Introduction

Top Abstract Introduction Methods Results Discussion References

It has been long known that compounds that increase ACh activity at muscarinic receptors, such as the acetylcholinesterase inhibitors physostigmine, neostigmine and THA, and several muscarinic agonists such as oxotremorine, arecoline, RS86 and pilocarpine, can produce antinociception in animals (Swedberg, 1994; Swedberg et al., 1993; Green and Kitchen, 1986) and man (Peterson et al., 1986; Flodmark and Wramner, 1945). Antinociception so produced in animals was reversed by atropine (Hartvig et al., 1989; Green and Kitchen, 1986). However, the small separation between the antinociceptive effects and side effects such as tremor and salivation have made clinical use of the classical muscarinic agonists unfavorable. The discovery of a novel nonopioid analgesic based on muscarinic mechanisms with an improved therapeutic window over opioids and presently available muscarinics would be clinically very valuable.

We have recently shown that a series of muscarinic agonists, the alkylthio/oxy-TZTPs, can produce antinociception in rodents, many with therapeutic windows far superior to those of the classical muscarinics (Sauerberg et al., 1995; Swedberg et al., 1993). The alkylthio/oxy-TZTPs were equieffective to morphine and in some cases severalfold more potent than morphine (Swedberg et al., 1993; Sheardown et al., 1997).

The present studies were conducted to assess and characterize the antinociceptive potential of the muscarinic agonist/antagonist butylthio[2.2.2] (NNC 11-1053/LY297802; fig. 1), the (+)-isomer of a racemic mixture chosen from a series of azabicyclic 1,2,5-thiadiazoles, based on its favorable structure-activity relationship in terms of antinociceptive and side-effect profile (Olesen et al., 1996). Butylthio[2.2.2] was an agonist/antagonist at muscarinic receptors in vitro (Shannon et al., 1997). The antinociceptive effects were assessed after s.c. and p.o. administration in mice with the grid-shock (Swedberg, 1994), tail-flick (D'Amour and Smith, 1941), hot-plate (Eddy and Lambach, 1953) and acetic acid-induced writhing (Ward and Takemori, 1983) tests, and in rats with use of a modified version of the mouse grid-shock test (Swedberg, 1994). The therapeutic index of butylthio[2.2.2] was assessed relative to that of oxotremorine by comparing the ED₅₀ values for antinociception, salivation and tremor. The motor-impairing potential of butylthio[2.2.2] was assessed in a rotarod apparatus, and incidence of lethality was also scored. The receptor mechanisms underlying the antinociceptive effects of butylthio[2.2.2] were investigated by coadministration of the opiate antagonist naltrexone and the muscarinic antagonist scopolamine (Heller Brown, 1990). The issue of central versus peripheral origin of the antinociceptive effects was investigated by use of the centrally acting muscarinic antagonist scopolamine and its quaternary form, methscopolamine, which does not readily cross the blood-brain barrier (Heller Brown, 1990). The duration of the antinociceptive effect was also determined. Because one major problem with currently used opioid analgesics in the clinic is the rapidly developing tolerance to the analgesic effects (Jaffe and Martin, 1990), we also assessed whether tolerance would develop to the antinociceptive effects of butylthio[2.2.2].

View larger version (8K): [in this window] [in a new window]   Fig. 1.   Chemical structure of butylthio[2.2.2] (NNC 11-1053/LY297802).

It was concluded that butylthio[2.2.2] was effective in several rodent models after s.c. and p.o. administration. Butylthio[2.2.2] was equieffective to, and more potent than morphine, and had a therapeutic index far superior to that of classical muscarinics. The antinociceptive effect was mediated by central muscarinic but not opioid receptors. There was little, if any, tolerance to the antinociceptive effects in the mouse. These data show that butylthio[2.2.2] has a very promising pharmacological and antinociceptive profile and suggest it as an interesting candidate for clinical development.

