Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The rostral ventrolateral medulla (RVLM), which includes the nucleus reticularis gigantocellularis (NRGC)/nucleus reticularis gigantocellularis  (NRGC) and the lateral reticular nucleus, regulates part of the physiological function of antinociceptive and cardiovascular control via the cholinergic system (Ossipov and Gebhart, 1986; Zhuo and Gebhart, 1992; Kubo et al., 1997). Anatomical evidence has demonstrated that the pedunculopontine tegmental nucleus in the brainstem is a major source of cholinergic afferents to the RVLM (Mitani et al., 1988; Yasui et al., 1990; Sherriff and Henderson, 1994). In addition, other studies have shown the existence of a descending cholinergic system from the RVLM to the spinal cord and the presence of small to medium-sized cholinergic neurons and choline acetyltransferase mRNA in small cells of the NRGC/nucleus reticularis paragigantocellularis (NRPG) (Bowker et al., 1983; Jones et al., 1986; Tago et al., 1989; Ruggiero et al., 1990; Lauterborn et al., 1993). These observations suggest that the cholinergic system plays an important role in the RVLM.

The interaction of morphine with cholinergic neurons in the central nervous system is well known. Using in vivo microdialysis, morphine has been shown to depress the release of acetylcholine (ACh) in the brain of rat (Rada et al., 1991; Taguchi et al., 1993). On the other hand, also using in vivo microdialysis, i.v. administration of morphine has been shown to produce a dose-dependent increase in ACh in human cerebrospinal fluid and sheep dorsal horn (Bouaziz et al., 1996). Furthermore, the cholinergic system in the central nervous system is considered part of an endogenous pain control system, activation of which can produce antinociception and analgesia in a variety of animals including humans (Christensen and Gross, 1948; Gillberg et al., 1990; Zhuo and Gebhart, 1990, 1991; Iwamoto and Marion, 1993; Hood et al., 1995). Pharmacological studies provide evidence that acetylcholine or carbachol administered into particular brainstem nuclei can produce pronounced antinociception that is reversed by muscarinic antagonists (Brodie and Proudfit, 1984; Yaksh et al., 1985). In addition, there is evidence that the descending cholinergic system, spinal cholinergic receptors, and anticholinesterase are involved in the mechanisms of opioid analgesia (Dirksen and Nijhuis, 1983; Naguib and Yaksh, 1994; Fang and Proudfit, 1996; Hood et al., 1997). For example, the antinociception produced by systemic administration of morphine is antagonized by intrathecal administration of the cholinergic muscarinic receptor antagonist, atropine (Chiang and Zhuo, 1989). Moreover, various combinations of morphine with anticholinesterase have been demonstrated to result in an increased antinociceptive effect (Clement and Copeman, 1986; Green and Kitchen, 1986; Beilin et al., 1997). These findings suggest that supraspinally or spinally released ACh may play a role in analgesia produced by systemic morphine. This is consistent with our notion that morphine analgesia may be caused, at least in part, by enhancement of ACh release within the RVLM.

The purpose of the present study was to examine the ACh release induced by systemic administration or local application of morphine in the RVLM and the antinociceptive effect of perfusion morphine through a microdialysis probe using the responses of both the hot-plate test and tail immersion test in freely moving rats.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Male adult Wistar rats (300-350 g) were anesthetized with pentobarbital sodium (50 mg/kg i.p.) and positioned in a stereotaxic apparatus. In each rat, the skull was exposed, and a hole was drilled for a microdialysis probe (CMA 10; Carnegie Medicin AB; diameter, 0.5 mm; dialysis membrane, 2.0 mm), which was implanted into the NRGC (bregma, 11.5 mm; lateral, 0.9 mm; ventral, 10.5 mm; Paxinos and Watson, 1986). The microdialysis probe was held firmly in place by dental acrylic and anchored to the skull using stainless steel screws. All experiments were performed 48 h after surgery.

In Vivo Microdialysis. The animals were placed in a Plexiglas cage (30 × 30 × 38 cm) and were connected by polyethylene inflow and outflow tubes to a syringe pump and collection vials. The perfusate was collected at 15-min intervals. ACh was measured by HPLC using electrochemical detection, as described previously (Taguchi et al., 1993). The HPLC-electrochemical detection system included a pump (PM-60; BAS, Lafayette, IN), guard, and chromatographic column (5 × 4 mm and 2 × 110 mm; BAS) and electrochemical detection (LC-4B; BAS). The mobile phase (pH 8.4) consisted of 50 mM Na₂HPO₄, 0.5 mM EDTA 2 Na, and 0.45 mM sodium octanesulfonate. The applied potential at the working electrode was +450 mV (versus Ag/AgCl). Both the chromatographic column and enzyme reactor column were maintained at 37°C using a column heater (LC-22A; BAS).

