An Inducible BRCA1 Expression System with In Vivo Applicability Uncovers Activity of the Combination of ATR and PARP Inhibitors to Overcome Therapy Resistance

Round 1

Reviewer 1 Report (New Reviewer)

Comments and Suggestions for Authors

In this manuscript, the authors developed a doxycycline-inducible system to express a series of BRCA1 mutants into the BRCA1-deficient MDA-MB-436 cell line. They used them for functional analysis of different BRCA1 hypomorphs to find that the ∆exon 11 mutant conferred resistance to PARPi and carboplatin. In addition, they found that combination of PARPi and an ATR inhibitor abrogated resistance against PARPi, suggesting that such a combination would be effective in treating malignancies with the BRCA1 ∆exon 11 mutant.

 

Overall, this manuscript is well written, and its conclusions are supported by their data. This reviewer endorses its publication as it is.

Author Response

We thank the reviewer for their endorsement.

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

Please refer to the following comments for the revision:

1. The MDAMB436 cell line utilized in this investigation possesses a homozygous deletion of the homologous recombination repair (HRR) gene RAD51B and necessitates dual complementation with RAD51B and BRCA1 to attain complete resistance. In this context, how might the presence of secondary HRR-related alterations in clinical specimens influence the generalizability of the resistance mechanism mediated by the Delta exon 11 variant observed in your model?
2. The data presented indicate a significant correlation between the extent of the exon 11 deletion and the capacity of BRCA1 to interact with its partners, such as PALB2 and BARD. Could you provide further insights into whether exon 11 serves a structural or "chaperone-like" function in stabilizing the BRCA1-PALB2-BARD1 complex, particularly given its prior characterization as a predominantly unstructured domain?
3. The findings of this study reveal that the Delta exon 11 variant of BRCA1 confers "full" resistance to carboplatin—comparable to that provided by full-length BRCA1—while only offering "partial" resistance to olaparib. Could you elucidate the mechanistic rationale behind the observation that the homologous recombination repair (HRR) threshold necessary for platinum resistance appears to be lower than that required for resistance to PARP inhibitors in this hypomorphic context?
4. Given that resistance in your inducible system was strictly dose-dependent and necessitated substantial overexpression of the Delta exon 11 variant to achieve statistical significance, how do the protein levels attained under your "high-dox" conditions quantitatively compare to the endogenous levels of Delta exon 11 typically observed in clinical samples or patient-derived xenograft (PDX) models of treatment-resistant breast cancer?
5. Your results indicate that cells expressing the Delta exon 11 variant exhibit a pronounced "specific reliance" on ATR activity for RAD51 foci formation, significantly more so than cells expressing full-length BRCA1. What specific downstream targets or phosphorylation events mediated by ATR are hypothesized to be critical for the limited HRR functionality of the Delta exon 11 isoform?
6. As illustrated in Figure 3A, tumors complemented with full-length BRCA1 demonstrated a markedly accelerated baseline growth rate compared to the parental BRCA1-mutant MDAMB436 cells, even in the absence of treatment. Does this observation imply that BRCA1 restoration confers a fitness or proliferative advantage independent of its role in DNA repair? Furthermore, how might these divergent growth rates complicate the interpretation of Tumor Growth Inhibition (TGI) percentages?

Author Response

• The MDAMB436 cell line utilized in this investigation possesses a homozygous deletion of the homologous recombination repair (HRR) gene RAD51B and necessitates dual complementation with RAD51B and BRCA1 to attain complete resistance. In this context, how might the presence of secondary HRR-related alterations in clinical specimens influence the generalizability of the resistance mechanism mediated by the Delta exon 11 variant observed in your model?

We thank the reviewer for this question. Co-occurrence of deleterious mutations in more than one HRR gene in the same tumour are very rare events (see for example https://doi.org/10.1200/PO.22.00258). As such, we believe that results in our RAD51B-complemented model system, which effectively carries a single deleterious HRR mutation (in BRCA1) thus representing the more general form of HRR deficiency observed in the clinic, are generalizable to the wider patient population carrying BRCA1 mutations in the exon 11 of the gene.

