Linking PAR polarity proteins to cell fate regulation: analysis of MEX-5 localization in Caenorhabditis elegans embryos

Abstract

Specification of somatic and germline lineages in the nematode Caenorhabditis elegans requires the establishment of anterior-posterior polarity in early embryos. Polarization begins by a sperm-induced cue in 1-cell embryos that cortically localizes PAR polarity proteins, including the Ser/Thr kinases PAR-1 and PAR-4/LKB1. Capping at the anterior pole of non-muscle myosin and several PAR proteins leads to asymmetric localization of proteins such as MEX-5 and MEX-6. MEX-5,-6 are closely-related CCCH zinc finger proteins required for germline specification that are anteriorly localized in 1-cell embryos. While no direct targets of the PAR proteins have been described in C. elegans, MEX-5,-6 are proposed to function as key intermediaries in the transduction of polarity cues from PAR proteins to downstream cell fate regulators. To understand how MEX-5 asymmetry is established, I constructed a series of fusion proteins containing Green Fluorescent Protein (GFP) fused to all or part of the MEX-5 protein; these fusion proteins allowed me to monitor asymmetry in living embryos. Deletion analysis of GFP:MEX-5 identified a single residue, Ser458, that is necessary for anterior localization of GFP:MEX-5 in 1-cell embryos. MEX-5 is phosphorylated at Ser458 in vivo, and this phosphorylation occurs at the onset of MEX-5 expression in the gonad. In a screen of 41 Ser/Thr kinases, I found that only PAR-1 and PAR-4 are necessary for MEX-5 phosphorylation. PAR-1 kinase activity is required for the initial phosphorylation of MEX-5, as a kinase-dead allele of par-1 abolished all staining with anti-MEX-5(pSer458) in oocytes and embryos. PAR-4 kinase activity is required to maintain MEX-5 phosphorylation in mature oocytes; anti-MEX-5(pSer458) staining decreased progressively in par-4 mutant oocytes, and was present only at low levels in 1-cell embryos. While phosphorylation of MEX-5 is necessary for its asymmetry in 1-cell embryos, it is not sufficient; in par-1 alleles with mutations outside the kinase domain, MEX-5 was phosphorylated but remained symmetric. In summary, my research has described multiple roles for the PAR-1 and PAR-4 kinases in MEX-5 phosphorylation and localization. These results provide the first evidence of a role for the PAR kinases in setting up embryonic asymmetries prior to fertilization.