Li C et al. (1988) Worm Breeder's Gazette "development of the egg-laying circuit."

Li C, Chalfie M

[Worm Breeder's Gazette, 1988]

The VC motoneurons branch from the ventral nerve cord to innervate the vulval muscles with a complex tuft of processes or arborization. We have been examining the assembly of this egg-laying circuit. We have previously reported that the formation of this arborization (as assayed immunocytochemically with an anti-FMRFamide antibody) is not dependent on neuron-target muscle or neuron-neuron (HSN) interactions, but is dependent on the presence of vulval cells. Vulval cells, irrespective of their position or their lineage, are capable of inducing VC branching into the vulval region. Only vulval cells derived from a primary lineage, however, are necessary to induce the branching and form the arborization. The presence of vulval cells alone, however, is not sufficient for the formation of the arborization. We have found that the somatic gonad also plays a role in the formation of the arborization. Because the anchor cell normally induces the vulva (1), gonadal ablations were performed in lin-15(n309) Multivulva mutants in which vulval lineages can be expressed independent of the gonad (2). Ablation of Z1 and Z4, the somatic gonad precursors, in lin-15 results in animals in which VC cell branches are present at the vulva and in pseudo vulvae, but the arborization is absent (n=12). By contrast, ablation of Z2 and Z3, the germline precursors, has no effect on the arborization (n=4). Induction of VC branching, therefore, can occur independently of the gonad, but the presence of the gonad is required for the formation of the arborization. We have begun to examine the question of how the vulval muscles are organized into their final positions around the vulva. Previous work suggested that the final position of the sex myoblasts, the vulval muscle precursors, is regulated by the somatic gonad (3, P. Sternberg, pers. comm.). We examined whether the vulval muscles were under the same instructive cues as the VC neurons. Vulval muscles were visualized immunocytochemically with an antimyosin antibody (kind gift from D. Miller), and were scored for their position around the vulva and/or between VC4 and VC5 and for their orientation. Our results suggest that the position of the sex myoblasts largely determines the position of the vulval muscles. The final position of the muscles, however, is dependent on gonadal and vulval cues. When the somatic gonad is ablated in wild-type animals, the sex myoblasts migrate anteriorly to a variable extent (3) but usually to between P6. pp (about VC4) and P7.pp (about VC5) (M. Stern and H.R. Horvitz, pers. comm.). In similarly ablated animals (n=9), the muscles are observed scattered throughout the length of the animal, as far anteriorly as VC2 and as far posteriorly as halfway between VC6 and the anus. In one animal muscles were detected dorsally. These muscles do not appear to be body wall muscle and look like vulval muscles, and will be referred to as vulval-like muscles. In Vulvaless mutants, the sex myoblasts migrate to their normal position (P. Sternberg, pers. comm). In these animals vulval-like muscles develop in the correct position between VC4 and VC5, but are randomly oriented. The muscles, therefore, can appear in their correct position with cues from the gonad and without input from the vulval cells. Whether vulval muscle orientation is lost due to the loss of the vulval attachment point or whether vulval cues are also important for orientation can not be distinguished from these experiments. The vulval cells, however, can also play a role in the positioning of the vulval muscles. Ablation of the somatic gonad precursors in lin-15(n309) Multivulva animals results in a protruding vulva. In such ablated animals (n=14) vulval-like muscles in their correct orientation are detected at the protruding vulva and at 2-3/4 pseudo vulvae. No vulval-like muscles are detected except at the pseudo vulvae. The vulval-like muscles are derived from the M lineage. Ablation of Z1, Z4, and M in lin-15 animals (n=2) results in the absence of any vulval or vulval-like muscles. lin-15 animals in which no ablations have been performed or in which the germline precursors were ablated (n=3) have correctly positioned vulval muscles and no ectopic vulval-like muscles. These results suggest that in the absence of gonadal cues, the vulval cells are capable of directing the position and orientation of the vulval muscles. Nematode muscle differs from muscle in other animals since it sends muscle arms to nerve cells rather than nerve processes seeking the muscle. The mispositioning of the muscles in these experiments could be caused by muscles projecting to ectopic VC branches in the pseudo vulvae. We think this possibility unlikely, however, because vulval- like muscles can be present at pseudo vulvae in the absence of any ectopic VC branches, and, conversely, ectopic VC branches are seen projecting into pseudo vulvae with no vulval-like muscles. These results are consistent with the hypothesis that the VC cells do not provide cues to the vulval muscles. We propose that the components of this neural circuit, the VC cells and the vulval muscles, are organized by the somatic gonad and the vulva. Preliminary experiments suggest that the HSNs may be under similar control.

