Animal ethics
All animal procedures were performed according to the Laboratory Animal Guideline for Ethical Review of Animal Welfare (GB/T 35892-2018). All animal procedures were approved in advance by the Animal Care and Use Committee of Kunming University of Science and Technology (Appr. No: PZWH(Dian) K2022-0014) and Kunming Medical University (Appr. No: KMMU20221593). All animals used in the study are treated with care and respect, ensuring that their welfares are protected throughout the study process. Adult SPF-grade Sprague Dawley (SD) female rats, weighing 220-280 g, were obtained from the Animal Center, Kunming Medical University, China. Healthy female rhesus monkeys, ranging in age from 13 to 20 years, were selected for use in this study. The rhesus monkeys were housed with a 12-hour light/dark cycle between 06:00 and 18:00 in a temperature-controlled room (22 ± 1°C) with free access to water and food.
Culture of NESCs and MSCs
Two human embryonic stem cell (hESC) lines (H1 and H4) were used for neuroepithelial stem cell (NESC) production. The two hESC lines were isolated and expanded on Vitronectin (ThermoFisher, A14700) coated plates in the AIC medium according to our previously developed protocol 34. The AIC medium consisted of a modified N2B27 medium supplemented with 10 ng/ml Activin A (Peprotech, 120-14E), 2 μM IWP-2 (Selleck, S7085) and 0.6 μM CHIR99021 (Selleck, S2924) 34. hESCs were routinely passaged every 3–4 days at a ratio from 1:10 to 1:20 by single-cell dissociation. A GMP system of differentiation and expansion of NESCs was developed according to our developed differentiation protocols 23 and expanded in large-scale expansion 25. Digested hNESCs were inoculated into a 6-well low-adhesion cell culture plate (Corning, 08022047) with serum-free NESC growth medium as previously reported 25. Every 3 days, the NESC spheres were dissociated and passaged using TrypLE™ Express Enzyme (Gibco, 12605028): PBS (1:2). The NESCs were fed 2 days after inoculation by replacing 50% of the medium with fresh NESC growth medium. NESCs were amplified for at least 3 passages before use.
Human umbilical cord mesenchymal stem cells (MSCs) were isolated from umbilical cords after a normal full-term delivery collected from the People's Hospital of Yunnan Province of China, with informed consent from pregnant mothers. To extensively expand MSCs in vitro, MSCs were digested into single cells and cultured in a 100 mm cell culture dish (Corning, 40167). Cells were cultured in a serum-free mesenchymal stem cell culture medium consisting of IMDM media and other mesenchymal stem cell growth factor surpluses (Patent # CN112608891A). After 7 days of culture, adherent primary cells were passaged, and tissue pieces were removed from the culture. At passage 3, MSCs were stained with MSC-specific marker antibodies, including CD90, CD105, CD73, CD45, and CD34 (BD,562245), and analyzed using a FACS AriaII flow cytometer (BD). The medium was changed every 2 days and passaged every 3 days. MSCs from passages 4 through 7 were used in the study. NESCs and MSCs were housed in a humidified incubator with 5% CO2 at 37 °C.
Synthesis and characterization of GelMA and HAMA
10g gelatin (porcine skin, Sigma-Aldrich) was dissolved in 100 mL of PBS solution at 55°C. Then, 6 mL methacrylic anhydride (Sigma-Aldrich) was added dropwise, and the acylation reaction occurred under continuous stirring for 3 h at 50 °C. The reaction was terminated by dilution with the addition of 500 mL of warm PBS (50 °C). Then, the mixture was dialyzed for ~7 days against deionized water to remove unreacted reagents and by-products with 8 ~14 kDa cut-off dialysis membrane (under ambient temperature). Finally, GelMA was obtained after lyophilization and stored at -20 °C. The preparation process of HAMA is similar to that of GelMA. 4g hyaluronate sodium was completely dissolved in 300 ml of deionized water and stirred at 25°C until fully dissolved. 10 mL methacrylic anhydride was added dropwise at 25 °C while stirring and the pH of the solution was maintained at 8.0-9.0 with 0.25 M NaOH. Afterward, the reaction mixtures were stirred continuously and allowed to react overnight at room temperature. Next, 1500 mL cold ethanol was added in the reaction mixtures and stirred for 15 min, and filtered by buchner funner (10 cm diameter) to collect HAMA crude products. Then, HAMA crude products were fully dissolved in deionized water to form HAMA solution. The HAMA solution was dialyzed for 7 days to remove unreacted methacrylic anhydride with 12~14 kDa cut-off dialysis membrane at 25°C. Finally, the dialyzed solution was lyophilized to obtain the resulting product, which is termed HAMA.
