In Silico Modeling of Metabolic State in Single Th17 Cells Reveals Novel Regulators of Inflammation and Autoimmunity

Abstract Cellular metabolism is a major regulator of immune response, but it is difficult to study the metabolic status of an individual immune cell using current technologies. Here, we present Compass, an algorithm to characterize the metabolic landscape of single cells in silico based on single-cell RNA-Seq and flux balance analysis. We used Compass to study the landscape of metabolic heterogeneity in Th17 cells and search for novel metabolic regulators of their inflammatory function. In central carbon metabolism, Compass predicted a metabolic switch between glycolysis and fatty acid oxidation that mirrors the Th17 vs. Treg phenotype, which we validated through transcriptomic, metabolic and functional assays. The TCA cycle was predicted to break at two points, both of which have been independently identified by other groups in M1 macrophage polarization. Surprisingly, and contrary to common immunometabolic understanding, Compass predicted that glycolysis too was divided into functional modules, and that one of them supported an anti-inflammatory phenotype. We validate the paradoxical prediction and demonstrate that inhibition of this module promotes a pro-inflammatory transcriptional program in Th17 cells, resulting in neuroinflammation in an adoptive transfer model of autoimmune disease. In conclusion, Compass is a widely applicable algorithm to characterize metabolic states at single cell resolution. It allows associating cellular metabolic states with effector functions and detection of metabolic targets that regulate effector phenotypes. We expect it to become a widely used tool with the increasing availability of single-cell RNA-Seq data, spurred by efforts such as the human cell atlas.