Clark SG et al. (1990) Worm Breeder's Gazette "lin-39 mutants are Supervuls"

Clark SG, Horvitz HR

[Worm Breeder's Gazette, 1990]

Mutations that prevent the formation of competent Pn.p cells or the expression of vulval lineages confer a vulvaless (Vul) phenotype. Previously (WBG 10(2): 45 1988; WBG 10(3): 134 1988), we described the isolation of Vul mutations as suppressors of lin-15 multivulva mutations. In a screen of 76,000 haploid genomes, we have isolated 156 mutations that suppress lin-15(n765ts), of which 80 are Vul. These include mutations in unc-83, let-23, let-341, lin-10, nSix allelic lin-15 suppressor mutations ( n1490, n1760, n1872ts, n1880, n2010, and n2110 III) cause P(3-8).p to join the hypodermal syncytium without dividing. This Vul phenotype is similar to that caused by n300, a mutation associated with the translocation nT1(IV,V). In N2, the vulval equivalence group cells P( 3-8).p divide at least once and can express vulval-specific cell lineages, while the non-vulval equivalence group cells, P(1,2,9-11).p, join the syncytium without dividing. Unlike most Vul mutants, in which one Pn.p cell division occurs, these new mutants appear to lack the vulval equivalence group. We refer to these mutations as Supervul mutations to distinguish them from other kinds of Vul mutations that allow Pn.p cells to join the vulval equivalence group but not to express vulval lineages. Chris Li (personal communication) observed that n1760 animals lack the anti-FMRFamide staining normally displayed by the six P(3-8).a- derived VC neurons of the ventral cord. In addition, ced-1; n1760 animals possess extra cell corpses in the ventral cord. Together, these observations suggest that the FMRFamide-staining VC neurons die in n1760 animals. VC neuron death and vulval abnormalities were observed in lin-39(n709) animals by Hilary Ellis. n709 fails to complement n1490 for at least the Vul phenotype, indicating that these six lin-15 suppressors are new alleles of lin-39. lin-39 maps between sma-3 and unc-36 on LGIII. Three of the new lin-39 mutants (n1760, n1880, n2010) are 100% Vul and slightly Unc, whereas n1490 and n2110 are less than 100% penetrant for the Vul and Unc phenotypes. n1872ts is 100% Vul and Unc at 25 C and 80% Vul and Unc at 15 C. n709 is only partially egg-laying defective and not Unc; however n7091nDf16 and n7091n1490 animals are Vul. Surviving VC neurons are often seen in n709 animals, but Chris has not seen any FMRFamide-staining VC cells in n1760 animals. These data indicate that n709 is probably a weak allele of lin-39.That both the Pn.a and Pn.p lineages are abnormal suggests that P(3-8) could be expressing a P(1,2,9,10) type of cell lineage in lin-39 animals (see figure below). We are currently investigating further the nature of the P(3-8) cell lineages and the cause of the Unc phenotype in lin-39 animals. [See Figure 1]

Li C et al. (1986) Worm Breeder's Gazette "FMRFamide-like immunoreactivity in C elegans."

