Engineering functional vasculatures to reconstruct sympathetic-parasympathetic circuits for bladder function after spinal cord injury

Haiyan Weng; Longyou Xiao; Ting Li; Hengrui Chang; Pengfei Xie; Liyan Zhu; Baobao Zhang; Shiqin Lv; Zijun Hu; Yu Dai; Pui Man Hoi; Wanguo Liu; Jialin Liu; Limin Rong; Yi Ren; Liumin He

doi:10.1016/j.bioactmat.2026.07.046

. 2026 Jul 31;67:59–79. doi: 10.1016/j.bioactmat.2026.07.046

Engineering functional vasculatures to reconstruct sympathetic-parasympathetic circuits for bladder function after spinal cord injury

Haiyan Weng ^a,¹, Longyou Xiao ^a,¹, Ting Li ^a,¹, Hengrui Chang ^c,¹, Pengfei Xie ^a, Liyan Zhu ^a, Baobao Zhang ^a, Shiqin Lv ^a, Zijun Hu ^d, Yu Dai ^a, Pui Man Hoi ^e, Wanguo Liu ^f, Jialin Liu ^a, Limin Rong ^a,^⁎, Yi Ren ^g,^h,^⁎⁎, Liumin He ^a,^b,^⁎⁎⁎

PMCID: PMC13452385 PMID: 42571573

Abstract

Spinal cord injury (SCI) leads to severe neurogenic bladder dysfunction, and its recovery highly depends on the restoration of bladder innervation of sympathetic-parasympathetic nerves. Stable blood perfusion is a prerequisite for sustaining neural regeneration, yet functional microvasculature rarely forms within injured spinal tissue. In this study, we engineered an integrating hydrogel-based spinal cord substitute to support vascularization through temporally staged angiogenic signaling combined with immune modulation. Early delivery of vascular endothelial growth factor (VEGF) promoted endothelial activation and sprouting, whereas delayed release of platelet-derived growth factor-bb (PDGFbb) supported pericyte recruitment and vessel maturation. Sustained immune modulation further stabilized nascent vasculature within the implanted segment. The spinal cord substitute supported the formation of structurally mature, non-leaky, and functionally perfused microvascular networks in vivo, accompanied by improved metabolic capacity within the implanted tissue. This vascularized environment enabled autonomic axonal regeneration, myelination, and synaptic organization, leading to re-engagement of bladder-related neural circuits and significant recovery of bladder function after SCI. These findings demonstrate that engineering functional vasculature within an implanted spinal cord substitute is sufficient to support tissue integration and selective autonomic recovery, establishing vascular reconstruction as a critical foundation for restoring complex neural functions after SCI.

Keywords: Spinal cord injury, Neural regeneration, Functional vascular networks, Sympathetic-parasympathetic circuits, Bladder function recovery

Graphical abstract

Highlights

•
Functional vascularization enables autonomic circuit reconstruction after SCI.
•
Temporal VEGF/PDGFbb signaling drives mature and perfused vessel formation.
•
Vascular maturation sustains metabolic integration and neural regeneration.
•
Rebuilt sympathetic-parasympathetic circuits restore bladder function after SCI.

1. Introduction

Spinal cord injury (SCI) causes a tissue environment that is profoundly unfavorable for repair. Vascular disruption, inflammatory infiltration, and metabolic insufficiency jointly compromise tissue integrity and constrain neural survival. Loss of stable blood supply results in sustained hypoxia and nutrient deprivation, conditions that limit axonal integration even when pro-regenerative signals are present. As a result, functional recovery after SCI remains poor [[1], [2], [3]].

Attempts to improve repair have largely focused on stimulating axonal regeneration through delivery of growth factors, cytokines, or biomaterials. Although these approaches have produced encouraging experimental outcomes, their efficacy remains limited in vivo [[3], [4], [5], [6]]. One major obstacle is the failure to restore functional vasculature. Angiogenic stimulation can increase vessel density, yet newly formed vessels are frequently immature, poorly perfused, and highly permeable. Such vessels fail to support metabolic homeostasis and can aggravate inflammation through immune cell leakage, reinforcing a hostile microenvironment that undermines tissue integration [7,8].

These considerations suggest that vascular regeneration in SCI is not limited by the availability of angiogenic factors alone, but by the absence of a permissive structural and immunological context in which functional vessels can form [14]. A spinal cord substitute offers a means to define such a context. By occupying a defined segment within the injured cord, an implanted substitute creates a space where vascular growth, immune responses, and tissue integration can be programmed and evaluated [[15], [16], [17]].

In our present study, we developed an integrating hydrogel-based spinal cord substitute designed to support functional vascularization through temporally staged angiogenic signaling combined with immune modulation. The platform incorporates early delivery of vascular endothelial growth factor (VEGF) to initiate endothelial activation and sprouting, followed by delayed release of platelet-derived growth factor BB (PDGFbb) to promote mural cell recruitment and vessel maturation. Curcumin was included to limit excessive inflammation and stabilize the immune environment during vascular morphogenesis. This strategy tests whether an implanted spinal cord substitute can itself support the formation of stable, perfused microvasculature in vivo. Using a complete transection model as a stringent test, we examined whether functional vascularization within the substitute is sufficient to restore metabolic support, permit axonal entry and synaptic organization, and enable recovery of autonomic circuit function. Our findings establish functional vascularization as a central design requirement for spinal cord substitutes and provide a framework for engineering neurovascular integration after injury.

2. Results

2.1. Fabrication of a biomimetic IntHydrogel platform for coordinated angiogenic stimulation

To enable temporally controlled angiogenic signaling, we engineered an IntHydrogel-based sequential delivery platform (IntHydrogel platform). VEGF was incorporated directly into the hydrogel to support early endothelial activation. PDGFbb was encapsulated within liposomes (Lip@PDGFbb) to delay its release and preferentially support vessel maturation. Curcumin was included as an anti-inflammatory component to explore whether coordinated regulation of angiogenesis and inflammation could yield enhanced therapeutic effects rather than being considered a structural component of the angiogenic system itself (Fig. 1A).

Fig. 1 — **Design and characterization of an IntHydrogel-based sequential release platform for coordinated angiogenic stimulation**. (A) Schematic illustration of the IntHydrogel platform loaded with VEGF, PDGFbb and curcumin (IntHydrogel + VEGF + Lip@PDGFbb + Curcumin), showing the strategy for sequential release to coordinate angiogenesis. Panel A was created with BioRender. (B) Size distribution profiles of blank liposome and Lip@PDGFbb. (C) TEM images of blank liposome and Lip@PDGFbb. (D) SEM images of blank IntHydrogel and the drug-loaded hydrogel (IntHydrogel + VEGF + Lip@PDGFbb + Curcumin), illustrating homogeneous distribution of Lip@PDGFbb nanoparticles along the hydrogel pore walls without apparent aggregation. The upper-right panel shows a zoom-in of the area marked in the full-size left panel. (E) Rheological properties of blank IntHydrogel and the drug-loaded hydrogel measured at angular frequencies ranging from 0.1 to 10 Hz, indicating that incorporation of bioactive cargos does not compromise the viscoelastic properties required to match native spinal cord tissue mechanics. (F) Cumulative *in vitro* release profiles of Curcumin, VEGF, and PDGFbb from the IntHydrogel, demonstrating an early release of curcumin and VEGF followed by delayed PDGFbb delivery, thereby achieving sequential angiogenic modulation.

The IntHydrogel matrix is composed of hyaluronic acid-graft-dopamine (HADA) and the designer peptide HGF-(RADA)₄-DGDRGDS (HRR), forming a three-dimensional porous network suitable for cargo incorporation and implantation into SCI lesion. These materials were selected based on our previous work demonstrating its close mechanical compatibility with native spinal cord tissue, self-healing capacity, strong tissue adhesion [5], and favorable blood compatibility [18]. Mechanistically, HADA serves as the primary structural scaffold and mechanical regulator of the IntHydrogel system. Its dopamine moieties enable oxidative crosslinking to form a stable network and also confer tissue-adhesive properties that anchor the hydrogel at the lesion site. The mechanical stiffness is tuned by HADA concentration and crosslinking density to match native spinal cord tissue. Notably, HADA does not directly control drug release kinetics. The staged release of VEGF and PDGFbb is achieved through distinct loading strategies, with HADA providing the framework that accommodates these delivery systems. These properties are critical for maintaining structural stability for infiltrating cells, angiogenesis and axonogenesis after implantation.

Building on these validated features, we next examined whether the IntHydrogel matrix could incorporate liposomal carriers without compromising its structural or mechanical integrity. Dynamic light scattering assay showed that PDGFbb encapsulation slightly increased liposome diameter from 112.1 ± 3.7 nm to 130.5 ± 6.5 nm, indicating successful loading while preserving nanoparticle uniformity (Fig. 1B–S1A). Transmission electron microscopy (TEM) confirmed that both empty liposomes and Lip@PDGFbb retained a well-defined core-shell morphology (Fig. 1C). Scanning electron microscopy (SEM) revealed that the IntHydrogel platform possessed an interconnected porous architecture with smooth pore walls. Lip@PDGFbb nanoparticles were uniformly distributed along the pore surfaces of curcumin-VEGF-Lip@PDGFbb IntHydrogel without detectable aggregation, confirming that IntHydrogel platform provides a stable microenvironment for liposomal incorporation (Fig. 1D).

Mechanical compatibility is essential for vascular and neural repair after SCI. Rheological analysis showed that the storage modulus of factor-loaded IntHydrogel formulations remained comparable to the blank hydrogel, indicating minimal impact that of VEGF, curcumin and Lip@PDGFbb on the mechanical properties (Fig. 1E). The stiffness remained within the range of native spinal cord tissue. In vitro degradation assays also showed no significant differences between loaded and unloaded hydrogels (Fig. S1B).

The encapsulation efficiency of PDGFbb within liposomes reached 71.3 ± 4.2%. The release kinetics analysis revealed a distinct temporal release pattern. VEGF and curcumin were released rapidly, while PDGFbb exhibited a delayed release from Lip@PDGFbb (Fig. 1F). This staged release behavior confirms that the IntHydrogel platform enables programmed delivery of angiogenic cues, allowing angiogenic initiation and maturation to be regulated as distinct biological processes.

Overall, these results establish the IntHydrogel platform as a structurally stable, mechanically compatible, and capable of temporally controllable cargo delivery. This IntHydrogel platform provides the necessary foundation for investigating how staged angiogenic signaling, together with immune modulation, shapes inflammation, neurovascular reconstruction after SCI.

2.2. Implanted spinal cord substitute (ICS) with the IntHydrogel platform defines a standardized substitute segment and shapes the immune microenvironment after SCI

Excessive and persistent inflammation is a critical barrier to spinal cord repair, as sustained pro-inflammatory signaling impedes axonal regeneration and exacerbates tissue damage [[19], [20], [21]]. To test how a spinal cord substitute behaves in vivo under controlled conditions, we used a segmental replacement SCI model involving complete transection and removal of a 2 mm spinal cord segment. In a complete thoracic transection model, we surgically established a stable and fully reproducible complete transection defect and immediately injected IntHydrogel formulations as a substitute segment. At the T10 level, an approximately 2 mm spinal cord segment corresponding to the lesion core was completely resected to create well-defined gap. Within the same surgical session, this gap was immediately filled with IntHydrogel formulations, which rapidly bridged the two severed ends of the host spinal cord and restored structural continuity across the defect. The dura was then closed to confine the implanted spinal cord substitute (ICS) and prevent extrusion (Fig. 2A). This approach generated a spatially well-defined substitute segment, enabling consistent analysis of immune, vascular and neural responses within the ICS.

Fig. 2 — **The implanted spinal cord substitute shapes a reparative immune microenvironment after SCI**. (A) Schematic of the experimental design. A 2 mm segment of rat spinal cord was transected and removed at T10 level, and the IntHydrogel were grafted immediately after SCI. The implanted segment was collected for analysis at 2 wpi. Panel A was created with BioRender. (B) Representative immunofluorescence images of CD206 and CD68 in the substitute segment at 2 wpi in the Control and the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin groups. (C) Quantification of the ratio of CD206/CD68 immunoreactivity in substitute segment at 2 wpi. (D) Representative immunofluorescence images of iNOS and CD68 in the substitute segment at 2 wpi in the Control and the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin groups. (E) Quantification of the ratio of iNOS/CD68 immunoreactivity in the substitute segment at 2 wpi. (F) Representative array map showing differential cytokine expression profiles at 2 wpi in the Control group and the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group. (G) Heatmap of relative cytokine normalized expression levels in different groups, illustrating a coordinated shift toward an inflammation-resolving and pro-angiogenic immune milieu in the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group. Color scale indicates expression intensity from low (blue) to high (red).

Because the substitute segment represents the primary site of biomaterial-host interaction, tissue collections for molecular and immunological analyses was focused on the ICS, included IntHydrogel together with infiltrated cells and newly formed tissue within it. Animals received IntHydrogel alone or IntHydrogel loaded with VEGF, PDGFbb (via Lip@PDGFbb), sequential VEGF plus PDGFbb, or sequential VEGF plus PDGFbb combined with curcumin, as indicated in Fig. 2A. Tissues were collected at 2 weeks post-injury (wpi), corresponding to the peak phase of macrophage-mediated inflammation.

Analysis of immune cell composition and signaling revealed coordinated immune modulation within the substitute segment. Immunostaining showed distinct macrophage polarization across treatment formulations. Compared with untreated animals, all IntHydrogel-based treatments exhibited a significant increase in CD206⁺ macrophage density and higher CD206⁺/CD68⁺ ratio, together with reduced iNOS⁺ macrophage density and iNOS⁺/CD68⁺ ratios. These changes indicate a shift toward an anti-inflammatory macrophage polarization within the ICS (Fig. 2B–E). Among all treatments, the most pronounced polarization was observed when sequential angiogenic signaling was combined with curcumin, exceeding that IntHydrogel delivering VEGF or PDGFbb alone, or by sequential release of VEGF plus PDGFbb without curcumin (Fig. S2). These changes in macrophages polarization markers were further corroborated by Western blot analysis of same ICS which showed reduced CD68 and iNOS expression together with increased CD206 levels (Fig. S3A–D). Moreover, qRT-PCR profiling revealed downregulation of pro-inflammatory mediators (iNOS, IL-3, TNF-α, IL-12) and a significant upregulation of anti-inflammatory markers (IL-4, IL-10, Arg1, TGF-β) across treatment groups compared with control (Fig. S3E–F). Expression of IL-1β, IL-6, IL-13 and IL-1ra did not differ significantly among treatment conditions (data not shown).