 

• The data presented indicate a significant correlation between the extent of the exon 11 deletion and the capacity of BRCA1 to interact with its partners, such as PALB2 and BARD. Could you provide further insights into whether exon 11 serves a structural or "chaperone-like" function in stabilizing the BRCA1-PALB2-BARD1 complex, particularly given its prior characterization as a predominantly unstructured domain?

We thank the reviewer for this comment, which we have now acknowledged in the Conclusions section by adding the following text and reference: “In addition, our analyses point towards the actual length of the deletion inside the exon 11 of BRCA1, rather than any specific regions encoded within, as the main determinant of the hypomorphic nature of these variants, describing a new potential function for this predominantly unstructured domain [32]”. Lines 582-583.

 

• The findings of this study reveal that the Delta exon 11 variant of BRCA1 confers "full" resistance to carboplatin—comparable to that provided by full-length BRCA1—while only offering "partial" resistance to olaparib. Could you elucidate the mechanistic rationale behind the observation that the homologous recombination repair (HRR) threshold necessary for platinum resistance appears to be lower than that required for resistance to PARP inhibitors in this hypomorphic context?

We can only speculate about the reason, which we already do in the Conclusions section: “The apparent different behaviour of the BRCA1 ∆exon11 hypomorph versus the full length protein between the olaparib (partial resistance) and carboplatin (full resistance) experiments may be driven by the reduced therapeutic index of carboplatin between fully HRR proficient and partially proficient settings”. Lines 598-602. While olaparib is a targeted drug that requires HRR deficiency to be fully effective, cisplatin is a DNA-damaging chemotherapy that is efficacious even in HRR proficient settings, which may explain the difference.

 

• Given that resistance in your inducible system was strictly dose-dependent and necessitated substantial overexpression of the Delta exon 11 variant to achieve statistical significance, how do the protein levels attained under your "high-dox" conditions quantitatively compare to the endogenous levels of Delta exon 11 typically observed in clinical samples or patient-derived xenograft (PDX) models of treatment-resistant breast cancer?

This is a very interesting question that cannot be addressed right now, as it would require a quantitative assay to measure BRCA1 protein levels in tumour samples that is currently not available.

 

• Your results indicate that cells expressing the Delta exon 11 variant exhibit a pronounced "specific reliance" on ATR activity for RAD51 foci formation, significantly more so than cells expressing full-length BRCA1. What specific downstream targets or phosphorylation events mediated by ATR are hypothesized to be critical for the limited HRR functionality of the Delta exon 11 isoform?

As the reviewer is probably aware, ATR (and/or ATM) can phosphorylate a substantial number of targets involved in HRR, including over a dozen “S/TQ” consensus sites in BRCA1 itself, making it difficult to pinpoint which phosphorylation events could be the most critical ones. Following this suggestion, we have now included the following text and reference in the Conclusions section: “The reason why the BRCA1 ∆exon11 hypomorph is specifically reliant on ATR activity to exert its HRR function is unclear, but it is important to note that there is a cluster of a dozen ATM/ATR phosphorylation sites on BRCA1 in the PALB2 interacting region located outside the exon 11 of the gene [36]”. Lines 625-629.

 

• As illustrated in Figure 3A, tumors complemented with full-length BRCA1 demonstrated a markedly accelerated baseline growth rate compared to the parental BRCA1-mutant MDAMB436 cells, even in the absence of treatment. Does this observation imply that BRCA1 restoration confers a fitness or proliferative advantage independent of its role in DNA repair? Furthermore, how might these divergent growth rates complicate the interpretation of Tumor Growth Inhibition (TGI) percentages?

