Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Obesity is a serious and increasingly prevalent health risk that impacts both the quality of life and life expectancy (Calle et al., 1999; Khaodhiar et al., 1999). Many pharmacological strategies for the treatment of obesity are under evaluation, including agents that reduce appetite and promote satiety (e.g., Sibutramine), inhibit nutrient absorption (e.g., Orlistat), and increase energy expenditure through activation of ₃-adrenergic receptors (Bray and Greenway, 1999). ₃-Adrenergic receptor activation stimulates lipolysis in white and brown adipocytes and increases energy expenditure in brown adipocytes (Cawthorne, 1992). This latter effect is mediated through the mitochondrial uncoupling protein UCP-1, whose activity is regulated in part by the interaction of free fatty acids with an allosteric binding site on the molecule (Himms-Hagen, 1992). Increasing energy expenditure has the potential to produce a negative shift in energy balance, and concomitant weight loss. However, besides producing metabolic changes, intravenous administration of ₃-adrenergic receptor agonists in several species results in cardiovascular changes, including a decrease in peripheral resistance, an increase in heart rate, and peripheral vasodilatation (Tavernier et al., 1992; Shen et al., 1994, 1996; Cohen et al., 1995).

Understanding the mechanism of ₃-adrenergic receptor agonist-induced changes in metabolic and cardiovascular function has been complicated by differences in the profile of ₃-adrenergic receptor agonist function in different species. Thus, several of the originally identified ₃-adrenergic receptor agonists (e.g., BRL 26830A, BRL 35135, CL 316243) were identified on the basis of their ability to stimulate lipolysis in rat adipocytes in the absence of ₁- or ₂-adrenergic receptor activation (Howe, 1993). When tested in rodent and canine models of obesity, these compounds caused an increase in metabolic rate, weight loss, and improved glucose tolerance (Arch and Ainsworth, 1983; Cawthorne et al., 1992; Himms-Hagen et al., 1994). However, weight loss studies with these same agents in humans were inconclusive and in some studies were complicated by tremors (a ₂-effect) and tachycardia (a ₁-effect) (Cawthorne et al., 1992). Subsequent studies in our laboratories (Naylor et al., 1998) using cloned human ₁-, ₂-, and ₃-adrenergic receptors demonstrated that these compounds are only weak partial agonists of the human ₃-adrenergic receptor. Moreover, none of the compounds were selective for the human ₃-adrenergic receptors but exhibited agonist activity at human ₁- and ₂-adrenergic receptors consistent with the reported side effect profile in human subjects.

Rodent-selective ₃-adrenergic receptor agonists have been reported to produce effects on systemic hemodynamics and regional blood flow distribution in the rodent (Takahashi et al., 1992; Cohen et al., 1995; Shen et al., 1996) and in the dog (Berlan et al., 1994; Shen et al., 1994, 1996). In the conscious rat, administration of the ₃-adrenergic receptor agonists BRL 37344 and CL 316243 produced significant reductions in mean arterial pressure and increases in heart rate (Cohen et al., 1995; Shen et al., 1996). Furthermore, in the dog, BRL 37344 elicits hypotension, tachycardia, and peripheral vasodilatation, particularly in skin and fat (Shen et al., 1994). In both rats and dogs ₃-adrenergic receptor agonist-induced tachycardia is believed to be an indirect effect, mediated via activation of a baroreflex mechanism. Thus, in pithed rats, which effectively lack baroreflex pathways, BRL 37344 at low doses elicits hypotension in the absence of significant tachycardia (Cohen et al., 1995). Some tachycardia was evident at higher doses of BRL 37344, which may reflect a direct activation of ₁-adrenergic receptors. In addition, disruption of baroreflex pathways in dogs via sino-aortic denervation abrogates the tachycardia evoked by the ₃-adrenergic receptor agonists BRL 37344 and CGP 12177 without altering cutaneous blood flow or systemic hypotension (Tavernier et al., 1992; Berlan et al., 1994).

