A cell motility-based selective hydrogel enables rapid generation of nerve-repairing blood clots

Abstract

Blood clots containing nutrients can promote multiple tissue repair, but their use in nerve repair is limited due to the risk of red blood cell-related neurotoxicity. We presented a cell motility-based selective hydrogel for the rapid generation of nerve-repairing blood clots with negligible red blood cell toxicity. This hydrogel, derived from gelatin and featuring a nanocolloidal structure, permitted the migration of neural stem cells (NSCs) while blocking red blood cells, which was mediated by differential cell motility within its nanostructure. Following the rapid generation of blood clots, the hydrogel with blood-derived growth factors promoted the recruitment of endogenous NSCs. The nanocolloidal structure in the hydrogel facilitated the migration and differentiation of NSCs to repair the neural tissue. In rats and porcine models, the hydrogel could induce rapid hemostasis and promote nerve repair in vivo, leading to improved neurological function. This work provides a proof of concept for the generation of nerve-repairing clots using a cell motility-based selective hydrogel, which would inspire future methods for nerve repair.

Graphical abstract

Upon blood contact, the cell motility-based selective hydrogel rapidly generates a nerve-repairing blood clot with minimal RBC entrapment. Subsequently, it recruits endogenous NSCs, promoting their migration and differentiation into neurons to facilitate nerve repair.

A cell motility-based selective hydrogel for rapid generation of nerve-repairing blood clots.

Highlights

• •We designed a nanostructured hydrogel with a cell-selective ability.
• •RBCs and NSCs could be selected due to their differences in motility.
• •The hydrogel rapidly generated nerve-repairing blood clots with low RBC toxicity.
• •Blood clots and the nanostructure synergistically influenced endogenous NSCs.
• •The hydrogel effectively repaired neural tissue and restored neurological function.

1. Introduction

Nerve repair is a major clinical need in the treatment of patients with neurological diseases. Researchers continuously explore innovative therapies to combat neuro-degenerative, ischemic, and traumatic diseases by introducing cells, neurotrophic factors, and genes with regenerative potential into the nervous system [1,2]. Blood clots are derived from blood, function in hemostasis, and are rich in platelets and growth factors that facilitate tissue regeneration [3,4]. Previous studies have demonstrated that blood clots have favorable effects in promoting the regeneration of tissues such as skin and bone [5,6]. Clinically, it is necessary to retain blood clots at the wound site for a period of time. However, blood clots with red blood cells (RBCs) release hemoglobin, iron, peroxiredoxin-2, and carbonic anhydrase-1 upon hemolysis, which have been shown to induce neural tissue damage [[7], [8], [9], [10]]. The retention of a large number of RBCs from blood clots in neural tissue can induce neurotoxicity and edema [11]; therefore, there is an urgent need for new methods to generate blood clots specifically designed for nerve repair.

Bioactive materials provide new approaches for the treatment of various diseases and are now widely used in the fields of tissue regeneration [12]. Pro-regenerative biomaterials have an effect on nerve repair by reducing inflammatory responses, promoting vascular regeneration, and regulating the fate of neural stem cells (NSCs) [[13], [14], [15]]. Current research has utilized biomaterials loaded with nutritional factors to activate and recruit endogenous NSCs for nerve repair [16]. Biomaterials typically form blood clots after hemostasis, which are rich in nutritional components and can be utilized to activate stem cells and promote regeneration. However, RBCs constitute the main component of blood clots and are detrimental to nerve repair. Therefore, there is a need for a cell-selective material that blocks RBCs while facilitating the migration of NSCs to repair nerves. To achieve cell selectivity, previous biomaterials typically employed specific ligands such as peptides and antibodies to enhance the activity of target cells or chemically modify the material surfaces to achieve differential cell adhesion [17,18]. Common physical strategies involve regulating mechanical properties like the internal pore size and material stiffness for selective cell recruitment [19,20]. However, there is currently no strategy for cell sorting based on differences in motility, though such an approach would be non-immunogenic and more cost-effective.

In practical clinical applications, the rapid generation of blood clots via biomaterials for hemostasis and the achievement of postoperative neural repair are equally important. Inadequate hemostasis may lead to severe consequences including hematoma and brain herniation, posing a life-threatening risk to patients [21,22]. Meanwhile, neural injury can cause critical neurological functions (including sensory and motor capacities) of patients to undergo progressive decline [23]. Therefore, new methods capable of simultaneously achieving rapid blood clot generation and leveraging blood clots to promote neural repair are highly anticipated. In this study, we showed a cell motility-based selective hydrogel with a nanocolloidal structure that achieves cell selectivity by leveraging the motility differences between RBCs and NSCs (Fig. 1). By blocking RBCs in the shallow layer and enhancing the migration of NSCs into the interior to differentiate into neural cells, this cell motility-based selective hydrogel achieved rapid generation of nerve-repairing blood clots and subsequent neural tissue regeneration.

Fig. 1.

A cell motility-based selective hydrogel for rapid generation of nerve-repairing blood clots. Upon blood contact, the cell motility-based selective hydrogel rapidly generates a nerve-repairing blood clot with minimal RBC entrapment. Subsequently, it recruits endogenous NSCs, promoting their migration and differentiation into neurons to facilitate nerve repair.

2. Results

2.1. Preparation and characterization

To develop a biomaterial for hemostasis and subsequent nerve repair, we designed a nanostructured gelatin (nano-gelatin) cryogel through the self-assembly and subsequent lyophilization of Gelatin Methacryloyl (GelMA) and Pluronics F68 (termed F68) (Fig. 2A). The GelMA cryogel, porous GelMA cryogel, and the gelatin sponge (commonly used in neurosurgery) were employed as the controls (Fig. 2B, Fig. S1A). The microstructures were characterized by scanning electron microscopy (SEM) (Fig. 2C, D, Fig. S1B). The nano-gelatin exhibited a unique nanostructured network with interconnected micro-nano scale pores, which was significantly different from those of GelMA and porous GelMA. After lyophilization, the nano-gelatin cryogel displayed a network with increased roughness and porosity. Notably, the nano-gelatin cryogel recovered to its original nanostructure in water, showing distinctive differences compared with the controls. The results indicated that the ice crystals formed during the lyophilization process could compress the interconnected nanostructures but not break them. The lyophilization and rehydration of the nanostructured network could be reversible.

Fig. 2.

The preparation and characterization of the nano-gelatin. (A) Flowchart illustrating the preparation of the nano-gelatin and its representative nanostructures. (B) Representative photos of the nano-gelatin cryogel. (C) Representative SEM images showing the microstructures of the nano-gelatin cryogel. (D) Representative SEM images showing the microstructures of the rehydrated nano-gelatin hydrogel. (E) Changes in the water contact angle of the nano-gelatin over 0–0.5 s. (F) The water contact angle of the cryogels. (G) Quantitative analysis of the open-cell content of the cryogels (N = 6). (H) The Zeta potential of the cryogels. (I) Quantitative analysis of the swelling rate of the hydrogels (N = 6). (J) Storage modulus (G′) and loss modulus (G″) of the hydrogels under strain sweep. (K) G′ and G″ of the hydrogels under frequency sweep. (L) Elastic modulus of the hydrogels at a strain of 1% and a frequency of 1 rad/s (N = 6). Statistical analysis was performed using (for F-I) one-way ANOVA followed by Tukey's multiple comparisons test and (for L) an unpaired t-test.

