2-Substitution of N⁶-Benzyladenosine-5′-uronamides Enhances Selectivity for A₃ Adenosine Receptors

Abstract

Adenosine derivatives bearing an N⁶-(3-iodobenzyl) group, reported to enhance the affinity of adenosine-5′-uronamide analogues as agonists at A₃ adenosine receptors (J. Med. Chem. 1994, 37, 636–646), were synthesized starting from methyl β-d-ribofuranoside in 10 steps. Binding affinities at A₁ and A_2a receptors in rat brain membranes and at cloned rat A₃ receptors from stably transfected CHO cells were compared. N⁶-(3-Iodobenzyl)adenosine was 2-fold selective for A₃ vs A₁ or A_2a receptors; thus it is the first monosubstituted adenosine analogue having any A₃ selectivity. The effects of 2-substitution in combination with modifications at the N⁶- and 5′-positions were explored. 2-Chloro-N⁶-(3-iodobenzyl)adenosine had a K_i value of 1.4 nM and moderate selectivity for A₃ receptors. 2-Chloro-N⁶-(3-iodobenzyl)adenosine-5′-N-methyluronamide, which displayed a K_i value of 0.33 nM, was selective for A₃ vs A₁ and A_2a receptors by 2500- and 1400-fold, respectively. It was 46,000-fold selective for A₃ receptors vs the Na⁺-independent adenosine transporter, as indicated in displacement of [³H]N⁶-(4-nitrobenzyl)-thioinosine binding in rat brain membranes. In a functional assay in CHO cells, it inhibited adenylate cyclase via rat A₃ receptors with an IC₅₀ of 67 nM. 2-(Methylthio)-N⁶-(3-iodobenzyl)-adenosine-5′-N-methyluronamide and 2-(methylamino)-N⁶-(3-iodobenzyl)adenosine-5′-N-methyluronamide were less potent, but nearly as selective for A₃ receptors. Thus, 2-substitution (both small and sterically bulky) is well-tolerated at A₃ receptors, and its A₃ affinity-enhancing effects are additive with effects of uronamides at the 5′-position and a 3-iodobenzyl group at the N⁶-position.

Introduction

The novel A₃ adenosine receptor¹ may be important in the regulation of CNS, cardiac, inflammatory, and reproductive functions. The expression of A₃ adenosine receptors in normal vs asthmatic lung tissue has been studied.² The A₃ adenosine receptor was cloned from rat brain and rat testes cDNA libraries.^1,3 Its activation stimulates phosphatidylinositol metabolism in antigen-exposed mast cells⁴ and inhibits adenylate cyclase in transfected CHO cells.¹ Activation of A₃ receptors enhances the release of inflammatory mediators from mast cells,^4,5 lowers blood pressure,⁶ and depresses locomotor activity.⁷. A cerebroprotective effect of chronic administration of an A₃ agonist has been discovered.⁸ The activation of A₃ receptors is also thought to be related to the cardioprotective preconditioning response following exposure to adenosine agonists.⁹

The structure-activity relationships of adenosine derivatives and xanthine derivatives at the rat A₃ versus A₁ and A_2a receptors have been explored.^10,11 The affinity of various ligands at sheep¹² and human¹³ A₃ receptors has been reported to be very different from rat. At rat A₃ receptors, most xanthines known to bind to A₁ and A₂ receptors do not act as antagonists.¹⁰

We recently reported new adenosine agonist derivatives of moderate A₃ selectivity.^7,10;11 The 5′-methyluronamide modification of adenosine and the N⁶-benzyl group, either alone or in combination, increases affinity in binding to A₃ receptors relative to A₁ and A_2a receptors.¹⁰ Optimization of substituent groups has led to the development of the highly potent A₃ agonist N⁶-(3-iodobenzyl)adenosine-5′-N-methyluronamide (IB-MECA, 1, Figure 1) which is 50-fold selective for A₃ vs either A₁ or A₂ receptors. A closely related, but less selective radioligand, [¹²⁵I]AB-MECA, 2, was developed for characterization of A₃ receptors and found to have a K_d value of 3.6 nM in binding to rat A₃ receptors in the RBL-2H3 mast cell line.¹⁴

Figure 1.

Structures of high-affinity A₃ receptor agonists.

In this study we have extended our previous development of A₃ selective agonists. By combining the two modifications at 5′- and N⁶-positions, which were found earlier to result in moderate selectivity, with a third site of modification, the 2-position, we have dramatically increased selectivity. This study presents the first compounds that combine very high potency and selectivity, which should make them very useful as pharmacological tools and potential therapeutic agents.

Results

There remains a need for the development of highly selective A₃ agonists. Although the new A₃ radioligand [¹²⁵I]AB-MECA, 2,¹⁴ is of nanomolar potency at A₃ receptors and more potent than the previously used [¹²⁵I]APNEA, 4,¹ it is not very selective for A₃ vs A₁ or A_2a receptors (Table 1). The presence of the 4-amino group of 2 decreases selectivity in comparison to moderately A₃ selective agonist, IB-MECA, 1.^7,11 The N⁶-derivative of adenosine, APNEA, 3, has been used recently in pharmacological studies⁶ to stimulate A₃ receptors, although it is actually A₁-selective. Until present, no monosubstituted adenosine derivatives have been reported to be selective for A₃ receptors.¹⁰ In our previous study of high-affinity 5′,N⁶-disubstituted adenosine derivatives,¹¹ only 50–70-fold selectivity for A₃ vs A₁ receptors had been achieved.

Table 1.

Affinities of 5′-Uronamide Derivatives in Radioligand Binding Assays at Rat Brain A₁, A_2a, and A₃ Receptors^a–^c

				Ki (nM)
compd	R1	R2	R3	A1a	A2ab	A3c	A1/A3	A2a/A3
1d	CH3NHCO	H	3-I-Bz	54	56	1.1	49	51
2d	CH3NHCO	H	3-I-4-NH2Bz	18	197	1.3	14	160
3	HOCH2	H	4-NH2Ph(CH2)2	14f	172 ± 50	116 ± 18	0.16	1.5
4	HOCH2	H	3-I-4-NH2Ph(CH2)2	2.1f		15.5g	0.14
5e	C2H5NHCO	NH(CH2)2-p-Ph-(CH2)2COOH	H	2600	15	584	4.4	0.026
6	C2H5NHCO	NH(CH2)2-p-Ph-(CH2)2CONH(CH2)2NH2	H	400	5.7	50 ± 24	8	0.11
7e	HOCH2	Cl	H	9.3	63	1890	0.0049	0.033
8d	HOCH2	H	cyclopentyl	0.59	462	240	0.0025	1.9
9d	HOCH2	Cl	cyclopentyl	0.6	950	237	0.0025	4.0
10	HOCH2	H	3-I-Bz	20.0 ± 8.5	17.5 ± 0.5	9.5 ± 1.4	2.1	1.8
11	HOCH2	Cl	3-I-Bz	18.5 ± 4.7	38.5 ± 2.0	1.41 ± 0.17	13	27
12	HOCH2	NH2	3-I-Bz	63.8 ± 15.1	117 ± 15	181 ± 30	0.35	0.65
13	CH3NHCO	Cl	3-I-Bz	820 ± 570	470 ± 365	0.33 ± 0.08	2500	1400
14	CH3NHCO	CH3NH	3-I-Bz	4890 ± 2580	4120 ± 210	3.12 ± 0.64	1600	1300
15	CH3NHCO	CH3S	3-I-Bz	2140 ± 100	3210 ± 1360	2.30 ± 0.96	930	1400

^aDisplacement of specific [³H]PIA binding, unless noted, in rat brain membranes expressed as K_i ± SEM in nM (n = 3–6).

