Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Central adrenergic pathways fulfill a diversity of functional roles. They are, for example, implicated in the modulation of nociception, motor behavior, and thermoregulatory mechanisms (Myers and Lee, 1989; Millan, 1997). The actions of norepinephrine (NE) are expressed via three families of adrenoceptor (AR): ₁, ₂, and  (Bylund et al., 1994; Hieble et al., 1995). ₁-ARs are principally coupled via G_q to activation of phospholipase C, and -ARs are positively coupled via G_s to adenylyl cyclase. ₂-ARs are predominantly and negatively coupled via G_i/G_o to adenylyl cyclase: in addition, they can suppress and enhance voltage-dependent calcium currents and inwardly rectifying potassium currents, respectively (Bylund et al., 1994; Aantaa et al., 1995; Hieble et al., 1995; Wade et al., 1999).

Three subtypes of ₂-AR have been cloned (_2A, _2B, and _2C),¹ which show contrasting patterns of distribution in mammalian central nervous system. Presynaptically, _2A-ARs predominate as autoreceptors, although there exists a minor population of _2C-ARs in the locus ceruleus [Nicholas et al., 1997; Millan et al., 2000a (accompanying article)]. At the postsynaptic level, _2B-ARs are largely restricted to the thalamus, although they are also expressed in the striatum and cerebellum of mice (Wang et al., 1996; Nicholas et al., 1997). _2C-ARs are present in high densities in the hippocampus, the olfactory system, the striatum, and the cerebral cortex (Rosin et al., 1996; Wang et al., 1996; Nicholas et al., 1997). However, _2A-ARs are the most prominent and broadly distributed ₂-AR subtype, being concentrated in the cortex, autonomic centers, amygdala, septum, hippocampus, and hypothalamusin the preoptic nucleus of which they mediate hypothermic actions (Myers and Lee, 1989; Quan et al., 1992; Rosin et al., 1996; Wang et al., 1996; Nicholas et al., 1997).

_2A-ARs also predominate in the dorsal horn, where they are largely expressed by intrinsic neurons, although an additional population is localized on the terminals of nocisponsive primary afferent fibers (Rosin et al., 1996; see Millan, 1997; Nicholas et al., 1997; Stone et al., 1998; Shi et al., 2000). _2B-ARs are present in only a low density in rat dorsal root ganglia and spinal cord (Millan, 1997; Nicholas et al., 1997; Shi et al., 2000), although they are more prominent in human dorsal horn (Stafford-Smith et al., 1995). Furthermore, in both species, there is relatively a minor population of _2C-ARs (Stafford-Smith et al., 1995; Stone et al., 1998; Shi et al., 2000). Like _2A-ARs, _2C-ARs are largely expressed by intrinsic neurons, although they also occur on primary afferent fiber terminals (Rosin et al., 1996; see Millan, 1997, 1999; Nicholas et al., 1997; Shi et al., 2000). Consistent with this organization, ₂-ARs fulfill an important role in the descending inhibition of nociceptive transmission (see Millan, 1997; Martin et al., 1999). Furthermore, spinal or systemic administration of ₂-AR agonists elicits antinociception both in animals (Fischer et al., 1991; Sullivan et al., 1992; see Millan, 1997) and humans (Eisenach et al., 1996, 2000; Hall et al., 2000). Studies in transgenic mice (Hunter et al., 1997; Lakhlani et al., 1997; see Kable et al., 2000) support pharmacological analyses (Millan et al., 1994; Zhang et al., 1998) indicating a primary role of (segmental) _2A-ARs in the mediation of antinociception, although _2C-ARs may also contribute (see Millan, 1997; Graham et al., 2000; Kable et al., 2000).

The above considerations underpin continuing interest in ₂-ARs as a target for novel therapeutic agents (Hieble et al., 1995). Concerning agonists, in addition to the treatment of hypertension and glaucoma, they have attracted much interest as potential analgesics (Hayashi and Maze, 1993; Millan, 1997; Williams et al., 1999). Furthermore, the combined antinociceptive, sedative, anxiolytic, hemodynamic-stabilizing, and myoprotective properties of ₂-AR agonists render them attractive in the perioperative environment for induction of anesthesia and sedation, as well as a reduction in opioid requirements (Hayashi and Maze, 1993; Martin et al., 1997; Hall et al., 2000). Although the prototypical, imidazoline ₂-AR partial agonist, clonidine (Bucaffusco, 1992; Eisenach et al., 1996), remains the most intensely studied agent in this context, the imidazole, dexmedetomidine, is of special interest in view of its more pronounced potency, selectivity, and efficacy at ₂-ARs (Fischer et al., 1991; Sullivan et al., 1992; Aantaa et al., 1993; Hall et al., 2000).

In the present studies, we compared the novel, spiroimidazoline, S18616, to dexmedetomidine and clonidine (Fig. 1). Herein, its receptor profile was characterized, including functional actions at cloned, h₂-AR subtypes. Furthermore, agonist properties at (postsynaptic) ₂-ARs mediating hypothermia and antinociception were examined (Bill et al., 1989; Millan, 1997). In the accompanying article, the influence of S18616 at ₂-AR autoreceptors and heteroceptors modulating monoaminergic transmission is described in relation to its sedative and anxiolytic properties. The involvement of ₂-ARs in the actions of S18616 was evaluated employing several, chemically diverse antagonists displaying marked selectivity for ₂- versus ₁-ARs: atipamezole, RX821,002, idazoxan (O'Rourke et al., 1994; Haapalinna et al., 1997; Newman-Tancredi et al., 1998), and BRL44408 (Young et al., 1989; Bylund et al., 1994; Millan et al., 1994; Renouard et al., 1994). Their actions were compared with those of two selective ₁- versus ₂-AR antagonists, prazosin and ARC219. Interestingly, BRL44408 behaves as a preferential antagonist of _2A- versus _2B/2C-ARs, whereas prazosin and ARC219 show an opposite pattern of preference for _2B/2C- versus _2A-ARs (Young et al., 1989; Bylund et al., 1994; Millan et al., 1994; Renouard et al., 1994). The use of these agents may, thus, provide insights concerning potential roles of individual ₂-AR subtypes in the actions of S18616, although definitive characterization will require further study.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Chemical structures of S18616, dexmedetomidine, and clonidine.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Binding Studies: Determination of Drug Affinities. The affinities of S18616 as compared with dexmedetomidine and clonidine at diverse binding sites were determined employing conventional procedures, most of which were extensively documented elsewhere (Millan et al., 2000b): conditions for the key sites are summarized in Table 1. In addition, affinities at cloned, human muscarinic receptor subtypes transfected into Chinese hamster ovary (CHO) cells were determined by use of the following radioligands: hM₁ ([³H]pirenzapine, 2.0 nM); hM₂ ([³H]AFDX384, 2.0 nM); hM₃ ([³H]4-DAMP, 0.2 nM); and hM₄ ([³H]4-DAMP, 0.2 nM). Isotherms were analyzed by nonlinear regression analysis, and IC₅₀ values were calculated using the program PRISM (Graphpad Software, San Diego, CA). IC₅₀ values were converted into K_i values in accordance with the equation: K_i = IC₅₀/(1 + L/K_d), where L corresponds to the radioligand concentration and K_d is the dissociation constant.

