Zarkower D et al. (1991) International C. elegans Meeting "THE tra-1 GENE ENCODES TWO ZINC FINGER PROTEINS."

Hodgkin JA, Zarkower D

[International C. elegans Meeting, 1991]

The tra-l gene is the terminal gene in the regulatory pathway controlling sexual differentiation in C. elegans. Mutational studies have suggested that tra-l is necessary and sufficient for female somatic development and that it has a complex role in the germ line. We have examined transcription from a region shown to contain tra-l based on the detection by clones from the region of rearrangements that cosegregate with several tra- 1 mutations. These rearrangements suggest that the gene spans at least 25 kb. The tra-l region encodes at least two transcripts. One is a 1.4 kb RNA that is present only during L2. The second is a 5 kb RNA present during all hermaphrodite stages and in adult males. Both RNAs are of low abundance. There may also be a third RNA of 1 kb in adult males. We have not yet examined larval males. Northern blot analysis suggests that the 1.4 kb RNA is spliced from a primary transcript of about 13 kb and the 5 kb RNA from a primary transcript of at least 23 kb. We have isolated and sequenced cDNAs representing the 1.4 kb RNA from the Kim library and by PCR. The 1.4 kb RNA can encode a protein of about 340 amino acids with two C- terminal 'zinc-finger' motifs. The fingers are 45% identical to those of GLI, a human gene amplified in certain glioblastomas, and to cubitus interruptus Dominant, a Drosophila segment polarity gene. The non-finger sequence is serine-rich and shows no similarity to other proteins in the databases. Genomic sequencing and PCR suggest that the 5 kb RNA is colinear with the 1.4 kb RNA until the end of the second finger. It is then spliced to include two more fingers that are 80% identical to the corresponding GLI and cubitus interruptus fingers. By analogy to the other two genes, there is likely to be a fifth finger in the tra-l product. cDNA clones of the 5 kb RNA have proved elusive. To help define the functions of the tra-l products, we have examined RNA from tra-l mutants. One mutation, el488, a rearrangement upstream of the predicted start of transcription, appears to reduce levels of the 1.4 kb RNA. Since el488 animals have a hermaphrodite gonad and germline in a masculinized soma, the product of the 1.4 kb RNA may function primarily in the nongonadal soma. We have also analyzed a number of stronger mutations in strains carrying eDp6 (a free duplication covering tra-l) and have detected transcripts of altered size. These are harder to interpret due to the presence of wild type transcripts from eDp6. To overcome this problem we are analyzing RNA from single homozygous animals by PCR.

Yandell MD et al. (1991) International C. elegans Meeting "PSORALEN CAUSES SMALL UNC-22 DELETIONS AT A HIGH FREQUENCY"

Edgar LG, Wood WB, Yandell MD

[International C. elegans Meeting, 1991]

We have analysed 23 trimethylpsoralen (TMP)-induced alleles of unc- 22 by Southern blot for RFLPs, using the probes DM17, DM18, DM20, DM22, SL13, and SL14 (many thanks to Carol Trent and Guy Benian!), which cover the entire coding region and about 11 kb of 5' upstream sequence. Eighteen of these mutations were obtained in a single mutagenesis, screening for both homozygotes and heterozygotes (mutation rate of lx10-4). In this mutagenesis, 15 strains were homozygous-viable, and 8 were originally homozygous-lethal. Two of these, however, were homozygous-viable after backcrossing; a third was successfully balanced over sDf22, and 5 were lost before analysis. Of the mutants analysed (the 18 remaining strains and an additional 4 homozygous- viable mutations isolated in previous TMP screens), 9 of the homozygous-viable mutants and the balanced homozygous-lethal mutant show visible polymorphisms within the unc-22 locus. All but one, which remains-ambiguous, were deletions ranging from 0.15 to 10 kb. The average deletion size was 2.2 kb; if the single 10-kb deletion is not included the average size was only 0.94 kb. Therefore, among mutations analyzed from the inclusive screen for both homozygotes and heterozygotes, 44% (8/18) were deletions of less than 10 kb in size. These data suggest that TMP produces small deletions at a higher frequency than any other mutagen in current use, which should be useful in generating mutations causing complete loss of function and likely to be associated with DNA polymorphisms.

Lee RC et al. (1991) International C. elegans Meeting "CLONING OF THE HETEROCHRONIC GENE lin-4 BY POLYMORPHISM MAPPING AND TRANSFORMATION RESCUE."

Ambros VR, Lee RC, Feinbaum RL

[International C. elegans Meeting, 1991]

We genetically mapped the lin4 locus to a position on LGII closely linked and to the left of bristol/bergerac polymorphism rnaP13, detected by YAC Y42A4. Using YAC and cosmid clones in the region as probes to southern blots, we analysed DNA from the only lin4 allele available at the time, e912, (see Feinbaum and Ambros, these abstracts) and found that the YAC Y18Cl detected a polymorphism in a 5 kb EcoRI band. Cosmids drawn under Y18Cl on the physical map did not detect this band, so we suspected that lin-4 was at the right end of Y18Cl, which extended into a cosmid-free gap. Therefore, to clone the sequences corresponding to the 5 kb EcoRI band, we used Y18Cl, which detects the 5 kb band, and Y42A4, which almost completely overlaps Y18Cl but does not detect the band, in a differential screen of an EcoRl lambda library prepared by Ann Rougvie. In this way we isolated two different 5 kb EcoRI fragments that both turn out to be partially deleted in lin4(e912) DNA. These EcoRI clones lie not in the gap, but in the contig and to the left of the previous end of Y18Cl. We have therefore revised the extent of Y18Cl to include the cosmids containing these two 5 kb EcoRI fragments. One of the polymorphic 5 kb EcoRI clones, pRL2D, rescues the mutant phenotype of lin-4(e912J, as does an approximately 2.9 kb Cla-l subclone of pRL2d, pRL2DCla. (Two overlapping cosmids, C02B6 and C31Gll, also rescue lin4). Therefore, pRL2DCla contains a functional lin4 gene, although the lin-4(e912) lesion, originally generated by 32P decay, is somewhat complex, involving deletion of more than 5 kb of additional sequence in the region, and partial duplication of another sequence. We are currently sequencing pRL2DCla and testing smaller subclones for rescue. cDNAs that hybridize to the lin-4 rescuing fragments are also under analysis.

Labouesse M et al. (1991) International C. elegans Meeting "THE lin-26 PUTATIVE ZINC FINGER PROTEIN MIGHT BE A HYPODERMAL DIFFERENTIATION FACTOR."

Labouesse M, Horvitz HR

[International C. elegans Meeting, 1991]

During C. elegans development, many blast cells generate both neurons and hypodermal cells. What factors make neurons and hypodermal cells dlfferent? To address this question. we are studying the mutatlon Ifn-26tnl56), whlch transforms the normal hypodermal cell fate of the Pn.p cells to a neuronal fate slmllar to that of thelr llneal slsters. the Pn.a cells (see Labouesse and Horvitz, 1990, WBG 11 (4), p. 70). We have cloned the gene Ifn-26 by mapplng deflclency endpolnts and by germilne transformatlon. A 9 kb reglon of cosmid C18C9 rescues nl56. This region encodes at least four transcripts: three RNAs (1.4 kb, 1.45 kb and 1.5 kb) arise via altemative splicing near thelr 3' ends; a fourth RNA, 2.2 kb long, overlaps in its 5' untranslated region with the 3' untranslated region of the other three transcripts. The two sets of transcripts encode proteins with similar structure which are likely to be transcription factors. Both sets of proteins contain: a serine/threonine-rich amino-terminal region, a glutamine-rich central region, two copies of a cysteine-histidine motif related to but distinct from the zinc fingers present in the TFIIIA transcription factor, and an acid-rich carboxy-terminal region; ln the first set of transcripts, it is the acid-rich that is modified by alternative splicing. By injecting constructs carrying a frameshift mutation either in the first set of transcripts or in the 2.2 kb transcript, we have found that the latter ls responsible for the nl56- resculng activlty. We do not know the function of the other set of RNAs. Sequence analysis shows that the mutation nl56 changes a conserved leucine residue in the first zinc finger of the 2.2 kb RNA to phenylalanine. A Ifn-26/1acZ gene fuslon has been made by insertlng an 8 kb DNA fragment (1 kb of the 2.2 kb RNA coding sequence and 7 kb of upstream sequence) lnto Andy Fire's vector pPD22.04. Thls construct has been injected along wlth the marker rol-6(sulO06J lnto the gonad of N2 hermaphrodltes. In transformed llnes with an extrachromosomal array, stalning with X-gal ls observed ln most hypodermal cells (hyp4 through hypll) and most ectoblast cells (the P cells, B, F, Y, U). starting in embryogenesis at the time these cells are bom. In addition, we have begun a preliminary characterization of nl56/mnDf88 animals ( mnDf88 is a deflclency which takes out the Ifn-26 locus). Such anlmals dle as Ll larvae with hypodermal defects: the tail is grossly deformed, the cuticle is detached. alae are usually absent, some anlmals have no anus, others dlsplay protruslons (as lf the body was not encompassed by the hypodermis). The Ifn-26/1acZ staining pattern, the presence of transcription factors motifs in lin-26 proteins and the terminal phenotype of nl56/mnDf88 animals together suggest that Ifn-26 could be a regulator of hypodermal difrerentiation.

L'Hernault SW et al. (1991) International C. elegans Meeting "SPE-17, UNC-22 AND THEIR NEIGHBORS."

Benian GM, Tobin K, Ward S, L'Hernault SW

[International C. elegans Meeting, 1991]

The null phenotype of unc-22 is strong twitching of body wall muscle and disorganization of muscle structure. Two exceptional unc-22 mutations that cause twitching and sterility due to defective sperm have been analyzed. Both of these mutations cause similar spermatogenesis-defective (spe) phenotypes: abberent localization of organelles in sperm and retention of ribosomes by the mature, normally ribosome-free, gamete (Shakes, D. and S. Ward 1989 Dev. Biol 134:307). Both of these mutations are deletions that affect several transcription units. We have sequenced genomic DNA from the entire region deleted by the smaller of these two deficiencies (ebDf1) and localized a number of RNA's with probes prepared from this sequenced region. We prepared male vs female Northern for these studies [male= RNA from Mog fem-3(q23gf); female= RNA from fem-1(hc171f)j. ebDf1 ( from D. Baillie) deletes at least 55 kb of chromosome IV, including all of unc-22, the gene for one female specific 1.4 kb RNA and the genes for two male specific RNA's of 1.2 kb and 0.58 kb. hcDf1 (from P. Anderson), which is 15 kb, deletes the 5' end of unc-22 and the gene that encodes the 0.58 kb male specific RNA. The common defect that apparently accounts for the defective sperm in ebDf1 and hcDfl homozygotes is deletion of the gene for the 0.58 kb mRNA, which we have named spe-17. Strains carrying two copies of either deletion are fully wild type when transformed with a small free duplication ( created from the cosmid C1 3G4 and the phage DM23 by A. Fire) that covers this region. We have sequenced one spe-17 cDNA and one cDNA for the female specific RNA; neither gene shows significant homology to anything in either Genbank or PIR. Our cDNA screens also recovered one unc-22 cDNA that included the 5' end of this gene, which encodes a 21. 6 kb transcript. We are now microinjecting worms with small genomic subclones in order to create transgenic deletion homozygotes that are fertile but still unc-22.; this should confirm that the Spe phenotype discussed above is caused by disruption of only one gene.

Gunther CV et al. (1998) Midwest Worm Meeting "Does an alternatively-spliced daf-4 transcript encode a dominant inhibitor of the DAF-7 signaling pathway?"

Gunther CV, Riddle DL, Estevez M

[Midwest Worm Meeting, 1998]

When C. elegans develops in a favorable environment, a signaling pathway activated by DAF-7 (a member of the TGF-b family of peptide growth factors) prevents the formation of developmentally-arrested dauer larvae. Genetic and molecular evidence suggests that the putative receptors for DAF-7 are DAF-1 and DAF-4, transmembrane receptor kinases most similar in sequence to Type I and Type II TGF-b receptors, respectively. Northern blots indicate that two transcripts are produced from the daf-4 gene: A 2.9 kb message containing eleven exons that encode the full-length receptor kinase, and a 2.0 kb message that lacks exons encoding the kinase domain and carboxy terminus. In mixed-stage worms grown in plentiful food, these transcripts are equally abundant. However, in dauer larvae steady-state levels of the 2.0 kb transcript exceed levels of the 2.9 kb transcript approximately five-fold. This difference suggests that alternative splicing of daf-4 might play a role in regulating the dauer state. Translation of the 2.0 kb transcript would yield a dominant inhibitory form of DAF-4, capable of binding ligand but inactive in kinase-dependent intracellular signaling. Structure:function analyses of daf-4 mutants provides genetic evidence supporting a daf-4 inhibitory function. Three alleles carry mutations in exons that encode the kinase domain, exons which are absent in the 2.0 kb transcript. An additional mutation deletes sequences that encode the extracellular domain and leads to a shift in the reading frame of the 2.9 kb and the 2.0 kb messages. Interestingly, the latter mutant formed approximately half as many dauer larvae at 15!C and 20!C as strains carrying mutations in the kinase domain. In the absence of wild-type DAF-4, we propose that this difference in phenotype is due to the 2.0 kb translation product inhibiting a DAF-4-independent DAF-1 signaling activity. We are currently investigating the functional role of the 2.0 kb transcript in greater detail by isolating and sequencing the 3' end of the transcript from RT-PCR products. Antibodies to the DAF-4 extracellular domain will be used to determine whether the transcript encodes a stable protein. We also are overexpressing a truncated daf-4 gene to look for enhancement of dauer formation. Our findings suggest that there are at least three ways to fine-tune DAF-7 signaling activity: environmental control of daf-7 transcription, weak DAF-1 growth-promoting activity in the absence of DAF-4, and regulated splicing of daf-4 mRNA to generate a transcript that encodes a putative inhibitory protein. Taken together, these mechanisms afford C. elegans a remarkably sensitive system for assessing environmental suitability during larval development. This work was supported by NIH grants F32 GM 17491 to C.V.G. and HD 11239 to D.L.R.

Li WQ et al. (1991) International C. elegans Meeting "MOLECULAR CHARACTERIZATION OF THE unc-33 GENE."

Li WQ, Shaw JE

[International C. elegans Meeting, 1991]

Mutation in unc-33 results in severely uncoordinated movement, presumably because of neuronal defects, which include aberrant axonal outgrowth and superabundance of microtubules (Hedgecock etal. 1985). Two spontaneous unc-33 mutations (mn260 and rhl O30) obtained from a TR679 background were shown to cosegregate with a specific Tc4 insertion which was absent in revertants. DNA adjacent to the Tc4 insertion was cloned by transposon tagging and revealed that an EMS- induced allele and a garnma ray-induced deletion allele contained RFLP's in the same region. Unique worm sequence flanking the Tc4 insertions was used as a probe to screen an N2 genomic library (from C. Link). One of the genomic clones was located on the physical map by A. Coulson and J. Sulston. The 3' part of the gene was not represented by any of the cosmids known in the region, however. Therefore, an overlapping YAC clone (from R. Waterston) was used to make a mini library, and one lambda clone covering the 3' portion of unc-33 was identified. Several overlapping cDNA clones were recovered from two cDNA libraries (from J. Ahringer and S. Kim), and a 3.7 kb sequence was found that corresponds tO a 10 kb unc-33 genomic region. The putative amino acid sequence from the 2.6 kb ORF within the 3.7 kb cDNA did not show significant similarity to any other protein sequences in the databank. Northern analysis of polyA+ RNA from N2 worms revealed three transcripts: 2.6 kb, 3.3 kb and 3.8 kb in size, with the middle band much fainter than the other two. It seems that the transcripts differ only at their 5' ends. Stage specific Northem analysis indicated that the three bands were present in all developmental stages, and their abundance did not seem to change throughout development. A 2.0 kb coding fragrnent from one of the cDNA clones was fused with the trpE expression vector, pATH10. A 110 kD trpE-unc33 fusion protein was partially purifed by preparative SDS-PAGE and injected into rabbits. The postimmune serum was used in Western blot analysis with total worm protein. Three protein bands, corresponding approximately to the sizes expected from the translation of the three transcripts, were found in wild-type worms at different stages of development. Different protein band patterns were detected in various unc-33 mutants.

Perry MD et al. (1991) International C. elegans Meeting "DECIPHERING HER-1'S SEXUAL MESSAGES"

Robertson B, Wood WB, Perry MD, Li WQ, van Auken KM, Trent C, Hageman JM, Ippensen DC

[International C. elegans Meeting, 1991]

The her-l gene, which is required for normal male sex determination, has been cloned using Tcl transposon tagging and shown to have two XO- specific transcripts (1). We are analyzing its functions by determining the structure of the gene and its products. We have sequenced approximately 8 kb of genomic DNA including the her-l gene as well as several cDNAs corresponding to the two transcripts. The first open reading frame in the rare 1.2 kb transcript is predicted to encode a 20 kD polypeptide whose hydrophobic amino terminus could be a secretion signal sequence. The relatively abundant 0.8 kb transcript is predicted to encode the Cterminal (approximately) half of the same polypeptide, with no signal sequence. RNAse protection mapping experiments of the region confirm our earlier evidence that the 1.2 kb and 0.8 kb transcripts include 4 and 2 exons, respectively. DNA- mediated transformation to rescue her-l (lf) mutations has been recalcitrant, and we are now testing smaller and smaller cosmid fragments in hope of removing noxious sequences. To define cisacting sequences in the her-l locus, we have fused different lengths of genomic DNA upstream of her-l to Andy Fire's lacZ constructs and injected them with a rol-6 plasmid. We see poor staining in the resulting Fl male rollers, suggesting that the 2.5 kb of DNA exarnined so far may be insufficient for expression. In an oblique attempt to identify essential elements of the her-l locus, we have used coding region probes in low stringency hybridizations to genomic DNAs and libraries from the male/female species C remanei and WS9-6 and have detected highly conserved fragments. Lack of definitive signals from parallel experiments on the hermaphroditic species C. briggsae suggests that even lower stringencies may be necessary. We are also continuing efforts to obtain anti-her-l antisera using bacterial fusion proteins.

Miller LM et al. (1991) International C. elegans Meeting "ALTERNATIVE SPLICING AND DEVELOPMENTAL EXPRESSION OF XOL-1 TRANSCRIPTS."

Meyer BJ, Miller LM

[International C. elegans Meeting, 1991]

The X-linked gene, xol-l (XO lethal) is required for proper X chromosome expression and sexual development in XO males. Loss- offunction mutations in xol-l shift XO animals toward their XX modes of sex detemlination and dosage compensation, but have no effect on otherwise wild-type hermaphrodites. xol-l is the earliest known gene in the hierarchy controlling the male/hermaphrodite decision and is perhaps the gene nearest to the primary sex-determining signal itself. The current model proposes that xol-l promotes male development by ensuring that genes (or gene products) directing hemmaphrodite sex determination and dosage compensation are inactive in XO animals. In certain genetic backgrounds, xol-l mutations can have an effect on the sexual phenotype of XX animals that is opposite to their effect on the sexual phenotype of XO animals; they can further masculinize XX animals already partially masculinized by a mutation in a hemlaphrodite sex determination gene. RFLP mapping and germline transfommation experiments led to the identification of a molecular clone of the xol-l gene. These experiments revealed that the spontaneous xol-l(y9) allele is a 25-50 kb deletion that spans the entire xol-l locus. Using a 7 kb rescuing fragment as a probe in Northern experiments, 3 transcripts (2.5 kb, 2. 2 kb and 1.5 kb) were identified in the region, one of which (2.2 kb) is greatly enriched in XO versus XX embryos. The other two, although more abundant in XO than in XX embryos, are less enriched. All three transcripts are expressed only in embryos. A cDNA library of 3.5 million independent clones was constructed from him-8(el489) embryonic polyA+ RNA. 600,000 unamplified clones were probed with the 7 kb xol-l rescuing clone and 4 xol-l cDNA clones were identified. Comparison of cDNA and genomic sequences revealed that the 4 cDNA clones represent 3 classes of transcripts, all of which undergo the same five splices at their S' ends, but differ at their 3' ends due to three alternative splicing fates of a single 5' donor sequence. The alternative splice that creates the XO-enriched transcript (2.2 kb) predicts a translation product with an extremely acidic region at its C-terminus. None of the xol-l proteins show any significant sequence similarities to any protein in the current data bases. *current address: Department of Developmental Biology, Stanford University, Stanford, CA 94305 (Stuart Kim's lab).

Cathy V Gunther et al. (2001) International C. elegans Meeting "DOES DAF-4 ALTERNATIVE POLYADENYLATION GENERATE A RECEPTOR ANTAGONIST?"

Cathy V Gunther, Donald L Riddle

[International C. elegans Meeting, 2001]

DAF-4 is the only type II TGF-beta/BMP receptor encoded in the C. elegans genome. daf-4 mutants are Daf, Sma and Mab, indicating a role for this receptor kinase in continuous (non-dauer) development, normal body size and male tail development. There are two daf-4 transcripts. During continuous growth, a 2.0 kb message is between 1.2 and 1.9-fold more abundant than the full-length 2.9 kb mRNA. In dauer larvae this ratio increases to 5-fold. However, when dauer larvae recover in fresh medium and resume growth, the relative abundance of 2.0 kb/2.9 kb mRNA declines to that of well-fed populations. This correlation suggests that alternative processing of daf-4 may be environmentally regulated. The 2.9 kb mRNA contains an SL1-spliced 11-exon open reading frame encoding the 744-amino acid DAF-4 receptor kinase. Using RT-PCR, we cloned and sequenced a cDNA of the 2.0 kb transcript, which includes an SL1 leader, all of exons 1-5, and intron 5 sequence upstream of an alternative poly(A) addition site. The open reading frame in the 2.0 kb cDNA encodes a putative 206-amino acid polypeptide that lacks transmembrane and kinase domains, but maintains 90% of the extracellular domain. To measure the effect of overexpressing the short daf-4 transcript on dauer formation, a daf-4 genomic fragment containing upstream sequence through intron 5, was subcloned and microinjected with the rol-6 marker into a weak daf-4 mutant. In three transgenic lines, constitutive dauer formation was enhanced in Rol progeny compared to non-Rol siblings. In one line at 22.5degC, 70% of the Rol progeny formed dauer larvae, whereas only 2% of the non-Rol siblings entered diapause. These results indicate that overexpression of the short daf-4 transgene increases dauer formation. We are currently testing whether this may be due to antagonism of DAF-4 signaling by the truncated receptor. For example, the secreted extracellular domain may bind ligand in competition with full-length receptor. Its abundance in the dauer stage may prevent dauer recovery in response to transient ligand expression.