The valosin-containing protein is a novel mediator of mitochondrial respiration and cell survival in the heart in vivo

The valosin-containing protein (VCP) participates in signaling pathways essential for cell homeostasis in multiple tissues, however, its function in the heart in vivo remains unknown. Here we offer the first description of the expression, function and mechanism of action of VCP in the mammalian heart in vivo in both normal and stress conditions. By using a transgenic (TG) mouse with cardiac-specific overexpression (3.5-fold) of VCP, we demonstrate that VCP is a new and powerful mediator of cardiac protection against cell death in vivo, as evidenced by a 50% reduction of infarct size after ischemia/reperfusion versus wild type. We also identify a novel role of VCP in preserving mitochondrial respiration and in preventing the opening of mitochondrial permeability transition pore in cardiac myocytes under stress. In particular, by genetic deletion of inducible isoform of nitric oxide synthase (iNOS) from VCP TG mouse and by pharmacological inhibition of iNOS in isolated cardiac myocytes, we reveal that an increase of expression and activity of iNOS in cardiomyocytes by VCP is an essential mechanistic link of VCP-mediated preservation of mitochondrial function. These data together demonstrate that VCP may represent a novel therapeutic avenue for the prevention of myocardial ischemia.

Acute myocardial ischemia (MI) and coronary artery disease remain among the top ranking causes of death and disability worldwide 1 . Mitochondrial dysfunction contributes to cell damage during ischemia/reperfusion (IR) and is central to cardiomyocyte death [2][3][4] . It has been shown that ischemic preconditioning (IPC), the gold standard for cardioprotection, attenuated the decline of mitochondrial function induced by ischemic injury, including oxidative phosphorylation, respiratory chain coupling and mitochondrial efficiency [5][6][7][8][9] . There is also increasing evidence that the mitochondrial permeability transition pore (mPTP) opening plays a central role in mediating both the necrotic and apoptotic components of IR injury, particularly at the onset of reperfusion, and inhibition of mPTP opening is considered to be a critical target of cardioprotection by IPC 3,10-14 . We identified previously that the valosin-containing protein (VCP), a member of a family that includes ATP-binding proteins, promotes a significant reduction in apoptosis in isolated cardiomyocytes under cell stress 15 . These data support the hypothesis that over-expression of VCP in vivo in the heart may provide protection against ischemic injury.
To test such hypothesis, we generated a transgenic (TG) mouse with cardiac-specific overexpression of VCP. Our results demonstrate that overexpression of VCP significantly reduced the infarct size (IS) by 50% after IR compared to wild type (WT) mice. We also showed that VCP exhibited a new role on activation of mitochondrial respiration efficiency and inhibition of the mPTP opening. Using both genetic and pharmacological approaches, we also show that the effect of VCP on the cardiac mitochondrial function was dependent upon the increase of inducible isoform of nitric oxide synthase (iNOS) conferred by VCP.

Mitochondrial distribution of VCP and iNOS was increased in VCP TG hearts. Our previous study
showed that not only the increase of total endogenous iNOS expression but also its subsequent translocation to the mitochondria in cardiac myocytes is crucial for its effects on both mitochondrial respiration and cardioprotection 18 . To further test this concept in VCP TG mouse, subcellular fractions were extracted from both VCP TG and WT mouse hearts as described in the Methods. To determine the cellular distribution of iNOS as accurately as possible, the purity of each cellar fraction was confirmed by specific protein markers (Fig. 3a). As shown in Fig. 3b, VCP and iNOS exhibited a similar subcellular distribution in which both VCP and iNOS were localized primarily in the mitochondria in both WT and TG mouse hearts. Compared with WT, TG mice showed a 3.1-fold increase in VCP and 4.5-fold increase in iNOS in the mitochondrial fraction (P < 0.01 vs. WT, Fig. 3c and d). iNOS activity in isolated mitochondria from VCP TG mouse hearts was also increased by 2.1-fold compared to WT mice, indicating its functional activity in heart mitochondria in VCP TG mice (Fig. 3e).
