NSCs isolation and culturing
NSCs cultures were obtained from the fetal brains of embryonic day 14 rats, which were extracted from pregnant Sprague-Dawley (SD) rats (Laboratory Animal Center of Sun Yat-sen University, Guangzhou, China) and identify as previously described [22, 23]. Briefly, the brain tissue was mechanically dissected and dissociated in Hanks Balanced Salt Solution, and the cell suspension was centrifuged at 1,000 rpm for 5 minutes. The supernatant was discarded, and the cell pellet was diluted to a single-cell suspension. NSCs were plated on a T25 culture flask (Corning, Acton, MA) containing Dulbecco modified eagle medium (DMEM) /F-12 nutrient mixture, 2% B27, 1% penicillin/streptomycin, 1% l-glutamine (Gibco, Grand Island, NY), 20 ng/mL fibroblast growth factor-2 (FGF-2) and 20 ng/mL epidermal growth factor (EGF) (Peprotech, Rocky Hill, NJ). NSCs were cultured at 37°C in 5% CO2 and were passaged via weekly digestion with accutase (Millipore, Bedford, MA) in the medium described above. All NSCs used in this study were between passages 2 and 4.
To induce neural differentiation, cells were plated at a density of 2x105 cells/well in 6- or 12-well tissue-culture plates and allowed to adhere for 24 h at 37°C, at which time cells were switch to neural differentiated medium consisting of basic medium supplemented with 2% B27, 1% penicillin/streptomycin, 1% l-glutamine. The medium was changed every 2-3 days.
Monocytes were isolated from the femurs of adult SD rats (Laboratory Animal Center of Sun Yat-sen University, Guangzhou, China) as previously described [24]. To induce M1 and M2 polarization of macrophage, cells were plated at a density of 2x105 cells/well in 6- or 12-well tissue-culture plates and allowed to adhere for 24 h at 37°C, at which time cells were switch to polarization medium consisting of basic medium supplemented with LPS (10μg/ml) for M1 and IL-4 (10ng/ml) for M2 polarization (Fig. S1a to c) [25, 26].
Lentiviral vector construction
The lentiviral vectors carrying green fluorescent protein (GFP) and a sequence that specifically overexpression the Wnt4. Lentiviruses carrying GFP and Wnt4 gene overexpression were constructed as previously described [22] (Fig. S2a). Briefly, the full-length Rat Wnt4 gene was encoded into vectors. The vectors and corresponding packaging plasmids were co-transfected into 293T cells using lipofectamine 2000 (Invitrogen, Camarillo, CA). The medium was changed with complete medium after 8 hours of incubation. After 48 hours, the supernatant was harvested from 293 T cells, filtered using a 0.45-mm pore size filter, and concentrated via ultracentrifugation at 96,500g for 2 hours at 4 °C. After resuspension, the serially diluted lentivirus was used to transduce 293T cells. Then, 4 days later, the labeled 293T cells were counted to calculate the viral titer, and high-titer recombinant lentiviral vectors carrying Wnt4 were harvested.
Surgical Procedures and Cell Transplantation
Adult female Sprague-Dawley (SD) rats (weighing 200-220 g, supplied by the Experimental Animal Center of Sun Yat-sen University, Guangzhou, China) were divided into four groups for this study: Sham group (n = 10), SCI group (n = 10), NSCvector group (n = 10), and NSCWnt4 group (n = 10). After 72 hours of lentiviral transfection, NSCs were collected for transplantation. Briefly, animals were anesthetized with 1% pentobarbital sodium (40-45mg/kg), and the spinal cord was exposed at the 10th thoracic vertebral level (T10) via laminectomy. The animals in the sham group underwent spinal cord exposure but no injury. In the other three groups, the exposed spinal cord was contused with a weight-drop device by dropping a 10g rod from a precalibrated height of 12.5 mm [27]. Once the bleeding was stopped, 5μl NSCs were implanted at a density of 1x105 cells/μl to the rostral and caudal of injured site using microsyringe [28, 29]. The T8-T11 spinal cord segments were dissected at 8 weeks after SCI and NSCs transplantation.
