The experiment was approved by the Institutional Animal Care and Use Committee at Chung-Ang University (No. 2015-00063). All experiments were performed in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals. Adult male Sprague–Dawley rats (250–300 g; Coretec, Seoul, Korea) were single-housed in cages in a temperature-controlled room (22℃), and fed a standard laboratory diet and tap water. They were kept under a 12-hour light/dark cycle (lights on from 8:00 a.m. to 8:00 p.m.) and acclimated to the housing facilities for one week prior to the experimental procedures. Rats showing any abnormalities were excluded.

All surgical procedures were performed under sterile conditions by one investigator who was unaware of the group allocation. General anaesthesia was induced with 6% isoflurane in 100% oxygen inside a sealed clear plastic chamber until the rats became immobile. They were then maintained on a non-rebreathing anaesthetic circuit mask using 1% to 2% isoflurane in 100% oxygen. Cefazolin (20 mg/kg; Chong Kun Dang Pharmaceutical Co., Seoul, Korea) was administered subcutaneously prior to incision. The plantar surface of the left hind paw of each rat was prepared aseptically for surgery. The incisional pain model was created as previously described [10]. In brief, a 1 cm longitudinal skin incision extending towards the digits was made with a blade at a point approximately 0.5 cm distal to the tibiotarsal joint on the plantar surface of the left hind paw. The plantaris muscle was isolated, elevated slightly, and then incised longitudinally. The incision was closed with two interrupted horizontal mattress sutures of 5-0 nylon.

All the medications used in the present experiment were purchased from Sigma-Aldrich (St. Louis, MO). PPF was dissolved in 2 mL sterile endotoxin-free isotonic saline. The control groups were prepared with syringes containing 2 mL normal saline. The syringes were covered with opaque tape and numbered sequentially according to the randomization list of a respective experiment for allocation concealment. Computer-generated randomization was performed using PASS^TM 11 software (NCSS, Kaysville, UT). Drugs in prepared syringes were intraperitoneally administered according to the study protocol. All experimental procedures were conducted with operatives blinded to the group allocation.

Evaluation of the anti-nociceptive effect of intraperitoneally administered PPF against mechanical allodynia was performed after plantar incision. Thirty-two rats were randomly assigned to one of four groups of eight: the control group and POST-PPF 3 mg/kg, 10 mg/kg and 30 mg/kg groups according to the administered dose of PPF. Either normal saline (the control group) or various doses of PPF (the experimental groups) were injected intraperitoneally 2 hours after plantar incision. The dose level of PPF was based on the amount used in a previous experimental study [7].

Evaluation of the anti-nociceptive effect of intraperitoneally administered PPF before plantar incision was performed. Thirty-two rats were randomly assigned to one of four groups of eight rats: the control group; and PRE-PPF 3 mg/kg, 10 mg/kg, 30 mg/kg groups by administered doses of PPF. Normal saline or various doses of PPF were injected intraperitoneally 30 minutes before plantar incision.

The observed effects of PPF against mechanical allodynia induced by plantar incision were examined in order to determine whether they are mediated by the following receptors: alpha (1 and 2) adrenergic, cholinergic (nicotinic and muscarinic), and opioid. Forty-two rats were randomly assigned to one of seven groups of 6: a PPF-only group as the control, and PPF with one of the study drugs: yohimbine (2 mg/kg), dexmedetomidine (50 μg/kg), prazosin (1 mg/kg), atropine (5 mg/kg), mecamylamine (1 mg/kg), or naloxone (5 mg/kg)) for the other groups. The study drug or normal saline was intraperitoneally injected 2 hours after the plantar incision, and 10 minutes after that, 30 mg/kg of PPF was injected intraperitoneally.

We tested the behavioural responses to mechanical stimuli in order to evaluate the antiallodynic effect and possible mediated mechanisms of PPF, respectively. Individual rats were placed on an elevated plastic mesh floor (8 × 8 mm perforations) under an overturned clear plastic cage (21 × 27 × 15 cm) and allowed to acclimatise for 15 minutes. The rats were then tested to determine their mechanical withdrawal thresholds (MWTs) to stimuli using von Frey filaments (Stoelting Co., Wood Dale, IL). Filaments with bending forces of 4, 9, 20, 59, 78, 98, 147, and 254 mN were applied vertically to the plantar aspect of the hind paw by administering sufficient pressure to gently bend them until either the hind paw was withdrawn or a bending force of 254 mN (the cut-off value) was reached. Each filament was applied three times at intervals of 3 minutes. The lowest bending force that caused paw withdrawal after application of the filament was used to determine the MWT of the hind paw. After a response was observed, filaments with higher and lower bending forces were applied to confirm the MWT level.

