UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The role of the let-268 and spc-1 genes in the early stages of muscle development in Caenorhabditis elegans Norman, Kenneth Rich 2000

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

Item Metadata

Download

Media
831-ubc_2000-565963.pdf [ 22.52MB ]
Metadata
JSON: 831-1.0099579.json
JSON-LD: 831-1.0099579-ld.json
RDF/XML (Pretty): 831-1.0099579-rdf.xml
RDF/JSON: 831-1.0099579-rdf.json
Turtle: 831-1.0099579-turtle.txt
N-Triples: 831-1.0099579-rdf-ntriples.txt
Original Record: 831-1.0099579-source.json
Full Text
831-1.0099579-fulltext.txt
Citation
831-1.0099579.ris

Full Text

The role of the let-268 and spc-1 genes in the early stages of muscle development in Caenorhabditis  elegans  by Kenneth Rich Norman B . S c , The University of New Hampshire, 1992  A THESIS SUBMITTED IN P A R T I A L F U L F I L L M E N T OF T H E REQUIREMENTS OF THE D E G R E E 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 (Department of Zoology)  We accept this thesis as conforming to the required standard  T H E UNIVERSITY OF BRITISH C O L U M B I A March 2000 <Q Kenneth Rich Norman, 2000  In  presenting  degree at the  this  thesis  in  University of  partial  fulfilment  of  this thesis for  department  or  by  his  or  requirements  British Columbia, 1 agree that the  freely available for reference and study. I further copying of  the  representatives.  an advanced  Library shall make it  agree that permission for extensive  scholarly purposes may be her  for  It  is  granted  by the  understood  that  head of copying  my or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department The University of British Columbia Vancouver, Canada  DE-6 (2/88)  Abstract Myogenesis in Caenorhabditis elegans is composed of several stages: the generation of presumptive myoblasts; the migration of these cells dorsally and ventrally from their lateral position to form the muscle quadrants; the accumulation of muscle specific proteins in these cells; the polarization of the cells; and, finally, the organization of sarcomeres and attachment structures within the cells. The goal of this thesis work has been to identify genes involved in this process. From a mutant screen designed to identify mutations that disrupt muscle development during C. elegans embryogenesis, eight mutations have been isolated and placed into five distinct classes.  Two of the mutants identified in this screen have been characterized in greater detail and the affected genes have been cloned. One of the mutants, let-268 (a class IV mutant), encodes a procollagen processing enzyme, procollagen lysyl hydroxylase, required for post-translational modifications of collagen. In let-268 mutants the processing and secretion of type IV collagen is disrupted. An examination of the body wall muscle in these mutant animals reveals normal myofilament assembly prior to contraction. However, once body wall muscle contraction commences the muscle cells separate from the underlying epidermal layer (hypodermis) and the myofilaments become disorganized. These observations indicate that type IV collagen is required in the basement membrane for mechanical support and not for organogenesis of the body wall muscle. The other mutant examined in greater detail is spc-1 (class V mutant), spc-1 encodes the only a spectrin gene in the C. elegans genome. Animals lacking functional a spectrin die just ^ after hatching and have defects in myofilament organization. More specifically, when  ii  compared to myofilaments in wild type animals, the myofilaments in the mutant animals are abnormally oriented relative to the longitudinal axis of the embryo. In cross section, the myofilaments appear to be loosely associated with the sarcolemma as compared to wild type. In addition, analysis of the basement membrane and the hypodermis of spcl(ra409) mutants provides evidence that the body wall muscle directs where the basement membrane is established and signals the hypodermis to determine where muscle anchoring structures are assembled in this tissue.  iii  Table of Contents Abstract  ii  Table of Contents  iv  List of Tables  viii  List of Figures  ix  List of Abbreviations  xii  Acknowledgment  xiv  Dedication  xv  Chapter 1. Introduction  1  Background  4  Summary ofC. elegans adult muscle  4  Myofilament assembly during C. elegans embryogenesis  8  Mutations affecting the body wall muscle in C. elegans  Chapter 2. Materials and Methods  12  18  General maintenance of nematodes and strains  18  Mutant screen  19  Genetic mapping  21  Deficiency mapping oflet-268(ra414)  22  Duplication mapping of spc-1 (ra409) and ra406  iv  24  Complementation tests  24  Double mutant constructions  26  Immunofluorescence staining  29  Phalloidin staining  -.32  Immunofluorescence microscopy  32  Differential interference contrast microscopy Molecular biology techniques  33 34  Gene constructs  35  RNA mediated interference  37  Worm transformation  38  Fusion protein production and purification  38  Antibody production  39  Western analysis  40  Chapter 3. Mutant screen  41  Background  41  Results  44  Mutant screen  44  Wild type embryonic development Phenotypic analysis  45 50  Class 1  54  Class II  57  Class III  64  Class IV  68  v  Class V  74  Discussion  82  Chapter 4. The let-268 locus of Caenorhabditis  elegans  encodes an enzyme essential  for type I V collagen processing.  93  Background  93  Results  98  Genetic analysis of let-268  98  let-268 encodes a lysyl hydroxylase  98  let-268 is expressed in cells that produce type IV collagen Type IVcollagen  secretion is absent in let-268 mutants  108 Ill  Mutations in let-268 affect the stability but not the assembly of the body wall muscle and the underlying basement membrane  116  Pharyngeal myofilaments appear normal in let-268 mutants  123  Discussion  129  let-268 encodes a lysyl hydroxylase in C. elegans essential for type IV collagen processing  129  Distinct functional roles for perlecan and type IV collagen in myofilament assembly and muscle contraction in C. elegans  134  Chapter 5. Analysis of alpha spectrin function during muscle development in Caenorhabditis  138  elegans  Background  138  Results  142  vi  spc-1 (ra409) encodes the C. elegans a spectrin  142  Myofilament organization in spc-l(ra409)  156  animals  Basement membrane and hypodermis organization in spc-1(ra409) mutants  162  Pharyngeal muscle organization in spc-l(ra409)  170  Immunolocalization  mutants  of a spectrin  173  Genetic interactions  183  spc-1(ra409) behaves as a genetic null  183  Genetic interactions between the different spectrin mutants and ankyrin  184  Analysis of the body wall muscle in the fi spectrin mutants Discussion  187 189  spc-1 encodes an a spectrin and the spectrin cytoskeleton localizes to the cell membranes of most tissues  189  Genetic interactions  190  The involvement of the spectrin cytoskeleton in body wall muscle  191  The body wall muscle directs where the basement membrane is established and where the hypodermal attachment structures are assembled  196  Pharyngeal defects in spc-1 (ra409) mutants  197  Future direction  201  Chapter 6. Summary and Conclusion  203  References  210  vii  List of Tables Table 1. Methodology for three factor mapping  23  Table 2. Antibodies used in this study  31  Table 3. Classification of mutants  51  Table 4. Three Factor Mapping  63  Table 5. Deficiency mapping of ra414  73  Table 6. Complementation results for ra414  73  Table 7. Genetic analysis of the let-268 locus  99  Table 8. Deficiency break points determined by PCR amplification of predicted genes. 101  Table 9. Genetic interactions  186  viii  List of Figures Figure 1. A n overview of C. elegans muscle organization Figure 2.  6  C. elegans embryonic muscle development  11  Figure 3. The process of elongation during wild type embryogenesis  47  Figure 4. Muscle development in the wild type embryo  49  Figure 5. Subcellular organization of myosin and perlecan in the wild type embryo  53  Figure 6. Characterization of the Class I mutant, ra401  56  Figure 7. Characterization of the Class II mutants, ra407 and ra408  59  Figure 8. Phalloidin Staining of wild type and ra407 mutant embryos  62  Figure 9. Characterization of the Class III mutant, ra402  66  Figure 10. Characterization of the Class IV mutants, ra406 and  71  ra414  Figure 11. Characterization of the Class V mutant, ra409  76  Figure 12. Genetic map positions of genes identified in this study  81  Figure 13. Genetic map position, genomic structure and GFP fusion protein of  let-268.  105  Figure 14. Amino acid sequence alignment of LET-268 with the three human lysyl hydroxylase isoforms  107  Figure 15. Localization of let-268::GFP expression  110  Figure 16. Type IV collagen localization in wild type, let-268 and emb-9 mutant embryos  115  Figure 17. Immunofluorescence microscopy of the body wall muscle and the underlying basement membrane of wild type, let-268(ra414) and emb-9(hc70) mutant embryos prior to body wall muscle contraction (-400 mpf)  ix  119  Figure 18. Immunofluorescence microscopy of the body wall muscle and the underlying basement membrane and the pharyngeal basement membrane of wild type, let268(ra414), let-268(mnl89) and emb-9(hc70) mutant embryos after body wall muscle contraction has initiated (-470-500 mpf)  121  Figure 19. Localization of Type IV collagen in wild type and unc-52  (null) mutant  embryos  125  Figure 20. The musculature of the pharynx in wild type and let-268  (ra414) mutant  embryos examined by staining with FITC conjugated phalloidin  128  Figure 21. The genetic and physical map of the left arm of the X chromosome  144  Figure 22. Morphological comparison of wild type, sma-1 mutant and spc-1 mutant animals  146  Figure 23. Gene and protein structure of a spectrin Figure 24. Deduced amino acid sequence of C. elegans  149 a  spectrin aligned to Drosophila  a spectrin and human nonerythroid a spectrin  151  Figure 25. SPC-1 is orthologous to Drosophila a spectrin and Human nonerthyroid a spectrin  Figure 26. Myosin filament organization in wild type, spc-1 (ra409) and embryos  155 sma-l(rul8)  159  Figure 27. Immunofluorescence and FITC-phalloidin staining of wild type and spc-1 (ra409) mutant animals  Figure 28. Body wall muscle cells are highly polarized  161 164  Figure 29. The myofilaments are not properly localized to the basal face of the body wall muscle cells in spc-1 mutants  166  Figure 30. Analysis of perlecan and the hypodermal hemidesmosomes in wild type and spc-l(ra409)  mutant embryos  169  Figure 31. The actin cytoskeleton of the hypodermis is normal in spc-1  (ra409) mutant  embryos  172  Figure 32. Thin filament structure of the pharynx in spc-l(ra409) mutant embryos... 175 Figure 33. Western analysis of an a spectrin fusion protein and worm extracts with AS 1. 178  Figure 34. The spectrin cytoskeleton localizes to the cell membrane of most cells during embryogenesis  180  Figure 35. The spectrin cytoskeleton localizes to the I bands in adult body wall muscle. 182  Figure 36. Schematic illustration of the interaction of the body wall muscle and the overlying hypodermis  199  xi  List of Abbreviations bp Ca  basepairs 2+  calcium  D A B CO  1,4-Diazabicyclo[2.2.2]octane  Df  deficiency  dH 0  Sterile distilled water  DIC  differential interference contrast microscopy  dpy  dumpy  DTT  1,4-Dithio-DL-threitol  ECM  extracellular matrix  emb  embryogenesis abnormal  ER  endoplasmic reticulum  2  EMS  ethylmethanesulfonate  FITC  fluorescein isothiocyanate  GLR  glial like cells  IPTG  isopropyl p-D-thiogalactopyranoside  kb  kilobase  kDa  kilodalton  let  lethal  Ion  long  M9  Minimal Salt Buffer  mAb  monoclonal antibody  ml  milliliter  mM  millimolar  mpf  minutes post fertilization  mu  map unit  xii  mua  muscle-attachment defective  map  muscle-positioning defective  Na  sodium  +  NGM  nematode growth media  ORF  open reading frame  PAGE  polyacrylamide gel electrophoresis  pat  paralyzed, arrested elongation at two-fold  PBS  phosphate buffered saline  PCR  polymerase chain reaction  PMSF  phenylmethylsulfonyl fluoride  RNAi  R N A interference  SDS  sodium dodecyl sulfate  spc  a spectrin  sma  small  TBS  tris buffered saline  TBS-T  tris buffered saline with 0.1 % Tween 20  ul  microliter  Ug  microgram  unc  uncoordinated  vab  variable abnormal morphology  xiii  Acknowledgment I would like to thank my supervisor Don Moerman for his guidance and support throughout my studies and for providing a wonderful environment in which to conduct research. Additionally, I would like to thank the past and present members of the Moerman Lab, Jason Bush, Shaun Cordes, Danelle Devenport, Greg Mullen, Poupak Rahmani and Teresa Rogalski, and my supervisory committee: Dr. Bruce Crawford, Dr. David Holm, Dr. Linda Matsuuchi and Dr. Terrance Snutch for their advice and encouragement. Finally, I send many thanks to Jennifer Bonner for her editing talents and for being incredibly supportive during my studies.  xiv  Dedication This thesis is dedicated with love and thanks to the memory of my mother, Dorothy Mae Norman.  xv  Chapter 1. Introduction  A major question in biology concerns how precursor cells of various tissue types become organized and arranged into highly specialized functioning units. Cells interact with each other during gastrulation and morphogenesis to give rise to these specialized tissues. Muscle is an example of a highly differentiated tissue with a complex and stable arrangement of myofilaments that are specialized to undergo rapid contraction and relaxation. Although muscle has been intensely studied for many years, the mechanisms underlying how the myofilament lattice assembles, is maintained, and grows are not well understood. In vitro studies have demonstrated that actin and myosin, the major cytoskeletal elements involved in muscle contraction, are each capable of self-assembly (reviewed in Obinata, 1993). Regardless, a structured myofilament lattice is not formed, which indicates that other components are necessary to form a highly ordered sarcomere array in muscle.  Studies in vertebrates and invertebrates are beginning to identify proteins involved in myofilament lattice assembly. Work by Holtzer's group on cultured cardiac and skeletal muscle cells has demonstrated the importance of attachment plaque-like structures in the assembly of the myofilament lattice. By treating cells with ethyl methanesulfonate (EMS), which inhibits differentiation, and allowing the cells to recover, Antin et al. (1986) discovered the gradual accumulation of stress fiber-like structures at the membrane and the subsequent organization of the myofilaments. In a later experiment, Holtzer and colleagues examined the role of adhesion plaque molecules in the earliest  1  stages of myofibril assembly by using a battery of antibodies. During the assembly of the myofibers, they found a strict colocalization of membrane and muscle proteins resembling the formation of stress fibers in fibroblasts (Lu et al., 1992). These results suggest that the interactions of the muscle cytoskeleton and the membrane are important for the initiation of myofibril assembly.  While the majority of vertebrate studies have focused on perturbation of cells in culture with drugs and antibody blocking, invertebrate studies have focused on genetic dissection. Research on Drosophila melanogaster mutants defective for a or p integrin have implicated these molecules as essential components in sarcomere assembly (Volk et al., 1990; Bloor et al., 1998; Prokop et al., 1998). Additionally, mutations in the a actinin locus in Drosophila have shown that a actinin is required for maintaining the integrity of the Z-discs in the adult flight muscles (Roulier et al., 1992). In the nematode Caenorhabditis elegans, several genes have been identified that are essential for the formation of the myofilament lattice. These genes encode proteins that are present in the basement membrane, such as perlecan (Rogalski et al., 1993), the sarcolemma, such as the a and (3 subunits of integrin (Williams and Waterston, 1994; Gettner et al., 1995), and the cytoplasm, such as myosin heavy chain A and vinculin (Waterston, 1989; Barstead and Waterston, 1991). Taken together, the studies in flies and worms and the experiments in cultured cells point to the importance of the interaction between the cytoskeleton, the membrane and the extracellular matrix (ECM) in myofilament assembly.  2  Drosophila  and C. elegans, owing to their ease of manipulation and the conservation of  biological mechanisms throughout evolution, are used as model organisms for investigating biological phenomena. The biological conservation found in these organisms is highlighted by the recent sequencing of the Drosophila  and the C. elegans  genome (C. elegans Sequencing Consortium, 1998; Adams et al., 2000; for a review see Rubin et al., 2000). For instance, many genes involved with human disease are found in both C. elegans and Drosophila (Rubin et al., 2000). We have chosen the small soil nematode C. elegans to investigate muscle development. C. elegans has proven to be an excellent organism for genetic and molecular studies of many biological processes including muscle development and function for the following reasons. (1) Genetic manipulation is very straight forward due to the ease of isolating mutations, the extensive genetic map, hermaphroditic lifestyle and the use of males that arise spontaneously by X chromosome nondisjunction (Wood, 1988). (2) The complete cell lineage has been determined allowing the identification of all cells throughout development (Sulston et al., 1983). (3) The ultrastructure of the entire adult has been characterized (White et al., 1986). (4) The cloning of genes is greatly simplified since -98% of the genome has been sequenced (C. elegans Sequencing Consortium, 1998; Coulson et al., 1991). Along with more than 80 muscle affecting mutants (see http://www.zoology.ubc.ca/labs/moerman/muscle_affecting_genes.html), about 30 monoclonal antibodies (mAbs) exist that recognize muscle and muscle-associated molecules (Miller and Shakes, 1995; Moerman and Fire, 1997). These mAbs can be used to characterize muscle of wild type and mutants at various developmental stages. These reagents together with the wealth of information accumulating about C. elegans make it  3  one of the best model organisms for studying the interaction of the muscle cytoskeleton with the membrane and the E C M . Furthermore, studies of the body wall muscle of C. elegans have provided the larger muscle research community with several "firsts" in muscle biology (for a review see Moerman and Fire, 1997). For example, the first D N A sequence of a myosin heavy chain and genetic analysis of this locus was done in C. elegans (Epstein and Thomson, 1974; Karn et al., 1984). Additionally, the identification of the myosin associated molecule, twitchin, was the first muscle protein to be discovered through genetic analysis (Brenner, 1974; Moerman et al., 1986; Moerman et al., 1988; Benian et al., 1989).  Background Summary of C. elegans adult muscle C. elegans undergoes several developmental events before becoming a mature adult. These events are composed of embryogenesis and four larval stages (L1-L4). Embryogenesis is devoted to cell proliferation, organogenesis and morphogenesis (see below). The larval stages are devoted to growth and maturation of the animal. During the larval stages additional muscle is added, resulting in a total of 95 body wall muscle cells, and the muscle cells grow in size and mass. The 95 body wall muscles of the adult reside in four quadrants, two dorsal and two ventral that run parallel with the longitudinal axis of the animal (Fig. 1). Each quadrant contains 24 mononucleate muscle cells, except the left ventral quadrant, which contains only 23 (Sulston and Horvitz, 1977). Adjacent to the body wall muscle cells is a thin basement membrane that separates the muscle cells from an epidermal layer, known as the hypodermis, which secretes the cuticle. The  4  Figure 1. A n overview of C. elegans muscle organization. (A) The body wall muscle run the length of the animal. (B) In cross section, the body wall muscle is organized into quadrants. (C) A polarized light micrograph from a live animal illustrating the high ordered structure of the myofilaments within the body wall muscle. The arrowhead indicates a dense body and the arrow indicates the M-line (H-zone). (D) A single body wall muscle cell is schematically cut away to indicate the cellular organization. (E) A single sarcomere (contractile unit of muscle) and its structural components. IF indicates the intermediate filaments of the hemidesmosomal-like structures in the hypodermis (adapted from Hobert, et al., 1999)  5  OUTSIDE CELL  6  exoskeleton of the worm is composed of a collagen rich cuticle (Kramer, 1994; 1997). Within the muscle and lying directly under the basement membrane, the myofilament lattice forms obliquely striated arrays that are at a six-degree angle to the longitudinal axis of the worm (Fig. 1). This lattice is structurally analogous and homologous to the vertebrate sarcomere (Waterston, 1988; Moerman and Fire, 1997). In C. elegans, the thin filaments attach to the Z-line-like dense body and the thick filaments attach to an M-line analog. The dense body and the M-line are attached to the membrane (sarcolemma) and extend the depth of the lattice anchoring all filaments to the membrane. Several of the major proteins that constitute the dense body have been identified (Francis and Waterston, 1985; Barstead and Waterston, 1991; Barstead et al., 1991; Williams and Waterston, 1994; Moulder et al., 1996; Hobert et al., 1999; B . Williams, pers. comm.). Localized at the membrane, a and (3 integrin form the base of the dense body. Further into the muscle cytoplasm, vinculin, talin, PINCH (UNC-97), I L K (integrin linked kinase) and a actinin are found. These observations suggest that an attachment plaquelike system anchors the myofilament lattice to the membrane and the extracellular matrix (Fig. 1). Many of these molecules are found in vertebrate muscle where they share an orthologous function (reviewed in Berthier and Blaineau, 1997).  In order for the force of contraction to result in movement, a mechanism must exist to anchor the body wall muscle cells to the cuticle. Electron micrographs of C. elegans reveal densely staining material in the basement membrane that is concentrated under the dense bodies and M-lines (Waterston, 1988). Additionally, filamentous structures (hemidesmosomes) that extend processes into the cuticle are observed in the hypodermal  7  cells (Francis and Waterston 1985; Francis and Waterston, 1991). These hemidesmosome structures in the hypodermis are only found adjacent to the muscle cells. These observations suggest that the attachment of the body wall muscle cells to the cuticle may occur via the following mechanism: the dense bodies and the M-lines of the muscle cells interact with a specialized basement membrane, and this specialized basement membrane interacts with the hemidesmosome structures of the hypodermis, connecting the muscle to the cuticle (Fig. 1). The interconnectedness of the contractile apparatus with the cuticle, through interactions with the sarcolemma, the basement membrane, and the hypodermis, can be likened to tendons found in vertebrate muscle.  Myofilament assembly during C. elegans embryogenesis One benefit of studying C. elegans is that the development of the embryo has been well characterized and documented. The first stage of embryogenesis is composed of gastrulation and cell proliferation, and the second stage is devoted to morphogenesis and organogenesis. The entire process, from fertilization to hatching, takes -800 minutes (Sulston et al., 1983). Gastrulation and cell proliferation utilize only a small proportion of this time while the formation of the cylindrical worm-like body and its organs from a ball of cells occupies the majority of embryogenesis.  The molecular and cellular interactions that direct the development of the body wall muscle are the focus of this thesis. Muscle development in C. elegans occurs in several stages. Between 240-290 minutes after fertilization, structural proteins of muscle begin to accumulate in the presumptive myoblasts and hemidesmosomal components are  8  randomly located in the hypodermal cells (Fig. 2B-1). Between 290-350 minutes after fertilization, the muscle cells start migrating from their lateral position and organize into the two dorsal and two ventral body wall muscle quadrants. Additionally, the hemidesmosomal components start to organize in the region of the hypodermis associated with the muscle quadrants. After the quadrants have formed, at -350 minutes after fertilization, the structural components of the muscle localize to the membrane adjacent to the hypodermis (Fig. 2B-2). Therefore, the muscle cells become polarized with the structural proteins located in the basal half of the cell. Concurrently, the basement membrane components organize under the muscle cells. Approximately 420 minutes after fertilization, the muscle cells flatten against the hypodermis as elongation begins (Fig. 2B-3). The organization of the sarcomeres in the muscle cells and the attachment structures in the hypodermis occurs between 420-450 minutes after fertilization (Fig. 2B4). The first movement of the embryo also occurs at this time. After 450 minutes post fertilization, the sarcomere structure is fully functioning even though the organization of actin is not complete until -550 minutes (Hresko et al., 1994).  Normal differentiation of the body wall muscle requires the formation of specific interactions with the E C M and the hypodermis and also molecular interactions within the muscle cell itself. Presumptive myoblasts will accumulate muscle components but without the formation of the attachment structures in the muscle and the hypodermal cells, myofilament assembly is blocked. Gossett et al. (1982) examined muscle cell differentiation in C. elegans mutants defective for cell and nuclear division. They found that muscle specific proteins accumulate but do not undergo any organization. The  9  Figure 2. C. elegans embryonic muscle development. (A) Time scale of embryonic elongation and muscle function. Embryonic elongation initiates at -290 minutes post fertilization (mpf) and continues until the embryo reaches the 4 fold stage (800 mpf). Body wall muscle contraction is first observed just prior to the two fold stage (-400 mpf). Pharyngeal contraction starts at -760 mpf (adapted from Williams and Waterston, 1994). (B) Cartoon illustrations of cross sections through C. elegans embryos at different developmental times. (1) is depicting a 290 minute embryo where the muscle cells (circles) are located in their lateral position and the structural proteins (dots) have accumulated within the muscle cells. The arrows indicate the direction the muscle cells will migrate to form the four muscle quadrants. (2) is illustrating the muscle cells of a 350 minute embryo after they have migrated to their respective quadrants. Also, the muscle and hypodermal structural proteins have started to colocalize around the basement membrane. The diagrams in (3 and 4) depict only one dorsal muscle quadrant and the surrounding area. (3) is illustrating an embryo at 420 mpf in which the muscle cells have flattened against the hypodermis and that the structural proteins in the muscle cells, the basement membrane and the hypodermis are coextensive. (4) demonstrates that the structural components are well organized into functional sarcomeres in the 450 minute embryo (Adapted from Hresko et al., 1994).  10  B  (1) 290 min  (3) 420 min  Dorsal  Dorsal  hypodermis  Ventral  (2) 350 min  (4) 450 min  ll  mutants used in this study generate approximately 200 cells whereas wild type embryos generate over 500 cells and the mutant embryos eventually arrest development with gross abnormalities. Since fewer cells are generated, these mutants lack a completely developed basement membrane and hypodermis. This further supports the notion that the interaction of the muscle cytoskeleton with the sarcolemma, a specialized basement membrane and the hypodermal cells is crucial for the normal assembly of the myofilament lattice. Moreover, the spatial and temporal concurrence of the muscle cell structures and the hypodermal cell structures suggests that important developmental signals may be passed between the two cell types. Two experiments have been conducted where certain body wall muscle cells have been ablated by laser microsurgery and the organization of the basement membrane and the hypodermal attachment structures were examined (Moerman et al., 1996; Hresko et al., 1999). In both cases, neither the basement membrane nor the hypodermal attachment structures organized in the area adjacent to the missing muscle cells. This provides further evidence that cross talk must occur between these two tissue types. To date no molecules involved in this cross talk have been identified.  Mutations affecting the body wall muscle in C. elegans  Approximately eighty genes have been identified by mutational analysis that affect body wall muscle structure and function in C. elegans (reviewed in Moerman and Fire, 1997; see http://www.zoology.ubc.ca/labs/moerman/muscle_affecting_genes.html).  These  mutant phenotypes range from embryonic lethality to mild uncoordinated movement and can be placed into four phenotypic categories. The first and largest phenotypic category  12  is the uncoordinated movement phenotype, known as Unc (for uncoordinated: Brenner, 1974). This phenotypic category ranges from mild uncoordinated movement to complete paralysis. The second category is composed of mutants with defects in muscle cell attachment (Mua; Plenefisch et al., 2000). In this category, the attachments between neighboring muscle cells and the hypodermis are defective and the muscle cells separate from each other and the hypodermis during the larval stages. The third category is composed of mutants with muscle cell position abnormalities (Mup; Hedgecock et al., 1987; Goh and Bogaert, 1991; Myers et al., 1996; Gatewood and Bucher, 1997). It is thought that mutants in this category never position their muscle cells properly into the four muscle quadrants. In many cases the mutants that fall into categories II and III die during larval development and therefore would have not been identified in Unc screens. The final category is composed of mutants that affect muscle development during embryogenesis. This category is known as the Pat mutants (paralyzed, arrested elongation at two-fold; Williams and Waterston, 1994). During the two fold stage of embryogenesis, the wild type embryo has organized sarcomeres and is undergoing body wall muscle contraction. In contrast, the Pat mutants fail to display any muscle contraction and arrest at this stage of embryogenesis.  In addition to these phenotypic differences, muscle mutations have distinct cellular consequences. Morphological studies indicate that muscle-affecting mutations can influence the organization of the thin filaments or the organization of the thick filaments or both the thin and thick filaments. For example mutations in deb-1, which encodes vinculin, displays defects in thin filament organization; however, the thick filaments are  13  normally organized (Barstead and Waterston, 1991; Williams and Waterston, 1994). On the other hand, mutations in the unc-45 locus, which encodes a protein related to the CR01/She4p proteins that have been implicated in the assembly of cytoplasmic myosin (Venolia et a l , 1999), result in abnormal organization of the thick filaments but the thin filaments appear normal (Epstein and Thomson, 1974; Barral et al; 1998). Lastly, mutations in the basement membrane molecule perlecan result in the disorganization of both thick and thin filaments (Rogalski et al., 1993; Williams and Waterston, 1994). A list of additional examples of these distinct cellular consequences caused by mutations is listed on the following web site: http://www.zoology.ubc.ca/labs/moerman/muscle_affecting_genes.html.  These examples  demonstrate that the mechanisms of muscle development can be carefully dissected using the model organism C. elegans.  In this study we hypothesized that new genes involved in muscle development could be isolated by conducting a mutant screen based on the Pat screen carried out by Williams and Waterston (1994). Instead of screening for animals that are paralyzed at the two fold stage, we screened for late embryonic (300-800 minutes after fertilization) to early L l arrested animals and analyzed these animals by immunofluorescence with antibodies to myosin and perlecan. By analyzing mutants from this large window of time, we hypothesized that we could identify a wide range of muscle-affecting mutants. This includes muscle cell positioning defects, subtle to severe myofilament assembly defects, stability and attachment defects of the myofilament lattice to the muscle cell membrane and the attachment of the body wall muscle cells to the underlying basement membrane  14  or hypodermis. In a small scale screen using this approach we have isolated eight mutants that display defects in muscle development (see Chapter 3). These eight mutants, representing eight complementation groups, have been placed into five phenotypic classes. Class I is the most severe class isolated and is represented by a single mutant (ra401). This mutant fails to assemble myofilaments and never displays any body wall muscle contraction. The Class Tl mutants (ra407 and ra408) are less severe than Class I. These mutant strains initiate myofilament assembly but the myofilaments never become completely organized. The Class III mutants (ra402 and ra404) also have defects in their ability to organize myofilaments although not as severe as in the Class II mutants. Additionally, the Class III mutants have defects in their ability to anchor the body wall muscle to the hypodermis. The Class IV mutants (ra406 and ra414), unlike the Class I—III mutants, are able to assemble muscle normally. However, once muscle contraction commences the body wall muscle separates from the hypodermis and large gaps arise in the muscle quadrants. The single class V mutant (ra409) has abnormally wider muscle quadrants than normal suggesting a cell morphological or growth defect. Two of the loci identified in this screen are new alleles of previously identified genes (unc-52(ra401) and let-268(ra414)).  The other six loci appear to be newly identified  genes (see Chapter 3 for further details).  In addition, in Chapters 4 and 5, two of the mutants identified in this screen have been characterized in greater detail and the affected genes have been cloned. In Chapter 4, the gene identification and characterization of one of the class IV mutants (ra414) is presented. This mutation is an allele of a previously identified locus called let-268. let-  15  268 encodes a collagen processing enzyme, procollagen lysyl hydroxylase, required for post-translational modifications. In let-268 mutants type IV collagen is expressed; however, it is retained within the type IV collagen producing cells. This observation indicates that LET-268 is required for the normal processing and secretion of type IV collagen. The examination of the body wall muscle in let-268 mutant animals reveals normal myofilament assembly prior to contraction. However, once body wall muscle contraction commences the muscle cells separate from the underlying epidermal layer (the hypodermis) and the myofilaments become disorganized. These observations indicate that type IV collagen is required in the basement membrane for mechanical support and not for organogenesis of the body wall muscle (see Chapter 4 for further details).  In Chapter 5, the cloning and characterization of the Class V mutant (ra409) is presented. ra409 is a mutation in the gene encoding the only a spectrin gene (spc-1) in the C. elegans genome (C. elegans Sequencing Consortium, 1998). We have localized the a spectrin protein to the body wall muscle cell membrane, the dense bodies (Z-line analog) and the I bands in the body wall muscle of C. elegans. Animals lacking functional a spectrin have defects in myofilament organization and die just after hatching. More specifically, when compared to myofilaments in wild type animals, the myofilaments in the mutant animals are abnormally oriented relative to the longitudinal axis of the embryo. Additionally, defects are observed in the musculature of the pharynx. This muscle tissue is abnormally shaped, has a disorganized arrangement of myofilaments and labors during contraction. From this work, we conclude that  16  a spectrin is an important component of muscle and that it is involved in myofilament organization (see Chapter 5 for further details).  17  Chapter 2. Materials and Methods  General maintenance of nematodes and strains Nematode maintenance and culture was carried out as described by Brenner (1974). In short, C. elegans are grown on N G M (Nematode Growth Media) agar plates with a lawn of OP50, a leaky uracil requiring strain of E. coli. The N2 Bristol strain was used as the control wild type strain. Unless otherwise indicated, all genetic experiments were carried out at 20°C. This study followed standard nomenclature as described by Horvitz et al. (1979). Strains for this study were provided by V . Vijayaratnam, C. Thacker and A . Rose (University of British Columbia), H. Kagawa (Japan), D.L. Baillie (Simon Fraser University), M . Costa (Exelixis Pharmaceuticals), R. Herman (University of Minnesota), J. Austin (University of Chicago) and the Caenorhabditis Genetics Center (CGC). The following mutant alleles were used in this study: Chromosome I: dpy-5(e61), lin-1 l(n566), pat-1 l(st541), unc-14(e57), unc-40(e271), unc54(el90), unc-75(e950), unc-87(el216), Chromosome II: dpy-10(el28), 268(mnl89),  let-268(mnl98),  unc-94(sul77),  vab-10(e698).  let-242(mn90), let-244(mn97), let-245(mnl85), unc-4(el20),  unc-52(e444), eDf21, mnDfl2,  mnDf29, mnDf30, mnDf57, mnDf58, mnDf60, mnDf62, mnDf68, mnDf71,  let-  mnDf28, mnDf83,  mnDfl05. Chromosome III: dpy-17(el74), pat-3(st564), unc-32(el89),  dpy-18(e364), emb-9(hc-70ts), lon-l(el85),  unc-32(el89),  unc-36(e251).  Chromosome IV: unc-22(e66), unc-44(e362), unc-44(ell97),  18  unc-44(el260).  pat-2(st538),  Chromosome V : dpy-ll(e224),  sma-l(e30), sma-l(rul8),  unc-70(n493), unc-70(n493nll71), 70(sll5),  unc-70(sl502),  unc-70(n493nll72),  unc-23(e25),  unc-70(e524),  unc-70(n493nll73),  unc-  unc-70(sl639).  Chromosome X : dpy-3(e27), dpy-7(sc27ts), let-2(e!53), let-2(g25ts), lon-2(e678), unc2(e55), unc-3(e!51),  unc-20(ell2ts),  unc-78(e!217),  mnDplO, mnDp31, mnDp33,  stDp2,  syDfl.  