2-Triazole-Substituted Adenosines: A New Class of Selective A₃ Adenosine Receptor Agonists, Partial Agonists, and Antagonists

Abstract

“Click chemistry” was explored to synthesize two series of 2-(1,2,3-triazolyl)adenosine derivatives (1–14). Binding affinity at the human A₁, A_2A, and A₃ARs (adenosine receptors) and relative efficacy at the A₃AR were determined. Some triazol-1-yl analogues showed A₃AR affinity in the low nanomolar range, a high ratio of A₃/A_2A selectivity, and a moderate-to-high A₃/A₁ ratio. The 1,2,3-triazol-4-yl regiomers typically showed decreased A₃AR affinity. Sterically demanding groups at the adenine C2 position tended to reduce relative A₃AR efficacy. Thus, several 5′-OH derivatives appeared to be selective A₃AR antagonists, i.e., 10, with 260-fold binding selectivity in comparison to the A₁AR and displaying a characteristic docking mode in an A₃AR model. The corresponding 5′-ethyluronamide analogues generally showed increased A₃AR affinity and behaved as full agonists, i.e., 17, with 910-fold A₃/A₁ selectivity. Thus, N⁶-substituted 2-(1,2,3-triazolyl)-adenosine analogues constitute a novel class of highly potent and selective nucleoside-based A₃AR antagonists, partial agonists, and agonists.

Introduction

Adenosine receptors (AR) are G-protein-coupled receptors and consist of four subtypes classified as A₁, A_2A, A_2B, and A₃. Among the four AR subtypes, the A₃AR is the most recently identified.¹ The distribution of A₃AR is species-dependent, and in humans this subtype occurs in the lungs, liver, heart, kidneys, and brain.^2–4 Activation of this receptor subtype inhibits adenylyl cyclase activity, increases phosphatidylinositol-specific phospholipase C activity, and stimulates Ca²⁺ mobilization.³ Adenosine A₃ receptor stimulation induces cardioprotection through the activation of K_ATP channels⁴ and is also involved in neuroprotection, suggesting the possibility of using A₃AR agonists to treat cardial and cerebral ischemia.⁵ A₃AR agonists also exhibit systemic anticancer and chemoprotective effects.⁶ A₃AR modulators have been proposed as antiinflammatory and antiasthmatic drugs.^7,8 Selective A₃AR antagonists promise to be useful in the regulation of cell growth^8,9 and as cerebroprotective agents.^10,11 They also seem to enhance anticancer treatment by counteracting P-glycoprotein efflux in multidrug resistance.¹² A₃AR antagonists are also proposed as potential therapeutics for the treatment of glaucoma; application of A₃-AR antagonists externally to the eye substantially lowers intraocular pressure in mice and monkeys.^13–15

Although diverse in structure, most AR antagonists share some common structural features. In general, they are planar, aromatic, or π-electron-rich and nitrogen-containing heterocycles. Additionally, most AR antagonists lack the ribose moiety, which seems essential for agonist activity.¹⁶ Various heterocyclic classes have been identified as promising leads for A₃AR antagonists, among them 1,4-dihydropyridines, pyridines, deazaadenines, pyrazolopyrimidines, adenines, and 1,2,4-triazolo[ 4,3-a]quinoxalin-1-ones.^4,7,17,18

However, the A₃AR, more than other AR subtypes, is amenable to the design of nucleoside-based antagonists. The efficacy of nucleoside derivatives in activation of the A₃AR is particularly sensitive to molecular substitution of the ligand.¹⁹ A wide range of adenosine derivatives have been shown to antagonize this receptor, including the highly potent A₁AR agonist 2-chloro-N⁶-cyclopentyladenosine. N⁶-Benzyl groups are associated with reduced A₃AR efficacy, leading to partial agonists and antagonists. However, many of the nucleosides so far demonstrated to be antagonists of the A₃AR are not highly subtype-selective.²⁰ Nucleoside-based A₃AR antagonists maintaining an intact ribose moiety were reported by Volpini et al.,²¹ with a series of 8-alkynyladenosine derivatives that exhibited A₃AR selectivity, but suffered from weak A₃AR affinity. A spirolactam derivative, in which the 5′-alkyluronamide group was cyclized onto the 4′ carbon, was found to potently and selectively antagonize the A₃AR.^15,19 An advantage of nucleoside-based A₃AR antagonists over other heterocyclic antagonists is the ability to achieve high affinity at murine species.

