INTRODUCTION

TOP ABSTRACT INTRODUCTION Methods RESULTS DISCUSSION REFERENCES

One of the most common pains in the body is acute toothache resulting from injury or inflammation of the tooth pulp (Byers 1992; Narhi et al. 1994; Sessle 1987, 2000; Sharav 1994). However, little is known about the central neural consequences of pulpal inflammation. The tooth pulp is richly innervated by small-diameter nerve fibers, and we have recently demonstrated that application of the small-fiber excitant and inflammatory irritant mustard oil (MO) to the maxillary molar tooth pulp produces an acute pulpal inflammation and reflexly induces significant increases in jaw muscle electromyographic (EMG) activity as well as profound neuroplastic changes [i.e., increases in neuronal mechanoreceptive field (RF) and response properties] in brain stem nociceptive neurons of trigeminal subnucleus caudalis (Vc) (Chiang et al. 1998; Sunakawa et al. 1999). Furthermore, intrathecal application of MK-801, a noncompetitive NMDA receptor antagonist, can block these neuroplastic changes occurring in Vc nociceptive neurons and the associated increased EMG activity (Chiang et al. 1998; Hu et al. 1997; Sunakawa et al. 1999; Yu et al. 1993, 1995, 1996).

While the Vc has been generally considered to serve as an essential brain stem relay of orofacial nociceptive information (Dubner et al. 1978; Sessle 1987, 2000), several clinical, behavioral, and electrophysiological findings also implicate the rostral components of the V spinal tract nucleus, including subnucleus oralis (Vo), in orofacial pain mechanisms and in particular tooth pulp pain (Azerad et al. 1982; Dallel et al. 1988, 1989, 1990; Greenwood and Sessle 1976; Raboisson et al. 1989, 1995; Young and Perryman 1984; for review see Sessle 2000). In support of these findings, anatomical evidence has shown that primary afferents from the tooth pulp project to all subdivisions of the ipsilateral V brain stem nuclear complex, and indeed Vo appears to receive a heavier pulp afferent projection than Vc (Arvidsson and Gobel 1981; Marfurt and Turner 1984; Shigenaga et al. 1986; Takemura et al. 1991). Therefore the present study was carried out to test if MO application to the rat molar pulp also induces neuroplastic changes in Vo neurons and if intranuclear application to Vo of MK-801 influences the MO-induced neuroplastic changes. The data have been presented in abstract form (Park et al. 1998a,b).

	Methods

TOP ABSTRACT INTRODUCTION Methods RESULTS DISCUSSION REFERENCES

Animal preparation

This study was carried out in 71 Sprague-Dawley adult male rats weighing 250-400 g. The methods used for animal preparation, stimulation, and neuronal recording and classification were similar to those described previously in detail (Chiang et al. 1998; Hu 1990; Kwan et al. 1996) and so will only be briefly outlined. The animal was anesthetized by an injection (ip) of a mixture of alpha-chloralose (50 mg/kg) and urethan (1 g/kg). The tracheal cannula was inserted and the left external jugular vein cannulated. To expose the pulp of the right maxillary first molar, the cavity was prepared with a low-speed dental drill and the cavity was temporarily filled with a small piece of cotton pellet soaked with physiological saline. The animal was placed in a stereotaxic apparatus, and a craniotomy was performed to expose the cerebellum. One hour after surgery, a supplemental dose of urethan (200-300 mg/kg, iv) was administered just before the recording session. The animal was immobilized with gallamine triethiodide (initial dose, 35 mg/kg; maintenance dose, 14 mg/h; iv) and then artificially ventilated throughout the experimental period. An adequate level of anesthesia was confirmed periodically by the lack of spontaneous movements and of responses to pinching the paw when the gallamine-induced muscle paralysis was allowed to wear off. In addition, pupil size and heart rate were routinely monitored to ensure their stability when noxious pinch stimuli were applied. The expired %CO₂ and rectal temperature were also continuously monitored and maintained at physiological levels of 3.5-4.5 and 37-37.5°C, respectively.

Recording and stimulation procedure

Single neuronal activity was recorded extracellularly by means of an epoxy resin-coated tungsten microelectrode, which was held by a microdrive with a caudal inclination of 23.5°. As the microelectrode was advanced through the cerebellum into the brain stem, natural stimuli (see next paragraph) were applied to the orofacial tissues to search for brain stem neurons receiving an orofacial sensory input. The brain stem was explored 2.4-3.0 mm lateral to the midline and between frontal planes P2.0 and P2.6 referred to interaural line (Paxinos and Watson 1986). Neuronal activity was amplified, displayed on oscilloscopes, and also led to a window discriminator connected to a A/D converter (CED 1401plus, CED, UK) and a personal computer.