For evaluation of the POST-PPF groups, the MWT was assessed according to the following schedule: 1 day before incision (BL); 2 hours after plantar incision (i.e., immediately before PPF administration) (AI); and 15, 30, 45, 60, 80, 100, and 120 minutes; 24 and 48 hours; and 7 days after the injection of 0.9% saline or PPF. For evaluation of the PRE-PPF groups, the MWT was assessed according to the following schedule: 1 day before incision; 2 hours after plantar incision; and 15, 30, 45, 60, 80, 100, and 120 minutes; 24 and 48 hours; and 7 days after the first measurement of MWT.

In order to identify the effect of PPF on motor function (the sedative effect of PPF), we used an accelerating Rotarod treadmill (Jeung Do Bio & Plant Co., Ltd., Seoul, Korea). Twelve rats were randomly assigned to one of two groups of six rats: the PPF and control groups. The rats were injected intraperitoneally with 30 mg/kg PPF or normal saline 2 hours after the plantar incision. The Rotarod test was performed before injection of PPF or normal saline, and at 2 and 24 hours after the injection. Specifically, the rats were placed on the treadmill running at a speed with a gradual increase from 1 to 18 rotations per minute (rpm) for 120 seconds and maintained for another 30 seconds at 18 rpm [12]. The time at which the rat fell off the treadmill was noted.

The rats were injected intraperitoneally with 30 mg/kg of PPF or normal saline 2 hours after plantar incision. At 1 and 48 hours after PPF or normal saline injection, blood samples were collected from the lateral tail vein of the rats into a chilled, sterile tube containing EDTA (EDTA vacutainer, Becton Dickinson, Franklin Lakes, NJ) and centrifuged at 2,000 g for 15 minutes. The plasma was harvested and stored at -80℃ until it was assayed for tumor necrosis factor (TNF)-α, interleukin (IL)-1β, and IL-6; their plasma concentrations were assessed with commercially available enzyme-linked immunosorbent assay kits (R&D Systems, Minneapolis, MN). Individual experimental procedures were performed according to the manufacturer’s instructions.

Ketorolac was used as the reference analgesic in order to compare the PPF groups with a positive control group [13]. Ketorolac (30 mg/kg) was intraperitoneally injected 2 hours after plantar incision in six rats before behavioural measurements. The area under curve (AUC) of the MWT over time, between the time points of AI and 120 minutes for each rat, was calculated for this experiment.

The primary outcome measure of this study was the MWT to stimulation using von Frey filaments. In order to estimate the group size for a study assessing the pre-emptive or treatment antiallodynic activity of PPF, a pilot study was conducted for measuring MWTs to von Frey filament stimulation in eight allodynia-induced rats. The average MWT at the baseline (BL); immediately after incision (AI); and at 15, 30, 45, 60, 80, 100, and 120 minutes; 24 and 48 hours; and 1-week post-operation were 78.4, 17.5, 17.5, 12.9, 12.9, 13.7, 12.9, 12.9, 12.5, 9.7, 9.7, and 38.0 mN, respectively, with standard deviations ranging from 0.9 to 11.5 mN, and an autocorrelation between adjacent measurements on the same individual of 0.6 mN. For the power calculation, we assumed that first-order autocorrelation adequately represented the autocorrelation pattern. In order to compare between-group differences, we used the Geisser-Greenhouse Corrected F-test for a repeated-measures analysis of variance (rANOVA). We wanted to detect 10%, 20%, and 30% increases in the MWT in the PPF 3 mg/kg, PPF 10 mg/kg, and PPF 30 mg/kg groups, compared with allodynia-induced rats. The standard deviation was 3.60 mN, and the actual effect standard deviation was 2.31 mN, thus the effect size was 0.64 mN. Therefore, eight rats per group were needed for α = 0.05 and a power of 80%. The PASS 11^TM software (NCSS) was used to calculate the sample size.

The Shapiro–Wilk test was used to test for the normality of the variables. IL-1β and IL-6, and the AUC of the MWT over time between AI and 120 minutes passed the normality test, but TNF-α and the MWT did not. We additionally checked q-q plots for TNF-α and the MWT, which did not show marked deviation from linearity. Therefore, we assumed that the normal distribution requirement for parametric testing had not been violated, and so decided that rANOVA was applicable. Because IL-1β, IL-6, TNF-α, and the Rotarod test passed sphericity testing, they were compared using rANOVA, followed by t-tests with Bonferroni correction (α = 0.05/2 = 0.025 or α = 0.05/3 = 0.017). Because applying Mauchly’s sphericity test indicated that the assumption of sphericity had been violated in the pre-emptive study (χ² (65) = 302.23, P < 0.001), the treatment study (χ² [65] = 468.16, P < 0.001), the mechanism study (χ² [65] = 303.09, P < 0.001), and the positive control study (χ² [65] = 302.13, P < 0.001), we used a one-way Wilk’s lambda multivariate analysis of variance (MANOVA), a generalized form of univariate analysis of variance (ANOVA), comparing two or more dependent variables. Moreover, univariate ANOVA with Bonferroni correction (α = 0.05/12 = 0.0042) was used to compare the MWT at each time point.

