Synergistic mitochondrial homeostasis regulation and cholinergic circuits reconstruction via a one-step synthesized multifunctional hydrogel facilitates spinal cord injury repair

Abstract

Neural Stem Cells (NSCs) possess significant potential to form new neural networks. However, following spinal cord injury (SCI), mitochondrial dysfunction leads to the excessive accumulation of reactive oxygen species (ROS), which severely impairs the neuronal differentiation of endogenous NSCs and thus hinders neural regeneration. Here, we report a multifunctional hydrogel, Poly(LA-Cho)/SS31 (PLCS), synthesized in one step using lipoic acid (LA), choline bicarbonate, and elamipretide (SS31). PLCS hydrogel exhibits injectability, self-healing ability, tissue adhesion, and sequential drug release. Initially, SS31 is released preferentially to scavenge mitochondrial ROS and alleviate mitochondrial dysfunction. Subsequently, LA is continuously to scavenge ROS. Notably, PLCS hydrogel not only promotes NSCs differentiation into cholinergic neurons but also increases acetyl-CoA levels and supplies choline, offering necessary substrates for acetylcholine synthesis in newly formed cholinergic neurons to support their functional maturation. The PLCS hydrogel achieves robust nerve regeneration and significantly improves motor, sensory, and bladder functions in rat models of SCI. RNA sequencing suggests the PI3K-Akt pathway may contribute to spinal cord repair. This one-step synthesis method without catalysts and organic solvents can effectively integrate physical and biological functions of hydrogel, through simple mixing, offering a highly promising strategy for the clinical translation of SCI treatment and other central nervous system injuries.

Graphical abstract

Highlights

• •This catalyst- and solvent-free, one-step synthesis is a promising non-toxic, cost-effective, and clinically translatable method.
• •This PLCS hydrogel is injectable, adhesive, and self-healing, constructing a microenvironment for neural regeneration.
• •LA and Cho, as bioactive molecules, constitute the PLCS’s backbone, enabling an integrated "scaffold-as-medicine" strategy.
• •Synergistic regulation of mitochondrial homeostasis and reconstruction of cholinergic circuits contribute to repairing SCI.

1. Introduction

Spinal cord injury (SCI), a severe trauma to the central nervous system, has identified mitochondrial dysfunction as a critical pathological mechanism and a key therapeutic target for breakthroughs [1]. Following SCI, disruption of the mitochondrial electron transport chain triggers explosive accumulation of mitochondrial reactive oxygen species (mtROS), which not only leads to collapse of cellular energy metabolism but also exacerbates neuronal apoptosis and neuroinflammation through oxidative damage to biomolecules such as deoxyribonucleic acid (DNA), proteins, and lipids [2,3].

Cholinergic neurons play a central role in spinal motor control. The synthesis of acetylcholine (ACh) from choline and acetyl-CoA, catalyzed by choline acetyltransferase (ChAT) within cholinergic neurons, is a critical step in regulating muscle movement [4]. After SCI, mtROS-triggered cascades lead to massive death of cholinergic neurons and hinder ACh synthesis, causing disruption of chemical signal transmission between neurons [5,6]. Activation of endogenous neural stem cells (NSCs) to differentiate into cholinergic neurons has emerged as a potential solution to replace lost cholinergic neurons and reconstruct signal conduction pathways, while avoiding the immunological rejection and ethical concerns associated with exogenous NSC transplantation [7,8]. In addition, restoring ACh synthesis by increasing the content of endogenous acetyl-CoA and supplementing exogenous choline is regarded as a potential strategy for reconstructing the cholinergic nervous system. However, the mtROS-triggered pathological microenvironment after SCI not only induces NSCs to deviate toward glial lineages but also oxidatively damages supplemented exogenous choline, leading to failure of SCI repair [9]. Although the importance of eliminating mtROS has been widely recognized, traditional drugs targeting mitochondria have low efficiency and high toxicity, which limits their clinical application [10] The SS-31 peptide is a new class of antioxidant with mitochondrial targeting properties, which mainly accumulates in the inner mitochondrial membrane. This peptide exerts its effects by protecting mitochondrial structure, promoting structural repair, boosting the synthesis of adenosine triphosphate (ATP), and reducing electron leakage and lipid peroxidation, with no negative impacts on healthy mitochondria [11,12].

Nerve regeneration is a complex pathological process after SCI, where the formation of cavities, oxidative stress, and inflammation in the injured area constitute physical and biochemical barriers to neural regeneration [13,14]. Hydrogels, with their superior physical properties and biological functions, offer a strategic solution to these challenges: their injectability enables filling of irregular injury cavities, adhesive groups mediate tight binding with spinal tissue, and self-healing capacity maintains hydrogel structural integrity - three features that synergistically construct a stable microenvironment for neural regeneration. Additionally, hydrogels can be endowed with tailored biological functions as needed [15,16]. However, the preparation of traditional multifunctional hydrogels requires complex multi-step processes and relies on organic solvents and catalysts, leading to high costs, poor reproducibility, and the generation of toxic residues, which severely hinder their clinical application [17,18]. Furthermore, conventional hydrogels exhibit limited drug-loading capacity and sustained-release efficiency, failing to meet the demand for continuous ROS scavenging in the SCI microenvironment.

In summary, constructing an integrated multifunctional hydrogel using strong antioxidants as the matrix through a simple process will become an effective strategy to overcome the above-mentioned problems. Lipoic acid (LA) is a natural strong antioxidant that can efficiently scavenge ROS and upregulate the activity of pyruvate dehydrogenase (PDH) to promote the synthesis of acetyl-CoA [19]. Additionally, the disulfide bond in LA's five-membered ring enables ring-opening polymerization (ROP) via simple thermal initiation [20,21]. Upon cooling, carboxylic acid side chains crosslink through hydrogen bonding to form Poly(lipoic acid) (PolyLA) with a supramolecular polymer network. However, instability of intermolecular disulfide bonds triggers cyclization reactions, leading to material depolymerization and formation of semi-crystalline PolyLA oligomers [22,23]. Therefore, how to construct a multifunctional integrated system with stability, excellent physical properties, targeted scavenging of mtROS and reconstruction of the cholinergic nervous system by using LA as the matrix through a simple process is a major challenge.

In this study, focusing on the synergistic regulation of mtROS scavenging and cholinergic nervous system reconstruction, a multifunctional Poly(LA-Cho)/SS31 (PLCS) hydrogel with injectability, adhesiveness, and self-healing properties was constructed via a one-step synthesis method (Scheme 1). The PLCS hydrogel uses LA and choline bicarbonate (Cho) as the matrix, and is loaded with SS31 to achieve temporal regulation of drug release. The PLCS hydrogel preferentially releases SS31 in the early stage to specifically scavenge mtROS and alleviate mitochondrial dysfunction, followed by sustained release of LA and choline in the later stage, providing a favorable microenvironment and material basis for the survival and functional maintenance of cholinergic neurons. Furthermore, research results demonstrate that PLCS hydrogels significantly promote the differentiation of endogenous NSCs into cholinergic neurons, upregulating the activity of PDH to catalyze the oxidative decarboxylation of pyruvate into acetyl-CoA, thereby effectively ensuring the biosynthesis of ACh. Compared with traditional hydrogels, this one-step synthesis method enables efficient integration of antioxidant and mechanical properties via simple mixing, avoiding the high costs associated with the complex fabrication processes. This strategy provides a novel solution with both high efficiency and clinical translation potential for motor function repair after SCI.

Scheme 1.

Preparation of PLCS hydrogels and their multimodal therapeutic effects in sci repair.

2. Results and discussion

2.1. Synthesis and characterization of PLC and PLCS hydrogel

The disulfide bond within the five-membered ring of LA can undergo thermally initiated ROP to form a crosslinked linear PolyLA hydrogel. However, PolyLA is metastable at room temperature, where molecular thermal motion induces spontaneous cleavage of the disulfide bonds, leading to depolymerization into semi-crystalline oligomers [24,25]. To address this, we used Cho as an efficient and benign deprotonating agent, which reacts in situ with LA to generate a biocompatible ionic liquid. This system stabilizes the disulfide network of PolyLA via hydrogen bonding, effectively suppressing depolymerization [22,26]. As choline is a direct precursor for ACh synthesis, its supplementation not only elevates ACh levels but also modulates neuronal differentiation, proliferation, and migration, thereby supporting the reconstruction of cholinergic neural circuits.

In this study, LA and Cho were mixed in phosphate-buffered saline (PBS) at a molar ratio of 1:0.6 and heated at 65 °C to form the Poly(LA-Cho) (PLC) hydrogel. Upon cooling to 45 °C, SS31 was incorporated homogeneously, yielding the final PLCS hydrogel (Fig. 1A). Notably, neither organic solvents nor catalysts are used in this process, endowing the hydrogel with excellent biocompatibility. The PLCS hydrogel appeared as a transparent yellow gel (Fig. 1B). Scanning electron microscopy (SEM) revealed a porous, loosely interconnected three-dimensional (3D) architecture (Fig. 1C). This structure provides abundant sites for cell adhesion and mechanical support for tissue ingrowth, thereby guiding organized cellular proliferation and tissue remodeling. Furthermore, the interconnected pores facilitate the diffusion and transport of bioactive molecules, thereby enhancing their bioavailability and functional efficacy. In the context of SCI repair, such a 3D scaffold can guide and support regenerating axons across the lesion site, promoting neural regeneration.

Fig. 1.

Synthesis and characterization of the PLCS hydrogel. (A) Schematic illustration of the preparation process of the PLCS hydrogel. (B) Visual representation of hydrogel formation. (C) TEM image revealing the microstructure of the PLCS hydrogel. (D) Time-dependent evolution of the G′ and G″ of the PLCS hydrogel. (E) Temperature sweep rheology showing the thermal stability of G′ and G″ from 25 °C to 80 °C. (F) Photographs demonstrating the excellent adhesion and stretchability of the PLCS hydrogel on a spinal cord. (G) Shear-thinning behaviour and injectability of the PLCS hydrogel, evidenced by the continuous extrusion and formation of the patterned letters “ABC”. (H) Strain sweep of the PLCS hydrogel (I) The rheological properties of the PLCS hydrogel subjected to alternate step strain switched from 1 % to 1000 %. (J) Photographs capturing the self-healing process of the PLCS hydrogel at 37 °C. (K) In vitro degradation profile of the PLCS hydrogel in PBS and H₂O₂ at 37 °C, showing accelerated mass loss under oxidative conditions. (L) Cumulative release of choline from the PLCS hydrogel in PBS and H₂O₂. (M) Cumulative release of SS31 from the PLCS hydrogel in PBS and H₂O₂.

To verify the homogeneous distribution of SS31 within the PLCS hydrogel, scanning electron microscopy–energy-dispersive X-ray spectroscopy (SEM-EDS) was employed for elemental mapping. To avoid interference from the nitrogen element in choline bicarbonate with the characteristic nitrogen (N) signal of SS31, sodium bicarbonate, which contains no nitrogen, was used instead of choline bicarbonate during hydrogel preparation. The distribution of SS31 in the hydrogel was indirectly evaluated by detecting the characteristic sulfur (S) signal of LA and the characteristic nitrogen (N) signal of SS31. The EDS results show that the characteristic element S of the hydrogel matrix and the characteristic element N of SS31 exhibit a completely consistent uniform distribution pattern inside the hydrogel (Fig. S1). This result directly confirms that SS31 has been successfully loaded and uniformly dispersed in the lipoic acid-based hydrogel system.

The synthesis of PLC was monitored and the structure confirmed by ¹H NMR spectroscopy (Fig. S2). Briefly, LA and Cho first undergo a neutralization reaction to form the ionic liquid LA-Cho, which then undergoes ROP to yield the PLC hydrogel. To investigate the internal interaction mechanisms within the PLCS hydrogel, Fourier transform infrared (FTIR) spectroscopy was employed for structural characterization. The FTIR spectra revealed that, after incorporating SS31, the characteristic stretching vibration peaks of carboxyl (–COOH) and hydroxyl (–OH) groups in the hydrogel showed a decrease in intensity or a slight redshift. Additionally, the absorption peaks corresponding to the amide (–NH) and carbonyl (C=O) groups in the SS31 molecules also exhibited corresponding changes (Fig. S3). These spectral shifts suggest the formation of hydrogen bond interactions within the hydrogel network. Furthermore, when considered in conjunction with the ionic structures of PLCS and SS31, indicate the potential presence of electrostatic interactions. In addition, the UV spectrum of the PLC hydrogel shows a new absorption peak at 318 nm, absent in LA (Fig. S4). This peak is attributed to the characteristic absorption of disulfide bonds in the polymer backbone, providing further evidence for successful ROP.

To evaluate the mechanical properties of the PLCS hydrogel, rheological measurements were performed. The results show that the storage modulus (G′) consistently exceeds the loss modulus (G″) throughout the measurement, indicating the formation of a stable viscoelastic solid with stiffness closely matching that of native spinal cord tissue (Fig. 1D). Temperature sweep analysis further shows that the moduli remain stable over a broad temperature range from 25 °C to 80 °C, demonstrating excellent thermal stability (Fig. 1E). Additionally, frequency sweep tests within the angular frequency range of 0.1–100 rad/s reveal that G′ remains significantly higher than G″, confirming the presence of a robust, percolating 3D network capable of elastic recovery under deformation, rather than viscous flow (Fig. S5). These characteristics collectively indicate that the PLCS hydrogel possesses superior mechanical stability and is well-suited for applications in physiological environments.

The gelation temperature range of the PLCS hydrogel was 50–80 °C, and a clear correlation was observed between gelation time and temperature: gelation occurred in approximately 360 min at 50 °C, shortened to about 180 min at 65 °C, and further reduced to around 30 min at 80 °C (Fig. S6A). These results indicate that the PLCS hydrogel has a broad gelation temperature range, making it adaptable to various clinical operating conditions. Additionally, the effect of stirring speed on gelation was negligible. At preset stirring speeds of 400 rpm, 800 rpm, and 1200 rpm, the resulting hydrogels showed no significant difference in elastic modulus, suggesting that the gelation process is not sensitive to stirring conditions (Fig. S6B). To further evaluate the preparation stability and batch-to-batch consistency of the PLCS hydrogel, three parallel batches were prepared at 50 °C, 65 °C, and 80 °C, and their elastic modulus was tested. The results showed that the variation in elastic modulus among batches at the same temperature was ≤5 % (Fig. S6C). This study provides an initial exploration of the clinical applicability and batch-to-batch consistency of the PLCS hydrogel. However, further optimization of preparation and application methods is needed in future work to improve ease of use and ensure greater consistency of samples, thereby ensuring stable and controllable product properties.

