Pulsed electromagnetic field-assisted reduced graphene oxide composite 3D printed nerve scaffold promotes sciatic nerve regeneration in rats

Peripheral nerve injuries can lead to sensory or motor deficits that have a serious impact on a patient’s mental health and quality of life. Nevertheless, it remains a major clinical challenge to develop functional nerve conduits as an alternative to autologous grafts. We applied reduced graphene oxide (rGO) as a bioactive conductive material to impart electrophysiological properties to a 3D printed scaffold and the application of a pulsed magnetic field to excite the formation of microcurrents and induce nerve regeneration. In vitro studies showed that the nerve scaffold and the pulsed magnetic field made no effect on cell survival, increased S-100β protein expression, enhanced cell adhesion, and increased the expression level of nerve regeneration-related mRNAs. In vivo experiments suggested that the protocol was effective in promoting nerve regeneration, resulting in functional recovery of sciatic nerves in rats, when they were damaged close to that of the autologous nerve graft, and increased expression of S-100β, NF200, and GAP43. These results indicate that rGO composite nerve scaffolds combined with pulsed magnetic field stimulation have great potential for peripheral nerve rehabilitation.


Introduction
Trauma and surgery, including tumor resection, can cause peripheral nerve injury (PNI) that results in long-term disability if not treated promptly and appropriately and is a significant financial burden that affects the patient physically and psychologically [1,2].Long-distance nerve injuries are often accompanied by a substantial loss of nerve tissue, which cannot be anastomosed end-to-end and requires surgical graft repair to restore function [3].Unlike solid organ transplants, such as the heart, liver, kidney, or lung, transplanted nerves cannot function per se but provide an optimal scaffold for axon lengthening.Autologous nerve grafts have been regarded as the gold standard of the surgical solution for peripheral nerve defects [4]; nevertheless, autologous nerve grafts have restrictions, including donorarea complications, limited donors, and mismatched canal diameters, that make the overall repair less than optimal [5].To avoid this problem, a variety of artificial nerve-guiding conduits (NGCs) that are structurally and compositionally similar to natural nerves have been fabricated using natural or synthetic materials [6,7].Among them, three-dimensional (3D) bioprinting is a pioneering technology that helps to reproduce the unique characteristics of complex human tissues and organs, allowing for precise positioning of biomaterials, molecules, and living cells, as well as spatial control of the placement of functional components [8].However, the simple physical connection of NGCs does not lead to good nerve regeneration [9, 10]; therefore, researchers are exploring the application of biostimulation.Decellularized extracellular matrix (dECM) plays an important role in supporting a wide range of cells by promoting specific physiological properties that can recapitulate the microenvironmental ecological niche inherent in 3D cell-printed structures [11].A 3D printed peripheral nervous system dECM scaffold has shown favorable functional effects in regenerating sciatic nerve [12].
Growth factors or cellular activators are the most widely used for in vivo biostimulators [13][14][15], but their slow-release systems are complicated to construct.Conversely, physical stimuli can be used as biostimulators to provide 'bioactivity' under a certain level of intensity, and most physical stimuli are non-invasive and have the advantages of being easily accessible and highly controllable [16].Electrical stimulation plays an important role in nerve recovery [17,18] and is used as a form of biostimulation in nerve regeneration.Electrical stimulation uses both electromagnetic and magnetic forces that interact with each other.The flux of the electromagnetic force can change with displacement, which is called as the electromagnetic field (EMF) [19].A pulsed EMF (PEMF) is an electromagnetic field with a specific amplitude and waveform.Cell proliferation, DNA replication, wound healing, cytokine expression, and cell differentiation are all modulated by various types of EMF, including PEMFs [20][21][22][23].The primarily cause PEMFs can cause biological changes refers to that they can easily pass through cells, and a major component of their biological basis is the synthesis of proteins, ion channel regulation, as well as the secretion of growth factors [23].
Because of the important role of electrical stimulation in clinical medicine, many conductive materials have been used in neural tissue engineering to obtain electrophysiological properties [24,25].Graphene, a two-dimensional crystal consisting of six-membered rings of sp2 hybridized carbon atoms and their derivatives, has become a popular material in the biomaterials field owing to its unique physicochemical, mechanical, electrical, thermal, and biological properties [26,27].However, the hydrophobicity and cytotoxicity of graphene limit its application in organization engineering [24].As a result of the carboxyl groups on its surface, graphene oxide (GO) is an excellent dispersant in certain polar solvents.However, oxidation of graphene reduces its aromaticity, which weakens its electrical properties [28].Through restoring the sp2-carbon bond, the conversion of graphene oxide to reduced GO (rGO) using chemical or thermal treatments can enhance the electrical properties of GO [29].
To date, the family of graphene and its related materials has also been shown to promote osteogenesis [30], neural [31] differentiation, and cardiac differentiation of stem cells [32].As a result of graphene's functionality, the field of cell and tissue engineering has been fascinated by it in recent years owing to its synergistic effect observed in osteogenesis and neural differentiation in which a pulsed magnetic field irradiating graphene-based nanosheets caused free electron residues at their outer edges to move, generating an electric current via the nanosheets [33].Lim et al determined the behavior of GO and rGO in a pulsed magnetic field by showing that rGO exhibited a positive effect, whereas GO exhibited a negative effect, and that the magnetic moments generated in rGO nanosheets could be converted into a current of 13.9 × 10 mA m −2 .The microcurrent generated by PEMF irradiation can affect cells [33].This combined application of rGO and PEMF for osteogenesis has been reported previously that can promote osteogenic differentiation ability with more calcium deposition [34] but has rarely been combined for the repair of peripheral nerve defects.
Currently, 3D bioprinting has emerged as a potential and effective approach for fabricating complex tissue structures for tissue engineering and regenerative medicine [35,36].Owing to the complexity of the anatomical structure of peripheral nerves, 3D bioprinting was used to manufacture the nerve tract and the surrounding membrane.One study showed that the porous walls of the tubular structure accelerated nutrient diffusion and vascularization, and parallel channels were conducive to oriented nerve growth [37].Therefore, we used 3D bioprinting to construct the topological structures of nerve scaffolds.
In this study, we focused on the combined effect of rGO contained scaffold and PEMF in inducing nerve regeneration.Physical and mechanical properties of the hydrogel were first evaluated.Biocompatibility analysis and 3D bioprinting was carried out with RSC96 cells.The synergistic effect of rGO-scaffold and PEMF was investigated through cell viability, S-100β expression, cell extension and gene expression level.In addition, we measured morphometric and electrophysiological data in a Sprague-Dawley (SD) rat sciatic nerve defect model and examined GAP43, NF-200, S-100β in the regenerated nerves by immunohistochemistry.

