Review Reports - Identification of Resistance Genes in Breast Cancer Cells Treated with Fulvestrant and Ribociclib via Retroviral Screening and Integration Site Sequencing

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript with ID: cells-4063176 titled “Identification of Resistance Genes in Breast Cancer Cells Treated with Fulvestrant and Ribociclib via Retroviral Screening and Integration Site Sequencing” by Huang, Z.; et al. is a scientific work where the authors sequenced the DNA extracted from breast cancer cells treated with tamoxifen, fulvestrant, ribociclib, abemaciclib or palbociblib therapies to identify resistance biomarkers. This is a topic of growing interest and the manuscript is generally well-written.

However, it exists some points that need to be addressed (please, see them below detailed point-by-point) to improve the scientific quality of the submitted manuscript paper before this article will be consider for its publication in Cells.

1) Introduction. Could the authors provide quantitative data details according to the worldwide global burdens of breast cancer and the associated disability-adjusted life years (DALYs)? This will significantly aid the potential readers to better undestand the significance of this research study.

2) “Understanding the molecular mechanisms underlying therapy resistance is therefore a priority in breast cancer research” (page 2). It may be advisable to discuss about the emerging nanoscale imaging tools [1] to ascertain the mechanical performance of biomarkers of breast cancer diseases [2]. This will strengthen the importance to invest for the early prognosis of this devastating malignancy.

[1] https://doi.org/10.1002/smsc.202500351

[2] https://doi.org/10.1016/j.actbio.2023.01.011

3) Materials & Methods. 2.1. Retroviral Cell Clones and Culture. “(…) CO2 incubator until reaching (…)” (page 3). The stoichiometry should be depicted in subscript. This comment needs to be taken into account for the rest of the main manuscript body text.

4) 2.2.Compounds and Retroviral Screen Design. “Compounds were dissolved in 100% DMSO (…) and added to the culture medium (…)” (page 3). Did the authors assess the potential toxic effects of DMSO on the cell viability? Some insights need to be furnished in this regard.

5) Results. Did the authors carry out genomic-wide association study (GWAS) analysis to identify single nucleotide polymorphisms that could contribute to breast cancer risk?

6) Figure 2 (page 8). The lettering is slightly blurry. Could the authors enlarge the font size?

7) “4. Discussoin” (pages 13-15). This section perfectly remark the most relevant outcomes found by the authors in this work and the promising future prospectives. It may be remarkable to also briefly discuss about the potential future action lines to pursue the topic covered in this research.

Author Response

Dear Editor and Reviewers,

We appreciate and want to thank the editors and all reviewers for their valuable comments and suggestions to improve our manuscript. Below, we have replied to each point raised and revised the manuscript as indicated in this response letter.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript titled “Identification of Resistance Genes in Breast Cancer Cells Treated with Fulvestrant and Ribociclib via Retroviral Screening and Integration Site Sequencing” addresses a biologically significant and timely topic in cancer biology. Overall, the topic holds merit, though it needs some improvements before possible publication in the journal.

The abstract currently includes sub-section headings; please remove them.
The final paragraph of the introduction currently serves as a procedural summary of the study's steps, not as a statement of conceptual significance. The introduction's conclusion should not merely repeat "what was done"; instead, it must explicitly articulate why the study was necessary (its conceptual gap) and what its successful completion will establish (its anticipated contribution to the field).
The discussion section would benefit significantly from the inclusion of relevant and recent citations to support the interpretation of findings.
In section 4, the method's utility should be demonstrated by the results themselves. Secondly, the final paragraph of section 4 should focus on the significance of the findings, not the hypothetical findings you could have had with a better experiment. Listing these technical caveats as primary discussion points highlights the study's weaknesses rather than its contributions. Therefore, it is recommended to move the technical limitations to a brief, final "Limitations" subsection or a short paragraph. State them succinctly without excessive defensive justification.
Write a conclusion to integrate the findings, to acknowledge limitations, and to articulate the broader significance of the work.
Few recent references have been found in the manuscript, and most of them are outdated. Updating the references will ensure the manuscript remains timely and relevant.
The manuscript requires extensive editing for English language grammar and clarity to meet the standards of a scientific publication.

