The Universality of Nucleosome Positioning: From Yeast to Human

Abstract

Nucleosome organization is important for the understanding of promoter accessibility and transcription regulation. In the past decade, in vivo nucleosome mapping suggested a stereotypical and conserved nucleosome organization, consisting of nucleosome-depleted regions (NDRs) around the transcription start sites (TSSs), flanked by regular arrays of nucleosomes on gene bodies. We study the factors that determine the nucleosome positioning, and their universality across different organisms, and we develop biophysical models that incorporate the common rules of nucleosome organization.We show that in vitro reconstituted nucleosomes occupy different locations from those observed in vivo. Thus, the DNA sequence has a limited role in nucleosome positioning. Using ChIP-seq and MNase-seq techniques, we show that non-histone proteins bind to gene promoters and compete with the histones, creating NDRs. We go on to apply rigorous biophysical models to explain the in vivo nucleosome phasing relative to the TSS, incorporating the ATP-dependent chromatin remodelers.In collaboration with other labs, we compare chromatin organization in yeast, fly, mouse and human. We find that most of yeast genes have a nucleosome depleted and DNase I accessible promoter. However, surprisingly, using MNase-seq, DNase-seq and RNA-seq data, we find that in higher organisms genes fall into two distinct, tissue-specific classes: the promoters of inactive genes are tightly packed with nucleosomes, not phased relative to the TSS and stabilized by H1 linker histones. By contrast, the active genes have DNase I accessible promoters and phased arrays of nucleosomes along the gene bodies. Using data available from the ENCODE project, we verify that most of the characterized epigenetic marks also have this bimodal distribution, correlated with the nucleosome organization of the gene promoters.