Li C, Chalfie M

[Worm Breeder's Gazette, 1986]

The neuropeptide FMRFamide (Phe-Met-Arg-Phe-amide) and peptides homologous to FMRFamide have been found throughout the animal kingdom where they have been implicated as neurotransmitters, neurohormones, and neuromodulators Using polyclonal antibodies raised against synthetic FMRFamide (donated by R Calabrese and T. O'Donohue) and a whole mount staining procedure modified from S McIntire, C. Desai, and R H Horvitz, we mapped the distribution of FMRFamide-like immunoreactivity in N2 adult hermaphrodites There are at least nine immunoreactive cells (some paired, some unpaired) around the nerve ring, six cells along the ventral cord, one pair of lateral cells near the position of the vulva, one cell in the dorsal-rectal ganglion, and one cell in the pre-anal ganglion In addition, the nerve ring is intensely labelled, there are numerous immunoreactive processes running throughout the length of the ventral cord (all but one of these processes run on the same side of the hypodermal ridge as the immunoreactive cells in the cord), one process, which runs the entire length of the animal, in the dorsal cord and each lateral cord, and a complex arborization of immunoreactive processes that may innervate the vulva muscles Preincubation of the antiserum with synthetic FMRFamide eliminates staining in the animal Tentative identification of many of the FMRFamide-like immunoreactive cells have been deduced from analysis of mutants and from the morphology and position of the cells. For example, the number and position of the immunoreactive cell bodies in the ventral cord corresponds to the VC motoneurons Furthermore, staining of lin-39 mutants, in which many of the VC motoneurons die, reveals fewer cell bodies and processes in the ventral cord, supporting the identification of the ventral cells as the VC motoneurons Other identified cells include the RID motoneuron, either the RIF or RIG interneurons, the DVB interneuron/motoneuron, and the PVT cell We have recently discovered that staining of neurons and processes in the cords can be optimized by using animal fragments, which are derived by gentle homogenization of formaldehyde-fixed animals This procedure has allowed more consistent visualization of immunoreactivity in two cell types, the ALA cell and the CAN cells Acetone extracts of whole animals contain 120 nmoles of FMRFamide- like peptide per gram wet weight, as measured by radioimmunoassay (RIA) The antibody displays a similar recognition to the C. s) and to synthetic FMRFamide Based on antibody specificities (Marder et al , 1986) these data suggest that the antigen(s) contains at least a terminal -ArgPhe-amide We intend to purify and sequence the FMRFamide-like peptide(s) from C elegans.We are continuing to screen unc mutants to identify animals that have FMRFamide-related defects or that are FMRFamide-deficient. In addition, we are examining target cell (vulval muscles) influences on the induction or stabilization of (VC and PVT) processes in Vulvaless, Multi vulva, and Egghead mutants We will confirm that the CAN cells contain FMRFamide by examining vab-8 mutants, which have misplaced CAN cells. The CAN cells are believed to be secretory cells and have been shown to be critical for the survival of young animals and for the proper development of the posterior half of the animal (J Sulston and J. Hodgkin, WBG 5(1), 1980) . As neuropeptides function as neurohormones in many instances, the FMRFamidelike peptide may be one of the neurohormones secreted by the CAN cells during the development of the animal; animals that lack FMRFamide may have a phenotype similar to animals that have CAN cell- related defects Lastly, we have obtained an APlysia cDNA clone encoding FMRFamide from R Scheller Using this Aplysia clone to probe a C. lambda library (a gift from S Emmons), we have isolated two positive clones, which will be subcloned and

Duerr JS et al. (1996) Worm Breeder's Gazette "A C. elegans synaptic vesicle amine transporter"

Duke A, Rand JB, Duerr JS

[Worm Breeder's Gazette, 1996]