Original ink preparation
Gelatin (Sigma-Aldrich, V900863), sodium alginate (Sigma-Aldrich, 18 097), and silk fibroin (SF) (Aladdin, W293432) were sterilized using Co60 irradiation prior to the use.The rGO (Xianfeng Nano, 100 022) was sterilized using UV irradiation for 30 min prior to use.Before configuring the materials, a cell count of 2000 for each sample was cultured with various concentrations of rGO in the culture medium, and cell growth was detected to determine the cytotoxicity of rGO.A concentration of 0.2% rGO was selected for experimentation.The materials were dissolved in DMEM containing 10% FBS and 1% penicillin-streptomycin to configure original inks containing different concentrations of SF.The concentrations of gelatin, sodium alginate, and rGO in the inks were 8.89%, 2.22%, and 0.22%, respectively, and the concentrations of SF were 0%, 5.56%, 11.11%, 13.89%, and 16.67%.The inks were mixed with an appropriate amount of medium to obtain different inks of gelatin, sodium alginate, and rGO at concentrations of 8%, 2%, and 0.2%, respectively, and at concentrations of 0%, 5%, 10%, 12.5%, and 15% SF.

Material physical and chemical properties testing 2.3.1. Scanning electron microscopy (SEM) detection
Each bioink was sprayed with gold after freeze-drying for SEM (SU8020, Hitachi) imaging of the fine structure of the hydrogels.The porosities of the different hydrogels were calculated using ImageJ software, and the average values were used for statistical comparison.The 12.5% SF was selected as the optimal concentration.

FTIR spectral analysis
A NICOLET IS10 FTIR spectrometer (Thermo Fisher Scientific, USA) was adopted for investigating chemical composition of each bioink.All the samples were lyophilized, ground with 200 mg potassium bromide in an onyx mortar at around 1:20 ratio, and then compressed into thin slices.The slices were positioned into a sample chamber with an automatic recognition function, and the spectrum was defined in the range of 4000-400 cm −1 with a resolution of 4 cm −1 and a scan time of nearly 100 s.The FTIR spectra were analyzed after background and baseline corrections.

Compressive modulus analysis
A universal testing machine (UTM4103, SUNS, China) was applied for the evaluation of the compressive properties.Briefly, the cross-linked hydrogels of each group were placed between load-bearing sensors with circular metal plates at room temperature.The samples were compressed at a rate of 0.5 mm min −1 , resulting in a deformation of 80% from their initial thickness.Throughout the compressive testing process, meticulous data recording was conducted.The modulus was determined by calculating the slope of the stress-strain curve.

Viscosity-shear rate and viscosity-temperature tests
A rotational rheometer (Discovery HR2; TA Instruments, Inc., USA) was adopted for testing the viscosity of each group of bioinks.The composite viscosity-shear rate curves were obtained by scanning the bioinks for 200 s at 1% strain, 20 • C, and a shear rate of 0.1-1000 s −1 in order to identify the correlation between the composite viscosity and shear rate of each bioink group.With the aim of further evaluating the correlation between the composite viscosity of the bioinks and temperature, the bioinks were scanned at a frequency of 1 Hz, a strain of 1%, and a temperature of 25.81 • C-45.73 • C for 245 s to obtain composite viscosity-temperature curves.

Thixotropy assay
The thixotropic behavior of each bioink group was assessed using a rotational rheometer.The experiments were conducted in three steps at a temperature of 20 • C and a frequency of 1 Hz.Firstly, a shear rate of 0.1 s −1 was used to simulate the hydrogel state prior to 3D bioprinting for 60 s.Subsequently, the shear rate was altered to 100 s −1 and maintained for 30 s to simulate the bio-3D printing process.At last, after the completion of bio-3D printing, the shear rate was reverted to 0.1 s −1 for 60 s, aiming to simulate the shear force exerted by the scaffolds during fluid exchange, and viscosity-time curves were created and analyzed.

Material magnetism detection
To detect the magnetic properties of the bioinks, an MPMS-XL5 superconducting quantum magnetic measurement system (SQUID, Quantum Design, USA) was applied.The rGO-free, and rGO ink were lyophilized, ground into powder, and placed in a sample chamber with a magnetic field ranging from −2 mT to 2 mT and a resolution of 0.01 mT to plot hysteresis lines.

Generation of PEMFs
Briefly, PEMFs were generated using a solenoid, and a sinusoidal electrical signal (±1 V, 50 Hz) was generated with the use of a function generator (VICTOR, VC2015H).The PEMFs obtained were 2.00 ± 0.10 mT, 50 Hz, measured by a Gaussiometer (Senjie, SJ700) and the samples were stimulated for 20 min d −1 .

