BMC Genomics (May 2019)

Histone H3 lysine 4 methyltransferase is required for facultative heterochromatin at specific loci

  • Qiaoqiao Zhu,
  • Mukund Ramakrishnan,
  • Jinhee Park,
  • William J. Belden

DOI
https://doi.org/10.1186/s12864-019-5729-7
Journal volume & issue
Vol. 20, no. 1
pp. 1 – 19

Abstract

Read online

Abstract Background Histone H3 lysine 4 tri-methylation (H3K4me3) and histone H3 lysine 9 tri-methylation (H3K9me3) are widely perceived to be opposing and often mutually exclusive chromatin modifications. However, both are needed for certain light-activated genes in Neurospora crassa (Neurospora), including frequency (frq) and vivid (vvd). Except for these 2 loci, little is known about how H3K4me3 and H3K9me3 impact and contribute to light-regulated gene expression. Results In this report, we performed a multi-dimensional genomic analysis to understand the role of H3K4me3 and H3K9me3 using the Neurospora light response as the system. RNA-seq on strains lacking H3 lysine 4 methyltransferase (KMT2/SET-1) and histone H3 lysine 9 methyltransferase (KMT1/DIM-5) revealed some light-activated genes had altered expression, but the light response was largely intact. Comparing these 2 mutants to wild-type (WT), we found that roughly equal numbers of genes showed elevated and reduced expression in the dark and the light making the environmental stimulus somewhat ancillary to the genome-wide effects. ChIP-seq experiments revealed H3K4me3 and H3K9me3 had only minor changes in response to light in WT, but there were notable alterations in H3K4me3 in Δkmt1/Δdim-5 and H3K9me3 in Δkmt2/Δset-1 indicating crosstalk and redistribution between the modifications. Integrated analysis of the RNA-seq and ChIP-seq highlighted context-dependent roles for KMT2/SET1 and KMT1/DIM-5 as either co-activators or co-repressors with some overlap as co-regulators. At a small subset of loci, H3K4 methylation is required for H3K9me3-mediated facultative heterochromatin including, the central clock gene frequency (frq). Finally, we used sequential ChIP (re-ChIP) experiment to confirm Neurospora contains K4/K9 bivalent domains. Conclusions Collectively, these data indicate there are obfuscated regulatory roles for H3K4 methylation and H3K9 methylation depending on genome location with some minor overlap and co-dependency.

Keywords