Author Response

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The work by Huang et al. addresses an important question in HR+/HER2 breast cancer using retroviral insertional mutagenesis and VIS‑NGS to nominate candidate genes for endocrine and CDK4/6 inhibitor resistance. The approach is technically sound and recovers both known and plausible novel resistance‑associated loci. Still, its impact is limited by a small sample size, lenient and uncorrected statistics, and a lack of functional validation. A clearer statistical framework and a more cautious, explicitly exploratory interpretation would substantially strengthen the work.

N.B. It was hard to specify areas for improvement by line numbers throughout the text because they were absent. Hence, the comments appear lengthy to specify the exact context that needs improvement.

General comments

- Throughout the manuscript, the authors should ensure that all gene names are italicized to match the standards of HUGO for gene nomenclature.

- Start sentences with words, not numerals (e.g., “Thirty-seven VIS loci…” instead of “37 VIS loci…”), etc.

Abstarct

- Please specify whether the 30% refers to progression on first-line ET+CDK4/6i in the metastatic setting, and ideally add “up to” or “around” to avoid over-precision for a single number.

- Also, please clarify the biological/system context in the study aim (i.e., “We aimed to identify genes associated with acquired resistance……. therapy in HR+ breast cancer cells.”

- Methods: Specify HR status of ZR75.1 (e.g., “ER-positive ZR75.1 breast cancer cells”) to align with HR+ context in the Background. Also, clarify whether “retroviral vector” was a gene-trap or overexpression library, and if space permits, state multiplicity (“library of randomly integrating retroviral vectors”) to explain the screening principle.

- Methods: The authors should clarify whether statistical thresholds (fold-change, p value/FDR) were applied (even briefly, e.g., “using enrichment over controls and statistical cut-offs”).

- Results: The authors should avoid calling VIS loci directly “resistance genes” without functional validation; use more cautious language: “37 VIS loci were associated with resistance to fulvestrant and ribociclib monotherapies” or “identified as candidate resistance loci.”

- By the end of the abstract, the authors are recommended to consider adding one short, impact-oriented clause: for example, “highlighting the complexity of resistance pathways and providing candidate targets for further functional validation and biomarker development.”

- The authors also should ensure the conclusion does not imply clinical validation; keep it clearly at the preclinical/cell-line level.

Introduction

2^nd paragraph:

“Despite these advances, therapy resistance remains a substantial clinical concern since about 15–20%… an additional 30–40% acquire resistance over time….⁴.” The authors should ensure that the percentages and the cited source match the clinical setting (e.g., adjuvant vs. metastatic HR+/HER2‑ disease); recent reviews emphasise that resistance is ultimately common/inevitable but may not give exactly these ranges. [PMID: 37511548 , PMID: 37035239].

Also, consider softening to “approximately” or “in the range of” and explicitly state whether these figures refer to endocrine therapy alone or ET+CDK4/6i.

3^rd paragraph

- Regarding “While the objective response rate for these combination therapies ranges from 20% to 40%, approximately 30% of patients derive no benefit” Plesae clarify setting (advanced/metastatic HR+/HER2‑ breast cancer) and which trials the 20–40% response rate and 23–28 month PFS range are drawn from, to avoid overgeneralisation across all CDK4/6i trials.

- “Moreover, both primary and secondary resistance ultimately develop in nearly all patients” is strong; consider rephrasing to “the majority of patients” or “resistance eventually develops in most patients” in line with recent reviews [PMID: 37511548, PMID: 36980649].

- Specify whether medians refer to first‑line ET+CDK4/6i in the metastatic setting, as reported in pivotal trials [PMID: 37035239, PMID: 37511548].

4^th paragraph

- Please briefly contrast retroviral screening with more contemporary tools (e.g., CRISPR or RNAi) to justify its use in the current study and highlight its specific advantages (e.g., stable genomic integration, saturation mutagenesis,…..).

