PDF, Englisch Download (35MB) | Nutzungsbedingungen

Abstract

Numerous mutations in histone H3 lysine 4 (H3K4) and H3 lysine 36 (H3K36) modifying enzymes have been reported in human disease, yet the role of the H3K4 and H3K36 residues in mammals remain unclear due to the clustered arrays of many histone genes. Replication-dependent canonical H3 (H3.1/H3.2) exists as multiple gene copies and supplies nucleosomes for packaging of newly synthesized DNA during replication. The histone variant H3.3 differs from canonical H3 by only 4 to 5 amino acids, which allow nucleosome assembly independent of DNA replication throughout the cell cycle and in post-mitotic cells. In this study, I set out to investigate the role of the K4 and K36 residues in the histone H3.3 variant, which is enriched at active regions of the mammalian genome and encoded by two isolated genes; therefore amenable to functional analysis. Using CRISPR-Cas9, I mutated the K4 or K36 residue of endogenous H3.3 to unmodifable alanine (A) in mouse embryonic stem cells (ESCs) and revealed that the K4A mutation, but not K36A, resulted in widespread gene expression changes and impairment of neuronal differentiation into glutamatergic neurons. Furthermore, K4A resulted in significant H3.3 protein depletion at transcription start sites and active enhancers of ESCs - without effects at other sites. Genomic regions depleted of H3.3K4A showed concerted alterations of histone modifications (decreased K27 acetylation and increased K4 methylation) regardless of gene expression changes. In differentiated neurons, the K4A mutation impacted protein stability and resulted in widespread proteasomal degradation of the mutant histone. Thus, H3.3K4 is required for site-specific nucleosome maintenance at regulatory regions, histone stability and cellular differentiation of ESCs. H3.3K36 is not required for H3.3 deposition and turnover inside coding regions, and the K36A mutation affected gene expression at later stages of neurodevelopment. Furthermore, the K36A mutation globally depleted H3K36 di-metylation levels in ESCs, which resulted in a spread of the repressive mark H3K27me3, suggesting that H3K36 di-methylation is required to restrict the activity of PRC2. This study demonstrates a direct link between a specific histone residue (H3K4) and histone maintenance at promoters and enhancers, and that H3.3 provides a platform for analyzing the role of histone residues in mammals.