Bidirectional modulation of nociception by GABA neurons in the dorsolateral pontine tegmentum that tonically inhibit spinally projecting noradrenergic A7 neurons

Pontine GABA neurons modulate nociception

Pergamon PII: S0306-4522(99)00603-X

Neuroscience Vol. 96, No. 4, pp. 773–783, 2000 773 Copyright q 2000 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306-4522/00 $20.00+0.00

www.elsevier.com/locate/neuroscience

BIDIRECTIONAL MODULATION OF NOCICEPTION BY GABA NEURONS IN THE DORSOLATERAL PONTINE TEGMENTUM THAT TONICALLY INHIBIT SPINALLY PROJECTING NORADRENERGIC A7 NEURONS K. NUSEIR and H. K. PROUDFIT* Department of Pharmacology, University of Illinois at Chicago, Chicago, IL 60612, U.S.A.

Abstract—The A7 catecholamine cell group in the dorsolateral pontine tegmentum constitutes an important part of the descending pathways that modulate nociception. Evidence from immunocytochemical studies demonstrate that noradrenergic A7 neurons are densely innervated by GABA terminals arising from GABA neurons that are located in the dorsolateral pontine tegmentum medial to the A7 cell group. GABAA receptors are also located on the somata and dendrites of noradrenergic A7 neurons. These findings suggest that noradrenergic neurons in the A7 cell group may be under tonic inhibitory control by GABA neurons. To test this hypothesis, the GABAA antagonist bicuculline methiodide in doses of 0.2 or 1.0 nmol was microinjected into sites located dorsal to the A7 cell group and the resulting effects on tail flick and nociceptive foot withdrawal responses were measured. Both doses of bicuculline produced significant increases in tail flick latencies and small, but significant, increases in foot withdrawal latencies. Intrathecal injection of the alpha2-adrenoceptor antagonist yohimbine, in a dose of 76.7 nmol (30 mg), attenuated the antinociceptive effect of bicuculline on both the tail and the feet. In contrast, the alpha1-adrenoceptor antagonist WB4101, in a nearly equimolar dose of 78.6 nmol (30 mg), increased the antinociceptive effect of bicuculline on both the tail and the feet. Intrathecal injection of the antagonists alone did not consistently alter nociceptive responses of either the feet or the tail. These findings suggest that noradrenergic neurons in the A7 cell group are tonically inhibited by local GABA neurons. Furthermore, these findings suggest that inhibition of GABAA receptors located on spinally-projecting A7 noradrenergic neurons disinhibits, or activates, two populations of A7 neurons that have opposing effects on nociception. One of these populations facilitates nociception by an action mediated by alpha1-adrenoceptors in the spinal cord dorsal horn and the other population inhibits nociception by an action mediated by alpha2-adrenoceptors. q 2000 IBRO. Published by Elsevier Science Ltd. Key words: antinociception, bicuculline, yohimbine, WB4101.

ventrolateral periaqueductal gray 3 and the region of the ventromedial medulla that includes the nucleus raphe magnus and nucleus gigantocellularis pars a. 11,28 Some of these neurons that project from the ventromedial medulla 28 and the periaqueductal gray 26,27 contain methionine-enkephalin and activation of these enkephalin neurons may produce antinociception by activating spinally projecting A7 neurons. This possibility is supported by the finding that microinjection of morphine near the A7 cell group produces antinociception that can be blocked by intrathecal injection of alpha2adrenoceptor antagonists. 29 It is unlikely that morphine directly activated descending A7 neurons in that study because met-enkephalin-containing axons do not directly contact A7 noradrenergic neurons. Rather, the majority of met-enkephalin-immunoreactive axon terminals are located on non-catecholamine neurons in the A7 region. 28 Some of these non-catecholamine neurons may be GABA interneurons that are abundant in the region medial to the A7 neurons. 44,48,49 These observations suggest that activation of enkephalin neurons in the ventromedial medulla or periaqueductal gray may inhibit local GABA interneurons, which would disinhibit A7 neurons that are tonically inhibited by the GABA interneurons. This possibility is supported by several observations: (i) noradrenergic A7 neurons are densely innervated by GABA terminals; 44,48,49 (ii) the results of anterograde tract tracing studies demonstrated that the terminals of neurons in the ventromedial medulla, some of which contain met-enkephalin, 28 are located on the somata and proximal dendrites of GABA neurons located medial to the A7 neurons; 48 and (iii) GABAA receptors, demonstrated by immunocytochemical localization of a3 subunits, are

The A7 catecholamine cell group in the dorsolateral pontine tegmentum (DLPT) is an important component of the descending pathways that modulate nociception. 55 The A7 cell group provides the major noradrenergic innervation of laminae I–IV in the spinal cord dorsal horn, 10 which is a major terminal field of primary afferent nociceptors and contains second order spinothalamic tract neurons. 36 These descending noradrenergic neurons appear to modulate directly the activity of spinothalamic tract neurons because noradrenergic terminals form synapses with identified spinothalamic tract neurons 73,74 and unidentified dorsal horn neurons. 15,16,23 This conclusion is supported by several reports, which demonstrate that electrical 78 or chemical 79 stimulation of sites near the A7 cell group produces antinociception that is partially blocked by intrathecal administration of alpha2-adrenoceptor antagonists. These findings are consistent with anatomical studies, which demonstrate that the majority of the alpha2-adrenoceptors located in the dorsal horn are of the alpha2A and alpha2C type. 45,59,66,69,77,81 The A7 cell group receives convergent projections from other brainstem areas involved in modulating nociception such as the periaqueductal gray and the ventromedial medulla. For example, anterograde tract tracing studies demonstrate significant projections from neurons in the *To whom correspondence should be addressed. Tel.: 1 1-312-996-2349; fax: 1 1-312-996-1225. E-mail address: [email protected] (H. K. Proudfit). Abbreviations: DLPT, dorsolateral pontine tegmentum; FWL, foot withdrawal latency; NGC, nucleus reticularis gigantocellularis; PBS, phosphate-buffered saline; TFL, tail flick latency. 773

774

K. Nuseir and H. K. Proudfit

located on the somata and dendrites of noradrenergic A7 neurons. 49 Taken together, this evidence leads to the proposal that GABA neurons in the DLPT tonically inhibit spinally ;projecting noradrenergic neurons in the A7 cell group. To test this hypothesis, the GABAA antagonist bicuculline was microinjected near the noradrenergic neurons in the A7 cell group and the resulting effects on nociception were measured using the tail flick and foot withdrawal responses. Alphaadrenoceptor antagonists were injected intrathecally to determine whether the effects of bicuculline were produced by activation of noradrenergic neurons. A preliminary account of these results has been published as an abstract. 49 EXPERIMENTAL PROCEDURES

