Establishment of murine myocardial I/R model

We obtained 8–10-week-old male wild-type C57BL/6J mice from the Experimental Animal Center of Southern Medical University. Myocardial I/R model was established in mice as previously described [1]. Briefly, the adult mice were anesthetized with isoflurane (5% for induction and 2% for maintenance) and ventilated using a rodent respirator. Mice in the GW4869 group received intraperitoneal injection of 2.5 mg/kg GW4869 (D1692, Sigma, USA) 1 h before I/R surgery [21]. Subsequently, left thoracotomy was performed and the left anterior descending (LAD) coronary artery was ligated for 45 min using an 8-0 silk suture. Immediately after removing the silk suture, the reperfusion region was injected with 5 × 106 neutrophils or 1 × 1010 particles EVs in accordance with the experimental design [22, 23]. The periods of reperfusion varied according to the experiments. Mice in the sham-operated group underwent the same surgical procedure but the LAD was not ligated. Before or after surgery, all the mice were maintained in a pathogen-free room under a 12-h light–dark cycle at 22 ± 1 °C and 65%–70% humidity, and were provided with free access to food and water.

Measurement of cardiac infarct size

Cardiac infarct sizes were assessed as previously described with minor modifications [24]. Briefly, after 24 h reperfusion, the LAD was re-ligated again and 0.5% Evans Blue dye (AAPR215-500, Pythonbio, China) was injected into the ascending aorta to delineate the ischemic area (area-at-risk, AAR) from the non-ischemic area. Then, the hearts were harvested, frozen at − 80 °C for 20 min, and sectioned into 5 slices. The heart tissue sections were incubated in 1.5% triphenyltetrazolium chloride (TTC) buffer (T8877, Sigma, USA) for 30 min at 37 °C to delineate the infarct size (IS). We measured the left ventricular (LV) region, AAR, and IS using the Image-Pro Plus 6.0 software (Media Cybernetics, Bethesda, MD, USA). The ischemic and infarct areas were expressed as percentages of AAR/LV and IS/AAR, respectively.

Measurement of cardiac functional parameters

Cardiac function was analyzed with Vevo 2100 high-resolution echocardiography (RMV-707B, VisualSonics, Toronto, ON, Canada) before surgery and 3 days after reperfusion. After anesthesia with inhaled isoflurane, the left ventricular end diastolic diameter (LVEDD) and the left ventricular end systolic diameter (LVESD) were measured along the short axis views. The left ventricular ejection fraction (LVEF), left ventricular fraction shortening (LVFS), left ventricular internal diameter in systole (LVIDs), and left ventricular internal diameter in diastole (LVIDd) were also calculated to determine the cardiac functional status.

Estimation of in vivo CM necrosis

CM necrosis is characterized by the rupture of the plasma membrane and estimated using the Evans blue dye uptake assay as described previously [25]. Briefly, the mice were administered 16 h before I/R with the Evans blue dye (0.1 mL/10 g) by intraperitoneal injections. The mice were sacrificed at 24 h post I/R. The hearts were harvested and cut into 4.5 μm thick slices after embedding in the optimal cutting temperature (OCT) compound. The heart sections were then incubated with the mouse anti-cardiac troponin T antibody (cTnT, ab33589, Abcam, USA). The stained sections were imaged under a confocal microscope. The percentage of necrotic CMs (Evans blue+ cTnT+) were quantified with the Image-Pro Plus 6.0 software. In addition, necroptotic molecular markers including receptor-interacting protein 3 (RIP3), mixed-lineage kinase domain-like protein (MLKL), phosphorylated RIP3 (p-RIP3), and phosphorylated MLKL (p-MLKL) were determined by western blotting.

Estimation of serum cTn-I and CK-MB levels

Blood samples were collected from the mice subjected to different treatments. The samples were processed and the serum samples were separated and stored at − 80 °C. The levels of cTn-I (JM-03071M2, Jingmei Biotechnology, China) and CK-MB (JM-03084M2, Jingmei Biotechnology, China) in the serum were measured using the ELISA kits according to the manufacturer’s instructions.

