Male Wistar rats (weighing ~200 g on arrival) were obtained from the animal center of the Chinese Academy of Sciences and housed two per cage with water and food available ad libitum. A 12-h light/12-h dark cycle was used, with lights on at 08:00. All animal experiments were carried out with the approval of the Animal Care Committee at Zhejiang University and consistent with the ethical guidelines for the investigation of experimental pain in animals [27]. All efforts were made to minimize the number of animals used and their suffering. Animals were acclimated to the housing facility for three days before the experiments began.

The rats were implanted with an intrathecal catheter under intraperitoneal anesthesia (pentobarbital, 60 mg/kg). A polyethylene-10 (Smiths Medical, UK) catheter was inserted into the subarachnoid space at the level of the spinal cord lumbar enlargement according to the method described by Størkson et al. [28]. The animals were allowed to recover for 3 days before being randomly divided into groups. Rats showing neurological deficits postoperatively were discarded from the study.

Animals were randomly divided into naïve, sham, and CCI groups (n=6). To create a neuropathic pain state, we performed a surgical procedure according to the description reported by Bennett and Xie [29] and our recent study [30]. Briefly, under pentobarbital (60 mg/kg, ip) anesthesia, the left sciatic nerve was exposed and freed from the underlying connective tissues. Three ligations were placed around the nerve with 4-0 chromic gut sutures, and a typical twitch of the hind paw was seen when the nerve was constricted. The overlying muscle and skin were closed with the same silk sutures. An identical surgical procedure was performed on sham-operated animals without ligation of the sciatic nerve. All animals received antibiotics (0.5 ml penicillin, 96 mg/ml, hypodermic injection) to reduce the possibility of infection. To reduce variability, all animal model procedures were carried out by one investigator who was proficient in modeling.

Animals were divided into naïve, sham, CCI+normal saline (NS), and CCI+DEXmedetomidine (DEX) groups (n=8 for each group). Dexmedetomidine (Sigma, Missouri) was diluted in NS. After neuropathic pain was successfully established, a dose of 2.5 µg DEX (in 20 µl volume) or NS (20 µl) was intrathecally administered once daily for three consecutive days from day 8 to 10 after CCI, followed by an injection of 10 µl NS to flush the catheter.

To minimize the stress induced by handling and measurements, we exposed the rats to the test condition for 3 days before the experiment. The behavioral test was performed 1 h after drug administration.

Mechanical paw withdrawal threshold (MT) testing. Animals were placed in a cage with a wire mesh floor and allowed to explore and groom until they settled. The von Frey filament stimulation has been described previously [31]. A calibrated series of von Frey hairs was applied to the plantar surface of hindpaws in ascending order (0.4, 0.6, 1, 1.4, 2, 4, 6, 8, 10, 15, and 26 g) for up to 6 s per filament. Once a withdrawal response was established, the paw was re-tested, starting with the next descending von Frey hair until no response occurred. The lowest amount of force required to elicit a response was recorded as the paw withdrawal threshold. The withdrawal threshold was measured three times in each animal, and the values from the three measurements were averaged.

Thermal paw withdrawal latency (TL) testing. Paw withdrawal latency to noxious heat stimuli was evaluated using a Hargreaves' test with the appropriate apparatus (Model 336; IITC/life Science, CA) [32,33]. Briefly, rats were placed in a Plexiglas chamber on a glass plate under which a light box was located. A radiant heat stimulus was applied using a beam of light aimed through a hole in the light box to the heel of each hind paw through the glass plate. The light beam was turned off when the rat lifted the foot, and the interval between the start of the light beam and the lifting of the foot was measured and defined as the paw withdrawal latency. Each heat stimulus was applied three times at 5-min intervals. A cutoff time of 20 s was applied to prevent potential tissue damage.

