A total of 36 Male Sprague-Dawley rats weighing 180-200 g (6-8 weeks old) were purchased from Hubei Province Experimental Animal Center. Animals were housed in a temperature-controlled room (22±1˚C) and 55±5% humidity with a 12/12 h light-dark cycle regime and access to water and food ad libitum. All efforts were made to minimize the number of animals used and their suffering.

OXA (Selleck Chemicals) was dissolved in 5% glucose solution (21). Res (3,5,40-trihydroxystilbene; Sigma-Aldrich; Merck KGaA) was dissolved in DMSO and diluted with 0.9% NaCl. Rats were randomly divided into four groups: Control, Res, OXA and OXA + Res. Each group contained six rats. For OXA treatment, the rats from the OXA and OXA + Res groups received intraperitoneal injection of 4 mg/kg OXA once daily for 5 consecutive days. The rats from the Control and Res groups received intraperitoneal injection of the same volume of 5% glucose solution once daily for 5 consecutive days. The pain behavioral test was performed 6 days after the first OXA injection. At 7 days after the first OXA injection, the rats from the Res and OXA + Res groups were intrathecally injected with Res (1 mg/kg). The Control and OXA groups were intrathecally injected with the same volume of vehicle (DMSO and 0.9% NaCl). Intrathecal injection was performed as previously described (22,23). Briefly, rats were anesthetized through an intraperitoneal injection of 50 mg/kg sodium pentobarbital and the skin was sterilized with 75% alcohol. Then, a 25-µl Hamilton syringe with a 30-gauge needle was held at an angle of ~20˚ above the vertebral column. The needle was inserted into the intervertebral space between L5 and L6. A puncture stimulating a tail-flick reaction was considered as successful. The Res solution (10 µl) was injected and the needle was left in the dosing position for >30 sec after injection. Subsequently, the mechanical allodynia assay was performed at 0, 2, 4 and 6 h after Res administration.

Anti-NF-κB p65 mouse monoclonal antibody (cat. no. BF8005) and β-actin rabbit antibody (cat. no. AF7018) were obtained from Affinity Biosciences. GFAP rabbit polyclonal antibody (cat. no. A0237), IL-1β rabbit monoclonal antibody (cat. no. A19635), TNF-α rabbit polyclonal antibody (cat. no. A0277) and COX-2 rabbit polyclonal antibody (cat. no. A1253) were purchased from ABclonal Biotech Co., Ltd.. The Reactive Oxygen Species Assay kit (cat. no. S0033S) and Mito-Tracker Red CMXRos (C1049B) were obtained from Beyotime Biotechnology. H&E staining solution (cat. no. BL735B) was purchased from Biosharp Life Sciences. Res (cat. no. 501-36-0) was purchased from Sigma. OXA (cat. no. 61825-94-3) was obtained from Shanghai Aladdin Biochemical Technology Co., Ltd. The secondary antibodies used for western blotting were HRP Goat Anti-Rabbit IgG (H+L) (AS014) and HRP Goat Anti-Mouse IgG (H+L) (AS014), purchased from ABclonal Technology. The secondary antibodies used for immunofluorescence analysis were Goat anti-mouse IgG H&L (FITC) (ab6785), Goat Anti-Rabbit IgG H&L (TRITC; cat. no. ab6718) and Goat Anti-Rabbit IgG H&L (FITC) (ab6717), purchased from Abcam.

The paw withdrawal threshold (PWT) was determined using the modified up-down method (24). Rats were accustomed individually for 30 min in a transparent plastic box with a 5x5 cm wire mesh grid floor. Subsequently, the von Frey filaments (0.4-26 g; Stoelting Co.) were used to stimulate the hind paw. The filaments were pressed vertically against the mid-plantar surface of the hind paw. The force caused filaments bent and this state was maintained for 3-5 s with a 2-min interval between two stimulations. A positive response was defined as paw withdrawal was observed within 3 sec of the stimulated hind paw. Six measurements were taken for each rat. If a positive response was obtained, a lower level of von Frey hair was used; conversely, the next higher force was used. The pattern of the positive and negative withdrawal response was converted to PWT. After the behavioral test, three rats were perfused and fixed for morphological analysis, including H&E staining and immunofluorescence. The other three rats were sacrificed and the spinal cord were collected for western blot analysis.

C6 glial cells (Jennio Biotech Co., Ltd.) were cultured in DMEM supplemented with 10% fetal bovine serum, 50 U/ml penicillin and 50 µg/ml streptomycin (all Gibco; Thermo Fisher Scientific, Inc.) at 37˚C with 5% CO₂. For ROS and mitochondrial membrane potential assessment, C6 glial cells (1.5x10⁵) were seeded on a 24-well plate and incubation with 5 ng/µl TNF-α at 37˚C for 4 h. Subsequently, cells were treated with 0 and 1 µM Res at 37˚C for 24 h and then the cells were used for ROS assessment and mitochondrial membrane potential detection. TNF-α was diluted with 0.9% NaCl. Res was dissolved in DMSO (Beyotime Biotechnology).