Hindlimb locomotor scale
The 22-point (0-21) Basso, Beattie, and Bresnahan (BBB) open-field locomotor scale was used to assess hindlimb locomotor function, including joint movements, stepping ability, coordination, and trunk stability. 21 score indicates unimpaired locomotion as observed in sham rats. All animals underwent behavioural testing, and the duration of each session was 5 minutes per rat. Finally, an overall score was calculated. The evaluation was performed by observers who were blinded to the treatment group of the tested animals [30].
Footprint analysis
Hindlimb behavior and motor coordination were assessed at 8 weeks post-injury. The fore limbs and hind limbs were coated with different colors dyes, and the rats of different groups were placed on a 10 cm x 100 cm runway covered with paper. The rats were encouraged to run straight so that the gait of the rats in different groups could be obtained and assessed. The footprint pattern was then digitalized and representative pictures were shown to assess coordination [31].
Spinal Cord–Evoked Potential (SCEP) Recording
At 8 weeks post-injury, total 20 rats (n = 5 in the sham group and n = 5 in each cell-transplanted group) were anesthetized with 1% pentobarbital sodium (40-45mg/kg) and stereotaxically fixed. The T5-L1 vertebrae were completely exposed. Briefly, the stimulation electrode was inserted into the T5-T6 interspinous ligaments, and a pair of needle electrodes was inserted into the interspinous ligaments of T12-L1 for SCEP recording. Then, the electrodes were connected to a BL-420 Biological Function Experiment System (Taimeng, Chengdu, China). The variables of the SCEP signals were set according to previous reports as follows: gain of 2000 time constant of 0.01 seconds, and filtering at 300 Hz. To elicit a SCEP, a single pulse stimulation (50 ms in duration at a frequency of 5.1 Hz and a voltage increase of 1 mV) was transmitted through the electrodes until a mild twitch of the vertebral body of the animal was observed. One hundred SCEP responses were averaged for each rat to obtain high-quality waveforms for the SCEP signals [32].
Histological analysis
At 8 weeks after SCI, all rats were deeply anesthetized with an adequate dose of 10% chloral hydrate (5 mL/kg) and were transcranial perfused with 250 mL of 0.9% normal saline. Animals were perfused with 300 mL of 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer solution (PBS, pH 7.4). The T8-T11 cord segments were dissected based on the dorsal spinal root count, post fixed overnight in 4% PFA, and soaked at 4℃ overnight in 10% sucrose followed by 30% sucrose. The specimens were embedded in optimal cutting temperature compound, frozen at -20℃ and sliced at a thickness of 20 μm in the longitudinal or transverse plane.
To visualize the cavity area, animals (n = 5 per group) were killed for haematoxylin-eosin (HE) staining. The T8-T11 longitudinal spinal cord sections from each group were stained with HE according to standard protocols and observed under bright field microscope [22, 32].
For neuron counting, animals (n = 5 per group) were killed, and transverse sections of the injured spinal cord were used to stain neurons with Nissl. Sections at 2 mm rostral and caudal to the lesion epicentre were counted for each rat. The numbers of positively stained cells were counted and averaged per section in a blinded manner [22, 32].
For axonal tract tracing, dorsal laminectomy was performed at T12 and Fluorogold (FG, Biotium, Fremont, CA) was injected into the spinal cord at seven weeks after operation. One week after injection, the animals were perfused and T8 segment of the spinal cord were removed, cryopreserved, embedded in OCT compound, and sliced into 10μm frozen sections. A fluorescence microscope (Olympus, Tokyo, Japan) was used to detect FG-labeled neurons [28, 29].
Statistical analysis
Data obtained from experiments in triplicate and repeated at least three times was represented as mean ± SD. Statistical evaluations was analyzed via one-way analysis of variance (ANOVA) with Levene’s homogeneity of variance test and, subsequently, Bonferroni’s post hoc test or Dunnett’s T3 post hoc test based on the comparison to be made and the statistical indication of each test. The analyses were performed using Statistical Package for Social Sciences, Version 16.0 for Windows (SPSS, Chicago, IL, USA). A statistical probability of ρ<0.05 was considered significant.