5^th paragraph

- “developed in chapter 3a12” is unclear.

- “VIS-NGS allows for…… increasing the sensitivity of detection.” “If possible, give a brief, quantitative sense of improvement (e.g., “enabling detection of low‑frequency integration events not readily captured by earlier methods”), supported by a methodological reference(s).

Methods

Section 2.1

- Please clarify “LN-defective” (e.g., “LN‑defective murine retroviral vector (replication‑defective)”) and give a brief indication of the vector backbone or key features (e.g., neomycin resistance cassette) if not already in the referenced paper.

- Since clone IX3 is very old, consider adding a sentence noting that identity and viability were confirmed (e.g., STR profiling, mycoplasma testing) after thawing.

- Please clarify whether pLN1 and pLN2 derive from independent thawings or subcultures of IX3, and note passage number range if available, as this can matter for integration stability and phenotype.

- Regarding the medium and supplements, please indicate whether bovine calf serum was heat‑inactivated and specify if 100 pM oestrogen is physiologic or previously optimised, perhaps citing any prior optimisation.

- G418 concentration “8 μl/ml” is ambiguous; the authors should give it in

μg/ml or a final molar concentration and specify stock concentration to avoid confusion.

- Please consider indicating whether E2 and G418 were present continuously during drug treatments.

- Regarding the culture conditions, 70–80% confluency is fine, but you might state whether cells were subcultured before reaching confluence to avoid integration‑independent adaptations.

- “Untreated cells served as negative controls (NC) in all analyses” is good; you may note whether they contained the same DMSO vehicle as treated wells.

Section 2.2.

“Retroviral cell clones were cultured for two months in the presence or absence of therapeutic compounds.”

- Please state whether compounds were added continuously (media refreshed with drug) and how often medium and drug were changed (e.g., every 2–3 days), as this affects selection pressure.

- Please clarify here whether two months represents a fixed duration for all compounds or whether some were harvested earlier due to toxicity (this is partly addressed later, but could be foreshadowed).

- “Compounds were dissolved in 100% DMSO (except for palbociclib, which was dissolved in 10% DMSO)” is unclear: does “10% DMSO” refer to the stock solution or final DMSO in medium? Explicitly state both stock and final DMSO percentages to ensure controls are matched.

- “added to the culture medium at a fixed final concentration of 1,000 nM.”

Please indicate briefly how 1,000 nM was chosen (e.g., “approximately corresponding to IC50–IC70 in ZR75.1 cells from preliminary dose–response experiments,” or “similar to concentrations used in prior resistance studies”).

- Regarding the flask/plate formats and replicates, the use of different formats (T25 for 4‑OH‑tamoxifen vs 24‑well plates for other compounds) should be justified or at least acknowledged as a practical consideration; indicate that seeding densities were adjusted to achieve comparable cell numbers.

- “Cells from each biological replicate were harvested after one or two months of culture depending on yield.” This is ambiguous; specify criteria for early vs late harvesting (e.g., time to reach a stable proliferative state or minimum cell number).

- If both 1‑ and 2‑month samples were analysed by VIS‑NGS, state whether these were treated as separate timepoints or combined in downstream analyses.

Section 2.3

- Up to 1 × 10⁷ cells” is clear; state actual cell numbers per sample range if they differed substantially between conditions, as this may affect VIS diversity.

- Please indicate whether RNase treatment was performed (either as part of the kit or an added step).

- Regarding the Qubit quantification, the authors might specify that all DNA inputs for VIS‑NGS were normalised to a fixed amount (e.g., “XX ng per library”) if that is the case; otherwise, the link between quantification and downstream library prep remains implicit.

Section 2.4

- Regarding the fragmentation and library preparation, the fragment size “approximately 1 kb” is relatively long for Illumina MiSeq 300‑cycle reads; please consider clarifying whether the fragment size distribution was validated (e.g., Bioanalyzer) and how this relates to read length.