Mutant screen Two mutant screens were carried out in this study. In both cases L4 larvae were exposed to 25 m M ethylmethanesulfonate (EMS) in M9 buffer for four hours and then allowed to recover on N G M agar plates with freshly streaked OP50 (Sulston and Hodgkin, 1988). The first screen involved mutagenizing N2 hermaphrodites and transferring approximately 600 FI progeny to individual plates. The FI animals were allowed to lay eggs for 24 hours and then picked off the plates. The F2 offspring were scored for the presence of arrested embryos or L I larvae in -1/4 of the population. These animals were subjected to further analysis (see below). The second screen involved mutagenizing a population of animals with the genotype unc-4(el20)l mnCl [dpy-10(el28) unc52(e444)]. mnCl is a cross-over suppressor that is used as a genetic balancer for the central region of chromosome Tl (Herman, 1978). In this case -1000 FI animals were plated to individual plates and the F2 were examined for the absence of Unc-4 animals. This would indicate that a lethal mutation linked to unc-4 has been isolated. Embryonic or L I lethal mutants identified in this screen were subjected to further analysis. To determine if these lethal mutants had defects in body wall muscle development, they were  19  subjected to immunofluorescence analysis with antibodies that recognize a body wall specific myosin (MHC A) and the basement membrane protein perlecan (GM1).  A l l mutant strains were all outcrossed three times before further analysis. Outcrossing was carried out by crossing N2 males with heterozygous lethal hermaphrodites (m/+). Single male (genotype m/+ or +/+) offspring were crossed into N2 hermaphrodites and brooded. Single male progeny (genotype m/+ or +/+) was then crossed back into the N2 background for one or two more rounds and hermaphrodites that segregated the lethal mutation were retained for genetic mapping.  For X-linked lethal mutations (X-linked mutations were identified by the failure of these mutations to be transmitted through males, see below), dpy-5(e61)/+ males were crossed with hermaphrodites that were heterozygous for the X linked mutation (m/+). 10-15 FI progeny were selected and placed on individual plates. In the F2, Dpy animals were selected from plates that were segregating the X linked mutant phenotype and their progeny were examined for the presence of the X linked mutation. These animals had the genotype dpy-5(e61)/ dpy-5(e61); m/+ and were used for further outcrossing. This involved mating wild type males with dpy-5(e61)/ dpy-5(e61); m/+ hermaphrodites and selecting phenotypically wild type progeny of the genotypes dpy-5(e61)/+; m/+ or dpy5(e61)/+; +/+. Animals that segregated the lethal phenotype were outcrossed one more time following the same procedure. These animals were further outcrossed once they were genetically balanced (see below).  20  Genetic mapping A l l mutations isolated in the two screens were recessive. The genomic location of these mutations was determined by mapping the lethal mutations with respect to the following recessive visible genetic markers: dpy-5(e61) I, unc-4(e!20) II, dpy-18(e364) III, unc36(e251) III, unc-22(e66) IV, unc-23(e25) V , dpy-ll(e224)  V . Wild type males were  crossed to single m/+ (where m is the lethal mutation) hermaphrodites and offspring males that arose were selected. Single males from these plates, that had genotypes of either +/+ or m/+, were crossed to strains homozygous for the visible markers (v/v). 1015 phenotypically wild type animals were selected from the offspring of these crosses that would have the genotype +/+; v/+ or m/+; v/+ and placed on individual plates. On plates that segregated the lethal mutation, 20-30 animals homozygous for the visible marker were then selected and placed on individual plates (genotype +/+, v/v or m/+, v/v). The offspring of these animals were then scored for the presence of the embryonic lethal phenotype. The presence of the mutant embryo phenotype in -50% of the populations would indicate that the two loci are not linked. However, the embryonic phenotype appearing in less than 50% of the tested individuals indicates that the two loci are linked to the same chromosome. This linkage was then be tested by two and three factor mapping (see below). Mutants that mapped to the X chromosome were identified by males that failed to carry the mutation since males only possess one X chromosome (m/0 = lethal male).  Where linkage was established, recombinant animals were selected (genotype vm/v+) and cis two factor mapping was carried out. These animals were crossed with wild type  21  males and several wild type progeny (genotype vm/++ or v+/++) were transferred individually to single plates and allowed to self fertilize. Plates that segregated lethal mutants were scored for the presence of the viable mutation and the recombination frequency was calculated by the following formula (Rogalski et al., 1982): p = 1- [l-(3 V / V + W)]  1/2  where p is the recombination frequency, V is the number of visible marker  progeny and W is the number phenotypically wild type progeny.  For three factor mapping the lethal mutation was placed in trans to a double viable mutant chromosome (genotype +m+/vl+v2).  Each viable recombinant progeny was  transferred to individual plates and allowed to self fertilize. Their offspring were scored for the presence of the lethal mutation (m). This mapping scheme is indicated in Table 1 and the data are listed in Table 4. Table 1 (A) indicates the results to expect if the lethal mutation was located outside of the visible markers and (B) shows the results that would arise if the lethal mutation was located in between the two visible markers.  Deficiency mapping of  let-268(ra414)  let-268(ra414) was genetically mapped by using a series of deficiencies.  let-268(ra4J4)  was balanced by mnCJ, a region covered by a large number of deficiencies (Sigurdson et al., 1984). Strains carrying the let-268(ra414) mutation isolated had the following genotype: unc-4(el20)  let-268(ra414)/mnCl[dpy-10(el28)  unc-52(e444)].  N2 males  were crossed into this strain and single males of the genotype unc-4(el20) let268(ra414)/++ or mnCl[dpy-10(el28)  unc-52(e444)]/++ males were crossed into several  deficiency (Df) bearing strains (see strains listed above and Table 5). The Df strains were genotypically unc-4(el20)  DfX/mnCl[dpy-10(el28)  22  unc-52(e444)].  Therefore,  Table 1. Methodology for three factor mapping. Heterozygote  Recombinant  Predominant  Phenotypes of  phenotype selected  genotype of  progeny  recombinants A  VI  + vl +/+ vl v2  V I , V1V2  V2  m + v2/+ vl v2  V2, V1V2, lethal  VI  vl + +/vl + v2  V I , V1V2  vl m +/vl + v2  V I , V1V2, lethal  + + v2/vl + v2  V2, V1V2  + m v2/vl + v2  V2, V1V2, lethal  + vl v2/m + +  B  vl + v2/+ m +  V2  m is a lethal mutation vl is a visible mutation v2 is another visible mutation  23  complementation was detected when Unc-4 animals arose in the progeny of this cross. Failure to complement was detected when no Unc-4 progeny arose from the cross. Plates that segregated DpyUnc males were not scored.  Duplication mapping of spc-l(ra409) and ra406 Animals with the genotype spc-1 (ra409) dpy-7(sc27ts)/+dpy-7(sc27ts) or dpy-7(sc27ts) ra4067dpy-7(sc27ts) + were mated with duplication bearing males (mnDplO,  mnDp31,  mnDp33 and stDp2). Outcrossed worms were transferred to individual plates and their progeny were scored for the presence of the lethal mutation and viable Dpy-7 animals. The presence of Dpy-7 animals in greater proportions than recombinants indicated the rescue of the linked lethal.  Complementation tests Complementation tests were conducted for several of the lethal mutations isolated in this thesis. The complementation test conducted between unc-52(ra401) and unc-52(e444) was carried out by crossing unc-52(ra401)/+ males with unc-52(e444) homozygous hermaphrodites. Their offspring were scored for the presence of Unc-52 males. Heteroallelic animals [unc-52(ra401)7unc-52(e444)] were not fertile.  A complementation test was carried out between ra407 and unc-94(sul77).  Both of these  mutations map near dpy-5 and both disrupt the myofilament lattice. ra407 heterozygous males were crossed to unc-94(sul77) homozygous hermaphrodites and their offspring were scored for the ra407 mutant phenotype or Unc males.  24  A complementation test was carried out between ra.408 and pat-2(st538). Both mutants are tightly linked and map to the left of unc-32. t o test complementation, ra408 unc32(e251)/++ males were mated with pat-2(st538) unc-32(el89);  stExlO[rol-6(sul006)]  (stExlO carries D N A that rescues the Pat phenotype and a dominant roller mutation). Self progeny of hermaphrodites with the genotype pat-2(st538) stExlO[rol-6(sul006)]  unc-32(el89);  produce animals that are Unc Rol (animals with the transgene)  and Pat (animals that have lost the transgene). Complementation of pat-2(st538) and ra408 was determined by the presence of Unc non Rol outcrossed offspring.  A complementation test was carried out between ra402 and vab-10(e698). Both of these mutations map to the left arm of the chromosome I near unc-75. ra402 heterozygous males were crossed to vab-10(e698) homozygous hermaphrodites and their offspring were scored for the ra407mutant phenotype or Vab males.  Complementation tests for let-268(ra414) were carried out as previously described (Sigurdson et al., 1984). In short, unc-4(el20) let-268(ra414)/mnCl[dpy-10(el28)  unc-  52(e444)] animals were mated with N2 males. Single males from this cross were mated with hermaphrodites carrying lethal mutations in the let-268 genetic region. The genotypes of the lethal mutations tested are unc-4(e!20) let-X/mnCl[dpy-10(el28)  unc-  52(e444)]. Therefore, complementation was detected by the presence of Unc-4 animals and non complementation was detected by the lack of Unc-4 animals in the progeny of the cross. Plates that segregated DpyUnc males were not scored.  25  The complementation between spc-l(ra409) and unc-78(el217) was carried out by mating +/+; mnDp33/+;spc-l(ra409)/0  78(el217)/unc-78(el217)  males to dpy-5(e61)7 dpy-5(e61); +/+; unc-  hermaphrodites. Their offspring were scored for the presence  of the spc-1 lethal phenotype or Unc non Dpy animals.  Additionally, spc-l(ra409) animals were tested to see if they fall into the deficiency syDfl.  This was carried out by mating mnDp33/+; spc-1 (ra409)/0 males into unc-2(e55)  lon-2(e678)/syDfl  hermaphrodites. Their progeny were scored for the presence of spc-  l(ra409) animals.  The T e l induced allele of spc-1 was tested for noncomplementation with This was carried out by crossing mnD 33/+; P  mnDp33/+; spc-l(ra417::Tcl)  spc-l(ra409)unc-20(el  unc-20(ell2ts)/spc-l(ra417::Tcl)  spc-l(ra409).  12ts)/0 males with unc-20(ell2ts)  hermaphrodites. Their offspring were scored for the spc-1 (ra409) phenotype and for the lack of Unc-20 animals.  Double mutant constructions  Several double-mutant strains were constructed for genetic analyses. The ra408 and unc32(el89) double was constructed by crossing ra408/+ males to dpy-17(el64) unc32(el89) homozygous hermaphrodites. Wild type heterozygous progeny (FI) were selected and placed on individual plates. Unc-32 recombinant progeny (F2) were  26  selected and the individuals that segregated ra408 mutants were used for complementation tests.  spc-l(ra409) 20(ell2ts)  and spc-l(ra417::Tcl)  [spc-1 (mutant)] were placed in cis with unc-  by crossing mnDp33/+; spc-l(mutant)  males to dpy-3(e27)  unc-20(ell2ts)  homozygous hermaphrodites. Wild type heterozygous progeny (FI) were selected and placed on individual plates. Unc-20 recombinant progeny (F2) were selected and the individuals that segregated the spc-1 (mutant) animals were used for complementation tests.  spc-1(ra409) and ra406 were placed in cis with dpy-7(sc27ts) by mating dpy-7(sc27ts)/0 males that were raised at the permissive temperature (15°C) with hermaphrodites carrying the ra409 or ra406 mutation (ra409/+ or ra406/+). Wild type heterozygous FI individuals were selected and placed on separate plates that were maintained at the restrictive temperature (25°C). Approximately 100 Dpy-7 (F2) offspring were transferred to individual plates and recombinant animals that segregated either the ra409 or ra406 mutation were maintained for further genetic analysis.  The dpy-5(e61); unc-78(el217) double mutant was constructed by mating dpy-5(e61)/+ males with unc-78(el217) homozygous hermaphrodites. Wild type heterozygous progeny (FI) were selected and placed on individual plates. DpyUnc animals selected in the F2 were used for subsequent genetic analysis.  27  Several double mutants were constructed with the spc-1 (ra409) mutation. These include sma-l(e30),  70(sl639).  sma-l(rul8),  unc-44(e362), unc-70(e524), unc-70(n493nl 172) and unc-  None of these mutations show genetic linkage with spc-l(ra409).  mutant construction with the viable mutations [sma-l(e30), sma-l(rul8),  Double  unc-44(e362)  and unc-70(e524)] were all constructed following the same method. For example, mnDp33/+; spc-1 (ra409)/0 males were mated with sma-l(e30) homozygous hermaphrodites. Wild type heterozygous offspring (FI) were selected and placed on individual plates. Sma-1 offspring (F2) were selected and animals that segregated spc-1 mutants were retained for further analysis. The double mutants constructed with spcl(ra409) and the lethal alleles of unc-70(lethal) [unc-70(n493nl 172) and  were constructed by mating mnDp33/+; spc-1(ra409)/0 males with +  unc-70(sl639)]  sma-1(ml8)/unc-  70(lethal) + hermaphrodites (see below). Approximately 50 wild type heterozygous offspring (FI) were selected and placed on individual plates. Half of the FI offspring have the genotype sma-1 (ml 8)/+; spc-1 (ra409)/+ and the other half have the genotype unc-70(lethal)/+;  spc-l(ra409)/+  (the genotypes were confirmed by phenotypic analysis  of the following generation). These heterozygous animals were examined for developmental and morphological abnormalities (see Chapter 5).  The sma-l(rul8)  and unc-70(sl639) or unc-70(n493nl 172) double mutants were  constructed by mating unc-70/+ males with sma-l(ru!8) homozygous hermaphrodites. Wild type heterozygous offspring (FI) were selected and placed on individual plates (Plates that contained both Sma-1 and Unc-70 animals were maintained for further analysis). Approximately 100 Sma-1 animals were selected from the F2 population that  28  segregated unc-70 lethal animals (see Chapter 5). Sma-1 recombinants that segregated unc-70 mutants were retained for further analysis [genotype unc-70(lethal) sma1(ml8)7+  sma-l(rul8)].  Several other double mutants were used in this study. These double mutants were obtained either from the Caenorhabditis elegans stock center or from other researchers that kindly provided them.  Immunofluorescence staining Animals were prepared for immunofluorescence by two methods. In the first, gravid adults were collected by washing a mixed populations of nematodes off N G M plates with d H 0 . Collected animals were treated with alkaline hypochlorite (1% NaOCl, 0.5 M 2  KOH) for 5-10 minutes at room temperature to dissolve the adults and then the embryos were washed three times with M9 buffer (Minimal salt buffer; Sulston and Hodgkin, 1988). The washed embryos were then fixed in PBS containing 3.2% formaldehyde for 10 to 15 minutes at room temperature. After fixation the embryos were washed once with d H 0 and stored in 100% methanol at -20°C for 10 minutes to 7 days. The embryos are 2  rehydrated in 75% methanol, 50% methanol, 25% methanol and TBS (Tris buffered saline). The embryos are then incubated in TBS-T (Tris buffered saline with 0.1% Tween 20) plus 0.1% milk powder for 5-10 minutes at room temperature. After this incubation, the primary antibodies are added and incubated overnight at 4°C (see Table 2 for antibodies used and concentrations). The other method involved washing a mixed population of nematodes off N G M plates with d H 0 . These animals were then placed on 2  29  0.1% to 1% polylysine (Sigma Chemical Company) coated microscope slides in 50pJL aliquots and a cover slip was placed on top to distribute the animals over the slide. The slides were placed at -80°C for 10 minutes to 7 days. The cover slips were cracked off using a razor blade and the slides were placed in -20°C 100% methanol for four minutes followed by -20°C 100% acetone for four minutes. The embryos were rehydrated in 75% acetone, 50% acetone, 25% acetone and TBS for a minute each. The rehydrated embryos were then placed in TBS-T and incubated for 5-10 minutes at room temperature. Primary antibodies were added in TBS-T with 0.1% B S A and incubated at 4°C overnight. For both antibody staining procedures, the primary labeled embryos were then washed 4 times in TBS-T for 5-10 minutes and then the secondary antibodies were added. The secondary labeling was carried out at room temperature for 2-3 hours. Following secondary labeling the embryos were washed 4 time in TBS-T for 5-10 minutes, 1 time in TBS for 5 minutes and resuspended in a Glycerol D A B CO (1,4Diazabicyclo[2.2.2]octane)solution (90% glycerol, 2.5% D A B C O in TBS).  Antibodies and concentrations used in this study have been listed in Table 2. The secondary antibodies used were FITC labeled donkey anti-rabbit IgG F(ab')2 and Texas Red labeled donkey anti-mouse IgG F(ab')2 (Jackson ImmunoResearch Laboratories) and were diluted 1:200. For control, each lot of secondary antibodies was tested for cross reactivity with nematodes. This was carried out by preparing animals as described above except the primary antibodies were omitted. In addition, the AS1 preimmune serum was tested for cross reactivity with nematodes. The AS 1 preimmune was used at a concentration of 1:100. Finally, to determine the specificity of AS1 antiserum, AS1 was  30  Table 2. Antibodies used in this study. Antibody  Specificity  Concentration*  Reference  AS1*  a spectrin  1:1000  K . R . N , and D. Moerman, unpublished data  DM5.6  f  body wall muscle myosin heavy  1:50  Miller etal., 1983  1:50  Miller et al., 1983  1:50  Moerman et al., 1996;  chain A DM5.8  f  body wall muscle myosin heavy chain B  GM1*  A l l isoforms of perlecan, major basement membrane component  GM3*  Long isoforms of perlecan,  Mullen etal., 1999 1:1000  Mullen etal., 1999  1:25  Francis and Waterston,  major basement membrane component MH2/3  f  perlecan, major basement membrane component  1991; Rogalski etal., 1993  MH4  Hemidesmosome; intermediate  f  1:100  filament subunit MH5  1991  Hemidesmosome  f  Francis and Waterston,  1:100  Francis and Waterston, 1991  MH27  f  Epithelial cell boundaries  1:100  Francis and Waterston, 1991  NW#68  Type IV collagen a2  1:50  Graham etal., 1997  NW#155  Type IV collagen a l  1:50  Graham et al., 1997  * rabbit polyclonal f  mouse monoclonal  ¥ Dilution of antibody used for immunofluorescence microscopy  31  incubated with 10 (Xg/ml of fusion protein produced from DM#154 (see below) and was used for indirect immunofluorescence.  Phalloidin staining Embryos for phalloidin staining were collected by washing gravid adults off N G M plates with d H 0 . These animals were treated with an alkaline hypochlorite (1% NaOCl, 0.5 M 2  KOH) for 5-10 minutes at room temperature and then the embryos were washed three times with M 9 buffer. The eggshells of the washed embryos were digested for 5 minutes by chitinase (Sigma) dissolved in egg salts (118 m M NaCl, 3.4 m M CaC12, 3.4 m M M g C l , 5 m M Hepes, pH 7.2). The digested embryos were fixed for 2 minutes in a 2  formaldehyde solution [4% formaldehyde, 60 m M Pipes, 25 m M Hepes, 10 m M E G T A , 2 m M M g C l , pH 6.8, containing 0.1 mg/mL L-a-lysolecithin (Sigma Chemical 2  Company). Fresh fixative without the lysolecithin was added and the embryos were fixed for an additional 20-30 minutes. After fixation, embryos were washed once in TBS, followed by another wash in TBS-T (TBS, 0.1% Tween). Embryos were then stained for 1 hour to overnight with 1 | i M FITC conjugated phalloidin (Sigma). Stained embryos were washed 3 times in TBS-T and then once in TBS and resuspended in mounting media (90% glycerol, 2.5% D A B C O , TBS). This protocol was adapted from the phalloidin staining protocol described by Costa and colleagues (Costa et al. 1997).  Immunofluorescence microscopy Immunofluorescence staining images were collected on either a Zeiss Axiophot microscope equipped for epifluorescence or a Nikon Optiphot-2 microscope using the  32  M R C 600 Confocal system (Bio-Rad) equipped with a Krypton/Argon laser. The images collected from the Zeiss Axiophot were photographed using Kodak Tmax 400 35 mm film or a Dage-MTI CCD-100 digital camera. The images collected from the confocal microscope were captured in a 400X400 pixel field of view with the optical sections collected at 0.2 |im intervals. The confocal images were composed of 100 to 150 optical sections for each embryo. Data collected from the confocal microscope were analyzed in NIH Image 1.61 and Adobe Photoshop 4.0 was used for presentation.  Differential interference contrast microscopy Differential interference contrast (DIC) images were collected on a Zeiss Axiophot microscope equipped with DIC optics. In short, embryos were washed off plates in d H 0 2  and treated with an alkaline hypochlorite solution to dissolve any larvae or gravid adults. The embryos were then washed I X in M9 buffer and placed on a microscope slide. The images were collected using a Dage-MTI CCD-100 digital camera and Scion image software. Adobe Photoshop 4.0 was used to present images.  Several mutants were examined by time-lapse video recordings. For time lapse video microscopy, circular 2% agarose pads (prepared in M9 buffer) were made on microscope slides. Several embryos were transferred to the agar pad in a drop of M9 buffer. Silicon grease or Vaseline was used to surround the agar pad (to hold the cover slip in place and to prevent evaporation). Excess M9 buffer was added to fill the space between the agar pad and the silicon grease. A cover slip was gently placed on top and time lapse video recordings were initiated. Embryos were observed using DIC optics and time lapse video  33  recording were made using a Panasonic Super VHS V C R (AG-6730). Images were taken every 3 minutes over an 4-8 hour period.  Molecular biology techniques PCR was conducted as described by Barstead et al (1991). In brief, several (3-5) wild type embryos, homozygous mutant embryos or homozygous deficiency embryos were placed in 3 uL of lysis buffer (50 m M K C L , 10 m M Tris (pH 8.0), 2.5 m M M g C l , 0.45% 2  Tween 20, 0.45% NP-40, 60 }Xg/mL Proteinase K) in a 0.5 mL microfuge tube. The tubes were incubated at 60°C for 45 minutes to lyse the embryos and then heated to 95°C for 15 minutes to inactivate the Proteinase K . For each reaction tube, 2 | i L of 10 m M dNTPs (Boehringer Mannheim), 2.5 uX of 10X PCR buffer (BRL), 0.4 | i L of 50 m M M g C l , 1 2  (il primers (25 pmole/p:L, 0.5 fiL of Taq D N A polymerase (2.5 Units) were added and the total volume was brought up to 25 p,L with sterile d H 0 . Amplification of the PCR 2  mixtures was carried out in a Perkin Elmer Cetus 480 thermocycler using standard procedures (Lundquist et al., 1997).  Long range PCR using the Boehringer Mannheim Expand Long Range Template PCR System was used to amplify the full length let-268 gene, including 2.2 Kb of 5' upstream sequence, from N2 genomic D N A following the protocol described by Barstead et al. (1991).  D N A sequencing reactions were carried out directly on PCR amplified genomic D N A (several independent reaction) and to c D N A clones within the pBluescript vector  34  (Stratagene) by using the B R L dsDNA Cycle Sequencing System as described by Rogalski et al. (1993, 1995). In brief, the sequencing primers were end labeled with (Y^P) A T P (25 uCi). Sequencing primers (1 pmole/uJ) were mixed with T4 polynucleotide kinase, 5X kinase buffer, labeled ATP and dH 0 in a 0.5 mL microfuge 2  tube and incubated at 37°C for 30 minutes, then heated to 55°C to end the reaction. The end labeled primers were then used for cycle sequencing as described by manufacturer (BRL). After cycle sequencing, the reactions were heated to 95°C for 5 minutes before being loaded onto a standard 6% acrylamide sequencing gel.  To determine the molecular lesion in each of the let-268 alleles, PCR products from homozygous let-268 mutant D N A were directly sequenced using internal primers. The entire coding region and the majority of the introns were sequenced except the 5' and 3' untranslated regions. Also, cDNAs from the let-268 locus, yk318c8 and yk476hl (GenBank accession number C63295 and C50032, respectively; kindly provided by Y . Kohara, National Institute of Genetics, Mishima, Japan) were sequenced to determine the 5' and 3' ends and the intron/exon boundaries.  Sequence alignments and comparisons were carried out using B L A S T (NCBI server, Altschul et al., 1997) and ClustalW (Mac Vector, Oxford Molecular Group).  Gene constructs General molecular biological techniques were carried out as described in Sambrook et al. (1989). The bacterial strains DH5cc (BRL) and X L l - B l u e (Stratagene) were used for  35  subcloning and fusion protein expression. To generate a reporter construct for let-268, a 2.7 kb Bgl II fragment was isolated from the cosmid F52H3 (kindly provided by A . Coulson, The Sanger Centre, Wellcome Trust Genome Campus, Hinxton, U K ; GenBank accession number Z66512) and was cloned into the green fluorescent protein (GFP) expression vector pPD95.79 (kindly provided by A . Fire, S. Xu, J. Ahnn and G. Seydoux, Carnegie Institute of Washington, Baltimore, MD). This clone DM#212 includes 2 kb upstream of the let-268 ORF, all of the first exon, all of the first intron and 408 base pairs of the second exon that is fused to GFP (Fig. 13 D).  To generate the full length a spectrin clone, a 4.7 kb Avr II Spe I fragment from the cosmid M01F12 (kindly provided by A . Coulson, The Sanger Centre, Wellcome Trust Genome Campus, Hinxton, U K ) was subcloned into Spe I digested pBluescript (Stratagene) to produce DM#213. Subsequently, a 9.1 kb Spe I fragment from the cosmid M01F12 was subcloned into the Spe I site of DM#213 producing a clone (DM#214) that contains the full length C. elegans a spectrin gene including 2 kb upstream and 1 kb downstream.  To generate a fusion protein encoding part of the C. elegans a spectrin gene, PCR was carried out to amplify a 957 bp region of the third exon from genomic D N A (PCR involved a 30:1 mixture of Taq (Gibco BRL) and Vent (New England Biolabs) D N A polymerases to reduce incidence of PCR-induced errors). Subsequently, the clone DM#215 was produced by digesting this PCR product with Apo I and the resultant 724 bp fragment was subcloned into the EcoRI site of the p G E X 4T-1 vector (Pharmacia).  36  RNA mediated interference The production of dsRNA was carried following the protocol previously described (Fire et al., 1998). In short, template cDNAs in pBluescript (Stratagene) were linearized with the appropriate restriction enzyme (Sac II and Apa I) and transcription was carried out using the T3 and the T7 R N A polymerases (BRL). This consisted of mixing 1 u\g of template D N A with 10 uL of 10 m M NTPs (BRL), 20 uL 5X Reaction Buffer (BRL), 5 UL 0.1 M DTT (1,4-Dithio-DL-threitol) and 2 uL of the R N A polymerase brought up to a total volume of 100 mL with d H 0 . This mixture was then incubated at 37°C for 1-3 2  hours. Once the transcription reaction was completed, the mixtures were extracted twice with phenol chloroform and ethanol precipitated. The ssRNA was resuspended in 20 (J,L d H 0 . Equimolar amounts of sense and antisense ssRNA were annealed to yield dsRNA 2  (-1-2 kb) which then was injected into wild type animals as previously described (Fire et al., 1998). The injected animals were placed at 20°C and allowed to lay eggs for 24 hours and then transferred to a fresh plate for another 24 hours. The progeny from injected animals were scored and compared to the progeny of uninjected wild type, let-268(ra414) (for yk476hl) and spc-l(ra409) (for yk205f3) animals on a Zeiss Axiophot compound microscope using DIC optics. The templates used for R N A production were the EST clones yk476hl and yk205f3. yk476hl was used to generate dsRNA homologous to the let-268 ORF and yk205f3 was used to generate dsRNA homologous to the spc-1 ORF.  37  Worm transformation D N A was injected directly into the gonad of N2 hermaphrodites as described by Mello and Fire (1995). N2 animals were injected with three different cocktails of D N A . The first cocktail consisted of 80 p,g/mL of pRF4 [dominant rol-6 mutation; rol-6(sul006dm)] and 20 p_/mL of DM#212 (let-268::G¥P expression vector). Roller progeny from N2 animals injected with this cocktail were selected and analyzed for GFP expression. The second cocktail consisted of 80 pg/mL of pRF4 and 10 p,g/mL of M01F12 (cosmid that contains the full length a spectrin). Roller progeny from N2 animals injected with this cocktail were selected and tested for their ability to rescue the spc-l(ra409) mutant phenotype. The third cocktail consisted of 80 (ig/mL of pRF4 and 20 p,g/mL of DM#214 (subclone of M01F12 containing only the a spectrin ORF). Roller progeny from N2 animals injected with this cocktail were selected and tested for their ability to rescue the spc-l(ra409) mutant phenotype. Finally, a Long range PCR product containing the full length let-268 gene plus 2.2 Kb upstream was injected into the strain, unc-4 let-268 (ra414)/mnCl (dpy-10 unc-52). Rescued animals were recognized by the presence of Unc-4 animals.  Fusion protein production and purification The GST a spectrin fusion protein was produced and purified by the procedure previously described (Smith and Johnson, 1988). The clone DM#215 (construction discussed above) was used to produce fusion protein. First, a 3 mL bacterial culture, grown at 37°C for 8 hours, was used to inoculate a 20 mL overnight culture. The overnight culture was grown in L Broth containing appropriate antibiotics for 14 hours at  38  37°C. This culture was then used to inoculate a 250 mL culture of L broth with appropriate antibiotics and was grown for 3 hours at 37°C. After the 3 hours had elapsed, IPTG (Isopropyl (3-D-thiogalactopyranoside)was added to a 1 m M final concentration and incubated at 37°C for an additional 3 hours. The cells were then harvested by centrifugation and resuspended in 9 mL MT-PBS (150 m M NaCl, 16 m M N a H P 0 , 4 2  4  m M NaH P0 ) containing 0.5 mM PMSF (phenylmethylsulfonyl fluoride) and 0.5 m M 2  4  E D T A . The cells were then lysed by sonication (6X 30 second pulses) and kept on ice. 1 mL of a 10% Triton X-100 in MT-PBS solution was then added to the lysed cells to bring the final volume to 10 mL and was incubated on ice for 5-10 minutes. The mixture was then subjected to centrifugation at 10,000 R P M to pellet insoluble material. The supernatant was transferred to a sterile 15 ml culture tube containing 3 mL pre-swollen glutathione-agarose beads (Sigma Chemical Company). The tube was gently nutated at room temperature for 5-10 minutes. The glutathione beads were washed 5 times in M T PBS with 1% Triton X-100, followed by 2 times in MT-PBS. The bound fusion protein was eluted from the beads 3 times at room temperature with 3 mL of elution buffer (50 m M Tris-Cl, 10 m M glutathione; Sigma Chemical Company) and stored at -80°C.  Antibody production To generate polyclonal antisera, two New Zealand White rabbits were injected subcutaneously with purified fusion protein emulsified in Freund's complete adjuvant (-0.5 mg protein/rabbit). Rabbits were boosted by intramuscular injections at 4-6 week intervals with fusion protein emulsified in Freund's incomplete adjuvant (-0.3 mg protein/rabbit) and blood samples taken 10 to 12 days post injection. Immune response  39  was monitored by Western blotting of purified fusion proteins and worm extracts and by immunofluorescence staining of nematodes.  Western analysis Proteins were separated by SDS-PAGE following standard procedures (Sambrook et al., 1989). Fusion proteins were separated on 10-12% acrylamide gels. Worm extracts were separated on 7-8% acrylamide gels. The proteins were then transferred to Hybond E C L grade nitrocellulose membranes (Amersham) for 15 - 30 min. at 12 - 15 V in a Trans-Blot SD Electrophoretic Transfer Cell (Bio-Rad). The Blots were blocked in 5% milk powder-TBS-T (TBS-T: 20 m M Tris (pH 7.6), 137 m M NaCl, 0.1% Tween 20) overnight at 4°C. The primary antibodies were added in TBS-T with 1% milk powder and incubated for 3 hours at room temperature. After washing 4 X in TBS-T, the blots were incubated with horseradish peroxidase-labelled secondary antibodies (ECL kit, Amersham) in TBS-T with 1% milk powder for ~1 hour at room temperature. After washing in 4 X in TBS-T and I X in TBS, the blots were incubated for 1 min. in E C L detection reagents (Amersham) and exposed to film (Kodak X - O M A T ) .  For Western blotting, rabbit polyclonal sera were diluted as follows: AS1 (1:25000 1:100000) and G M 2 (1:1000 - 1:2000). Mouse monoclonal antibody DM5.6 (Miller et al., 1983) was diluted 1:500 (for ascites fluid). AS1 preimmune was diluted 1:5000. Secondary antibodies, horseradish peroxidase-labelled goat anti-rabbit IgG or goat antimouse IgG (Amersham), were both diluted 1:10000.  40  Chapter 3. Mutant screen  Background Through the application of genetic dissection -80 genes have been identified that affect muscle in the nematode C. elegans. This list includes such molecules as myosin, paramyosin, actin, vinculin and twitchin (for a complete list see http://www.zoology.ubc.ca/labs/moerman/muscle_affecting_genes.html).  Most of the  early genetic screens have focused on phenotypes typified by abnormal locomotion or abnormal myofilament structure as observed in mature adults (Waterston, 1988; Moerman and Fire, 1997). This class of muscle mutant is most likely near saturation but since there are still several muscle proteins that have not been identified by this mutant phenotype (i.e. integrin and troponins), a different method is required for the identification of new components of the body wall muscle of C. elegans. Several years ago, a screen carried out by Williams and Waterston (1994) focused on identifying embryonic lethal mutations that affect muscle development. This screen had been based on the mutant phenotype associated with mutations identified in the genes encoding myosin heavy chain A (Waterston, 1989) and vinculin (Barstead and Waterston, 1991). Null mutations in these genes result in animals that are paralyzed, arrested development at the two fold stage during embryogenesis. This mutant class is known as Pat for paralyzed arrested at two fold. Development commences normally in these mutant embryos until body wall muscle contraction initiates. At this point the mutant embryos fail to display any body wall muscle contraction. These embryos also cease elongation at the two fold length and fail to become worm shaped larvae (see Figs. 2 and 3).  41  In this study a mutant screen similar to the one described previously by Williams and Waterston (1994) was developed to identify mutants with defects in myofilament assembly and stability. In the Williams and Waterston (1994) screen, the primary focus was to identify animals that were strictly paralyzed and arrested at the two fold stage of embryogenesis (Pat). In this study, we hypothesized that mutations in components important for embryonic development of muscle could be mutated and result in a less severe defect in myofilament assembly or stability. These mutants could arrest slightly after two fold to just after hatching. In addition, screening for mutants that arrest prior to the two fold stage may identify mutants with defects in the migration of the laterally positioned body wall muscle cells. The screen carried out by Williams and Waterston (1994) did not identify all loci that have a Pat phenotype (i.e. unc-45, Venolia and Waterston, 1990; etr-1 Milne and Hodgkin, 1999) and only one allele was isolated for several Pats mutants indicating that the screen was not large enough to identify all possible Pat mutants. Therefore, a screen was devised that would search for muscle defects in animals that arrested around the 1.5 fold stage to just prior to hatching (290800 minutes post fertilization). This screen should allow for a wider array of mutants at different developmental times as well as more Pat mutants to be identified. The body wall muscle and the tightly associated underlying basement membrane were analyzed in all of the mutants by immunofluorescence microscopy and the mutants with abnormal body wall musculature were retained for further analysis. Eight mutations were identified in a relatively small screen, which indicates that a future larger scale screen may be useful for identifying additional new components required for muscle  42  development. At least two of these mutations, ra401 and ra4J4, appear to be new alleles of previously identified genes, unc-52 (Rogalski et al., 1993) and let-268 (Sigurdson et al., 1984; Chapter 4), respectively. However, the other mutations isolated this screen appear to be newly identified genes that have a role in C. elegans muscle development.  43  Results Mutant screen Two different mutagenesis screens were conducted to test if our strategy would identify mutants with defects in muscle development. In the first, N2 hermaphrodites were mutagenized and the FI progeny produced by self fertilization were picked and placed individually on separate plates and allowed to lay eggs. After 50 eggs had been laid, the FI animal was removed and plates that contained approximately 25% animals that had arrested development between the 1.5 fold stage to freshly hatched larvae were retained for further analysis. From 600 FI populations, a total of twenty one mutants were selected in this genome wide screen (see below).  A second screen was carried out to identify mutations in the central region of chromosome II which can be balanced by the well characterized genetic balancer, mnCl (Herman, 1978; Sigurdson et al., 1984). In this screen unc-4/mnCl  [dpy-10 unc-52]  hermaphrodites were mutagenized and single heterozygous hermaphrodite FI offspring were transferred to individual plates. The progeny of the -1000 FI hermaphrodites were scored for the absence of Unc-4 animals. Animals that segregated only wild type, DpyUncs and arrested embryos or L l larvae were retained for further analysis (presumptive genotype unc-4 let-XJ'mnCl [dpy-10 unc-52]). Two mutants were isolated that fit this criteria. This mutant screen was more restrictive than the genome wide screen described above and only identified mutations in the central region of chromosome II. This region of chromosome II contains -2000 genes compared to the whole genome that contains -19,000 genes (C. elegans Sequencing Consortium, 1998). However, the  44  mutants isolated in this screen were genetically balanced when identified whereas the mutations isolated in the genome wide screen needed to be genetically mapped and balanced and were therefore more labor intensive to maintain.  