Nociceptive testing procedures Nociception was assessed by determining the latencies of the tail flick and foot withdrawal responses to a noxious thermal stimulus as described in detail elsewhere. 18,47,80 Response latency values for three different sites on the tail were averaged and defined as the tail flick latency (TFL). Response latencies for the left and right feet were averaged and defined as the foot withdrawal latency (FWL). In the absence of a response, skin heating was terminated after 10 s to minimize tissue damage. Microinjection of bicuculline or vehicle near the A7 cell group Female Sprague–Dawley rats (300–400 g, Charles River, Portage, MI, or Sasco, Madison, WI, U.S.A.) were lightly anesthetized with sodium pentobarbital (35 mg/kg, i.p.) and immobilized in a stereotaxic instrument. An incision was made along the midline of the scalp and a burr hole was drilled in the skull to allow insertion of a 24-gauge microinjection guide cannula into the DLPT. The tail and feet were blackened with India Ink to provide more uniform skin heating during the application of the noxious thermal stimulus. A 30-gauge stainless steel microinjection cannula was lowered through the guide cannula into the DLPT using the following stereotaxic coordinates with the incisor bar set to 22.5 mm: posterior ˆ 0.1 mm, lateral ˆ 2.3 mm, dorsal ˆ 2.3 mm. After establishing baseline response latencies, the GABAA receptor antagonist bicuculline methiodide (Sigma Chemical Co., St Louis, MO, U.S.A.) was microinjected into sites near the dorsal border of the A7 cell group in a dose of either 0.1 mg (0.2 nmol) or 0.5 mg (1.0 nmol). TFLs and FWLs were then measured 5, 10, 15, or 30, 45 and 60 min after the injection. In a separate group of animals, saline or phosphate-buffered saline (PBS) was microinjected in the A7 cell group to determine the effect of the vehicle on the tail and foot responses. Saline was used as the vehicle for the initial studies, but the pH of saline was found to be variable and since bicuculline is unstable at basic pH values, 33 PBS (50 mM, pH 7.4) was used as the drug vehicle. No significant difference was noticed in the effects of bicuculline when either vehicle was used. Bicuculline solutions were freshly prepared immediately before each injection to minimize inactivation of the drug. 33,50 Sprague–Dawley rats from both Sasco and Charles River were used in this study because rats were not available from Sasco for several months. Sasco rats were derived from Charles River rats and both of these rat substrains are genetically similar. In addition, no differences in drug responses between the rats from these two vendors were observed. Intrathecal injection of noradrenergic antagonists In two separate groups of rats, an intrathecal catheter made of polyethylene tubing (PE-10) was inserted through an incision in the cisterna magna to the level of the lumbar enlargement. Baseline response latencies were determined 15, 10 and 5 min before bicuculline was microinjected in the A7 cell group in a dose of 0.5 mg (1.0 nmol) and 15 min later nociceptive response latencies were again determined. Immediately after the measurement of response latencies, 10 ml of the alpha2-adrenoceptor antagonist yohimbine, in a dose of 76.7 nmol (30 mg), or the alpha1 antagonist WB4101, in a dose of 78.6 nmol (30 mg), was administered intrathecally followed by 10 ml saline to flush the catheter. TFLs and FWLs were measured again

at 5, 10, 15, 30, 45, and 60 min after the intrathecal injection. Each antagonist was dissolved in physiological saline. The antagonist drug doses were chosen on the basis of published reports that used similar doses of these drugs to block the antinociceptive effects produced by activating spinally-projecting noradrenergic neurons. 42,72,78,79 A third group of rats was used to determine the effects of an intrathecal injection of either yohimbine or WB4101 alone on nociceptive response latencies. The protocol for drug injections, antagonist drug doses, and testing procedures was identical to the previous experiment except bicuculline was not microinjected in the DLPT. A fourth group of rats was used to determine the effects of pretreatment with an intrathecal injection of yohimbine on the antinociceptive effect of bicuculline. This experiment was done because the duration of the bicuculline effect on FWLs was very brief and recovered before the peak effect of yohimbine, which did not allow antagonist effects to be determined. Baseline FWLs were determined and yohimbine in a dose of 76.7 nmol (30 mg) was injected intrathecally. Bicuculline in a dose of 1.0 nmol was injected 10 min after the yohimbine injection and FWLs were measured at 5, 10, 15, 30, 45, and 60 min after the bicuculline injection. Methodological considerations The effects of bicuculline were assessed using lightly-anesthetized rats, which allows more direct comparisons to electrophysiological studies done in the same preparation, and reduces variability due to factors such as extraneous uncontrollable environmental stimuli and state-dependent drug effects. However, these advantages favor the use of an anesthetized preparation only if the effects of bicuculline are comparable in conscious and anesthetized preparations. One factor that may result in different drug actions in anesthetized preparations is the common ability of many anesthetics such as halothane, isoflurane, and barbiturates to enhance GABA-mediated inhibition of neurons. 33,40 However, this action of anesthetics appears to produce quantitative rather than qualitative differences in some cases and no effect in other cases. For example, methohexital increases the antinociceptive potency of bicuculline microinjected in the ventromedial medulla 24 compared to conscious rats. 17 These findings suggest that if the magnitude of GABA-mediated inhibition is greater in barbiturateanesthetized rats, then the magnitude of the antinociceptive effect produced by microinjection of bicuculline in the A7 cell group may be greater than that produced in conscious rats. However, this conclusion is not supported by two reports which demonstrated that the potency of the GABAA receptor agonist THIP microinjected into sites in the ventromedial medulla were the same in conscious 17 and barbiturate-anesthetized rats. 24 Similar ambiguous findings have been reported for the effects of barbiturates on the antinociceptive effects of opioids, which are mediated in part by a reduction of GABA-mediated inhibition. 12,53,57,65 If the magnitude of GABA-mediated inhibition is greater in barbiturateanesthetized rats, then the potency of morphine should be less than that produced in conscious rats. This prediction is supported by the observation that pentobarbital decreases the potency of morphine microinjected in the periaqueductal gray compared to unanesthetized rats. 1,51 However, the antinociception produced by systemically 9 or intrathecally 51 administered morphine is not affected by pentobarbital. Thus, the antinociception produced by microinjection of bicuculline in the A7 cell group probably reflects a similar action in conscious rats, but the magnitude of the effect may be greater. Localization of microinjection sites in immunocytochemically stained sections At the end of each experiment, the animal was deeply anaesthetized with pentobarbital (50 mg/kg, i.p.) and transcardially perfused with 100 ml of saline, 200 ml of 4% phosphate-buffered paraformaldehyde (pH 7.4), and 100 ml of 10% sucrose in phosphate buffer. Brains were removed, postfixed overnight, and stored for at least three days in 20% sucrose in 0.1 M phosphate buffer at 48C for cryoprotection. Freefloating sections were processed for tyrosine hydroxylase immunoreactivity to allow the identification of noradrenergic neurons that comprise the A7 cell group. These methods have been previously described in detail. 28 Briefly, 40 mm transverse sections were cut on a cryostat microtome, washed three times in PBS (pH 7.4), incubated for 60 min in a blocking solution that contained 0.8% bovine serum albumin and 0.2% gelatin in PBS, then incubated for two to three days in primary antisera directed against tyrosine hydroxylase (mouse

Pontine GABA neurons modulate nociception

775

not different under either control and drug-treated conditions and these values for the left and right feet were averaged. Statistical comparisons among treatment groups were made using two-way analysis of variance (ANOVA) for repeated measures 35 and P # 0.05 was considered to be statistically significant. Each drug treatment group consisted of five to 20 rats. Animal care and use The care and use of animals in these experiments were conducted in accordance with the National Institute of Health guide for the care and use of laboratory animals (NIH Publications No. 90-23, revised 1978). All efforts were made to minimize animal suffering, reduce the number of animals used, and use alternatives to in vivo experiments. RESULTS

Antinociceptive effect of bicuculline microinjected near the A7 cell group

Fig. 1. Microinjection of bicuculline into sites in the DLPT near the A7 noradrenergic cell group increased TFL (A) and FWL (B). Mean baseline response latencies (^S.E.M. s) are plotted between 215 min and time zero. The arrowhead indicates the time at which bicuculline or vehicle was microinjected in the DLPT. The closed circles (X) represent the mean response latencies before and after microinjection of vehicle (VEH) into sites in the DLPT. The open squares (A) represent the mean response latencies before and after microinjection of 0.2 nmol of bicuculline (BIC 0.2) in the DLPT. The filled squares (B) represent the mean response latencies before and after microinjection of 1.0 nmol of bicuculline (BIC 1) in the DLPT. The numbers of animals in each group are indicated by the numbers in parentheses in the graph legend. Some of the error bars are obscured by the plotting symbols in this and subsequent figures.

