Review Reports - Kimchi Fermentation-Driven Detoxification of Flaxseed: Impact on Physicochemical Quality and Antioxidant Potential

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors repeatedly attribute the observed detoxification and hydrolysis of CGs and lignan disaccharides to the β-glucosidase activity of “kimchi-derived lactic acid bacteria” or “Lactobacillaceae consortium”, but do not present any experimental data to support this claim at the microbiological level. Please answer the following questions:
1. What is the taxonomic composition of the consortium used?
2. Which genera or species dominate during fermentation?
3. Whether the consortium is stable or changes over time;
4. Which specific microorganisms could be responsible for the degradation of cyanogenic glycosides?

This is particularly problematic because β-glucosidase activity is not a universal feature of all Lactobacillaceae, nor is it uniform in strength or substrate specificity across different genera and species (e.g., Leuconostoc, Lactiplantibacillus, Weissella, Pediococcus).

Metagenomic analysis (16S rRNA amplicon sequencing) or isolation and characterization of the dominant bacteria, followed by in vitro tests for β-glucosidase activity and/or CGs degradation capacity, would provide direct evidence of a causal relationship.

If experimental approaches are not possible, the authors should present a significantly more in-depth overview of the known microbial species in kimchi, clearly arguing which specific species are known to have β-glucosidase activity, including Tables, references, and descriptions: whether there are published data on detoxification of cyanogenic glycosides by these microorganisms; to what extent is it justified to extrapolate these data to the consortium used in the present study.
The lack of microbiological data significantly weakens the article's conclusions. In its current form, the study demonstrates an effect but does not provide a convincing explanation for the mechanism by which it is realized. This contradicts the ambitious claims for the “One Health” framework and the role of microbial ecology, which require precisely the integration of microbiological, biochemical, and functional data.

Author Response

Reviewer #1:

The article presents an interesting and potentially significant approach for biological detoxification of cyanogenic glycosides (CGs) in flaxseed by fermentation using kimchi as a functional matrix. The obtained results of the complete elimination of CGs and the concomitant increase in antioxidant potential are convincing at the phenotypic and metabolic levels. However, a significant methodological and conceptual shortcoming of the study is the lack of a characterization of the microbial community used. The authors repeatedly attribute the observed detoxification and hydrolysis of CGs and lignan disaccharides to the β-glucosidase activity of “kimchi-derived lactic acid bacteria” or “Lactobacillaceae consortium”, but do not present any experimental data to support this claim at the microbiological level. Please answer the following questions:

General Response: We sincerely appreciate the reviewer’s critical observation regarding the microbiological characterization of the study. We agree that attributing the detoxification mechanism solely to a "consortium," without taxonomic specification, weakens scientific depth. To address this, we have significantly revised the Materials and Methods and Discussion sections. While we did not perform new metagenomic sequencing for this specific kimchi batch, the CGDF used was prepared from a well-characterized consortium described in our previous work (Tse et al., 2020; Huang et al., 2023), which we have now explicitly detailed. Furthermore, to strengthen the phenotypic data, we have upgraded our statistical analysis to Duncan's Multiple Range Test (Figure 3) to ensure the reliability of the observed physicochemical changes. And we have added a comprehensive discussion and a new Table/Figure description referencing the specific β-glucosidase activities of Kimchi-derived genera (Leuconostoc, Lactiplantibacillus, Weissella) to explain the enzymatic mechanism of cyanogenic glycoside (CG) hydrolysis.

[4] Huang, C.; Tse, T.J.; Purdy, S.K.; et al. Depletion of cyanogenic glycosides in whole flaxseed via Lactobacillaceae fermentation. Food Chem. 2023, 403, 134441. https://doi.org/10.1016/j.foodchem.2022.134441.

[24] Tse, JT, J Shen, YY Shim, MJT Reaney (2020) Changes in bacterial populations and their metabolism over Tse, J.T.; Shen, J.; Shim, Y.Y.; Reaney, M.J.T. Changes in bacterial populations and their metabolism over ninety sequential cultures on wheat-based thin stillage, J. Agric. Food Chem. 2020, 68(16), 4717–4729, https://doi.org/10.1021/acs.jafc.9b07414.

Here are our point-by-point responses to your specific questions:

Comment: 1. What is the taxonomic composition of the consortium used?

Response: The consortium used for flaxseed fermentation (CGDF preparation) is a specific, stable culture isolated from wheat-based thin stillage and characterized in our previous studies. We have updated the manuscript to explicitly state its composition. It is dominated by the family Lactobacillaceae, specifically the genera Lactiplantibacillus (formerly L. plantarum), Limosilactobacillus, and Lentilactobacillus.

Manuscript Revision: We have clarified this in Section 2.2 (Preparation of Ground Flaxseed and CGDF). "To remove CGs, a specialized fermentation process was conducted using a defined Lactobacillaceae consortium previously isolated and characterized in our laboratory [4,24]. This consortium is predominantly composed of L. plantarum, Limosilactobacillus fermentum, and Lentilactobacillus spp., all of which possess β-glucosidase activity." [Line 112-115]

Comment: 2. Which genera or species dominate during fermentation?

Response: We have clarified the distinction between the pre-fermentation of flaxseed (CGDF production) and the secondary fermentation in the Kimchi matrix.

In CGDF Production: Lactiplantibacillus spp. maintain dominance, executing the "lock-and-key" hydrolysis of linustatin.
In Kimchi Fermentation: The process follows the classical succession: Leuconostoc mesenteroides dominates the early stage, followed by Lactiplantibacillus plantarum in the later stages (Weeks 4–6). We have expanded the Discussion to attribute the detoxification to their robust β-glucosidase activity.

Line 312–315 (Section 3.5): " The complete elimination of CGs is attributed to the robust β-glucosidase activity of the specific starter culture (e.g., L. plantarum and Leuconostoc mesenteroides), which selectively cleaves the glycosidic bonds of cyanogenic precursors [14,19,30].