Strong adhesion between the hydrogel and spinal cord tissue is critical for therapeutic efficacy. This property allows tight anchoring at the injury site, effectively preventing treatment failure due to displacement. Moreover, it preserves local biochemical exchange, enhancing retention of released bioactive molecules at the lesion site while shielding the injured area from harmful external infiltration [27]. This stabilization of the microenvironment provides sustained support for neural regeneration. PLCS hydrogel adheres tightly to spinal cord tissue and can support the weight of the spinal cord without detachment (Fig. 1F1). Upon gentle pulling of the spinal cord ends, the hydrogel exhibits notable extensibility, maintaining structural integrity and robust adhesion without delamination (Fig. 1F2). This mechanical resilience ensures that the hydrogel remains firmly attached to the spinal cord tissue in vivo, even under physiological movements or postural changes. The superior tissue adhesion of PLCS hydrogel arises from multiple synergistic interactions, which include hydrogen bonding, electrostatic interactions, and metal coordination. These interactions are formed between the –COOH and –COO^- groups in the hydrogel network and functional moieties on the spinal cord tissue surface.

Following SCI, the lesion site typically exhibits an irregular morphology, and the spinal cord tissue is inherently fragile, being surrounded by bony vertebrae and soft tissues, which poses significant challenges for the delivery of therapeutic agents [28]. Injectable hydrogels offer a promising solution by enabling minimally invasive, precise delivery via needle injection directly into the injury site, where they conform intimately to the irregular lesion cavity to exert therapeutic effects [29]. The practical injectability of the PLCS hydrogel is demonstrated by its pronounced shear-thinning behavior: viscosity decreases from approximately 1000 Pa s to 10 Pa s as the shear rate increases from 0 to 100 s⁻¹. Furthermore, the hydrogel can be smoothly extruded through a 20-gauge needle and used to easily write legible “ABC” patterns (Fig. 1G). This behavior arises from the dynamic, shear-induced breaking and rapid reformation of the hydrogel network, a hallmark of shear-thinning materials. The robust shear-thinning response demonstrates that PLCS possesses excellent injectability, making it highly suitable for clinical translation in SCI therapy.

The microenvironment at the SCI site is inherently complex and dynamic. Implanted hydrogels are prone to structural fragmentation under physiological stresses, which can severely limit their therapeutic effectiveness [30]. Furthermore, cracks and voids within the graft may allow uncontrolled infiltration of inflammatory cells, exacerbating local inflammation. Therefore, maintaining the structural integrity of the hydrogel during neural regeneration is crucial. Self-healing hydrogels offer a promising solution by dynamically repairing damage and restoring their original structure, thereby overcoming these challenges [31].

The self-healing behavior of the PLCS hydrogel was systematically evaluated using a combination of rheological analysis and macroscopic observation. Rheological measurements revealed a reversible gel-to-sol transition at a critical strain of approximately 695 %, where the loss modulus (G″) exceeded the storage modulus (G′), indicating network breakdown (Fig. 1H). To simulate extreme mechanical damage, cyclic strain tests were conducted with alternating strains of 1 % and 1000 %. When the strain increased to 1000 %, G′ dropped below G″, reflecting partial dissociation of the network due to disruption of hydrogen bonding and electrostatic interactions. Notably, upon returning the strain to 1 %, G′ recovered to nearly its initial value. The stability of both moduli over multiple cycles further demonstrates the robust self-healing capability of the PLCS hydrogel (Fig. 1I). Macroscopic observations further corroborated this property: when two separate pieces of PLCS hydrogel were brought into contact and incubated at 37 °C, they fully integrated within 4 h, showing seamless fusion at the interface (Fig. 1J). These results collectively demonstrate that the PLCS hydrogel exhibits excellent self-healing performance, enabling it to maintain structural continuity in the dynamic and mechanically challenging environment of the injured spinal cord.

Following SCI, the local microenvironment becomes highly hydrated due to the accumulation of cerebrospinal fluid, interstitial fluid, and inflammatory exudate. Traditional hydrogels are prone to structural damage under specific conditions. Excessive water absorption causes them to swell excessively, which in turn leads to the collapse of their three-dimensional network and ultimately results in a significant decrease in mechanical strength [32]. This compromises their ability to provide sustained physical support for neural regeneration, undermining both spatial stabilization of the lesion cavity and guidance for axonal regrowth. To address this challenge, the swelling behavior of the PLCS hydrogel was evaluated. The results show that, in aqueous solution, the swelling ratio remains below 20 % over 12 h and reaches near equilibrium by 8 h (Fig. S7). This indicates that PLCS hydrogel effectively restricts excessive water penetration, balancing moderate hydration with exceptional structural stability. The hydrogel's resistance to over-swelling enables it to maintain robust mechanical properties in the highly hydrated post-SCI environment. Consequently, PLCS hydrogel can provide durable physical support and sustained topographical guidance within the injury site, fulfilling a critical requirement for neural regeneration and functional recovery. These attributes position PLCS as a promising and reliable biomaterial platform for SCI repair.

The degradation profile of a hydrogel can transform its role from a “static scaffold” to a “dynamic repair partner” by synchronizing with the timeline of neural regeneration, minimizing long-term foreign body reactions, and facilitating host tissue integration. To characterize the degradation behavior of the PLCS hydrogel, its mass loss was monitored in PBS and hydrogen peroxide (H₂O₂) at 37 °C over time. The results show a gradual decrease in hydrogel mass over time, likely attributed to PBS-induced erosion and progressive chain depolymerization. Notably, when H₂O₂ was introduced to simulate the oxidative stress conditions typical of the injured spinal cord microenvironment, the degradation rate was significantly accelerated. After 24 days of incubation, the residual mass of the PLCS hydrogel in H₂O₂-containing PBS was significantly lower than in PBS alone, demonstrating its redox-responsive degradation behavior (Fig. 1K). This redox-sensitive degradation profile suggests spatiotemporal compatibility with the ROS-rich milieu after SCI, enabling environment-responsive material clearance and supporting the dynamic process of neural repair.

To evaluate the drug loading performance of the hydrogel, the initial encapsulation efficiency (EE₀ %) of SS31 was determined. The results showed that the initial encapsulation efficiency of the PLCS hydrogel was approximately 95 %, indicating that almost all of the SS31 was successfully encapsulated (Fig. S8). These results confirm that the PLCS hydrogel possesses high drug loading capacity and formulation stability, providing a solid foundation for subsequent SS31 release and bioactivity studies.

Spatiotemporally controlled drug release from hydrogels, exemplified by rapid delivery during the acute phase and sustained release in the chronic phase, enables on-demand therapeutic delivery that aligns with the dynamic pathophysiology of SCI. This temporal precision helps avoid suboptimal therapeutic outcomes resulting from poorly timed drug availability. To evaluate the release kinetics of the PLCS hydrogel under SCI-mimicking conditions, we exposed it to H₂O₂ to simulate the ROS-rich microenvironment and monitored the release of SS31 and choline. In both PBS and H₂O₂ solutions, SS31 exhibited a marked early preferential release in the initial phase, with accelerated release kinetics under ROS conditions ((Fig. 1L). This phase corresponds to the key therapeutic window during the acute phase of SCI, where mtROS are explosively produced. The targeted scavenging of mitochondrial ROS by SS31 helps alleviate mitochondrial dysfunction, creating a more favorable microenvironment for neuroregeneration. The porous structure of the hydrogel provides diffusion channels for the small molecule SS31; SS31 forms partial non-covalent binding on the hydrogel surface through hydrogen bonding and electrostatic interactions, thus preventing early explosive release.

The choline release experiment results showed that, in a PBS environment, the cumulative release of PLCS hydrogel over 24 days was approximately 70 %; whereas in a H₂O₂ environment, the release rate was significantly accelerated, with a cumulative release of about 90 % over 24 days (Fig. 1M). The sustained release of choline primarily relies on the gradual degradation of the PLCS hydrogel network and the dissociation of ionic bonds, thereby achieving a controlled and sustained release effect. This stable and continuous release characteristic ensures the stable supply of choline over an extended period, providing sufficient precursors for acetylcholine synthesis and supporting the functional recovery of cholinergic neurons. Notably, the choline release duration of approximately 4 weeks aligns to some extent with the repair phase of SCI repair. Yang et al.'s study demonstrated that the number of neural stem cells at the injury site reaches a peak on day 30 post-surgery, with their differentiation efficiency into neurons also simultaneously attaining the highest level [33]. This suggests that the sustained-release properties of PLCS hydrogel not only maintain local drug concentration stability but also provide continuous physiological support during the critical repair window, contributing to neural regeneration and functional recovery. The release kinetics of both SS31 and choline collectively indicate that PLCS hydrogel can dynamically adapt to changes in the injury environment, enhancing regenerative outcomes through sequential and synergistic drug delivery.

2.2. The antioxidant performance of PLC and PLCS hydrogel

Following SCI, elevated levels of ROS at the lesion site initiate and exacerbate secondary injury cascades. Given the potent antioxidant properties of both SS31 and LA, we assessed the radical-scavenging activity of PLC and PLCS hydrogels using the 2, 2′-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) radical cation (ABTS⁺•) assay. Both hydrogels induced a time-dependent fading of solution color at 0, 5, 10, 15, and 20 min (Fig. 2A). Quantitative analysis by UV–vis spectroscopy at 734 nm showed a significant time-dependent reduction in absorbance for PLC (Fig. 2B) and PLCS hydrogels (Fig. 2C). Notably, the PLCS hydrogel exhibited superior radical scavenging efficiency, reducing ABTS⁺• levels by approximately 90 % within 20 min (Fig. 2D).

Fig. 2.

ROS scavenging capabilities of the PLCS hydrogel. (A) Visual color changes of the ABTS⁺• solution after incubation with PLC and PLCS hydrogels over time. (B) and (C) UV–vis absorption spectra of ABTS⁺• solution following incubation with PLC and PLCS hydrogels at different time points. (D) Quantitative comparison of ABTS⁺• scavenging activity between PLC and PLCS hydrogels over time (n = 3). (E) and (F) UV–vis absorption spectra of •OH scavenging activity of PLC, PLCS hydrogels at different time points. (G) Statistical comparison of •OH scavenging efficiencies of PLC and PLCS hydrogels (n = 3). (H) O₂•^- scavenging ability of PLC and PLCS hydrogels over time (n = 3). (I) H₂O₂ scavenging ability of PLC and PLCS hydrogels over time (n = 3). (J) Flow cytometric analysis of intracellular ROS levels using DCFH-DA staining in different treatment groups. (K) Quantification of DCFH-DA fluorescence intensity (n = 3). (L–N), Fluorescent microscopy images of PC12 cells stained with DCFH-DA, DHE, and DAPI after various treatments, along with corresponding fluorescence intensity quantification (n = 3). Statistically significant at ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.

The hydroxyl radical (•OH) is one of the most reactive and damaging ROS, contributing significantly to oxidative tissue injury and impeding functional recovery following SCI. To evaluate the •OH scavenging capacity of PLC and PLCS hydrogels, we used 3,3′,5,5′-tetramethylbenzidine (TMB) as a colorimetric probe that exhibits a characteristic absorbance peak at 652 nm upon oxidation. UV–vis analysis revealed a time-dependent decrease in absorbance at 652 nm for both hydrogels, indicating progressive •OH scavenging. Notably, the PLCS hydrogel demonstrated superior antioxidant activity, scavenging over 85 % of •OH within 20 min. In contrast, the SS31-free PLC hydrogel showed lower scavenging efficiency. These results indicate that the PLCS hydrogel possesses potent •OH scavenging activity and significantly outperforms the PLC hydrogel (Fig. 2E–G).

Superoxide anion (O₂⁻•) is a reactive oxygen species that contributes to inflammation, induces cellular damage, and impedes tissue regeneration. In this study, we evaluated the O₂⁻• scavenging efficiency of PLC and PLCS hydrogels. O₂⁻• was generated via photoirradiation of riboflavin in the presence of methionine, and its consumption was quantified using nitroblue tetrazolium (NBT) as a probe. During co-incubation of the hydrogels with the O₂⁻•-generating system, UV–vis spectra were recorded at 0, 5, 10, 15, and 20 min. Compared to PLC hydrogel, the PLCS hydrogel significantly reduced absorbance at 450 nm, reflecting NBT formazan formation, and scavenged approximately 65 % of O₂⁻• within 20 min, demonstrating its superior and rapid radical clearance capacity (Fig. S9 and 2H).

Furthermore, the H₂O₂ scavenging ability of the hydrogels was assessed using titanium(IV) oxysulfate (TiOSO₄) as a colorimetric probe. Results showed that PLCS cleared H₂O₂ significantly more effectively than PLC (Fig. S10 and 2I). Collectively, these findings highlight the potent broad-spectrum antioxidant activity of the PLCS hydrogel, particularly against key ROS such as O₂⁻• and H₂O₂. This capability positions PLCS as a promising therapeutic platform for modulating the oxidatively stressed microenvironment after spinal cord injury, thereby fostering conditions conducive to neural regeneration.

To further evaluate the ability of PLC and PLCS hydrogels to scavenge intracellular ROS, PC12 cells were used as a neuronal model. To mimic the oxidative stress environment following SCI, PC12 cells were exposed to H₂O₂, and the ROS-scavenging capacities of the hydrogels were assessed using multiple detection methods. Flow cytometric analysis using the DCFH-DA probe revealed that both hydrogels significantly reduced the proportion of ROS-positive cells compared to the H₂O₂ group. Notably, the PLCS group showed a significantly lower proportion of ROS-positive cells than the PLC group (Fig. 2J and K). Fluorescence microscopy confirmed that H₂O₂ exposure markedly increased intracellular green fluorescence, indicating elevated ROS levels. Both hydrogels reduced fluorescence intensity, with PLCS exhibiting significantly greater suppression than PLC (Fig. 2L and M). To assess scavenging of specific ROS subtypes, dihydroethidium (DHE) was used to detect O₂⁻•. DHE-derived fluorescence was significantly reduced in both groups, with PLCS showing greater suppression of O₂⁻• than PLC (Fig. 2L and N).

To further verify the mtROS-scavenging capacity of the PLCS hydrogel, a cellular model with mtROS production induced by rotenone was established for detection. The results showed that the mtROS level in the PLCS hydrogel-treated group was significantly reduced, with a statistically significant difference compared to the rotenone group (Fig. S11). This further indicates that the PLCS hydrogel can effectively scavenge mtROS. These findings directly confirm that the PLCS hydrogel possesses mitochondria-targeted antioxidant activity at the cellular level, which is consistent with the mitochondrial localization characteristic of SS31.

