An injectable SFMA/HA-E hydrogel for sustained delivery of hucMSC-derived exosomes promotes myocardial repair after infarction Authors: Gang Deng, Shuoji Zhu, Qing Ouyang, Xuyu He, Yu-Sen Ou, Haotao Su, Linhui Jiang, Jinsong Huang, Xiang Long, Min Wu, Ge Li, Changjiang Yu, Chen Lu, Ping Zhu, Moussa Ide Nasser Published: 2026 ABSTRACT BackgroundMyocardial infarction (MI) causes permanent loss of cardiomyocytes and heart failure. Human umbilical cord mesenchymal stem cell-derived exosomes (hucMSCs-Exos) have shown therapeutic potential. Still, their clinical utility is limited by rapid systemic clearance, highlighting the need for a delivery system to achieve sustained release and effect.MethodsWe fabricated an injectable, photopolymerizable hydrogel consisting of silk fibroin methacrylate (SFMA) and epigallocatechin gallate-grafted hyaluronic acid (HA-E) to encapsulate hucMSC-Exos (SFMA/HA-E+Exos). Therapeutic effects were assessed in a rat MI model using echocardiography, histology, and molecular analysis.ResultsThe SFMA/HA-E hydrogel exhibited favorable mechanical properties, controllable degradation, and sustained exosome release. In vitro, hucMSC-Exos promoted cell survival, migration, and tube formation under hypoxic conditions. In vivo, a single intramyocardial injection of SFMA/HA-E+Exos significantly improved cardiac function compared with the PBS-treated MI group (LVEF: 70.71 ± 3.04% vs. 30.75 ± 3.55%), reduced fibrosis, and suppressed cardiomyocyte apoptosis. The treatment also reinforced angiogenesis, as indicated by increased numbers of CD31⁺ and α-SMA⁺ vessels, induced a pro-reparative immune microenvironment by shifting macrophages polarization towards the M2 phenotype, and up-regulated the gap junction protein Connexin 43.ConclusionThe SFMA/HA-E hydrogel-mediated delivery of hucMSC-Exos provides potent dual therapy through structural support and bioactive signaling, inhibiting adverse remodeling and inducing cardiac repair after MI, and represents a potential therapeutic approach for ischemic heart disease.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13287-026-05097-z. Introduction Myocardial infarction (MI) is a worldwide health problem with high mortality, featuring acute occlusion.and irreversible loss.The subsequent wound healing process is frequently accompanied by the appearance of a non-contractile fibrotic scar, which leads to deleterious ventricular remodeling, abnormal electrical conduction, and ultimately heart failure [1–3]. Although there have been significant improvements in reperfusion-based therapies that can rescue the ischemic border zone, none can regenerate lost myocardium, underscoring an unmet need for novel therapeutic approaches. Stem cell-based therapy represents a potential advancement in cardiac repair. Among the various cell types, mesenchymal stem cells (MSCs) have been widely studied for their ease of isolation, multilineage potential, and decisive paracrine action [4, 5]. hucMSCs are especially appealing for their robust proliferative potential, low immunogenicity, and potent immunomodulatory effects [6, 7]. It is now widely acknowledged that the beneficial effects of MSCs are predominantly mediated by secreted paracrine factors, in particular exosomes. These small extracellular vesicles deliver bioactive proteins, lipids, and nucleic acids into target cells [8]. hucMSC-Exos have been reported to attenuate ischemia-reperfusion injury, promote angiogenesis, and regulate inflammatory responses [9, 10]. Nevertheless, direct injection suffers from rapid systemic clearance, necessitating repetitive doses that diminish therapeutic activity. Biomaterial-based delivery systems provide an elegant solution to this limitation. Injectable hydrogel can be applied to the target area in a minimally invasive manner. It may temporarily serve as a three-dimensional sponge and mechanical support, assisting in the formation of the infarct wall while providing sustained release of the therapeutic payload [11, 12]. In this work, we developed a new hydrogel platform combining two functional biopolymers, silk fibroin methacrylate (SFMA) and epigallocatechin gallate-grafted hyaluronic acid (HA-E). Silk fibroin (SF), known for its superior mechanical strength and biocompatibility, was functionalized with methacrylate groups to enable photopolymerization [13, 14]. Hyaluronic acid (HA) is a naturally occurring polysaccharide distributed across a broad range of organisms, from lower prokaryotes such as bacteria to complex eukaryotes [15]. Alongside collagen, HA is one of the most abundant components of the extracellular matrix (ECM). When implanted into infarcted myocardial tissue, HA-based scaffolds improve both structural and functional cardiac parameters through multiple mechanisms, including promoting cell survival, attenuating inflammatory responses, enhancing neovascularization, and facilitating the retention or delivery of bioactive components, such as exosomes and stem cells [16]. As a major glycosaminoglycan in the native cardiac ECM, HA also contributes to improved biocompatibility and enhanced cell–matrix interactions [17]. To further potentiate these beneficial effects, HA was functionalized with epigallocatechin gallate (EGCG), a naturally occurring polyphenol derived from green tea. EGCG possesses well-documented anti-inflammatory and antioxidant properties that may synergize with exosome-based therapies to mitigate the hostile post-infarct microenvironment [18, 19]. Another natural polyphenol, gallic acid (3,4,5-trihydroxybenzoic acid), is obtained by acid hydrolysis of hydrolyzable tannins. Previous investigations have reported a range of therapeutic properties for gallic acid, including antibacterial, antiviral, and anti-inflammatory activities [20–22]. We hypothesized that the SFMA/HA-E hydrogel would provide a supportive scaffold for the infarcted myocardium while enabling localized and sustained delivery of hucMSC-Exos. We hypothesized that this hybrid strategy would engineer a multilevel reparative response targeting cardiomyocyte survival, angiogenesis, fibrosis, and inflammation. Here, we synthesized and characterized the SFMA/HA-E hydrogel, demonstrated its feasibility as an exosome delivery platform, and revealed its remarkable therapeutic efficacy and associated cellular and