View this table: [in this window] [in a new window]   TABLE 1 Affinities (pKi values) of S18616, dexmedetomidine, and clonidine at diverse adrenoceptor subtypes and related sites Data are means of two to four determinations. In all cases, the S.E.M. (n  3) or range (n = 2) was 10% of the mean; i.e., 0.2 log units (pKi).

Agonist Properties at h_2A-, h_2B-, and h_2C-ARs: Induction of [³⁵S]GTPS Binding. The procedure employed has been described in detail previously (Newman-Tancredi et al., 1998; Millan et al., 2000b) for h_2A-ARs, and an essentially similar procedure was adopted herein for h_2B- and h_2C-AR subtypes expressed in CHO cells. [As determined with [³H]RX821,002, the binding density (B_max) for h_2A-, h_2B-, and h_2C-ARs, was 2.0, 1.3, and 1.0 pmol/mg, respectively.] Guanosine-5'-O-(3-[³⁵S]thio)triphosphate ([³⁵S]GTPS; 1000 Ci/mmol, Amersham Pharmacia Biotech, Les Ulis, France) was employed at a concentration of 0.1 nM. Samples (containing 50 µg of protein) were incubated for 60 min at 22°C. The buffer composition was as follows: HEPES (20 mM, pH 7.4), NaCl (100 mM), GDP (3 µM), and MgSO₄ (3 mM). Incubations were terminated by rapid filtration through GF/B filters (Whatman, Clifton, NJ) using a harvester (Brandel, Gaithersburg, MD). Radioactivity retained on the filters was quantified by liquid scintillation counting. The stimulatory actions of S18616, dexmedetomidine, and clonidine alone were compared with those of NE, the maximal effect of which was defined as 100%. Log effective concentrations₅₀ (pEC₅₀) were calculated. Antagonist properties of clonidine against fixed concentrations of NE were also evaluated at h_2B- and h_2C-ARs in view of its low efficacy at these sites: the IC₅₀ was determined and the K_b calculated (Newman-Tancredi et al., 1998). In additional antagonist studies, the influence of the selective antagonist, atipamezole, was determined against fixed concentrations of S18616, dexmedetomidine, andwhere sufficiently active, clonidineat h_2A-, h_2B-, and h_2C-ARs. Atipamezole concentration curves were analyzed to yield K_b values.

Agonist Actions at p_2A-AR/Fusion Proteins. The methodologies employed for construction of the p_2A-AR-containing fusion proteins, as well as for their expression and characterization, were detailed previously (Wise et al., 1997). Briefly, rat (wild-type, Cys-351) Gi1 was coupled to the p_2A-AR (generously provided by L. E. Limbird, Vanderbilt University, Nashville, TN) and ligated into the KpnI and EcoRI sites of the eukaryotic expression vector pcDNA to yield the p_2A-AR-Gi1 fusion protein in pCDNA3. COS-7 cells were grown to confluency (18-24 h) before transfection (over 5 h) with the pCDNA3 (2.5-2.8 µg) using LipofectAMINE. Cells were harvested 48 h after transfection. In certain experiments, cells were treated for the 24 h before harvesting with pertussis toxin (50 ng/ml). Cells were stored at 80°C, and high affinity GTPase assays were performed on membrane-containing particulate fractions as done previously (Wise et al., 1997). Nonspecific GTPase activity was assessed by parallel assays containing GTP (100 µM). All experiments were carried out at least three times on membranes prepared from individual cell transfections.

Agonist Properties at h_1A-ARs: Depletion of Membrane-Bound [³H]PI. The influence of S18616 as compared with NE, dexmedetomidine, and clonidine on phospholipase C activity was characterized in CHO cells transfected with h_1A-ARs essentially as described previously for h5-HT_2C cells (Cussac et al., 2000). Cells were grown in adherent culture in 225-cm² flasks with UltraCHO medium (BioWhittaker, Walkersville, MD) containing sodium pyruvate (1 mM), dialyzed fetal calf serum (0.1%), and geneticin (400 µg/ml). At 80% to 90% confluency, cells were labeled with 2 µCi/ml [³H]myo-inositol (10-20 Ci/mmol) for 24 h in serum-free UltraCHO medium. Adherent cells were rinsed twice in Krebs-LiCl buffer (NaH₂PO₄, 15.6 mM, pH 7/NaCl, 120 mM/KCl, 4.8 mM/MgSO₄, 1.2 mM/CaCl₂, 1.2 mM/glucose, 0.6% (w/v)/bovine serum albumin, 0.04% (w/v)/LiCl, 10 mM), scraped from the flask, and washed twice again by slow centrifugation. The cells were then resuspended in Krebs-LiCl and left to stand for 15 min at 37°C before use. [Inclusion of LiCl in the buffer allows measurement of depletion of a fixed pool of membrane-bound [³H]phosphatidyl inositol ([³H]PI): Thus, the present methodology detects primarily the membrane content of phosphatidyl inositol, phosphatidyl inositol-4-phosphate, and phosphatidyl inositol-4,5-biphosphate, the sum of which is referred to herein as "[³H]PI" for simplicity.] Cells were incubated in 96-well plates at 37°C for 30 min with incremental concentrations of NE, S18616, dexmedetomidine, or clonidine in Krebs-LiCl buffer. In the absence of agonists, ~40,000 dpm were typically detected, as compared with ~25,000 in the presence of a maximally effective concentration of NE (30 µM). Agonist efficacies were expressed as a percentage of the effect observed with NE (30 µM). For antagonist studies, cells were preincubated for 15 min with prazosin before the addition of NE (30 µM) or S18616 (1 µM). Assays were stopped by addition of 0.4 ml of methanol/HCl (88 ml of 100% methanol + 12 ml of HCl, 1 N), and cells were stored at 20°C for at least 2 h to facilitate cell lysis. The 96-well plates were then sonicated for 2 min, and membranes were recovered using a harvester (Filtermate; Packard, Meriden, CT) by filtration through GF/B filters impregnated with 0.1% (v/v) polyethyleneimine followed by three washes with distilled, deionized water. Radioactivity was determined using a Top-Count microplate scintillation counter (Packard).