Recently, researchers from CV Therapeutics described a series of 2-pyrazolyladenosine analogues.²² Several representative compounds containing an N⁶-methyl substituent proved to display high affinity and selectivity for the A₃AR. This study confirms the former finding²³ that introduction of a methyl group into the N⁶ position increases the affinity for the human A₃AR and enhances the selectivity versus A₁ and A_2AARs. On the basis of these results, we explored the versatile “click chemistry” approach²⁴ to synthesize two series of N⁶-methyl-2-(1,2,3-triazolyl)adenosine derivatives and evaluated their affinity, selectivity, and efficacy at the A₃AR. Although a number of 1,2,3-triazole nucleoside derivatives have been described,²⁵ most involve base replacement with 1,2,3-triazole or introduction of 1,2,3-triazole at C8 or at the sugar moiety.

Results and Discussion

Chemistry

The synthesis of the 1,2,3-triazol-1-yladenosine derivatives is depicted in Scheme 1. 2-Iodo-N⁶-methyladenosine 22²³ was prepared by reacting 21²⁶ with 2.0 M CH₃NH₂ in THF. Since the reaction conditions²⁷ for a Cu(I)-catalyzed nucleophilic substitution are very similar to those used in the “click” variant of Huisgen’s 1,3-dipolar cycloaddition, we initially attempted to perform a one-pot conversion of 22 to the desired 1,4-disubstituted 1,2,3-triazoles. The 2-azido derivative 23 was isolated as the main reaction product, and only a minor amount of the appropriate triazole was formed. This event forced us to perform the reaction in two steps. First the azido intermediate 23 was prepared in 66% yield from 22. ¹H and ¹³C NMR in DMSO-d₆ proved the presence of a tautomeric fused tetrazole form (17%) of the 2-azidoadenosine derivative 23, due to a spontaneous cyclization (Scheme 2). Such azido/tetrazole tautomerism has been previously reported for 2-azidoadenine and for 2-azidoadenosine.^28,29 Next we applied a Cu(I)-catalyzed 1,3-cycloaddition reaction of azide 23 with the appropriate alkyne to generate the triazole analogues 1–11 (Scheme 1).³⁰ Generally, the use of a water/butanol mixture as a solvent for the 1,3-cycloaddition allowed simple isolation of the desired compounds, which precipitated from the reaction medium.

Scheme 1.

Synthesis of 1,2,3-Triazol-1-yl Analogues of N⁶-Methyladenosine 1–11^a

^a Reagents and conditions: (a) CH₃NH₂ in THF, 2 days; (b) CuSO₄·5H₂O, sodium ascorbate, L-proline, Na₂CO₃, NaN₃, H₂O/^tBuOH 1:1, 60 °C; (c) CuSO₄·5H₂O, sodium ascorbate, alkyne, H₂O/^tBuOH 3:1, room temp.

Scheme 2.

Azido/Tetrazole Tautomerism of a 2-Substituted Adenosine Derivative 23

Similarly, the 1,2,3-triazol-4-yl analogues 12–14 (Scheme 3) were prepared by a Cu⁺-catalyzed Huisgen 1,3-dipolar cycloaddition reaction of 2-ethynyl-N⁶-methyladenosine (25) with the appropriate azide.

Scheme 3.

Synthesis of 1,2,3-Triazol-4-yl Analogues of N⁶-Methyladenosine 12–14^a

^a Reagents and conditions: (a) trimethylsilylacetylene, CuI, (Ph₃P)₃PdCl₂, DMF; (b) 7 N NH₃ in MeOH, 0 °C; (c) CuSO₄·5H₂O, sodium ascorbate, alkyne, H₂O/^tBuOH 3:1, room temp.

The synthesis of the 5′-N-ethylcarbamoyl 2-(1,2,3)-triazol-1-yladenosine analogues 15–19 was carried out starting from 2-iodo-9-(2′,3′-O-isopropylidene-β-D-ribofuranosyl)-N⁶-methyladenine (26). After permanganate oxidation, carboxylic acid 27 was converted into its p-nitrophenyl ester 28, which upon treatment with ethylamine gave uronamide 29. Deprotection with 80% trifluoroacetic acid yielded 5′-ethylcarbamoyl-N⁶-methyl-2-iodoadenosine (30).³⁰ The conversion of this 2-iodo derivative into the azido intermediate 31 was performed in 79% yield. The presence of a tautomeric fused tetrazole form (20%) of the 2-azidoadenosine derivative 31, due to a spontaneous cyclization, was here also observed in the NMR spectrum. Finally we applied the Cu(I)-catalyzed 1,3-cycloaddition reaction of azide 31 with the appropriate alkyne to generate the triazole analogues 15–19 (Scheme 4).

Scheme 4.

