A novel rat model of temporomandibular disorder with improved face and construct validities

Abstract

Aims:

Temporomandibular disorders are a cluster of orofacial conditions that are characterized by pain in the temporomandibular joint (TMJ) and surrounding muscles/tissues. Animal models of painful temporomandibular dysfunction (TMD) are valuable tools to investigate the mechanisms responsible for symptomatic temporomandibular joint and associated structures disorders. We tested the hypothesis that a predisposing and a precipitating factor are required to produce painful TMD in rats, using the ratgnawmeter, a device that determines temporomandibular pain based on the time taken for the rat to chew through two obstacles.

Materials and Methods:

Increased time in the ratgnawmeter correlated with nociceptive behaviors produced by TMJ injection of formalin (2.5%), confirming chewing time as an index of painful TMD. Rats exposed only to predisposing factors, carrageenan-induced TMJ inflammation or sustained inhibition of the catechol-O-methyltransferase (COMT) enzyme by OR-486, showed no changes in chewing time. However, when combined with a precipitating event, i.e., exaggerated mouth opening produced by daily 1-hour jaw extension for 7 consecutive days, robust function impairment was produced.

Key findings:

These results validate the ratgnawmeter as an efficient method to evaluate functional TMD pain by evaluating chewing time, and this protocol as a model with face and construct validities to investigate symptomatic TMD mechanisms.

Significance:

This study suggests that a predisposition factor must be present in order for an insult to the temporomandibular system to produce painful dysfunction. The need for a combined contribution of these factors might explain why not all patients experiencing traumatic events, such as exaggerated mouth opening, develop TMDs.

Graphical abstract

Introduction

Temporomandibular (TM) disorders (TMDs) are a cluster of orofacial conditions that adversely affect 6-12% of the general population[1][2][3] involving the masticatory muscle complex, the TM joint (TMJ) and associated structures, characterized by impaired/limited jaw function, often accompanied by pain [4][5][6][7][8]. The pain can be debilitating and even affect a person’s ability to speak, chew, swallow and make facial expressions. Stretching of the TMJ-associated structures, as during endotracheal intubation[9] or routine dental procedures[10][11], has been proposed to increase risk for TMD initiation. Our understanding of the interactions between mechanical forces, biochemical mediators, peripheral and central nociceptive signals, and psychosocial factors affecting the use of the TMJ and the experience of pain is limited because animal models of orofacial pain are less studied than pain in other areas of the body[12].

Another limitation is that there are two critical deficiencies in current preclinical models of TMDs. The first is a need for models with increased face validity (similarity in clinical signs and symptoms). Most TMD rodent models rely on endpoint measurements of thermal hyperalgesia or mechanical allodynia using von Frey fibers[12]. However, chewing induces the highest levels of pain in TMD patients, not superficial touch, and patients are frequently prescribed jaw rest during episodes of TMJ or masticatory muscle pain[13][14][8][15][16]. Therefore, an objective animal assay quantifying nociceptive-induced dysfunction of the TMJ while chewing would improve face validity.

The second deficiency is a need for models with improved construct validity (similarity in pathophysiological mechanisms). Current rodent models of painful TMD rely on chemical interventions like injection of inflammogens - carrageenan (CARR) or Complete Freud’s Adjuvant (CFA) - into the TMJ or associated muscles[17][12]. In contrast, the pathophysiology of TMDs indicates that predisposing factors interact with initiating factors to produce pain. Predisposing factors include 1) catechol-O-methyltransferase (COMT) gene polymorphisms reportedly implicated in changes in pain tolerance[18][19][20][21]; 2) aberrant pain and sensory processing caused by previous inflammation[22][23] or 3) alterations in central nervous system function[24]. When one of these is combined with a triggering factor like jaw overloading, persistent or long-term sensitization of the TM system is produced[25][23][26][27]. In this sense, developing an animal model that allows the investigation of the contribution of these factors to TMDs would be extremely important to advance our understanding of how different conditions interact and lead to pain in humans.

To address these deficiencies, we developed a protocol combining precipitating and predisposing factors, allowing the study of their interactions on the development of TMDs. Our protocol uses chewing function as endpoint surrogate for pain, utilizing a modified version of the previously described “dolognawmeter”, a device designed to assess nociceptive-induced gnawing dysfunction in mice[28]. Our version, named the “ratgnawmeter”, was adapted for use in rats, and allowed the evaluation of the effects of exaggerated jaw opening on chewing function in the presence and absence of predisposing inflammation or COMT inhibition[18][19]. We hypothesized that clinically significant painful masticatory dysfunction would only manifest when both the predisposing factor (inflammation or decreased COMT activity) and the triggering insult (jaw extension) were combined.