Partial Agonist Properties at h5-HT_1A and h5-HT_1D Receptors: Induction of [³⁵S]GTPS Binding. The procedures employed were described in detail previously (Newman-Tancredi et al., 1998; Millan et al., 2000b). Membranes of CHO cells expressing h5-HT_1A or h5-HT_1D (15 µg of protein) receptors were incubated for 20 min at 22°C with S18616 or 5-HT in a buffer containing [³⁵S]GTPS (0.1 nM), GDP (3 µM), HEPES (20 mM, pH 7.40), MgCl₂ (3 mM), and NaCl (100 mM). The efficacy of S18616 was defined relative to that of a maximally effective concentration of 5-HT (10 µM, 100%). For antagonist studies, h5-HT_1A membranes were preincubated for 30 min with WAY100,635, then S18616 was added at a concentration of 100 µM. Experiments were completed by rapid filtration through Whatman GF/B filters. Radioactivity retained on the filters was quantified by liquid scintillation counting.

Antagonist Properties at hM₁ and hM₄ Receptors: Inhibition of Oxotremorine and Carbachol-Induced [³⁵S]GTPS Binding. Membranes (4-7 µg of protein) of hM₁ or hM₄ receptors were incubated for 60 min at 22°C with oxotremorine (hM₁), carbachol (hM₄), and/or S18616 in a buffer containing HEPES (20 mM, pH 7.4), GDP (0.1 µM and 1 µM for hM₁ and hM₄ receptors, respectively), MgCl₂ (50 mM and 10 mM for hM₁ and hM₄ receptors, respectively), and NaCl (150 mM) for 15 min before addition of [³⁵S]GTPS (0.5 nM). Nonspecific binding was defined with [³⁵S]GTPS (10 µM). Incubations were performed in 96-plates, and membranes were recovered using the Filtermate harvester (Packard) by filtration through GF/B filters impregnated with water followed by three washes with cold incubation buffer. Radioactivity was determined using a the Top-Count microplate scintillation counter (Packard).

Animals. The in vivo studies employed male Wistar rats of 200 to 250 g and NMRI mice of 20 to 25 g (Iffa Credo, l'Arbresles, France) housed in sawdust-lined cages with unrestricted access to standard chow and water. There was a 12-h/12-h light/dark cycle with lights on at 7:30 AM. Laboratory temperature and humidity were 21 ± 0.5°C and 60 ± 5%, respectively. Animals were adapted to laboratory conditions for at least a week before testing. All animal use procedures conformed to international European ethical standards (86/609-EEC) and the French National Committee (décret 87/848) for the care and use of laboratory animals.

Influence of S18616 Compared with Dexmedetomidine and Clonidine on Core Temperature. As detailed previously (Millan et al., 2000b), core (rectal) temperature (CT) was determined in lightly restrained mice or rats with a digital thermistoprobe. For studies of induction of hypothermia, basal CT was measured; S18616, dexmedetomidine, clonidine, or vehicle was administered; and CT was redetermined 90 min later. In antagonist studies, mice received injections of vehicle or antagonist followed, 30 min later, by vehicle or S18616 (0.005 mg/kg, s.c.) and CT was redetermined 30 min later. The difference to pretreatment values was calculated in each case. Data were analyzed by ANOVA followed by Dunnett's test. For antagonist studies, the percentage inhibition was computed as follows:
and ID₅₀ plus 95% confidence limits (CL) were calculated.

Antinociceptive Properties of S18616 Compared with Dexmedetomidine and Clonidine in Mice. All algesiometric procedures in mice were performed as described in detail elsewhere (Millan et al., 1994, 1996). They may be briefly summarized as follows. For evaluation of the tail-flick (TF) response, mice were lightly restrained in paper wadding. For thermal stimuli, a light beam was focused on the tail (Tail-Flick Apparatus, Columbus Instruments, Columbus, OH) at two points 1 and 2 cm from the tip. The mean latency to TF was determined with an upper limit of 8 s. For mechanical stimuli, an incremental weight was applied to the tail at two points 3 and 4 cm, respectively, from its tip via a blunt piston (Randall and Sellito apparatus, Ugo-Basile, Varese, Italy). The mean pressure to removal was determined with a cut-off of 250 g. The formalin test was performed on mice placed in observation cylinders 11 cm in diameter and 15 cm in height. Formalin (20 µl, 5% in saline) was injected into the skin of the dorsal surface of the right hind-paw using a 26-gauge needle. Mice were immediately placed in the cylinders, and the time spent licking the paw was counted both 0 to 5 min ("Early Phase") and 35 to 50 min ("Late Phase") after injection. For the abdominal constriction procedure, mice were placed in the observation cylinders immediately after i.p. injection of acetic acid (10 ml/kg, 0.6% in distilled water), and the number of abdominal constrictions were counted over 5 to 15 min. In the hot-plate protocol, mice were placed on the center of a circular copper plate (diameter 19 cm) maintained at 55 ± 0.05°C and surrounded by a perplex cylinder (height, 18 cm). The latency to lick either hind-paw was determined with a cut-off of 60 s. S18616, dexmedetomidine, clonidine, or vehicle were administered 30 min before each procedure. For antagonist studies with the hot-plate procedure, vehicle or antagonists were administered 30 min before S18616 or vehicle. Data were analyzed by ANOVA followed by Dunnett's test and ID₅₀ values (95% CL) or active dose (AD)₅₀ values (95% CL) were calculated. For the time course of action on oral administration in the abdominal constriction test, data were analyzed by a two-way ANOVA with "drug" and "time" as the factors.

