Research

Our genomes are organized in a polymeric complex called chromatin with a repeating unit called the nucleosome. Long thought of as a repressive complex, it is now clear that the nucleosome serves as a vibrant signaling hub for genome-templated processes. We use structural biology, protein chemistry, and proteomics to better understand epigenetic chromatin signaling in health and disease.

Establishing paradigms of nucleosome recognition

Current understanding of the patterns for nucleosome recognition by chromatin enzymes has been slowly and anecdotally built through structural characterization of individual chromatin enzyme-nucleosome complexes. Using these structures, we have proposed an emerging paradigm of nucleosome recognition in which arginines from the chromatin enzymes (that we have named arginine anchors) bind to an acidic patch on the nucleosome surface. However, the pervasiveness of the arginine anchor-nucleosome acidic patch interaction and existence of other paradigms for nucleosome recognition were unclear. We have used proteomics to show that the majority (hundreds) of nucleosome interacting proteins required the acidic patch for chromatin binding and mapped these interactions at amino acid resolution. We are now studying how acidic patch interactions are regulated by epigenetic signaling pathways.

Workflow for nucleosome interactome proteomics.

Determining the structural basis of chromatin signalling

A large part of our research program centers on using cryo-electron microscopy (cryo-EM) to understand nucleosome recognition by chromatin enzymes at near atomic resolution. A primary goal is to define mechanisms governing specificity, multivalency, and crosstalk in chromatin signaling. Through understanding how these enzymes function, we gain insight into their roles in genome-templated processes and their misregulation in disease, especially cancer. We are uniquely poised to study chromatin signaling due to our ability to build site-specifically modified nucleosomes using protein chemistry tools like expressed protein ligation. For example, we solved a cryo-EM structure of the histone H3 lysine methyltransferase DOT1L bound to a site-specifically, chemically ubiquitylated nucleosome. DOT1L requires prior ubiquitylation of histone H2B for optimal activity. Misregulation of this histone crosstalk pathway is pathogenic in a subset of leukemias and DOT1L inhibitors are now in clinical trials. Our structure shows how Dot1L binds its nucleosome substrate in a poised state and is activated by ubiquitin. More recently, we have turned our attention to enyzmes that remove lysine methylation. We developed nucleosome-demethylase inhibitor conjugates to stabilize KDM2 and KDM6 demethylases on the nucleosome for cryo-EM. Our structures suggest that these demethylases have evolved to function in unique chromatin contexts. Finally, we are exploring the role of chromatin in regulating cGAS, an innate immune sensor of cytosolic DNA that has enigmatic nuclear functions.

Cryo-EM structures of methyltransferase DOT1L (left), demethylase KDM2A (middle), and cGAS (right) bound to nucleosomes.

Creating tools to enable epigenetics research

We also are developing new tools to probe epigenetic signaling both in vitro and in model systems. Such tools will allow precise hypothesis-driven interrogation of epigenetic signaling networks. A recent example is a Time-Resolved FRET assay for high-throughput quantitation of nucleosome interaction.

Time-Resolved FRET assay for quantitation of nucleosome interactions.

Our funding sources