Live, Probiotic, or Neither? Microbial Composition of Retail-Available Kombucha and “Hard” Kombucha in the Pacific Northwest of the United States

Round 1

Reviewer 1 Report

Dear Authors,

I reviewed the revised manuscript entitled “Live, probiotic, or neither? Microbial composition of retail-available kombucha and “hard” kombucha in the Pacific Northwest of the United States”. This study is based on microbiological and chemical analysis of kombucha products and derivates from a region of the USA. The aim was to assess the existence of correlation between the composition (microbial and chemical) and the type or claims of the products. In essence, the results have shown many discrepancies with more of a spectrum of product profiles than actual clear categories. This highlights low control of kombucha producers over their products’ quality, especially after commercialization.

Overall, the manuscript is well written, and the context of the kombucha market is especially well understood and explained. My main criticism addresses some tables and figures that could benefit some clarifications. Please refer to my specific comments in the pdf attached.

I must mention how I appreciated reviewing this manuscript for its quality. This is a very valuable contribution for the scientific knowledge of kombucha and other similar beverages. This study should be particularly useful and important for the growth and maturation of the kombucha branch.

Comments for author File: Comments.pdf

Author Response

Figure 1. Misspelt “pasteurization”; Change “EtOH” to “ethanol”

Author: These corrections have been made

 

Table 1: Remove “…”

Author: These corrections have been made

 

Line 115-116: Please add composition of the agar media, or cite the appropriate. Please add developed name (WLN and NA).

Author: The composition of the agar had previously been described in section 2.1. Full names and citations of media have been added to sections 2.1 and 2.3.

 

Line 236: In my opinion Figure S2, the MANOVA analysis does not bring much new information. I actually believe that the following interpretation you make based on Table S4 also fits with Figure 2. If Figure S2 brings key new information, please stress it more. In Table S4, it is unclear to me what the values presented are, as well as the intervals. Are those average values and standard deviation?

Author: The reviewer brought to our attention that the original description of Figure 2 did not fully distinguish the information it provides from that presented in Table S4. In response, we have revised the text to provide a more comprehensive explanation of the distinct perspective that Figure 2 offers on the results presented in the paper. We appreciate the reviewer's helpful feedback in prompting us to clarify and expand upon this aspect of the analysis.

We amended the following in response: “The PCA plot shows a clear separation of samples based on their anaerobic fermentation products (lactic acid and glycerol) versus residual sugars and acetic acid. Hard kombucha samples cluster to the left of the biplot while soft kombucha samples are clustered to the right. Interestingly, the major analytes present in soft-aberrant kombucha do not appear to be intermediates between hard and soft kombucha. Instead, some soft-aberrant samples cluster with hard kombucha samples, while others cluster with soft kombucha samples, along the first principal component.”

 

Line 236: In Table S4, it is unclear to me what the values presented are, as well as the intervals. Are those average values and standard deviation?

Author: The caption was edited to clarify the meaning of the values depicted. We amended to the following: Results of NMR H1 analysis, comparing the chemical composition of hard, soft, and soft-aberrant Kombucha products based on the mean and standard deviation (X +/- SD) of key analytes. The Kruskal-Wallis test was used to determine significant differences in means of these analytes among the different types of Kombucha.

 

Figure 2: This is a minor suggestion, but would it be possible, for all relevant figures, to move the "type" / "value" /"claim" labels above or beneath the legend?  I find it confusing to have everything on the same line. I believe this modification could improve the readability.

Author: The reviewer draws attention to improvements that could be made to enhance the clarity of Figure 2. In considering the suggestion, we determined that across all figures where legends were used the “title” of each legend was superfluous. Each figure panel was sufficiently described in figure captions. Therefore, we removed the legend title elements across all figures, which we feel has simplified their presentation.

 

Line 257: I have troubles understanding discernable growth and non-discernable growth concepts as well as the table itself because of empty line labels. Please clarify.

Author: We thank the reviewer for pointing out the lack of clarity in Table S5 making it difficult to follow our reporting of “discernable growth” and “non discernable growth”. Supplemental Table 5 was adjusted to include line labels and the caption was adjusted to clarify the measures as follows: Non-discernable growth (NDG) was determined as the proportion of colony counts (CFU/ml) that were less than 3 log CFU/ml. To quantify NDG for different claim types (i.e., “none”, “live”, and “probiotic”), the number of plates (NDG|n) and the percentage of total plates (NDG|%) were calculated. Discernable growth (DG) was determined as the proportion of colony counts (CFU/ml) that were equal to or greater than 3 log CFU/m. To quantify DG for different claim types (i.e., “none”, “live”, and “probiotic”), the number of plates (DG|n) and the percentage of total plates (DG|%) were calculated.

 

Line 258: Please check Table S6 for errors

Author: After checking the table as suggested by the reviewer, we noticed that there was a discrepancy with how the comparison of colony counts on MRS media between products with organisms listed on the label and those without (yes-no). The results from the Dunn Post-Hoc test were not included. The table has been adjusted to correct that oversight.