^bDisplacement of specific [³H]CGS 21680 binding, unless noted, in rat striatal membranes, expressed as K_i ± SEM in nM (n = 3–6).

^cDisplacement of specific binding of N⁶-[[¹²⁵I]-4-amino-3-iodobenzyl]adenosine-5′-N-methyluronamide¹⁴ from membranes of CHO cells stably transfected with the rat A₃–cDNA, expressed as K_i ± SEM in nM (n = 3–7).

^dValues are from Gallo-Rodriguez et al.¹¹

^eValues are from van Galen et al.¹⁰ A₃ affinity measured by displacement of specific binding of [¹²⁵I]APNEA in membranes of CHO cells stably transfected with the rat A₃–cDNA.¹ K_i values at A₁ receptors are vs specific binding of [³H]–N ⁶–cyclohexyladenosine or [³H]R-PIA. K_i values at A_2a receptors are vs specific binding of [³H]NECA in the presence of 50 nM N ⁶-cyclopentyladenosine or vs specific binding of [³H]CGS 21680 in rat striatal membranes.

^fIC₅₀ values (nM) vs displacement of specific binding of [¹²⁵I]APNEA in rat brain membranes.²³

^gK_d value (nM) from saturation of binding of [¹²⁵I]APNEA in membranes of CHO cells stably transfected with the rat A₃–cDNA.¹

New adenosine analogues (compounds 10–15, Table 1) were synthesized according to Schemes 1 and 2 and characterized (Table 2) and tested in radioligand binding assays^14–16 for affinity at rat brain A₁, A_2a, and A₃ adenosine receptors. The compounds were assayed as follows: at A₁ receptors in rat cortical membranes using [³H]-N⁶-[(R)-phenylisopropyl]adenosine¹⁵; at A_2a receptors in rat striatal membranes using [³H]-CGS 21680¹⁶; at A₃ receptors using [¹²⁵I]AB-MECA, 2,¹⁴ in membranes of CHO cells stably transfected with cDNA for rat brain A₃ receptors.¹

Scheme 1^a.

^a Reagents: (a) TBDPSiCl, DMAP, DMF, room temperature; (b) Bz₂O, py; (c) n-Bu₄NF, THF; (d) RuO₂, NaIO₄, CHCl₃–CH₃CN–H₂O (2:2:3); (e) EDAC, DMAP, MeOH; (f) MeNH₂, THF, 75 °C; (g) BzCl, py–CH₂Cl₂; (h) Ac₂O, H₂SO₄, AcOH.

Scheme 2.

Table 2.

Characterization of Intermediates and 2-Substituted-N ⁶-(3-iodobenzyl)adenosine Derivatives

compd	mp (°C)	formula	analysis
10	172	C17H18N5O4I	C, H, N
11	foam	C17H17N5O4ClI•0.3MeOH	C, H, N
12	152–154	C17H19N6O4I•1.2MeOH	C, H, N
13	206–207	C18H18N6O4ClI•0.5MeOH	C, H, N
14	190	C19H23N7O4I	a
15	179	C19H21N6O4IS	a
17	syrup	C22H30O5Si	C, H
18	syrup	C36H38N7Si	C, H
19	syrup	C20H20N7	C, H, N
20	syrup	C20H18N8•0.63H2O	C, H
21	92.2–93.7	C21H20N8	C, H
22	syrup	C21H21NO7•0.5H2O	C, H, N
23a,b	foam	C22H21NO8•0.3H2O	C, H, N
24a	222–224	C12H19N5ClI	C, H, N
25	foam	C32H26N6O6ClI•1.0C6H14	C, H, N
26	foam	C38H29N5O7ClI0.2C6H14	C, H, N

^aHigh-resolution MS (m/z) measured in FAB⁺ mode. 14: calcd for C₁₉H₂₃N₇O₄I 540.0856, found 540.0867. 15: calcd for C₁₉H₂₁N₆O₄I₁S₁ 557.0468, found 557.0482.

Substitution at the 2-position is often associated with selectivity of adenosine agonists for A_2a vs A₁ receptors. For example, CGS 21680, 5, and APEC, 6, both having sterically bulky 2-substituents, were reported to be highly selective for A_2a receptors in models of adenylate cyclase.²² Compound 6 was more potent at all three adenosine receptor subtypes than 5, and both compounds displayed a potency order of A_2a > A₃ > A₁, consistent with the findings of van Galen et al.¹⁰ that 2-substitution of adenosine is well-tolerated in binding to A₃ receptors. Among monosubstituted derivatives of adenosine, 2-(phenylamino)- and 2-chloroadenosine, 7, have K_i values for inhibition of binding of [¹²⁵I]APNEA (N⁶-[2-(4-aminophenyl)ethyl]adenosine) at rat A₃ receptors of 4.4 and 1.9 µM, respectively.¹⁰ Substitution at the 2-position is also compatible with N⁶-substitution for affinity at A₃ receptors. For example, 2-chloro-N⁶-cyclopentyladenosine, 9 (Table 1), is nearly identical in its receptor binding profile to N⁶-cyclopentyladenosine, 8.

N⁶-(3-Iodobenzyl)adenosine, 10, was prepared from 6-chloropurine riboside and 3-iodobenzylamine hydrochloride in the presence of triethylamine in ethanol at 80 °C. Compound 10 was 2-fold selective for A₃ vs A₁ or A_2a receptors, making it is the first monosubstituted adenosine analogue with any selectivity for A₃ receptors. 2-Chloro-N⁶-(3-iodobenzyl)adenosine, 11, was 7-fold more potent than 10 at A₃ receptors and of moderate selectivity. 2-Amino substitution of adenosine analogues is also compatible with N⁶-substitution in A₃ receptor binding but is not as favorable as 2-chloro for potency and selectivity. For example, 2-amino-N⁶-(3-iodobenzyl)-adenosine, 12, was less potent than the 2-H analogue, 10, by factors of 3.2 (A₁ receptors), 6.7 (A_2a receptors), and 19 (A₃ receptors).

