Probing Adenosine and P2 Receptors: Design of Novel Purines and Nonpurines as Selective Ligands

Abstract

Recent approaches to the design of selective agonists and antagonists at adenosine (AR) and P2 receptors include both modifying known receptor ligands and searching for structurally diverse antagonists. The ribose-like moiety of nucleoside/nucleotide derivatives was rigidified with a methanocarba (mc) modification, to constrain the ring in a conformation that was favored in binding to ARs or P2Y receptors. (N)-mc analogs of various N⁶-substituted adenosine derivatives, including cyclopentyl and 3-iodobenzyl, in which the parent compounds are potent agonists at either A₁ or A₃ARs, respectively, retained high receptor affinity and selectivity. For nucleotides acting as P2Y₁ receptor antagonists, the (N)-mc analog MRS 2279 ((1R,2S,4S,5S)-1-[(phosphato)methyl]-4-(2-chloro-6-methylaminopurin-9-yl) bicyclo [3.1.0]-hexane-2-phosphate) proved to be a selective antagonist, with an IC₅₀ of 52 nM. Other ribose substitutions possible in P2Y₁ receptor antagonists were 4- and 6-membered rings and acyclic derivatives. High affinity for the A_2BAR was achieved through the formation of anilides and benzylamides of XCC (8-[4-[[[carboxy]methyl]oxy]phenyl]-1,3-dipropylxanthine). A p-cyanoaniline derivative (MRS 1754, K_i value 1.97 nM) was 205-, 255-, and 289-fold selective for the human A_2BARs vs. human A₁/A_2A/A₃ ARs, respectively. A template approach based on the pyridine family, i.e., 1,4-dihydropyridine nucleus and the corresponding 3,5-diacylpyridines, was used for the design of novel adenosine antagonists. The pyridine derivative MRS 1523 (5-propyl-2-ethyl-4-propyl-3-(ethylsulfanylcarbonyl)-6-phenylpyridine-5-carboxylate) was shown to be a selective antagonist at the rat A₃AR as well as the human A₃AR. Chemical libraries were screened computationally and using binding assays to identify novel AR antagonists. Molecular modeling of ARs and P2Y receptors provided hypotheses for ligand docking. Drug Dev. Res. 52:178–186, 2001.

INTRODUCTION

Modulation of adenosine (P1) and nucleotide (P2) receptors by selective agonists and antagonists has the potential for the treatment of inflammation and diseases of the cardiovascular and central nervous systems [Fredholm et al., 1997]. Adenosine is released in large amounts during myocardial ischemia and is capable of exerting potent cardioprotective effects in the heart [Strickler et al., 1996]. Thus, a synthetic adenosine receptor (AR) agonist, selective for either the A₁ or A₃ subtype, might be beneficial to the survival of the ischemic heart [Liang and Jacobson, 1998]. In the asthmatic lung, however, adenosine acts as an irritant and bronchoconstrictor, and an antagonist, possibly one having selectivity for the A_2B subtype, is highly desirable [Auchampach et al., 1998; Feoktistov et al., 1998].

In comparison to the ARs, much less is known about the effects of P2 receptors. While selective agonists and antagonists are available for most of the ARs, a current challenge to purine medicinal chemists is to design such agents for the ca. 14 subtypes of P2 receptors recently cloned [Burnstock and King, 1996]. The therapeutic potential of selective P2 receptor ligands is already evident. For example, in platelets there exist three subtypes of P2 receptors and these are involved in the process of aggregation leading to thrombus formation. A selective antagonist of the platelet P2Y₁ receptor might be useful as an antithrombotic agent [Fagura et al., 1998; Hechler et al., 1998; Jin et al., 1998].

The Molecular Recognition Section of the National Institute of Diabetes and Digestive and Kidney Diseases, NIH, is developing novel ligands for both ARs and P2 receptors for application to new therapeutic modalities. Recent approaches to analog design utilized in these investigations include: rigidifying the ribose-like moiety of nucleoside/nucleotide derivatives, to constrain the ring in a conformation that provides favorable affinity and/or selectivity at ARs and P2 receptors; modifying known receptor antagonists using a template approach based on the pyridine family for the design of novel adenosine antagonists; and screening chemical libraries in conjunction with molecular modeling.