We next profiled cytokine and chemokine secretion within the ICS at 2 wpi (Fig. 2F–S3G). ELISA analysis confirmed that anti-inflammatory cytokines such as IL-4, IL-10, and IL-1ra were significantly increased, whereas pro-inflammatory mediators including IL-3 and TCK-3 were suppressed following IntHydrogel-based treatments (Fig. S3H). In addition, several chemokines associated with immune cell recruitment and vascular regulation, including CINC-1, CINC-2, CINC-3, and CCL11, were markedly up regulated (Fig. S3I–L). Heatmap analysis confirmed a coordinated shift in the cytokine landscape toward an inflammation resolution and tissue healing profile (Fig. 2G). These secreted signals established a paracrine environment through which macrophage remodeling can directly influence endothelial behavior and host vessel growth.

Notably, delivery of VEGF or PDGFbb alone each exerted anti-inflammatory effects. However, combined delivery of VEGF or PDGFbb did not yield an additive or synergistic anti-inflammatory benefit. Curcumin was therefore specifically incorporated into to the VEGF + PDGFbb formulation to evaluate its synergistic contribution to inflammatory modulation in the context of sequential agniogenic stimulation. This combination consistently produced the strongest shift toward a reparative immune state.

These findings indicate that IntHydrogel substitute segment supports macrophage remodeling into a functional paracrine immune environment. This immune conditioning provides the signaling context necessary for subsequent endothelial activation and host-derived vessel ingrowth, linking inflammation resolution to downstream vascular repair.

2.3. IntHydrogel substitute segment recruits host vessels and promotes early vascular maturation after SCI

SCI causes profound vascular collapse and disruption of the neurovascular unit (NVU), resulting in a fragmented microarchitecture and insufficient metabolic support that constrains axonal regeneration and neuronal survival [22]. Having established that IntHydrogel implant reshapes the early immune environment within a defined spinal cord defect, we next examined whether this newly reconstructed tissue niche could support effective vascular regeneration during the early stage following SCI. At the time of injection, the IntHydrogel segment occupied a surgically defined defect and was completely avascular. We therefore reasoned that sustained local release of VEGF and PDGFbb within this avascular segment would establish a spatially confined angiogenic gradient, thereby actively recruiting host-derived blood vessels from adjacent spinal cord into the IntHydrogel-filled defect.

To assess early vascular responses, analyses were performed at 2 wpi. Quantification focused on the interface between host spinal cord tissue and the ICS. This interface was anatomically delineated by GFAP-labeled reactive astrocytic boundaries. All measurements were restricted to the host spinal cord side to avoid potential interference from the IntHydrogel material itself.

In untreated control animals, Glut1⁺ vessels were largely confined to the host tissue cord border and accumulated near the GFAP-defined boundary. These vessels failed to extend beyond the astrocytic boundary into the defect region, indicating the absence of effective vascular ingrowth. In contrast, animals receiving IntHydrogel formulations containing angiogenic factors exhibited robust increase in both the density and length of Glut1⁺ vessels at the host-implant interface. Importantly, these vessels extended directionally from the host spinal cord toward the ICS (Fig. 3A). This pattern indicates active recruitment and ingrowth of host-derived vasculature into the ICS rather than passive peripheral accumulation, providing a valid starting point for a good vascular organization. Among all treatment groups, the IntHydrogel formulations delivering sequential VEGF and PDGFbb together with curcumin showed the most pronounced vascular response. Noteworthy, astrocytes were found to migrate into the ICS in continual contact with newly formed vessels, indicating the formation of NVU. Quantitative analysis revealed a significant increase in both Glut1⁺ vessel number and average vessel length compared with untreated control animals and all single-factor treatment groups (Fig. 3C and D).

Fig. 3 — **The IntHydrogel-based substitute drives host vessel in growth and facilitates early vascular maturation following SCI**. (A) Representative images of longitudinal spinal cord sections at 2 wpi in different groups. Boxed border regions were magnified with GFAP and Glut1 immunoreactivity. The white dotted lines indicated the border between host spinal cord and injury site. (B) Representative images of VEGFR2 (green), Glut1 (red), and DAPI (blue) immunofluorescence at the injury border and within the spinal cord at 2 wpi in different groups. (C) Quantification of the number of Glut1⁺ blood vessels at the injury border (n = 3 per group). (D) Average length of Glut1⁺ blood vessels at the injury border (μm). (E) Quantification of VEGFR2 and Glut1 co-expression intensity at the injury border (%). (F) Representative immunostaining images showing the expression of NF200, Glut1 and GFAP along the longitudinal sections of spinal cord at 2 wpi. The boxed regions were shown at higher magnification to display the NVU structure. White dashed lines indicate the injury boundary. (G) Representative Western blot and (H, I) quantitative densitometry results of VEGFR2 and Glut1 protein expressions in the substitute segment at 2 wpi from different groups (n = 3 per group). (J) High-resolution images with Occludin, Glut1, and DAPI in the Control group and the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group. (K) TEM image of vasculature in the Control group and the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group at 2 wpi. Right panel: Zoom-in view of the rectangular area indicated in the left panel. EC: Endothelial cell lining the vascular lumen. PC: Pericyte processes enveloping the abluminal surface. L: vascular lumen. BL: basement membrane. J: Tight junctions between endothelial cells (arrowheads). (L) Thirty minutes after tail vein injection of 4 kDa FITC-dextran, spinal cord sections were analyzed by immunofluorescence in the Control group and the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group at 2 wpi. Blood vessel leakage at the injury site was evaluated, with Glut1 marking endothelial cells and DAPI for nuclear counterstaining. FITC-dextran extravasation was used to assess vascular permeability at the injury site. (M) Quantitative analysis was performed on values for FITC-dextran mean intensity in different groups. Data were shown as the mean ± SD; P values were determined by two-tailed one-way ANOVA with Tukey's multiple-comparisons test.

Endothelial activation was further examined by examining VEGFR2 expression. Glut1⁺ vessels adjacent to the ICS delivering VEGF, PDGFbb and curcumin exhibited the high level of VEGFR2 co-expression, consistent with enhanced endothelial tip-cell activation and angiogenic sprouting [23,24]. VEGFR2⁺ vessels were less prominent in formulations delivering VEGF alone or PDGFbb alone, and intermediate in the sequential VEGF + PDGFbb group lacking curcumin (Fig. 3B–E). Western blot analyses of the tissue harvested from the host-implant interface confirmed elevated expression of VEGFR2 (4.54-fold vs control) and Glut1 (6.48-fold vs control) in the curcumin-containing IntHydrogel group (Fig. 3G–I). In comparison, curcumin alone induced only a modest increase in these angiogenic markers, indicating that anti-inflammatory modulation by itself was insufficient to drive robust angiogenic activation (Fig. S5).

Beyond endothelial activation, newly formed vessels must acquire barrier-related properties to function effectively. Consistent with this requirement, IntHydrogel-based treatments promoted centripetal extension of vascular structures from the host spinal cord into the implanted segment (Fig. 3F), indicating that vascular ingrowth progressed directionally into substitute tissue rather than remaining confined to the host border. Vessel remodeling was further supported by robust endothelial proliferation, as evidenced by the presence of BrdU-incorporating Glut1⁺ cells within regenerating vessels (Fig. S4A–B). In parallel, vessels forming within the ICS particularly in the curcumin-containing formulation, displayed key features of vascular maturation, including continuous occludin⁺ endothelial junctions and robust pericyte coverage, as confirmed by immunostaining and TEM (Fig. 3J–K, S4C-D).

Vessel barrier function was directly assessed using FITC-dextran leakage. Compared with control animals, vascular leakage was markedly reduced across all IntHydrogel treated groups and was minimal in animals receiving IntHydrogel delivering VEGF plus PDGFbb and curcumin. Quantification revealed an approximately 90.5% reduction in leakage relative to control (Fig. 3L–M, S4E). These findings indicate that the ICS, particularly when combined with curcumin, supports formation of vessels with improved barrier integrity.

Taken together, these results show that a spinal cord substitute formed by IntHydrogel implantation establishes a defined tissue niche that supports immune conditioning and enables host vessel recruitment and early vascular maturation within a defined defect. Sequential release of VEGF and PDGFbb is sufficient to initiate robust vascular ingrowth, whereas the addition of curcumin further stabilized nascent vessels and markedly improved barrier integrity. The formation of functional, non-leaky vasculature within the ICS therefore provides foundation for subsequent tissue remodeling and regeneration.

2.4. Immune conditioning within the ICS instructs NRP1-mediated endothelial activation and vascular growth

To explore how immune remodeling within the ICS influences vascular regeneration, we examined changes in pro-angiogenic mediators at the early post-implantation stage. ELISA analysis revealed that treatment with the IntHydrogel platform markedly increased the expression of multiple pro-angiogenic factors, including PDGF-AA, VEGF, HGF, Gas1, JAM-A within the defect region. Among these factors, Neuropilin-1 (NRP1) showed the most pronounced induction, with expression increased by nearly twentyfold compared with control tissue (Fig. 4A). This coordinated increase suggests that immune conditioning within the ICS is coupled to activation of an endothelial angiogenic program. Immunostaining revealed strong NRP1 expression within Glut1⁺ regenerating vessels located at the host-implant interface and extending into the ICS (Fig. 4B). NRP1 localization was largely restricted to endothelial structures, identifying that endothelial NRP1 may serve as a potential molecular link between immune-derived cues and host vessel growth within the substitute tissue.

Fig. 4 — **Immune conditioning within the ICS promotes vascular regeneration via NRP1-mediated endothelial activation**. (A) Quantitative analysis of significantly altered cytokines measured by ELISA (n = 4 per group), including pro-angiogenic cytokines associated with vascular repair. (B) Representative images of longitudinal spinal cord sections at 2 wpi in the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group, showing robust co-localization of NRP1 with Glut1⁺ regenerating endothelial structures at the substitute segment, indicating endothelial enrichment of NRP1 during vascular regeneration. Boxed border regions were magnified with Glut1 and NRP1 immunoreactivity. (C) Schematic illustrating the workflow for the macrophage-BMEC co-culture system used to investigate the direct regulation of endothelial NRP1 by macrophages. (D) Representative immunofluorescence images showing Phalloidin-labeled cytoskeleton and NRP1 expression in BMECs co-cultured with differentially polarized macrophages, demonstrating selective induction of endothelial NRP1 by anti-inflammatory macrophages. (E) Quantification of NRP1 fluorescence intensity in BMECs under different co-culture conditions, normalized to the control group. (F) Representative Western blot and (G) quantitative densitometry results of NRP1 protein expressions in BMECs co-cultured with differentially polarized macrophages (n = 3 per group). (H) Representative immunofluorescence images of VEGFR2 expression in BMECs following co-culture with polarized macrophages, indicating activation of pro-angiogenic signaling downstream of immune modulation. (I) Quantitative analysis of vascular branch points under different treatment conditions in BMECs co-cultured with differentially polarized macrophages. (J) Quantitative analysis of VEGFR2 fluorescence intensity in BMECs co-cultured with differentially polarized macrophages. (K) Representative immunofluorescence images showing VEGFR2 expression in BMECs co-cultured with anti-inflammatory macrophages with or without NRP1 knockdown, assessing the requirement of endothelial NRP1 for macrophage-induced angiogenic activation. (L) Quantitative analysis of vascular branch points in BMECs co-cultured with anti-inflammatory macrophages with or without siNRP1 transfection. (M) Quantitative analysis of VEGFR2 fluorescence intensity in BMECs co-cultured with anti-inflammatory macrophages with/without siNRP1 transfection. Data were shown as the mean ± SD; P values were determined by two-tailed one-way ANOVA with Tukey's multiple-comparisons test.

To determine whether macrophages directly regulate endothelial NRP1 expression, we employed a macrophage-brain microvascular endothelial cells (BMECs) co-culture system (Fig. 4C). Macrophages were polarized toward an anti-inflammatory phenotype by IL-4 plus IL-10 prior to co-culture. Exposure to anti-inflammatory macrophages significantly increased NRP1 expression in BMECs, as confirmed by immunostaining and immunoblotting (Fig. 4D–G). This increase was accompanied by an elevated VEGFR2 expression (Fig. 4H–J), consistent with activation of the NRP1-VEGFR2 angiogenic signaling pathway. Furthermore, anti-inflammatory macrophage stimulation significantly promoted vascular branching (Fig. 4I), indicating that macrophage-derived signals can directly enhance endothelial angiogenic behavior.

To determine whether NRP1 is required for transducing macrophage-derived signals, NRP1 expression was selectively knockdown in BMECs using siRNA (Fig. S6). Knockdown of NRP1 markedly attenuated macrophage-induced VEGFR2 upregulation and abolished the enhancement in vascular branch formation (Fig. 4K–M). These findings demonstrated that endothelial NRP1 is required for converting macrophage-derived reparative signals into functional angiogenic responses.

Therefore, these results identified NRP1 as a key molecule linking immune conditioning within the ICS. By converting macrophage phenotype toward a reparative phenotype and activating NRP1-meidated VEGFR2 signaling in endothelial cells, the IntHydrogel platform established an immune-vascular regulatory axis that supports robust and functional vascular growth within the implanted substitute segment.

2.5. ICS supports long-term vascular maturation and functional perfusion after SCI

For newly formed vessels to provide sustained metabolic and trophic support, they must undergo proper maturation and integrate functionally with surrounding tissue [25,26]. Having observed robust host vessel recruitment into the ICS at early stages, we next examined whether these vessels could achieve long-term stability and functional perfusion within the spinal cord substitute. At 8 wpi, untreated animals in which the defect was filled with hydrogel but lacked active formulations showed minimal vascularization within the ICS. Only sparse and fragmented Glut1⁺ endothelial structures were detected near the host-lesion border, indicating a failure of spontaneous revascularization within the substitute tissue.

