FDXR Mutations Cause Sensorial Neuropathies and Expand the Spectrum of Mitochondrial Fe-S-Synthesis Diseases

Abstract

Hearing loss and visual impairment in childhood have mostly genetic origins, some of them being related to sensorial neuronal defects. Here, we report on eight subjects from four independent families affected by auditory neuropathy and optic atrophy. Whole-exome sequencing revealed biallelic mutations in FDXR in affected subjects of each family. FDXR encodes the mitochondrial ferredoxin reductase, the sole human ferredoxin reductase implicated in the biosynthesis of iron-sulfur clusters (ISCs) and in heme formation. ISC proteins are involved in enzymatic catalysis, gene expression, and DNA replication and repair. We observed deregulated iron homeostasis in FDXR mutant fibroblasts and indirect evidence of mitochondrial iron overload. Functional complementation in a yeast strain in which ARH1, the human FDXR ortholog, was deleted established the pathogenicity of these mutations. These data highlight the wide clinical heterogeneity of mitochondrial disorders related to ISC synthesis.

Keywords: ARH1; Auditory neuropathy; FDXR; Fe-S cluster synthesis; iron overload; iron-sulfur cluster; mitochondria; optic atrophy.

Figure 1

Pedigree of the Investigated Families FDXR variants are reported according to GenBank: NM_024417.4.

Figure 2

Ferredoxin Reductase Mutations (A) Three-dimensional representation of the crystal structure of beef ferredoxin reductase (PDB: 1CJC). Blue and green sticks illustrate flavin adenine dinucleotide and sulfate ion, respectively. The mutated residues are highlighted in pink. (B–E) H-bond modification resulting from p.Leu215Val, p.Arg242Trp, p.Arg306Cys, and p.Arg327Ser changes. (F) Sequence alignment of ferredoxin proteins from various species. The arrows indicate the amino acid changes.

Figure 3

A Defect in Ferredoxin Reductase Results in Deregulation of Iron Homeostasis (A) Decreased steady-state levels of ferredoxin reductase (upper panel) and increased steady-state levels of TfR1 and SOD2 (lower panel) in fibroblasts of subjects 1 and 5 (S1, S5). Controls (C1 and C2) are representative of three controls. Fibroblasts were grown in DMEM with 10% FBS. Ferredoxin reductase was studied in reducing conditions, and SOD2 and TfR1 were studied in non-reducing conditions so that TfR1 dimerization would be maintained. GAPDH was used as a loading control. (B) Upper panel: in-gel aconitase assays revealed decreased mitochondrial (m-aco) and cytosolic (c-aco) aconitases in S1 and S5 fibroblasts compared to controls (C1–3). Lower panel: decreased HO-1 steady-state levels in S1 and S5 fibroblasts. C1–3: controls 1–3. (C) Measurement of mitochondrial reactive oxygen species (ROS). Fibroblasts were grown in low- (-FAC) or high- (+FAC) iron conditions, and ROS amounts were similar in both conditions for control (C is representative of three controls) and subjects 1 and 5. The figure presents the ROS content in high-iron conditions. The numbers represent the percent of fluorescence intensity of subjects’ fibroblasts compared to controls. ROS were measured as previously described. (D) Steady-state levels of proteins involved in iron homeostasis. TfR1, SOD1, and SOD2 were studied in non-reducing conditions, ferritin (FTH, FTL), ACO1, ACO2, and SDH were studied in reducing conditions in controls (C1–2), and FDXR fibroblasts grown in FBS-free DMEM medium were studied in low (-FAC) and high- (+FAC) iron conditions. GAPDH was used as a loading control. (E) Iron quantification using the ferrozine-based colorimetric assay in fibroblasts of controls (C1–C4) and FDXR subjects (S1, S5) grown in FBS-free DMEM medium and low- (-FAC) or high- (+FAC) iron conditions. The data are the means ± SEM of three independent experiments. A Student-Newman-Keuls ANOVA multifactorial test was used. ^∗∗∗p < 0.001. We harvested cultured skin fibroblasts on ice by scraping cells in reducing (2 μM DTT, denaturation at 95°C for 5 min) or non-reducing (no DTT, no heat denaturation) cell lysis buffer. SDS-PAGE was performed on whole-cell protein extracts in 10% or 7.5% acrylamide gel for reducing and non-reducing conditions, respectively. Immunodetections were performed in PBS with 5% milk and 0.05% Tween 20 (SIGMA) and the following antibodies: mouse anti-TfR1 (ab108985), rabbit anti-ferritin (ab75973), rabbit anti-FDXR (ab204310), rabbit anti-HO1 (Enzo Life Sciences, ADI-SPA-896), mouse anti-SOD1 (ab86454), mouse anti-SOD2 (ab125702), rabbit anti-GAPDH (ab62489), rabbit anti-ACO1 (ab126595), rabbit anti-ACO2 (ab129069). HRP-conjugated goat anti-mouse IgG H&L (ab6789), or goat anti-rabbit IgG (ab6721) were used as secondary antibodies before ECL-based detection (SuperSignal West Dura, Thermo Scientific). Images were visualized on a CCD camera (BioRad) and analyzed with ImageLab software (BioRad).

Figure 4

Functional Complementation Assay of the Yeast arh1Δ Mutant Strain Expressing Either the Wild-Type FDXR or the FDXR Mutants (A) Growth on synthetic dextrose (SD) with or without 5-fluoroorotic acid (5-FOA) of arh1Δ and not D cells expressing the yeast wild-type ARH1 and transformed with the wild-type FDXR (wt), the FDXR mutants, or the corresponding empty plasmid (-) under the control of the strong PGK promoter. 5-FOA treatment, as described by Boeke, eliminates the plasmid encoding the yeast ARH1 and allows one to evaluate growth capability of the arh1Δ cells expressing only FDXR or its mutated forms. (B) Growth after 5-FOA treatment of wt cells (BY4741) and arh1Δ cells expressing wtARH1, FDXR, or the FDXR mutants. Cells were spotted onto fermentable (YPD) or non-fermentable (YPG) plates. Drop dilution growth tests were performed at 1/10 dilution steps, and plates were incubated for 3 days at 28°C. (C) FDXR mutations affect aconitase activity. Aconitase assays on crude yeast extracts from arh1Δ cells expressing the wild-type FDXR (wt) or the mutant forms Arg306Cys and Leu215Val are shown. Cells were grown to mid-logarithmic phase, and total crude extracts were prepared and analyzed for aconitase activity as previously described. Consumption of cis-aconitate was monitored spectrophotometrically at 240 nm at a molar absorption of 3,600 M⁻¹ × cm⁻¹. Activities shown are an average of two independent experiments. Error bars indicate SD.