Introduction

Top Summary Introduction Materials and Methods Results Discussion References

₂-Adrenergic receptors (ARs) play a prominent role in the regulation of the sympathetic nervous system (Ruffolo et al., 1991). Activation of ₂-ARs in the brainstem leads to a reduction in sympathetic tone, with a resultant decrease in heart rate and blood pressure. This effect is augmented by stimulation of ₂-ARs on sympathetic nerve terminals. These presynaptic ₂-ARs serve as autoreceptors regulating catecholamine release. There are three ₂-AR subtypes (_2A, _2B, and _2C), and studies using gene-targeting strategies indicate independent functions for each (Link et al., 1996; MacMillan et al., 1996; Sallinen et al., 1997). Resting blood pressure and heart rate were not significantly altered by disruption of either the _2B or _2C gene, indicating that neither receptor is necessary for normal sympathetic regulation (Link et al., 1996). Intra-arterial administration of clonidine-like ₂ agonists produced a biphasic blood pressure response in wild-type (WT) mice so an initial brief pressor effect was followed by a sustained fall in arterial blood pressure (Link et al., 1996; MacMillan et al., 1996). This is a characteristic cardiovascular response pattern of clonidine-like drugs in other mammals (Hoefke and Kobinger, 1966). The vasopressor response to ₂ agonists was absent in the _2B knockout (KO) mice, indicating that the _2B-AR mediates vasoconstriction in some vascular beds. The response to ₂ agonist was not altered in the _2C-deficient mice. Recent studies indicate that the _2C-AR plays a role in several aspects of behavior (Sallinen et al., 1998).

Much has been learned about the function of the _2A-AR from mice with a targeted mutation of the _2A-AR in the second transmembrane at position 79 (_2AD79N) (MacMillan et al., 1996). In cultured cell lines, the _2AD79N mutant receptor failed to activate K⁺ currents but exhibited normal inhibition of voltage-gated calcium channels and cAMP production (Surprenant et al., 1992). The _2AD79N mice were developed to study the physiological importance of K⁺ current regulation by the _2A-AR. Surprisingly, targeted mutation of the _2A-AR gene reduced expression of the _2AD79N mutant receptor by 80% as determined by radioligand binding assays in brain (MacMillan et al., 1996). _2AD79N mutant mice had normal resting heart rate and blood pressure. The initial hypertensive response to an ₂ agonist was similar in _2AD79N and WT mice; however, the hypotensive response to ₂ agonists was absent, demonstrating that the _2A-AR mediates this brainstem response.

We now report the physiological effects of disrupting the _2A-AR gene and describe differences between these mice and the _2AD79N mice. Cardiovascular studies show that unlike the _2AD79N mice, _2AKO mice have a resting tachycardia. This difference can be attributed to the loss of presynaptic autoregulation in _2AKO mice, which is preserved in _2AD79N mutant mice.

	Materials and Methods

Top Summary Introduction Materials and Methods Results Discussion References

Gene Targeting. The murine _2A-AR gene (4 kb) and flanking regions (5' arm, 3 kb; 3' arm, 6 kb) was subcloned into pBluescript II SK(). A 1672-bp neomycin resistance cassette (neo) containing the PGK promoter, neomycin resistance gene, and bovine growth hormone poly(A) sequence was inserted into a unique BglII site within the _2A-AR gene. The neo sequence was inserted in the opposite orientation relative to the _2A-AR gene, resulting in a premature termination codon within the third transmembrane domain. A herpes simplex virus thymidine kinase cassette (hsv-tk) was inserted downstream of the 6-kb 3' flanking region. R1 embryonic stem (ES) cells were electroporated with 40 µg of linearized targeting construct and placed under selection with G418 and ganciclovir. Resistant ES cell colonies were screened, and 37 of 132 were correctly targeted. Targeted ES cells clones were aggregated with eight cell FVB/N mouse embryos and cocultured overnight. The resulting blastocysts were transferred to pseudopregnant mice. Twenty-two male chimeric mice were generated and distinguished by agouti coat color carried by the R1 ES cell genome. Three of these chimeric mice transmitted the R1 ES cell genome with the disrupted _2A-AR locus to offspring. The generation of _2BKO mice, _2CKO mice, and _2AD79N mice has been described previously (Link et al., 1995, 1996; MacMillan et al., 1996).