A wide range of mechanical (brush, pressure, and pinch), electrical, and noxious thermal (radiant heat, 51-53°C) stimuli were applied to orofacial skin or intraoral mucosa to classify (Chiang et al. 1998; Hu 1990; Kwan et al. 1996) each neuron into low-threshold mechanoreceptive (LTM), wide dynamic range (WDR), or nociceptive-specific (NS). A neuron was classified as a deep nociceptive (D) neuron if its RF involved deep tissues solely. A small number of neurons with WDR and NS neuronal properties (based on their cutaneous RF features) also had a deep RF, and so in accordance with earlier findings (Chiang et al. 1994; Yu and Mense 1990; Yu et al. 1993), these neurons were termed skin/mucosa and deep (S + D) neurons. Electrical stimuli of constant-current single pulses (0.2 ms and <1 mA for A-fiber inputs; 2 ms and <5 mA for C-fiber inputs) were applied within the delineated RF at tooth pulp to determine the existence of A or C fiber input. A discharge consistently evoked at a latency >30 ms was attributed to C-fiber inputs on the basis of a 40-50 mm conduction path and allowances of 1 ms for peripheral activation time, central narrowing of the afferents in the V spinal tract, and synaptic delay.

The spontaneous activity, RF size, mechanical threshold, and responses to suprathreshold mechanical stimuli were assessed at the time intervals specified below (see Experimental paradigm section). Spontaneous activity was measured as the total number of spikes occurring for 2 min. As mentioned in our previous studies (Chiang et al. 1997, 1998), the RF of each neuron was determined through the use of a brush, blunt probe, and a pair of forceps. The extent of the cutaneous RF was manually defined by brushing (in the case of a tactile RF) or pinching (in the case of a pinch RF) the RF; as previously described (Chiang et al. 1994), the noxious stimulation was used sparingly so as to avoid damage to the skin and neuronal sensitization. A burst response consisting of at least two spikes during each stimulus trial (touch or pinch) was accepted as the criterion for the RF boundary of the neuron tested. A neuron was considered to have a deep RF when the neuron responded to stimulation applied by a blunt probe to muscle, bone, tendon, or temporomandibular joint (TMJ) and had a mechanical threshold above 5 g but no response could be evoked by the wide range of cutaneous stimuli used (Iggo 1960; Schaible and Schmidt 1983; Yu et al. 1993). Then the extent of the RF was outlined on a life-size drawing of the rat's head and the area of the RF was measured by a computer-aided device (SigmaScan, Jandel, CA). For intraoral RF measurement, the intraoral region was divided arbitrarily into 11 parts (e.g., upper and lower lips, upper and lower buccal mucosae, maxillary incisor and molar teeth including gingiva, mandibular incisor and molar teeth including gingiva, anterior and posterior part of hard palate, and tongue), and the extent of the intraoral RF was assessed in terms of the number of parts from which the response was evoked by light touch or firm pressure applied with a dental explorer. The mechanical threshold was determined at the most sensitive part of the RF with the use of a set of von Frey nylon monofilaments (0.007-92 g) applied to the RF. The threshold was defined as the monofilament with the lowest value that elicited 1-2 spikes/trial in at least four of six trials. Responses to graded suprathreshold mechanical stimuli applied to the RF were quantified as the number of spikes produced per stimulus. For LTM neurons, the responses to suprathreshold mechanical stimuli (0.04-0.32 g, 2 s, 6 trials) were determined with von Frey nylon monofilaments. Responses of nociceptive (NS, S + D, and D) neurons to graded pinch stimuli (50, 100, and 200 g applied in ascending order, each for 3 s at 15 s intervals) were assessed by the use of a modified forceps with an attached strain gauge that monitored force levels up to 600 g/mm² or by graded pressure stimuli applied to the RF by von Frey monofilaments. Responses of WDR neurons to graded pinch or pressure stimuli were evaluated in the same manner but with less intense stimuli (10, 20, and 40 g). A stimulus that produced a suprathreshold response was selected from these graded stimuli as the one to be repeated every 10-20 min before and after MO or mineral oil (MIN) application, at the time intervals specified below (see Experimental paradigm section).

To establish an acute chemical irritant-induced pulpal inflammation, MO (allyl isothiocyanate, 95%; Aldrich Chemical Co.) was applied locally to the exposed molar pulp. A similar application of MIN served as a control. To test for possible NMDA mechanisms, a freshly prepared solution of (+)-MK-801 (10 µg/µl in saline; RBI), a noncompetitive blocker of the NMDA receptor-ion channel, was intranuclearly applied as a pretreatment into the right Vo. A similar application of saline served as a control.

Experimental paradigm

In each animal, only one neuron was tested with MK-801 or saline application to Vo. For this purpose, each animal received pretreatment with intranuclear microinjection of saline or MK-801 into the Vo followed by the pulpal application of MO or MIN. Experimental animals were divided into four groups according to the drug application: pretreatment of saline followed by MO application to the pulp (SAL/MO); pretreatment of saline followed by MIN application to the pulp (SAL/MIN); pretreatment of MK-801 followed by MO application to the pulp (MK-801/MO); and pretreatment of MK-801 followed by MIN application to the pulp (MK-801/MIN).

Before starting neuronal recording, a fine #7000 Hamilton syringe (1 µl) was filled with MK-801 or saline. This syringe was held by a microinjector (model 5000, David Kopf) and placed at a rostral inclination of 15° into the rostromedial part of the right Vo (L: 2.6; A-P: 2.0 referred to interaural line; V: 8.0-8.5 from the cortex) (Paxinos and Watson 1986). Then, the recording microelectrode was lowered at a caudal inclination of 23.5° into the right Vo. At the end of the experiment, the distance in Vo between the tip of the injection needle and the recording microelectrode, which ranged between 0.3 and 1.2 mm, was measured after removing the rat from the stereotaxic apparatus and resetting the injection needle and recording microelectrode to the original position.

