Repression of PRMT activities sensitize homologous recombination-proficient ovarian and breast cancer cells to PARP inhibitor treatment

Reviewer #2 (Public review):

Summary:

The authors show that a combination of arginine methyltransferase inhibitors synergize with PARP inhibitors to kill ovarian and triple negative cancer cell lines in vitro and in vivo using preclinical mouse models.

Strengths and weaknesses

The experiments are well-performed, convincing and have the appropriate controls (using inhibitors and genetic deletions) and use statistics.

They identify the DNA damage protein ERCC1 to be reduced in expression with PRMT inhibitors. As ERCC1 is known to be synthetic lethal with PARPi, this provides a mechanism for the synergy. They use cell lines only for their study in 2D as well as xenograph models.

Author response:

The following is the authors’ response to the original reviews

Public Reviews:

Reviewer #1 (Public Review):

Summary:

The authors aimed to enhance the effectiveness of PARP inhibitors (PARPi) in treating high-grade serous ovarian cancer (HGSOC) and triple-negative breast cancer (TNBC) by inhibiting PRMT1/5 enzymes. They conducted a drug screen combining PARPi with 74 epigenetic modulators to identify promising combinations.

Zhang et al. reported that protein arginine methyltransferase (PRMT) 1/5 inhibition acts synergistically to enhance the sensitivity of Poly (ADP-ribose) polymerase inhibitors (PARPi) in high-grade serous ovarian cancer (HGSOC) and triple-negative breast cancer (TNBC) cells. The authors are the first to perform a drug screen by combining PARPi with 74 well-characterized epigenetic modulators that target five major classes of epigenetic enzymes. Their drug screen identified both PRMT1/5 inhibitors with high combination and clinical priority scores in PARPi treatment. Notably, PRMT1/5 inhibitors significantly enhance PARPi treatment-induced DNA damage in HR-proficient HGSOC and TNBC cells through enhanced maintenance of gene expression associated with DNA damage repair, BRCAness, and intrinsic innate immune pathways in cancer cells. Additionally, bioinformatic analysis of large-scale genomic and functional profiles from TCGA and DepMap further supports that PRMT1/5 are potential therapeutic targets in oncology, including HGSOC and TNBC. These results provide a strong rationale for the clinical application of a combination of PRMT and PARP inhibitors in patients with HR-proficient ovarian and breast cancer. Thus, this discovery has a high impact on developing novel therapeutic approaches to overcome resistance to PARPi in clinical cancer therapy. The data and presentation in this manuscript are straightforward and reliable.

Strengths:

(1) Innovative Approach: First to screen PARPi with a large panel of epigenetic modulators.

(2) Significant Results: Found that PRMT1/5 inhibitors significantly boost PARPi effectiveness in HR-proficient HGSOC and TNBC cells.

(3) Mechanistic Insights: Showed how PRMT1/5 inhibitors enhance DNA damage repair and immune pathways.

(4) Robust Data: Supported by extensive bioinformatic analysis from large genomic databases.

Weaknesses:

(1) Novelty Clarification: Needs clearer comparison to existing studies showing similar effects.

(2) Unclear Mechanisms: More investigation is needed on how MYC targets correlate with PRMT1/5.

(3) Inconsistent Data: ERCC1 expression results varied across cell lines.

(4) Limited Immune Study: Using immunodeficient mice does not fully explore immune responses.

(5) Statistical Methods: Should use one-way ANOVA instead of a two-tailed Student's t-test for multiple comparisons.

We sincerely thank Reviewer #1 for the insightful and constructive feedback, as well as for the kind acknowledgment of the significance of our work: “These results provide a strong rationale for the clinical application of a combination of PRMT and PARP inhibitors in patients with HR-proficient ovarian and breast cancer. Thus, this discovery has a high impact on developing novel therapeutic approaches to overcome resistance to PARPi in clinical cancer therapy. The data and presentation in this manuscript are straightforward and reliable.” We greatly appreciate the reviewer #1’s thoughtful comments, which have significantly improved the quality of our manuscript. In response, we conducted additional experiments and analyses, and made comprehensive revisions to the text, figures, and supplementary materials. In the “Recommendations for the authors” sections, we have provided point-by-point responses to each of the reviewer’s comments, which were immensely helpful in guiding our revisions. We believe these updates have substantially strengthened the manuscript and have fully addressed all reviewer concerns.

