A Limited Number of Caenorhabditis Elegans Genes Are Readily Mutable to Dominant, Temperature-Sensitive Maternal-Effect Embryonic Lethality

Dominant gain-of-function mutations can give unique insights into the study of gene function. In addition, gain-of-function mutations, unlike loss-of-function alleles, are not biased against the identification of genetically redundant loci. To identify novel genetic functions active during Caenorhabditis elegans embryogenesis, we have collected a set of dominant temperature-sensitive maternal-effect embryonic lethal mutations. In a previous screen, we isolated eight such mutations, distributed among six genes. In the present study, we describe eight new dominant mutations that identify only three additional genes, yielding a total of 16 dominant mutations found in nine genes. Therefore, it appears that a limited number of C. elegans genes mutate to this phenotype at appreciable frequencies. Five of the genes that we identified by dominant mutations have loss-of-function alleles. Two of these genes may lack loss-of-function phenotypes, indicating that they are nonessential and so may represent redundant loci. Loss-of-function mutations of three other genes are associated with recessive lethality, indicating nonredundancy.     

Molecular analysis of mutations in the Caenorhabditis elegans collagen gene dpy-7.

Collagens are a family of proteins contributing to the body structure of eukaryotes. They are encoded by a large and diverse gene family in the nematode Caenorhabditis elegans but by only a few genes in vertebrates. We have studied mutant alleles of the C. elegans dpy-7 gene, one of a large group of genes whose mutant phenotype is altered body form and several of which have previously been shown to encode cuticular collagens. We made use of the C. elegans physical map to screen specifically for collagen genes in the region of the X chromosome to which dpy-7 maps. This yielded a wild-type collagen gene clone which we showed, by micro-injection, could repair the dpy-7 mutant phenotype in transgenic animals. We cloned the homologous sequence from four dpy-7 mutant strains and by sequence analysis identified a single mutation in each case. All four mutations result in the substitution of a glycine with a larger residue in the conserved Gly-X-Y collagen domains. Similar substitutions in vertebrate collagens cause the heritable brittle bone disorder osteogenesis imperfecta. Whereas the human mutations are dominant, the dpy-7 mutations are recessive, and this may reflect different levels of complexity of collagenous macromolecular structures in the two organisms.     

Analysis of Dominant Mutations Affecting Muscle Excitation in Caenorhabditis Elegans

We examined mutations that disrupt muscle activation in Caenorhabditis elegans. Fifteen of 17 of these genes were identified previously and we describe new mutations in three of them. We also describe mutations in two new genes, exp-3 and exp-4. We assessed the degree of defect in pharyngeal, body-wall, egg-laying, and enteric muscle activation in animals mutant for each gene. Mutations in all 17 genes are semidominant and, in cases that could be tested, appear to be gain-of-function. Based on their phenotypes, the genes fall into three broad categories: mutations in 11 genes cause defective muscle activation, mutations in four genes cause hyperactivated muscle, and mutations in two genes cause defective activation in some muscle types and hyperactivation in others. In all testable cases, the mutations blocked response to pharmacological activators of egg laying, but did not block muscle activation by irradiation with a laser microbeam. The data suggest that these mutations affect muscle excitation, but not the capacity of the muscle fibers to contract. For most of the genes, apparent loss-of-function mutants have a grossly wild-type phenotype. These observations suggest that there is a large group of genes that function in muscle excitation that can be identified primarily by dominant mutations.     

Genetic Studies of Unusual Loci That Affect Body Shape of the Nematode CAENORHABDITIS ELEGANS and May Code for Cuticle Structural Proteins

In Caenorhabditis elegans, four loci (sqt-1, sqt-2, sqt-3 and rol-8) in which mutations affect body shape and cuticle morphology have unusual genetic properties. Mutant alleles of sqt-1 can interact to produce animals with a variety of mutant phenotypes: left roller, right roller, dumpy and long. At least three mutant phenotypes are specified by mutations in the sqt-3 locus. Most alleles at these loci are either dominant or cryptic dominant (i.e., are dominant only in certain genetic backgrounds). Most alleles of these loci exhibit codominance. Two putative null alleles of the sqt-1 locus produce a wild-type phenotype. Many alleles of these genes demonstrate unusual intergenic interactions that are not the result of simple epistasis: animals doubly heterozygous for mutations at two loci often display unexpected and unpredictable phenotypes. We suggest that these genetic properties might be expected of genes, such as the collagen genes, the products of which interact to form the animal's cuticle, and which are member genes of a gene family.     