A similar experimental paradigm was applied for all four groups. After a neuron was identified, the extent of the RF and neuronal response properties including spontaneous activity, mechanical threshold, and responses to suprathreshold mechanical stimuli were first determined and served as baseline values. A complete determination of the neuronal RF and response properties took around 7 min (2 min for spontaneous activity, 1 min for threshold, 1-2 min for tactile and/or pinch or pressure RF size, and 1-2 min for responses to suprathreshold stimulation). A bolus (0.3 µl) of either saline or MK-801 was then slowly microinjected into the right Vo over a period of 6-8 min. The selected dose of MK-801 (3 µg/0.3 µl) was comparable to that shown to be effective in our previous study in Vc (Chiang et al. 1998) and also in accordance with that shown to be effective in other recent studies (Coderre and Van Empel 1994; Dunbar and Pulai 1998; Dunbar and Yaksh 1997; Fairbanks and Wilcox 1997). Three minutes after the termination of drug or saline injection, the determination of the spontaneous activity, RF, and neuronal response properties were repeated. Then, the saline-soaked cotton pellet in the prepared cavity of the maxillary first molar was carefully replaced by a segment (0.5 mm) of dental paper point (X-fine, Absorbent Point, Kerr Co., MI) soaked with either MO or MIN (0.2 µl), and the cavity was sealed with CAVIT (ESPE, Germany). To ensure that the amount of MO or MIN solution was standardized, the 0.5-mm segment was cut off from the tip of the paper point and then placed on a concave glass surface and soaked with a 0.2 µl MO or MIN solution delivered by a 1-µl Hamilton syringe. This freshly prepared soaked paper point was immediately used for pulpal application of the solution. After the MO or MIN application, the spontaneous activity, RF, and response properties were determined at 5-10 min intervals over a 40-60 min period, except for quantified pinch-evoked responses which were determined every 10-20 min.

Histological and statistical analysis

Recording sites were marked by an electrolytic lesion (anodal current 8 µA, 10 s) at the end of each experiment and verified with conventional histological procedures. The recording sites were represented on schematic drawings of the right Vo (Fig. 1). Differences between baseline (predrug) values and values at different postdrug time points in each group were treated a priori by a Wilcoxon signed rank test, and differences between groups were treated also a priori by a two-way analysis of variance (ANOVA) (Denenberg 1984). A nonparametric ANOVA on ranks was used for comparison of neuronal threshold data between groups. All the data except spontaneous activity and threshold were normalized because different baseline values were observed between groups. Values were expressed as mean ± SE, and those for threshold as median with 25 and 75%. The level of significance was set at P < 0.05. 

View larger version (26K): [in this window] [in a new window]   Fig. 1. Two diagrammatic transections at the level of the rostral medulla, 1.8 and 2.6 mm behind the interaural line, that show the location of histologically confirmed recording sites within the trigeminal subnucleus oralis (Vo). A: the nociceptive neurons of the saline followed by MO application to the pulp (SAL/MO) group are represented by filled symbols, while those of the other three [pretreatment of saline followed by MIN application to the pulp (SAL/MIN), pretreatment of MK-801 followed by MO application to the pulp (MK-801/MO), and pretreatment of MK-801 followed by MIN application to the pulp (MK-801/MIN)] groups are represented by open symbols. Circle, nociceptive-specific (NS) neuron; triangle, wide dynamic range (WDR) neuron; square, skin/mucosa and deep (S + D) neuron; double border square, deep (D) neuron. B: the nociceptive neurons with only an intraoral mechanoreceptive field (RF) are represented by diamonds, the low-threshold mechanoreceptive (LTM) neurons by cross symbols. 7: facial nucleus; DM: dorsomedial part of Vo.

	RESULTS

TOP ABSTRACT INTRODUCTION Methods RESULTS DISCUSSION REFERENCES

A total of 83 single neurons responding to orofacial stimulation were recorded in the right Vo. These neurons were 22 nonnociceptive (LTM) neurons and 61 nociceptive (26 WDR, 25 NS, 2 D, 8 S + D) neurons, and all of them had an ipsilateral orofacial RF. Due to the technical difficulties of maintaining a stable single unit recording for more than 2 h required to identify and examine drug effects, only half of the neurons were studied with the full range of tests. The locations of neurons tested in the right Vo are shown in Fig. 1. The majority of the nociceptive neurons were found in the dorsal part of Vo.

General properties

NONNOCICEPTIVE NEURONS. As shown in Table 1, only 13% of the LTM neurons showed a spontaneous activity >0.1 Hz when they were initially encountered. Most of LTM neurons had a RF that was localized within one V division and involved the intraoral or perioral region, whereas very few LTM neurons had a RF involving the ophthalmic division. All the LTM neurons tested (n = 14) received electrically evoked A-fiber inputs from their RF, but no responses attributable to electrically evoked C-fiber inputs from their RF were observed. Other features of the LTM neurons are shown in Table 1.