We agree with the reviewer that the implication is that BRCA1 restoration confers a proliferative advantage to MDAMB436 cells, but we have no reason to believe this is independent of the role of BRCA1 in DNA repair: it could reflect the need for BRCA1 functions in dealing with endogenous DNA damage. The different proliferation rate of these cells does not complicate the interpretation of TGI as these values are always represented relative to the starting tumour volume – essentially, providing an internal control for the proliferation rate. In addition, we never compare treatment effects between the different cell lines – all the in vivo studies are independent experiments.

Reviewer 3 Report (New Reviewer)

Comments and Suggestions for Authors

1. Beyond the intrinsic qualities of the triple-negative MDA-MB-436 cell line as a model for BRCA1 mutation, why was this specific cellular model utilised instead of alternatives, such as the SUM149PT line which harbours a mutation in exon 11? Please explain.
2. Regarding the generation of the inducible stable MDA-MB-436 + BRCA1 cell lines: how many clones were screened prior to selecting the specific clone utilised in this manuscript?
3. Furthermore, the control system—lacking BRCA1 expression—is not shown. This is a concern, given that the integration of the inducible stable system into the genome is stochastic; such insertions can occur at any locus, potentially yielding results that are not truly representative of the parental control.
4. Owing to the use of inducible expression systems, the study is difficult to follow and several points remain ambiguous. Specifically, it is not clearly established whether the doxycycline-inducible cells express wild-type (WT) or mutant BRCA1; the manuscript lacks a sufficiently rigorous explanation of this distinction.
5. How is the correlation established between BRCA mutations and the interaction with HCC-associated proteins (BARD1, PALB2, BRCT, BRIP1 and CtIP), as well as their implications for drug resistance? It remains unclear why these specific interactions were not further evaluated within the MDA-MB-436 cell line. This aspect of the study requires further elucidation.
6. What is the justification for utilising HEK293T cells transfected with FLAG-tagged BRCA1 constructs, rather than using tumour cell lines where BRCA variants could interact with endogenous proteins in a more physiologically relevant context?
7. What was the rationale for the specific carboplatin concentrations utilised in the in vitro assays? Furthermore, what served as the basis for the olaparib dosages administered in the in vivo study? Finally, why was carboplatin excluded from the in vivo experimental arm?

Author Response

• Beyond the intrinsic qualities of the triple-negative MDA-MB-436 cell line as a model for BRCA1 mutation, why was this specific cellular model utilised instead of alternatives, such as the SUM149PT line which harbours a mutation in exon 11? Please explain.

The main reason to not use SUM149PT cells was the fact that they express the BRCA1 exon 11 hypomorph, which makes it a not-so-clean system compared to MDAMB436 cells, which do not express any detectable BRCA1 protein. See for example https://doi.org/10.1158/0008-5472.CAN-16-0186.

 

• Regarding the generation of the inducible stable MDA-MB-436 + BRCA1 cell lines: how many clones were screened prior to selecting the specific clone utilised in this manuscript?

We thank the reviewer for this comment and apologise for not making it clearer in our initial submission. None of the cell lines was clonally selected and we always worked with polyclonal populations to avoid unwanted selection pressures. We have now included a sentence in the Materials and methods section to clarify this: “Cells were then expanded in the presence of antibiotic and used for the experiments described here as polyclonal populations to avoid clonal selection issues”. Line 153.

 

• Furthermore, the control system—lacking BRCA1 expression—is not shown. This is a concern, given that the integration of the inducible stable system into the genome is stochastic; such insertions can occur at any locus, potentially yielding results that are not truly representative of the parental control.

We apologise but cannot fully understand the reviewer’s comment. All the in vitro experiments reported show results in our control cell lines (MDAMB436-TR + LacZ or MDAMB436-B-TR + LacZ), which lack BRCA1 expression. Furthermore, the IC50 values for olaparib in the MDAMB436-TR + LacZ cell line (4.3 nM) are similar to the values obtained in the parental MDAMB436 cell line (10 nM; AZ internal data not shown).

 

• Owing to the use of inducible expression systems, the study is difficult to follow and several points remain ambiguous. Specifically, it is not clearly established whether the doxycycline-inducible cells express wild-type (WT) or mutant BRCA1; the manuscript lacks a sufficiently rigorous explanation of this distinction.

