Animals
Eighty Sprague-Dawley rats (male:female ratio of 1:1, weight 270±20 g) were purchased from Xinjiang Experimental Animal Research Center (Xinjiang Uygur Antonomous Region, Urumqi, China; Certificate No. SCXK2011-0003).
Methods
Preparation of rat models of chronic sciatic nerve injury
Rats were intraperitoneally injected with 10% chloral hydrate (3 mg/kg) for induction of surgical anesthesia. The hair on the lower back and thigh was shaved and the skin was sterilized. The rats were secured tightly, and neuropathic pain was induced by CCI as described previously[7]. Briefly, an approximately 2 cm incision in length was cut parallel to the femur in the rear middle thigh, gluteus maximus, and biceps flexor cruris directly dissociated to expose the sciatic nerve in the spatium intermusculare. The surrounding tissues were gently separated to expose the sciatic nerve, and three ligatures (silk 4-0) were placed around the nerve proximal to the trifurcation, with approximately 1 mm between each ligature. The ligatures were slowly tightened until a brief flick of the ipsilateral hind limb was observed. After nerve ligation, muscular and skin layers were immediately sutured with thread. Penicillin sodium was injected intramuscularly at 3 × 104 U per rat for 3 successive days after surgery to prevent infection. All surgical procedures were performed under normal sterile conditions by the same experimenter.
Standards for the success of sciatic nerve injury
The main parts of sciatic nerve were subjected to a loose ligation with the first line before bifurcation. Three ligations were performed respectively with 1 mm spacing. Taking a mild calf muscles quiver as a symbol, sciatic nerve and arterial mild compression on nerve surface were visible under dissecting microscope, without blood flow interruption.
Grouping and drug administration
Eighty adult Sprague-Dawley CCI rats were randomly divided into five groups using a random number generator: low-dose Ferula. sinkiangensis K.M. Shen (CCI+A1), moderate-dose Ferula. sinkiangensis K.M. Shen (CCI+A2), high-dose Ferula. sinkiangensis K.M. Shen (CCI+A3), physiological saline (CCI+NS), and celecoxib groups (CCI+S).Ferula sinkiangensis K.M. Shen (Xinjiang Institute of Materia Medica, Urumqi, China) was mixed with distilled water and ground evenly in a mortar. Rats were intragastically administered Ferula sinkiangensis K.M. Shen at 0.075, 0.15 and 0.3 g/kg in the CCI+A1, CCI+A2 and CCI+A3 groups respectively, while physiological saline was intragastically administered (CCI+NS group) at 3 mL/100 g. The drug doses used were extrapolated from human studies. The moderate drug dose (1.5/60×6.5 g/kg) was calculated by conversion from a human study using the body surface area (BSA)[8]. The low dose was one half of the moderate dose, and the high dose was double of the moderate dose.
Behavioral pain scores
Behavioral pain assessment was performed at 1 day before operation and at 1, 2, 3, 5, 7, 14 days after operation. Spontaneous pain was assessed based on the position of the paw on the surgical side according to Attal et al [9]. Score 0, normal paw position; score 1, the paw is positioned slightly on the cage bottom and appears slightly flexed; score 2, only the internal edge of the paw touches the cage bottom; score 3, only the foot heel touches the cage bottom, with paw inversion observed; score 4, paw suspension; score 5, rat licks the paw on the surgical side.
Thermal withdrawal latency measurement (TWL)
The TWL of the bottom foot of rat was measured with a PL-200 thermal pain stimulator (Chengdu Taimeng Technology Co., Ltd., Sichuan Province, China)[10]. At 1 day before CCI and at 1, 2, 3, 5, 7 and 14 days after CCI, rats were placed gently in a Plexiglass box on the upper glass plate of the thermal pain stimulator for 10 minutes. Until the rats were habituated to the environment, the sole of the posterior limb was exposed to thermal light at an intensity of 60%, and rat paw TWL was recorded. The duration of exposure to light was limited to 25 seconds and the interval between two measurements was 15 minutes. The measurement was repeated for three times to yield the mean TWL response.