Reviewer #2 (Public Review):

Summary:

The authors show that a combination of arginine methyltransferase inhibitors synergize with PARP inhibitors to kill ovarian and triple-negative cancer cell lines in vitro and in vivo using preclinical mouse models.

PARP inhibitors have been the common targeted-therapy options to treat high-grade serous ovarian cancer (HGSOC) and triple-negative breast cancer (TNBC). PRMTs are oncological therapeutic targets and specific inhibitors have been developed. However, due to the insufficiency of PRMTi or PARPi single treatment for HGSOC and TNBC, designing novel combinations of existing inhibitors is necessary. In previous studies, the authors and others developed an "induced PARPi sensitivity by epigenetic modulation" strategy to target resistant tumors. In this study, the authors presented a triple combination of PRMT1i, PRMT5i and PARPi that synergistically kills TNBC cells. A drug screen and RNA-seq analysis were performed to indicate cancer cell growth dependency of PRMT1 and PRMT5, and their CRISPR/Cas9 knockout sensitizes cancer cells to PARPi treatment. It was shown that the cells accumulate DNA damage and have increased caspase 3/7 activity. RNA-seq analysis identified BRCAness genes, and the authors closely studied a top hit ERCC1 as a downregulated DNA damage protein in PRMT inhibitor treatments. ERCC1 is known to be synthetic lethal with PARP inhibitors. Thus, the authors add back ERCC1 and reduce the effects of PRMT inhibitors suggesting PRMT inhibitors mediate, in part, their effect via ERCC1 downregulation. The combination therapy (PRMT/PARP) is validated in 2D cultures of cell lines (OVCAR3, 8 and MDA-MB-231) and has shown to be effective in nude mice with MDA-MB-231 xenograph models.

Strengths and weaknesses:

Overall, the data is well-presented. The experiments are well-performed, convincing, and have the appropriate controls (using inhibitors and genetic deletions) and statistics.

They identify the DNA damage protein ERCC1 to be reduced in expression with PRMT inhibitors. As ERCC1 is known to be synthetic lethal with PARPi, this provides a mechanism for the synergy. They use cell lines only for their study in 2D as well as xenograph models.

We sincerely thank Reviewer #2 for the insightful and constructive feedback, as well as for the kind acknowledgment of the significance of our work: “Overall, the data are well-presented. The experiments are well-performed, convincing, and supported by appropriate controls (using inhibitors and genetic deletions) and statistics.” We greatly appreciate the reviewer #2’s thoughtful comments, which have significantly improved the quality of our manuscript. In response, we conducted additional experiments and analyses, and made comprehensive revisions to the text, figures, and supplementary materials. In the “Recommendations for the authors” sections, we have provided point-by-point responses to each of the reviewer’s comments, which were immensely helpful in guiding our revisions. We believe these updates have substantially strengthened the manuscript and have fully addressed all reviewer concerns.

Recommendations for the authors:

Reviewer #1 (Recommendations for the authors):

(1) Recent studies have revealed promising synergistic effects between PRMT inhibitors and chemotherapy, as well as DDR-targeting drugs (ref. 89-92). In the discussion, the authors should highlight what is novel in this study compared to the reported studies.