Synthesis of 1,2,3-Triazol-1-yl Analogues of N⁶-Methyladenosine-5′-N-ethyluronamide 15–19^a

^a Reagents and conditions: (a) KMnO₄, KOH, room temp, 20 h; (b) p-nitrophenol, EDCI, DMF, room temp; (c) ethylamine; (d) 80% TFA/H₂O; (e) CuSO₄·5H₂O, sodium ascorbate, L-proline, Na₂CO₃, NaN₃, H₂O/^tBuOH 1:1, 60 °C; (f) CuSO₄·5H₂O, sodium ascorbate, alkyne, H₂O/^tBuOH 1:1, room temp.

2-Azido-N⁶-(5-chloro-2-methoxybenzyl)-2-(4-cyclopentylmethyl-1,2,3-triazol-1-yl)-9-(β-D-ribofuranosyl)adenine (20) was prepared in three steps starting from intermediate 21, as depicted in Scheme 5.

Scheme 5.

Synthesis of Compound 20, 2-(4-Cyclopentylmethyl-1,2,3-triazol-1-yl)-N⁶-(2-chloro-5-methoxybenzyl)adenosine^a

^a Reagents and conditions: (a) 2-chloro-5-methoxybenzylammonium chloride, Et₃N, EtOH, reflux; (b) CuSO₄·5H₂O, sodium ascorbate, L-proline, Na₂CO₃, NaN₃, H₂O/^tBuOH 1:1, 60 °C; (c) CuSO₄·5H₂O, sodium ascorbate, alkyne, H₂O/^tBuOH 1:1, room temp.

Biological Evaluation

The binding affinities of the newly synthesized adenosine derivatives were measured at the hA₁, hA_2A, and hA₃ARs expressed in CHO (Chinese hamster ovary) cells as previously described,²⁰ and their relative efficacy in the activation of the A₃AR was determined (Table 1). The binding affinity of more potent compounds at the ARs was evaluated with full competition curves, while the weaker compounds at the hA₁ and hA_2A ARs were measured at a fixed concentration of 10 μM. Several compounds showed affinity for the A₃AR in the low nanomolar range, a very high ratio of A₃/A_2A selectivity, and a moderate-to-high A₃/A₁ selectivity ratio. A functional assay of A₃AR activation consisted of the ability of a single, high concentration of the nucleoside (10 μM) to inhibit forskolin-stimulated adenylyl cyclase measured by the method of Nordstet and Fredholm,³² in comparison to the full agonist NECA (10 μM). Cl-IB-MECA (2-chloro-N⁶-(3-iodobenzyl)-5′-N-methylcarboxamidoadenosine) was also a full agonist (100%) in this assay.¹⁹ The range of efficacies observed depended on the nature of the groups at the 2 and N⁶ positions.

Table 1.

Binding Affinities of Adenosine Derivatives at Human A₁, A_2A, and A₃ARs Expressed in CHO Cells and Relative Efficacy at the A₃AR^a

	R1	R2	R3	Ki (nM) or % inhibition (in parentheses) at 10 μM			% efficacyb hA3
				hA1	hA2A	hA3
1	CH2OH	H	CH3	1000 ± 30	(13 ± 3)	10.4 ± 0.2	41 ± 6
2	CH2OH	ethyl	CH3	2920 ± 910	(18)	13.8 ± 3.3	23c
3	CH2OH	butyl	CH3	848 ± 76	(23)	11.7 ± 3.1	3 ± 8
4	CH2OH	2-hydroxyethyl	CH3	1270 ± 260	(14)	45.0 ± 4.4	25c
5	CH2OH	dimethylaminomethyl	CH3	3800 ± 600	(6)	117 ± 25	8c
6	CH2OH	phenyl	CH3	(36)	(5)	14.9 ± 1.7	14 ± 8
7	CH2OH	pyridin-2-yl	CH3	1970 ± 210	(40)	10.3 ± 1.5	11 ± 4
8	CH2OH	4-propoxyphenyl	CH3	(49)	(14)	25.2 ± 2.6	31c
9	CH2OH	benzyl	CH3	589 ± 55	(20)	9.5 ± 0.7	−1 ± 5
10d	CH2OH	cyclopentylmethyl	CH3	335 ± 13	(39)	1.3 ± 0.4	−5 ± 7
11	CH2OH	cyclohexylmethyl	CH3	1430 ± 60	(16 ± 1)	21.3 ± 8.1	2 ± 5
12	CH2OH	benzyl	CH3	770 ± 210	(21 ± 5)	53.9 ± 6.6	2 ± 3
13	CH2OH	3-methoxybenzyl	CH3	957 ± 65	(43 ± 10)	86.1 ± 3.8	−1 ± 3
14	CH2OH	3-Cl-benzyl	CH3	956 ± 6	(39 ± 10)	81.1 ± 5.0	0 ± 5
15d	C2H5NHCO	H	CH3	590 ± 70	(18 ± 3)	2.1 ± 0.1	102 ± 5
16	C2H5NHCO	butyl	CH3	750 ± 110	(43 ± 1)	5.6 ± 0.2	89 ± 3
17d	C2H5NHCO	pyridin-2-yl	CH3	1640 ± 90	(45 ± 12)	1.8 ± 0.6	90 ± 7
18d	C2H5NHCO	benzyl	CH3	510 ± 50	(33 ± 2)	2.8 ± 1.3	86 ± 5
19	C2H5NHCO	cyclopentylmethyl	CH3	1250 ± 150	(36 ± 7)	11.5 ± 1.4	83c
20	CH2OH	cyclopentylmethyl	2-Cl–5-MeO–Bn	830 ± 40	6000	18 ± 11	−6 ± 3
23	CH2OH		CH3	230 ± 10	(23)	10.8 ± 3.1	84 ± 9
31	C2H5NHCO		CH3	429 ± 55	(18 ± 3)	11.4 ± 4.2	112c
33	CH2OH		2-Cl–5-MeO–Bn	60 ± 10	1800 ± 500	1.4 ± 0.1	44 ± 5

