Abstract
Microorganisms are central to ecological and biotechnological processes that sustain food systems and human health. Although the nutritional and health potential of traditional fermented foods and plant-based resources are well recognized, the mechanisms underlying microbiome metabolome interactions that drive these benefits remain poorly characterized. Advances in multi-omics now enable in-depth analysis of microbial ecosystems such as spontaneously fermented foods and the human gut, revealing their metabolic networks and ecological dependencies. In this context, this thesis aimed to elucidate microbiome-metabolome synergies through integrative and mechanistic approaches to guide predictive and functional understanding of microbial activity, metabolite production, and their roles in fermented food systems and human gut health. To establish the conceptual foundation, a review paper synthesized current advances on synthetic microbial communities (SynComs) as model systems for understanding resistance, resilience, and functionality in spontaneously fermented foods. It emphasized the need for standardized design and integrated multi-omics approaches to ensure stability and reproducibility, positioning SynComs as essential models for steering microbial interactions. Building on this, the first research study applied an ecological role-based SynCom design in a sourdough model. Sourdough is a dynamic ecosystem in which microbial players are continuously shaped and challenged by environmental and spatiotemporal stimuli. Frequent back-slopping and the acidic environment sustain a complex community, yet this dynamism can compromise metabolic stability and resilience. To investigate whether functional robustness could be maintained despite compositional variability, twenty de novo consortia, each comprising six strains, were assembled at two different growth phases from pre-defined ecological roles of a metabolically robust SynCom - including ‘core dominant’, ‘subdominant’, and ‘satellite’ species. Serial substitutions of the stratified microbial players were performed at both the species and strain level. In vitro and in situ fermentations demonstrated that functional stability, acidification dynamics, and volatile organic compound profiles were conserved even after taxonomic substitutions, provided that ecological balance was maintained. These findings established that function-centric, rather than species-centric, design principles ensure reproducibility and robustness in food microbiomes, marking a paradigm shift in SynCom engineering toward predictable and resilient fermentation systems. Another milestone concerned the tailored co-metabolism of lactic acid bacteria (LAB) and yeast for the valorization of Moringa oleifera leaf powder (MOLP), a nutrient- and polyphenol-rich substrate. Binary and ternary cocultures significantly increased lactic acid, total peptides, and free polyphenols (rutin, quercetin, kaempferol), improving antioxidant potential. This demonstrates that metabolic complementarity among co-fermenting microbes can be leveraged to generate multifunctional ingredients for skin care and functional food applications. Extending the exploration of microbiome-metabolome crosstalk to the human gut ecosystem, the final study explored the prebiotic potential of a polyphenol- and fiber-rich elderberry extract (EBE) using the Simulator of the Human Intestinal Microbial Ecosystem (SHIME®) in vitro gut model, with a focus on inter-individual variability. EBE supplementation modulated microbiota composition and predicted functionality in a donor specific manner, increasing potentially beneficial taxa such as Akkermansia muciniphila and Bifidobacterium longum, along with elevated short-chain fatty acids (SCFAs) production. Shotgun metagenomics revealed an increased abundance of genes involved in amino acid metabolism, vitamin B6 biosynthesis, and polyphenol catabolism, collectively suggesting overall physiological benefits. Spearman correlation analyses further highlighted putative interactions between specific microbial species and pathway-associated genes. Overall, this thesis integrates ecological theory and multi-omics to demonstrate how microbial communities can be predictively steered, from food fermentations to the gut ecosystem, toward enhanced functionality and health-promoting outcomes.