Increase in Cardiac Ischemia-Reperfusion Injuries in Opa1+/- Mouse Model

Background Recent data suggests the involvement of mitochondrial dynamics in cardiac ischemia/reperfusion (I/R) injuries. Whilst excessive mitochondrial fission has been described as detrimental, the role of fusion proteins in this context remains uncertain. Objectives To investigate whether Opa1 (protein involved in mitochondrial inner-membrane fusion) deficiency affects I/R injuries. Methods and Results We examined mice exhibiting Opa1delTTAG mutations (Opa1+/-), showing 70% Opa1 protein expression in the myocardium as compared to their wild-type (WT) littermates. Cardiac left-ventricular systolic function assessed by means of echocardiography was observed to be similar in 3-month-old WT and Opa1+/- mice. After subjection to I/R, infarct size was significantly greater in Opa1+/- than in WTs both in vivo (43.2±4.1% vs. 28.4±3.5%, respectively; p<0.01) and ex vivo (71.1±3.2% vs. 59.6±8.5%, respectively; p<0.05). No difference was observed in the expression of other main fission/fusion protein, oxidative phosphorylation, apoptotic markers, or mitochondrial permeability transition pore (mPTP) function. Analysis of calcium transients in isolated ventricular cardiomyocytes demonstrated a lower sarcoplasmic reticulum Ca2+ uptake, whereas cytosolic Ca2+ removal from the Na+/Ca2+ exchanger (NCX) was increased, whilst SERCA2a, phospholamban, and NCX protein expression levels were unaffected in Opa1+/- compared to WT mice. Simultaneous whole-cell patch-clamp recordings of mitochondrial Ca2+ movements and ventricular action potential (AP) showed impairment of dynamic mitochondrial Ca2+ uptake and a marked increase in the AP late repolarization phase in conjunction with greater occurrence of arrhythmia in Opa1+/- mice. Conclusion Opa1 deficiency was associated with increased sensitivity to I/R, imbalance in dynamic mitochondrial Ca2+ uptake, and subsequent increase in NCX activity.


Objectives
To investigate whether Opa1 (protein involved in mitochondrial inner-membrane fusion) deficiency affects I/R injuries.

Methods and Results
We examined mice exhibiting Opa1 delTTAG mutations (Opa1 +/-), showing 70% Opa1 protein expression in the myocardium as compared to their wild-type (WT) littermates. Cardiac left-ventricular systolic function assessed by means of echocardiography was observed to be similar in 3-month-old WT and Opa1 +/mice. After subjection to I/R, infarct size was significantly greater in Opa1 +/than in WTs both in vivo (43.2±4.1% vs. 28.4 ±3.5%, respectively; p<0.01) and ex vivo (71.1±3.2% vs. 59.6±8.5%, respectively; p<0.05). No difference was observed in the expression of other main fission/fusion protein, oxidative phosphorylation, apoptotic markers, or mitochondrial permeability transition pore (mPTP) function. Analysis of calcium transients in isolated ventricular cardiomyocytes demonstrated a lower sarcoplasmic reticulum Ca 2+ uptake, whereas cytosolic Ca 2+ removal from the Na + /Ca 2+ exchanger (NCX) was increased, whilst SER-CA2a, phospholamban, and NCX protein expression levels were unaffected in Opa1 +/-

Opa1 delTTAG mice and breeding
As previously described, our group generated a knock-in Opa1 mouse model carrying the recurrent Opa1 delTTAG mutation [15]. Briefly, C57Bl6/J mice were genetically modified to obtain the Opa1 delTTAG heterozygous mutation (Opa1 +/-) that is observed in 30% of patients affected by dominant optic atrophy [15]. Homozygous Opa1 -/is lethal in the early stages of fetal development. Male Opa1 +/mice and their counterpart wild-type (WT) controls exhibiting Opa1 +/+ were fed and hydrated without any restrictions. They were held in the animal facility of the UMR INSERM 1083-CNRS 6214 in Angers, France. All experiments were performed in compliance with the European Union and French guiding principles on the protection of animals used for scientific purposes (EU Directive 2010/63/EU; French Decree no. 2013-118). The protocol was approved by the local ethics committee (Comité d'Ethique en Expérimentation Animale des Pays de la Loire) and by the national committee (MENESR, 2015101511544187/APAFIS 3723).