The above results indicate that the PLCS hydrogel exhibits potent in vitro antioxidant activity and significantly outperforms the SS31-free PLC hydrogel in ROS scavenging. This enhanced capacity is likely attributable to the combined effects of SS31 and lipoic acid within the hydrogel matrix, underscoring the therapeutic potential of PLCS in mitigating oxidative stress, attenuating secondary damage, and fostering a more favorable microenvironment for neural repair after SCI.

2.3. PLCS and PLCS hydrogel alleviates mitochondrial dysfunction

Mitochondria play a central role in cellular aerobic respiration. The electron transport chain (ETC), embedded in the inner mitochondrial membrane, transfers electrons derived from nutrient oxidation to oxygen [34]. This process is coupled with proton pumping from the mitochondrial matrix, establishing a transmembrane proton gradient that drives ATP synthesis. During mitochondrial dysfunction, ETC impairment causes electron leakage, leading to the one-electron reduction of oxygen and ROS generation. These reactive species can attack unsaturated fatty acids in cellular membranes, triggering lipid peroxidation and ultimately cell death [35,36]. As a mitochondria-targeted antioxidant peptide, SS31 requires rapid release from the hydrogel and precise subcellular localization to exert its full therapeutic effect. In vitro release experiments confirmed that SS31 is rapidly released from PLCS hydrogels in the initial phase, enabling prompt therapeutic action during the early stages of injury [12,37]. However, whether SS31 retains its mitochondrial targeting capability after release remains unclear.

Following SCI, mitochondrial dysfunction contributes to neuronal death and functional impairment, which underlie severe sequelae such as paralysis and sensory deficits. Therefore, promoting neuronal survival is essential for enabling subsequent neural regeneration. To systematically evaluate the neuroprotective effect of the PLCS hydrogel, PC12 cells were subjected to H₂O₂-induced oxidative stress, and cell viability was assessed by calcein-AM/PI staining. The results showed that both the PLC and PLCS hydrogel groups exhibited significantly greater green fluorescence compared to the H₂O₂-only control, with a marked reduction in red fluorescence (Fig. S12). These findings demonstrate that the PLCS hydrogel protects neural cells from oxidative stress-induced damage, an effect potentially mediated by its ROS-scavenging activity, providing a rationale for its therapeutic application in SCI repair.

To evaluate the mitochondrial targeting of SS31 in PLCS hydrogels, FITC-labeled SS31 was loaded into PLCS hydrogels and co-cultured with PC12 cells. Co-localization of SS31 and mitochondria was assessed by confocal microscopy. After 0.5 h, FITC-SS31 co-localized with mitochondrial fluorescence, indicating rapid targeting of mitochondria. Under H₂O₂-induced oxidative stress, mitochondrial targeting efficiency was significantly enhanced. Time-course analysis (2 and 4 h) showed progressive accumulation of FITC-SS31 in mitochondria (Fig. 3A). Further co-localization analysis showed that the Manders' correlation of SS31-FITC with mitochondria was close to 1, indicating that SS31-FITC has a very high specificity in targeting mitochondria (Fig. S13). To confirm the specific mitochondrial-targeting ability of SS31, fluorescence colocalization analysis was performed between SS31 and markers of the endoplasmic reticulum (ER) and lysosomes (Lyso). The results demonstrated that after 4 h of co-incubation, SS31 showed no obvious colocalization with either ER or Lyso markers, thereby effectively ruling out the possibility of non-specific targeting (Fig. 3B). To further quantitatively analyze the enrichment level of SS31 in mitochondria, liquid chromatography-mass spectrometry (LC-MS) was employed for the accurate quantification of SS31 concentration in mitochondrial fractions. The results showed that in the PBS environment, the concentration of SS31 in the mitochondrial components was approximately 550 pg/mg prot; while under H₂O₂ treatment conditions, the enrichment of SS31 in the mitochondria significantly increased, reaching approximately 670 pg/mg prot (Fig. 3 E and S14). These results confirm the mitochondrial targeting capability of SS31 delivered by the PLCS hydrogel and suggest that oxidative stress enhances its targeting efficiency, highlighting the responsiveness of this delivery system to pathological microenvironments.

Fig. 3.

PLCS hydrogel targets mitochondria and scavenges mtROS to alleviate mitochondrial dysfunction. (A) Confocal images showing the mitochondrial targeting capability of the PLCS hydrogel under different environmental conditions over time. (B) Confocal imaging and colocalization analysis of SS31 with Lyso and ER in PC12 cells after 4 h co-incubation with PLCS hydrogel under PBS and H₂O₂ environments. (C) Confocal images depicting mitochondrial morphology in PC12 cells across treatment groups. (D) TEM analysis revealing ultrastructural changes in mitochondria from PC12 cells in each group. (E) LC-MS analysis of SS31 content in mitochondria following 4 h of co-incubation of PLCS with PC12 cells under PBS and H₂O₂ environments. (F) Intracellular ATP levels in PC12 cells following different treatments (n = 3). (G) Representative confocal images of JC-1 and MitoSOX staining: red fluorescence indicates JC-1 aggregates, green fluorescence indicates JC-1 monomers; MitoSOX signals reflect mtROS levels. (H) Immunofluorescence staining of TOMM20 and dsDNA in PC12 cells after various treatments. Statistically significant at ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.

Excessive ROS can damage mitochondrial membrane integrity, triggering the opening of mitochondrial permeability transition pores. This leads to mitochondrial swelling, morphological disruption, and loss of membrane potential, ultimately promoting apoptosis and functional failure. To assess the protective effect of PLCS hydrogels on mitochondrial structure, changes in mitochondrial morphology were evaluated using MitoTracker Deep Red staining. In the control group, mitochondria exhibited elongated rod-like or elliptical shapes, evenly distributed throughout the cell body, axons, and dendrites, with perinuclear enrichment commonly observed. In the H₂O₂ group, cells adopted a rounded morphology with retracted neurites, and mitochondria displayed marked swelling, loss of structural integrity, and a shift from elongated to spherical morphology. Treatment with the PLC hydrogel led to partial axonal regeneration, reduced mitochondrial swelling, and partial restoration of rod-like mitochondrial morphology. In contrast, the PLCS hydrogel treatment resulted in complete axonal regeneration and significant alleviation of mitochondrial swelling, with most mitochondria regaining their elongated, rod-like morphology (Fig. 3C).

Given the limited resolution of confocal fluorescence microscopy in resolving mitochondrial ultrastructure, transmission electron microscopy (TEM) was used to examine mitochondrial morphology at the ultrastructural level. The results showed that mitochondria in the control group maintained intact rod-like structures with distinct inner and outer membranes, and the inner membrane folded inward to form densely packed mitochondrial cristae. In the H₂O₂ group, mitochondria exhibited a sparse matrix with reduced electron density and evident vacuolization. The cristae were shortened, flattened, or even absent, likely due to inner membrane expansion. In the PLC hydrogel group, mitochondrial swelling was reduced, and cristae structures were partially restored in some mitochondria. Notably, mitochondria in the PLCS hydrogel group exhibited near-normal ultrastructural features, including well-organized cristae and intact inner and outer membrane integrity (Fig. 3D).

Nerve regeneration following SCI is a highly energy-intensive process, with key events such as axonal elongation, myelin sheath reconstruction, and synapse formation relying heavily on mitochondrial energy production. To further evaluate the effects of PLC and PLCS hydrogels on neural cell energy metabolism, ATP levels and mitochondrial membrane potential (ΔΨm) were measured to investigate their protective effects on mitochondrial function. Following H₂O₂ treatment, ATP levels in PC12 cells were significantly reduced, consistent with the observed structural and functional impairment of mitochondria. Treatment with PLC and PLCS hydrogels significantly increased ATP levels with the PLCS group showing values closer to those of the control group (Fig. 3F). These results suggest that both hydrogels help restore cellular energy metabolism, likely by preserving mitochondrial integrity. The superior recovery observed with the PLCS hydrogel may be attributed to the mitochondrial-targeting property of SS31, which enhances its protective efficacy under oxidative stress.

The mitochondrial membrane potential (ΔΨm) is a key parameter of mitochondrial function, and its stability directly influences energy production, metabolite transport, cell survival, and other essential cellular processes. To assess ΔΨm, the JC-1 dye was used to visualize changes in mitochondrial membrane potential. In the control group, high ΔΨm promoted JC-1 accumulation and formation of J-aggregates in the mitochondrial matrix, which emit red fluorescence. In the H₂O₂ group, mitochondrial dysfunction caused membrane depolarization, leading to the dissociation of J-aggregates into monomers. As a result, most PC12 cells exhibited predominant green fluorescence, indicating severe loss of mitochondrial membrane potential. In the PLC hydrogel group, green fluorescence intensity decreased and the proportion of red fluorescence increased, indicating partial restoration of ΔΨm. The PLCS hydrogel group exhibited a further increase in red fluorescence, with levels approaching those of the control group, indicating a more pronounced restoration of mitochondrial membrane potential compared to the PLC group (Fig. 3G and S15A).

The above results demonstrated that both PLC and PLCS hydrogels effectively scavenged cellular ROS. However, it remained unclear whether mtROS production is suppressed during the recovery process. To address this, mtROS levels were assessed using the MitoSOX Red probe. Both hydrogels significantly reduced mtROS levels, with the PLCS group showing a more pronounced decrease in fluorescence intensity (Fig. 3G and S15B).

Integrated analysis of energy metabolism and membrane potential demonstrates that both hydrogels enhance ATP production by mitigating mitochondrial damage and preserving electron transport chain function. Notably, the PLCS hydrogel shows superior efficacy in restoring membrane potential and improving energy metabolism, attributable to the mitochondrial-targeting capability of SS31. These findings provide critical insights into the "mitochondrial protection - energy recovery" axis underlying hydrogel-mediated neural repair and suggest a promising strategy for targeting mitochondrial restoration in SCI therapy.

ROS attack polyunsaturated fatty acids in cellular membranes, initiating a cascade of lipid peroxidation reactions that disrupt membrane integrity and promote cell necrosis. To evaluate the hydrogels' ability to inhibit this process, the C11-BODIPY fluorescent probe was used, with a decreased red-to-green fluorescence ratio indicating elevated lipid peroxidation. Compared to the H₂O₂ group, both PLC and PLCS hydrogel groups exhibited reduced green fluorescence intensity and a higher red-to-green ratio, indicative of attenuated lipid peroxidation. Notably, the PLCS hydrogel group showed the lowest green signal and highest ratio, suggesting superior protective efficacy against membrane lipid peroxidation (Fig. S16).

Mitochondrial dysfunction can lead to the release of mitochondrial DNA (mtDNA) into the cytoplasm, where it acts as a damage-associated molecular pattern (DAMP). Once recognized by glial cells as non-self nucleic acid, cytosolic mtDNA activates the cGAS-STING pathway, thereby triggering neuroinflammation. To assess whether hydrogels suppress mtDNA leakage, double immunofluorescence staining for TOMM20 and dsDNA was performed. The PLCS hydrogel can significantly inhibit the release of mtDNA from mitochondria (Fig. 3H and S15C).

Activation of the cGAS-STING pathway induces the release of pro-inflammatory cytokines, driving a secondary inflammatory cascade that exacerbates spinal cord injury and neuronal apoptosis. To evaluate the hydrogels' ability to mitigate this response, immunofluorescence was used to assess expression of the pro-inflammatory cytokine IL-1β and the anti-inflammatory marker Arg1. Compared with the H₂O₂ group, both hydrogels significantly reduced IL-1β signal and increased Arg1 expression, with the PLCS hydrogel showing a significantly stronger anti-inflammatory effect than the PLC hydrogel (Fig. S17 and S18).

Macrophages, as key immune effector cells following spinal cord injury, play a crucial role in shaping the inflammatory microenvironment and influencing the subsequent tissue repair process. To investigate the immunomodulatory effects of PLCS hydrogel, the transcription levels of inflammation-related genes in RAW264.7 cells were analyzed using quantitative real-time polymerase chain reaction (qRT-PCR). The results showed that, compared with the H₂O₂-treated group, the expression levels of M1-type macrophage pro-inflammatory cytokine genes (TNF-α, IL-1β, and IL-6) were significantly downregulated in the PLCS hydrogel-treated group. Meanwhile, the expression levels of M2-type macrophage-associated anti-inflammatory cytokine genes (IL-10 and ARG-1) were significantly upregulated (Fig. S19). These findings indicate that PLCS hydrogel effectively exerts anti-inflammatory and immunoregulatory effects by suppressing macrophage polarization toward the pro-inflammatory M1 phenotype and promoting activation toward the anti-inflammatory M2 phenotype, thereby creating a favorable inflammatory microenvironment for SCI repair. These findings reveal that PLCS hydrogel achieves efficient regulation of the neuroinflammatory microenvironment after SCI through a cascade mechanism involving "mitochondrial protection - inhibition of mtDNA release - blockage of inflammatory pathways".

2.4. The early antioxidant and anti-inflammatory effects of PLC and PLCS hydrogels in vivo

In vitro experiments demonstrated that the PLCS hydrogel possesses potent antioxidant and anti-inflammatory activities. To further evaluate its in vivo effects, PLC and PLCS hydrogels were implanted at the SCI lesion site, and ROS levels in the injured tissue were assessed 7 days post-implantation. DCFH-DA and DHE staining revealed markedly elevated ROS levels in the Injury group, indicating a robust oxidative burst following SCI. Both hydrogels significantly reduced ROS levels compared to the Injury group, with the PLCS hydrogel showing a more pronounced scavenging effect (Fig. 4A, B, E, and F).

Fig. 4.

In vivo antioxidant and anti-inflammatory effect of PLCS hydrogel at 7 days post-implantation in a SCI model. (A) and (B) Representative confocal images of DCFH-DA and DHE staining in spinal cord tissues, indicating ROS levels at the lesion site. (C) and (D) Immunofluorescence staining of IL-1β and Arg1 showing inflammatory responses at the injury site. (E–H) Quantitative analysis of fluorescence intensity for DCFH-DA, DHE, IL-1β, and Arg1 (n = 5). (I) Confocal images of cleaved caspase-3 (C-casp3) and NeuN co-staining in spinal cord tissue sections across treatment groups. (J) Quantitative analysis of (I) (n = 5). (K) Representative image of the Western blot analysis of TOMM20, NeuN, cleaved caspase-3 (C-casp3), and GAPDH at the injury site in spinal cord tissues. (L–N) quantification of relative protein expression levels to GAPDH: TOMM20, NeuN, and C-casp3 (n = 3). Statistically significant at ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001. (ns = not significant).

