Targeting Polyamine Metabolism in Colorectal Cancer: Apigenin Dismantles the HIF-1α/SMOX Positive Feedback Loop to Suppress Tumor Progression

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This research provides significant insights into the molecular mechanisms of Apigenin as an antitumoral agent in colorectal cancer. The proposal of Spermine oxidase (SMOX) participation in HIF-1α/SMOX pathway is an important contribution to knowledge. The authors provide a comprehensive description of field background justifying this particular choice for therapy development, and I consider the research design is well-founded for achieving the main objectives. Nevertheless, certain recommendations could improve the quality and presentation of the results.

First, the methodology requires further detail, including technical information on reagents and equipment. The text lacks the approval number from Ethics Committee for animals use and experimentation. The RKO cell line is duplicated in the methodology section. Bioinformatics is poorly described in the text, which is a growth opportunity to enhance the reach of this study. Additionally, the KEGG pathway analysis lacks a proper description of the search criteria; authors should select the key information from the KEGG map and create a clearer, well-annotated figure. Line 232 includes the hydrogen peroxide formula as H2O2; the correct chemical formula should use subscripts (H₂O₂).

Moreover, all histological images are poorly described and lack scale bars. The figures are difficult to interpret as the images and labels are too small. Authors should increase the figure size or consider to split them into multiple figures. Regarding the statistical analysis, authors should include a more in-depth analysis of the results, including the normality tests used. Finally, references should be updated to include sources from 10 years to the present, as some current citations date back to 2002.

Author Response

Comments 1: First, the methodology requires further detail, including technical information on reagents and equipment.

Response 1: Thank you for pointing this out. We agree with this comment. We have expanded the methodology section to include more specific technical information regarding reagents and equipment. Please refer to Section 4.1 in the revised manuscript.

Comments 2: The text lacks the approval number from Ethics Committee for animals use and experimentation.

Response 2: Agree. We have added the specific approval number to the main text to ensure full transparency. 

Comments 3: The RKO cell line is duplicated in the methodology section.

Response 3: We apologize for this typographical error. We have revised it.

Comments 4: Bioinformatics is poorly described in the text, which is a growth opportunity to enhance the reach of this study. Additionally, the KEGG pathway analysis lacks a proper description of the search criteria;

Response 4: Thank you for this constructive feedback. We completely agree that a robust bioinformatics description enhances the study's reproducibility. Thus, we have enriched Section 4.2 (Network Pharmacology Analysis). And we have provided the exact search criteria for the KEGG and GO analyses.

Comments 5: authors should select the key information from the KEGG map and create a clearer, well-annotated figure.

Response 5: We deeply appreciate the reviewer's suggestion regarding the KEGG maps. After careful consideration, we respectfully request to retain the original KEGG map formats for both Figure 1F  and Figure 3D. Our opinion is that these figures represent the direct, objective outputs of the initial untargeted metabolomics screening and network pharmacology analyses. Retaining the original KEGG outputs preserves the macroscopic, unbiased landscape of the global metabolic and signaling networks without premature subjective filtering. We think that Figure 10 extracts the key nodes from these massive KEGG networks, providing illustration of exactly how API intervenes in the HIF-1α/SMOX positive feedback loop. We hope this approach atisfies the reviewer's underlying concern.

Comments 6: Line 232 includes the hydrogen peroxide formula as H2O2; the correct chemical formula should use subscripts (H₂O₂).

Response 6: We apologize for these mistakes. We have checked again and revised them.

Comments 7: Moreover, all histological images are poorly described and lack scale bars.

Response 7: We agree with this comment. We have modified all histological images to include clear scale bars. And we have enriched the histological descriptions in Section 2.4.

Comments 8: The figures are difficult to interpret as the images and labels are too small. Authors should increase the figure size or consider to split them into multiple figures.

Response 8: Thank you for the suggestion. We have optimized the layout and and split the figure 2.

Comments 9: Regarding the statistical analysis, authors should include a more in-depth analysis of the results, including the normality tests used.

Response 9: Agree. We have updated the statistical analysis section to provide more depth.

Comments 10: Finally, references should be updated to include sources from 10 years to the present, as some current citations date back to 2002.

Response 10: We completely agree with this recommendation. We have renewed our reference list.