Comment: 3. Whether the consortium is stable or changes over time

Response: The specific inoculum used for CGDF production has demonstrated remarkable stability. In our previous work (Tse et al., 2020), this consortium was re-cultured over 80 times on wheat-based thin stillage media without significant loss of metabolic functionality or shift in dominant taxa. This stability ensures the reproducibility of the detoxification step (72 h <10 mg/kg HCN). We have clarified the defined nature and stability of this consortium in the Materials and Methods section.

We have noted the stability of the inoculum in the Methods section (Section 2.2).

Line 112–113 (Section 2.2): "...using a defined Lactobacillaceae consortium previously isolated and characterized in our laboratory."
Line 230–232 (Discussion): "...starter culture stability tests... are essential... to guarantee that the β-glucosidase activity remains consistent without genetic drift."

Comment 4: Which specific microorganisms could be responsible for the degradation of cyanogenic glycosides?

Response: Based on the reviewer's suggestion, we have synthesized the literature to attribute the β-glucosidase activity to specific species.

Primary Candidate: Lactiplantibacillus plantarum (abundant in inoculum and Kimchi).
Secondary Candidates: Leuconostoc mesenteroides and Weissella cibaria. We have cited genomic evidence (Michlmayr et al., 2014) confirming that these strains possess the specific gene clusters (GH1/GH3 families) required for hydrolyzing amygdalin and linamarin.

Line 313–315 (Section 3.5): The complete elimination of CGs is attributed to the robust β-glucosidase activity of the specific starter culture (e.g., L. plantarum and Leuconostoc mesenteroides), which selectively cleaves the glycosidic bonds of cyanogenic precursors [14,19,30].

Comment 5 (Critique): The lack of microbiological data significantly weakens the article's conclusions. In its current form, the study demonstrates an effect but does not provide a convincing explanation for the mechanism by which it is realized. This contradicts the ambitious claims for the “One Health” framework and the role of microbial ecology, which require precisely the integration of microbiological, biochemical, and functional data.

Response: We acknowledge this limitation. To compensate for the lack of new sequencing data, we have provided a comprehensive literature-based overview linking the known metabolic capabilities of the Lactobacillaceae consortium (used in our CGDF preparation) to the observed phenotypic results (HCN reduction). We have significantly expanded the Materials and Methods and Discussion sections to explicitly attribute the observed detoxification to the robust β-glucosidase activity of specific genera (Lactiplantibacillus plantarum and Leuconostoc mesenteroides), and we reference established genomic studies (Michlmayr et al., 2014; Jung et al., 2014). Additionally, we have restructured Table 2 to clearly demonstrate the "Reduction Status," thereby proving the efficacy of the biological process based on phenotypic evidence.

Strengthened Statistical Rigor (Figure 3): We acknowledge the reviewer's concern regarding the rigorous verification of our data. To strengthen the reliability of the physicochemical changes observed (e.g., pH buffering, antioxidant protection), we have upgraded our statistical methodology. Instead of the previously used LSD test, we applied Duncan’s Multiple Range Test (DMRT) to rigorously validate the differences between the control and flaxseed-fortified groups.

As shown in the revised Figure 3, distinct letters (a–f) have been added above the bars to indicate statistical significance (p < 0.05). For instance, in the Vitamin C and Glucose data at Week 6, the CGDF 0.5% group is labeled with specific letters (e.g., 'a' or 'b') that are distinct from the Control group, statistically proving that the "Antioxidant Shield" effect is significant and not merely an observational trend.

Literature-Based Validation: We significantly expanded the Discussion to link the observed HCN reduction to the known metabolic capabilities of the Lactobacillaceae consortium used.
Explicit Limitations: We added a statement in the Conclusion that future work will include targeted multi-omics to fully characterize the ecosystem dynamics.

Line 199–200 (Statistical Analysis): "...followed by Duncan’s multiple range test for multiple comparisons to ensure statistical rigor."
Line 327–329 (Limitations): "...the lack of specific microbial enumeration and direct radical scavenging assays (e.g., DPPH, ABTS) represents a limitation of the present study."

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

1. It is recommended that the title be aligned solely with the detoxification process and physicochemical effects, so statements about broad health impacts should be avoided (in its current version, the title is quite ambitious, suggesting an impact on "One Health" that is not scientifically proven, but only discussed conceptually).
2. The abstract should not mix measured results with functional and Health inferences that were not experimentally evaluated in the study (the guidelines of the journal Foods (MDPI) should be reviewed for correct abstract writing).
3. In the case of keywords, some repeat the same terms as the title (more technical and specific terms should be included).
4. The introduction should include the research gap defined in relation to similar previous studies. On the other hand, the hypothesis is still vague; it should be formulated explicitly (including what is expected to be demonstrated beyond the elimination of CGs). The "One Health" framework is introduced before being methodologically justified, functioning more as a narrative than as a scientific axis. Correctly write and define the specific objectives in the final part of the introduction.
5. In the materials and methods, the flaxseed fermentation process is based on previous work; however, its limitations and industrial reproducibility are not discussed.
6. Given that CGDF proportions are evaluated within a kimchi formulation with a constant sum, the system is compositional, and a mixture design would have been methodologically more appropriate than a one-way ANOVA, as it would allow interactions to be modeled and support the assumed optimization.
7. Based on the previous comment, it can be inferred that the experimental design is limited (n = 3) for the number of treatments and times evaluated. The statistical analysis is insufficient; a one-way ANOVA is not appropriate for a design with multiple factors (e.g., flaxseed type × concentration × time). Statistical assumptions, effect size, and statistical power are also not reported.
8. The elimination of CGs is demonstrated by ¹H-NMR, but is not validated by an orthogonal method (e.g., HPLC or enzymatic method).
9. The interpretation of "antioxidant shield" is speculative and is not supported by direct mechanistic assays.
10. Cardiovascular bioactivity is inferred without any in vitro/in vivo biological or functional evaluation.
11. The figures are mostly informative but overloaded and do not always clearly indicate statistical comparisons.
12. The discussions emphasize potential benefits but minimize inconsistencies and possible sensory or technological adverse effects (which were not included in the study).
13. Key instrumental analyses are missing in this type of study (such as fatty acid profiling (GC-FID) to confirm ALA preservation, quantification of lignan aglycones (HPLC-DAD/LC-MS), complementary antioxidant assays (ORAC, FRAP), and microbiological counts of LAB to correlate biochemistry and fermentation).
14. The conclusions exceed the data presented, especially in terms of human Health and One Health. The study's actual limitations and barriers to industrial scaling are not discussed.
15. Although the bibliography is extensive, several key references are outdated for certain concepts (antioxidants, functionality, among others).
16. It is recommended to reduce the similarity index in the materials and methods and results and discussions (according to the Ithenticate report 16%).

