Harmful DNA:RNA hybrids are formed in cis and in a Rad51-independent manner
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
Abstract
DNA:RNA hybrids constitute a well-known source of recombinogenic DNA damage. The current literature is in agreement with DNA:RNA hybrids being produced co-transcriptionally by the invasion of the nascent RNA molecule produced in cis with its DNA template. However, it has also been suggested that recombinogenic DNA:RNA hybrids could be facilitated by the invasion of RNA molecules produced in trans in a Rad51-mediated reaction. Here, we tested the possibility that such DNA:RNA hybrids constitute a source of recombinogenic DNA damage taking advantage of Rad51-independent single-strand annealing (SSA) assays in the yeast Saccharomyces cerevisiae . For this, we used new constructs designed to induce expression of mRNA transcripts in trans with respect to the SSA system. We show that unscheduled and recombinogenic DNA:RNA hybrids that trigger the SSA event are formed in cis during transcription and in a Rad51-independent manner. We found no evidence that such hybrids form in trans and in a Rad51-dependent manner.
Article activity feed
-
-
###This manuscript is in revision at eLife
The decision letter after peer review, sent to the authors on April 16, 2020, follows.
Summary
Using a clever genetic system in the budding yeast Saccharomyces cerevisiae the authors test whether R-loop can form in trans, meaning that a transcript from locus A could lead to R-loop formation in locus B. Moreover, they test whether R-loop formation is dependent on Rad51, the eukaryotic RecA family recombinase. Using their genetic system and cytological analysis of Rad52 foci and the S9.6 antibody to detect R-loops in wild type and strains with mutations known to affect R-loop, conclusive data are shown that R-loops only form in cis and that R-loops in this genetic system are independent of Rad51. Overall, this work significantly enriches the discussion in the R-loop field and provides an …
###This manuscript is in revision at eLife
The decision letter after peer review, sent to the authors on April 16, 2020, follows.
Summary
Using a clever genetic system in the budding yeast Saccharomyces cerevisiae the authors test whether R-loop can form in trans, meaning that a transcript from locus A could lead to R-loop formation in locus B. Moreover, they test whether R-loop formation is dependent on Rad51, the eukaryotic RecA family recombinase. Using their genetic system and cytological analysis of Rad52 foci and the S9.6 antibody to detect R-loops in wild type and strains with mutations known to affect R-loop, conclusive data are shown that R-loops only form in cis and that R-loops in this genetic system are independent of Rad51. Overall, this work significantly enriches the discussion in the R-loop field and provides an alternative view point of an earlier publication that suggested R-loop formation in trans being catalyzed by Rad51.
Essential Revisions
The pGal promoter induces very high transcription in the presence of galactose (often 500X or more induction). The level is likely very different (much less) for the tet promoter, which is generally only induced 2-3X upon addition of doxycycline. This could significantly affect the results - e.g. the cis vs trans effects could really be a matter of different transcription levels. Transcription levels from each promoter really need to be determined- this is a very important control. The exact induction conditions used, including concentrations and induction times, need to be spelled out in the methods and should be consistent with those used during the RT-PCR experiment to test transcript levels. In the absence of being able to do the experiment on the constructs used (which would be optimal), at least it could be cited if this lab has used the same promoters and induction conditions in the past, and a caveat inserted if transcription levels are different. It would also be good to switch the promoters and make sure the result holds, as there could be issues of differences in timing of transcription as well.
In Figure 2 the authors relate recombination frequencies in their assays to RNA:DNA hybrid formation without measuring hybrids directly. This is a major weakness that significantly limits data interpretation. For instance, I am very surprised that the "cis" recombination frequency of the inverted LacZ reporter is essentially as high as the regular lacZ construct. This result implies that hybrid formation is insensitive to the orientation of the reporter when in many reported cases, R-loop formation is strongly orientation-dependent. Of course, another hypothesis is that (stalled) transcription itself triggers recombination, not R-loops. Without data on R-loop formation, one cannot disentangle transcription from co-transcriptional R-loop formation. The authors must use DRIP-based assay to quantify R-loop levels in the various sequence contexts and under the various genetic backgrounds to establish that their assay is reflective of R-loop levels. Using bisulfite-based readouts to measure R-loop distributions and lengths across the LacZ region would be even better. Without this data, the claim that this new genetic assay can "infer the formation of recombinogenic DNA:RNA hybrids" is unsubstantiated.
Source data. The source data file should be labeled better. Missing are:
- what the numbers in the table are (rates of Leu+ x 10^-4?)
- which data goes with which Figure panel
- average and SEM numbers should be shown in the data table
The exact p values not reported and could be added to source data file. N values can be discerned from the source data file but it would be nice for them to be stated in the figure legends.
-