Materials and Methods

Experimental Animals:

Adult male Sprague Dawley rats (Charles River Laboratory, USA) weighing 250 - 350g were used in all experiments (n = 120). Rats were housed 2-3 per cage and kept in a temperature-controlled environment on a 12:12 hour light-dark cycle (lights on at 6:00 A.M., lights off at 6:00 P.M.) with ad libitum access to food and water. Tests were performed between 9:00 A.M. and 5:00 P.M. All animal procedures were reviewed and approved by the Institutional Animal Care and Use Committee of the University of Utah and conformed to National Institutes of Health Guidelines for the Care and Use of Laboratory Animals. Animals were used only once and, after the experiments were completed, euthanized by carbon dioxide inhalation followed by cardiac puncture as recommended by the University guidelines.

Design of the Ratgnawmeter:

The device used in this study to evaluate gnawing function and the impact of treatments on chewing activity is based off a similar apparatus previously used for study in mice[28]; We termed it “ratgnawmeter” to distinguish it from the dolognawmeter device used in mice. Its framework is constructed out of a polyvinyl chloride (PVC) material (Sheet Stock PVC Type 1, Grainger, Lake Forest, IL, USA), and is customized to fit in a standard facility rat cage (Maxi-Miser® #8, Thoren Caging Systems, Hazleton, PA, USA). The ratgnawmeter has a set of openings that can accommodate a removable PVC tube (length = 28 cm). A large PVC tube (diameter = 6 cm) is used for rats weighing greater than 350g. An adapter can be used to secure a smaller tube (diameter = 5 cm) to the apparatus and is used for rats weighing less than 350g. The tubes are designed to be large enough for the rat to climb in, however, is small enough that the rat does not have enough room to turn around. Two sets of holes lie near the exit of each PVC tube. These square (2 x 2 cm) and radial holes (diameter = 1.3 cm) allow obstacles to be placed across the PVC tube opening. The polystyrene and resin obstacles are in turn connected to a spring-loaded mechanism that retracts the obstacles and starts and stops a timer (Arduino Uno R3 board, Arduino, Somerville, MA, USA) located on the top of the ratgnawmeter and is used to determine when chewing begins and ends. Figure 1 shows the ratgnawmeter in detail.

Figure 1:

The ratgnawmeter. (A) General view of the ratgnawmeter device. The rat is put into the tube located under the apparatus (C). Once in the tube, the rat will chew through the foam stick (panel B, b) releasing the spring-loaded mechanism (c), which activates the timer.

Ratgnawmeter Behavioral Testing:

The ratgnawmeter is a behavioral assay that measures the time taken for a rat to completely chew through a set of obstacles. The general experimental approach for behavioral testing in the ratgnawmeter is as follows:

1. The rat is held at the base of the tail and introduced to the rear end of the ratgnawmeter. The rat will instinctually climb into the tunnel, and the apparatus can be placed in the standard facility cage.

2. Once inside the ratgnawmeter, the first obstacle the rat encounters is a polystyrene foam stick (l = 10 cm, w = 1.5 cm, h = 1.5 cm, Figure 1, panels B and C, b). Breaking this obstacle activates a spring-loaded mechanism (Figure 1, panels B and C, c) and the two broken pieces laterally retract to begin an electronic timer.

3. The rat can then climb forward and access the next obstacle, an ethylene-vinyl acetate dowel commercially available as a glue stick (AdTech Hot Sticks, 10in x 0.44in, clear, Hampton, NH, USA, Figure 1, panels B and C, a). Figure 2 (panel A) shows a rat chewing through the resin dowel. Once broken, the spring-loaded mechanism stops the timer.

4. The assay is now complete, and the rat can exit the tube into the front enclosed portion of the cage. The time taken to chew through the dowel is reported in seconds. (A video of the ratgnawmeter can be found in the Supplementary Material)

5. The rat can then be returned to its facility cage.

6. The ratgnawmeter is cleaned between each set of experiments.

Figure 2: Ratgnawmeter behavioral testing.

(A) shows a rat chewing through the resin dowel. (B) Average chewing time during the five required ratgnawmeter behavioral training sessions. Chewing time on the first training session averaged 339 seconds and plateaued out by the fifth session at 196 seconds. On the fifth training session, the baseline chewing time was determined. Comparison of the mean chewing time on the first day of training (339.6 ± 159 seconds, mean ± SD of n = 100) with that on the 5^th day (195.6 ± 102 seconds) shows a decrease in the variability and habituation of the rats to the method.