ADP-dependent oxygen consumption was enhanced in VCP in TG mouse heart. Since the majority of VCP is localized in mitochondria, we tested the effect of VCP on mitochondrial respiration in the heart. We first tested the complex I-dependent mitochondrial respiration by measuring oxygen consumption in the presence of pyruvate and malate, two Complex I substrates, with and without ADP, as illustrated in Fig. 4a. Respiration rates of states 2 to 4 were determined by the oxygen consumption per minute normalized by mitochondrial proteins. As shown in Fig. 4b, although there was a modest increase in state 2 respiration rate in the mitochondria from VCPTG mice vs WT, the difference did not reach statistical significance. Mitochondria from VCP TG mouse heart tissues showed a significant increase in oxygen consumption at the ADP-dependent state 3 under complex I stimulation compared to WT, while no significant difference was seen in oxygen consumption states 4 (upon the addition of oligomycin, a known inhibitor of the ATP synthase) (Fig. 4b). The efficiency of mitochondrial respiration, as measured by the respiratory control ratio (RCR: state 3/state 4) was significantly increased in VCP TG mice versus WT (Fig. 4c). In addition, VCP TG mice also exhibited a significant increase WT (n = 6) TG (n = 6)  in maximum respiration capacity versus WT as measured after the addition of the uncoupling agent carbonyl cyanide 4-trifluoro methoxyphenylhydrazone (FCCP) (Fig. 4d).
In addition, we tested the effect of VCP on complex II dependent respiration in the isolated mitochondria by adding succinate, a known Complex II substrate, in the presence of ADP, with and without the Complex I inhibitor rotenone. As shown in Fig. 4e and f, there was a remarkable increase in respiration rate from both WT and VCP TG mice after the addition of succinate compared to Complex I alone, suggesting that respiration under Complex II was additive. However, the difference of respiration rates between VCP TG and WT observed under stimulation of Complex I (Plot 2 in Fig. 4e and f) lost significance upon the simulation of Complex I and II after the addition of succinate (Plot 4 in Fig. 4e and f). We tested Complex II-specific respiration by adding the Complex I inhibitor, rotenone, and found that there was no longer a difference in respiration rate between VCP TG and WT (Plot 5 in Fig. 4e and f). These data together suggest that VCP may be predominantly affecting Complex I dependent respiration in VCP TG mice.
To determine whether the increase of oxygen consumption in VCP TG mice was the result of enhanced mitochondrial respiration, we used Cytochrome C to test the integrity of the mitochondrial membrane. As shown in Plot 3 of Fig. 4e and f, addition of cytochrome C did not increase the respiration rates of either group's samples, indicating a high membrane integrity. Next, we added potassium cyanide (KCN), a known inhibitor of cytochrome oxidase. Our data showed that addition of KCN inhibited the respiration rates for both WT and VCP TG in a similar manner, indicating a direct effect on mitochondrial respiratory chain (Plot 6 in Fig. 4e and f). Moreover, there were no significant difference observed between VCP TG and WT upon the addition of oligomycin after the inhibition of cytochrome oxidase by KCN (Plot 7 in Fig. 4e and f).

Genetic deletion of iNOS abolished VCP-mediated increase in mitochondrial respiration in vivo.
We have shown that VCP dose-dependently increases the expression of iNOS in myocytes 15 . However, it is unknown if iNOS mediates the effect of VCP on mitochondrial respiration. To further determine whether the stimulation of mitochondrial respiration observed in VCP TG mouse heart is mediated by iNOS in vivo, a bigenic VCP TG/iNOS KO −/− mouse was generated, in which iNOS was deleted from VCP TG mice (see Methods for mating strategy). Oxygen consumption was measured in these VCP TG/iNOS KO −/− mice and compared with their litter-matched VCP TG/iNOS +/+ mice and WT/iNOS +/+ counterparts. As shown in Fig. 4a to c, the increase in mitochondrial respiration state 3 and in RCR observed in VCP TG mouse heart was abolished upon deletion of iNOS. There was also a significant decrease in maximum respiration capacity as measured after the addition of the uncoupling agent FCCP in VCP TG/iNOS KO −/− mice compared to VCP TG mice (Fig. 4d). These data indicate that iNOS is necessary for the stimulation of respiration by VCP.

