Introduction

Summary Introduction Materials & Methods Results Discussion References

₂-ARs mediate many physiological functions and pharmacological effects in the central nervous system, mainly by inhibiting neuronal firing and release of NE and other neurotransmitters. ₂-ARs are also involved in a wide range of functions in peripheral tissues (e.g., in the regulation of NE release from sympathetic nerves, smooth muscle contraction, platelet aggregation, insulin secretion, glomerular filtration, and energy metabolism) (1). Activation of ₂-ARs with the highly specific ₂-AR agonist dexmedetomidine results in bradycardia, hypotension, hypothermia, locomotor inhibition, anxiolysis, analgesia, sedation, and, with higher doses, anesthesia. Dexmedetomidine also reduces the turnover of the monoamine neurotransmitters NE, DA, and 5-HT (serotonin) in brain (2).

Recent pharmacological and biochemical research has led to a subdivision of ₂-ARs into three distinct subtypes: _2A-, _2B-, and _2C-ARs. This classification was first based on the pharmacological properties of the receptors and was confirmed through the cloning of three distinct ₂-AR genes in humans, rats, mice, and other species (3). Each receptor has a distinct tissue distribution. In the central nervous system of the rat, _2A-ARs are widely expressed, whereas the other ₂-AR subtypes have more limited distributions. _2C-ARs are present in the basal ganglia, olfactory tubercle, hippocampus, and cerebral cortex, but _2B-ARs are expressed only in thalamic nuclei (4, 5). The _2A-AR seems to have an important role as a presynaptic and somatodendritic autoreceptor in noradrenergic nerve cells because this receptor subtype has been identified through both in situ hybridization (4, 5) and immunohistochemistry (6, 7) in the cells of the locus ceruleus and other noradrenergic centers.

Pharmacological experiments have not revealed the functions of the individual receptor subtypes because suitable subtype-selective agonists or antagonists have not been available. An alternative approach to this problem is to use transgenic techniques to genetically alter the expression of a single receptor gene and thereby gain knowledge about the functions of a particular receptor protein (8-13). Recent results obtained with transgenic mice with dysfunctional _2A-ARs and mice with targeted disruption of _2B- and _2C-AR genes demonstrate the involvement of _2A-AR in central cardiovascular regulation (14) and of _2B-AR in the peripheral vasoconstriction induced by ₂-AR agonists (15). We studied the physiological and pharmacological functions of _2C-AR in mice with tissue-specific overexpression of this receptor and in mice with targeted disruption of the _2C-AR gene and lack of functional _2C-AR protein. We concentrated on locomotor functions and temperature control because the distribution of _2C-AR in rodent brain points to an involvement of these receptors in striatal functions. In addition, we studied whether the neurochemical effects of a potent, subtypenonselective ₂-AR agonist on monoamine turnover in brain would be influenced by altered _2C-AR expression.

	Materials and Methods

Summary Introduction Materials & Methods Results Discussion References

Animals

Two strains of genetically engineered mice were generated at Stanford University (Stanford, CA). One had a targeted inactivation of the _2C-AR gene (16), and the other had tissue-specific overexpression of _2C-ARs. A total of 60 mice were shipped to the Central Laboratory Animal Facility of the University of Turku, Finland, where they were maintained and bred according to standard procedures.

The generation of mice with targeted disruption of the gene encoding _2C-AR (KO mice) has been described previously (16). Briefly, the mutation was made by inactivating one copy of the Adra2c gene (the murine gene of the _2C-adrenergic receptor) in R1 129/Sv embryonic stem cells (17). Cells containing a targeted mutation of the Adra2c locus were injected into C57BL/6J blastocysts, and the resulting chimeric mice were bred to F₁ (C57BL/6J × DBA/2J) animals. These animals were back-crossed for several generations to C57BL/6J mice. The resulting offspring from heterozygous intercrosses formed the colony maintained in Turku. The genetic background of the strain is thus mainly C57BL/6J with a small contribution from 129/Sv and DBA/2J strains. Adra2c^/ mice are viable and fertile and appear grossly normal. The germline transmission of the mutation was continuously monitored from mouse tail DNA by Southern blot analysis.

