Bimodal quantitative relationships between histone modifications for X-linked and autosomal loci.

Journal Article (Journal Article)

Gene expression is controlled by coordinated action of many epigenetic mechanisms including covalent histone modifications. Although numerous recurrent patterns of colocalized histone modifications have been associated with specific gene expression states, interrelationships between individual modifications are largely unknown. Here, we analyze quantitative relationships between colocalized histone marks during embryonic stem cell (ESC) differentiation and find that, for autosomal genes, these densities follow bimodal patterns. Analysis of repressive H3K27me3 and activating H3K4me3 modifications reveals the expected anticorrelation between them at active promoters but an unexpected positive correlation at inactive promoters. The two trends connect in a region corresponding to bivalent genes. Interestingly, this region is characterized by maximal H3K27 methylation. Resolving gene bivalency during ESC differentiation does not conform to the expected model of two marks as counteracting and competing forces. Although activated genes acquire H3K4me3 and lose H3K27me3, repressed genes lose H3K4me3 without gaining H3K27me3. The behavior of X-linked genes also deviates from expected models. Allele-specific analysis of chromatin modifications during X-chromosome inactivation (XCI) suggests that the silencing machinery focuses on active genes and depletion of H3K4me3 and that H3K27me3 is most significant during establishment of gene silencing. Our analysis reveals nontrivial relationships between H3K4me3 and H3K27me3, reveals unique aspects of gene bivalency, and demonstrates that XCI does not conform neatly to autosomal models.

Full Text

Duke Authors

Cited Authors

  • Sadreyev, RI; Yildirim, E; Pinter, SF; Lee, JT

Published Date

  • April 23, 2013

Published In

Volume / Issue

  • 110 / 17

Start / End Page

  • 6949 - 6954

PubMed ID

  • 23564346

Pubmed Central ID

  • PMC3637770

Electronic International Standard Serial Number (EISSN)

  • 1091-6490

Digital Object Identifier (DOI)

  • 10.1073/pnas.1216449110


  • eng

Conference Location

  • United States