Stepwise Regulation of Cellular Oxidative Stress via Conductive‐Piezoelectric Integrated Microstructured Conduits for Enhanced Nerve Regeneration

Abstract

Peripheral nerve injury (PNI) poses significant challenges due to the complex structure and regenerative microenvironment of peripheral nerves, which limit self‐repair capabilities. Artificial nerve conduits have been widely used for nerve repair. Here, a conductive‐piezoelectric integrated microstructured conduit is designed, using poly(lactic glycolic acid) (PLGA) and poly(vinylidene fluoride) (PVDF) via electrostatic spinning to obtain an implantable, biodegradable piezoelectric nanofibrous membrane. This membrane is further enhanced with a reduced graphene oxide/methacrylated gelatin (rGO/GelMA) gel, which synergistically promotes peripheral nerve repair. In vitro assessments reveal that the microgroove surface pattern of the conduit effectively stimulated the directional migration of cells. Moreover, using a rat sciatic nerve injury model, rGO is demonstrated to significantly modulate cellular oxidative stress, thereby facilitating nerve repair. Additionally, mild electrical stimulation induced by low‐intensity pulsed ultrasound (LIPUS) is found to enhance the recovery of motor function. These findings demonstrate the multifaceted benefits of the rGO/GelMA@PVGA composite conduit, which integrates physical guidance, oxidative stress inhibition, and ultrasound‐activated electrical stimulation, providing an unprecedented multimodal synergistic strategy with great potential for clinical treatment of peripheral nerve injury.

This study presents a conductive‐piezoelectric integrated microstructured conduit for peripheral nerve regeneration. The conduit combines reduced graphene oxide with a piezoelectric nanofiber membrane, enhancing nerve repair through physical guidance, oxidative stress inhibition, and ultrasound‐activated electrical stimulation. In vivo and in vitro results demonstrate its multifaceted benefits, offering a novel strategy for treating peripheral nerve injuries.

1. Introduction

Peripheral nerve injury (PNI) is a common neurotraumatic disorder caused by trauma‐induced loss of peripheral nerve function. This can result in partial or complete loss of sensory, motor, and autonomic functions, as well as neuropathic pain.^{[ 1} , ² , ^{3 ]} It is estimated that more than 5 million cases of PNI occur worldwide each year.^{[ 4 ]} Although PNI has some regenerative capacity, this natural healing process is often insufficient to achieve a fully functional recovery.^{[ 5} , ^{6 ]} Currently, autologous nerve grafts are considered the gold standard for the treatment of severe nerve gaps. However, challenges such as limited donor availability and donor site morbidity severely limit their therapeutic efficacy.^{[ 7} , ⁸ , ^{9 ]} Therefore, the repair of PNI remains a major therapeutic challenge that requires further attention.

In recent years, implantable piezoelectric nanogenerators (PENG) have attracted much attention in PNI treatment due to their self‐powered properties.^{[ 10 ]} Piezoelectric materials can convert mechanical forces into electrical energy in the absence of electrical stimulation,^{[ 11 ]} among which, polyvinylidene fluoride (PVDF) is an organic piezoelectric polymer that promotes the activity and function of fibroblasts.^{[ 12 ]} Low‐intensity pulsed ultrasound (LIPUS) can be used as a vibration source for piezoelectric materials, driving the deformation of piezoelectric materials to change their piezoelectric potential and induce charge release to stimulate cells on the surface.^{[ 13} , ¹⁴ , ¹⁵ , ^{16 ]} In addition, it has been shown that surface topography is an important parameter affecting cell growth,^{[ 17} , ¹⁸ , ^{19 ]} especially the grooved surface structure is beneficial for nerve repair. Methacrylated gelatin (GelMA) can be combined with PDMS micro‐pattern mold and UV curing technology to prepare GelMA gel with a grooved structure. Additionally, its gelatin backbone contains natural cell‐binding motifs such as arginine‐glycine‐aspartic acid (RGD) and matrix metalloproteinase (MMP) sequences, which are favorable for providing adhesion sites for cell proliferation and migration. Therefore, this GelMA gel has been widely prepared as a scaffold for neural repair and achieved good results.^{[ 20} , ²¹ , ²² , ^{23 ]}

After PNI, a dramatic secondary injury response occurs at the affected site, characterized by uncontrolled oxidative stress, a severe inflammatory response, and a rapid influx of glutamate and calcium ions.^{[ 24} , ²⁵ , ^{26 ]} Together, these responses lead to cytotoxic neuroexcitation, pathological necrosis, and neuronal apoptosis.^{[ 27} , ^{28 ]} Reduced graphene oxide (rGO) is a neurotransmission modulator that can be oxidized by adsorbing reactive oxygen species (ROS), thus reducing ROS levels. In addition, rGO has good electrical conductivity, which is beneficial for nerve repair, the restoration of the bioelectric signal plays an important role in the regeneration of peripheral nerves due to the electrically active properties of nerve tissue.^{[ 29} , ^{30 ]}

In this study, a biodegradable bilayer nerve conduit was prepared by using the piezoelectric PVDF/PLGA (PVGA) nanofiber membrane as the outer layer, and a conductive rGO/GelMA gel with a microgroove surface pattern was laid on PVGA as the inner layer (Figure 1 ). PLGA has good biocompatibility and biodegradability and can provide excellent mechanical support for nerve regeneration.^{[ 31} , ^{32 ]} In addition, as an emerging and highly promising research direction, ultrasound precision activation demonstrates the synergistic advantages of deep tissue penetration, remote spatiotemporal selectivity, and mechanical‐electrical energy interconversion.^{[ 33 ]} Furthermore, by utilizing rGO to enhance electrical conductivity and scavenge ROS, our conductive‐piezoelectric integrated microstructured conduits demonstrated remarkable regenerative potential in PNI repair. This innovative approach effectively promotes nerve regeneration without the aid of any percutaneous leads, providing an efficient and relatively safe solution for the treatment of nerve injury.

