Materials and Methods

Animals.

Experimental subjects were 20- to 25-g male ICR mice (Harlan, Madison, WI) or 15- to 20-g male and female mice (gender-matched) with a mixed C57BL/6 × 129/Sv genetic background (designated WT for wild type). We also used male and female mice with this same background (C57BL/6 × 129/Sv) with a “hit and run” gene-targeted mutation (D79N) that renders the α_2aAR dysfunctional (MacMillan et al., 1996;Stone et al., 1997). These experiments were approved by the Institutional Animal Care and Use Committee. Subjects were housed in groups of 5 to 10 in 25 × 48 × 15-cm plastic cages in a temperature- and humidity-controlled environment. Subjects were maintained on a 12-h light/dark cycle and had free access to food and water. In the experiment using the noradrenergic neurotoxin 6-OHDA, the mice that were tested 3 days after treatment (saline or 6-OHDA) were randomized and retested 9 days later on day 14 after treatment. In all other experiments, each animal was used only once.

Chemicals.

Moxonidine [4-chloro-5-(2-imidazolin-2-ylamino)-6-methoxy-2-methylpyrimidine] chloride was a generous gift of Solvay Pharma GmbH (Hannover, Germany). Norepinephrine (NE) was purchased from Sigma Chemical (St. Louis, MO) and was prepared fresh for each experiment in acidified saline. Smith Kline & French (King of Prussia, PA) donated the SK&F 86466 [6-chloro-2,3,4,5-tetrahydro-3-methyl-1-H-3-benzazepine]. Zeneca (Wilmington, DE) donated the dexmedetomidine [(+)-(S)-4-[1-(2,3-dimethylphenyl) ethyl]1H-imidazole]. Efaroxan [2-(2-ethyl-2,3-dihydrobenzofuranyl)-2-imidazoline] hydrochloride was purchased from Research Biochemicals International (Natick, MA). SP was purchased from Sigma Chemical Co. (St. Louis, MO). Clonidine HCl (2-[2,6-dichloroaniline]-2-imidazoline) was from Boehringer-Ingelheim Ltd. Moxonidine was dissolved in 1% acetic acid and diluted with acidified saline (pH 3.2–4.0, 0.01 N acetic acid). NE and SP were dissolved in acidified saline. All other drugs were dissolved in 0.9% saline. All drugs were administered intrathecally in a 5-μl volume in conscious mice according to the method of Hylden and Wilcox (1980) as modified by Wigdor and Wilcox (1987).

Thermal Nociception.

Thermal nociceptive responsiveness was determined using the warm water (52.5°C) immersion tail-flick test. The latency to the first rapid tail-flick represented the behavioral endpoint (Janssen et al., 1963). Baseline measurements of tail-flick latencies were collected on all mice before testing. The mean baseline tail-flick latency of the ICR mice was 3.6 s (S.D. = 0.69 s,n = 31); ICR mice that failed to respond within 5 s to baseline tests were excluded from analysis. The mean baseline tail-flick latency of the WT and D79N-α_2amutant mice was 6.4 s (S.D. = 2.6 s, n = 250). The baseline tail-flick latency did not differ (unpairedt test: p > .05) between the WT and D79N-α_2a mice (WT: mean = 6.5 s, S.D. = 2.5 s, S.E.M. = 0.27 s, n = 133 mice; D79N-α_2a mice: mean = 6.2 s, S.D. = 2.7 s, S.E.M. = 0.24 s, n = 85 mice). WT and D79N-α_2a mice that failed to respond within 11.5 s (a value of more than 2 S.D. from the mean) were eliminated from analysis (5.5%) (Horan et al., 1991). The percent of maximum possible antinociceptive effect (% MPE) of injected (drug or saline control) was determined according to the following formula:%MPE=[(Postdrug Latency−Predrug latency)/ Cutoff−Predrug latency)]×100 In all three lines of mice, a maximum score of 100% was assigned to those animals not responding before a 12-s (ICR) or 15-s (WT/D79N-α_2a) cutoff to avoid tissue injury. In each case in which the animal did not respond before the cutoff, the tail was examined for loss of motor control. Only animals capable of movement in their tails were included in analysis.

SP Nociceptive Test.

Nociceptive responsiveness was also tested in the SP nociceptive test. The SP assay is a sensitive indicator of milder analgesics (Hylden and Wilcox, 1981). A constant dose of SP (10–20 ng) was injected intrathecally to produce approximately 40 to 60 behaviors (scratches and bites directed to the hindlimbs) in the first minute after injection. The dose of SP required to produce this number of behaviors was confirmed with each new experiment. Coadministration of opioid or adrenergic analgesics dose-dependently inhibits those behaviors (Hylden and Wilcox, 1983). To test the ability of moxonidine, dexmedetomidine, clonidine, and NE to inhibit SP-induced behavior, the drugs were coadministered with SP and inhibition was expressed as a percent of the mean response of the control group (determined with each new experiment) according to the following equation:%Inhibition=[(control−experimental)/control]×100 In some experiments, antagonists were coadministered with the moxonidine-SP combinations. In the case of the SK&F 86466 compound and efaroxan, dose-antagonism curves were determined for their respective abilities to antagonize the inhibition of SP behavior by an 80% effective dose of moxonidine. The ED₈₀ value was calculated from the regression line of the dose-inhibition curve. This measure was expressed by the equation:%Antagonism=[(80−%inhibition)/80]×100 Each dose selected from the calculated ED₈₀ value was tested experimentally on the day of the percent antagonism experiment. In the event that the experimental dose did not produce an exact 80% inhibition, the equation was adjusted to reflect exact percent inhibition that the antagonist was acting against (e.g., 70 or 90 for 70% or 90%, respectively).