We then characterized the materials in the lyophilized state. In this study, the nano-gelatin cryogel exhibited optimal hydrophilicity and open-cell content, which would be beneficial for rapid blood adsorption (Fig. 2E, F, G; Fig. S1C). Besides, the negative surface potential of the nano-gelatin cryogel was helpful for adsorbing coagulation factors and facilitating effective blood clot formation (Fig. 2H). To evaluate the materials after they transition to a hydrogel state upon implantation, subsequent tests measured hydrogel properties including swelling ratio and viscoelasticity. The nano-gelatin demonstrated significantly lower swelling ratio than that of other materials (Fig. 2I), thereby mitigating brain tissue compression and post-implantation damage. Additionally, rheological testing demonstrated that the nano-gelatin exhibited a storage modulus (G′) of 301-483 Pa, loss modulus (G″) of 26-65 Pa, and elastic modulus of 369 ± 57 Pa (Fig. 2J, K, L). These mechanical parameters closely matched those of native brain tissue, and such biomimetic properties could enhance the recruitment of endogenous NSCs for nerve repair [20]. These results indicated that the nano-gelatin could serve as a scaffold occupying the surgical wound site to promote nerve repair after hemostasis.

2.2. Cell motility-based selective ability

The nanocolloidal structure of the nano-gelatin hydrogel enables it to selectively screen cells based on their motility (Fig. 3A). To explore the cell motility-based selective ability of the nano-gelatin, we conducted experiments on three types of cells. The first was its blocking effect on RBCs. SEM micrographs revealed that RBCs remained passively arrested exclusively on the nano-gelatin surface (Fig. 3B). Then, we investigated S16 cells as the model cell responding to the nano-gelatin hydrogel. Cell migration assays showed that cells were able to infiltrate and migrate from the surface into the nano-gelatin hydrogel. In contrast, a negligible number of cells were observed in the interior of the GelMA hydrogel (Fig. 3C; Fig. S1D). Moreover, time-lapse imaging showed that cells within the nano-gelatin hydrogel could migrate dynamically accompanied by the deformation of the hydrogel nanostructure (Fig. 3D; Fig. S1E), indicating the cell adaptability of the nano-gelatin hydrogel. Next, we examined the migration of NSCs, which play vital roles in nerve repair, within the nano-gelatin hydrogel. Cell migration assays showed that NSCs effectively migrated from the surface to the interior of the nano-gelatin (Fig. 3E), while time-lapse imaging revealed their robust protrusive activity within the nano-gelatin hydrogel (Fig. 3F).

Fig. 3.

Cell motility-based selective ability of the nano-gelatin. (A) Schematic illustration of the motility-based cell selectivity in the nano-gelatin. (B) Representative SEM images showing RBCs resting on the nano-gelatin surface. (C) Representative SEM images showing S16 cells migration from the surface into the interior of the nano-gelatin. (D) Representative time-lapse images showing S16 cells within the nano-gelatin. (E) Representative SEM images showing NSCs migration from the surface into the interior of the nano-gelatin. (F) Representative time-lapse images showing NSCs within the nano-gelatin. (G) Representative time-lapse images via confocal 3D reconstructions showing NSCs, CB-treated NSCs, and RBCs within the nano-gelatin. (H) Time-dependent quantitative penetration depth of NSCs, CB-treated NSCs, and RBCs. (I) Pore size distribution histogram (by Feret diameter) of the nano-gelatin.

To elucidate whether cell motility drives the selective infiltration, we pretreated NSCs with the actin cytoskeleton inhibitor Cytochalasin B (CB, 2 μg/mL). Cytochalasin B is a classic actin cytoskeleton inhibitor that effectively blocks cell migration by disrupting actin polymerization. We seeded NSCs, CB-treated NSCs, and RBCs onto the surface of the nano-gelatin hydrogels. Over time, untreated NSCs actively penetrated deep into the nano-gelatin, reaching a penetration depth of 91.15 ± 4.44 μm after 8 h. In contrast, the motility-deficient, CB-treated NSCs remained confined to the surface (8.48 ± 3.20 μm), a behavior comparable to that of non-motile RBCs (5.09 ± 1.07 μm) (Fig. 3G–H). Finally, we quantified the pore size distribution of the nano-gelatin hydrogel. Given that the Feret diameter of the nano-gelatin pores was 0.94 (0.45–1.92) μm (median, IQR), cellular infiltration is restricted to the surface in the absence of active motility (Fig. 3I). Collectively, these results demonstrated that the nano-gelatin selectively blocked RBCs in the shallow layer while allowing cells such as NSCs to migrate into its interior, indicating its potential to reduce RBC-related toxicity and support subsequent nerve repair functions.

2.3. Evaluation of the hemostatic capacity

To assess the coagulation performance in vitro, a blood coagulation index (BCI) test was performed by measuring the heme content in the supernatant after coagulation (Fig. 4A). A lower BCI value was observed in the nano-gelatin cryogel compared with other cryogels from 0 min to 5 min (Fig. 4B), indicating that the nano-gelatin cryogel could promote blood coagulation. Meanwhile, the clotting time of the nano-gelatin cryogel was less than 5 min, and stable clots were formed 30 min post-treatment with the nano-gelatin. In contrast, the gelatin sponge group still exhibited partially uncoagulated clots within 30 min (Fig. S2A). In vitro coagulation profiles, including PT, APTT, and TT, similarly demonstrated the pro-coagulant activity of the nano-gelatin (Fig. S2B–D). The rapid blood absorption capacity of the nano-gelatin was validated through a 5-s blood uptake test, demonstrating significantly higher absorption than other cryogels (Fig. S2E). Additionally, a hemolysis test revealed that the nano-gelatin had good hemocompatibility (Fig. S2F). Our results suggested that the nano-gelatin cryogel had effective hemostatic properties in vitro.

Fig. 4.

Hemostatic performance of the nano-gelatin. (A) The representative images of a continuous observation of the hemostatic capability of the cryogels. (B) The BCI of the cryogels at different time points in vitro. (C) Representative images showing the hemostatic process with the nano-gelatin in rat models. (D) Quantitative analysis of the blood loss in the rat liver hemorrhage model (N = 6). (E) Quantitative analysis of the blood loss in the rat sagittal sinus hemorrhage model (N = 6). (F) A diagram of the porcine liver hemorrhage model. (G) Representative images of the hemostatic process with the cryogels in the porcine liver hemorrhage model. (H) Quantitative analysis of the blood loss in the porcine liver hemorrhage model (N = 8). (I) A diagram of the porcine sagittal sinus hemorrhage model. (J) Representative images of the hemostatic process with the cryogels in the porcine sagittal sinus hemorrhage model. (K) Statistical analysis of the hemostatic success rate within 1 min in the porcine sagittal sinus hemorrhage model (N = 10). Statistical analysis was performed using (for D-E) one-way ANOVA followed by Tukey's multiple comparisons test, (for H) an unpaired t-test, and (for K) a Fisher's exact test.

