This study investigated the antinociceptive effects of ( S )-(−)-perillyl alcohol (PA) on orofacial nociception in Swiss male mice using formalin-, capsaicin-, and glutamate-induced pain tests. For each test, eight animals per group were pre-treated intraperitoneally by a blinded investigator with PA (50 or 75 mg/kg), morphine, or vehicle (saline + 0.2% Tween 80). The treatment was performed before the induction of orofacial nociception by injecting formalin, capsaicin, or glutamate solution into the right area of the upper lip. The orofacial nociceptive behaviour was timed in all tests by an investigator who was blinded to the treatments. The statistical analysis was performed using confidence intervals (CI), the effect size, and power. PA blocked the orofacial nociceptive behaviour at both doses tested ( P < 0.05) similarly to morphine ( P > 0.05), in all tests. The effect size was high in the phase 1 formalin test for 50 mg/kg PA (95% CI 0.48–2.31, power 84.6%) and 75 mg/kg PA (95% CI 0.82–2.76, power 96.2%), in phase 2 for 75 mg/kg PA (95% CI 0.44–2.26, power 82.3%), and in the glutamate test for 75 mg/kg PA (95% CI 1.11–3.16, power 99.2%). These findings show strong evidence for the antinociceptive properties of PA in the orofacial region.
The current pharmacological therapy for orofacial pain includes the use of analgesics and corticosteroids for acute conditions, and non-steroidal anti-inflammatory drugs (NSAIDs), local anaesthetics, and muscle relaxants for acute and chronic conditions . Clinicians should prescribe these drugs with caution, as their continued use can lead to drug tolerance and dependence . Attention should also be drawn to contraindications or variations in effects depending on the region where pain is referred . The long-term use of NSAIDs is not indicated in view of the high incidence of exacerbation of hypertension and side effects in the digestive (gastric ulcers) and urinary (osmotic imbalance) systems . Furthermore, local anaesthetics have variable efficacy, as well as inconvenience of administration (injection) , and muscle relaxants have also been reported to have side effects.
In this background, the perspective of using other agents with analgesic activity in the orofacial region gains importance, especially the investigation of natural products. ( S )-(−)-perillyl alcohol (PA) is a monoterpene found in essential oils from many plants, such as Mentha , Cerasus , citrus, and Cymbopogon citratus . In addition to its reported anti-inflammatory and antioxidant properties, a great number of published studies have described the antitumor activity of this monoterpene, including in vitro and in vivo data . PA stands as a promising natural agent for the treatment of orofacial pain, given that a previous study demonstrated analgesic activity in the spinal system . Nevertheless, despite the differences in the afferent pattern of the nervous system, the antinociceptive activity of PA in the trigeminal system remains to be investigated.
Different to the spinal system, the nociception conducted by trigeminal nerves includes a group of organs not located in other parts of the body, i.e. special senses, and represents a particular level of the facial region. The processing of facial nociception occurs at the level of the brain stem in the spinal trigeminal nuclear complex, particularly in the subnucleus caudalis . It involves different properties from those observed in the neurons of the dorsal horn of the spinal cord in relation to the other parts of the body .
The trigeminal and spinal afferent systems have particularities in the processing of pain upon many types of injury, making the trigeminal system unique with regard to features that contribute to the responses to injuries when compared to the spinal afferents. The uniqueness of the trigeminal system is the integration of the nociceptive inputs at more than one level, i.e. at the brain stem, the caudal subnucleus, and the other components of the trigeminal nuclear complex, which altogether also process the properties of epicritic touch and proprioception of the face . Among the animal models developed or adapted for the study of orofacial pain with inflammatory and nociceptive agents are complete Freund’s adjuvant , carrageenan , formalin , capsaicin , and glutamate .
The aim of this study was to investigate the orofacial antinociceptive effect of PA in mice using formalin, capsaicin, and glutamate animal models, considering the null hypothesis that PA has no orofacial antinociceptive effect in animal models (no peripheral or central activity) and that there is no correlation in ordinal scale between different concentrations of PA and control substances (vehicle and morphine).
Materials and methods
This was a randomized, controlled, triple-blind study performed using healthy adult male Swiss albino mice ( Mus musculus ) obtained from the university vivarium. All animals weighed between 25 g and 35 g and were maintained in a 12-h light–dark cycle under controlled temperature, with food and water available ad libitum.