Antagonism of Moxonidine

Dose-response curves to moxonidine and dexmedetomidine were generated in ICR mice in the SP nociceptive test. ED₈₀ doses were calculated according to the method of Tallarida and Murray (1987). ICR, WT, and D79N-α_2a mutant mice were coadministered with either SK&F 86466 (0.01–10 nmol i.t.) or efaroxan (0.001–10 nmol i.t.) and an ED₈₀ dose of either moxonidine (ICR: efaroxan, 0.2 nmol; SK&F 86466, 0.3 nmol; WT: 1.5 nmol; D79N: 8.6 nmol i.t.) or dexmedetomidine (ICR: efaroxan, 0.2 nmol; SK&F 86466, 0.8 nmol) in the SP nociceptive test. Inhibition curves for SK&F 86466 or efaroxan were generated and the ID₅₀ values were calculated according to the method of Tallarida and Murray (1987).

6-OHDA Treatment.

ICR mice were injected intrathecally with the noradrenergic neurotoxin 6-OHDA (5 μg i.t.) (Fig.6). This dose of 6-OHDA was selected because it has been demonstrated to be effective when administered intrathecally in mice; Fasmer and colleagues (1986) demonstrated that uptake of ³H-NE was dramatically reduced in mouse spinal cord 14 days after the i.t. injection of 5 μg of 6-OHDA. At 3 and 14 days after treatment, saline-treated mice and 6-OHDA-treated mice were injected with different doses of moxonidine or NE. Dose-inhibition curves were collected and ID₅₀values calculated for each drug in each treatment group. The 14-day time point represents the same mice used for the 3-day time point. Mice from the first group were randomized for the subsequent treatment (NE or moxonidine) to reduce potential confound from previous drug exposure. The time points of 3 and 14 days treatment were selected to be comparable to the testing days used by Fasmer et al. (1986); this group observed differences in the responses to nociceptive stimuli of 6-OHDA-treated and control mice on day 3 but not on day 14 after treatment.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6

Pretreatment with the noradrenergic neurotoxin 6-OHDA increases norepinephrine potency but not moxonidine potency in ICR mice. ICR mice were injected either with saline or 5 μg of 6-OHDA. Three days later, dose-response curves to NE and to moxonidine were collected in both groups of mice using the SP test. A, 3 days pretreatment: dose-response curves to NE were generated in saline-treated [ED₅₀ = 32 pmol (23–44) ▵, dashed line] and 6-OHDA-treated [ED₅₀ = 12 pmol (10–13), left ▴, solid line] mice; dose-response curves to moxonidine in saline-treated [ED₅₀ = 130 pmol (100–170), ○, dashed line] and 6-OHDA-treated [ED₅₀ = 140 pmol (100–180), ●, solid line] mice. B, 14 days pretreatment: dose-response curves to NE were generated in saline-treated [ED₅₀ = 72 pmol (40–130), ▵, dashed line] and 6-OHDA-treated [ED₅₀ = 32 pmol (25–41), leftward ▴, solid line] mice; dose-response curves to moxonidine in saline-treated [ED₅₀ = 327 pmol (197–549), ○, dashed line] and 6-OHDA-treated [ED₅₀ = 282 pmol (145–551), ●, solid line] mice.

Statistical Analysis.

Data describing antinociception are expressed as % MPE or percent inhibition with S.E.M. In experiments in which full dose-response curves were generated, a minimum of three doses was used for each drug or combination of drugs. Statistical comparisons of potencies are based on the confidence limits of the ED₅₀ values. A shift in a dose-response curve is considered significant when the calculated ED₅₀value of one curve falls outside the confidence limits of the ED₅₀ value of the curve to which it is being compared. The ED₅₀ values and confidence limits were calculated according to the method of Tallarida and Murray (1987).

Results

Moxonidine-Induced Antinociception

Moxonidine Produces Antinociception in ICR Mice in Tail-Flick Test and SP Nociceptive Test.