Mice with overexpression of the _2C-AR gene were generated by pronuclear injection using standard techniques (18, 19). In brief, one-cell fertilized eggs were harvested from 5-week-old superovulated FVB/N female mice. The microinjected constructs contained 4.5 kb of 5 flanking sequence and 5 kb of 3 flanking sequence surrounding the Adra2c open reading frame, which encodes the murine _2C-AR (20). In addition, we attached the influenza hemagglutinin epitope recognized by the antibody 12CA5 to the extreme amino terminus of the Adra2c protein. Previous research has demonstrated that the amino-terminal epitope does not impair expression of the Adra2c receptor or alter its ligand-binding properties to a significant degree (21). A tyrosinase minigene construct was coinjected for visual identification of the transgenic progeny on the basis of different coat color. This construct contained 2.2 kb of 5, 0.7 kb of 3, and 1.3 kb of coding sequence from pBS-Tyrosinase (22). Eggs that survived microinjection, as judged by morphology, were transferred to the oviduct of pseudopregnant foster mothers. DNA was collected from tail biopsies of weanling animals from these transfers and analyzed by Southern blot analysis or PCR for the presence of the adrenergic transgene (Fig. 1). For transgene-specific PCR, the sense primer (5-AGC TTC CAT GGG CTA CCC ATA CGA CGT CCC AGA CTA CGC CAG-3) was located in the 12CA5 epitope, whereas the antisense primer (5-TTT CTC GCT GAG CGT ACG-3) was located immediately after the fifth transmembrane domain of the Adra2c protein. Southern blot analysis was performed as described previously (16) using a 0.72-kb NcoI/MluI Adra2c probe. The adrenergic transgene and dark coat color were found to cosegregate during breeding. In addition, the intensity of dark coat color was correlated with the number of copies of both the tyrosinase minigene and the adrenergic transgene as detected by Southern blot analysis. The transgenic mouse line (281#17) that had the greatest amounts of gene copies was chosen for further breeding and experimentation (23).

View larger version (35K): [in this window] [in a new window]   Fig. 1.   PCR and Southern blot characterization of transgenic mouse lines carrying the Adra2c overexpression transgene. a, Ethidium-stained agarose gels showing detection of transgenic founder animals using the PCR assay described in Materials and Methods. Arrowhead, position of the 777-bp transgene-specific fragment. Left, sizes of relevant size standards. Standards were HindIII cut  and HaeIII cut PhiX174 DNA. Open arrows, PCR-positive founder lines (numbered). The coat color positive line no. 6 (filled arrow), which carries only integrated copies of the tyrosinase minigene, was confirmed to be PCR negative in this assay. b, Quantitative Southern blot analysis comparing transgene copy number from different transgenic lines with the endogenous Adra2c gene from a wild-type (wt) mouse. Equal amounts of genomic DNA from tail biopsies were digested with EcoRI and detected as described in Materials and Methods. With this strategy, the Adra2c coding sequences derived from the wild-type locus and the transgene concatamers comigrate as a 6.3-kb fragment.

Mice were housed in groups of 7-10 at 22 ± 1°, with a 12:12-hr light/dark cycle (lights on at 6:00 a.m. and off at 6:00 p.m.). Standard certified pelleted food (Rat/Mouse 1 Maintenance Expanded SQC; Special Diet Services, Essex, UK) and water were available ad libitum. The mice were transferred to the neuropharmacological test laboratory for behavioral experiments at least 1 week before testing. All tested KO mice were 2.5-4.5 months old and homozygous for the mutation, and all OE mice were 1.5-2.5 months old and heterozygous. Wild-type control animals for each strain were age- and sex-matched littermates or closely related animals. Each animal was used only once.

