All protocols and procedures adhered to the Institutional Animal Care Committee of Tianjin Medical University in line with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. The protocols and procedures were approved by the Committee on the Ethics of Experiments of Tianjin Medical University (Tianjin, China; permit no. 2014-X62-27) and the Institutional Animal Care Committee of Tianjin Medical University (Tianjin, China). All procedures were performed under sevoflurane anesthesia and considerable efforts were made to minimize animal suffering.

A total of 72 male Sprague Dawley (SD) rats (age, 8–9 weeks; weight, 250–280 g) and 24 male SD rat pups (age, 14–21 days; weight, 50–70 g), were provided by the Laboratory Animal Center of the Military Medical Sciences Academy of the Chinese PLA (Beijing, China). Rats were housed in cages with a 12-h light/dark cycle at a temperature of 22±2°C. Rats had access to water and food ad libitum.

An intrathecal catheter implantation was performed as described by Yaksh and Rudy (25). After being anesthetized by sevoflurane inhalation (3.0% for induction, 1.0% for maintenance; Abbott Japan Co., Ltd., Tokyo, Japan), the rats were implanted with 8.5 cm polyethylene (PE-10) catheters via the atlantooccipital membrane down to the lumbar enlargement at L1-L2 of the spinal bony structure. The level of the L1-L2 spinal bony structure corresponds to the spinal cord segments of L5-S3, which are responsible for hind limb pain. The presence of the catheter in the subarachnoid space was confirmed at 24 h by paralysis of the hind limb. Rats exhibiting neurological deficits, including paralysis or motor function impairment were excluded from the study and euthanized immediately.

The rat model of incisional pain was performed according to a previously described procedure (26). In brief, rats (weight, 250–280 g) were anesthetized with sevoflurane (3.0% for induction, 1.0% for maintenance) via a nose mask. A 1-cm longitudinal incision was made through the skin, fascia and muscle of the right hind paw, starting 0.5 cm from the proximal edge. The skin was sutured with 4-0 silk sutures after the underlying flexor muscle was divided. To avoid infection, an erythromycin ointment was applied to the incision.

The rats were anesthetized with sevoflurane (3.0% for induction, 1.0% for maintenance) via a nose mask. Next, a 24-gauge over-the-needle Teflon catheter was inserted into the caudal vein and flushed with heparinized saline. The caudal vein catheter was used to infuse the experimental drugs. The control group (C) received saline infusion (0.1 ml/kg/min; 60 min), the remifentanil group (R) received remifentanil infusion (1.0 ml/kg/min; 60 min), the incision group (I) received a surgical incision, the remifentanil plus incision group (RI) underwent surgical incision and remifentanil infusion (1.0 ml/kg/min; 60 min), the remifentanil, incision and DOR agonist deltorphin II group (RD) received a surgical incision, remifentanil infusion (1.0 µg/kg/min i.v.; 60 min) and deltorphin II (10 µl of a 4 nM solution by intrathecal injection), and the remifentanil, incision and naltrindole group (RN) received a surgical incision, remifentanil infusion (1.0 µg/kg/min; 60 min) and naltrindole (10 µl of a 30 nM solution by intrathecal injection); n=12 per group.

In the patch clamp recording experiment, spinal slices from rat pups of the C group were incubated in artificial cerebrospinal fluid (ACSF) for 60 min. In the R group, spinal slices were incubated in ACSF with 4 nM remifentanil for 60 min. In the RD group, spinal slices were incubated in ACSF with 4 nM remifentanil and 4 nM deltrophin for 60 min. Finally, in the RN group, spinal slices were incubated in ACSF with 4 nM remifentanil and 30 nM naltrindole for 60 min.

The paw withdrawal threshold (PWT) was measured using von Frey filaments (cat. no. BSEVF3; Harward Apparatus Co., Holliston, MA, USA) where the rats were placed in individual cages with a wire grid bottom. Each trial was repeated 5 times at 5-min intervals with a cut-off values of 60 g. The paw withdrawal latency (PWL) was measured using a hot plate (cat. no. YLS-6B; Zheng Hua Biological Instrument Co., Suixi, China). Each trial was repeated 5 times at 5-min intervals and a cut-off time of 20 sec to avoid tissue damage.