Comments on the Quality of English Language

Professional academic editing is recommended to adjust the tone to the Foods standards (MDPI).

Author Response

Reviewer #2:

General Response: We sincerely appreciate the reviewer’s thorough and critical evaluation. Your comments regarding the statistical design, the scope of the claims (One Health), and the need for rigorous validation have significantly improved the scientific quality of our manuscript. We have revised the title, abstract, and discussion sections to be more conservative and data-driven, and clarified the study's limitations.

Comment 1: It is recommended that the title be aligned solely with the detoxification process and physicochemical effects, so statements about broad health impacts should be avoided (in its current version, the title is quite ambitious, suggesting an impact on "One Health" that is not scientifically proven, but only discussed conceptually).

Response: Following Reviewer 2's feedback, we have reduced overly emphatic language (e.g., “proving One Health”) and shifted the focus to data-driven content. The title has been revised as follows:

Revised Title: “Kimchi Fermentation–Driven Detoxification of Flaxseed: Impact on Physicochemical Quality and Antioxidant Potential” [Line 1–3]

Comment 2: The abstract should not mix measured results with functional and Health inferences that were not experimentally evaluated in the study (the guidelines of the journal Foods (MDPI) should be reviewed for correct abstract writing).

Response: We have rewritten the abstract to focus solely on the measured results (HCN reduction, changes in acidity, nutrient retention). We removed speculative statements regarding specific anti-inflammatory or immune-supportive effects that were not tested in vivo/in vitro [Line 18-36] (Abstract section).

Comment 3: In the case of keywords, some repeat the same terms as the title (more technical and specific terms should be included).

Response: We have updated the keywords to include more specific technical terms related to the mechanism and analysis, avoiding repetition from the title [Line 17-39].

Comment 4: The introduction should include the research gap defined in relation to similar previous studies. On the other hand, the hypothesis is still vague; it should be formulated explicitly (including what is expected to be demonstrated beyond the elimination of CGs). The "One Health" framework is introduced before being methodologically justified, functioning more as a narrative than as a scientific axis. Correctly write and define the specific objectives in the final part of the introduction.

Response: We have rewritten the Introduction to clearly define the research gap: while physical/chemical detoxification methods exist, they often compromise nutritional integrity (the "nutritional trade-off"), and the direct application of biologically detoxified flaxseed within a complex kimchi matrix to create a bioactive shield has not been optimized. We have explicitly reformulated the hypothesis to extend beyond simple detoxification, postulating that the enzyme-mediated fermentation creates an "Antioxidant Shield" that protects unstable linolenic acid (ALA). Furthermore, we have methodologically justified the "One Health" framework as the integration of plant chemical safety, microbial ecology, and human metabolic well-being, rather than using it merely as a narrative device.

"Building on our previous research, which established a 72-h fermentation protocol using a specific Lactobacillaceae consortium to reduce HCN to undetectable levels, we incorporated this detoxified substrate into a standardized kimchi matrix. We hypothesize that the enzyme-mediated fermentation will not only ensure toxicological safety but also create an "Antioxidant Shield" that protects the unstable ALA from oxidation during storage. This approach aligns with the "One Health" framework by integrating plant chemical safety, microbial ecology, and human metabolic well-being into a unified, sustainable food innovation [22] [Line 90-97] (Last paragraph of Introduction)

Comment 5: In the materials and methods, the flaxseed fermentation process is based on previous work; however, its limitations and industrial reproducibility are not discussed.

Response: We have added a brief description in the Discussion section 3.1 acknowledging that while the 1 kg scale-up was successful, further industrial-scale validation (ton-scale) and starter culture stability tests are required for commercialization. We explicitly mention critical factors such as material-to-liquid ratio and inoculum density maintenance that must be controlled for successful industrial implementation.

Results and Discussion, Section 3.1, End of paragraph

In scaling this process from bench-scale (35 g) to industrial-scale (1 kg and beyond), several factors become critical: the material-to-liquid ratio, the initial inoculum density, and maintaining anaerobic conditions. While this study successfully demonstrated the detoxification of CGs in a 1 kg batch, further industrial-scale validation (ton-scale) is required to ensure consistent heat transfer and microbial stability. Additionally, starter culture stability tests during repeated subculturing are essential for commercial implementation to ensure that β-glucosidase activity remains consistent without genetic drift. [Line 226-232]

Comment 6: Given that CGDF proportions are evaluated within a kimchi formulation with a constant sum, the system is compositional, and a mixture design would have been methodologically more appropriate than a one-way ANOVA, as it would allow interactions to be modeled and support the assumed optimization.

Response: We acknowledge the reviewer’s expert observation regarding compositional data analysis and experimental design. To address this, we have clarified in the Materials and Methods that our experimental design was "additive" rather than a constrained "mixture" design. The CGDF (or CGDF) was added as a supplement to a fixed seasoning base (calculated relative to the seasoning weight), rather than replacing a major component to maintain a constant sum; thus, treating concentration levels as independent factors is methodologically valid.

