-13 KR2650 dpy-5 let-532(h715) unc-13 ;hEx26 Dpy-5 Unc-13 KR2509 dpy-5 let-601(h281) unc-13 ; hEx32 Dpy-5 Unc-13 117 (ii) Partial Rescue of Lethal Phenotypes: Seven mutations, representing seven genes, showed partial rescue of their lethal phenotypes when complemented with transgenic strains. Both lethal (non-rescued) (dpy-5 let-x unc-13) and partially rescued (dpy-5 let-x unc-13;hExz) Dpy-5 Unc-13 hermaphrodites were examined under Nomarski optics to help characterize the different phenotypes. Adult Dpy-5 Unc-13 hermaphrodites were used for comparison (Figure 10). When complemented with the rescuing transgenic strains, four early arresting lethal strains survived to adulthood and produced arrested eggs or dead LI larvae, one mid arresting lethal strain survived to adulthood but was sterile, and two sterile adult lethal strains were able to produce arrested eggs or dead LI larvae. The change in phenotype of the partially rescued lethal strains are summarized in Table 14. Each of the partially rescued strains produced two types of Dpy-5 Unc-13 progeny: those that arrested in development at the stage characteristic for that mutation (not rescued), and those that arrested at a later stage in development (partially rescued). The proportion of early to later arresting Dpy-5 Unc-13 progeny correlated to the array transmission frequency expected for the rescuing transgenic strain. These results suggested a correlation between the presence of the array and the increased survival of the arresting mutant. The transgenic strain bearing the extrachromosomal array hExlO partially rescued the lethal phenotypes of three mutations, let-376(hl30), let-395(h271) and let-530(h798). These mutations failed to be complemented by the adjacent arrays, hEx26 and hEx33, which shared cosmids in common with the hExlO array. These results suggested that the cosmid that is unique to the hExlO array, M01A12, was responsible for the altered lethal phenotype of all three mutations. The complementation of other lethal mutations with the nExiO-bearing strain suggested that the hExlO array carried all three cosmids (see Table 12). The lethal mutation let-376(h!30) caused developmental arrest at the early larval stage in homozygous animals (Figure 3 ld,e,f). The presence of the array hExlO, enabled Dpy-5 Let-376 Unc-13 hermaphrodites to develop into adults, most of which produced eggs (Figure 31a,b,c). The eggs were rarely laid (two animals laid two eggs each) and the majority of eggs did not hatch. 118 Occasionally, internal hatching occurred, where five or six early larvae were detected inside the adult (Figure 31a). These early larvae died within the adult, and no 'bag of worms' phenotype was seen. Some hermaphrodites showed more severe defects in gametogenesis (Figure 31b). Homozygous let-530(h798) animals normally arrested in development as early larvae (Figure 32c). The presence of hExlO allowed Dpy-5 Let-530 Unc-13 hermaphrodites to develop into adults that produced eggs, however some variability in phenotype was observed (Figure 32a,b). If eggs were produced by the adult, then they were never laid and failed to hatch, but did develop past cleavage to the two-fold stage (Figure 32a-2). Other adults showed more severe defects in gametogenesis (Figure 32b), and lacked the organization of germ cell development characteristic of Dpy-5 Unc-13 animals (Figure 10). Occasionally morphological defects were seen in the adults, such as increased girth in the center of the animal (Figure 32b-2), a defect-that was also seen when too high a DNA concentration was injected into Unc-11 hermaphrodites. Homozygous let-395(h271) animals normally arrested in their development as sterile adults with no recognizable gonads (Figures 33b,c; Chapter 2, Figure 11). There appeared to be no variability in expressivity of this mutation. The presence ofhExlO allowed Dpy-5 Let-395 Unc-13 hermaphrodites to develop into adults that produced eggs (Figure 33a). The eggs were never laid and failed to hatch, but did developed past cleavage to the two-fold stage. The transgenic strain bearing the extrachromosomal array hEx!6 partially rescued the lethal phenotype of the mutation let-603(h289). This mutation failed to be complemented by other transgenic strains. The strain bearing hExl6 carried only one cosmid, C32E7, since there were gaps in the cosmid contig on either side of this cosmid. The complementation results suggested that the C32E7 cosmid was responsible for the altered lethal phenotype of this mutation. Homozygous let-603(h289) animals normally arrested in their development as sterile adults with a tumorous mass and no recognizable gonad (Figures 34b,c; Chapter 2, Figure 10). The presence of the hEx!6 array in Dpy-5 Let-603 Unc-13 hermaphrodites abolished the tumorous phenotype and enabled the development of a gonad (Figure 34a). These adults laid many eggs, some of which hatched, but the larvae did not survive past the LI stage. Other eggs showed no 119 sign of morphogenesis and appeared to be developing abnormally, since they displayed grossly unequal cell sizes (Figure 34a-1). The transgenic strain bearing the extrachromosomal array hEx36 partially rescued the lethal phenotype of the mutation let-602(h283). This mutation failed to be complemented by the adjacent array, hEx34, which shared one cosmid, C32G12, in common with hEx36. A stable line was not obtained for the set of cosmids to the left of those in hEx36. These results suggested that either/both of the cosmids unique to the hEx36 array, C46H8 and C27D2, was responsible for the partially rescued lethal phenotype of this mutation. The lethal mutation let-602(h283) caused developmental arrest at the early larval stage in homozygous animals (Figure 35c). The presence of the hEx36 array enabled Dpy-5 Let-602 Unc-13 hermaphrodites to survive to adulthood (Figure 35a,b). Most hermaphrodites were completely sterile, failing to produce any eggs (Figure 35a). In these animals, gametogenesis did not occur, and the gonads appeared atrophied. A few adults produced eggs (Figure 35b). These eggs were not laid, and showed no sign of morphogenesis. Even though eggs were produced, these animals still lacked the characteristic organization of germ cell development. The transgenic strain bearing the extrachromosomal array hEx34 partially rescued the lethal phenotype of the mutation let-611(h850). This mutation failed to be complemented by the adjacent arrays, hEx36 or hEx39, each of which shared one cosmid in common with hEx36. These results suggested that the cosmid unique to the hEx36 array, C40A4, was responsible for the change in the lethal phenotype of this mutation. The lethal mutation let-61 l(h850) caused developmental arrest at the mid larval stage in homozygous animals (Figure 36b,c). The presence of the hEx34 array enabled Dpy-5 Let-611 Unc-13 hermaphrodites to survive to adulthood (Figure 36a). The hermaphrodites were completely sterile, failing to produce any eggs. Gametogenesis did not occur in any of the mutants observed, and they appeared to lack normal gonad development. 120 The transgenic strains bearing the extrachromosomal array hEx28 or hEx33 both partially rescued the lethal phenotype of the mutation let-393(h225). These two arrays shared a single cosmid in common, F40D12. These results suggested that the cosmid F40D12 was responsible for the altered lethal phenotype of this mutation. The lethal mutation let-393(h225) caused developmental arrest at the early larval stage in homozygous animals (Figure 37d). The presence of the hEx28 or the hEx33 array enabled Dpy-5 Let-393 Unc-13 hermaphrodites to develop into adults that produced eggs, however some variability in phenotype was observed (Figure 37a,b,c). In general, hermaphrodites that carried the array hEx33 developed further than those carrying the array hEx28. Most hermaphrodites carrying the array hEx33 laid many eggs (Figure 37c). Some of the eggs hatched, giving rise to larvae that died in the LI stage. However, none of the larvae survived past the LI stage. A few adults carrying hEx33 were completely sterile, failing to produce any eggs. By contrast, the majority of adult hermaphrodites carrying hEx28 were completely sterile, showing little gonad development (Figure 37a). Some of the hermaphrodites had developed gonads that laid eggs (Figure 37b), but they rarely hatched. The few larvae that hatched died at the LI stage. The majority of eggs displayed no sign of morphogenesis. The difference in development of hEx28 and hEx33 array-bearing hermaphrodites could either be due to sequences in the overlapping cosmid M01E11 found only in the array hEx33, or to differences in expression of the two arrays. 121 Table 14: Partial Cosmid Rescue of Essential Genes Gene (allele) Arrest Stage No array present Rescuing Array Arrest Stage Array present let-376 (hi30) Early larval hExlO Adult - few eggs laid: most do not hatch; some hatch to produce LI-arresting progeny - some internal hatching (die inside adult) let-393 (h225) Early larval hEx28 hEx33 Adult - variable phenotype: some sterile; others produce Ll-arresting progeny - hEx33 enabled further development than hEx28 let-395 (h271) Sterile adult (no recognizable gonad) hExlO Adult - eggs develop to the 2-fold stage - eggs are not laid and do not hatch let-530 (h798) Early larval hExlO Adult - variable phenotype: some show no gametogenesis; others produce eggs that develop to 2-fold stage, but are not laid let-602 (h283) Early larval hEx36 Adult - variable phenotype: some have no gametogenesis and atrophied gonads; others produced eggs that show no morphogenesis. let-603 (h289) Sterile adult (tumorous mass) hExl6 Adult - many eggs laid: some show no morphogenesis and unequal cell size; others hatch to produce Ll-arresting progeny - no tumorous mass let-611 (h850) Mid larval hEx34 Adult - no gametogenesis - lack normal gonad development Figure 31: Partial Rescue of lei -376( hi 30) Hermaphrodites viewed under Nomarski optics displaying partially rescued (a, b, c) or arrested (d, e, f) phenotypes. (a, b, c) dpy-5 let-376 unc-13;hExlO (partial rescue). The presence of hExlO enables these animals to develop to adulthood. Most produced eggs (a-1, c-1), some of which were laid and a few hatched to give dead larvae. The animal in (a) laid two eggs. Internal hatching occasionally occurred (a-2), but the larvae died wi th in the adult. Other adults showed more severe defects in gametogenesis (b-1). Distal germ cells (c-2) and oocytes (c-3) are marked. (d) Dpy-5 Let 376 Unc-13s arrested in early larval stage, picked from the hExlO complementation plate (Figure 30). 63% of Dpy-5 Let-376 Une-13 animals arising from a self-cross are expected to lack the hExlO array (Table 8), and arrest in the early larval stage characteristic for this mutation. (e) dpy-5 let-376 unc-13;hEx33. (f) dpy-5 let-376 unc-13;hEx26. The presence of either hEx33 or hEx26 failed to rescue the lethal mutation, resulting in early larval arrest. 123 Figure 32: Partial R e s c u e o f let-53()(h798) Hermaphrodites viewed under Nomarski optics displaying partial rescue (a, b) or arrested (c) phenotypes. (a, b) dpy-5 let-530 unc-13;hExlO. The presence of hExlO enabled these animals to develop into adults that produced eggs (a-1, b-1). The eggs were never laid, nor hatched, but some developed to the two-fold stage (a-2). Other adults (b) showed more severe defects in gametogenesis, lacking the characteristic organization of germ cell development. Some animals had an increased girth in their middle (b-2). Distal germ cells (a-3), and oocytes (a-4) are marked, (c) dpy-5 let-530 unc-13 animals arrested in the early larval stage, characteristic for this lethal mutation. Animals homozygous for let-530 and carrying either hEx33 or hEx26 also arrested during the early larval stage, and were similar in appearance to the animal in (c) (not shown). F i g u r e 33: Partial Rescue of let-395(h271) Hermaphrodites viewed under Nomarski optics displaying partial rescue (a) or arrested (b, c) phenotypes. (a) dpy-5 let-395 unc-I3;hExlO. The presence of hExlO enabled these animals to develop into adults that produced eggs, which developed to the two-fold stage (a-1), but failed to hatch. Oocytes (a-2) are marked, (b) Dpy-5 Let-395 Unc-13s arrested as sterile adults that lacked gonad development. These animals were picked from the hExlO complementation plate (Figure 27), from which 63% of the Dpy-5 Let-395 Unc-13 animals were expected to lack the hExlO array (Table 8), and arrest as sterile adults, (c) dpy-5 let-395 unc-13;hEx26. The presence of hEx26 failed to rescue the lethal mutation, resulting in sterile adults lacking gonad development. 125 F i g u r e 34: Partial Rescue of let-6()3(/i289) Hermaphrodites viewed under Nomarski optics displaying partially rescued (a) or arrested (b, c) phenotypes. (a) dpy-5 let-603 unc-13;hExl6. The presence of hExl6 enabled these animals to develop into adults that produced eggs, many of which were laid and a few hatched to give dead larvae. Other eggs showed no sign of morphogenesis and had grossly unequal cell sizes (a-1). No tumour-like growth can be seen in these animals, (b) Dpy-5 Let-603 Unc-13s arrested as sterile adults with a tumour-like growth (b-1), and no recognizable gonads. These animals were picked from the hEx!6 complementation plate (Figure 30), from which 46% of the Dpy-5 Let-603 Unc-13 animals were expected to lack the hEx]6 array (Table 8), and arrest as sterile adults, (c) dpy-5 let-603 unc-13;hEx26. The presence of hEx26 failed to rescue the lethal mutation, resulting in sterile adults with a tumour-like growth (c-1) and no recognizable gonad. 126 F i g u r e 35: Partial Rescue of let-6()2(h283) Hermaphrodites viewed under Nomarski optics displaying partially rescued (a, b) or arrested (c) phenotypes. (a, b) dpy-5 let-602 unc-13;hEx36. The presence of hEx36 enabled these animals to develop to adulthood. Most were completely sterile, showing no gamete development (a). A few adults produced eggs, which showed no sign of morphogenesis (b-1). However, these animals still lacked the characteristic organization of germ cell development, (c) Dpy-5 Let-602 Unc-13s arrested in early larval stage, picked from the hEx36 complementation plate (Figure 30), from which 62% of the Dpy-5 Let-602 Unc-13s were expected to lack the hEx36 array (Table 8). Animals homozygous for let-602 and carrying hEx34 also arrested during the early larval stage, and were similar in appearance to the animal in (c) (not shown). 127 Figure 36: Partial Rescue of let-61 l(h850) Hermaphrodites viewed under Nomarski optics displaying partially rescued (a) or arrested (b, c) phenotypes. (a) dpy-5 let-611 unc-I3;hEx34. The presence of hEx34 enabled these animals to develop to adulthood. However, they appear to lack any gonad development, (b) dpy-5 let-611 unc-13. Arrested in the mid larval stage, characteristic for this mutation, (c) dpy-5 let-611 unc-13; hEx36. The presence of hEx36 failed to rescue the lethal mutation, resulting in arrest during the mid larval stage. 