Nine newly purchased experimental Bama pigs (3-month-old, 11.40 ± 1.68 kg, China) were adaptatively fed for 1 week. All pigs were housed and underwent experiments in Large Animal Laboratory, Center of Experimental Animals of Capital Medical University, and kept in a humidity- and temperature-controlled environment with a 12-h light-dark cycle. Animal experiments complied with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health (NIH Publication No. 8523, revised 2011, United States). All experimental projects and protocols were approved by the Medical Ethics Committee of Capital Medical University (AEEI-2019-098).

Prior to surgery, all pigs were randomly divided into three groups (Figure 1A): sham group (n = 3), osteotomy only with normal motor evoked potential (MEP); incomplete distraction spinal cord injury (IDSCI) group (n = 3): osteotomy with DSCI, MEP amplitude decreased by approximately 75%; and complete distraction spinal cord injury (CDSCI) group (n = 3): osteotomy with DSCI, MEP amplitude decreased to 0. Any abnormal MEP signals compared with the baseline amplitude were considered to imply accidental iatrogenic SCI, and we excluded these animals from the experimental group. The spinal cord-evoked potential was monitored using an electrophysiological monitor (Cadwell, United States) during the operation. The stimulating electrodes were placed in the C3-C4 cortical motor area, and the single stimulation intensity was 100–200 V. Needle-like recording and reference electrodes were placed on the gastrocnemius muscles of both hind limbs, and the distance between the two needle electrodes was approximately 1.5 cm. The compound muscle action potential induced by a single electrical stimulus was recorded. MEP was recorded and a baseline was established after successful anesthesia. MEP was monitored throughout the spinal cord during placing screws, osteotomy, distraction (Figure 1A). Applying longitudinal distraction force to the vertebra to stretch the spinal cord until the MEP amplitude decreased by approximately 75% or 100%, and maintained the amplitude changes for 10 min. Animals were sacrificed 7 days after magnetic resonance imaging (MRI), and spinal cords were collected and processed for histological, molecular and biochemical analyses.

As shown in Figures 1B–E, the schematic of DSCI surgical procedure was displayed. A midline longitudinal incision extending from T13 to L2 was made, and the laminae from T14, T15, and L1 were exposed with coagulation cautery; 4.0 mm*25 mm short coccygeal pedicle screws (Weigao Orthopedic Materials Co., Ltd, China) were infixed into the vertebrae of T14 and L1. SINO rods of suitable length (Weigao Orthopedic Materials Co., Ltd, China) were installed on the pedicle screws on the right and left sides for temporary fixation. Global column osteotomy was performed at T15 under the protection of a root retractor, and the T15 vertebral body and adjacent discs were completely resected. A spreader was used to pull the spine at 1 mm intervals gradually. The distraction operation was performed according to MEP changes. The distance between the pedicle screws of T14 and L1 was measured before and after the DSCI procedure. After the muscles and skin were sutured, rigorous wound care was performed every 3 days. 1.5 g Cefuroxime sodium was administered for 3 days after surgery to prevent incision infection. Additionally, the bladder of each pig was manually emptied twice per day, and rigorous care of the perineum was taken to prevent urinary infection.

The animals were made to fast for 24 h before anesthesia. We performed the operation under anesthesia with a solution of 3% pentobarbital sodium injected intramuscularly. Subsequently, the pigs were treated with an oxygen mask and isoflurane inhalation anesthesia. Each pig was intubated with an endotracheal tube and maintained under inhalation anesthesia without a muscle relaxant (Kaiser et al., 2006). The depth of anesthesia was monitored using pain and corneal reflex tests. The depth of anesthesia was adjusted at all times to ensure that the pig breathed smoothly to prevent convulsions during image acquisition. Three vital signs, including rectal temperature, respiratory rate, and heart rate, were continuously monitored using a multifunctional bedside physiological monitor (PVM-2701; Nihon Kohden Corporation, Tokyo, Japan).

At the end of the DSCI model generation surgery, a wake-up test was carried out to prevent any false-negative or false-positive MEP signals from occurring. Approximately 30 min after osteotomy stabilization in animal model generation surgery, the pigs were awakened from anesthesia. We tested and observed lower extremity movements and sensory responses to mechanical stimulation.

Neurologic function assessment for pigs was performed at 1, 3, and 7 days postoperatively. According to the modified Tarlov score, we assessed pairs of hind limbs of the pigs. The modified Tarlov scoring was as follow (Assina et al., 2008): 0 point: no voluntary movement; 1 point: barely perceptible movement; 2 points: frequent movement of hind limbs, no weight support; 3 points: alternate stepping or propulsive movement, no weight support; 4 points: hind limbs can support weight; 5 points: ambulation with mild deficit; and 6 points: normal ambulation.

In addition, we assessed the muscle strength in each group. Muscle strength was divided into five grades according to the Medical Research Council (MRC) scale for muscle strength (Compston, 2010): grade 5: normal muscle strength; grade 4: muscle strength is reduced, but muscle contraction can still move the joint against resistance; grade 3: muscle strength is further reduced such that the joint can be moved only against gravity with the examiner’s resistance completely removed; grade 2: muscle can move only if the resistance of gravity is removed; grade 1: only a trace or flicker of movement is seen or felt in the muscle or fasciculations are observed in the muscle; grade 0: no movement is observed.