Isolation, culturing, and treatment of primary murine cells

We obtained 1-day-old neonatal wild-type C57BL/6J mice from the Experimental Animal Center of Southern Medical University. The primary neonatal CMs were isolated as previously described [26]. Briefly, we isolated hearts from the 1-day-old C57BL/6 mice, cut into small pieces, and digested with 0.25% trypsin (Gibco, USA) at 4 °C for 12 h. Then, the heart tissues were further digested with 0.1% type II collagenase (17101-015, Gibco, USA) in 1% BSA (Sigma, USA) at 37 °C for 15 min. Subsequently, the tissue extracts with CMs were centrifuged at 300×g for 5 min at 25 °C. The cell pellet was resuspended in DMEM/F12 medium (Gibco, USA) supplemented with 10% FBS and 1% penicillin/streptomycin (Gibco, USA). To purify the CMs, the cell suspension was subsequently cultured at 37 °C and 5% CO2 in a humidified incubator for 2 h. Then, the supernatant with CMs was carefully collected and centrifuged. The CMs were then resuspended with the DMEM/F12 medium and seeded in petri dishes at an appropriate density for subsequent assays.

Bone marrow-derived neutrophils (BMDNs) were isolated using the Mouse Neutrophil Negative Selection Kit (19762, Stemcell, Canada) according to the manufacturer’s instructions. Briefly, the adult wild-type C57BL/6 mice were sacrificed. The femurs and tibias were harvested immediately and placed in a petri dish with sterile pre-cooled PBS. Bone marrow cells were flushed out into a 5 mL tube with a syringe using the EasySep™ Buffer (20144, Stemcell, Canada) and sequentially incubated with the enrichment cocktail, biotin selection cocktail, and the magnetic particles (all kit components) at 4 °C for 15 min each. BMDNs were then separated using a magnet, resuspended with the RPMI 1640 medium (Gibco, USA) containing 10% FBS, and cultured at 37 °C and 5% CO2 in a humidified incubator. Freshly isolated BMDNs were defined as N0 neutrophils. BMDNs were stimulated with 1 µg/mL LPS and 20 ng/mL IFN-γ for 4 h to induce N1 polarization (CD11b+Ly6G+CD206−) and incubated with 20 ng/mL IL-4 to induce N2 polarization (CD11b+Ly6G+CD206+).

For the hypoxia/reoxygenation (H/R) experiments, the CMs were washed twice with PBS, resuspended in serum-free DMEM medium without glucose (Gibco-BRL, USA) and incubated in a hypoxic environment with 1% O2 and 5% CO2 for 12 h. Then, they were resuspended in DMEM/F12 medium supplemented with 10% FBS and 1% penicillin/streptomycin and then incubated in a normoxic environment with 21% O2 and 5% CO2 for 6 h.

For the co-culture experiments, CMs or CFs were seeded into the 0.4 μm pore inserts of a 6-well plate (Corning, USA) and subjected to H/R treatment. Subsequently, the CMs or CFs were co-cultured with the BDMNs, which were previously seeded into the 6-well plates and assessed after 24 h. CMs were treated with 10 μM GW4869 (D1692, Sigma, USA) in the 6-well plate inserts for 24 h to inhibit production of EVs. Then, they were subjected to H/R treatment by resuspending the cells in glucose-free DMEM containing 10 μM GW4869 [24]. Subsequently, the CMs were co-cultured with the BMDNs. DMSO was used as the vehicle control.

Cell viability assays

Lactate dehydrogenase (LDH) release by CMs in the culture medium was spectrophotometrically measured using a kit (BL1405A, Biosharp, China) according to the manufacturer’s instructions. The viability of BMDNs were determined by trypan blue staining (AAPR162-20, Pythonbio, China).

Isolation and characterization of EVs

EVs were isolated from the CM-conditioned medium by ultracentrifugation as previously described [27]. In brief, CM-conditioned media from different treatments were sequentially centrifuged at 300×g for 10 min, 2000×g for 20 min, and 10,000×g for 40 min at 4 °C to eliminate the cells. Then, the supernatant was filtered through a 0.22 μm filter (Millipore) to remove the cellular debris. The filtered supernatant was then centrifuged at 120,000×g for 90 min at 4 °C (Optima L-80 XP, Beckman Coulter, Brea, CA, USA) to pellet the EVs. The supernatant was removed. The pelleted EVs were resuspended with pre-cooled PBS and recentrifuged again at 120,000×g for 90 min at 4 °C. The supernatant was removed and the EV pellets were resuspended in PBS and stored at − 80 °C for subsequent use in the experiments.