Antinociceptive Properties of S18616 Compared with Dexmedetomidine and Clonidine in Rats. The algesiometric procedures employed in rats have all been detailed previously (Millan et al., 1994, 1996). The TF procedures were performed essentially as described above for mice employing the same apparatus. The cut-off was 8 s and 250 g for thermal and mechanical stimuli, respectively. Spontaneous tail-flicks (STFs) were determined in rats loosely restrained in horizontal, opaque, plastic cylinders over a period of 5 min. They were defined as the upward, rapid movement of the tail to a level higher than that of the body axis. S18616, dexmedetomidine, dopamine, and vehicle were injected 30 min before 8-OH-DPAT (0.63 mg/kg, s.c.) and STFs measured 10 min thereafter. For the vocalization test, rats were likewise restrained in cylinders and a bipolar electrode applied to the tail. Square-wave current pulses of incremental voltage were delivered (1.0-s duration, 20 ms, 50 Hz), and the threshold to vocalization was determined. S18616, dexmedetomidine, clonidine, or vehicle were administered 30 min before each procedure. Data were analyzed by ANOVA followed by Dunnett's test. For the TF procedure, AD₅₀ values (95% CL) were calculated, and for the vocalization and TF procedures, ID₅₀ values (95% CL) were determined.

Drugs. All drug doses are shown in terms of the base. Drugs were dissolved in sterile water, and if necessary, a few drops of lactic acid were added, and pH was adjusted to as close to normality (>5.0) as possible. Drugs were administered in an injection volume of 1 ml/kg (rats, s.c.) or 10 ml/kg (mice, s.c. and rats, p.o.). Drug sources, salts, and structures were as follows. ARC239 HCl {2-(2-(4-O-methoxyphenyl)piperazine)-1-ylethyl-4,4-dimethyl-1,3-(2H,4H)-isoquinolinedione} was from Boehringer (Ingelheim am Rhein, Germany). Carbachol, clonidine HCl, norepinephrine HCl, oxotremorine methiodide, prazosin HCl, and serotonin were from Sigma (Chesnes, France). 8-OH-DPAT HBr [(±)-8-dihydroxy-2-(di-n-propylamino)tetralin] and RX821,002 HCl [2-(2-methoxy-1,4-benzodioxan-2-yl)-2-imidazoline] were from Research Biochemicals International (Natick, MA). Atipamezole HCl, BRL44408 (2-(2H-(1-methyl-1,3-dihydroisoindole)methyl)-4,5-dihydroimidazoline), dexmedetomidine tartrate, idazoxan HCl, and WAY100,635 3HCl (N-{2-[4-(2-methoxyphenyl)-1-piperazinyl]ethyl}-N-(2-pyridinyl)cyclohexanecarboxamide) were all synthesized by Servier chemists (A. Cordi, C. Malen, and J.-L. Péglion). (S)-S18616 HCl {(S)-spiro[(1-oxa-2-amino-3-azacyclopent-2-ene)-4,2'-(8'-chloro-1',2',3',4'-tetrahydronaphthalene)]} was prepared by resolution of the racemic mixture (S17824; Cordi et al., 1995) through repeated ethanol crystallizations (2 ×) in the presence of half an equivalent each of (+)-di-para-toluoyl tartaric acid and HCl. The R enantiomer (S18574) was obtained by the same procedure using ()-di-para-toluoyl tartaric acid as resolving agent.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

S18616: An Optically Pure ₂-AR Agonist. The structure of S18616 is indicated in Fig. 1 in comparison to those of dexmedetomidine and clonidine. It may be seen that S18616 is optically pure. S18616 was selected for study in preference to the other enantiomer, S18574, because it presents higher affinity for _2A-ARs. Thus, the affinities at native rat _2A-ARs for S18616, S18574, and racemic S17824 were 9.8, 8.3, and 9.5, respectively.

Affinities at Native Rat and Cloned Human ₂-ARs (Table 1). S18616 displayed high affinity for both cortical rat ₂-ARs, predominantly of the r_2A-AR subtype (Renouard et al., 1994), and native r_2A-ARs, the species homolog of h_2A-ARs (see the Introduction). The affinity of dexmedetomidine for cortical ₂-ARs was ~10-fold lower than that of S18616, whereas clonidine showed about 100-fold lower affinity. At both native h_2A-ARs in platelets, as well as cloned h_2A-ARs expressed in both mammalian (CHO) and nonmammalian (Sf9) systems, the affinity of S18616 was very pronounced. At these sites, dexmedetomidine was 10- to 20-fold less potent than S18616, and clonidine was 100-fold less potent. The affinity of S18616 for CHO-expressed h_2B- and h_2C-ARs was nearly as pronounced as its affinity for CHO-expressed h_2A-ARs, although its affinity at Sf9-expressed h_2C-ARs was slightly lower than for h_2A-ARs expressed in this system. Dexmedetomidine showed similar affinity at CHO-expressed h_2B-ARs and h_2A-ARs and similar affinity for h_2C- and h_2A-ARs in both CHO and Sf9 expression systems. A comparable pattern of data was obtained for clonidine, although it was a markedly less potent ligand.

