Mutations in mitochondrial ferredoxin FDX2 suppress frataxin deficiency

Abstract

Frataxin is a key component of an ancient, mitochondrial iron-sulfur cluster biosynthetic machinery, serving as an allosteric activator of the cysteine desulfurase NFS1 (refs. ^1-5). Loss of frataxin levels underlies Friedreich's ataxia⁶, the most common inherited ataxia. Yeast, Caenorhabditis elegans and human cells can tolerate loss of frataxin when grown in 'permissive' low oxygen tensions⁷. Here we conducted an unbiased, genome-scale forward genetic screen in C. elegans leveraging permissive and non-permissive oxygen tensions to discover suppressor mutations that bypass the need for frataxin. All mutations act dominantly and are in the ferredoxin FDX2/fdx-2 or in the cysteine desulfurase NFS1/nfs-1 genes, resulting in amino-acid substitutions at the FDX2-NFS1 binding interface. Our genetic and biochemical analyses show that the suppressor mutations boost iron-sulfur cluster levels in the absence of frataxin. We also demonstrate that an excess of FDX2 inhibits frataxin-stimulated NFS1 activity in vitro and blocks the synthesis of iron-sulfur clusters in mammalian cell culture. These findings are consistent with structural and biochemical evidence that frataxin and FDX2 compete for occupancy at the same site on NFS1 (refs. ^8,9). We show that lowering levels of wild-type FDX2 through loss of one gene copy can ameliorate the growth of frataxin mutant C. elegans or the ataxia phenotype of a mouse model of Friedreich's ataxia under normoxic conditions. These genetic and biochemical studies indicate that restoring the stoichiometric balance of frataxin and FDX2 through partial knockdown of FDX2 may be a potential therapy for Friedreich's ataxia.

Fig. 1. Mutations in NFS1/nfs-1 or FDX2/fdx-2 partially rescue the growth defect caused by frataxin loss in C. elegans.

a, A forward genetic screen using random chemical mutagenesis revealed particular substitution mutations in fdx-2 and nfs-1 that can rescue the loss of frataxin. b, Multiple sequence alignment of NFS1 (C. elegans residues 239–251) and FDX2 (C. elegans residues 117–127) including homologues from mammals, fish and invertebrates made using ClustalW. c, Synchronized frataxin-null animals grown at 7% oxygen for 4 days with or without suppressor mutations in fdx-2 and nfs-1. Scale bar, 1 mm. d–f, Growth of C. elegans at 7% oxygen (4 days) (d), 1% or 10% oxygen (2 days) (e) or 7% oxygen (2 days) (f) quantified by body length measurements. The number of individual worms in each group was: all groups n = 12 (d), all groups n = 12–14 (e), frh-1 n = 27, frh-1; nfs-1 n = 22, frh-1; nfs-1/+ n = 8, frh-1; fdx-2 n = 17, frh-1; fdx-2/+ n = 8 (f). For all panels, statistical significance was calculated using one-way analysis of variance (ANOVA) followed by Dunnett’s (d,f) or Sidak’s (e) multiple comparison test. Error bars represent mean ± s.d. NS, not significant, *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 2. Frataxin suppressor mutations restore levels of Fe–S cluster-containing ETC complexes.

a, Mean intestinal fluorescence of age-matched day 1 adult animals containing hsp-6::gfp exposed to 21% oxygen for 24 h. The number of individual worms in each group was n = 11–12. b, Quantitative TMT proteomics of complex I subunits in wild-type animals, frataxin mutants and frataxin mutants with suppressor mutations nfs-1(R244K) or fdx-2(A126V) grown continuously at 1% oxygen. Values are normalized to wild type; each line represents one protein from n = 1 experiment. Proteins shown from which at least two peptides were quantified. c, Growth of C. elegans at 1% or 10% oxygen for indicated durations. The number of individual worms in each group was n = 20. For all panels, statistical significance was calculated using one-way ANOVA followed by Sidak’s multiple comparison test (a,c). Error bars represent mean ± s.d. *P < 0.05, **P < 0.01, ***P < 0.001. a.u., arbitrary units.