NOCICEPTIVE NEURONS. As shown in Table 1, only 9% of the NS neurons showed spontaneous activity >0.1 Hz when they were initially encountered, whereas 63% of the WDR neurons and 50% of the D and S + D neurons had spontaneous activity >0.1 Hz. All of the WDR and NS neurons had an RF that involved the maxillary and/or mandibular divisions and the intraoral and/or perioral regions. Of note is the finding that 12% of the WDR neurons and 17% of the NS neurons had only an intraoral RF; these seven neurons were termed intraoral nociceptive neurons. None of the WDR and NS neurons had a RF involving the ophthalmic division. Two D neurons had only a deep RF which involved the TMJ; the eight S + D neurons had a RF involving the TMJ and/or jaw muscle in combination with a perioral and/or intraoral RF that was usually in the mandibular division and included the tongue. The majority of the WDR and NS neurons were excited by electrically evoked A-fiber inputs only, but 38% of the WDR and 29% of the NS neurons received C-fiber as well as A-fiber electrically evoked inputs from their RF. In contrast, only 25% of the S + D neurons tested received electrically evoked A-fiber, but not C-fiber, inputs from deep tissues or their perioral or intraoral RF; however 33% of the D and S + D neurons were excited by electrically evoked A- or C-fiber inputs from the pulp. A similar proportion (42-48%) of the WDR and NS neurons responded to noxious heat applied to the RF. The mechanical activation threshold of the WDR neurons was significantly lower than that of the NS, D, and S + D neurons (P < 0.05, Dunn's method), but significantly higher than that of LTM neurons (P < 0.05, Dunn's method; see Table 1).

Pulpal application of MO induces neuroplastic changes in oralis nociceptive neurons

In the group of rats pretreated with saline microinjection into the Vo followed by the pulpal application of MO (SAL/MO), significant MO-induced neuroplastic changes occurred in all tested nociceptive neurons (3 WDR, 2 NS, 3 S + D) having their RF involving cutaneous and/or deep tissues: one WDR and two NS neurons had both perioral and intraoral RFs, one WDR neuron had only a perioral RF, and another WDR neuron had a facial cutaneous RF and a masseter muscle RF; the remaining three S + D neurons had TMJ and intraoral RFs. The neuroplastic changes included increases in spontaneous activity, RF size, and responses to suprathreshold mechanical stimuli, and a decrease in mechanical threshold (see Figs. 3-5, Table 2). An example is shown in Fig. 2.

View larger version (24K): [in this window] [in a new window]   Fig. 2. An example of a Vo NS neuron showing changes in RF and response properties following mustard oil (MO) application to the right maxillary molar pulp. A: neuronal responses to mechanical stimuli applied to the cutaneous lower lip RF. Top: marker of Br (brush), Pr (pressure), and Pi (pinch). Each stimulus lasted for 3 s. Middle: neuronal responses in control conditions (i.e., Pre-MO, before MO application). Bottom: neuronal responses to same stimuli 40 min after MO application (i.e., Post-MO, after MO application). B: neuronal responses to 50 g mechanical pinch stimuli. Each stimulus lasted for 3 s. Binwidth is 0.5 s in both A and B. C: neuronal RF mechanical threshold. Note that this was tested with a von Frey monofilament at the lower incisor gum rather than at the cutaneous lower lip, because the soft consistency of the lip was not conducive for reproducible testing. A moderate decrease in mechanical threshold occurred at 40 min. The value at 0 min represents the baseline level before MO application; arrow represents the time of MO application to the molar pulp. D: change of spontaneous firing rate. The value at 0 min represents the baseline level before MO application (arrow). E: MO induced an expansion of cutaneous pinch RF. The value at 0 min represents the baseline level before MO application (arrow). Insets illustrate histologically retrieved recording site and cutaneous pinch RF sizes before (0 min) and 20 min after MO application.

However, the seven intraoral nociceptive neurons (4 NS, 3 WDR) having their RF only in the intraoral region showed no significant changes in RF size, spontaneous activity, mechanical threshold, and responses to suprathreshold mechanical stimuli; the recording sites of these neurons were all located within the rostral dorsomedial portion of Vo (see Fig. 1). Therefore these nociceptive neurons were not included in further comparisons between the SAL/MO group and the other three groups.

INCREASE IN SPONTANEOUS ACTIVITY. Of the eight nociceptive neurons (3 WDR, 2 NS, 3 S + D) tested, pulpal application of MO produced an immediate and dramatic increase in firing rate in only one neuron (a S + D neuron); its activity peaked at 3,468 spikes/2 min at 5 min and gradually settled down to a level of 800 spikes/2 min of the next 55 min. Of the remaining neurons, 2 WDR neurons showed an increase in spontaneous activity above baseline (41 and 821 spikes/2 min) that had a slow onset (>1 min), peaked around 5 min (2,136 and 1,908 spikes/2 min), and lasted for 40-60 min. The remaining neurons (1 WDR, 2 NS, 2 S + D) had little or no spontaneous activity and showed no clear change in spontaneous activity after MO application to the pulp. Overall, the mean spontaneous firing rate of this SAL/MO group (8 neurons) increased significantly (Wilcoxon signed rank test, P < 0.05) at 5 min after MO application and then declined to an insignificant moderate level of tonic firing (Fig. 3, Table 2). In comparison, four of the seven intraoral nociceptive neurons had low baseline spontaneous activity (1 spike/2 min) and the remaining three had no spontaneous activity, and there was no significant change in the mean spontaneous firing rate of these seven neurons following MO application (baseline value versus peak value, 0.43 ± 0.2 spikes/2 min versus 0.86 ± 0.55 spikes/ 2 min).

