The 2009 discovery of somatic mutations affecting isocitrate dehydrogenase 1 and 2 (IDH1/2) in about 20 percent of karyotypically normal acute myeloid leukemia (AML) cases raised puzzling questions about how mutations in a pair of isoenzymes so fundamental to normal cellular metabolism could induce cancer.1 Normal IDH enzymes convert isocitrate to α-ketoglutarate (2-oxoglutarate, 2-OG), which is an essential nitrogen transporter and co-factor for oxidation reactions. As subsequent investigations revealed, the specific mutations in IDH1 and IDH2 observed in leukemia (and in malignant gliomas) generate neomorphic enzyme activity, resulting in production from α-ketoglutarate of the (R)-enantiomer of 2-hydroxyglutarate ((R)-2-HG), which inhibits enzymes such as TET2 (itself commonly mutated in AML) and augments the activity of the family of prolyl hydroxylases that down-regulate hypoxia-inducible factor (HIF).2,3 While these observations were suggestive that (R)-2-HG is an oncometabolite, it has been unclear whether (R)-2-HG is actually leukemogenic, since IDH mutants have several other effects on the cellular metabolome besides (R)-2-HG generation.
Julie Losman, in Bill Kaelin’s lab in Boston, and her colleagues have now shown that (R)-2-HG is indeed sufficient to promote leukemogenesis, and they have also provided new mechanistic insights into how that transformation might occur. The investigators transfected the peculiar TF-1 erythroleukemia cell line, which remains cytokine-dependent and retains differentiation potential, with tagged versions of IDH1 R132 and with control wild-type and catalytically inactive versions of IDH1. Losman and her colleagues observed that the IDH1 R132 mutation conferred cytokine independence and resulted in differentiation block. Hematopoietic growth factor independence and differentiation-failure are hallmarks of leukemia.
Similar behavior with IDH1 R132 transfection was also observed in an immortalized murine hematopoietic progenitor cell line (SCF ER-Hoxb8), confirming that these changes are not unique to TF-1 cells. Furthermore, incubation of non-infected TF-1 cells with (R)-2-HG showed that this oncometabolite (we can now safely call it that) itself induced cytokine independence and loss of differentiation similar to that promoted by the mutant IDH1 that generates (R)-2-HG; TET2 knockdown resulted in similar effects. A lentiviral shRNA knockdown screen of all known 2-OG-dependent dioxygenases pointed to TET2 as a candidate for conferring cytokine independence when inhibited.
In Brief
Importantly, the effects of (R)-2-HG on cell behavior were reversible in this model system, despite early fears that the dangerous epigenetic patterns resulting from altered function of TET2 and other enzymes under the influence of (R)-2-HG might be permanent. Incubating the transformed TF-1 cells and the murine cell line with a specific IDH1 R132 inhibitor (AGX-891) led to cell differentiation and loss of growth-factor independence. Presumably, the same would be true of mutation-specific inhibitors of IDH2 R140 and R172 mutants. Collectively, this work suggests that inhibition of IDH mutants by small molecules may be a potent therapeutic strategy in patients with AML who have these mutations. IDH1 and IDH2 inhibitors are in an advanced phase of preclinical development and will hopefully soon be ready for early-phase clinical trials.
References
Competing Interests
Dr. Steensma indicated no relevant conflicts of interest.