Identification and Characterization of 22 Genes That Affect the Vulval Cell Lineages of the Nematode CAENORHABDITIS ELEGANS

Ninety-five mutants of the nematode Caenorhabditis elegans altered in the cell lineages of the vulva have been isolated on the basis of their displaying one of two phenotypes, Vulvaless or Multivulva. In Vulvaless mutants, which define 12 genes, no vulva is present. In Multivulva mutants, which define ten genes, one or more supernumerary vulva-like protrusions are located along the ventral side of the animal. A single recessive mutation is responsible for the phenotypes of most, but not all, of these strains. Fifteen of these 22 genes are represented by multiple alleles. We have shown by a variety of genetic criteria that mutations that result in a Vulvaless or Multivulva phenotype in six of the 22 genes most likely eliminate gene function. In addition, Vulvaless or Multivulva mutations in seven of the other genes most likely result in a partial reduction of gene function; the absence of the activity of any of these genes probably results in lethality or sterility. Our results suggest that we may have identified most, or all, genes of these two classes.     

The Multivulva Phenotype of Certain Caenorhabditis Elegans Mutants Results from Defects in Two Functionally Redundant Pathways

We previously identified Caenorhabditis elegans mutants in which certain of the six vulval precursor cells adopt fates normally expressed by other vulval precursor cells. These mutants define genes that appear to function in the response to an intercellular signal that induces vulval development. The multivulva (Muv) phenotype of one such mutant, CB1322, results from an interaction between two unlinked mutations, lin-8(n111) II and lin-9(n112) III. In this paper, we identify 18 new mutations, which are alleles of eight genes, that interact with either lin-8(n111) or lin-9(n112) to generate a Muv phenotype. None of these 20 mutations alone causes any vulval cell lineage defects. The "silent Muv" mutations fall into two classes; hermaphrodites carrying a mutation of each class are Muv, while hermaphrodites carrying two mutations of the same class have a wild-type vulval phenotype. Our results indicate that the Muv phenotype of these mutants results from defects in two functionally-redundant pathways, thereby demonstrating that redundancy can occur at the level of gene pathways as well as at the level of gene families.     

Genes That Implement the Hermaphrodite Mode of Dosage Compensation in Caenorhabditis Elegans

We report a genetic characterization of several essential components of the dosage compensation process in Caenorhabditis elegans. Mutations in the genes dpy-26, dpy-27, dpy-28, and the newly identified gene dpy-29 disrupt dosage compensation, resulting in elevated X-linked gene expression in XX animals and an incompletely penetrant maternal-effect XX-specific lethality. These dpy mutations appear to cause XX animals to express each set of X-linked genes at a level appropriate for XO animals. XO dpy animals are essentially wild type. Both the viability and the level of X-linked gene expression in XX animals carrying mutations in two or more dpy genes are the same as in animals carrying only a single mutation, consistent with the view that these genes act together in a single process (dosage compensation). To define a potential time of action for the gene dpd-28 we performed reciprocal temperature-shift experiments with a heat sensitive allele. The temperature-sensitive period for lethality begins 5 hr after fertilization at the 300-cell stage and extends to about 9 hr, a point well beyond the end of cell proliferation. This temperature-sensitive period suggests that dosage compensation is functioning in XX animals by mid-embryogenesis, when many zygotically transcribed genes are active. While mutations in the dpy genes have no effect on the sexual phenotype of otherwise wild-type XX or XO animals, they do have a slight feminizing effect on animals whose sex-determination process is already genetically perturbed. The opposite directions of the feminizing effects on sex determination and the masculinizing effects on dosage compensation caused by the dpy mutations are inconsistent with the wild-type dpy genes acting to coordinately control both processes. Instead, the feminizing effects are most likely an indirect consequence of disruptions in dosage compensation caused by the dpy mutations. Based on the cumulative evidence, the likely mechanism of dosage compensation in C. elegans involves reducing X-linked gene expression in XX animals to equal that in XO animals via the action of the dpy genes.     

