Yochem JJ et al. (1988) Worm Breeder's Gazette "a lin-12 homolog."

Greenwald IS, Yochem JJ

[Worm Breeder's Gazette, 1988]

We have isolated clones from a genomic library of N2 that cross- hybridize at reduced stringency with probes from the lin-12 gene. We sent one of the genomic clones, which had been partially sequenced to confirm that it did indeed contain a lin-12 related gene, to Alan Coulson and John Sulston, who located the sequence on the physical map just to the right of lin-12 by fingerprinting. The combination of map position and sequence data made us very suspicious that the cloned sequence might correspond to glp-1, a gene that, like lin-12, has been implicated in cell-cell interactions (Austin and Kimble, 1987; Priess et al., 1987). We sent the clone to Judith Austin and Judith Kimble, who had at about the same time by Southern blot analysis found glp-1 allele alterations associated with cosmids in the region, and they then provided convincing evidence that our new sequence does correspond to the glp-1 gene (see their article in this Newsletter). We also sent the clone to Andy Fire, who had been microinjecting cosmid DNA from the region in collaboration with Jim Priess (see his article in this Newsletter). DNA sequencing of both genomic and cDNA clones reveals a predicted glp-1 product that is 50-60% identical with lin-12 over extensive regions. Like lin-12, the new sequence contains both EGF-like and nonEGF-like cysteine repeats in the predicted extracellular part of the protein as well as another repeated motif (also found in cdc10, SWI6 and Notch ) in the presumed cytoplasmic portion . For example, we compare in the Figure part of the sequence encoded within a restriction fragment identified by Austin and Kimble as containing glp- 1 allele alterations with part of the ORF A exon of lin-12 (an exon containing EGF-like repeats). These and more extensive sequence data not shown here suggest that lin-12 and glp-1 evolved from a common ancestor by gene duplication. Perhaps other developmentally important genes are members of this family. [See Figure 1]

Yochem JJ et al. (1988) Worm Breeder's Gazette "lin-12 and glp-1 encode very similar potential transmembrane proteins."

Greenwald IS, Yochem JJ

[Worm Breeder's Gazette, 1988]

The lin-12 and glp-1 genes have each been implicated in certain cell- cell interactions during C. elegans development (Greenwald et al., 1983; Austin and Kimble, 1987; Priess et al., 1987). Evidence was presented in the previous issue of the Gazette that these genes are related at the sequence level (Yochem and Greenwald, p. 84; Austin and Kimble, p. 85; Fire and Priess, p. 86). The complete DNA sequencing of genomic and cDNA clones of glp-1 has now been finished, permitting an overall comparison with the complete sequence of lin-12 ( Yochem et al., 1988). Iin-12 and glp-1 appear to have evolved from a common ancestor, since several splice junctions are conserved and their products are strikingly similar (the overall identity is about 50%), although the glp-1 product contains only ten copies of the EGF- like motif found 13 times in lin-12, and is therefore smaller (1295 versus 1429 amino acids). The predicted primary products are aligned below in schernatic form relative to a strongly hydrophobic stretch of amino acids which could serve as a membrane spanning domain. Both products contain a somewhat similar sequence ('T + Y-encoded sequence') that does not completely resemble EGF after their first copy of the EGF-like motif, three copies of a cysteine-rich sequence ('LNR') that does not resemble EGF, and six copies of a motif (Breeden and Nasmyth, 1987) in their cytoplasmic regions that is also present in two yeast gene products (Schizosaccharomyces pombe cdc10, Saccharomyces cerevisiae SWI6). The lin-12 and glp-1 products also have an intriguing overall resemblance to the product predicted for the Drosophila Notch gene (Wharton et al., 1985; Kidd et al., 1986 ), which has also been implicated in cell-cell interaction during development. The lin-12 and glp-l gene products are much more closely related to each other than to the Notch product, and the extent of divergence suggests that Notch is not a counterpart of either. A more reasonable consideration is that these genes define a new family with individual members possessing related but distinct functions during development. [See Figure 1]

Buechner M et al. (1997) Worm Breeder's Gazette "exc Gene Interactions"

