Utilization of a-Ketoglutarate for Synthesis of 2-Hydroxyglutarate Oncometabolite Promotes Catabolic Flexibility, Redox Perturbation and Mitochondrial Activity That Supports Chemoresistance in IDH1 Mutant Acute Myeloid Leukemia

Abstract

Metabolic reprogramming is now considered one of the major characteristics of cancer cells as they must adapt their metabolism to fuel energetic and biosynthetic needs and ensure proliferation. Changes in intermediary and energy metabolism are also a key hallmark of acute myeloid leukemia (AML), and targeting glycolysis, glutaminolysis, fatty acid β-oxidation or mitochondrial oxidative phosphorylation (OxPHOS) are promising anti-leukemic approaches. Furthermore, recurrent mutations in two crucial metabolic genes, isocitrate dehydrogenase 1/2, have been discovered in more than 15% of AML patients and could negatively impact prognosis as well as represent new therapeutic targets. Whereas wild-type IDH1 catalyzes the conversion of isocitrate to α-ketoglutarate (α-KG) generating NADPH in the cytosol, mutant IDH1 generates the oncometabolite 2-hydroxyglutarate (2-HG) from α-KG, thereby limiting the availability and utilization of this key metabolite central to intermediary metabolism, and oxidizes NADPH. NADPH provides reducing equivalents necessary to maintain cellular redox balance and serves as an electron donor in biosynthetic reactions. However, the contribution of IDH1 and its variants to metabolism has thus far been overlooked in AML. Using a computational and biological system-based approach combining complementary methods such as transcriptomics, metabolomics, proteomics, enzymatic assays, nutrient deprivations and metabolic pathway inhibitors, we have determined that IDH1 mutant cells display a high capacity to use a broad spectrum of substrates and are thus not dependent on any single one. Indeed, when glutamine is the main source of production of -2-HG, it only provides ~65% of the total amount of 2-HG with glucose, pyruvate, fatty acid and proline oxidation providing the balance. Using 13C-isotope nutrient-based fluxomics, we also observed that substrate flexibility is exploited to provide α-KG, the precursor of 2-HG, and substrate in many metabolic pathways. Hydroxylated peptides and hypermethylated gene promoters in several key metabolic genes were specifically identified in IDH1 mutant AML cells. We further observed that this leads to an altered redox state and increased mitochondrial activity as confirmed by in silico modeling using 3 different algorithms. Measurement of mitochondrial activity in IDH1 mutant AML cells showed that these cells exhibit higher fatty acid β-oxidation as well as a higher respiration and produced increased amounts of TCA cycle intermediates and mitochondrial ATP. All of these features are consistent with a high oxidative phosphorylation (OxPHOS) status and activity that promotes AML chemoresistance and predicts treatment response in PDX and AML patients. Accordingly, IDH1 mutants are less sensitive to cytarabine but more vulnerable to hypoxia and OxPHOS or mitochondrial BCL2 inhibitors (metformin, ABT-199) in vitro and in vivo. Surprisingly, while ABT-199 and AraC markedly decrease 2-HG, metformin increases production of 2-HG, revealing a dissociation between 2-HG production and apoptosis in AML cells. Altogether, our results suggest that IDH1 mutation and production of 2HG in AML cells confers catabolic flexibility and increases their capacity for consumption of multiple substrates and to drug resistance. This metabolic redirection toward the 2-HG synthesis induces a dependency on OxPHOS metabolism and leads to greater resistance to genotoxic agents and enhances sensitivity to mitochondrial inhibitors. Thus, this work provides a strong rationale for clinical trials that incorporate a triple therapy combining AGI, OXPHOS inhibitor and AraC in AML patients with IDH mutations. DisclosuresRecher:Celgene, Sunesis, Amgen, Novartis: Research Funding; Novartis, Celgene, Jazz, Sunesis, Amgen: Consultancy.