molecular changes in a rat MI model. Materials and methods Isolation and characterization of huMSCs huMSCs were isolated from the Wharton’s jelly of umbilical cord tissue from healthy full-term fetuses in accordance with institute-approved ethical guidelines (No. S2025-613-02). Briefly, the umbilical cord tissue was washed with sterile PBS, disinfected using 75% alcohol for 3 min, and cut into small segments. After separating the arteries and veins, the residual tissue was minced into pieces and placed in the cell incubator at 37 °C with 5% CO2. After about 2 weeks, huMSCs were observed to migrate from the tissues. hucMSCs at passages 3–6 were used in the experiments. hucMSCs were cultured in MEM-Alpha (Hyclone, USA) mixed with 2% fetal bovine serum (FBS, Hyclone, USA) and 0.1% penicillin-streptomycin (Gibco, USA). Surface marker expression, including CD44, CD73, CD90, CD105, CD34, CD45, CD11b, CD19, and HLA-DR (all from BD Biosciences), was analyzed as previously described [23]. Isolation and characterization of huMSCs-exosomes For exosome isolation, hucMSCs were cultured in exosome-depleted medium for 48 h. After the culture medium was gradient centrifuged to remove dead cells and cell debris, the supernatant was ultracentrifuged at 100,000 × g for 70 min (Beckman Colter, Optima XE-90), and the isolated exosomes were resuspended in 50 µL PBS (Gibco) for further use. Protein concentration was determined by BCA assay, and particle size/concentration was analyzed by nanoparticle tracking analysis (NTA, ZetaView). In vitro studies To establish in vitro hypoxia injury models of H9C2 (Cat: CBP60588, Cobioer Biosciences Co., Ltd., China) and EA.hy926 (Cat: CBP60714, same source) cells. H9C2 and EA.Hy926 cells were seeded into 96-well plates at a density of 5,000 cells/well in 100 µL of medium. After 24 h of culture, when cell confluence reached 70%-80%, the cells were divided into different treatment groups for an additional 24 h. The normal group was treated with 100 µL of high-glucose DMEM complete medium; the hypoxia group, with 100 µL of medium containing 400 µM CoCl₂; and the treatment group, with 100 µL of medium containing 400 µM CoCl₂ and hucMSC-Exo at 5 µg/mL (selected as the minimum concentration providing significant protection without saturation effects based on preliminary optimization experiments and published literature) [10]. After 24 h of treatment, the medium was replaced with fresh medium. Then, 10 µL of CCK-8 solution was added to each well, followed by incubation at 37 °C for 2 h, and the OD was measured at 450 nm. Scratch wound assay The migration-promoting effect of exosomes under hypoxic conditions was assessed using a scratch wound assay. H9C2 (Cat: CBP60588, Cobioer Biosciences Co., Ltd., China) and EA.hy926 (Cat: CBP60714, same source) cells were seeded in a 6-well plate at 90–95% confluency (n = 3). hucMSCs-Exos (50 µg/mL, based on the preliminary wound closure kinetics that indicated the optimum concentration for maximal migration without inhibition of cell proliferation [9, 10] ) was added to a medium containing 400 µM CoCl₂, and the cell monolayer was scraped with a pipette tip. The control group received medium only, and the hypoxia group received 100 µL of medium containing 400 µM CoCl₂. At 24 h, cell migration was observed under an Olympus CKX53 optical microscope. Quantitative analysis of the images was performed and expressed as the percentage of wound remaining. Tube formation assay EA.hy926 cells with a cell density of 1 × 106 cells/mL were grown for Matrigel. The normal group was treated with 100 µL of DMEM high-glucose complete medium; the hypoxia group, with 100 µL of medium containing 400 µM CoCl₂; and the treatment group, with 100 µL of medium containing 400 µM CoCl₂ and hucMSCs-Exos at 50 µg/mL. At 24 h, cells were observed under an Olympus CKX53 optical microscope. Exosome uptake assay hucMSC-Exo internalization was determined in H9C2 cardiomyoblasts using confocal laser microscopy. The hucMSC-Exo were labeled with PKH67 dye according to the manufacturer’s protocol and incubated with cells at 37 °C and 20 µg/mL for 3 h. Confocal microscopy with appropriate filters was used to visualize exosome uptake by cells. Preparation of SFMA/HA-E hydrogel SFMA/HA-E hydrogels were prepared at concentrations of 10%, 15%, and 20% (w/v) SFMA with a constant HA-E content of 0.5% (w/v). SFMA was synthesized by reacting a 1% (w/v) silk fibroin solution with 8% (v/v) glycidyl methacrylate at 60 °C for 3 h under stirring at 1000 rpm, followed by dialysis (MWCO 3500 Da) for 72 h and lyophilization. HA-E was synthesized by dissolving HA (100 kDa) in MES buffer (0.1 M, pH 5.5), activating with EDC/NHS for 2 h, then adding EGCG-EDA conjugate (synthesized separately) and stirring overnight at room temperature, followed by dialysis for 5 days and lyophilization. For hydrogel formation, SFMA and HA-E were dissolved in PBS containing 0.5% (w/v) LAP photoinitiator, and 100 µL of precursor solution was exposed to 366 nm UV light (15 mW/cm²) for 60 s in cylindrical molds. Characterization of SFMA/HA-E hydrogel The nuclear magnetic resonance (NMR) spectrum of the sample was recorded on a Bruker AVANCE III 600 MHz spectrometer. 5 mg of the sample was dissolved in DMSO‑d6 and transferred into an NMR tube. For Fourier transform infrared (FTIR) measurements, the sample powder was mixed with KBr, milled, and pressed into pellets under vacuum. FTIR spectra were obtained on a Thermo Scientific Nicolet iS10 FT-IR spectrometer (wavenumber range: 400–4000 cm⁻¹). For scanning electron microscopy (SEM) observation, the 10%, 15%, and 20% SFMA/HAE samples were frozen at -80 °C, lyophilized, and then coated with gold before SEM analysis using a Hitachi SU3500. The mechanical properties of the hydrogels were measured using a universal testing machine (AGS-X-50 N, Shimadzu Co., Japan) at a displacement rate of 1 mm min− 1, with a maximum strain of 90%. Both compressive force and strain were measured, and the slope of the straight line in stress-strain curves determined the compressive modulus. The rheological behavior of the hydrogel was determined using a Malvern Kinexus rheometer equipped with 25 mm parallel plates. The storage modulus (G’) and loss modulus (G’’) were recorded for 10%, 15%, and 20% SFMA/HAE hydrogels over 300 s at 1 rad/s. Dynamic