Sprayable bioadhesive microcarriers loaded with Tβ4-Engineered ADSC exosomes for diabetic wound healing

Abstract

Stem cell-derived therapeutics show strong potential to recalibrate diabetic wound immunity, yet their stability, retention, and practical usability remain major barriers to effective application. Here, we report a novel microcarrier platform loaded with thymosin β4 (Tβ4)-overexpressing stem cell-derived exosomes for a sprayable diabetic wound dressing. Adipose-derived stem cells (ADSCs) were genetically engineered to overexpress Tβ4, generating potent immunoregulatory exosomes that were efficiently encapsulated into uniform, micron-scale hydrogel microcarriers via microfluidic fabrication and further functionalized with a mesoporous polydopamine (mPDA) coating to enhance wet adhesion and tissue retention. The resulting EXOs^Tβ4/mPDA@MS system stabilizes the exosome payload and enables convenient spray-based wound administration. These microcarriers provide sustained, localized exosome release, significantly enhance macrophage efferocytosis, suppress inflammatory signaling, and accelerate wound repair in diabetic models. Thus, our engineered, sprayable, and adhesive microcarrier platform offers a stable, minimally invasive, and clinically adaptable strategy for advancing stem cell-derived exosome therapies in chronic diabetic wound repair.

Graphical abstract

Highlights

• •Sprayable mPDA-functionalized microcarriers deliver engineered ADSC exosomes.
• •Tβ4-enriched exosomes enhance macrophage efferocytosis and immune remodeling.
• •MerTK/Pi3k/Akt signaling mediates inflammation resolution and tissue repair.
• •A translational exosome-biomaterial strategy for chronic wound therapy.

1. Introduction

Diabetic wounds remain a formidable clinical challenge, largely caused by the persistent and dysregulated inflammatory microenvironment [1,2]. A key pathological feature is impaired macrophage efferocytosis [3,4]. This defect leads to insufficient clearance of apoptotic cells and sustains inflammatory signaling [[5], [6], [7]]. Consequently, the wound becomes trapped in a non-resolving inflammatory state that ultimately obstructs the progression of tissue repair. Nowadays, stem cell therapy has attracted considerable attention in wound management due to its potent immunomodulatory and pro-healing secretome [8,9]. Notably, adipose-derived stem cells (ADSCs) offer advantages in accessibility, scalability, and paracrine potency [[10], [11], [12]]. However, the local application of stem cells typically requires coverage with gauze or conventional dressings to prevent cell loss from the wound surface, which inevitably causes pain during dressing changes. In recent years, various bioactive materials have been developed for targeted stem cell delivery and wound coverage [9,13,14]. Yet, under the hostile pathological conditions of diabetic wounds, these systems still struggle to maintain stem cell viability and preserve their biological functions [15,16]. These challenges underscore the need for a more flexible, convenient, and clinically compatible platform for stem cell-based therapy in chronic wound care.

Here, we propose a sprayable adhesive microcarrier system loaded with engineered ADSC-derived products for diabetic wound treatment, as illustrated in Fig. 1. Stem cell engineering has emerged as a powerful strategy to enhance specific cellular functions, enabling the purposeful augmentation of immunomodulation, angiogenesis, or cytoprotection to meet the therapeutic needs of different wounds [17]. Especially, to avoid the rapid inactivation of living cells within complex physiological environments, increasing attention has shifted toward engineered stem cell-derived exosomes [18,19]. As nanoscale lipid vesicles enriched with stem cell specific bioactive molecules, engineered exosomes offer superior stability and resilience, making them particularly advantageous for treating refractory diabetic wounds [6,20]. In contrast, microcarriers serve as a promising delivery system to safeguard sensitive bioactive payloads while enabling controlled release [[20], [21], [22], [23], [24]]. Notably, hydrogel microcarriers feature a hydrated macromolecular network that can encapsulate exosomes efficiently while preserving their integrity and bioactivity [[25], [26], [27]]. Moreover, these hydrogel networks can be easily functionalized with additional chemical groups to introduce new functionalities, such as wet adhesion properties suitable for complex wound environments [[28], [29], [30]]. Therefore, we integrated engineered exosomes with an adhesive microcarrier platform to create a more practical and user-friendly stem cell-derived therapeutic strategy.

Fig. 1.

Schematic overview of the preparation and therapeutic application of EXOs^Tβ4/mPDA@MS. (a) Preparation of Tβ4-engineered exosomes (EXOs^Tβ4). (b) Microfluidic fabrication of EXOs^Tβ4/mPDA@MS microspheres. (c) Overview of the application and mechanism by which EXOs^Tβ4/mPDA@MS facilitates diabetic wound healing through enhanced macrophage efferocytosis.

In this study, we encapsulated thymosin β4 (Tβ4)-overexpressing ADSC-derived exosomes within functionalized hydrogel microcarriers via microfluidic technology for sprayable wound dressing. To address the immunological dysregulation characteristic of diabetic wounds, we genetically engineered ADSCs via lentiviral transduction to overexpress Tβ4. Tβ4 is a multifunctional bioactive peptide involved in anti-inflammatory regulation and tissue repair [31]. However, free Tβ4 is prone to rapid degradation and inactivation in the harsh diabetic microenvironment [32]. Our previous findings demonstrated that Tβ4 is downregulated in diabetic wounds and is associated with impaired cellular function [33]. Therefore, Tβ4-enriched exosomes represent a rational strategy to restore deficient pro-reparative signaling in the diabetic microenvironment. To achieve stable exosome delivery in a sprayable format, we used microfluidic technology to fabricate micron-scale hydrogel microcarriers functionalized with mesoporous polydopamine (mPDA), forming a bioadhesive EXOs^Tβ4/mPDA@MS platform. These microcarriers exhibit uniform, monodisperse micron-sized architectures and can efficiently encapsulate engineered exosomes with high structural stability, thereby providing a robust platform for localized, on-demand therapeutic release. When deployed as a sprayable formulation, these microcarriers markedly enhance macrophage efferocytosis, attenuate inflammatory activation, and accelerate wound closure in diabetic mouse models. Mechanistic analyses revealed activation of the MerTK/Pi3k/Akt axis, which facilitated efferocytosis-driven inflammation resolution and subsequent tissue regeneration. Collectively, this engineered, sprayable, and adhesive microcarrier system provides a clinically adaptable platform that integrates targeted exosome delivery with immune microenvironment reprogramming for chronic diabetic wound therapy.