Material toxicity assay and screening of pulsed magnetic field intensity
In this study, a cell counting kit (CCK8, Japan, LK815) was adopted for detecting cell proliferation on days 1, 3, 5, and 7 following the treatment with the rGO scaffold extract and 24 h after magnetic field stimulation of different densities.The cell counts of the initial samples were all 2000 (2 × 10 5 cells for MF stimulation screening), and the control was not treated by the extraction solution or subjected to magnetic field stimulation.Briefly, the working solution was obtained by mixing CCK8 solution with culture medium at the ratio of 1:9, and the sample of each time point (n = 3) was incubated with 550 µl CCK8 working solution for 1 h.Then, a 110 µl of the supernatant was transferred to a 96-well plate.In addition, absorbance (OD) values were measured at 450 nm using an enzyme labeling instrument (BioTek ELX800, VT, USA).

3D bioprinting
Prior to 3D bioprinting, a cylinder of 8 mm height and 4 mm diameter was pre-engineered in the Medprin bioprinter (BMP-C300-T300-IN3) application.The graphic was layered at 200 µm height per layer.Cell concentration was adjusted to 9 × 10 6 ml −1 , after which the original ink and cell suspension were mixed by a ratio of 9:1 to reach the final concentration of 0.2% rGO-8% Gel-2% Alg-12.5% SF and the cell suspension of 1 × 10 6 ml −1 .The bioink above was used in all in vitro evaluation.Cell-hydrogel mixtures were filled into 1 ml syringes prior to printing.With the consideration of the bionic structure and cellular activity, a Musashi printing nozzle of 300 µm inner diameter was selected.According to the previous study [38] and experience [39], following the repetition of the test, the room temperature was defined at 16 • C, and the hydrogel syringe was maintained at 28.3 • C (figure 1).With these parameters, the bioink was maintained in a liquid state and smoothly extruded, and the deposition could be rapidly gelled on the platform, whereas the cells kept a high viability level.The extrusion speed was defined at 0.15 ml min −1 .The nozzle scanning speed was of 3.5 mm s −1 .For crosslinking, the scaffolds were soaked in 50 mM sterilized CaCl2 solution for 5 min after printing.Subgroups requiring pulsed magnetic field stimulation were stimulated daily for 20 min within a 2 mT, 50 Hz pulsed magnetic field.

Calcein AM and propidium iodide (PI) staining
Cell viability in printed scaffolds was evaluated with the use of a fluorescent live/dead viability assay kit (KeyGEN Bio-TECH, China, KGAF001).Briefly, the printed scaffolds were immersed in 8 µM PI and 2 µM Calcein-AM diluted 1:1000 in serum-free DMEM on different days after printing.After 30 min incubation, the scaffolds were rinsed three times with PBS and the live (calcein-AM, 490 nm) and dead (PI, 535 nm) cells were visualized with fluorescence microscopy.

S-100β immunofluorescence staining
S-100β immunofluorescence staining of 3D bioprinted hydrogel scaffolds was adopted for the detection of the characteristic protein expression in rat Schwann cells.On the 7th day of culture, the scaffolds were crosslinked with 50 mM calcium chloride solution for 5 min, fixed with 4% paraformaldehyde for half an hour, and configured with a closure solution (0.3% Triton X-100 + 5% BSA) and an antibody dilution solution (0.3% Triton X-100 + 1% BSA).After sealing for half an hour, the scaffolds were gently washed thrice with PBS; besides, the scaffolds were immersed in rabbit anti-rat S100β antibody solution (1:500, Abcam, ab52642) at 4 • C overnight.On the following day, the antibody solution was aspirated.The scaffolds were washed thrice gently with PBS.The samples were subject to incubation with the sheep anti-rabbit secondary antibody solution (1:500, Abcam, ab150077) for 1 h and then rinsed thrice with PBS.The samples were then stained for 10 min with DAPI staining solution (Zgb-bio, ZLI-9557) for 10 min and visualized with a fluorescence confocal microscopy.

Cytoskeletal staining
On day 7 of culture, the scaffolds were crosslinked with 50 mM calcium chloride solution for 5 min, fixed with 4% paraformaldehyde on ice for 30 min, permeabilized with a permeabilizing solution (0.2% Triton X-100 in PBS) for 30 min, and gently rinsed with PBS three times.The scaffolds were immersed in ghost pen cyclic peptide staining reservoir solution (1:100, Proteintech, PF00003) for 20 min, the residual liquid was aspirated.In addition, the scaffolds were gently washed three times with PBS.Then, the scaffolds were stained with DAPI staining solution for 10 min and observed using a fluorescence confocal microscope.

PCR array analyses
Cells were cultured on 3D bioprinted scaffolds for 7 d and explored for gene expression (n = 3).The scaffolds were soaked in 50 mM sodium citrate for 10 min and gently stirred to depolymerize the hydrogel.Following de-crosslinking, the cells were subject to centrifugation at 9168 g for 10 min, washed with PBS, and centrifuged again at 9168 g for 3 min.Cells The hydrogel is extruded from the syringe and arranged on a low-temperature platform to form a printed structure.The scaffolds for in vitro and in vivo experiments were obtained using this construction method.The rGO in the scaffolds generated microcurrent under the stimulation of an external pulsed magnetic field that promotes a better functional state in the cultured cells, such as promoting synapse formation, axon extension, neuron survival, and transcription facilitation.Nerve regeneration can be induced after implantation in rats.Alg: alginate, Gel: gelatin, SF: silk fibroin, rGO: reduced graphene oxide.
were lysed with Trizol, and total RNA was extracted with chloroform and isopropanol.Reverse transcription was carried out with the use of the PrimeScript™ RT Reagent Kit (TaKaRa, Japan).Based on the rat neurogenesis PCR array (Wcgene Biotech, Shanghai, China), 92 mRNA related to neuroregeneration were detected and analysed.