Delivery of VEGF or PDGFbb alone induced limited vascular ingrowth. Glut1⁺ vessels appeared discontinuous, poorly organized, and largely confined to peripheral regions of the ICS. By contrast, IntHydrogel platform delivering sequential angiogenic signaling together with curcumin stimulated robust microvascular formation throughout the substitute tissue. Dense and interconnected Glut1⁺ vascular network extended deeply into regions that were previously avascular, forming continuous intrasegmental networks rather than remaining restricted to host border (Fig. 5A and B, S7A). Quantitative analysis confirmed a significant increase in endothelial density and vessel caliber compared with formulations delivering VEGF alone, PDGFbb alone or sequential VEGF plus PDGFbb without curcumin, which aligned with upregulated Glut1 protein expression (Fig. 5C–D, S8A-B).

Fig. 5 — **ICS drives long-term vascular maturation and functional perfusion after SCI**. (A) Representative images of longitudinal spinal cord sections at 8wpi in the Control group and the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group. The treatment group showed extensive vascular ingrowth into the lesion core, whereas Control lesions remained largely avascular. Boxed border regions were magnified with GFAP and Glut1 immunoreactivity. The white dotted lines indicated the border between host spinal cord and the injury site. (B) High-resolution images of Glut1 and DAPI at 8wpi in the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group. (C) Quantification of Glut1 expression intensity in the substitute segment (%). (D) Average diameter of vessels in the substitute segment (μm). (E) High-resolution images with VE-cadherin, Glut1 and DAPI at 8wpi in the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group. Continuous VE-cadherin junctions indicate stabilized endothelial cell-cell adhesion. The image on the right is a 3D reconstruction. (F, G) Representative immunostaining images of α-SMA (F) and PDGFRβ (G) co-labeled with Glut1 at 8 wpi in the substitute segment of the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group, demonstrating robust recruitment of pericytes surrounding regenerated vessels. (H) Western blot analysis of α-SMA and PDGFRb protein levels in the substitute segment of different groups. (I, J) Quantification of α-SMA (I) and PDGFRb (J) protein expression (n = 3 per group), confirming enhanced perivascular coverage indicative of late-stage vascular maturation. (K) TEM image of vasculature at 8wpi in the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group. Ultrastructural analysis reveals continuous endothelial linings, closely apposed pericytes, organized basement membranes, and distinct intercellular junctions, consistent with near-native vascular architecture. Right panel: Zoom-in view of the rectangular area indicated in the left panel. EC: Endothelial cell lining the vascular lumen. PC: Pericyte processes enveloping the abluminal surface. L: vascular lumen BL: basement membrane. (L) Ultrasound localization microscopy imaging results of spinal cord injury area at 8wpi in the Control group and the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group. The Control group displayed sparse, fragmented perfusion, whereas IntHydrogel treatment restored dense, well-aligned microvascular networks with coherent flow patterns. Boxed border regions were magnified. Intensity: Intensity map of spinal cord. Direction: direction map of spinal cord. Velocity: velocity map of spinal cord. (M) Quantification of blood vessel density in the substitute segment (%) based on ULM. (N) Quantification of blood vessel velocity in the substitute segment (cm/s) based on ULM. Data were shown as the mean ± SD; P values were determined by two-tailed one-way ANOVA with Tukey's multiple-comparisons test.

Vessel maturation markers further indicated enhanced vessel stability within the ICS. VE-cadherin staining revealed continuous and well-organized endothelial junctions along newly formed Glut1⁺ vessels (Fig. 5E–S8C), indicating stabilized endothelial cell-cell adhesion. This was supported by increased VE-cadherin protein levels detected by Western Blot analysis (Fig. S8D–E). Furthermore, newly formed vessels exhibited markedly increased perivascular coverage, with numerous PDGFRβ⁺ and α-SMA⁺ mural cells closely associated with newly formed Glut1⁺ endothelium within the substitute tissue (Fig. 5F–J). TEM provided ultrastructural confirmation of vascular maturation, showing continuous endothelial linings, intact basement membranes, well-defined intercellular junctions, and closely associated perivascular cell processes (Fig. 5F–K).

To determine whether these structural features translated into functional perfusion, vascular flow was assessed using ultrasound localization microscopy. SCI caused severe perfusion failure, characterized by sparse and fragmented microvasculature, low vessel density (4.59 ± 0.59%), and a markedly reduced blood vessel velocity (0.34 ± 0.05 cm/s). Delivery of VEGF or PDGFbb alone induced only modest improvements in vascular density (6.54 ± 0.76% with VEGF and 7.70 ± 0.82% with PDGFbb, respectively) and minimal elevations in blood vessel velocity (0.35 ± 0.05 cm/s and 0.43 ± 0.01 cm/s). Sequential delivery of VEGF and PDGFbb further increased vessel density to 12.91 ± 1.16% and flow velocity to 0.51 ± 0.02 cm/s. However, perfusion within the ICS remained heterogeneous and partially disorganized. In contrast, IntHydrogel platform delivered VEGF and PDGFbb together with curcumin generated dense and well-aligned microvascular networks with coherent perfusion throughout the substitute tissue. Quantitative analysis revealed a microvascular density of 23.71 ± 3.71% and a flow velocity of 0.63 ± 0.03 cm/s, corresponding to a 1.83-fold increase in flow velocity compared with untreated controls (P = 0.0002) and a significant improvement over all formulations lacking curcumin (Fig. 5L–N, S9).

Therefore, these observations demonstrate that angiogenic signaling is sufficient to initiate vessel ingrowth into a spinal cord substitute, while immune conditioning is required to stabilize vascular maturation and establish sustained functional perfusion. This structurally and functionally perfused microvascular networks within the ICS provides a metabolically supportive microenvironment for neural regeneration.

2.6. Vascularization of the ICS enhances local metabolic support after SCI

Having established that vessels formed within the ICS are structurally mature and functionally perfused, we next asked whether this vascularization supports metabolic activity within the substitute segment. To address this question, we conducted untargeted metabolomics analysis during the early post-injury phase. Metabolomic profiling was performed on tissue harvested from the ICS after 2 wpi. We focused on comparisons between untreated control animals and animals receiving IntHydrogel sequential angiogenic cues with curcumin, as this formulation exhibited the most robust vascular ingrowth, the highest degree of vascular maturation, and lowest vascular permeability among all conditions. This design allowed us to assess whether effective vascular formation and functional perfusion of the ICS was associated with coordinated metabolic activity.

Principal component analysis revealed a clear separation between control and ICS in both positive and negative ion modes, indicating that a distinct metabolic state within the ICS following IntHydrogel treatment (Fig. 6A). Volcano plot analysis showed that most significantly altered metabolites were increased in the IntHydrogel-treated group, rather than bidirectional distribution (Fig. 6B). These changes spanned multiple biochemical classes, including amino acids, nucleotides, coenzymes, phenolic acids, and flavonoids, consistent with a global elevation in biosynthetic and energy-related metabolic processes within the ICS (Fig. 6C).

Fig. 6 — **Vascularization of the ICS enhances local metabolic support after SCI**. (A) PCA plot showed clustering of substantial disparities in metabolism across the Control group and the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group, indicating a global metabolic shift induced by the treatment following SCI. (B) Volcano plots displayed the differential intensity of metabolites between the Control group and the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group. (C) Bubble charts illustrated the variations in metabolites between the Control group and the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group, indicating enhanced local biosynthetic and energetic metabolism in the spinal cord. (D) KEGG enrichment analysis of significantly differential metabolites identified between the Control group and the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group, with bubble size representing pathway impact and color indicating statistical significance. (E) Clustering heatmap highlighted expression differences of metabolites between the Control group and the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group. (F-I) Box plots showed the relative intensity levels of significantly altered metabolites, including citric acid (F), oxalacetic acid (G), isocitric acid (H) and ADP (I), reflecting enhanced tricarboxylic acid (TCA) cycle activity and improved cellular energy turnover in the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group. (J) KEGG enrichment analysis revealed significant enrichment of neural-relevant pathways in the IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group compared to the Control group, suggesting that metabolic reprogramming supports downstream neural repair and functional recovery.

Pathway enrichment analysis identified cellular energy metabolism and redox pathways as the most prominently affected. Metabolites involved in nucleotide metabolism, the tricarboxylic acid (TCA) cycle, taurine and hypotaurine metabolism, and fatty acid metabolism were significantly enriched in the ICS (Fig. 6D). A clustering heatmap highlighted coherent metabolic differences between the two groups, supporting coordinated metabolic enhancement rather than isolated metabolite alterations (Fig. 6E). Consistent with pathway analysis, several key mitochondrial intermediates, including citric acid, oxaloacetic acid, and isocitric acid, were significantly elevated in the ICS, accompanied by reduced levels of adenosine diphosphate (ADP) (Fig. 6F–I). This metabolite profile is consistent with enhanced TCA cycle activity and more efficient ATP turnover within the ICS.

Importantly, these metabolic features aligned with the presence of functionally perfused host-derived vasculature within the ICS, supporting the notion that newly formed vessels within the ICS improved local delivery of nutrients and oxygen. Beyond energy metabolism, KEGG enrichment analysis also revealed significant enrichment of neural-related metabolic pathways, including those related to neurotransmitter synthesis and synaptic function (Fig. 6J). Notably, these metabolites were measured from tissue homogenates of the implanted graft, suggesting a correlation between improved vascularization and local metabolic availability within the graft.

Taken together, host vessel recruitment into the ICS is accompanied by coordinated metabolic enhancement, characterized by increased mitochondrial energy metabolism and activation of neuro-supportive pathways. This vascular-metabolic coupling supports local circulatory and metabolic capacity, providing a permissive biochemical foundation for subsequent neural regeneration and functional recovery after SCI.

2.7. ICS supports axonal entry and synaptic formation after SCI

In a complete transection model, long-range descending motor pathways are largely refractory to regeneration [27,28]. In line with this constraint, our data showed that 5HT⁺ axons were observed to regenerate into the ICS but failed to extend into the caudal host (Fig. S10A). Host inputs into grafts were both locally and supraspinally derived. Although host supraspinal axons regenerated into grafts, in no case did they regenerate beyond the graft into host spinal cord parenchyma caudal to the lesion site. Like graft-derived axons, host axons penetrating grafts colocalized with the synaptic marker synaptophysin, establishing a mechanism for host-to-graft connectivity.

ICS intervention exhibited significant improvement on hindlimb locomotion after SCI determined by the Basso, Beattie, and Bresnahan (BBB) scaling. The BBB score reached 5.6 ± 1.4 (Fig. S10B–D), falling within a range similar to that reported in the studies by Lu et al. [29] and Yang et al. [30]. These studies involved implanting neural stem cells embedded in fibrin matrices containing growth factor cocktails and neurotrophin-3-coupled chitosan, respectively, following complete spinal cord transection. However, locomotor recovery remained limited and was often debated [31,32], reflecting the challenge of restoring long-distance motor connectivity after complete transection with a large gap.

Giving this limitation, we examined whether reconstruction of a vascularized, metabolically active niche could support axonal entry and local circuit features that are more sensitive to microenvironmental normalization. Functionally, although locomotor recovery remains an important outcome, behavioral improvement in the best-performing formulation was modest. We therefore focused on subsequent analyses on bladder dysfunction, one of the most urgent clinical consequences of SCI, whose neural control relies heavily on autonomic circuits that are particularly sensitive to local vascular, metabolic, and inflammatory conditions.

To assess neural responses within the ICS at later stages after injury, we analyzed tyrosine hydroxylase-positive (TH⁺) axons, which labels catecholaminergic fibers involved in autonomic regulation, including bladder control. At 8 wpi, TH⁺ axons were largely absent from the ICS in control animals, indicating minimal spontaneous axonal entry into the ICS. In contrast, animals receiving with IntHydrogel formulations delivering angiogenic cues promoted extensive ingrowth of TH⁺ axons across the host-replacement interface, with fibers extending into the ICS (Fig. 7A). Quantitative analysis confirmed a significant increase in TH⁺ axon number within the ICS. Delivery of VEGF or PDGFbb alone produced moderate increases, whereas their combined delivery showed a synergistic effect. Importantly, incorporation of curcumin into the sequential angiogenic formulation yielded the highest TH⁺ axon density within the ICS (Fig. 7B).

High-resolution imaging further demonstrated continuous myelin basic protein (MBP) labeling along TH⁺ axons within the ICS (Fig. 7C). TEM revealed the myelinated axonal structure at 8 wpi, providing ultrastructural evidence of successful myelination within the ICS (Fig. 7D). These structural observations were supported by Western blot analysis, which demonstrated a stepwise increase in TH and MBP protein levels from single-factor treatments to sequential angiogenic formulations, with the highest expression observed in the VEGF + PDGFbb + curcumin condition (Fig. 7E–G).

Confocal imaging revealed close spatial association between TH⁺ axons and Glut1⁺ vessels within the ICS, particularly in animals receiving the sequential angiogenic and anti-inflammatory formulation (Fig. 7H). This spatial coupling suggests that the newly formed vasculature provides both structural and metabolic support for axonal growth within the ICS.

Synaptic organization within the ICS was further examined. Presynaptic and postsynaptic markers, including synaptophysin (Syn), vesicular glutamate transporter (VGLUT), and vesicular GABA transporter (VGAT), showed numerous synaptic puncta throughout the ICS (Fig. 7I–K). These signals indicate the presence of both excitatory and inhibitory synaptic contacts. TEM confirmed the synapse formation within the ICS, revealing type I and type II synapses characterized by defined postsynaptic densities, presynaptic vesicles, and clear synaptic clefts (Fig. 7L). Consistent with this findings, Western blot analyses showed marked increases in Syn (6.7-fold vs Control) and postsynaptic density protein 95 (PSD95; 1.6-fold vs Control) in ICS treated with sequential angiogenic and immune-modulatory cues, exceeding the levels observed in single-factor or dual-factor conditions (Fig. 7M − O).

These data demonstrated that a vascularized and metabolically active ICS supports TH⁺ axonal entry, local myelination and synaptic organization at later stages after SCI.

