NDUFAB1 confers cardio-protection by enhancing mitochondrial bioenergetics through coordination of respiratory complex and supercomplex assembly

Abstract

The impairment of mitochondrial bioenergetics, often coupled with exaggerated reactive oxygen species (ROS) production, is a fundamental disease mechanism in organs with a high demand for energy, including the heart. Building a more robust and safer cellular powerhouse holds the promise for protecting these organs in stressful conditions. Here, we demonstrate that NADH:ubiquinone oxidoreductase subunit AB1 (NDUFAB1), also known as mitochondrial acyl carrier protein, acts as a powerful cardio-protector by conferring greater capacity and efficiency of mitochondrial energy metabolism. In particular, NDUFAB1 not only serves as a complex I subunit, but also coordinates the assembly of respiratory complexes I, II, and III, and supercomplexes, through regulating iron-sulfur biosynthesis and complex I subunit stability. Cardiac-specific deletion of Ndufab1 in mice caused defective bioenergetics and elevated ROS levels, leading to progressive dilated cardiomyopathy and eventual heart failure and sudden death. Overexpression of Ndufab1 effectively enhanced mitochondrial bioenergetics while limiting ROS production and protected the heart against ischemia-reperfusion injury. Together, our findings identify that NDUFAB1 is a crucial regulator of mitochondrial energy and ROS metabolism through coordinating the assembly of respiratory complexes and supercomplexes, and thus provide a potential therapeutic target for the prevention and treatment of heart failure.

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1

Cardiac-specific knockout of Ndufab1 (cKO) causes dilated cardiomyopathy leading to heart failure in mice. a Western blots of NDUFAB1 in wild-type (WT) and cKO cardiomyocytes. ATPB served as the loading control. b Left, photograph of WT and cKO mice at 14 weeks of age. Right, effect of cardiac NDUFAB1 ablation on mouse body weight (mean ± s.e.m.; n = 3–16 male mice per group per time point; **p < 0.01 versus WT at the same time point). c Kaplan–Meier survival curves for WT and cKO mice (n = 11 male mice per group; **p < 0.01 versus WT). d Representative longitudinal sections of hearts from male mice at 6–16 weeks (w) old. Scale bar, 5 mm. e Ratios of heart weight (HW) to body weight (BW) or tibia length (TL) (mean ± s.e.m.; n = 3–14 male mice per group; *p < 0.05, **p < 0.01 versus WT at the same age). f Progressive hypertrophy of cardiomyocytes from cKO mice. Left: representative confocal micrographs of cardiomyocytes stained with DCF (scale bar, 20 μm). Right: statistical analysis (mean ± s.e.m.; n = 97–611 cells from 4 male mice per group; ** p < 0.01 versus WT at the same age). g Fibrosis of left ventricular tissue (Masson’s trichrome staining; scale bar, 100 μm). h Echocardiographic analysis of cardiac function. EF, ejection fraction; FS, fractional shortening (mean ± s.e.m.; n = 3–9 male mice per group; **p < 0.01 versus WT at the same age)

Fig. 2

Defective mitochondrial bioenergetics and ROS metabolism in Ndufab1 cKO mice. a Electron micrographs of mitochondria in left ventricular tissue from 6- and 16-week-old male mice (scale bars, 1 μm). b Decline of ΔΨm in cKO myocardium assessed by TMRM staining and confocal imaging of isolated cardiomyocytes. The intensity of FCCP-sensitive TMRM fluorescence was calculated and the intensity in cKO group was normalized to that of WT at the same age (n = 13–26 cells from three male mice per group; *p < 0.05, **p < 0.01 versus WT). c Progressive elevation of mitochondrial ROS measured with mitoSOX in isolated cardiomyocytes. Antimycin A (AA, 5 μg/mL) was used to induce massive ROS production. The intensities of mitoSOX fluorescence in the groups of cKO, WT + AA and cKO + AA were normalized to that of WT at the same age (n = 31–179 cells from 4–5 male mice per group; ** p < 0.01 versus WT). d ATP levels of isolated cardiomyocytes measured with luciferin assay (n = 3–4 male mice per group; ** p < 0.01 versus WT). e Top 5 enriched Gene Ontology terms of downregulated and upregulated genes. RNA-seq analysis was performed on cKO and WT hearts at 8–16 weeks (n = 4 male mice per group) and the genes upregulated or downregulated were analyzed