RECENTLY INTRODUCED AR AGONISTS AND ANTAGONISTS

Conformationally Constrained Adenosine Derivatives As Agonists

Conformational constraint of the ribose-like ring of purine nucleosides and nucleotides can lead to enhanced receptor affinity and/or selectivity. The ribose ring of adenosine and ATP can adopt a range of conformations as described by a “pseudorotational cycle” [Marquez et al., 1996], while the ARs and P2 receptors likely prefer specific conformations of the ring.

In collaboration with Dr. Victor Marquez of the National Cancer Institute, we recently reported that a carbocyclic (bicyclic) modification of the ribose moiety, which incorporates ring constraints, is a general approach to the design of AR agonists having favorable pharmacodynamic properties [Jacobson et al., 2000]. While simple carbocyclic substitution of adenosine derivatives often diminishes potency at ARs, methanocarba-adenosine analogs have preserved or enhanced pharmacological properties as agonists at two AR subtypes. In such methanocarba analogs a fused cyclopropane moiety constrains the pseudosugar (cyclopentyl) ring of the nucleoside to either a Northern (N)- or Southern (S)- conformation (Fig. 1), as defined in the pseudorotational cycle. Such analogs help to define the role of sugar puckering in stabilizing the active AR-bound conformation, and thereby allow identification of a favored isomer.

Fig. 1.

Examples of methanocarba adenosine analogs evaluated at ARs [Jacobson et al., 2000]. The two structural variations shown for simple adenosine analogs (A) restrict the ring pucker, i.e., hold the ribose-like ring (pseudosugar) in either Northern (N) or Southern (S) envelope conformations. The (N)-conformation, favored at the A₁ and A₃ ARs, was then applied to more selective analogs (B), containing cyclopentyl or 3-iodobenzyl substitution. A = adenine.

In binding assays at A₁, A_2A, and A₃ARs, (N)-methanocarba-adenosine had a higher affinity than the (S)-analog, particularly at the human A₃AR ((N)-/(S)-affinity ratio of 150). (N)-Methanocarba analogs of various N⁶-substituted adenosine derivatives, including cyclopentyl and 3-iodobenzyl, in which the parent compounds are potent agonists at either A₁ or A₃ARs, respectively, were synthesized (Fig. 1). The N⁶-cyclopentyl derivatives were A₁AR-selective and maintained high efficacy at recombinant human, but not rat brain A₁ARs, as indicated in a functional assay based on stimulation of binding of [³⁵S]GTP-γ-S. The 2-chloro analog, MRS 1761, was a full, selective agonist at the human A₁AR. The (N)-methanocarba-N⁶-(3-iodobenzyl)adenosine and its 2-chloro derivative, MRS 1760, had equilibrium inhibition constant (K_i) values of 4.1 and 2.2 nM at A₃ARs, respectively, and were highly selective partial agonists. Partial agonism combined with high functional potency at A₃ARs (EC₅₀ < 1 nM) may lead to tissue selectivity. In conclusion, the introduction of an N-methanocarba modification of the ribose ring of nucleoside ligands represents a fruitful approach to the design of agonists acting at A₁ and A₃ARs. The relevance of this approach for P2Y receptors will be discussed below.

Purines as A_2BAR Antagonists

Until recently no selective ligands for the A_2B subtype were known, among either xanthines or non-xanthines. Thus, it has been difficult to characterize the pharmacological actions of this AR subtype using only nonselective agonists and antagonists.

The alkylxanthine theophylline is a weak, nonselective AR antagonist, used therapeutically for the treatment of asthma. The mechanism of action of theophylline in asthma has been controversial for decades [Feoktistov et al., 1998] and its use has been associated with unpleasant side effects, such as insomnia and diuresis. The development of a theophylline-like drug with reduced side effects is desirable.

At therapeutically relevant doses, theophylline and caffeine, its closely related analog, blocked activation of ARs by endogenous adenosine. Only recently did antagonism of the A_2B subtype become a plausible mechanism for the antiasthmatic action of xanthines [Auchampach et al., 1998]. The A_2BAR is expressed in some mast cells, such as canine mastocytoma cells and HMC-1 cells, in which they appear to be responsible for triggering acute Ca²⁺ mobilization and degranulation.

There is evidence that both the A_2BAR and A₃AR may play a role in asthma [Fozard and Hannon, 1999; Salvatore et al., 2000]. The A₃AR mediates the degranulation of rat RBL mast-like cells and is present in high density in human blood eosinophils. The availability of antagonists selective for the A_2BAR shall provide an opportunity to explore the importance of these two AR subtypes in asthma and other inflammatory diseases.