128 F i g u r e 3 7 : Partial Rescue of let-393(h225) Hermaphrodites viewed under Nomarski optics displaying partially rescued (a, b, c) or arrested (d) phenotypes. (a, b) dpy-5 let-393 unc-J3;hEx28. The presence of hEx28 enabled these animals to develop to adulthood. Most adults were completely sterile, lacking normal gonad development (a). Other adults produced eggs (b), some of which hatched to give dead larvae. Most eggs showed no sign of morphogenesis (b-1). Even though the animal in (b) produced eggs, it still lacks the characteristic organization of germ cell development, (c) dpy-5 let-393 unc-13;hEx33. Most adults carrying hEx33 laid many eggs (c-1), some of which hatched to give dead larvae. Distal germ cells (c-2) and oocytes (c-3) are marked, (d) Dpy-5 Let-393 Unc-13 arrested in the early larval stage, picked from the hEx33 complementation plate (Figure 27), from which 62% of the Dpy-5 Let-393 Unc-13 animals were expected to Idck the hEx33 array (Table 8). Animals homozygous for let-393 and carrying hExlO also arrested during the early larval stage, and were similar in appearance to the animal in (d) (not shown). 129 (iii) Alignment of the Genetic and Physical Maps The complementation results obtained between visible/lethal mutations and extrachromosomal array-bearing strains has given a right-left position to loci that were previously inseparable by complementation to duplication strains (Figure 38). As shown in Figure 5 (Chapter 1), three lethal mutations map between the right breakpoints of the duplications hDpl6 and hDp31: let-396, let-602 and let-611. Complementation to array-bearing strains has tentatively placed let-602 to the left of let-611. The relative position of let-396 has not been determined. One visible and five lethal mutations map between the right breakpoint of the duplication hDp31 and the left breakpoint of the deletion hDf8: bli-4, tet-380, let-381, let-387, let-601 and let-606. Complementation results with array-bearing strains tentatively placed let-381 to the left of bli-4, and let-601 to the right of bli-4. The physical positions of the remaining three lethal mutations have not been determined, however there is a gap in the cosmid map to the right of the cosmid C03F7, which rescues the lethal phenotype of let-601. The first cosmid lying after the gap is C44D11, which was shown by K. Peters (unpublished results) to carry the right breakpoint of the deletion hDJB. The three lethal mutations could either lie in this gap, or the cosmids carrying these loci may have failed to rescue the lethal phenotypes of these mutations. Complementation of the visible/lethal mutations with array-bearing strains has also given physical positions for certain duplication and deletion breakpoints (Table 15). The right breakpoint of the deletion sDf4 separates the lethal mutations let-367 and let-530, with let-367 lying within and let-530 lying outside the deletion (A. Howell, unpublished results) (Figure 5, Chapter 1). Since the phenotype of let-367 is rescued by the cosmid C12H4 and the phenotype of let-530 is partially rescued by the overlapping cosmid M01A12, the right breakpoint of sDf4 can be tentatively positioned to C12H4 and/or M01A12. The right breakpoint of the duplication hDp57 lies between let-530 (on the left) and let-395 (on the right). The left breakpoint of the duplication hDp48 lies between let-395 (on the left) and let-376 (on the right) (S. McKay, unpublished results). Since the cosmid M01A12 partially rescues the phenotype of all three lethal mutations, let-530, let-395 and let-376, the right 130 breakpoint of hDp57 and the left breakpoint of hDp48 can be tentatively positioned to the cosmid M01A12. The right breakpoint of the duplication hDp41 lies between let-376 (on the left) and let-378 (on the right) (S. McKay, unpublished results). Since the phenotype of let-376 is partially rescued by the cosmid M01A12, and the phenotype of let-378 is rescued by the overlapping cosmid C07F10, the right breakpoint of hDp41 can be tentatively positioned to M01A12 and/or C07F10. The right breakpoint of the duplication hDp72 lies between let-378 (on the left) and let-379 (on the right). The breakpoint of the duplication hDp37 lies between let-379 (on the left) and unc-40 (on the right) (S. McKay, unpublished results). The phenotype of let-378 was rescued by the cosmid C07F10. The uncoordinated phenotype of unc-40 was rescued by the cosmid T02D7 (J. Culotti, personal communication), which completely overlaps C07F10 (J. Hodgkin, A C E D B release 4). Therefore, the right breakpoints of hDp72 and hDp37 can be tentatively positioned to C07F10 (T02D7). The right breakpoint of the duplication hDp31 lies between let-611 (on the left) and let-381 (on the right). Since the phenotypes of let-611 is partially rescued by the cosmid C40A4, and the phenotype of let-381 is rescued by the cosmids C38F7 and/or F26B1, the right breakpoint of hDp31 may lie on one of the cosmids C40A4, C48E7, C38F7 or F26B1. From the data presented in Table 15, three cosmids, C12H4, M01A12, and C07F10, are good candidates for carrying six of the seven rearrangement breakpoints tentatively positioned on the cosmid map. To determine the precise location of these rearrangement breakpoints on the physical map, the ends of the duplications must be mapped molecularly. let-531 , let-532 C28A7 F40D12 M01A12 C04F1 C03E6 C30A11 F53D11 ZK411 C27D2 C40A4 C38F7 K04F10 C24H1 sD/4 hDplS Wpl3 hDp57 hDp48 hDp41 hDp72 hDpS7 hDpl7 hDp61 HDpl2 Wpl4, hDpl6, hDpl8, hDp32 hDpSl Figure 38: Map Showing the Cosmid Rescue of Essential Genes in the dpy-5 bli-4 Region. The unc-40 locus was mapped, to cosmid T 0 2 D 7 (J Culotti, personal communication), which completely overlaps the C 0 7 F 1 0 cosmid. 132 Table 15: Alignment of Rearrangement Breakpoints with the Cosmid Contig Most probable cosmid(s) containing Rearrangement rearrangement breakpoints sDf4 C12H4 and/or M01A12 hDp57 M01A12 hDp48 M01A12 hDp41 M01A12 and/or C07F10 hDp72 C07F10 hDp37 C07F10 hDp31 C40A4 and/or C48E7 and/or C38F7 and/or F26B1 133 Confirmation of the Presence of Cosmid Sequences: Polymerase chain reaction (PCR) was used to confirm the presence of transforming DNA in transgenic animals, by amplifying sequences specific to the Lorist vector (see Materials and Methods). PCR on whole worms was performed on the following strains: N2, KR2225 hEx7, KR2276 hEx4, KR2356 dpy-5 bli-4(h754)unc-13;hEx5, KR2366 dpy-5 bli-4(h42) unc-13;hEx4 (Tables 7 and 13), as well as on no worms as a negative control. Primers specific to adenosyl homocysteine gene (AHH) amplified sequences in N2 worms, as well as the four transgenic strains, KR2225, KR2276, KR2356, and KR2366. No sequences were amplified in the vial lacking worms (negative control). The A H H gene is not present in the cosmids, but is present in the C. elegans genome, and was used as a positive control. Primers specific to the Lorist vector only amplified sequences in the reactions involving transgenic worms carrying the cosmid K04F10: KR2276, KR2356 and KR2366. However, the Lorist-specific primers failed to amplify sequences in the transgenic strain KR2225, which was transformed with the plasmid pRF4 alone. Negative results were also obtained with N2 worms and when no worms were used (negative controls). These results confirm the presence of Lorist vector in the three transgenic strains tested, including those that rescued the two bli-4 mutations. 134 Chapter 3: DISCUSSION This chapter described the phenotypic rescue of mutations in 13 essential genes mapping in the dpy-5 unc-13 region by complementation to strains bearing extrachromosomal arrays. A total of 30 transgenic strains were,constructed. These rescues provided physical positions for 13 essential genes, and predicted the positions of six duplication breakpoints and one deficiency breakpoint, thereby aiding in the alignment of the genetic and physical maps in the region under investigation. The data presented on the frequencies of transmission of extrachromosomal arrays indicated that for several arrays, their transmission in hermaphrodites occurred predominantly through the sperm. However, after several generations, three arrays showed an increase in stability during oogenesis. Germline Transformation: Transformation in C. elegans using cosmid and plasmid DNA has been achieved through injections to oocyte nuclei (Fire 1986), as well as to the cytoplasm in the distal arms (Stinchcomb etal. 1985; Mello etal. 1991). In this study, injections were made to the cytoplasm in the distal arms of the gonads. The injected cosmid and plasmid clones contained shared regions of homology to help ensure their co-transformation. Mello etal. (1991) showed that shared regions of homology greater than 500 base pairs aided in the co-assembly of injected sequences into arrays, possibly via homologous recombination reactions. They demonstrated that the co-injection of plasmids bearing non-overlapping deletions that inactivated the rol-6 gene underwent homologous recombination between the injected molecules during array formation, thereby restoring the expression of the Rol-6 phenotype. Restoration of gene function was dependent upon the length of the homologous region. By contrast, non-homologous molecules showed a much lower frequency of co-assembly. Recombination between homologous DNA plasmids has 135 also been demonstrated in mammalian cells (Folger et al. 1985). Linear DNA molecules were found to be preferred substrates for homologous recombination with mammalian chromosomes (Wake et al. 1985), and the degree of recombination was found to increase with increasing homology (Lin et al. 1984). Homologous recombination was found to occur during yeast transformation as well, since the recombination and repair gene rad52-l blocked the integration of linear DNA plasmids into yeast chromosomes (Orr-Weaver etal. 1981). In contrast, Fire and Waterston (1989) suggested that although homologous recombination occurred between injected plasmids during transformation, integration of the injected DNA into the C. elegans chromosome is usually a non-homologous event. Homologous recombination not only ensures the co-transformation of overlapping sequences, but enables the reconstruction of genes that are only partially contained in overlapping cosmids, thereby restoring their expression. The reconstruction of an intact gene from two defective genes by recombination betweeninjected DNA fragments has been demonstrated in many systems, including C. elegans (Mello et al. 1991), yeast (Kunes etal. 1990), and mammalian (Lin etal. 1984; Wake etal. 1985). Transformation of C. elegans germline can produce three types of transformants: transiently transformed, heritably transformed with extrachromosomal arrays, and stable integrated lines (rare). The concentration of injected DNA influences the heritability of extrachromosomal arrays. DNA concentrations of around lOOug/ml were found to result in a larger percentage of heritable transformation over DNA concentrations of 50-80ug/ml. As the DNA concentration was increased further to 120ug/ml, morphological defects were seen in the progeny of injected animals, and transformants became rare. These results are consistent with other researchers (Fire 1986; Mello etal. 1991). In this study, only one integrated line was obtained, and it arose from a line carrying an extrachromosomal array. Work with mammalian cells has demonstrated that extrachromosomal sequences can integrate into the genome upon prolonged culture in selective media (Fournier and Ruddle 1977; Klobutcher etal. 1980). All DNA injections in this study were made into gonad 136 cytoplasm. In previous work on C. elegans transformation, DNA sequences injected directly into the nucleus (Fire 1986) showed a much higher level of integration than sequences injected into gonad cytoplasm (Mello etal 1991). Mello etal. (1991) suggested that nuclear injections allowed the DNA to be exposed directly to chromosomes, whereas cytoplasmic injections may result in DNA modification prior to exposure to the chromosomes. In particular, oocyte cytoplasm contains all the structural components necessary to assemble nuclei. Modifications to injected DNA sequences have been shown to occur in Xenopus. Laskey et al. (1977) demonstrated that chromatin can assemble from purified DNA in one hour using a stored histone pool from Xenopus oocytes. In addition, structures resembling typical eukaryotic cell nuclei assembled around naked DNA 60-90 minutes after being injected into Xenopus oocytes (Forbes etal. 1983; Newport 1987). The assembly of plasmid DNA into nucleus-like structures in Drosophila embryos is thought to substantially protect it from degradation, since cytoplasmic nucleases were found to degrade unprotected DNA (Steller and Pirrotta 1985). In this study, it was not considered necessary to obtain integrated lines, because extrachromosomal arrays were as suitable for complementation analysis as integrated sequences. Germline Transmission of Extrachromosomal Arrays: In this study, 85% of transformed progeny showed transient expression of the extrachromosomal array, failing to segregate the array to their offspring. The remaining 15% of transformed progeny were heritably transformed. Heritable transformation is dependent upon the size of an array. Mello et al. (1991) suggested that at appropriate DNA concentrations, homologous recombination reactions drive the assembly of large contiguous arrays, which are capable of being transmitted in the germline, while smaller arrays can be expressed in the progeny of injected animals, but are not heritable. Stinchcomb et al. (1985) estimated that transformed cells contained between 80 and 300 plasmid copies per cell, assembled in one or a few arrays. In mammalian systems, small, cytologically undetectable transgenic fragments were lost from mouse 137 cells at a faster rate than larger fragments, presumably through irregular segregation (Klobutcher et at. 1980). Faithful germline transmission of extrachromosomal arrays is dependent upon accurate chromosome replication and segregation, as well as nuclear retention. Loss of an array could be caused by failures in replication, nondisjunction, or loss from the nucleus. Efficient replication of injected DNA appears to be dependent upon the size of the DNA, rather than upon specific sequences (Blow and Laskey 1986). In Xenopus embryos, Marini etal. (1986) showed that high molecularweightconcatemersof plasmid DNA were replicated very efficiently, while monomelic DNA remained unreplicated. Similarly, McMahon et al. (1985) demonstrated that in the sea urchin, injected linear DNA formed concatemers that were efficiently replicated, while injected circular DNA did not form concatemers and remained unreplicated. Harland and Laskey (1980) found that the extensive replication of long concatemers occurred irrespective of the sequence of the injected DNA. When prokaryotic or eukaryotic DNA was injected into Xenopus oocytes, it underwent semiconservative replication and observed cellular regulatory signals, even if it lacked a eukaryotic origin of replication. The meiotic and mitotic segregation of exogenous DNA is also influenced by the size of the array. The effect of physical length on the mitotic segregation of artificial chromosomes has been demonstrated in yeast (Hieter et al. 1985; Murray et al. 1986). They showed that chromosomal fragments must be at least 100 kb to be mitotically stable. Smaller chromosomal fragments could attach to the mitotic spindle, but segregated randomly and were frequently lost during mitosis. They proposed that as DNA molecules increase in size, they become large enough to form several domains of supercoiling, enabling the sister chromatids to become catenated, which directed their segregation. McKim and Rose (1990) showed that shortening duplications in C. elegans decreases their mitotic stability, possibly because smaller chromosomal fragments would have insufficient spindle attachment for proper segregation. They suggested that structural differences between duplications, such as linear or circular forms, could alter their stability. 