The rats must first be trained to chew through the obstacles, requiring 5 consecutive days to reach a consistent chewing time. This training must be conducted prior to starting experiments. Chewing through the resin dowel (panel B, a) releases the mechanism that stops the timer. (C): bottom view of the apparatus, with the foam stick (b) and the resin dowel (a) installed (left side), and the start/stop timer mechanism (c). The red arrow indicates the entrance of the tube. (D): the ratgnawmeter placed inside the facility cage.

Drugs, Doses and Administration Routes:

The following drugs were used in this study: A 10% formalin solution (Fisher Scientific) diluted in 0.9% saline to a concentration of 2.5%[29]; the inflammatory agent λ-carrageenan plant mucopolysaccharide (CARR, from Sigma-Aldrich) diluted in 0.9% saline; the glucocorticosteroid dexamethasone (Tocris) diluted in dimethyl sulfoxide (DMSO, Sigma-Aldrich); and the COMT inhibitor OR-486 (Tocris) diluted in a vehicle containing DMSO:ethanol:distilled water[30], in a proportion of 5:3:2, to a concentration of 15 mg/kg/day for use in an Alzet Osmotic mini-pump (model 2002, Durect Corporation, Cupertino, CA, USA). The doses and routes of administration were based in previous studies: Formalin (50μl)[29] or CARR (100 μg/25 μl)[31]; Solutions were injected into the TMJ using a 27G needle attached to a Hamilton syringe; Dexamethasone was injected either intraperitoneally (i.p., 10 mg/kg, in the experiment using formalin)[32] or subcutaneously in the nape of the neck (s.c., 1 mg/kg, in the experiment using CARR)[33].

Jaw Extension Procedure:

Jaw extension was achieved by use of a custom sling-and-pulley system built to produce prolonged, calibrated jaw opening in rats, as previously described [34]. Only a 3.5N force was used for these sets of experiments (350g weight). However, the amount of weight attached to the pulley can be interchanged to vary the vertical force placed on the TMJ and surrounding muscles. Rats were initially placed in an induction chamber saturated with 4% isoflurane until deeply anesthetized. The animals were then put on a loading platform where the lower mandible is held static by a lower hook and the maxilla attached to the sling-and-pulley system (see Figure 4 for details). The platform was then placed into the induction chamber saturated with 2-3% isoflurane for 1 h. A heating pad was used to minimize heat loss during prolonged anesthetic administration. This protocol was repeated for 7 consecutive days as previously described[35][34]. All rats were continuously monitored throughout the protocol for proper respiratory rate. Following the 1 h jaw loading period, animals were recovered at room air on a heated pad in their home cage. On days in which the jaw loading procedure was combined with testing in the ratgnawmeter, testing was performed 2 to 4h after anesthetic administration to allow adequate recovery from the anesthesia.

Figure 4: Jaw extension apparatus, used to model the repeated exaggerated opening of the rat mouth, and overload of the TMJ system.

On the left (upper panel) the loading platform, with the sling-and-pulley system attached to the adjustable weights, used to support and keep the rat’s mouth open, and the heating pad, that keeps the rat body warm during the anesthesia, are shown. Once the rat is put to sleep and loaded in the platform, with the upper and lower jaws secured open by the sling-and-pulley system (as shown in the right panel), the platform is put into the anesthesia chamber (left, lower panel). The jaw extension protocol used in this study lasted 1 h and was repeated daily for 7 consecutive days.

Catecholamine-O-Methyltransferase (COMT) Inhibitor Treatment:

Alzet Osmotic minipumps were used to deliver a 14-day sustained dose of 15 mg/kg/day of OR-486. After training in the ratgnawmeter and reaching behavioral baseline, rats had the osmotic pumps implanted subcutaneously slightly posterior to the scapulae as per the manufacturer guidelines. Each rat was given two days to recover from surgery and then re-tested in the ratgnawmeter; no significant deviations from the presurgical baseline were observed (not shown). The placement of the pump was continuously monitored throughout the experiment, and drug release was confirmed by monitoring urine color as well as inspection of the pump after explantation.