iNOS inhibition prevents VCP-induced increase in mitochondrial respiration in isolated cardiomyocyte in vitro.
To further determine whether the enhancement of mitochondrial respiration observed in VCP TG mice originated directly from cardiomyocytes, rat neonatal cardiac myocytes (RNCMs) were transfected with adenoviruses harboring the full-length VCP sequence (Ad-VCP) and compared to the Ad-β -Gal control (Ad-β -Gal). iNOS activity was increased by 2-fold upon the overexpression of VCP versus β -Gal. Addition of iNOS inhibitor 1400 W abolished VCP-mediated activation of iNOS (Fig. 4g). Consistent with the change in VCP TG mice, compared to Ad-β -Gal, RNCMs treated with Ad-VCP exhibited a 2.2-fold increase in oxygen consumption rates (OCR), which represents an index of mitochondrial respiration capacity. This effect was abolished by addition of the iNOS inhibitor, 1400 W (Fig. 4h). Therefore, these results in cultured cardiomyocytes offer further support to the results obtained from genetically-modified mice showing that the effect of VCP on mitochondrial respiration is dependent upon increased iNOS activity.
Over-expression of VCP prevents mPTP opening. In order to further investigate the protective role of VCP in IR, we next examined its potential effect on the mPTP opening, which is a critical mechanism to promote cell damage during IR, particularly at the early stage of reperfusion 3 . Mitochondria were isolated from the heart of WT, VCP TG and VCP TG/iNOS KO −/− mice. mPTP opening was induced by high concentration of calcium (600 µ M CaCl 2 ) and determined by the calcium-overload swelling assay. As showed in Fig. 5a and b, compared to the WT, the mitochondria isolated from the heart of VCP TG mice exhibits a significantly less decrease in absorbance at 540 nm upon the addition of CaCl 2 , indicating the inhibition of calcium induced mPTP opening. This protection was abolished by the deletion of iNOS in VCP TG/iNOS KO −/− mice (Fig. 5b). Furthermore, addition of cyclosporine A (CsA), a known mPTP inhibitor, prevented the mPTP opening in all the three groups and eliminated the difference between WT and VCP TG ( Fig. 5a and b). These data further support that VCP protects the mitochondria from the calcium-load induced mPTP opening. mPTP opening was also determined in isolated cardiomyocytes by a reduction in mitochondrial calcein fluorescence signal after addition of FCCP, which is an uncoupler of the mitochondrial respiratory chain, at the concentration of 1 µ mol/L or 10 µ mol/L in myocytes treated for 48 hours with Ad-β -Gal and Ad-VCP with or without the iNOS inhibitor 1400 W (Fig. 5c). Calcein fluorescence signaling was traced over time after addition of FCCP (Fig. 5d). Compared to the untreated control, the calcein fluorescence signal in Ad-β -Gal-treated RNCMs declined upon FCCP treatment (p < 0.05), reflecting mPTP opening. RNCMs treated with Ad-VCP preserved the same calcein fluorescence signal as control cells upon FCCP stimulation, indicating that VCP prevented the induced mPTP opening in RNCMs. This prevention conferred by VCP was abolished by the addition of the iNOS inhibitor 1400 W (Fig. 5e). Therefore, increased expression of iNOS by VCP is necessary for the reduction in mPTP pore opening in cardiomyocytes.

Discussion
As presented in Fig. 6, our results demonstrate that over-expression of VCP in a genetically-modified mouse model in vivo provides cardiac protection in ischemic heart equivalent to that provided by the SWOP 16,17 ; This effect is linked to a preservation of the capacity of mitochondrial respiration and a prevention of mPTP opening under the stress; The mechanistic link between VCP-mediated cardioprotection and preservation of mitochondrial function under the cardiac stress is an increased expression and activity of iNOS. Taken together, these results demonstrate that VCP is a novel agonist of iNOS-mediated mechanisms of cardioprotection against ischemia.