Blood pressure assessment
Blood pressure was measured by means of tail-cuff plethysmography (Visitech BP2000 System) in awake mice. For each mouse, the mean systolic blood pressure was averaged from 15 measurements recorded for 15 minutes every day for five consecutive days.
Isolated cardiomyocytes were loaded with fluo-4 AM (4μM for 20 min, Molecular Probes, Eugene, Oregon, USA) and then placed on the stage of an inverted confocal microscope (LSM510 Meta Zeiss, Carl Zeiss, Jena, Germany) equipped with a 63x water-immersion objective (NA: 1.2). Cells were field-stimulated at 1Hz in a tyrode solution containing: 135mM NaCl, 4mM KCl, 1.8mM CaCl 2 , 1mM MgCl 2 , and 2mM HEPES (NaOH-adjusted pH of 7.4). Triggered calcium transients were recorded in line-scan mode (1.5ms/line, 3,000 lines/cell) along the longitudinal axis of the cell. Fluo-4 was excited at 488nm, and emitted light was collected using a 505nm-long pass filter. The laser intensity employed (<10% of maximum) had no noticeable detrimental effect on the fluorescent signal or on cell function over the course of the experiment.
Analysis was performed using the software ImageJ (NIH, Bethesda, Maryland, USA). In order to enable comparison between cells, the change in fluorescence (ΔF) was divided by the fluorescence detected immediately prior to the 1Hz stimulation pulse (F0). For assessment of sarcoplasmic reticulum (SR) Ca 2+ content, cardiomyocytes were paced 1-2min at 1Hz to reach steady-state Ca 2+ transients and SR Ca 2+ load. Stimulation was subsequently stopped, and caffeine (10mM) applied to empty the SR.
The course of field-stimulated Ca 2+ transients and caffeine-induced SR Ca 2+ release was assessed by analyzing the maximum amplitude and time constant (τ) of the exponential part of the late decay phase. The time-to-peak normalized to the peak amplitude was also determined for stimulated Ca 2+ transients.

Mitochondrial calcium measurement and ventricular action potential
Mitochondrial Ca 2+ movements and ventricular action potential (AP) were simultaneously recorded by means of a whole-cell patch-clamp technique, as previously described [18,19]. Briefly, after loading with Rhod-2 AM (5μM for 40 min at 37°C, TEFLabs, Austin, Texas, USA), the left-ventricular cardiomyocytes were placed on the stage of an inverted confocal microscope (LSM510 Meta Zeiss, 63x objective, NA: 1.2) and superfused with a standard tyrode solution containing 135mM NaCl, 4mM KCl, 1.8mM CaCl 2 , 1mM MgCl 2 , 2mM HEPES, and 10mM glucose at an NaOH-adjusted pH of 7.4. Subsequently, isolated cardiomyocytes were whole-cell-patch-clamped using an Axopatch 200B (Molecular Devices, Sunnyvale, California, USA), and patch pipettes (2-3MO) were filled with an internal solution containing 130mM KCl, 25mM HEPES, 3mM MgATP, 0.4mM NaGTP, 5mM NaCl, and 0.5mM EGTA at a KOH-adjusted pH of 7.2. Such an approach allows the dialysis of cytosolic rhod-2 [18,19]. APs were elicited using a current clamp by means of 0.2ms current injections of suprathreshold intensity and stimulated routinely at 1Hz as previously described [20]. Shifts in dye fluorescence were recorded in line-scan mode (1.54ms/line) along the short axis of the cell. As regards fluo-4 signals, confocal images were analyzed using the software ImageJ, and the change in fluorescence (ΔF) was divided by the fluorescence detected immediately prior to the 1Hz stimulation pulse (F0). APs were recorded and analyzed using the software program Pclamp 10 (Axon Instruments). AP durations were measured at 20%, 30%, 50%, 90%, and 95% of the repolarization phase.

