Longitudinal adipose tissue single cell transcriptomics reveals genes and variants regulating weight loss after bariatric surgery

Abstract

Ability to lose weight during different obesity treatments shows substantial variability between individuals and is likely under genetic control; however, the underlying predictive variants and weight loss genes remain unknown. Here we profiled longitudinal, single cell level adipose transcriptomes of individuals undergoing bariatric surgery to elucidate genes and their regulatory variants contributing to interindividual variability in weight loss outcomes. We identified wide-spread cellular and subcellular transcriptional changes to weight loss with most profound responses in adipocyte subtypes. By clustering the weight loss genes based on their cell-type level co-expression profiles, we uncovered functionally distinct subsets of genes reflecting altered adipocyte expression of central adipocyte function enriched genes. Next, we discovered that body mass index (BMI) polygenic risk score (PRS) built using the cis regulatory variants in these 45 adipocyte weight loss genes significantly predict the magnitude of the achieved weight loss and are strongly enriched for variance explained in the change of BMI. Taken together, this longitudinal single nucleus adipose data establishes gene signatures for weight loss and discovers genetic regulators underlying the interindividual variability of weight loss.