Polydatin (PD), a resveratrol (RES) glycoside, has a stronger antioxidative effect than RES. It is known that RES is an autophagic enhancer and exerts a cardioprotective effect against ischaemia/reperfusion (I/R) injury. However, the effect of PD post-treatment on myocardial I/R injury remains unclear. In the present study, we investigated the influences of PD post-treatment on myocardial I/R injury and autophagy. C57BL/6 mice underwent left coronary artery (LCA) occlusion and cultured neonatal rat cardiomyocytes (NRCs) subjected to hypoxia were treated with vehicle or PD during reperfusion or re-oxygenation. We noted that PD enhanced autophagy and decreased apoptosis during I/R or hypoxia/reoxygenation (H/R), and this effect was antagonized by co-treatment with adenovirus carrying short hairpin RNA for Beclin 1 and 3-methyladenine (3-MA), an autophagic inhibitor. Compared with vehicle-treated mice, PD-treated mice had a significantly smaller myocardial infarct size (IS) and a higher left ventricular fractional shortening (LVFS) and ejection fraction (EF), whereas these effects were partly reversed by 3-MA. Furthermore, in the PD-treated NRCs, tandem fluorescent mRFP-GFP-LC3 assay showed abundant clearance of autophagosomes with an enhanced autophagic flux, and co-treatment with Bafilomycin A1 (Baf), a lysosomal inhibitor, indicated that PD promoted the degradation of autolysosome. In addition, PD post-treatment reduced mitochondrial membrane potential and cellular reactive oxygen species (ROS) production in NRCs, and these effects were partially blocked by Baf. These findings indicate that PD post-treatment limits myocardial I/R injury by promoting autophagic flux to clear damaged mitochondria to reduce ROS and cell death.
PD pre-treatment has cardioprotective effects, but the effects of PD post-treatment and its underlying mechanisms during I/R are unknown.
The present paper provides evidence that PD post-treatment promotes autophagic flux to clear damaged mitochondria to reduce ROS and cell death, and thus limits IS.
The findings in the present study suggest that PD post-treatment is likely to be an alternative strategy to attenuate myocardial I/R injury in patients with MI.
Acute myocardial infarction (AMI) is one of the main causes of morbidity and mortality in coronary heart disease. Timely reperfusion is required to prevent cardiomyocyte loss and limit infarct size (IS), but the accompanying ischaemia/reperfusion (I/R) injury is associated with poor prognoses and causes mitochondrial oxidative stress and cell death [1,2]. Both ischaemic pre- and post-conditioning can confer cardioprotection against I/R injury , although post-conditioning is a more attractive approach which is possible for clinical manipulation. Pharmacological intervention at the onset of reperfusion is feasible to mimic ischaemic post-conditioning.
Polydatin (PD) and resveratrol (RES) are both natural monocrystalline compounds extracted from Polygonum cuspidatum. The difference between PD (3,4,5-trihydroxystilbene-3-β-mon-D-glucoside) and RES (3,4,5-trihydroxystilbene) is the substitution of a glucoside group at the position C-3 of PD instead of a hydroxy group  (Figure 1A). Similar to RES, PD exerts multiple pharmacological effects, such as antioxidant and anti-inflammatory activity [4,5], and alleviation of pressure overload-induced cardiac remodelling [6,7]. Further, PD has more potent antioxidant effects than RES due to its specialized biological properties resulting from the conformational difference from RES [4,8]. Previous studies indicate that RES and PD have cardioprotective effects if administrated before ischaemia [9,10], however it remains unclear whether PD post-treatment can potentially protect against myocardial I/R injury.
Effects of PD on cardiomyocyte apotosis and signatures of autophagy and apoptosis
Autophagy is a controlled lysosomal-dependent catabolic process, which is involved in the degradation of long-lived proteins, as well as removing excess or damaged organelles such as mitochondria . Previous studies indicate that RES pre-treatment increased cardiomyocyte survival via autophagic induction , whereas the role of PD post-treatment on autophagy during myocardial I/R is unknown. Considering the close association between oxidative stress and autophagy  and the powerful antioxidative property of PD, we hypothesized that PD treatment during reperfusion would attenuate I/R injury by enhancing autophagy.