strain sweeps (0.1–10% strain) were conducted to determine the linear viscoelastic region. The stability of the hydrogels was also assessed through 50 compression cycles (0–60% strain). For swelling analysis, 400 mg of the hydrogel was immersed in water at 37 °C for 15 min to determine the initial weight of the hydrogel (W₀), which was then placed in PBS buffer at different times (0, 1, 2, 3, and 4 h). The samples were removed at each time point, blotted dry with a paper towel, and weighed (Wt). The swollen ratio (Q) was determined by Q = (Wt - W₀)/W0 × 100%. Immersing hydrogel samples in deionized PBS or lysozyme solution followed by shaking at 37 °C (70 rpm) on days 1, 4, 7, 14, 21, and 28. Weight was determined after washing and lyophilization. The degradation rate (%) was calculated as (W₀ - Wt)/W₀ × 100%. Establishment of MI model This study was performed in compliance with the National Institutes of Health’s Guide for the Care and Use of Laboratory Animals and presented according to ARRIVE guidelines 2.0 [24]. Male specific pathogen-free Sprague-Dawley rats (275–300 g, 8–10 weeks old) were procured from Zhuhai Baitriantong Biotechnology Co., LTD. (Production License No.: SYXK(Guangdong), 2020-0051) and maintained under barrier conditions in the Laboratory Animal Centre of Guangzhou Huateng Biomedical Technology Company, LTD., China (NB: Male rats are used because they have reasonably stable physiological conditions as well as hormone levels). Anesthesia was achieved by subcutaneous injection of a combination of medetomidine hydrochloride, midazolam, and butorphanol (MMB) at doses of 0.3 mg/kg, 5.0 mg/kg, and 5.0 mg/kg, respectively [25]. MI was produced by ligation of the left anterior descending coronary artery. For all in vivo studies, the 20% SFMA/HA-E formulation was used based on optimal properties identified in characterization studies. The hydrogel was injected at multiple points along the infarct border using a 26-gauge needle, then exposed to 366 nm UV light for 60 s to induce gelation. Rats received either Exos alone or Gel+Exos. For the Gel+Exos group, hucMSC-Exos were concentrated and mixed with the hydrogel precursor solution to achieve a final exosome protein content of 150 µg per 100 µL of injectable mixture (based on a conversion factor of 2 × 10⁹ particles/µg protein, determined by correlating NTA and BCA). A 100 µL volume of this mixture was injected per rat. For the Exos-only group, the same total exosome dose (150 µg protein) was delivered in 100 µL of PBS [26, 27], showing efficacy in rat MI models. The chest was closed up with 6 − 0 sutures, and the muscles and skin layers were sutured with 4 − 0 silk. After surgery, conventional postoperative analgesia and infection management were administered. Rats were randomly divided into five groups (n = 12 per group): Sham, MI (PBS), MI+SFMA/HA-E (Gel), MI+hucMSC-Exos (Exo), and MI+hucMSC-Exo+SFMA/HA-E (Exo + Gel). At experimental endpoints, rats were euthanized by intraperitoneal injection of sodium pentobarbital (150 mg/kg body weight), followed by cervical dislocation to ensure death. All euthanasia procedures were performed by trained personnel in accordance with institutional guidelines. Inclusion and exclusion criteria Immediate ST-segment elevation on electrocardiography and blanching of the anterior left ventricular wall confirmed successful MI induction. Animals that died during surgery or within 24 h after the operation were excluded and replaced, since it was considered an acute surgical model. A total of 15 rats per group underwent MI surgery, resulting in a final sample size of n = 12 per group after exclusions. Blinding Surgeons performing the treatments were not blinded to the interventions, as the treatments were visibly distinguishable during and after surgery. All subsequent data acquisition and analyses, including echocardiography, histological staining, and quantitative image analysis, were performed in a blinded manner with respect to group allocation. Exosome labeling and in vivo tracking To assess the retention of exosomes in the myocardium, hucMSC-Exos were fluorescently labeled with the lipophilic dye PKH26 (Sigma‑Aldrich, USA) according to the manufacturer’s instructions with minor modifications. Briefly, exosomes (150 µg protein, equivalent to 3.0 × 10¹¹ particles) were suspended in 1 mL of Diluent C. Separately, 4 µL of PKH26 dye was mixed with 1 mL of Diluent C. The two solutions were combined and incubated at 25 °C for 5 min with gentle agitation. The labeling reaction was stopped by adding an equal volume of 1% bovine serum albumin (BSA) in PBS. Labeled exosomes were washed twice by ultracentrifugation at 100,000 × g for 70 min at 4 °C to remove unbound dye, and the pellet was resuspended in 100 µL of PBS (for the Exo‑only group) or mixed with the SFMA/HA-E hydrogel precursor solution (for the Gel + Exo group) immediately before injection. The final dose of PKH26‑labeled exosomes was 150 µg protein per rat in a total injection volume of 100 µL. At 1 and 7 days post‑MI, rats were euthanized, and hearts were harvested, embedded in optimal cutting temperature (OCT) compound, and snap‑frozen in liquid nitrogen. Cryosections (10 μm thickness) were cut from the infarct border zone, mounted on glass slides, and counterstained with DAPI (4′,6‑diamidino‑2‑phenylindole) to visualize nuclei. Fluorescence images were captured using a confocal microscope (Leica TCS SP8, Germany) with identical acquisition settings for all samples. The relative fluorescent intensity (RFI) of PKH26 in the infarct border area was quantified using ImageJ software (NIH, USA) by measuring the integrated density of the PKH26 signal in five random fields per section (three sections per animal). The mean RFI was calculated for each group and normalized to the background fluorescence of the corresponding negative control sections. Cardiac echocardiography The cardiac function was assessed on post-MI day 28 using high-resolution ultrasound (VisualSonics). Anesthesia was induced with 3.5% isoflurane and maintained with 2% isoflurane after the rats became unconscious. After being marked by ear tags, they were laid supine on a plate kept at a constant temperature. 