Buechner M, Hedgecock EM

[Worm Breeder's Gazette, 1997]

We have continued genetic studies on exc mutants, in which the long hollow processes (canals) of the excretory cell form large fluid-filled cysts. We previously classified these mutants into five groups based on cyst morphology: exc-2, exc-4, let-4, and let-653 mutants form large septate cysts near the excretory cell body; exc-1 and exc-5 mutants form extremely large septate cysts (often visible via dissecting microscope) predominantly at the shortened canal termini; exc-3, exc-6, and exc-8 mutants show regional canal widenings and meanderings, and occasional inappropriate extra branchings to form multiple canal lumena; alleles of sma-1 exhibit a short, very wide, unseptate canal. Finally, the lone exc-7 allele rh252 exhibits small cysts along the length of the canal, terminating in a large group of tiny cysts. We have now made double mutants of several of the genes from various groups. Double mutants made of genes from within any one group showed no increase in severity in canal phenotype. For example, exc-2 (rh90) exc-4 (rh133) animals showed no changes in cyst size or morphology compared to rh90 or rh133 alone, and no increase in larval lethality. Similarly, exc-6 (rh103) exc-8 (rh210) canals appeared no shorter or more meandering than those of the single mutants, and exc-1 (rh26) exc-5 (rh232) mutant canals had similar cyst size and placement as the single mutants alone. Doubles constructed with exc-7 (rh252) showed interesting effects, however. exc-7 is epistatic to exc-6; the double mutants showed no abnormal branchings, only a series of small cysts along the entire canal length. Perhaps not surprisingly, the large cysts of exc-2 and exc-4 are epistatic to exc-7. Most strikingly, the doubles exc-7 (rh252) sma-1 (e30) and especially exc-7 (rh252) exc-3 (rh207) show canal phenotypes different from that of the single mutants; instead, these animals show large cysts near the excretory cell body, like those of exc-4 mutants. Finally, we have been unable so far to isolate doubles of exc-7 (rh252) and either exc-1 (rh26) or exc-5 (rh232); this could indicate that these mutants are lethal in combination. Our previous ultrastructural data, along with sequencing results from the labs of David Baillie (Jones & Baillie, Mol. Gen. Genet. 248, 719-726, !95) and Judith Austin (see McKeown, Patel, Praitis, and Austin, WBG 14 (2), p. 83) suggest that the exc genes encode proteins necessary for the proper placement of structural elements, both extracellular and cytoskeletal, at the apical epithelial surface. We believe that these new genetic results indicate that the differing morphologies of the various groupings of exc genes reflect different elements that interact with each other to shape this surface; Exc-7 may play a central position in these interactions.

Austin JA et al. (1988) Worm Breeder's Gazette "a preliminary molecular analysis of glp-1."

Austin JA, Kimble JE

[Worm Breeder's Gazette, 1988]