To get better MRI images of the central spinal cord lesions, we removed internal fixation in pigs. Pigs were maintained under isoflurane inhalation anesthesia throughout the examination. MRI of the porcine spinal cord was performed using a 3.0-T MRI scanner (Siemens, Berlin, Germany) 1 week after DSCI. Fifteen consecutive sagittal T2-weighted images were obtained by scanning the surgical regions of the pigs using a double-tuned volume radiofrequency coil. The parameters were set as follows: repetition time/echo time (TR/TE), 3500/103 ms; slice thickness, 3 mm; and slice gap, 10 percent. T2-weighted images were used to calculate the T2 intensity at the central DSCI lesion region (T15) using ImageJ software (National Institutes of Health, United States) (Yao et al., 2021).

Seven days after DSCI, spinal cord samples were acquired after sacrifice, and the specimens were fixed in buffered formalin. After being immersed in 4% paraformaldehyde for 24 h and embedding in paraffin, a serial cross-section was made at the central DSCI lesion region. Hematoxylin and eosin (H&E) staining and Luxol fast blue (LFB) staining were used to evaluate morphological and structural changes in the white and gray matter, nerve sheath, and general spinal cord conditions in the pigs of different groups.

An enzyme-linked immunosorbent assay (ELISA) was carried out on samples from central DSCI lesions and cerebrospinal fluid (CSF) at 7 days after DSCI. Twenty milliliters of CSF were collected for ELISA before the pigs were sacrificed (Duan et al., 2018). After the spinal cord tissue was homogenized and centrifuged at 4000 rpm for 10 min, the liquid supernatant and cerebrospinal fluid were detected at 450 nm wavelength according to the manufacturer’s instructions for the IL-1β, IL-6, and TNF-α ELISA kits (Shanghai Jianglai Biological Technology Co., Ltd, China). The ELISA samples were run in triplicates.

After the pigs were sacrificed, spinal cord sections (1 cm) from the central DSCI lesion region were collected and embedded in paraffin. Spinal cord sections (5 μm thick) from each specimen were deparaffinized with xylene and incubated in graded concentrations of ethanol. They were then washed with phosphate-buffered saline (PBS) for 3 × 5 min. The sections were incubated for blocking with a blocking solution (0.1% Triton X-100 in PBS and 10% normal goat serum) at room temperature for 2 h. The sections were incubated overnight with primary antibodies at 4°C. The primary antibodies used were anti-GFAP (1: 2000, GeneTex), anti-Iba-1 (1:200, Affinity), and anti-NeuN (1:100, Abcam). After rinsing with PBS for 3 × 5 min, the sections were incubated with secondary antibodies for 2 h at room temperature. The secondary antibody used in this study was goat anti-rabbit antibody (Alexa Fluor^®594) (1:200, Abcam). Following three rinses with PBS, a drop of antifade mounting medium containing DAPI (Solarbio Biotechnology, China) was placed on each slide. Finally, a coverslip was placed on top of each sample. Immunofluorescence imaging was carried out using an Olympus fluorescence microscope.

The spinal cord tissues were removed from the liquid nitrogen, and the spinal cord was cut into small pieces and subsequently homogenized in radioimmunoprecipitation assay lysis buffer. The supernatant was obtained after centrifugation at a low temperature, and the protein concentration was determined using a bicinchoninic acid kit. Proteins were separated by gradient sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electrophoretically transferred to polyvinylidene difluoride (PVDF) membranes. PVDF membranes were blocked with 5% non-fat dry milk for 2 hours and subsequently incubated with primary antibodies overnight at 4°C. The primary antibodies used were as follows: anti-Bcl-2 (1:2000, GeneTex), anti-Bax (1:1000, Biorbyt), anti-Caspase-3 (1:2000, Abcam), and anti-β-actin (1:5000, Proteintech). The washed membrane was incubated with goat anti-rabbit or goat anti-mouse IgG HRP-conjugated secondary antibody (1:5000). At last, PVDF membranes were exposed using a Tanon 5200 chemiluminescence image analysis system. Following exposure, ImageJ software (National Institutes of Health, United States) was used to analyze the gray values of bands.

Statistical analyses were performed using GraphPad Software (United States). A student’s t-test was used to compare group means, and one-way analysis of variance (ANOVA) was used to compare multiple samples. Data are presented as the mean ± standard deviation (SD). Changes in muscle strength and modified Tarlov scores from baseline to day 7 were assessed with a generalized linear mixed model using PROC GLIMMIX in SPSS 22.0 (United States). A Bonferroni adjustment was carried out for multiple comparisons of the three groups. The interaction between time and degree of DSCI was also analyzed in the model. Statistical significance was considered *p < 0.05, **p < 0.01, and ***p < 0.001 vs. the sham group.