2.8. The ICS enables re-engagement of supraspinal-spinal autonomic circuits and restoration of bladder function after SCI

Coordinated bladder storage and voiding depend on the integrated function of sympathetic, parasympathetic, and somatic components organized as a multisynaptic reflex circuit [33,34]. To assess whether neural integration within the ICS contributes to re-engagement of this system after SCI, we conducted pseudorabies virus (PRV)-mediated transsynaptic retrograde tracing. PRV-RFP was injected into bladder walls musculature following established paradigms for mapping bladder-related autonomic circuits after SCI (Fig. 8A). In aminals treated with Inthydrogel-based formulation delivering sequential angiogenic cues with immune modulation, abundant PRV⁺ neurons were detected in multiple cortical and subcortical regions implicated in bladder control (Fig. 8B), suggesting reactivation of supraspinal centers. Robust PRV labeling was also observed in thoracic spinal segments (Fig. 8C), consistent with the restoration of communication between spinal autonomic relays and higher-order control regions.

Within the C5-T1 spinal segments, a substantial fraction of PRV⁺ neurons co-localized with TH and Syn (Fig. 8D), indicating the involvement of catecholaminergic interneurons that are essential for sympathetic regulation of bladder outlet and storage function. High-resolution imaging further revealed numerous PRV⁺/Choline acetyltransferase⁺ (ChAT⁺) neurons, indicating re-engagement of parasympathetic circuits responsible for detrusor contraction during voiding (Fig. 8E). Furthermore, PRV-labeled neurons co-expressing parvalbumin (PV) were detected (Fig. 8F), identifying inhibitory interneurons that coordinate sphincter-detrusor coupling. Collectively, combined TH, ChAT, and PV labeling strategy allowed simultaneous evaluation of sympathetic, parasympathetic and inhibitory components, demonstrating reassembly of an anatomically and neurochemically coherent autonomic network.

To determine whether circuit-level reengagement contributed to functional recovery, bladder function was evaluated at 8 wpi. Untreated SCI animals exhibited markedly elevated residual urine volumes, reflecting severe neurogenic bladder dysfunction. Delivery of VEGF or PDGFbb alone produced only modest improvements in residual volume, whereas sequential delivery of VEGF and PDGFbb resulted in partial recovery. In contrast, animals receiving the IntHydrogel-based formulation incorporating sequential angiogenic cues with immune modulation showed a pronounced reduction in residual urine volume, corresponding to an approximate 72.6% decrease compared with control (Fig. 8G and H), indicating substantial improvement in voiding efficiency.

Histological analysis of bladder tissue further supported functional recovery. Hematoxylin and eosin staining revealed pronounced thinning and disorganization of bladder smooth muscle in control animals, whereas treatment preserved wall thickness and maintained well-organized smooth muscle bundles (Fig. 8I and J). Examination of bladder wall innervation showed sparse TH⁺ sympathetic fibers and weak botulinum toxin (BTX) labeling of neuromuscular junctions in controls (Fig. 8K). In contrast, bladders from treated animals exhibited dense TH⁺ fibers and robust BTX signals (Fig. 8L and M), demonstrating restoration of sympathetic innervation and neuromuscular junction integrity. ChAT/Syn co-staining showed abundant cholinergic synaptic structures within bladder wall (Fig. 8N–P), indicating the restoration of the parasympathetic drive necessary for coordinated detrusor contraction.

Collectively, these findings demonstrated that vascularized ICS using sequentially programmed formulation re-established a functionally integrated supraspinal-spinal-peripheral autonomic network, enabling coordinated bladder storage and voiding after SCI.

3. Discussion

Functional recovery after SCI remains limited, not because of a lack of growth factors or regenerative signals, but because injured tissue fails to regain stable vascular support. Disruption of the spinal microvasculature leads to persistent hypoxia, impaired nutrient delivery, and metabolic insufficiency. These deficits constrain tissue integration and neural repair, even when pro-regenerative cues are supplied. Addressing vascular function, rather than angiogenic signaling alone, remains a central challenge in spinal cord repair.

Angiogenesis is commonly pursued as a strategy to improve tissue perfusion after injury. Over past decades, our understanding of angiogenesis mechanisms has increased at an explosive rate. Meanwhile, a variety of interventions have been developed to construct vascular networks in lesion site for SCI repair, including the administration of proangiogenic factors, gene modulation, cell transplantation, and biomaterial implantation [[35], [36], [37], [38]]. However, new vessels induced in the injured spinal cord are frequently immature and unstable [38,39], with recurrent features of high permeability of the blood-spinal cord barrier, poor perfusion, and leakage of circulating immune cells [40,41]. Malfunctioning vasculatures fail to support sustained anatomic repair and instead prevent functional recovery [39,42]. These observations indicate that vascular growth alone is insufficient. Vessel structure, maturation, and functional integration are critical.

Angiogenesis is a multi-step process that requires not only endothelial proliferation and sprouting, but also subsequent vessel maturation involving pericyte recruitment, basement membrane assembly, and stabilization of endothelial junctions [43,44]. Accordingly, sequential induction of these physiological progresses is especially important. Nevertheless, most of research remains descriptive, without delving into the intricate detailed mechanistic insights that drive this process. In this study, we developed an IntHydrogel platform for staged angiogenic signaling in vivo. Using this system, we tested whether separating early endothelial activation from subsequential maturation improves vascular outcomes within an implanted spinal cord substitute. We found that host-derived vessels can be recruited into the implanted segment and mature into stable, perfused microvascular networks. When angiogenic cues were delivered without staging, non-staged delivery supported only limited or produced unstable vessels.

Importantly, immune conditioning emerged as a necessary component of this process, a role that was rarely documented in previous studies. Within the implanted segment, macrophage polarization shifted toward a reparative state that facilitates angiogenesis with a significantly higher vessel density. Moreover, this immune environment directly influenced endothelial behavior, favoring functional integration of blood-spinal cord barrier. Thus, the administration of signaling molecules for angiogenesis and inflammatory regulatory factors is essential to construct a functional vascular network. Endothelial NRP1 was strongly induced and functionally required for translating macrophage-derived signals into angiogenic activation and vascular branching. Disruption of NRP1 signaling abolished these effects. These data identify a direct immune-vascular coupling mechanism that governs vessel formation within the spinal cord substitute.

NVU formation requires coordinated integration of endothelial cells, pericyte and astrocytes [[45], [46], [47]]. Although the ICS robustly supported endothelial ingrowth and pericyte recruitment, astrocyte incorporation into the core of the implanted segment remained limited. This indicates that vascularization alone is insufficient for full NVU reconstruction. Astrocyte recruitment may be restricted by both insufficient integration time and the lack of astrocyte-specific guidance cues within the substitute matrix. These findings define a structural limitation of the current IntHydrogel platform and point to future directions, including astrocyte-targeted matrix engineering [48,49] and combinatorial strategies incorporating cell transplantation [50,51], to achieve more complete neurovascular unit assembly.

Vascularization within the implanted segment was accompanied by marked metabolic changes. Metabolomic profiling revealed enhanced mitochondrial energy metabolism, increased nucleotide turnover, and activation of neurotransmitter-associated pathways. Previous angiogenic approaches in SCI have reported limited metabolic benefits, which often reflect increased vessel density without effective tissue perfusion. Immature or leaky vessels frequently fail to deliver oxygen and nutrients to the parenchyma, constraining metabolic recovery [39,52]. In contrast, the metabolic changes observed in our study extended beyond the vascular lumen and were evident within surrounding tissue, indicating improved metabolic availability driven by functional perfusion rather than vessel density alone. Nevertheless, metabolic restoration was incomplete. The absence of fully reconstituted NVU, particularly limited astrocyte integration, likely restricted metabolic coupling. These findings suggest that while functional vascularization is necessary for metabolic recovery, additional strategies will be required to achieve full metabolic normalization.

At later stages, the vascularized implanted segment supported axonal entry, local myelination, and synaptic organization. Long-range corticospinal regeneration was not observed, consistent with the constraints of complete transection [53,54]. However, catecholaminergic axons entered the implanted segment, became myelinated, and formed structured synaptic contacts. Neural elements aligned closely with the vascular network, indicating that vascular support shaped the spatial organization of regenerating tissue.

These tissue-level changes translated into recovery of autonomic circuit function. Bladder control depends on coordinated supraspinal, spinal, and peripheral pathways [55,56] and is highly sensitive to local metabolic and vascular conditions [57,58]. Re-engagement of supraspinal bladder circuits, restoration of spinal relay neurons, and reinnervation of the bladder wall were observed following implantation of the vascularized spinal cord substitute. Functional recovery of voiding followed. This level of repair reflects system-level reorganization of supraspinal innervation of bladder through sympathetic, parasympathetic and autonomic nerves rather than local axonal growth alone.

Several limitations should be acknowledged. First, this study used a complete transection model with a 2 mm spinal cord segment removed in all animals, which represents an even less clinically relevant condition than simple transection and is far less frequent clinically than contusion injuries. The model was chosen to provide a stringent test of vascularization and tissue integration in the absence of spared connectivity, which eliminates the confounding influence of spontaneous functional recovery often seen in contusion or hemisection models due to residual neural tissue preservation. The pathological features such as vascular collapse, immune dysregulation, and metabolic failure are also present in other SCI models such as contusion injuries [59,60]. Whether partial implantation or targeted replacement can synergize with spared tissue in those settings remains to be determined. Second, significant locomotor recovery is not detected, consistent with accumulating evidence that long-range descending pathways are largely refractory to regeneration after complete transection [27,61]. Although several studies have shown that biomaterials combined with growth factors and stem cells induced locomotor recovery [29,30], the magnitude and reproducibility of such recovery remain debated [31,32]. Host supraspinal axons can regenerate into grafts implanted within the large gap; however, they never regenerated beyond the graft into the host spinal cord parenchyma caudal to the lesion site [29]. Required growth facilitators propelled axon regeneration across the crushed lesion with a small gap, but no detectable improvement of locomotor function was acquired [27]. Restoring long-distance motor connectivity likely requires additional mechanisms beyond vascular and metabolic reconstruction, which were not addressed in this study. Additionally, although we observed reduced bladder residual volumes and detrusor muscle recovery, we did not directly measure voiding efficiency or urodynamic parameters. Thus, we cannot exclude the possibility that the reduced residual volumes partly reflect increased detrusor spasticity. Another limitation is that the hydrogel was implanted immediately after injury, which does not reflect the clinical reality of delayed intervention. Thus, the effects of secondary injury events such as inflammation and glial scar formation were not fully captured in this study. Whether this strategy would be effective at delayed time points remains to be determined in future studies.

In summary, this work demonstrates that a spinal cord substitute can support robust in vivo vascularization, leading to the formation of stable and functional microvasculature. This vascularization is sufficient to restore metabolic support, enable tissue integration, and support selective neural and functional recovery. These findings establish vascular engineering as an essential component of spinal cord repair and define functional vascularization as a prerequisite for meaningful tissue reconstruction.

4. Conclusion

In the segmental replacement model used in this study, SCI creates a cavity devoid of native ECM within the resected gap, which prevents neural regeneration [59,62]. Severe neurogenic bladder impairments are caused, and, as a result, the life quality of SCI patients is impacted, both physically and psychologically. This severe injury leads to neurogenic bladder dysfunction, which substantially impacts the quality of life of SCI patients. Restoration of supraspinal innervation of the bladder is essential for recovering lower urinary tract function. However, neural regeneration is constrained by metabolic insufficiency resulting from extensive vascular disruption, highlighting the need to reconstruct functional microvascular networks to support circuit reorganization.

Despite recent advances in revascularization strategies, the formation of mature, functionally perfused vessels remains a major challenge, as newly formed vessels are often immature and leaky. To address this, we engineered an integrating hydrogel-based spinal cord substitute that delivers VEGF for early endothelial activation and sprouting, followed by delayed release of PDGFbb to promote pericyte recruitment and vessel maturation. This staged delivery system mimics the natural progression of vascularization. Sustained immune modulation further stabilizes nascent vasculature within the implanted segment. As a result, mature, perfused, pericyte-supported vascular networks were established in the SCI lesion, which recalibrated the inflammatory milieu, restored metabolic sufficiency, and created a stable nutritional microenvironment for axonal regeneration, remyelination, and synaptic reconstruction.

Crucially, the restored vasculature enabled the reassembly of multisynaptic autonomic pathways spanning the brain, spinal cord, and peripheral organs, contributing to recovery of bladder storage and voiding. Together, these findings establish functional vascularization as a promising strategy to reorganize complex sympathetic and parasympathetic circuits for SCI repair.

5. Methods

5.1. Materials

HA was obtained from Toronto Research Chemicals Inc. (Toronto, Canada). DA was purchased from J&K Scientific Ltd. (Beijing, China). 1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC) and N-hydroxysuccinimide (NHS) were supplied by Meryer Chemical Technology Co. Ltd. (Shanghai, China). The self-assembling peptide HRR was synthesized by Qiangyao Biotechnology Co. Ltd. (Hubei, China).

5.2. Synthesis and characterization of HADA

HADA was synthesized by conjugating DA to HA through a conventional EDC/NHS-mediated coupling reaction. In brief, 758.6 mg of HA (2 mmol) was dissolved in 120 ml of deionized water and stirred at 35°C under a nitrogen atmosphere. NHS (690.5 mg, 6 mmol) and EDC (1.1502 g, 6 mmol) were introduced into the solution and activated for 15 min while maintaining the pH at 5.0-5.5. Subsequently, 189.6 mg of DA (1 mmol) was added, and the mixture was allowed to react at 35°C in the dark for 24 h. The product was then dialyzed against deionized water for 3 days and finally lyophilized to obtain purified HADA.

5.3. Preparation of PDGFbb-loaded sustained-release liposomes (Lip@PDGFbb)

Lip@PDGFbb were prepared using a thin-film hydration method. Phosphatidylcholine and cholesterol were mixed at a molar ratio of 4:1, with 303.2 mg phosphatidylcholine (0.4 mmol) and 38.7 mg cholesterol (0.1 mmol) dissolved in 2 mL dichloromethane. The solution was transferred to a round-bottom flask and rotary-evaporated to form a uniform lipid film. After complete solvent removal, the film was hydrated with phosphate-buffered saline (PBS containing recombinant PDGFbb (Thermo, USA), followed by vigorous vortexing to obtain multilamellar liposomes. The suspension was then sonicated in an ice bath to reduce particle size and achieve uniformity. Free PDGFbb was removed by dialysis (100 kDa MWCO) against PBS, and the resulting Lip@PDGFbb suspension was stored at 4°C.