Mitochondrial morphology and function
Electron microscopy. Hearts from 3-month old mice were fixed by a retrogradely perfusion with 2% glutaraldehyde in cacodylate buffer (100mM sodium cacodylate and 2 mM MgCl 2 ; pH 7.3). Then, left-ventricular papillary muscles were isolated and post-fixed as previously described [21]. Ultra-thin longitudinal sections were cut, and examined using an electron microscope. 20,000x-magnified images were used in the software ImageJ in order to assess the number and area of mitochondria. A mean of 20 fields were analyzed per mouse.
Mitochondrial respiration. Oxidative phosphorylation was determined as previously described [22]. The hearts were excised, and mitochondria isolated. Complex activity was assessed using 350μg of mitochondrial proteins. Complex I was determined by adding 5mM Glutamate/Malate/Puryvate (GMP), complex II by adding rotenone in conjunction with succinate, and complex IV by adding antimycine with PMDP ascorbate.
Mitochondrial permeability transition pore experiments. Mitochondrial permeability transition pore (mPTP) opening was assessed as previously described [23]. Calcium retention capacity (CRC) measurement using 250μg of mitochondrial proteins was performed at baseline and after the adjunction of cyclosporine A (1μM) in Opa1 +/and WT mice.

I/R experiments
Ex vivo I/R. Opa1 +/and WT mice were anesthetized by means of an intraperitoneal injection of sodium pentobarbital (100mg/kg, Ceva Santé Animal 1 , Libourne, France). Heparin (1,000 IU/kg, Heparine Choay 1 ) was also administered to prevent intracoronary clot formation. The heart was rapidly excised and immediately immersed in ice-cold modified Krebs-Henseleit buffer containing: 118.5mmol/l NaCl; 4.7mmol/l KCl; 25mmol/l NaHCO 3 ; 1.2mmol/l MgSO 4 ; 1.2mmol/l KH 2 PO 4 ; 1.8mmol/l CaCl 2 ; and 11mmol/l glucose (pH 7.4). The heart was mounted on a Langendorff-perfusion apparatus (ADInstruments, Dunedin, New Zealand) and retrogradely perfused through the aorta with non-recirculating buffer saturated with 95% O 2 and 5% CO 2 at 37°C. The heart was maintained in a thermostatic chamber at 37°C. Perfusion was maintained at a constant pressure of 80mmHg. After a 20-minute stabilization period, the hearts were subjected to 30 minutes of global ischemia and 2 hours of reperfusion.
In vivo I/R. Opa1 +/and WT mice were anesthetized by means of intraperitoneal sodium pentobarbital injections (90mg/kg; Ceva Santé Animale) prior to endotracheal intubation and mechanical ventilation using MiniVent 845 1 (Hugo-Sachs Elektronik-Harvard Apparatus GmbH, March, Germany). The mice received an injection of heparin (1,000 IU/kg, Heparine Choay 1 , Sanofi, Gentilly, France) prior to thoracotomy. Temperatures were monitored during the procedure and strictly maintained at 36.5°to 38°C. The thorax was opened via a left lateral thoracotomy, and the pericardium removed. Coronary ligation was performed on the left coronary artery using a monofilament (PROLENE 7.0 1 , Ethicon, LLC, Cincinnati, Ohio, USA), placed in a polyethylene tube to create a reversible snare, as previously described [24]. Ischemia was induced by clamping the tube and confirmed through observations of dyskinesia and cyanosis of the myocardial region below the suture. It was maintained for 45 minutes, followed by 120 minutes of reperfusion by loosening the snare.
Infarct size assessment. For in vivo experiments, coronary ligation was repeated at the end of the reperfusion period, and 400μl of Evans Blue (Sigma 1 ) were injected through the left-ventricular apex, coloring the perfused (non-ischemic) myocardium blue. The area-at-risk (AAR) was then identified as the non-blue myocardium area and expressed as a percentage of left-ventricular area (LV). After heart extraction, the left ventricle was cut into six to seven slices before staining it with 2,3,5-triphenyltetrazolium chloride (TTC). The slices were subsequently submerged in 4% paraformaldehyde for 10 minutes, thereby delimiting the viable myocardium in red and area of necrosis (AN) in white. Quantification (planimetry) was performed using the software ImageJ 1.47. Infarct size was expressed as a percentage of the AAR (AN/ AAR).
For isolated perfused hearts, infarct size was calculated after TTC staining and expressed as a percentage of the LV (AN/LV).

