Dapagliflozin was obtained from MCE (Cat. HY-10450). Ferrostatin-1 (Fer-1) was purchased from Selleck Chemicals (Cat# S7243). Drugs were separately dissolved in DMSO and added to the culture medium 12 h before H/R. For in vivo experiments, DAPA was purchased from commercially available dapagliflozin tablets (Forxiga Tab 10 mg, AstraZeneca AB, United States). DAPA was diluted with sterile water for injection (WFI) and given to animals by intragastric administration, as previously shown (Nikolaou et al., 2022). Primary antibodies against β-tubulin (1:5,000, Cat. AC021), GPX4 (1:1,000, A1933), SLC7A11 (1:1,000,A2413), ACSL4 (1:1,000, A6826), PTGS2 (1:1,000, A1253), and FTH1 (1:1,000, A19544) were purchased from ABclonal (Wuhan, China); FTMT (1:200,aa65-227) was purchased from LifeSpan BioSciences (Seattle, WA, United States); p-ERK (1:1,000,#4370), ERK (1:1,000, #4695), p-P38 (1:1,000, #4511), and P38 (1:1,000, #8690) were purchased from Cell Signaling Technology (Beverly, MA, United States); p-JNK (1:1,000, ab76572)), and JNK (1:1,000, ab208035) were purchased from Abcam (San Francisco, CA).

H9C2, the rat cardiomyocyte cell line, was purchased from the Shanghai Institute of Cell Biology, Chinese Academy of Science (Shanghai, China). The cells were cultured with Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum (FBS), 100 U/mL penicillin, and 100 mg/mL streptomycin (Beyotime, Shanghai, China) at 37°C with 5% CO₂. The cells were treated with the H/R model as reported previously (Son et al., 2020). H9C2 cardiomyocytes were replaced with FBS-free and glucose-free DMEM and exposed to a hypoxic environment of 95% N₂ and 5% CO₂ at 37°C for 4 h, while the control groups were cultured under normal conditions. H9C2 cells were then replaced with ordinary medium and placed in an ordinary incubator for 2 h to simulate reoxygenation injury. Two concentrations of DAPA (2.5 or 5 μM)were preadministered for 12 h.

Cell viability was measured using a Cell Counting Kit-8 (CCK-8) (CK04, Dojindo). After treatment with H/R on a 96-well plate, 10 μL CCK-8 solution was added and incubated at 37°C for 1 h. Then, the OD of each well was measured at 450 nm using a microplate reader (Thermo Fisher Scientific, United States of America). A cytotoxicity lactate dehydrogenase (LDH) Assay Kit-WST (CK12, Dojindo) was used to detect the release of lactate dehydrogenase to further detect the degree of cell damage. The reagent was added according to the product instructions, and the absorbance at 490 nm was measured immediately to calculate the cell damage rate.

H9C2 cardiomyocytes were washed twice with ice-cold PBS and then lysed. The glutathione (GSH) content was measured using a GSH and GSSG assay kit (S0053, Beyotime), and the optical density was measured at 405 nm. A lipid oxidation detection kit (S0131S, Beyotime) was used to detect the level of malondialdehyde (MDA), and the optical density value was measured at 535 nm. SOD enzyme activity was detected by a total SOD activity detection kit (WST-8 method) (S0101S, Beyotime), and the optical density value was measured at 450 nm. The Fe²⁺ concentrations of cardiac tissue and cell samples were determined using a Ferrous Iron Colorimetric Assay Kit (E-BC-K773-M, Elabscience, Wuhan, China), measuring absorbance at 590 nm using a microplate reader.

Intracellular ROS were detected by fluorescence 2′,7′-dichlorofluorescein diacetate (DCFH-DA) (S0033M, Beyotime, Shanghai, China) and Dihydroethidium (S0063, Beyotime). The cells were incubated with 10 μM DCFH-DA or 5 μM Dihydroethidium diluted with FBS-free DMEM for 20 min at 37°C and then washed twice with FBS-free medium. Fluorescence at an excitation wavelength of 488 nm was measured by fluorescence microscopy or flow cytometry.

Cells were seeded in 6-well plates and pretreated with dapagliflozin as indicated. After H/R, the cells were incubated with 10 μM C11 BODIPY 581/591 (RM02821, ABclonal) for 1 h at 37°C in the dark and then analyzed using a flow cytometer. BODIPY emission was detected using the FL one channel. As a conservative estimate, 10,000 cells were collected for data analysis.

