Animals

A total of 54 adult (8-week-old) female Sprague–Dawley (SD) rats obtained from Beijing Vital River Laboratory Animal Technology Co., Ltd., were used for the study. The rats were housed two per cage in a specific pathogen-free housing facility with the temperature maintained at 22–24 °C, a relative humidity of 45%–60%, and a 12 h day/night cycle (lights on at 8:00 am). Food and water were provided ad libitum. The animals were left undisturbed before they were acclimated to the environment for at least one week. Animals were allowed to be fully adapted to the researchers and experimental setups. Necessary sedatives and anesthetics including tiletamine-zolazepam (Zoletil 50; Virbac S.A., Carros, France) (20 mg/kg) and xylazine (2.5 mg/kg) were administered to minimize their suffering. The animals were grouped via simple randomization (assigning subjects to groups randomly by generating random numbers, ranking them, and assigning them to predetermined groups to ensure unbiased allocation in an experiment) and their group identities were blinded to the researchers conducting functional assessments and image analysis (Bespalov et al. 2020). All the animal procedures were preapproved by the Institutional Animal Care and Use Committee at Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology (ethics approval number, TJH-201903007).

Spinal Cord Injury

At the age of 9–11 weeks, the rats were randomly assigned to the SCI or sham group. The procedure for spinal cord contusion at the low thoracic level was similar to that described previously (Gong et al. 2021; Guo et al. 2021). Briefly, rats were anesthetized by intramuscular injection of tiletamine–zolazepam (Zoletil 50; Virbac S.A., Carros, France) (20 mg/kg) and xylazine (2.5 mg/kg). The back fur was shaved, and the bladder was evacuated before the animal was placed in a prone position on the operating table. Ceftriaxone (50 mg/kg) was administered intraperitoneally to prevent infection. Both eyes were covered with eye ointment to prevent xerophthalmia during anesthesia. Under aseptic conditions, the 9th thoracic vertebra (T9) was palpated and marked on the skin of the back. After skin disinfection, a longitudinal midline incision was made at the marked site, and the paraspinal muscles were detached from the vertebrae. T9 laminectomy was performed to expose the dorsal surface of the spinal cord, which was then stabilized in a small animal spinal holder (RWD, Inc., China) by clamping the T8 and T10 vertebrae with stabilizing forceps. A modified NYU impactor was used to drop a 10 g metal rod from a height of 25 mm on the exposed spinal cord. The rod was removed immediately after the drop. Swing of both hindlimbs and flutter of the tail indicated successful impact. Edema and congestion could be observed in the spinal cord tissue. For rats in the sham group, the spinal cord was similarly exposed and stabilized without contusion. Then, the wound was rinsed with normal saline and sutured in layers. The animal was placed on a 37 °C heating pad throughout the surgery until arousal. After the operation, the rats received 5 ml of Ringer’s solution and 5 mg/kg carprofen subcutaneously for 3 days to prevent dehydration and pain, respectively. Ceftriaxone (50 mg/kg) was administered intraperitoneally for the following week to prevent infection. Bladders were expressed manually three times daily until spontaneous voiding was restored.

Functional Assessments

The first batch consisted of 10 sham rats and 14 SCI rats and underwent behavioral tests for motor and sensory function assessments. One sham rat with a BBB score of 10.5 at 1 day post-surgery and four SCI rats with BBB scores below 9 (incapable of weight support that is required for von Frey tests) were excluded, reducing the animal numbers to n = 9 for sham group and n = 10 for SCI group. All behavior tests were conducted by researchers blinded to the group allocation.

Basso, Beattie, and Bresnahan Locomotor Scores

To monitor locomotor function, the BBB open field 21-point rating scale (Basso et al. 1995) was applied before the injury, 1 day after SCI and weekly thereafter until 6 weeks post injury (Fig. 1A). The rats were acclimated to the open field twice daily for 5 consecutive days before the first examination. Each examination was conducted by two individual observers with one responsible for each side.