To determine whether gelatin and hyaluronate sodium were conjugated by the photosensitive methacrylate, the synthetic GelMA and HAMA were dissolved in D2O (10mg/ml-1) and characterized by 1H NMR (400 MHz, Bruker, Germany), which successfully conjugation was confirmed by the spectra analysis (Supplementary Figs. 1e). By conjugating the photosensitive methacrylate (MA) on Gelatin and HA, the double-crosslinked hydrogel of hybrid GelMA and HAMA (GH) obtained a simultaneous higher elasticity and outstanding bioactivity, while not much negative effect of functional assemblies of Gelatin and HA. The GH samples (5% GelMA and 1%HAMA) with 0.5 wt% LAP photoinitiator were crosslinked for 45 seconds using a 405 nm wavelength blue light, and form the GH hydrogel at room temperature (Fig. S1g). On the contrary, the mixture of gelatin and HA cannot be crosslinked to form the scaffold with stable elasticity. To monitor the gelation process of GH hydrogels, a dynamic time sweep rheological experiment was conducted by a rheometer (TA, HR10, USA). The GH hydrogel was rapidly formed when the storage modulus (G’) surpassed the loss modulus (G”) at t (time) ≈ 100 s after UV irradiation (Fig. S1H). In contrast, during the shear process, the storage modulus (G’) of the mixture (gelatin and HA) was observed to be smaller than the loss modulus (G”) as the samples were still in the liquid state. The final storage modulus (G’) of the GH hydrogel can reach a plateau of ∼2138 Pa, which was much higher than that of the uncrosslinked GH (∼0.102 Pa) or the mixture of gelatin and HA materials (∼0.037 Pa), signifying a structurally robust network that maintains its 3D shape (Fig. S1I).
To present the microstructure of GH microspheres, the lyophilized GH microspheres were observed by scanning electron microscopy (Nova-Nano450, FEI) after being fixed on a sample table and sprayed with gold for 90 s (Fig. S1g).
Preparation and characterization of stem cell-loaded hydrogels:
The synthetic GelMA (G) and HAMA (H) were dissolved in PBS solution or IMDM medium including 0.5% (w/v) LAP photoinitiator (lithium phenyl-2,4,6-trimethylbenzoylphosphinate, R101A, STEMEASY) at a final concentration of 10% GelMA and 2% HAMA (w/v), respectively. The GelMA or HAMA solution was then filtered through a 0.22μm filter for sterility before use, respectively.
To generate GH hydrogel microsphere-encapsulated stem cells for spinal cord injury (SCI) rat transplantation, the following methods were used to obtain GH hydrogel microsphere-encapsulated MSCs (MSC@GH), microsphere-encapsulated NESCs (NESC@GH), and microsphere-coencapsulated NESCs and MSCs (NESC+MSC@GH).
1ml 10% GelMA and 1ml 2% HAMA solution were thoroughly mixed in an equal ratio in a test tube. The final mixed GH hydrogel concentration is 5% GelMA (G), 1%HAMA (H), and 0.5% LAP (w/v). For the MSC@GH group or NESC@GH group, a total of 1×106 MSCs or 1×106 NESCs were resuspended in 1ml mixed GH hydrogel solution with 5% GelMA and 1% HAMA, respectively. For the NESC+MSC@GH group, a total of 5×105 NESCs and 5×105 MSCs were resuspended in 1ml mixed GH hydrogel solution with 5% GelMA and 1% HAMA.
The cell mixed hydrogel solution was immediately added with a 10μl multichannel pipette drop by drop at 0.5ul into 4℃ cooled silicone oil in a dish, formed the scattered droplets, which were then preliminarily physical-crosslinked and settled onto the bottom of the dish to form microspheres due to the gravity. The microspheres were crosslinked for 45 seconds using a 405 nm wavelength blue light and collected by filtering through 70 μm mesh.