Figure 1.

Schematic representation of the fabrication process for a conductive‐piezoelectric integrated microstructured conduit for targeted dynamic modulation of oxidative stress and self‐generation, and its impact on accelerated nerve repair in vivo and in vitro. The micro‐groove structure is shown to enhance cell adhesion and proliferation. Graphene oxide and GelMA hydrogel are depicted as modulators of cellular oxidative stress, while piezoelectric activation via ultrasound is highlighted for its role in promoting nerve regeneration.

2. Results and Discussion

2.1. Preparation and Characterization of PVDF, PLGA, PVGA Nanofibrous Membranes

Previous research demonstrated that a 20% glycolic acid (GA) content in polymers induces a transition from a highly crystalline to an amorphous crystalline structure, thereby enhancing the hydrophilicity and degradation characteristics of the materials. In this study, we selected a PLGA copolymer with a ratio of L‐lactic acid (L‐LA) to GA (L‐LA: GA = 80:20) for its biodegradable properties. As shown in Figure S1 (Supporting Information), the hydrogen spectrum nuclear magnetic resonance (¹H NMR) confirmed the successful synthesis of PLGA.^{[ 34 ]} Consequently, these biodegradable PLGA materials were employed to fabricate a nanofiber membrane using electrospinning technology, which was then utilized as the outer layer of a nerve guidance conduit, offering superior mechanical support.^{[ 35 ]} The micro/nano topologies of various nanofibrous membranes were characterized by scanning electron microscopy (SEM), revealing that the high surface area ratio and porosity of these membranes were instrumental in promoting cell adhesion and proliferation.^{[ 36} , ^{37 ]} The elemental distribution within the nanofibrous membranes was further characterized by energy dispersive spectroscopy (EDS). The results indicated that fluorine (F) was exclusively present in PVDF and PVGA nanofibrous membranes, and the oxygen (O) content in PVGA nanofibrous membranes was notably higher compared to PVDF membranes, confirming the efficacy of the preparation process (Figure 2A). Of note, the fiber diameter of PVDF is mainly distributed in 400–500 nm, the fiber diameter of PLGA is mainly distributed in 700–800 nm, and PVGA exhibited an intermediate trend, with fiber diameter is mainly distributed in 600–700 nm (Figure S2, Supporting Information), the fiber diameter statistics show in Table S1 (Supporting Information).

Figure 2.

Characterizations of PVDF, PLGA, and PVGA nanofibrous membranes. A) SEM and EDS elemental mapping images. B) FTIR spectra. C) XRD patterns. D) TGA curving. E) d₃₃ piezoelectric constant. F) Piezoelectric amplitude and G) phase of PLGA nanofibrous membrane. H) Piezoelectric amplitude and I) phase of PVGA nanofibrous membrane. J) The output voltage of the PVGA nanofibrous membrane without or with LIPUS excitation.

The composite film's structure, pivotal to its performance, was elucidated through Fourier transform infrared (FTIR) and X‐ray diffractometry (XRD). FTIR analysis revealed characteristic peaks at 839 and 1274 cm⁻¹, assignable to the C–F out‐of‐plane deformation and –CH₂ rocking vibrations, respectively, which are diagnostic of the β phase of PVDF.^{[ 38 ]} PLGA showed a telescopic vibrational peak belonging to C = O at 1748 cm⁻¹ and a telescopic vibrational peak belonging to C–O at 1085 cm⁻¹.^{[ 39 ]} In the PVGA nanofiber membrane, two β‐phase characteristic bands were retained (Figure 2B). In the XRD spectra, both PVDF and PVGA nanofiber membranes showed diffraction peaks at 2θ = 20.3°, which corresponded to the β phase of PVDF.^{[ 40 ]} These results indicated that PVGA could well maintain the β phase of PVDF (Figure 2C). We evaluated the thermal stability of different nanofiber membranes by thermogravimetric analysis (TGA). At 600 °C, PLGA was almost completely thermally decomposed, while both PVDF and PVGA had more residuals. From the thermogravimetric curve, it could be verified that the mass ratio of PVDF to PLGA was close to 1:1 (Figure 2D).

The piezoelectric constants of PVDF, PLGA, and PVGA nanofiber membranes were measured using a quasi‐static d₃₃ meter after being prepared by electrostatic spinning. The results indicated that PVDF has a d₃₃ piezoelectric constant of 34.94 ± 0.83 pC N⁻¹, while no piezoelectric constant was detected for PLGA, indicating that it does not possess piezoelectric properties. The PVGA nanofiber membrane obtained after blending exhibited a d₃₃ piezoelectric constant of 17.4 ± 0.79 pC N⁻¹, indicating excellent piezoelectric properties (Figure 2E). The piezoelectric force microscopy (PFM) was used to investigate the piezoelectric effect of PVGA and PLGA nanofiber membranes. Compared with PLGA (Figure 2F,G), PVGA exhibited a classic butterfly‐shaped amplitude loop (Figure 2H). The piezoelectric hysteresis loop confirms the high piezoelectric properties of the PVGA nanofiber membrane, as the phase angle changes by 180° under a 10 V bias field (Figure 2I).^{[ 41 ]}

To validate the application of PVGA nanofiber membrane in ultrasound‐activated electrical stimulation, we examined the output voltage in the presence/absence of LIPUS. Without LIPUS stimulation, the output voltage of the PVGA nanofiber membrane was close to 0 V, while in the presence of LIPUS stimulation, the output voltage could reach ≈0.5 V, a voltage level sufficient to stimulate and guide the regeneration of nerve cells^{[ 42} , ^{43 ]} (Figure 2J). This showed the real‐time response of the PVGA nanofiber membrane to ultrasound stimulation.