5.4. Preparation of hydrogels

For IntHydrogel, HRR peptide was first dissolved in deionized water to make a 0.1% (w/v) solution, then lyophilized HADA was added to reach 2% (w/v). The mixture was adjusted to pH 7 with 1 M Tris-HCl to induce hydrogel formation.

For IntHydrogel loaded with VEGF, Lip@PDGFbb, and curcumin (IntHydrogel + VEGF + Lip@PDGFbb + curcumin), curcumin was first dissolved in DMSO at 10 mM under stirring until completely dissolved, and then diluted to 100 μM in PBS, followed by filtration through a 0.22 μm syringe filter. The curcumin solution was then mixed with the IntHydrogel solution. At the same time, VEGF (100ug/mL, MCE, USA) and Lip@PDGFbb were added. PH of the final mixture was adjusted to 7.4 using 1 M Tris-HCl to promote hydrogel formation.

5.5. Transmission electron microscopy (TEM)

Liposome morphology was examined using transmission electron microscopy. A drop (10 μL) of the liposome suspension (diluted 1:10 with PBS) was placed onto a carbon-coated copper grid (300 mesh) and allowed to adsorb for 2 min at 25°C. Excess liquid was carefully removed with filter paper. The grid was then negatively stained with 10 μL of 2% (w/v) uranyl acetate solution for 1 min, and the excess stain was gently blotted off with filter paper. The grid was air-dried at 25°C for 10 min before imaging. Samples were examined under a transmission electron microscope (Philips CM 10, Philips, Eindhoven, The Netherlands) operating at 80 kV. Images were acquired to assess the size, morphology, and uniformity of the liposomes.

For spinal cord tissue, after anesthesia, rats underwent cardiac perfusion with a fixative solution containing equal volumes of 2% glutaraldehyde and 4% paraformaldehyde (PFA). A 2 mm segment of the spinal cord was carefully excised from the surgical site and immersed in the fixative solution at 4°C for 1 h. The harvested tissues were post-fixed in 1% osmium tetroxide for 1 h, followed by dehydration using a graded ethanol series. The tissues were then embedded in Epon resin. Ultra-thin sections (80 nm) were prepared using a Reichert ultramicrotome (Reichert E, Co, Vienna, Austria), stained accordingly, and examined under a Philips CM 10 transmission electron microscope (Philips, Eindhoven, The Netherlands).

5.6. Dynamic light scattering (DLS)

The size distribution and polydispersity index (PDI) of blank liposomes and Lip@PDGFbb were determined by dynamic light scattering (DLS) using a Zetasizer Nano ZS instrument (Malvern Panalytical, UK). Liposome samples were diluted 1:100 with filtered PBS (pH 7.4, 0.22 μm filter) to avoid multi-scattering effects. Measurements were performed at 25°C with a scattering angle of 173° (backscatter mode). Each sample was equilibrated for 2 min prior to measurement, and each measurement consisted of 12 runs.

5.7. Scanning electron microscopy (SEM)

The microstructure of the hydrogels was examined by SEM. Blank IntHydrogel and VEGF + Lip@PDGFbb-loaded hydrogels were cut into small pieces and fixed in 2.5% glutaraldehyde for 4 h at 4°C. Samples were dehydrated through a graded ethanol series (30%-100%), followed by critical-point drying. Dried specimens were mounted on stubs and sputter-coated with a thin gold layer. SEM imaging was performed using a field-emission microscope (Zeiss GeminiSEM 300) at 5-10 kV. Representative images were acquired at low and high magnifications to visualize overall pore morphology and embedded nanoscale particles. Pore size and particle distribution were qualitatively assessed based on images.

5.8. Rheological tests

The mechanical properties of the hydrogels were assessed using a Kinexus Pro rheometer (Malvern Instruments) equipped with a 10-mm parallel plate at 25°C. Approximately 300 μL of hydrogel was loaded between the plates with a fixed gap of 0.3 mm. Frequency sweep measurements were then conducted, beginning at 0.1 Hz and progressing to 10 Hz, under a constant shear strain of 0.5%. The data acquisition density was set to 10 points per frequency decade.

5.9. In vitro release of VEGF, curcumin, and PDGFbb from the hydrogel

A volume of 500 μL the composite IntHydrogel + VEGF + Lip@PDGFbb + curcumin hydrogel was dispensed into each well of a 24-well plate, followed by the addition of 1 mL PBS. Three replicate samples were incubated at 37°C for each experimental group. At predetermined intervals, 500 μL of the supernatant was collected and immediately replaced with an equal volume of fresh PBS. VEGF and PDGFbb concentrations in the collected samples were quantified using an ELISA kit according to the manufacturer's protocol. Cumulative release was calculated based on the initial loading amount and plotted over time.

Curcumin (MedChemExpress, HY-N0005, USA) was quantified by UV-visible (UV-vis) spectrophotometer (Shimadzu, Japan). A standard curve was generated using known curcumin concentrations (3.125, 6.25, 12.5, 25, and 50 μM) measured at 429 nm. Hydrogels containing curcumin were prepared as described above, and 500 μL of gel was immersed in 1 mL PBS at 37°C. At predetermined time points, 1 mL of the release medium was collected and replaced with fresh PBS. Each experiment was performed in triplicate, and mean values were used to plot cumulative release curves.

5.10. Thoracic spinal cord transection surgery and animal care

In this study, adult female Sprague-Dawley (SD) rats (180-200 g) were supplied by the Guangdong Provincial Medical Laboratory Animal Center (license no. SYXK(Yue)2022-0002). The experimental protocol was approved by the Institutional Animal Care and Use Committee (IACUC) at Sun Yat-sen University (Approval No. SYSU-IACUC-2026-000035), and all procedures followed the “Regulations for the Administration of Affairs Concerning Experimental Animals” within a specific pathogen-free (SPF) facility. After one week of acclimatization, rats were anesthetized with intraperitoneal injection of pentobarbital sodium (1%, 3.5 mL/kg body weight) and underwent a laminectomy under sterile conditions to expose the thoracic spinal segments (T9-T10). A 2-mm spinal cord segment was removed at the T10 level. Rats were randomly assigned to five experimental groups (24 rats per group): (a) Control group: Spinal cord was completely transected over a 2 mm segment without treatment. (b) IntHydrogel + VEGF group: Injection of IntHydrogel + VEGF hydrogel into the injury site. (c) IntHydrogel + Lip@PDGFbb group: Injection of IntHydrogel + Lip@PDGFbb hydrogel into the injury site. (d) IntHydrogel + VEGF + Lip@PDGFbb group: Injection of IntHydrogel + VEGF + Lip@PDGFbb hydrogel into the injury site. (e) IntHydrogel + VEGF + Lip@PDGFbb + Curcumin group: Injection of IntHydrogel + VEGF + Lip@PDGFbb + Curcumin hydrogel into the injury site. Following the surgery, the muscle and skin layers were closed using sutures. Postoperative care included administration of subcutaneous buprenorphine (0.05 mg/kg every 12 h for 3 days) for pain relief (Tianjin Pharmaceutical Research Institute, China) and intramuscular penicillin G (50,000 IU/kg/day for 7 days) to prevent infection (Jiangxi Keda Animal Pharmaceutical Co., Ltd., China). For analgesia, buprenorphine was initiated immediately after recovery from anesthesia and continued for 72 h post-surgery. Pain levels were assessed twice daily using the Rat Grimace Scale and behavioral observations (e.g., guarding, locomotion, and vocalization); additional rescue analgesia (subcutaneous buprenorphine 0.01 mg/kg) was administered if signs of moderate-to-severe pain persisted beyond 30 min after the scheduled dose. All animals were monitored for signs of distress, and no adverse events related to analgesia were observed. Bladder expression was performed twice daily, and soft bedding was provided to avoid pressure sores. The bladder was gently compressed until completely empty, and skin over bony prominences was inspected daily; additional heating and nutritional support were given as needed. The bedding was fluffed twice daily, and the animals were gently repositioned during each routine check to minimize sustained pressure on the abdomen and hindlimbs. The rats were housed in a Controlled environment (25°C, 12-h light/dark cycle).

5.11. Immunofluorescence staining and imaging

At specified time points, rats were euthanized, and spinal cord tissues surrounding the injury site, as well as bladder tissues, were collected. For tissue fixation, cardiac perfusion with saline followed by 4% PFA was performed. After dehydration in sucrose solutions, tissues were embedded in OCT compound and stored at −80°C. Longitudinal sections (10 μm) were cut using a cryostat, mounted on adhesive slides, and stored at −80°C. Sections were baked at 37°C for 1 h, washed with PBS, and blocked with 0.1% Triton X-100 and 5% BSA. Primary antibodies (Table S1, Supporting Information) were applied overnight at 4°C, followed by incubation with secondary antibodies (Table S1, Supporting Information) in the dark. The slides were mounted with anti-fade medium containing DAPI (ab104139, Abcam, UK). Fluorescent images were captured using an automated fluorescence microscope (Nikon ECLIPSE Ti2-E) and a confocal microscope (Andor Dragonfly CR-DFLY-202). Data analysis was conducted with ImageJ, ensuring a minimum of three biological replicates and multiple fields of view for each group.

5.12. Quantitative reverse transcription polymerase chain reaction (RT-PCR)

RNA extraction was performed using the E.Z.N.A. Total RNA Kit I (Omega, R6834-02, USA), according to the manufacturer's protocol. RNA concentration and purity were measured using a NanoDrop One Spectrophotometer (Thermo Fisher Scientific, PA, USA). For cDNA synthesis, 1 μg of RNA was reverse-transcribed using the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, K1622, USA). Quantitative RT-PCR was performed in a 10 μL reaction volume using PowerUp SYBR Green Master Mix (Thermo Fisher Scientific, A25742, USA) on the QuantStudio Real-Time PCR System (Applied Biosystems), with triplicate repeats to ensure reproducibility. Melting curve analysis was performed to confirm specificity. β-actin was used as the reference gene for data normalization, and relative gene expression was calculated using the ΔΔCt method. Primer sequences are provided in Table S3 (Supporting Information).

5.13. Western blotting

Tissue samples were lysed in RIPA buffer containing 1% PMSF, protease inhibitors, and EDTA on ice for 30 min. The lysates were centrifuged at 12000g for 20 min at 4°C to collect the supernatant. Protein concentrations were determined using the BCA Protein Assay Kit. Samples were diluted to 30 μg/10 μL with 5X loading buffer, boiled for 10 min, and centrifuged. Proteins were separated by SDS-PAGE (90 V for 30 min, 120 V for 60 min), transferred to PVDF membranes using a semi-dry transfer system (25 V, 1.0 A for 30 min), and blocked with 5% BSA for 60 min. Membranes were incubated overnight at 4°C with primary antibodies (Table S2, Supporting Information), followed by secondary antibody incubation at 25°C for 1 h. Protein bands were visualized using enhanced chemiluminescent reagents, and the relative expression levels were quantified using ImageJ, with GAPDH as the reference protein.

5.14. Evaluating blood vessel permeability at SCI site

To assess blood vessel permeability at the SCI site, a 4 kDa FITC-dextran (HY-128868A, MCE, USA) was intravenously injected into rats. First, the rats were anesthetized with isoflurane. Once the animals were fully anesthetized, a tail vein injection of 4 kDa FITC-dextran (50 mg/kg) was administered. The dye was allowed to circulate for 30 min to ensure adequate distribution throughout the vasculature. Following the circulation period, the rats were euthanized by cervical dislocation without prior perfusion to preserve the blood vessels and the integrity of the injured area. The spinal cord tissue was carefully excised and immediately placed in 4% PFA solution for fixation. After overnight fixation at 4°C, the tissues were dehydrated through a graded sucrose solution and then embedded in OCT compound for cryosectioning. Tissue was longitudinally sectioned into 10 μm thick sections using a cryostat. Sections were stained with Glut1 antibody labeled with endothelial cells and mounted with DAPI-containing mounting medium to preserve fluorescence. The stained sections were observed under a fluorescence microscope. FITC-dextran leakage was quantified by measuring fluorescence intensity in the damaged area using ImageJ software (green fluorescence indicates leakage).

5.15. Metabolomics and data analysis

At two weeks post-spinal cord injury modeling, spinal cord injury site tissues were harvested from rats in each experimental group. The tissues were immediately rinsed with ice-cold PBS to remove residual blood. Subsequently, the samples were promptly snap-frozen in liquid nitrogen to preserve metabolic profiles and prevent degradation. The frozen tissues were then transported and stored at −80°C until analysis. All samples were sent to LC-Bio Technology for absolute quantitative non-targeted metabolomic analysis and subsequent data processing. The top 20 differential metabolites with their p-values, FDR, and log₂FC are provided in Table S4 of the Supporting Information.

5.16. Quantification of in vivo cytokines

Spinal cord injury site tissues were collected and ground, then centrifuged at 300 g for 15 min at 4°C to obtain the supernatant. Cytokine levels in the supernatant were analyzed using a Rat Cytokine Array Kit (RayBiotech, GSR-CAA-67). Additionally, the concentrations of various cytokines were quantified using specific rat ELISA kits for PDGF-AA, VEGF, Galectin, Gas1, HGF, JAM-A, NRP-1, CINC-1, CINC-2, CINC-3, CCL11, IL-4, IL-10, IL-13, IL-1ra, IL-2ra, IL-3, and TCK-3.

5.17. Macrophage polarization and co-culture with BMECs

Primary bone marrow-derived macrophages (M0) were isolated from the femurs of adult rats and cultured in complete RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. To induce polarization into M2 macrophages, M0 cells were treated with interleukin-4 (IL-4, 20 ng/mL) and interleukin-10 (IL-10, 20 ng/mL) for 48 h. The media containing IL-4 and IL-10 was replaced every 24 h during the treatment period. Following polarization, the M2 macrophages were harvested and transferred to the upper chamber of a transwell insert (0.4 μm pore size, Corning, USA) placed in a 24-well plate. In the lower chamber, rat BMECs were seeded at a density of 1 × 10⁵ cells/well in complete endothelial cell growth medium (MSI012Ra11, cloud-clone, China). The co-culture system was maintained for 24-48 h at 37°C in a 5% CO₂ incubator, allowing for the interaction between M2 macrophages in the upper chamber and BMECs in the lower chamber.

