Phenolic Compounds of Grape Pomace Skin Released During SHIME Colonic Fermentation Shape the Transcriptomic Profile of Tight Junctions, Improving the Barrier Properties in Caco-2 Cells

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript addresses an question: whether phenolic compounds released from grape pomace skin during simulated colonic fermentation influence intestinal epithelial barrier-related responses. The integration of in vitro digestion, TWINSHIME fermentation, TEER measurements, and RNA-seq is a strength. However, the current version doesn't support the conclusions. I recommend major revision on clarifying the design limitations, and strengthening the barrier-mechanism validation.

1. The RNA-seq evidence needs more careful validation. RNA was not collected immediately after a single exposure, but only after two sequential apical challenges and a subsequent recovery period.
2. The manuscript states that PCA separated cells exposed to phenolic-rich ferments from control cells, but these data are reported as “data not shown.”
3. Although the authors report increased expression of several junction- or apical-domain-related genes, not all changes point in the same direction. Should discuss those changes.
4. The comparison between “phenolic-rich” and “phenolic-depleted” fractions is central to the manuscript, but the depleted fraction was generated using extensive extraction steps. This raises the possibility that the depleted material differs from the rich fraction in more than phenolic content alone, including matrix structure and accessibility of other components. Maybe add more control setups. 
5. The RNA-seq methods should specify which statistical framework was used. 
6. Should justify the two-challenge design more explicitly in the methods or discussion.
7. The manuscript should state much more clearly what counts as an independent replicate for SHIME, TEER, and RNA-seq, how many truly independent upstream fermentation replicates support the downstream conclusions.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Authors,

Your manuscript presents a coherent study on processed grape pomace skin, subjected to in vitro digestion followed by simulated human colonic fermentation in SHIME bioreactors, aiming to characterize phenolic compounds and SCFA, while also producing a physiologically relevant ferment suitable for apical exposure in differentiated Caco-2 cells. The research topic is both timely and relevant, offering valuable insights into the biochemical characterization of grape pomace fermentation and its potential effects on intestinal cell models. Overall, the manuscript is clearly written and well-organized. Nevertheless, several aspects could benefit from clarification and improvement. My comments and suggestions are outlined below:

1. In the Introduction, please add a few sentences describing the Caco-2 cell line and clearly explain the rationale for selecting this particular model for the study.
2. It would be helpful to clarify which standards were used for the HPLC analysis and the range of the calibration curves. Additionally, please specify the limits of detection (LOD) and quantification (LOQ) for the individual compounds analyzed by HPLC.
3. Two different figures are currently labeled as Figure 1. This needs to be corrected to ensure each figure has a unique and sequential label.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript is clearly written, methods are reproducible, and data support the conclusions without overclaiming. With moderate revisions (mainly validation and limitations), it is suitable for publication.
Major Comments
Add functional validation of barrier improvement.
Perform a complementary paracellular permeability assay (e.g., FITC-dextran 4 kDa flux) at key time points to confirm that TEER recovery reflects genuine tight-junction strengthening.
Improve RNA-seq transparency and validation.
Make the GEO dataset (GSE317808) public upon acceptance; provide full DEG list and GO/KEGG tables as supplementary material; validate key genes (CLDN3/4, MARVELD3, CFTR, NHERF1) by qPCR.
Deepen phenolic/metabolite profiling of the ferments.
Use targeted LC-MS/MS (or untargeted metabolomics) on the 24 h ascending-colon ferments to identify specific released phenolics or microbial derivatives responsible for the observed effects.
Address microbiota donor variability.
Explicitly discuss the single-donor limitation in the Discussion (and Abstract) and note that multi-donor or pooled-feces studies would strengthen generalizability.
Clarify SCFA-independence.
Acknowledge that other microbiota-derived metabolites (beyond phenolics) cannot be entirely ruled out as contributors.
Minor Comments
Abstract: Mention the repeated-challenge design and that the phenolic-rich ferment prevented the TEER drop on the second challenge.
Figures: Add error bars, p-values, and clear labels to Fig. 1 and Fig. 2.
Methods: Standardize NDF-g dose terminology; clarify “FNDg”; specify RNA-seq pipeline used (limma or DESeq2).
Discussion: Briefly compare with recent grape-pomace SHIME studies.
Language: Minor proofreading for consistency (“fermenta”, long sentences).

Comments on the Quality of English Language

The English language is generally understandable; however, it would benefit from moderate editing for clarity, conciseness, and grammatical consistency. Some sentences are overly long and could be simplified to improve readability.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for the revised manuscript and the detailed point-by-point response. The authors have addressed my major concerns in a satisfactory manner. I recommend acceptance after minor editorial revision. 

Author Response

Thank you for your positive feedback and for recommending the acceptance of our work.

In this final version, we have implemented minor editorial improvements to enhance the manuscript’s clarity. Specifically, we have refined the text for better flow, unified terminology, and improved the presentation of the tables and the volcano plot to facilitate the identification of key genes.

We appreciate your valuable guidance throughout this process.

Reviewer 3 Report

Comments and Suggestions for Authors

The revised manuscript shows clear improvement in language, structure, and overall readability compared to the previous version. The abstract is now more coherent, and the introduction provides a clearer narrative linking dietary fiber, phenolic compounds, and gut microbiota interactions. The methods section is also easier to follow, particularly the description of the SHIME system and RNA-seq workflow.

While the manuscript has improved in presentation, further improvements in scientific depth would strengthen the study, such as incorporating orthogonal validation experiments (e.g., paracellular permeability assays), validating transcriptomic results using qPCR or protein-level analysis, and providing deeper insight into the specific phenolic or microbial metabolites driving the observed biological effects.

RNA-seq analysis provides useful information on gene expression changes, the study does not further explore the underlying mechanisms driving these effects or validate key findings using independent methods (e.g., qPCR or protein-level analysis).

In addition, conclusions regarding improved barrier function are based solely on TEER measurements and would be strengthened by complementary permeability assays. Similarly, the interpretation of SCFA-independent effects should be presented more cautiously, as other microbiota-derived metabolites may also contribute.

The manuscript would benefit from a more explicit and critical discussion of study limitations. In particular, the use of a single microbiota donor should be clearly acknowledged, as inter-individual variability may influence phenolic metabolism and the resulting biological effects. In addition, the reliance on in vitro models (SHIME and Caco-2) should be discussed more thoroughly, as these systems do not fully capture the complexity of the human gastrointestinal environment, thereby limiting the generalizability and physiological relevance of the findings.

Although the manuscript has improved considerably in clarity and presentation, it may still benefit from a final round of revision to address the reviewer comments. Overall, the revised version represents a meaningful improvement in readability, the enhancement in scientific depth remains modest.

Author Response

Dear Reviewer, please find the attached document for your review.

Author Response File: Author Response.pdf