Analysis of _2C mRNA Expression in Brain

In situ hybridization. Four KO and OE mice and four respective control animals (age, 3-4 months) were used for in situ hybridization. The mice were killed with carbon dioxide, and the brains were rapidly excised and cooled in ice-cold 0.9% saline. The brains from KO and OE animals and their controls were frozen together on the same specimen holder. The tissues were sectioned in a Microm HM 500 cryostat at 14 µm and thaw-mounted onto Probe-On glass slides (Fisher Scientific, Pittsburgh, PA). The sections were stored at 20° until use.

Hybridization probes. Two oligonucleotides [nucleotides 186-233 and 815-862 of the Adra2c coding sequence (20)] were labeled to specific activity of 1 × 10⁹ cpm/µg at the 3 end with [³³P]dATP (New England Nuclear Research Products, Boston, MA) using terminal deoxynucleotidyltransferase (Amersham, Buckinghamshire, UK). Both probes produced similar results. Several control probes with the same length and similar GC content and specific activity were used to determine the specificity of the hybridizations. The addition of 100-fold excess of respective unlabeled probe abolished all hybridization signals.

Analysis of _2C-AR Binding Sites in Brain

Receptor autoradiography. Three 7-week-old mice from each group were killed by decapitation, and their brains were rapidly dissected and frozen by immersion into cold isopentane in a dry ice bath. Coronal 14-µm sections were cut on a Microm cryostat and thaw-mounted onto gelatin-coated slides. The slides were first dried at room temperature for 2 hr and subsequently stored at 70° with desiccant in sealed containers.

Striatal homogenates. The 7-week-old mice were killed by decapitation, and their striata were rapidly dissected and homogenized in 10 volumes of potassium phosphate buffer. Crude membrane fractions were prepared by centrifugation. The membranes were washed once and used for binding assays using the ₂ subtype-nonselective radioligand [³H]RX821002 (0.1-4 nM) or the _2C-AR-preferring radioligand [³H]rauwolscine [0.8 nM with or without 100 nM oxymetazoline (Sigma) to block _2A-ARs] (29). ()-Epinephrine (100 µM) was used to determine specificity of binding.

Spontaneous Motor Activity Tests

Spontaneous locomotor activity was measured by placing individual animals into a polypropylene animal cage (38 × 22 × 15 cm) with a transparent polypropylene lid and a ~1.5-cm-thick layer of granulated aspen bedding on the floor. The cages were surrounded with a photobeam frame system designed for activity measurements (Photobeam Activity System PAS, Cage Rack, San Diego Instruments, San Diego, CA). The system consisted of 16 separate frames connected to a computer control unit. The system enabled the simultaneous measurement of the activity of eight mice at two levels (3 and 6 cm from the bottom of the cage). Three different parameters were recorded: 1) ambulations [large movements (breaks of two adjacent beams of light at the lower level)], 2) fine movements (two consecutive breaks of a single beam of light at the lower level), and 3) rearings (one break of a beam of light at the upper level).

The base-line locomotor activity of all mouse strains (eight 12-16-week-old male mice per group) was measured over 1-hr intervals for 22 hr after a 3-hr habituation period to detect possible differences in spontaneous locomotor activity or its diurnal rhythm.

The effects of the potent and specific ₂-AR agonist dexmedetomidine (Orion Corporation, Orion-Pharma, Turku, Finland) (2) on locomotor activity were tested in two separate experiments with adult males. In the first experiment, three different doses (5, 10, or 20 µg/kg) or control solution (distilled water) was administered subcutaneously to OE and OE-wt mice (10 mice per group; total, 80). In the second experiment, groups of KO and KO-wt mice (11-14 mice per group; total, 116) were administered dexmedetomidine (5, 10, 20, or 30 µg/kg) or vehicle. The mice were weighed on the day of the experiment, and allocation of the mice to different treatment groups was randomized according to the Latin square principle. Dexmedetomidine or vehicle was administered 20 min before the animal was placed into the measurement cage, and activity was measured over 10-min intervals for 30 min. The mice were decapitated 1 hr after drug administration, and their brains were rapidly dissected and frozen (70°) for neurochemical assays.