After 5 days of OXA administration and subsequent Res treatment, the rats from the Control, OXA and OXA + Res groups (n=3/group) were deeply anesthetized with 60 mg/kg sodium pentobarbital and perfused transcardially with saline containing heparin. The clearing of the liver (red to light brown in color) is an indicator of a good perfusion. Then switched perfusate to 4% paraformaldehyde (PFA, 0.1 M phosphate buffer, pH 7.4) and perfused until the animal body was stiff and rigid. After perfusion was complete, spinal cords were removed and post-fixed in 4% PFA (0.1 M phosphate buffer, pH 7.4) for 12 h at 4˚C and embedded in paraffin. Subsequently, the tissues were cut into 4-µm sections using a microtome (RM 2165; Leica Microsystems GmbH). The sections were stained using the standard H&E method. Briefly, paraffin sections were treated with xylene (5 min, two times), 100% ethanol (10 min, two times), 90% ethanol (10 min), 70% ethanol (10 min) for dewaxing and rinsing with tap water for 2 min. Then the sections were dyed with hematoxylin solution for 3 min and washed with tap water for 10 sec. The sections were stained with eosin for 3 min and washed with tap water for 10 sec. The dehydration and transparent treatment were conducted by putting the slices into 80% ethanol (5 min), 90% ethanol (5 min), 95% ethanol (5 min), 100% ethanol (5 min, two times), xylene (5 min, times). Finally, the sections were sealed with neutral balsam and observed using a fluorescence microscope (Olympus IX73; Olympus). H&E images were analyzed using the ImageJ 1.51j8 software (National Institutes of Health). The scoring criteria of inflammation cell infiltration (25) is: 0 (normal); 1 (lymphocyte infiltration around meninges and blood vessels); 2, 1-10 lymphocytes in a field); 3 (11-100 lymphocytes in a field); 4 (>100 lymphocytes in a field).

For immunofluorescence analysis, the paraffin sections were treated with xylene (5 min, two times), 100% ethanol (10 min, two times), 90% ethanol (10 min), 70% ethanol (10 min) for dewaxing and rinsing with ddH₂O (5 min, two times). Then the sections were conducted to antigen retrieval (Improved Citrate Antigen Retrieval Solution, P0083, Beyotime Biotechnology). Immerse the slices in antigen retrieval solution and heat at 95-100˚C for 20 min. The antigen retrieval solution was preheated to 95-100˚C before use. Then cooled to room temperature and washed 1-2 times with distilled water for 3-5 min. After incubation with 3% hydrogen peroxide for 10 min, immunostaining was performed. The tissue sections were blocked with goat serum (Gibco, 10%, room temperature) for 1 h and then incubated with primary antibody overnight at 4˚C. Subsequently, the sections were incubated with fluorescent secondary antibody at room temperature for 1 h and observed under a fluorescence microscope (Olympus IX73; Olympus Corporation). The fluorescence intensities were analyzed using ImageJ 1.51j8 (National Institutes of Health). The following primary antibodies were used: Anti-IL-1β (1:100), anti-GFAP (1:100), anti-COX-2 (1:100) and anti-NF-κB (1:100).

6 h after the Res treatment, three rats were euthanized with an overdose of pentobarbital sodium (100~150 mg/kg) by intraperitoneal injection and sacrificed through decapitation. Lumber spinal cord samples were collected and homogenized in RIPA lysis buffer containing 1% protease inhibitors (Sigma-Aldrich; Merck KGaA). After centrifugation at 12,000 g, 4˚C for 20 min, the supernatant was collected. For cells, after treatment, the cells in the cell culture dish were collected in a centrifuge tube. The cells were lysed in ice-cold RIPA buffer, after cell lysis, one third volume of 4x sample loading buffer was added and heated in boiling water bath for 10 min and ultrasonic treatment of 10~15 sec. After centrifugation at 12,000 g, 4˚C for 20 min, the supernatant was collected. Protein concentration (tissue and cells) was quantified using a BCA analysis kit (Beyotime Biotechnology). Protein lysates (40 µg/lane) were separated on 10% SDS-PAGE and transferred to 0.22 µm PVDF membranes. Membranes were blocked with 5% milk in TBST (0.1% Tween-20) for 1 h at room temperature. Then the membranes were incubated with specific primary antibodies at 4˚C overnight and HRP-conjugated secondary antibodies in TBST (1:5,000) at room temperature for 1 h. Protein bands were visualized using ECL detection reagent (Biosharp Life Sciences) and detected with an iBright 1500 instrument (Invitrogen; Thermo Fisher Scientific, Inc.). The grey values of western blot bands were analyzed using ImageJ 1.51j8 software (National Institutes of Health). β-actin was used as a loading control. The following primary antibodies were used: Anti-IL-1β (1:1,000), anti-GFAP (1:1,000), anti-COX-2 (1:1,000), anti-NF-κB (1:1,000), anti-TNF-α (1:1,000) and anti-β-actin (1:1,000).