The hemostatic ability of the nano-gelatin cryogel in vivo was first investigated in rat liver and superior sagittal sinus hemorrhage models (Fig. 4C). In the rat models, the blood loss of the nano-gelatin cryogel group was significantly reduced compared to the controls (Fig. 4D–E). Next, we verified the hemostatic efficacy of the nano-gelatin cryogel using a porcine model. In liver injury hemostasis, the nano-gelatin group exhibited significantly less blood loss (20.8 ± 11.9 mg) compared to the gelatin sponge group (224.4 ± 83.0 mg), demonstrating excellent in vivo hemostatic effectiveness (Fig. 4F, G, H). More importantly, the nano-gelatin cryogel showed superior hemostatic performance in neurosurgery. It achieved a 100% (10/10) hemostasis rate within 1 min, which was significantly higher than the 20% (2/10) rate observed in the gelatin sponge group (p = 0.0007) (Fig. 4I, J, K; Video. S1). Our findings highlighted the superior hemostatic capability of the nano-gelatin cryogel both in vitro and in vivo.

2.4. Hemostatic mode and mechanism

Nano-gelatin formed a unique shallow hemostatic mode (Fig. 5A). Post-hemostatic appearance revealed that a thin layer of RBCs was observed at the shallow part of the nano-gelatin, while serum seeped into its deeper layer. Further SEM observations revealed that the longitudinal section of the nano-gelatin cryogel exhibited a distinctive three-phase structure following hemostasis. The network without blood adsorption possessed abundant pores, including macropores, micropores, and capillaries, which were conductive to rapid blood absorption, while after hemostasis the network recovered to the nanocolloidal structure with significantly reduced pore size. Concurrently, blood cells accumulated in the shallow layer while serum infiltrated the deep layer. In contrast, gelatin sponges formed full-layer blood clots upon blood absorption, while GelMA and porous GelMA cryogels only developed surface clots (Fig. 5B).

Fig. 5.

Hemostatic mode and mechanism of the nano-gelatin. (A) Schematic illustration of the shallow hemostatic mode elicited by the nano-gelatin. (B) Representative photographs of the cryogels after hemostasis and SEM micrographs of their longitudinal sections. (C) Representative SEM images of the cryogels incubated with blood. Red arrow indicates RBC, green arrow indicates fibrin, and yellow arrow indicates platelet.

To investigate the effects of nano-gelatin on key components during the coagulation process, we employed SEM to examine the morphology of blood cells and fibrin on the materials (Fig. 5C). A substantial presence of blood cells and abundant fibrin on the surface of the nano-gelatin cryogel was observed, indicating a robust proactive hemostatic effect. Notably, the nano-gelatin cryogel exhibited the highest degree of platelet aggregation and activation. Whereas non-activated platelets predominantly maintained a discoid morphology without pseudopodia extension (Fig. S2G), those adhered to the nano-gelatin cryogel displayed prominent pseudopodia, demonstrating the successful platelet-activating functionality within the nano-gelatin.

2.5. Reducing the RBC-related neural injury

To investigate the role of the nano-gelatin in hemostasis and nerve repair, we established a rat model of surgical brain injury (SBI). Brains implanted with different cryogels were harvested, and macroscopic images revealed minimal blood retention in the nano-gelatin group compared to controls (Fig. 6A). Furthermore, H&E staining demonstrated a shallow hemostasis pattern in the nano-gelatin group, characterized by the aggregation of RBCs predominantly occurring within the superficial cryogel matrix. In contrast, porous GelMA and gelatin groups exhibited significant RBC infiltration into the injury site (Fig. 6B and C). Over time, infiltrated RBCs in brain tissue metabolized into hemosiderin—a neurotoxic agent. Correspondingly, histological sections showed substantially fewer hemosiderin deposits in the nano-gelatin group versus controls (Fig. 6D and E).

Fig. 6.

The nano-gelatin mitigates neural injury by blocking RBCs. (A) Representative macroscopic images of brains implanted with different cryogels. (B) Representative H&E staining images of brain sections on day 3. (C) Statistical analysis of the RBC layer thickness (N = 4). (D) Representative H&E staining images showing the hemosiderin (red arrow) deposition at the injured site. (E) Statistical analysis of the hemosiderin deposition (N = 4). (F) Representative MRI of the lesion area on day 3. (G) Representative images of TUNEL and CD68 immunofluorescence staining in brain sections on day 7. (H-I) Quantification of the relative density of TUNEL and CD68 (N = 5). Statistical analysis was performed using one-way ANOVA followed by Tukey's multiple comparisons test.

To further validate the mitigation of neurotoxicity by the nano-gelatin, we first performed whole-brain magnetic resonance imaging (MRI) scans 3 days post-implantation. T2-weighted MRI indicated the minimal cerebral edema in the nano-gelatin group compared to controls (Fig. 6F). Subsequently, on day 7 post-surgery, we assessed apoptosis and inflammation levels using TUNEL assay and CD68 immunofluorescence staining. Compared to the control groups, the nano-gelatin group exhibited the lowest apoptosis level and mildest inflammatory cell infiltration (Fig. 6G, H, I). By day 28 post-surgery, the nano-gelatin group maintained a low inflammation level (Fig. S3). These findings demonstrated that the nano-gelatin restricted RBCs to its superficial layer, reduced RBC infiltration into brain tissue and the subsequent hemosiderin deposition, and mitigated neural tissue injury.

2.6. Facilitating NSCs for nerve repair

To explore the selectivity of the nano-gelatin for NSCs and its subsequent impact on their fate, we first isolated and identified NSCs (Fig. S4), followed by an assessment of the effects and mechanisms of the nano-gelatin on the recruitment, migration, survival, and differentiation of NSCs. Following hemostasis, the nano-gelatin formed nerve-repairing clots. Its shallow layer contained a small number of RBCs, while its deeper layer was enriched with bioactive components derived from blood (Fig. 7A). SEM revealed distinct serum infiltration within the nano-gelatin, which was further confirmed by the detection of albumin (ALB), a major component of serum (Fig. 7B and C). Besides, SEM demonstrated an abundance of platelet-derived extracellular vesicles (PEVs) distributed within the nano-gelatin post-hemostasis. PEVs are released by platelets and carry a wealth of bioactive substances, which can promote tissue repair [24]. Next, we investigated the reservoir capacity for growth factors and chemokines. The results revealed that the negatively charged nano-gelatin could effectively adsorb nerve growth factor (NGF) and stromal cell-derived factor-1 (SDF-1), which may contribute to the recruitment and survival of NSCs (Fig. 7D and E).

Fig. 7.