^aAll binding experiments were performed using cells stably transfected with cDNA encoding one of the human ARs. Binding at human A₁, A_2A, and A₃ARs in this study was carried out as described in Experimental Section using [³H]CCPA, [³H]CGS 21680, or [¹²⁵I]IAB-MECA as a radioligand. Values from the present study are expressed as K_i values (mean ± SEM, n = 3, unless otherwise noted) or as percent displacement of radioligand.

^b% activation at 10 μM, relative to cyclic AMP inhibitory effect of 10 μM NECA (=100%). Cl-IB-MECA was also a full agonist (100%) in this assay.

^cn = 2.

^d10, LC 153; 15, LC 260; 17, LC 257; 18, LC 259.

The 2-azido precursor 23 showed high binding affinity at the A₃AR (K_i = 10.8 nM) and modest selectivity in comparison to the A₁AR. The 1,2,3-triazol-1-yl derivatives obtained by 1,3-dipolar cycloaddition of azide 23 with acetylene (1), butyne (2), and hexyne (3) maintained high affinity for the A₃AR and increased selectivity. They displayed K_i values of 10.4, 13.8, and 11.7 nM, respectively. Also, aromatic triazole substituents (6, 7, 9) resulted in similar K_i values of about 10 nM and even greater selectivity. Introducing nitrogen or oxygen including substituents at position 4 of the 1,2,3-triazole ring (4, 5, and 8) reduced the A₃AR affinity. Among the investigated analogues, the 4-cyclopentylmethyl derivative 10 exhibited the highest affinity for the A₃AR (K_i = 1.3 nM) and 260-fold selectivity in comparison to the A₁AR. Replacement of the cyclopentyl ring with a phenyl (9) or cyclohexyl (11) moiety adversely affected A₃AR affinity. Remarkably, the 1,2,3-triazol-4-yl regiomers (12–14) showed decreased affinity for the A₃AR in comparison to similar 1,2,3-triazol-1-yl regiomers. In particular, a comparison of homologous compounds 12 and 9 indicated a 6-fold loss of affinity at the A₃AR for the 4-yl isomer, approximately the same affinity at the A₁AR, and no significant measurable gain in affinity at the A_2AAR.

Replacement of the ribose 4′-hydroxymethyl moiety of the 2-azido derivative 23 by a 5′-N-ethyluronamide did not appreciably affect affinity at any of the AR subtypes. However, a similar substitution in the 2-(1,2,3-triazol-1-yl)-substituted series provided a modest (2- to 5-fold) increase in A₃AR affinity and small or no changes in A₁AR affinity, as demonstrated for the unsubstituted, 4-butyl, 4-pyridin-2-yl, and 4-benzyl substituted 1,2,3-triazol-1-yl combinations (15, 16, 17, and 18, in comparison to 1, 3, 7, and 9, respectively). Curiously, one 5′-uronamide, compound 19, exhibited decreased A₃AR affinity (K_i = 11.5 nM) compared to its potent 5′-OH analogue 10 (K_i = 1.3 nM).