- When describing linkers and primers, the authors should consider indicating the approximate distance from the integration site that can be captured, if known, as this informs resolution.

- Regarding the PCR, the authors should indicate how many PCR cycles were used for initial and nested PCR and whether unique indexes/barcodes were added at this stage or during TruSeq library preparation, as these can affect bias and multiplexing.

- It would be helpful to mention any steps taken to minimise PCR duplicates or chimeras (e.g., limited cycle number, use of high‑fidelity polymerase).

Regarding the MiSeq micro V2 300 cycle: please specify read configuration (e.g., “paired‑end 2 × 150 bp”) and whether both ends were used for mapping.

- The cutadapt/BWA/BEDTools pipeline is standard; please consider adding mapping parameters (e.g., minimum mapping quality, removal of multi‑mapping reads) and how unique VIS were defined (e.g., collapsing reads mapping within X bp as a single integration site).

- “Reads per VIS locus were normalised to reads per million (RPM).” Plesae consider noting whether any further normalisation or batch correction between runs was applied.

Section 2.5

- “Of the 37 candidate genes, only 21 were recognised… based on a minimum expression threshold of 1 CPM in at least one cell.” Please give a rationale or citation for the 1 CPM threshold, and specify how CPM was calculated at the single‑cell level (e.g., library size normalisation).

- Please clarify whether genes not detected (16/37) were entirely absent or below threshold in all cells, and mention this explicitly as a limitation of the validation.

- Please indicate whether any adjustment for multiple testing was applied across the 21 genes; if not, state that results are exploratory.

Section 2.6

- Please consider specifying what variables were analysed with Student’s t‑tests (e.g., RPM‑normalised VIS counts across replicates) and whether assumptions (normality, equal variances) were assessed.

- Differences with p-values below 0.10 were defined as potential exploratory significance, p-values below 0.05 were considered statistically significant.” This dual threshold requires careful justification; consider adding a sentence earlier in Methods or at the start of Results to explain that the 0.10 threshold is used for hypothesis‑generating findings only.

- If multiple comparisons are substantial (e.g., many VIS loci/genes), consider whether false discovery rate (FDR) control is needed; if not applied, explicitly acknowledge that the analysis is exploratory and discuss false‑positive risk.

Results

- Regarding the sample size and VIS filtering clarity, please clarify the exact counting of samples (60 total; breakdown by treatment; handling of 1‑ vs 2‑month timepoints) and how many independent libraries per condition were analysed

- Also, please define precisely the criteria used to filter 3902 VIS to 426, then to 218 loci (e.g., “detected in ≥2 samples and in both biological replicates”) and justify exclusion of singleton loci, which may remove rare but relevant events.

- The use of p<0.1 as a threshold for both Fisher’s exact and t‑tests across many loci and treatments greatly inflates false positives; please either apply multiple‑testing correction (e.g., FDR) or very clearly label p<0.1 findings as exploratory and focus main conclusions on p<0.05 results.

- The authors should justify the use of Student’s t‑test on VIS RPM (normality/variance assumptions) and consider non‑parametric or transformed approaches if needed.

- Clearly specify which tests (Fisher, t‑test, Chi‑square, Fisher exact in 3.5) are used for which analyses, and provide effect sizes (e.g., odds ratios, fold‑changes) in supplementary tables.

- Throughout this section, the authors should tone down causal language: replace “resistance genes” with “candidate resistance genes” or “resistance‑associated loci” unless supported by functional perturbation data.

- Statements that specific loci “underlie common mechanisms,” or that position of VIS “modulates gene function and resistance,” should be clearly framed as hypotheses, not conclusions, and supported (if possible) with gene expression/orientation information.

- For BCAR1/BCAR3/EGFR, the authors should provide quantitative data (fold‑enrichment, p‑values) and acknowledge that these genes are known antiestrogen resistance mediators, not limited to 4‑OH‑tamoxifen.

