UC Santa Cruz

DNA in eukaryotic organisms is complexed with histone and non-histone proteins to form chromatin. Gene activities are determined by chromatin states that are either permissive or restrictive to transcription. In the budding yeast Saccharomyces cerevisiae, chromatin states at the cryptic mating loci HML and HMR are mediated by silencer elements and the Sir proteins (Sir1, Sir2, Sir3, and Sir4). The Sir proteins bind unacetylated tails of the histones in chromatin and silence transcriptional activity by sterically hindering transcription factor access to chromatin. The silent state, once established, is transmitted stably through multiple cell divisions, but the factors contributing to the silent state are in constant flux, suggesting that the system can tolerate fluctuations in levels of these factors without functional consequences.

In the first part of this work, we sought to determine the levels of the individual factors necessary for silencing. We developed methods to measure the thresholds of individual proteins and modifications that lead to loss of the silent state. We show that silencing loss is not observed until more than half of the unacetylated histone complement at a silent locus is acetylated. For Sir proteins, our data suggest that silencing can tolerate significant reductions in the levels of Sir2 and Sir3 but cannot tolerate even a twofold reduction in Sir4 levels.

Transcriptional silencing functionally operates on enhancers and promoters of genes. In the second part of this work, we tested the susceptibility of various enhancer and core promoter elements to silencing at HMR using a panel of different enhancers and promoters of differing transcriptional strengths. Our data suggest that silencing of these elements occurs in a probabilistic manner dependent on properties of individual enhancers and core promoters. We find that constitutively active enhancers/promoters escape silencing to a large degree and the silencing machinery is only able to suppress transcription from moderately weak enhancers. However, our data also show that increased Sir1 binding to silencers can counteract the effect of strong enhancers/promoters. Using fluorescent reporters of nascent transcription, we show that the Sir proteins function in silencing by modulating the transcription frequency of enhancers. Altogether, this work suggests a model in which the Sir proteins function by reducing the probability that transcription factors bind to target sites at the enhancers of silenced domains, thereby affecting transcription frequency.