In collaboration with Dr. Joel Linden of the University of Virginia, we chose xanthines as suitable leads for developing selective A_2BAR antagonists [Kim et al., 1999, 2000] based on the high affinity of their analogs at A_2B and other AR subtypes. For example, a xanthine synthesized in our laboratory in 1984, XAC (8-[4-[[[[(2-aminoethyl)amino]carbonyl]methyl]oxy]phenyl]-1,3-dipropyl-xanthine), although nonselective, has a K_i value of 12.3 nM in binding to the human A_2BAR [Ji and Jacobson, 1999]. We therefore probed the structure–activity relationships (SAR) of both known and novel xanthine derivatives at the four subtypes of ARs, A₁/A_2A/A_2B/A₃. An 8-phenyl group was associated with increased affinity at the A_2BAR. The 8-phenyl modification of theophylline produced a 22-fold enhancement of binding affinity at the A_2BAR.

Leads have been identified among derivatives of 8-[4-[[[carboxy]methyl]oxy]phenyl]-1,3-dipropylxanthine (XCC, xanthine carboxylic congener) for achieving moderate A_2BAR selectivity. New derivatives, aryl, alkyl, and aralkyl amides, of XCC, were synthesized and their affinities compared in radioligand binding assays at the four subtypes of ARs. High affinity for the A_2BAR was achieved through the formation of anilides and benzylamides of XCC [Kim et al., 2000b]. Simple anilides, particularly those substituted in the p-position with electron-with-drawing groups, such as nitro, cyano, and acetyl, bound to the human A_2BAR with K_i values in the range of 1–3 nM. An unsubstituted anilide had a K_i value at A_2BAR of 1.48 nM, but was only moderately selective vs. human A₁/A_2A ARs and nonselective vs. the rat A₁AR. Substitution of the xanthines at the p-position resulted in modulation of the affinity at A₁/A_2A/A₃, but not at the A_2BAR (Fig. 2), at which affinity was generally high. A p-cyanoaniline derivative (MRS 1754, K_i value 1.97 nM) and the corresponding p-aminoacetophenone derivative (K_i value 1.39 nM) were highly potent and selective A_2BAR antagonists. MRS 1754 was 205, 255-, and 289-fold selective for the human A_2BAR vs. human A₁/A_2A/A₃ ARs, respectively, while appearing less selective in comparison of human A_2BAR with rat A₁/A_2A ARs. Affinity at the rat A_2BAR has not been thoroughly studied. Substitution of the p-carboxymethyloxy group of XCC and its amides with an acrylic acid group decreased affinity at the A_2BAR and increased affinity at the A₁AR.

Fig. 2.

Effect of ring substitution at the p-position on the binding affinity of anilides of the xanthine carboxylic congener (XCC) as adenosine receptor antagonists. The p-cyano analog is MRS 1754 [Kim et al., 2000b]. K_i values at the following human ARs are shown: A₁(■), A_2A (✦), A_2B (●), and A₃ (▲).

Xanthine derivatives displaying selectivity for the A_2B subtype in binding assays had comparable antagonistic effects in functional assays [Kim et al., 2000b]. Selective A_2BAR antagonists inhibited NECA-stimulated calcium mobilization in HEK cells expressing the human A_2BAR. At a concentration of 100 nM, MRS 1754 completely inhibited the rise in intracellular calcium elicited by 10 nM NECA. High-affinity compounds, such MRS 1754, may be the first selective pharmacological probes needed to investigate the physiological role of this receptor subtype.

This series of xanthines also provided the first high-affinity, selective radioligand for the A_2BAR. [³H]MRS 1754 (150 Ci/mmol) was synthesized and found to bind to a single class of receptors in HEK-293 cells expressing the human A_2BAR. This binding displayed pharmacological characteristics of the A_2BAR [Ji et al., 2000] and with a K_D value of 1.13 ± 0.12 nM. Binding to other AR subtypes was not significant.

Known Pharmacophores vs. Molecular Diversity

Attempts to identify leads for selective xanthine-based antagonists of the A₃AR have not yet been productive, due to the relatively low affinity of xanthines, the classical antagonists of A₁, A_2A, and A_2B subtypes, at the A₃AR. At human, dog, and sheep A₃ARs, certain xanthines display moderate affinity (K_i < 1 μM); however, none have been found to be highly selective.

