Hierarchical piezoelectric conduit coordinates bioelectric cues and microRNA regulation for functional neurological regeneration

Abstract

Current clinical interventions for critical-sized nerve defects face dual challenges of inadequate microenvironmental guidance and passive regenerative response in synthetic neural interfaces. We present an electrotopographic nerve guidance conduit with triaxial architectural hierarchy that actively coordinates mechanoelectrical signaling, biochemical modulation, and directional axonotropism through self-sustained piezoelectric dynamics. The trilaminar architecture comprises: an inner piezoelectric PLL-PLLA nanofibrous layer generating endogenous electrical fields under physiological deformation while providing contact guidance via aligned grooves, a collagen-PLLA transitional layer maintaining mechanical compliance matching native perineurium, and an outer genipin-gelatin cross-linked sheath. In rat sciatic nerve defects, the conduit achieved accelerating functional recovery, outperforming autografts in myelinated axon density and motor endplate reinnervation. Through integrated transcriptomics and functional genomics, we identified a piezoelectric-triggered miR-30a-5p/PI3K-AKT axis driving BMSCs neurodifferentiation. The directional electroactivity enhanced schwann cell migration and axonal alignment precision. This work establishes a paradigm-shifting strategy for autologous neuromodulation by harnessing physiological motion-derived electromechanical energy, circumventing external power dependency while achieving spatiotemporal precision in neural repair. The self-sustaining therapeutic system holds translational promise for addressing critical challenges in long-gap nerve reconstruction.

Graphical abstract

Highlights

• •Hierarchical piezoelectric conduit enables self-powered bioelectrical guidance and ectopic sprouting prevention.
• •Topological-electrical gradient design promotes repair of sciatic nerve defects in rats.
• •Mechanistically modulates miR-30a-5p/PI3K-AKT axis via endogenous electromechanical cues.

1. Introduction

Nerve guiding conduits (NGCs), as a fundamental raw material for tissue engineering, are primarily employed to facilitate nerve regeneration by creating a regenerative microenvironment post-nerve injury, accordingly enhancing the adhesion, proliferation, and differentiation of nerve cells [1,2]. In recent years, many researchers have focused on the advancement of inducibly active and functionalized NGCs to enhance the efficacy of repair [1,3]. The topography of the materials is a crucial physical signal in neural tissue engineering, and parallel topographical morphology may act as an independent factor to enhance the neural differentiation of stem cells into neurons [4,5]. The microenvironment created only by topological substrates is restricted by the absence of neural inductive factors and transmission of electricity, which impacts the efficacy of nerve-guided conduits for nerve wound repair.

The external introduction of neurotrophic substances enhances regeneration after injury to the nerves. The substantial expense of exogenous induction factors and their propensity to diffuse into adjacent tissues have significantly restricted the utilization of exogenous neurotrophic factors [6]. L-Polylysine (PLL) exhibits thermal resistance and enhances cellular adhesion [7], exhibits no toxicity to cells and possesses excellent biocompatibility [8]. L-Lysine, the degradation product of poly-L-lysine, is an essential amino acid and a neurotrophic factor; its lack may lead to central nervous system problems and cognitive impairment [9]. Concurrently, PLL can serve as a platform for drug delivery to facilitate sustained drug release [10].

Cellular electrical activity is crucial in the initial development of the central and peripheral nervous systems, governing the differentiation and migration of neural stem/progenitor cells [11]. Optimal electrical stimulation (20 Hz) enhances axonal development and promotes regeneration innervation of end-organs [12]. Nonetheless, contemporary electroactive materials necessitate the implantation of electrodes, thereby elevating the risk of surgical complications and infections [11]. Piezoelectric materials can convert mechanical pressure into pulsed electrical signals without the need for external electrodes. The electrical signals are modulated through the autonomic nervous system, synchronizing with physiological electrical stimulation [13]. Polyvinylidene fluoride (PVDF) and its copolymers exhibit strong piezoelectric properties and can effectively promote neuronal axon growth and peripheral nerve regeneration [14,15]. However, their non-biodegradable nature poses long-term risks of fibrotic encapsulation and chronic nerve compression, limiting their translational potential for resorbable conduits. In contrast, Poly(3-hydroxybutyrate) (PHB) combines piezoelectric properties with degradability, and PHB/Fe₃O₄/CA composite conduits have been shown to effectively enhance nerve regeneration [16,17]. Nevertheless, its inherent brittleness and high crystallinity (common feature among inherently piezoelectric biopolymers) lead to a mechanical mismatch with soft neural tissues. Furthermore, its degradation rate is highly dependent on both enzymatic activity and pH levels within the local microenvironment [[18], [19], [20]]. Poly(L-lactic acid) (PLLA) can be derived from renewable sources such as corn or sugar [21,22]. Its piezoelectricity originates from the collective orientation of C=O dipoles upon uniaxial stretching [23]. Notably, electrospinning inherently induces this chain stretching, generating a d14 shear piezoelectric response without the need for post-fabrication electrical poling. This represents a distinct advantage over traditional piezopolymers [24]. Additionally, the high-speed rotation of the collector during electrospinning not only aligns the fibers macroscopically to provide topographical guidance, but also induces molecular chain orientation along the fiber axis through extensional flow, thereby amplifying the shear piezoelectric response of the PLLA membrane [25]. Meanwhile, PLLA membranes exhibit favorable biocompatibility and degradability [26]. However, their hydrophobic surface is unfavorable for cell adhesion. To promote close attachment between the cell membrane and the material surface, the positive charge inherent to poly-L-lysine (PLL) can be utilized. It binds tightly to the cell membrane through electrostatic adsorption, thereby enabling cells to sense the piezoelectric signal more sensitively [27,28].

This study proposes a three-layer structured nerve guidance conduit to promote repair of sciatic nerve defects. PLL-PLLA served as the inner layer for the construction of the nerve conduit, featuring an electrically active and parallel topography. The cross-linking of the middle layer (Gelatin) with genipin was employed to create the outermost part of the conduit, imparting a specific degree of toughness. This approach yielded a triple-layered nerve conduit capable of effectively promoting axonal regeneration while exhibiting superior biocompatibility. Utilizing the piezoelectric properties of the material surface to simulate cellular electrical signaling and simulate peripheral nerve structure through highly parallel rotating morphology, this approach provides a favorable regenerative microenvironment and a guiding scaffold for peripheral nerve injury. Furthermore, this study examined the potential molecular mechanisms and the impact of PLL concentration on neural differentiation of stem cells. We investigated the role of miR-30a-5p in PLL-PLLA electrospun membrane-induced neural differentiation and the associated changes in the PI3K/AKT signaling pathway.

2. Materials and methods

2.1. Preparation of scaffolds generated from PLL-PLLA via electrostatic spinning technology

Parallel-type PLL-PLLA membranes were prepared by magnetic stirring of PLL-PLLA solutions with varying PLL concentrations (using a 7% PLLA base solution), followed by electrospinning under the parameters: 15 cm injection distance, 16 kV voltage, 23# metal needle, 0.7 mL/h flow rate, and a high-speed roller collector. The electrospun films were vacuum-dried at room temperature to remove residual solvents.

The liquid thickness of the smooth PLLA material (flat) measured 75 μm. A smooth surface material of PLLA was created by uniformly applying 7% PLLA spinning solution onto a quartz glass plate. The glass plate was subsequently heated to 35 °C on a platform overnight.

2.2. Characterisation of PLLA/PLL-PLLA electrospun membrane and flat membrane

SEM was employed to evaluate the material's surface topography, as well as the uniformity of the alignment and diameter, among other parameters. The elemental composition of the material's surface was determined through the application of EDS. The elemental bromine content of the PLL-PLLA electrospun membrane was determined through ICP-MS to evaluate the release of PLL from the material. Subsequently, piezoelectric devices were constructed using nanofibers, and the evaluation of the piezoelectric properties of the nanofiber membranes was conducted with a CHI760E electrochemical workstation (Chenhua, Shanghai).