Affinities at Native Rat and Cloned Human ₁-ARs, -ARs, and Imidazoline Binding Sites (Table 1). The affinity of S18616 for native, rat cerebral ₁-ARs was substantially lower than its affinity for rat, cerebral ₂-ARs. Dexmedetomidine also showed marked selectivity, whereas the degree of separation was only modest for clonidine. For S18616, dexmedetomidine, and clonidine, in each case, their affinities were somewhat higher for native, rat _1A-ARs versus cerebral ₁-ARs. On the other hand, their affinities for native rat _1B-ARs were similar to those observed at cerebral ₁-ARs. At cloned h_1A-ARs, S18616 showed marked affinity, which was 22-fold lower than that for h_2A-ARs expressed in the same system (CHO cells). The affinities of dexmedetomidine and clonidine at h_1A-ARs relative to their affinities at h_2A-ARs were 80- and 11-fold lower, respectively. The affinity of S18616 at cloned h_1B- and h_1D-ARs was higher than those of dexmedetomidine and clonidine. Nevertheless, for both of these sites, S18616 still showed a separation to affinities at h_2A-ARs of ~100-fold. For h_1D-ARs, this selectivity ratio was superior to those of both dexmedetomidine and clonidine. S18616 showed modest affinity for native I₁ and I₂ sites, in each case >100-fold lower than its affinity for native, rat ₂-ARs. The affinity of dexmedetomidine for imidazoline I₁ and I₂ sites was also markedly lower than its affinity for native, rat ₂-ARs, although, in the later case, it was less selective than S18616. Although clonidine was weakly active at I₂ sites, it showed equivalent affinity for I₁ and rat ₂-ARs. S18616 did not show significant affinity for cloned, h₁- or h₂-ARs. S18616, dexmedetomidine, and clonidine all showed negligible (<6.0) affinities for NE uptake sites, as well as for monoamine oxidase (MAO) A and MAO B.

Affinities at Nonadrenergic Receptor Types. S18616 displayed negligible affinity (pK_i < 6.0) for cloned hD₁, hD₂, hD₃, and hD₄ receptors and for dopamine uptake sites. It showed mild affinity for native rat 5-HT_1A receptors (6.7 ± 0.1) and cloned h5-HT_1A receptors (6.8 ± 0.1), although still >100-fold lower than its affinity at ₂-ARs and h_2A-ARs, respectively. The affinity of S18616 for h5-HT_1D receptors was modest (7.6 ± 0.2), and it showed low affinity (<5.0-6.3) at all other 5-HT receptor types examined (5-HT_1B, h5-HT_1B, 5-HT_2A, h5-HT_2A, h5-HT_2C, 5-HT₃, and 5-HT₄). The affinity of S18616 was negligible (<6.0) for GABA_A, GABA_B, sigma₁, µ-opioid, histamine₁, histamine₂, as well as NK₁ and NK₂ receptors. It showed mild affinity for cloned hM₁ (6.7 ± 0.1), hM₂ (6.3 ± 0.1), hM₃ (6.4 ± 0.1), and hM₄ (6.2 ± 0.1) receptors, albeit >100-fold lower than at h_2A-ARs. At a further range of receptors, ion channels, and enzymes (>20 sites), the affinity of S18616 was negligible (<5.0).

Agonist Properties of S18616 Compared with Dexmedetomidine and Clonidine at Cloned h₂-AR Subtypes (Table 2 and Fig. 2). At cloned h_2A-ARs, NE elicited a robust, maximal (2.5-fold) stimulation of [³⁵S]GTPS binding, which was defined as "100%" for the purposes of quantifying the actions of other drugs. S18616 elicited a concentration-dependent increase in [³⁵S]GTPS binding. Its maximal effect (E_max) was ~50% that of NE, although it was >100-fold more potent. Similarly, dexmedetomidine behaved as a partial agonist and was substantially more potent that NE. The efficacy of clonidine was slightly (although not significantly) less than that of S18616 and dexmedetomidine, and it was substantially less potent. At h_2B-ARs, S18616 and dexmedetomidine potently and fully mimicked the stimulatory influence of NE. In contrast, clonidine showed low efficacy (significantly weaker than S18616) at h_2B-ARs. Indeed, in additional experiments, clonidine concentration dependently suppressed NE-induced stimulation of [³⁵S]GTPS binding at h_2B-ARs with a pK_b of 7.02 ± 0.08 (n = 4, not shown). At h_2C-ARs, S18616 behaved as a potent partial agonist relative to NE. Dexmedetomidine was likewise a potent agent and showed efficacy intermediate between those of NE and S18616. In distinction, compared with S18616, clonidine displayed significantly weaker (partial) agonist activity at h_2C-ARs. Indeed, it blocked the actions of NE at these sites with a pK_b of 6.9 ± 0.06 (n = 4, not shown).

View this table: [in this window] [in a new window]   TABLE 2 Efficacies of S18616, dexmedetomidine, and clonidine at cloned human 2-ARs as determined by [35S]GTPS binding The effect of norepinephrine (10 µM) was defined as 100%. Data are means ± S.E.M. of at least three determinations.

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Agonist actions of S18616 compared with norepinephrine, dexmedetomidine, and clonidine at cloned CHO-expressed h2A-, h2B-, and 2C-ARs, as determined by stimulation of [35S]GTPS binding. A, B, and C, concentration-dependent activation of [35S]GTPS binding by norepinephrine (NE), S18616, dexmedetomidine (DMT), and clonidine alone. D, E, and F, concentration-dependent blockade of the actions of NE, S18616, dexmedetomidine, and clonidine by the selective 2-AR antagonist atipamezole. Data are from a representative experiment repeated on at least three occasions.

Blockade of the Agonist Actions of S18616, Dexmedetomidine, and Clonidine at h_2A-, h_2B-, and h_2C-ARs by Atipamezole (Table 3 and Fig. 2). The specificity of drug actions in stimulating [³⁵S]GTPS binding was confirmed by employing the selective antagonist, atipamezole. Atipamezole concentration dependently and completely blocked the stimulation of [³⁵S]GTPS binding at h_2A-ARs by S18616, dexmedetomidine, clonidine, and NE with potencies corresponding to its high affinity (pK_i, 9.5) at these sites. Similarly, at h_2B-ARs, for which atipamezole also showed high affinity (pK_i = 9.0), the stimulatory actions of S18616, dexmedetomidine, and NE were abolished. At h_2C-ARs, for which the affinity of atipamezole was 9.2, the stimulation of [³⁵S]GTPS binding by S18616, dexmedetomidine, clonidine, and NE was likewise potently abolished.