To evaluate neuroinflammation in the injured spinal cord, protein expression of the anti-inflammatory marker Arg1 and the pro-inflammatory cytokine IL-1β was assessed 7 days post-treatment. Compared with the Injury group, both hydrogels significantly reduced IL-1β expression (Fig. 4C and G) and increased Arg1 levels (Fig. 4D and H), with the PLCS hydrogel exhibiting a stronger anti-inflammatory effect than the PLC hydrogel. This enhanced efficacy is likely attributable to the PLCS hydrogel's ability to rapidly release SS31 during the early phase, which effectively scavenges mtROS and thereby suppresses the inflammatory cascade initiated by mitochondrial dysfunction.

To assess the neuroprotective effect of the PLCS hydrogel, double immunofluorescence staining for cleaved caspase-3 and NeuN was performed. The Injury group exhibited extensive neuronal apoptosis at the lesion border, as indicated by numerous cleaved caspase-3⁺/NeuN⁺ cells. In contrast, both hydrogel-treated groups showed significantly fewer apoptotic neurons, with the PLCS hydrogel demonstrating the greatest neuroprotective effect (Fig. 4I and J). Western blot analysis revealed that cleaved caspase-3 expression was significantly upregulated in the Injury group, while levels of the mitochondrial outer membrane protein TOMM20 and the neuronal marker NeuN were downregulated. Treatment with PLC and PLCS hydrogels reversed these changes, with the PLCS hydrogel more effectively restoring NeuN and TOMM20 protein levels (Fig. 4K–N).

Mechanistically, the PLCS hydrogel may inhibit neuronal death through two complementary pathways: first, SS31 accumulates in neuronal mitochondria and scavenges mtROS, thereby suppressing the intrinsic apoptotic pathway; second, it inhibits mtDNA release, blocking the cGAS-STING pathway, attenuating neuroinflammation, and reducing extrinsic apoptotic signaling. In summary, the PLCS hydrogel exerts significant neuroprotection after SCI by combining mitochondrial protection with inflammation modulation, offering a multifaceted therapeutic strategy for SCI.

2.5. The potential mechanism of PLCS hydrogel in repairing SCI

To gain mechanistic insights into the therapeutic effects of the PLCS hydrogel in SCI repair, RNA sequencing (RNA-seq) analysis was performed. Heatmap and volcano plot analyses revealed a substantial number of differentially expressed genes (DEGs) between the PLCS hydrogel-treated group and the injury control group at 4 weeks post-surgery. Using stringent criteria for differential expression (|log₂ fold change| > 1 and adjusted PValue <0.05), we identified numerous DEGs, among which 909 significantly upregulated and 1655 downregulated representative genes are displayed in Fig. 5A and B. Functional annotation indicated that these genes are predominantly enriched in key biological processes critical for neural regeneration, including neuronal migration, differentiation, axonal growth, and synaptogenesis, suggesting that the PLCS hydrogel promotes functional recovery by modulating these regenerative pathways.

Fig. 5.

RNA sequencing identifies potential mechanisms underlying PLCS hydrogel-mediated spinal cord injury repair at 4 weeks post-surgery. (A) Heatmap showing the expression profiles of differentially expressed genes across treatment groups. (B) Volcano plot illustrating the significance and fold change of differentially expressed genes; upregulated genes (red), downregulated genes (blue), and non-significant genes (gray). (C) GO enrichment analysis of biological processes, cellular components, and molecular functions associated with differentially expressed genes. (D). KEGG pathway enrichment analysis displaying significantly enriched pathways among differentially expressed genes. (E) Representative Western blot images of key proteins in the PI3K/AKT signaling pathway. (F) and (G) Quantitative analysis of p-PI3K/PI3K and p-AKT/AKT (n = 3). Statistically significant at ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.

To functionally annotate DEGs in spinal cord tissues, gene ontology (GO) enrichment analysis was performed comparing the PLCS hydrogel-treated and Injury groups. The results revealed that numerous DEGs were enriched in biological processes such as neural stem cell proliferation, central nervous system differentiation, and cholinergic synaptic transmission, as well as molecular functions including protein binding, synapse assembly, and axonogenesis (Fig. 5C). To identify signaling pathways associated with these DEGs, KEGG pathway enrichment analysis was performed. This analysis revealed significant enrichment of DEGs in the PI3K-Akt signaling pathway. Notably, this pathway is known to promote NSC differentiation into neurons and suppress astrocytic over proliferation. The extracellular matrix (ECM)-receptor interaction pathway, also enriched, regulates key neural processes such as cell migration, differentiation, and regeneration. These findings suggest that PI3K-Akt and ECM-receptor signaling may represent key mechanisms mediating the therapeutic effects of the PLCS hydrogel (Fig. 5D). Western blot analysis further confirmed that the expression levels of P-PI3K/PI3K and P-Akt/Akt proteins in the PLCS hydrogel group were significantly higher than those in the Injury group (Fig. 5E–G). These results indicate that PLCS hydrogel may induce the differentiation of NSCs into neurons by activating the PI3K-Akt signaling pathway.

Under oxidative stress conditions, the differentiation of NSCs tends to skew toward glial cells, while their differentiation into neurons is inhibited. To validate the conclusion from RNA sequencing that PLCS hydrogel can promote the differentiation of NSCs into neurons, this study first investigated the regulatory capacity of PLC and PLCS hydrogels on NSC differentiation into neurons at the cellular level. Specifically, NSCs isolated from the spinal cords of neonatal mice were co-cultured with PLC or PLCS hydrogel for 5 days in a differentiation medium under H₂O₂-induced oxidative stress conditions, and the expression differences of neural differentiation-related markers were detected by immunofluorescence staining. The results showed that few Tuj1-positive cells were observed in NSCs of the H₂O₂ treated group; in contrast, the number of Tuj1-positive neurons increased significantly in the PLC and PLCS hydrogel groups. Statistical analysis indicated that PLCS hydrogel exerted a stronger effect in promoting the differentiation of NSCs into the neuronal lineage (Fig. 6A and S20A).

Fig. 6.

PLCS hydrogel promotes the neuronal differentiation of endogenous NSCs and the biosynthesis of cholinergic neurons. (A) Representative confocal images of in vitro NSC differentiation: double immunofluorescence labeling for Nestin (green, NSC marker) and Tuj1 (red, neuronal marker) after 5 days of treatment. (B) Representative confocal images of in vitro NSC differentiation: double immunofluorescence labeling for Nestin and ChAT (red, cholinergic neuronal marker) after 5 days of treatment. (C) Mechanism diagram of PLCS hydrogel promoting ACh synthesis in cholinergic neurons. (D–G) Measurement of cholinergic biosynthesis key markers in spinal cord tissue at the injury site 4 weeks post-implantation: ACh, ChAT, acetyl-CoA, and choline levels (n = 8). (H) and (I) Representative confocal images showing endogenous NSC differentiation into neurons (Nestin⁺/Tuj1⁺) and cholinergic neurons (Tuj1⁺/ChAT⁺) after 4 weeks of PLCS hydrogel implantation. (J) Quantification of Nestin⁺/Tuj1⁺ double-positive cells in (H), (n = 5). (K) Western blotting analysis of the levels of Nestin, ChAT, Tuj1, and GAPDH in spinal cord tissue at SCI sites after 4 weeks of implantation. (L), (M), and (N) Quantification of the relative expression of Nestin, ChAT, and Tuj1 to GAPDH, (n = 3). Statistically significant at ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.

Cholinergic neurons in the spinal cord anterior horn regulate skeletal muscle contraction via acetylcholine release and serve as key mediators of somatic motor function. To further investigate whether there are cholinergic neurons among the newly generated neurons, double immunofluorescence staining for Nestin and ChAT was performed. Compared with the H₂O₂ group, both PLC and PLCS hydrogel groups contained a higher number of ChAT-positive cells, with a relatively higher proportion in the PLCS hydrogel group. These findings suggest that the PLCS hydrogel enhances the generation of cholinergic neurons, potentially by promoting NSC differentiation through microenvironmental improvement (Fig. 6B and S20B).

To further verify the role of the PI3K-Akt signaling pathway, a specific PI3K-Akt pathway inhibitor was used for intervention. Additionally, to specifically maintain the ROS environment in mitochondria, rotenone, a mitochondrial ROS inducer, was used to replace hydrogen peroxide. The results showed that the use of rotenone did not reverse the effect of PLCS on the differentiation of NSCs into cholinergic neurons, further supporting the role of PLCS in regulating mitochondrial homeostasis. After inhibitor treatment, the directed differentiation of NSCs into the cholinergic neuronal subtype induced by PLCS was significantly inhibited (Fig. S21). Collectively, these results indicate that the PLCS hydrogel regulates the differentiation of NSCs into cholinergic neurons by activating the PI3K–Akt signaling pathway. Notably, the transcriptome analysis in this study has certain limitations. the bulk RNA-seq method used in this study provides a gene expression profile at the population level, which cannot directly distinguish gene expression changes in different cell subpopulations within the tissue. To address this limitation, future research will incorporate single-cell RNA sequencing to more precisely understand the molecular responses of different cell types (such as NSCs, neurons, and glial cells) after PLCS treatment. This will provide a comprehensive view of how PLCS affects gene expression across different cell types.

Mechanistically, the PLCS hydrogel ameliorates energy metabolism disorders and cholinergic neuron loss after SCI, while also restoring impaired ACh synthesis. ACh biosynthesis requires two essential precursors: acetyl-CoA and choline. Acetyl-CoA is primarily generated through the oxidative decarboxylation of pyruvate, catalyzed by the PDH complex. Notably, LA, a key component of the E2 subunit of the PDH complex, may support PDH function and thereby promote Acetyl-CoA production when supplied exogenously. Meanwhile, choline released from Cho can form ionic bonds with the COO⁻ of the PolyLA side chains via electrostatic interactions. This enables sustained choline release, ensuring a stable supply for ACh synthesis (Fig. 6C).

The primary function of cholinergic neurons is to synthesize and release ACh to mediate synaptic transmission in motor pathways. To evaluate the functional maturation of newly generated cholinergic neurons, ACh levels in spinal cord tissue at the injury site were measured. Results showed that ACh concentrations were significantly higher in both the PLC and PLCS hydrogel groups compared to the Injury group, with the PLCS group exhibiting levels approaching those of the Sham group (Fig. 6D). These results suggest that the PLCS hydrogel promotes both the generation of cholinergic neurons and the restoration of ACh synthesis capacity in the injured spinal cord.

To verify the potential mechanism by which the hydrogel promotes the synthesis of ACh, PDH activity and acetyl-CoA levels in spinal cord tissues at the injury site were measured 4 weeks post-treatment. Both parameters were significantly higher in the PLC and PLCS hydrogel groups than in the Injury group, with the PLCS group showing levels approaching those of the Sham group (Fig. 6E and F). Acetyl-CoA serves not only as a precursor for acetylcholine synthesis but also as the primary substrate for the tricarboxylic acid cycle. Within mitochondria, it initiates the TCA cycle by condensing with oxaloacetate, leading to NADH and FADH₂ production, which drive oxidative phosphorylation and ATP generation. This pathway may contribute to the observed increase in cellular ATP levels mediated by the hydrogels. To assess choline delivery, tissue choline levels were measured. The PLCS hydrogel significantly increased choline content at the injury site, suggesting effective release and retention that may help counteract choline deficiency after SCI (Fig. 6G).

In summary, the PLCS hydrogel enhances PDH activity to boost acetyl-CoA production, supporting acetylcholine synthesis and fueling the TCA cycle for ATP generation. Sustained choline release further ensures precursor availability for neurotransmitter biosynthesis. These metabolic and biochemical improvements create a positive feedback loop: enhanced energy supply promotes NSC differentiation into functional neurons, whose restored activity may in turn stabilize local metabolic homeostasis and neural circuit function.

The ability of the PLCS hydrogel to promote NSC differentiation into neurons has been demonstrated at the cellular level. To further evaluate its efficacy and long-term effects in promoting endogenous NSC differentiation in vivo, spinal cord tissues at the injury site were analyzed by immunofluorescence staining 4 weeks after treatment. In the Injury group, extensive cavitation was observed, with only a few Nestin⁺ cells at the lesion edge that failed to migrate inward; Tuj1⁺ neurons were nearly absent. In contrast, numerous Nestin⁺ cells were present at the injury site and in rostral and caudal regions in both the PLC and PLCS groups, indicating activation and bidirectional migration of endogenous NSCs toward the lesion for tissue repair. Compared with the PLC group, the PLCS group exhibited significantly more Nestin⁺ cells and a greater abundance of Tuj1⁺ neurons. Notably, Nestin⁺/Tuj1⁺ double-positive cells were observed, suggesting that endogenous NSCs undergo neuronal differentiation in vivo under PLCS hydrogel induction (Fig. 6H and J).

To assess the regeneration of cholinergic neurons, double immunofluorescence staining for Tuj1 and ChAT was performed on spinal cord tissues. Tuj1⁺/ChAT⁺ double-positive neurons were observed at the injury site in both the PLC and PLCS hydrogel groups, indicating the formation of cholinergic neurons at the injury site (Fig. 6I and S20C). To further confirm whether NSCs at the injury site differentiated into cholinergic neurons, double immunostaining for Nestin and ChAT was performed. The appearance of Nestin⁺/ChAT⁺ double-positive cells further supported the differentiation of NSCs into cholinergic neurons at the injury site (Fig. S22). These results suggest that the PLCS hydrogel promotes the regeneration of cholinergic neurons by enhancing the migration and differentiation of endogenous neural stem cells, thus aiding in spinal cord repair after injury.

To assess the expression of neuroregeneration-related proteins, Western blotting was performed to detect Nestin, Tuj1, and ChAT. Nestin expression was low in the Sham group and progressively increased from the Injury group to the PLC hydrogel group and further in the PLCS hydrogel group. This indicates that spinal cord injury induces a certain level of NSC proliferation, and hydrogel treatment enhances this response, with the PLCS group showing the most robust activation. Tuj1 and ChAT expression remained low in the Injury group but was significantly higher in both the PLC and PLCS hydrogel groups. Notably, Tuj1 levels in the PLCS group approached those observed in the Sham group (Fig. 6K–N).

2.6. PLCS hydrogels promoted the recovery of nerve conduction and motor function

To systematically evaluate the efficacy of the PLCS hydrogel in promoting neurological recovery after SCI, multidimensional assessments were performed, including behavioral tests, neuroelectrophysiological evaluations, and locomotor functional scoring. Footprint analysis, a well-established behavioral method for assessing motor recovery, quantifies key gait parameters such as hindlimb step length, step width, and interlimb coordination, providing an objective measure of functional improvement during locomotion.

