Regulation of the Centromeric Histone H3 Variant Cse4 by the E3 Ubiquitin Ligase, Pshl

Aneuploidy, one of the most common hallmarks of cancer, results from defects in chromosome segregation during cell division. Once cells replicate their genome, sister chromatids must be accurately segregated into daughter cells during mitosis in order to ensure genomic stability. Chromosome segregation IS directed by the kinetochore, a multi-protein complex that assembles onto specialized chromosomal locus called the centromere. Because centromere DNA is not conserved, there are epigenetic mechanisms that ensure its propagation. Cenp-A, a conserved histone H3 variant appears to be the epigenetic mark that specifies centromere identity and is essential for kinetochore assembly and function. Exclusive localization of Cenp-A to the centromere is critical to prevent formation of di-centric chromosome that can be broken during segregation. Proteolysis is one of the mechanisms that cells utilize to regulate Cenp-A levels and prevent its mislocalization to non-centromeric loci in budding yeast and flies. However, the molecular machinery and the underlying mechanisms are not known. In my research, I focused on identifying proteins involved in degradation of Cenp-A and elucidate the meehanism by which the proteolytic machinery prevents eetopic localization of Cenp-A using budding yeast as a model system. Cse4 is the Cenp-A homolog in budding yeast. To identify Cse4 degradation machinery, I focused on known and novel Cse4 interacting proteins that were previously identified in the lab. Of the various candidates, I focused on Pshl, a protein of unknown function. Pshl contains a R1NG domain, a hallmark of all ruNG type E3 ligases. I demonstrated that Pshl is a bonafide E3 ligase in vitro and functions to ubiquitinate and degrade Cse4 in vivo. Moreover, Pshl mediated degradation of Cse4 prevents mislocalization of Cse4 to non-centromeric loci. Consistent with this, overexpression of Cse4 is toxic to pshl 11 cells and leads to mislocaliztion of Cse4 to the euchromatin. The toxicity of Cse4 overexpression is correlated with the levels of Cse4 protein and can be partly attributed to the activation of the spindle checkpoint. Since Cse4 is about 60% identical to the canonical H3, I also investigated if Pshl is a specific E3 ligase for Cse4 and how it is able to discriminate between the two histones. Pshl co-purifies with Cse4 but fails to interact with histone H3. Consistent with a specific function of Pshl for Cse4 proteolysis, overexpression of H3 had no effect on the viability of pshl 11 cells. Domain swap experiments reveal the centromere targeting domain (CATD) in Cse4, which is absent in H3, to be the recognition motif by which Pshl distinguishes Cse4 from H3. In addition, this domain appears to be required and sufficient for Pshl mediated degradation. Taken together, my work has shown that the CA TD has a previously unknown role in maintaining the exclusive localization of Cse4 by preventing its misloealization to euchromatin via Pshl-mediated degradation