PARP Inhibitors as a Therapeutic Agent for Homologous Recombination Deficiency in Breast Cancers
Round 1
Reviewer 1 Report
JCM review of manuscript jcm_462403, March 5, 2019
“PARP inhibitors as a therapeutic agent for homologous recombination deficiency in breast cancers” by Keung MYT, Wu Y, and Vadgama JV.
The authors outline the known cellular mechanisms of PARP inhibitors, review different platforms available to predict PARP inhibitor sensitivity and present a comprehensive review of the currently active or ongoing clinical trials involving PARP inhibitors alone or in combination.
This review article provides broad information on the current knowledge on PARP inhibitors and the clinical experience with currently approved PARP inhibitors in the context of breast cancer.
The review is well structured and written. However, there are comments/ suggestions for changes listed below that would improve the clarity or content of this review.
Major comments:
1. A list of frequently used abbreviations should be provided to make it easier for the reader to follow the text.
2. HR stands for hormone receptors (line 38) and for homologous recombination (lines 46, 67, 72, 77 and more); it is suggested to use HRR for homologous recombination repair instead throughout the manuscript.
3. First paragraph in the introduction: it is worth mentioning that hormone receptor positive breast cancers can recur as receptor negative tumors and are equally challenging for therapies than TNBC.
4. Section: PARP inhibitors as therapeutic intervention
5. Lines 62/63: the post-translational modification is not the “addition of PAR to sites of DNA strand breaks”, but rather to proteins that are functionally involved in the recognition or repair of DNA damage. This should be specified.
6. Line 66: other excellent PARP reviews on the detailed structure-function relation should be mentioned here: e.g. [Gibson and Kraus; June 2012 NATURE REVIEWS, Molecular Cell Biology];
7. Line 69/70: “…… resulting in repair of DNA damage and cell viability” is too simplified a statement and needs explanation. The mechanisms of PARP-1, -2, -3 involvement in DNA repair processes are detailed in the following review and could help the reader to follow up in detail: Beck C et al., Experimental Cell Research, 2014.
8. Line 65: a consistent description of the PARP1 molecular domains throughout the review manuscript should be achieved: domain structure listed in line 65 is different from line 95 (ref 16)
9. Line 75: the term “PARP trapping” should be further explained – at least briefly in this section. Different mechanisms are discussed to cause PARP trapping to DNA that involve the lack of negative charge repulsion from the DNA or reversed allosteric mechanisms resulting in stronger Zn finger binding to DNA.
10. Figure 1: this figure only depicts the cell death resulting from PARP inhibitor treatment in BRCA1 non-functional cells with lack of efficient DSB repair; this should be mentioned in the legend.
11. Lines 241 ff: the HRDetect method lists 6 predictive parameters for sensitivity to PARP inhibitors in BRCA1/2 mut tumors. These predictors appear to be related to genomic changes resulting from alternative DNA repair mechanisms in BRCA1/2 mutated cancers. It would be very helpful for the reader of this review if the authors explained alternative DNA repair mechanisms (either as Figure 2 or as a separate paragraph in the chapter “PARP inhibitors as therapeutic interventiuon”) that BRCA1/2-deficient cancer cells can utilize, for example single-strand annealing, alternative end joining (which also uses PARP1), and classical NHEJ. This could also help to explain resistance to PARP inhibitors and the concept of Co-synthetic lethality to PARP inhibitors in the BRCA-mut context.
12. Line 251: how is the BRACAnalysis CDx test different from the other sequencing-based methods that are described in this chapter?
13. Lines 301-305: this paragraph addresses the immune-regulatory functions of PARP1 that are not related to the DNA repair functions addressed in this review. This should be clearly mentioned with a reference to other literature to PARP1 immuno-modulatory roles or the paragraph is to be omitted from this review.
