Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

In 1989, the cloning of a gene encoding for a third -adrenoceptor (-AR) designated as the ₃-AR (Emorine et al., 1989) offered a putative explanation for those effects of catecholamines that could not be related to ₁- or ₂-AR stimulation. Since then, ₃-ARs have been characterized pharmacologically in a variety of tissues from human and laboratory mammals including dogs, rats, rabbits, guinea-pigs, and monkeys. ₃-ARs have been reported in white and brown adipose tissues, where they induced lipolytic and thermogenic effects (Arch et al., 1984; Zaagsma and Nahorski, 1990; Langin et al., 1991). They also have been identified in a number of gastrointestinal smooth muscle (Manara and Bianchetti, 1990; Koike et al., 1994), in the airways (Martin and Advenier, 1995), and in the blood vessels (Tavernier et al., 1992; Berlan et al., 1994; Shen et al., 1994), where they produced relaxation. However, in a given tissue, the pharmacological profile of the ₃-AR and the efficiency of the ₃-AR agonists were found to vary markedly depending on the species studied.

In the heart muscle, four populations of -AR potentially modulate the cardiac function. The effects of ₁- and ₂-AR are well established both in humans and other mammals. Their stimulation produces positive chronotropic and inotropic effects. Concerning the two other -ARs described more recently, less data are available. The atypical -AR, termed by some authors as the ₄-AR, caused positive inotropic and chronotropic effects. This receptor, which has not yet been cloned, has been characterized pharmacologically in vitro in cardiac preparations from rats, guinea pigs, cats (for review, see Kaumann, 1989, 1997), and humans (Kaumann, 1996) and in situ in the pithed rat (Malinowska and Schlicker, 1996, 1997). We have demonstrated previously the presence of ₃-AR in the human ventricle. In human cardiac tissues, ₃-AR stimulation by catecholamines in the presence of nadolol (a ₁- and ₂-AR antagonist) or by preferential ₃-AR agonists induces a pronounced concentration-dependent, negative inotropic effect (Gauthier et al., 1996). By contrast, the effects of ₃-AR stimulation in the hearts from nonhuman mammalian species are unknown. This has important pharmacological implications because the vast majority of therapeutic drugs are developed for human use but selected in animal models. The present study was designed to address this issue. According to the adipocyte responses to -AR agonists as reported by Lafontan (1994), the rat, dog, and guinea pig species were selected for this study. In addition, the ferret species also was selected because its ventricular myocardium exhibits contractile characteristics and adrenergic responses similar to those of the human myocardium (Cook et al., 1992). We found that ₃-AR stimulation produces very different effects in humans and in other mammal hearts and that the negative inotropic effects produced by ₃-AR agonists in human and dog ventricles are associated with the presence of ₃-AR transcripts. These findings are discussed in light of species- and tissue-specific effects.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Human Ventricular Biopsies. All protocols were approved by the Ethics Committee of the Centre National de la Recherche Scientifique (France). Twenty-four human endomyocardial biopsies were obtained from the right interventricular septum of cardiac transplant patients (20 men and 4 women; mean age, 51 ± 3 years) during right jugular vein catheterization performed routinely to detect possible rejection. None of the patients had evidence of cardiac rejection. All received immunosuppressive therapy (azathioprine, prednisolone, and cyclosporine). In addition, eight patients were given a calcium antagonist, and one patient was receiving a diuretic. Sixteen patients had no treatment known to possess cardiovascular effects. The effects of ₃-AR agonists in biopsies from patients treated with calcium antagonists were similar to those in biopsies from patients not receiving these drugs. Tissues were placed in a transport solution containing HEPES as the extracellular buffer and conveyed quickly to the laboratory.

Mammal Ventricular Tissues. The hearts of male Wistar rats (240-260 g), male guinea pigs (350-450 g), ferrets (0.8-1 kg), and male mongrel dogs (10-18 kg) were isolated under general anesthesia with sodium pentobarbital. Papillary muscles were dissected out from right and left ventricles.

