Multiomic analysis of microbiome and host identifies features differentiating diet and surgical induced weight loss

Objective: Gastric bypass surgery (GBP) is a proven effective treatment for correcting metabolic dysfunction in individuals with severe obesity, though the physiological changes induced by surgery are incompletely understood. This study aimed to characterize the differences between diet and surgery induced weight loss by multi-omic analysis of microbiome and host features.Design: The COCKTAIL study was an open, non-randomized, single-center study. 88 participants with severe obesity and a wide range of glucose tolerance were recruited into two arms - GBP versus very low energy diet (VLED). Longitudinal stool and blood samples were collected during the 2-year follow up period. Multi-omic analyses were performed at 6-weeks post GBP/start of VLED.Results: There was overlap in the microbiome composition at baseline, but a dramatic shift was seen in the GBP group after surgery both in terms of taxonomic abundances and functional capacity. Differential enrichment comparisons between the GBP and VLED was performed at both Week 6 and Year 2 revealing an enrichment of oral commensals such as Steptococcus salivarius and Streptococcus parasanguinis at both timepoints in the GBP group. Our final signature of 120 features derived from multi-omic integration at Week 6 achieved excellent accuracy in classifying patients to the GBP or VLED groups and further identified metabolites such as phenol sulphate and tryptophan as characterizing the GBP group. Lastly, these analyses also helped identify critical links for S. parasaguinis which was inversely associated with tryptophan and positively linked to phenol sulphate.Conclusions: These data may have important implications as we observed significant restructuring of the microbiome away from baseline dysbiosis in the GBP group along with altered functional capacity. Such effects are not seen in the VLED group with matched weight loss up to 6 weeks. Multi-omic analyses revealed potentially novel mechanistic insights at the intersection of the microbiome and host.