2.3. Phalloidin staining, Transwell migration assay

Human bone marrow stem cells (hBMSCs) were purchased from Cyagen Biosciences (Guangzhou, China). hBMSCs were cultured on the surfaces of PLL-PLLA electrospun membranes and flat membranes. Phalloidin staining was utilized on these hBMSCs. Cells were fixed with 4% paraformaldehyde (PFA), permeabilized using Triton X-100, and subsequently stained with phalloidin on the electrospun membrane surface. To assess the effect of PLL-PLLA electrospun membranes on hBMSC migration, a Transwell co-culture system was used. PLLA and PLL-PLLA membranes were placed in the lower chamber, and approximately 10⁴ hBMSCs were seeded in the upper chamber. After 24 h, cells were fixed with 4% paraformaldehyde and nuclei were stained with Hoechst.

2.4. Cell counting kit (CCK-8) experiments

The proliferation capacity of hBMSCs was assessed utilizing CCK-8. After cultivating hBMSCs on the spinning membrane surface for various times (0d, 1d, 3d, and 5d), 10% CCK-8 solution was added to each well. The OD values were subsequently quantified at the wavelength of 450 nm utilizing a microplate reader. Each group underwent four replicates.

2.5. Calcein AM/PI assay

The influence of each group on the bioactivity of hBMSCs was evaluated utilizing Calcein AM/PI (Beyotime, C2015S). The staining kit was employed in accordance with the manufacturer's guidelines to assess intracellular esterase activity and cell membrane integrity using Calcein AM/PI, respectively. The staining process was conducted for 30 min, followed by imaging of the samples using immunofluorescence microscopy.

2.6. PLL cell affinity assay

PLLs were labeled with FITC fluorescence and systematically imaged utilizing CellMask™ Cell Membrane Stain (Invitrogen™, C10045) and Hoechst 33342 solution (Thermo Scientific™, 62249) in accordance with the manufacturer's instructions. The image was obtained using a live cell workstation module from Keyence (BZ-X800LE).

2.7. Immunofluorescence

The nerve induction solution was configured according to prior literature [29]. Neurogenic differentiation was induced using Neurobasal A (Gibco, USA) supplemented with B27 (Gibco, USA), 40 ng/mL FGF, 20 ng/mL EGF, 2 mM glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin (Gibco, USA). The hBMSCs were applied to the surface of PLL-PLLA electrospun membranes with varying concentrations of PLL-PLLA (0 mg/ml, 1 mg/ml, 5 mg/ml, 10 mg/ml) and the flat membrane. Flat membranes supplemented with neurogenic inducers served as the positive control. The cells were immobilized with 4% PFA at room temperature and subsequently treated with 0.1% Triton X-100 (Sigma-Aldrich) for approximately 5 min to facilitate permeabilization. The cells were subsequently blocked with QuickBlock™ immunostaining blocking solution (Beyotime, P0260) for 10 min at room temperature. The samples underwent treatment with primary antibodies TUBB3 (1:200, ab78078, Abcam), AKT (1:50, 60203-2, Proteintech), and p-AKT (1:100, 66444-1, Proteintech), followed by overnight incubation at 4 °C. After that, the samples were incubated with a secondary antibody the next day (ab150113, Abcam; goat anti-mouse IgG H&L, Alexa Fluor® 488, 1:200 dilution). The incubation occurred on a shaker at room temperature for 2 h. The samples were stained with rhodamine-labeled phalloidin (Solarbio, CA1610) and a DAPI solution (10 μg/ml, ready-to-use) (Solarbio C0065). The specimen underwent examination via fluorescence microscopy.

2.8. Real-time fluorescence quantitative PCR (RT-qPCR)

The RT-qPCR assay is based on previous literature [30]. RNA isolation was performed utilizing TRIzon Reagent (CWBIO, CW0580), followed by conversion to complementary DNA (cDNA) with the PrimeScript RT reagent Kit (TaKaRa, RR037A). The cDNA was analyzed using RT-qPCR with Hieff UNICON® Universal Blue qPCR SYBR Master Mix (Yeason, 1184ES08) for detection. The specific procedures undertaken in the experiment were as follows: 95 °C for 2 min, 1 cycle. 95 °C for 10 s, 60 °C for 30 s, repeated for 40 cycles, and concluding with 95 °C for 15 s, 60 °C for 60 s, followed by 95 °C for 15 s, 1 cycle. The primer sequences are displayed in Table S1.

2.9. Western Blot assay

Protein expression was determined through Western Blot analysis, utilizing RIPA lysis solution (Beyotime, P0013B) augmented with protease and phosphatase inhibitors. The proteins in the collected cells were quantified using the BCA Protein Quantification Kit (Thermo Scientific™, 23227). Electrophoresis was conducted, followed by the transfer of proteins onto membranes and subsequent blocking with 5% skimmed milk. Subsequently, the membranes were incubated with primary antibodies: TUBB3 (1:1000, ab78078, Abcam), NSE (1:500, A19091, ABclonal Technology), β-actin (1:500, ab8227, Abcam), p-AKT (1:1000, 4060S, Cell Signaling Technology), AKT (1:1000, 9272S, Cell Signaling Technology), p-FAK (1:1000, 381143, Zenbio), and FAK (1:1000, R24276, Zenbio) at 4 °C overnight. On the following day, the samples were incubated with goat anti-rabbit IgG-HRP (1:1000, ab205718, Abcam) for 2 h. Subsequently, the membranes underwent incubation with an NCM Biotechnology P10060 Extremely Sensitive ECL Chemiluminescence Kit for observation.

2.10. Transwell assay for neurotrophic function of PLL-PLLA electrospun membrane

The neurotrophic effect of PLL-PLLA was evaluated using the Transwell method. Progenitor cells from the dorsal root ganglia(DRG) of rats were initially isolated and positioned in the lowest compartment of a 24-well plate, adhering to the methodology outlined in previous research [31]. Stereomicroscopy facilitated the extraction of the lateral myelin sheath from isolated dorsal root ganglia obtained from rat intervertebral foramina. The extracted myelin sheath was then seeded into 24-well plates. The upper chamber of the Transwell system was populated with PLLA and PLL-PLLA nanofiber membranes. Three days later, the cells were fixed with 4% PFA at room temperature, permeabilized, and subsequently treated with the primary antibody NF-200 (1:100, 2836S, Cell Signaling Technology) at 4 °C overnight. DAPI solution (10 μg/ml) (Solarbio, C0065) was subsequently employed to stain the samples. Fluorescence microscopy was employed to analyze the morphology of the cells.

2.11. Preparation and characterization of neural guide conduits produced from PLL-PLLA using electrostatic spinning technology

First, based on our previous research, an electrospun membrane with suitable inductive properties was constructed [30]. An inner layer with aligned morphology was fabricated using electrospinning at a rotational speed of 2800 rpm from PLLA/PLL-PLLA. Subsequently, using this electrospun inner layer as a substrate, an intermediate gelatin-PLLA electrospun layer(GEL) was constructed. To ensure uniformity between the inner and outer layers of the conduit, the aluminum foil on the collector was cut longitudinally along the long axis of the mandrel and half of it was removed, after which the gelatin-PLLA electrospun layer was fabricated on the remaining part. A gelatin outer layer was then formed using the crosslinking properties between genipin(GNP) and gelatin [32]. Based on previous literature, 10% (w/v) gelatin and 0.3% (w/v) genipin can reduce porosity while achieving a moderate swelling ratio and appropriate degradation time [32]. To preserve the alignment of the inner electrospun layer, the side of the electrospun membrane that was peeled off from the aluminum foil was used as the inner layer to minimize disruption from disordered fibers. Subsequently, a water-soluble poly(vinyl alcohol) (PVA) polymer was used as an inner core, and the nerve conduit was prepared by peeling the material off the mandrel [[33], [34], [35]]. The conduit was rinsed multiple times with deionized water to remove the residual PVA. The surface morphology of the nerve conduit components, including the cross-section, outer surface, and inner surface of the conduit, was observed using scanning electron microscopy (SEM).

