The chemical structure of roflumilast was determined by PubChem (pubchem.ncbi.nlm.nih.gov/; Fig. 1A).

The experimental protocol for animal studies was reviewed and approved by the Committee for the Ethics of Animal Experiments, Shenzhen Peking University-The Hong Kong University of Science and Technology Medical Center (approval no. 2021-807; Shenzhen, China). The MI/R rat model was established as previously described (21). A total of 20 male Sprague-Dawley rats (age, 12 weeks; weight, 180-250 g) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. The rats were raised in an environmentally controlled room (22±2˚C, humidity of 55±5%, 12/12-h light/dark cycle) with free access to standard animal feed and filtered tap water for 7 days. Rats were anesthetized by intraperitoneal injection of 10% chloral hydrate (0.3 ml/100 g) and the intercostal space was opened under mechanical ventilation. The animals exhibited no signs of peritonitis, pain or discomfort. To establish the MI/R rat model, the left anterior descending coronary artery was ligated for 45 min, followed by reperfusion for 2 h. Rats in the control group were not ligated. The rats were randomly divided into four groups (n=5/group): Control; MI/R, MI/R + roflumilast (1 mg/kg; Adooq Bioscience) and MI/R + roflumilast (3 mg/kg) group. Prior to MI/R operation, roflumilast was administered to rats by gavage at 1 or 3 mg/kg once daily for 7 consecutive days. Rats were humanely sacrificed under anesthesia by intraperitoneal injection of 1% pentobarbital sodium (150 mg/kg). The blood samples (~6 ml) collected from the hearts and cardiac tissues were stored at -80˚C until further analysis.

H9C2 cells were obtained from the American Tissue Culture Collection (ATCC; cat. no. CRL-1446). H9C2 cells were cultured in ATCC-formulated Dulbecco's Modified Eagle's Medium (cat. no. ATCC 30-2002) supplemented with 10% fetal bovine serum (cat. no. ATCC 30-2020) in 95% air and 5% CO₂ at 37˚C.

For H/R stimulation, H9C2 cells were grown in an anoxic chamber with 5% CO₂ and 95% N₂ for 6 h and then in a normal chamber with 95% air and 5% CO₂ for 12 h at 37˚C.

For roflumilast treatment, cultured cells were pre-incubated with roflumilast (1.0, 2.5 and 5.0 µM; Adooq Bioscience) or compound C (10 µM; an inhibitor of the AMPK signaling pathway; MedChemExpress, United States) for 30 min at 37˚C before H/R treatment.

The MI area in each group was observed by TTC staining. Following storage at -18˚C for 15 min, the heart tissue perpendicular to the coronary sulcus was cut into five equal pieces. Following incubation with 1% TTC solution (Sigma-Aldrich; Merck KGaA) for 10 min at 37˚C and 10% neutral buffered formalin (Thermo Fisher Scientific, Waltham, MA) for 90 min at 37˚C in the dark, all slices (1 mm) were imaged using a digital camera. The MI areas were determined by Image-Pro Plus image analysis software (version 4.1; Media Cybernetics, Inc.). Finally, calculation of MI area was conducted according to the following formula: Myocardial infarct size (%)=(infarct area/whole heart area) x100%.

The cardiac tissue samples collected from the left ventricle were fixed in 4% paraformaldehyde overnight at 4˚C, dehydrated in ascending ethanol gradient and embedded in paraffin for 50-60 min at room temperature. Subsequently, the embedded cardiac tissues were cut into 4 µm slices, followed by staining with 0.5% hematoxylin for 5 min and eosin for 2 min at room temperature. Then, the sections were mounted and observed under a light microscope (magnification, x100/400; Olympus Corporation BX53).

The mitochondrial membrane potential in cardiac tissue and H9C2 cells was detected using the JC-1 staining kit (cat. no. C2006; Beyotime Institute of Biotechnology) according to the manufacturer's instructions. In brief, the isolated cardiomyocyte suspension and H9C2 cells were stained with 2.5 mg/ml JC-1 solution for 20 min at 37˚C. Subsequently, the cells were washed with JC-1 staining buffer twice and observed using a fluorescence microscope (magnification, x200; Olympus Corporation BX63).

The collected peripheral blood (15 ml) was centrifuged at 2,072 x g at 4˚C for 10 min to separate the serum. Heart muscle damage indicators, including aspartate transaminase (AST), creatine kinase-myocardial band (CK-MB) and lactate dehydrogenase (LDH) in serum were detected by AST (cat. no. C010-2-1), CK-MB (cat. no. A032-1-1) and LDH (cat. no. A020-2-2) assay kits (all Nanjing Jiancheng Bioengineering Institute) according to the manufacturer's instructions, respectively.