In vivo experiments 2.11.1. Establishment of rat sciatic nerve defect model and nerve scaffold implantation
Twenty-four SD rats (male, SPF grade, weighing 200-250 g) were chosen for in vivo experiments.In a random manner, Patients were equally divided into four groups: autotransplantation, rGO catheter, pulsed magnetic field stimulation, and rGO catheter + pulsed magnetic field stimulation.Before the right sciatic nerve was exposed by gluteal muscle incision to isolate a 15 mm-long nerve, an intraperitoneal injection of sodium pentobarbital (30 mg kg −1 ) was provided as anesthesia.The nerve was clipped to create a 10 mm nerve defect area.The clipped nerve was inverted in the defect and sutured with a 9-0 nylon suture for autotransplantation.The muscles and skin were sequentially closed with 4-0 nylon sutures.In all nerve scaffold groups, 10 mm acellular prints were placed in the nerve defect area and closed with a 9-0 nylon suture.The muscles and skin were sutured as described above.Rats that received pulsed magnetic field stimulation were stimulated daily for 20 min in a 2 mT, 50 Hz pulsed magnetic field.Gait evaluations were performed at 2, 4, and 8 w postoperatively in each group.All animals were housed at 20 • C-25 • C under a 12 hour light/dark cycle.All animals were injected intraperitoneally with 105 units of penicillin immediately after surgery.Following the instructions of the Animal Ethics Committee of the Department of Medicine, Peking University, animal care and use were carried out.All rats were sacrificed eight weeks postoperatively.

Analysis of regenerative nerve function
Walking trajectory analysis was conducted by bilaterally applying black ink to the hind limbs of the rats to collect their footprints when they walked across white paper.Toe spread width (TS), paw length (PL), and intermediate toe spread width (IT) were measured on the experimental side (E) and normal side (N), and the sciatic nerve function index (SFI) was calculated: SFI = 109.5(ETS-NTS)/NTS − 38.3 (EPL-NPL)/NPL + 13.3 (EIT-NIT)/NIT-8.8

Electrophysiological and histologic analyses
Electrophysiological analyses were performed on SD rats at 8 w postoperatively.The right sciatic nerve was visualized under anesthesia.A bipolar electrode was fixed to the proximal end of the regenerated nerve to deliver a single electrical signal.Electromyography (EMG) was performed with an electrode implanted in the abdomen of the gastrocnemius muscle.Different latencies and distances between the ends of the stimulus were recorded, and nerve conduction velocity (NCV) and CMAP were measured.We repeated the experiment for five times.Regenerated nerves were immediately isolated midway between the distal and proximal end after electrophysiological testing for TB and immunohistochemical staining while the slices were obtained from the distal part of the sample.For TB staining, all nerve samples were fixed with 4% PFA for 48 h and treated with 2% osmium tetroxide (Sigma-Aldrich, USA).TB staining was performed after embedding the specimens.The number of the myelinated nerve fibers in the middle of the regenerating sciatic nerve were observed under a light microscope and calculated.Immunohistochemical staining of GAP43 (Proteintech, 16971-1-AP), NF-200 (Proteintech, 18934-1-AP), and S-100β (Proteintech, 15146-1-AP) was carried out on the midsection of the nerve grafts.Antigen repair was carried out in sodium citrate buffer at 95 • C for 20-25 min.Nonspecific antigens were blocked with 1% BSA.Sections were subject to incubation with primary antibody at 4 • C overnight, rinsed in PBS, and stained with a secondary antibody at RT for 1 h.The samples were rinsed again, stained with DAB, and re-stained with hematoxylin.Bilateral gastrocnemius muscles were obtained from the bone attachments 8 w after surgery.The middle muscle was fixed in 10% formalin for hematoxylin and eosin (HE) and Masson staining.

A suitable concentration of ingredients is helpful for porous structure and cell survival
SF has been widely used in the synthesis of hydrogels because it enhances their mechanical strength for a variety of tissue engineering applications and has a positive effect on nerve regeneration [38,40,41].In the SEM images, each group of hydrogels presented a porous and interconnected structure, whose morphology and size changed significantly after adding 15% SF (figure 2(a)).As reported previously, porous structures are conducive to cell growth and vascularization.Generally, the porosity increased slightly from 0% SF to 12.5% SF in the hydrogels tested but declined 50.98% when the SF concentration increased to 15% (figure 2(b)), with the highest porosity in the 12.5% SF group and the lowest in the 15% SF group.However, these differences between the groups were not of statistical significance.There was no significant difference in average pore size of the material with different SF content, but the standard deviation of pore size was significantly different among the groups.Our observation of the SEM images of the hydrogels revealed that the pore size of the 15% SF group was very variable, with the largest variance in pore size (165.14± 136.42 µm) observed in the 15% SF group (figure 2(c)), which was not conducive to the even distribution of cells.The compression modulus of different hydrogels was tested, resulting in a slight increasement from 0% to 12.5% SF, reaching a maximum of 60.0 ± 9.79 kPa, with a sharp drop when the SF concentration went to 15% (figure 2(d)).Therefore, 12.5% SF has a favorable mechanical strength.Meanwhile, rGO concentration was screened with a CCK8 cytotoxicity test, which showed that 0.4% rGO significantly reduced the cell proliferation curve, indicating an obvious cytotoxicity at 0.4% (figure 2(e)), so 0.2% concentration with no prominent toxicity was selected for further experiment.