5.18. BMEC vascular tube formation assay

To evaluate the angiogenic potential of BMECs under co-culture conditions, the vascular tube formation assay was performed. Briefly, 48 h after co-culturing with M2 macrophages, the BMECs were transferred to a 24-well plate pre-coated with 50 μL of Matrigel (354234, Corning, USA) and allowed to polymerize at 37°C for 30 min. After Matrigel polymerization, the co-cultured BMECs were added on top at a density of 1 × 10⁵ cells/well. The BMECs were incubated for 6-8 h, during which time they were monitored for the formation of capillary-like structures under an inverted microscope (Nikon ECLIPSE Ti2-E). Images of the tube structures were captured at regular intervals, and the total tube length and branch points were quantified using ImageJ software. The angiogenic response was compared between BMECs cultured alone and BMECs co-cultured with M2 macrophages to assess the influence of M2 polarization on endothelial cell tube formation.

5.19. Spinal cord microvascular imaging using ultrafast Ultrasound Localization Microscopy (ULM)

Spinal cord microvascular imaging was performed using an ULM system with an L22-14vX linear array transducer (15.625 MHz, PodaMed, China). The system utilized multiple tilted plane-wave illuminations for ultrafast imaging, achieving high frame rates and resolution. The data were acquired by applying robust principal component analysis (RPCA)-based spatial-temporal clutter filtering, which separated the microbubble signals from tissue and noise. The microbubbles were localized and tracked through their motion across consecutive frames. In this study, microbubbles were injected via the tail vein to serve as contrast agents. The motion trajectories of individual microbubbles were then tracked using the RS (radial symmetry) algorithm, and sub-wavelength resolution images of the microvascular network were reconstructed. The velocity of blood flow was quantified by calculating the displacement of microbubbles between consecutive frames, providing a detailed map of blood flow velocity. Three types of maps were generated: (a) blood flow density, showing the distribution and density of microbubbles within the spinal cord's microvascular network; (b) blood flow direction, which distinguished the direction of blood flow in adjacent vessels; and (c) blood flow velocity, which represented the speed of blood flow at various points within the vessels. These maps were used to analyze the microvascular dynamics and evaluate the blood flow characteristics.

5.20. Bladder ultrasonography assessment

At 8 weeks post-SCI, residual urine volume was measured by ultrasound. The probe of the ultrasound system (Mindray Animal M90 SCI) was placed on the suprapubic region of the rats, and bladder images were acquired in both sagittal and transverse planes. Maximum width, depth and height were measured, and residual urine volume was automatically calculated using the ellipsoid formula.

5.21. Hematoxylin-Eosin (HE) staining

Bladder sections were immersed in hematoxylin solution (PHYGENE, China) for 5 min, followed by a 10 s rinse with distilled water. Sections were then differentiated in hydrochloric acid-ethanol solution for 3 s and rinsed with distilled water for 30 s. Subsequently, sections were incubated in eosin solution (PHYGENE, China) for 5 min and rinsed with distilled water for 5 s. The sections were dehydrated through an ethanol gradient: 80% ethanol for 10-20 s, 90% ethanol for 10-20 s, 95% ethanol for 1-2 min, and anhydrous ethanol for 1-2 min. After clearing in xylene for 5 min, sections were mounted with neutral resin. Images were captured using an automatic slide scanning system (KFBIO/KF-FL-400) for further analysis.

5.22. Retrograde trans-synaptic PRV tracing

Retrograde trans-synaptic tracing was performed 8 weeks after spinal cord injury. Rats were deeply anesthetized with isoflurane (5% for induction, 2-3% for maintenance in 100% O₂). The bladder was exposed through a midline abdominal incision. Two injection sites were selected: one near the dome and the other on the lateral wall of the bladder detrusor muscle, with a needle insertion depth of 0.5-1.0 mm. A glass micropipette (tip diameter 20-30 μm) was used to inject 1 μL of PRV carrying red fluorescent protein (PRV-RFP, 1 × 10¹⁰ PFU/mL) at each site. The injection was performed at a rate of 100 nL/min using a microinjection pump. After each injection, the micropipette was left in place for 5 min to allow for sufficient diffusion and to minimize backflow, then slowly withdrawn over 1 min. Any tracer leakage at the puncture site was immediately absorbed with sterile cotton swabs, and the area was gently rinsed with sterile saline. Following the injections, the bladder was carefully repositioned into the abdominal cavity, and the incision was closed in layers with absorbable sutures. Animals were allowed to recover and were housed individually with standard postoperative care. Rats were sacrificed 5-7 days after PRV injection for tissue processing and fluorescence analysis.

5.23. Statistical analysis

Statistical analyses were performed using GraphPad Prism 9.0 software (GraphPad Software, San Diego, CA, USA). Data are presented as the mean ± standard deviation (SD). Comparisons between groups were made using either a two-way repeated measures analysis of variance (ANOVA) with Bonferroni's post hoc test or a one-way repeated measures ANOVA followed by Tukey's post hoc comparisons. Statistical significance was determined at a p-value of <0.05.

CRediT authorship contribution statement

Haiyan Weng: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Supervision. Longyou Xiao: Data curation, Funding acquisition, Investigation, Methodology, Resources, Software, Writing – original draft. Ting Li: Data curation, Formal analysis, Methodology. Hengrui Chang: Methodology, Writing – review & editing. Pengfei Xie: Methodology, Software. Liyan Zhu: Software, Validation. Baobao Zhang: Software, Visualization. Shiqin Lv: Investigation. Zijun Hu: Supervision. Yu Dai: Data curation. Pui Man Hoi: Writing – review & editing. Wanguo Liu: Supervision. Jialin Liu: Software. Limin Rong: Funding acquisition, Resources, Writing – review & editing. Yi Ren: Writing – review & editing. Liumin He: Conceptualization, Funding acquisition, Resources, Software, Supervision, Validation, Writing – original draft, Writing – review & editing.

Ethics approval and consent to participate

All animal experiments were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of Sun Yat-sen University (approval No. SYSU-IACUC-2026-000035). This study used 120 CD(SD)IGS rats for research on spinal cord injury repair. Experiments were conducted from 8 December 2025 to 9 February 2026 at the Sun Yat-sen University East Campus Laboratory Animal Center (Permit SYXK (Yue) 2023-0112). The protocol was permitted after revision on 4 January 2026, complying with national laboratory animal welfare and ethical standards. No human participants were involved in this study, so human informed consent is not required. All animal operations followed the approved ethical protocol to minimise animal suffering.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Liumin He reports financial support was provided by National Key Research and Development Program of China. Liumin He reports financial support was provided by National Natural Science Foundation of China. Liumin He reports financial support was provided by Municipality-Hospital Co-founded Special Projects of Guangzhou Basic Research Program. Liumin He reports financial support was provided by Excellent Youth Foundation of Guangdong Scientific Committee. Liumin He has patent #202111236338.3 issued to China National Intellectual Property Administration. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was supported by the National Key Research and Development Program of China (2024YFF1206400), National Natural Science Foundation of China (92468101, 32271417, 825B2078, and U22A20297), the Municipality-Hospital Co-founded Special Projects of Guangzhou Basic Research Program (2025A03J3216), Excellent Youth Foundation of Guangdong Scientific Committee (2022B1515020083).

Footnotes

Peer review under the responsibility of editorial board of Bioactive Materials.

^{Appendix A}

Supplementary data to this article can be found online at https://doi.org/10.1016/j.bioactmat.2026.07.046.

Contributor Information

Limin Rong, Email: ronglm@mail.sysu.edu.cn.

Yi Ren, Email: yren05@gzhmu.edu.cn.

Liumin He, Email: helm9@mail.sysu.edu.cn.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Multimedia component 1