We determined dose-response curves (Fig.1) for the effects of moxonidine in the tail-flick test [0.1, 0.3, and 1 nmol, i.t; ED₅₀= 0.5 nmol (0.3–0.7)] and in the SP nociceptive behavioral test in [(0.01, 0.03, 0.1, and 0.3 nmol, i.t; ED₅₀ = 0.04 nmol (0.03–0.07)]. The lower ED₅₀ value of moxonidine in the SP test versus the tail-flick test is consistent with comparisons made between the tests for other antinociceptive compounds. We have observed that both opioid and adrenergic agonists produce antinociception in the SP nociceptive test with greater potency compared with their respective potencies in the tail-flick assay (Hylden and Wilcox, 1983). Furthermore, the ED₅₀ values of moxonidine in both tests are comparable to those observed with spinally administered morphine or clonidine (see Discussion).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Intrathecally administered moxonidine produces spinal antinociception. A, tail-flick test: dose-response curves to moxonidine were generated in ICR mice in the warm water immersion tail-flick assay [ED₅₀ = 0.5 nmol (0.3–0.7) ○]. B, SP test: dose-response curves to moxonidine were generated in ICR mice in the SP nociceptive behavioral assay [ED₅₀ = 0.04 nmol (0.03–0.07) ●].

Antagonism of Moxonidine-Induced Antinociception

SKF-86466 and Efaroxan Dose-Dependently Antagonize Moxonidine-Induced Antinociception in SP Nociceptive Test.

We compared the abilities of the α₂AR-selective antagonist SK&F 86466 (Hieble et al., 1986) and the I₁/α₂AR-selective mixed antagonist efaroxan (Haxhiu et al., 1994) to antagonize the effects of moxonidine and dexmedetomidine in the SP test in ICR mice.

SK&F 86466 and Efaroxan Antagonism of Dexmedetomidine in ICR Mice.

As a positive control, we antagonized dexmedetomidine antinociception with both SK&F 86466 and efaroxan. On each experimental day, we generated a dose-response curve to dexmedetomidine in ICR mice and the calculated the ED₈₀ value from the regression line. For the SK&F 86466 antagonism experiment, the ED₅₀ value of dexmedetomidine was calculated to be 0.0013 nmol and the ED₈₀ dose was calculated to be 0.8 nmol. This constant dose of dexmedetomidine was administered with varying doses of SK&F 86466 (0.1, 0.3, 1, 3, 6, and 8 nmol i.t.) to generate a dose-inhibition curve (Fig.2A). A high dose of SK&F 86466 (8 nmol i.t.) alone did not affect the number of SP-elicited behaviors (data not shown). Similarly, we observed that a single dose of the nonimidazoline α₂AR-selective antagonist yohimbine (2.5 nmol) effectively antagonized high efficacy doses of moxonidine (3 nmol, data not shown) in the tail-flick test. For the efaroxan antagonism experiment, the ED₅₀ value of dexmedetomidine was calculated to be 0.003 nmol (0.001–0.1) and the ED₈₀ dose was calculated to be 0.2 nmol. This constant dose of dexmedetomidine was administered with varying doses of efaroxan (0.01, 0.1, 0.3, and 1 nmol i.t.) to generate a dose-inhibition curve (Fig. 2A). A high dose of efaroxan (1 nmol i.t.) did not affect the number of SP-elicited behaviors (data not shown). The ID₅₀ value of SK&F 86466 antagonism of dexmedetomidine was 1.5 nmol (0.87–2.6), which is 15-fold greater than the ID₅₀ value of efaroxan (0.1 nmol, 0.028–0.38) antagonism of dexmedetomidine.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

SK&F 86466 and efaroxan antagonize dexmedetomidine- and moxonidine-induced inhibition of SP-elicited behavior in ICR mice. A, dose-response curves to dexmedetomidine [SKF experiment: ED₅₀ = 0.0013 nmol (0.0002–0.01); EFA experiment: ED₅₀ = 0.003 nmol (0.001–0.01); ED₈₀= 0.2 nmol] were generated to determine the appropriate ED₈₀ doses (data not shown). Dose-antagonism curves (shown) for the ability of SK&F 86466 [ID₅₀ = 1.5 nmol (0.87–2.6), ▴] and efaroxan [ID₅₀ = 0.1 nmol (0.028–0.38), ○] to inhibit the effect of the ED₈₀ dose of dexmedetomidine were generated. B, dose-response curves to moxonidine [SKF experiment: ED₅₀ = 0.04 nmol (0.03–0.07); ED₈₀ = 0.3 nmol; EFA experiment: ED₅₀ = 0.06 nmol (0.04–0.08); ED₈₀ = 0.2 nmol)] were generated to determine the appropriate ED₈₀ doses (data not shown). Dose-antagonism curves (shown) were generated for the ability of SK&F 86466 [ID₅₀ = 0.15 nmol (0.079–0.29), ▴; efaroxan: ID₅₀ = 0.05 nmol (0.014–0.18 ○] to inhibit the effect of the ED₈₀doses of moxonidine.

SK&F 86466 and Efaroxan Antagonism of Moxonidine in ICR Mice in SP Test.