ΔΨm was detected using a JC-1 fluorescent probe (S2003, Beyotime, China), a non-toxic cell-permeable cationic fluorescent dye. Briefly, after treatment, H9C2 cells were incubated with 1 mL JC-1 staining solution in a 37°C incubator for 20 min, washed with JC-1 staining buffer twice, cultured with 1 mL of culture medium according to the experimental protocols, and then removed in an ice bath. Fluorescence was immediately captured with a fluorescence microscope. The JC-1 monomers were excited with a 488 nm helium-neon laser for green fluorescence and imaged through a 525 nm-long path filter. In addition, the JC-1 aggregates were excited with red fluorescence by a 543 nm He-Ne laser line and imaged through a 590 nm-long path filter.

Based on “Practical guidelines for Rigor and Repeatability in Preclinical and clinical studies of cardiac protection” (Botker et al., 2018), the study was conducted in adult male SD rats aged 8–12 weeks weighing 250–300 g. The experimental protocols of the animals involved in this study were approved by the Laboratory Animal Research Committee of Soochow University. Rats were kept in pathogen-free, temperature-controlled environments (20°C–25°C) and specific facilities with 12-h light/dark cycles, with a maximum of six per cage and free feeding on conventional laboratory animal feed. The rats were randomly divided into three groups (n = 5 per group): Sham group (saline + sham operation), I/R group (saline + I/R); Dapa group (dapagliflozin 10 mg/kg/day + I/R). DAPA or saline was intragastrically administered once daily for 5 days. On the sixth day, the rats underwent myocardial ischemia for 30 min followed by reperfusion for 2 h. Briefly, adult male SD rats were anesthetized with 50 mg/kg sodium pentobarbital and placed in the supine position on a 37°C heating pad. During the experiment, a standard limb II lead electrocardiogram was performed continuously. The tracheal incision was intubated, and mechanical ventilation was connected with a ventilator. After the left thoracic incision, 6-0 silk thread was sutured at the root of the anterior descending branch of the left coronary artery (LAD), positioned 2 mm below the intersection of the left atrial appendage and arterial conus, and a slip-knot was made. After 30 min of ischemia, the knot was released, and reperfusion was performed for 2 h. In the sham group, the same thoracotomy was performed without ligating the coronary arteries. The animals were then sacrificed for subsequent experiments.

Transthoracic echocardiography was performed at baseline (before experiments) and after reperfusion in anesthetized rats using a Visual Sonic Vevo2100 system with an MS-250 transducer. Heart rate and left ventricular (LV) parameters, including diastolic and systolic wall thickness, LV end-diastolic and end-systolic diameters, and LV functional parameters, including left ventricular ejection fraction (LVEF) and LV fractional shortening (LVFS), LV end-systolic interior diameter (LVID.s), end-diastolic interior diameter (LVID. d), were measured from a 2D short axis at the papillary muscle level in M-mode images using Vevo 2,100 software.

The heart was removed and fixed with 4% paraformaldehyde overnight, embedded in paraffin, and sliced into 5-μm-thick sections. Left ventricular specimens were stained with hematoxylin and eosin (H&E) according to a previously reported method (Fang et al., 2019) and photographed by light microscopy.

Blood samples were collected after I/R, and serum was obtained after centrifugation at 2000 × g for 20 min. Cardiac troponin T (TNT) and brain natriuretic peptide (BNP) levels were determined using commercial ELISA kits (LunChangshuo Biotech, China) according to the manufacturer’s instructions. Infarct size determination was performed using Evans blue and 2,3,5-triphenyl tetrazolium chloride monohydrate (TTC) (G3005, Solarbio, Beijing, China) double staining. Briefly, following 2 h of reperfusion, the hearts were excised and 2 ml of a 1% solution of Evans Blue dye was retrogradely perfused into the aorta. After perfusion, the heart was frozen at −20°C for 30 min, and sliced into five cross-sections from apex to base. The sections were incubated with 1% TTC at 37°C for 10 min and then fixed in 10% formalin solution for 24 h. The heart sections were analyzed using ImageJ software, the blue area determined the non-ischemic region, and areas were summed and calculations of the White Area (Infarct area, IA)/total red + white area (Area at risk, AAR) for each heart were determined. The myocardial infarction size is a percentage of the area at risk.