Eight weeks after hydrogel implantation, hindlimb footprints during walking were collected from rats in each group. The results showed the follows. In the Sham group, step lengths were uniform and stable, step widths were narrow and symmetrical, and toe spread was moderate. Full plantar contact and linear walking trajectories were observed, with intact forelimb-hindlimb coordination. In the Injury group, hindlimbs were completely paralyzed, leaving only smear marks from dragging, with no discernible footprints. In the PLC group, hindlimbs were mobile but non-weight-bearing, with irregular drag patterns, absent interlimb coordination, increased step width, and loss of toe spread. In the PLCS group, hindlimbs showed partial weight-bearing, with clear landing imprints and visible pressure distribution on the soles. Step length was significantly improved, although mild dragging, compensatory gait patterns, and impaired coordination persisted (Fig. 7A–C).

Fig. 7.

Motor function recovery and spinal cord tissue repair in SCI rats treated with PLCS hydrogel over 8 weeks. (A) Representative footprint images for assessing hindlimb motor function recovery. (B and C) Quantitative analysis of stride length and sway distance to evaluate locomotor recovery. (D) Representative electrophysiological traces from left and right sciatic nerves after cortical stimulation. (E and F) Latency and amplitude of MEPs in the left hindlimb. (G and H) Latency and amplitude of MEPs in the right hindlimb. (I) Weekly BBB scores over 8 weeks post-injury (n = 5). (J) Representative images showing rat behavioral responses to photothermal stimulation. (K) Response time to photothermal stimulation (n = 5). (L) H&E staining of longitudinal spinal cord sections. (M) H&E and α-SMA immunofluorescence staining of bladder tissues. (N) Bladder volume across groups. (O) Bladder wall thickness. (P) α-SMA fluorescence intensity in bladder tissue. (Q) Wet weight of gastrocnemius muscle. Statistically significant at ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.

Motor evoked potential (MEP) recording assesses the integrity and functional status of motor pathways and reflects the degree of functional recovery after spinal cord injury. MEP amplitude correlates with the number of functional axons, while latency reflects conduction velocity. In the Injury group, bilateral MEP waveforms were flat, indicating complete disruption of motor pathways and absence of functional reconnection. Compared with the Injury group, both the PLC and PLCS hydrogel groups exhibited higher MEP amplitudes and reduced latencies. Notably, the PLCS group showed significantly higher amplitudes and shorter latencies than the PLC group, although both parameters remained below Sham group levels (Fig. 7D–H). These neurophysiological improvements suggest that the PLCS hydrogel enhances axonal conduction across the injury site, likely through promotion of synaptic and myelinated axon regeneration. Collectively, these findings demonstrate that the PLCS hydrogel promotes functional reconstruction of neural circuits after SCI, as evidenced by improved signal intensity and conduction velocity.

The Basso-Beattie-Bresnahan (BBB) locomotor rating scale was used to assess motor function recovery weekly after SCI, based on hindlimb joint movement, weight-bearing capacity, and coordination. The Injury group maintained a BBB score of approximately 2 throughout the 8-week period, indicating minimal spontaneous motor recovery. In contrast, both the PLC and PLCS hydrogel-treated groups exhibited progressively increasing scores. Specifically, rats in the PLC group achieved weight support but lacked coordinated hindlimb stepping, reaching a final score of approximately 6. In the PLCS group, animals displayed improved ankle mobility and intermittent plantar contact during locomotion. Despite persistent gait incoordination and hindlimb adduction, the peak score reached approximately 8, indicating significantly greater functional improvement (Fig. 7I).

To evaluate the recovery of sensory function, the hot plate test was performed. In this experiment, a constant thermal stimulus temperature of 55 °C was set, and the reaction latency was recorded, with the endpoint defined as the occurrence of pain-evasive responses in rats such as frequent paw lifting and licking (Fig. 7J). The results showed that rats in the Injury group had no response to thermal pain stimulation, indicating that their sensory function was not restored. In contrast, rats in both the PLC and PLCS hydrogel groups responded to thermal pain stimulation, and the reaction time of the PLCS hydrogel group was closer to that of the Sham group (Fig. 7K). These findings suggest that PLCS hydrogel can more effectively promote the recovery of sensory function and contribute to the reconstruction of damaged neural sensory conduction pathways.

To evaluate spinal cord tissue repair, hematoxylin-eosin (HE) staining was performed. The Sham group showed intact spinal cord architecture with well-organized gray matter and parallel white matter tracts along the longitudinal axis. In contrast, the Injury group exhibited complete disruption of gray and white matter, disorganized fiber bundles, and large cystic cavities lacking neural tissue. In the PLC group, regenerated tissue filled the lesion site and showed partial longitudinal alignment of neurons and fibers, although disorganization persisted. The PLCS group demonstrated more advanced structural recovery, with restored continuity of gray and white matter, densely packed, well-organized neurons and highly aligned, abundant nerve fiber. Tissue repair in the PLCS group was significantly superior to that in the PLC group and closely resembled normal spinal cord morphology (Fig. 7L).

Bladder dysfunction is a common and severe consequence of SCI, often leading to incontinence, urinary tract infections, and even renal failure. To assess bladder tissue recovery, HE staining was performed on bladder sections from each group. In the Sham group, the bladder exhibited a well-defined histological architecture, with numerous irregular longitudinal and reticular mucosal folds creating a corrugated surface, and detrusor muscle fibers were densely and regularly arranged. Both the Injury and PLC groups showed severe bladder wall stretching, marked thinning, and disorganization of mucosal and muscular layers. In contrast, the PLCS group showed significant mitigation of these changes, with partial restoration of mucosal folds and improved organization of detrusor muscle fibers, indicating superior tissue repair compared to the PLC group (Fig. 7M and N).

To assess the distribution and morphology of bladder smooth muscle cells, α-smooth muscle actin (α-SMA) immunofluorescence staining was performed. In the Sham group, strong and uniform red fluorescence indicated well-organized detrusor muscle fibers with clear boundaries and minimal interstitial tissue, reflecting structural integrity. In the Injury group, denervation led to severe neurogenic bladder remodeling. Gaps between muscle bundles were markedly widened, and fiber architecture was disorganized. The α-SMA signal was heterogeneous and reduced, with a decreased positive staining area and focal smooth muscle atrophy. These changes indicate significant structural and phenotypic disruption. The PLC group showed partial recovery of α-SMA expression and improved fiber organization. In contrast, the PLCS group exhibited significantly higher fluorescence intensity and a larger positive area, with marked restoration of fiber alignment and thickness (Fig. 7M and P). Quantitative analysis showed a significant increase in bladder volume after SCI. Intervention with PLC/PLCS hydrogels markedly reduced the volume. This is attributed to injury-induced impairment of bladder innervation, which causes detrusor dysfunction, abnormal micturition reflex, urine retention, and compensatory bladder enlargement. The hydrogels restore neural innervation and normal function by promoting repair and remodeling the spinal cord microenvironment, thereby alleviating bladder dilatation. (Fig. 7O).

The gastrocnemius, a major muscle in the posterior leg compartment, is essential for ankle plantar flexion, lower limb stability, and locomotion. Following SCI, prolonged denervation leads to muscle atrophy, characterized by fiber thinning, fat infiltration, and connective tissue hyperplasia, contributing to persistent lower limb dysfunction and impaired rehabilitation. To assess atrophy, gastrocnemius wet weight was measured in all groups. The Injury group showed severe weight loss due to denervation. Both the PLC and PLCS groups exhibited partial recovery, with the PLCS group restoring muscle weight to a level closer to that of the Sham group (Fig. 7Q).

2.7. PLCS hydrogel restored the microstructure of the injured spinal cord

To systematically evaluate neural regeneration 8 weeks post-treatment, analysis of key indicators such as NSC proliferation, differentiation, and glial scar formation was conducted in spinal cord tissues. In the Injury group, few scattered Nestin-positive cells were observed in the lesion area, with sparse neurite outgrowth, indicating impaired self-repair capacity and stalled neuroregeneration. In the PLC and PLCS groups, although Nestin-positive cell numbers decreased compared to 4 weeks post-treatment, they remained detectable, and more Nestin/Tuj1 double-positive cells were observed in the PLCS group. These findings indicate that 8 weeks post-treatment, the PLCS hydrogel continues to activate endogenous NSCs, promoting their proliferation and directed neuronal differentiation (Fig. 8A and D).

Fig. 8.

PLCS hydrogel promotes neuronal differentiation of endogenous NSCs and nerve fiber regeneration after 8 weeks of implantation. (A) Representative immunofluorescence image showing NSC differentiation into neurons, co-labeled for Nestin (green) and Tuj1 (red). (B) Representative image showing inhibition of NSC differentiation into astrocytes, co-labeled for Nestin (green) and GFAP (red). (C) Representative image of synaptic regeneration, showing Synapsin (Syn, red), NF200+ axonal filaments (green), and nuclei (DAPI, blue). (D) Quantitative analysis of Nestin and Tuj1 double-positive cells in (A). (E) Western blot analysis of MBP, NF200, NeuN, and GAPDH protein expression in spinal cord tissues at the injury site. (F–H) Relative protein levels of MBP, NF200, and NeuN to GAPDH in (E) (n = 3). Statistically significant at ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.

Further analysis via Nestin/GFAP double staining revealed that in the injured areas of the PLC and PLCS-treated groups, glial scar formation was significantly inhibited, with few Nestin/GFAP double-positive cells. This indicates that the differentiation of NSCs into astrocytes was suppressed, preventing excessive glial scar hyperplasia from hindering neuroregeneration (Fig. 8B and S23A). In summary, from multiple dimensions including cell proliferation, differentiation, and microenvironment regulation, these findings systematically demonstrate the persistence of PLCS hydrogel in promoting neuroregeneration after SCI.

Synapses are critical for neuronal communication. After SCI, disruption of nerve fibers and synaptic structures at the lesion site impairs neural signal transmission, resulting in paralysis and sensory deficits. Therefore, synaptic repair is essential for functional recovery. Double immunofluorescence labeling of nerve fibers and synaptophysin revealed extensive colocalization, suggesting synaptic reorganization. Compared with the Injury group, both PLC and PLCS groups exhibited dense regenerating nerve fibers and synapses extending into the lesion site. Quantitative analysis showed that the PLCS group had significantly higher neurite density and synapse formation than the PLC group, demonstrating its superior efficacy in reconstructing neural circuits (Fig. 8C and S23B).

In the central nervous system, the myelin sheath surrounding axons forms the structural basis for rapid nerve impulse conduction. Western blot analysis confirmed significantly increased expression of NeuN, NF200, and MBP in the PLCS group compared to other groups, indicating its broad effects on neuronal integrity and myelination (Fig. 8E–H). In summary, by creating a 3D regenerative microenvironment that promotes axonal growth, synaptogenesis, and remyelination, the PLCS hydrogel overcomes the three major barriers to neural regeneration after SCI, offering a strategy that bridges structural repair with functional recovery in central nervous system injury.

After SCI, immune cell infiltration and fibrosis are major obstacles to repair. Early immune responses help clear cell debris and pathogens from the injury site, but excessive immune cell infiltration, particularly the sustained presence of M1 macrophages, exacerbates inflammation and inhibits neuronal regeneration. Fibrosis, caused by the overactivation of fibroblasts and collagen deposition, further impedes the regeneration of neuronal axons. Effectively controlling immune cell infiltration and fibrosis is crucial for SCI repair. To assess the long-term immune response and fibrosis after PLCS hydrogel implantation, Western blot analysis was performed on spinal cord tissue samples from each group. Compared to the injury group, the PLCS hydrogel group showed significantly reduced expression of the fibrosis marker Col1 and the M1 macrophage marker CD86, while the expression of the M2 macrophage marker CD206 was significantly increased. These results indicate that PLCS hydrogel promotes the polarization of macrophages toward the anti-inflammatory M2 phenotype and effectively suppresses post-injury tissue fibrosis. Additionally, compared to the sham group, the expression levels of Col1 and CD86 in the PLCS hydrogel group did not show significant elevation, suggesting that the material does not induce abnormal immune responses in vivo (Fig. S24). Collectively, these findings demonstrate that PLCS hydrogel possesses good biocompatibility, modulates the local immune microenvironment, alleviates fibrosis, and provides a favorable tissue environment for SCI repair.

2.8. Biocompatibility of the PLCS hydrogels

To evaluate the biocompatibility of the PLCS hydrogel, PC12 cells were seeded on its surface for in vitro culture, and cell viability was assessed over time using Annexin V-FITC/PI double staining. The results showed no significant difference in either early apoptosis rate or late apoptosis/necrosis rate between the PLCS and control groups, confirming that PLCS hydrogel does not induce cell apoptosis (Fig. 9A and D). Further evaluation using calcein-AM/PI staining. The results showed that green fluorescence intensity generated by the hydrolysis of calcein-AM by intracellular esterases increased significantly over time (days 1, 3, and 5), accompanied by a gradual increase in cell density. No obvious red PI fluorescence was detected at any time point, suggesting that the PLCS hydrogel is non-cytotoxic and supports cell proliferation (Fig. 9B and E). In addition, CCK-8 assay results demonstrated that cell viability remained unaffected after co-culture with the PLCS hydrogel for 1, 3, and 5 days, indicating that the hydrogel provides a favorable environment for cell growth (Fig. 9F).

Fig. 9.

Biocompatibility evaluation of the PLCS hydrogel. (A) Apoptosis analysis of PC12 cells treated with extract solutions containing different concentrations of PLCS hydrogel. (B) Live/dead staining of PC12 cells co-cultured with PLCS hydrogel for 1, 3, and 5 days. (C) Scratch wound-healing assay and cytoskeletal staining after co-incubation with PLCS hydrogel. (D) Quantitative analysis of apoptosis rates corresponding to (A). (E) Quantitative analysis of cell viability corresponding to (B). (F) Cytotoxicity assessment of PLCS hydrogel using the CCK-8 assay. (G) Quantitative analysis of cell migration rates corresponding to (C). (H–M) Blood biochemical parameters of rats, including WBC, HGB, AST, ALT, BUN, and CREA, following different treatments. (N) H&E staining of major organs harvested from rats after 8 weeks of treatment. Statistical significance: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, ns, not significant.

To comprehensively evaluate the effect of PLCS hydrogel on cell migration activity, scratch assays were performed. The PLCS group exhibited significantly accelerated wound healing with nearly complete closure at 24 h, which was markedly superior to the control group (Fig. 9C and G). Meanwhile, through F-actin staining, it was found that the cells in the PLCS group could all spread well (Fig. 9C). Collectively, these results demonstrate that the PLCS hydrogel supports high cell viability without impairing cell migration, exhibiting excellent cytocompatibility and regenerative potential.