Wild type embryonic development Before describing the mutant phenotypes isolated in the two mutant screens, it is first necessary to describe the development of the wild type embryo (reviewed in Wood, 1988; Hresko et al., 1994). The first 350 minutes post fertilization (mpf) is mainly composed of cell proliferation. After this point the embryo undergoes morphogenesis and organogenesis. At -350 mpf, the embryo begins to elongate converting the ovoid shaped embryo into a worm shaped embryo. The process of elongation occurs in small increments and lasts until the embryo hatches from the egg. Each of these increments are shown in Fig. 3. First, the embryos begins to constrict in the posterior end producing an embryo that is known as the lima bean stage embryo (360 mpf; Fig. 3 A). As the embryo continues to elongate, it passes through the comma stage (390 mpf; Fig. 3 B) and the 1 1/2 fold stage (420 mpf; Fig. 3 C). At the 1 1/2 fold stage the embryo begins to display some body wall muscle contraction as detected by the twitching of the embryo. At this point the myofilaments and other structural proteins have localized to the basal membrane of the body wall muscle cells creating a highly polarized cell (Fig. 2 B). At 450 mpf, the embryo has elongated to the 2 fold stage (Fig. 3 D) and is now beginning to display more vigorous movements within the egg. The myofilaments at this point are organized into sarcomeres (Fig. 4 B). The thick filaments (myosin filaments) are organized into double rows of A bands and the thin filaments (actin filaments) have  45  Figure 3. The process of elongation during wild type embryogenesis. (A) Wild type embryo at the lima bean stage (350 mpf) beginning to undergo elongation. (B) Wild type embryo at the comma stage (390 mpf). (C) Wild type embryo at the 1 1/2 fold stage (420 mpf). At the 1 1/2 fold stage the embryo begins to display body wall muscle contraction. (D) Wild type embryo at the two fold stage (450 mpf). At the two fold stage the sarcomeres have assembled (see Figure 2) and the embryo is starting to move around within the egg. (E) Wild type embryo approaching the 2 1/2 fold stage (-480 mpf). (F) Wild type embryo between three fold and four fold in length preparing to hatch (~700mpf). Scale bar is 10 (im.  46  47  Figure 4. Muscle development in the wild type embryo. (A-D) Wild type embryo at the two fold stage and (E-H) wild type embryo at the three/four fold stage. (A) DIC micrograph of a wild type embryo at the 2 fold stage. (B, D) Wild type embryo at the two fold stage doubled labeled with antibodies to myosin, DM5.6 (B), and perlecan, GM1 (D). (B) Myosin has localized to the membrane where adjacent muscle cells contact each other and the hypodermis. The myosin filaments are organized into double rows of A bands. (D) Perlecan is evenly distributed in the basement membrane underlying the body wall muscle quadrants. A n even distribution of perlecan is found surrounding the pharynx (D). (C) Two fold wild type embryo labeled with FITCphalloidin. The thin filaments have polarized and are organized into double rows of I bands. (E) DIC micrograph of a wild type embryo at the three/four fold stage. (F, H) Wild type embryo at the three/four fold stage double labeled with antibodies to myosin, DM5.6 (F), and perlecan, GM1 (H). Myosin is organized into mature sarcomeres (F), and perlecan is evenly distributed in the basement membrane underlying the body wall muscle and surrounding the pharynx (H). (G) Wild type embryo at the three/four fold stage labeled with FITC-phalloidin. The thin filaments are organized into sarcomeres in the body wall muscle and the thin filaments are organized into half I bands in the pharynx (see text). Additionally, filamentous actin within the intestine can be seen extending from the posterior end of the pharynx to the posterior of the embryo (G). Scale bar is 10 urn.  48  49  organized into double rows of I bands within each body wall muscle cell (Fig. 5 A , B). By 550 mpf, the embryo has reached the three fold stage (Fig. 3 F, 4 F) and is vigorously rolling within the egg. Contraction of the pharynx is not observed until -750 mpf when the embryo is close to four fold in length and is near to hatching (Fig. 4 G , H and 5 C). Pharyngeal contraction is required for the secretion of digestive enzymes that soften the egg shell allowing the embryo to hatch (Williams and Waterston, 1994; Plenefisch et al., 2000).  Phenotypic analysis The mutations isolated have been placed into 5 classes depending upon the severity of the mutant phenotype (Table 3). Determination of the mutant phenotypes were based on observations using differential interference contrast microscopy (DIC) and immunofluorescence microscopy. Antibodies specific to body wall muscle myosin (DM5.6; see Table 2) and the basement membrane protein perlecan (GM1; see Table 2) were used as indicators of muscle organization, namely the myofilaments and the tightly associated basement membrane (see the wild type staining pattern in Fig. 5). Some of the mutants were further characterized with fluorescently labeled phalloidin (F-actin binding toxin) and all were genetically mapped. Class I, the most severe class (most similar to the Pat mutants), contains one mutant ra401. In this mutant the body wall muscle cells fail to polarize and the myofilaments never assemble. The class II mutants, ra407 and ra408, are less severe than class I. In these mutants, the muscle structural proteins polarize normally; however, normal myofilaments are never made. The class III mutants, ra402 and ra404, appear to have defects in anchoring the body wall muscle to the  50  Table 3. Classification of mutants. Class  Mutants  Time of  Myosin staining  Perlecan staining  Arrest  Genetic map position  Class I  ra401  2 fold  Fails to assemble  absent  unc-52  Class II  ra407  2 fold  Fails to assemble  Normal  Middle of II  ra408  2 fold  ra402  2 fold  Class III  ra404  Class  ra406  2 fold  2-3 fold  IV ra414  Middle of III Fails to assemble  Normal until  and the muscle  separates from the  quadrants separate  hypodermis and  from the  remains attached  hypodermis  to the muscle  the muscle  Gaps in basement  quadrants separate  membrane  2-2.5  from the  fold  hypodermis and  Right arm of I  Left arm of II  Middle o f X  Middle of II  gaps arise in the muscle quadrants Class V  ra409  2 fold  Muscle cells wider and abnormal organization  51  2 fold wider  Left arm of X  Figure 5. Subcellular organization of myosin and perlecan in the wild type embryo. Wild type embryo at the three/four fold stage double labeled with antisera to myosin, DM5.6 (A, B) and to perlecan, GM1 (C, D). (A) The same embryo shown in Fig. 4F is shown here depicting the normal organization of the myosin filaments. (B) The subcellular organization of the myosin filaments are shown from the right dorsal muscle quadrant of the embryo shown in (A). The filaments are organized into double rows of A bands that run nearly parallel with the longitudinal axis of the animal. The myosin filaments are assembled against the body wall muscle cell membrane adjacent to neighboring muscle cells and the hypodermis. (C) The same embryo shown in Fig. 4 H is shown here depicting the normal distribution of perlecan in the basement membrane underlying the body wall muscle and surrounding the pharynx. (D) The basement membrane underlying the right dorsal quadrant of the embryo in (C) is shown. Perlecan is evenly distributed under and is in close association with the body wall muscle. Scale bar is 10 p,m (A and C) and is 5 |im (B and D).  52  53  underlying hypodermis. In these mutants, the muscle proteins polarize and partially assemble into myofilaments (see below). However, the muscle quadrants separate from the hypodermis. The class IV mutants, ra406 and ra414, assemble normal and functional myofilaments, but the anchoring of the body wall muscle to the hypodermis and attachments between neighboring muscle cells are defective (i.e. gaps arise in the body wall muscle quadrants). The class V mutant, ra409, has a unique phenotype that has not previously been described: the body wall muscle is twice as wide as normal.  Class I Class I is composed of one mutant, ra401. ra401 is the only true Pat identified in our mutant screens. This mutant fails to elongate beyond the two fold stage (Fig. 6 A) and displays no body wall muscle contraction. However, pharyngeal muscle contraction occurs and the mutant embryos are capable of hatching. Analysis of ra401 mutants with antibodies to myosin indicate that the myosin filaments fail to polarize and never assemble (Fig. 6 B), which is consistent with these mutants failing to display body wall muscle contraction. Moreover, ra401 mutants fail to stain with a polyclonal antibody directed to the basement membrane protein perlecan (UNC-52; Moerman et al., 1996; Fig. 6 C). Pat alleles of unc-52 have been identified previously and have demonstrated the essential requirement of perlecan in the basement membrane underlying the body wall muscle for myofilament assembly (Rogalski et al., 1993). To determine if ra401 is a new allele of unc-52, we conducted a complementation test between ra401 and a viable unc52 mutant, unc-52(e444). ra401 failed to complement unc-52(e444) and gave rise to Unc animals that are sterile. This is consistent with previous reports that heteroallelic  54  Figure 6. Characterization of the Class I mutant, ra401. (A) DIC micrograph of a ra401 mutant embryo. Embryonic elongation ceases at the two fold length and these mutants display no body wall muscle contraction. The pharynx continues to develop (arrow head) and is capable of some muscle activity. (B and C) another ra401 mutant embryo double labeled with a myosin antibody, DM5.6 (B), and a perlecan antibody, GM1 (C). (B) Myosin has failed to polarize to the basal membrane and has failed to assemble into thick filaments. (C) ra401 mutants stained for perlecan failed to recognize any protein indicating that no perlecan is present in ra401 mutant embryos. Scale bar is 10 um.  55  56  combinations between unc-52 Pat alleles and viable alleles produced sterile Unc animals (Gilchrist and Moerman., 1992). This mutant has been subsequently balanced with mnDp34 (Fig. 12) and has been useful for studying the biology of perlecan (Mullen et al., 1999).  Class II Class II is composed of two mutants, ra407 and ra408. Both ra407 and ra408 resemble mild Pats (Williams and Waterston, 1994). These two mutants fail to elongate beyond the two fold stage of embryogenesis (Fig. 7 A , D) and display weak body wall muscle contraction. The initiation of body wall muscle contraction occurs at the normal developmental time (350 minutes post fertilization), but these mutant animals never display the vigorous movement within the egg as observed in wild type. ra407 and ra408 mutant embryos display only minor twitching until they arrested development. Analysis of ra407 and ra408 mutants by immunofluorescence with antibodies to myosin and perlecan indicate that the myosin filaments polarize normally but never organize into double rows of A bands as seen in wild type (compare Fig. 5 A , B and 7 B , E). Nonetheless, the distribution of perlecan appears normal in both mutants (Fig. 7 C, F). To investigate the organization of the thin filaments, ra407 homozygotes were labeled with phalloidin-FITC. Similarly, the thin filaments have polarized; however, they do not assemble into distinct I bands (Fig. 8 B). Therefore, it appears that the ra407 mutation affects both the organization of the thick and thin filaments. Interestingly, the ra407 homozygotes are capable of some weak body wall muscle contraction even though the  57  Figure 7. Characterization of the Class II mutants, ra407 and ra408. Both ra407 (A-C) and ra408 (D-F) were examined by DIC and immunofluorescence microscopy with antisera to myosin, DM5.6 (B, E), and perlecan, GM1 (C, F). (A) DIC micrograph of a ra407 mutant embryo. The process of elongation ceases at the two fold length; however, these mutants are capable of some minor body wall muscle contraction. The pharynx differentiates (arrowhead) and appears fairly normal (see fig 6), but no muscle activity is observed and ra407 mutants fail to hatch. (B-C) another ra407 mutant embryo double labeled with antisera to myosin (B) and perlecan (C). (B) myosin polarizes to the basal membrane of the muscle cells (arrow) but fails to assemble into double rows of A bands. (C) Perlecan is distributed evenly within the basement membrane surrounding the pharynx and underlying the body wall muscle. (D) DIC micrograph of ra408 mutant embryo. Similar to ra.407, the process of elongation ceases at the two fold length in ra408 mutants. However, as is observed in ra407 mutant embryos, ra408 mutants are capable of some minor body wall muscle contraction. The pharynx differentiates completely (arrowhead) and is capable of muscle contractions. (E-F) another ra408 mutant embryo double labeled with antisera to myosin (E) and perlecan (F). (E) myosin polarizes to the basal membrane of the muscle cells (arrow) but fails to assemble into double rows of A bands. (F) Perlecan is distributed evenly within the basement membrane underlying the body wall muscle and surrounding the pharynx. The pharynx of ra.408 mutants appears more normal than the pharynx of ra407 mutants (compare C and F with Fig. 2 H). Scale bar is 10 \im.  58  59  myofilaments have not organized into proper sarcomeres. The organization of the thin filaments in ra408 animals have not been investigated.  The pharynx, the other major muscle tissue in C. elegans, does not display any muscle contraction in ra407 mutants. Examination of the myofilament organization in the pharynx of ra407 mutants had been conducted by staining mutants with FITC-phalloidin. In wild type, the thin filaments organize on the basal and luminal membrane of the pharynx (forming two sets of half I bands; Fig. 4 G and 8 A) and a small gap appears between the two half I bands where the myosin filaments assemble (known as the H zone). The organization of the thin filaments in ra407 mutant embryos appears normal (Fig. 8 B). Even though the morphology of the pharynx is not normal (all Pat mutants have shorter than normal pharynges), the thin filaments of the pharynx are organized into two half I bands with a normal spaced H zone between the two sets of thin filaments (Fig. 8 B). The pharynx in ra408 mutants is capable of muscle contraction and therefore, the myofilaments of the pharynx were not investigated.  ra407 maps to chromosome I and is tightly linked to unc-40. Three factor mapping positions ra407 just to the right of unc-40 (Fig. 12 and Table 4). One known muscle affecting gene falls in this region, unc-94 (Zengel and Epstein, 1980). A complementation tests has been carried out between ra407 and unc-94(sul77).  unc-  94(sul77) complements ra407. Therefore, ra407 and unc-94(sul77) affect two separate loci. ra408 maps to the center of chromosome III. By three factor mapping, ra408 is positioned between unc-36 and unc-32 (Fig. 12 and Table 4). A complementation test  60  Figure 8. Phalloidin Staining of wild type and ra407 mutant embryos. Wild type (A) and ra407 mutant embryos (B) were labeled with FITC-phalloidin. (A) A three/four fold wild type embryo is shown (similar developmental time as ra407 mutant shown in B). The thin filaments have localized to the basal membrane of the body wall muscle quadrants and are organized into sarcomeres. Also, the thin filament have organized into half I bands in the pharynx. (B) The thin filaments in ra407 mutants have polarized but have failed to assemble into double rows of I bands. The thin filaments have not formed into mature sarcomeres as seen in wild type embryos (compare A and B). Additionally, the thin filaments in the pharynx of ra407 mutant embryos have organized into half I bands similar to those observed in wild type (compare A and B). Scale bar is 10 pm.  61  62  Table 4. Three Factor Mapping  Mutant  Genotype  Summary of recombination events  ra402  + ra402 +/dpy-5 + unc-54  dpy-5 (24/45) ra408 (21/45) unc-54  + + ra402/lin-ll unc-75 +  lin-11 (17/17) [unc-75 ra402]  ra404  ra404 + +/+ dpy-10 unc-4  [ra404 dpy-10] (14/14) unc-4  ra406  + + ra406/unc-2 lon-2 +  unc-2 (21/21) [lon-2 ra406]  + + ra4067lon-2 unc-18 +  lon-2 (11/11) [unc-18 ra406]  + ra406 +/dpy-7 +unc-3  dpy-7 (4/17) ra406 (13/17) unc-3  + ra407 +/dpy-5 + unc-54  dpy-5 (2/26) ra407 (24/26) unc-54  + ra407 +/dpy-5 + unc-87  dpy-5 (16/28) ra407 (12/28) unc-87  + + ra407/dpy-5 unc-40 +  dpy-5 (8/8) [unc-40 ra407]  + + ra4087lon-l unc-36 +  lon-1 (8/8) [unc-36 ra408]  + ra408 +/dpy-17 + unc-32  dpy-17 (21/22) ra408 (1/22) unc-32  + ra409 +/unc-2 + lon-2  unc-2 (12/48) ra409 (36/48) lon-2  + ra409 +/dpy-3 + unc-20  dpy-3 (45/51) ra409 (6/51) unc-20  ra407  ra408  ra409  63  has been carried out between ra408 and pat-2(st538). Both of these mutations map close to unc-32 and the similarity in phenotypes suggested that they could be allelic (Williams and Waterston, 1994). These two mutants complemented each other indicating that they are mutations in different loci.  Class III Class III is composed of two mutants, ra402 and ra404. These mutant animals display normal body wall muscle "twitching" around the 1 1/2 fold stage of embryogenesis. However, the body wall muscle fails to undergo any vigorous body wall muscle contraction as observed in wild type embryos. Both ra402 and ra404 mutant embryos fail to complete elongation and arrest at the two fold stage (Fig. 9 A) and can also be classified as mild Pats. Pharyngeal contraction occurs in these mutants and they are able to hatch. Antibody analysis of myosin in the body wall muscle of these mutants reveals that the myosin filaments appear to polarize normally to the basal side of the cell and adjacent to the neighboring body wall muscle cell in the same quadrant (9 B , C). However, the myosin filaments never assemble completely. In wild type animals, the myosin filaments are organized in double rows forming two A-bands in each body wall muscle cell (Fig. 5 A , B). In ra402 and ra404 mutants the double row of myosin is never observed indicating the incomplete assembly of the myosin filaments. Examination of the thin filaments in ra402 animals with FITC-phalloidin indicates that these filament structures assemble normally. Rows of I bands comparable to those in wild type embryos are observed in each muscle cell of ra402 mutants (compare Fig. 5 C, G and 9 D). Therefore, the assembly of the myosin filaments and not the actin filaments are affected  64  Figure 9. Characterization of the Class III mutant, ra402. (A) DIC micrograph of ra402 mutant embryo. Although the process of elongation ceases at the two fold length, these mutants are capable of some minor body wall muscle contraction. The pharynx differentiates (arrowhead) and appears fairly normal. (B, E and C,F) Two other ra402 mutant embryos double labeled with antisera to myosin, DM5.6 (B and C), and to perlecan, GM1 (E and F). The ra402 mutant in (B and E) is slightly older than the ra402 mutant shown in (C and F). In both (B) and (C), myosin has polarized to the basal side of the muscle cell (horizontal arrow); however, proper myofilaments are never organized. The embryo in (C) shows that the left dorsal muscle quadrant is in the process of separating from the hypodermis (arrowhead). The vertical, arrow in (C) indicates the region where the muscle quadrant has separated from the hypodermis. The embryo in (B) shows that both dorsal muscle quadrants have separated from the hypodermis (arrowhead). As in (C), the vertical arrow in (B) indicates where the body wall muscle quadrants have separated from the hypodermis. In both (E) and (F), perlecan distribution is evenly distributed throughout the basement membrane surrounding the pharynx and the basement membrane underlying the body wall muscle. The basement membrane containing perlecan also separates from the hypodermis along with the body wall muscle quadrants (arrowhead). The vertical arrows in (E) and (F) indicate where the basement membrane has separated from the hypodermis. (D) another ra402 mutant embryo labeled with FITC-phalloidin. The thin filaments of ra402 mutants polarize and are capable of forming into double rows of I bands as seen in wild type (see fig. 4 G). Scale bar is 10 urn  65  66  in ra402 mutant animals. Examination of the thin filament organization in ra404 mutants had not been carried out.  Antibody analysis in slightly older ra402 and ra404 mutant embryos demonstrates that the muscle and the underlying basement membrane separates from the hypodermis (Fig. 9 B , E compared to Fig. 9 C, F). This indicates that these mutant animals are capable of some body wall muscle contraction and that the force of contraction pulls the body wall muscle and the underlying basement membrane away from the hypodermis. This suggests the presence of a weak link in the interconnections of the body wall muscle and the hypodermis. The body wall muscle is anchored to the cuticle via attachments to the basement membrane and the hypodermis interacts with the basement membrane and the cuticle through hemidesmosomal-like structures (Fig. 1). It is possible that there is a defect in one of these attachment structures.  To investigate the integrity of the basement membrane, ra402 and ra404 mutant embryos were stained with an antiperlecan antiserum. Both mutants reveal a normal distribution of perlecan (Fig. 9 E , F). However, when the muscle quadrant separates from the hypodermis perlecan remains associated with the body wall muscle cells (Fig. 9 E). These results suggest that a defect did not exist in the attachment of perlecan to the body wall muscle but that an attachment defect arose somewhere between perlecan and the underlying hypodermis.  67  The ra402 mutation genetically maps to the right arm of chromosome I. Three factor mapping places ra402 to the left of unc-75 (Fig. 12 and Table 4). There are no known mutations affecting muscle in this area except vab-10, a mutant that has been classified as a Mua (defective muscle attachment, Plenefisch et al., 2000). vab-10 mutants display a degenerate head musculature, that is bent dorsally or ventrally and these defects have been associated with defects in body wall muscle attachment (Plenefisch et al., 2000). These two mutants complemented each other indicating that they fall into two separate loci. ra404 maps to chromosome II and falls outside the chromosome II balancer, mnCl. Three factor mapping also positions ra404 to the left of dpy-10 (Fig. 12 and Table 4).  Class IV Class IV consists of two mutants, ra406 and ra414, both of which arrest between the two and three fold stage of embryogenesis. Although most ra406 mutants arrest at the two fold stage (Fig. 10 A), the two fold arrest is not fully penetrant, since some mutant embryos can elongate to three fold. The majority of ra414 mutants cease elongation at the two fold stage, but some progress to the 2 1/2 fold stage (Fig. 10 D). Both mutants initiate body wall muscle contraction at the correct developmental time; however, contraction never progresses to more than minor twitching. In contrast, in wild type embryos soon after the initial body wall muscle contraction, the embryo begins to move vigorously within the egg. The body wall muscle twitching in ra406 and ra414 animals eventually ceases and the ra406 animals hatch as paralyzed L l larvae, whereas, ra414 animals fail to hatch. Analysis of ra406 and ra414 mutant embryos with antibodies to myosin indicates that prior to two fold these mutant embryos are indistinguishable from  68  Figure 10. Characterization of the Class IV mutants, ra406 and ra414. Both ra.406 (AC) and ra414 (D-F) have been examined by DIC and immunofluorescence microscopy with antisera to myosin, DM5.6 (B, E), and perlecan, GM1 (C, F). (A) DIC micrograph of a ra406 mutant embryo. In most cases, the process of elongation ceases at the two fold length, but this phenotype is not fully penetrant and some mutants can reach three fold in length. These mutants are capable of some minor body wall muscle contraction. The pharynx differentiates (arrowhead) and displays pharyngeal contraction. (B-C) ra406 mutant embryo double labeled with antisera to myosin (B) and perlecan (C). Myosin polarizes normally to the basal membrane of the muscle cells and assembles into double rows of A bands. However, once muscle contraction commences the body wall muscle cells separate from the hypodermis and gaps develop within the muscle quadrant (arrows in B). Perlecan is initially distributed evenly within the basement membrane underlying the body wall muscle. However, once muscle contraction commences the basement membrane becomes severely disorganized and large gaps develop within the basement membrane underlying the body wall muscle (arrows in C). In contrast, the distribution of perlecan in the basement membrane surrounding the pharynx appears normal (C). (D) DIC micrograph of a ra414 mutant embryo. ra414 mutants arrest elongation between 2 and 2 1/2 fold. These mutants also display some minor body wall muscle contraction. The pharynx in these mutants differentiates (arrowhead) but never displays any muscle contractions and ra414 mutant embryos fail to hatch. (B-C) ra414 mutant embryo double labeled with antisera to myosin (E) and perlecan (F). Myosin assembles normally in ra414 mutants. However, once muscle contraction commences the body wall muscle cells separate from the hypodermis and gaps develop within the muscle quadrant (arrows  69  in E). Perlecan is initially distributed evenly within the basement membrane surrounding the pharynx and the underlying the body wall muscle. However, once muscle contraction commences the basement membrane associated with both tissues becomes severely disorganized and patchy (F). In addition, large gaps develop within the basement membrane underlying the body wall muscle (arrows in F) as seen in ra406 mutants (arrows in C). Scale bar is 10 u\m.  70  71  wild type. The myosin filaments have polarized and have formed two double rows of A bands (data not shown; see Chapter 4). However, after body wall muscle has initiated contractions (two fold stage) these mutants begin to develop gaps in their muscle quadrants (Fig. 10 B, E). Similarly, perlecan distribution appears wild type in ra406 and ra414 mutant embryos prior to two fold (data not shown; see Chapter 4). However, after body wall muscle contraction commences large gaps are evident in the basement membrane and perlecan is patchy and disorganized (Fig. 10 C, F). It appears that after body wall muscle contraction initiates the body wall muscle and the underlying basement membrane are physically torn apart. This mutant phenotype is reminiscent of the type IV collagen mutants (see Chapter 5; Gupta et al., 1997). The distribution of perlecan around the pharynx of ra.406 mutant embryos is normal. In contrast, perlecan distribution around the pharynx of ra414 mutant embryos is patchy and disorganized. Pharyngeal muscle contraction has been observed in ra406 mutants but is not observed in ra414 mutants. This suggests that the gene mutated in ra414 mutants may have a developmental role in both the pharynx and the body wall muscle (see Chapter 5), whereas ra406 appears only to participate in body wall muscle development.  ra406 maps to the X chromosome and is balanced by stDp2. Three factor mapping places ra406 to the right of dpy-7 (Fig. 12 and Table 4). ra414 is the only mutant identified in the mutant screen of the central region of Chromosome II. Therefore, ra414 already has been mapped to the central region of Chromosome II. By the use of deficiency mapping, ra414 maps to the right of unc-4 (Fig. 12). Three overlapping deficiencies, eDf21, mnDf7] and mnDf83 fail to complement ra414 positioning ra414 to  72  Table 5. Deficiency mapping of ra414. Deficiency  Result  eD/21  Includes ra414  mnDfl2  Does not include ra414  mnDf28  Does not include ra414  mnDf29  Includes ra414  mnDf30  Does not include ra414  mnDf57  Includes ra414  mnDf58  Includes ra414  mnDf60  Does not include ra414  mnDf62  Includes ra414  mnDf68  Does not include ra4J4  mnDf71  Includes ra414  mnDf83  Includes ra414  mnDflOS  Does not include ra414  Table 6. Complementation results for ra414. Gene tested  Results  let-242  Complemented ra414  let-244  Complemented ra414  let-245  Complemented ra414  let-268  Failed to complement ra414  73  a small region between egl-43 and daf-19 (Table 5). Further complementation tests with several lethal mutants in this region (Herman, 1978; Sigurdson et al., 1984) reveal that ra414 is a new allele of let-268 (Table 6; see Chapter 4).  Class V Class V is composed of one mutant, ra409. Although ra409 mutant embryos cease embryonic elongation at two fold (Fig. 11 A), they are capable of some movement. ra409 animals display limited movement within the egg shell compared to similarly staged wild type animals. Additionally, some pharyngeal contraction takes place in ra409 mutants; however, the contraction is infrequent and appears irregular. After hatching, ra409 animals display some muscle contraction, but their movement is extremely uncoordinated. These mutants are capable of the most movement of any of the mutations analyzed in this study. An intriguing phenotype was observed in these mutant embryos when analyzed by immunofluorescence microscopy with antibodies to body wall specific myosin and perlecan. The body wall muscle quadrants of ra409 mutant embryos are twice as wide as the body wall muscle quadrants in similarly staged wild type embryos (Fig. 11 B). Additionally, the underlying basement membrane is twice as wide as wild type as indicated by perlecan staining (Fig. 11 C). This increase in size of the muscle quadrant and underlying basement membrane as seen by antibody staining of myosin and perlecan could result from one of two possibilities: an increase in body wall muscle cell number or an increase in the size of the muscle cells.  74  Figure 11. Characterization of the Class V mutant, ra409. (A) DIC micrograph of a ra409 mutant embryo. The process of elongation ceases at the two fold length but these mutants display some movement within the egg (compare with wild type in D). The pharynx continues to develop (arrowhead) and is capable of some muscle activity. (B and C) ra409 mutant embryo double labeled with a myosin antibody, DM5.6 (B), and a perlecan antibody, GM1 (C). (B) Myosin polarizes and assembles into thick filaments. However, the muscle quadrants in these mutants are twice as wide as the muscle quadrants in similarly staged wild type embryos (compare with wild type in E). (C) Perlecan is distributed evenly in the basement membrane surrounding the pharynx and underlying the body wall muscle. Interestingly, the basement membrane underlying the body wall muscle is wider than what is observed in wild type animals (compare with wild type in F). Scale bar is 10 um.  75  76  To investigate the possibility of an error occurring in an early cell fate decision leading to excess body wall muscle cells and the absence of another tissue, we investigated the integrity of the intestine and the pharynx, two major tissues of C. elegans. A l l of the intestine and part of the pharyngeal tissue arise from one cell of the 4 cell embryo (Schnabel and Priess, 1997). This cell is called the EMS cell. A large portion of the body wall muscle is also derived from the EMS cell. Therefore, if an error occurred in a subsequent cell fate determination step of the EMS cell lineage, it is possible that more body wall muscle cells could be determined at the expense of the intestine or pharynx. The intestine was examined by differential interference contrast (DIC) microscopy and by the autofluorescence of the gut granules. The gut granules and the morphology of the intestine appeared normal in ra409 mutant embryos. Next, the pharynx was investigated using DIC microscopy. Examination of ra409 animals with DIC microscopy revealed the presence of pharyngeal tissue that appeared to be fully differentiated (Fig. 11). However, unlike the wild type pharynx, the pharynx of ra409 animals struggles to pump and has an abnormal morphology. The pharynges of ra409 mutants fail to elongate and are half the length of a similarly staged wild type pharynx (Fig. 11). This examination indicates that both the pharynx and the intestine are present in ra409 animals and that pharyngeal and intestinal cells have not been mistakenly determined as body wall muscle cells. However, it is still possible that extra cell division could have occurred in the body wall muscle cell lineage producing an excess of body wall muscle cells.  To investigate the presence of additional cells in the body wall muscle quadrants, ra409 homozygotes were stained with antibodies specific to the major body wall muscle  77  myosin, M H C B (DM5.8). This myosin makes up 75% of the myosin in the body wall muscle cells with the remaining 25% consisting of the minor body wall muscle myosin, M H C A . Since M H C B is present in great abundance within the body wall muscle cells, the immunofluorescence is found throughout the muscle cell cytoplasm and one can distinguish the nuclei within the muscle cells by the exclusion of immunofluorescence. Therefore, a nuclear count can be carried out to determine the number of cells present in each muscle quadrant. This analysis established that ra409 homozygotes have 81 body wall muscle cells, the identical number as similarly staged wild type embryos (Sulston et al., 1983). This result has been corroborated with the use of a GFP expression vector that used the myo-3 promoter (body wall muscle myosin promoter) and contained a nuclear localization signal (A. Fire, pers comm.). Based on these results, the extra wide body muscle quadrants of ra409 mutants does not result from the presence of extra body wall muscle cells in the each quadrant. This suggests that the muscle cells are larger than wild type muscle cells. Since this is a novel phenotype, this mutant has been retained for further analysis (see Chapter 5).  ra409 maps to the X chromosome. Two factor mapping with a temperature sensitive allele of dpy-7(sc27ts) found that ra409 maps -10 map units (mu) from dpy-7. Two crosses have been conducted with strains carrying the duplications mnDplO and mnDp33. Both of these duplications serve as genetic balancers of the X chromosome and cover a region -10 mu from dpy-7. mnDp33 covers the left arm of the X chromosome and mnDplO covers the right arm of the X chromosome. mnDp33 is capable of balancing ra409 (Fig. 12) and this strain has been used for further genetic mapping. Three factor  78  mapping of ra409 positioned ra.409 just left of unc-20 (Fig. 12 and Table 4). A complementation test was conducted between ra409 and unc-78(e!217) mutants, which maps to the same genetic region as ra409. unc-78(eJ217) are slow moving animals because they have an abnormal muscle ultrastructure with large aggregates of thin filaments (Waterston et al., 1980). Both ra409 and unc-78(e!2J7) complemented one another indicating that these two mutations fall into two separate loci. For further analysis of this mutant see Chapter 5.  79  Figure 12. Genetic map positions of genes identified in this study. Genes identified in this study are in bold type set and are listed below each chromosome. The genes listed above each chromosome were used for mapping. Deficiencies and duplications used in this study are shown below chromosome II and X . Duplications are thick lines whereas the deficiencies are thin lines. The region balanced by mnCl is indicated by the thick line under chromosome II. Iet-268(ra414) is uncovered by eDf21 and mnDf83, placing let-268(ra414) in the region of overlap between these two deficiencies. Iet-268(ra414) was identified in the screen balanced by mnCl. ra409 is balanced by mnDp33 and is uncovered by syDfl. ra406 is balanced by stDp2. Chromosomes IV and V have not been included since none of the mutations identified in this study genetically map to these chromosomes.  80  81  Discussion The relatively small screen described here has identified several mutants that affect many aspects of muscle development and suggest that a larger scale mutant screen following the same parameters may prove to be very fruitful. The mutants identified have been placed into five different classes (Table 3). The Class I and II mutants appear to be required for proper myofilament assembly whereas the Class III and IV mutants appear to be important for the proper attachment of body wall muscle cells to neighboring muscle cells and to the basement membrane or hypodermis. The class V mutant appears to be required for more subtle organization of the myofilament lattice as well as muscle cell morphology (see Chapter 5).  The mutations isolated in this study have been placed into five different classes according to their body wall muscle defect. Class I contains one mutant, unc-52(ra401). This class represents the severe Pat class where the myofilaments fail to assemble and no body wall muscle contraction is detected. This mutant is the only true Pat isolated in the screen described here. In unc-52(ra401) mutants, myosin fails to polarize and does not organize into thick filaments (Fig. 6 B). Although these mutants fail to stain with perlecan antisera, a few small punctate structures are observed (Fig. 6C). This is most likely background signal since the sensitivity of the photo detector was enhanced several fold over normal conditions. However, we can not rule out the possibility that some protein product is produced since the molecular lesion in unc-52(ra401) mutants has not been determined.  82  The role of unc-52 in the development of the body wall muscle has been previously reported (Rogalski et al., 1993; Mullen et al., 1999). Lethal alleles of unc-52 were identified in by Williams and Waterston (1994) in their Pat screen, unc-52 encodes perlecan, a major component of the basement membrane. This molecule has been shown to be essential for the assembly of both the thick and the thin filaments in the body wall muscle, unc-52 was first identified by a class of viable mutants, which display a progressive paralysis phenotype (Waterston et al., 1980). During embryogenesis and early larval stages in unc-52 viable mutants, the myofilament lattice is normal and mutants are indistinguishable from wild type (Mullen et al., 1999). However, during the late larval stages in unc-52 viable mutant animals the myofilament lattice begins to deteriorate. One study has indicated that several UNC-52 isoforms are temporally and spatially expressed (Mullen et al., 1999). The viable mutants contain mutations in isoforms expressed in later developmental stages whereas the isoforms expressed in earlier developmental stages appear normal. The lethal allele of unc-52 isolated here and the lethal alleles of unc-52 isolated in the Williams and Waterston (1994) screen are null alleles since no unc-52 product is present (Fig. 6 C). This implies that isoforms essential for the early stages of development are not produced and this in turn leads to a failure of myofilament lattice assembly and embryonic arrest as Pats (Rogalski et al., 1993; Williams and Waterston, 1994).  Class II contains two mutants ra407 and ra408. Both of these mutants disrupt the organization of the thick filaments (Fig. 7 B , E) and ra407 disrupts the organization of the thin filaments (Fig. 8 B). In both mutants the myosin filaments appear to polarize  83  normally, but they fail to assemble into proper filaments. The thin filaments in ra407 mutant embryos also appear to polarize normally. Thin filament assembly has not been investigated for ra408. Therefore, thin filament organization in ra408 mutants is not known and should be investigated. Additionally, pharyngeal contraction appears to be defective in the ra407 mutants since no pharyngeal contraction is observed and the embryos fail to hatch, which is indicative of pharyngeal paralysis. However, examination of the pharyngeal thin filament organization indicates a normal filament structure (Fig. 8 B). Therefore, the failure of pharyngeal contraction is not a result of the lack of thin filament assembly. The pharyngeal thick filaments, however, have not been investigated so it cannot be concluded that all the myofilaments of the pharynx have assembled normally in ra407 mutants.  