Materials and Methods (Sections 2.4 and 2.7) [Line 156-159, 195, 199, 200]

The specific amounts added were 4.65 g, 9.30 g, and 18.60 g, respectively, per 930 g of seasoning mixture, not the final product weight.

All experiments were performed in triplicate (n = 3).

The differences between means were assessed using a one-way analysis of variance (ANOVA), followed by Duncan’s multiple range test for multiple comparisons to ensure statistical rigor.

Comment 7: Based on the previous comment, it can be inferred that the experimental design is limited (n = 3) for the number of treatments and times evaluated. The statistical analysis is insufficient; a one-way ANOVA is not appropriate for a design with multiple factors (e.g., flaxseed type × concentration × time). Statistical assumptions, effect size, and statistical power are also not reported.

Response: We apologize for the lack of clarity. We have explicitly stated that all experiments were performed in independent biological triplicates (n = 3). To ensure robust statistical power and control for false positives in multiple comparisons, we have upgraded our post-hoc analysis from LSD to Duncan’s multiple-range test. We utilized one-way ANOVA at each specific fermentation time point to clearly isolate the effect of CGDF concentration on physicochemical parameters.

[Line 190-195] Materials and Methods (Section 2.7)

All experiments were performed in triplicate (n = 3).
The differences between means were assessed using a one-way analysis of variance (ANOVA), followed by Duncan’s multiple range test for multiple comparisons to ensure statistical rigor.

Comment 8: The elimination of CGs is demonstrated by ¹H-NMR, but is not validated by an orthogonal method (e.g., HPLC or enzymatic method).

Response: We acknowledge the reviewer's comment regarding validation. However, 1H-NMR spectroscopy is a highly specific and sensitive method for detecting the distinct resonance peaks of linustatin (1.5 ppm) and neolinustatin (0.9 ppm), as established and validated in our previous analytical method development studies (Shim et al., 2016). Unlike HPLC, which often requires complex derivatization of these glycosides, NMR allows direct quantification without chemical modification, thereby minimizing artifact formation. In this study, the complete disappearance of these specific diagnostic peaks in the fermented samples provides conclusive evidence of degradation. We have cited the validated methodology to support the robustness of these findings.

[Line 203-212] (Results section 3.1)

The fermentation process successfully detoxified the flaxseed substrate. ¹H-NMR analysis revealed distinct resonance peaks characteristic of CGs at 1.5 ppm and 0.9 ppm in the untreated flaxseed extract (Figure 2). The raw flaxseed initially contained high concentrations of diglucoside, specifically linustatin (2,414 ± 24 mg/kg) and neolinustatin (3,300 ± 30 mg/kg), with a total HCN of 371 ± 25 mg/kg (Table 2). However, as Lactobacillaceae consortium fermentation proceeded, the area of these resonances decreased sharply. After 72 h of fermentation and the degumming process, even in a scaled-up 1 kg sample, ¹H-NMR analysis showed that resonance signals corresponding to CGs (linustatin and neolinustatin) were reduced to levels below 10 mg/kg HCN potential, indicating near-complete removal of cyanogenic precursors.

Comment 9: The interpretation of "antioxidant shield" is speculative and is not supported by direct mechanistic assays.

Response: We agree that "shield" is a metaphorical term. We have revised the text to explicitly propose this as a "sacrificial antioxidant" hypothesis (putative mechanism) rather than a proven fact. We explain this logically based on the observed data: the simultaneous retention of labile Vitamin C and glucose in the CGDF group compared to the control suggests a protective effect from the added phenolic aglycones. We have toned down the language to be less definitive and added a specific paragraph detailing this mechanism.

"Based on these observations, we propose a 'Sacrificial Antioxidant' hypothesis to explain the physical stability of the CGDF-kimchi (Section 3.3). Unlike liquid oil additives, which often accelerate lipid oxidation in acidic kimchi brines due to direct exposure, the cellular matrix of ground CGDF provides physical encapsulation. This structural integrity, combined with the continuous release of antioxidant aglycones, creates a stable 'lignan-protein-lipid matrix' that buffers lipid droplets against acid hydrolysis and intercepts oxidation chain reactions. This combined enzymatic and physical mechanism offers a plausible explanation for the extended optimal ripening window and enhanced oxidative stability observed in the CGDF group.” [357–365]

Comment 10: Cardiovascular bioactivity is inferred without any in vitro/in vivo biological or functional evaluation.

Response: We have removed direct claims of cardiovascular efficacy from the Abstract and Results section. In the Introduction and Discussion, we now strictly refer to previous literature establishing the link between ALA/lignans and cardiovascular health, framing our product as a source of these compounds (providing the potential for health benefits) rather than asserting it is a clinically proven remedy in this specific study.

Overall, this work provides foundational evidence for developing safe, nutritionally enhanced functional foods within the "One Health" framework, integrating food safety, microbial ecology, and improved bioactive compound availability. [Line 33-36]
The successful application of this bioprocess highlights kimchi fermentation as a versatile platform for activating other glycoside-rich botanicals (e.g., ginseng saponins, soy isoflavones) [30]. By converting plant defence compounds into bioavailable nutrients, this approach aligns with the "One Health" framework, bridging plant chemistry, microbial biotechnology, and human health [34]. [Line 366-371]

Comment 11: The figures are mostly informative but overloaded and do not always clearly indicate statistical comparisons.

Response: 1. Justification for ANOVA: regarding the experimental design, we clarified that the ground flaxseed and CGDF were added as supplements (0.5%, 1.0%, 2.0%) based on the weight of the seasoning mixture (930 g), rather than replacing a component in a closed mixture design (Lines 140–141). Therefore, treating these concentrations as independent factors in a one-way ANOVA is methodologically valid for this additive study.