In mammals, monoamine neurotransmitters (including dopamine, norepinephrine, epinephrine, and serotonin) are transported into synaptic vesicles by one of 2 vesicular transporters, designated VMAT1 and VMAT2. These transporters contain 12 transmembrane domains and are part of a family that also includes the synaptic vesicle acetylcholine transporters (e.g., the C. elegans UNC-17 protein). The biological significance of the differences between the two transporters is unclear. The Genome Sequencing Project recently identified part of a C. elegans VMAT homolog, and we have now cloned a 2.0 kb cDNA (from Bob Barstead's LambdaZAP library) which appears to encode a full-length VMAT protein. The C. elegans homolog has comparable similarity to both of the mammalian sequences, suggesting that it is likely to be able to transport dopamine, serotonin, and other putative neurotransmitters, such as octopamine. We have prepared antibodies against a peptide derived from a region near the C-terminus of the VMAT homolog. Antibody staining indicates that the C. elegans VMAT homolog is primarily localized to synaptic regions of a subset of neurons. These neurons are (with one clear exception) distinct from those that are immunopositive for UNC-17 (the acetylcholine transporter). VMAT-positive cells include the dopamine containing ADE, CEP, and PDE neurons, the serotonin containing NSM, HSN, and RIH neurons, some of the cholinergic and FMRFaimde positive VC cells, as well as a few, still unidentified cells in hermaphrodites, and numerous cells in the male tail. Based on map position, the cat-1 gene is the best candidate to be the C. elgans VMAT homolog. In cat-1 mutants, dopamine (visualized by formaldehyde-induced fluorescence) and serotonin (visualized by immunohistochemistry) are no longer localized to neuronal processes, but are only found in cell bodies. Using our anti-VMAT antibodies, we have now shown that cat-1(e1111) mutants are completely deficient for synaptic VMAT immunoreactivity. We now plan to attempt rescue of the cat-1 mutant phenotype with cloned VMATs. Supported by research grants from NIH and OCAST.

Hobert O et al. (1996) Worm Breeder's Gazette "LIM-homeobox genes in the worm"

Hobert O, Liu Yanxia, Ruvkun GB

[Worm Breeder's Gazette, 1996]

LIM-homeobox (LHX) genes, which encode proteins with two Zn-finger-like LIM domains and a DNA-binding homeodomain, have been shown to play roles in nervous system development of both invertebrate and vertebrate organisms. Originally, three LIM-homeobox genes were characterised in C.elegans: lin-11 and mec-3 (founding members of the LIM-homeobox family) were identified genetically (Freyd et al., 1990, Way et al., 1988), while the ceh-14 LHX gene was identified in a sequence based screen for homeobox genes (B!rglin et al., 1989). By doing lots of GENEFINDER and BLAST searches, we found that the C.elegans genome project has sequenced 4 more LHX genes, located on cosmids ZC64, C40H5, K03E6 and C04F1, which we termed CeLIM-4,-5,-6 and -7, respectively. We are interested in characterising the function of these genes by expression pattern analysis, and phenotypic analysis of animals bearing mutations in these genes. CeLIM-4 is located on the X chromosome right next to ceh-18. We constructed a CeLIM-4-GFP translational fusion and found CeLIM-4 to be expressed exclusively in several pairs of head neurons of which at least one is an amphid sensory neuron. We are in the process of identifying these cells. CeLIM-4 maps close to the let-43 gene; we are trying to rescue this mutant with CeLIM-4 but think it is unlikely that a mutant head neuron-specific transcription factor would cause lethality. In collaboration with R.Plasterk we have started to search for Tc1-insertions. CeLIM-5 is the C.elegans ortholog of the Drosophila apterous and vertebrate LH2/LHX2-genes. We found CeLIM-5 to be encoded by the ttx-3 gene (see accompanying abstract by Hobert et al.). In adults, CeLIM-5/ttx-3 is exclusively expressed in the AIY interneuron. CeLIM-6 is located on the X chromosome (next to unc-1) and is most closely related to the Lmx-subclass of LHX genes. Like CeLIM-4 and CeLIM-5, CeLIM-6 also shows a neuronal specific expression pattern: A promoter fusion to GFP demonstrates CeLIM-6 to be expressed most strongly in two pairs of head sensory neurons. which we preliminarily identified as the ASG and AWA sensory neurons. In collaboration with R.Plasterk we have started to search for Tc1-insertion for this gene as well. CeLIM-7 maps on chromsome I close to several let-mutations (which we are trying to rescue) and is the ortholog of the islet-class of vertebrate LHX genes. We are in the process of subcloning its promoter to determine its expression pattern. The function of the vertebrate islet-1 gene in motorneuron generation (Pfaff et al., 1996) suggests the possibility of a neuronal expression pattern for this gene as well. We constructed a new lin-11-GFP fusion containing 10 kB of lin-11 promoter sequence and preliminary expression analysis reveals a similar expression pattern similar to a lin-11-lacZ expression pattern reported by Freyd & Horvitz (several pairs of head neurons, including one pair of sensory neurons, VC neurons, vulval cells). We will use this construct to monitor axon morphology of these neurons in the lin-11 mutant background. Preliminary analysis with a different, smaller lin-11-GFP construct generously provided by Sawa & Horvitz revealed VC process morphology defects. So far, for every known subclass of vertebrate LHX genes a C.elegans homolog has been identified. Given this observation and the number of genes that have been estimated to be sequenced already by the GSC (~80%), we don't expect (m)any more LHX genes to exist in the worm. The overlapping expression patterns of several vertebrate LHX genes in the nervous system has led to the proposal of a "LIM-code", in which distinct combinationsof LHX gene expression specify different motor neuron fates (Tsuchida et al., 1994). Although our expression pattern analysis is not complete yet, we note that so far there is no significant overlap of C.elegans LHX gene expression. If this holds true, it would suggest that C. elegans LHX genes may specify the simpler nervous system of the worm in manner that is distinct from the postulated combinatorial LHX interactions that may specify the vertebrate nervous system.