After drug treatment, rats were sacrificed by decapitation under sevoflurane anesthesia, and the lumbar enlargement of spinal cord (L3-L5) was immediately removed and stored at −80°C. To prepare a total lysate, the tissue was homogenized in ice-cold lysis buffer (2% Triton X-100, 150 mM NaCl, 50 mM Tris, 100 g/ml phenylmethylsulfonyl fluoride, 1 g/ml aprotinin and phosphatase inhibitors; pH 7.5). The lysate was centrifuged at 12,000 × g for 30 min at 4°C and the supernatant was used for a western blot analysis. To prepare membrane and cytosolic fractions, the tissue was separated into membrane, cytosolic and nuclear fractions by using a membrane, cytosolic and nuclear compartment protein extraction kit (cat. no. K3014005; Biochain Institute, Inc., Hayward, CA, USA). The protein concentration was measured using a Bio-Rad protein assay (cat. no. 5000006; Bio-Rad Laboratories, Inc., Hercules, CA, USA) and the samples were then heated to 99°C for 5 min. A total of 15 µg protein was loaded onto a 4–20% Criterion^™ Tris-HCl Protein Gel (cat. no. 3450032, Bio-Rad Laboratories, Inc.). The proteins then underwent electrophoresis and were transferred onto Immun-Blot^® polyvinylidene difluoride membranes (cat. no. 1620177; Bio-Rad Laboratories, Inc.). All membranes were blocked in a buffer of 5% non-fat milk, 50 mM Tris-HCl, 154 mM NaCl and 0.05% Tween-20 (pH 7.4) for 1 h and subsequently incubated overnight at 4°C with antibodies directed against GluR1, phosphorylated-GluR1 (Ser845), PKA (all 1:1,000 dilution; cat. nos. sc-7609, sc-24,593 and sc-28315, respectively. (all, Santa Cruz Biotechnology, Inc., Dallas, TX, USA), DOR (1:1,000 dilution; cat. no. ab63536; Abcam, Cambridge MA, USA) and β-actin (1:5,000 dilution; cat. no. A5316; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany). The membranes were subsequently incubated with horseradish peroxidase-conjugated anti-mouse or anti-rabbit secondary antibodies (1:3,000 dilution; cat. nos. 115-035-003 and 111-035-003, respectively; Jackson Immuno Research Laboratories, Inc., West Grove, PA, USA) at room temperature for 2 h. They were then visualized using Clarity^™ Western ECL Substrate (cat. no. 170-5060; Bio-Rad Laboratories, Inc.) and exposured on the ChemiDoc XRS System with Image Lab software (Bio-Rad Laboratories, Inc.). The intensity of the blots was quantified by densitometry using Image Lab software. β-actin was used as the loading control. The western blot analysis was repeated 3 times.

The spinal cord slices were prepared as described previously (27,28). Lumbar enlargements of 14–21 day-old rats were separated under sevoflurane anesthesia (3.0% for induction, 1.0% for maintenance). Spinal cords (L₁-L₅) were sliced into transverse slices (350 µm) with a vibratome (VT1000S; Leica Microsystems, Wetzlar, Germany). The slices were incubated in ACSF under specific conditions (22–25°C, 5% CO₂ and 95% O₂, pH 7.4) for 1 h prior to incubation. ASCF was composed of (in mM): 126 NaCl, 2 MgCl₂, 1.25 NaH₂PO₄, 3.5 KCl, 26 NaHCO₃, 10 D-glucose and 2 CaCl₂. Whole-cell patch-clamp recordings were performed in the medial prefrontal cortex through an upright microscope (BX51W1; Olympus, Tokyo, Japan) equipped with a television monitor, which was connected to a low light sensitive charge-coupled device camera (710M; Leica Microsystems, Inc., Buffalo Grove, IL, USA). The recording pipette was produced by a vertical electrode puller (PIP5; HEKA, Lambrecht, Germany) with tip openings of 1–2 µm and a series resistance of 4–9 MΩ. The pipettes were filled with an intracellular solution containing (in mM): 130 KCl, 10 4-(2-hydroxyethyl) piperazine-1-ethanesulfonic acid, 0.5 CaCl₂, 10 ethylene-bis (oxyethylenenitrilo) tetraacetic acid, 2 MgCl₂, 2 Mg-ATP and 0.3 Na-GTP (pH 7.3). To isolate the AMPAR-mediated miniature excitatory postsynaptic current (mEPSC), D-2-amino-5-phosphonovalerate (40 µM), tetrodotoxin (0.5 µM) strychnine (2 µM) and bicuculline (5 µM) were added to the perfusion solution. The electric signals were amplified and filtered using pCLAMP 9.2 software (Molecular Devices, Sunnyvale, CA, USA). The amplitude and frequency of AMPAR-mediated mEPSCs was analyzed using Clampfit 9.0 (Axon Instruments, Union City, CA, USA).

Statistical analysis was performed using SPSS 19.0 (IBM Corp., Armonk, NY, USA). Values are expressed as the mean ± standard deviation. Two-way analysis of variance (ANOVA) was used for the results of the behavioral tests and one-way ANOVA was used for the western blot analysis and whole-cell patch-clamp recordings, followed by post-hoc comparisons using the Tukey-Kramer test. P<0.05 was considered to indicate a statistically significant difference.

As in a previous study by our group (29), remifentanil treatment increased the sensitivity to heat pain and mechanical pain at the rate of 1 µg/kg/min from 2 to 48 h after discontinuation of remifentanil infusion. Remifentanil alone produced higher heat pain and mechanical pain sensitivity from 2 h to 5 days after remifentanil infusion. As presented in Fig. 1, thermal and mechanical hyperalgesia appeared at 2 h, reached a peak level on day 2 and persisted for 5 days after remifentanil infusion and/or incision (P<0.05). The results indicated that the effect of incision was about as high as that of remifentanil treatment regarding the induction of hyperalgesia, while there appears to be an enhancing effect in the RI group, which means remifentanil infusion and incision had an additive effect.