2. Results and discussion

2.1. Engineering and characterization of Tβ4-overexpressing ADSCs and derived exosomes

In this study, exosomes enriched with Tβ4 were produced by lentiviral transduction of ADSCs (Fig. 2a). The isolated ADSCs were validated by flow cytometry, showing a typical mesenchymal stem cell phenotype (CD90⁺, CD73⁺, CD44⁺, and CD105⁺) and the absence of hematopoietic markers CD45 and CD34 (Fig. S1). Efficient lentiviral transduction was verified by fluorescence microscopy, as both ADSCs^vector and ADSCs^Tβ4 exhibited strong green fluorescent protein (GFP) signals, with transduction efficiencies exceeding 90% according to flow cytometric analysis (Fig. 2b, c, h). RT-qPCR analysis showed a marked increase in Tβ4 transcript levels in ADSCs^Tβ4 relative to ADSCs^vector. (Fig. S2). Immunofluorescence staining further confirmed elevated Tβ4 expression in ADSCs^Tβ4 at the protein level, as evidenced by stronger fluorescence signals and quantitative intensity analysis (Fig. 2d–i). Consistently, Western blot analysis corroborated these results, showing markedly increased Tβ4 protein levels in ADSCs^Tβ4 (Fig. 2f and g).

Fig. 2.

Characterization of Tβ4-enriched ADSCs and EXOs^Tβ4. (a) Schematic of Tβ4 lentiviral transduction in ADSCs and exosomes isolation. (b, c) Fluorescence and bright-field images of ADSCs^vector and ADSCs^Tβ4, scale bar 100 μm. (d, i) Immunofluorescence images(d) and quantitative analysis(i) of Tβ4 in ADSCs^vector and ADSCs^Tβ4, scale bar 50 μm, n = 3. (e) TEM images of EXOs^vector and EXOs^Tβ4, scale bar 200 nm. (f, g) Western blot(f) and quantitative analysis(g) of Tβ4 in ADSCs^vector and ADSCs^Tβ4, n = 3. (h) Flow cytometry of GFP expression in ADSCs after transduction. (j, k) Western blot (j) and quantitative analysis (k) of Tβ4 in EXOs^vector and EXOs^Tβ4, n = 3. (l) NTA of the size distribution of EXOs^vector and EXOs^Tβ4. (m) Western blot of TSG101, CD9, CD63 and Calnexin.

Transmission electron microscopy (TEM) verified the characteristic cup-shaped morphology of both EXOs^vector and EXOs^Tβ4 (Fig. 2e). Western blot analysis further confirmed a pronounced enrichment of Tβ4 protein in EXOs^Tβ4 compared with EXOs^vector (Fig. 2j and k). Nanoparticle tracking analysis (NTA) demonstrated that both exosome populations exhibited comparable size distributions (∼150 nm) and similar zeta potentials, indicating uniform colloidal stability (Fig. 2l–S3). In addition, no significant difference in particle concentration was observed between EXOs^vector and EXOs^Tβ4, suggesting that Tβ4 overexpression does not affect exosome secretion yield (Fig. S4). To confirm exosome identity, the characteristic markers CD63, CD9, and TSG101 were detected in both EXOs^vector and EXOs^Tβ4, whereas the endoplasmic reticulum protein Calnexin was absent from exosome preparations but present in parental ADSCs, verifying the high purity of the isolated vesicles (Fig. 2m). Collectively, these results confirm the successful generation and comprehensive characterization of Tβ4-enriched ADSC-derived exosomes, providing a solid foundation for subsequent therapeutic application. Compared with native exosomes, engineered EXOs^Tβ4 exhibit enhanced and targeted immunomodulatory activity [34]. In addition, exosomes possess intrinsic biocompatibility and efficient cellular uptake, offering advantages over synthetic nanoparticle systems [35]. Together, these features support the rationale of using engineered exosomes as a biologically optimized delivery platform.

2.2. Fabrication and bioadhesive characterization of EXOs^Tβ4/mPDA@MS

To achieve controlled delivery of EXOs^Tβ4 in a bioadhesive and sprayable formulation, a microcarrier delivery system was fabricated using a microfluidic technique (Fig. 3a). TEM imaging confirmed the successful synthesis of mPDA nanoparticles, and dynamic light scattering analysis revealed a uniform particle diameter of 165 ± 6.9 nm (Fig. S5). Concurrently, the successful synthesis of GelMA was confirmed by ¹H NMR spectroscopy, showing characteristic peaks of methacryloyl groups (Fig. S6). In this study, a GelMA pregel solution containing suspended EXOs^Tβ4 was used as inner phase, with oil serving as outer phase, to fabricate EXOs^Tβ4@MS via a single-emulsion microfluidic method. Subsequently, mPDA deposition was performed on the surface of the prepared microcarriers to obtain EXOs^Tβ4/mPDA@MS. Optical microscopy and electron microscopy further confirmed the successful mPDA functionalization of the microcarriers, which was accompanied by an evident alteration in their optical characteristics. In contrast to pristine EXOs^Tβ4@MS, mPDA-modified EXOs^Tβ4/mPDA@MS exhibited a darkened and opaque appearance (Fig. 3b–d). Quantitative size analysis showed narrow and monodisperse distributions for EXOs^Tβ4@MS and EXOs^Tβ4/mPDA@MS, with mean diameters of 99.8 ± 5.8 μm and 100.4 ± 6.2 μm, respectively. Scanning electron microscope (SEM) imaging confirmed the successful deposition of mPDA onto the microcarrier surface and revealed the intrinsic porous internal architecture characteristic of the hydrogel microspheres (Fig. 3c–e, S7). Photographs and a spraying movie directly demonstrated the uniform sprayability and favorable handling properties of the EXOs^Tβ4/mPDA@MS formulation (Fig. S8; Movie S1), confirming that mPDA modification does not compromise the sprayability of the microsphere system.

Fig. 3.

Fabrication and characterization of EXOs^Tβ4/mPDA@MS. (a) Schematic of the synthesis of EXOs^Tβ4/mPDA@MS. (b) Bright-field image and size distribution of EXOs^Tβ4@MS, scale bar 100 μm. (c) SEM image and magnified view of EXOs^Tβ4@MS, scale bars 50 μm (left) and 10 μm (right). (d) Bright-field image and size distribution of EXOs^Tβ4/mPDA@MS, scale bar 100 μm. (e) SEM image and magnified view of EXOs^Tβ4/mPDA@MS, scale bars 50 μm (left) and 5 μm (right). (f) Representative images of EXOs^Tβ4@MS on porcine skin before and after PBS washing, scale bar 200 μm. (g) Representative images of EXOs^Tβ4/mPDA@MS on porcine skin before and after PBS washing, scale bar 200 μm. (h) FTIR spectra of GelMA and mPDA@GelMA. (i) Swelling ratio of EXOs^Tβ4@MS and EXOs^Tβ4/mPDA@MS, n = 3. (j) Cumulative release curves of EXOs^Tβ4 from EXOs^Tβ4@MS and EXOs^Tβ4/mPDA@MS, n = 3.

