Sprague-Dawley (SD) male rats (weight, 200–250 g; age, 6–8 weeks; n=50) and neonatal SD rats (age, 1–3 days; n=30) were obtained from the Animal Experimental Center of Shandong University. Protocols for the care and use of laboratory animals were approved by the Local Ethics Committee at the Medical College of Shandong University. Laboratory animals were reared according to the Guide for the Care and Use of Laboratory Animals from the National Institutes of Health (NIH) (22). At the end of animal experiments, rats were euthanized using excessive anesthesia administration as conditionally acceptable (intraperitoneal injection with sodium pentobarbital at a dosage of 200 mg/kg). Metformin was purchased from Sigma-Aldrich (Merck KGaA).

Rats were anesthetized by sodium pentobarbital (50 mg/kg) via intraperitoneal injection and then ventilated with a rodent respirator (ALCV9A; Shanghai Alcott Biotech Co., Ltd.) after intubation. Then, the heart was exposed, and the left anterior descending coronary artery (LAD) was ligated by a 6-0 silk suture for 45 min. Ischemia was determined by blanching of the myocardium, dyskinesia of the ischemic region and ST segment elevation on the ECG. Then, the heart was reperfused for 4 h by loosening the knot, and a marked hyperemic response indicated reperfusion (16).

The rats were randomly divided into three groups: i) Sham (sham operated) group, in which the heart was exposed but the LAD was not occluded; ii) ischemia-reperfusion (IR) group, in which the LAD was occluded for 45 min and reperfused for 4 h; and iii) IR + metformin (IR + met) group, in which the LAD was occluded for 45 min, and metformin (5 mg/kg) was intravenously injected via the jugular vein, subsequently followed by 4 h of reperfusion. The dose of metformin used in the present study was selected based on a previous study (16).

Staining was performed as previously described (23). At the end of the reperfusion, the LAD was ligated again, and the heart was retrogradely infused with Evans blue (0.25% in phosphate buffer; Sigma-Aldrich; Merck KGaA) from the aorta. The non-ischemic area was stained blue, indicating the area at risk (AR, non-blue region). Then, the heart was frozen at −20°C for 30 min and sectioned into 6 slices (2 mm/slice). The slices were incubated in triphenyltetrazolium chloride (TTC; 1% in phosphate buffer; Sigma-Aldrich; Merck KGaA) for 10 min at 37°C and then immersed in 4% paraformaldehyde for 4 h at room temperature. TTC staining can differentiate the infarct size (IS; white region) from the non-infarct area at risk (AR; red region). These heart slices were used only for Evans-blue-TTC staining. Finally, the slices were arranged from apex to base and digitally photographed. Digital images of the slices were analyzed using ImageJ software (version 1.47; National Institutes of Health). The final result was presented as the IS/AR as previously described (24).

The left ventricles of rat hearts were fixed in 4% paraformaldehyde at room temperature for 24 h and embedded in paraffin at 56°C for 1.5 h, then sectioned (thickness, 6 µm). According to the manufacturer's instructions, terminal deoxynucleotidyl transferase-mediated dUTP nick end labelling (TUNEL) assays were performed to detect apoptotic cells in heart tissue sections using an in situ cell death detection kit (Roche Applied Science). The sections were incubated in TUNEL reaction mixture for 60 min at 37°C in a humidified atmosphere in the dark. Sections were then stained with hematoxylin (Beijing Solarbio Science & Technology, Co., Ltd.) for 2 min at room temperature, washed with PBS, and mounted with mounting medium (Abcam). The nuclei were counted in 10 random fields of each section using a light microscope (magnification, ×400; Leica Microsystems GmbH), and the results are shown as a percentage of TUNEL-positive nuclei compared with the total number of cell nuclei.

