In C. elegans, Eph signaling functions in embryonic neuroblast movements after gastrulation, and in epidermal enclosure. The C. elegans genome encodes one Eph receptor, VAB-1, and four ephrins: VAB-2/EFN-1, EFN-2, EFN-3, and EFN-4.
mab-26 mutants have defects in male tail morphogenesis (Chow and Emmons 1994). The predicted ephrin gene
efn-4 (aka F56A11.3), was located in the same region as
mab-26. A genomic clone containing F56A11.3 rescued Mab-26 phenotypes in transgenic arrays. These data combined with identification of mutant alleles demonstrate that
mab-26 and
efn-4 are the same gene. The EFN-4 protein encodes a divergent ephrin. EFN-4 overall is most similar to murine Ephrin-B2 (27% identity and 45% similarity). Like other worm ephrins, EFN-4 is predicted to be GPI-anchored. Unlike all other ephrins EFN-4 contains a 24 amino acid insert in the putative receptor-binding domain. The mutation
bx80 is a deletion of second exon of
efn-4, and thus is likely a null mutation. Three other alleles from the Cambridge collection cause alterations in
efn-4:
e36 causes an early nonsense mutation, and the weak alleles
e660 and
e1746 (provided by Jonathan Hodgkin) cause missense alterations in the receptor binding domain. The phenotypes of
efn-4 mutants indicate that like other eph signaling proteins it is required for embryonic morphogenesis. Genetic interactions with the other ephrins and
vab-1 are inconsistent with EFN-4 acting only in the VAB-1 pathway (see abstract by Moseley and Chisholm). These observations have led us to examine the the EFN-4/VAB-1 binding interaction using cell culture experiments. Previously the EFN-1 ephrin was shown to bind VAB-1::AP fusion proteins with high affinity in mammalian cell culture assays (Chin-Sang et al 1999). We are using parallel approaches to test EFN-4/VAB-1 interactions. An EFN-4::GFP reporter gene is expressed in a subset of neurons. This expression pattern suggests that EFN-4/VAB-1 signaling in the embryo occurs between neurons and is thus required non-autonomously for epidermal development. We will test this idea using genetic mosaic analysis and tissue-specific rescue experiments.