Nucleosomes, transcription and probability

14 Feb 2014

Hinrich Boeger
Department of Molecular, Cell and Developmental Biology
UC Santa Cruz


I shall describe theoretical efforts to understand recent experimental observations on the chromatin structure of single gene molecules, the long path of a molecular biologist toward probabilistic theories and quantitative model building.

Electron microscopic analysis of the PHO5 gene of budding yeast indicated that its promoter assumes alternative nucleosome configurations under transcriptionally inducing conditions, including the fully nucleosomal configuration, which prevails under non-inducing conditions, and the nucleosome-free configuration. What is the origin of this structural variation?

Surprisingly, the observed configurational frequencies are obtained by a random walk on a directed graph with nodes corresponding to nucleosome configurations, and edges indicating possible transitions between them. As each node is connected to more than two neighbors, structural variation results from the randomness of choice between alternative nucleosome configuration. This result suggests that the promoter transitions between alternative nucleosome configurations in a stochastic manner; as configurations are either transcriptionally conducive or inconducive, transcription is expected, therefore, to occur in randomly spaced "bursts". Analysis of gene expression at the single cell level corroborates this conjecture. Gene expression and its regulation may be understood only in probabilistic terms. I shall touch on theoretical challenges, and perhaps some epistemological problems (which might not be novel to physicists).

current theory lunch schedule