To evaluate the effects of triple substitution of adenosine, i.e. at 5′-, 2-, and N⁶-positions on the affinity at A₃ receptors, we developed a general synthetic strategy in which a 5′-uronamide sugar moiety (Scheme 1) was condensed with a purine moiety, such as a substituted adenosine derivative. The key sugar intermediate 23 was synthesized starting from methyl β-d-ribofuranoside (16), which was commercially available or could be synthesized¹⁷ from d-ribose, in eight steps. The primary alcohol of 16 was selectively protected with tert-butyldiphenylsilyl chloride¹⁸ to provide 17, and the remaining alcohols were followed with benzoyl protection to provide 18. Desilylation of 18 with TBAF/THF gave compound 19. The 5′-position of 19 was oxidized using ruthenium tetroxide¹⁹ to give compound 20 which was purified after methylation by silica gel column chromatography. The methylamide at 5-position was introduced by nucleophilic displacement of 21 with methylamine in THF and benzoyl reprotection of resulting 2,3-diol to give the sugar intermediate 23.

In order to synthesize N⁶-(3-iodobenzyl)-2-substituted-adenosine derivatives, it was necessary to prepare the corresponding adenine derivative (Schemes 2 and 3). 2,6-Dichloropurine reacted with 3-iodobenzylamine hydrochloride in the presence of triethylamine in ethanol at room temperature to provide N⁶-(3-iodobenzyl)-2-chloroadenine, 24a, which was silylated before coupling to give 24b. The glycosidic bond was formed upon treatment of the 1′-O-acetyl riboside derivative 23 with the 9-silylated adenine derivative 24b in the presence of TMSOTf as a Lewis acid catalyst (Scheme 2). Condensation of 22 with 24b produced ribose ring opened product. Benzoyl groups of 25 were deprotected with NH₃/MeOH to produce 13, which reacted with various nucleophiles such as methylamine/THF and sodium thiomethoxide/DME to yield compounds 14 and 15. The benzoyl groups of 26 (Scheme 3) were similarly deprotected to produce the riboside derivative 11.

Scheme 3^a.

^a Reagents: (a) 3-iodobenzylamine–HCl, triethylamine, EtOH; (b) HMDS, (NH₄)₂SO₄; (c) TMSOTf, ClCH₂CH₂Cl; (d) NH₃/MeOH.

The assignments of the anomeric structure of compounds 25 and 26 were performed based on the comparison of the coupling pattern of anomeric proton of compounds 11 and 13 with compound 10, which is of known anomeric structure being derived from 6-chloropurine riboside.

The combination of 2-substitution with the substituent groups of compound 1 resulted in very high potency and selectivity for A₃ receptors. The A₃ affinity of the 2-chloro analogue, 2-chloro-N⁶-(3-iodobenzyl)adenosine-5′-N-methyluronamide, 13, was 3-fold greater than for IB-MECA, 1. The affinity at A₁ and A_2a receptors was diminshed relative to 1, by 15- and 9-fold, respectively. Thus, selectivities of approximately 2500-fold vs A₁ receptors and 1400-fold vs A_2a receptors were achieved. 2-(Methylamino)-N⁶-(3-iodobenzyl)adenosine-5′-N-methyluronamide, 14, was less potent (K_i value 3 nM), but still highly selective for A₃ receptors. 2-(Methylthio)-N⁶-(3-iodobenzyl)adenosine-5′-N-methyluronamide, 15, was also highly selective for A₃ receptors.

The selectivity of several of the adenosine derivatives vs a nucleoside transporter previously characterized in brain²⁴ was probed. These experiments were carried out because of the structural similarity of the present adenosine derivatives to various 6-benzyl ethers or thioether derivatives of purine ribosides, known to be high affnity antagonists of adenosine uptake via this transporter.²⁴ The simple N⁶-benzyl derivative of adenosine was not selective for the receptors vs the adenosine transporter. In contrast, 2-chloro-N⁶-3-iodobenzyl)adenosine-5′-N-methyluronamide, 13, was 46 000-fold selective for A₃ receptors vs the Na⁺-independent adenosine transporter, as indicated in displacement of [³H]-S-(4-nitrobenzyl)-6-thioinosine binding in rat brain membranes. Thus, in this series of 2,6,5′-trisubstituted adenosine derivatives there was a high degree of selectivity for A₃ receptors vs potential antagonism of adenosine uptake.

The agonist properties of the selective ligands were also examined (Figure 2). In a functional assay using membranes from CHO cells stably transfected with rat A₃ receptors, compounds 1 and 13 inhibited adenylate cyclase with IC₅₀ values of 90.0 ± 22.5 and 66.8 ± 9.0 nM (n = 4), respectively. Both derivatives were full agonists, with a maximal 41% inhibition of forskolin-stimulated adenylate cyclase. These two derivatives were considerably more potent in the A₃ receptor functional assay than were either N⁶-benzylNECA (IC₅₀ of 1.61 µM) or NECA (IC₅₀ of 5.6 ± 1.9 µM, n = 3). These findings establish a rank order of potency similar to that observed in binding assays, but at higher concentrations. The K_i values for N⁶-benzylNECA and NECA at rat A₃ receptors in CHO cells vs [¹²⁵I]APNEA were 6.8 and 113 nM, respectively.¹⁰ Thus, the IC₅₀ values for these four agonists to inhibit adenylate cyclase were 50–240-fold higher than the respective K_i values at A₃ receptors.

Figure 2.

Inhibition of adenylate cyclase in membranes from CHO cell stably transfected with rat A₃ receptors. The assay was carried out as described in the Experimental Procedures in the presence of 1 µM forskolin. Each data point is shown as mean ± SEM for four to seven determinations. Adenosine derivatives were (number of separate experiments in parentheses): solid triangles, 1, N⁶-(3-iodobenzyl)adenosine-5′-N-methyluronamide (4); open circles, 13, 2-chloro-N⁶-(3-iodobenzyl)adenosine-5′-N-methyluronamide (4); solid circles, N⁶-benzylNECA (7); and open squares, NECA (3). IC₅₀ values were 1, 90.0 ± 22.5 nM; 13, 66.8 ± 9.0 nM; N⁶-benzylNECA, 1.61 µM; NECA, 5.6 ± 1.9 µM.

Discussion

In two earlier studies^10,11 we demonstrated that combined modification of adenosine at 5′- and at N⁶-positions with groups that enhanced A₃ potency resulted in moderate A₃ selectivity. We previously showed N⁶-benzyladenosine-5′-N-ethyluronamide (N⁶-benzyl-NECA) to be a full agonist in inhibiting adenylate cyclase via rat A₃ receptors.¹⁰ However, that derivative was only 1 order of magnitude selective for rat A₃ receptors vs either A₁ or A_2a receptors in binding assays. In this study we have introduced triple substitution of adenosine as a means of enhancing the degree of A₃ selectivity, and selectivity in binding assays of 3 orders of magnitude has now been achieved. 2-Chloro-N⁶-(3-iodobenzyl)adenosine-5′-N-methyluronamide, the most potent and selective agent in binding assays, was also shown to be a full agonist in the inhibition of adenylate cyclase, with an IC₅₀ of 67 nM. The agonist potency was also greater than that of other agonists, indicating a parallel between binding affinities and relative potencies in this functional assay. The agonist properties in another relevant functional assay, stimulation of A₃-mediated phosphoinositide metabolism, are currently being examined.