View this table: [in this window] [in a new window]   TABLE 3 Antagonism by atipamezole of the induction of [35S]GTPS binding by S18616, dexmedetomidine, clonidine, and norepinephrine at h2A-, h2B-, and h2C-ARs Values are pKb. Data are means ± S.E.M. of at least three determinations.

Agonist Properties of S18616 Compared with Dexmedetomidine and Clonidine: p_2A-Gi1 Fusion Protein (Fig. 3). In COS-7 cells transiently expressing a fusion protein constructed between p_2A-ARs and a (wild-type) Gi1-protein (Wise et al., 1997; Milligan, 2000), NE markedly enhanced GTPase activity, an action expressed with a pEC₅₀ (see Fig. 4, mean ± range) of 6.31 ± 0.30. S18616 mimicked the stimulation by NE of GTPase activity. Although its maximal effect was less pronounced (64.0 ± 5.6) than that of NE (defined as 100%), it was considerably more potent, yielding a pEC₅₀ of (mean ± range) 8.94 ± 0.07. Dexmedetomidine was active with an efficacy (69.2 ± 5.6) comparable to that of S18616, but with lower potency: pEC₅₀ (mean ± range) = 7.85 ± 0.22. Clonidine also behaved as a partial agonist (57.1 ± 7.0 efficacy), exerting its actions with a pEC₅₀ (mean ± range) of 7.47 ± 0.19. After pretreatment with pertussis toxin, the actions of NE, S18616, dexmedetomidine, and clonidine were all markedly attenuated. Residual percentage stimulation was 24.1 ± 5.8, 12.6 ± 0.01, 12.9 ± 2.8, and 12.3 ± 0.4, respectively.

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Agonist actions of S18616 compared with norepinephrine, dexmedetomidine, and clonidine at a COS-7-transfected p2A-AR-Gi1 fusion protein, as determined by a GTPase assay. A, norepinephrine; B, S18616; C, dexmedetomidine; and D, clonidine. Closed and open symbols indicate no pretreatment or pretreatment with pertussis toxin (25 µg), respectively. Data are from a representative experiment. Similar data were obtained in a separate experiment.

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Agonist actions of S18616 compared with norepinephrine, dexmedetomidine, and clonidine at cloned, CHO-expressed h1A-ARs as determined by depletion of membrane-bound [3H]PI. A, concentration-dependent depletion of [3H]PI; B, concentration-dependent blockade of the actions of S18616 compared with norepinephrine by the selective 1-AR antagonist prazosin. Data are from a representative experiment repeated on three occasions.

Agonist Properties of S18616 Compared with Dexmedetomidine and Clonidine at Cloned h_1A-ARs (Fig. 4). At cloned h_1A-ARs, NE elicited a robust and concentration-dependent reduction in [³H]PI levels with a pEC₅₀ of 6.51 ± 0.09. S18616 similarly provoked [³H]PI depletion, displaying a pEC₅₀ of 7.80 ± 0.05 and a maximal effect of 79.0 ± 1.9% relative to that of NE (100.1 ± 2.7). The influence of dexmedetomidine on [³H]PI levels was less pronounced with a pEC₅₀ of 6.67 ± 0.06 and a maximal effect of 63.2 ± 6.1%. Clonidine depleted [³H]PI levels with a pEC₅₀ of 6.69 ± 0.13 and a maximal effect of 40.7 ± 5.9%. Prazosin, which itself did not modify PI levels, abolished the actions of NE and S18616 with pK_b values of 9.11 ± 0.11 and 8.92 ± 0.10, respectively, indicating specific mediation of the actions of agonists by _1A-ARs.

Partial Agonist Properties at Cloned h5-HT_1A and at h5-HT_1D Receptors. S18616 concentration dependently enhanced [³⁵S]GTPS binding at h5-HT_1A receptors with an efficacy (63.7 ± 3.5%, 100 µM) inferior to that of 5-HT (defined as 100%) and at markedly higher concentrations, pEC₅₀ ~5.2 (S18616) as compared with 7.40 ± 0.05 for 5-HT. The stimulatory influence of S18616 was abolished by WAY100,635 with a pK_b of 9.1 ± 0.1, close to that for blockade of the action of 5-HT: pK_b = 8.5 ± 0.1. This action of S18616 was, thus, only expressed at concentrations markedly higher than those required for activation of h₂-AR subtypes. Similarly, relative to 5-HT (pEC₅₀ = 8.9 ± 0.1, 100%) S18616 acted as a weak partial agonist at h5-HT_1D receptors (pEC₅₀ = 7.7 ± 0.1; efficacy = 56.8 ± 3.3).

Antagonist Properties at Cloned hM₁ and hM₄ Receptors. In contrast to the prototypical muscarinic agonist, carbachol, which elicited a 1.6-fold elevation in [³⁵S]GTPS binding (pEC₅₀ = 5.62 ± 0.08) over a broad range of concentrations (10¹⁰-10³), at hM₄ receptors stably transfected into CHO cells, S18616 did not show agonist activity. Rather, it concentration dependently and completely abolished the action of carbachol (10 µM) with a potency (pK_b = 5.90 ± 0.04) lower than that of the prototypical antagonist, scopolamine (9.35 ± 0.06). Similarly, S18616 did not enhance [³⁵S]GTPS binding at hM₁ receptors relative to the agonist, oxotremorine (pEC₅₀ = 5.19 ± 0.06, 1.4-fold elevation), the actions of which were blocked by S18616 with a pK_b of 6.33 ± 0.14, compared with scopolamine = 9.48 ± 0.14. The antagonist properties of S18616 at hM₁ and hM₄ receptors were, thus, expressed only at concentrations considerably superior to those required for activation of h_2A-ARs.

Influence of S18616 Compared with Dexmedetomidine and Clonidine on CT (Fig. 5). S18616 potently elicited a pronounced reduction in CT in both mice and rats. This action was mimicked by dexmedetomidine. It was also less potently mimicked by clonidine, which elicited only a mild hypothermia in mice. Pretreatment of mice with atipamezole, idazoxan, RX821,002, and BRL44408 dose dependently inhibited the hypothermic action of S18616 with ID₅₀ values (95% CL) as follows: 0.08 (0.02-0.35), 0.12 (0.04-0.34), 0.05 (0.01-0.17), and 0.9 (0.4-2.1) mg/kg, s.c., respectively. Idazoxan and BRL44408 themselves elicited a modest decrease in CT. This hypothermic action of S18616 in mice was not attenuated by prazosin nor ARC239, however, which both reduced CT alone (not shown).