We apologise if it was not completely clear in the manuscript that full length (FL) BRCA1 protein refers to the wild-type version of it. We have now included the following text when the FL protein is firstly described in the Results section to help the reader: “All constructs carrying point mutations were expressed at similar levels to the wild-type, full length (FL) BRCA1 protein…”. Line 306.

 

• How is the correlation established between BRCA mutations and the interaction with HCC-associated proteins (BARD1, PALB2, BRCT, BRIP1 and CtIP), as well as their implications for drug resistance? It remains unclear why these specific interactions were not further evaluated within the MDA-MB-436 cell line. This aspect of the study requires further elucidation.

The interaction studies were performed to confirm that the mutations introduced in our constructs behaved as previously described, as their effect is well known and independent of the cellular background (see for example doi:10.7554/eLife.21350). For this reason, we believe no additional insights will be gained by repeating the immunoprecipitation experiments in the MDAMB436 cell line.

 

• What is the justification for utilising HEK293T cells transfected with FLAG-tagged BRCA1 constructs, rather than using tumour cell lines where BRCA variants could interact with endogenous proteins in a more physiologically relevant context?

We used HEK293T cells as the quickest way to assess the utility of our constructs, given their easily transfectable nature and their known capacity to produce high protein levels. We do not fully understand the reviewer’s comment regarding the physiologically relevant context, as the interactions between BRCA1 and its protein partners is not exclusive to tumour cell lines (see for example https://doi.org/10.1038/s41467-021-25346-4).

 

• What was the rationale for the specific carboplatin concentrations utilised in the in vitro assays? Furthermore, what served as the basis for the olaparib dosages administered in the in vivo study? Finally, why was carboplatin excluded from the in vivo experimental arm?

We assessed the anti-proliferative activity of carboplatin in our MDAMB436-BTR + LacZ cell line to define an IC50 for it, with a value around 5 μM. Complete inhibition of proliferation was achieved at 10 μM carboplatin, and as such we defined a dose range covering up to 10 μM carboplatin. The doses of olaparib used in vivo are based on previous experience, with 100 mg/kg QD being clinically relevant when the exposure is compared to the 300 mg BID clinical dose (see for example https://doi.org/10.1158/1078-0432.CCR-22-0301), as stated in the Results section of the manuscript: “…the parental MDAMB436 cell line showed exquisite sensitivity to olaparib in a dose-dependent manner, with the top dose of olaparib used (the clinically relevant 100 mg/kg dose) causing complete tumour regressions (Figure 3C)”. Lines 456-458. We did not perform any experiments with carboplatin in vivo as our main goal was to identify drug combinations to overcome resistance to PARPi.

Round 2

Reviewer 3 Report (New Reviewer)

Comments and Suggestions for Authors

The answers are correct. Acceptance of the manuscripts is suggested.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Irving and colleagues test PARPi resistance for various patient-relevant mutations in BRCA1 and identify BRCA1(delta)exon11 as a hypomorphic allele with partial resistance to olaparib in vitro and in vivo. This resistance can be overcome through combination ATR inhibitor (ceralasertib) treatment. This study demonstrates the importance of precision modelling cancer variants and functionally testing specific mutations in well-known tumor suppressors like BRCA1. This reviewer’s comments focus on clarifications of certain interpretations and suggestions to increase impact.