Analysis of apoptosis
TUNEL staining was carried out on thin slices of heart tissue from AAR fixed in 4% paraformaldehyde, using the DeadEnd™ Fluorimetric TUNEL System (Promega 1 , Fitchburg, Wisconsin, USA), whilst adhering to manufacturers' instructions. Propidium iodide (red) and fluorescein (green) were used to color total and apoptotic nuclei, respectively. Slides were observed using a confocal fluorescence microscope. Caspase-3 activity was determined using the Caspase 3 Colorimetric Assay Kit (Abcam 1 plc, Cambridge, UK), thereby adhering to manufacturer's instructions. Absorbance of ρ-nitroaniline was determined via spectrophotometry at 405nm and employed as an indirect indicator of the amount of substrate cleaved by caspase-3.

Statistical analysis
Statistical analyses were performed using the software SPSS Version 15.0 for Windows (SPSS Inc., Chicago, Illinois, USA). Results were expressed as mean±standard error of the mean (SEM). After testing linear variable distribution, group differences were assessed using ANOVA, followed by Student's t-test. A non-parametric Mann-Whitney test was employed when appropriate. A p-value <0.05 was considered statistically significant.

Results
Cardiac structure and function in Opa1 +/mice We characterized cardiac function by means of echocardiography in 3-and 6-month-old Opa1 +/and WT mice. At 3 months of age, BW, HW, were not significantly different between Opa1 +/-(n = 11) and WT (n = 6) mice, as well as left-ventricular size and function (n = 9 per group) (Fig 1A-1D). At 6 months of age, Opa1 +/mice exhibited an alteration of left-ventricular systolic function (FS = 45±1% in Opa1 +/vs. 59±1% in WT [n = 10 per group], p<0.01), associated with significant dilation of the left ventricle (LVEDD = 3.35±0.13mm in Opa1 +/vs. 2.97±0.1mm in WT group, p<0.05). As at 3 months, no statistical difference in BW and indexed HW was observed between Opa1 +/and WT groups at 6 months. It is important to note that heart rates recorded at the time of echocardiography did not differ significantly between the groups at the same age (542±17bpm in WT and 548±20bpm in Opa1 +/at 3 months, and 580±15bpm in WT and 545±20bpm in Opa1 +/at 6 months). Systolic blood pressure likewise displayed no significant difference between groups at 3 months (120±2mmHg in WT and 118±1mmHg in Opa1 +/-) and 6 months (120±1mmHg in WT and 122±2mmHg in Opa1 +/-; n = 12/group).
Given that we previously reported aberrant structural conformation of myofibrils with large punctuated mitochondria and sarcomere disorganization, with large zones of autophagic and mitophagic materials in 5-month-old Opa1 +/- [15], we performed electron microscopy analysis in 3-month-old mice (n = 6/group). No difference was observed in structural conformation, nor the number and area of mitochondria between WT and Opa1 +/mice (Fig 1E and 1F).

Response of Opa1 +/mice to I/R
In order to investigate whether a decreased expression of the fusion protein Opa1 affects I/R sensitivity, the following experiments were performed on 3-month-old mice exhibiting preserved cardiac morphological and functional parameters. First, hearts from 11 WT and 12 Opa1 +/mice were subjected to 30 minutes of ex vivo global ischemia and 2 hours of reperfusion. As shown in Fig 2A-2C, infarct size was significantly greater in the Opa1 +/group as compared to the WT group (AN/LV = 71.1±3.2% vs. 59.6 ±8.5%, respectively; p<0.05). Similarly, Opa1 +/mice displayed significantly greater infarct size than the control group when subjected in vivo to coronary artery ligation for 45 min, followed by 2 hours of reperfusion (AN/AAR = 43.2±4.1% in OPA1 +/vs. 28.4±3.5% in WT, n = 8/ group, p<0.01), whilst AAR/LV did not differ between groups (Fig 2D-2F). Thus, decreased Opa1 expression increased susceptibility to myocardial I/R.