Fig. 1

Development of below-level NP following SCI. A Experimental design. B Representative images of brains and spinal cords. C Enlarged views of the laminectomy or injury site from (B). D Representative images of H&E staining. E Quantification of lesion volume. Mann Whitney test, p = 0.0022; n = 6 per group. F Body weight change from baseline to 6 weeks post-injury (wpi). Two-way repeated measures ANOVA, F (1, 17) = 0.0019, p = 0.9661, ηp2 = 0.0001. G Quantification of body weight at 6 wpi. Sidak’s test: p > 0.9999. H Assessment of locomotion via the BBB score. Two-way repeated measures ANOVA, F (1, 17) = 4291, p < 0.0001, ηp2 = 0.9961. I Quantification of the BBB score at 6 wpi. Sidak’s test: p < 0.0001. J Assessment of mechanical allodynia via the von Frey test. Two-way repeated measures ANOVA, F (1, 17) = 50.16, p < 0.0001, ηp2 = 0.8701; Sidak’s test: pre-injury, p = 0.9077; 4 wpi, p < 0.0001; 6 wpi, p < 0.0001. K Assessment of thermal hyperalgesia via the Hargreaves test. Two-way repeated measures ANOVA, F (1, 17) = 65.67, p < 0.0001, ηp2 = 0.7055; Sidak’s test: pre-injury, p = 0.3338; 4 wpi, p < 0.0001; 6 wpi, p = 0.0001. Scale bar, 1 mm in (B and D). ns, not significant, **p < 0.01, ***p < 0.001, ****p < 0.0001 SCI vs. Sham

von Frey Test

The mechanical withdrawal threshold was evaluated with an electronic von Frey Anesthesiometer (IITC Life Sciences Inc., 2391) as previously described (Vivancos et al. 2004). Briefly, rats were placed in the plexiglass chamber on a wire mesh floor in a quiet room for 45 min of habituation starting 3 days before testing. On the testing day, the rats were allowed to acclimate for 15 min. A tilted mirror was placed under the mesh floor to provide a clear view of the hind paw, at the center of which a rigid tip (diameter 1 mm) attached to the recording apparatus was applied with a gradual increase in pressure. When the rats felt pain and withdrew their hind paws, the stimulus was automatically discontinued, and the force was recorded and displayed by the device. The procedure was repeated six times with an interval of at least 1 min on each paw. With the maximum and minimum excluded to decrease variability, the remaining four values were averaged. To monitor injury-induced mechanical allodynia, rats were assessed before surgery and biweekly until 6 weeks post injury (Fig. 1A).

Hargreaves Test

The thermal withdrawal latency was evaluated with a Hargreaves plantar test apparatus (IITC Life Sciences Inc., 390) as previously described (Hargreaves et al. 1988). Briefly, rats were placed in the plexiglass chamber on a transparent glass stand in a quiet room for 45 min of habituation starting 3 days before testing. On the testing day, the rats were allowed to acclimate for 15 min. A tilted mirror was placed under the glass to provide a clear view of the hind paw, at the center of which a light beam (30% of maximum intensity) was focused continuously. When the rats felt pain and withdrew their hind paws, the light was turned off, and the stimulation time was recorded and displayed by the device. A 20 s automatic cutoff was set to prevent tissue damage. The procedure was repeated six times with an interval of at least 1 min on each paw. With the maximum and minimum excluded to decrease variability, the remaining four values were averaged. To monitor injury-induced thermal allodynia, rats were assessed before surgery and biweekly until 6 weeks post injury (Fig. 1A).

Peripheral Stimulation Paradigms

At 6 weeks post injury, the bilateral hind paws of the rats were stimulated before sacrifice (Fig. 1A). There were five different types of stimuli that varied in sensory modality and intensity: (1) no stimulus, (2) innocuous mechanical stimulus, (3) innocuous thermal stimulus, (4) noxious mechanical stimulus and (5) noxious thermal stimulus. Accordingly, rats were randomly assigned to five groups receiving one specific type of stimulus. Each group consisted of equal numbers of SCI rats and sham rats. During the study, two out of 56 animals were lost: one sham rat and one SCI rat in the noxious thermal stimulation group, due to anesthesia-related complications and urinary tract infection, respectively. All the stimulating manipulations were performed in the afternoon, and the rats were habituated to the experimental environment three days in advance to avoid excessive locomotor activity due to stress. Care was taken to manipulate the SCI and sham rats under the same conditions for each stimulation type.

