Multi-omics characterization of the microbial populations and chemical space composition of a water kefir fermentation

In recent years, the popularity of fermented foods has strongly increased based on their proven health benefits and the adoption of new trends among consumers. One of these health-promoting products is water kefir, which is a fermented sugary beverage based on kefir grains (symbiotic colonies of yeast, lactic acid and acetic acid bacteria). According to previous knowledge and the uniqueness of each water kefir fermentation, the following project aimed to explore the microbial and chemical composition of a water kefir fermentation and its microbial consortium, through the integration of culture-dependent methods, compositional metagenomics, and untargeted metabolomics. These methods were applied in two types of samples: fermentation grains (inoculum) and fermentation samples collected at different time points. A strains culture collection of ∼90 strains was established by means of culture-dependent methods, mainly consisting of individuals of Pichia membranifaciens, Acetobacter orientalis, Lentilactobacillus hilgardii, Lacticaseibacillus paracasei, Acetobacter pomorum, Lentilactobacillus buchneri, Pichia kudriavzevii, Acetobacter pasteurianus, Schleiferilactobacillus harbinensis, and Kazachstania exigua, which can be further studied for their use in synthetic consortia formulation. In addition, metabarcoding of each fermentation time was done by 16S and ITS sequencing for bacteria and yeast, respectively. The results show strong population shifts of the microbial community during the fermentation time course, with an enrichment of microbial groups after 72 h of fermentation. Metataxonomics results revealed Lactobacillus and Acetobacter as the dominant genera for lactic acid and acetic acid bacteria, whereas, for yeast, P. membranifaciens was the dominant species. In addition, correlation and systematic analyses of microbial growth patterns and metabolite richness allowed the recognition of metabolic enrichment points between 72 and 96 h and correlation between microbial groups and metabolite abundance (e.g., Bile acid conjugates and Acetobacter tropicalis). Metabolomic analysis also evidenced the production of bioactive compounds in this fermented matrix, which have been associated with biological activities, including antimicrobial and antioxidant. Interestingly, the chemical family of Isoschaftosides (C-glycosyl flavonoids) was also found, representing an important finding since this compound, with hepatoprotective and anti-inflammatory activity, had not been previously reported in this matrix. We conclude that the integration of microbial biodiversity, cultured species, and chemical data enables the identification of relevant microbial population patterns and the detection of specific points of enrichment during the fermentation process of a food matrix, which enables the future design of synthetic microbial consortia, which can be used as targeted probiotics for digestive and metabolic health.