2.12. Establishment of rat model for sciatic nerve defect repair

The Animal Ethics Committee at Beijing Stomatological Hospital, affiliated with Capital Medical University, approved all animal-related procedures and ensured compliance with the committee's ethical standards(KQYY-202403-003). Each of the 4 groups—autograft, sham, GNP/GEL-PLLA/PLL-PLLA-rBMSCs(PLL-PLLA-rBMSCs), and separation—contained six 8-week-old SD rats. We anesthetized rats using isoflurane and a gas anesthesia machine. Subsequently, we performed an incision in the gluteal muscle to expose the sciatic nerve of each rat. The exposed sciatic nerve was rotated in the autograft group and sutured at an angle of 180°. In the sham group, the muscle and skin were sutured in layers over the exposed sciatic nerve. In the PLL-PLLA-rBMSCs group, the sciatic nerve was resected, and both ends were subsequently sutured into the nerve guidance conduit using the horizontal mattress suture technique. The sciatic nerve in the separation group was dissected and left untreated. In the week following surgery, around 10⁵ units of penicillin were administered intraperitoneally each day.

2.13. Rat gait analysis

We assessed the sciatic nerve function index (SFI), tibial nerve function index (TFI), and peroneal nerve function index (PFI) in each group of rats postoperatively utilizing the Rodent Gait Analysis System, DigiGait. Refer to the manufacturer's instructions for specific test procedures. The SFI, TFI, and PFI exhibit negative values ranging from 0 to −100. The equations for multiple linear regression analysis are computed to offer a non-invasive, quantitative evaluation of hindlimb locomotor function in rats with targeted hindlimb nerve injury. The indices represent coefficients derived from the measurement of walking trajectories in rats with nerve damage. The experiment was conducted three times for each rat.

2.14. Histological analysis

All animals were euthanized, and their primary functional organs (heart, liver, spleen, lungs, kidneys), gastrocnemius muscle, nerve conduits, and developing nerves were harvested. Fixation was initially performed through cardiac perfusion, utilizing approximately 300 ml of saline. The primary functional organs were subsequently dissected using a scalpel and fiber tools, followed by fixation of the gastrocnemius muscle samples and each group of functional organ samples in 4% PFA. Hematoxylin-eosin (HE) staining was employed to evaluate the toxicity in paraffin sections of key functional organs, including the heart, liver, spleen, lungs, and kidneys. Additionally, Masson staining and HE staining were utilized to stain muscle fibers in the gastrocnemius, with subsequent examination under a light microscope to assess the extent of muscle atrophy.

The sciatic nerves from each group designated for electron microscopic observation were fixed using 2.5% glutaraldehyde (Biosharp, BL911A) and subsequently stained with 1% toluidine blue (TB). The remaining sciatic nerves were fixed with 4% PFA, and the regeneration of the nerves within the grafted conduits was assessed through immunofluorescence analysis. Schwann cells were identified using S100β, while myelin sheaths were characterized by myelin basic protein (MBP). The primary antibody concentrations utilized were as follows: NF-200 (1:100, 2836S, Cell Signaling Technology), MBP (1:100, ab40390, Abcam), S100β (1:100, 90393S, Cell Signaling Technology), and TUBB3 (1:100, 4466S). Subsequent manipulations were conducted as previously outlined.

2.15. RNA-seq

The hBMSCs cultured on the PLL-PLLA surface for 14 days were designated as the experimental group, while those cultured on the smooth membrane for the same duration were designated as the control group. RNA-seq was conducted by Capitalbio Technology Corporation (Beijing, China). To investigate the signaling pathways involved in this process, we employed screening conditions of PLL-PLLA/Flat≥1.6 and p ≤ 0.05 for KEGG enrichment analysis to identify the enriched pathways.

2.16. Transfection of miRNA mimic, negative control for mimic, miRNA inhibitor, and negative control for inhibitor

Negative controls (NC mimics), miR-30a-5p mimics (Mijia Scientific, Beijing, China), inhibitor negative controls (NC inhibitor), and miR-30a-5p inhibitor (Mijia Scientific, Beijing, China) were utilized at a final concentration of 50 nM.

2.17. Statistical analysis

Each experiment was independently repeated at least three times (n = 3). The sample size (n) denotes the number of independent biological replicates: for in vitro experiments, each replicate came from independently prepared cell cultures or scaffold batches; for in vivo experiments, n represents the number of animals per group.

Data normality was assessed using the Shapiro–Wilk test. For two-group comparisons, the unpaired two-tailed Student's t-test was used when data met normality and homoscedasticity; otherwise, the Mann–Whitney U test was applied. For multi-group comparisons, one-way ANOVA with Tukey's post hoc test was used for normally distributed data, and the Kruskal–Wallis test with Dunn's post hoc test for non-normal data.

All analyses were performed using GraphPad Prism, with statistical significance set at p < 0.05. For in vivo experiments, animals were randomly assigned to groups using a computer-generated randomization schedule. Data collection and analysis were performed blinded to group allocation to minimize bias.

3. Results and discussion

3.1. Preparation and characterization of PLL-PLLA nanofiber membranes and nerve conduits

The utilization of nerve-guided conduits for the reconstruction of nerve injuries is a prominent area of research in peripheral nerve damage restoration [36]. To simulate the physiological structure of axons and nerve fibers, we fabricated a nerve guidance conduit with an inner layer of electrospun membrane featuring an aligned fiber configuration and an outer layer of GNP/GEL. We investigated the biomechanical differences of conduits with varying thicknesses, with a particular focus on the inductive effect of the inner layer on stem cell fate.

In this study, electrospinning technology was used to combine the topography and biological properties of materials to construct a porous three-dimensional fibrous scaffold (GNP/GEL-PLLA/PLL-PLLA) for promoting the healing and regeneration of nerve defects (Fig. 1A and B). Studies have found that using a mixture of GEL-PLA can form a three-dimensional assembled nanofiber scaffold with good biocompatibility [37]. However, the crosslinking agent glutaraldehyde used in that study exhibits high cytotoxicity. In contrast, genipin (GNP), which is mainly derived from gardenia fruit, has weaker cytotoxicity and stronger crosslinking ability, making it a suitable alternative to glutaraldehyde [38]. In this study, a GNP/GEL outer shell with a specific diameter and resilience was fabricated using the crosslinking reaction between genipin and gelatin (Fig. 1C and D). In general, it is apparent that the conduit length can be modified according to the defect size, while the diameter of the constructed nerve conduit assembly is approximately 2 mm (Fig. 1D), aligning with the diameter reported in the literature, which is suitable for rat sciatic nerve diameter [39]. Fig. 1E presents scanning electron microscopy(SEM) pictures of several sections of the nerve conduit. The findings indicate that the conduit comprises a multilayered structure, characterized by an exterior layer that is less dense, and the interior layer maintains a more uniform spinning shape. Spatiotemporal patterns of nerve conduction signals originate in intricate short-range interactions among cells and extensive long-range connections of axons [40].To imitate the physiological architecture of axons and nerve fibers, we fabricated electrospun PLLA membranes with a parallel fiber configuration and incorporated varying concentrations of PLL to improve their capacity to modulate the biological behavior of stem cells. This study aims to investigate the roles of PLLA electrospun membranes loaded with PLL in regulating stem cell function, lineage specification, and the associated mechanisms. A smooth PLLA membrane (Flat), produced via the casting procedure, served as the negative control. SEM results showed that the fiber orientation on the surface of the electrospun membranes was consistent across all groups, and there was no significant change in the diameter of the electrospun fibers with increasing PLL concentration (Fig. 1F).

Fig. 1.