Visualizing Significance (Figure 3): We agree that the previous figures lacked clarity regarding statistical comparisons. We have completely revised Figure 3 to include statistical significance markers derived from Duncan’s Multiple Range Test.

◦ Visual Update: Letters (a, b, c, d, e, f) are now placed above each bar.

◦ Interpretation: Bars sharing the same letter within the same fermentation week are not significantly different, whereas those with different letters differ at p < 0.05.

◦ Example: In the Total Acidity graph (Week 6), the CGDF 0.5% group shows a distinct statistical grouping compared to the Control, confirming its role in modulating acidification kinetics.

[Line 264–265, Figure 3 Caption] "Means with different letters within the same time point are significantly different (p< 0.05) according to Duncan's Multiple Range Test."

Comment 12: The discussions emphasize potential benefits but minimize inconsistencies and possible sensory or technological adverse effects...

Response: We have added a section in the Discussion acknowledging the technological limitations observed at higher concentrations. We explicitly state that while 0.5–1.0% addition levels improved fermentation stability and sensory scores, the 2.0% inclusion level led to undesirable texture (increased viscosity) and strong herbal flavors, which are limiting factors for consumer acceptance* [29]. *This clarifies that 1.0% is the practical upper limit for commercial application. The next instalment of this manuscript is planned for submission to the Journal of Foods, which will present the sensory evaluation results.

[29] Oh, S.C.; Shim, Y.Y.; Huang, C.; Ding, K.; Lee, S.J.; Kim, H-J.; Reaney, M.J.T.; Kim, Y.J. Removal of cyanogenic glycosides (CG) in flaxseed through fermentation: creating CG-free functional products, In Proceedings of the 2025 Korean Society of Food Science and Nutrition (KFN), International Symposium and Annual Meeting, Busan, Republic of Korea, 29–31 October 2025.

[Line 283-286] (Sensory discussion)

Conversely, concentrations exceeding 2.0% were observed to induce excessive viscosity and strong herbal notes, potentially limiting consumer acceptability despite high antioxidant levels [29]. Therefore, the 0.5–1.0% concentration range is suggested as the optimal inclusion level for balancing functional efficacy with sensory quality.

Comment 13: Key instrumental analyses are missing... (fatty acid profiling, lignan quantification, microbiological counts).

Response:

Fatty Acids: We have explicitly referenced our previous work (Huang et al., 2023), which utilized GC-FID and NMR lipid profiling to confirm that the fatty acid composition remains stable under these specific fermentation conditions. This justifies omitting redundant profiling in this study.
Lignans: Instead of HPLC, we used H-NMR spectroscopy to quantify increases in SDG and aglycone resonance signals, which served as a reliable proxy for lignan bioconversion.
Microbial Counts: We focused on metabolic outputs (acidity, pH, metabolite production) as the primary indicators of fermentation efficacy. We have added a statement in the Discussion section explicitly acknowledging the lack of specific microbial enumeration as a limitation to be addressed in future studies.
Regarding the analytical scope, the stability of the fatty acid profile during this specific fermentation process has been previously validated [4], justifying the focus on ALA retention in this study. This aligns with the findings of Son et al. [17], indicating that the enzymatic activity of kimchi LAB effectively hydrolyzes plant glycosides while the acidic environment facilitates the volatilization of HCN (Figure 4). [Line 318-323] (Discussion on limitations)

Comment 14: The conclusions exceed the data presented, especially in terms of human Health and One Health.

Response: We have rewritten the Conclusions section to focus solely on the study's successful technological achievements, specifically the biological detoxification of flaxseed and the preservation of unstable nutrients within the kimchi matrix. We have removed direct claims regarding human health outcomes that were not measured. The "One Health" concept is now framed solely as a theoretical framework for future functional food development and industrial implications, rather than a direct experimental result of this study.

In conclusion, this study demonstrates that 72-h Lactobacillaceae-mediated fermentation is a viable strategy to resolve the nutritional trade-off of flaxseed. This process achieved complete detoxification (HCN < 10 mg/kg) while significantly preserving ALA and concentrating lignans. When incorporated into Napa cabbage kimchi at 0.5–1.0% concentration, the resulting CGDF provided a distinct "Antioxidant Shield" effect. Contrary to the typical acceleration of acidification observed with nutrient enrichment, the CGDF matrix paradoxically extended the optimal ripening window by mitigating oxidative stress, thereby preserving labile nutrients such as vitamin C and glucose. A critical mechanistic observation is that kimchi fermentation facilitates the dual hydrolysis of both toxic CGs and lignan diglucoside, effectively transforming plant defence compounds into bioavailable nutrients. These findings lay the foundation for next-generation functional foods within the "One Health" framework. However, as direct radical scavenging assays (e.g., DPPH, ABTS) were not included in this study, the antioxidant potential was inferred primarily from metabolite profiles. Consequently, future research should focus on validating these findings through direct antioxidant kinetics, industrial-scale validation, in vivo bioavailability assessments, and targeted multi-omics approaches to fully characterize the breadth of this enzymatic activity. [Line 373-389] (Conclusion)

Comment 15: Although the bibliography is extensive, several key references are outdated...

Response: We have updated the bibliography by replacing older citations with recent reviews and experimental studies from 2020–2025 regarding antioxidants, fermented foods, and the One Health framework. Notable additions include recent publications in Foods (2025), Heliyon (2024), and One Health Outlook (2021) to ensure the manuscript reflects the current state of the art. Five or more recent references have been newly inserted.

[8] Shim, Y.Y.; Kim, J.H., Cho, J.Y.; Reaney, M.J.T. Health benefits of flaxseed and its peptides (Linusorbs), Crit. Rev. Food Sci. Nutr. 2024, 64(7), 1845–1864, https://doi.org/10.1080/10408398.2022.2119363.

