Electrical gastrodin-polyurethane spiral conduits with micro/nano-structure for accelerating peripheral nerve regeneration

Abstract

Peripheral nerve injuries tend to cause the proximal nerve unable to contact the corresponding target organ, resulting in sensory and motor dysfunction. The simple filling materials within nerve conduits are often inadequate for axonal ingrowth and directional regeneration. In this study, to enhance the guidance effect and achieve physiologically adaptive function, a 3D nanofibrous polyurethane (PU) scaffold with oriented microchannels was engineered using electrospinning and manual curling techniques. The electrospun fibrous membranes can be manually curled up into tubular structures with spiral and longitudinal multi-channels. The immunoregulatory and conductive properties were developed by being grafted gastrodin and aniline trimer (AT, 2.6 % and 5 %). Gastrodin stimulated proliferation of neural cells and expression of neuroblast-related genes. Electroactive AT produced an electrical signal in combination with electrical stimulation (ES) to accelerate the elongation and growth of Schwann cells (SCs) and neurite outgrowth of PC12 cells. The in vivo experiments revealed that the releasing gastrodin and electrical signals created prohealing microenvironment for alleviating inflammation and promoting vascularization. Of note, the topological structure provided well-organized internal support for the cells to spread, as well as the migration of SCs and the directional elongation of regenerating axons. The adaptive electroactivity of gastrodin-PU-AT5 % further ensured nerve signal transmission, ultimately promoted remyelination through upregulation of Rap1 and mTOR signaling pathways; thereby enhancing functional and structural regeneration. This scaffold design strategy will push forward the application of nerve conduits in long-distance peripheral nerve injury.

Graphical abstract

Highlights

• •A spiral conduit is designed to enhance directional guidance.
• •Nano-fibers and micro-channels provide well-organized internal support.
• •Released gastrodin alleviates inflammation and promotes vascularization.
• •The electroactivity activates Rap1 and mTOR signal transmission.
• •The conduit accelerates functional and structural regeneration of nerve.

1. Introduction

The dispersion and mismatch of regenerating axons are the key constraints for the effective regeneration of nerves in peripheral nerve injury (PNI) [1]. The incidence of PNI caused by trauma, compression, ischemia, or iatrogenic factors is becoming increasingly common in modern society [2]. PNI often leads to dysfunctions of sensory and motor nerves [3]. Nerve reconstruction is difficult to achieve autonomously when the PNI is longer than 5 cm (long-distance nerve defect) [4]. Autologous nerve transplantation, which is regarded as the “golden standard” for treating large gap PNI in the clinic [5], is still restricted by second surgery, potential or diameter mismatch between injured nerve and donor nerve, as well as neuroma formation and infection [6,7]. Nerve guidance conduits (NGCs) are the alternative and innovative therapeutic strategy to nerve autografts, owing to their ability to guide nerve growth along the conduit axis [8]. The prevailing NGCs predominantly adopt single-channel architectures, failing to mimic intraluminal microenvironment to guide Schwann cells (SCs) migration, thus resulting in inappropriately targeted re-innervation [9] (see Scheme 1).

Scheme 1.

Schematic illustration of an electrical spiral conduit with oriented multi-channels was designed using electrospinning and manual curling techniques for peripheral nerve regeneration

Peripheral nerves have a unique multi-fascicle structure, and the axons in each fascicle are aligned in orientation [10]. In pursuit of intraluminal ultrastructure, topographical cues, including highly oriented structure, micro/nano arrays, and multi-channels, have gained much attention in developing NGCs [11]. The conventional hollow conduits could transport water and nutrients over long distances, but lacking linear guidance effect [12]. This disordered structure resulted in misalignment of regenerated nerve positioning. To enhance directional axon extension, the multi-channel structure with a similar bundle structure of nerves emerged as a candidate [13,14]. Particularly, micro/nano-channels with rough surfaces actively promote protein absorption and stimulate cellular responses, further regulate the directional growth of SCs and preserve their physiological functions well [15]. Maximum spatial utilization of the multi-channel ensures effective bridging between the injured nerves with the NGCs. Therefore, it is of great significance to integrate rich topographical cues to NGCs to guide peripheral nerve regeneration. Electrospinning nanoscale substrates show greatly improved efficiency in contact surface area, topographical cues, permeability, degradability, and sufficient mechanical strength to maximize nerve growth, which is prevalent in the field of peripheral nerve [16,17]. Such conduits are effective for repairing short-gap nerve defects [18]. For long-segment nerve defects, however, there is an urgent need for functional regulation to accelerate axonal maturation and long-term nerve regeneration.

Bioelectricity, a key factor in preserving the internal electrophysiological microenvironment, plays an important role in regulating cell activity and tissue function in electrically active tissues [19]. In the process of nerve development, nerve tissue completes the transmission of bioelectrical signals through the conversion of action potential and resting potential, then regulates various life activities [20]. PNI causes abnormal local bioelectric signal and delays the healing of nerve tissue [21]. To overcome the above-mentioned issues, NGCs with the conductive property are crucial to enable appropriate transmission of bioelectric signal [22]. The conductive matrix can alter the local electrostatic charge of the conduits and promote more serum proteins to be adsorbed onto the conduits, accelerating cell adhesion, proliferation, and migration [23]. Additionally, the electroactivity within the conduits also serves as an electrostatic binding site, combining with the negatively charged cell membrane, bringing the cells closer to the conduits surface, and establishing stronger attachment sites [24]. This helps restore the conduction of action potentials in damaged tissues, promoting cell differentiation and tissue growth. Aniline oligomers and oligoaniline-based conductive polymers have been widely recognized due to their excellent electrical conductivity, desirable degradability [25,26], and modifiability, which can be cleared and excreted through the kidneys [27]. As a type of aniline oligomer, aniline trimer (AT) exhibits remarkable electrical activity to support the repair of bone [28], nerve [29], and cardiac tissue [30]. In this way, AT enhanced migration, neurotrophic factor levels, and myelination of SCs, accelerating nerve healing and functional recovery.

When the nerve graft is implanted at the site of a nerve defect, it will inevitably lead to different degrees of inflammatory response in the tissue. Excessive inflammation ultimately affects nerve regeneration and repair [31]. Gastrodin, the main bioactive components of gastrodia elata Blume, has been reported as a potent regulator of inflammation by normalizing tumor necrosis factor-α (TNF-α), interleukin 1β (IL-1β), inducible nitric oxide synthase (iNOS), and nuclear factor-kappa B (NF-κB), etc. [32]. Gastrodin can exert neuroprotective effects via improving brain energy metabolism disorders in rats, and reducing mitochondrial dysfunction induced by H₂O₂ oxidative damage to cells [33]. Recent studies have revealed that gastrodin can accelerate angiogenesis [34,35], which may ameliorate the hypoxia in the center of the implant, especially in long-distance repair. Previous studies have shown that gastrodin functionalized polyurethane (PU) conduit could effectively regulate M2 macrophage polarization, growth of SCs, and vascularized nerve regeneration, although its repair efficiency is still not as good as autograft [36]. To achieve accurate regeneration and functional recovery of injured nerves, it is very necessary to further develop functional artificial nerve grafts presenting multiple cues that synergistically direct axonal advance.

Based on the above mentioned, a spiral tubular fibrous structure combined with electroactive stimulation and anti-inflammation is expected to greatly improve the regeneration of nerve grafts for long-distance PNI. To achieve this, we constructed biomimetic membranes by electrospinning and then curled up into spiral structure to simulate the topography of microchannels. Such a design provides oriented structure and porosity, creates adequate spatial support for axonal growth and ensures bridging between the injured nerves with grafts. We first used PU as the main body to prepare copolymers with excellent strength. Further incorporation of AT and gastrodin simultaneously enhance the viability, migration, and signal transduction of nerve cells in vitro, as well as their differentiation. To demonstrate the potential of our electroactive spiral conduit in the field of repairing large-distance nerve defects, we created a 10-mm defect in the rat sciatic nerve and comprehensively analyzed their effects on nerve regeneration. Efficient bridging along the channels would open up a route for supporting the migration and spatial organization of nerve cells, representing a promising strategy for the improvement of comprehensive healing of the nerve defect.