Evaluation of mitochondrial physiology
To gain insight into the cellular mechanisms responsible for I/R susceptibility, we assessed mitochondrial parameters. First, we assessed whether decreased Opa1 expression affects the amounts of the other proteins involved in mitochondrial dynamics. Myocardial expression of Opa1, Mfn2 fusion and Drp1, Fis1 fission proteins were assessed in 3-month-old mice. Opa1 expression was significantly lower in Opa1 +/mice (p<0.01), whilst expression of Drp1, Fis1, and Mfn2 was similar between the WT (n = 6) and Opa1 +/-(n = 6) groups (Fig 3). Similar results were obtained after 45 minutes of ischemia and 2 hours of reperfusion. We further characterized mitochondrial respiration in Opa1 +/and WT mice, by assessing the respirations of complex I, II and IV in baseline conditions and state 3 / state 4 respiration ratio, without evidencing significant difference (n = 6/group). After 45 minutes of ischemia and 2 hours of reperfusion, oxidative phosphorylation was clearly reduced in both groups, with no significant difference between Opa1 +/and WT mice (Fig 4). In addition, we evaluated the parameters related to apoptosis, and found that the expression of the pro-apoptotic Bax after I/R was significantly higher in the Opa1 +/group (n = 4) compared to the WT group (n = 6) (p<0.05) ( Fig  5A-5D), although Bcl-2 expression and Bax/Bcl-2 ratio were not different between groups. Moreover, the number of apoptotic TUNEL-positive nuclei did not differ between Opa1 +/and WT groups (respectively 11.5±3.5%, n = 8 vs. 9.1±2.7%, n = 7; p = 0.56) nor the caspase-3 activity (n = 7/group; p = 0.21).

mPTP function
Finally, we assessed the opening of the mPTP. Under baseline conditions, calcium retention capacity was not significantly different in the Opa1 +/group than in the WT group (Fig 5E). Administration of cyclosporine A, a selective mPTP inhibitor, significantly increased calcium retention, but to similar degrees in both groups. After 45 minutes of ischemia and 2 hours of reperfusion, Opa1 +/mice displayed similar calcium retention capacities to WT mice. Moreover, there was no significant difference in baseline and post-I/R calcium retention capacity. When compared to baseline, the post-I/R effect of cyclosporine A adjunction was reduced.

Calcium transients
We subsequently compared calcium transients in isolated cardiomyocytes stimulated at 1Hz. In Opa1 +/left-ventricular cells, the amplitude of calcium transients was significantly lower and exhibited a slower decay time constant compared to WT cardiomyocytes (Fig 6A-6C). This indicates that SR calcium uptake is affected in Opa1 +/cells. Similarly, the extend of caffeine-induced SR calcium release was comparable in WT and Opa1 +/myocytes. That said, the decay time constant, which depends primarily upon sarcolemmal Na + /Ca 2+ exchanger (NCX) activity, proved to be much faster in Opa1 +/myocytes (Fig 6D-6F). Overall, these results suggest that, in Opa1 +/-, SR Ca 2+ uptake was decreased, whereas cytosolic Ca 2+ removal due to NCX increased. It is important to note that myocardial expressions of SERCA2a, total and phosphorylated phospholamban, and NCX proteins were not significantly different between WT (n = 6) and Opa1 +/-(n = 4) groups in 3-month-old mice.
We subsequently determined whether mitochondrial Ca 2+ movements and AP durations were affected in Opa1 +/ventricular cardiomyocytes. To this end, we simultaneously recorded rhod-2 signals and AP using a whole-cell patch-clamp technique. As shown in Fig 7, and in accordance with increased Ca 2+ extrusion through NCX, AP durations in Opa1 +/cardiomyocytes were significantly increased during the late phase of repolarization. Furthermore, 50% of Opa1 +/cardiomyocytes were triggered soon after depolarization (6/12 cells), whereas no such arrhythmia was observed in WT cardiomyocytes (Fig 7). On the other hand, mitochondrial Ca 2+ uptake during steady-state AP was lower and exhibited a slower rate of increase. The mitochondrial Ca 2+ decay time constant was not significantly affected in Opa1 +/cardiomyocytes compared to the WT group (Fig 8). Altogether, these results indicate that in Opa1 +/cardiomyocytes, mitochondrial Ca 2+ uptake is impaired, whereas the late repolarization phase of APs is markedly increased with an enhanced arrhythmia triggering.