Total mRNA was extracted from H9C2 cells or heart homogenates using commercial RNA extraction kits (RC112-01, Vazyme, China). The concentration and purity of the samples were measured using a NanoDrop One spectrophotometer (Thermo Scientific, Waltham, MA). RNA was reverse-transcribed by ABScript III RT Master Mix for qPCR with gDNA Remover (RK20429, ABclonal) according to the manufacturer’s instructions. Quantitative PCR was performed with an AB7500 Fast Real-Time System using Genious 2X SYBR green (RK21204, Abclonal). The sequences of the primers are shown in Supplementary Table S1. The amplification conditions for PCR were as follows (in a total volume of 10 μl): 95°C for 10 min followed by 36 cycles of denaturation at 65°C for 5 min. Cycle threshold (Ct) values were determined by the comparative Ct method and normalized to GAPDH levels.

For Western blotting, cells or tissues were lysed using RIPA buffer containing 1% protease inhibitor and phosphatase inhibitor for 30 min and then centrifuged at 12000 × g for 15 min at 4°C to collect the supernatant. The protein concentrations were measured using a Barford protein assay kit (P0006C, Beyotime). The samples were heated with a loading buffer at 95°C for 15 min. Approximately 20 µg of protein was separated by 10%–12.5% SDS‒PAGE and then transferred to PVDF membranes. After blocking with 5% BSA for 1 h, the membranes were incubated overnight with a specific primary antibody at 4°C. The immunoreactive bands were incubated with horseradish peroxidase secondary antibody for 1 h and exposed using a gel imager (Bio-Rad, CA, United States). Protein bands were visualized using an ECL kit (FUDE Biological Technology, Hangzhou, China) and quantified using ImageJ analysis software version 1.8.0. The relative quantity of the target protein was normalized to β-tubulin. For immunofluorescence, samples were treated with primary antibody followed by incubation with Alexa Fluor 488-conjugated anti-rabbit IgG (1:200, Ab150117).

The data in this study represent at least three independent experiments, and the values are expressed as the mean ± standard deviation. Statistical analysis was performed using GraphPad Prism 8.0 software (San Diego, CA, United States). First, the normal distribution was tested, followed by a homogeneity test of variance using Levene’s test. Data from more than two groups were compared using one-way ANOVA followed by Tukey’s correction for post hoc multiple comparisons (equal variances) or Welch ANOVA tests followed by Dunnett’s T3 multiple comparisons test (Unequal variances). Two-way ANOVA for repeated measures (cardiac echocardiography parameters) followed by Sidak’s multiple comparison tests. p < 0.05 was considered to be statistically significant.

To investigate MIRI, we performed H/R in H9C2 cardiomyocytes. We initially performed hypoxia for 4 h followed by a time gradient of reoxygenation to measure cell activity using CCK-8. The results revealed that the cell activity of H9C2 cells decreased significantly after H/R (Figure 1A). To prove that cardiac H/R injury may be related to ferroptosis, we treated H9C2 cells with ferrostatin (Fer-1) before H/R. The results confirmed that Fer-1 improved H9C2 H/R cell activity (Figure 1B) and reduced ROS production (Figures 1C, D). In addition, in H/R-induced H9C2 cell death, treatment with Fer-1 significantly reduced the mRNA level of the molecular marker of ferroptosis and the most critical lipid peroxidase, prostaglandin endoperoxide synthase 2 (PTGS2) and acyl-CoA synthetase long-chain family member 4 (ACSL4) (Figures 1E, F). We also examined the mRNA expression of genes associated with iron metabolism in ferroptosis, such as TFRC and HMOX1, which showed similar changes (Figures 1G, H). These results suggest that ferroptosis activation may play a critical role in the pathogenesis of myocardial ischemia/reperfusion.

The results showed that DAPA had no cytotoxic effect on H9C2 cells under normal culture up to 200 μM (Figure 1I). Meanwhile, the H/R model resulted in a significant decrease in H9C2 cell viability, which was reversed by DAPA treatment in a dose-dependent manner, indicating that DAPA had a protective effect on H/R in vitro (Figure 1J). On the other hand, under H/R conditions, cell viability decreased further when DAPA reached a maximum of 40 μM (Figure 1K). As shown in Figure 1L, DAPA prevented the decrease in SOD activity, and LDH release in H/R-treated cells was also prevented by DAPA (Figure 1M). The protein level of PTGS2 was significantly decreased in the DAPA-treated group compared with the H/R group (Figures 1N, O). The above results indicated the protective effect of DAPA on H/R in cardiomyocytes.