Nociceptive Test. The hot-plate test and tail immersion test were performed in conscious rats to assess the effects of local application of morphine in the RVLM. The hot-plate test was performed by placing a rat on an aluminum plate maintained at 55°C (Muromachi kikai Co. Ltd.). The latency to a nociceptive response, identified as either licking of a hindpaw or attempts to escape by jumping from the heated surface, was measured. Nonresponding rats were removed from the heated surface at 30 s to prevent tissue damage.

5

4

5

4

Histological Procedures. At the end of the experiments, the animals were sacrificed by an overdose of pentobarbital-sodium. The brain was fixed in 10% formaldehyde solution (Formalin), and frozen 60-µm-thick sections were cut using a freezing microtome. The tracks of the dialysis cannula were verified microscopically in histological sections.

Data Analysis. Dialysis data are shown as the means ± S.E.M. of the percentage of baseline level obtained from each rat before drug treatment. Hot-plate latency and tail immersion latency times were converted to the percentage of maximum possible effect according to the following formula: % Maximum possible effect = [(Postdrug latency)  (Predrug latency)]/[(Maximum latency)  (Predrug latency)] × 100. Data were analyzed by repeated measurement two-way ANOVA, followed by Tukey-Kramer honest significant difference test. Differences were considered significant when P values were < .05.

Drugs. The drugs used were morphine hydrochloride (Sankyo), tetrodotoxin (Sigma Chemical, St. Louis, MO), and naloxone hydrochloride (Endo Laboratories). Morphine and naloxone were dissolved in sterile saline. All solutions were sterilized by filtering through a Millipore filter (0.2 µm). The control group of animals was administered physiological saline only.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Mapping of Medullary Spontaneous ACh Release. The amount of extracellular ACh recovered from the RVLM with chronically implanted microdialysis probes was 303.3 ± 28.2 fmol/20-µl sample. Basal release of ACh was stable over 3 h after the beginning of perfusion with perfusion solution containing 10 µM neostigmine (Fig. 1). For the mapping studies, Fig. 1 shows the location of the microdialysis membrane at the tips of the probes. The release of ACh was registered at points 1.0 to 2.0 mm distant in the vertical, medial, dorsal, and rostral directions from the center of the NRGC within the RVLM (Fig. 1). The average concentrations of ACh in the five regions are shown in Fig. 1. The basal ACh concentration in the RVLM including the NRGC was approximately two to six times higher than that collected from the caudal (1; n = 4), medial (2; n = 6), vertical (3; n = 6), and rostral (5; n = 5) medulla regions (Fig. 1).

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Locations of the microdialysis probe in the rostral ventrolateral medulla and mean quantity of ACh (in femtomoles) recovered in 20-µl samples from brain sites (1-5). The values are basal release of ACh in the rostral ventrolateral medulla and not following systemic administration of morphine. Sites are illustrated on representative coronal brainstem sections. The rectangles show the estimated positions of the tips of the dialysis membranes. NRM, nucleus raphe magnus, Pyr, pyramidal tract; sp5, spinal trigeminal tract (Paxinos and Watson, 1986).

Effects of Tetrodotoxin and Calcium-Free Perfusion Solution on ACh Release. To demonstrate that the ACh detected in the RVLM dialysate originated from neuronal terminals, we examined whether release was dependent on sodium channel- or calcium channel-dependent mechanisms and cellular depolarization. We first assessed the effects of the voltage-dependent sodium channel blocker TTX on ACh release in the RVLM. Perfusion solution containing 1.0 µM TTX was infused for 30 min. TTX significantly decreased the output of ACh in the RVLM by 45.2 ± 2.2% at 15 min and by 62.2 ± 1.7% at 30 min [F_(1,120) = 93.18, P < .001; n = 5]. The effect of ACh release returned to baseline levels 45 min after the removal of TTX (Fig. 2A).

View larger version (23K): [in this window] [in a new window]   Fig. 2.   A, effects of TTX perfusion on ACh release in the RVLM. TTX (1 µM) significantly decreased the release of ACh (n = 5). The shaded area depicts the time of TTX perfusion through the microdialysis probe. Each point represents the mean ± S.E.M. B, effects of calcium-free perfusion solution on ACh release into the RVLM. Perfusion of calcium-free solution (shaded area) significantly decreased the release of ACh (n = 5). Each point represents the mean ± S.E.M. C, effects of potassium depolarization on the ACh release in the RVLM. A high concentration of potassium ions (30 mM) in the solution (shaded area) significantly increased ACh release (n = 5). Each point represents the mean ± S.E.M.

Potassium Depolarization. We next assessed the extent to which depolarizing concentrations of high potassium ion (30 mM for 30 min) affect the extracellular concentration of ACh in the RVLM. As shown in Fig. 2C, high potassium rapidly increased the release of ACh by 33.5 ± 1.4% at 15 min and by 37.2 ± 3.8% at 30 min [F_(1,120) = 12.92, P < .001; n = 5]. ACh release returned to predrug levels within 30 min after termination of this treatment.

Effects of Systemic-Administered Morphine on the Release of ACh in the RVLM. Figure 3 shows the effects of morphine (5 and 10 mg/kg i.p.) on the release of ACh in the RVLM. At a dose of 5 mg/kg, morphine induced a significant increase in ACh release (42.4 ± 7.2% at 75 min) compared with the saline group [F_(1,170) = 107.79, P < .001; n = 6] (Fig. 3). Recovery was observed 175 min later. At a dose of 10 mg/kg, morphine significantly increased the release of ACh by 39.1 ± 8.1% at 30 min, with a peak of 86.4 ± 4.9% at 75 min [F_(1,170) = 314.75, P < .001; n = 6]. Naloxone (1.0 mg/kg i.p.; n = 5), administered i.p. 30 min before administration of morphine (5 mg/kg), attenuated the morphine-induced increase in ACh release in the RVLM [F_(1,170) = 135.01, P < .001; n = 6] (Fig. 3).

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Effects of systemic morphine on extracellular ACh release in the RVLM. Morphine (5 and 10 mg/kg i.p.) significantly increased the release of ACh. Pretreatment with naloxone (1 mg/kg i.p.), administered 30 min before morphine (5 mg/kg), antagonized the morphine-induced increase in ACh release in the RVLM. Administration of morphine and naloxone was performed at the times indicated by arrows. Each point represents the mean ± S.E.M.

Effects of Locally Applied Morphine on the Release of ACh in the RVLM. The effects of local application of morphine (10⁵ and 10⁴ M) on ACh release into the RVLM is shown in Fig. 4. Morphine (10⁵ M) had no effect on the release of ACh in the RVLM [F_(1,170) = 1.39, P = .241; n = 5]. Local application of morphine (10⁴ M) significantly increased the release of ACh by 52.9 ± 8.2% at 30 min, with a peak of 85.8 ± 2.5% at 45 min [F_(1,170) = 46.59, P < .001; n = 5]. A significant increase in ACh release was observed between 30 and 60 min during local application of morphine; this increase recovered after drug removal. Pretreatment with systemic administration of naloxone (1.0 mg/kg i.p.) 30 min before local application of morphine (10⁴ M) significantly attenuated the morphine-induced increase in ACh release in the RVLM [F_(1,170) = 61.67, P < .001; n = 5] (Fig. 4).

View larger version (48K): [in this window] [in a new window]   Fig. 4.   Effects of local application of morphine on ACh release in the RVLM. Morphine (105 M) infusion via a dialysis probe did not affect the release of ACh in the RVLM (n = 5). Morphine (104 M) significantly increased the release of ACh (n = 5). Pretreatment with naloxone (1.0 mg/kg i.p.), administered 30 min before perfusion of morphine (104 M), antagonized morphine-induced increase in ACh release. The shaded area depicts the time of morphine perfusion (60 min) via the microdialysis probe. Administration of naloxone was performed at the times indicated by arrows. Each point represents the mean ± S.E.M.

Effects of Local Application of Morphine into the RVLM on the Hot-Plate Response. Figure 5A shows the effects of morphine (10⁵ and 10⁴ M) on the hot-plate nociceptive response. The hot-plate control latency was 8.7 ± 0.3 s. At a concentration of 10⁵ M, morphine did not affect the hot-plate response compared with the control group [F_(1,54) = 2.03, P = .160; n = 4] (Fig. 5A). At a concentration of 10⁴ M, morphine significantly increased the hot-plate response by 50.6 ± 6.1% at 60 min [F_(1,54) = 83.71, P < .001; n = 4]. A significant increase in the hot-plate response was observed between 30 and 90 min. Systemic administration of naloxone (1.0 mg/kg i.p.), 30 min before local application of morphine (10⁴ M), completely blocked the morphine-induced antinociceptive response [F_(1,54) = 29.68, P < .001; n = 4] (Fig. 5A).