Saturation and Competition Binding. Brain membranes were prepared by homogenizing whole brain in lysis buffer (10 mM Tris · HCl, 5 mM EDTA, pH 7.4), followed by centrifugation at 10,000g. The pellet was washed in Tris · HCl buffer (75 mM Tris · HCl, 12.5 mM MgCl₂, 1 mM EDTA, pH 7.4), followed by centrifugation at 10,000g. The pellet was resuspended in potassium acetate (KAC) binding buffer (50 mM KAC, pH 7.4), and protein concentration was determined. For ₂-AR saturation binding experiments, 180 to 250 µg of homogenate protein was used in a 500-µl reaction containing 1 to 10 nM [³H]RX81002, with or without 1 µM atipamezole and KAC binding buffer. For ₂-AR competition binding experiments, 200 to 250 µg of homogenate protein was used in a 500-µl reaction containing 1 nM [³H]RX81002, 1 to 1000 nM yohimbine, with or without 1 µM atipamezole, and KAC binding buffer. All binding assay mixtures were incubated at room temperature for 1 h. For -AR saturation binding, heart homogenates were prepared by Polytron homogenization of whole heart in lysis buffer, followed by centrifugation at 10,000g. The pellet was washed, and 50 to 100 µg of homogenate protein was used in a 500-µl reaction containing 1 to 300 pM [¹²⁵I]iodocyanopindolol ([¹²⁵I]CYP), with or without 1 µM (dl)-propranolol, in Tris · HCl binding buffer. The binding assay mixtures were incubated for 2 h. Binding reactions were terminated by filtration using a Brandel cell harvester. Membrane-bound [³H]RX821002 was determined by scintillation counting, and [¹²⁵I]CYP was determined by gamma emission. The results were analyzed with a nonlinear least-squares curve-fitting technique (Prism; GraphPAD, San Diego, CA).

In Vivo Cardiovascular Physiology. Studies were performed on eight WT and eight _2A-AR KO adult mice (10-20 weeks old) that were generated from _2A-AR heterozygote breeding. The mice were anesthetized with isoflurane (1-3%), and a polyethylene (PE10) catheter that had been stretched (0.5 mm o.d.) was inserted into the left internal carotid artery. The catheter was tunneled s.c. to exit at the base of the neck and placed within a s.c. pouch. After 24 h of recovery, the catheter was removed from the s.c. pouch, flushed with saline, and connected to a Spectramed DTX Plus pressure transducer. Heart rate and mean aortic blood pressure were recorded with a Gould eight-channel recorder and digitized on the Crystal Biotech Dataflow system (Hopkinton, MA). Baseline hemodynamics were continuously recorded for 1 h after the animal was placed in the study cage. To examine the role of vagal tone on baseline heart rate, atropine sulfate (1 mg/kg i.a.) was administered, and hemodynamics were recorded. Similarly, to examine the role of sympathetic tone on baseline heart rate, hemodynamics were recorded after propranolol administration (3 mg/kg i.a.). The next day, hemodynamic responses to dexmedetomidine (5 µg/kg i.a.) were recorded.

Tissue Norepinephrine Levels. Tissue norepinephrine concentrations were measured from the supernatants of whole heart and kidney homogenates. Tissue samples were homogenized on ice in 0.1 M sodium phosphate (pH 7.4), and a small sample was removed for protein determination. Perchloric acid was added (0.6 M final concentration), and samples were centrifuged at 10,000g for 3 h at 4°C. The resulting supernatants were analyzed by HPLC.

Vas Deferens Contractions. Vasa deferentia were isolated after sacrifice by cervical dislocation. The tissue was suspended on a force transducer and placed in an organ bath filled with physiological buffer solution consisting of 118 mM NaCl, 4.7 mM KCl, 3.0 mM CaCl₂, 1.22 mM KH₂PO₄, 25 mM NaHCO₃, and 10 mM glucose, oxygenated with a mixture of 95% O₂/5% CO₂. The samples were allowed to equilibrate for 30 min, and then 200 mg of resting tension was applied. Electrical field stimulation was applied to produce nerve terminal depolarization and neurotransmitter release (two electrical pulses every 10 s: 30 V, 0.9-ms width, 100-ms interval). The force of muscle contraction was digitized and displayed on a MacLab data analysis package. After contractions had equilibrated, dexmedetomidine was added to the organ chamber in a cumulative fashion every 1.5 min without changing the bath solution. In a separate group of animals, yohimbine was administered to the organ baths, also in a cumulative manner every 1.5 min without changing the medium.