The occurrence of significant cardiovascular and hemodynamic responses consequent upon ₃-adrenergic receptor agonist administration would clearly limit the clinical utility of ₃-adrenergic receptor agonists in the treatment of human obesity. To assess the cardiovascular sequelae of human ₃-adrenergic receptor agonist administration in appropriate susceptible species we have developed in the rhesus monkey a model for the concurrent assessment of cardiovascular and metabolic function. In this report we describe the in vivo profile of two potent and selective ₃-adrenergic receptor agonists, (R)-4-[4-(3-cyclopentylpropyl)-4,5-dihydro-5-oxo-1H-tetrazol-1-yl]-N-[4-[2-[[2-hydroxy-2-(3-pyridinyl)ethyl]amino]ethyl]phenyl]- benzenesulfonamide (compound A, Shih et al., 1999) and (R)-N-[4-[2-[[2-hydroxy-2-(3-pyridinyl)ethyl]amino]ethyl]phenyl]-1-(4-octylthiazol-2-yl)-5-indolinesulfonamide (compound B, Mathvink et al., 1999) that were identified using cloned and expressed human -adrenergic receptors. Furthermore, we have sought to investigate the underlying mechanisms of ₃-adrenergic receptor agonist-induced peripheral vasodilatation and tachycardia.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

General Procedure. All animal procedures performed in this study were reviewed and approved by the Merck Research Laboratories Institutional Animal Care and Use Committee.

Experimental Protocols. The dose dependence and kinetics of ₃-adrenergic receptor agonists for changes in cardiovascular and metabolic parameters were determined. Following collection of baseline values, compound A, compound B, or an equivalent volume (0.1 ml/kg) of vehicle (60% polyethylene glycol-400, 20% ethanol, 20% saline) was administered intravenously as a single bolus administration over a 2-min period. Blood pressure and heart rate data were collected and stored to computer every 5 min for 60 min postcompound administration. Arterial blood samples were collected before and 5, 15, 30, and 60 min postcompound administration for measurement of serum glycerol. Facial color was determined before and 3, 5, 7, 10, 12, 15, 30, 45, and 60 min postcompound administration. Metabolic rate was monitored continuously and reported as average values over 5-min intervals. Sixty minutes after compound administration animals were administered an infusion of isoproterenol (0.1 mg/kg) over 15 min to elicit a maximal increase in serum glycerol and an additional blood sample was collected.

Transfection of Chinese Hamster Ovary (CHO) Cells with Adrenergic Receptors, Binding, and Adenylate Cyclase Activity. Binding experiments were performed using cell membranes prepared from CHO cells stably transfected with cloned rhesus monkey -adrenoceptors. Rhesus monkey ₁-, ₂-, and ₃-adrenoceptors were cloned as described previously (Searles et al., 1994; Amend and Guan, 1995; Walson et al., 1997) and expressed in CHO cells. Stable transfectants expressing the cloned rhesus -adrenoceptors were grown in selective media for 3 days and membranes prepared by hypotonic lysis in 1 mM Tris, pH 7.2. Receptor binding assays were carried out in a final volume of 250 µl containing 5 to 10 µg of membrane protein, the radioligand [¹²⁵I]cyanopindolol at a concentration of 45 pM, and the compound of interest at various concentrations. Binding reactions were carried out for 1 h at 23°C, and terminated by filtration over GF/C filters using a 96-well cell harvester from Inotech (Lansing, MI).

Drugs and Chemicals. Compound A and compound B were prepared at Merck Research Laboratories, Rahway, NJ. DL-Isoproterenol hydrochloride, DL-propranolol hydrochloride, hexamethonium chloride, histamine hydrochloride, human CGRP, human CGRP_8-37, arachidonic acid, and mepyramine were from Sigma-Aldrich (St. Louis, MO). Ketamine hydrochloride (Ketaset) was obtained from Aveco Co., Inc. (Fort Dodge, IA) and sodium pentobarbital (Nembutal) was from Abbott Laboratories (North Chicago, IL). Atropine sulfate was from Radix Laboratories, Inc. (Eau Claire, WI), indomethacin from Merck and Co., Inc., (Rahway, NJ), and heparin from Elkins-Sinn, Inc. (Cherry Hill, NJ).