To evaluate the systemic biosafety of PLCS hydrogel, blood biochemical post-implantation. The results showed that hemoglobin (HGB), white blood cell (WBC) counts and renal function indicators (Blood Urea Nitrogen, BUN and creatinine, CREA) remained within normal physiological ranges. Although slight variations were observed among groups, there were no statistically significant differences between the PLCS, PLC, and sham groups. Similarly, liver function indicators, including alanine aminotransferase (ALT) and aspartate aminotransferase (AST), showed no significant intergroup differences and remained within normal ranges (Fig. 9H–M). Histopathological analysis of major organs (heart, liver, spleen, lung, kidney) from SCI rats treated with PLC or PLCS hydrogels for 8 weeks revealed no pathological abnormalities in any group, further supporting the good biocompatibility of the PLC and PLCS hydrogel (Fig. 9N). These findings indicate that PLCS hydrogel implantation did not induce hepatic or renal toxicity or hematological abnormalities in vivo, demonstrating its excellent systemic biocompatibility and biosafety.

3. Conclusion

SCI is a debilitating condition that can lead to neurological deficits ranging from transient to permanent, representing a growing challenge in global healthcare. Traumatic SCI initiates a cascade of secondary injury processes, including direct neuronal damage, progressive loss of neurons and glial cells, tissue ischemia, and sustained inflammation. To address the synergistic regulation of mtROS clearance and cholinergic system reconstruction, we developed a multifunctional PLCS hydrogel using a one-step fabrication strategy. The PLCS hydrogel promotes the directed differentiation of endogenous NSCs into cholinergic neurons and facilitates ACh biosynthesis by upregulating PDH activity, thereby enhancing acetyl-CoA production. This dual approach establishes a synergistic "cell regeneration–metabolic remodeling" strategy for ameliorating motor dysfunction following SCI, offering a highly efficient and translationally promising therapeutic avenue.

This PLCS hydrogel enables SS31 delivery via local injection at the SCI site, thereby avoiding the off-target effects and metabolic burden associated with systemically administered mitochondria-targeted antioxidants (e.g., MitoQ, SS peptides). Notably, the uptake of SS31 is independent of mitochondrial membrane potential (ΔΨm), allowing it to adapt to the pathological microenvironment of SCI. Furthermore, the hydrogel is biodegradable, does not induce obvious inflammatory responses, and exhibits excellent biocompatibility.

Compared with hydrogel-based SCI therapies (e.g., PCL, GelMA), the PLCS hydrogel features a green and simple preparation process without organic solvents or catalysts, facilitating large-scale production. It integrates three synergistic functions, thus overcoming the limitations of traditional single-functional hydrogels and achieving a balance between "easy preparation" and "high therapeutic efficacy."

Unlike existing cholinergic regeneration strategies, which primarily induce NSC differentiation into ChAT-positive cholinergic neurons at the phenotypic level while overlooking ACh synthesis, a core prerequisite for functional circuit reconstruction, the PLCS hydrogel achieves dual regulation of "differentiation + functional maturation" through the synergy of mitochondrial protection and microenvironment remodeling. It enhances ACh synthesis by ensuring acetyl-CoA supply (via PDH upregulation) and maintaining ChAT activity, addressing a critical gap in current research.

A limitation of this study is the absence of a control group for individual components in the in vivo experiments. To address this, future research will include single-component intervention groups to systematically compare the biological effects of the PLCS composite with those of each individual component. This will help identify the key active ingredients and verify whether they exert their effects through the same "mtROS scavenging - PI3K-AKT pathway modulation" mechanism. Such studies will provide more comprehensive experimental support for target analysis, the elucidation of the pharmacokinetic basis, and the clinical translation and application of PLCS.

4. Experimental section

4.1. Materials of the PLCS hydrogel

α-Lipoic acid was purchased from Macklin Biochemical Technology Co., Ltd. (Shanghai, China). Choline bicarbonate was obtained from Shanghai KEIWEI Chemical Technology Co., Ltd (Shanghai, China). SS31 peptide and SS31-FITC peptide were supplied by Hangzhou All Peptide Biotechnology Co., Ltd. (Hangzhou, China).

4.2. Synthesis and characterization of PLC, PLCS, and PLCS-FITC hydrogels

Briefly, 1 mmol of LA and 0.6 mmol of Cho were dissolved in 1 mL of PBS. The solution was then transferred to a 65 °C water bath and stirred magnetically at 800 rpm for 3 h to ROP. After the reaction, the system was cooled to 45 °C, at which point 2 mg of SS31 peptide or FITC-labeled SS31 was added for the preparation of drug-loaded hydrogels. The mixture was gently stirred and further cooled to room temperature until gelation was complete, yielding PLCS or PLCS-FITC hydrogels. The SS31-free control, prepared without adding SS31, was designated as PLC hydrogel. All hydrogels were stored at 4 °C in the dark prior to use. The chemical structures of the LA, Cho, and PLC were confirmed by ¹H nuclear magnetic resonance (Bruker Avance NEO 400 MHz, Germany) in DMSO-d6 as the solvent. Chemical shifts are reported in ppm versus deuteroxide as an internal standard. To further examine the ROP of LA, UV–vis spectroscopy was performed on a UV-2450/2250 (Shimadzu) spectrophotometer. The absorbance of LA and PLC solutions was measured from wavelength 300–500 nm. FTIR spectroscopy was employed to characterize the chemical structure of the PLCS hydrogel and to verify the formation of its internal crosslinked network and structural stability. Freeze-dried PLCS hydrogel samples were ground into fine powder and evenly spread on the sample stage. Spectra were recorded at room temperature within the wavenumber range of 4000–400 cm⁻¹.

4.3. Scanning electron microscope detection

The PLCS hydrogels were rapidly frozen in liquid nitrogen and sectioned using a cryomicrotome to obtain cross-sectional slices. The sections were then freeze-dried and sputter-coated with gold prior to scanning electron microscopy (TESCAN MIRA LMS, TESCAN, Czech Republic) observation at an accelerating voltage of 3 kV. The internal microstructure of the scaffolds was examined. To study the distribution of SS31 in PLCS, we performed EDS elemental analysis. To avoid interference from the nitrogen element in choline bicarbonate, we replaced it with nitrogen-free sodium bicarbonate, while keeping the other steps unchanged. The HCO₃⁻ in sodium bicarbonate completely replaced that in choline bicarbonate, and its aqueous solution is consistent with the original system, without affecting the stability and distribution of SS31. Sodium bicarbonate does not interfere with the crosslinking of lipoic acid, and through the principle of equimolar cations, ensures the stability of the gel structure without affecting gel strength. Subsequently, the distribution signals of sulfur (S, from LA) and nitrogen (N, from SS31 peptide) in the hydrogel slices were collected, followed by elemental mapping and quantitative analysis to evaluate the dispersion and loading uniformity of SS31.

4.4. Rheological detection

The viscoelastic properties of PLCS hydrogels were evaluated using a rotational rheometer (Haake Mars40, Germany) equipped with a 20 mm parallel plate geometry. A 500 μL aliquot of hydrogel was loaded onto the lower plate, with a gap of 1000 μm. Silicone oil was applied around the sample to prevent dehydration during testing. Time sweep tests were conducted immediately after sample loading at 25 °C, using a 1 % strain and an angular frequency of 10 rad/s, to monitor gelation kinetics and ensure mechanical stability before subsequent measurements. Frequency sweeps (0.1–100 rad/s) at 1 % strain were performed at 25 °C to assess the storage modulus (G′). Temperature sweeps were conducted from 25 °C to 80 °C at a heating rate of 2 °C/min, with a fixed strain of 1 % and angular frequency of 10 rad/s, to evaluate thermal stability. Additionally, shear rate sweeps (1–1000 s⁻¹) at 25 °C were carried out to characterize the shear-thinning behavior of the hydrogels.

4.5. Self-healing performance detection

To assess the self-healing capability of the PLCS hydrogel, cylindrical samples were prepared and cut into two pieces. The freshly cut surfaces were brought into contact and held together at 37 °C without applying external pressure. Reintegration was monitored over time by visual inspection, with complete healing typically observed visually. To quantitatively evaluate the self-healing dynamics, rheological recovery tests were conducted at 37 °C using a rotational rheometer. Strain amplitude sweeps (0.1 %–1000 %, 10 rad/s) were first performed to identify the critical strain, defined as the onset of nonlinear viscoelastic behavior. Subsequently, three-cycle alternating step-strain tests were carried out: the strain was alternated between a small amplitude (γ = 1.0 %, to assess elastic recovery) and a large amplitude (γ = 1000 %, to disrupt the network structure). During each low-strain phase, the recovery of the storage modulus (G′) was monitored to evaluate the hydrogel's ability to self-repair under dynamic mechanical stress.

4.6. Swelling rate detection

The water absorption and swelling behavior of the PLCS hydrogel scaffolds was evaluated using a gravimetric method. Freeze-dried samples were weighed to obtain their initial dry weight (W₀). The samples were then immersed in PBS (Gibco, USA) and incubated at 37 °C. At predetermined time intervals (2, 4, 6, 8, and 12 h), the samples were removed, gently blotted with absorbent paper to remove surface moisture, and immediately weighed to record the swollen weight (W_t). The swelling ratio (SR) was calculated using the following equation: SR(%) = (W_t − W₀)/W₀ × 100. The experiment was conducted three times. This method effectively characterized the water absorption capacity and swelling kinetic behavior of the hydrogel at different time points.

4.7. In vitro degradation

To evaluate the degradation behavior of the PLCS hydrogel under oxidative conditions, a gravimetric method was employed. Briefly, cylindrical hydrogel samples (10 mm diameter × 10 mm height) were weighed to obtain their initial dry weight (W₀). Each sample was then placed in 1 mL of PBS containing 200 μM H₂O₂ to simulate an inflammatory microenvironment. The samples were incubated at 37 °C in a shaker at 100 rpm. At predefined time intervals, the samples were removed, gently blotted with absorbent paper to remove surface liquid, and immediately weighed (W_t). The mass loss ratio (MLR) was calculated using the following equation: MLR(%) = (W₀ − W_t)/W₀ × 100. All experiments were performed in triplicate (n = 3). This approach enabled quantitative assessment of the hydrogel's degradation kinetics and structural stability over time.

4.8. Initial encapsulation efficiency of SS31

To evaluate the drug loading performance of the PLCS hydrogel, the initial encapsulation efficiency (EE₀ %) of SS31 was determined using an indirect method. The SS31-containing precursor solution was first prepared and allowed to gel completely. After the gelation process, the supernatant was carefully collected, and the equipment walls were quickly rinsed with PBS to remove any unencapsulated SS31. The encapsulated SS31 in the hydrogel was then quantified using a FITC-SS31 fluorescence assay. The initial encapsulation efficiency was calculated using the following formula: Encapsulation Efficiency (EE₀ %) = (Total drug amount −Amount of free drug)/Total drug amount × 100 %. This formula calculates the encapsulation efficiency as the percentage of the drug successfully incorporated into the hydrogel, relative to the total amount of drug initially added, subtracting the amount of free drug remaining after the gelation process.

4.9. The release of SS31 and Cho from the PLCS hydrogel

To systematically investigate the release kinetics of SS31 and Cho from the PLCS hydrogel, quantitative analysis was performed using fluorescence spectroscopy. Cylindrical PLCS-FITC hydrogel samples (10 mm diameter × 10 mm height) were fabricated with FITC-labeled SS31 (SS31-FITC). Each sample was immersed in 1 mL of release medium—either PBS or PBS containing 200 μM H₂O₂—and incubated at 37 °C in a shaker at 100 rpm. At predetermined time points (2, 6, 12, 24, 36, 48, 60, and 72 h), the supernatant was collected, and the fluorescence intensity was measured using a multi-mode microplate reader (Multiskan FC, Thermo). The cumulative release of SS31-FITC was calculated based on a pre-established standard curve. In a separate experiment, supernatants were collected from PLCS hydrogel over a longer period (2, 4, 8, 12, 16, 20, and 24 days) to evaluate the release of choline, a degradation product of the PLCS hydrogel, using a choline/acetylcholine fluorescence assay kit. All experiments were performed in triplicate (n = 3).

4.10. ROS scavenging assays

To evaluate the antioxidant properties of PLC and PLCS hydrogels, a quantitative analysis was performed using the 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (Macklin, China) radical (ABTS•⁺) scavenging assay. The specific procedures were as follows: First, 7 mM ABTS solution was thoroughly mixed with 2.45 mM potassium persulfate (K₂S₂O₈) (Macklin, China) solution and reacted overnight at 4 °C in the dark to generate a stable ABTS•⁺ radical stock solution. Subsequently, 1 mL of the ABTS•⁺ radical solution was co-incubated with PLC hydrogel and PLCS hydrogel, separately. Supernatants were collected at 0, 5, 10, 15, and 20 min under 37 °C. The absorbance of the supernatants was measured at 734 nm using a multi-mode microplate reader (Multiskan FC, Thermo), and the scavenging rate of ABTS•⁺ radicals was calculated based on the change in absorbance.

To evaluate the hydroxyl radical (•OH)-scavenging capacity of PLC and PLCS hydrogels, a quantitative assay based on the Fenton reaction was performed. Briefly, •OH generated by the Fenton reaction oxidizes 3,3′,5,5′-tetramethylbenzidine (Macklin, China) (TMB, 0.3 mM) to its oxidized product (oxTMB). The specific procedure was as follows: 10 μM ferrous sulfate (FeSO₄) (Macklin, China) was mixed with 50 μM H₂O₂ and reacted at room temperature for 10 min. The resulting Fenton reaction system was then mixed with TMB solution, followed by co-incubation with PLC or PLCS hydrogel separately. At 0, 5, 10, 15, and 20 min under 37 °C constant temperature, supernatants were collected, and their absorbance was measured at 652 nm using a multi-mode microplate reader (Multiskan FC, Thermo). The •OH scavenging rate was calculated based on absorbance changes. This assay effectively characterizes the hydrogels' antioxidant properties and •OH-scavenging efficiency.

To evaluate the superoxide anion radical (O₂⁻)-scavenging capacity of PLC and PLCS hydrogels, a quantitative assay was performed using the nitroblue tetrazolium (NBT) (Macklin, China) reduction method. The specific procedure was as follows: First, detection solutions were prepared containing 6.67 μM riboflavin (Macklin, China), 4.33 mM L-methionine (Macklin, China), 25 μM NBT, and either PLC or PLCS hydrogel. The mixtures were then irradiated under a 30 W illumination tube for 300 s to generate O₂⁻ via photochemical reduction. Supernatants were collected at 0, 5, 10, 15, and 20 min. The generated O₂⁻ reacts with NBT to form a blue formazan precipitate, which exhibits a characteristic absorption peak at 560 nm. The absorbance of the reaction system was measured at 560 nm using a multi-mode microplate reader (Multiskan FC, Thermo), and the O₂⁻ scavenging rate was calculated based on absorbance changes. This assay effectively characterizes the hydrogels’ O₂⁻-scavenging performance and antioxidant capacity by leveraging the blue product formed from the reaction between NBT and O₂⁻.

