To investigate the therapeutic effect of recombinant HAP1 (r-HAP1) in SCI, the SCI model was established with Allen’s weight drop apparatus. Then, r-HAP1 (2 mg/kg/day) was administered by intrathecal injection, respectively, on the day of SCI and on the seventh day after SCI (Figure 1A). The recovery of neurological function was analyzed by the restoration process of the Basso Mouse Scale (BMS) score, tenderness threshold, Warmth sensation threshold, and neurotic electrophysiology by cortical motor evoked potential (cEMP). The results showed that r-HAP1 can effectively improve the BMS score (Figure 1B), pain sensation (Figure 1C), warmth sensation (Figure 1D) and cEMP manifestation (Figures 1E,F) after SCI in mice, but has little effects on normal mice (Supplementary Figure S1).

To further investigate the pathological alterations of SCI-injured local spinal cord tissue after r-HAP1 administrations, by immunofluorescent the neuronal characteristic marker Tuj was identified, and we characterized the status of injured spinal cord surrounding neurons. The outcomes revealed that r-HAP1 significantly increased the expression of local Tuj in the area of injury (Figure 1G; Student test value of p < 0.05).

To elucidate the effect of r-HAP1on growth of the axons, we isolated NSCs from spinal cord tissue from E14 mice. When NSCs were maintained with stemness in the proliferation phase, they grew as neurospheres and floated in the culture medium (Supplementary Figure 2A). As expected, these cells were positive for expression of the typical NSC marker nestin (Supplementary Figure 2B). R-HAP1 was added to the culture medium at a final concentration of 10 ng/mL, and the cells were incubated for 7–14 days for differentiation, the medium was changed every other day. The neurite length was calculated on the 7th day of neuronal differentiation. The results confirmed that r-HAP1 administration effectively promoted neurite outgrowth (Student test value of p < 0.05; Figure 2). Based on these findings, one may conclude that r-HAP1 is avail of the growth of neurite during neuronal differentiation in NSCs.

To gain more information on the effects of HAP1 in neurite growth, we further investigated the signaling mechanisms involved in HAP1-mediated regulation of NSCs. We screened the levels of signaling pathway changes in r-HAP1-treated NSCs with a Cignal Finder Reporter Array. The results showed that certain pathways were significantly activated, including the mitogen-activated protein kinase/extracellular regulated protein kinase (MAPK/ERK) pathway, the Sox2 pathway the Glucocorticoid pathway (log2 > 1). Most of the signaling pathway changes were not significant (Figure 3A).

The MAPK/ERK pathway was further probed by Western blotting as it has been reported that HAP1 activates the tyrosine kinase receptor TrkA and stimulates its phosphorylation (Rong et al., 2006; Su et al., 2018). To verify the hypothesis that HAP1 promotes TrkA phosphorylation by activating the MAPK/ERK pathway, we examined the phosphorylation of TrkA and ERK1/2 in our experimental conditions. The western blotting results showed a significant increase in the phosphorylation levels of TrkA and ERK1/2 in r-HAP1-treated NSCs than the PBS group (Student test value of p < 0.05, Figure 3B).

To further define the molecular mechanism by which HAP1 promotes neurite elongation after neuronal differentiation of neural stem cells, TrkA selective inhibitor GW441756 (2 μM; Selleck, China) (Jung et al., 2013), TrkA selective activator gambogic amide (0.5 μM, Santa Cruz Animal Health, USA) (Jang et al., 2007) and ERK activator U-46619 (10 nM; Santa Cruz, USA) (Nishihara et al., 2000) were used. The lengths of neurites after 7 days of neuronal differentiation were calculated. In the r-HAP1 + GW441756 group, the neurite length was significantly lower than gambogic amide group and r-HAP1 + GW441756 + U-46619 group (Figure 4A; one-way ANOVA with Turkey’s multiple comparisons test value of p < 0.05); Results indicated that the effect of HAP1 on neurite elongate can be inhibited by GW441756, and reversed by ERK1/2 activation. At the same time, gambogic amide could mimic the function of HAP1 in regulating neurite elongate after neuronal differentiation of neural stem cells.