138 Loss of extrachromosomal arrays from the nucleus has been demonstrated in C. elegans. Stinchcomb etal. (1985) used in situ hybridization of labeled plasmid sequences to show loss of foreign DNA during cell proliferation of the germ line. They calculated a rate of loss of 5% per cell division. Similar values have been reported for extrachromosomal D N A loss in both yeast (Sakaguchi and Yamamoto 1982) and mammalian cells (Scangos and Ruddle 1981). Transmission Frequencies of Extrachromosomal Arrays: The different transgenic lines produced in this study that were transformed with the same cosmid/plasmid DNA were found to have different characteristics with regard to transmission frequency of the array, regardless of whether these lines originated from the same injected hermaphrodite, or from different injected hermaphrodites. For example, hEx33 and hEx38 originated from the same injected hermaphrodite, yet hEx33 segregated Ex gametes at the frequency of 21%, while hEx38 segregated only 6% Ex gametes (Table 9). In addition, both transient and stable transgenic animals can be found among the progeny of one injected hermaphrodite. Mello etal. (1991) suggested that newly injected D N A sequences are initially highly reactive, and assemble into independent transforming elements. They demonstrated that separate progeny from a single injected animal contained arrays with unique DNA rearrangements. These rearrangements were as different from each other structurally as were the arrays found from separate injected animals. (i) Transmission Frequencies in Oocytes and Sperm: For several array-bearing strains, the transmission of extrachromosomal arrays in hermaphrodites was found to occur predominantly through the sperm. Twelve out of .16 extrachromosomal arrays were recovered at a frequency 2- to 7-fold higher in hermaphrodite sperm than oocytes. In addition, extrachromosomal arrays that had a high overall transmission frequency in gametes were not necessarily more stable in oocytes than arrays that had a low overall 139 transmission frequency in gametes. Some strains that produced a high frequency of Ex gametes showed a greater discrepancy between the frequency of Ex oocytes and Ex sperm. This discrepancy between array stability in oocytes and sperm in hermaphrodites was found to vary between strains injected with the same cosmid DNA, which argues against the influence of specific cosmid DNA sequences in this phenomenon. On the other hand, the two strains hEx5 and hEx32 that displayed the greatest discrepancy between array transmission in oocytes and sperm (7-fold difference), both carried DNA sequences in common. If there is a sequence carried by these two arrays that can influence the transmission or expression of extrachromosomal DNA, then it must either be absent or unexpressed in the strains hEx4 and hEx31, which were injected with the same DNA as hEx5 and hEx32, respectively. Such sequences could be absent or altered in different arrays composed of the same cosmid DNA, because cosmids can recombine when they form concatemers (Stinchcomb et al. 1985; Mello etal. 1991). For example, even though both hEx4 and hEx5 rescue all three classes of bli-4 mutations indicating that they carry an intact bli-4 locus, the sequences affecting array transmission could be present in hEx5 and altered or lacking in hEx4. This can only be determined by isolating the sequences that affect array transmission in the hEx5 and hEx32 arrays. If such sequences exist, their biological function could be elucidated by analyzing the mutant phenotype induced by an extrachromosomal array. Of course, it is possible that array transmission could be affected both by specific sequences, as well as more general sequence-independent factors. Whether influenced by specific sequences or not, the lower frequency of Ex oocytes could represent either a more frequent loss of exogenous DNA from oocytes, or more restrictions on their replication. In Xenopus oocytes, the packaging of naked DNA into chromatin (Laskey et al. and its assembly into nucleus-like structures (Forbes etal. 1983; Newport 1987) was found to be necessary for the replication of exogenous DNA (Newport 1987). Similar processes have not been described in sperm. It is unknown if such processes occur in C. elegans, however it is possible that exogenous DNA in C. elegans oocytes could be sequestered away from endogenous DNA. In addition, oocytes carry many cellular components in their cytoplasm, including DNA 140 modifying and packaging proteins, for the later development of the embryo. It is possible that the replication of exogenous DNA in oocytes is subject to more stringent regulation than in sperm. In regards to the loss of exogenous DNA from oocytes, C. elegans oocytes undergo more mitotic cell divisions than hermaphrodite sperm, thereby providing more opportunity for array loss during cell division. Hermaphrodite sperm are made for a brief period between L3 and late L4, when about 80 mitotic nuclei become committed to spermatogenesis to produce about 320 sperm, irrespective of whether the hermaphrodite is mated or not (Hodgkin 1987). All the remaining mitotic nuclei are committed to oogenesis, which occurs throughout adulthood, producing 320-1400 oocytes (Kimble and Hirsh 1979). Each mitotic nucleus will give rise to only one oocyte, and three polar bodies, whereas one mitotic nucleus can give rise to four functional sperm. Furthermore, at least a third of the germ cells that are committed to oogenesis are believed to undergo cell death, so as to increase the cytoplasmic to nuclear ratio in the developing oocytes (B. Horvitz, unpublished results). Therefore, for a hermaphrodite to produce the same number of oocytes as sperm, germ nuclei must undergo more mitotic cell divisions during oogenesis in order to compensate for those lost through programmed cell death. Only 80 mitotic nuclei are required to produce 320 sperm, compared to at least 480 mitotic nuclei that would be required to produce 320 oocytes. Stinchcomb et al. (1985) estimated a 5% loss of extrachromosomal arrays per cell division. The increased number of cell divisions occurring during oogenesis as compared to spermatogenesis could provide more opportunity for loss of extrachromosomal arrays from the developing oocytes. This loss may be accentuated by mating the hermaphrodite to a male. Hermaphrodites are stimulated to produce more oocytes when fertilized by a male (up to 1400), than when self-fertilized (320) (Hodgkin 1986). Consequently, there is an increase in the number of mitotic cell divisions occurring during oogenesis as a result of mating, thereby further increasing the opportunity for loss of the array. In contrast, the number of mitotic cell divisions during spermatogenesis remains constant in mated or unmated hermaphrodites. The decrease in transgenic progeny observed after mating as compared to self-fertilization, may in part be attributed to an increase in the proliferation of oocytes as a result of the mating. These results cannot be 4 141 extended to male sperm, since transmission frequencies were not examined in males. Furthermore, the process of spermatogenesis differs between hermaphrodites and males, with spermatogenesis occurring continuously in males to generate about 3000 sperm (Hodgkin 1987). Extrachromosomal arrays are capable of being transmitted through male sperm in C. elegans (Stinchcomb etal. 1985), but no comparison of array loss in hermaphrodite and male sperm has been made. A similar difference in transmission between oocytes and hermaphrodite sperm has been observed with duplications in C. elegans. Some free X chromosome duplications (Herman et al. 1976) and chromosome I duplications (McKim and Rose 1990) have been shown to have a lower frequency of recovery in hermaphrodite oocytes and male sperm than in hermaphrodite sperm. McKim and Rose (1990) attributed this loss to the increased number of cell divisions occurring during oogenesis and male spermatogenesis, as compared to hermaphrodite spermatogenesis. Other chromosome I duplications were recovered at similar frequencies in hermaphrodite oocytes and sperm, indicating that some duplications are less sensitive to loss during cell division. In general, the segregation stability of duplications is considerably higher than for extrachromosomal arrays. The difference in stability may be due to their larger size, as well to the presence of specific sequences and their structure, which may be important for their segregation, replication, and maintenance. In contrast, the integrated strain nisi, showed an equal frequency of transmission of the integrated sequences in oocytes and hermaphrodite sperm. The integrated sequences were expected to be highly stable, because they should be replicated and maintained along with the endogenous DNA. Integrated sequences have been found to be highly stable in C. elegans (Fire 1986), yeast (Orr-Weaver etal. 1981), and mammalian systems (Klobutcher and Ruddle 1979). (ii) Changes in Transmission Frequencies: The different arrays examined in this study had different transmission frequencies, however, the stability of an array for a given line remained characteristic for that array over several 142 generations, fluctuating only slightly. After an array is formed, it becomes less recombinogenic, and is propagated with little further rearrangement (Stinchcomb al. 1985; Mello et al. 1991). A limited period of reactivity for injected DNA fragments has been shown for mammalian systems as well. Folger etal. (1985) suggested that homologous recombination events between injected plasmid molecules occur within one hour after injection, based on the inability of two separately injected plasmids to recombine when the interval of successive injections exceeds one hour. However, several transgenic strains examined in this study showed a sudden increase in transmission frequency after prolonged culture, resulting in a new, consistent transmission frequency for that strain. The increase in array stability was found to occur primarily in oocytes. Three strains showed a 4- to 13-fold increase in the frequency of Ex oocytes. A decrease in transmission frequency in oocytes was never observed. Array stability in hermaphrodite sperm was more variable, since both increases and decreases in stability were seen (up to 1.5-fold either way). However, in each case the array transmission frequency to sperm still remained higher than to oocytes. The increase in array stability occurring predominantly in oocytes argues against array loss during cell division as being the only reason for the lower transmission of arrays through oocytes as compared to hermaphrodite sperm. If there was a general decrease in array loss per cell cycle, one would expect to see a similar increase in array stability during spermatogenesis, as was seen during oogenesis. The increase in array stability in oocytes could be due to an increase in length of the arrays, however this might be expected to cause a more gradual increase in array stability. Alternatively, the addition of endogenous sequences to extrachromosomal arrays could cause a sharp increase in array stability. Such endogenous sequences could be introduced into the array by unequal crossing over between the array and the chromosome. In yeast, the addition of functional centromeres, telomeres and origin of replication sites to DNA sequences has resulted in stable artificial chromosomes (Murray and Szostak 1983). In Drosophila, functional centromeres have been shown to be sufficient for completely normal transmission of minichromosomes (Murphy and Karpen 1995). Whether such sequences are added in vivo to extrachromosomal 143 arrays is unknown. A third possibility could be a change in the conformation of the DNA, or its association with stabilizing factors, such as proteins or envelope. The technique of fluorescence in situ hybridization (FISH) has been used to analyze the replication of cosmids and yeast artificial chromosomes in yeast (Rosenberg etal. 1995), and could provide valuable information on the position, replication, composition and structure of exogenous DNA in C. elegans. For example, such technology could yield information on the structure of extrachromosomal arrays in oocytes and sperm, that may help to explain the difference in array stability between these germ cells that was found in this study. (ii) Early Mitotic Loss Extrachromosomal arrays were occasionally lost early in germline development, prior to gametogenesis. When this occurred, a transgenic hermaphrodite would lack the extrachromosomal array in its germline and produce no transgenic progeny. Each extrachromosomal array had a characteristic frequency of early mitotic loss in the germline, ranging from 1% to 23%. In general, the frequency of early mitotic loss correlated to the frequency of germline transmission, whereby arrays that were mitotically stable showed a higher frequency of transmission in the gametes. Mitotically unstable arrays are potentially important tools for mosaic analysis. The integrated line, hlsl, never showed early mitotic loss of the Rol-6 sequences, consistent with the DNA being integrated into the genome. Fire et al. (1991) also found a correlation between the lack of detectable mitotic loss in certain extrachromosomal arrays with less mosaicism and increased meiotic stability. The same behaviour has been described with regard to duplications. McKim and Rose (1990) described duplication loss that occurred early in germline development, resulting in progeny that failed to carry the duplication. They suggested that these duplications might exhibit a higher level of somatic duplication loss over those with similar transmission frequencies in gametes. 