Characterization and manufacturing of PLL-PLLA electrospun membranes. A) Schematic illustration of the manufacturing process of the Genipin-Gelatin-PLL/PLLA(GNP/GEL-PLLA/PLL-PLLA) nerve conduit for peripheral nerve regeneration. This picture was created in BioRender. Liu, S. (2024) BioRender.com/l31f996. B) Schematic of the cross-section of the GNP/GEL-PLLA/PLL-PLLA. C,D) Photographs of the GNP/GEL-PLLA/PLL-PLLA nerve conduit. E) SEM image of the nerve conduit, including lateral and medial views. F) Representative photographs of PLL-PLLA with different concentrations of PLLs. G) PLL release from PLL-PLLA (5 mg/ml) PLL-PLLA electrospun membranes. H) Open-circuit voltage outputs of the electrospun membranes and smooth film under the press-release stimulation with an applied force of 5N.

An ideal nerve conduit material should also possess biodegradability and should not cause burst release of small molecules from the loaded drugs [41].The three-layer nerve conduit (GNP/GEL/PLL-PLLA) we constructed is biodegradable. PLLA undergoes hydrolytic degradation to produce lactic acid, which is then eliminated through excretion in the form of water and carbon dioxide. Its degradation time is 40 weeks in vitro and approximately 30 weeks in vivo [42].To further investigate the degradation behavior of the GNP/GEL/PLL-PLLA conduit, we fabricated GNP/GEL membranes to observe the degradation of the electrospun surface. Meanwhile, we prepared GNP/GEL/PLL-PLLA conduits using PLL loaded with the fluorescent dye FITC to monitor the sustained release of PLL-FITC from the conduit. The results showed that the membrane exhibited minimal degradation within the first 7 days, after which the degraded area gradually increased until day 23, followed by stabilization (Fig. S1A and B). To further examine the surface morphology, membranes at day 0 and day 27 were air-dried and observed by SEM. The results revealed that the membrane surface exhibited a relatively dense porous structure at day 0, whereas at day 27, the surface presented a loose and highly porous morphology. Consistent with the above findings, the release of PLL-FITC from the GNP/GEL/PLL-PLLA conduit was relatively low before day 7, but gradually increased from day 11 to day 14 (Fig. S1C). To more thoroughly investigate the release characteristics of PLL from the PLL-PLLA electrospun scaffold, we immersed PLL-PLLA electrospun membranes in PBS and analyzed the concentration of Br in the immersion solution using ICP-MS. The results indicated that Br remained consistently present in the immersion solution over time, demonstrating that PLL-PLLA possesses a capacity for sustained release (Fig. 1G).

An ideal nerve guidance conduit material should also provide a nerve-mimicking biomechanical environment. Therefore, we fabricated GNP/GEL/PLL-PLLA conduits with different thicknesses and examined their biomechanical properties. The overall conduit thickness, thickness of each layer, porosity, and inner layer fiber diameter are shown in Fig. S1D. SEM observation of the layer-specific thickness of the nerve conduits revealed that while the inner layer (PLL-PLLA) thickness showed no significant difference among groups, and the outer layer (GNP/GEL-PLLA) thickness increased only slightly with total thickness, the middle layer (GEL-PLLA) thickness increased significantly with total conduit thickness, findings that align with the intended design of our nerve conduits. Consistent with other studies, the fiber diameter in this study was relatively uniform, which may be attributed to the low concentration of the spinning solution [43]. Our results revealed that the 200 μm conduit exhibited the highest porosity among the three thickness groups, and porosity values decreased progressively with increasing total conduit thickness (Fig. S1D). This is consistent with the finding that the 10% gelatin used reduces porosity [32]. Additionally, a universal testing machine was used to evaluate the tensile strength (dry/wet), compressive strength and elastic modulus of conduits with different thicknesses, using rat sciatic nerve as a control. The results showed that under dry conditions, the tensile strength of the 400 μm-thick GNP/GEL/PLL-PLLA conduit was comparable to that of the sciatic nerve, whereas the tensile strength of the 1000-μm thick conduit, the thickest among the groups, was lower than that of the 400 μm conduit (Fig. S1E and F). We speculate that this result may be related to uneven fiber packing and the complex accumulation and distribution of internal stresses in the thicker group [25]. Under wet conditions, the mechanical properties of all conduit groups slightly decreased, which is associated with the gelatin content. Furthermore, compressive strength testing revealed that the compressive resistance of the conduits increased with increasing thickness. In summary, the GNP/GEL/PLL-PLLA conduit with a thickness of 400 μm was selected for subsequent experiments in this study.

In addition, electrical stimulation plays a crucial role in facilitating tissue regeneration, as piezoelectric materials may reduce the necessity for open surgery and enhance damage recovery and stem cell differentiation via mechanoelectrical response [44]. PLLA possesses piezoelectric properties, and common approaches for fabricating piezoelectric PLLA materials include melt spinning, thermal annealing, and material blending [21,45]. Among these, melt spinning induces orientation of PLLA molecular chains along the fiber axis and increases the proportion of the crystallinity and chain orientation by controlling parameters such as draw ratio and cooling rate, though it offers limited control over fiber diameter (micrometer scale) [46]. Thermal annealing promotes the transformation of the amorphous phase of PLLA into a more ordered crystalline structure, but the process is complex [47]. In contrast, during the fabrication of PLLA nanofibers, electrospinning technology not only significantly promotes molecular chain orientation due to high tensile stress and rapid solvent evaporation, thereby enhancing its piezoelectric effect, but also macroscopically aligns the fibers to provide contact guidance [21,24]. Reducing the solution concentration and increasing the thickness of the nanofiber membrane can further enhance the piezoelectric performance of the membrane [48,49]. Additionally, studies have shown that the crystallinity of PLLA is closely related to its piezoelectric properties; an increase in crystallinity from 34.8% to 57.4% resulted in a substantial enhancement of the VPFM amplitude signal [47]. This study designed an assembly for measuring the piezoelectric capabilities of various groups of electrospun membrane structures. The assembly's composition is illustrated in the right figure of Fig. 1 H. Copper foils were affixed to both sides of the tested membranes as electrodes, followed by the application of PU films on both sides of the copper foils. A force of 5N was exerted on one side of the assembly via a linear motor and quantified using a dynamometer. The results showed that the electrospun PLLA/PLL-PLLA membrane with aligned morphology exhibited superior piezoelectric properties compared to the smooth group, which is attributed to the fabrication method and the crystallization characteristics of PLLA [50]. The piezoelectric performance of our fabricated aligned electrospun membrane was lower than the 7.9 V of PLLA/ZnO aligned fibers, which may be attributed to differences in material composition and spinning solution concentration (Fig. 1H).

The inner layer of the nerve conduit exerts the strongest contact guidance effect on stem cells. Therefore, to simulate the in vivo environment of animals, this study tested the piezoelectric generation of the PLL-PLLA membrane under cyclic loading, bending, and torsion. The results showed that the PLL-PLLA membrane exhibits piezoelectric properties under various mechanical loading conditions (Fig. S1G). To further investigate the stimulation perceived by cells under resting conditions, we measured the open-circuit voltage generated by the PLL-PLLA membrane under fixed loading conditions (PLL-PLLA) and non-loading (piezoelectric shielding) conditions. The results (Fig. S1H) showed that the PLL-PLLA membrane possesses favorable piezoelectric capability and can generate a relatively weak open-circuit voltage within the cap of a centrifuge tube.

3.2. Biosafety evaluation of PLL-PLLA electrospun membrane and GNP/GEL-PLLA/PLL-PLLA conduits

The microenvironment of tissues and biomaterials significantly influences cellular behavior, including cell morphology, orientation, migration, and subsequent tissue development [51]. To assess the biocompatibility of this scaffold, we performed Calcein AM/PI staining and CCK-8 assays. Cells cultured in normal culture plates were used as a negative control, and the DMSO group was used as a positive control. The results indicated that the cellular activity of all groups was not significantly inhibited by the cells induced by electrospun membrane culture medium (Fig. S2A). Concurrently, hBMSCs were directly cultured on the surface of the electrospun membrane within all groups, and the quantity of red fluorescent cells (PI staining) was minimal, indicating that electrospun membranes did not significantly impact cell viability in every group (Fig. S2B). The CCK-8 results indicated that cell proliferation could be stimulated by PLL concentrations of 0.5 mg/ml, 1 mg/ml, 2.5 mg/ml, and 5 mg/ml, with the exception of the 10 mg/ml group. Nevertheless, the promotion of cell proliferation was substantially diminished by the higher concentration of PLL (10 mg/ml) (Fig. S2C).

