Cocoa Powder Modulates HIF-1α Stability and Inhibits Ocular Angiogenic and Degenerative Pathology

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

  The manuscript is well written; however, several questions need to be addressed before publication. It is unclear whether the authors performed two separate analyses using HPLC and LC–MS. If not, the authors should clearly explain the instrumentation and how the tandem systems were used. In addition, how was the same concentrated sample injected into the quadrupole MS if it was first detected using a UV detector, as such concentrations typically oversaturate the MS. The manuscript also does not clearly state how many mice were used in the study. Finally, the authors performed a total proteomic assay using BSA; it would be more informative to quantify and identify the upregulated and downregulated proteins associated with immune function.

Author Response

The revised part in the text is in blue color.

 

Reviewer #1

Comments and Suggestions for Authors

The manuscript addresses a highly interesting topic in the field of nutrition and functional food science. However, several aspects require clarification and improvement.

 

1. The title refers to “Flavanol-Rich Cocoa Powder”; however, no chemical profiling is provided to support this claim. Instead, the authors emphasize the presence of alkaloids. The title should therefore be reconsidered to better reflect the chemical composition reported.

- We thank the reviewer for this important comment. We agree that the original title overstated the chemical characterization of the test material, as the initial manuscript did not provide sufficient compositional evidence to support the term “flavanol-rich.” In response, we have revised the title to “Cocoa Powder Modulates HIF-1α Stability and Inhibits Ocular Angiogenic and Degenerative Pathology” to more accurately reflect the scope of the data and to avoid unsupported compositional claims. In addition, we added a chemical characterization section describing the LC–MS profiling of the cocoa powder used in this study and identified two major constituents, theobromine and caffeine. We also clarified in the Results and Discussion that the tested material was Dutch-processed cocoa powder, for which flavanol content is expected to be reduced during alkalization (see lines 497-505). These revisions better align the title and text with the analytical data presented in the revised manuscript.

[Line 497-505] However, an important consideration in interpreting the present data is that the cocoa powder used in this study was Dutch-processed cocoa powder, for which flavanol content is expected to be reduced during alkalization. Consistent with this, LC–MS analysis identified theobromine and caffeine as the major detectable constituents of the tested material (Fig. 1). Accordingly, the biological effects observed here should not be attributed solely to flavanols, but rather to the processed cocoa matrix as a whole, potentially involving remaining polyphenolic constituents. This compositional distinction should be considered when comparing the present findings with studies using flavanol-rich cocoa extracts or purified flavanol preparations.

 

2. It is difficult to understand why the ethanolic extract yielded only two prominent peaks in the LC–MS analysis (alkaloids), especially considering that the extract was obtained directly from 100% pure cocoa powder (as indicated by authors). A broader metabolite profile would be expected. The authors should clarify this result and discuss possible reasons (e.g., extraction conditions, detection parameters, or data processing).

- We thank the reviewer for the insightful comment. Cocoa is a rich source of flavanols and other polyphenolic compounds. For commercial use, cocoa is often treated with alkali, a process known as Dutch processing, to improve its color, dispersibility, and flavor. This process substantially reduces flavanols due to oxidation and polymerization reactions [1,2]. In this study, the cocoa powder was purchased from Dutch Cocoa BV, which supplies Dutch-processed cocoa. Therefore, only two major alkaloids were observed in the LC-MS chromatogram of the extract. We have added this explanation to the revised manuscript (see lines 122, 307-311).