Very little information is available about the function of VCP in the heart, at the opposite of other tissues. VCP is a member of the type I AAA (ATPases associated with various cellular activities). By interacting with several sets of adaptor proteins 19 , VCP is involved in a variety of cellular pathways that are essential for cell homeostasis, such as cell cycle control, transcriptional regulation, apoptosis, protein degradation, and cellular stress response [20][21][22][23] . Despite this information gained from other tissues, the physiology of VCP in the heart remains largely unknown. We first identified VCP in the heart when we showed that it acts as a novel downstream effector of the cardioprotective signaling mechanism conferred by Hsp22 15 . Our initial studies in vitro showed that VCP represents the link between Hsp22-mediated activation of Akt and nuclear factor-kappa B (NF-ĸB)-induced expression of iNOS in cardiac myocytes, thereby playing a central role in the mechanisms of cardiac cell survival promoted by Hsp22. We also showed that overexpression of VCP protects cardiomyocytes against apoptosis 15 . Based upon this evidence obtained in vitro, we tested the physiological relevance of our findings in vivo by generating a cardiac-specific VCP TG mouse.
In the present study, we identified for the first time a specific physiological function of VCP in the mammalian heart, in both normal and stress conditions. Our results show that no change was observed in TG mouse compared with WT in baseline conditions when considering heart mass, ventricular structure, contractile function, and the size and viability of cardiomyocytes. These data demonstrate that chronic stimulation of VCP at this overexpression level has no toxic effects on the heart tissue. Thus, the VCP TG mouse provides an original biological tool for the investigation of the mechanisms underlying VCP-mediated actions on cardiac stress.
A major observation of the present study is that overexpression of VCP protects the heart against IR-induced damage. Acute MI remains one of the leading causes of morbidity, mortality and disability worldwide 1 . Early and successful reperfusion is the most effective strategy for reducing the IS and for improving the clinical outcome. However, the process of restoring blood flow to the ischemic myocardium can create injury itself, for which no effective therapy is currently available 5,24 . IPC is well recognized as the most powerful endogenous mechanism of cardioprotection, however the clinical translation of IPC remains limited because of the problematic requirement of pre-emptive ischemic episodes 16,17,24,25 . Our results defined VCP as a new and powerful mediator of cardiac protection during IR, demonstrating that VCP may represent a novel therapeutic avenue for the prevention of MI.
Cardiomyocyte death induced by IR is the major cause of myocardial infarct and loss of cardiac function. VCP has been identified as an apoptosis regulator in other mammalian cells and in yeast [26][27][28] , Recent studies have identified that loss of VCP activity due to mutations is linked to cell death in different human diseases, including a variety of neurodegenerative diseases, such as Alzheimer's disease, Pakinson's disease and amyotrophic lateral sclerosis 26,29,30 . VCP has also been shown to be associated with cell survival of cancer cells [31][32][33][34] . We demonstrated previously the protective role of VCP against apoptosis in mammalian cardiomyocytes in vitro 15 . However, whether this anti-apoptotic effect of VCP protects the heart against ischemia-induced cell death remained unclear. Our data further show that overexpression of VCP by 3.5-fold is sufficient to provide prophylactic cardioprotection against cell death in the heart under ischemic stress.