The blank hydrogel microspheres (GH) were obtained by a similar protocol but the microspheres do not contain cells. All hydrogel microspheres were washed with PBS three times. The hydrogel microspheres encapsulating stem cells were temporarily stored at 4℃ before transplantation or used for in vitro cell culture.
For cells used for transplantation of rhesus monkey SCI models, a total of 1×106 NESCs and 1×106 MSCs were resuspended in 1ml GH mixed hydrogel solution with 5% GelMA and 1% HAMA. The cell solution was temporarily stored at 4℃ before transplantation.
Cell viability and proliferation assay of MSCs and NESCs encapsulated by hydrogels
Approximately 20 MSC or NESC-loaded hydrogel microspheres were seeded into a 24-well cell culture plate (Corning, 3524) with NESC or MSC growth media. The morphologies of stem cell in hydrogel microspheres were observed under an optical microscope, and their diameter were measured.
To evaluate the MSC adhesion ability in GH hydrogel microspheres, 100 μl of GH hydrogel microspheres containing 1×106 MSCs were seeded into a 6-well low-adhesion cell culture plate. Cells growing on the surface of blank hydrogel microspheres were stained with DAPI (nuclei, blue) and Alexa 488 Phalloidin (F-actin, Thermo Fisher, A12379) after 3 days of cell culture.
A Calcein-AM/PI Double Stain living/dead cell kit (YEASEN, 40747ES76) was applied to measure cell viability in the microspheres according to the manufacturer's instructions. After 1, 3, and 5-day cultures, cell-loaded microspheres were stained in a 37 °C incubator for 30 mins by the living/dead working solution. The living cells could be labeled by green Calcein AM and the dead cells were stained red with PI. Fluorescent images were captured using an inverted fluorescence microscope to assess cell viability.
A cell counting kit (CCK)-8 (Dojindo, CK04) was used to evaluate cell proliferation activity of MSCs or NESCs in the hydrogel microspheres following the CCK8 kit manufacturer’s protocol. Briefly, 20 μl of CCK-8 working solution was added on day 1, day 3, and day 5. After 1h of incubation, the absorbance density of 100 μl supernatant medium in a 96-well plate was recorded by a multifunctional microplate reader.
Completely transected SCI rat models and hydrogel-encapsulated stem cell transplantation
A total of 90 adult SPF-grade SD female rats weighing 220-280 g were used for performing complete transection of spinal cords. Rats were anesthetized with 5% isoflurane and maintained on 1.8% isoflurane anesthesia during the surgery. To prepare a rat model of thoracic spinal cord transection, we created a laminectomy, followed by transection at the thoracic T4, with the removal of a 1.0-mm spinal cord segment with micro-scissors. The spinal cord was carefully flushed using PBS saline and a 1mm gap was observed. After surgeries, similar lesions were confirmed in all experimental groups. About 330 μl hydrogel microsphere-encapsulated stem cells were immediately transplanted into the injured spinal space in SCI rats using a 5ml injection syringe after SCI surgery, which was sufficient and enough to fully fill the space of the injured spinal cord (Fig. S2A). Each rat received about 3.3×105 stem cells. The wound was sutured layer by layer with a 4-0 absorbable surgical suture. Penicillin was administered for three days post-surgery. Rats were divided into five groups: the lesion control group, in which the thoracic T4 spinal cord was surgically completely transected and then the saline solution was injected into the lesion gap; the empty hydrogel (GH) group, in which an empty GH (no cell encapsulation) microspheres were inserted into the lesion gap; the GH-encapsulated MSC group (MSC@GH; 3.3x105 MSCs); the GH-encapsulated NESC group (NESC@GH; 3.3x105 NESCs); and the GH-co-encapsulated NESC and MSC group (NESC+MSC@GH; 1.65 x105 cells each NESCs and MSCs).