1. Miller, K. B.; Hurst, W. J.; Payne, M. J.; Stuart, D. A.; Apgar, J.; Sweigart, D. S.; Ou, B. Impact of Alkalization on the Antioxidant and Flavanol Content of Commercial Cocoa Powders. Journal of Agricultural and Food Chemistry 2008, 56 (18), 8527–8533. https://doi.org/10.1021/jf801670p.
2. ‌ Greño, M.; Herrero, M.; Cifuentes, A.; Marina, M. L.; Castro-Puyana, M. Assessment of Cocoa Powder Changes during the Alkalization Process Using Untargeted Metabolomics. LWT 2022, 172, 114207. https://doi.org/10.1016/j.lwt.2022.114207

[Line 307-311] Notably, the CP used in this study was Dutch processed CP, which is treated with alkali. This process has been re-ported to markedly reduce flavanols due to oxidation and polymerization [28,29]. Accordingly, these alkaloids represent characteristic constituents of CP and support the chemical identity of the material used in this study (Figure 1 and Table 1).

 

3. In line 124, the authors state that a water extract was used for cell-based experiments, whereas the chemical characterization was performed using an ethanolic extract. The reasons for characterizing an extract that was not used in the biological assays should be clearly explained

- We thank the reviewer’s valuable comment. The ethanolic extract was used for chemical characterization to enable the detection of a broader range of compounds, including moderately polar constituents. In contrast, the aqueous extract was used for cell-based experiments to reflect physiologically relevant conditions.

As per the reviewer’s comment, we prepared an extract of cocoa powder using 100% aqueous solvent. We then performed LC–MS analysis to ensure consistency between the analytical and biological experiments. As a result, the chemical profiles of the 30% EtOH extract and the 100% aqueous extract were comparable, with both showing consistent detection of theobromine and caffeine (see Figure 1). Therefore, the use of different extraction solvents does not affect the interpretation of the bioactivity results. We have added this explanation to the revised manuscript (see lines 302-307).

[Line 302-307] For chemical characterization, both the 30% ethanolic extract and the aqueous preparation of cocoa powder were analyzed by LC–MS. In both preparations, two predominant peaks corresponding to theobromine and caffeine were identified by comparison with authentic standards. These results indicate that the major detectable constituents were consistent between the analytically characterized extract and the preparation used for cell-based assays.

 

Supplementary Figure S1. Comparison of LC-MS chromatogram of 30% EtOH extract and the 100% aqueous extract

Extraction solvent		LC-MS chromatogram
30% EtOH extract	195 nm
	TIC
100% aqueous extract	195 nm
	TIC

 

4. Related to this, if the chemical characterization is intended to represent the composition of the tested material, it is inconsistent that the title highlights flavanols (“Flavanol-Rich Cocoa Powder”) instead of the compounds actually identified (mainly alkaloids). This discrepancy should be addressed.

- We appreciate the reviewer’s comment. In response, we have revised the title by removing “flavanol-rich,” as it does not reflect the compounds identified in this study.

 

5. Lines 533–534: No supplementary material was found. 

-We thank the reviewer for this comment. The supplementary materials were not properly included in the previous submission. We have now prepared and uploaded the supplementary file, which includes Fig. S1 (peak information of major compounds analyzed by LC-MS) and Table S1 (primer sequences and expected product sizes used for qPCR analysis). The corresponding statement has also been revised in the manuscript.

 

6. Additionally, the authors should justify the use of a 30% ethanol extraction (line 150), including its relevance for the compounds of interest.

- We thank the reviewer for this comment. The 30% ethanol extraction was selected for chemical characterization of cocoa powder under the LC–MS conditions used in this study. This solvent system was chosen to provide an aqueous-organic environment suitable for recovering the major readily detectable small molecules present in the sample, while maintaining compatibility with chromatographic analysis. In particular, it enabled reliable detection of the two major compounds identified in our study, theobromine and caffeine, by comparison with authentic standards.

 

7. Finally, if additional phytochemical information is available, it is strongly recommended to include it in a table to improve the characterization and understanding of the cocoa powder constituents.

- We thank the reviewer’s insightful comment. However, because Dutch-processed cocoa powder was used in this study, many phytochemical constituents were either significantly reduced or not detectable under our experimental conditions. Therefore, no additional phytochemical data were available to be presented in a table. To support this, the raw LC–MS chromatograms of the cocoa extract have been provided (see Figure 1).

