Changes in Mitochondrial Morphology and Bioenergetics in Human Lymphoblastoid Cells With Four Novel OPA1 Mutations

Abstract

Purpose.: Mutations in the optic atrophy 1 gene (OPA1) have been reported in patients with autosomal dominant optic atrophy (ADOA). OPA1 plays important roles in mitochondrial dynamics and cell apoptosis. The link between OPA1 mutations and changes in bioenergetics is still not fully resolved. The aim of this study was to investigate the effects of OPA1 mutations on the mitochondrial tubular network and bioenergetics.

Methods.: We established lymphoblastoid cell lines from four ADOA families harboring different OPA1 mutations, unaffected relatives (internal control cell lines), and unrelated normal controls (normal control cell lines). OPA1 splice variants and mRNA were analyzed by reverse transcription-PCR and quantitative real-time PCR. Protein isoforms were examined by Western blotting. The mitochondrial network was visualized by confocal microscopy. Mitochondrial bioenergetics were assessed using a Seahorse XF24 flux analyzer. Mitochondrial membrane potential and oxidative damage were analyzed by flow cytometry.

Results.: OPA1 mutant cell lines showed significant decreases in OPA1 mRNA and protein expression, mitochondrial membrane potential, and ATP synthesis. A marked deficiency of the long isoform of OPA1 was observed in cells with OPA1 mutations in the middle domain and GTPase effector domain. Confocal microscopy revealed increased mitochondrial fragmentation in OPA1 mutant cells. OPA1 mutant cells also displayed reduced oxygen consumption and underwent glycolysis to produce ATP. Moreover, OPA1 mutations caused the accumulation of oxidative damage.

Conclusions.: Our experiments demonstrated that OPA1 mutations induced mitochondrial fragmentation, uncoupled mitochondrial respiration, and elicited dysfunctional bioenergetics. However, there were no significant differences among the various OPA1 mutations.

Autosomal dominant optic atrophy (ADOA) is characterized by an insidious loss of visual acuity,¹ which begins in early childhood with a central or cecocentral visual field defect, color vision deficits, and temporal or diffuse pallor of the optic disc.^2,3 Histopathological studies have shown that ADOA involves the primary degeneration of retinal ganglion cells, which is accompanied by ascending optic atrophy.^4,5 

The optic atrophy 1 (OPA1) gene was found to be the cause of ADOA.^6,7 OPA1 is a dynamin-like GTPase protein required for mitochondrial fusion,⁸ the structure of the cristae membrane, maintenance of inner membrane integrity, and regulation of cell apoptosis.⁹ OPA1 overexpression in mouse embryo fibroblasts promotes mitochondrial tubulation.¹⁰ OPA1 down-regulation leads to mitochondrial fragmentation and cell apoptosis.¹¹ 

The OPA1 gene (Online Mendelian Inheritance in Man [OMIM] identifier 605290) has 31 exons that encode a 960-amino acid protein.^7,12 The OPA1 protein includes a highly basic domain, a transmembrane domain, a GTPase domain, a middle domain, and a GTPase effector (GED) domain. OPA1 is ubiquitously expressed in all human tissues, as determined by Northern blotting.⁶ The highest transcript level is observed in the retina, followed by the brain, testis, heart, and skeletal muscle.⁷ Several layers of the neural retina and the optic nerve express OPA1 protein,^13,14 which affects retinal ganglion cell synaptic architecture and neural cell connectivity.¹⁵ 

OPA1 undergoes complex posttranscriptional regulation and posttranslational proteolysis. OPA1 is regulated by proteolytic cleavage, which degrades long OPA1 isoforms into short isoforms. An apoptotic insult or dissipation of the ΔΨm induces cleavage of the long OPA1 isoform into short isoforms and triggers mitochondrial fragmentation.^16–18 Recovery of the mitochondrial fusion and ΔΨm was demonstrated through restoration of the long OPA1 isoforms.^19–21 

