R-Loop Formation in Meiosis: Roles in Meiotic Transcription-Associated DNA Damage

Round 1

Reviewer 1 Report

This review manuscript nicely summarizes the possible functions of R-loops in meiosis. Because an R-loop-related protein SETX is specifically required for male meiosis, the possible functions of R-loops in meiosis is an important unanswered question in the field. The manuscript comprehensively covers two possible functions of R-loops in meiotic DNA repair and transcriptions. I enjoyed reading the manuscript, and I believe that the manuscript will be an important asset in the field. The figures are very clear. I especially like Fig. 4; this is a good summary of how R-loops are regulated. Overall, I support the publication of this manuscript. I have two major suggestions and other minor points to improve the manuscript.

 

Major points

1.     Lines 111-124: The manuscript well covered an undated knowledge in the studies of R-loops and introduced a new concept of two classes of R-loops. Overall, this is a good summary. At the end of the paragraph, a caveat for the ex vivo method was described. Because the concept of two classes was introduced in a recent review (Ref#19), Ref#19 is sufficient for citation here. In addition, both "ex vivo" and "in vivo" methods have distinct advantages. These respective advantages and differences should be highlighted instead of showing a caveat of "ex vivo" methods. Also, “ex vivo” and “in vivo” should be clearly defined (“ex vivo” and “in vivo” are often described differently by different people).

2.     Figure 1 and lines 197-202; The description of NHEJ in meiosis is not accurate. NHEJ is generally suppressed in germ cells, and Ref# 99 and 100 do not conclude that NHEJ is turned on during the late prophase. In Ref #99, the authors show that NHEJ factors accumulated on damage sites, but they found RAD51-dependent somatic HR operates after the accumulation of NHEJ factors at that sites. Ref #100 also showed somatic HR operates in later prophase. The chart in Fig. 1: this should be described as the localization of NHEJ factors (not activity). There is no study showing that NHEJ dominates after meiotic prophase.

 

Minor points:

1.     Abstract Line 11-13, I do not understand why meiosis is a strategy of meiotic transcription.

2.     Line 67: “Fewer genes”; compared to what?

3.     Line 79-80: MYBL-1 (A-MYB) can be described in the main text in the contest of transcription burst.

4.     Line 100: how do DNA lesions compromise transcription fidelity via Pol II?

5.     Line 103: I do not think meiotic DSBs and HR activities are restricted to recombination hotspots.

6.     Line 132: "Impaired resolution" should be "Loss-of-function"

7.     Some more information can be added to describe Dhx9 and Dhx39b

8.     Line 176 “RNaseH2 subunit-encoding genes” should be described mutant mice for these genes.

9.     Line 179: Setx should be indispensable for meiosis.

10.  Line 199: the possible function of SETX in NHEJ in meiosis can be carefully described (see my major point #2).

11.  Line 209: Ref #104 only explained female meiosis. This may not be relevant here because SETX is only required for male meiosis.

12.  Line 245: Why do the authors think RNaseH should be suppressed in meiosis? 

Author Response

We highly appreciate the reviewer's insightful comments. Attached below is our comment-by-comment response. 

 

Major points

 

1. Lines 111-124: The manuscript well covered an undated knowledge in the studies of R-loops and introduced a new concept of two classes of R-loops. Overall, this is a good summary. At the end of the paragraph, a caveat for the ex vivo method was described. Because the concept of two classes was introduced in a recent review (Ref#19), Ref#19 is sufficient for citation here. In addition, both "ex vivo" and "in vivo" methods have distinct advantages. These respective advantages and differences should be highlighted instead of showing a caveat of "ex vivo" methods. Also, “ex vivo” and “in vivo” should be clearly defined (“ex vivo” and “in vivo” are often described differently by different people).

Thank you for his/her suggestion. We removed Ref#57 and #60 from this paragraph as #19 is sufficient. Also, we removed “in vivo” and “ex vivo” and explained the difference between Class I and Class II more clearly.

 

2. Figure 1 and lines 197-202; The description of NHEJ in meiosis is not accurate. NHEJ is generally suppressed in germ cells, and Ref# 99 and 100 do not conclude that NHEJ is turned on during the late prophase. In Ref #99, the authors show that NHEJ factors accumulated on damage sites, but they found RAD51-dependent somatic HR operates after the accumulation of NHEJ factors at that sites. Ref #100 also showed somatic HR operates in later prophase. The chart in Fig. 1: this should be described as the localization of NHEJ factors (not activity). There is no study showing that NHEJ dominates after meiotic prophase.