- For the 20 “common” fulvestrant/ribociclib/CDK4/6i genes (including TRIM24), please soften claims to “may be involved in shared resistance pathways and warrant functional validation,” and explicitly relate TRIM24 findings to existing functional literature.

- For RPS14P7, DPM3, TRPS1 and comparison with clone VIII‑18, the authors should avoid speculative statements about clones being “overlooked accidentally”; rather, acknowledge limitations of screen coverage and selection bottlenecks and comment on differences in VIS coordinates (relative to gene structure).

- Regarding single‑cell RNA‑seq “independent validation,” only 21/37 genes were evaluable and only a subset showed directionally consistent changes; please temper the term “independent validation” and explicitly state that validation is partial and exploratory.

- Make p‑value handling consistent: TRIM24 at p=0.076 should be labelled as a trend rather than “significant,” and multiple‑testing issues across 21 genes should be addressed (FDR or explicit acknowledgement).

Discussion

- Throughout the Discussion, please replace “resistance genes” with “candidate resistance genes” or “resistance‑associated loci” unless supported by direct functional perturbation data.

- Statements that specific loci “underlie common mechanisms of resistance,” “modulate gene function,” or “are novel therapeutic targets” should be framed explicitly as hypotheses and linked to the correlative nature of VIS‑NGS data.

- The differential patterns of BCAR3, EGFR, and CCND1 (frequency vs read depth) are intriguing but do not yet establish which metric is more informative; current text (“frequency could matter more…”) is speculative and should be toned down or supported by additional analyses.

- TRPS1 and TRIM24 have strong independent evidence for roles in ER‑driven breast cancer and endocrine resistance. Please clearly distinguish what is novel here (their identification as VIS‑associated loci in fulvestrant/CDK4/6i resistance in ZR75.1) from what is already known, and avoid suggesting that their role as therapeutic targets is newly established.

- The discrepancy between TRPS1/RPS14P7/DPM3 detection in earlier tamoxifen‑resistant clones and absence in current tamoxifen‑treated cells should be discussed more cautiously (e.g., library complexity, stochastic sampling, different selection conditions) instead of attributing this mainly to “chance” or oversight.

- Although the limitations are partially acknowledged, the conclusion section should more clearly state that the current findings are hypothesis‑generating and require independent validation (in additional models and ideally clinical cohorts).

- Given the p<0.1 cut‑off and no FDR correction, please temper language about “stringent selection criteria” and “significance” and explicitly note that some reported associations may represent false positives.

- Only 21/37 genes were evaluable and only a subset showed concordant direction of change; please describe this as partial, exploratory support rather than strong independent validation.

- For genes with opposite directions between VIS and RNA‑seq (e.g., DPM3, MNT, NXPH1), the authors should provide a brief mechanistic explanation (disruptive vs activating insertions, clonal heterogeneity, different drug/regimen context) rather than leaving this discrepancy unexplored.

- The authors are recommended to move the more translational statements to a short “future directions” paragraph that clearly separates confirmed findings from anticipated applications.

- Phrases such as “predictive biomarkers that can guide treatment selection,” “personalized treatment strategies,” and “novel therapeutic targets” should be toned down or explicitly framed as future possibilities contingent on functional and clinical validation, given that all data derive from a single in vitro model.

Comments on the Quality of English Language

The manuscript is generally written in understandable English, but there are frequent issues with grammar, word choice, and typographical/encoding errors that impede readability. I recommend careful language editing by a fluent scientific English speaker or professional service to improve sentence structure, consistency of tense, and clarity, particularly in the Results and Discussion sections.

Author Response

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors did a great work to deal all the suggestions raised by the Reviewers. For this reason, the scientific manuscript quality was greatly improved. Based on the kovelty and significance of the gathered results, I warmly endorse this work for further publication in its current form.

Reviewer 2 Report

Comments and Suggestions for Authors

None

Reviewer 3 Report

Comments and Suggestions for Authors

I thank the authors for thoroughly addressing the reviewers' concerns. The manuscript now meets the journal's publication standards.