Animals

Newborn (1 d) and adult female Sprague Dawley (SD) rats about 200 g were purchased from the experimental animal center of Nantong university. The anesthesia of adult female SD rats was under intraperitoneal injection of 3% sodium pentobarbital solution according to 30 mg/kg of body weight. For groups and numbers for the animal studies please refer to previous study [24]. All experimental protocols were approved by the Administration Committee of Experimental Animals, Jiangsu Province, China, in accordance with the guidelines of the Institutional Animal Care and Use Committee, Nantong University, China (Inspection No: 20190225-004).

Isolation of SKPs and their induced differentiation into SKP-SC

SKPs were isolated from the backs of 1 to 3-day-old Sprague Dawley and green fluorescent protein (GFP)-transgenic rats as previously reported with minor modification [25]. Backs were firstly waxed and cleaned with ethanol before a patch of skin was removed. The skin was cleaned of fat, fascia, and blood vessels and minced into small pieces a few millimeters in size with razor blades. Minced skin was digested in collagenase type XI (1 mg/ml), inverting at 37 ºC for 1.5 to 2 h. Intermittent mechanical mashing was done with a 10 ml pipette every 40 min. Digested skin was triturated with a P1000 pipette and diluted with DMEM to stop digestion. The skin solution was filtered by a 40 μm cell strainer (BD Falcon, Bedford, MA, USA) to remove undigested skin and collect only single cells. Single cells were grown in “SKPs proliferation media” which contained DMEM/F12 (3:1), 1% penicillin/streptomycin, 2% B27 supplement (Invitrogen, Carlsbad, CA, USA), 20 ng/ml EGF, and 40 ng/ml FGF-2 (BD Bioscience, San Diego, CA, USA). To prevent contamination from dissection, 1 µg/ul Fungizone Antimycotic was added. The cells were seeded from 25 000 to 50 000 cells/ml media in a vented-cap flask. SKPs were fed with SKPs proliferation media every 5 days and passaged when floating spheres turned big. To passage SKPs, the spheres were separated from the conditioned media by centrifugation. Conditioned media filtered through a 20 μm syringe filter was saved for feeding. Spheres were digested with 1 mg/ml collagenase for 10 min in a 37ºC water bath, followed by trituration with P1000 and P200 pipettes. Dissociated spheres were resuspended in SKPs proliferation media that contained half DMEM/F12 (3:1) and half conditioned media.

SKP-SCs were differentiated from passage three SKPs as previously described [25, 26]. SKPs were dissociated to single cells, as during passaging. Single cells were resuspended in “SKP adherence media” containing DMEM/F12 (3:1), 1% penicillin/streptomycin, 10% fetal bovine serum (FBS, HyClone, Logan, UT, USA), 1% B27 supplement, 1% N2 supplement, 20 ng/ml EGF, and 40 ng/ml FGF-2. Cells were plated at 25 000 cells/ml in petri dishes. After 3 days, the media was changed to SC proliferation media and was replenished every 3–4 days. SC colonies that could be seen after 2–3 weeks were isolated with cloning cylinders (Corning). When confluent, cells were expanded and passaged with trypsin-EDTA (Thermo Fisher Scientific, Carlsbad, CA, USA).

Immunocytochemistry

To check for the expression of typical SC markers, immunocytochemistry was performed. Cells were fixed in 4% paraformaldehyde (PFA) for 10 min or in 100% methanol (chilled to -20 ºC) for 5 min at -20ºC. Cells were blocked with 5% normal goat serum (NGS) and permeabilized with 0.1% Triton-X for 30 min. Primary antibody was added to the samples for 2 h at room temperature or overnight at 4 ºC, followed by a secondary antibody added for 1 h at room temperature. The samples stained with DAPI ( 1:5000, Sigma) was added for 5 min. The following primary antibodies were used at the stated dilutions: chicken anti-P0 (1:500, Aves Labs), mouse anti-S100β (1:500, Sigma), rabbit anti-GFAP (1:500, Dako), mouse anti-p75NTR (1:500, Chemicon), mouse anti-NF200 (1:4000, Abcam), and rat anti-MBP (1:500, Millipore). The following secondary antibodies were used at the stated dilutions: Alexa 555 goat anti-chicken, Alexa 555 goat anti-rabbit, Alexa 488 goat anti-mouse, Alexa 555 goat anti-mouse, (all 1:1000, all Invitrogen).