Locomotor function testing. The effects of medication and surgery on locomotor function were examined using the following methods [34]. The rats were divided randomly into five groups (n=6): naïve, sham, CCI, CCI+NS, and CCI+DEX. Placing reflex: The rat was held with the hind limbs slightly lower than the forelimbs, and the dorsal surfaces of the hind paws were brought into contact with the edge of a table. The experimenter recorded whether the hind paws were reflexively placed on the table surface. Grasping reflex: The rat was placed on a wire grid, and the experimenter recorded whether the hind paws grasped the wire on contact. Righting reflex: The rat was placed on its back on a flat surface, and the experimenter noted whether it immediately assumed the normal upright position. Scores for placing, grasping, and righting reflexes were based on counts of each normal reflex shown in five trials.

Western blot analysis was conducted as described in our previous reports [35]. Rats were sacrificed immediately under deep anesthesia, followed by isolation of the lumbar 4~5 spinal cords. The spinal cord was divided into ipsilateral and contralateral halves, and then homogenized in ice-cold homogenization buffer (Beyotime Institute of Biotechnology, China). The homogenates were centrifuged at 10,000 × g for 10 min at 4℃, the supernatant was collected, and the protein concentrations were determined by the Mocro BCA protein assay reagent kit (Thermo Fisher Scientific Inc., MA). Equivalent protein samples were subjected to 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electrophoretically transferred onto polyvinylidene difluoride membranes (Thermo Fisher Scientific Inc.) using a wet transfer apparatus (Bio-Rad Laboratories Inc., CA). Proteins bound to the membrane were stained with Ponceau S solution (Beyotime Institute of Biotechnology) to determine the quality of the transfer. Membranes were blocked with 5% non-fat milk in Tris-buffered saline at room temperature for 2 h and incubated with the primary antibodies against P2X7R (1:1000; Cell Signaling Technology, MA), PERK (1:1000; Millipore, CA), ERK (1:1000; Cell Signaling Technology), and β-tubulin (1:2000; Beyotime Institute of Biotechnology) overnight at 4℃. After the membranes were extensively washed with TBST buffer, a 1:2000 dilution of goat anti-rabbit or goat anti-mouse horseradish peroxidase secondary antibody (Beyotime Institute of Biotechnology) was used as appropriate and incubated for 2 h at room temperature. After extensive washing, signals were detected with enhanced chemiluminescence reagent (Thermo Fisher Scientific Inc.) and captured using a ChemiDoc MP System (Bio-Rad Laboratories Inc.). Immunoreactive density was analyzed using Image Lab software (Bio-Rad Laboratories Inc.). The density of the specific bands was normalized against that of the loading control (β-tubulin). Commercial markers (Thermo Fisher Scientific Inc.) were used as molecular weight standards.

Immunohistochemical staining was conducted in accordance with our previous reports [35]. Seven days after CCI or sham operation, rats were anesthetized and perfused thoroughly with NS, followed by perfusion with 4% paraformaldehyde in 0.1 M phosphate buffer. Lumbar 4~5 segments were then harvested, postfixed, and dehydrated in 30% sucrose. Subsequently, 30-µm free-floating transverse sections were cut in a freezing microtome. Sections were collected and incubated for 48 h at 4℃ in rabbit anti-P2X7R (1:400; ProteinTech, IL) and rabbit anti-PERK (1:200; Millipore). The sections were then incubated with Cy3-conjugated anti-rabbit IgG (1:200; ProteinTech) for 2 h at room temperature. Positive signals were viewed with a fluorescence microscope (x-Cite 120; OLYMPUS, Japan) with the appropriate filters.

Data are expressed as the mean±S.E.M and were analyzed using one-way or two-way ANOVA followed by Fisher's least significant differences (LSD) test or Dunnett's multiple comparisons as appropriate. p<0.05 was considered statistically significant.