Fig. 3. Excess FDX2 is detrimental to NFS1 activity and Fe–S cluster biosynthesis.

a, Growth of animals for 3 days at room temperature exposed to 21% or 1% oxygen. The number of individual worms in each group was n = 20. b, Cryo-EM structure of the human Fe–S cluster assembly complex containing FDX2 (ref. ) (Protein Data Bank 8RMC) with boxes indicating homologous residues to C. elegans suppressor mutations. c, Cysteine desulfurase activities of SDA_ec (0.5 µM; NFS1–ISD11–ACP_ec) complexes containing ISCU2 (1.5 µM; white) and increasing equivalents of FXN (0.25–30 µM; grey), FDX2_ox (0.25–30 µM, oxidized ferredoxin 2; blue) and FDX2_ox E131K (0.25–30 µM; purple). Reactions were initiated with 2 mM l-cysteine, quenched after 3 min and the sulfide was converted to methylene blue to determine activities. For all groups n = 3 independent experiments. d, Iron–sulfur assembly activities on ISCU2 (100 µM) for complexes containing SDA_ec (1 µM), FXN (0–60 μM), FDX2 or FDX2 E131K (0–60 μM), FDXR (1 μM) and NADPH (500 μM). Reactions were initiated with 500 μM l-cysteine, and the change in ellipticity at 430 nm was fit to a linear equation to determine the rate of Fe–S assembly on ISCU2. The numbers on the x axis represent protein equivalents compared with SDA_ec. ND, not detectable. For all groups n = 3 independent experiments. The initial and final circular dichroism spectra and averaged change in ellipticity at 430 nm for determining the reaction rate for each experiment are shown in Extended Data Fig. 5a–c. e, In wild-type conditions frataxin and FDX2 compete for binding to NFS1, promoting persulfide transfer and donating electrons, respectively, at distinct steps in the biosynthesis of Fe–S clusters. In the context of low or zero frataxin, or FDX2 overexpression, Fe–S synthesis is blocked due to frataxin displacement by FDX2 and/or FDX2 directly inhibiting NFS1 activity. Point mutations that weaken the NFS1–FDX2 interaction, or simply lowering FDX2 levels to generate NFS1 unbound by FDX2, can partially restore Fe–S synthesis and support growth in C. elegans. f, Three-day growth assay with K562 or 293T cells overexpressing (O/E) GFP or FDX2 cDNA, grown in 21% or 1% oxygen tensions. Experiments were conducted in three biological replicates in technical duplicate for a total n = 6. g, Immunoblots for FDX2, lipoic acid, OXPHOS, NFS1 and loading control tubulin in K562 (left) and 293T (right) cell lysates collected from f. For gel source data, see Supplementary Data 1. For western blotting n = 3 biological replicates; all individual replicates for lipoic acid are shown and quantified in Supplementary Data 2. Statistical significance was calculated using one-way (a) or two-way (f) ANOVA followed by Sidak’s multiple comparison test. Error bars represent mean ± s.d. *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 4. Partial loss of wild-type FDX2 activity suppresses frataxin mutants.

a, C. elegans animals carrying a CRISPR–Cas9-generated fdx-2 null mutation (mut) (8-base-pair deletion), predicted to produce no functional protein, were propagated as heterozygotes (het) and through self-fertilization produced wild types, heterozygotes and homozygous mutants. b, Growth of animals for 3 days at 21% oxygen. All animals were healthy and fertile except for fdx-2 null homozygotes. The number of individual worms in each group was n = 20. c, Growth of animals for 3 days at 1% oxygen (left) or 6 days at 10% oxygen (right). Frataxin-null animals are rescued by a heterozygous null mutation in fdx-2. The number of individual worms in each group was n = 20 (left) or n = 12–15 (right). d, Schematic of mouse cross and experimental design. e, Grip strength (left) and latency to fall on a rotarod test (right) in mice on doxycycline treatment for 12 weeks. The number of individual mice in each group was: wild type n = 22, Fdx2 het n = 20, shFxn het n = 13, double het n = 17. For all panels, statistical significance was calculated using one-way ANOVA followed by Dunnett’s (b) or Tukey’s (c) multiple comparison test, or a two-tailed unpaired t-test (e). Error bars represent mean ± s.d. *P < 0.05, **P < 0.01, ***P < 0.001. Schematics in a and d were created using BioRender (https://biorender.com).

Extended Data Fig. 1. Mutations in NFS1/nfs-1 or FDX2/fdx-2 partially rescue the growth defect caused by frataxin loss in C. elegans.

a, Phylogenetic tree of FDX1 and FDX2 homologs from E. coli, S. cerevisiae, Amphimedon queenslandica, Nematostella vectensis, C. elegans, D. melanogaster, D. rerio, and H. sapiens. Alignments were made with Clustal Omega, maximum-likelihood trees made using PhyML and visualized with FigTree. b, Multiple sequence alignment of C. elegans FDX-2 with human FDX2 and FDX1. Sequences necessary and sufficient for the activity of FDX2 are boxed in blue, and a residue critical for the activity of FDX1 is boxed in orange. c,d,e,f, Growth of C. elegans at 10% oxygen (5 days) (c), 21% oxygen (2 days) (d), 21% oxygen (3 days) (e), or 1% oxygen (2 days) (f) quantified by body length measurements. The number of individual worms in each group was: all groups n = 20 (c), all groups n = 13-14 (d), all groups n = 20 (e), all groups n = 12 (f). For all panels statistical significance was calculated using one-way ANOVA followed by Dunnett’s (e,f) or Sidak’s (c,d) Multiple Comparison Test. Error bars represent mean ± standard deviation. n.s. = not significant, * = p value < 0.05, ** = p value < 0.01, *** = p value < 0.001.