Minor comments:
a. Line 158/159: it is accurate to say “BRCA1- and BRCA2-mutated cancers” instead of BRCA1 cancers etc.
b. Figure 2 legend: mention the reference for the “signature 3” in (e).
c. Line 185: does miR-182 sensitize to PARP1 inhibitors because it silences BRCA1? This should be added.
d. Line 280: refer to the chapter “Acquired resistance to PARP inhibition” here.
Author Response
Responses to Reviewer 1 Comments:
Major comments:
Point 1: A list of frequently used abbreviations should be provided to make it easier for the reader to follow the text.
Response 1: A list of frequently used abbreviations has been added to the text in alphabetical order following the section “Conflicts of Interest” in lines 650 - 697.
Point 2: HR stands for hormone receptors (line 38) and for homologous recombination (lines 46, 67, 72, 77 and more); it is suggested to use HRR for homologous recombination repair instead throughout the manuscript.
Response 2: We have modified the abbreviations so that HR stands for hormone receptors, HRR stands for homologous recombination repair, and HRD stands for homologous recombination deficiency. Abbreviations are noted in lines 669 - 671. The use of HR in the text as it previously referred to homologous recombination has been changed to either HRR or HRD (lines 49, 85, 91, 92, 101, 182, 188, 194, 195, 239, 241, 249, 259, 276, 333, 565, 568, 570, 572, 574, 58, 587, 595, 621) to eliminate confusion with HR referring to hormone therapy.
Point 3: First paragraph in the introduction: it is worth mentioning that hormone receptor positive breast cancers can recur as receptor negative tumors and are equally challenging for therapies than TNBC.
Response 3: We have included information which reflects the aforementioned challenge in lines 40 – 44 and included 3 references.
Point 4: Section: PARP inhibitors as therapeutic intervention
Response 4: Reviewer comments for this section are noted in points 5 through 9.
Point 5: Lines 62/63: the post-translational modification is not the “addition of PAR to sites of DNA strand breaks”, but rather to proteins that are functionally involved in the recognition or repair of DNA damage. This should be specified.
Response 5: We agree with Reviewer 1 and have corrected the sentence to emphasize that upon sensing a DNA strand break, PARP catalyzes the addition of PAR chains, and this is a post-translational modification which helps recruit other DNA repair proteins. This correction is included in lines 65 – 68. We’ve added additional references from 2016 to include more current knowledge of PARylation.
Point 6: Line 66: other excellent PARP reviews on the detailed structure-function relation should be mentioned here: e.g. [Gibson and Kraus; June 2012 NATURE REVIEWS, Molecular Cell Biology];
Response 6: We have included information from this 2012 review on PARP in lines 74 – 76, reference 20. We have added additional reviews which highlights the role of PARP1 in chromatin modification and the effect of NAD+ concentration (lines 76 – 79).
Point 7: Line 69/70: “…… resulting in repair of DNA damage and cell viability” is too simplified a statement and needs explanation. The mechanisms of PARP-1, -2, -3 involvement in DNA repair processes are detailed in the following review and could help the reader to follow up in detail: Beck C et al., Experimental Cell Research, 2014.
Response 7: We have incorporated Reviewer 1’s recommendations and provided additional explanation on recruitment of DNA repair factors. This explanation emphasizes that DNA repair is a complex process involving various components. The changes are reflected in lines 80 – 84. Referenced is the suggested review paper from Beck et al., reference 22.
Point 8: Line 65: a consistent description of the PARP1 molecular domains throughout the review manuscript should be achieved: domain structure listed in line 65 is different from line 95 (ref 16)
Response 8: We agree that there is a slight discrepancy in the molecular domains listed in these lines. The domain structures listed in line 65 (now lines 68 – 71) are the known functional units of PARP-1 as referenced in references 13, 15 – 18. The domain structures as listed in line 95 (now line 132 - 133) are domains which are specifically mentioned in Langelier et al’s research (reference 19). These specific domains form a network of connections linking the DNA damage to the catalytic domain (CAT).
