UC Berkeley

Heterochromatin is key to the appropriate regulation of cell identity. Saccharomyces cerevisiae silent mating loci, HML and HMR, provide a simple, yet elegant model for the regional control of transcriptional repression and its contribution to cell identity. Yeast can be of mating type a or &alpha as is dependent on the alleles at the actively transcribed MAT locus. However, additional copies of the &alpha and a genes are present elsewhere in the genome to facilitate mating-type switching. Interestingly, even though HML and HMR have the exact same promoters as when the genes are at the MAT locus, the genes are not expressed. This promoter-independent repression is due to the Silence Information Regulator (Sir) proteins. Current models of yeast silencing posit that Sir proteins are recruited by transcription factors to HML and HMR. Sir2, whose catalytic activity is necessary for silencing and is the founding member of the sirtuin enzymatic class, deacetylates the neighboring nucleosome creating a high-affinity binding site for the Sir complex. Iterative rounds of this deacetylation results in spreading of Sir proteins across the silenced locus, which is key to silencing. Through a similar mechanism, the Sir-protein complex is also important for silencing telomeres.

This nucleation and spreading model for silencing makes testable prediction of silencing. For example, if iterative rounds of deacetylation by Sir2 are necessary for silencing, then a Sir complex would occupy each nucleosome at silenced loci. Additionally, this model only attributes Sir-protein association to HML, HMR and telomeres and not elsewhere in the genome. Lastly, if Sir2's only known role is to deacetylate nucleosome tails then its catalytic activity should be bypassed by preemptively removing these histone acetylations.

To test these predictions of the silencing model I have definitively characterized the distribution of Sir proteins throughout the genome and explored the unaccounted role for Sir2's catalytic activity in silencing. By characterizing the distributions of Sir2, Sir3 and Sir4 at high-resolution I have revealed unappreciated nucleation sites for Sir proteins at HML and HMR and a possible higher-order structure critical to silencing. Characterization of Sir proteins across the genome revealed an unanticipated phenomenon of highly-expressed genes being vulnerable to non-specific ChIP enrichment. Lastly, I performed a forward genetic screen to restore silencing independent of Sir2 catalytic activity. This screen identified a non-sirtuin deacetylase as refractory to silencing and suggested the possible role of a non-histone deacetylase substrate as important for silencing. This dissertation work has refined the Saccharomyces cerevisiae silencing model, contributing to our understanding of regional repression across organisms.