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript by Zhang et al. presents a multi-omics approach to determine the mechanism of Apigenin in colorectal cancer. The identification of transcriptional interaction between HIF-1α and SMOX is well-supported by presented data. The use of a high-throughput endogenous metabolite library is a significant methodological contribution to determine the role of API.

Major Comments

1. The manuscript emphasizes that API functions on TLR4/MyD88 and  tumor microenvironment. However, the in vivo validation was performed using BALB/c nude mice that are immunodeficient and lack functional T-cells. The authors should acknowledge this limitation in the Discussion section.
2. The authors present untargeted metabolomics data in Figure 1, but do not describe the number of patient samples used for the tumor and adjacent normal tissue groups in both the Results and Methods. Furthermore, there is a lack of clinical and pathological information regarding the tumor samples used. The authors must state the sample size and provide the patient demographics and tumor characteristics.
3.  The authors demonstrate that adding exogenous polyamines rescues CRC cells from API-induced cytotoxicity. To definitively prove that API's anti-tumor effects are dependent on SMOX suppression, the authors should test whether the SMOXOE cells are resistant to API's anti-proliferative OR anti-migratory effects.

Minor Comments

• Figures presented are low resolution and hard to assess. 
• The scatter plots for the metabolite library screening (Figure 5B and 5D) are difficult to interpret. The manuscript should describe statistical threshold used to determine what constituted a "significant" rescue or sensitization among the 860 metabolites.
• Consistent formatting of gene versus protein symbols - HIF1A is italicized when referring to the gene and HIF-1α is used for the protein.

Author Response

Comments 1: The manuscript emphasizes that API functions on TLR4/MyD88 and tumor microenvironment. However, the in vivo validation was performed using BALB/c nude mice that are immunodeficient and lack functional T-cells. The authors should acknowledge this limitation in the Discussion section.
Response 1: Agree. We have added a paragraph to the Discussion section acknowledging that our BALB/c nude mice model lacks functional T-cells and cannot fully capture the complex interactions with the adaptive immune system.

Comments 2: The authors present untargeted metabolomics data in Figure 1, but do not describe the number of patient samples used for the tumor and adjacent normal tissue groups in both the Results and Methods. Furthermore, there is a lack of clinical and pathological information regarding the tumor samples used. The authors must state the sample size and provide the patient demographics and tumor characteristics.
Response 2: Agree. We have updated both the Results (Section 2.1) and Methods (Section 4.1) sections to provide the exact clinical details. The study utilized 11 paired tissue specimens from patients with rectal adenocarcinoma, with adjacent normal tissues taken>5 cm from the tumor margin.

Comments 3: The authors demonstrate that adding exogenous polyamines rescues CRC cells from API-induced cytotoxicity. To definitively prove that API's anti-tumor effects are dependent on SMOX suppression, the authors should test whether the SMOXOE cells are resistant to API's anti-proliferative OR anti-migratory effects.
Response 3: Agree. We actually performed these validation experiments previously. We conducted both CCK-8 and cell scratch assays on NC/SMOXOE cells treated with API. The results clearly demonstrate that SMOX overexpression resist to API's effects. However, to maintain the narrative flow of the main text, we deleted these data at first. We now propose providing these results as Supplementary Figure 1 rather than inserting them into the main manuscript. We request the reviewer to review this supplementary data and advise if it is acceptable.

Minor Comments

Comments 4: Figures presented are low resolution and hard to assess.
Response 4: Agree. We have regenerated all the figures and split the figure 2 into two pieces.

Comments 5: The scatter plots for the metabolite library screening (Figure 5B and 5D) are difficult to interpret. The manuscript should describe statistical threshold used to determine what constituted a significant rescue or sensitization among the 860 metabolites.
Response 5: Agree. We have defined the statistical threshold in Section 4.5. Metabolites were considered significant "rescue" only if they achieved an empirical dual-threshold, which are a relative viability ratio≥1.20 combined with statistical significance (p<0.05). Furthermore, we have added a red dashed line at the 1.20 mark in the corresponding scatter plots to visually represent this threshold and updated the figure legends.

Comments 6: Consistent formatting of gene versus protein symbols - HIF1A is italicized when referring to the gene and HIF-1α is used for the protein.

Response 6: Agree. We have checked again and revised them.