 

Line 259: I suggest to rephrase " CFU counts were significantly lower in hard kombucha samples .... only on M13 media "

Author: The suggestion was taken and the main text was revised.

 

Line 266: Please replace "N" by "Number" in the table for more clarity.

Author: The reviewers suggested revision to Supplemental Table 7 was taken.

 

Figure 3: Please define max SPC. it is difficult to understand what the light blue bars based on the Figure title.

Author: The figure caption was revised to include “Maximum (MAX) growth determined by the highest SPC count across the media types”.

 

Figure 4: I am confused by the way the barplots are organized. They seem to be sorted by similarity ? So the "recommended" reading would be from the inside towards the outside. This leads me to interprete that no obvious correlation can be seen between microbial profiles and product features. Is it the right way to approch this figure? I believe more explanation on how to read the figure should be included in the main text. As it is I believe that this figure does not have the emphasis that it deserves.

Better readability might be achieved with more consistent color associations. For exemple : acetic acid bacteria genra could be associated to hues of warm colors (red or orange) and lactic acid bacteria to cold colors etc... This could ease the reading a little bit. This comment is not fundamental thought.

Author: The reviewer’s comment has been received and was taken in consideration.

To clarify how the samples are organized, the methods section 2.7. was amended to include the following: “The samples in the NMDS plot are sorted based on their pairwise similarities, with closer samples representing more similar microbial communities. The ordering is determined by minimizing the stress value, which measures the discrepancy between the pairwise distances in the plot and the original distances between the samples.”

Considering the reviewer’s comments about figure 4 not receiving sufficient emphasis we rewrote sections 3.3 and 3.4 to better incorporate description of how samples clustered alongside PERMANOVA statistical analyses of beta-diversity.

 

Figure 6: I believe the ellipses are not present.

Author: The caption has been changed to reflect figure accurately.

 

Line 363: Please mention key parameters used in the cited study, namely the culture medium.

Author: The main text was revised following the reviewers’ comments to: “We found that hard kombucha contained more glycerol, and it is notable that the most prevalent fungal species associated with soft kombucha production, batch-growth of Brettanomyces bruxellensis [10,[11, produces negligible glycerol during alcoholic fermentation under anaerobiosis in minimal culture medium (5% glucose, 0.2% (NH4)2SO4, 0.5% KH2PO4, 0.04% MgSO4·7H2O and 0.1% yeast extract) [29.”

 

Line 364: I believe that mentionning key parameters  are important as well since they can be very different from those of kombucha fermentation

Author: In this instance the article we cite is a review which discusses observations across many different experimental conditions. We updated the text to acknowledge that no studies have directly assessed glycerol production by Brettanomyces or Saccharomyces under conditions of kombucha fermentation:

“While no studies have directly assessed whether these yeast species exhibit similar behavior under conditions of kombucha fermentation, we speculate that hard Kombucha producers inoculate a base Kombucha tea with starter cultures of S. cerevisiae given its widespread application in other alcoholic fermentations [16]”

 

Line 408: I would modulate the sentence, since perhaps not all LAB have poor survivability in kombucha. Indeed some may be more adapted to kombucha than others, especially if the added strains do not originate from kombucha.

Author: The main text was revised following the reviewers’ comments to:

“The lack of association between Lactobacilli on ingredient lists with Lactobacillus reads may reflect poor survivability of some lactic acid bacteria in kombucha, although it should be noted that not all LAB may have poor survivability, especially those strains that are more adapted to kombucha or those that are naturally present in the pellicle.”

 

Line 425: Since no kombucha signature species are present in gut microbiome profiles (no Brettanomyces, no acetic acid bacteria for example), I am very doubtful about the establishment of kombucha micro-organisms in  the gut microbiome. It is now well established that consumed live or even dead micro-organism do stimulate the gut microbiome through their cellular constituants ( from the cell wall namely) or their metabolites. Such established mecanisms are relevant in the modulation of the gut microbiome with associated impacts on the host's health. This applies very nicely to the consumption of fermented food. I suggest additional comments regarding this aspects in the main text and with citation of the appropriate references.

Author: The main text was revised following the reviewers’ comments to:

Given the acid pH range (2.5-4.0) of kombucha, it is reasonable to speculate that it contains microbial communities with greater potential to survive the gastric environment, though to date there is no evidence kombucha microbes become resident in the gut microbiome. Moreover, there is a lack of evidence that food is a more effective probiotic delivery than supplements, though some research has focused upon acid-adaptation of probiotic strains for this purpose [41]. Regardless, it is well-established that live or even dead microorganisms, through their cellular constituents or metabolites, can stimulate the gut microbiome, modulating its composition with associated impacts on the host's health [42]. Therefore, further research is needed to understand the extent to which kombucha consumption meets consumer expectations, and strategies should be developed to maximize the native microbial content of finished kombucha while addressing stability issues.”

 

Line 454: I believe it is important to mention that this constitutes quite a challenge.

Author: The main text was revised following the reviewers’ comments to: “The extent to which this meets consumer expectations is worthy of further research, along with strategies to maximize native microbial content of finished kombucha while addressing stability issues, which constitutes quite a challenge.”