The contact angles of the three groups of nanofiber membranes were not significantly different. They were more hydrophobic, which was not conducive to gel adhesion on the surface (Figure S3A, Supporting Information). Plasma treatment is a widely reported method to make the material surface more hydrophilic.^{[ 44 ]} Therefore, we treated the surface of the PVGA nanofiber membrane with plasma. There was an obvious decrease in the contact angle, and the surface became hydrophilic after 0.5 min of treatment. This change facilitated stronger adhesion of the gel to the nanofiber membrane (Figure S3B, Supporting Information). Furthermore, the nanofibers remained undamaged after plasma treatment, with only slight curvature observed. The fiber surface became rougher, which also contributed to the increase in hydrophilicity (refer to Figure S4, Supporting Information). Therefore, the optimal condition for subsequent experiments was chosen to be a plasma treatment of 0.5 min.

2.2. Topography, Mechanical, and Degradation Properties of rGO/GelMA@PVGA Composite Conduits

GelMA was obtained by modifying gelatin with methacrylic anhydride (MA). The chemical structures of gelatin and GelMA were verified by ¹H NMR and FTIR. The characteristic functional groups of MA (δ = 5.32 and 5.56 ppm) demonstrated the successful preparation of GelMA as shown by ¹H NMR comparative analysis (Figure S5A, Supporting Information). FTIR showed that the characteristic peaks observed at 1647, 1542, and 1238 cm⁻¹ belonged to the stretching vibration of the C = O bond of Amide III, the stretching vibration of the N–H bond of Amide II, and the stretching vibration of the N–H bond of Amide I (Figure S5B, Supporting Information). These FTIR analyses confirmed the successful synthesis of GelMA.^{[ 45 ]}

The FTIR spectra of GO showed that the characteristic peaks at 3423, 1717, 1617, 1384, and 1048 cm⁻¹ corresponded to O–H stretching, C = O stretching, C = O, O–H bending, and C–O stretching vibration, respectively. The C = O vibrational peak at 1617 cm⁻¹ and the O–H deformation peak at 1384 cm⁻¹ of the rGO obtained after reduction were significantly reduced, indicating that GO was successfully reduced to rGO (Figure S6A, Supporting Information). The Raman spectra at 1350 and 1589 cm⁻¹ correspond to the characteristic vibrational peaks in the D and G bands, respectively. The D/G intensity ratio was compared to evaluate the degree of reduction from GO to rGO. The I _D/I _G of rGO was 1.06 based on the Raman spectra, and the increase in the I _D/I _G ratio compared to the GO I _D/I _G of 0.97 indicated the successful reduction of GO to rGO (Figure S6B, Supporting Information).^{[ 46} , ^{47 ]}

In addition, the pores inside the gel also increased with the increase of rGO concentration, but after increasing to 1.0 mg mL⁻¹, due to the solution color being too black with strong light avoidance ability, it was difficult to achieve complete photocrosslinking inside the hydrogel, leading to the inhomogeneity of the internal pores (Figure S7, Supporting Information). We further analyzed the surface morphology of the conduit using SEM. It is clear that the ridges and grooves on the conduits were parallel, and the widths of the ridges and grooves were ≈10/15 µm. In addition, the depth of the grooves was ≈5 µm, and the thickness of the conduits was ≈90 µm (Figure 3A). From the section image, it could be seen that there was no obvious interface between the gel and the nanofiber membrane, which acted as an integrated membrane (Figure 3B).

Figure 3.

The microtopography of micropatterned rGO/GelMA@PVGA conduit membrane: A) the surface and B) the section. Mechanical properties of PVGA, GelMA@PVGA, and rGO/GelMA@PVGA conduit membrane were presented: C) stress–strain curve, D) tensile modulus, E) elongation at the break, and F) tensile strength. Circuit of G) rGO/GelMA@PVGA conduit membrane and H) PVGA nanofibrous membrane with LED light. Degradability of PVGA, GelMA@PVGA, and rGO/GelMA@PVGA conduit membrane: I) weight loss and J–M) morphological changes with 14, 28, 42, and 56d of degradation. Data are presented as mean ± standard deviation (SD) (n = 3 biological replicates).

Comparing the stress–strain curves of the three materials in Figure 3C, it could be observed that the difference in mechanical properties was not significant and that the decrease in both Young's modulus and elongation at break does not change much (Figure 3D,E). Although Figure 3F showed that the tensile strength of PVGA, GelMA@PVGA, and rGO/GelMA@PVGA are 1.98 ± 0.65 MP, 2.95 ± 0.40 MPa, and 3.58 ± 0.51 MPa. It can be concluded that the addition of GelMA and rGO has increased the maximum tensile strength to fracture the whole material subjected to static tension.