Experimental Protocol. Preparations were placed in an experimental chamber and superfused at a flow rate of 5 ml/min with oxygenated (95% O₂; 5% CO₂) Tyrode's solution warmed at 37 ± 0.5°C. The Tyrode's solution for human tissues had the following composition: 116 mM NaCl; 5 mM KCl; 2.7 mM CaCl₂; 1.1 mM MgCl₂; 0.33 mM NaHPO₄; 24 mM NaHCO₃, and 5 mM glucose. For the other mammal cardiac tissues, this Tyrode's solution was modified slightly in agreement with the literature. Tissues were allowed to recover for at least 60 min and then submitted to field stimulation at a pacing cycle length of 1700 ms. Stimulus pulse width was 1 to 2 ms, and amplitude was twice the diastolic threshold. Tension was recorded by using a mechanoelectric force transducer (Akers, AE 801; SensoNor, Horten, Norway), as described previously (Gauthier et al., 1994). Ventricular tissues were stretched stepwise (10-µm increments) to a length at which contraction force was maximal. Studies then were performed at 90% of maximal tension. After equilibration, cumulative concentration-response curves for the various -AR agonists were determined by superfusion with increasing concentrations of the drugs. For all concentrations, tension was recorded at steady state on a digital storage oscilloscope (Gould 400; Gould, Les Ulis, France), a strip chart recorder (Gould 8188), and a digital tape recorder (DTR-1200; Biologic, Claix, France).

Statistics. Results are expressed as means ± S.E.M. of n number of experiments. The statistical significance of the effects of a drug was assessed by using one-way ANOVA followed by a Dunnett's test. To determine agonist potencies from the concentration-response curves, the EC₅₀ values were determined by fitting curves with the Bolzmann equation. pD₂ values then were calculated according to the equation pD₂ = log(EC₅₀).

Drugs. CGP 12177 (4-[3-t-butylamino-2-hydroxypropoxy]benzimidazol-2-one) was a gift from CIBA-Geigy (Basel, Switzerland), SR 58611 (N[2s)7-carb-ethoxymethoxy-1,2,3,4-tetra-hydronaphth]-(2r)-2-hydroxy-2(3-chlorophenyl) ethamine hydrochloride) was a gift from Sanofi Research (Montpellier, France), and CL 316 243 (5-(2-{[2-(3-chlorophenyl)-2-hydroxyethyl]-amino}propyl)-1,3-benzodioxole-2,2-dicarboxylate] was a gift from American Cyanamid Company (Pearl River, NY). Bupranolol was a gift from Schwartz Pharma (Mannheim, Germany). BRL 37344 (4-[[2-hydroxy-(3-chlorophenyl)ethyl-amino]propyl]phenoxyacetate) was obtained from Research Biochemicals International (Natick, MA), and nadolol was obtained from Sigma Chemical Co. (St. Louis, MO).