RIPA solution (Abcam) was used to homogenize the frozen tissue samples and cultured cardiomyocytes and BCA assay (Pierce; Thermo Fisher Scientific, Inc.) was used to detect the concentration of protein samples. Then, 50 µg/lane extracted protein was separated via 10% SDS-PAGE and transferred onto PVDF membranes (EMD Millipore). Non-specific reactivity was blocked with 5% milk for 2 h at room temperature and the membrane was incubated with primary antibody overnight at 4°C in buffer (10 mM Tris-HCl; pH 7.5; 150 mM NaCl, 2% Tween-20, 4% bovine serum albumin). The membrane was then incubated with secondary antibody for 1.5 h at room temperature. ECL chemiluminescence (Cell Signaling Technology, Inc.) was used for detection by an imaging system (ChemiDoc XRS; Bio-Rad Laboratories, Inc.), and densitometry was performed using ImageJ software (version 1.47; National Institutes of Health) semi-quantitatively (4). All protein levels were normalized to that of GAPDH. The antibodies used in the study were as follows: Phosphorylated (p)-AMPK (Thr172; 1:1,000; cat. no. 50081; rabbit monoclonal antibodies; Cell Signaling Technology, Inc.), AMPK (1:1,000; cat. no. 2532; rabbit monoclonal antibodies; Cell Signaling Technology, Inc.) and NOX4 (1:1,000; cat. no. ab109225; rabbit monoclonal antibody; Abcam); GAPDH (1:10,000; cat. no. AB2302; mouse monoclonal antibody; EMD Millipore) and corresponding horseradish peroxidase (HRP)-conjugated secondary antibodies (OriGene Technologies, Inc.).

NRVMs were isolated from 1–3-day-old SD rats as previously described (25). The rats were euthanized by decapitation and the hearts were rapidly excised, minced and dissociated with 0.1% trypsin and 0.03% collagen II solution. The dispersed cells were then plated at a density of 2×10⁵ cells/cm² on a 6-well plate with 2 ml/well DMEM supplemented with 10% fetal bovine serum (FBS; both HyClone; GE Healthcare Life Sciences). Cytosine arabinoside (10 M) was used to suppress non-cardiomyocytes. After serum-starved cultivation (37°C, 95% O₂, 5% CO₂, 24 h), the myocytes were put in a hypoxia (95% N₂, 5% CO₂) incubator for 4 h at 37°C then, the culture medium was replaced with fresh oxygenated DMEM (10% FBS) and the plates were transferred to a normoxic incubator (5% CO₂) for 6 h of reoxygenation.

The NRVMs were randomly divided into four groups: i) Control (CON) group, cells were incubated with normal oxygen; ii) hypoxia-reoxygenation (HR) group, in which the cells were incubated in hypoxic conditions for 4 h and reoxygenated for 6 h; iii) HR + metformin (HR + met) group, at the onset of reoxygenation, metformin (0.1 mM) was added to the DMEM (the dose of metformin used in the present study was selected based on a previous study) (26); and iv) HR + metformin + compound-C (HR + met + compound-C), at the onset of reoxygenation, metformin and compound-C (10 nM; cat. no. ab120843; Abcam) were added into the medium simultaneously (the dose of compound-C used in the present study was selected based on a previous study) (26).

Cell viability was assessed by an MTT assay. The MTT assay was performed according to the manufacturer's protocols (cat. no. C0009; Beyotime Institute of Biotechnology). Cardiomyocytes were plated on 96-well dishes at a density of 1×10⁴ cells/well. Following reoxygenation, cells were incubated with MTT reaction solution (0.5 mg/ml) for 4 h and the medium was removed. Then, 100 µl DMSO was added to each well for 10 min and mixed thoroughly with a mechanical plate mixer. At last, the absorbance was measured with a microplate reader (Molecular Devices, LLC) at a wavelength of 530 nm.

Blood samples were centrifuged at 1,800 × g for 10 min at 4°C. LDH (cat. no. A020-1) and CK-MB (cat. no. E006-1-1) levels in plasma were detected according to the manufacturer's protocols (Nanjing Jiancheng Bioengineering Institute). The optical density of the tetrazolium product was determined spectrophotometrically (Molecular Devices, LLC) at a wavelength of 490 nm.

Lipid peroxidation was estimated by measuring the concentration of MDA in the isolated myocytes using a Lipid Peroxidation Assay kit (cat. no. MAK085; Calbiochem; Merck KGaA). Briefly, the cells were lysed and mixed with reagent R1 (N-methyl-2-phenylindole in acetonitrile and methanol) for a total of 60 min at 45°C. The samples were then centrifuged at 21,130 × g for 10 min at room temperature. The optical density of the supernatant was determined spectrophotometrically at a wavelength of 586 nm (Molecular Devices, LLC). The MDA concentration was displayed as µmol/g protein.

