IDH2 Mutations in Patients with d-2-Hydroxyglutaric Aciduria

See allHide authors and affiliations

Science  15 Oct 2010:
Vol. 330, Issue 6002, pp. 336
DOI: 10.1126/science.1192632


Heterozygous somatic mutations in the genes encoding isocitrate dehydrogenase-1 and -2 (IDH1 and IDH2) were recently discovered in human neoplastic disorders. These mutations disable the enzymes’ normal ability to convert isocitrate to 2-ketoglutarate (2-KG) and confer on the enzymes a new function: the ability to convert 2-KG to d-2-hydroxyglutarate (D-2-HG). We have detected heterozygous germline mutations in IDH2 that alter enzyme residue Arg140 in 15 unrelated patients with d-2-hydroxyglutaric aciduria (D-2-HGA), a rare neurometabolic disorder characterized by supraphysiological levels of D-2-HG. These findings provide additional impetus for investigating the role of D-2-HG in the pathophysiology of metabolic disease and cancer.

Recent studies in human cancer genetics have led to a resurgence of interest in a group of metabolic enzymes called isocitrate dehydrogenases. A recurrent heterozygous somatic mutation in the gene encoding cytosolic isocitrate dehydrogenase-1 (IDH1) is present in glioblastoma multiforme and alters residue Arg132 (R132) in the enzyme’s active site (1). This mutation disables the enzyme’s normal ability to convert isocitrate to 2-ketoglutarate (2-KG) and confers on it a new function: the ability to convert 2-KG to d-2-hydroxyglutarate (D-2-HG) (2). Heterozygous mutations that alter residues R140 and R172 of mitochondrial isocitrate dehydrogenase-2 (IDH2), the latter corresponding to R132 of IDH1, have been detected in other tumor types, acute myeloid leukemia and gliomas. These mutations also lead to abnormal production of D-2-HG (3, 4).

This unusual pathophysiological mechanism prompted us to explore whether mutations in IDH1 or IDH2 are associated with d-2-hydroxyglutaric aciduria (D-2-HGA) (Mendelian Inheritance in Man no. 600721). D-2-HGA is a rare inherited neurometabolic disorder with a wide clinical spectrum. Although some children with D-2-HGA are asymptomatic, others exhibit characteristics that can include developmental delay, epilepsy, hypotonia, cardiomyopathy, and dysmorphic features. All affected individuals have consistently increased D-2-HG levels in urine, plasma, and cerebrospinal fluid (5). About 50% of patients with this disorder, denoted D-2-HGA type I, have autosomal recessive mutations in the gene D2HGDH encoding D-2-hydroxyglutarate dehydrogenase (6), but the genetic basis of the disease in the remaining patients is unknown (5).

We sequenced the open reading frames of IDH1 and IDH2 in 17 unrelated idiopathic D-2-HGA patients (i.e., normal D-2-HGDH enzyme activity or no mutations in D2HGDH and consistently increased D-2-HG levels in body fluids). No mutations were detected in IDH1. In 15 patients, germline mutations were detected in IDH2: the known heterozygous G-to-A substitution at position 419 (c.419G>A), resulting in the replacement of Arg140 with Gln140 (p.R140Q) (4), and a novel heterozygous C-to-G substitution at position 418 (c.418C>G), Arg140→Gly140 (p.R140G) (Fig. 1 and table S1). Although the D-2-HGDH enzyme functions normally in these patients, the active protein appears to lack the catalytic capacity to oxidize all D-2-HG formed by IDH2 containing the R140 mutation; we thus denote the disorder in these patients as D-2-HGA type II. The higher urinary excretion of D-2-HG in the type II patients compared with that of type I patients (Fig. 1) is best explained by hyperproduction of this metabolite. The involvement of mitochondrial IDH2 is also consistent with the finding that D-2-HG is derived from mitochondrial 2-KG (7).

In eight of nine sets of parents, the mutation could not be detected, indicating that the heterozygous mutation arose de novo and that D-2-HGA type II is an autosomal dominant trait. In one family, however, three subsequent affected pregnancies were diagnosed by increased D-2-HG levels in amniotic fluid, suggesting germline mosaicism in the mother who herself had normal urinary D-2-HG levels and showed somatic mosaicism in her blood (Fig. 1).

Fig. 1

Characteristics of patients with D-2-HGA types I and II. The living D-2-HGA type II patients (n = 6) range in age from 3 to 22 years. The age of death of the remaining patients (n = 9) ranged from a few months up to 14 years. To date, none of the patients has been diagnosed with cancer. Shown below the table are sequence chromatograms from patients 5 and 15 (table S1), who have heterozygous IDH2 mutations. Somatic mosaicism for the R140Q mutation was detected in DNA of the mother of patient 5, unlike the other sets of parents in whom the mutation was not detected.

The pathophysiological consequences of increased D-2-HG in both cancer and D-2-HGA remain to be determined. Dang et al. have hypothesized that D-2-HG is an “onco-metabolite” that contributes to the formation of gliomas (2). However, patients with malignant gliomas and anaplastic astrocytomas that harbor IDH1 or IDH2 mutations show improved survival in comparison to patients whose tumors lack these mutations (1, 4). The absence of cancer diagnoses in our D-2-HGA patient population (>85 patients) is also not consistent with the proposed role of D-2-HG as an onco-metabolite, although the 15 D-2-HGA type II patients are young, which may preclude firm conclusions about cancer susceptibility (Fig. 1 and table S1).

An increased incidence of brain tumors has been noted among patients with l-2-hydroxyglutaric aciduria (L-2-HGA) (8). l-2-hydroxyglutarate (L-2-HG) is the stereoisomer of D-2-HG, but L-2-HGA is a distinct neurometabolic disease. In contrast to D-2-HGA, L-2-HGA is a leukodystrophy. L-2-HGA manifests in early childhood with slowly progressive neurological symptoms, including psychomotor retardation, cerebellar ataxia, variable macrocephaly, and epilepsy (9). The biochemical defect in L-2-HGA is caused by mutations in the L2HGDH gene, which encodes an enzyme that specifically degrades the l enantiomer of 2-HG (9).

Now that disease-associated mechanisms have been described for nearly all D-2-HGA patients, genetic counseling is expected to be enhanced. Our findings provide additional impetus for investigating the role of D-2-HG in the pathophysiology of inborn errors of metabolism and neoplastic disorders.

Supporting Online Material

Materials and Methods

Tables S1 and S2



View Abstract

Navigate This Article