[³H]Norepinephrine Release. The release of [³H]norepinephrine from mouse vas deferens was determined as described previously for other mouse tissues (Limberger et al., 1995; Wahl et al., 1996), with minor modifications. Briefly, small pieces of the vas deferens were incubated with [³H]norepinephrine (0.1 µM) in physiological buffer (Wahl et al. 1996) for 30 min. Tissue pieces were then superfused with [³H]norepinephrine-free medium containing 1 µM desipramine at a rate of 1.2 ml/min. Transmitter release was elicited by rectangular pulses of 1-ms width and 47-V/cm voltage, giving a current strength of 80 mA. There were six stimulation periods in each experiment, applied at intervals of 18 min (S₁ to S₆). Each period consisted of one train of 20 pulses at 50 Hz. Medetomidine was added at cumulatively increasing concentrations 12 min before S₂ to S₆. At the end of experiments, tissues were solubilized, and tritium was determined in superfusate samples and tissues. Electrically evoked overflow of total tritiated compounds was calculated as the difference between total tritium outflow and estimated basal outflow and was expressed as a percentage of tissue tritium at the time of stimulation (Wahl et al., 1996). A logistic curve was fitted to the concentration-inhibition data of medetomidine (Trendelenburg et al., 1993). The electrically evoked overflow of total tritium reflects exocytotic release of [³H]norepinephrine (Taube et al., 1977) and is termed thus in this report. Medetomidine (racemic) was used in these experiments because of the limited availability of the dextroisomer dexmedetomidine. Both compounds are nonselective ₂ agonists.

Statistical Analysis. All results are expressed as mean ± S.E.M. Baseline hemodynamics and tissue catecholamine levels were analyzed by independent t test. Responses to dexmedetomidine were analyzed by two-way ANOVA for repetitive measures. The Mann-Whitney test was used for statistical comparison between experimental groups in the catecholamine release studies.

	Results

Top Summary Introduction Materials and Methods Results Discussion References

Generation of _2A-AR-Deficient Mice. To disrupt the _2A-AR gene (Adra2a) in embryonic stem cells, a targeting vector was constructed that interrupted the coding region of the _2A-AR by insertion of a neomycin resistance cassette (neo; Fig. 1a). The disrupted _2A gene could be distinguished from the WT allele as a 5.1-kb HindIII fragment by a 5' Southern probe that lies outside the region of homology with the targeting vector (Fig. 1a). From the correctly targeted stem cell clones, three germline transmitting chimeras were generated by morula aggregation. One hundred thirty-one F₂ mice were generated from heterozygous _2A-AR (FVB/N, 129/SV) intercrosses and genotyped to determine whether disruption of the _2A-AR gene had a significant impact on development or viability before weaning. The distribution of WT, heterozygous, and homozygous (_2AKO) mice did not deviate significantly from that predicted by Mendelian genetics, indicating that there was no significant increase in mortality rates in _2A-AR-deficient mice (data not shown). KO mice could not be distinguished from WT or heterozygous littermates by appearance, body weight, fertility, or viability.

View larger version (37K): [in this window] [in a new window]   Fig. 1.   Generation of mice lacking a functional 2A-AR (Adra2a) gene. a, targeting vector for disruption of the Adra2a gene in mouse embryonic stem cells by homologous recombination. Insertion of the neomycin resistance gene (neo) disrupts the coding sequence (cds) of the 2A-AR. hsv-tk, herpes simplex virus thymidine kinase gene. b, Southern blot analysis of progeny from crosses of heterozygous mice carrying the mutant 2A allele. Using the external probe depicted in a, the WT allele can be detected as a 6.9-kb band. The disrupted allele (KO) is detected as a 5.1-kb band after HindIII digestion of genomic DNA.

