Research on GM1 bound bMSCs loaded with SF hydrogel for spinal cord injury repair Authors: Qin Yu, Hai-Yan Zhang, Yan Song, Ling Wang, Xiao-Yong Zhang, Yang-Jun Xu, Chen Chen Published: 2026 ABSTRACT AbstractBackgroundMonosialotetrahexosylganglioside (GM1), a major ganglioside component, is expressed in the spinal cords of mammals. Gangliosides have been shown to play important roles in various physiological processes, including neuronal protection, neurorepair, and stem cell differentiation. Bone marrow-derived mesenchymal stem cells (bMSCs) were expected to be a good source of cells for clinical transplantation and repair due to their abundance, easy accessibility and lack of ethical restrictions. They can also be induced to differentiate into neurons. However, there is still no ideal delivery material or method that allows the sustained and controlled release of GM1 and bMSCs for clinical use in spinal cord injury (SCI). Silk fibroin (SF)-based hydrogels have emerged as a promising strategy for local, controlled, and sustained drug release in the treatment of various diseases.MethodsIn this study, we developed a SF-based hydrogel for the sustained delivery of GM1 and bMSCs (GM1 + bMSCs@SF hydrogel) and evaluated its therapeutic effect in a rat model of SCI.ResultsThe GM1 + bMSCs@SF hydrogel showed sustained GM1 release in vitro. Treatment with the GM1 + bMSCs@SF hydrogel increased cell cycle exit and neuronal differentiation 1 (Cend1) expression, reduced neuronal pyroptosis and apoptosis, and promoted the differentiation of bMSCs into neurons. Double immunofluorescence staining confirmed a decrease in cell pyroptosis. Moreover, rats treated with the GM1 + bMSCs@SF hydrogel showed significant improvement in motor function recovery, with the Basso, Beattie and Bresnahan (BBB) score at 8 weeks reaching about 71.4% of the normal level, compared with about 30% in the SCI group.ConclusionsThis study will minimize neuronal damage by innovatively using SF hydrogel loaded with GM1 and bMSCs for spinal cord injury repair. It will also explore the molecular mechanisms underlying post-traumatic spinal cord injury repair, opening new therapeutic avenues for spinal cord injury treatment.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12967-026-08387-5.Keywords: Spinal cord injury, GM1 + bMSCs @SF hydrogel, Cend1, Pyroptosis, Differentiation Introduction Spinal cord injury (SCI) is one of the most complex medical conditions and remains a major social and economic burden worldwide [1]. The main causes of SCI are traffic accidents, falls from heights, and violent trauma [2]. The estimated lifetime cost of a single traumatic SCI case ranges from 500,000 to 2 million USD [3, 4]. Each year, 250,000 to 500,000 new cases occur globally according to the World Health Organization (WHO), with an incidence rate of about 40.1 cases per million people. The condition is more common in males than in females [5]. SCI usually leads to lifelong disability and is characterised by two stages of injury, namely the primary and secondary phases [2, 6]. The primary injury results from direct mechanical trauma to the spinal cord, leading to tissue contusion, compression, hemorrhage, or even transection. These events disrupt cell membrane integrity, cause ionic and neurotransmitter imbalances, and lead to demyelination and rupture of sensory and motor nerve fibers. As a result, neurons in the affected region undergo rapid death, causing an immediate loss of neurological function [7]. Initial mechanical damage also compromises the blood-spinal cord barrier. This results in haemorrhaging, ischaemia, microvascular loss, oedema, inflammation, oxidative stress, lipid peroxidation and the release of toxic compounds from injured neurons and glial cells [8]. These processes trigger neuroinflammation. Proinflammatory cytokines induce different types of programmed cell death, including apoptosis, pyroptosis, and necroptosis. Activated microglia and infiltrating immune cells such as lymphocytes, monocytes, and neutrophils amplify the inflammatory response. The release of cytokines and chemokines further promotes neural stem cell differentiation into astrocytes, leading to glial scar formation, demyelination, and excessive local fibrosis [9]. In the early stage, glial scars have protective functions. They restrict lesion expansion, contain inflammation, and prevent the spread of toxins to nearby tissues [10]. However, glial scars pose a significant barrier to regeneration at a later stage. In contrast to peripheral nerve injury, scars in the central nervous system persist and inhibit axonal regrowth through physical and chemical means. One example of this is the presence of chondroitin sulfate proteoglycans (CSPGs). They also alter the microenvironment by suppressing neurotrophic factors and limiting the repair of neurons, astrocytes, oligodendrocytes, and their precursor cells [11]. Although decades of medical progress have greatly improved the survival of patients with SCI, there is still no clinically effective treatment that can restore neural conductivity [12]. This highlights the urgent need for more effective therapeutic strategies. With the rapid progress in stem cell research, stem cell transplantation has gained attention as a promising therapy for SCI repair. After SCI, reconstructing neural pathways and restoring motor function remain major medical challenges. The key difficulties involve reducing cell death after injury and replacing lost cells. These goals can be achieved either by activating endogenous stem cells or by transplanting exogenous stem cells capable of differentiating into functional neural cells. Many types of stem cells are currently being studied in animal models and have entered clinical trials. Embryonic stem cells, which are derived from the inner cell mass of blastocysts and possess pluripotent differentiation potential across the three germ layers, theoretically represent an ideal cell source for transplantation therapy. However, their clinical use is limited by several concerns, including the risk of tumor formation, genetic instability, and ethical controversies. In contrast, bone marrow-derived mesenchymal stem cells (bMSCs) offer clear advantages for clinical transplantation and neural repair. They can be obtained through minimally invasive procedures, are available in abundance, have no ethical restrictions, and possess the ability to differentiate into multiple cell types [13, 14]. Although stem cell transplantation can supply a large number of new cells, it is difficult to reproduce the complex structure and stability of the spinal cord. Moreover, the survival rate of transplanted cells is low in the harsh microenvironment of the injured area, and they often fail to guide axons to grow in the correct direction. Hydrogels are hydrophilic three-dimensional polymer networks that can be classified as natural or synthetic polymer hydrogels according to their material composition. Natural polymer hydrogels have attracted increasing research interest because of their excellent biocompatibility, wide availability, and low cost. In recent years, hydrogels have been widely used in neural tissue repair and regeneration due to their controllable biodegradation properties and high biocompatibility [15]. Silk fibroin (SF), a natural protein polymer approved by the Food and Drug Administration (FDA), shows unique advantages in biomedical