Impact of rGO on material chemical composition
The FTIR analysis reported a hydroxyl bond (-OH) of sodium alginate (3291 cm −1 ), CH group stretching vibration (2953 cm −1 ), C=O group stretching vibration of amide I (1631 cm −1 ), NH group curvature vibration of amide II (1547 cm −1 ), and C-N group stretching vibration of fatty amide (1241 cm −1 ) in the 0% SF hydrogel.When SF and rGO were added, all the above peaks shifted to a lower range.After the addition of rGO, the light transmittance of the hydrogel decreased, and the characteristic epoxy C-O stretching vibration observed at 1241 cm −1 weakened (figure 2(f)).This analysis indicated a simple mixture of rGO and the hydrogel with no chemical bond formation.

rGO composite hydrogel is a shear-thinning and biocompatible material
To understand how the hydrogels form fine structures via 3D printi ng, their viscosity and shearthinning properties were measured versus temperature.The shear-thinning property allows the shear viscosity to decrease as hydrogels pass through the print nozzle, thus preventing the pulling and dragging of the printed hydrogels at the printing position, which can deform the entire printed structure.A decrease in the viscosity of each hydrogel was observed with an increase in shear rate (figure 2(g)).
Temperature is a determinant of hydrogel application during 3D bioprinting, and in our experiments, the viscosity of the different gels decreased as temperature went up (figure 2(h)).The thixotropic experiments with each group showed that their viscosities decreased significantly as the shear rate was raised to 100 s −1 ; nevertheless, when the shear rate recovered to 0.1 s −1 , the hydrogels of each group returned to the first stage viscosity (figure 2(i)).
These results indicated that all hydrogels were materials of high thixotropy, which was significant for their application in extruded 3D bioprinting, and the appearance of rGO slightly increased the overall viscosity of the hydrogel.Additionally, the Alg-gel-SF-rGO hydrogel extract made no obvious impact on the proliferation of RSC96 cells on days 1, 3, 5, and 7 (figure 2(j)), suggesting that the cells had good compatibility with the hydrogel and did not show cytotoxicity.

rGO composite hydrogel has good printability
The syringe temperature for the hydrogels were tested from 25 • C to 30 • C, corresponding to different gel status from thick (25 • C), proper mobility (around 28 • C) to water-like (30 • C), where the shape of the printed body was incomplete, well-shaped, and blurred into the surrounding substrate (figure 3(a)).Generally, the inner diameter of the printing nozzle was 300 µm, room temperature was 16 • C, syringe of hydrogel was kept at 28.3 • C, extrusion speed was defined at 0.15 ml min −1 , and the nozzle speed was 3.5 mm s −1 .The structure of nerve scaffold was designed and split into layers (figure 3(b)).After 3D bioprinting, the printed structures maintained their original geometry with clear edges, and no blockages occurred.No obvious collapse of the material was observed, and the diameter of the printed body was maintained at 4 mm and the height at 8 mm, meanwhile, the morphology of the designed model was maintained (figure 3(c)).

Appropriate magnetic field intensity is beneficial for cell viability
The RSC96 cells were treated under the PEMF generating device (figure 3(d)) with pulsed magnetic fields of 1, 2, 4, and 8 mT for 20 min.CCK8 was adopted to detect cell viability 24 h after magnetic stimulation.The cell proliferation rate of the blank control group was set at 100% 24 h after stimulation, according to which the other rates were calculated.The cell viability after 2 mT pulsed magnetic field stimulation was the highest, showing statistical significance with the control cells (figure 3(e)); therefore, the stimulation intensity was set to 2 mT for further experimentation.The hysteresis loop showed that the rGO-gel gained paramagnetism similar to that of rGO alone (figure 3(f)).

Combination of magnetic field and rGO composite hydrogel promotes S-100β protein expression
S-100β is a typical cytoplasmic protein marker for Schwann cells associated with the proliferation and functional protein expression of neural cells.S-100β is also stably expressed in RSC96 cells.RSC96 cells seeded into 3D bioprinted gelatin-alginate neural scaffolds for 7 d were shown to stably express S100β [39].On d 7 after printing, S-100β expression was investigated in each group of bioprinted scaffolds.The MF + rGO group had more S-100β expression than the other groups or the control, implying that neither the 3D bioprinting process and 3D cell culture nor rGO and pulsed magnetic field stimulation inhibited S-100β expression (figures 5(a)-(d)).Quantification of the S-100β expression rate indicated that it was promoted by the simultaneous impact of the MF and rGO (figure 5(e)), which indicated that cell function was enhanced by MF + rGO stimulation.

Combination of magnetic field and rGO composite hydrogel improves RSC96 cell adhesion
Cytoskeletal staining helps observe the morphological characteristics of the printed cells.Cells printed directly onto the rGO-free hydrogel showed a tendency to grow into clusters that were poorly stretched and had no obvious polarity (figure 5(f)).By adding rGO to the hydrogel or providing daily pulsed magnetic field stimulation after printing, some cells exhibited long spindle shapes (figures 5(g) and (h)).When rGO hydrogel prints were stimulated with pulsed magnetic fields, the vast majority of the cells exhibited wellstretched, long, spindle-shaped features (figure 5(i)).The cells were much longer in MF + rGO group than those in the other three groups, with the shortest ones on the control group, showing statistical difference (figure 5(j)).The extensional growth of Schwann cells contributes to the formation of axonal structures and facilitates peripheral nerve regeneration more than clustered growth.These results suggested that combination of magnetic field and rGO was helpful for Schwann cells morphogenesis and may enhance nerve extension aided by Schwann cells.