View larger version (42K): [in this window] [in a new window]   Fig. 5.   A, time course of the effects on the hot-plate response of local application of morphine into the RVLM. Morphine at 105 M did not affect the hot-plate response, whereas morphine at 104 M significantly increased the hot-plate response. Antinociception was produced by local application of morphine (104 M) into the RVLM. Systemic administration of naloxone (1.0 mg/kg i.p.), 30 min before local application of morphine (104 M), completely blocked the morphine-induced antinociceptive response. B, time course of the effects on the tail-immersion withdrawal response of local application of morphine into the RVLM. Morphine at 105 M did not affect the tail immersion response, whereas morphine at 104 M significantly increased the tail immersion response. The shaded area depicts the time of morphine perfusion (60 min) via the microdialysis probe. Systemic administration of naloxone (1.0 mg/kg i.p.), 30 min before local application of morphine (104 M), completely blocked the morphine-induced antinociceptive response. Each point represents the mean ± S.E.M.

Effects of Local Application of Morphine into the RVLM on the Tail Immersion Response. Figure 5B shows the effects of morphine (10⁵ and 10⁴ M) on the tail immersion nociceptive response. The tail immersion withdrawal control latency was 1.9 ± 0.1 s. At a concentration of 10⁵ M, morphine did not affect the withdrawal latencies compared with the control group [F_(1,54) = 3.01, P = .089; n = 4] (Fig. 5B). At a concentration of 10⁴ M, morphine significantly increased the withdrawal latencies by 58.5 ± 5.0% at 60 min [F_(1,54) = 220.14, P < .001; n = 4]. A significant increase in the tail immersion response was observed between 30 and 90 min. Naloxone (1.0 mg/kg i.p.), 30 min before local application of morphine (10⁴ M), antagonized the morphine-induced antinociceptive response [F_(1,54) = 119.26, P < .001; n = 4] (Fig. 5B).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present findings demonstrate that the basal ACh concentration in the RVLM, including the NRGC, is higher than that in four medulla regions as assessed using in vivo microdialysis. Anatomical studies have shown that a dense pleux of retrogradely labeled cholinergic fibers and terminals exists in the gigantocellular reticular field following injection of Fluoro-gold into the pedunculopontine tegmental nucleus (Yasui et al., 1990). Large- and giant-sized cells in the NRGC/NRGC are retrogradely labeled by wheat germ agglutinin-horseradish peroxidase from the spinal cord and stain positively for acetylcholinesterase (Bowker et al., 1983). Jones et al. (1986) found that most neurons labeled following injection of [³H]choline into the upper cervical spinal cord of rat were concentrated in the more ventral part of the NRGC and NRGC. In addition, in situ hybridization studies have found the presence of choline acetyltransferase mRNA in small cells in the NRGC (Lauterborn et al., 1993). The present study's mapping of the medulla showed that the cholinergic terminal lies in the RVLM. In combination, these results support the suggestion that ACh plays an important role in the RVLM, including the NRGC.

We found that perfusion, via a dialysis probe, of solution containing TTX (a voltage-dependent Na⁺ channels blocker) decreased output of ACh within the RVLM, similar to our previous report in the striatum (Taguchi et al., 1993). Similarly, removal of calcium ions produced a decrease in ACh output in the RVLM. In addition, potassium depolarization produced an increase in ACh output. Thus, the dialysate ACh collected from the RVLM resulted from spontaneous, sodium channel- and calcium channel-dependent neuronal release.

