Altered branched-chain α-keto acid metabolism is a feature of NAFLD in individuals with severe obesity.

Journal Article (Journal Article)

Hepatic de novo lipogenesis is influenced by the branched-chain α-keto acid dehydrogenase (BCKDH) kinase (BCKDK). Here, we aimed to determine whether circulating levels of the immediate substrates of BCKDH, the branched-chain α-keto acids (BCKAs), and hepatic BCKDK expression are associated with the presence and severity of nonalcoholic fatty liver disease (NAFLD). Eighty metabolites (3 BCKAs, 14 amino acids, 43 acylcarnitines, 20 ceramides) were quantified in plasma from 288 patients with bariatric surgery with severe obesity and scored liver biopsy samples. Metabolite principal component analysis factors, BCKAs, branched-chain amino acids (BCAAs), and the BCKA/BCAA ratio were tested for associations with steatosis grade and presence of nonalcoholic steatohepatitis (NASH). Of all analytes tested, only the Val-derived BCKA, α-keto-isovalerate, and the BCKA/BCAA ratio were associated with both steatosis grade and NASH. Gene expression analysis in liver samples from 2 independent bariatric surgery cohorts showed that hepatic BCKDK mRNA expression correlates with steatosis, ballooning, and levels of the lipogenic transcription factor SREBP1. Experiments in AML12 hepatocytes showed that SREBP1 inhibition lowered BCKDK mRNA expression. These findings demonstrate that higher plasma levels of BCKA and hepatic expression of BCKDK are features of human NAFLD/NASH and identify SREBP1 as a transcriptional regulator of BCKDK.

Full Text

Duke Authors

Cited Authors

  • Grenier-Larouche, T; Coulter Kwee, L; Deleye, Y; Leon-Mimila, P; Walejko, JM; McGarrah, RW; Marceau, S; Trahan, S; Racine, C; Carpentier, AC; Lusis, AJ; Ilkayeva, O; Vohl, M-C; Huertas-Vazquez, A; Tchernof, A; Shah, SH; Newgard, CB; White, PJ

Published Date

  • August 8, 2022

Published In

Volume / Issue

  • 7 / 15

PubMed ID

  • 35797133

Pubmed Central ID

  • PMC9462486

Electronic International Standard Serial Number (EISSN)

  • 2379-3708

Digital Object Identifier (DOI)

  • 10.1172/jci.insight.159204


  • eng

Conference Location

  • United States