More than 300 unique DNA variants of OPA1 from ADOA patients are reported in the MITOchondrial DYNamics database (formerly eOPA1; http://lbbma.univ-angers.fr/eOPA1).⁸ The effects of OPA1 mutations on mitochondrial oxidative phosphorylation (OXPHOS) have been inconsistent.^22–25 Most of the studies have been performed in yeast or cells manipulated with knockdown technology or in OPA1-deficient animal models.^26,27 The effects of OPA1 mutations on human OXPHOS are debatable.^22–25 Spinazzi et al.²³ reported that OPA1 mutations do not cause mitochondrial functional defects in fibroblasts from an ADOA patient. However, Zanna et al.²⁴ observed that OXPHOS was significantly affected in ADOA fibroblasts. These conflicting results may be attributed to cells with different OPA1 mutations. We hypothesized that different OPA1 mutations might have different effects on mitochondrial morphology and bioenergetics. Here, we analyzed lymphoblastoid cell lines with four different OPA1 mutations to determine how OPA1 affects mitochondrial morphology and bioenergetics. 

Materials and Methods

Patients

Eight ADOA patients with OPA1 mutations (six males and two females) and eight unaffected relatives with wild-type OPA1 (internal controls) from four unrelated pedigrees were included (Fig. 1). Pedigree A had a c.1065+2T>C mutation in the OPA1 gene at the splice site of intron 10 with skipping of exon 10. Pedigree C had a c.1212+2insT mutation in the OPA1 gene at the splice site of intron 12 with skipping of exon 12. Pedigree D had a c.1776_1778delACT mutation in exon 19 of the OPA1 gene, which generated deletion of the amino acid Leu593. Pedigree E had a c.2846 T>C mutation in exon 28 of the OPA1 gene, resulting in an amino acid change Leu949Pro. The pedigrees were the same as in a previous report.²⁸ Two unrelated normal controls were also included. This study was performed according to the tenets of the Declaration of Helsinki for research involving human subjects. The protocol was approved by the Institutional Review Board/Ethics Committee of Taipei Veteran General Hospital. Because there were limitations on the use of retinal ganglion cells from the ADOA patients, we established lymphoblastoid cell lines from four ADOA families. After obtaining informed consent, we collected 15 mL whole blood in an EDTA-containing tube. Epstein-Barr virus-transformed lymphoblastoid cell lines were generated from all participants at the Food Industry Research and Development Institute (Hsinchu, Taiwan). 

Figure 1

 

Pedigrees of the four studied families with autosomal dominant optic atrophy (ADOA). (A) Schematic of the OPA1 gene and the four mutations analyzed in this study. Four mutations are considered disease-causing mutations: two at the splice site in the GTPase domain, one deletion in the middle domain, and one missense mutation in the GTPase effector domain (GED). (B) Families A, C, and E showed a pattern of inheritance compatible with autosomal dominant inheritance. Family D is a sporadic case with a de novo mutation. Eight ADOA patients affected with OPA1 mutations and eight unaffected relatives without OPA1 mutation (internal controls) from four unrelated pedigrees were included. #, cells from ADOA patients AI-1 and AII-2 were difficult to culture for most of the assays.

Chemicals and Cell Culture

Unless otherwise specified, chemicals and media were purchased from Sigma-Aldrich Corp. (St. Louis, MO, USA). Cell lines were maintained in RPMI 1640 medium supplemented with 15% fetal calf serum (FCS; Gibco, Grand Island, NY, USA); 24 hours before the assays, medium was changed to RPMI 1640 supplemented with 2% FCS. Cells from ADOA patients AI-1 and AII-2 were notably difficult to culture for use in most of the assays (see Supplementary Fig. S1). 

Reverse Transcription-Polymerase Chain Reaction (RT-PCR) for OPA1 Splice Variants

Total RNA was extracted with an RNeasy mini-kit (Qiagen, Valencia, CA, USA). First-strand cDNA synthesis was performed with 5 U Moloney Murine leukemia virus reverse transcriptase (Epicentre, Madison, WI, USA), 1 μg RNA, and 50 pmol primers (Promega, Madison, WI, USA). OPA1 was amplified using primers located in exon 3 (OPA1S: 5′-GGATTGTGCCTGACATTGTG-3′) and exon 9 (OPA1AS: 5′-CACTCAGAGTCACCTTAACTGG-3′). ACTB expression was amplified using the forward primer 5′-CCAACCGCGAGAAGATGA-3′ and the reverse primer 5′-CCAGAGGCGTACAGGGATAG-3′. 