Myelination of DRG axons in vitro

In vitro myelination of SKP-SCs was assessed using the DRG myelination assay [27]. DRGs were isolated from E14.5 SD rats and dissociated with 0.25% trypsin (Gibco). Dissociated ganglia were plated at 300 000 cells/ml on matrigel and poly-D-lysine (Sigma)- coated chamber slides in “neuron growth media” composed of 2% B27 supplement, 1% GlutaMAX supplement (Gibco), 50 ng/ml NGF, and 1% penicillin/streptomycin in Neurobasal media (Gibco). Media was changed every other day. To obtain pure neuron cultures, cytosine arabinoside was added to the media for one media change. One week after obtaining pure neurons, SKP-SCs were added in media containing 1% ITS supplement (Sigma), 1% GlutaMAX supplement, 0.2% bovine serum albumin, 4 g/l D-glucose (Sigma), 50 ng/ml NGF, and 1% penicillin/streptomycin in Basal Media Eagle (BME, Invitrogen). The SKP-SCs were grown in DMEM with 10% FBS (instead of in SC proliferation media) for three days. Media was changed every other day. After 6–8 days, SCs were differentiated into a myelinating state by changing to “myelination media” containing 1% ITS supplement, 1% GlutaMAX supplement, 15% FBS, 4 g/l D-glucose, 50 ng/ml NGF, 1% penicillin/streptomycin, and 50 ug/ml L-ascorbic acid (Sigma) in BME. Myelination media was exchanged every other day for 10–14 days. Immunocytochemistry was performed on the cultures at the end point for MBP, P0, MAG, and NF200.

Construction of TENG in vitro

The chitosan/silk fibroin-fabricated neural scaffold was prepared as previously reported [28]. According to the Chinese patent ZL 0110820.9 A, chitosan conduit was fabricated of chitosan gel by injection molding, and silk fibers (Bombyx mori silk) were processed by degumming in aqueous boiling Na2CO3 solution. Then, a chitosan/silk fibers (SFs)-fabricated neural scaffold was obtained by inserted 120 of SFs in a chitosan conduit. SKP-SCs were seeded to the neural scaffold at a cell density of 1 × 107 cells/ml. The culture was transferred to a perfusion rotatory cell culture system (RCCMax) (Synthecon Inc, Friendswood, TX) 24 h later to facilitate continuous incubation for seven days with the addition of fresh ascorbic acid (50 µg/ml, Sigma, St. Louis, MO, USA) and 15% FBS (Fig. 1d) to stimulate ECM secretion. After washing with PBS, a SKP-SCs seeded TENG was fabricated for further surgery.

Sciatic nerve surgical procedure

After anesthesia, shaving and disinfection, a skin incision and separation of nearby muscles in the left lateral thigh was carried out to expose sciatic nerve were in adult female SD rats. To obtain a 10 mm long nerve defect after retraction of the nerve stumps, a 6 mm long segment of sciatic nerve was transected and removed. All rats were randomly divided into 4 groups as TENG, autograft, scaffold, and sham groups. The sciatic nerve defect was separately bridged via 3 types of nerve grafts: (1) A TENG prepared by 3D culturing SKP-SCs to a chitosan/SFs neural scaffold in vitro; (2) an autologous nerve graft prepared by a reversed nerve segment; (3) a chitosan/SFs neural scaffold consisting of a hollow chitosan conduit (i.d. 2.0 mm) and 120 of SFs (12 mm long, diameter 8 μm). Moreover, rats obtaining a sham surgery were the control group that underwent the same procedures without injuring the sciatic nerve.