	Methods

Top Abstract Introduction Methods Results Discussion References

Subjects. Male NMRI mice (Bomholdtgaard, Ry, Denmark) weighing 18 to 22 g were used for the grid-shock, hot-plate and tail-flick studies, and male Crl:CF1^RBR mice (Charles River Laboratories, Portage, MI) weighing 18 to 32 g were used for the writhing studies. Male Wistar rats (Moellegaard, Lille Skensved, Denmark) weighing 130 to 160 g were used for the rat grid-shock studies. Animals were group housed in a large colony room with food and water available at all times, and lights on between 6:00 A.M. and 6:00 P.M. In the morning of the day of the experiments, the animals were brought into the laboratory. Each animal was weighed and was administered a dose according to body weight. The mice used in the grid-shock test were also used for scoring side effects. All other animals were used in one test only.

Grid-shock. Apparatus and testing procedures were as described earlier (Swedberg, 1994), or a computerized version thereof with identical parameters for shock deliveries and sound measurements. A transparent acrylic chamber (13 × 13 × 13 cm for mice, and 22 × 29 × 22 cm for rats) was equipped with a stainless steel grid floor through which electric shocks could be delivered. The top of the chamber was covered by an acrylic plate that had a decibel meter, or a dynamic microphone connected to an amplifier with an electronic filter with a 2.4 kHz center frequency (computerized version) attached to it. A shock generator for delivery of scrambled shocks was connected to the grid floor. The shock generator delivered an output current starting at 0 µA and increasing by 10 µA (20 µA for rats) every 0.01 min up to a maximal 0.5 mA (1 mA for rats) at 0.50 min (30 sec). Shocks at 160 V were presented as square-wave pulses of 2-msec duration at 30 Hz. Shocks were terminated when a 70-dB squeak was emitted, or after 30 sec, whichever occurred first; a latency score was determined. A predrug latency to vocalization was generated by placing each animal individually in the chamber and starting the shock generator. Animals were then administered vehicle and/or drug and re-placed in the holding cages. After the appropriate pretreatment interval, a postdrug latency to vocalization was generated. Ten animals were used per dose and for the controls.

Acetic acid-induced writhing. Separate groups of 10 mice each were administered vehicle and/or drug. After the appropriate pretreatment interval, an i.p. injection of 0.5% acetic acid was administered. Each mouse was then placed in an individual clear plastic observational chamber, and the total number of writhes made by each mouse was counted between 5 and 10 min after acetic acid administration.

Hot-plate. A predrug latency to jumping off a 55°C hot plate was generated in each mouse. Animals were then administered the appropriate drug and dose, and were re-placed in the holding cages. Thirty minutes later, a postdrug latency was generated. Ten animals were used per dose.

Tail-flick. A predrug latency to removal of its tail from a 55°C water bath was generated for each mouse. Animals were then administered the appropriate drug and dose, and were re-placed in the holding cages. Thirty minutes later, a postdrug latency was generated. Ten animals were used per dose.

Tolerance testing (grid-shock). In the afternoon of the first day (day 1) of antinociceptive testing, and in the morning and afternoon of the following 6 days (days 2-7), all mice received a dose of either butylthio[2.2.2] (3.0 mg/kg) or morphine (40.0 mg/kg) s.c., doses corresponding to ten times the ED₅₀ as determined on day 1. On day 8, the mice were tested again. For the repeated dosing regimen each mouse was weighed daily to allow accurate dosing.

Side-effect scoring in mice. Immediately before being tested postdrug in the grid-shock test, the incidence and severity of salivation and tremor was assessed in each mouse by assigning each animal a score from 0 to 3 corresponding to the severity of the effect (0 = no effect, 3 = maximal effect). For salivation, a score of 0 indicates no observed signs; 1, wet around the mouth; 2, wet around mouth, throat and/or neck; 3, wet around mouth, throat and/or neck, and on chest and/or belly. For tremor, a score of 0 indicates no observed signs; 1, weak tremor upon handling; 2, tremor and flexing of neck upon handling; 3, spontaneously occurring pronounced clonic tremor.

