Accelerated diabetic wound healing via microenvironmental modulation treated by hydrogel encapsulated with neural stem cells

Abstract

Introduction

Diabetic wound healing is a highly coordinated, multi-stage process that relies critically on the pro-regenerative microenvironment. Recently, stem cell-based therapies have emerged as a promising paradigm in regenerative medicine, largely attributable to their inherent self-renewal capacity, multilineage differentiation potential, and pro-repair secretome. Despite these advances, the therapeutic potential of neural stem cells (NSCs) in cutaneous wound repair remains largely unexplored.

Methods

We systematically evaluated the pro-angiogenic, antioxidant, and mitochondrial modulatory effects of NSCs on endothelial cells. A bio-inspired hydrogel was designed for NSC encapsulation with favorable preliminary biocompatibility. The in vivo pro-healing efficacy of NSCs-armed hydrogel was then assessed in diabetic mice.

Results

NSCs enhanced angiogenesis, antioxidant capacity, and mitochondrial function in endothelial cells. The engineered hydrogel exhibited favorable preliminary biocompatibility for the encapsulation of NSCs. This platform accelerated diabetic wound healing by regulating inflammation, promoting angiogenic, and exerting neural-supportive effects.

Conclusions

This NSC-loaded hydrogel platform offers a promising and clinically relevant strategy for diabetic wound repair.

1. Introduction

Diabetic foot ulcers are among the most prevalent and severe complications of diabetes mellitus, presenting a significant clinical challenge due to their high rates of disability and mortality [[1], [2], [3], [4]]. These ulcers also exhibit a substantial recurrence rate, reported at 65% within 3 to 5 years [4]. Although conventional treatments such as surgery, negative pressure therapy, and hyperbaric oxygen therapy are available, their clinical utility is often limited by high costs and suboptimal patient compliance. Diabetic wound healing is frequently impaired by multiple interrelated pathological features, chief among them being inadequate angiogenesis and heightened oxidative stress [5]. A growing body of evidence further underscores the importance of neural regulation in wound repair. For instance, Schwann cells at the wound edge promote the migration of neighboring cells through the secretion of TGF-β3 [6]. Calcitonin gene-related peptide (CGRP) sensory neurons also contribute by orchestrating a neuro-immune axis, inducing TSP-1 expression in neutrophils and macrophages to advance the healing response [7].

Stem cell-based regenerative therapies have consequently emerged as a promising therapeutic avenue. The efficacy of stem cells primarily stems from their capacity for self-renewal, multi-lineage differentiation, and the secretion of pro-repair factors. Various stem cell types, including adipose-derived stem cells [8,9], bone marrow-derived mesenchymal stem cells [10], ectodermal mesenchymal stem cells [11], and human umbilical cord mesenchymal stem cells [12], have demonstrated substantial efficacy in promoting wound closure in diabetic models. However, neural stem cells (NSCs), a distinct stem cell population, remain largely underexplored in the context of diabetic wound healing, despite their well-documented efficacy in the management of neurological diseases such as traumatic brain injury [13] and spinal cord injury [14,15]. A potential advantage of NSCs lies in their inherent capacity to exert neural-supportive effects.

A critical translational hurdle for cell-based therapy is the rapid loss of cells within the hostile pathological wound environment [16,17]. Hydrogel scaffolds, with their high content of water and tunable three-dimensional architecture, offer an ideal solution [[18], [19], [20]]. As biomimetic extracellular matrix (ECM) that supports cell viability and functions, hydrogel can be engineered to respond to specific pathological cues. Hyaluronic acid, a naturally occurring linear polysaccharide and key component of the human ECM, exhibits low immunogenicity [[21], [22], [23]]. The abundant functional groups along its molecular chains, particularly carboxyl and hydroxyl moieties, facilitate chemical modifications and multiple gelation strategies. Moreover, the mechanical properties of hyaluronic acid hydrogels can be finely tuned via gelation conditions, allowing customization for diverse clinical applications.

