Abstract
The α2A-adrenergic receptor (AR) subtype mediates antinociception induced by the α2AR agonists clonidine, dexmedetomidine, norepinephrine, and 5-bromo-N-(4,5-dihydro-1H-imidazol-2-yl)-6-quinoxalinamine (UK-14,304) as well as antinociceptive synergy of UK-14,304 with opioid agonists [d-Ala2,N-Me-Phe4,Gly5-ol]-enkephalin and deltorphin II. Differential localization of α2-adrenergic (α2A-, α2B-, α2C-) and opioid (μ-, δ-, κ-) subtypes suggests differential involvement of subtype pairs in opioid-adrenergic analgesic synergy. The present study applies a novel imidazoline1/α2-adrenergic receptor analgesic, moxonidine, to test for involvement of α2B- and α2CARs in antinociception and antinociceptive synergy, because spinal antinociceptive activity of moxonidine shows minimal dependence on α2AAR. Intrathecal administration of moxonidine produced similar (2–3-fold) decreases in both mutant mice with a functional knockout of α2AAR (D79N-α2AAR) and α2CAR knockout (KO) mice. The potency of moxonidine was not altered in α2BKO mice, indicating that this subtype does not participate in moxonidine-induced spinal antinociception. Moxonidine-mediated antinociception was dose dependently inhibited by the selective α2-receptor antagonist SK&F 86466 in both D79N-α2A mice and α2CKO mice, indicating that α2AR activation is required in the absence of either α2A- or α2CAR. Spinal administration of antisense oligodeoxynucleotides directed against the α2CAR decreased both α2CAR immunoreactivity and the antinociceptive potency of moxonidine. Isobolographic analysis demonstrates that moxonidine-deltorphin antinociceptive synergy is present in the D79N-α2A mice but not in the α2CAR-KO mice. These results confirm that the α2CAR subtype contributes to spinal antinociception and synergy with opioids.
Several central nervous system physiological processes, including cardiovascular regulation, sedation, and analgesia are mediated by the α2-adrenergic receptor (α2AR) family of G protein-coupled receptors. The α2ARs are divided into three distinct, but highly homologous, subtypes, α2A, α2B, and α2C, which share common signal transduction pathways (Bylund et al., 1994). Discreet central nervous system localization of each subtype (Rosin et al., 1996; 1998; Talley et al., 1996; Stone et al., 1998; Rosin, 2000;Shi et al., 2000) implies that different subtypes may mediate different processes. Identification of separate physiological roles for different α2AR subtypes could improve design of novel compounds for specific therapeutic goals. Resolution of the functions specific to each α2AR subtype has been difficult due to lack of sufficiently selective pharmacological tools. However, genetic manipulation has yielded mouse lines with a dysfunctional α2AAR (MacMillan et al., 1996) or deleted α2BAR or α2CAR (Link et al., 1996), which has permitted improved evaluation of the specific physiological roles of each α2AR subtype (see Discussion). Interestingly, clarification of the physiological roles of α2CAR has reportedly been difficult (MacDonald et al., 1997). Initially, only minimal differences in α2AR agonist-induced responses (Link et al., 1996) were observed between α2CAR knockout (KO) and wild-type (WT) mice. Comparing the subtle differences among α2CAR-KO, α2CAR over-expresser (OE), and their respective WT control mice provided converging evidence that α2CARs contribute to cardiovascular function (MacDonald et al., 1997) and several physiological processes (see Discussion, Bjorklund et al., 1998; Sallinen et al., 1998a,b). A role for α2CAR in analgesia has been previously suggested (Takano and Yaksh, 1993; Guo et al., 1999; Fairbanks and Wilcox, 1999; Graham et al., 2000) but not clearly established. Identification of spinal α2CAR-mediated analgesia has been elusive, perhaps because most of the commonly used α2AR agonists appear to require the α2AAR to achieve full antinociceptive potency; functional knockout of the α2AAR dramatically reduced the potency or efficacy of these agents (clonidine, norepinephrine, dexmedetomidine, UK-14,304) (Hunter et al., 1997; Lakhlani et al., 1997; Stone et al., 1997; Fairbanks and Wilcox, 1999). Unlike these agonists, spinally administered moxonidine produces α2AR-mediated antinociception that appears largely α2AAR-independent (Fairbanks and Wilcox, 1999). The present study capitalized on the apparent α2AAR independence of moxonidine in spinal antinociception to test for an analgesic role for the α2CAR. The studies presented here apply gene substitution (α2AAR; MacMillan et al., 1996), gene knockout (α2BAR, α2CAR; Link et al., 1996), and gene knockdown (α2CAR) strategies to demonstrate a subtle but clear role for the α2CAR in spinal antinociception and α2AR-opioid antinociceptive synergy in the mouse spinal cord.