Tissue sections (thickness, 6 μm) were deparaffinized and then blocked with CAS-Block (Invitrogen; Thermo Fisher Scientific, Inc.) for 1 h at 37°C. The sections were incubated with anti-4HNE primary antibody (1:200; cat. no. ab48506; Abcam) for 2 h at 37°C followed by incubation with HRP-conjugated secondary antibody (1:500; cat. no. 7074; Cell Signaling Technology, Inc.) for 1 h at 37°C. Sections were developed in 3,3′-diaminobenzidine solution for 3 min at room temperature, then stained with hematoxylin for 2 min at room temperature, washed with PBS, and mounted with mounting medium. Sections were visualized under a light microscope (magnification, ×400; Leica Microsystems GmbH).

Total RNA of myocardial tissues and NRVMs was extracted using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.). The TaqMan^® Reverse Transcription Reagents (cat. no. N8080234; Applied Biosystems; Thermo Fisher Scientific, Inc.) was used to generate the first-strand cDNA using the following thermocycling conditions: 70°C for 5 min, 42°C for 1 h and 70°C for 15 min. A Mx3000 Multi-plex Quantitative PCR System (Stratagene; Agilent Technologies, Inc.) with SYBR-Green fluorescence (Molecular Probes, LLC) was used to amplify the cDNA in 35 cycles. Each cycle consisted of heating denaturation for 30 sec at 94°C, annealing for 30 sec at 56°C and extension for 30 sec at 72°C. All samples were quantitated using the comparative Cq method for relative quantitation of gene expression, normalized to GAPDH (27). The primers used in this study were as follows: NOX4, forward, 5′-TGGCCAACGAAGGGGTTAAA-3′ and reverse, 5′-CACTGAGAAGTTCAGGGCGT-3′; AMPK, forward, 5′-GATCGGACACTACGTGCTGG-3′ and reverse, 5′-TAGTTGCTCGCTTCAAGGGG-3′; and GAPDH, forward, 5′-TGATGACATCAAGAAGGTGGTGAAG-3′ and reverse, 5′- TCCTTGGAGGCCATGTAGGCCAT-3′.

Transfection of siRNA was performed with Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocols. NRVM (6×10⁵ cells per well of a 6-well plate at) were transfected with siRNA directed to AMPK (cat. no. sc-74461; Santa Cruz Biotechnology, Inc.) at 10 µM for 6 h in serum-free medium. Afterwards, it was changed to standard medium containing 10% FBS for 24 h. Non-specific (NS)-siRNA (cat. no. sc-37007; Santa Cruz Biotechnology, Inc.) was used as negative control.

The NRVMs were randomly divided into four groups: i) CON + NS-siRNA group, in which the cells were pretreated with NS-siRNA and then incubated with normal oxygen; ii) HR + NS-siRNA group, in which the cells were pretreated with NS-siRNA and then subjected to 4 h of hypoxia and 6 h of reoxygenation; iii) HR + metformin + NS-siRNA (HR + met + NS-siRNA) group, in which the cells were pretreated with NS-siRNA and then subjected to 4 h of hypoxia and 6 h of reoxygenation, and at the onset of reoxygenation, metformin (0.1 mM) was added; and iv) HR + metformin + AMPK-siRNA (HR + met + AMPK-siRNA) group, in which the cells were pretreated with AMPK-siRNA and then subjected to 4 h of hypoxia and 6 h of reoxygenation, and at the onset of reoxygenation, metformin (0.1 mM) was added into DMEM.

All data are presented as the mean ± SEM (n≥3). Statistical comparisons between the groups were performed using one-way ANOVA followed by Newman-Keuls or Tukey's post hoc test. Kruskal-Wallis test followed by Dunn's post hoc test was used for non-parametric data. The statistical analyses were performed by GraphPad Prism 5.0 (GraphPad Software, Inc.). P≤0.05 was considered to indicate a statistically significant difference.