2. Materials and methods

2.1. Materials

Isophorone diisocyanate (IPDI) was purchased from Aladdin Co. Ltd. (China). Poly (ε-caprolactone)-diol (PCL-diol, Mn = 2000) was purchased from Acros. Tin (Ⅱ) 2-ethylhexanoate (Sn(Oct)₂, 92.5–100.0 %) was purchased from Sigma-Aldrich (Germany). Lysine ethyl ester dihydrochloride (Lys·OEt-2HCl), p-phenylenediamine, aniline, aqueous ammonia (NH₃∙H₂O), polyethylene oxide (PEO), 2,2,2-trifluoroethanol (TFEA, purity ≥99.5 %) were obtained from Macklin (China). Deuterated dimethyl sulfoxide (DMSO-d₆) was purchased from Energy Chemical (China). Ammonium persulfate was obtained from Sangon Biotech (China). Hydrochloric acid (HCl) was purchased from Chongqing Chuandong Chemical Group (China). Calcium chloride dihydrate and lipase were obtained from Shanghai Yuanye Bio-Technology Co., Ltd. (China). Dimethyl sulfoxide (DMSO) and ethanol were obtained from Chengdu Kelong Chemical Co., Ltd. (China). Gastrodin (with a purity exceeding 99.0 %) was obtained from Kunming Pharmaceutical Co. Ltd. (China). Phosphoric acid aqueous solution was purchased from Tianjin Fengchuan Chemical Reagents Technology Co., Ltd. (China), and acetonitrile was obtained from Thermo Fisher Scientific (USA).

2.2. Synthesis of AT

AT was prepared according to the “one-step method” with minor modification [37]. In brief, 10.00 g p-phenylenediamine and 5.80 g aniline were dissolved in a three-neck flask containing a mixed solution of 10 mL of 1 M HCl and 90 mL ethanol, with chilled in a salty ice bath. After the temperature was under −5 °C, 12.25 g ammonium persulfate in 40 mL of 1 M HCl was slowly added into the aforementioned solution to react for 6 h. The resulting solution was vacuum filtered and washed with 1 M HCl, followed by deionized water. The product was then treated with 1 M ammonium hydroxide for 12 h, and the solution was again vacuum filtered and washed with deionized water until the pH of the filtrate was neutral. Finally, the dark blue solid was obtained and dried by lyophilizer (TFD-1A-95, Boyikang (Beijing) Instrument Co., Ltd., China).

2.3. Synthesis of gastrodin-PU-AT copolymers

The prepolymers were prepared according to the literature [27]. In brief, dehydrated PCL2000 and IPDI (0.02 % Sn(Oct)₂) were mixed in anhydrous DMSO in a 250 mL three-necked flask under nitrogen at 70 °C and stirred for 6 h. Lys·OEt-2HCl dispersed in DMSO solution was then added as a chain extender and reacted for another 4 h. Subsequently, AT dissolved in anhydrous DMSO was added dropfully and stirred for 1.5 h. Gastrodin dissolved in DMSO were added finally and continued to react for 1.5 h. After this period, copolymers were precipitated in anhydrous ethanol and washed with deionized water prior to vacuum-dried for 3 days. The samples were named as PU, gastrodin-PU, PU-AT2.6 %, PU-AT5 %, gastrodin-PU-AT2.6 %, and gastrodin-PU-AT5 % (Table S1).

2.4. Fabrication of membranes and spiral conduits

The copolymers containing 5 wt% PEO were dissolved in TFEA at a concentration of 12 w/v% to form a spinning solution and stirred for 12 h. After standing for 4 h, the solution was fed into a 10 mL plastic syringe with a 21-gauge needle and used for electrospinning by electrospinning machine (Beijing Yongkang Leye Technology Development Co., Ltd., China). The flow rate of the solution was 1 mL/h. A voltage of 12 kV was applied between the needle tip and the collector (rotating speed at 140 rpm), with a collection distance of 15 cm. Then, the fibrous membranes were air dried at room temperature to eliminate residual solvent adequately. The membranes with a thickness of 300 μm (electrospinning for 36 h) were cut into discs (Ф 14 mm × 300 μm) for further cell experiments. The membranes with a thickness of 150 μm (electrospinning for 18 h) were cut into rectangles (20 mm-long, 12 mm-wide) and manually curled into spiral conduits with micro/nano structures for animal experiments.

2.5. Characterization

2.5.1. Chemical studies

Fourier transform infrared spectroscopy (FTIR) of AT and copolymers were observed by FTIR spectrometer (SP3, PerkinElmer, USA) at the range of 4000–600 cm⁻¹.¹H nuclear magnetic resonance (¹H NMR) spectra was obtained by using a Bruker AVANCE NEO 600 MHz spectrometer. DMSO-d₆ was used as the solvent and internal standard (δ 2.50 ppm).

2.5.2. Morphologies observation

The surface morphologies of electrospun membranes were observed by scanning electron microscopy (SEM, FEI Quanta-200, Switzerland) at an accelerating voltage of 20 kV. Prior to observation, all samples were sputter-coated with gold. The diameter of electrospun fibers were measured using Nano Measurer 1.2 software (National Institutes of Health, Bethesda, MD, USA). At least 50 nanofibers from each sample were analyzed in different SEM images.

2.5.3. Surface hydrophilicity evaluation

The contact angle of membranes was measured using a video contact angle instrument (JY-82 A, Chengde Dingsheng, China). Each drop of deionized water was dispensed at a volume of 3.0 μL, followed by the controlled descent of the titration system. The initial contact angle was recorded and five replicates were carried out for each group.

2.5.4. Electroactivity analysis

The ultraviolet–visible spectroscopy (UV–vis) spectra of the AT and copolymers in TFEA were recorded on an UV–vis spectrophotometer (PerkinElmer Lambda 35, USA). Cyclic voltammograms (CV) of copolymers were recorded on an electrochemical workstation (CHI 660 E, Shanghai Chenhua Electronic Technology Co., Ltd., China) with a scanning rate of 10 mV/s in 1 mol/L HCl solution. A graphite electrode was employed as the working electrode, while a platinum wire electrode served as the counter electrode, and an Ag/AgCl electrode was used as the reference electrode. The graphite electrode was immersed separately in the copolymer solutions for 3 h, and upon drying, a distinct film formation was observed for subsequent testing. The conductivity of the film was measured using four-probe technique, and the conductivity was recorded using an RTS-8 digital four-probe tester (Shenzhen Junda Instrument Co., Ltd., China).

2.5.5. Mechanical testing

The uniaxial tensile properties of the spiral conduits were determined by universal testing machine (Model 34TM10, Instron Company, USA). The conduits (Ф 3.0 mm × 40 mm) were placed in between two parallel splints with initial gauge length of 20 mm and then stretched at a constant strain rate of 5 mm/min until the conduit ruptured. Both the tensile strength and modulus were calculated from stress-strain curves by averaging at least eight specimens.

2.5.6. Degradability

The degradation of conduits (∼0.04 g) in vitro was performed in 1.5 mL of 0.05M NaOH solution at 37 °C, shaking at 30 rpm. The media was refreshed every week. Specimens were tested in quintuplicate.

Mass loss: At the determined time points (3 days, 1, 2, 4, 8, and 12 weeks), samples were taken out from solutions, gently rinsed with distilled water, and dried in an oven at 40 °C until constant weight. The percentage of weight loss was computed according to the following equation:

$$\text{Mass}\text{loss}(\%)=\frac{\left(W0-W1\right)}{\mathrm{W}0}\times 100\%$$

Where W₀ and W₁ represent the weights of conduits before and after degradation, respectively.

Gastrodin release: The gastrodin released from conduits was quantified by high-performance liquid chromatography (HPLC, Shimadzu, Kyoto, Japan). The collected degradation liquid was filtered through a 0.22 μm filter, separated on a C18 column (250 mm × 4.6 mm, 5 μm, USA) using a mobile phase composed of a 97:3 (v/v) mixture of 0.05 % phosphoric acid aqueous solution and acetonitrile. The sample injection volume was 10 μL and the flow rate was 0.8 mL/min. The detection of gastrodin was set at 220 nm, and a calibration curve of gastrodin was obtained under the same conditions.

2.6. Cytocompatibility of membranes in vitro

SCs and PC12 were obtained from the Bei Na Chuanglian Biotechnology (BNCC, Wuhan, China). They were cultured in Dulbecco's modified Eagle's medium (DMEM, Gibco, USA) containing 10 % fetal bovine serum (FBS, Gibco, USA) and 1 % penicillin-streptomycin (Gibco, USA) in a 5 % CO₂ incubator at 37 °C. The culture medium was refreshed every two days. The discs were sterilization with γ-ray irradiation with 25 kGy before incubated overnight in well plates using complete DMEM, and then seeded with cells.