Discussion
The susceptibility of the myocardium to develop I/R injury is highly related to the mitochondrial functions [26]. Indeed, during ischemia, cardiac mitochondria are subjected to hypoxia, calcium overload, low pH, and ATP depletion. Reperfusion abruptly subjects mitochondria to a rapid pH recovery, oxidative stress, restoration of mitochondrial membrane potential, and calcium overload, all of which inducing mPTP opening [27]. mPTP opening at reperfusion onset is a primary trigger of cardiomyocyte death [28]. Of note, modulation of mitochondrial dynamics may affect susceptibility to I/R injury by preventing mitochondrial dysfunction and mPTP inhibition [6]. Whilst excessive fission has been associated with I/R injury and cell death, it was unclear whether loss of the fusion protein Opa1 likewise contributes to the occurrence of I/R injury in vivo. Our study demonstrated that Opa1 deficiency is associated with increased sensitivity to I/R in mice exhibiting the Opa1 delTTAG mutation, both ex vivo and in vivo.
Several studies have reported mitochondrial fragmentation in cardiac cells subjected to simulated I/R [6,11,29], a phenomenon associated with a reduction in Opa1 protein expression [11]. However, most of these studies were performed in vitro. Ong et al. have shown that suppression of fission and stimulation of fusion by Drp1 downregulation, expression of dominantnegative Drp1, treatment with mdivi-1, or overexpression of Mfn1 or Mfn2 in HL-1 cardiac cell, prevented mPTP opening and reduced cell death after I/R [6].
In order to investigate in vivo the role of Opa1 deficiency in cardiac I/R injury, we studied our previously generated knock-in OPA1 mouse model [15]. This model carrying the recurrent groups. E: mPTP opening sensitivity at baseline and after I/R (n = 6/group). Required calcium overload for mPTP opening in WT and Opa1 +/groups at baseline, after I/R, with or without use of cyclosporine analog (CsA). The I/R protocol included 45 minutes of ischemia and 2 hours of reperfusion. Values are mean ± SEM. *p<0.05, **p<0.01, ***p<0.001. $ p< 0.05 compared to the same group at baseline + CsA. Non-parametric Mann-Whitney test (TUNEL and Bax/Bcl2 ratio) and t-test (Bax expression and caspase 3 activity). doi:10.1371/journal.pone.0164066.g005 Opa1 delTTAG heterozygous mutation affected one third of patients exhibiting dominant optic atrophy. The mouse displayed a multi-systemic poly-degenerative phenotype and presented a combination of visual failure, deafness, encephalomyopathy, peripheral neuropathy, ataxia, and cardiomyopathy [15]. Whilst cardiac muscles from these mice displayed aberrant structural conformation of myofibrils with large punctuated mitochondria but normal cristae organization and sarcomer disorganization, with large zones of autophagic and mitophagic materials at 5 months [15], we observed no difference with the WT littermates at 3 months of age, indicating age-related onset of cardiomyopathy.
In line with the age-related development of cardiac muscular dysfunction and mitochondrial structure abnormalities in this mouse model, we observed normal cardiac function at 3 months and moderate alteration of left-ventricular systolic function at 6 months. In another knock-in Opa1 model, Chen et al. reported similar, late-onset, cardiac dysfunction, associated with no significant cardiac functional or gross structural abnormalities at 3 months, but impaired systolic function at 12 months [12]. In our Knowledge, cardiac involvement in patients carrying OPA1 mutation has only been recently reported for the first time in two sisters harboring a homozygous OPA1 mutation leading to lethal encephalopathy and hypertrophic cardiomyopathy [30]. As we wished to explore the role of Opa1 deficiency prior to the occurrence of cardiac functional impairment, we opted to subject the mice to I/R at 3 months. At this age, Opa1 expression had decreased by 70% in the heart and, importantly, none of the other major fusion/fission proteins (Drp1, Fis1, and Mfn2) were affected.
Similarly to the data observed in vitro in other Opa1-deficient models, we recorded greater infarct size in Opa1-deficient mice subjected to I/R in vivo. Although acute I/R injury has been linked to mPTP opening at reperfusion onset following a period of ischemia [27], we observed no increased sensitivity to mPTP in Opa1-deficient mice either at baseline or after I/R. Furthermore, mitochondrial respiration was also not impaired in Opa1-deficient mice at baseline, and oxidative phosphorylation was lower both in OPA1-deficient mice and controls after I/R.