Figure 1. raw LC–MS chromatograms of cocoa extract, showing UV detection at 195 nm and TIC in positive ion mode (ESI+)

195 nm
TIC (+)

 

Reviewer #2

The manuscript is well written; however, several questions need to be addressed before publication.

1. It is unclear whether the authors performed two separate analyses using HPLC and LC–MS. If not, the authors should clearly explain the instrumentation and how the tandem systems were used.

- Thank you for the reviewer’s comment. The analysis was conducted using an LC-MS system and was not performed as separate experiments. Therefore, we have revised the manuscript by replacing “HPLC” with “LC-MS” to clarify the instrumentation. (see Fig. S1.)

 

2. In addition, how was the same concentrated sample injected into the quadrupole MS if it was first detected using a UV detector, as such concentrations typically oversaturate the MS.

- We thank the reviewer’s valuable comment. The analysis was performed using an LC-MS system, allowing simultaneous UV and MS detection from a single injection. While the sample concentration (1 mg/mL) may appear high for MS analysis, no evidence of detector saturation or ion suppression was observed under our experimental conditions. When a lower concentration (0.1 mg/mL) was tested, no detectable caffeine peak (t_R = 7.4 min) was observed in the UV chromatogram at 195 nm or in the TIC (+) chromatogram (see Figure 2). Therefore, a higher concentration was required to ensure sufficient signal intensity for both UV and MS detection.

Figure 2. UV (195 nm) and TIC chromatograms at two concentrations (1 mg/mL and 0.1 mg/mL)

Concentration		Expanded LC-MS chromatogram (3–10 min)
1 mg/mL	195 nm
	TIC (+)
0.1 mg/mL	195 nm
	TIC (+)

 

2. The manuscript also does not clearly state how many mice were used in the study.

- We apologize for the lack of clarity regarding the animal numbers. As suggested by the reviewer, we have now explicitly stated the number of mice used in each experimental group in the Materials and Methods section of the revised manuscript. Specifically, we have added that n = 9 mice per group were used for the MNU-induced model (Section 2.6) and n = 7–9 mice per group were used for the alkali-induced corneal injury model (Section 2.9).

 

3. Finally, the authors performed a total proteomic assay using BSA; it would be more informative to quantify and identify the upregulated and downregulated proteins associated with immune function.

Before 3

After

 

-We sincerely appreciate the reviewer’s insightful comment regarding protein analysis. We would like to clarify that Fig. 5B does not represent a total proteomic assay, but rather targeted Western blot analyses of selected proteins (MMP2, MMP9, VEGF, α-SMA, and Ninj1) to evaluate key pathways related to inflammation and tissue remodeling. We apologize for any confusion caused by this presentation.

To more directly address the reviewer’s suggestion regarding the quantification of immune-related responses, we additionally performed TNF-α ELISA and included these results in Fig. 5C. As shown in Fig. 5B, 1N NaOH exposure significantly increased the expression of proteins associated with inflammation and extracellular matrix remodeling, whereas CP treatment reduced their expression in a dose-dependent manner. Consistently, TNF-α levels were markedly elevated following NaOH treatment and were significantly decreased by CP treatment, further demonstrating its immunomodulatory effect.

These protein-level findings are in agreement with the mRNA expression data presented in Fig. 5A, collectively supporting that CP effectively suppresses NaOH-induced inflammatory and immune responses. Accordingly, we have revised the manuscript and figure legends to clearly distinguish between targeted protein analysis and immune-related cytokine quantification.

 

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript addresses a highly interesting topic in the field of nutrition and functional food science. However, several aspects require clarification and improvement.

The title refers to “Flavanol-Rich Cocoa Powder”; however, no chemical profiling is provided to support this claim. Instead, the authors emphasize the presence of alkaloids. The title should therefore be reconsidered to better reflect the chemical composition reported.