Quantitative Real-Time PCR (qPCR) for OPA1 mRNA Quantification

Quantitative PCR was performed with an OPA1S/OPA1AS primer pair that covered the common region of exons 21 and 22, which was present in all eight splice variants. Quantitative PCR was performed using a LightCycler FastStart DNA Master SYBR Green kit and a LightCycler apparatus (Roche Diagnostics GmbH, Manheim, Germany). The threshold cycle numbers for the ACTB and OPA1 genes were determined. The following primers were used: OPA1qS 5′-TGTGATTGAAAACATCTACCTTCCA-3′ and OPA1qAS 5′-TTTAAGCTTGATATCCACTGTGGTGT-3′. 

Western Blot Analysis

Protein samples were subjected to 8% SDS-PAGE and then transferred to a polyvinylidene fluoride membrane (GE Healthcare Bioscience, Fribourg, Switzerland). Immunoblotting was performed with an anti-OPA1 (BD Biosciences, San Jose, CA, USA) or anti–β-actin (Santa Cruz Biotechnology, Dallas, TX, USA) primary antibody, a horseradish peroxidase-conjugated anti-mouse immunoglobulin G (IgG) secondary antibody (Cell Signaling Technologies, Beverly, MA, USA), and enhanced chemiluminescence detection (ECL; GE Healthcare Bioscience). 

Mitochondrial Network Analysis

Dual fluorescent staining was performed to detect mitochondrial heat shock protein 60 (mtHSP60; Santa Cruz Biotechnology) and nuclei, using 4′,6-diamidino-2-phenylindole (DAPI; Molecular Probes, Eugene, OR, USA). Cells were cultured on IBIDI Micro-Slides (Applied BioPhysics, Troy, NY, USA), fixed in 4% paraformaldehyde for 15 minutes, permeabilized with 0.1% Triton X-100, blocked with bovine serum albumin (BSA) for 1 hour, and immunostained with an anti-mtHSP60 antibody overnight at 4°C. The cells were then incubated with fluorescein isothiocyanate (FITC)-conjugated AffiniPure goat anti-mouse IgG (Jackson ImmunoResearch, West Grove, PA, USA), stained with DAPI for 3 minutes, and mounted with Fluoromount (Sigma-Aldrich Corp.). Cells were examined using a confocal fluorescence microscope (model TCS-SP5; Leica Microsystems CMS GmbH, Mannheim, Germany) with three lasers (argon, 488 nm; diode, 405 nm; Leica Microsystems). Mitochondrial morphology was determined in 100 individual cells and scored based on the filamentous and fragmented forms in each of five different experiments. 

Flow Cytometry Analysis of Mitochondrial Membrane Potential (ΔΨm)

JC-1 dye forms a green fluorescent monomer with emission at 530 nm at lower ΔΨm but forms red fluorescent aggregates with emission at 590 nm at higher ΔΨm. All analyses were performed using FACScan (BD Biosciences) equipped with a 100 mW argon laser. Cells were incubated for 10 minutes with 5 μM JC-1 (Molecular Probes) at 37°C. Data were acquired in list mode, and the relative proportions of cells within different areas were quantified using LYSYS II software (BD Biosciences). 

ATP Assay

Cellular ATP levels were determined by luciferin- and luciferase-based assays. Cells were lysed with ATP-releasing buffer and quantified using an ATP assay kit (Molecular Probes). The supernatant was analyzed on a Wallac Victor 1420 multilabel counter (Perkin Elmer, Waltham, MA, USA). ATP levels were calculated as nanomoles of ATP per milligram of protein and were normalized to the ATP levels in the control cells. 

Seahorse XF24 Metabolic Flux Analysis

Intact cellular respiration was detected using the Seahorse XF24 Metabolic Flux Assay (Seahorse Bioscience, Chicopee, MA, USA). Cells were cultured in XF24-well microplates coated with CELL-TAK (BD Biosciences). Baseline measurements were recorded before the addition of 3 μM oligomycin, 1 μM carbonyl cyanide m-chlorophenylhydrazone (CCCP), and 5 μM rotenone. The oxygen consumption rate (OCR), extracellular acidification rate (ECAR; an indicator of lactic acid production or glycolysis), spare respiratory capacity, and proton leakage were automatically calculated and recorded using Seahorse XF24 software. The percent change compared to the basal rate was calculated as the change in value divided by the average value of the baseline readings. 