Effect and mechanism of the nano-gelatin on NSCs. (A) A diagram of the layered composite structure of the nano-gelatin after hemostasis. (B) Representative SEM images showing the platelet-derived extracellular vesicles in the nano-gelatin after hemostasis. (C) ALB contents in the cryogels following hemostasis (N = 4). (D) NGF contents in the cryogels (N = 3). (E) SDF-1 contents in the cryogels (N = 3). (F) Representative images showing NSCs migrating through the Transwell membrane into the plate with cryogels. (G) Quantification of the NSC numbers that migrated through the Transwell membrane into the plate (N = 4). (H) Representative images of the live/dead staining showing the survival and morphology of NSCs on the cryogels. (I) Cytotoxicity of the cryogels on NSCs by CCK-8 assay. (J) Representative images of immunostaining against F-actin, paxillin, and vinculin for cells encapsulated in the nano-gelatin and the GelMA hydrogel. (K) Representative images of immunostaining against Tuj-1 and GFAP. (L) Volcano plot analyzing DEGs between the nano-gelatin group and the control group. (M) The enriched GO pathways. (N) The enriched KEGG pathways. (O) The heatmaps of DEGs associated with Focal adhesion. (P) Schematic diagram of the potential mechanism by which the nano-gelatin regulates NSC migration and differentiation to promote nerve repair. Statistical analysis was performed using one-way ANOVA followed by Tukey's multiple comparisons test.

We further conducted a Transwell migration assay to verify the recruitment effect of the nano-gelatin on NSCs. Compared to the number (99 ± 5) of migrated NSCs in the nano-gelatin group, the numbers in the GelMA, porous GelMA, and gelatin sponge groups were 50 ± 4, 53 ± 5, and 47 ± 3, respectively (Fig. 7F and G). These findings indicate that the nano-gelatin after hemostasis has the potential to attract endogenous NSCs to the injury site, which is a crucial step in nerve repair [25,26]. Subsequently, the live/dead and CCK-8 assays showed that NSCs survived well with a spreading morphology, while the gelatin sponge failed to support the adhesion of NSCs (Fig. 7H and I). Meanwhile, staining for focal adhesion proteins revealed that the nano-gelatin could promote the spreading, adhesion and migration of cells. The areas of paxillin, vinculin and F-actin were increased in the nano-gelatin hydrogels, indicating that the nanostructures could enhance cell-matrix interaction and cell migration (Fig. 7J). Additionally, glial fibrillary acidic protein (GFAP) and neuronal class III β-tubulin (Tuj-1) functioned as markers for astrocytes and neuronal differentiation, respectively. The immunofluorescent results suggested that NSCs in the nano-gelatin group primarily differentiated into neurons, while those in the control group mainly differentiated into astrocytes (Fig. 7K).

To further investigate the underlying mechanisms by which the nano-gelatin influences NSCs, we extracted RNA from NSCs cultured on nano-gelatin for 7 days and performed whole-transcriptome sequencing analysis. This was compared against NSCs cultured under conventional conditions. As shown in the volcano plot, the nano-gelatin group exhibited 2161 upregulated genes and 2035 downregulated genes compared to the control group (Fig. 7L). Subsequently, we conducted Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses on the differentially expressed genes (DEGs) between the nano-gelatin group and the control group. GO enrichment analysis revealed that DEGs were significantly enriched in functional categories related to "cell division" and "positive regulation of cell migration" (Fig. 7M). KEGG enrichment analysis indicated that DEGs were predominantly enriched in signaling pathways including "Focal adhesion", "ECM-receptor interaction", and "MAPK" (Fig. 7N). The DEGs involved in the "Focal adhesion" pathway were subsequently analyzed (Fig. 7O). Based on the aforementioned results, we postulated that nano-gelatin enhances NSCs’ motility via the focal adhesion pathway through its nanocolloidal structure, while concurrently promoting their neuronal differentiation through the MAPK signaling pathway (Fig. 7P). Our results demonstrated that the nano-gelatin hydrogel provided a favorable microenvironment for facilitating NSC recruitment, migration and differentiation for nerve repair.

2.7. Nerve repair efficacy in vivo

Firstly, we evaluated the recovery of neurological function in SBI model rats after the material implantation over a 28-day period. Garcia scores, the percentage of error steps, and balance beam tests were employed as comprehensive methods to assess functional recovery (Fig. 8A). On the first postoperative day, the Garcia score (14.5 ± 1.2) and balance beam score (0.8 ± 0.7) of the nano-gelatin group were lower than those of the normal rats, with a higher incidence of missteps (55.4% ± 7.0%). Over time, the neurological function of rats in the nano-gelatin group gradually recovered. On day 28, the Garcia score of the nano-gelatin group recovered to 18.0 ± 0.8. The balance beam score improved to 4.4 ± 0.7, and the error step rate decreased to 10.6% ± 6.0%, showing significant differences with the other control groups (Fig. 8B). To assess brain injury healing at the 28-day time point, 7T MRI revealed a smaller lesion cavity area in the nano-gelatin group compared to the control groups (Fig. 8C and D). Further analysis was conducted to evaluate neural regeneration at the lesion site. The double fluorescence staining of the neurofilament (NF200) and myelin basic protein (MBP) in the injured area reflected the formation of newborn nerve fibers and their myelination. The results showed that the nano-gelatin group exhibited higher expression of NF200 and MBP compared with the controls (Fig. 8E, F, G).

Fig. 8.

Neurological function evaluation and nerve repair over 28 days. (A) Documentary photographs of Garcia scoring, percentage of foot-fault errors, and balance beam testing sessions. (B) Behavioral assessments in the different groups following surgery within 28 days (N = 8). (C) Representative MRI of the lesion area on day 28. (D) Quantification analysis of the lesion area (N = 4). (E) Representative images of immunofluorescence staining for NF200/MBP in the brain sections. The yellow box was a magnified image of the injury center. (F-G) Quantification of the relative density of MBP and NF200 (N = 5). Statistical analysis was performed using two-way (for B) or one-way (for D, F, G) ANOVA followed by Tukey's multiple comparisons test.

We subsequently investigated the evolution of the interface between the blood clot and neural tissue in different materials. Gross specimen photography of harvested brain samples during the early repair stage showed that the surgical cavities across all groups were completely filled. However, compared to other groups, the aggregation of RBCs around the nano-gelatin was significantly less. By day 28, blood clots around the injury site in the nano-gelatin group had completely resolved, whereas residual clots in the other groups remained undegraded. Additionally, the nano-gelatin hydrogel underwent progressive degradation with low toxicity, creating space for nerve repair at the material-tissue interface (Fig. 9A and G, Fig. S5A–G). Histopathological analysis of H&E-stained brain sections from the lesion sites from day 7 to 28 demonstrated more evident signs of progressive blood clot resolution and nano-gelatin degradation (Fig. 9B and H). We further examined nerve repair processes at the material-tissue interface. Nestin and Doublecortin (DCX) immunofluorescence staining was performed 7 days post-surgery to evaluate early nerve repair. The nano-gelatin group showed significantly higher Nestin expression than the other groups, indicating enhanced endogenous NSC recruitment (Fig. 9C and E). This recruited NSC population was associated with an increase in primary neuron formation. Accordingly, DCX immunofluorescence revealed the highest expression levels in the nano-gelatin group. (Fig. 9D and F). We further evaluated nerve repair at 28 days post-surgery using immunofluorescent labeling with GFAP and NeuN. A positive NeuN stain indicated the formation of mature neurons, and the nano-gelatin group expressed higher NeuN expression levels than the control group (Fig. 9I and K). Additionally, GFAP staining revealed glial scar formation at the injury site, which was not conducive to nerve repair. The results demonstrated that the nano-gelatin group exhibited less glial scar formation than the gelatin sponge group (Fig. 9J and L). Collectively, our findings suggest that the nano-gelatin and the blood clot it generated can degrade over time and promote nerve repair at the material-tissue interface.