Data Analysis and Statistics:

The main (primary) outcome of this study was the evaluation of chewing time, which was then correlated with the presence of painful masticatory dysfunction in rats. The group sizes were based on an apriori power analysis using the chewing data from our preliminary experiments. Sample size calculations were made using G*Power 3.1.9.7 software; A two-tailed Wilcoxon-Mann-Whitney means test was employed for each sample size calculation. For the experiments with jaw extension vs anesthesia control (Figure 5), we used a group N1 mean of 501 with a variance of 359, and a group N2 mean of 143 with a variance of 59 yielding an effect size d = 1.4. With power (1-β err prob) = 0.95, α = 0.05, and an allocation ratio of N2/N1 = 0.5, the calculated total sample size was 22 rats with n = 15 for the jaw extension group and n = 7 for the control group. For experiments comparing two groups/conditions [effect size d = 1.3, power (1-β err prob) = 0.95, α = 0.05], the calculated total sample size was also 22 rats. In all the behavior experiments (except the experiment shown in Figure 3A; n = 4), the dependent variable was time in seconds to chew through the resin dowel in the ratgnawmeter. To determine the impact of a pharmacological agent [formalin (Figure 3B; n = 11); CARR (Figure 6A; n = 10); OR-486 (Figure 7; n = 10)], jaw extension alone (Figure 5; n = 15), or jaw extension + treatment [CARR+jaw extension (Figure 6; n = 12); OR-486+jaw extension (Figure 7; n = 12)], one-way repeated measures analysis of variance followed by Bonferroni post hoc testing was used (intra-group analysis). To compare groups submitted to different conditions/treatments with their respective control groups [Figure 3B: Formalin vs Formalin+Dexamethasone (n = 11) or vs control (n = 8); Figure 5: jaw extension vs control (n = 7); Figure 6A: CARR+jaw extension vs CARR-only; Figure 6B: CARR+jaw extension with (n = 10) or without (n = 10) dexamethasone; Figure 7: OR-846+jaw extension vs OR-846-only], two-way repeated measures analysis of variance followed by Bonferroni post hoc testing was used (inter-group analysis). Statistical analysis was performed using GraphPad Prism 8.4.2 (GraphPad Software, Inc., San Diego, CA, USA); Results are presented as mean ± 95% confidence interval, with p < 0.05 considered statistically significant.

Figure 5: Effect of repeated jaw extension on chewing time.

Rats underwent 1h of jaw extension with a 3.5N force daily for 7 consecutive days (black circles). A control group (empty circles) was similarly anesthetized, but no load was applied to the jaw. Ratgnawmeter testing occurred two hours after jaw extension was completed. While chewing times trended higher in the jaw extension group during the loading period (F_{(3.100, 43.40)} = 2.564, p = 0.0652, NS, one-way repeated measures analysis of variance followed by Bonferroni posttest, when the chewing times over time are compared to baseline), comparison with the control group showed no statistical difference (F_{(7, 140)} = 0.6321, p = 0.7287, NS, two-way repeated measures analysis of variance followed by Bonferroni posttest, when both groups are compared). Similarly, no difference between the jaw extension and the control groups was observed in the follow-up period (F_{(1, 20)} = 0.7661, p = 0.3918). These findings indicate that jaw extension alone is not enough to impact TMJ function, neither acutely nor persistently. Jaw extension group, n = 15; Control group, n = 7

Figure 3:

Rats received an injection of formalin (2.5%) into the TMJ and the nociceptive behaviors rubbing (panel A, empty squares) and head flinching (panel A, empty circles) were subsequently evaluated for up to 45 minutes. As previously described, the injection of formalin produces behaviors compatible with the presence of nociceptive stimulation and inflammatory pain. When rats received similar formalin injection and underwent ratgnawmeter testing (panel B, empty triangles), a significant increase in chewing time was observed, with the animals still not completing the ratgnawmeter assay by the 20 min time point after injection, when compared to a group treated with vehicle (empty diamonds; F_{(2, 34)} = 32.84, ****p < 0.0001, two-way repeated measures analysis of variance followed by Bonferroni posttest). To confirm that the increase in time was caused by the formalin-induced inflammatory pain, a separate group of rats received an i.p injection of the anti-inflammatory drug dexamethasone (10 mg/kg) 60 min before formalin (black triangles). The pretreatment with dexamethasone significantly attenuated the effect of formalin, decreasing chewing time (F_{(2, 40)} = 7.671, *p = 0.0015, when formalin-only and dexamethasone + formalin groups are compared). Of note, when the dexamethasone + formalin group was compared to the vehicle-treated group, chewing times were not significantly different (p = 0.0770). Panel A, n = 4; Panel B: formalin group, n = 11; dexamethasone + formalin group, n = 11; control group, n = 8

Figure 6: Local inflammation works as a predisposition factor for jaw extension to induce chewing dysfunction:

(A) Rats received an injection of carrageenan (CARR, 100 μg/25 μl) into the TMJ area. In sequence, they were divided in 2 groups, that were submitted to a 7-day 3.5N jaw loading period (black squares), or a control, CARR only, group (open squares). Both groups were evaluated in the ratgnawmeter daily until the 7^th day, then on days 9 and 13 (follow-up period). Of note, the ratgnawmeter evaluations were performed on the day following the jaw extensions, prior to the next extension. In the control group, CARR injection produced a slight, non-significant increase in chewing time (F_{(1.165, 10.49)} = 3.662, p = 0.0792, one-way repeated measures analysis of variance followed by Bonferroni posttest, when the chewing times over time are compared to baseline), whereas in the CARR + jaw extension group there was a significant increase in chewing time, when compared to baseline time (F_{(3.913, 35.49)} = 8.974, p < 0.0001), that was different from the control group until the 3^rd day post-CARR (F_{(8, 144)} = 6.749, ***p = 0.0009, **p = 0.0020, ##p = 0.0040, *p = 0.0283, two-way repeated measures analysis of variance followed by Bonferroni posttest), and there was no difference between the groups after the loading period has ended (F_{(2, 36)} = 1.667, p = 0.2031, NS); (B) Rats that received CARR injection in the TMJ and were submitted to jaw extension were pretreated with dexamethasone (dark gray bars, 1 mg/kg, s.c.) or vehicle (light gray bars) 30 min previously. We observed that the increase in chewing time produced by jaw extension after CARR injection was significantly attenuated in the group pretreated with dexamethasone (F_{(2, 36)} = 6.347, #p = 0.0217, ###p = 0.0409, two-way repeated measures analysis of variance followed by Bonferroni posttest), suggesting that a robust inflammation such as that produced by CARR potentiates the impact caused by exaggerated opening of the mouth, leading to the TMJ dysfunction. BL: baseline chewing time prior to jaw extension. CARR only group, n = 10; CARR + jaw extension group, n = 12; vehicle group, n = 10; dexamethasone group, n = 10

Figure 7: COMT inhibition potentiates the effect of jaw extension and produces persistent painful TMD.

Rats received the COMT inhibitor OR-486 (15 mg/kg/day) for 16 consecutive days through osmotic pumps implanted subcutaneously two days before starting the 7-day 3.5N force jaw extension protocol (black circles). A control group (open circles) received OR-486 only, without jaw extension. No significant change in chewing time was observed in the control group during the whole experiment (F_{(1.330, 11.97)} = 2.747, p = 0.1171, one-way repeated measures analysis of variance followed by Bonferroni posttest, when the chewing times over the experiment are compared to the baseline). In contrast, the group that received OR-486 + jaw extension showed a marked increase in chewing time (F_{(2.454, 27.00)} = 14.04, ****p < 0.0001, when the chewing times during the 7-day jaw extension protocol are compared to baseline), that persisted after the loading period had ended (F_{(1.206, 13.27)} = 11.64, **p = 0.0032, when the chewing times during the follow-up period are compared to baseline), indicating that decrease in COMT activity contributes to the impairment in TMJ function, and to its persistence. Of note, when both groups are compared, significant difference between the groups is observed both during the jaw extension protocol (F_{(8, 160)} = 5.949, p < 0.0001, two-way repeated measures analysis of variance followed by Bonferroni posttest) and during the follow-up period (F_{(1, 20)} = 12.59, p = 0.0020). OR-486 control group, n = 10; OR-846 + jaw extension, n = 12

Results

Of the 177 initially screened rats, 120 (67.8%) learned to chew through both obstacles. These 120 rats took 340 ± 16s to chew through the resin dowel in their first training session. Chewing behavior closely resembled that of mice, and rats eventually learned to chew in a characteristic “V” pattern as previously described[28]. Average chewing times plateaued on the fourth and fifth training session (approx. 196 ± 10s), and thus the fifth training session was considered the behavioral baseline (Figure 2, panel B). Ratgnawmeter training sessions occurred on five separate, consecutive days between 9:00 A.M. – 3:00 P.M. Previous studies have shown that rats placed in confined spaces for extended periods exhibit anxiety-like behaviors[36], therefore rats that did not complete the assay in 20 minutes were taken out of the apparatus.