Materials and Methods
Animals.
Experimental subjects included several strains and lines of mice. First, male ICR mice (20–25 g; Harlan, Madison, WI) were used for the antisense experiments represented in Fig. 2. Second, three separate strains of mice, each mutated for one of the α2ARs, were used for experiments represented in Figs. 1, 2, 3, and 5. In each experiment, equal numbers of male and female mice (15–20 g) were used for comparison with the corresponding WT line with the same genetic background. The genetic backgrounds of each mutated mouse line were as follows. For α2AAR, mice with a gene-targeted mutation (D79N) that renders the α2AAR dysfunctional (MacMillan et al., 1996;Stone et al., 1997) were used for the experiments represented in Figs.1A and 4. Both the mutated mouse line (designated D79N-α2A) and the corresponding wild-type line (designated α2AWT) were generated on a combined 129Sv/J × C57BL/6 genetic background. For α2BAR, mice with a gene-targeted mutation that knocks out the α2BAR (Link et al., 1996) were used for the experiments represented in Fig. 1B. Both the mutated mouse line (α2BKO) and the corresponding WT line (WT-α2B) were generated on a combined 129Sv/J × C57BL/6J genetic background. For α2CAR, mice with a gene-targeted mutation that knocks out the α2CAR (Link et al., 1996) were used for the experiments represented in Figs. 1C, 3, and 5. Both the mutated mouse line (α2CKO) and the corresponding WT line (WT-α2C) were generated on a combined 129Sv/J × FVB/N genetic background. Subjects were housed in groups of 5 to 10 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. Each animal was used only once. These experiments were approved by the University of Minnesota Institutional Animal Care and Use Committee.
Chemicals.
Moxonidine [4-chloro-5-(2imidazolin-2-ylamino)-6-methoxy-2-methylpyrimidine] chloride was a generous gift of Solvay Pharma GmbH (Hannover, Germany). Substance P was purchased from Sigma Chemical (St. Louis, MO). SmithKline Beecham (King of Prussia, PA) generously donated [6-chloro-2,3,4,5-tetrahydro-3-methyl-1-H-3-benzazepine] (SK&F 86466). Efaroxan [2-(2-ethyl-2,3-dihydrobenzofuranyl)-2-imidazoline] hydrochloride and deltorphin II were purchased from Sigma/RBI (Natick, MA). Moxonidine was dissolved in 1% acetic acid and diluted with acidified saline (pH 3.2–4.0, 0.01 N acetic acid). 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).
Substance P Nociceptive Test.
Nociceptive responsiveness was tested in the substance P nociceptive test, a sensitive indicator of milder analgesics (Hylden and Wilcox, 1982). 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 postinjection. The dose of SP required to produce this number of behaviors was confirmed with each new experiment. Coadministration of opioid or α2-adrenergic analgesics dose dependently inhibits those behaviors (Hylden and Wilcox, 1981). To test the ability of moxonidine and deltorphin II to inhibit SP-induced behavior, the drugs were coadministered with SP and inhibition was expressed as a percentage of the mean response of the control group (determined with each new experiment) according to the following equation:
Antisense Oligonucleotide Treatment.