Caenorhabditis Elegans Mutants Defective in the Functioning of the Motor Neurons Responsible for Egg Laying

We have isolated and characterized 45 Caenorhabditis elegans mutants presumed to be defective in the functioning of the hermaphrodite-specific neurons (HSNs). Like hermaphrodites that lack the HSN motor neurons, these mutants are egg-laying defective and do not lay eggs in response to exogenous imipramine but do lay eggs in response to exogenous serotonin. Twenty of the 45 mutations define 10 new egl genes; the other 25 mutations are alleles of five previously defined genes, four of which are known to affect the HSNs. Seven mutations in three genes cause the HSNs to die in hermaphrodites, as they normally do in males. These genes appear to be involved in the determination of the sexual phenotype of the HSNs, and one of them (egl-41) is a newly identified gene that may function generally in sex determination. Five of the 15 genes are defined only by mutations that have dominant effects on egg laying. One gene egl(n1108), is defined by a temperature-sensitive allele that has a temperature-sensitive period after HSN development is complete, suggesting that egl(n1108) may be involved in HSN synaptic transmission. Four of the genes are defined by single alleles, which suggests that other such genes remain to be discovered. Mutations in no more than 4 of the 15 genes specifically affect the HSNs, indicating that there are few genes with functions needed only in this single type of nerve cell.     

The Unc-8 and Sup-40 Genes Regulate Ion Channel Function in Caenorhabditis Elegans Motorneurons

Two Caenorhabditis elegans genes, unc-8 and sup-40, have been newly identified, by genetic criteria, as regulating ion channel function in motorneurons. Two dominant unc-8 alleles cause motorneuron swelling similar to that of other neuronal types in dominant mutants of the deg-1 gene family, which is homologous to a mammalian gene family encoding amiloride-sensitive sodium channel subunits. As for previously identified deg-1 family members, unc-8 dominant mutations are recessively suppressed by mutations in the mec-6 gene, which probably encodes a second type of channel component. An unusual dominant mutation, sup-41 (lb125), also co-suppresses unc-8 and deg-1, suggesting the existence of yet another common component of ion channels containing unc-8 or deg-1 subunits. Dominant, transacting, intragenic suppressor mutations have been isolated for both unc-8 and deg-1, consistent with the idea that, like their mammalian homologues, the two gene products function as multimers. The sup-40 (lb130) mutation dominantly suppresses unc-8 motorneuron swelling and produces a novel swelling phenotype in hypodermal nuclei. sup-40 may encode an ion channel component or regulator that can correct the osmotic defect caused by abnormal unc-8 channels.     

spe-29 encodes a small predicted membrane protein required for the initiation of sperm activation in Caenorhabditis elegans

Caenorhabditis elegans spermatids complete a dramatic morphogenesis to crawling spermatozoa in the absence of an actin- or tubulin-based cytoskeleton and without synthesizing new gene products. Mutations in three genes (spe-8, spe-12, and spe-27) prevent the initiation of this morphogenesis, termed activation. Males with mutations in any of these genes are fertile. By contrast, mutant hermaphrodites are self-sterile when unmated due to a failure in spermatid activation. Intriguingly, mutant hermaphrodites form functional spermatozoa and become self-fertile upon mating, suggesting that spermatids can be activated by male seminal fluid. Here we describe a mutation in a fourth gene, spe-29, which mimics the phenotype of spe-8, spe-12, and spe-27 mutants. spe-29 sperm are defective in the initiation of hermaphrodite sperm activation, yet they maintain the ability to complete the morphogenetic rearrangements that follow. Mutant alleles of spe-12, spe-27, and spe-29 exhibit genetic interactions that suggest that the wild-type products of these genes function in a common signaling pathway to initiate sperm activation. We have identified the spe-29 gene, which is expressed specifically in the sperm-producing germ line and is predicted to encode a small, novel transmembrane protein.     