Formalin acutely affects chewing time:

To determine if changes in chewing time correlated with the presence of pain, we used the established formalin-induced rat TMJ pain model[29] and compared pain behaviors with chewing times using the ratgnawmeter. Rats received an injection of formalin into the TMJ and acute nociceptive behaviors (flinches and orofacial rubbing) were observed for 45 min, clearly showing the presence of nociceptive behaviors (Figure 3, panel A). A separate group of rats also received TMJ formalin but were evaluated using the ratgnawmeter 20 min and 4h following the injections (Figure 3, panel B). A significant increase in chewing time was observed after formalin injection, when compared to rats that received vehicle (F_{(2, 34)} = 32.84, p < 0.0001), showing a correlation between nociceptive behaviors (observed in panel A) and an increase in chewing time. To verify if this increase in chewing time was mediated by an inflammatory response, a separate group of rats was treated with dexamethasone prior to formalin injection. When compared to the formalin-only group, a marked reduction in chewing time was observed in the group pretreated with dexamethasone (F_{(2, 40)} = 7.671, p = 0.0015). Together, these data show that formalin-induced inflammation affects chewing behavior in the ratgnawmeter and confirmed the correlation between chewing time and observed spontaneous pain behaviors in rats.

Jaw loading and extension:

A 7-day repeated 3.5N jaw loading protocol was performed to evaluate the effect of repeated mechanical stress of the temporomandibular system on chewing behavior using the sling-and-pulley system built to produce prolonged, calibrated jaw extension in rats (Figure 4), as previously described[34][35]. No significant changes in chewing time were produced by the jaw extension, neither during the loading nor the follow-up period when compared to the control group, which was only anesthetized but not submitted to jaw extension (F_{(1, 20)} = 3.249, p = 0.0865). This observation suggests that the stimulus resulting from a mechanical jaw overloading was not strong enough to induce TM pain in SD rats (Figure 5).

Inflammation contribution to TM dysfunction:

To determine the impact of multiple insults on chewing behavior, the jaw extension protocol was coupled with different predispositions to TM disorders (i.e. a two-insult model of pain). Inflammation is known to contribute to the severity of pain originating from joints or muscles[37][38][39][40][41]. We therefore tested the effects of inflammation (induced by 100μg CARR injected into the TM area) on chewing function in the presence and absence of jaw extension (Figure 6, panel A). While the injection of CARR by itself did not produce significant effect in the control group using the ratgnawmeter test (p = 0.0792), there was a significant increase in chewing time when combined with the jaw extension protocol (F_{(8, 144)} = 6.749, p < 0.0001, when both groups are compared). However, the chewing times in the follow-up period were not different between the groups (F_{(2, 36)} = 1.667, p = 0.2031). To confirm that the inflammation enhanced by jaw extension was the factor responsible for the increased chewing time, we treated rats with dexamethasone (1 mg/kg, s.c.) before injecting CARR and performing the jaw extension (Figure 6, panel B). We observed that the group receiving dexamethasone had faster chewing times when compared to the vehicle-treated group (F_{(2, 36)} = 6.347, p = 0.0044). These findings indicate that the presence of inflammation induced by CARR exacerbated the impact of jaw extension on chewing function and suggests that a predisposing component must be involved in the development of a painful condition associated with stress of the TMJ system.

COMT activity plays a role in the effect of jaw hyperextension on chewing time:

Another factor associated with increased susceptibility to painful TMDs is impaired function of the COMT enzyme[42], possibly due to gene polymorphisms[18][19][20][21]. To mimic this condition, we produced a sustained inhibition of COMT by systemic delivery of the small molecule antagonist OR-486 via osmotic minipumps. OR-486 infusions started two days prior to the 7-day jaw extension protocol to assure that the level of COMT inhibition achieved would be robust. A control group received OR-486 alone without jaw extension. We found that treatment with OR-486 by itself did not significantly alter chewing times (F_{(1.330, 1197)} = 2.747, p = 0.1171). However, when associated with jaw extension, a robust increase in chewing time was observed (F_{(2.454, 27.00)} = 14.04, p < 0.0001) that persisted after jaw extension was terminated (F_{(1.206, 13.27)} = 11.64, p = 0.0032). These results indicate that decreased COMT activity by itself is not sufficient to produce changes in the TMJ system but requires a precipitating factor such as jaw extension to produce painful TM dysfunction (Figure 7).

Discussion

In this study, we report that painful behaviors initiated by the injection of formalin into the TMJ correlated with an increase in chewing time in male SD rats, as assessed using a novel device, the ratgnawmeter. Pretreatment with dexamethasone confirmed the correlation between increased chewing time and inflammatory hyperalgesia. We then performed 7-consecutive days jaw extension and found that it failed to induce increase in chewing time. When jaw loading was associated with an additional inflammatory insult using CARR, painful TMD was evident for up to 72 hours. The increase in chewing time was attenuated by pretreatment with dexamethasone. Likewise, persistent TMD was observed once jaw extension was accompanied by COMT inhibition. These results are consistent with clinical findings from patients with TM disorders wherein a combination of predisposing and precipitating factors induce TM pain and dysfunction.

