Loss of OPA1 Perturbates the Mitochondrial Inner Membrane Structure and Integrity, Leading to Cytochrome c Release and Apoptosis

OPA1 encodes a large GTPase related to dynamins, anchored to the mitochondrial cristae inner membrane, facing the intermembrane space. OPA1haplo-insufficiency is responsible for the most common form of autosomal dominant optic atrophy (ADOA, MIM165500), a neuropathy resulting from degeneration of the retinal ganglion cells and optic nerve atrophy. Here we show that down-regulation of OPA1 in HeLa cells using specific small interfering RNA (siRNA) leads to fragmentation of the mitochondrial network concomitantly to the dissipation of the mitochondrial membrane potential and to a drastic disorganization of the cristae. These events are followed by cytochromec release and caspase-dependent apoptotic nuclear events. Similarly, in NIH-OVCAR-3 cells, the OPA1 siRNA induces mitochondrial fragmentation and apoptosis, the latter being inhibited by Bcl2 overexpression. These results suggest that OPA1 is a major organizer of the mitochondrial inner membrane from which the maintenance of the cristae integrity depends. As loss of OPA1 commits cells to apoptosis without any other stimulus, we propose that OPA1 is involved in the cytochrome c sequestration and might be a target for mitochondrial apoptotic effectors. Our results also suggest that abnormal apoptosis is a possible pathophysiological process leading to the retinal ganglion cells degeneration in ADOA patients. OPA1 encodes a large GTPase related to dynamins, anchored to the mitochondrial cristae inner membrane, facing the intermembrane space. OPA1haplo-insufficiency is responsible for the most common form of autosomal dominant optic atrophy (ADOA, MIM165500), a neuropathy resulting from degeneration of the retinal ganglion cells and optic nerve atrophy. Here we show that down-regulation of OPA1 in HeLa cells using specific small interfering RNA (siRNA) leads to fragmentation of the mitochondrial network concomitantly to the dissipation of the mitochondrial membrane potential and to a drastic disorganization of the cristae. These events are followed by cytochromec release and caspase-dependent apoptotic nuclear events. Similarly, in NIH-OVCAR-3 cells, the OPA1 siRNA induces mitochondrial fragmentation and apoptosis, the latter being inhibited by Bcl2 overexpression. These results suggest that OPA1 is a major organizer of the mitochondrial inner membrane from which the maintenance of the cristae integrity depends. As loss of OPA1 commits cells to apoptosis without any other stimulus, we propose that OPA1 is involved in the cytochrome c sequestration and might be a target for mitochondrial apoptotic effectors. Our results also suggest that abnormal apoptosis is a possible pathophysiological process leading to the retinal ganglion cells degeneration in ADOA patients. outer mitochondrial membrane inner mitochondrial membrane intermembrane space cytochrome c horseradish peroxidase z-Val-Ala-Asp-fluoromethylketone cis-diamine dichloroplatinum phosphate-buffered saline poly(ADP-ribose) polymerase fluorescence-activated cell sorter 4′,6-diamidino-2-phenylindole tetramethyl rhodamine ethyl ester dinitrophenol small interfering RNA Large GTPases from the dynamin family are involved in various processes related to membrane dynamic (6Danino D. Hinshaw J.E. Curr. Opin. Cell Biol. 2001; 13: 454-460Google Scholar, 7van der Bliek A.M. Trends Cell Biol. 1999; 9: 96-102Google Scholar). Two dynamin subfamilies are specifically involved in the mitochondrial network morphology. One comprises the Dnm1/Drp1 protein that is located on the cytoplasmic side of the outer mitochondrial membrane (OM)1 to control the network fission process (8Bleazard W. McCaffery J.M. King E.J. Bale S. Mozdy A. Tieu Q. Nunnari J. Shaw J.M. Nat. Cell Biol. 1999; 1: 298-304Google Scholar, 9Smirnova E. Shurland D.L. Ryazantsev S.N. van der Bliek A.M. J. Cell Biol. 1998; 143: 351-358Google Scholar). The second one includes the intra-mitochondrial Mgm1p/Msp1/OPA1 protein, which is required for the maintenance of the mitochondrial structure and genome in yeasts (10Jones B.A. Fangman W.L. Genes Dev. 1992; 6: 380-389Google Scholar, 11Pelloquin