ra407 has been given the gene designation pat-13 because its mutant phenotype is reminiscent of the mild Pat class of mutants described by Williams and Waterston (1994). ra407 was found to be tightly linked to unc-40 (Fig. 12) and examination of the sequenced genome in this region suggested a candidate locus for ra407 that encodes tropomodulin. Tropomodulin is a pointed-end, actin-capping molecule involved in the regulation of actin filament length (Gregorio et al., 1995; Littlefield and Fowler, 1998). Mutations in tropomodulin may result in abnormal thin filaments and, in turn could be detrimental to the organization of the myofilament lattice leading to the phenotype observed in ra407 mutant embryos. The possibility was tested by genetically introducing four cosmids, one containing the tropomodulin ORF (McDowall and Rose, 1997), into the ra407 background. The cosmids failed to rescue the ra407 mutant phenotype  84  indicating that the ra407 mutation does not fall into any of the genes present on the 4 cosmids, including the tropomodulin ORF. However, it is difficult to determine if the proper dosage of each cosmid is present for rescue or if all 4 cosmids are present in the transgenic strain. Therefore, it maybe possible, although unlikely, that the tropomodulin ORF is mutated in ra407 animals. This could be tested by transformation rescue with only the cosmid containing the tropomodulin gene or a subclone containing only the tropomodulin gene. Also R N A interference (RNAi; see Chapter 4 and 5 for details; Fire et al., 1998) could be used to test if dsRNA directed to the tropomodulin gene can phenocopy the ra407 mutant animals when injected into wild type animals. It should be noted that pat-11 maps to the same region where pat-13(ra407) has been positioned. Complementation tests between these two genes were never conducted because the pat11 strain obtained from the C. elegans stock center (three separate occasions) never segregated the pat mutant phenotype.  Because ra408 is also reminiscent of the mild Pat class of mutants described by Williams and Waterston (1994), ra408 has been designated as a pat-14. ra408 falls in a region spanned by -20 cosmids and is closely linked to the cloned gene unc-32 (Fig. 12). This region is covered by several deficiencies, which would allow for the rapid cloning of this gene by using a PCR/Deficiency strategy (see Chapter 2 and 4). One candidate in the region is the C. elegans flightless ovtholog. flightless was initially identified in  Drosophila  where two different classes of flightless alleles have been isolated. One leads  to the atrophy of the flight muscles and the other more severe alleles affect the process of cellularization during embryogenesis (Campbell et al., 1993). flightless encodes a  85  molecule with an actin binding domain similar to gelsolin and an N-terminal domain consisting of a repetitive leucine-rich motif (Campbell et al., 1993). flightless has been implicated in the regulation of the actin cytoskeleton during Drosophila  development  (Straub et al., 1996). The failure of myofilament assembly in ra408 mutants and the predicted involvement of flightless in the regulation of the cytoskeleton makes the C. elegans flightless ortholog a good candidate gene for ra408. This possibility could be tested by transformation rescue with a clone containing the C. elegans flightless gene or by R N A i .  Class III contains two mutants, ra402 and ra404. These mutants have defects in the attachment of the body wall muscle to the basement membrane or the hypodermis. The myosin filaments polarize normally, but they do not form a double row of A bands as observed in similarly staged wild type embryos. Additionally, as development progresses, the muscle quadrants separate from the hypodermis as evidenced by myosin and perlecan staining (Fig. 9). Analysis of the thin filaments in ra402 mutants verify that the actin filaments polarize and assemble into a double row of I bands. This indicates that the thin filaments assemble normally. However, the myosin filament organization and the anchoring of the muscle quadrants to the hypodermis are defective in ra402 mutant embryos. The thin filaments have not been investigated in ra404 mutants but these mutants also have defects in myosin filament organization and the attachment of the muscle quadrants to the hypodermis.  86  Mutations in myo-3, the gene encoding M H C A (Waterston, 1989), and in the unc-45 gene, a gene encoding a novel molecule thought to be important for myosin filament assembly (Venolia et al., 1999; A o and Pilgrim, 2000), result in a similar phenotype as the class III mutants described here, myo-3 and unc-45 mutants do not form proper myosin filaments and the muscle quadrants separate from the hypodermis. Therefore, it is likely that the class III mutants are involved in the organization of the myosin filaments. How the organization of the myofilaments can result in a weakened link to the hypodermis is not known. However, it is possible that to allow for stable attachment of the body wall muscle to the hypodermis, the normal forces of body wall muscle contraction are required to stabilize the attachment structures in the hypodermis. Perhaps in the myo-3, unc-45 and class III mutants, the force of body wall muscle contraction required to form the proper attachment structure in the hypodermis is insufficient. Therefore, proper attachment structures are not formed and the body wall muscle separates from the hypodermis owing to the minor amount of contraction that is occurring in the body wall muscle. In congruence with this hypothesis, the attachment defect does appear to occur in the hypodermis since in the class III mutants, the basement membrane protein perlecan remains associated with the muscle quadrant and not the hypodermis. This indicates that the attachment structures within the body wall muscle cells are strong enough to anchor the muscle cells to the basement membrane and suggests that the weak link is located in the hypodermis. This could be investigated further by using the antisera, MH4, MH5 and MH46, which recognize different components of the hemidesmosomal-like structures of the hypodermis (Francis and  87  Waterston, 1991; Hresko et al., 1994), to determine if these structures have assembled properly in the myo-3, unc-45 and the class III mutants.  The class III mutants, ra402 and ra404, have been given the gene designation mua-11 and mua-12, respectively, for muscle attachment abnormal (Plenefisch et al., 2000). ra402 was genetically positioned just to the right of unc-75 on chromosome I and ra404 has been genetically positioned to the left of dpy-10 on chromosome II. The only Mua mutation found on chromosome I is vab-10 and its genetic location is fairly close to ra402. A complementation test between these two mutations indicates that they are two separate loci, mua-1 is found on chromosome II, however, its genetic position is to the right of dpy-10 whereas ra404 has been position to the left of dpy-10 indicating that these two mutations represent two different loci. ra402 and ra404 are localized to large genomic regions and no candidate genes have been identified.  Class IV contains two mutants, ra414 and ra406. These mutants are similar to class III mutants in that they appear to have defects in the attachment of the body wall muscle to the basement membrane or defects in hypodermal attachment to the basement membrane. However, the class IV mutants differ from class III mutants by the additional defect in the attachment of body wall muscle cells to each other. This is apparent by the large gaps that arise within the body wall muscle quadrants. Additionally, the initial assembly of the myosin filaments in ra406 and ra414 is indistinguishable from wild type whereas in the class III mutants the myosin filaments never assemble properly. The normal organization of the myosin filaments is short lived once body wall muscle contraction commences. At  88  this point the body wall muscle cells separate from the basement membrane or hypodermis and large gaps arise in the body wall muscle quadrants. This mutant phenotype most likely results from defects in the ability of the muscle cells to adhere to one another and to the hypodermis when body wall muscle contraction initiates. Since the myofilaments appear to assemble normally, the force of body wall muscle contraction is most likely similar to the force generated in wild type animals. Therefore, the weakened attachment of the muscle cells to one another and to the hypodermis in the class IV mutants cannot withstand the force of body wall muscle contraction.  In addition, ra414 mutants fail to display any pharyngeal contraction and fail to hatch. Examination of perlecan distribution in the basement membrane surrounding the pharynx of ra414 mutant embryos indicates that the perlecan distribution is patchy and disorganized (Fig. 10). This suggests that the ra414 mutation, in addition to affecting body wall muscle development, may affect pharyngeal muscle development. Unlike ra414 mutants, ra406 mutants do not display any pharyngeal defects (Fig. 10). The distribution of perlecan in the basement membrane surrounding the pharynx is normal and ra406 mutants are capable of pharyngeal muscle contraction. The body wall muscle of ra406 mutants, however, resembles ra414 mutants. Perhaps ra406 encodes a molecule involved in the stabilization of the basement membrane and/or hypodermal attachment structures for the body wall muscle quadrants, whereas ra414 encodes a molecule involved in the basement membrane and/or hypodermal attachment structures for the body wall muscle quadrants and for the stabilization of the basement membrane surrounding the pharynx.  89  ra.406 has been given the gene designation mua-13 and has been genetically positioned to the right of dpy-7 in the center of the X chromosome (Fig. 12). This is a large genomic region and no gene candidates have been identified for this locus. ra414 was genetically positioned to the left of unc-4 on chromosome II and was found to be an allele of let-268 (Fig. 12 and Table 6). The let-268 region of chromosome II has been thoroughly characterized, i.e. many cloned genes and several deficiencies span the region, which aided in the cloning of let-268 (see Chapter 4).  The other alleles of let-268, let-268(mnl89) and let-268(mnl98), were isolated in a large lethal screen of an extensive region of chromosome II (Sigurdson et al., 1984). let268(mnl89) and let-268(mnl98) have weaker phenotypes and would not have been detected as muscle mutants (see Chapter 4). As described above, the mutant isolated here, let-268(ra414), arrests development slightly after two fold and fails to display any vigorous body wall muscle contraction as observed in wild type. Also, large gaps in the body wall muscle quadrant arise after contraction commences (see Chapter 4). These defects are not detected in the weaker alleles of let-268 (see Chapter 4).  The Class V mutant, ra409, has a unique phenotype resulting in animals with body wall muscle quadrants twice as wide as wild type (Fig. 11). This affect appears to result from the abnormal morphology or abnormal growth of body wall muscle cells, as no additional body wall muscle cells are present and all tissue types analyzed are present. This indicates that cell fate decisions have occurred properly in ra409 animals and no  90  additional body wall muscle cell have been added to generate an extra wide muscle quadrant. ra409 has been genetically positioned just to the left of unc-20 on the X chromosome (Fig. 12 and Table 4). For further characterization of ra409 see Chapter 5.  The screen described here was designed to isolate mutants with defects in muscle development during embryogenesis. Eight mutants were found in this small scale screen which indicates the power of this screen. Two of the loci isolated have been previously identified; one with a known role in muscle development and the other without a known role in muscle development. The other six loci most likely represent newly identified genes that have an essential role in muscle development. However, mutants affecting the migration of the body wall muscle cells were not isolated. Since this was a small scale screen, it seems most likely that a large enough population was not examined to identify all types of muscle affecting mutants. Alternatively, this mutant class may be composed of genes that have redundant functions and would be difficult if not impossible to obtain.  For example, unc-6 is a gene known to be required for the dorsal and ventral migration of a subset of cells in C. elegans (Hedgecock et al., 1990). Examination of the body wall muscle cells in unc-6 animals reveals slight defects in body wall muscle cell migration (Hedgecock et al., 1990). However, these defects would be very hard to identify in a large mutagenesis screen since they are mild and not fully penetrant. Perhaps conducting a screen on a population of worms that are already compromised in cell migration would yield more severe defects in muscle migration that could be identified with greater ease in large scale mutagenesis screens. For example, a forward mutagenesis screen conducted  91  on a large population of unc-6 animals may identify genes involved in the migration of the body wall muscle cells from their initial lateral position to their final dorsal and ventral positions.  With -98% of the C. elegans genome sequenced (C. elegans Sequencing Consortium, 1998), the identification of genes involved in different biological process and their rapid cloning using R N A i and transformation technologies will provide invaluable insight into 'many of the different biological processes examined. Mutagenesis screens such as the one described here and other screens examining other developmental events in C. elegans will identify new components involved in these processes and provide a greater understanding of these developmental events.  92  Chapter 4. The let-268 locus of Caenorhabditis elegans encodes an enzyme essential for type IV collagen processing.  Background Basement membranes are thin polymeric sheets of specialized extracellular molecules found in all multicellular animals. They are important for supplying mechanical support to surrounding tissues, for separating different tissue types or adjacent interstitial extracellular matrix and for providing an interactive surface for cell adhesion, cell shape and cell migration (Yurchenco and Schittny, 1990). Basement membranes are composed of several molecules including, type IV collagen, laminin, perlecan and nidogen/entactin (Yurchenco and Schittny, 1990). The type IV collagen molecule is a trimer composed of two a chains, namely a 1 and a 2 [al(IV) a 2(IV)](Yurchenco and O'Rear, 1994). The 2  majority of type IV collagen consists of the central G - X - Y repeat domain, which folds into a triple helical structure (Kiihn, 1994). Laminin is a highly modular molecule, composed of an a, a (3 and a y chain, that is capable of many protein interactions (Timpl, 1996). Both laminin and type IV collagen are capable of forming macromolecular networks independent of each other. Perlecan is a large heparan sulfate proteoglycan that is also highly modular (Timpl and Brown, 1996). Nidogen contains three globular domains, two that are located at either end of the molecule, and one located in the middle (Timpl and Brown, 1996). This molecule is involved in bridging the laminin and type IV collagen macromolecular networks (Yurchenco and O'Rear, 1994; Timpl and Brown, 1996).  93  For a basement membrane to be established, these molecules must undergo a complex arrangement of interactions with one another and with transmembrane molecules, such as integrin and dystrogylcan, to ultimately form a stable lattice network (Timpl and Brown, 1996; Hynes and Lander, 1992; Henry and Campbell, 1998). Not only are these interactions important for basement membrane assembly but also for the structure of the surrounding tissues. For example, the importance of basement membrane molecules in maintaining mechanical support of skeletal musculature has been demonstrated by analyzing mutations in the laminin oc2 chain in mice. Mutations in the dy locus (encoding the laminin oc2 chain) cause a congenital muscular dystrophy which manifests as necrotic muscle fibers (Helbling-Leclerc et al., 1995). In another example, studies in C. elegans demonstrate that the basement membrane molecule perlecan is essential for the assembly of the body wall muscle myofilaments (Rogalski et al., 1993). Perlecan mutant animals fail to assemble myosin and actin filaments within the body wall muscle and arrest development as paralyzed embryos. The phenotype of animals lacking functional perlecan is identical to mutants in the extracellular matrix receptor integrin (Gettner et al., 1995; Williams and Waterston, 1994). These two examples illustrate the importance of basement membrane and transmembrane protein interactions during morphogenesis and for maintaining tissue integrity.  We are using the small soil nematode, Caenorhabditis elegans, to explore how the basement membrane interacts with myoblasts to determine muscle organization and anchoring during morphogenesis. In C. elegans, a thin basement membrane lines the  94  pseudocoelomic cavity and separates the body wall muscle cells from the hypodermis and nervous system (White et al., 1976). A similar basement membrane covers the gonad and intestine, while a thicker basement membrane surrounds the pharynx (White, 1988; Albertson and Thomson, 1976). Homologs of several mammalian basement membrane proteins have been identified in C. elegans (Kramer, 1997). Type IV collagen localizes to the basement membrane separating the body wall muscles from the hypodermis and localizes to the basement membranes surrounding the pharynx, the intestine and the gonad (Graham et al., 1997). Perlecan and osteonectin/SPARC are also localized in the basement membrane separating the body wall muscle from the hypodermis and the basement membrane that surrounds the pharynx (Hresko et al, 1994; Mullen et al., 1999; and Fitzgerald and Schwarzbauer, 1998). In addition, the sequenced C. elegans genome reveals four laminin chains, 2 a, 1 (3 and 1 y, and nidogen (C. elegans Sequencing Consortium, 1998). Several studies on mutations in genes encoding basement membrane molecules illustrate the importance of these components in morphogenesis (reviewed in Kramer, 1997; Moerman and Fire, 1997).  In C. elegans, type IV collagen and perlecan are both expressed by the body wall muscle cells (Graham et al., 1997; Moerman et al., 1996). Before these molecules are secreted into the basement membrane, they must pass through the secretory system where they undergo post-translational modifications. These modifications have been shown to be important for the function of basement membrane molecules. For instance, posttranslational modification of collagen by proline hydroxylation of procollagen is important for the stable folding of the collagen trimer at physiological temperatures  95  (Fessler and Fessler, 1978). Additionally, lysine hydroxylation is important for providing sites for collagen to form intermolecular crosslinks and sites for glycosylation (Kivirikko and Pihlajaniemi, 1998). Inhibition of both prolyl hydroxylase and lysyl hydroxylase with the iron chelator a, a ' dipyridyl leads to collagens that are under-hydroxylated and are unable to assemble into a functional trimer (Harwood, et a l , 1976). Under these conditions collagen fails to be secreted. This effect is thought to arise by blocking the prolyl hydroxylases. The hydroxylated proline residues are important for the stable formation of the collagen triple helices (Brodsky and Ramshaw, 1997).  In this study, we demonstrate that one of the mutations identified in the mutant screen described in Chapter 3, let-268(ra414), falls in the gene encoding the C. elegans lysyl hydroxylase. Lysyl hydroxylase is an endoplasmic reticulum (ER) enzyme that is essential for adding an hydroxyl group to the 8 carbon of lysine residues found in the Y position of the G - X - Y repetitive region of collagen chains, a structural motif found in all collagens (Kivirikko and Pihlajaniemi, 1998). These hydroxylysine residues can be further glycosylated within the E R and such secondary modifications are believed to be essential for the stabilization of intermolecular collagen crosslinks in the extracellular matrix (Kivirikko and Pihlajaniemi, 1998). The C. elegans lysyl hydroxylase is highly similar to the vertebrate lysyl hydroxylase, containing all essential motifs required for enzymatic activity, and it is the only lysyl hydroxylase found in the C. elegans sequenced genome (C. elegans Sequencing Consortium, 1998). In the absence of C. elegans lysyl hydroxylase function, type IV collagen is expressed; however, it is retained within the type IV collagen producing cells. This observation indicates that in let-268 mutants the  96  processing and secretion of type IV collagen is disrupted, let-268 mutants fail to complete embryogenesis and arrest at the two fold stage shortly after elongation commences. Analysis of let-268 mutants reveals that although functional body wall muscle is made, once contraction initiates the body wall muscle cells separate from the underlying hypodermis. These results indicate that type IV collagen is required in the basement membrane for mechanical support and is not required for the assembly of the myofilament lattice within the body wall muscle.  97  Results Genetic analysis of let-268 As described in Chapter 3, ra414 has been identified as an allele of let-268. Herman and colleagues (Sigurdson et al., 1984) previously identified two alleles of let-268 (mnl89 and mnl98) in a large lethal screen of chromosome II. Although both of these alleles are also lethal homozygotes, they have milder phenotypes than ra414. mnl89 arrests development during the first larval stage ( L l ) and mnl98 arrests development during late L l to L2. On occasion, let-268 (mnl98) homozygotes escape larval lethality and survive to become sterile adults. Heteroallelic combinations between the three let-268 alleles and deficiencies that delete the let-268 locus have been constructed to determine if ra414 behaves as a null or a hypomorphic allele. When either homozygous or over a deficiency, let-268 (ra414) animals arrest as two fold embryos (450-470 minutes after first cleavage). Also, when heteroallelic with ra414 or over a deficiency, mnl89 and mnl98 both arrest as two fold embryos (450-470 minutes after first cleavage). This suggests that ra4J4 represents the null state of let-268 whereas mnl89 and mnl98 are hypomorphs (Table 7). The sequencing of the relevant mutations supports this conclusion (see below).  let-268  encodes a lysyl hydroxylase  In order to identify a candidate gene for the let-268 locus, the genetic and physical map in the let-268 region were aligned. A PCR based approach utilizing the genomic sequence (C. elegans Sequencing Consortium, 1998) was conducted to determine the molecular break points of the three deficiencies that fail to complement let-268 mutations (see  98  Table 7. Genetic analysis of the let-268 locus. Genotype  Phenotype  let-268(ra414) homozygotes  Arrest development at the 2 fold stage of embryogenesis  let-268(mnl89) homozygotes  Arrest as L l larvae  let-268(mnl98) homozygotes  Arrest as L1-L2 larvae  let-268(ra414)/mnDf71  Arrest development at the 2 fold stage of embryogenesis  let-268(mnl89  Arrest development at the 2 fold stage of  )/mnDf71  embryogenesis Arrest development at the 2 fold stage of  let-268(mnl98)/mnDf71  embryogenesis Arrest development at the 2 fold stage of  Iet-268(mnl89)/let-268(ra414)  embryogenesis Arrest development at the 2 fold stage of  let-268(mnl98)Aet-268(ra414)  embryogenesis Arrest prior to embryonic morphogenesis  mnDjll/mnDfil  99  Material and Methods; Table 8). This strategy mapped the overlap between the two deficiencies, mnDf83 (Sigurdson et al., 1984) and eDfll (Raymond et al., 1995), to one cosmid, F52H3 (GenBank accession number Z66512; Coulson et al., 1995; Fig. 13 A and B; Table 8). The putative open reading frame (ORFs) from this cosmid were examined for a candidate gene. One of the seven putative ORFs on F52H3 encodes a procollagen lysyl hydroxylase (procollagen lysine, 2-oxoglutarate 5-dioxygenase precursor, F52H3.1, GenBank accession number CAA91321). A gene that encodes a procollagen processing enzyme seemed a likely candidate, since type IV collagen mutant animals have been reported to arrest development at a similar stage during embryogenesis as let-268 mutants (Sibley et al., 1994; Gupta et al., 1997).  To determine if let-268 encodes the C. elegans lysyl hydroxylase, wild type genomic D N A encoding let-268 was amplified using PCR and introduced into let-268 mutant animals. The PCR product included the entire lysyl hydroxylase gene plus 2.2 Kb of the 5' putative promoter region (see Material and Methods). This PCR product was microinjected into animals with the genotype unc-4(el20) let-268(ra414)/mnCl  [dpy-  10(el28) unc-52(e444)]. Rescued let-268(ra414) animals were recognized by the presence of Unc-4 animals. Although the PCR fragment was capable of rescuing the embryonic lethality phenotype, the rescued animals were sterile. To verify that let-268 encodes the lysyl hydroxylase, R N A mediated interference (RNAi) was used to determine the loss of function phenotype of F52H3.1. R N A i involves injecting dsRNA homologous to a gene of interest into wild type hermaphrodites and results in progeny with the loss of function phenotype of the gene (Fire et al., 1998). In this case, dsRNA  100  Table 8. Deficiency break points determined by P C R amplification of predicted genes. Gene Identification  mnDf83  ZK892.2 (nlt-1)  +  F52H3.7 (lec-2)  +  mnDjVl  eDfll  F52H3.2 F52H3.1 (let-268) C18D1.3  -  -  +  C18D1.1  -  -  +  ZK945.9  -  -  +  R05H5.3/.6*  -  -  +  + indicates that the open reading frame tested is present in homozygous deficiency. - indicates that the open reading frame tested is not present in homozygous deficiency. * The forward primer was designed to R05H5.6 and the reverse primer was designed to R05H5.3.  101  was made from the c D N A clone yk476hl, which contains the complete C. elegans lysyl hydroxylase ORF. Examination of the progeny from wild type hermaphrodites injected with this dsRNA resulted in animals with an identical phenotype as let-268(ra414) in approximately 75% of the progeny. These results suggest that let-268 encodes the C. elegans lysyl hydroxylase.  To further substantiate that let-268 encodes the C. elegans lysyl hydroxylase, the entire coding region plus most of the introns of the three let-268 mutants were sequenced. Single mutations were identified in the lysyl hydroxylase gene for each let-268 mutant. Consistent with EMS induced mutagenesis, all three mutants have G C to A T substitutions. ra414 has a substitution in the first base of the 5th intron which results in the loss of a splice donor site (Fig. 13 C). This would introduce a premature stop codon indicating that the let-268(ra414) mutation would result in a truncated gene product lacking the putative catalytic C O O H terminus (Pirskanen et al., 1996). Sequencing of the mnl98 allele revealed a change from a glycine to an aspartic acid at amino acid position 682 (Fig. 13 C and 14). This glycine is conserved in all lysyl hydroxylases identified (Fig. 14). Lastly, the mnl89 allele is a change from an aspartic acid to an arginine at amino acid position 668 (D668N). Surprisingly, the amino acid at this position in the mammalian ortholog is an arginine whereas in the C. elegans lysyl hydroxylase it is an aspartic acid (Fig. 14). This alteration in let-268 (mnl89) converts the C. elegans lysyl hydroxylase to be more similar to the wild type mammalian ortholog, yet it has deleterious effects and produces a lethal phenotype. Genetic rescue, R N A i and  102  sequencing of mutant alleles confirms that let-268 encodes the C. elegans lysyl hydroxylase.  The C. elegans lysyl hydroxylase gene encodes a predicted 730 amino acid polypeptide that contains a putative ER localization signal in the first 16 amino acids of the N terminus (Fig. 14, Hautala et al., 1992). LET-268 is highly similar to the mammalian lysyl hydroxylases. To date, three lysyl hydroxylases have been identified in humans (Hautala et al. 1992; Yeowell et al., 1992; Valtavaara et al., 1997; Valtavaara et al., 1998; Passoja et al., 1999) and each show approximately 45% overall identity and 65% overall similarity to the C. elegans ortholog (Fig. 14). Higher identity is seen in the C O O H terminus; amino acid 630-730 of the C. elegans polypeptide is 65% identical and 80% similar to the mammalian lysyl hydroxylases (Fig. 14). The C O O H terminus is believed to be important for the catalytic activity of lysyl hydroxylase (Kivirikko and Pihlajaniemi, 1998). Several key amino acids essential for the function of the mammalian lysyl hydroxylase (Pirskanen et al., 1996) are conserved in the C. elegans lysyl hydroxylase (Fig. 14). These include (1) a putative Fe binding site, consisting of 2+  His 659, Asp 661 and His 711 (Fig. 14), which is a necessary cofactor for lysyl hydroxylase activity (Pirskanen et al., 1996; Kivirikko and Pihlajaniemi, 1998) and (2) an arginine at position 721 which is essential for binding to the C-5 group of 2-oxoglutarate (a second cofactor required for enzymatic activity, Passoja et al., 1998). Analysis of the genomic sequence of C. elegans indicates that let-268 is the only lysyl hydroxylase in C. elegans (C. elegans Sequencing Consortium, 1998).  103  Figure 13. Genetic map position, genomic structure and GFP fusion protein of  let-268.  (A) Display of the genetic map of the central region of chromosome II of C. elegans. The three deficiencies, eDf21, mnDf71 and mnDf83, are indicated by thick solid line. (B) Contig map illustrating the cosmids in the let-268 region (see Table I). The physical break points of the deficiencies and the localization of the egl-43, let-268 and daf-19 open reading frames (ORFs) are shown (thin vertical lines). (C) Genomic structure of the let-268 ORF with the putative promoter region included. The exons are indicated by the black boxes and the introns and the 5' and the 3' U T R are indicated by the thin horizontal lines. The localization of the three identified point mutations are indicated. The exon/intron boundaries as well as the 5' and 3' termini of the wild type let-268 O R F were determined by sequencing two full length cDNAs, yk318c8 and yk476hl (Genbank accession number C63295 and C50032, respectively) verifying that indeed ra414 alters the splice donor site of the fifth intron. (D) Illustration of the GFP expression construct used in this study which includes the putative promoter region and the putative ER localization signal (see Fig. 3).  104  105  Figure 14. Amino acid sequence alignment of LET-268 with the three human lysyl hydroxylase isoforms. Three human isoforms of lysyl hydroxylase have been identified and are indicated by Hs L H 1 (GenBank accession number Q02809), Hs L H 2 (GenBank accession number AAD40977) and Hs L H 3 (GenBank accession number AAC39753). Identity is boxed and similarity is shaded. The N H terminal underlined region indicates the putative E R localization signal for all four lysyl hydroxylases. The * marks the amino acids important for Fe binding (Piskanen et al., 1996). The • indicates the 2+  amino acid important for binding the C-5 carboxyl group of 2-oxoglutarate (Passoja et al., 1998). The amino acid changes found in let-268 (mnl89) and (mnl98) are indicated at position 668 and 682, respectively. The mutation in let-268 (ra414) results in the loss of a splice donor at the end of the fifth exon (see Fig. IC) and would likely produce a truncated protein at the position indicated by the arrow.  106  LET-268 Hs LH3 H3LH2 Hs LH)  Em^v  L KlRk L E  P V L L|A]T T I TD L P • • M R ' MTSSGPGPRFt —P UPPAASIAlS DRPRSRDPVNPEK LL P .L MGGC T V K PQ L..LL I A l ? LIH P.W.N PC LGAD SEK PS S I PTDK  LfC T  L UUffl  limnjMt  A K  G0AK  - - - PEDN  LET-268 Hs LH3 Hs LH2 Hs LH1 LET-268 Hs LH3 Hs LH2 Hs LH1 LET-268 Hs LH3 Hs LH2 Hs LH1 LET-268 Hs LH3 H3LH2 Hs LH1  YpjYLyJsDIGKR  KllRQARlT MIGYIGIPIEMHKimK L K S V E | D KID D D Q L | Y J Y T ] M MIGTJ! I H Q | I V R QpY KYKDDDDDQL FYT|Rl I V Q Q WN LQDNDDDQL I GYAW V N R I G V A P I N L S K L _ A E[VVJ E i Q _ S _ S _ _ l E D I Y ] E [ D Q [ F ] K E D _ T P EC D E Y V L < " D E Y V L K Flf P E V V L K Fit r  G L E Y K E S E N Q D R R T  LET-268 Hs LH3 Hs LH2 Hs LH1  F S I K N Q FH E P H : A v H E KD _ _ K H H K A Q Y E E  LET-268 Hs LH3 Hs LH2 Hs LH 1  AY AY  AYN [ T M K. P I L RlNlYlAYD T L P K N T FJY T L P V IRINIIIAYIP T L M Y  I  L N G|Y I LNGAlL  F Q Q J L N G A Y  I F Q N LIDIGAIL  LN Y LGN Y L G J N I R R N S Q LGC Y H G N G iPlsjKlS - GGC Y | H G N G P T K LQjLNYLGNY Y[PN G « TlP E I N ] G N G P T K i L LNYQDGNY N. S W T Q D N GC I IH G N G P T K L CflL N V LGNV F E TlGC T Y | IPjR f\&I  v|p t  AE F J D T P K E K I A[T1 V | T | Y [ N ] K I Q P L D Y P P D R Y T L FF _ |H TN "|N TL D Y P , K E A L K L F I H NI L LIRILIHIY P I Q K H M R L F I H NlHlEl'  E YIP. • GGQP P E F D T Y D • A Y D Y H DEG tRSLUKG I GDEALlPjT  I  D [ F ) L Q K H G K S Y Y T K R V D S W P Q L Q D H FSAYfF Y [ F ] F D K A K H E I K T IK | F J L A Q H G S E Y Q S Y [ K J  F S  NG YTE IGDR E ARNlE A I E W N - K AR N V E F^jFlLMp|_D] EALSPG E A R | D M | A M I D ' tCRQD P EC E F Y F S L D A D EN LSQAlEARNMlGMp FCRQD EKCDY Y F S YD AD Y R M A N A D I A R N MJG A T D I L I C R QDR sIclT Y Y F S YlD AD R  A P | M I GQPlGKTlFTlN FWGAll AANlG V Y A R S E D YlMAMl M L S I R H G K L W S N L Y T R H G K L W S N LMTIRHGIRILWSN  LET-268 Hs LH3 Hs LH2 Hs LHt  F W G A L S ] P | D I | ] Y Y A R SE D Y Y ] E L FW G A L S P D G Y Y A R S E D Y Y D L T ] F W G A L S I A I D G Y Y A R S E D Y V I D I I I  EAMKDAYSYNK Nil LPQRDY SlElMN ERN Y  D P D M[S  MC K F ARD  N G  SG S D TJD P DMA F C K S FR ID K V R D K[L D P D M A L CR N AR EM|fi I S^KLkJ  LET-268 H 3 LH3 Hs LH2 HsLHI  LET-268 Hs LH3 HsLH2 Hs LH1  I Y A L  F LIGISGG F  I F QN I F QN  fl GlfjAlTT  LET-268 Hs LH3 Hs LH2 Hs LH1  LET-268 Hs LH3 Hs LH2 HsLHI  E FIG K R F L N S G _ 7 G T G K R F L N S G G F ... H IG K R 0 L N S G G F  FEH YJSJE KR L L YQSG_- -R L L flQKANH - - KYY  T Y T EG KWlPilP EMlWQ I F E P R E ' L H E A R - L | H | P 6 LWQ I YDWK E N P YDWK E NDtWQI io L W T E V I F I S I N P ElDWKE  FM F J M  H _ l M E P E H S R A L E G E G  sfRll F T EN TIKJAL A G K-  JpMElGFlG RWSD SNND lAKSGYENYP T RD I H M V[Y1DJQ|AJCPD Y Y[DJFP|LMTSEIR F | C Y E Q P C P D V Y W F PL L S E Q M C D E L Y [ A | E M E H YGQNYSGGRHE I A G G Y E N Y P T YD I H M V EQPC PD Y_WF P I F|S E K A C D E L Y E E M E H Y G KftYS G G K I SkJGYENYPTDD I HM YEfflPCPDVYWFP I F T I E I V A I C D E L Y E E M E H F_Q__ LlfijNN KIDJNIBJI QIGGYENVPT  H Hp  YBMDTIYIYR JlY D O J W L Q L L R TlY Y YG PMT E S L  YwjjDfr-11  N F[EJRE_H  N  G Y Y I H Q P Y E S N M M I F V V R Y • A R A V M N FYVRY • G F A L L N F V V_Y  F I A G Y Y T K R E_F_ I A|P|Y|T[IIK1V K|FJL L E _ I A|PJMIU J U L Y P g v v T P  D  PEE Q P SLR PHHD> Pp_Q PSLRPHHDO E R ) Q | R ) S LRPHHD*  -A Q F D L AIFYYRYIKIPIDIEQPSLIMIPHHP  LET-268 Hs LH3 Hs LH2 Hs LH1  A L N|K KlGlRlD YEGG G I V I R I Y I I R T N C | T V P A D E V Y AMMFlPGR I TH L H E G L|AJTT|KG ST FT_N[VjALNHK LDYEGGGCR F L R Y_C|V SSlPR KGWlA L J J H P G R L T H Y H E G L P T _ W G | STFTINIALNNY EpfFQlG G G C|K1 F L R Y N C S E S P R K G W S F I M H P G R L T H L H E G L P [ V T N GI  LET-268 Hs LH3 Hs LH2 HsLHI  T R Y I MY S F i M P ] T R Y I MV S F P T R Y I AVSF I D P TRY llAlVS FYlDP  A__R  VtelVlDYEGGGCR F L R Y N CIS!  vp  107  I IR A|P R K GWlT LlMH P G R L T H I Y I H E G L P T T I R I G I  let-268 is expressed in cells that produce type IV collagen  To examine the expression pattern of let-268, a D N A construct encoding a GFP fusion protein was injected into wild type animals. The construct consisted of 2 Kb of the 5' promoter region upstream of let-268 and all of the first exon, including the putative E R localization signal, and part of the second exon fused to the coding region of GFP (Fig. 13 D; see Material and Methods). Expression of let-268::GFP begins at the 1.5 fold stage and continues throughout development. let-268::G¥?  is detected in the body wall muscle  cells and in the glial like cells (GLR cells; White, 1988). let-268::GFP expression in the body wall muscle cells and the G L R cells coincides spatially and temporally with type IV collagen expression (Graham et al., 1997; data not shown). However, let-268::GFP was not observed in the distal tip cells of the gonad, the spermatheca or the vulva muscles where type IV collagen expression has been reported (Graham et al., 1997). Interestingly, the GFP signal produced from the let-268::GFP fusion protein appears to localize to a subcellular organelle that surrounds the nucleus in the body wall muscle cells and the G L R cells (Fig. 15). Since the mammalian lysyl hydroxylase 1 has been shown to be expressed in the E R (Kellokumpu et al., 1994), this is the most likely site of the let-268\.GFV fusion protein. The coincidental expression pattern of let-268 and type IV collagen in the body wall muscle cells and the subcellular localization of the GFP fusion protein to an intracellular compartment (possibly the ER) suggests that LET-268 does process type IV collagen within the cell prior to secretion.  108  Figure 15. Localization of let-268::G¥V  expression. (A) Wild type expression and  localization of let-268::GFP in two dorsal body wall muscle quadrants of a young adult. (B) Subcellular localization of let-268::G¥P  in two body wall muscle cells within an  adult. The signal is localized within the cell in a pattern surrounding the nucleus in all of the let-268::GFP expressing cells. Anterior is to the top in both (A) and (B). Scale bar is 20 u\m in (A) and is 5 u\m in (B).  109  B  A  m  t  •  v  i  f mm  __P' :  •  —  110  Type IV collagen secretion is absent in let-268 mutants Since let-268 encodes a lysyl hydroxylase and the function of lysyl hydroxylase is necessary for procollagen processing (Kivirikko and Pihlajaniemi, 1998), we examined type IV collagen distribution in let-268 mutant animals, let-268 mutants were analyzed using polyclonal antibodies to the type IV collagen a l (NW#155) and a2 (NW#68) chains (Graham et al., 1997; Gupta et al., 1997). These antibodies stain most of the basement membranes in the wild type embryo including the basement membrane underlying the body wall muscle, the pharynx and the intestine (data shown for NW#68; Fig. 16 A and B , Graham et al., 1997; Gupta et al., 1997). In let-268(ra414) mutant animals both a l and a2 type IV collagens accumulate in a subcellular pattern surrounding the nucleus in the body wall muscle cells and in the G L R cells and little to no type IV collagen is secreted into the basement membrane in these mutants (data shown for NW#68, Fig. 16 C and C ) . This staining pattern appears identical to the GFP expression of lysyl hydroxylase (Fig. 15). Staining of the hypomorphic alleles of let-268, mnl89 and mnl98 indicates that most of the type IV collagen is held up within the body wall muscle cells and the G L R cells. However, trace amounts of type IV collagen can be seen in the basement membrane (data shown for NW#68, Figure 16 E and E'). This is consistent with the weaker phenotype observed in the two hypomorphic alleles and suggests that the lysyl hydroxylase in these mutants may have some residual enzymatic activity.  