View larger version (25K): [in this window] [in a new window]   Fig. 3. Time course of MO-induced changes in mean spontaneous activity of Vo nociceptive neurons in four groups of rats. The first group (filled circle; n = 8) received saline (0.3 µl) pretreatment followed by pulpal application of MO (SAL/MO group). The second group (open circle; n = 7) received saline (0.3 µl) pretreatment followed by pulpal application of mineral oil (SAL/MIN group). The third group (filled square; n = 7) received MK-801 (3 µg/0.3 µl) pretreatment followed by pulpal application of MO (MK-801/MO group). The fourth group (open square; n = 6) received MK-801 (3 µg/0.3 µl) pretreatment followed by pulpal application of mineral oil (MK-801/MIN group). All drug or saline pretreatments were performed by local injection into Vo nearby the neuronal recording sites. Arrows represent the saline or drug injection time and also the time for pulpal application of MO or mineral oil. Note that differences between baseline (predrug) values at 8 min time point and values at different postdrug time points in each group are treated by a Wilcoxon signed rank test (*P < 0.05). Significant differences in time-course values between the SAL/MO group and the other three groups are treated by two-way analysis of variance (ANOVA) and indicated in Table 2. This legend and all symbols apply also to following figures.

INCREASE IN RF SIZE. Pulpal application of MO produced a significant and long-lasting increase in size of the facial cutaneous and/or deep RFs in all eight nociceptive neurons (3 WDR, 2 NS, 3 S + D) tested. As shown in Fig. 4 and Table 2, the pinch or deep RF size increased significantly throughout the 40-min period following MO application, with its peak occurring around 30 min (Wilcoxon signed rank test, P < 0.05). In the three WDR neurons tested, the mean tactile RF size also increased to 121 ± 18 and 120 ± 13% at 5 and 10 min, respectively; however, these changes were not significant. Unlike NS neurons tested in Vc (Chiang et al. 1998), a novel area responsive to both tactile and pinch stimuli did not appear within the previous pinch RF in the two NS neurons and one S + D neuron following pulpal application of MO, although in one NS neuron having an intraoral RF, a novel deep RF in the TMJ region appeared after MO application. The other two S + D neurons had intraoral tactile and pressure RFs, both of which increased after MO application.

View larger version (19K): [in this window] [in a new window]   Fig. 4. Time course of MO-induced changes in mean cutaneous pinch or deep RF size of Vo nociceptive neurons. Note that the pinch or deep RF sizes of the SAL/MO group (n = 8) increased throughout 40-min observation period following MO application, while those of the other three groups (MK-801/MO group, n = 7; SAL/MIN group, n = 7; MK-801/MIN group, n = 6) remained unchanged. Differences between baseline (predrug at 8 min time point) values and values at different postdrug time points in each group are treated by a Wilcoxon signed rank test (*P < 0.05). Significant differences in time-course values between these four groups are treated by two-way ANOVA and indicated in Table 2.

DECREASE IN MECHANICAL THRESHOLD. Pulpal application of MO produced a marked decrease in mechanical threshold of the seven nociceptive neurons (2 WDR, 2 NS, 3 S + D) tested. The two WDR neurons, which had a baseline cutaneous threshold of 0.09 and 0.32 g, showed a decrease in their thresholds to 0.04 and 0.17 g, respectively, for 20-40 min after MO application. The cutaneous threshold of the two NS neurons was 1.43 and 2.27 g at baseline and decreased to 0.59 and 1.43 g, respectively, throughout the 40- to 60-min period after MO application. The three S + D neurons all had a high baseline mechanical threshold (12.75, 53.2, 53.2 g), but their threshold decreased to 7.01, 19.5, and 19.5 g, respectively, for 20-40 min after MO application. The median mechanical threshold of the two NS and three S + D neurons was 12.8 g and significantly decreased to 7.1 g at 40 min after MO application (Wilcoxon signed rank test, P < 0.05; Table 2).

INCREASE IN RESPONSE TO MECHANICAL STIMULI. After pulpal application of MO, neuronal responses to suprathreshold mechanical stimuli were increased in all eight nociceptive neurons (3 WDR, 2 NS, 3 S + D) tested at 20, 40, and 60 min. The mean response to suprathreshold mechanical stimuli was gradually increased and reached a peak at 40 min that was significantly different from baseline (192 ± 27% of control; Wilcoxon signed rank test, P < 0.05) (Fig. 5, Table 2). In the seven intraoral nociceptive neurons tested, the response to suprathreshold mechanical stimuli showed no significant changes following MO application (115 ± 20 and 137 ± 28% of control at 20 and 40 min, respectively, P > 0.05).

View larger version (19K): [in this window] [in a new window]   Fig. 5. Time course of MO-induced changes in mean responses to 50 g stimuli applied to RF of Vo nociceptive neurons. Note that mean responses gradually but significantly increased by MO application in the SAL/MO group (n = 8), while those in the other three groups remained unchanged (n = 6-7). Note that differences between baseline (predrug) values and values at different postdrug time points in each group are treated by a Wilcoxon signed rank test (*P < 0.05). Significant differences in time-course values between these four groups are treated by two-way ANOVA and indicated in Table 2.