L. Belenguer P. Menon Y. Gas N. Ducommun B. J. Cell Sci. 1999; 112: 4151-4161Google Scholar, 12Shepard K.A. Yaffe M.P. J. Cell Biol. 1999; 144: 711-720Google Scholar). Localization of this dynamin in the different species remains controversial and so does its precise function on the mitochondrial membranes. Mgm1 has been first localized anchored to the OM facing the cytoplasm (12Shepard K.A. Yaffe M.P. J. Cell Biol. 1999; 144: 711-720Google Scholar), then in the intermembrane space (IMS) associated to the inner membrane (IM) (13Wong E.D. Wagner J.A. Gorsich S.W. McCaffery J.M. Shaw J.M. Nunnari J. J. Cell Biol. 2000; 151: 341-352Google Scholar). Msp1 has been located in the matrix anchored to the IM (11Pelloquin L. Belenguer P. Menon Y. Gas N. Ducommun B. J. Cell Sci. 1999; 112: 4151-4161Google Scholar). Finally OPA1 has been recently located to the IMS anchored on the cristae IM (1Olichon A. Emorine L.J. Descoins E. Pelloquin L. Brichese L. Gas N. Guillou E. Delettre C. Valette A. Hamel C.P. Ducommun B. Lenaers G. Belenguer P. FEBS Lett. 2002; 523: 171-176Google Scholar). We have previously shown that mutations in the OPA1 gene induce the most common form of autosomal dominant optic atrophy (2Alexander C. Votruba M. Pesch U.E. Thiselton D.L. Mayer S. Moore A. Rodriguez M. Kellner U. Leo-Kottler B. Auburger G. Bhattacharya S.S. Wissinger B. Nat. Genet. 2000; 26: 211-215Google Scholar, 3Delettre C. Lenaers G. Griffoin J.M. Gigarel N. Lorenzo C. Belenguer P. Pelloquin L. Grosgeorge J. Turc-Carel C. Perret E. Astarie- Dequeker C. Lasquellec L. Arnaud B. Ducommun B. Kaplan J. Hamel C.P. Nat. Genet. 2000; 26: 207-210Google Scholar) (ADOA, MIM165500) and affect the structure of the mitochondrial network in monocytes from ADOA patients (3Delettre C. Lenaers G. Griffoin J.M. Gigarel N. Lorenzo C. Belenguer P. Pelloquin L. Grosgeorge J. Turc-Carel C. Perret E. Astarie- Dequeker C. Lasquellec L. Arnaud B. Ducommun B. Kaplan J. Hamel C.P. Nat. Genet. 2000; 26: 207-210Google Scholar). The possible function of OPA1 as a mechano-enzyme (14Stowell M.H. Marks B. Wigge P. McMahon H.T. Nat. Cell Biol. 1999; 1: 27-32Google Scholar) or as a molecular switch (15Sever S. Damke H. Schmid S.L. Traffic. 2000; 1: 385-392Google Scholar) acting on the IM suggests that it might be involved in the structure and dynamic of the cristae (16Mannella C.A. Pfeiffer D.R. Bradshaw P.C. Moraru I.I. Slepchenko B. Loew L.M. Hsieh C.E. Buttle K. Marko M. Dig. Liver Dis. 2001; 33: 795-802Google Scholar) and therefore on mitochondrial activity. In metazoan vertebrates, cristae structures have a crucial role as they define intra-cristae volumes that segregate the majority of the cytochrome c (cyt c) from the IMS (17Bernardi P. Azzone G.F. J. Biol. Chem. 1981; 256: 7187-7192Google Scholar). In vitro, upon addition of the pro-apoptotic factor tBid on purified mitochondria, this pool of cyt c is released in the IMS by a major cristae remodeling process, which allows in vivo the extensive release of the cyt c in the cytoplasm to trigger apoptosis (18Scorrano L. Ashiya M. Buttle K. Weiler S. Oakes S.A. Mannella C.A. Korsmeyer S.J. Dev. Cell. 2002; 2: 55-67Google Scholar). To address OPA1 function on the mitochondrial membranes and gain insight into the pathophysiological mechanism of ADOA, we have analyzed the OPA1 knock-down in culture cells and found that it disrupts the mitochondrial IM structure and functional integrity and leads to cyt c release and apoptosis. OPA1 antibodies have been described previously (1Olichon A. Emorine L.J. Descoins E. Pelloquin L. Brichese L. Gas N. Guillou E. Delettre C. Valette A. Hamel C.P. Ducommun B. Lenaers G. Belenguer P. FEBS Lett. 2002; 523: 171-176Google Scholar). Commercial antibodies were from the sources indicated: anti-HSP60 (LK2, Sigma), anti-Actin (Chemicon), anti-cytochrome c(Santa Cruz), anti-cleaved PARP (Promega), Alexa-594 anti-rabbit IgG and Alexa-488 anti-mouse IgG (Molecular Probes), and anti-rabbit IgG-HRP and anti-mouse IgG-HRP (New England Biolabs). HeLa cells were cultured in DMEM, 10% fetal calf serum, 5% CO2. NIH-OVCAR-3 cells (19Lafon C. Mathieu C. Guerrin M. Pierre O. Vidal S. Valette A. Cell Growth Differ. 