applications. It offers tunable mechanical properties, good biocompatibility, precise process control, and adjustable biodegradation rates. SF can be fabricated into different forms such as foams, films, fibers, powders, hydrogels, sponges, and scaffolds [16]. In particular, SF hydrogels are highly useful as injectable scaffolds for cell delivery because they rapidly recover stiffness after injection, which helps maintain structural integrity. They also serve as efficient drug delivery systems that stabilize therapeutic compounds and prolong tissue retention through controlled release [17]. Biomaterials can physically bridge the gap between the two ends of the injured spinal cord, but they cannot replace the cells that were lost during injury [18, 19]. Therefore, the combination of tissue engineering techniques, biomaterials and stem cells has emerged as a promising strategy for promoting nerve regeneration after SCI. However, the limited capacity of neural stem cells and other cell types to develop into functional neurons and oligodendrocytes remains a significant challenge. At present, incorporating drugs that induce directional differentiation into biomaterials containing stem cells may be a more effective approach [20]. Monosialotetrahexosylganglioside (GM1) is widely distributed in the mammalian central nervous system and is an important structural component of cell membrane domains. GM1 regulates multiple cellular processes by enhancing the effects of neurotrophic factors to influence cell differentiation, suppressing excessive activation of excitatory amino acid receptors, and reducing cytotoxicity to protect against acute neural injury. It also plays important roles in cell proliferation, neuronal differentiation, and signal transduction [21, 22]. Based on these findings, our study aims to develop a novel three-component biomaterial, an SF-based hydrogel for the sustained delivery of GM1 and bMSCs (GM1 + bMSCs@SF hydrogel). This biomaterial is intended for local intrathecal administration to enable continuous GM1 release and precise regulation of stem cell differentiation, thereby supporting neural repair and promoting functional recovery after spinal cord injury. Materials and methods Preparation of injectable GM1@SF hydrogel Silk degumming Raw silk fibers were immersed in 0.05% (w/v) sodium carbonate (Na₂CO₃) solution at a ratio of 1:50 (w/v). The degumming process was carried out in a boiling water bath for 30 min. The fibers were then rinsed three times with deionized water, 5 min each time, to completely remove residual sericin [23, 24]. SF extraction The degummed SFs were dissolved in 9.3 M lithium bromide (LiBr) solution at a ratio of 1:4 (w/v) and stirred constantly at 60 °C for 4 h. The solution was dialyzed using a 3 kDa molecular weight cut-off (MWCO) membrane against deionized water for 72 h, with water replaced every 8 h. After dialysis, the solution was centrifuged at 9000×g for 15 min and filtered through a 0.22 μm membrane to obtain a clear SF solution. Adjustment of SF solution The SF solution was diluted to 0.75% (w/v), and the pH was adjusted to 9.5 using 0.5 M sodium hydroxide (NaOH). Anhydrous ethanol was then added slowly to a final concentration of 7% (v/v) under constant stirring [25]. Hydrogel formation The prepared mixture was incubated at 25 ± 2 °C for 7 days to allow self-assembly of SF nanofibers. After complete gelation, the hydrogel was dialyzed using a 3 kDa MWCO membrane against deionized water for 24 h, with the water changed every 4 h, to remove ethanol residues. Preparation of injectable SF hydrogel The purified silk nanofiber hydrogel was mixed with 50 mM sodium chloride (NaCl) solution at a ratio of 1:9 (v/v) and centrifuged at 8000×g for 10 min to obtain an injectable SF hydrogel with a final concentration of approximately 1.5% (w/v). Fabrication of GM1@SF composite hydrogel GM1 (TargetMol, 37758-47-7, USA) was dissolved in normal saline to prepare a 5 mg/ml stock solution and stirred magnetically for 30 min. The GM1 solution was then blended with the 1.5% (w/v) SF hydrogel at predetermined ratios and homogenized using a vortex mixer at 2000 rpm for 2 min to obtain the injectable GM1@SF composite hydrogel. Characterization of GM1@SF hydrogel Morphological observation The surface morphology, pore interconnectivity, and internal structure of the hydrogels were examined using field emission scanning electron microscopy (SEM, Hitachi S-4800, Tokyo, Japan). Representative samples from each group (SF and GM1@SF) were selected for imaging. Before observation, the samples were coated with a thin layer of gold, and images were captured at an accelerating voltage of 30 kV. Raman spectroscopy Raman spectroscopy was performed using an XploRA laser Raman spectrometer (HORIBA, France). The spectra were collected using a grating of 1200, an excitation wavelength of 785 nm, and a laser power of 90 mW. Each spectrum was acquired with an accumulation time of 10 s and scanned 10 times to improve signal quality. Rheological analysis The rheological properties of the SF and GM1@SF hydrogels were measured using a rotational rheometer (Gemini HR nano 200, Malvern, UK). A parallel plate geometry with a diameter of 20 mm and a gap of 1 mm was used. The viscoelastic behavior was evaluated by performing an oscillatory frequency sweep from 0.1 Hz to 10 Hz under a constant strain of 1.5% at 25 °C. Evaluation of physical state The physical state of the samples, including SF solution, SF gel, and GM1@SF hydrogel, was determined using the vial tilting method. One milliliter of each sample was placed in a conical vial and tilted at a 45° angle. The sample was considered to have formed a gel when the meniscus remained unchanged upon tilting to a 90° angle. The self-supporting property of the gels was further confirmed by inverting the tubes to observe whether the gels maintained their shape. Drug release assay and GM1 release profile Lyophilized SF hydrogel (md = 1 g) was immersed in deionized water until swelling equilibrium was reached. The swollen SF hydrogel was incubated in GM1 solution (initial concentration C0 = 5 mg/mL, V0 = 10 mL) for 48 h to achieve adsorption equilibrium. After loading, the residual GM1 solution (V mL) was collected. The absorbance of the solution was measured using a UV-visible spectrophotometer, and the residual concentration (C mg/mL) was determined from a standard calibration curve. The drug loading capacity (m, mg/g) was calculated using the following formula: m= (C0V0-CV)/md. GM1-loaded hydrogels were immersed in deionized water. At specific time points (4 h, 12 h, 1 d, 3 d, 5 d, 7 d, 14 d, and 28 d), the release medium was sampled and analyzed by UV-visible spectrophotometry to determine the cumulative amount of GM1 released (mR, mg/g). The cumulative release percentage (R, %) was calculated as: R=(mR/m)×100%. The composition of the GM1@SF hydrogel used in the main experiments was selected based on pretests of drug release behavior and cytocompatibility (see Supplementary Fig. S1). A GM1 concentration of 5 mg/mL provided sustained release for up to 28 days, and 150 µg/mL was identified as the optimal concentration for maintaining maximal cell viability and biosafety in both HT22 and bMSC assays. Detailed experimental procedures and data are provided in the Supplementary Materials. Cell culture and cytotoxicity assessment of GM1@SF hydrogel Isolation and culture of bMSCs