The concentrations of IL-1β and IFN-γ in myocardial tissue and in cell supernatant were measured using ELISA kits for IL-1β (cat. no. E-EL-R0012c; Elabscience Biotechnology, Inc.) and IFN-γ (cat. no. E-EL-R0009c; Elabscience Biotechnology, Inc.) according to the manufacturer's instructions.

The activity of malondialdehyde (MDA) and superoxide dismutase (SOD) in myocardial tissue and H9C2 cells were measured using MDA (cat. no. S0131S; Beyotime Institute of Biotechnology) and SOD assay kits (cat. no. S0109; Beyotime Institute of Biotechnology) according to the manufacturer's instructions.

The detection of ATP concentration in H9C2 cells was conducted using ATP assay kit (cat. no. S0026; Beyotime Institute of Biotechnology) according to the manufacturer's instructions. Briefly, H9C2 cells were collected and mixed with cell lysis buffer for 10 min at 4˚C, followed by centrifugation at 12,000 x g at 4˚C for 5 min. Subsequently, cell supernatant was incubated with 100 µl kit solution at room temperature for 5 min and the ATP levels in the cell supernatant were detected using a LuminMax-C luminometer (Maxwell Sensors Inc.).

Protein from cardiac tissues and H9C2 cells was obtained using RIPA lysis buffer (Beyotime Institute of Biotechnology) and qualified with a BCA detection kit (Beyotime Institute of Biotechnology). A total of 25 µg/lane protein was separated by 10% SDS-PAGE and transferred to a PVDF membrane. Following blocking with 5% BSA (Beyotime Institute of Biotechnology) at room temperature for 2 h, the membrane was incubated with primary antibodies targeting AMP-activated protein kinase alpha (AMPKα; cat. no. 5831; 1:1,000; Cell Signaling Technology, Inc.), phosphorylated (p-)AMPKα (cat. no. 50081; 1:1,000; Cell Signaling Technology, Inc.), SIRT1 (cat. no. ab189494; 1:1,000; Abcam), Bcl-2 (cat. no. ab196495; 1:1,000; Abcam), Bax (cat. no. ab32503; 1:1,000; Abcam), PINK1 (cat. no. ab186303; 1:1,000; Abcam), DRP1 (cat. no. 8570; 1:1,000; Cell Signaling Technology, Inc.), p-DRP1 (cat. no. 4867; 1:1,000; Cell Signaling Technology, Inc.) and β-actin (cat. no. 93473; 1:1,000; Cell Signaling Technology, Inc.) overnight at 4˚C. Then, membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG (cat. no. ab6721; 1:2,000; Abcam) or goat anti-mouse IgG (cat. no. ab6789; 1:2,000; Abcam) for 1 h at room temperature. Finally, the bands were examined with ECL reagent (Beyotime Institute of Biotechnology) and band density was quantified using ImageJ Software (version 1.46; National Institutes of Health).

H9C2 cells were inoculated into a 96-well plate at a density of 1x10³ cells/well. H9C2 cells were treated with roflumilast (1.0, 2.5 and 5.0 µM) for 24 h at 37˚C and induced by H/R. H9C2 cells in each well were mixed with 10 µl CCK-8 solution (Beyotime Institute of Biotechnology) and incubated at 37˚C for 1 h. The optical density at 450 nm was detected by a spectrophotometer (Bio-Rad Laboratories, Inc.).

The apoptosis of H9C2 cells was determined by In Situ Cell Death Detection kit (cat. no. 11684795910; Roche) for the TUNEL assay. Briefly, H9C2 cells in a 24-well plate (1x10⁵ cells/well) were fixed in 4% paraformaldehyde at room temperature for 15 min, then 0.1% Triton-X-100 at room temperature for 10 min. Then, cells on the slides were stained with TUNEL reaction mixture (50 µl terminal deoxynucleotidyl-transferase and 450 µl fluorescein-labeled deoxyuridine triphosphate) at 37˚C for 1 h in the dark and nuclei were stained with 10 µg/ml DAPI at room temperature for 5 min. The cells were mounted with PBS and glycerol (ratio, 1:2) and observed by fluorescence microscopy in five randomly selected high-power microscope fields (magnification, x200). The formula used to calculate the percentage of cell apoptosis was as follows: Cell apoptosis (%)=(number of apoptotic H9C2 cells/total number of H9c2 cells) x100%.

