The KH domain protein MEX-3 is central to the temporal and spatial control of PAL-1 expression in the C. elegans early embryo. PAL-1 is a caudal-like homeodomain protein required to specify the fate of posterior blastomeres. While
pal-1 mRNA is present throughout the oocyte and early embryo, PAL-1 protein is expressed only in posterior blastomeres starting at the 4-cell stage. MEX-3 protein is present uniformly throughout the newly fertilized embryo but is enriched in the anterior relative to the posterior by the 4-cell stage. This anterior enrichment is consistent with its role in repressing PAL-1 translation and requires the CCCH Zn-finger protein MEX-5, and the RNA Recognition Motif protein SPN-4, which were identified as MEX-3 binding proteins. Genetic evidence indicates that
mex-5 is required for MEX-3 stability in the anterior, while
spn-4 is required for MEX-3 degradation. In
mex-5(-) embryos, MEX-3 is rapidly degraded and PAL-1 translation is no longer restricted to the posterior. In
spn-4(-) embryos, MEX-3 is present at unusually high levels throughout the embryo, but is largely unable to repress PAL-1 translation. MEX-5 protein is present in the anterior, while SPN-4 protein is present throughout the early embryo. Intriguingly, when
par-4 is depleted, the abundant MEX-3 in
spn-4(-) embryos retains the ability to repress PAL-1 translation throughout the embryo. PAR-4 is a serine/threonine kinase enriched in the cell cortex, but not asymmetrically localized along the anterior-posterior axis. These data suggest the following model: MEX-5 binds to MEX-3 in the anterior and protects it from degradation, allowing MEX-3 to bind the
pal-1 3''UTR and repress translation. MEX-3 that is not bound to MEX-5 becomes inactivated in a
par-4 dependent fashion, possibly through phosphorylation, then targeted for degradation through binding to SPN-4. In
spn-4(-) embryos, inactivated MEX-3 binds to and interferes with active MEX-3. To determine which regions of the MEX-3 protein are required for stability and which for degradation, different regions of the protein were fused to GFP and introduced into the C. elegans embryo. The N-terminus of MEX-3, which contains the two KH domain RNA binding domains, confers unusual
mex-5-dependent stability to GFP. In contrast, the C-terminus of MEX-3, which contains potential phosphorylation sites, confers unusual
spn-4-dependent instability to GFP. Experiments are in progress to determine which phosphorylation sites, if any, are required for MEX-3 degradation.