Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Section 2.3 states the IC50 for HT-29 is 28.4 µM (the 48 h value) and for HCT-116 is 65.3 µM, but Section 2.4 uses 33.2 µM for HT-29 and 84.8 µM (the 24 h values). Please clarify which IC50 (24 h or 48 h) was used for each downstream assay and why different time-point IC50s were chosen for different experiments this needs explicit justification or correction.

Figure 2 legend states "**, P < 0.01; ****, P < 0.0001" but the text (Section 3.2) only reports P < 0.01 for both cell lines — clarify which comparisons reached which significance level.

9,832 DEGs out of what is presumably ~20,000 detected genes is roughly half the transcriptome called "significant." This is unusually high even for a strong perturbation and raises questions about the DESeq2 model design (was this n=2 vs n=2, as Figure 5A suggests with only 4 ADP and 3-4 NC points?). With such low replicate numbers, please discuss whether multiple testing correction and dispersion estimation were robust, and ideally report exact biological replicate numbers in the RNA-seq/proteomics methods (Section 2.7–2.8 currently don't state n per group).

The conclusion that ADP "promotes... necrosis" rests on qualitative microscopic description of YO-PRO-1/PI staining (Section 3.4) without quantification (e.g., percentage of PI+/YO-PRO-1– cells vs. double-positive cells, which would distinguish late apoptosis from primary necrosis). Flow cytometry quantification of this assay, or at least cell counts per field across replicates, would substantially strengthen this claim.

The text (Section 3.5) states 254 molecules overlap between DEGs and DEPs, but Figure 5E appears to show different overlap and set values (e.g., "316" and unlabeled totals are hard to reconcile with "9516" DEGs vs. the 9832 DEGs stated in text, and 405/721 for DEPs). Please double check and correct the figure/legend/text for internal consistency.

The paper's title and abstale emphasize "integrated multi-omics," but the mechanistic validation (Figure 6) only follows up on five canonical apoptosis proteins (BAX, BID, CASP3, CYCS, BCL2) that could have been hypothesized without any omics data. The KEGG analysis flags ferroptosis and the p53 pathway as top hits (Section 3.5)  these are not explored at all. Consider either validating one of these less-expected pathways (ferroptosis marker like GPX4 or ACSL4) to justify the omics framing, or tempering the title/abstract to better reflect that the multi-omics data were primarily descriptive/hypothesis-generating rather than mechanistically pursued.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript investigates the anticancer activity of andrographolide (ADP) against colorectal cancer using a combination of in vitro assays, integrated transcriptomic and proteomic analyses, validation by Western blotting, and an in vivo xenograft model. The topic is relevant because the identification of natural compounds with anticancer potential remains an active area of research in colorectal cancer. The inclusion of multi-omics analyses represents a potentially valuable aspect of the study.

The results indicate that ADP suppresses proliferation and migration, induces G1-phase arrest, modulates apoptosis-related proteins, and reduces xenograft growth. However, several important methodological and interpretative limitations reduce the mechanistic strength of the conclusions. 

Therefore, substantial revision is required before the manuscript can be considered for publication.

Major Comments

• The conclusion that ADP induces apoptosis is primarily based on qualitative YO-PRO-1/PI fluorescence microscopy images. Although these images suggest increased cell death, they do not provide robust quantitative evidence. The authors should perform quantitative apoptosis analysis using flow cytometry (Annexin V/PI or Annexin V-FITC/7-AAD). Furthermore, cleaved caspase-3 and cleaved PARP should be evaluated, as total caspase-3 expression alone is not sufficient to demonstrate activation of the apoptotic pathway.
• The wound-healing assay was performed using concentrations close to the IC50 values. Under such conditions, reduced wound closure may simply reflect reduced proliferation or increased cell death rather than a true anti-migratory effect. The authors should repeat migration experiments using sub-cytotoxic concentrations
• The integrated transcriptomic and proteomic analyses represent one of the major strengths of the manuscript. However, despite identifying thousands of differentially expressed genes and hundreds of proteins, validation is restricted to a small set of apoptosis-related markers.

• The manuscript repeatedly concludes that ADP exerts its effects through apoptosis-related pathways. However, the omics analyses identified multiple enriched pathways, including ferroptosis and p53 signaling. The current data demonstrate association rather than direct mechanistic causality. The discussion and conclusion sections should be revised to avoid overstatement.