anti-tyrosine hydroxylase; DiaSorin., Stillwater, MN, U.S.A.) diluted 1:1000 in PBS that contained 0.5% Triton X-100. Sections were then incubated for 60 min in donkey anti-mouse secondary antisera (1:100; Jackson Immunoresearch Laboratories, Inc., West Grove, PA, U.S.A.) and for an additional 60 min in mouse peroxidase anti-peroxidase (Cappel, Organon Teknika Corp., Durham, NC, U.S.A.) diluted 1:150. Sections were incubated for 3–5 min in 0.05% diaminobenzidine and 0.009% H2O2 in PBS. Finally, the sections were mounted on gelatin-coated slides and coverslipped. The injection sites and the tyrosine hydroxylase-immunoreactive A7 neurons were plotted on drawings of transverse brainstem sections made using a digital camera lucida (NeuroLucida, MicroBrightfield, Colchester, VT, U.S.A.). The locations of the microinjection sites, with respect to the tyrosinehydroxylase-immunoreactive A7 cell group neurons, were determined by plotting the most ventral position of the cannula tip on camera lucida drawings of serial sections using bright-field microscopy. The minimum area affected by the microinjections of bicuculline was assumed to be a sphere of tissue centered on the plotted point with a radius of 500 mm. 54,61,62 Statistical analysis Tail flick and foot withdrawal response latencies are presented as the mean ^ S.E.M. Mean response latencies of the left and right feet were

Microinjection of the GABAA receptor antagonist bicuculline in doses of 0.2 or 1.0 nmol in the DLPT near the A7 cell group produced statistically significant increases in mean tail flick latencies compared to mean response latencies after vehicle injections into similar sites (two-way ANOVA, P , 0.05; Fig. 1A). The antinociceptive effects of both doses reached a peak within 5 min and lasted for more than 60 min. In contrast to the effects on tail flick latencies, the high dose of bicuculline (1.0 nmol) produced only a small, but statistically significant increase in mean foot withdrawal latencies compared to mean response latencies after vehicle injections into similar sites near the A7 cell group (two-way ANOVA, P , 0.05; Fig. 1B). The duration of this effect was short and the response latency values returned to control levels by 30 min. Most of the sites at which microinjection of bicuculline was effective were less than 500 mm dorsal to A7 noradrenergic neurons (Fig. 2). One site at which microinjection of the 0.2 nmol dose produced antinociception was located more than 1000 mm rostral to the A7 cell group and was not included in the analysis of the effects of bicuculline microinjected near the A7 cell group. Two injection sites that were near the A7 neurons produced no effect and were also excluded from the data analysis. Three additional microinjection sites that had no effect on nociception were located 650, 1000, and 2000 mm dorsolateral to the center of the A7 cell group. Most of the sites where microinjection of the 1.0 nmol dose was effective (15/20) were just dorsal to the A7 group, 4/20 were located at the level of the rostral A7 group, and one site was caudal at the level of the subcoeruleus nucleus, but less than 1000 mm from the A7 neurons. Seven injection sites did not produce any effect and were not included in the data analysis. Three of these sites were located less than 500 mm from the center of either the rostral or the caudal pole of the A7 cell group and the other four were located between 650 and 1300 mm from the A7 neurons. All vehicle injections were made into sites comparable to those at which bicuculline was injected and were located either dorsal or lateral to the A7 noradrenergic neurons at distances of less than 500 mm (Fig. 2). Effect of intrathecal injection of alpha-adrenoceptor antagonists on bicuculline-induced antinociception Another set of experiments determined whether microinjection of bicuculline near the A7 cell group produces

776

K. Nuseir and H. K. Proudfit

Fig. 2. Distribution of sites in the DLPT at which bicuculline was microinjected in doses of 0.2 nmol (A) and 1.0 nmol (B). The filled circles (X) on the left side of each section represent effective microinjection sites and the stars (w) represent noradrenergic neurons identified by tyrosine hydroxylase immunoreactivity. The open circles (W) represent microinjection sites that were not effective. The open squares (A) represent microinjection sites at which vehicle was injected. Scale bar ˆ 1 mm. A7, A7 catecholamine cell group; A10 dc, A10 cell group, dorsocaudal part; Aq, cerebral aqueduct; IC, inferior colliculus; PYR, pyramidal tract; SubCA, subcoeruleus nucleus alpha; SubCD, dorsal subcoeruleus nucleus; SubCV, ventral subcoeruleus nucleus.

antinociception by disinhibiting and activating spinally projecting A7 neurons. In these experiments bicuculline was microinjected near the A7 cell group and either yohimbine, an alpha2-adrenoceptor antagonist, or WB4101, an alpha1adrenoceptor antagonist, was injected intrathecally. Bicuculline, in a dose of 1.0 nmol, produced antinociception (Fig. 3) when microinjected into sites located immediately dorsal to the A7 noradrenergic neurons (Fig. 4). One site located in the inferior colliculus did not alter nociceptive responses. Intrathecal injection of yohimbine in a dose of 76.7 nmol (30 mg) attenuated the antinociceptive effect of bicuculline as measured by TFL within 5 min and completely blocked the antinociception by 15 min (Fig. 3A). In contrast, WB4101 in a dose of 78.6 nmol (30 mg) potentiated the antinociceptive effect of bicuculline within 5 min. This enhancement persisted for more than 45 min (two-way ANOVA, P , 0.05). Bicuculline also produced a small, but statistically significant, elevation in mean FWLs that lasted less than 15 min. Intrathecal injection of yohimbine (76.7 nmol) did not have any significant effect, presumably because the peak effect of yohimbine occurs between 15 and 30 min, but the effect of bicuculline lasted less than 30 min. However, the antinociceptive effects of bicuculline on both the feet and the tail were blocked by an intrathecal injection of yohimbine (76.7 nmol) that preceded the microinjection of bicuculline by 10 min (Fig. 5). In contrast, intrathecal injection of WB4101 (78.6 nmol) potentiated the effect of bicuculline and increased the mean FWL to nearly four times baseline values by 5 min after the intrathecal antagonist injection (two-way ANOVA,

P , 0.05, Fig. 3B). Intrathecal injection of neither yohimbine (76.7 nmol) nor WB4101 (78.6 nmol) produced any statistically significant changes in the response latencies of either the feet or the tail at any time-point after antagonist injection (Fig. 6).