Body Temperature Experiments

Core body temperatures were measured with a rectal probe and a digital thermometer (Ellab, Roedovre, Denmark). The probe was inserted 2.5 cm inside the anal sphincter. The animals were familiarized to the procedure 1 day before the pharmacological experiments. Three doses (10, 20, and 30 µg/kg) of dexmedetomidine or vehicle (distilled water) was administered subcutaneously, and body temperature was measured at base-line and 30, 60, and 90 min after the injection. KO mice and controls (10 mice per group; total, 80) were female, and OE mice and controls (8-10 mice per group; total, 76) were male. Similar groups of male KO and KO-wt mice were also challenged with three different hypothermic doses of the nonselective DA agonist ()-apomorphine (0.1, 0.3, and 1.0 mg/kg; Research Biochemicals, Natick, MA) and the 5-HT_1A receptor agonist (±)-8-OH-DPAT (0.1, 0.3, and 1.0 mg/kg; Research Biochemicals).

Neurochemical Assays

Biogenic amines and their metabolites were determined from striatal and frontal cortical samples and from whole-brain homogenates in 0.1 M perchloric acid using electrochemical detection (ESA Coulochem, 5011, Bedford, MA) after separation by HPLC on a reversed-phase C18 column (Ultrasphere ODS, 4.6 × 250 mm; Beckman Instruments, Fullerton, CA). The buffer systems described by Mefford (30) were used, with separate assays for indoles and the DA metabolite HVA and for catecholamines after their purification on activated alumina. The minor modifications to the original published procedures have been described elsewhere (31). The NE metabolite MHPG was determined in a separate HPLC assay using a modified procedure previously validated for plasma samples (32). The brain homogenates were centrifuged, and the supernatants were neutralized before extraction with Bond Elut PH columns (Analytichem International, Harbor City, CA). The intra-assay coefficient of variation was 6.6% at 0.40 nmol/g, and the interassay coefficient of variation was 10.0% at 0.28 nmol/g.

Statistical Analysis

The statistical significance of differences between groups was determined by ANOVA using STATISTICA 4.5 computer software (StatSoft, Tulsa, OK). Two-way ANOVA was used for the locomotor activity results and neurochemical measurements to test the differences between genotypes and drug doses, and a comparison of different doses was performed with Tukey's post hoc t test. The body temperature measurements were evaluated using three-way ANOVA for repeated measurements using genotype, dose, and time as independent factors. Values of p < 0.05 were considered statistically significant. The results are presented as mean ± standard error.

	Results

Summary Introduction Materials & Methods Results Discussion References

Analysis of _2C mRNA expression in brain by in situ hybridization. No clear differences were observed in the intensity or distribution of the hybridization signals between KO and KO-wt animals (Fig. 2, top). Brain sections from OE mice had clearly increased _2C mRNA in regions that also in OE-wt mice were labeled for _2C mRNA (Fig. 2, bottom). In addition, very strong hybridization was seen in the ependyma of the ventricles and choroid plexus of OE mice (Fig. 2, bottom).

View larger version (113K): [in this window] [in a new window]   Fig. 2.   Top, in situ hybridization of 2C mRNA in the brains of KO-wt and KO mice. No clear difference in the expression of 2C mRNA can be seen. Bottom, in situ hybridization of 2C mRNA in the brains of OE-wt and OE mice. The hybridization signal is increased in OE mice. The signal in OE animals is in the same brain areas that in OE-wt mice exhibit 2C mRNA except that a clear signal is seen in the ependyma of the ventricles and choroid plexus. Acb, accumbens nucleus; CA1 and CA3, fields of Ammon's horn; DG, dentate gyrus; CPu, caudate-putamen; Tu, olfactory tubercle; Lv, lateral ventricle; D3v, dorsal third ventricle.