Pettitt J et al. (1996) Worm Breeder's Gazette "cdh-3 IS REQUIRED FOR HYP10 MORPHOGENESIS"

Wood WB, Plasterk RHA, Pettitt J

[Worm Breeder's Gazette, 1996]

cdh-3 encodes a member of the cadherin superfamily. We have previously reported the determination of the expression pattern of cdh-3 in embryos undergoing morphogenesis, as well as the isolation of a deletion allele of cdh-3 (WBG 13 (2), 76, 1994; 1995 Meeting Abstracts, #416). We have continued our analysis of cdh-3 expression using a GFP reporter construct and have further characterized the phenotype resulting from the pk87 deletion allele. Expression of the cdh-3::GFP reporter construct is confined to ectodermally derived cells. We observe expression, beginning just prior to the start of embryonic morphogenesis, in the seam cells, both arcade cells, F, U, the excretory cell, and hyp10 and hyp11. Expression in these cells continues throughout postembryonic development, with the exception of hyp10, hyp11 and the excretory cell; expression in these cells continues only into the L1. During the L3 stage we see cdh-3::GFP expression in the anchor cell and the invaginating vulval epidermis, continuing throughout the development of the vulva. At the same time we also see expression in the VC neurons and the two HSNs. More careful examination of animals homozygous for the pk87 mutation revealed abnormalities in the morphogenesis of hyp10 (see table). These defects can be rescued by a wild-type copy of cdh-3 and are consistent with the expression of cdh-3::GFP in this cell during the period that it is undergoing a change in cell shape. However, we have not been able to detect any other defects in pk87 homozygotes. The absence of defects in the other cells that express cdh-3 reporter constructs hints that other genes may substitute for cdh-3 function. Future work will address this possibility, both by isolating mutations in the other cdh genes that have been identified and by carrying out screens for mutations that cause additional phenotypes in a cdh-3(pk87) background. ____________________________________________________________________ hyp10 morphology# ____________________________________________________________________ Genotype N wild-type abnormal ____________________________________________________________________ cdh-3(+) 115 100% 0.0% cdh-3(pk87) 170 24% 76.0% cdh-3(pk87; pkIs243)* 112 97% 2.5% ____________________________________________________________________ # Abnormalities include ectopic protrusions as well as complete failure to elongate. * Strain carrying an integrated array of cdh-3(&#43).