144 Phenotypic Rescue of Lethals using Extrachromosomal Arrays: (i) Partial Rescue of Lethal Phenotypes Of the 15 mutations rescued by complementation to transgenic strains, eight showed complete phenotypic rescue, while seven showed only partial phenotypic rescue. In addition, some of the lethal mutations that were genetically mapped to the region under study were not rescued. It should be noted that while positive results provide evidence that the injected cosmids carry a wild type copy of the endogenous lethal mutation, negative results are not conclusive. In regards to the complete phenotypic rescue of mutations, it has been demonstrated that re-introduced genes can be expressed in the same cell types as endogenous loci, and can restore wild-type gene function (Way and Chalfie 1988; Fire and Waterston 1989; Spencer al. 1990; Clark and Baillie 1992). On the other hand, partial or absence of rescue could be due to the type of transcript produced from the transgene, or to the level of expression and regulation of the transgene. If the transgene does not produce a functional product, then the lethal mutation will not be rescued. There are several ways in which an extrachromosomal array might produce a dysfunctional product. The transgene could be mutant, producing an aberrant protein product, or no product at all. An aberrant protein product could function poorly to enable only partial rescue of the lethal phenotype, or may be dysfunctional resulting in no detectable rescue. Even if the cosmid carries a wild-type transgene, some copies of the cosmid could become altered upon formation of the array, which could again result in an aberrant protein product. The aberrant transcript/protein could compete with any wild-type copies produced from the array, resulting in partial or no rescue. Fire (1986) observed some disruption of DNA introduced into C. elegans, with novel joints found between plasmid sequences, in addition to intact plasmid sequences. Alternatively, the orientation of cosmid molecules within extrachromosomal arrays could result in the production of antisense RNA transcripts that could interfere with sense transcripts from the same array. This could reduce the number of sense transcripts that are able to be translated, resulting in partial or no rescue. By the injection of unc-22 clones into wild-type animals, Fire etal. (1991) showed that antisense RNA 145 transcripts could be produced from extrachromosomal arrays, and that these transcripts could interfere with the expression of the wild-type endogenous locus (Moerman and Baillie 1979). Complete phenotypic rescue of a lethal mutation also depends upon the level of transcription from an extrachromosomal array. Arrays can carry a few to many copies of a DNA clone (Stinchcomb etal. 1985; Mello etal. 1991). Hence, the level of expression from different extrachromosomal arrays can be highly variable. Some loci can tolerate a high level of expression and still produce a wild-type phenotype, whereas the level of expression at other loci must be stringent if wild-type function is to be restored. Mello etal. (1991) showed that with a rol-6 transgene, phenotypic expression is identical over a broad range of 10 to over a 100 copies of the rol-6 plasmid per cell. In contrast, Fire (1986) observed that with the amber suppressor sup-7, transformed strains could only tolerate between one and ten copies of the sup-7 plasmid. Higher doses of the sup- 7 plasmid were found to be lethal. Some of the lethal mutations in this study that were partially rescued, or failed to be rescued, could be sensitive to the copy number of the cosmid sequences. Consistent with this hypothesis is the observation that deletions have been difficult to isolate within the chromosome I gene cluster (A Rose, unpublished results), suggesting that certain loci may be haplo-insufficient, indicating a sensitivity to gene dose. A cosmid occurring in high copy number in an array may turn a null or hypomorphic endogenous mutation into a hypermorph. High doses of a wild-type product from a transgene may interfere with the proper functioning of that protein, or even have a lethal effect. Fire and Waterston (1989) observed a wild-type pattern of tissue-specific expression of two muscle genes, myo-3 and unc-54, when integrated at low copy number in the C. elegans genome, even though they integrated at arbitrary points. However, when unc-54 occurred in high copy number, it resulted in incomplete rescue, presumably due to the interference of muscle function from excess unc-54 product. Partial or lack of rescue could also result from an absence or an insufficient level of expression of the transgene. For instance, a transgene with a low level of expression may be insufficient to fully rescue an endogenous locus with a null phenotype. 146 The level of transcription from extrachromosomal arrays can be influenced by several factors, including topology, the presence of control regions, methylation, and interference from vector sequences. The topology of the exogenous DNA can influence its expression. In Xenopus oocytes, the level of transcription from linear exogenous molecules was shown to be 500 to 1,000-fold less than that produced by circular DNA, which had the same level of efficiency as endogenous loci (Harland et al. 1983). They suggested that linear molecules could dissipate tortional stress and may experience interference from the free ends. The level of condensation of DNA also determines its accessibility to transcription factors. Secondly, the injected DNA may lack the regulatory sequences required for the proper expression of the transgenes. If such sequences are not included in the cosmid, the regulation of transgene expression may be capricious. For example, the human (5-globin gene contains five different control regions located 50 kb 5', inside, and 20 kb 3' of the structural gene (Grosveld etal. 1987). The presence of the 5' control region results in near wild-type levels of transgene transcription in mice, and is independent of the position of integration. By contrast, in constructs lacking the 5' control region, expression of the (3-globin transgene is suppressed (regardless of copy number), and under the influence of control factors at the site of integration. Grosveld etal. (1987) suggested that the 5' control region could enable /ram-acting factors to access the promoter and enhancer sequences surrounding the p-globin locus. In C. elegans, as many as 26% of genes may occur in operons, producing polycistronic mRNA precursors that are processed by transplicing (Spieth et al. 1993; Zorio et al. 1994). In such an operon, a promoter could be some distance from the gene under study, and if it is not contained in the cosmid, mutations in the gene may not be rescued by that cosmid. Thirdly, DNA methylation can alter transgene expression (MacGregor et al. 1987; Swain etal. 1987). Methylation reduces transcription from a locus. MacGregor et al. (1987) demonstrated that the treatment of transgenic mammalian cells with a nucleoside analog that inhibits methylation resulted in elevated transgene transcription. Fourthly, vector sequences could interfere with the recognition of transgenic sequences, thereby affecting its expression (Palmiter and Brinster 1986; Kothary et at. 1989). 147 Partial phenotypic rescue cou ld also be a consequence o f the l eve l o f interaction between a transgene and an endogenous locus. T h e transgenic D N A may be expressed, but f a i l to interact w i th an endogenous locus at a wi ld- type l e v e l . In Drosophila, H a z e l r i g g et al. (1984) showed that transduced white D N A segments differed i n their interactions w i t h endogenous teste mutat ions. These interactions ranged f rom a greater than normal repression by zeste, to complete insensi t ivi ty to zeste. H a z e l r i g g etal. (1984) suggested that such variable interactions c o u l d be a consequence o f the amount o f f l a n k i n g D N A transferred w i t h white, the c o p y number o f white, and/or the pos i t ion o f in tegrat ion o f white, zeste is k n o w n to be a t ranscr ipt ion factor that decondenses chromat in to enable transcription (Judd 1995). Par t ia l phenotyp ic rescue is not necessar i ly ove rcome by the in tegra t ion o f t ransgenic sequences. M e l l o etal. (1991) observed a comparable l eve l o f express ion o f a transgene f rom extrachromosomal arrays as f rom chromosomal integration. T h e y showed that the rol-6 p l a smid occurred i n h igh copy number i n both ex t rachromosomal arrays and chromosomal ly integrated w i t h no apparent difference i n phenotypic expression. S i m i l a r l y , the sup-7 p l a smid was found to be sensitive to copy number i n both array-bearing and integrated strains (F i r e 1986). T h e m a i n advantage o f integrated transgenic sequences i n regards to their expression is that a s ingle copy can be stably integrated, whereas arrays require large copy numbers to be heri table. T h i s is important for the rescue o f genes that are sensit ive to copy number . H o w e v e r , ch romosoma l in tegrat ion c o u l d impose other constraints upon express ion o f a transgene not found i n ex t rachromosomal arrays. Transgene expression is often dependent upon its posi t ion i n the chromosome ( L e v i s et al. 1985). Pos i t ion effect cou ld cause the transgene to be dormant, overexpressed, or have variegated expression. In Drosophila, H a z e l r i g g etal. (1984) correlated the variegated expression o f a white transgene w i t h its in tegra t ion near he terochromat in . T h e express ion o f integrated transgenic sequences can also be influenced by endogenous promoters or enhancers located near the pos i t ion o f integration (Palmiter and Brinster 1986). Integration can also pose part icular problems, such as integrat ion in to an endogenous l o c u s , caus ing muta t ion or even le tha l i ty (Perry et al. 1995; Schr ick etal. 1995). 148 (ii) Phenotypic Variability of Partially Rescued Animals Most of the partially rescued lethal mutations (5/7) displayed a range of rescued phenotypes. For example, partially rescued let-376 homozygotes ranged from having severe defects in gametogenesis, to being capable of producing eggs that hatched internally to give rise to dead LI larvae. The variability in phenotypic rescue was confined to partially rescued animals, since the eight lethal strains that were completely rescued displayed no variability in their phenotypes. Other researchers have found that the partial rescue of endogenous mutations can result in variable phenotypes of the rescued animals (Fire and Waterston 1989; Kim and Horvitz 1990). Loci that are only partially rescued may be sensitive to gene dose. The phenotypic variability of partially rescued animals could arise from fluctuations in the amount of transgene product available. These fluctuations could arise from the somatic loss of transgenic sequences, or to variable expressivity of transgenic sequences in different tissues. Mutations that can be completely rescued may be less affected by fluctuations in the amount of transgene product available than mutations exhibiting partial phenotypic rescue. Somatic cell loss of an extrachromosomal array can result in somatic mosaic animals that display a variegated expression of a transgene. Variable expression of a transgene in the tissue(s) in which it is required could cause partial phenotypic rescue, as well as variability in the rescued phenotype, especially with a non-diffusible product. Stinchcomb etal. (1985) used in situ hybridization of embryos and adults to show that extrachromosomal arrays could be lost in somatic cells to yield mosaic transformants. They found a wide range of mosaicism in embryos, with between 2/30 to 28/30 cells carrying the array. Fire et al. (1991) described transgenic animals displaying phenotypic mosaicism in muscle function, and demonstrated array mosaicism at the single cell level. In a similar fashion, mosaicism has been shown to occur from the somatic loss of unstable duplications (McKim and Rose 1990). A second possible explanation for the variability in the phenotypes of partially rescued transgenic animals could be a dissimilar expression of transgenic sequences in different tissues, or in different cells within the same tissue. Tissue-specific patterns of transgene expression have 149 been demonstrated in C. elegans. Fire and Waterston (1989) achieved muscle-specific expression of an unc-54 transgene. However, high copy number led to variable phenotypes, ranging from minor defects in movement to complete paralysis. In mammalian cells, clonal populations have been found to be mosaic for the expression of p-galactosidase fusion genes, regardless of copy number or the site of integration (MacGregor et al. 1987; Kothary et al. 1989). Stuart et al. (1990) suggested that the variegated patterns of transgene expression that they observed at the cellular level in zebrafish fin epithelial cells could be a consequence of selective activation or repression of stable transgene inserts. Variable transgene expression could arise in many ways through either transcriptional or translational regulation. For instance, DNA methylation may alter transgene expression in certain tissues (MacGregor et al. 1987; Swain et al. 1987). Swain etal. (1987) demonstrated in mice that a c-myc transgene inherited from the male parent is expressed exclusively in the heart, while it fails to be expressed at all when inherited from the female parent. They correlated this pattern of expression with parentally imprinted methylation. (iii) Artificial Positive Results In general, the order of the cosmids that rescued lethal mutations corresponded to the order of the mutations on the genetic map. However, in one case, let-393, phenotypic rescue was achieved with a set of cosmids lying further to the left than expected. Five lethal mutations, let-367, let-376, let-395, let-530, and let-532, all lie to the left of let-393 on the genetic map (Figure 5, Chapter 1), yet all five mutations were rescued with cosmids that physically mapped to the right of those used to rescue let-393. let-393 was partially rescued by two extrachromosomal arrays, hEx28 and hEx33. let-393 normally arrests in the early larval stage, while the presence of either array enabled let-393 homozygotes to develop into adults, with phenotypes ranging from complete sterility to being able to produce dead L I larvae. There are two possible explanations for the disparity between the genetic mapping and cosmid rescue results. Firstly, the position of let-393 on the genetic map could be further to the left than first postulated in Chapter one of this thesis, if the duplication hDp41 that failed to complement let-393 carried a mutant copy of let-393, 150 or failed to carry let-393 due to a deletion or rearrangement. Since the chromosome I duplications used in this study were initially derived from sDp2 by gamma radiation (McKim and Rose 1990), it is possible that a gamma ray-induced lethal mutation could be present in the hDp41 duplication. Deletions have been found within gamma ray-induced, chromosome I duplications (McKim and Rose 1990). At least five chromosome I duplications were found to contain more than one breakpoint. Two of these duplications failed to complement loci expected to lie within them, suggesting that they had internal deletions. A second possible explanation for the disparity in data concerning the position of let-393, could be that the cosmid rescue was an artificial positive result. Artificial positive results could be obtained for a number of reasons. A gene product from a second related locus covered by hEx28 and hEx33 could compensate for the lack of let-393 product when produced in higher amounts by the extrachromosomal arrays. Such compensation may not occur with a duplication, because the duplication only carries one copy of each locus, while extrachromosomal arrays usually carry many copies. Functional compensation has been demonstrated for different transcripts produced from the same locus. With regards to the bli-4 locus on chromosome I in C. elegans, Thacker etal. (1995) demonstrated that one transcript could compensate for another that is not present, when produced in higher amounts from an extrachromosomal array. Mutations affecting blisterase A cause a blistering phenotype, while mutations affecting the other three isoforms, blisterases C, B, or D, result in lethality. They found that transgenics encoding for only blisterase A could rescue lethal mutations affecting exclusively blisterases B, C, or D, suggesting a functional redundancy between the isoforms. A similar phenomenon could occur between two related loci. Alternatively, let-393 could interact with a locus present in hEx28 and hEx33. For instance, hEx28 and hEx33 could encode for a transcription factor involved in the expression of let-393. If this were the case, then extra copies of a transcription factor could enable increased amounts of let-393 gene product, if the defect was in a let-393 control region. There are several other plausible explanations for the data obtained in this study, however, further experimentation is necessary to discern which is the most likely. 