More than 30 TMDs have been identified through the Diagnostic Criteria for Temporomandibular disorders[43][3], loosely organized into 3 main groups; 1) TM joint disorders, 2) masticatory muscle disorders and 3) headaches. Various surgical/mechanical models have endeavored to mimic the different pathologies of the TMJ, including disc displacement, condylectomy, disc perforation, and mechanical perturbation[44]. In contrast, models of masticatory muscle disorders are predominately chemical models, wherein CFA, formalin, or CARR are injected into the masseter muscle to produce inflammation and pain[45]. Less frequently, mechanical trauma such as repeated mouth opening, jaw hyperextension, intraoral appliance placement or ligation of the tendon of the masseter muscle have also been used[46][47][34][48]. However, due to the high heterogenicity of TMDs, mimicking clinical syndromes in animal models has proven challenging. Thus, the goal of the present study was to develop a model of TMD pain with increased face and construct validities.

Face Validity: the ratgnawmeter as an operant method to assess TM dysfunction -

Studies of pain using rodents have relied on reflexive measures such as tail, hind paw, or face retraction from thermal or mechanical stimuli. However, these methods have come under increasing scrutiny and criticism because new therapies that appear to be promising in animal models have failed in human clinical trials[22][49][50]. A proposed solution is to move to operant behavioral assays. The operant assay most commonly used in orofacial pain studies is the OroFacial Pain Assessment Device (OPAD) which provides an automated measurement of hot, cold or mechanical orofacial pain in the distribution of the trigeminal nerve in rats and mice[51][52]. As this device was designed and optimized to test cutaneous neuropathic pain, we found it inadequate in reproducibly assessing deeper muscle and joint pain around the TM area.

A second operant assay used for orofacial pain is the dolognawmeter. This device uses chewing as a surrogate to assess pain from deep in the muscle or the joint in mice. Despite its effectiveness, we found only 3 publications utilizing this device[28][53][54]. It seems possible that the reason for the lack of more widespread use of the dolognawmeter is that it has never been made commercially available. In this study, we introduced modifications to the original device, adapting it for rats by 1) increasing the resin dowel diameter, 2) increasing the device dimensions to fit inside of a standard rat cage, and 3) the addition of an adaptor to fit tubes of differing diameters to accommodate rats of various sizes. Since there is a much greater discrepancy in size of rats based on sex and age compared with mice, greater flexibility in adapting to size was necessary.

To correlate impaired gnawing function to the presence of pain using the ratgnawmeter, we injected formalin into the TMJ, as described by Roveroni et al.[29] and extensively used by others[55][56][57][58], and compared the onset of nociceptive behaviors with changes in chewing time. We observed that the painful behaviors initiated by injection of formalin into the TMJ correlated with an increase in chewing time. Since pretreatment with dexamethasone decreased the time for the rats to chew through the resin dowels, we concluded that the presence of inflammatory pain played a role in the gnawing dysfunction, confirming a correlation between chewing time and the presence of inflammatory mechanical hyperalgesia reported previously[28]. Of note, in contrast with the two-phase effect produced by formalin injection in the paw[59][32][60] or upper lip[61], injection into the TMJ induces only one phase of nociceptive behaviors[29]. Thus, the effect of dexamethasone, reported to affect only the second phase of the formalin test[32], suggests that, in this model, the nociceptive and inflammatory phases overlap, indicating the impact of the inflammatory component of the TMJ injection of formalin on chewing activity.

The greatest limitation of the ratgnawmeter is the relatively low percentage of rats that were successfully trained to chew through the dowel (67.8%). The motivation to chew appears to be a desire to escape the narrow tube[62]. We attempted to entice rats that would not spontaneously chew by fasting them, then providing a food reward both on the dowel and outside the tube to try to motivate and train them to complete the chewing task. These efforts were unsuccessful (not shown). There is the possibility that inherent motivational differences exist between the rats that chewed and those that did not which could bias the final results. Another limitation is that some rats in extreme pain failed to chew, despite passing the initial training. In all cases when a painful stimulus was stopped, or when the pain was treated, these rats would again chew through the dowel. Other rats in pain made chewing efforts but were unable to chew completely through the resin dowel. We therefore stopped the assay at 1200s to minimize the distress of the rats being confined in the tube. Time cut-offs are utilized in other pain assays, such as the Hargreaves thermal paw withdrawal method, which limits the time a radiant heat source is applied to the hind paw to minimize the chances of producing thermal burns[63]. A time limit was also implemented in the original dolognawmeter experiments[28].