Rats receiving no stimulus were lightly restrained in a rodent restraint bag for 10 min and then allowed to stay awake in the experimental environment for 2 h just before euthanasia. For innocuous stimulations, rats were lightly restrained in a rodent restraint bag while being stimulated. To induce an innocuous mechanical stimulation, the hind paws of the rats were brushed from heel to toe with a paintbrush. Each stroke lasting 2 s was delivered once every 4 s for 10 min (Catheline et al. 1999; Liu et al. 2018; Cao et al. 2022). To induce an innocuous thermal stimulus, both hind paws of the rats were immersed up to the ankle joints in warm water maintained at 45 °C in a water bath (HH.S11-1, Shanghai Boxun Industry Co., Ltd.) for 20 s each time, which has been proven to be innoxious for rodents (Abbadie et al. 1994). The stimulation was repeated at 20 s intervals for 10 min in total.

To reduce animal suffering, noxious stimulation was administered to the rats under anesthesia. Tiletamine–zolazepam (Zoletil 50; Virbac S.A., Carros, France) (50 mg/kg) and xylazine (2.5 mg/kg) were intraperitoneally injected to induce a state of deep anesthesia with only the corneal reflex and slight pain reflexes maintained (Williams et al. 1990; Abbadie et al. 1994; Catheline et al. 1999; Rahman et al. 2002). Noxious mechanical stimulation consisted of repeated pinching of folds of skin (6 points each on the dorsal and ventral surface of the hind paw) for 10 s with a hemostat (J31010, Shanghai Medical Instruments (Group) Ltd.) (Spike et al. 2002; Polgár et al. 2013). To ensure that the same stimulus was applied each time and to each rat, the hemostat were locked to the tightest ratchet position for each application. Noxious thermal stimulation was induced by immersion of both hind paws up to the ankle joints in hot water maintained at 52 °C in a water bath (HH.S11-1, Shanghai Boxun Industry Co., Ltd.) for 20 s (Williams et al. 1990; Catheline et al. 1999; Spike et al. 2002; Polgár et al. 2013). Following noxious stimulation, anesthetized rats were placed on a heating pad at 37 °C until euthanasia.

Tissue Processing

Rats were anesthetized 2 h after stimulus initiation and transcardially perfused with heparinized (100,000 U/L) normal saline followed by 4% paraformaldehyde (PFA). The brains and spinal cords were harvested, postfixed in 4% PFA at 4 °C overnight and then immersed in 30% sucrose solution for 72 h. The lumbar enlargements (L4, L5) were dissected, flash frozen in optimal cutting temperature compound (Sakura, 4853) and sectioned transversely at a thickness of 30 μm using a cryostat (Leica, CM1950). Similarly, the brains were frozen and coronally sectioned at a thickness of 40 μm. Free-floating sections were stored in cryoprotectant consisting of 30% sucrose and 30% ethylene glycol in phosphate-buffered saline (PBS) at − 80 °C before immunostaining. For lesion analysis, serial horizontal Sects. (1.2 cm long, 20 μm thick) of thoracic segments containing the lesion site were collected spanning the dorsal to ventral axis and mounted on coated glass slides before histology analysis.

Lesion Analysis

For lesion analysis, 20 μm thick horizontal sections containing the injury epicenter (160 μm apart) were stained with hematoxylin and eosin (H&E) and imaged via brightfield microscopy. The lesion area and total spinal cord area within the 1.0 cm long sections were measured using the free-hand tool in Fiji software (NIH, Bethesda, MD; RRID:SCR_002285) (Schindelin et al. 2012) by a researcher blinded to the group allocation. The total lesion volume (lesion volume) and total spinal cord volume (total volume) were calculated using the Cavalieri method (Fan et al. 2013). This method is a summation of the measured area of each section multiplied by the intersection distance. The percent lesion volume was calculated using the following equation: % lesion volume = lesion volume/total volume × 100%.