1. Figure 1 - Why do un-transfected HEK293 cells not express BRCA1? Or are Western blots in panel B-C against FLAG protein (comment regarding antibody epitope on the bottom of page 8 makes this reviewer think that the blots are anti-BRCA1 as indicated)?
2. In Figure 1, could it be more relevant to show protein expression in the MDAMB436 cell line tested for PARPi resistance in this study (or alongside HEK293 cells)?
3. If protein expression in the BRCA1 model cell lines is higher than untransfected (and therefore undetectable in untransfected HEK293 cells at the exposure shown in Fig1), could over-expression induce artificial sensitivities? How close does this expression re-capitulate endogenous BRCA1 expression in wild-type and mutant cells? This could be commented on in the discussion.
4. While resistance to PARPi with several variants are described as “minimal”, many are shown in Fig2B to display increased IC50 with high significance. It may be worth commenting in the results section on what this might mean clinically and or indicate the minimal free concentration in patient plasma (mentioned in the discussion) in Fig2 and/or the figure legend?
5. The authors observe dose-response behaviours with respect to the BRCA1 exon11 deletion variant and this variant appears to be expressed more highly that full-length. If the authors boost expression of the RING-less and/or BRCT-less variants, is increased resistance observed here as well? What could this mean for patients with amplifications? This could be discussed more in the discussion section.
6. Along these lines, as BRCA1 is a regulator of the cell cycle, could the rate of proliferation affect expression levels and/or resistance in vitro or in vivo? Have differential effects on cell growth been assessed and have the authors accounted for this in their assays?
7. Figure 3 - No panel (F) is observed in the figure, while it is described in the figure legend as “Relative FLAG-BRCA1 mRNA expression in tumour samples from individual mice implanted with MDAMB436-BTR cells expressing the full length (FL) or Δexon11 BRCA1 Expression is normalized to that of GAPDH.” It is interesting that mRNA levels may be differentially impacted, suggesting that differentiation expression maybe not be at the level of protein stability. It may be helpful to comment on this in the manuscript and/or cite relevant literature regarding the potential of regulation at the mRNA level.

Minor comment. The authors could consider increasing the size of figure fonts. Axis labels for certain graphs (ex. Fig2C,D, Fig4A heat-map quantifications) are difficult to see.

 

 

Author Response

Comments and Suggestions for Authors

Irving and colleagues test PARPi resistance for various patient-relevant mutations in BRCA1 and identify BRCA1(delta)exon11 as a hypomorphic allele with partial resistance to olaparib in vitro and in vivo. This resistance can be overcome through combination ATR inhibitor (ceralasertib) treatment. This study demonstrates the importance of precision modelling cancer variants and functionally testing specific mutations in well-known tumor suppressors like BRCA1. This reviewer’s comments focus on clarifications of certain interpretations and suggestions to increase impact.

We thank the reviewer for acknowledging the interest of our study, and for the suggestions to improve our manuscript.

1.

Figure 1 - Why do un-transfected HEK293 cells not express BRCA1? Or are Western blots in panel B-C against FLAG protein (comment regarding antibody epitope on the bottom of page 8 makes this reviewer think that the blots are anti-BRCA1 as indicated)?

We apologize for not making this clearer. The reviewer is right in that it is a BRCA1 blot shown in Fig 1B. Un-transfected HEK293 do express BRCA1 but at much lower levels. So, essentially, all the FLAG-BRCA1 constructs are over-expressed compared to the endogenous protein. For clarity, we have included the following in the main text: “All constructs carrying point mutations were expressed at similar levels to the full length (FL) BRCA1 protein and over-expressed compared to the endogenous BRCA1 protein levels (Fig 1B).” Lines 303-305.

For the reviewer’s reference, we attach here 2 longer exposures from the same blot that clearly show expression in the un-transfected cells. We preferred to use the shorter exposure in the figure to make clear that expression of the different mutants was similar in all cases.

 

2.

In Figure 1, could it be more relevant to show protein expression in the MDAMB436 cell line tested for PARPi resistance in this study (or alongside HEK293 cells)?

We thank the reviewer for the comment, but we think showing them alongside is not particularly helpful, as over-expression in HEK293 cells is not doxycycline controlled, while it is in the MDAMB436 system. BRCA1 protein expression in the MDAMB436 cell lines generated in this study is shown in Supplementary Figure 2F.

 

3.