We thank the reviewer for this important comment and fully agree that prior studies have demonstrated the potential of PRMT inhibitors to enhance the efficacy of DNA damage-targeting agents and certain chemotherapies[1-4]. In response to the reviewer’s constructive suggestion, we have now revised the discussion to highlight the novel aspects of our study compared to previously reported findings. Specifically, our work presents several key advances that go beyond prior studies. Below, we would like to emphasize the novelty of our current study as follows:

In the clinic, a strategy termed “induced PARP inhibitor (PARPi) sensitivity by epigenetic modulation” is being evaluated to sensitize homologous recombination (HR)-proficient tumors to PARPi treatments. Together with other groups, we reported that repression of BET activity significantly reduces the expression levels of essential HR genes by inhibiting their super-enhancers[5]. This preclinical discovery is now being assessed in a Phase 1b/2 clinical trial combining the BET inhibitor ZEN-3694 with the PARPi talazoparib for the treatment of patients with metastatic triple-negative breast cancer (TNBC) who do not carry germline BRCA1/2 mutations. Promising anti-tumor activity has been observed in this ongoing trial[6]. Importantly, gene expression profiles from paired tumor biopsies demonstrated robust target engagement, evidenced by repression of BRCA1 and RAD51 mRNA expression, consistent with our preclinical findings in xenograft models. Based on these encouraging results, the trial is being expanded to a Phase 2b stage to enroll additional TNBC patients. Moreover, other combination strategies[7-13] based on this “induced PARPi sensitivity by epigenetic modulation” approach have also shown promising clinical responses in both intrinsic and acquired HR-proficient settings. Notably, these clinical studies indicate that the strategy is well-tolerated, likely due to cancer cells being particularly sensitive to epigenetic repression of DNA damage response (DDR) genes, compared with normal cells.

However, two key clinical challenges remain for broader application of this strategy in oncology: 1) which clinically actionable epigenetic drugs can produce the strongest synergistic effects with PARPi? and 2) can a BRCA-independent approach be developed? To address these questions, we performed a drug screen combining the FDA-approved PARPi olaparib with a panel of clinically relevant epigenetic drugs. This panel includes 74 well-characterized epigenetic modulators targeting five major classes of epigenetic enzymes, comprising 7 FDA-approved drugs, 14 agents in clinical trials, and 54 in preclinical development. Notably, both type I PRMT inhibitors (PRMTi) and PRMT5 inhibitors (PRMT5i) achieved high combination and clinical prioritization scores in the screen. Functional assays demonstrated that PRMT inhibition markedly enhances PARPi-induced DNA damage in HR-proficient cancer cell lines. In line with a strong positive correlation between PRMT and DDR gene expression across primary tumors, we observed that PRMT activity supports the transcription of DDR genes and maintains a BRCAness-like phenotype in cancer cells. These findings provide strong rationale for clinical development of PRMT/PARPi combinations in patients with HR-proficient ovarian or breast cancers. Mechanistic characterization from our study further supports PRMTi clinical development by elucidating mechanisms of action, identifying rational combinations, defining predictive biomarkers, and guiding dosing strategies.

We believe our studies will be of significant interest to the cancer research community for several reasons. First, they address major clinical challenges in women’s cancers, specifically, high-grade serous ovarian cancer (HGSOC) and TNBC, both of which are aggressive malignancies with limited therapeutic options. Second, they offer a novel solution to overcome PARPi resistance. Our earlier discovery of “induced PARPi sensitivity by epigenetic modulation” has already shown promising clinical results and represents a new path to overcome both primary and acquired resistance to PARPi and platinum therapies. Third, they focus on a clinically translatable drug class. Selective and potent PRMT inhibitors have been developed by leading pharmaceutical companies, with more than ten currently in advanced clinical trials. Fourth, they support mechanism-driven combination strategies. Preclinical evaluation of PRMTi-based combinations with other therapeutic agents is urgently needed for future clinical success. Finally, our work highlights understudied but therapeutically relevant mechanisms in cancer biology. In-depth mechanistic analysis of the PRMT regulome is essential, and our studies provide important new insights into how PRMTs regulate transcription, RNA splicing, DNA damage repair, and anti-tumor immune responses in the context of HGSOC and TNBC.