Genetic studies have shown that glp-1 is necessary for at least two regulatory cell-cell interactions during C. elegans development ( Austin and Kimble, 1987; Priess, Schnabel and Schnabel, 1987). As a first step towards understanding the molecular mechanism by which glp- 1 affects these interactions, we have cloned the gene. Our strategy relied on the fact that glp-1 maps 0.02% to the right of lin-12. Since the ratio of physical and recombinational distance in this region is about 830 kb/map unit (Greenwald et al. 1987), we estimated that the glp-1-lin-12 intergenic distance would be approximately 17kb. Therefore, we screened DNAs from 8 EMS and 2 gamma-ray induced glp-1 alleles with two hybridization probes: T26E12, a cosmid that carries lin-12, and ZK506, a cosmid that overlaps and is to the right of T26E12 (both cosmids kindly provided by J. Priess). None of the glp- 1 alleles showed altered restriction fragment patterns when probed with T26E12 but 1/8 EMS and 2/2 gamma-ray induced alleles showed alterations in the same 1.5 kb R1 fragment when probed with ZK506. q172 has a 300bp deletion within the 1.5 kb R1 fragment, qDf2 has a large deletion with one endpoint in this fragment, and q339 has a DNA rearrangment where this fragment is replaced by novel fragments of 1.1 and 1.9 kb. At the same time, John Yochem, in Iva Greenwald's lab, had identified a sequence with strong similarity to the lin-12 sequence (J. Yochem and I. Greenwald, personal communication; see also Yochem and Greenwald, this issue) that was then localized to ZK506 (J. Sulston and A. Coulson, personal communication). Using a phage containing this lin-12 homologue as a probe for Southern blots of q172 and qDf2, we were able to determine that the DNA alterations associated with these two mutations lay within the region of similarity to lin-12. In particular, the 1.5 kb R1 fragment is in a region that corresponds to the extracellular domain of lin-12 (J. Yochem and I. Greenwald, personal communication). On northerns, glp-1 probes hybridize to a single abundant transcript of 4.5 kb in the RNA of adult hermaphrodites. This is the same size as the major lin-12 transcript (J. Yochem and I. Greenwald, personal communication). Which raises the issue of whether crosshybridization is occurring. We believe that we presently have gene specific hybridization but plan to test this using glp-1 and lin-12 alleles that produce transcripts of altered size. Comparison of the level of glp-1 transcript in wildtype hermaphrodites and hermaphrodites where the germ line is absent suggests that the glp-1 message is located primarily in the germ line. The presence of glp-1 message in the germ line is consistent with earlier results that indicated that glp-1 acts in the germ line to control the switch of germ cells from mitosis to meiosis (Austin and Kimble, 1987).

Austin JA et al. (1988) Worm Breeder's Gazette "a new balancer for chromosome III."

Austin JA, Kimble JE

[Worm Breeder's Gazette, 1988]

We have isolated an allele of glp-1 that suppresses recombination on the third chromosome between at least dpy-1 and dpy-18 9) was generated by gamma-ray mutagenesis and is linked to e1259. In the strain dpy-19(e1259)glp-1(q339)/unc-32(e189) dpy-18(e164) we observed that the expected unc-32 and dpy-18 recombinants were absent. To investigate this phenomenom further we looked for Dpy and Unc recombinants in strains heterozygous for e1259q339 and a series of Dpy on Chromosome III: dpy-1(e1)unc-93(e1500), 500)dpy-17(e164), 64)unc-32(e189), 89)dpy-18(e364) and dpy-18(e364)unc-64(e246). For each strain six broods were scored The only recombinant found was a single dpy-18 recombinant in the strain dpy-19(e1259)glp-1(q339)/dpy-18(e364)unc-64(e246) We think that the suppression of recombination by the e1259q339 chromosome is specific to Chromosome III rather than a generalized suppression of recombination because dpy-19(e1259)glp-1(q339)/+ + III; unc-60(e723) dpy-11(e224)/+ + V animals segregate both unc-60 and dpy-11 recombinants at the expected frequency. The e1259q339 chromosome contains a DNA rearrangement with a breakpoint in the glp-1 gene (Austin and Kimble this issue) This rearrangment may be the cause of the suppression of recombination Although we have not yet determined the nature of this rearrangment, no dead eggs are laid by e1259 q339 heterozygotes, suggesting that it is not a reciprocal translocation

Norman KR et al. (1998) Worm Breeder's Gazette "A mutation in alpha-spectrin affects elongation of the embryo and myofilament organization."

Norman KR, Moerman DG

[Worm Breeder's Gazette, 1998]