Another important finding in the present study is a novel role for VCP in mitochondria of the mammalian heart. Previous mechanistic studies on VCP-mediated cell survival focus on its involvement in endoplasmic reticulum (ER) stress-triggered apoptotic pathway 35,36 . VCP deficiency is associated with decreased mitochondrial membrane potential in human dopaminergic neuroblastoma cell line [37][38][39] . It has been shown that mutations or deficiency of VCP cause profound mitochondrial dysfunction which results in a significant reduction of ATP synthesis, making neuronal cells more vulnerable to ischemia and cell death [37][38][39] . However, the potential role of VCP in cardiac mitochondria has not been tested. We first showed that, in a mouse model, VCP exhibits a preferred accumulation in cardiac mitochondria as compared to other sub-cellular fractions, which highlights the possibility of an important mitochondrial function for VCP under stress conditions. This is supported by our observation that VCP TG mice exhibit a significant elevation of the capacity of mitochondrial respiration in the state 3, with no change in states 2 and 4. State 3 is defined as ADP-stimulated respiration. When ADP binds ATP synthase, protons are driven into the matrix from the outside of the inner membrane. The energy released by this proton flux directly drives ATP synthesis. Therefore, an increase in state 3 respiration indicates an increased capacity of ATP synthesis 6 , while states 2 and 4 are viewed as the steady states of mitochondrial respiration. Therefore, an increase in RCR, which reflects the state 3/state 4 ratio, is an excellent indicator of improved respiratory activity and efficiency 6 . Additionally, the increase of oxygen consumption was blocked by the inhibitor of cytochrome oxidase, KCN, further supporting the enhancement of mitochondrial respiratory chain by VCP. These data together indicates that overexpression of VCP promotes mitochondrial respiration efficiency, which enhances cellular resistance to oxidative damage, and thus may act as a mechanism of the prevention of cardiomyocyte death during ischemia 6 .
There is also increasing evidence that mPTP opening plays a central role in mediating both the necrotic and apoptotic components of IR injury particularly at the onset of reperfusion 3,12,40,41 . Importantly, it has been shown that IPC attenuates IR-induced opening of the mPTP 12 . Recent studies also demonstrated that the use of mPTP opening inhibitors, such as cyclosporin A, reduces IS in animal models of acute IR injury 41 . Despite intensive investigation, the molecular identity of the mPTP remains undefined. Our findings showed that overexpression of VCP reduced mPTP opening, indicating that VCP targets an important mitochondrial regulator of cardioprotection against reperfusion injury during IR.
Our next observation is that VCP regulates mitochondrial function in an iNOS-dependent manner. The effect of VCP on mitochondria is related to iNOS for the following reasons: (1) VCP induces the expression of iNOS not only in cardiac myocytes in vitro 15 but also in the heart of VCP TG mice in vivo. (2) The cytoprotection conferred by VCP was abolished by the addition of a selective iNOS inhibitor 15 . (3) iNOS is necessary for Hsp22-related mitochondrial regulation and cytoprotection 18 . To further investigate if the effect of VCP on mitochondrial function is mediated through the increase of iNOS expression, both pharmacological and genetic approaches were used. We first generated a bigenic mouse in which iNOS was deleted from the VCP TG mouse. This model confirms that enhanced mitochondrial respiration by VCP overexpression in the mouse heart was abolished by the deletion of iNOS. Importantly, VCP exhibited no stimulatory effect on another NOS isoform, such as eNOS. In addition, we also showed in isolated RNCMs that VCP overexpression results in stimulation of mitochondrial respiration, an effect which was eliminated upon treatment with the iNOS inhibitor 1400 W. Although the precise mechanism of VCP on mitochondrial function is largely unknown, our results imply that iNOS is a crucial mediator of such effect. Furthermore, it has been shown previously that mitochondria extracted from iNOS TG mouse hearts subjected to IR have a decreased mPTP opening compared to WT 42 . We demonstrated that overexpressing VCP decreased the rate of mPTP opening, and that such effect was completely abolished by the deletion or inhibition of iNOS, further indicating that iNOS is the critical mediator of the effect of VCP in the inhibition of mPTP opening.
