Worm Breeder's Gazette 11(5): 47

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.

Fingering tra-1

David Zarkower and Jonathan Hodgkin

The tra-1 gene acts as a genetic switch controlling sexual 
differentiation in C.  elegans.  Loss-of-function mutations in tra-1 
cause XX animals to develop as males, while gain-of-function mutations 
transform both XO and XX animals into fertile females.  A small contig 
was identified as likely to contain at least part of the tra-1 gene 
based on the detection by one cosmid of rearrangements cosegregating 
with several tra-1 mutations (WBG vol.  10, #2).
We have examined transcription from the tra-1 region by northern 
analysis and have detected two groups of transcripts.  One correlates 
precisely with the region delineated by the rearrangements, while the 
second is encoded by the region immediately adjacent to it.  We have 
focussed so far on the first group of transcripts, but it will be 
important to establish what role, if any, the adjacent transcripts 
have in tra-1 function.
There are at least three transcripts within the region, all of low 
abundance.  One is about 1.5 kb and is present only during L2.  A 
second RNA is about 5 kb and present during all hermaphrodite stages 
and in adult males.  The third is shorter than 1 kb and has been 
detected only in adult males.  We have not yet examined larval males.  
In addition there may be one or two other, less reproducibly detected, 
transcripts from the region.
We have isolated cDNAs corresponding to the L2 RNA both from the Kim 
library and by PCR.  The mature message is spliced from a primary 
transcript of more than 11 kb and can encode a protein of at least 340 
amino acids (the 5' end has not yet been unambiguously determined) 
which includes two C-terminal ''zinc-finger'' motifs.  The fingers are 
45% identical to those of human GL1, a gene amplified in glioblastoma 
tumors, while the non-finger region (which is serine-rich) shows no 
similarity to other proteins in the databases.  After the second 
finger there is a 'link' sequence and a cysteine at the correct 
position to be part of a third finger, so it is likely that at least 
one other transcript encodes a protein containing additional zinc 
fingers.  Indeed, preliminary data suggest that the 5 kb message 
contains additional sequences in the region of the fingers.  We are 
currently isolating cDNA clones of the 5 kb and male-specific messages.

To help define the functions of the putative tra-1 RNAs, we have 
examined transcription in tra-1 mutants.  Only the weaker mutations 
can be easily analyzed in homozygotes, because the stronger alleles 
eliminate self-fertility.  Several alleles produce novel transcripts 
in addition to the normal length RNAs.  One mutation, e1488, a 
rearrangement upstream of the predicted start of transcription, 
appears to have low levels of the L2 message.  Since e1488 animals 
have a hermaphrodite gonad and germ line and are self-fertile (albeit 
not very), but have a strongly masculinized soma, the L2 product may 
function primarily in the non-gonadal soma.
We also have analyzed some stronger mutations in strains carrying 
eDp6, a free duplication covering tra-1.  Several have aberrant 
transcripts, further suggesting that these RNAs are tra-1 products, 
but the presence of wild type transcripts from eDp6 makes it difficult 
to assign functions to particular RNAs.  To overcome this problem, we 
are analyzing cDNA from single homozygous animals using PCR.  One 
allele, e2334, contains a large insertion.  In the e2334 L2 transcript 
all but about 100 bases of the insertion are spliced out.