2.6.1. SCs proliferation and viability assay

SCs at a density of 2 × 10⁴ cells/well were seeded onto discs in a 24-well plate. Proliferation was assessed using the CCK-8 assay (Dojindo Molecular Technologies, Japan). At intervals of 1 and 3 days, the culture medium was refreshed with 450 μL fresh DMEM containing 50 μL CCK-8 solution. Following incubated at 37 °C for 2 h, 100 μL supernatant was transferred to 96 well plates, and the absorbance at wavelengths of 450 nm was quantified using Multiskan Spectrum (SpectraMax 190, Molecular Devices Corporation, USA). The cell cytotoxicity was analyzed using a Live/dead Viability/Cytotoxicity Assay Kit (Thermo Fisher Scientific, USA). The discs with adherent SCs were rinsed with Phosphate Buffered Saline (PBS, Servicebio) three times and then stained with 400 μL of combination dye for 0.5 h, followed by observation under an inverted fluorescence microscope (IX73, Olympus, Japan). For EdU staining, SCs at a density of 3 × 10⁴ cells/well were seeded onto discs in a 12-well plate. After incubation for 3 days, cells were stained with BeyoClick™ EdU Cell Proliferation Kit with Alexa Fluor 555 (C0075S, Beyotime Biotechnology, China). Images of positively stained cells were captured by a light microscope (BX53, Olympus, Japan) and calculated by Image J software. Meanwhile, the gene expression of insulin receptor substrate 53 (IRSp53), glial cell-derived neurotrophic factor (GDNF), early growth response 2 (Egr2), and peripheral myelin protein 22 (PMP22) were analyzed by RT-qPCR (Table S2).

Migration: SCs (2 × 10⁴/well) were added to the upper compartment of transwell inserts with an 8 μm polycarbonate membrane pore size (Corning, USA), while the disc was positioned in the lower 24-well plate. After 3 days of incubation, the cells in the upper chamber were fixed with paraformaldehyde and stained with crystal violet (Solarbio, China). The non-invaded cells on the upside of the transwell were gently wiped with a cotton swab, and the migrated cells to the underside were observed under a light microscope.

Myelination: To examine the combined effects of electrospun membranes and electrical stimulation on SCs, discs seeded with SCs (3 × 10⁴ cells/well) in 12-well plate were subjected to electrical stimulation (ES) at 100 mV/cm and 20 Hz frequency, for 2 h per day over consecutive periods of 1 and 3 days, based on the parameter optimization shown in Fig. S11. Prior to staining with S100 for the cytoplasm and DAPI (Sigma, Germany) for the nuclei, cells were fixed with 4 % paraformaldehyde. Stained cells were observed under a light microscope.

2.6.2. Differentiation of PC12

PC12 cells at a density of 3 × 10⁴ were seeded onto discs in 12-well plates. The parameters of ES were 400 mV/cm and 20 Hz frequency, for 1 h per day over consecutive periods of 1 day and 3 days, with the stimulation parameters selected according to Fig. S12. At indicated time points, the cells were double-stained with NF-200 and βIII-Tubulin in the cytoplasm, and DAPI in the nucleus. The images were captured by a light microscope and the length of neural synapses were calculated using Image J software. For Western blotting (WB), discs seeded with PC12 cells (5 × 10⁵/well) in 6-well plates were incubated for 3 days to detect protein expression of growth associated protein 43 (GAP43) and cyclic adenosine monophosphate (cAMP) (Table S3).

2.7. Animal subjects

Forty-eight adult male Sprague Dawley (SD) rats (weighing 280–300 g) were used to investigate the repair effects of conduits on sciatic nerve defects. The animal experiments were approved by the Experimental Animal Ethics Committee of Kunming Medical University (Protocol No. 20221734). The rats were randomly divided into three groups (sixteen rats/group): autograft, gastrodin-PU-AT2.6 %, and gastrodin-PU-AT5 %. Before implantation, the rats were adaptively fed with a standard diet for one week, then fasted for 12 h without water deprivation. The conduits were sterilized with γ-ray irradiation with 25 kGy.

2.7.1. Surgical procedure

Rats were anesthetized with 1 % pentobarbital sodium (30 mg/kg body weight) via intraperitoneal injection. The right hind leg hair was shaven and the skin was disinfected. All surgical procedures were conducted under aseptic conditions using sterile tools. A 30 mm incision was made parallel to the femoral axis and the muscle tissue was exposed using a blunt dissection. Then the right sciatic nerve was exposed carefully and a 10 mm nerve segment was sharply transected and removed. For the autograft group, 10 mm nerve was cut off from the surgical side nerve, inverted and sutured with the nerve ends using 8–0 absorbable nylon sutures immediately. In the other two conduit groups, 12 mm conduits were used and sutured, ensuring that the proximal and distal nerve stump had no torsion and was in its original position (Fig. S15). Then, the muscle layer and skin were closed layer-by-layer with 5–0 and 4–0 absorbable surgical sutures, respectively.

2.7.2. Walking track analysis and sciatic functional index (SFI)

The functional recovery of the sciatic nerve was quantitatively evaluated at 4 and 12 weeks post-surgery using an automated gait analysis system (Animal Behavior Recognition Technology, XR-FP101, Shanghai XinRuan Technology Development Co., Ltd., Shanghai, China). Each rat was trained to walk freely along a transparent walkway equipped with a high-speed camera positioned below the lane. The system automatically captured dynamic gait videos and reconstructed footprint patterns in both two-dimensional (walking tracks) and three-dimensional (3D heatmap) formats. Additionally, the posterior limbs were dipped with nontoxic black inkpads and the footprints of the rats were recorded on the blank sheet of paper. The rat black footprints were obtained and calculated the SFI as follows:

$$\text{SFI}=-38.3\times \frac{\left(\text{EPL}-\text{NPL}\right)}{\text{NPL}}+109.5\times \frac{\left(\text{ETS}-\text{NTS}\right)}{\text{NTS}}+13.3\times \frac{\left(\text{EITS}-\text{NITS}\right)}{\text{NITS}}-8.8$$

Where E represents the experimental side, and N represents the normal side. Print length (PL): Distance between heel and third tiptoe. Toe spread (TS): Distance between first and fifth toes. Intermediary toe spread (IT): Distance between second toe and fourth toe.

2.7.3. Electrophysiological test

At 12 weeks postoperatively, electrophysiological testing was conducted on SD rats. Following abdominal anesthesia, the skin on the operated side of the rats was disinfected with iodine solution and the sciatic nerve was exposed. The stimulation electrode of the electromyography/evoked potential instrument (Keypoint 9033A07, Dandy, USA) was positioned proximally to the experimental site on the right side, while parallel recording electrodes were placed on the target gastrocnemius muscle. A stimulation current of 5 mA was performed and compound muscle action potential (CMAP) was recorded. The CMAP was also measured on the normal contralateral side as intrinsic control.

2.7.4. Evaluation of nerve regeneration

After the assessment of motor function recovery, animals were sacrificed and the right regenerated nerves were isolated at 4 and 12 weeks. Next, the nerve samples were fixed in 4 % paraformaldehyde, dehydrated in 10 %, 20 %, and 30 % gradient sucrose solution, then embedded in optimum cutting temperature (OCT, Beijing Pulilai Gene Technology Co., Ltd., China) and sectioned in 3–10 μm thickness either longitudinal or transverse sections using a freezing microtome (Thermo Fisher Scientific, USA). For histological observation, cross-sections were stained with Hematoxylin/Eosin (H&E Staining Kit, Solarbio, China) and toluidine blue (Toluidine Blue O Solution, Solarbio, China) to observe tissue morphology.

For transmission electron microscope (TEM), the isolated samples were immediately fixed in 3 % cold-buffered glutaraldehyde for 3 days, dehydrated in gradient acetone, embedded by epoxyresin, cut into ultrathin cross-sections around 50–70 nm thickness. After treated with uranyl acetate and lead citrate, sections were observed with a TEM (JEM-100 CX, JEOL, Japan). Myelin thickness distribution was calculated by Origin 2025b. And myelin thickness (by G-ratio) was measured using MyelTracer software (GitHub at the following URL: https://github.com/HarrisonAllen/MyelTracer).

For multiplex immunofluorescence staining, the sections were permeabilized in 0.3 % Triton X‐100 and blocked with 5 % goat serum in PBS solution for 3 h. Then, the sections were incubated with S100/NF-200, CD31, Arg-1, iNOS and anti-laminin antibody at 4 °C overnight. After three washes with PBST, sections were incubated with the secondary antibody for 2 h in the dark at room temperature. The secondary antibodies used were as follows: CoraLite488-conjugated Goat Anti-Rabbit IgG (H + L), CoraLite594-conjugated Goat Anti-Mouse IgG (H + L) (Tables 4 and 5). Cell nuclei were stained with DAPI and then imaged using a light microscope.