In the present study, i.p. administration of morphine increased the release of ACh in the RVLM. In addition, local application of morphine was shown to enhance ACh release in the RVLM by in vivo microdialysis. Chiang and Zhuo (1989) have suggested that a descending cholinergic system is involved in antinociception produced following the systemic administration of morphine. In addition, microdialysis experiments suggest that increased ACh levels in cerebrospinal fluid result from systemic morphine-induced activation of bulbospinal pathways (Bouaziz et al., 1996). Recently, systemic morphine increased ACh and norepinephrine in dorsal horn microdialysates, and these increases were attenuated by naloxone or cervical spinal cord transection (Xu et al., 1997). These observations lead to the conclusion that morphine activates descending cholinergic neurons from the brainstem. In contrast, cat spinal cord transection does not reduce the amount of choline acetyltransferase in the spinal cord (Kanazawa et al., 1979). The antinociceptive effect of morphine microinjected into the ventrolateral periaqueductal gray is reduced by intrathecal administration of atropine (Fang and Proudfit, 1996). This antinociceptive effect of microinjected morphine into the ventrolateral periaqueductal gray is considered to be the result of activation of the spinal cholinergic muscarinic receptors. Thus, there is no direct evidence that activation of the pathway from the brainstem including the RVLM to the spinal cord may induce antinociception that is mediated by the descending cholinergic system. Consequently, morphine may act on both spinal cholinergic neurons and descending cholinergic neurons from the brainstem to cause ACh release. The present data demonstrate that systemic administration of morphine increases the concentration of ACh in the extracellular space in the RVLM. In addition, morphine was applied via a dialysis membrane directly into the RVLM with a resultant increase in ACh release. The release of ACh induced by systemic administration and local application of morphine was attenuated by naloxone, an antagonist for the opiate receptor. Our observations suggest that the increase in ACh release induced by morphine activated the cholinergic neurons via an opiate receptor located in the RVLM. However, an important question to address is whether the effect of morphine in increasing ACh release is mediated directly by cholinergic terminals or cholinergic cell bodies. In our experiment, the microdialysis probe measured the release of ACh from cholinergic terminals within the RVLM region of the brainstem. Therefore, the present results encourage further experiments designed to determine whether the actions of morphine are exerted on the cholinergic terminals or the cholinergic cell bodies within the RVLM. Furthermore, the terminals responsible for ACh release within the RVLM may arise from cholinergic interneurons intrinsic to the region or from other nuclei-containing cholinergic neurons such as pedunculopontine tegmental nucleus. The exact origin of the cholinergic innervation responsible for the morphine-induced ACh release in the RVLM remains to be determined.

With respect to the antinociceptive effect of morphine, microinjection of morphine or glutamate into, and electrical stimulation of, the RVLM including the NRGC/NRGC has been shown to inhibit the spinal nociceptive reflex and spinal dorsal horn neuron response to peripheral stimulation (Satoh et al., 1979; Azami et al., 1982; Sandkuhler and Gebhart, 1984; Zhuo and Gebhart, 1990, 1997). Furthermore, central administration of morphine inhibits the tail-flick reflex and response to the hot-plate test, suggesting that morphine induces activation of the descending inhibitory system, primarily the serotonergic and noradrenergic pathways (Satoh et al., 1983; Tseng and Tang, 1989). In the present experiments, local application of morphine into the RVLM via a microdialysis probe produced increases in the hot-plate response and tail immersion withdrawal response. These antinociceptive effects were reduced by naloxone. These findings support the notion that the antinociceptive effects of morphine are mediated by release of ACh in the RVLM. Cholinergic activation in the RVLM is important in the relay of the descending inhibitory system from midbrain stem to spinal cord. Therefore, the morphine-induced release of ACh in the RVLM most likely results in activation of supraspinal antinociceptive mechanisms as well as the descending inhibitory system. Recently, systemic or intrathecal administration of cholinergic drugs, particularly anticholinesterase, has been shown to enhance antinociception of opioid administration in animals including humans (Beilin et al., 1997; Hood et al., 1997). In addition, electrical stimulation or glutamate microinjection in the NRGC and NRGC produce antinociception that is attenuated by intrathecal administration of atropine (Zhuo and Gebhart, 1990). Thus, the activation of the cholinergic system in the RVLM is, in part, involved in antinociception. However, the possibility that descending monoaminergic neurons modulate the antinociceptive response to noxious stimulation cannot be excluded, as intrathecal injection of either serotonergic or noradrenergic agonists can reduce the nociceptive responses of the tail-flick test and hot-plate test (Jensen and Yaksh, 1984; Fang and Proudfit, 1996). These observations lead to the suggestion that local application of morphine via a microdialysis probe into the RVLM causes the activation of a descending serotonergic system and a descending noradrenergic system.

In conclusion, our results show that antinociceptive morphine treatments increase the release of acetylcholine in the RVLM. Because other studies have indirectly suggested a role for cholinergic systems in morphine analgesia, these studies support the hypothesis that activation of brain stem cholinergic mechanisms plays an important role in morphine antinociception in the rat.