According to the guideline of the International Society of Extracellular Vesicles [28], EVs from the CM-conditioned medium were characterized by transmission electron microscopy (TEM), nanoparticle tracking analysis (NTA), and western blotting. Morphology of the freshly isolated EVs was assessed by TEM. We pipetted 15 μL of the EVs onto a carbon-coated copper grid and air dried for 1 min. Then, the sample was incubated with 2% uranyl acetate for 1 min at room temperature. After removing excess uranyl acetate and drying, images were captured using the FEI Tecnai G2 Spirit TEM (Tecnai). The size distribution and concentration of the EVs were estimated by NTA (NanoSight NS300, Malvern, UK). EV sample was prepared by diluting 1 μL of the EVs with 29 μL of 0.22 μm filtered sterile PBS, and then samples were injected into the NanoSight instrument. The motility of the EVs were recorded for 30 s at least 3 times with the screen gain of 10 and the camera level of 12, and the following parameters were used for the analysis: a detection threshold of 4 and a screen gain of 10. The results were analyzed by the NTA analytical software (NanoSight NTA3.4). Western blotting analyses of the CMs and EV samples were performed using antibodies against EV-specific markers such as CD9, CD81, and TSG101.

EV uptake assay

CM-derived EVs were labeled with the red fluorescent dye PKH26 (MINI26, Sigma, USA) according to the manufacturer’s instructions. The PKH26-labeled EVs were isolated using ultracentrifugation again to clear out the free PKH26 dye. BMDNs (1 × 107 cells) were seeded in the poly-l-lysine–coated confocal dishes. Then, after 3 h, BMDNs were incubated with 1 × 1010 PKH26-labeled EVs at 37 °C for 6 h. The nuclei of BMDNs were stained with DAPI (10 μg/mL) for 10 min before the BMDNs were observed by confocal microscopy (Leica, Germany). The PBS was also incubated with PKH26 and purified to serve as a negative control.

For assessing the uptake of EVs by the cardiac-infiltrating neutrophils in vivo, CM-derived EVs were labeled with the red fluorescent dye DiR (UR21017, Umibio, China) according to the manufacturer’s instructions. The DiR-labeled CM-derived EVs were isolated using ultracentrifugation again to eliminate the unbound DiR dye. Then, immediately after removing the silk suture, the reperfusion region was injected with 1 × 1010 DiR-labeled EVs. The PBS was also incubated with DiR and purified to serve as a negative control. The fluorescence of DiR-labeled EVs was measured using the Bruker In Vivo FX Pro system (Bruker, MA, USA) at 0 h and 24 h after the injection of DiR-labeled EVs. Furthermore, heart samples were collected after 24 h and single cell suspensions were prepared. Flow cytometry was performed to determine the uptake of DiR-labeled EVs in the cardiac-infiltrating neutrophils.

Preparation of single-cell suspensions from the murine hearts

Single-cell suspensions of the murine hearts were prepared as described previously [10]. Briefly, LV tissues were harvested, minced, and incubated with a cocktail of collagenase II (600 U/mL, Worthington) and DNase I (60 U/mL, AppliChem) at 37 °C for 45 min. The tissue debris and cell clumps were removed by filtering the cell suspension through a 70 μm cell strainer. The single cell suspensions were placed on ice and used immediately to isolate the EVs or perform flow cytometry.