The opening of the mPTP was detected using a calcein-loading/cobalt chloride (CoCl₂)-quenching system. Briefly, 2x10⁵ H9C2 cells seeded in a 6-well plate were treated with 1 µM calcein and 2 mM CoCl₂ for 20 min at 37˚C in the dark. After washing with PBS, H9C2 cells were observed and imaged using a confocal laser scanning microscope (model no. LSM 880; Carl Zeiss AG; magnification, x100). The mean green fluorescence intensities in the mitochondria were quantified using ImageJ Software (version 1.46; National Institutes of Health). The changes of green fluorescence intensity in the mitochondria were the index of mPTP opening.

GraphPad (version 8.0.1; GraphPad Software, Inc.; Dotmatics) was used to analyze the experimental data. Data are shown as the mean ± the standard deviation from three independent experiments. The comparisons between multiple groups were conducted by one-way ANOVA followed by Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference.

The chemical structure of roflumilast is shown in Fig. 1A. TTC staining results showed that the infarct area in the MI/R group was significantly increased compared with that in the control group, indicating that the rat model of MI/R was successfully constructed. Moreover, roflumilast (1 and 3 mg/kg) significantly decreased the increased infarct size in rats subjected to MIRI in a dose-dependent manner (Fig. 1B). Furthermore, the increased levels of CK-MB (Fig. 1C), LDH (Fig. 1D) and AST (Fig. 1E) in the MI/R group were decreased by treatment of roflumilast.

The H&E staining results revealed that the myocardial fiber structure was damaged, vascular walls were broken and hemocyte infiltration was apparent in the MI/R group. However, MI/R rats pre-injected with roflumilast exhibited mild tissue damage (Fig. 2A). The inflammatory factors (IL-1β and IFN-γ) in cardiac tissues were significantly increased in the MI/R group but significantly reduced by roflumilast treatment (Fig. 2B).

The ratio of green and red fluorescence intensity was significantly increased in MI/R rats compared with control rats; this was reversed by roflumilast treatment, indicating that roflumilast decreased depolarization of the mitochondrial membrane (Fig. 3A). In addition, MDA was significantly increased and SOD was significantly decreased in the cardiac tissues of rats subjected to MI/R; these effects were subsequently reversed by roflumilast (Fig. 3B).

The expression of p-AMPK and SIRT1 in the cardiac tissue of rats subjected to MI/R were significantly decreased compared with the control group, while roflumilast treatment promoted the expression of p-AMPK and SIRT1 in a dose-dependent manner (Fig. 4).

After H9C2 cells were treated with roflumilast, their viability was unchanged, indicating that roflumilast has no significant effect on H9C2 cells at these concentrations (Fig. 5A). H/R-induced H9C2 cells showed significantly decreased viability, which was reversed by roflumilast in a dose-dependent manner (Fig. 5B). H/R induction significantly decreased the expression of p-AMPK and SIRT1, while roflumilast promoted the expression of p-AMPK and SIRT1 in H/R-induced H9C2 cells (Fig. 5C). H9C2 cells subjected to H/R induction demonstrated a significant increase in the number of TUNEL-positive/apoptotic cells; this was significantly decreased following treatment with roflumilast, while compound C weakened the effect of roflumilast (Fig. 5D). Roflumilast resulted in the increased expression of Bcl-2 and decreased expression of Bax in H/R induced H9C2 cells, which was then reversed by compound C (Fig. 5E).

Roflumilast significantly suppressed MDA and significantly upregulated SOD in H/R-induced H9C2 cells, while compound C significantly impaired the function of roflumilast (Fig. 6A). ELISA revealed a significant increase in levels of IL-1β and IFN-γ in the H/R group compared with the control group, which were significantly decreased by roflumilast treatment. However, the decreased levels of IL-1β and IFN-γ in the H/R + roflumilast (5 µM) group were partially elevated by compound C (Fig. 6B).

Roflumilast treatment significantly decreased the ratio of green and red fluorescence intensity in H/R-induced H9C2 cells compared with that in the H/R group; this was increased following the administration of compound C (Fig. 7A). H/R induction significantly decreased the ATP levels, which were promoted by roflumilast. Compound C decreased the ATP levels in H/R + roflumilast + compound C group compared with the H/R + roflumilast group (Fig. 7B). Compared with the control, the fluorescence intensity of mitochondrial calcein was significantly decreased in the H/R group, indicating that the extent of mPTP opening was enhanced following H/R. Moreover, roflumilast significantly increased fluorescence intensity compared with the H/R group, indicating that roflumilast inhibited H/R-induced mPTP opening in H9C2 cardiomyocytes, which was reversed by compound C (Fig. 7C and D). The enhanced expression of PINK1 and p-DRP1 in H9C2 cells due to H/R induction were suppressed by roflumilast treatment; these effects were then reversed by compound C (Fig. 7E).