Poly (ADP‐ribose) polymerase inhibition protects against myocardial ischaemia/reperfusion injury via suppressing mitophagy

Abstract Myocardial ischaemia/reperfusion (I/R) injury attenuates the beneficial effects of reperfusion therapy. Poly(ADP‐ribose) polymerase (PARP) is overactivated during myocardial I/R injury. Mitophagy plays a critical role in the development of myocardial I/R injury. However, the effect of PARP activation on mitophagy in cardiomyocytes is unknown. In this study, we found that I/R induced PARP activation and mitophagy in mouse hearts. Poly(ADP‐ribose) polymerase inhibition reduced the infarct size and suppressed mitophagy after myocardial I/R injury. In vitro, hypoxia/reoxygenation (H/R) activated PARP, promoted mitophagy and induced cell apoptosis in cardiomyocytes. Poly(ADP‐ribose) polymerase inhibition suppressed H/R‐induced mitophagy and cell apoptosis. Parkin knockdown with lentivirus vectors inhibited mitophagy and prevented cell apoptosis in H/R‐treated cells. Poly(ADP‐ribose) polymerase inhibition prevented the loss of the mitochondrial membrane potential (ΔΨm). Cyclosporin A maintained ΔΨm and suppressed mitophagy but FCCP reduced the effect of PARP inhibition on ΔΨm and promoted mitophagy, indicating the critical role of ΔΨm in H/R‐induced mitophagy. Furthermore, reactive oxygen species (ROS) and poly(ADP‐ribosylation) of CypD and TSPO might contribute to the regulation of ΔΨm by PARP. Our findings thus suggest that PARP inhibition protects against I/R‐induced cell apoptosis by suppressing excessive mitophagy via the ΔΨm/Parkin pathway.


| INTRODUC TI ON
Reperfusion therapy is the most effective treatment for acute myocardial infarction, reduces ischaemic injury and limits the infarct size.
However, reperfusion can independently induce myocardial injury including cardiomyocyte death, leading to expansion on the scope of myocardial infarction. 1 Myocardial ischaemia/reperfusion (I//R) injury may account for up to 50% of the infarct size. 1 Despite great effort to optimize reperfusion conditions, the means of reducing I/R injury-associated cell death are very limited. 2 Several critical factors mediate the detrimental effects of I/R-induced injury, including oxidative stress, intracellular Ca 2+ overload, inflammation and mitochondrial permeability transition pore (MPTP) opening. 1,2 The opening of the MPTP is related to poly(ADP-ribose) polymerase (PARP) activation, and PARP inhibition prevents myocardial I/R injury. 3,4 PARP is a highly conserved DNA-binding nuclear enzyme family that can be activated by DNA damage. Poly(ADP-ribose) polymerase plays an important role in the regulation of multiple physiological cellular functions including DNA repair, transcription, cell cycle, cell death and genomic integrity. 5 The activation of PARP initiates an energy-consuming cycle by transferring ADP-ribose units from nicotinamide adenine dinucleotide (NAD) to form long branches of ADP-ribose polymers (PAR) on glutamic acid residues of a number of target proteins. 6 Mitochondria generate most of the energy for the heart via oxidative phosphorylation. To maintain a healthy and functional mitochondrial network, dysfunctional or damaged mitochondria are eliminated via a process known as mitochondrial autophagy or mitophagy, which is triggered by starvation, hypoxia and reactive oxygen species (ROS). 7,8 Mitophagy has been classified into canonical and non-canonical pathways in the heart. 8 The Parkin-dependent pathway is the main form of canonical mitophagy. 9 Poly(ADP-ribose) polymerase activation has been shown to prevent mitophagy in xeroderma pigmentosum group A-deficient cells, and PARP inhibition by the inhibitor olaparib induces mitophagy in BRCA1 and BRCA2 mutant breast cancer cells. 10,11 However, whether PARP activation regulates mitophagy in I/R-injured cardiomyocytes remains unclear. This study showed that PARP inhibition attenuated I/R-or hypoxia/oxygenation (H/R)-induced mitophagy and cell apoptosis in vivo and in vitro.