Flow cytometry

Peripheral white blood cells of mice were obtained after lysis of peripheral blood with red blood cell lysis buffer (AAPR27-A500, Pythonbio). Peripheral white blood cells, single-cell suspensions of LV tissues, and BMDNs subjected to different treatments were washed with pre-chilled PBS. To label different neutrophil phenotype, they were incubated with PE-tagged anti-CD11b, APC-tagged anti-Ly6G and Alexa Fluor 700-tagged CD206 antibodies (all from BD Biosciences, USA) in pre-chilled PBS containing 1% BSA for 15 min on ice in the dark. In addition, peripheral white blood cells and single-cell suspensions of LV tissues were incubated with PerCP/Cyanine5.5-tagged anti-CD3, FITC-tagged anti-CD4, Alexa Fluor 647-tagged anti-CD8 and PE-tagged anti-NK-1.1 (all from BioLegend, USA) to label different lymphocyte subsets (CD4+ T cells, CD8+ T cells, and natural killer cells). Flow cytometry analysis was performed using the FACScan flow cytometer (BD Biosciences, USA). The flow cytometry data was analyzed using the FlowJo version 7.6.1. software.

In vitro transfection and in vivo administration

The miRNA mimic, inhibitor, antagomir, and agomir used in this study were synthesized by Ribobio (Guangzhou, China). BMDNs were isolated from the mouse bone marrow and seeded in 6-well plates. After 3 h, they were transfected with the miRNA mimic or inhibitor using the RNA transfection reagent (HB-RF-1000, HANBIO, China). The transfection efficiency was analyzed by real-time quantitative polymerase chain reaction (RT-qPCR). The primary neonatal CMs were isolated from the hearts of 1-day-old C57BL/6 mice, and CMs were transfected with the miRNA mimic or inhibitor using the RNA transfection reagent. For knockdown or overexpression of miR-9-5p in vivo, the reperfusion region was injected immediately with antagomiR-9-5p (3 mg/kg) or agomiR-9-5p (1 mg/kg) after removing the silk suture [29].

Sequencing analyses of EV-derived miRNAs

Total RNA was extracted from the normal control (NC) or H/R-treated CM-derived EVs using the exoRNeasy Maxi Kit (Qiagen, Valencia, CA, USA). MiRNA sequencing was performed using the Illumina NextSeq 500 system (Aksomics, China). The differentially expressed miRNAs were screened based on the expression change ratio ≥ 1.5 or ≤ − 1.5 and P value < 0.05 as threshold parameters. The miRanda, miRTarbase, Targetscan, and miRDB databases were screened to identify the candidate target genes of miR-9-5p. The KOBAS-i (http://bioinfo.org/kobas), an online web tool, was used for gene functional annotation of the candidate target genes of miR-9-5p [30].

RNA pull-down assay

The probe-coated beads were generated by incubating the streptavidin-coupled magnetic beads (15942-050, Invitrogen) with the biotin-coupled miR-9-5p probes and the oligo probes overnight at 4 °C. Then, the cell lysates were prepared from approximately 1 × 107 BMDNs and incubated with the probe-coated beads for 1 h at room temperature. The probe-coated beads were then washed with the buffer and RT-qPCR was performed to detect the pulled down RNAs.

Real-time quantitative polymerase chain reaction

Total RNA was extracted from the cell or tissue lysates using the TRIzol reagent (R6830-01E.Z.N. A, OMEGA). Total RNA was extracted from the EVs using the exoRNeasy Maxi Kit (77164, QIAGEN). Total RNA was transcribed into cDNA using the PrimeScript RT Master Mix (Takara, Japan) or the miRNA 1st strand cDNA synthesis kit (AG11717, AG) according to the manufacturer’s instructions. Then, qPCR analysis was performed with the SYBR Green PCR Master Mix (Takara, Japan) using the LightCycler 480 System (Roche, Germany). The relative expression levels of the miRNAs were normalized to the expression levels of U6 RNA. The expression levels of target mRNAs were normalized to the expression levels of GAPDH mRNA. The relative expression levels were calculated using the 2−∆∆Ct method. The primers used for the RT-qPCR analysis in this study are listed in Tables S1, S2.