Fig. 9.

Evolution of the interface between the blood clot and neural tissue. (A) Representative photographs of the brain morphology during the early repair stage. (B) Representative H&E-stained brain sections on day 7. (C-D) Representative immunofluorescence staining images of Nestin and DCX in brain sections on day 7. (E-F) Quantification of the relative density of Nestin and DCX (N = 5). (G) Representative photographs of the brain morphology during the late repair stage. (H) Representative H&E-stained brain sections on day 28. (I-J) Representative immunofluorescence staining images of NeuN and GFAP in brain sections on day 28. (K-L) Quantification of the relative density of NeuN, and GFAP (N = 5). Statistical analysis was performed using one-way ANOVA followed by Tukey's multiple comparisons test.

3. Discussion

In this study, we engineered a nanostructured cell motility-based selective hydrogel and lyophilized it into a cryogel for neurosurgical hemostasis and subsequent nerve repair. Recently, cryogels have rapidly emerged as attractive platforms for multifaceted biomedical applications, such as rapid hemostasis and tissue regeneration [[27], [28], [29]]. Typically, cryogels are prepared by freeze-drying hydrogels, but during the freezing process, ice crystal growth displaces the polymeric network of rigid hydrogels, creating structural discontinuities. As a result, rehydrated cryogels often struggle to regain their original structure, potentially undermining their therapeutic efficacy. Conventional hydrogels such as nanofibrillated cellulose (NFC) require the addition of lyoprotectants, including sucrose or trehalose, to ensure reconstitution into their original structure after lyophilization [30]. However, the nano-gelatin cryogel developed in this study originates from a nanocolloidal hydrogel featuring a dynamic network composed of interconnected GelMA nanoparticles. During lyophilization, when subjected to localized mechanical stress, these nanocolloidal structures undergo deformation-driven displacement to rapidly accommodate space for growing ice crystals, thereby substantially mitigating structural damage to the hydrogel's original architecture. Upon rehydration, the cryogel permits water permeation throughout its matrix, enabling rapid restoration to its initial hydrogel morphology. This free and reversible transition between the cryogel and hydrogel states provides a new approach for the design of multifunctional cryogels in the future.

The hemostatic modes employed by existing biomaterials prove incompatible with post-surgical nerve repair requirements. The hydrophilic hemostatic material, such as porous gelatin-based matrix, has a highly porous structure that confers exceptional blood absorption capacity [31]. However, RBCs are often trapped within these clots, leading to the accumulation of hemosiderin, which can be toxic to neural tissue [32]. Hydrophobic hemostatic materials typically do not interact extensively with blood components, but they are often removed post-hemostasis and thus provide no sustained contribution to subsequent tissue repair [33,34]. In this study, the nano-gelatin cryogel exhibits a novel shallow hemostatic mode that simultaneously addresses potential neurotoxicity related to RBCs and harnesses blood to promote regeneration. In neurosurgical clinical applications, compared to other commonly used hemostatic materials (such as gelatin sponge, oxidized cellulose-based products, and microfibrillar collagen), the nano-gelatin offers the advantages of rapid hemostasis combined with low RBC-related neurotoxicity [35]. Meanwhile, its low swelling rate, viscoelasticity matching that of brain tissue, and ability to preserve pro-regenerative components in the blood are conducive to subsequent neural tissue regeneration (Table S1). The nano-gelatin and its unique hemostatic mode provide new insights for the development of hemostatic materials in neurosurgery.

Nanostructures offer potential solutions to achieve hemostasis and neural tissue regeneration [[36], [37], [38]]. The previously developed hemostatic materials with nanostructures can interact with components of the coagulation cascade to achieve rapid hemostasis [39,40]. Meanwhile, our nano-gelatin has been confirmed to have an effect on activating platelets, thereby promoting hemostasis. Besides, nanostructures can modulate cellular behaviors via mechanotransduction-mediated mechanisms to facilitate tissue regeneration [41,42]. However, while current nanomaterials can target a single stage of stem cell migration, proliferation, or differentiation [43,44], they fail to comprehensively promote the entire process from stem cell recruitment to final neuronal maturation. To address this limitation, current strategies require the integration of exogenous cells and bioactive molecules [45]. In this study, the nano-gelatin enhances the migration of NSCs and their differentiation into neurons through the synergistic effects of its nanocolloidal structure and absorption of bioactive components in plasma, demonstrating significant potential for nerve repair after hemostasis. Additionally, nanostructures in contemporary biomaterials exhibit post-synthesis static configurations, which fail to dynamically coordinate with RBCs and NSCs, thereby failing to meet the requirements for reducing RBC-related toxicity and promoting neural regeneration. Our developed nano-gelatin, featuring a dynamic nanocolloidal structure, achieves cell selectivity by leveraging the motility differences between different cells. RBCs rely on hemodynamic forces and have poor active motility, thus being trapped in the shallow layer of the nano-gelatin. In contrast, NSCs possess intrinsic active motility and are further enhanced by the nanocolloidal structure, enabling them to penetrate into a deeper layer. The dynamic nanocolloidal structure addresses the dual requirements of rapid hemostasis and enhanced neural tissue regeneration through its interactions with cells such as platelets, RBCs, and NSCs.

Current clinical techniques can separate plasma and utilize it for tissue regeneration [46,47]. In our study, the nano-gelatin with a nerve-repairing clot enriches the plasma during hemostasis while reducing the entry of RBCs, thereby significantly reducing neurotoxicity. The sequestered plasma enriches endogenous growth factors, creating a pro-regenerative niche that orchestrates neurogenesis and axonal remodeling. When applied in surgical procedures, the nano-gelatin can be implanted into the surgical cavity post-operation to achieve effective hemorrhage control while harnessing blood-derived factors for neural tissue regeneration, ultimately improving the patient's neurological function. The cell motility-based selective hydrogel with the nerve-repairing clots presents a promising method for achieving surgical hemostasis and subsequent nerve repair.

This study still has several limitations that merit further studies. First, controlled pressure-driven perfusion experiments could be conducted to further elucidate the hemostatic and cell-selective mechanism of the nano-gelatin. Second, while this study demonstrated that the nano-gelatin can regulate the function of NSCs, techniques such as lineage and axonal tracing are still needed to further verify the NSC-derived neuronal and functional integration. Finally, for future clinical translation, combining the nano-gelatin with thrombin could be considered to enhance hemostatic efficacy, and to be compared with the gelatin-thrombin matrix commonly used in neurosurgery.

