Waddle JA et al. (1990) Worm Breeder's Gazette "Candidate CapZ genes from C."

Waterston RH, Waddle JA, Cooper JA

[Worm Breeder's Gazette, 1990]

CapZ is one of a family of actin binding proteins that may regulate actin dynamics in eukaryotic cells. This heterodimeric protein is located at the Z-line of vertebrate skeletal muscle in situ (1). Since the ends of actin filaments are also located at the Z-line, CapZ could function in the assembly or tethering of thin filaments at this location. The Z-lines of vertebrate skeletal muscle also contain the proteins alpha-actinin and vinculin. These two proteins have been localized to the dense bodies of nematode body wall muscles (2,3). Collectively, the above observations suggest that a CapZ homolog might also be present at the dense bodies or attachment plaques of nematode muscle. On this premise, we have initiated a search for CapZ homologs in C. elegans.Sequence data from cDNA clones encoding the CapZ subunits of chicken, (4,5), D. discoideum (6) and S. cerevisiae (7,8) were used to design degenerate oligonucleotide primers for the polymerase chain reaction. Direct amplification of nematode CDNA libraries (kindly provided by R. Barstead) generated candidate fragments for both subunits. A comparison between the deduced amino acid sequence from the nematode PCR products and the homologous regions in other organisms yielded the following table of

Sarkis GJ et al. (1986) Worm Breeder's Gazette "cathepsins Ce1, Ce2, and Ce3 in C elegans."

Sarkis GJ, Jacobson LA, Ashcom JD

[Worm Breeder's Gazette, 1986]

Two acid thiol proteases previously designated as cathepsins L 1 and L2 [Kurpiewski et al., Gazette Vol. 8, No. 3, 1984] have now been renamed cathepsins Ce1 and Ce2, since further studies have shown that they are not precisely analogous to the mammalian cathepsin L. Both proteases prefer Z-phe-arg-aminomethylcoumarin (AMC) to Z-arg-arg-AMC, and in this respect resemble mammalian cathepsin L rather than cathepsins B or H. On the other hand, both Ce1 and Ce2 show carboxydipeptidase activity toward the model substrate leu-trp-met-arg- phe-ala, and in this respect resemble cathepsin B rather than cathepsin L. Like the mammalian thiol proteases, both Ce1 and Ce2 are inhibited by EP-64, leupeptin and antipain. Both are insensitive to the trypsin inhibitors from soybean and pancreas, to alpha- 1 - antitrypsin and to pepstatin. Three lines of evidence indicate that Ce1 and Ce2 are distinct gene products. First, Ce1 has much more endoprotease activity (using FITC-casein as substrate) than Ce2, relative to the activity of each against Z-phe-arg-AMC. Second, isoelectric focusing in nondenaturing slab gels shows that Ce1 has a pI of 6.8, whereas Ce2 has a pI of 4.7. This seems too large a difference to attribute to variant post-translational processing. Finally, the age-dependences of Ce1 and Ce2 are different (Sarkis et al., this issue). We are still working on the development of differential screens for mutants deficient in each of these enzymes. A third protease has also been separated during purification of Ce1 and Ce2 on DEAE-cellulose. We call this enzyme cathepsin Ce3. It has strong endoproteolytic activity toward FITC-casein at pH 5.5, but no activity toward Z-phe-arg-AMC, Z-arg-arg-AMC, arg-AMC, N-succinyl-(ala) 3-AMC, glutaryl-phe-AMC, or any other model substrate we have tried. It does not hydrolyze the model hexapeptide leu-trp-met-arg-phe-ala. Unlike Ce1 and Ce2, it does not require a thiol for activity, but is inhibited by millimolar concentrations of DTT. It is insensitive to leupeptin, EP-64, pepstatin, antipain, STI and PTI. Its pI is about 7. 4. The major protein band in our Ce3 preparations reacts covalently with alpha-2-macroglobulin, a general protease-trapping inhibitor. This suggests that the major band is cathepsin Ce3. We have looked at crude extracts -- which have been depleted of cathepsin D by affinity chromatography -- by IEF and solid-phase casein zymography. The predominant activities are in the positions of cathepsins Ce 1 and Ce3, with a minor band at the position of cathepsin Ce2. Thus, most of the endoprotease activity at acid pH is attributable to cathepsins D, Ce1 and Ce3. For those of you whose interest in proteases is limited to protecting your favorite proteins or enzymes during isolation or assay, we recommend a mixture of pepstatin (an inhibitor of cathepsin D) and EP-64. The former is a queer pentapeptide; the latter is an active- site directed inhibitor based upon an active epoxide group. Neither is likely to interfere with your pet enzyme or protein. Both are available from Sigma. We suggest concentrations of about 10 M.