PVGA is capable of generating voltage under ultrasound stimulation, whereas it is not conductive by itself and cannot transfer the generated voltage to the membrane surface. Therefore, we covered the surface of the membrane with a GelMA gel containing rGO. When rGO was introduced, we noticed a significant change in the conductivity of the conduit. As the concentration of rGO increased, the resistance value decreased from 74.67 ± 3.06 to 15.67 ± 3.21 kΩ (Figure S8A, Supporting Information), and the conductivity was calculated according to Equation (1), which shows that the conductivity increased from 0.2 × 10⁻⁴ to 1.02 × 10⁻⁴ S cm⁻¹ (Figure S8B, Supporting Information). Interestingly, compared with the PVGA nanofiber membrane, the rGO/GelMA@PVGA conduit membrane on the surface could light up the LEDs by electric current, which vividly demonstrated the conductivity of the rGO/GelMA@PVGA conduit for nerve repair (Figure 3G,H),

$$\sigma =\frac{L}{R\times S}$$ (1)

where, σ is the conductivity, S cm⁻¹; L is the height of the gel, cm; R is the resistance of the gel, Ω; and S is the area of the gel, cm².

The degradation of PVGA, GelMA@PVGA, and rGO/GelMA@PVGA conduit membranes was examined by immersing them in DMEM medium for 14, 28, 42, and 56 days. After 56 days in DMEM, the mass of the PVGA membrane decreased by only 16%, indicating a slow degradation rate. In contrast, the mass of the GelMA@PVGA and rGO/GelMA@PVGA nerve repair membranes decreased by 70% and 74%, respectively, with a faster degradation rate, indicating that the major weight loss during degradation was the rGO/GelMA gel (Figure 3I). In addition, we also observed changes in the surface morphology of rGO/GelMA@PVGA conduit membrane during the degradation process by scanning electron microscopy, as shown in Figures 3J–M. In particular, after 28 days of immersion in the medium, defects appeared in the gel grooves, and the PVGA membrane at the bottom was exposed. At day 56, more defects appeared in the gel grooves on the surface, the width of the ridges decreased significantly, most of the PVGA nanofiber membrane on the bottom was exposed, and the gel on the surface was degraded. The degradation of the gel can release the internal rGO, thus eliminating ROS. To investigate whether the PVGA nanofiber membrane still has a piezoelectric effect after 56 days of degradation, we examined it under low‐frequency pulsed ultrasound. The signal peaks were affected and become no longer homogeneous, but a high output voltage could still be detected, indicating that it still has a good piezoelectric effect after 56 days of degradation (Figure S9, Supporting Information).

2.3. Cellular Toxicity, Proliferation, and Adhesion of the Anti‐Oxidative Piezoelectric Nerve Conduit Membrane

To assess the capacity of the nerve conduit to support cellular growth, Schwann cells (SCs) were seeded onto the gel layer surface. Following a 7‐day culture period, results from live/dead cell staining demonstrated robust cell proliferation on the gel layer surface for all experimental groups (Figure 4A). This finding was further corroborated by the CCK‐8 assay, revealing no significant differences in cellular proliferation among the various groups (Figure 4B). Subsequently, scanning electron microscopy was employed to directly observe the morphology of cells adhering to the gel surface. Notably, the cells can be observed adhering to the gel surface with grooves, while maintaining their normal morphology (Figure 4C). As can be seen in Figure S10 (Supporting Information), the Schwann cells also adhered well to the outer membrane of PVDF, PLGA, and PVGA nanofiber membranes, indicating that the rGO/GelMA@PVGA compliant conduit had great biocompatibility. Additionally, the conduit membrane can also effectively support the adhesion, proliferation, and migration of neuronal cells (PC12) (Figure S11A,B, Supporting Information). PC12 cells cultured on the conduit membrane exhibited enhanced neurite growth and pattern arrangement along the surface of micro‐grooves (Figure S11C, Supporting Information), further confirming the conduit's potential for nerve regeneration. These results collectively support the ability of the conduit to promote neural regeneration at both the Schwann cell and neuron levels.

Figure 4.

Adhesion and proliferation of SCs. A) Live/dead fluorescence staining of SCs after being cultured for 1 and 7 days. B) CCK‐8 assay results of culture SCs for 1–7 days. C) SEM images of SCs adhesion after being cultured on the with/without grooved surface of the rGO/GelMA@PVGA membrane. Data are presented as mean ± SD (n = 3 biological replicates). The statistical significance of the data was assessed using one‐way ANOVA followed by Tukey's multiple comparisons test. ^* p < 0.05, ns, no significance.