The present study aimed to assess the effects of PD post-treatment on myocardial I/R injury and to clarify its underlying mechanisms involving autophagy.
MATERIALS AND METHODS
I/R mouse model
All experiments were performed in accordance with the Guidelines for the Care and Use of Laboratory Animals (National Institutes of Health publication, No. 86-23, revised 1996) and were approved by the Bioethics Committee of the Southern Medical University. All animals used received ethical and humane care and were housed under standard conditions with a light/dark cycle of 12 h and free access to food and water. Adult male wild-type C57BL/6 mice (6–8 weeks; 20–25 g) were provided by the Laboratory Animal Center of Guangdong Province. Mice were anaesthetized with a mixture of 5 mg/kg xylazine and 100 mg/kg ketamine by intraperitoneal injection (IP) and the adequacy of anaesthesia was monitored by the disappearance of the pedal withdrawal reflex . After incubation with PE-90 tubing and ventilation using a mouse miniventilator with air, surgical procedures were initiated and the left coronary artery (LCA) was ligated (no ligation for sham) for 30 min followed by 120 min of reperfusion as previously described . The drug was directly dissolved in 1 ml normal saline (NS) at 37°C as described elsewhere  and administrated intraperitoneally (IP). The same amount of NS was administrated in sham mice. At the indicated time points, the mice were killed by overdose anaesthesia with pentobarbital sodium (150 mg/kg, IP) and cervical dislocation, and the hearts were extracted for further analysis.
PD used in the present study is an injection developed by Neptunus Co. which can be applied after dilution with NS. 3-Methyladenine (3-MA) was purchased from Sigma–Aldrich.
We designed the following groups: sham group; I/R+NS group; I/R+PD group; I/R+PD+3-MA group (15 mg/kg of 3-MA); I/R+3-MA group.
Cardiomyocyte culture and cell viability assay
Neonatal rat cardiomyocytes (NRCs) were isolated from 1 to 3 day Sprague–Dawley rats and cultured as previously described . To mimic an in vivo I/R model, hypoxia/reoxygenation (H/R) treatments were used. Cells were washed with PBS and incubated in complete-medium or starvation medium (lacking amino acids and serum) for 3 h at 37°C in a hypoxic chamber (Modular Incubator Chamber) equilibrated with 95% N2 and 5% CO2. Reoxygenation was then initiated by buffer exchange to normoxic complete medium alone or supplemented with 1, 10 or 100 μM PD.
Cell viability was measured by MTS assay (Promega) according to the manufacturer's instructions. Briefly, after drug administration, NRCs were treated with MTS and incubated for 4 h at 37°C in the dark. Absorbance was then measured at 490 nm using a microplate reader (Bio-Rad Laboratories) and readings were normalized with vehicle control. NRCs were exposed to 3 h hypoxia/3 h reoxygenation with or without PD, 3-MA (10 mM), Bafilomycin A1 (Baf, 100 nM) (Sigma–Aldrich) and MnTMPyP (50 μM, Merck Millpore). In some experiments, NRCs were pretreated with 10 mM N-acetylcysteine (NAC) (Sigma–Aldrich) for 1 h before exposure to H/R. Cells were harvested for analysis after 3 h reoxygenation.
Apoptosis was analysed using TUNEL (Invitrogen) assay according to the manufacturer's instructions. Apoptotic nucleuses were visualized with light microscopy or fluorescent microscopy. TUNEL-stained cell (%) was calculated according to the distribution of myocardial cells under microscopy (×100), five views were chosen in each section, and 200 cells were counted in each view, then the average percentage of apoptotic cell was calculated as the apoptosis index.
Samples were obtained from NRCs or mouse hearts. Protein concentrations were measured via a BCA method (Promega), and equal amounts of protein were resolved via SDS/PAGE and transferred to PVDF membranes (Millipore). After blocking in 5% non-fat milk at room temperature for 2 h, membranes were incubated with the following primary antibodies overnight at 4°C: anti-cleaved caspase-3 (Asp175), anti-LC3B, anti-p62 (all from Cell Signaling Technology) and anti-GAPDH (glyceraldehyde 3-phosphate dehydrogenase) (Kangchen). All dilutions were 1:1000. After incubation for 2 h at room temperature with secondary antibodies (Boster), the bands were visualized using enhanced chemiluminescence (Millipore). The blots were quantified by densitometry using Image Pro-Plus 6.0 software (Media Cybernetic) and the relative protein expression was compared with GAPDH.
