Spinal Endogenous Acetylcholine Contributes to the Analgesic Effect of Systemic Morphine in Rats

Male rats (Harlan Sprague-Dawley, Indianapolis, IN) weighing 250–275 g were used in this study. The surgical preparations and experimental protocols were approved by the Animal Care and Use Committee of the Pennsylvania State University College of Medicine (Hershey, PA). The rats were anesthetized with 2% halothane, and the right jugular vein was cannulated with PE-50 tubing. Intrathecal catheters (PE-10 tubing) were inserted through an incision in the cisternal membrane and advanced 8 cm caudal so that the tip of each catheter was positioned at the lumbar spinal level. Both intravenous and intrathecal catheters were externalized to the back of the neck and sutured to the musculature and skin at the incision site. The rats were used for analgesic studies after recovery for 4 or 5 days after cannulation.

To quantitatively assess the nociceptive threshold of the hind paw, rats were placed on the glass surface of a thermal plantar testing apparatus (Model 336; IITC Inc./Life Science Instruments, Woodland Hills, CA). The rats were allowed to acclimate for 30 min before testing. The temperature of the glass surface was maintained constant at 30°C. A mobile radiant heat source located under the glass was focused onto the hind paw of the rats. The paw-withdrawal latency was recorded by a digital timer. The withdrawal latencies for the left and right paws were averaged, and the mean value was used to indicate the sensitivity to noxious heat stimulation. 15The apparatus was adjusted at the beginning of the study in six separate rats so that the baseline paw-withdrawal latency was approximately 10 s. This setting (i.e.,  the light beam intensity) was kept unchanged for the remainder of the study. The cut-off of 30 s was used to prevent potential tissue damage. 15 

Motor function was evaluated by the placing or stepping reflex and the righting reflex. 18The former was evoked by drawing the dorsum of either hind paw across the edge of the table. The latter was assessed by placing the rat horizontally with its back on the table, which normally gives rise to an immediate, coordinated twisting of the body to an upright position. Changes in motor function was scored as follows: 0, normal; 1, slight deficit; 2, moderate deficit; and 3, severe deficit.

After the baseline was measured, three groups of rats were treated intrathecally with 30 μg atropine (a specific muscarinic receptor antagonist, n = 7), 50 μg mecamylamine (a specific nicotinic receptor antagonist, n = 6), or the vehicle (saline, n = 6). The maximal effective doses of these antagonists upon intrathecal injection in rats have been systematically determined in our previous studies. 12Fifteen minutes later, rats received intravenous injection of 2.5 mg/kg morphine, and the paw-withdrawal latency was determined every 10 min for 60 min after morphine injection. This dose of morphine was selected based on our previous study showing that it produces 60% maximal possible effect in rats. 15To assess the possibility that blockade of both muscarinic and nicotinic receptors may attenuate the antinociceptive action of morphine to a greater extent, we also tested the effect of intravenous morphine in six other rats 15 min after intrathecal injection of both 30 μg atropine and 50 μg mecamylamine. Drugs for intrathecal injections were dissolved in normal saline and administered in a volume of 5 μl followed by a 10-μl flush with normal saline. Atropine and mecamylamine were obtained from Sigma Chemical Company, St. Louis, MO. Morphine was supplied by Astra Pharmaceutical, Westborough, MA.

To further demonstrate the role of spinal endogenous acetylcholine in the effect of morphine on antinociception, separate rats were injected intrathecally with either 5 nm ethylcholine mustard aziridinium ion (AF64A; n = 6) or its vehicle (n = 6). This dose of AF64A has been shown to effectively reduce the acetylcholine content in the dorsal spinal cord in rats. 19AF64A is a specific cholinergic neurotoxin that disrupts the high-affinity choline transporter (HAChT). 20Because the HAChT is a rate-limiting step for the synthesis of acetylcholine, 21AF64A is capable of inducing irreversible inhibition of acetylcholine synthesis. AF64A was prepared according to the method described previously. 22Briefly, acetylethylcholine mustard HCl (RBI, Natick, MA) was diluted in distilled water and brought to pH 11.5 for 30 min with 1 N NaOH. The AF64A was formed by decreasing the pH to 7.0 with 0.1 N HCl and then adjusting to pH 7.4 with NaHCO³. The cyclized solution was kept at 4°C and injected within 2 h. The vehicle control solution was distilled water to which an equivalent amount of 1 N NaOH was added, followed by pH adjustment with HCl and NaHCO³, as with the AF64A solution. One week after intrathecal AF64A treatment, the analgesic effect produced by intravenous injection of 2.5 mg/kg morphine was tested. In addition, 1 μg hemicholinium-3 (HC-3; Sigma) dissolved in saline, a reversible inhibitor of choline transporter, 23,24was injected intrathecally in six other rats. Three hours later, the antinociceptive effect of 2.5 mg/kg intravenous morphine in these rats was tested, as described earlier.