Although the role of iNOS in cardioprotection has been firmly established, the effect of NO in the mitochondria has been controversial. Indeed, the detrimental effect of NOS on mitochondrial respiration were observed in several conditions related to excessive formation of NO 43,44 , such as during hypoxia [45][46][47] or under treatment with exogenous NO stimulators 48 . However, our data clearly show that the moderate increase of endogenous iNOS by 2-to 3-fold, which reproduces the extent of overexpression during SWOP, provides cardiac protection during IR injury and promotes mitochondrial respiration. These data are also consistent with our previous observation in a Hsp22 TG mouse model 18,49 as well as in ischemic preconditioning 50 . In addition, different conclusions were made when comparing the increased iNOS expression originating from resident inflammatory cells 51 or from circulatory blood cells 52 compared to the iNOS generated by cardiac myocytes 52 . Also, we recently showed that it, instead of the total cellular iNOS content; it is rather the abundance of iNOS in the mitochondria that is crucial for Hsp22-mediated stimulation of oxidative phosphorylation 18 . Accordingly, the present study shows that iNOS has a preferred location in mitochondria of VCP TG mouse heart, which further supports the concept that stimulation of endogenous iNOS expression and its subsequent translocation to the mitochondria in cardiac myocytes is crucial for its effects on both mitochondrial respiration and cardioprotection. This may also explain the divergent results observed in previous studies in which NO production originated from "non-myocyte" iNOS 47,52 or from exogenous NO donors 48 . The cellular origin and subcellular distribution of iNOS are likely major determinants of its beneficial effect against ischemic injury and mitochondrial respiration. The VCP TG mouse that we generated provides a biological tool for the further study of the mechanisms underlying endogenous iNOS mediated protection.
In summary, the data presented in this study provide the following novel information in the heart in vivo: (1) Cardiac-specific overexpression of VCP in a TG mouse did not change the overall physiological and morphological characteristics of the mouse in baseline conditions. (2) The VCP TG mouse provides a significant reduction in IS during IR injury, which is quantitatively comparable to that observed in IPC. (3) Overexpression of VCP in the heart in vivo leads to an increase in endogenous iNOS expression and activity. (4) Increased iNOS by VCP shows a preferential localization to the mitochondria in TG mouse heart. (5) VCP overexpression leads to an increase in mitochondrial respiratory capacity and an inhibition of the rate of mPTP opening, which are dependent upon iNOS expression and activity. Altogether, these data further highlight a novel cardioprotective mechanism for VCP in vivo. Therefore, pre-emptive activation of VCP and its downstream targets may represent a novel approach to prevent ischemic damage in the heart at risk of suffering from subsequent ischemic stress.

Methods
Animal Models. Generation of the VCP TG mouse. A construct harboring a 2.4 Kb coding sequence of VCP was generated, in which the transgene expression was under the control of the cardiac-specific promoter of the α -myosin heavy chain (α MHC). The plasmid was digested with Bam HI, and introduced by pronuclear microinjection in zygotes of FVB mice, as described previously 53 . Positive mice were mated with wild type, and their offspring were screened. Culture of RNCMs. RNCMs were prepared from Sprague-Dawley rat pups (Charles River Laboratories, Wilmington, MA) as described previously 15,18,49,54 . Recombinant adenoviruses (Ad) harboring the full-length VCP sequence (Ad-VCP) or β -Gal control (Ad-β -Gal) was generated with the AdEasy XL adenoviral vector system (Agilent, Santa Clara, CA). Myocytes were infected with Ad-VCP and Ad-β -Gal for 48 hours after 24 hours of serum-free starvation. Inhibition of iNOS was initiated 24 hours before the collection of the cells upon addition of 100 µ mol/L 1400 W (Sigma-Aldrich, St. Louis, MO) 15,18 . Echocardiography. Cardiac function and LV structure were measured in the VCP TG mice and WT littermates by two-dimensional echocardiography as described previously 49 . After determination of body weight, mice were anesthetized with 2% isoflurane (JD Medical, AZ). Echocardiography was performed using a Logiq E vet with a 13-MHz probe (12L-RS). Heart rate, LVEDD and LVESD, wall thickness, and contractility (EF and FS) were measured. Surgical procedure. IR was performed on both VCP TG and WT mice and the infarct area and the area at risk were measured as described previously 55 . Briefly, mice were anesthetized with ketamine 65 mg/kg, xylazine 1.2 mg/kg, and acepromazine 2.17 mg/kg intraperitoneally and ventilated via tracheal intubation connected to a rodent ventilator with a tidal volume of 0.2 ml and a respiratory rate of 110 breaths per minute. The heart was exposed through a thoracotomy and the left anterior descending (LAD) artery was located and was occluded for a period of 45 min to induce ischemia followed by 24 hours of reperfusion. At the end of each experiment, the LAD was re-ligated to determine the area at risk by low-pressure retrograde injection of 0.5 mg/ml Alcian blue through the aorta. The infarct size was measured following incubation in 1% Triphenyl Tetrazolium Chloride (TTC) at 37 °C for 15 min. Image-Pro software was used to measure and analyze of the infarct area and the area at risk from each section.
