SDHAF4 promotes mitochondrial succinate dehydrogenase activity and prevents neurodegeneration

Abstract

Succinate dehydrogenase (SDH) occupies a central place in cellular energy production, linking the tricarboxylic cycle with the electron transport chain. As a result, a subset of cancers and neuromuscular disorders result from mutations affecting any of the four SDH structural subunits or either of two known SDH assembly factors. Herein we characterize an evolutionarily conserved SDH assembly factor designated Sdh8/SDHAF4, using yeast, Drosophila, and mammalian cells. Sdh8 interacts specifically with the catalytic Sdh1 subunit in the mitochondrial matrix, facilitating its association with Sdh2 and the subsequent assembly of the SDH holocomplex. These roles for Sdh8 are critical for preventing motility defects and neurodegeneration in Drosophila as well as the excess ROS generated by free Sdh1. These studies provide insights into the mechanisms by which SDH is assembled and raise the possibility that some forms of neuromuscular disease may be associated with mutations that affect this SDH assembly factor.

Figure 1. Yeast Sdh8 is a conserved mitochondrial matrix protein required for maximal SDH activity and assembly

(A) Ten-fold serial dilutions of WT yeast transformed with an empty vector (EV) or sdh8Δ mutants transformed with either EV or a plasmid expressing SDH8-His₆HA₃, all in a BY4741 background. (B) GC/MS was used to measure the abundance of metabolites in WT and sdh8Δ yeast (BY4741 background. N=6 biological replicates. ****p < 0.0001; **p < 0.005; *p < 0.05). (C) SDH and (D) MDH enzyme assays were performed on mitochondrial extracts of WT, sdh8Δ, and sdh1Δ strains, normalized to total protein (±SEM. N=3 biological replicates. **p < 0.005). (E) WT and sdh8Δ mitochondria were solubilized in digitonin, fractionated by BN-PAGE, and analyzed by immunoblotting to detect Atp2 (complex V), Cor1/2 (complex III), Cox3 (complex IV), Sdh1 (complex II), or Porin. (F) Ten-fold serial dilutions of the indicated strains were grown as in (A). SDHAF4 and dSdhaf4 are the human and Drosophila orthologs of SDH8, respectively, and were each expressed under the control of the endogenous SDH8 promoter and terminator.

Figure 2. Mammalian SDHAF4 is required for maximal SDH activity

(A) C2C12 cells were transfected with either a control non-targeting siRNA or either of two siRNAs directed against SDHAF4 (si1 and si2), and SDHAF4 mRNA abundance was assayed by quantitative RT-PCR and normalized to tubulin mRNA (±SEM. N=3 biological replicates. ***p < 0.0005). (B) SDH enzyme activity was measured in mitochondria harvested from C2C12 cells treated as in (A) (±SEM. N=3 biological replicates. ***p < 0.0005; *p < 0.05). (C) Mitochondria extracted C2C12 cells treated as in (A) were solubilized in digitonin, fractionated by BN-PAGE, and analyzed by immunoblotting for ATP5A (complex V monomers (CV_I) or dimers (CV_II), QCR2 (complex III), VDAC, or SDHA (complex II).

Figure 3. Drosophila Sdhaf4 is required for SDH activity

(A) Either LC/MS (acetyl-CoA, succinyl-CoA, ATP, AMP, NAD, NADH, NADP, and NADPH) or GC/MS (citrate, isocitrate, alpha-ketoglutarate, succinate, fumarate and malate) was used to measure the abundance of metabolites in trans-heterozygous dSdhaf4^1/2 (dSdhaf4⁻) mutants compared to genetically matched w¹¹¹⁸ controls (dSdhaf4⁺). 8–12 biological replicates from two independent experiments were combined per genotype. (B) SDH or (C) citrate synthase enzymatic activity was measured in extracts from mitochondria isolated from either control or dSdhaf4 mutant flies (±SEM. N=4–6 biological replicates. **p < 0.01 and ***p <0.001). Two independent experiments were performed with similar results.