Li C et al. (1990) Worm Breeder's Gazette "Update on the FMRFamide-like peptides in C."

Burglin TR, Li C

[Worm Breeder's Gazette, 1990]

About 25 neurons, many of which have been identified, stain with an antibody against the neuropeptide FMRFamide [see WBG 9(3):110]. The onset of FMRFamide-like expression in these neurons, as determined immunocytochemically, appears to be developmentally regulated. The first visible staining occurs in the nerve ring of young L1 animals. Shortly afterwards in L1, the cell bodies of about 8 neurons in the nerve ring region and anterior ventral cord processes become visible. During L2, about 13 cell bodies can be seen around the nerve ring, and ventral cord processes now extend to the tail. The posterior cells DVB and PVT, a dorsal process, and lateral processes begin to stain during mid to late L2. A few additional neurons around the nerve ring first become immunoreactive during L3 or L4. The VC cell bodies become visible during early adulthood. All neurons continue to express the peptide into adulthood, and no neuron is stained in larval animals that is not present in adult animals. We have recently isolated a cDNA encoding multiple FMRFamide-like peptides. Degenerate oligonucleotides against Phe-Met-Arg-Phe-Gly-Lys( Arg) were used to screen Stuart Kim's cDNA library; 16 hybridizing clones were selected. Because neuropeptides are synthesized as part of a large precursor polypeptide that is posttranslationally cleaved to yield many smaller peptides, we examined Sau3A restriction digests of the cDNA clones, and selected a clone that showed multiple hybridizing bands, as would be expected if several copies of FMRFamide are present in the precursor gene. This clone was subcloned and sequenced (shown below). The deduced amino acid sequence of the translation product reveals 8 potential neuropeptide sequences. Each of these peptide sequences is flanked by possible endoproteolytic cleavage sites (single, double, or triple basic residues), and ends with a C-terminal Gly, a potential amide donor. Seven of the peptides terminate with the sequence -Pro-Asn-Phe-Leu-Arg-Phe-Gly and extend 0- 4 residues towards the N-terminus by one of 5 separate sequences. These FMRFamide-like peptides would be recognized by the FMRFamide antiserum, and are likely to account for some, if not all, of the immunoreactivity seen. We are currently using the cDNA fragment as a probe to screen the MRC and MIT cosmid libraries to find the gene and position it on the physical and genetic maps. [See Figure 1]

Sommer RJ et al. (1995) Worm Breeder's Gazette "Evolution of vulva formation: Part VII: Mutants of Pristionchus with extra or missing Pn.p-ectoblasts."

Sommer RJ, Sternberg PW

[Worm Breeder's Gazette, 1995]

A dramatic reorganization in the ventral cord of Pristionchus pacificus with respect to presumed ancestors involves cell death of P(1-4).p and P(9-11).p. The vulva is formed by P(5-7).p and the remaining ectoblast, P8.p, is not a vulva precursor cell. In a genetic screen for egg laying defective mutants we found mutants with increased or decreased numbers of Pn.p-ectoblasts. Some of the ped (for P-Ectoblast Determination) mutant phenotypes are described below. We have assigned these mutations to separate genes based on complementation analysis (except for ped-2). ped-l(sy312): Some of the posterior Pn.p ectoblasts, P(9-11).p, survive. This phenotype is variable. If the ectoblasts are present, they form vulva-like structures in the posterior body region in addition to the normal central vulva. AC ablation does not influence vulva differentiation of these celIs . ped-2(sy319): The central ectoblasts P(5-8).p are absent, whereas the hypl2 cell is present. We have observed a few embryonic cell deaths in the central ventral cord. However, this analysis is complicated by high embryonic lethality. More than 95% of the eggs are lethal, and most of the remaining larvae die as L1. In the few surviving larvae, the VC neurons (P(5-8).aap in wild-type) die, like the corresponding cells in the anterior and posterior body region. Altogether, the ped-2 phenotype resembles the C. elegans ceh-20 phenotype (Chen & Stern, C. elegans Mtg. 1995, pp20). ped - 6(sy345): P(3,4).p survive and become ectoblasts. This phenotype is highly penetrant. The two additional cells form vulva-like structures with one or two additional anterior invaginations. AC ablations reveal that in ped-6 not only P(3,4).p but also P(5-7).p differentiate in an AC independent manner, indicating that ped-6 is a multivulva mutation, possibly similar to lin-15. ped-5(sy344): P(3,4).p survive as in ped-6, but do not form vulva-like structures. A series of VPC ablation experiments indicate that P(3,4).p are vulva precursor cells. After ablation of P(5,6).p, a normal vulva can be formed by P(3,47).p. Thus, for the anterior region ped-5 reconstitutes the vulva equivalence group seen in C. elegans, indicating a homeotic transformation . ped-4(sy346): P8.p undergoes additional vulva differentiation in appr. 40% of animals. A weaker effect is present on P7.p also. AC ablation show, that both cell can differentiate in an AC independent manner. Thus, both cells of the original ventrolateral pair P7/8L,R are affected. In summary, we can use a genetic approach to identify molecular changes accounting for cellular diversity.