Reverse Transcription-Polymerase Chain Reaction (RT-PCR) Assay. Total RNA was prepared from white adipocyte and cardiac samples by using a single-step guanidinium thiocyanate/phenol/chloroform extraction (Chomczynski and Sacchi, 1987). Total RNA was treated with DNase I (Gibco/BRL, Cergy Pontoise, France). For myocardial samples, poly(A)⁺ RNA was purified by using the Dynabeads mRNA purification kit (Dynal, Skoyen, Norway). Isolated adipocyte total RNA (300 ng) and cardiac poly(A)⁺ RNA (10-50 ng) were treated with Superscript II RNase H reverse transcriptase (Gibco/BRL) and oligo(dT)12-18 primer (Pharmacia, St. Quentin en Yvelines, France) as described previously (Gauthier et al., 1996). A control without reverse transcriptase was performed to verify that amplification did not proceed from residual genomic DNA. cDNA was amplified by 40 cycles (92°C, 1 min; 55°C, 1 min; 72°C, 1 min) in PCR buffer containing 2.5 U Taq DNA polymerase (Perkin-Elmer, Courtaboeuf, France), 2.5 mM MgCl₂, 10% dimethyl sulfoxide (v/v), and 2.5% (v/v) formamide. Sense (5'-GCC TCC AAC ATG CCC TA-3') and antisense (5'-GCC TGC GGC AGT AGA TG-3') primers common to the rat and canine ₃-AR DNA (Lenzen et al., 1998) were expected to yield a 480-bp amplicon. Another combination of sense (5'-CGC TGA CGG GCC GCT GGC CTC TG-3') and antisense (5'-GCC ACC ACT TGC TCA TGA TGG GCG C-3') primers was used to amplify ₃-AR cDNA from dog myocardium. The expected length of the fragment was 243 bp. PCR products were separated by electrophoresis through 2% agarose ethidium bromide-stained gels. Gels were blotted onto nylon membranes (HybondN⁺; Amersham, Little Chalfont, UK) that were hybridized at 65°C to the human ₃-AR cDNA. Blots were washed at a final stringency of 15 mM NaCl, 1.5 mM sodium citrate, and 0.1% SDS at 65°C and subjected to digital imaging (Molecular Dynamics, Sunnyvale, CA).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effects of ₃-AR Agonists in Rat and Guinea Pig Hearts. Figure 1 shows the effects on rat papillary muscle contractility of BRL 37344, SR 58611, CL 316 243, and CGP 12177 at a concentration range of 0.1 nM to 1 µM. Among the drugs tested, only BRL 37344 significantly decreased peak tension with a pD₂ of 7.67 ± 0.64 (n = 6). The maximum negative inotropic effect was obtained for a concentration of 0.1 µM (Fig. 1A), which decreased peak tension by 21.6 ± 6.6% below control value (P < .05; n = 6). At a higher concentration (1 µM), BRL 37344 restored the contractile force to the control level. This effect was associated with a 6.4 ± 2.3% reduction in time-to-peak (P < .05; n = 6). Other contractile parameters were not modified significantly when compared with control values. The other ₃-AR agonists tested did not produce significant effects on contractile function in rat ventricular tissues. In papillary muscles from the guinea pig, only CL 316 243 induced a concentration-dependent reduction in peak tension with a pD₂ of 7.76 ± 0.85 (n = 5). At 1 µM, CL 316 243 decreased peak tension by 31.3 ± 2.4% (P < .05; n = 5) below control (Fig. 2A). This effect was associated with a 7.7 ± 1.5% reduction in time-to-peak (P < .05; n = 5). The other ₃-AR agonists caused no inotropic effects (Fig. 2B).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   The negative inotropic effect of the 3-AR agonists in rat ventricular myocardium. Papillary muscles were perfused with Tyrode's solution until a stable baseline contraction was obtained. Cumulative concentrations of the 3-AR agonist then were perfused for 10 min for each concentration to obtain a steady-state effect. A, superimposed twitches obtained with control solution at baseline and after 0.1 µM BRL 37344 (BRL) perfusion. B, concentration-response curves for the inotropic effect of BRL 37344, SR 58611, CL 316 243, and CGP 12177. Values are the means ± S.E.M. of n experiments. Response is expressed as the percentage of peak tension measured at baseline. *P < .05 versus control.

View larger version (16K): [in this window] [in a new window]   Fig. 2.   The negative inotropic effect of the 3-AR agonists in guinea pig ventricular myocardium. A, superimposed twitches obtained with control solution and after 1 µM CL 316 243 (CL) perfusion. B, concentration-response curves for the negative inotropic effect of BRL 37344, SR 58611, CL 316 243, and CGP 12177. Values are the means ± S.E.M. of n experiments. *P < .05 versus control.