Reactive Oxygen Species Assay Kit uses fluorescent probe DCFH-DA for ROS detection. The DCFH-DA was diluted with serum-free medium at 1:1,000 to a final concentration of 10 µM. The cell culture medium was removed and diluted DCFH-DA was added. Cells were incubated at 37˚C for 30 min. After incubation, cells were washed three times with serum-free medium to adequately remove DCFH-DA that did not enter the cell. Then the cells were placed under a fluorescence microscope (Olympus IX73; Olympus) for observation. The fluorescence intensities were analyzed using ImageJ 1.51j8 software.

Following Res treatment, the cell culture medium is removed and the Mito-Tracker Red CMXRos working solution was added, and the cells are incubated at 37˚C for 20 min. Then the Mito-Tracker Red CMXRos working solution was removed and the fresh cell culture medium was added. The cells were observed by fluorescence microscope (Olympus IX73; Olympus). The fluorescence intensities were analyzed using ImageJ 1.51j8 software (National Institutes of Health).

The X-ray crystal structure of COX-2 was obtained from the Protein Data Bank (PDB ID: 5f19; rcsb.org/structure/5F19). The structure of Res was downloaded from the PubChem database (pubchem.ncbi.nlm.nih.gov/compound/445154) and optimized using ChemBio3D Ultra 14.0 software (PerkinElmer Informatics). Auto Dock Vina 1.2.0 software (Center For Computational Structural Biology) was used to dock conformation between COX-2 and Res. PyMOL 2.2.3 (PyMOL by Schrödinger, DeLano Scientific LLC; pymol.org/installers/) was used to visualize the conformation.

All statistical analyses were performed using SPSS 21.0 statistics software (IBM Corp.). Data obtained through H&E staining, immunofluorescence and western blotting were analyzed using one-way analysis of variance followed by Tukey's test and presented as the mean ± SD. PWT data were analyzed using unpaired Student's t-test and presented as the mean ± SEM. P<0.05 was considered to indicate a statistically significant difference. All the experiments were repeated three times.

To investigate the effect of Res on neuropathic pain, an OXA-induced neuropathic pain rat model was established by once-daily intraperitoneal injection of OXA for 5 consecutive days. On day 6 after the OXA administration, the PWT value indicating mechanical pain sensitivity was recorded. The PWT values in the OXA group were significantly decreased compared with those in the Control group (P<0.05; Fig. 1A). The PWT values of the Control and OXA groups were 17.0±0.6 and 7.7±0.3, respectively. Res was intrathecally injected into the lumbar spinal cord of OXA rats. The PWT values of OXA-induced rats were significantly increased at 2-6 h after Res treatment (Fig. 1B). The PWT values of the OXA and OXA + Res groups at post-treatment time points of 2, 4 and 6 h were 7.5±1.8 vs. 13.8±1.8 (P<0.05 vs. OXA group), 7.4±1.4 vs. 14.2±1.7 (P<0.05 vs. OXA group) and 7.2±1.5 vs. 13.9±1.6 (P<0.05 vs. OXA group), respectively. However, Res treatment had no significant effect on the mechanical pain behavior of control rats (Fig. 1C).

Histological characterization was performed to analyze spinal inflammation. Subsequent to OXA administration, severe infiltration of inflammatory cells (red arrow) was observed in the spinal dorsal horn, while Res treatment decreased the inflammatory response induced by OXA administration (Fig. 2A). The relative inflammation scores in the OXA and OXA + Res groups were 2.9±0.03 (P<0.05 vs. Control; Fig. 2C) and 2.5±0.04 (P<0.05 vs. OXA group; Fig. 2C), respectively. Pro-inflammatory cytokines, such as TNF-α and IL-1β, promote the inflammatory reaction and are associated with the process of pathological pain (26). OXA enhanced the fluorescence intensities of spinal IL-1β, while this increase was reduced by Res treatment (Fig. 2B). The relative intensities of the OXA and OXA + Res groups were 1.28±0.06 (P<0.05 vs. Control) and 1.06±0.09 (P<0.05 vs. OXA group; Fig. 2D), respectively. Western blot analysis indicated upregulation of spinal TNF-α and IL-1β protein expression in the OXA group (Fig. 2E). The relative gray values of TNF-α and IL-1β in the OXA group were 2.24±0.09 and 1.84±0.02, respectively (P<0.05 vs. Control; Fig. 2F). Following Res treatment, the levels of TNF-α and IL-1β were decreased to 1.59±0.02 and 1.32±0.04, respectively (P<0.05 vs. OXA group; Fig. 2F).