2D echocardiographic images were obtained using B-mode imaging, and the long-axis view was positioned. The short-axis view was positioned by adjusting the papillary muscle section to 90° and activating M-mode for recording and measurement. Left ventricular end-diastolic and end-systolic dimensions were calculated from three cardiac cycles. Histology and immunofluorescence staining The hearts were quickly removed, fixed in 4% paraformaldehyde, and embedded in paraffin. Paraffin sections were then fixed and stained with masson’s trichrome or hematoxylin and eosin (H&E) according to the manufacturer’s instructions. Images were obtained using fluorescence microscopy (Leica DMI8, Germany), and infarct size was analyzed in ImageJ. TUNEL staining was performed with the Beyotime One Step TUNEL Apoptosis Assay Kit according to the manufacturer’s instructions. For immunofluorescence stainings, antibodies against Ki67 (Cat. No. GB121141), α-SMA (Cat. No. GB12045), and CD31 Cat. No. GB11063-2), cTnT (Cat. No. GB122408), Connexin 43 (Con43) (Cat. No. GB11234), Collagen I (Col I, Cat. No. GB11022-3 ), and Col III (Cat. No. GB111629) were obtained from Wuhan Servicebio Technology co., Ltd. Antibodies against CD206 (Cat. No. Ab64693) and CD68 (Cat. No. Ab31630) were obtained from Abcam. Dilutions of primary antibodies were prepared in blocking solution and incubated overnight at 4 °C. After staining with the corresponding secondary antibodies, the sections were counterstained with DAPI and examined by fluorescence microscopy. Western blotting Western blot was conducted according to previously published protocols [28]. The primary antibodies include anti-connexin 43 (Cx43, Servicebio, GB11234-100), anti-Bax (Servicebio, GB15690), anti-Bcl-2 (Servicebio, GB154830), anti-β-tubulin (Servicebio, GB152667), anti-cleaved caspase 3 (CST, 9664), and anti-cleaved caspase 9 (CST, 9505). The secondary antibodies used were anti-mouse HRP (Abcam, ab6789) and anti-rabbit HRP (Abcam, ab205718). Real-time quantitative PCR (RT-qPCR) analysis Cardiac tissue was homogenized in Trizol, and RNA was isolated. The resulting RNA was reverse-transcribed into cDNA using the cDNA Synthesis Kit (Cat. No. AT341-01, TransGen Biotech, China), according to the manufacturer’s instructions. Both the cDNA and the transcription mix were later used for real-time PCR with SYBR Green Master Mix. GAPDH was referred to as the internal control for mRNA. Amplification conditions were 94 °C for 30 s, followed by 45 cycles of 94 °C for 5 s, and then 72 °C for 10 s. The relative gene expression levels were determined using the 2^-ΔΔCt method, with GAPDH as the internal control. The applied primers are referred to in Supplementary Table S1. ELISA assay Blood samples were collected from the orbital venous plexus of rats in each experimental group for analysis. After incubation for 2 h at 4 °C, the blood samples were centrifuged at 3,000 rpm for 15 min to obtain the supernatant. Serum concentrations of interleukin-6 (IL-6, Cat. No. 88-50629), interleukin-10 (IL-10, Cat. No. 88-50625), and tumor necrosis factor α (TNF-α, Cat. No. 88-7340) in serum were detected according to the ELISA kits manufacturer’s instructions from Wuhan Servicebio Technology CO., Ltd. Statistical analysis All quantitative results in the present study are represented as mean ± standard deviation (SD). GraphPad Prism 9 was used to compare data across groups, and analysis was performed using one-way ANOVA, followed by post hoc Tukey’s test. Values of p < 0.05 were considered statistically significant. Results Identification of hucMSCs and hucMSCs-exosomes Primary hucMSCs exhibited a typical spindle-shaped fibroblast-like morphology (Fig. 1A), and expressed the characteristic MSC surface markers CD44, CD105, CD90, and CD73 (> 95%); while showing minimal expression of CD34, HLA-DR, CD19, CD11b, and CD45 (< 2%) by flow cytometry (Fig. 1C). The extraction of hucMSC-exos is illustrated in Fig. 1B. Successful isolation of hucMSCs-Exos was confirmed by: (1) cup-shaped morphology by TEM (Fig. 1D); (2) a mean particle size of 71.38 nm and concentration of 6.3 × 10⁹ particles/mL by NTA (Fig. 1E); and (3) expression of CD9 and CD81 on the nanoparticle surface (Fig. 1F). Fig. 1Characterization of hucMSC. A After hucMSCs were cultured for 48 h, they were visualized using a 40× light microscope and showed a long spindle shape, comparable size, and a parallel or whirlpool-like pattern. B Schematic of hucMSCs and exosome isolation. C Flow cytometry was used to evaluate the expression of surface protein markers in hucMSC. The results demonstrated negative expression for CD34 (0.36%), HLA-DR (0.43%), CD19 (0.32%), CD11b (0.32%), and CD45 (1.65%), while positive expression was noted for CD44 (99.7%), CD105 (99.9%), CD90 (99.8%), and CD73 (99.8%). D Exosomal morphology was assessed by transmission electron microscopy (TEM) with a 100 nm scale bar. E Exosome fraction particle size analysis revealed a peak at 71.38 nm. F Exosomes were detected using nanoflow fluorescence analysis for their surface-specific protein markers Characterization of hucMSC. A After hucMSCs were cultured for 48 h, they were visualized using a 40× light microscope and showed a long spindle shape, comparable size, and a parallel or whirlpool-like pattern. B Schematic of hucMSCs and exosome isolation. C Flow cytometry was used to evaluate the expression of surface protein markers in hucMSC. The results demonstrated negative expression for CD34 (0.36%), HLA-DR (0.43%), CD19 (0.32%), CD11b (0.32%), and CD45 (1.65%), while positive expression was noted for CD44 (99.7%), CD105 (99.9%), CD90 (99.8%), and CD73 (99.8%). D Exosomal morphology was assessed by transmission electron microscopy (TEM) with a 100 nm scale bar. E Exosome fraction particle size analysis revealed a peak at 71.38 nm. F Exosomes were detected using nanoflow fluorescence analysis for their surface-specific protein markers hucMSC-exos protect and repair cells in vitro Hypoxia in infarcted hearts causes widespread cardiac cell death. To determine the potential of hucMSCs-Exos to enhance cell viability, we established a CoCl2-induced hypoxia model (Fig. S1). The results showed that under CoCl₂-induced hypoxia, hucMSC-Exos significantly enhanced the viability of H9C2 cardiomyoblasts and EA.hy926 endothelial cells (Fig. S2). Vascularization helps heart tissue recover from MI. We next examined the effects of hucMSCs-Exos