Behavioral analysis

The CatWalk XT 9.0 gait analysis system (Noldus, Wageningen, Netherlands) was used to assess functional recovery of motor skills. 4, 8, and 12 weeks after nerve grafting, animals (n = 10) in 3 groups were placed on the right side of a runway consisting of a glass surface and black plastic walls. The animals were motivated to traverse the runway toward the left end where food pellet rewards were located. The locomotion of rats was recorded. The evaluation system is based on a combination of posture, hind limb movement, hind limb force, and joint motion. The sciatic function index (SFI) value was calculated by the following formula: SFI = 109.5(ETS-NTS)/NTS-38.3(EPL-NPL)/NPL + 13.3(EIT-NIT)/NIT-8.8. TS is the total toe spread, IT is the intermediate toe spread, and PL is the footprint length. The label N refers to the contralateral uninjured (normal) side, and the label E refers to the injured side. SFI value fluctuates from 0 to -100, with 0 corresponding to normal function and − 100 to complete dysfunction.

Electrophysiological assessment

12 weeks after surgery, the rats (n = 12) in each group were subjected to electrophysiological tests as described in the literature [29, 30]. Briefly, the sciatic nerve on the injured side was re-exposed under sodium phenobarbital anesthesia. Electrical stimuli were applied to the proximal and distal nerve stumps and the compound muscle action potentials (CMAPs) were recorded on the belly of target gastrocnemius muscle. The conduction velocity of motor nerves was calculated by dividing the CMAP amplitude by the distance between two stimulation sites. The CMAPs recorded on the contralateral uninjured side was used for normalization.

FluoroGold retrograde tracing

FluoroGold (FG) retrograde tracing was performed according to the protocols described previously [29]. Briefly, after the sciatic nerve on the injured side was re-exposed under anaesthesia, 100 µl of 5% FG solution (Fluorochrome Inc., Denver, CO) was injected into the regenerated nerve trunk 10 mm from the distal end followed by closure of the incision. After 2 weeks, rats were transcardially perfused with saline and 4% (v/v) paraformaldehyde in 0.1 M PBS. Spinal cord segments at L6, L7, and S1, along with the corresponding DRGs, were excised, and cut into 30-µm-thick (for spinal cords) and 20-µm-thick (for DRGs) longitudinal sections on a cryostat. This was followed by observation under a DMR fluorescent microscope (Leica, Wetzlar, Germany) with ultraviolet illumination and photography under a confocal laser scanning microscope (TCS SP2, Leica). The percentage of FG-labeled DRG sensory neurons was calculated by dividing the FG-positive cell number by the total neuron number. The number of FG-labeled spinal cord motoneurons was directly counted.

Histological assessment and morphometric analysis of regenerated nerves and target muscles

After the functional evaluations described above, animals were transcardially perfused. The transverse section of the regenerated nerve in the distal end was subjected to Meyer’s modified trichrome staining, immunofluorescence staining, and transmission electron microscopy as described previously [31]. Meyer’s modified trichrome staining was conducted before visualization and photography under light microscopy (Axio Imager 2, ZEISS). For immunofluorescent double-staining, rabbit anti-S100β polyclonal antibodies (1:50 dilution) and mouse anti-neurofilament-200 (NF200) monoclonal antibodies (1:200 dilution, both antibodies were from Sigma) were applied to nerve sections and allowed to incubate overnight at 4 °C. The labelling was finished with a secondary antibody (Goat anti-Mouse IgG-Alex-488, 1:500 and Donkey anti-Rabbit IgG-Cy3, 1: 1000) at 4 °C overnight and nerve sections were observed under a confocal laser scanning microscope (TCS SP2, Leica).

Moreover, the gastrocnemius and anterior tibial muscles on the injured side and contralateral uninjured side were harvested from anesthetized rats and weighed to calculate the wet weight ratio (the wet weight of muscle on the injured side/the wet weight of muscle on the uninjured side). The muscle samples were harvested from the mid-belly of the gastrocnemius on the injured side, fixed in 4% paraformaldehyde, and cut into transverse Sect. (10 μm thick) to undergo Masson’s trichrome staining and observation by light microscopy. The morphometric analysis was performed as previously reported. The longitudinal sections of the gastrocnemius muscles on the injured side were subjected to double immunostaining with α-bungarotoxin (1:800 dilution, Sigma) and mouse anti-NF200 monoclonal antibody (1:600 dilution, Sigma), followed by observation under a light macrograph. The contralateral, uninjured muscle samples were used as a control.