Flow Cytometry Analysis of Reactive Oxygen Species (ROS) and Oxidative Damage

Aliquots of 1 × 10⁶ cells were gently stained with fluorescent probes in the dark for 15 minutes at room temperature. All analyses were performed using a FACScan unit (BD Biosciences). Intracellular ROS were stained with CM-H₂DCFDA (Molecular Probes) and were measured using an excitation of 490 or 500 nm and an emission of 525 nm. Lipid peroxidation products were stained with C11-BODIPY^581/591 (Molecular Probes). The fluorescence excitation/emission of C11-BODIPY^581/591 shifted from 581/591 nm to 490/510 nm after oxidation. Mitochondrial superoxide was measured using MitoSox Red (Molecular Probes) with an excitation of 510 nm and an emission of 580 nm. To measure 8-hydroxy-2′-deoxyguanosine (8-OHdG), cells were fixed with 4% paraformaldehyde, incubated for 1 hour with 2% BSA, and incubated for 30 minutes with an anti-8-OHdG mouse monoclonal antibody (Genesis Biotech, Taipei, Taiwan). The cells were subsequently treated with FITC-conjugated goat anti-mouse IgG (ImmunoResearch), which was measured at 488 nm excitation and 525 nm emission. The levels of 8-OHdG were quantified by flow cytometry, and the accumulation of 8-OHdG was visualized by confocal imaging. 

Statistics

All results are presented as means ± standard deviations (SD). Statistical significance was determined by t-test analysis. A P value of <0.05 was considered statistically significant. 

Results

Reduced OPA1 mRNA and Protein Expression in OPA1 Mutant Cells

At least eight mRNA splice variants and five distinct proteolytically processed isoforms have been identified.¹² A schema of the OPA1 splice variants is presented in Figure 2A. Our OPA1 transcript analysis with the OPA1S/OPA1AS primers revealed six OPA1 splice variants that migrated as a complex mixture during gel electrophoresis, including products of 762 bp, 708 bp, 651 and/or 654 bp, 597 and/or 600 bp, 543 bp, and 489 bp (Fig. 2B). The pattern of OPA1 splice variants in AII-1, CI-1, and CII-1 cells looked similar to those in normal and internal controls. Deficiencies in the long variants of OPA1 were found in DII-3, EI-1, and EII-1 cells (Fig. 2B). The OPA1qS/OPA1qAS primer pair was used to generate an 82-bp PCR product. OPA1 mRNA levels were reduced in all OPA1 mutant cells (Fig. 2C). Using qPCR, we observed a significant reduction in OPA1 mRNA in all OPA1 mutant cells, especially in those from DII-3 and EII-1 (Fig. 2D). Western blotting detected five OPA1 isoforms, including two higher-molecular-weight long isoforms (L1 and L2) and three lower-molecular-weight short isoforms (S3, S4, and S5), ranging from approximately 86 to 92 kDa. All the OPA1 isoforms were down-regulated in all the OPA1 mutant cells. In DII-3 and EII-1, only the short isoforms were found (Fig. 2E). 

Figure 2

 

Differential expression of OPA1 mRNA and OPA1 isoforms. (A) Schema of the eight splice variants of OPA1 generated by alternative splicing involving exons 4, 4b, and 5b. (B) Amplification using the OPA1S/OPA1AS primers resulted in six different fragments of 762, 708, 651/654, 597/600, 543, and 489 bp. Global decreases in all splice variants of OPA1 were observed in all the OPA1 mutant cells. Deficiencies in the long splice variants of OPA1 were found in DII-3, EI-1, and EII-1. (C) The OPA1qS/OPA1qAS primer pair was used to generate an 82-bp PCR product. Reduced OPA1 mRNA levels were observed in all OPA1 mutant cells. (D) A significant reduction in OPA1 mRNA expression was found in all OPA1 mutant cells by qPCR using the OPA1qS/OPA1qAS primers. (E) Western blotting for OPA1 revealed five distinct OPA1 isoforms. In all OPA1 mutant cells, expression of all OPA1 isoforms was reduced. In DII-3 and EII-1, only the short isoforms were found. β-Actin served as the loading control. Data are means ± SD from three independent experiments. *P < 0.05, **P < 0.01 compared to the normal control; #P < 0.05, ##P < 0.01 compared to the internal controls in the same pedigree.

OPA1 Mutations Induced Mitochondrial Fragmentation

Representative images of the mitochondrial network are shown in Fig. 3A. Normal control and internal control cells (Fig. 3, AI-2, CI-2, DI-2, DII-2, and EI-2 [white letters]) exhibited a balanced mitochondrial network between the filamentous and fragmented states. The mitochondria displayed a distinct punctate pattern in all OPA1 mutant cells (Fig. 3, AI-1, AII-1, CI-1, CII-1, DII-3, EI-1, and EII-1 [red letters]). In all OPA1 mutant cells, the filamentous mitochondrial network was less than 2% (Fig. 3B). The proportions making up the filamentous, intermediate, and fragmented networks were 37%, 44%, and 19%, respectively, in the normal control cells and 1%, 22%, and 77%, respectively, in EII-1 cells. 

Figure 3

 

Effect of OPA1 mutations on the mitochondrial network. (A) Mitochondrial morphology was visualized by immunofluorescent labeling with an antimitochondrial HSP60 antibody conjugated to FITC (green) and was analyzed by confocal microscopy. DAPI-stained nuclei were identified by their blue fluorescence. Normal control and internal control cells (AI-2, CI-2, DI-2, DII-2, and EI-2 [white]) exhibited a balanced mitochondrial network of the filamentous and fragmented states. Mitochondria in cells carrying OPA1 mutations (AI-1, AII-1, CI-1, CII-1, DII-3, EI-1, and EII-1 [red]) displayed a distinct punctate pattern. (B) A balanced mitochondrial network was visualized in normal cells and internal controls, but the presence of the fragmented network was increased in OPA1 mutant cells. Data are means ± SD from five independent experiments. **P < 0.01 compared to the normal control; ##P < 0.01 compared to the internal controls of the same pedigree.

OPA1 Mutations Lowered the ΔΨm and Caused ATP Dissipation

Cells with a lower ΔΨm exhibited more monomers than aggregates with the JC-1 stain. The JC-1 aggregate-to-monomer ratio was significantly decreased in all OPA1 mutant cells compared to the normal and internal controls (Fig. 4A, Supplementary Fig. S2). The ATP content was significantly decreased in all the OPA1 mutant cells relative to that in the normal and internal controls (Fig. 4B). The ATP content relative to that in the normal control was 56% in AII-1, 74% in CI-1, 58% in CII-1, 63% in DII-3, 63% in EI-1, and 58% in EII-1 cells (Fig. 4B). 

Figure 4

 

OPA1 mutations induced mitochondrial dysfunction. (A) The JC-1 aggregate-to-monomer ratio was significantly decreased in all OPA1 mutant cells compared to that in normal and internal controls. (B) Adenosine triphosphatase content was significantly decreased in all OPA1 mutant cells compared to that in normal and internal controls. Data are mean ± SD from three independent experiments. *P < 0.05, **P < 0.01 compared to the normal control; #P < 0.05, ##P < 0.01 compared to the internal controls of the same pedigree.

OPA1 Was Involved in the Maintenance of Mitochondrial Bioenergetics and Coupled Respiration

Data in Figure 5A represent the time course of the OCR and ECAR measurements under the basal condition, followed by the sequential addition of oligomycin, CCCP, and rotenone. The OPA1 mutant cells showed significant decreases in OCR, maximum respiration, and spare respiratory capacity and significant increases in ECAR and proton leakage rate (Figs. 5A, 5B). The OCR-to-ECAR proportion was 89%:11% in the normal control. Decreased OCR and increased ECAR were observed in AII-1, DII-3, EI-1, and EII-1 cells. The lowest ATP-related OCR occurred in AII-1 cells. The rate of proton leakage was significantly increased in all OPA1 mutant cells (Fig. 5B). The proton leakage rate was 11.0% ± 3.0% in the normal control and was increased 5.5-fold in DII-3 and 3.4-fold in EI-1. 