[24] Tse, J.T.; Shen, J.; Shim, Y.Y.; Reaney, M.J.T. Changes in bacterial populations and their metabolism over ninety sequential cultures on wheat-based thin stillage, J. Agric. Food Chem. 2020, 68(16), 4717–4729, https://doi.org/10.1021/acs.jafc.9b07414.

[28] Jung, S.; Hwang, I.M.; Lee, J.H. Temperature impact on microbial and metabolic profiles in kimchi fermentation. Heliyon 2024, 10, e27174.

[20] Oh, S.C.; Shim, Y.Y.; Huang, C.; Ding, K.; Lee, S.J.; Kim, H-J.; Reaney, M.J.T.; Kim, Y.J. Removal of cyanogenic glycosides (CG) in flaxseed through fermentation: creating CG-free functional products, In Proceedings of the 2025 Korean Society of Food Science and Nutrition (KFN), International Symposium and Annual Meeting, Busan, Republic of Korea, 29–31 October 2025.

[30] Renchinkhand, G.; Park, Y.W.; Cho, S.-H.; Song, G.-Y.; Bae, H.C.; Choi, S.-J.; Nam, M.S. Identification of β-glucosidase activity of Lactobacillus plantarum CRNB22 in Kimchi and its potential to convert ginsenoside Rb1 from Panax ginseng. J. Food Biochem. 2015, 39, 155–163.

[32] Feng, C.; Wu, Y.; Cai, Z.; Song, Z.; Shim, Y.Y.; Reaney, M.J.T.; Wang, Y.; Zhang, N. A comparative study on flaxseed lignan biotransformation through resting cell catalysis and microbial fermentation. J. Sci. Food Agric. 2024, 104, 5869–5881. https://doi.org/10.1002/jsfa.13407.

[33] Lee, M.A.; Seo, H.Y.; Yang, J.H.; Jang, M.S. Effect of perilla oil addition on the quality and stability of kimchi during fermentation. J. Agric. Life Sci. 2013, 47, 255–266. https://doi.org/10.14397/jals.2013.47.6.255.

[34] Destoumieux-Garzón, D.; Mavingui, P.; Boetsch, G.; Boissier, J.; Darriet, F.; Duboz, P.; Fritsch, C.; Giraud, P.; Le Roux, F.; Morand, S.; et al. The One Health Concept: 10 Years Old and a Long Road Ahead. Front. Vet. Sci. 2018, 5, 14. https://doi.org/10.3389/fvets.2018.00014.

Comment 16: It is recommended to reduce the similarity index in the materials and methods and results and discussions (according to the Ithenticate report, 16%).

Response: We have carefully paraphrased the text throughout the manuscript to address the similarity index. Specifically, we have rewritten the Materials and Methods section, including the protocols for Kimchi Formulation (Section 2.4) and NMR-Based Metabolomic Analysis (Section 2.6), to reduce overlap with our previous publications. Additionally, the Introduction and Results and Discussion sections have been substantially revised to include new mechanistic insights (e.g., the 'biochemical switch' and 'antioxidant shield' concepts), which has naturally lowered the similarity with existing literature.

Location in Manuscript:

◦Section 1. Introduction: Paraphrased research gap and objectives.

◦Section 2. Materials and Methods: Specifically, Section 2.4 (Standardized Kimchi Formulation) and Section 2.6 (NMR-Based Metabolomic Analysis).

◦Section 3. Results and Discussion: Extensive revisions throughout the section.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The paper needs extensive revisions. Please find below major points of improvements:

Introduction

The distinction between scientific novelty and the mere application of an ingredient that has already been characterised is not adequately argued. The purpose and extent of the work's innovation must be clarified, particularly in relation to the One Health approach. The knowledge it aims to address is unclear; the objective appears more applicative than exploratory. The authors are encouraged to clarify this.

Material

How many species of the Lactobacillaceae family were used to obtain the consortium? Were the same strains used in the cited work? As the consortium has not been sufficiently characterised, it would be appropriate to supplement the information with details of the species composition and the specificity and enzymatic activity of the prevalent strains. The choice of CGDF concentrations (0.5%, 1.0% and 2.0%) is not justified. Please explain why these concentrations were chosen.

The number of experimental replicates is to be explained in detail in section 2.4.

The use of the LSD post hoc test is unjustified, as the limited number of replicates increases the risk of false positives.

There is an error in the numbering of section 2.5. “NMR-Based Metabolomic Analysis” due to the omission of section 2.6.

Results

The Results and Discussion section is too long and repetitive. Some concepts, such as the antioxidant role of CGDF, β-glucosidase and vitamin C protection, are repeated several times. Please make the results clearer.

The reduction in cyanogenic glucosides is reported as "almost complete" (line 187), despite the data only showing a decrease below the regulatory threshold. Please specify the relevant international regulatory threshold and cite any applicable legislation.

The increase in polyphenols and flavonoids is attributed to conversion to aglycones, but the specific aglycones are not directly measured. Please specify this statement more clearly.

The discussion overinterprets the data by proposing mechanistic explanations, such as pH, polyphenols, antioxidant protection and the effects induced by β-glucosidase. However, statistical analysis is not included in the graphs to support these assertions.

Chapter 2.5 discusses antioxidant enhancement, but it has been determined that this is not specified in the Materials and Methods section, since not all polyphenols and flavonoids have the same antioxidant activity or potential. The hypotheses described in the text should be discussed in more detail.

The discussion on proportional correlation (line 216) calls for validation through statistical analysis to support this conclusion.

The discussion (lines 243–246) states that 'NMR data confirm that CGDF acts as a functional buffer during microbial community transition'. However, within this study, the role of the microbial community in detoxification and metabolite modulation remains hypothetical due to a lack of specific microbiological or enzymatic measurements. This section should be revised and improved with better bibliographical support or by adding relevant data.