Pulpal application of MO does not induce any significant changes in oralis nonnociceptive (LTM) neurons

None of the LTM neurons tested showed an immediate increase in firing rate following pulpal application of MO. In six LTM neurons tested, the mean baseline and peak spontaneous activity were 0.8 ± 0.5 and 13.3 ± 12.2 spikes/2 min, respectively, and the mean peak values of tactile RF size and mean response to suprathreshold mechanical stimuli were 102.5 ± 1.6 and 89.3 ± 15.4% of control, respectively. None of these changes were significant. The median mechanical threshold remained unchanged after MO application.

Pulpal application of mineral oil does not induce any significant neuroplastic changes in oralis nociceptive neurons

In the group of rats pretreated with saline microinjection into the Vo followed by the pulpal application of mineral oil (SAL/MIN), there were no significant changes in spontaneous activity, pinch or deep RF size, mechanical threshold, or responses to suprathreshold mechanical stimuli in any of the seven nociceptive neurons (2 WDR, 3 NS, 2 S + D) with a RF involving facial cutaneous, intraoral, or deep tissues. Except for their responses to suprathreshold mechanical stimuli and mechanical threshold, the spontaneous activity and the size of pinch or deep RF of this group were significantly different from those of the SAL/MO group (two-way ANOVA, P < 0.001-0.002; Figs. 3-5, Table 2).

Microinjection of MK-801 into the Vo antagonizes MO-induced neuroplastic changes in nociceptive neurons

In the group of rats pretreated with MK-801 (3 µg/0.3 µl intranuclearly injected into Vo) followed by the pulpal application of MO (MK-801/MO), no significant MO-induced neuroplastic changes occurred in the seven nociceptive neurons (3 WDR, 1 NS, 1 D, and 2 S + D) with a RF involving facial cutaneous, intraoral, or deep tissues.

Following pretreatment with MK-801, pulpal application of MO did not produce any significant increases in mean pinch or deep RF size of the seven nociceptive neurons tested (Fig. 4, Table 2). The difference between the MK-801/MO group and the SAL/MO group was significant (two-way ANOVA, P < 0.001; Table 2). The tactile RF size also did not increase in the three WDR neurons tested.

MO application following pretreatment with MK-801 also failed to induce any significant change in mean spontaneous activity throughout the 60-min observation period, although pulpal application of MO did produce an immediate increase in firing rate in two of the seven nociceptive neurons tested (Fig. 3, Table 2). The mechanical threshold and mean response to suprathreshold mechanical stimuli also did not change significantly after MK-801 pretreatment followed by MO application (Fig. 5, Table 2). Differences in spontaneous activity and responses to suprathreshold mechanical stimuli between the MK-801/MO group and the SAL/MO group were all significant (two-way ANOVA, P < 0.001; Table 2).

Microinjection of MK-801 into the Vo does not induce any significant changes in oralis nociceptive neurons

In the group of rats pretreated with MK-801 (3 µg/0.3 µl) microinjected into the Vo followed by pulpal application of mineral oil (MK-801/MIN), all six nociceptive (3 WDR, 3 NS) neurons with a RF involving cutaneous, intraoral, or deep tissues showed no significant changes in spontaneous activity, pinch RF size, mechanical threshold, and responses to suprathreshold mechanical stimuli (Figs. 3-5). There were no significant differences between this group and the MK-801/MO group (two-way ANOVA, P > 0.05; Table 2).

	DISCUSSION

TOP ABSTRACT INTRODUCTION Methods RESULTS DISCUSSION REFERENCES

Application of the small-fiber excitant and inflammatory irritant MO to the rat molar pulp produced significant neuroplastic changes in Vo nociceptive neurons, including increases in spontaneous activity, RF size, and responses to suprathreshold mechanical stimuli. Intranuclear microinjection into Vo of MK-801, a noncompetitive blocker of NMDA receptor-ion channels, reduced the MO-induced neuroplastic changes of Vo nociceptive neurons. These findings indicate that the application of MO to the tooth pulp can induce significant neuroplastic changes in Vo nociceptive neurons and that central NMDA receptor mechanisms may be involved in these neuroplastic changes.

Of the 10 neurons activated from deep tissues, two D neurons had only a deep RF; this low incidence of D neurons in the present study is consistent with earlier electrophysiological findings in Vo (Davis and Dostrovsky 1988; Sessle and Greenwood 1976). The other eight neurons were S + D neurons that had a deep RF in combination with a perioral and/or intraoral RF mostly involving the mandibular division (including the tongue); only three of these S + D neurons were tested by MO application, which produced similar neuroplastic changes in these neurons as in the WDR and NS neurons. Consistent with the location of TMJ afferent projections to Vo (Capra 1987), six of the histologically retrieved eight neurons (1 D, 7 S + D) in our study were located at the caudal and dorsolateral part of Vo (see Fig. 1).

