Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

-Adrenoceptors are cell membrane-bound G protein-coupled receptors that mediate physiological responses to the endogenous catecholamines epinephrine and norepinephrine. These receptors were initially classified into ₁- and ₂-adrenoceptor subtypes with each subtype being implicated in a major physiological role (Lands et al., 1967). The ₂-adrenoceptor is predominantly found in the lung and vasculature, where it mediates bronchorelaxation and vasodilation, respectively. On the other hand, the ₁-adrenoceptor is predominantly present in the heart, where it mediates positive inotropic and chronotropic effects. The ₁-adrenoceptor subtype was also thought to mediate lipolysis in adipocytes. However, the discovery of novel phenethanolamine agonists that potently stimulate lipolysis without significant inotropic and chronotropic effects suggested the presence of a novel -adrenoceptor subtype (Arch et al., 1984). This adrenoceptor was later cloned from human, rodent, and a number of other animal species, and is currently designated as the ₃-adrenoceptor (Emorine et al., 1989; Granneman et al., 1991; Strosberg, 1997a).

The ₃-adrenoceptor is predominantly found in white adipose tissue of rats where it stimulates lipolysis, and brown adipose tissue where it stimulates lipolysis and thermogenesis (Arch and Kaumann, 1993). In addition to ₃-adrenoceptors, rat adipocytes express ₁-adrenoceptors that can also stimulate lipolysis (Hollenga and Zaagsma, 1989). Selective stimulation of the ₃-adrenoceptor results in significant weight loss in obese rodents without adverse effects on ₁-adrenoceptor-mediated inotropic and chronotropic effects. In humans, the ₃-adrenoceptor is present in abundance in brown adipocytes of newborns. Adult humans lack significant quantities of this thermogenic tissue (Strosberg and Pietri-Rouxel, 1996; Danforth and Himms-Hagen, 1997). However, recent evidence suggests that adrenergic stimulation of white adipocytes can result in activation of dormant brown adipocytes or the white adipocytes themselves take on a brown adipocytic phenotype (Picó et al., 1998; Yoshida et al., 1998). A polymorphism of the ₃-adrenoceptor in humans has been associated with an increase in the body mass index, suggesting that this receptor plays a significant role in energy expenditure (Strosberg, 1997b). Recently, several ₃-adrenoceptor-selective agonists have been synthesized that lead to a remarkable decrease in body fat and improvements in insulin sensitivity in animal models of obesity (Arch and Wilson, 1996). ₃-Adrenoceptor agonists also mediate lipolysis in human white and immortalized brown adipocytes (Sennitt et al., 1998; Jockers et al., 1998). These studies suggest that ₃-adrenoceptor-selective agonists are promising candidates for the management of obesity and type II diabetes.

