Evolution of silencers and silenced DNA in budding yeasts.

摘要

Traditionally, heterochromatin has been viewed as transcriptionally-silent DNA with no important sequence information. This stereotype has recently been challenged in a number of eukaryotes, and my work contributes to the new and exciting perspective on heterochromatin. I observed an unexpected level of divergence in heterochromatic sequences of closely-related budding yeasts of the sensu stricto complex. Looking within species, I also detected an increase in polymorphisms in these sequences. Critically, by analyzing SNPs in degenerate codon positions inside genes, I showed that the hyperdivergence was due to a faster mutation rate in silenced DNA, rather than due to a relaxed selective constraint. The implications of my discoveries may be particularly significant for researchers studying host-immune evasion by eukaryotic pathogenes, as most of the antigenic variation genes of such pathogens are located in heterochromatin.;I characterized the heterochromatic evolution because it was important for my interest in the DNA elements, called silencers. I have been fascinated by the ability of DNA-binding proteins that activate gene expression in euchromatin to act in the opposite direction when bound at these silencers. How does the cell avoid spurious silencing throughout the genome, and how does it ensure nucleation of heterochromatin at the silencers? My work on hyperdivergence underscored a surprising conservation of a protein binding site in one of these silencers. An extension of this analysis suggested that the specific sequence of the binding site at the silencer may be important for directing the protein to recruit the silencing complex, instead of activating transcription. This contrasts with the accepted model of protein interaction with binding sites where only the overall affinity to the site is important for the functional output. Subsequently, I analyzed the co-occurrence of silencer-binding proteins across the genome and found a bias against juxtaposition of these binding sites outside of the silencers, implying a possible selection against spurious silencing in euchromatin.;While I was exploring the impact of heterochromatin on the evolution of the underlying DNA and on genome organization, my classmate and colleague Bilge Ozaydin discovered that silencing at the yeast mating cassette can affect the shearing of DNA in molecular experiments. This result was unanticipated because shearing has been assumed to be equally efficient in heterochromatin and euchromatin. Using high-throughput datasets, I extended Bilge's findings and showed that heterochromatin shears poorly throughout the genome, not just at the silenced mating cassettes. I also observed that in euchromatic regions, the chromatin state can enhance or interfere with shearing, independently of silencing. This is an important conclusion because it has relevance for the exploding use of the ChIP-seq technique. My results highlighted the importance of proper controls for ChIP-seq experiments. Moreover, my analysis revealed that the ChIP-seq controls are informative for detecting difference in chromatin states across the genome, and this is likely to be a powerful tool for many eukaryotes.