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Induction of hypothermia by S18616 compared with dexmedetomidine and clonidine in mice and rats. A, induction of hypothermia in rats; B, induction of hypothermia in mice; C to F, dose-dependant blockade of S18616-induced hypothermia in mice by atipamezole (C), idazoxan (D), RX821,002 (E), and BRL44408 (F). Data are means ± S.E.M. N = 5 per value. ANOVA results are as follows. A: S18616, F(4,32) = 56.6, P < .001; dexmedetomidine, F(4,29) = 56.9, P < .001; and clonidine, F(4,40) = 67.9, P < .001. B: S18616, F(4,21) = 66.2, P < .001; dexmedetomidine, F(5,26) = 37.0, P < .001; and clonidine, F(4,21) = 15.6, P < .001. C: influence of atipamezole versus S18616, F(3,26) = 16.7, P < .001, and effect of atipamezole alone, F(3,24) = 0.3, P > .05. D: influence of idazoxan versus S18616, F(4,23) = 19.0, P < .001, and effect of idazoxan alone, F(4,22) = 5.3, P < .01. E: influence of RX821,002 versus S18616, F(4,21) = 9.1, P < .001, and effect of RX821,002 alone, F(4,21) = 1.0, P > .05. F: influence of BRL44408 versus S18616, F(4,28) = 17.6, P < .001, and effect of BRL44408 alone, F(4,26) = 14.8, P < .001. Asterisks indicate significance of differences to respective control values in Dunnett's test. *P < .05.

Antinociceptive Properties of S18616 Compared with Dexmedetomidine and Clonidine (Fig. 6; Tables 4 and 5). In algesiometric models involving chemical stimuli, S18616 potently, dose dependently, and completely suppressed both the early and late phases of formalin-induced hind-paw licking and reduced abdominal constrictions in mice. These actions were expressed over a similar dose range. Dexmedetomidine was active in these procedures at slightly higher doses. Clonidine was also effective, although at doses some 10- to 20-fold higher than those of S18616. In the vocalization test, S18616 and dexmedetomidine displayed robust activity at similar doses, whereas clonidine was effective only over a substantially (20-fold) higher dose range. For each drug, active doses were considerably higher than those required in tests involving chemical stimuli. S18616 and, at slightly higher doses, dexmedetomidine dose dependently suppressed the TF response to mechanical stimulation in both mice and rats. At doses higher than those of S18616, clonidine was similarly active in these paradigms. 8-OH-DPAT-induced STFs in rats were potently and dose dependently diminished by both S18616 and dexmedetomidine. In this model, clonidine was effective only at doses 100-fold higher than those of S18616, and it attained only submaximal inhibition. S18616 and dexmedetomidine were fully active in the mouse hot-plate test, whereas clonidine displayed only submaximal activity. On the other hand, S18616, dexmedetomidine, and clonidine showed similar potency and maximal effects in inhibiting the TF response to a thermal stimulus in mice and rats.

View larger version (29K): [in this window] [in a new window]   Fig. 6.   Activity of S18616 compared with dexmedetomidine and clonidine in diverse models of antinociceptive activity. A, formalin test, late phase; B, abdominal constriction test; C, vocalization test; D, tail-flick response to a mechanical stimulus, mice; E, tail-flick response to a mechanical stimulus, rat; F, spontaneous tail-flicks, rat; G, hot-plate test, mice; H, tail-flick response to a thermal stimulus, mice; and I, tail-flick heat response to a thermal stimulus, rat. All data are expressed in absolute values, with the exception of the vocalization test, for which percentage values are expressed relative to basal thresholds (defined as 100%). Data are means ± S.E.M. N = 4 per value. ANOVA results are as follows. A: S18616, F(4,17) = 57.1, P < .001; dexmedetomidine, F(3,17) = 19.6, P < .001; and clonidine, F(4,20) = 10.6, P < .001. B: S18616, F(3,27) = 103.8, P < .001; dexmedetomidine, F(3,15) = 21.0, P < .001; and clonidine, F(3,17) = 73.2, P < .001. C: S18616, F(4,24) = 12.6, P < .001; dexmedetomidine, F(3,24) = 10.5, P < .001; and clonidine, F(4,25) = 23.3, P < .001. D: S18616, F(4,24) = 40.0, P < .001; dexmedetomidine, F(5,27) = 5.8, P < .001; and clonidine, F(3,18) = 9.6, P < .001. E: S18616, F(5,22) = 12.8, P < .001; dexmedetomidine, F(5,25) = 15.7, P < .001; and clonidine, F(5,55) = 15.4, P < .001. F: S18616, F(3,20) = 16.1, P < .001; dexmedetomidine, F(5,29) = 5.9, P < .001; and clonidine, F(4,22) = 1.5, P > .05. G: S18616, F(5,39) = 30.1, P < .001; dexmedetomidine, F(4,25) = 42.1, P < .001; and clonidine, F(5,47) = 6.7, P < .001. H: S18616, F(4,24) = 10.9, P < .001; dexmedetomidine, F(5,27) = 19.7, P < .001; and clonidine, F(3,18) = 24.0, P < .001. I: S18616, F(5,19) = 19.7, P < .001; dexmedetomidine, F(5,27) = 9.2, P < .001; and clonidine, F(5,53) = 7.7, P < .001. Asterisks indicate significance of differences to respective vehicle values in Dunnett's test. *P < .05.