151 Alignment of Genetic and Physical M a p s : The cosmid rescue experiments in this study have enabled an alignment of the genetic and physical maps in the region of study, by the physical placement of 13 lethal mutations. These placements provide guidepost between the genetic and physical maps. From the positions of these anchored loci, the physical positions of six duplications and one deficiency can be predicted. In addition, the remaining loci in this region with placements on the genetic map alone, can now be placed within specific physical intervals defined by these anchored loci. The cosmid rescue results have also had an impact on the genetic map, by giving a right-left positioning to loci that were inseparable by complementation to duplications, or recombination experiments. It is interesting to note that the predicted positions on the physical map for six of the seven rearrangement breakpoints involved only three overlapping cosmids, C12H4, M01A12 and C07F10, all of which are carried by the hExlO array. These cosmids might have contained sequences that were more susceptible to breakage. Furthermore, the three cosmids in the hExlO array rescued the most genes (5/13) in this study (Figure 38), in addition to carrying the unc-40 locus (J. Culotti, personal communication). Three of the rescued loci, let-376, let-395 and let-530, were rescued by hExlO alone, placing them all on the cosmid M01A12 that is unique to hExlO. It is interesting to note that all three loci could only be partially rescued by hExlO. Genes that are physically clustered together are good candidates for a polycistronic transcription unit (Zorio et al. 1994). It is estimated that 26% of the genes in C. elegans are coordinately expressed in operons (Speith etal. 1993; Zorio etal. 1994). These three loci could be part of an operon. If they were part of an operon and their promoter was located some distance from these genes on a different cosmid, then their expression could fluctuate, thereby accounting for their partial rescue. In several cases the genes within an operon are functionally related, but in most cases this has not been evident. The co-regulation of genes may be advantageous for an organism with a small genome size. Operons can serve to reduce genome size, which is consistent with the apparent constraints upon the size of the gene clusters. The gene clusters have been shown to be approximately equal in size in all the autosomes, and are very compact, with presumably less than 152 50% intergenic DNA (Barnes et al. 1995). This compactness is presumably maintained by recombination suppression within the gene clusters (Kim and Rose 1987). If any of these genes are grouped within an operon, then it could be revealed with the completion of the sequencing of this region (Wilson et al. 1994; Waterston and Sulston 1995). From the genome sequencing project, it has been estimated that cosmids in the chromosome III gene cluster should carry on average seven genes per cosmid (Sulston etal. 1992; Wilson etal. 1994). From a Poisson calculation (McDowall 1990), the sDp2 region is estimated to contain approximately 225 essential genes, or one quarter of the total expected loci (Chapter 1 discussion). If the cosmids in the chromosome I gene cluster carry a similar number of genes per cosmid, then there should be approximately 2 essential genes per cosmid. Comparing the number of genetically mapped loci (27) to the number of contiguous cosmids (27) in this region (Figure 38), there are approximately one essential gene per cosmid represented in the sDp2 lethal set. Hence, we may have identified fewer than 80% of the essential genes in the dpy-5 unc-13 region, as originally predicted by a truncated Poisson calculation (McDowall 1990). In the dpy-5 unc-13 region, there are 15 genetic markers that have been positioned on the physical map (Figure 38), 13 from the lethal rescue data in this study, one involving the visible locus unc-40 (J. Culotti, personal communication), and one from the RFLP hP5 (Starr et al. 1989). From this collective data, there are 11 unique placements on a cosmid contig composed of 27 overlapping cosmids (this study alone provides 10 unique placements). If these cosmids are assumed to contain the average size of insert (34 kb) as determined by A Coulson et al. (1986), then this interval of the physical map consists of approximately 920 kb. Hence, there are links between the genetic and physical map on average every 85 kb over a 0.5 map unit region as a result of this and previous studies. This is comparable with the resolution obtained in Drosophila using a P element gene disruption method, which resulted in links between the genetic and physical maps at 100 kb intervals (Spradling etal. 1995). The future cloning of these genes should be greatly facilitated by the integration of the genetic and physical maps in this region. 153 Extrachromosomal A r r a y s as Mapping Tools: Germline transformation was used to create heritable extrachromosomal arrays, to be used as small duplications for fine scale mapping, and the alignment of the genetic and physical maps. Extrachromosomal arrays composed of small regions of DNA make very fine mapping tools, capable of separating closely mapping genes, as well as giving the genes a placement on the physical map. In this study, the cosmids were introduced into non-lethal strains so as to obtain stocks suitable for complementation to several genes. Both extrachromosomal arrays and duplications can be used for complementation analysis of lethal mutations. Duplications were used in Chapter one of this study to position essential loci in the region of study. Duplications are an effective way to position large numbers of lethal mutations on the genetic map (Howell et al. 1987; McDowall 1990; McKim etal. 1992). When duplication analysis is followed by complementation to extrachromosomal arrays, the smaller amount of DNA contained in cosmids enables a more precise localization of the genes. In this regard, cosmids are more useful than yeast artificial chromosomes (YACs), because in general YACs carry more DNA (up to 600 kb) than cosmids (around 34 kb) (Coulson et al. 1986; Coulson etal. 1988; Sulston etal. 1992), and are consequently more prone to shearing during transformation. The circular nature of cosmid DNA may also decrease shearing. In terms of stability, somatic loss of DNA can occur with duplications (McKim and Rose 1990), YACs (Hieter etal. 1985; Murray etal. 1986), and extrachromosomal arrays (Stinchcomb etal. 1985; Mello et cd. 1991), however, duplications are the most stable. The increased loss of exogenous DNA through oocytes as compared to sperm has been found for both duplications (McKim and Rose 1990) and extrachromosomal arrays (this study), however, the greater loss of arrays through oocytes makes them more difficult to manipulate genetically than duplications. Since duplications are chromosomally derived (Rose etal. 1984), the regulation of replication and transcription of a duplication should be closer to a wild-type level than for an extrachromosomal array. In addition, duplications usually carry only one copy of each locus, whereas extrachromosomal arrays can carry many copies. Hence, negative results are more likely to occur when a mutation is 154 complemented to an extrachromosomal array, than to a duplication, in cases where they both carry the locus under investigation. Yet, cosmid transformation has some unique advantages. The small size of cosmid DNA enables their use in investigating the expression of one or a few loci in isolation. Furthermore, by the use of cosmid subclones, a locus can be dissected for fine-structure mapping, or to determine its expression patterns. Thacker et al. (1995) used transgenics carrying various cosmid subclones to determine the functional activity of bli-4 isoforms. This would not be possible with duplications. Furthermore, the decreased stability of extrachromosomal arrays over duplications works to their advantage for mosaic analysis (Stinchcomb etal. 1985). The ease of co-transformation of cosmids allows cloned markers, such as ncl-1 (enlarged nucleoli) (E. M . Hedgecock, personal communication), to be incorporated into an array for mosaic analysis. In addition, extrachromosomal arrays can be linked to free duplications, thereby facilitating the introduction of a cloned gene or a marker construct onto the duplication (R. K. Herman, personal communication). This permits considerable flexibility in designing extrachromosomal elements for mosaic analysis. Hence, the use of duplications and extrachromosomal arrays for genetic analysis can augment one another. In Drosophila, P elements have been used for germline transformation (Rubin and Spradling 1983). P element integration in Drosophila displays some similarities to that of cosmids in C. elegans. Both can integrate at many site within the genome, with no evidence for homologous recombination (Rubin and Spradling 1983; Fire 1986). However, homologous recombination between plasmids or cosmids can occur to produce large P elements in Drosophila of around 54 kb (Rubin and Spradling 1983), or large extrachromosomal arrays of up to 200 cosmid copies in C. elegans (Stinchcomb et al. 1985; Mello et al. 1991). While injected P elements integrate into the Drosophila chromosome, the integration of injected cosmids in C. elegans is rare (Fire 1986; Mello etal. 1991). Both can confer advantages. Extrachromosomal arrays enable genes to be examined without concern for any position effects arising from integration. Conversely, P elements can be induced to change their location in the genome, facilitating the analysis of position effect mutants (Levis et al. 1985). DNA can be inserted into a 155 non-autonomous P element, which can integrate into the genome, but cannot catalyze its own transposition. Levis etal. (1985) demonstrated that the injection of a second P element, such as 'wings-clipped', can induce the transgenic P element to transpose, without itself being integrated into the genome. To study position effects on a transgenic locus in C. elegans, several integrated lines would need to be established. There are different approaches to studying gene expression and function in eukaryotes. In this regard, P elements are also useful for site-directed mutagenesis (Spradling et al. 1995). Cosmid transformation in C. elegans will invariably augment similar studies in other organisms. 156 CONCLUSIONS A lethal mutant library provides a valuable tool for discerning the biological function of essential genes. The sDp2 region of chromosome I has been screened quite extensively for mutations affecting essential genes, resulting in the recovery of 495 lethal mutations (Howell etal. 1987). This approach was taken to generate a 'mutant library' representing a large fraction of the essential genes in a 15 map unit region. This mutant library is estimated to be 80% saturated for mutations in essential genes (McDowall 1990). However, to attain the goal of correlating the biological function of these essential genes with the upcoming sequence data, the mutants in the sDp2 mutant library must be placed into complementation groups, and positioned on both the genetic and physical maps. Together with Howell et al. (1987), Howell and Rose (1990), McDowall (1990), and McKim etal. (1992), this thesis completes the analysis of 495 lethal mutations recovered using sDp2 with respect to the dpy-5 unc-13 region, as well as to other genetic markers. From this collaborative effort, a total of 256 mutations, defining 97 essential genes, have been positioned to the sDp2 region on the genetic map. In the dpy-5 unc-13 region, 61% of the sDp2-rescued essential genes have been identified by more than one allele. A range of mutant alleles, including nulls, is important for elucidating the biological function of a gene. The majority of mutations in the 97 essential genes represented in the mutant library have been characterized with respect to the time of their developmental arrest (Howell and Rose 1990; McDowall 1990; McKim etal. 1992). In this study, the mutant phenotypes of 15 adult steriles were further defined by microscopic analysis. Three of these sterile strains have been phenotypically rescued (ten probably lie outside the cosmids used to construct the transgenics). These characterizations give direction to further studies concerning the biological function of these genes. Future functional analysis of these genes could be approached by determining their spatial and temporal expression patterns using reporter gene fusions that measure protein levels (Lynch et al. 1995), or by fluorescence in situ hybridization (FISH) that measures mRNA levels (Birchall et 157 al. 1995). Using these methods, it is possible to identify individual cells that express a particular gene, and to determine the expression of that gene within those cells. Such technologies will provide a complementary set of genetic tools that can be used to determine the biological function of these genes, and their homologs in other species. Ultimately, the goal is to discern the biological function of every essential gene in this mutant library. However, by the shear magnitude of such an undertaking, this work must involve the entire C. elegans community. To make these mutations available to other researchers, it is important to integrate their genetic map position with the physical map, so as to gain access to the sequencing information currently being generated (Wilson etal. 1994; Waterston and Sulston 1995). The sequence data itself should shed light on some of the functions of these genes by comparisons to databases, once the mutations have been cloned. By the use of such databases, information can be gained from the sequencing of a wide variety of organisms, including vertebrates (Adams et al. 1995; Dietrich et al. 1995), invertebrates (Spradling etal. 1995; Waterston and Sulston 1995), plants (Goodman etal. 1995; Carels etal. 1995), yeast (Oliver etal. 1992; Dujon etal. 1994), and bacteria (Fleischmann etal. 1995; Fraserefa/. 