After 24 h reperfusion in IR rat, the IS and AR were determined using Evans blue and TTC. The IS/AR was used to evaluate damage to the heart as previously described (16). The results showed that treatment with metformin significantly reduced the infarct size in IR rats by ~20% (P<0.05; Fig. 1A and B). There was no infarct area in the Sham group, so the IS/AR was 0. Plasma LDH and CK-MB levels in the IR group were increased compared with the Sham group; however, the increases were significantly attenuated by metformin (P<0.001; Fig. 1D and E). The viability of cardiomyocytes decreased significantly after HR treatment compared with the cells in CON group (P<0.001); treatment with metformin increased the viability of HR-treated cardiomyocytes (P<0.05; Fig. 1C).

Oxidative stress plays an important role in the process of MIRI (14). 4-hydroxynonenal (4HNE) is the final product of lipid oxidation, the amount of which indicates the severity of oxidative damage (28). Immunohistochemistry results showed that the levels of 4HNE in the IR group were notably increased compared with the Sham group (Fig. 2B). However, the content of 4HNE in the IR + met group was markedly decreased compared with the IR group (Fig. 2B). In addition to 4HNE, another lipid peroxidation product, MDA, was measured in cells following HR. The results showed that the level of MDA in the HR + met group was significantly decreased compared with the HR group (Fig. 2D; P<0.001). A TUNEL assay was performed in tissue sections to evaluate apoptosis. It was found that apoptosis was increased in the IR group compared with the Sham group (P<0.05; Fig. 2A and C). When treated with metformin, the percentage of TUNEL-positive cells was significantly decreased compared with the IR group (P<0.01; Fig. 2C).

NOX4 plays an important role in regulating the content of ROS in cardiac myocytes (6). A significant increase in NOX4 mRNA was observed in the IR group compared with the Sham group; after metformin was administered, the elevation in NOX4 was significantly attenuated (P<0.05; Fig. 3A). The protein level of NOX4 in the IR group was also significantly increased compared with that in the Sham group; after metformin treatment, the increase of NOX4 was significantly reversed (P<0.05; Fig. 3B and C). The activation of AMPK, which is the downstream factor of metformin (26), was also evaluated. The phosphorylation of AMPK in the IR group was significantly increased compared with that in the Sham group. In the IR + met group, metformin significantly promoted the phosphorylation of AMPK to a higher level than that in the IR group (P<0.05; Fig. 3B and D).

In the NRVM HR model, treatment with metformin increased p-AMPK levels (P<0.01; Fig. 4A and B) and significantly reduced NOX4 expression (P<0.01; Fig. 4A and C). The AMPK inhibitor compound-C was used to investigate the relationship between AMPK and NOX4; following treatment with compound-C, metformin did not induce a further increase in the phosphorylation of AMPK (P<0.05; Fig. 4B), and the expression of NOX4 was significantly increased compared with the HR + met group (P<0.05; Fig. 4C).

Following transfection with AMPK-siRNA, the expression of AMPK was significantly downregulated in both the CON + AMPK-siRNA group and IR + AMPK-siRNA group compared with the corresponding NS-siRNA groups (P<0.05; Fig. 5F). In the HR + NS-siRNA NRVM model, treatment with metformin significantly enhanced AMPK phosphorylation (P<0.05; Fig. 5A and B), while significantly reducing NOX4 expression (P<0.01; Fig. 5C). AMPK-siRNA was used to explore the relationship between AMPK and NOX4. In the HR + met + AMPK-siRNA group, the expression of AMPK was significantly downregulated compared with the HR + met + NS-siRNA group (P<0.01; Fig. 5A and D). The p-AMPK/AMPK ratio was used to determine the phosphorylation level of AMPK; it was found that p-AMPK/AMPK was significantly increased in the HR + met + NS-siRNA group compared with in the HR + NS-siRNA group (P<0.05; Fig. 5A and E. No significant difference, however, was observed between the HR + met + NS-siRNA and HR + met + AMPK-siRNA groups (P>0.05; Fig. 5A and E). Following transfection with AMPK-siRNA, metformin did not induce a further increase in p-AMPK (P<0.05; Fig. 5B), and the expression of NOX4 was significantly increased compared with the HR + met + NS-siRNA group (P<0.01; Fig. 5A and C).