Histopathology. Hearts form both VCP TG and WT were collected and weighed. The LV/BW, LV/TL and LW/TL ratios were measured. Heart samples were fixed in 10% formalin and cut into 7-µ m-thick sections. Collagen accumulation, myocyte cross-sectional area and myocyte apoptosis were measured as described previously 49 . qPCR and Western blotting. mRNA was extracted from mouse heart tissues and qPCR was performed as described previously 49,54 . Protein extraction and subcellular fractions were performed as described previously 49,54 . Targeted proteins including VCP, eNOS and iNOS were detected by western blotting as previously described 15,18 . The primary antibodies used for western blotting were anti-VCP (catalog no. 2648S, rabbit, dilution of 1:1000, Cell Signaling Technology), anti-iNOS (catalog no. ab136918, rabbit, dilution of 1:1000, Abcam), and anti-eNOS (catalog no. Sc-654, rabbit, dilution of 1:1000, Santa Cruz Biotechnology). Anti-GAPDH (Catalog no. ab9484, mouse, dilution of 1:10,000, Abcam) and anti-voltage-dependent anion-selective channel protein 1(VDAC) (Catalog no. 4866, rabbit, dilution of 1:5,000, Cell Signaling Technology) were used as loading controls.

iNOS activity assay. iNOS activity was measured with the Nitric Oxide (NO) Assay Kit (Oxford Biomedical
Research, Oxford, UK) as described previously 18 . Mitochondrial isolation. Mitochondria were isolated from mouse heart as described previously 15,18,49 .
Oxygen consumption rates (OCR) were also measured in intact RNCMs with a Clark-type electrode as described previously 18 . Briefly, RNCMs infected with Ad-β -Gal and Ad-VCP were collected and suspended in extracellular buffer as described 18,57 . Glucose was used as a substrate. OCR was determined after addition of 6 µ mol/L oligomycin or 5 µ mol/ FCCP. The ratio of FCCP-stimulated to oligomycin-inhibited OCR in the myocytes was calculated 18 .
Scientific RepoRts | 7:46324 | DOI: 10.1038/srep46324 Mitochondrial swelling assays. Mitochondria were isolated as described above. mPTP opening was induced by calcium loading via addition of 600 µ M of CaCl 2 . The mitochondrial swelling was monitored by the decrease in light-scattering at 540 nm in a spectrophotometer (Thermo Scientific Multiskan GO) every 6 second for 10 minutes in the presence or absence of 10 µ M Cyclosporin A 58 . mPTP opening assay in intact cardiomyocytes. mPTP opening was determined by analyzing the mitochondrial calcein leak as previously described 59,60 . Briefly, RMCMs were loaded with 1 µ mol/L calcein-AM and 1 mmol/L CoCl2 for 30 min at room temperature, and calcein fluorescence quenching was measured. mPTP opening was evaluated after addition of 1 µ mol/L or 10 µ mol/L FCCP, which uncouples the mitochondrial respiratory chain, and depolarizes mitochondrial membrane potential. Opening of mPTP promotes mitochondrial calcein quenching. During the whole experiments, the loaded cells were excited at 470 nm and the emitted light was collected at 510 nm. The mPTP opening was indicated by a reduction in mitochondrial calcein fluorescence signal.
Statistical analysis. Results are the mean ± SEM for the number of samples indicated in the figure legends.
A one-way ANOVA was used, and a Student-Newman-Keuls post hoc correction was applied for multi-group comparisons. A value of p < 0.05 was considered significant.