In a previous study by our group, the membrane fraction of the GluR1 subunit reached a peak level on day 2 after remifentanil-induced postoperative hyperalgesia (30). Therefore, spinal cords (L₁-L₅) were harvested at 2 days after remifentanil infusion. As presented in Fig. 2, the expression of membrane GluR1 was increased in the remifentanil group and the incision group on day 2 (P=0.003 and 0.004, respectively). The membrane GluR1 in the RI group demonstrated a notable increase compared with the R and I groups. However, no change in the expression of total GluR1 was detected in all groups (P>0.05 for each). The ratio of membrane and total GluR1 was increased in the remifentanil and the incision group on day 2 (P=0.009 and 0.002, respectively). The ratio of membrane and total GluR1 in RI group demonstrated a notable increase compared with the R and I groups. These results suggested that remifentanil infusion or/and incision promoted GluR1 membrane trafficking in the spinal cord. Remifentanil infusion and incision had an additive effect on promoting GluR1 membrane trafficking.

Spinal cords (L₁-L₅) were separated at 2 days after remifentanil infusion. Compared with that in the control group, the membrane fraction of DOR exhibited a significant increase after remifentanil infusion and/or incision, whereas total protein levels of DOR were not affected (P<0.01; Fig. 3). The ratio of membrane to total DOR increased in the remifentanil and the incision group on day 2 (P=0.007 and 0.003, respectively). The membrane DOR expression in the RI group demonstrated a notable increased compared with the R and I groups. The ratio of membrane to total DOR also significantly increased compared with the R and I groups. Together, these results suggested that remifentanil infusion and/or incision promoted DOR membrane trafficking in the spinal cord. Remifentanil infusion and incision had an additive effect on promoting DOR membrane trafficking.

The selective DOR agonist deltorphin or the antagonist naltrindole were used to investigate whether DOR participates in remifentanil-induced postoperative hyperalgesia. Compared with that in the RI group, the DOR agonist deltorphin had no effect on either mechanical or thermal hyperalgesia. However, it was observed that DOR antagonist naltrindole significantly decreased remifentanil-induced increases in PWT and PWL (P<0.05; Fig. 4). The anti-hyperalgesic effect of naltrindole was demonstrated in the rat model of remifentanil-induced postoperative hyperalgesia. These results suggested that inhibition of DOR had an anti-hyperalgesia effect in the remifentanil and incision-induced hyperalgesia model.

Compared with that in the control group, the expression of membrane GluR1 increased in dorsal horn neurons in the remifentanil and incision group (P=0.0009). The membrane GluR1 expression levels were not significantly different between the RI group and the RD group. Naltindole treatment inhibited remifentanil and incision-induced increases in mGluR1 and the m/t GluR1 ratio (P=0.0013, 0.0022). However, no change in total protein levels of GluR1 was observed in all groups (P=0.463; Fig. 5). The increased ratio of membrane to total DOR in the RI group and RD group was restored by naltrindole treatment (P=0.0065 and 0.0035). These results suggested that DOR inhibition alleviated GluR1 membrane insertion in rats with remifentanil and incision-induced hyperalgesia.

To examine the role of DOR in the function of AMPAR in spinal cord neurons, whole-cell patch-clamp recordings were made in the presence of remifentanil to detect the changes in the AMPAR-mediated mEPSC. Representative traces of AMPAR-mediated mEPSCs in different groups are presented in Fig. 6A. Incubation with remifentanil significantly increased the amplitude of the AMPAR-mediated mEPSC, while decreasing the inter-event interval (P=0.0012 and 0.0004, respectively; Fig. 6), suggesting that the function of AMPAR in the spinal cord was enhanced after treatment with 4 nM remifentanil. No significant difference in the amplitude and frequency of the AMPAR current was observed between the remifentanil and the remifentanil + deltorphin group. However, naltrindole decreased the remifentanil-induced changes in the amplitude and the frequency of the AMPAR-mediated mEPSC when compared with that in the R group (P<0.001; Fig. 6B and C). These results suggested that the increased AMPAR function in the remifentanil-induced hyperalgesia state is consistent with the activation of DOR in the spinal cord.

To further examine the role of DOR in remifentanil-induced enhancement of AMPAR expression, the phosphorylation of GluR1 at Ser845 and the expression of PKA in total protein were measured using western blot analysis. Remifentanil infusion and incision increased the phosphorylation of GluR1 at Ser 845 and the expression of PKA (P=0.0005 and 0.0012, respectively). This increase was inhibited with the application of naltrindole, indicating that the phosphorylation of GluR1 and PKA was dependent on the activity of DOR in rats with remifentanil-induced hyperalgesia (P=0.0023 and 0.0042; Fig. 7).