Combination of magnetic field and rGO composite hydrogel enhances nerve regeneration gene expression
To explain the mechanism of neural regeneration in RSC96 cells in each group of biological materials, a PCR microarray was used for identifying 92 genes associated with neural regeneration in each group on d 7 (figure 6(a)).The results showed that the expression levels of ACHE (figure 6 were higher in the rGO + MF group than in the other groups.These genes are closely related to synapse formation, axon extension, promoting neuronal survival.These results showed that MF + rGO enhanced transcriptional processes and promoted neuronal differentiation and axonal extension.

Combination of magnetic field and rGO composite hydrogel supports sciatic nerve regeneration In vivo
We assessed the recovery of sciatic nerve function after NGC implantation using a gait analysis in rats.In general, a decrease in sciatic nerve function causes toe extension and intermediate toe extension to narrow.Footprint analysis at 8 w (figure 7(e)) showed that toe extension and intermediate toe extension were greater in the autograft and rGO + MF groups than in the rGO and MF groups, with no significant difference found between the autograft and rGO + MF groups.After complete sciatic nerve dissection, the SFI was significantly lower in all animals.At 2 w post operation (figure 7(f)), the SFIs in the magnetic field (−85.54 ± 6.66) and the rGO (−85.63 ± 1.47) groups were still low, while the SFIs in the rGO + MF group (−73.48 ± 4.47) showed a rebound.At 4 w post-operation, the SFIs in the rGO + MF group (−69.65 ± 0.27) were notably higher than those in both the magnetic field and rGO groups.At 8 w, the SFIs in the rGO + MF group still showed a significant improvement of −46.11 ± 13.80, and the recovery in the rGO group was similar to that in the MF group (figure 7(f)).
To investigate the function of sciatic nerve regeneration after NGC implantation in more detail, we performed EMG in postoperative animals at 8 w (figure 7(g)).The autograft (5.47 ± 0.78 mV) and rGO + MF (1.66 ± 0.65 mV) groups were significantly higher than the rGO (0.56 ± 0.18 mV) and MF (0.63 ± 0.35 mV) groups (figure 7(h)).In addition, NCV was significantly higher in the autograft (41.77 ± 5.21 m s −1 ) and rGO + MF (38.27 ± 5.08 m s −1 ) groups than in the MF group (30.99 ± 10.15 m s −1 ), while NCV was significantly lower in the rGO group than in the other groups at 8 w (figure 7(i)).Together, these results indicated a better electrophysiological recovery in MF + rGO group than that in MF or rGO group.
Sciatic nerve injury can lead to loss of innervation of the target muscles and eventually to muscle atrophy.Eight weeks after surgery, muscle morphology was observed using HE (figure 8(a)) and Masson trichrome staining (figure 8(b)), with all groups showing recovery from atrophy.Regenerated nerves were evaluated using histological analysis.Semi-thin toluidine blue staining of the middle portion of the MF + rGO group revealed many well-myelinated axons (figure 8(c)) when compared with that of the MF and rGO groups.
The levels of GAP43, NF200, and S-100β in the four groups were detected by immunohistochemistry (figures 9(a)-(c)).The MF group showed the lowest expression of GAP43, NF200, and S-100β of the tested groups.We also found an obvious elevation in the expression of GAP43, NF200, and S-100β in the autograft and rGO groups in relative to that in the MF or rGO groups.The differences in GAP43, NF200, and S100 expression in the rGO + MF group were statistically significant when compared with those in the MF and rGO groups (figures 9(d)-(f)).