Rhesus monkey SCI models and hydrogel-loaded stem cell transplantation
A total of 9 female rhesus monkeys (13-20 years old, weighing 4±2kg) were used in this study. 6 rhesus monkeys were anesthetized with intramuscular injection of ketamine hydrochloride (10 mg/kg) and hydroxyzine (5 mg/kg), intubated, and maintained with sodium pentobarbital (20 mg/kg). A laminectomy was performed to expose the thoracic spinal cord at the T10. Approximately one-quarter of the T10 left dorsal spinal cord was cut off and removed using a scalpel under an operating microscope (2mm along the rostral-caudal direction, 2 mm along the left and right direction, and 2 mm along the dorsoventral direction), generating a 2× 2× 2 mm injury cavity.
Monkeys were allowed to recover from anesthesia at their home cages (relative humidity 28–29%, ambient temperature 23-26℃). Penicillin was continuously administered for three days post-surgery. Buprenorphine (50μg/100g) was subcutaneously administered twice daily for 4 days after surgery.
Since quarter-section only produced limited space that were unable to be fully filled by transplanted microspheres with 400-500 μm diameter, after hemostasis was achieved, we injected the mixtures of hydrogel and stem cells (200 µl) individually into the lesion gap of 3 animals followed by crosslinking into solid with 405 nm wavelength light, and the remaining 3 animals were used as lesion controls by injection of PBS (Fig. 1A and S7A). After the hydrogels were immediately cross-linked exposure to 405nm wavelength blue light for 45s, the muscles and skin were sutured layer by layer. Every rhesus monkey received 4×106 stem cells (2×106 MSCs and 2×106 NESCs). Cell concentration in the monkey is 20 times that in rats. No any immunosuppressant was used in the whole study. 3 monkeys were randomly injected with PBS as the control group (without GH hydrogels or cells).
Rat behavioral tests
The recovery of open-field motor function of SCI rats was assessed by two individuals using the Basso, Beattie, and Bresnahan (BBB) scoring after and before stem cell transplantation in double-blind condition following the previous report 39 . Rats were made to move freely and the movements of their hindlimbs were recorded for 5 min according to the BBB open-field locomotor scale. The BBB test was carried out every 5 days from day 0 to day 180 following a surgical procedure. Each animal phenotype was recorded through videos. Results were analyzed using GraphPad prism.
Monkey behavioral tests
The open-field test was conducted to assess monkey behaviors. Briefly, within 180 days of SCI monkey model treatment, the behaviors of each monkey in the control and NESC+MSC@GH group were analyzed using open-field equipment (XZ-XR301, Shanghai Xin Run Information Technology Co, Ltd, China). The equipment included an open-field opaque cage (240cm ×240cm×240cm) and a data acquisition system for automatic control of data collection and analysis. Monkeys were transferred to the open-field opaque cage in a quiet environment, and each monkey was allowed to move freely, such as climbing, jumping and walking in the field for 30 minutes. High-resolution video equipment (HDRXR260, SONY) was used to record all behaviors of monkeys in the open-field opaque cage for at least 20 min. The following items were used to quantify indices of locomotion activity, including average moving speed (mm/s), total moving distance (mm), total climbing time (s), immobility time (s), standing time (s), sitting time (s), top speed (mm/s), freezing time ratio (%), immobility time ratio (%) and climbing time ratio (%).
In addition, on days 2, 7, 14, 21, 30, 60, 90, and 180 post-injury, the hindlimb scores (MHS) of each monkey were observed, averaged and recorded by three trainees in double-blind conditions. The items for MHS included joint movement, walking ability, and hindlimb digital function after treatment 53. Finally, the behavioral differences of SCI monkeys in open-field assays before and after treatment were analyzed using a one-way analysis of variance (one-way ANOVA) by IBM SPSS Statistics 25.
Transcriptome analysis after treatment
On days 3, 7, and 14 post-treatment, three rats per time point were randomly chosen from each treatment group (control group, MSC@GH group, NESC@GH group, and NESC+MC@GH group) and humanely euthanized by intraperitoneal injection of pentobarbital after induction of anesthesia. The spinal cords were rapidly exposed after perfusion with physiological saline to remove tissue blood residues. A 5mm tissue sample was excised from the central region of the spinal cord injury site, which remained non-continuous at the time of excision. To avoid the interference of transplanted cells with the results of endogenous host cells, we removed grafted stem cells in GH microspheres from the host tissues since GH microspheres maintained the integrity of the structure before 14 days. The extracted spinal cord tissue was promptly placed into a 1.5 ml microtube and flash-frozen in liquid nitrogen for preservation. The samples were then transported with dry ice to Annoroad Gene Technology Co., Ltd. (Beijing, China) for RNA extraction and quality assessment. Library preparation was then performed, and the resulting libraries were analyzed using high-throughput RNA sequencing on an Illumina 6000 platform.