DISCUSSION

Evidence that GABA tonically inhibits spinally projecting noradrenergic A7 neurons The present results support the proposal that GABA neurons tonically inhibit spinally projecting A7 neurons and that blockade of GABAA receptors disinhibits A7 neurons and produces antinociception mediated by alpha2-adrenoceptors in the spinal cord. Specifically, the observation that microinjection of the competitive GABAA receptor antagonist bicuculline 33,38 into sites dorsal to the A7 cell group produces antinociception that can be blocked by intrathecal injection of the alpha2-adrenoceptor antagonist yohimbine indicates that blockade of GABAA receptors in the DLPT disinhibits and activates A7 noradrenergic neurons. The activation of A7 neurons by bicuculline appears to be mediated by tonically active excitatory inputs to A7 neurons and not by an increased intrinsic excitability of the A7 neurons themselves. This conclusion is supported by the observation that blockade of synaptic transmission to A7 neurons by microinjection of cobalt into the A7 cell group does not have any effect on nociception. 47 These results suggest that A7 neurons receive tonic inhibitory and excitatory inputs, but the inhibition

Pontine GABA neurons modulate nociception

777

Fig. 4. Distribution of sites in the DLPT at which microinjection of 1.0 nmol of bicuculline was followed by intrathecal injection of an alphaadrenoceptor antagonist. The stars (w) represent noradrenergic neurons identified by tyrosine hydroxylase immunoreactivity. The filled circles (X) on the left side of the section represent sites at which microinjection of bicuculline was followed by an intrathecal injection of the alpha1 antagonist WB4101, and the filled triangles (O) represent sites at which bicuculline microinjection was followed by an intrathecal injection of the alpha2 antagonist yohimbine. Scale bar ˆ 1 mm. A7, A7 catecholamine cell group; Aq, cerebral aqueduct; IC, inferior colliculus; PYR, pyramidal tract. Fig. 3. The effects of intrathecal injection of alpha-adrenoceptor antagonists on the antinociception produced by microinjection of 1.0 nmol of bicuculline into sites in the DLPT near the A7 cell group. Graph A illustrates mean TFLs and graph B illustrates mean FWLs. Mean baseline response latencies (^S.E.M. s) are plotted between 215 min and time zero. The left arrowhead indicates the time at which bicuculline or vehicle was microinjected in the DLPT and the right arrowhead indicates the time at which the alpha2 antagonist yohimbine (76.7 nmol) or the alpha1 antagonist WB4101 (78.6 nmol) was injected intrathecally. The filled circles (X) represent the mean response latencies before and after microinjection of vehicle (VEH) into sites in the DLPT. The filled squares (B) represent the effect of microinjecting 1.0 nmol bicuculline (BIC) into sites near the A7 cell group. The open triangles (D) represent the mean response latencies for the yohimbinetreated group (YHB). The filled triangles (O) represent the mean response latencies for the WB4101-treated group.

produced by tonically active GABA neurons predominates such that the A7 neurons are inactive or nearly inactive. This conclusion is further supported by the observation that intrathecal injections of alpha-adrenoceptor antagonists do not alter nociception (see Fig. 6 and previous reports 18,19,29). Neurons in other brainstem nuclei that modulate nociception are also tonically inhibited by GABA neurons. For example, direct electrophysiological evidence has demonstrated that bicuculline disinhibits and activates neurons in both the ventromedial medulla 25 and the periaqueductal gray. 4 The disinhibition of neurons in the ventromedial medulla and the periaqueductal gray by GABAA receptor blockade appears to produce antinociception because microinjection of the GABAA antagonists bicuculline 17,24 or gabazine 24 into the ventromedial medulla significantly elevates nociceptive response latencies. In contrast, inhibition of neurons in the ventromedial medulla by microinjection of the GABAA agonists muscimol or THIP into these nuclei

produces hyperalgesia. 17,24 Similar antinociceptive and hyperalgesic effects are produced by microinjection of bicuculline 43,52,60 or THIP, 14,58 respectively, into sites in the ventrolateral periaqueductal gray. The antinociception produced by microinjection of bicuculline in the periaqueductal gray appears to be mediated in part by inhibition of nociceptive neurons in the spinal cord dorsal horn. 7,8,63 Although direct electrophysiological evidence for disinhibition of noradrenergic A7 neurons by GABAA receptor blockade is not available, the presence of GABA terminals and GABAA receptors on the somata and dendrites of these A7 neurons 49 and the reduction of bicuculline-induced antinociception by intrathecal alpha2 antagonists (Figs 3 and 5), strongly supports the conclusion that noradrenergic A7 neurons are tonically inhibited by GABA and disinhibited by local injection of GABAA receptor antagonists. Bicuculline produces bidirectional effects on nociception that are mediated by alpha-adrenoceptors in the spinal cord The participation of spinally-projecting noradrenergic A7 neurons in the antinociception produced by microinjection of bicuculline into sites near the A7 cell group was examined by determining the effects of intrathecal injection of alphaadrenoceptor antagonists on bicuculline-induced antinociception. The alpha2 antagonist yohimbine blocked the antinociceptive effect of bicuculline on both the tail and the feet (Figs 3 and 5), which provides evidence that A7 neurons activated by bicuculline produce antinociception that is mediated by alpha2-adrenoceptors in the spinal cord. In contrast, intrathecal injection of the alpha1 antagonist WB4101 potentiated the antinociceptive effect of bicuculline on both the feet and

778

K. Nuseir and H. K. Proudfit

Fig. 5. The effects of the alpha2-adrenoceptor antagonist yohimbine injected intrathecally before the microinjection of 1.0 nmol of bicuculline into sites in the DLPT near the A7 cell group. Graph A illustrates mean TFLs and graph B illustrates mean FWLs. Mean baseline response latencies (^S.E.M. s) are plotted between 215 min and time zero. The left arrowhead indicates the time at which yohimbine (76.7 nmol) was injected intrathecally and the right arrowhead indicates the time at which bicuculline was microinjected in the DLPT. The filled circles (X) represent the mean response latencies before and after microinjection of vehicle (VEH) into sites in the DLPT. The filled squares (B) represent the effect of microinjecting 1.0 nmol bicuculline (BIC) into sites near the A7 cell group. The open triangles (D) represent the mean response latencies for the yohimbine-treated group (YHB).

the tail (Fig. 3). These observations indicate that bicuculline activates another population of A7 neurons that facilitates nociception by an action at alpha1-adrenoceptors in the spinal cord dorsal horn. Thus, bicuculline causes disinhibition of two different populations of spinally projecting A7 neurons. The bidirectional modulation of nociceptive transmission in the spinal cord that is produced by activation of bulbospinal A7 neurons appears to be a feature common to many afferent projections to the A7 cell group. For example, microinjection of morphine into sites near the A7 cell group also produces both inhibition and facilitation of nociception that is mediated by alpha2- and alpha1-adrenoceptors, respectively. 29 These observations, and the existence of met-enkephalin neurons in the ventromedial medulla that project to the area of the dorsolateral pontine tegmentum that includes the A7 cell group, 29 predict that activation of these enkephalin neurons should also produce bidirectional effects on nociception. This prediction has been confirmed by several reports, which demonstrate that microinjecting cholinergic muscarinic agonists in the ventromedial medulla produces opposing

Fig. 6. The effects of an intrathecal injection of alpha-adrenoceptor antagonists on mean TFL (A) and FWL (B). Mean baseline response latencies (^S.E.M. s) are plotted between 215 min and time zero. The arrowhead indicates the time at which the alpha2 antagonist yohimbine (76.7 nmol), the alpha1 antagonist WB4101 (78.6 nmol), or vehicle was injected intrathecally. The symbols represent the mean response latencies before and after intrathecal injection of vehicle (X, VEH), yohimbine (D, YHB), or WB4101 (O, WB4101).

effects on nociception. More specifically, intrathecal injection of the alpha1 antagonists prazosin or WB4101 enhances the antinociception produced by microinjecting the cholinergic agonists carbachol 6 or N-methyl carbachol 32 into sites in the nucleus raphe magnus. In contrast, the antinociception produced by these agonists is blocked by intrathecal injection of alpha2adrenoceptor antagonists such as yohimbine or idazoxan. 6,32 Similar bidirectional effects on nociception are produced by microinjection of morphine in the ventrolateral periaqueductal gray. Thus, the antinociception produced by morphine is enhanced by intrathecal injection of alpha1-adrenoceptor antagonists. 19 However, this effect is apparent only for nociceptive foot, but not tail responses. In contrast, the antinociceptive effect of morphine is reduced by intrathecal injection of alpha2-adrenoceptor antagonists using nociceptive tail, but not foot responses. 18 These findings have been partially confirmed by electrophysiological studies of sacral nociceptive dorsal horn neurons that were activated by noxious tail heating. The inhibition of these neurons produced by stimulation of neurons in the ventrolateral periaqueductal gray was reduced by iontophoretic application of alpha2-adrenoceptor antagonists, but was not affected by alpha1-adrenoceptor antagonists. 8 Taken together, these results demonstrate that activation of neurons in the PAG produces opposing effects