Rat mechanical allodynia latency measurement
The 50% paw mechanical allodynia latency in rats (WMT) was assessed using von Frey filaments (Stoelting; Wood Dale, IL, USA) and the up-and-down method at 1 day before treatment and at 1, 2, 3, 5, 7 and 14 days after treatment[11]. Rats were placed in a Plexiglass box on a metal grid for 15 minutes. Then, the bottom of medial foot of the posterior limb was stimulated successively with 2, 4, 6, 8, 10 and 15 g von Frey filaments for less than 6 seconds. A positive reaction was defined as lifting or licking the stimulated foot. If stimulating with a lower force did not induce positive response, a higher force was used. If a positive response was present, stimulation with a lower force was applied to a neighboring point on the bottom of foot. The stimulation was applied 10 times with 30 seconds interval between each of stimulation. Mechanical hyperalgesia was defined as the presence of ≥ 5 positive responses in 10 times of stimulation, and the percentage of positive response occurring in 10 times of stimulation was defined as the positive rate of maximum folding force. If < 5 positive responses occurred in 10 times of stimulation, the positive rate of the minimal folding force that induced the positive response was estimated, and such a folding force indicated the minimal folding force. The maximum intensity was 15 g, and intensity more than 15 g was also recorded as 15 g. The folding force that induces a positive response was calculated using the following formula[12]: folding force to induce positive response = the maximum folding force-(maximum folding force-minimal folding force)/(positive rate of maximum folding force-positive rate of minimal folding force) × (positive rate of maximum folding force-50%), with 50% of this folding force used as the WMT.
Determination of Fos protein expression in spinal cord
At 3, 7 and 14 days after treatment, three rats were randomly selected from each group and were sacrificed. The chest was rapidly opened and a catheter was inserted into the ascending aorta. The remaining blood was flushed with 100 mL NaCl (9 g/L), followed by perfusion of 500 mL phosphate buffered saline (PBS; 0.1 mol/L) containing 40 g/L paraformaldehyde for 1 to 2 hours. The fixed spinal cord was excised and embedded in paraffin for slice preparation. Spinal cord samples were cut into cross sections, and Fos protein expression was determined by immunohistochemistry method. Briefly, slices were dewaxed, hydrated, washed three times with 0.01 mol/L PBS (pH < 7.4) (5 minutes/wash) and incubated in H2O2 for 15 minutes to quench endogenous peroxidase activity. Specimens were washed three times with PBS (5 minutes/wash) and blocked in normal goat serum for 5-10 minutes. The excess serum was removed and slices incubated in rabbit anti-Fos antibody (1:8 000) at 4 ℃ for 48 hours, followed by three PBS washes (5 minutes/wash). Immunolabeled slices were then incubated in biotin-labeled goat anti-rabbit intravenous gamma-globulin at room temperature for 30 minutes and wash three times with PBS (5 minutes/wash). Immunolabeling was visualized by 3,3'-Diaminobenzidine (DAB) staining (10 minutes). Labeled slices were flushed with running water, counterstained with hematoxylin and eosin (H&E) for approximately 30 seconds, dried, dehydrated, made transparent, and sealed under coverslips. Cells stained positive for Fos were counted under an Olympus BH2 microscope (Olympus, Tokyo, Japan) at 400× magnification. Six sections were sampled from the spinal cord of each rat and the average number of neurons with Fos-like immunoreactivity (FLI) was reported.
Determination of astrocyte activation in the spinal cord
The spinal cord samples were cut into sections and stained for glial fibrillary acidic protein (GFAP). Briefly, paraffin-embedded spinal cord sections were dewaxed, made transparent in dimethylbenzene, hydrated in a graded alcohol series, and treated with citric acid for antigen retrieval. The sections were washed three times with PBS (3 minutes/wash), and blocked in (animal) serum in a humidity box at 37 ℃ for 30 hours. The blocking solution was removed, and slices incubated in mouse anti-GFAP (1:400; Sigma) at 4 ℃ overnight. Immunolabeled sections were washed three times with PBS (3 minutes/wash), and incubated in fluorophore-conjugated secondary antibody (1:100; Sigma) in a humidity box at room temperature for 1 hour. The sections were washed three times with PBS (10 minutes/wash), preserved with fluorescent mounting medium, and visualized under a confocal microscope. The spinal cord gray matter was demarcated into 4 Rexed layers: the superficial layer (Rexed layers I-II), laminae propria (Rexed layers III-IV), neck of the dorsal horn (Rexed layers V-VI), and ventral horn (Rexed layers VII-X). Six sections were sampled from the spinal cord of each rat, and the number of immunopositive astrocytes on each section was counted. The mean number of GFAP-positive astrocytes was calculated as an estimate of astroglial reactivity (activation). The staining intensity in individual cells was not taken into account for this estimation.
Main outcome measures
Behavior pain scores; TWL measurement results; WMT measurement results; Fos protein expression in rat spinal cord; astroglial GFAP expression in rat spinal cord.
All measurement data are expressed as the mean ± SD, and all statistical analyses were performed using the Statistic Package for Social Science 13.0 software (SPSS, Chicago, IL, USA). Repeated-measures data were tested for statistical significance by analysis of variance, and treatment groups were compared by two independent samples t-tests. Categorical values were expressed as proportions and compared by Chi-square test. A P-value < 0.05 was considered statistically significant.