The bioadhesive performance of the microcarriers was first evaluated using an ex vivo porcine skin model. After PBS rinsing, EXOs^Tβ4@MS readily detached from the tissue surface, whereas EXOs^Tβ4/mPDA@MS remained firmly adhered even after repeated washing (Fig. 3f and g; Movies S2, S3). This robust adhesion is attributed to the catechol-rich chemistry of mPDA, which enables strong interfacial interactions under wet conditions. To quantitatively assess adhesion, lap shear strength tests were performed. EXOs^Tβ4@MS exhibited relatively low adhesive strength, while mPDA modification significantly enhanced adhesion, confirming the critical role of mPDA in reinforcing interfacial bonding (Fig. S9). Fourier-transform infrared (FTIR) spectroscopy further verified successful mPDA functionalization, as evidenced by the characteristic C=N imine peak (∼1670 cm⁻¹) associated with Schiff base reactions and the broadened bands at 3300-3400 cm⁻¹ corresponding to hydrogen bonding (Fig. 3h). These results collectively elucidate the chemical basis underlying the enhanced adhesion. Both microcarrier systems exhibited pronounced swelling behavior, indicating high water absorption capacity and structural integrity (Fig. 3i). Notably, EXOs^Tβ4/mPDA@MS displayed a more sustained exosome release profile compared to EXOs^Tβ4@MS, likely due to diffusion modulation by the mPDA layer (Fig. 3j). To further evaluate degradation behavior, in vitro studies were conducted under different temperature conditions (Fig. S10). At 37°C, EXOs^Tβ4/mPDA@MS exhibited rapid degradation, reaching ∼80% within 3 days, whereas degradation at 4°C was markedly slower. These results indicate efficient biodegradation under physiological conditions while maintaining stability during low-temperature storage. Zeta potential analysis revealed a shift toward more negative surface charge after mPDA modification (Fig. S11), confirming successful surface functionalization and supporting enhanced exosome retention. Furthermore, cellular uptake assays demonstrated efficient internalization of EXOs^Tβ4 released from EXOs^Tβ4/mPDA@MS by macrophages, as evidenced by strong intracellular fluorescence signals (Fig. S12), confirming effective payload delivery.

Collectively, these results demonstrate the successful development of a sprayable, bioadhesive, and sustained-release exosome delivery system with robust structural integrity and functional performance. The incorporation of mPDA endows the microcarriers with catechol-mediated wet adhesion, enabling stable retention on moist wound surfaces and effectively overcoming the rapid detachment commonly observed in conventional hydrogel systems [36]. Nevertheless, several challenges remain. In particular, long-term storage stability and the preservation of exosome bioactivity following spray delivery require further investigation. Moreover, EXOs^Tβ4@MS without mPDA was not included in subsequent in vitro and in vivo studies due to its rapid loss under wet conditions, and its independent contribution warrants further exploration.

2.3. EXOs^Tβ4/mPDA@MS enhances macrophage efferocytosis and suppresses inflammatory activation

The biosafety of the microsphere systems was evaluated by CCK-8 assay in Raw264.7 macrophages. All formulations, including PBS, mPDA@MS, EXOs^vector/mPDA@MS and EXOs^Tβ4/mPDA@MS, exhibited negligible cytotoxicity, indicating favorable biocompatibility (Fig. S13). Consistently, live/dead staining revealed minimal toxicity in wound-relevant cells, including HUVECs and HDFs (Fig. S14).

The immunomodulatory effects of EXOs^Tβ4/mPDA@MS were evaluated by assessing macrophage efferocytosis and inflammatory activation in vitro. Apoptosis of Jurkat cells was induced by UV irradiation, yielding more than 55% apoptotic cells as confirmed by Annexin V/PI staining (Fig. S15). Co-incubation assays revealed that mPDA@MS enhanced efferocytosis in Raw264.7 macrophages compared with PBS, as evidenced by increased uptake of apoptotic cells (Fig. 4a), likely due to its antioxidant-mediated regulation of macrophage redox homeostasis [[37], [38], [39]]. EXOs^vector/mPDA@MS further augmented phagocytic activity, whereas EXOs^Tβ4/mPDA@MS achieved the highest efferocytosis efficiency among all groups (Fig. 4b). This enhancement was further validated by flow cytometry (Fig. S16). Consistent trends were observed in primary BMDMs. While PBS-treated cells exhibited limited efferocytosis, mPDA@MS and EXOs^vector/mPDA@MS induced moderate uptake. Notably, EXOs^Tβ4/mPDA@MS markedly enhanced apoptotic cell clearance, confirming its robust efficacy across macrophage subtypes (Fig. 4c and d; Figs. S17 and S18). Importantly, no significant difference was observed between free EXOs^Tβ4 and EXOs^Tβ4/mPDA@MS (Fig. S19), indicating that microcarrier encapsulation preserves the intrinsic bioactivity of EXOs^Tβ4.

Fig. 4.

EXOs^Tβ4/mPDA@MS enhances macrophage efferocytosis and suppresses inflammatory activation. (a, b) Fluorescence images(a) and quantitative analysis of efferocytosis index(b) from Raw264.7 in different treatments, scale bar 50 μm, n = 3. (c) Fluorescence images from BMDMs in different treatments, scale bar 10 μm. (d) Flow cytometry of CM-Dil-labeled apoptotic cell uptake by CMFDA-labeled BMDMs in different treatments. (e, f) Immunofluorescence images(e) and quantitative analysis of iNOS expression(f) in Raw264.7 in different treatments, scale bar 100 μm, n = 3. (g) Western blot of iNOS, IL-18, and TNF-α in different treatments (i: PBS, ii: mPDA@MS, iii: EXOs^vector/mPDA@MS, iv: EXOs^Tβ4/mPDA@MS). (h, i) Quantitative analysis of iNOS and TNF-α protein expression, n = 3. (∗P < 0.05 vs PBS, ^$P < 0.05 vs mPDA@MS, ^@P < 0.05 vs EXOs^vector/mPDA@MS).

To determine whether EXOs^Tβ4/mPDA@MS-mediated efferocytosis attenuates macrophage inflammatory activation, key inflammatory markers were assessed under efferocytosis conditions. Immunofluorescence revealed a progressive reduction in iNOS expression from PBS to mPDA@MS to EXOs^vector/mPDA@MS, with the lowest levels observed in EXOs^Tβ4/mPDA@MS group (Fig. 4e and f). Consistently, Western blot analysis showed significant downregulation of iNOS, IL-18 and TNF-α, which were further supported by reduced mRNA levels of iNOS, TNF-α, IL-6, and IL-18 (Fig. 4g–i; Figs. S20 and S21).