Localization, distribution and expression of the astrocytic marker GFAP were measured using an immunofluorescence assay and western blot analysis. The fluorescence intensity of spinal GFAP was increased in the OXA group, while it was decreased in the OXA + Res group compared with the Control group (Fig. 3A). The relative fluorescence intensities of the OXA and OXA + Res groups were 1.23±0.05 (P<0.05 vs. Control; Fig 3B) and 0.86±0.09 (P<0.05 vs. OXA group; Fig. 3B), respectively. Western blot analysis indicated that relative GFAP expression was increased to 1.70±0.01 in the OXA group (P<0.05 vs. Control; Fig. 3C and D), whereas Res treatment reduced the relative spinal GFAP expression in the OXA + Res group with the relative grey value decreased to 0.63±0.01 (P<0.05 vs. OXA group; Fig. 3C and D).

COX-2 is involved in inflammatory responses and COX-2 inhibitors are used as nonsteroidal anti-inflammatory drugs (22). To identify if Res is a potential COX-2 inhibitor, a molecular docking assay was performed on the X-ray crystal structures of COX-2 and the ligand Res (Fig. 4A-C). Auto Dock data showed that Res formed six hydrogen bonds with COX-2 at residues H39, R44, D125 and G135. H39 and R44 belong to the epidermal growth factor-like domain of COX-2 and D125 and G135 belong to the C-terminal globular catalytic domain of COX-2. An immunofluorescence assay and western blot analysis were used to detect spinal COX-2 location and expression, respectively. The fluorescence intensity of spinal COX-2 was increased in the OXA group with a relative fluorescence intensity of 1.26±0.08 (P<0.05 vs. Control; Fig. 4D and E). Res exerted an inhibitory effect on COX-2 expression, which showed a relative fluorescence intensity of 0.93±0.09 (P<0.05 vs. OXA; Fig. 4D and E). Western blot analysis indicated that COX-2 expression was upregulated in the OXA group (Fig. 4F) showing a relative gray value of 1.76±0.03 (P<0.05 vs. Control; Fig. 4G). The present data indicated that spinal COX-2 was upregulated after OXA administration. The upregulated COX-2 expression was reduced following Res treatment (Fig. 4F) with a relative gray value decreased to 1.06±0.04 (P<0.05 vs. OXA; Fig. 4G).

The effect of Res on NF-κB was detected using an immunofluorescence assay and western blot analysis. The relative fluorescence intensity of spinal NF-κB in the OXA group was significantly increased to 1.45±0.12, while Res treatment in the OXA + Res group reduced the increase in relative fluorescence intensity to 1.17±0.13 (P<0.05 vs. OXA; Fig. 5A and B). The relative expression levels of the spinal NF-κB protein were significantly increased to 1.55±0.02 in the OXA group (P<0.05 vs. Control; Fig. 5C and D), whereas these were significantly decreased to 1.12±0.01 in the OXA + Res group compared with the OXA group (P<0.05; Fig. 5C and D).

C6 rat glioma cells were treated with TNF-α to induce inflammation. Western blot analysis indicated an increase in COX-2 and IL-1β expression in TNF-α-treated cells that was significantly reduced in the TNF-α + Res group (Fig. 6A). The relative expression levels of COX-2 and IL-1β in the TNF-α-treated group were 1.49±0.32 and 1.51±0.08, respectively (P<0.05 vs. Control; Fig. 6B), while in the TNF-α + Res group these were 1.04±0.05 and 0.99±0.04, respectively (P<0.05 vs. TNF-α; Fig. 6B). The ROS production marked by DCFH-DA was used to analyze the effect of Res on ROS production. Res treatment reduced the TNF-α-induced high ROS levels (Fig. 6C) and the relative ROS fluorescence intensities in the TNF-α and TNF-α + Res groups were 1.47±0.17 (P<0.05 vs. Control; Fig. 6D) and 1.11±0.10 (P<0.05 vs. TNF-α; Fig. 6D), respectively.

The present results demonstrated that intrathecal injection of Res inhibited COX-2 expression, reduced high ROS levels, decreased spinal inflammation and relieved OXA-induced neuropathic pain (Fig. 7).