When the homoscedasticity requirement, using Levene’s test for homogeneity of variances, was not met using ANOVA, we used Welch’s corrected ANOVA. Tukey’s or Tamhane’s T2 post-hoc test was used when ANOVA or Welch’s corrected ANOVA was significant in identifying the groups with statistically significant mean differences.

Individual measurements were expressed as the mean ± standard deviation and analysed with SPSS 23.0 (IBM Co., Armonk, NY). A P value of 0.05 or less was considered statistically significant.

A total of 124 rats completed the present study without follow up loss. Throughout the experimental period, the rats remained well-groomed and appeared to ingest normal amounts of food and water. Except for impaired weight-bearing on the area of the incision, their gaits appeared unaffected. None of the rats developed complications in the surgical wound.

Fig. 1 shows the effect of PPF on mechanical allodynia administered after plantar incision. The results of MANOVA showed statistically significant difference among the groups (F[36.0, 56.96] = 2.532, P = 0.001: Wilk’s lambda = 0.048, partial η² = 0.636). The MWT values at 45, 60, 80, 100, and 120 minutes for the PPF 30 mg/kg group, and at 60 minutes for the PPF 10 mg/kg group, were significantly increased compared to the control group. The MWT values at 45, 60, and 80 minutes in the PPF 30 mg/kg group were significantly increased compared to the PPF 10 mg/kg group. On the other hand, PPF administered before plantar incision showed no significant change in the MWT (data not shown). The MANOVA results show no statistically significant differences among the groups (F[33.0, 36.06] = 0.769, P = 0.776: Wilk’s lambda = 0.208, partial η² = 0.407).

Fig. 2 show the possible mechanism mediated in PPF-induced analgesia. The MANOVA results show statistically significant differences among the groups (F[33.0, 36.06] = 0.769, P = 0.776: Wilk’s lambda = 0.208, partial η² = 0.407). Compared with group PPF 30 mg/kg as the control, the MWT values at 15, 30, 45, and 60 minutes for the PPF 30 mg/kg group with dexmedetomidine were significantly increased, those at 30, 45, 60, 80, 100, and 120 minutes for the PPF 30 mg/kg group with yohimbine were significantly decreased (Fig. 2A). Similarly, compared with group PPF 30 mg/kg as the control, the MWT values at 30, 45, 60, and 80 minutes for group PPF 30 mg/kg with atropine and those at 45 and 80 minutes for group PPF 30 mg/kg with mecamylamine were significantly decreased (Fig. 2B).

Intraperitoneal injection of 30 mg/kg PPF did not have a significant effect on the motor performance of the treatment group rats compared to those injected with the control vehicle. Fig. 3 shows that there was no difference among the groups with respect to latency before falling off the Rotarod (F[1, 0.222] = 0.002, P = 0.964, partial η² = 0.001).

There was a statistically significant difference between the PPF 30 mg/kg and control groups in the serum level of TNF-α (F[1, 13973.15] = 56.87, P < 0.001, partial η² = 0.850), IL-1β (F[1, 11443.76] = 11.99, P = 0.006, partial η² = 0.545) and IL-6 (F[1, 2610.40], P = 0.042, partial η² = 0.353). The serum level of TNF-α was significantly reduced 1 hour after injection of PPF 30 mg/kg compared to that in the control group (Fig. 4A). The serum level of IL-1β was significantly reduced 48 hours after injection of PPF 30 mg/kg compared to that in the control group (Fig. 4B). However, there was no difference in the serum levels of IL-6 between the two groups at any particular time point (Fig. 4C).

Fig. 5 shows a comparison of the control, the PPF 30 mg/kg, and the ketorolac 30 mg/kg groups. The results of MANOVA showed statistically significant difference among the groups (F[24.0,20.0] = 5.985, P < 0.001, partial η² = 0.878). The MWT values at 30, 45, 60, and 80 minutes for the PPF 30 mg/kg group and at 15, 30, 45, and 60 minutes for the ketorolac 30 mg/kg group were significantly increased compared to the control group. The MWT values for the ketorolac 30 mg/kg group were significantly increased at 15 minutes and significantly decreased at 80 minutes compared to the PPF 30 mg/kg group.

Fig. 6 shows the AUC for the comparison of the three groups. There was a significant difference in the AUC of the MWT over time between AI and 120 minutes (F[2.21] = 17.594, P < 0.001). The AUCs of the MWT over time between AI and 120 minutes were significantly lower in the control group than those in groups PPF 30 mg/kg and ketorolac 30 mg/kg (P < 0.001 and P < 0.001, respectively), but there was no evidence of difference between groups PPF 30 mg/kg and ketorolac 30 mg/kg (P = 0.999).