Replacement of the N⁶-methyl substituent of the 2-azido precursor 23 by a sterically demanding 2-chloro-5-methoxybenzyl group yielded 33, which manifested very high A₃AR affinity (K_i = 1.4 nM). A similar replacement of the N⁶-methyl group of an analogue 10, also having a bulky 2-position substituent, to yield 20 reduced A₃AR affinity but not appreciably. This was in accordance with previous observations that a simultaneous substitution at the 6 and 2 positions did not improve A₃AR affinity.^22,33 Thus, the effects of substitution at the 2 and N⁶ positions were not independent; however, it was possible to retain considerable A₃AR selectivity (46-fold in compound 20). This was not representative of findings in a previous study in which double substitution greatly diminished the affinity and selectivity at the human A₃AR.²²

Whereas some previously synthesized 2-substituted adenosine derivatives^22,23 displayed selective A₃AR agonist activity, all 2-triazol-1-yl-N⁶-methyladenosine analogues synthesized with an unmodified ribose moiety (1–11 and 20) behaved as antagonists or weak partial agonists. Similar findings were reported for 2-ester derivatives of adenosine, in which a combination of 2 and N⁶ substitution reduced efficacy.³⁴ Direct ring substitution at the 4-position of the 1,2,3-triazole with alkyl or aryl groups resulted in weak partial agonists (1, 2, 4–8), but subtle changes of structure resulted in a loss of efficacy, e.g., the 4-butyl derivative 3. 2-Triazol-1-yl-N⁶-methyladenosine analogues with a methylene spacer between the 1,2,3-triazole moiety and a ring system yielded full A₃AR antagonists (9–11, 20), since they bound to the receptor but did not activate it. The 2-triazol-4-yl-N⁶-methyladenosine derivatives (12–14) also behaved as full A₃AR antagonists. Thus, the 5′-OH derivatives 3, 9–14, and 20 appeared to be A₃AR antagonists with the following order of decreasing selectivity for the A₃AR in comparison to the A₁AR: 4-cyclopentylmethyl-N⁶-methyl 10 (260-fold) > 4-butyl-N⁶-methyl 3 (72-fold), 4-cyclohexylmethyl-N⁶-methyl 11 (67-fold) > 4-cyclopentylmethyl-N⁶-(5-chloro-2-methoxybenzyl) 20.

Interestingly, the 5′-N-ethyluronamide modification was able to reestablish the A₃AR agonist activity in analogues with sterically bulky substitution at the 2 position. This is consistent with previous findings that similar 5′-uronamides overcome the efficacy-reducing activity of substitution at the adenine 2 and N⁶ positions but not at the ribose 3′ position.^19,35,36 Indeed, all 5′-N-ethyluronamide analogues studied here (15–19) proved to be full agonists at the A₃AR. Among them are highly selective A₃AR agonists, N⁶-methyladenosine-5′-N-ethyluronamide 2-(1,2,3-triazol-1-yl) derivatives: pyridin-2-yl 17 (910-fold) > unsubstituted 15 (280-fold) > benzyl 18 (180-fold). The 2-azido-N⁶-methyl precursors 23 and 31 also showed full agonist activity, whereas azide 33 having a bulky N⁶ group showed partial agonist activity.

Selected potent agonists in this series were measured in a functional assay of the human A_2BAR. At 10 μM, compounds 3–7, 15, and 16 did not significantly stimulate adenylyl cyclase in human A_2BAR-expressing CHO cells (<10% of the effect of 10 μM NECA, as a full agonist). Compounds 2, 8–14, 17, 19, and 23 at 10 μM stimulated adenylyl cyclase by <50%. Compounds 18 and 31 produced approximately 50% stimulation at 10 μM. Thus, selectivity for the A₃AR was demonstrated; these nucleosides that activate the A₃AR at low nanomolar concentrations activated the A_2BAR only at substantial micromolar concentrations.

On the basis of previous findings, it is predicted that only the analogues containing the substituted N⁶-benzyl group (5-chloro-2-methoxy), i.e., full agonist 20 and partial agonist 33, would be expected to bind in the nanomolar range to the rat A₃AR. Small alkyl groups at the N⁶ position, such as methyl and ethyl, although conducive to high affinity at the human A₃-AR, led to negligible affinity at the rat homologue of the receptor. Selected compounds were measured in binding to the rat A₃AR expressed in CHO cell membranes using [¹²⁵I]I-AB-MECA. The K_i values determined were as follows: compounds 5, 7, and 8, K_i > 10 μM; compound 9, K_i = 1.79 μM; compound 10, K_i = 0.312 μM.

Molecular Modeling

To explain the structural basis for the high binding affinity of the nucleoside 2-(4-cyclopentylmethyl-1,2,3-triazol-1-yl)-N⁶-methyl-9-(β-D-ribofuranosyl)adenine 10 at hA₃AR, we performed a computational study of ligand docking in a previously derived A₃AR model based on the high-resolution structure of bovine rhodopsin.^37,38 Various bound conformations of the C2-substituent and χ₁ angles for the adenine ring were generated for an energetically favorable binding location and orientation, and the resulting conformations were compared energetically in the putative binding site.