The One Health perspective is reiterated in the discussion, but it is not clear which outcomes are the most decisive.

Author Response

Reviewer #3:

The paper needs extensive revisions.

General Response: We sincerely appreciate the reviewers’ in-depth analysis and constructive criticism regarding the experimental design and interpretation of our study. We have carefully addressed each point by 1) clarifying the scientific novelty, 2) providing detailed information on the microbial consortium, and 3) upgrading the statistical analysis to Duncan's Multiple Range Test.

Please find below the major points of improvement:

Introduction

Comment 1: The distinction between scientific novelty and the mere application of an ingredient that has already been characterised is not adequately argued. The purpose and extent of the work's innovation must be clarified, particularly in relation to the One Health approach. The knowledge it aims to address is unclear; the objective appears more applicative than exploratory. The authors are encouraged to clarify this.

Response: We clarified that the novelty lies in the interaction between the detoxified matrix and the fermentation environment. Specifically, CGDF functions as a "bioactive stabilizer" (Antioxidant Shield) that paradoxically extends the ripening window and preserves labile nutrients.

[Line 72–74, Introduction] "In contrast, biological bioprocessing using lactic acid bacteria (LAB) offers a superior alternative by efficiently depleting CGs while preserving the substrate's primary nutritional integrity..."

Comment 2: Material

How many species of the Lactobacillaceae family were used to obtain the consortium? Were the same strains used in the cited work? As the consortium has not been sufficiently characterised, it would be appropriate to supplement the information with details of the species composition and the specificity and enzymatic activity of the prevalent strains. The choice of CGDF concentrations (0.5%, 1.0% and 2.0%) is not justified. Please explain why these concentrations were chosen.

Response:

Consortium Composition: We have clarified in the Materials and Methods that the consortium corresponds to the stable culture characterized in our previous study (Tse et al., 2020; Huang et al., 2023). It is dominated by the family Lactobacillaceae, specifically the genera Lactiplantibacillus (formerly L. plantarum), Limosilactobacillus (e.g., L. fermentum), and Lentilactobacillus. These strains were selected and maintained for their high β-glucosidase activity, which is the critical enzymatic mechanism for the "lock-and-key" hydrolysis of cyanogenic glycosides (linustatin/neolinustatin) into detoxified metabolites.

“To remove CGs, a specialized fermentation process was conducted using a defined Lactobacillaceae consortium previously isolated and characterized in our laboratory (Tse et al., 2020; Huang et al., 2023[4]). This consortium is predominantly composed of Lactiplantibacillus (formerly Lactobacillus) plantarum, Limosilactobacillus fermentum, and Lentilactobacillus species, which are known to possess high β-glucosidase activity.”

Justification of Concentrations: The 0.5–2.0% range was chosen to balance functional efficacy with sensory acceptability.

[Line 114–115, Methods] "This consortium is predominantly composed of L. plantarum, Limosilactobacillus fermentum, and Lentilactobacillus spp., all of which possess β-glucosidase activity."

[Line 285–286, Discussion] "Therefore, the 0.5–1.0% concentration range is suggested as the optimal inclusion level for balancing functional efficacy with sensory quality."

Comment 3: The number of experimental replicates is to be explained in detail in section 2.4.

Response: We have explicitly stated in Section 2.7 (Statistical Analysis) that all analytical measurements and fermentation batches were performed in triplicate (n = 3) to ensure data reliability and statistical validity.

“All experiments were performed in triplicate (n = 3). Data were analyzed using the Statistical Package for the Social Sciences..” [Line 195, 196 (Replicates)]

Comment 4: The use of the LSD post hoc test is unjustified, as the limited number of replicates increases the risk of false positives.

Response: We sincerely appreciate this statistical insight. We have removed the LSD post-hoc test to eliminate the risk of Type I errors (false positives).

[Line 199–200, Statistical Analysis] "...followed by Duncan’s multiple range test for multiple comparisons to ensure statistical rigor."

Comment 5: There is an error in the numbering of section 2.5. “NMR-Based Metabolomic Analysis” due to the omission of section 2.6.

Response: Thank you for pointing out this error. We have corrected the section numbering and organization in the revised manuscript to ensure logical flow. Section 2.5 now covers "Analytical Procedures" (including pH, acidity), Section 2.6 is dedicated to "NMR-Based Metabolomic Analysis," and Section 2.7 covers "Statistical Analysis."

Line 166, 173, 194: Corrected section numbering (2.5, 2.6, 2.7).

Comment 6: The Results and Discussion section is too long and repetitive. Some concepts, such as the antioxidant role of CGDF, β-glucosidase and vitamin C protection, are repeated several times. Please make the results clearer.

Response: We have streamlined the Results and Discussion section. We merged the repeated explanations of the "Antioxidant Shield" mechanism into a single, comprehensive discussion within Section 3.3, and removed redundant sentences linking β-glucosidase activity to vitamin C preservation in other sections.

Line 270–274: Consolidated discussion on the "Antioxidant Shield" and nutrient preservation.

Comment 7: The reduction in cyanogenic glucosides is reported as "almost complete" (line 187), despite the data only showing a decrease below the regulatory threshold. Please specify the relevant international regulatory threshold and cite any applicable legislation.

Response: We have corrected the phrasing to be quantitatively precise. The text now states that CGs were "reduced to concentrations equivalent to less than 10 mg/kg hydrogen cyanide potential," rather than "almost complete." We have explicitly cited the regulatory threshold of 10 mg/kg HCN, which aligns with the standards enforced by the Food Sanitation Act of Japan and the Ministry of Food and Drug Safety (MFDS) of South Korea for processed flaxseed products. This ensures the claims are directly supported by the analytical data (<10 mg/kg).

Section: 3.1. Efficiency of Biological Detoxification via Fermentation

◦ Line 309–310: "...reduced to concentrations equivalent to less than 10 mg/kg hydrogen cyanide potential..."