mmc1.docx^{(4.5MB, docx)}

References

1.David G., Mohammadi S., Martin A.R., et al. Traumatic and nontraumatic spinal cord injury: pathological insights from neuroimaging. Nat. Rev. Neurol. 2019;15:718–731. doi: 10.1038/s41582-019-0270-5. [DOI] [PubMed] [Google Scholar]
2.Freund P., Seif M., Weiskopf N., et al. Mri in traumatic spinal cord injury: from clinical assessment to neuroimaging biomarkers. Lancet Neurol. 2019;18:1123–1135. doi: 10.1016/S1474-4422(19)30138-3. [DOI] [PubMed] [Google Scholar]
3.Hu X., Xu W., Ren Y., et al. Spinal cord injury: molecular mechanisms and therapeutic interventions. Signal Transduct. Targeted Ther. 2023;8:245. doi: 10.1038/s41392-023-01477-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Ma Y., Yu X., Pan J., et al. Exosomes: a promising microenvironment modulator for spinal cord injury treatment. Int. J. Biol. Sci. 2025;21:3791–3824. doi: 10.7150/ijbs.115242. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Tan Z., Xiao L., Ma J., et al. Integrating hydrogels manipulate ecm deposition after spinal cord injury for specific neural reconnections via neuronal relays. Sci. Adv. 2024;10:eado9120. doi: 10.1126/sciadv.ado9120. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Alvarez Z., Kolberg-Edelbrock A.N., Sasselli I.R., et al. Bioactive scaffolds with enhanced supramolecular motion promote recovery from spinal cord injury. Science (New York, NY) 2021;374:848–856. doi: 10.1126/science.abh3602. [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Ahuja C.S., Wilson J.R., Nori S., et al. Traumatic spinal cord injury. Nat. Rev. Dis. Primers. 2017;3 doi: 10.1038/nrdp.2017.18. [DOI] [PubMed] [Google Scholar]
8.Ruiz de Almodovar C., Dupraz S., Bonanomi D. Neurovascular dynamics in the spinal cord from development to pathophysiology. Neuron. 2025;113:4134–4157. doi: 10.1016/j.neuron.2025.09.017. [DOI] [PubMed] [Google Scholar]
9.Gianni-Barrera R., Butschkau A., Uccelli A., et al. Pdgf-bb regulates splitting angiogenesis in skeletal muscle by limiting vegf-induced endothelial proliferation. Angiogenesis. 2018;21:883–900. doi: 10.1007/s10456-018-9634-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Peng W., Xie Y., Liu Y., et al. Targeted delivery of cd163(+) macrophage-derived small extracellular vesicles via rgd peptides promote vascular regeneration and stabilization after spinal cord injury. J. Contr. Release. 2023;361:750–765. doi: 10.1016/j.jconrel.2023.08.025. [DOI] [PubMed] [Google Scholar]
11.Cao L., Mooney D.J. Spatiotemporal control over growth factor signaling for therapeutic neovascularization. Adv. Drug Deliv. Rev. 2007;59:1340–1350. doi: 10.1016/j.addr.2007.08.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Richardson T.P., Peters M.C., Ennett A.B., et al. Polymeric system for dual growth factor delivery. Nat. Biotechnol. 2001;19:1029–1034. doi: 10.1038/nbt1101-1029. [DOI] [PubMed] [Google Scholar]
13.Carmeliet P., Jain R.K. Molecular mechanisms and clinical applications of angiogenesis. Nature. 2011;473:298–307. doi: 10.1038/nature10144. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Duan Y., Yang F., Zhang Y., et al. Role of mitophagy in spinal cord ischemia-reperfusion injury. Neural Regen. Res. 2026;21:598–611. doi: 10.4103/NRR.NRR-D-24-00668. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Gao Z., Xing H., Shen Y., et al. Neuronal hdac3 knockdown promotes propriospinal detour pathway formation and locomotor recovery in a mouse model of spinal cord injury. Sci. Transl. Med. 2025;17:eadp1873. doi: 10.1126/scitranslmed.adp1873. [DOI] [PubMed] [Google Scholar]
16.Zhang D., Zhou X., Zhao C., et al. Spatiotemporal-controlled ultrasound-driven li-pda@zno nanoparticles promote neural stem cell differentiation synergy with biohydrogel repair spinal cord injury. Bioact. Mater. 2026;55:446–463. doi: 10.1016/j.bioactmat.2025.07.050. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Bai J., Zhang Y., Fan Q., et al. Reactive oxygen species-scavenging scaffold with rapamycin for treatment of intervertebral disk degeneration. Adv. Healthcare Mater. 2020;9 doi: 10.1002/adhm.201901186. [DOI] [PubMed] [Google Scholar]
18.Chen L., Xiao L., Ma Y., et al. Bioengineered composite hydrogel scaffold for accelerated skin regeneration and wound repair. Chem. Eng. J. 2025;504 [Google Scholar]
19.Bradbury E.J., Burnside E.R. Moving beyond the glial scar for spinal cord repair. Nat. Commun. 2019;10:3879. doi: 10.1038/s41467-019-11707-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Cao Y., Zhu S., Yu B., et al. Single-cell rna sequencing for traumatic spinal cord injury. FASEB J. 2022;36 doi: 10.1096/fj.202200943R. [DOI] [PubMed] [Google Scholar]
21.Liu Z., Xiang C., Zhao X., et al. Regulation of dynamic spatiotemporal inflammation by nanomaterials in spinal cord injury. J. Nanobiotechnol. 2024;22:767. doi: 10.1186/s12951-024-03037-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Zhou R., Li J., Wang R., et al. The neurovascular unit in healthy and injured spinal cord. J. Cereb. Blood Flow Metab. 2023;43:1437–1455. doi: 10.1177/0271678X231172008. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Hoeben A., Landuyt B., Highley M.S., et al. Vascular endothelial growth factor and angiogenesis. Pharmacol. Rev. 2004;56:549–580. doi: 10.1124/pr.56.4.3. [DOI] [PubMed] [Google Scholar]
24.Olsson A.K., Dimberg A., Kreuger J., et al. Vegf receptor signalling - in control of vascular function. Nat. Rev. Mol. Cell Biol. 2006;7:359–371. doi: 10.1038/nrm1911. [DOI] [PubMed] [Google Scholar]
25.Menezes K., Rosa B.G., Freitas C., et al. Human mesenchymal stromal/stem cells recruit resident pericytes and induce blood vessels maturation to repair experimental spinal cord injury in rats. Sci. Rep. 2020;10 doi: 10.1038/s41598-020-76290-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Roolfs L., Hubertus V., Spinnen J., et al. Therapeutic approaches targeting vascular repair after experimental spinal cord injury: a systematic review of the literature. Neurospine. 2022;19:961–975. doi: 10.14245/ns.2244624.312. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Anderson M.A., O'Shea T.M., Burda J.E., et al. Required growth facilitators propel axon regeneration across complete spinal cord injury. Nature. 2018;561:396–400. doi: 10.1038/s41586-018-0467-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Courtine G., Sofroniew M.V. Spinal cord repair: advances in biology and technology. Nat. Med. 2019;25:898–908. doi: 10.1038/s41591-019-0475-6. [DOI] [PubMed] [Google Scholar]
29.Lu P., Wang Y., Graham L., et al. Long-distance growth and connectivity of neural stem cells after severe spinal cord injury. Cell. 2012;150:1264–1273. doi: 10.1016/j.cell.2012.08.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Yang Z., Zhang A., Duan H., et al. Nt3-chitosan elicits robust endogenous neurogenesis to enable functional recovery after spinal cord injury. Proc. Natl. Acad. Sci. U. S. A. 2015;112:13354–13359. doi: 10.1073/pnas.1510194112. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Schaffran B., Bradke F. Reproducibility - the key towards clinical implementation of spinal cord injury treatments? Exp. Neurol. 2019;313:135–136. doi: 10.1016/j.expneurol.2018.12.010. [DOI] [PubMed] [Google Scholar]
32.Sharp K.G., Yee K.M., Steward O. A re-assessment of long distance growth and connectivity of neural stem cells after severe spinal cord injury. Exp. Neurol. 2014;257:186–204. doi: 10.1016/j.expneurol.2014.04.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.de Groat W.C., Griffiths D., Yoshimura N. Neural control of the lower urinary tract. Compr. Physiol. 2015;5:327–396. doi: 10.1002/cphy.c130056. [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Fowler C.J., Griffiths D., de Groat W.C. The neural control of micturition. Nat. Rev. Neurosci. 2008;9:453–466. doi: 10.1038/nrn2401. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Kim K.T. The importance of vascular repair as the first step in spinal cord injury treatment: commentary on "therapeutic approaches targeting vascular repair after experimental spinal cord injury: a systematic review of the literature". Neurospine. 2022;19:976–977. doi: 10.14245/ns.2245022.511. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Li G., Zhang B., Sun J.H., et al. An nt-3-releasing bioscaffold supports the formation of trkc-modified neural stem cell-derived neural network tissue with efficacy in repairing spinal cord injury. Bioact. Mater. 2021;6:3766–3781. doi: 10.1016/j.bioactmat.2021.03.036. [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Marrotte E.J., Johnson K., Schweller R.M., et al. Induction of neurogenesis and angiogenesis in a rat hemisection spinal cord injury model with combined neural stem cell, endothelial progenitor cell, and biomimetic hydrogel matrix therapy. Crit Care Explor. 2021;3 doi: 10.1097/CCE.0000000000000436. [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Yao C., Cao X., Yu B. Revascularization after traumatic spinal cord injury. Front. Physiol. 2021;12 doi: 10.3389/fphys.2021.631500. [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Tsivelekas K., Evangelopoulos D.S., Pallis D., et al. Angiogenesis in spinal cord injury: progress and treatment. Cureus. 2022;14 doi: 10.7759/cureus.25475. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Winkler E.A., Sengillo J.D., Bell R.D., et al. Blood-spinal cord barrier pericyte reductions contribute to increased capillary permeability. J. Cereb. Blood Flow Metab. 2012;32:1841–1852. doi: 10.1038/jcbfm.2012.113. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Zhou R., Li J., Chen Z., et al. Pathological hemodynamic changes and leukocyte transmigration disrupt the blood-spinal cord barrier after spinal cord injury. J. Neuroinflammation. 2023;20:118. doi: 10.1186/s12974-023-02787-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Liu X., Liu X., Luo S., et al. Quercetin promotes angiogenesis and protects the blood-spinal cord barrier structure after spinal cord injury by targeting the pi3k/akt signaling pathway. J. Transl. Med. 2025;23:958. doi: 10.1186/s12967-025-06973-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Takakura N. Role of intimate interactions between endothelial cells and the surrounding accessory cells in the maturation of blood vessels. J. Thromb. Haemostasis. 2011;9(Suppl 1):144–150. doi: 10.1111/j.1538-7836.2011.04275.x. [DOI] [PubMed] [Google Scholar]
44.Wilhelm I., Gyori F., Dudas T., et al. Eppur si muove: the dynamic brain pericyte. Fluids Barriers CNS. 2025;22:95. doi: 10.1186/s12987-025-00706-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Gong Y., Wu M., Huang Y., et al. Research developments in the neurovascular unit and the blood-brain barrier. Biomed Rep. 2025;22:88. doi: 10.3892/br.2025.1966. (review) [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Rowsthorn E., Pham W., Nazem-Zadeh M.R., et al. Imaging the neurovascular unit in health and neurodegeneration: a scoping review of interdependencies between mri measures. Fluids Barriers CNS. 2023;20:97. doi: 10.1186/s12987-023-00499-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Zarekiani P., Nogueira Pinto H., Hol E.M., et al. The neurovascular unit in leukodystrophies: towards solving the puzzle. Fluids Barriers CNS. 2022;19:18. doi: 10.1186/s12987-022-00316-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Li H., Liu Y., Sun Y., et al. Targeting astrocytes polarization after spinal cord injury: a promising direction. Front. Cell. Neurosci. 2024;18 doi: 10.3389/fncel.2024.1478741. [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Wu H., Feng E., Yin H., et al. Biomaterials for neuroengineering: applications and challenges. Regen. Biomater. 2025;12:rbae137. doi: 10.1093/rb/rbae137. [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Jagrit V., Koffler J., Dulin J.N. Combinatorial strategies for cell transplantation in traumatic spinal cord injury. Front. Neurosci. 2024;18 doi: 10.3389/fnins.2024.1349446. [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Pieczonka K., Fehlings M.G. Incorporating combinatorial approaches to encourage targeted neural stem/progenitor cell integration following transplantation in spinal cord injury. Stem Cells Transl. Med. 2023;12:207–214. doi: 10.1093/stcltm/szad008. [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Entenmann C.J., von Bronewski E.J., Waldmann L., et al. Analysis of the spatiotemporal dynamics of vascular injury and regeneration following experimental spinal cord injury. Brain Spine. 2025;5 doi: 10.1016/j.bas.2025.104191. [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Couvrette L.J., Walker K.L.A., Bayat A., et al. Poly-l-ornithine coated plant scaffolds support motor recovery in rats after traumatic spinal cord injury. Sci. Rep. 2025;15 doi: 10.1038/s41598-025-21856-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Shrestha B., Coykendall K., Li Y., et al. Repair of injured spinal cord using biomaterial scaffolds and stem cells. Stem Cell Res. Ther. 2014;5:91. doi: 10.1186/scrt480. [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Ferreira A., Chambel S.S., Reguenga C., et al. Development of urinary impairment after spinal cord injury courses with altered urethral serotonin signalling in the female mice. Sci. Rep. 2025;16:8. doi: 10.1038/s41598-025-29193-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Pavese C., Kessler T.M. Prediction of lower urinary tract, sexual, and bowel function, and autonomic dysreflexia after spinal cord injury. Biomedicines. 2023;11 doi: 10.3390/biomedicines11061644. [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Hogg F.R.A., Kearney S., Solomon E., et al. Acute, severe traumatic spinal cord injury: improving urinary bladder function by optimizing spinal cord perfusion. J. Neurosurg. Spine. 2022;36:145–152. doi: 10.3171/2021.3.SPINE202056. [DOI] [PubMed] [Google Scholar]
58.Keller E.E., Bauer S., Roider K., et al. Montelukast improves urinary bladder function after complete spinal cord injury in rats. Int. J. Mol. Sci. 2025;26 doi: 10.3390/ijms26125606. [DOI] [PMC free article] [PubMed] [Google Scholar]
59.Kjell J., Olson L. Rat models of spinal cord injury: from pathology to potential therapies. Dis. Model. Mech. 2016;9:1125–1137. doi: 10.1242/dmm.025833. [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Verstappen K., Aquarius R., Klymov A., et al. Systematic evaluation of spinal cord injury animal models in the field of biomaterials. Tissue Eng., Part B. 2022;28:1169–1179. doi: 10.1089/ten.teb.2021.0194. [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Lu P. Stem cell transplantation for spinal cord injury repair. Prog. Brain Res. 2017;231:1–32. doi: 10.1016/bs.pbr.2016.11.012. [DOI] [PubMed] [Google Scholar]
62.Ahuja C.S., Nori S., Tetreault L., et al. Traumatic spinal cord injury-repair and regeneration. Neurosurgery. 2017;80:S9–S22. doi: 10.1093/neuros/nyw080. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Multimedia component 1

mmc1.docx^{(4.5MB, docx)}

[bib1] 1.David G., Mohammadi S., Martin A.R., et al. Traumatic and nontraumatic spinal cord injury: pathological insights from neuroimaging. Nat. Rev. Neurol. 2019;15:718–731. doi: 10.1038/s41582-019-0270-5. [DOI] [PubMed] [Google Scholar]

[bib2] 2.Freund P., Seif M., Weiskopf N., et al. Mri in traumatic spinal cord injury: from clinical assessment to neuroimaging biomarkers. Lancet Neurol. 2019;18:1123–1135. doi: 10.1016/S1474-4422(19)30138-3. [DOI] [PubMed] [Google Scholar]

[bib3] 3.Hu X., Xu W., Ren Y., et al. Spinal cord injury: molecular mechanisms and therapeutic interventions. Signal Transduct. Targeted Ther. 2023;8:245. doi: 10.1038/s41392-023-01477-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib4] 4.Ma Y., Yu X., Pan J., et al. Exosomes: a promising microenvironment modulator for spinal cord injury treatment. Int. J. Biol. Sci. 2025;21:3791–3824. doi: 10.7150/ijbs.115242. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib5] 5.Tan Z., Xiao L., Ma J., et al. Integrating hydrogels manipulate ecm deposition after spinal cord injury for specific neural reconnections via neuronal relays. Sci. Adv. 2024;10:eado9120. doi: 10.1126/sciadv.ado9120. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib6] 6.Alvarez Z., Kolberg-Edelbrock A.N., Sasselli I.R., et al. Bioactive scaffolds with enhanced supramolecular motion promote recovery from spinal cord injury. Science (New York, NY) 2021;374:848–856. doi: 10.1126/science.abh3602. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib7] 7.Ahuja C.S., Wilson J.R., Nori S., et al. Traumatic spinal cord injury. Nat. Rev. Dis. Primers. 2017;3 doi: 10.1038/nrdp.2017.18. [DOI] [PubMed] [Google Scholar]

[bib8] 8.Ruiz de Almodovar C., Dupraz S., Bonanomi D. Neurovascular dynamics in the spinal cord from development to pathophysiology. Neuron. 2025;113:4134–4157. doi: 10.1016/j.neuron.2025.09.017. [DOI] [PubMed] [Google Scholar]

[bib9] 9.Gianni-Barrera R., Butschkau A., Uccelli A., et al. Pdgf-bb regulates splitting angiogenesis in skeletal muscle by limiting vegf-induced endothelial proliferation. Angiogenesis. 2018;21:883–900. doi: 10.1007/s10456-018-9634-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib10] 10.Peng W., Xie Y., Liu Y., et al. Targeted delivery of cd163(+) macrophage-derived small extracellular vesicles via rgd peptides promote vascular regeneration and stabilization after spinal cord injury. J. Contr. Release. 2023;361:750–765. doi: 10.1016/j.jconrel.2023.08.025. [DOI] [PubMed] [Google Scholar]

[bib11] 11.Cao L., Mooney D.J. Spatiotemporal control over growth factor signaling for therapeutic neovascularization. Adv. Drug Deliv. Rev. 2007;59:1340–1350. doi: 10.1016/j.addr.2007.08.012. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib12] 12.Richardson T.P., Peters M.C., Ennett A.B., et al. Polymeric system for dual growth factor delivery. Nat. Biotechnol. 2001;19:1029–1034. doi: 10.1038/nbt1101-1029. [DOI] [PubMed] [Google Scholar]

[bib13] 13.Carmeliet P., Jain R.K. Molecular mechanisms and clinical applications of angiogenesis. Nature. 2011;473:298–307. doi: 10.1038/nature10144. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib14] 14.Duan Y., Yang F., Zhang Y., et al. Role of mitophagy in spinal cord ischemia-reperfusion injury. Neural Regen. Res. 2026;21:598–611. doi: 10.4103/NRR.NRR-D-24-00668. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib15] 15.Gao Z., Xing H., Shen Y., et al. Neuronal hdac3 knockdown promotes propriospinal detour pathway formation and locomotor recovery in a mouse model of spinal cord injury. Sci. Transl. Med. 2025;17:eadp1873. doi: 10.1126/scitranslmed.adp1873. [DOI] [PubMed] [Google Scholar]

[bib16] 16.Zhang D., Zhou X., Zhao C., et al. Spatiotemporal-controlled ultrasound-driven li-pda@zno nanoparticles promote neural stem cell differentiation synergy with biohydrogel repair spinal cord injury. Bioact. Mater. 2026;55:446–463. doi: 10.1016/j.bioactmat.2025.07.050. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib17] 17.Bai J., Zhang Y., Fan Q., et al. Reactive oxygen species-scavenging scaffold with rapamycin for treatment of intervertebral disk degeneration. Adv. Healthcare Mater. 2020;9 doi: 10.1002/adhm.201901186. [DOI] [PubMed] [Google Scholar]

[bib18] 18.Chen L., Xiao L., Ma Y., et al. Bioengineered composite hydrogel scaffold for accelerated skin regeneration and wound repair. Chem. Eng. J. 2025;504 [Google Scholar]