Trimetoquinol (TMQ) is a prototype of the tetrahydroisoquinoline class of compounds that is structurally distinct from catecholamines (Fig. 1). The compound lacks a -hydroxy group essential for the agonist activity of catecholamines and the amino nitrogen is constrained within a tetrahydroisoquinoline nucleus. However, a trimethoxybenzyl ring at the 1-carbon of the tetrahydroisoquinoline nucleus compensates for the lack of the -hydroxy functional group, and TMQ is a potent nonselective agonist of the three -adrenoceptor subtypes in functional assays (Feller et al., 1978; Konkar et al., 1996). Interestingly, 3',5'-diiodoTMQ, a 1-benzyl halogen-substituted analog of TMQ, was a potent activator of the ₃-adrenoceptor in rat tissues (Konkar et al., 1996). Consequently, additional modifications were made to increase the size (1-benzyl ring substituents) and perturb electronic characteristics (halogenation of the tetrahydroisoquinoline nucleus) of the parent TMQ molecule. Additionally, bioisosteric tetrahydrothiazolopyridine (THP) substitutions were designed to mimic the structural features of the reference ₃-adrenoceptor-selective arylpropanolamine 4-(3-t-butylamino-2-hydroxypropoxy)benzimidazol-2-one (CGP 12177). In this study we have characterized the -adrenoceptor activities of the resulting series of 1-benzyl ring halogen-substituted and tetrahydroisoquinoline ring-modified TMQ analogs on human -adrenoceptor subtypes. Our results indicate that halogenation of the 6,7-dihydroxy-catechol ring or replacement of the 6,7-dihydroxy-catechol ring (see Fig. 1) results in compounds that selectively activate the ₃-adrenoceptor as compared with ₁- and ₂-adrenoceptors. Thus, these compounds represent promising leads for the development of antiobesity and antidiabetic drugs.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Structures of TMQ, 1-benzyl-halogen-substituted analogs, 6,7-dihydroxy-catechol halogen-substituted analogs, and bioisosteric THP analogs. The  within a chemical structure indicates the presence of a center of asymmetry.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Radioligand Binding Assays. Radioligand binding assays were carried out as described previously (Fraundorfer et al., 1994). Briefly, Chinese hamster ovary (CHO) cells expressing either human ₁-, ₂-, or ₃-adrenoceptors were harvested into Ham's F-12 solution after digestion with trypsin. Cells were pelleted by centrifugation and washed three times with Tris-EDTA (composition: TRIZMA HCl, 50 mM; NaCl, 150 mM; and disodium EDTA · 2H₂O, 20 mM) buffer (pH 7.4). After centrifugation, the cells were suspended in Tris-EDTA buffer. Competition binding assays were performed in duplicate by incubating 1 to 5 × 10⁴ cells/18 to 70 pM [¹²⁵I]iodocyanopindolol (human ₁- and ₂-adrenoceptors) and 3 to 5 × 10⁵ cells/200 to 500 pM [¹²⁵I]iodocyanopindolol (human ₃-adrenoceptors), and varying concentrations of competing drugs for 1 h at 37°C. Nonspecific binding was determined in the presence of ()-propranolol [10⁶ M (human ₁ and ₂) and 10⁴ M (human ₃)]. The reactions were terminated by rapid filtration over Whatman GF/B (human ₁ and ₂) or Whatman GF/C (human ₃; presoaked in 0.1% polyethylenimine) glass fiber filters using a Brandel model 12-R cell harvester. The filters were washed three times with Tris-EDTA buffer (4°C) and dried under cell harvester vacuum. The radioactivity in the filters was determined by gamma scintillation counting (Beckman gamma counter model 8000; Beckman Instruments, Berkeley, CA).

cAMP Accumulation Assays. cAMP accumulation was determined in confluent cultures of CHO cells expressing either human ₁-, ₂-, or ₃-adrenoceptor subtypes as described previously (Fraundorfer et al., 1994). Briefly, CHO cells grown to confluence in 60-mm dishes were washed with Hanks' balanced salt solution. Cells were then incubated with Hanks' balanced salt solution (pH 7.4) containing 20 mM HEPES and 1 mM 3-isobutyl-1-methylxanthine and 1 mM L-ascorbic acid for 30 min at 37°C. Various concentrations of the drugs were then added and cells were incubated for an additional 30 min at 37°C. The Hanks' buffer was removed, the cAMP generated within the cells was extracted by the addition of trichloroacetic acid (6% w/v), and the precipitated protein was dissolved in 0.1 N NaOH. cAMP content was determined as the amount of ¹²⁵I-labeled succinyl-cAMP tyrosine methyl ester/antibody precipitated by radioimmunoassay as described by Brooker et al. (1979). Protein content was determined by the method of Lowry et al. (1951) using BSA as the standard.