Promising leads for A₃AR antagonists have appeared among nonxanthine heterocyclic compounds. For example, the triazoloquinazoline CGS 15943 (9-chloro-2-(2-furanyl)[1,2,4]triazolo[1,5-c]quinazolin-5-amine), bound nonselectively to human ARs (almost equipotent at all four subtypes, with K_i values of ~1–10 nM), and thus served as a lead for antagonists of the less well-characterized subtypes. A N⁵-phenylacetyl derivative, MRS 1220, was a moderately A₃ selective antagonist with very high affinity (K_i 0.65 nM) at the human A₃AR [Kim et al., 1998], but was of limited use in rat and other species due to reduced A₃AR affinity in those species. Baraldi and coworkers [Varani et al., 2000] recently introduced MRE-3008F20 (5-N-(4-methoxyphenylcarbamoyl)-amino-8-propyl-2-(2-furyl)pyrazolo[4,3-e]-1,2,4-triazolo-[1,5-c]pyrimidine), a nonxanthine heterocycle in a similar chemical class, as a potent and selective antagonist of the human A₃AR.

Identification of novel leads — screening of libraries

Since xanthines and other known AR antagonists provide only limited leads for A₃AR selective antagonists, an alternate strategy to identify selective antagonists is to exploit molecular diversity through screening chemical libraries.

Novel structures have been identified through library screening in both industrial and academic laboratories [Jacobson, 1998]. Among the first high-affinity, A₃ selective antagonists to be defined by screening of chemically diverse compound collections were a triazolonaphthyridine (L-249313), a thiazolopyrimidine (L-268605), and pyridylisoquinolines (e.g., VUF 8504) [van Muijlwijk et al., 1998]. Other novel structures of moderate affinity and selectivity at the human A₃AR were recently reported [Webb et al., 2000].

We assembled a small, biased library of pure substances containing many known pharmaceutical agents and structures that resemble purines in one or more features. The primary screen at the human A₃AR consisted of single point binding assays at a fixed concentration (10 μM) using the radioligand ¹²⁵I-AB-MECA. Once we discovered structural principles of recognition by the A₃AR in these novel antagonist classes, we subjected the leads to structural optimization through iterative chemical synthesis and pharmacological testing.

Naturally occurring phenolic derivatives, e.g., flavones and flavonols, provided a structural lead for A₃AR antagonists. The affinity of common phytochemicals at ARs suggested that a wide range of natural substances in the human diet may potentially antagonize the effects of endogenous adenosine, including those mediated via the A₃ subtype. The flavonoid class was chemically optimized in the form of MRS 1067 (3,6-dichloro-2′-(isopropoxy)-4′-methylflavone), which was 200-fold selective for human A₃- vs. A₁-ARs [Jacobson et al., 2001].

Optimization of leads through a “multidimensional” template approach — the pyridine family

The family of substituted pyridines and 1,4-dihydropyridines (DHPs) has given rise to a large number of novel AR antagonists. DHPs are privileged structures in medicinal chemistry, i.e., they display low-affinity binding at a variety of receptor and ion channel sites [Triggle, 1985]. Varying substituents on the DHP template can have dramatic effects on their interactions with these receptors. The corresponding, oxidized pyridines have provided a separate library of A₃AR antagonists, which have SAR in some ways similar to, but overall distinct from the SAR of DHPs.

Dihydropyridines

DHPs, known as potent blockers of L-type calcium channels [Coburn et al., 1988] and used widely in treating coronary heart disease, were found to bind appreciably to human adenosine A₃ARs. Common DHP drugs typically bound either nonselectively (for example, nifedipine, with a K_i value at hA₃AR of 8.3 μM) or in some cases with selectivity for the A₃ vs. other AR subtypes (for example, S-niguldipine, with a K_i value at hA₃AR of 2.8 μM).