Small fragments of myocardium sized ∼1 mm3 and NRCs were fixed overnight by immersion in 2.5% glutaraldehyde, 0.01% picric acid, 0.1 M cacodylate buffer, pH 7.4. After rinsing in the same buffer, the tissues were immersed in 1% osmium tetroxide in 0.1 M cacodylate buffer for 1 h followed by block incubation with 2% aqueous uranyl acetate for 2 h. The samples were dehydrated with a graded series of ethanol (50%, 70%, 90%, 100%), propylene oxide and then infiltrated with a 1:1 mixture of propylene oxide and EMbed 812 (Electron Microscopy Sciences). Ultrathin sections (75–80 nm) were cut with ultramicrotome, collected on 200-mesh copper grids and contrasted with in 5% uranyl acetate in ethanol (10 min) and lead citrate (5 min). Grids were examined with a Philips CM 10 electron microscope operated at 80 kV.
Myocardial infarct size measurement
2,3,5-Triphenyltetrazolium chloride (TTC) and Evans Blue dye were used to detect myocardial IS as previously described  with minor modifications. After reperfusion, the ligature was re-tightened and 0.25% Evans Blue was injected to stain the normally perfused region of the heart. The heart was then frozen and cut into 2-mm slices, which were stained with TTC at 37°C for 20 min. The area stained with Evans Blue represented the non-I/R myocardium, whereas the unstained area was the I/R myocardium, or area at risk (AAR). Within the AAR, I/R but viable myocardium stained brick red by TTC, whereas dead myocardium (infarct) was white. The IS was calculated as IS/AAR (%). For sham-operated hearts, the thread was passed around the LCA but left untied during I/R period. Prior to detecting the AAR, the LCA was occluded before Evans Blue infusion after thoracotomy and threading. Therefore, the Evans Blue does not enter the LCA perfused myocardium. Then TTC staining was used to evaluate the IS in sham-operated hearts. Images were captured using a camera and the area of the infarcted myocardium was analysed with Image Pro-plus 6.0 software.
Cardiac function was evaluated by non-invasive transthoracic echocardiography using a Visual Sonics Vevo 770 system with a RMV707B probe (Visual Sonics) 7 days after surgery. Mice were anaesthetized using 2% isoflurane, and the left chest was denuded. M-mode echocardiographic views of the mid-ventricular short axis were obtained at the level of the papillary muscle tips below the mitral valve . The left ventricular end-diastolic (Did) and systolic (Dis) diameter, end-diastolic (Vid) and systolic dimension (Vis) were measured. Left ventricular fractional shortening (LVFS) and ejection fraction (EF) were calculated as follows:
Adenovirus-short-hairpin RNA of Beclin 1(Ad-sh-Beclin 1) infection
Ad-sh-Beclin 1 and Ad-negative control (Ad-NC) were generated by a professional company (Vigene Biosciences). NRCs were cultured for 3 days, then directly transfected with the Ad-sh-Beclin 1 or Ad-NC . Multiplicity of infection (MOI) was 10 and the transduction efficiency exceeded 90%.
Evaluation of fluorescent LC3 puncta
NRCs cultured on coverslips were transfected with adenovirus of tandem fluorescent mRFP-GFP-LC3 (MOI=15). Twenty-four hours after adenoviral transfection, cells were washed with PBS, fixed with 4% paraformaldehyde, mounted with a reagent containing DAPI (Sigma–Aldrich), and viewed under a confocal laser scanning microscope (OLYMPUS FV1000). Punctuate localization of LC3 on autophagosomes had both red and green fluorescence and appeared yellow in merged images due to autophagosome induction. Subsequently, GFP instability in the acidic lysosomal compartment leads to loss of the green fluorescent signal on fused autophagosome–lysosomes and formation of autolysosomes . Therefore, red dots that do not overlay green dots and appear red in merged images indicate autolysosome formation. The transfection efficiency was more than 90%, and the subsequent transfection-induced cell death was less than 10%.