Thank you for his/her suggestion. We modified Figure1 and its legend as suggested. “Activity” was changed to “localization of repair factors”, and the lines were changed to dashed lines after late pachytene stage because of uncertainties.

 

Minor points:

 

1. Abstract Line 11-13, I do not understand why meiosis is a strategy of meiotic transcription.

We are sorry about the confusing sentence. We now changed it to “mammalian meiosis is thought to employ a strategy of exclusively transcribing meiotic or post-meiotic genes”.

 

2. Line 67: “Fewer genes”; compared to what?

We corrected this sentence accordingly.

 

3. Line 79-80: MYBL-1 (A-MYB) can be described in the main text in the contest of transcription burst.

We added a description of MYBL1 in the main text.

 

4. Line 100: how do DNA lesions compromise transcription fidelity via Pol II?

We apologize if “(DNA lesions) around the genic region” confused you. We intended to say “DNA legions WITHIN genes”, and we changed it so.

 

5. Line 103: I do not think meiotic DSBs and HR activities are restricted to recombination hotspots.

We changed “restricted” to “preferentially observed”.

 

6. Line 132: "Impaired resolution" should be "Loss-of-function"

We appreciate your indication. Because several mutations have been identified in SETX and RNASEH proteins, both loss-of-function and gain-of-function, we changed the sentence to “Failure of R-loop resolution due to the defects of these molecules can….…..”.

 

7. Some more information can be added to describe Dhx9 and Dhx39b

We added their involvement in cancers.

 

8. Line 176 “RNaseH2 subunit-encoding genes” should be described mutant mice for these genes.

We corrected this sentence accordingly.

 

9. Line 179: Setx should be indispensable for meiosis.

We corrected this sentence accordingly.

 

10. Line 199: the possible function of SETX in NHEJ in meiosis can be carefully described (see my major point #2).

We appreciate your comment. Because the possible function of SETX in NHEJ is still ambiguous, we removed the sentence. In addition, we modified Figure 1, so that repair pathways later than late pachytene are still not known. 

 

11. Line 209: Ref #104 only explained female meiosis. This may not be relevant here because SETX is only required for male meiosis.

We removed Ref#104 and related sentences.

 

12. Line 245: Why do the authors think RNaseH should be suppressed in meiosis?

Thank you for your comment. Because it appears that one of the main purposes of producing a large number of transcripts in pachytene spermatocytes is to use these RNAs during the subsequent spermatogenic stages, RNAs in R-loops must be unwound rather than degraded by SETX.

 

Reviewer 2 Report

Review has been written well with suitable subtopics organization. This article will help everyone to find right information and to understand role of R loops in meiosis. I recommend this article for the publication.

Authors can also include potential DNA damage response factors involved in regulation of R loops during meiosis e.g. role of check point kinases.

Author Response

Review has been written well with suitable subtopics organization. This article will help everyone to find right information and to understand role of R loops in meiosis. I recommend this article for the publication.

We appreciate the reviewer's positive comments and recommendation for publication.

 

Authors can also include potential DNA damage response factors involved in regulation of R loops during meiosis e.g. role of check point kinases.

 

Thank you for the suggestion. We also assume that ATM and ATR play important roles in transcription-associated R-loop formation and subsequent DNA damage response. Thus we considered including this point in the revised manuscript. Unlike studies in somatic cells, however, very little is known about the roles of ATM and ATR kinases in the meiotic transcription. We are also concerned that the statements about ATM and ATR may lead the reader to think that the topic of this manuscript is meiotic DNA replication rather than meiotic transcription. Based on these reasons, we decided not to mention this in this review manuscript. We hope the reviewer will understand.

Reviewer 3 Report

The review by Fijiwara et al tries to gather the information on R-loops in meiosis and, more importantly, pretends to provide a perspective on their putative role or effect (positive or negative) in this process. This is mainly reasoned based on observations in mouse mutants in genes that are critical for R-loop maintenance that cause infertility. Although this is potentially a very interesting topic, in the current form, this manuscript is unfocussed and preliminary. It is difficult to understand and the sections seem disconnected between themselves.  The citations are not specifically well-chosen and some are missing. Figures are not clear and figure legends are not sufficiently explanatory by themselves.