We used the 1,4-DHP nucleus as a template for probing structure–activity relationships (SAR) at the subtypes of ARs (Fig 3). We observed that at adenosine receptors the affinity of many DHPs, even commercial L-type calcium channel blockers, was in the μmolar range at the human A₃AR, and we optimized this lead in the design of more selective antagonists. The essential modifications leading to A₃ selectivity and preventing binding to L-type calcium channels and other sites were: a phenyl ring at the 6-position and either a styryl or phenylethynyl group (e.g., second row of Fig. 3) at the 4-position. Such DHPs are similarly inactive at human A_2B receptors [Ji and Jacobson, 1999]. For example, a trisubstituted 1,4-dihydro-6-phenylpyridine analog, 3-ethyl 5-benzyl 2-methyl-6-phenyl-4-phenylethynyl-1,4-(±)-dihydropyridine-3,5-dicarboxylate (MRS 1191), inhibited radioligand binding at the human A₃AR with a K_i value of 31 nM, while the same derivative was nearly inactive in binding at A₁ and A_2AARs (i.e., >1,300-fold selective). Even in rat tissue, MRS 1191 was pharmacologically selective in various paradigms, e.g., it bound with 28-fold higher affinity for A₃ vs. A₁ARs. In an electrophysiological study [Dunwiddie et al., 1997], at a 10 μM concentration MRS 1191 antagonized only the A₃ subtype in the CA1 region of the hippocampus. Furthermore, MRS 1191 antagonized the effects of the A₃ receptor selective agonist IB-MECA on inhibition of adenylate cyclase via recombinant human (K_B values of 92 nM) or rat A₃ARs [Jacobson et al., 1997]. In chick ventricular myocyte cultures, MRS 1191 antagonized the antiischemic effects of Cl-IB-MECA [Liang and Jacobson, 1998]. Thus, DHP derivatives, such as MRS 1191, appear to be useful as A₃AR antagonists across species, although there is still a need for high-affinity antagonists of rat A₃ARs.

Fig. 3.

Use of 1,4-dihydropyridines as a molecular template for antagonists of the human A₃AR. Representative members of a library of DHP derivatives and their affinity (K_i in μM) in binding to the human A₃AR are shown [Jiang et al., 1998].

The structure–activity relationships of analogs of MRS 1191, containing both subtle and drastic structural changes at various positions of the DHP ring (its 3- and 5-acyl substituents, the 4- and 6-aryl/alkyl substituents, and the 2-methyl group), were systematically investigated. Substitutions of a 5-benzyl ester group provided the greatest versatility for achieving A₃ receptor selectivity of >30,000-fold and nanomolar potency. Affinity and selectivity for the human A₃AR within this series was optimal in MRS 1334, 3-ethyl 5-(4-nitrobenzyl) 2-methyl-6-phenyl-4-phenylethynyl-1,4-(±)-dihydropyridine-3,5-dicarboxylate, which had a K_i value of 2.7 nM.

While the stereospecificity of the binding of chiral ligands is a general property of receptors, the enantiomeric ratios of affinities may vary widely. At L-type calcium channels, the potency ratios between (4-R)- and (4-S)-DHPs are generally 10–100 fold, with the (4-S)-enantiomer being more potent. For most enantiomeric pairs of calcium channel ligands, the key factor determining stereospecificity is a bulky ester substituent at the 5-position. Initially, the AR affinity of several pairs of enantiomers of common 1,4-DHPs, mainly 4-aryl derivatives, was compared. DHPs bearing 2,6-dimethyl groups (i.e., the antagonist niguldipine and the antagonist/agonist pair BayK 8644) demonstrated a slight preference for the (4-R)-enantiomers (2- and 8-fold, respectively, at the human A₃AR). In a subsequent study, the effects of the 6-phenyl substituent of 1,4-dihydropyridines on stereospecificity of binding to ARs was studied [Jiang et al., 1999].

Since racemic 6-phenyl-4-phenylethynyl-1,4-dihydropyridine derivatives displayed exceptionally high A₃AR selectivity, compounds in this series were chemically resolved in order to study the biological properties of pure stereoisomers. Such a method for resolving the enantiomers at the C4 position consisted of either the selective crystallization or HPLC purification of diastereomeric ester derivatives. The chiral ester moieties, derivatives of 2,2-dimethyl-1,3-dioxolane-4-methanol and 2,3-O-isopropylidene-D-threitol, could have been selectively replaced with an ethyl moiety in two steps, i.e., deprotection of a diol followed by transesterification in ethanol. Binding studies with the pure enantiomers of MRS 1191 indicated a 35-fold stereoselectivity for the (4-S)-isomer at the human A₃AR [Jiang et al., 1999].