 

Line 468: Please change to exponent

Author: This correction has been made.

 

Line 470: Do you mean "16S"?

Author: This correction has been made.

 

Line 475: Please put in italic

Author: This correction has been made.

 

Reviewer 2 Report

Commercially available Kombucha was collected and classified into soft, soft-aberrant and hard, and the characteristics were evaluated.

1. Clearly explain the criteria for soft, soft-aberrant and hard kombucha.

2. Why is B. anomalus relatively abundant in soft kombucha compared to hard kombucha? Does the soft-aberrant kombucha show an intermediate abundance of the main flora between the hard and soft types?

3. Organic acids mainly found in kombucha are acetic acid, glucuronic acid, and lactic acid. According to Table S4, gluconic acid but not glucuronic acid was analyzed. Does commercially available kombucha not contain glucuronic acid? Or is it not analyzed?

4. Lines 300-302. Notably, S. cerevisiae was more abundant in hard than soft-aberrant kombucha, while B. anomalus was more abundant in soft kombucha relative to hard kombucha samples.

Is it possible to estimate the species by the metabarcoding sequence used in the thesis? Genus level seems to be suitable.

Author Response

Commercially available Kombucha was collected and classified into soft, soft-aberrant and hard, and the characteristics were evaluated. Clearly explain the criteria for soft, soft-aberrant and hard kombucha.

Author: The criteria for classifying kombucha samples as either 'hard', ‘soft’ or 'soft aberrant' are clearly explained in the manuscript. Commercial producers typically label their products as either traditional (i.e., 'soft') or alcoholic (i.e., 'hard'). For hard kombucha, the classification criterion is maintained consistently throughout the study. Soft kombucha samples are divided into two groups based on their alcohol-by-volume (ABV) content, as measured by nuclear magnetic resonance spectroscopy and weighed against the theoretical legal limit of 0.5% +/- 0.05%, as described in Section [insert section number. The justification for this division is provided in the Results section [insert section number. Samples with ABV above the legal limit are classified as 'soft-aberrant', while those with ABV within the legal limit are categorized as 'soft' kombucha. We respectfully disagree with the reviewer's suggestion that this classification scheme requires further clarification, as the criteria and their rationale are explained in detail within the manuscript.

 

Why is B. anomalus relatively abundant in soft kombucha compared to hard kombucha? Does the soft-aberrant kombucha show an intermediate abundance of the main flora between the hard and soft types?

 

Author: The reviewer touches upon discussion point not elaborated upon enough. The main text was refined to place further emphasis on this point, as follows:

“Niche-engineering by S. cerevisiae through production of ethanol is known to cause a collapse in fungal diversity of wine fermentations and the dominance of S. cerevisiae in hard kombucha likely reflects its capacity to outcompete B. bruxellensis and B. anomalus [36]. Bacterial communities differed according to relative abundance of several genera, including the two most abundant; Gluconobacter and Bacillus. It was interesting to note that Gluconobacter comprised more sequencing reads than Komagataeibacter in retail-available soft kombucha, whereas Komagataeibacter is the consensus dominant bacterial genera in kombucha pellicle as well as in the liquid phase at the end of kombucha fermentation [13], [14],[37].  These observations indicate that different production practices during hard kombucha fermentation are responsible for the differences in the abundance of main taxa between hard, soft and soft-aberrant kombucha. Therefore, there is no evidence to suggest that soft-aberrant kombucha shows an intermediate abundance of the main fungal or bacterial taxa between hard and soft types.”

3. Organic acids mainly found in kombucha are acetic acid, glucuronic acid, and lactic acid. According to Table S4, gluconic acid but not glucuronic acid was analyzed. Does commercially available kombucha not contain glucuronic acid? Or is it not analyzed?

Author: The reviewer touches upon discussion point not elaborated upon enough.

The main text was refined to place further emphasis on this point, as follows: “We found that across soft, soft-aberrant and hard kombucha samples the main organic acids were acetic acid, gluconic acid, and lactic acid. We did not detect glucuronic acid, a finding in line with recent studies which may have implications for some health claims associated with consuming kombucha [32],[34]. However, it is important to note that our study only analyzed a limited number of commercially-available kombucha brands.

In Section 2.4.2., we have clarified which organic acids were included in our analysis, and defined our cutoff point to be 0.01 mg/L. We noted that glucuronic acid did not meet this cutoff.

 

4. Lines 300-302. Notably, S. cerevisiaewas more abundant in hard than soft-aberrant kombucha, while B. anomaluswas more abundant in soft kombucha relative to hard kombucha samples.

Is it possible to estimate the species by the metabarcoding sequence used in the thesis? Genus level seems to be suitable.

Author: The objective of our study is to evaluate the diversity of yeast and bacteria in different types of commercially-available kombucha. Our sequencing approach allows for the resolution of ITS at a species level, which provides a more comprehensive understanding of the variations in the fungal population. While it is true that genus level classification is often used in metabarcoding studies, the increased resolution afforded by species-level identification provides more nuanced insight into microbial communities.

Round 2

Reviewer 2 Report

All questions from the previous review have been resolved.