Effects of ₃-AR Agonists in the Ferret Heart. In ferret, none of the ₃-AR agonists tested induced a negative inotropic effect as shown in Fig. 3B. Inversely to the rat and guinea pig hearts, BRL 37344 induced in the ferret ventricle a marked, positive inotropic effect at concentrations greater than 10 nM. At 1 µM, peak tension was increased by 240.0 ± 91.5% (P < .05; n = 5) as compared with control (Fig. 3A). At this concentration, the positive inotropic effect varied markedly depending on the preparation. This effect was associated with an abbreviation of the twitch. At 1 µM, BRL 37344 decreased total twitch duration by 12.7 ± 2.3% (P < .05; n = 5) and time-to-peak by 8.5 ± 5.0% (P < .05, n = 5), half-contraction time by 12.7 ± 6.6% (P < .05; n = 5), and half-relaxation time by 19.5 ± 3.0% (P < .05; n = 5).

View larger version (17K): [in this window] [in a new window]   Fig. 3.   The inotropic effect of the 3-AR agonists in ferret ventricular myocardium of ferret. A, superimposed twitches obtained with Tyrode's solution at baseline and after 1 µM BRL 37344 (BRL) perfusion. B, concentration-response curves for the inotropic effect of BRL 37344, SR 58611, CL 316 243, and CGP 12177. Values are the means ± S.E.M. of n experiments. *P < .05 versus control.

Effects of ₃-AR Agonists in Dog Hearts. The cardiac effects of ₃-AR agonists in dog are illustrated in Fig. 4. In this species, the ₃-AR agonist that induced the most efficient negative inotropic effect was CGP 12177, with a maximum effect obtained at 0.1 µM (Fig. 4A). At this concentration, CGP 12177 decreased peak tension by 26.9 ± 4.0% (P < .05; n = 5). The pD₂ for CGP 12177 was 9.33 ± 0.13 (n = 5). Other contractile parameters were not modified significantly. At higher concentrations, CGP 12177 increased peak tension. Both BRL 37344 and SR 58611 induced similar concentration-dependent negative inotropic effects (Fig. 4B) with pD₂ values of 8.78 ± 0.33 (n = 6) and 8.67 ± 0.24 (n = 6), respectively. At 0.1 µM, BRL 37344 reduced peak tension by 26.3 ± 4.6% (P < .05, n = 6), total twitch duration by 8.4 ± 3.0% (P < .05, n = 6), and time-to-peak by 5.0 ± 3.1% (P < .05, n = 6). At higher concentrations of BRL 37344, the control contractile force was restored. At 1 µM, SR 58611 decreased peak tension by 33.7 ± 5.7% (P < .05, n = 6) without significant alteration in the twitch kinetics. Finally, CL 316 243 induced no significant modification in cardiac contractility in the dog species (Fig. 4B). Therefore, in the canine ventricular myocardium, the rank order of potency of ₃-AR agonists to decrease peak tension was CGP 12177 > BRL 37344 = SR 58611  CL 316 243. 

View larger version (16K): [in this window] [in a new window]   Fig. 4.   The negative inotropic effect of the 3-AR agonists in the dog ventricular myocardium. A, superimposed twitches obtained with control solution at baseline and after 0.1 µM CGP 12177 (CGP) perfusion. B, concentration-response curves for the inotropic effect of BRL 37344, SR 58611, CL 316 243, and CGP 12177. Values are the means ± S.E.M. of n experiments. *P < .05 versus control.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Concentration-response curves for the negative inotropic effect of CGP 12177 in the presence of -AR antagonists in dog ventricular myocardium. Antagonists were perfused alone until a steady state was reached (20 min), which was taken as control (C). Cumulative concentrations of CGP 12177 were perfused in the presence or absence of the antagonist. Values are the means ± S.E.M. of n experiments. *P < .05 versus control.