_2A-AR Expression in Brain. Saturation binding with [³H]RX821002, a nonsubtype-selective ₂ antagonist, was performed on membranes prepared from brains of _2AKO and WT mice (Fig. 2A). _2AKO mice showed a 90% reduction in specific [³H]RX821002 binding (B_max: _2AKO 29 ± 2 fmol/mg protein; WT, 281 ± 27 fmol/mg protein). The residual ₂-AR binding in _2AKO mice was anticipated based on previous reports suggesting that ~10% of the ₂-AR in the brain is of the _2C subtype (Ordway et al., 1993). To confirm that the residual binding in the KO mice was not of the _2A-AR subtype, competition assays were performed with yohimbine (Fig. 2B). Yohimbine, an ₂ antagonist, has an unusually low affinity for the rodent _2A-AR subtype (Link et al., 1992). In the WT mice, yohimbine inhibited binding of [³H]RX821002 in a concentration-dependent manner displaying a small population (4 ± 6%) of high-affinity receptors (K_i = 1 nM) and large population of low-affinity receptors (K_i = 37 ± 3 nM). In KO mice, only high-affinity sites were present (K_i = 4.0 ± 0.3 nM), confirming that the residual ₂-AR in the KO mice is not of the _2A-AR subtype.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Characterization of 2-AR radioligand binding in brain membranes from 2AKO and WT mice. A, saturation binding of the 2-AR ligand [3H]RX821002 in WT () and 2AKO () membranes. Maximal specific binding of [3H]RX821002 is reduced by 90% in 2-deficient membranes compared with WT membranes. Nonspecific binding was determined in the presence of 1 µM atipamezole. B, competition of [3H]RX821002 binding to brain membranes by yohimbine. Each point represents the mean ± S.E.M. derived from three mice. Yohimbine displaced specifically bound [3H]RX821002 with higher affinity in 2A-deficient membranes than in membranes from WT mice.

Cardiovascular Physiology. Mean aortic blood pressure and heart rate were recorded 24 h after the insertion of a left carotid artery catheter and while the mice were quietly resting (Fig. 3). Heart rate was significantly higher in _2AKO mice (KO, 581 ± 21 min¹; WT, 395 ± 21 min¹). Mean aortic blood pressure was not significantly different (KO, 131 ± 8 mm Hg; WT, 128 ± 5 mm Hg). Atropine was then administered to determine whether the observed tachycardia in _2AKO mice was due to a reduction in vagal tone (Fig. 3). Atropine significantly increased heart rate in both WT and _2AKO mice. This increase was greater in WT mice, although heart rate remained significantly higher in KO mice. Propranolol was then administered to block sympathetic influence on heart rate. Propranolol produced a greater reduction in heart rate in KO mice so that heart rates were no longer significantly different between genotypes (KO, 479 ± 13 min¹; WT, 437 ± 23 min¹). In a separate group of WT and _2AKO mice, propranolol was administered without pretreatment with atropine. This was done to determine whether propranolol alone could normalize the heart rate. The reduction in heart rate was greater in KO mice, so the heart rates were not significantly different after -blockade (before propranolol: KO, 520 ± 46 min¹; WT, 409 ± 29 min¹; after propranolol: KO, 370 ± 32 min¹; WT, 365 ± 44 min¹, n = 3).

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Heart rate regulation in unrestrained, conscious WT and 2AKO mice. Baseline heart rate was significantly elevated in 2AKO mice compared with WT mice (*p < .001, KO versus WT). Injection of atropine led to an increase in heart rate in both groups of animals (¶p < .001, atropine versus control), but 2AKO mice were still tachycardic compared with WT mice (*p < .001, KO versus WT). Subsequent injection of the  antagonist propranolol abolished the difference in heart rate between WT and 2AKO mice.

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Hemodynamic effects of the 2 agonist dexmedetomidine on mean arterial blood pressure (a) and heart rate (b) in unrestrained, conscious WT and 2AKO mice. At time 0, a bolus of dexmedetomidine (5 µg/kg) was administered through the arterial catheter. In 2AKO mice, the initial hypertensive effect of the 2 agonist was greater than that in WT mice. However, the hypotensive effect of the 2 agonist was completely abolished in 2AKO mice. The bradycardic response to dexmedetomidine was significantly blunted in 2AKO mice compared with WT mice. Data are derived from eight animals per group (mean ± S.E.M.).

