Histone Recycling by FACT and Spt6 during Transcription Prevents the Scrambling of Histone Modifications [tiling]

The genetic information encoded in DNA is framed by additional layers of information, referred to as the epigenome. Epigenetic marks such as DNA methylation, histone modifications and histone variants are concentrated on specific genomic sites as means to instruct, but also sometimes as a consequence of, gene expression. How this information is maintained, notably in the face of transcription, is not understood. Here we show that the histone chaperones FACT and Spt6 are required for maintaining proper localization of several histone modifications including H3K4me1,2,3, H3K36me3, H3K79me3, H3K14ac, H3K18ac and H2Bub in Saccharomyces cerevisiae. In the absence of functional FACT or Spt6, transcription generates massive nucleosome loss which is partially compensated by increased histone assembly by histone chaperones such as Asf1 and HIR. Because re-incorporation of histones by these histone chaperones in not coupled to transcription, the modified histones are randomly incorporated, leading to scrambling of the epigenetic information. Hence, our work highlights the importance of local nucleosome recycling by FACT and Spt6 during transcription in the maintenance of a proper epigenetic landscape. We profiled eight histone modifications in temperature-sensitive mutants for the histone chaperones FACT and Spt6 (spt16-197 and spt6-1004 respectively) in yeast Saccharomyces cerevisiae by ChIP-chip on tiling arrays. These histone modifications included H3K4me1, H3K4me2, H3K4me3, H3K36me3, H3K79me3, H3K14ac, H3K18ac and H2Bub. Abrogating the function of FACT or Spt6 leads to redistribution of the modified histones. The contribution of FACT and Spt6 on H3K4me3 mislocalization was also confirmed by nuclear depletion of Spt16 and Spt6 using the anchor-away system. In addition, we profiled histones (H4, H3, H2B) and RNAPII occupancy in wild type, spt16-197 and spt6-1004 cells. RNAPII occupancy was also measured in a temperature-sensitive mutant for RNAPII (rpb1-1). In order to investigate whether nucleosome loss may lead to mislocalization of histone modifications, we profiled H4, RNAPII, FACT and histone modifications in cells where H4 expression was shutdown using a system developed by the Grunstein lab (Han and Grunstein, 1988). To test if the chromatin defects observed in the FACT and Spt6 mutants are caused by transcription, we inhibited transcription (or not) by treating cells with thiolutin, concomitantly to anchoring-away FACT or Spt6, and performed H3K4me3, H3K36me3, H4 and RNAPII ChIP-chip experiments. In addition, H3K4me3 ChIP-chip experiments were performed in htz1Δ, spt16-197/htz1Δ and spt6-1004/htz1Δ cells in order to rule out that the defects in histone modifications may occur because of the histone variant H2A.Z mislocalization in FACT and Spt6 mutants. H3K4me3 occupancy was also measured in ATP-dependent chromatin remodeler mutants including chd1Δ, spt16-197/chd1Δ, isw1Δ, spt16-197/isw1Δ, isw2Δ, spt16-197/isw2Δ. Finally, the localization of the histone chaperones Asf1 and HIR (FLAG –tagged) was determined in wild type and in spt16-197 cells. All ChIP samples were labeled with Cy5 and hybridized against appropriate samples that range from Input DNA, H2B, H3, H4 or NoTag ChIP samples (labeled with Cy3). The microarrays used were described before (Jeronimo and Robert, 2014). The data were normalized by setting the median of log2 ratios to zero. Each experiment was done in duplicate.