Midland Certified Reagent (Midland, TX) generated the unmodified 18-base antisense oligodeoxynucleotide (ODN) directed against the 5′ end of the coding sequence of α2CAR. Bases 1, 5, 11, and 16 were shuffled to create a mismatch control ODN sequence. The sequences were 5′-CCA-TTC-GCC-CGC-GTC-GCT-CC-3′ (antisense) and 5′-GCA-TGC-GCC-CTC-GTC-CCT-CC-3′ (mismatch). The ODNs were injected intrathecally (12.5 μg/5-μl injection) by direct lumbar puncture twice a day for 3 days before testing according to the method of Lai et al. (1996). On day 4, the animals received one more injection in the morning several hours before testing or perfusion. Each study included antisense ODN, mismatch ODN, and vehicle control groups. Several animals from each treatment group were chosen at random, anesthetized (75 mg/kg ketamine, 5 mg/kg xylazine, and 1 mg/kg acepromazine mixture i.m.), and perfused transcardially with 4% paraformaldehyde and 0.2% picric acid in 0.1 M phosphate-buffered saline, pH 6.9, by vascular perfusion as previously described (Wessendorf and Elde, 1985). Spinal cord tissue was processed for immunohistochemistry to confirm that knock down of receptor expression had occurred in a manner similar to that previously described (Lai et al., 1996). Full dose-response curves for moxonidine-induced inhibition of SP-elicited behavior were constructed for each treatment group.
Immunohistochemistry.
Spinal cords were removed and rinsed overnight with 10% sucrose in phosphate-buffered saline. Spinal segments were frozen and thaw-mounted cryostat sections (14 μm) prepared for indirect immunofluorescence histochemistry. Cryostat sections were preincubated for 1 h at room temperature in diluent containing 1% normal donkey serum, 0.3% Triton X-100, 0.01% sodium azide, and 1% bovine serum albumin. Sections were then incubated overnight at 4°C in a humid chamber with primary antisera and rinsed several times with phosphate-buffered saline. Sections were then incubated with secondary antisera for 1 h at room temperature, rinsed, and coverslipped in glycerol and p-phenylenediamine in phosphate-buffered saline with sodium bicarbonate. The primary antisera used were rabbit-derived anti-α2AAR and guinea pig-derived anti-α2CAR was used at a dilution of 1:1000 (Stone et al., 1998). The sequences against which both the α2AAR and α2CAR antisera were directed are the same in mouse as in rat. Preparations were visualized with cyanine 3.18-conjugated secondary antisera 1:200 (Jackson Immunoresearch Laboratories, Inc., West Grove, PA) and examined with a Bio-Rad MRC-1000 confocal imaging system (Bio-Rad Microscience Division, Cambridge, MA). Micrographs used in plates were assembled using Photoshop 5.0 (Adobe Systems, Mountain View, CA).
Statistical Analysis.
The ED50 values and 95% confidence limits of drugs in nanomoles were calculated using the graded dose-response curve method of Tallarida and Murray (1987). A minimum of three doses was used for each drug or drug combination. To determine differences in agonist or antagonist potency between treatment groups, nonoverlapping 95% CL were considered to represent statistically significant differences. When evaluating the extent of a potency shift between treatment groups, a potency ratio representing the ratio of the respective ED50 values was calculated.
To test for drug interactions, the 95% confidence intervals of all dose-response curves were arithmetically arranged around the corresponding ED50 values by using the equation (ln(10) × ED50) × (S.E. of log ED50). Isobolographic analysis (the appropriate method for evaluating synergistic interactions; Tallarida and Murray, 1987; Tallarida, 1992) necessitates this manipulation. When testing an interaction between two drugs given in combination for synergy, additivity, or subadditivity, a theoretical additive ED50 value is calculated for the combination based on the dose-response curves of each drug administered separately. This theoretical value is then compared by a t test (p < 0.05) with the observed experimental ED50 value for the combination of deltorphin II and moxonidine. These values are based on total dose of both drugs; in other words, the total dose of moxonidine plus the total dose of deltorphin. For the purpose of comparison with the drug doses administered separately, we have separated the moxonidine and deltorphin II components of the observed and theoretical ED50 values (Tables1 and 2). An interaction is considered synergistic if the observed combined ED50 value is significantly (p < 0.05) less than the calculated theoretical additive combined ED50 value (Tallarida and Murray, 1987;Tallarida, 1992). Additivity is indicated when the combined theoretical and experimental ED50 values do not differ.