4. Conclusions

This work demonstrates a cell motility-based selective hydrogel for the rapid generation of nerve-repairing blood clots. Through its nanocolloidal structure, the hydrogel blocks RBCs at the shallow layer to form blood clots with low neurotoxicity while simultaneously recruiting endogenous NSCs to migrate into the hydrogel and differentiate into neural cells. The hydrogel effectively induces rapid hemostasis and promotes nerve repair in vivo, leading to significant improvements in neurological function. This cell motility-based selective hydrogel holds potential for surgical applications and will inspire the development of future biomaterials.

5. Materials and methods

5.1. Preparation and characterization of GelMA

GelMA was synthesized using a reported methodology [48]. Type A gelatin (Sigma, USA) was dissolved in sodium carbonate buffer (0.25 M, pH 9.0) and was reacted with methacrylic anhydride (MAA, Sigma, USA). Following dialysis, the obtained GelMA monomer was lyophilized and stored at −20 °C in a dry environment for future use. The substitution degree of GelMA was quantified by the trinitrobenzene sulfonic acid (TNBS, Xiya reagent, C1056) assay using a UV spectrophotometer (UV-2300, Shimadzu Corp., Japan). Additionally, fluorescein isothiocyanate (FITC) or rhodamine-labeled GelMA was synthesized as described in our previous report [49].

5.2. Preparation of the nano-gelatin

The nano-gelatin hydrogel was prepared according to the previous report [49]. GelMA was dissolved in PBS solution containing 0.75% (w/v) lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) to obtain a 15% (w/v) solution. Then, this solution was mixed with a 20% (w/v) F68 solution at a volume ratio of 2:1 (V_{GelMA solution}: V_{F68 solution} = 2:1) to generate the prepolymer solution. Meanwhile, GelMA and porous GelMA hydrogels were prepared and used as controls. The GelMA solution was mixed with PBS or polyethylene oxide (PEO, 1.5% w/v) solution at a volume ratio of 2:1 to prepare GelMA and porous GelMA prepolymer solutions, respectively. The obtained solutions were exposed to UV light (130 mW cm⁻² for 30 s) to generate hydrogels. The hydrogels were immersed in distilled water, and the cryogels were obtained by further lyophilization for 36 h. Additionally, commercial gelatin sponge (FUKANGSEN, Guilin, China) commonly used in neurosurgical practice was included as another control group.

5.3. Characterization of the nano-gelatin

The microstructures of the hydrogels and cryogels were observed by SEM (JSM7500F, JEOL) at a voltage of 15.0 kV. The open-cell content of the cryogels was measured using an Automatic Open/Closed-Cell Content Analyzer (UltraFoam 1200e, Anton Paar Quantachrome, USA). The water contact angles of the cryogels were measured with a sessile drop method by an optical contact angle goniometer (DSA25E, Kruss, Germany). Briefly, a distilled water droplet was dropped onto the surface of the cryogels, the contact angle images were recorded, and the static contact angles were measured within 0.5 s. The Zeta potential of the cryogels was measured using a Malvern Zetasizer (Nano-ZS, Malvern Instruments, Malvern, U.K.).

The cryogels were homogenized and then dispersed in deionized water to obtain the samples. Meanwhile, the cryogels were immersed in deionized water for 1 min to record the initial wet weight. Then, the samples were immersed in deionized water for 12 h for swelling equilibrium, and the final wet weight was recorded. The swelling ratio was calculated using the following formula (Wt is the final wet weight, and Wd is the initial wet weight):

$$\text{Swelling}\text{rate}(\%)=\frac{Wt-Wd}{Wd}\times 100\%$$

A rheological test was performed using a rheometer (MCR 302, Anton Paar) equipped with parallel plates (25 mm diameter, 1.0 mm gap). Hydrogels were tested in oscillatory mode at 25 °C. An amplitude sweep test within a shear strain range of 0.01-100% was performed at a constant frequency of 10 rad/s. Then, frequency sweep test was carried out at a constant shear strain of 0.1% within 1-100 rad/s.

The degradation rate experiment was conducted as follows: nano-gelatin cryogels were p laced in a freshly prepared collagenase I solution (0.4 mg/mL) and incubated with shaking at 37 °C. The collagenase solution was replaced every 2 days. At predetermined time points, the samples were washed with deionized water, freeze-dried for 24 h, and then weighed. The degradation rate was calculated using the formula (Wt is the weight at each time point, and W0 is the initial weight):

$$\mathrm{D}\text{egradation}\text{rate}(\%)=\frac{Wt-W0}{W0}\times 100\%$$

5.4. In vitro hemostatic evaluation

Whole blood was collected from the abdominal aorta of rats into sodium citrate anticoagulant tubes. Blood (9 mL) was mixed with CaCl₂ (1 mL, 0.1 M). Then, 100 μL of blood was added onto the surface of the cryogel. Subsequently, 1 mL of deionized water was added to dissolve the uncoagulated blood, and the solutions were collected at the designed time points (0, 3, 5 min). The optical density (OD) values at 540 nm were determined using a microplate reader (Bio-Rad iMark, California, USA). The blood clotting index (BCI) was calculated according to the following formula:

$$\text{BCI}(\%)=\frac{ODofTestSample}{ODofBlankControl}\times 100\%$$

Meanwhile, 100 μL of recalcified blood was added into 1.5 mL tubes containing cryogels, and 1 mL of deionized water was added at the predesigned time points. The tubes were then inverted, and images were acquired to analyze the blood clot time, which was defined as the time without blood flow.

Hemostatic materials were placed into centrifuge tube respectively, followed by the addition of 1 mL normal saline. Additionally, centrifuge tubes without cryogels served as a negative control followed by addition of 1 mL of normal saline, as well as a positive control followed by addition of 1 mL deionized water. Subsequently, 200 μL of blood was added to each tube, and the samples were incubated at 37 °C for 1 h. After centrifugation, the OD values of the supernatant at 540 nm were measured. The hemolysis rate was calculated as follow:

$$Hemolysisrate=\frac{ODofTestSample-ODofNegativeControl}{ODofPositiveControl-ODofNegativeControl}\times 100\%$$

To investigate the platelet activation capacity of hemostatic materials, the cryogels were incubated with platelets. The platelets were obtained from the peripheral blood of SD rats using a platelet isolation kit. Meanwhile, cryogels used for hemostasis in liver bleeding were harvested to assess their blood cell adsorption performance. Then, the samples were processed for fixation, gradient dehydration, and supercritical drying. Finally, the metal-spraying samples were observed by SEM (JSM7500F, JEOL).

5.5. Cell culture

The primary NSCs isolation was performed according to reported methodology [50]. In brief, pregnant rats at 14-16 days of gestation were anesthetized with isoflurane before aseptic dissection to retrieve embryos. Subsequently, the cerebral cortex of the fetal rats was carefully excised, washed, and gently triturated. The suspension was filtered through a 200-mesh cell strainer. After centrifugation at 1500 rpm for 5 min, the pelleted cells were resuspended in NSC culture medium (DMEM containing 20 ng/mL bFGF, 20 ng/mL EGF, and 2% B27, NeuroCult) at a density of 1 × 10⁶ cells/mL. Cells were then seeded into T25 flasks and cultured in a humidified and 5% CO₂ incubator.