Analysis of Mutations in the Sqt-1 and Rol-6 Collagen Genes of Caenorhabditis Elegans

Different mutations in the sqt-1 and rol-6 collagen genes of Caenorhabditis elegans can cause diverse changes in body morphology and display different genetic attributes. We have determined the nucleotide alterations in 15 mutant alleles of these genes. Three mutations in sqt-1 and one in rol-6 that cause dominant right-handed helical twisting (RRol) of animals are arginine to cysteine replacements. These mutations are all within a short conserved sequence, on the amino terminal side of the Gly-X-Y repeats, that is found in all C. elegans cuticle collagens. A recessive RRol mutation of rol-6 is a replacement of one of the same conserved arginines by histidine. In contrast, three sqt-1 mutations that cause recessive left-handed helical twisting (LRol) are replacements of a conserved carboxyterminal cysteine residue with either tyrosine or serine. These results suggest that disulfide bonding is important in collagen organization and that a deficit or surplus of disulfides may cause cuticle alterations of opposite handedness. In contrast to other collagens, glycine replacement mutations in the Gly-X-Y repeats of sqt-1 cause very mild phenotypes. Nonsense mutations of both sqt-1 and rol-6 cause nearly, but not totally, wild-type phenotypes. A nonsense mutation in sqt-1 suppresses the phenotype of rol-6 RRol mutations, suggesting that rol-6 collagen function is dependent on the presence of sqt-1 collagen. Mutations of sqt-1 are not suppressed by a rol-6 nonsense mutation, however, indicating that sqt-1 collagen can function independently of rol-6.     

Properties of a Class of Genes Required for Ray Morphogenesis in Caenorhabditis Elegans

We have identified eight mutations that define at least five terminal differentiation genes (ram genes) whose products are required during the extension of the male-specific ray sensilla in Caenorhabditis elegans. ram gene mutations result in morphological abnormalities in the sensory rays but do not appear to interfere with ray functions. A similar ray morphology phenotype was observed in males harboring mutations in three previously defined genes, dpy-11, dpy-18 and sqt-1, that also affect body shape. One of these genes, sqt-1, is known to encode a collagen. Mutations in different ram genes failed to complement, from which we infer that their gene products functionally interact. For one ram gene, failure to complement was shown to result from haploinsufficiency. Intergenic noncomplementation did not extend to the body morphology genes. The temperature-sensitive periods of both ram and body morphology mutations corresponded to the period of development in which ray extension occurs. We propose that ram gene products act together in a critical interaction between the rays and the cuticle required for wild-type ray morphology.     

C. ELEGANS unc-105 Mutations Affect Muscle and Are Suppressed by Other Mutations That Affect Muscle

Certain mutations in the unc-105 II gene of the nematode Caenorhabditis elegans have dominant effects on morphology and behavior: animals become small, severely hypercontracted and paralyzed. These unc-105 mutants revert both spontaneously and with mutagens at high frequencies to a wild-type phenotype. Most of the reversion events are intragenic, apparently because the null (loss-of-function) phenotype of unc-105 is wild type. One revertant defined an extragenic suppressor locus, sup-20 X. Such suppressor alleles of sup-20 are rare, and the apparent null phenotype of sup-20 is embryonic lethality. By constructing animals genetically mosaic for sup-20, we have shown that the primary effect of sup-20 is in muscle cells. In addition to mutations in sup-20, other mutations causing muscle defects, such as unc-54 and unc-22 mutations, suppress the hypercontracted phenotype of unc-105. The ease of identifying nonhypercontracted revertants of unc-105 mutants greatly facilitates the isolation of new mutants defective in muscle structure and function.     

Multiple levels of redundant processes inhibit Caenorhabditis elegans vulval cell fates

Many mutations cause obvious abnormalities only when combined with other mutations. Such synthetic interactions can be the result of redundant gene functions. In Caenorhabditis elegans, the synthetic multivulva (synMuv) genes have been grouped into multiple classes that redundantly inhibit vulval cell fates. Animals with one or more mutations of the same class undergo wild-type vulval development, whereas animals with mutations of any two classes have a multivulva phenotype. By varying temperature and genetic background, we determined that mutations in most synMuv genes within a single synMuv class enhance each other. However, in a few cases no enhancement was observed. For example, mutations that affect an Mi2 homolog and a histone methyltransferase are of the same class and do not show enhancement. We suggest that such sets of genes function together in vivo and in at least some cases encode proteins that interact physically. The approach of genetic enhancement can be applied more broadly to identify potential protein complexes as well as redundant processes or pathways. Many synMuv genes are evolutionarily conserved, and the genetic relationships we have identified might define the functions not only of synMuv genes in C. elegans but also of their homologs in other organisms.     

Charcot-Marie-Tooth-linked mutant GARS is toxic to peripheral neurons independent of wild-type GARS levels

Charcot-Marie-Tooth disease type 2D (CMT2D) is a dominantly inherited peripheral neuropathy caused by missense mutations in the glycyl-tRNA synthetase gene (GARS). In addition to GARS, mutations in three other tRNA synthetase genes cause similar neuropathies, although the underlying mechanisms are not fully understood. To address this, we generated transgenic mice that ubiquitously over-express wild-type GARS and crossed them to two dominant mouse models of CMT2D to distinguish loss-of-function and gain-of-function mechanisms. Over-expression of wild-type GARS does not improve the neuropathy phenotype in heterozygous Gars mutant mice, as determined by histological, functional, and behavioral tests. Transgenic GARS is able to rescue a pathological point mutation as a homozygote or in complementation tests with a Gars null allele, demonstrating the functionality of the transgene and revealing a recessive loss-of-function component of the point mutation. Missense mutations as transgene-rescued homozygotes or compound heterozygotes have a more severe neuropathy than heterozygotes, indicating that increased dosage of the disease-causing alleles results in a more severe neurological phenotype, even in the presence of a wild-type transgene. We conclude that, although missense mutations of Gars may cause some loss of function, the dominant neuropathy phenotype observed in mice is caused by a dose-dependent gain of function that is not mitigated by over-expression of functional wild-type protein.     

The Let-60 Locus Controls the Switch between Vulval and Nonvulval Cell Fates in Caenorhabditis Elegans

During induction of the Caenorhabditis elegans hermaphrodite vulva by the anchor cell of the gonad, six multipotent vulval precursor cells (VPCs) have two distinct fates: three VPCs generate the vulva and the other three VPCs generate nonspecialized hypodermis. Genes that control the fates of the VPCs in response to the anchor cell signal are defined by mutations that cause all six VPCs to generate vulval tissue (Multivulva or Muv) or that cause all six VPCs to generate hypodermis (Vulvaless or Vul). Seven dominant Vul mutations were isolated as dominant suppressors of a lin-15 Muv mutation. These mutations are dominant alleles of the gene let-60, previously identified only by recessive lethal mutations. Our genetic studies of these dominant Vul recessive lethal mutations, recessive lethal mutations, intragenic revertants of the dominant Vul mutations, and the closely mapping semi-dominant multivulva lin-34 mutations suggest that: (1) loss-of-function mutations of let-60 are recessive lethal at a larval stage, but they also cause a Vul phenotype if the lethality is rescued maternally by a lin-34 gain-of-function mutation. (2) The dominant Vul alleles of let-60 are dominant negative mutations whose gene products compete with wild-type activity. (3) lin-34 semidominant Muv alleles are either gain-of-function mutations of let-60 or gain-of-function mutations of an intimately related gene that elevates let-60 activity. We propose that let-60 activity controls VPC fates. In a wild-type animal, reception by a VPC of inductive signal activates let-60, and it generates into a vulval cell type; in absence of inductive signal, let-60 activity is low and the VPC generates hypodermal cells. Our genetic interaction studies suggest that let-60 acts downstream of let-23 and lin-15 and upstream of lin-1 and lin-12 in the genetic pathway specifying the switch between vulval and nonvulval cell types.     

Mutations in the alpha 2(IV) basement membrane collagen gene of Caenorhabditis elegans produce phenotypes of differing severities.

Type IV collagen forms a network that provides the major structural support of basement membranes. We have determined the nucleotide alterations and phenotypes of 17 mutant alleles of the Caenorhabditis elegans alpha 2(IV) collagen gene let-2. All 17 mutations are within the triple helical (Gly-X-Y) repeat domain of the molecule. Fifteen of the mutations are replacements of Gly-X-Y repeat glycines with aspartate, glutamate or arginine, and they cause a wide range of phenotypes. The mildest alleles are nearly wild-type at 15 and 20 degrees C but embryonic lethal at 25 degrees C, while the most severe allele is embryonic lethal at all three temperatures. Mutations resulting in severe phenotypes are generally located in areas of lower calculated thermal stability of the type IV collagen molecule. An alanine to threonine substitution at position X of a Gly-X-Y triplet immediately following an interruption results in a severe phenotype. This mutation is unusual because substitutions at positions X or Y have not generally been found to cause strong phenotypes in C. elegans or human collagens. An intron splice acceptor mutation causes a strict embryonic lethal phenotype, but does not completely abolish gene function. Pairs of independent mutations affect each of three glycines, indicating a non-random distribution of mutations in the molecule. It is suggested that this clustering results because many glycine substitutions may cause dominant lethal or sterile phenotypes.     

Wild-type and mutant actin genes in Caenorhabditis elegans

We have sequenced the four actin genes of Caenorhabditis elegans. These four genes encode typical invertebrate actins and are highly homologous, differing from each other by, at most, three amino acid residues. As a first step toward an understanding of the developmental regulation of this gene set we have also sequenced mutant actin genes. The mutant genes were cloned from two independent revertants of a single dominant actin mutant. For both revertants, reversion was accompanied by an actin gene rearrangement. The accumulation of actin mRNA during development in these two revertants is different from that of wild-type animals. We present here a correlation between actin gene structure and expression in wild-type and mutant animals. The results, suggest that co-ordinate regulation of actin genes is not essential for wild-type muscle function. In addition, it appears that changes in the 3' region of at least one of the actin mRNA may affect its steady-state regulation during development.     

Lethal and Amanitin-Resistance Mutations in the Caenorhabditis Elegans Ama-1 and Ama-2 Genes

Mutants of Caenorhabditis elegans resistant to alpha-amanitin have been isolated at a frequency of about 1.6 x 10(-6) after EMS mutagenesis of the wild-type strain, N2. Four new dominant resistance mutations have been studied genetically. Three are alleles of a previously identified gene, ama-1 IV, encoding the largest subunit of RNA polymerase II. The fourth mutation defines a new gene, ama-2 V. Unlike the ama-1 alleles, the ama-2 mutation exhibits a recessive-lethal phenotype. Growth and reproduction of N2 was inhibited at a concentration of 10 micrograms/ml amanitin, whereas ama-2/+ animals were inhibited at 100 micrograms/ml, and 800 micrograms/ml was required to inhibit growth of ama-1/+ larvae. We have also determined that two reference strains used for genetic mapping, dpy-11(e224)V and sma-1(e30)V, are at least four-fold more sensitive to amanitin that the wild-type strain. Using an amanitin-resistant ama-1(m118) or ama-1(m322) strain as a parent, we have isolated amanitin-sensitive mutants that carry recessive-lethal ama-1 alleles. The frequency of EMS-induced lethal ama-1 mutations is approximately 1.7 x 10(-3), 1000-fold higher than the frequency of amanitin-resistance alleles. Nine of the lethal alleles are apparent null mutations, and they exhibit L1-lethal phenotypes at both 20 degrees and 25 degrees. Six alleles result in partial loss of RNA polymerase II function as determined by their sterile phenotypes at 20 degrees. All but one of these latter mutations exhibit a more severe phenotype at 25 degrees C. We have also selected seven EMS-induced revertants of three different ama-1 lethals. These revertants restore dominant resistance to amanitin. The selection for revertants also produced eight new dominant amanitin resistance alleles on the balancer chromosome, nT1.