These results indicate that EXOs^Tβ4/mPDA@MS enhances efferocytosis while suppressing pro-inflammatory activation of macrophages. In diabetic wounds, defective efferocytosis leads to apoptotic cell accumulation and sustained inflammation, disrupting immune homeostasis and impairing repair [40]. As a key initiator of inflammation resolution, efferocytosis likely serves as an upstream trigger for macrophage reprogramming and tissue regeneration restored by EXOs^Tβ4/mPDA@MS [41].

2.4. EXOs^Tβ4/mPDA@MS induces M2 macrophage polarization and reduces intracellular oxidative stress

Impaired macrophage efferocytosis in diabetic wounds leads to apoptotic cell accumulation and persistent inflammatory activation, disrupting immune homeostasis and delaying tissue repair [41,42]. This sustained pro-inflammatory milieu further hinders macrophage transition toward the reparative M2 phenotype. To determine whether EXOs^Tβ4/mPDA@MS can reverse this dysfunction, we next evaluated its ability to promote M2 polarization. Flow cytometry showed a progressive increase in CD206⁺ M2 macrophages from PBS to mPDA@MS to EXOs^vector/mPDA@MS, with the highest level in the EXOs^Tβ4/mPDA@MS group (Fig. 5a–c). Immunofluorescence analysis confirmed a marked elevation of CD206⁺ cells following EXOs^Tβ4/mPDA@MS treatment (Fig. 5b–d). Given the ROS-rich diabetic microenvironment, intracellular ROS was further assessed [22]. Both immunofluorescence and flow cytometry revealed a stepwise reduction in ROS levels across treatment groups, reaching the lowest level in EXOs^Tβ4/mPDA@MS (Fig. 5e–g).Consistently, Western blot analysis showed increased expression of M2 markers CD206 and Arg1, while RT-qPCR confirmed upregulation of IL-10 and CD206 mRNA, with the strongest effects observed in EXOs^Tβ4/mPDA@MS (Fig. 5h and i; Figs. S22 and S23).

Fig. 5.

EXOs^Tβ4/mPDA@MS promotes M2 macrophage polarization and reduces intracellular ROS levels. (a, c) Flow cytometry (a) and quantitative analysis (c) of CD206⁺ M2 macrophages in different treatments, n = 3. (b, d) Immunofluorescence images (b) and quantitative analysis (d) of CD206 in different treatments, scale bar 100 μm, n = 3. (e) Flow cytometry of intracellular ROS levels in different treatments. (f, g) Immunofluorescence images(f) and quantitative analysis(g) of intracellular ROS levels in different treatments, scale bar 100 μm, n = 3. (h) Western blot analysis of CD206 and Arg-1 in different treatments (i: PBS, ii: mPDA@MS, iii: EXOs^vector/mPDA@MS, iv: EXOs^Tβ4/mPDA@MS). (i) Quantitative analysis of Arg-1 protein expression, n = 3. (∗P < 0.05 vs PBS, ^@P < 0.05 vs EXOs^vector/mPDA@MS).

Collectively, EXOs^Tβ4/mPDA@MS promotes M2 polarization while attenuating oxidative stress. As downstream outcomes of enhanced efferocytosis, these changes reflect coordinated immune reprogramming, linking apoptotic cell clearance to macrophage phenotypic transition and microenvironmental restoration to support tissue repair [41]. Together, these findings establish a coherent mechanistic framework in which EXOs^Tβ4/mPDA@MS initiates immune resolution via efferocytosis and subsequently drives macrophage reprogramming and microenvironmental improvement to support tissue repair.

2.5. EXOs^Tβ4/mPDA@MS promotes healing and regeneration of diabetic wounds

Based on our in vitro findings that EXOs^Tβ4/mPDA@MS significantly enhances macrophage efferocytosis, promotes M2 polarization, and reduces intracellular ROS accumulation, we further evaluated its therapeutic efficacy in vivo. Full-thickness wounds were created on the dorsal skin of streptozotocin (STZ)-induced type 1 diabetic mice to assess the regenerative potential of the microcarrier system. As shown in Fig. 6a and b, wounds treated with EXOs^Tβ4/mPDA@MS exhibited markedly smaller wound areas throughout the healing process (days 3, 7, 11, and 14) compared with PBS, mPDA@MS, and EXOs^vector/mPDA@MS. Quantitative analysis further demonstrated that EXOs^Tβ4/mPDA@MS achieved the highest wound closure efficiency at all evaluated time points (Fig. 6c). Hematoxylin and eosin (H&E) staining on day 7 revealed that EXOs^Tβ4/mPDA@MS-treated wounds exhibited thicker granulation tissue, reduced inflammatory edema, and a narrower wound gap compared with control groups, indicating an accelerated transition from the inflammatory to proliferative phase (Fig. 6d–f). By day 14, EXOs^Tβ4/mPDA@MS-treated wounds displayed a structurally intact neoepidermis and well-organized dermal architecture (Fig. 6g), with the highest epithelial regeneration rate among all groups (Fig. 6h). Although EXOs^vector/mPDA@MS and mPDA@MS partially promoted epidermal reconstruction, their effects were markedly weaker than those of EXOs^Tβ4/mPDA@MS. Masson's trichrome staining further showed that EXOs^Tβ4/mPDA@MS induced more abundant and well-aligned collagen deposition, whereas collagen in other groups appeared sparse and disorganized (Fig. 6i and j).

Fig. 6.

EXOs^Tβ4/mPDA@MS enhances diabetic wound repair in type 1 diabetic wounds. (a) The photographs of wounds on different days post-treatment, scale bar 5 mm. (b) Schematic wound morphology across different groups at indicated time points. (c) Quantitative evaluation of wound closure dynamics over time, n = 5. (d) Wound tissue thickness measurements on day 7 based on H&E, n = 5. (e) H&E representative images of wound tissues on day 7, scale bars 1 mm (up) and 200 μm (down). (f) Measurement of wound gap width at day 7 from H&E sections, n = 5. (g) H&E representative images of wound tissues on day 14, scale bars 1 mm (up) and 200 μm (down). (h) Quantification of epithelial thickness on day 14, n = 5. (i) Masson's trichrome-stained representative images of wound tissues on day 14, scale bars 1 mm (up) and 200 μm (down). (j) Quantification of collagen deposition, n = 5. (∗P < 0.05 vs PBS, ^$P < 0.05 vs mPDA@MS, ^@P < 0.05 vs EXOs^vector/mPDA@MS).

Importantly, these findings collectively demonstrate the favorable therapeutic efficacy and short-term biosafety of EXOs^Tβ4/mPDA@MS in vivo. Although no obvious adverse responses were observed within 14 days, the long-term degradation behavior and biosafety, including potential material accumulation and delayed foreign body responses, require further investigation. In addition, a limitation of this study is the lack of a systematic dose-response analysis of the engineered exosomes. The selected concentration was based on our previous experience and demonstrated effective therapeutic outcomes [33]. Nevertheless, further studies are needed to optimize the dosing regimen to maximize efficacy while minimizing potential waste or adverse effects.