In summary, our study identifies PRMT1 and PRMT5 as key epigenetic regulators of DNA damage repair and shows that their inhibition sensitizes HR-proficient tumors to PARP inhibitors by repressing transcription and altering splicing of BRCAness genes. Distinct from prior strategies, dual inhibition of type I PRMT and PRMT5 exhibits strong synergy, allowing for lower-dose combination treatments that may reduce toxicity. Our findings also nominate ERCC1 as a potential predictive biomarker and suggest that MYC-driven tumors may be particularly responsive to this approach. Collectively, these results offer a mechanistic rationale and translational framework to broaden the clinical application of PARP inhibitors.

(2) In Figures 3H-J, MYC targets were likely to correlate with the expression levels of PRMT1/PRMT5 in various public datasets, supporting previous reports that the Myc-PRMT loop plays critical roles during tumorigenesis (ref. 45). "Myc-targets" signatures were also the most significant signatures correlated with the expression of PRMT1 and PRMT5. The authors suggest that under MYC-hyperactivated conditions, tumors may be extremely sensitive to PRMT inhibitors or PRMTi/PARPi combination. However, the underlying mechanism remains unclear.

We sincerely thank the reviewer for the critical and insightful comments. We fully agree that more direct evidence is needed to establish the regulatory relationship between MYC and PRMT1/5. To investigate the effect of c-Myc on PRMT1 and PRMT5 expression, we analyzed RNA-seq data from P493-6 Burkitt lymphoma cells, which harbor a tetracycline (Tet)-repressible MYC transgene. In this system, MYC expression can be suppressed to very low levels and then reactivated, enabling a gradual increase in c-Myc protein levels[14]. Upon Tet removal to induce MYC expression, we observed a robust upregulation of both PRMT1 (4.3-fold) and PRMT5 (3.6-fold) RNA levels within 24 hours, as measured by RNA-seq. These findings indicate that MYC activation can transcriptionally upregulate PRMT1 and PRMT5. To determine whether this regulation is directly driven by MYC, we further analyzed MYC ChIP-seq profiles from the same cell line following 24 hours of MYC induction. Consistently, we observed remarkably increased MYC binding at the promoter regions of both PRMT1 and PRMT5 genes. Interestingly, MYC’s regulatory influence was not limited to PRMT1 and PRMT5, we also observed transcriptional upregulation of other PRMT family members, including PRMT3, PRMT4, and PRMT6, in response to MYC activation. Together with the data presented in Figure 3H, these new results strongly suggest that MYC directly upregulates the expression of PRMT family genes by binding to their promoter regions. Consequently, increased PRMT expression may facilitate MYC’s regulation of target gene expression and splicing in cancer cells. In cancers with MYC hyperactivation, this feed-forward loop may be amplified, creating a potential therapeutic vulnerability. In response to the reviewer’s insightful suggestion, we have further explored how MYC regulates PRMT1/5 and whether this regulation modulates the efficacy of PRMT inhibitors in oncology. These unpublished observations are currently being prepared for a separate manuscript, and we have now incorporated a discussion of these unpublished findings into the revised version of this manuscript. We thank the reviewer again for the thoughtful and constructive comments regarding the MYC–PRMT regulatory axis.

(3) In Figure 5F, ERCC1 expression was unlikely to be reduced in cells treated with GSK025, especially in OVCAR8 cells, although other cells, including TNBC cells, are dramatically changed after treatment.

We sincerely thank the reviewer for the critical and insightful comments. We agree with the reviewer that in Figure 5F, although GSK025 treatment reduced ERCC1 expression, the loading control Tubulin also showed a notable decrease in the OVCAR8 cell line. This may be because Tubulin expression is not specifically affected by the chemical inhibitor GSK025 in this particular cell line, or it may be secondarily reduced as a consequence of PRMT inhibitor-induced cell death. As the reviewer pointed out, this phenomenon was not observed in other cell lines, suggesting that the effect on Tubulin is not specific to PRMT inhibition. To further investigate, we employed CRISPR/Cas9-mediated knockout of PRMT1 or PRMT5 in OVCAR8 cells, a more specific genetic approach to inhibit PRMT activity. In both cases, ERCC1 expression was significantly reduced, whereas Tubulin levels remained stable (Figure 5G). These results support the conclusion that PRMT1 and PRMT5 specifically regulate ERCC1 expression in OVCAR8 cells. The inconsistent effect on Tubulin is likely due to nonspecific cellular responses to chemical inhibition, which are generally more variable and less precise than those induced by genetic perturbation.