To evaluate the H₂O₂-scavenging capacity of PLC and PLCS hydrogels, a quantitative assay was performed using the titanium sulfate (Ti(SO₄)₂) (Macklin, China) colorimetric method. The specific procedure was as follows: First, 0.5 mM H₂O₂ solution was co-incubated with PLC or PLCS hydrogel separately. Supernatants were collected at 0, 5, 10, 15, and 20 min. Subsequently, Ti(SO₄)₂ solution was added to the supernatants. Once the color of the reaction system stabilized, the absorbance was measured at 405 nm using a UV spectrophotometer, and the H₂O₂ scavenging rate was calculated based on absorbance changes. This assay effectively characterizes the hydrogels’ H₂O₂-scavenging performance capacity by leveraging the formation of a yellow complex from the reaction between H₂O₂ and Ti(SO₄)₂. All of the above experiments were repeated three times.

4.11. Detection of total ROS in cells and mitochondrial ROS

DCFH-DA (Servicebio, China) and DHE (Solarbio, China) fluorescent probes were used to evaluate the intracellular ROS-scavenging capacity of PLC and PLCS hydrogels, while MitoSox Red (MCE, USA) probe was applied to detect their scavenging effect on mtROS. The specific procedures were as follows: PC12 cells were seeded at 5 × 10⁴ cells/well into 24-well plates pre-coated with PLC or PLCS hydrogel and cultured overnight in a 37 °C, 5 % CO₂ incubator. Subsequently, the cells were randomly divided into four groups: (1) Control; (2) H₂O₂ (200 μM); (3) H₂O₂ + PLC hydrogel; and (4) H₂O₂ + PLCS hydrogel, with each group cultured for another 24 h. After cultivation, the cells were washed three times with pre-cooled PBS (pH 7.4). For ROS detection, 10 μM DCFH-DA or DHE working solution was added, followed by incubation at 37 °C in the dark for 30 min. For mtROS detection, PC12 cells was incubated with 5 μM MitoSox Red working solution under the same conditions in the dark for 15 min. Fluorescence images were acquired using a Nikon inverted fluorescence microscope (Nikon, Japan). All images were quantitatively analyzed using NIS-Elements AR software (Version 5.11, Nikon) to ensure the accuracy and reproducibility of experimental results. All experiments were repeated three times.

To verify the specific capability of PLCS hydrogel in scavenging mtROS, a rotenone-induced mtROS generation model was employed to evaluate its scavenging effect. PC12 cells were seeded into 24-well plates at a density of 5 × 10⁴ cells per well and incubated overnight at 37 °C with 5 % CO₂. The cells were then subjected to different treatments, including normal culture, rotenone-induced mtROS generation, and rotenone induction with PLCS hydrogel co-treatment (final concentration of rotenone: 1 μmol/L). After treatment, the culture medium was discarded, and the cells were washed three times with pre-cooled PBS. MitoSOX Red probe was then added for incubation, followed by PBS washing to remove unbound probe, and Hoechst 33342 was used for nuclear staining. Fluorescence images were captured using a Nikon inverted fluorescence microscope, and the MitoSOX Red fluorescence intensity was quantitatively analyzed using NIS-Elements AR software.

4.12. Cell activity and cytotoxicity detection

To assess the biocompatibility of the hydrogel, PC12 cells were seeded at 5 × 10⁴ cells/well into 24-well plates pre-coated with PLCS hydrogel and cultured in a 37 °C, 5 % CO₂ incubator. On days 1, 3, and 5 of culture, Calcein/PI working solution (Beyotime, China) was added, and the cells were incubated at 37 °C in the dark for 30 min. Fluorescent images were then captured using fluorescence microscope (Nikon, Japan) and analyzed with NIS-Elements AR software.

The biocompatibility of the hydrogel was further evaluated using a CCK-8 assay. PC12 cell suspensions were seeded into 96-well plates pre-coated with PLCS hydrogel, ensuring uniform coverage. The culture plates were incubated overnight at 37 °C with 5 % CO₂ to allow for cell adhesion. At predetermined time points (1, 3, and 5 days), 10 μL of CCK-8 reagent (Beijing, Solarbio Co., Ltd.) was added to each well and incubated for 4 h. The absorbance was then measured at 450 nm using a microplate reader.

To evaluate the effect of the PLCS hydrogel on cell migration, cells were seeded into six-well plates pre-coated with PLCS hydrogels and cultured to approximately 90 % confluence. A straight linear scratch was created across the cell monolayer using a sterile 200 μL pipette tip. The culture medium was removed, and the cells were gently washed with PBS (2–3 times) to remove debris. Low-serum medium (1 % FBS) or treatment medium was then added. Images were captured at 0, 12, and 24 h at the same marked positions. The wound area was quantified using ImageJ software, and the wound closure (%) was calculated as follows: Wound closure (%) = (A₀h – A_t)/A₀h × 100 was quantified using ImageJ. Each group had triplicates.

To assess the neuroprotective effects of the hydrogel, PC12 cells (5 × 10⁴ cells/well) were seeded into 24-well plates pre-coated with PLC or PLCS hydrogel, and incubated overnight (37 °C, 5 % CO₂). The next day, cells were randomized into four groups: control (normal culture), H₂O₂ injury (200 μM), H₂O₂ + PLC, and H₂O₂ + PLCS, followed by 24-h culture. After washing three times with pre-cooled PBS (pH 7.4), cells were incubated with Calcein/PI working solution (37 °C, dark, 30 min). Fluorescent images were captured using a fluorescence microscope (Nikon, Japan) and analyzed with NIS-Elements AR software. All of the above experiments were repeated three times.

4.13. Mitochondrial targeting assay of PLCS hydrogel

PC12 cells were seeded into confocal dish at 5 × 10⁴ cells/dish and cultured overnight in a 37 °C, 5 % CO₂ incubator. Pre-prepared PLC/SS31-FITC hydrogel was then added, and the cells were randomly divided into two groups: (1) PLCS-FITC hydrogel group; (2) H₂O₂ (200 μM) + PLCS-FITC hydrogel group. Cultivation was terminated at preset time points (30 min, 2 h, 4 h). After termination, mitochondria, endoplasmic reticulum, and lysosomes were stained, followed by nuclear staining with Hoechst 33342 (Beyotime, China) for 5 min. Fluorescence images were acquired using a confocal laser scanning microscope (Nikon, Japan). All images were quantitatively analyzed using NIS-Elements AR software to ensure the accuracy and reproducibility of experimental data. The mitochondrial targeting ability of PLCS-FITC hydrogel in different environments was evaluated based on changes in fluorescence intensity.

4.14. Detection of SS31 in mitochondria

Homogenization buffer, wash buffer, and chromatographic-grade reagents are prepared in advance, SS31 standards and the isotopically labeled internal standard are readied, the high-speed refrigerated centrifuge is pre-cooled, and the LC–MS system is tuned. Fresh tissue/cell samples are rinsed with ice-cold PBS, mixed with homogenization buffer and homogenized on ice, then centrifuged at 4 °C, 1000–1500 g for 10 min to remove nuclei and large debris. The supernatant is further centrifuged at 4 °C, 10,000–12 000 g for 15 min to collect the mitochondrial pellet, which is resuspended in wash buffer and centrifuged again under the same conditions to obtain purified mitochondria. Methanol containing the internal standard is added to the pellet to lyse mitochondria and extract SS31; after solid-phase extraction, the eluate is dried under nitrogen, reconstituted in an appropriate solvent, and filtered through a 0.22 μm membrane. A series of SS31 calibration standards containing a fixed amount of internal standard is prepared and analyzed on a C18 column with gradient elution in ESI positive-ion MRM mode. Injections are performed in the order of blank, standards, and samples, and mitochondrial SS31 levels are calculated from the calibration curve with correction for dilution factors and sample amount.

4.15. Detection of mitochondrial morphology

Mitochondrial morphological changes were observed using a high-resolution confocal microscope. The specific procedures were as follows: PC12 cells were seeded at 5 × 10⁴ cells/well into 24-well plates pre-coated with PLC or PLCS hydrogel and cultured overnight in a 37 °C, 5 % CO₂ incubator. Subsequently, the cells were randomly divided into four groups: (1) Control; (2) H₂O₂ (200 μM); (3) H₂O₂ + PLC hydrogel; and (4) H₂O₂ + PLCS hydrogel, with each group cultured for another 24 h. After cultivation, the cells were washed three times with pre-cooled PBS (pH 7.4), then incubated with MitoTracker Deep Red (Beyotime, China) working solution at 37 °C in the dark for 30 min, followed by nuclear staining with Hoechst 3224. Finally, mitochondrial morphological images were observed and acquired using a confocal microscope (Nikon, Japan). All experiments were repeated three times to ensure result reliability.

To further observe mitochondrial ultrastructural changes, TEM was used to examine mitochondrial morphology. The specific procedures were as follows: After cell culture under the aforementioned grouping and treatment conditions, cells were collected and fixed in 2.5 % glutaraldehyde at 4 °C for 2 h. Subsequently, cell samples were fixed in 1 % osmium tetroxide (OsO₄) for 2 h, followed by three washes with 0.075 M PBS. Next, dehydration was performed using gradient ethanol (50 %, 70 %, 90 %, 100 %) and acetone, prior to embedding in epoxy resin. Semi-thin sections (1 μm) were stained with hematoxylin and observed under an optical microscope. Based on these observations, 10–15 semi-thin sections were selected for ultra-thin sectioning (70 nm). After double staining with uranyl acetate and lead citrate, the ultra-thin sections were observed and imaged for mitochondrial ultrastructure using TEM (FEI Tecnai F20, USA).

4.16. Mitochondrial membrane potential detection

Mitochondrial membrane potential was measured using a JC-1 fluorescent probe kit. PC12 cells were seeded at 5 × 10⁴ cells/well into 24-well plates pre-coated with PLC or PLCS hydrogel and incubated overnight in a 37 °C, 5 % CO₂ incubator. The next day, cells were randomly divided into four groups: (1) Control; (2) H₂O₂ (200 μM); (3) H₂O₂ + PLC hydrogel; and (4) H₂O₂ + PLCS hydrogel, with further cultivation for 24 h. After cultivation, cells were washed three times with pre-cooled PBS (pH 7.4), then incubated with JC-1 working solution at 37 °C in the dark for 30 min. Finally, fluorescent images were acquired using a fluorescence microscope (Nikon, Japan), and image analysis was performed using NIS-Elements AR software (Version 5.11, Nikon).

4.17. ATP detection

The culture medium was aspirated, and lysis buffer was added at a ratio of 200 μL per well of a 6-well plate. The solution was pipetted up and down repeatedly or shake the culture plate to ensure the lysis buffer fully contacted and lysed the cells. After lysis, the mixture was centrifuged at 12,000g for 5 min at 4 °C, and the supernatant was collected for subsequent assays. 100 μL of ATP detection working solution was added to each detection well or tube and left at room temperature for 3–5 min to consume background ATP and reduce noise. To save time, 100 μL of ATP detection working solution could be added to 10–20 detection wells or tubes simultaneously. 20 μL of sample or standard was added to the detection well or tube, mixed rapidly with a micropipette, and after at least 2 s, the relative light unit (RLU) value was measured using a multifunctional microplate reader (Multiskan FC, Thermo).

4.18. Extraction of NSCs

The isolation and culture of NSCs were performed as follows: newborn rats were sacrificed by cervical dislocation, their body surfaces were disinfected with 75 % alcohol, and spinal cord tissues were dissected under aseptic conditions. The spinal cord tissues were minced into ∼1 mm³ pieces using ophthalmic scissors, transferred to a centrifuge tube containing pre-cooled medium, and repeatedly pipetted with a narrow-bore pipette until the suspension became turbid. The tissue fragments were then mixed with 1.0 mL of 0.25 % trypsin (Servicebio, China), digested in a 37 °C water bath for 15 min, and centrifuged at 1000 rpm for 5 min. After discarding the supernatant, the cell pellet was resuspended in fresh NSC-specific medium (DMEM/F12 (Gibco, USA) supplemented with B27(Gibco, USA), 20 ng/mL EGF (Servicebio, China), and 20 ng/mL bFGF (Aladdin, China)) and transferred to an uncoated culture flask for suspension culture in a 37 °C, 5 % CO₂ incubator (Thermo, USA). After 7 days of primary culture, neurospheres with a diameter of ∼150 μm were observed. Following 2 passages, the NSCs were used for subsequent experiments.

4.19. Differentiation of NSCs in vitro

The specific procedures were as follows: NSCs in neurosphere form were seeded into 24-well plates pre-coated with PLC or PLCS hydrogel and cultured overnight in a 37 °C, 5 % CO₂ incubator. Subsequently, the cells were randomly divided into four groups: (1) Control; (2) H₂O₂ (200 μM); (3) H₂O₂ + PLC hydrogel; and (4) H₂O₂ + PLCS hydrogel. All groups were cultured in differentiated medium for 5 days. After cultivation, the cells were washed three times with pre-cooled PBS (pH 7.4) for 5 min each, then fixed with 4 % paraformaldehyde at room temperature for 15 min. Following fixation, the cells were washed three times with PBS and permeabilized with 0.3 % Triton X-100 at room temperature for 15 min. Next, non-specific binding was blocked with 5 % bovine serum albumin (BSA) (Biofroxx, Germany) solution at room temperature for 1 h. Primary antibodies (Anti-Nestin 1:500; Anti-ChAT 1:200; Anti-Tuj1 1:500) were then added and incubated overnight at 4 °C. The next day, after three PBS washes, fluorescently labeled secondary antibodies (Alexa Fluor 488 or 594, 1:500; Abcam) were added and incubated in the dark at room temperature for 1 h. Unbound secondary antibodies were washed away, and cell nuclei were stained with DAPI for 5 min. Finally, the cells were washed three times with PBS, mounted with anti-fluorescence quenching medium, and fluorescence images were acquired using a Nikon inverted fluorescence microscope (Nikon, Japan). All images were quantitatively analyzed using NIS-Elements AR software (Version 5.11, Nikon) to ensure data accuracy and reproducibility. All experiments were repeated three times. To verify whether the effects of PLCS could be reversed, rotenone (Rote), a specific mitochondrial ROS inducer, was used instead of hydrogen peroxide, with the remaining steps unchanged. Additionally, to assess the importance of the PI3K-Akt pathway, inhibition experiments were conducted using a specific inhibitor (LY294002, LY) of this pathway, with the other steps remaining the same.