Figure 4. Drosophila Sdhaf4 is required for SdhA and SdhB stability

(A) Mitochondrial extracts from either w¹¹¹⁸ controls (dSdhaf4⁺) or dSdhaf4^1/2 mutants (dSdhaf4) were fractionated by BN-PAGE and analyzed by immunoblotting to detect Atpα (Complex V) or SdhB (Complex II). (B) Total protein from mitochondria isolated from either control or dSdhaf4 mutant flies was resolved using SDS-PAGE, followed by immunoblotting to detect SdhA, SdhB or Atpα. (C) Flies carrying either control or dSdhaf4 mutant chromosomes over a balancer chromosome were crossed to flies carrying a strong loss-of-function allele for either SdhA or Cytochrome c oxidase Va subunit (CoVa) over a balancer. The resulting progeny were scored for the absence or presence of the marker linked to the balancer chromosome. The proportion of the expected progeny that eclosed, based on Mendelian ratios of the indicated genotypes, is shown. (±SEM. N = number of progeny assayed from each cross. ***p <0.001 (Student’s t test).

Figure 5. Drosophila Sdhaf4 mutants display neurodegenerative phenotypes, early-adult lethality, and sensitivity to oxidative stress

(A) The number of surviving w¹¹¹⁸ control (dSdhaf4⁺) or dSdhaf4^1/2 mutant (dSdhaf4⁻) females maintained on standard laboratory food at normoxia was assayed at 2–4 day intervals. Each data point is the percentage of surviving flies±SEM, with at least 110 flies assayed per genotype. (B) Control or dSdhaf4 mutant females aged 6–8 hours, 2–4 days, or 8–12 days were vortexed for 15 seconds and the percentage of paralyzed flies was counted as described (Ganetzky and Wu, 1982). (±SEM. N>32 flies per age per genotype. **p <0.01 and ***p <0.001 (ANOVA)). (C) The length of time required for the dSdhaf4 mutant flies in (A) to recover from paralysis is presented as a bar graph, with the number of paralyzed dSdhaf4 mutants analyzed from each age group depicted above the bar. (D) Cross sections of compound eyes from 8–12 day old control or dSdhaf4 mutant flies visualized by transmission electron microscopy. Arrows point to aberrant vesicularization of dSdhaf4 mutant photoreceptors, arrowheads point to mitochondria, and asterisks denote vacuoles (E) Control or dSdhaf4 mutant flies at 8–12 days of age carrying either the ubiquitous Daughterless-Gal4 (Da-GAL4) driver, the pan-neuronal Elav-Gal4 driver, or no driver, in the presence or absence of either a UAS-dSdhaf4 transgene or a UAS-human Sdhaf4 (hSdhaf4) transgene, were vortexed as in (A). Note that UAS-dSdhaf4 alone has leaky dSdhaf4 activity since it partially rescues in the absence of a GAL4 driver. (±SEM. N=15–40. ***p <0.001 (ANOVA)). (F) Control or dSdhaf4 mutant females at 2–4 days of age were transferred to 100% oxygen and the number of surviving animals was counted each day. The median lifespan after transfer to hyperoxia is 9 days for controls and 1 day for dSdhaf4 mutants. Data is shown as the percentage of surviving flies±SEM. At least 40 flies were assayed per genotype. P values for both survival curves (panels A, F) were calculated using a log-rank test.

