The majority of behavioral pain tests currently in use are only applicable to the hindpaws or tail. Thermal tests such as the hotplate/cold plate or hot-water bath immersion are very difficult to perform in the facial region. The commonly used Hargreaves plantar test, which provides a thermal stimulus with the aid of a movable infrared source is a bulky machine – a small adaptor is required for this type of stimulation to be applied in the facial region. Moreover, in order for the heat intensity delivered to be even, the heat source should always be placed at the same distance from the animals' face, which in freely moving animals is virtually impossible.
On the other hand, the specific characteristics of the orofacial region allow for certain functional tests that cannot be performed with other body parts; in particular, gnawing, chewing, and willingness to chew can be observed and quantified. Thus, we can observe food intake decrease following a TMJ inflammation (Harper et al. 2000
), reduction in the bite force following masseter muscle injections of CFA (Ro 2005
), a decrease in food-pellet-releasing lever pressing and feeding following both TMJ and masseter muscle inflammation (Thut et al. 2007
), and decrease in gnawing through objects following similar inflammation (Dolan et al. 2010
). All these reflect symptoms in human orofacial pain patients who avoid pain-potentiating chewing. However, in some cases the observable changes in these behaviors may be subtle and it is of interest to also be able to quantify orofacial hyperalgesia in response to a stimulus.
Pain-related spontaneous behavior
As most of the currently available pain-testing devices prove impossible to use in the facial region, only a relatively small number of studies has been performed to date and most focus on spontaneous responses (see and ). The most used to date and the most simple, is the formalin test, which involves the injection of the irritant chemical into the upper lip of the rodent and observing the licking and scratching behavior. This model has been first described by Clavelou et al. (1989
) and further used by a number of groups in both rats (Luccarini et al. 2004
; Raboisson and Dallel 2004
) and mice (Luccarini et al. 2006
; Bornhof et al. 2011
). In TMJ injections of formalin, a head-flinching behavior and chewing-like motions of the mandible were also observed (Roveroni et al. 2001
). Formalin is particularly useful for evaluating primary and secondary hyperalgesia alterations in transgenic mice. Capsaicin mustard oil and glutamate are other substances that elicit spontaneous nocifensive behaviors and also have been applied in the orofacial region in rats (Pelissier et al. 2002
; Hartwig et al. 2003
; Ro and Capra 2006
) and mice (Quintans-Junior et al. 2010
In a study of chronic constriction of the IoN in rats, Vit et al. (2006
) measure the “eye-closure response” as an indication of pain, based on the paroxysms of pain in TN. They show that such eye-closure response can be temporarily blocked with an analgesic dose of morphine and demonstrate an analgesic effect of an interfering-RNA directed against Cx43, a protein found in satellite glial cells, thought to be implicated in neuropathic pain. Such method, once sufficiently validated, could be useful for the study of spontaneous neuropathic responses.
Other spontaneous behaviors such as changes in weight, spontaneous grooming, aggression, and other changes in behavior can be monitored in pain studies (Mogil 2009
). Vos et al. (1994
) have quantified some of such behavior in their seminal article on the chronic constriction of the IoN. They found that animals with the constriction explored less, exhibited freezing like behavior, defecated more, and gained less weight compared with controls. However, such behavioral studies tend to be time consuming and difficult to quantify, and also it is difficult to ascertain whether they indicate stress, pain, paresthesia, or avoidance behavior and most studies performed in orofacial pain do not include measurements of such spontaneous behavior.
The newly developed Rat and Mouse grimace scales, which measure facial “grimaces” of the rodents following a painful stimulus (so far, only used in nonhead areas (Langford et al. 2010
; Sotocinal et al. 2011
), may prove to be useful in trigeminal pain models. However, it remains to be seen whether the presence of inflammation in the face would affect the quality of the “grimace”. Also, this method is only valuable for pain of short-to-moderate duration and would not be useful for chronic studies.
Stimulus-evoked behavior testing methods
The whisker pad region of rodents is a tricky area to study stimulus-evoked behavior. This region has a rich mechanosensory receptor sheet, which is stimulated in nearly continuous haptic activities during exploratory behavior, and these complex whisker movements can complicate the testing. On the other hand, the IoN is a large and relatively easily accessible sensory-only nerve, and innervates a large area which has been the region of choice for many studies.