The isolation of bMSCs was carried out on 3-4-week-old male Sprague-Dawley (SD) rats, in accordance with well-established protocols [26]. Primary cells were cultured in Dulbecco’s modified Eagle medium (DMEM)/F12 medium (11320033, Thermo Fisher Scientific) supplemented with 10% fetal bovine serum (FBS; C0235, Beyotime), 2 ng/mL epidermal growth factor (EGF; HY-P7109, MedChemExpress), and 1% penicillin-streptomycin (PS; C0222 Beyotime). Cells were maintained at 37 °C in a humidified atmosphere containing 5% CO₂. The culture medium was replaced every 72 h, and cells were passaged once or twice per week when they reached 80–90% confluence. All experiments were performed using bMSCs from passages 3 to 8 to guarantee the consistency of the phenotype. HT22 cell culture HT22 cells were obtained from the Chinese Academy of Sciences. Cells were seeded in polystyrene culture dishes and maintained in high-glucose DMEM containing 5% FBS (Gibco, USA), 10% horse serum (Gibco, USA), and 1% PS (Gibco, USA). Cultures were incubated at 37 °C in an atmosphere of 5% CO₂. For cell experiments, eight groups were established: Control group (cells cultured without treatment); lipopolysaccharide (LPS) group (cells treated with 100 ng/mL LPS derived from Escherichia coli 055:B5; Solarbio, China); Control+GM1 groups (cells treated with 100 µg/mL, 150 µg/mL, or 200 µg/mL GM1); LPS+GM1 groups (cells co-treated with 100 ng/mL LPS and 100 µg/mL, 150 µg/mL, or 200 µg/mL GM1). All treatments were conducted under identical culture conditions. Cell viability assay BMSCs were seeded in 96-well plates and cultured at 37 °C in an atmosphere containing 5% CO₂. Cell viability was assessed using the Cell Counting Kit-8 (CCK-8; A311-01, Vazyme). For the colorimetric assay, 100 µL of CCK-8 reagent was added directly to each well. The plates were then incubated at 37 °C for 3–4 h. The optical density (OD) of each well was measured at 450 nm using a microplate spectrophotometer (BioTek Instruments, Winooski, VT, USA). Cell viability was expressed as the percentage of viable cells relative to the control group. Each experiment was performed with six replicate wells per group and repeated independently three times. Encapsulation of bMSCs in hydrogel vesicles The prepared SF hydrogel solution was placed into a glass-bottom dish and mixed with a cell suspension containing 2 × 106 cells/mL under sterile conditions. The culture medium was then added and the samples were incubated at 37 °C in a humidified atmosphere containing 5% CO₂. The medium was replaced every 2–3 days. After 12 days of culture, the bMSCs-encapsulated SF hydrogels (bMSCs@SF hydrogels) were collected for in vivo experiments. Nuclear staining for assessment of cell death Morphological changes and chromatin condensation in the nuclei of bMSCs were examined using DAPI (P0131, Beyotime) and propidium iodide (PI; C1008M, Beyotime) staining. Cells were first fixed with 4% paraformaldehyde for 10 min, followed by three washes with phosphate-buffered saline (PBS). The cells were then stained with 10 µg/mL DAPI for 10 min at room temperature. For PI staining, a stock solution of PI (10 mg/mL; Sigma-Aldrich) was diluted with 0.9% NaCl to a final concentration of 40 µg/mL. Cells were fixed in 100% ethanol for 10 min at room temperature, stained with PI for 30 min, washed three times with PBS, and mounted with Permount. Stained samples were observed under a Nikon Eclipse Ti-S fluorescence microscope (Nikon Corporation, Tokyo, Japan). Viable cells displayed normal nuclear morphology and uniform fluorescence in the DAPI channel, whereas dead cells showed both PI and DAPI staining with condensed nuclei. Colocalization and quantitative analysis were performed using ImageJ software version 1.6 (NIH, USA). Animals and drug treatments Male SD rats with an average body weight of 250 g (Vital River, Beijing, China) were used for all experiments. The animals were housed in a pathogen-free facility under a 12-hour light/dark cycle with free access to food and water. Rats were randomly assigned to the following groups: Sham, SCI, SCI + SF, SCI+bMSCs, SCI+GM1, SCI+bMSCs@SF, SCI+GM1@SF, SCI+GM1 + bMSCs, and SCI+GM1 + bMSCs@SF. For behavioral assessments, six rats were included in each group (n = 6). For molecular and histological analyses, including western blotting (WB), immunofluorescence, and hematoxylin-eosin (H&E) staining, three rats were used per group (n = 3). The sample size was determined by statistical power analysis. At one hour after SCI, 20 µL of SF hydrogel, GM1@SF hydrogel, bMSCs@SF hydrogel, or GM1 + bMSCs@SF hydrogel was injected into the lesion site at a rate of 1 µL per minute. The needle was kept in place for one minute to ensure proper diffusion. GM1 was dissolved in saline. Sham animals underwent laminectomy and saline injection without SCI induction. All procedures were performed by investigators blinded to group allocation. All animal experiments were approved by the Institutional Animal Care and Use Committee of Nantong University and conducted in accordance with the national institutes of health (NIH) guide for the care and use of laboratory animals and the ARRIVE guidelines. All efforts were made to minimize animal numbers and reduce suffering. SCI model The SCI model was established as previously described [27]. Rats were deeply anesthetized with pentobarbital (50 mg/kg, intraperitoneal injection), and surgery was performed under aseptic conditions. A dorsal laminectomy was carried out at the ninth thoracic vertebra (T9), followed by a contusion injury using the NYU impactor. A 10 g rod with a diameter of 2.0 mm was dropped from a height of 10 cm onto the exposed spinal cord. Body temperature was maintained between 36.5 °C and 37.5 °C throughout the procedure using a heating pad and lamp. The overall mortality rate was less than 2%. After surgery, animals were carefully monitored and allowed to recover under continuous observation. WB analysis For WB analysis, spinal cord tissues were collected and stored at -80 °C until use. A 10 mm segment of spinal cord, including 5 mm rostral and 5 mm caudal to the injury center, was dissected for protein extraction. Frozen samples were minced on ice and homogenized in lysis buffer containing 1% NP-40, 50 mmol/L Tris (pH 7.5), 5 mmol/L EDTA, 1% sodium deoxycholate, 1% Triton X-100, 1% SDS, 1 mmol/L PMSF, 10 µg/mL aprotinin, and 1 µg/mL leupeptin. The homogenates were centrifuged at 4 °C for 20 min, and the supernatants were collected. Protein concentrations were determined using the Bradford assay (Bio-Rad). Equal amounts of protein (50 µg) were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene difluoride (PVDF) membranes (FFP28, Beyotime) using a wet transfer system at 300 mA for 2 h. Membranes were blocked with 5% nonfat milk at room temperature for 2 h, followed by incubation with the following primary antibodies at 4 °C for 18–24 h: Cend1 (Rabbit, 1:10,000; EPR3739; Abcam), active caspase-3 (Mouse, 1:500; sc-56053; Santa Cruz), Bax (Mouse, 1:500; sc-20067; Santa Cruz), Bcl-2 (Mouse, 1:500; sc-7382; Santa Cruz), NLRP3 (Mouse, 1:500; sc-134306; Santa Cruz), GSDMD (Mouse, 1:500; sc-393581; Santa Cruz), ASC (Mouse, 1:500; sc-514414; Santa Cruz), caspase-1 (Mouse, 1:500; sc-56036; Santa Cruz), p-ERK1/2 (Rabbit, 1:10,000; #9101; Cell Signaling Technology), cyclin D1 (Mouse, 1:500; sc-8396; Santa Cruz), NeuN (Mouse, 1:500; 66836-1-Ig; Proteintech), and GAPDH (Mouse, 1:1,000; G9545; Sigma). After primary antibody incubation, membranes were washed three times with Tris-buffered saline containing 0.1% Tween-20 (TBST; ST828, Beyotime) for 5 min each. The membranes were then incubated with horseradish peroxidase-conjugated secondary antibodies (31460, Thermo Fisher Scientific) at room temperature for 2 h, followed by three additional washes with TBST. Protein bands were visualized using an enhanced chemiluminescence detection system (P0018S, Beyotime). Double immunofluorescent staining Tissue sections were removed from storage and placed in an oven at 37 °C for 40 min. To prevent nonspecific staining, sections were blocked with 10% normal serum from the same species as the secondary antibody, containing 3% bovine serum albumin (BSA), 0.1% Triton X-100, and 0.05% Tween-20, for 2 h at room temperature. After blocking, sections were incubated overnight at 4 °C with the following primary antibodies: anti-GSDMD (Mouse, 1:200; sc-393581; Santa Cruz) and one of the cell-specific markers, NeuN (Rabbit, 1:200; ab177487; Abcam), GFAP (Rabbit, 1:200; ab7260; Abcam), or Iba-1 (Rabbit, 1:200; ab178846; Abcam). The next day, sections were washed with PBS and incubated with CY2- and CY3-conjugated secondary antibodies (R37116; Thermo Fisher Scientific) for 2 h at 4 °C. The stained sections were examined using a Leica spectral confocal microscope (Leica Microsystems, Germany). Small interfering RNA (siRNA) and transfection The siRNA constructs targeting Cend1 were designed with the following nucleotide sequences: 5′-CCUUGACCAAGCCUGUGAATT-3′, 5′-CCUGGCCUUGUGAGAACUUTT-3′, 5′-GCAGAUCCUGUACUUCUCATT-3′ (GIEE0346762, GeneChem). HT22 cells were transfected with either Cend1-specific siRNA or a negative control plasmid using Lipo8000™ Transfection Reagent (C0533, Beyotime) according to the manufacturer’s instructions. The efficiency of Cend1 knockdown was confirmed by western blot analysis. Transfected cells were used for subsequent experiments 48 h after transfection. H&E staining Rats were deeply anesthetized before perfusion. The following steps were performed in sequence. Pre-perfusion was carried out with 0.01 M PBS. Fixative perfusion was performed using 4% paraformaldehyde in PBS. Post-fixation was conducted by rinsing the tissue in running water for 30 min. The samples were then dehydrated through a graded ethanol series of 70%, 80%, 90%, and 100%, with each step lasting one hour. Finally, the tissues were cleared in xylene for 30 min twice. Tissues were immersed in melted paraffin at 60 °C three times for 1 h each. Orientation and embedding were performed by casting in paraffin, followed by solidification at 4 °C for 30 min. Paraffin blocks were stored at -20 °C until sectioning. Paraffin blocks were sectioned at a thickness of 5–8 μm using a rotary microtome. Sections were floated on a 40–45 °C water bath for 3–5 min and then mounted onto poly-L-lysine-coated glass slides at a 15° angle. Slides were dried in an oven at 37–45 °C for 2 h. Slides were baked at 60 °C for 30 min, followed by deparaffinization in xylene twice for 10 min each. Rehydration was performed through a graded ethanol series: 100% ethanol (two changes, 5 min each), 95%, 90%, 80%, and 70% ethanol (5 min each), followed by rinsing in distilled water. Slides were stained with Harris hematoxylin for 3–8 min at room temperature. Differentiation was performed in 1% acid ethanol for 1–3 s under microscopic observation. Bluing was carried out using 0.6% ammonia water for 30 s or running water for 5 min. Slides were stained with 0.5-1% eosin for 1–3 min at room temperature. Sections were dehydrated in 95% ethanol (two changes, 5 min each) followed by 100% ethanol (two changes, 5 min each). Clearing was performed in xylene twice for 5 min each. Slides were air-dried until the xylene odor disappeared, then mounted with neutral balsam. Care was taken to avoid air bubbles. Stained slides were examined under a light microscope, and digital images were captured. Histomorphometric analysis was performed on the acquired images. Behavioral function tests The Basso, Beattie, and Bresnahan (BBB) locomotor rating scale test, inclined plane test, and beam walk test were used for behavioral evaluation as previously described [28]. All rats were tested at the same time of day on days 1, 3, 7, 14, 21, 28, 35, and 42 after surgery (n = 6 per group). Each rat was allowed to rest for 30 min after completing one test before proceeding to the next [29]. Statistical analysis All data were expressed as mean ± SEM. The number of PI-positive cells was analyzed using the rank-sum test. Data from motor function were analyzed by two-way repeated measures analysis of variance (ANOVA) with factors for group and time, followed by Bonferroni’s post hoc test for multiple comparisons. Data from the probe trial were analyzed using one-way ANOVA followed by Tukey’s post hoc test. Western blot results were analyzed using one-way ANOVA with Dunnett’s t-test. All analyses were performed using SPSS software (version 13.0 for Windows, SPSS Inc., USA). A value of P < 0.05 was considered statistically significant. Results Characterization of GM1@SF hydrogel The microstructure of the GM1@SF hydrogel was examined using scanning electron microscopy (SEM). The SF hydrogel showed an open and interconnected porous structure, with pore sizes larger than 100 μm. Since the diameter of bMSCs is about 20–30 μm, this pore size is suitable for supporting cell survival (Fig. 1a). To confirm the physical state of the material, the aqueous SF solution and the hydrogel were tilted at a 45° angle. The meniscus did not deform, indicating successful gel formation. Loading GM1 into the SF hydrogel did not cause visible changes in morphology (Fig. 1b). Characterization of GM1@SF hydrogel. (a) SEM images show the microstructure of the SF hydrogel and the GM1@SF hydrogel (Scale bars: upper images, 500 μm; lower images, 200 μm). (b) The inversion test shows the gelation state of the SF hydrogel. (c) The injection test confirms the injectability of the SF hydrogel. (d) The in vitro release profile of GM1 from the GM1@SF hydrogel compared with the GM1 aqueous solution (n = 3 independent experiments for each group) For further testing, the hydrogel precursor solution was placed in a 1 mL syringe and allowed to gel. The formed hydrogel could be smoothly extruded through the syringe needle to form the letters “SF” (Fig. 1c), showing good injectability. UV-Vis spectrophotometric analysis demonstrated that the SF hydrogel allowed a slow and continuous release of GM1. About 25% of GM1 was released on the first day, followed by a gradual increase in release over time. GM1 release was still detectable after 28 days (Fig. 1d). Cytotoxicity evaluation of GM1@SF hydrogel The cytotoxicity of GM1@SF hydrogel was assessed using a CCK-8 assay. HT22 cells were cultured for seven days in three groups: (1) DMEM high glucose with 10% FBS (control group), (2) SF hydrogel (SF group), and (3) GM1-loaded SF hydrogel (GM1@SF group). As shown in Fig. 2a, cell viability in both the SF and GM1@SF groups was close to 100%, similar to the control group, indicating that these hydrogel systems were not cytotoxic. Biocompatibility evaluation of SF and GM1@SF hydrogels. (a) The CCK 8 assay shows that cell viability remains close to 100% in all groups (n = 3 independent experiments, each performed in triplicate; ns indicates no