Data Analysis. Data are expressed as the means ± S.E. of the given number of experiments. All concentration-response and competition binding curves were analyzed using GraphPad Prism (GraphPad Software, San Diego, CA). pK_act values are expressed relative to the maximal effect for each compound or effect at the highest concentration tested (for compounds with limited solubility). Relative efficacies (e_r) were calculated from plots of fractional percent occupancy versus response (% increase in cAMP accumulation) as described by Furchgott and Bursztyn (1967). The efficacies are expressed relative to ()-isoproterenol, a reference -adrenoceptor agonist. K_i values were calculated from IC₅₀ values obtained in competition binding experiments according to the method of Cheng and Prusoff (1973). Statistical difference between two data sets was determined by unpaired Student's t test at a significance level of 5%.

Drugs and Cell Culture. Chemicals were obtained from the following sources: ()-3-[¹²⁵I]iodocyanopindolol (2000 Ci/mmol) (Amersham, Arlington Heights, IL); fetal bovine serum, Geneticin, L-glutamine, Ham's F-12 medium, Hanks' balanced salt solution, penicillin-streptomycin solution, and trypsin-EDTA solution (Life Technologies, Gaithersburg, MD); BSA, 3-isobutyl-1-methylxanthine, ()-isoproterenol (+)-bitartrate, ()-propranolol HCl, and TRIZMA HCl (Sigma Chemical Co., St. Louis, MO). The isomers of TMQ [()-(S)-TMQ, _D = 28.5 (99% stereochemical purity) and (+)-(R)-TMQ, _D = +29.0 (99.78% stereochemical purity)] and 7-chloroTMQ were generous gifts from Tanabe Seiyaku Co. Ltd. (Osaka, Japan). The isomers of TMQ used in this study came from a single batch. (±)-TMQ and the 1-benzyl ring halogen-substituted analogs and bioisosteric aminothiazole analogs of TMQ were provided by Dr. Duane D. Miller (Department of Pharmaceutical Sciences, University of Tennessee, Memphis, TN). Chemical structures of the compounds used in this study are illustrated in Fig. 1. Unless specifically stated in the text and figure legends, the compounds used in cAMP accumulation and radioligand binding assays were racemic mixtures. The TMQ analogs and reference standards were dissolved at their highest concentrations in double distilled water and diluted appropriately. All other chemicals were of reagent grade.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

-Adrenoceptor Activities of TMQ and Isomers. Racemic - and ()-TMQ were found to be potent nonselective activators of the human -adrenoceptor subtypes in cAMP accumulation assays (Table 1). The isomers of TMQ exhibited high stereoselective activation of the -adrenoceptor subtypes and ()-TMQ was 214-, 281-, and 776-fold more potent than (+)-TMQ at stimulating ₁-, ₂-, and ₃-adrenoceptor subtypes, respectively (Table 1). The isomers in general displayed slightly lower stereoselective receptor affinities and ()-TMQ displayed 123-, 331-, and 5-fold greater affinity than (+)-TMQ for ₁-, ₂-, and ₃-adrenoceptor subtypes, respectively (Table 2). A ratio of efficacy of the test compound to that of ()-isoproterenol exhibits the relative efficacy (e_r) of the compound. TMQ exhibited efficacies at the ₁(()- and (±)-TMQ)- and ₂(()-TMQ)-adrenoceptors that were lower relative to ()-isoproterenol (e_r = 1). However, the efficacies of ()-TMQ and (±)-TMQ at the ₃-adrenoceptor were 8.2- and 3.4-fold greater relative to the efficacy of ()-isoproterenol (e_r = 1) (Fig. 2; Table 2).

View this table: [in this window] [in a new window]   TABLE 1 Agonist potencies (pKact valuesa) for stimulation of cAMP accumulation and intrinsic activities of standard -adrenoceptor ligands, trimetoquinol, selected halogen-substituted and bioisosteric aminothiazole analogs of TMQ at human -adrenoceptor subtypes expressed in CHO cells

View this table: [in this window] [in a new window]   TABLE 2 Relative efficacies of TMQ isomers at -adrenoceptor subtypes expressed in CHO cells

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Effects of TMQ isomers on activation of human 1 (A)-, 2 (B)-, and 3 (C)-adrenoceptor subtypes expressed in CHO cells. cAMP accumulation elicited by the ()-isomer () and (+)-isomer () is expressed as a percentage of maximal response to ()-isoproterenol. The data are presented as means ± S.E. of n = 5 to 11 experiments.