2.4. Anti‐Oxidative Piezoelectric Nerve Conduit Inhibits Apoptosis by Alleviating Mitochondrial Oxidative Stress

After nerve injury, Wallerian degeneration occurs distally, leading to an elevation in local inflammation and oxidative stress levels.^{[ 48} , ^{49 ]} Due to the presence of numerous functional groups on its surface, rGO exhibits the ability to adsorb superoxide ions (ROS). Therefore, we employed mitochondrial JC‐1 staining to assess cellular oxidative stress levels and evaluated the proportion of cell apoptosis. After exposure to oxidative stress, the intracellular mitochondrial membrane potential undergoes a decrease, leading to the entry of JC‐1 dye into the cytoplasm in its unpolymerized form, which exhibits a green fluorescence signal. Conversely, under conditions of low oxidative stress levels, JC‐1 exists in its polymerized state and emits a red fluorescence signal.^{[ 50} , ⁵¹ , ^{52 ]} As shown in Figure 5A, cells in each group were stimulated with TNF‐α (20 ng mL⁻¹) for 24 h. In the absence of reduced rGO in the hydrogel, cells exhibited higher levels of oxidative stress. Similarly, ROS staining results revealed a significant reduction in intracellular ROS release in the rGO/GelMA@PVGA+L and rGO/GelMA@PVGA groups (Figure 5B). Flow cytometry‐based apoptosis assays further indicated that reducing mitochondrial damage and inhibiting oxidative stress could potentially mitigate post‐injury apoptosis (Figure 5C). Quantitative analysis of ROS revealed that rGO adsorption significantly reduced ROS expression and alleviated mitochondrial damage (Figure 5D,E). Excessive oxidative stress is known to impair cellular function and trigger programmed cell death. Figure 5F illustrates that compared with the GelMA@PVGA+L control, the rGO‐modified group exhibited a more pronounced decrease in early apoptotic cells (Q2 quadrant). Figure 5G–I demonstrate that the expression of antioxidant genes was significantly lower in the GelMA@PVGA+L and rGO/GelMA@PVGA+L groups, and the related proteins were also significantly lower after rGO modification (Figure 5J,K). These results confirm that rGO can effectively reduce ROS expression and mitigate mitochondrial damage.

Figure 5.

Oxidative stress and apoptosis in SCs. A) Mitochondrial JC‐1 staining. B) Fluorescence images of ROS staining. C) Flow cytometry apoptosis during Q2 period. D) Semi‐quantitative analysis of mitochondrial JC‐1 staining. E) Semi‐quantitative analysis of the total ROS level. F) Quantitative analysis of apoptosis ratio by flow cytometry. Expression of antioxidant gene: G) SOD1, H) SOD2, I) NRF2. J) Western blot analysis of SOD1, SOD2, and NRF2 in SCs. K) Semiquantitative SOD1, SOD2, and NRF2 expression profiles of gray scale values of Western blots, normalized to β‐actin. Data are presented as mean ± SD (n = 3 biological replicates). The statistical significance of the data was assessed using one‐way ANOVA followed by Tukey's multiple comparisons test. ^* p < 0.05, and ^** p < 0.01, ns, no significance.

2.5. Anti‐Oxidative Piezoelectric Nerve Conduit Alleviates the Oxidative Stress and Promotes Nerve Regeneration In Vivo

A rat sciatic nerve clamp injury model was constructed ≈1 cm proximal to the bifurcation of the sciatic nerve. rGO/GelMA@PVGA nerve conduit membrane measuring 1 cm in length and 0.5 cm in width was curled into a tubular shape to cover the injury site and fixed with sutures to evaluate the effect of a nerve repair scaffold on sciatic nerve regeneration in rats (Figure S12, Supporting Information). After nerve injury, local production of large amounts of free radicals exacerbates inflammation and impedes nerve regeneration and repair.^{[ 53 ]} As the gel degrades, rGO in the conduit gradually exposes itself to the damaged nerve surface, alleviating oxidative stress levels through its upper cap energy group's interaction with free radicals. Figure 6A illustrates ROS levels at clamp 2 weeks post‐surgery. Both rGO/GelMA@PVGA+L and rGO/GelMA@PVGA (containing rGO in the gel) exhibited a remarkable ability to clear ROS and effectively alleviated oxidative stress following nerve injury. However, in comparison to the rGO/GelMA@PVGA group, it was observed that the rGO/GelMA@PVGA+L group did not exhibit a significantly enhanced capacity for ROS clearance. This suggests that the combined effect of low‐frequency pulsed ultrasound‐generated electrical stimulation does not impact the level of oxidative stress. At week 8 post‐surgery, NF200 immunofluorescence staining was performed on the nerves to assess regeneration degree based on fluorescence area intensity and proportion (Figure 6B). According to semi‐quantitative results, the group treated with rGO for oxidative stress relief combined with low‐frequency pulsed ultrasound for electrical stimulation displayed significantly higher NF200 positive area compared to other groups. These findings indicate its strong potential in promoting nerve regeneration (Figure 6C,D).

Figure 6.

Neuronal ROS and regeneration‐specific protein fluorescence staining. A) ROS fluorescence staining in 2 weeks (the first line bar = 500 µm, the second line bar = 100 µm). B) NF200 fluorescence staining in 8 weeks. Semi‐quantitative analysis of C) ROS and D) NF200. Data are presented as mean ± SD (n = 3 biological replicates). The statistical significance of the data was assessed using one‐way ANOVA followed by Tukey's multiple comparisons test. ^* p < 0.05, and ^** p < 0.01, ns, no significance.

2.6. Anti‐Oxidative Piezoelectric Nerve Conduit Promotes the Recovery of Nerve and the Target Muscle Structure

The quality of the regenerated nerve is indicated by the thickness of the myelin sheath and its proportion to the cross‐sectional area of the nerve. Using transmission electron microscopy, we analyzed regenerated nerve cross sections eight weeks post‐surgery. In Figure 7A, the basic nerve structure was restored in each group. The myelin sheath thickness and area proportion in the rGO/GelMA@PVGA+L group showed significant differences compared to the rGO/GelMA@PVGA group, as illustrated in Figure 7B,C. It is suggested that electrical stimulation of low‐intensity pulsed ultrasound can enhance the regeneration of nerve effectively. Concurrently, the GelMA@PVGA+L group exhibited reduced thickness and proportion of myelin sheath compared to the rGO/GelMA@PVGA+L group, possibly due to the antioxidative properties of rGO during the initial phase and enhancement of the early repair environment.

Figure 7.