Freyd G et al. (1991) Worm Breeder's Gazette "lin-11-lacZ FUSION GENE IS EXPRESSED IN VULVAL CELLS AND NEURONS"

Horvitz HR, Freyd G

[Worm Breeder's Gazette, 1991]

Mutations in the gene lin-11 cause the 2 vulval precursor cells to divide symmetrically instead of asymmetrically. Mutant animals cannot lay eggs and are slightly uncoordinated. We previously reported that lin-11 encodes a protein with both a homeodomain and two copies of a cysteine-rich region that potentially binds metal ions, called LIM ( Nature 344, 876,1990). Two other C. elegans proteins, mec-3 (Way & Chalfie, Cell 54, 5,1988) and ceh-14 (Burglin & Ruvkun, WBG 11:2, 52, 1990) and one mammalian protein, Isl-1 (Karlsson et al., Nature 344, 879,1990), also contain a homeodomain and two copies of the LIM motif. We have determined the minimal region of genomic DNA required for lin-11 function as assayed by germline transformation. We coinjected the rol-6 plasmid pRF4 (Mello et al., WBG 11:1,18,1989) with either the genomic cosmid clone ID6 or subclones of ID6 into lin-11 mutant animals and looked for roller progeny that have the ability to lay eggs. We have identified an approximately 10.5 kb region of the cosmid ID6 that is sufficient for lin-11 rescue. To identify cells in which lin-11 is expressed, we constructed a translational fusion between lin-11 and lacZ. The fusion was made using the vector pPD22.04 (obtained from Andy Fire), which contains a nuclear localization sequence, a synthetic intron, and unc-54 3' sequences in addition to lacZ. The clone used for microinjection contained most of the minimal region of genomic DNA required for lin- 11 rescue, but lacked 3 kb of genomic DNA encoding the carboxyl end of the lin-11 protein. The expected fusion product contains two copies of the lin-11 LIM domain, but not the homeodomain. The lin-11-lacZ fusion plasmid was coinjected with pRF4 into wild- type worms and stably transmitting lines were examined. Animals stained with X-gal (according to the procedure of Jeff Way, personal communication) have blue nuclei in the head, ventral nerve cord, vulva, and tail. Staining appears in embryos and the pattern changes during larval development as described below. More than 20 cells in the vulval region stain during the L3 and L4 larval stages, the time period in which the vulva is being formed in the hermaphrodite and the 2 cells undergo an asymmetric cell division. This staining disappears by the time animals reach adulthood. We are in the process of trying to determine the identities of the cells in the vulval region. We suspect that both daughters of the 2 cells express the fusion protein. In addition, 1 cell descendants and/or gonadal cells and/or muscle cells might also express the fusion protein. At least one nucleus in the pharynx stains and preliminary identification suggests it might be the neuron I5. Eight neurons in the lateral ganglion stain and are probably the left-right pairs of the neurons RIC, AIZ, ADF, and ADL (identified by Cori Bargmann). ADF and ADL are amphid sensory neurons and AIZ and RIC are interneurons. These neurons stain embryonically, and the staining persists throughout adulthood. Six cells in the ventral nerve cord begin to stain during larval development and continue to stain through adulthood. We believe that these cells are the VC neurons because of their number and positions. It is possible that lin-11 expression in some of these six classes of neurons is required for the animal to move properly, accounting for the slightly Unc phenotype of lin-11 mutants. We are testing the tentative identifications and functions of these cells by crossing the lin-11-lacZ fusion into mutants and by laser ablation.