Ill  Type IV collagen mutants arrest during embryogenesis at a similar time in development as let-268 (ra414). Many of the type IV collagen mutations isolated have missense mutations in the G - X - Y repeat and are predicted to alter the assembly of the heterotrimer (Engel and Prockop, 1991). A study by Gupta and others (1997) demonstrated that animals carrying mutations in type IV collagen failed to secrete this protein into the basement membrane. To compare the subcellular localization of type IV collagen in let268 mutants and the type IV collagen mutants, two different temperature sensitive type IV collagen alleles were examined for type IV collagen distribution. emb-9(hc70), which encodes the a l chain, and let-2(g25), which encodes the a2 chain, were grown at the restrictive temperature (25°C) and stained with the anti-type IV collagen sera (NW#68). We found that these mutants display a different staining pattern within the muscle cells and G L R cells than that observed in let-268 mutants. The subcellular localization of type IV collagen in the emb-9 and let-2 mutants is not in a pattern surrounding the nucleus but instead it appears packed in vesicles (perhaps the golgi apparatus) that are retained within the cell (data shown for emb-9, Fig. 16 D and D'). The mutations in emb-9(hc70) and let2(g25) alter a glycine in the G - X - Y amino acid repeats found in collagens and most likely disrupt the stable triple helix formation of collagen (Guo et al., 1991 and Sibley et al., 1994). Both emb-9 (hc70) and let-2 (g25) animals stained with either the antisera to the a l or the a2 chain appear identical, indicating that both type IV collagen chains are retained within the cell in both mutants (data not shown and Gupta et al., 1997). Therefore, these collagen chains are not secreted into the basement membrane. The comparison of subcellular localization between let-268 mutants and type IV collagen  112  Figure 16. Type IV collagen localization in wild type, let-268 and emb-9 mutant embryos. (A and B) A wild type embryo stained with type IV collagen antiserum (NW#68) is shown at two different focal planes. Figure (A) highlights the staining of the basement membrane separating the body wall muscle cell from the hypodermis (arrow) and figure (B) shows the basement membrane surrounding the pharynx (arrow) and the intestine. (C) let-268(ra414) embryo stained with the type IV collagen antisera (NW#68). Type IV collagen in these mutant animals is not secreted into the basement membrane but is retained within the collagen expressing cells. The staining pattern is seen surrounding the nucleus within the body wall muscle cells and the G L R cells. The G L R cells are indicated by the arrow whereas the rest of the staining is within the body wall muscle cells and the arrowhead indicates the cell that is enlarged in ( C ) . (D) emb9(hc70) embryo stained with the type IV collagen antisera. Type IV collagen also fails to be secreted into the basement membrane and is retained within the body wall muscle and G L R cells. The staining pattern within these cells is different from the cellular staining pattern seen in let-268(ra414) mutant animals. Instead of surrounding the nucleus, the staining pattern in emb-9(hc70) mutants appears as two or more clumps within the collagen expressing cells (D'). The arrowhead in (D) indicates the cell that is enlarged in (D') and the arrow indicates the G L R cells. (E) let-268(mn!98) stained with the type IV collagen antisera (NW#68). A similar staining pattern is observed between the different alleles of let-268. Iet-268(mnl98) shows strong staining within the cell in a pattern surrounding the nucleus. However, some type IV collagen is secreted in these mutant animals (arrow). Faint type IV collagen can also be seen surrounding the pharynx and the intestine in let-268(mnl98) mutant animals. The arrowhead in (E) indicates the cell  113  enlarged in (E'). ( C ) A cell enlarged from the let-268(ra414) mutant animal shown in (C). Type IV collagen fails to be secreted and is retained in the expressing cell in a ring like pattern that surrounds the nucleus. (D') A cell enlarged from the emb-9(hc70) mutant animal shown in (D). Type IV collagen fails to be secreted and is retained within the expressing cells. The staining pattern in emb-9(hc70) mutants is different than let268(ra414) mutants (compare C and D'). In emb-9(hc70) mutant animals, type IV collagen appears to form aggregates within the collagen expressing cells. (E') A cell enlarged from the let-268(mnl98) embryo shown in (E). A ring like staining pattern of type IV collagen is observed and this staining pattern is very similar to the staining pattern observed in let-268(ra414) mutants (compare E ' to C ) . However, some type IV collagen is secreted in the underlying basement membrane (arrow). The asterisk in (A)(E) indicates the anterior end of the embryos and the asterisk in (C')-(D') mark the location of the nucleus. Scale bar in (A)-(E) is 10 urn and the scale bar in (C')-(E') is 5p.m.  114  115  mutants suggests that type IV collagen is retained within different compartments of the secretory pathway.  Mutations in let-268 affect the stability but not the assembly of the body wall muscle and the underlying basement membrane Type IV collagen is a major component of the basement membrane and it interacts with many other components to form a complex network. It is therefore possible that the loss of type IV collagen could lead to the loss or mislocalization of other basement membrane components. Defects in the basement membrane, in turn, might negatively affect tissues intimately associated with it. To address these issues body wall muscle function and basement membrane integrity were examined in the let-268 mutants. Wild type and let268 animals were examined by immunofluorescence microscopy with antibodies specific to body wall muscle myosins (Miller et al., 1983) and to the basement membrane protein perlecan (Moerman et al., 1996) to determine if there were any defects in the body wall muscle architecture and basement membrane assembly or stability. This analysis was conducted on embryos prior to (Fig. 17) and after body wall muscle contraction had commenced (Fig. 18). Wild type embryos double labeled with myosin antibodies (DM5.6 and DM5.8, Miller et al., 1983) and perlecan antibodies (GM1, Moerman et al., 1996) display myosin filaments organized into double rows of A bands in each muscle cell in the four body wall muscle quadrants (Fig. 17 D and 18 A). Perlecan in these animals distributes evenly throughout the basement membrane underlying the muscle quadrants (Fig. 17 G and 18 E) and the pharynx (Fig. 18 E). Examination of let268(ra414) mutant animals double-labeled with the myosin and perlecan antibodies prior  116  to contraction (400 minutes after first cleavage) resembled the wild type staining pattern (Fig. 17 E and H). The myosin filaments are organized into double rows of A bands (Fig. 17 E) and perlecan is distributed evenly under the body wall muscle quadrants (Fig. 17 H). In contrast, after body wall muscle contraction initiates, let-268(ra414) mutants display disrupted muscle and basement membrane. In let-268(ra414) animals, the body wall muscle cells separate from the hypodermis resulting in severely disorganized myofilaments (Fig. 18 B). In addition, the distribution of perlecan is patchy and discontinuous throughout the basement membrane underlying the body wall muscle and the pharynx (Fig. 18 F). The let-268(ra414) animals also have large aggregates of perlecan within the body cavity (Fig. 18 F, arrows). The hypomorphic alleles of let-268, mnl89 and mnl98, were also analyzed by immunofluorescence and similar defects were observed just prior to hatching (data shown for mnl89, Fig. 18 C and G). Unlike let268(ra414) embryos, the hypomorphs are able to complete elongation and hatch. However, just prior to hatching defects in myosin organization and perlecan distribution are observed (Fig. 18 C and G). This observation is consistent with the weaker phenotype displayed by these alleles.  Since type IV collagen is a target of LET-268, we examined the body wall muscle and basement membrane of type IV collagen mutants to determine if similar basement membrane and body wall muscle defects arise in these mutants. The temperature sensitive mutants in the a l chain, emb-9 (hc70), and a2 chain, let-2 (g25), of type IV collagen were examined by immunofluorescence microscopy with antibodies specific to myosin and perlecan. Immunofluorescence staining of these mutants raised at the  117  Figure 17. Immunofluorescence microscopy of the body wall muscle and the underlying basement membrane of wild type, let-268(ra414) and emb-9(hc70) mutant embryos prior to body wall muscle contraction (-400 mpf). The same region of the dorsal muscle quadrant and underlying basement membrane is shown for all embryos. Animals in (AF) are double labeled with type IV collagen polyclonal antiserum (NW#68) and myosin monclonal antisera (DM5.6 & DM5.8). Animals in (G-I) are stained with the perlecan monoclonal antisera (MH3). Wild type embryos are shown in Panels (A, D and G). let268(ra414) mutant embryos are shown in panels (B, E and H). emb-9(hc70) mutant embryos are shown in panels (C, F and I). The wild type embryo has type IV collagen (A) and perlecan (G) distributed normally under the body wall muscle cells. The myofilaments in the wild type embryo are organized into two double rows of A bands (D). The let-268(ra414) mutant embryo retains type IV collagen within the body wall muscle cells and secretes little to no type IV collagen into the basement membrane (B). However, both the myofilaments appear to be organized into A bands (E) and perlecan is distributed normally under the body wall muscle cells (H). The emb-9(hc70) mutant, also, retains type IV collagen within the body wall muscle cells and fails to secrete type IV collagen into the basement membrane (C). Similarly, emb-9(hc70) mutant embryos are capable of assembling myosin into A bands (F) and perlecan is normally distributed in the basement membrane (I). Scale bar is 5 |im.  118  119  Figure 18. Immunofluorescence microscopy of the body wall muscle and the underlying basement membrane and the pharyngeal basement membrane of wild type, let268(ra414), let-268(mnl89) and emb-9(hc70) mutant embryos after body wall muscle contraction has initiated (-470-500 mpf). Embryos were double labeled with monoclonal antiserum to myosin (DM5.6, A-D) and polyclonal antiserum to perlecan (GM1, E-H). In wild type embryos, the myosin filaments are organized into two double rows of A bands in all four body wall muscle quadrants (A). Perlecan is evenly distributed under the body wall muscle cells and is distributed evenly around the pharynx (E). After contraction initiates, let-268(ra414) mutant embryos display gaps in the muscle quadrants where the muscle cell have separated from the hypodermis (arrowheads in B). Perlecan staining is patchy and is associated with the muscle cells that have separated from the hypodermis and the myofilaments have become severely disorganized (F). Also, aggregates of perlecan are observed in the body cavity of these mutants (arrow in F). let268(mnl89) mutant embryos complete elongation; however, prior to hatching the muscle cells start to separate from the hypodermis (arrowheads in C and G) and perlecan distribution becomes patchy (G). emb-9(hc70) mutant embryos resemble let-268(ra414) mutant embryos (compare with B and F with D and H). When the body wall muscle contraction initiates in emb-9(hc70) mutants, the muscle cells separate from the hypodermis and the myofilaments become severely disorganized leaving large gaps in the muscle quadrants (arrow heads in D). Additionally, the distribution of perlecan becomes disorganized and patchy (H). Perlecan appears to form large aggregates (arrow) in the body cavity and some perlecan appears to still be associated with the body wall muscle cells that have separated from the hypodermis. Scale bar is 12 p;m.  120  121  restrictive temperature (25°C) with the antimyosin and the antiperlecan antibodies looked strikingly similar to the let-268 mutants (compare Fig. 17 and 18, data shown for emb-9). In the type IV collagen mutants the myofilaments assemble and perlecan distributes normally (Fig. 17 F and I) but once contraction initiates the muscle cells separate from the hypodermis disrupting the basement membrane and the myofilament lattice (Fig. 18 D and H). This mutant phenotype is identical to the let-268(ra414) mutant phenotype. Detachment of the body wall muscles from the hypodermis in type IV collagen mutants has been previously reported (Williams and Waterston, 1994; Gupta et al., 1997). However, the normal assembly of the myofilaments prior to body wall muscle contraction has not been previously reported and indicates that type IV collagen is not required for the assembly of the myofilaments but is required for anchoring the body wall muscle to the underlying hypodermis once muscle contraction begins.  Interactions between perlecan and type IV collagen in vitro have been reported (Laurie et al., 1986; reviewed in Timpl and Brown, 1996). The relevance of these studies to what occurs in vivo is still unclear, including any possible interdependence of these major basement membrane constituents on each other for assembly. The data presented here indicate that perlecan is distributed normally in the absence of type IV collagen prior to body wall contraction (Fig. 17 H and I). Also, the myofilaments in let-268 and the type IV collagen mutants assemble normally prior to contraction (Fig. 17 E and F). This is evidence that perlecan is functioning normally in the absence of type IV collagen since it has been shown in C. elegans that perlecan is essential for the assembly of the myofilaments in the body wall muscle cells (Rogalski et al., 1993). To investigate the  122  localization of type IV collagen in the absence of perlecan, null mutations in the locus encoding perlecan, unc-52, were stained with type IV collagen antisera. Type IV collagen was evenly distributed in the basement membranes underlying the body wall muscle, the pharynx and the intestine (Fig. 19 A and B) in both wild type and unc-52 null mutant embryos. Therefore, the localization of type IV collagen was not affected by the lack of perlecan, indicating that perlecan is not necessary for type IV collagen localization and assembly. These data agree with reports that have demonstrated that type IV collagen is capable of forming a macromolecular network without the presence of any other basement membrane molecules (reviewed in Timpl and Brown, 1996) and that the formation of the basement membrane occurs normally in perlecan null mice (Costell et al., 1999).  Pharyngeal myofilaments appear normal in let-268 mutants Evidence thus far suggests that type IV collagen is essential for the stability of the basement membrane during body wall muscle contraction. The other major muscle tissue in C. elegans is the pharynx and it is also associated with a basement membrane containing type IV collagen (Graham et al., 1997). The role of type IV collagen in this muscle tissue is unknown. Examination of pharyngeal pumping in let-268(ra414) mutants and both emb-9(hc70) and let-2(g25) mutants raised at the restrictive temperature (25°C) demonstrates that these mutants fail to display any pharyngeal pumping before they arrest development. To determine if type IV collagen has a role in myofilament assembly in the pharynx in let-268(ra414),  emb-9(hc70) and let-2(g25) mutant animals,  the thin filaments have been examined by staining these mutants with FITC conjugated  123  Figure 19. Localization of Type IV collagen in wild type and unc-52 (null) mutant embryos. Wild type and unc-52 (null) mutant embryos were stained with antisera to type IV collagen (NW#68). In the wild type embryo, type IV collagen is distributed in the basement membrane surrounding the pharynx (arrowhead) and the intestine, and the basement membrane separating the body wall muscle cells from the hypodermis (arrowhead). In the unc-52 (null) mutant, type IV collagen staining appears normal. Type IV collagen is distributed in the basement membrane surrounding the pharynx (arrowhead) and the intestine, and in the basement membrane separating the body wall muscle cells from the hypodermis. Scale bar is 10 (xm.  124  125  phalloidin. The thin filaments in wild type, let-268(ra414), emb-9(hc70) and let-2(g25) appear very similar (data shown for wild type and let-268, Fig. 20 A and B). These filaments are organized into half I bands that are anchored to the basal and apical faces of the pharynx. This demonstrates that myofilaments of the pharynx can assemble in the absence of type IV collagen in the basement membrane. However, examination of let268(ra414), emb-9(hc70) and let-2(g25) mutant embryos in which body wall muscle contraction has initiated reveals disorganization of pharyngeal tissue (Fig. 18 F, H and 20 C, F). As mentioned above, contraction of the body wall muscle in these mutants results in the separation of the muscle cells from the hypodermis. The displacement of the muscle cells from the hypodermis disrupts the body cavity and compacts the pharynx. The pharynx normally commences contraction around 760 minutes after first cleavage whereas the body wall muscle initiates contraction at 420 minutes after first cleavage. In these mutants the pharynx and surrounding tissues are severely perturbed by the time pharyngeal contraction initiates. In animals that have not initiated body wall muscle contraction, the morphology of the pharynx as well as the pharyngeal myofilaments appear wild type (compare Fig. 20 D and E).  126  Figure 20. The musculature of the pharynx in wild type and let-268 (ra414) mutant embryos examined by staining with FITC conjugated phalloidin. The images in (A)-(C) have been included to show the entire embryo whose pharynx is examined. The images in (D)-(F) are the pharynges from the embryos shown in (A)-(C). (A) is a wild type embryo at the three fold stage. The body wall muscle quadrants, the intestine and the pharynx are labeled with phalloidin. (B) and (C) are two let-268(ra414) mutant embryos. Although these embryos are at very similar stage of development, the separation of the body wall muscle cells from the hypodermis has progressed slightly further in the embryo in (C) than in (B). (B) The body wall muscle cells can be seen starting to separate from the hypodermis (arrow in B). (C) The dorsal body wall muscle cells in this animal have completely separated from the hypodermis (arrow). As the muscle cells separate from the hypodermis, the body cavity of the worm becomes disorganized and the pharynx becomes compressed. The pharynx of the wild type embryo is shown in (D). The thin filaments are organized into half I bands (arrow) that surround the lumen of the pharynx. (E) The pharynx from the let-268(ra414) mutant embryo in (B) has properly organized thin filaments (arrow). (F) The pharynx from the let-268 (ra414) mutant embryo from (C) shows a pharynx that is extremely disorganized from the defects that arose from the body wall muscle cells separating from the hypodermis. There are no indications of a filament structure in this pharynx (F). Scale bar is 10 urn (A-C) and 5 um (D-E).  127  Discussion let-268  encodes a lysyl hydroxylase in C. elegans essential for type IV collagen  processing Lysyl hydroxylase functions as a homodimer that acts in the E R to convert certain lysine residues found in the G - X - Y repeats to hydroxylysine in collagen and collagen like domains found in other proteins (Kivirikko and Pihlajaniemi, 1998). Specifically, lysines found in the Y position of the G - X - Y repeat are possible substrates for lysyl hydroxylase. These hydroxylysine residues are further processed by the addition of a mono or disaccharide. These post-translational modifications are essential for the stabilization of intermolecular crosslinks between collagens in the extracellular matrix (Kivirikko and Pihlajaniemi, 1998). A human disease, Ehlers Danlos syndrome type VI, is generally caused by a deficiency in lysyl hydroxylase activity (Prockop and Kivirikko, 1995). The disease manifests as hyperextensible skin and joints, scoliosis and ocular fragility. Three lysyl hydroxylase genes have been identified in humans and in most Ehlers Danlos syndrome type V I patients one of these genes, lysyl hydroxylase 1, is deficient, let-268 encodes the only lysyl hydroxylase ortholog found in the C. elegans genome (C. elegans Sequencing Consortium, 1998) and its function is necessary for the completion of embryogenesis.  The results presented here indicate that the C. elegans lysyl hydroxylase acts on type IV collagen in the body wall muscle cells and in the G L R cells. In let-268 mutants, both type IV collagen chains, a l and a2, accumulate intracellularly within the cells expressing type IV collagen, indicating that the collagen is retained in the cell and is not secreted  129  into the basement membrane (Fig. 16). Previous studies using the iron chelator a, a'dipyridyl, which inhibits prolyl-4-hydroxylase and lysyl hydroxylase, have shown that collagen remains within the ER and is not secreted in cultured cells (Harwood et al., 1976; Bonfanti et al., 1998). The failure of collagen to be secreted has been attributed to the inhibition of prolyl hydroxylase. Hydroxyproline is thought to be essential for stable triple helix formation within the collagen trimer (Bromsky and Ramshaw, 1997). Here we demonstrate that the lack of lysyl hydroxylase alone induces a similar effect as a, a'dipyridyl on cultured cells, let-268 mutants result in the accumulation of type IV collagen in an area surrounding the nucleus, resulting in the failure of type IV collagen secretion. Therefore, our data indicates that the hydroxylysine residues, in addition to the hydroxyproline residues, are also required for the proper modification, folding and secretion of type IV collagen in vivo.  Not surprisingly, expression of a let-268::GFP fusion protein in the body wall muscle cells and the G L R cells indicates that LET-268 function is required within the cells producing type IV collagen. Since it has been shown in mammalian cells that lysyl hydroxylase functions in the ER (Kellokumpu et al., 1994) and our GFP expression pattern indicates a subcellular localization of let-268::GF? surrounding the nucleus, it is most likely that the C. elegans lysyl hydroxylase acts in the ER. However, further experiments are required to conclusively determine that LET-268 functions in the ER, such as immuno-electron microscopy with an antibody to LET-268. This subcellular localization of let-268::GFP is identical to type IV collagen retained within cells in let268(ra414) mutants. Therefore, it is probable that type IV collagen is not undergoing  130  proper modification in the E R causing it to accumulate possibly by "quality control mechanisms". Also, it is posssible that lysyl hydroxylase may function in "quality control" since prolyl 4-hydroxylase has recently been implicated in the E R retention of improperly processed collagen molecules. Prolyl 4-hydroxylase remains associated with incorrectly folded collagen molecule within the E R and only when the molecules are correctly processed and folded are they released by prolyl 4-hydroxylase and allowed to continue through the secretory system (Walmsley et al., 1999). Therefore, it is possible that lysyl hydroxylase may have a similar function in the ER.  let-268(ra414) mutants and type IV collagen temperature sensitive mutants raised at the restrictive temperature all failed to secrete type IV collagen into the basement membrane. However, the subcellular accumulation of type IV collagen differed between let-268 mutants and the type IV collagen mutants. Type IV collagen accumulated in a subcellular pattern surrounding the nucleus (perhaps the ER) in let-268 mutants, whereas type IV collagen localized in subcellular compartments (perhaps the golgi apparatus or vesicles) more distant from the nucleus in emb-9 and let-2 mutants. The mutations present in the type IV collagen mutants examined alter a glycine in the G - X - Y repeat which is predicted to affect heterotrimer assembly before secretion (Engel and Prockop, 1991; Guo et al., 1991; Sibley et al., 1994). This distinction in intracellular accumulation of type IV collagen suggests that quality control mechanisms at different points of the secretory pathway are functioning to ensure only properly modified and folded type IV collagen is secreted into the basement membrane. Interestingly, one class of type IV collagen mutants, represented by emb-9(st540) and emb-9(st545) (William and  131  Waterston, 1994) has a similar staining pattern to let-268 mutants (see Gupta et al., 1997) indicating that the mutation present in these alleles may affect the processing of type IV collagen within the ER. Unfortunately, Gupta et al. (1997) reported that the nucleotide alterations for these alleles were not identified even after the sequencing of the entire emb-9 genomic region. The identification of the molecular lesion in these mutants should provide insight on the intracellular processing of type IV collagen.  Besides the type IV collagens, C. elegans has an estimated 150 genes encoding cuticular collagens (Cox et al., 1984; C. elegans Sequencing Consortium, 1998; Johnstone ,2000). The cuticle is produced and secreted by the hypodermis. However, the  let-268::G¥P  reporter did not indicate any hypodermal expression. Previous reports have indicated that the cuticle is devoid of hydroxylysine residues whereas hydroxyproline is detected in substantial quantities (Cox et al., 1981). Nevertheless, a small quantity of hydroxylysine is detected in the dauer larvae cuticle (Cox et al., 1981). The dauer larvae is an alternative larval stage that arises due to environmental stress, e.g. when food is in low quantities (Riddle, 1988). The cuticle of the dauer larvae is especially thick and is partly composed of dauer specific collagens (Cox et al., 1981). A n analysis of animals in the dauer stage failed to indicate any let-268::GFP expression in the hypodermis of dauer larvae. The reporter construct contains only 2 kb of the upstream region of the let-268 ORF and does not contain the entire genomic region of the let-268 ORF. Therefore, it is possible that the let-268.:GFP reporter construct lacks important elements required for hypodermal expression or insufficient signal to be detected. In support of this possibility,  132  GFP expression was not observed in the distal tip cells, the spermatheca or the vulva muscles where type IV collagen expression has been reported (Graham et al., 1997).  Three mutations in the C. elegans lysyl hydroxylase gene have been identified. Of these three mutations, one behaves as a null mutation and the other two as hypomorphs. The let-268(ra414) mutation alters a conserved splice donor site and most likely results in a truncated lysyl hydroxylase missing a large part of the catalytically important C O O H terminus (Pirskanen et al., 1996). Homozygous let-268(ra414) animals behave as genetic nulls and little or no type IV collagen is secreted from the body wall muscle cells or from the G L R cells. The other alleles of let-268, mnl89 and mnl98, are both missense mutations and they behave as hypomorphic alleles. The affect the hypomorphic mutations have on the processing of type IV collagen and embryonic development is less severe than let-268(ra414).  A small amount of type IV collagen is secreted into the basement  membrane of the hypomorphic mutant animals and this presumably allows them to live to early larval stages whereas let-268 (ra414) mutants arrest before hatching. This conclusion is supported by our immunofluorescence analysis of the hypomorphs, where defects in the myofilament lattice and perlecan organization are detected just prior to hatching (Fig. 18). By contrast, in let-268(ra414) mutants these defects occur earlier in development. This suggests that the mnl89 and the mnl98 mutations in the C. elegans lysyl hydroxylase gene mildly affect the enzyme's function and therefore some type IV collagen is processed correctly and is secreted into the basement membrane. However, the amount of type IV collagen secreted into the basement membrane in the hypomorphs is not sufficient for complete larval development and they arrest during early larval  133  development. A large increase in muscle cell growth occurs during the larval stages and after hatching the movement of the larvae increases as it forages for food. So perhaps the low amount of type IV collagen in the basement membranes of the hypomorphs is unable to provide sufficient structural support at these early larval stages, resulting in larval lethality.  Distinct functional roles for perlecan and type IV collagen in myofilament assembly and muscle contraction in C. elegans The basement membrane underlying the body wall muscle is important for transmitting the forces of muscle contraction to the cuticle to allow the animal to move (Waterston, 1988). Also, the basement membrane has been shown to be critical for the assembly of the myofilaments within the body wall muscle. For example, in the absence of the basement membrane protein perlecan, myofilament assembly does not occur (Rogalski et al., 1993). Although type IV collagen is a major component of the basement membrane and has been implicated in muscle development (Williams and Waterston, 1994; Gupta et al., 1997), its role in this process remains elusive. From the analysis of emb-9, let-2 and let-268 mutants, this present study demonstrates that in the absence of type IV collagen in the basement membrane, the myofilaments undergo normal assembly (Fig. 17). However, once contraction starts the body wall muscle cells separate from the hypodermis, and the myofilaments become disorganized (Fig. 18). Consistent with observations that demonstrated the importance of perlecan for myofilament assembly (Rogalski et al., 1993), perlecan localization in type IV collagen mutants or let-268 mutants is normal prior to body wall muscle contraction (Fig. 17). After muscle  134  contraction initiates, perlecan distribution becomes disorganized and remains associated with the body wall muscle cells that have separated from the hypodermis (Fig. 18). This suggests that prior to contraction perlecan deposition into the basement membrane and the assembly of the myofilament lattice are normal. However, once muscle contraction commences, the weakened basement membrane lacking type IV collagen cannot withstand the forces of muscle contraction and the body wall muscle cells pull away from the hypodermis. In summary, the basement membrane appears to have two distinct roles in the development of the body wall muscle in C. elegans. In an initial phase, the basement membrane protein, perlecan, is required for the proper assembly of the myofilaments. In the second phase, type IV collagen is required to stabilize the basement membrane and anchor the body wall muscle and basement membrane to the underlying hypodermis.  The pharynx is also covered by a specialized basement membrane that is approximately two fold thicker than the basement membrane underlying the body wall muscle (Albertson et al., 1976). This basement membrane also contains type IV collagen and perlecan. Unlike the body wall muscle, the musculature of the pharynx does not require perlecan for its assembly (Mullen et al., 1999). Nor does the pharynx require perlecan for stability, since the pharynx is capable of pumping and no abnormalities arise in the pharyngeal myofilaments (animals die early in development so if pharyngeal contraction occurred longer maybe then an abnormality would arise). Analysis of the pharynx in type IV collagen mutants and in the let-268 mutants indicated normal assembly of the myofilaments of the pharynx. However, pharyngeal pumping is not observed in these  135  animals. This may occur because the body cavity in these animals is severely disorganized by the body wall muscle pulling away from the hypodermis, resulting in pharynges that are compacted and are therefore not capable of contraction. Analysis of let-268 and type IV collagen mutants after body wall muscle contraction shows that the morphology of the pharynges in these animals is extremely abnormal and this defect likely results in the failure of pharyngeal muscle contraction.  The basement membrane is a highly complex tissue composed of several multivalent proteins that have the potential to interact with multiple proteins to form a dynamic network structure (Timpl and Brown, 1996). In vitro, type IV collagen and laminin are capable of forming matrices independent of one another (Yurchenco and Furthmayr, 1984; Yurchenco et al., 1992). However, the function of these protein interactions in vivo is not well understood. For example, it is not known if type IV collagen is required for the incorporation of other components into the basement membrane or if type IV collagen requires other basement membrane molecules to become incorporated into the basement membrane in vivo. While type IV collagen has been reported to interact with perlecan in vitro (Laurie et al., 1986; reviewed in Timpl and Brown, 1996), it is not known if they require one another for proper localization in vivo. In this study we have shown that perlecan functions normally in the absence of type IV collagen in the basement membrane (Fig. 17). In addition, type IV collagen appears to localize normally in the absence of perlecan (Fig. 19), which suggests that perlecan is not required in the basement membrane for type IV collagen localization. This is consistent with prior reports that have shown that type IV collagen is capable of self assembly in vitro (Timpl  136  and Brown, 1996) and a recent report that illustrates the normal formation of the basement membrane in homozygous mice with a null mutation in the perlecan gene (Costell et al., 1999). Our in vivo results indicate that two major components of the basement membrane, type IV collagen and perlecan, are positioned and organized independent of one another.  This study demonstrates that let-268 encodes the C. elegans lysyl hydroxylase and that this enzyme is essential for type IV collagen processing and secretion. In addition, we have shown that type IV collagen is a critical component required for the stability of the basement membrane and for the transmission of the contractile force of body wall muscle to the cuticle. However, type IV collagen is not required for the organogenesis of the body wall muscle. This is the first report of a model organism with mutations in a lysyl hydroxylase gene. C. elegans may provide an excellent in vivo system for analysis of lysyl hydroxylase biology.  137  Chapter 5. Analysis of alpha spectrin function during muscle development in Caenorhabditis elegans  Background The spectrin-based membrane cytoskeleton is a highly ubiquitous structure found in metazoans. Members of the spectrin gene family and their associated proteins are widely distributed in most tissues in both vertebrates and invertebrates (Pesacreta et al., 1989; Bennett and Gilligan, 1993; Thomas and Kiehart, 1994). The spectrin-based membrane cytoskeleton first characterized in erythrocytes, has subsequently been studied in great detail in these cells. In the erythrocyte, spectrin forms a submembrane cytoskeletal network that is important for maintaining the structural integrity of the membrane and for maintaining cell shape (Lux and Palek, 1995). However, the role of the spectrin-based membrane cytoskeleton in non-erythrocytes has remained elusive. Functional models of the vertebrate spectrin-based membrane cytoskeleton have been proposed based on protein localization studies in cells and tissues and in vitro protein interaction studies. From its localization at cell-cell junctions and its association with cadherins, the spectrinbased membrane cytoskeleton has been implicated in the establishment and maintenance of epithelial polarity (Drubin and Nelson, 1996; Yeaman et al., 1999). Additionally, nonerythroid spectrin is thought to be involved in the organization of membrane subdomains, such as the organization of synaptic vesicles in presynaptic nerve endings (Sikorski et al., 1991, 2000). Recently, spectrin has been localized to the Golgi apparatus and has been implicated in Golgi function and vesicle transport (Devarajan et al., 1996; Fath et al., 1997; De Matteis and Morrow, 1998).  138  In vertebrate skeletal and cardiac muscle, spectrin is found associated with the cell membrane in structures known as costameres (Craig and Pardo, 1983). Costameres are vinculin and integrin containing structures that lie on the cytoplasmic face of the muscle cell membrane (sarcolemma) overlying the Z lines and M lines of striated muscle (Pardo et al., 1983; McDonald et al., 1995). Costameres have been implicated in linking the contractile apparatus to the sarcolemma and in the stabilization of the sarcolemma undergoing contractile forces. During rat and mouse embryogenesis, spectrin localizes to the sarcolemma in early muscle development (Zhou et al., 1998). Therefore, it is possible that spectrin is involved in organizing the myofibrils as the myotubes are becoming fully differentiated. In addition, the spectrin cytoskeleton is associated with neuromuscular junctions (Berthier and Blaineau, 1998). Although, the role of spectrin in neuromuscular junctions is not known, it has been shown that (3 spectrin colocalizes with voltage-gated N a channels at postsynaptic folds of mammalian neuromuscular junctions along with +  dystrophin (Wood and Slater, 1998). The voltage-gated N a channels are known to +  facilitate action potentials at the neuromuscular synapse (Kaplan et al., 1997). Therefore, spectrin may have a role in creating a membrane domain where synaptic proteins associate to form a functional synaptic complex. Further research is required to determine the roles of spectrin in striated muscle.  Spectrin belongs to a large super family of actin cross linking molecules that includes a actinin and dystrophin (for a review see Hartwig, 1995). The spectrin molecule is a rod shaped tetramer composed of two a and two (3 subunits (Bennett, 1990; Bennett and  139  Gilligan, 1993). Each subunit consists largely of a series of homologous 106 amino acid spectrin repeats (see Fig. 23). This repeated domain is predicted to consist primarily of a helices organized as a triple-helical coiled-coil (Spreicher and Marchesi, 1984; Yan et al., 1993). There are also non-repetitive sequences in each subunit that are responsible for some unique protein interactions of spectrin. For example, the a spectrin molecule contains an SH3 domain near the middle of the protein and two EF hand C a binding 2+  domains near the carboxyl terminus. P spectrin contains an actin binding domain in the amino terminus and a pleckstrin homology domain in the carboxyl terminus.  