(4) In Figure 7H-L, MDA-MB-231 cells were implanted subcutaneously in nude immunodeficient mice to confirm the synergistic therapeutic action of the PRMTi/PARPi combination in vivo. Although PRMT inhibition activates intrinsic innate immune pathways in cancer cells, suggesting that PRMTi treatments may enhance intrinsic immune reactions in tumor cells, the use of nude immune deficient mice means that changes in the tumor immune microenvironment remain unknown.

We sincerely thank the reviewer for the critical and insightful comments. We fully agree with the reviewer that our in vivo experiments using the human cancer cell line MDA-MB-231 in immunodeficient nude mice limit our ability to assess changes in the tumor immune microenvironment. We thank the reviewer for highlighting this important limitation. While the primary goal of the current study was to investigate the therapeutic synergy between PRMT inhibition and PARP inhibition in cancer cells, we would like to take this opportunity to share additional unpublished data that further support and extend the reviewer’s point regarding the immunomodulatory effects of PRMT inhibitors. In syngeneic mouse tumor models, we have observed that the combination of PRMT inhibition and PARP inhibition leads to a more robust anti-tumor immune response compared to either treatment alone. Specifically, we found increased infiltration of CD8⁺ cytotoxic T cells within the tumor microenvironment, suggesting enhanced immune activation and tumor immunogenicity. Furthermore, we have also obtained preliminary evidence that PRMT inhibition can potentiate immune checkpoint blockade therapy. Mechanistically, this may be mediated through the activation of the STING1 pathway and the upregulation of splicing-derived neoantigens, both of which have been implicated in promoting tumor immune visibility. These findings indicate that beyond enhancing DNA damage response, PRMT inhibition may have a broader impact on tumor-immune interactions and could serve as a promising strategy to sensitize tumors to immunotherapy. A separate manuscript detailing these results is currently in preparation and will be submitted for publication as an independent research article. In light of the reviewer’s thoughtful suggestions and in consideration of feedback from Reviewer #2, who recommended removing Figure 6 from the manuscript, we have carefully reevaluated the overall organization of the manuscript. Given the scope and focus of the current work, as well as the desire to maintain a concise and coherent narrative, we decided to move the content originally presented in Figure 6 to the supplementary materials. This figure is now included as Supplementary Figure S5 in the revised version of the manuscript. We believe this change helps streamline the main text while still making the additional data available for interested readers.

(5) In Figures 6-7, a two-tailed Student's t-test was used to determine the statistical differences among multiple comparisons, which should be performed by one-way ANOVA followed by a post hoc test.

We thank the reviewer for this thoughtful and important comment regarding the choice of statistical method. We fully agree with the reviewer that one-way ANOVA followed by a post hoc test is one of the standard approaches for multiple group comparisons. In response to the suggestion, we have performed one-way ANOVA on our data and found that the statistical conclusions are consistent with those obtained from the two-tailed Student’s t-tests. For example, in the first panel of Figure 6A (OVCAR8 treated with GSK715), one-way ANOVA (p = 1.1 × 10^-6), followed by Tukey’s HSD test, confirmed significant differences between control and Olaparib (p = 0.000165), control and GSK715 (p = 0.000145), control and combination (p = 6.067 × 10^-7), Olaparib and combination (p = 0.0003523), and GSK715 and combination (p = 0.0004015), consistent with the conclusions from the two-tailed t-test shown in Figure 6H. Additionally, we would like to explain why two-tailed Student’s t-tests were used in our current study. When comparisons are predefined and conducted pairwise (i.e., two groups at a time), a two-tailed Student’s t-test is statistically equivalent to one-way ANOVA for those comparisons. In our study, each comparison involved only two groups, and we therefore chose t-tests for hypothesis-driven, specific comparisons rather than exploratory multiple testing. This approach aligns with valid statistical principles. All statistical analyses presented in Figures 6-7 were designed to evaluate specific, biologically meaningful comparisons (e.g., treatment vs. control or treatment A vs treatment B). The study was hypothesis-driven, not exploratory, and did not involve simultaneous comparisons across multiple groups. In such cases, the t-test provides a more direct and interpretable result for targeted comparisons. The use of Student’s t-tests reflects the focused nature of the analysis, where each test directly addresses a specific biological question rather than a global group comparison. We sincerely appreciate the reviewer’s thoughtful comments on the statistical methods.

