This is a really exciting and thoroughly investigated study, kudos to everybody involved! It's so exciting to see that fermentation-derived metabolites alone have such stark impacts on inflammation, gut microbiome composition/CFU count, and GLP-stimulation. Particularly for impacting the gut microbiome, this adds to the growing body of literature that just adding substrates (fiber) and/or probiotic strains alone don't really do anything in most cases, it really depends on what metabolites are produced.
Some comments: - This paper is incredibly acronym-heavy, and I found myself constantly scrolling up to remind myself what certain acronyms were. In paragraph for there is reference to "microbial ASVs" and I don't think the acronym was defined previously. Since it only appears once (I believe), I think it would be best just to spell out the full definition, especially because this could be confused with 16S ASVs. In the headers for each result I think it would be fine to say "Wild sauerkraut metabolites" instead of the acronym, because then the main heading/result reads a lot easier. Additionally in the methods there is SK-FDMs and wGS-FDMs that I think are used interchangeably, but in the figures it's referred to as wGS-FDMs. - It would be really interesting if you had metagenomic sequencing of each of the vegetable ferments to see how they differed in microbial composition and if that in combination with the substrate how that led to the metabolite profile. Particularly if there are large differences in species composition, or in contrast not really and it's more at the strain level or dependent on species interactions for producing certain metabolite profiles. Because not only do I think this study highlights that metabolites are the important piece of fermented foods for health effects, but also not all sauerkraut/fermented foods are the same, and possibly not even the same for health effects. - I'm not entirely sure how to think about comparing abundance of specific metabolites, and this is more because I'm new to metabolomics and don't know this specific method. Can the abundance of different metabolites/compound classes be compared against each other because of the normalization to the internal standard? So is it fair to compare the abundance of a metabolite in different samples, and different metabolites against each other? Such as looking at the wild green sauerkraut used for the in vivo studies and wanting to understand the most abundant metabolites, is it accurate to compare the abundance of different metabolites in a single sample? - I'm super interested in forthcoming studies which I think are coming based on the last paragraph of the discussion about deciphering if specific individual molecules are responsible for the effects through limited pathways, or from diverse molecules through many pathways. I'm curious what would be simple experiments to try to start answering this question. The first thing that comes to mind is fractioning the whole metabolite extract and performing in vitro cell-based assays to narrow down the candidates and then see if you can recapitulate in mice once you have either a narrowed down pool or single molecule. I'm not sure how feasible this would be however, but was the first thing that came to mind. - In Supplementary Figure 11 I don't quite understand why some of the figures have headers with 0 ng/mL LPS vs the human THP-1 cells figure says 200 ng/mL LPS. Weren't they all stimulated with LPS? Or is that the concentration of LPS when the measurements were taken? - It would be great once the data is available or publication accepted to update the Data Availability section with those repository links so everything is in one place.
Great work! Excited to see what comes next.