Structures and Mechanisms of ATP-dependent Chromatin Remodeling Machines

The genome is functionally compartmentalized into regions that undergo rapid transcriptional regulation and regions that are transcriptionally silent through several cell generations. These two types of functional regions have different underlying chromatin structures. The regions that are rapidly regulated have more accessible chromatin structures and contain nucleosomes that exchange their histone content rapidly. In contrast, the regions that are heritably silent are more condensed and have less accessible chromatin structures. Two different classes of molecular machines catalyze the conformational changes that define each type of chromatin domain. The SWI/SNF complex generates a distribution of nucleosomal states, which contain nucleosomes with altered positions, altered composition and altered DNA paths. These enable efficient exposure of short regions of DNA for localized binding of activators or repressors. The ACF motor only generates chromatin with regularly spaced nucleosomes, which enables long-range folding of chromatin into silent chromatin. The different products are unexpected because the two machines have highly homologous motor domains. To understand how these two different complexes alter chromatin structure and why they generate such different products, we are comparing and contrasting their mechanisms on a variety of model chromatin systems using a combination of different biophysical approaches.


Molecular Mechanisms of Heterochromatin Spread

The spread of heterochromatin is crucial for heritably silencing large regions of the genome and consequently for generating and maintaining cell identity during development. A hallmark of heterochromatin is its ability to spread to adjacent regions and cause gene silencing. Aberrant heterochromatin spread is thought to promote cancer by permanently repressing genes required for normal differentiation. A major type of heterochromatin formation requires methylation of histone H3 on lysine 9 (H3K9me3), which is catalyzed in humans by Suvar3-9 and is recognized by the chromodomain (CD) of HP1. We are studying how the highly homologous core components in fission yeast, the methylase, Clr4 and the HP1 protein, Swi6 act together to spread heterochromatin. Many additional factors, including RNAi–based complexes have been identified as being key for nucleating heterochromatin. Yet the fundamental mechanisms by which Clr4 and Swi6 spread heterochromatin outward from a nucleation center are still unknown.

There are several key questions:

(i) Dimerization of Swi6 through its chromoshadow domain (CSD) is important for Swi6 spread yet it is not clear how dimerization alone can lead to long-range spreading of Swi6;

(ii) the spread of Swi6 is hypothesized to condense underlying chromatin yet the molecular basis for such condensation is not known;

(iii) closely spaced nucleosomes are strongly correlated with stable heterochromatin, but the underlying mechanism is not known;

(iv) Clr4 and Swi6 cooperate in vivo, yet beyond the methylase activity of Clr4, additional levels of cooperation have not been demonstrated. We are addressing these questions using a combination of biochemical, structural and in vivo approaches.