Fig. 3

Defective respiration and impaired assembly of ETC complexes I-III and SCs in cKO hearts. a Effect of NDUFAB1 ablation on state III oxygen consumption rate (OCR) in isolated mitochondria at the age of 6 or 10 weeks. For each measurement, 100 μg mitochondria were used. Different respiratory substrates were used: malate/glutamate (Mala/glu, 5 mM) for complex I, succinate (Succ, 5 mM) for complex II, glycerol-3-phosphate (G3P, 5 mM) for complex III, and ascorbate (Asc, 2.5 mM)/N,N,N’,N’-tetramethyl-p-phenylenediamine (TMPD, 0.5 mM) for complex IV, and ADP (100 μM) (mean ± s.e.m.; n = 3–14 male mice of per group; *p < 0.05, **p < 0.01 versus WT at the same age). b In-gel activity of SCs and complex I at the age of 6 or 16 weeks. c BN-PAGE immunoblots of individual ETC complexes and SCs at the age of 6 or 16 weeks. SCs were visualized by antibodies against subunits of complex I (CI, NDUFB8), complex III (CIII, UQCRFS1 or UQCRC1), and complex IV (CIV, COX IV). Complex II was visualized by antibody against SDHA, and complex V with antibody against ATPB. d–f Western blots of subunits of complex I (d), complex II (e), and complex III (f) at the age of 6 or 16 weeks. Dagger represents FeS-containing subunits. ATP5A served as the loading control

Fig. 4

Enhanced mitochondrial bioenergetics and SC assembly in the heart of Ndufab1 transgenic mice (TG). TG and WT littermates of 2–4 months old were used. a Western blots for NDUFAB1 expression in WT and TG hearts. ATPB served as the loading control. b Decreased mitochondrial ROS level in TG cardiomyocytes. The intensity of mitoSOX fluorescence of TG was normalized to that of WT (n = 34–41 cells from 3 male mice; **p < 0.01 versus WT). c Increased ΔΨm in TG cardiomyocytes. The intensity of FCCP-sensitive TMRM fluorescence of TG was normalized to that of WT (n = 12–15 cells from three male mice per group; *p < 0.05 versus WT). d, e Whole-cell OCR measured with Seahorse using glucose/pyruvate (d) or palmitate (e) as the substrate. Arrows indicate the time of adding palmitate (PAL, 200 µM), oligomycin (Oligo, 1 µM), FCCP (0.5 µM), and rotenone/antimycin A (Rot/AA, 1 µM for each). f, g Statistical analysis of maximal OCR in d and e (mean ± s.e.m.; n = 10–13 wells from 3 male mice per group; *p < 0.05 versus WT). h Changes of respiratory control ratio (RCR) with different substrates in isolated mitochondria (mean ± s.e.m.; n = 3–8 male mice per group; *p < 0.05, **p < 0.01 versus WT). i In-gel activity of SCs and complex I. j Statistical analysis of i. The activity of TG was normalized to WT (mean ± s.e.m.; n = 5 male mice per group; *p < 0.05, **p < 0.01 versus WT). k BN-PAGE immunoblots of SCs and individual ETC complexes as in Fig. 3c. l Statistical analysis of k. The expression level of TG complexes and supercomplexes was normalized to that of WT (mean ± s.e.m.; n = 4–8 male mice per group; *p < 0.05 versus WT)

Fig. 5

NDUFAB1 overexpression protects the heart against IR injury. TG and WT male littermates of 2–4 months old were used. a Alleviated myocardial ROS production in TG hearts. Confocal images of DCF-stained myocardium after Langendorff perfusion and IR injury. Ischemia (IS) was mimicked by stopping perfusion. See Supplementary information, Fig. S19a for experimental protocol. Scale bars, 20 μm. b Statistical analysis of a (mean ± s.e.m.; n = 32–51 image frames from 4–7 male mice per group; **p < 0.01 versus WT, #p < 0.01 after IR versus basal). c Rotenone (Rot)-sensitive H2O2 production at complex I in WT and TG cardiac mitochondria (mean ± s.e.m.; n = 6 male mice per group; **p < 0.01 versus WT). Succinate (Succ, 5 mM) was used as the substrate of complex II and H2O2 was measured with Amplex red. d Images for simultaneous measurement of ∆Ψm (TMRM, pseudocolor: red) and cardiomyocyte death (EBD uptake, pseudocolor: green) in Langendorff-perfused hearts subjected to IR injury. Note that all cells with intact ∆Ψm are EBD-negative, and all EBD-positive cells display loss of ∆Ψm, but not vice versa. Scale bars, 50 μm. e, f Percentages of cells with loss of ∆Ψm (e) and EBD-positive cells (f) (mean ± s.e.m.; n = 9–10 male mice per group; *p < 0.05 versus WT, #p < 0.01 after IR versus basal). g Representative cross-sections of WT and TG hearts after IR injury. See Supplementary information, Fig. S19b for the experimental protocol. White, infarcted area (IF); red, rest of the area at risk (AAR); blue, tissue not at risk. The AAR was demarcated by dotted line. Scale bar, 1 mm. h, i Statistical analysis of AAR (h) and IF (i) reported as the ratios of AAR to left ventricular area (LV) and IF to AAR (n = 5–8 male mice per group; **p < 0.01 versus WT). j Measurement of serum LDH release after IR (n = 10–13 male mice per group; *p < 0.05 versus WT, #p < 0.01 after IR versus basal). k Echocardiographic analysis of cardiac function 48 h after IR injury. EF, ejection fraction; FS, fractional shortening (mean ± s.e.m.; n = 4 male mice per group; *p < 0.05 versus WT)