Minor Comments

• The manuscript should clarify whether the concentrations used in apoptosis assays correspond exactly to the experimentally determined IC50 values.
• Information regarding biological replicates for RNA-seq and proteomics should be provided more clearly.
• The Western blot images appear relatively low in resolution. Higher-quality images should be provided.
• Statistical analysis should specify whether normality testing was performed before application of parametric tests.
• Figure legends should clearly indicate the number of biological replicates used for each experiment.
• Additional discussion of andrographolide bioavailability and translational limitations would improve the manuscript.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Guan et. al. analyzed the results obtained after treatment with Andrographolide (ADP) on two colorectal cancer lines. Apparently, the study looks solid, as they disclose results on two distinct cell lines (HT-29 and HCT-116), provide data regarding IC50, cell-cycle, migration, apoptosis and necrosis studies, transcriptomics and proteomics analysis with some Western Blot confirmation and even disclose some preliminary results on HT-29 Xenograft mice. However carefully analysis of the manuscript revealed some flaws which could raise some questions regarding their conclusions. Their topic is not new, previous publications revealing similar analysis with sometimes distinct results (PMID: PMIDs: 33030050, 35599281). Thus, I would strongly encourage the authors to present their results in light with the already published data in the field. Another problem in my opinion is that the authors do not provide any result on a control cell line (healthy cell line). As it is currently known selectivity in chemotherapy is one of the major burning issues in oncology and providing results along such a control would actually disclose essential information regarding the ADP adoption in therapy. Alternatively, the authors could have provided results of ADP adoption in conjunction with other chemotherapies. In this aspect, it is not clear if the authors would like to suggest using ADP in chemotherapy or as an adjuvant for example in immunotherapy since ADP stimulates CD8+ T cell-infiltration and enhanced anti-PD-1 monoclonal therapy as they noted in the manuscript introduction section. Beyond this topic there are also other major issues which raises some questions. For example, it is not clear why do the authors choose to perform the transcriptomic and proteomics analysis only on a single cell line, although all the other parts of the manuscript present data on both cell lines (HT-29 and HCT-116). Solid motivation should have been provided for such an option. There are also some serious questions regarding their both transcriptomics and proteomics analysis. Although these were performed in dedicated facilities, no details are provided about how the samples were processed, on which instruments were used, how the data was obtained and analyzed. Moreover, there are no supplemental tables that should provide both the readers and the reviewers the possibility to assess the quality and relevance of their results. Clearly this is not with nowadays publication standards. Thus, considering all these aspects I cannot recommend publication. Below I have detailed some of the most important issues related to the manuscript:

 

 

1. The authors should include also a control (healthy cell line). Just studying the outcome on CRC cell lines does not add valuable information in relation to the current therapeutic options concerning CRC chemotherapy, as the selectivity is one of the main limiting factors in the field.
2. ADP study on colon-cancer cell line is not something new. That being said the authors should comment their obtained IC50 values and cell cycle results in light of previous results (PMIDs: 33030050, 35599281) and the observed discrepancies.
3. Section 3.4: Why did the authors used the IC50 values obtained at 24h and not the 48h values? The test for cell apoptosis and necrosis detection clearly states that the cell lines were incubated for 48h (see Materials and Methods).
4. Figure 4: The authors should add some quantification data relevant for the results in the figure, besides the shown representative images. The results could show % apoptotic, % of necrotic, % of live cells etc.
5. RNA Sequencing and analysis: Why do the authors study ADP impact only on HT-29 cells and not on both cell lines? The authors do not motivate their rationale for using only one of the cell lines and how did they choose this cell line. The authors should disclose similar data for HCT-116 cell line along a healthy control. Similar for the proteomics experiments.
6. RNA Sequencing and analysis: No details are provided about how the samples were processed and analyzed for the sequencing experiment. This is not acceptable. The authors should disclose all the details about the experimental part: library prep, sequencing instrument depth of the data, data qc and pre-processing etc.
7. Figure 5C: The PCA plot suggests a proportional higher variability of the data between the replicates of the same condition compared with the variability induced by ADP treatment (compare PC1 dimension of 39.44% with PC2 of only 18.12%, which includes the effect of ADP treatment). This suggest that their data is highly variable, beyond ADP treatment and there are other factors contributing to this. However, without access to their data is difficult to conclude. Thus, I would advise the authors to clearly denote the contributing factors to these variability beyond ADP treatment (technical, biological, analytical replicates etc).
8. Proteomic analysis: Again, no details are provided on how the experiments were conducted. This is unacceptable. The authors should provide the full details of how the samples were processed, injected, on how the data was acquired, qc, filtered and analyzed to obtain the results in Figure 5.
9. Figure 5E: There is a contradiction between the text and the figure. The authors state: “Integrated multi-omics analysis identified 254 molecules that were significantly altered at 235 both the gene and protein levels (Figure 5E)”. However the VennDiagram in Figure 5E shows 316 elements in the union.
10. Section 3.6: Which adjusted p values and fold change were obtained in the -omics experiments for BAX, CASP3, BID, and CYCS? These should be mentioned in the manuscript text.