2.6. EXOs^Tβ4/mPDA@MS reprograms immune responses and promotes angiogenesis in diabetic wounds

To characterize the immune reprogramming induced by EXOs^Tβ4/mPDA@MS in vivo, we first evaluated macrophage efferocytosis within wound tissues. Immunofluorescence staining of CD68 (macrophages) and cleaved caspase-3 (apoptotic cells) revealed markedly enhanced efferocytic activity in EXOs^Tβ4/mPDA@MS-treated wounds, with progressively lower levels in the EXOs^vector/mPDA@MS and mPDA@MS groups, and minimal activity in PBS (Fig. 7a and b). Conversely, the number of free apoptotic cells (cleaved caspase-3⁺, CD68⁻) was highest in PBS and lowest in EXOs^Tβ4/mPDA@MS, further indicating improved apoptotic cell clearance (Fig. S24). These results confirm that Tβ4-enriched exosomes effectively enhance efferocytosis in vivo.

Fig. 7.

EXOs^Tβ4/mPDA@MS enhances macrophage efferocytosis in type 1 diabetic wounds. (a, b) Immunofluorescence images (a) and quantitative analysis (b) of wound sections stained for CD68 (green) and cleaved caspase-3 (red) across treatment groups, scale bars 200 μm (up) and 50 μm (down), n = 5. (c, d) Immunofluorescence images (c) of IL-1β (green) in wound sections and quantitative analysis (d) of fluorescence intensity across treatment groups, scale bar 100 μm, n = 5. (e, f) Immunohistochemistry images (e) of TNF-α in wound sections and quantitative analysis (f) of TNF-α densities across treatment groups, scale bar 100 μm, n = 5. (g, h) Immunofluorescence images (g) of wound sections co-stained for F4/80 (green) and CD206 (red) and quantitative analysis (h) of M2 macrophage proportion across treatment groups, scale bar 100 μm, n = 5. (i, j) Immunofluorescence images (i) of wound sections stained for α-SMA (green)/CD31 (red) and quantitative analysis (j) of vessel density across treatment groups, scale bar 100 μm, n = 5. (k, l) Western blot analysis (l) and quantitative analysis (k) of CD206 and Arg-1 in wound tissues across treatment groups (i:PBS, ii: mPDA@MS, iii: EXOs^vector/mPDA@MS, iv: EXOs^Tβ4/mPDA@MS), n = 3. (∗P < 0.05 vs PBS, ^$P < 0.05 vs mPDA@MS, ^@P < 0.05 vs EXOs^vector/mPDA@MS).

We next assessed inflammatory status within the wound microenvironment. Immunofluorescence staining showed a marked reduction in IL-1β expression in EXOs^Tβ4/mPDA@MS-treated wounds (Fig. 7c and d). Similarly, immunohistochemical analysis demonstrated the lowest TNF-α levels in this group, indicating attenuated inflammatory activation (Fig. 7e and f). Consistently, Western blot and RT-qPCR analyses revealed significant downregulation of key pro-inflammatory mediators (IL-18, TNF-α, IL-1β, IL-6, and iNOS) following EXOs^Tβ4/mPDA@MS treatment (Figs. S25 and S26).

Given the critical role of macrophage polarization in tissue repair, we further examined the M2 phenotype. Immunofluorescence staining of CD206 and F4/80 showed a higher proportion of M2 macrophages in the EXOs^Tβ4/mPDA@MS group (Fig. 7g and h). Western blot and RT-qPCR analyses confirmed upregulation of M2-associated markers (Arg-1, CD206, and IL-10), consistent with these observations (Fig. 7k and l; Fig. S27). To evaluate oxidative stress in vivo, dihydroethidium (DHE) staining was performed. EXOs^Tβ4/mPDA@MS-treated wounds exhibited markedly reduced ROS levels compared with all control groups (Fig. S28), indicating effective attenuation of oxidative stress. Together, these findings suggest that EXOs^Tβ4/mPDA@MS reprograms macrophages toward an anti-inflammatory, pro-reparative phenotype. In addition to immune modulation, angiogenesis was assessed as a key indicator of tissue regeneration. Immunofluorescence staining of α-SMA and CD31 revealed significantly enhanced neovascularization in the EXOs^Tβ4/mPDA@MS group compared with all controls (Fig. 7i and j). This improvement is likely driven by coordinated inflammatory resolution and M2 polarization, which collectively establish a pro-angiogenic microenvironment.

Taken together, EXOs^Tβ4/mPDA@MS alleviates excessive inflammation by enhancing efferocytosis and suppressing pro-inflammatory signaling, thereby promoting macrophage polarization toward a reparative M2 phenotype and establishing a regenerative immune microenvironment. Consequently, angiogenesis is enhanced and tissue repair is accelerated, highlighting the therapeutic potential of this bioadhesive exosome-microcarrier spray for diabetic wound treatment. Notably, unlike previous Tβ4-engineered extracellular vesicle strategies that primarily emphasize general pro-healing effects, this study identifies impaired efferocytosis as a key upstream pathological target [33,43]. Moreover, compared with conventional or microparticle based systems, the sprayable mPDA-functionalized microcarriers exhibit robust wet adhesion [[20], [21], [22]]. It improves local retention and enables sustained release. These features facilitate more effective exosome delivery within the diabetic wound microenvironment.

2.7. EXOs^Tβ4/mPDA@MS enhances macrophage efferocytosis through activation of the MerTK/Pi3k/Akt

To elucidate the molecular mechanisms underlying EXOs^Tβ4/mPDA@MS enhanced efferocytosis, transcriptomic profiling was performed. MerTK is a key efferocytosis receptor that is functionally suppressed in diabetic wounds and critically regulates macrophage clearance of apoptotic cells [41]. Its activity depends on downstream Pi3k/Akt signaling [44]. We hypothesized that EXOs^Tβ4 restores the MerTK/Pi3k/Akt axis to enhance macrophage efferocytosis.