To further clarify the safety of the degradation products of the GNP/GEL/PLL-PLLA conduit and their impact on surrounding tissues, supernatants were collected at 0, 3, 7, 14, and 21 days, and their effects on the viability of RSC96s were investigated. The results of CCK-8 and Calcein-AM/PI assays showed that the degradation products within 21 days had no significant inhibitory effect on the viability of RSC96s (Fig. S2D and F). Meanwhile, the results of in vivo implantation in mice showed no obvious inflammatory tissue encapsulation within 21 days, with only slight connective tissue adhesion (Fig. S2E). In addition, HE staining of the visceral organs (heart, liver, spleen, lung, and kidney) collected after in vivo implantation of the material showed no obvious inflammatory or damage responses, confirming the satisfactory in vivo biocompatibility of the material (Fig. S2G).

The proliferation of cells on the surface of each category for membrane materials can be illustrated through phalloidin staining. The results indicated that the cells expanded in parallel on the electrospun membrane, and the growth direction was more uniform. Initially, the hBMSCs that were induced by flat membrane exhibited a variety of morphologies (Fig. 2A). Cell elongation can regulate directed growth, gene expression, lineage differentiation, which could influence tissue regeneration and biochemical feedback [52].This study reveals that parallel topography triggers stem cells to elongate and is compatible. The increased intercellular spacing on the spinning surface of the parallel topography (Fig. 2A–C) promotes more isolated growth compared to the flat group, which exhibits a random growth pattern. In line with our findings, cell membrane potential ranges from −10 to −90 mV, therefore, positively charged PLL can promote cell adhesion, proliferation, and spreading [24]. This serves as a physically modified model for investigating cell patterns and enhancing the understanding of localized single-cell growth (Fig. 2D).

Fig. 2.

Bioactivity of human bone marrow-derived stem cells (hBMSCs) on PLL-PLLA, PLLA electrospun membranes and smooth membrane surfaces(Flat). A)-C) Phalloidin staining of hBMSCs on the membrane surfaces of PLL-PLLA(5 mg/ml), PLLA electrospun membranes and smooth membrane surfaces. D) Schematic representation of the morphology of cells growing on each group of membranes(Flat). This picture was created in BioRender. Liu, S. (2024) BioRender.com/l43i056. E) Labeling PLL with FITC and measuring its affinity for hBMSCs using a confocal live cell workstation.∗p< 0.05, ∗∗p< 0.01, ∗∗∗p < 0.001.

3.3. Effect of PLL-PLLA membrane on BMSC migration and nuclear entry

We established a Transwell co-culture system by inoculating hBMSCs in the upper chamber and placing parallel-shaped PLLA or PLL-PLLA in the lower chamber. The number of cells crossing the chamber was monitored at 24 h using Hoechst staining, and the results indicated that the PLL-PLLA group had a significantly higher number of cells than the PLLA group (Fig. S3A and B). This suggests that the PLL-PLLA electrospun membrane has the potential to promote the migration of hBMSCs. In this investigation, PLL was infused with a FITC fluorescent moiety (PLL-FITC) to examine the affinity between PLL and cells and its localization within cells. Initially, the co-localization of PLL-FITC with cells was examined utilizing a confocal live cell workstation. The results indicated that PLL-FITC progressively infiltrated the cells within approximately 5 min and then concentrated in the nucleus over time (25 min) (Fig. S3C). This demonstrated that PLL-FITC was able to penetrate the cells and gradually accumulate in the nucleus, potentially correlating with the directional influence of the PLL-PLLA electrostatic spinning membrane on the fate differentiation spectrum of hBMSCs. To exclude the impact of cell species on the results, we isolated rat bone marrow mesenchymal stem cells (rBMSCs) from 4-week-old rats and characterized them using flow cytometry (Fig. S3D). Following the culture of rBMSCs with PLL-FITC for 24 and 48 h, confocal imaging revealed a progressive accumulation of PLL-FITC within the cells (Fig. 2E), in conjunction with a marked rise in fluorescence intensity, suggesting that PLL exhibits a robust affinity for cells.

3.4. Effect of PLL-PLLA membrane on regulating neurogenic differentiation of BMSCs

To identify the optimal concentration of PLL in PLL-PLLA electrospun membranes, we seeded hBMSCs onto PLL-PLLA membrane with different PLL concentrations and assessed the expression of neural-related genes (TUBB3, MAP2, NeuroD1, and NSE). Flat + neurogenesis is a neural induction medium supplemented with EGF, bFGF, B27 that facilitate the differentiation of stem cells into neuron-like cells [53]. The results of this research (Fig. 3A and B) indicated that the expression of TUBB3 and MAP2 increased substantially and even exceeded that of Flat + neurogenesis group at PLL concentrations of 5 mg/ml and 10 mg/ml. MAP2 are markers of dendrites and myelin sheaths, respectively, and they indicate the formation of neuron-like cells [53]. NeuroD1 functions as a transcription factor that facilitates the direct neuronal transformation of microglia [54].NSE inhibits neuronal damage, contributes to neuronal cell lineage-specific differentiation, and is essential for neuronal growth and function [11]. Despite finding that 10 mg/ml of PLL-PLLA did not result in a significant increase in NeuroD1 and NSE expression in hBMSCs, it was still higher than that of the control group (Flat). These results indicate that PLL-PLLA has the potential to facilitate the neural differentiation of hBMSCs. To further investigate the neuronal differentiation of hBMSCs induced by PLL-PLLA over time, cells were inoculated onto the surface of PLL-PLLA electrospun membranes for 7, 14, and 21 days for RT-qPCR. The findings indicated that PLL inclusion at all time intervals enhanced the expression of neurulation-related genes TUBB3 and MAP2, while TUBB3 expression peaked at 14 days (Fig. 3C).

Fig. 3.

Concentration-responsive effects of PLL-PLLA electrospun membranes on neural differentiation of hBMSCs. A,B) mRNA expression levels of neurogenic markers in hBMSCs cultured on the surface of PLL-PLLA electrospun membranes at different concentrations for 14d. C) mRNA expression levels of neurogenic markers in hBMSCs cultured on the surface of PLL-PLLA electrospun membranes at different time points (7d,14d,21d). D) Immunofluorescence was used to detect TUBB3 (green), F-actin (red) and nuclear DNA (DAPI, blue) in hBMSCs. (Scale bar:20 μm). E) Protein expression levels of neurogenic markers in hBMSCs cultured on the surface of PLL-PLLA electrospun membranes at different concentrations for 14d. F) Immunofluorescence detection of DRG axon stretching changes in smooth and PLL-PLLA electrospun membranes (Scale bar:100 μm).ns = not significant, ∗p< 0.05, ∗∗p< 0.01, ∗∗∗p < 0.001. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

TUBB3 is expressed in youthful neuronal cell bodies, dendrites, axons, and axon terminals, indicating the characteristic shape of dendrites and axons in neural cells [55]. Consequently, this research utilized immunofluorescence to explore cell shape and the expression of the neural-associated protein TUBB3. Immunofluorescence results indicated that hBMSCs on the smooth membrane surface were sparsely dispersed with a polygonal shape, and the fluorescence intensity of TUBB3 was inadequate. The elongated spindle-shaped morphology of cells grown on the surface of PLL-PLLA membranes was aligned along the longitudinal axis of the fibers. The fluorescence intensity of TUBB3 was significantly higher on the PLL-PLLA surface than that of the Flat group, and only slightly lower than that of the positive group (Flat + neurogenesis) (Fig. 3D). Furthermore, Western blot analysis revealed a substantial increase in the expression levels of the neural-associated proteins TUBB3 and NSE after 14 days following neural induction (Fig. 3E). Following the occurrence of acute peripheral nerve damage, neurotrophic factors are released from nerve terminals, subsequently activating the healing mechanism to promote synapse formation and neuronal differentiation [56]. To further improve the simulation of this biological process, we employed the Transwell system to position PLLA and PLL-PLLA membranes in the top chamber, while inoculating the bottom chamber with dorsal root ganglion cells (DRG) sourced from isolated rat dorsal root ganglia. DRG in the PLLA group exhibited a more disorganized growth pattern and displayed a polygonal morphology, whereas the cell morphology in the PLL-PLLA electrospun group was elongated, and the axon length was significantly greater compared to the PLLA group (Fig. 3F).