Ellis HM et al. (1982) Worm Breeder's Gazette "on death and dying in C elegans."

Horvitz HR, Ellis HM

[Worm Breeder's Gazette, 1982]

Programmed cell death is a prominent feature of both the embryonic and postembryonic cell lineages of C. elegans. We have recently isolated two new classes of mutations affecting cell death. In one class cells that normally die instead survive, and in the other class cells that normally survive instead die. These mutants were isolated by mutagenizing ced-1(e1735) I, a mutation that prolongs the highly refractile stage of cell death (kindly given to us by Ed Hedgecock; see Newsletter Vol. 5, No. 2), and screening F2 progeny for extra or missing deaths. Two mutations in the gene ced(n717, n718) IV define the first class of mutations mentioned above. The late embryonic and early postembryonic deaths, which can normally be seen in L1 and L2 larvae of ced-1, are not seen in ced-1; n717 and ced-1; n718. Instead, extra neuronal-like nuclei are generated in several lineages that normally include deaths: the Q1 and V5.pa lineages of ced-1; n717 are identical to those of N2 except that cells that normally die form small compact nuclei. Thus, n717 and n718 appear to suppress the phenotype of ced-1 by acting prior to ced-1.Further characterization of the fate of the extra cells in ced(n717) may aid us in understanding the role of cell death in C. elegans development. For instance, it is conceivable that cell death functions to eliminate blast cells. However, no extra divisions occur in the Q1, V5.pa or Pn.a lineages of these mutants. Although embryonic lineages have not been determined in ced(n717), there is no evidence at hatching of grossly abnormal cell proliferation during embryonic growth. It seems more likely that programmed cell death eliminates cells that would otherwise differentiate. FIF staining of ced-1; n717 reveals an extra dopamine- containing cell in the postdeirid; the extra dopamine cell is probably V5.paapp, the cell that normally dies. We have also examined the fate of the normally male-specific cephalic companion cells in ced-1; n717 hermaphrodites. The cephalic companions have been suggested to play a role in male attraction to hermaphrodites. John Sulston has found that in N2 hermaphrodites the cells lineally equivalent to the cephalic companions die during embryonic development. In ced-1; n717 L4 hermaphrodites, the cephalic companions can be identified by their characteristically large nucleoli. However, adult ced-1; n717 hermaphrodites do not track toward other hermaphrodites. Based on these observations, it seems likely that cells that fail to die in ced(n717) generally differentiate. Given the large number of programmed cell deaths in the embryonic and postembryonic lineages, it is interesting that neither n717 nor n718 has any noticeable defects in behavior or morphology visible at the level of the dissecting microscope. The recessive mutation n709(III) results in extra cell deaths in the ventral cord. In N2, Pn.aap dies in the lineages of P1, P2, and P9- P12, while P3.aap-P8.aap survive and differentiate into VC neurons. In n709, Pn.aap deaths occur in the P3-P8 lineages as well as in the P1, P2, and P9-P12 lineages. By Nomarski criteria, the extra deaths appear indistinguishable from those that normally occur in the ventral cord. Counting ventral cord neuronal nuclei in DAPI-stained animals shows that the number of extra deaths in n709 ranges from one to six. Lineage studies have confirmed the variable expressivity of n709. Extra deaths have not been observed in any other postembryonic lineages of n709.