An L1 lethal mutant ra409 was isolated from a screen designed to isolate embryonic lethals with defects in muscle development. ra409 animals have a unique phenotype in that mutant animals do not complete morphogenesis and do not elongate entirely. Although all tissues types examined have differentiated in ra409 homozygotes, many do not appear wild type. In particular, the body wall muscle and the overlying basement membrane are 2-fold wider than in wild type animals. In addition, muscle filaments in ra409 animals are oriented at a steeper angle to the longitudinal axis of the embryo than what is seen in wild type. This may be a consequence of either the lack of complete elongation or the expansion of muscle cells or basement membrane. We have also observed that the pharynx is much shorter in these mutant animals and labors during contraction. The actin filament organization and the antigen recognized by MH4 (intermediate filament like antigen) in the hypodermis both appear normal, with the exception that the MH4 staining pattern is two fold wider corresponding to the wider muscle quadrants in this mutant. Recent characterization and cloning of sma-1 (McKeown, C., Praitis, V. And Austin, J. 1997 Worm Meeting) provided a clue to aid in the identification of the ra409 gene. sma-1 is a viable mutant that has defects during morphogenesis similar to ra409 mutants. At hatching sma-1 mutants are half the length of wild type, a phenotype that is identical to ra409 homozygotes. Transformation rescue of sma-1 animals revealed that sma-1 encodes a beta-spectrin (McKeown, C., Praitis, V. And Austin, J. 1997 Worm Meeting). We have genetically mapped ra409 to a region on the X chromosome containing an alpha-spectrin (C. elegans genome sequencing consortium). Spectrins are a major components of the membrane cytoskeleton and are important for maintaining cell shape and polarity. Since beta-spectrin forms heterodimers with alpha-spectrin, we felt this alpha-spectrin might be a likely candidate gene for ra409. To determine if the ra409 lesion is in the alpha-spectrin gene we initially took an antisense RNA approach using a 2.8 kb partial cDNA obtained from Y. Kohara. Antisense as well as RNA injection into the gonad of wild type worms has been shown to phenocopy the mutant phenotype of several genes in the progeny of injected worms (Driver, S.E., Ali, M., Mello, C.C. 1997 Worm Meeting). The 2.8 kb cDNA to the 3! end of the alpha spectrin gene was used to generate RNA and this RNA was injected into wild type animals and the progeny were screened for the presence of the ra409 phenotype. Several L1 larvae from the injected wild type hermaphrodite resembled ra409 homozygotes, but unlike ra409 animals these were viable. Next, in an attempt to rescue ra409 animals we transformed animals with a purified cosmid clone, M01F12, that contains the entire alpha-spectrin gene (supplied from the Sanger Center). We obtained transformation rescue of the ra409 mutants thus demonstrating that ra409 is an allele of the alpha-spectrin gene. This alpha-spectrin, asp-1 (alpha spectrin), is 58% identical to the vertebrate non-erythroid alpha-spectrin and 63% identical to the Drosophila alpha-spectrin. The non-erythroid alpha-spectrins and ASP-1 contain an SH3 domain, a calmodulin binding domain and a calcium binding region. Now that the gene is identified the goal will be to determine how it functions during development. One area we will investigate is what role spectrin plays in the patterning of the muscle cytoarchitecture.

Seydoux GC et al. (1988) Worm Breeder's Gazette "lin-12 is required in the somatic gonad for germ line development."

Seydoux GC, Greenwald IS

[Worm Breeder's Gazette, 1988]

lin-12(0) hermaphrodites are essentially sterile. As reported in the last edition of the Gazette, Tim Schedl first recognized that, in lin-12(0) hermaphrodites, the proximal germ cells do not enter meiosis but seem to continue to proliferate in the mitotic cycle, like the distal germ cells (see Figure), although gametes are present in the loop region. We were interested in ascertaining whether lin-12 is required in the somatic gonad or in the germ line for proper germ line development. Mosaic hermaphrodites obtained from the strain qDp3: 65) 51) 41) were useful in distinguishing between these two possibilities. [qDp3 ( Austin and Kimble, 1987) complements ncl-1, 41) is a null allele of lin- 12.] Five mosaics were obtained in which half of the gonad ( either the Z1 or the Z4 descendants) was Ncl- and therefore presumably lin-12( 0). The presence of the Dp in the germ line was ascertained by looking for WT animals among the progeny of the mosaic hermaphrodite. In 5/5 mosaics, the half of the gonad not expressing lin-12 exhibited the lin-12(0) defect in the germ line, while the other wild-type half contained a normally differentiated germ line (see Figure); all mosaics produced some wild-type progeny. These results are consistent with lin-12 function being required in the somatic gonad for wild-type development of the germ line. (Our previous tentative inference of a requirement for lin-12(+) in the germ line was based on a mosaic phenotypic class that is also consistent with the interpretation presented here.) One possibility is that lin-12(0) animals have another, as yet unidentified, cell fate transformation in the gonad. This transformation would create a cell in the proximal part of the gonad that can promote mitosis in the germ line, a role normally associated only with the distal tip cell. Another possibility is that lin-12 plays a more direct role in germ line development. We hope to distinguish between these possibilities by laser ablation experiments.