significant difference). (b) Live and dead staining with the Calcein AM and PI kit shows excellent cell survival in the SF hydrogel and the GM1@SF hydrogel. No dead cells are observed (scale bar: 50 μm; n = 3 independent experiments) To further confirm these results, cell viability was evaluated using a Calcein-AM/PI double-staining assay. Calcein-AM, a cell-permeant dye, is converted by intracellular esterases into green-fluorescent calcein in live cells. PI is a red-fluorescent dye that only enters cells with damaged membranes, labeling dead cells. HT22 cells were cultured under the same three conditions for 24 h, and live/dead cells were quantified. The results showed strong cell survival in both hydrogel groups, with no detectable dead cells, further confirming that SF and GM1@SF hydrogels were non-cytotoxic (Fig. 2b). GM1 effectively reduces neuronal apoptosis and pyroptosis The neuronal apoptosis and pyroptosis model was successfully established by stimulating HT22 cells with 100 ng/ml lipopolysaccharide (LPS). Western blot (WB) analysis was used to examine the expression levels of cell cycle exit and neuronal differentiation 1 (Cend1; a neural-specific regulator involved in promoting cell cycle exit and neuronal differentiation), the apoptosis-related protein Caspase-3, and the pyroptosis-related proteins NLRP3, GSDMD, Caspase-1, IL-1β, and ASC. Protein expression was compared among three experimental conditions: injury group (LPS-induced damage), non-injured groups treated with different concentrations of GM1, and injured groups treated with different concentrations of GM1. As shown in results, Cend1 expression did not change significantly following neuronal injury. However, treatment with different concentrations of GM1 led to varying degrees of Cend1 upregulation in both injured and non-injured groups (Fig. 3a, b). The apoptosis-related protein Caspase-3 increased significantly in the injury group, while GM1 treatment did not alter Caspase-3 expression in non-injured cells. Notably, GM1 treatment reduced Caspase-3 expression in injured neurons in a concentration-dependent manner, indicating effective inhibition of LPS-induced neuronal apoptosis (Fig. 3a, c). GM1 reduces neuronal pyroptosis and apoptosis. (a) Western blot analysis shows the expression of Cend1 and proteins related to apoptosis and pyroptosis. (b-h) Quantitative analysis of the expression levels of Cend1 and apoptosis related proteins in each group (n = 3 for each group; P < 0.05 compared with the Control group and the LPS group) Similarly, the expression of pyroptosis-related proteins (NLRP3, GSDMD, Caspase-1, IL-1β, and ASC) increased significantly after injury. In non-injured cells, GM1 treatment did not significantly affect these proteins. In contrast, GM1 reduced the expression of these pyroptosis-related proteins in injured neurons, suggesting that GM1 inhibits LPS-induced neuronal pyroptosis (Fig. 3a, d-h). Overall, these results indicate that GM1 provides multi-target neuroprotection by regulating Cend1 expression and inhibiting proteins associated with apoptosis and pyroptosis. GM1 inhibits neuronal apoptosis through the Cend1 signaling pathway Lip8000 transfection technology was used to introduce Cend1 siRNA plasmids into HT22 cells. The interference efficiency of different Cend1 siRNAs was evaluated by WB analysis, and the plasmid with the highest interference efficiency was selected for subsequent experiments. The results showed that Cend1-siRNA#2 most effectively reduced Cend1 expression (Fig. 4a, b). GM1 inhibits neuronal apoptosis through the Cend1 signaling pathway. (a-b) Western blot analysis shows that Cend1 siRNA number 2 produces the strongest interference effect. (c-f) Western blot analysis shows that Cend1-siRNA#2 reduces the protective effect of GM1 on neuronal apoptosis. It does not change the protective effect of GM1 on neuronal pyroptosis (n = 3 independent experiments for each group; P < 0.05) To explore the mechanism by which GM1 inhibits neuronal apoptosis and pyroptosis and its relationship with Cend1, HT22 cells were transfected with Cend1-siRNA#2. After three days of culture, the cells were stimulated with 100 ng/ml LPS. Six hours later, 200 µg/ml GM1 was added and incubated for 24 h. Cells were then collected for WB analysis to evaluate protein expression across the groups. The results showed that Caspase-3 expression in the LPS+GM1 group was significantly higher than in the control group and did not differ significantly from the LPS group. Meanwhile, GSDMD expression in the LPS+GM1 group did not change significantly compared with the Cend1-interfered group (Fig. 4c-f). These results indicate that Cend1 interference inhibits the protective effect of GM1 on neuronal apoptosis, but does not affect its inhibition of pyroptosis. This suggests that GM1 suppresses neuronal apoptosis by upregulating endogenous Cend1 and activating downstream signaling pathways. However, GM1’s inhibition of neuronal pyroptosis does not appear to depend on Cend1. GM1 promotes neuronal differentiation of bMSCs bMSCs were treated with different concentrations of GM1 (100, 150, and 200 µg/ml) for seven days to establish differentiation models. WB analysis was used to detect the expression of Cend1, the neuronal marker NEUN, Cyclin D1, and P-ERK, a key protein regulating cell proliferation and differentiation. As shown in results, Cend1 expression increased to varying degrees after GM1 treatment (Fig. 5.1a, b). NEUN expression also increased (Fig. 5.1a, c). In contrast, Cyclin D1 and P-ERK expression decreased following GM1 stimulation (Fig. 5.1a, d, e). These results suggest that GM1 promotes Cend1 expression, inhibits bMSC proliferation, and induces neuronal differentiation. GM1 promotes neuronal differentiation of bMSCs. (a) Western blot analysis shows the expression of Cend1, NEUN, Cyclin D1 and P-ERK. (b-e) Quantitative analysis of the expression levels of Cend1, NEUN, Cyclin D1 and P- ERK (*P < 0.05) To further validate these findings, immunofluorescence analysis was performed to identify cell types after differentiation. Different concentrations of GM1 stimulated bMSCs to differentiate into neurons, and at 150 µg/ml and 200 µg/ml, nearly all bMSCs differentiated into neurons (Fig. 5.2a, b). Overall, these results indicate that GM1 promotes bMSC differentiation into neurons, potentially through upregulation of Cend1 and downregulation of cell cycle and proliferation-related proteins. GM1 promotes neuronal differentiation of bMSCs. (a-b) Immunofluorescence images show that bMSCs differentiate into neurons after treatment with different concentrations of GM1 (Scale bar: 50 μm) Lip8000 was used to transfect Cend1-siRNA#2 into bMSCs. After one week of culture under 200 µg/ml GM1 stimulation, Cend1 downregulation significantly impaired neuronal differentiation. As shown in Fig. 5.3a-e, NEUN expression was markedly reduced, while Cyclin D1 and P-ERK expression increased after Cend1 interference. These results indicate that GM1 promotes bMSC differentiation into neurons primarily through Cend1. When Cend1 is inhibited, bMSCs stimulated by GM1 tend to remain in a proliferative state rather than differentiate. This supports the hypothesis that GM1-induced neuronal differentiation of bMSCs depends on Cend1 regulation. GM1 promotes the neuronal differentiation of bMSCs through the Cend1 signaling pathway. (a) Western blot analysis shows the expression