1996; 7: 1095-1104Google Scholar) were cultured in RPMI, 5% fetal calf serum, 5% CO2. siRNA experiments were carried out as described previously (20Elbashir S.M. Harborth J. Weber K. Tuschl T. Methods. 2002; 26: 199-213Google Scholar). The sequence of the region targeted by the OPA1-siRNA (Dharmacond Research), 5′-AAGTTATCAGTCTGAGCCAGGTT-3′, corresponds to nucleotides 1810–1833 of OPA1 open reading frame (GenBankTM accession numberAB011139) and is present in all OPA1 alternate splicing variants (21Delettre C. Griffoin J.M. Kaplan J. Dollfus H. Lorenz B. Faivre L. Lenaers G. Belenguer P. Hamel C.P. Hum. Genet. 2001; 109: 584-591Google Scholar). Transfections of HeLa cells were performed with Oligofectamine® reagent (Invitrogen). Final concentration of the siRNA duplex in culture medium was 100 nm. Control experiments included treatments of cells with 100 μm z-VAD-fmk (Calbiochem) for 48 h or 5 μm CDDP (Sigma) for 24 h. To detect OPA1 and actin by Western blot, transfected cells were trypsinized, washed once in ice-cold PBS, and harvested. Equal amounts of cells were solubilized in 50 μl of Laemmeli sample buffer and boiled for 10 min. Samples were run on 8% polyacrylamide gels and transferred to nitrocellulose. Immunodetection (anti-OPA1: 1/300; anti-actin: 1/10000; IgG-HRP: 1/10,000) was carried out using ECL (Amersham Biosciences). PARP detection was performed according to the protocol from Promega (anti-PARP: 1/1000). Cells grown on glass coverslips were fixed in PBS, 3.7% paraformaldehyde (30 min, 4 °C), permeated in 100% methanol (1 min, −20 °C), then in PBS, 0.2% Triton X-100 (10 min, room temperature) and immunolabeled in PBS, 2% bovine serum albumin, using the following antibodies (HSP60: 1/100; cytochrome c: 1/100; Alexa-594 anti-rabbit IgG: 1/500; Alexa-488 anti-mouse IgG: 1/500) and stained with DAPI (0.1 μg/ml). To quantify phenotypes, cells were stained directly in the culture using 100 nm CMXros Mitotracker® Red (Molecular Probes) for 30 min, then fixed and DAPI-stained. To analyze the ΔΨm, cells were incubated 20 min with 5 μg/ml JC-1 (Molecular Probes) in culture medium and observed by confocal microscopy. Fluorescence images were captured and processed using a Leica DMIRE-2 microscope or a LSM-410 Zeiss confocal microscope. For FACS analyses of the ΔΨm, cells were removed from the culture dishes and then pelleted by low centrifugation, resuspended in culture medium with 50 nm TMRE (Molecular Probes) (22Castedo M. Ferri K. Roumier T. Metivier D. Zamzami N. Kroemer G. J. Immunol. Methods. 2002; 265: 39-47Google Scholar), and analyzed 15 min later by flow cytometry (FACS-Calibur, BD Biosciences) using the Cellquest software. Control experiments included treatment with 50 mm DNP (Sigma) for 60 min or 2 μm Oligomycin (Sigma) for 30 min or 0.1% Nonidet P-40 for 30 min. For transmission electron microscopy, cells were fixed for 2 h with 4% glutaraldehyde in sodium cacodylate buffer, post-fixed for 1 h with 1% osmium tetroxide, dehydrated, and embedded in Epon (EMS). Thin sections adsorbed onto nickel grids were stained with 1% uranyl acetate and 0.3% lead citrate and imaged in a JEOL-1200 EX electron microscope at 80 kV. Mitochondria surfaces were assessed with the Image Tools 3.0 software. We have designed a specific human OPA1 siRNA that corresponds to a sequence situated downstream of the GTPase coding sequence and analyzed its effect in HeLa cells comparatively with a control scramble siRNA (Dharmacond Research). After transfection, cells were collected every day for 3 days and Western blots performed. The different forms of OPA1 (1Olichon A. Emorine L.J. Descoins E. Pelloquin L. Brichese L. Gas N. Guillou E. Delettre C. Valette A. Hamel C.P. Ducommun B. Lenaers G. Belenguer P. FEBS Lett. 2002; 523: 171-176Google Scholar, 21Delettre C. Griffoin J.M. Kaplan J. Dollfus H. Lorenz B. Faivre L. Lenaers G. Belenguer P. Hamel C.P. Hum. Genet. 2001; 109: 584-591Google Scholar) were slightly affected at 24 h, declined at 48 h, and were almost not detectable at 72 h (Fig. 1 a). In parallel, the mitochondrial morphology was monitored on cells stained with the Mitotracker® Red (Fig. 1 b). 