When studying stimulus-evoked behavior in the orofacial region, one of the major pitfalls is the criterion of the “response”. In the paw region, a reflex-like withdrawal of the paw from the stimulation source is usually considered as the response. In the facial region, the responses may vary from scratching and blinking to grimaces and removing the head. All possible responses need to be classified before testing and analyzing. Vos et al. (1994
) have set a standard for orofacial pain testing in the first report of IoN-CCI in rats. They have thoroughly studied the rat's behavior following the CCI intervention, including spontaneous activity (face grooming, exploratory behavior) and evoked behavior which included stimulation with various thicknesses of von Frey filaments and a pin prick. Based on the responses, a “response score” was attributed, combining the various criteria. We have recently adapted a simplified version of such quantification in mice, where face-grooming behaviors, withdrawal and aggressiveness toward the probe have been totaled to achieve a response score (Krzyzanowska et al. 2011
Apart from the challenges of approaching the testing probes to the area of interest, the facial region is tricky to stimulate as rodents tend to actively move their heads, which is especially pronounced in mice. In addition, mice are particularly active and escape when the stimulating object (such as a von Frey hair) is approached. Rats, on the other hand, are much calmer and it is possible to perform stimulations with von Frey hairs, as demonstrated in numerous publications (Vos et al. 1994
; Idanpaan-Heikkila and Guilbaud 1999
; Deseure et al. 2002
; Martin and Avendano 2009
; Martin et al. 2010
In mice, to date, only a few publications have reported the use of von Frey hairs in the orofacial region. Recently, in a study involving partial IoN ligation, the authors have behaviorally tested mice placed on a mesh floor, restricted within a 8-cm-diameter plastic cup from the top, and stimulated by von Frey hairs from underneath (Xu et al. 2008
; Aita et al. 2010
). With this approach, it can be difficult to see exactly where the filament is stimulating; moreover, the filament advances parallel to the skin rather than at a 90°, as recommended. In this case, it would be impossible to press the filament against the whisker pad region exerting a bend in the filament. In two other studies of mouse neuropathic facial pain, the animal was held by the experimenter during testing with either von Frey filaments (Seino et al. 2009
) or a heat source (Luiz et al. 2010
). The holding method requires numerous habituation procedures, is stressful for the mouse and results in the animal being held in an unnatural position, restricting its movements, thus limiting the scope of response. In contrast, in studies involving application of an inflammatory agent (carrageenan) to the orofacial area, the mice were allowed to freely move in a steel tank, with the von Frey filaments being applied from above (Yeo et al. 2004
; Vahidy et al. 2006
; Poh et al. 2009
; Tang et al. 2009
). Although relatively unstressful, due to the active nature of the animals it would be challenging to stimulate them and, importantly, it would be difficult to ascertain where exactly the probe touched the face or what response was obtained. We recently proposed an alternative way of restraining the mice, which involves the mouse being placed in a box, with its tail being attached to a special device (Krzyzanowska et al. 2011
). Although not entirely stress-free, this set-up allows the animal to move its head and forepaws freely and allows the examiner to observe various types of responses. Also, plasma corticosterone measurements showed this type of set-up to be less stressful than the hand-held method.
While von Frey hairs can be used for determining mechanical thresholds, the air puff method is a useful tool for studying the effect of a completely non-noxious stimulus. Ahn et al. have used this method in several facial neuropathy (Ahn et al. 2009a
) and inflammation (Ahn et al. 2004
; Jung et al. 2006a
) models in rats to test whether the animals develop mechanical allodynia. They showed that while naïve animals do not respond to an air puff of 40 psi, animals which had an IL-1β induced inflammation or TG compression responded to air puffs of much lower pressure (5 psi). Our group has observed similar results with the air-puff method in mice which underwent an IoN-CCI or CFA inflammation (Krzyzanowska et al. 2011
Thermal testing of the orofacial region is even more complicated. The machinery needed for the thermal stimulation, such as a tube with the heat beam, is much larger than the von Frey hairs, and approaching such apparatus may scare the animal. Furthermore, the light shining in the animals eye may be unpleasant. The skin of the snout is covered by hair – unlike the paw which has a glabrous surface – which makes it difficult to apply a specific desired temperature. In addition, a thermal probe will first touch the facial hairs and vibrissae, thus activating the low-threshold mechanoreceptors before producing the thermal sensation, thus a radiant heat source is more suitable (Imamura et al. 1997
), or the animal should be shaved (Eriksson et al. 2005
; Neubert et al. 2005b
). The latter situation is not entirely physiological as some of the normal sensory information is transmitted through the facial hair.