The result of docking 10 in the putative binding site of the A₃AR is shown in Figure 1A. The purine ring was surrounded by a hydrophobic pocket, defined by L91 (3.33) and L246 (6.51). In addition, the H-bonds formed between the exocyclic amine and the hydroxyl group of S247 (6.52) and between the purine N¹ atom and the side chain of N250 (6.55). The 2′-OH group of the ribose moiety formed a H-bond with the side chain of Q167 (EL2), and the 3′-OH group formed an intramolecular H-bond with the 5′-OH group. Unlike the previously reported docking models of N⁶-substituted adenosines,³⁷ here the 5′-OH group H-bonded with the side chain of H272 (7.43) and the backbone carbonyl group of S271 (7.42). The cyclopentyl moiety interacted with aliphatic hydrophobic residues, M177 (5.38) and V178 (5.39), through a hydrophobic interaction and were situated in proximity to F168 (EL2) F182 (5.43), consistent with the optimized binding affinity of compound 10.

Figure 1.

(A) Docking complexes of compound 10, 2-(4-cyclopentylmethyl-1,2,3-triazol-1-yl)-N⁶-methyladenosine. (B) Superimposition of Cl-IB-MECA in red and compound 10 in color by atom type. Residues that were within 5 Å to the ligand in this putative binding site were L91 (3.33), T94 (3.36), H95 (3.37), Q167 (EL2), F168 (EL2), M172 (EL2), S181 (5.42), M177 (5.38), V178 (5.39), F182 (5.43), W243 (6.48), L246 (6.51), S247 (6.52), N250 (6.55), C251 (6.56), I268 (7.39), S271 (7.42), and H272 (7.43). The ligand is represented by a ball-and-stick model. The H-bonds are indicated with yellow dots. By use of the MOLCAD ribbon surface program, the A₃AR is shown in a ribbon model with different colors for each TM (TM1, red; TM2, orange; TM3, yellow; TM4, green; TM5, cyan; TM6, blue; TM7, purple; H8, violet).

A comparison of the docking modes of Cl-IB-MECA and compound 10 in the putative binding domain showed considerable overlap of the ribose rings and of the adenine moieties, although in compound 10 both were situated a little closer to extracellular loop 2 (Figure 1B). Previously, it was noted that 5′-uronamide analogues, typically of derivatives having bulky N⁶-subsitituents, generally gain affinity in comparison to the corresponding 5′-hydroxyl analogue. Here, the 5′-uronamide analogue 19 (agonist) of the most potent 5′-hydroxyl analogue 10 (antagonist) displayed a lower binding affinity, which could be explained by the shift of the ribose position in adenosine analogue having a bulky 2-(4-cyclopentylmethyl-1,2,3-triazol-1-yl) substituent in comparison to those having N⁶ bulky substituents. The binding of the cyclopentylmethyl group in 10 was directed more toward the upper part of TM5, partially overlapping with the binding site of the 3-iodophenyl ring in Cl-IB-MECA. Curiously, other closely related triazolo derivatives displayed a higher potency of the 5′-uronamide analogues; thus, compound 10 must bind to the receptor in a very distinct manner. There was a subtle difference in orientation between the 2-cyclopentyl group and bulkier groups like benzyl (9) or cyclohexylmethyl (11), which were associated with unfavorable van der Waals interactions and resulted in a decrease of binding affinity of 7- and 16-fold, respectively. In addition, the same preferred χ₁ angles of the energetically favorable bound conformation, common to Cl-IB-MECA and compound 10, were consistent with the empirical finding that the combination of bulky N⁶ and C2 substituents was unfavorable for A₃AR selectivity because of competitive interaction of these bulky substituents. Thus, the modeling has demonstrated that the human A₃AR preference of 2-(4-cyclopentylmethyl-1,2,3-triazol-1-yl) derivatives in the 5′-OH series might be explained by optimal van der Waals interactions.