Pontine GABA neurons modulate nociception

on nociception that appear to be mediated by two different populations of spinally projecting A7 noradrenergic neurons. They also suggest that the nociceptive responses of the feet and tail in the rat are modulated by different pathways from the ventrolateral PAG to the spinal cord dorsal horn. Electrical or chemical stimulation of neurons in the gigantocellular reticular nucleus (NGC) also produces bidirectional modulation of nociception82,84 and the activity of nociceptive dorsal horn neurons. 85 The inhibitory actions of such stimulation may be mediated in part by descending noradrenergic neurons in the A7 cell group because intrathecal injection of alpha2-adrenoceptor antagonists significantly reduces the inhibition of nociceptive responses. 83 However, the facilitation of nociception produced by stimulation of sites in the NGC is mediated by the release of serotonin that acts at serotonin 5-HT1 receptors, but not by the release of norepinephrine in the spinal cord. 84 These observations may indicate that neurons in the NGC selectively activate the population of spinally-projecting noradrenergic A7 neurons that inhibits nociception by activating alpha2-adrenoceptors. A particular subset of neurons in the rostral ventromedial medulla, termed “ON” and “OFF” cells also produce bidirectional modulation of nociception. 20 However, these neurons project directly to the spinal cord 21 and their excitatory and inhibitory actions on nociceptive dorsal horn neurons do not appear to involve activation of descending noradrenergic neurons. Collectively, these findings provide strong evidence that some A7 neurons activate alpha1-adrenoceptors in the spinal cord dorsal horn and facilitate nociception, while others activate alpha2-adrenoceptors in the dorsal horn and inhibit nociception. A more detailed discussion of the evidence for the existence of two populations of spinally-projecting A7 neurons that produce opposite effects on nociception was presented in a previous report. 29 Evidence that alpha1-adrenoceptors in the spinal cord dorsal horn mediate facilitation of nociception The hypothesis that activation of alpha1-adrenoceptors in the spinal cord dorsal horn facilitates nociception is supported by the results of electrophysiological studies, which demonstrate that iontophoretic application of the alpha1 agonist phenylephrine activates nociceptive dorsal horn neurons. 13 Furthermore, the activation of these neurons by phenylephrine is blocked by the alpha1 antagonist WB4101. Norepinephrine also activates a population of unidentified neurons in the dorsal horn 5,13,31,41,46,70,71 that is mediated by alpha1-, 13,46 but not by alpha2-adrenoceptors. 46 Additional evidence that alpha1 agonists activate dorsal horn neurons is the finding that intrathecal injection of norepinephrine increases the expression of c-fos in the spinal cord dorsal horn; an effect that is blocked by alpha1-, but not alpha2adrenoceptor antagonists. 34 The activation of nociceptive dorsal horn neurons that is mediated by alpha1-adrenoceptors may result from a membrane depolarization produced by the reduction of a potassium current. For example, alpha1agonists applied in vitro to thalamocortical relay neurons produce a large slow depolarization due to the reduction of a resting potassium current. 39 This ionic mechanism is in contrast to alpha2 agonists that produce neuronal inhibition by membrane hyperpolarization that results from increasing inwardly rectifying potassium currents. 64,67,68 Although the majority of evidence supports the conclusion that

779

alpha1-adrenoceptor agonists activate neurons in the spinal cord dorsal horn, there is conflicting evidence which indicates that phenylephrine inhibits, rather than excites, nociceptive dorsal horn neurons. 22 Although a majority of the evidence indicates that alpha1-adrenoceptor agonists activate some nociceptive dorsal horn neurons, the effects of these agonists on nociception are less clear. For example, several reports indicate that intrathecal injection of alpha1 agonists such as phenylephrine 56 or methoxamine 30,37 produces antinociception rather than facilitation of nociception. However, we have reported that relatively high doses of phenylephrine produce both antinociception and severe motor effects such as serpentine tail movements. 2 In contrast, lower doses of phenylephrine, that do not produce overt motor effects, produce hyperalgesia, i.e. facilitation of nociception. Thus, the reported antinociceptive effects produced by alpha1-adrenoceptor agonists, such as phenylephrine, appear to result from motor effects and not from reduced nociceptive transmission. Finally, it may be argued that the apparent facilitation of nociception by alpha1 agonists is due to an enhancement of motor reflexes rather than activation of nociceptive dorsal horn neurons. For example, there are several observations which demonstrate that alpha1 agonists depolarize motoneurons 75 and enhance the monosynaptic reflex. 76 These excitatory effects on motoneurons could reduce the latencies of nociceptive responses and lead to the false conclusion that alpha1 agonists facilitate nociception. It may be similarly argued that alpha1-adrenoceptor antagonists increase the antinociception produced by the injection of bicuculline near the A7 cell group, because motor reflexes are reduced which would increase the latency of nociceptive responses. However, this argument is not valid because noradrenergic A7 neurons project almost exclusively to the spinal cord dorsal horn, with virtually no projections to motoneurons in the ventral horn. 10 Thus, activation of descending A7 neurons would have no direct effect on the activity of motoneurons. Furthermore, since intrathecal injection of an alpha1adrenoceptor antagonist alone does not alter nociceptive responses (Fig. 6 and previous reports 19,29), the enhancement of bicuculline-induced antinociception by alpha1 antagonists is most likely due to an action in the dorsal horn, and not the ventral horn. Thus, the hypothesis that activation of alpha1adrenoceptors in the spinal cord dorsal horn facilitates nociception is supported by the results of a variety of studies using different experimental approaches. The physiological function of GABA neurons in the dorsolateral pontine tegmentum Several anatomical and behavioral studies provide converging evidence that GABA neurons and spinally projecting noradrenergic A7 neurons in the dorsolateral pontine tegmentum are integral components of the brainstem pain modulation system that includes the ventromedial medulla and periaqueductal gray (Fig. 7). For example, the antinociception produced by activating neurons in the ventromedial medulla 47 and periaqueductal gray 18,19 is mediated in part by spinally projecting noradrenergic A7 neurons. Furthermore, neurons in both the ventromedial medulla 11,28 and the periaqueductal gray 3 have major projections to both noradrenergic and noncatecholamine neurons in the area of the dorsolateral pontine tegmentum that includes the A7 cell group. Neurons in the ventromedial medulla project to nearly all GABA neurons in