Pyridines

We explored SARs for the oxidized forms of the 1,4-DHPs, i.e., 3,5-diacylpyridines, as highly selective antagonists for the human A₃AR (Fig. 4) [Li et al., 1999a]. Moderate selectivity for the rat A₃AR was also present. As for the DHPs, a 6-aryl substituent present in the oxidized pyridines favored A₃AR selectivity. However, in contrast to DHPs, at 4-position and 5-ester position only small alkyl groups such as propyl were tolerated (Fig. 4). Furthermore, a 3-thioester enhanced A₃AR selectivity. Fluoro-, hydroxy-, and other polar groups were introduced in this series, which otherwise had a highly hydrophobic character, and affinity as high as 4 nM was achieved. Sterically constrained (bicyclic) analogs were designed and prepared for purposes of conformational analysis, demonstrating a preference for the 3-carbonyl group in the “down” orientation. New pyridine derivatives with higher affinity and selectivity for the A₃AR and potentially improved pharmacodynamic properties were obtained. The pyridine derivative MRS 1523 (5-propyl-2-ethyl-4-propyl-3-(ethylsulfanylcarbonyl)-6-phenylpyridine-5-carboxylate, Fig. 4) was selective at the rat A₃ (K_i 113 nM) as well as the human A₃AR (K_i 18.9 nM).

Fig. 4.

Use of 3,5-diacylpyridine derivatives, prepared by oxidation of the corresponding DHPs, as a molecular template for antagonists of the human A₃AR. Representative members of a library of such pyridine derivatives and their affinity (K_i in μM) in binding to the human A₃AR are shown [Li et al., 1999].

Pyridinium salts — activated from prodrugs through oxidation

We further extended the pyridine family of templates through quaternization and found that 3,5-diacyl-1,2,4-trialkyl-6-phenylpyridinium derivatives (Fig. 5) constituted a novel class of selective A₃AR antagonists of moderate affinity [Xie et al., 1999]. The SARs of this class of antagonists, incorporating the 3-thioester, were explored. The pyridinium salts were generated through oxidation of the corresponding 3,5-diacyl-1,2,4-trialkyl-6-phenyl-1,4-dihydropyridine derivative, with iodine or in the presence of rat brain homogenates. A 6-cyclopentyl analog was shown to increase affinity at the human A₃AR upon oxidation from the 1-methyl-1,4-dihydropyridine analog (Fig. 5) to the corresponding pyridinium derivative, MRS 1651 (K_i 695 nM), suggesting a prodrug scheme. In inflamed tissues, oxidative enzymes would be upregulated, thus the generation of a more potent antagonist from the prodrug might occur in a site-specific manner. The most potent analog in this group was 2,4-diethyl-1-methyl-3-(ethylsulfanylcarbonyl)-5-ethyloxycarbonyl-6-phenylpyridinium iodide (MRS 1649), which had K_i values of 219 nM at the recombinant human A₃AR, and >10 μM at rat brain A₁ and A_2A ARs, and at the recombinant human A_2BAR. Homologation of the pyridinium derivatives to N-ethyl and N-propyl at the 1-position caused a progressive reduction in the affinity at the A₃AR. Modifications of the alkyl groups at the 2-, 3-, 4-, and 5-positions failed to improve potency in binding at the A₃AR. The pyridinium antagonists were not as potent as other recently reported, selective A₃AR antagonists; however, they displayed uniquely high water-solubility (43 mM for MRS 1649). MRS 1649 antagonized the inhibition of adenylate cyclase elicited by IB-MECA in CHO cells expressing the human A₃AR, with a K_B value of 399 nM, and did not act as an agonist, demonstrating that the pyridinium salts were pure antagonists.

Fig. 5.

A prodrug scheme for the oxidative generation of pyridinium salts which act as antagonists of the human A₃AR from the corresponding 1-alkyldihydropyridine derivatives. Such antagonists are highly water-soluble in contrast to most other classes of A₃AR antagonists, which are hydrophobic [Xie et al., 1999].

Pyrans

DHP derivatives were synthesized through standard three-component condensation/oxidation reactions, which permitted versatile ring substitution at five positions, i.e., the central ring served as a molecular scaffold for structurally diverse substituents. We extended this template approach from the DHP series to a parallel chemical library of stable pyran derivatives [Li et al., 1999b], in which the ring NH was replaced by O, similarly derived from a stepwise reaction of three components. Since the orientation of substituent groups may be conformationally similar to the 1,4-DHPs, a direct comparison between the SARs of key derivatives in binding to ARs was carried out. The most potent ligands in this group in binding to the human A₃AR were 6-methyl and 6-phenyl analogs of 3,5-diethyl 2-methyl-4-phenyl-4H-pyran-3,5-dicarboxylate, which had K_i values of 381 and 583 nM, respectively. These two derivatives were selective for the human A₃AR vs. rat brain A₁ receptors by 57-fold and 24-fold, respectively. These derivatives were inactive in binding at the rat brain A_2AAR, and at the recombinant human A_2BAR displayed K_i values of 17.3 and 23.2 μM, respectively. The selectivity, but not affinity, of the pyran derivatives in binding to the A₃AR was generally enhanced vs. the corresponding DHP derivatives.