| Reagents and antibodies
The Parkin-siRNA lentivirus and control lentivirus were con-

| Mice experiments
Acute myocardial I/R model was performed on adult male C57BL/6 mice (10-12 weeks) by ligating the left anterior descending artery (LAD). Mice were anaesthetized with isoflurane and ventilated using a Rodent Anesthesia Machine. After taped to a heating pad in the supine position, mice chest was opened at the third intercostal space and the heart was exposed by squeezing. A 6-0 silk suture was passed under the LAD 1-2 mm from the tip of the left atrium. Left anterior descending artery was ligated with a slipknot. The occlusion was maintained for 30 minutes and then the knot was released to reperfuse the heart for 120 minutes. 12 To determine the myocardial infarct size, hearts were collected and sectioned into 2-3 mm slices.

| Electron microscopy
Sections of myocardium or cell on coverslips were fixed in 2.5% glutaraldehyde overnight. After rinsing in 0.1 mmol/L cacodylate buffer with 1% tannic acid, the samples were immersed in 1% osmium tetroxide in 0.1 mmol/L cacodylate buffer for 1 hour. After being rinsed again, the samples were dehydrated with alcohol and embedded in Epon 812. Then, the samples were examined using a transmission electron microscope.

| Mitochondria isolation
Mitochondria from cultured H9C2 cells were isolated with the Mitochondria Isolation Kit (Thermo Scientific) referred to the manufacturer's instruction.

| Measurement of mitochondrial membrane potential (ΔΨm)
ΔΨm was determined using a commercial assay kit by incubation with JC-1(Beyotime) in serum-free medium for 20 minutes at 37°C. Then cells were washed with JC-1 staining buffer and imaged under a fluorescence microscope (Olympus). Normal mitochondria produce red fluorescence, and depolarized or inactive mitochondria produce green fluorescence. ΔΨm was calculated by the red/green fluorescence ratio.

| ROS production
Reactive oxygen species were determined with the Reactive Oxygen Species Kit (Beyotime). According to the instruction, cells were incubated with DCFH-DA for 20 minutes at 37°C. The ROS level was examined in a Thermo Fisher Varioskan Flash spectral scanning multimode reader.

| Immunoprecipitation
Poly(ADP-ribosylation) of CypD and TSPO was detected using immunoprecipitation method. 13 Lysates of H9C2 cells were prepared with NP-40 Lysis Buffer (Bosterbio). For the immunoprecipitation studies, 800 μg proteins were incubated with 10 μg PAR antibody followed by incubation with 60 μl protein A/G agarose (Santa Cruz Biotechnology) and the pellets were washed four times. Then, beads were added 40 μl 2 × SDS-PAGE loading buffer and boiled for 5 minutes to elute the immunocomplexes. Supernatants were subjected to SDS-PAGE and analysed for CypD and TSPO. Similar procedures were performed to determine the protein-protein interaction between PARP-1 and CypD or TSPO.

| NAD and ATP measurements
Nicotinamide adenine dinucleotide was measured with a NAD/ NADH Assay Kit (Beyotime), and ATP was detected with an ATP Assay Kit (Beyotime). The level of NAD and ATP was normalized to total protein content which was determined by the bicinchoninic acid method.

| Statistical analysis
All data were presented as the mean ± SEM and analysed by either one-way ANOVA or a two-tailed Student's t test. The null hypothesis was rejected at P < .05.

| PARP inhibition reduces infarct size and suppresses mitophagy in I/R-injured hearts
The role of PARP inhibition in myocardial I/R injury was investigated using the LAD ligation model. I/R-induced PARP activation and the inhibitor, DPQ, effectively inhibited PARP activity, as indicated by PAR expression ( Figure 1A). As expected, PARP inhibition limited the increase of the infarct size compared with the I/R group ( Figure 1B).
Moreover, PARP inhibition improved the cardiac function after IR injury, as indicated by LVEF ( Figure 1C)

| PARP inhibition prevents cell apoptosis and mitophagy in H/R-treated cardiomyocytes
To test the effects of PARP inhibition on cell apoptosis and mitophagy after I/R injury in vitro, we subjected H9C2 cells to hypoxia/reoxygenation (H/R) treatment. Figure 2A

| Knockdown of Parkin prevents mitophagy and cell apoptosis in H/R-treated cardiomyocytes
To determine the relationship between mitophagy and cell apoptosis in H/R-treated H9C2 cells, Parkin was knocked down with lentivirus-RNAi in H9C2 cells. Figure 3A