To preliminarily explore the relationship between the mechanism of PLL-PLLA membrane-induced stem cell fate and piezoelectricity, we established a non-loading group (electrically shielded group) and a normally cultured PLL-PLLA group (piezoelectric group). After 14 days of culture in the two setups, immunofluorescence results showed that cells in both the electrically shielded group and the PLL-PLLA group elongated along the fiber direction, but TUBB3 expression was significantly lower in the electrically shielded group than in the PLL-PLLA group(Fig. S3E). RT-qPCR results also showed that the expression levels of neurogenic-related genes MAP2, MBP, and NSE in the PLL-PLLA group were significantly higher than those in the electrically shielded group (Fig. S3F). This suggests that the PLL-PLLA membrane promotes neuron-like differentiation of stem cells through the synergistic effects of piezoelectricity, chemical induction, and topographical cues. In addition to piezoelectric properties, the aligned surface morphology also plays an important role in inducing cell differentiation. Aligned microgrooves or nanofibers provide contact guidance, causing cells to orient along the direction of the topography [23],while simultaneously regulating the differentiation fate of stem cells through pathways involving mechanical stress [57]. These findings are consistent with our results, in which BMSCs cultured on the PLL-PLLA electrospun surface exhibited an elongated morphology, along with increased expression of neuron-like differentiation-related genes and proteins. However, aligned topography alone exhibited a weak capacity to induce stem cell differentiation (Fig. S3E and F), and effective induction of neuron-like differentiation from stem cells required combination with piezoelectric stimulation.

3.5. Restorative outcomes following sciatic nerve defect in rats

To investigate the nerve regeneration capacity, we established a rat model of sciatic nerve damage, illustrated in Fig. 4A. The autologous nerve transplantation group served as a positive control for the evaluation of nerve regeneration. The Sciatic Functional Index (SFI) is a quantitative instrument for evaluating motor function following sciatic nerve damage by examining toe positioning to determine the functionality of muscles innervated by the sciatic nerve [58]. The gait analysis results indicated that the SFI, tibial functional index (TFI), and peroneal functional index (PFI) of the rats in the PLL-PLLA-rBMSCs group(GNP/GEL-PLLA/PLL-PLLA-rBMSCs) experienced a substantial recovery. However, the recovery was not as significant as that of the sham group. It is important to mention that the plantar contact area of the rats in the PLL-PLLA-rBMSCs group was significantly restored, surpassing that of the separation group and even approaching that of the sham group (Fig. 4B). The above functional results indicate that the nerve guidance conduit constructed in this study significantly promotes functional recovery after sciatic nerve injury in rats. This material exhibits electroactivity, topographical guidance, and immunomodulatory effects. Studies have shown that different electrical signals (0.07875 V, 1.69 V), oriented growth guidance for injured nerve axons, and various inductive factors can effectively promote sciatic nerve regeneration after injury [13,59]. This fully demonstrates that the approach to promoting nerve regeneration is not limited to a single fixed standard but rather operates within a certain therapeutic threshold. Specifically,the piezoelectric output generated by mechanical deformation during normal locomotion in rats is not a single fixed value but gradually fluctuates. This may correspond to the reduced motor activity in the early stage of injury (resulting in weaker piezoelectric effects) and the gradual increase in motor activity in the later stage (resulting in stronger piezoelectric effects).

Fig. 4.

Evaluation of in vivo sciatic nerve regeneration. A) Schematic diagram of the construction of animal model of sciatic nerve defects in rats and photographs of implantation of Sham, PLL-PLLA-rBMSCs, autograft groups. B) Functional reconstruction-related indices (SFI,TFI, PFI, Paw area variability) were calculated based on walking footprint analysis in rats. C) Schematic diagram of each group of injured and regenerated sciatic nerves examined by electron microscopy. This picture was Created in BioRender. Liu, S. (2022) BioRender.com/v90f693. D) Toluidine blue(TB) staining of regenerated sciatic nerve. (Scale bar:50 μm). E) Transmission electron microscopy (TEM) images of the regenerated sciatic nerve. F) Statistical analyses of the area. (G) Statistical analyses of the thickness of myelin sheath. H) Representative images of the gastrocnemius muscle in different scaffold and autologous nerve graft groups in vivo. I) Hematoxylin eosin (HE) and Masson's trichrome staining of gastrocnemius muscle from each group (Scale bar:100 μm). J) Statistical analyses of weight ratios of gastrocnemius muscle from different groups. K) Statistical analyses of myofiber diameters. ns = not significant, ∗p< 0.05, ∗∗p< 0.01, ∗∗∗p < 0.001. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Toluidine blue staining (TB staining) and transmission electron microscopy (TEM) examination of peripheral nerve sections represent the gold standard for morphological assessment of peripheral nerve cross-sections, with commonly utilized evaluation metrics encompassing myelin roundness, myelin thickness, myelin density, and myelin area(Fig. 4C). TB staining indicated that the majority of the regenerated nerves exhibited proper organization, devoid of severe scarring or adipose tissue (Fig. 4D). The density of myelinated axons and the area of myelin were greatly restored in the PLL-PLLA-rBMSCs group, surpassing those in the autograft group (Fig. 4E–G). This finding demonstrates the efficacy of GNP/GEL/PLL-PLLA nerve conduit in nerve regeneration. The sciatic nerve innervates the gastrocnemius muscle, and damage to this nerve is likely to result in denervation atrophy of the muscle [31]. The gastrocnemius muscle weight in the PLL-PLLA-rBMSCs group was much greater than that in the separation group, indicating that GNP/GEL/PLL-PLLA treatment inhibited gastrocnemius atrophy (Fig. 4H). Furthermore, the weight ratio of the gastrocnemius muscle in the PLL-PLLA-rBMSCs group exceeded that of the separation group (Fig. 4J). In addition, HE and Masson staining results indicated that the average diameter of muscle fibers in the PLL-PLLA group was greater than that in the separation group (Fig. 4I–K).

After axonal injury, distal remnant axons lose contact with schwann cells, leading to demyelination. Meanwhile, schwann cells (both myelinating and nonmyelinating) undergo reprogramming into a repair phenotype, thereby promoting and directing axonal repair [60]. Upon regeneration following nerve damage, neurons transition from a signaling mode to a growth mode, progressively expressing proteins associated with growth to facilitate axonal regeneration (Fig. 5A). To investigate axonal and myelin regeneration capabilities of nerves following GNP/GEL-PLLA/PLL-PLLA nerve conduit-induced defect, we assessed the expression of neural-specific proteins S100β, NF-200, MBP, and TUBB3 in both transverse and longitudinal sections. TUBB3 is a neuronal marker that is used to determine axonal recovery, whereas myelin basic protein (MBP) is a marker for myelin formation [61]. S100 β is a cell-specific marker for schwann cells [62]. Neurofilament-200 (NF-200) is a marker exclusive to axons that has a positive correlation with axonal content [31,63]. The immunofluorescence results indicated that the expression levels of S100β, NF-200, MBP, and TUBB3 in the PLL-PLLA-rBMSCs group were markedly elevated compared to the separation group and the autograft group, although they remained weaker than those in the sham group (Fig. 5B–E). However, rodents and humans exhibit significant differences in peripheral nerve regeneration rates and tissue remodeling mechanisms. Therefore, although this study found that the PLL-PLLA membrane effectively promotes nerve recovery in a rat model of sciatic nerve defect, validation in larger animal models is still required prior to clinical translation. While the present study demonstrates the promising regenerative efficacy of the GNP/GEL-PLLA/PLL-PLLA conduit in a rat sciatic nerve injury model, several limitations should be acknowledged. In particular, although we observed no gross neuroma-like pathology in the subcutaneous implantation model and favorable functional recovery without overt aberrant growth in the nerve defect model, quantitative histomorphometric analysis of ectopic nerve growth was not performed in this study. Such quantitative assessment, including the measurement of ectopic axonal area, fiber counts outside the epineurial boundary, and serial section analysis to exclude neuroma-like reconnections, would provide a more rigorous evaluation of the conduit's specificity in guiding targeted axonal regeneration and preventing undesirable side branching. The absence of these quantitative data represents a limitation of the current work. Future studies incorporating these quantitative analyses, particularly in larger animal models with extended observation periods, will be essential to further validate the safety and specificity of this conduit for clinical translation.