It is difficult to understand why the ethanolic extract yielded only two prominent peaks in the LC–MS analysis (alkaloids), especially considering that the extract was obtained directly from 100% pure cocoa powder (as indicated by authors). A broader metabolite profile would be expected. The authors should clarify this result and discuss possible reasons (e.g., extraction conditions, detection parameters, or data processing).

In line 124, the authors state that a water extract was used for cell-based experiments, whereas the chemical characterization was performed using an ethanolic extract. The reasons for characterizing an extract that was not used in the biological assays should be clearly explained

Related to this, if the chemical characterization is intended to represent the composition of the tested material, it is inconsistent that the title highlights flavanols (“Flavanol-Rich Cocoa Powder”) instead of the compounds actually identified (mainly alkaloids). This discrepancy should be addressed.

Lines 533–534: No supplementary material was found. 

Additionally, the authors should justify the use of a 30% ethanol extraction (line 150), including its relevance for the compounds of interest.

Finally, if additional phytochemical information is available, it is strongly recommended to include it in a table to improve the characterization and understanding of the cocoa powder constituents.

Author Response

The revised part in the text is in blue color.

 

Reviewer #1

Comments and Suggestions for Authors

The manuscript addresses a highly interesting topic in the field of nutrition and functional food science. However, several aspects require clarification and improvement.

 

1. The title refers to “Flavanol-Rich Cocoa Powder”; however, no chemical profiling is provided to support this claim. Instead, the authors emphasize the presence of alkaloids. The title should therefore be reconsidered to better reflect the chemical composition reported.

- We thank the reviewer for this important comment. We agree that the original title overstated the chemical characterization of the test material, as the initial manuscript did not provide sufficient compositional evidence to support the term “flavanol-rich.” In response, we have revised the title to “Cocoa Powder Modulates HIF-1α Stability and Inhibits Ocular Angiogenic and Degenerative Pathology” to more accurately reflect the scope of the data and to avoid unsupported compositional claims. In addition, we added a chemical characterization section describing the LC–MS profiling of the cocoa powder used in this study and identified two major constituents, theobromine and caffeine. We also clarified in the Results and Discussion that the tested material was Dutch-processed cocoa powder, for which flavanol content is expected to be reduced during alkalization (see lines 497-505). These revisions better align the title and text with the analytical data presented in the revised manuscript.

[Line 497-505] However, an important consideration in interpreting the present data is that the cocoa powder used in this study was Dutch-processed cocoa powder, for which flavanol content is expected to be reduced during alkalization. Consistent with this, LC–MS analysis identified theobromine and caffeine as the major detectable constituents of the tested material (Fig. 1). Accordingly, the biological effects observed here should not be attributed solely to flavanols, but rather to the processed cocoa matrix as a whole, potentially involving remaining polyphenolic constituents. This compositional distinction should be considered when comparing the present findings with studies using flavanol-rich cocoa extracts or purified flavanol preparations.

 

2. It is difficult to understand why the ethanolic extract yielded only two prominent peaks in the LC–MS analysis (alkaloids), especially considering that the extract was obtained directly from 100% pure cocoa powder (as indicated by authors). A broader metabolite profile would be expected. The authors should clarify this result and discuss possible reasons (e.g., extraction conditions, detection parameters, or data processing).

- We thank the reviewer for the insightful comment. Cocoa is a rich source of flavanols and other polyphenolic compounds. For commercial use, cocoa is often treated with alkali, a process known as Dutch processing, to improve its color, dispersibility, and flavor. This process substantially reduces flavanols due to oxidation and polymerization reactions [1,2]. In this study, the cocoa powder was purchased from Dutch Cocoa BV, which supplies Dutch-processed cocoa. Therefore, only two major alkaloids were observed in the LC-MS chromatogram of the extract. We have added this explanation to the revised manuscript (see lines 122, 307-311).