Several studies in Drosophila are beginning to provide new information on the function of the spectrin cytoskeleton in non-erthyrocytes. In Drosophila, three spectrin genes have been identified. One gene encodes an a spectrin and the other two genes encode P spectrins. The a and one of the P spectrins are highly similar to the non-erythroid spectrins (Dubreuil et al., 1987). However, the other  P spectrin, P  Hea  v  1 S r  unusual in that  it encodes a very large spectrin molecule -400 kDa (Dubreuil et al., 1990; Thomas and Kiehart, 1994) compared to the non-erythroid p spectrin which is -220 kDa (Dubreuil et al., 1987). The spectrins are widely expressed in most tissues of Drosophila, including the musculature, and associate with the plasma membrane (Pesacreta et al., 1989; Thomas and Kiehart, 1994). Spectrin exists as two populations in Drosophila,  one  isoform is composed of the a and P subunits and the other is composed of the a and the P avy He  subunits (Dubreuil et al., 1997). Both of these isoforms form tetramers and  resemble the vertebrate spectrin in appearance by electron microscopy and actin crosslinking activity (Dubreuil et al., 1987). In Drosophila epithelial tissues the two spectrin  140  isoforms associate with different membrane domains; the a(3 isoform localizes to the lateral and basal membrane and the a[3  Heavy  isoform localizes to the apical membrane  (Dubreuil et al., 1997). Mutations in the a spectrin gene reveal that in the absence of a spectrin the |3 isoforms can still associate with the membrane (Dubreuil and Yu, 1994). However, several striking development defects arise in these mutants implicating spectrin in the processes of cell growth, differentiation and specification (Lee et al., 1997a; de Cuevas et al., 1996; Thomas et al., 1998). The role of the spectrin cytoskeleton in Drosophila  muscle development has not been investigated.  This chapter describes the identification of ra409 as a mutation in the a spectrin gene, called spc-1, of C. elegans and provides a description of the role spectrin plays in muscle development. C. elegans a spectrin localizes to cell membranes in all tissues examined during embryogenesis. In the body wall muscle of larvae and adults, spectrin is localized to the I bands. Further characterization of the spc-l(ra409) mutant embryos reveals that in addition to extra wide body wall muscle quadrants (described in Chapter 3), the myofilaments are abnormally positioned during the early stages of muscle development. This implicates spectrin in the assembly and organization of the myofibrils. The wider cell shape of the body wall muscle is mirrored in the underlying basement membrane and in the organization of the attachment structures of the hypodermis. In addition, the thin filaments of the pharynx are disorganized, which leads to a poorly functioning pharynx. These results indicate that the spectrin-based membrane cytoskeleton is required in several tissues and is essential for coordinating cell shape and growth in the developing embryo, similar to a spectrin in Drosophila (Lee et al., 1997a).  141  Results spc-1  (ra409)  encodes the C.  elegans  a  spectrin  In order to identify the mutated gene in ra409 animals, the genetic mapping data presented in Chapter 3 and the genomic sequence (C. elegans Sequencing Consortium, 1998) have been utilized. ra409 is located on the left arm of the X chromosome between two cloned genes, unc-2 and unc-20, and is tightly linked to unc-20 (-0.47 map units to the left of unc-20; see Table 4 in Chapter 3). B . Wightman (Muhlenberg College) identified cosmid C15C7 as the positive clone containing the unc-20 ORF. unc-2 has been cloned and is found on cosmid T02C5 (Schafer and Kenyon, 1995). This places ra409 in an interval that contains 20 cosmids or -700 kb (Fig. 21). The genome sequence of this region has been examined for candidate genes that may help in narrowing down the 700 kb. Serendipitously, at the same time we conducted this analysis, Austin and colleagues identified mutations in a (3 spectrin gene of C. elegans (McKeown et al., 1998). This mutant, sma-1, fails to complete the process of embryonic elongation resulting in small animals. This phenotype is identical to the elongation defect observed in ra409 mutants (see Chapter 3; Fig. 22). However, unlike ra409 mutants, sma-1 animals are viable, sma-1 encodes a protein that is highly similar to Drosophila pHeavy  spectrin (McKeown et al., 1998), which is known to form heterotetramers with a  spectrin. This observation suggests that a likely candidate gene for ra409 may be a spectrin. Analysis of the genomic sequence between unc-2 and unc-20 revealed a gene encoding the only C. elegans a spectrin on the neighboring cosmid containing unc-20. This data is consistent with the tight linkage found between ra409 and unc-20.  142  Figure 21. The genetic and physical map of the left arm of the X chromosome. The genetic map is illustrated by the thick horizontal line, unc-2 and unc-20 were used for the genetic mapping of spc-1. The physical map is indicated by the contig map below the genetic map. Each thin black horizontal line represent a single cosmid and the name of the cosmid is indicated below the line. The unc-2 ORF is found on cosmid T02C5 (Schafer and Kenyon, 1995). The unc-20 ORF is found on C15C7 (B. Wightman, unpublished). The fax-1 ORF is found on C02A6 (Much et al., 2000).  143  < CN ©  0)  co W  S o  U  O  s o u  •  s fa  o  5J "  X  w  QJ  CN  o  Q co co  s  U  -  fi  cs m PH  S  u O  H  144  Figure 22. Morphological comparison of freshly hatched wild type, sma-1 mutant and spc-1 mutant larvae. Wild type animals elongate to almost 4 fold the length of the egg. sma-1 animals fail to elongate completely and have a distinctive rounded anterior end (nose). Similarly, spc-1 mutants fail to complete elongation and have a rounded nose. Scale bar is 20 \im.  145  146  To obtain further evidence that ra409 represents a mutation in the a spectrin gene, R N A mediated interference (RNAi) was conducted. This technique involves injecting dsRNA homologous to a gene of interest into wild type hermaphrodites and results in progeny with the loss of function phenotype of the gene (Fire et al., 1998). dsRNA was produced from a c D N A clone, yk205f3 (kindly provided by Y. Kohara, Mishima, Japan), containing ~1 kb of the a spectrin coding region and introduced into wild type animals (Fig. 23). Introduction of dsRNA from the a spectrin gene into wild type hermaphrodites produced an identical phenotype to ra409 homozygotes in approximately 50% of the progeny of injected animals. Some animals with the Sma (small') phenotype were observed in the progeny of injected animals, suggesting that a weak mutation in the a spectrin gene could result in animals resembling sma-1 mutants. To further substantiate that the ra409 lesion falls in the a spectrin gene, the ra409 phenotype was rescued by the introduction of a cosmid (MOIF 12) containing the full length a spectrin gene, followed by rescue with a 15 kb subclone containing only the a spectrin gene (DM#153; see Chapter 2). In addition, a transposon induced allele of spc-1 was isolated with the aid of J. Culloti (Mt. Sinai, Toronto). We have localized the transposable element (Tel) to the 13 exon of the a spectrin gene (Fig. 23). The T e l allele fails to complement spcth  l(ra409).  These results indicate that the ra409 mutation falls in the a spectrin gene. This  locus has been named spc-1 for a spectrin.  147  Figure 23. Gene and protein structure of a spectrin. (A) The genomic structure of the spc-1 gene. The exons are indicated by the black boxes and the introns are indicated by the thin black lines. The T e l insert is shown by the inverted triangle in the 10 exon. th  The c D N A clone (yk205f3) that was used for generating dsRNA is shown below the genomic structure. (B) Protein structure of the C. elegans spectrin family, spc-1 encodes the non-erythroid a spectrin (see Fig. 24), unc-70 encodes the non-erythroid (3 spectrin and sma-1 encodes the  P  Heavy  spectrin. The spectrin repeats are illustrated by the ovals.  Both SPC-1 and SMA-1 have SH3 domains. UNC-70 and SMA-1 contain an actinbinding domain in their N-terminus and a pleckstrin homology domain in their Cterminus. SPC-1 contains a C a binding domain in its C-terminus. The * indicates 2+  where the T e l would introduce a premature stop codon.  148  149  Figure 24. Deduced amino acid sequence of C. elegans a spectrin aligned to Drosophila a spectrin and nonerythroid a spectrin. The C. elegans spectrin (Ce_alpha; Genbank accession # AAB53876) is 61% identical and 76% similar in its entirety to the Drosophila a spectrin (Dm_alpha; Genbank accession # PI3395) and it is 57% identical and 72% similar in its entirety to the human nonerythroid a spectrin (Hs_alpha; Genbank accession # NP_003118), indicating that SPC-1 is the ortholog of human nonerythroid a spectrin. Identity is boxed and similarity is shaded.  150  Ce_alpha Dm_alpha Hs_alpha  MADSNDTHAPPEPVLEVPP M_|N F M D  Ce_alpha Dm_alpha Hs_alpha Ce_alpha Dm_alpha Hs_alpha  D R|L E|EA|R|_F Q Y F K R D A D E E|K L E D S R R F Q Y F K R D A D E K LEDSIYIR FQTFIFIQIRD AIEIE 101 87 84  |KTH_JA F E A E V G A H|A K T| ||A N K H Q A F E A E V [ S J A H S ] N [ A I MS K H Q A F E A E V Q A[N1S1GIA I VIK  Ce_alpha Dm_alpha Hs_alpha Ce_alpha Dm_alpha Hs_alpha Ce. _ alpha Dm_alpha Hs_alpha  VEIIILQRK FDIDFFILIKIE LGNIH1Q[YR1I[N~E1I V E V L Q R K FDE FQKDMA VEV LQIKIK FIEIE FQPTIDMA 251 237 234  Ce_alpha Dm_alpha Hs_alpha Ce_alpha Dm_alpha Hs_alpha Ce_alpha Dm^alpha Hs_alpha  401 387 384  Ce^alpha Dm_alpha Hs_alpha  451 437  Ce_alpha Dm_alpha Hs_alpha  501 N KQE A 487 A K Q E A 484 S, K Q E A  D E L A K D V A G A E A L LEISJHQEHKGE I o AR|AJDS FINQTTIASAJGQK L|_E|MS| E L A K D V A GA E A L L E R H O E H K G E I D A R E D S F WLIJJTIES GQ_,L L _ | R _ | H Y| NIA D E L AISID V A G A E A L LIDIR H Q E H K G E ID AfinlE D s F KIS A DIE S G(  SlAD  ElSlP E V Q E K L [A]A A EQTQE K LA 434 IA SID E VTRIEK L  D H | E K S S L L|G|LWE! R R I L Y E Q C M D L Q L F YRD TEQA_|TWH E N D k S S L LS L WE D l R R I L Y E Q C M D L Q L F Y R D T E Q A D T W I I S ElElR A AIL L L WE IRRIQQIYEQCMD LQ L F YRDTEQMDINIWI*  F L A N[TjD L G D S L DS V E_|L I KK HE D F E K S LA A Q E E K I N A L D E F A T K L F L A N E D L G D S L D S V E A L I K K H E D F E K S L A A Q E E K 1 K A LDLUFA TK L| F LIUN E D L G D S L D S V E A LIDX. K H E D F E K S LISIA Q E E K I TIA L D E F A T K L  Ce_alptia Dm_alpha Hs_alpha Ce_alpha Dm_alpha Hs_alpha Ce_alpha Dm_alpha Hs_alpha Ce_alpha Dm_alpha Hs_alpha Ce_alpha Dm_alpha Hs_alpha Ce_alpha Om_alpha Hs_alpha  L l A W I R E K E l Q _ _ A _ I S TNRGRD L I GVQN L I K K|QJQAJJ I lA E I A|AW I R E K E P I A A S T N R GJID L I GVONLJ,KKHQA[VL]AE I R E K EP  Ce_alpha Dm_alpha Hs_alpha Ce_alpha Dm_alpha Hs_alpha  I A AS TNRGIKIP L I G V Q N LMIK K H Q A LIQIA E I  I Q Q l G H F l L A l P E l I RD|K L A Q|L RDNWR I L K S K A E KlR Rl L | Q E Q W N Tl L K E K_|S Q Rm N Q K W E Ai LKIAIKASQRRI  LK pQPIFAISDE I RIQRIL v E EIGH FA AIEIDIVK A IK L 901 8S7  _ D L D D S L Q A H O Y[L_SJD A N E AIP_AJW^SJE K E P I V _ _ T D Y G K D E D S A E A L L K K H[R QD L D D S L Q A H Q Y F A D A N E AE S W M R E K E P I _ _ G _ _ D YG K D E Ds[_E A L L K KH -psi. — . . ., _ .. _ »» •« i _ i • ^ i u r \ u i . u O I O I C ry i _ i _ r \ r \ n c LJ IQD L l E l D S L Q A l Q l Q Y F A D A N E A E S W M R E K E P I V G S T D Y G K D E D S A E A L L K K H E w  Ce_alpha Dm_alpha Hs_alpha  884 951 937 934  Ce_alpha Dm_alpha Hs_alpha  1000 R E V S M K K G D V L T L L N|A S[N_IDWWK S N VE V N D R Q G FVP A A Y v KlRrirE[p~G| - TTAIQIQIH 987 R E V S M K K G D V L T L L N S l N N N K D W W K V E V N D R Q G F V P A A YQJKK I DIA]GIL[SASQ[Q 984 R E VITJMK K GDITIL T L L N SITIN K DWWK v E V N D R Q G F V P A A Y V K KFOD PIA QIS A SIR E  A LjLS L\\  - MGQ I J C V LA L Y D YQE YQE K S P T)p|v Vjoj i f f G K E C , v [ _ A L Y D Y _ ] E K S P, YQE T DIDIEIT G K E f D v L A L YD YQE K S P  IA  tJMs  m  151  s E | T _ D K [ Y ] Q R[L1M M L  049 A Q Q O V N 037JN~L)VDNH 034 N L L E E Q GlS  E Q  I D N Q T R  I T K E A G S V S  N [ Q \MS Q Y D N L L ] A | I R M K [ O J \ _ ] E LMHSIL LIE  'jofElLlR  E A N D L Aigjwl] 3V L V R E A 0 D L A Q W I  D L E Q V E V L Q K K D L E_)v E V t Q K K D  K F MILIFIR E A NIE1L[QIQW I F D D F K G | D L K A N E V R LjOE|MN F D D F N D|D L K A N E V R L F D O F Q K D L K A N E I S I R L  T A L T S V G Q T E V Q L T S . LGQJLE E D _ [ E _ | [GJLMA E E_]QA  '67  ITIEID L N1A|R W R I A  156 183  MQ|D LN EJ_V_"N] E T D S K T A S P W K S A R L M V H T V A T F N S I T I K E I L N E R W R I S H E V Q R F H R D V  D E T_|DW  ||Q|E  H E V Q R F H R D I D E T K D W I H E V Q R F H R D I A DETKIEIW I A A  A A L G D KI"V  AWVlA A WV)  RR N I  [ E A  L A N D V T G A E A L L E R H Q  IS L V L l l S D E  L A N D V T G A E A L [ D E R H Q A L  L E R H Q  _Ah IUJDQEIR  i Y A s p E i KIQIK U P t T W M S  A R E A  F L[AQ[ED  MS W M S A R E A F L M A W D INIW IVIAIA R E A F L NITIE D  - G O N V EU G|G]N V E / LIDISIV E /  I N EEK I l K G I L l K L l F A l E S l L l | K N N H Y _ ) S i|[p A I V I T R I K ^ I I N E [ _ K I A A L q T VjA D O L I AIQIN H Y A S jJN N LjVjD L | V ) D E|K Rll I NV l Q E E K I A A L QIAIF A D Q L I A U G I H Y A [ K G D I S S R _ | l  L HQlK RI F K L G E S Q T L Q Q F S R D A D E I E K R S _ ) L G _ _ ) QT L Q Q F S R D A D E  ILKIAQMI  L AIKIDV T G A E  L N S H HF Y A I A G P f n E N R [ D Q G V N E p ^ l K j L Q A N HI Y A S P E I K ] E | K | T E DJLjA K l A R  D O C L E L Q L F]Y R DC E ( AV_1 R W L O J L1_) G _ ] L | D ) L O L [ _ M R D C E|1 A WW V lvCl U R R I M M L D Q C L E L Q I R D C E Q A  !|L K E ] G [ L  L G S Ai  I O J L V S SlQlE I G I L V S S D E  I DISJRAIAJT F Q A F [ D ) Q F G N Q L I DARAGTFIGIA FEQFGNIEIL I D A R A G T F Q A F E Q F G K  L I K K H E D F D K A L I K K H E D F D K A L I K K H E D F D K A  Q L G S A, E E RlSlQlLlL G S A  E L N E CMN R [ L ) T S ( K A ] N NIFTK E | T ) N _ M W D Q I I TIKIS T A R K E L N j Q A _ 1 S S[T1G K RIAID QIR K  292 | E K L L D S Y D L Y j Q R F L S D Y R D L I 281 k K L L D S Y D L Q R F L S D Y R D L [ ] 333 A I K LIGID SIHID L Q R F L S D [ F ] R D L I IEIYJRTE ] E H R T E E H R T E  V Y G M M P R D  D L RSVQ [ _ L Q R K H E G V E R D L Y _ | H ID L I A I S V Q A L Q R K H E G f F l E R D L  )E T A N R L OE T AIEIR L H Q S I  f  V J _ _ Q E  F|G|R|D L R S V Q A L Q R K H E G V E R D L K A I  L G D K  A A L[_D K  L E Q V E V I O K K  E K R S K  L G E S Q T L Q Q F S R D M D E  I ENV_1T  tQE E _ J Y _ | D P  I E N W|]A  ITVEES  Y K D P  I EIATW M S  LQlTlAiS D I E S  Y K D P  KHQK|QJQA F E A E LHAN|Sj0R I K H Q K H Q A F E A E L 0 A N A D R I K H Q K H Q A F E A E L H A N A D R I  3 N N L I EJN A K JIGIN L I E K K Q 3NTS"1L I E I R G A  Y R J L K E Ai N K Q | K S _ ) M | A A V K D L | £ ) F W L < 3 E V L E L K E A NI K Q J RI T Y I I A A V K D L D F W L G E V lK L K E A N K Q i o T G llK D[F1P F W L < S 1 E v i  NfFlN  D Y G K D L A S V N L _ K K H|S)L L E A D I V A H_Ib~R . T j E D _ ) G K D L A S V Q|N „ K K HQ L_lE A D I V A H E D R JAISED Y G K D L A S V N L L K K H Q L L E A D l l s l A H E D R D(RY|D  GVJKlKlMAl  H Q F F R D I D D E E  R Y|E RfilC N T ) A A ] H  L H Q F F R D |JA]D E E L H Q F F R D I ^ D D E E  GIBJF Q KI I K I S I M A A I S  S W I K E K K L L V l S j S D D Y G R D L[PJG V O N L R ] R [ K ~ H ] R [ R ] I D T l E L A S H E P i Q V S L l V l K L Kl S W I K E K K L L V G S D . D Y G R D L T G V Q N L 0 K K H K R L E A E L _ J s H E P A I Q 1 A | V | Q E A|G| SW I K E K K L L V GSfilD Y G R D L T G V Q N L R K K H K R L I L AIAIHI E P A I QlGIVl L D TjGj  iG  L R S A A E A GV G E D] - - - V S N L G V F Slpl D N T I ISJK E l l  I  L D E S Q I R D LLTGN[R_)QR L AAJT RGqKjL D E S QlL A AlA R G Q R L H I E S  ifLlGlslD  E E E E A vitAN E K Q Q I N FIGIE N [M MA | G J V Q G L L K K H P l T l F l Q V l D l L E E E E E A W w I L X J E K Q Q L L J S V J E D Y G D ] S |M MA A V Q G L L K K H D A F E T D F T ) A ! E E E E A W W I N E K l M TILIV A J S E D Y G P I T L A I U Q G L L K K HIEIA F E T D F  IFT_A1_L I NK  GlD  _ . I J CC|D |l d G S T N _ | Q D| NSA  NLUS  YLQFMWKlSjDVVESW  I A l E K E t a Q V R S _ l D F G R D L S S VQLUL  N S A Y L Q F M W K A D V V E S W I N s AIFIL Q F[N!WK A D V V E S W I  D DIK E N^YlV R S DlEJF G R D L s f f l v Q T I I  T K Q E T F D A G  " E K E NIS L K TID DTYIG R D L S S V Q T L L T K Q E T F D A G  L N A F E | _ E G I Q R j l T{EjL K D Q L V S S E H O Qp S P A HlN I T A L K D Q L i NTAIS H AQ S P A 2030 ILIQIA F I Q I Q E G I A [ N I T A L K D Q L L A | A ] K | H | V IQSIKIA 1973  LTKQE|AjFDAG  L N A F E Q E G I  152  I E K R H  E0  IIDKRHGDIV  N T  lElAlRHlAS  N _ | A _ |  Ce_alpha Dm_alpha Hs_alpha  2035 2023 2060  Ce_alpha Dm_alpha Hs_alpha  2085 2073 2130  EIRALRDAHGEFC JE IRALRDAHAQFC IE IIKIA L RHIA HID AIFIF  Ce_alpha Dm_alpha Hs_alpha  2135 2123 2180  [LE1PJTWRNLQ_II IKEREJGJE L^_[E]HQ|RQE ENDKLRR E FAK|_AN AFH[AjWL T|NJT LEETWRNLQK I IHEREG|E LAKE RQEENDK LRJK E FAKHAN_1FHQWLTE T LEE TWRN LQK I IK E R IE LIQIK E RQEENDK L HOE F AIGIH A N A F H QWTTQIE T  Ce_alpha Dm_alpha Hs_alpha Ce_alpha Dm_alpha Hs_alpha  F_)Q I E E L Y Ll_lF A K K A S _ | F N S W F E N A E E D L T D P V R C N S L E FRQI E E L Y L T F A K K A S A F N S W F E N A E E D L T D P V R C N S_|E S H[F RIKVIEIDILIFIL T F A K K A S A F N S W F E N A E E D L T D P V R C N S LE  QE|MME|A TISIMMEG] jjYL LD_|S  S L S S A E|E|D F| SLSSAEADF| s LSSAJQIAD  G)G T L |SG[S)L CMVE EIS G T L  I K S FN VGP NP YTWF TM_TA IKSFNVGPNPYTWFTMEA I KS FIR]V|AS1NP YTWFTMEA  IN K G_T4R__ I E E_IGA] Rrav"5)LKK I EE LGA IMIRISIQ L K K I E{51L  2229 Ni L I L D N R Y T E H S 7 V G_IA~C_JW D Q L D Q L | A | M R M Q H N LEQQ 2217 H| L I L D N R Y T E H S T V G L A Q Q W D Q LDO L l S l M R M Q H N L EQQ 2280 A' L i LDNIKIYTEHSTVGLAQQWDQLDQ  Ce_alpha Dm_alpha Hs_alpha  I QARNQSGVSEE Al I Q A R N H|_G V jS_E_J LIGIMRMQHNLEQQ IQARNI' 2279 •R)E FSMMFKHFD KIEJKIJJG R LIPJHQIGJF K S C L R A L G Y D L P M VIDIE G Q PIEFFFIQ RHI 2267 'LKEFSMMFKHFDKDKS G_IL NHQE F K S C L R A L G Y D L P M V E E GQP DP E FE A I 2330 L K E F S M M F K H F D K D K S G R L N H Q E F K S C L RISIL GYDLPMVEE GJEIP DP E FE A I  Ce_alpha Dm_alpha Hs_alpha  2329 2317 2380  Ce_alpha Dm_alpha Hs_alpha  2379 f T _ l E E L YlAlN L T | P _ Q AIJLFJCII R|R M K PY|M[5)A l i ^ s r i l O G G f L D " 2367 T K E E L YCNLT K"D_|A D Y C VlQlR M K PFS E P RI^GEPLLIK D|A LD Y | 2430 T K E E L YIQIN L TIRIEQ A D Y C HIM K P YIVIBIG K^GWE L P  VDPNRDGYVLDLQEYMAFMISKE TE NLLIQSS EE I EMlA F R A L S1K[E1F|R I R PP YYVV D V V D P N R D G Y V S L Q E Yf_A FM I S K E T E N V Q S Q E E I E F R A l I T A A D[R P Y V DITIV D P N R DGIHIVS LQE YMA FM I STRIE TEN VIKISSEE I E A F R ATSISIEIG KIP Y V  vis  153  T_fFlc  Figure 25. SPC-1 is orthologous to Drosophila a spectrin and Human nonerthyroid a spectrin. Dot matrix alignments of SPC-1 (Ce alpha spectrin) with the Drosophila a spectrin (Dm alpha spectrin; A) and the Human nonerthyroid a spectrin (Hs alpha Fodrin; B). The addition diagonals that arise indicate the multiple spectrin repeats that make up the majority of these proteins.  154  -si •  200  ^X  400 600 800  .....x....  1400 1600 1800  ..A..5»; •-•  < X  V  ^  * *  -  ^  •••<••••'••-•••  ' 1',' •  \ ,  • S  •s\V "X  •••••'•<:••  "  •  •  -  •  X  (  -  2000  -  2200 2400  •"7  % i  1000 1200  .•i ::i... ..jw...•?». •v. \ ^ -  ; 200  *  <  00 240 60C8C 0 1000 12 00 14 00 16 00 18 00 2000 22  400  Ce alpha spectrin 200 400  ^ . - •> .>»^„* ^  1000 1200  ., - " ^ - '*  Jw  1800 «J"-*- |  2200  ".  '  -  ' %  »  200  i  i  -. • - -. ^ •  >  ; .v^ <  V  •  %  .'l.v.....:..  .  ^. ....?x...  r»  V. .  v  •s  h  ". ;Ai  -•  s  >  . X  > • •*  • >  . -  • -  V_  *v-  i  ^ ...  _  .". .....  +  . i U . .  i  1  2400  X  X  X  ^J,.....^  2000  f. Sy...  v ""  - v..  1400 1600  v  s  •  600 800  -  ' •>  i...:  >  j i H i T i —i — T 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400  Ce alpha spectrin  155  spc-1 encodes the only a spectrin in the C. elegans genome. The C. elegans a spectrin is 61% identical and 76% similar in its entirety to the Drosophila a spectrin and it is 57% identical and 72% similar in its entirety to the vertebrate non-erythroid a spectrin (Fig. 24 and 25). The vertebrate non-erythroid a spectrin and the C. elegans a spectrin both contain an SH3 domain and a calcium binding region. The overall conservation of the C. elegans a spectrin indicates that SPC-1 is the true ortholog of vertebrate non-erythroid a spectrin (Fig. 24 and 25).  Myofilament organization in spc-l(ra409) animals As described in Chapter 3, spc-1 (ra409) mutants have muscle quadrants that are twice as wide as wild type. To investigate the organization of the myofilaments within spcl(ra409) animals, mutant embryos were stained with a body wall muscle myosin specific antiserum, DM5.6. In addition to the muscle quadrants being larger, a closer examination of myosin organization within the body wall muscle cells reveals that the myosin filaments in these animals are abnormally oriented. In wild type the A bands (the striation of the myosin filaments) are oriented at an oblique angle (6 degree angle) to the longitudinal axis of the nematode. In spc-l(ra409) mutants this oblique striation is exaggerated to an almost 20 degree angle especially in the anterior muscle cells (Fig. 26 B). Also, there is a large gap between the myofilament lattice in neighboring muscle cells within the same muscle quadrant that is not observed in the muscle quadrants of wild type animals. Normally, in wild type, the myofilament lattice in neighboring cells is closely associated (Fig. 26 A). However, in spc-l(ra409) animals the myofilament lattice is not tightly associated and there is a several micron gap between the myofilament lattice  156  in neighboring cells (Fig. 26 B). The abnormal myosin filament positioning and the large gap in the center of the muscle quadrants are observed in the initial stages of muscle development (comma to 1 1/2 fold stage; Fig. 26 D) indicating that the initial assembly of the myosin filaments is abnormal in spc-l(ra409) mutants. Interestingly, the body wall muscle cells of spc-l(ra409) mutants still only contain 2 A bands (the normal amount, see Fig. 26 A , B) even though the cell size could allow for 3 or more A bands.  This defect is not only associated with the thick filaments, as similar defects are seen when the thin filaments of spc-l(ra409) mutants are stained with FITC labeled phalloidin. Similar to and consistent with the thick filaments, the thin filaments are abnormally oriented within the body wall muscle cells (Fig. 27). This filament organization suggests that the myofilaments are not properly assembled at the body wall muscle membrane. In addition, a similarly sized gap, that has been observed in the myosin analysis, is present between the myofilament lattice in neighboring body wall muscle cells within the same muscle quadrant (Fig. 27 B). Since the spectrin-based membrane cytoskeleton has been reported to be involved with cell polarization (Hu et al., 1995; Drubin and Nelson, 1996) and the body wall muscle cells of C. elegans are highly polarized cells, perhaps this abnormal arrangement of myofilaments in spc-1 (ra409) mutants is because there is a loss of cellular polarization. To determine if there are defects in the polarization of the body wall muscle cells in spc-l(ra409)  animals, mutant  embryos labeled with FITC-phalloidin were examined by confocal microscopy and computer generated cross sections were constructed using the software package N I H image (available at http://rsb.info.nih.gov/nih-image). In wild type, the thin filaments are  157  Figure 26. Myosin filament organization in wild type, spc-l(ra409) and sma-l(rul8) embryos. Wild type (A), spc-l(ra409) (B and D), sma-l(rul8) (C) embryos were labeled with DM5.6, an antimyosin antibody. The right dorsal muscle is shown in (A-C) with its anterior up. In the wild type, the myosin filaments are organized into a double row of A bands that run -parallel with the longitudinal axis of the embryo. In both spc-1 (B and D) and sma-1 (C) mutant embryos the myosin filaments are organized into double rows of A bands but the filaments are abnormally oriented. The myosin filaments in these mutants are almost oriented at a 20° angle to the longitudinal axis of the embryo. Additionally, the muscle quadrants of the spc-1 (B and D) and sma-1 (C) mutant embryos are twice as wide as normal (compare to A), z, y and z are indications of body wall muscle quadrant width for wild type (z), spc-1 (y) and sma-1 (z). Arrowheads in (B) and (C) indicate large gap between the myofilament lattice in neighboring muscle cells. (D) is a spc-l(ra409) mutant embryo at the comma stage (390 mpf; see Fig. 2 and 3). This image illustrates that at early stages of muscle development the myosin filaments are positioned abnormally in spc-l(ra409) mutants (especially in the anterior). Scale bar in (A-C) is 5 fim and 10 p:m in (D).  158  sixy  iBmpun^i^uoq  159  Figure 27. Thin filament organization in wild type and spc-1 (ra409) embryos. Wild type (A) and spc-l(ra409) (B) were labeled with FITC-phalloidin. In wild type embryos, the thin filaments are organized into a rows of I bands that run approximately parallel with the longitudinal axis of the embryo (A). In spc-1 mutant embryos the thin filaments are organized into rows of I bands but these filaments are abnormally oriented. The thin filaments in these mutants are almost oriented at a 20° angle to the longitudinal axis of the embryo (B). Additionally, the muscle quadrants of the spc-1 (B) are wider (compare to A). Similar defects are observed in thick filaments (see Fig. 25). Scale bar is 10 Jim.  160  s i x y [tmipnjiguoq  161  restricted to a small region adjacent to the neighboring muscle cell membrane facing the hypodermis (Fig. 28 B and 29 D, E). The thin filaments in spc-1 (ra409) embryos are localized to the basal half of the muscle cell but they fail to become tightly associated with the membrane adjacent to the neighboring muscle cell or to the hypodermis (Fig. 29 F). These results indicate that the body wall muscle cells of spc-1 (ra409) mutants have polarized. However, the spatial restriction of the myofilaments to the basal membrane face adjacent to the neighboring body wall muscle cell has not occurred.  Basement membrane and hypodermis organization in spc-l(ra409) mutants As stated in Chapter 3, the localization of the basement membrane protein perlecan was examined in spc-l(ra409) embryos. The distribution of perlecan matched the expansion of the body wall muscle quadrants and results in a basement membrane that is twice as wide as wild type (Fig. 30 A , B). Aside from the larger size of the basement membrane, the distribution of perlecan appears wild type, perlecan is evenly distributed under the muscle quadrant. This indicates that the neighboring muscle cells must be in contact with one another since perlecan fails to be stabilized without close association with the muscle cell membrane (Moerman et al., 1996). Therefore, the gap observed in the muscle quadrants by immunofluorescence with a myosin antibody or with phalloidin must arise in the failure of the myofilaments to be completely restricted to the basal membrane face. This is further evidence that the spectrin-based membrane cytoskeleton is important for the organization of the myofilaments. The wider basement membrane in spc-1 (ra409) mutants, led to the speculation that other components involved in the attachment of the body wall muscle quadrants to the cuticle may also be affected in spc-l(ra409)  162  mutants.  Figure 28. Body wall muscle cells are highly polarized. (A) Wild type embryo at the 2 fold stage labeled with FITC-Phalloidin. (B) A computer generated cross section of the embryo shown in (A). The vertical line shown in (A) represents the position of the cross section. In (B) two body wall muscle cells of the left dorsal quadrant are shown. In this figure the highly polarized nature of the body wall muscle cell is readily apparent. The thin filaments are localized to the basal surface of the muscle cell that is in direct contact with an adjacent muscle cell. Small arrows indicate the thin filaments and the large arrows indicate the muscle cell bodies. The arrowhead indicates the cell boundary between the two neighboring body wall muscle cells. Scale bar is 10 UMTI (A) and 5 [im (B).  163  164  Figure 29. The myofilaments are not properly localized to the basal face of the body wall muscle cells in spc-1 mutants. FITC-phalloidin staining of wild type and spc-1 (ra409) embryos. A wild type embryo just prior to hatching is shown in (A) and the cross section (indicated by the white line) of this embryo is shown in (D). A wild type embryo at the two fold stage is shown in (B) and the cross section (indicated by the white line) of this embryo is shown in (E). A spc-l(ra409) mutant embryo is shown in (C) and the cross section (indicated by the white line) of this embryo is shown in (F). The myofilaments of the wild type embryos have localized to the basal membrane face of the body wall muscle cells near the neighboring muscle cell (arrows in D and E; see Fig. 27). The myofilaments in the spc-1 mutant embryo appear to not be properly localized to the basal face of the muscle cell membrane (arrow in F). Also, note the large gap between the adjacent muscle cells in the spc-1 mutant (arrowhead). Scale bar is 10 urn (A-C) and 5 urn (D-E).  165  166  Therefore, the hypodermal attachment structures associated with the muscle cells were investigated in spc-1 (ra409) mutant embryos using the antiserum M H 5 . Although the antigen has not been molecularly identified, the MH5 antiserum recognizes a hemidesmosomal-like structure in the hypodermis that assembles adjacent to the body wall muscle cells in the dorsal and ventral hypodermis (Fig. 1 and 2; Francis and Waterston, 1991). The MH5 antigen is thought to be a component of the attachment structures required for transmitting the force of body wall muscle contraction to the cuticle. In wild type, these hemidesmosomal-like structures mimic the spatial and temporal distribution of the muscle components and are tightly associated with the hypodermal membrane adjacent to the muscle cells where the myofilament lattice is positioned (Fig. 1 and 2; Francis and Waterston, 1991; Hresko et al., 1994). Staining with M H 5 reveals that the hypodermis in spc-l(ra409) animals closely mimics the myofilament lattice, the hemidesmosomal-like structures are also twice as wide as normal (Fig. 30 C, D). This indicates that the complete linkage from muscle to the cuticle is affected in these mutants. This analysis was repeated using the M H 4 antiserum that detects intermediate filament like antigen in the hemidesmosomal structures of the hypodermis (Fig. 1 and 2; Francis and Waterston, 1991). Identical results were obtained indicating that two components of the hemidesmosomes are affected in spc-1 (ra409) mutants.  The hypodermis was examined in further detail to determine if there were any other defects in the cytoarchitecture. Since spc-l(ra409) mutants fail to complete elongation and the process of elongation during embryogenesis is thought to be coordinated by  167  Figure 30. Analysis of perlecan and the hypodermal hemidesmosomes in wild type and spc-l(ra409) mutant embryos. (A) wild type and (B) spc-1(ra409) mutants stained with GM3, an antibody that recognizes the basement membrane protein perlecan. (C) wild type and (D) spc-l(ra409) mutants stained with M H 5 , an antibody that recognizes an unknown component of the hypodermal hemidesmosomes. In wild type embryos perlecan is evenly distributed under each of the body wall muscle quadrants (A). Similarly, in spc-l(ra409) mutants perlecan is evenly distributed under each of the body wall muscle quadrants (B). However, the spatial distribution of perlecan in spc-l(ra409) mutants is twice as wide as normal (arrowheads in A and B), reflecting the defects observed in the body wall muscle (see Fig. 25 and 26). In wild type embryos the hemidesmosomes have assemble under each of the body wall muscle quadrants (C). In spc-l(ra409) mutant embryos the hemidesmosomes assembled normally under each body wall muscle quadrant (D). However, similar to the defects seen with perlecan (A and B) and the body wall muscle (see Fig. 25 and 26), the hemidesmosomes are assembled in an area twice as wide as normal (arrowheads in C and D). Scale bar is 10 p:m.  168  169  circumferentially oriented actin filaments within the hypodermis (Priess and Hirsh, 1986), the organization of the actin cytoskeleton in the hypodermis was examined in spcl(ra409) mutant embryos. Examination of spc-l(ra409) mutant embryos labeled with FITC-phalloidin indicated that the circumferential oriented actin filaments of the hypodermis assemble and are present in a pattern similar to that observed in wild type embryos (Fig. 31). Another event that occurs in the hypodermis during embryogenesis is that several hypodermal cells fuse to create a large syncytium cell (hyp7) that covers a large part of the embryo (Sulston et al., 1983). To investigate the occurrence of cell fusions in the hypodermis, spc-1 (ra409) mutant embryos were labeled with an antibody, MH27 that recognizes belt desmosomes within the hypodermis. This antiserum demarcates the cell boundaries of the hypodermis and allows for the identification of a failed cell fusion event (Podbilewicz and White, 1994). Analysis of spc-l(ra409)  mutant  embryos labeled with MH27 indicates that the cell fusions that produce hyp7 occur normally (data not shown). These data indicate that the development of the hypodermis is normal except for the spatial assembly of the hemidesmosomal-like attachment structures.  Pharyngeal muscle organization in spc-l(ra409) mutants Since the morphology of the pharynx was abnormal and pharyngeal contraction appeared irregular in spc-l(ra409) mutants (see Chapter 3), the myofilaments of the pharynx were examined in spc-l(ra409)  mutant embryos. The thin filaments of the pharynx were  labeled with FITC-phalloidin and their organization was investigated by fluorescence microscopy. First, the pharynx in spc-1 (ra409) mutants fails to elongate, resulting in a  170  Figure 31. The actin cytoskeleton of the hypodermis is normal in spc-1 (ra409) mutant embryos. Wild type (A), spc-1(ra409) (B) and sma-l(rul8)  (C) embryos were stained  with FITC-phalloidin to visualize the circumferentially oriented actin cytoskeleton in the hypodermis. In spite of the increased hemidesmosomal containing compartment of the hypodermis (see figure 29), the circumferentially oriented actin network within the hypodermis appears normal in spc-1 mutants. In both wild type and spc-1 mutant embryos the actin filaments can be seen organized in a circumferential pattern (arrows) running perpendicular to the body wall muscle quadrant (arrowheads). The sma-1 mutant displays defects in the patterning of the hypodermal actin cytoskeleton (note the spacing and arrangement abnormalities). The sma-1 mutant has been included to indicate how defects in the circumferential actin cytoskeleton of the hypodermis would appear. Scale bar is 10 |im.  171  172  pharynx approximately half the size of the pharynx in wild type embryos of a similar age. Second, a large gap is observed between the thin filaments of the pharynx, indicating the abnormal organization of the thin filaments of the pharynx. In wild type, the thin filaments organize on the basal and luminal membrane of the pharynx (forming two sets of half I bands) and a small gap appears between the two half I bands where the myosin filaments assemble (known as the H zone). The H zone in the spc-l(ra409) mutants is much larger than normal (Fig. 32). Lastly, the lumen of the pharynx appears not to close completely in spc-l(ra409) mutants. As visualized by phalloidin staining, the lumen of the pharynx in wild type is tightly closed and no gap is observed between neighboring luminal facing half I bands (Fig. 32). This is not the case in spc-l(ra409)  mutant  embryos, where a gap is present between neighboring luminal facing half I bands (Fig. 32). This indicates that the lumen is not completely closed which would affect the ability of this mutant to feed (see discussion).  Immunolocalization of a spectrin To determine the localization of a spectrin in the developing embryo, a rabbit polyclonal antiserum was produced. The complete third spectrin repeat plus part of the second and fourth spectrin repeat (AA 208-449) encoded by exon 2 was cloned into p G E X 4T-1 to produce DM#154. In an attempt to avoid antiserum cross reactivity with the other spectrin molecules (SMA-1 and UNC-70), the spectrin repeat chosen was the most distantly related to the other spectrins found in the C. elegans a spectrin protein. The antiserum has been designated AS 1 and was shown on Western blots to specifically recognize the fusion protein it was raised against (Fig. 33 A).  173  Figure 32. Thin filament structure of the pharynx in spc-l(ra409) mutant embryos. Wild type (A) and spc-1 (ra409) embryos (B) were stained with FITC-phalloidin to visualize the thin filaments of the pharynx. (A) The thin filaments of the wild type pharynx extend from the basal to the apical face throughout the length of the pharynx (the few gaps observed are the positions of nuclei). (B) The thin filaments of the spc-l(ra409) mutant pharynx never extend from the basal to the apical face (large gap between apical and basal face for the entire length of the pharynx). Additionally, the lumen has not sealed (gap in center of pharynx) in the spc-1 mutant. Arrows indicate the lumen and the basal membrane, and the arrowheads indicate the apical membrane of the pharynx. Scale bar is 10 |im.  174  175  To determine if AS 1 detects the C. elegans a spectrin, Western blots were conducted on wild type worm extracts. As shown in Fig. 33, AS1 diluted 1:50,000 recognizes two bands of -220 kDa and -240 kDa, whereas the preimmune had minimal cross reactivity with the worm extract (Fig. 33 B). The -220 kDa and -240 kDa species detected correspond to the sizes of the P and a spectrin isoforms, respectively. Therefore, it appears that AS1 recognizes both the a and (3 spectrin isoforms. However, when AS1 is diluted to 1:100,000, a single band of -240 kDa is detected on Western blots carried out on wild type worm extracts (Fig. 33 B). This indicates that it is possible for AS 1 to specifically recognize only a spectrin. The  P avy He  spectrin isoform is -400 kDa and was  not detected by AS1. Since it is possible that a protein of this size would have difficulties entering into the separating gel, the stacking gel was also transferred and probed with AS1 antisera. AS1 did not detect the presence of any protein species -400 kDa indicating that AS1 does not interact with  p av He  y  spectrin (data not shown). As a control  AS1 was incubated with 10 P-g/ml of fusion protein produced from DM#154 and subsequent Western blot analysis on wild type worm extracts failed to detect any protein product (Fig. 33 B) indicating that AS1 reacts specifically with the -220 and ~240kDa proteins.  