Extended Data Fig. 2. Frataxin suppressor mutations do not fully alleviate stress-response pathway induction.

a,b, Mean intestinal fluorescence of age-matched day 2 adult animals containing hsp-6::gfp (a) or gst-4::gfp (b) grown continuously at 1% oxygen or exposed to 21% oxygen for 48 h. The number of individual worms in each group was: all groups n = 19-26 (a), all groups n = 21-24 (b). For all panels statistical significance was calculated using one-way ANOVA followed by Sidak’s Multiple Comparison Test (a,b). Error bars represent mean ± standard deviation. n.s. = not significant, * = p value < 0.05, ** = p value < 0.01, *** = p value < 0.001.

Extended Data Fig. 3. Frataxin suppressor mutations restore levels of Fe–S cluster-containing ETC complexes.

a, TMT quantitative proteomics from animals grown continuously at 1% oxygen or animals shifted from 1% to 21% for 2 days. Plotted are log 2-fold ratios relative to wild type at 21% oxygen of all proteins from which at least two peptides were quantified. Fe–S cluster (also known as ISC)-containing proteins are annotated based on Andreini et al.. supplemented with recently identified new Fe–S cluster proteins^,. n = 1 experimental replicate. b,c, SDS-PAGE followed by western blot of whole worm lysate from animals exposed to continuous 1% oxygen (b) or shifted from 1% to 21% for 2 days (c). Arrow indicates ATP5A, star indicates non-specific band. For gel source data, see Supplementary Data 1.

Extended Data Fig. 4. FDX2 and NFS1 suppressor mutation interact genetically and physically.

a, SDS-PAGE followed by western blot of whole worm lysate from animals grown continuously at 1% oxygen or shifted from 1% to 21% for 2 days. For gel source data, see Supplementary Data 1. b, Cryo-EM structure of the human Fe–S cluster assembly complex containing FDX2 (PDB: 8RMC) or FXN (PDB: 6NZU) with NFS1/FDX2 and NFS1/FXN interfaces exposed and electrostatic potential indicated.

Extended Data Fig. 5. Excess FDX2 is detrimental to NFS1 activity and Fe–S cluster biosynthesis.

a, Initial (dashed line) and final CD spectrum of Fe-S assembly reactions. b, Iron-sulfur assembly reactions on ISCU2 were monitored by the change in ellipticity at 430 nm. Spectra (300–700 nm) were measured every 3 min for 30 min, and the average change in ellipticity was plotted. The change in ellipticity at 430 nm was fit to a linear equation to determine the rate of Fe-S assembly on ISCU2. n = 3 independent experimental replicates. Numbers represent the equivalents of frataxin (F), and ferredoxin (X) compared to SDAec. Graphs display mean ± standard deviation (left) or individual data points (right). c, Raw CD spectrum of ISCU2 assembly reactions without X and F (initial spectrum, red; final spectrum, blue). Spectra were recorded every 3 min for 30 min (black lines).

Extended Data Fig. 6. Partial loss of wild type FDX2 activity suppresses frataxin mutants.

a, Broodsize of wild type and fdx-2(null) animals grown continuously at 21% or 1% oxygen. The number of individual worms in each group was: wild type 21% and 1% O₂ n = 5, fdx-2 21% and 1% O₂ n = 10. b, SDS-PAGE followed by western blot of brain tissue from wild type mice or mice heterozygous for Fdx2 mutation. Arrow indicates FDX2, star indicates non-specific band. For gel source data, see Supplementary Data 1. c, Quantification of western blot using Fiji. n = 3 individual mice representing biological replicates for each group (b,c). d, Body weight measurements of male (left) and female (right) mice following Doxycycline treatment. The number of individual mice in each group was: wild type males n = 12, Fdx2 het males n = 9, shFxn het males n = 10, double het males n = 12, wild type females n = 10, Fdx2 het females n = 11, shFxn het females n = 10, double het females n = 14. Mean ± standard deviation is plotted, individual data points are available in Supplementary Table 4. e, Survival of mice following Doxycycline treatment. The number of individual mice in each group was: wild type n = 22, Fdx2 het n = 20, shFxn het n = 20, double het n = 26. For all panels statistical significance was calculated using one-way ANOVA followed by Sidak’s (a) Multiple Comparison Test or a two-tailed unpaired t-test (c). Error bars represent mean ± standard deviation. n.s. = not significant, * = p value < 0.05, ** = p value < 0.01, *** = p value < 0.001.