Rotarod ataxia. These tests were conducted as described earlier (Swedberg et al., 1995a). Mice were trained to remain on a rotating rod (6 rpm) for 2 min as the rod rotated toward the animal. After the 2-min training period, the mice were administered vehicle or drug and 30 min later placed on the rotating rod. Each mouse was observed for 2 min, and a mouse that fell off the rod twice or more was considered ataxic. Eight animals were used per dose and for the controls.

Lethality. The incidence of deaths were observed in the mice (n = 8/dose) tested in the rotarod apparatus. Mice were observed for 30 min after a s.c. administration of vehicle or drug.

Data analysis. For grid-shock data a percent increase in latency to vocalization score [(postdrug  predrug/cutoff  predrug) × 100] was generated for each animal, and the average was then calculated for each dose. Writhing data are expressed as the mean number of writhes ± 1 S.E. during the 5-min observation period. Tail-flick and hot-plate data represent the percent increase in latency to tail withdrawal or jumping off the hot plate with use of the same formula as with the grid-shock test. Salivation and tremor data were calculated by multiplying each score by the number of mice which received that score, and expressing the sum of scores for each group (n = 10) as a percentage of the maximum possible sum of scores (30). Rotarod data are expressed as the percentage of mice falling off the rotating rod during the observation period. Lethality data are expressed as the percentage of mice dying within the observation period. ED₅₀ values, t_1/2 values and estimated standard errors were calculated by sigmoid nonlinear regression equations (GraphPAD Prism, GraphPAD Software Inc., San Diego, CA).

Drugs. Butylthio[2.2.2] tartrate (NNC 11-1053/LY297802), naltrexone HCl (Endo Laboratories, Garden City, NY), oxotremorine sesquifumarate salt (Sigma, St. Louis, MO), ()-scopolamine hydrobromide (Sigma), morphine HCl (Mecobenzon, Copenhagen, Denmark), morphine sulfate (Eli Lilly and Company, Indianapolis, IN) and ()-scopolamine methyl bromide (Sigma) were dissolved in distilled or deionized water, and injected s.c. or p.o. in a volume of 10.0 and 1.0 ml/kg to mice and rats, respectively.

	Results

Top Abstract Introduction Methods Results Discussion References

Butylthio[2.2.2] and morphine produced dose-dependent antinociception in mice after s.c. as well as p.o. administration in the grid-shock (fig. 2, upper left section), writhing (fig. 2, upper right section), tail-flick (fig. 2, lower left section) and hot-plate (fig. 2, lower right section) tests. Butylthio[2.2.2] and morphine were equieffective (75-100%) in all tests after s.c. administration. When administered p.o., butylthio[2.2.2] was equieffective in all tests. Morphine after p.o. administration was less effective in the grid-shock (64%) and tail-flick (54%) tests.

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Antinociceptive effects of butylthio[2.2.2] (circles) and morphine (triangles) 30 min after subcutaneous (closed symbols) or oral (open symbols) administration in the mouse grid-shock (top left), writhing (top right), tail-flick (bottom left) and hot-plate (bottom right) tests. x-axes, doses in milligrams per kilogram; y-axes, percent change in response latency (top left, bottom left and right), number of writhes (top right). V, vehicle.

The ED₅₀ values of butylthio[2.2.2] were 0.23, 0.19, 0.39 and 1.47 mg/kg after s.c., and 1.51, 3.77, 10.59 and 12.23 mg/kg after p.o. administration in the grid-shock, writhing, tail-flick and hot-plate tests, respectively (table 1). In the various antinociceptive assays, butylthio[2.2.2] was approximately 3 to 24 and 2 to 47 times more potent than morphine after s.c. and p.o. administration, respectively. Ratios of p.o. to s.c. potencies in the various assays ranged from 7 to 27 for butylthio[2.2.2] and from 8 to 32 for morphine (table 1).