Processing and analysis of these collected data were conducted with the R software. To map reads and quantify transcript expression levels, the workflows of HISAT2 and StringTie-Ballgown were used. The paired-end clean reads were aligned to the Rattus norvegicus (UCSC rn6) reference genome using HISAT2 software (v2.2.1). Following alignments, raw counts of total genes were normalized to FPKM.
To investigate the gene expression profiles of both the control and the three treatment groups, the expression patterns were compared using the TCseq (Time course sequencing data analysis) clustering method, and genes used for clustering were selected by maximum FPKM ≥ 40. Hence, the gene expression patterns of the control and treatment groups were divided into 12 categories. To facilitate further analysis and interpretation, the raw data (FPKM) was standardized into z-scores. This transformation provides valuable insight into the magnitude of changes in gene expression levels, enabling identification of significant increases or decreases across different time points or samples. Z-scores allow for more accurate comparison of gene expression levels between different groups and can identify outlier data points that may be of particular interest or significance. A membership value of 1 indicates a perfect match for the cluster, while a value of 0 indicates no association with the cluster at all.
To gain a thorough understanding of the functions of genes within clusters exhibiting similar expression patterns across different treatment groups, genes from such clusters with comparable up- or downregulation patterns were aggregated and subsequently analyzed via a KOBAS GO (Gene Ontology) biological process enrichment analysis (http://kobas.cbi.pku.edu.cn/). Meaningful enrichment of biological processes that are relevant to spinal cord injury diseases have been selected and presented for each cluster. The enriched biological processes s with P-values <0.05 were finally visualized with a bubble chart generated with packages ggplot on RStudio. The bubble size typically represents the number of genes associated with a particular GO term.
SEPs and MEPs examination
Rhesus monkeys were anesthetized with intramuscular injection of ketamine hydrochloride (10 mg/kg) and hydroxyzine (5 mg/kg), intubated, and maintained with sodium pentobarbital (20 mg/kg). Electrophysiological examinations of rhesus monkeys were performed over 180 days before and after the treatment each group. Six months after surgery, both the lesion control group and treatment group were evaluated using somatosensory evoked potentials (SEPs) and transcranial electrical stimulation-motor evoked potentials (TES-MEP). The experiment utilized the Nicolet EDX system (Natus), a US electrophysiological diagnostic device, which had an electric stimulator probe capable of producing a maximum output voltage of 400V. The experimental monkeys were anesthetized and their limbs were secured to a wooden board.
The examination of transcranial electrical stimulation-motor evoked potentials (TES-MEP) was conducted at a temperature of 23-26°C. The reference electrode was positioned on the forehead, the grounding electrode on the abdomen, and the recording electrode on the ventral surface of the hind limbs' quadriceps femoris muscle. Electrical stimulation was delivered repeatedly to the motor area of the monkey cerebral cortex's lateral side using the electric stimulator probe. The signal was amplified and recorded using an amplifier with a frequency band range of 2Hz-10kHz, amplification sensitivity of 200uv/Div, and stimulation intensity of 200-300, until contraction of the recorded quadriceps femoris muscle was observed. The scanning speed was 10ms/Div, and the duration of the continuous current was 0.5ms. The latency (P1) and amplitude from negative to positive (from N to P) were measured.
We also measured the somatosensory evoked potentials (SEPs) of the previously mentioned monkeys. A reference electrode was positioned on the forehead, and the recording electrode was placed on the coronal aspect of the skull, outside of the somatosensory cortex of the brain. Additionally, a ground electrode was placed on the abdomen. SEPs were measured using an electrophysiological device, while a bipolar stimulation probe with both positive and negative poles was used to stimulate the bilateral tibialis anterior muscles. The stimulation intensity was 20 mA, and we applied a continuous current duration of 0.2 ms, even if there was a slight movement in the hind limb toes. The amplification sensitivity was 5 μV/div, and the frequency band range was set at 20 Hz to 3 kHz. We used a scanning speed of 8 ms/div and obtained an average of 120 measurements. The recording electrode was kept at a constant distance from the stimulation electrode, and we recorded the SEPs using the recording electrode placed above the somatosensory area of the cortex, measuring the latency (P1) and amplitude (P1-n1).