In this study, we first revealed that NSCs capably protect endothelial cells against oxidative stress and restore mitochondrial function. Subsequently, we designed a bio-inspired hydrogel system for NSC encapsulation to modulate the diabetic wound microenvironment. The hydrogel was constructed by crosslinking phenylboronic acid-modified hyaluronic acid methacryloyl with polyvinyl alcohol, forming a three-dimensional network with dynamic bonding properties. This NSC-loaded hydrogel system, serving as a novel combinatory therapeutic strategy, underwent comprehensive evaluation for in vitro cytocompatibility and in vivo biosafety. In diabetic mouse models, our formulation modulated the immune microenvironment, promoted angiogenesis, and exerted neural-supportive effect, and ultimately facilitated wound closure. Collectively, our work presents a hydrogel platform for NSCs, offering a novel strategy for efficient diabetic wound repair (Scheme 1).

Scheme 1.

Modulation of diabetic wound microenvironment by NSC-laden hydrogel. The hydrogel was developed by grafting methacryloyl and phenylboronic acid groups onto a hyaluronic acid backbone, followed by mixing with polyvinyl alcohol (PVA) and exposure to blue light, to achieve gelation and efficient encapsulation of NSCs. Upon local application to diabetic wounds, this system effectively attenuated inflammation, reduced oxidative stress damage, preserved mitochondrial homeostasis, and enhanced angiogenesis. In addition, it promoted collagen deposition and supported nerve regeneration. Collectively, this integrated strategy translates into accelerated wound closure and improved tissue regeneration.

2. Results

2.1. Therapeutic functionality of NSCs

NSCs, isolated from the embryos of C57BL/6 mice, exhibited characteristic neurosphere formation under bright-field microscopy (Fig. S1). Their phenotypic identity was further confirmed by immunofluorescence staining for the canonical markers Nestin and GFAP (Fig. 1A).

Fig. 1.

Characterization and functional evaluation of NSCs. (A) Identification of neural stem cells by immunofluorescence with specific markers including Nestin and GFAP (scale bar = 200 μm). (B) Representative tube formation image of HUVECs by different treatments (scale bar = 200 μm). (C) Flow cytometric analysis of ROS in HUVECs with indicated treatments. (D) Fluorescence images of intracellular ROS in HUVECs after various treatments (scale bar = 100 μm). (E) Flow cytometry analysis of JC-1 staining of HUVECs. (F) Representative fluorescent images JC-1 staining in HUVECs (scale bar = 100 μm). (G) Quantification of total branching length from tube formation assay (n = 4). (H) Quantification of mean fluorescence intensity (MFI) in HUVECs by flow cytometry (n = 4). (I) Quantification of MFI of intracellular ROS by fluorescent images (n = 4). (J) Quantitative analysis of JC-1 aggregates in HUVECs (n = 3). Statistical significance was determined by one-way ANOVA and is presented as ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

To further characterize their functionality, NSCs were subjected to a series of in vitro assays. Angiogenesis is essential for effective wound healing, fulfilling two critical functions: restoring local blood perfusion to deliver oxygen and essential nutrients to the injured tissue, and supporting recruitment of repair-associated cells to the wound site [24,25]. To evaluate the pro-angiogenic potential of our hydrogel system, we performed tube formation assay using human umbilical vein endothelial cells (HUVECs) treated by NSCs. Relative to the control group, NSC treatment significantly increased total branching length with well-organized tubular network formation (Fig. 1B and 1G).