Relative antinociceptive potencies for butylthio[2.2.2], morphine and oxotremorine after s.c. administration are shown in table 2. Butylthio[2.2.2] was 3 to 24 times more potent than morphine and 4 to 14 times less potent than oxotremorine (table 2).

In the rat grid-shock test, butylthio[2.2.2] produced a dose-dependent increase in antinociception after s.c. as well as p.o. administration (fig. 3, left section), yielding ED₅₀ values of 0.26 and 25.28 mg/kg (table 1). Morphine produced a dose- dependent increase in antinociception after s.c. administration, but was less effective after p.o. administration (fig. 3, right section), yielding ED₅₀ values of 1.46 and >100 mg/kg after s.c. and p.o. administration, respectively (table 1). Oxotremorine after s.c. administration produced an ED₅₀ in the rat grid-shock test of 0.11 mg/kg (table 2, data not shown).

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Antinociceptive effects of butylthio[2.2.2] (circles; left) and morphine (triangles; right) 30 min after subcutaneous (closed symbols) or oral (open symbols) administration in the rat grid-shock test. x-axes, doses in milligrams per kilogram; y-axes, percent change in response latency.

Butylthio[2.2.2] produced a dose-dependent increase in antinociception in the mouse grid-shock test as well as in salivation and tremor after s.c. (fig. 4, left section) and p.o. (fig. 4, right section) administration. The ED₅₀ values for salivation and tremor were >30 and 12.31 mg/kg, and >60 and >60 mg/kg after s.c. and p.o. administration, yielding ratios of salivation and tremor to antinociception of >130 and 54, and >40 and >40, after s.c. and p.o. administration, respectively (table 3 for s.c. data). In comparison, oxotremorine (s.c.) produced dose-dependent increases in antinociception in the grid-shock test, as well as in salivation and tremor (fig. 5). The ED₅₀ values were 0.033, 0.22 and 0.41 for antinociception, salivation and tremor, respectively (table 3). The ratios of salivation and tremor to antinociception were 7 and 12, respectively (table 3).

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Antinociceptive effects of butylthio[2.2.2] in the mouse grid-shock test (closed circles), salivation (open squares), tremor (open triangles) 30 min after subcutaneous (left section) or oral (right section) administration. x-axes, doses in milligrams per kilogram; y-axes, percent effect. V, vehicle.

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Antinociceptive effects of oxotremorine in the mouse grid-shock test (closed circles), salivation (open squares), tremor (open triangles) 30 min after subcutaneous administration. x-axis, doses in milligrams per kilogram; y-axis, percent effect. V, vehicle.

Rotarod ataxia and lethality produced by butylthio[2.2.2] increased dose dependently (fig. 6), yielding approximate ED₅₀ values of 26.6 and 58.5 mg/kg, respectively (table 3). The ratios of rotarod ataxia and lethality to antinociception for butylthio[2.2.2] were 116 and 254, respectively, after s.c. administration (table 3). Whereas some motor incoordination was seen at 10 and 30 mg/kg of butylthio[2.2.2], deaths occurred only at doses of 60 and 100 mg/kg. At the highest dose of butylthio[2.2.2] (100 mg/kg) all mice had myoclonic seizures within 20 min after administration. Rotarod ataxia and lethality produced by oxotremorine increased dose dependently (fig. 6), yielding approximate ED₅₀ values of 0.07 and 10.2 mg/kg, respectively (table 3). The ratios of rotarod ataxia and lethality to antinociception for oxotremorine were 2 and 309, respectively (table 3).

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Incidence of rotarod ataxia (closed symbols) and lethality (open symbols) produced by subcutaneous administration of butylthio[2.2.2.] (circles) and oxotremorine (triangles) in mice. Rotarod ataxia was assessed 30 min after drug administration. Incidence of lethality was observed up to 30 min after administration. x-axis, doses in milligrams per kilogram; y-axis, percent effect.