780

K. Nuseir and H. K. Proudfit

The neuronal model illustrated in Fig. 7 is based on these anatomical observations and shows the neuronal connections of enkephalin neurons in ventromedial medulla and neurons in the dorsolateral pontine tegmentum that contain GABA and norepinephrine. This model illustrates the hypothesis that enkephalin neurons in the ventromedial medulla inhibit GABA neurons in the dorsolateral pontine tegmentum which tonically inhibit spinally projecting noradrenergic A7 neurons. The inhibition of tonically active GABA neurons disinhibits A7 neurons and results in antinociception mediated by alpha2-adrenoceptors in the spinal cord dorsal horn. This model is supported by the results of a previous study that examined the functional significance of the enkephalin neurons that project from the ventromedial medulla to the A7 cells group. 29 These studies demonstrated that microinjection of morphine into the A7 cell group can produce antinociception that is blocked by intrathecal injection of alpha2-adrenoceptor antagonists. This observation suggests that morphine inhibits local GABA interneurons or the synaptic release of GABA, which disinhibits A7 neurons that are tonically inhibited by the GABA interneurons. Enkephalin neurons in the periaqueductal gray may serve a similar function to those in the ventromedial medulla. Neurons in the ventrolateral periaqueductal gray project to A7 neurons 3 and form synapses with both noradrenergic and non-catecholamine neurons in the region that includes the A7 cell group (Bajic and Proudfit, unpublished observations). Some of these projection neurons contain enkephalin 28 and activation of these neurons may inhibit GABA neurons in the dorsolateral pontine tegmentum that tonically inhibit noradrenergic A7 neurons. SUMMARY AND CONCLUSIONS

Fig. 7. Model that illustrates the proposed neuronal pathways that mediate the bidirectional effects on nociception produced by microinjecting the GABAA receptor antagonist bicuculline into sites in the DLPT near the A7 catecholamine cell group. a1, alpha1-adrenoceptor; a2, alpha2-adrenoceptor; ENK, enkephalin; NE, norepinephrine; PAG, periaqueductal gray; STT, spinothalamic tract; VMM, ventromedial medulla.

the area of the dorsolateral pontine tegmentum that includes the A7 cell group and the area medial to A7 neurons (Nuseir and Proudfit, unpublished observations). In addition, enkephalin neurons in the ventromedial medulla project primarily to non-catecholamine neurons in the dorsolateral pontine tegmentum 28 and many of these non-catecholamine neurons may contain GABA. This conclusion is based on the coexistence of mu opioid receptors and GABA-immunoreactivity in neurons in the dorsolateral pontine tegmentum (Jones and Proudfit, unpublished observations). Finally, a large number of GABA neurons located in the medial dorsolateral pontine tegmentum project to the region that contains noradrenergic A7 neurons, and glutamate decarboxylaseimmunoreactive terminals and GABAA receptors are located on the somata and dendrites of A7 neurons. 49

In summary, spinally projecting noradrenergic neurons in the A7 catecholamine cell group receive a moderate to dense innervation by GABA neurons and GABAA receptors are located on the somata and dendrites of these A7 neurons. The results of the present study demonstrate that microinjection of the competitive GABAA antagonist bicuculline dorsal to the A7 cell group produces antinociception that is mediated in part by disinhibition or activation of spinally projecting noradrenergic neurons in the A7 cell group. The antinociception produced by bicuculline was reduced by intrathecal injection of an alpha2-adrenoceptor antagonist and enhanced by intrathecal injection of an alpha1-adrenoceptor antagonist. These results suggest that the antinociception produced by bicuculline is the net effect of two opposing actions; facilitation of nociception that is mediated by alpha1 receptors in the spinal cord dorsal horn and inhibition of nociception that is mediated by alpha2-adrenoceptors. The results of these studies and the localization of GABAA receptors and GABA terminals on A7 neurons support the conclusion that GABA neurons in the dorsolateral pontine tegmentum tonically inhibit two different populations of spinally projecting noradrenergic neurons in the A7 cell group that have opposing effects on nociception. Finally, these results confirm the conclusions of several previous reports from this laboratory that noradrenergic neurons in the A7 cell group constitute an important system of neurons that can exert bidirectional control of nociceptive responses. Acknowledgement—This work was supported by USPHS grant DA03980 from the National Institute on Drug Abuse.