Further, the Western blot results were in accordance with the results for neurite elongation. GW441756 significantly blocked TrkA and ERK1/2 phosphorylation (Figure 4B; one-way ANOVA with Tukey’s multiple comparison test value of p < 0.05 versus r-HAP1 group). From previous results, we concluded that gambogic amide could simulate the function of r-HAP1. Western blot results further illustrated that gambogic amide significantly overregulated the phosphorylation of TrkA and ERK1/2 (Figure 4B; one-way ANOVA with Tukey’s multiple comparisons test value of p < 0.05 versus PBS group). U-46619 caused ERK phosphorylation when TrkA was inactivated. These data strongly support and confirm our hypothesis that the HAP1-TrkA-MAPK/ERK signaling pathway regulates the neuronal differentiation of NSCs and promotes neurite outgrowth during neuronal differentiation.

To further validate the role of the HAP1-TrkA-MAPK/ERK signaling pathway in axonal neurite elongation in vivo, GW441756, gambogic amide, and U-46619 were grouped according to in vitro tests and injected in intralumbar. Tuj was identified on the 14th day after SCI modeling. The result was similar to the in vitro outcomes, in the r-HAP1 + GW441756 group the Tuj was significantly lower than gambogic amide group and the r-HAP1 + GW441756 + U-46619 group (Figure 5; one-way ANOVA with Tukey’s multiple comparisons test value of p < 0.05).

To verify the role of TrkA-MAPK/ERK in neurological function, The BMS score, pain sensation, warmth sensation, and cEMP manifestation were examined again with the administration of GW441756, gambogic amide, and U-46619. The BMS score result shows that r-HAP1 administrated with the GW441756 group was significantly lower than the gambogic amide group and r-HAP1 + GW441756 + U-46619 group (Figure 6A). Pain sensation, warmth sensation, and cEMP manifestation (Figures 6B–D) are consistent with BMS score results. Meanwhile, the administration of GW441756, gambogic amide, and U-46619 had no impact on the neurological function of healthy mice (Supplementary Figure S2). Combining the above-described results, it can be concluded that TrkA selective inhibitor GW441756 will block the effect of r-HAP1 to promote neurological rehabilitation, while TrkA selective activator gambogic amide can mimic the therapeutic effect of r-HAP1 on SCI mice; ERK activator U-46619 can restore the blocking effect of TrkA selective inhibitor GW441756 on r-HAP1.

C57BL/c mice aged approximately 6–8 weeks were provided by the Department of Laboratory Animals of Central South University. All animal breeding, care, and testing procedures were approved by the Animal Care and Use Ethics Committee of Central South University Xiangya Hospital. All mice were anesthetized with ketamine (60 mg/kg) and xylazine (10 mg/kg) by intraperitoneal injection. With skin preparation, disinfection, and positioning, the skin, subcutaneous muscle, and lamina were cut at the T10 spinous process to reveal the spinal cord. A model of incomplete spinal cord contusion was established using a modified Allen’s method with an impact force of 10 g × 25 mm. The wound was rinsed, the incision was closed layer by layer, and the wound was covered. R-HAP1 (2 μg r-HAP1 diluted in 0.5 μL PBS) was administrated by intralumbar delivery. All mice were housed in an animal laboratory at an ambient temperature of 22 to 24°C and a relative humidity of 60 to 80%. After the operation, the mice were kept in separate cages, with standard food and drinking water offered ad libitum. An intramuscular injection of penicillin 2 WU/Bid was routinely given to prevent infection 3 days after surgery. The bladder was squeezed 2 to 4 times daily to aid urination until the micturition reflex was restored.