of Cend1, NEUN, Cyclin D1 and P-ERK. (b-e) Quantitative analysis of the expression levels of Cend1, NEUN, Cyclin D1 and P-ERK (*P < 0.05) GM1 + bMSCs@SF hydrogel suppresses the expression of proteins associated with pyroptosis and apoptosis After establishing a SCI model in SD rats, the expression of Cend1 and proteins related to apoptosis and pyroptosis (Caspase 3, GSDMD) was measured by WB in the Sham group and in injured spinal cord tissue collected at 6 h, 12 h, 1 d, 3 d, 5 d, 7 d and 14 d after injury (Fig. 6.1a). The results showed that Cend1 and the apoptosis and pyroptosis related proteins rose after 12 h, peaked at 3 d, and then gradually declined (Fig. 6.1b-d). These findings indicate that apoptosis and pyroptosis occur after spinal cord injury and may be associated with Cend1 expression. The peak at day 3 suggests a temporal pattern that has important implications for the timing of therapeutic interventions. Expression of apoptosis and pyroptosis related proteins after SCI. (a) Western blot analysis shows the expression of Cend1 and the proteins related to apoptosis and pyroptosis, including Caspase-3 and GSDMD, at different time points after SCI (b-d) Quantitative analysis shows the expression levels of Cend1, Caspase-3 and GSDMD in spinal cord tissue from each group (*P < 0.05 compared with the Sham group) Following successful model establishment, seven treatment groups (SF, bMSCs, GM1, bMSCs@SF, GM1@SF, GM1 + bMSCs, GM1 + bMSCs@SF) were treated. After 3 days of treatment, Cend1 and proteins related to apoptosis and pyroptosis (Caspase 3, NLRP3, GSDMD, Caspase 1, IL-1beta, ASC) in spinal cord tissue were assessed by WB (Fig. 6.2a). Cend1 expression increased significantly in the GM1, GM1@SF and GM1 + bMSCs@SF groups, with the largest increases in GM1@SF and GM1 + bMSCs@SF (Fig. 6.2b). At the same time, levels of apoptosis and pyroptosis markers decreased to varying degrees in all treatment groups, with the greatest reductions observed in the GM1 + bMSCs@SF group (Fig. 6.2c). Notably, the SF group alone showed a modest reduction compared with control, although this effect was smaller than that seen in the combination groups. These results suggest that GM1@SF and GM1 + bMSCs@SF may improve pathological outcomes after SCI by upregulating Cend1 and reducing apoptosis and pyroptosis. Expression of apoptosis and pyroptosis related proteins after 3 days of SCI treatment. (a) Western blot analysis shows the expression of Cend1 and proteins related to apoptosis and pyroptosis, including Caspase-3, NLRP 3, GSDMD, Caspase-1, IL-1β and ASC, in each treatment group after 3 days of SCI. (b-h) Quantitative analysis of the expression levels of Cend1, Caspase-3, NLRP 3, GSDMD, Caspase-1, IL-1β and ASC in spinal cord tissue from each group (*P < 0.05 compared with the Sham group and the SCI3d group) The GM1 + bMSCs@SF hydrogel inhibits neuronal pyroptosis To extend the above findings, spinal cord tissue was collected at day 3 after SCI and analyzed by immunofluorescence to determine the cellular localization of the pyroptosis marker GSDMD. Co-staining with the neuronal marker NeuN was performed, and as shown in Fig. 7.1, GSDMD signal was detected in NeuN-positive neurons after injury. These results indicate that neuronal pyroptosis occurs following SCI and provide a basis for therapeutic strategies targeting the pyroptosis pathway. In addition, co-staining with antibodies against GFAP and IBA1 was performed to identify astrocytes and microglia, respectively; the results are shown in Supplementary Fig. 2. Cellular localization of the pyroptosis related protein GSDMD in spinal cord tissue. (a) Representative immunofluorescence images show the colocalization of GSDMD with NeuN in spinal cord tissue 3 days after SCI. High magnification insets show colocalization within the same cells. (Scale bar: 50 μm; Inset scale bar: 10 μm). (b) Quantification was performed using 3 animals for each group. Three random fields of view were selected for each animal Having identified neurons as the main cell type exhibiting pyroptotic markers three days after injury, we examined the impact of various treatments on neuronal pyroptosis using immunofluorescence. As shown in Fig. 7.2, all treatment groups reduced neuronal GSDMD expression to varying degrees at 3 days after injury, with the GM1 + bMSCs@SF hydrogel producing the most pronounced suppression. This result suggests a potential therapeutic benefit of the composite. The combined treatment may act through several mechanisms, including direct downregulation of pyroptosis-related proteins, activation of pro-survival signaling pathways, and provision of a more favorable microenvironment for neuronal repair and regeneration. Changes in pyroptotic neurons in spinal cord tissue in different treatment groups. (a-b) Dual immunofluorescence staining shows the expression of GSDMD and the neuronal marker NeuN in spinal cord tissue 3 days after treatment. (c) Quantitative analysis shows that all treatment groups suppress neuronal pyroptosis to different degrees at 3 days after injury. The GM1 and bMSCs at SF hydrogel composite group shows the strongest inhibition (*P < 0.05 compared with the Sham group and the SCI3d group; n = 3 animals for each group; three random fields of view were analyzed for each animal; scale bar: 50 μm) The GM1 + bMSCs@SF hydrogel promotes neurogenesis and reduces syringomyelia formation after SCI Neurogenesis was evaluated by immunofluorescence staining four weeks after SCI (Fig. 8.1). Sections from four groups, Control, bMSCs, GM1 + bMSCs and GM1 + bMSCs@SF hydrogel, were examined. Compared to the control group, the bMSCs group showed a nearly twofold increase in the proportion of Tuj-1-positive cells, which are indicative of newly formed neurons. The GM1 + bMSCs group showed a roughly threefold increase. The GM1 + bMSCs@SF hydrogel group exhibited more than a sevenfold increase. All increases were statistically significant. These data indicate that combining bMSCs with GM1 enhances neurogenesis, and that inclusion in the SF hydrogel strongly amplifies this effect. Tuj-1 immunofluorescence staining of spinal cord tissue 4 weeks after SCI in four treatment groups. (a) Representative Tuj-1 immunofluorescence images of spinal cord tissue from the Control group, the bMSCs group, the GM1 + bMSCs group and the GM1 + bMSCs@SF hydrogel group at 4 weeks after injury. (b) Quantitative analysis shows that the bMSCs group and the GM1 + bMSCs group have a larger Tuj-1 positive area than the Control group. The GM1 + bMSCs@SF hydrogel group shows a significantly larger Tuj-1 positive area than all other groups (*P < 0.05; n = 3 animals for each group; three sections were analyzed for each animal; scale bar: 250 μm) Tissue repair was assessed by H&E staining (Fig. 8.2). The GM1 + bMSCs@SF hydrogel group showed a greater than 50% reduction in cystic cavity area compared with the untreated control group. bMSCs and GM1 + bMSCs produced moderate decreases in cavity size, but only the composite group showed a marked and statistically significant reduction. Thus, the combined treatment of GM1, bMSCs and SF hydrogel not only promotes formation of new neurons but also effectively limits syringomyelia, supporting improved spinal cord tissue repair. H&E staining of spinal cord tissue 4 weeks after spinal cord injury in four treatment groups. (a-d) Representative HE staining images of spinal cord tissue