Reviewer #2 (Recommendations for the authors):

(1) If the authors kept the tumors of various sizes in Figure 7I, it would be important to assess the protein and/or mRNA level of ERCC1 to further support their mechanism.

We sincerely thank the reviewer for the insightful comments. We fully agree that evaluating ERCC1 expression in drug-treated tumor samples is critical to support the proposed mechanism. Due to the limited volume of tumor specimens and extensive necrosis observed after three weeks of treatment in the condition used for Figure 7I, we were unable to obtain sufficient material for expression analysis in the original cohort. To address this, we conducted an additional experiment using xenograft-bearing mice (MDA-MB-231 model), initiating treatment when tumors reached approximately 200 mm³ to ensure adequate tissue collection. We also shortened the treatment duration to 7 days to assess early molecular responses to therapy, rather than downstream effects. Consistent with our in vitro results, both GSK715 and GSK025 significantly reduced ERCC1 RNA expression (0.79 ± 0.17, p = 0.03; 0.82 ± 0.11, p = 0.02, respectively), and the combination treatment further decreased ERCC1 expression (0.49 ± 0.20, p = 0.0003), as determined by qRT-PCR. A two-tailed Student’s t-test was used for statistical analysis. In this experiment, we used the same dosing regimen as in the three-week treatment shown in Figure 7I. Importantly, the shorter treatment period and moderate tumor size at treatment initiation minimized necrosis and did not significantly affect tumor growth, allowing for reliable molecular evaluation. We sincerely thank the reviewer for highlighting this important point.

(2) Figure 2G: please explain why two bands remain for sgPRMT1.

We greatly appreciate the reviewer for raising this insightful and important question. As the reviewer pointed out, an additional band appeared after PRMT1 knockdown in OVCAR8 cells using two sequence-independent gRNAs. Notably, this band was not observed in MDA-MB-231 cells. The antibody used to detect PRMT1 (clone A33, #2449, Cell Signaling Technology) is widely adopted in PRMT1 research, with over 65 citations supporting its specificity. Interestingly, previous studies[15] have identified seven PRMT1 isoforms (v1–v7), generated through alternative splicing and exhibiting tissue-specific expression patterns. Of these, three isoforms are detectable using the A33 antibody. We believe the additional band observed upon sgRNA treatment likely represents a PRMT1 isoform that is normally expressed at low levels in OVCAR8 cells. Upon knockdown of the major isoforms by CRISPR/Cas9, expression of this minor isoform may have increased as part of a compensatory feedback mechanism, rendering it detectable by immunoblotting. Because PRMT1 isoform expression is largely tissue-type specific, it is not surprising that the same band was absent in MDA-MB-231 cells, which are derived from a different lineage than OVCAR8 cells. The reviewer raised an important question regarding the role of PRMT1 isoforms in regulating DNA damage response in cancer. We agree this is an intriguing direction and will investigate it further in future studies.

(3) Figure 4D: Please correct the figure legend so the description matches the color in the figure. Red and blue are absent.

We sincerely thank the reviewer for the critical and insightful comments. The figure legend for Figure 4D has been corrected in the revised version of the manuscript to accurately match the colors shown in the figure. We thank the reviewer for pointing out this issue.

(4) Figure 7A and B: please indicate the cell lines used.

We sincerely thank the reviewer for the critical and insightful comments. In Figure 7A and 7B, human embryonic kidney 293T (HEK293T) cells were used due to their high transfection efficiency and widespread application in reporter assays. This information has been incorporated into the figure legend for Figures 7A and 7B.