2.7.5. Gastrocnemius muscles evaluation

Morphology of gastrocnemius muscles was evaluated at weeks 12. Bilateral gastrocnemius muscles were dissected from individual rats and weighed. The muscle wet weight ratio rate was calculated according to: the weight of the experimental side divided by the weight of normal side × 100 %. Then, the gastrocnemius muscles were fixed, dehydrated, cross-sectioned, and subjected to Masson's trichrome staining (Solarbio, China).

2.7.6. Proteomic analysis

The harvested samples of rats were collected at 4 weeks after surgery and immediately stored in liquid nitrogen for proteomic investigation, with 3 replicate samples in each group. The samples were conducted by Beijing novogene Biotechnology Co., Ltd. Differentially expressed proteins (DEPs) were identified by comparative analysis between the nerve conduit and autograft groups. Subsequent functional enrichment analysis was performed using the Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) databases to explore the biological roles and potential regulatory mechanisms associated with the observed expression changes. Enrichment results were filtered based on statistical thresholds, including adjusted p-values and fold-change criteria, to ensure biological relevance. Circular heatmaps were further constructed to compare the expression distribution of proteins involved in key signaling pathways; and the protein-protein interaction (PPI) networks were generated based on the identified DEPs. To further confirm the molecular mechanisms revealed by transcriptomic analysis, WB analyses were performed on sciatic nerve tissues collected from SD rats at 4 weeks post-surgery in the autograft, gastrodin-PU-AT2.6 %, and gastrodin-PU-AT5 % groups. The detailed WB procedure was provided in the Supplementary Materials.

2.8. Statistical analysis

The images were processed using Image J software, while GraphPad Prism 9 (GraphPad Software, Inc. La Jolla, CA) was employed for plotting and statistical analysis. All experimental data were analyzed based on at least three parallel samples, with results presented as mean ± standard deviation. An unpaired t-test was used for comparisons between independent samples. One-way analysis of variance (ANOVA) followed by Tukey's post-hoc test was conducted for multiple comparisons. The levels of significance were *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

3. Results

3.1. Design and characterization of the materials

3.1.1. Chemical structure of copolymers

Previous study designed PU with appropriate contents of gastrodin to make elastomer. To endow the conduit with electrical activity, we further developed a series of gastrodin-PU-AT copolymers (Fig. S1). A FTIR spectra (Fig. 1A) demonstrated that a strong C=O stretching at 1724 cm⁻¹ and -NH stretching at 1568 cm⁻¹ belonged to the urethane (OCONH) groups of PU. Obviously, the characteristic benzene ring-related absorption peaks of AT were only observed in PU-AT and gastrodin-PU-AT polymers, with the band at 805 cm⁻¹ corresponding to the 1,4-disubstituted benzenoid ring, and the peak at 1163 cm⁻¹ attributed to benzenoid ring vibrational modes, indicating the successful incorporation of AT into the polyurethane backbone.

Fig. 1.

Characterization of copolymers and membranes. (A) FT-IR spectra of each copolymer. (B) ¹H NMR spectrum of gastrodin-PU-AT5 %. (C) UV–vis spectra of copolymers in TFEA solution. (D) CV scanning of copolymers in 1 M HCl. (E) Electrical conductivity of copolymers (n = 3). (F) SEM observation for surface morphology of fibrous membranes. (G) Diameter distribution of membranes. (H) Water contact angles of membranes (n = 5). a: PU; b: PU-AT2.6 %; c: PU-AT5 %; d: gastrodin-PU; e: gastrodin-PU-AT2.6 %; f: gastrodin-PU-AT5 %. **p < 0.01; ****p < 0.0001.

¹H NMR spectrum further confirmed the molecular structure of the copolymers (Fig. 1B and Fig. S2). The peaks at 1.29 ppm (h), 1.59–1.51 ppm (i,g), 2.70 ppm (j), 3.98 ppm (f), were assigned to PCL segment. The peaks at 0.98–0.86 ppm (c,d,e) of IPDI, could be clearly identified. As expected, the peaks of gastrodin presented at 4.41 ppm (b) and 7.2 ppm (a), which was absent in the PU. In addition, the peaks appearing at 7.5–6.5 ppm (m, n, o) were related to the aromatic protons in AT structure. The above results demonstrated that AT and gastrodin were successfully introduced into the backbone of PU.

3.1.2. Electroactivity of the copolymers

The UV–visible spectra (Fig. 1C) showed a highly electroactive AT in PU-AT and gastrodin-PU-AT copolymers, wherein two absorption peaks appeared at 300 and 552 nm corresponding to the π-π* transition of benzenoid ring and the excitonic transition from benzenoid to quinoid ring, respectively. However, due to the absence of AT in both PU and gastrodin-PU, no absorption peak appeared in the wavelength range of 300–900 nm.

The electroactivity of the copolymers was verified by CV measurements. Aniline oligomers have different oxidation state, including leucoemeraldine, pernigraniline, and emeraldine state, which change reversibly under different voltages or the action of oxidants/reducers. As shown in Fig. 1D and Fig. S3, the AT segment in the copolymer underwent two oxidation reactions. The first oxidation occurred at a voltage of 0.38 V, causing the leucoemeraldine state to oxidize to the emeraldine state. The other oxidized from the emeraldine state to the pernigraniline state at a voltage of 0.65 V. Nevertheless, no redox peaks were observed in PU and gastrodin-PU. The region enclosed by the CV curve reflected the electronic storage capacity, and the gastrodin-PU-AT5 % offered unparalleled advantages in electrical activity. As further confirmed by conductivity result, the conductivity of the copolymer gradually increased with the increase of AT content, most obvious in gastrodin-PU-AT5 % (Fig. 1E).

3.1.3. Morphology of membranes

The surface topography of electrospun membranes was analyzed by SEM. Fibers with uniform structure and smooth surface were obtained, forming a 3D interconnected network in Fig. 1F. Statistical data displayed that PU fibers incorporated gastrodin had higher diameter (1.44 ± 0.88 μm) than PU (0.58 ± 0.23 μm). Similarly, gastrodin-PU-AT2.6 % (1.49 ± 0.66 μm) and PU-AT2.6 % (1.15 ± 0.60 μm) also elevated, and the AT further prompted the diameter. However, the diameter in PU-AT5 % (0.76 ± 0.42 μm) dramatically decreased, especially in gastrodin-PU-AT5 % (0.44 ± 0.14 μm) (Fig. 1G). Furthermore, the measured water contact angle indicated increased hydrophilicity of the membranes (Fig. 1H), with PU grafted with gastrodin (69.57° ± 3.17°) exhibiting smaller angles than those PU (76.91° ± 4.65°). The incorporation of the AT increased the water contact angle, but there was no difference from PU-AT5 % to PU-AT2.6 %; however, it can be noted that grafted gastrodin slightly reduced the water contact angle, indicating the presence of gastrodin could improve the hydrophilic character of membranes.

3.1.4. Mechanical properties of conduits

The tensile strength of conduits was calculated from stress-strain curves (Fig. 2B). The tensile modulus of PU conduit was significant increased from 0.50 ± 0.14 MPa to 1.45 ± 0.22 MPa after grafted gastrodin. AT content had also an intense effect on the tensile stress. The synergy of gastrodin and AT remarkably enhanced tensile modulus, being high up to 2.18 ± 0.26 MPa for gastrodin-PU-AT5 % (Fig. 2C). Although PU was the lowest (0.50 ± 0.14 MPa), it was still higher than the native nerve of SD rat (0.12 ± 0.02 MPa, Fig. S5). The elongation at break followed an opposite trend to that of tensile modulus (Fig. 2D).

Fig. 2.