24 h after transfection of the OPA1 siRNA, cells retained a filamentous mitochondrial network as control cells (Fig. 1 b, top left). At 48 h, when OPA1 levels started to decline, half of the cells showed a filamentous mitochondrial network and half showed a fragmented one (Fig. 1 b, bottom left). At 72 h, when OPA1 levels were barely detectable, the filamentous phenotype had virtually disappeared to the benefit of the fragmented phenotype (60%), and of an additional phenotype (35%) that we described in Fig. 4. Fragmentation of the mitochondrial network induced by the loss of OPA1 was further visualized by immunofluorescence experiments using HSP60 antibodies and confocal microscopy three-dimensional reconstructions (Fig. 1 c).Figure 4Loss of OPA1 triggers apoptosis. a, kinetic of appearance of the apoptotic phenotype in cells transfected with the scramble (−) or the OPA1 (+) siRNA stained with DAPI and Mitotracker® Red. A Western blot detecting the PARP-cleaved form (PARP p89) and actin, 24, 48, and 72 h after transfection. As control, OPA1 siRNA-transfected cells were treated with z-VAD-fmk, or untransfected cells were treated with CDDP.b, fluorescence microscopy using DAPI and Mitotracker® Red or cytochrome c antibodies on cells, 72 h after transfection with scramble (control) or OPA1 siRNA.View Large Image Figure ViewerDownload (PPT) As a correlation between the IM potential dissipation and the fragmentation of the mitochondrial network has been recently established (23Legros F. Lombes A. Frachon P. Rojo M. Mol. Biol. Cell. 2002; 13: 4343-4354Google Scholar), we checked the effect of OPA1 down-regulation on the mitochondrial membrane potential (ΔΨm) using JC1 and TMRE dyes. 72 h after OPA1 siRNA transfection, fluorescence microscopy observations revealed that the fragmented cells were all devoid of the JC1 ΔΨm-sensitive red fluorescence (Fig.2 a), as cells treated with the uncoupling drug DNP, and in deep contrast to cells treated with oligomycine, that increases the ΔΨm and consequently the amount of red fluorescence. FACS analyses performed with TMRE dye revealed that the ΔΨm in cells treated with the OPA1 siRNA was decreased to a level similar to that estimated for cells treated with the DNP drug and slightly higher than that observed for cells treated with the Nonidet P-40 detergent (Fig. 2 b). To gain insight on the possible function of OPA1 on the mitochondrial IM, we performed transmission electron microscopic observations on cells transfected or not by the OPA1 siRNA and grown for 72 h. Classical long tubular cylinders (mean mitochondrial surface area: 0.66 ± 0.49 μm2,n = 69) were observed in control cells (Fig.3 a), while circular vesicles of reduced diameter (0.115 ± 0.059 μm2,n = 194) predominated in treated cells (Fig. 3,b and c), confirming the photonic observations. Mitochondria from cells devoid of OPA1 further presented electron denser matrix and cristae completely unstructured, adopting unusual shapes consisting of vesicle-like structures with abnormal increased space between the membranes (Fig. 3, b, c, andb′), providing evidence suggesting that OPA1 has a major function in structuring the cristae membranes. Our electron microscopy pictures are consistent with descriptions of mitochondrial pyknosis (24Desagher S. Martinou J.C. Trends Cell Biol. 2000; 10: 369-377Google Scholar) and of the switch from wild-type to apoptotic mitochondria in liver cells (18Scorrano L. Ashiya M. Buttle K. Weiler S. Oakes S.A. Mannella C.A. Korsmeyer S.J. Dev. Cell. 2002; 2: 55-67Google Scholar). In addition, our previous photonic observations, revealed 72 h after transfection of the OPA1 siRNA a high proportion (35%) of cells, with aggregated mitochondria, condensed chromatin, and fractionated nucleus reminiscent to apoptotic phenotype (Fig.4 a). To assert apoptosis, we estimated the amounts of the cleaved PARP protein (25Sallmann F.R. Bourassa S. Saint-Cyr J. Poirier G.G. Biochem. Cell Biol. 1997; 75: 451-456Google Scholar) by Western blot (Fig. 4 a) and found that they correlated with the percentage of apoptotic cells. PARP cleavage did not occur in the presence of the caspases inhibitor z-VAD-fmk, but well in cells treated with CDDP (Fig.4 a). Cytochrome c release was further monitored by immunofluorescence using cyt c antibodies 72 h after the OPA1 siRNA transfection. Barely half of the cell population had released cyt c in the cytoplasm (Fig. 4 b), while no cyt c release or apoptotic figure was observed in control cells. To address the issue whether the anti-apoptotic Bcl2 protein could counteract the OPA1 induced apoptotic process, the OPA1 and scramble siRNA were transfected in an ovarian adenocarcinoma cell line (NIH-OVCAR-3) expressing constitutively or not Bcl2 (19Lafon C. Mathieu C. Guerrin M. Pierre O. Vidal S. Valette A. Cell Growth Differ. 