Roubin R et al. (1996) Worm Breeder's Gazette "AN FGF GENE IN THE WORM?"

Thierry-Mieg D, Birnbaum D, Roubin R

[Worm Breeder's Gazette, 1996]

A putative gene coding for a heparin-binding growth factor protein and presenting 30 % similarity with the Fibroblast Growth Factor 5 (FGF5) has been identified on chromosome III of C.elegans (Wilson, R et al, Nature, 1994, 368, 32-38, and Du, Z, unpublished data). The genomic clone CO5D11 carries a complete copy of the gene but no cDNA has yet been identified. In mammals, FGFs are important for embryonic development and have been implicated in some pathologies. Recently, Stern's group has shown that egl-15 encodes a member of the FGF receptor family in C.elegans (DeVore, D. l. et al, Cell, 1995, 83, 611-620). In both C.elegans and D.melanogaster, FGF receptors are involved in cell migration.

Holdeman R et al. (1995) Worm Breeder's Gazette "MES-2 Contains a Highly Conserved Motif Characteristic of Proteins Involved in Regulating Gene Expression via Chromatin Structure"

Srome S, Nehrt S, Holdeman R

[Worm Breeder's Gazette, 1995]

We have been studying maternal-effect sterile (mes) mutants in an effort to understand how development of the germ line of C. elegans is maternally controlled. Four genes, mes-2, mes-3, mes-4, and mes-6, all give a similar mutant phenotype. Specifically, progeny of mes/mes mothers undergo apparently normal embryogenesis and develop into heathy appearing adults but fail to produce a functional germ line. Mes-2 mutant larvae can produce as many as 80 germ nuclei by the L4 larval stage. There is then an obvious and extensive die-off of germ nuclei, which results in adult worms having only ~ nine germ nuclei and no mature gametes (Capowski et al., 1991; Paulsen et al., 1995). mes-3 and mes-6 have previously been cloned (mes-3: Paulsen et al., 1995); they encode proteins that lack significant similarity to any described proteins. Transformation rescue of Mes-2 mutants was accomplished by injecting overlapping genomic fragments of 7 and 13 kb in length from cosmid R06A4 on LGIIR. These fragments, which both recognize a 2.5 kb Northern transcript, were used to screen a cDNA library. A candidate cDNA clone of 2.5 kb, which also recognizes the 2.5 kb transcript, was identified in the screen. Injection of both sense and antisense RNA produced from this cDNA phenocopied the Mes-2 mutant phenotype in wild-type worms. Worms injected with control RNA did not show this phenotype. The mes-2 cDNA encodes a protein that contains a 150 amino acid motif known as the SET domain. This motif gets its name from three Drosophila proteins in which it is found: Trithorax (Trx), Suppressor of variegation 3-9 [Su(var)3-9], and Enhancer of zeste [E(z)] (see WBG for figure). SET domains have also been found in proteins from plants, yeast, and mammals. The E(z) protein, to which MES-2 is most similar, is a member of the Polycomb group (PC-G) genes in Drosophila. These genes are believed to be needed for the long-term maintenance of transcriptional inactivation of specific genes, such as the homeotic genes. PC-G proteins are thought to work by affecting chromatin structure, although they do not seem to bind DNA directly. Current thinking is that these proteins form large complexes with other gene products and control chromatin structure at specific sites (reviewed in Paro, 1995). Genetic evidence suggests that the Mes mutant phenotype may be due to defects in X-chromosome gene expression, specifically in the germ line. The similarity of MES-2 to E(z) and other SET domain-containing proteins, which have a long-term role in regulation of expression, is intriguing in light of our suspicion that mes genes may affect gene expression on the X chromosome. We are in the process of raising antibodies to MES-2 in order to further analyze this possibility. References: E. Capowski, P. Martin, C. Garvin, and S. Strome (1991) Genetics 129, 1061-1072. J. Paulsen, E. Capowski, and S. Strome (1995) Genetics, in press. R. Paro, TIG (1995) 11 (8), 295-297.