-Adrenoceptor Activities of 1-Benzyl Ring Halogen-Substituted Analogs. Previous studies demonstrated that halogenation of the 1-benzyl ring resulted in compounds that show ₃-adrenoceptor selectivity in rat tissues (Konkar et al., 1996). We therefore evaluated the activities of a series of 1-benzyl ring halogen-substituted analogs of TMQ, namely, the 3',5'-diiodo-4'-methoxy and 3',5'-diiodo-4'-desmethoxy and the 3',4',5'-triiodo derivatives of TMQ (Fig. 1). All three analogs exhibited high potencies and receptor affinities for the three -adrenoceptor subtypes (Tables 1 and 3, respectively). Interestingly, iodination of the benzyl ring resulted in compounds that were potent yet partial agonists at the ₂-adrenoceptor, whereas at the ₁- and ₃-adrenoceptors these analogs displayed maximal activities equal to or greater than that of ()-isoproterenol (Fig. 3; Table 1). A comparison of their potencies suggests that these compounds exhibit moderate selectivity for activation of ₃- over ₁-adrenoceptor, and maximal activities at the ₃-adrenoceptor that are significantly higher than those for ₂-adrenoceptors (Table 1). The affinities of these compounds at ₁- and ₂-adrenoceptors were in general greater than those at the ₃-adrenoceptor (Table 3). Comparison of the potency to affinity values shows that these compounds have greater efficacy (pK_act-pK_i values) at the ₃-adrenoceptor as compared with ₁- and ₂-adrenoceptors.

View this table: [in this window] [in a new window]   TABLE 3 Binding affinities (pKia values) of standard -adrenoceptor ligands, TMQ, selected halogen-substituted and bioisosteric aminothiazole analogs of TMQ for human -adrenoceptor subtypes expressed in CHO cells

View larger version (22K): [in this window] [in a new window]   Fig. 3.   cAMP accumulation stimulated by 1-benzyl halogen-substituted analogs of TMQ [3',5'-diiodoTMQ (A), 3',5'-diiodo-4'-desmethoxyTMQ (B), and 3',4',5'-triiodoTMQ (C)] in CHO cells expressing human 1 ()-, 2 ()-, and 3 ()-adrenoceptor subtypes. cAMP accumulation stimulated by the compounds is presented as a percentage of the maximal response to ()-isoproterenol. The data are presented as means ± S.E. of n = 3 to 12 experiments.