Analysis of _2C-binding sites in brain by receptor autoradiography. Fig. 3 shows the autoradiographic distribution of [³H]RX821002 and [³H]rauwolscine binding in forebrain sections of OE mice and their controls (OE-wt). Both [³H]RX821002 and [³H]rauwolscine clearly labeled a larger population of sites in many regions of OE mouse brain compared with corresponding regions in control (OE-wt) mice (Table 1). This difference was largest in the caudate-putamen and in the stratum radiatum of the CA1 region of the hippocampus (~3-fold). In addition, clearly increased _2C-AR binding was noted overlying all brain ventricles. There was no difference in B_max values between the strains in frontal cortex and in the dentate gyrus. Receptor affinity for [³H]RX821002 was equal in both strains and in all regions, ~0.4 nM. In contrast, calculated K_D values for [³H]rauwolscine were clearly smaller in OE mice than in OE-wt mice, especially in brain regions in which _2C-ARs were abundant (fundus striatum, caudate-putamen, stratum radiatum of CA1: 0.3 versus 1.3 nM). Rosenthal plots of [³H]rauwolscine binding could not be reliably resolved into two populations of binding sites because of the small number of radioligand concentrations used in the assays. The reduction in total ₂-AR density and high affinity [³H]rauwolscine binding in brains of KO mice has been demonstrated previously (16).

View larger version (109K): [in this window] [in a new window]   Fig. 3.   Images of autoradiographic binding of (top) [3H]RX821002 and (bottom) [3H]rauwolscine in coronal sections from two levels of forebrain of (left) OE and (right) OE-wt mice. Acb, accumbens nucleus; CA1 and CA3, fields of Ammon's horn; LMol, lacunosum moleculare; Rad, stratum radiatum; DG, dentate gyrus; CPu, caudate-putamen; Lv, lateral ventricle; D3v, dorsal third ventricle; Tu, olfactory tubercle; LSD, lateral septal nucleus.

View this table: [in this window] [in a new window]   TABLE 1 Density (Bmax) of 2C-AR binding sites in different regions of brains of OE and OE-wt mice determined by in vitro autoradiography with [3H]RX821002 and [3H]rauwolscine

Radioligand binding to striatal membranes. Striatal membrane preparations from four animals from each group were used to determine receptor affinities (K_D) and receptor densities (B_max) in this brain region. Receptor affinities for [³H]RX821002 were similar in membranes from OE, OE-wt, KO, and KO-wt mice: 0.23-0.37 nM. The binding capacity was clearly lower in membranes from KO mice (81 ± 8 fmol/mg of protein) than in membranes from KO-wt mice (122 ± 20 fmol/mg of protein); an intermediate value was obtained with membranes from mice heterozygous for the inactivating mutation (94 ± 7 fmol/mg). Compared with [³H]RX821002, binding of [³H]rauwolscine to striatal membranes was more markedly reduced in KO mice, especially in experiments conducted in the presence of 100 nM oxymetazoline to block binding to _2A-AR (10 ± 4 versus 35 ± 3 fmol/mg; 20 ± 5 fmol/mg in heterozygous mice). The results for OE mice and their controls are shown in Table 2 separately for male and female animals. The capacity of [³H]RX821002 binding to striatal membranes of OE mice was more than double that of membranes from OE-wt mice (162 versus 67 fmol/mg). Binding of [³H]rauwolscine to striatal membranes of OE mice indicated a ~3-fold higher density of _2C-adrenergic binding sites in striatal membranes from OE mice than in membranes from OE-wt mice (109 versus 36 fmol/mg).

View this table: [in this window] [in a new window]   TABLE 2 2-AR density in membrane preparations from striata of 6-7-week-old OE and OE-wt mice of either sex [3H]RX821002 was used to detect total 2-AR binding. The results obtained with the 2C-AR-preferring radioligand [3H]rauwolscine, in the presence of 100 nM oxymetazoline to block 2A-ARs, reflect 2C-AR density.

Spontaneous motor activity and the effects of dexmedetomidine. OE mice and their controls were more active in the 22-hr base-line activity measurements compared with KO mice and their controls. The average total counts (ambulations plus fine movements) over 22 hr were 4750 ± 136/4910 ± 116 for the OE and OE-wt mice and 3680 ± 92/3440 ± 73 for KO and KO-wt mice. No differences were noted between the mice with altered _2C-AR expression and their respective controls in total counts, ambulations or fine movements, or the diurnal pattern of locomotor activity (Fig. 4).