In the present study, a model of CCI was used to simulate a chronic neuropathic pain state. Using behavioral measurements, we found that the paw withdrawal thresholds for both mechanical and thermal stimuli were lower in the CCI group than in the sham-operated group on day 7, the time point at which the CCI group usually reached the maximal behavioral hyperalgesia after nerve injury [36]. As shown in Fig. 1, compared with the sham group, the CCI group exhibited a significantly lower MT on day 7 after nerve injury (n=6, p<0.001; Fig. 1A), with similar results for the TL values (n=6, p<0.01; Fig. 1B), suggesting the development of robust mechanical and thermal hyperalgesia in the CCI rats. We further analyzed whether CCI-induced hyperalgesia is associated with spinal P2X7R activation. On day 7 after CCI, western blot analyses revealed that, in parallel with the behavioral hyperalgesia, P2X7R expression in the ipsilateral spinal cord was significantly higher than that in the sham group (n=4, p<0.05; Fig. 1D), while the expression in the contralateral spinal dorsal horn was not significantly different from that in the sham group (n=4, p=0.131; Fig. 1D). Similarly, 7 days after CCI, PERK expression in the ipsilateral spinal dorsal horn was significantly greater than that in the sham group (n=4, p<0.05; Fig. 1E), while the expression in the contralateral spinal dorsal horn was not significantly different from that in the sham group (n=4, p=0.234; Fig. 1E). The results of immunohistochemical assessments confirmed the activation of spinal P2X7R/ERK signaling after CCI injury (Fig. 1C).

To investigate the effect of α₂-adrenoceptor agonists on CCI-associated neuropathic pain, which was reflected by mechanical and thermal hyperalgesia, we administered intrathecal dexmedetomidine (2.5 µg/20 µl) daily from day 8 to day 10 after CCI, when the pain behaviors were well established. Basal MT and TL were not significantly different among the groups before CCI surgery; however, the MT in the CCI+DEX group was significantly higher than that in the NS control (CCI+NS) groups from day 8 to day 10 after CCI (n=8, p<0.001; Fig. 2A), but it was still lower than that in the sham group (n=8, p<0.001; Fig. 2A). In contrast, the TL in the CCI+DEX group was also obviously higher than that in the CCI+NS group from day 8 to day 10 after CCI (n=8, p<0.001; Fig. 2B) and similar to that in the sham group, except on day 10 (p>0.05, from day 8 to 9; p<0.05, day 10, n=8; Fig. 2B).

Importantly, CCI surgery and dexmedetomidine administration did not produce significant effects on locomotor function. As shown in Fig. 3, CCI surgery and intrathecal dexmedetomidine administration at the dosage used in our study did not produce obvious effects on locomotor function. Furthermore, convulsions and hypermobility were not observed in any of the groups.

Next, to further explore the involvement of spinal P2X7R and ERK activation in dexmedetomidine analgesia, we examined the expression of P2X7R and ERK and its phosphorylated form (PERK) by western blot analyses. We chose to perform western blot analysis on day 10 after CCI on the basis of the results of behavior tests. After three days of dexmedetomidine or NS administration, the rats subjected to CCI were sacrificed 1 h after drug administration on day 10 after CCI surgery. In parallel with the changes in mechanical and thermal hyperalgesia, western blot analysis revealed that P2X7R expression in the CCI+NS group was highly upregulated compared with the expression in the sham group (n=5, p<0.01; Fig. 4A). However, in contrast to the CCI+NS group, the CCI+DEX group exhibited significant downregulation of P2X7R expression (n=5, p<0.01; Fig. 4A). Furthermore, western blot analysis showed that although there was no obvious difference in the total ERK expression in the spinal cord on day 10 following CCI among the naïve group, CCI+NS group, or CCI+DEX group (n=4, p>0.05; Fig. 4B), PERK expression in the CCI+NS group was significantly upregulated in comparison with that in the sham group (n=5, p<0.05; Fig. 4C); however, in parallel with the changes in mechanical and thermal hyperalgesia, the expression of PERK in the CCI+DEX group was significantly lower than that in the CCI+NS group (n=5, p<0.05; Fig. 4C).