Construct Validity: Two-insult mechanism of TM function impairment -

The etiology of TM disorders is not one-dimensional, and patients will often have precipitating events such as trauma from TM joint/muscle/ligament hyperextension (e.g. endotracheal intubations, laryngoscopy, routine dental procedures, yawning)[64][9][11][65] occurring in parallel with predisposing factors (e.g. bio-psychosocial stress, impaired sleep quality, genotypic and biological variations, among others[66][67][68][69][70] that increase the risk for TM pain and exacerbate symptoms. We hypothesized that the combination of two different insults is needed to trigger painful TMDs (as proposed in Figure 8). One contributory factor of jaw muscle pain is elevated inflammatory cytokine levels[71]. To mimic this inflammation, we administered CARR in the TM area. To our surprise, when CARR was administered alone, there was no indication that it induced chewing dysfunction. However, when associated with jaw extension, robust increases in chewing time were observed. This is in line with the study from Oliveira et al. in which inflammation alone did not produce nociceptive behaviors but required the additional injection of histamine[31]. CARR-induced nociceptor sensitization lasts around 3-4 days, due to the ongoing presence of inflammation[72]. In our experiment, we observed that the impact of jaw extension and CARR was stronger during the first three days of the procedure, and gradually returned to baseline values, indicating that the presence of active inflammation is necessary to produce an impact on chewing function. This was confirmed by the attenuating effect of dexamethasone on the dysfunction produced by CARR and jaw extension. These results indicate that CARR caused latent sensitization of the nociceptors in the TMJ prior to jaw loading and provide convincing evidence that a subclinical, asymptomatic inflammation acts as a predisposing agent to painful TMD produced by exaggerated jaw opening.

Figure 8: Proposed Two-hit model hypothesis.

Significant, persistent chewing dysfunction only develops when a precipitating factor occurs in the context of a predisposing factor. (1) Rats trained in the ratgnawmeter (2) received a predisposing insult, i.e., injection of CARR or treatment with a COMT inhibitor. (3) When subjected to jaw extension (the precipitating factor) (4) chewing activity was impaired due to the presence of pain during function.

Genetic variations in catecholamine metabolizing enzymes have also been reported to enhance pain sensitivity and increase the risk for functional pain syndromes such as fibromyalgia and TMDs[18][73][74]. In fact, previous preclinical models show sustained mechanical nociceptive hypersensitivity following COMT inhibition[30]. However, sustained COMT inhibition alone was not enough to induce painful masticatory dysfunction in our model. In line with our two-insult hypothesis, we found that rats that coupled sustained COMT deficiency using the inhibitor OR-486 with jaw extension experienced painful TMD on all 7 days of the acute loading phase. More significantly, chewing times remained elevated after the cessation of the jaw extension protocol (on day 6), providing evidence for the role of COMT deficiency in the transition to persistent masticatory dysfunction.

A key reason that the additive effects of predisposing and precipitating factors could be detected was due to the use of the ratgnawmeter. The jaw extension method was based on the protocol described by Nicoll et al.[34]. While they found that daily jaw extension for 7 consecutive days with 3.5N jaw loading alone produced persistent pain as assessed by von Frey fibers[35], we found that this same stimulus was insufficient to produce painful chewing dysfunction. Sprague Dawley rats have no known predisposition for developing chronic pain. Likewise, most patients that go to the dentist or are intubated also do not report the development of persistent TM pain. Therefore, while von Frey fibers are very sensitive in identifying changes in skin sensitivity, these differences may not translate into clinically significant pain and TMD. Thus, using the ratgnawmeter as a pain endpoint may provide increased face, construct and predictive validities to preclinical orofacial pain studies.

Conclusion

Here we introduce the ratgnawmeter, used to determine chewing function in the rat as endpoint surrogate for pain, and present a protocol combining precipitating and predisposing factors to show that their interaction impacts the susceptibility to develop TM disorders. Our findings indicate that the predisposition factor is necessary for an insult to the TM system to trigger painful dysfunction. The need for a combined contribution of these factors might explain why not all patients submitted to traumatic events such as exaggerated mouth opening develop TMDs.

Highlights:

• We present a model with face/construct validities to evaluate temporomandibular pain

• The ratgnawmeter determines temporomandibular pain based in chewing time and efficacy

• A predisposing factor allows the expression of pain in temporomandibular disorders

• Subclinical inflammation contributes to pain caused by exaggerated mouth opening

• Susceptibility to develop chronic temporomandibular pain after trauma depends on COMT

Abbreviations:

TM

temporomandibular

TMJ

temporomandibular joint

TMD

temporomandibular dysfunction

COMT

catechol-O-methyltransferase

CARR

carrageenan

CFA

Complete Freud’s Adjuvant