Effects of ₃-AR Agonists in Human. In the human ventricular myocardium, all the ₃-AR agonists tested produced a concentration-dependent, negative inotropic effect (Fig. 6). BRL 37344 decreased peak tension at concentrations ranging from 0.1 nM to 1 µM. The maximum effect was observed at 1 µM, which decreased peak tension by 55.7 ± 3.7% (P < .05, n = 8) as compared with control (Fig. 6A). This effect was associated with an abbreviation of the twitch. At this concentration, BRL 37344 decreased total duration of the twitch by 12.5 ± 2.7% (P < .05, n = 8), time-to-peak by 11.3 ± 1.6% (P < .05, n = 8), half-contraction time by 11.4 ± 2.4% (P < .05, n = 8), and half-relaxation time by 17.5 ± 4.9% (P < .05, n = 8). The pD₂ for BRL 37344 was 8.75 ± 0.20 (n = 8).

View larger version (17K): [in this window] [in a new window]   Fig. 6.   The negative inotropic effect of the 3-AR agonists in human endomyocardial biopsies. A, superimposed twitches obtained with Tyrode's solution and after 1 µM BRL 37344 (BRL) perfusion. B, concentration-response curves for the negative inotropic effect of BRL 37344, SR 58611, CL 316 243, and CGP 12177. Values are the means ± S.E.M. of n experiments. *P < .05 versus control.

Detection of ₃-AR mRNA in Rat and Dog Ventricular Myocardium. To determine whether the presence of a ₃-AR-mediated negative inotropic effect was associated with the expression of ₃-AR transcripts, we used a RT-PCR assay. RT-PCR was performed with primers located in the first exon of the ₃-AR gene. RNA was treated with DNase I to prevent contamination by genomic DNA. We previously reported expression of ₃-AR mRNA in human myocardium (Gauthier et al., 1996). Using a similar protocol, two amplicons of expected sizes were obtained with different combinations of primers in dog myocardium (Fig. 7). ₃-AR mRNA expression was detected in four different animals. Hybridization to human ₃-AR cDNA confirmed the identity of the amplified products. No amplification was observed in rat myocardium (n = 3). As a control of the assay, reverse-transcribed ₃-AR mRNA was readily amplified by PCR in rat and dog white adipocytes (Fig. 7).

View larger version (32K): [in this window] [in a new window]   Fig. 7.   Detection of 3-AR mRNA in dog and rat myocardium. Poly(A)+ RNA from ventricle and total RNA from white adipocytes (WAT) were subjected (+) or not () to RT. PCR amplification was performed with two combinations of primers in dog heart that yielded the 243- and 480-bp amplicons. The sequences of the primers used to amplify the 480-bp product are conserved in the 3-AR genes of the two species and were used for PCR with dog WAT and rat heart and WAT cDNA. Top and bottom, negatives of agarose ethidium bromide-stained gel electrophoreses and Southern blots hybridized to a human 3-AR cDNA probe, respectively.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study reveals the complexity of ₃-AR pharmacology in the myocardium as illustrated by the pronounced interspecies variability and the heterogeneous pharmacological profiles of ₃-AR agonists in a given species. In human, all the ₃-AR agonists tested induced a marked negative inotropic effect. In dog, the negative inotropic effects of three ₃-AR agonists were less pronounced than in the human heart and virtually absent for CL 316 243. In addition, the rank order of potency of the various ₃-AR agonists differs markedly in the human and dog hearts. In rat and guinea pig, a significant decrease in peak tension (20-30%) was observed with only one agonist: BRL 37344 for the rat and CL 316 243 for the guinea pig. Finally, in ferret, none of the agonists induced negative inotropic effects. Based on these findings, we have classified species within three groups of responders defined as hyper-responders (human and dog), hyporesponders (rat and guinea pig), and nonresponders (ferret) to ₃-AR agonist stimulation in the myocardium (see Table 1). For the group of hyper-responders, the functional effects induced by ₃-AR agonists were associated with the presence of ₃-AR transcripts.