Seydoux GC et al. (1993) Worm Breeder's Gazette "A lac-Z Fusion Expressed in the C Lineage"

Fire A, Seydoux GC

[Worm Breeder's Gazette, 1993]

During the course of our screen for embryonic expression patterns (WBG 12#4, p20 ),we have isolated a lac-Z fusion that appears to be expressed specifically in the C lineage. This fusion (pGS 15.02) consists of a 3.2 kb fragment of C. elegans genomic DNA fused to the lacZ gene in the pPD21 .28vector. We found that 1 kb of the 3.2 kb is sufficient to drive C specific lac-Z expression (pGS 15.24) and have used three lines carrying an integrated array of pGS15 .24to characterize its expression pattern. To facilitate the identificadon of the lac-Z expressing cells, we have developed a modified method to stain lineaged embryos (thanks to Lois Edgar and Deborah Cowing for their helpful suggestions). Briefly, 2 to 4 cells embryos are transfered in a drop of modified lacZ staining mix (see below) onto a 1.5% agarose pad (also containing modified staining mix) and covered by a coverslip for Nomarski observadon. Embryos are allowed to develop and their lineage is recorded until a desired stage is reached (see WBG 12#2, p 19 for a descripdon of our recording system). At that time, the egg shell and vitelline membrane are disrupted using a laser microbeam to allow the staining mix to enter the embryo. 12 hours later, cells staining for lacZ can be idendfied based on their lineage idendty. In this manner, we found that in the 28 cell-stage embryo the four C-derived blastomeres (Caa, Cap, Cpa, Cpp) all express the fusion in their cytoplasm. (In nonlineaged embryos, we sometimes saw two lacZ-expressing cells in 24 to 26 cellembryos, suggesting that Cp and Ca may occasionally start expressing the fusion towards the end of their cell cycle). The four C-derived blastomeres, as well as their 8 progeny continue to express the fusion during their endre cell-cycle. By the 16 C-cellstage (172 cells), the expression of the fusion weakens but still remains confined to the C lineage. During morphogenesis, no expression of the fusion has been observed in lines carrying the integrated array. Curiously, lines carrying pGS15 .24on an extrachromosomal array continue to express the fusion up to the comma stage. At this dme, expression of the fusion is seen in the nuclei of C-derived hypodermal cells. (We note, however, that expression of the fusion appears somewhat deregulated in non-integrated lines: in the 28 cell-stage embryo, many more cells besides the C-derived blastomeres often express the fusion.) Sequencing pGS15 .24,we have found that it encodes a translational fusion between a novel C. elegans gene and lacZ. We have obtained and sequenced cDNAs corresponding to this gene and found that it has the potential to encode for a 38 kD protein containing an acidic region rich in glutamic acid. No homologies or other features have been noticed as of yet. By Northern analysis, this gene appears to be expressed predominantly in embryos, with no detectable expression in larvae and a low level of expression in adults. Modified Staining Mix: O.O5M NaPi pH 7.5 2mM MgCl2 5mM KFe/5mM KFo 1 ug/ml DAPI 0.004% SDS 5% Sucrose 0.125% X-gal 1% glutaraldehyde

Kurpiewski MR et al. (1984) Worm Breeder's Gazette "two distinct thiol proteases in C elegans."