Adenosine agonists of high selectivity, such as 13, 14, and 15, are needed for defining the role of A₃ receptors in vivo. We have demonstrated that selective agonists may have therapeutic potential as cerebroprotective agents.^8,20 Recently Downey and colleagues⁹ have demonstrated the cardioprotective potential of A₃ receptor activation, based on use of APNEA coadministered with a xanthine antagonist that does not act at A₃ receptors. In this study we have shown that APNEA is 8-fold A₁ selective, and its pharmacological use is limited to such combination with antagonists of both A₁ and A_2a receptors. Clearly, the availablility of ligands such as 13 could be critical in pharmacological studies of A₃ receptors. A highly selective A₃ ligand would be useful as a radioligand, since the currently used high affinity ligand [¹²⁵I]AB-MECA, is not sufficiently selective for general application in tissue.¹⁴

It will be necessary to establish the selectivities of these novel A₃ agonists in different species, due to the unusually large species dependence in ligand affinity at this subtype, although differences appear to be more pronounced for antagonists than for agonists.^12,13,25 It is to be noted that 2-chloroadenosine is 17-fold less potent than NECA at rat A₃ receptors,¹⁰ whereas at sheep A₃ receptors 2-chloroadenosine is only 1.7-fold less potent than NECA.¹² Thus, since the most selective compound in the present series, 13, contains the 2-chloro substitution, it is likely that the selectivity will not be substantially diminshed in other species, such as sheep and human. We have shown a high degree of correlation in the relative affinities of adenosine derivatives at rat vs human A₃ receptors.²⁵

The selectivity of compound 1 for adenosine receptors vs other neurotransmitter/modulator receptors was shown.¹¹ In this study, we have shown a high degree of selectivity of the doubly-substituted derivative, 1, and the present triply-substituted adenosine derivatives for A₃ receptors vs the NBTI-sensitive adenosine uptake site. We have not tested the adenosine derivatives at the normally low-affinity A_2b receptor, but substitution at the 2-position of adenosine has been shown not to be well-tolerated at the mouse fibroblast A_2b receptor.²⁶

In conclusion, 2-substitution is well-tolerated at A₃ receptors, whether it be with a small group (e.g., 11) or a large group (e.g., 6). The potency-enhancing effects of 2-substituents appeared to follow the order: chloro > thioether > amine. The effects of 2-substitution to enhance A₃ affinity are also additive with effects of uronamides at the 5′-position and a 3-iodobenzyl group at the N⁶-position. The A₃ affinity-enhancing effect of a 2-chloro group was not additive with an N⁶-cyclopentyl group. The combination of most favorable modifications at three positions has led to very potent and highly selective agonist ligands, compounds 13–15.

Experimental Procedures

Chemistry

New compounds were characterized (and resonances assigned) by 300 MHz proton nuclear magnetic resonance mass spectroscopy using a Varian GEMINI-300 FT-NMR spectrometer. Unless noted, chemical shifts are expressed as ppm downfield from tetramethylsilane. Synthetic intermediates were characterized by chemical ionization mass spectrometry (NH₃) and adenosine derivatives by fast atom bombardment mass spectrometry (positive ions in a noba or m-bullet matrix) on a JEOL SX102 mass spectrometer. In the EI mode accurate mass was determined using a VG7070F mass spectrometer. C, H, and N analyses were carried out by Atlantic Microlabs (Norcross, GA), and ±0.4% was acceptable. All adenosine derivatives were judged to be homogeneous using thin-layer chromatography (silica, 0.25 mm, glass backed, Alltech Assoc., Deerfield, IL) following final purification. 2-Chloroadenosine, NECA, CGS 21680, N⁶-cyclopentyladenosine, and 2-chloro-N⁶-cyclopentyladenosine were obtained from Research Biochemicals International (Natick, MA). IB-MECA and APEC were synthesized as reported.^11,27 APNEA and iodo-APNEA were the gift of Prof. Ray A. Olsson (University of South Florida, Tampa, FL).

N⁶-(3-Iodobenzyl)-9-β-d-ribofuranosyladenine (10)

A mixture of 6-chloropurine riboside (purchased from Aldrich Chemical Co., 100 mg, 0.35 mmol), triethylamine (0.146 mL, 1.05 mmol), and 3-iodobenzylamine hydrochloride (103 mg, 0.38 mmol) in ethanol (2 mL) was heated for 18 h at 85 °C in a sealed bottle. After the reaction mixture was concentrated to dryness, the residue was purified by silica gel column chromatography (CHCl₃–MeOH, 10:1) to give compound 10 (148 mg, 88%) as a colorless solid: ¹H NMR (DMSO-d₆) δ 3.54 (m, 1 H, H-5′a), 3.67 (m, 1 H, H-5′b), 3.96 (d, J = 3.3 Hz, 1 H, H-4′), 4.14 (m, 1 H, H-3′), 4.60 (m, 1 H, H-2′), 4.66 (br s, 2 H, CH₂), 5.16 (d, J = 4.4 Hz, 1 H, exchangeable with D₂O, 3′-OH), 5.34 (br s, 1 H, exchangeable with D₂O, 5′-OH), 5.43 (d, J = 6.1 Hz, 1 H, exchangeable with D₂O, 2′-OH), 5.89 (d, J = 6.0 Hz, 1 H, H-1′), 7.11 (pseudo t, J = 8.0 and 7.8 Hz, 1 H, H-5″), 7.36 (d, J = 7.6 Hz, 1 H, H-4″ or -6″), 7.58 (d, J = 7.8 Hz, 1 H, H-4″ or -6″), 7.72 (s, 1 H, H-2″), 8.21 (s, 1 H, H-2 or -8), 8.40 (s, 1 H, H-2 or -8), 8.48 (br s, 1 H, exchangeable with D₂O, N⁶-H).

2-Chloro-N⁶-(3-iodobenzyl)-9-β-d-ribofuranosyladenine (11)

A mixture of compound 26 (760 mg, 0.916 mmol) and NH₃/MeOH (15 mL) was stirred for 66.5 h at room temperature. After the reaction mixture was concentrated to dryness, the residue was purified by silica gel column chromatography (CHCl₃–MeOH, 20:1) to yield compound 11 (445 mg, 94%) as a foam: ¹H NMR (DMSO-d₆) δ 3.55 (m, 1 H, H-5′a), 3.65 (m, 1 H, H-5′b), 3.94 (d, J = 3.6 Hz, 1 H, H-4′), 4.12 (m, 1 H, H-3′), 4.51 (q, J = 5.5 Hz, 1 H, H-2′), 4.60 (br d, J = 5.7 Hz, 2 H, CH₂), 5.04 (pseudo t, J = 5.7 and 5.5 Hz, 1 H, exchangeable with D₂O, 5′-OH), 5.19 (d, J = 4.9 Hz, 1 H, exchangeable with D₂O, OH), 5.47 (d, J = 6.0 Hz, 1 H, exchangeable with D₂O, OH), 5.83 (d, J = 5.5 Hz, 1 H, H-1′), 7.13 (pseudo t, J = 7.9 and 7.6 Hz, 1 H, H-5″), 7.36 (d, J = 7.5 Hz, 1 H, H-4″ or -6″), 7.60 (d, J = 7.9 Hz, 1 H, H-4″ or -6″), 7.74 (s, 1 H, H-2″), 8.43 (s, 1 H, H-8), 8.94 (br t, J = 6.0 Hz, 1 H, exchangeable with D₂O, NH).