| PARP inhibition prevents H/R-induced mitophagy by regulating the mitochondrial membrane potential (ΔΨm)
Opening of the MPTP causes the loss of ΔΨm, which triggers PINK1/ Parkin-mediated mitophagy. 15,16 ΔΨm was obviously reduced by H/R injury, and mitochondrial membrane depolarization was restored by PARP inhibition (Figure 4A,B). To ensure the critical role of ΔΨm in mitophagy in H/R-injured H9C2 cells, we pre-incubated H9C2 cells with cyclosporin A, which is a potent inhibitor of the MPTP. Figure 4C shows that cyclosporin A prevented the disruptive effect of H/R on ΔΨm. Similar to PARP inhibition, cyclosporin A prevented mitophagy in H/R-injured H9C2 cells ( Figure 4D,E). FCCP is an oxidative phosphorylation uncoupler that depolarizes the mitochondrial membrane. 17 It was found that the effect of PARP inhibition on ΔΨm

| ROS contributes to the PARP-mediated decline in ΔΨm
Excessive ROS trigger the opening of the MPTP and activate PINK1/ Parkin-mediated mitophagy. 18 To investigate whether ROS play a role in PARP-mediated changes in ΔΨm, we measured ROS production in H/R-treated H9C2 cells. The production of ROS was significantly increased in H/R-treated H9C2 cells compared with that in control cells, and PARP inhibition partially prevented the increase in ROS production after H/R ( Figure 5A). The reductant NAC reduced the production of ROS ( Figure 5B). More importantly, we found that NAC prevented the H/R-induced decline in ΔΨm ( Figure 5C,D). Thus, ROS have an effect, at least partially on the changes in ΔΨm mediated by PARP after H/R.

| Poly-ADP-ribosylation of CypD and TSPO may also contribute to the PARP-mediated decline in ΔΨm
Poly(ADP-ribose) polymerase activation induces poly-ADP-ribosylation of mitochondrial proteins after I/R. 19 Immunoblotting indicated that more PAR was found in proteins of isolated mitochondria after H/R injury ( Figure 6A)  Figure 6B,C). However, we did not find a direct interaction between PARP-1 and CypD or TSPO ( Figure S4). These data demonstrate that PARP might regulate the opening of the MPTP by directly modifying CypD and TSPO by poly-ADP-ribosylation.

| D ISCUSS I ON
In the present study, we found that PARP inhibition prevents I/R injury-induced mitophagy and cell apoptosis in cardiomyocytes. Consistent with previous studies, 23,24 we also found that PARP inhibition prevented I/R-induced cell apoptosis in vivo and in vitro.

Mitophagy inhibition due to knockdown of Parkin decreases
Mitophagy facilitates the normal turnover of mitochondria and becomes more important during exposure to stress including I/R injury. Although the role of mitophagy in myocardial I/R injury has been discussed in several studies, it needs to be further explored because of conflicting conclusions. 25  ΔΨm. 18 = 3). B, Immunoprecipitation using a PAR antibody showed that CypD was modified by poly(ADP-ribosylation). (n = 5). C, Immunoprecipitation using a PAR antibody showed that TSPO was modified by poly(ADP-ribosylation). (n = 5) However, it is unclear whether MPTP opening is related to the poly-ADP-ribosylation of relative proteins. The protein components of the MPTP still need to be explored. CypD and TSPO are potent components that act as regulators of MPTP opening. 21,34 Down-regulation of CypD prevents MPTP opening and protects against the loss of ΔΨm. 35 Conversely, the loss function of TSPO causes MPTP opening and leads to a reduced ΔΨm. 21,36 Our results showed that both CypD and TSPO could be poly-ADP-ribosylated, but the effects of this modification on the functions of CypD and TSPO were not studied. As reported, poly-ADP-ribosylation can serve as a marker for ubiquitinproteasomal system (UPS)-dependent protein degradation. 37 Thus, we speculate that the function of CypD and TSPO might be changed by poly-ADP-ribosylation, which should be investigated in the future.
In conclusion, our study is the first to report that PARP inhi-

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest. Pan and Feng Xu analysed the data. All authors read and approved the version of the manuscript to be published.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.