Tissue Norepinephrine Levels. The increased heart rate in _2AKO mice suggested that norepinephrine release from sympathetic terminals was enhanced in vivo. Because reliable determinations of plasma catecholamines are difficult to obtain in mice, tissue norepinephrine concentrations were determined in WT and _2AKO heart and kidney (Fig. 5a). The concentrations of norepinephrine in heart were significantly reduced in _2AKO compared with WT mice, suggesting that the sympathetic catecholamine stores are depleted due to enhanced transmitter release. Similarly, norepinephrine concentration were reduced in the kidney, but this achieved only borderline significance (p < .1).

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Tissue norepinephrine levels and cardiac -AR density are decreased in 2AKO mice. a, norepinephrine concentrations in heart and kidney of 2AKO mice are decreased compared with WT mice (*p < .01). Open columns, WT; solid columns, 2AKO. b, enhanced release of norepinephrine from sympathetic terminals leads to down-regulation of cardiac -ARs. Saturation isotherms for the -AR ligand [125I]CYP revealed a reduction in -AR density in heart membranes from 2AKO compared with WT mice. Data are derived from five or six animals per group (mean ± S.E.M.).

Cardiac -AR Down-Regulation. The results presented above suggest that loss of presynaptic autoregulation in sympathetic nerves of _2A-AR KO mice leads to baseline tachycardia and depletion of catecholamine stores. To investigate the effect of the loss of presynaptic _2A-AR function on postsynaptic ARs, we measured the density of cardiac -ARs using the nonselective  antagonist [¹²⁵I]CYP. It has previously been shown that chronic agonist infusion leads to down-regulation of cardiac -AR (Chang et al., 1982; Nanoff et al., 1989). Saturation binding studies revealed a significant reduction in cardiac -AR density in _2AKO mice compared with WT mice (Fig. 5b).

Presynaptic ₂-Autoreceptor Function. A vas deferens contraction assay was used to examine ₂-AR regulation of sympathetic transmitter release. Electrical stimulation of the vas deferens suspended in an organ bath leads to release of norepinephrine and ATP from intramural sympathetic nerve endings and subsequent contraction due to activation of ₁-ARs and P₂-purinergic receptors. Although the smooth muscle of the mouse vas deferens contains ₂-ARs, they play only a minor role in mediating neurogenic contractions (Bültmann et al., 1991). Stimulation of presynaptic ₂-AR on the sympathetic nerve terminals, however, markedly inhibits transmitter release and therefore the neurogenic contraction. Concentration-inhibition curves for the nonselective ₂ agonist dexmedetomidine in vas deferens isolated from the _2B-AR KO, _2C-AR KO, _2AD79N mutant, and WT mice are shown in Fig. 6. Dexmedetomidine at sufficient concentrations completely blocked the contractile response to electrical stimulation in vas deferens from all of these mice in a similar manner. The data indicate that disruption of the _2B-AR and _2C-AR genes or mutation of the _2A-AR (D79N) does not alter presynaptic function in sympathetic nerves. In contrast, the inhibition by dexmedetomidine in the _2AKO mice (Fig. 7a), although not abolished, was markedly impaired. The maximal inhibition of vas deferens contraction by the ₂ agonist in _2AKO mice was only 42 ± 9% compared with nearly complete inhibition in the WT mice. The concentration-inhibition curve of dexmedetomidine was also shifted to the right in _2AKO mice (EC₅₀: _2AKO, 4.2 ± 1.6 nM; WT, 0.7 ± 0.2 nM).

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Inhibition of electrically evoked contractions of isolated vas deferens preparations from WT and 2-AR-deficient mice. Stimulation of presynaptic 2-ARs by dexmedetomidine decreased the amplitude of the evoked twitch response. No difference was found for the inhibitory effect of dexmedetomidine on twitch contractions in WT and 2B- or 2C-deficient mice (a) or between WT and 2AD79N mice (b). Data represent mean ± S.E.M. from six to eight vas deferens preparations.