Since a and P spectrin are known to form heterotetramers, analysis of protein localization using the AS1 antiserum should not be complicated by the cross reactivity with P spectrin . Therefore, to determine where the spectrin-based membrane cytoskeleton functions in C. elegans, wild type embryos were stained with AS1 and examined by immunofluorescence microscopy. In early embryos, staining is observed at  176  Figure 33. Western analysis of an a spectrin fusion protein and worm extracts with AS 1. (A) Western blot against purified SPC-1 ::GST (pAS 1; produced from DM#156) and GST. p A S l was loaded into lanes 1, 3 and 5 and GST was loaded into lanes 2, 4 and 6. Lane 1 and 2 were probed with AS 1 antiserum. AS 1 detects both pAS 1 and GST. Lanes 3 and 4 were probed with AS1 preabsorbed against pASl(-pASl). Neither p A S l nor GST were detected. Lanes 5 and 6 were probed with AS 1 preabsorbed against a GST expressing bacterial acetone powder (ASl(-GST)). ASl(-GST) only recognizes p A S l and not GST. For all experiments AS1 was used at 1:50,000. (B) Western blot of worm extract probed with preimmune (1:2000; lane 1). ASl(-GST) (1:50,000; lane 2), ASICGST) (1:100,000; lane 3) and A S l ( - p A S l ) (1:50,000; lane 4). The preimmune did not recognize any nematode proteins (lane 1). ASl(-GST) at 1:50,000 dilution recognized an -220 kDa protein and an -240 kDa protein. These sizes correspond with the sizes of the P and a spectrin, respectively. ASl(-GST) at 1:100,000 dilution only recognized an ~240kDa protein corresponding with the size of a spectrin. The A S l ( - p A S l ) did not recognize any nematode protein.  177  1 2  3 4  5 6 — 68  pASl  — 29  GST  B  12 3 4 -200  -97  -68  178  Figure 34. The spectrin cytoskeleton localizes to the cell membrane of most cells during embryogenesis. 1 1/2 fold stage embryos (A - E) stained with D M 5.6 (myosin, red) and AS1 (green). AS1 stains the cell membrane of many tissues. (A and C) indicate that hypodermal and muscle membrane staining. (B) indicates where the body wall muscle is located. (D - E) is a high magnification indicating that the body wall muscle is demarcated by AS 1 staining (arrowheads). (G) is a three fold wild type embryo labeled with AS1. Pharyngeal staining is indicated by the small arrows. The intestinal staining is indicated by a large arrowhead. Additionally, the nervous system stains with AS1. The ventral nerve cord (large arrow) is seen extending from the nerve ring (small arrowhead). Scale bar is 10 urn.  179  180  Figure 35. The spectrin cytoskeleton localizes to the I bands in adult body wall muscle. The body wall muscle of an adult wild type animal with DM5.6 (B), a myosin antibody, and AS1 (C). (A) is a merged panel of (B) and (C) (red is DM5.6 and green is AS1). DM5.6 indicates the localization of the A bands in the body wall muscle and does not overlap with the I bands. AS1 staining does not overlap with the DM5.6 staining indicating that the spectrin cytoskeleton localizes with the I bands. Scale bar is 10 p:m.  181  182  or near the cell membrane in most if not all cells (Fig. 34) starting at the earliest stages of embryogenesis, i.e. two cell stage (data not shown). After morphogenesis (elongation) and organogenesis initiates AS 1 staining is observed in the body wall muscle, hypodermis, pharynx, intestine and the developing nervous system (Fig. 34 G). In all these tissues, spectrin is localized to the cell membrane. Moreover, in the body wall muscle AS1 staining demarcates the muscle cells within the muscle quadrants (Fig. 34 DF), indicating that spectrin may maintain the cell shape of the body wall muscle cells.  To determine if spectrin colocalizes with components of the myofilament lattice in the body wall muscle, adults were stained with AS 1 and examined by immunofluorescence microscopy. Since the structure of the body wall muscle is much larger in adults, the spectrin cytoskeleton can be examined in greater detail than in embryos and its subcellular localization can be determined. Analysis of adult animals stained with AS 1 reveals that spectrin localizes to a region near the body wall muscle cell membrane overlying the I bands (Fig. 35). This indicates that the spectrin cytoskeleton localizes to an area containing the thin filaments and overlapping with the dense bodies.  Genetic interactions spc-1 (ra409) behaves as a genetic null As described earlier (Chapter 3), spc-l(ra409) mutants die as L l larvae that have failed to elongate beyond the two fold stage during embryogenesis. The spc-1 (ra417::Tcl) mutants arrest as L l larvae; however, they have elongated slightly longer during embryogenesis than spc-l(ra409) mutants. Since the T e l insert in spc-l(ra417::Tcl)  183  is  in the 3' end of the a spectrin gene (Fig. 23), is possible that a truncated or abnormal protein is produced and it is capable of some function to enable the  spc-l(ra417::Tcl)  mutants to partially elongate. To determine if either spc-1 (ra409) or  spc-l(ra417::Tcl)  are null mutants, both mutants were crossed into a strain carrying a deficiency that deletes the spc-1 locus (syDfl) and the heterozygous progeny were examined. Both spcl(ra409)/syDfl  and spc-l(ra417::Tcl)/syDfl  progeny arrested elongation at the two fold  stage and died as L I larvae. The heterozygous animals were indistinguishable from spcl(ra409) homozygotes, suggesting that spc-l(ra409)  is the genetic null of the spc-1  locus. In addition, heterozygous animals of the genotype l(ra409) and spc-l(ra417::Tcl)/syDfl  spc-l(ra417::Tcl)/spc-  were indistinguishable from spc-1 (ra409)  homozygotes, providing further support that spc-1 (ra409) is indeed a null mutation in the spc-1 locus and that spc-l(ra417::Tcl)  behaves as a hypomorphic allele.  Genetic interactions between the different spectrin mutants and ankyrin In addition to the single a spectrin gene, there are two (3 spectrin genes present in the C. elegans genome, encoding a non-erythroid (3 spectrin and a |3  Heavy  spectrin. Several  mutations have been isolated in both (3 spectrin genes. Null mutations in the (3 spectrin gene (unc-70) result in uncoordinated larvae that arrest as mid stage larvae (Park and Horvitz, 1986; Hammarlund et al., 2000). Similar to the spc-1 mutants, (3  Heavy  mutants  (sma-1) fail to complete elongation, however they are viable (McKeown et al., 1998). Therefore, sma-1 animals grow into short adults. Since the a and (3 subunits of spectrin are known to form heterotetramers (Bennett, 1990; Bennett and Gilligan, 1993), potential genetic interactions between these loci were tested. To accomplish this, double  184  heterozygotes and double homozygotes of the genotypes indicated in Table 9 were constructed between the three spectrin mutants and their phenotypes were examined for genetic enhancement (a more severe phenotype). The mutant alleles used for this analysis are recessive and are wild type when heterozygous (Table 9). Surprisingly, no double heterozygous or double homozygous animals constructed with the spc-1(ra409) enhanced the mutant phenotype of the single mutant (Table 9). For example, double heterozygous animals of the genotype sma-l(rul8)/+;  spc-1 (ra409)/+ were wild type.  Further, double homozygous animals of the genotype sma-1 (ml8); spc-1(ra409) were indistinguishable from homozygous spc-1 (ra409) mutants. The only double mutant that displayed any enhancement in phenotype was the double homozygous animal with the genotype unc-70(n493nl171),  sma-1(ml8) which resulted in animals with the identical  mutant phenotype as spc-1(ra409), L l arrested larvae that have failed to elongate beyond the two fold stage. Interestingly, animals with the genotype unc-70(n493nl 171)/+ , smal(ml8)  were not as healthy (i.e. smaller brood size) than sma-l(rul8)  unc-70(n493nl 171)/+ animals. unc-70(n493nll71)/+ than 20 progeny whereas sma-l(ml8)  animals have less  homozygotes and unc-70(n493nl 171)/+ animals  have well over a 100 progeny (sma-l(rul8) 70(n493nll71)/+  , sma-l(ml8)  homozygotes and  animals have -100 progeny and unc-  animals have -250 progeny). This indicates a genetic interaction  between unc-70 and sma-1 and illustrates the importance of the two (3 spectrins. Surprisingly, sma-l(rul8),  spc-1(ra409)/+ did not display any phenotypic enhancement  of the sma-1 (ml8) phenotype.  185  Table 9. Genetic interactions. Genotype  Phenotype  spc-l(ra409)  Arrest as two fold L I larvae  spc-l(ra409)/+  Wild type  sma-l(rul8)  Small viable animals  sma-l(ru!8)/+  Wild type  unc-70(n493nll71)  Arrest as L l / 2 larvae  unc-70(n493nl 171)7+  Wild type  unc-70(sl639)  Arrest as L l / 2 larvae  unc-70(sl639)7+  Wild type  unc-44(e362)  Uncoordinated viable animals  unc-44(e362)/+  Wild type  sma-1 (ml 8) /+; sma-l(ml8); sma-l(ru!8);  spc-l(ra409)/+  spc-l(ra409)/+  unc-70(n493nll71);  spc-l(ra409)/+ spc-l(ra409)  unc-70(n493nl 171) sma-1 unc-70(n493nll71)/+, unc-70(n493nll71), unc-44(e362)/+;  Small viable animals Arrest as two fold L I larvae  spc-1 (ra409)  unc-70(n493nl 171)/+;  Wild type  (ml8)/++  sma-l(rul8) sma-l(rul8)  spc-l(ra409)/+  Wild type Arrest as two fold L I larvae Wild type Unhealthy Small semi-viable animals Arrest as two fold L I larvae Wild type  unc-44(e362);  spc-l(ra409)/+  Uncoordinated viable animals  unc-44(e362);  spc-l(ra409)  Arrest as two fold L I larvae  186  To analyze another spectrin interacting protein, double mutants were constructed with animals harboring a mutation in the C. elegans ankyrin homologue,  unc-44(e362)(Otmka  et al., 1995). Ankyrin is a peripheral membrane protein involved in anchoring the spectrin cytoskeleton to the membrane (Bennett and Gilligan, 1993). unc-44(e362) mutants are viable, paralyzed animals that are coiled and slightly dumpy. The dumpy phenotype is similar to the small body length defect observed in sma-1 animals. Therefore, the reported interaction of ankyrin with the spectrin-based membrane cytoskeleton and the small body length phenotype suggests a possible interaction. To test this prediction, double mutants of the genotype unc-44(e362)/+; spc-1 (ra409)/+ and unc44(e362); spc-l(ra409)  were constructed (unc-44(e362) is a recessive mutation and  appears wild type when heterozygous). Again, no enhancement of phenotype was observed (Table 9). The double heterozygous animals were wild type and the double homozygotes were indistinguishable from single spc-l(ra409) homozygous animals. These data indicate that there is no effect by halving the gene dosage of two spectrin subunits that are known to form heterotetramers (Table 9).  Analysis of the body wall muscle in the (3 spectrin mutants To determine if the (3 spectrin isoforms have a role in muscle development, unc-70 and sma-1 mutant embryos were stained with antibodies specific to body wall myosin (DM5.6). Several different mutant alleles of unc-70 were stained including the putative null allele (unc-70(s!639); Hammarlund et al., 2000) and all were indistinguishable from wild type. In contrast, sma-1 mutants appeared to have a similar phenotype as spc-  187  I(ra409) mutants. The muscle quadrants of sma-1 mutant embryos  (P avy He  spectrin) were  wider than the muscle quadrants of similarly stage wild type embryos (Fig. 26 C). Additionally, the myofilaments were abnormally oriented in the body wall muscle cells (Fig. 26 C). Two mutant alleles of sma-1, ru!8 and e30, were analyzed and the same results were obtained for both. The mutant phenotype of sma-1 mutant embryos was not as severe as the muscular defects observed in spc-l(ra409) mutants such as the abnormal orientation is less affected in sma-1 mutants. However, the similarity in the mutant phenotypes suggests that the spectrin heterotetramer involved in this developmental event must be composed of the a and P spectrin. H  188  Discussion spc-1 encodes an a spectrin and the spectrin cytoskeleton localizes to the cell membranes of most tissues The C. elegans a spectrin (K10B3.10) is a 2427 A A polypeptide that is 72% similar to the human non-erythroid a spectrin that contains a centrally located SH3 domain and a carboxyl terminal C a binding domain (Hartwig, 1995; Fig. 23 and 24). This is the only 2+  a spectrin found within the C. elegans genome sequence (C. elegans Sequencing Consortium, 1998). Three lines of evidence indicate that the spc-1 locus on the X chromosome encodes the C. elegans a spectrin gene. First, introduction of dsRNA into wild type hermaphrodites produced offspring with a similar phenotype as  spc-l(ra409)  mutants. Second, a 15 kb genomic construct containing the entire a spectrin gene, including 5' and 3' flanking sequences rescues the spc-1(ra409) mutant phenotype. Lastly, the identification of a mutant (spc-l(ra417::Tcl)  with a transposable element  inserted into the 13 exon of the gene encoding the C. elegans a spectrin gene failed to th  complement spc-1 (ra409) mutants.  During C. elegans embryogenesis the spectrin cytoskeleton localizes near the plasma membrane of most, if not all, cell types beginning at the two cell stage. In the adult muscle, spectrin localizes to the sarcolemma overlying the I bands. Since a and P spectrin form a complex (Bennett and Gilligan, 1993) and results in Drosophila  indicate  that a spectrin consistently localizes with one of the P spectrins (Dubreuil et al., 1997), the results presented here using the AS 1 antiserum most likely indicate the true distribution of a spectrin even though the antiserum may cross react with the P subunit.  189  However, we can not rule out the possibility that (3 spectrin can localize to a unique location in the absence of a spectrin which would complicate the interpretation of a spectrin localization with the AS 1 antisera. Consistent with the immunolocalization results presented here, a study of the distribution of the P spectrin protein (UNC-70) indicates that P spectrin localizes to the cell membrane of most tissues starting at the earliest stages of embryogenesis (Moorthy et al., 2000). In addition, p spectrin was found to associate with the sarcolemma overlying the I bands (Moorthy et al., 2000). The other P isoform, sma-1, has been found to be predominantly expressed in the hypodermis and the pharynx (McKeown et al., 1998).  Genetic interactions The results from the genetic analysis indicate that spc-1 (ra409) behaves as a genetic null and that spc-l(ra417::Tcl)  behaves as a hypomorph. The genetic combinations tested  reveal that most likely C. elegans is composed of ap and aP  Heavy  tetramers since the  double homozygote constructed between unc-70 and sma-1 had an identical phenotype as spc-1(ra409).  If spectrin in C. elegans is composed of a and P, and a and P  avy  He  heterotetramers, then it is probable that if both P spectrins are removed, the outcome would be similar to removing the a spectrin. This indeed is what we observed when the unc-70 and sma-1 double homozygous mutants are constructed. Furthermore, reducing the functional gene copy of p spectrin in the p 70(n493nll71)/+,  Heavy  spectrin mutant background {unc-  sma-l(ru!8)) enhanced the sma-1 phenotype. Surprisingly, no  phenotypic enhancements were observed in the sma-l(rul8);  spc-1 (ra409)/+ animals.  Since these molecules form a complex, one would predict that reducing the functional a  190  spectrin gene copy in the sma-1 background would result in an observable enhanced phenotype. This implies that either a spectrin gene expression is increased to accommodate the lowered a spectrin protein level in relation to the P spectrin protein level or excessive amounts of a spectrin protein is normally produced so a defect is not detected or a combination both.  In addition, a genetic interaction was tested between the gene encoding the C. elegans ankyrin homolog, unc-44 (Otsuka et al., 1995). Ankyrin is known to be involved in anchoring the spectrin cytoskeleton to the membrane (Bennett and Gilligan, 1993). This analysis failed to detect any phenotypic enhancement (Table 9). Therefore, it is likely that proteins other than ankyrin are involved in the function of the spectrin-based membrane cytoskeleton in C. elegans. This is consistent with other reports that demonstrate that spectrin can be tethered to the membrane via other molecules (Bennett and Gilligan, 1993).  The involvement of the spectrin cytoskeleton in body wall muscle Mutations in the spc-1 locus result in animals that fail to complete the process of elongation during embryogenesis and arrest soon after hatching. Analysis of the body wall muscle of spc-1 mutants revealed that the muscle quadrants are twice as wide as normal and the myofilaments are abnormally organized. Normally, the myofilaments are organized parallel to the longitudinal axis of the worm. In spc-1(ra409) mutants the myofilaments are oriented at a steeper angle to the longitudinal axis. This likely results from the cell size difference between wild type animals and spc-1 (ra409) animals. Since  191  the organization of the myofilaments in spc-1 (ra409) mutants is abnormal from the initial stages of myofilament assembly, this suggest that either the cells are wider at this initial stage and that the spectrin cytoskeleton is involved in determining the size of the body wall muscle cells or the spectrin cytoskeleton is involved in the positioning of the myofilament attachment structures and this affects the size of the body wall muscle cells. Consistent with these models of spectrin function, the immunolocalization of the spectrin cytoskeleton was found to be associated with the muscle cell membrane in the body wall muscle of developing embryos. This staining pattern suggests that spectrin could be involved with maintaining the body wall muscle cell shape (in the absence of spectrin the cells can expand beyond their normal size) and/or the positioning of the myofilament attachment structure in the membrane. In the adult body wall muscle, which provides more spatial resolution due to the larger size of the muscle cells, the spectrin cytoskeleton was found to localize to the same region that contains the dense bodies (structures required to anchor the thin filaments to the membrane). This observation is consistent with the notion that the spectrin cytoskeleton may be involved in the positioning of the myofilament attachment structures.  During embryogenesis, the wild type embryo contains only 2 A bands per body wall muscle cell (Fig. 26 A). Once the wild type embryo hatches new A bands are added during the larval and adult stages to reach 8-10 A bands per body wall muscle cell (Fig. 35). Interestingly, spc-l(ra409) mutant embryos have 2 A bands per body wall muscle cell (the same as wild type). The body wall muscle cells of spc-l(ra409)  mutant embryos  are wide enough to have additional rows of A bands. This implies that body wall muscle  192  cell size does not restrict the number of A bands that are assembled during embryogenesis but suggests that there is another developmental mechanism that determines when new sarcomeres are assembled.  Interestingly, mutations in (3 spectrin (unc-70) do not affect the assembly or the organization of the myofilaments and the morphology of the body wall muscle is normal. This result was surprising since this tissue is reported to have a high level of (3 spectrin. However, in animals at later stages of development the myofilament lattice is reported to be disorganized (Moorthy et al., 2000; Hammarlund et al., 2000). These results indicate that the initial assembly and organization of the myofilaments in the body wall muscle occur normally but in later larval stages the body wall muscle is fragile. This suggests that the (3 isoform is required for the stability of the myofilament lattice and/or sarcolemma. The related molecule dystrophin has been implicated in the stability of the sarcolemma in skeletal and cardiac muscle (Berthier and Blaineau, 1997). Duchenne and Becker muscular dystrophy patients lack functional dystrophin and the disease manifests as the wasting of muscle (Michalak and Opas, 1997). Interestingly, spectrin colocalizes with dystrophin in skeletal and cardiac muscle (Williams and Bloch, 1999). A mutation in the C. elegans homologue of dystrophin/utrophin (dys-1) was recently identified and its mutant phenotype results in viable animals that are hyperactive and slightly hypercontracted (Bessou et al., 1998). Moreover, the dys-1 mutants have been found to have cholinergic transmission defects (Gieseler et al., 1999; Giugia et al., 1999). The organization of the myofilaments in the body wall muscle of dys-1 mutants is normal (Bessou et al., 1998), indicating that dystrophin does not have a role in the maintenance  193  of the body wall muscle in C. elegans. However, it does appear that the spectrin cytoskeleton is important for the maintenance of the body wall muscle (data presented here; Moorthy et al., 2000; Hammarlund et al., 2000). Since dystrophin and spectrin (Hartwig, 1995) are related molecules and they are known to colocalize in vertebrate striated muscle (Berthier and Blaineau, 1997), perhaps the spectrin cytoskeleton in C. elegans is providing the role that dystrophin and the spectrin cytoskeleton provide in vertebrate muscle.  Analysis of the  P avy He  spectrin mutant embryos (sma-1) revealed a similar body wall  muscle phenotype as spc-l(ra409) mutants. The body wall muscle quadrants and the underlying basement membrane of sma-1 mutant embryos are wider than normal. The data thus far indicates that the cq3 spectrin isoform is involved in the organization and H  the morphogenesis of the body wall muscle and the aP spectrin isoform is involved in the maintenance of the body wall muscle. A caveat to this model is that P  Heavy  spectrin has  been reported to be primarily expressed in the hypodermis and the pharynx (McKeown et al., 1998). This suggests that it is unlikely that txP  Heavy  spectrin is functioning in the body  wall muscle. How then can the muscle defect observed in these two mutants be explained? Here we propose two possibilities that are not mutually exclusive. First, since spc-1 and sma-1 mutants both fail to complete elongation and they have the same affect on body wall muscle development, it is possible that the body wall muscle defect may arise from the lack of elongation. Alternatively,  P avy He  spectrin is expressed in the  body wall muscle and it was not detected in the in situ hybridization experiments (McKeown et al., 1998). Interestingly, several Pat mutants that affect gene products that  194  are involved in the regulation of sarcomere contraction have myofilaments that have assembled normally and are not abnormally oriented. For example, animals harboring mutations in the genes encoding troponin C, tropomyosin and a voltage gated C a  2+  channel, all have normal body wall muscle architecture even though they arrest at the two fold stage (Williams and Waterston, 1994; Kagawa et al., 1997; Lee et al., 1997b; Terami et al., 1999). This indicates that the failure to elongate beyond two fold alone can not account for the abnormally positioned myofilaments as observed in spc-1 and sma-1 mutants, suggesting that the P  H e a v y  spectrin may function with a spectrin in the body wall  muscle to maintain muscle cell shape and/or myofilament positioning.  On the other hand, the failure of the Pat mutants to elongate is presumably due to the lack of body wall muscle contraction. The influence of body wall muscle contraction on the process of elongation is not well understood but it is known that if the body wall muscle does not contract then the embryo fails to elongate beyond the two fold stage (Williams and Waterston, 1994). In addition to body wall muscle contraction, circumferential constriction of the cytoskeleton within the hypodermis is important for embryonic elongation (Priess and Hirsh, 1986; Costa et al., 1998). Since the body wall muscle in sma-1 and spc-1 is capable of muscle contraction, the failure to elongate in these mutants most likely arises from a defect in the hypodermis. As mentioned, the body wall muscle cells of sma-1 and spc-1 are wider than normal at the initial stages of myofilament assembly. This may occur because the hypodermis is failing to undergo normal cytoskeleton rearrangements that are required for embryonic elongation or the timing of these events is retarded (see below). Normally, the hypodermal cytoskeletal contractions  195  convert the embryo from a ball of cells into a cylindrical worm shape. This results in the redistribution of the volume of the embryo, lengthening the anterior-posterior axis and reducing the dorsal-ventral axis. This morphogenic event could result in the establishment of spatial restrictions on the body wall muscle cells dictating the maximum size of the body wall muscle cells. If the process of elongation fails or is retarded, then this spatial restriction would not be established and the muscle cells could become wider leading to the abnormal positioning of the myofilaments. Since the spectrin cytoskeleton is widely expressed, more studies are required to further elucidate the role of the spectrin cytoskeleton (see below).  The body wall muscle directs where the basement membrane is established and where the hypodermal attachment structures are assembled In C. elegans a physical linkage between the myofilament lattice of the body wall muscle and the cuticle is required to allow myofilament contraction to result in the movement of the animal. This complex linkage system is composed of integrin containing adhesioncomplexes (Fig. 1 ) within the body wall muscle cells that anchor the myofibrils to the muscle cell membrane, a specialized basement membrane underlying the muscle quadrant, and hemidesmosomal-like structures present in the hypodermis. The hypodermis has specialized faces, one to anchor the hypodermis to the basement membrane and the other to secrete and anchor the cuticle. In spc-1 (ra409) mutants the basement membrane underlying the body wall muscle and the hypodermal attachment structures associated with the body wall muscle are wider than normal, mirroring the expansion of the body wall muscle. These results suggest that the body wall muscle is  196  determining where the basement membrane is established and through an inductive event the body wall muscle is creating a zone where the hypodermal attachment structures are assembled (Fig. 36). Results from two previous studies are consistent with this model. These two studies involved the laser ablation of certain body wall muscle cells, creating a gap in the body wall muscle quadrant. One study found that the basement membrane failed to assemble in the region lacking muscle cells (Moerman et al., 1996), indicating that the body wall muscle is important in the establishment of the basement membrane (Fig. 36). The second study found that the attachment structures in the hypodermis fail to assemble over the body wall muscle gap (Hresko et al., 1999), suggesting that the body wall muscle is the source of an inductive signal that directs where the attachment structures of the hypodermis are to be assembled (Fig. 36).  Pharyngeal defects in spc-l(ra409)  mutants  The pharynx, the other major muscle tissue of C. elegans, is also abnormal in spc-1 mutants. The thin filaments are abnormally arranged and the lumen of the pharynx never appears fully closed. Contraction of the pharynx is never observed, although pharyngeal contraction is inferred since spc-1 mutants are able to hatch. It is likely that weak pharyngeal contractions occur, but it is doubtful that these pharyngeal contractions are able to pump food into the intestine. C. elegans is under internal pressure, and the pharynx is required to pump food into the digestive tract and is essential to seal off the pharyngeal lumen to prevent food from exiting out of the mouth (Avery and Thomas, 1996). In spc-1 mutants it is likely that either food cannot be properly pumped into the intestine or the pharynx can not seal off the lumen and the internal pressure does not  197  Figure 36. Schematic illustration of the interaction of the body wall muscle and the overlying hypodermis. Spectrin (red circles) is localized to the muscle cell boundaries in the wild type animal and the green and yellow squares indicate the myofilament attachment sites. The hemidesmosomal structures in the hypodermis are indicated by the red stripped bar. When the body wall muscle cells are ablated ('X'), as in B , the hemidesmosomal structures fail to assemble (red stippled bar; Hresko et al., 1999) and the basement membrane is not formed (blue stipples; Moerman et al., 1996). In spc-1 mutants, the body wall muscle cells become wider. The wider muscle cells result in a wider muscle quadrant, which leads to a wider basement membrane and the hemidesmosomes are assembled in a wider pattern. These results suggests that the muscle cell signals to the hypodermis to determine where the hemidesmosomes will be assembled.  198  199  allow any food enter the digestive tract or a combination of both. The inability of these mutants to feed is what most likely leads to the arrest of the spc-1 mutants.  The involvement of the spectrin cytoskeleton in embryonic morphogenesis Time lapse video recording of spc-l(ra409) mutants during embryogenesis indicates that these mutants have a slow rate of elongation and that this process ceases once it reaches the two fold length resulting in L I larvae that are half the size of wild type L I larvae (Fig. 22). Contraction of the apically localized actin cytoskeleton in the hypodermis is thought to be the driving force behind the process of elongation during morphogenesis (Priess and Hirsh, 1986). This cytoskeleton appears to assemble normally in spcl(ra409) mutants (Fig. 31 B); however, they remain unable to complete elongation. Other mutants have a similar defect, such as mutations in let-502 and mlc-4. Although the actin cytoskeleton of these mutants has not been examined, these mutants arrest at the two fold stage and appear similar to spc-l(ra409).  let-502 encodes a Rho dependent  kinase and is thought to regulate the process of elongation by interacting with contractile elements in the seam cells of the hypodermis (Wissmann et al., 1997). mlc-4 encodes a non-muscle myosin light chain and is also proposed to be involved in the process of elongation by regulating contractile elements in the seam cells of the hypodermis (Shelton et al., 1999). Therefore, it is possible that the spectrin cytoskeleton is involved in the localization of regulatory molecules to the apically localized actin cytoskeleton and/or required for the stabilization of the hypodermal membrane. Another molecule known to be involved in the process of elongation is (3  200  Heavy  spectrin. Mutations in P  H e a v y  spectrin (sma-1) result in retarded elongation that ceases around two fold stage of embryogenesis. The (3  Heavy  subunit contains an N-terminal actin binding domain, a  centrally located SH3 domain and a C-terminal Pleckstrin homology domain. Therefore, one could speculate that the actin binding domain in the (3  Heavy  spectrin could also be  involved in regulating the actin cytoskeleton in the hypodermis. This is consistent with the examination of the actin cytoskeleton in sma-1 animals, which revealed an abnormal arrangement of the actin filaments (Fig. 31 C).  Future direction In order to clear up some of the issues brought up in the discussion, further experimentation is needed. One of these issues is that since a spectrin is widely expressed it is difficult to interpret whether defects observed in spc-1 (ra409) mutants is the result of the loss of a spectrin function in that particular tissue or a secondary affect arising from an abnormal neighboring tissue. For example, the body wall muscle defect in spc-1(ra409) mutants could result from the loss of a spectrin function in the body wall muscle, or a secondary defect caused by the loss of a spectrin function in the hypodermis or it could be a combination of both. One line of experiments that would address the function of a spectrin in the body wall muscle without the complications of secondary affects arising from spectrin function lacking in neighboring tissues would be to generate genetic mosaic animals. Genetic mosaics are animals that bear genetically distinct cell types, some cells of an individual are genotypically mutant and other cells are genotypically wild type. This type of analysis is carried out routinely in C. elegans (Lackner et al., 1994; Miller et al., 1996; Baum et al., 1997). In addition, spectrin is  201  highly expressed in the nervous system, therefore, defects in the nervous system should be investigated in spc-l(ra409) mutants and/or genetic mosaic animals lacking a spectrin function in certain neuronal cell lineages. These mosaic experiments, when added to the information presented here on spectrin function in C. elegans development, should provide a greater understanding of spectrin function in non-erythroid cells.  The spectrin-based membrane cytoskeleton is widely expressed and provides a network of proteins that act as structural support to the cell membrane. In addition, spectrin is a multidomain protein providing many sites for protein interactions, such as ankyrin, Band 4.1, adducin and filamentous actin to name only a few. Ankyrin has been shown to interact with ion channels and other transmembrane molecules (Bennett and Gilligan, 1993). Band 4.1, a member of the F E R M family of cytoskeletal regulating molecules, contains a domain that interacts with the actin cytoskeleton and another domain that interacts with membrane (Vaheri et al.,1997). Furthermore, two P D Z - L I M proteins have recently been identified that interact with a spectrin repeat in a actinin (Xia et al., 1997; Vallenius et al., 2000). This large and diverse number of interacting protein partners suggests that spectrin may act as a nucleating site for the regulation of many cellular events, thus highlighting the importance of understanding the role the spectrin-based membrane cytoskeleton has in non-erythrocyte cells.  202  Chapter 6. Summary and Conclusion  Through the process of genetic dissection and cell biological techniques, many of the components of the contractile unit of muscle, the sarcomere, have been identified. The emerging view of how sarcomeres are assembled is similar in many aspects in vertebrate and invertebrate striated muscle. Assembly is initiated at the muscle cell membrane and the two filament structures are assembled in a compartmental nature. Although thin and thick filaments are intimately associated in mature muscle, they are able to assemble independent of one another. Perturbation of either the thin or thick filaments does not interfere with the reciprocal in assembly (Waterston, 1989; Venolia and Waterston, 1990; Barstead and Waterston, 1991; L u et al., 1992). The membrane and associated components have been implicated in the initial assembly of the sarcomere (for review, see Epstein and Fischman, 1991; Franzini-Armstrong, 1996). For example, in cultured cardiac cells myofibrillar components are found adjacent to the muscle cell membrane and begin to form stress fiber like structures, indicating that sarcomere assembly is initiated at the membrane (Antin et al., 1986). Moreover, mutations in the transmembrane molecule integrin, in both C. elegans and Drosophila,  result in the failure  of myofilament assembly (Volk et al., 1990; Williams and Waterston, 1994; Gettner et al., 1995). Although the process of myofilament assembly is beginning to be unraveled, many unanswered questions remain. Some of these questions include what determines myofilament length and myofilament number, how are the myofilaments positioned in the membrane and how are these processes regulated. Additionally, many components of  203  striated muscle have been identified; however, mutational analyses to determine what is the function of these components have not been conducted.  In an attempt to answer some of these questions we conducted a mutant screen to identify new genes involved in the development of the body wall muscle of C. elegans. From this screen we have isolated eight mutants that affect different stages of early muscle development. These mutations have been placed into five different classes based on their mutant phenotypes. The class I mutant, unc-52(ra401), fails to complete the organogensis of the body wall muscle. In these mutants, the myofilaments never assemble and body wall muscle contraction is not observed (see Chapter 3). unc-52 encodes the basement membrane molecule perlecan (Rogalski et al., 1993). Perlecan (UNC-52) has been shown to be essential for the initial stages of myofilament assembly. Interestingly, unc-52 mutants share an identical mutant phenotype as the C. elegans integrin mutants, pat-2 and pat-3 (Williams and Waterston, 1994, Gettner et al., 1995; B . Williams pers. comm.). Furthermore, an investigation of integrin localization in unc-52 mutants reveals that integrin fails to become stabilized in the membrane (G. Mullen and D. Moerman, pers. comm.). These observations suggest that perlecan is either the integrin ligand or is required for the function of the integrin ligand. In addition, these results provide further support of the importance of the muscle cell membrane and its associated components during myofilament assembly.  The class II mutants, pat-13(ra407) and pat-14(ra408), initiate myofilament assembly but the filaments never become entirely organized. The muscle cells become polarized and  204  nascent myofilaments (or immature myofilaments; see Epstein et al., 1993) are assembled; however, mature myofilaments are never observed in these mutants (see Chapter 3). Perhaps filament length is defective or auxiliary proteins required for proper filament assembly are lacking in these mutants. Interestingly, pat-13(ra407) maps close to a gene encoding the actin capping protein tropomodulin and pat-14(ra408) maps near a gene encoding the C. elegans ortholog of flightless (Claudianos and Campbell, 1995). Both of these proteins have reported roles in the regulation of actin filament length (Straub et a l . , 1996; Cooper and Schafer, 2000). Therefore, if these mutations do affect these genes one could speculate that these mutants have thin filaments with abnormal lengths. In this case, the myofilaments would be able to undergo initial assembly to form nascent myofilaments (and the muscle cell could undergo polarization), however, when the thick and thin filaments interdigitate to form mature sarcomeres the thin filaments of abnormal length would interfere with this process.  The class III mutants, mua-1 l(ra402) and mua-12(ra404), initiate myofilament assembly normally but, as in the class III mutants, the filaments never become entirely organized. Additionally, the muscle quadrants of these mutants fail to make proper attachments with the hypodermis (see Chapter 3). Since perlecan remains associated with the surface of the separated muscle quadrant(s), this defect in these mutants most likely arises in the E C M or the hypodermis. Consistent with this hypothesis, three hypodermally expressed genes have been identified whose mutant phenotypes are similar to the class III mutant phenotypes. These genes are mua-1, mua-3 and mup-4 and they are exclusively  205  expressed in the hypodermis (E. Bucher pers. comm. and J. Plenefisch pers comm.). A l l three of these mutants display progressive muscle detachment from the hypodermis.  