Eight-week-old male Sprague Dawley rats, weighing 220–250 g, were housed under standard conditions at the experimental animal center of Zhengzhou University (Zhengzhou, Henan, China) at room temperature (25°C), with a humidity of 40 ± 60% and 12-h light-dark cycles. Rats were fed a standard diet and water ad libitum. All animal experiments were approved by the Animal Experiments Committee of Zhengzhou University and performed according to the guideline for the Care and Use of Laboratory Animals (NIH Publication, No. 85–23, revised 1996).

The protocol for the Sprague Dawley rat MI model and drug rationale (SB216763, a potent irreversible and cell-permeable pharmacological inhibitor of GSK-3 which is highly selective for GSK-3β and has no significant influence on the activity of other kinases (Coghlan et al., 2000), 0.6 mg/kg⁻¹, administered intravenously 1 h before surgery) were based on our recent report (Wang et al., 2020). Rats were randomly divided into three groups (n = 10 each) namely Sham, MI and SB + MI. Following the loss of corneal reflex in all rats, we opened the thoracic cavity and ligated the left anterior descending artery (Sham group: The thread was inserted without ligation after SB216763; the SB + MI group: SB216763 was injected through the tail vein 1 h before ligation). Three heart samples were fixed in 4% formalin for histopathological examination, and seven samples were used for quantitative real-time (qPCR) and western blotting analyses at 7 days.

Cardiac function was evaluated at seven post-operation by transthoracic echocardiography. The left ventricular ejection fraction (LVEF), left ventricular fractional shortening (LVFS), left ventricular end-diastolic (LVED), and stroke volume (SV) were calculated using M-mode tracing (Vevo 2,100; Visual Sonics, Toronto, Canada).

Hearts were individually excised and immediately immersed in 4% formaldehyde for 24 h. After fixation and paraffin-embedding, 4-μm-thick sections were cut and stained with H&E for overall morphological evaluation using an optical microscope (BX60; Olympus, Japan). Image acquisition and analysis were performed using ImageJ Launcher (National Institutes of Health, Bethesda, United States). All measurements were performed in a double-blinded manner by two independent researchers.

Immunohistochemistry staining of paraffin sections was performed using a microwave-based antigen retrieval method. The heart tissues were fixed in 4% paraformaldehyde and embedded in paraffin. The sections were subsequently cut at 6-μm intervals perpendicular to the long axis of the heart. The primary antibody-rat CD47 (1:200, Abcam, Cambridge, MA, United States) was visualized using Alexa Fluor 550 secondary antibody (1:200, Servicebio, Wuhan, China). The primary antibody-rat CD68 (1:200, Abcam) for macrophages detection was visualized using HRP-labeled secondary antibody (1:200, Servicebio). The nuclei were stained with 4′, 6-diamidino-2-phenylindole dihydrochloride (Invitrogen, New York, United States). CD47/CD68-positive cells per square millimeter were counted under a 20× power field of the microscope in five random areas of LV tissues.

For cell apoptosis assays, the FITC-Annexin V apoptosis detection kit (Tianjin Sungene Biotech Co., Ltd.) was used according to the manufacturers’ instructions. Briefly, cells (2 × 10⁵ cells/plate) were incubated in 6-well plates for 48 h. Cells were subsequently collected by mild trypsinization, stained with FITC-Annexin V and propidium iodide on ice for 5 min, and subjected to flow cytometric analysis using analytical flow cytometry (BD FACSymphony™ A5, New Jersey, United States).

The H9c2 (rat embryonic ventricle) cell line was purchased from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China) and cultured in Dulbecco’s modified Eagle medium (DMEM, Corning Inc., Corning, NY, United States of America) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin.

Rat neonatal cardiomyocytes were harvested according to a previously described method (Wang et al., 2020). Briefly, heart ventricles from newborn rats were used within 24 h of birth. The ventricles were minced and digested with 0.1 mg/ml trypsin (Solarbio, Beijing, China) and 0.1 mg/ml collagenase II (Worthington, Lakewood, NJ, United States). The cell suspensions were plated for 2 h at 37°C to separate rat cardiac fibroblasts from RCMs. The supernatant containing RCMs was collected and inoculated into a 6-well plate, which was treated experimentally 24 h later. The cells were then cultivated in DMEM with 10% FBS and 1% penicillin/streptomycin.

Rat GSK-3β siRNA and negative control siRNA were obtained from Thermo Fisher Scientific (Invitrogen). H9c2 cells were transiently transfected with 40 pM siRNA using 3.75 µL lipofectamine RNAiMAX (Invitrogen), following the manufacturer’s instructions.