Pathological assessment of the regenerating nerves and target muscle. A) TEM analysis of regenerated nerves 8 weeks post‐surgery. Semi‐quantitative analysis of B) myelin sheath thickness and C) area ratio. D) H&E and Masson's trichrome staining of triceps surae muscle 8 weeks after operation. Semi‐quantitative analysis of E) muscle wet weight ratio and F) collagen area ratio. Data are presented as mean ± SD (n = 3 biological replicates). The statistical significance of the data was assessed using one‐way ANOVA followed by Tukey's multiple comparisons test. ^* p < 0.05, ^** p < 0.01 and ^*** p < 0.001, ns, no significance.

The nerve exerts a nutritional influence on the target muscle, and during nerve injury, the target muscle inevitably undergoes atrophy due to transient disinnervation. Therefore, it is crucial to assess the recovery of the target muscle. The structure and collagen accumulation in the specific muscle were examined through H&E and Masson's staining, with the muscle's weight ratio measured at week 8 after the surgery. Figure 7D clearly showed significant muscle atrophy without any treatment. H&E staining revealed the fundamental muscle fibers, while Masson's trichrome staining demonstrated extensive fiber fissures accompanied by collagen deposition, which could potentially impede motor function recovery. At 8 weeks post‐surgery, the rGO/GelMA@PVGA+L group exhibited comparable sizes of triceps surae on both surgical and control sides. Masson's staining indicated reduced collagen deposition, with significantly higher wet‐weight ratios observed in the triceps surae on the surgical side compared to other groups (Figure 7E). Moreover, the area of collagen deposition was notably lower than that observed in other groups (Figure 7F).

2.7. Anti‐Oxidative Piezoelectric Nerve Conduit Improves Motor Function In Vivo

The CatWalk system was used to evaluate the motor function of the rats at 4 and 8 weeks after implantation (Figure 8A). The fourth week's findings revealed that the rGO/GelMA@PVGA+L group did not differ significantly from the rGO/GelMA@PVGA group, but both outperformed the control group. This suggests that rGO's management of oxidative stress enhanced motor function recovery (Figure 8B). In the eighth week, this difference was more significant. At the same time, the pressing depth ratio of the surgical side (Left Hind) and the control side (Right Hind) in the rGO/GelMA@PVGA+L group was closest to 1 (Figure 8C). The improvement compared to the rGO/GelMA@PVGA group was substantial, indicating that the addition of low‐frequency pulsed ultrasound enhanced motor function recovery through weak electrical stimulation. This was directly confirmed by the sciatic nerve index (SFI) results (Figure 8D). Compound muscle action potential (CMAP) and nerve conduction velocity (NCV) measurements (Figures S13 and S14, Supporting Information) indicated that the rGO/GelMA@PVGA+L group could significantly promote the recovery of neuromuscular function. Lastly, according to the results of H&E staining of major organs in rats 8 weeks after surgery (Figure S15, Supporting Information), there was no significant inflammatory infiltration in the lungs, liver, kidneys, spleen, and heart. Therefore, the composite scaffold exhibited good biocompatibility during in vivo nerve regeneration and has potential application value.

Figure 8.

Behavioral analysis of rats. A) Footprint intensity of the surgical side (Left Hind) in CatWalk at 4 and 8 weeks. B,C) Analysis of footprint intensity ratio between surgical and control sides at 4 and 8 weeks. D) Sciatic nerve index analysis. Data are presented as mean ± SD (n = 3 biological replicates). The statistical significance of the data was assessed using one‐way ANOVA followed by Tukey's multiple comparisons test. ^* p < 0.05 and ^** p < 0.01, ns, no significance.

3. Conclusion

In this study, we have successfully developed a novel bio‐conduit by integrating a gel containing reduced rGO with a piezoelectric electrospun nanofiber membrane, thereby facilitating peripheral nerve regeneration. Microgroove topology on the conduit surface to induce cell proliferation and migration. By utilizing the piezoelectric property of PVGA and the conductivity of rGO, a conductive‐piezoelectric integrated material system can be constructed synergistically. In addition, the antioxidant property of rGO significantly reduces the level of ROS at the injury site, providing a “protective microenvironment” for nerve regeneration. In vitro experiments have confirmed the biocompatibility of the composite membrane and its ability to support cell adhesion. Importantly, the presence of rGO within the membrane effectively counters oxidative stress by mitochondrial damage, thereby reducing cell apoptosis in the damaged area. Moreover, the outer PVGA membrane also demonstrated excellent piezoelectric effects that synergistically enhanced nerve regeneration. In vivo experiments further validated its capacity to restore neural structure and promote motor function in rats. To sum up, the conductive‐piezoelectric integrated microstructured conduits show significant potential for treating peripheral nerve damage and other conditions related to nerve injuries in the future.

4. Experimental Section

Materials Preparation

L‐lactide (L‐LA) and glycolic acid (GA) were purchased from Daigang Biotech Co., Ltd. Stannous octoate (SnOct₂, purity 95%) was purchased from Aldrich Biochemical Technology Co., Ltd. (Shanghai, China). Methylene chloride (DCM) (purity ≥ 99.5%) and anhydrous ethanol (purity ≥ 99.7%) were both purchased from Jiangsu Anway Chemical Technology Co., Ltd. (Jiangsu, China). Polyvinylidene fluoride (PVDF) (Mw = 400 kDa) was provided by Yuanye Biotechnology Co., Ltd. (Shanghai, China). N, N‐Dimethylformamide (DMF) and tetrahydrofuran (THF) were purchased from Taitan Scientific Co., Ltd. (Shanghai, China). Graphene oxide (GO) was purchased from Alab Chemical Technology Co., Ltd. (Shanghai, China). Gelatin (type A) was provided by Sigma (St. Louis, MO, USA). Methacrylic anhydride (MA) was provided by Aladdin Industrial Corporation. 2‐hydroxy‐4′‐(2‐hydroxyethoxy)‐2‐methylpropiophenone (I2959) was purchased by Yuanye Biotechnology Co., Ltd. (Shanghai, China).

