Fig 2: 2HG production by IDH2 in the glucose-free medium conditions. (a) GC-MS analysis of 2HG levels in glucose-containing (+Glc) and glucose-free medium (-Glc) in control cell line SHSY5Y, or SHSY5Y cells expressing IDH2 WT or K413Q. Below there are depicted respective western blots for IDH2, M2, and actin. N = 3. ***p < 0.0001 using one-way ANOVA, ##p < 0.01 (p = 0.0055 and 0.0012, respectively) calculated using Unpaired t-test. (b) Respiration rates of SHSY5Y cells cultivated in glucose-containing (+Glc) and glucose-free medium (-Glc) measured by Oroboros 2 K oxygraph; routine respiration refers to endogenous respiration of million cells, maximal respiration refers to maximally uncoupled respiration using the mitochondrial uncoupler FCCP. N > 5. ***p = 0.0006, *p = 0.0150, calculated using one-way ANOVA Tukey’s multiple comparisons test. (c) mRNA expression analysis of SIRT3 in SHSY5Y and HepG2 cells grown in glucose-free medium for 72 hours. N > 4. ns p = 0.3960 and p = 0.6406, respectively, calculated using Unpaired t-test.

Fig 3: Role of GCN5L in acetylation of mitochondrial IDH2. (a) Cellular 2HG analysed by LC-MS from the control (ctrl), pcDNA3.1 (EV), and GCN5L-transfected cells overexpressing IDH2 WT or K413Q, respectively. N > 5. Underneath there is depicted western blot of the corresponding cell lysates of the analysed samples demonstrating transfection of GCN5L along with IDH2 variants (both detected by anti-M2 antibody). Western blot detecting IDH2 expression in stable cell lines. (b) Michaelis-Menten plots for NADP+-driven IDH2-mediated NADPH formation as dependent on concentrations of NADP+ and isocitrate (IC) for purified protein from the IDH2-overexpressing SHSY5Y cells transfected with GCN5L. Box contains values of Vmax and Km. Unit of Kcat for IDH2 WT/K413Q is s-1; data are expressed as mean ± s.e.m. (c) Confocal microscopy images of GCN5L transfected cells, GCN5L detected by primary anti-M2 antibody and labelled by secondary antibody conjugated with Alexa-Fluor 568 (red) or Alexa-Flour 488 (green); mitochondria-targeted GFP (mito-GFP, green); mitochondria-targeted Keima-Red (red), ER-targeted Tomato fluorescent protein (tdTomato-ER-3, red); overlay of the red and green channels (merge). The scale corresponds to 10 µm. (d) Western blot depicting mitochondrial and cytosolic localization of GCN5L (detected by Flag-tag, M2), whole-cell lysate, cytosolic fraction, and mitochondrial fraction.

Fig 4: IDH2 reaction of IDH2 WT and IDH2 K413Q. (a) Western blot of the exemplar purification of IDH2; whole cell lysate (lysate), flow-through (FT) indicating that significant pool of IDH2-Flag is attached to the resin; three sequential steps of elution by Flag-peptide (EL1, EL2, EL3). (b) Michaelis-Menten plots for NADP+-driven IDH2-mediated NADPH formation as dependent on concentrations of NADP+ (left) and IC (right). Box contains values of Vmax and Km. Unit of Kcat for IDH2 WT/K413Q is s-1. Data are expressed as mean ± s.e.m. (c) GC-MS evaluation of reaction products after 120 s of a purified enzyme in the reaction mixture, 2OG (left) and 2HG (right). N > 10, ***p < 0.0001 calculated using Unpaired t-test with Welch´s correction. (d) Reaction schema depicting 2HG formation in the in-vitro reaction; reactions of isolated IDH2 WT and K413Q as recorded by NADPH formation depicting similar reaction rates; underneath, 2OG and 2HG evaluation of the actual reaction products corresponding to presented reaction rates. N = 3, ***p < 0.0001 calculated using One Way ANOVA Tukey´s multiple comparisons test. (e) GC-MS evaluation of metabolites 2HG, 2OG and citrate extracted from the 293LTV cells transfected with no vector (“ctrl”), with the empty vector (“EV”), with vector encoding wild-type IDH2 (“IDH2 WT”) or K413Q mutant (“IDH2 K413Q”) with corresponding western blots. N = 3, ***p < 0.001 (p = 0.0001, p < 0.0001, p < 0.0001, respectively) calculated using One Way ANOVA Tukey’s multiple comparisons test. (f) Citrate levels in SHSY5Y expressing IDH2 WT and K413Q, respectively (left), and % incorporation calculated as M+0 and M+1 ratio of the respective ion form (middle), and the experimental scheme depicting 13C labeling of citrate from 1-13C-glutamine. N = 3, *p < 0.05 (p = 0.0186) calculated by Unpaired t-test.

Fig 5: SIRT3 deacetylates IDH2 in vitro. (a) Western blot analysis of the purified IDH2, indicating that treatment of acetylated IDH2 (“acetylation”) with recombinant SIRT3 deacetylates IDH2, performed with anti-IDH2 and anti-acetyl-lysine antibodies. Western blot was performed using equimolar amount of protein. The lanes contain following fractions: whole cell lysate (lysate), flow-through (FT), three sequential steps of elutions (EL1, EL2, EL3), combined three elutions (ELcombined), following recovery by 10 K Amicon (ELrecovered), protein after acetylation step (acetylation), protein after treatment with recombinant SIRT3 (+SIRT3), and parallel sample without SIRT3 (-SIRT3). (b) Coomassie blue-stained PAGE of the purified IDH2 used for LC-MS analysis of acetylated lysines with indicated recombinant SIRT3. C) Model depicting IDH2 structure with NADPH. The reaction center is depicted in orange, K413 in red, K106, 166, and 384 in magenta, other possibly acetylated/deacetylated lysines in pink.