Mechanical properties and degradation behavior of conduits. (A) Digital photos of conduits (a, b) and their deformation under tensile stretching at different stages (c, d). (B) Tensile stress-strain curves. (C, D) Tensile modulus and elongation at break (n = 8). (E) Degradation curves and (F) Cumulative release of gastrodin over 12 weeks (n = 3). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

3.1.5. In vitro degradation behavior

The degradation behavior of the conduit in 0.05M NaOH solution was investigated. As can be seen in Fig. 2E, the conduit exhibited a gradient degradation behavior over a prolonged duration. The addition of gastrodin and AT were favorable for the degradation of PU. Compared with the PU group, the degradation of PU-AT2.6 % and PU-AT5 % increased, and the latter degradation was lower. For gastrodin grafted PU groups, the mass loss in gastrodin-PU-AT2.6 % was always superior to that in the gastrodin-PU-AT5 % and gastrodin-PU. Thereinto, the degradation in gastrodin-PU-AT5 % preceded gastrodin-PU at initial stages; however, gastrodin-PU accelerated degradation at week 4, there was no statistically significant differences between gastrodin-PU-AT5 % and gastrodin-PU at both subsequent time points. After 12 weeks, the gastrodin-PU-AT2.6 % and gastrodin-PU groups completely degraded, surpassing that of the gastrodin-PU-AT5 % group (81.85 ± 8.28 %). A sustained gastrodin release profile was consistent with degradation. Fig. 2F showed that gastrodin-PU-AT2.6 % had higher cumulative release levels than the other groups. The gastrodin-PU, which had the slowest release, caught up with gastrodin-PU-AT5 % at 4 weeks and surpassed gastrodin-PU-AT5 %. The release profile of AT in the gastrodin-PU-AT conduits was also plotted. Notably, the gastrodin-PU-AT5 % group exhibited higher AT concentration than the gastrodin-PU-AT2.6 % group over a 12-week time period, consistent with its higher initial AT loading (Fig. S7).

3.2. Electroactive membranes regulate cell behavior in vitro

SCs have significant roles in maintaining the microenvironment of the sciatic nerve and are involved in the regeneration of peripheral nerves following damage. The proliferation of SCs on the different membranes was evaluated by CCK-8 (Fig. 3A). Although the cell proliferation of all groups increased with time, the cell proliferation of gastrodin-PU and gastrodin-PU-AT groups appeared relatively high, while both PU-AT2.6 % and PU-AT5 % groups was slightly higher than PU. Moreover, increasing the AT content favored cell proliferation. The comparison revealed that the cell proliferation of gastrodin-PU-AT5 % group was the most significant, which indicated that gastrodin and AT might synergistically promote a positive environment conducive to cell survival. Live/dead results (Fig. 3B) showed that cells adhered and grew on the surface of each membrane on day 3. Compared with PU-AT2.6 % and PU-AT5 % groups, the number of living cells (green) in the PU group was worse than the former two groups. In addition, with the addition of gastrodin groups, such as gastrodin-PU, gastrodin-PU-AT2.6 %, and gastrodin-PU-AT5 %, the number of dead cells (red) was significantly smaller than that of the group without gastrodin. This difference was also reflected in the EdU results (Fig. 3C and D). The migration efficiency of SCs in response to the membranes was next studied via transwell assay. As shown in Fig. 3E, the crystal violet staining was enhanced in PU incorporating gastrodin or AT, of which both gastrodin-PU-AT groups were better than gastrodin-PU and PU-AT groups, suggesting optimal migration capacity of SCs by synergy between gastrodin and AT. Consistently, both gastrodin-PU-AT2.6 % and gastrodin-PU-AT5 % groups showed a higher level of IRSp53 gene compared to other groups (Fig. 3F), which exerted greater migration activity.

Fig. 3.

In vitro evaluation of nerve cells behavior. (A) CCK-8 assay for the proliferation of SCs on membranes (n = 3). (B) Live/dead fluorescence staining of SCs on membranes. (C, D) Representative EdU staining at 3 days and semiquantification of positively stained cells (n = 3). (E) Migrated cells in transwell were stained by crystal violet at 3 days. (F, G) Relative gene-expression activity of IRSP53, PMP22, Egr2, and GDNF. (H) GAP43 and cAMP proteins expression of PC12 cells as measured by WB (n = 5). Error bars represent the standard deviation from the mean. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

The effect of membranes on SCs’ myelination is also important because of its correlation to nerve regeneration. In view of this, the mRNA expression level of related markers was detected. As shown by RT-qPCR, the higher myelination genes of Egr2 and PMP22 was detected in gastrodin-PU-AT5 % than gastrodin-PU-AT2.6 %, followed by both PU-AT2.6 % and PU-AT5 % groups and gastrodin-PU, and least in PU. The GDNF gene were also enriched in gastrodin-PU-AT5 %, being higher than any other groups (Fig. 3G). Likewise, co-expression of S100 (green) and MBP (red, expressed in mature SCs) by immunofluorescence in gastrodin grafted PU groups was enhanced compared to PU or PU-AT groups, especially booming in gastrodin-PU-AT groups at day 1 (Fig. S9). The significant difference appeared to be pronounced after 3 days of incubation, thus indicating the promoting maturation of SCs by both gastrodin-PU-AT5 % and gastrodin-PU-AT2.6 % groups. The neuronal differentiation in PC12 cells were further quantified by WB (Fig. 3H). The levels of GAP43 and cAMP expression in all gastrodin-PU and gastrodin-PU-AT groups were higher than these in PU-AT groups, while PU group was the lowest. The results suggested that both gastrodin and AT had a potent enhancement in neurogenic signaling.

To better emulate the electrophysiological microenvironment, cells seeded onto various membranes with or without ES were explored. As displayed in Fig. 4B, the SCs cultured with membranes for 1 day, all showed great cell activity and myelination capacity (positive S100). The expression of S100 was comparable across the different groups. Upon induced by ES, number of spindle-shaped SCs was significantly higher compared to those without ES, particularly in the gastrodin-PU-AT5 % group. This trend consistently continued to the 3rd day, that SCs within the gastrodin-PU-AT5 % showed significant elongated cell bodies and axons. Then we carried out ES experiment based on PC12 cells, which exhibited neuron-like differentiation behavior. PC12 cells of PU-AT groups and gastrodin-PU-AT groups exhibited fusiform-shaped and elongated neurites from the body, confirming the differentiation of PC12 cells (Fig. 4C). Moreover, PC12 cells with ES showed more clearly differentiated with a higher percentage of differentiated cells and longer synapses than those without ES, suggesting that AT had a certain electrical conductivity and significantly promoted PC12 cell differentiation. The gastrodin-PU group also showed comparable neurite length, while PU group was discouraging. Notably, the outstanding neurite length of gastrodin-PU-AT5 % group over the other goups suggested the combined effect of ES-triggered and controllable gastrodin release further promoted neurite growth of PC12 cells. Obviously, the positive NF-200 cells (red) on gastrodin-PU-AT5 % without ES were dominant while other groups mainly possessed positive βIII-Tubulin cells. This expression trend of NF-200 was more prominent under ES (Fig. 4D). After 3 days of culture, the percentage of differentiated cells and the neurite length of PC12 cells on gastrodin-PU-AT5 % were significantly greater than those on other membranes with ES (Fig. S10). The above studies demonstrated that conductive gastrodin-PU-AT5 % and gastrodin-PU-AT2.6 % with ES could effectively regulate the maturation and differentiation of SCs and PC12 cells.

Fig. 4.

The morphology of SCs and PC12 cells on membranes in response to electrical stimulation. (A) Schematic diagram of cell culture under electrically stimulated (+ES) and non-stimulated (-ES) conditions. (B) Representative fluorescence images of SCs cultured on the membranes for 1 and 3 days with or without ES (green: S100; blue: DAPI). (C) Representative images of co-expressing NF200 and βIII-Tubulin of PC12 cells on the membranes for 1 day with or without ES. (D) Statistical analysis of the neurite length of differentiated PC12 cells (n = 3). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

3.3. In vivo sciatic nerve regeneration

To explore the effectiveness of the conduit for supporting nerve repair in vivo, the SD rats were used, in which, a 10-mm nerve gap was excised from the right sciatic nerve and bridged with autograft or gastrodin-PU-AT2.6 % or gastrodin-PU-AT5 %. There was no obvious inflammation, swelling, and neurofibroma in all rats at 4- and 12-weeks post-surgery. As shown in Fig. S14, the expression of iNOS (M1) weakened with the extension of implantation time in all groups. Significantly, the persistent high expression of Arg-1 (M2) in gastrodin-PU-AT5 %, that was comparable to autograft, could facilitate the neuronal extension of the defective nerve.