View this table: [in this window] [in a new window]   TABLE 1 Classification of species according to the intensity of their response to the 3-AR agonist stimulation

In the dog, the negative inotropic effect of CGP 12177, the most potent ₃-AR agonist identified in this species, was not modified by pretreatment with nadolol (a ₁- and ₂-AR antagonist), indicating that negative inotropy was not mediated by ₁- and ₂-AR. Inversely, bupranolol, which possesses ₃-AR antagonist properties, abolished CGP 12177 negative inotropic effect. These data are in agreement with our previous data obtained in human ventricular biopsies. The negative inotropic effect of BRL 37344 was not modified by pretreatment with nadolol, but was shifted to the right by bupranolol (Gauthier et al., 1996). SR 59230, a compound described as a selective ₃-AR antagonist in some tissues (De Ponti et al., 1996), did not antagonize at 0.1 µM the effect of SR 58611 in the human ventricle (data not shown). In the dog, CGP 12177 at the highest concentrations (1-100 µM) restored contractile force close to the control level. In addition, when the ₁-, ₂-, and ₃-AR were inhibited by bupranolol, a slight increase in contractility was observed. This latter effect could be attributed to stimulation of a putative ₄-AR as reported previously by Kaumann (1996, 1997). The existence of ₄-AR in the heart was deduced from the effects of partial agonists such as CGP 12177, cyanopindolol, and pindolol, which exert positive inotropic effects in vitro in atrial tissues from rat, guinea pig, cat, and human (Kaumann, 1989, 1996, 1997) and cause tachycardia in vivo in the rat (Malinowska and Schlicker, 1996, 1997). Positive chronotropic effects also have been reported with selective ₃-AR agonists such as BRL 37344 and CL 316 243 in dog and rat (Tavernier et al., 1992; Shen et al., 1994, 1996). However, this effect was not related to a direct stimulation of cardiac ₃-AR but, rather, to a reflex mechanism because it was abolished after sinoaortic denervation in conscious dogs (Tavernier et al., 1992) or after ₁- and ₂-AR blockade (Shen et al., 1996). In our study, another ₃-AR agonist, BRL 37344, when used at high concentrations, produced in some species an increase in contractile force or restored contractility to the control level. This effect likely resulted from a nonspecific activation of the ₁- and ₂-AR at high concentrations (Muzzin et al., 1992; Ida et al., 1996; Oriowo et al., 1996).

In the ventricular myocardium of human and dog, the ₃-AR stimulation produced a negative inotropic effect in stark contrast to ₁- and ₂-AR stimulation. In other tissues such as adipocytes, ₃-AR but also ₁-AR and ₂-AR stimulation produces lipolysis. However, as in the heart, the lipolytic effects of ₃-AR stimulation markedly vary depending on the species. As also reported in Table 1, three groups were defined previously by Lafontan (1994) in the adipose tissue: 1) a group of hyperresponders composed of rodents and hibernating mammals (Syrian hamster, dormouse) in which adrenergic stimulation produces lipolysis mainly through ₃-AR stimulation; 2) a group of hyporesponders, including rabbit, dog, and marmoset monkey, in which the three classes of -AR are equally involved in the lipolytic effect of catecholamines; and 3) a third group composed of guinea pig, baboon and macaque monkeys, and human, in which ₃-AR agonists induce a very weak response or even no response at all. Strong species-related differences also have been reported in other tissues. In the canine bronchi, selective ₃-AR agonists induced relaxation whereas these drugs produce almost no effect in human, guinea pig, and sheep bronchi (Martin and Advenier, 1995). In the colon, motility inhibition by ₃-AR agonists is more pronounced in the dog than in the rat or the guinea pig (Manara et al., 1995). In the vessels, ₃-AR agonists produce a much stronger vasodilator effect in dog than in rat. By contrast, ₃-AR agonists produce no vasodilator effects in nonhuman primates (Shen et al., 1996). Thus, interspecies variability in ₃-AR pharmacology not only concerns the heart but also other tissues. It is important to note that a species classified as a hyperresponder to cardiac ₃-AR agonist stimulation may be a weak responder to ₃-AR agonist stimulation in adipocyte (e.g., human). Conversely, rodents that are hyperresponders to adipocyte ₃-AR agonist stimulation are hyporesponders to cardiac ₃-AR agonist stimulation. Thus, interspecies variability differs depending on the organ.