-Adrenoceptor Activities of TMQ Analogs with Modifications of the Tetrahydroisoquinoline Nucleus and Bioisosteric THP Derivatives. The tetrahydroisoquinoline ring containing the 6,7-dihydroxy catechol is critical for the activity of TMQ at -adrenoceptors (Fraundorfer, 1993). We examined the effects of halogen substituents at this ring on -adrenoceptor affinity and activation (Tables 1 and 3). All analogs exhibited affinities comparable to TMQ at each of the three -adrenoceptor subtypes. However, substitution of fluoro groups at the 8- and 5,8-positions of the catechol ring resulted in compounds (8-fluoroTMQ and 5,8-difluoroTMQ; Fig. 1) that potently activated ₃-adrenoceptors with approximately 10-fold greater selectivity over ₁- and ₂-adrenoceptors (Fig. 4; Tables 1 and 3). Replacement of the 7-hydroxy group with a chloro moiety (7-chloroTMQ) resulted in a highly selective ₃-adrenoceptor agonist (Table 1). This compound lacked significant ₁- and ₂-adrenoceptor activities at the highest concentrations tested, and its potency was comparable to the prototypical ₃-adrenoceptor agonist (RR,SS)-(±)-4-[2-(2-(3-chlorophenyl)-2-hydroxyethyl)amino)propyl] phenoxyacetate (BRL 37344; Fig. 4; Table 1). Similarly, modification of the 6,7-dihydroxy catechol nucleus to a corresponding bioisosteric aminothiazole derivative resulted in 3', 4', 5'-trimethoxybenzyl THP, a selective agonist of the ₃-adrenoceptor which exhibited weak ₁- and ₂-adrenoceptor activity (Table 1; Fig. 5). Halogen atom replacement of two of the three methoxy groups on the 1-benzyl ring of this THP nucleus yielded a 3',5'-diiodo derivative with increased potency (65-fold) and maximal activity at the ₃-adrenoceptor. Moreover, this 3',5'-diiodo THP analog was inactive at ₁- and ₂-adrenoceptors at the highest concentrations tested (Fig. 5; Table 1).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   cAMP accumulation stimulated by 6,7-dihydroxy halogen-substituted analogs of TMQ [8-fluoroTMQ (A), 5,8-difluororimetoquinol (B), and 7-chloroTMQ (C)] in CHO cells expressing human 1 ()-, 2 ()-, and 3 ()-adrenoceptor subtypes. cAMP accumulation stimulated by the compounds is presented as a percentage of the maximal response to ()-isoproterenol. The data are presented as means ± S.E. of n = 5 to 6 experiments.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   cAMP accumulation stimulated by bioisosteric aminothiazole analogs of TMQ [3',4',5'-trimethoxybenzylTHP (A) and 3',5'-diiodo-4'-methoxybenzylTHP (B)] in CHO cells expressing human 1 ()-, 2 ()-, and 3 ()-adrenoceptor subtypes. cAMP accumulation stimulated by the compounds is presented as a percentage of the maximal response to ()-isoproterenol. The data are presented as means ± S.E. of n = 3 to 6 experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

TMQ is a prototype of the tetrahydroisoquinoline class of compounds, and the ()-isomer of TMQ is used as a bronchorelaxant in Japan due to its agonist activity at the ₂-adrenoceptor. Previous studies have extensively characterized the effects of modifications of this molecule on ₂- versus ₁-adrenoceptor activities in attempts to improve the ₂-adrenoceptor selectivity profile of TMQ (Clark et al., 1987; Fedyna et al., 1987; Shams et al., 1990). Recent functional studies have demonstrated that, in addition to being a potent ₁- and ₂-adrenoceptor agonist, TMQ is a potent stimulant of ₃-adrenoceptor-mediated activities in rat tissues. These include lipolysis in brown adipocytes and relaxation of the esophageal smooth muscle (Konkar et al., 1996; Lezama et al., 1996). 1-Benzyl ring halogen-substituted analogs of TMQ were also found to be potent and preferential activators of ₃-adrenoceptors in rat tissues (Konkar et al., 1996). The role of ₃-adrenoceptors in reduction of obesity and improvement of glucose tolerance makes this receptor an attractive target for selective ₃-adrenoceptor ligands that may be of value in the treatment of obesity and diabetes (Arch and Wilson, 1996; Strosberg, 1997a). We have characterized the biochemical activity and receptor affinity of TMQ and a selected series of TMQ analogs at the three human -adrenoceptor subtypes with the aim of developing ₃-adrenoceptor-selective compounds.