Jacobson LA, Kurpiewski MR, Jen-Jacobson L

[Worm Breeder's Gazette, 1984]

We had previously observed that thiol-dependent, leupeptin-sensitive proteases make a significant contribution to proteolysis by crude extracts of C. elegans at pH 4.5-5.0 (approximating the intralysosomal pH). We have now resolved two distinct thiol proteases, which we designate cathepsin L1 and cathepsin L2, by chromatography on DEAE-Sephadex. The properties of cathepsin L2 from C. elegans correspond quite closely to those of vertebrate cathepsin L with respect to substrate specificity, pH optimum and molecular weight. Both proteases are optimally active toward Z-phe-arg-7-amino-4-methyl coumarin amide (MCA) at pH 5. Both are strictly thiol-dependent and leupeptin-sensitive, and neither is inhibited by pepstatin, bestatin or PMSF. Neither shows any activity toward Z-arg-arg-MCA, a substrate for vertebrate lysosomal cathepsin B, nor towards succinyl-(ala)(ala)( ala)-MCA, a substrate for the elastase-like protease of C. elegans. Cathepsins L1 and L2 are clearly distinct enzymes, since L1 is inhibited by soybean trypsin inhibitor (STI), whereas L2 is not. L2 digests denatured FITC-casein, whereas L1 does not. Cathepsin L2 preparations show a single major band on SDS gels with M=25,000, as detected after labeling the enzyme with [125I]. L2 acitivity emerges from a gel filtration column at a position corresponding to 24 kDa, so it appears likely that cathepsin L2 is active as a monomer. We are presently developing screening procedures for isolating mutants deficient in cathepsin L2. to contain several additional MSP genes. Thus in a 50Kb region there are at least five MSP genes. Two of them correspond to one of our cDNA sequences and two others correspond to one of Michael Klass's cDNA sequences so at least four of these genes must be expressed. If we can learn where these genes are in the genome it might be possible to eliminate them with a small deletion and so discover what happens to sperm with insufficient MSP. We are currently sequencing the genomic MSP genes to see if there are conserved regions outside the coding sequence. These might be either regulatory sites or the remnants of residual translocatable elements which could have participated in dispersing these genes throughout the genome.

Gilleard JS et al. (1996) Worm Breeder's Gazette "Regulation of cell and stage specific expression of the cuticle collagen gene dpy-7."

Barry JD, Johnstone IL, Gilleard JS

[Worm Breeder's Gazette, 1996]

We are interested in understanding how the spatial and temporal patterns of the cuticle collagen genes are regulated. To this end we have been studying the regulation of the cuticle collagen gene dpy-7. dpy-7/lac-Z translational reporter gene fusions with as little as 310 bp of 5! flanking sequence produce exactly the same pattern of expression as larger constructs (up to 1265bp). However translational reporter gene fusions containing just 165 bp of 5! flank produce no b-galactosidase expression showing sequences between -310 and -165 are necessary for expression. b-galactosidase expression is first seen in hypodermal cells at the late comma stage and is seen in all postembryonic developmental stages including the adult. The precise pattern of expression has been resolved by an IFA double labelling experiment using anti-lac Z monoclonal antibodies to detect the nuclear localised b-galactosidase and monoclonal antibody MH27 to delineate hypodermal cell membranes. The dpy-7 reporter constructs are expressed in hyp-7 , P cells and most of the hypodermal cells of the head and tail (hyp-5, hyp-6 and hyp-10 stain particularly strongly). However there is no expression in the V (or other) seam cells. Throughout development there is an increasing number of hyp-7 nuclei staining, with expression presumably beginning some time after each nuclei joins the syncitium. Since dpy-7 is trans-spliced with SL1 we have mapped the transcription initiation site by using a 5! RACE strategy designed to be specific for pre-mRNA by using an oligo immediately 5! to the SL1 splice acceptor site in conjunction with the anchor primer in the final round of PCR. Transcription appears to initiate from numerous sites between -210 and -159 relative to ATG (-206 and -155 relative to SL1 splice acceptor). We have also cloned the C.briggsae dpy-7 homolog and can repair the C.elegans dpy-7 mutant phenotype by microinjection of the C.briggsae gene showing that the regulation (and also the function) of dpy-7 is conserved between the two species. Similarly C.elegans reporter gene constructs produce the same pattern of expression when transformed into C.briggsae and sequence between -310 and -165 is also necessary for expression in C.briggsae. Comparison of the 5! flanking sequences of dpy-7 between the two species shows a block of homology (H1) in which 66 out of 72 residues are conserved. This lies within the region necessary for reporter gene expression and immediately upstream of the transcription initiation sites ( - 202 to -273 C.elegans and -185 to -256 C.briggsae). This is the only significantly conserved region of sequence between the dpy-7 ATG and the next predicted upstream gene at -346 (by genefinder analysis). Using RT-PCR we cannot detect any dpy-7 transcripts which are SL2 trans-spliced which, together with our analysis of the dpy-7 functional promoter and transcription initiation sites, show that dpy-7 is not a downstream gene of a polycistronic transcription unit. Transcriptional lac/Z reporter gene fusions with as little as 145bp encompassing the H1 homology and the transcription initiation sites are sufficient to produce the same pattern of expression as the translational fusions although expression in the later stages (post L1) is poor. A number of constructs have been examined and the low level of later stage expression does not appear to be due to the loss of a particular region of sequence but seems to be a general feature of our transcriptional fusion constructs. The transcriptional fusions are all orientation dependant and so the element appears to function as a true tissue specific promoter element which drives expression in all hypodermal cells except the seam cells. A series of fine 5! deletions through the H1 region shows that a single 15bp deletion completely ablates the b-galactosidase expression pattern. This deletion disrupts an AGATAA motif (which is conserved in C.briggsae) suggesting that a member of the GATA transcription factor family may play a role in the regulation of dpy-7 expression.