Compared to normoglycemic wounds, diabetic wounds are characterized by significantly elevated oxidative stress [26,27]. To assess the cytoprotective effect of NSCs, we established a hydrogen peroxide (H₂O₂)-induced oxidative injury model in HUVECs co-cultured with NSCs (Fig. S2). Treatment with the NSCs markedly reversed the H₂O₂-induced rise in intracellular ROS, as evidenced by substantially weakened green fluorescence in DCFH-DA staining and a leftward shift in the flow cytometry peak (Fig. 1C, 1D, 1H, 1I). Also, mitochondria serve as the cellular powerhouses that generates adenosine triphosphate via oxidative phosphorylation to meet energy demands and also orchestrates key metabolic pathways essential for cell survival. Mitochondrial membrane potential (MMP), a well-established marker of mitochondrial integrity, is known to be disrupted by oxidative stress. Using JC-1 staining combined with flow cytometry and fluorescence microscopy, we observed that H₂O₂ treatment increased the proportion of JC-1 monomers (green fluorescence), indicating depolarization of MMP. In contrast, application of NSCs restored the ratio of JC-1 aggregates (red fluorescence), as reflected by enhanced red fluorescence and diminished green fluorescence (Fig. 1E, 1F, 1J). This finding confirms the restoration of MMP and further highlights the mitochondrial protective function conferred by our NSC therapy. In parallel, we assessed changes in mitochondrial superoxide levels. Consistently, NSC treatment effectively reversed the H₂O₂-induced upregulation of MitoSOX in HUVECs, indicating a recovery of mitochondrial redox homeostasis (Fig. S3).

Collectively, NSCs simultaneously promote angiogenesis, mitigates oxidative stress and preserves mitochondrial integrity, thereby enhancing the reparative capacity of vascular endothelial cells in a simulated diabetic wound microenvironment.

2.2. Preparation and characterization of hydrogel loaded with neural stem cells

Despite the confirmed functionality of NSCs, direct injection of cell suspension into wound sites is hampered by rapid cell loss. To overcome this limitation, we sought to employ a hydrogel-scaffold for NSCs encapsulation. Phenylboronic acid-modified methacrylated hyaluronic acid (HAMA-PBA) was selected as the backbone material. Through crosslinking with polyvinyl alcohol (PVA), a hydrogel system was therefore constructed. The introduced methacryloyl groups also enabled photo-crosslinking under 405 nm blue light in the presence of the photo-initiator lithium phenyl-2,4,6-trimethylbenzoylphosphinate, allowing precise modulation of the hydrogel's mechanical properties.

At macroscopic level, the hydrogel precursor solution remained in a liquid state prior to crosslinking. After mixing HAMA-PBA with PVA and exposing the mixture to 405 nm blue light, rapid gelation occurred, forming a stable hydrogel that retained its shape without flowing when inverted, indicating the formation of a robust three-dimensional network (abbreviated as Gel) (Fig. 2A). 3D reconstruction verified the successful encapsulation and distribution of NSCs within hydrogel matrix (Fig. 2B). ¹H Nuclear magnetic resonance (NMR) spectroscopy confirmed successful grafting, showing characteristic hydrogen signals assigned to the methacryloyl (labeled 1) and phenylboronic acid (labeled 2) groups in the spectrum (Fig. 2C). Fourier-transform infrared spectroscopy (FTIR) further showed the expected characteristic absorption peaks, consistent with the designed chemical structure. Typically, a distinct stretching vibration at around 1712 cm⁻¹ is observed in HAMA and HAMA-PBA but not in HA; meanwhile, a characteristic peak at around 1151 cm⁻¹ corresponding to the B–O bond appears exclusively in HAMA-PBA (Fig. 2D). Compression tests were conducted to evaluate the mechanical performance of the hydrogel, and the results showed that the gel withstood a compressive strain of 75.3% before failure, corresponding to a compressive stress of 18,874 Pa (Fig. 2E). Scanning electron microscopy images revealed a honeycomb-like porous morphology, and energy-dispersive X-ray spectroscopy validated the presence of C, N, O, P, and B elements, verifying the intended compositional distribution (Fig. 2F and 2G). Moreover, in vitro degradation tests demonstrated that approximately 60% of the hydrogel mass remained after 15 days, confirming that its degradation profile aligns with the required timeframe for diabetic wound therapy (Fig. S4).

Fig. 2.

Preparation and characterization of composited hydrogels. (A) Gelation of hydrogel by mixing HAMA-PBA and PVA under 405 nm blue-light exposure. (B) 3D reconstruction of confocal laser scanning microscope image of neural stem cells encapsulated in hydrogel (scale bar = 200 μm). (C) ¹H NMR analysis of HA, HAMA and HAMA-PBA. (D) FTIR of HA, HAMA, HAMA-PBA and PVA. (E) Compression test of hydrogel. (F) Scanning electron microscopy images of hydrogel (scale bar = 100 μm). (G) Energy dispersive X-ray spectroscopy element mapping of hydrogel matrix. HA, hyaluronic acid. HAMA, hyaluronic acid methacryloyl. Phenylboronic acid modified hyaluronic acid methacryloyl, HAMA-PBA. Poly(vinyl alcohol), PVA.

