Worm Breeder's Gazette 14(1): 30 (October 1, 1995)
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.
Dept. of Medical Genetics, University of Toronto, 1 King's College Circle,
Toronto, Canada Male development in C. elegans requires the activities of the three fem genes. Current models of sex determination suggest that in X0 animals, they promote male cell fates in somatic tissues by negatively regulating tra-1, and that in XX animals, they are prevented from interfering with tra-1 activity by the transmembrane protein encoded by the tra-2 locus, TRA-2A. We presented evidence at the last C. elegans meeting that a direct interaction between the C-terminal domain of TRA-2A and FEM-3 is important for negatively regulating the masculinizing activity of the fem genes. [See abstract by Spence et al.] On the hypothesis that when liberated from the influence of TRA-2A, FEM-3 engages in new specific interactions to bring about male development, we used the yeast two-hybrid system to screen for other proteins that interact with FEM-3. We used a "bait" protein consisting of FEM-3 fused to the DNA- binding domain of GAL4 to screen a library of C. elegans cDNAs fused to the GAL4 transcriptional activation domain. The library was constructed and generously provided by R. Barstead, and the fem-3 cDNA was a gift from J. Kimble. A screen of 2.3 million yeast transformants yielded three cDNAs encoding proteins which interacted specifically with the FEM-3 bait. One encoded a C-terminal fragment of TRA-2A, as expected from our earlier observations. The other two mapped to the immediate vicinity of fem-2 on the genomic physical map. Expression of the larger of the two from the heat-shock promoter partially rescued male somatic development in X0 animals homozygous for the putative null mutation fem-2(e2105). Both cDNAs can encode a member of the Type 2C protein serine/threonine phosphatase family identical in sequence to that predicted from the sequence of fem-2, and provided before publication, by D. Pilgrim and colleagues. Coimmunoprecipitation of FEM-2 and myc epitope-tagged FEM-3 following their synthesis in rabbit reticulocyte lysates confirmed their ability to associate. Bacterially expressed GST-FEM-2 fusion protein was also able to bind radiolabelled FEM-3 from in vitro translation reactions in batch affinity experiments. We have used the latter assay in a deletion analysis to locate the region of FEM-2 responsible for binding FEM-3, and our results to date suggest that a region of about 190 amino acids near the N-terminus contains sequences required for FEM-3 binding. Bacterially expressed GST-FEM-2 exhibits phosphatase activity in vitro with 32P-labelled casein as substrate. The activity is completely magnesium-dependent, confirming that FEM-2 is a Type 2C protein phosphatase. A report that other Type 2C phosphatases require Mg2+ for substrate binding led us to test the magnesium dependence of the interaction between FEM-2 and FEM-3. The interaction does not require magnesium, but we cannot exclude the possibility that FEM-3 may nevertheless be a substrate for the FEM-2 phosphatase. We have made several point mutations which alter conserved residues in FEM-2 and abolish its phosphatase activity in vitro without affecting its ability to bind FEM-3. We are currently testing alleles carrying these mutations for their ability to rescue the phenotype of fem-2 mutants. We propose that a direct interaction between FEM-3 and FEM-2 protein phosphatase is required for male development in C. elegans. We cannot discriminate among several plausible hypotheses concerning the biochemical consequences of the interaction. FEM-3 may become the target of an (unknown) inhibitory kinase upon binding to TRA-2A and require dephosphorylation by FEM-2 for its activation. Alternatively, FEM-3 may regulate the localization, activity or substrate specificity of the FEM-2 phosphatase.