ZC3H4 restricts non-coding transcription in human cells

The human genome encodes many thousands of non-coding RNAs. A majority of these are curtailed by early transcriptional termination followed by rapid degradation. The mechanisms and factors responsible for restricting non-coding transcription in this fashion are poorly characterised, but these RNAs vastly outnumbering those with coding potential. We identify ZC3H4 as a metazoan-specific factor controlling the level and extent of hundreds of unstable transcripts. The most striking effects are for antisense RNAs and those transcribed from super-enhancers, but a subset of protein-coding RNAs are also affected. ZC3H4 occupies these genomic loci and its level of recruitment correlates with RNA extension observed in its absence. Finally, engineered targeting of ZC3H4 to an otherwise unaffected RNA supresses its production and leads to rapid degradation, endowing it with the features associated with its natural targets. We conclude that ZC3H4 is a hitherto uncharacterised factor that restricts unproductive transcription. Overall design: Nuclear selected RNA was isolated from auxin-inducible degron (AID) tagged HCT116 cell lines treated with or without auxin, or HCT116 wild type treated with siTARGET or siCONT and sequenced using illumina total RNA sequencing. RNA Polymerase II and ZC3H4 associated DNA fragments were also analysed in the form of a ChIP-Seq experiment. We integrated an in-frame insertion of AID tag, P2A cleavage site, antibiotic resistance marker and SV40 polyA site to the 3' end of CPSF30. CPSF30-AID cells also ectopically expressed TIR1 from a sleeping beauty transposon integrated cassette. TIR1 protein is the E3 ubiquitin ligase required to induce degradation of AID containing proteins in the presence auxin.