2-Amino-N⁶-(3-iodobenzyl)-9-β-d-ribofuranosyladenine (12)

A mixture of 2-amino-6-chloropurine riboside (purchased from Aldrich Chemical Co., 80 mg, 0.26 mmol), 3-iodobenzylamine hydrochloride (71.5 mg, 0.265 mmol), and triethylamine (0.11 mL, 0.79 mmol) in ethanol (1.6 mL) was heated for 24 h at 80 °C. After the reaction mixture was concentrated to dryness the residue was purified by silica gel column chromatography (CHCl₃–MeOH, 20:1 → 10:1) to yield compound 12 (99 mg, 75%) as a colorless solid: ¹H NMR (DMSO-d₆) δ 3.52 (m, 1 H, H-5′a), 3.63 (m, 1 H, H-5′b), 3.89 (m, 1 H, H-4′), 4.10 (m, 1 H, H-3′), 4.50 (m, 1 H, H-2′), 4.60 (br s, 2 H, CH₂), 5.08 (d, J = 4.6 Hz, 1 H, exchangeable with D₂O, 3′-OH), 5.35 (m, 2 H, exchangeable with D₂O, 5′- and 2′-OH), 5.73 (d, J = 6.2 Hz, 1 H, H-1′), 5.83 (br s, 2 H, exchangeable with D₂O, NH₂), 7.11 (pseudo t, J = 7.9 and 7.8 Hz, 1 H, H-5″), 7.36 (d, J = 7.8 Hz, 1 H, H-4″ or -6″), 7.58 (d, J = 7.8 Hz, 1 H, H-4″ or -6″), 7.70 (s, 1 H, H-2″), 7.94 (s, 1 H, H-8).

2-Chloro-N⁶-(3-iodobenzyl)-9-[5-(methylcarbmoyl)-β-d-ribofuranosyl]adenine (13)

A mixture of compound 25 (27 mg, 0.036 mmol) and NH₃/MeOH (15 mL) was stirred for 16 h at room temperature. After rotary evaporation of the volatiles, the residue was purified by silica gel column chromatography (CHCl₃–MeOH, 20:1 → 10:1) to give compound 13 (13.4 mg, 68.7%) as a colorless solid: ¹H NMR (DMSO-d₆) δ 2.72 (d, J = 4.3 Hz, 3 H, NHCH₃), 4.17 (br s, 1 H, H-3′), 4.32 (s, 1 H, H-4′), 5.55 (m, 1 H, H-2′), 4.61 (br d, J = 5.5 Hz, 2 Hz, CH₂), 5.56 (d, J = 6.4 Hz, 1 H, exchangeable with D₂O, 2′-OH), 5.72 (d, J = 4.3 Hz, 1 H, exchangeable with D₂O, 3′-OH), 5.92 (d, J = 7.2 Hz, 1 H, H-1′), 7.13 (pseudo t, J = 7.9 and 7.6 Hz, 1 H, H-5″), 7.36 (d, J = 7.5 Hz, 1 H, H-4″ or -6″), 7.61 (d, J = 7.8 Hz, 1 H, H-4″ or -6″), 7.75 (s, 1 H, H-2″), 8.27 (br d, J = 4.3 Hz, 1 H, exchangeable with D₂O, NH), 8.49 (s, 1 H, H-8), 9.02 (br t, J = 6.2 and 5.7 Hz, 1 H, exchangeable with D₂O, N⁶H).

N⁶-(3-Iodobenzyl)-2-(methylamino)-9-[5-(methylcarbamoyl)-β-d-ribofuranosyl]adenine (14)

A solution of 13 (10 mg, 0.018 mmol) in 2 N CH₃NH₂/THF (1.5 mL) was heated for 3 days at 90 °C. After the reaction mixture was concentrated to dryness, the residue was purified on a preparative TLC (CHCl₃–MeOH, 10:1) to give 14 (7 mg, 70%) as a colorless solid: ¹H NMR (DMSO-d₆) δ 2.66 (d, J = 4.7 Hz, 3 H, -NHCH₃), 2.76 (d, J = 4.3 Hz, 3 H, NHCH₃), 4.18 (m, 1 H, H-3′), 4.25 (s, 1 H, H-4′), 4.57 (br s, 2 H, CH₂), 4.69 (m, 1 H, H-2′), 5.47 (d, J = 6.5 Hz, 1 H, exchangeable with D₂O, 2′-OH), 5.59 (d, J = 4.6 Hz, 1 H, exchangeable with D₂O, 3′-OH), 5.84 (d, J = 7.2 Hz, 1 H, H-1′), 6.28 (br d, J = 4.4 Hz, exchangeable with D₂O, NH), 7.11 (pseudo t, J = 8.0 and 7.8 Hz, 1 H, H-5″), 7.38 (d, J = 7.9 Hz, 1 H, H-4″ or -6″), 7.58 (d, J = 7.9 Hz, 1 H, H-4″ or -6″), 7.70 (m, 1 H, exchangeable with D₂O, NH), 7.76 (s, 1 H, H-2″), 8.02 (s, 1 H, H-8), 8.05 (br s, 1 H, exchangeable with D₂O, NH).

N⁶-(3-Iodobenzyl)-2-(methylthio)-9-[5-(methylcarbamoyl)-β-d-ribofuranosyl]adenine (15)

A solution of 13 (15 mg, 0.029 mmol) and sodium thiomethoxide (4.0 mg, 0.057 mmol) in anhydrous ethylene glycol dimethyl ether (2 mL) was heated at 80 °C, under nitrogen atmosphere, for 3 days. After cooling to room temperature, the reaction mixture was neutralized with glacial acetic acid and evaporated to dryness. The residue was purified on a preparative TLC (CH₂Cl₂–MeOH, 9.5:0.5) to give 15 (5.6 mg, 36.5%) as a yellow solid: ¹H NMR (DMSO-d₆) δ 2.43 (s, 3 H, SH₃), 2.74 (d, J = 4.3Hz, 3 H, NHCH₃), 3.48 (br s, 2 H, 2 × OH), 4.19 (m, 1H, H-3′), 4.31 (s, 1 H, H-4′), 4.62 (br s, 3 H, CH₂ & H-2′), 5.87 (d, J = 7.9 Hz, 1 H, H-1′), 7.11 (pseudo t, J = 8.0 and 7.8 Hz, 1 H, H-5″), 7.90 (d, J = 7.9 Hz, 1 H, H-4″ or -6″), 7.58 (d, J = 7.9 Hz, 1 H, H-4″ or -6″), 7.76 (s, 1H, H-2″), 8.24 (br s, 1 H, NH), 8.35 (s, 1 H, H-8), 8.68 (br s, 1 H, NH).