Intralumbar injection was referred to Bey et al. (2017). Briefly as follows, a 30-gauge beveled needle was connected to a 10-μl syringe (Hamilton, Ghiroda, Romania), and the needle was inserted between lumbar vertebra 4 and lumbar vertebra 5; when the needle was properly inserted there will be a tail flick reflex, r-HAP1, gambogic amide, and PBS were epidurally administered via a syringe pump injection.

The behavioral tests were performed as previously described. Briefly, the locomotive function was evaluated by Basso Mouse Scale (BMS). The von Frey filament test was used to test the sensation threshold of pain, and the unilateral Hargreaves thermal test was used to test the sensation threshold of warmth. The cortical evoked motor potential examination was used to identify the Neuro-conduction capacity. All of this was tested by two independent assessors who were blinded to the experimental design and who were familiar with those assessment methods.

NSCs were isolated from the spinal cords of fetal mice on embryonic day 14.5 (E14.5). The spinal cords harvested from 3–4 E14.5 embryos were cut up, dissociated with 1 mL micropipettes, digested with the same volume of Accutase (StemPro, US) for 10 min, and centrifuged at 1,000 rpm for 5 min. The pellets were resuspended in a mouse NSC culture medium (Cyagen, China). The cells were seeded into T-25 culture flasks and incubated in a cell incubator at 37°C in a humidified atmosphere with 5% CO2. Single cells grew and accumulated into neurospheres for 5 days. For passaging, neurospheres were collected by centrifugation and then digested with accutase for 1–2 min at 37°C. The neurospheres were aspirated with 1 mL micropipettes approximately 20 times. The NSCs were reseeded with NSC culture medium at a density of 20 × 10⁴ cells/ml into T-25 culture flasks.

For differentiation, NSCs were seeded onto poly-lysine-coated (Sigma, USA) slides in 24-well cell culture plates and grown with different differentiation media for 7 to 14 days. For neuronal differentiation, NSCs were cultured in neuronal differentiation medium composed of neurobasal medium (Gibco, US), 2% B-27 (Gibco, US), 2 mM L-glutamine (Gibco, US), 2 μM all-trans retinoic acid (Sigma, US), 5 μM forskolin (MedchemExpress, China) and 0.05 g/L penicillin/streptomycin (Gibco, US). To identify the neurite outgrowth efficiency, NSCs were seeded at a density of 2.5 × 10⁴ cells/well and to verify the efficiency of differentiation, NSCs were seeded at a density of 10 × 10⁴ cells/well. To differentiate into astrocytes, NSCs were cultured in Dulbecco’s modified Eagle’s medium (DMEM)-F12/GlutaMAX (Gibco; US), 2% B-27, and 10% fetal bovine serum (Gibco, US). For oligodendrocyte differentiation, NSCs were cultured in an oligodendrocyte differentiation medium containing DMEM-F12/GlutaMAX, 2% B-27, and 200 ng/mL insulin-like growth factor 1 (Cyagen, China).

Cells were seeded onto cell culture plates; frozen sections were fixed with 4% paraformaldehyde for 10 min at room temperature, washed with phosphate-buffered saline (PBS) three times, and permeabilized with 0.1% Triton X-100 in PBS for 15 min. For BrdU (5-bromo-2′-deoxyuridine) labeling, the cells and sections were treated with 2 M HCl for 30 min and neutralized with 0.1 M barium tetraborate for 20 min. The cells and sections were blocked with 4% BSA in PBST for 1 h and then incubated with the appropriate primary antibody against the following: nestin (ab6142, Abcam, 1:200 dilution), Tuj (ab18207, Abcam, 1:400 dilution), O4 (MAB345, Millipore, 10 μg/mL), GFAP (ab7260, Abcam, 1:1,000 dilution), NeuN (ab104225, Abcam, 1:1,000 dilution), and BrdU (ab1893, Abcam, 1:400 dilution) at 4°C overnight. The cells were washed three times with PBS, incubated with the appropriate secondary antibody (Alexa Fluor^® 647 Goat Anti-Rat IgG, ab150167, Abcam, 1:500 dilution; Alexa Fluor^® 488 Goat Anti-Rabbit IgG, ab150077, Abcam, 1:500 dilution; Alexa Fluor^® 647 Goat Anti-Rabbit IgG, ab150083, Abcam, 1:500 dilution; Alexa Fluor^® 488 Rabbit Anti-Sheep IgG, ab1501817, Abcam, 1:500 dilution; Alexa Fluor^® 488 Goat Anti-Mouse IgG H&L, ab150113 Abcam, 1:500 dilution) at room temperature for 90 min, and stained with DAPI.

