(598b) Physical Modeling of the Spreading of Epigenetic Modifications through DNA Looping

Epigenetics refers to variations in gene expression that are caused not by changes in the sequence of DNA, but by chemical modifications in nucleosomes. These epigenetic marks often spread along the chromatin fiber through enzymatic processes that require genomically distant DNA segments to come into close spatial proximity. Such DNA looping is a common phenomenon in many critical biological processes within the nucleus, such as recombination events and gene regulation. Gaining knowledge of the dynamics of loop formation and the kinetics of mark transfer is therefore an essential step towards a better understanding of epigenetic processes.

Loops in DNA occur at a range of length scales, from short DNA lengths on the order of 100 basepairs for the binding of regulatory proteins, to intermediate lengths of kilobases for chromosome folding, to long lengths of hundreds of kilobases for genetic recombination. It is important to recognize that different physics govern looping at different length scales. At short lengths, the rigidity of the DNA must be handled explicitly, whereas at long lengths, the overall chain dynamics dominate. Because of the range of length and time scales involved in the motion of DNA, gaining insight into looping dynamics has proved a particularly complicated pursuit. Previous theoretical studies have addressed the problem through physical modeling and simulation of DNA as a Rouse or semiflexible polymer and have obtained qualitative agreement with experimental results.

To build upon this research, we are working to develop a comprehensive kinetic model for the spreading of epigenetic modifications through DNA looping. We use both analytical theory and computer simulations to calculate the probabilities that various lengths of DNA will loop, and in both cases, we account for the dynamics of DNA looping over the entire range of relevant length scales. We consider the spreading of methyl marks along chromosomal DNA, as dictated by these looping probabilities and rates of methylation and demethylation. Given an initial methylation profile, we determine the steady-state probabilities that other sites will become marked. We examine the effects of the type of transfer (i.e. local vs. at-a-distance), the initial profile (i.e. presence or absence of a nucleation site), the overall topology of the DNA (i.e. linear vs. ring), and the DNA configuration (i.e. static vs. dynamic).