Hierarchical clustering and volcano plot analyses identified 2026 differentially expressed genes between EXOs^Tβ4/mPDA@MS and PBS groups, including 1228 upregulated and 798 downregulated genes (Fig. 8a and b). Notably, efferocytosis-related receptors (MerTK, CD36) and bridging molecules (C1q) were significantly upregulated, suggesting enhanced apoptotic cell recognition and clearance. GO enrichment analysis indicated that upregulated genes were associated with extracellular regions, cell surface components, and immunological synapses, while biological processes were enriched in phagocytosis and negative regulation of inflammation (Fig. 8c). Consistently, KEGG analysis highlighted significant enrichment in Pi3k/Akt signaling, clathrin-mediated endocytosis, and apoptotic cell clearance pathways (Fig. 8d). Western blot analysis confirmed marked upregulation of MerTK in EXOs^Tβ4/mPDA@MS-treated macrophages (Fig. 8e and f). Pharmacological inhibition using the MerTK inhibitor UNC2025 significantly attenuated EXOs^Tβ4/mPDA@MS-induced efferocytosis, as demonstrated by immunofluorescence and flow cytometry (Fig. 8g–j). Mechanistically, EXOs^Tβ4/mPDA@MS promoted phosphorylation of Pi3k and Akt, whereas UNC2025 abrogated this activation, confirming that efferocytosis enhancement is mediated via the MerTK/Pi3k/Akt axis (Fig. 8k, l, n). Importantly, in vivo analyses corroborated these findings, showing elevated expression of MerTK, phosphorylated Pi3k, and phosphorylated Akt in wound tissues following EXOs^Tβ4/mPDA@MS treatment, thereby validating the activation of MerTK/Pi3k/Akt in the wound microenvironment (Fig. 8m–o, p).

Fig. 8.

EXOs^Tβ4/mPDA@MS enhances macrophage efferocytosis via MerTK/Pi3k/Akt signaling axis. (a) Hierarchical clustering heatmap of DEGs between EXOs^Tβ4/mPDA@MS and PBS. (b) Volcano plot depicting upregulated (1,228) and downregulated (798) of DEGs. (c) GO enrichment analysis of genes upregulated in macrophages treated with EXOs^Tβ4/mPDA@MS versus PBS. (d) KEGG enrichment analysis of upregulated genes treated with EXOs^Tβ4/mPDA@MS. (e, f) Western blot images (e) and quantification (f) of MerTK following different treatments. (g, h) Fluorescence images (g) and quantification (h) of macrophage efferocytosis after pre-treatment with the MerTK inhibitor UNC2025, scale bar 50 μm. (i, j) Flow cytometry (i) and quantification (j) of macrophage efferocytosis following pre-treatment with the MerTK inhibitor UNC2025. (k, l, n) Western blot analysis (k) and quantification (l, n) of p-Pi3k, Pi3k, p-Akt, and Akt following different treatments. (m, o, p) Western blot analysis (m) and quantification (o, p) of MerTK, p-Pi3k, Pi3k, p-Akt, and Akt in wound tissues in vivo. All n = 3. (∗P < 0.05 vs PBS, ^@P < 0.05 vs EXOs^Tβ4/mPDA@MS).

To exclude potential off-target effects of UNC2025, MerTK was silenced in Raw264.7 cells using siRNA, with efficient knockdown confirmed by RT-qPCR and Western blotting (Figs. S29 and S30). Consistently, MerTK silencing markedly reduced Pi3k and Akt phosphorylation, and EXOs^Tβ4/mPDA@MS failed to rescue this effect, indicating that Pi3k/Akt activation occurs downstream of MerTK (Fig. S30). Functionally, both fluorescence imaging and quantitative analysis showed that MerTK knockdown significantly impaired EXOs^Tβ4/mPDA@MS-mediated efferocytosis, which was further confirmed by flow cytometry (Figs. S31 and S32). Together with pharmacological inhibition, these results demonstrate that EXOs^Tβ4/mPDA@MS enhances macrophage efferocytosis primarily through the MerTK/Pi3k/Akt signaling axis.

While this pathway was identified as a key mediator, upstream regulatory mechanisms remain to be elucidated. In particular, the specific role of Tβ4 within exosomes, including its interaction with other exosomal cargos, was not investigated and warrants further study. Moreover, emerging evidence indicates substantial heterogeneity among stem cell-derived vesicles, with distinct subpopulations exhibiting diverse biological functions beyond classical exosomes [[45], [46], [47]]. Although vesicle subtypes were not distinguished in this study, the enhanced therapeutic efficacy observed here suggests that functional cargo engineering, such as Tβ4 overexpression, can effectively improve pro-reparative outcomes. These findings highlight the potential of integrating vesicle engineering with subtype-specific characterization to further optimize therapeutic strategies for diabetic wound healing.

2.8. EXOs^Tβ4/mPDA@MS accelerates wound healing in db/db mice via enhanced macrophage efferocytosis

To further evaluate the therapeutic efficacy of EXOs^Tβ4/mPDA@MS under type 2 diabetic conditions, full-thickness wounds were created in db/db mice. Representative wound images at days 0, 3, 7, 11, and 14 were shown in Fig. 9a and b. Consistent with the STZ-induced type 1 diabetic model, EXOs^Tβ4/mPDA@MS-treated wounds exhibited markedly smaller residual areas throughout the healing process compared with EXOs^vector/mPDA@MS, mPDA@MS, and PBS. Quantitative analysis confirmed that EXOs^Tβ4/mPDA@MS achieved the fastest wound closure among all groups (Fig. 9c). Histological analysis further demonstrated enhanced tissue regeneration. H&E staining revealed a thicker neoepidermis and narrower wound width in the EXOs^Tβ4/mPDA@MS group at day 14 (Fig. 9d–f). Masson's trichrome staining showed more abundant and well-organized collagen deposition, indicating improved extracellular matrix remodeling (Fig. 9g and h).

Fig. 9.

EXOs^Tβ4/mPDA@MS accelerates wound healing in db/db mice by enhancing macrophage efferocytosis. (a) Representative images of db/db mouse wounds treated with PBS, mPDA@MS, EXOs^vector/mPDA@MS, or EXOs^Tβ4/mPDA@MS on different days, scale bar 5 mm. (b) Macroscopic wound morphology across different treatment groups at indicated time points. (c) Quantitative evaluation of wound closure dynamics over time, n = 5. (d) Representative H&E images of wound tissues collected, scale bar 1 mm (up) and 200 μm (down). (e, f) Quantification of tissue thickness and tissue width on day 14 based on H&E staining, n = 5. (g) Representative Masson's trichrome staining on day 14, scale bars 1 mm (up) and 200 μm (down). (h) Quantitative assessment of collagen deposition, n = 5. (i) Representative immunofluorescence images of wound tissues stained for CD68 (green, macrophages) and cleaved caspase-3 (red, apoptotic cells), scale bars 200 μm (up) and 50 μm (down). (j) Quantification of macrophage efferocytosis index, n = 5. (∗P < 0.05 vs PBS, ^$P < 0.05 vs mPDA@MS, ^@P < 0.05 vs EXOs^vector/mPDA@MS).

