sup-39 mutations have been identified as allele-specific dominant suppressors of
unc-73(
e936) [1]. Two alleles of
sup-39 were isolated by the ability of these mutations to suppress the
unc-73(
e936) coordination defect. By themselves,
sup-39 mutations cause a maternal effect of 50% embryonic lethality. The sequence of UNC-73 has been determined and the gene encodes a rac GTPase guanine nucleotide exchange factor required for axon guidance [2]. The
e936 allele of
unc-73 contains a mutation in intron 15 that changes the canonical G found at the first base of the intron to a T. This mutation should disrupt the ability of the splicing machinery to use the wild type 5' splice site of this intron. Since
sup-39 mutations are involved in the allele-specific suppression of a splice site defect, perhaps mutations in
sup-39 alter splicing regulation. Splice sites are determined by a combination of RNA:RNA, RNA:protein and protein:protein interactions between the pre-mRNA, the small nuclear RNPs and splicing factor proteins. Mutations of 5' splice sites, like those in
unc-73(
e936), generally lead to two possible outcomes in splicing; skipping of the exon bordered by the mutated splice junction or activation of cryptic splice sites. Using reverse transcription and PCR, I have found that in
unc-73(
e936) animals, the 5' splice site mutation in intron 15 leads to activation of three cryptic 5' splice sites, not exon skipping. One cryptic site occurs 23 bases into the intron and another occurs one base prior to the intron. Splicing to either of these cryptic 5' splice sites leads to splicing into the same nonsense reading frame. The third cryptic splice site is unusual and represents an unprecedented finding; the wild type splice site is used even though this intron begins with T instead of the canonical G. By doing RT-PCR with 32P labeled primers and separating the products on a sequencing polyacrylamide urea gel, I have found that for
unc-73(
e936) worms, the "wild type" splice site is used in 10% of the stable
unc-73 mRNAs. This is consistent with the finding that
unc-73(
e936) worms make a very low level of full length UNC-73 protein relative to wild type worms and that
unc-73(
e936) is not a null mutant [2]. I then tested for the relative use of the cryptic splice sites in the presence of the allele-specific suppressor,
sup-39. In
unc-73(
e936);
sup-39 animals, the same three cryptic 5' splice sites of
unc-73 intron 15 are used. However, the relative levels of use of the different splice sites change. In
unc-73(
e936);
sup-39 animals, the use of the in-frame splice site increases to 35% of the stable
unc-73 message. Based on these data, a model for the mechanism of suppression of the
unc-73(
e936) allele by
sup-39 mutations can be made.
sup-39 mutants change the preference of the splicing machinery for different cryptic splice sites, which can lead to an increase in the amount of in-frame
unc-73 message. This could in turn lead to production of sufficient amounts of the UNC-73 gene product required for development of a coordinated worm. Sup-39 may represent a new type of informational suppressor involved in regulation of cryptic splice site usage. Currently I am testing this model for
unc-73 suppression, trying to identify other genes whose splicing may be altered by
sup-39 mutations, and trying to identify the
sup-39 gene.