1. Miller, K. B.; Hurst, W. J.; Payne, M. J.; Stuart, D. A.; Apgar, J.; Sweigart, D. S.; Ou, B. Impact of Alkalization on the Antioxidant and Flavanol Content of Commercial Cocoa Powders. Journal of Agricultural and Food Chemistry 2008, 56 (18), 8527–8533. https://doi.org/10.1021/jf801670p.
2. ‌ Greño, M.; Herrero, M.; Cifuentes, A.; Marina, M. L.; Castro-Puyana, M. Assessment of Cocoa Powder Changes during the Alkalization Process Using Untargeted Metabolomics. LWT 2022, 172, 114207. https://doi.org/10.1016/j.lwt.2022.114207

[Line 307-311] Notably, the CP used in this study was Dutch processed CP, which is treated with alkali. This process has been re-ported to markedly reduce flavanols due to oxidation and polymerization [28,29]. Accordingly, these alkaloids represent characteristic constituents of CP and support the chemical identity of the material used in this study (Figure 1 and Table 1).

 

3. In line 124, the authors state that a water extract was used for cell-based experiments, whereas the chemical characterization was performed using an ethanolic extract. The reasons for characterizing an extract that was not used in the biological assays should be clearly explained

- We thank the reviewer’s valuable comment. The ethanolic extract was used for chemical characterization to enable the detection of a broader range of compounds, including moderately polar constituents. In contrast, the aqueous extract was used for cell-based experiments to reflect physiologically relevant conditions.

As per the reviewer’s comment, we prepared an extract of cocoa powder using 100% aqueous solvent. We then performed LC–MS analysis to ensure consistency between the analytical and biological experiments. As a result, the chemical profiles of the 30% EtOH extract and the 100% aqueous extract were comparable, with both showing consistent detection of theobromine and caffeine (see Figure 1). Therefore, the use of different extraction solvents does not affect the interpretation of the bioactivity results. We have added this explanation to the revised manuscript (see lines 302-307).

[Line 302-307] For chemical characterization, both the 30% ethanolic extract and the aqueous preparation of cocoa powder were analyzed by LC–MS. In both preparations, two predominant peaks corresponding to theobromine and caffeine were identified by comparison with authentic standards. These results indicate that the major detectable constituents were consistent between the analytically characterized extract and the preparation used for cell-based assays.

 

Supplementary Figure S1. Comparison of LC-MS chromatogram of 30% EtOH extract and the 100% aqueous extract

Extraction solvent		LC-MS chromatogram
30% EtOH extract	195 nm
	TIC
100% aqueous extract	195 nm
	TIC

 

4. Related to this, if the chemical characterization is intended to represent the composition of the tested material, it is inconsistent that the title highlights flavanols (“Flavanol-Rich Cocoa Powder”) instead of the compounds actually identified (mainly alkaloids). This discrepancy should be addressed.

- We appreciate the reviewer’s comment. In response, we have revised the title by removing “flavanol-rich,” as it does not reflect the compounds identified in this study.

 

5. Lines 533–534: No supplementary material was found. 

-We thank the reviewer for this comment. The supplementary materials were not properly included in the previous submission. We have now prepared and uploaded the supplementary file, which includes Fig. S1 (peak information of major compounds analyzed by LC-MS) and Table S1 (primer sequences and expected product sizes used for qPCR analysis). The corresponding statement has also been revised in the manuscript.

 

6. Additionally, the authors should justify the use of a 30% ethanol extraction (line 150), including its relevance for the compounds of interest.

- We thank the reviewer for this comment. The 30% ethanol extraction was selected for chemical characterization of cocoa powder under the LC–MS conditions used in this study. This solvent system was chosen to provide an aqueous-organic environment suitable for recovering the major readily detectable small molecules present in the sample, while maintaining compatibility with chromatographic analysis. In particular, it enabled reliable detection of the two major compounds identified in our study, theobromine and caffeine, by comparison with authentic standards.

 

7. Finally, if additional phytochemical information is available, it is strongly recommended to include it in a table to improve the characterization and understanding of the cocoa powder constituents.