Taken together, our hydrogel platform potentially represents a compatible scaffold for encapsulation of NSCs.

2.3. Biocompatibility and functionality of composite hydrogels

Biosafety is a fundamental prerequisite for any material intended for therapeutic application. To assess the potential cytotoxicity of our hydrogel system, we co-cultured HUVECs with different treatments. Cell viability was evaluated using live/dead staining at three timepoints: 24, 48, and 72 h. Across all different groups, most of cells exhibited strong green fluorescence (indicative of live cells), while minimal red fluorescence (indicative of dead cells) was observed throughout the culture period (Fig. 3A). The results were in line with the CCK-8 assay data, demonstrating great cell viability (Fig. S5). Favorable preliminary biosafety profile of the composite material supports its suitability for further therapeutic evaluation.

Fig. 3.

Biocompatibility of NSC-laden hydrogels. (A) Representative live/dead staining of HUVECs with different treatments for 24h, 48h and 72h (scale bar = 200 μm). (B) HE staining of major organs after hydrogel implantation (Scale bar = 100 μm).

In addition, we conducted a subcutaneous implantation study in mice to evaluate the in vivo biocompatibility of the hydrogel. Following 15 days of continuous implantation, major organs (including heart, liver, spleen, lung, and kidney) were harvested and subjected to hematoxylin and eosin (H&E) staining. Histopathological analysis revealed no signs of significant structural damage or pathological abnormalities in the examined organs, indicating no significant organ toxicity (Fig. 3B). Furthermore, blood routine analysis showed no significant deviations in key hematological parameters, confirming that the hydrogel did not induce systemic hematological alterations (Fig. S6). In summary, these results confirm that the developed hydrogel exhibits no obvious short-term toxicity under the tested conditions, underscoring its potential for further therapeutic applications.

Furthermore, we performed a preliminary assessment of the functional performance of the composite hydrogel. The results showed that, compared with the control and hydrogel-only groups, the NSC-loaded hydrogel modulated oxidative stress and mitochondrial membrane potential, and promoted angiogenesis, establishing a foundation for subsequent in vivo experiments (Fig. S7).

2.4. Hydrogel platform promotes wound healing in diabetic mice

Building on the demonstrated pro-angiogenic and stress-resistant properties of NSCs in vitro, we next sought to evaluate its efficacy in a diabetic mouse wound model. A full-thickness excisional wound was created on the dorsum of each mouse, and animals were divided into four groups: untreated control, NSCs only, hydrogel only (abbreviated as Gel), and NSC-loaded hydrogel (abbreviated as NSC@Gel). Wound area was monitored over 12 days through digital imaging and analysis (Fig. 4A).

Fig. 4.

Accelerated diabetic wound healing in mice treated by composite hydrogel. (A) Schematic illustration of animal experiments in this study. (B) Representative macroscopic images of wound beds. (C) Traced wound area progression. Brown: Day 0; Blue: Day 3; Orange: Day 6; Green: Day 12. (D) Quantification of the residual wound area (n = 5). (E) Quantification of wound length at 12 days post-wounding (n = 5). (F) Quantification of collagen deposition at 12 days post-wounding (n = 5). (G) HE staining of wound tissues at 12 days post-wounding (scale bar = 1000 μm). (H) Masson staining of wound tissues at 12 days post-wounding (scale bar = 100 μm). Statistical significance was determined by one-way ANOVA and presented as ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. Simple size n refers to number of wounds included in each evaluation.