Discussion
Better nerve repair was achieved using 3D-printed SF-collagen scaffolds to repair damaged spinal cords in rats [42].Therefore, we combined gelatinsodium alginate hydrogels with SF to prepare compliant hydrogels.The rGO was included as a conductive component in the hydrogels, and NGCs and (h) Th.Data are presented as mean ± standard deviation (n = 3).1, 2, 3: control group, cells were cultured without rGO or magnetic field.MF: magnetic field group, cells were cultured without rGO but under magnetic field stimulation.rGO: rGO group, cells were cultured with rGO but without magnetic field.MG: MF + rGO group, cells were cultured with both rGO and magnetic field stimulation.* p < 0.05, * * p < 0.01.
were prepared using 3D bioprinting to produce cell-containing 3D structures favorable for the spatial distribution of cells and, by pre-designing them, to obtain a simulation of the outer membrane structure with an internal growth channel structure to facilitate the growth of neurovascular bundles.rGO is widely used in 3D bioprinting of various cells because it promotes functional expression in cells [43].A previous study has reported the magnetism of different graphene derivatives, with rGO exhibiting paramagnetism, indicating that it could enhance the impact of an external magnetic field.The SQUID test of the rGO hydrogel showed paramagnetism similar to that of rGO, further indicating the probability of an analogous effect when co-cultured with cells.
We found that rGO content (0%, 0.1%, 0.2%, and 0.4%) affected cellular activity at different culture times, with a high rGO ratio showing significant cytotoxicity.This result was consistent with previous studies [44][45][46]; however, we found that rGO and MF together accelerated nerve regeneration.Therefore, we selected an appropriate rGO dose that synergized with MF in the range of non-cytotoxic concentrations.RSC96 cells were treated with leachate, exerting no influence on cell proliferation.This suggested that the material concentration was biocompatible with RSC96 cells.
Once the ingredients of the bioink were confirmed, their viscosity versus temperature and shear rate were measured, which showed that the viscosity of each hydrogel reduced with the increase of shear rate.This shear-thinning property allowed the shear viscosity to decrease as the hydrogels passed through the print nozzle, achieving a fluent printed line and preventing the printing nozzle from being obstructed.In our experiments, the viscosity decreased when different gels were heated, with a relatively low optimal temperature for the syringe and bioprinting compared to the culture temperature, which maintained cell metabolism at a low level during the nutrientpoor printing process and thus, also maintained cell viability.Thixotropy tests showed that all the hydrogels were highly thixotropic materials.This was vital for extruded 3D bioprinting.
Magnetic fields have been used in osteogenicand neuroregeneration-related studies.In different studies and application conditions, magnetic fields have only been reported to have positive effects on regeneration within a certain intensity range, with strong magnetic fields over 5 mT leading to cell apoptosis [47,48].Therefore, cell survival was tested after stimulation with different field strengths, which led to the selection of a 2 mT magnetic field strength for our experiments.Although the cell proliferation rate was measured 24 h after magnetic stimulation, no significant changes in cell survival were observed during live/dead staining after 7 d, which was consistent with the literature [49,50], possibly because the prescreened material concentration and magnetic field intensity had no cytotoxic or survival-facilitation effects.Moreover, the similar proliferative status of RSC96 cells in the MF stimulation and control groups indicated the biocompatibility of the 2 mT magnetic field.
S-100β is a cytoplasmic soluble calcium-binding protein that exerts a vital role in growth, cell signaling, motility, and metabolism.Previous studies demonstrated that gelatin-sodium alginate scaffolds affected S-100β expression [39,51].In this study, although S-100β expression was not directly affected by rGO-assisted pulsed magnetic field stimulation, we observed a significant upregulation of S-100β under the influence of magnetic field and rGO, suggesting the improvement in cell functions.More work is required to understand how S-100β protein expression is regulated under rGO-assisted magnetic field stimulation.
Compared with the round cell morphology observed in the control group, the cells were elongated in the MF and rGO groups.Moreover, the cells in the MF + rGO group appeared spindleshaped, indicating improved cell adhesion.Lim et al [52] combined rGO with PEMFs to induce MSC differentiation and suggested that this combined treatment enhanced cell viability by increasing adhesion and ion transport.These results provided evidence that the synergistic effect of rGO and PEMFs produced new cellular responses, such as enhanced cell proliferation and intercellular communication through microcurrents.In the present study, we also found that administering rGO or PEMFs alone did not significantly change AChE, DRD2, TH, HEYL, BMP2, NRP2, and TBP expression.However, the application of both increased the expression of nerve regeneration-related genes, which is consistent with previous report [53].
Studies have investigated the functions of these genes in nerve regeneration; the application of AChE to the growth medium of hippocampal cell cultures induced neurite protrusion extension and synapse formation [54,55].DRD2 regulates synaptic pruning through a cell-autonomous mechanism involving mTOR signaling activation [56].TH encodes a key enzyme involved in catecholamine synthesis that is important in the physiology of adrenergic neurons, while HEYL encodes a transcription factor that promotes neuronal differentiation in neural precursor cells both in vivo and in the embryonic brain [57][58][59].HEYL expression can be increased by BMP signaling, which may be mediated by Notch1 signaling.BMP2 encodes a secreted ligand that binds a variety of TGF-β receptors, contributing to the recruitment and activation of SMAD family transcription factors regulating gene expression and whose neurotrophic effects are relied on the activation of the BMP receptor (BMPR) and the SMAD 1/5/8 signaling pathway [60].TBP encodes TATA-binding proteins that participate in basal transcription, act as co-activators, participate in promoter recognition, and modify general transcription factors to promote complex assembly and transcription initiation.NRP2 is a member of the neuropilin receptor family encoding a transmembrane protein that interacts with vascular endothelial growth factor (VEGF).NRP2/VEGF controls somatic migration, axon formation, and synaptogenesis [61] and helps direct the neural crest cell precursors of neurons and glial cells in the peripheral nervous system [62,63].
It could be hypothesized that under the mutual effect of rGO and PEMFs, AchE, Drd2, and TH function to regulate the biosynthesis and levels of neurotransmitters and participate in synaptic chemotransmission and synapse formation.Another part of the synergistic effect of rGO and PEMFs was the promotion of HEYL expression via BMP2 activation of its downstream effector, Notch1 [64,65], which promotes adult neural differentiation.We noticed a certain degree of elevated Notch1 expression in MF + rGO group, which may hold responsible for the upregulation of HEYL.In addition, Nrp2 expression is upregulated to promote axonal growth by promoting Schwann cell migration through the greater acceptance of VEGF signaling [66].Further studies should be conducted to confirm this signaling pathway.
PEMFs have been used for spinal cord and sciatic nerve regeneration.Investigations in spinal cord injury suggest that magnetic stimulation can reverse synaptic function at the injured site and protect neurons against degeneration and necrosis by improving the release of neurotransmitters and neurotrophic factors, such as AchE and BDNF [67,68].Moreover, rGO was reported to promote neural differentiation by improving cell adhesion and cytokine secretion [24,69].In this study, the rGO hydrogels were combined with PEMFs.The gel scaffold was transplanted into the injury site of sciatic nerveinjured rats with continuous additional PEMF stimulation.This therapy significantly improved gait recovery and promoted nerve regeneration in rats.Electrophysiological analysis also showed that the autologous nerve graft group had the best CMAP recovery, far exceeding the other three treatment groups.However, in the gait analysis, we found that the rats with the best recovery on the affected side were in the rGO + MF group.We speculate that rGO + MF treatment plays an important role in the regeneration of neuromuscular function.Several genes related to axon extension and synaptic junctions were upregulated under the combination of magnetic field and rGO composite hydrogel, which agrees with a previous study that reported that MF improved muscle regeneration and function by preventing muscle atrophy and inducing hypertrophy [21], which could also explain the differences between electrophysiological and functional recovery.Conversely, the MF group had better functional recovery than the rGO group at week 8, while the rGO + MF group had the best functional recovery, which was significantly different from the rGO group but not from the MF group, suggesting that continuous stimulation with PEMFs had a better effect on nerve function recovery and was facilitated by the presence of rGO.
S-100β expression reflects Schwann cells proliferation and functional status [70].We showed that S-100β expression was induced by rGO-assisted daily magnetic stimulation, improving the functional recovery of the sciatic nerve after 8 w.GAP43 is involved in regulating neurite outgrowth during nerve regeneration [71].We found that GAP43 expression was upregulated in the MF + rGO group, suggesting a positive effect on axonal extension.Moreover, the high expression of NF200 in the MF + rGO group indicated a high rate of vessel formation [72], which supports nutritional supplementation for rapid nerve growth.
In summary, the present study found that the combination of PEMF stimulation and rGO could promote RSC96 cell function and adhesion, with the recovery effect on rat sciatic nerve generally comparable to that of autograft during in vivo experiments, and even functionally slightly superior, which provides a path to address PNI.