On each experimental day, we generated a dose-response curve to moxonidine in ICR mice and calculated the ED₈₀value from the regression line. For the SK&F 86466 antagonism experiment, the ED₅₀ value of moxonidine was 0.04 nmol (0.03–0.07) and the ED₈₀ dose was calculated to be 0.3 nmol. This constant dose of moxonidine was administered with varying doses of SK&F 86466 (0.01, 0.1, and 1 nmol i.t.) to generate a dose-antagonism curve (Fig. 2B). Similarly, we observed that a single dose of the yohimbine (0.3 nmol) effectively antagonized high-efficacy doses of moxonidine (0.3 nmol, data not shown) in the SP test. For the efaroxan antagonism experiment, the ED₅₀ value of moxonidine was 0.06 nmol (0.04–0.08) and the ED₈₀ dose was calculated to be 0.2 nmol. This constant dose of moxonidine was administered with varying doses of efaroxan (0.001, 0.01, 0.1, and 1 nmol i.t.) to generate a dose-inhibition curve (Fig. 2B). The ID₅₀ value of efaroxan antagonism of moxonidine was 0.05 nmol (0.014–0.18), which did not differ significantly from the ID₅₀ value of SK&F 86466 (0.15 nmol, 0.079–0.29) antagonism of moxonidine.

Adrenergic Receptor Subtype Participation in Moxonidine-Induced Antinociception

Moxonidine Produces Antinociception in WT Mice and D79N-α_2a Mutant Mice in Tail-Flick and SP Nociceptive Tests.

The ability of SK&F 86466 to antagonize moxonidine in ICR mice strongly suggested the participation of an α₂AR in moxonidine-induced antinociception. Other studies suggest that the α_2aAR subtype is the primary mediator of α₂AR agonist-mediated antinociception (Stone et al., 1997). To examine the participation of the α_2aAR subtype in moxonidine-mediated antinociception, we tested moxonidine for its ability to produce antinociception in mice with a mutation (D79N) in the α_2aAR. This mutation results in a functional disruption of the α_2aAR and an 80% knock-down in receptor binding in membrane preparations from brain (MacMillan et al., 1996). To begin, we tested for moxonidine-induced antinociception in the D79N-α_2aAR mutant mouse and WT counterpart line and compared potency to that of clonidine. Moxonidine dose-dependently (0.01, 0.03, 0.06, 0.1, and 3 nmol i.t.) produced potent antinociception in WT mice in the tail-flick test [Fig.3A, ED₅₀ = 0.17 nmol (0.09–0.32)]. Clonidine also dose-dependently [1, 10, 60, and 100 nmol i.t. ED₅₀ = 28 nmol (11–72)] produced potent antinociception in WT mice in the tail-flick test (Fig. 3B). We also tested both moxonidine and clonidine for the inhibition of SP-induced behavior. Moxonidine [0.1, 0.3, and 1 nmol i.t. ED₅₀ = 0.4 (0.3–0.5)] (Fig.4A), clonidine [0.1, 1, and 10 nmol i.t. ED₅₀ = 2.7 (1.3–5.4)] (Fig. 4B), and NE (0.001, 0.03, 0.1, 0.3, 1 pmol i.t. ED₅₀ = 0.087 pmol (0.016–0.48), Fig. 4C) all dose-dependently inhibited the SP-induced nociceptive behavior. We then evaluated moxonidine antinociception in D79N-α_2aAR mice in the tail-flick test [0.01, 0.1, 1, and 3 nmol, ED₅₀ = 0.32 (0.074–1.6)] and in the SP nociceptive test [0.1, 1, and 10 nmol, ED₅₀ = 1.1 (0.7–1.7)]. In both tests, moxonidine potency was approximately 2-fold lower in the α_2a mutant mice than in their WT counterparts. In contrast, clonidine did not produce antinociception in the D79N-α_2aAR mice in either the tail-flick (Fig.3B) or the SP test (Fig. 4B) at doses up to 100 (SP test) or 300 (tail-flick test) nmol. In contrast to both moxonidine and clonidine, the nonselective catecholamine α₂AR agonist NE produced antinociception in the D79N-α_2aAR mice [0.3, 100, and 300 pmol i.t. ED₅₀ = 68 pmol (8.1–564), Fig. 4C] but with 780-fold decreased potency relative to their WT counterparts.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Moxonidine produces spinal antinociception in the tail-flick test in WT and D79N-α_2a mutant mice. A, tail-flick test: moxonidine. Dose-response curves to moxonidine were generated in WT [ED₅₀: 0.17 nmol (0.09–0.32), ▪] and D79N-α_2a [ED₅₀: 0.32 nmol (0.074–1.6), ●] mice. B, tail-flick test: clonidine. Varying doses of clonidine were administered intrathecally to WT [ED₅₀: 28 nmol (11–72), ▪] and D79N-α_2a (ED₅₀ not calculable, ●) mice.

• Download figure
• Open in new tab
• Download powerpoint

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Moxonidine produces spinal antinociception in the SP test in WT and D79N-α_2a mutant mice. A, SP test: moxonidine. Dose-response curves to moxonidine were generated in WT [ED₅₀ = 0.4 (0.3–0.5), ▪] and D79N-α_2a [ED₅₀ = 1.1 (0.7–1.7), ●] mice. B, SP test: clonidine. Varying doses of clonidine were administered intrathecally to WT [ED₅₀ = 2.7 (1.3–5.4), ▪] and D79N-α_2a (ED₅₀ not calculable, ●) mice. C, SP test: norepinephrine. Varying doses of NE were administered intrathecally to WT [ED₅₀ = 0.087 pmol (0.016–0.48), ▪] and D79N-α_2a[ED₅₀ = 68 pmol (8.1–564)] mice.