To estimate the effect of intrathecal AF64A on HAChT, the spinal cords of AF64A- and vehicle-treated rats (n = 6 in each group) were rapidly removed under 2–3% halothane anesthesia 1 week after intrathecal treatment. The dorsal halves of the spinal cord were dissected and used for the [³H]HC-3 binding experiments to confirm the neurotoxic effect of AF64A on the dorsal spinal HAChT. The HC-3 binds specifically to the cholinergic transporter, and the [³H]HC-3 binding has been used as a marker of cholinergic neuronal activity and of the integrity of cholinergic terminals. 25The [³H]HC-3 binding experiments were performed using membrane preparations. 26Briefly, an aliquot of membrane was added to the tubes containing 50 mm glycylglycine buffer, 200 mm NaCl, and 10 nm [³H]HC-3 (NEN, Boston, MA). After 30 min of incubation at 25°C, the reaction was terminated by adding ice-cold Tris-HCl buffer (pH 7.4) followed by rapid filtration using a Brandel Harvester (Model M48, Gaithersburg, MD). Radioactivity on Whatman GF/B filters was determined in a liquid scintillation counter (Model LS 6500; Beckman Coulter, Inc., Fullerton, CA). The nonspecific binding was determined by adding 10 μm HC-3 to the reaction mixture.

Data are presented as mean ± SEM. The effects of atropine and mecamylamine on morphine-produced antinociceptive action and the difference between treatment with AF64A/HC-3 and the vehicle on the effect of morphine were determined by repeated-measures analysis of variance followed by Tukey post hoc  test. The saturation binding data were processed using nonlinear regression analysis (Prism; GraphPad Software, San Diego, CA) to calculate maximal specific binding (B^max) and dissociation constant (K^D). Differences in B^maxand K^Dbetween the control and AF64A-treated groups were compared by the Student paired t  test. P < 0.05 was considered to be statistically significant.

The withdrawal latency of the hind paw in response to the radiant heat stimulation before morphine injection was 10.3 ± 0.6 s (n = 6, saline group). The paw-withdrawal latency increased significantly (25.6 ± 1.4 s, P < 0.05) 10 min after intravenous injection of 2.5 mg/kg morphine in rats being given intrathecal saline (fig. 1). The effect of intravenous morphine on the paw-withdrawal latency lasted for 40–50 min. Intrathecal pretreatment with 30 μg atropine or 50 μg mecamylamine attenuated significantly the effect of morphine (fig. 1). Furthermore, the inhibitory effect of atropine on the effect of morphine was significantly greater than that of mecamylamine (fig. 1). In rats subjected to intrathecal injection of a combination of atropine and mecamylamine, the antinociceptive effect produced by subsequent intravenous injection of 2.5 mg/kg morphine was similar to that in rats pretreated with intrathecal atropine. Intrathecal administration of atropine or mecamylamine was not associated with any overt behavioral or motor function changes, which were assessed by testing the animals’ ability to stand and ambulate in a normal posture and to place and step with the hind paw. 18 

The baseline withdrawal threshold in response to the radiant heat stimulus was not significantly altered by intrathecal injection of AF64A or HC-3, compared to the vehicle group (fig. 2). The motor function, based on the placing/stepping reflex and the righting reflex, appeared normal in AF64A- and HC-3-treated rats. In rats subjected to intrathecal injection of AF64A, the effect of morphine on the paw-withdrawal latency was attenuated significantly, compared to that in rats pretreated with intrathecal vehicle (fig. 2). Intrathecal pretreatment with HC-3 also significantly reduced the effect of intravenous morphine to the level similar to that observed in the AF64A treated group (fig. 2).