If protein expression in the BRCA1 model cell lines is higher than untransfected (and therefore undetectable in untransfected HEK293 cells at the exposure shown in Fig1), could over-expression induce artificial sensitivities? How close does this expression re-capitulate endogenous BRCA1 expression in wild-type and mutant cells? This could be commented on in the discussion.

We thank the reviewer for the comment, as it is precisely the reason why we developed the system as inducible in the MDAMB436 cell line. No survival assays were performed in HEK293 cells due to this reason – all the efficacy studies were run in MDAMB436 or other relevant cancer cells. We cannot know how close our system is to endogenous BRCA1 expression in MDAMB436 cells, as they are naturally deficient in BRCA1 expression, but the levels of resistance to olaparib caused by expression of the full-length protein at the top doses of doxycycline (with IC50 values close to 4 μM) suggest that this is physiologically relevant – as these are values close to the IC50 range of olaparib in homologous recombination proficient cell lines (see for example Supplementary Figure 2 in https://doi.org/10.1158/1078-0432.CCR-22-0301). To further clarify this, we have added the following sentence in the Results section: “The IC50 value of olaparib in the MDAMB436-B-TR + BRCA1 system is consistent with those in HRR proficient cell lines (23), suggesting that our expression system re-capitulates restoration of BRCA1 expression.”. Lines 365-367.

 

4.

While resistance to PARPi with several variants are described as “minimal”, many are shown in Fig2B to display increased IC50 with high significance. It may be worth commenting in the results section on what this might mean clinically and or indicate the minimal free concentration in patient plasma (mentioned in the discussion) in Fig2 and/or the figure legend?

We thank the reviewer for the suggestion, which we have now implemented in all the panels with olaparib data on Figure 2.

 

5.

The authors observe dose-response behaviours with respect to the BRCA1 exon11 deletion variant and this variant appears to be expressed more highly that full-length. If the authors boost expression of the RING-less and/or BRCT-less variants, is increased resistance observed here as well? What could this mean for patients with amplifications? This could be discussed more in the discussion section.

Our doxycycline-inducible system maxes out expression between 100-500 nM regardless of the BRCA1 protein variant tested (see example with FL and Δexon11 in Fig 2F). That is why we used 1 μM doxycycline on our cell survival experiments, where the RING-less and BRCT-less variants are clearly over-expressed with respect to the FL protein (Supp Fig 2F). We agree with the reviewer that we can make this clearer in the text and we have included the following sentence in the Discussion: “Our data are in accordance with previous reports showing that overexpression of RING or BRCT mutant BRCA1 proteins caused modest resistance to chemotherapy and PARPi both in vitro and in vivo (11,34), even at expression levels higher than those of the FL protein (Supp Fig S2F)”. Lines 595-598.

 

6.

Along these lines, as BRCA1 is a regulator of the cell cycle, could the rate of proliferation affect expression levels and/or resistance in vitro or in vivo? Have differential effects on cell growth been assessed and have the authors accounted for this in their assays?

We thank the reviewer for this comment, as we took this into consideration but did not include the data in the initial submission. We have now produced a new panel in Supplementary Figure S2 (panel 2G) that shows that growth rate and incorporation of the nucleotide analogue EdU are not significantly affected by introduction of the different BRCA1 variants. We have now added a new sentence in the Results section to explain this: “Of note, expression of FL or different hypomorphic variants of BRCA1 did not affect proliferation or DNA replication rates (Figure S2G)”. Lines 428-430.

 

7.

Figure 3 - No panel (F) is observed in the figure, while it is described in the figure legend as “Relative FLAG-BRCA1 mRNA expression in tumour samples from individual mice implanted with MDAMB436-BTR cells expressing the full length (FL) or Δexon11 BRCA1 Expression is normalized to that of GAPDH.” It is interesting that mRNA levels may be differentially impacted, suggesting that differentiation expression maybe not be at the level of protein stability. It may be helpful to comment on this in the manuscript and/or cite relevant literature regarding the potential of regulation at the mRNA level.