Fitzgerald K et al. (1992) Worm Breeder's Gazette "Structure/Function Analysis of the lin-12 Protein"

Fitzgerald K, Greenwald IS

[Worm Breeder's Gazette, 1992]

We are currently trying to establish the relationship of various lin -12 structural motifs to their function in vitro. As a first step in this process we have attempted to map functional domains of the lin -12 protein using lin 12/glp-1 protein chimeras under lin-12 regulation. Two chimeras were constructed (see the figure below) by fusing corresponding introns or exons of lin-12 and glp-1 (Yochem and Greenwald Cell 58: 565, 1989; Austin and Kimble Cell 58: 565, 1989). Each construct contains 3.4 kb of lin-12 5' flanking sequence, which is sufficient to allow a lin-12 genomic DNA clone to complement a lin-12 null mutation (Wilkinson and Greenwald WBG 12: 61, 1992). Each construct was injected at a concentration of 6ug/ml along with a dominant rol-6 marker (Mello et al, EMBO J 10: 3959, 1991) at lOOug/ml into N2 wild type hermaphrodites. Arrays from the resulting Roller lines were then crossed into a lin-12 (n676n909am)background and assayed for rescue of the Lin-12 (0)sterility and protruding vulva defects. In each case the chimera containing array was able to complement lin 12(n676 n909 ).These results indicate that lin-12 /glp-1 chimeras most likely will not be useful for mapping functional domains of lin-12 (or glp-1 ).However, these results are interesting in that they strongly support the hypothesis of Lambie and Kimble (Development 112: 231, 1991), who suggested that lin-12 and glp-1 are functionally interchangeable based on the novel "Lag" phenotype of a lin-12 (0)glp-1 (0) double mutant. (Additional support for this hypothesis was also provided by Mango et al, Nature 352: 811, 1991.) We have now begun a structure/function analysis of the lin-12 protein by making specific deletions and point mutations, starting with deletions of various EGF like repeats in the extracellular portion of the protein.

Lee RYN et al. (1993) Worm Breeder's Gazette "An eat With a Sequence"

Avery L, Lee RYN

[Worm Breeder's Gazette, 1993]

We are interested in how genes regulate the contractile activity of the pharynx to bring about efficient trapping and intake of bacteria. eat-4 seems to participate in the fine tuning of pharyngeal muscle contraction and relaxation. Worms with eat-4 mutated have partial contractions of the corpus and slow relaxations of the corpus as well as terminal bulb. As a consequence, mutant worms are unable to swallow bacteria efficiently. They are pale, slow growing and produce smaller broods (Avery 1993, Genetics, in press). No other phenotypes have been observed. We have four alleles of eat-4 ,with ad572 being the strongest. The phenotype of eat-4 mutants is probably due to lost-of-function because all four alleles are fully recessive; ad572 /ad572 /qDp3 is wild type; and ad572 /qDf2 is Eat. We suspect that ad572 may be a null since ad572 /qDf2 does not seem to be more Eat than ad572 homozygote. In addition, new alleles of eat-4 have been isolated at high frequency (2 / 2653 EMS mutagenized chromosomes) in a non-complementation Fl screen. This screen did not bias towards viable mutants. We believe that eat-4 is in the cosmid ZK512 based on three lines of evidence. First, eat-4 has been mapped between unc-32 and emb-9 by 3-point-crosses. Moreover, since eat-4 is uncovered by qDf2 and since qDf2 has its left breakpoint in glp-1 (Austin et al 1989. Development Suppl: 53), eat-4 should be between glp-1 and emb-9 on LGIII. Second, we have shown that ZK512 can rescue two eat-4 mutant alleles by germ-line transformation. Third, we have detected an RFLP (probably an insertion) in ad572 by Southerns probed with ZK512 .The restriction pattern is changed back to the wild type in an ad572 intragenic revertant. (ad572 reverts at a low frequency (<1 in ~6800)). ZK512 has recently been sequenced by the C. elegans genomic sequencing project. (Trevor Hawkins kindly provided us with the sequence). Using the GENEFINDER program, we searched for potential genes around the segment in which the RFLP was detected. We found a candidate gene which does not seem to have any significant homology with known proteins. It is intriguing that eat-4 affects both contraction and relaxation. Since the strongest mutant has no phenotype outside of the pharynx, we think that it probably is not a house-keeping gene. We are continuing the molecular analysis of eat-4 to find out if it indeed corresponds to the locus found by GENEFINDER.