Pontine GABA neurons modulate nociception

781

REFERENCES

1. Aimone L. D. and Gebhart G. F. (1986) Stimulation-produced spinal inhibition from the midbrain in the rat is mediated by an excitatory amino acid neurotransmitter in the medial medulla. J. Neurosci. 6, 1803–1813. 2. Aran S. and Proudfit H. K. (1990) Antinociceptive interactions between intrathecally administered alpha noradrenergic agonists and 5 0 -N-ethylcarboxamide adenosine. Brain Res. 519, 287–293. 3. Bajic D. and Proudfit H. K. (1999) Neurons in the ventrolateral periaqueductal gray project to noradrenergic neurons in the brainstem that are involved in modulating nociception. J. comp. Neurol. 405, 359–379. 4. Behbehani M. M., Jiang M. R., Chandler S. D. and Ennis M. (1990) The effect of GABA and its antagonists on midbrain periaqueductal gray neurons in the rat. Pain 40, 195–204. 5. Belcher G., Ryall R. W. and Schaffner R. (1978) The differential effects of 5-hydroxytryptamine noradrenaline and raphe stimulation on nociceptive and non-nociceptive dorsal horn interneurones in the cat. Brain Res. 151, 307–321. 6. Brodie M. S. and Proudfit H. K. (1986) Antinociception induced by local injection of carbachol into the nucleus raphe magnus: alteration by intrathecal injection of noradrenergic antagonists. Brain Res. 371, 70–79. 7. Budai D. and Fields H. L. (1998) Endogenous opioid peptides acting at mu-opioid receptors in the dorsal horn contribute to midbrain modulation of spinal nociceptive neurons. J. Neurophysiol. 79, 677–687. 8. Budai D., Harasawa I. and Fields H. L. (1998) Midbrain periaqueductal gray (PAG) inhibits nociceptive inputs to sacral dorsal horn nociceptive neurons through alpha2-adrenergic receptors. J. Neurophysiol. 80, 2244–2254. 9. Cicero T. J., Meyer E. R. and Smithloff B. R. (1974) Alpha adrenergic blocking agents: antinociceptive activity and enhancement of morphine-induced analgesia. J. Pharmac. exp. Ther. 189, 72–82. 10. Clark F. M. and Proudfit H. K. (1991) The projection of noradrenergic neurons in the A7 catecholamine cell group to the spinal cord in the rat demonstrated by anterograde tracing combined with immunocytochemistry. Brain Res. 547, 279–288. 11. Clark F. M. and Proudfit H. K. (1991) Projections of neurons in the ventromedial medulla to pontine catecholamine cell groups involved in the modulation of nociception. Brain Res. 540, 105–115. 12. Cohen G. A., Doze V. A. and Madison D. B. (1992) Opioid inhibition of GABA release from presynaptic terminals of rat hippocampal interneurons. Neuron 9, 325–335. 13. Davies J. and Quinlan J. E. (1985) Selective inhibition of responses of feline dorsal horn neurones to noxious cutaneous stimuli by tizanidine (DS103282) and noradrenaline: involvement of a2-adrenoceptors. Neuroscience 16, 673–682. 14. Depaulis A., Morgan M. M. and Liebeskind J. C. (1987) Gabaergic modulation of the analgesic effects of morphine microinjected in the ventral periaqueductal gray matter of the rat. Brain Res. 436, 223–228. 15. Doyle C. A. and Maxwell D. J. (1991) Catecholaminergic innervation of the spinal dorsal horn: a correlated light and electron microscopic analysis of tyrosine hydroxylase-immunoreactive fibres in the cat. Neuroscience 45, 161–176. 16. Doyle C. A. and Maxwell D. J. (1991) Ultrastructural analysis of noradrenergic nerve terminals in the cat lumbosacral spinal dorsal horn: a dopamine-ßhydroxylase immunocytochemical study. Brain Res. 563, 329–333. 17. Drower E. J. and Hammond D. L. (1988) GABAergic modulation of nociceptive threshold: effects of THIP and bicuculline microinjected into the ventral medulla of the rat. Brain Res. 450, 316–324. 18. Fang F. and Proudfit H. K. (1996) Spinal cholinergic and monoamine receptors mediate the antinociceptive effect of morphine microinjected in the periaqueductal gray on the rat tail, but not the feet. Brain Res. 722, 95–108. 19. Fang F. and Proudfit H. K. (1998) Antinociception produced by microinjection of morphine in the periaqueductal gray is enhanced in the rat foot, but not the tail, by intrathecal injection of alpha1-adrenoceptor antagonists. Brain Res. 790, 14–24. 20. Fields H. L., Heinricher M. M. and Mason P. (1991) Neurotransmitters in nociceptive modulatory circuits. A. Rev. Neurosci. 14, 219–245. 21. Fields H. L., Malick A. and Burstein R. (1995) Dorsal horn projection targets of on and off cells in the rostral ventromedial medulla. J. Neurophysiol. 74, 1742–1759. 22. Fleetwood-Walker S. M., Mitchell R., Hope P. J., Molony V. and Iggo A. (1985) An a2 receptor mediates the selective inhibition by noradrenaline of nociceptive responses of identified dorsal horn neurones. Brain Res. 334, 243–254. 23. Hagihira S., Senba E., Yoshida S., Tohyama M. and Yoshiya I. (1990) Fine structure of noradrenergic terminals and their synapses in the rat spinal dorsal horn: an immunohistochemical study. Brain Res. 526, 73–80. 24. Heinricher M. M. and Kaplan H. J. (1991) GABA-mediated inhibition in rostral ventromedial medulla: role in nociceptive modulation in the lightly anesthetized rat. Pain 47, 105–113. 25. Heinricher M. M. and Tortorici V. (1994) Interference with GABA transmission in the rostral ventromedial medulla: disinhibition of off-cells as a central mechanism in nociceptive modulation. Neuroscience 63, 533–546. 26. Holden J. E. and Proudfit H. K. (1994) Projection of substance P and met-enkephalin neurons to the dorsolateral pontine tegmentum: a possible role in the modulation of nociception. Soc. Neurosci. Abstr. 20, 757. 27. Holden J. E. and Proudfit H. K. (1995) Projection of met-enkephalin neurons to noradrenergic cells in the dorsolateral pontine tegmentum: a possible role in modulating nociception. Soc. Neurosci. Abstr. 21, 1638. 28. Holden J. E. and Proudfit H. K. (1998) Enkephalin neurons that project to the A7 catecholamine cell group are located in brainstem nuclei that modulate nociception: ventromedial medulla. Neuroscience 83, 929–947. 29. Holden J. E. and Proudfit H. K. (1999) Microinjection of morphine in the A7 catecholamine cell group produces opposing effects on nociception that are mediated by alpha1- and alpha2-adrenoceptors. Neuroscience 91, 979–990. 30. Howe J. R., Wang J.-Y. and Yaksh T. L. (1983) Selective antagonism of the antinociceptive effect of intrathecally applied a-adrenergic agonists by intrathecal prazosin and intrathecal yohimbine. J. Pharmac. exp. Ther. 224, 552–558. 31. Howe J. R. and Zieglga¨nsberger W. (1987) Responses of rat dorsal horn neurons to natural stimulation and to iontophoretically applied norepinephrine. J. comp. Neurol. 255, 1–17. 32. Iwamoto E. T. and Marion L. (1993) Adrenergic, serotonergic and cholinergic components of nicotinic antinociception in rats. J. Pharmac. exp. Ther. 265, 777–789. 33. Johnston G. A. R. (1996) GABAA receptor pharmacology. Pharmac. Ther. 69, 173–198. 34. Jones S. L. (1992) Noradrenergic modulation of noxious heat-evoked fos-like immunoreactivity. J. comp. Neurol. 325, 435–445. 35. Keppel G. (1973) Design and Analysis, A Researcher’s Handbook. Prentice–Hall, Englewood Cliffs. 36. Light A. R. (1992) The initial processing of pain and its descending control: spinal and trigeminal systems. In Pain and Headache, (ed. Gildenberg P. L.), Vol. 12, pp. 1–306. Karger, Basel. 37. Loomis C. W. and Arunachalam K. D. (1992) Intrathecal St-587: effects on nociceptive reflexes and blood pressure in the rat. J. Pharmac. exp. Ther. 263, 428–435. 38. Macdonald R. L. and Olsen R. W. (1994) GABAA receptor channels. A. Rev. Neurosci. 17, 569–602. 39. McCormick D. A., Pape H.-C. and Williamson A. (1991) Actions of norepinephrine in the cerebral cortex and thalamus: implications for function of the central noradrenergic system. Prog. Brain Res. 88, 293–305. 40. Mehta A. K. and Ticku M. K. (1999) An update on GABAA receptors. Brain Res. Rev. 29, 196–217.

