EXPLORING THE GUT MICROBIOME IN HEALTHY AGING: FROM METABOLIC ASSOCIATIONS TO AN EXPERIMENTAL FRAMEWORK FOR NOVEL HEALTH-ASSOCIATED MICROBIAL SPECIES ISOLATION

Chronological aging is coupled with a progressive decline in health and a perturbation of physiological homeostasis. The gut microbiome, a key modulator of host metabolism and immunity, can be altered by this process and, in turn, influence its progression. While extensive evidence links microbial features to cardiometabolic health, a comprehensive strategy for assessing the gut microbiome's holistic association with an individual's health status remains a key need. Here, we developed a novel, quantitative framework to address this gap. First, we computed a Metabolic Risk Index (MRI) for prevalent species-level genome bins (SGBs) by leveraging a key set of validated and emerging metabolic biomarkers in 1,043 middle-aged to elderly participants from the PLIC cohort. Our analysis revealed that SGBs with the strongest and most consistent association with a healthy metabolic profile were largely uncharacterized species, lacking cultured representatives in public repositories. Building on the MRI, we created a Metabolic Deviation (MetaDev) score for each individual's gut microbiota profile. The MetaDev score showed strong correlations with systemic cardiometabolic markers and lifestyle factors, and was also highly predictive of a wide range of diseases. Its most pronounced effects were observed at the extremes of the score’s distribution, including conditions such as heart failure, cerebral stroke, lung cancer, and type 2 diabetes. These associations were robustly validated in an independent cohort. To move from statistical associations to mechanistic understanding, we established a metagenomic-guided culture-based workflow called MetaCulture to isolate these health-associated uncharacterized microbial species. By developing rapid, cost-effective, high-throughput molecular tools for targeted detection, our workflow overcomes the significant challenges of low species abundance and complex culture requirements, enabling the efficient isolation of pure colonies. As a proof-of-concept, we successfully applied MetaCulture to isolate strains of two uncharacterized clades. First, we isolated and assembled the draft genome of a representative strain of SGB4706, a species with a low MRI value and promising anti-inflammatory potential. Second, we isolated two strains of Catenibacterium mitsuokai subsp. tridentinum subsp. nov., a clade associated with non-urbanized populations, and investigated its engraftment and function in a mouse model. Altogether, this study provides a new framework for identifying key health-associated microbes and a powerful workflow for their isolation, thereby enabling their experimental validation. This work offers a clear path to identify individuals with unfavorable gut microbiome profiles, and can thus guide targeted lifestyle interventions and the development of next-generation biotherapeutics, including novel health-associated microbial strains and strategies to enhance their presence or function.