The timing of agonist delivery can influence the nature of the apparent interaction between drugs; therefore, both agonists should be delivered in such a way that both drugs are at or near their time of peak effect during the assessment of effects on the dependent measure. Time-response experiments (data not shown) indicate that both moxonidine and deltorphin II exert their peak effect on SP-induced behavior within the first 2 min after their coinjection with SP. The same doses of either moxonidine or deltorphin II provide the same percentage of inhibition whether given as a coinjection with SP or as a 5-min pretreatment, and 10-min pretreatment yields reduced inhibition. Therefore, coinjection of the agonists with SP provides the most efficient method of delivery with minimum variability by limiting exposure of the animal to one injection.
Results
Moxonidine Produces Antinociception in D79N-α2AMutant and α2CKO Mice.
To determine the relative importance of α2AAR vis a vis α2CAR activation in moxonidine-mediated spinal antinociception, we evaluated the ability of moxonidine to inhibit SP-evoked behavior in mice mutated for α2A-, α2B-, and α2CAR and their respective wild-type counterparts. Moxonidine produced dose-dependent inhibition of SP-evoked behavior and was significantly (3.8-fold; CL, 2.1–6.8) less potent in D79N-α2A mice (Fig. 1A) than in their respective wild-type counterparts (WT-α2A). We have previously demonstrated that moxonidine-induced antinociception produced in the D79N-α2A mice is reversed by a selective α2AR antagonist (SK&F 86466) and therefore requires α2AR activation (Fairbanks and Wilcox, 1999). These results (retention of efficacy with small rightward shift) implicate the participation of either α2B- or α2CAR in moxonidine-mediated antinociception. To test for the involvement of those receptors, we evaluated antinociception produced by moxonidine in both α2BAR and α2CAR knockout mice in comparison with the respective wild-type counterparts. Moxonidine inhibited SP-evoked behavior with equal potency in α2B-WT and α2BKO mice (Fig. 1B), suggesting that α2B-adrenergic receptors do not participate in moxonidine-mediated antinociception. The potency of moxonidine was decreased in α2CKO mice (Fig. 1C) as evidenced by a moderate (2-fold; CL, 1.5–3.1) but significant parallel rightward shift in dose-response curve compared with that of their wild-type counterparts (WT-α2C). This result was confirmed in the repeat experiment represented in Fig. 5A where moxonidine showed significantly decreased potency (2.8-fold; CL, 1.4–5.4) in α2CKO versus α2C-WT mice. These results show that the α2CAR contributes to, but is not absolutely required for, moxonidine-mediated antinociception. These data indicate that activation of both the α2AAR and α2CAR (but not α2BAR) contributes to the expression of moxonidine's full antinociceptive potency.
α2CAR Participates in Moxonidine-Mediated Antinociception.
To determine whether compensatory changes accompanying α2CAR knockout accounted for KO/WT differences, we evaluated the analgesic potency of moxonidine in ICR mice treated with antisense ODN directed against the α2CAR (Fig. 2A). Moxonidine inhibited SP-evoked behavior with significantly lower potency in α2CAR antisense-treated mice relative to control mice treated with vehicle (5.8-fold difference; CL, 3.8–11) or mismatch ODN (5.1-fold difference; CL, 3.0–8.5). To confirm the integrity of the knockdown, immunohistochemistry was performed on treated tissue by using subtype-selective antisera directed against the α2CAR and the α2AAR. α2CAR immunoreactivity was observed as previously reported (Stone et al., 1999) in the superficial dorsal horn of spinal cords extracted from vehicle-treated animals (Fig. 2B). In contrast, tissue from mice given α2CAR antisense showed a substantial decrease in α2CAR-immunoreactivity (Fig. 2C); such a decrease in α2CAR-immunoreactivity was not observed in tissue from mismatch-treated (Fig. 2D) controls. Furthermore, the antisense-mediated knockdown appears to be specific because no change was observed in α2AAR-ir (Stone et al., 1998) after α2CAR antisense (Fig. 2E) or mismatch treatment (Fig. 2F). These results confirm the involvement of the α2CAR in antinociception.