Although the domains mentioned in the former describes known domains involved in PARP-1 activity, the domains mentioned in the latter are specific for linking the site of DNA damage to the catalytic domain. The changes made in the text reflect these notions. We have also included a notation to clarify that some researchers include the WGR domain as part of the CAT domain, whereas other researchers say WGR and the CAT domain are distinct domains (lines 72 – 74, lines 133 - 134). We have now clarified the difference in domain structures, as per the reviewer’s suggestions.
Point 9: Line 75: the term “PARP trapping” should be further explained – at least briefly in this section. Different mechanisms are discussed to cause PARP trapping to DNA that involve the lack of negative charge repulsion from the DNA or reversed allosteric mechanisms resulting in stronger Zn finger binding to DNA.
Response 9: We have elaborated briefly on the PARP trapping mechanism, with reference to the paper by Murai et al. This modification is reflected in lines 105 - 108.
Point 10: Figure 1: this figure only depicts the cell death resulting from PARP inhibitor treatment in BRCA1 non-functional cells with lack of efficient DSB repair; this should be mentioned in the legend.
Response 10: We have added to this figure legend according to the reviewer’s suggestion, mentioning synthetic lethality and the role of BRCA in homologous recombination repair. This change is reflected in lines 120 - 125.
Point 11: Lines 241 ff: the HRDetect method lists 6 predictive parameters for sensitivity to PARP inhibitors in BRCA1/2 mut tumors. These predictors appear to be related to genomic changes resulting from alternative DNA repair mechanisms in BRCA1/2 mutated cancers. It would be very helpful for the reader of this review if the authors explained alternative DNA repair mechanisms (either as Figure 2 or as a separate paragraph in the chapter “PARP inhibitors as therapeutic intervention”) that BRCA1/2-deficient cancer cells can utilize, for example single-strand annealing, alternative end joining (which also uses PARP1), and classical NHEJ. This could also help to explain resistance to PARP inhibitors and the concept of Co-synthetic lethality to PARP inhibitors in the BRCA-mut context.
Response 11: A new paragraph with extensive information has been added which further elaborates upon alternative DNA repair mechanisms that come into play. This paragraph is in lines 85 – 102.
Point 12: Line 251: how is the BRACAnalysis CDx test different from the other sequencing-based methods that are described in this chapter?
Response 12: We have included a description of BRACAnalysis CDx in lines 290 – 292. The specific utility of BRACAnalysis CDx is in identifying germline BRCA1/2 mutations. However, the other sequencing-based methods listed in this review have broader functions, such as MSK-Impact in detecting various mutation types in over 300 cancer-associated genes. We’ve provided a discussion of those approaches in this review.
Point 13: Lines 301-305: this paragraph addresses the immune-regulatory functions of PARP1 that are not related to the DNA repair functions addressed in this review. This should be clearly mentioned with a reference to other literature to PARP1 immuno-modulatory roles or the paragraph is to be omitted from this review.
Response 13: We have decided to eliminate at this time the discussion of immune checkpoint inhibitors in combination with PARP inhibitors, as they are outside the scope of this review.
Minor comments:
Point A: Line 158/159: it is accurate to say “BRCA1- and BRCA2-mutated cancers” instead of BRCA1 cancers etc.
Response A: We have modified the sentence structure so that it is more general to say that germline mutations are more common than somatic mutations in cancers and eliminated the phrase “(11% and 2.5%, respectively, in 560 breast cancers surveyed)”. This change is reflected in lines 196 - 197.
Point B: Figure 2 legend: mention the reference for the “signature 3” in (e).
Response B: We have added the reference for signature 3 in line 221.
Point C: Line 185: does miR-182 sensitize to PARP1 inhibitors because it silences BRCA1? This should be added.
Response C: Line 224 includes a brief description of how miR-182 sensitizes to PARP inhibitors, by downregulating BRCA1.
Point D: Line 280: refer to the chapter “Acquired resistance to PARP inhibition” here.
