| Summary: | Through a structure-function analysis, I demonstrated the requirement of kinase activity for <i>Drosophila </i>PAR-1 function. Mammalian PAR-1 homologues, the MARKs, phosphorylate microtubule-associated proteins of the tau family to inhibit their interaction with microtubules. This biochemical activity is consistent with the microtubule defects observed in <i>par</i>-1 mutant oocytes, and prompted me to clone, characterise and generate protein null mutants in the sole <i>Drosophila </i>tau homologue. Remarkably, <i>tau</i> mutants are viable and fertile, indicating that this conserved cytoskeletal regulator is dispensable for <i>Drosophila</i> development. As Tau is unlikely to be a relevant germline substrate for PAR-1, this kinase must therefore function by a distinct mechanism to that of the MARKs. To define this mechanism, I isolated PAR-1 interactors by yeast tow-hybrid screening, and recovered the <i>Drosophila</i> 14-3-3 homologues, <i>14-3-3e</i> and <i>leonardo (leo).</i> 14-3-3 proteins interact with diverse targets in a phosphorylation-dependent manner to modulate their activity or localisation. The PAR-1 kinase domain binds to a region of 14-3-3 distinct from the phosphoepitope-recognition pocket and can phosphorylate a 14-3-3-binding site in Raf. PAR-1 may therefore promote 14-3-3/target interactions by generating the phosphoepitope essential for 14-3-3 binding and by directly delivering 14-3-3 to these targets. I show that mutations in <i>14-3-3e</i> and <i>leo</i> produce phenotypes identical to <i>par-1</i> mutants in both oocyte determination and A-P axis formation, indicating that 14-3-3 mediates PAR-1 function in vivo. This parallels the requirement for the <i>C. elegans</i> 14-3-3 protein, PAR-5, in A-P polarisation, and predicts that the binding of 14-3-3/PAR-5 to PAR-1 substrates is a conserved mechanism by which these kinases generate polarity. In computational and biochemical screens for <i>Drosophila </i>14-3-3 targets, I identified the PAR-3 homologue Bazooka (BAZ), and show that PAR-1 phosphorylates BAZ on its two 14-3-3-binding sites. In epithelial cells, BAZ and PAR-1 localise to complementary apical and lateral membrane domains. I show that PAR-1 and 14-3-3 are required to polarise the follicular epithelium, and to restrict BAZ apically.
|
|---|