Preparation of PLGA Copolymer

The preparation of PLGA (L‐LA: GA = 80: 20, M_w = 270 kDa) was performed via Ring‐Opening polymerization. L‐LA and GA were mixed in a certain ratio, the catalyst stannous octanoate was added, and the reaction was carried out under argon atmosphere at 140 °C for 1 h. The product was dissolved in DCM and purified with anhydrous ethanol. The collected products were dried in an oven at 40 °C to obtain pure PLGA.

Preparation of Nanofiber Membrane

PLGA and PVDF (1:1 by weight) were dissolved in a mixture of DMF and THF (1:1 by volume) at a concentration of 10% (w/v), and the mixed solution was transferred into a syringe with an 18‐gauge needle. The nanofibers were collected on a drum. The following electrospinning parameters were used: voltage 25 kV; work distance 13 cm; injection rate 1.0 mL h⁻¹; roller rotation speed 1000 rpm. The same process was used in PLGA and PVDF nanofiber membrane preparation.

Preparation of GelMA and rGO

GelMA and rGO were prepared according to previously reported methods.^{[ 54} , ^{55 ]} 10 g of gelatin was dissolved in 100 mL of PBS at 50 °C, and 12 mL of MA was added slowly dropwise and reacted at 50 °C for 4 h. The resulting mixture was dialyzed with deionized water (DI) at 45 °C for one week, and the final product was obtained after freeze‐drying. GO was immersed in 2 mg mL⁻¹ of ascorbic acid solution and reacted at 37 °C protected from light. After washing and lyophilization, rGO was obtained.

Preparation of rGO/GelMA@PVGA Composite Scaffold

The rGO was ultrasonically dispersed in PBS to form 0, 0.3, 0.5, and 1.0 mg mL⁻¹ suspensions, and then 15% (w/v) GelMA and 0.5% (w/v) photoinitiator I2959 were prepared to form the precursor solutions of GelMA, 0.3 rGO/GelMA, 0.5 rGO/GelMA, and 1.0 rGO/GelMA. After treating the PVGA nanofibrous membranes with plasma for 0.5 min, a PDMS mold was used to form a layer of rGO/GelMA gel with grooves on the surface of PVGA nanofibrous membranes, with a UV curing time of 5 min. The dimensions of the mold were as follows: grooves of 15 µm in length, with a spacing of 10 µm, and a depth of 5 µm.

Characterization of rGO/GelMA@PVGA Conduits with Microgroove Structure

The functional presentation and chemical structure of the conduits were characterized by ¹HNMR with tetramethylsilane as the internal standard and FTIR in the wavelength range of 400–4000 cm⁻¹. The surface morphology of the micropatterned conduits was analyzed by SEM (HITACHI, Japan), after sputter coating with gold. The diffraction patterns of conduits were analyzed via an automated X‐ray diffractometer with angles from 10° to 80° (RIGAKU, Japan). The thermal properties of conduits were investigated by thermogravimetric analysis (TGA, NETZSCH, Germany). Measurement of piezoelectric constants through a quasi‐static d₃₃ measuring device (ZJ‐3AN, Shenzhen Yice Medical Test Co., Ltd). The piezoelectric amplitude and phase were characterized by piezoresponse force microscopy (PFM, Bruker Dimension Icon Scanning Probe Microscope, Germany). Using the Keithley DMM6500 (USA) to measure voltages generated by ultrasonic (Shenzhen WELL.D Medical Electronics Co., Ltd) stimulation of piezoelectricity. Surface wettability was examined by measurement of the water contact angle (Shanghai Zhongchen Digital Technology Equipment Co., Ltd). Raman spectroscopic analysis was performed to examine the chemical reduction of rGO. The conductivity of rGO/GelMA(G)@PVGA conduits was characterized by LED bulb lighting tests and by measuring the resistance of films with a digital multimeter. The degradation properties were analyzed by recording the mass loss and microstructure change of the conduits at 37 °C.

Cell Culture, Cell Viability, and Oxidative Stress Assay

Schwann cell line (SCs) was obtained from the National Collection of Authenticated Cell Cultures (Chinese Academy of Science, RRID:CVCL_4694, CSTR:19375.09.3101RATGNR6). Rat pheochromocytoma cells (PC12) were obtained from the National Collection of Authenticated Cell Cultures (Chinese Academy of Science, RRID:CVCL_0481, CSTR:19375.09.3101RATTCR8). The cells were cultured in Dulbecco's modified eagle medium (DMEM) with high glucose, supplemented with 10% fetal bovine serum and 1% penicillin‐streptomycin solution under standard conditions (37°C, 5% CO₂). The anti‐oxidative stress ability of the materials in vitro was evaluated by mitochondrial JC‐1 staining, total ROS assay, and apoptosis flow analysis (Beyotime C2006). After treated with TNF‐α (20 ng mL⁻¹) for 24 h, RSC96 cells (1 × 10⁵ cells well⁻¹) were seeded on the surface of each group of materials that had been partially degraded in 6‐well plates and 12‐well plates. After three days of co‐culture, the level of mitochondrial damage in cells in 12‐well plates was assessed using the JC‐1 kit, cells in 6‐well plates were collected, and the apoptotic rate was detected using the flow apoptosis kit. Total ROS levels were determined using the ROS kit. Flowjo V10 was the program used for analyzing flow cytometry data.