Calcium flux constitutes one of the major cellular factors in myocardial I/R injury [31,32]. As stated, Opa1 +/ventricular cardiomyocytes displayed a reduced calcium transient amplitude with a slower decay time constant, whereas the rate of calcium transient increase remained unaffected compared to WT cardiomyocytes [12]. Furthermore, our data suggests enhanced cytosolic Ca 2+ removal by means of the NCX, characterized by a faster decline of the caffeine-induced Ca 2+ transients and a drastic increase in the late repolarization phase of the Aps [33,34]. Increased NCX activity is associated with greater occurrence of arrhythmias, such as delayed after depolarization (DADs) and/or early after depolarization (EADs) [35]. The involvement of NCX in DADs is well-established and is related to increased NCX-dependent Ca 2+ extrusion during diastole, which generates an inward current and the resting membrane potential depolarization, whereas the NCX contribution to EADs genesis is more complex. Increased Ca 2+ extrusion during the AP repolarization phase slows down AP repolarization, favoring l-type Ca2+ current reactivation and triggering of EADs [36]. Enhanced NCX activity and increased AP were not associated with an increased expression of NCX in Opa1 +/or a shift in SR Ca 2+ load and SERCA2a 2+ /PLB, but with a decrease in beat-to-beat mitochondrial Ca 2+ uptake. Decreased dynamic mitochondrial Ca 2+ uptake may thus destabilize Ca 2+ flux equilibrium between SR Ca 2+ uptake, NCX Ca 2+ removal, and mitochondria in controlling cytosolic Ca 2+ homeostasis during Aps [37]. Whilst no definitive explanation prevails as to why Opa1 deficiency is associated with increased susceptibility to cardiac I/R injury, it may be postulated that an imbalance in Ca 2+ fluxes and the subsequent increase in NCX activity may contribute to the observed reperfusion injury. Indeed, increased NCX Ca 2+ extrusion may raise steady-state intracellular Na + levels during normoxia, which, during ischemia, may combine with the intracellular Na + rise mediated by Na + /K + ATPase inhibition and acidosis. Increased ischemic Na + levels would thus raise Ca 2+ influx via NCX reverse mode, thereby increasing cytosolic Ca 2+ overload during ischemia and subsequent reperfusion injuries [35].
On the other hand, Varanita T et al. recently reported genetically mild Opa1 overexpression to be associated with reduced occurrence of cardiac injury in Langendorff-perfusedhearts subjected to I/R [38]. It is worth noting that protection from I/R injury was not limited to the heart, as infarct size was likewise smaller in brains subjected to I/R. The commonly accepted conclusion is that fused mitochondria are more tolerant to ischemia and better adapt to factors causing mPTP opening at reperfusion onset. That said, a recent report has been the subject of contention regarding this concept. In a cardiac-specific Drp1 heterozygous knockout mouse, Ikeda et al. reported cardiac mitochondrial elongation associated with increased susceptibility to I/R [39]. Drp1 downregulation induced accumulation of damaged mitochondria and impaired mitochondrial autophagy. This data underscores that mitochondrial dynamics is an active process requiring a tight equilibrium between fusion and fission in order to maintain efficient cardiac cell functions, especially in critical situations such as I/R.

Limits
We employed a knock-in Opa1 mouse model carrying the recurrent Opa delTTAG mutation in order to assess the role of mitochondrial dynamics in the development of I/R injury in vivo as well as ex vivo. As a consequence of this chronic Opa1 deficiency, several unsuspected compensation mechanisms may have occurred in addition to the increase in NCX activity.
Moreover, we observed no significant increase in apoptosis in Opa1-deficient mice as an underlying cause for their higher infarct size. Due to the dynamic nature of apoptosis, it could be postulated that we overlooked such an increase by assessing apoptosis at single point in time. It is worth noting that Chen et al. likewise did not find any marker of increased apoptosis activity in their OPA1-deficient model using TUNEL, histological, Bak1, bcl2, bcl2/1, Bnip3, bk, and bax analysis [12]. Finally, calcium regulation was investigated in normoxic cardiomyocytes. It would be of interest to obtain similar information after hypoxia/reoxygenation.

Conclusion
Opa1 deficiency was associated with an increased sensitivity to I/R in mice with the Opa1 delT-TAG mutation, both ex vivo and in vivo. The mechanism seems to be related to an imbalance in dynamic mitochondrial Ca 2+ uptake and the subsequent increase in NCX activity.