Measurement of mitochondrial membrane potential (ΔΨm)
ΔΨm was estimated using a commercial kit (Beyotime) as described elsewhere . Briefly, after treatment, cells were incubated with the JC-1 staining solution (5 μg/ml) for 20 min at 37°C, and then washed three times with JC-1 staining buffer and assayed by confocal laser scanning microscopy (OLYMPUS FV1000, λex 488 nm; λem 530 and 590 nm). ΔΨm was determined by the red/green ratio. Rhodamine 123 (Sigma–Aldrich) also was used for ΔΨm detection in accordance with the manufacturer's instructions. Briefly, NRCs were washed and incubated with Rhodamine 123 (1 μM) for 30 min after experimental manipulations, then they were trypsinized and flow cytometry was used to assay cells with emission in the FL1 channel (BD, FACS Calibur). FlowJo software was used for analysis.
Intracellular ROS (reactive oxygen species) measurements
Intracellular ROS was quantified with 2,7-dichlorofluoresein diacetate (DCFH-DA) (Sigma–Aldrich) . Confocal laser scanning microscope (OLYMPUS FV1000) and flow cytometer (BD, FACS Calibur) were used for assays. For fluorescent microscopy analysis, NRCs were seeded in laser confocal petri dishes, incubated with culture medium containing 20 μM DCFH-DA for 30 min at 37°C and washed three times with PBS. Then, cells were visualized microscopically (λex 488 nm, λem 525 nm). For flow cytometry, NRCs were cultured in 6-well flat-bottom tissue culture plates, treated and incubated with 20 μM DCFH-DA for 30 min at 37°C, then washed and trypsinized and subjected to flow cytometry with emission in the FL1 channel. FlowJo software was used for analysis. NRCs treated with 150 μM hydrogen peroxide (H2O2) for 2 h was used as positive control .
Measurement of hydrogen peroxide
H2O2 was determined with an Amplex red hydrogen peroxide assay kit (Molecular Probes) according to the manufacturer's instructions. Briefly, the harvested NRCs after treatments were lysed by three cycles of freeze–thawing . The supernatant collected after centrifugation was reacted with Amplex red (100 μM) and horseradish peroxidase (0.2 unit/ml) in 96-well plates for 30 min at room temperature. The absorbance was determined spectrophotometrically (Molecular Devices) at 560 nm. The results were expressed as micromoles per gram of protein.
Quantification of superoxide anion release
Superoxide anion production was determined by using the superoxide dismutase-inhibitable (SOD-inhibitable) cytochrome c reduction assay as previously described . Briefly, 500 μl culture media were mixed with 50 μl cytochrome c (40 μM, Sigma–Aldrich), followed by adding 250 μl Hank's balanced salt solution. The mixtures were incubated at room temperature for 10 min with or without 50 μl SOD (100 μg/ml, Sigma–Aldrich). The absorbance was measured spectrophotometrically at 550 nm. The molar extinction coefficient for reduced cytochrome c is 21000 mM/cm and the calculated results were expressed as micromoles per gram of protein.
All analyses were performed using SPSS 17.0 software. All data were expressed as means ± S.E.M. and P<0.05 was considered to be statistically significant. Statistical differences were evaluated using one-way ANOVA followed by Bonferroni's or Dunnett's T3 multiple comparison exact probability tests.
Effects of PD on cardiomyocyte apotosis and signatures of autophagy and apoptosis
In cultured NRCs exposed to H/R, PD (1, 10, 100 μM) dose-dependently improved cell survival, but there was no significant difference between 10 and 100 μM groups (Figure 1B). Thus, 10 μM of PD was applied in subsequent in vitro experiments. In mice insulted with I/R, TUNEL assay showed that PD (5, 7.5, 10 mg/kg) dose-dependently decreased apoptosis, but there was no significant difference between the 7.5 and 10 mg/kg groups (Figures 1C and 1D). Therefore, the concentration of 7.5 mg/kg was chosen for subsequent in vivo experiments.
