Computational and experimental insights into the structure and dynamics of heterochromatin

Abstract
Chemical, molecular and structural transformations of chromatin are intimately involved in critical cellular
phenomena, including differentiation, signaling, and pathogenesis. A detailed knowledge of how molecular
complexes involving multiple kilobases of DNA and hundreds of proteins respond to the finest changes in
chemical structure is key to elucidating the role of chromatin transformations in life and disease. The overarching
goal of this project is to develop and apply computational tools to investigate how the structure and dynamics of
chromatin determine its functional states. Our central hypothesis is that physical properties and behavior of the
chromatin fiber and associated proteins lend themselves to encoding into efficient and useful ultra-coarse-
grained (UCG) representations. Our strategy to reach the goal is by bridging together several computational and
experimental methodologies. We initiated the development of Molecular Biosystems (MB), a computational
platform for UCG simulations specifically adapted to the chromatin biology. MB methodology represents a blend
of physics-based mechanisms, such as dynamics of the chromatin fiber, with stochastic processes encompassing
protein-protein interactions and enzymatic reactions. MB studies will be complemented by all-atom MD and CG
simulations and experimentally tested using a unique chromatin in vivo assay (CiA) methodology.
Specifically, we will investigate the chromatin-mediated repression of Oct4, a key gene regulating embryonic
stem (ES) cell pluripotency at defined points in mammalian development. This is important because the ability
to reverse the Oct4 repression would streamline production of induced pluripotent cells (iPSC) and advance
regenerative medicine. The CiA technology at the Oct4 locus in mouse ES cells will be used for the exploration of
changes to chromatin structure, as well as for testing the adequacy of MB simulations. Experimental endpoints
that are directly comparable to computational hypotheses will be produced: (1) fraction of Oct4-repressed cells
in cell culture; (2) H3K9 methylation patterns on Oct4 promoter; and (3) chromatin conformation capture.
Three main components of our research are: (i) Extending and enhancing the UCG MB approach; (ii) Multi-
scale simulations of chromatin processes to elucidate the structure and dynamics of heterochromatin of Oct4
regulatory elements; (iii) Experimental real-time monitoring of heterochromatin molecular signatures using
Chromatin in vivo Assay (CiA) to study mechanisms and time course of Oct4 de-repression and provide feedback
for the computational models.
This work is important because of its focus on the physics of the gene repression, whose understanding will
bring us one step forward toward the promise of regenerative medicine and new prospects for cancer therapy.