Shivakumar S et al. (1994) Worm Breeder's Gazette "wnt Homologs in Caenorhabditis elegans"

Kenyon CJ, Shivakumar S, Varmus HE, Jongeward GD

[Worm Breeder's Gazette, 1994]

wnt homologs in Caenorhabditis elegans Supriya Shivakumar, Gregg Jongeward, Cynthia Kenyon and Harold Varmus. University of California, San Francisco, CA 94143 The wnt gene family encodes cysteine-rich secreted proteins which have been shown to act as important regulators of early development in the frog, fly and mouse. wnt genes have a crucial role in intercellular signaling in the mouse CNS development and in pattern formation in the fly. We chose to study the normal role of Ce-wnt-Z in signaling in C. elegans development. Two members of the wnt gene family have been identified in C elegans, Cc-wnt 1 and Ce-wnt-22. The Ce-wnt-1 gene encodes a 372 amino acid protein and shares 22 of the 24 cysteines found among other members of the wnt family. A 1.4 kb transcript (including an SL1 transpliced leader sequence) is detected at high levels in embryos and at lower levels at all other stages. The 1.4 kb Ce-wnt-2 cDNA is also expressed in a similar temporal pattern to Ce-wnt-1. The Cc-wnt-2 cDNA encodes a protein of 362 amino acids with 22 of the cysteines shared among other Wnt proteins. Neither Ce-wnt 1 and Ce-wnt-2 are direct homologs of any specific wnt genes in other organisms. To elucidate the functions of the Ce-wnt-Z gene product, we have isolated mutations in the gene. The Ce- wnt-l locus maps to the left arm of chromosome II, but does not appear to correspond to any existing mutations. We made the assumption that the Ce-wnt-1 gene is required for viability based on wnt mutant phenotypes in other species. We screened for psoralen induced lethal mutations balanced by an extrachromosomal array comprising a 15 kb lambda clone containing the Ce-wnt-1 gene and a marker (rol-6D) in a F2 clonal screen (see diagram below). 1500 F2 animals were screened and 2 embryonic lethal alleles were isolated. Both contain deletions in the Ce wnt-1 coding region. These mutations are rescued by the lambda done, but not by rol-6. A 53 kb subdone also rescues the lethality. We are now trying to demonstrate conclusively that this early embryonic lethality is due to the disruption of the Ce-wnt-1 gene. We have also screened for psoralen induced mutations (in an identical screen) that require a cosmid clone containing Ce- wnt-2. We have identified a number of mutations in the screen which we are nDw analyzing in more detail. 1Shackleford et al., Oncogene 8:1857 (1993) 2Waterson et al., Nature genetics 1:114 (1992)