1995). A systematic integration of the genetic and physical maps in the 0.5 map unit region between dpy-5 and bli-4 has resulted from the positioning of 13 essential loci from the sDp2 mutant library, by identifying the cosmid DNA needed to rescue mutations in these genes. These rescues provide ten unique placements on the physical map. Combined with the placement of the polymorphism hP5 (Starr et al. 1989), there are links between the genetic and physical maps on average every 85 kb over a 0.5 map unit region, thereby providing convenient guideposts between the maps. These links will facilitate the cloning not only of the anchored loci, but of all the loci that have been genetically mapped to this region, because unanchored loci can be placed within specific physical intervals as defined by these guideposts. This will have a considerable impact on any future approach to the cloning of loci in this region. With links provided approximately every second cosmid, the cloning of an unanchored locus in this region would best be approached by subcloning the cosmids lying within the physical interval defined by the guideposts, thereby 158 narrowing down the DNA fragments used for rescue. With this approach, a single open reading frame could more readily be correlated with the phenotypic rescue of a mutant. As such, the transgenics generated in this study have served their purpose as vehicles used to position the lethal mutations, and the value now lies in the integrated maps to provide access to sequencing data, and the mutants themselves to correlate this sequence with their biological function. The integration of genetic and physical maps will also yield insights into genome organization. A comparison of recombination frequency with the physical distance between loci will enable the correlation of the genetic and physical distribution of essential loci. In the sDp2 region, the genetic distribution of essential genes was found to be analogous to the distribution of visible loci, with the clustering of loci in a recombination suppressed region near the center of the chromosome (Howell etal. 1987; Kim and Rose 1987; Starr etal. 1989; Howell and Rose 1990; McDowall 1990; McKim et al. 1992). This clustering of loci is not due to recombination suppression alone, but represents a physical clustering as well. By analyzing the distribution of cDNA clones, Waterston etal. (1992) showed that gene density was higher in the central gene clusters than in the arms of the autosomes. A study by Barnes et al. (1995) revealed a striking similarity between the autosomes in terms of gene density, arrangement, and the number of crossovers per bivalent, despite variations in their length. Using the position of cloned genes and random cDNA clones, two-thirds of loci were found to occur in central clusters, where there was little variation in size (~7 Mb), gene density (~190 genes/Mb) and recombination rates between the autosomes, suggesting that the gene clusters may be under constraints for compactness. In contrast, the chromosome arms were found to carry eight-fold fewer genes (~25 genes/Mb), interspersed with single-copy noncoding DNA, and had high recombination rates. Consequently, in C. elegans, noncoding DNA appears to be a preferred target for crossing-over (Barnes et al. 1995). This is in contrast to most other organisms, such as Drosophila (Ashburner 1989), maize (Civardi etal. 1994; Carels etal. 1995), and yeast (Oliver et al. 1992), where crossing-over is associated with gene-rich regions. It has been proposed that repetitive elements like CeRep3 and 159 RcS5, that occur preferentially outside the gene clusters, could act to promote recombination in the autosome arms (Cangiano and La Volpe 1993; Barnes et al. 1995). From the sequencing data on chromosome III (Wilson etal. 1994), the genes within the central gene cluster were found in general to be evenly distributed. If the autosomal gene clusters bear striking similarity (Barnes etal. 1995), then the same even distribution of genes found within the chromosome III gene cluster (Wilson etal. 1994) should also occur within the chromosome I gene cluster. This appears to be true for the region analyzed in this study. There were no large gaps found between lethal mutations positioned on the physical map. Instead, ten of the 14 transgenic strains rescued 13 lethal mutations. The only apparent clustering of essential loci occurred in the cosmids carried by the array hExlO, which rescued mutations in five essential genes. This cluster of essential loci could represent an operon, within which genes are coordinately expressed. It has been estimated that 26% of genes in C. elegans are organized into operons comprised of two or more genes (Spieth et al. 1993; Zorio et al. 1994). The co-transcription of functionally, or temporally related genes may facilitate their coordinate expression. In addition to the correlation of genetic and physical map data, this study has yielded information on the genetic behaviour of extrachromosomal arrays, in regards to their stability in oocytes and sperm. In 12/16 transgenic strains, the arrays were found to be transmitted 2- to 7-fold more frequently in hermaphrodite sperm than oocytes, indicating that extrachromosomal DNA is predominantly transmitted through sperm in hermaphrodites. This phenomenon could be due to a difference between the expression and maintenance of extrachromosomal DNA in oocytes and sperm, or to the increased number of cell divisions that occur during oogenesis as compared to spermatogenesis. Alternatively, array stability could be affected by specific cosmid sequences, since the greatest disparity in array transmission between oocytes and sperm (7-fold) occurred in two strains that carried cosmid sequences in common. An array-induced mutant phenotype could provide information on the function of specific sequences. To conclude, this research has increased the accessibility of a set of lethal mutations within the sDp2 mutant library as a resource for structure-function correlation, as well as for 160 investigations into genome organization. Furthermore, these essential loci have become accessible on different levels. For instance, the genetic map data enables researchers to look for new alleles of the gene they are investigating. The integration of the genetic and physical maps enables researchers to place unanchored loci between the anchor points to determine which physical region to investigate, as well as facilitating the search for candidate mutations in open reading frames identified on the basis of sequence information. Finally, the phenotypic characterization of adult steriles, which can be extended to the other phenotypic classes in this region, will aid researchers looking for specific classes of mutations, such as those involved in germline development, germline sex determination, or embryogenesis. The ability to approach the mutants, or the information they generate, from multiple levels greatly enhances the value of a mutant library by making it more accessible. These mutations will gain in value with the completion of the C. elegans sequencing project, since a mutational approach is the best method of correlating sequence data with biological function, given that current databases can at best predict the biological role of only 50% of the open reading frames identified by sequencing (Hodgkin et al. 1995). As such the integration of genetic and molecular techniques are synergistic for whole genome analysis. The elucidation of structure-function relationships for C. elegans genes should help define the roles played by homologous genes in other organisms, since approximately 45% of open reading frames sequenced in C. elegans display significant similarity to non-C. elegans genes (Waterston and Sulston 1995). By extending this integrative approach to the remaining lethal mutations in the sDp2 set, this mutant library will become more accessible to the C. elegans community as a rich source of mutations in a wide variety of genes. With the cooperative use of this mutant library, we will begin to address the question of how these genes act collectively to regulate the whole organism. 161 BIBLIOGRAPHY ADAMS, M. D., A. R. KERLAVAGE, R . D. FLEISCHMANN, R . A . FULDNER, C. J . BULT etal., 1995. Initial assessment of human gene diversity and expression patterns based upon 83 million nucleotides of cDNA sequence. Nature 377: 3-17. AHRINGER, J., and J. KIMBLE, 1991. Control of the sperm-oocyte switch in Caenorhabditis elegans hermaphrodites by fem-3 3'untranslated region. Nature 349: 346-348. ASHBURNER, M . , 1989. Mapping and exchange, pp.451-501 in Drosophila: A Laboratory Handbook. Cold Spring Harbor Press, Cold Spring Harbor, N.Y. AUSTIN, J., and J. KIMBLE, 1987. glp-1 is required in the germline for regulation of the decision between mitosis and meiosis in C. elegans. Cell 51: 589-599. BARSTEAD, R . J . , L . KLEIMAN, and R . H. WATERSTON, 1991. Cloning, sequencing and mapping of an alpha-actin gene from the nematode Caenorhabditis elegans. Cell. Motil. BARNES, T. M . , Y . KOHARA, A . COULSON, and S. H E K 3 M I , 1995. Meiotic recombination, noncoding DNA and genomic organization in C. elegans. Genetics 141: 159-179. 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. BARTON, M . K., T. B. SCHEDL, and J. KIMBLE, 1987. Gain-of-function mutations of fem-3, a sex-determining gene in C. elegans. Genetics 115: 107-119. B E A N A N , M . J., and S. S T R O M E , 1992. Characterization of a germ-line proliferation mutation in C. elegans. Dev. 116: 755-766. B l R C H A L L , P. S., R. M . F lSHPOOL, and D. G. A L B E R T S O N , 1995. Expression patterns of predicted genes from the C. elegans genome sequence visualized by FISH in whole organisms. NatureGenet.il: 314-320. 20: 69-78. B L O W , J. J., and R. A. L A S K E Y , 1986. Initiation of DNA replication in nuclei and purified DNA by a cell-free extract of Xenopus eggs. Cell 4 7: 577-587. 162 B O S S Y , B . , L . M . C. HALL, and P. S P I E R E R , 1984. Genetic activity along 315 kb of the Drosophila chromosome. EMBO J. 3: 2537-2541. B O W E R M A N , B . , B . A . E A T O N , and J. R. P R I E S S , 1992. skn-1, a maternally expressed gene required to specify the fate of ventral blastomeres in the early C . elegans embryo: Cell 68: 1061-1075. B R E E D E N , L . , and K. N A S M Y T H , 1987. Similarity between cell cycle genes of budding yeast and fission yeast and the Notch gene of Drosophila. Nature 329: 651-654. B R E N N E R , S., 1974. The genetics of Caenorhabditis elegans. Genetics 7 7: 71-94. B R E S N A H A N , P. A. , R. L E D U C , L . T H O M A S , J. T H O R N E R , H. L . G I B S O N , A. J. B R A K E , P. J. B A R R , and G . T H O M A S , 1990. Human fur gene encodes a yeast KEX2-like endoprotease that cleaves pro-beta-NGF in vivo. J. Cell. Biol. I l l : 2851-2859. B R O W N I N G , H., and S . S T R O M E , 1996. A sperm-supplied factor required for embryogenesis in C. elegans. Dev. 122: 391-404. B U C H E R , E. A., and I. G R E E N W A L D , 1991. A genetic mosaic screen of essential zygotic genes in C. elegans. Genetics 128: 281-292. B U L L E R J A H N , A. M . E„ and D. L . R I D D L E , 1988. Fine-structure genetics of ama-1, an essential gene encoding the amanitin-binding subunit of RNA polymerase II in C . elegans. Genetics 120: 423-434. C A N G I A N O , G . , and A. L A V O L P E , 1993. Repetitive DNA sequences located in the terminal portion of the C. elegans chromosomes. Nuc. Acid Res. 21: 1133-1139. C A P O W S K I , E . E . , P. M A R T I N , C . G A R V I N , and S . S T R O M E , 1991. Identification of grandchildless loci whose products are required for normal germline development in the nematode C . elegans. Genetics 129: 1061-1072. C A R E L S , N., A. B A R A K A T , and G. B E R N A R D I , 1995. The gene distribution of the maize genome. Proc. Natl. Acad. Sci. USA 92: 11057-11060. C H A L F I E , M.1984. Neuronal development in C . elegans. Trends Neurosci. 7: 197-202. C H U M A K O V , I. M . , P. RlGAULT, I. L E G A L L , C . B E L L A N N E - C H A N T E L O T , A. BlLLAULT, etal., 1995. A YACcontig map of the human genome. Nature 3 7 7: 175-183. 163 ClVARDI, L . , Y . X lA, K . J. EDWARDS, P. S. SCHNABLE, and B. J. NlKOLAU, 1994. The relationship between genetic and physical distances in cloned al-sh.2 interval in the Zea mays L . genome. Proc. Natl. Acad. Sci. USA 9 1 : 8268-8272. CLANDIMN, R. T., and P. E . MAINS, 1993. Genetic studies of mei-1 gene activity during the transition from meiosis to mitosis in Caenorhabditis elegans. Genetics 134: 199-210. CLARK, D. V. , and D. L . BAILLIE, 1992. Genetic analysis and complementation by germ-line transformation of lethal mutations in the unc-22 IV region of C. elegans. M o l . Gen. Genet. 2 3 2 : 97-105. CLARK, D. V. , T. M . ROGALSKI, L . M . DONATI, and D. L . BAILLIE, 1988. The unc-22(IV) region of C. elegans: genetic analysis of lethal mutations. Genetics 1 1 9 : 345-353. COLLINS, J. E . , C. G. COLE, L . J. SMINK, C. L . GARRETT, M . A . LEVERSHA, etal., 1995. A high density Y A C contig map of human chromosome 22. Nature 3 7 7 : 367-371. CORNALL, R. J., T. J. AlTMAN, C. M . HEARNE, and J. A . TODD, 1991. The generation of a library of PCR-analyzed microsatellite variants for genetic mapping of the mouse genome. Genomics 10: 874-881. COULSON, A . , J. SULSTON, S. BRENNER, and J. KARN, 1986. Toward a physical map of the genome of the nematode C. elegans. Proc. Natl. Acad. Sci. USA 8 3 : 7821-7825. COULSON, A . , R. WATERSTON, J. KlFF, J. SULSTON, and Y . KOHARA, 1988. Genome linking with yeast artificial chromosomes. Nature 3 3 5: 184-186. CRITTENDEN, S. L . , E . R. TROEMEL, T. C. EVANS, and J. KIMBLE, 1994. GLP-1 is localized to the mitotic region of the C. elegans germline. Dev. 1 2 0 : 2901-2911. DIETRICH, W. F., N . G. COPELAND, D. J. GILBERT, J. C. MILLER, N . A . JENKINS, and E . S. LANDER, 1995. Mapping the mouse genome: current status and future prospects. Proc. Natl. Acad. Sci. USA 9 2 : 10849-10853. DOGGETT, N . A. , L. A . GOODWIN, J. G. TESMER, L . J. MEINCKE, D. C. BRUCE, etal. 1995. An integrated physical map of human chromosome 16. Nature 3 7 7 : 335-339. DONIACH, T., 1986. Activity to the sex-determining gene tra-2 is modulated to allow spermatogenesis in the C. elegans hermaphrodite. Genetics 114: 53-76. 164 DUJON, B., D. ALEXANDRAS, B. ANDRE, W. ANSORGE, V . BALADRON, et al., 1994. Complete DNA sequence of yeast chromosome XI. Nature 3 6 9 : 371-378. EDGAR, L . G. , and D. I. HIRSH, 1985. Use of a psoralen-induced phenocopy to study genes controlling spermatogenesis in Caenorhabditis elegans. Dev. Bio. 1 1 1 : 108-118. EDGLEY, M . L. , and D. L . RIDDLE, 1993. pp.3.281-3.318 in Genetic Maps: Locus Maps of Complex Organisms, Lower Eukaryotes, edited by S. J. O'Brien. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. ELLIS, H. M . , and H. R. HORVITZ, 1986. Genetic control of programmed cell death in the nematode Caenorhabditis elegans. Cell 44 : 817-829. ELLIS, R. E. , and J. KIMBLE, 1995. The fog-3 gene and regulation of cell fate in the germ line of C. elegans. Genetics 1 3 9 : 561-577. EMMONS, S., W., L . YESNER, K. RUAN, and D. KATZENBERG, 1983. Evidence for a transposon in C. elegans. Cell 3 2: 55-65. FIRE, A., 1986. Integrative transformation of Caenorhabditis elegans. EMBO J. 5 : 2673-2680. FIRE A. , D. ALBERTSON, W. HARRISON, and D. MOERMAN, 1991. Production of antisense RNA leads to effective and specific inhibition of gene expression in C. elegans muscle. Dev. 113 : 503-514. FIRE, A , and R. H. WATERSTON, 1989. Proper expression of myosin genes in transgenic nematodes. EMBO J. 8 : 3419-3428. FLEISCHMANN, R. D., M . D. ADAMS, O. WHITE, R. A. CLAYTON, E . W. KIRKNESS, etal, 1995. Whole-genome random sequencing and assembly of Haemophilus influenzae Rd. Science-269: 496-512. FODOR, A. , and P. D E A K , 1985. The isolation and genetic analysis of a C. elegans translocation (szTl) strain bearing an X-chromosome balancer. J. Genet. 6 4 : 143-157. FOLGER, K. R., K. THOMAS, and M . R. CAPECCHI, 1985. Nonreciprocal exchanges of information between DNA duplexes coinjected into mammalian cell nuclei. Mol. Cell. Biol. 5 : 59-69. 