1996; 7: 1095-1104Google Scholar) (Fig.5 d). Loss of OPA1 (Fig.5 d) triggered fragmentation of the mitochondrial network in both cell lines (Fig. 5, a and b), while no morphological change was found using the scramble siRNA. In addition, 72 h after transfection, apoptosis (Fig. 5 c) reached 27% in the OVCAR cells, but was reduced by a 3-fold ratio in cells overexpressing Bcl2 (Fig. 5 d). Thus, in a second transformed cell line, loss of OPA1 triggers fragmentation of the mitochondrial network and an apoptotic process that can be inhibited by Bcl2 overexpression. Results presented here have shown that loss of OPA1 induces structural and functional modifications of the mitochondria. Indeed, down-regulation of OPA1 expression changes the mitochondrial network fusion-fission balance toward fission, as already noticed for its orthologs Mgm1p (13Wong E.D. Wagner J.A. Gorsich S.W. McCaffery J.M. Shaw J.M. Nunnari J. J. Cell Biol. 2000; 151: 341-352Google Scholar) and Msp1 2E. Guillou, unpublished data. in yeast. Thus OPA1 could either contributes as a fusion effector or as a fission inhibitor. Alternatively, the fragmentation of the mitochondrial network could be the consequence of the dissipation of the IM potential induced by loss of OPA1. In this respect, we further observed that the cristae were completely disorganized in the absence of OPA1. This later phenomenon could be the central process that would first dissipate the IM potential, then open the cristae junctions, allowing mobilization of cyt c and its release in the cytoplasm to trigger caspase-dependent apoptosis. How in this situation the cytc crosses the OM remains unclear, but this process and the downstream apoptotic events are inhibited by the overexpression of Bcl2. These data converge to propose that OPA1 participates in structuring the cristae to maintain their functional integrity. Therefore, OPA1 might be a key protein that modulates the IM dynamic to either maintain cell homeostasis or commit them to apoptosis and, consequently, could be a target for pro- or anti-apoptotic effectors. Thus coordinate structural changes of both mitochondrial membranes during apoptosis could be performed by both families of mitochondrial dynamin, as Drp1 was recently shown to be involved in the fragmentation of the mitochondrial network during apoptosis (26Frank S. Gaume B. Bergmann-Leitner E.S. Leitner W.W. Robert E.G. Catez F. Smith C.L. Youle R.J. Dev. Cell. 2001; 1: 515-525Google Scholar). Finally, the link between OPA1 and apoptosis gives clues on the possible pathophysiological effect of mutations in the OPA1gene found in ADOA patients that leads to the neurodegeneration of the retinal ganglion cells (5Kjer P. Jensen O.A. Klinken L. Acta Ophthalmol. (Copenh.). 1983; 61: 300-312Google Scholar). Indeed, this cell population undergoes two rounds of apoptotic process during normal life: one during development when setting up the retina, which eliminates 70% of them, and one during every day life, which accounts for a loss of few thousands of them per year (27Farkas R.H. Grosskreutz C.L. Ophthalmol. Clin. North Am. 2002; 15: 61-68Google Scholar). Haplo-insufficiency generated by the dysfunction of one OPA1 allele (4Thiselton D.L. Alexander C. Taanman J.W. Brooks S. Rosenberg T. Eiberg H. Andreasson S. Van Regemorter N. Munier F.L. Moore A.T. Bhattacharya S.S. Votruba M. Invest. Ophthalmol. Vis. Sci. 2002; 43: 1715-1724Google Scholar) might increase their susceptibility to apoptogen stimuli and in particular the one generated by the daily exposure to the UV light. We are indebted to the IFR109 technical resources and Brice Roncin for help in electron and confocal microscopy and members of the UMR5088 for helpful discussions.