HE-staining areas represented the growth of new nerve cells and tissues, and the formation of new blood capillary was crucial for neural tissue regeneration. At 4 weeks post-operation, the gastrodin-PU-AT5 % group demonstrated a more pronounced positive regions than gastrodin-PU-AT2.6 % group, suggesting more newly nerve cells and tissue extended into the entire spiral conduit. The newly formed nerve tissue within the gastrodin-PU-AT5 % conduit was similar to that in the autograft group, with well-arranged fibers. Moreover, the gastrodin-PU-AT5 % group had neovascularization density of approaching to autograft. In the gastrodin-PU-AT2.6 % group, only a small number of new blood vessels containing red blood cells (indicated by green stars) growing into tissues were observed, while the nerve fibers appeared disorganized. At 12 weeks, the gastrodin-PU-AT5 % group contained a higher level of regenerative tissue density as compared with the gastrodin-PU-AT2.6 % group. In magnified cross-sections, the regenerated tissue area of organized arrangement in gastrodin-PU-AT5 % group was significantly larger than that of the gastrodin-PU-AT2.6 % group, while vessel signal comparable to autograft group but significantly larger than gastrodin-PU-AT2.6 % group (Fig. 5A).

Fig. 5.

Neovascularization of regenerated nerve fibers at 4- and 12- weeks post-implantation. (A) H&E staining of cross-sections of regenerated nerve tissue (red blood cells indicated by green stars). (B) Immunofluorescent staining of CD31 (red) and nuclei (blue) (neovascularization indicated by green arrows). C: conduit; N: nerve tissue.

The expression levels of the vessel-specific marker CD31 were further detected by immunofluorescent staining in the middle cross-sections of the regenerating nerves (Fig. 5B and Fig. S16). The autograft group exhibited the highest density of positive area for CD31 at each detection time point. At 4 weeks post-surgery, the gastrodin-PU-AT5 % conduit supported for infiltration of red blood cells, with most of the forming vessel in line with the fiber direction, and the density of regenerative vessels was significantly higher than that of the gastrodin-PU-AT2.6 % conduit. In addition, the morphology of regenerating vessel in the gastrodin-PU-AT5 % group was the same as that in the autograft group. At 12 weeks post-surgery, a more substantial capability for regenerating vessels were improved in both autograft and gastrodin-PU-AT5 % groups, but was less in gastrodin-PU-AT2.6 % group; indicated that the gastrodin-PU-AT5 % group excelled in promoting vascular tissue regeneration.

The biological signals (NF-200 (biomarker for neurofilament) and S100 (biomarker for SCs)) were detected by immunofluorescent staining in regenerating nerves. At 4 weeks postoperatively, gastrodin-PU-AT5 % group showed much stronger signals compared with gastrodin-PU-AT2.6 % group, but weaker than the autograft group. Notably, both gastrodin-PU-AT5 % and gastrodin-PU-AT2.6 % groups displayed positive signals within the spiral channels, and the intensive signal was detected in the gastrodin-PU-AT5 % group (Fig. 6A). This finding was further confirmed by the longitudinal section of the regenerating nerves, where new neurofilaments and SCs were occurred at the site of nerve defect. Moreover, all groups guided ingrowth and bridge of nerves from proximal and distal ends, with most of the nerves in line with the spiral conduit (Fig. 6B). The density of nerves and S100⁺ SCs in gastrodin-PU-AT5 % group was significantly higher than that of the gastrodin-PU-AT2.6 % group.

Fig. 6.

In vivo sciatic nerve regeneration assisted by spiral conduits at 4 weeks after implantation. Double immunofluorescent staining of (A) transverse and (B) longitudinal sections of explanted conduits, showing the distribution of neurofilaments (NF200, red), SCs (S100, green), and nuclei (DAPI, blue). Take the explanted conduits and divide them into three equal zones marked as proximal, middle, and distal segments. C: conduit; N: nerve tissue.

IF staining of the regenerated nerve tissue at 12-weeks post-operation further revealed that autografts completely supported the rapid convergent extension of nerve fibers (Fig. 7 and Fig. S17). The gastrodin-PU-AT5 % group supported nerve regeneration accompanied by migration of SCs, while the gastrodin-PU-AT2.6 % group had the lowest level of nerve regeneration among the three groups. Meanwhile, transverse images of the middle segment of the conduit showed that the extensive neurofilament ingrew into the conduit and higher nerve regeneration density covered the center (Fig. 7A). Moreover, in terms of longitudinal sections, the gastrodin-PU-AT5 % group almost completely bridged the defect with regenerated nerve fibers, and possessed substantial colocalization of SCs and neurofilaments throughout the spiral conduits. In contrast, there were few regenerated nerves and only a few SCs infiltrated the middle segment of gastrodin-PU-AT2.6 % group (Fig. 7B). These results indicated that gastrodin-PU-AT5 % incorporating spiral conduits with micro/nano-structures could effectively promote neurofilament ingrowth and myelin sheath formation in newly formed nerve tissue, further facilitating axonal regeneration. In addition, the conductive and gastrodin-grafted properties were also of considerable importance for nerve defect repair.

Fig. 7.

In vivo sciatic nerve regeneration assisted by spiral conduits at 12 weeks after implantation. Double immunofluorescent staining of (A) transverse and (B) longitudinal sections of explanted conduits, showing the distribution of neurofilaments (NF200, red), SCs (S100, green), and nuclei (DAPI, blue). Take the explanted conduits and divide them into three equal zones marked as proximal, middle, and distal segments. C: conduit; N: nerve tissue.

Myelination is the hallmarks of successful sciatic nerve regeneration. The TEM and TB staining were performed on the regenerated middle nerves at the 12th week to observe the formation of axons and myelin sheath. Fig. 8A showed that most of the regenerated axons were enveloped by thick and dense myelin sheaths. The gastrodin-PU-AT2.6 % group had the thinnest myelin sheath, while the gastrodin-PU-AT5 % group presented a slightly thicker myelin sheath but still less than the autograft group (Fig. 8B and C). The average diameter of regenerated axons showed no statistical difference among three groups (Fig. 8E). Besides, the G-ratio of the regenerated axons was elevated in the gastrodin-PU-AT groups, especially gastrodin-PU-AT2.6 % (Fig. 8D). SCs growing into myelin sheaths within the newly generated nerves were stained blue with TB (indicated by yellow arrows). The myelin sheath-positive area (dark blue area) in gastrodin-PU-AT5 % group was comparable to that of the autograft, somewhat higher than that in gastrodin-PU-AT2.6 % group (Fig. 8F). The results jointly demonstrated the remarkable repair effect of the gastrodin-PU-AT5 % conduit and advantages in guiding nerve regeneration.

Fig. 8.

The remyelination of newly formed nerves at 12 weeks after implantation. (A) TEM images of the transverse sections at the middle of the regenerated nerve segment. Statistical analysis of regenerated nerve fibers was performed by counting (B) myelin thickness distribution, (C) thickness of myelin sheath (n = 40), (D) G-ratio (n = 11) and (E) axon diameter (n = 6). (F) Toluidine blue staining. **p < 0.01; ****p < 0.0001. Values are presented as mean ± SD.

The repair effect of the conduit was synchronously evaluated by monitoring the restoration of motor function in rats. As shown in Fig. 9A, the gastrodin-PU-AT5 % group exhibited better functional recovery compared with gastrodin-PU-AT2.6 % group throughout the 4–12 weeks period, approaching those in the autograft group. The SFI results showed a trend of recovery in all three groups. When extended to 12 weeks, the SFI in gastrodin-PU-AT5 % and gastrodin-PU-AT2.6 % groups progressively increased although remained worse than that in the autograft group (Fig. 9B). The denervation of the gastrocnemius muscle was observed by the digital photographs (Fig. 9C), of which gastrocnemius muscles became smaller and stiffer on the experimental side than the normal contralateral side and had not recovered to the normal level at 12 weeks. Compared with autograft group, the gastrocnemius muscles atrophy in the gastrodin-PU-AT5 % group was more severe, followed by the gastrodin-PU-AT2.6 % group. The wet weight of gastrocnemius muscle was also an important index for evaluating nerve regeneration. As shown in Fig. 9E, the wet weight ratio in the gastrodin-PU-AT5 % group showed no statistical difference from that in the gastrodin-PU-AT2.6 % group at 4 weeks post-surgery and was lower than that observed in the autograft group. All groups maintained an upward trend and significant differences appeared to be pronounced at the 12-week implantation timepoint, exerting a more promoting effect on tissue regeneration. Masson's trichrome staining of cross-sectional gastrocnemius muscle in injured site displayed that some blue-colored collagen deposition appeared in all groups, while the muscle tissue in the both gastrodin-PU-AT groups was more compact and adhesive than autograft (Fig. 9D). Subsequently, complete muscle fiber interfaces and less collagen infiltration were achieved in all groups, remarkedly relieving denervated atrophy and degeneration. Collectively, the morphology of the muscle in the autograft and both gastrodin-PU-AT groups were relatively regular and uniform. Similar results were further confirmed through immunofluorescent staining using anti-laminin (Fig. 9F).

Fig. 9.

Functional evaluation of the regenerated nerves at 4- and 12- weeks post-implantation. (A) Automated gait analysis of rats in different experimental groups at 4 and 12 weeks postoperatively. 3D footprint heatmaps and corresponding 2D walking tracks. LH: normal side; RH: experimental side. (B) Sciatic functional index (SFI) in each group. (C) Gross images of the isolated gastrocnemius muscles. (D) The cross-sectional view of the gastrocnemius muscles highlighted by Masson's trichrome stain. red: muscle fibers; blue: collagen fibers. (E) Wet weight ratio of gastrocnemius muscles. (F) Anti-laminin immunofluorescent staining images of cross-sections of gastrocnemius muscles. (G) Illustration of the CMAP testing protocol. (H) Comparison of electrophysiological recordings of CMAP. (I) The peak amplitude of CMAP and (J) statistical analysis of latency. Error bars represent the standard deviation from the mean (n = 4). *p < 0.05; **p < 0.01; ***p < 0.001.

The electrical signal reaches the neuromuscular junction can induce contraction of the target muscle, indicating the recovery of motor function. Neuroelectrophysiological detection was then performed and the CMAP were recorded using recording electrodes (Fig. 9G). Although autograft had superiority at 4 weeks, these difference between autograft and gastrodin-PU-AT5 % groups was relatively small (Fig. 9H and I). At 12 weeks post-surgery, the autograft group had higher peak of CMAP than the gastrodin-PU-AT5 % group, followed by gastrodin-PU-AT2.6 % group. Statistics on conduction latency was shown in Fig. 9J, no significant difference of latency was observed among three groups. The peak potential of CMAP directly reflects neuromotor function reconstruction after re-innervation, with a degree of myelination of the regenerated nerve through the conduction speed of the nerve signal, indicating a good communication between regenerated nerves and muscle at the injured site.

To reveal the underlying mechanism and related bioevents of conduit-mediated peripheral nerve regeneration, an integrated assessment of hierarchical clustering and functional enrichment was conducted (Fig. 10). Quantitative proteomic analysis uncovered distinct molecular profiles among the autograft and gastrodin-PU-AT groups, as shown in the hierarchical clustering heatmap (Fig. 10A). Notably, the gastrodin-PU-AT5 % group exhibited a distinct proteomic profile compared to the autograft group, with a moderate divergence observed from the gastrodin-PU-AT2.6 % group as well, indicating a potential dose-dependent modulation of the regenerative microenvironment. GO enrichment analysis (Fig. 10B and Table S7) indicated that these proteins were involved in biological processes (BP), key enriched terms included immune-related responses (e.g., defense response, activation of immune response), actin cytoskeleton dynamics (e.g., Arp2/3 complex-mediated actin nucleation), and cellular regulation (e.g., regulation of TOR signaling, intracellular signal transduction). In cellular components (CC), proteins were significantly enriched in the extracellular matrix and Arp2/3 complex. In molecular functions (MF), oxidoreductase activity, cytokine activity, and various peptidase functions were dominant, suggesting an active remodeling of the tissue microenvironment and redox regulation during repair. KEGG enrichment analysis (Fig. 10C and Table S6) showed immune and inflammatory pathways were enriched in gastrodin-PU-AT groups, such as Fc gamma R-mediated phagocytosis, lysosome, phagosome, NOD-like receptor signaling, and Toll-like receptor signaling. In screening for differentially expressed proteins associated with nerve regeneration, we noted a significant elevation in the expression related to Rap1 (the markers of actin cytoskeleton remodeling and cell adhesion, Itgal, Thbs1, Rac2, and Pfn1) and mTOR signaling pathways (the markers of axonal elongation and remyelination, Atp6v1a, Pik3r1, and Lamtor3) in gastrodin-PU-AT5 % compared to other groups (Fig. 10D–G). The shared components such as Pik3r1 and Map2k1 in both pathways further testified signaling crosstalk that coordinated immune modulation with structural repair, thereby promoting the nerve repair. The result was evidenced by WB analyses of sciatic nerve samples at 4 weeks post-surgery (Fig. 10H and I). Both gastrodin-PU-AT groups exhibited elevated phosphorylation of ERK and MEK relative to the autograft group, reflected by significantly increased p-ERK/ERK and p-MEK/MEK ratios. To a certain extent, the p-mTOR/mTOR ratio in gastrodin-PU-AT5 % group was comparable to that in autograft, higher than the gastrodin-PU-AT2.6 % group, suggesting the activated Rap1/mTOR axis during nerve regeneration.

Fig. 10.

Proteomic profiling in various nerve repair strategies. (A) Heatmap visualization of differentially expressed proteins among the autograft, gastrodin-PU-AT2.6 %, and gastrodin-PU-AT5 % groups. (B) KEGG pathway enrichment analysis of differentially expressed proteins between the autograft and conduit groups. Pathways with adjusted p-values <0.05 were considered significantly enriched. (C) GO enrichment analysis of the same protein sets. GO terms were categorized into biological process, cellular component, and molecular function, and enrichment significance was determined by p-values. (D, E) Circular heatmap of protein expression profiles in the Rap1 (D) and mTOR (E) signaling pathways. (F, G) PPI network of Rap1 (F) and mTOR (G) signaling pathways. (H) Representative WB images showing the expression and phosphorylation levels of ERK, MEK, and mTOR. GAPDH was used as the internal loading control. (I) Quantitative analysis of the relative protein expression ratios of p-ERK/ERK, p-MEK/MEK, and p-mTOR/mTOR. Data are presented as mean ± SD (n = 3). *p < 0.05; **p < 0.01.

4. Discussion

In neural tissue engineering, the biomimetic fascicular architecture within a hollow nerve conduit has proven to be a simple yet effective strategy for nerve regeneration. Spiral NGCs may provide adequate spatial cues for guiding axon extension while preventing the invasion of surrounding connective tissue [38]. Therefore, we developed a 3D fiber conduit with oriented spiral structure, which were manually rolled up using a 2D fiber membrane. It can circumvent the limitation that excessive fillering materials within the conduit hinder the migration of regenerating axons and permeability [39,40]. To promote axon regeneration, the design of our gastrodin-PU-AT conduit has considered the hierarchical structure, including the molecular modification (AT and gastrodin) of PU matrix, the introduction of fibers, and the formation of spiral structure. On a microscale, the oriented spiral layers and interlayer spacing with robust surface area to volume ratio were appropriate for the directional growth of cells and the regeneration of axons (Fig. 6, Fig. 7). On a nanoscale, electrospun fibers simulated the microstructure of extracellular matrix and provided appropriate pore size for cell adhesion, while maintaining the mechanical stability of conduits (Fig. 1, Fig. 2B-D). On a molecular scale, the introduction of gastrodin not only improved hydrophilicity (Fig. 1H) and degradability of conduits (Fig. 2E and Fig. S6) but also provided immune microenvironment in defective site (Fig. S14). Equally important was the conductivity of the AT (Fig. 1D and E), especially the effective bioelectricity to talk with nerve cells and tissues (Fig. 4B and C). This unique hierarchical structure is crucial for high-performance nerve conduit that empowers neural pathway for axonal extension.