Data Analysis. Data are expressed as mean ± standard error of the mean. Data were analyzed by an analysis of covariance wherein the change from baseline for the drug treatment group was compared with the change from baseline for the vehicle control group. Data were considered significant at p < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Binding and Adenylate Cyclase Activity of ₃-Adrenoceptor Agonists in CHO Cells Stably Transfected with Rhesus Monkey -Adrenoceptors. Previously published studies demonstrated that compound A (Shih et al., 1999) and compound B (Mathvink et al., 1999) are potent and selective agonists in vitro for activation of the human ₃-adrenoceptor. Here we demonstrate, using cloned and expressed rhesus monkey -adrenoceptors, an equivalent profile of -adrenoceptor binding and adenylate cyclase activation with compound B and compound A to that seen using human -adrenoceptors. Thus, the EC₅₀ values for activation of adenylate cyclase by compound B and compound A, in cells expressing rhesus monkey ₃-adrenoceptors, were 0.43 ± 0.06 (mean ± S.D., n = 3) and 11 ± 6 nM (n = 7), respectively, and percentage of maximal activation with respect to isoproterenol of 80 ± 5 and 67 ± 11%, respectively. Furthermore, compound B and compound A demonstrated relatively weak binding to and activation of rhesus monkey ₁- (IC₅₀ values were 800 ± 156, n = 2, and 2439 ± 452 nM, n = 5, respectively, and percentage of maximal activation with respect to isoproterenol of 0 and 18 ± 1% at 10 µM, respectively) and ₂-adrenoceptors (IC₅₀ values were 315 ± 35, n = 2, and 1265 ± 761 nM, n = 4, respectively, and percentage of maximal activation with respect to isoproterenol of 9 ± 16 and 14 ± 11% at 10 µM, respectively). Thus, compound B is more than 700-fold and compound A more than 100-fold selective for activation of the ₃-receptor over activation or binding at ₁- and ₂-receptors.

Dose Dependence of 3AR Agonist-Induced Cardiovascular and Metabolic Responses in Rhesus Monkeys. Intravenous administration of compound A or compound B to anesthetized rhesus monkeys evoked dose-dependent increases in heart rate (Fig. 1A), peripheral vasodilatation, manifest as facial flushing (Fig. 1B), lipolysis (Fig. 1C), and metabolic rate (Fig. 1D). No significant changes in mean arterial pressure or serum potassium (an index of 2AR activation) were observed in response to either compound (data not shown). Consistent with the greater in vitro potency of compound B versus compound A at rhesus 3AR, the dose-response curves for compound B-evoked responses lie to the left of those for compound A-evoked responses. Thus, the ED₅₀ values for compound B- and compound A-evoked lipolysis are 0.2 and 0.6 mg/kg, respectively. Furthermore, the dose-response curves for compound B- and compound A-induced tachycardia lie to the right of their respective dose-response curves for stimulation of lipolysis. Thus, the ED₁₅ values (ED₁₅ is the dose of agonist producing a 15% increase in heart rate, approximately half of the maximal increase in heart rate seen under these conditions) for compound B- and compound A-induced tachycardia are 0.9 and 3 mg/kg, respectively.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Effects of compound A and compound B, administered by intravenous bolus, on heart rate (A), peripheral vasodilatation (B), lipolysis (C), and metabolic rate (D) in anesthetized rhesus monkeys. Compound A (0.1, 0.3, 1, or 10 mg/kg), compound B (0.03, 0.1, 0.3, 1, or 3 mg/kg), or vehicle (20% ethanol, 60% polyethylene glycol-400, 20% saline) was administered as a single intravenous bolus (0.1 ml/kg). Heart rate (% change from baseline), peripheral vasodilatation (change in colorimeter units), lipolysis (serum glycerol expressed as % of maximal response to isoproterenol), and metabolic rate (% change from baseline) were measured over a 60-min period postcompound administration. Data, reported for the time point 30 min postcompound administration, represent the mean ± S.E.M (n = 3-4 animals/dose group).

Kinetics of 3AR Agonist-Induced Cardiovascular and Metabolic Responses in Rhesus Monkeys. Following bolus intravenous administration compound B-induced tachycardia (Fig. 2A) was biphasic with an initial rapid increase in heart rate followed by a more slowly developing tachycardia. Peripheral vasodilatation (Fig. 2B), was rapid in onset with maximal changes occurring within 15 min of compound administration. Increases in lipolysis and metabolic rate developed more slowly than peripheral vasodilatation, were maximal within 15 min of compound administration, and were contemporaneous with the second phase increase in heart rate. The kinetics of compound A-evoked responses (data not shown) was similar to that of compound B. 

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Kinetics of the effects of compound B, in the presence and absence of propranolol, on heart rate (A), peripheral vasodilatation (B), lipolysis (C), and metabolic rate (D) in anesthetized rhesus monkeys. Compound B (1 mg/kg) or vehicle (20% ethanol, 60% polyethylene glycol-400, 20% saline) was administered as a single intravenous bolus (0.1 ml/kg) to rhesus monkeys that had received either propranolol (0.3 mg/kg) or saline. Heart rate (% change from baseline), peripheral vasodilatation (change in colorimeter units), lipolysis (serum glycerol expressed as % of maximal response to isoproterenol), and metabolic rate (% change from baseline) were measured over a 60-min period postcompound administration. Data represent the mean ± S.E.M. (n = 6 animals/dose group).

Effects of Propranolol on 3AR Agonist-Induced Cardiovascular and Metabolic Responses in Rhesus Monkeys. To explore the possibility that some or all of the responses evoked by compound B administration were mediated through activation of 1AR, the effects of propranolol (0.3 mg/kg i.v.) were investigated. We have shown that propranolol at this dose produces an approximately 10-fold shift to the right of the dose-response curve for isoproterenol-induced tachycardia with no effect on 3AR agonist-induced lipolysis in anesthetized rhesus monkeys (Forrest et al., 2000). In the presence of propranolol, there was significant attenuation of compound B-induced tachycardia (Fig. 2A) but no effects on compound B-induced peripheral vasodilatation (Fig. 2B), lipolysis (Fig. 2C), or increase in metabolic rate (Fig. 2D).

Effects of Ganglion Blockade on 3AR Agonist-Induced Cardiovascular and Metabolic Responses in Rhesus Monkeys. Ganglion blockade was achieved in anesthetized rhesus monkeys following the combined administration of hexamethonium (10 mg/kg i.v.) and atropine (1 mg/kg i.v.). The effectiveness of this regimen was indicated by attenuation of sodium nitroprusside-induced reflex tachycardia and phenylephrine-induced reflex bradycardia (Table 1A). Ganglion blockade did not significantly alter the vasodepressor efficacy of sodium nitroprusside nor the pressor efficacy of phenylephrine (Table 1B). Tachycardia evoked by compound B administration was significantly attenuated in ganglion-blocked animals (Fig. 3A), whereas peripheral vasodilatation (Fig. 3B), lipolysis (Fig. 3C), and increased metabolic rate (Fig. 3D) were equivalent in control and ganglion-blocked animals. Interestingly, whereas ₃-agonist administration has no significant effect on mean arterial pressure in control animals, compound B administration to ganglion-blocked monkeys evoked a rapid and sustained decrease in mean arterial pressure (Fig. 4).