Seydoux GC et al. (1989) Worm Breeder's Gazette "ectopic mitosis in the germline of lin-12(0) and WT hermaphrodites."

Seydoux GC, Schedl TB, Greenwald IS

[Worm Breeder's Gazette, 1989]

In lin-12(0) hermaphrodites, proximal germ cells fail to enter meiosis and instead proliferate in the mitotic cycle as distal germ cells do. Meiosis and gametogenesis are observed only in the loop region where both oocytes and sperm are formed (WBG 10(1):108). Mosaic analysis has shown that lin-12 function is required in the somatic gonad to prevent proximal mitosis (WBG 10(2):82). One possibility is that in lin-12(0) hermaphrodites, a cell in the proximal part of the somatic gonad promotes mitosis in the germline, a role normally associated only with the distal tip cell (DTC). Ablation of such a cell would restore meiosis proximally in lin-12(0) hermaphrodites. Indeed, we have found that ablation of all anchor cells (ACs) restores proximal meiosis in lin-12(0) [unc-36(e251)lin-12( n941)] hermaphrodites. Conversely, ablation of all somatic gonadal cells except for the DTCs and one AC results in a gonad with proximal mitosis. Thus, an AC is both necessary and sufficient to promote proximal mitosis in lin-12(0) hermaphrodites. In addition, we have found that even in wild type, an AC can promote proximal germline mitosis when certain of its somatic neighbors Z1.p(a/p) and Z4.a(a/p) descendants] are ablated. We conclude that, in wild type, cell-cell interactions among certain Z1.p(a/p) and Z4.a(a/p) descendants are necessary to keep the AC from promoting proximal mitosis in the germline. Since proximal mitosis is observed in lin-12(0) mutants, we infer that these cell-cell interactions require lin-12 activity. Keeping in mind that the AC is in close contact with proximal germ cells at least up to the L2 molt, we propose two models to account for our findings. In one model, we postulate that cell-cell interactions prevent the AC from expressing the 'DTC-to-germline' signal. Either ablation of somatic cells neighboring the AC or removal of lin-12 activity would result in the expression of the 'DTC-to-germline' signal by the AC, which would promote mitosis in proximal germ cells. Another model relies on the proposal that lin-12 functions as the receptor for the 'AC-to-VU' signal, which emanates from the presumptive AC (Seydoux and Greenwald, 1989) and that glp-1 functions as the receptor for the 'DTC-to-germline' signal (Austin and Kimble, 1987; Yochem and Greenwald, 1989). In view of the functional and molecular similarities between lin-12 and glp-1 (see Yochem and Greenwald, 1989), we postulate that the 'AC-to-VU' signal can activate glp-1 in germ cells. In wild type, the 'AC-to-VU' signal preferentially binds to lin-12 and not glp-1. In lin-12(0) hermaphrodites, however, the 'AC-to-VU' signal is available to activate glp-1 and promote mitosis in proximal germ cells. Similarly, in operated wild type hermaphrodites in which cells we hypothesize are expressing lin-12 have been ablated, the 'AC-to-VU' signal can activate glp-1 in proximal germ cells. We are currently testing the requirement for glp-1 for the proximal mitosis observed in operated lin-12(+) hermaphrodites by repeating our ablation experiments in glp-1(ts) animals. We are also reverting the lin-12(0) germline defect to help us distinguish between different models and perhaps to define the 'AC-to-VU' signal.