SKF-86466 and Efaroxan Dose-Dependently Antagonize Moxonidine-Induced Antinociception in SP Nociceptive Test.

The ability of moxonidine to produce antinociception in both WT and D79N-α_2a mice suggested the participation of some α₂AR, although perhaps not α_2aAR. However, moxonidine is also thought to act at the imidazoline receptor, which could explain its effectiveness in the D79N-α_2a mutant mice. To address this issue, we tested the ability of the α₂AR-selective antagonist SK&F 86466 as well as the mixed I₁/α₂AR-selective antagonist efaroxan to antagonize the effects of moxonidine and dexmedetomidine WT and D79N-α_2a mice in the SP test.

WT Mice.

ED₈₀ doses of moxonidine were estimated from previous dose-response curves and confirmed by administering that dose (1.46 nmol) against SP. This constant dose of moxonidine was administered with varying doses of SK&F 86466 (0.1, 0.2, 0.5, and 1 nmol i.t.) to generate a dose-inhibition curve [ID₅₀ = 0.4 nmol (0.3–0.5)] (Fig.5A). Similarly, we observed that a single dose of the yohimbine (0.3 nmol) effectively antagonized high-efficacy doses of moxonidine in WT mice (0.1 nmol, data not shown). Efaroxan (0.1, 0.3, and 1 nmol i.t.) antagonized this same ED₈₀ dose of moxonidine with an ID₅₀ value of 0.24 nmol (0.18–0.33) (Fig. 5B); this value does not differ significantly from the ID₅₀ value of SK&F 86466.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

SK&F 86466 and efaroxan antagonize moxonidine-induced inhibition of SP-elicited behavior in WT and D79N-α_2amutant mice. A, WT mice: dose-antagonism curves were generated to SK&F 86466 [ID₅₀ = 0.4 nmol (0.3–0.5), ●] and efaroxan [ID₅₀ = 0.24 (0.18–0.33), ▪] against a constant ED₈₀ dose of moxonidine (1.46 nmol) B, D79N-α_2a mutant mice: dose-antagonism curves were generated to SK&F 86466 [ID₅₀ = 1 nmol (0.68–1.5 nmol i.t.) ●] and efaroxan [ID₅₀ = 4.6 nmol (3.1–6.8), ▪] against a constant ED₈₀ dose of moxonidine (8.56 nmol for SK&F 86466 and 10 nmol for efaroxan) (two different experimental days). Efaroxan (10 nmol i.t.) did not produce a behavioral effect when administered alone to D79N-α_2amice, nor did it affect SP-elicited behavior (data not shown).

D79N-α_2a Mutant Mice.

ED₈₀doses of moxonidine were estimated from previous dose-response curves (Fig. 2B) and confirmed by administering that dose against SP. This constant dose of moxonidine (8.6 nmol i.t.) was administered with varying doses of SK&F 86466 (0.3, 1, and 3 nmol i.t.). SK&F 86466 dose-dependently antagonized moxonidine with an ID₅₀ value of 1 nmol (0.68–1.5) (Fig. 5A). Similarly, we observed that a single dose of yohimbine (2.5 nmol) effectively antagonized high-efficacy doses of moxonidine (6 nmol) in D79N-α_2a mutant mice (data not shown). Efaroxan dose-dependently (1, 3, and 10 nmol i.t.) antagonized an ED₈₀ dose of moxonidine (10 nmol i.t.) with an ID₅₀ value 4.6 nmol (3.1–6.8); the ID₅₀ value of efaroxan is significantly (4.5-fold) greater than the ID₅₀ value of SK&F 86466.

Exclusion of Autoreceptor-Containing Noradrenergic Terminals.

Dose-response curves to moxonidine- and NE-induced inhibition of SP-elicited behavior were collected 3 and 14 days after i.t. treatment with either saline or 6-OHDA (Fasmer et al., 1986). At both time points, the potency of NE was significantly increased in mice treated with 6-OHDA [ED₅₀ = 3 days, 12 pmol (10–13); 14 days, 32 pmol (25–41)] compared with saline-treated controls [ED₅₀ = 3 days, 32 pmol (23–44); 14 days, 72 pmol (40–130)]. This may indicate a supersensitization of postsynaptic α₂ARs after the destruction of presynaptic noradrenergic terminals by the 6-OHDA toxin (Swerdlow and Koob, 1989). In contrast, moxonidine potency did not differ between 6-OHDA [ED₅₀ = 3 days, 140 pmol (100–180); 14 days, 282 pmol (145–551)] compared with saline-treated controls [ED₅₀ = 3 days, 130 pmol (100–170); 14 days, 327 pmol (197–549)] at either time point.