Butylthio[2.2.2] at 3.0 mg/kg and morphine at 30.0 mg/kg had a similar duration of action after s.c. administration in the mouse grid-shock test, which yielded half-lives of approximately 160 and 180 min, respectively (fig. 7, left section). After p.o. administration in the mouse writhing test, butylthio[2.2.2] produced dose-dependent antinociception, yielding ED₅₀ values (mg/kg) of 3.77 (95% CL, 1.87-7.61), 5.72 (95% CL, 4.02-8.16), 17.79 (95% CL, 7.29-43.42) and >30 at 0.5, 1.0, 2.0 and 3 hr after administration (fig. 7, right section).

View larger version (23K): [in this window] [in a new window]   Fig. 7.   Duration of action of the antinociceptive effect of 3.0 mg/kg butylthio[2.2.2] (closed circles) and 30.0 mg/kg of morphine (open circles) at various time points after subcutaneous administration in the mouse grid-shock test (left section), and dose-response curves for the antinociceptive effects of butylthio[2.2.2] at 0.5 h (closed circles), 1 h (open circles), 2 h (open squares) and 3 h (open triangles) after oral administration in the mouse writhing test. x-axes, time in minutes (left section), doses in milligrams per kilogram (right section); y-axes, percent change in response latency (left section), and number of writhes (right section).

Naltrexone at 10.0 mg/kg in the mouse grid-shock test did not affect the antinociceptive dose-response curve of butylthio[2.2.2] as compared with vehicle (fig. 8, left section), yielding ED₅₀ values (mg/kg) for butylthio[2.2.2] of 0.78 (95% CL, 0.18-3.37) and 0.37 (95% CL, 0.10-1.32) after vehicle and naltrexone coadministration, respectively. In the mouse writhing test, scopolamine produced a dose-dependent parallel shift to the right of the butylthio[2.2.2] dose-response curve, yielding ED₅₀ values (mg/kg) of 0.21 (95% CL, 0.15-0.29), 0.95 (95% CL, 0.28-3.21), 3.30 (95% CL, 2.77-3.94) and 7.32 (95% CL, 3.51-15.25) after vehicle, 0.1, 0.3 and 1.0 mg/kg of scopolamine, respectively (fig. 8, right section). In addition, whereas increasing doses of scopolamine (0.003-1.0 mg/kg) dose dependently reversed the antinociceptive effects of 1.0 mg/kg of butylthio[2.2.2] in the mouse writhing test, which yielded an ED₅₀ of 0.10 mg/kg (95% CL, 0.08-0.13), the antinociceptive effects of 1.0 mg/kg of butylthio[2.2.2] were partly reversed by methscopolamine only at a dose of 10.0 mg/kg, with an ED₅₀ of 9.77 mg/kg (95% CL, 9.66-9.88; fig. 9).

View larger version (22K): [in this window] [in a new window]   Fig. 8.   Antinociceptive effects of butylthio[2.2.2] after coadministration with saline (closed circles; left section) or 10.0 mg/kg of naltrexone (open circles; left section) in the mouse grid-shock test 30 min after subcutaneous administration, and dose-dependent shifts to the right of the butylthio[2.2.2] dose-response curve in the mouse writhing test 30 min after subcutaneous administration of scopolamine vehicle (open circles), scopolamine 0.1 mg/kg (open squares), scopolamine 0.3 mg/kg (open triangles) and scopolamine 1.0 mg/kg (open inverse triangles). x-axes, doses in milligrams per kilogram; y-axes, percent change in response latency (left section), and number of writhes (right section). V, butylthio[2.2.2] vehicle and the vehicle of the antagonist.

View larger version (16K): [in this window] [in a new window]   Fig. 9.   Reversal of the antinociceptive effects of 1.0 mg/kg of butylthio[2.2.2] in the mouse writhing test 30 min after subcutaneous administration of scopolamine (open circles) or methscopolamine (open squares). x-axis, doses of the antagonists in milligrams per kilogram; y-axis, number of writhes. V+V, butylthio[2.2.2] vehicle and the vehicle of the antagonist. V+1.0, butylthio[2.2.2] vehicle and 1.0 mg/kg of the antagonist.