, Rosenfeld et al. designed a facial nociception device which was mounted onto the skull of the animals and delivered heat to the cheek. The responses measured were scratching or face-rubbing by fore or hind limbs. This apparatus, however, requires surgery to install the device and is clearly uncomfortable for the animal and has not been widely adapted. A more practical test, developed by Imamura et al. (1997
), involves placing a rat in a restrainer so that only the snout is visible for noxious radiant heat-beam stimulation, at the same time shielding the eyes animals from the heat light. With this apparatus, they showed significant decreases in withdrawal latencies after a constriction of the IoN. In this set-up, the animals had to be thoroughly habituated to the apparatus before behavioral testing in order to avoid any stress-facilitated changes in behavior and analgesia. A similar contraption was reported by Ahn et al. (2009a
) who induced neuropathic pain with an injection of the demyelinating agent LPA into the trigeminal ganglion of the rat. They restrained the rats in a cylindrical acrylic restrainer and applied heat stimulus using an infrared thermal stimulator (diode laser) placed 10 cm away from the vibrissal pad. However, they have failed to observe any differences in responses to this stimulus between the vehicle- and LPA-treated groups (Ahn et al. 2009a
). This could be due to the nature of the model, which is more sensitive to mechanical stimuli. Other recent studies used infrared irradiation to thermally stimulate the face of mice and rats held by the investigator (Luiz et al. 2010
) or of mice restrained in a plastic tube (Shinoda et al. 2011
). Both groups, however, do not specify the type of thermal source machine used and the restraint of the animal by the investigator is not optimal (see above). Moreover, Shinoda and colleagues repeatedly anesthetized the animals in order to place them in the plastic tube for the behavioral testing. While behavioral procedures were performed 30 min after anesthesia, one cannot exclude some residual effects of the isofluorane. Several other studies using thermal stimulus have been reported using lightly anesthetized rats (Tzabazis et al. 2005
; Niv et al. 2008
; Cuellar et al. 2010
); however, little more has been published in awake animals.
Operant behavior paradigms
A new type of mechanical and thermal stimulation has been proposed by the group of Neubert et al. They have developed a set-up which allows for the observation of operant responses to painful stimuli. In this paradigm, the rodent has a choice between receiving a reward (sweet condensed milk) or preventing receiving an aversive (painful) stimulus. In order to receive the reward, the rodent must poke its snout through an opening equipped with a thermode so that the aversive stimulus is obtained at the same time as the reward. The painful stimuli can be heat (Neubert et al. 2005b
), cold (Rossi et al. 2006
), or a mechanical stimulus (Nolan et al. 2011
), resulting in the reduction of the reward-seeking behavior following peripheral inflammation – an observation which has been demonstrated to be reversed with analgesic drugs (Neubert et al. 2005b
). This testing system has also been adapted for studies on mice, showing that TRPV1−/− mice are insensitive to the 37–52°C heat range (Neubert et al. 2008
Another recent study proposes an alternative way of estimating trigeminal pain based on the rodents' natural tendency to gnaw on objects obstructing their passage in a narrow tube (Dolan et al. 2010
). They hypothesize that nociception-induced gnawing dysfunction can be used as an index of orofacial nociception in an animal model, reflecting the trigeminal pain-induced unwillingness to chew in humans, and demonstrate this in three different orofacial pain models in mice.
The operant behavior paradigms allow to observe a more spontaneous type of behavior when compared with stimulus-evoked studies. However, they require considerable training and importantly, have a motivational component which makes the interpretation of the pain-related behavior more complicated (Mogil 2009