mise à jour du
20 novembre 2005
Arch Oral Biol
2005; 50; 3; 333-340

Behavioural responses following tooth injury in rats
Chudler EH, Byers MR
Department of Anesthesiology, University of Washington, Seattle


Dental pain is estimated to affect 12-14% of the US population annually. Despite its high incidence, dental pain relies on century-old methods for its diagnosis. Although clinical treatment of dental pain has improved with new understanding of psychological, sensory and pharmacological mechanisms of trigeminal pain, there are still significant areas for improvement.
Several experimental models of orofacial nerve injury have been developed in recent years. Such methods have provided a wealth of behavioural, neurophysiological, neuroanatomical and neurochemical data concerning peripheral and central nervous system changes subsequent to injuries of the trigeminal system. Although acute dental pain models have been developed in rats and ferrets, few investigations concerning the behavioural consequences of tooth injuries in animals have been conducted.
Most studies investigating the behavioural consequences of dental pain in animals have relied on single indices of pain often in response to electrical, mechanical or chemical stimulation of the tooth. For example, the jaw-opening reflex evoked by electrical stimulation of the tooth has been used extensively to investigate pain in rats. More recently, Chattipakorn et al. used bacterial lipopolysacchande to inflame the ferret tooth pulp to evaluate pain behaviour based on the number of tongue protrusions and Chidiac et al. devised a pain scale based on abnormal grooming and movements following intrapulpal capsaicin or formalin applied to the rat incisor. Tooth extraction has also been proposed to model dental pain, but questions have been raised about its usefulness.
The present study sought to determine the shortterm behavioural effects of tooth injuries with crystals of Fluorogold (FG) implanted into molar teeth of adult rats. We were interested in developing a set of quantitative measures of behavioural changes that could be used with a variety of pain models and retrograde neural labelling techniques.
Changes in spontaneous behaviour and body weight following injuries to the infraorbitat nerve, sciatic nerve and experimental arthritis have been used as indices of pain in rats. Alterations in steep patterns have also been observed within the first few days after chronic constriction of the sciatic nerve. Tooth injuries, as produced in the present study, result in behaviouraL changes similar to those observed in these other pain models: reduced body weight gain, increased time spent freezing and decreased time spent exploring as welt as inducing yawning behaviour in some of the injured rats. Comparable behavioural observations, including changes in movement, steep and food intake, have been used to evaluate human pain behaviour. Our data showing greater behavioural changes when three to four teeth are involved compared to the group with two injured teeth are consistent with our previous observation that larger injuries for rat molar teeth caused a slower return to normal weight gain and more extensive ganglionic satellite cell can central c-Fos responses. Experiments can therefore be designed to study dental injuries and their effects on trigeminal neuronal cytochemistry and function without behavioural changes (two teeth) or with behavioural changes (three or more teeth).
The behavioural tests presented here can now be used to establish the behavioural consequences of the various dental neuronal retrograde labelling procedures. This would be especially important for investigations that label dental neurons for subsequent patch clamp in vitro analyses, in order to know whether those neurons come from animals with significant pain behaviour. Previous studies have often assumed that rats return to normal behaviour when they resume feeding after surgery. Our data, however, show that body weight, yawn and motor behaviour can show delayed recovery.
FG labelled many dental neurons in the corresponding trigeminal division of the ipsilateral ganglion, and we estimate that more than 500 dental neurons were routinely labelled in right trigeminal ganglia in the three tooth labelling group. We were surprised to find that less than 20% of the FG neurons co-localized CGRP, since others have strong evidence that FG is an excellent tracer for following dental neuronal changes after tooth injury. It is likely that our implantation of FG crystals directly onto pulp exceeded toxicity limits for some of the neurons. Furthermore, we were analysing the neurons at three days after implant, while Pan et al. placed 200 µl of 5% FG into shallow cavities along the buccal surface of the first maxillary molar, sealing it in place, and then waited three to five weeks before analysing the cytochemistry of neurons. Such dentinal labelling procedure and another FG technique that shows penetration of FG from oral sulcus into intact rat molars did not cause significant weight loss, showing that exposure of rats to FG does not by itself cause reduced feeding.