5.6. Identification of NSCs

Neurospheres at the third passage were enzymatically dissociated using Accutase (Yeasen, 40506ES60), and the cell suspension was seeded onto the poly-L-lysine-coated slides and incubated until cells began to extend neurites. Then, the samples were gently rinsed and fixed with 4% paraformaldehyde. Immunofluorescence staining with Anti-Nestin antibody (1:1000, Abcam, ab221660) was performed to identify NSCs.

5.7. Motility-based cell selectivity evaluation

Cell infiltration into the hydrogels was investigated with the 24-well Transwell plate. 50 μL prepolymer solutions were added in the upper chamber and photo-crosslinked to form hydrogels, and cells (1.5 × 10⁴) were seeded onto the hydrogels. Then, 600 μL DMEM medium containing NGF (50 ng/mL, Solarbio, P00056) was added into the lower chamber. Finally, the hydrogels were fixed and longitudinally fractured in liquid nitrogen for SEM observation (JSM7500F, JEOL).

Time-lapse imaging was performed to investigate cell spreading behavior within the nano-gelation hydrogels. FITC labeled GelMA was used to regenerate GelMA nanodroplets and encapsulate Dil (Beyotime Biotechnology, C1991S) labeled cells at a density of 3 × 10⁶ cells/mL. After culture for 12 h, the samples were observed by a microscope (Nikon N-STORM, Japan).

Time-lapse 3D confocal reconstruction was employed to monitor cell migration behavior. Dil-labeled RBCs, Dio-labeled (Beyotime Biotechnology, C1993S) NSCs, and Cytochalasin B (Yeasen, 53373ES03) treated NSCs were seeded onto the surface of the nano-gelatin, to quantify the penetration depth over time.

To investigate cell adhesion within hydrogels, cells were encapsulated in rhodamine B labeled hydrogels. After culture for 48 h, the samples were fixed and stained with anti-paxillin antibody (1:100, Abcam, ab32115) or anti-vinculin antibody (1:100, Abcam, ab129002), followed by anti-rabbit IgG (H + L) Alexa Fluor 647 (1:500, Abcam, ab150079), Alexa Fluor 488 Phalloidin (1:400, Invitrogen, A12379) or anti-rabbit IgG (H + L) Alexa Fluor 488 (1:50, Zsbio, ZF-0311), and DAPI (1 μg/mL, Invitrogen). Images were acquired by confocal laser scanning microscope.

5.8. The protein adsorption performances

The ALB adsorption of cryogels following hemostasis was determined by a blood chemistry analyzer. The cryogels after hemostasis were homogenized and dispersed in normal saline. After centrifugation, the supernatants were obtained as the samples. To investigate the growth factor and chemokines adsorption, NGF and SDF-1 solution (2.0 ng/mL) were incubated with the cryogels instead of blood, thereby avoiding the interference of the blood components for determination. After incubation for 0.5 h, the NGF and SDF-1 concentration remained in the solution was assayed by ELISA to analyze the NGF and SDF-1 adsorption of cryogels.

5.9. Cell recruitment

Cell migration assays were performed to investigate the cell recruitment potential of the cryogels following hemostasis in a 24-well Transwell plate. NSCs were seeded onto the top of the Transwell membrane at a density of 1 × 10⁷ cells/mL. On day 5, the chambers were fixed with paraformaldehyde and stained with crystal violet. The samples were observed by a microscope (CKX31, Olympus).

5.10. Cytotoxicity evaluation

NSCs were seeded onto the sterile cryogels and cultured in DMEM/F12 medium (Gibco, Grand Island, NY, USA) supplemented with serum. The cell viability was evaluated on days 1, 3, and 5 post-culture using a live/dead staining kit (Beyotime Biotechnology, C2015M). The samples were observed by a cell imaging system (Delta Vision Ultra, GE Healthcare). The hemostatic cryogels were immersed in DMEM/F12 medium supplemented with serum at a concentration of 30 mg/mL. After 24 h of incubation at 37 °C with shaking, extract samples of each group were obtained. NSCs were seeded into a 96-well plate at a density of 4000 cells per well, and then cultured with the obtained extract samples. The DMEM/F12 medium was used as a control. Cell viabilities on days 1, 2, and 3 were determined by CCK-8 (MeilunBio, MA0218) following the manufacturer's instruction.

5.11. Cell differentiation assay

Neurospheres at the third passage were dissociated into single cells by using Accutase and resuspended in serum-containing medium at a density of 1 × 10⁷ cells/mL. Cells were seeded onto the surface of cryogels and cultured for 7 days. Then, the samples were fixed and stained with astrocyte marker GFAP (1:500, CST, Rabbit mAb #80788) and neuron marker Tuj-1 (1:200, HUABIO, SP06-00) to assess differentiation.

5.12. RNA transcriptome analysis

In transcriptome sequencing analysis, a total RNA isolation kit (Foregene, RE-03011) was used for RNA extraction from NSCs cultured within the nano-gelatin hydrogel at day7. The extracted total RNA was subsequently sent to Annoroad Gene Technology Co., Ltd. (Beijing) for RNA sequencing. In the sequencing data analysis, DEGs were determined by |log2Fold Change| ≧ 1 and Q-value ≤0.05. The GO analysis and KEGG pathway analysis were performed by Annoroad Gene Technology Co., Ltd.

5.13. Animal experiments

All animals were approved (No. 20230803001) and monitored by the Animal Welfare and Ethics Committee of West China Hospital, Sichuan University. Animals were procured from Beijing Huafukang Bioscience Co., Ltd. The animals were housed in a controlled environment with regulated humidity and temperature for at least 7 days prior to surgery. Randomization was using the online tool QuickCalcs provided by GraphPad Prism 9.0 software, employing the''Random Number'' function to code and allocate animals randomly into four groups (treated with nano-gelatin, GelMA, Porous GelMA, and gelatin sponge groups). The experiments and analyses were conducted blinded to group allocation. Animals that failed to meet the criteria for successful model establishment or died during the surgical procedure would be excluded. The animals were anesthetized via isoflurane inhalation (RWD Life Science Corp, China). The pain relief after surgery was provided with MediGel CPF (Moldiets, China). Vital parameters were closely monitored and documented throughout surgery and recovery. Three experimental models were involved in this study: liver incision injury, superior sagittal sinus injury and SBI [[51], [52], [53]]. The first two models were used to assess the hemostatic efficiency, while the last one was used to evaluate the nerve repair.

For liver incision injury, the liver of the rats was first fully exposed. A pre-weighed filter paper was placed beneath the liver, and a 1 cm² wound was made 1 cm below the lower edge of the liver. Hemostatic cryogels were applied for 5 min, and blood loss was calculated by the weight difference of the filter paper before and after hemostasis.