Fig. 5.

Immunofluorescence analysis of the sciatic nerve. A) Schematic diagram of fluorescent sections for repair of sciatic nerve defects in rats.Sciatic nerve horizontal section of MBP/TUBB3(B). (This picture was created in BioRender. Liu, S. (2024) BioRender.com/q31i388.) and NF-200/S100β (C) by immunofluorescence staining. In vertical section, Sciatic nerve was detected by immunofluorescence staining, including MBP/TUBB3 (D) and NF-200/S100β (E). TUBB3 (green) / MBP (red) and S100β (green) / NF-200 (red). (Scale bar: 50 μm). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

3.6. PLL-PLLA membrane exerts its neuroregenerative effect through activation of the PI3K/AKT signaling pathway

To investigate the molecular mechanism of PLL-PLLA-induced neural differentiation of hBMSCs, we cultivated the cells on the surface of PLL-PLLA membranes for 14 days. Cells from the flat membranes and PLL-PLLA membranes were harvested for transcriptome investigation by RNA-seq. The heatmap findings indicated well-defined intra-group clustering along with enhanced inter-group separation. A total of 431 genes were up-regulated (red) and 408 genes were down-regulated (blue), as illustrated in Fig. 6A. Following this, the differentially expressed genes associated with brain development in the RNA-seq data were verified via RT-qPCR. The results revealed a substantial increase in the expression of neural-related genes (INA, NOG, PACSIN3, BEX-1) (Fig. 6B). All these findings indicated that the PLL-PLLA membranes might enhance the neuronal development of hBMSCs.

Fig. 6.

Activation of PI3K/AKT signaling pathway by PLL-PLLA electrospun membranes promotes nerve regeneration. A) Heat maps of up-regulated and down-regulated genes in hBMSCs after 14 days of incubation on smooth membrane surface and PLL-PLLA electrospun membrane surface. B) RT-qPCR validation of significantly differentially expressed neurogenic genes in RNA-Seq heatmaps. C) KEGG enrichment analysis bubble diagram. D) Heatmap showing up-regulation of genes promoted and down-regulation of genes negatively regulated by the PI3K/AKT signaling pathway. E) RT-qPCR validation of differentially expressed genes in the PI3K/AKT signaling pathway. F) Expression levels of PI3K/AKT signaling pathway-related proteins on smooth membrane surface and PLL-PLLA electrospun membrane surfaces. ns = not significant, ∗p< 0.05, ∗∗p< 0.01, ∗∗∗p < 0.001.

The KEGG enrichment analysis results indicated that genes associated with the PI3K/AKT pathway were more prevalent among the differential genes, with the differences being statistically significant (Fig. 6C). The PI3K/AKT pathway is crucial for sustaining fundamental physiological functions, including cell growth, proliferation, apoptosis, and metabolism. Phosphatase and tensin homologs (PTEN), the primary antagonists of PI3K, exert negative regulation on this system [64].The heatmap revealed a substantial increase in the expression of activation genes associated with the PI3K/AKT pathway (AKT1, AKT2), but PTEN expression was reduced (Fig. 6D). Furthermore, the RT-qPCR results indicated that the expression of the related positively activated genes IGF2 and VEGFA was increased, while the expression of the negative regulators OSMR and PTEN was decreased (Fig. 6E). Activation of PI3K protein is a necessary condition for axonal regeneration [65,66]. FAK serves as a molecular hub that links integrins with downstream signaling molecules within the integrin signaling cascade [11]. Focal adhesion kinase (FAK, also known as PTK2), integrins, and talin co-localize and assemble to create adhesion patches that transduce topological cues into biochemical signals, influencing cell fate differentiation patterns [31]. FAK is a critical protein that is involved in the activation of the PI3K/AKT pathway [67]. The Western blot results indicated an increase in the expression of phosphorylated focal adhesion kinase (p-FAK) and p-AKT in the PLL-PLLA group, while FAK expression remained relatively stable and AKT expression was down-regulated; furthermore, the expression of the negative regulator PTEN was diminished (Fig. 6F). The results validate the activation of the PI3K/AKT signaling pathway during PLL-PLLA-induced neural development of hBMSCs. Furthermore, PLL is frequently employed as a chain segment motif to augment stability and hydrophilicity [68].Our investigation demonstrated that PLL successfully penetrated the nucleus within a brief duration of 25 min (Fig. 2E), indicating that PLL may function as an intracellular transport carrier to modulate intracellular signaling.

3.7. miR-30a-5p downregulation in the PLL-PLLA membrane group regulates the PI3K/AKT pathway and neuronal expression

MicroRNAs (miRNAs) are prevalent in astrocytes and neurons, and the suppression of miR-30a-5p mitigates motor impairments and pathological alterations in Parkinson's animals [69]. Simultaneously, miR-30a-5p can mitigate the advancement of traumatic brain damage by suppressing microglial inflammation and apoptosis [70].Likewise, miR-30a-5p can mitigate LPS-induced neuronal death by regulating NeuroD1, hence influencing pro-inflammatory cytokines, reactive oxygen species, and other factors [71]. The miRNA microarray sequencing results of the study indicated a reduced expression of miR-30a-5p on the surface of PLLA electrospun membranes exhibiting parallel shape. Additionally, PLL exhibits superior cell affinity performance (Fig. 2E, Fig. S3C). In this investigation, the expression of miR-30a-5p on the surface of PLL-PLLA membranes was assessed by RT-qPCR, revealing a decrease in miR-30a-5p expression in the PLL-PLLA electrospun group (Fig. 7A). To further investigate the correlation between miR-30a-5p and PLL-PLLA-induced neuronal differentiation of hBMSCs, we acquired miR-30a-5p mimics and miR-30a-5p inhibitors using transfection. The transfection efficiency was assessed by RT-qPCR (Fig. 7B). The EdU results for each group of hBMSCs indicated that the suppression of miR-30a-5p markedly enhanced the proliferation of hBMSCs (Fig. 7C). Furthermore, RT-qPCR results indicated that the suppression of miR-30a-5p markedly increased the expression of neural-associated genes NSE and TUBB3 (Fig. 7D and E). Consequently, miR-30a-5p mimics induced a significant reduction in the expression of the neural-associated gene NSE (Fig. 7D).Western Blot analysis revealed increased expression of miR-30a-5p inhibitory neural-associated proteins (NSE, TUBB3, NF-200), indicating a correlation between miR-30a-5p and neural development of cells. Notably, Western Blot analysis revealed that the overexpression of miR-30a-5p diminished the levels of p-PI3K and p-AKT, whereas the levels of PI3K and AKT remained relatively constant; conversely, the inhibition of miR-30a-5p resulted in an increase in p-PI3K and p-AKT levels, accompanied by a decrease in AKT expression (Fig. 7F).

Fig. 7.