Immunofluorescent Staining

1 mm sciatic nerve tissue was collected according to the clamp tags for the preparation of a frozen section. Sciatic nerve sections were blocked with PBS containing 5% serum and were incubated with primary antibodies overnight at 4 °C. After rinsing three times with PBS, the samples were incubated with secondary antibodies for 2 h at room temperature. Cell nuclei were stained with DAPI (G1407‐25ML, Servicebio). Stained sections were observed with a confocal laser scanning microscope (FV1000, Olympus).

Sciatic Nerve Clamp Model

The Ethics Committee of Zhongshan Hospital, Fudan University, approved the experimental sciatic nerve clamp model procedure (reference number SYXK (Hu) 2023‐0027). Fifty male Sprague‐Dawley rats, aged 6 weeks and weighing 120 to 180 g, were housed separately in stainless steel cages at 22 °C with a 12‐h light and dark cycle. These rats were obtained from Shanghai Jiesijie Laboratory Animal Co., Ltd., China, and all cages were located in the same room. Ten animals were assigned to each of the five groups randomly, including the control group, GelMA@PVGA group, GelMA@PVGA+L group, rGO/GelMA@PVGA group, and rGO/GelMA@PVGA+L group. Animals were anesthetized with an intraperitoneal injection of 30 mg kg⁻¹ pentobarbital sodium before exposure to the left sciatic nerve. The peripheral nerve injury was simulated by applying a vascular clamp 1mm away from the sciatic nerve bifurcation for two 15‐s intervals. The clamp width was ≈1mm, and the distal end was marked with a 10‐0 suture after clamping. Subsequently, the site was carefully wrapped and sutured using sutures. Animals in the GelMA@PVGA+L and rGO/GelMA@PVGA+L groups received daily 20‐min sessions of low‐intensity pulsed ultrasound. The settings for low‐intensity pulsed ultrasound included a frequency of 1 MHz, a power density of 0.5 W cm⁻², and a duty cycle of 20%.

Nerve Morphometric Analysis

After 8 weeks post‐surgery, rats were sacrificed, and three nerves from each of the control group, GelMA@PVGA group, GelMA@PVGA+L group, rGO/GelMA@PVGA group, and rGO/GelMA@PVGA+L group were preserved in 10% formalin for 24 h before being dehydrated with ethanol solutions of varying concentrations. Immunohistochemical staining with primary antibodies targeting ROS (PA5‐67241, ThermoFisher) and NF200 (18934‐1‐AP, Proteintech) was conducted to evaluate the number of cells showing positive staining, and this was compared across the different groups. Additionally, nerves that had been regenerated in each group were isolated and placed in a 2.5% glutaraldehyde solution for 24 h. Following this, the sciatic nerves were stained with 1% osmium tetroxide and examined using transmission electron microscopy (TEM, JEM‐2100) after undergoing dehydration and embedding. Five randomly selected regions of each sample were imaged using TEM, and quantitative analysis was performed by measuring the myelin sheath thickness and the ratio of myelin area to nerve fiber cross‐sectional area.

Histological Assessment

The animals were sacrificed by overdose anesthesia 8 months after surgery. The tissue was collected and the specimens were stored in a solution containing 4% paraformaldehyde. Next, the samples were subjected to a standard paraffin embedding process and then transformed into slides of 5 µm in thickness. To perform H&E staining, the slides were initially placed in a hematoxylin solution, followed by an alcohol rinse, and then submerged in Eosin dye for 5 min. Massons' Trichrome staining was carried out using the slides from identical samples, according to the guidelines provided by the manufacturer.

Behavioral Tests

Researchers utilized the CatWalk XT system from Noldus Information Technology in the Netherlands to observe the behavior of rats at 4/8 weeks post‐surgery. The SFI was determined by applying the formula following the footprint measurement, with “E” representing the “experimental side” and “N” representing the “non‐experimental side”, and the SFI was calculated according to Equation (2).

$$\begin{array}{ccc}\hfill SFI& =& -38.3\times \frac{EPL-NPL}{NPL}+109.5\times \frac{ETS-NTS}{NTS}\hfill \\ & & +13.3\times \frac{EIT-NIT}{NIT}-8.8\hfill \end{array}$$ (2)

The major internal organs (lung, liver, kidney, spleen, and heart) were collected and processed into slides embedded in paraffin eight months post‐surgery. Next, the samples underwent standard H&E staining and were examined using a light microscope.

Statistical Analysis

All statistical analyses were conducted with GraphPad Prism 9.0 software (GraphPad Software). Experimental data were presented as the mean ± standard deviation (SD), derived from at least 3 independent experiments. The statistical significance of the data was evaluated using one‐way ANOVA followed by Tukey's multiple comparisons test in GraphPad Prism 9.0. The level of significant difference was set as ^* p < 0.05 and ^** p < 0.01, and ^*** p < 0.001, ns, not significant.

Conflict of Interest

The authors declare no conflict of interest.

Author Contributions

D.Z. and M.B. contributed to writing the original draft, methodology, and data curation. L.W., L.X., J.H., and J.Z. contributed to writing, review, and editing, with L.W. also handling project administration. B.Y. and Q.C. were responsible for validation, and Y.L. and L.J. oversaw supervision and project administration.

Supporting information

Supporting Information

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.