Several explanations could account for interspecies differences related to ₃-AR agonist stimulation. One concerns ₃-AR expression. In the dog, the negative inotropic effect induced by ₃-AR agonists is associated with the expression of ₃-AR transcripts. These results agree with those obtained in human ventricle (Gauthier et al., 1996). By contrast, we did not detect ₃-AR mRNA in rat ventricular myocardium as shown previously in the rat right ventricle (Evans et al., 1996). The expression level may not be the sole factor that governs the response to ₃-AR stimulation. Sequences of ₃-ARs show either deletions or substitutions of key amino acids between species. The human, monkey, and bovine ₃-ARs have a higher interspecies homology in their amino acid sequence than rodents. In particular, in the first transmembrane region, a three-residue (valine-leucine-alanine) deletion is observed in the smaller but not in the larger mammals (Strosberg and Petri-Rouxel, 1996). However, deletion of these three residues from the human ₃-AR gene does not restore the "rodent-like" pharmacological profile, suggesting that these residues are not critically involved in interspecies specificity (Gros et al., 1998). In addition, the human ₃-AR is structurally different from its known homologs in other species. These differences concern transmembrane regions that are considered critical for ligand binding and G protein interaction (Strosberg and Petri-Rouxel, 1996). Clearly, further investigations are needed to clarify the consequences of the difference in ₃-AR structure across species. Another hypothesis is based on the exon-intron boundary of the ₃-AR gene. The ₃-AR gene contains two introns (Granneman et al., 1992; Lelias et al., 1993) in opposition to ₁- and ₂-AR genes, which are intronless. This structure leads to splice variants. The B and C isoforms contain 12 and 6 additional amino acids, respectively, at their C terminus in comparison with the A isoform (Granneman et al., 1992; Lelias et al., 1993). In rat adipocytes, a unique isoform is expressed that is close to the B isoform, whereas in human brown adipocytes, the C isoform is predominant (Lelias et al., 1993; Van Spronsen et al., 1993). It could be hypothesized that the physiological response to ₃-AR stimulation differs depending on the isoform expressed in a given species (Levasseur et al., 1995). Interestingly, the response to ₃-AR stimulation is variable in a given species depending the tissue. Again, human is a hyperresponder to ₃-AR stimulation in the heart but a weak responder in the adipose tissue. This discrepancy could result from expression of different isoforms in the heart and the adipose tissue. Further investigations are needed to determine the expression levels of the three isoforms in the various tissues, as well as their coupling pathway and their physiological roles. Indeed, differences in structure or isoforms between species and tissues could be responsible for a poor or high coupling of the ₃-AR or lead to a different coupling pathway. In adipose tissue, the three -ARs are linked to adenylyl cyclase through G_s proteins (Blin et al., 1993; Strosberg and Petri-Rouxel, 1996), whereas in the human ventricular muscle the ₃-AR, but not the ₁- and ₂-AR, is coupled to G_i/o proteins (Gauthier et al., 1996) and stimulates a nitric oxide synthase, leading to an increase in nitric oxide production and intracellular cGMP (Gauthier et al., 1998).

The present study demonstrates that caution should be taken in using animal models for the development of therapeutic compounds active on the human ₃-AR and that any novel drug targeted to ₃-AR should be evaluated on cardiac human tissue at least for safety reasons.

Acknowledgments

We thank Mr. Mortéza Erfanian for skillful technical assistance and Drs. Thierry Petit and Hervé L'Henaff from the Nantes Department of Cardiology for providing endomyocardial human biopsies.