View larger version (46K): [in this window] [in a new window]   Fig. 4.   Base-line locomotor activity of (a) OE and OE-wt mice and (b) KO and KO-wt mice. The total activity (ambulations and fine movements) was measured at 1-hr intervals for 22 hr starting at 2:00 p.m. after a 3-hr habituation. Filled bars, OE mice. Open bars, KO mice. Hatched or cross-hatched bars, respective wt controls. Error bars, standard error for eight measurements.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Locomotor activity measured as ambulation counts (large horizontal movements) in two separate drug experiments. a, Effect of dexmedetomidine on OE and OE-wt mice. b, Effect of dexmedetomidine on KO and KO-wt mice. Filled bars, OE mice. Open bars, KO mice. Hatched or cross-hatched bars, respective wt controls. Error bars, standard error for 10-14 measurements.

Effects of dexmedetomidine on body temperature. The dose-dependent hypothermic effect of dexmedetomidine was more potent in OE animals than in their controls (three-way ANOVA, time × strain interaction: p = 0.0017). Correspondingly, the KO mice were not as sensitive to the hypothermic effect of the drug as their controls (p = 0.0029) (Fig. 6). The average differences from the controls in the reduction of body temperature calculated at 30 min after the injection were 0.80° (17.3%) in KO mice and 0.43° (11.8%) in OE mice when all doses of active drug were included in the analysis. There were no differences in the hypothermic responses of KO and KO-wt mice to the DA agonist apomorphine or the 5-HT_1A agonist (±)-8-OH-DPAT. Apomorphine (1 mg/kg) lowered the mean body temperature 30 min after the injection by 4.86 ± 0.32° and 5.57 ± 0.49° in KO and KO-wt mice, respectively (p = 0.10). After 0.3 mg/kg, the reductions were 3.41 ± 0.65° and 3.28 ± 0.53° (p = 0.99). The corresponding temperature reductions induced by 1 mg/kg (±)-8-OH-DPAT were 3.17 ± 0.30° and 2.90 ± 0.36° (p = 0.90). In addition, the two strains of control mice showed differing sensitivity to dexmedetomidine (p < 0.0001), with the OE-wt mice less hypothermic after dexmedetomidine than the KO-wt mice. This difference may have been due to sex or strain differences or both.

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Time course of effects of different doses (0, 10, 20, and 30 µg/kg) of dexmedetomidine on body temperature. a, OE and OE-wt mice (male; 8-10 mice per group). b, KO and KO-wt mice (female; 10 mice per group). Values are mean ± standard error. , , , and , wt mice. , , , and , (a) OE or (b) KO mice. , 0 µg/kg; , 10 µg/kg; , 20 µg/kg, , 30 µg/kg.

Effects of dexmedetomidine on brain monoamines and their metabolites. Table 3 shows the results of brain neurochemistry analyses from whole-brain homogenates. As expected, dexmedetomidine dose-dependently reduced the concentrations of the measured monoamine metabolites in whole brain. There were, however, no significant strain × dose interactions in the two-way ANOVA, indicating that the drug-induced reductions in the metabolism of NE, DA, and 5-HT were not influenced by altered _2C-AR expression. When all treatment groups were included in the ANOVA, statistically significant strain differences were observed as follows: OE mice had higher concentrations of HVA (p < 0.0001) and DA (p = 0.049) in brain than their controls, and KO mice had lower concentrations of HVA (p = 0.013), 5-HIAA (p = 0.041), and MHPG (p = 0.007) in brain than their controls. In brain samples from another group of naive mice, KO mice, compared with KO-wt mice, had lower concentrations of HVA in striatum (15.1%) (p = 0.021) but not in frontal cortex, and OE mice, compared with OE-wt mice, had higher concentrations of HVA (+19.2%) (p = 0.038) in frontal cortex but not in striatum (Table 4). Dexmedetomidine had no significant effects on the levels of NE, DA, and 5-HT in brains of OE and OE-wt mice, but it increased the amine concentrations in brains of KO and KO-wt mice. It also significantly reduced the brain concentrations of the 5-HT precursor amino acid tryptophan in all strains.