As shown in Figure 2, the muscle strength of the hind limbs and the modified Tarlov score in the sham group maintained nearly normal at the 1, 3, 7 days after DSCI. At each time point, both the levels of muscle strength and modified Tarlov score in the two DSCI groups showed a significant decrease compared with the sham group (p < 0.001). Over time, the muscle strength of pigs in the IDSCI and CDSCI groups did not recover, and time was a factor influencing muscle strength in the two groups (F = 600.00; p < 0.001). For the two DSCI groups, muscle strength and modified Tarlov score of the CDSCI group reduced more significantly. ANOVA for repeated measures indicated significant differences among the three groups reciprocally in hind limb muscle strength (p < 0.001). Differences in the modified Tarlov score defects among the three groups were statistically significant. The modified Tarlov scores of the IDSCI and CDSCI groups did not improve over time (F = 852.76; p < 0.001).

We detected the imaging differences in the DSCI lesion area using MRI in live pigs of the three groups. As revealed in Figure 3A, T2-weighted MRI showed that hyperintense areas corresponded to central DSCI lesions areas of spinal cord in the IDSCI and CDSCI groups. Compared with the sham group, the increased enhancement in T2-weighted MRI was observed at the central DSCI lesions in the IDSCI (p < 0.001) and CDSCI groups (p < 0.01) (Figure 3B). And the relative intensity in T2-weighted MRI was higher in the CDSCI group with a higher degree of DSCI. The results may indicate that edema, inflammation, demyelination, axon loss, and astrogliosis occurred following DSCI.

As displayed in Figure 4, H&E staining showed a normal gray and white matter structure and normal neuronal morphology in the sham group. In the IDSCI group, it was revealed that damaged tissue structure, a decreased number of neurons, cell edema, and hyperemia in the spinal cord tissue. Injury to the spinal cord structure in the CDSCI group was more severe than that in the IDSCI group. The number of neurons decreased, hyperemia and edema were significant, alongside significant inflammatory cell infiltration.

Moreover, LFB staining stained myelin fiber deep blue. It was shown that a clear boundary differentiated from the surrounding structures in the sham group. The spinal tissue of IDSCI group showed disordered white matter arrangement, decreased density of nerve fiber bundles, and edema of the myelin sheath. In the CDSCI group, the white matter structure was further destroyed, and nerve fiber bundles showed a sparse network; in addition, myelin edema was obvious.

To detect the effect of different degrees of DSCI on the level of inflammatory cytokines in spinal cord tissues and CSF at 7 days after surgery, inflammatory cytokines were detected by ELISA (Figure 5). IL-1β, IL-6, and TNF-α levels in the two DSCI groups were significantly increased in spinal cord tissues after DSCI compared to the sham group (p < 0.001) (Figures 5A–C). Compared with the IDSCI group, the levels of IL-1β (p = 0.04), IL-6 (p = 0.03), and TNF-α (p < 0.001) were significantly higher in the CDSCI group.

Similarly, the levels of IL-1β, IL-6, and TNF-α in the CSF of the CDSCI group. These results indicate that the levels of inflammatory cytokines in CSF were elevated significantly in the two DSCI groups (p < 0.05) (Figures 5D–F). Among them, the concentration of IL-1β (p = 0.001), IL-6 (p < 0.001), and TNF-α (p = 0.002) were higher in the CSF of CDSCI group. These results indicate that the levels of inflammatory cytokines in spinal cord tissues and CSF increased with the increase in DSCI degree.

To assess the neuroinflammatory responses and neuron survival in the central DSCI lesions 7 days after DSCI, we conducted immunofluorescence staining to detect the expression levels of GFAP, Iba-1, and NeuN, as presented in Figure 6. We measured the expression of GFAP, a major component of scar matrices, to study the gliosis process after DSCI (Figure 6A). Iba-1, a major cell type involved in neuroinflammation, was also used to explore microglial activation (Figure 6B). GFAP and Iba-1 fluorescence was slightly expressed in the sham group. The optical density of GFAP and Iba-1 in the IDSCI and CDSCI groups was significantly increased compared with that in the sham group (p < 0.05) (Figures 6D, E). The optical density of GFAP and Iba-1 were higher in the CDSCI group than that in the IDSCI group. The number of GFAP-positive astrocytes increased with the DSCI degree increasing, pointing to gliosis hyperplasia. NeuN staining was used to assess neuronal survival after DSCI (Figure 6C). The expression level of NeuN in the IDSCI and CDSCI groups was significantly lower than that in the sham group (p < 0.05), and the CDSCI group presented a lower level between the two DSCI groups (Figure 6F).

As represented in Figure 7, the expression levels of P53/Bcl-2/Bax/Caspase-3 apoptosis signaling pathway-related proteins were detected by western blotting. The results revealed that the expression levels of P53, Bax, and Caspase-3 proteins were significantly increased in the two DSCI groups (p < 0.05) except Bax protein in the IDSCI group, and the CDSCI group having a higher expression level. While Bcl-2 protein expression was significantly decreased in the two DSCI groups (p < 0.05), the decrease degree was more significant in the CDSCI group.