Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Spinal application of cholinomimetics has been shown to produce antinociception in animals as well as in humans (Yaksh et al., 1985; Gillberg et al., 1989; Iwamoto and Marion, 1993a; Qian et al., 1993; Naguib and Yaksh, 1994; Bouaziz et al., 1995; Hood et al., 1995; Naguib and Yaksh, 1997) but the specific neuronal mechanisms underlying spinal cholinergic antinociception are not understood. There is good evidence, however, that cholinergic antinociception is modulated, in part, through an activation of muscarinic cholinergic receptors (mAChRs) (Yaksh et al., 1985; Gillberg et al., 1989; Iwamoto and Marion, 1993b; Naguib and Yaksh, 1997). In addition, stimulation of nicotinic receptors has been shown to elicit both antinociceptive and algogenic responses (Qian et al., 1993; Khan et al., 1997, 1998).

Pharmacological studies have suggested an antinociceptive role for m1, m2, m3, and m4 mAChRs (Iwamoto and Marion, 1993b; Bouaziz et al., 1995; Naguib and Yaksh, 1997; Ellis et al., 1999). It is, however, unlikely that the m1 receptor is of importance because this subtype is not detectable in spinal cord (Höglund and Baghdoyan, 1997). In addition, the m1 receptor has been shown not to be a requirement for muscarinic antinociception (Sheardown et al., 1997). Two different nicotinic-binding sites have been proposed to exist in the spinal cord for epibatidine of which the low-affinity site is suggested to mediate antinociception (Khan et al., 1997).

Several transmitter systems have been proposed to influence the release of ACh. Recently, it was proposed that i.v.-administered morphine activates spinal release of norepinephrine, which in turn activates the release of ACh from spinal interneurons in sheep (Xu et al., 1997). It also has been shown that serotonin might be involved in the control of antinociception produced by muscarinic agonists (Iwamoto and Marion, 1993b). The release of ACh possibly causes a release of nitric oxide (Xu and Tseng, 1994; Xu et al., 1996) and -aminobutyric acid (Baba et al., 1998). According to these findings cholinergic neurons play a key role in the spinal processing of nociceptive information.

Because data are now available showing that m2, m3, and m4 mAChRs and nicotinic receptors are present in the spinal cord (Gillberg and Aquilonius, 1985; Höglund and Baghdoyan, 1997) it is of interest to establish a model to study how these receptor classes and subtypes control ACh release in the spinal cord. Microdialysis has successfully been used to study mAChR regulation of ACh release in striatum (Damsma et al., 1987; Billard et al., 1995), in the septo-hippocampal system (Moor et al., 1995), and in pons (Baghdoyan et al., 1998). With microdialysis, data were obtained to suggest that the m2 mAChRs function as autoreceptors in both pons and striatum (Billard et al., 1995; Baghdoyan et al., 1998) because subtype-specific inhibition of these subtypes causes an increased release of ACh. Agonists were in these studies shown to inhibit the release of ACh.

The aim of the present study was to use the spinal loop dialysis catheter (Marsala et al., 1995) to establish a model in rat to further study the cholinergic receptor mechanisms regulating the release of ACh in spinal cord. Because the cholinesterase inhibitor neostigmine (Naguib and Yaksh, 1994; Bouaziz et al., 1995) and mAChR agonists such as carbachol (Iwamoto and Marion, 1993b; Abram and O'Connor, 1995), oxotremorine (Sheardown et al., 1997), and epibatidine (Qian et al., 1993) have been found to produce antinociception it was hypothesized that the agonists would increase the release of ACh in the spinal cord. The mAChR antagonist atropine has been found to elicit an algogenic effect (Ghelardini et al., 1990), and thus, it was hypothesized that an antagonist administered to the spinal cord would decrease the release of ACh. The effects of these substances on ACh release in the spinal cord would thus be opposite to what has been reported from dialysis studies in the brain (Damsma et al., 1987; Moor et al., 1995; Baghdoyan et al., 1998).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

All experiments were conducted after approval of the Animal Ethics Committee in Uppsala, Uppsala, Sweden. Male outbred Sprague-Dawley rats (n = 53) weighing 340 to 480 g (B&K Universal, Sollentuna, Sweden) were provided with free access to food (R36; Ewos, Vadstena, Sweden) and tap water at all times. The animals were acclimatized after delivery at a room temperature of 20 ± 2°C and were kept on a 12-h light/dark cycle 1 week before use.