from the Control group, the bMSCs group, the GM1 and bMSCs group and the GM1 and bMSCs at SF hydrogel composite group at 4 weeks after injury. (e) Quantitative analysis shows that the bMSCs group and the GM1 and bMSCs group have a significantly smaller cystic cavity area than the Control group. GM1+bMSCs@SF hydrogel group shows a much smaller cystic cavity area than all other groups (*P < 0.05; n = 3 animals for each group; three sections were analyzed for each animal; scale bar: 250 μm) GM1 + bMSCs@SF hydrogel improves motor function in rats after SCI To assess motor recovery after SCI, we used the BBB scale. The BBB assessment is an established open field locomotor rating scale ranging from 0 (no hindlimb movement) to 21 (normal hindlimb function in healthy control rats), which evaluates joint movement, limb coordination, and paw placement. BBB scoring was performed weekly from 1 to 8 weeks after injury. Functional recovery was expressed as a percentage of the maximum score (BBB score/21 × 100%). In the first week after surgery, all injured groups except the uninjured control showed severe motor impairment, with BBB scores below 10% of the maximum (Fig. 9). During weeks 2 and 3, all treatment groups showed gradual improvement. By week 4, the GM1 + bMSCs@SF hydrogel group reached approximately 36% of the maximum score (mean ~ 7.5 of 21), while the untreated SCI group remained below 20% (mean ~ 3.5 of 21). BBB scores of hindlimb functional recovery during 8 weeks after SCI in rats. BBB scores were recorded for each group for 8 weeks (n = 6 animals for each group) Differences became more pronounced from weeks 5 to 8. At week 8, the GM1 + bMSCs@SF group achieved a mean BBB score of about 15 of 21, corresponding to 71.4% of normal function. The untreated SCI group reached only about 6.3 of 21 (30.0%). Groups treated with single agents, such as GM1 or bMSCs alone, showed intermediate recovery, typically 9 to 13 of 21 (43% to 62%). Statistical analysis showed that from week 4 onward the GM1 + bMSCs@SF group had significantly higher functional recovery than all other groups (P < 0.05). These data demonstrate that the GM1 + bMSCs@SF hydrogel markedly enhances motor recovery after SCI. The results support a key role for bMSCs in repair and indicate a synergistic effect of GM1 and the SF hydrogel in promoting functional neuroregeneration. Discussion Treatment strategies for SCI are generally divided into two main approaches. The first approach aims at neural regeneration. It seeks to reconstruct neural pathways by using cell transplantation and tissue engineering to promote neuron and axon regrowth and thereby restore function [27, 30]. The second approach focuses on neuroprotection. It aims to limit secondary damage by using neuroprotective agents such as corticosteroids including methylprednisolone and dexamethasone, recombinant human granulocyte colony stimulating factor (rhG-CSF), riluzole, fibroblast growth factor 2 (FGF-2), erythropoietin (EPO), ethylene oxide derivatives, polyethylene glycol (PEG)and GM1. These agents act to mitigate secondary injury cascades [31, 32]. Although many of these methods are effective in animal models, clinical evidence remains inconclusive. Given the complex and multifactorial nature of SCI, single agent therapies have clear limitations. Therefore, combined strategies that integrate multiple treatment modalities may offer greater therapeutic benefit [33]. Stem cell therapy has attracted considerable attention because of its ability to self renew and to differentiate into multiple cell types. Transplanted stem cells can replace lost neurons and supporting cells and thereby promote repair after SCI [6, 34]. The potential of bMSCs as a cell type for clinical translation is particularly promising. They are relatively easy to obtain by minimally invasive bone marrow aspiration. They show low immunogenicity after allogeneic transplantation and have a low risk of tumor formation. bMSCs can differentiate into several lineages, including osteocytes, chondrocytes, neural cells and adipocytes. In addition, they secrete trophic factors such as brain derived neurotrophic factor (BDNF), vascular endothelial growth factor (VEGF) and insulin-like growth factor 1 (IGF-1). These paracrine effects modulate inflammation, promote angiogenesis and support cell survival and differentiation, which together enhance tissue repair. bMSCs can be expanded in vitro to obtain therapeutic cell numbers and they are suitable for autologous use without major ethical concerns [35, 36]. These characteristics make bMSCs a strong candidate for cell based therapies in SCI. GM1 is a sialic acid containing glycosphingolipid that has important biological functions in the nervous system. It regulates neural cell growth and development and facilitates interneuronal communication and signal transduction. GM1 also promotes neuronal differentiation and guides axonal pathfinding. In addition, it helps maintain neural homeostasis and function. GM1 has immunomodulatory effects and can influence immune activity. Clinical studies have reported therapeutic benefits of GM1 in a range of conditions, including traumatic brain injury, cerebral hemorrhage, Parkinson’s disease, Alzheimer’s disease, and certain immune related disorders and cancer adjuvant therapies [37, 38]. Despite its clinical use, the precise molecular mechanisms underlying GM1’s effects remain to be clarified. Hydrogels are three dimensional, crosslinked polymer networks with high water content, tunable physicochemical properties and an extracellular matrix (ECM) like microenvironment. They can be engineered to adjust mechanical strength and degradation rate. While other natural polymers such as collagen and hyaluronic acid are frequently used, they often exhibit rapid degradation and limited mechanical strength, which can be suboptimal for the dynamic and harsh environment of an SCI lesion. In contrast, SF-based hydrogels have attracted attention due to their superior biodegradability, good biocompatibility, and highly tunable mechanical propertie [39–41]. This tunability allows them to better mimic the native spinal cord tissue while providing robust structural support. Importantly, SF hydrogels can serve as drug delivery platforms and enable controlled release of therapeutic agents, which offers a promising strategy for treating SCI [42]. In this study, we developed a SF based hydrogel that provided sustained delivery of GM1 and bMSCs (GM1 + bMSCs@SF hydrogel) and evaluated its effect in a rat model of SCI. Treatment with GM1 + bMSCs@SF hydrogel increased the expression of Cend1. Cend1 is a protein closely linked to neural development. It promotes neural stem cell differentiation into neurons and is involved in nervous system maturation. Previous studies have shown that Cend1 can regulate p75 neurotrophin receptor (p75NTR) expression, inhibit neuronal apoptosis, and drive neuronal differentiation of neural stem cells [43, 44]. Given the established pro-apoptotic function of p75NTR following SCI, it is plausible that the neuroprotection we observed may be linked to the Cend1-p75NTR axis. Specifically, the upregulation of Cend1 likely suppresses p75NTR’s pro-apoptotic signaling, thereby reducing neuronal cell death. In our experiments, GM1 + bMSCs@SF hydrogel not only upregulated Cend1, but also reduced neuronal pyroptosis and apoptosis and promoted bMSCs differentiation toward