The isomers of TMQ exhibited stereoselective interaction with all three human -adrenoceptors studied, and ()-TMQ was the more potent isomer in both -adrenoceptor binding and activation assays. Interestingly, the difference in the affinity of the isomers for the ₃-adrenoceptor was significantly lower than the difference observed for activation of the ₃-adrenoceptor. There was no such difference in the isomers between receptor binding and receptor activation of ₁- and ₂-adrenoceptors. The ₃-adrenoceptor binding site is less sterically hindered than the ₁- and ₂-adrenoceptor binding sites, as there are fewer bulky amino acid side chains that point into the ligand-binding pocket. This allows fairly large molecules to be accommodated in the binding site (Blin et al., 1993; Granneman et al., 1998), and could account for the low difference (5-fold) in stereoselectivity of the TMQ isomers observed in binding assays. However, the high difference (776-fold) observed for activation suggests that the ₃-adrenoceptor does exhibit stringent chiral requirements for activation. These results are consistent with the stereoselective activation of -adrenoceptor subtypes by TMQ isomers observed in rat tissues (Fraundorfer et al., 1994; Konkar et al., 1996). The ₃-adrenoceptor is generally regarded to display low stereoselectivity for interaction with chiral compounds (Strosberg and Pietri-Rouxel, 1996). Our results with isomers of high isomeric purity demonstrate that the ₃-adrenoceptor does in fact display high stereoselective activation. The large trimethoxybenzyl ring at the 1-carbon of TMQ, which is essential for -adrenoceptor activity, seems to be optimally oriented for potent ₃-adrenoceptor activation in the S-configuration. In contrast, R-TMQ only weakly activates the receptor even though it binds to ₃-adrenoceptors with nearly the same affinity as the S-isomer. The low stereoselectivity exhibited by the ₃-adrenoceptor can be exploited for the synthesis of ₃-adrenoceptor antagonists through the chiral synthesis of derivatives exhibiting high receptor affinity but poor receptor activation. Furthermore, it is important to note that the active isomer possesses greater efficacy at the human ₃-adrenoceptor than ()-isoproterenol. This observation is of significance because most selective ₃-adrenoceptor agonists exhibit poor efficacy at the human ₃-adrenoceptor (Arch and Wilson, 1996). Synthesis of -adrenoceptor agonists that possess high efficacy and selectivity for the human ₃-adrenoceptor may yield compounds of therapeutic value in the treatment of obesity and diabetes (Arch and Wilson, 1996). Therefore, we have modified TMQ, a highly efficacious ₃-adrenoceptor agonist, at the 1-benzyl ring and the 6,7-dihydroxy-catechol nucleus with the aim of increasing ₃-adrenoceptor selectivity.

Halogen substitution at the 1-benzyl ring of TMQ, as in 3', 5'-diiodoTMQ resulted in a compound with partial ₁- and ₂-adrenoceptor activity and ₃-adrenoceptor selectivity in rat tissues (Konkar et al., 1996). However, results from rat tissues do not always translate into a similar activity profile in human receptor systems. A number of highly potent rat ₃-adrenoceptor-selective agonists exhibit only partial agonist activity and lack selectivity for the human ₃-adrenoceptor (Arch and Wilson, 1996). Hence, we investigated the activities of the 1-benzyl ring halogen-substituted analogs and compared their activities at all three -adrenoceptor subtypes. The substitution of iodine atoms for methoxy groups at the 1-benzyl ring of TMQ provides an increase in bulk, electronegativity, and lipid solubility. Because ₃-adrenoceptors are theorized to accommodate large N-substituted lipophilic ligands as opposed to ₁- and ₂-adrenoceptors (Blin et al., 1993), substitution of an iodo group was predicted to result in a potent ₃-adrenoceptor-selective agonist. However, these substitutions resulted in compounds with high potency and affinity for all three -adrenoceptor subtypes. Notably, the 1-benzyl iodo-substituted compounds exhibited partial agonist activity at the ₂-adrenoceptor. These compounds also elicited "suprafull" agonist responses, as compared with ()-isoproterenol, at ₁- and ₃-adrenoceptors and were selective ₃-adrenoceptor agonists. We propose that: 1) functional group modifications at the 1-benzyl ring such as substitution of fluoro, chloro, or bromo atoms change the electronic characteristics of the ring; or 2) the substitution of an acidic moiety at the 4'-position of the 1-benzyl ring, as in the prototypical ₃-adrenoceptor-selective agonists disodium (RR)-5-[2-[[2-(3-chlorophenyl)-2-hydroxyethyl]-amino]propyl]-1,3-benzodiaoxazole-2,2-dicarboxylate (CL 316243) or BRL 37344, may improve the ₃-adrenoceptor selectivity profile without a loss of efficacy of these derivatives.