Moxonidine-Mediated Antinociception Is α2AR-Dependent in α2CKO Mice.
Moxonidine's high affinity for the imidazoline (I1) receptor raises the possibility that the I1 receptor mediates moxonidine-induced antinociception in mice with disrupted α2AAR or deleted α2CAR. To address this question in a previous study (Fairbanks and Wilcox, 1999), we compared the abilities of the α2AR-selective antagonist SK&F 86466 (Hieble et al., 1986) and the mixed I1/α2AR antagonist efaroxan (Haxhiu et al., 1994) to antagonize the effects of moxonidine in D79N-α2A mice (Fairbanks and Wilcox, 1999). We observed that moxonidine-mediated antinociception in D79N-α2A mice was dose dependently reversed by both antagonists and concluded that moxonidine produced an α2AAR-independent but α2AR-dependent antinociception. In the present study, we used these same antagonists to test for α2AR dependence of moxonidine-induced antinociception in α2CKO mice and their wild-type counterparts. In α2CWT mice, a dose of moxonidine (1 nmol) was used that provided a 76 ± 3.3% antinociceptive response (n = 8 mice); SK&F 86466 dose dependently antagonized moxonidine's antinociceptive effect with a comparable but 2-fold higher potency than that of efaroxan (Fig.3A). The comparable potency of these antagonists to inhibit moxonidine-mediated antinociception confirms the requirement for α2AR activation in this mouse line. In the α2CKO mice, the magnitude of the antinociceptive response (67 ± 5.4%) to moxonidine (1.5 nmol) did not differ from that of WT-α2C mice (Student's t test, p > 0.05). Similar to the results observed in the WT-α2C mice, SK&F 86466 and efaroxan dose dependently antagonized moxonidine with comparable ED50 values (Fig. 3B); these results confirm the requirement of α2AR activation for moxonidine-mediated antinoception in α2CAR KO mice. Unambiguous demonstration of a role of the imidazoline receptor in this process is not possible in the absence of more selective antagonists for that receptor.
Moxonidine and Deltorphin II Produce Antinociceptive Synergy in α2AAR-Mutant Mice.
When agonists to both α2AR and opioid receptors are coadministered with SP, they act synergistically to inhibit SP-elicited behavior (Roerig et al., 1992). Although α2AAR mediates opioid synergy with UK-14,304, the α2CAR involvement of moxonidine-induced antinociception raised the possibility that α2CAR contributes to α2AR-opioid receptor synergy. Intrathecally administered moxonidine and deltorphin II both dose dependently inhibited SP-evoked behavior in α2AAR WT mice (Fig. 4A). The moxonidine-deltorphin II equi-effective dose ratio used (1:6) was based on their ED50 values. Combination of moxonidine and deltorphin II at this dose ratio resulted in significant leftward shifts in the dose-response curves (i.e., increased potency) compared with those of each agonist administered separately (Fig. 4A), with ED50 values significantly less than the calculated theoretical additive values (Fig. 4B; Table 1). This result indicates a synergistic interaction. Intrathecally administered moxonidine and deltorphin II both inhibited SP-evoked behavior (Fig.4C) in D79N-α2A mice. The moxonidine-deltorphin equi-effective dose ratio used was 1:1.5. Combination of moxonidine and deltorphin at this dose ratio resulted in increased potency compared with that of each agonist administered separately (Fig. 4C; Table 1). The coadministration of moxonidine-deltorphin II combinations in mice resulted in antinociceptive dose-response curves with ED50 values significantly less than the calculated theoretical additive values (Fig. 4D; Table 1), confirming a synergistic interaction in mice with dysfunctional α2AAR. The dependence of moxonidine-mediated antinociception on α2AR activation (Fairbanks and Wilcox, 1999; Fig. 3) together with the observation of moxonidine-deltorphin II synergism in D79N-α2Amice suggests an important role for α2CAR in α2AR-opioid receptor antinociceptive synergy.
Deletion of α2CAR Impairs Analgesic Synergism between Moxonidine and Deltorphin II α2CWT.