MRI test
The rhesus monkeys were anesthetized in accordance with the above-mentioned method, then we carried out a study using a Philips Elition 3.0T scanner to perform magnetic resonance imaging (MRI) of the spinal cord in non-human primates. To ensure consistency, we made no modifications or alterations to the hardware or software configuration of the MR system during the course of the study.
For image acquisition, we utilized a 32-channel head receiver coil array, and each anesthetized monkey was positioned in a fixed system in a supine position. This approach minimized respiratory motion and ensured a consistent position during image capture. We conducted spinal cord MRI examinations at four-time points: prior to surgery, and at 1-, 2-, and 6 months post-surgery. We employed a standardized imaging protocol, as detailed in Table S1, which encompassed imaging the entire length of the spinal cord using a fast spin-echo sequence (TSE) to obtain diffusion tensor imaging (DTI) data. The DTI axial imaging was performed with a b-value of 800 s/mm2 and 15 diffusion gradient directions. To minimize the impact of physiological motion artifacts, we applied saturation bands to the chest and abdomen regions. The total acquisition time for the imaging sequence was 40 minutes.
MRI image processing and data analysis
The monkey MRI images were processed with the MR Diffusion software (https://diffusionkit.readthedocs.io/en/latest/) of the Philips IntelliSpace Portal workstation for DTI post-processing54. The eddy currents and head motions of the diffusion MRI were corrected by affine transformations to minimize gradient coil eddy current distortions. Diffusion tensor deterministic tractography was performed using the fiber assignment by continuous tracking (FACT) method. Seed points were selected as voxels with an fractional anisotropy (FA) value greater than 0.1 in spinal cord. The tractography procedure was terminated when the angle between adjacent steps greater than 45 or reached a voxel with an FA less than 0.1. The details for fiber tractography can be found in a previous study55. The FA value was calculated by selecting the injured spinal cord segment from the T2 images and computing the apparent diffusion coefficient (ADC) value. Prior to image post-processing, diffusion correction was performed to improve the accuracy of the generated FA maps, leading to more precise fiber tracking results. Subsequently, the required anatomical images were dragged into the DTI post-processing module to fuse with the FA pseudocolor maps, and multiple regions of interest were delineated to facilitate the tracking of the fiber bundles.
Tissue Processing and immunohistochemistry
On day 30, 90, 180 post-injuries, rats were anesthetized by an overdose injection of pentobarbital sodium. Anesthetized rats were perfused with 150ml physiological saline followed by 100ml 4% paraformaldehyde in PBS. The bladders, spinal cords encompassing the lesion sites, and brains were harvested in 4% paraformaldehyde. On day 180 post-injury, rhesus monkeys were deeply anesthetized with a high dose of intramuscular ketamine hydrochloride (50 mg/kg), and then perfused transcardially with 1500 ml of normal saline, followed by 4000 mL of 4% paraformaldehyde in phosphate-buffered saline.
To assess the bladder histopathological alteration, rat bladder tissues were processed, embedded in paraffin, and cut into 4 μm sections. The sections were then performed hematoxylin-eosin (H&E) staining with hematoxylin and 25% eosin. Digital images of stained sections were scanned and exported using a Pannoramic Digital Slide Scanner and analyzed using Pannoramic Viewer software (1000, 3D HISTECH). The bladder wall thickness and cross-sectional area were measured using ImageJ software. Thirty locations were randomly selected to measure the bladder wall thickness, and the average value was calculated by GraphPad Prism.
