Chromatin endogenous cleavage provides a global view of RNA polymerase II transcription kinetics

Reviewer #1 (Public review):

Summary:

In this study, the authors use ChEC-seq, an MNase-based method to map yeast RNA pol II. Part of the reasoning for this study is that earlier biochemical work suggested pol II initiation and termination should involve slow steps at the UAS/promoter and termination regions that are not well visualized by formaldehyde-based ChIP methods. Here the authors find that pol II ChIP and ChEC give complementary patterns. Pol II ChIP signals are strongest in the coding region (where ChIP signal correlates well with transcription (rho = 0.62)). In contrast, pol II ChEC signals are strongest at promoters (rho = 0.52) and terminator regions. Weaker upstream ChEC signals are also observed at the STM class genes where biochemical studies have suggested a form of Pol (and maybe other general factors) is recruited to UAS sites. ChEC of TFIIA and TFIIE give promoter-specific ChEC signals as expected. Extending this work to elongation factors Ctk1 and Spt5 unexpectedly give strong signals near the PIC location and little signals over the coding region. This, and mapping CTD S2 and S5 phosphorylation by ChEC suggests to me that, for some reason, ChEC isn't optimal for detecting components of the elongation complex over coding regions.

Examples are also presented where perturbations of transcription can be measured by ChEC. Modeling studies are shown where adjustment of kinetic parameters agree well with ChEC data and that these models can be used to estimate which steps in transcription are affected by various perturbations. No tests were performed to see if the predictions could be validated by other means. Finally, the role of nuclear pore binding by Gcn4 is explored, although the results do not seem convincing. Overall, the authors show that pol II ChEC is a valuable and complementary method for investigating transcription mechanisms and slow steps at the initiation and termination regions.

Reviewer #2 (Public review):

Summary:

The study by VanBelzen et. al. compares chromatin immunoprecipitation (ChIP-seq) and chromatin endogenous cleavage sequencing (ChEC-seq2) to examine RNA polymerase II (RNAPII) binding patterns in yeast. While ChIP-seq shows RNAPII enrichment mainly over transcribed regions, ChEC-seq2 highlights RNAPII binding at promoters and upstream activating sequences (UASs), suggesting it captures distinct RNAPII populations that the authors speculate are linked more tightly to active transcription. The authors develop a stochastic model for RNAPII kinetics using ChEC-seq2 data, revealing insights into transcription regulation and the role of the nuclear pore complex in stabilizing promoter-associated RNAPII. The study suggests that ChEC-seq2 identifies regulatory events that ChIP-seq may overlook.

Strengths:

(1) This is a carefully crafted study that adds significantly to existing literature in this area. Transgenic MNase fusions with endogenous Rpb1 and Rpb3 subunits were carefully performed, and complemented by fusions with several additional proteins that help the authors to dissect the transcription cycle. Both the S. cerevisiae lines and the sequencing data are likely to be of significant use to the community.

(2) The validation of ChEC-seq2 and its comparison with ChIP-seq is highly valuable technical information for the community.

(3) The kinetic modeling appears to be thoughtfully done.

Weaknesses:

(1) The term "nascent transcription" is all too often used interchangeably for NET-seq, PRO-seq, 4sU-seq, and other assays that often provide different types of information. The authors should make it clear their use of the term refers to SLAM-seq data.

(2) The authors do not perform any comparison to run-on (PRO-seq) data. My impression is that the distribution of PRO-seq signal in S. cerevisiae agrees better with the distribution the authors observe by ChIP-seq. PRO-seq only captures RNAPII that is engaged and actively transcribing. If PRO-seq does indeed provide a similar profile as ChIP-seq, wouldn't this indicate that the high frequency of association between RNAPII and either the promoter or UAS reflects RNAPII that has not yet started transcription elongation? Perhaps this could help sort out what types of activities are occurring at the UAS (which does not appear to require a full PIC) or at the promoter (which does)?

Author response:

Response to Reviewer #1:

We agree with the reviewer that ChIP is much better than ChEC at recovering RNA polymerase II and elongation factors associated with the transcribed regions. We believe that this is due to cross-linking, which enriches for these interactions. However, as we highlight in the manuscript, cross-linking may not accurately report on the occupancy of RNA polymerase II and elongation factors over all regions. Although, by ChEC, we observe elongation factors upstream of the transcribed region, compared with total RNA polymerase II, the relative enrichment of elongation factors or phosphorylated RNA polymerase II is significantly higher over transcribed regions, with a bias to the 3’ end (Figure 4B & C). This is consistent with these proteins and modifications functioning in elongation.

Regarding how convincing the results with the gcn4-pd mutant are, we would highlight that the phenotype of this mutant is a quantitative decrease in transcription and this leads to a quantitative decrease, rather than qualitative loss, of RNA polymerase II over the promoter, without impacting the association of RNA polymerase II over the UAS region. This effect was small but statistically significant (p = 4e-5). Obviously, more mechanistic studies will need to be performed, but this result supports a role for the interaction with the nuclear pore complex in either enhancing the transfer of RNA polymerase II from the enhancer to the promoter or in preventing its dissociation from the promoter.

Response to Reviewer #2:

Thank you for your supportive comments and suggestions. We will clarify our use of Nascent RNA in the text. We agree that the stronger enrichment of the transcribed region from Rpb1 ChIP-seq experiments should correlate well with actively transcribing RNA polymerase II observed by PRO-seq; enrichment by PRO-seq reported in a paper from John Lis’ lab strongly favors transcribed regions with a modest peak over the terminator (PMID: 27197211, Figure 2A). ChEC reveals functionally important forms of RNA polymerase II that are not engaged in transcription. This manuscripts highlights the potential utility of ChEC-seq2 in measuring these interactions - suggested by the recent work from Buratowski and Gelles’ single molecule studies - globally.