Unified pharmacophore-based screening

In collaboration with Dr. Thomas Webb of ChemBridge Corp., we set out to use information derived from known A₃AR antagonists to identify new leads for this receptor family and to explore both the pharmacophore relationships within the receptor family and the utility of pharmacophore database queries for the discovery of new leads [Webb et al., 2000]. The ability to find new structural series from pharmacophore information for an existing bioactive structural series would be advantageous in the development of new drug leads. An unexpected added benefit of our approach was the discovery of new selective leads for related AR subtypes, A₁, A_2A, and A_2B.

The general strategy that we chose used the structures of known potent antagonists to produce pharmacophore queries generated using the modeling program Chem-X, and then used these pharmacophores to search a Chem-X pharmacophore database. We then utilized in vitro receptor assays to determine potency and selectivity within the receptor family. Since our primary goals were the simultaneous discovery of both useful search methods and new active compounds, we required a diverse set of compounds that had not necessarily been previously evaluated for specific biological activity. A database of this type was available to us, ChemBridge Corporation’s CNS-SetTM, a collection of pharmacophore diverse compounds with defined properties (calculated logP 0–5, MW < 500, etc.). The pharmacophores were generated from approximately 19,000 structures in the initial selection for the CNS-SetTM collection, which resulted in the acquisition of ~9,600 compounds. From 186 compounds selected computationally and screened in receptor binding, 26 bound to the human A₃AR, while additional structures bound to the three other AR subtypes.

NOVEL P2Y RECEPTOR AGONISTS AND ANTAGONISTS

The P2Y₁ receptor is a G-protein-coupled receptor (GPCR) that stimulates phospholipase C (PLC) in response to adenine nucleotides and is present in the heart, smooth muscles, prostate, ovary, brain, and platelets. An antagonist may be useful as an antithrombotic agent.

Ribose-Modified Nucleotide Analogs

In collaboration with Drs. José L. Boyer and T.K. Harden and the University of North Carolina, we modified the native P2Y₁ receptor ligands, leading to a series of deoxyadenosine 3′,5′-bisphosphate derivatives that acted as selective P2Y₁ antagonists, or in some cases with small structural changes, agonists, or partial agonists [Camaioni et al., 1998; Nandanan et al., 1999, 2000]. Using molecular modeling of the receptor binding site we predicted that there were no essential H-bonds formed with the ribose moiety, and we therefore substituted this moiety with carbocylics, smaller and larger rings, conformationally constrained rings, and acyclics, with retention of affinity for the receptor.

Figure 6 shows various types of modifications of nucleotide-based ligands in the development of P2Y₁ receptor antagonists. Initial findings by Boyer et al. [1997] indicated that adenosine bisphosphates (at either 3′,5′ or 2′,5′ positions) were antagonists or partial agonists at the P2Y₁ receptor. The 2-position accommodated Cl or thioethers, while the N⁶-position was limited to Me or Et. A 2′-substitution (OH or ether) increased agonist efficacy over 2′-H. The analog MRS 2179 (2′-deoxy-N⁶-methyladenosine-3′,5′-bisphosphate), a more potent and competitive antagonist, contained an intact 2′-deoxyri-bose ring with adenine modifications. The N⁶-methyl and 2-chloro modifications lead to lower agonist efficacy and higher affinity, respectively. MRS 2179 also antagonized rat P2X1 receptors in an electrophysiological model with an IC₅₀ of 1.1 μM [Brown et al., 2000].

Fig. 6.

Structural modifications of the adenosine moiety of P2Y₁ receptor antagonists [Nandanan et al., 1999, 2000; Kim et al., 2000a]. The endogenous agonists of P2 receptors are nucleotides; however, bisphosphate analogs which act as antagonists have been identified. IC₅₀ values at the turkey erythrocyte P2Y₁ receptor for antagonism of phospholipase C activation by 30 nM 2-methylthio-adenosine 5′-diphosphate are indicated.