To explore the underlying mechanism, macrophage efferocytosis was assessed by immunofluorescence staining for CD68 and cleaved caspase-3 (Fig. 9i and j). EXOs^Tβ4/mPDA@MS significantly enhanced apoptotic cell clearance, yielding the highest efferocytosis index among all groups. Notably, these results were consistent with those observed in STZ-induced mice, confirming that enhanced efferocytosis contributes to improved wound repair under type 2 diabetic conditions. In addition, macrophage polarization and angiogenesis were evaluated. Immunofluorescence staining of CD206 and F4/80 revealed a higher proportion of CD206⁺ M2 macrophages in the EXOs^Tβ4/mPDA@MS group (Fig. S33a and b), consistent with a pro-reparative immune phenotype. CD31 staining further demonstrated increased vascular density in EXOs^Tβ4/mPDA@MS-treated wounds, whereas control groups showed sparse and discontinuous microvessels (Fig. S34a and b), indicating enhanced angiogenesis.

Collectively, EXOs^Tβ4/mPDA@MS accelerates wound healing in db/db mice by enhancing macrophage efferocytosis, promoting pro-regenerative immune remodeling, and improving tissue reconstruction, highlighting its therapeutic potential across both type 1 and type 2 diabetic wounds.

3. Conclusion

In summary, we developed a sprayable and bioadhesive microcarrier platform (EXOs^Tβ4/mPDA@MS) that enables stable, localized, and immune-targeted delivery of Tβ4-engineered ADSC-derived exosomes for chronic diabetic wound therapy. Leveraging microfluidic micron-scale fabrication and mPDA-functionalized hydrogel architectures, the microcarriers exhibit uniform morphology, strong wet adhesion, high exosome stability, and efficient deposition on moist wound surfaces. Mechanistically, EXOs^Tβ4/mPDA@MS reprograms macrophage behavior by activating the MerTK/Pi3k/Akt signaling axis, thereby enhancing efferocytosis, promoting M2 polarization, and reducing intracellular ROS accumulation. These coordinated immunoregulatory actions translate into accelerated wound closure, improved epithelial regeneration, organized collagen remodeling, and augmented angiogenesis in diabetic wounds. Collectively, this engineered exosome-microcarrier spray provides a multifunctional and clinically adaptable strategy that disrupts chronic inflammatory circuits while promoting tissue regeneration, offering a promising direction for advancing next-generation exosome-based biomaterials toward translational application.

4. Methods and materials

4.1. Materials

Methacrylic anhydride (MA), sodium bicarbonate, and polydopamine were obtained from Aladdin (Shanghai, China). Gelatin methacryloyl (GelMA) was synthesized in-house. The GFP-tagged Tβ4 lentivirus was purchased from Yanmingsheng Biotechnology (Shenzhen, China). Fetal bovine serum was purchased from Gibco (USA). A Live/Dead cell staining kit was supplied by Beyotime Biotechnology (Shanghai, China). Antibodies against CD206, CD86, F4/80, CD44, CD73, CD90, CD34, CD45, and CD105 were acquired from BD Pharmingen™. Antibodies against Tβ4, TSG101, CD63, and CD9 were provided by Proteintech (Wuhan, China). C57BL/6 and db/db mice were obtained from Sellingmice Biotechnology (Nanjing, China).

4.2. Cell culture

ADSCs (Cat. No. 7510, ScienCell, USA) were used in this study. ADSCs were transduced with lentivirus to overexpress Tβ4 and maintained in DMEM/F-12. Raw264.7 macrophages were sourced from the Cell Bank of the Chinese Academy of Sciences. Bone marrow-derived macrophages (BMDMs) were harvested from the femurs and tibias of C57BL/6 mice.

4.3. Isolation and characterization of EXOs^Tβ4

ADSCs^Tβ4 were cultured in exosome-depleted medium, and the conditioned medium was collected. EXOs^Tβ4 were then isolated by sequential ultracentrifugation according to established protocols [33]. TEM and NTA were employed to characterize the morphology and size distribution of EXOs^Tβ4.

4.4. Preparation and characterization of EXOs^Tβ4/mPDA@MS sprayable microspheres

Microfluidic device fabrication: The microfluidic system was assembled by coaxially aligning two circular capillaries. The inner capillary was pulled and polished to a tip diameter of ∼100 μm using a capillary puller (Sutter Instrument, P-97), while droplets were collected with a capillary (OD 1 mm, ID 550 μm). During operation, fluids were delivered into the device via syringe pumps (Harvard Apparatus, PHD 2000 series), and the flow rates were precisely regulated to modulate droplet size.

Preparation of EXOs^Tβ4/mPDA@MS: Specifically, EXOs^Tβ4/mPDA@MS were fabricated using a microfluidic method, in which the inner phase consisted of 60 μL of a solution containing 1 wt% HMPP, 10 wt% GelMA, and 200 μg EXOs^Tβ4, while the outer phase comprised paraffin oil supplemented with 10 wt% Span 80, according to our previous study [33]. For in vitro studies, macrophages were treated with EXOs^Tβ4/mPDA@MS at an exosome concentration of 50 μg/mL. For in vivo experiments, EXOs^Tβ4/mPDA@MS was locally applied to the wound area at a dose of 50 μg exosomes per wound at each treatment time point via the sprayable microcarrier system. Typical flow rates for the inner and outer phases were 1 μL/min and 25 μL/min, respectively. The resulting droplets were collected at 4°C and crosslinked under UV irradiation (OmniCure S1000). Solidified microspheres were harvested in culture dishes, washed with PBS to remove oil, and centrifuged (100 × g, 3 min). The collected microspheres were then dispersed in PBS containing 1.0% (w/v) mPDA and gently stirred at 4°C for 30 min, followed by washing to eliminate unpolymerized mPDA and centrifugation (100 × g, 3 min). The morphology and sphericity of the microspheres were examined using an optical microscope, and their size distribution was determined by optical microscopy. The microstructure was further characterized by SEM.

Swelling assay: For the swelling test, 0.01 g of lyophilized EXOs^Tβ4/mPDA@MS was pre-weighed and placed in a 5 mL centrifuge tube. The tube was then incubated on a shaker at 37°C and 200 rpm with the addition of 1 mL PBS. At designated time intervals, the supernatant was removed and the residual microspheres were weighed. The swelling ratio was calculated according to the following equation:

$$\text{Swelling}\text{ratio}(\%)=(\text{Wt}‐\mathrm{W}0)\mathrm{W}0\times 100\%$$

where Wt represents the weight of the swollen microspheres and W0 represents the initial dry weight.

Drug release assay: To evaluate the release kinetics of Dil-labeled exosomes from EXOs^Tβ4/mPDA@MS, the microspheres were immersed in PBS in 24-well plates. At designated time points, the culture supernatant was harvested. The fluorescence intensity of the released exosomes was measured using a microplate reader, with an excitation wavelength set at 570 nm.