The Vo has traditionally been considered to be involved in transmitting orofacial tactile information (Dubner et al. 1978; Sessle 1987). However, several studies point to the involvement of rostral components of the V spinal tract nucleus, including Vo, in orofacial pain mechanisms (see Sessle 2000). Lesioning of Vc or V tractotomy at obex level does not completely abolish nociceptive behavioral, reflex responses, or ventrobasal thalamic neuronal responses evoked by noxious intraoral or perioral stimuli, including tooth pulp stimulation (Dallel et al. 1988, 1989; Greenwood and Sessle 1976; Raboisson et al. 1989; Vyklicky et al. 1977; Young and Perryman 1984; c.f., Yokota et al. 1986). In line with these findings, disruption of rostral components including Vo has been reported to interfere with nociceptive sensation or behavior evoked by noxious intraoral and perioral stimulation (Broton and Rosenfeld 1986; Graham et al. 1988; Luccarini et al. 1995; Pickoff-Matuk et al. 1986; Young and Perryman 1984). Electrophysiological studies also have demonstrated that nociceptive neurons responding to noxious stimuli applied to the intraoral or perioral region also occur in the Vo, and some of them are sensitive to diffuse noxious inhibitory controls and show a long latency discharge that can be related to C-fiber input evoked by high-intensity electrical stimulation applied to their RF (Azerad et al. 1982; Dallel et al. 1990, 1998; Hu and Sessle 1984; Hu et al. 1992; Parada et al. 1997; Raboisson et al. 1995). These various findings support the view that a dual organization may exist in the processing of nociceptive information from the V region and that Vo may be concerned in orofacial pain mechanisms, especially those related to intraoral or perioral tissues.

The present study supports the accumulating evidence for the involvement of Vo in orofacial nociceptive mechanisms. Many nociceptive neurons classified as WDR, NS, D, and S + D were documented in the rat Vo. They had a RF involving maxillary and/or mandibular divisions mainly located in the intraoral and/or perioral regions and could be activated by noxious mechanical and thermal stimuli applied to their RF. Many of the WDR and NS neurons responded to noxious heat applied to the RF and received C-fiber input evoked by electrical stimulation of the RF as well as A-fiber electrically evoked input.

Only 1 of 15 neurons tested in the present study showed an immediate response to application to the pulp of the small-fiber excitant MO. This finding is surprising since the tooth pulp is predominantly innervated by small myelinated (A) and unmyelinated C fibers (Dubner et al. 1978; Nahri et al. 1994; Sessle 1987) and the Vo does receive a heavy primary afferent projection from the tooth pulp, including molar afferents (Arvidsson and Gobel 1981; Marfurt and Turner 1984; Shigenaga et al. 1986; Takemura et al. 1991), and many Vo neurons can be activated by tooth pulp stimulation (for review, see Dubner et al. 1978; Sessle 1987). Moreover, some c-fos expression has been reported in the dorsomedial part of Vo following noxious electrical or chemical (capsaicin) stimulation of intraoral tissues including the tooth pulp (Allen et al. 1996; Oakden and Boissonade 1998; Sugimoto et al. 1998), although in most studies, tooth pulp stimulation or noxious stimulation of other orofacial tissues predominantly elicits c-fos expression in Vc (Carstens et al. 1995; Coimbra and Coimbra 1994; Eberberger et al. 1995; Hathaway et al. 1995; Iwata et al. 1995; Shepheard et al. 1995; Wakasaki et al. 1992). Nonetheless, the types of afferents and the pathway(s) involved have not been elucidated (see Sessle 2000), and possible explanations for our finding are that the great majority of Vo neurons were not receiving a relatively direct MO-evoked afferent input, or some types of pulp nociceptive afferents projecting to Vo may not have been activated by MO.

Central sensitization reflected in neuroplastic changes in spinal and V nociceptive pathways can be induced by nociceptive afferent inputs, e.g., inflammation or injury-induced RF expansions and heightened neuronal excitability, and has been explained by unmasking or strengthening of the convergent afferent inputs to the central neurons (Chiang et al. 1998; Coderre and Katz 1997; Dubner and Basbaum 1994; Woolf 1992). MO injection into the tooth pulp, masseter muscle, or temporomandibular joint produced significant increase in jaw muscle EMG activity, and neuroplastic changes in nociceptive neurons in Vc (Chiang et al. 1998; Hu et al. 1992; Sunakawa et al. 1999; Yu et al. 1993, 1995). In some Vo nociceptive neurons, injection of MO into the deep masseter muscle induced a prolonged but reversible increase in responses to electrical stimulation of cutaneous afferent inputs (Hu et al. 1992). We found that MO application to the pulp also produced neuroplastic changes in Vo nociceptive neurons, such as significant increases in spontaneous activity, facial cutaneous, intraoral or deep RF size, and responsiveness to suprathreshold mechanical stimuli, and a marked decrease in the mechanical threshold. The pulpal application of MO did not produce any significant neuroplastic changes in Vo LTM neurons, consistent with findings in spinal dorsal horn (Woolf and King 1990) and Vc (Chiang et al. 1998). The neuroplastic changes in Vo nociceptive neurons appear to reflect a central sensitization induced by noxious stimulation of the tooth pulp and may be related to the hyperalgesia and spread of pain that can often be seen clinically after injury or inflammation of the tooth pulp (Grushka and Sessle 1984; Sessle 2000; Sharav 1994).