PD-induced autophagy decreased myocardial apoptosis
in vivo and in vitro
In mice subjected to I/R, PD increased LC3 II/I and LC3 II and decreased cleaved caspase-3, and these effects were abrogated by co-treatment with 3-MA, that inhibits class III PI3K activity  (Figure 2A). Similar results also were noted in NRCs exposed to H/R (Supplementary Figure S1A). Next, Ad-sh-Beclin 1 was applied to NRCs. After 24 h infection, the transduction efficiency was over 90% (Figure 2B). Beclin 1 was reduced by approximately 70% in the Ad-sh-Beclin 1 treated cardiomyocytes (Figure 2C). In NRCs exposed to H/R, co-treatment with Ad-sh-Beclin 1 led to a significantly lower LC3 II/I protein ratio and LC3 II/GAPDH in H/R group and H/R+PD group respectively (Figures 2D and 2E). MTS assay indicated that Ad-sh-Beclin 1 decreased the viability of NRCs with or without H/R treatment, whereas the effect of PD reducing cell loss response to H/R was abolished by co-treatment with Ad-sh-Beclin 1 (Figure 2F). TUNEL assay revealed that cell death was increased in the H/R group compared with control, and the effect of PD on decreasing cell death was partly blocked by 3-MA (Supplementary Figures S1B and S1C).
PD-induced autophagy decreases myocardial apoptosis
in vivo and in vitro
PD reduced myocardial infarct size and improved heart function in I/R mice
To examine the efficacy of PD post-treatment on I/R injury, myocardial cell ultrastructure was examined by TEM. Numerous dense and tight mitochondria, and neatly arranged and intact myocardial fibrils could be seen in the cytoplasm of sham group. In contrast, I/R-treated group had loose and swollen mitochondria with ruptured or disappearing swollen myocardial fibrils (Figure 3A). Cardiomyocytes in PD treated mice had less pathological changes: less ruptured mitochondria and myocardial fibrils, and more highly visible vacuoles with/without content, whereas I/R-induced mitochondrial swelling was alleviated by PD as evidenced by decreased mitochondrial mean area (Figure 3A).
Effects of PD post-treatment on mitochondrial damage, infarct size and heart function in mice subjected to I/R
The IS in I/R+PD group was significantly reduced in comparison with I/R group (59±4.5% compared with 30±4.0%; Figures 3B and 3C), whereas co-treatment with 3-MA abrogated the effect of PD on reducing myocardial IS (42±2.8% compared with 30±4.0%) (Figures 3B and 3C). There were no significant differences in AAR/LV among the groups (Figures 3B and 3D).
Two D- and M-mode of echocardiographic examinations were performed (Figures 3E and 3F). LVFS and EF were significantly lower in I/R group than in the sham group, which were improved by PD post-treatment, whereas autophagic inhibitor 3-MA blunts this beneficial effect by PD post-treatment (Figures 3G and 3H).
PD enhanced autophagic flux in NRCs exposed to H/R
To monitor autophagic flux, tandem fluorescent mRFP-GFP-LC3 was performed on NRCs (Ad-LC3-NRCs). The normal Ad-LC3-NRCs had basal autophagy with few autolysosomes and few autophagosomes (Figures 4A and 4B). Ad-LC3-NRCs subjected to H/R had accumulated autophagosomes and few autolysosomes (Figures 4A and 4B), suggesting that autophagosome clearance was inhibited and autophagic flux was blocked or impaired during myocardial H/R. In PD-treated group, Ad-LC3-NRCs subjected to H/R had more autolysosomes and less autophagosomes than in untreated group (Figures 4A and 4B), indicating PD treatment induces the consumption of autophagosomes and enhanced autophagic flux. However, co-treatment with PD and autophagic inhibitor 3-MA decreased autolysosomes compared with PD. In addition, data showed that the ratio of LC3 II/I and LC3 II had no significant difference between H/R groups in the presence and absence of Baf (a lysosomal inhibitor), whereas co-treatment with Baf and PD induced the accumulation of LC3 II in NRCs exposed to H/R (Figure 4C), indicating that PD promoted the degradation of autolysosome. Thus, autophagosomes accumulated may be due to impaired autophagy during H/R, and PD treatment induced a significant clearance of autophagosomes with the degradation of autolysosome. TUNEL assay indicated that PD post-treatment decreased the apoptosis of NRCs exposed to H/R, whereas Baf partly abolished this effect of PD (Figure 4D).
