Worm Breeder's Gazette 15(2): 37 (February 1, 1998)

These abstracts should not be cited in bibliographies. Material contained herein should be treated as personal communication and should be cited as such only with the consent of the author.

An HMG box protein encoded by cog-2 and a novel protein encoded by cog-4 are involved in vulva morphogenesis

Wendy Hanna-Rose, Min Han

Department of MCD Biology and HHMI, University of Colorado, Boulder, CO 80309-0347

        In screens for mutants defective in vulva morphogenesis, we have
isolated multiple alleles, defining at least three new genes, that cause
a defect such that the uterus and the vulva fail to make the proper
connections resulting in an Egl- phenotype.  These mutants have been
named cog or connection of gonad defective.  
        In animals carrying lesions in the cog-2 gene, the anchor cell
(AC) appears to fail to fuse with the uterine seam cell (utse).  The AC
remains at the apex of the vulva, blocking the entrance from the uterus.
cog-2 animals also show a low penetrance of gonad migration defects.  We
mapped cog-2 to the X chromosome between mec-2 and unc-6, and the Egl-
phenotype of cog-2 was rescued by a plasmid containing the 3' end of
cosmid T22B7.  Genefinder predicts two genes in the rescuing region,
T22B7.1 and W01C8.2.  However, analysis of cDNA clones obtained from a
C. elegans mixed stage cDNA library obtained from Stratagene proved that
portions of the two predicted genes and additional upstream sequences
are actually part of a single transcript that we are calling cog-2.  The
predicted protein encoded by cog-2 belongs to a large family of proteins
called SOX, SRY related HMG box, proteins.  COG-2 is most similar to the
mammalian SOX-5 and SOX-6 proteins.  COG-2 is greater than 70% identical
to SOX-5 and SOX-6 over the 110 amino acid HMG box region.  The mouse
SOX-5 protein is a sequence specific DNA binding protein with the
ability to severely bend DNA upon binding, and many SOX proteins are
thought to play architectural roles in the transcription process.  
        We have identified lesions in four of our five cog-2 alleles. 
The ku205 allele has a G-A transition in the 5' splice site following
the fifth of ten exons.  Failure to splice at this site would lead to a
truncated protein and deletion of the HMG box.  ku194 and ku209
introduce stop codons prior to the HMG box, and ku207 introduces a
missense mutation in an HMG box residue that is highly conserved
throughout the HMG family not just within the SOX subfamily.
        In animals carrying lesions in the cog-4 gene, the thin process
at the top of the vulva that is normally formed by the utse is replaced
by a thick layer of tissue with undetermined origin.  This thick layer
of tissue appears to block the entrance to the vulva from the uterus. 
The anchor cell in cog-4 mutants appears to successfully fuse to the
utse.  cog-4 mutant males mate inefficiently and appear to have defects
in late stage male tail morphogenesis.  We mapped cog-4 to the left of
the physical marker, stP100 on LGII.  The Egl- phenotype of cog-4 can be
rescued with cosmid C50D2 or a plasmid containing a single gene,
C23H3.1, as predicted by Genefinder.  A cDNA obtained from the same
library as above confirmed that the four predicted exons of C23H3.1
reside in a single transcript.  However, the 5' end of the cDNA was
incomplete and it is still unknown if additional upstream sequences are
included in the transcript we are calling cog-4.  Additional sequences
are suspected since a probe that includes only sequences in and near the
predicted C23H3.1 gene identifies a much larger transcript on a
northern.  cog-4 appears to encode a totally novel protein.  The two
alleles of cog-4 we have isolated fail to complement n481, an allele of
egl-26.  We have identified lesions in n481 and an additional allele,
ku211.  n481 and ku211 each introduce a missense mutation into an exon
of the novel cog-4 gene.