Figure 6. Yeast Sdh8 is a co-chaperone for unbound and covalently flavinated Sdh1

(A) Sdh8-His₆HA₃ was immunoprecipitated from digitonin-solubilized mitochondria isolated from the yeast strains described as follows. Lanes 1 and 6, WT cells; Lanes 2 and 7, sdh8Δ cells expressing Sdh8-His₆HA₃; Lanes 3 and 8, sdh2Δ sdh8Δ cells expressing Sdh8-His₆HA₃; Lanes 4 and 9 sdh1Δ sdh8Δ cells expressing Sdh8-His₆HA₃ and Sdh1^H90A; Lanes 5 and 10, sdh1Δ sdh2Δ sdh8Δ cells expressing Sdh8-His₆HA₃ and Sdh1^H90A. (B) Mitochondrial lysates isolated from the following strains were fractionated by SDS-PAGE. Lane 1, sdh8Δ cells expressing Sdh8-His₆HA₃; Lane 2, sdh1Δ sdh8Δ cells expressing Sdh8-His₆HA₃; Lane 3, sdh2Δ sdh8Δ cells expressing Sdh8-His₆HA₃; Lane 4, sdh8Δ cells; Lane 5, sdh2Δ sdh8Δ cells (BY4741 background). Steady state protein abundance was assessed by immunoblot using antibodies against Sdh1, Sdh2, and HA (Sdh8-His₆HA₃). (C) Mitochondria isolated from the following yeast strains were solubilized in digitonin and fractionated by BN-PAGE. Lane 1, sdh8Δ cells expressing Sdh8-His₆HA₃; Lane 2, sdh2Δ sdh8Δ cells expressing Sdh8-His₆HA₃; Lane 3, sdh1Δ sdh2Δ sdh8Δ cells expressing Sdh8-His₆HA₃; Lane 4, sdh2Δ sdh8Δ; Lane 5, sdh4Δ sdh8Δ cells expressing Sdh8-His₆HA₃; Lane 6, sdh1Δ sdh4Δ sdh8Δ cells expressing Sdh8-His₆HA₃, Lane 7, sdh4Δ sdh8Δ cells. Immunoblotting was used to detect Atp2 (complex V) as a control and complexes containing Sdh1, Sdh2, and HA (Sdh8-His₆HA₃). SDH complex (~232 kDa), Sdh1-Sdh8 sub-complex (~150 kDa).

Figure 7. Sdh8 protects the cell from oxidative stress

(A) Ten-fold serial dilutions of WT and sdh8Δ yeast either transformed with an empty vector or over-expressing Sdh1 were plated on media containing the indicated carbon source (W303 background). (B) Ten-fold serial dilutions of WT yeast transformed with an empty vector, WT yeast overexpressing Sdh1, and WT yeast overexpressing Sdh1 and Sdh8-His₆HA₃ were plated on media containing acetate (W303 background). (C) Ten-fold serial dilutions of WT yeast transformed with an empty vector, sdh8Δ cells transformed with an empty vector, and sdh8Δ cells expressing Sdh8-His₆HA₃ were plated on media containing glycerol with and without 1mM paraquat and incubated at 30°C (W303 background) as in (A). (D) Aconitase activity was measured in WT and sdh8Δ cells grown to stationary phase in synthetic complete media containing 2% glucose and normalized to total protein (±SEM; N=4 biological replicates. **p < 0.005). (E) WT cells (solid line), sdh8Δ cells (dashed line), and sdh8Δ cells expressing Sdh8-His₆HA₃ (dotted line) transformed with mito-roGFP were plated on synthetic media containing 2% acetate, allowed to grow for 48 hours and analyzed by flow cytometry. Shown are representative histograms measuring the ratio of oxidized/reduced mito-roGFP in 30,000 cells (BY4741 background). (F) The median value of the ratio of oxidized/reduced mito-roGFP is plotted for WT cells, sdh8Δ cells, and sdh8Δ cells expressing Sdh8-His₆HA₃ (±SEM; N=4 biological replicates. ****p < 0.0001). (G) Ten-fold serial dilutions of WT yeast transformed with an empty vector, sdh8Δ cells transformed with an empty vector, and sdh8Δ cells over-expressing YAP1 were plated on media containing the indicated carbon source and incubated at 30°C (BY4741 background) as in (A). (H) SDH activity enzyme assays were performed using mitochondria harvested from WT and sdh8Δ cells carrying an empty vector or overexpressing YAP1. Activity was normalized to total protein (±SEM; N=3 biological replicates. **p < 0.005).