Intrathecally administered moxonidine and deltorphin II both dose dependently inhibited SP-evoked behavior in α2CAR WT mice (Fig. 5A). The moxonidine-deltorphin II equi-effective dose ratio used was 24:1. Combination of moxonidine and deltorphin II at this dose ratio resulted in increased potency compared with that of each agonist administered separately (Fig. 5A). The coadministration of moxonidine-deltorphin II in mice resulted in antinociceptive dose-response curves with ED50values significantly less than the calculated theoretical additive values (Fig. 5B; Table 2). This result indicates a synergistic interaction in α2CAR WT mice.
α2CKO.
Intrathecally administered moxonidine and deltorphin II both inhibited substance P-evoked behavior (Fig. 5C) in α2CAR KO mice. The moxonidine-deltorphin equi-effective dose ratio used was 46:1. Although the combination of moxonidine and deltorphin II shifted each dose-response curve significantly, the ED50 values did not differ significantly from the theoretical additive ED50values (Fig. 5D; Table 2). This result indicates that the interaction between moxonidine and deltorphin II was additive in mice with deleted α2CAR, which contrasts with the synergistic interaction shown in the corresponding WT mice. This result suggests that α2CAR activation is required for moxonidine-deltorphin II synergy; in contrast, although moxonidine may produce antinociception through α2AAR receptors, it appears that α2AAR activation is insufficient for moxonidine-deltorphin II antinociceptive synergy. These observations confirm a role for α2CAR in α2AR-opioid antinociceptive synergy.
Discussion
We have previously demonstrated substantial α2AAR dependence for the antinociceptive action of a panel of α2AR agonists [Fairbanks and Wilcox, 1999 (52.5°C warm water tail immersion and substance P tests); Stone et al., 1997 (substance P test)]. Functional knockout of α2A-adrenergic receptors in the D79N-α2A mouse line decreased potency or efficacy of these agonists with the following rank order from most to least affected: clonidine > dexmedetomidine > norepinephrine > UK-14,304. This mutation also blocked synergy of the least α2A-dependent agonist, UK-14,304, with the opioid receptor agonist deltorphin II and DAMGO. These two findings suggested that spinal α2AR-opioid receptor synergy, as well as adrenergic antinociception itself, relies on intact α2A-receptor function (Stone et al., 1997). However, our recent characterization of the spinal analgesic action produced by a novel imidazoline1/α2AR receptor agonist, moxonidine, suggested that another α2AR subtype must participate in spinal α2AR antinociception. Unlike the other α2AR-selective agonists, moxonidine-induced antinociception demonstrated minimal decrease (2–3-fold) in antinociceptive potency in the D79N-α2A mice (Fairbanks and Wilcox, 1999). However, moxonidine-mediated analgesia in the D79N-α2A mice was fully reversed by the α2AR-selective antagonist SK&F 86466, confirming an antinociceptive role of another α2-receptor subtype. Therefore, we tested for changes in moxonidine potency in α2BAR and α2CAR KO mice. In the present study, moxonidine potency decreased moderately but significantly in α2CAR KO mice compared with their wild-type counterparts; no such change was seen with the α2BAR KO. To further probe this apparent contribution of α2CAR to moxonidine-induced antinociception, we applied an antisense strategy (Lai et al., 1996) to knock down α2CAR expression. Multiple intrathecal injections of antisense ODN reduced the α2CAR immunoreactivity and significantly decreased the antinociceptive potency of moxonidine relative to saline-treated and mismatch-treated controls. These complementary knockout and knockdown observations unequivocally demonstrate that the α2CAR must have an analgesic function. The participation of α2AAR and α2CAR in moxonidine-induced antinociception in mice with dysfunctional (D79N-α2A) or deleted (α2CKO) appears to be equivalent: small decreases in the potency of moxonidine are observed in both mouse lines, suggesting that both receptors participate in the functional outcome of moxonidine treatment. However, the studies of moxonidine-deltorphin synergy in these two lines indicate an important distinction. Unlike the UK-14,304-deltorphin II combination (Stone et al., 1997), moxonidine-deltorphin antinociceptive synergy is present in the D79N-α2A mice but not in the α2CAR KO mice (Fig. 5). These results extend our previous studies on the role of α2AAR in spinal analgesia by demonstrating that the α2CAR also contributes to α2AR opioid synergy induced by certain agonists and identifies the potential of the α2CAR-δ-opioid receptor pair as a participant in spinal analgesia.