Effects of PD on autophagic flux in cultured NRCs
PD-induced autophagic flux attenuated mitochondrial damage
TEM examination showed that NRCs in H/R+PD group produced more autolysosomes than in H/R group, indicating that PD can lead a shift from early autophagic vacuoles to late autolysosomes, whereas the effect of PD was partly abolished by 3-MA (Figure 5A). ΔΨm was dramatically reduced in NRCs exposed to H/R, and partially restored by PD. However, the effect of PD was abolished by Baf (Figures 5B and 5C).
Effects of PD treatment on mitochondrial function in NRCs exposed to H/R
PD-induced autophagic flux reduced cellular ROS
As damaged mitochondria are the major source of reactive oxygen species (ROS) production, we investigated effects of PD on cellular ROS production in NRCs. Exposing NRCs to H/R significantly increased nonspecific ROS production, and PD markedly inhibited ROS production (Figures 6A, 6B and Supplementary Figure S3). However, the inhibitory effect of PD on ROS generation was blocked by Baf (Figures 6A, 6B and Supplementary Figure S3).
Effects of PD on production of cellular ROS
In addition, to confirm whether PD post-treatment had effects on H2O2, NRCs were quantified using Amplex Red assay. H2O2 was significantly increased in the H/R group compared with control, and the effect of PD on decreasing H2O2 was partly eliminated by Baf (Figure 6C). Similar results were also found by quantifying with SOD-inhibitable cytochrome c reduction assay to detect superoxide anion (Figure 6D). These results showed that PD-induced autophagic flux reduced the generation of H2O2 and superoxide anions.
The present study indicates that PD treatment during reperfusion exerts significant protective effects against myocardial I/R injury by enhancing autophagic flux. We illustrated the major findings in Figure 7. PD post-treatment promotes autophagic flux to clear damaged mitochondria to reduce ROS and cell death, and thus limits myocardial IS.
Illustration of actions and mechanisms of PD on myocardial I/R injury
Previous studies suggest that autophagy is increased in mouse heart during both ischaemia and reperfusion [23,24], whereas the effect of autophagy on myocardial I/R injury is controversial and may be context-dependent. Some researchers have proposed that whether up-regulation of autophagy is beneficial or detrimental may depend on the extent of autophagy. Moderate autophagy is reported to be beneficial for cardiomyocyte survival but excessive autophagy exacerbates cardiomyocyte death [13,25]. Sadoshima's group reported that autophagy was amplified during reperfusion and exacerbated cardiomyocyte death after ischaemia, suggesting that excessive autophagy was detrimental to the heart . In our study, the protective effects of PD may be attributed to enhanced autophagic flux. Our tandem fluorescent mRFP-GFP-LC3 results indicate that autophagosomes were accumulated and autophagic flux was impaired during H/R, whereas PD induced autophagosome clearance. Consistent with our data, a study revealed that autophagic flux was impaired at the phase of autophagosome–lysosome fusion as well as autolysosome degradation during reperfusion, which did not show protective effects against myocardial cell death during I/R injury . This phenomenon highlighted the importance of identifying ‘intact autophagic flux’ to accurately characterize autophagy, which is a dynamic process.
Given that undigested autophagosome and damaged mitochondria may release harmful substances such as ROS to aggravate cell death , elimination of damaged mitochondria and autophagosomes would prevent further injury to neighbouring mitochondria and the wholesale opening of mitochondrial permeability transition pore . Our results indicate that ΔΨm was decreased and ROS was increased in myocardial cells exposed to H/R, whereas ΔΨm was restored and ROS production was inhibited by PD post-treatment. Enhanced myocardial cell viability by PD may result from fewer damaged mitochondria and production of less ROS.
PD exerts antioxidant effects by increasing SOD expression which is an antioxidant enzyme . The partial restoration of ROS production after co-treatment with autophagic inhibitor and PD suggested that the protective effects of PD were associated with scavenging ROS. Our results have confirmed that PD reduced the generation of H2O2 and superoxide anions. This is consistent with a recent study reporting that PD pre-treatment exerted cardioprotective effects via eliminating free oxidative radical .