The class IV mutants, mua-13(ra406) and let-268(ra414), initially assemble normal sarcomeres, however, once muscle contraction commences the body wall muscle separates from the hypodermis. Additionally, some of the muscle cells become detached creating large gaps in the body wall muscle quadrants (see Chapter 3). We found that let268 encodes an enzyme, procollagen lysyl hydroxylase, required for the proper modification of collagen (see Chapter 4). In let-268(ra414) mutants, type IV collagen is not secreted from the body wall muscle cells into the basement membrane. Therefore, let-268(ra414) phenocopies mutations in type IV collagen. The mutant phenotype of let268(ra414) indicates that type IV collagen is not required for myofilament assembly, however, type IV collagen is required in the basement membrane to anchor the body wall muscle cells to the hypodermis. The similarity of the mua-13(ra406) mutant phenotype with the type IV collagen and the let-268 mutant phenotypes (see Chapter 4) suggests that the mua-13 gene product may act in the basement membrane for stability.  The class V mutant, spc-l(ra409),  assembles myofilaments into sarcomeres but the  myofilaments are not properly positioned and the muscle quadrants are twice as wide as normal (see Chapter 5). The wider muscle quadrants observed in these mutants likely results from the body wall muscle cells taking on an abnormal morphology or excess cell growth. We found that spc-1 gene encodes an a subunit of spectrin (see Chapter 5). In the erythrocyte, spectrin is a major membrane cytoskeletal protein that is involved in  206  maintaining cell shape and maintaining the structural integrity of the membrane (Lux and Palek, 1995). Here, we have shown that in the absence of a spectrin the body wall muscle cells are wider than normal. Therefore, spectrin may have a analogous function in the body wall muscle cells of C. elegans as it does in erythrocyte cells. Additionally, the spectrin cytoskeleton may have a role in determining where the myofilaments are anchored to the muscle cell membrane. Spectrin has been implicated in generating membrane subdomains in many cells types (Craig and Pardo, 1983; Bloch and Morrow, 1989; Black et al., 1988; Sikorski et al., 1991) and regulating lateral diffusion in the erythrocyte membrane (Tomishige et al., 1998). In an analogous scenario in the body wall muscle cells of C. elegans, the spectrin based membrane cytoskeleton may regulate the distribution of proteins required for anchoring the myofilaments in the membrane (such as integrins). A n analysis of the basement membrane and the hypodermal attachment structures in spc-l(ra409) mutants implicates the body wall muscle cells in determining where the basement membrane is established and through an inductive event where the hypodermal attachment structures are assembled.  This study has provided 8 new mutations with different affects on muscle assembly. Furthermore, two of these genes have been cloned and this has led to the examination of the role of type IV collagen and the role of the spectrin cytoskeleton during muscle development in C. elegans. As mentioned earlier, the cell membrane is the nucleating point for sarcomere assembly. Interestingly, sarcomeres are able to assemble in the type IV collagen, let-268 and the spc-1 mutants (although the myofilaments are abnormally positioned in the spc-1 mutants). Spectrin is a major membrane cytoskeletal component  207  that is tightly associated with the membrane, yet spc-1 mutants are capable of making sarcomeres. This indicates that integrin and associated molecules must localize to the membrane and allow for myofilament assembly in the absence of a spectrin. Type IV collagen, a major basement membrane molecule, is found in the thin specialized basement membrane between the body wall muscle cells and the hypodermis. In let-268 and type IV collagen mutants, type IV collagen is not secreted into the basement membrane, yet these mutants are capable of assembling a normal myofilament lattice. It appears that type IV collagen and a spectrin "bracket" the key elements involved in sarcomere assembly. While spectrin is required for the proper positioning and type IV collagen is required for the stability of the basement membrane, these molecules are not required for sarcomere assembly. In type IV collagen and let-268 mutants perlecan, a component required for sarcomere assembly, initially localizes normally under the body wall muscle cells leading to the assembly of the myofilaments. In spc-1 mutants, perlecan is distributed under the body wall muscle cells, albeit in a spatially wider pattern, which allows the myofilaments to assemble into sarcomeres. These two examples emphasize the importance of the membrane and the molecules it is intimately associated with in the process of myofilament assembly.  The further characterization and cloning of the other mutations described in this study may identify new components that are closely associated with the membrane and are required for the proper building of a sarcomere, such as the identification of unc52(ra401). In addition, we have potentially identified mutants that are involved in regulating filament length (pat-13 and pat-14) and spc-1 appears to influence where the  208  myofilaments are positioned in the membrane. Moreover, the identification of a mutation in a spectrin allows for the first mutational analysis of the function of a spectrin in striated muscle. Likewise, the mutation isolated in let-268 provided new information on the role of type IV collagen during muscle development. A large mutant screen based on the conditions described here should help answer many of the questions remaining of the process of myofilament assembly (discussed above and in Chapter 1).  209  References Adams, M.D., Celniker, S.E., Holt, R.A., Evans, C.A., Gocayne, J.D., Amanatides, P.G., Scherer, S.E., Li, P.W., Hoskins, R.A., Galle, R.F., et al. (2000). The genome sequence of Drosophila  melanogaster.  Science. 287: 2185-2195.  Albertson, D.G. and Thomson, J.N. (1976). The pharynx of Caenorhabditis  elegans.  Philos. Trans. R. Soc. Lond. B. Biol. Sci. 275: 299-325.  Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z., Miller, W. and Lipman, D.J. (1997). Gapped B L A S T and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25: 3389-3402.  Antin, P.B., Tokunaka, S., Nachmias, V.T. and Holtzer, H. (1986). Role of stress fiber-like structures in assembling nascent myofibrils in myosheets recovering from exposure to ethyl methanesulfonate. J. Cell Biol.102: 1464-1479.  Ao, W. and Pilgrim, D. (2000).  Caenorhabditis elegans UNC-45 is a component of muscle thick filaments and colocalizes with myosin heavy chain B , but not myosin heavy chain A. J. Cell Biol. 148: 375-384.  Avery, L. and Thomas, J.H. (1997). Feeding and Defecation. In C. elegans II. D.L. Riddle, T. Blumenthal, B.J. Meyer, and J.R. Priess, editors. Cold Spring Harbor, N Y : Cold Spring Harbor Laboratory Press. 679-716.  Barral, J.M., Bauer, C.C., Ortiz, I. and Epstein, H.F. (1998).  unc-45 mutations in  Caenorhabditis elegans implicate a CR01/She4p-like domain in myosin assembly. J. Cell Biol. 143: 1215-1225.  Barstead, R.J., Kleiman, L. and Waterston, R.H. (1991). Cloning, sequencing, and mapping of an alpha-actinin gene from the nematode Caenorhabditis elegans. Cell Motil Cytoskeleton 20: 69-78.  Barstead, R.J. and Waterston, R.H. (1991). Vinculin is essential for muscle function in the nematode. J. Cell Biol. 114: 715-724.  Baum, P.D. and Garriga, G. (1997). Neuronal migrations and axon fasciculation are disrupted in ina-1 integrin mutants. Neuron. 19: 51-62.  Benian, G.M., Kiff, J.E., Neckelmann, N., Moerman, D.G. and Waterston R.H. (1989). Sequence of an unusually large protein implicated in regulation of myosin activity in C. elegans. Nature. 342: 45-50.  Bennett, V. (1990). Spectrin-based membrane skeleton: a multipotential adaptor between plasma membrane and cytoplasm. Physiol. Rev. 70: 1029-1065. 210  Bennett, V. and Gilligan, D.M. (1993). The spectrin-based membrane skeleton and micron-scale organization of the plasma membrane. Annu. Rev. Cell Biol. 9: 27-66.  Berthier, C. and Blaineau, S. (1997). Supramolecular organization of the subsarcolemmal cytoskeleton of adult skeletal muscle fibers. A review. Biol. Cell. 89: 413-434.  Bessou, C , Giugia, J.B., Franks, C.J., Holden-Dye, L. and Segalat, L. (1998). Mutations in the Caenorhabditis elegans dystrophin-like gene dys-1 lead to hyperactivity and suggest a link with cholinergic transmission. Neurogenetics. 2: 61-72.  Black, J.D., Koury, S.T., Bankert, R.B. and Repasky, E.A. (1988). Heterogeneity in lymphocyte spectrin distribution: ultrastructural identification of a new spectrin-rich cytoplasmic structure. J. Cell Biol. 106:97-109. Bloch, R.J. and Morrow, J.S. (1989). A n unusual beta-spectrin associated with clustered acetylcholine receptors. J. Cell Biol. 108:481-493.  Bloor, J.W. and Brown, N.H. (1998).Genetic analysis of the Drosophila alphaPS2 integrin subunit reveals discrete adhesive, morphogenetic and sarcomeric functions. Genetics. 148 :1127-1142.  Bonfanti, L., Mironov, A.A. Jr., Martinez-Menarguez, J.A., Martella, O., Fusella, A., Baldassarre, M., Buccione, R., Geuze, H.J., Mironov, A.A., Luini, A. (1998). Procollagen traverses the Golgi stack without leaving the lumen of cisternae: evidence for cisternal maturation. Cell. 95: 993-1003.  Brenner, S. (1974). The genetics of Caenorhabditis  elegans. Genetics.  77: 71-94.  Brodsky, B. and Ramshaw, J.A. (1997). The collagen triple-helix structure.  Matrix  Biol. 15: 545-554. C. elegans  Sequencing Consortium. (1998). Genome sequence of the nematode  C.  elegans: a platform for investigating biology. Science. 282: 2012-2018.  Campbell, H.D., Schimansky, T., Claudianos, C , Ozsarac, N., Kasprzak, A.B., Cotsell, J.N., Young, I.G., de Couet, H.G. and Miklos, G.L. (1993). The Drosophila melanogaster flightless-I gene involved in gastrulation and muscle degeneration encodes gelsolin-like and leucine-rich repeat domains and is conserved in Caenorhabditis elegans and humans. Proc. Natl. Acad. Sci. USA.  90: 11386-11390.  Claudianos, C. and Campbell, H.D. (1995). The novel flightless-I gene brings together two gene families, actin-binding proteins related to gelsolin and leucine-rich-repeat proteins involved in Ras signal transduction. Mol. Biol. Evol. 12: 405-414.  211  Cooper, J.A. and Schafer, D.A. (2000). Control of actin assembly and disassembly at filament ends. Curr. Opin. Cell Biol. 12: 97-103.  Costa, M., Draper, B.W. and Priess, J.R. (1997). The role of actin filaments in patterning the Caenorhabditis elegans cuticle. Dev. Biol. 184: 373-84.  Costa, M., Raich, W., Agbunag, C , Leung, B., Hardin, J. and Priess, J.R. (1998). A putative catenin-cadherin system mediates morphogenesis of the Caenorhabditis embryo. J. Cell Biol. 141:297-308.  elegans  Costell, M., Gustafsson, E., Aszodi, A., Morgelin, M., Bloch, W., Hunziker, E., Addicks, K., Timpl, R. and Fassler, R. (1999). Perlecan Maintains the Integrity of Cartilage and Some Basement Membranes. J. Cell Biol. 147: 1109-1122.  Coulson, A., Kozono, Y., Lutterbach, B., Shownkeen, R., Sulston, J. and Waterston, R. (1991). Y A C s and the C. elegans genome. Bioessays. 13: 413-417.  Coulson, A., Huynh, C , Kozono, Y. and Shownkeen, R. (1995). The physical map of the Caenorhabditis  elegans genome. Methods Cell Biol. 48: 533-550.  Cox, G.N., Kramer, J.M. and Hirsh, D. (1984). Number and organization of collagen genes in Caenorhabditis  elegans. Mol. Cell Biol. 4: 2389-2395.  Cox, G.N., Staprans. S. and Edgar, R.S. (1981). The cuticle of  Caenorhabditis  elegans. II. Stage-specific changes in ultrastructure and protein composition during postembryonic development. Dev. Biol. 86: 456-470.  Craig, S.W. and Pardo, J.V. (1983). Gamma actin, spectrin, and intermediate filament proteins colocalize with vinculin at costameres, myofibril-to-sarcolemma attachment sites. Cell Motil. 3: 449-462. de Cuevas, M., Lee, J.K. and Spradling, A.C. (1996). alpha-spectrin is required for germline cell division and differentiation in the Drosophila ovary. Development. 122: 3959-3968. De Matteis MA, Morrow JS. (1998). The role of ankyrin and spectrin in membrane transport and domain formation. Curr. Opin. Cell Biol. 10: 542-549.  Devarajan, P., Stabach, P.R., Mann, A.S., Ardito, T., Kashgarian, M. and Morrow, J.S. (1996). Identification of a small cytoplasmic ankyrin (AnkGl 19) in the kidney and muscle that binds beta I sigma spectrin and associates with the Golgi apparatus. J. Cell Biol. 133: 819-830.  Drubin, D.G. and Nelson, W.J. (1996). Origins of cell polarity.  212  Cell.  84: 335-344.  Dubreuil, R., Byers, T.J., Branton, D., Goldstein, L.S. and Kiehart, D.P. (1987). Drosophilia  spectrin. I. Characterization of the purified protein.  J. Cell Biol. 105: 2095-2102.  Dubreuil, R.R., Byers, T.J., Stewart, CT. and Kiehart, D.P. (1990). A beta-spectrin isoform from Drosophila (beta H) is similar in size to vertebrate dystrophin. J. Cell Biol. Ill: 1849-1858.  Dubreuil, R.R. and Yu, J. (1994). Ankyrin and beta-spectrin accumulate independently of alpha-spectrin in Drosophila.  Proc. Natl. Acad. Sci. U S A. 91: 10285-10289.  Dubreuil, R.R., Maddux, P.B., Grushko, T.A. and MacVicar, G.R. (1997). Segregation of two spectrin isoforms: polarized membrane-binding sites direct polarized membrane skeleton assembly. Mol. Biol. Cell. 8: 1933-1942.  Engel, J. and Prockop, D.J. (1991). The zipper-like folding of collagen triple helices and the effects of mutations that disrupt the zipper. Annu. Rev. Biophys. Biophys. Chem. 20: 137-152.  Epstein, H.F., Casey, D.L. and Ortiz, I. (1993). Myosin and paramyosin of Caenorhabditis elegans embryos assemble into nascent structures distinct from thick filaments and multi-filament assemblages. /. Cell Biol. 122: 845-858.  Epstein, H.F. and Fischman, D.A. (1991). Molecular analysis of protein assembly in muscle development. Science. 251: 1039-1044.  Epstein, H.F. and Thomson, J.N. (1974). Temperature-sensitive mutation affecting myofilament assembly in Caenorhabditis elegans. Nature. 250: 579-80.  Fath, K.R., Trimbur, G.M. and Burgess, D.R. (1997). Molecular motors and a spectrin matrix associate with Golgi membranes in vitro. J. Cell Biol. 139: 1169-1181.  Fessler, J.H. and Fessler, L.I. (1978). Biosynthesis of procollagen. Annu.  Rev.  Biochem. 47: 129-162.  Fire, A., Xu, S., Montgomery, M.K., Kostas, S.A., Driver, S.E. and Mello, C.C. (1998). Potent and specific genetic interference by double-stranded R N A in Caenorhabditis  elegans. Nature. 391: 806-811.  Fitzgerald, M.C. and Schwarzbauer, J.E. (1998). Importance of the basement membrane protein SPARC for viability and fertility in Caenorhabditis elegans. Curr. Biol. 8: 1285-1288.  213  Francis, G.R. and Waterston, R.H. (1985). Muscle organization in Caenorhabditis elegans: localization of proteins implicated in thin filament attachment and I-band organization. J. Cell Biol. 101: 1532-1549. Francis, R. and Waterston, R.H. (1991). Muscle cell attachment in Caenorhabditis elegans. J. Cell Biol. 114: 465-479. Franzini-Armstrong, C. (1996). Functional significance of membrane architecture in skeletal and cardiac muscle. Soc. Gen. Physiol. Ser. 51: 3-18. Gatewood, B.K. and Bucher, E.A. (1997). The mup-4 locus in Caenorhabditis elegans is essential for hypodermal integrity, organismal morphogenesis and embryonic body wall muscle position. Genetics. 146: 165-183. Gettner, S.N., Kenyon, C. and Reichardt, L.F. (1995). Characterization of beta pat-3 heterodimers, a family of essential integrin receptors in C. elegans. J. Cell Biol. 129: 1127-1141. Gieseler, K., Bessou, C. and Segalat, L. (1999). Dystrobrevin- and dystrophin-like mutants display similar phenotypes in the nematode Caenorhabditis elegans. Neurogenetics. 2: 87-90. Gilchrist, E.J. and Moerman, D.G. (1992). Mutations in the sup-38 gene of Caenorhabditis elegans suppress muscle-attachment defects in unc-52 mutants. Genetics. 132: 431-442. Giugia, J., Gieseler, K., Arpagaus, M. and Segalat, L. (1999) Mutations in the dystrophin-like dys-1 gene of Caenorhabditis elegans result in reduced acetylcholinesterase activity. FEBS Lett. 463: 270-272. Goh, P.Y. and Bogaert, T. (1991). Positioning and maintenance of embryonic body wall muscle attachments in C. elegans requires the mup-1 gene. Development. I l l : 667681. Gossett, L.A., Hecht, R.M. and Epstein, H.F. (1982). Muscle differentiation in normal and cleavage-arrested mutant embryos of Caenorhabditis elegans. Cell. 30: 193-204. Graham, P.L., Johnson, J.J., Wang, S., Sibley, M.H., Gupta, M.C. and Kramer, J.M. (1997). Type IV collagen is detectable in most, but not all, basement membranes of Caenorhabditis elegans and assembles on tissues that do not express it. / . Cell Biol. 137: 1171-1183. Gregorio, C.C., Weber, A., Bondad, M., Pennise, CR. and Fowler VM. (1995). Requirement of pointed-end capping by tropomodulin to maintain actin filament length in embryonic chick cardiac myocytes. Nature. 377: 83-86.  214  Guo, X.D., Johnson, JJ. and Kramer, J.M. (1991). Embryonic lethality caused by mutations in basement membrane collagen of C. elegans. Nature 349: 707-709.  Gupta, M.C., Graham, P.L. and Kramer, J.M. (1997). Characterization of alphal(IV) collagen mutations in Caenorhabditis elegans and the effects of alphal and alpha2(IV) mutations on type IV collagen distribution. J Cell Biol. 137: 1185-1196.  Hammarlund, M., Davis, W. S. and Jorgensen, E. M. (2000). Mutations in 8-Spectrin Disrupt Axon Outgrowth and Sarcomere Structure. J. Cell Biol. 149: 931-942.  Hartwig, J.H. (1995). Actin-binding proteins. 1: Spectrin super family. CCQ. 2: 703-800.  Harwood, R., Grant, M.E. and Jackson, D.S. (1976). The route of secretion of procollagen. The influence of alphaalpha'-bipyridyl, colchicine and antimycin A on the secretory process in embryonic-chick tendon and cartilage cells. Biochem. J. 156: 81-90. Hautala,T., Byers, M.G., Eddy, R.L., Shows, T.B., Kivirikko, K.I. and Myllyla, R. (1992). Cloning of human lysyl hydroxylase: complete cDNA-derived amino acid sequence and assignment of the gene (PLOD) to chromosome lp36.3-—p36.2. Genomics 13: 62-69.  Hedgecock, E.M., Culotti, J.G., Hall, D.H. and Stern, B.D. (1987). Genetics of cell and axon migrations in Caenorhabditis elegans. Development. 100: 365-382.  Hedgecock, E.M., Culotti, J.G. and Hall, D.H. (1990). The unc-5, unc-6, and  unc-40  genes guide circumferential migrations of pioneer axons and mesodermal cells on the epidermis in C. elegans. Neuron. 4: 61-85.  Helbling-Leclerc, A., Zhang. X., Topaloglu, H., Cruaud, C , Tesson, F., Weissenbach, J., Tome, F.M., Schwartz, K., Fardeau, M., Tryggvason, K., et al. (1995). Mutations in the laminin alpha 2-chain gene (LAMA2) cause merosin-deficient congenital muscular dystrophy. Nat. Genet. 11: 216-218.  Henry, M.D. and Campbell, K.P. (1998). A role for dystroglycan in basement membrane assembly. Cell. 95: 859-870. Herman, R.K. (1978). Crossover suppressors and balanced recessive lethals in Caenorhabditis  elegans. Genetics 88: 49-65.  Hobert, O., Moerman, D.G., Clark, K.A., Beckerle, M.C. and Ruvkun, G. (19990. A conserved L I M protein that affects muscular adherens junction integrity and mechanosensory function in Caenorhabditis elegans. J. Cell Biol. 144: 45-57.  Horvitz, H.R., Brenner, S., Hodgkin J. and Herman, R.K. (1979). A uniform genetic nomenclature for the nematode Caenorhabditis elegans. Mol. Gen. Genet. 175: 129-133.  215  Hresko, M.C, Williams, B.D. and Waterston, R.H. (1994). Assembly of body wall muscle and muscle cell attachment structures in Caenorhabditis elegans. J. Cell Biol. 124: 491-506.  Hresko, M.C, Schriefer, L.A., Shrimankar, P. and Waterston, R.H. (1999). Myotactin, a novel hypodermal protein involved in muscle-cell adhesion in Caenorhabditis  elegans. J. Cell Biol. 146: 659-672.  Hu, R.J., Moorthy, S. and Bennett, V. (1995). Expression of functional domains of beta G-spectrin disrupts epithelial morphology in cultured cells. J. Cell Biol. 128: 10691080  Hynes, R.O. and Lander, A.D. (1992). Contact and adhesive specificities in the associations, migrations, and targeting of cells and axons. Cell 68: 303-322.  Kagawa, H., Takuwa, K. and Sakube, Y. (1997). Mutations and expressions of the tropomyosin gene and the troponin C gene of Caenorhabditis  elegans.  Cell Struct. Funct. 22:213-218.  Kaplan, M.R., Meyer-Franke, A., Lambert, S., Bennett, V., Duncan, I.D., Levinson, S.R. and Barres, B.A. (1997). Induction of sodium channel clustering by oligodendrocytes. Nature. 386: 724-728.  Karn, J., Brenner, S. and Barnett, L. (1983). Protein structural domains in the Caenorhabditis elegans unc-54 myosin heavy chain gene are not separated by introns. Proc. Natl. Acad. Sci. USA. 80: 4253-4257.  Kellokumpu, S., Sormunen, R., Heikkinen, J. and Myllyla, R. (1994). Lysyl hydroxylase, a collagen processing enzyme, exemplifies a novel class of luminallyoriented peripheral membrane proteins in the endoplasmic reticulum. J. Biol. Chem. 269: 30524-30529.  Kivirikko, K.I. and Pihlajaniemi, T. (1998). Collagen hydroxylases and the protein disulfide isomerase subunit of prolyl 4-hydroxylases. Adv. Enzymol. Relat. Areas Mol. Biol. 72: 325-398.  Kramer, J.M. (1994). Genetic analysis of extracellular matrix in C.  elegans.  Annu. Rev. Genet. 28: 95-116.  Kramer, J.M. (1997). Extracellular Matrix. In C. elegans II. D.L. Riddle, T. Blumenthal, B.J. Meyer, and J.R. Priess, editors. Cold Spring Harbor, N Y : Cold Spring Harbor Laboratory Press. 471-500.  Kiihn, K. (1994). Basement membrane (type IV) collagen.  216  Matrix Biol.  14: 439-445.  Lackner, M.R., Kornfeld, K., Miller, L.M., Horvitz, H.R. and Kim,S.K. (1994). A M A P kinase homolog, mpk-1, is involved in ras-mediated induction of vulval cell fates in Caenorhabditis  elegans.  Genes Dev. 8: 160-173.  Laurie, G.W., Bing, J.T., Kleinman, H.K., Hassell, J.R., Aumailley, M., Martin, G.R. and Feldmann, R.J. (1986). Localization of binding sites for laminin, heparan sulfate proteoglycan and fibronectin on basement membrane (type IV) collagen. J. Mol. Biol. 189: 205-216.  Lee, J.K., Brandin, E., Branton, D. and Goldstein, L.S. (1997a). alpha-Spectrin is required for ovarian follicle monolayer integrity in Drosophila Development. 124: 353-362.  melanogaster.  Lee, R.Y., Lobel, L., Hengartner, M., Horvitz, H.R. and Avery, L. (1997b). Mutations in the alphal subunit of an L-type voltage-activated Ca2+ channel cause myotonia in Caenorhabditis  elegans. EMBO J. j : 6066-6076.  Littlefield, R. and Fowler, V.M. (1998). Defining actin filament length in striated muscle: rulers and caps or dynamic stability? Annu. Rev. Cell Dev. Biol. 14: 487-525.  Lu, M.H., DiLullo, C., Schultheiss, T., Holtzer, S., Murray, J.M., Choi, J., Fischman, D.A. and Holtzer, H. (1992). The vinculin/sarcomeric-alpha-actinin/alpha-actin nexus in cultured cardiac myocytes. J. Cell Biol. 117: 1007-1022.  Lux, S.E. and Palek, J. (1995). Disorders of the Red Cell Membrane. Zn Blood: Principles and Practice of Hematology. R.I. Handin, S.E. Lux and T.P. Stossel, editors. J.B. Lippincott Company, Philadelphia, Pennsylvania. 1701-1818.  McDonald, K.A., Lakonishok, M. and Horwitz, A.F. (1995). Alpha v and alpha 3 integrin subunits are associated with myofibrils during myofibrillogenesis. J. Cell Sci. 108: 975-983.  McDowall, J.S. and Rose, A. (1997). Alignment of the genetic and physical maps in the dpy-5 bli-4 (I) region of C. elegans by the serial cosmid rescue of lethal mutations. Mol. Gen. Genet. 255: 78-95.  McKeown, C , Praitis, V. and Austin, J. (1998).  sma-1 encodes a betaH-spectrin homolog required for Caenorhabditis elegans morphogenesis. Development. 125: 20872098.  Mello, C. and Fire, A. (1995). D N A transformation.  Methods Cell Biol.  48: 451-482.  Michalak, M. and Opas, M. (1997). Functions of dystrophin and dystrophin associated proteins. Curr. Opin. Neurol. 10: 436-442.  217  Miller, D . M . 3d, Ortiz, I., Berliner, G . C . and Epstein, H . F . (1983). Differential localization of two myosins within nematode thick filaments. Cell 34: 477-490. Miller, D . M . and Shakes, D . C . (1995). Immunofluorescence microscopy. Methods Cell Biol. 48: 365-394.  Miller, L . M . , Waring, D.A. and K i m , S.K. (1996). Mosaic analysis using a ncl-1 (+) extrachromosomal array reveals that lin-31 acts in the Pn.p cells during Caenorhabditis elegans vulval development. Genetics. 143: 1181-1191. Milne, C A . and Hodgkin, J . (1999). ETR-1, a homologue of a protein linked to myotonic dystrophy, is essential for muscle development in Caenorhabditis elegans. Curr. Biol. 9. 1243-1246.  Moerman, D.G., Benian, G . M . and Waterston, R . H . (1986). Molecular cloning of the muscle gene unc-22 in Caenorhabditis elegans by T e l transposon tagging. Proc. Natl. Acad. Sci. USA.  83: 2579-2583.  Moerman, D.G., Benian, G . M . , Barstead, R.J., Schriefer, L . A . and Waterston, R . H . (1988). Identification and intracellular localization of the unc-22 gene product of Caenorhabditis  elegans. Genes Dev. 2: 93-105.  Moerman, D . G . and Fire, A . (1997). Muscle: structure, function and development. In C. elegans II. D.L. Riddle, T. Blumenthal, B.J. Meyer, and J.R. Priess, editors. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. 417-470. Moerman, D.G., Hutter, H . , Mullen, G.P.and Schnabel, R. (1996). Cell autonomous expression of perlecan and plasticity of cell shape in embryonic muscle of Caenorhabditis  elegans. Dev. Biol. 173: 228-242.  Moorthy, S., Chen, L . and Bennett V . (2000). Caenorhabditis elegans 6-G Spectrin Is Dispensable for Establishment of Epithelial Polarity, but Essential for Muscular and Neuronal Function. J. Cell Biol. 149: 915-930. Moulder, G . L . , Huang, M . M . , Waterston, R . H . and Barstead, R J . (1996) Talin requires beta-integrin, but not vinculin, for its assembly into focal adhesion-like structures in the nematode Caenorhabditis  elegans. Mol. Biol. Cell. 7: 1181-1193.  M u c h , J.W., Slade, D.J., Klampert, K . , Garriga, G . and Wightman, B . (2000). The fax-1 nuclear hormone receptor regulates axon pathfinding and neurotransmitter expression. Development. 127:703-712. Mullen, G.P., Rogalski, T . M . , Bush J.A., Gorji, P.R. and Moerman, D . G . (1999). Complex patterns of alternative splicing mediate the spatial and temporal distribution of perlecan/UNC-52 in Caenorhabditis  elegans. Mol. Biol. Cell. 10: 3205-3221.  218  Myers, C D , Goh P Y , Allen TS, Bucher E A and Bogaert T. (1996). Developmental genetic analysis of troponin T mutations in striated and nonstriated muscle cells of Caenorhabditis elegans. J. Cell Biol. 132: 1061-1077. Obinata, T. (1993). Contractile proteins and myofibrillogenesis. Int. Rev. Cytol. 143: 153-189. Otsuka, A . J . , Franco, R., Yang, B., Shim, K . H . , Tang, L . Z . , Zhang, Y . Y . , Boontrakulpoontawee, P., Jeyaprakash, A., Hedgecock, E., Wheaton, V . L , et al. (1995). A n ankyrin-related gene (unc-44) is necessary for proper axonal guidance in Caenorhabditis elegans. J. Cell Biol. 129: 1081-1092. Pardo, J.V., Siliciano, J.D. and Craig, S.W. (1983). A vinculin-containing cortical lattice in skeletal muscle: transverse lattice elements ("costameres") mark sites of attachment between myofibrils and sarcolemma. Proc. Natl. Acad. Sci. USA. 80: 10081012. Park, E . C . and Horvitz, H.R. (1986). Mutations with dominant effects on the behavior and morphology of the nematode Caenorhabditis elegans. Genetics. 113: 821-852. Passoja, K . , Myllyharju, J . , Pirskanen, A . and Kivirikko, K.I. (1998). Identification of arginine-700 as the residue that binds the C-5 carboxyl group of 2-oxoglutarate in human lysyl hydroxylase 1. FEBS Lett. 434: 145-148. Passoja, K . , Rautavuoma, K . , Ala-Kokko, L . , Kosonen, T. and Kivirikko, K.I. (1998). Cloning and characterization of a third human lysyl hydroxylase isoform. Proc. Natl. Acad. Sci. USA 95: 10482-10486. Pesacreta, T.C., Byers, T.J., Dubreuil, R., Kiehart, D.P. and Branton, D . (1989). Drosophila spectrin: the membrane skeleton during embryogenesis. J. Cell Biol 108: 1697-1709. Pirskanen, A . , Kaimio, A . M . , Myllyla, R. and Kivirikko, K.I. (1996). Site-directed mutagenesis of human lysyl hydroxylase expressed in insect cells. Identification of histidine residues and an aspartic acid residue critical for catalytic activity. J. Biol. Chem. 271: 9398-9402. Plenefisch, J.D., Zhu, X . and Hedgecock, E . M . (2000). Fragile skeletal muscle attachments in dystrophic mutants of Caenorhabditis elegans: isolation and characterization of the mua genes. Development. Ill: 1197-1207. Podbilewicz, B . and White, J . G . (1994). Cell fusions in the developing epithelial of C. elegans. Dev. Biol. 161: 408-424. Priess, J.R. and Hirsh, D.I. (1986). Caenorhabditis elegans morphogenesis: the role of the cytoskeleton in elongation of the embryo.  219  Dev. Biol. 117: 156-173.  Prockdp, D.J. and Kivirikko, K.I. (1995). Collagens: molecular biology, diseases, and potentials for therapy. Annu. Rev. Biochem. 64: 403-434.  Prokop, A., Martin-Bermudo, M.D., Bate, M. and Brown, N.H. (1998) Absence of PS integrins or laminin A affects extracellular adhesion, but not intracellular assembly, of hemiadherens and neuromuscular junctions in Drosophila embryos.Dev. Biol. 196: 5876.  Raymond, C.S., Shamu, C.E., Shen, M.M., Seifert, K.J., Hirsch, B., Hodgkin, J . and Zarkower, D. (1998). Evidence for evolutionary conservation of sex-determining genes. Nature 391: 691-695.  Riddle, D.L. (1988). The Dauer Larvae. In The Nematode Caenorhabditis elegans. W.B. Wood, editor. Cold Spring Harbor, N Y : Cold Spring Harbor Laboratory Press. 393-412.  Rogalski, T.M., Gilchrist, E.J., Mullen, G.P. and Moerman, D.G. (1995). Mutations in the unc-52 gene responsible for body wall muscle defects in adult Caenorhabditis elegans are located in alternatively spliced exons. Genetics 139: 159-169.  Rogalski TM, Moerman DG, Baillie DL. (1982). Essential genes and deficiencies in the unc-22 IV region of Caenorhabditis elegans. Genetics.  102: 725-736.  Rogalski, T.M., Williams, B.D., Mullen, G.P. and Moerman, D.G. (1993). Products of the unc-52 gene in Caenorhabditis elegans are homologous to the core protein of the mammalian basement membrane heparan sulfate proteoglycan. Genes Dev. 7: 1471-1484.  Roulier, E.M., Fyrberg, C . and Fyrberg, E . (1992) Perturbations of Drosophila alphaactinin cause muscle paralysis, weakness, and atrophy but do not confer obvious nonmuscle phenotypes. /. Cell Biol. 116: 911-922.  Rubin, G.M., Yandell, M.D., Wortman, J.R., Gabor Miklos, G.L., Nelson, C.R., Hariharan, I.K., Fortini, M.E., Li, P.W., Apweiler, R., Fleischmann, W., et al. (2000). Comparative genomics of the eukaryotes. Science. 287: 2204-2215.  Sambrook, J . , Fritsch, E.F. and Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, N Y : Cold Spring Harbor Laboratory Press. Schafer, W.R. and Kenyon, C J . (1995). A calcium-channel homologue required for adaptation to dopamine and serotonin in Caenorhabditis elegans. Nature. 375: 73-78.  220  Schnabel, R. and Priess, J.R. (1997). Specification of cell fates in the early embryo. In C. elegans II. D.L. Riddle, T. Blumenthal, B.J. Meyer, and J.R. Priess, editors. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. 361-382.  Shelton, C.A., Carter, J.C., Ellis, G.C. and Bowerman, B. (1999). The nonmuscle myosin regulatory light chain gene mlc-4 is required for cytokinesis, anterior-posterior polarity, and body morphology during Caenorhabditis elegans embryogenesis. / . Cell Biol. 146: 439-451.  Sibley, M.H., Graham, P.L., von Mende, N. and Kramer, J.M. (1994). Mutations in the alpha 2(IV) basement membrane collagen gene of Caenorhabditis elegans produce phenotypes of differing severities. EMBO J. 13: 3278-3285.  Sigurdson, D.C, Spanier, G.J. and Herman, R.K. (1984).  Caenorhabditis  elegans  deficiency mapping. Genetics. 108: 331-345.  Sikorski, A.F., Terlecki, G., Zagon, LS. and Goodman, S.R. (1991) Synapsin Imediated interaction of brain spectrin with synaptic vesicles. J. Cell Biol. 114: 313-318.  Sikorski, A.F., Sangerman, J., Goodman, S.R. and Critz, S.D. (2000). Spectrin (betaSpIIsigmal) is an essential component of synaptic transmission. Brain Res. 852: 161-166.  Smith, D.B. and Johnson, K.S. (1988). Single-step purification of polypeptides expressed in Escherichia coli as fusions with glutathione S-transferase. Gene. 67: 31-40.  Speicher, D.W. and Marchesi, V.T. (1984). Erythrocyte spectrin is comprised of many homologous triple helical segments. Nature. 311: 177-180.  Straub, K.L., Stella, M.C and Leptin, M. (1996). The gelsolin-related flightless I protein is required for actin distribution during cellularisation in Drosophila. 109: 263-270.  J. Cell Sci.  Sulston, J.E. and Hodgkin, J. (1988). Methods. In The Nematode Caenorhabditis elegans. W.B. Wood, editor. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. 587-606. Sulston, J.E. and Horvitz, H.R. (1977). Post-embryonic cell lineages of the nematode, Caenorhabditis  elegans. Dev. Biol. 56: 110-156.  Sulston, J.E., Schierenberg, E., White, J.G. and Thomson, J.N. (1983). The embryonic cell lineage of the nematode Caenorhabditis elegans. Dev. Biol. 100: 64-119.  Terami, H., Williams, B.D., Kitamura, S., Sakube, Y., Matsumoto, S., Doi, S., Obinata, T. and Kagawa, H. (1999). Genomic organization, expression, and analysis of the troponin C gene pat-10 of Caenorhabditis elegans. J. Cell Biol. 146: 193-202.  221  Thomas, G.H. and Kiehart, D.P. (1994). Beta heavy-spectrin has a restricted tissue and subcellular distribution during Drosophila embryogenesis. Development. 120: 20392050.  Thomas, G.H., Zarnescu, D.C., Juedes, A.E., Bales, M.A., Londergan, A., Korte, C.C. and Kiehart, D.P. (1998). Drosophila betaHeavy-spectrin is essential for development and contributes to specific cell fates in the eye. Development. 125: 2125-2134.  Timpl, R. (1996). Macromolecular organization of basement membranes.  Curr. Opin.  Cell Biol. 8: 618-624.  Timpl, R. and Brown, J.C. (1996). Supramolecular assembly of basement membranes. Bioessays 18: 123-132.  Tomishige, M., Sako, Y. and Kusumi, A. (1998). Regulation mechanism of the lateral diffusion of band 3 in erythrocyte membranes by the membrane skeleton. / . Cell Biol. 142: 989-1000.  Vaheri, A., Carpen, O., Heiska, L., Helander, T.S., Jaaskelainen, J., MajanderNordenswan, P., Sainio, M., Timonen, T. and Turunen, O. (1997). The ezrin protein family: membrane-cytoskeleton interactions and disease associations. Curr. Opin. Cell Biol. 9: 659-666.  Vallenius, T., Luukko, K. and Makela, T.P. (2000). CLP-36 P D Z - L I M Protein Associates with Nonmuscle alpha-Actinin-1 and alpha-Actinin-4. /. Biol. Chem. 275: 11100-11105.  Valtavaara, M., Papponen, H., Pirttila, A.M., Hiltunen, K., Helander, H. and Myllyla, R. (1997). Cloning and characterization of a novel human lysyl hydroxylase isoform highly expressed in pancreas and muscle. J. Biol. Chem. 272: 6831-6834.  Valtavaara, M., Szpirer, C , Szpirer, J. and Myllyla, R. (1998). Primary structure, tissue distribution, and chromosomal localization of a novel isoform of lysyl hydroxylase (lysyl hydroxylase 3). J. Biol. Chem. 1998 273: 12881-12886.  Venolia, L. and Waterston, R.H. (1990). The unc-45 gene of Caenorhabditis  elegans is  an essential muscle-affecting gene with maternal expression. Genetics. 126: 345-353.  Venolia, L., Ao, W., Kim, S., Kim, C. and Pilgrim D. (1999). Caenorhabditis  unc-45 gene of  elegans encodes a muscle-specific tetratricopeptide repeat-containing  protein. Cell Motil. Cytoskeleton. 42: 163-77.  Volk, T., Fessler, L.I. and Fessler, J.H. (1990). A role for integrin in the formation of sarcomeric cytoarchitecture. Cell. 63 :525-536.  222  Walmsley, A.R., Batten, M.R., Lad, U . and Bulleid, NJ. (1999). Intracellular retention of procollagen within the endoplasmic reticulum is mediated by prolyl 4hydroxylase. J. Biol. Chem. 274: 14884-14892.  Waterston, R.H., Thomson, J.N. and Brenner, S. (1980). Mutants with altered muscle structure of Caenorhabditis  elegans. Dev. Biol. 77: 271-302.  Waterston, R.H. (1988). Muscles. In The Nematode Caenorhabditis elegans. W.B. Wood, editor. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. 281-336. Waterston, R.H. (1989). The minor myosin heavy chain, mhcA, of  Caenorhabditis  elegans is necessary for the initiation of thick filament assembly. EMBO J. 8: 34293436. White, J.G. (1988). The Anatomy. In The Nematode Caenorhabditis elegans. W.B. Wood, editor. Cold Spring Harbor, N Y : Cold Spring Harbor Laboratory Press. 81-122.  White, J.G., Southgate, E . , Thomson, J.N. and Brenner, S. (1976). The structure of the ventral nerve cord of Caenorhabditis elegans. Philos. Trans. R. Soc. Lond. B. Biol.  Sci. 275: 327-348.  White, J.G., Southgate, E . , Thomson, J.N. and Brenner, S. (1986). The structure of the nervous system of Caenorhabditis elegans. Philos. Trans. R. Soc. Lond. B. Biol. Sci.  314: 1-340.  Williams, B.D. and Waterston, R.H. (1994). Genes critical for muscle development and function in Caenorhabditis elegans identified through lethal mutations. J. Cell Biol. 124: 475-490.  Williams, M.W. and Bloch, RJ. (1999). Extensive but coordinated reorganization of the membrane skeleton in myofibers of dystrophic (mdx) mice. J. Cell Biol. 144: 12591270. Wissmann, A., Ingles, J., McGhee, J.D. and Mains, P . E . (1997).  Caenorhabditis  elegans LET-502 is related to Rho-binding kinases and human myotonic dystrophy kinase and interacts genetically with a homolog of the regulatory subunit of smooth muscle myosin phosphatase to affect cell shape. Genes Dev. 11: 409-422.  Wood, W.B. (1988). Embryology. In The Nematode Caenorhabditis elegans. W.B. Wood, editor. Cold Spring Harbor, N Y : Cold Spring Harbor Laboratory Press. 215-241.  Wood, SJ. and Slater, CR. (1998). beta-Spectrin is colocalized with both voltagegated sodium channels and ankyrinG at the adult rat neuromuscular junction. J. Cell Biol. 140: 675-684.  223  Xia, H., Winokur, S.T., Kuo, W.L., Altherr, M.R. and Bredt, D.S. (1997). Actininassociated L I M protein: identification of a domain interaction between PDZ and spectrinlike repeat motifs. J. Cell Biol. 139: 507-515.  Yan Y, Winograd E, Viel A, Cronin T, Harrison SC, Branton D. (1993). Crystal structure of the repetitive segments of spectrin. Science. 262: 2027-2030.  Yeaman, C , Grindstaff, K.K., Nelson, WJ. (1999). New perspectives on mechanisms involved in generating epithelial cell polarity. Physiol. Rev. 79: 73-98.  Yeowell, H.N., Ha, V., Walker, L.C., Murad, S. and Pinnell, S.R. (1992). Characterization of a partial c D N A for lysyl hydroxylase from human skin fibroblasts; lysyl hydroxylase mRNAs are regulated differently by minoxidil derivatives and hydralazine. J. Invest. Dermatol. 99: 864-869.  Yurchenco, P.D., Cheng, Y.S. and Colognato, H. (1992). Laminin forms an independent network in basement membranes. I. Cell Biol. 117: 1119-1133.  Yurchenco, P.D. and Furthmayr, H. (1984). Self-assembly of basement membrane collagen. Biochemistry 2: 1839-1850.  Yurchenco, P.D. and O'Rear, JJ. (1994). Basal lamina assembly.  Curr. Opin. Cell  Biol. 6: 674-681.  Yurchenco, P.D. and Schittny, J.C. (1990). Molecular architecture of basement membranes. FASEB1.  4: 1577-1590.  Zengel, J.M. and Epstein, H.F. (1980). Identification of genetic elements associated with muscle structure in the nematode Caenorhabditis elegans. Cell Motil. 1: 73-97.  Zhou, D., Ursitti, J.A. and Bloch, RJ. (1998). Developmental expression of spectrins in rat skeletal muscle. Mol. Biol. Cell. 9: 47-61.  224  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0099579/manifest

Comment

Related Items