Conclusions

Several 2-(1,2,3-triazol-1-yl)-N⁶-methyl-substituted adenosine derivatives described in the present study displayed A₃AR affinities in the low nanomolar range, showed very high A₃/A_2A, and a moderate to high A₃/A₁ selectivity. Contrary to what we expected, the 2-triazole analogues with an unmodified ribose moiety (1–14) showed antagonist or weak partial agonist activity at the A₃AR. A 2-(4-cyclopentylmethyl-(1,2,3-triazol-1-yl))-N⁶-methyl derivative 10 was 260-fold selective in binding in comparison to the A₁AR. The binding of the 4-cyclopentylmethyl group in 10, in distinction to the binding of closely related bulky groups pendent on the triazole ring, was directed more toward the upper part of TM5 partially overlapping with the binding site of the 3-iodophenyl ring in Cl-IB-MECA. The 5′-N-ethyluronamide modification was dominant over the efficacy reducing effects at the 2 position and was capable of fully re-establishing the A₃AR agonist activity, resulting in highly potent and selective A₃AR agonists 15–19. The most selective agonist derivative was compound 17, 9-(5-ethylcarbamoyl-β-D-ribofuranosyl)-N⁶-methyl-2-(4-pyridin-2-yl-1,2,3-triazol-1-yl)adenine, which was 910-fold selective in binding to the A₃AR in comparison to the A₁AR. The retention of high human A₃AR affinity in compound 20 was not typical of previous findings that double bulky substitution at the 2 and N⁶ positions tended to reduce A₃AR affinity markedly. Thus, the 2-triazol-1-yl-N⁶-methyladenosine analogues 1–11 constitute a novel class of highly potent and selective nucleoside-based A₃AR partial agonists and antagonists (all of which maintain an intact ribose in the molecular structure) and agonists. Since the reported analogues show excellent affinity for the A₃AR and span the full intrinsic activity range, they might be useful as pharmacological tools or as leads for further optimization.

Experimental Section

All reagents were from standard commercial sources and of analytic grade, except for the benzylic azides, which were prepared by treating the corresponding benzylic bromides with NaN₃ in DMF. Precoated Merck silica gel F254 plates were used for TLC, and spots were examined under UV light at 254 nm and further visualized by sulfuric acid–anisaldehyde spray. Column chromatography was performed on ICN silica gel (63–200 μm, 60 Å, ICN Biochemicals, Eschwege, Germany). NMR spectra were obtained with a Varian Mercury 300 MHz spectrometer. Chemical shifts are given in ppm (δ) relative to the residual solvent signals, which in the case of DMSO-d₆ were 2.54 ppm for ¹H and 40.5 ppm for ¹³C. Structural assignment was confirmed with COSY and DEPT. All signals assigned to hydroxyl groups were exchangeable with D₂O. Exact mass measurements were performed on a quadrupole/orthogonal-acceleration time-of-flight (Q/oaTOF) tandem mass spectrometer (qToF 2, Micromass, Manchester, U.K.) equipped with a standard electrospray ionization (ESI) interface. Samples were infused in a 2-propanol/water (1:1) mixture at 3 μL/min. For the compounds that precipitated from the reaction medium, the yields were calculated from the amount obtained after filtration and are lower than the actual yields, since in most cases a considerable amount remained in solution.

N⁶-Methyl-9-(β-D-ribofuranosyl)-2-(1,2,3-triazol-1-yl)adenine (1)

In a pressure tube was added 23 (165 mg, 0.51 mmol), trimethylsilylacetylene (292 μL, 2.05 mmol), and 4 mL of DMF. The mixture was stirred at 105 °C for 15 h. After solvent evaporation, the yellowish residue was dissolved in 2 mL of a 1.0 M solution of tetrabutylammonium fluoride in THF and stirred for 5 h. The reaction was monitored by NMR. After evaporation of the solvent, the residue was dissolved in ethyl acetate. Water was added, and the triazole product was precipitated in the water layer. After overnight cooling and filtration, the precipitate was further purified on a silica gel column (CH₂Cl₂/MeOH, 92:8) and yielded compound 1 as a white solid (82 mg, 46%). ¹H NMR (300 MHz, DMSO-d₆): δ 3.06 (d, 3H, J = 4.5 Hz, N⁶-CH₃), 3.54–3.61 (m, 1H, H-5′A), 3.65–3.72 (m, 1H, H-5′B), 3.96 (dd, 1H, J = 3.8 and 7.9 Hz, H-4′), 4.20 (dd, 1H, J = 4.7 and 8.2 Hz, H-3′), 4.65 (app q, J = 5.9 Hz, H-2′), 4.98 (t, 1H, J = 5.6 Hz, 5′-OH), 5.23 (d, 1H, J = 5.0 Hz, 3′-OH), 5.48 (d, 1H, J = 6.2 Hz, 2′-OH), 5.95 (d, 1H, J = 5.9 Hz, H-1′), 7.92 (d, 1H, J = 1.2 Hz, H-4′′), 8.37 (d, 1H, J = 4.6 Hz, N⁶-H), 8.46 (s, 1H, H-8), 8.82 (br s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 27.84 (N⁶-CH₃), 62.19 (C-5′), 71.13 (C-3′), 74.30 (C-2′), 86.42 (C-4′), 88.01 (C-1′), 119.75 (C-5), 124.67 (C-5′′), 134.25 (C-4′′), 141.13 (C-8), 149.56 and 149.85 (C-2 and C-4), 156.082 (C-6). HRMS (ESI-MS) C₁₃H₁₇N₈O₄ [M + H]⁺: 349.1367 found; 349.1372 calcd. Anal. (C₁₃H₁₆N₈O₄·½H₂O) C, H, N.