Inflammation plays a significant role in myocardial I/R injury  and it is involved not only in acute tissue damage, but also in subsequent tissue repair and healing . Monocyte/macrophage infiltration after myocardial infarction has been confirmed, suggesting that inhibition of excessive inflammation could limit IS and strengthen heart function . Studies have confirmed that PD has anti-inflammatory effects via inhibiting expression of interleukin-17  and various cell adhesion molecules . Our data revealed that PD decreased polymorphonuclear leukocytes (PMNs) in mice exposed to I/R (Supplementary Figure S2, Supplementary Materials). Thus, the cardioprotection of PD found in the present study may depend on multiple mechanisms, including scavenging ROS, eliminating damaged mitochondria, promoting autophagosome clearance and restricting inflammation.
Apoptosis may be indirectly regulated by autophagy via alleviating ROS production. Two major mechanisms activate caspases: the extrinsic pathway which activates caspase-8 and the intrinsic pathway or the mitochondrial-dependent mechanism which activates caspase-9. Both pathways activate executioner caspase-3. Kamata et al.  reported that reduction of ROS accumulation deactivated the extrinsic pathway and promoted cell survival by transiently activating JNK. In regard to the mitochondria-dependent mechanism, pro-apoptotic Bcl-2 family proteins cause transient mitochondrial permeability transition pore opening, which promotes ROS accumulation  and Bax- or Bak-mediated outer mitochondrial membrane permeabilization along with subsequent generation of ROS . Consequently, it facilitates the release of mitochondrial-resident apoptogenic factors (such as cytochrome c)  and these activities ultimately lead to apoptotic cell death. As cytochrome c triggers a post-mitochondrial pathway forming an ‘apoptosome’ of cytochrome c, Apaf-1 and caspase-9, which subsequently cleaves the effector caspases-3 and -7 [35,36], the deficiencies of cytochrome c display defects in apoptosis following intrinsic signals . Therefore, promoting autophagic flux and antioxidative action by PD to reduce ROS production may decrease apoptosis through the above-mentioned mechanisms.
The smooth process of autophagic flux accompanied by the attenuation of apoptotic cell death after PD treatment and the up-regulation of apoptosis induced by autophagic inhibitor highlight the role of autophagy in apoptotic regulation. It was reported that inhibiting autophagy triggered apoptosis , and that the up-regulation of autophagy protected cardiac tissue against apoptotic stimuli . Furthermore, autophagy was reported not only to degrade damaged mitochondria and cleave caspases such as caspase-3, but also to provide a membrane-based intracellular platform to process caspases during apoptotic regulation . Our findings suggest that up-regulation of autophagy by PD is accompanied with reduced apoptosis and the administration of autophagic inhibitor disturbs optimal autophagy and leads to apoptosis. In addition, a previous study revealed that autophagy and apoptosis could be cooperative or competitive pathways depending on the cellular environment . However, the mechanisms of cross-talk between autophagy and apoptosis remain unclear and should be explored in future studies.
The present study was conceived and designed by Aihua Chen, Yuanna Ling, Xianbao Wang and Guiming Chen. Data and results were analysed and interpreted by Yuanna Ling, Aihua Chen, Xianbao Wang, Yulin Liao, Masafumi Kitakaze, Guiming Chen, Yi Deng, Huixiong Tang, Long Ling, Xiaoming Zhou, Xudong Song, Pingzhen Yang, Yingfeng Liu, Zhiliang Li, Cong Zhao and Yufei Yang. Experiments were performed by Yuanna Ling, Guiming Chen, Xianbao Wang and Yi Deng. The manuscript was written, edited and revised by Yuanna Ling, Yulin Liao, Aihua Chen, Masafumi Kitakaze, Xianbao Wang and Guiming Chen.
We are grateful to Dr Yan Wang (Institute of Regenerative Medicine, Zhujiang Hospital, Southern Medical University, China) for generosity in sharing valuable technical assistance and laboratory equipment and Professor Kesen Zhao (Department of pathophysiology, Southern Medical University, China) for the generous gift of PD.
This work was supported by the National Natural Science Foundation of China [grant numbers 81270218 (to A.C.) and 81400190 to (X.W.)].
area at risk
left coronary artery
left ventricular fractional shortening
multiplicity of infection
neonatal rat cardiomyocytes
reactive oxygen species
These authors contributed equally to this article.