To calculate the length of the neurite (Wang et al., 2017), representative regions were harvested by Zeiss microscope. The neurite was manually traced by ImageJ software, and a schematic diagram of neurons and neurites was portrayed. The total neurite length of each neuron was measured by ImageJ software, and the average length of a random zone from each biological replicate was calculated. The target regions of the cell slide or the representative sections were randomly captured, and the positive cells were counted independently by two people. Each experiment was biologically repeated three times.

To investigate signal changes, a Cignal Finder Reporter Array plate (Qiagen, US) was used according to the manufacturer’s instructions. Briefly, 50 μL of NSC complete medium was added to each well of the Cignal Finder Array plate. The side of the plate was gently tapped to resuspend the constructs. Then, the plate was slightly rocked back and forth and left and right before being incubated for 5 min at room temperature. Next, 0.6 μL of attractions in 50 μL of NSC complete medium was added to each individual transfection. The plate was incubated for 20 min at room temperature to allow complex formation to occur. Simultaneously, NSCs were isolated as described previously. The cells were suspended in 5,400 μL of complete medium at a density of 20 × 10⁴ cells/ml. A total of 50 μL of cell suspension was added to each well containing construct-attractive complexes. The plate was mixed gently by rocking. The cells were incubated at 37°C in a 5% CO2 incubator for 12 h, and then recombinant HAP1 (r-HAP1, 15 ng/mL; BIOMATIK) was added. Twelve hours later, a luciferase assay was performed using a dual-luciferase reporter assay system. The luminescence signal was quantified by following the manufacturer’s instructions by normalizing the ratios of cellular responses to r-HAP1 (firefly luminescence signal) to the nonspecific responses to PBS (renilla luminescence signal). The r-HAP1-treated cells were compared with the PBS-treated control cells, and the results were plotted as a log2-fold change. If log2 > 1, the difference was considered significant.

After cells were harvested, ice-cold Dulbecco’s PBS (DPBS) was used to wash cells three times, and RIPA buffer with protease inhibitor and phosphatase inhibitor was used for cell lysis. Cellular protein was collected by centrifugation of the cell lysates at 12,000 rpm for 10 min. The protein concentration was measured with a BCA protein quantitation kit (Thermo Scientific, USA). Total proteins were separated by 10% SDS-PAGE and then transferred to nitrocellulose membranes. The membranes were blocked with 5% nonfat dry milk in tris-buffered saline/0.5% Tween. Then, the cells were incubated with the appropriate primary antibody against the following: ERK 1/2 (ab184699, Abcam, 1:10,000 dilution), phosphorylated ERK1/2 (ab76299, Abcam, 1:10,000 dilution), TrkA (ab76291, Abcam, 1:5,000 dilution), and phosphorylated TrkA (ab193229, Abcam, 1:1,000 dilution) before being incubated with the appropriate secondary antibody (Goat Anti-Rabbit IgG, ab216773, Abcam). Pierce™ ECL plus western blotting substrate was used to detect the proteins.

All data are expressed as the mean ± standard deviation (SD) of three independent experiments. The results were analyzed by a two-tailed Student’s test for comparisons between two groups, and one-way ANOVA with Tukey’s multiple comparisons test for multiple groups. Statistical and graphical analyses were performed using the software GraphPad Prism 6 (GraphPad Software, USA). p < 0.05 indicated statistical significance.