An anaerobic workstation was maintained to establish the cell hypoxia model (Research Scientific Services, Derwood, MD, United States). RCMs or H9c2 cells were placed in the transfer chamber and exposed to high-purity N₂. Then, the cells were transferred to the anaerobic working chamber, followed by introduction of mixed gas (10% H₂ + 10% CO₂ + 80% N₂). The cells were subsequently divided into three groups, control (CT); hypoxia model, hypoxia model with SB216763 (10 µM) (Kirby et al., 2012), or SN50 (an NF-κB inhibitor, 18 μM) (Zhao et al., 2013), or CD47 monoclonal antibody (10 μM) (Mohanty et al., 2019) pretreatment for 1 h before hypoxia was performed at various time points (6, 12, and 24 h).

Total RNA was isolated from rat heart tissues using TRIzol (Roche, Germany) and reverse transcribed, following the manufacturer’s protocol (RR047A, Takara Bio, China). Subsequently, qPCR was performed using SYBR Green Master mix (Thermo Fisher Scientific, United States) on the 7,500 Fast Real-Time PCR system (Applied Biosystems; Thermo Fisher Scientific, United States). The thermocycling conditions were as follows: holding at 50°C for 2 min; pre-denaturation at 95°C for 2 min, followed by 15 s at 95°C and then 1 min at 60°C for 40 cycles. The melt curve was at 95°C for 15 s, 1 min at 60°C, 15 s at 95°C, and 15 s at 60°C. The mRNA levels were normalized to GAPDH levels. The relative expression was calculated using the change-in-quantification (2^−ΔΔCT) method. The following primers were used: rat CD47: forward, 5′-GCT​TGC​TGG​ATA​CCC​CTG​TT-3′; reverse, 5′-TGC​ATA​GGA​AGT​AGG​CGT​GAG-3′; rat NF-κB: forward, 5′-TCT​TGA​GGT​GGC​TGC​TTA​CC-3′; reverse, 5′- CAC​CGT​GTT​CAT​TCC​AGT​GTC-3′; rat GAPDH: forward, 5′-TCC​CTC​AAG​ATT​GTC​AGC​AA-3′; reverse, 5′-AGA​TCC​ACA​ACG​GAT​ACA​TT-3′.

Samples were solubilized in ice-cold RIPA lysis buffer (Solarbio, Beijing, China) containing a protease inhibitor cocktail (MedChemExpress, Shanghai, China). The protein concentrations were determined using the bicinchoninic acid method (Beyotime, Shanghai, China). Membrane proteins were subjected to SDS-PAGE (Solarbio, Beijing, China) and electrophoretically transferred onto polyvinylidene fluoride membranes (Millipore, United States) and blocked with 5% non-fat dry milk in tris-buffered saline containing 0.1% tween-20 (TBST) for 1 h before incubation with rabbit anti-CD47 antibody (1:1,000, Abcam), anti-NF-κB (p65; 1:1,000, Cell Signaling, Danvers, MA, United States), anti-p-NF-κB (p-p65; 1:1,000, Cell Signaling), anti-Bcl-2 (1:1,000; Proteintech, Rosemont, IL, United States), anti-GSK-3β (1:1,000, Cell Signaling), anti-p- GSK-3β (Ser9; 1:1,000, Cell Signaling), anti-Bax (1:1,000; Proteintech), anti-caspase-3 (1:2,000; Invitrogen), and mouse anti-GAPDH (1:10,000, Proteintech) in 5% non-fat dry milk/TBST overnight. Then, the membranes were incubated with alkaline phosphatase conjugated Affinipure goat anti-rabbit IgG (H + L; 1:10,000, Proteintech) in TBST for 2 h at 37°C. The membranes were visualized using an ECL detection kit (Pierce Biotech, Rockford, IL, United States). The blots were analyzed and quantified using the ImageJ analysis software.

Whole-cell lysates were obtained through RIPA buffer lysis and incubated with anti-CD47/anti-p65 at 4°C for 12 h, and with protein A/G magnetic beads (Invitrogen) for an additional 2 h. After three washes with cold PBS, the immunocomplexes were analyzed using western blotting.

Data are presented as mean ± SD. All data were analyzed using SPSS 21.0 (SPSS Inc., Chicago, IL, United States). Statistical comparisons were conducted using unpaired t tests between two groups. Statistical differences between multiple groups were compared using ANOVA test followed by Bonferroni post hoc tests. p values <0.05 were considered statistically significant. This study followed the principles of blinding and randomization.

As shown in Figure 1A, the heart cells of the Sham group were regular in shape and arranged in a tight, orderly configuration. Alternatively, the myocardial structure of the MI group exhibited significant damage, with a loose cell arrangement, broken fibers and blood-filled intercellular spaces. SB216763 pretreatment significantly reduces the myocardial damage caused by the above-mentioned MI.

The number of CD68⁺ cells (a marker of macrophages) in the MI group showed increased compared to those in the Sham group (Supplementary Figure S1A). SB216763 pretreatment significantly reduces the number of CD68⁺ cells (Supplementary Figure S1B).