We apologize for this oversight, as we uploaded the wrong version of the figure. The right version containing panel F is now included in the manuscript. We agree with the reviewer that it is important to link this finding to previous literature reports. We have now added a new sentence in the Results section and a reference to highlight this: “Interestingly, expression of FLAG-tagged ∆exon11 BRCA1 transcript was significantly higher than that of the FL transcript (Fig 3F), suggesting that increased expression of the BRCA1 ∆exon11 protein could be due to increased mRNA levels, as reported before (14)”. Lines 458-462.

 

Minor comment. The authors could consider increasing the size of figure fonts. Axis labels for certain graphs (ex. Fig2C,D, Fig4A heat-map quantifications) are difficult to see.

We thank the reviewer for this suggestion, that we have now implemented throughout.

Reviewer 2 Report

Comments and Suggestions for Authors

Minor Comments

 

 

• The figure legends should include details of the statistical analyses performed (e.g., tests used, number of replicates, and definition of significance thresholds). For instance, Figures 1E and 2 appear to involve quantitative comparisons but do not specify the statistical test applied. Clearly stating the analysis method (e.g., Pearson correlation, one-way ANOVA with Holm–Sidak correction, p-value criteria) would improve transparency and reproducibility.
• The discussion effectively contextualizes Δexon11 as a hypomorphic variant with dose-dependent resistance properties. The authors also propose an interesting ATR dependency as a synthetic vulnerability. To further strengthen the clinical relevance, it would be valuable to briefly discuss whether ATR inhibition could be a generalizable strategy beyond the Δexon11 context—for example, in other hypomorphic BRCA1 settings or partial HRR-deficient tumors.

Author Response

Comments and Suggestions for Authors

Minor Comments

• The figure legends should include details of the statistical analyses performed (e.g., tests used, number of replicates, and definition of significance thresholds). For instance, Figures 1E and 2 appear to involve quantitative comparisons but do not specify the statistical test applied. Clearly stating the analysis method (e.g., Pearson correlation, one-way ANOVA with Holm–Sidak correction, p-value criteria) would improve transparency and reproducibility.

We apologize for this oversight, which we have now resolved. Statistical analyses performed in Fig 1E are now stated in the figure legend. Number of replicates and statistical analyses, as well as p value criteria, were already included in the legend of Figs 2, 3 and 4.

 

• The discussion effectively contextualizes Δexon11 as a hypomorphic variant with dose-dependent resistance properties. The authors also propose an interesting ATR dependency as a synthetic vulnerability. To further strengthen the clinical relevance, it would be valuable to briefly discuss whether ATR inhibition could be a generalizable strategy beyond the Δexon11 context—for example, in other hypomorphic BRCA1 settings or partial HRR-deficient tumors.

We agree with the reviewer that this is a valuable point to make. We have now included an additional sentence in the Discussion to highlight this: “As the combination of ceralasertib and olaparib is being tested in the clinic (35), supported by pre-clinical data showing activity of this combination in HRR-deficient settings and beyond (2,3), our work provides a supportive mechanistic platform to help explain the benefit of this combination in patients with tumours harbouring mutations in the exon 11 of BRCA1”. Lines: 619-623.

 

Irving and colleagues present the design and validation of a doxycycline-inducible BRCA1 expression system in BRCA1-mutant TNBC cells. Using this system, the authors systematically evaluate distinct BRCA1 hypomorphic variants and demonstrate that expression of the Δexon11 variant partially restores homologous recombination repair (HRR) and confers intermediate resistance to PARP inhibitors and carboplatin. Importantly, the study identifies that co-treatment with the ATR inhibitor ceralasertib effectively overcomes Δexon11-mediated resistance, both in vitro and in vivo.

Overall, this is a well-designed and clearly presented study. The experimental approach is elegant, the in vivo validation adds translational value, and the findings provide mechanistic insight into a clinically relevant form of PARPi resistance. While the work may not be groundbreaking conceptually, it represents a high-quality preclinical contribution suitable for publication in Cancers.

We thank the reviewer for their support.