Response D: We have addressed this comment by adding the phrase: please refer to section 8, “Acquired resistance to PARP inhibition” in line 320 - 321.
Reviewer 2 Report
This review article on the roles of ubiquitin in the degradation of aggregation-prone proteins by Keung et al. manages to take a complex subject and summarize it into a very interesting and comprehensive review that is difficult to put down until one has finished reading it.
I only have a few suggestions:
1 Given the well-established function of 53BP1 and BRCA1 in deciding between Homologous Recombination Repair (HRR) and Non-Homologous end joining (NHEJ), the authors should elaborate more on the role of 53BP1 in promoting PARPi cytotoxicity. Specifically, the authors should cite the literature demonstrating that 53BP1 blocks HRR by inhibiting 5’ end resection (PMID: 20362325; PMID: 27670884; PMID: 20453858).
2 The authors discuss how the loss of PARG is a mechanism of resistance to PARPi citing reference 122, which is incorrect. The correct reference is PMID: 29894693.
3 Resistance of BRCA1/2-mutant cancers to PARP inhibitors can result from changes in DNA repair pathway proteins that abrogate the action of the PARP inhibitors. For example, overexpression of RAD51, a key HRR-protein, is an important mechanism of PARPi resistance. Therefore, factors that increase RAD51 levels or activity can potentially lead to a resistance to PARPi. In particular, a recent study demonstrated that EMI1, a substrate receptor of an SCF ubiquitin ligase (PMIDs: 23657496 and 25394868), constitutively targets RAD51 for degradation. A subset of breast cancers develop resistance to PARPi by downregulating EMI1 and reconstitution of EMI1 expression reestablishes PARPi sensitivity in vitro and in vivo (PMID: 30554948). This mechanism should be discussed and cited accordingly.
Author Response
Responses to Reviewer 2 comments:
Point 1: Given the well-established function of 53BP1 and BRCA1 in deciding between Homologous Recombination Repair (HRR) and Non-Homologous end joining (NHEJ), the authors should elaborate more on the role of 53BP1 in promoting PARPi cytotoxicity. Specifically, the authors should cite the literature demonstrating that 53BP1 blocks HRR by inhibiting 5’ end resection (PMID: 20362325; PMID: 27670884; PMID: 20453858).
Response 1: The aforementioned literature was included in a more detailed discussion of 53BP1 and resistance to PARP inhibitors. Reviewer 2’s suggested literature is added as references 118, 119, 122. Additional references are from studies with in vivo models. These changes are reflected in lines 565 – 579.
Point 2: The authors discuss how the loss of PARG is a mechanism of resistance to PARPi citing reference 122, which is incorrect. The correct reference is PMID: 29894693.
Response 2: The corrected reference was added in line 628.
Point 3: Resistance of BRCA1/2-mutant cancers to PARP inhibitors can result from changes in DNA repair pathway proteins that abrogate the action of the PARP inhibitors. For example, overexpression of RAD51, a key HRR-protein, is an important mechanism of PARPi resistance. Therefore, factors that increase RAD51 levels or activity can potentially lead to a resistance to PARPi. In particular, a recent study demonstrated that EMI1, a substrate receptor of an SCF ubiquitin ligase (PMIDs: 23657496 and 25394868), constitutively targets RAD51 for degradation. A subset of breast cancers develop resistance to PARPi by downregulating EMI1 and reconstitution of EMI1 expression reestablishes PARPi sensitivity in vitro and in vivo (PMID: 30554948). This mechanism should be discussed and cited accordingly.
Response 3: The role of EMI1 in modulating PARPi resistance was included in greater detail with the suggested citations, listed as references 124 – 126. Changes made are reflected in lines 580 – 589.
Round 2
Reviewer 1 Report
the authors have addressed the previously raised concerns and incorporated suggested changes.
the review manuscripts reads well and provides valuable information on the basic and clinical aspects of PARP inhibitor treatments in breast cancer.