165 F O R B E S , D. J., M . W . K I R S C H N E R , and J. W . N E W P O R T , 1983. Spontaneous formation of nucleus-like structures around bacteriophage DNA microinjected into Xenopus eggs. Cell 34: 13-23. FOURNIER, R. E. K , and F . H. R U D D L E , 1977. Stable association of the human transgenome and host murine chromosomes demonstrated with trispecific microcell hybrids. Proc. Natl. Acad. Sci. USA 7 4: 3937-3941. F R A N C I S , R., M . K . B A R T O N , J. K I M B L E , and T. S C H E D L , 1995a. gld-1, a tumor suppressor gene required for oocyte development in C . elegans. Genetics 139: 579-606. F R A N C I S , R., E . M A I N E , and T. S C H E D L , 1995b. Analysis of the multiple roles of gld-1 in germline development: interactions with the sex determination cascade and the glp-1 signaling pathway. Genetics 139: 607-630. F R A S E R , C. M . , J. D. G O C A Y N E , O . W H I T E , M . D. A D A M S , R. A. C L A Y T O N , etal., 1995. The minimal gene complement of Mycoplasma genitalium. Science 2 7 0: 397-403. GEMMILL, R. M . , I. CHUMKOV, P. SCOTT, B. WAGGONER, P. RIGAULT, et al., 1995. A second-generation Y A C contig map of human chromosome 3. Nature 377: 299-302. G l B E R T , M . , J. S T A R C K , and B. B E G U E T , 1984. Role of the gonad cytoplasmic core during oogenesis of the nematode Caenorhabditis elegam. Biol. Cell 50: 77-86. GOFFEAU, A., 1995. Life with 482 genes. Science 270: 445-446. G O O D M A N , H. M . , J. R. E C K E R S , and C. D E A N , 1995. The genome of Arabidopsis thaliana. Proc. Natl. Acad. Sci. USA 92: 10831-10835. G R A H A M , P. L . , and J. K I M B L E , 1993. The mog-1 gene is required for the switch from spermatogenesis to oogenesis in Caenorhabditis elegans. Genetics 133: 919-931. G R O S V E L D , F., G. B. van A S S E N D E L F T , D. R. G R E A V E S , and G. K O L L I A S , 1987. Position-independent, high-level expression of the human p-globin gene in transgenic mice. Cell 51: 975-985. G U Y E R , M . S., and F. S. COLLINS, 1995. How is the human genome project doing, and what have we learned so far? Proc. Natl. Acad. Sci. USA 9 2: 10841-10848. 166 HARLAND, R. M . , and R. A. LASKEY, 1980. Regulated replication of DNA microinjected into eggs of Xenopus laevis. Cell 21: 761-771. HARLAND, R. M . , H. WEINTRAUB, and S. L . McKNlGHT, 1983. Transription of DNA injected in to Xenopus oocytes is influenced by template topology. Nature 3 0 2: 38-43. HAWKINS, N. C , J. THORPE, and T. SCHUPBACH, 1996. encore, a gene required for the regulation of germ line mitosis and oocyte differentiation during Drosophila oogenesis. Dev. 122: 281-290. HAZELRIGG, T., R. LEVIS, and G. M . RUBIN, 1984. Transformation of the white locus in Drosophila: dosage compensation, zeste interaction, and position effects. Cell 36: 469-481. HEDGECOCK, E. M . , J. G. CULOTTI, D. H. HALL, and B. D. STERN, 1987. Genetics of cell and axon migrations in Caenorhabditis elegans. Dev. 100: 365-382. HENDERSON, S. T., D. GAO, E. J. LAMBIE, and J. KIMBLE, 1994. lag-2 may encode a signal ligand for the GLP-1 and LIN-12 receptors of C . elegans Dev. 120: 2913-2924. HERMAN, R. K . , 1984. Analysis of genetic mosaics of the nematode C. elegans. Genetics 108: 165-180. HERMAN, R. K . , D. G. ALBERTSON, and S, BRENNER, 1976. Chromosome rearrangements in C. elegans. Genetics 8 3: 91-105. HIETER, P., C. MANN, M . SNYDER, and R. W. DAVIS, 1985. Mitotic stability of yeast chromosomes: a colony color assay that measures nondisjunction and chromosome loss. Cell 40: 381-392. HIRSH, D., K . J. KEMPHUES, D. T. STINCHCOMB, and R. JEFFERSON, 1985. Genes affecting early development in C . elegans. Cold Spring Harbor Symp. Quant. Biol 50: 69-78. HIRSH, D., D. OPPENHEIM, and M . KLASS, 1976. Development of the reproductive system of Caenorhabditis elegans. Dev. Biol. 49: 200-219. HOCHMAN, B., 1971. Analysis of chromosome 4 in D.melanogaster. II: EMS-induced lethals. Genetics 6 7: 235-252. 167 HODGKIN, J. , 1986. Sex determination in the nematode C. elegans: analysis of tra-3 suppressors and characterization offern genes. Genetics 114: 15-52. HODGKIN, J . , 1987. Sex determination and dosage compensation in C. elegans. Ann. Rev. Genet. 21: 133-154. HODGKIN, J . , 1990. Sex determination compared in Drosophila and Caenorhabditis. Nature 344: 721-728. HODGKIN, J . , H. R. HORVITZ, and S. BRENNER, 1979. Nondisjunction mutants of the nematode C. elegans. Genetics 9 1: 67-94. HODGKIN, J . , R. H. A. PLASTERK, and R. H. WATERSTON, 1995. The nematode C. elegans and its genome. Science 270: 410-414. HOWELL, A. M . , S. GlLMOUR, R. MANCEBO, and A . M . ROSE, 1987. Genetic analysis of a large autosomal region in C. elegans by the use of a free duplication. Genet. Res. 49: 207-213. HOWELL, A. M . , and A. M . ROSE, 1990. Essential genes in the hDf6 region of chromosome I in C. elegans. Genetics 126: 583-592. ISNENGHI, E . , R. CASSANDA, K . SMITH, K . DENICH, K . RADNIA, and G. von EHRENSTEIN, 1983. Maternal effects and temperature-sensitive period of mutations affecting embryogenesis in Caenorhabditis elegans. Dev. Biol. 98: 465-480. JOHNSEN, R. C . , and D. L . BAILLIE, 1991. Genetic analysis of a major segment [LGV left)] of the genome of C. elegans. Genetics 129: 735-752. JUDD, B. H. , 1995. Mutations of zeste that mediate transvection are recessive enhancers of position-effect variegation in D. melanogaster. Genetics 141: 245-253. KEMPHUES, K . J . , M . KUSCH, and N. WOLF, 1988. Maternal-effect lethal mutations on linkage group II of C. elegans. Genetics 120: 977-986. KIM, S. K . , and H . R. HORVITZ, 1990. The C. elegans gene lin-]0 is broadly expressed while r-required specifically for the determination of vulval cell fates. Genes and Dev. 4: 357-371. 168 KIM, J. S., and A . M . ROSE, 1987. The effect of gamma radiation on recombination in Caenorhabditis elegans. Genome 29: 457-462. KIMBLE, J. E. , 1981. Alterations in cell lineage following laser ablation of cells in the somatic gonad of Caenorhabditis elegans. Dev. Biol. 87: 286-300. KIMBLE, J., L . EDGAR, and D. HIRSH, 1984. Specification of male development in C. elegans: the/em genes. Dev. Biol. 105: 189-196. KIMBLE, J., and D. HlRSH, 1979. Postembryonic cell lineages of the hermaphrodite and male gonads in Caenorhabditis elegans. Dev. Biol. 7 0: 396-417. KIMBLE, J., J. HODGKIN, T. SMITH, and J. SMITH, 1982. Suppression of an amber mutation by microinjection of suppressor tRNA in C. elegans. Nature 299: 456-458. KIMBLE, J. E . , and W . J. SHARROCK, 1983. Tissue-specific synthesis of yolk proteins in Caenorhabditis elegans. Dev. Biol. 96: 189-196. KIMBLE, J. E . , and J. G. WHITE, 1981. On the control of germ cell development in Caenorhabditis elegans. Dev. Biol. 81: 208-219. KLOBUTCHER, L . A. , C. L . MILLER, and F. H . RUDDLE, 1980. Chromosome-mediated gene transfer results in two classes of unstable transformants. Proc. Natl. Acad. Sci. USA 7 7: 3610-3614. KLOBUTCHER, L . A., and F. H . RUDDLE, 1979. Phenotype stabilization and integration of transferred material in chromosome-mediated gene transfer. Nature 280: 657-660. KOTHARY, R., S. CLAPOFF, S. DARLING, M . D. PERRY, L. A. MORAN, and J. ROSSANT, 1989. Inducible expression of an hsp68-lacZ hybrid gene in transgenic mice. Dev. 105: 707-714. KOZLOVA, A. , V. F. SEMESHIN, I. V. TRETYAKOVA, E. B. KOKOZA, V. PlRROTTA, V. E . GRAFODATSKAYA, E . S. BELYAEVA, and I. F. ZHIMULEV, 1994. Molecular and cytogenetic characterization of the 10A1-2 band and adjoining region in the D. melanogaster polytene X chromosome. Genetics 13 6: 1063—1073. KRAUTER, K . , K . MONTGOMERY, S. YOON, J. LEBLANC-STRACESKI, B RENAULT, etal., 1995. A second-generation Y A C contig map of human chromosome 12. Nature 377: 321-324. 169 KUNES, S., D. BOTSTEIN, and M . S. FOX, 1990. Synapsis-mediated fusion of free DNA ends form inverted dimer plasmids in yeast.. Genetics 124: 67-80. KUWABARA, P. E. , and J. KIMBLE, 1992. Molecular genetics of sex determination in C. elegans. Trends Genet. 8: 164-168. LAMBIE, E. J., and J. KIMBLE, 1991. Two homologous regulatory genes, lin-12 and glp-1, have overlapping functions. Dev. 112: 231-240. LASKEY, R. A., A. D. MILLS, and N. R. MORRIS, 1977. Assembly of SV40 chromatin in a cell-free system from Xenopus eggs. Cell 10: 237-243. LEICHT, B. G., and J. J. BONNER, 1988. Genetic analysis of chromosomal region 67A-D of Drosophilamelanogaster. Genetics 119: 579-593. LEVIS, R., T. HAZELRIGG, and G. M . RUBIN, 1985. Effects of genomic position on the expression of transduced copies of the white gene of Drosophila. Science 229: 558-560. L'HERNAULT, S. W., D. C. SHAKES, and S. WARD, 1988. Developmental genetics of chromosome I spermatogenesis-defective mutants in the nematode Caenorhadhitis elegans. Genetics 120: 435-452. LIN, F., K. SPERLE, and N. STERNBERG, 1984. Model for homologous recombination during transfer of DNA into mouse L cells: role for DNA ends in the recombination process. Mol. Cell. Biol. 4: 1020-1034. LYNCH, A. S., D. BRIGGS, and I. A . HOPE, 1995. Developmental expression pattern screen for genes predicted in the C. elegans genome sequencing project. Nature Genet. 11: 309-313. MacGREGOR, G. R., A . E. MOGG, J. L . BURKE, and C. T. CASKEY, 1987. Histochemical staining of clonal mammalian cell lines expressing E. coli p-galactosidase indicates heterogeneous expression of the bacterial gene. Som. Cell Mol. Genet. 13: 253-265. MAINS, P. E . , I. A. SULSTON, and W. B. WOOD, 1990. Dominant maternal-effect mutations causing embryonic lethality in C. elegans. Genetics 125: 351-369. MARINI, N. J., L . D. ETKIN, and R. M . BENBOW, 1988. Persistence and replication of plasmid DNA microinjected into early embryos of Xenopus laevis. Dev. Biol. 127: 421-434. 170 McDOWALL, J. , 1990. Essential genes in the hDpl6/hDpl9 Region of chromosome I in C. elegans. M.Sc. thesis, University of British Columbia, Vancouver, B.C. McKlM, K . , A . M . HOWELL, and A . M . ROSE, 1988. The effects of translocations on recombination frequency in C. elegans. Genetics 120: 987-1001. McKlM, K. , and A. M . ROSE, 1990. Chromosome I duplications in C. elegans. Genetics 124: 115-132. McKlM, K . , T. STARR, and A. M . ROSE, 1992. Genetic and molecular analysis of the dpy-14 region in C. elegans. Mol. Gen. Genet. 233: 241-251. McLAREN, A., 1992. Testis determination and the H - Y hypothesis. Curr. Top. Dev. Biol. 23: 163-183. M C M A H O N , A . P., C. N. FLYTZANIS , B. R. HOUGH-EVANS, K . S. KATULA , R. J. BRITTEN, and E. H . DAVIDSON, 1985. Introduction of cloned DNA into sea urchin egg cytoplasm: replication and persistence during embryogenesis. Dev. Biol. 108: 420-430. MELLO, C. C , J. M . K R A M E R , D. STINCHCOMB, and V. AMBROS, 1991. Efficient gene transfer in C. elegans: extrachromosomal maintenance and integration of transforming sequences. EMBO J. 10: 3959-3970. MENEELY , P., and R. HERMAN, 1979. Lethals, steriles and deficiencies in a region of the X chromosome of C. elegans. Genetics 9 2: 99-115. MENEELY , P., and R. H E R M A N , 1981. Suppression and function of X-linked lethal and sterile mutations in C. elegans. Genetics 9 7: 65-84. MOERMAN, D. G., and D. L . BAILLIE, 1979. Genetic organization in Caenorhabditis elegans: Fine-structure analysis of the unc-22 gene. Genetics 91: 95-103. MOURNIER, N., and J. BRUN, 1980. A cytogenetical analysis of sterile mutants in Caenorhabditis elegans. Can. J. Genet. Cytol. 22: 391-403. MULLER, H . J., 1932. Further studies on the nature and causes of gene mutation, pp. 213-255 in Proceedings of the Sixth International Congress of Genetics, edited by D. Jones. Brooklyn Botanical Gardens, Menasha, Wisconsin. 171 MURPHY, T . D., and G. H . KARPEN, 1995. Localization of centromere function in a Drosophila minichromosome. Cell 8 2: 599-609. MURRAY, A . W., N . P. SCHULTES, and J . W. SZOSTAK, 1986. Chromosome length controls mitotic chromosome segregation in yeast. Cell 45: 529-536. MURRAY, A . W., and J . W. SZOSTAK, 1983. Construction of artificial chromosomes in yeast. Nature 305: 189-193. NEWPORT, J . , 1987. Nuclear reconstitution in vitro: stages of assembly around protein-free DNA. Cell 48: 205-217. NURSE, P. and Y. BlSSETT, 1981. Gene required in Gi for commitment to cell cycle and in G2 for control of mitosis in fission yeast. Nature 292: 558-560. OLIVER, S. G . , Q . J . M . VAN DER AART, M . L . AGNOSTONI-CARBONE, M . AIGLE, L . ALBERGHINA, et al., 1992. The complete D N A sequence of yeast chromosome III. Nature 357: 38-46. ORR-WEAVER, T . L . , J . W. SZOSTAK, and R. J . ROTHSTEIN, 1981. Yeast transformation: a model system for the study of recombination. Genetics 7 8: 6354-6358. PALMITER, R. D., and R. L . BRINSTER, 1986. Germ-line transformation in mice. Ann. Rev. Genet. 20: 465-499. PARK, E. C , and H . R. HORVITZ, 1986. Mutations with dominant effects on the behaviour and morphology of the nematode C. elegans. Genetics 113: 821-852. PAULI, D., B. OLIVER, and A . P. MAHOWALD, 1995. Identification of regions interacting with ovo mutations: potential new genes involved in germline sex determination or differentiation in D. melanogaster. Genetics 139: 713-732. PERRIMON, N . , D. MOHLER, L . ENGSTROM, and A . P. MAHOWALD, 1986. X-linked female-sterile loci in D. melanogaster. Genetics 113: 695-712. PERRY, W. L . , T . J . VASICEK, J . J . LEE, J . M . ROSSI, L . ZENG, T . ZHANG, S. M . TlLGHMAN, and F . CONSTANTINI, 1995. Phenotypic and molecular analysis of a transgenic insertional allele of the mouse fused locus. Cell 141: 321-332. 172 PETERS, K., J. S. McDOWALL, and A . M . ROSE, 1991. Mutations in the bli-4 (I) locus of Caenorhabditis elegans disrupts both adult cuticle and early larval development. Genetics 129: 95-102. PLASTERK, R. H. A. , and J. T. M . GROENEN, 1992. Targeted alterations of the C. elegans genome by transgene instructed DNA double strand break repair following Tel excision. EMBO J. 11: 287-290. PRASAD, S. S., T. STARR, and A. M . ROSE, 1993. Molecular characterization in the dpy-14 region identifies the S-adenylhomocysteine hydrolase gene in Caenorhabditis elegans. Genome 3 6: 57-65. PRIESS, J. R., H. SCHNABEL, and R. SCHNABEL, 1987. The glp-1 locus and cellular interactions in early C. elegans embryos. Cell 5 1: 601-611. RIDDLE, D. L . , 1982. Developmental biology of C. elegans: Symposium introduction. J. Nematol. 14: 238-239. ROBERTS, T. M . , and S. W A R D , 1982. Centripedal flow of pseudopodial surface components could propel the amoeboid movement of Caenorhabditis elegans spermatozoa. J. Cell Biol. 92: 132-138. RODESCH, C , P. K. GEYER, J. S. PATTON, E . B A E , and R. N. NAGOSHI, L995. Developmental analysis of the ovarian tumor gene during Drosophila oogenesis. Genetics 141: 191-202. ROEHL, H. , and J. KIMBLE, 1993. Control of cell fate in C. elegans by a Glp-1 peptide consisting primarily of ankyrin repeats. Nature 2 6 4: 632-635. ROGALSKI, T., and D. L. BAILLIE, 1985. Genetic organization of the unc-22 IV gene and the adjacent region in C. elegans. Mol. Gen. Genet. 201: 409-414. ROGALSKI, T., A. M . E. BULLERJAHN, and D. L . RIDDLE, 1988. Lethal and amanitin-resistant mutations in the C. elegans ama-1 and ama-2 genes. Genetics 120: 409-422. ROGALSKI, T., D. MOERMAN, and D. L . BAILLIE, 1982. Essential genes and deficiencies in the unc-22 IV region of C. elegans. Genetics 102: 725-736. ROSE, A. M . , and D. L. BAILLIE, 1979. Effect of temperature and parental age on recombination and nondisjunction in Caenorhabditis elegans. Genetics 9 2: 409-418. 173 ROSE, A. M . , and D. L . BAILLIE, 1980. Genetic organization of the region around unc-15(f), a gene affecting paramyosin in C. elegans. Genetics 9 6: 639-648. 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 C. elegans. Mol. Gen. Genet. 188: 286-291. 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. ROSENBERG , C , R. J. FLORIJN, F . M . V A N D E RIJKE, L. A. J. BLONDEN , T. K. R A A P , G.-J. B . OMMEN, and J. T. D E N DUNNEN, 1995. High resolution DNA fiber-fish on yeast artificial chromosomes: direct visualization of DNA replication. Nature Genet. 