Methyl 5-(tert-Butyldiphenylsilyl)-β-d-ribofuranoside (17)

To a mixture of methyl β-d-ribofuranoside (16, purchased from Sigma Chemical Co., 460 mg, 2.8 mmol) and anhydrous methylene chloride (20 mL) were added triethylamine (0.468 mL, 3.36 mmol), tert-butyldiphenylchlorosilane (0.9 mL, 3.46 mmol), and DMAP (13.7 mg, 0.112 mmol) successively at room temperature. The reaction mixture was stirred for 18 h at room temperature under nitrogen. The reaction mixture was washed with water (20 mL), saturated ammonium chloride (20 mL), and brine (20 mL), dried over anhydrous MgSO₄, filtered, and concentrated to dryness. The residue was separated by silica gel column chromatography (CHCl₃–MeOH, 50:1) to yield compound 17 [R_f = 0.48 (CHCl₃–MeOH, 10:1), 618 mg, 54.8%] as a thick syrup: ¹H NMR (DMSO-d₆) δ 0.96 (s, 9 H, t-Bu), 3.22 (s, 3 H, OCH₃), 3.61 (dd, J = 11.0 and 5.3 Hz, 1 H), 3.74 (d, J = 4.4 Hz, 1 H), 3.81 (dd, J = 11.0 and 2.7 Hz, 1 H), 3.90 (m, 1 H), 4.00 (m, 1 H), 4.68 (s, 1 H, H-1′), 4.84 (br s, 1 H, exchangeable with D₂O, OH), 5.05 (br s, 1 H, exchangeable with D₂O, OH), 7.45 and 7.67 (m, 10 H, Ph₂).

Methyl 5-(tert-Butyldiphenylsilyl)-2,3-dibenzoyl-β-d-ribofuranoside (18)

To a solution of compound 17 (579 mg, 1.44 mmol) in methylene chloride–pyridine (4:1, 12.5 mL) was added dropwise benzoyl chloride (0.367 mL, 3.16 mmol) at 0 °C. The reaction mixture was stirred for 2.5 h at 0 °C and for 14.5 h at room temperature. Ice was added to quench the reaction, and the mixture was stirred for 1 h. Methylene chloride (100 mL) was added, and two phases were separated. Organic layer was washed with water, saturated ammonium chloride, and brine, dried over anhydrous MgSO₄, filtered, and concentrated to dryness to give crude compound 18, which was then purified by silica gel column chromatography (Hx–EtOAc, 5:1 → 1:1) to yield compound 18 [R_f = 0.75 (CHCl₃–MeOH, 10:1), 869 mg, 99%] as a thick syrup: ¹H NMR (CDCl₃) δ 1.05 (s, 9 H, t-Bu), 3.42 (s, 3 H, OCH₃), 3.86 (dd, J = 11.1 and 4.8 Hz, 1 H, H-5a), 3.92 (dd, J = 11.1 and 4.7 Hz, 1 H, H-5b), 4.48 (dd, J = 10.6 and 4.6 Hz, 1 H), 5.15 (s, 1 H), 5.63 (d, J = 5.0 Hz, 1 H), 5.82 (pseudo t, J = 5.8 and 5.4 Hz, 1 H), 7.29–8.18 (m, 20 H, Ar).

Methyl 2,3-Dibenzoyl-β-d-ribofuranoside (19)

A solution of compound 18 (849 mg, 1.39 mmol) and 1.0 M tetrabutylammonium fluoride in THF (1.53 mL, 1.53 mmol) was stirred for 2 h at room temperature. After evaporation of the solvent, the residue was purified by silica gel column chromatography (Hx–EtOAc, 1:1) to yield compound 19 [R_f = 0.50 (Hx-EtOAc, 1:1), 461 mg, 89%] as a thick syrup: ¹H NMR (DMSO-d₆) δ 3.38 (s, 3 H, OCH₃), 3.62 (m, 2 H, H-5), 4.37 (q, J = 5.2 Hz, 1 H, H-4), 5.02 (pseudo t, J = 6.0 and 5.3 Hz, 1 H, exchangeable with D₂O, 5-OH), 5.18 (s, 1 H, H-1), 5.45 (m, 1 H, H-2), 5.53 (t, J = 5.2 Hz, 1 H, H-3), 7.42–7.88 (m, 10 H, Ar).

1-O-Methyl 2,3-Dibenzoyl-β-d-ribofuronic Acid (20)

A mixture of compound 19 (374.8 mg, 1.01 mmol), ruthenium-(IV) oxide (10 mg), and sodium periodate (1161 mg, 5.43 mmol) in CHCl₃–CH₃CN–H₂O (2:2:3, 14 mL) was stirred vigorously for 2.5 h at room temperature. Chloroform (20 mL) was added, and semisolid was removed by filtration. The two layers of filtrate were separated, and aqueous layer was extracted with chloroform (2 × 40 mL). Combined organic layer and extracts were washed with brine, dried over anhydrous MgSO₄, filtered, and concentrated to dryness. After drying in vacuo overnight, 1-O-methyl-2,3-dibenzoyl-β-d-ribofuranuronic acid (20, 340 mg, 89.5%) was obtained as a thick syrup: ¹H NMR (CDCl₃) δ 3.56 (s, 3 H, OCH₃), 4.90 (d, J = 6.1 Hz, 1 H, H-4), 5.25 (s, 1 H, H-1), 5.66 (d, J = 5.0 Hz, 1 H, H-2), 6.00 (pseudo t, J = 5.7 and 5.4 Hz, 1 H, H-3), 7.32–7.99 (m, 10 H, Ar).

Methyl 1-O-Methyl-2,3-dibenzoyl-β-d-ribofuranuronate (21)

N-Ethyl-N′-(diaminopropyl)carbodiimide (EDAC, 198 mg, 1.04 mmol) was added to a solution of acid (20, 0.16 g, 0.414 mmol) in MeOH (3 mL), and the reaction mixture was stirred for 3 h at room temperature. After the solvent was removed by rotary evaporation, the residue was dissolved in chloroform (50 mL), washed with water (30 mL) and brine (30 mL), dried over anhydrous MgSO₄, filtered, and concentrated to dryness. The residue was purified on a preparative TLC (Hx–EtOAc, 1:1) to yield methyl 1-O-methyl-2,3-dibenzoyl-β-d-ribofuronate [21, R_f = 0.77 (Hx–EtOAc, 1:1), 120 mg, 72.3%] as a colorless solid. ¹H NMR (CDCl₃) δ 3.52 (s, 3 H, 1-OCH₃), 3.82 (2, 3 H, 5-OCH₃), 4.84 (d, J = 6.2 Hz, 1 H, H-4), 5.21 (s, 1 H, H-1), 5.62 (d, J = 4.8 Hz, 1 H, H-2), 6.01 (pseudo t, J = 5.7 and 5.5 Hz, 1 H, H-3), 7.32–7.99 (m, 10 H, Ar).