Downregulation of miR-30a-5p activates the PI3K/AKT pathway and is associated with enhanced adult neural differentiation. A) miR-30a-5p expression levels on smooth membrane surface and PLL-PLLA electrospun membrane surface. B) Validation of the efficiency of transfection of hBMSCs. C) EdU assay of miR-30a-5p mimics, miR-30a-5p inhibitor and its control on hBMSCs activity. Effects of miR-30a-5p mimics, miR-30a-5p inhibitor and its control pair on the expression of mRNA levels of neural-related markers, NSE(D),TUBB3(E) in hBMSCs. F) Effects of miR-30a-5p mimics, miR-30a-5p inhibitor and its control on the protein levels of PI3K/AKT signaling pathway-related markers as well as neurogenesis-related markers in hBMSCs. G) Schematic illustration of the down-regulation of miR-30a-5p to activate the PI3K/AKT pathway.This picture was created in BioRender. Liu, S. (2024) BioRender.com/t60d807. H) Immunofluorescence assay for the effect of miR-30a-5p inhibitor and its control on the expression of neural-related markers (TUBB3) and markers related to the PI3K/AKT signaling pathway in hBMSCs. AKT, p-AKT, TUBB3 (green),F-actin (red),DAPI (blue). (Scale bar:200 μm). I,J) Quantitative analysis of the expression of markers related to TUBB3 and PI3K/AKT signaling pathway. ns = not significant, ∗p< 0.05, ∗∗p< 0.01, ∗∗∗p < 0.001. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

The current work posits that the parallel-shaped PLL-PLLA membrane may facilitate neural differentiation of stem cells by stimulating the PI3K/AKT signaling pathway via the suppression of miR-30a-5p expression (Fig. 7G). The findings aligned with the immunofluorescence data (Fig. 7H–J, Figure S4), indicating that the miR-30a-5p inhibitor may facilitate neuronal development of hBMSCs via activating the PI3K/AKT signaling pathway.

In a related study, the expression of neurulation-related markers was significantly reduced as a consequence of the LY294002 treatment of ADSCs in rGO/PCL conduits [31]. This study employed the PI3K inhibitor LY294002 to inhibit the PI3K/AKT signaling pathway and to observe the changes in the phenotypes related to the proliferation and neuronal differentiation of hBMSCs in order to further investigate the role of the PI3K/AKT signaling pathway in PLL-PLLA membrane-induced neural differentiation of hBMSCs. The proliferative ability of hBMSCs decreased gradually as the concentration of the PI3K inhibitor LY294002 increased, and the cellular activity of hBMSCs decreased to a certain extent (Fig. S5A–C). These findings were found in the EdU experimental results. The fluorescence intensity of p-AKT/AKT showed a progressive decrease as the concentration of LY294002 increased in the immunofluorescence results of AKT and p-AKT (Fig. 8A and B). We selected 20 μM LY294002 for subsequent experiments in conjunction with the CCK-8 results(Fig. S5C). The addition of LY294002 resulted in a decrease in the expression of neural-associated transcripts (NSE, TUBB3) (Fig. 8C and D). Consequently, this investigation hypothesized that the PI3K/AKT signaling pathway was activated to induce neuronal differentiation of hBMSCs through the use of PLL-PLLA electrospun membranes (Fig. 8D). Additionally, the immunofluorescence results of TUBB3 demonstrated a substantial decrease in TUBB3 expression following the addition of LY294002. The fluorescence intensity of hBMSCs cells induced by the nerve-forming inducer was significantly higher than that of the control group, and the fluorescence intensity of TUBB3 was significantly reduced following the addition of LY294002 to the nerve-forming inducer (Fig. 8E). These findings confirm that activation of the PI3K/AKT pathway mediates the neural repair function of PLL-PLLA electrospun fibers.

Fig. 8.

Inhibition of PI3K suppresses neural differentiation of hBMSCs. A) Effect of different concentrations of LY294002 on the expression of hBMSCs PI3K/AKT signaling pathway-related markers by immunofluorescence detection. AKT, p-AKT (green),F-actin (red), DAPI (blue). (Scale bar:200 μm). B) Quantitative analysis of the expression of markers related to PI3K/AKT signaling pathway. C) Effect of PI3K inhibitor LY294002 on the expression of mRNA levels of nerve-related markers (NSE, TUBB3) in hBMSCs. D) Activation of PI3K pathway by PLL-PLLA electrospun membranes promotes neural differentiation of BMSCs. This picture was created in BioRender. Liu, S. (2024) BioRender.com/e58z971. E) Immunofluorescence detection of the effect of LY294002 on the expression of nerve-related marker (TUBB3). TUBB3 (green), F-actin (red), DAPI (blue). (Scale bar: 200 μm).ns = not significant, ∗p< 0.05, ∗∗p< 0.01, ∗∗∗p < 0.001. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

4. Conclusion

This research utilized electrospun technology to produce the PLL-PLLA nanofibrous membrane, which functioned as the inner layer of the three-layer neural cannula structure. The neurogenic differentiation of hBMSCs was enhanced by the parallel topography of the nanofibrous membrane, which displayed superior piezoelectric properties, as evidenced by SEM and electrochemical instrument results. The genipin-gelatin composite outer layer exhibits good biocompatibility. The rat sciatic nerve disarticulation defect can be effectively repaired and axon growth can be directed by this nerve conduit. Furthermore, the expression of miR-30a-5p decreased following the induction of PLL-PLLA electrospun membranes, which was associated with the activation of the PI3K/AKT signaling pathway. Knockdown of miR-30a-5p activates the PI3K/AKT signaling pathway and enhances the expression of neural-related genes and proteins. Our results indicate that PLL-PLLA electrospun membranes can inhibit the expression of miR-30a-5p, thereby activating the PI3K/AKT signaling pathway to enhance neural differentiation of stem cells. This research is designed to to enhance understanding of multidimensional induced biomaterials for peripheral nerve defect repair and to identify new targets for regeneration and repair.

Consent for publication

All the authors agree to the publication of the article.

Ethics approval and consent to participate

All protocols involving the use of animals followed the ethical guidelines of the Animal Ethics Committee of Capital Medical University and were approved by the Animal Ethics Committee of Beijing Stomatological Hospital affiliated with Capital Medical University (KQYY-202403-003).

Funding

This study was Supported by Beijing Natural Science Foundation (7242063), Young Scientist Program of Beijing Stomatological Hospital, Capital Medical University (YSP202411), R&D Program of Beijing Municipal Education Commission (KM202110025010), Beijing Natural Science Foundation (L242154), Beijing Hospitals Authority Innovation Studio of Young Staff Funding Support (202133), Beijing Municipal Science & Technology Commission (Z211100002921066), Innovation Research Team Project of Beijing Stomatological Hospital, Capital Medical University (CXTD202204).

CRediT authorship contribution statement

Sirui Liu: Conceptualization, Funding acquisition, Investigation, Methodology, Validation, Writing – original draft. Lin Song: Conceptualization, Investigation, Methodology, Visualization, Writing – original draft. Xueyan Zhang: Formal analysis, Investigation, Methodology. Weishu Dai: Formal analysis, Investigation, Methodology. Qiang Sun: Formal analysis, Investigation, Methodology. Shuai Zhang: Funding acquisition, Investigation, Methodology, Software, Visualization. Xiu Liu: Funding acquisition, Investigation, Methodology, Software, Visualization. Hongyu Li: Investigation, Methodology, Software. Chao Liang: Investigation, Methodology, Software. Congchong Yang: Investigation, Methodology, Software. Mengdie Liu: Investigation, Methodology, Software. Huijing Zhang: Investigation, Methodology, Software. Shengnan Wang: Investigation, Methodology. Luling Li: Investigation, Methodology, Software. Ying Huang: Conceptualization, Project administration, Writing – review & editing. Ying Hu: Conceptualization, Project administration, Writing – review & editing. Yan Lv: Conceptualization, Funding acquisition, Project administration, Writing – review & editing.

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.

Supporting information

Supporting Information is available from the author.

Appendix A. Supplementary data

The following is the Supplementary data to this article.

Data availability

Data will be made available on request.