Sokol SB et al. (1996) Worm Breeder's Gazette "Characterization of a C. elegans homologue of the S. cerevisiae gene STE20"

Sokol SB, Way JC

[Worm Breeder's Gazette, 1996]

We have isolated by PCR a 400-bp fragment of the kinase domain of an apparent C. elegans homologue of the yeast gene STE20, a kinase in the mating pheromone response pathway acting immediately downstream of the G-protein (upstream of the kinase cascade of STE7, STE11 and FUS3/KSS1). This fragment displays an even higher degree of similarity with the mammalian brain kinase p65PAK, which appears to be activated by Cdc42 (Manser et al., 1994). (See figure below.) We are trying to isolate an animal deleted for this gene as well as a full-length cDNA clone. The region containing this gene, which maps to the X chromosome, has recently been sequenced as part of the Genome Project. This gene spans an apparent cosmid gap, but it actually is located over two cosmids that do overlap by 4 kb. The presence of this "gap" plus the difficulty we have had in producing a full-length cDNA clone and in growing the cosmids cause us to think that this DNA sequence may be difficult to grow in E. coli. We would thus be grateful for any assistance with obtaining E. coli strains that will accept these difficult-to-grow sequences. Reference: Manser, E., T. Leung, H. Salihuddin, Z.-s. Zhao and L. Lim. "A brain serine/threonine protein kinase activated by Cdc42 and Rac1." Nature 367: 40-46 (1994).

Miller DM et al. (1989) Worm Breeder's Gazette "unc-4: 2 kb from ennui."

Dubois ML, Ghee M, Ruvkun GB, Burglin TR, Miller DM

[Worm Breeder's Gazette, 1989]

unc-4 appears to define the wiring diagram for a specific portion of the ventral nerve cord. In unc-4(e120), most VAn motor neurons assume a pattern of presynaptic connections normally reserved for the VBn motor neuron class; VAn's receive input from AVB and PVC interneurons as opposed to the normal sets of connections with AVA, AVD and AVE which are absent in e120 (White et al., WBG 9, p81, 1985). We have now shown that unc-4 may correspond to yet another homeobox locus in C. elegans.A spontaneous allele, unc-4(wd1), was isolated from the mutator strain TR679 in a general screen for movement-defective mutants. Blots of recombinants with flanking markers rol-6 and bli-1 identified two closely linked TC1-containing EcoRI fragments; wdP1 (2. 5 kb) is 0.12 m.u to the right and wdP3 (3.2 kb) is 0.1 m.u to the left. A phage clone overlapping wdP1 was fingerprinted in Cambridge and maps to a large contig near unc-4 on chromosome II. wdP3 includes repetitive DNA and a corresponding phage clone to mark the left side boundary is apparently not represented in our genomic library. The identification of a homeobox sequence in the region, however, suggested a likely location for unc-4. Cosmid C07E2 overlaps, ceh-4, one of 50 different homeobox loci that hybridize with a degenerate oligonuclcotidc probe (T. Burglun and G. Ruvkun, WBG 10, p35, 1988). We probed Southern blots of DNA isolated from 15 independent unc-4 alleles with random primer labeled C07E2. Alterations were detected in five of these alleles: e2314, e2320, e2308 (from M. Shen), e887 and wd-1. All of these mutations affect a 1.8 kb EcoRI fragment (see below). The e2320 mutation deletes 200 bp from the 1.8 kb fragment. In wd-1, the 1.8 kb as well as several other EcoRI fragments are absent suggesting a large deletion (~25kb). The alterations observed in e2314, e2308, and e887 have not been reconstructed but all of these mutations delete the 1.8 kb EcoRI fragment. A restriction map of the region indicates that a 1.05 kb EcoRI fragment containing the homeobox 'third helix' consensus sequence (.WFQNRR.K.R) is adjacent to the region that is altered in unc-4 mutants. We are now searching for a cDNA to determine if in fact the putative homeodomain corresponds to portion of the unc-4 coding sequence. [See Figure 1]