For superior sagittal sinus injury, a drill was used to create a 3 mm diameter circular opening at the intersection of the sagittal and coronal sutures in each rat, exposing the dura mater. The superior sagittal sinus was carefully isolated, and after being transected with a limbal knife, the wound was immediately covered with hemostatic cryogels and pressed with gauzes for 2 min. The difference in gauze weights before and after application was used to evaluate the hemostatic performance of cryogels.

For SBI, the right frontal bone was first fully exposed. A craniotomy was performed to create a 5 mm² bone window, followed by an incision of the dura mater. A partial resection of the right frontal lobe was then carried out 1 mm anterior to the coronal suture and 1 mm lateral to the sagittal suture. Hemostatic cryogels were placed into the cavity after resection, and the wound was sutured after achieving complete hemostasis.

The surgical hemostatic procedure for porcine is detailed in Supplementary Video 1.

Supplementary data related to this article can be found online at https://doi.org/10.1016/j.bioactmat.2026.05.015

The following are the Supplementary data related to this article.

5.14. Neurological function assessment

The Modified Garcia Test (mGarcia) [54], Balance Beam Test [55], and Foot Fault Test [56] were employed to evaluate neurological function. Pre-operatively, animals underwent training sessions for a week to gradually accustom them to the various activities involved in behavioral scoring. The specific tests comprised the following:

The Modified Garcia Test was performed to assess sensory-motor deficits by evaluating cage activity, bilateral body sensation, whisker tactile response, symmetry of hindlimb movement, turning ability, forepaw extension, and climbing. The total score ranges from 3 to 21, in which a higher score indicates better performance in neurological function.

In the Balance Beam Test, each animal's score ranged from 0 to 5, depending on the distance traveled and the time taken to traverse the beam. Higher scores reflect better performance.

In the Foot Fault Test, animals were placed on a horizontal elevated ladder to assess their motor function, with a recording made of their 2-min walk on the ladder. The total number of steps taken by each limb and instances of paw misplacement were counted.

5.15. Magnetic resonance imaging (MRI)

Rats were anesthetized with isoflurane and subjected to MRI using a 7 T small animal MRI system (Bruker BioSpin 7T/20 USR, Germany) on day 3 and day 28 postoperative. Once anesthetized, the rats were positioned in a prone orientation on a dedicated fixation system tailored for the MRI procedure. T2-weighted scans were employed to visualize brain edema and hemorrhage, with scan parameters set as follows: matrix size of 256 × 256, slice thickness of 0.8 mm, inter-slice gap of 1 mm, echo time (TE)/repetition time (TR) of 33/2500 ms, a rapid acquisition with relaxation enhancement (RARE) factor of 8, and a flip angle of 90°. T2-weighted images were acquired in both sagittal and axial planes using ParaVision 6.0.1 software.

5.16. Histological analysis

Rats were euthanized at 1, 3, 7, 14 and 28days post-surgery. All samples retrieved from rats were embedded in paraffin and sectioned coronally into slices of 5 μm thickness. These sections were stained with Hematoxylin and Eosin (H&E) to examine the state of damaged cavities, the presence of hemostatic materials, and the degree of neural tissue regeneration.

For immunofluorescence analyses, dehydrated samples were embedded in optimal cutting temperature (OCT, Epreida, Neg-50) compound and sectioned coronally into 10 μm thick slices. On postoperative day 7, immunofluorescence analyses were conducted using antibodies against Nestin (1:100, Abcam, ab221660), Doublecortin (DCX; 1:1000, Oasis biofarm, OB-PGP019), and CD68 (1:200, Abcam, ab31630). On day 28, additional antibodies were employed, including myelin basic protein (MBP; 1:1000, Abcam, ab7349), anti-200 kDa subunit of neurofilament (NF200,1:500, Abcam, ab207176), glial fibrillary acidic protein (GFAP; 1:200, CST, Rabbit mAb #80788), Anti-NeuN (NeuN; 1:200, Abcam, ab177487) and CD68. The protocol entailed pre-incubating tissue samples with goat serum working solution at 37 °C for 2 h to minimize non-specific interactions. A dual immunostaining method was adopted, involving overnight incubation of the sections with primary antibodies at 4 °C. Following the removal of primary antibodies and subsequent washes, secondary antibodies were applied. Each secondary antibody was tagged with a distinct fluorophore and incubated with the samples at 37 °C for 2 h. Finally, slides were stained with 4′,6-diamidino-2-phenylindole (DAPI) for nuclear visualization before mounting for microscopy examination.

To assess the apoptosis level of brain tissue at postoperative day 7, a TUNEL assay kit (BIOSSCI, #BBL-0117) was utilized. Briefly, cryosections of the tissue were gently washed with phosphate buffered saline (PBS) for 5 min, after which they were blocked for 30 min in a solution containing 1% donkey serum and 0.3% Triton X-100 (Solarbio, Beijing, China). The TUNEL detection solution was then prepared and cautiously applied to the samples, which were subsequently incubated in a dark environment at 37 °C for 2 h. Fluorescence images were acquired using a Nikon confocal laser scanning microscopy (Nikon, Japan), and Image-Pro Plus software (Media Cybernetics) was employed for quantitative analysis.

5.17. Statistical analysis

All experiments were repeated at least three times, and the results were subjected to statistical analyses. Data are presented as the mean ± standard deviation. Data analysis and graphing were performed using R (Version 4.5.1, The R Foundation for Statistical Computing) and GraphPad Prism 9 software. Comparisons between two groups were conducted using unpaired t-tests, while multiple-group comparisons were conducted via one-way analysis of variance (ANOVA) test and post-hoc Tukey's test. Behavioral statistical analyses were performed using two-way ANOVA. For comparisons involving categorical variables, Fisher's exact test was used. Specific statistical tests are detailed in the Results section and figure legends. Differences are considered statistically significant if P < 0.05.

CRediT authorship contribution statement

Wenbo He: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis. Yi Zhang: Writing – review & editing, Methodology, Investigation, Conceptualization. Wenbi Wu: Writing – review & editing, Investigation, Formal analysis. Datong Zheng: Methodology, Investigation, Formal analysis. Ming Peng: Methodology, Investigation. Qi Zhu: Methodology, Investigation. Li Li: Methodology, Investigation. Yongchao Zhao: Methodology, Investigation. Yinchu Dong: Methodology. Boya Li: Writing – review & editing. Haofan Liu: Investigation. Shuai Yang: Methodology. Xue Zhang: Investigation. Wentao Li: Investigation. Liansha Tang: Validation. Ludwig Cardon: Methodology. Mariya Edeleva: Methodology. Jianguo Xu: Supervision, Methodology. Yu Hu: Supervision, Funding acquisition, Conceptualization. Maling Gou: Writing – review & editing, Supervision, Funding acquisition, Conceptualization.

Ethics approval and consent to participate

This study and included experimental procedures were approved by the Animal Welfare and Ethics Committee of West China Hospital, Sichuan University (Ethics Approval No. 20230803001). All animal housing and experiments were conducted in strict accordance with the institutional guidelines for care and use of laboratory animals.

Declaration of competing interest

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

Appendix A. Supplementary data

The following are the Supplementary data to this article.

Data availability

Data will be made available on request.