Influence of Antagonists on the Antinociceptive Actions of S18616 in the Hot-Plate Procedure (Fig. 7). The hot-plate procedure was selected for a characterization of the involvement of ₂-ARs in S18616-induced antinociception. The antinociceptive properties of S18616 (0.08 mg/kg, s.c.) were potently and dose dependently abolished by atipamezole, idazoxan, RX821,002, and BRL44408 with ID₅₀ values (95% CL) as follows: 0.3 (0.2-0.5), 0.3 (0.1-0.7), 0.07 (0.02-0.25), and 1.5 (0.5-4.1) mg/kg, s.c., respectively. In contrast, prazosin (0.63 mg/kg, s.c.) and ARC239 (10.0) did not (% maximal observed inhibition = 0 in both cases) modify the antinociceptive actions of S18616 (not shown). Furthermore, in the presence of atipamezole and RX821,002, dose-response curves for S18616-induced antinociception were significantly displaced to the right. AD₅₀ (95% CL) for S18616: vehicle, 0.05 (0.01-0.23); atipamezole, 0.5 (0.2-1.8); and RX821,002, 0.3 (0.1-0.5) mg/kg, s.c.

View larger version (24K): [in this window] [in a new window]   Fig. 7.   Influence of 2-AR antagonists on the antinociceptive properties of S18616 in the hot-plate test in mice. A and B, dose-dependent antagonism by the 2-AR antagonists, idazoxan, atipamezole, RX821,002, and BRL44408; C and D, displacement to the right of the dose-response curve for S18616-induced antinociception by atipamezole and RX821,002, respectively. Data are means ± S.E.M. N = 5 per value. ANOVA results are as follows. A: RX821,002, F(5,27) =19.7, P < .001, and idazoxan, F(4,24) = 6.1, P < .01. B: atipamezole, F(3,19) = 20.2, P < .001, and BRL44408, F(5,29) = 30.5, P < .001. C and D: influence of S18616 alone, F(6,35) = 11.3, P < .001; influence of atipamezole versus S18616, F(6,28) = 9.9, P < .001; and influence of RX821,002 versus S18616, F(6,28) = 18.7, P < .001. Asterisks indicate significance of the difference to respective vehicle values in Dunnett's test. *P < .05.

Antinociceptive Actions of S18616 Compared with Dexmedetomidine and Clonidine on Oral Administration in Mice (Table 6 and Fig. 8). In preliminary metabolic studies, it was established that S18616 has an intense and distinctive route of degradation in the rat but not the mouse or higher species. Total bioavailability was, thus, ~40% in the mouse compared with ~4% in the rat (M. Bertrand, unpublished observation). For this reason, the activity of S18616 upon oral administration was characterized in mice. As shown in Table 5, S18616 potently expressed its antinociceptive properties upon oral injection. The separation between p.o. and s.c. doses over the six parameters varied from 1.3 to 13.3 with a median of 2.5. Dexmedetomidine was also active by the oral route, albeit less potently. Its range of p.o. to s.c. potencies was 1.5 to 42.0 with a median of 19. Clonidine was active at doses slightly higher than those active via the s.c. route (range = 2.0-22.7 and median = 2.5). In Fig. 8, it may also be seen that S18616 expressed its antinociceptive properties upon oral administration over a prolonged time course, with a significant effect from 30 min to 8 h, administered at a dose of 0.16 mg/kg.

View larger version (25K): [in this window] [in a new window]   Fig. 8.   Antinociceptive properties of S18616 compared with dexmedetomidine and clonidine upon oral administration in mice. A, formalin, late phase; B, abdominal constriction test; C, time course of action in the abdominal constriction test; D, tail-flick response to a mechanical stimulus; E, tail-heat response to a thermal stimulus; and F, hot-plate test. Data are means ± S.E.M. N = 4 per value. ANOVA results are as follows. A: S18616, F(4,22) = 14.5, P < .001; dexmedetomidine, F(4,20) = 7.3, P < .001; and clonidine, F(5,30) = 4.0, P < .01. B: S18616, F(4,31) = 61.7, P < .004; dexmedetomidine, F(4,26) = 23.2, P < .001; and clonidine, F(6,36) = 29.4, P < .001. C: S18616, influence of S18616, F(1,72) = 539.2, P < .001, influence of time, F(6,72) = 8.5, P < .001, and interaction, F(6,72) = 13.3, P < .001; dexmedetomidine, influence of dexmedetomidine, F(1,69) = 22.1, P < .001, influence of time, F(6,69) = 5.7, P < .001, and interaction, F(6,69) = 5.2, P < .001. D: S18616, F(4,31) = 16.5, P < .001; dexmedetomidine, F(3,12) = 2.5, P > .05; and clonidine, F(3,17) = 43.9, P < .001. E: S18616, F(4,31) = 9.9, P < .001; dexmedetomidine, F(3,12) = 2.8, P > .05; and clonidine, F(3,17) = 46.4, P < .001. F: S18616, F(5,25) = 22.1, P < .001; dexmedetomidine, F(4,44) = 29.0, P < .001; and clonidine, F(5,30) = 6.2, P < .001. Asterisks indicate significance of differences to respective vehicle values in Dunnett's test. *P < .05.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Agonist Properties at h₂-ARs. S18616 was a highly potent ligand at h_2A-ARs, displaying ~10- and 100-fold higher affinity than dexmedetomidine (Aantaa et al., 1993) and clonidine (Bucaffusco, 1992), respectively. Interestingly, S18616 displays high affinity for h_2A-ARs despite the fact that, distinct from dexmedetomidine and clonidine, it lacks a "spacer" between the imidazoline ring and the aromatic moiety (Fig. 1). As discussed elsewhere (Cordi et al., 1995), structure-activity relationships show that the carbon 1 of the tetrahydronaphthalene residue fulfills this function. In a common (CHO) expression system, like dexmedetomidine and clonidine (Jasper et al., 1998), S18616 showed similar affinity for h_2A-, h_2B-, and h_2C-ARs. Activation of _2A-ARs, which couple principally (although not exclusively) via pertussis-sensitive G_i proteins to diverse intracellular transduction pathways (Bylund et al., 1994; Aantaa et al., 1995; Wade et al., 1999; Milligan, 2000), can be quantified by [³⁵S]GTPS binding (Jasper et al., 1998). The atipamezole-reversible stimulation by S18616 of [³⁵S]GTPS binding at h_2A-, h_2B-, and h_2C-ARs demonstrates, thus, agonist properties. Furthermore, these [³⁵S]GTPS studies underpin the superior potency of S18616 versus dexmedetomidine and clonidine. In manifesting full and partial agonist activity at h