For transmission electron microscopy (TEM), nerve specimens were fixed in pre-cooled 2.5% glutaraldehyde for 3 h, post-fixed with 1% osmium tetraoxide solution for 1 h, washed, dehydrated, embedded in Epon 812 epoxy resin, and cut into ultra-thin sections of 60 nm thickness for staining with lead citrate and uranyl acetate. The stained sections were observed under a transmission electron microscope (JEOL Ltd., Tokyo, Japan) and images were taken from 20 random fields of each section. The number of myelin sheath layers, the thickness of the sheaths, and the diameter of myelinated nerve fibers were quantified using Q550 IW image analysis system (Leica Imaging Systems Ltd., Cambridge, England) and the Leica QWin software package as described previously.

Sample preparation and microarray

Graft segments with both nerve stumps (0.2 cm) and nerve segments in the shame group were collected at 1 d, 4 d, 7 d, 1 w, 2 w, 3 w, 4 w, 8 w, and 12 w post-surgery. According to the manufacturer’s instructions, total RNAs were extracted by using Trizol (Life technologies, Carlsbad, CA). Agilent Bioanalyzer 2100 (Agilent technologies, Santa Clara, CA) and Nanodrop ND1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE) were applied to determine the RNA quality of each sample. Microarray analysis was executed by an Agilent Microarray Scanner (Agilent Technologies) and the raw data compiled with Agilent feature extraction software. National Engineering Center for Biochip at Shanghai (China) helped us to execute all steps from RNA amplification to the final scanner output. The mixed samples were obtained from three independent animals for each group. Each group included three biological replicates.

Bioinformatics analysis

The Venn diagrams was done by a free online tool [32]. The dynamic expression level of the selected genes clustered in the heatmap were done by a free online platform for data analysis and visualization (https://www.bioinformatics.com.cn). For weighted correlation network analysis (WGCNA), the R package was used to find sets of co-regulated genes, defined as a module [33]. Gene modules were labeled in different colors. WGCNA started with a gene expression matrix and calculates pairwise correlations between genes. These correlations were then raised to a power (soft thresholding) to distinguish strong connections between genes. Using hierarchical clustering based on the topological overlap measure (TOM), genes were grouped into modules with similar expression patterns. Modules were often represented by a module eigengene (ME), which captures the average expression profile of the genes within the module. WGCNA assessed the correlation between module eigengenes and sample traits. Intramodular hub genes, which are highly connected within their module and may represent key regulators or biomarkers were identified. Network concepts include whole network connectivity (degree), intramodular connectivity, and the clustering coefficient, etc. The mean clustering coefficient has been used to measure the extent of module structure present in a network. PPI network and pathway analysis of the expression data was done by Ingenuity Pathway Analysis (IPA, QIAGEN, Redwood City) for genes with fold change (FC) ± 2.0 [34].

Real-time quantitative RT-PCR

Quantification was performed with a two-step reaction process: reverse transcription (RT) and PCR. Reactions were performed in a GeneAmp® PCR System 9700 (Applied Biosystems, USA). Real-time PCR was performed using LightCycler® 480 Real-time PCR Instrument (Roche, Swiss). Each sample was run in triplicate for analysis. At the end of the PCR cycles, melting curve analysis was performed to validate the specific generation of the expected PCR product. The expression levels of mRNAs were normalized to GAPDH and were calculated using the 2−ΔΔCt method.

Statistical analysis

One-way analysis of variance (ANOVA) was used for multiple comparisons among groups. Statistical analysis was performed by using Graph-Pad Prism 8.0 software (GraphPad Software Inc., La Jolla, CA, USA). A p-value < 0.05 was considered as statistically significant.