View larger version (37K): [in this window] [in a new window]   Fig. 7.   Presynaptic 2-AR function in vas deferens from 2AKO mice. a, the inhibitory effect of dexmedetomidine on electrically evoked twitch contractions was reduced by 56% in 2AKO mice compared with WT mice. b, the addition of the 2 antagonist yohimbine caused an increase in the twitch amplitude of vas deferens from WT and 2AKO mice. c, the concentration-response curve for yohimbine is shifted to the left in vas deferens from 2AKO compared with WT mice. d, the postsynaptic contractile effect of the 1 agonist phenylephrine is not different in 2AKO and WT vas deferens. Contractions were elicited by adding phenylephrine in a cumulative manner to nonelectrically stimulated vas deferens preparations. Contraction responses to phenylephrine were not altered in the presence of the 2 agonist dexmedetomidine (0.1 µM Dex). Data represent mean ± S.E.M. from six to nine vas deferens preparations.

View larger version (20K): [in this window] [in a new window]   Fig. 8.   Effect of medetomidine on [3H]norepinephrine release in mouse vas deferens. Pieces of vas deferens from WT or 2A-AR-deficient mice were preincubated with [3H]norepinephrine, superfused, and, beginning after 54 min of superfusion, subjected to six periods of electrical stimulation 18 min apart (S1 to S6). Each stimulation period consisted of a train of 20 pulses delivered at 50 Hz. Increasing concentrations of medetomidine were added in a cumulative manner 12 min before S2 to S6. a, tritium efflux-versus-time curves from single tissue pieces. Interrupted lines, controls without medetomidine; uninterrupted lines with  (WT) or  (2AKO) circles, medetomidine was added as indicated. b, concentration-response curves of medetomidine. The effect is corrected for time-matched controls. Mean ± S.E.M. from eight or nine tissue pieces.

	Discussion

Top Summary Introduction Materials and Methods Results Discussion References

Disruption of _2A-AR Gene. _2A-AR gene disruption was confirmed by ligand-binding experiments in whole brain. Disruption of the _2A-AR gene resulted in a 90% reduction in total _2A-AR binding in the mouse brain. The 10% residual binding is similar to previous estimates of the extent of _2C-AR expression in brain (Ordway et al., 1993). We found that the residual ₂-AR binding in _2AKO mice had high affinity for yohimbine. This finding is consistent with the higher affinity of yohimbine for the rodent _2B and _2C subtypes than for the _2A subtype (Link et al., 1992).

Role of _2A-AR in Regulating Heart Rate and Blood Pressure. ARs form the interface between the sympathetic nervous system and the cardiovascular system. The ₂-ARs also play an important role in regulating the sympathetic nervous system both centrally, by regulating sympathetic tone, and peripherally, by regulating transmitter release from presynaptic nerve terminals. We have used strains of genetically engineered mice to investigate the regulatory functions of specific ₂-AR subtypes. Previous studies of _2AD79N mice demonstrated that the _2A-AR subtype mediates the hypotensive effects of nonselective ₂ agonists (MacMillan et al., 1996). Several lines of evidence suggest that the hypotensive effect of ₂ agonists could result from actions at sites within the central and/or the peripheral sympathetic nervous system (DeJonge et al., 1981; Urban et al., 1995). However, our results indicate that activation of cardiac and vascular presynaptic ₂-ARs alone is not sufficient to produce hypotension in response to ₂ agonists. The data presented in Fig. 6 demonstrate that presynaptic regulation of catecholamine release is preserved in _2AD79N mice; however, these mice failed to become hypotensive in response to administered ₂ agonists (MacMillan et al., 1996).

More Than One Presynaptic ₂-Autoreceptor. The _2A-AR has been suggested to be the main presynaptic ₂-AR in mammalian tissues (Trendelenburg et al., 1993). However, previous studies indicated that at least in certain tissues such as the rat heart, a second ₂ subtype might also regulate neurotransmitter release (Limberger et al. 1992; Trendelenburg et al., 1997). Ho et al. (1998) studied the release-enhancing effect of ₂ antagonists and concluded that in addition to _2A-ARs, either _2B- or _2C-ARs mediate presynaptic inhibition in rat heart atria.