Some of neuroplastic changes observed in Vo nociceptive neurons following pulpal MO application had a different time course compared with those we have previously documented in Vc nociceptive neurons (Chiang et al. 1998), despite the fact that Vo and Vc neuronal recordings were performed under similar experimental conditions. For example, the pinch or deep RF size expansion reached its peak at 20 min in Vc and at 30 min in Vo following MO application, and significant increases in responses to suprathreshold mechanical stimuli occurred throughout the 60-min observation period in Vo, whereas significant changes were no longer apparent at 40 min in Vc. In addition, following MO application, Vc WDR neurons often had afterdischarges in response to pinch, whereas Vo WDR neuronal responses increased moderately. The often slower but longer expression of most neuroplastic changes in Vo nociceptive neurons raises the possibility that these MO-induced changes in Vo may be indirectly activated via Vc. It is noteworthy that Vo does not possess many of the features normally viewed as crucial in nociceptive processing in Vc, e.g., lack of laminae I-II and limited density of putative neurotransmitters or receptors associated with nociceptive transmission (Mansour et al. 1994; Petralia et al. 1994; Sessle 1987, 2000; Tallaksen-Greene et al. 1992). Neurons of Vc project to rostral V sensory nuclei including Vo via ascending intersubnuclear pathways (Hu et al. 1981; Ikeda et al. 1982, 1984; Jacquin et al. 1990; Nasution and Shigenaga 1987; Panneton and Burton 1982), and Vc exerts a net facilitatory modulating influence on Vo neurons (Greenwood and Sessle 1976; Khayyat et al. 1975; Young and King 1972). Dallel et al. (1998) recently have shown that C-fiber evoked responses of Vo nociceptive neurons can be depressed by morphine injection into the superficial laminae of Vc, but not into Vo itself, suggesting that morphine might exert its antinociceptive action on Vo nociceptive neurons by blocking the C-fiber inputs that relay in the Vc superficial laminae. This is further supported by our recent findings that CoCl₂ microinjected into Vc can significantly and reversibly block the neuroplastic changes in Vo nociceptive neurons evoked by pulpal application of MO (Chiang et al. 2000). Thus it is likely that Vc provides an input to Vo that may contribute to the role of Vo in orofacial pain mechanisms by virtue of its ascending projections to Vo.

It is interesting to note that the pulpal application of MO did not produce any significant neuroplastic changes in Vo nociceptive neurons that were located in the rostral dorsomedial part of Vo and that had a RF localized to the intraoral region. At the moment, the functional significance of the marked intraoral primary afferent projections to the dorsomedial part of Vo (Arvidsson and Gobel 1981; Falls 1988; Kruger et al. 1988; Marfurt and Turner 1984; Shigenaga et al. 1986; Sugimoto et al. 1997; Takemura et al. 1991) remains unclear. Recent studies (Shigenaga et al. 2000; Yoshida et al. 1994) have indicated that the intraoral nociceptive neurons of Vo may contribute to reflex motor functions related to pain, rather than sensory aspects of pain, since some Vo nociceptive neurons having their RF in intraoral regions directly project to the V motor nucleus. The direct afferent projections to Vo also may have little if any role in inducing the Vo neuroplastic changes, because morphine, MK-801, or CoCl₂ injected into Vc can eliminate most C-fiber-evoked responses, "wind-up," and neuroplastic changes in Vo (Chiang et al. 2000; Dallel et al. 1998; Woda et al. 1998). Previous studies have implicated an intersubnuclear pathway originating in Vc in the modulatory effects that Vc exerts on Vo (Greenwood and Sessle 1976; Ikeda et al. 1984; Jacquin et al. 1990; Khayyat et al. 1975; Nasution and Shigenaga 1987; Panneton and Burton 1982; Young and King 1972). It is conceivable that this intersubnuclear pathway might provide an important substrate involved in the central sensitization in Vo, and the lack of neuroplastic changes in the Vo nociceptive neurons with a RF localized to the intraoral region could then be explained by a differential distribution of the intersubnuclear bundles within Vo since these bundles are sparse within the rostral dorsomedial part of Vo compared with other Vo regions (Gobel and Purvis 1972).

NMDA receptor mechanisms play a critical role in the central sensitization process associated with inflammatory conditions (Basbaum 1999; Chiang et al. 1998; Coderre and Katz 1997; Dubner and Basbaum 1994; Kolhekar et al. 1994; Ren et al. 1992; Urban et al. 1994; Woolf and Thompson 1991; Yu et al. 1996). The present study also has indicated that NMDA receptor mechanisms are involved in the central sensitization induced by pulpal application of MO in Vo since pretreatment with MK-801 intranuclearly injected into Vo could antagonize the MO-induced neuroplastic changes in the Vo nociceptive neurons. Parada et al. (1997) have also shown that systemic MK-801 can antagonize C-fiber-related responses and wind-up produced by orofacial stimulation in Vo nociceptive neurons. They have argued that this MK-801 antagonizing effect may depend on the ascending projection from Vc. This was further supported by a recent finding (Woda et al. 1998) that local application of MK-801 to Vc can effectively block the C-fiber-evoked responses of some Vo nociceptive neurons. Nonetheless, in the present study a dose of 0.3 µl MK-801 (3 µg equivalent to 8.9 nmol) applied directly into Vo itself can antagonize Vo neuroplastic changes induced by nociceptive afferent inputs from the tooth pulp, although the diffusion range of MK-801 was not determined. Our findings indicate that NMDA receptor mechanisms may also be locally involved in Vo. Indeed, this hypothesis is supported by anatomical findings of NMDA receptor expression in Vo as well as in Vc (Dohrn and Beitz 1994; Petralia et al. 1994). Collectively, these results suggest a role for NMDA receptor mechanisms within Vo itself in V central sensitization and hyperalgesia that can result from pulpal injury or inflammation.