 

Some examples and specific comments follow:

 

Line 36 uses reference 5 for R-loops but this reference refers to DNA-RNA hybrids as an intermediate of in vitro genetic transcription and not to R-loops. Indeed, hybrids can form inside the RNAPII holoenzyme but R-loops are a different thing that should not be confused with these natural hybrids. Reference 5 could be removed and substituted by many others (19-26). Indeed, there are too many references that talk about the same thing or almost. Reference 6 and 21 are actually from the same group…citations seem somehow repetitive and not well-chosen. Furthermore, some critical references are missing: Line 39, Reference PMID: 21211720 should be cited when describing Sen1 as it was the first paper describing Sen1 role in hybrids., Line 56, R-loops can contribute to meiotic progression but also affect it (impairing genome stability). Reference PMID: 22878416 should be added and discussed here or somewhere else in the manuscript.

Some concepts are misleading:

R-loops do not only cause DNA damage due to the displaced ssDNA but, more importantly due to replication impairment and this is not understood from line 129.

The physiological and pathological R-loops should be distinguished as the reader might understand from this review that all types of R-loops cause DNA damage and this is not the case.

Figure 3. What are tpm and dpp?

Figure 4 Transcribed is spelled with b and not v.

Author Response

The review by Fijiwara et al tries to gather the information on R-loops in meiosis and, more importantly, pretends to provide a perspective on their putative role or effect (positive or negative) in this process. This is mainly reasoned based on observations in mouse mutants in genes that are critical for R-loop maintenance that cause infertility. Although this is potentially a very interesting topic, in the current form, this manuscript is unfocussed and preliminary. It is difficult to understand and the sections seem disconnected between themselves.  The citations are not specifically well-chosen and some are missing. Figures are not clear and figure legends are not sufficiently explanatory by themselves.

We appreciate his/her criticism. We agree that the review article is a little immature. The reason is partly because while R-loop studies have made remarkable progress over the past several years, the details are still not fully understood. And more critically, little is known about the function of R-loop during meiosis in male germ cells, except for a report of SETX KO mice. Therefore, this manuscript is more like a call for interest or problem statement regarding the function of R-loop during meiosis rather than a comprehensive review. The importance of making this attempt is also evident from the phenotype of the SETX KO mice. That is, the pathological phenotype of systemic SETX KO is limited to late meiotic arrest and male infertility and later neuro-muscular symptoms. These findings suggest that at least SETX-dependent R-loop resolution in these non-replicating cells is distinct from functions suggested in somatic cells, such as the balance between DNA replication and transcription. We hope that this review will raise more pathophysiological questions for R-loop research.

 

Some examples and specific comments follow:

Line 36 uses reference 5 for R-loops but this reference refers to DNA-RNA hybrids as an intermediate of in vitro genetic transcription and not to R-loops. Indeed, hybrids can form inside the RNAPII holoenzyme but R-loops are a different thing that should not be confused with these natural hybrids. Reference 5 could be removed and substituted by many others (19-26). Indeed, there are too many references that talk about the same thing or almost. Reference 6 and 21 are actually from the same group…citations seem somehow repetitive and not well-chosen. Furthermore, some critical references are missing: Line 39, Reference PMID: 21211720 should be cited when describing Sen1 as it was the first paper describing Sen1 role in hybrids., Line 56, R-loops can contribute to meiotic progression but also affect it (impairing genome stability). Reference PMID: 22878416 should be added and discussed here or somewhere else in the manuscript.

Thank you for the suggestion. We removed Ref#5 and added a brief description about the difference between physiological and pathological hybrids as below. We also added or deleted the references according to your suggestion.

“While DNA-RNA hybrids are physiologically formed at certain regions with a physiological function, R-loops are generally considered as pathological and detrimental products that interfere with the transcription process and subsequently contribute to genome instability because unstructured single-stranded DNA is targeted for damage.”

 

Some concepts are misleading:

R-loops do not only cause DNA damage due to the displaced ssDNA but, more importantly due to replication impairment and this is not understood from line 129.

Thank you for your concern. In this sentence, we specifically intended about meiotic transcription-associated R-loop formation, which occurs after DNA replication is completed. To make our intention clear, we changed the subtitle as “R-loops and meiotic transcription” and changed the sentence as “Transcription-associated R-loop structures contain ssDNA susceptible to environ-mental stress and, thus, could be a source of DNA damage.”

 

The physiological and pathological R-loops should be distinguished as the reader might understand from this review that all types of R-loops cause DNA damage and this is not the case.

We added a brief description about the difference between physiological and pathological hybrids, as explained above.

 

Figure 3. What are tpm and dpp?

Thank you for pointing out this. TPM stands for transcripts per million. DPP stands for days post partum.  We added this explanation in the legend.

 

Figure 4 Transcribed is spelled with b and not v.

We corrected the figure 4.