For immunohistochemistry of rat and monkey spinal cord tissue, spinal tissues were fixed in 4% paraformaldehyde for additional 24 hours (48 hours for monkeys), sequentially infiltrated with a series of 10, 20, and 30% (w/v) sucrose solutions; embedded in OCT (optimal cutting temperature compound; Sakura, 4583); and sectioned at the thickness of 20 μm on a Leica cryostat. According to the section thickness per slice and the initial tissue damage depth, the lesion area was demarcated, and the location of every section in the spinal cords was recorded to ensure that each section used for further staining and histopathological analysis contained the lesion region (Fig. 7B).
Sectioned spinal tissues on slides were washed three times in PBS (phosphate-buffered saline), and treated with 0.2% Triton X-100 (Gibco, 85112) overnight, washed three times in PBS, incubated in blocking buffer (3% BSA (Solarbio, A8020) in PBS) for 2 hours at room temperature, and washed three times in PBS. Tissue slides were then incubated with primary antibodies overnight at 4℃, washed three times in PBS, and incubated for 2 hours at room temperature in blocking buffer containing Alexa Fluor 488, 594, or 647 conjugated secondary antibodies (1:600; A-21202, A-31571, A31573, A-31572, A-21447, A11055, Invitrogen). Nuclear counterstain was visualized with 4′,6-diamidino-2-phenylindole (DAPI, 32670, Sigma). The following primary antibodies were used at 1:200 to 1:2000: NF200, neurofilament mouse (N5389, Sigma, 1:1000); SMI312, mouse (837904, Biogelend, 1:1000); TUJ1, rabbit (T2200, Sigma-Aldrich, 1:600); MBP, rabbit (HPA049222, Sigma-Aldrich, 1:600); CGRP, mouse (ab81887, Abcam, 1:100); GFAP, rabbit (z033429-2, Dako, 1:800); Iba1, rabbit (019-19741, Dako, 1:800); GFP, chicken (ab13970, Abcam, 1:600); HN, human nuclei, mouse (MAB1281, Millipore, 1:600); DCX, rabbit (ab18723, Abcam, 1:600); PSA-NCAM, mouse (MAB5324, Millipore, 1:500); NEUN, rabbit (ABN78, Millipore, 1:600).
Histochemical Analyses
Rat axons labeled with NF200 were counted by two researchers in double-blind conditions. For each section, a start line was drawn at the rostral boundary of the lesion, and then 16 vertical lines at regular distances with 0.4 mm were drawn from rostral to caudal orientation. The regenerated axon intensity was presented as the number of labeled axons. In brief, lines were drawn across the longitudinal spinal cord sections of the SCI lesion site and the number of axons intercepting the line drawn was counted.
BDA tracing
The rats in the NESC+MSC@GH group and the control group were anesthetized in accordance with the above-described method 180 days after treatment. For anterograde BDA tracing, a total of eight holes were drilled on the rat skull using a micro drill, then 2.5μl of 10% BDA (Dextran, Biotin, 10,000 MW, Lysine Fixable, D1956, ThermoFisher) was slowly injected into 8 sites of the sensorimotor or motor cortex with a micro-syringe. The bregma was used as the zero point. The coordinates of these 8 points were: 0.2mm, -0.2mm anteroposterior coordination from the bregma; lateral to the bregma: 2.5, 3.5, -2.5, and -3.5mm. And the injection depths were 1 mm and 2.5 mm, respectively. The needle was left in place for an additional 3 min to ensure the BDA solution adequately penetrated the tissue before moving to the next site, and then the skin was closed with sutures. For retrograde BDA tracing, a laminectomy was performed at the T10 level to expose the 3mm dorsal spinal cord, and 10μl of 10% BDA was slowly injected into the spinal cord with a micro-syringe.
Two weeks after the injection of BDA, rats wereanesthetized and perfused with 4% paraformaldehyde. The spinal cord tissue was sliced longitudinally to obtain sections with 20μm thickness. The tissue sections were incubated with streptavidin Alexa Flour 568 (1:1000 with PBS, S11226, Invitrogen) overnight at 4℃ to detect the BDA-labeled axon fibers.
Statistical analysis
The data are presented as mean ± standard deviation (SD). Statistical analyses were performed using GraphPad Prism software (version 8.0) with one-way ANOVA and two-way repeated measures ANOVA. A p-value less than 0.05 was considered statistically significant.
Reporting summary
Further information on research design is available in the NatureResearch Reporting Summary linked to this article.