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Kinetics of the effects of compound B, in the presence and absence of ganglion blockade, on heart rate (A), peripheral vasodilatation (B), lipolysis (C), and metabolic rate (D) in anesthetized rhesus monkeys. Compound B (1 mg/kg) or vehicle (20% ethanol, 60% polyethylene glycol-400, 20% saline) was administered as a single intravenous bolus (0.1 ml/kg) to rhesus monkeys that had received either hexamethonium (10 mg/kg) plus atropine (1 mg/kg) or saline. Heart rate (% change from baseline), peripheral vasodilatation (change in colorimeter units), lipolysis (serum glycerol expressed as % of maximal response to isoproterenol), and metabolic rate (% change from baseline) were measured over a 60-min period postcompound administration. Data represent the mean ± S.E.M. (n = 3 animals/dose group).

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Kinetics of the effects of compound B, in the presence and absence of ganglion blockade, on mean arterial pressure in anesthetized rhesus monkeys. Compound B (1 mg/kg) or vehicle (20% ethanol, 60% polyethylene glycol-400, 20% saline) was administered as a single intravenous bolus (0.1 ml/kg) to rhesus monkeys that had received either hexamethonium (10 mg/kg) plus atropine (1 mg/kg) or saline. Mean arterial pressure (change from baseline in mm Hg) was measured over a 60-min period postcompound administration. Data represent the mean ± S.E.M. (n = 3 animals/dose group).

Compound A Following Pretreatment with Saline, Propranolol, Indomethacin, Mepyramine, and CGRP_8-37 To determine whether the peripheral vasodilatation evoked following 3AR agonist administration was a direct effect of compound on the peripheral vasculature or was mediated through generation of a secondary endogenous vasodilator, a series of studies were performed comparing responses to 3AR agonist in the presence and absence of antagonists or inhibitors of the action or generation of known vasodilator agents. Accordingly, peripheral vasodilatation in response to the 3AR agonist, compound A (10 mg/kg i.v.) was determined in control animals and in animals pretreated with either the histamine H1 receptor antagonist mepyramine (5 mg/kg), the CGRP antagonist CGRP_8-37 (0.2 mg/kg/min), or the cyclooxygenase inhibitor indomethacin (5 mg/kg). The extent and duration of compound A-evoked peripheral vasodilatation was equivalent in control animals treated with saline and in animals administered either mepyramine, CGRP_8-37, or indomethacin (Fig. 5). That the doses of antagonist or inhibitor were appropriate to antagonize or inhibit their respective targets was established in simultaneous control studies (Table 2).

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Kinetics of the effects of compound B, in the presence and absence of pharmacological agents, on peripheral vasodilatation in anesthetized rhesus monkeys. Compound B (1 mg/kg) or vehicle (20% ethanol, 60% polyethylene glycol-400, 20% saline) was administered as a single intravenous bolus (0.1 ml/kg) to rhesus monkeys that had received either indomethacin (5 mg/kg), CGRP8-37 (0.3 mg/kg loading dose plus continuous infusion at 0.2 mg/kg/min), mepyramine (5 mg/kg), or saline. Peripheral vasodilatation (change in colorimeter units) was measured over a 60-min period postcompound administration. Data represent the mean ± S.E.M. for indicated (n) number of animals per dose group.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Administration of primate selective 3AR agonists to anesthetized rhesus monkeys evoked dose-dependent stimulation of lipolysis, increases in metabolic rate, peripheral vasodilatation, and tachycardia. 3AR-agonist administration did not change mean arterial pressure in control animals but evoked significant hypotension in animals that had been ganglion blocked. The effects of compound A and compound B on lipolysis and thermogenesis are believed to be mediated via activation of 3ARs. Thus, propranolol, at a dose that blocks isoproterenol-induced tachycardia in rhesus monkeys (Forrest et al., 2000), did not attenuate either lipolysis or thermogenesis evoked by 3AR agonist administration. Furthermore, activation of 2AR is unlikely given the absence of hypokalemia following compound administration. Shen et al. (1996) previously reported that the 3AR agonists BRL 37344 and CL 316243 failed to stimulate lipolysis in baboons when compounds were administered intravenously at a dose 10 times greater than required for stimulation of lipolysis in dogs. However, given species differences in 3AR agonist potency and effectiveness, and the low potency (approximately 1 µM) of BRL 37344 and CL 316243 for activation of the cloned human 3AR (Arch and Wilson, 1996), it remains a distinct possibility that the failure of BRL 37344 and CL 316243 to stimulate lipolysis in baboons is attributed to low potency of these compounds for activation of baboon 3AR.

