We used 6-week-old male C57BL/6 mice (20–25 g) purchased from SPF (Beijing) Biotechnology Co., Ltd. [License No.: SCXK (Beijing) 2019–0010] in this study. The mice were housed in a temperature-controlled environment (21 ± 2 °C) with a 12-h light/dark cycle (lights on at 06:00) and ad libitum access to food and water. To establish the type I diabetes mouse model, the mice (n = 12) were intraperitoneally administered streptozotocin (STZ) (50 mg/kg) in citrate buffer (0.05 mol/L; pH 4.5) for 5 d; control mice (n = 30) received an equivalent amount of citrate buffer solvent. Two weeks after the initial administration of STZ, the mice with ≥ 16.7 mM (300 mg/dL) blood glucose level were considered diabetic and included in the diabetic cohort [43].

The mice were anesthetized using isoflurane (2%; O2 2 L/min), and the surgery was performed after confirming that the pedal reflexes of the mice were absent. The mice underwent left thoracotomy at the fourth intercostal space; their hearts were smoothly and gently extracted, and the left anterior descending branch of the coronary artery was distally ligated to its main bifurcation using a 7 − 0 ophthalmic suture. The success of the coronary occlusion was confirmed based on the pallor and regional-wall motion abnormality of the left ventricle or ST-segment elevation ≥ 0.25 mV on an ECG monitor [44]. Sham mice underwent the same time-matched surgical procedure without a ligation step. For cardiac-specific expression or silencing of TFPI2 in the MI mice, adenovirus (2 × 109 plaque-forming units (PFU/L, 30 µL) was directly injected into the myocardium at three positions along the margin of the ischemic area when inducing MI; another 2 × 109 PFU/L of adenovirus was injected into the tail vein after 1 week [45, 46]. Each mouse was injected with adenovirus carrying TFPI2 cDNA or the empty vector, sh-TFPI2, or sh-NC.

Consequently, there are seven groups of mice in this study (n = 6 for each group), including Sham group, MI mice with sh-TFPI2 knockdown (MI + sh-TFPI2), MI mice transfected with sh-NC (MI + sh-NC), MI mice with TFPI2 overexpression (MI + TFPI2), MI mice transfected with empty vector (MI + vector), diabetic MI mice with TFPI2 overexpression (DM MI + TFPI2), and diabetic MI mice transfected with cDNA vector (DM MI + vector). The mice were euthanized at 3 weeks post MI according to their groups. The study protocol was approved by the Ethics Committee of Qingdao University School of Medicine (QYFY WZLL 27,656, Qingdao, China).

In the 3rd week after MI surgery, transthoracic echocardiography was performed using the Vevo2100 imaging system (Visual Sonics, Toronto, Canada). Moreover, 2D echocardiography and M-mode echocardiography were used to measure EF, LV, FS, LVDd, LVAWd, and LVPWd. All measurements were performed by the same observer, and the values were averaged over five consecutive cardiac cycles [47].

BMDMs were isolated from C57BL/6 mice and cultured in Dulbecco’s modified Eagle medium (DMEM) containing 20 ng/mL mouse macrophage-colony-stimulating factor for 5–7 d to induce M0 macrophages. CFs were isolated by enzymatic digestion and cultured in full DMEM as previously described [16]. The CFs isolated from male neonatal C57BL/6J mice (1–3 d) were cultured under normal glucose conditions with 10% foetal bovine serum till the second generation before treatment.

Next, the cells were transfected with TFPI2 cDNA or the empty vector, sh-TFPI2, or sh-NC, following which the cells were incubated in a medium with 5 mM D-glucose (normal control, NC), 5 mM D-glucose + 27.5 mM mannose (osmotic control, OC), 33 mM D-glucose (HG), or 33 mM D-glucose + 3 µM PPARγ antagonist (GW9662, Selleck Chemicals, United States) for 48 h [48].

The mouse hearts were harvested and fixed in 4% formalin for at least 24 h, followed by paraffin embedding and sectioning (5 μm). Inflammatory cell infiltration was assessed using haematoxylin-eosin staining (H&E), and the extent of post-infarct fibrosis was assessed using Masson’s trichrome staining. For immunohistochemistry, the slides were incubated overnight at room temperature with primary antibodies against collagen I (1:50, ab138492, Abcam), collagen III (1:50, ab184993, Abcam), MMP2 (1:50, ab92536, Abcam), and MMP9 (1:50, ab283575, Abcam). For immunofluorescence analysis, the sections were simultaneously labelled using unconjugated primary antibodies against CD68 (1:200, 14-0681-82, eBioscience) and the M1 markers CD86 (1:200, ab242020, Abcam)/iNOS (1:500, ab129372; Abcam), or M2 markers CD206 (1:200, ab64693, Abcam)/Arg-1(1:500, ab91279; Abcam) and incubated overnight, followed by incubation with a fluorophore-conjugated secondary antibody for 30 min. The stained sections were mounted using DAPI-containing Vector Shield mounting medium (Vector). All pathological sections were scanned and photographed using the Qingdao University Hospital Digital Cross Section Scanning System. Data were collated for analysis using the ImageJ software.

The cells or extracted heart tissue samples were homogenized using RIPA lysis buffer (Elabscience) and centrifuged at 15,000 × g for 10 min at 4 °C. The supernatant was collected, and the protein concentration was determined using a BCA assay kit (Thermo Fisher Scientific). The proteins were separated using 10% sodium dodecyl sulphate polyacrylamide gel electrophoresis and transferred onto nitrocellulose membranes (Merck Millipore, USA). The membranes were blocked with 5% skim milk for 2 h and incubated overnight at 4 °C with the primary antibodies against the following proteins: TFPI2 (1:1000, ab186747; Abcam), iNOS (1:1000, ab129372; Abcam), Arg-1 (1:1000, ab91279; Abcam), PPAR-γ (1:1000, Santa Cruz), collagen I (1:50, ab138492, Abcam), collagen III (1:50, ab184993, Abcam), MMP2 (1:50, ab92536, Abcam), MMP9 (1:50, ab283575, Abcam), and β-actin (1:4000, 8H10D10, CST). Horseradish peroxidase (HRP)-labelled goat anti-rabbit IgG (1:10,000, Absin, Shanghai, China) was used as the secondary antibody and incubated for 1 h at room temperature. Target bands were visualized using chemiluminescent ECL (Merck Millipore, USA) and detected using an Amersham Imager 600 (GE Healthcare, Little Chalfont, UK). The images were analysed using the ImageJ data acquisition software.

We investigated cell migration using wound healing and Transwell assays. For the wound healing assay, CFs were seeded in six-well plates. Wounds were made through the cell monolayer using 1000-mL plastic tips after the cells were incubated for 12 h, and images were captured using a Nikon Ti-S inverted phase-contrast microscope (Nikon, Tokyo, Japan) at 24 h to calculate the healing rates.

For the Transwell assay, the cells (1 × 105 cells) were trypsinized and seeded into the upper chambers (8-mm pores, 24-well, Corning Life Sciences, Corning, NY, USA) in FBS-free DMEM (200 mL). The lower chambers contained DMEM supplemented with HG or the OC, and incubation was performed for 24 h. Next, the cells that migrated into the lower chambers were fixed using 4% paraformaldehyde after removal of the medium and stained with crystal violet for 0.5 h. Images of five stochastic fields per membrane were obtained using a Nikon Ti-S inverted phase-contrast microscope.

All experiments were repeated at least three times, and data are presented as mean ± SD (standard deviation). Statistical analysis was performed using ANOVA followed by Tukey’s post hoc test (GraphPad Prism 9, USA). P < 0.05 was considered significant.