◦ Table 2 (Line 220): Distinctly shows "Reduction Status" and "< 10 mg/kg".

Comment 8: The increase in polyphenols and flavonoids is attributed to conversion to aglycones, but the specific aglycones are not directly measured. Please specify this statement more clearly.

Response: While we did not isolate specific aglycones via HPLC, our H-NMR metabolomics data showed spectral shifts in the aromatic region (6.2–7.8 ppm), consistent with the hydrolysis of SDG (glycoside) and the appearance of phenolic aglycones. We have revised Section 3.4 to state that the increase in TPC is attributed to this enzymatic conversion, supported by NMR spectral evidence.

Line 292–303: Discussion on bioconversion supported by NMR signals (6.20–7.80 ppm).

Comment 9: The discussion overinterprets the data by proposing mechanistic explanations... However, statistical analysis is not included in the graphs to support these assertions.

Response: 1. New Statistical Method: We re-analyzed all data using Duncan’s Multiple Range Test (DMRT), which provides stricter control for multiple mean comparisons.

Figure 3 Revision: The revised Figure 3 now explicitly displays these statistical results. As seen in the updated graph:

◦ In Fructose and Glucose panels (Week 6), the CGDF 0.5% group is marked with the letter 'a' or 'b', while the Control is marked with 'd' or 'e'.

◦ This statistically confirms that the retention of sugars in the CGDF group is significantly higher ( < 0.05) than in the Control, supporting our hypothesis of an "Antioxidant Shield" protecting labile nutrients.

Lines 198–199 (Statistical Analysis): "...assessed using a one-way analysis of variance (ANOVA), followed by Duncan’s multiple range test..."

◦ Lines 269–270 (Results): Paradoxically, the CGDF 0.5% group maintained significantly higher fructose and glucose concentrations at Week 6 than the control (p < 0.05).

◦ Lines 264–265 (Figure 3 Caption): Means with different letters (a-f) within the same time point are significantly different (p < 0.05) according to Duncan's Multiple Range Test.

Comment 10: Chapter 2.5 discusses antioxidant enhancement, but it has been determined that this is not specified in the Materials and Methods section, since not all polyphenols and flavonoids have the same antioxidant activity or potential. The hypotheses described in the text should be discussed in more detail. The discussion on proportional correlation (line 216) calls for validation through statistical analysis to support this conclusion.

Response: We respectfully acknowledge the reviewer’s comment. In this study, the discussion of "antioxidant enhancement" was primarily inferred from quantitative profiling of known antioxidant precursors (total polyphenolics, flavonoids, and SDG lignans) determined by 1H-NMR and colorimetric assays, rather than from direct radical-scavenging assays. Consequently, methods for DPPH and ABTS analysis were not included, and this specific kinetic data is currently unavailable in the presented dataset. We have revised the Discussion to clarify that our conclusions are based on the accumulation of bioactive substrates. Furthermore, we have explicitly stated in the Conclusion that direct validation of radical scavenging activity (DPPH & ABTS) is a limitation of the current work and will be conducted in future studies to robustly correlate metabolite profiles with functional antioxidant capacity.

Regarding the analytical scope, the stability of the fatty acid profile during this specific fermentation process has been previously validated (Huang et al., 2023), justifying the focus on ALA retention in this study. While H-NMR provided robust data on lignan bioconversion and metabolite flux, the lack of specific microbial enumeration and direct radical-scavenging assays (e.g., DPPH, ABTS) remains a limitation. Consequently, future research should focus on industrial-scale validation, in vivo bioavailability assessments, and targeted microbial enumeration to fully characterize the ecosystem dynamics.

◦ Line 327–329, 385: Explicit acknowledgement of the lack of DPPH/ABTS assays.

◦ Line 386–389 (Conclusion): "Future research should focus on validating these findings through direct antioxidant kinetics..."

Comment 11: The discussion (lines 243–246) states that 'NMR data confirm that CGDF acts as a functional buffer during microbial community transition'. However, within this study, the role of the microbial community in detoxification and metabolite modulation remains hypothetical due to a lack of specific microbiological or enzymatic measurements. This section should be revised and improved with better bibliographical support or by adding relevant data.

Response: We agree with the reviewer that using the word "confirm" is too definitive given the absence of 16S rRNA sequencing for this specific batch. We have revised the statement to be more scientifically prudent: "The metabolic profile obtained via NMR suggests that CGDF may modulate the fermentation environment, potentially influencing the microbial succession." To support this inference, we have cited recent literature (e.g., Jung et al., 2024) which links specific metabolite patterns, such as the sustained levels of glucose and fructose observed in our study, to the dominance of Lactobacillaceae and the suppression of heterofermentative spoilage organisms.

Line 272-274 (Discussion - Microbial Community)

"The metabolic profile obtained via ¹H-NMR suggests that CGDF may modulate the fermentation environment, potentially influencing the microbial succession."

Comment 12: The One Health perspective is reiterated in the discussion, but it is not clear which outcomes are the most decisive.

Response: We have sharpened the Conclusion to explicitly define the decisive One Health outcome: the study demonstrates a reproducible bioprocess that transforms a plant-based environmental hazard (CGs) into a safe, health-promoting food resource using beneficial microbial ecology. [Line 383–384, Conclusion] "These findings lay the foundation for next-generation functional foods within the 'One Health' framework."

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have responded to all my comments with arguments and have significantly improved the presentation. All missing details have been added, and I am completely satisfied with the manuscript in its current form. I propose acceptance.

Author Response

We hope these changes improve the manuscript's overall quality for publication.

Reviewer 2 Report

Comments and Suggestions for Authors

Accepted in the present form.

Author Response

Thank you for your patience and recommendations for strengthening our manuscript.

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript has been sufficienty improved for pubblication .

Author Response

Thank you for your patience and recommendations for strengthening our manuscript .