Extended Data Fig. 3: Mutant IDH protein inactivation by inhibitors alters NADPH fluxes induced by mutant IDH.
From: Cytosolic and mitochondrial NADPH fluxes are independently regulated
To confirm that the observed changes in cytosolic and mitochondrial NADPH metabolism were caused by compartmentalized (D) 2-hydroxyglutarate (2HG) production, we repeated the analysis in the presence of pharmacological inhibitors that selectively inactivate mutant IDH protein. We used AGI-5198 to inhibit 2HG production by mutant IDH1 and enasidenib to inhibit 2HG production by mutant IDH2. As anticipated, AGI-5198 reduced intracellular 2HG levels in IDH1 mutants but did not influence intracellular 2HG levels in IDH2 mutants (a). Enasidenib had the opposite effect, reducing intracellular 2HG levels in IDH2 mutants but not IDH1 mutants (a). We treated cells with either AGI-5198 or enasidenib for 48 hours, while they were being labeled with 2H glucose. Neither inhibitor influenced the distribution of cytosolic or mitochondrial NADPH fluxes in wild-type cells. AGI-5198, but not enasidenib, did lead to a statistically significant change in the distribution of cytosolic NADPH fluxes in IDH1 mutants (b-d). Only enasidenib, on the other hand, influenced the distribution of mitochondrial NADPH fluxes in IDH2 mutants (e-h). These rescue experiments provide additional evidence that localized production of 2HG has compartment-specific effects on NADPH metabolism. a, Relative levels of intracellular 2HG from HCT116 wild-type cells, IDH1 mutants, and IDH2 mutants treated with AGI-5198 or Enasidenib. b, Ratio of proline to G6P enrichment in HCT116 wild-type cells labeled with 3-2H glucose is not affected by inhibitors of mutant IDH. c, Ratio of proline to G6P enrichment in IDH1 mutants labeled with 3-2H glucose. Inhibition of mutant IDH1 with AGI-5198 alters the distribution of cytosolic NADPH fluxes, but enasidenib does not. d, Ratio of proline to G6P enrichment in IDH2 mutants labeled with 3-2H glucose is not affected by inhibitors of mutant IDH. e, Ratio of P5C to malate enrichment in HCT116 wild-type cells labeled with 4-2H glucose is not affected by inhibitors of mutant IDH. f, Ratio of P5C to malate enrichment in IDH1 mutants labeled with 4-2H glucose is not affected by inhibitors of mutant IDH. g, Ratio of P5C to malate enrichment in IDH2 mutants labeled with 4-2H glucose. Inhibition of mutant IDH2 with enasidenib alters the distribution of mitochondrial NADPH fluxes, but AGI-5198 does not . WT, wild-type HCT116 cells; mIDH1, mutant IDH1 cells; mIDH2, mutant IDH2 cells. A concentration of 0.2 μM AGI-5198 and 0.1 μM enasidenib was used. Values are mean ± s.d; n = 3 biologically independent samples (a-g). Statistically significant differences were calculated by using a one-way ANOVA followed by Dunnett’s multiple comparison test (a-g).NS= no significant difference.