Before repeated morphine treatment, the ED₅₀ (mg/kg) for morphine in the mouse grid-shock test was 3.81 (95% CL, 1.53-9.52), whereas after 6.5 days of morphine (40.0 mg/kg) twice daily, the dose-response curve was shifted to the right, yielding an ED₅₀ of 22.41 mg/kg (95% CL, 8.26-60.82; fig. 10, left section). Before repeated butylthio[2.2.2] treatment, the antinociceptive ED₅₀ for butylthio[2.2.2] in the mouse grid-shock test was 0.36 mg/kg (95% CL, 0.0003-376.11), whereas after 6.5 days of butylthio[2.2.2] (3.0 mg/kg) twice daily, the dose-response curve was depressed at the low and high doses, but not at the middle dose, yielding an ED₅₀ of 1.31 mg/kg (95% CL, 0.003-546.42; fig. 10, right section).

View larger version (20K): [in this window] [in a new window]   Fig. 10.   Antinociceptive effects of morphine after 6.5 days twice daily treatment with 40.0 mg/kg of morphine subcutaneously (left), and antinociceptive effects of butylthio[2.2.2] after 6.5 days twice daily treatment with 3.0 mg/kg of butylthio[2.2.2] subcutaneously (right), 30 min after subcutaneous administration in the mouse grid-shock test. x-axes, doses in milligrams per kilogram; y-axes, percent change in response latency. Open symbols, dose-response curves before repeated treatment; closed symbols, dose-response curves after repeated treatment.

	Discussion

Top Abstract Introduction Methods Results Discussion References

Consistent with previous reports that cholinesterase inhibitors and muscarinic agonists produced potent antinociception, the muscarinic agonist butylthio[2.2.2] produced potent and efficacious antinociception in mouse and rat. Butylthio[2.2.2] produced antinociception as efficaciously as morphine in the mouse after s.c. as well as p.o. administration, and was 2 to 47 times more potent. In the rat, butylthio[2.2.2] was six times more potent than morphine after s.c. administration. After p.o. administration to rats, butylthio[2.2.2] was at least four times more potent than morphine (see below). Compared with other clinically effective opioids, butylthio[2.2.2] was 150 times more potent than codeine and 3.5 times less potent than fentanyl in the mouse grid-shock test, which showed a very good correlation between potencies in the mouse and clinical doses (Swedberg, 1994). These data suggest that butylthio[2.2.2] may be a therapeutically useful analgesic in humans.

Butylthio[2.2.2] was antinociceptive at doses well below those causing typical muscarinic side effects such as salivation and tremor. The antinociceptive potency of butylthio[2.2.2] compares very favorably with what has been reported for other muscarinics. For example, as an antinociceptive butylthio[2.2.2] was equipotent to the cholinesterase inhibitor physostigmine (Swedberg, 1994; Harris et al., 1969) but 70-fold more potent than arecoline (Swedberg, 1994). Butylthio[2.2.2] was less potent than oxotremorine in producing antinociception as shown in this study and previously (Swedberg, 1994; Harris et al., 1969), and was also less potent in producing salivation or tremor, but it produced a larger separation between the therapeutic and side effects than oxotremorine. Thus, the ratios of the ED₅₀ values for mouse grid-shock antinociception to salivation or tremor, respectively, yielded therapeutic windows for butylthio[2.2.2] of >130 and 54, respectively, as compared with 7 and 12, respectively, for oxotremorine. Regarding motoric side effects and lethality, therapeutic ratios of 116 and 254 were obtained for butylthio[2.2.2], and 2 and 309 for oxotremorine, respectively. Thus, while butylthio[2.2.2] had a wide separation between antinociception and ataxia, oxotremorine produced ataxia at doses only slightly higher than those producing antinociception. The present findings suggest that butylthio[2.2.2] may be less prone to produce motoric disturbances at antinociceptive doses than typical muscarinic agonists such as oxotremorine.

