Back

BRD4 recruitment desilences transcription without erasure or depletion of repressive chromatin

Brandon, C. J.; Robinson-Thiewes, S.; Kaulage, M.; Rosikiewicz, W.; Cuneo, M. J.; Brett, J.; Kandola, J.; Lang, W. H.; Low, J.; Mohammed, A.; Danda, A.; Rider, S.; Valentine, M.; Ochoada, J.; Young, B.; Nguyen, T.; Kietlinska, S. J.; Taylor, A. B.; Hoeckendorf, B.; Rodrigues, P.; Lin, W.; Khairy, k.; Xu, B.; Shelat, A.; Chen, T.; Mittag, T.; Ansari, A. Z.

2026-04-13 cell biology
10.64898/2026.04.10.717856 bioRxiv
Show abstract

How genes are desilenced within mesoscale repressive chromatin is a crucial yet poorly understood phenomenon. Prevailing models posit that methylation of lysine-9 of histone H3 (H3K9me3) engages heterochromatin protein 1 (HP1) to drive chromatin compaction and transcriptional silencing. The erasure of this repressive mark and its replacement with acetyl/acyl groups recruits positive factors such as BRD4/BET to elicit gene transcription. We report that in Friedreichs ataxia, a synthetic gene regulator (SynGR1) drives transcription across repressive chromatin without removal or replacement of H3K9me3 or HP1. By selectively recruiting BRD4/BET into repressive GAA-repeats in frataxin (FXN), SynGR1 creates a paradoxical state where gene transcription and repressive chromatin co-exist. Contrary to convention, we find that BRD4 readily partitions into phase separated HP1 condensates in vitro and into HP1 puncta in patient-derived cells, thus presenting a mechanistic explanation for desilencing transcription without the dispersal of mesoscale repressive chromatin. Epigenetic drugs that gate sequential steps in transcription, synergistically stimulate FXN expression while concomitantly increasing, rather than eliminating, repressive H3K9me3 and HP1 levels. More broadly, this study highlights the dynamic nature of repressive chromatin and the context-dependence of epigenetic marks in regulating gene expression.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.