DNA-directed termination of RNA Polymerase II transcription II

RNA polymerase II (RNAPII) transcription involves initiation from promoters, transcript elongation through the gene body, and cessation of transcription in the downstream terminator regions. In contrast to bacteria, where terminators often contain specific DNA elements to direct RNAP dissociation1, termination by RNAPII is thought to be driven entirely by protein co-factors1-3. Here we use biochemical reconstitution to shed new light on RNAPII termination. Unexpectedly, transcription through a terminator region by pure RNAPII results in a significant amount of intrinsic polymerase dissociation at specific sequences containing T-tracts. A combination of biochemistry and single molecule analysis indicates that such intrinsic termination involves pausing without backtracking prior to spontaneous RNAPII dissociation from the DNA template. Importantly, while the ‘torpedo’ Rat1-Rai1 RNA exonuclease (XRN2 in humans) works inefficiently on paused or stopped polymerases, it greatly stimulates intrinsic termination. By contrast, elongation factor Spt4-Spt5 (DSIF in humans) suppresses such termination.  Genome-wide analysis in yeast using 3’-end sequencing further supports the idea that transcriptional termination occurs by transcript cleavage at the polyA site exposing a new RNA-end  that allows loading of the Rat1-Rai1 torpedo, which then catches up with a destabilised RNAPII at intrinsic termination sites containing T-tracts to terminate transcription. In this study, we explored whether S.cerevisiae RNAPII intrinsic termination sites (T-tracts on the non-template strand) identified in vitro plays a function in RNAPII termination in vivo. To do so, we applied 3'end sequencing method to determine the 3'end of RNA transcripts. We carried out 3'end sequencing experiment in both wild type yeasts and rrp6 depletion yeasts, with or without in vitro polyadenylation by E.coli polyA polymerase (EPAP). After determining that transcripts are indeeded preferentially released at T-tracts, we performed the same 3'end sequencing experiments in Rat1(Xrn2 in humans) anchor away (Rat1AA) and Ysh1 (CPSF73 in humans) anchor away (Ysh1-AA) yeast strains, further confirming that transcripts released at T-tracts are indeed due to transcription termination rather than transcript cleavage by cleavage and polyadenylation factors (CPF in yeast, CPSF in humans). De novo motif discovery confirmed the presence of T-tracts at 3'end of transcripts from 3'end sequencing datasets.