(5) What is the link with ERCC1 splicing because reduced overall ERCC1 expression is clear?

We sincerely thank the reviewer for the critical and insightful comments. As the reviewer pointed out, although the direct impact of ERCC1 alternative splicing on its protein expression remains to be fully elucidated, it is likely that PRMT inhibition induces aberrant splicing events that result in the production of alternative ERCC1 isoforms with impaired or altered function. These splicing changes may compromise ERCC1’s role in DNA repair pathways. Furthermore, as shown in Figure 4G, we observed a reduction in the total ERCC1 mRNA reads following PRMTi treatment. This decrease may be attributed, at least in part, to the instability of the alternatively spliced ERCC1 transcripts, which could be more prone to degradation. In combination with the transcriptional downregulation of ERCC1 induced by PRMT inhibition, these alternative splicing events may lead to a further reduction in functional ERCC1 protein levels. This dual impact on ERCC1 expression, through both decreased transcription and the generation of unstable or non-functional isoforms, likely contributes to the enhanced cellular sensitivity to PARP inhibitors observed in our study. We believe this represents an important mechanistic insight into how PRMT inhibition modulates the DNA damage response in cancer cells, and further studies are warranted to investigate the precise role of ERCC1 splicing regulation in this context. We thank the reviewer for pointing out this interesting future research direction.

(6) Figure 7J: From the graph, it seems like Olaparib+G715 and G715+G025 have a similar effect on tumor volume (two curves overlap). Please discuss.

We sincerely thank the reviewer for the critical and insightful comments. In the current study, the doses used for single-agent treatments were selected based on prior publications. For example, the dose of GSK715 was guided by a recent study from the GSK group[16]. Our in vitro and in vivo findings, together with previously published data, consistently demonstrate that GSK715 is more potent than both GSK025 and Olaparib. Notably, treatment with GSK715 alone led to significantly greater inhibition of tumor growth compared to either GSK025 or Olaparib administered individually. This higher potency of GSK715 also explains the comparable levels of tumor suppression observed in the combination groups, including GSK715 plus Olaparib and GSK715 plus GSK025. These results suggest that GSK715 is likely the primary driver of efficacy in the two drug combination settings. Importantly, this observation provides a valuable opportunity to further refine and optimize the dosing strategy for GSK715. Specifically, because GSK715 is highly potent, its dose may be reduced when used in combination regimens without compromising therapeutic efficacy. This approach could significantly improve the safety profile of GSK715 by minimizing potential dose-related toxicities, thereby enhancing its suitability for future clinical development in combination therapy contexts.

(7) Discussion: "PRMT5i increased global sDMA levels"-> "... aDMA levels.".

We sincerely thank the reviewer for the critical and insightful comments. In response, we have corrected the sentence in the discussion from “PRMT5i increased global sDMA levels, which suggested that type I PRMT and PRMT5 share a substrate (i.e., MMA) and/or their functions are compensatory” to “PRMT1i increased global sDMA levels, which suggested that type I PRMT and PRMT5 share a substrate (i.e., MMA) and/or their functions are compensatory.” We apologize for the misstatement and have corrected this error in the revised version of the manuscript.

(8) In addition to the methods, add that nude mice were used in the body of the results and the figure legend for Figure 7J.

We sincerely thank the reviewer for the critical and insightful comments. In the revised version of the manuscript, we have added that immunodeficient nude mice were used in both the body of the Results section and the figure legend for Figure 7J, in addition to the Methods section. We thank the reviewer for this helpful suggestion.

(9) Figure 6 can be deleted to focus the manuscript. It does not add to the PARP inhibition story, but only suggests a link to immunotherapy where this has been reported previously PMID: 35578032 and 32641491.