Butylthio[2.2.2] had a favorable bioavailability (p.o./s.c. ratio) and duration of antinociceptive action in the mouse. The p.o./s.c. ratio of the antinociceptive potencies in the mouse was similar to that of morphine, whereas in the rat, lack of oral efficacy of morphine precluded comparison. Butylthio[2.2.2] produced full antinociceptive efficacy up to 2 hr after oral administration, and at equieffective doses after s.c. administration, butylthio[2.2.2] and morphine had a similar duration of action.

Muscarinic but not opioid mechanisms mediate the antinociceptive effects of butylthio[2.2.2] inasmuch as the muscarinic antagonist scopolamine produced a dose-dependent and parallel shift to the right of the antinociceptive dose-response curve of butylthio[2.2.2], and the opiate antagonist naltrexone was ineffective. These data confirm in vitro receptor binding results showing lack of binding to opiate receptors (Shannon et al., 1997). The lack of antagonism by the opiate antagonist naltrexone and lack of opioid binding suggest that butylthio[2.2.2] is likely devoid of opioid effects commonly seen with opioid analgesics, such as abuse liability and physical dependence (Jaffe and Martin, 1990). Furthermore, these and other findings (Pleuvry and Tobias, 1971) showing lack of antagonism of oxotremorine-induced antinociception by the opiate antagonist nalorphine, and lack of antagonism of antinociception induced by the muscarinic agonist (+)-cis-methyldioxolane by the opiate antagonist naloxone (Iwamoto and Marion, 1993), suggest that muscarinic antinociception is independent of the opioid systems.

Butylthio[2.2.2] appears to exert its antinociceptive effects via central muscarinic mechanisms, because the peripherally acting muscarinic antagonist methscopolamine only reversed the antinociceptive effects at doses more than 100 times higher than effective doses of the centrally acting muscarinic antagonist scopolamine. Moreover, the lack of antinociceptive effects by the peripherally acting non-steroid antiinflammatory, aspirin, in tests used in this study (Swedberg, 1994) further support the conclusion that butylthio[2.2.2] does not exert its antinociceptive effects via peripheral mechanisms.

Tolerance to the antinociceptive effects of butylthio[2.2.2] developed to a lesser extent, if at all, as compared with the tolerance seen with equipotent doses of morphine. After repeated morphine exposure, the morphine dose-response curve was shifted approximately 6-fold to the right in a parallel fashion. After repeated butylthio[2.2.2] exposure, the butylthio[2.2.2] dose-response curve overlapped with the dose-response curve obtained before repeated exposure.

However, comparing the ED₅₀ values before and after repeated exposure suggests an almost 4-fold nonparallel shift, and the separation at the high and low ends suggests that some degree of tolerance development cannot be ruled out even though it appears to be less pronounced than for morphine.

In summary, the muscarinic agonist/antagonist butylthio[2.2.2] was an efficacious and potent antinociceptive in mouse and rat. Butylthio[2.2.2] was more potent than morphine and had a larger therapeutic window than the muscarinic agonist oxotremorine. The bioavailability and duration of antinociceptive action was similar to what was seen with morphine. The antinociceptive effects of butylthio[2.2.2] were mediated by central muscarinic but not opioid receptors, which suggests that butylthio[2.2.2] may be less susceptible to produce opiate-like side effects. There was also evidence to suggest that butylthio[2.2.2] may cause less tolerance to the antinociceptive effects than morphine. Butylthio[2.2.2]'s nonopioid antinociceptive pharmacology, and its superior therapeutic window compared with oxotremorine's, suggests that butylthio[2.2.2] may be devoid of opiate-type abuse and dependence liability, and demonstrates the possibility to develop muscarinic analgesics without the classic muscarinic side-effect profile at analgesic doses. These data suggest that butylthio[2.2.2] would be a promising candidate for clinical development.