N,1-O-Dimethyl-2,3-dibenzoyl-β-d-ribofuranuronamide (22)

A mixture of methyl ester 21 (35 mg, 0.087 mmol) and 2.0 M methylamine in THF (3 mL) was heated for 15 h at 50 °C in a sealed tube. The volatiles were removed by evaporation, and the residue was reacted with benzoyl chloride (0.15 mL, 1.29 mmol) in methylene chloride–pyridine (2:1, 6 mL) for 3 h at room temperature. After workup as procedure for compound 18, the residue was separated by preparative TLC (Hx–EtOAc, 1:1) to yield compound 22 [R_f = 0.28 (Hx–EtOAc, 1:1) or 0.77 (CHCl₃–MeOH, 10:1), 25 mg, 72%] as a syrup: ¹H NMR (CDCl₃) δ 2.92 (d, J = 5.0 Hz, 3 H, NHCH₃), 3.58 (s, 3 H, 1-OCH₃), 4.82 (d, J = 5.5 Hz, 1 H, H-4), 5.24 (s, 1 H, H-1), 5.60 (m, 1 H, H-2), 5.87 (t, J = 5.1 Hz, 1 H, H-3), 6.68 (br m, 1 H, NH), 7.33–7.99 (m, 10 H, Ar).

N-Methyl-1-O-acetyl-2,3-dibenzoyl-α-d-ribofuronamide (23a) and N-methyl-1-O-acetyl-2,3-dibenzoyl-β-d-ribofuranuronamide (23b)

To a solution of 22 (1.533 g, 3.84 mmol) and acetic anhydride (3.8 mL, 40.3 mmol) in glacial acetic acid (19 mL) was added dropwise concentrated H₂SO₄ (1.125 mL, 21.1 mmol), and the reaction mixture was stirred for 15 h at room temperature. After water (30 mL) was added slowly, the mixture was extracted with methylene chloride (150 mL × 3) and the organic layer was washed with saturated NaHCO₃ and brine, dried over anhydrous MgSO₄, filtered, and connectrated to dryness. The residue was purified by silica gel column chromatography (CHCl₃–MeOH, 20:1) to give a mixture of 23a and 23b [R_f = 0.71 and 0.76 (CHCl₃–MeOH, 20:1), respectively, 0.55 g, 33.5%] as a foam. Analytical samples were separated by preparative TLC (CHCl₃–MeOH, 20:1): ¹H NMR (CDCl₃) (compound 23a) δ 2.10 (s, 3 H, OAc), 2.91 (d, J = 4.9 Hz, 3 H, -NHCH₃), 4.96 (s, 1 H, H-4), 5.45 (pseudo t, J = 5.5 and 5.0 Hz, 1 H, H-2), 6.08 (d, J = 5.9 Hz, 1 H, H-3), 6.71 (d, J = 4.7 Hz, 1 H, H-1), 6.72 (br s, 1 H, NH), 7.28 (pseudo t, J = 7.8 and 7.7 Hz, 2 H, Ar), 7.48 (q, J = 7.8 Hz, 3 H, Ar), 7.62 (pseuso t, J = 7.7 and 6.9 Hz, 1 H, Ar), 7.79 (d, J = 7.4 Hz, 2 H, Ar), 8.13 (d, J = 7.8 Hz, 2 H, Ar); (compound 23b) δ 2.17 (s, 3 H, OAc), 2.90 (d, J = 4.9 Hz, 3 H, NHCH₃), 4.89 (d, J = 6.2 Hz, 1 H, H-4), 5.73 (d J = 4.9 Hz, 1 H, H-2), 5.96 (t, J = 5.9 Hz, 1 H, H-3), 6.43 (s, 1 H, H-1), 6.50 (br s, 1 H, NH), 7.37 (pseudo t, J = 7.8 and 7.6 Hz, 4 H, Ar), 7.48–7.58 (m, 2 H, Ar), 7.93 (d, J = 8.1 Hz, 2 H, Ar), 7.98 (d, J = 7.3 Hz, 2 H, Ar).

2-Chloro-N⁶-(3-iodobenzyl)adenine (24a)

A solution of 2,6-dichloropurine (purchased from Aldrich Chemical Co., 1 g, 5.3 mmol), 3-iodobenzylamine hydrochloride (1.7 g, 5.8 mmol), and triethylamine (2.2 mL, 15.35 mmol) in ethanol (10 mL) was stirred for 5 days at room temperature. The colorless solid formed was collected by suction, washed with small amount of cold ethanol, and dried to give compound 24a (1.16 g, 60%): mass (EI) 385 (M⁺); ¹H NMR (DMSO-d₆) δ 4.59 (br s, 2 H, CH₂), 7.13 (pseudo t, J = 8.2 and 7.5 Hz, 1 H, Bn), 7.36 (d, J = 7.5 Hz, 1 H, Bn), 7.61 (d, J = 7.5 Hz, 1 H, Bn), 7.74 (s, 1 H, Bn), 8.14 (s, 1 H, H-8), 8.76 (br s, 1 H, exchangeable with D₂O, NH), 13.14 (br s, 1 H, exchangeable with D₂O, NH); UV (MeOH) λ_max, 281.7, 257.5, 232.5 nm.

2-Chloro-N⁶-(3-iodobenzyl)-9-[5-(methylcarbamoyl)-2,3-di-O-benzoyl-β-d-ribofuranosyl]adenine (25)

A mixture of 2-chloro-N⁶-(3-iodobenzyl)adenine (24a, 165 mg, 0.43 mmol), ammonium sulfate (catalytic amount), and HMDS (15 mL) was refluxed for 4 h under nitrogen to provide the silylated derivative 24b. The clear solution was concentrated to dryness in vacuo with exclusion of moisture, and the residue was dissolved in dry dichloroethane (6 mL). A solution of 23 (141 mg, 0.33 mmol) in dry dichloroethane (6 mL) and TMSOTf (83 µL, 0.43 mmol) were added, and the reaction mixture was stirred for 0.5 h at room temperature and refluxed for 62 h under nitrogen. Saturated NaHCO₃ (10 mL) was added, and the mixture was stirred for 15 min. Two layers were separated, and the aqueous layer was extracted with methylene chloride (50 mL × 3), washed with brine, dried over anhydrous MgSO₄, filtered, and concentrated to dryness. The residue was purified on a preparative TLC (CHCl₃–MeOH, 20:1) to give 25 [R_f = 0.58 (CHCl₃–MeOH, 20:1), 83 mg, 33 %] as a foam: MS (CI NH₃) 753 (M⁺+l); ¹H NMR (CDCl₃) δ 3.10 (d, J = 4.6 Hz, 3 H, NHCH₃), 4.79 (br s, 2 H, CH₂), 4.97 (s, 1 H, H-4′), 6.08 (m, 1 H, H-3′), 6.15–6.25 (m, 3 H, H-2′, 1′, NH), 7.06–8.06 (m, 15 H, Ar), 8.52 (br s, 1 H, NH).