 

Minor:

1. Figure 1B&C: For a robust fitting, IC50 should be calculated from a 8-10 points dilution as it requires usually fitting a sigmoidal curve. Also, their current data displays imbalanced dilutions, usually these should be half-log dilutions.
2. What concentration of ADP did the authors used for the transcriptomic analysis? The Materials and Methods section mentions IC50, but which IC50 value: the 24h or 48h value?
3. Figure 5A: the distribution of NC replicates on the PCA plot suggests that these were processed and analyzed in batches (NC1&2 and respectively NC3&4). What data did the authors used to obtain the PCA plot: raw, normalized, log-transformed, variance-stabilized etc?
4. Figure 5B: Which p value correction did the authors used to assess significance for the transcriptomic data?
5. Proteomic analysis: What drug concentration did the authors used? Again, the Materials and Methods the authors mention only IC50. But which value of IC50?
6. Figure 5D: The authors identified: 11k protein groups, which apparently it is a very deep proteome analysis. However carefully analysis of the figure reveals multiple protein groups which all display the same log2 fold change but distinct -log10 p values. These can be observed on both sides of the volcano plot, suggesting that all share the same either negative or positive ratio. This is at least strange and suggests that the authors kept protein groups with missing values during their data analysis. They should provide solid arguments for keeping these.
7. Figure 5B&D: Please add labels for the up-/down-regulated genes/proteins.
8. Figure 5F: How was the “Rich Factor” (x-axis) calculated? This is not described.
9. P9, L254-256: The authors state: “Consistent with the multi-omics findings, ADP treatment for 48 h significantly increased the expression of the pro-apoptotic proteins BAX, CASP3, BID, and CYCS in both cell lines…”. This is inaccurate. The authors did not present any multi-omics results on HCT-116 cell line.
10. Animal experiments: Why did the authors used only male mice?
11. No raw and analyzed data are available for the multi-omics experiments. The author should disclose the raw and analyzed data by deposition under dedicated repositories such as (but not necessarily limited to): NCBI GEO, EMBL-EBI ArrayExpress for transcriptomics, or PRIDE, MassIVE repositories for proteomics data. The data should be available to the reviewers and made publicly available after manuscript publication.
12. Full uncropped Western Blots should be made available as supplemental data accompanying the manuscript and also all the data (processed) from transcriptomics and proteomics experiments in .xlsx or tabular or .csv formatted tables to assess their results.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have successfully addressed the concerns. However, there are still a few minor typographical errors that should be corrected before publication.

Author Response

We sincerely thank the reviewer for the positive assessment of our revised manuscript. We appreciate the reviewer’s careful reading and valuable suggestion. In response, we have carefully checked the manuscript and corrected the minor typographical errors throughout the text.

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript can be acceptable in this form.

Author Response

We sincerely thank the reviewer for the positive assessment and are pleased that the manuscript is considered acceptable in its current form. We greatly appreciate the reviewer’s valuable comments and time.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors provided a new version of their manuscript which unfortunately does not meet the requirements for publication standards in my opinion. The authors refuse to add any additional experiments from the list of my requests and even more, they refuse to share their raw data based on which they draw their conclusions. Apart, from the methodological aspects there is no significant modification to the manuscript to support recommendation for publication. The authors cand find below a point-by-point comment to their responses:

 

 

1. The authors failed to address my request and they refuse to include additional experiments as requested.
2. The authors did not motivate why they obtained different IC50 values compared with those already reported. I specifically asked to include a comment on this aspect beside citing the relevant literature.
3. The authors modified accordingly the methodology section of the manuscript.
4. Figure 4: Although the authors added the number of biological replicates, they refused to provide quantitative data and statistical significance as I’ve requested.
5. Again the authors refused to provide any additional data.
6. The authors provided the requested details regarding the sequencing experiment.
7. The authors addressed only partial my request by further clarifying the experimental design setup, but not providing any relevant motivation for their result. Moreover, in the cover letter they mention that an additional experiment was done, but no additional data is provided in the revised manuscript.
8. The authors indeed provided details about the proteomic analysis. However these also include comments about the quality of the data (Pearson correlation coefficients distribution, peptide length distribution, % of 0 missed cleaveage peptides etc). However, these should go into the Results section of the manuscript accompanied by supplemental figures to sustain their claims.
9. This was indeed modified.
10. The table only lists the values for the transcriptomic analysis and not also for the proteomics as I have requested. Moreover it was not included into the revised manuscript as I have requested.
11. No additional points were included as I have requested.
12. This information was added indeed.
13. The authors added the requested details.
14. The authors clarified the p-value correction.
15. The section now contains the required concentration.
16. The authors acknowledge the error in plotting. However, they refused to change the figures as requested.
17. The authors added the labels only for one up-regulated protein and one down-regulated. Did they find Albumin in Figure 5D as significantly downregulated? And it is the only one down-regulated? This is strange.
18. This was added indeed.
19. The authors modified the text section.
20. Well regarding variability, going further with this idea the author could have just use a single animal to reduce variability. Moreover, as far as I know male mice can also display hormone variability (PMID: 30560152).
21. This is at least strange. I did not ask a phrase under which the authors to agree with me, but their raw data. Clearly, they refuse to share the data and are not aware of the nowadays publication standards. For example, the authors should follow closely the international recommended guidelines for reporting LC-MS/MS based proteomic data (see for example https://hupo.org/HUPO-Minimum-Information-Publication-Guidelines ).

Author Response

Please see the attachmen

Author Response File: Author Response.pdf