Conclusion
In this study, a nerve scaffold consisted of gelatin, sodium alginate, SF, and rGO was successfully constructed using 3D bioprinting, and PEMF was applied to the scaffold system non-invasively.The printed construct could interact with PEMF to promote RSC96 cell adhesion and the expression of AchE, Drd2, HEYL, BMP2, NRP2, TBP, and TH.Preliminary results showed that the interaction between the 3D bioprinted rGO composite scaffold and PEMF provided a suitable microenvironment for RSC96 and promoted better biological functions.This combined effect improved sciatic nerve regeneration in rats.Therefore, the gelatin-alginate-SF-rGO composite hydrogel is expected to become a candidate material for neural tissue engineering and can be applied to more scenarios in the field of nerve regeneration combined with PEMF stimulation.

Figure 1 .
Figure1.Schematic representation of nerve scaffold fabrication and characterization.The hydrogel is extruded from the syringe and arranged on a low-temperature platform to form a printed structure.The scaffolds for in vitro and in vivo experiments were obtained using this construction method.The rGO in the scaffolds generated microcurrent under the stimulation of an external pulsed magnetic field that promotes a better functional state in the cultured cells, such as promoting synapse formation, axon extension, neuron survival, and transcription facilitation.Nerve regeneration can be induced after implantation in rats.Alg: alginate, Gel: gelatin, SF: silk fibroin, rGO: reduced graphene oxide.

Figure 2 .
Figure 2. Selection and characterization of hydrogel components.(a) SEM images of Alg-Gel hydrogel after adding different concentrations of SF.(b) Porosity and (c) average pore size of the material with different SF content.(d) Compression modulus of different SF content hydrogels.(e) The rGO concentration screening test.(f) FTIR detection of different hydrogels.(g) The shear-viscosity curve, (h) temperature-viscosity curve, and (i) time-viscosity curve of the composite hydrogels.(j) The cytotoxicity test of the extract of the composite hydrogel with the determined components (0.2% rGO, 12.5% silk fibroin, 8% gelatin, 2% alginate), control: cells that were not treated by the extraction solution.

Figure 3 .
Figure 3. Material printability and magnetic field intensity test.(a) Exploration of hydrogel printing conditions.(b) Computer-aided design of the printed path of the neural scaffold model.(c) Composite hydrogel with and without rGO was used to print the 3D structure of the nerve scaffold.(d) Physical drawing of the pulsed magnetic field generating device.In the cell experiment, the culture dish was placed in the center of the two coils.(e) Screening of the magnetic field intensity.(f) Magnetic moment detection of the composite hydrogel.* * * p < 0.001.

Figure 6 .
Figure 6.PCR array of genes involved in nerve regeneration.(a) Heat map of the relative expression of 92 genes in different culture conditions 7 d in vitro.The differentially expressed genes were (b) Ache, (c) Heyl, (d) Drd2, (e) Nrp2, (f) Bmp2, (g) Tbp,and (h) Th.Data are presented as mean ± standard deviation (n = 3).1, 2, 3: control group, cells were cultured without rGO or magnetic field.MF: magnetic field group, cells were cultured without rGO but under magnetic field stimulation.rGO: rGO group, cells were cultured with rGO but without magnetic field.MG: MF + rGO group, cells were cultured with both rGO and magnetic field stimulation.* p < 0.05, * * p < 0.01.

Figure 7 .
Figure 7. Animal experiments and functional recovery after nerve damage using hydrogel scaffolds.(a) The right sciatic nerve in rats was cut to cause a 10 mm defect.In the defect area: the autogenous nerve was (b) anastomosed by inverting the nerve, (c) implanted with rGO composite scaffold, and (d) implanted with no rGO scaffold, followed by wound closure.(e) Functional recovery of sciatic nerves was measured using walking track analysis after 8 w.(f) SFI functional scores at 2, 4, and 8 w after nerve damage and implantation.(g) The action evoked electromyography (EMG) of the affected side 8 w after the operation, and (h) CMAP and (i) NCV on the affected side 8 w after the operation were analyzed and measured.CMAP: compound muscle action potential, NCV: nerve conduction velocity.* p < 0.05, * * p < 0.01.

Figure 8 .
Figure 8. Histological examination of the affected side 8 w after the operation.(a) HE and (b) Masson staining of the affected gastrocnemius muscle, and (c) nerve TB staining after repair of the defect area.(d) The muscle fiber area ratio of the affected gastrocnemius muscle was calculated to evaluate the atrophy recovery of the affected muscle.(e) The density of the myelinated nerves was assessed to evaluate the recovery of the damaged nerve.And (f) the diameter of myelinated axons.Scale bar: 20 µm.* p < 0.05, * * * p < 0.001.Scale bar: 50 µm.