Insights from retinitis pigmentosa into the roles of isocitrate dehydrogenases in the Krebs cycle

Abstract

Here we describe two families with retinitis pigmentosa, a hereditary neurodegeneration of rod and cone photoreceptors in the retina. Affected family members were homozygous for loss-of-function mutations in IDH3B, encoding the beta-subunit of NAD-specific isocitrate dehydrogenase (NAD-IDH, or IDH3), which is believed to catalyze the oxidation of isocitrate to alpha-ketoglutarate in the citric acid cycle. Cells from affected individuals had a substantial reduction of NAD-IDH activity, with about a 300-fold increase in the K(m) for NAD. NADP-specific isocitrate dehydrogenase (NADP-IDH, or IDH2), an enzyme that catalyzes the same reaction, was normal in affected individuals, and they had no health problems associated with the enzyme deficiency except for retinitis pigmentosa. These findings support the hypothesis that mitochondrial NADP-IDH, rather than NAD-IDH, serves as the main catalyst for this reaction in the citric acid cycle outside the retina, and that the retina has a particular requirement for NAD-IDH.

Figure 1

Scatter plot of probe-set hybridization intensities from individual 003-053. Data are organized according to percentage of respective mean signal across all subjects (x axis) and number of s.d. below this mean (y axis). The plot shows the 306 probe sets detecting hybridization signals two or more s.d. below the mean of all other subjects and controls. Of the 306 probe sets, 3 were clear outliers, with hybridization intensities more than 6 s.d. below the respective means; the 3 probe sets detected IDH3B (IDH3B-1, probe set 210014_x_at; IDH3B-2, 210418_s_at; IDH3B-3, 201509_at.) These were the only probe sets in the microarray designed to hybridize to mRNA from IDH3B.

Figure 2

IDH3B showing three known RNA splicing variants (see URLs section below) and the location of the m.p.Ile163fs and m.p.Leu98Pro mutations. The m.p.Leu98Pro mutation is a missense mutation and alters IDH3B in variants 1 and 2. There is no predicted effect on variant 3 since the mutation is located in the 5′ untranslated region of that variant. Splice variant 3 is missing exons 1-5 which encode 132 amino acid residues out of 385 in the full-length protein. It is not likely to encode a functional beta subunit that could compensate for the defective m.p.Leu98Pro mutation in the two other variants. The m.p.Ile163fs mutation causes a frameshift with a premature stop codon in all three variants presumably leading to mRNA degradation by nonsense-mediated decay.

Figure 3

Pedigrees with the IDH3B m.p.Ile163fs and m.p.Leu98Pro mutations. Black arrows point to the index patients. Beneath each family member’s symbol is that person’s IDH3B genotype; family members for which no genotype is shown did not submit a blood sample for analysis. Red arrows indicate the location of the mutation in the DNA sequence tracings. Nucleotides corresponding to the normal sequence (top line) and mutated sequences (lower line) are indicated below the sequence alignments.

Figure 4

Activity of NAD-dependent isocitrate dehydrogenase. Average results of at least duplicate measurements with standard deviations of Km for NAD and Km for isocitrate (with or without ADP), Vmax and NAD-IDH activity under standard conditions in lymphoblast lysates from three normal subjects (+/+); two heterozygotes (m1/+) and one homozygote (m1/m1) with IDH3B-m.p.Ile163fs; one heterozygote (m2/+) and one homozygote (m2/m2) with the IDH3B-m.p.Leu98Pro mutation. The mutant homozygotes have a 267-298-fold increase in the Km for NAD (upper left panel) and a resulting loss of NAD-IDH activity (lower right panel). The normal effect of ADP in lowering the Km for isocitrate is partially (for m1 = m.p.Ile163fs) or completely (for m2 = m.p.Leu98Pro) lost in homozygote patients (lower left panel).

Figure 5

Frequency of SAGE tags from IDH3 and IDH2 in the retina and retinal pigment epithelium (retina/RPE) in comparison to that from all other human tissues using EyeSAGE database. IDH3: mitochondrial NAD-dependent isocitrate dehydrogenase, alpha subunit - IDH3A (gene ID 3491), beta subunit - IDH3B (gene ID 3420), gamma subunit - IDH3G (gene ID 3421); IDH2: mitochondrial NADP-dependent IDH - IDH2 (gene ID 3418). The low ratio of tags from NADP-IDH suggest that the retina/RPE has a relatively low abundance of NADP-IDH and may be unusually dependent on NAD-IDH.