Immunofluorescence Staining

For immunostaining, free-floating sections were rinsed three times in Tris-buffered saline containing 0.3% Triton-X100 (TBST) for 5 min each, blocked with Quick Block ™ blocking buffer (Beyotime, P0260) for 15 min, and then incubated overnight with primary antibodies in Quick Block ™ primary antibody dilution buffer (Beyotime, P0262) at 4 °C. Thereafter, the sections were rinsed three times with TBST for 5 min each and incubated with fluorochrome-conjugated secondary antibodies and Hoechst 33258 (Sigma, 94403) in Quick Block ™ secondary antibody dilution buffer (Beyotime, P0265) for 1 h in the dark. Finally, after being rinsed three times in TBST for 5 min each, the sections were mounted on glass slides, air-dried thoroughly, and cover-slipped with Prolong Glass Antifade Mount (Invitrogen, P36980). Antibody specificity was verified in prior research, with references provided in Supplementary Information.

The following primary antibodies were used: guinea pig anti-c-Fos (1:1000; Synaptic Systems Cat# 226308, RRID: AB_2905595), chicken anti-NeuN (1:1000; Millipore Cat# ABN91, RRID: AB_11205760), rabbit anti-PKCγ (1:1000; Abcam Cat# ab71558, RRID: AB_1281066), rabbit anti-Lmx1b (1:500; Abcam Cat# ab259926, RRID: N/A), mouse anti-Brn3a (1:200; Millipore Cat# MAB1585, RRID: AB_94166), rabbit anti-Pax2 (1:400; Biolegend Cat# 901001, RRID: AB_2565001), rabbit anti-NK1R (1:2000; Sigma Cat# S8305, RRID: AB_261562), rabbit anti-calretinin (1:1000; Swant Cat# 7697, RRID: AB_2721226), rabbit anti-calbindin D-28K (1:2000; Swant Cat# CB38, RRID:AB_10000340), rabbit anti-Vglut1 (1:500; Abcam Cat# ab227805, RRID: AB_2868428), guinea pig anti-Vglut2 (1:500; Synaptic Systems Cat# 135404, RRID: AB_887884), rabbit anti-Gad65/67 (1:1000; Abcam Cat# ab183999, RRID: N/A), guinea pig anti-Vgat (1:1000, Synaptic Systems Cat# 131004, RRID: AB_887873), rabbit anti-dopamine beta hydroxylase (DBH, 1:500; Abcam Cat# ab209487, RRID: AB_2892178).

The following secondary antibodies were used: donkey anti-guinea pig Alexa Fluor 488 (1:1000; Jackson ImmunoResearch Cat# 706–545-148, RRID: AB_2340472), donkey anti-rabbit Alexa Fluor 488 (1:1000; Invitrogen Cat# A-21206, RRID: AB_2535792), donkey anti-chicken Cy3 (1:1000; Jackson ImmunoResearch Cat# 703-165-155, RRID: AB_2340363), donkey anti-rabbit Cy3 (1:1000; Jackson ImmunoResearch Cat# 711-165-152, RRID: AB_2307443), donkey anti-mouse Cy3 (1:1000; Jackson ImmunoResearch Cat# 715-165-150, RRID: AB_2340813), donkey anti-rabbit Alexa Fluor 647 (1:1000; Invitrogen Cat# A-31573, RRID: AB_2536183), donkey anti-guinea pig Alexa Fluor 647 (1:1000; Jackson ImmunoResearch Cat# 706-605-148, RRID: AB_2340476), donkey anti-mouse Alexa Fluor 647 (1:1000; Invitrogen Cat# A-31571, RRID: AB_162542), and donkey anti-chicken Alexa Fluor 647 (1:1000; Jackson ImmunoResearch Cat# 703-605-155, RRID: AB_2340379).