[bib19] 19.Bradbury E.J., Burnside E.R. Moving beyond the glial scar for spinal cord repair. Nat. Commun. 2019;10:3879. doi: 10.1038/s41467-019-11707-7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib20] 20.Cao Y., Zhu S., Yu B., et al. Single-cell rna sequencing for traumatic spinal cord injury. FASEB J. 2022;36 doi: 10.1096/fj.202200943R. [DOI] [PubMed] [Google Scholar]

[bib21] 21.Liu Z., Xiang C., Zhao X., et al. Regulation of dynamic spatiotemporal inflammation by nanomaterials in spinal cord injury. J. Nanobiotechnol. 2024;22:767. doi: 10.1186/s12951-024-03037-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib22] 22.Zhou R., Li J., Wang R., et al. The neurovascular unit in healthy and injured spinal cord. J. Cereb. Blood Flow Metab. 2023;43:1437–1455. doi: 10.1177/0271678X231172008. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib23] 23.Hoeben A., Landuyt B., Highley M.S., et al. Vascular endothelial growth factor and angiogenesis. Pharmacol. Rev. 2004;56:549–580. doi: 10.1124/pr.56.4.3. [DOI] [PubMed] [Google Scholar]

[bib24] 24.Olsson A.K., Dimberg A., Kreuger J., et al. Vegf receptor signalling - in control of vascular function. Nat. Rev. Mol. Cell Biol. 2006;7:359–371. doi: 10.1038/nrm1911. [DOI] [PubMed] [Google Scholar]

[bib25] 25.Menezes K., Rosa B.G., Freitas C., et al. Human mesenchymal stromal/stem cells recruit resident pericytes and induce blood vessels maturation to repair experimental spinal cord injury in rats. Sci. Rep. 2020;10 doi: 10.1038/s41598-020-76290-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib26] 26.Roolfs L., Hubertus V., Spinnen J., et al. Therapeutic approaches targeting vascular repair after experimental spinal cord injury: a systematic review of the literature. Neurospine. 2022;19:961–975. doi: 10.14245/ns.2244624.312. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib27] 27.Anderson M.A., O'Shea T.M., Burda J.E., et al. Required growth facilitators propel axon regeneration across complete spinal cord injury. Nature. 2018;561:396–400. doi: 10.1038/s41586-018-0467-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib28] 28.Courtine G., Sofroniew M.V. Spinal cord repair: advances in biology and technology. Nat. Med. 2019;25:898–908. doi: 10.1038/s41591-019-0475-6. [DOI] [PubMed] [Google Scholar]

[bib29] 29.Lu P., Wang Y., Graham L., et al. Long-distance growth and connectivity of neural stem cells after severe spinal cord injury. Cell. 2012;150:1264–1273. doi: 10.1016/j.cell.2012.08.020. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib30] 30.Yang Z., Zhang A., Duan H., et al. Nt3-chitosan elicits robust endogenous neurogenesis to enable functional recovery after spinal cord injury. Proc. Natl. Acad. Sci. U. S. A. 2015;112:13354–13359. doi: 10.1073/pnas.1510194112. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib31] 31.Schaffran B., Bradke F. Reproducibility - the key towards clinical implementation of spinal cord injury treatments? Exp. Neurol. 2019;313:135–136. doi: 10.1016/j.expneurol.2018.12.010. [DOI] [PubMed] [Google Scholar]

[bib32] 32.Sharp K.G., Yee K.M., Steward O. A re-assessment of long distance growth and connectivity of neural stem cells after severe spinal cord injury. Exp. Neurol. 2014;257:186–204. doi: 10.1016/j.expneurol.2014.04.008. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib33] 33.de Groat W.C., Griffiths D., Yoshimura N. Neural control of the lower urinary tract. Compr. Physiol. 2015;5:327–396. doi: 10.1002/cphy.c130056. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib34] 34.Fowler C.J., Griffiths D., de Groat W.C. The neural control of micturition. Nat. Rev. Neurosci. 2008;9:453–466. doi: 10.1038/nrn2401. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib35] 35.Kim K.T. The importance of vascular repair as the first step in spinal cord injury treatment: commentary on "therapeutic approaches targeting vascular repair after experimental spinal cord injury: a systematic review of the literature". Neurospine. 2022;19:976–977. doi: 10.14245/ns.2245022.511. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib36] 36.Li G., Zhang B., Sun J.H., et al. An nt-3-releasing bioscaffold supports the formation of trkc-modified neural stem cell-derived neural network tissue with efficacy in repairing spinal cord injury. Bioact. Mater. 2021;6:3766–3781. doi: 10.1016/j.bioactmat.2021.03.036. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib37] 37.Marrotte E.J., Johnson K., Schweller R.M., et al. Induction of neurogenesis and angiogenesis in a rat hemisection spinal cord injury model with combined neural stem cell, endothelial progenitor cell, and biomimetic hydrogel matrix therapy. Crit Care Explor. 2021;3 doi: 10.1097/CCE.0000000000000436. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib38] 38.Yao C., Cao X., Yu B. Revascularization after traumatic spinal cord injury. Front. Physiol. 2021;12 doi: 10.3389/fphys.2021.631500. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib39] 39.Tsivelekas K., Evangelopoulos D.S., Pallis D., et al. Angiogenesis in spinal cord injury: progress and treatment. Cureus. 2022;14 doi: 10.7759/cureus.25475. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib40] 40.Winkler E.A., Sengillo J.D., Bell R.D., et al. Blood-spinal cord barrier pericyte reductions contribute to increased capillary permeability. J. Cereb. Blood Flow Metab. 2012;32:1841–1852. doi: 10.1038/jcbfm.2012.113. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib41] 41.Zhou R., Li J., Chen Z., et al. Pathological hemodynamic changes and leukocyte transmigration disrupt the blood-spinal cord barrier after spinal cord injury. J. Neuroinflammation. 2023;20:118. doi: 10.1186/s12974-023-02787-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib42] 42.Liu X., Liu X., Luo S., et al. Quercetin promotes angiogenesis and protects the blood-spinal cord barrier structure after spinal cord injury by targeting the pi3k/akt signaling pathway. J. Transl. Med. 2025;23:958. doi: 10.1186/s12967-025-06973-7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib43] 43.Takakura N. Role of intimate interactions between endothelial cells and the surrounding accessory cells in the maturation of blood vessels. J. Thromb. Haemostasis. 2011;9(Suppl 1):144–150. doi: 10.1111/j.1538-7836.2011.04275.x. [DOI] [PubMed] [Google Scholar]

[bib44] 44.Wilhelm I., Gyori F., Dudas T., et al. Eppur si muove: the dynamic brain pericyte. Fluids Barriers CNS. 2025;22:95. doi: 10.1186/s12987-025-00706-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib45] 45.Gong Y., Wu M., Huang Y., et al. Research developments in the neurovascular unit and the blood-brain barrier. Biomed Rep. 2025;22:88. doi: 10.3892/br.2025.1966. (review) [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib46] 46.Rowsthorn E., Pham W., Nazem-Zadeh M.R., et al. Imaging the neurovascular unit in health and neurodegeneration: a scoping review of interdependencies between mri measures. Fluids Barriers CNS. 2023;20:97. doi: 10.1186/s12987-023-00499-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib47] 47.Zarekiani P., Nogueira Pinto H., Hol E.M., et al. The neurovascular unit in leukodystrophies: towards solving the puzzle. Fluids Barriers CNS. 2022;19:18. doi: 10.1186/s12987-022-00316-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib48] 48.Li H., Liu Y., Sun Y., et al. Targeting astrocytes polarization after spinal cord injury: a promising direction. Front. Cell. Neurosci. 2024;18 doi: 10.3389/fncel.2024.1478741. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib49] 49.Wu H., Feng E., Yin H., et al. Biomaterials for neuroengineering: applications and challenges. Regen. Biomater. 2025;12:rbae137. doi: 10.1093/rb/rbae137. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib50] 50.Jagrit V., Koffler J., Dulin J.N. Combinatorial strategies for cell transplantation in traumatic spinal cord injury. Front. Neurosci. 2024;18 doi: 10.3389/fnins.2024.1349446. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib51] 51.Pieczonka K., Fehlings M.G. Incorporating combinatorial approaches to encourage targeted neural stem/progenitor cell integration following transplantation in spinal cord injury. Stem Cells Transl. Med. 2023;12:207–214. doi: 10.1093/stcltm/szad008. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib52] 52.Entenmann C.J., von Bronewski E.J., Waldmann L., et al. Analysis of the spatiotemporal dynamics of vascular injury and regeneration following experimental spinal cord injury. Brain Spine. 2025;5 doi: 10.1016/j.bas.2025.104191. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib53] 53.Couvrette L.J., Walker K.L.A., Bayat A., et al. Poly-l-ornithine coated plant scaffolds support motor recovery in rats after traumatic spinal cord injury. Sci. Rep. 2025;15 doi: 10.1038/s41598-025-21856-z. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib54] 54.Shrestha B., Coykendall K., Li Y., et al. Repair of injured spinal cord using biomaterial scaffolds and stem cells. Stem Cell Res. Ther. 2014;5:91. doi: 10.1186/scrt480. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib55] 55.Ferreira A., Chambel S.S., Reguenga C., et al. Development of urinary impairment after spinal cord injury courses with altered urethral serotonin signalling in the female mice. Sci. Rep. 2025;16:8. doi: 10.1038/s41598-025-29193-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib56] 56.Pavese C., Kessler T.M. Prediction of lower urinary tract, sexual, and bowel function, and autonomic dysreflexia after spinal cord injury. Biomedicines. 2023;11 doi: 10.3390/biomedicines11061644. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib57] 57.Hogg F.R.A., Kearney S., Solomon E., et al. Acute, severe traumatic spinal cord injury: improving urinary bladder function by optimizing spinal cord perfusion. J. Neurosurg. Spine. 2022;36:145–152. doi: 10.3171/2021.3.SPINE202056. [DOI] [PubMed] [Google Scholar]

[bib58] 58.Keller E.E., Bauer S., Roider K., et al. Montelukast improves urinary bladder function after complete spinal cord injury in rats. Int. J. Mol. Sci. 2025;26 doi: 10.3390/ijms26125606. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib59] 59.Kjell J., Olson L. Rat models of spinal cord injury: from pathology to potential therapies. Dis. Model. Mech. 2016;9:1125–1137. doi: 10.1242/dmm.025833. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib60] 60.Verstappen K., Aquarius R., Klymov A., et al. Systematic evaluation of spinal cord injury animal models in the field of biomaterials. Tissue Eng., Part B. 2022;28:1169–1179. doi: 10.1089/ten.teb.2021.0194. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib61] 61.Lu P. Stem cell transplantation for spinal cord injury repair. Prog. Brain Res. 2017;231:1–32. doi: 10.1016/bs.pbr.2016.11.012. [DOI] [PubMed] [Google Scholar]

[bib62] 62.Ahuja C.S., Nori S., Tetreault L., et al. Traumatic spinal cord injury-repair and regeneration. Neurosurgery. 2017;80:S9–S22. doi: 10.1093/neuros/nyw080. [DOI] [PubMed] [Google Scholar]

PERMALINK

Engineering functional vasculatures to reconstruct sympathetic-parasympathetic circuits for bladder function after spinal cord injury

Haiyan Weng

Longyou Xiao

Ting Li

Hengrui Chang

Pengfei Xie

Liyan Zhu

Baobao Zhang

Shiqin Lv

Zijun Hu

Yu Dai

Pui Man Hoi

Wanguo Liu

Jialin Liu

Limin Rong

Yi Ren

Liumin He

Abstract

Graphical abstract

Highlights

1. Introduction

2. Results

2.1. Fabrication of a biomimetic IntHydrogel platform for coordinated angiogenic stimulation

Fig. 1.

2.2. Implanted spinal cord substitute (ICS) with the IntHydrogel platform defines a standardized substitute segment and shapes the immune microenvironment after SCI

Fig. 2.

2.3. IntHydrogel substitute segment recruits host vessels and promotes early vascular maturation after SCI

Fig. 3.

2.4. Immune conditioning within the ICS instructs NRP1-mediated endothelial activation and vascular growth

Fig. 4.

2.5. ICS supports long-term vascular maturation and functional perfusion after SCI

Fig. 5.

2.6. Vascularization of the ICS enhances local metabolic support after SCI

Fig. 6.

2.7. ICS supports axonal entry and synaptic formation after SCI

Fig. 7.

2.8. The ICS enables re-engagement of supraspinal-spinal autonomic circuits and restoration of bladder function after SCI

Fig. 8.

3. Discussion

4. Conclusion

5. Methods

5.1. Materials

5.2. Synthesis and characterization of HADA

5.3. Preparation of PDGFbb-loaded sustained-release liposomes (Lip@PDGFbb)

5.4. Preparation of hydrogels

5.5. Transmission electron microscopy (TEM)

5.6. Dynamic light scattering (DLS)

5.7. Scanning electron microscopy (SEM)

5.8. Rheological tests

5.9. In vitro release of VEGF, curcumin, and PDGFbb from the hydrogel

5.10. Thoracic spinal cord transection surgery and animal care

5.11. Immunofluorescence staining and imaging

5.12. Quantitative reverse transcription polymerase chain reaction (RT-PCR)

5.13. Western blotting

5.14. Evaluating blood vessel permeability at SCI site

5.15. Metabolomics and data analysis

5.16. Quantification of in vivo cytokines

5.17. Macrophage polarization and co-culture with BMECs

5.18. BMEC vascular tube formation assay

5.19. Spinal cord microvascular imaging using ultrafast Ultrasound Localization Microscopy (ULM)

5.20. Bladder ultrasonography assessment

5.21. Hematoxylin-Eosin (HE) staining

5.22. Retrograde trans-synaptic PRV tracing

5.23. Statistical analysis

CRediT authorship contribution statement

Ethics approval and consent to participate

Declaration of competing interest

Acknowledgements

Footnotes

Contributor Information

Appendix A. Supplementary data

References

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles