Decoding the Molecular Drivers of Epithelial to Mesenchymal Transition in Breast Cancer: Insights into Epithelial Plasticity and Microenvironment Crosstalk

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript presents a well-designed and comprehensive study investigating EMT and epithelial plasticity in breast cancer using integrated proteomic and bioinformatic analyses. The work is timely, clearly written, and the identification of an EMT-related gene signature with prognostic relevance is a notable strength.

I believe the manuscript would benefit from addressing and clarifying the following points, which I believe will further improve its rigor and clarity:

1. Can the authors more clearly state the core hypothesis and explain how the observed proteomic EMT heterogeneity motivated the subsequent in silico EMT gene signature analysis?

 

2. How do the authors distinguish true hybrid epithelial/mesenchymal states within tumor cells from mixed cellular populations, particularly stromal or immune cell contributions, given the use of bulk tissue lysates?

 

3. The term “alfa-Smooth Muscle Actin” used in the Western blotting methods appears incorrect. Please correct to “alpha-smooth muscle actin (α-SMA)” and ensure consistency throughout the manuscript.

 

4. How does the proposed 37-gene EMT signature compare with existing EMT or prognostic signatures in breast cancer, and what is its potential clinical or translational application?

Author Response

REVIEWER 1

This manuscript presents a well-designed and comprehensive study investigating EMT and epithelial plasticity in breast cancer using integrated proteomic and bioinformatic analyses. The work is timely, clearly written, and the identification of an EMT-related gene signature with prognostic relevance is a notable strength.

Reply: We thank the reviewer for the positive comments and for considering our study to be well structured, clearly written, and of high scientific merit.

I believe the manuscript would benefit from addressing and clarifying the following points, which I believe will further improve its rigor and clarity:

1. Can the authors more clearly state the core hypothesis and explain how the observed proteomic EMT heterogeneity motivated the subsequent in silico EMT gene signature analysis?

Reply: We thank the reviewer for this important comment, which allowed us to better clarify both the rationale and the core hypothesis of the study. Our proteomic analyses revealed that individual patients displayed highly variable expression patterns of both epithelial and mesenchymal markers, which often co-existed within the same tumor sample. Moreover, the identification of distinct protein isoforms with patient-specific expression patterns uncovered an unexpected level of regulatory complexity. Collectively, these findings indicate that EMT in breast cancer is far more complex and finely regulated than predicted by the classical binary epithelial–mesenchymal model and strongly support the existence of dynamic epithelial plasticity rather than uniform EMT states. Based on this evidence, we hypothesized that the biological and clinical relevance of EMT heterogeneity cannot be adequately captured by single markers or limited proteomic panels. So, we extended our study toward a comprehensive in silico approach aimed at capturing the EMT regulatory landscape. Accordingly, we complemented our experimental data with a systematic bioinformatic analysis to identify EMT-associated transcriptional programs and derive a robust EMT gene signature with prognostic relevance. To address the reviewer’s comment, we have revised the Abstract and the Introduction to more clearly state the core hypothesis and rationale of the study. A detailed explanation of the transition from proteomic findings to the in silico analysis is also provided at the beginning of the Results 3.3 section, entitled “In silico analysis”.

2. How do the authors distinguish true hybrid epithelial/mesenchymal states within tumor cells from mixed cellular populations, particularly stromal or immune cell contributions, given the use of bulk tissue lysates?

Reply: We thank the reviewer for raising this important point. As stated, while the tissue specimens analyzed are enriched for tumoral cells, we acknowledge that contributions from the tumor microenvironment cannot be excluded. Importantly, stromal and immune cells are known to actively influence epithelial plasticity, and their presence may therefore be biologically relevant rather than merely confounding. Accordingly, the Conclusion section has been expanded to emphasize the need for future spatially resolved and single-cell analyses to more precisely define hybrid epithelial/mesenchymal states.

3. The term “alfa-Smooth Muscle Actin” used in the Western blotting methods appears incorrect. Please correct to “alpha-smooth muscle actin (α-SMA)” and ensure consistency throughout the manuscript.

Reply: Done

4. How does the proposed 37-gene EMT signature compare with existing EMT or prognostic signatures in breast cancer, and what is its potential clinical or translational application?

Reply: We thank the reviewer for this important question. A systematic comparison of the proposed 37-gene EMT signature with existing breast cancer signatures would require additional bioinformatic integration with large datasets, which is beyond the scope of the present study. Regarding potential clinical applications, our study primarily serves as an exploratory analysis, identifying a candidate EMT gene signature based on bioinformatics and proteomics data. Functional and clinical validation of these genes will be essential to confirm their role in breast cancer progression and to establish their potential as reliable biomarkers. Once their functional relevance is confirmed, diagnostic panels could be developed, and potential therapeutic strategies involving small molecule inhibitors or antibody-drug conjugates could be explored. Moreover, understanding the relationship between this EMT gene signature and clinical treatments, including chemotherapy, targeted therapy, and immunotherapy, will require further investigation. We have clarified this aspect in the Conclusion section, emphasizing the importance of functional validation as the next step before considering clinical translation.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This study analyzed the expression of EMT-associated proteins in a cohort of 95 BC tissue samples, revealing significant intratumor and intertumor heterogeneity. Proteomic data supported the presence of cellular plasticity and hybrid phenotypes, while in silico analyses revealed deregulation of 144 EMT-associated genes and presented a novel EMT signature of 37 genes with potential prognostic and therapeutic significance.

The methodology of the study is generally appropriate, and the interpretations are generally consistent with the data. However, some points could be addressed as minor revisions. The statistical methods used in the correlation analyses with marker expressions and clinical parameters could be explained more clearly. Immunohistochemical analyses, as suggested in the study, would be valuable in elucidating the cellular origin of the findings. Comparing breast cancer types might be more helpful in illuminating the diversity. In addition, confirming a few of the selected key genes with Western blot or IHC and/or discussing their MET association would strengthen the study.

Author Response

REVIEWER 2

This study analyzed the expression of EMT-associated proteins in a cohort of 95 BC tissue samples, revealing significant intratumor and intertumor heterogeneity. Proteomic data supported the presence of cellular plasticity and hybrid phenotypes, while in silico analyses revealed deregulation of 144 EMT-associated genes and presented a novel EMT signature of 37 genes with potential prognostic and therapeutic significance. The methodology of the study is generally appropriate, and the interpretations are generally consistent with the data. However, some points could be addressed as minor revisions.

Reply: We thank the reviewer for this positive assessment of our work.

The statistical methods used in the correlation analyses with marker expressions and clinical parameters could be explained more clearly.

Reply: We thank the reviewer for this helpful comment. We have expanded in the Materials and Methods section the description of the statistical tests used for the correlation analyses including the thresholds for statistical significance.

 

Immunohistochemical analyses, as suggested in the study, would be valuable in elucidating the cellular origin of the findings. Comparing breast cancer types might be more helpful in illuminating the diversity.

Reply: We thank the reviewer for this insightful suggestion. We agree that immunohistochemical analyses could provide valuable information regarding the cellular origin of the observed findings, and that comparisons among different breast cancer subtypes would further elucidate biological diversity. However, in the context of the present study, the evaluation of EMT markers across breast cancer subtypes could increase analytical complexity without necessarily clarifying the relationship between epithelial plasticity and tumor behaviour. Nonetheless, we recognize that assessing the contribution of the tumor microenvironment cells is critical. To address this point, we complemented our analyses with in silico investigations using the TIMER database, which allows inference of cell-type–specific contributions through correlation analyses across large transcriptomic datasets. Specifically, we assessed the relationship between EMT marker expression and immune cell infiltration, including cancer-associated fibroblasts, macrophages, endothelial cells, CD8⁺ T cells, and CD4⁺ T cells. These analyses revealed significant positive correlations between Vimentin and α-SMA expression and multiple components of the tumor immune microenvironment. In contrast, E-cadherin and Cytokeratin-18 showed only weak correlations with immune cell infiltration, consistent with their predominantly epithelial expression. These findings support the potential role of stromal and immune cells into the EMT marker expression in bulk tissue lysates. The new results have been included in the Results section, with a new figure. In the conclusion section we also emphasize that immunohistochemical validation remains an important direction for future studies.

In addition, confirming a few of the selected key genes with Western blot or IHC and/or discussing their MET association would strengthen the study.

Reply: We thank the reviewer for this suggestion. As stated in our previous response, we will consider these important validation in a new study.

 

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Reviewer Report
This manuscript provides a comprehensive and well-structured investigation of epithelial-mesenchymal transition (EMT) and epithelial plasticity in breast cancer, combining proteomic data with extensive silico analyses. The study is timely, technically robust, and gives important information about EMT heterogeneity, hybrid epithelial/mesenchymal states, tumor microenvironment crosstalk, and cancer stem cell-associated traits. The discovery of a 37-gene EMT signature with predictive significance is particularly noteworthy.
The given results provide strong support for the conclusions, and the experimental design, data interpretation, and bioinformatic integration are generally sound. The work is well-referenced and written in an understandable manner.
Specific comments (minor revisions)
1. Clarity of language.
• The article has minor grammatical and typographical errors, including spacing, hyphenation, and line breaks in gene/protein names. Careful language change would increase readability.
• To improve clarity, consider shortening some long sentences in the introduction and discussion.
2. Tables and figures
• Consistency in table numbering should be verified (e.g., duplication use of "Table 1").
• Figure legends (e.g., Figures 1–4) could be further clarified to improve interpretability for readers unfamiliar with the experimental setup.
3. Proteomic data interpretation
The Discussion may benefit from a stronger statement recognizing the potential role of stromal cells (e.g., fibroblasts, myoepithelial cells, immune cells) in Vimentin and α-SMA signals in bulk tissue lysates.
4. Ensure consistent nomenclature.
Use consistent gene and protein nomenclature throughout the text and additional resources.
5. Statistical Reporting Provide clarification on: -Correction for multiple testing in survival analysis -Provide confidence intervals for hazard ratios (if available).
6. Recommendation
Include a short paragraph about limitations, such as:
Limitations of this study include the use of bulk tissue lysates, absence of spatial resolution, and lack of functional validation experiments. Future studies incorporating single-cell and spatial transcriptomic approaches will be essential.
Overall, the manuscript is of high scientific merit and suitable for publication after minor to moderate revision, primarily focused on improving clarity, presentation, and contextual interpretation.

Comments on the Quality of English Language

• Minor grammatical and typographical errors are present throughout the manuscript (e.g., line breaks in gene/protein names, inconsistent spacing, and hyphenation). A careful language and style revision is recommended.
• Some sentences, particularly in the Introduction and Discussion sections, are very long and could be split to improve readability.

Author Response

REVIEWER 3

This manuscript provides a comprehensive and well-structured investigation of epithelial-mesenchymal transition (EMT) and epithelial plasticity in breast cancer, combining proteomic data with extensive silico analyses. The study is timely, technically robust, and gives important information about EMT heterogeneity, hybrid epithelial/mesenchymal states, tumor microenvironment crosstalk, and cancer stem cell-associated traits. The discovery of a 37-gene EMT signature with predictive significance is particularly noteworthy. The given results provide strong support for the conclusions, and the experimental design, data interpretation, and bioinformatic integration are generally sound. The work is well-referenced and written in an understandable manner.

Reply: We thank the reviewer for this positive assessment of our work.

Specific comments (minor revisions)

1. Clarity of language: The article has minor grammatical and typographical errors, including spacing, hyphenation, and line breaks in gene/protein names. Careful language change would increase readability. To improve clarity, consider shortening some long sentences in the introduction and discussion.

Reply: We thank the reviewer for this suggestion. We have carefully revised the manuscript to correct grammatical and typographical errors.

2. Tables and figures: Consistency in table numbering should be verified (e.g., duplication use of "Table 1"). Figure legends (e.g., Figures 1–4) could be further clarified to improve interpretability for readers unfamiliar with the experimental setup.

Reply: We thank the reviewer for this suggestion. To avoid misinterpretation we added Supplementary for each Figure or Table, inserted as supplementary material.

3. Proteomic data interpretation: The Discussion may benefit from a stronger statement recognizing the potential role of stromal cells (e.g., fibroblasts, myoepithelial cells, immune cells) in Vimentin and α-SMA signals in bulk tissue lysates.

Reply: We thank the reviewer for this suggestion, which was also raised by another reviewer. To address this point, we complemented our analyses with in silico investigations using the TIMER database, which allows inference of cell-type–specific contributions through correlation analyses across large transcriptomic datasets. Specifically, we assessed the relationship between EMT marker expression and immune cell infiltration, including cancer-associated fibroblasts, macrophages, endothelial cells, CD8⁺ T cells, and CD4⁺ T cells. These analyses revealed significant positive correlations between Vimentin and α-SMA expression and multiple components of the tumor immune microenvironment. In contrast, E-cadherin and Cytokeratin-18 showed only weak correlations with immune cell infiltration, consistent with their predominantly epithelial expression. These findings support the potential role of stromal and immune cells into the EMT marker expression in bulk tissue lysates. The new results have been included in the Results section, with a new figure, and are further discussed in the Discussion section.

4. Ensure consistent nomenclature: Use consistent gene and protein nomenclature throughout the text and additional resources.

Reply: We thank the reviewer for this suggestion. We have carefully revised the manuscript and supplementary materials to ensure that gene and protein nomenclature is used consistently and correctly

5. Statistical Reporting Provide clarification on: -Correction for multiple testing in survival analysis; -Provide confidence intervals for hazard ratios (if available).

Reply: We thank the reviewer for this comment. Survival analyses were performed using KM Plotter, which applies appropriate statistical corrections for multiple testing as part of its standard workflow. Hazard ratios and correction for multiple testing are automatically provided by the software. In particular, HR and logrankP are automatically included in the generated plots. More important, these analyses can be readily replicated by other researchers using the same platform, ensuring transparency and consistency of the results. These information’s have now been added to the Material and Methods section.

6. Recommendation: Include a short paragraph about limitations, such as: Limitations of this study include the use of bulk tissue lysates, absence of spatial resolution, and lack of functional validation experiments. Future studies incorporating single-cell and spatial transcriptomic approaches will be essential.

Reply: We thank the reviewer for this comment. Accordingly, we add these limitations in the conclusion section

 

 

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

1. IHC is missing for spatial validation, it would be better if the localization of the markers (Vimentin, E-Cadherin and alpha-smooth muscle actin) to differentiate between tumor cells and stromal areas
2. The authors have mentioned 37 gene EMT signature which is entirely from public datasets. I would recommend validating from TCGA dataset
3. There could be an inclusion of certain functional assays such migration and invasion assays for mechanistic claims
4. Why you selected 14 studies from EMTome database and what was the inclusion criteria
5. Also mention the rational of selecting KRT18 over KRT19?

Author Response

REVIEWER 4

 

1. IHC is missing for spatial validation, it would be better if the localization of the markers (Vimentin, E-Cadherin and alpha-smooth muscle actin) to differentiate between tumor cells and stromal areas

Reply: We thank the reviewer for this comment. As stated in our previous response to Reviewer 2, we agree that immunohistochemical analyses could provide valuable information regarding the cellular origin of the observed findings. To address this point, we complemented our analyses with in silico investigations using the TIMER database, which allows inference of cell-type–specific contributions through correlation analyses across large transcriptomic datasets. Specifically, we assessed the relationship between EMT marker expression and immune cell infiltration, including cancer-associated fibroblasts, macrophages, endothelial cells, CD8⁺ T cells, and CD4⁺ T cells. These analyses revealed significant positive correlations between Vimentin and α-SMA expression and multiple components of the tumor immune microenvironment. In contrast, E-cadherin and Cytokeratin-18 showed only weak correlations with immune cell infiltration, consistent with their predominantly epithelial expression. These findings support the potential role of stromal and immune cells into the EMT marker expression in bulk tissue lysates. The new results have been included in the Results section, with a new figure. In the conclusion section we also emphasize that immunohistochemical validation remains an important direction for future studies.

 

2. The authors have mentioned 37 gene EMT signature which is entirely from public datasets. I would recommend validating from TCGA dataset

Reply: We thank the reviewer for this comment. We would like to clarify that the public databases used in this study (including those employed for survival and immune infiltration analyses) are based on TCGA breast cancer datasets or integrate TCGA-derived transcriptomic and clinical data. Therefore, the proposed 37-gene EMT signature has already been evaluated using TCGA-based data

3. There could be an inclusion of certain functional assays such migration and invasion assays for mechanistic claims

Reply: We thank the reviewer for this valuable suggestion. We agree that functional assays, such as migration and invasion assays, would provide important mechanistic insights. These experiments are planned as part of future studies aimed at functionally validating the EMT-related findings reported here.

4. Why you selected 14 studies from EMTome database and what was the inclusion criteria

Reply: We thank the reviewer for this question. The selection of the 14 studies was not arbitrary but was determined by the availability of data within the EMTome platform. Specifically, when filtering EMTome for breast cancer, only 14 studies met the platform-defined criteria and were available for analysis. These studies were therefore included in their entirety.

5. Also mention the rational of selecting KRT18 over KRT19?

Reply: We thank the reviewer for this comment. In the context of our study, KRT18 and KRT19 are both well-established epithelial markers and do not exhibit major functional differences with respect to the questions addressed here.

 

Author Response File: Author Response.pdf