Imaging and Quantification

Immunofluorescence images were captured using either an Olympus BX51 fluorescence microscope or an Olympus Fluor View FV3000 confocal microscope. Imaging settings remained unchanged for the same marker to avoid interference from variable acquisition conditions. Representative images are displayed as maximum projected z stacks or orthogonal projections. The co-labeling of cells for the expression of the c-Fos protein and other markers was verified in three dimensions, and the cells were counted by a blinded researcher using the plugin CellCounter in Fiji software independently of the c-Fos staining intensity (Coggeshall 2005). For the spinal cord, the numbers of c-Fos+ neurons were separately quantified in the conventional pain-recruited superficial DH (I–IIo) and innocuous-stimulus-activated deep DH (Iii–VI), which were divided by the dorsal border of the PKCγ immunolabeling band. The number of cells co-labeled for neuronal subtype markers and the c-Fos protein within the DH was counted. Six sections were selected from L4/L5 for immunofluorescence staining, of which the three most densely c-fos-labeled sections were quantified as the average value for each animal.

For response of the supraspinal nuclei, we explored the neuronal activity of these nuclei following noxious stimuli, under which a prominent SCI-induced effect at the spinal level was observed. The brain regions were defined from a reference atlas (Paxinos and Watson 2009), except for the LC, which was delineated via DBH immunostaining due to the major distribution of noradrenergic neurons. For supraspinal nuclei, the numbers of activated neurons and the area of the corresponding nuclei in coronal sections were quantified using Fiji software. The results were normalized and presented as the number of activated neurons per unit area (Brown et al. 2022). The values of three to six sections were averaged for one brain region.

For quantification of fluorescence intensity, images of the dorsal horn were acquired using an Olympus 10 × lens under the same scanning settings. Each was a z-stack consisting of 6 focal planes separated by 3 μm in the z-axis and processed as a 2D maximal projection in the 16-bit format using Fiji software. The threshold was adjusted to specifically select the positive area within the dorsal horn, and the mean fluorescence intensity was acquired from the selected region by the blinded researcher. For each animal, the values of 3 sections from L4/5 segments were averaged. The data were divided by the average values of all sham rats for normalization.

Statistical Analysis

GraphPad Prism Software 9.5 (San Diego, CA, USA; RRID: SCR_002798) was used to perform the statistical analyses and to generate the graphs. The sample size was determined based on prior studies involving c-fos analysis in the central nervous system (Brown et al. 2022; Cao et al. 2022; Targowska-Duda et al. 2023). For example, Brown et al. similarly investigated c-fos expression in the dorsal horn. Their two-way analysis of variance (ANOVA), with 3–4 biological replicates per group (4 groups total), yielded F(1, 10) = 8.53 and a partial eta squared of 0.46, confirming the statistical power of the sample size in this context (Brown et al. 2022). Parametric data are presented as the mean ± standard deviation (SD), and non-parametric data as median ± interquartile range. The Shapiro–Wilk test was used to evaluate the normality of the data, and the F test or the Brown-Forsythe test to assess the homogeneity of the variance. The analysis results were provided in the Supplementary Information. Nonparametric tests were applied when data were not normally distributed or the variance was not homogeneous. Unpaired Student’s t tests were used to analyze two groups of data, with Cohen’s d calculated to estimate the effect size (Lakens 2013). And Mann–Whitney U tests were applied for analysis of the non-parametric data with two groups. Behavioral data collected at multiple time points were analyzed using two-way repeated measures analysis of variance, followed by Sidak’s multiple comparison tests. To assess differences between the SCI and sham groups under each stimulation condition, two-way ANOVA followed by Sidak’s test was employed, with the effect size estimated via the partial eta squared (ηp2) (Richardson 2011; Lakens 2013). For data not meeting ANOVA assumptions (normality and homoscedasticity), the Scheirer–Ray–Hare test was used with eta squared (η2) indicating the effect size, followed by Mann–Whitney U tests with Bonferroni–Holm correction for post-hoc analysis. Differences were considered significant at p < 0.05 (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001).