We sincerely thank the reviewer for the critical and insightful comments. Reviewer #1 also raised a related concern regarding the relevance of this section to the main focus of the manuscript. In consideration of both reviewers’ comments, we have decided to move the data previously shown in Figure 6 to the supplementary section as Supplementary Figure S5. This revision allows us to streamline the main text and maintain a clear focus on the core findings related to PARP inhibition. At the same time, we believe the immunotherapy-related observation may still be of interest to some readers. By presenting these results in the supplementary materials, we ensure that this potentially relevant link remains accessible without distracting from the primary narrative of the manuscript. We greatly appreciate the reviewers’ guidance in helping us improve the clarity and focus of our work. We thank the reviewer for the thoughtful suggestion.

References

(1) Dominici, C., et al. Synergistic effects of type I PRMT and PARP inhibitors against non-small cell lung cancer cells. Clin Epigenetics 13, 54 (2021).

(2) O'Brien, S., et al. Inhibiting PRMT5 induces DNA damage and increases anti-proliferative activity of Niraparib, a PARP inhibitor, in models of breast and ovarian cancer. BMC Cancer 23, 775 (2023).

(3) Carter, J., et al. PRMT5 Inhibitors Regulate DNA Damage Repair Pathways in Cancer Cells and Improve Response to PARP Inhibition and Chemotherapies. Cancer Res Commun 3, 2233-2243 (2023).

(4) Li, Y., et al. PRMT blockade induces defective DNA replication stress response and synergizes with PARP inhibition. Cell Rep Med 4, 101326 (2023).

(5) Yang, L., et al. Repression of BET activity sensitizes homologous recombination-proficient cancers to PARP inhibition. Sci Transl Med 9(2017).

(6) Aftimos, P.G., et al. A phase 1b/2 study of the BET inhibitor ZEN-3694 in combination with talazoparib for treatment of patients with TNBC without gBRCA1/2 mutations. Journal of Clinical Oncology 40, 1023-1023 (2022).

(7) Karakashev, S., et al. BET Bromodomain Inhibition Synergizes with PARP Inhibitor in Epithelial Ovarian Cancer. Cell Rep 21, 3398-3405 (2017).

(8) Sun, C., et al. BRD4 Inhibition Is Synthetic Lethal with PARP Inhibitors through the Induction of Homologous Recombination Deficiency. Cancer Cell 33, 401-416 e408 (2018).

(9) Johnson, S.F., et al. CDK12 Inhibition Reverses De Novo and Acquired PARP Inhibitor Resistance in BRCA Wild-Type and Mutated Models of Triple-Negative Breast Cancer. Cell Rep 17, 2367-2381 (2016).

(10) Iniguez, A.B., et al. EWS/FLI Confers Tumor Cell Synthetic Lethality to CDK12 Inhibition in Ewing Sarcoma. Cancer Cell 33, 202-216 e206 (2018).

(11) Shan, W., et al. Systematic Characterization of Recurrent Genomic Alterations in Cyclin-Dependent Kinases Reveals Potential Therapeutic Strategies for Cancer Treatment. Cell Rep 32, 107884 (2020).

(12) Muvarak, N.E., et al. Enhancing the Cytotoxic Effects of PARP Inhibitors with DNA Demethylating Agents - A Potential Therapy for Cancer. Cancer Cell 30, 637-650 (2016).

(13) Abbotts, R., et al. DNA methyltransferase inhibitors induce a BRCAness phenotype that sensitizes NSCLC to PARP inhibitor and ionizing radiation. Proc Natl Acad Sci U S A 116, 22609-22618 (2019).

(14) Lin, C.Y., et al. Transcriptional amplification in tumor cells with elevated c-Myc. Cell 151, 56-67 (2012).

(15) Goulet, I., Gauvin, G., Boisvenue, S. & Cote, J. Alternative splicing yields protein arginine methyltransferase 1 isoforms with distinct activity, substrate specificity, and subcellular localization. J Biol Chem 282, 33009-33021 (2007).

(16) Fedoriw, A., et al. Anti-tumor Activity of the Type I PRMT Inhibitor, GSK3368715, Synergizes with PRMT5 Inhibition through MTAP Loss. Cancer Cell 36, 100-114 e125 (2019).