The drugs oxotremorine sesquifumarate, (±)-epibatidine HCl, carbamylcholine chloride (carbachol), neostigmine bromide, ()-scopolamine methyl bromide, ACh chloride, and choline were all purchased from Sigma-Aldrich Sweden AB (Stockholm, Sweden). Methohexital sodium (Brietal) was purchased from Eli Lilly (Indianapolis, IN). Isofluane was purchased from Abbot Scandinavia (Kista, Sweden). The salts NaCl, CaCl₂, Na₂HPO₄, and KCl were purchased from Kebo Lab (Spånga, Sweden). Spinal microdialysis probes were purchased from Marsil Enterprises (San Diego, CA).

For each experiment, anesthesia was induced with methohexital sodium (40 mg/kg) and rats were intubated and connected to a Harvard (Harvard Apparatus Inc., South Natick, MA) ventilator. A rat was placed on a heated pad to maintain core body temperature at 37.5°C. The methohexital sodium anesthesia was subsequently replaced by isoflurane in 100% oxygen. During surgery isoflurane was kept at about 2.5% and during microdialysis sampling isoflurane was maintained at 1.3%. The end-tidal pCO₂ was kept at 4 kPa.

For insertion of the microdialysis probe a midline incision was made at the back of the skull. The neck muscles were dissected to expose the cisterna magna. The dura and pia mater were cut and a spinal microdialysis probe (Marsala et al., 1995) was inserted intraspinally so that the tip was at about C5. The probe was perfused (3 µl/min) with Ringer's solution (147 mM NaCl, 2.4 mM CaCl₂, 4.0 mM KCl) containing 10 µM neostigmine to prevent degradation of ACh (Billard et al., 1995; Roth et al., 1996). After inserting the microdialysis probe, the rats rested for 40 min before starting spinal microdialysis.

ACh (pmol/10-min dialysis sample) was quantified by HPLC with electrochemical detection (Antech, Leiden, The Netherlands). The mobile phase was 50 mM Na₂HPO₄ (pH 9.0), which enabled detection of ACh at 4.25 min and choline at 5.5 min. The dialysis probe recovery of ACh was determined both before and after each experiment to ensure that all microdialysis measures accurately reflected the spinal ACh release and were not confounded by intraexperimental probe damage. A standard calibration curve ranging from 1 to 20 pmol of ACh was established before each experiment from two samples of each concentration. Twenty microliters of the collected 30-µl samples were analyzed for ACh directly after each sampling period was completed.

Concentrations of neostigmine ranging between 1 and 10 µM have been used in previous microdialysis experiments of ACh from brain tissue (Damsma et al., 1987; Billard et al., 1995; Moor et al., 1995; Baghdoyan et al., 1998) to prevent degradation of ACh and enable analysis. A concentration range between 0.1 and10 µM neostigmine was initially tested in the present experiments and it was found that the most consistent amount of ACh was obtained with 10 µM neostigmine. A too-high concentration of neostimine might influence the release of ACh especially if autoreceptors are present. Such influences were however not observed with 10 µM neostigmine. Thus 10 µM neostigmine was used in the Ringer's solution which served as vehicle in all experiments.

Data were analyzed by using ANOVA. Dunnett's post hoc test was used to determine the statistical difference against control. P values <.05 were considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Probe recovery averaged 38%, and for each experiment pre- and postexperimental probe recoveries were compared by t test. For all of the data described there were no statistically significant changes in dialysis probe recovery during any single experiment.