2-Chloro-N⁶-(3-iodobenzyl)-9-(2,3,5-tri-O-benzoyl-β-d-ribofuranosyl)adenine (26)

A mixture of 2-chloro-N⁶-(3-iodobenzyl)adenine (24a, 0.84 g, 2.18 mmol), ammonium sulfate (catalytic amount), and HMDS (20 mL) was refluxed for 5 h under nitrogen to provide the silylated derivative 24b. The clear solution was concentrated to dryness in vacuo with exclusion of moisture, and the residue was dissolved in dry dichloroethane (6 mL). A solution of acetyl-2,3,5-tri-O-benzoyl-β-d-ribofuranoside (purchased from Janssen Chimica Chemical Co., 1 g, 1.98 mmol) in dry dichloroethane (12 mL) and TMSOTf (0.42 mL, 2.18 mmol) were added, and the reaction mixture was stirred for 20 min at room temperature and refluxed for 14 h under nitrogen. After similar workup for compound 25, the residue was purified by silica gel column chromatography (Hx–EtOAc, 2:1) to give 26 [R_f = 0.11 (Hx–EtOAc, 3:1), 1.495 g, 91%] as a colorless foam: ¹H NMR (CDCl₃) δ 4.69–4.92 (m, 5 H, CH₂, H-4′, H-5′), 6.15 (m, 3 H, H-2′, H-3′, NH), 6.45 (d, J = 4.3 Hz, 1 H, H-1′), 7.07 (pseudo t, 1 H, Bn), 7.31–8.10 (m, 20 H, Ar).

Methods for Receptor Binding and Adenylate Cyclase Measurement

Procedures for preparation of rat brain membranes and CHO cell membranes were as reported.^10,11,14 For binding experiments, membrane homogenates were frozen and stored at −20 °C for ≤2 months. Adenosine deaminase (ADA) was from Boehringer Mannheim (Indianapolis, IN). [³H]R-PIA was from Amersham (Arlington Heights, IL), and [³H]CGS 21680 was from DuPont NEN (Boston, MA). [¹²⁵I]-AB-MECA was prepared as described by Olah et al.¹⁴

Binding of [¹²⁵I]AB-MECA to CHO cells stably transfected with the A₃ receptor clone was performed essentially as described.^11,14 Assays were performed in 50 mM Tris/10 mM MgCl₂/1 mM EDTA buffer (adjusted to pH 8.26 at 5 °C) in glass tubes and contained 100 µL of the membrane suspension, 50 µL of [¹²⁵I]AB-MECA (final concentration 0.3 nM), and 50 µL of inhibitor. Inhibitors were routinely dissolved in DMSO and were then diluted with buffer; final DMSO concentrations never exceeded 1%; this concentration did not influence [¹²⁵I]-AB-MECA binding. Incubations were carried out in duplicate for 1 h at 37 °C, and were terminated by rapid filtration over Whatman GF/B filters, using a Brandell cell harvester (Brandell, Gaithersburg, MD). Tubes were washed three times with 3 mL of buffer. Radioactivity was determined in a Beckman gamma 5500B γ-counter. Nonspecific binding was determined in the presence of 200 µM NECA. K_i values were calculated according to Cheng–Prusoff,²¹ assuming a K_d for [¹²⁵I]AB-MECA of 1.48 nM.⁶

Binding of [³H]PIA to A₁ receptors from rat cortical membranes and of [³H]CGS 21680 to A₂ receptors from rat striatal membranes was performed as described previously.^8,11 Adenosine deaminase (2 units/mL) was present during the preparation of brain membranes. Additional deaminase was not added during incubation with the radioligand.

Competition for binding of [³H]NBTI was carried out by a modification of the procedure of Marangos et al.²⁴ Rat striatal membranes, prepared as above, were subjected to incubation for 30 min at 23 °C with 0.3 nM [³H]NBTI and varying concentrations of the nucleoside derivative in Tris buffer, pH 7.4, in a total of 0.5 mL. For nonspecific binding 5 µM S-(p-nitrobenzyl)-6-thioguanosine (Sigma, St. Louis, MO) was added, and specific binding was 95% of total. A K_d value of 0.15 nM was used in the calculation of K_i values.²⁴ Specific binding was 95% of total.

Adenylate cyclase was assayed in membranes from CHO cells stably expressing the rat A₃ receptor, prepared as above, using a previously reported method.¹⁰ The method involved addition of [α-³²P]ATP to membranes in the presence of forskolin to stimulate adenylate cyclase and papaverine as a phosphodiesterase inhibitor. The reaction was terminated by addition of a stop solution containing 20 000 cpm/mL [³H]cyclic AMP. The total radiolabeled cyclic AMP was isolated on columns of Dowex 50 ion-exchange resin and alumina. Maximal inhibition of adenylate cyclase activity corresponded to ~40% of total activity under conditions of stimulation (typically by 6–8-fold) in the presence of 1 µM forskolin. IC₅₀ values were calculated using InPlot (Graphpad, San Diego, CA).

Table 3.

Inhibition by Various N ⁶-Benzyladenosine Derivatives of the Specific Binding of [³H]-S-(4-Nitrobenzyl)-6-thioinosiante at Adenosine Uptake Sites in Rat Brain Membranes and the Selectivity Ratio for Affinity at Cloned Rat A₃ Receptors (K_i values from Table 1)

compd	Ki (NBTI)a	Ki (NBTI)/Ki ([125I]AB-MECA
N 6-benzyladenosine	203 ± 93	1.69
1	28200 ± 10700	22000
13	15200 ± 5200	46000
15	49500 ± 633	22000

^aExpressed in nanomolar as K_i ± SEM for three or four determinations, each done in triplicate. Rat striatal membranes were incubated for 30 min at 23 °C with 0.3 nM [³H]NBTI and varying concentrations of the nucleoside derivative in Tris buffer, pH 7.4 in a total of 0.5 mL. Nonspecific binding was determined in the presence of 5 µM S-(p-nitrobenzyl)-6-thioguanosine.

Abbreviations

AB-MECA

N⁶-(4-amino-3-iodobenzyl)adenosine-5′-N-methyluronamide

APNEA

N⁶-[2-(4-aminophenyl)ethyl]adenosine

CGS 21680

2-[4-[(2-carboxyethyl)-phenyl]ethyl-amino]-5′-N-(ethylcarbamoyl)adenosine

CHO

Chinese hamster ovary

DMAP

4-(N,N-dimethylamino)pyridine

DMF

N,N-dimethylformamide

DMSO

dimethyl sulfoxide

EDAC

1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide hydrochloride

HMDS

1,1,1,3,3,3-hexamethyldisilazane

IB-MECA

N⁶-(3-iodobenzyl)adenosine-5′-N-methyluronamide

NBTI

S-(4-nitrobenzyl)-6-thioinosine

NECA

5′-N-(ethylcarbamoyl)adenosine;

PIA

(R)-N⁶-(phenylisopropyl)adenosine

TBAF

tetrabutylammonium fluoride

TBDPSiCl

tert-butyldiphenylsilyl chloride

THF

tetrahydrofuran

TMSOTf

trimethylsilyl trifuoromethanesulfonate

Tris

tris(hydroxymethyl)-aminomethane