α2C-Adrenergic Receptor in Antinociception.
Identification of a role for the α2CAR in analgesia is consistent with its localization and with physiological responses to adrenergic agonists. Adrenergic agonists inhibit release of peptides from spinal cord slices (Ono et al., 1991) and inhibit dorsal horn nociceptive neurons (Fleetwood-Walker et al., 1985). These actions suggest both presynaptic localization on primary afferent terminals in dorsal horn and postsynaptic localization on spinal neurons. In agreement with this physiological deduction, the α2AAR has been shown to be primarily localized on SP-containing primary afferent neurons (presynaptic sites), whereas the α2CAR appears to reside primarily in spinal dorsal horn neurons (postsynaptic sites; Stone et al., 1998). In situ hybridization studies have detected α2AAR mRNA in both dorsal root ganglion (DRG) and spinal cord neurons (Gold et al., 1997; Nicholas et al., 1993; Shi et al., 1999, 2000). Immunohistochemical studies by several independent groups have clearly demonstrated α2AAR-ir in the superficial dorsal horn (Rosin et al., 1993) and in DRG neurons (Gold et al., 1997; Birder and Perl, 1999). Stone et al. (1998) extended these results to show that spinal α2AAR-ir is primarily localized to the terminals of substance P-expressing, capsaicin-sensitive primary afferent terminals. There is, therefore, a strong case for both transcription and translation of the α2AAR gene in primary afferent neurons.
In situ hybridization studies have also detected α2CAR mRNA in a large number of DRG neurons and a subset of spinal cord neurons (Nicholas et al., 1993; Shi et al., 1999; Shi et al., 2000). In agreement with those studies, Stone et al. (1998) observed the expression of α2CAR protein by both primary afferent terminals and sources intrinsic to the spinal cord. However, whereas the mRNA studies would predict a significant contribution from primary afferent fibers, it was observed that the primary, albeit not exclusive, source of α2CAR-ir in the superficial dorsal horn is spinal neurons. This conclusion was drawn by two observations. First, in rats subjected to dorsal rhizotomy α2CAR-immunoreactivity was reduced only partially relative to much greater reductions for SP-ir and α2AAR-ir. This result suggests a smaller α2CAR expression in primary afferent neurons relative to intrinsic spinal neurons. Second, α2CAR-immunoreactivity was not reduced in adult rats that had been subjected to capsaicin treatment as neonates. This result indicates that (unlike the α2AAR) the α2C-adrenergic receptor is not expressed in capsaicin-sensitive C fiber primary afferent neurons. Given the difficulties associated with extrapolating relative levels of protein expression from relative levels of mRNA, it is not surprising that results between mRNA and receptor immunoreactivity studies might be qualitatively discordant.
Other Functions of α2CAR Receptors.