Figure 5

 

OPA1 mutation impaired mitochondrial bioenergetics and uncoupled mitochondrial respiration. (A) Mitochondrial respiration and glycolysis as indicators of cellular bioenergetics were assessed using a Seahorse XF24 metabolic flux analyzer. Data are time courses of the OCR and ECAR under basal conditions, followed by the sequential addition of oligomycin (1 μg/mL), CCCP (1 μM), and rotenone (10 μM). O, oligomycin; C, CCCP; R, rotenone. (B) OPA1 mutant cells exhibited significant decreases in OCR, maximal respiration, and spare respiratory capacity and significant increases in ECAR and proton leakage rate. Data are means ± SD from three independent experiments. *P < 0.05, **P < 0.01 compared to the normal control; #P < 0.05, ##P < 0.01 compared to the internal controls of the same pedigree.

OPA1 Mutations Elicited Accumulation of Oxidative Damage

High levels of ROS and oxidative damage were found in all OPA1 mutant cells (Figs. 6A–D). The production of hydrogen peroxide was 1.7- to 2.1-fold increased in AII-1, CII-1, DII-3, and EII-1 cells compared to that in the normal controls (Fig. 6A). The production of mitochondrial superoxide was 3.4- to 4.2-fold increased in the OPA1 mutant cells (Fig. 6B). A significant increase in lipid peroxide was detected in all OPA1 mutant cells, especially in DII-3, EI-1, and EII-1. Compared to the normal control, the lipid peroxide level was 7.9-fold increased in DII-3, 7.2-fold in EI-1, and 4.6-fold in EII-1 (Fig. 6C). An increase of approximately 3.2-fold in 8-OHdG was observed in DII-3 cells (Fig. 6D, Supplementary Fig. S3). Furthermore, enhanced 8-OHdG immunopositivity was visualized in the mitochondria in OPA1 mutant cells (Fig. 6E). 

Figure 6

 

OPA1 mutations induced ROS generation and oxidative insults. Four types of immunofluorescent stains were detected, and flow cytometry was performed: (A) CM-H2DCFDA, hydrogen peroxide; (B) MitoSOX Red, mitochondrial superoxide; (C) BODIPY^581/591, lipid peroxides; and (D) FITC-conjugated anti-8-OHdG antibody, 8-OHdG. High levels of ROS and oxidative damage were present in OPA1 mutant cells. (E) Accumulation and localization of 8-OHdG were visualized by confocal microscopy. Strong green fluorescent signals were observed in the mitochondria of OPA1 mutant cells. Data are means ± SD from three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 compared to the normal control; #P < 0.05, ##P < 0.01, ###P < 0.01 compared to the internal controls of the same pedigree.

Discussion

The effects of OPA1 mutations on mitochondrial morphology and bioenergetics have been shown previously in mice and humans. Here, we investigated four human OPA1 mutations. In this study, we used lymphoblastoid cell lines instead of fibroblasts, which have been used in previous human studies.²⁴ 

The present study showed that OPA1 mutant cells have decreased expression of OPA1 mRNA and protein. A global depletion in the expression of OPA1 protein isoforms was observed in all OPA1 mutant cells. A distinct deficiency in the long isoform of OPA1 was observed in DII-3 and EII-1 cells. Whether the long OPA1 isoform deficiency was due to the increased proteolysis of the long isoforms or to the structural instability of the long isoforms in DII-3 and EII-1 requires further clarification. 

Mitochondria continuously merge and divide to share solutes, metabolites, and proteins.^27,29 Our study showed that the control cells had a balanced mitochondrial network, but the OPA1 mutant cells had a more fragmented network. Studies have shown that mitochondria expressing only the long OPA1 isoforms or only the short OPA1 isoforms have little fusion activity, but there is substantial fusion activity when the long and short OPA1 isoforms are coexpressed.^19,21 Haploinsufficiency of OPA1 also decreases fusion activity. Overexpression of OPA1 restores and increases mitochondrial fusion.^10,19 The loss of fusion activity and the increase in mitochondrial network fragmentation in OPA1 mutant cells may be partly due to long isoform deficiency and partly due to haploinsufficiency. 