All research procedures were conducted in accordance with the Brazilian Guidelines for the Care and Use of Animals for Scientific and Teaching Purposes of the National Council for the Control of Animal Experimentation and the ARRIVE guidelines (Animal Research: Reporting In Vivo Experiments). This study was approved by the Ethics Committee on Animal Use of the Federal University of Paraíba. A reduced number of animals was considered for the experiments, and all measures aimed at controlling their suffering or stress after experimentation were undertaken in accordance with ethical guidelines for animal research .
The sample size was estimated using the data of a pilot study involving 16 animals. The sample size calculation determined that eight animals per group would provide 98% statistical power to detect differences in orofacial nociceptive behaviour in mice (considering a one-sided type I error of 0.01, effect size g = 2.43) .
Induced orofacial nociception tests
Three different tests were performed to evaluate the orofacial antinociceptive activity of PA in mice: (1) formalin-induced nociception, (2) capsaicin-induced nociception, and (3) glutamate-induced nociception. Each test was performed using four distinct groups of eight animals each ( n = 32). The animals were pre-treated with morphine (5 mg/kg; intraperitoneal injection), PA (50 mg/kg or 75 mg/kg; intraperitoneal injection), or vehicle (saline + 0.2% Tween 80; intraperitoneal injection) with the subsequent induction of orofacial pain. The test doses of PA were determined based on its activity for general pain . All test substances were purchased from Sigma Aldrich (Missouri, USA).
In order to prevent biases in data collection and analysis, randomization and blinding procedures were used during the experiments. As the treatment groups had the same sample size, a block randomization protocol was conducted prior to the experiments, as suggested by Suresh (available online at www.randomization.com , accessed April 31, 2015). Blinding was performed by designating arbitrary codes using the software Biostat 2009 for each test variable: (1) nociception test; (2) treatment group; (3) treatment substance and dosage; (4) follow-up time; and (5) sample size. The categorization and codification of the animals was conducted by a researcher prior to the experiments (researcher A). Another researcher, who was blinded to the substances, proceeded with the administration of treatments (researcher B). After induction with the algogenic substance by researcher A, the evaluation of nociception was performed by another researcher, who had been trained and was blinded to the treatment groups (researcher C).
The formalin test was initially proposed by Clavelou et al. to evaluate general pain induced in the plantar region ; it was later adapted to orofacial pain by Luccarini et al . The formalin test in the orofacial region is considered to be the only animal model of persistent cutaneous nociception in the trigeminal region. It allows the magnitude of nociception generated by a long-duration chemical stimulus to be assessed, in contrast to the majority of tests that involve a brief stimulus . In this study, the orofacial nociception was induced with the administration of 20 μl of a 20% formalin solution (Êxodo Científica, São Paulo, Brazil) into the right upper lip of the mouse (paranasal or vibrissa region) with a 27-gauge needle ; this was done at 0.5 h after treatment. After the administration of formalin, the animal was immediately placed in a mirrored box and another researcher (researcher), who had been trained and was blinded to the treatment groups, used a chronometer (Casio Co., Ltd, Japan) to record the time spent by the animal rubbing the injected region with its fore or hind paws. The measurement of nociception behaviour in this test is based on two distinct phases of formalin activity : phase 1 or neurogenic (0–5 min) and phase 2 or inflammatory (15–40 min).
The second test (capsaicin-induced nociception), described by Pelissier et al. and adapted by Quintans-Júnior et al. , consisted of the administration of 20 μl of a 2.5-μg solution (saline + 0.2% Tween 80) of capsaicin (Sigma Aldrich, Missouri, USA) into the right upper lip of the mouse at 0.5 h after treatment. Subsequently, the nociceptive behaviour was timed by a blinded researcher (researcher) for 42 min after the induction of nociception.
The third test (glutamate-induced nociception) was first described by Beirith et al. and was adapted to the orofacial region. This last test consisted of the subcutaneous administration of 40 μl of a 25-μM glutamate solution (Sigma Aldrich, Missouri, USA) into the right upper lip of the mouse with a 27-gauge needle. After the administration of glutamate, the animals were placed in an observation box and monitored for their orofacial nociceptive behaviour for 15 min by a blinded researcher (researcher) .