- We thank the reviewer’s insightful comment. However, because Dutch-processed cocoa powder was used in this study, many phytochemical constituents were either significantly reduced or not detectable under our experimental conditions. Therefore, no additional phytochemical data were available to be presented in a table. To support this, the raw LC–MS chromatograms of the cocoa extract have been provided (see Figure 1).

Figure 1. raw LC–MS chromatograms of cocoa extract, showing UV detection at 195 nm and TIC in positive ion mode (ESI+)

195 nm
TIC (+)

 

Reviewer #2

The manuscript is well written; however, several questions need to be addressed before publication.

1. It is unclear whether the authors performed two separate analyses using HPLC and LC–MS. If not, the authors should clearly explain the instrumentation and how the tandem systems were used.

- Thank you for the reviewer’s comment. The analysis was conducted using an LC-MS system and was not performed as separate experiments. Therefore, we have revised the manuscript by replacing “HPLC” with “LC-MS” to clarify the instrumentation. (see Fig. S1.)

 

2. In addition, how was the same concentrated sample injected into the quadrupole MS if it was first detected using a UV detector, as such concentrations typically oversaturate the MS.

- We thank the reviewer’s valuable comment. The analysis was performed using an LC-MS system, allowing simultaneous UV and MS detection from a single injection. While the sample concentration (1 mg/mL) may appear high for MS analysis, no evidence of detector saturation or ion suppression was observed under our experimental conditions. When a lower concentration (0.1 mg/mL) was tested, no detectable caffeine peak (t_R = 7.4 min) was observed in the UV chromatogram at 195 nm or in the TIC (+) chromatogram (see Figure 2). Therefore, a higher concentration was required to ensure sufficient signal intensity for both UV and MS detection.

Figure 2. UV (195 nm) and TIC chromatograms at two concentrations (1 mg/mL and 0.1 mg/mL)

Concentration		Expanded LC-MS chromatogram (3–10 min)
1 mg/mL	195 nm
	TIC (+)
0.1 mg/mL	195 nm
	TIC (+)

 

2. The manuscript also does not clearly state how many mice were used in the study.

- We apologize for the lack of clarity regarding the animal numbers. As suggested by the reviewer, we have now explicitly stated the number of mice used in each experimental group in the Materials and Methods section of the revised manuscript. Specifically, we have added that n = 9 mice per group were used for the MNU-induced model (Section 2.6) and n = 7–9 mice per group were used for the alkali-induced corneal injury model (Section 2.9).

 

3. Finally, the authors performed a total proteomic assay using BSA; it would be more informative to quantify and identify the upregulated and downregulated proteins associated with immune function.

Before 3

After

 

-We sincerely appreciate the reviewer’s insightful comment regarding protein analysis. We would like to clarify that Fig. 5B does not represent a total proteomic assay, but rather targeted Western blot analyses of selected proteins (MMP2, MMP9, VEGF, α-SMA, and Ninj1) to evaluate key pathways related to inflammation and tissue remodeling. We apologize for any confusion caused by this presentation.

To more directly address the reviewer’s suggestion regarding the quantification of immune-related responses, we additionally performed TNF-α ELISA and included these results in Fig. 5C. As shown in Fig. 5B, 1N NaOH exposure significantly increased the expression of proteins associated with inflammation and extracellular matrix remodeling, whereas CP treatment reduced their expression in a dose-dependent manner. Consistently, TNF-α levels were markedly elevated following NaOH treatment and were significantly decreased by CP treatment, further demonstrating its immunomodulatory effect.

These protein-level findings are in agreement with the mRNA expression data presented in Fig. 5A, collectively supporting that CP effectively suppresses NaOH-induced inflammatory and immune responses. Accordingly, we have revised the manuscript and figure legends to clearly distinguish between targeted protein analysis and immune-related cytokine quantification.

 

 

Author Response File: Author Response.pdf