A suitable electrical signal is necessary for the reconstruction of the physiological microenvironment that transmits information primarily through action potentials generated by synapses. Many studies have shown that conductive materials can transmit electricity or generate it upon external stimulation and activate critical intracellular signaling pathways, stimulating the proliferation and differentiation of nerve cells [[41], [42], [43]]. AT was chemically grafted onto PU matrix, and incorporating gastrodin further improved AT dispersion. As expected, increased conductivity was found in the gastrodin-PU-AT5 % than in the gastrodin-PU-AT2.6 %. The π-π* interaction and hydrogen-bonding would facilitate stronger cell attachment. Consequently, both gastrodin-PU-AT5 % and gastrodin-PU-AT2.6 % groups exhibited significantly SCs proliferation and migration (Fig. 3A, B, E), which was associated with upregulation of expression IRSp53 (Fig. 3F). IRSp53 is involved in the formation of neural synapses and is closely related to the coupling of filamentous pseudopodia formation and actin dynamics, affecting cell movement and migration processes [44]. Moreover, during peripheral nerve regeneration, SCs clear axonal debris and align to form Büngner bands that provide structural and biochemical cues for axonal regrowth [45,46]. Consistently, our PC12-SCs co-culture showed enhanced neurite outgrowth and SCs proliferation on gastrodin-PU-AT conduits, creating a favorable microenvironment that mimics neuron-Schwann cell interactions in vivo (Fig. S13). In the early stage of nerve injury, IRSp53 may promote migration of SCs to the injured site, recruit SCs to release cytokines including Egr2 (necessary for cell myelination) [47] and GDNF (protecting motor neurons and promoting myelination) [48]. And more crucially, myelination of the regenerated axons are critical events in the functional repair of the peripheral nerve damage [49]. The improvement on MBP and S100, two parameters reflecting SC-like cell differentiation, was found to be robust and rather close between the gastrodin-PU-AT5 % and gastrodin-PU-AT2.6 % (Fig. S9). Of note, the gastrodin-PU-AT5 % had elevated expression of Egr2 and GDNF, which was heavily depended on the preferable electroactivity and nanofiber. Such an electrical microenvironment effectively conducts nerve cells impulses through exogenous ES, further promoting the elongation and growth of SCs and neurite outgrowth of PC12 cells (Fig. 4B and C). This can be evidenced by Brianna C Thompson et al. [50] that conductive biomaterials cooperated with electrical stimulation could induce axonal elongation and promote axonal growth. Consistently, gastrodin-PU-AT5 % conduit enhanced actin cytoskeleton remodeling and cell adhesion proteins in regenerated nerves of rats via activating Rap1 signaling pathway compared to others, thus promising migration of Schwann cell and guidance of regenerating axons (Fig. 10). Another pathway for crosstalk with it, mTOR signaling, upregulated metabolic reprogramming and lysosomal activity for remyelination. Importantly, the conductive microenvironment modulated intracellular signaling crucial for axonal regeneration. Both gastrodin-PU-AT groups activated the Rap1/ERK-mTOR axis, with p-ERK upregulation and a dose-dependent biphasic change in p-mTOR ([51,52]). This suggests dynamic ERK-mTOR crosstalk, where enhanced conductivity at higher AT loading relieved feedback inhibition and restored mTOR activation [53,54], maintaining a balance between regenerative signaling and metabolic regulation. Uncontrollable inflammation upon implantation can impair cell function and nerve regeneration, ultimately affecting effect of repair [55]. Our previous study constructed a hollow gastrodin-PU conduit, the anti-inflammatory activity and neural protection of gastrodin stimulated function of SCs for facilitating nerve regeneration [36]. In addition, electrical microenvironment showed capability to regulate macrophage polarization. AT-incorporated conductive NGCs reported by Zhang et al. had manifested an immunomodulatory capacity that induced macrophage polarization to the pro-healing M2 phenotype [29]. A higher incidence of M2 macrophages within gastrodin-PU-AT5 % compared to gastrodin-PU-AT2.6 % group further confirmed the importance of electrical activity (Fig. S14). Immuno-inflammatory pathways, such as Fcγ receptor-mediated phagocytosis [56], NOD-like receptor signaling, and Toll-like receptor signaling [57], were better in gastrodin-PU-AT groups (Fig. 10). A significant protein enrichment in gastrodin-PU-AT5 % might orchestrate the early-phase tissue repair response and regulate the immune microenvironment following nerve injury. Furthermore, newborn capillaries with red blood cells inside were distributed in regenerated nerve tissues in all groups after 4 and 12 weeks, providing the requisite nutrition transportation and exchange. Such findings were consistent with previous studies that sustained release of gastrodin promoted angiogenesis [58]. The microvessel density in gastrodin-PU-AT5 % group was heightened significantly. Therefore, the local application of electroactive AT could be an effective complement to gastrodin-release system for regulating the repairing microenvironment.

Furthermore, the NGC with both structural guidance and electrical cues is of great benefit in contact of axons and the transmission of electrical signals. Here, the oriented spiral conduits with micro/nano fiber structure gave the conduit architectural modifications at both micro- and nanoscale levels, which could deliver directional induction for axonal extension. The multiple channels would facilitate the SCs attachment and arrangement. Usually, the aligned SCs tend to form a Büngner band to guide the axon to grow in the correct direction [59]. Our results indicated that the rolled-up spiral conduits enhance the migration, proliferation, and myelination of SCs and axon extension to promote functional regeneration of damaged nerves. The density of regenerated nerves significantly increased in the nerve bridges of gastrodin-PU-AT5 % at 4 and 12 weeks compared to gastrodin-PU-AT2.6 % in H&E staining (Fig. 5A). Such a notable difference implied that 2D fiber membrane with clear topological cue and small fiber size were more effective at cellular behavior. The nanofibers were more favorable attachment of nerve cells than microfibers [60]. Our data provided strong evidence showing that intraluminal guidance structures encouraged SCs growth and neurite outgrowth. Accordingly, nerve fibers indicated by NF-200 could shuttle through different spiral layers of the conduit in both the gastrodin-PU-AT groups after 4-week transplantation (Fig. 6). Meanwhile, positive S100 staining results supported that SCs migrated and proliferated from the nerve stump to the middle, assisting in nerve repair. This growth trend continued until 12 weeks, a larger amount of newborn nerve fibers and SCs infiltrated the entire nerve conduit (Fig. 7). The fluorescence staining results of the longitudinal section corroborate with these findings, that is, during nerve injury repair, the SCs firstly migrate from the broken end to the middle, and then nerve fibers are gradually guided and repair the nerve [45]. This was consistent with previous reports that have demonstrated that spiral nerve conduits greatly enhance axon extension and guidingly bridge nerve gap during sciatic nerve regeneration; and the regenerated axons extend more evenly along the conduit wall, providing an optimal strategy for uneven fiber distribution [61].

The myelination of regenerated axons greatly favors the signal transduction and functional recovery [62]. A large axon and a thick myelin sheath indicate successful regeneration of the sciatic nerve [63,64]. The regenerated axon in gastrodin-PU-AT5 % group had thicker and larger myelin sheath similar to the axon of autograft than that in gastrodin-PU-AT2.6 % group, and the enhancement was further confirmed by TB staining (Fig. 8F). This better explained that the utilization of gastrodin-PU-AT5 % improved functional recovery of the regenerated nerve and its affected muscles (Fig. 9). The conductive microenvironment created by AT promoted GDNF expression, a need for preventing muscle atrophy [65]. The myelinated nerve fibers with the restored microstructure, as well as myelinating SCs, revealed a significant role of the gastrodin-PU-AT5 % conduit in promoting the nerve regeneration. We also noticed that in the animals receiving the gastrodin-PU-AT treatment, the gastrocnemius muscle underwent more atrophy than the autograft group. This suggests a further optimized strategy that embedded neurotrophic factors as auxiliary fillers inside the spiral multi-channels, or incorporating oriented fiber patterns, is needed. Further extending recovery period may provide a synchronized process between material degradation and tissue regeneration. And decode of exact sensory and motor function mediated by gastrodin-PU-AT and the spiral channels will improve the nerve repair strategy.

5. Conclusion

With the aim of developing an effective scaffold for the repair of long-distance PNI, we fabricated a 3D fibrous conduit with oriented longitudinal micro-channels, by integrating electrospinning and manual curling techniques. The scaffolds provided the desired topological morphology, mechanical stability, sequential gastrodin release, and electrical conductivity, which further promoted the elongation and growth of SCs and neurite outgrowth of PC12 cells. Importantly, animal studies demonstrated that the scaffold guided long-distance ( > 10 mm) directional migration of cells and promoted axonal elongation and remyelination via activation of Rap1 and mTOR signaling pathways, being along with alleviating inflammatory microenvironment. Further, introducing bioelectrical AT in the scaffold enabled nerve impulse transmission and muscle receptivity, thereby preventing muscle denervation atrophy and promoting functional recovery. The results jointly make us believe that the scaffolds are a promising candidate for repairing longer nerve defects.

CRediT authorship contribution statement

Xiaoqian Lan: Writing – original draft, Software, Methodology, Conceptualization. Guangli Feng: Validation, Investigation. Qing Li: Validation, Investigation. Shiyi Qin: Visualization, Investigation. Yingrui Hu: Investigation. Shilin Pan: Investigation. Jianlin Jiao: Investigation. Di Lu: Writing – review & editing, Supervision, Project administration, Funding acquisition, Conceptualization. Lianmei Zhong: Writing – review & editing, Supervision, Funding acquisition.

Ethics approval and consent to participate

The animal experiments were approved by the Experimental Animal Ethics Committee of Kunming Medical University (Protocol No. 20221734).

Declaration of competing interest

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

Appendix A. Supplementary data

The following is the Supplementary data to this article: