mes-2, mes-3, mes-4 and mes-6 mutants produce sterile but otherwise healthy hermaphrodite progeny. The MES proteins are predicted to function in complexes to repress gene expression in the germline: 1) MES-2 and MES-6 are homologs of Polycomb Group (PcG) proteins, which in Drosophila function in complexes to repress gene expression. 2) MES-2 and MES-6 depend on each other and on MES-3 for correct nuclear localization. 3) Transgenes on extrachromosomal arrays are silenced in wild-type C. elegans germlines, but can be desilenced in mes mutant germlines (Kelly & Fire), suggesting that MES proteins normally participate in repressing germline gene expression. We have investigated the localization of MES-3 and MES-4. MES-3 is a novel protein, and MES-4 is a SET-domain protein with PHD fingers. Both proteins resemble MES-2 and MES-6 in being abundant in all nuclei of early embryos and remaining detectable only in the primordial germ cells in later stage embryos. However, MES-4 shows much more dramatic localization to chromosomes than any of the other MES proteins, and intriguingly is associated with the autosomes but not the X chromosomes. In the germline, MES-3 levels are high in the distal mitotic region, low in pachytene nuclei, and increasing in maturing oocytes, while MES-4 is only detectable in the distal mitotic region. Coimmunoprecipitation experiments have demonstrated that MES-2, MES-3, and MES-6 are in a complex in embryo extracts. The size of this complex is predicted to be ~300kD, based on sucrose gradient and gel filtration analyses. In vitro binding assays have shown that MES-2 and MES-6 interact directly with each other, but not with MES-3, suggesting that an additional component(s) mediates the association of MES-3 with MES-2 and MES-6. MES-4 protein did not coprecipitate with any of the other three MES proteins. Our finding that MES-4 protein spreads to the X chromosomes in mes-2, mes-3, or mes-6 mutants suggests that the MES-2,3,6 complex somehow regulates MES-4's binding pattern. These new results, along with our previous findings that the Mes mutant phenotype is highly sensitive to X-chromosome dosage, suggest that in wild-type worms, MES-2, MES-3, and MES-6 act in a complex to exclude MES-4 from the X chromosomes and that this leads to repression of X gene expression.