10: 477-479. ROSENBLUTH, R. E . , C . CUDDEFORD, and D. L . BAILLIE, 1983. Mutagenesis in C. elegans: A rapid eukaryotic mutagen test system using the reciprocal translocation, eTl(iii;v). Mut. Res. 110: 39-48. ROSENBLUTH, R. E . , T. ROGALSKI, R. JOHNSEN, L . ADDISON, and D. L . BAILLIE, 1988. Genomic organization in C. elegans: deficiency mapping on LGV(left). Genet. Res. 5 2 : 105-118. RUBIN, G. M . , and A. C . SPRADLING, 1983. Vectors for P element-mediated gene transfer in Drosophila. Nuc .Ac idRes . i l : 6341-6351. SAKAGUCHI, J . , and M . YAMAMOTO, 1982. Cloned ural locus of S. pombe propagates autonomously in this yeast assuming a polymeric form. Proc. Natl. Acad. Sci. USA 7 9: 7819-7823. SAMBROOK, J., E . F . FRITSCH, and T. MANIATIS, 1989. Molecular Cloning: A Laboratory Manual. Second ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York SCANGOS, G., and F . H. RUDDLE, 1981. Mechanisms and applications of DNA-mediated gene transfer in mammalian cells. Gene 14: 1-10. 174 SCHEDL, T., and J . KIMBLE, 1988. fog-2, a germ-line-specific sex determination gene required for hermaphrodite spermatogenesis in Caenorhabditis elegans. Genetics 119: 43-61. SCHRICK, J . J., M . E. DICKINSON, B . L . M . HOGAN, P. B . SELBY, and R. P. WOYCHIK, 1995. Molecular and phenotypic characterization of a new mouse insertional mutation that causes a defect in the distal vertebrae of the spine. Cell 140: 1061-1067. SCHUPBACH, T., 1987. Germline and soma cooperate during oogenesis to establish the dorso-ventral pattern of egg shell and embryo in Drosophilamelanogaster. Cell 49: 699-707. SEYDOUX, G., and I. GREENWALD, 1989. Cell autonomy of lin-12 function in a cell fate decision inC. elegans. Cell 57: 1237-1245. SEYDOUX, G., T. SCHEDL, and I. GREENWALD, 1990. Cell-cell interactions prevent a potential inductive interaction between soma and germline in C. elegans. Cell 6 1: 939-951. SlGURDSON, D. , G. SPANIER, and R. HERMAN, 1984. C . elegans deficiency mapping. Genetics 108: 331-345. SHANNON, M . P., T. C. K A U F M A N , M. W . SHEN, and B . H. JUDD, 1972. Lethality patterns and morphology of selected lethal and semi-lethal mutations in the zest-white region of Drosophilamelanogaster. Genetics 7 2: 615-638. SCHMIDT, R., J . WEST , K . LOVE , Z. LENEHAN , C . LISTER , H . THOMPSON , D. BOUCHEZ , and C. DEAN, 1995. Physical map and organization of Arabidopsis thaliana chromosome 4. Science 2 7 0: 480-483. SPENCE, A. M . , A. COULSON, and J . HODGKIN, 1990. The product of fem-1, a nematode sex-determining gene, contains a motif found in cell cycle control proteins and receptors for cell-cell interactions. Cell 60: 981-990. SPIETH, J., G. BROOKE , S. KUERSTEN, K . L E A , and T. BLUMENTHAL , 1993. Operons in C. elegans: polycistronic mRNA precursors are processed by transplicing of SL2 to downstream coding regions. Cell 7 3: 521-532. SPRADLING , A. C , D. M . STERN, I. K ISS, J . ROOTE, T. LAVERTY , and G. M . RUBIN , 1995. Gene disruptions using P transposable elements: an integral component of the Drosophila genome project. Proc. Natl. Acad. Sci. USA 92: 10824-10830. 175 STARR, T., A. M . HOWELL, J. McDOWALL, K. PETERS, and A . M . ROSE, 1989. Isolation and mapping of DNA probes within the LGI gene cluster of C. elegans. Genome 32: 365-372. STELLER, H. , and V. PlRROTTA, 1985. Fate of DNA injected into early Drosophila embryos. Dev. Biol. 109: 54-62. STINCHCOMB, D. T., J. E. SHAW, S. H. C A R R , and D. HlRSH, 1985. Extrachromosomal DNA transformation of Caenorhabditis elegans. Mol. Cell. Biol. 12: 3484-3496. STUART, G. W . , J. R. VlELKIND, J. V. McMURRAY, and M . WESTERFIELD, 1990. Stable lines of transgenic zebrafish exhibit reproducible patterns of transgene expression. Dev. 109: 577-584. SULSTON, J., and S. BRENNER, 1974. The DNA of C. elegans. Genetics 7 7: 95-104. SULSTON, J., Z. DU, K. THOMAS, R. WILSON, L . HILLIER, et al., 1992. The C. elegans genome sequencing project: a beginning. Nature 356: 37-41. SULSTON, J., and H . R. HORVITZ, 1977. Post-embryonic cell lineages of the nematode C. elegans. Dev. Biol. 56: 110-156. 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. SWAIN, J. L . , T. A. STEWART, and P. LEDER, 1987. Parental legacy determines methylation and expression of an autosomal transgene: a molecular mechanism for parental imprinting. Cell 50: 719-727. THACKER, C , K. PETERS, M . SRAYKO, and A. M . ROSE, 1995. The bli-4 locus of Caenorhabditis elegans encodes structurally distinct kex2/subtilisin-like endoproteases essential for early development and adult morphology. Genes and Dev. 9 : 956-971. TOMKIEL, J., L. FANTI, M . BERLOCO, L. SPINELLI, J. W . T A M K U N , B. T. WAKIMOTO , and S. PlMPINELLI, 1995. Developmental genetical analysis and molecular cloning of the abnormal oocyte gene of D.melanogaster. Genetics 140: 615-627. W A K E , C. T. , F. V E R N A L E O N E , and J. H . W I L S O N , 1985. Topological requirements for homologous recombination among DNA molecules transfected into mammalian cells. Mol. Cell. Biol. 5 : 2080-2089. 176 W A R D , S., D. J . BURKE , J . E . SULSTON, A . R . COULSON , D. G . ALBERTSON , D. A M M O N S , M . K L A S S , and E. HOGAN, 1988. Genomic organization of major sperm protein genes and pseudogenes in the nematode C. elegans. J . Mol. Biol. 199: 1-13. WARD, S., and J . S. CARREL , 1979. Fertilization and sperm competition in the nematode C. elegans. Dev. Biol. 73: 304-321. WATERSTON, R . H . , and G. R . FRANCIS, 1985. Genetic analysis of muscle development in C. elegans. Trends Neurosci. 8 : 270-276. WATERSTON, R . , C. MARTIN, M . CRAXTON, C. HUYNH, A. COULSON, etal, 1992. A survey of expressed genes in C. elegans. Nature Genet. 1: 114-123. WATERSTON, R . , and J . SULSTON, 1995. The genome of C. elegans. Proc. Natl. Acad. Sci. USA 9 2: 10836-10840. W A Y , J . C , and M . CHALFIE, 1988. mec-3, a homeobox-containing gene that specifies differentiation of the touch receptor neurons in C. elegans. Cell 5 4: 5-16. W l E S C H A U S , E . 1996. Embryonic transcription and the control of developmental pathways. Genetics 142: 5-10. WHARTON, K . A. , K . M . JOHANSON, T. X U , and S. ARTAVANIS-TSAKONAS, 1985. Nucleotide sequence from the neurogenic locus Notch implies a gene product that shares homology with proteins containing EGF-like repeats. Cell 43: 567-581. WILLIAMS, B. D., B. SCHRANK, C. HUYNH, R . SHOWNKEEN, and R . H . WATERSTON, 1992. A genetic mapping system in C. elegans based on polymorphic sequence-tagged sites. Genetics 131: 609-624. WILSON, R . , R . A l S N S C O U G H , K . ANDERSON, C. BAYNES, M . BERKS, etal, 1994. 2.2 Mv of contiguous sequence from chromosome III of C. elegans. Nature 368: 32-38. WOOD, W . B., R . HECHT, S. CARR, R . VANDERSLICE, N. WOLF , and D. HIRSH, 1980. Parental effects and phenotypic characterization of mutations that affect early development in Caenorhabditis elegans. Dev. Biol. 74: 446-469. YOCHEM, J., and I. GREENWALD, 1989. glp-J and lin-12, genes implicated in distinct cell-cell interactions in C. elegans, encode similar transmembrane proteins. Cell 5 8: 553-563. 177 ZARKOWER, D . , and J . HODGKIN, 1992. Molecular analysis of the C. elegans sex determination gene tra-1. C e l l 7 0: 237-249. ZORIO, D . A . R . , N . N . CHENG, T. BLUMENTHAL , and J . SPIETH, 1994. Operons as a common form of chromosomal organization in C. elegans. Nature 3 7 2: 270-272. ZWAAL, R . R . , A . BROEKS, J . VAN MEURS, J . T . M . GROENEN, and R . H . A . PLASTERK, 1993. Target-selected gene inactivation in C. elegans by using a frozen transposon insertion mutant bank. Proc. Natl. Acad. Sci . U S A 90: 7431-7435. APPENDIX 1: Strains Used S t r a i n s G e n o t y p e O r i g i n KR4 unc-11 (e47) D. Baillie KR236 dpy-5 unc-131 ++ ; sDp2 A. Rose KR281 dpy-5 let-390(h44) unc-13 ; sDp2 A. M . Howell KR290 dpy-5 bli-4 (h42) unc-13 ; sDp2 A. M . Howell KR350 dpy-5 let-391 (h91) unc-13 ; sDp2 A. M. Howell KR355 dpy-5 let-384 (h84) unc-13 ; sDp2 A. M . Howell KR357 dpy-5 let-382 (h82) unc-13 ; sDp2 A. M. Howell KR358 dpy-5 let-355 (h81) unc-13 ; sDp2 A. M . Howell KR359 dpy-5 let-380 (h80) unc-13 ; sDp2 A. M. Howell KR362 dpy-5 let-388 (h88) unc-13 ; sDp2 A. M . Howell KR363 dpy-5 let-387(h87) unc-13 ; sDp2 A. M. Howell KR426 dpy-5 let-370 (h!28) unc-13 ; sDp2 A. M. Howell KR429 dpy-5 let-381 (hJ07) unc-13 ;sDP2 A. M . Howell KR430 dpy-5 let-377 (hi 10) unc-13 ; sDp2 A. M . Howell KR432 dpy-5 let-386 (hi 17) unc-13 ; sDp2 A. M. Howell KR434 • dpy-5 let-367 (hi 19) unc-13 ; sDp2 A. M. Howell KR444 dpy-5 let-378 (h!24) unc-13 ; sDp2 A. M . Howell KR446 dpy-5 let-376 (hl30) unc-13 ; sDp2 A. M . Howell KR454 dpy-5let-379 (h!27) unc-13 ; sDp2 A. M. Howell KR531 dpy-5 let-396 (h217) unc-13 ; sDp2 A. M . Howell KR539 dpy-5 let-393 (h225) unc-13 ; sDp2 A. M . Howell KR612 dpy-5 let-395 (h271) unc-13 ; sDp2 A. M. Howell KR621 dpy-5 let-601 (h281)unc-13 ; sDp2 D. Pilgrim KR623 dpy-5 let-602 (h283) unc-13 ; sDp2 D. Pilgrim Appendix 1: continued Strains Genotype Origin KR646 dpy-5 let-605 (h312) unc-13 ; sDp2 J. McDowall KR633 dpy-5 let-603 (h289) nuc-13 ; sDp2 J. McDowall KR634 dpy-5 let-599 (H290) unc-13 ; sDp2 J. McDowall KR636 dpy-5 let-606 (h292) unc-13 ; sDp2 J. McDowall KR637 dpy-5 let-604 (1x293) unc-13 ; sDp2 J. McDowall KR727 dpy-5 let-607 (h402) unc-13 ; sDp2 J. McDowall KR1280 dpy-5 dpy-14 ;hDp!3 K. McKim KR1282 dpy-5 dpy-14 ; hDpl6 K. McKim KR1284 dpy-5 dpy-14 ; hDpl5 K. McKim KR1293 dpy-5 dpy-14 ; hDpl2 K. McKim KR1294 dpy-5 dpy-14; hDpll K. McKim KR1331 dpy-5 let-544 (h692) unc-13 ; sDp2 K. McKim KR1334 dpy-5 let-610 (h695) unc-13 ; sDp2 J. McDowall KR1347 dpy-5 let-608 (h706) unc-13 ; sDp2 J. McDowall KR1395 dpy-5 bli-4 (H754) unc-13 ; sDp2 J. McDowall KR1458 unc-11 dpy-14 ; + 1 szTl [++ ; lon-2] K. McKim KR1504 dpy-5 let-545 (h842) unc-13 ; sDp2 K. McKim KR1512 dpy-5 let-611 (h850) unc-13 ; sDp2 J. McDowall KR1598 dpy-5 let-538(h990) unc-13;szTl[unc-29;lon-2 ] K. McKim KR1772 dpy-5 dpy-14 ; hDp37 K. McKim KR1760 dpy-5 dpy-14 ; hDp41 K. McKim KR1745 dpy-5 dpy-14 ; hDp61 K. McKim KR1817 dpy-5 dpy-14 ; hDp72 K. McKim APPENDIX 2: Strains of New Lethal Mutations Used S t r a i n s G e n o t y p e O r i g i n KR672 dpy-5 h351 unc-13 •sDp2 * A. Rose Lab KR680 dpy-5 h359 unc-13 • sDp2 A. Rose Lab KR683 dpy-5 h362 unc-13 , • sDp2 A. Rose Lab KR695 dpy-5 h375 unc-13 • sDp2 A. Rose Lab KR713 dpy-5 h388 unc-13 • sDp2 A. Rose Lab KR726 dpy-5 h401 unc-13 • sDp2 A. Rose Lab KR733 dpy-5 h408 unc-13 • sDp2 A. Rose Lab KR741 dpy-5 h416 unc-13 • sDp2 A. Rose Lab KR757 dpy-5 h433 unc-13 • sDp2 A; Rose Lab KR761 dpy-5 h437 unc-13 • sDp2 A. Rose Lab KR778 dpy-5 h454 unc-13 • sDp2 A. Rose Lab KR806 dpy-5 h475 unc-13 •sDp2 A. Rose Lab KR890 dpy-5 h510 unc-13 • sDp2 A. Rose Lab KR1316 dpy-5 h677 unc-13 • sDp2 A. Rose Lab KR1339 dpy-5 h715 unc-13 • sDp2 A. Rose Lab KR1353 dpy-5 h712 unc-13 •sDp2 A. Rose Lab KR1370 dpy-5 h729 unc-13 • sDp2 A. Rose Lab KR1374 dpy-5 h733 unc-13 •sDp2 A. Rose Lab KR1377 dpy-5 h736 unc-13 •sDp2 A. Rose Lab KR1382 dpy-5 h741 unc-13 •sDp2 A. Rose Lab KR1393 dpy-5 h752 unc-13 • sDp2 A. Rose Lab KR1394 dpy-5 h753 unc-13 •sDp2 A. Rose Lab KR1407 dpy-5 h764 unc-13 • sDp2 A. Rose Lab KR1409 dpy-5 h766 unc-13 • sDp2 A. Rose Lab 181 Appendix 2: continued Strains Genotype Origin KR1410 dpy-5 h767 unc-13 ; sDp2 A. Rose Lab KR1418 dpy-5 h775 unc-13 ; sDp2 A. Rose Lab KR1437 dpy-5 h794 unc-13 ; sDp2 A. Rose Lab KR1440 dpy-5 h797 unc-13 ; sDp2 A. Rose Lab KR1441 dpy-5 h798 unc-13 ; sDp2 A. Rose Lab KR1447 dpy-5 h804 unc-13 ; sDp2 A. Rose Lab KR1448 dpy-5 h805 unc-13 ; sDp2 A. Rose Lab KR1451 dpy-5 h808 unc-13 ; sDp2 A. Rose Lab KR1455 dpy-5 h812 unc-13 ; sDp2 A. Rose Lab KR1485 dpy-5 h823 unc-13 ; sDp2 A. Rose Lab KR1490 dpy-5 h828 unc-13 ; sDp2 A. Rose Lab KR1494 dpy-5 h832 unc-13 ; sDp2 A. Rose Lab KR1505 dpy-5 H843 unc-13 ; sDp2 A. Rose Lab KR1518 dpy-5 h865 unc-13 ; sDp2 A. Rose Lab KR1523 dpy-5 h870 unc-13 ; sDp2 A. Rose Lab A. Rose lab participants: J. Babity, S. Gilmour, L . Harris, A . M . Howell, J. S. Kim, N. Mawji, J. McDowall, K. McKim, K. McNeil, K. Peters, T. Starr, B. Rattray, A. Rose and M . Zetka. 182 APPENDIX 3: Complementation Results Below is the complementation table summarizing the data used to assign gene names to the new essential genes in the hDpl3lhDpl6 region. In the table below, a '+' indicates complementation and a'-' indicates failure to complement. Heterozygous lethal-bearing males are listed across the top of the table and homozygous .vDp-2-rescued lethal-bearing hermaphrodites are listed down the side. h351 h359 h362 h375 h388 h401 h408 h416 h433 h43 Iet-367(hll9) + + + + + + -i- + + + let-376(hl 30) + + + + + + + - + + let-377(hl 10) + + + + + + + + - + let-378(hl24) + + + + + - + + + + let-379(hl27) + + + + + + + + + + let-384(h84) + + + + - + + + + + let-388(h88) + + + + + + + + + + let-391(h91) + + + + + + + + + + let-393(h225) + + + - + + + + + + let-395(h271) + + + + + + + + + + let-603(h289) + + + + + + - + + -h677 + - + h715 + h752 + h767 + + + h775 + - + h797 - + -h798 -h808 - + -Appendix 3: continued h.454 h475 h510 h677 h712 h715 h729 h733 h736 h741 let-367(hll9) + + + + + + + + + + let-376(hJ30) + + + + + + + + + + let-377(hll0) + + + + + + + + + + let-378(hl24) + + + + + + + + + + let-379(M27) + + + + - + + + + + let-384(h84) • - + + + • + + + + + + let-388(h88) + + + + + + - + + + let-391(h91) + - + + + + + + - + let-393(h225) + + + + + + + + + + let-395(h27J ) + + + + + + + + + + let-603(h289) + + + + + + + + + + h677 + + + + h767 + - + h775 + - + + h797 - + + + -h808 - + + + -Appendix 3: continued h752 h753 h764 h766 let-367(hll9) + + + + let-376(hl30) + + - + let-377(hll0) + + + -let-378(M24) + + + + let-379(hl27) + + + + let-384(h84) + + + + let-388(h88) + + + + let-391(h91) + + + + let-393(h225) + + + + let-395(h271) + + + + let-603(h289) + + + + h677 + + h715 + + h733 h752 + K767 + + h775 + + h797 + + h798 h808 + + h767 h775 h794 h797 h798 h804 + + + + + + + + + + + -+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + Appendix 3: continued h805 h808 h812 h823 H828 h832 h843 h865 h870 let-367(hll9) + + + + + + + + let-376(h!30) - + + - + - + + -let-377(hl 10) + + + + + + + + + let-378(hl24) + + + + + + + + + let-379(h!27) + + + + + + + + let-384(h84) + + + + + + + - + let-388(h88) + + + + + + - + + let-391(h91) + + + - + + + + + let-393(h225) + + + + + + + + + let-395(h271) + + + + + + + + + let-603(h289) + + - + + + + + + h677 + -h767 + h775 + -h797 - + h808 + 186 APPENDIX 4: Deriving the Equation for the Calculation of Gametic Frequency of Ex Transmission The frequency of transmission of extrachromosomal arrays to hermaphrodite gametes was calculated by scoring progeny of selfed transgenic (unc-11 ;hEx) hermaphrodites. In the Punnett square below, (p) = frequency of array-bearing gametes and (1 - p) = frequency of non-array-bearing gametes. (P) d-p) unc-1 l;hEx unc-11 ip) ip)2 Pd-P) unc-11 ;hEx unc-1 l;hEx unc-11 ;hEx d-p) Pd-P) (l-p)2 unc-11 unc-1 l;hEx unc-11 The ratio of transgenic : non-transgenic progeny = x 2p(\ -p) + p2 d - p ) 2 2p - 2p2 + p2 x = x (1 -P)2 2P - P2 "(1 -P) _P(2 - p) (1 -p)2 187 Appendix 4: continued Let y = 1 - p therefore p = 1 - y (1 - V) (2 - 11 - vi) X ~ v2 (1 - v) (1 + v) X = 1 3,2 Therefore, v = ± V(x + 1) Butp<\, thereforey is always +, so: 188 APPENDIX 5: Sequence of Primers Used for PCR Analysis Primer Sequence (5'-*3') Amplification j Target Sequence I Product j RL12 | CGT C C G T T C T T G A G G G T G RL14 | C T A A G A TGC T C G C C A A G G ! RL17 | CGT C C G GCG C A C A G A A G C RL18 j G T G C T G A G C C C G GCC A A A f 580bp | C. elegans A H H gene 880bp j Lorist cosmid vector | 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0087269/manifest

Comment

Related Items