RNA polymerase II coordinates histone deacetylation at active promoters

Abstract

Nucleosomes at promoters of active genes are marked by specific histone post-translational modifications and histone variants. These features are thought to promote the formation and maintenance of an "open" chromatin environment that is suitable for transcription. However, recent reports have drawn conflicting conclusions about whether these histone modifications depend on active transcription. To further interrogate this relationship, we inhibited transcription initiation using triptolide, which triggered degradation of RNA polymerase II, and examined the impact on histone modifications. Transcription initiation was not required for either hormone-induced or steady-state active histone modifications at transcription start sites (TSSs) and enhancers. Rather, blocking transcription initiation increased the levels of histone acetylation and H2AZ incorporation at active TSSs. P300 activity was dispensable for this effect, but inhibition of histone deacetylases masked the increased acetylation. Together, our results demonstrate that active histone modifications occur independently of transcription. Furthermore, our findings suggest that the process of transcription coordinates the removal of these modifications to limit gene activity.

Fig. 1.. Transcription is dispensable for steady-state and de novo H3K27 acetylation.

(A) Differential heatmap of K27ac ChIP-seq signal (log2 Dex/Veh) over class III GR peaks ± 2 hours triptolide. (B) Browser image of K27ac ChIP-seq and RNAP2 ChIP-seq over ZBTB16 gene. (C) Browser image of K27ac ChIP-seq and RNAP2 ChIP-seq over GLUL gene. (D) Western blot of whole-cell lysates from four biological replicates each of A1-2 cells ± 2 hours triptolide. (E) Meta-profiles of representative ChIP-seq and Cut&Tag replicates over Refseq TSSs. (F) Meta-profiles of representative RNAP2 ChIP-seq ± 2 hours triptolide over Start-seq defined active, inactive, and enhancer TSSs. (G) Meta-profiles of representative K27ac Cut&Tag ±2 hours triptolide over Start-seq defined active, inactive, and enhancer TSSs

Fig. 2.. Transcription initiation suppresses acetylation and H2AZ incorporation at active TSSs.

(A to J) Meta-profiles of representative Cut&Tag replicates ±2 hours triptolide over active TSSs. (K to M) Heatmaps depicting relative change [(triptolide − control)/max(control)] in Cut&Tag signal over Refseq TSSs in indicated cell lines.

Fig. 3.. RNAP2 degradation reproduces the effect of inhibiting transcription inhibition.

(A to D) Meta-profiles of ChIP-seq and Cut&Tag signal ±2 hours flavopiridol over active TSSs. (E) Western blots of whole-cell lysate from HCT116-mAID-POLR2A cells ± auxin or triptolide. (F to H) Heatmaps depicting relative change [(auxin − control)/max(control)] in Cut&Tag signal for indicated marks.

Fig. 4.. HDAC inhibition masks the effect of inhibiting transcription initiation.

(A) Western blots of whole-cell lysates from A1-2 cells treated ± triptolide, A485, and TSA. (B to D) Meta-profiles of Cut&Tag signal in A1-2 cells treated ± triptolide, A485, and TSA.