TMQ analogs with halogen substitution at the 6,7-dihydroxy-catechol nucleus have been investigated earlier in rat tissues to determine ₁- versus ₂-adrenoceptor activity (Clark et al., 1987; Markovich et al., 1992). Similar substitutions on the catechol nucleus of norepinephrine and epinephrine resulted in compounds that exhibit ₂-adrenoceptor selectivity due to electronic effects of the halogen substituents (Kirk and Creveling, 1984). These TMQ analogs have not been evaluated for their activity on the human -adrenoceptor subtypes, especially the ₃-adrenoceptor. In this study we analyzed compounds with halogen substitutions at the 8-, 5,8-, and 7-positions of the 6,7-dihydroxy-catechol nucleus for activity at human -adrenoceptor subtypes. The first two substitutions resulted in compounds that displayed slightly greater potency at the ₃-adrenoceptor as compared with ₁-adrenoceptor, and they displayed partial agonist activity at the ₂-adrenoceptor. Interestingly, the latter substitution of the 7-hydroxy group with a chloro group yielded a highly selective activator of the ₃-adrenoceptor that had insignificant activity at ₁- and ₂-adrenoceptors. However, the compound displayed partial agonist activity at the ₃-adrenoceptor. As discussed earlier, halogenation of the 1-benzyl ring of TMQ resulted in "suprafull" agonists; therefore, we propose that similar 1-benzyl ring modifications of the 6,7-hydroxy catechol-modified derivatives may yield potent compounds with high ₃-adrenoceptor selectivity and efficacy.

We investigated further the effect of modification of the 6,7-dihydroxy-catechol group of TMQ on -adrenoceptor activity and selectivity by synthesizing bioisosteric aminothiazole analogs. In these compounds the 6,7-dihydroxy-tetrahydroisoquinoline ring was replaced with a THP nucleus (see Fig. 1). It has been shown previously that aminothiazoles constitute bioisosteric replacement for the catechol ring of catecholamines (Schneider and Mierau, 1987). This substitution yielded compounds that are structurally similar to the ₃-adrenoceptor agonist CGP 12177 (see Fig. 1). As expected, replacement of the catechol-containing tetrahydroisoquinoline ring of TMQ with the THP nucleus resulted in a ₃-adrenoceptor-selective agonist with only weak activity at ₁- and ₂-adrenoceptors. Interestingly, halogen substitution of the benzyl ring of this THP analog resulted in a 65-fold increase in potency along with an increase in maximal activity at the ₃-adrenoceptor, accompanied by a suppression of ₁- and ₂-adrenoceptor activity.

Our results demonstrate that the ()-isomer of TMQ is a highly potent, nonselective -adrenoceptor agonist that possesses high efficacy at the ₃-adrenoceptor. 1-Benzyl ring halogen-substituted analogs are highly potent and efficacious activators of ₃-adrenoceptor that are moderately selective. Whereas halogen substitutions on the 6,7-dihydroxy-catechol nucleus resulted in compounds with similar activities, substitution of the hydroxyl group with a chloro atom as in 7-chloroTMQ, or replacement of the 6,7-dihydroxy-tetrahydroisoquinoline ring with a bioisosteric THP nucleus, yielded highly selective ₃-adrenoceptor agonists. Importantly, the 3',5'-diiodo-4'-methoxybenzyl THP analog exhibited a greater potency and maximal activity than the corresponding 3',4',5'-trimethoxybenzyl derivative (Fig. 5, Table 1). Additional modifications and stereoselective synthesis of these TMQ analogs will be expected to yield compounds that are highly selective and efficacious human ₃-adrenoceptor agonists.