	Discussion

Summary Introduction Materials & Methods Results Discussion References

The present in situ hybridization, receptor autoradiographic, and in vitro receptor binding results from the brains of OE mice showed that microinjections of the described DNA sequences could be successfully used to generate a strain of mice with tissue-specific overexpression of _2C-ARs. The in situ results and receptor autoradiograms from brains of OE mice showed the expected regional distribution of increased _2C-AR expression, confirming that the expression was tissue specific. The promoter elements present in the gene construct were capable of directing the increased receptor expression to those brain regions, and probably also to those cells, which normally express endogenous _2C-ARs. The only possible exception to this contention detected so far was the presence of _2C mRNA and [³H]rauwolscine binding in ependymal tissue in OE mice. Control (OE-wt) mice did not have detectable _2C mRNA hybridization or [³H]rauwolscine binding in this tissue. Results from receptor binding assays with striatal homogenates support the conclusion that OE mice have ~3-fold tissue-specific overexpression of _2C-AR in striatum.

As reported previously (16), based on evidence from Northern blot and reverse-transcription PCR analyses, Adra2c^/ (KO) mice do not produce gene transcripts capable of encoding functional _2C-AR protein. This was supported by receptor autoradiographic results from brain sections, in which [³H]rauwolscine binding (in the presence of oxymetazoline to mask _2A-AR) was markedly reduced in the caudate-putamen, the brain region normally expressing the highest proportion of _2C-AR (16). The brain regions that were found to have significantly increased binding of [³H]rauwolscine in OE mice were the same with reduced binding in KO mice (16).¹ Also, this distribution pattern agrees with our _2C mRNA in situ hybridization results from normal mice and previous, more-detailed in situ _2C-AR mRNA hybridization results in the rat brain (4, 5). In situ hybridization analysis of brain sections from KO mice and their controls yielded similar results, which is not unexpected because the probes were targeted to mRNA sequences upstream of the inactivating mutation.

Despite significantly altered _2C-AR expression, both KO and OE animals are viable and fertile and appear grossly normal. Clearly, the _2C-AR is not fundamental for mouse embryonic development or adult reproductive function, and a moderately increased receptor expression level does not confer a serious disadvantage for survival and reproduction. Also, altered _2C-AR expression did not have major effects on spontaneous locomotor activity or its diurnal rhythm. The altered _2C-AR expression modified the hypothermic response to the potent ₂-AR subtype-nonselective agonist dexmedetomidine because OE mice were more hypothermic and KO mice were less hypothermic after the drug than were their controls. However, altered _2C-AR expression had no effect on the sedative response to dexmedetomidine.

According to the HPLC analyses, there were no overall strain-dependent differences in the sensitivities of mice to the ₂-AR agonist-induced alterations in brain monoamine levels or metabolism. However, some minor differences were found. Interestingly, the concentrations of the DA metabolite HVA seemed to be influenced by altered _2C-AR expression, with OE mice having increased and KO mice having decreased concentrations of HVA. The similar drug effects on brain monoamine metabolism in the mutant and control mice indicate that the majority of the effects of ₂-AR agonists on brain monoamine turnover are mediated by _2A-ARs, as was expected for NE.

In the neurochemical assays performed in the course of this study, the expected reductions in the turnover rates of brain monoamine neurotransmitters after dexmedetomidine were similar within both pairs of strains. However, some minor but consistent differences in the measured levels were noted in all dose groups between mutant and control mice. In KO mice, MHPG (metabolite of NE), 5-HIAA (metabolite of 5-HT), and HVA (metabolite of DA) levels were lower compared with their controls, whereas OE mice had minor elevations in their levels of HVA and DA in brain. Thus, it seems that the lack of _2C-AR expression slightly decreases the rate of monoamine turnover in brain and that overexpression of _2C-AR increases DA stores and metabolism. The opposite findings for the HVA levels in KO and OE mice suggest a possible involvement of _2C-ARs in regulation of brain DA systems. This was supported by the assays of striatal and frontal cortical samples of drug-naive animals, in which KO mice had lower HVA concentrations in striatum and OE mice had higher HVA concentrations in frontal cortex, whereas other monoamine and metabolite levels were unaltered. The difference in frontal cortex was somewhat unexpected because cortical _2C-AR expression is quite low, but it suggests that the involvement of the _2C-ARs in dopaminergic regulation is not confined to striatum.

