The goal is to understand how the chromatin proteins, including the histones and regulatory proteins, regulate the biochemical behavior of DNA within the nucleus, including the processes of cellular differentiation and embryonic development.
Our group studies the relationship between chromatin structure and gene expression in eukaryotes. We focus on epigenetic mechanisms and the structure of both the individual nucleosomes (the fundamental chromatin subunits) and the folded polynucleosome fiber. We are particularly interested in the mechanisms that establish boundaries between large-scale chromatin domains within the nucleus. Some boundary elements prevent the spread of condensed chromatin into transcriptionally active regions. Other kinds of elements serve to create ‘loop’ domains that either bring together distant DNA sequence elements or segregate them into separate loops. These play an important role in regulation of gene expression. We are studying both kinds of boundary elements. In each case, we have identified proteins that bind to boundary sites, as well as the co-factors recruited by those proteins, and obtained important information about how they function. We have recently investigated the action of one such protein, VEZF1/BGP1. We have shown that this protein blocks the spread of DNA methylation, a silencing signal, and prevents the spread of an adjacent heterochromatic domain into a transcriptionally active region of the genome. We have shown an important property of this protein when bound to its site on DNA: it slow the passage of RNA polymerase II. One effect of this action is to alter the abundance of different RNA splice variants in the affected genes, an important regulatory function. We believe that the ability to impede the progress of the polymerase also serves as a boundary at other sites, preventing the encroachment of an adjacent heterochromatic domain. Other recent work in our laboratory focuses on the heterochromatic domains themselves. We initially studied the physical properties of such a domain upstream of the chicken beta globin locus. Recently, we reported that this domain is maintained in the condensed state by a mechanism involving the protein Argonaute 2. In many ways, it resembles the silencing mechanism used in fission yeast. We are now extending this analysis to other kinds of heterochromatic sequences in human cells.
Another important focus of our research concerns long-range chromatin organization and the boundaries between independently regulated domains. Some years ago, we demonstrated that the protein CTCF is responsible for establishing an ‘insulator’ activity to block selectively interactions between enhancers and promoters. We also showed that CTCF plays such a role at the imprinted Igf2/H19 locus in vertebrates. We suggested that CTCF insulating activity might reflect an ability to establish loop domains within the nucleus. More recent work in many laboratories shows that this is the case. Generally, CTCF helps organize the genome into such domains. One aspect of our recent studies has focused on identifying the co-factors that work with CTCF in domain organization. We are particularly interested in the protein p68 and its associated long non-coding RNA, SRA; both are required for some CTCF functions, and we are trying to understand more generally how p68/SRA functions genome-wide.
Our other main focus is the examination of long-range interactions within the nucleus in human pancreatic beta cells, between the insulin gene and other genes that may be coregulated. This is part of a larger effort to study chromatin structure of the insulin locus and its relationship to insulin gene expression and secretion. We have recently shown that the human insulin gene promoter makes physical contact in pancreatic islet cells with many distant sites on the same chromosome. Some of these contacts are to genes that are important in insulin secretion, and their expression is stimulated by the contacts, which increase when glucose is added to the islets. This is a complex physical regulatory network that may link insulin expression to secretion. We are extending this study to obtain a more complete picture of the patterns of insulin gene interactions with the entire genome.