At the end of the experiments, researcher A entered the data into the software PASW Statistics version 18.0 (SPSS Inc., Chicago, IL, USA) and another researcher (researcher D) performed a blind statistical analysis. Subsequently, all data were re-coded by researcher A.
Descriptive and inferential statistical analyses were performed using PASW. The treatment groups were compared using the Mann–Whitney test, with a 95% confidence level (α = 5%). The percentage of inhibition was calculated in Microsoft Excel 2013 for Windows using the formula (A − B)/A × 100, where A is the median for the control group and B is the median for the treatment group . Additionally, the hypothesis of a linear association between treatment substances was verified with Spearman’s test (α = 5%). The effect size was calculated for both tests (Mann–Whitney and Spearman) using Hedge’s g formula, described with its confidence intervals (CI) and power .
The administration of PA (50 mg/kg or 75 mg/kg) led to a significant decrease in the orofacial nociceptive behaviour ( P < 0.05) when compared to the negative control in both the neurogenic and inflammatory phases of the formalin test ( Figs. 1 and 2 ). For the neurogenic phase, the statistical differences showed high power: 84.6% (95% CI 0.48–2.31) for 50 mg/kg and 96.2% (95% CI 0.82–2.76) for 75 mg/kg. This sample size was not able to detect a difference between the doses of PA tested (Mann–Whitney test, P = 0.49, power 24.2%, 95% CI −0.33 to 1.32), despite the occurrence of the effect size compared with the control group. For the inflammatory phase, the 50 mg/kg dose showed moderate power (power 73.6%, 95% CI 0.30–2.08) and the 75 mg/kg showed high power (power 82.3%, 95% CI 0.44–2.26).
In the neurogenic phase of the formalin test, the 50 mg/kg dose of PA reduced the nociceptive behaviour by 51.76% and the 75 mg/kg dose led to a reduction of 60.00%. The percentage reduction of nociception in this phase using morphine was found to be 52.65%. An even greater inhibition was caused by PA in the inflammatory phase, with a 72.05% reduction for the 50 mg/kg dose and an 83.33% reduction for the 75 mg/kg dose; morphine decreased nociceptive behaviour by 69.11% in this phase.
A significant reduction in the capsaicin-induced orofacial nociceptive behaviour was observed for the two tested doses of PA ( P < 0.01) as compared to the negative control ( Fig. 3 ). However, a low effect size was detected on comparing the 50 mg/kg and 75 mg/kg doses of PA with the negative control group (power 43.3%, 95% CI −0.07 to 1.62, and power 53.6%, 95% CI 0.04–1.76, respectively). In this test, both doses of PA and morphine were included in the same homogeneous subset with no significant differences between their median values (Mann–Whitney test, P > 0.05) ( Fig. 3 ).
For the capsaicin test, PA inhibited the orofacial nociceptive behaviour by 62.30% (50 mg/kg) and 72.43% (75 mg/kg), while morphine resulted in an inhibition of 80.08%.
When compared to the control group, PA (50 mg/kg or 75 mg/kg) caused a significant reduction in the orofacial nociceptive behaviour induced by glutamate ( P < 0.05) ( Fig. 4 ). In this test, only the 75 mg/kg dose showed a significant difference (power 99.2%, 95% CI 1.11–3.16). Moreover, only in this test was a high power detected for the 75 mg/kg dose when compared to the 50 mg/kg dose (Mann–Whitney test, P = 0.02, power 97.8%, 95% CI 0.94–2.91).
The percentage of inhibition was 48.57% at 50 mg/kg and 71.43% at 75 mg/kg of PA, while morphine resulted in a reduction of 65.57%. The Mann–Whitney test showed no significant differences between medians of the other experimental groups ( P > 0.05) ( Fig. 4 ).
The occurrence of a linear antinociceptive effect comparing the treatment groups was also calculated for each test using Spearman’s correlation test. The sum of all treatment groups for each test ( n = 32) was able to provide 95% statistical power to detect an ordinal association between groups (vehicle, 50 mg/kg PA, 75 mg/kg of PA, and morphine). This sample size was verified, considering a one-sided type I error of 0.01. The treatment groups in each test (formalin, capsaicin, and glutamate) showed an association ( P < 0.01) with an ordinal decrease in orofacial nociceptive behaviour with high power analysis. An acceptable power of association was found for the formalin (inflammatory phase), capsaicin, and glutamate tests ( Table 1 ).