Biocompatibility assessment: The cytotoxicity of EXOs^Tβ4/mPDA@MS was evaluated using HUVECs and HDFs. Cells were cultured in standard medium in the control group, whereas those in the experimental group were co-cultured with EXOs^Tβ4/mPDA@MS. On days 1 and 2, the viability was examined using a Live/Dead staining kit (1:1000 dilution, 30 min), followed by confocal imaging.

4.5. Efferocytosis assay

Jurkat were subjected to UV irradiation (254 nm) for 20 min to induce apoptotic cells (ACs) and subsequently maintained in a humidified incubator for 3 h. The resulting ACs were stained with the lipophilic fluorescent dye CM-Dil, whereas macrophages (Mφ) were labeled with CMFDA. Macrophages were then co-incubated with ACs at a 1:1 cell ratio for 4 h. Following co-culture, uninternalized ACs were removed by PBS washing, and macrophages were fixed with 4% paraformaldehyde. Efferocytosis was visualized using fluorescence and confocal microscopy.

4.6. Flow cytometry

Cells were divided into four groups as described above. Raw264.7 and BMDM cells were prepared as single-cell suspensions. M2 macrophage polarization was assessed using FITC-conjugated F4/80 antibody and PE-conjugated CD206 antibody. Cells were incubated with the indicated stains for 30 min in the dark and subsequently analyzed by flow cytometry using a CytoFLEX instrument (Beckman Coulter, Miami, FL, USA).

4.7. RT-qPCR

The RNA was extracted from cultured cells or harvested animal tissues using a commercially available RNA isolation kit (Vazyme, Nanjing, China) according to the manufacturer's instructions. The procedure ensured efficient recovery of high-quality RNA suitable for subsequent molecular analyses. RNA quality and purity were evaluated using a spectrophotometer (Thermo Fisher, USA) based on A260/A280 and A260/A230 ratios. Purified RNA was then reverse transcribed to generate cDNA. RT-qPCR was performed using a SYBR Green-based detection system (Vazyme, China). The relative expression levels of target genes were calculated after normalization to the corresponding endogenous reference genes. Detailed primer information used for amplification is summarized in Table S1.

4.8. Cell immunofluorescence staining

Cells were incubated with the designated primary antibodies overnight at 4°C, enabling specific interaction with their corresponding antigens. After incubation, unbound antibodies were removed by multiple washes with PBST to minimize nonspecific fluorescence. Subsequently, cells were exposed to fluorescently conjugated secondary antibodies for 1 h at room temperature in darkness to preserve signal integrity. After additional PBST washes to remove excess secondary antibodies, cell nuclei were counterstained with DAPI to visualize nuclear morphology. Finally, fluorescent images were acquired using a high-resolution fluorescence microscope and subsequently analyzed to assess protein localization and expression patterns.

4.9. Antioxidant assay

Raw264.7 were plated in 24-well culture and subjected to H₂O₂ stimulation to induce oxidative stress-related damage. The cells were then treated with PBS, mPDA@MS, EXOs^vector/mPDA@MS, or EXOs^Tβ4/mPDA@MS for 24 h. A negative control group without H₂O₂ induction was included. After treatment, cells were incubated with DCFH-DA, and fluorescent signals were recorded using a fluorescence microscope. In parallel, the percentage of ROS-positive cells was determined by flow cytometric analysis following the same staining protocol.

4.10. Western blotting

The proteins were extracted using RIPA supplemented with phosphatase and protease inhibitor cocktails (1% each) to prevent protein degradation and dephosphorylation during lysis. Protein concentrations in the lysates were quantified using a BCA assay according to the manufacturer's instructions. Equal amounts of protein from each sample were then separated by SDS-PAGE and transferred onto PVDF membranes. The membranes were subsequently blocked and probed with specific primary and corresponding secondary antibodies, followed by detection using standard Western blotting procedures as previously described [43,48,49].

4.11. Animal experiments

Type 1 diabetes was induced in C57BL/6 mice according to an established protocol with minor modifications [40,48]. Briefly, mice were deprived of food for 12 h and then intraperitoneally administered streptozotocin (STZ, 60 mg/kg) once daily for five consecutive days. Successful induction of type 1 diabetes was verified by sustained hyperglycemia, defined as blood glucose concentrations exceeding 16.7 mmol/L. Type 2 diabetic models were generated using db/db mice. The wounds were surgically created on the dorsal skin of mice, which were subsequently randomized into four treatment groups: PBS, mPDA@MS, EXOs^vector/mPDA@MS, and EXOs^Tβ4/mPDA@MS. EXOs^Tβ4/mPDA@MS was applied once every 3 days, and the wound site was gently rinsed with saline prior to each administration to remove residual microcarriers. Wound healing was monitored by photographic documentation on days 0, 3, 7, 11, and 14. Wound tissues were harvested for histological and immunological evaluations, including H&E and Masson's trichrome staining, immunohistochemical analysis (TNF-α), and immunofluorescence staining (CD206, F4/80, IL-1β, CD31, α-SMA, CD68, and cleaved caspase-3). All animal procedures were conducted in compliance with the regulations approved by the Experimental Animal Ethics Committee of Drum Tower Hospital, Nanjing University Medical School (Approval No. 2024AE01104).

4.12. Statistical analysis

All quantitative data were analyzed using GraphPad Prism (version 10.2). Student's t-test was used for comparison between two groups, while analysis of variance (ANOVA) model was employed for comparison among multiple groups. The data are presented as mean ± Standard Error of the Mean (SEM) of the measured values. Statistical significance was considered at ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.001 ^$P < 0.05, ^$$P < 0.01, ^@P < 0.05, ^@@P < 0.01, ^@@@P < 0.001, while non-significant differences are denoted as ns (p > 0.05).

Data and materials availability

Data will be made available on request.

Ethics approval and consent to participate

All animal procedures were conducted in compliance with the regulations approved by the Experimental Animal Ethics Committee of Drum Tower Hospital, Nanjing University Medical School (Approval No. 2024AE01104).

CRediT authorship contribution statement

Youjun Ding: Conceptualization, Investigation, Methodology, Validation, Writing – original draft, Writing – review & editing. Danqing Huang: Methodology, Writing – review & editing. Zhiwei Zhao: Methodology, Writing – review & editing. Han Feng: Methodology, Writing – review & editing. Shuyin Zhou: Methodology, Validation. Danni Qian: Methodology, Validation. Yepeng Zhang: Supervision, Writing – review & editing. Yuanjin Zhao: Conceptualization, Funding acquisition, Project administration, Supervision. Min Zhou: Conceptualization, Funding acquisition, Supervision, Writing – review & editing.

Declarations of competing interest

The authors declare no competing financial interests. Yuanjin Zhao is an editorial board member for Bioactive Materials and was not involved in the editorial review or the decision to publish this article. All authors declare that there are no competing interests.

Appendix A. Supplementary data

The following are the Supplementary data to this article.