ARIP4 resolves R loops at promoter proximal regions of highly transcribed, androgen-dependent genes to promote PolII pause release and transcriptional induction

Abstract

Pausing of RNA Polymerase II (PolII) is a prevalent regulatory mechanism in transcription that primes genes with transcriptional elongation machinery, enabling rapid transcriptional output upon pause release signals. Pausing of PolII occurs downstream of the transcription start site (TSS), resulting in short, nascent RNA strands that can form R loops. Recently, it was speculated that these pausing-associated R loops are a major source of conflict between the replication and transcription machinery, which may lead to genomic instability. However, how the maintenance of pausing-associated R loops directly affects transcriptional induction is unclear. Furthermore, transcriptional induction can result in topoisomerase II beta (TOP2β)-mediated double-strand breaks (DSB), which have been suggested to be a consequence of pause release. Altogether, PolII pausing-related events encompassing R loop level, DSB formation, and transcriptional output have yet to be linked together as a regulatory axis in transcriptional induction. ARIP4 (androgen receptor-interacting protein 4) is an elusive protein known to be a coactivator of androgen receptors (AR), but how it is involved in regulating transcription remains obscure. In this study, we showed that ARIP4 mediates optimal induction of AR target gene expression in response to androgen. ChIP-seq analysis revealed that ARIP4 was mostly enriched at promoter proximal regions overlapping with PolII and preferentially occupied paused genes. Furthermore, ARIP4 was found to interact with TOP2β and was indispensable for the timely formation of transient DSBs upon androgen stimulation. Indeed, ARIP4 deficiency led to defective pause release of PolII at AR target genes. Given that it has putative helicase activity and a strong preference for the nucleosome-depleted promoter proximal region, we speculate that ARIP4 might have resolving activity for pausing-associated R loops. Interestingly, depletion of ARIP4 did not uniformly affect all pausing-associated R loops, while R loop levels in highly expressed, highly induced AR target genes were sensitive to ARIP4 loss. Finally, the in vitro unwinding assays validated that ARIP4 binds and preferentially unwinds R loops. Here, we propose a model in which androgen stimulation triggers PolII pause release and ARIP4-mediated unwinding of R loops formed at AR target genes with high transcriptional activity, which prevents excessive pausing-associated R loops that can impede subsequent PolII machineries, leading to TOP2β activation at the promoter proximal region and occasional DSB formation.