Administration of compound A or compound B to anesthetized rhesus monkeys evoked significant tachycardia. The increase in heart rate was attenuated by propranolol, at a dose that did not reduce the evoked lipolysis or increase in metabolic rate, suggesting the tachycardia was proximally mediated, in part, via 1AR activation. It is unlikely that the increase in heart rate is a result of direct activation of 1AR by either compound A or compound B. This contention is based on the following observations. The selectivity of compound A or compound B for activation of 3AR versus activation of 1AR, when measured in vitro using cloned rhesus receptors, is between 100- and 1000-fold. However, the dose-response curves for compound B- and compound A-induced lipolysis and tachycardia are separated by only 0.5 to 1 log unit of each other, far less than would be predicted from the compounds' in vitro selectivity profiles. Furthermore, the time course of compound A- or compound B-induced tachycardia exhibits a distinct biphasic profile that differs from the monophasic time course observed in response to administration of 1AR agonists such as isoproterenol. Since the positive chronotropic effects of 3AR agonists were unlikely to be mediated via direct 1AR activation, but were attenuated by propranolol, it was suggested that the evoked tachycardia was due to activation of a reflex pathway with concomitant liberation of endogenous catecholamines. This hypothesis was supported by the observation that 3AR agonist-induced tachycardia was abolished in ganglion-blocked animals, whereas the lipolytic and thermogenic effects of 3AR agonist administration remained intact. Further evidence that 3AR agonist-induced tachycardia in other species is mediated through a baroreflex pathway was provided by Tavernier et al. (1992) and Berlan et al. (1994) using sino-aortic denervated dogs and by Cohen et al. (1995) using pithed rats. In these studies, 3AR agonist-induced tachycardia was either absent or significantly attenuated compared with responses in intact animals.

3AR agonist administration to control animals did not change mean arterial pressure but in ganglion-blocked animals the administration of compound B evoked significant hypotension. The decline in blood pressure was rapid, within 5 min, and contemporaneous with the initial vasodilator response to 3AR agonist administration. After the initial decline in blood pressure there was a partial restoration of pressure but it remained significantly below that of control animals for the duration of the study. The absence of a vasodepressor response to 3AR agonist in control animals, despite an equivalent degree of facial flushing, suggests the increase in heart rate was able to compensate for the decrease in peripheral resistance and maintain mean arterial pressure.

Although compelling evidence indicates that the mechanism of 3AR agonist-induced tachycardia is, at least in part, via activation of a baroreceptor reflex, the precise trigger(s) for this reflexogenic mechanism are unknown. A biphasic increase in heart rate following 3AR administration, the first phase temporally associated with peripheral vasodilatation and a second phase contemporaneous with the evoked increase in metabolic rate, suggests the reflex is provoked by at least two stimuli. We propose that the initial trigger for reflexogenic tachycardia in response to 3AR agonist administration is peripheral vasodilatation. In the current study, peripheral vasodilatation in response to 3AR agonists has been demonstrated through changes in skin color and has been demonstrated more directly in dogs using radiolabeled microspheres (Shen et al., 1994) and laser Doppler flowmetry (Berlan et al., 1994). In conscious dogs, 3AR agonist-induced vasodilatation is accompanied by a reduction in mean arterial pressure and peripheral resistance (Tavernier et al., 1992; Berlan et al., 1994; Shen et al., 1996) sufficient to activate baroreceptors and initiate a reflexogenic increase in heart rate. Partial attenuation of 3AR agonist-induced tachycardia with propranolol in conscious dogs (Shen et al., 1996) suggests that tachycardia is proximally mediated via both increased sympathetic output and withdrawal of parasympathetic myocardial tone. This contention is supported by our observations using pentobarbital anesthetized rhesus, where propranolol partially blocked 3AR agonist-induced tachycardia. With respect to the second phase of tachycardia, the mechanism also is dependent on intact reflex pathways, evidenced by attenuation of tachycardia in ganglion-blocked animals. However, it is unknown whether the trigger for activation of the reflex is vascular or metabolic in origin.