Yu-Wai-Man et al.³⁰ reported that OXPHOS is deficient in the skeletal muscle of ADOA patients. Van Bergen et al.³¹ showed that ATP synthesis is affected in ADOA patients with poor vision. Agier et al.²⁵ reported that respiratory complex IV activity is decreased in ADOA fibroblasts. OPA1 is thought to modulate OXPHOS by interacting with the mitochondrial respiratory chain complexes and by facilitating mitochondrial coupling efficiency.²⁴ Our study showed that OPA1 mutant cells had a significant reduction in ATP production and OCR. However, Mayorov et al.³² found no significant effects of OPA1 mutation on OXPHOS in lymphoblastoid cell lines from ADOA patients. These conflicting results may be related to different culture conditions. We used 15% FCS for cell culture before switching to 2% FCS. There was decreased OXPHOS capacity in the OPA1 mutant cells, but the difference was not significant (data not shown). However, using 2% FCS to trigger metabolic stress to evaluate OPA1 mutant cell growth revealed a significant reduction in ATP production and OCR in OPA1 mutant cells. These data indicate that only under sub-optimal or “stress” conditions does this phenotype become important. 

The bioenergetic profiling of OPA1 mutant cells showed a significant decrease in OCR and increase in ECAR. These results indicated that OPA1 mutant cells preferred glycolysis rather than OXPHOS for cell metabolism. In addition, the spare respiratory capacity was significantly reduced in the OPA1 mutant cells. The spare respiratory capacity includes extra ATP that is produced by OXPHOS in the event of a sudden energy demand and is an indicator of the cell stress response.³³ The depletion of spare respiratory capacity has been related to some pathologies that affect high energy-requiring tissues and result in cell senescence or organ failure.^33,34 We also found a significant increase in the proton leak rate in OPA1 mutant cells. Mitochondria couple respiration to ATP synthesis through an electrochemical proton gradient. Protons leaking across the inner membrane adjust the coupling efficiency.³⁵ Functional mitochondrial proton leakage is an important component of cellular metabolism, including thermogenesis. This leakage protects against ROS and regulates the calcium flux.³⁶ A high proton leakage was observed that caused decreased mitochondrial bioenergetics and thereby decreased ATP synthesis. 

Some studies have elucidated the mechanisms by which OPA1 mutations induce mitochondrial dysfunction.^9,24,37 Zanna et al.²⁴ observed a link between OPA1 and OXPHOS capacity through a direct interaction between OPA1 and subunits of complexes I, II, and III, as well as apoptosis-inducing factor. In addition, OPA1 regulates cristae shape⁹ and modulates the assembly of respiratory chain supercomplexes (RCS) and mitochondrial respiratory efficiency.³⁷ Stabilization of OPA1 could correct RCS and mitochondrial dysfunction.³⁷ 

Our study found a significant increase in ROS generation and oxidative damage (hydrogen peroxide and 8-OHdG) in all OPA1 mutant cells. Excessive generation of ROS can cause impaired mitochondrial OXPHOS, leading to further oxidative modification of cell components and mitochondrial genomic instability.³⁸ 

In conclusion, our study of lymphoblastoid cells with all four pathogenic OPA1 mutations exhibited haploinsufficiency. A marked deficiency in the long isoform of OPA1 was observed in the cells with OPA1 mutations in the middle domain (c.1776_1778delACT) and GTPase effector domain (c.2846T>C). Our experiments demonstrated that under sub-optimal or metabolic stress conditions, OPA1 mutations induced mitochondrial fragmentation, uncoupled mitochondrial respiration, and elicited dysfunctional bioenergetics. However, there were no significant differences among the different OPA1 mutations. 

Acknowledgments

Supported by Taipei Veterans General Hospital Grant VGH 97C1-077 and National Science Council of Taiwan Grant NSC 97-2314-B-075-021-MY3 (MYY). 

Disclosure: S.-H. Kao, None; M.-Y. Yen, None; A.-G. Wang, None; Y.-L. Yeh, None; A.-L. Lin, None