At Day 3 and Day 6 after wounding, the untreated group still displayed pronounced unhealed areas. In contrast, both NSC and hydrogel monotherapy led to varying degrees of wound reduction, among which the composite hydrogel group showed the most obvious wound contraction. At Day 12, control group displayed substantial unhealed regions, while visible wound remnants were detected in groups treated with either NSCs or hydrogel alone. Remarkably, the NSC@Gel group showed superior wound closure efficacy at Day 12 (Fig. 4B, 4C, 4D, Fig. S8). Histological assessment by H&E staining revealed prominent unhealed areas in the control group. Treatment with NSCs or hydrogel alone partially promoted wound closure, while the NSC@Gel group exhibited markedly reduced wound length and more mature and uniformly formed epidermis. Compared with the sparse hair follicle structures observed in the control and hydrogel-only groups, treatment with NSC alone increased hair follicle density, whereas the composite hydrogel group displayed more densely distributed follicles (Fig. 4E and 4G). Furthermore, Masson's trichrome staining indicated markedly increased and better-aligned collagen deposition in wound tissues treated by NSC@Gel, suggesting improved extracellular matrix remodeling (Fig. 4F and 4H).

2.5. NSC-laden hydrogel exerts inflammation-regulating, angiogenic, and neural-supportive effects in diabetic wounds

To elucidate the mechanism by which the hydrogel promotes diabetic wound healing, this study focused on multiple pivotal events in tissue repair: inflammatory regulation, angiogenesis and neural-supportive effect.

Interleukin-6 (IL-6) was selected as a key inflammatory marker due to its central role in the inflammatory cascade. In regenerating wounds, the NSC@Gel group exhibited a marked reduction in IL-6 immunofluorescence intensity compared to the control group, whereas treatment with NSCs or hydrogel alone failed to achieve sufficient resolution of inflammation (Fig. 5A and 5F). Alpha-smooth muscle actin (α-SMA) was used as an indicator of vascular regeneration. In the control group, α-SMA expression was negligible. Both the hydrogel-alone group (which lacked therapeutic cargo) and the group treated with NSCs alone (which might suffer from rapid cell loss due to the absence of scaffold-mediated encapsulation) exhibited only limited α-SMA-positive signals. In contrast, the composite hydrogel (NSC@Gel) group showed abundant α-SMA expression. (Fig. 5B and 5G). CD31 specifically labels endothelial cells, allowing for preliminary quantification of micro-vessel density. Our results showed that vascular density was low in the control group and the hydrogel-only group, whereas the composite hydrogel group exhibited the highest vascular density, outperforming the NSC-only group (Fig. 5C and 5H).

Fig. 5.

Immunofluorescence staining of wound tissues by different treatments. Representative immunofluorescent images of (A) IL-6 (scale bar = 80 μm), (B) α-SMA (scale bar = 50 μm), (C) CD31 (scale bar = 50 μm), (D) CGRP (scale bar = 100 μm), and (E) NF200 (scale bar = 50 μm) on Day12 post-treatment. (F) Quantification of relative fluorescence intensity of IL-6 (n = 5). (G) Quantification of relative α-SMA-positive signal (n = 5). (H) Quantification of relative number of vessels (n = 5). (I) Quantification of relative fluorescence intensity of CGRP (n = 5). (J) Quantification of relative fluorescence intensity of NF200 (n = 5). Statistical analysis was conducted using one-way ANOVA, with significance defined as ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001. Simple size n refers to number of wounds included in each evaluation.

Given the emerging recognition of neural and sensory regulation in wound repair, we further assessed the neural-supportive effect of NSCs. CGRP immunofluorescence was barely detectable in the control and hydrogel-alone groups, whereas NSC treatment markedly enhanced CGRP-positive areas. The NSC@Gel group achieved increased CGRP expression, likely attributable to the neuronal differentiation and secretion of neurotrophic factors by NSCs (Fig. 5D and 5I). Additionally, immunostaining of NF200 was performed. The results showed that the NSC-only group exhibited relatively prominent NF200 expression, whereas it was rarely observed in the control group and the hydrogel-only group. Notably, the composite hydrogel group displayed the most pronounced NF200 expression, indicating an increase in nerve distribution (Fig. 5E and 5J).