Western blotting

Total protein extracts were prepared from the EVs, cells, and LV tissues using the RIPA Lysis buffer (BL504A, Biosharp, China) supplemented with the Protease Inhibitor Cocktail (C0001, Pythonbio) and the Phosphatase Inhibitor Cocktail (AAPR176-1, Pythonbio) according to the manufacturer’s instructions. Total protein concentrations were estimated using the BCA assay. Equal amounts of protein samples were separated by 10–12% SDS-PAGE electrophoresis. The separated proteins were transferred onto the PVDF membranes. The membranes were then blocked with 5% BSA. Subsequently, the membranes were incubated overnight at 4℃ with the following primary antibodies: anti-RIP3 (505431, Zen-bioscience, China), anti-MLKL (612662, Zen-bioscience, China), anti-pRIP3 (R30283, Zen-bioscience, China), anti-pMLKL (R30265, Zen-bioscience, China), anti-CD9 (AF1192, Beyotime, China), anti-CD81 (ab79559, Abcam, USA), anti-TSG101 (AF8259, Beyotime, China), anti-TNF-α (17590-1-AP, Proteintech, USA), anti-IL-1β (26048-1-AP, Proteintech, USA), anti-IL-6 (bs-0782R, Bioss, China), anti-IL-10 (82793-16-RR, Proteintech, USA), anti-Arg1 (16001-1-AP, Proteintech, USA), anti-TGF-β1 (26155-1-AP, Proteintech, USA), anti-SOCS5 (bs-13664R, Bioss, China), anti-SIRT1 (bs-0921R, Bioss, China), anti-JAK2 (YT2426, Immunoway, USA), anti-STAT3 (10253-2-AP, Proteintech, USA), anti-p65 (80979-1-RR, Proteintech, USA), anti-pJAK2 (YP0155, Immunoway, USA), anti-pSTAT3 (bs-1658R, Bioss, China), anti-acetylp65 (bs-23216R, Bioss, China), anti-β-actin (20536-1-AP, Proteintech, USA), and anti-GAPDH (10494-1-AP, Proteintech, USA). Next, the membranes were incubated with the HRP-conjugated anti-rabbit IgG secondary antibody (sc-2004, 1:10,000, SantaCruz, USA) for 1–2 h at room temperature. Then, the protein bands were developed using ECL and analyzed using the ImageJ software.

Serum EV-derived miR-9-5p in patients with STEMI undergoing PCI

Patients with STEMI undergoing PCI were enrolled in this study from Guangdong Provincial People’s Hospital, and the diagnostic criteria for STEMI were in accordance with the 2013 American College of Cardiology Foundation/American Heart Association guideline [31]. Patients were excluded when: (1) patients with cardiogenic shock; (2) pregnant women; (3) patients with severe liver dysfunction or end-stage nephropathy; (4) patients with severe organic disease, infection, or inflammatory disease; (5) patients with malignancy. The whole blood samples (10 mL from peripheral venous blood) were collected from enrolled patients within 24 h after PCI. The whole blood samples were centrifuged at 500×g for 8 min at 25 °C to acquire serum, and serum EVs were subsequently isolated using ultracentrifugation. Then, the serum EV-derived miR-9-5p levels were detected by RT‐qPCR analyses. The primary endpoint was follow-up cardiovascular mortality, which was defined as the deaths attributed to heart diseases and cerebrovascular diseases. The secondary endpoints were follow-up all-cause mortality and major adverse cardiovascular events (MACEs, defined as the composite of all-cause death, stroke, recurrent myocardial infarction, and target vessel revascularization). All clinical adverse events were obtained through telephone interviews conducted by trained nurses or physicians.

Statistical analysis

The data are represented as the mean ± standard deviation (SD) from at least three independent experiments. The data was compared between two groups using the unpaired t-test and for more than 2 groups using the one-way analysis of variance (ANOVA). The variables were analyzed at different time points using the Bonferroni-corrected repeated measures ANOVA. Univariate and Multivariate Logistic regression models were constructed to explore the associations between serum EV-derived miR-9-5p levels and the incidence of clinical adverse events. Non-linear correlations were explored using restricted cubic splines. Receiver operating curve (ROC) analysis was used to determine the prognostic value of serum EV-derived miR-9-5p. All the statistical analyses were performed using the SPSS 21.0 software (IBM, Armonk, NY, USA) and the R software (version 4.2.2). Two-tailed P < 0.05 was considered as statistically significant.