Initially, the basal release of ACh from spinal cord was established over time and found to be stable for at least 150 min (Fig. 1, control) at 3.98 ± 0.22 (mean ± S.E.) pmol/10-min sample when only Ringer's solution with 10 µM neostigmine was dialyzed. After this prerequisite had been fulfilled the mixed nicotinic and muscarinic agonist carbachol (1 mM), the muscarinic agonist oxotremorine (1 mM), and the nicotinic agonist epibatidine (50 µM), respectively, were dialyzed in time-response studies. In these studies five samples of 30 µl each were collected while dialyzing with Ringer's solution to establish the basal release of ACh. The mean ACh content in these five samples was used to calculate the percentage of change of ACh release in the following samples. It was found that all agonists caused a significant increase in the release of ACh (Fig. 1). In contrast, the mAChR antagonist scopolamine (100 nM) caused a decrease in ACh release during the 150-min sampling period. The effect of oxotremorine and epibatidine on ACh release was found to reach a plateau after 40 min, whereas carbachol acted at a slower rate.

View larger version (24K): [in this window] [in a new window]   Fig. 1.   The spinally placed microdialysis probe was perfused with Ringer's solution containing 10 µM neostigmine to prevent ACh degradation and make possible the HPLC analysis. Samples of 30 µl/10 min were collected of which 20 µl was assayed for ACh content. The first five samples in each experiment were averaged and the baseline release thus obtained was used to calculate the percentage of change of ACh release in the following sampling periods. The basal ACh amount was 5.32 ± 0.32 (mean ± S.E.) pmol/20 µl. The effects of oxotremorine (1 mM, n = 5, ), carbachol (1 mM, n = 5, ), epibatidine (50 µM, n = 5, ), and scopolamine (100 nM, n = 4, ) was studied over a period of 150 min and compared, with use of the one-way ANOVA with Dunnett's post hoc analysis, to dialysis of Ringer's solution only (n = 5, ). From 80 to 150 the effects of oxotremorine, carbachol, and epibatidine were significantly different (F4,19 > 32 at each point of time).

Because the effect of oxotremorine and epibatidine reached a maximum after four sampling periods it was considered feasible to obtain a dose-response relationship from experiments where several concentrations of these substances were administered to the same animal switching to a higher dose after five sampling periods. Statistics were performed on the last three samples at each dose level. Different concentrations of oxotremorine (30-300 µM), epibatidine (5-5000 µM), and scopolamine (25-200 µM) were administered after an initial period of five samples had been collected after dialysis with Ringer's solution only. No animal was given more than three different concentrations. Both oxotremorine (Fig. 2) and epibatidine (Fig. 3) were found to increase the release of ACh in a dose-dependent manner.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Effects of 30 (n = 4), 100 (n = 4), 300 (n = 4), and 1000 µM (n = 5) oxotremorine on the release of ACh in percentage of relation to baseline. The basal ACh amount was 6.2 ± 0.8 (mean ± S.E.) pmol/20 µl. The line shows the best fit to the hyperbolic function {maximum effect × [oxotremorine]/(ED50 + [oxotremorine])} of the averaged change in release at each dose level. Doses 100 to 1000 µM were significantly different from the control period (F4,78 = 15.7).

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Effects of 5 (n = 4), 16 (n = 4), 50 (n = 4), 160 (n = 4), 500 (n = 3), 1600 (n = 3), and 5000 (n = 3) µM epibatidine on the release of ACh in percentage of relation to baseline. The basal ACh amount was 3.8 ± 0.4 (mean ± S.E.) pmol/20 µl. The line shows the best fit to the hyperbolic function {maximum effect × [epibatidine]/(ED50 + [epibatidine])} of the averaged change in release at each dose level. Doses 50 to 5000 µM were significantly different from the control period (F7,108 = 11.6).

Scopolamine did not affect the release of ACh at 25 nM (n = 3) but inhibited the release of ACh significantly (F_4,61 = 6.8) at 50 nM (27.7 ± 4%, n = 3), 100 nM (21.6 ± 4.8%, n = 4), and 200 nM (28.4 ± 3.5%, n = 3). The basal ACh amount detected in control samples was in these experiments 4.6 ± 0.4 pmol/20 µl.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The intraspinal location of the dialysis probe was found to give measurable amounts of ACh when 10 µM neostigmine was added to the Ringer's solution. For technical reasons it is impossible to place the probe exactly at the same place in each experiment. Generally, the probes were found in the dorsal region of the spinal cord when dissected after completion of control experiments. Probably depending on the small differences in location of the probes a slight variation in basal release of ACh between experiments was observed. This variation prompted us to collect five samples at the beginning of each experiment to calculate the percentage of change in release of ACh from baseline. It is notable that the basal release was constant over a period of at least 150 min, making it possible to study the effects of pharmacological interventions.

Although the probes were placed in the dorsal part of the spinal cord it is highly likely that the effects we observed on ACh release are a summation of events in the whole diameter of the cord because the substances dialyzed probably diffuse readily. This phenomenon must be taken into consideration when results from these kinds of experiments are discussed from a mechanistical and locational point of view.

Although it has been shown that pain itself, opiate receptor activation, and 2-adrenergic receptor activation produce analgesia by increasing the release of ACh from spinal cord (Eisenach, 1999), the effects of direct application of muscarinic or nicotinic agonists have never been tested until now. Our study, which basically was initiated to establish a model to further study which receptors and which receptor subtypes are involved in the control of ACh release, clearly shows that both nicotinic and muscarinic agonists cause a significant increase in ACh release. Mechanistically, this is interesting because administration of agonists to other areas of the central nervous system such as hippocampus, pons, and striatum shows the opposite effect (Billard et al., 1995; Baghdoyan et al., 1998). Inhibition of ACh release after agonist administration has been suggested to be evidence of the presence of autoreceptors in these structures (Billard et al., 1995). Because both nicotinic and muscarinic agonist administrations to the spinal cord caused an increased release of ACh and because the antagonist scopolamine caused a decreased release of ACh, the present data does not support an autoreceptor function of either muscarinic or nicotinic receptors in the spinal cord. It is more likely that ACh release is under tonic control of other transmitter systems such as the norepineprinergic and endogenous opioid because activation of these systems has been shown to increase spinal ACh release (Bouaziz et al., 1996; Klimscha et al., 1997) or has been suggested to interact with cholinergic mechanisms (Detweiler et al., 1993). Serotonin also has been implicated in the control of cholinergic activity in the spinal cord (Iwamoto and Marion, 1993a,b).

We show that oxotremorine and epibatidine, which are well known for their analgesic properties (Qian et al., 1993; Sheardown et al., 1997), enhanced the release of ACh in a dose-dependent manner after microdialysis administration. The effect of these substances closely followed a hyperbolic function, indicating that a single receptor (of each muscarinic and nicotinic class) only is involved in the regulation of ACh release. These results were unexpected because of the data suggesting that at least m2, m3, and m4 mAChRs as well as high- and low-affinity nicotinic receptors exist in the spinal cord and may all be involved in ACh release control. Especially, the effect of epibatidine on ACh release is interesting because it has been shown that intrathecal administration of low doses of this substance causes a nociceptive response (Khan et al., 1998). In view of the suggested relationship between increased ACh in spinal cord and antinociception this would implicate that epibatidine in low doses should decrease the release of ACh. Although we administered epibatidine over a large dose range (5-5000 µM) we consistently observed increases in ACh release, indicating that low-dose administration of epibatidine must cause the algogenic responses through other mechanisms than by reducing ACh release. These mechanisms would include release of excitatory amino acids because intrathecally nicotinic agonists were found to release aspartate and glutamate (Khan et al., 1996).

In conclusion, the method described in the present paper is suitable for studies of intraspinal mechanisms regulating ACh release. The release of ACh is constant for at least 150 min in the anesthetized rat. Cholinergic agonists cause an increase of intraspinal ACh and the antagonist scopolamine caused a decreased release of ACh. The data do not support an autoreceptor function of neither nicotinic nor muscarinic receptors in the spinal cord, contrary to what has been observed in the brain.