We then assessed cardiac function after MI using echocardiography (Figure 1B). LVEF (Figure 1C), LVFS (Figure 1D), and SV (Figure 1F) were decreased in the MI group at 7 days post-surgery, while SB216763 treatment significantly reversed these effects. Additionally, the LVED (Figure 1E) was significantly increased in the MI group and decreased by SB216763.

To explore the effect of GSK-3β on CD47 expression in MI tissues, we determined CD47 expression by performing immunofluorescence assays (Figure 2A). The number of CD47-positive cells in the MI group was higher than that in the Sham group, however, decreased after SB216763 pretreatment (Figure 2B). CD47 mRNA was also significantly upregulated in ischemic tissues compared to the Sham group. SB216763 pretreatment significantly decreased CD47 mRNA levels compared to those in the MI group (Figure 2C). Similarly, SB216763 pretreatment also decreased the upregulated CD47 protein expression induced by MI (Figures 2D,E).

To further investigate the effects of GSK-3β on CD47 expression, we used RCMs and H9c2 cells to establish an in vitro cell hypoxia model. Western blotting results showed that hypoxic stimulation increased the expression of CD47 protein, while SB216763 pretreatment decreased its expression in RCMs (Figures 3A,B) and H9c2 cells (Figures 3H,I). In addition, GSK-3β inhibition downregulated apoptosis-related proteins (caspase-3 and Bax/Bcl-2) in RCMs (Figures 3C–E) and H9c2 cells (Figure 3J–L) under hypoxic conditions. Moreover, flow cytometric analysis showed that SB216763 pretreatment reduced the number of apoptotic RCMs (Figures 3F,G) and H9c2 cells (Figures 3M,N).

Next, we explored whether GSK-3β affects CD47 expression after MI via NF-κB signaling. NF-κB represents a group of structurally related and evolutionarily conserved proteins, with five members in mammals, namely Rel (c-Rel), RelA (p65), RelB, NF-κB1 (p50 and its precursor p105), and NF-κB2 (p52 and its precursor p100), forming homo- or heterodimers that bind the IκB family of proteins in unstimulated cells (Ghosh and Karin, 2002). When cells become stimulated, NF-κB is activated and translocates to the nucleus, through the exposed nuclear localization signal (subunit p65), where it functions to regulate inflammation, cell proliferation, and apoptosis (Silverman and Maniatis, 2001). We, therefore, sought to primarily quantify the expression of the p65 subunit.

As shown in Figure 4A, NF-κB mRNA expression increased 7 days after MI, while SB216763 pretreatment reduced this effect. Moreover, ischemic tissues, induced by MI, exhibited an upregulated p-p65/p65 ratio and NF-κB protein levels compared with those in the Sham group. This effect was reversed by SB216763 pretreatment 7 days after MI (Figures 4B,C). Similarly, the level of NF-κB protein increased under hypoxic conditions in vitro. Meanwhile, SB216763 pretreatment was found to effectively decrease the upregulated p-p65/p65 ratio and NF-κB protein expression in RCMs (Figures 4D,E) and H9c2 cells (Figures 4F,G).

Subsequently, SN50 pretreatment decreased CD47 and NF-κB protein expression (Figures 5A–C), without significantly affecting p-GSK-3β/GSK-3β (Figures 5D,E) under hypoxic conditions. Moreover, SN50 pretreatment reduced apoptosis-related protein expression in RCMs (Figures 5F–H), which was confirmed via flow cytometric analysis (Figures 5I,J).

In addition, SN50 elicited similar effects on H9c2 cells in inhibiting CD47 protein upregulation (Figures 6A–C), apoptosis-related proteins (Figures 6F–H), and cells apoptosis (Figures 6I,J) under hypoxic conditions. Meanwhile, SN50 pretreatment did not affect upregulation of hypoxia-induced GSK-3β protein expression (Figures 6D,E). Co-immunoprecipitation further demonstrated the protein interaction between CD47 and p65 NF-κB (Figure 6K). Hence, consistent with our assumption, inhibition of GSK-3β would reduce its interaction with p65.

To examine whether basal GSK-3β depletion exerts protective effects similar to those observed for SB216763, GSK-3β expression was knocked down through transfection with siRNA (siRNAGSK-3β-1, 2, and 3) in H9c2 cells. GSK-3β siRNA knockdown was confirmed by the significant depletion of GSK-3β (Figures 7A,B) and we selected siRNAGSK-3β-1 for further experiments. As shown in Figures 7C–F, knockdown of GSK-3β also decreased NF-κB and CD47 protein upregulation in H9c2 cells.

RCMs and H9c2 cells were treated with anti-CD47 antibody alone or in combination with SB216763 to evaluate the effect on cell apoptosis. When compared with the NC and 12-h-hypoxia groups, we observed that treatment with anti-CD47 antibody alone sensitized RCMs (Figures 8A–E) and H9c2 cells (Figures 8F–J), while combined anti-CD47 antibody and SB216763 pretreatment exhibited an additive effect on cell apoptosis inhibition.