Two genes,
deg-1 and
mec-4 ,can be mutated to induce neurodegeneration in C. elegans (Chalfie and Wolinsky, Nature 345: 410, 1990; Driscoll and Chalfie, Nature 349: 588, 1991). The proteins predicted from partial cDNAs from these genes, degenerins, are approximately 50% identical. Because additional cross-hybridizing sequences are found in C. elegans genomic DNA, we have been interested in determining whether additional degenerin genes exist. We were able to amplify several segments from genomic DNA using PCR and degenerate oligos for two conserved regions in
mec-4 and
deg-1 (CGDPRF...ADFGGQ). Although a
mec-4 probe hybridized with two of the fragments (one was the correct size for
mec-4 ),a
deg-1 probe did not. The lack of
deg-1 hybridization may signify that one of the conserved regions used to generate the oligos might be interrupted by an intron in
deg-1 (it is not in
mec-4 ).The second band showing
mec-4 hybridization was cloned and sequenced and had extensive homology to
mec-4 and
deg-1 .This clone was used to probe Chris Martin's 2-3 kb cDNA library and four cDNAs were found among 6x10 +E6screened. These cDNAs all represented the same transcript. The largest cDNA (2.3 kb) defines an ORF of 730 amino acids with two in frame stop codons about 150 bp from the 3' end. We are not sure if this cDNA clone contains the complete message since the two most upstream methionines are only 40 bp from the 5' end of the clone and no upstream stop codons are seen in frame. We are calling this gene
deg-2 .
deg-2 share extensive sequence identity with
mec-4 and
deg-1 (54% identity with
mec-4 over 469 aa sequence and 47% identity with
deg-1 over a 282 aa region - only partial cDNAs are available for these genes). Overall there is a better match (fewer gaps introduced and more extensive homology) between
mec-4 and
deg-2 than between either and
deg-1 .Like
mec-4 ,the
deg-2 predicted protein contains a hydrophobic domain near but not at the amino terminus (this was predicted for the
mec-4 protein from a comparison of genomic sequences of the C. elegans and C. briggsae genes by Monica Driscoll) and two Cys-rich domains and a second hydrophobic region; the second Cys-rich and hydrophobic regions are also predicted from the smaller
deg-1 cDNA. The C-termini of all three proteins are not similar, but all are highly charged. Thus, the degenerins may be membrane proteins that span membranes twice. One putative glycosylation site is conserved in all three predicted sequences. All three sequences also contain the Ala near the beginning of the second hydrophobic domain that is mutated in
mec-4 and
deg-1 to cause neurodegeneration. We are currently obtaining a full length genomic clone so that we can mutate this Ala to determine whether this
deg-2 mutation will also result in a dominant degeneration of neurons or other cells.
deg-2 was mapped to YACs Y51E1 ,Y75D9 ,and Y58B1 by Alan Coulson; this positions the gene between
xol-1 and
kin-9 on the X chromosome, approximately 1 map unit to the right of
deg-1 .Hybridization to the YAC grid revealed several weaker hybridizing spots (representing, minimally, DNA on chromosomes I, IV, and V). These YACs may identify additional homologues. Interestingly, the
deg-2 probe did not hybridize in these experiments to YACs containing
mec-4 or
deg-1 .We are pursuing the analysis of these other putative degenerin genes and examining the pattern of hybridization of
mec-4 and
deg-1 probes to the YACs.