Lipid peroxidation is described as a process in which oxidants, such as ROS, attack lipids, which is a key feature of ferroptosis, so we examined the effects of DAPA on ROS production and lipid peroxidation. H/R modelcaused a distinct increase in ROS and DHE. Nevertheless, DAPA pretreatment effectively prevented-reversed these changes (Figures 2A–D). In addition, DAPA alleviated the H/R-induced increase in MDA content (Figure 2E). ACSL4 is the most critical enzyme in lipid peroxidation in ferroptosis. As expected, the mRNA (Figure 2F) and protein levels (Figures 2G, H) of ACSL4 increased significantly under H/R, which was effectively relieved in the dapagliflozin treatment groups.

Since glutathione derivation was demonstrated to be a hallmark of ferroptosis, we next explored whether dapagliflozin affected glutathione depletion. As shown in Figures 3A–C, antioxidant enzyme GSH activity was significantly reduced during H/R, which was effectively ameliorated by DAPA. The SLC7A11/GPX4 axis is recognized as the primary defense mechanism against ferroptosis with the help of GSH. The protein levels of GPX4 and SLC7A11, glutathione metabolism genes, were decreased in the H/R group compared with the control group, and DAPA pretreatment ameliorated this alteration (Figures 3D–F), which was further evaluated by immunofluorescence and quantitative analyses (Figures 3G, H).

As the central mediator of ferroptosis, iron overload produces excess ROS by the Fenton reaction, leading to mitochondrial and cell damage (Li and Zhang, 2021). We investigated whether dapagliflozin affected iron metabolic indicators and mitochondrial membrane potential. The results showed that the intracellular Fe²⁺ level was significantly increased in the H/R group, and pretreatment with dapagliflozin reduced intracellular Fe²⁺ levels (Figure 4A). To unveil the intrinsic molecular mechanism of cumulative ferrous iron, we further explored the expression of genes related to extracellular iron absorption and intracellular iron storage. Transferrin receptor 1 (TFRC), ferritin, and mitochondrial ferritin (FTMT) play important roles in iron metabolism by regulating iron uptake and storage, respectively. The level of TFRC tended to be upregulated during H/R, but the difference was not statistically significant, while H/R significantly decreased FTH and FTMT levels (Figures 4B–E). To detect mitochondrial dynamics, JC-1 staining is shown in Figure 4F. A significant decrease in red fluorescence was remarkable in the H/R group, leading to a drop in the ratio of red to green fluorescence, compared to the control group. In contrast, dapagliflozin treatment reversed these changes (Figure 4G). These findings demonstrate that dapagliflozin regulates iron metabolism and improves mitochondrial function.

To evaluate the effect of dapagliflozin on myocardial I/R injury, rats were pretreated with DAPA for 5 days before myocardial I/R surgery (Figure 5A). ST-segment elevation and reperfusion arrhythmia confirmed the success of the I/R procedure. As shown in Figure 5B, I/R stimulated the ST segment to be significantly elevated, and arrhythmias such as premature ventricular contractions or ventricular tachycardia occurred after reperfusion. The DAPA group had decreased ST segment elevation and reperfusion arrhythmia and significantly ameliorated the increased cTnT and BNP levels (Figures 5C, D). Representative echocardiographic images in M-mode showed that the DAPA group had higher LVPWs, LVFS and LVEF but lower LVIDs than the control group (Figures 5E, F). I/R exposure resulted in myocardial structural damage, disordered fiber arrangement, reduced myocardial cells, nuclear shrinkage, and inflammatory cell infiltration. In addition, Figures 5G, H shows that myocardial infarct size (%) was significantly higher in the I/R group than in the sham group, and pretreatment of DAPA decreased the infarct size. The histopathological changes were significantly attenuated in the DAPA group (Figure 5I). These results indicate that dapagliflozin reduces myocardial I/R injury and improves cardiac function in rats.

In terms of molecular mechanisms, there was a significant increase in PTGS2, ACSL4, and NCOA4 mRNA levels in the I/R group compared to the control group, and the DAPA group exhibited reversed changes (Figures 6A–C). As depicted in Figure 6D, I/R surgery greatly facilitated the generation of Fe²⁺, and DAPA reduced intracellular ferrous iron overload. We found that PTGS2 was dramatically upregulated and SLC7A11/GPX4 protein levels were downregulated in the I/R group compared with the sham group. Non-etheless, DAPA successfully reduced PTGS2 levels and promoted SLC7A11/GPX4 axis expression (Figures 6E–H), suggesting that DAPA reduces ferroptosis in myocardial I/R rats.