neuronal lineages. Importantly, the SF hydrogel alone produced a modest neuroprotective effect. This was reflected by reduced levels of apoptosis and pyroptosis related proteins, including cleaved caspase-3, NLRP3 and GSDMD, compared with untreated controls. These effects may reflect the biocompatibility and three dimensional supportive environment provided by SF hydrogels, as reported previously [45]. Such a microenvironment can support endogenous neuroprotection and tissue repair and thus partially attenuate secondary injury after spinal cord trauma. Nevertheless, the therapeutic benefit of the SF hydrogel alone was clearly less pronounced than that of the combination treatments, and was markedly lower than that of the GM1 + bMSCs@SF group. In vitro, LPS was used to induce apoptosis and pyroptosis in HT22 cells. GM1 treatment increased Cend1 expression and reduced LPS-induced apoptosis and pyroptosis in these cells. Knockdown of Cend1 with siRNA markedly increased the apoptosis signal but did not significantly alter pyroptosis. In bMSCs, GM1 treatment induced neuronal differentiation, as shown by WB and immunofluorescence. Cend1 knockdown prevented GM1 from driving bMSCs neuronal differentiation. Together, these findings indicate that GM1 upregulates Cend1 and inhibits neuronal apoptosis through a Cend1-dependent pathway while promoting bMSCs neuronal differentiation via the same pathway. The inhibition of neuronal pyroptosis by GM1 appears to occur through a Cend1-independent mechanism. These results define two complementary mechanisms for GM1 action and support further development of multi-targeted strategies for SCI repair. Our results demonstrate that the GM1 + bMSCs@SF hydrogel group achieved superior tissue repair and functional recovery, evidenced by a smaller spinal cavity and more new neurons. Behaviorally, this translated to a significantly improved BBB score of 15.00 ± 0.89 at 8 weeks post-injury. This outcome compares favorably with other hydrogel-based strategies in the literature. For instance, a study utilizing a gelatin methacrylate (GelMA) hydrogel co-delivering neural and mesenchymal stem cells reported a BBB score of approximately 11 at the same time point [46]. These findings support the superior efficacy of the GM1 + bMSCs@SF hydrogel in promoting motor recovery after SCI. They are also consistent with meta-analytic evidence showing significant locomotor improvements in hydrogel-treated SCI models [47]. The SF hydrogel alone produced moderate improvements in tissue morphology and motor function, which highlights the potential of SF as a scaffold for spinal cord repair. The most pronounced effects were observed with the combined GM1, bMSCs, and SF hydrogel treatment, underscoring the benefit of synergistic multi-component interventions for SCI repair. This study has several limitations that should be acknowledged. First, our in vitro mechanistic studies utilized the HT22 hippocampal neuronal cell line. While this is a standard and robust model for exploring fundamental neuroprotective pathways due to its stability and reproducibility, it is not of spinal cord origin. Therefore, future validation using primary spinal cord neurons or more complex organotypic slice cultures is essential to confirm that the observed Cend1-mediated effects are directly translatable to the specific neuronal populations affected in SCI. Second, While our initial experiments demonstrated that both GM1@SF and bMSC@SF have beneficial effects on protein expression related to apoptosis and pyroptosis, the absence of these groups at later endpoints limits our ability to fully delineate the individual contributions and potential synergistic effects of each component in neurogenesis and tissue repair. Future studies incorporating these experimental groups will be necessary to comprehensively assess their respective therapeutic effects, clarify underlying mechanisms, and further optimize multilayered strategies for spinal cord injury repair. Addressing this limitation will strengthen the interpretability and translational relevance of our findings. Furthermore, another limitation of the present study is that functional and histological assessments were limited to eight weeks after injury, largely due to practical constraints. While this observation period is consistent with many preclinical studies in the field, it may not capture long-term functional recovery or delayed adverse effects. Future investigations should incorporate longer follow-up durations to better evaluate the persistence and safety of these therapeutic effects over extended periods. Conclusions In summary, this is the first report of the modulation of neuronal apoptosis and pyroptosis, and the promotion of bMSC differentiation into neurons after SCI, through the transplantation of GM1-loaded bMSCs within an SF hydrogel. These findings reveal a novel molecular pathway that may contribute to post-traumatic repair following SCI and support the potential of this composite as a therapeutic strategy. The precise mechanisms underlying the effects of the GM1 + bMSCs@SF hydrogel remain to be defined. Further studies are needed to elucidate these mechanisms and to assess long-term efficacy and safety. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements Not applicable. Abbreviations Monosialotetrahexosylganglioside Spinal Cord Injury Silk Fibroin Bone marrow-derived mesenchymal stem cells Cell cycle exit and neuronal differentiation 1 Neural Stem Cell Glial Fibrillary Acidic Protein Glyceraldehyde-3-phosphate dehydrogenase Lipopolysaccharide 4’,6-diamidino-2-phenylindole Propidium Iodide Ultraviolet-visible Spectroscopy NOD-like receptor protein 3 Gasdermin D Apoptosis-associated speck-like protein containing a CARD Extracellular signal-regulated kinase Neuronal nuclei Ionized calcium-binding adapter molecule 1 Epidermal growth factor Neuron-specific class III beta-tubulin Author contributions All authors listed have made a substantial, direct, and intellectual contribution to the work. All authors have read and approved the final manuscript. Funding This study was supported by grants from Nantong University Special Scientific Research Fund for Clinical Medicine (grant no. 2023JY022, 2023LY016), Yancheng Municipal Science and Technology Project (grant no. DYKJ-202101) and Nantong Health Commission (grant no. QNZ2023069). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. Data availability All the original data of this study are available from the corresponding author upon request. Declarations Ethics approval and consent to participate All animal procedures were approved by the Institutional Animal Care and Use Committee of Nantong University (Approval No. P20230311-043) and were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Consent for publication Not applicable. Competing interests Authors wish to confirm that there are no known conflicts of interest associated with this publication and there has been no significant financial support for this work that could have influenced its outcome. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Qin Yu, Hai-Yan Zhang and Yan Song contributed equally to this work. Contributor Information Yang-Jun Xu, Email: drzhaojian123@163.com. Chen Chen, Email: drcc3090@163.com. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement All the original data of this study are available from the corresponding author upon request.