Together, these findings demonstrate that NSC-loaded hydrogel system facilitates diabetic wound healing through the coordinated modulation of the inflammatory microenvironment, coupled with enhanced angiogenesis and neural-supportive effects.

3. Discussion

Diabetic wound healing is a highly complex biological process. Effective repair requires both robust angiogenesis to restore nutrient and oxygen supply and sufficient cellular resilience to withstand the hostile wound microenvironment [28,29]. Hyperglycemia elevates oxidative stress, which severely compromises cellular function, particularly mitochondrial homeostasis. The critical regulatory role of sensory and neural innervation in cellular behavior has gained considerable attention, inspiring strategies that target the neural microenvironment to actively accelerate wound repair [[30], [31], [32], [33]].

Stem cell-based therapies represent a significant advance in regenerative medicine. NSCs possess potent differentiation capacity and a powerful secretome that modulates inflammation, promotes vascularization, and protects against oxidative damage. Under oxidative stress, NSCs confer cytoprotection by reducing intracellular ROS and mitochondrial superoxide levels while restoring mitochondrial membrane potential, thereby enhancing cellular fitness in a simulated diabetic environment. To improve their translational utility for diabetic wounds, we embedded NSCs within a hydrogel matrix. We evaluated the therapeutic efficacy of this hydrogel in a full-thickness cutaneous wound model in diabetic mice, where it markedly accelerated wound healing. Histological analyses revealed that the composite hydrogel orchestrated a favorable microenvironment through immunomodulatory and pro-angiogenic activities. It concurrently upregulated the expression of CGRP and NF200. Compared to other stem cell types, such as adipose-derived stem cells or bone marrow mesenchymal stem cells, NSCs potentially offer advantages in neural regeneration-related functions, due partially to the ability to survive, integrate, and differentiate in vivo following transplantation. These integrated effects underscore the potential of NSCs to coordinate the repair program in chronic wounds.

Several aspects of this study could be further strengthened in future investigations. A more comprehensive understanding of the fate of transplanted NSCs could be achieved through the application of techniques such as in vivo cell tracking, which would offer deeper insight into their retention, long-term survival, integration, and differentiation potential. The therapeutic advantages of this platform could be further demonstrated through experimental comparisons with mesenchymal stem cells. Moreover, future investigations on hydrogel-based NSC delivery should include direct comparisons of cell persistence between free-cell injection and hydrogel-based approaches. Such knowledge is essential for optimizing delivery strategies and maximizing therapeutic efficacy. Additionally, more comprehensive mechanistic studies are needed to clarify the pathways underlying NSC-mediated therapeutic effects, particularly, the roles of inflammatory modulation, vascular remodeling, and neural regulation by expanding the panel of immune-neuro-vascular markers. Elucidating these mechanisms would not only strengthen the mechanistic basis of NSC-based therapies but also inform the rational design of combination strategies.

While murine diabetic wound model is widely accepted as a standard preclinical platform, it is important to note that wound closure in mice is heavily influenced by contraction, a process that differs substantially from healing observed in humans. Future studies would therefore benefit from employing a splinting strategy to minimize this confounding effect, thereby better recapitulating human wound healing, along with validation in infected chronic wound models. Finally, translating this NSC-loaded hydrogel platform toward clinical applications will require a more thorough evaluation of its long-term safety, functional outcomes, microenvironment responsiveness and in vivo degradation profile. Comprehensive assessment of these parameters with enlarged sample volume would provide critical evidence to support the feasibility and safety of this approach in clinical settings.

In summary, we present a multifunctional, translationally relevant hydrogel-based platform that harnesses the reparative functions of NSCs for diabetic wound therapy, underscoring the potential of integrating stem cell biology with biomaterial design to address the complex pathophysiology of chronic wounds.

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki and was approved by the ethics committee of the Fourth Military Medical University (Approval No. 20260004).

CRediT authorship contribution statement

Junying Song: Writing - original draft, Methodology, Investigation. Yifu Zhu: Investigation, Visualization. Yufei Zhang: Methodology. Zhou Yu: Writing - review & editing, Supervision. Baoqiang Song: Writing – review & editing, Methodology, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article.