Although we observed differences in body weight, exploration and freezing in tooth-injured and sham-operated rats two days after injury, these behavioural changes were statistically significant only at the three-day observation period. It is possible that the absence of significant differences between groups at the two-day period was due to residual anaesthetic effects. However, other movement assays (rearing, grooming, jaw motions) were normal at two to three days after these FG implants. Unlike changes in body weight, exploration and freezing, statistically significant alterations in jaw movements, rearing and facial and body grooming were not observed within the time frame of the present study. Using the infraorbital nerve constriction model, Vos et al. reported that rats displayed increased facial grooming within three days after surgery, but Benoliel et al did not observe such changes until two weeks after a similar surgery. It is possible that tooth injuries do not result in abnormal sensory phenomena (i.e., paresthesia) during the observation period of the current experiments.
A unique observation was the increased number of yawns seen after tooth injuries. Although the significance of yawning following tooth injuries is unclear, it may be related to stress. Yawning can be elicited in horses by abdominal stress and in rats by foot shock stress. Yawning is also a symptom of some migraine headaches in humans. It is unlikely that the yawns after tooth injuries were caused by damage to the jaw or temporomandibular joint or to stress of the novel testing environment because sham-operated rats did not yawn. Moreover, because rats do not yawn following infraorbital nerve injury, it is likely that yawning is made in direct response to sensory information from the tooth. The use of a rat strain that displays a high frequency of yawning may help resolve this question because such animals might display a higher number of yawns following tooth injuries and reduce the variability of this behavioural marker. Additional experiments to observe rats at other times of the day and for longer observation periods are also warranted. It is possible that yawning is an important behavioural consequence of tooth injury that has been overlooked in previous investigations.
Tooth injuries with implants of the tracer Di-I, similar to the injury procedure in the present study, increase the number of trigeminal ganglion neurons encircled by glial fibrillary acidic protein (GFAP) three days after surgery. Such alteration in satellite cell cytochemistry may not indicate altered electrophysiology, as no changes in trigeminal neuronal function were detected in the maxillary division where many uninjured neurons exhibited prolonged increase of GFAP expression, after injuries to the mandibular neurons that project into the inferior alveolar nerve. However, significant increases in the expression of P/Q-type (Ca2.1) calcium channels three days after tooth pulp exposure have also been observed 20 after pulp exposure injury.
The long-term behavioural consequences of tooth injury were not determined in this study. Moreover, motor behaviours were measured by the time spent in locomoting, rearing and freezing. A more detailed analysis of fine motor movement may reveaL additional changes. Benoliel et al. reported reduced bite force and altered bite patterns within a few days after application of a pro-inflammatory agent to the infraorbital nerve of the rat. A recent report suggests that bite force in trained rats is also a valid measure for craniofacial muscle inflammatory pain and it would be interesting to test such an assay for behavioural responses to tooth injury. Nevertheless, our results demonstrate that changes in normal behaviour patterns can be used to assess pain in rats.

Equine Vet J.1990; 22; 4; 241-243
Naloxone-induced abdominal distress in the horse.
Kamerling SG, Hamra JG, Bagwell CA.
Department of Veterinary Physiology, Pharmacology and Toxicology, School of Veterinary Medicine, Baton Rouge, Louisiana
Endogenous opioid peptides have been implicated in the regulation of pain perception, behaviour, gastrointestinal activity and other physiological responses. However, the functional role of these peptides in the horse has yet to be elucidated. The opioid antagonist, naloxone, is often administered to infer endogenous opioid effects. In the present study, naloxone (0.75 mg/kg bodyweight) was administered to eight Thoroughbred racehorses and a number of behavioural and autonomic responses were measured. Naloxone produced rapid onset diarrhoea, restlessness, abdominal checking, tachycardia, tachypnoea, paradoxical yawning and diaphoresis. These responses described an acute abdominal distress syndrome similar to spasmodic colic. Results from this study suggest that, in the horse, endogenous opioids: 1) influence behaviour, 2) modify intestinal activity and sensation, and 3) if perturbed, may be involved in pathophysiology of colic.