Discussion

Moxonidine produced potent antinociception in the tail-flick test and the SP nociceptive test in two different strains of mice: ICR and C57BL/6 × 129/Sv (WT). Two α₂AR antagonists [efaroxan (I₁/α₂) and SK&F 86466 (α₂)] dose-dependently antagonized moxonidine-induced antinociception, confirming the requirement for α₂AR. Moxonidine potency was decreased 2-fold in the D79N-α_2a mutant mice compared with their WT counterparts. This stands in sharp contrast to the complete lack of a spinal antinociceptive effect of clonidine in the D79N-α_2a mutant mice and substantially decreased potency for NE. Taken together, these data suggest that moxonidine-induced antinociception is governed primarily by an α₂AR subtype, but not likely the α_2aAR subtype.

Moxonidine-Induced Antinociception.

The ability of α₂-adrenergic agonists to prolong tail-flick latencies is well established (Reddy et al., 1980; Yasuoka and Yaksh, 1983; Milne et al., 1985; Solomon et al., 1989; Ossipov et al., 1990). Spinal administration of moxonidine dose-dependently increased tail-flick latency in two strains of mice: ICR (Fig. 1) and C57BL/6 × 129/Sv (Fig. 3A). Moxonidine and morphine have similar tail-flick ED₅₀ values in both ICR mice (Fairbanks and Wilcox, 1997) and C57BL/6 × 129/Sv (WT) mice (Stone et al., 1997): in ICR mice, moxonidine (0.5 nmol, 0.3–0.7), morphine (1.2 nmol, 0.7–2.4); and in WT mice, moxonidine (0.17 nmol, 0.09–0.32), morphine (0.52 nmol, 0.36–0.74). Morphine remains the gold standard with which other analgesics are compared. The observation that moxonidine prolongs tail-flick latencies with potency comparable to that of morphine illustrates the relevance of moxonidine-induced antinociception. Like morphine, clonidine represents the gold standard for α₂AR-mediated analgesia. Interestingly, we observed that the potency of moxonidine in the tail-flick test is 100-fold greater than that of clonidine [Fairbanks and Wilcox, 1999; ED₅₀ = 49 nmol (29–82) in ICR mice Fig. 1AB] and 164-fold greater than that of clonidine in WT mice (Fig. 3AB). That moxonidine is more potent than clonidine further attests to the relevance of moxonidine-induced antinociception.

These potency relationships observed in the tail-flick test generalize to a chemical test of nociception, the SP test. This test sensitively and reliably detects antinociception induced by both opioid and adrenergic agonists: morphine, [d-Ala²,N-MePhe⁴,Gly-ol⁵]-enkephalin, deltorphin II, UK-14,304, and dexmedetomidine (Hylden and Wilcox, 1983;Stone et al., 1997); all inhibit SP-induced nociceptive behavior with greater potency than they prolong tail-flick latencies (52.5°C). Like these other antinociceptive compounds, intrathecally administered moxonidine dose-dependently inhibited SP-elicited behavior in both strains of mice [ICR (Fig. 1) and C57BL/6 × 129/Sv (WT) (Fig.4A)] with comparable potency. In the SP test in ICR mice, the ED₅₀ value of moxonidine (0.04 nmol, 0.03–0.07) is lower than that of morphine (0.5 nmol, 0.4–0.6) but similar to that of clonidine (0.015 nmol, 0.004–0.05) (Hylden and Wilcox, 1981). In the SP test in WT mice, the ED₅₀ value of moxonidine (0.23 nmol, 0.18–0.31) is also comparable to that of morphine (0.13 nmol, 0.05–0.3) (Stone et al., 1997) and lower than that of clonidine (2.7 nmol, 1.3–5.4). Therefore, moxonidine is either more potent than or as potent as the prototypic opioid and adrenergic analgesics, morphine and clonidine, in both tests and both mouse strains tested here. However, predictive extrapolation from effective intrathecal doses in the rodent models reported here to analgesically effective human doses awaits further testing in other larger species. Further studies are also needed to evaluate potential hemodynamic changes evoked by spinal administration of moxonidine because systemically administered moxonidine has been reported to decrease blood pressure in WT mice (Eglen et al., 1998).

It is noteworthy that baseline nociceptive responsiveness differed to some degree in these two strains: the mean tail-flick latency in the ICR mice (3.6 ± 0.69 s, n = 31) was shorter than that of the C57BL/6 × 129/Sv strain (6.4 ± 2.6 s,n = 250). On the other hand, the C57BL/6 × 129/Sv strain appeared to be more sensitive to SP, requiring a smaller dose of SP (10 ng) than ICR mice (15 ng) to elicit 40 to 60 behaviors. Recent comparisons of nociceptive sensitivities across multiple inbred mouse lines in multiple pain tests reveal that a particular strain may show low sensitivity in one test (e.g., C7BL/6, carrageenan-induced thermal hypersensitivity) and high sensitivity in another (e.g., C57BL/6, formalin test, late phase) (Mogil, 1999). Although we did not systematically compare the nociceptive sensitivities between ICR and WT lines, we did observe that moxonidine was more potent in WT than in the ICR mice in the tail-flick test, whereas the opposite was true in the SP test. In the absence of further experiments, the relevance of these potency differences remains unclear; however, it is noteworthy that moxonidine antinociception is not specific to a particular mouse strain.