Physiological studies using the mouse lines with dysfunctional α2AAR (MacMillan et al., 1996) or deleted α2A-, α2B-, or α2CAR (Link et al., 1996; Altman et al., 1999) have provided strong evidence for discrete physiological functions for the respective α2AR subtypes and have been recently comprehensively reviewed (Kable et al., 2000). Collectively, studies originally indicated that the α2AAR primarily mediated centrally mediated hypotension (MacMillan et al., 1996), anesthesia (Lakhlani et al., 1997), analgesia (Hunter et al., 1997; Lakhlani et al., 1997; Stone et al., 1997), sedation (Hunter et al., 1997; Lakhlani et al., 1997; Sallinen et al., 1997), antiepileptogenesis (Janumpalli et al., 1998), and α2AR agonist-mediated inhibition of monoamine release and metabolism in brain (MacDonald et al., 1997). The α2BAR appears to be required for the initial peripheral hypertensive responses to α2AR agonists (Link et al., 1996), salt-induced hypertension (Makaritsis et al., 1999) and possibly development or reproduction (Makaritsis et al., 1999). Clarification of the physiological role of α2CAR has reportedly been difficult (MacDonald et al., 1997). Despite widespread central nervous system distribution, it was notable that the α2CAR did not prove critical for the cardiovascular effects mediated by a reportedly nonselective α2AR agonist, dexmedetomidine (Link et al., 1996). Further evaluation of small differences between α2CKO and α2CWT mice suggested α2CAR participation in dexmedetomidine-induced hypothermia, dopamine metabolism, and d-amphetamine-induced hyperlocomotion (Rohrer and Kobilka, 1998). Confirmation of these subtle physiological differences was greatly aided through comparison of α2AR agonist-mediated effects in α2CAR KO and α2CAR OE mice (Bjorklund et al., 1998; Sallinen et al., 1998a,b). These studies of subtle differences between α2CAR KO and α2CAR OE mice also revealed participation by the α2CAR in cardiovascular function (MacDonald et al., 1997), the startle reflex and aggression (Sallinen et al., 1998a,b), and complex navigation behavior (Bjorklund et al., 1999). The present study extends those findings to illuminate (through rigorous examination of moderate, but significant effects in α2CKO and α2CWT mice as well as α2CAR antisense-treated mice) a role for the α2CAR in antinociception.
Significance.
The present study directly demonstrates a requirement for α2CAR to mediate an antinociceptive action of an exogenously administered imidazoline1/α2AR agonist (moxonidine). These observations provide strong evidence that the α2CAR subtype can contribute to α2AR agonist-mediated analgesia and synergy with opioids in the mouse spinal cord. Based on the potential involvement of α2CAR in the action of endogenously released norepinephrine (Guo et al., 1999) we speculate that synergy between α2CAR and opioid receptors may mediate, at least in part, analgesia induced by systemic morphine. An interaction between morphine action at spinal sites and norepinephrine released spinally as a consequence of supraspinal morphine-mediated activation of descending noradrenergic pathways has been proposed (Wigdor and Wilcox, 1987). The present results support the assertion (Guo et al., 1999) that targeting the α2CAR for analgesic therapy may represent an improvement over α2AAR-selective agonists, because the latter receptor subtype likely mediates adrenergic agonist-induced sedation (Mizobe et al., 1996; Lakhlani et al., 1997). The validity of this target is further supported by clinical observations that antihypertensive doses of moxonidine produce significantly fewer side effects (sedation, dry mouth, rebound withdrawal) than clonidine, a strongly α2AAR-dependent agent (Fairbanks and Wilcox, 1999). The present study provides strong support for development of moxonidine or other α2CAR-selective agonists to be used either separately or in combination with opioid analgesics for the treatment of pain.
Acknowledgments
We extend our appreciation to Drs. Dieter Ziegler and Joerg Meil (Solvay Pharma GmbH) and Dr. Paul Hieble (SmithKline Beecham) for the gifts of moxonidine and SK&F 86466, respectively; to Drs. Lee Limbird and Leigh MacMillan for the donation of the D79N-α2A mutant mice and their wild-type counterparts for the start of the breeding colony; to Dr. John Hunter for the donation of the α2C- and α2B-mutant mice and their wild-type counterparts for the start of the breeding colony; and to Dr. Michael H. Ossipov for assistance with statistical analysis.
Footnotes
-
This study was supported by National Institutes of Health Grants R01-DA-01933 and R01-DA-11236 to G.L.W. National Institute on Drug Abuse training Grant T32A07234 supported C.A.F.
Abbreviations
- AR
- adrenergic receptor
- KO
- knockout
- OE
- over-expresser
- WT
- wild-type
- SP
- substance P
- ODN
- oligodeoxynucleotide
- CL
- confidence limits
- I1 imidazoline1
- -ir, immunoreactivity
- DRG
- dorsal root ganglion
- UK-14,304
- 5-bromo-N-(4,5-dihydro-1H-imidazol-2-yl)-6-quinoxalinamine
- Received June 19, 2001.
- Accepted September 12, 2001.
- The American Society for Pharmacology and Experimental Therapeutics