The _2C-AR distribution in brain and the observed differences in HVA concentrations in both mutant mouse strains compared with their controls point to the possibility that _2C-ARs participate in regulation of dopaminergic function. In addition to locomotor functions, striatal dopaminergic systems have been implicated in other physiological functions, including modulation of body temperature regulation. This is a possible mechanism for the opposite alterations in the hypothermic responses to dexmedetomidine in KO and OE mice. The preoptic area of the anterior hypothalamus is a central site of thermoregulation in rodents, but pharmacological experiments have demonstrated that nigrostriatal dopaminergic neurons also participate in the complex regulation of body temperature (33). It is also possible that the modulatory effects of _2C-ARs on body temperature are due to altered heat production by peripheral muscles, which is indirectly controlled by nigrostriatal pathways. Similar thermal responses of KO and KO-wt mice to apomorphine and (±)-8-OH-DPAT support the assumption that central DA or 5-HT receptors are not compensatorily regulated in KO mice and that the strain-dependent changes in body temperature after dexmedetomidine are consequences of differing _2C-AR activation. KO mice, however, had a tendency to have an attenuated response to the highest dose of apomorphine compared with KO-wt mice, which is in agreement with the suggested ₂ agonistic properties of apomorphine.

Our results of unaltered dexmedetomidine-induced locomotor inhibition in both KO and OE mice indicate that _2A-ARs are involved in mediation of the sedative effect of dexmedetomidine. This assumption is now directly testable in transgenic mice with dysfunctional _2A-ARs (14). The inhibition of spontaneous locomotor activity in the current study reflects mainly sedation, and it is not very sensitive to other possible drug-induced changes in motor performance. Thus, _2C-ARs may have more subtle modulatory effects on locomotor functions, as suggested by their presence in the striatum. Marked strain differences were observed between KO-wt and OE-wt mice in all behavioral experiments, complicating the study designs, because both mutant strains needed their own control strains and direct comparisons between KO and OE mice were not appropriate. It has recently been emphasized that phenotypical abnormalities observed in mice with targeted gene disruptions may in fact be due to chance effects attributable to background genes that are unevenly distributed in the breeding process (34). However, the opposite findings in our two gene-manipulated mouse strains, one with targeted disruption of the Adra2c gene and the other with tissue-specific overexpression of the receptor encoded by this gene, point to a specific role of the _2C-AR in mediating modulation of DA metabolism and the hypothermic effect of ₂-AR agonists.

In conclusion, genetic alteration of receptor expression is a powerful new method of studying the physiological and pharmacological functions of a particular receptor, especially if selective agonists or antagonists are not available. This study supports the concept that the sedative effects of dexmedetomidine are probably mediated by _2A-ARs, which inhibit the activity of noradrenergic neurons in the locus ceruleus. In addition, because _2A expression has also been detected in dopaminergic and serotonergic control centers in rodent brain (4, 5) and _2B expression is very limited in brain, it seems that dopaminergic and serotonergic systems also are mainly regulated by _2A-ARs. Striatal _2C-ARs may, however, have subtle modulatory effects on the dopaminergic systems in brain. This could be studied further in _2C-AR-lacking KO mice and in OE mice (e.g., by exploring interactions in locomotor tests between adrenergic and dopaminergic drugs). Also, methods other than simple motor activity measurements that are specifically designed to monitor more complex aspects of motor performance might reveal a possible role of _2C-AR in locomotor modulation and coordination. The availability of mice with altered _2C-AR expression permits the analysis of the involvement of _2C-AR in many other ₂-AR-mediated functions, such as blood pressure control (15), metabolic regulation, sensory modulation, and higher behavioral functions such as learning and memory.