782

K. Nuseir and H. K. Proudfit

41. Millar J. and Williams G. V. (1989) Effects of iontophoresis of noradrenaline and stimulation of the periaqueductal gray on single-unit activity in the rat superficial dorsal horn. J. comp. Neurol. 287, 119–133. 42. Miller J. F. and Proudfit H. K. (1990) Antagonism of stimulation-produced antinociception from ventrolateral pontine sites by intrathecal administration of a-adrenergic antagonists and naloxone. Brain Res. 530, 20–34. 43. Moreau J. L. and Fields H. L. (1986) Evidence for GABA involvement in midbrain control of medullary neurons that modulate nociceptive transmission. Brain Res. 397, 37–46. 44. Mugnaini E. and Oertel W. H. (1985) An atlas of the distribution of GABAergic neurons and terminals in the rat CNS as revealed by GAD immunohistochemistry. In Handbook of Chemical Neuroanatomy (eds. Bjorklund A. and Ho¨kfelt T.), Vol. 4, pp. 436–608. Elsevier, Amsterdam. 45. Nicholas A. P., Pieribone V. and Ho¨kfelt T. (1993) Distributions of mRNAs for alpha-2 adrenergic receptor subtypes in rat brain: an in situ hybridization study. J. comp. Neurol. 328, 575–594. 46. North R. A. and Yoshimura M. (1984) The actions of noradrenaline on neurones of the rat substantia gelatinosa in vitro. J. Physiol. 349, 43–55. 47. Nuseir K., Heidenreich B. A. and Proudfit H. K. (1999) The antinociception produced by microinjection of a cholinergic agonist in the ventromedial medulla is mediated by noradrenergic neurons in the A7 catecholamine cell group. Brain Res. 822, 1–7. 48. Nuseir K. and Proudfit H. K. (1997) Microinjection of the GABAA antagonist bicuculline in the A7 catecholamine cell group produces antinociception. Soc. Neurosci. Abstr. 23, 1814. 49. Nuseir K. and Proudfit H. K. (1999) Modulation of nociception by GABA neurons that tonically inhibit noradrenergic neurons in the A7 cell group. Soc. Neurosci. Abstr. 25, 1675. 50. Olsen R. W., Ban M., Miller T. and Johnston G. A. (1975) Chemical instability of the GABA antagonist bicuculline under physiological conditions. Brain Res. 98, 383–387. 51. Ossipov M. H. and Gebhart G. F. (1984) Light pentobarbital anesthesia diminishes the antinociceptive potency of morphine administered intracranially but not intrathecally in the rat. Eur. J. Pharmac. 97, 137–140. 52. Pan Z. Z. and Fields H. L. (1996) Endogenous opioid-mediated inhibition of putative pain-modulating neurons in rat rostral ventromedial medulla. Neuroscience 74, 855–862. 53. Pan Z. Z., Williams J. T. and Osborne P. B. (1990) Opioid actions on single nucleus raphe magnus neurons from rat and guinea-pig in vitro. J. Physiol., Lond. 427, 519–532. 54. Proudfit H. K. (1980) Reversible inactivation of raphe magnus neurons: effects on nociceptive threshold and morphine-induced antinociception. Brain Res. 201, 459–464. 55. Proudfit H. K. (1992) The behavioural pharmacology of the noradrenergic descending system. In Toward the Use of Noradrenergic Agonists for the Treatment of Pain (ed. Besson J.-M.), pp. 119–136. Elsevier, Amsterdam. 56. Reddy S. V. R., Maderdrut J. L. and Yaksh T. L. (1980) Spinal cord pharmacology of adrenergic agonist-mediated antinociception. J. Pharmac. exp. Ther. 213, 525–533. 57. Renno W. M., Mullett M. A. and Beitz A. J. (1992) Systemic morphine reduces GABA release in the lateral but not the medial portion of the midbrain periaqueductal gray of the rat. Brain Res. 594, 221–232. 58. Retz K. C. and Holaday L. M. (1986) Analgesia and motor activity following administration of THIP into the periaqueductal gray and lateral ventricle of rats. Drug Dev. Res. 9, 133–142. 59. Rosin D., Zeng D., Stornetta R., Norton F., Riley T., Okusa M., Guyenet P. and Lynch K. (1993) Immunohistochemical localization of a2A-adrenergic receptors in catecholaminergic and other brainstem neurons in the rat. Neuroscience 56, 139–155. 60. Roychowdhury S. M. and Fields H. L. (1996) Endogenous opioids acting at a medullary mu-opioid receptor contribute to the behavioral antinociception produced by GABA antagonism in the midbrain periaqueductal gray. Neuroscience 74, 863–872. 61. Sandku¨hler J. and Gebhart G. F. (1991) Production of reversible local blockage of neuronal function. In Methods in Neurosciences (ed. Conn P. M.), Vol. 7, pp. 122–138. Academic, New York. 62. Sandku¨hler J., Maisch B. and Zimmermann M. (1987) The use of local anesthetic microinjections to identify central pathways: a quantitative evaluation of the time course and extent of the neuronal block. Expl Brain Res. 68, 168–178. 63. Sandku¨hler J., Willmann E. and Fu Q. G. (1989) Blockade of GABAA receptors in the midbrain periaqueductal gray abolishes nociceptive spinal dorsal horn neuronal activity. Eur. J. Pharmac. 160, 163–166. 64. Shen K.-Z., North R. A. and Surprenant A. (1992) Potassium channels opened by noradrenaline and other transmitters in excised membrane patches of guinea-pig submucosal neurones. J. Physiol. 445, 581–599. 65. Stiller C.-O., Bergquist J., Beck O., Ekman R. and Brodin E. (1996) Local administration of morphine decreases the extracellular level of GABA in the periaqueductal gray matter of freely moving rats. Neurosci. Lett. 209, 165–168. 66. Stone L. S., Broberger C., Vulchanova L., Wilcox G. L., Ho¨kfelt T., Riedl M. and Elde R. (1998) Differential distribution of a2A and a2C adrenergic receptor immunoreactivity in the rat spinal cord. J. Neurosci. 18, 5928–5937. 67. Surprenant A., Horstman D. A., Akabarali H. and Limbird L. E. (1992) A point mutation of the a2-adrenoceptor that blocks coupling to potassium but not calcium currents. Science 257, 977–980. 68. Surprenant A. and North R. A. (1988) Mechanism of synaptic inhibition by noradrenaline acting at alpha-2 receptors. Proc. R. Soc. Lond. B234, 84–114. 69. Talley E. M., Rosin D. L., Lee A., Guyenet P. G. and Lynch K. R. (1996) Distribution of alpha2A-adrenergic receptor-like immunoreactivity in the rat central nervous system. J. comp. Neurol. 372, 111–134. 70. Todd A. J. and Millar J. (1983) Receptive fields and responses to iontophoretically applied noradrenaline and 5-hydroxytryptamine of units recorded in laminae I–III of cat dorsal horn. Brain Res. 288, 159–167. 71. Weight F. F. and Salmoiraghi G. C. (1966) Responses of spinal cord interneurons to acetylcholine, norepinephrine and serotonin administered by microelectrophoresis. J. Pharmac. exp. Ther. 153, 420–427. 72. West W. L., Yeomans D. C. and Proudfit H. K. (1993) The function of noradrenergic neurons in mediating antinociception induced by electrical stimulation of the locus coeruleus in two different sources of Sprague–Dawley rats. Brain Res. 626, 127–135. 73. Westlund K. N., Carlton S. M., Zhang D. and Willis W. D. (1990) Direct catecholaminergic innervation of primate spinothalamic tract neurons. J. comp. Neurol. 299, 178–186. 74. Westlund K. N., Zhang D., Carlton S. M., Sorkin L. S. and Willis W. D. (1991) Noradrenergic innervation of somatosensory thalamus and spinal cord. Prog. Brain Res. 88, 77–88. 75. White S. R., Fung S. J. and Barnes C. D. (1991) Norepinephrine effects on spinal motoneurons. Prog. Brain Res. 88, 343–350. 76. White S. R. and Neuman R. S. (1980) Facilitation of spinal motoneurone excitability by 5-hydroxytryptamine and noradrenaline. Brain Res. 188, 119–127. 77. Winzer-Serhan U. H., Raymon H. K., Broide R. S., Chen Y. and Leslie F. M. (1997) Expression of a2 adrenoceptors during rat brain development-1. a2A messenger RNA expression. Neuroscience 76, 241–260. 78. Yeomans D. C., Clark F. M., Paice J. A. and Proudfit H. K. (1992) Antinociception induced by electrical stimulation of spinally-projecting noradrenergic neurons in the A7 catecholamine cell group of the rat. Pain 48, 449–461. 79. Yeomans D. C. and Proudfit H. K. (1992) Antinociception induced by microinjection of substance P into the A7 catecholamine cell group in the rat. Neuroscience 49, 681–691. 80. Yeomans D. C. and Proudfit H. K. (1994) Characterization of the foot withdrawal response to noxious radiant heat in the rat. Pain 59, 85–94. 81. Zeng D. and Lynch K. R. (1991) Distribution of alpha2-adrenergic receptor mRNAs in the rat CNS. Molec. Brain Res. 10, 219–225.

Pontine GABA neurons modulate nociception

783

82. Zhuo M. and Gebhart G. F. (1990) Characterization of descending inhibition and facilitation from the nucleus reticularis gigantocellularis and gigantocellularis pars alpha in the rat. Pain 42, 337–350. 83. Zhuo M. and Gebhart G. F. (1990) Spinal cholinergic and monoaminergic receptors mediate descending inhibition from the nuclei reticularis gigantocellularis and gigantocellularis pars alpha in the rat. Brain Res. 535, 67–78. 84. Zhuo M. and Gebhart G. F. (1991) Spinal serotonin receptors mediate descending facilitation of a nociceptive reflex from the nuclei reticularis gigantocellularis and gigantocellularis pars alpha in the rat. Brain Res. 550, 35–48. 85. Zhuo M. and Gebhart G. F. (1992) Characterization of descending facilitation and inhibition of spinal nociceptive transmission from the nuclei reticularis gigantocellularis and gigantocellularis pars alpha in the rat. J. Neurophysiol. 67, 1599–1614. (Accepted 16 December 1999)