Receptor Subtype Participation in Moxonidine-Induced Antinociception.

Elucidation of the specific receptors that mediate antinociceptive drug action remains an area of intense investigation. Pharmacological antagonism has proved to be effective in determining the opioid receptor subtypes (μ, δ, or κ) activated by various opioid agonists. Unfortunately, α₂AR antagonists are insufficiently selective between the subtypes α_2a, α_2b, and α_2c to make similar distinctions. Generation of antisense oligonucleotides and mutant mouse lines with absent or disrupted α_2aAR, α_2bAR, and α_2cAR permitted studies to determine the selective function of these subtypes. These studies have concluded that the α_2aAR mediates opioid-adrenergic synergy (Stone et al., 1997), epileptogenesis (Janumpalli et al., 1998), and sedation (Mizobe et al., 1996; Hunter et al., 1997; Lakhlani et al., 1997). Moxonidine is an α₂AR agonist; however, both moxonidine and clonidine are also thought to act at a nonadrenergic receptor, the imidazoline receptor, to produce an antihypertensive action (Ziegler et al., 1996). Moxonidine binds preferentially to the putative imidazoline binding site over the α₂ARs: moxonidine was 33-fold selective for I₁ receptors versus α₂ARs (Ernsberger et al., 1993). In comparison, clonidine is much less (3.8-fold) selective for I₁ receptors over α₂ARs. Considering the selectivity of moxonidine for I₁ receptors together with immunohistochemical localization of imidazoline receptors in superficial dorsal horn of the spinal cord (Ruggiero et al., 1998), it is conceivable that imidazoline receptors participate in moxonidine-mediated antinociception in concert with α₂ARs.

To examine what receptor subtypes participate in moxonidine-induced antinociception, we challenged moxonidine with the nonimidazoline, α₂AR-selective antagonist SK&F 86466 (Hieble et al., 1986) and the mixed I₁/α₂AR antagonist efaroxan in both ICR (Fig. 2) and WT mice (Fig. 5). SK&F 86466 has demonstrated 2700-fold selectivity for α₂receptors over I₁ receptors (Ernsberger et al., 1990). The mixed I₁/α₂AR antagonist efaroxan is 40-fold selective for I₁ receptors over α₂ARs (Haxhiu, 1994). This experimental strategy has been used by other investigators to distinguish the α₂AR-mediated and I₁receptor-mediated antihypertensive mechanisms of clonidine and moxonidine (Hieble et al., 1986; Gomez et al., 1991; Haxhiu et al., 1994, 1995, Mest et al., 1995). We expected that an observed difference in the ID₅₀ value of either antagonist to inhibit moxonidine would suggest that moxonidine acts preferentially at one receptor over the other; however, there was no appreciable difference between the abilities of SK&F 86466 and efaroxan to antagonize an ED₈₀ dose of moxonidine in either ICR mice (Fig.2 B) or WT mice (Fig. 5A). The antagonism by SK&F 86466 confirmed that α₂AR activation is required for moxonidine-induced antinociception. The observation that efaroxan antagonized moxonidine with comparable potency to SK&F 86466 suggests that the two compounds act at the same site, specifically an α₂AR receptor. These data are consistent with those of another study by Monroe et al. (1995), who tested for imidazoline receptor mediation of clonidine-induced spinal antinociception. Similar to the efaroxan/SK&F 86466 strategy, this group (Monroe et al., 1995) tested idazoxan (imidazoline) and yohimbine (nonimidazoline) for antagonism of clonidine-mediated antinociception. They showed that these two antagonists completely and equipotently blocked clonidine-induced antinociception. From these results, they concluded that the imidazoline receptor does not participate in the antinociception induced by spinally administered clonidine.

