We want to understand eukaryotic transcriptional regulation at the molecular level. Problems in gene regulation underlie many human diseases. We study gene regulation in yeast because of the powerful genetic and molecular tools that are available. Importantly, the transcription regulatory machinery is conserved between yeast and vertebrates, and insights gained from studies in yeast are generally universal.

Changes in chromatin structure leading to gene activation. We have studied activation of the yeast HO gene, which is transiently expressed during the cell cycle in only one of the two cells following mitotic division. Chromatin immunoprecipitation (ChIP) experiments show that transcriptional coactivator complexes get recruited to one region of the HO promoter early in the cell cycle, and then migrate to a different promoter region. Chromatin disassembly occurs in waves both along the length of the promoter and during the cell cycle, and three different chromatin factors are required for disassembly of nucleosomes, each at different regions of the HO promoter. The SBF DNA-binding factor binds to the promoters of HO and the G1 cyclin genes, and recruits three distinct factors: the Rpd3(L) histone deacetylase that inhibits gene activation, the FACT chromatin reorganizing complex that stimulates transcription, and the cyclin dependent kinase (CDK) that activates these genes as cells pass the G1/S transition by expelling the Rpd3(L) inhibitor from the promoter. Current experiments are to understand how these chromatin changes are regulated and how they control gene activity.

Memory at the HO promoter. The Swi5 DNA-binding protein is the earliest factor that binds to the HO promoter, and it initiates changes in HO chromatin that propagate long after the unstable Swi5 factor is degraded. Importantly, one can experimentally extend the time between when Swi5 was last bound and when the gene is activated, and the promoter “remembers” the actions of the Swi5 activator. Thus the HO promoter has been described as having a “memory,” and we are currently defining memory this at the molecular level.

Promoter specificity of transcriptional factors. Specific gene expression is controlled by transcription factors binding to elements present in promoters and enhancers. Yeast has two transcription factors, Swi5 and Ace2, that show similar patterns of cell cycle regulation, that have nearly identical zinc finger DNA-binding domains, and recognize the same DNA sequences in vitro. Despite these similarities, Swi5 and Ace2 activate transcription of different genes in vivo. There are "Swi5-only" genes where Swi5 binds and activates in vivo, but Ace2 cannot bind in vivo. Since Ace2 binds to these sites in vitro, promoter specificity is determined by mechanisms that control factor binding. Experiments are in progress to determine how chromatin structure can prevent one factor from binding while allowing another protein, with the same DNA-binding domain, to bind. Both Swi5 and Ace2 bind to "Ace2-only" promoters in vivo, but only Ace2 is able to actually activate transcription of these genes; thus, Swi5 binds to these promoters but fails to activate. We have identified other proteins bound nearby at these promoters that function as selective “anti-activators,” blocking Swi5 from activating transcription while allowing Ace2 to activate. Current experiments are studying the mechanisms of these selective anti-activators. These studies are quite relevant to mammalian gene regulation, because in many cases it has been shown that multiple transcription factors recognize the same sequence, but that simple DNA-binding at a promoter is not sufficient for gene activation in vivo.

FACT in transcriptional elongation. We are studying the FACT chromatin reorganizing complex in collaboration with Tim Formosa’s laboratory. In addition to the functions at promoters (like HO) and DNA replication, FACT promotes transcriptional elongation. Current studies investigate how FACT functions to properly reassemble nucleosomes following passage of an elongating RNA polymerase.