Ring expansion of the ribose in the 2-Cl-N⁶-Me series (in the form of the anhydrohexitol derivative, MRS 2283, IC₅₀ 0.566 μM, at the turkey erythrocyte P2Y₁ receptor) and ring contraction (in the form of the cyclobutyl derivative, MRS 2264, IC₅₀ 0.805 μM, at turkey P2Y₁ receptor) further emphasized the flexibility of substitution for the ribose moiety in these antagonists [Nandanan et al., 2000]. An acyclic modification of the ribose ring preserved affinity at the P2Y₁ receptor. MRS 2286 (2-[2-(2-chloro-6-methylamino-purin-9-yl)-ethyl]-propane-1,3-bisoxy(diammonium-phosphate)) was an antagonist at turkey P2Y₁ receptor with an IC₅₀ value of 0.84 μM, and no agonist affinity was observed [Kim et al., 2000a]. Furthermore, the compound was inactive at P2X1 receptors [Brown et al., 2000].

Conformational constraints were built into nucleoside and nucleotide ligands using the methanocarba approach, i.e., fused cyclopropyl and cyclopentyl rings in place of the ribose moiety (see above). Rigid rings in the methanocarba series defined a preference for the Northern (N) conformation of ribose at the P2Y₁ receptor. MRS 2268, the (N)-methanocarba analog of 2′-deoxyadenosine-3′,5′-bisphosphate, was a potent P2Y₁ agonist (EC₅₀ = 155 nM), 86-fold more potent than the corresponding Southern (S) isomer, MRS 2266. However, the corresponding 2-Cl-N⁶-methyl-(N)-methanocarba analog, MRS 2279 ((1R,2S,4S,5S)-1-[(phosphato)methyl]-4-(2-chloro-6-methylaminopurin-9-yl) bicyclo [3.1.0]-hexane-2-phosphate tetraammonium salt), was a potent antagonist (IC₅₀ = 52 nM). In the (N)methanocarba analogs as with other ribose modifications, the presence of an N⁶-Me group in these bisphosphate analogs transformed either a partial or full agonist into a pure antagonist, and the 2-chloro modification enhanced affinity.

Receptor Modeling as a Tool in Ligand Development

The development of drugs that modulate GPCRs, which contain seven membrane-spanning helical domains (TMs), is well served by defining the putative binding site, using mutagenesis and molecular modeling in conjunction with organic synthesis of small molecules. Conformational considerations of both receptors and their ligands are important in structure-based drug design.

We have used alanine scanning mutagenesis of the human P2Y₁ receptor to show the importance of specific residues in molecular recognition. Molecular modeling of both ARs and P2Y receptors (using a rhodopsin template) and their docked ligands, based on mutagenesis results, have been carried out to interpret the findings and to suggest new ligands [Moro et al., 1998]. With simplified pharmacophores we are currently exploring the steric and electronic constraints of the receptor binding site and the structural basis of receptor activation.

Molecular modeling of cloned P2Y₁ receptor sequences has focused on both the seven membrane-spanning helical domains (TMs) and the three extracellular loops. The features of the putative binding site identified were consistent with both mutagenesis results and known ligand specificities. To obtain an energetically refined 3D structure of the complex, we introduced a new approach, a “cross-docking” procedure [Moro et al., 1998], which simulated the reorganization of the native receptor induced by the ligand. In order to ascertain which residues of the human P2Y₁ receptor were involved in ligand recognition, we mutated individual residues of both the TMs (3, 5, 6, and 7) and the extracellular loops (e.l. 2 and 3). A cluster of positively charged amino acids, Lys and Arg residues near the exofacial side of TMs 3 and 7, and to a lesser extent TM6, putatively coordinated the phosphate moieties of nucleotide agonists and antagonists [Hoffmann et al., 2000]. Two essential disulfide bridges in the extracellular domains were identified and several charged residues in the extracellular loops 2 (E209) and 3 (R287) were shown to be critical for receptor activation. This suggested that the role of the extracellular loops in ligand recognition was as important as that of the TMs. The presence of “meta-binding sites” in the P2Y₁ receptor, in which the nucleotide may bind to distal site(s) on its way to the principle TM binding site, were predicted using both mutagenesis and molecular modeling. These secondary binding sites may serve to guide the ligand in its approach to the TM binding site and reduce the energy barrier to complex formation.