on cell migration and endothelial tube formation under hypoxic conditions. We hypothesized that hucMSC-Exos would promote repair-related cellular responses under hypoxic conditions. H9C2 and EA·hy926 cells were further used for tube formation assays to determine whether hucMSC-Exos could enhance repair-related responses under hypoxic conditions. H9C2 and EA·hy926 cells migration increased significantly with hucMSC-Exo groups, as given in Fig. S3. Notably, the migration ratio of EA·hy926 cells in the Hypoxia + Exo group was roughly twofold higher than in the Hypoxia + PBS group after 24 h of culture. Similarly, hucMSC-Exo treatment increased endothelial lumen development. This rise was 3.7-fold greater than in the hypoxia-only control group (Fig. S4). In addition, confocal microscopy also demonstrated the effective uptake of PKH67-labeled exosomes into H9C2 and EA·hy926 cells (Fig. S5). Preparation and characterization of the SFMA/HA-E hydrogel To construct the SFMA/HA-E hydrogel, SFMA and HA-E were synthesized separately and then mixed, followed by UV irradiation at 366 nm for approximately 1 min to achieve rapid gelation (Figs. 2A and 3A). Successful methacrylation of SF was confirmed by 1H NMR, as evidenced by the appearance of characteristic peaks at 6.11 and 5.68 ppm assigned to the methacrylamide double bonds (Fig. 2B). Likewise, 1H NMR of HA-E showed new peaks at 7.00, 6.65, and 6.15 ppm, corresponding to the aromatic protons of EGCG, confirming successful conjugation of EGCG to HA (Fig. 2C). FTIR analysis further supported these structural modifications: SFMA exhibited a distinct absorption peak at 1733 cm-1, attributable to the ester carbonyl groups introduced by methacrylation, while HA-E displayed a characteristic band at 1293 cm-1, consistent with EGCG grafting (Fig. 2D). Together, these results verified the successful synthesis of both SFMA and HA-E. Fig. 2Synthesis of SFMA/HA-E Hydrogels. A Diagram illustrating the chemical synthesis structure of SFMA and HA-E. B Nuclear magnetic resonance (NMR) spectra of SF and SFMA. C NMR spectra of HA and HA-E. D FTIR spectra of SF, SFMA, HA, and HA-E Synthesis of SFMA/HA-E Hydrogels. A Diagram illustrating the chemical synthesis structure of SFMA and HA-E. B Nuclear magnetic resonance (NMR) spectra of SF and SFMA. C NMR spectra of HA and HA-E. D FTIR spectra of SF, SFMA, HA, and HA-E Fig. 3Characterization of SFMA/HA-E. A The SFMA/HA-E hydrogel composites are gelled by UV light at a wavelength of 366 nm. Three group images of SFMA/HA-E hydrogels with different concentrations were displayed. B SEM photographs and pore size distribution of 10% SFMA/HA-E, 15% SFMA/HA-E, and 20% SFMA/HA-E scaffolds. C-D Time- and frequency-dependent behavior of G′ (elastic modulus) and G″ (viscous modulus) of the hydrogel. E-F Plots of compressive modulus and strain differences. G Swelling test of the hydrogel. H-I In vitro degradation profile of the hydrogel under various conditions Characterization of SFMA/HA-E. A The SFMA/HA-E hydrogel composites are gelled by UV light at a wavelength of 366 nm. Three group images of SFMA/HA-E hydrogels with different concentrations were displayed. B SEM photographs and pore size distribution of 10% SFMA/HA-E, 15% SFMA/HA-E, and 20% SFMA/HA-E scaffolds. C-D Time- and frequency-dependent behavior of G′ (elastic modulus) and G″ (viscous modulus) of the hydrogel. E-F Plots of compressive modulus and strain differences. G Swelling test of the hydrogel. H-I In vitro degradation profile of the hydrogel under various conditions To determine the optimal formulation for subsequent studies, 10%, 15%, and 20% SFMA/HA-E hydrogels were systematically evaluated for microstructure, rheological behavior, mechanical performance, swelling, and degradation. SEM images showed that all hydrogels possessed a porous and interconnected internal architecture. At the same time, the pore size gradually decreased with increasing SFMA concentration, indicating the formation of a denser network structure at higher polymer contents (Fig. 3B). Rheological analysis demonstrated that the storage modulus (G′) was consistently higher than the loss modulus (G″) in all groups, confirming the predominantly elastic nature of the hydrogels (Fig. 3C, D). Moreover, both G′ and G″ increased with increasing SFMA concentration, suggesting an increase in crosslink density within the hydrogel network. Consistent with these findings, compressive testing revealed that the 20% SFMA/HA-E hydrogel exhibited the greatest mechanical strength, withstanding a compressive stress of over 80 kPa, compared with approximately 64 kPa and 56 kPa for the 15% and 10% groups, respectively. The compressive modulus of the 20% SFMA/HA-E hydrogel was also significantly higher than that of the lower-concentration groups (Fig. 3E-F). The swelling behavior of the hydrogels approached equilibrium within 3 h (Fig. 3G). The 10% SFMA/HA-E hydrogel exhibited the highest swelling ratio, reaching approximately 182%, whereas the swelling ratios decreased to 88% and 74% for the 15% and 20% SFMA/HA-E hydrogels, respectively. This reduction in swelling ratio with increasing SFMA concentration was likely attributable to a higher crosslinking degree, which led to a more compact network structure and consequently reduced water uptake. The in vitro degradation profiles further demonstrated that the hydrogels degraded more rapidly in the presence of lysozyme than under enzyme-free conditions. In addition, increasing the SFMA concentration prolonged the degradation period, which may be explained by a denser, more highly crosslinked network formed at higher SFMA contents. Among the tested formulations, the 20% SFMA/HA-E hydrogel exhibited the slowest degradation profile. It underwent complete degradation on day 23 under lysozyme-containing conditions and on day 25 under enzyme-free conditions (Fig. 3H, I). Collectively, these results indicate that increasing SFMA concentration led to a denser and mechanically stronger hydrogel network with reduced swelling and prolonged degradation. Based on its overall superior physicochemical and mechanical performance, the 20% SFMA/HA-E hydrogel was selected for subsequent experiments. In vivo encapsulation with hydrogel effectively prolonged the sustained release of exosomes After MI was induced by ligation of the left anterior descending coronary artery, as confirmed by ECG changes and visible blanching (Fig. S6), therapies were then administered. Analysis of PKH26 fluorescence in heart sections showed that rats treated with Gel + Exos retained significantly more exosomes in the infarct region than rats treated with exosomes alone (Fig. S7). At day 1 after injection, the relative fluorescence intensity (RFI) in the Gel + Exos group was approximately twofold higher than that in the Exos-only group (P < 0.0001), and at day 7, exosomes were still detectable in the Gel+Exos group. In contrast, the RFI of the Exos-only group had returned to background levels. This demonstrates that the SFMA/HA-E hydrogel effectively inhibits the rapid systemic clearance that limits the efficacy of free exosome therapy. Cardiac function and remodeling are markedly ameliorated by combined therapy The therapeutic efficacy of the exosome-loaded hydrogel was tested in a rat MI model, and the experimental design and timeline of the study are shown in Fig. 4A. To assess cardiac function non-invasively, we performed 2D echocardiography, which provides comprehensive cardiac function parameters. After MI induction and hydrogel implantation, echocardiography was performed at the 28-day time point, and all parameters were compared with the negative control (MI). M-mode echocardiographic images revealed that exosomes treated with Exo and Gel + Exo groups showed overall improvement in cardiac function (Fig. 4B). However, significant differences were observed in the Gel + Exo group with respect to ejection fraction, fractional shortening, and stroke volume (Figs. 4C-E). Fig. 4Effects of SFMA/HA-E hydrogel with hucMSC-Exo on heart function after MI. A A scheme for in vivo experiments. B Echo post-surgery of hearts from the sham, PBS, Gel, Exo, and Exo + Gel groups. C-E The sum of left ventricle SV, the percentages of LVEF, and LVFS at day 28 post-MI (n = 5 per group). Dates are presented as mean ± s.d., and statistical significance was evaluated via one-way ANOVA followed by Tukey’s post hoc test. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 Effects of SFMA/HA-E hydrogel with hucMSC-Exo on heart function after MI. A A scheme for in vivo experiments. B Echo post-surgery of hearts from the sham, PBS, Gel, Exo, and Exo + Gel groups. C-E The sum of left ventricle SV, the percentages of LVEF, and LVFS at day 28 post-MI (n = 5 per group). Dates are presented as mean ± s.d., and statistical significance was evaluated via one-way ANOVA followed by Tukey’s post hoc test. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 Histological analysis confirmed the functional data. Masson’s trichrome staining showed that the fibrotic scar area significantly decreased in the Gel+Exos group (20.99 ± 4.04%) compared with that in the MI group (67.40 ± 3.68%) (Fig. 5A, D). H&E staining disclosed that tissue architecture was maintained, cardiomyocyte edema was decreased, and inflammatory infiltration was ameliorated in the combination treatment group (Fig. 5B). Cardiac tissue comprises a maximum amount of collagen I (80%) with trace amounts of collagen III (20%), where the high amount of type I renders the contractile property [29]. Immunofluorescence analysis showed enhanced scar maturation, as evidenced by an increased ratio of stiff collagen I to the more permissive type III collagen, indicating a favored shift in scar composition with Gel+Exos treatment. The Col I/Col III ratio: 6.79 ± 0.47 vs. MI 0.31 ± 0.02, P < 0.0001 by MFI quantification (Fig. 5C, E). Fig. 5Effects of SFMA/HA-E hydrogel-hucMSC-Exo on myocardial remodeling following MI. A and D Masson’s staining and statistical analysis of the collagen volume fraction in each group at 28 days post-MI (n = 5 per group). (B) H&E staining of each group at 28 days post-MI. C Immunofluorescence images of Col I and Col III in each group 7 days after MI (scale bars = 20 μm). E The proportions of collagen I and type III were determined from immunofluorescence images (n = 5 per group). Dates are presented as mean ± s.d., and statistical significance was evaluated via one-way ANOVA followed by Tukey’s post hoc test. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 Effects of SFMA/HA-E hydrogel-hucMSC-Exo on myocardial remodeling following MI. A and D Masson’s staining and statistical analysis of the collagen volume fraction in each group at 28 days post-MI (n = 5 per group). (B) H&E staining of each group at 28 days post-MI. C Immunofluorescence images of Col I and Col III in each group 7 days after MI (scale bars = 20 μm). E The proportions of collagen I and type III were determined from immunofluorescence images (n = 5 per group). Dates are presented as mean ± s.d., and statistical significance was evaluated via one-way ANOVA followed by Tukey’s post hoc test. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 Gel+exos therapy suppressed cardiomyocyte apoptosis and induced proliferation To investigate the anti-apoptotic efficacy of Gel + Exo treatment, TUNEL staining was performed at 3 days after MI. The results showed that the number of TUNEL-positive cells in the border region of MI rats treated with Gel+Exos was markedly lower than in the other groups (Fig. 6A, C). Western blot analysis was consistent with this result. Bax and cleaved caspase-3/9 were down-regulated at the protein level, and anti-apoptotic Bcl-2 was up-regulated in the infarct border zone tissues (Fig. 6E-I). Cardiomyocyte renewal by cell cycle re-entry has been confirmed as a key strategy for cardiac regeneration. The expression of the proliferation marker Ki-67 was strongly enhanced in the infarct border zone in the Gel + Exo group, reflecting active tissue repair (Fig. 6B, D). These results indicated that Gel + Exo treatment attenuated apoptosis and was associated with increased proliferation-related responses in the infarct border zone. Fig. 6SFMA/HA-E hydrogel mixed with hucMSC-Exo increases rat cardiomyocyte survival by inhibiting apoptosis after MI. A TUNEL-stained sections of rats 3 days after MI, with nuclei counterstained with DAPI (scale bars = 20 μm). The arrows denote apoptotic cells. B Rats 7 days after MI, labeled with the cell proliferation marker Ki67, and nuclei were counterstained with DAPI (scale bars = 20 μm). C Apoptosis was measured as the proportion of cells positive for TUNEL staining (n = 3 per group). D Cell proliferation was measured by counting Ki67-positive cells per high-power field (n = 3 per group). E Protein expression of cleaved caspase-9/3, Bcl-2, Bax, and β-tubulin was determined via Western blotting. F-I Percentage of cleaved caspase-9/3, Bcl-2, and Bax from the western blot images (n = 3 per group). Dates are presented as mean ± s.d., and statistical significance was evaluated via one-way ANOVA followed by Tukey’s post hoc test. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 SFMA/HA-E hydrogel mixed with hucMSC-Exo increases rat cardiomyocyte survival by inhibiting apoptosis after MI. A TUNEL-stained sections of rats 3 days after MI, with nuclei counterstained with DAPI (scale bars = 20 μm). The arrows denote apoptotic cells. B Rats 7 days after MI, labeled with the cell proliferation marker Ki67, and nuclei were counterstained with DAPI (scale bars = 20 μm). C Apoptosis was measured as the proportion of cells positive for TUNEL staining (n = 3 per group). D Cell proliferation was measured by counting Ki67-positive cells per high-power field (n = 3 per group). E Protein expression of cleaved caspase-9/3, Bcl-2, Bax, and β-tubulin was determined via Western blotting. F-I Percentage of cleaved caspase-9/3, Bcl-2, and Bax from the western blot images (n = 3 per group). Dates are presented as mean ± s.d., and statistical significance was evaluated via one-way ANOVA followed by Tukey’s post hoc test. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 Gel+Exos therapy improved angiogenesis and immunomodulating Vessel density in the border area was examined on day 28 post-MI (Fig. 7A, B). Vessels were identified via immunostaining of CD31 and α-SMA. The result showed that the Gel+Exos group had a significantly higher density of mature microvessels than all control groups (Fig. 7D, E), indicating robust angiogenesis. Analysis of the immune microenvironment revealed that Gel+Exos treatment induced macrophage recruitment and polarisation. The ratio of M2-like (CD206⁺) macrophages to total (CD68⁺) macrophages in the infarcted area was significantly increased compared with the MI group 7 days after treatment (Fig. 7C, F, G). This pro-reparative switch was accompanied by a beneficial systemic cytokine profile, with significantly reduced serum TNF-α and IL-6 levels and increased anti-inflammatory IL-10 levels (Fig. S8). Together, Gel+Exos treatment enhances revascularization and modulates macrophage differentiation. Fig. 7Immunostaining analysis of vessel density and macrophage differentiation. A-B Rat 28 days after MI, labeled with fluorochrome-conjugated antibodies against CD31 and α-SMA, with DAPI as a nuclear counterstain (scale bars = 50 μm). C Typical confocal images of CD68+ and CD206+ macrophages in the infarcted area on day 7 post MI (scale bars = 40 μm). D-E Quantification analysis of cells submitted to immunofluorescence as α-SMA+ and CD31+ structure (n = 5 per group). F-G Percentage of CD68+ or CD206+ macrophages (n = 5 per group). Dates are presented as mean ± s.d., and statistical significance was evaluated via one-way ANOVA followed by Tukey’s post hoc test. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 Immunostaining analysis of vessel density and macrophage differentiation. A-B Rat 28 days after MI, labeled with fluorochrome-conjugated antibodies against CD31 and α-SMA, with DAPI as a nuclear counterstain (scale bars = 50 μm). C Typical confocal images of CD68+ and CD206+ macrophages in the infarcted area on day 7 post MI (scale bars = 40 μm). D-E Quantification analysis of cells submitted to immunofluorescence as α-SMA+ and CD31+ structure (n = 5 per group). F-G Percentage of CD68+ or CD206+ macrophages (n = 5 per group). Dates are presented as mean ± s.d., and statistical significance was evaluated via one-way ANOVA followed by Tukey’s post hoc test. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 Gel+Exos preserve the gap junction Collateral circulation helps restore blood supply to injured heart tissue after MI [30]. Cx43 is a key gap junction in ventricular myocardium, and was evaluated to assess electrical coupling-related remodeling after MI [31]. Gel+Exos treatment restored Cx43 expression after MI (Fig. 8A). In particular, Cx43 expression was substantially higher in the Gel+Exos group compared to Exos alone (Figs. 8B-D). This implies the treatment would also be beneficial to prevent arrhythmogenic sequelae in post-MI. Fig. 8SFMA/HA-E hydrogel, in conjunction with hucMSC-Exo, enhances gap junction expression. A Immunofluorescent staining for cTnT and Connexin 43 (Cx43) expression, with nuclei counterstained using DAPI (scale bars = 20 μm). B Western blot analysis of protein expression. C Densitometry analysis of western blot results normalized to tubulin protein levels (n = 3 per group). D Percentage of Connexin 43-positive cells calculated from the immunofluorescence images (n = 5 per group). Dates are presented as mean ± s.d., and statistical significance was evaluated via one-way ANOVA followed by Tukey’s post hoc test. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 SFMA/HA-E hydrogel, in conjunction with hucMSC-Exo, enhances gap junction expression. A Immunofluorescent staining for cTnT and Connexin 43 (Cx43) expression, with nuclei counterstained using DAPI (scale bars = 20 μm). B Western blot analysis of protein expression. C Densitometry analysis of western blot results normalized to tubulin protein levels (n = 3 per group). D Percentage of Connexin 43-positive cells calculated from the immunofluorescence images (n = 5 per group). Dates are presented as mean ± s.d., and statistical significance was evaluated via one-way ANOVA followed by Tukey’s post hoc test. ns: non-significant, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 Discussion The pathogenesis of post-MI heart failure is complex, resulting from the combined actions of cardiomyocyte fatalities, excessive fibrotic network maturation, insufficient vascular supply, and distorted immune response [32, 33]. Although traditional anti-inflammatory approaches, such as colchicine, have shown promise [34, 35], they do not possess the requisite pro-regenerative properties for effective myocardial repair. Here, we present a hydrogel-assisted exosome delivery system to address these pathologies. The main novelty is the preparation of the SFMA/HA-E hydrogel. As postulated, this system fused three components. Our work demonstrated that methacrylate groups in SFMA form stable three-dimensional networks via covalent crosslinks in the presence of the LAP photoinitiator during photopolymerization under ultraviolet (UV) light. In contrast, HA-E contains no methacrylate groups, but it will become physically trapped within the three-dimensional network. By being part of the network, HA-E will alter hydrogel characteristics: hydrogen bonds formed between HA-E and SFMA chains will increase hydrophilicity, while steric hindrance from EGCG groups will limit pore size and alter the materials’ degradation rates. Likewise, photopolymerized SFMA’s mechanical stability provided mechanical support to the infarcted region of the heart, which caused less dilation to occur in the ventricles while maintaining healthy levels of wall thickness [11, 12], and HA improved the compatibility of the gel with the body between uses (biocompatibility) and supported cells to attach and begin angiogenesis [17], thus allowing the cells to migrate into the area of treatment and become involved in tissue repair. Lastly, conjugating EGCG with the gel addressed known issues of free EGCG such as its (1) instability, (2) rapid oxidation, and (3) very short half-life through chemical crosslinking (i.e., “stabilizing” the compound) and providing a means for continuous release over a protracted period, which explains the enhanced anti-inflammatory effects and macrophage polarization towards the M2 phenotype observed in the Gel+Exos condition (Fig. 7). The SFMA/HA-E scaffold contributes to cardiac repair through both biophysical and biogenetic mechanisms. It may mechanically stabilize the infarct wall, provide a porous microstructure that permits host cell infiltration, and offer matrix-associated cues that support tissue integration [17]. In addition, the hydrogel provides a sustained-release depot for exosomes. The hydrogel has been shown to maintain therapeutic levels of exosomes for at least 7 days (Fig. S7) and to overcome the rapid clearance associated with exosome administration in their free form [26, 27]. The therapeutic benefit observed in this study was likely attributable to coordinated effects on several major aspects of post-infarction remodeling. These included attenuation of cardiomyocyte apoptosis, likely through modulation of the Bcl-2/Bax axis; enhanced angiogenesis and microvascular formation [10]; a shift in the immune microenvironment toward an M2-dominant reparative phenotype; and reduced fibrotic remodeling with changes in scar composition [9]. The approximately 40% absolute increase in LVEF with combination therapy supports the advantage of integrating biomaterial support with exosome-based treatment. The combination of early cellular changes following treatment and late functional recovery will strengthen the rationale for this approach. There was an apparent reduction in cardiomyocyte apoptosis, M2 macrophage polarization, and decreased levels of inflammatory cytokines, all of which occurred within 7 days post-treatment. In addition to these early cellular changes, there was an increase in proliferative activity at this time point. Thus, the beneficial effects of these four cellular changes provide a foundation for the substantial functional recovery observed at day 28 (LVEF = 70.71% ± 3.04%) [10]. The additional benefit of our strategy is the preservation of a gap junction protein that is incredibly important in ensuring electrical coupling (Cx43), which is often downregulated during arrhythmic remodeling. The preservation of Cx43 suggests that this approach may reduce the risk of post-infarction arrhythmias and thereby enhance its therapeutic value beyond contractile recovery alone. Taken together, our results align with the emerging paradigm that the therapeutic effects of MSC-derived exosomes are pleiotropic, targeting multiple cell types within the cardiac niche [26, 36]. Importantly, the hydrogel functions not merely as a passive delivery vehicle but actively contributes to therapeutic efficacy through mechanical reinforcement and EGCG-mediated antioxidant effects [37, 38]. This work extends the growing field of hydrogel-based exosome delivery for MI [39] by introducing a rationally designed platform that combines structural support with bioactive signaling. Study limitations While this investigation shows promise for the effectiveness of this hydrogel, more research is required. The absence of direct co-culture studies with hydrogel formulations and viable cells limits the conclusions we can draw from our in vivo efficacy and biocompatibility data, because each component of the hydrogel (SFMA, HA, and EGCG) has demonstrated acceptable biocompatibility in the literature [14, 40]. Thus, to understand the interactions between cells and the hydrogel, evaluate the potential cytotoxicity of the hydrogel formulation, and assess the influence of the hydrogel on cell behavior in a controlled hypoxic environment, future studies should use co-culture systems. The long-term efficacy and durability of the positive changes seen with 28 days should also be assessed in future studies. Furthermore, the hydrogel’s degradation profiles were evaluated only under in vitro conditions; direct in vivo degradation kinetics remain to be determined in future studies. In addition, identifying the exosomal cargo responsible for these effects (e.g., miRNAs/proteins, etc.) is important. Ultimately, to enable translation into the clinic, future studies must be conducted using hydrogel products manufactured in large quantities and demonstrating sterility and injectability. Conclusion We designed a new injectable hydrogel for the long-term release of hucMSC exosomes. The SFMA/HA-E+Exos construct effectively reverses the critical pathological processes in post-infarction remodeling apoptosis, fibrosis, suppressed angiogenesis, and maladaptive inflammation that translates into substantive functional recovery. This strategy of combinatory therapy is a collaborative and translational therapeutic modality for patients with myocardial infarction. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1. Supplementary Material 1.