The α_2aAR receptor subtype has been suggested to be the primary mediator of α₂AR-mediated spinal analgesia (Stone et al., 1997). We tested moxonidine for antinociceptive efficacy in a line of mice expressing a point mutation (D79N) in the α_2a receptor, which uncouples the receptor from K⁺ and Ca²⁺channels and reduces α_2a receptor binding (MacMillan et al., 1996). Taken together, these observations support the assertion that this mouse line can be considered a functional knockout of the α_2a receptor. The antinociceptive potency of three α₂AR agonists is substantially diminished in these animals (Stone et al., 1997) (Figs. 3 and 4). Dexmedetomidine inhibits SP-elicited behavior with a 2500-fold decreased potency in the D79N-α_2amice compared with their WT counterparts (Stone et al., 1997). Similarly, the antinociceptive potency of UK-14,304 is decreased 250-fold (Stone et al., 1997) and NE antinociceptive potency is decreased 780-fold in the D79N-α_2a mice (Fig.4C). Clonidine efficacy is absent (Fig. 3B, 4B) in these animals in both the SP (Fig. 4B) and tail-flick tests (Fig. 3B). In contrast, moxonidine is only 2-fold less potent in D79N-α_2a animals compared with their WT controls in both the SP (Fig. 4A) and tail-flick (Fig. 3A) tests. The relevance of this small shift is questionable in light of the large decreases in potency observed in D79N-α_2a mice with the other agonists; therefore, we assert that this small decrease in potency observed in the D79N-α_2a mice implicates the participation of receptors other than α_2aAR subtype in the mediation of moxonidine-induced antinociception. To indirectly test for participation by the α_2cAR or α_2bAR subtype and the I₁receptor, we combined the strategies of pharmacological antagonism and genetically mutated mice by antagonizing moxonidine antinociception in the D79N-α_2a mice with SK&F 86466 and efaroxan. SK&F 86466 effectively antagonized moxonidine-induced inhibition of SP-induced behavior in ICR, WT, and D79N-α_2a mice (Figs. 2 and 5, A and B). The observation that SK&F 86466 antagonized the antinociceptive action in the D79N-α_2a mice strongly implicates a role for some α₂AR in this effect, although probably not the α_2aAR subtype. In the D79N-α_2a mice, SK&F 86466 antagonized moxonidine-induced antinociception with significantly greater (4.5-fold) potency than did efaroxan. Furthermore, efaroxan antagonized moxonidine antinociception with 19-fold greater potency in the WT mice compared with the D79N-α_2a mutant mice. This result may indicate a functional selectivity of efaroxan for the α_2aAR. Given that efaroxan is a mixed I₁/α₂AR antagonist, the participation of imidazoline receptors in moxonidine-induced analgesia cannot be ruled out, but the antinociceptive effect of moxonidine clearly requires α₂AR. The principal finding from the present experiments is that the ability of SK&F 84666 to antagonize moxonidine-induced antinociception in the D79N-α_2a mutant mice solidly implicates participation of α₂ adrenergic subtypes, either α_2c or α_2b. Further studies are required to resolve the contribution of those α₂AR subtypes to moxonidine-induced antinociception.

Lack of Participation of Autoreceptors on Descending Noradrenergic Terminals in Moxonidine-Induced Antinociception.

It is conceivable that SK&F 86466-sensitive moxonidine-induced antinociception could require the participation of autoreceptors on descending noradrenergic terminals. In other words, moxonidine-induced antinociception could be mediated by action at autoreceptors on descending noradrenergic terminals, resulting in the release of NE to act on postsynaptic α₂ARs. In support of the participation of autoreceptors and subsequent NE release, Klimscha et al. (1997) have demonstrated that intrathecally administered clonidine increased the concentration of NE in spinal cord dorsal horn microdiasylate in sheep. Furthermore, Armah and colleagues (Ziegler et al., 1996) found that the hypotensive effects of moxonidine in rabbits were substantially diminished with 6-OHDA pretreatment, implicating the participation of catecholamine stores in the activity of moxonidine. The present experiments revealed that NE potency increased in animals treated with 6-OHDA compared with saline-treated control animals (Fig.6). This apparent supersensitivity after a catecholaminergic lesion suggests increased affinity of postsynaptic α₂ARs (Post et al., 1987; Swerdlow and Koob, 1989), indicating that the 6-OHDA treatment successfully destroyed descending noradrenergic terminals. If moxonidine-induced antinociception requires postsynaptic activation of α₂ARs by release of NE from descending terminal, the potency of moxonidine should increase in parallel with that of the exogenously applied NE. However, moxonidine potency did not differ between 6-OHDA-treated and saline-treated animals at either 3 or 14 days after treatment (Fig. 6). We interpret this result to indicate that that moxonidine-induced antinociception does not require the release of NE. This is in contrast to supersensitization observed with clonidine-mediated antinociception after 6-OHDA lesioning (Post et al., 1987). These results combined with the observation that (in these murine models of nociception) both clonidine and NE demonstrate selectivity for the α_2aAR suggests that the α_2aAR receptor is the subtype that becomes supersensitive after 6-OHDA administration. This selectivity could explain why moxonidine-induced antinociception does not display supersensitization after 6-OHDA lesioning. Further studies would be needed to confirm this proposal.

Conclusion.

To our knowledge, this study is the first to report an antinociceptive application for the first of a new generation of antihypertensive compounds, moxonidine. Based on the scientific experience with other adrenergic agonists such as clonidine and dexmedetomidine, we anticipate that the antinociceptive effect of moxonidine will generalize to other species. Unlike antinociception induced by clonidine, dexmedetomidine, UK-14,304, and NE, moxonidine-induced antinociception requires α₂AR activation independent of the α_2aAR subtype. Others have suggested that differential receptor selectivities (imidazoline receptor versus α₂AR) account for the improved side effect liabilities of moxonidine versus clonidine when given systemically for the treatment of hypertension. We speculate that our observed differences in α₂AR subtype requirements for the effects of moxonidine and clonidine could also result in differential side effect profiles in spinal administration of moxonidine for the treatment of pain.