3AR agonist-induced vasodilatation appears to be a result of direct 3AR activation because the extent of vasodilatation was not reduced following propranolol treatment, or after ganglion blockade. Similar findings were reported in dogs by Shen et al. (1994) and Berlan et al. (1994) where 3AR agonist-induced vasodilatation was resistant to propranolol, sino-aortic denervation, or ganglion blockade, suggesting a direct effect of compounds on the canine vasculature. However, the possibility exists for a vasodilator mechanism secondary to generation or release of an endogenous vasodilator autocoid. To explore this possibility, common known vasodilator pathways were blocked using mepyramine, a histamine H₁ receptor antagonist, CGRP_8-37, a CGRP receptor antagonist, and indomethacin to prevent generation of vasodilator prostaglandins. Under none of these conditions was there a reduction of 3AR agonist-induced vasodilatation, suggesting that 3AR agonists have a direct effect on the rhesus vasculature to produce vasodilatation. Similar findings have been reported previously in dogs (Shen et al., 1994).

The presence of alternative (non-₁/₂) AR in the heart has been proposed based upon both in vitro and in vivo studies with a range of AR agonists and antagonists. Gauthier et al. (1996) reported that 3AR agonists (BRL 37344, CL 316243, and SR 58611) or weak partial agonists (CGP 12177) in the presence of 1AR and 2AR antagonism could evoke negative inotropism of human ventricular endomyocardial biopsies. However, Kaumann and Molenaar (1997) did not observe negative inotropism with 3AR agonists using human ventricular trabeculae. Furthermore, administration of 3AR agonists to rats, dogs, or rhesus did not evoke negative inotropism in vivo (Shen et al., 1996). In addition to the putative cardiodepressant 3AR reported by Gauthier et al. (1996), a putative cardiostimulant AR termed the 4AR has been proposed by Kaumann (1996) and Kaumann and Molenaar (1996). In these studies, nonconventional partial 3AR agonists such as CGP 12177, in the presence of 1AR and 2AR antagonism, exert positive inotropism and chronotropism in rat and human atrial preparations in vitro and positive chronotropism in pithed rats in vivo (Malinowska and Schlicker, 1996). The positive chronotropism of compound A and compound B reported here in anesthetized rhesus monkeys is unlikely to be mediated via this putative 4AR for the following reasons. First, the dose-response curves for lipolysis evoked by compound A and compound B lie to the left of those for tachycardia, whereas CGP 12177 when administered to anesthetized rhesus evokes glycerolemia and tachycardia over the same dose range (data not shown). Furthermore, positive chronotropism evoked by compound A and compound B is abolished in ganglion-blocked animals, whereas the cardiostimulant effects of CGP 12177 are present in pithed rats (Malinowska and Schlicker, 1996).

In conclusion, two 3AR agonists, identified on the basis of their high potency, efficacy, and selectivity for activation of the cloned and expressed human 3AR, have been examined in anesthetized rhesus monkeys. Administration of the compounds evoked dose-dependent increases in lipolysis, energy expenditure, heart rate, and peripheral vasodilatation. Tachycardia in response to 3AR administration was shown to be reflexogenic in origin and triggered, at least in part, by the vasodilator response to 3AR administration, which may be a direct consequence of activation of vascular 3AR.