4.20. Animals and surgical operations

Female Sprague Dawley rats (6–8 weeks old, weighing 220–250 g) were used. All animal experiments were approved (Approval No. SYSU-IACUC-2025-001377) by the Institutional Animal Care and Use Committee (IACUC), Sun Yat-Sen University and strictly conducted in accordance with the Guidelines for Laboratory Animal Welfare and Ethics and relevant laws and regulations. All animals were randomly assigned to experimental groups using a simple randomization method generated by an independent researcher who was not involved in data collection or analysis. The experimental procedures, including surgery and transplantation, were as follows: First, rats were anesthetized with isoflurane (inhalation, 1.5 %–2 % concentration) to maintain an appropriate anesthetic depth. After anesthesia, the hair on the rat's back at the T9–T10 level was shaved, and the surgical site was disinfected three times with 0.5 % povidone-iodine solution to reduce infection risk. Under aseptic conditions, a 1.5 cm skin incision was made at the T9–T10 vertebral level. The lateral spinal muscles were bluntly dissected, and the spinous processes and laminae were removed with rongeurs to fully expose the spinal cord. Subsequently, a 2 mm segment of spinal cord tissue at the T10 level was completely resected using micro-scissors to establish a complete spinal cord transection model, ensuring a clear injury site with no residual tissue. PLC or PLCS hydrogel was then slowly injected into the injury cavity using a syringe to ensure uniform distribution and complete filling of the injured area. The experiment was divided into four groups: (1) Sham (only spinal cord exposure without injury); (2) Injury; (3) PLC hydrogel; (4) PLCS hydrogel. After surgery, the surgical field was irrigated with normal saline, disinfected, and muscles and skin were sutured layer by layer with absorbable sutures. Penicillin (50000unit kg⁻¹ day⁻¹, Jusheng, China) was administered intramuscularly for 7 consecutive days to prevent infection. The bladder was manually expressed twice daily (morning and evening) until the rats regained autonomous micturition. Meanwhile, the diet, activity, and wound healing of rats were closely monitored, with nutritional support provided as needed.

4.21. RNA sequencing and analysis

The specific procedures for RNA-seq are as follows: First, RNA extraction and quality assessment are performed. Samples are rapidly collected and stored in RNA protectants or liquid nitrogen. Total RNA is extracted using TRIzol reagent or commercial kits. Concentration and purity are determined using a NanoDrop (A260/A280 > 1.8, A260/A230 > 2.0), and RNA integrity is evaluated via an Agilent Bioanalyzer (RIN ≥7.0). Next, library construction is conducted: For eukaryotes, mRNA is enriched using poly(A) magnetic beads; for prokaryotes, rRNA is removed. mRNA is fragmented into 200–500 bp segments, followed by reverse transcription to synthesize cDNA, end repair, adapter ligation, and PCR amplification/purification to generate sequencing libraries. For high-throughput sequencing, platforms such as Illumina NovaSeq are used. Libraries are diluted, loaded, and sequenced, with base signals read via sequencing-by-synthesis technology. Finally, data analysis involves processing raw data with tools like Trimmomatic, aligning sequences to reference genomes using Hisat2, quantifying gene expression with Salmon, analyzing differentially expressed genes via DESeq2, and performing GO and KEGG enrichment analyses. Throughout the process, attention must be paid to biological replicates, batch effects, RNA integrity, and contamination control. Extended applications include differential splicing analysis, fusion gene detection, and non-coding RNA analysis.

4.22. The detection of ACh, acetyl-CoA, choline, and PDH

Following anesthesia and immobilization, rats were subjected to transcardial perfusion with pre-cooled PBS to remove residual blood. The target spinal cord segments were then dissected and isolated, rinsed with PBS, blotted dry with filter paper, and weighed. Pre-chilled extraction reagent containing protease inhibitors was added at a proportional volume, and the tissue was homogenized on ice. After centrifugation at 11,000×g for 10 min at 4 °C, the supernatant was collected and stored on ice or frozen. Throughout the process, cold conditions were maintained to preserve sample integrity. Subsequently, the ACh, acetyl-CoA, Choline, and PDH were measured using the corresponding commercial kits following the manufacturers' instructions.

4.23. Behavioral assessment

Hindlimb motor function recovery in rats was evaluated using the BBB scale and a footprint analysis system. The specific methods were as follows: The BBB scale (0–21 points) was used to quantitatively assess hindlimb motor function, with scores ranging from 0 (complete paralysis) to 21 (normal motor function). BBB scoring was performed weekly at fixed time points, with each observation lasting 10 min. The BBB scoring was independently performed by two uniformly trained experimenters, with the single-blind method implemented throughout the process: prior to scoring, a third party assigned anonymous numbers to animals in each group to mask group information; the experimenters observed the behavior of each rat without knowing the group assignments; if the score difference between the two experimenters exceeded 2 points, a third senior experimenter re-evaluated and determined the final score to avoid subjective bias. The footprint experiment further evaluated motor function recovery by analyzing the morphology and distribution of footprints. The experimental setup included a 100 cm-long, 7 cm-wide, and 6 cm-high cardboard channel, with white paper of the same size placed at the bottom and transparent plates on both sides to prevent rats from deviating. Before the experiment, the soles of the rats' hindlimbs were evenly coated with black ink, and the rats were then placed at the channel entrance to walk through at a constant speed. The number and morphology of footprints left on the white paper were recorded. Footprint data were analyzed using ImageJ software, with stride length and stride width measured to quantitatively assess motor function recovery.

4.24. Electrophysiological detection

At 8 weeks post-surgery, MEPs were measured to evaluate neural circuit reconstruction. The specific procedures were as follows: Rats were anesthetized via inhaled isoflurane (1.5 %–2 %), with anesthetic depth monitored by the disappearance of the toe-pinch reflex. After anesthesia, rats were immobilized on an experimental board using a custom cloth bag, and room temperature was maintained at 25–28 °C to stabilize body temperature.

For MEP detection, stimulation electrodes consisted of a positive and a negative electrode: The positive electrode (2-mm-diameter spherical electrode) was precisely placed on the skull surface over the midline of the motor cortex, at coordinates 2.5 mm posterior to bregma and 2 mm left/right of the midline; the negative electrode (4-mm-diameter disk electrode) was positioned on the hard palate skull surface. Recording electrodes were inserted 1.5 mm deep into the gastrocnemius muscles of both hindlimbs, with the reference electrode 2 cm distal to the recording electrode and the ground electrode between the stimulation and recording electrodes.

Monophasic square-wave stimulation was delivered to the motor cortex with the following parameters: intensity 5 mA, duration 0.2 ms, frequency 1 Hz, band-pass filter 2 Hz to 10 kHz, and amplifier sensitivity 0.1 mV/Div. Signals from the bilateral gastrocnemius muscles were recorded, with key indices including latency and peak-to-peak amplitude (from the negative peak to the adjacent positive peak). All data were collected and quantitatively analyzed using the BL-420 Data Acquisition Analysis System (TECHMAN SOFT, China). All the experiments were conducted five times.

4.25. Heat pain assessment

At 8 weeks after SCI, the sensitivity of rat hindlimbs to photothermal stimulation was assessed using a thermal pain detector (PL-200, Tekman Software, China) under quiet conditions with animals allowed to move freely. Rats were placed in a dark observation chamber for 30 min to acclimate. A focused heat source was directed precisely onto the plantar surface of the hindpaws. Stimulation was initiated at 50 % light and thermal intensity, with the start time automatically recorded. When the rat lifted its paw from the glass plate, the device detected the movement, recorded the paw withdrawal latency, and immediately terminated the stimulus.

4.26. Quantitative reverse transcription polymerase chain reaction (qRT-PCR) analysis

The expression of pro-inflammatory cytokines (IL-1β, TNF-α, IL-6) and anti-inflammatory cytokines (IL-10, Arg1) in RAW264.7 cells was quantified by qRT-PCR. Total RNA was extracted using TRIzol reagent (Invitrogen, USA) and reverse-transcribed into complementary DNA (cDNA) with a PrimeScript RT reagent kit (Takara, Japan). qPCR was performed in a 10 μL reaction system containing SYBR Green Master Mix (Takara, Japan), specific primers, and cDNA template. The thermal cycling conditions were as follows: initial denaturation at 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s, 60 °C for 25 s, and 72 °C for 30 s. The relative gene expression levels were normalized to the housekeeping gene B2m and calculated using the 2^−ΔΔCt method. Results are presented as fold changes relative to the control group. Primer sequences are listed in Supplementary Table 1.

4.27. Western blot analysis

Proteins were extracted using RIPA lysis buffer supplemented with protease and phosphatase inhibitors. Protein concentrations were determined using a BCA Protein Assay Kit. Equal amounts of protein were separated by SDS-PAGE and transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore, BUSA). Membranes were blocked with 5 % non-fat milk in TBST (Tris-buffered saline with Tween-20) for 90 min at room temperature. After blocking, membranes were incubated with primary antibodies overnight at 4 °C. The following day, membranes were washed five times with TBST and then incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit or anti-mouse IgG secondary antibodies (Proteintech, 1:10,000) for 1 h at room temperature. Immunoreactive bands were visualized using an ECL chemiluminescent substrate (Thermo Scientific, USA) and detected with using the Tanon system (TANON-5200CE). For sequential probing of multiple proteins, membranes were stripped using a primary antibody-specific stripping buffer, followed by re-blocking and reprobing with new primary and secondary antibodies. Densitometric analysis was performed using ImageJ software. The primary antibodies include Anti-Nestin (Abcam,1:1000) Anti-Tuj1(ABclonal,1:1000), Anti-TOMM20 (Proteintech, 1:2000), Anti-C-Casp3 (CST, 1:1000), Anti-GAPDH (Poteintech, 1:10,000), Anti-NF200 (Abcam, 1:1000), Anti-MBP (Abcam, 1:1000), Anti-NeuN (Abcam, 1:1000), Anti-PI3K (CST, 1:1000), Anti-pPI3K (Proteintech 1:1000), Anti-AKT (Proteintech, 1:1000), Anti-pAKT (Proteintech, 1:1000), and Anti-ChAT (WANLEIBIO, 1:1000).

4.28. HE and immunofluorescence staining of the spinal cord tissue

HE staining and immunofluorescence were used to evaluate spinal cord pathological structures and nerve regeneration. The specific procedures were as follows: Rats were anesthetized, and the heart was exposed with surgical scissors. PBS and 4 % paraformaldehyde (PFA) (Biosharp, China) were sequentially perfused through the left ventricle until complete fixation. Spinal cord tissues were then harvested and dehydrated in gradient sucrose solutions (10 %, 20 %, 30 %) at 4 °C. After dehydration, tissues were embedded in OCT compound (Sakura Finetek, USA) and longitudinally sectioned into 12-μm-thick slices using a cryostat.

For HE staining, the specific steps were as follows: After sectioning, slices were air-dried at room temperature for 30 min, then rehydrated through a graded ethanol series (100 %, 95 %, 70 %, 50 %) for 3 min each, followed by a 5-min rinse in distilled water. The slices were stained with hematoxylin solution for 1 min, then differentiated in 1 % hydrochloric acid-ethanol for 30 s to remove excess stain. After rinsing with tap water for 10 min to blue the nuclei, eosin solution was applied for 2 min for cytoplasmic staining. The slices were then dehydrated through the reverse ethanol series (50 %, 70 %, 95 %, 100 %) for 3 min each, cleared with xylene for 5 min, and mounted with neutral balsam. Stained sections were observed under a light microscope (Nikon, Japan) to evaluate cavity formation, inflammatory cell infiltration, and morphological changes at the lesion site.

For immunofluorescence, primary antibodies Nestin (Abcam,1:500), Tuj1(ABclonal, 1:8000), GFAP (Abcam, 1:500), NF200 (Abcam, 1:200), MBP (Abcam, 1:500), Syn1(Abcam, 1:500), NeuN (Abcam,1:800), IL-1β (affinity, 1:500), Arg1 (affinity, 1:500), and ChAT (WANLEIBIO, 1:200) were used to detect nerve regeneration. The specific steps were: Slices were warmed to room temperature for 30 min, washed three times with PBS, and permeabilized with 0.3 % Triton X-100 at room temperature for 15 min. Next, slices were blocked with 5 % BSA for 1 h at room temperature to reduce non-specific binding. Primary antibodies were then added and incubated overnight at 4 °C. The following day, after three 5-min PBS washes, fluorescently labeled secondary antibodies (Alexa Fluor 488, 594, or 647, 1:500; Abcam) were added and incubated in the dark at room temperature for 1 h. Unbound secondary antibodies were washed away, and slices were mounted with anti-fluorescence quenching medium containing DAPI (Servicebio, China). Fluorescence images were acquired using a panoramic digital scanner (Nikon, Japan), and all images were quantitatively analyzed with NIS-Elements AR software (Version 5.11, Nikon) to ensure data accuracy and reproducibility.

4.29. Blood biochemical analyses

Before blood collection, rats were anesthetized with isoflurane to minimize stress and ensure procedural safety. Blood samples were collected via orbital sinus puncture following standardized procedures, in strict accordance with ethical guidelines and operational standards for animal experiments, and were subsequently analyzed. Each experimental group consisted of five biologically independent rats with consistent genetic backgrounds and uniform housing conditions.

4.30. Statistical analysis

All experimental results in this study are expressed as mean ± standard deviation (Mean ± SD). Statistical analyses were performed using GraphPad Prism software (Version 10.1.2, GraphPad Software, USA). Statistical significance was determined by one-way ANOVA followed by the Tukey post hoc test. The significance levels were set as ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001, indicating statistically significant differences.

CRediT authorship contribution statement

Yiqian Luo: Writing – original draft, Supervision, Methodology, Conceptualization. Pan Jiang: Writing – original draft, Supervision, Investigation. Daoqiang Huang: Writing – original draft, Data curation. Hong Li: Writing – original draft. Jiale He: Data curation. Ruoqi Shen: Supervision. Yunheng Jiang: Writing – review & editing. Limin Rong: Writing – review & editing, Funding acquisition. Bin Liu: Writing – review & editing, Project administration, Funding acquisition, Conceptualization.

Ethics approval and consent to participate

All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC), Sun Yat-Sen University (Approval No. SYSU-IACUC-2025-001377) and strictly conducted in accordance with the Guidelines for Laboratory Animal Welfare and Ethics and relevant laws and regulations.

Declaration of competing interest

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

Appendix A. Supplementary data

The following is the Supplementary data to this article: