Abstract Body: Cellular metabolism is fundamental to cellular functions and energy production. The current understanding of the metabolic profiles of distinct kidney cell types is limited. In this study, we present a comprehensive, cell-type-specific metabolic landscape of the human kidney.
We utilized single-cell RNA-seq data from 369,590 human kidney cells available from Cellxgene as input to the constraint-based flux-balanced model, COMPASS. To improve computational efficiency, we reduced the data size by creating micro-pools within each “organ-tissue region-original cell type” condition. We then applied Wilcoxon test on the reaction scores computed by COMPASS. Using the reaction ranks, we developed new methods to calculate count matrices for both reactions and inferred metabolite levels for downstream analysis. We generated spatial metabolomic data from a rat kidney for validation.
Proximal tubule epithelial cells (PTEC) showed high activity of reactions catalyzed by phenylalanine-tetrahydrobiopterin oxidoreductase and tetrahydrobiopterin-carbinolamine dehydrogenase, and tyrosine metabolism. Tetrahydrobiopterin is an essential cofactor in the conversion of phenylalanine to tyrosine, and the production of nitric oxide, which plays a vital role in regulating blood flow and tubular function in the kidneys. PTEC and Connecting tubule cells (CTC) showed high activity of fatty acid elongation in mitochondria as their primary means of energy production. Distal convoluted tubule cells (DCTC) showed high activity of succinate metabolism underscoring their contribution to regulating blood pressure. Loop of Henle thin descending limb (LHTDL) showed high activity of the reaction catalyzed by NAD(P) oxidoreductase showing its vital role in the superoxide production contributing to maintaining the balance of free radicals in the kidney tubules. LHTDL showed higher activity of tyrosine metabolism compared to loop of Henle think ascending limb (LHTDL). We compared tissue-specific metabolite profiles inferred by the flux model with spatial metabolomic data and observed 57% similarity in metabolite differences between kidney cortex and kidney medulla.
In conclusion, we identified previously uncharacterized metabolic signatures in specific cell types including PTEC, DCTC, and LHEC. These metabolic characteristics may play an important role in maintaining and regulating physiological functions of these cell types.