Scatchard plot transformed from the saturation binding data demonstrates the effect of intrathecal AF64A on the B^maxand K^Dof [³H]HC-3 bindings in the dorsal spinal cord (fig. 3). Compared with the vehicle-treated group (B^max= 102.5 ± 5.58 fmol/mg protein; K^D= 8.86 ± 1.03 nm), intrathecal AF64A significantly re-duced the B^max(76.2 ± 3.86 fmol/mg protein) of the [³H]HC-3 binding without an apparent effect on its K^D(7.91 ± 0.73 nm) in the dorsal spinal cord.

In the present study, we examined the role of spinal endogenous acetylcholine and cholinergic receptors in the analgesic effect produced by intravenous morphine in rats. A previous study has reported that intrathecal treatment with atropine reduces the analgesic effect of intraperitoneal morphine. 14In the present study, we have provided new evidence that pretreatment with muscarinic or nicotinic receptor antagonists significantly attenuates the analgesic action of intravenous morphine. Another important finding of the current study is that reduction of endogenous spinal acetylcholine through inhibition of HAChT with AF64A or HC-3 significantly reduces the analgesic effect of morphine. Therefore, this study provides important information that endogenous acetylcholine in the spinal cord mediates the analgesic action of intravenous morphine through both spinal muscarinic and nicotinic receptors.

Systemic morphine causes spinal acetylcholine release in animals and humans. 5,6Also, intrathecal administration of cholinergic receptor agonists or cholinesterase inhibitors produces analgesia. 8,10,27Thus, we reasoned that endogenously released acetylcholine in the spinal cord is involved in morphine analgesia. A unique tool to study the role of endogenous acetylcholine in the spinal cord is the neurotoxin AF64A. The neurotoxin AF64A is an analog of choline and contains a highly reactive aziridinium ion capable of nucleophilic attack. It binds to catalytic sites of enzymes using choline as a substrate and reacts covalently causing irreversible inhibition. 20,26The HAChT system is a specific marker for cholinergic neurons and is important for the acetylcholine synthesis. 28–30Administration of AF64A in animals results in a long-lasting reduction in the number of functional Na⁺-dependent HAChT sites, leading eventually to a long-term reduction in the steady-state levels of tissue acetylcholine. 29The effect of intrathecal AF64A on the HAChT in the dorsal spinal cord has not been shown directly. In the present study, we found a significant reduction in the spinal [³H]HC-3 binding site after AF64A treatment, suggesting the spinal high-affinity choline transporters and cholinergic presynaptic terminals are probably damaged by AF64A. This is consistent with a recent report that spinal acetylcholine content is reduced by the similar treatment with AF64A. 19The decrease in the spinal endogenous acetylcholine concentration after AF64A treatment could be a result of an irreversible alkylation of nucleophilic active sites on the choline carrier. Thus, it could render the cholinergic nerve terminal deficient in newly synthesized acetylcholine because of its prevention of choline from gaining access to the site of acetylcholine synthesis. As an alternative, the decrease in acetylcholine may result from cholinergic nerve terminal degeneration. In this regard, AF64A could be accumulated in the cholinergic neurons through HAChT, which could disrupt the metabolic processes required for neuronal viability. 29 

We found that intrathecal treatment with either of the two structurally dissimilar choline transporter inhibitors, AF64A or HC-3, both reduced significantly the morphine analgesia to the same extent. These data strongly suggest that endogenous acetylcholine in the spinal cord is important for the full manifestation of morphine analgesia. Acetylcholine is formed in cholinergic neurons from the cosubstrates, choline and acetyl coenzyme A, through a reaction catalyzed by the enzyme choline acetyltransferase. Because cholinergic neurons cannot synthesize choline de novo , their function depends on choline uptake. 25,28The HAChT is considered to be present specifically in cholinergic neurons because a substantial portion of choline is converted to acetylcholine only when taken up through the high-affinity system. 28In the spinal cord, the high-affinity choline transporters are found at high density in the dorsal horn of the spinal cord, 25which could be the reason why intrathecal AF64A and HC-3 had no apparent effect on the motor function.

The role of spinal muscarinic receptors in morphine analgesia is supported by the finding in the present study that intrathecal atropine largely attenuated the effect of intravenous morphine. It has been reported that intrathecal atropine blocks the analgesic effect of intraperitoneal morphine in rats. 14We observed previously that the analgesic effect of intravenous morphine is not attenuated by intrathecal atropine if given 15–30 min after intravenous morphine. 15Thus, pretreatment of cholinergic antagonists appears to be important to demonstrate the role of spinal cholinergic receptors in morphine-produced analgesia. The discrepancy between this and our previous studies is most likely attributable to the fact that pretreatment of atropine can effectively block nitric oxide production caused by intrathecal morphine. We have shown that there is a rapid generation of nitric oxide in the spinal cord dorsal horn after intravenous morphine, which is completely blocked by pretreatment with intrathecal atropine. 5Intrathecal injection of nitric oxide synthase inhibitors eliminates the analgesic effect of intravenous morphine, indicating that spinal nitric oxide is essential for the analgesic action of systemic morphine. 15The important observation that pretreatment, but not posttreatment, with atropine can attenuate the effect of morphine supports an intermediate role of spinal cholinergic receptors in morphine-produced analgesia (i.e.,  morphine → cholinergic receptors → nitric oxide → analgesia). There are at least three different subtypes of muscarinic receptors in the spinal cord. 16Because of a lack of highly specific antagonists for muscarinic receptor subtypes, we were unable to further study precisely the subtypes of muscarinic receptors in intravenous morphine analgesia using pharmacologic approaches at present. Importantly, Duttaroy et al.  31have recently reported that the analgesic potency of morphine is reduced in M⁴and M²/M⁴receptor knockout mice. Therefore, the M²and M⁴receptor subtypes probably play an important role in the analgesic action of intravenous morphine.

Our study provides the first evidence that spinal nicotinic receptors mediate the analgesic action of intravenous morphine. Both spinal muscarinic and nicotinic receptors are present in the spinal cord. 16,17Stimulation of spinal nicotinic receptors can produce antinociception in animals. 27The α⁴and β²subunits of neuronal nicotinic receptors are directly related to the analgesia produced by nicotinic agonists and are also probably involved in morphine analgesia. 9,32We found that intrathecal pretreatment with mecamylamine significantly attenuated the effect of intravenous morphine. This observation is consistent with our recent finding that nicotinic receptors are involved in spinal nitric oxide release caused by clonidine and the antiallodynic effect of intrathecal clonidine in a rat model of neuropathic pain. 12,13Similar to our recent finding that spinal muscarinic receptors play a greater role in the clonidine-produced antiallodynia in neuropathic pain, 12we found that intrathecal atropine attenuated the morphine analgesia to a greater extent than intrathecal mecamylamine. A combination of muscarinic and nicotinic antagonists did not further reduce the morphine analgesia, suggesting that a common final pathway is responsible for both muscarinic and nicotinic receptor-mediated analgesia. In this regard, spinal nitric oxide has been shown to be a prerequisite mediator for morphine analgesia in rats. 5,15 

We have shown that spinal acetylcholine release after intravenous morphine depends on supraspinal sites. 5It has been shown that the noradrenergic neurons innervating the spinal cord dorsal horn are mainly located in the supraspinal site, especially the A7 cell group in the dorsolateral pontine tegmentum. 4,33On the other hand, the spinal cholinergic system appears to be intrinsic because spinal cholinergic neurons and nerve terminals are mainly located in the deep dorsal horn. Recent studies have suggested that systemically administered morphine probably acts on μ-opioid receptors in the ventromedial medulla and periaqueductal gray to cause disinhibition of noradrenergic neurons located in the A7 cell group, 34–36which leads to norepinephrine spillover in the spinal cord dorsal horn. 4,37Activation of spinal α²receptors could stimulate spinal cholinergic interneurons to increase acetylcholine, which then produces analgesia through spinal nitric oxide release. 12,13Consistent with this notion, it has been shown that cholinergic neurons are colocalized with neurons expressing nitric oxide synthase in the spinal cord dorsal horn. 38We have demonstrated that clonidine-induced spinal nitric oxide release is mediated by spinal muscarinic and nicotinic receptors. 13Therefore, it is likely that spinally released acetylcholine functions as an important mediator of spinal nitric oxide generation and the analgesic effect of systemic morphine.

In summary, the current study indicates that the spinal cholinergic system plays an important role in the analgesic action of intravenous morphine. Our data, together with previous studies, 5,14provide complementary functional evidence that intravenous morphine activates the descending inhibitory system leading to increased release of endogenous acetylcholine in the spinal cord, which produces analgesia through activation of spinal muscarinic and nicotinic receptors. Thus, the current study has an important clinical implication that intrathecal cholinergic receptor agonists or acetylcholinesterase inhibitors could potentiate the analgesic effect of systemically administered opioids.