The OPA1-dependent mitochondrial cristae remodeling pathway controls atrophic, apoptotic, and ischemic tissue damage

Abstract

Mitochondrial morphological and ultrastructural changes occur during apoptosis and autophagy, but whether they are relevant in vivo for tissue response to damage is unclear. Here we investigate the role of the optic atrophy 1 (OPA1)-dependent cristae remodeling pathway in vivo and provide evidence that it regulates the response of multiple tissues to apoptotic, necrotic, and atrophic stimuli. Genetic inhibition of the cristae remodeling pathway in vivo does not affect development, but protects mice from denervation-induced muscular atrophy, ischemic heart and brain damage, as well as hepatocellular apoptosis. Mechanistically, OPA1-dependent mitochondrial cristae stabilization increases mitochondrial respiratory efficiency and blunts mitochondrial dysfunction, cytochrome c release, and reactive oxygen species production. Our results indicate that the OPA1-dependent cristae remodeling pathway is a fundamental, targetable determinant of tissue damage in vivo.

Graphical abstract

Figure 1

Opa1^tg Mice Are Viable, Fertile, and Grow Normally (A) Schematic representation of targeted transgenesis strategy into the permissive hypoxanthine phosphoribosyltransferase (HPRT) locus. The targeted vector includes a 5′ HPRT homology arm, β-actin promoter, the transgene cDNA OPA1 isoform 1, and 3′ HPRT homology arm including a part of HPRT promoter and a homologous region to the mouse HPRT locus. The targeting construct is linearized before electroporation into BPES cells. Homologous recombination of the 3′ arm reconstitutes a functional HPRT gene, allowing the selection of targeted BPES cells on stringent hypoxanthine-aminopterin-thymidine (HAT) conditions. (B) PCR analysis of genomic DNA from wild-type (WT), heterozygous (Opa1^+/−), and homozygous (Opa1^+/+) Opa1^tg mouse tail. (C) Equal amounts of protein from tissues of the indicated genotypes were separated by SDS-PAGE and immunoblotted with the indicated antibodies. (D) Densitometric analysis of OPA1 protein level in tissue from heart and muscle. Data represent average ± SEM (n = 7 for each group). ^∗p < 0.05 in an unpaired two-sample Student’s t test. (E) Average ± SEM body weight of 5-month-old male (n = 34 for each group) and female (n = 13 for each group) and of 9-month-old male (n = 18 for each group) and female (n = 16 for each group) C57/Bl6 littermates of the indicated genotype. ^∗p < 0.05 in a two-tailed ANOVA test. (F) Kaplan-Meier censorial analysis of C57/Bl6 mice of the indicated genotype (n = 30 male and n = 26 female littermates). See also Figure S1.

Figure 2

Opa1^tg Mice Develop a Non-pathological Cardiac Hypertrophy (A) Representative photographs of hearts dissected from littermates of the indicated genotype and age. Scale bars, 0.5 cm. (B) Average ± SEM heart/body weight ratio in littermates of the indicated genotype and age (n = 8 in each group). ^∗p < 0.05 in a two-tailed ANOVA test versus 5 months Opa1^tg dataset. (C) Representative detail confocal images of left ventricular cryosections stained for dystrophin from 9-month-old mice of the indicated genotype. Scale bar, 25 μm. (D) Experiments were as in (C). Data represent mean ± SEM of four independent experiments. ^∗p < 0.05 in an unpaired two-sample Student’s t test. (E) Echocardiographic long axis view of hearts from 9-month-old littermates of the indicated genotype. LV, left ventricle; A, aorta. (F) Experiments were as in (E). Data represent mean ± SEM percentage of cardiac ejection fraction (EF) and shortening (SF) of four independent experiments. (G) H&E staining of ventricular cryosections from 9-month-old littermates of the indicated genotype. Scale bars, 100 μm. (H) Representative immunofluorescence images of ventricular cryosections stained for atrial natriuretic peptide (ANP) from 9-month-old littermates of the indicated genotype. Scale bar, 100 μm. (I) Data represent average ± SEM of ANP mRNA levels determined by RT-PCR in 4- and 9-month-old littermates of the indicated genotype (n = 4 for each group). (J) Representative immunofluorescence images of ventricular cryosections stained for collagen I from 9-month-old littermates of the indicated genotype. Scale bar, 100 μm. (K) Experiments were as in (J). Data represent mean ± SEM of four independent experiments. See also Figure S2.

Figure 3

Opa1^tg Mice Are Protected from Denervation-Induced Muscle Atrophy (A) H&E staining of control and 10-days-denervated gastrocnemial cryosections from 9-month-old littermates of the indicated genotype. Scale bars, 50 μm. (B) Experiments were as in (A). Muscle loss 10 days after denervation was quantified by cross-sectional area (CSA) measurement of innervated and denervated fibers. Data represent mean ± SEM of five independent experiments. ^∗p < 0.05 in an unpaired two-sample Student’s t test. (C–E) Data represent average ± SEM of the indicated mRNA levels determined by RT-PCR in littermates of the indicated genotype treated as indicated (n = 5 for each group). ^∗p < 0.05 in an unpaired two-sample Student’s t test. (F) Data are average ± SEM of five independent experiments of TMRM fluorescence analysis over mitochondrial regions in fibers isolated from FDB of the indicated genotype treated as indicated. (G) Equal amounts (30 μg) of protein from gastrocnemius isolated from littermates of the indicated genotype treated as indicated were separated by SDS-PAGE and immunoblotted using the indicated antibodies. Den., denervated. See also Figure S3.

Figure 4

Opa1^tg Mice Are Protected from Ischemic Damage (A) Schematic representation of the experimental ischemia-reperfusion protocol used. (B) Data represent average ± SEM of eight independent experiments of lactate dehydrogenase (LDH) release during reperfusion in 5-month-old littermates of the indicated genotype, sex, and background (n = 8 for each group). ^∗p < 0.05 in a two-tailed ANOVA test. (C) Equal amounts (30 μg) of heart protein from littermates of the indicated genotype treated as indicated were separated by SDS-PAGE and immunoblotted using the indicated antibodies. (D) Representative images of TTC-stained whole brains from 3-month-old Sv129 female littermates of the indicated genotype mice 72 hr after middle cerebral artery occlusion (MCAo). (E) Representative TTC-stained 250-μm brain sections from 3-month-old Sv129 female littermates of the indicated genotype 72 hr after MCAo. Slices are aligned from anterior to posterior. (F) Data represent average ± SEM of four independent experiments carried out as in (E). ^∗p < 0.05 in an unpaired two-sample Student’s t test.

Figure 5

Opa1^tg Mice Are Protected from Fas-Induced Apoptotic Liver Damage (A) Representative images of H&E-stained paraffin-embedded liver sections from littermates of the indicated genotype. Where indicated mice were tail vein injected with 0.25 μg/g anti-Fas antibody (Fas), sacrificed after 24 hr, and livers were explanted for histology. Scale bar, 50 μm. (B) Experiments were as in (A), except that paraffin-embedded liver sections were TUNEL stained. Arrows indicate TUNEL-positive hepatocytes. Scale bar, 25 μm. (C) Data represent average ± SEM of four independent experiments carried out as in (B). At least 30 images per group per experiment were analyzed. ^∗p < 0.05 in an unpaired two-sample Student’s t test. (D) Representative confocal images of liver cryosections from littermates of the indicated genotypes treated as indicated and immunostained for cytochrome c (green) and TOM20 (red). The boxed areas are magnified 5×. Scale bar, 20 μm. (E and F) Plasma ALT (E) or AST (F) levels in mice of the indicated genotypes at the indicated time points after tail vein injection of 0.25 μg/g anti-Fas antibody. Data represent average ± SEM (n = 7 in E, and n = 8 in F, for each group). ^∗p < 0.05 in an unpaired ANOVA test. See also Figure S4.

Figure 6

Opa1^tg Mitochondria Are Resistant to Apoptotic Cristae Remodeling and Cytochrome C Release (A) Representative confocal images of mitochondrially targeted YFP transfected in primary myoblasts of the indicated genotype. Scale bar, 20 μm. (B) Data represent average ± SEM of four independent experiments (30 cells/experiment) of mitochondrial major axis length calculation in experiments as in (A). ^∗p < 0.05 in an unpaired two-sample Student’s t test. (C) Representative electron micrographs of hearts of the indicated genotype. Scale bar, 500 nm. (D) Morphometric analysis of cristae width in 80 randomly selected mitochondria of 5-month-old hearts from WT and Opa1^tg mice. ^∗p < 0.05 in an unpaired two-sample Student’s t test. (E) Fold increase in MitoSOX fluorescence after treatment with antimycin A (10 μM, 20 min) in cells incubated for 24 hr in DMEM supplemented with the indicated monosaccharides. Data represent average ± SEM of four independent experiments. ^∗p < 0.05 in an unpaired two-sample Student’s t test. (F) Mitochondria isolated from livers of the indicated genotypes were treated for the indicated times with 40 pmol/mg cBID, and cytochrome c release was measured by ELISA. Data represent average ± SEM of five independent experiments. (G) Ascorbate/TMPD-driven respiration of purified liver mitochondria of the indicated genotypes treated where indicated for 15 min with cBID. Data represent average ± SEM of four independent experiments. ^∗p < 0.05 in an unpaired two-sample Student’s t test. (H) Purified muscle mitochondria of the indicated genotype were treated for the indicated times with cBID, and, where indicated, 10 mM EDC was added. Equal amounts (30 μg) of proteins were separated by SDS-PAGE and immunoblotted using the indicated antibodies. (I) Densitometric analysis of OPA1 oligomers. Experiments were as in (H). Data represent average ± SEM of five independent experiments. ^∗p < 0.05 in a two-way ANOVA with the corresponding time point WT or between the indicated samples. See also Figure S5.