2-(4-Ethyl-1,2,3-triazol-1-yl)-N⁶-methyl-9-(β-D-ribofuranosyl)-adenine (2)

Compound 23 (70 mg, 0.217 mmol), sodium ascorbate (8.6 mg, 0.043 mmol), and CuSO₄·5H₂O (2.2 mg, 0.009 mmol) were suspended in 20 mL of H₂O/^tBuOH (3:1). The mixture was saturated with 1-butyne and stirred for 4 days at room temperature in a Parr apparatus. Purification on a preparative TLC plate (CH₂Cl₂/MeOH, 90:10) resulted in compound 2 as a white solid in 40% yield. ¹H NMR (300 MHz, DMSO-d₆): δ 1.27 (t, 3H, J = 7.62 Hz, CH₃), 2.75 (q, 2H, J = 7.6 Hz, CH₂), 3.05 (d, 3H, J = 4.4 Hz, N⁶-CH₃), 3.51–3.59 (m, 1H, H-5′A), 3.61–3.70 (m, 1H, H-5′B), 3.94 (dd, 1H, J = 4.0 and 7.6 Hz, H-4′), 4.15 (dd, 1H, J = 4.4 and 7.9 Hz, H-3′), 4.60 (app q, J = 5.6 Hz, H-2′), 4.97 (t, 1H, J = 5.6 Hz, 5′-OH), 5.22 (d, 1H, J = 4.7 Hz, 3′-OH), 5.47 (d, 1H, J = 6.2 Hz, 2′-OH), 5.93 (d, 1H, J = 6.2 Hz, H-1′), 8.34 (d, 1H, J = 4.4 Hz, N⁶-H), 8.45 (s, 1H, H-8), 8.55 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 14.32 (CH₃), 19.08 (CH₂), 27.84 (N⁶-CH₃), 62.13 (C-5′), 71.15 (C-3′), 74.34 (C-2′), 86.43 (C-4′), 87.84 (C-1′), 119.57 (C-5), 120.962 (C-5′′), 140.92 (C-8), 149.31, 149.63, 149.88 (C-2, C-4, and C-4′′), 156.05 (C-6). HRMS (ESI-MS) C₁₅H₂₁N₈O₄ [M + H]⁺: 377.1682 found; 377.1685 calcd. Anal. (C₁₅H₂₀N₈O₄) C, H, N.

General Procedure for the Synthesis of 4′′-Substituted 2-(1,2,3-Triazol-1-yl)adenosine Derivatives 3–11

Compound 23 (70 mg, 0.217 mmol), sodium ascorbate (8.6 mg, 0.043 mmol), and CuSO₄·5H₂O (2.2 mg, 0.009 mmol) were suspended in 2 mL of H₂O/^tBuOH (3:1). The appropriate alkyne (2 equiv) was subsequently added, and the mixture was stirred overnight at room temperature. The 2-triazol-1-yl compounds (generally) precipitated from this reaction medium and were isolated by filtration with water.

2-(4-Butyl-1,2,3-triazol-1-yl)-N⁶-methyl-9-(β-D-ribofuranosyl)-adenine (3)

The reaction of 23 (70 mg, 0.217 mmol) with 1-hexyne (50 μL, 0.435 mmol) gave compound 3 in 59% yield. ¹H NMR (300 MHz, DMSO-d₆): δ 0.93 (t, 3H, J = 7.3 Hz, CH₃), 1.38 (m, 2H, CH₂), 1.66 (m, 2H, CH₂), 2.71 (t, 2H, J = 7.2 Hz, C4′′-CH₂), 3.05 (d, 3H, J = 4.0 Hz, N⁶-CH₃), 3.52–3.62 (m, 1H, H-5′A), 3.62–3.72 (m, 1H, H-5′B), 3.95 (dd, 1H, J = 3.6 and 7.2 Hz, H-4′), 4.18 (dd, 1H, J = 4.8 and 8.1 Hz, H-3′), 4.62 (app q, 1H, J = 5.7 Hz, H-2′), 5.01 (t, 1H, J = 5.2 Hz, 5′-OH), 5.29 (d, 1H, J = 4.0 Hz, 3′-OH), 5.54 (d, 1H, J = 5.6 Hz, 2′-OH), 5.94 (d, 1H, J = 5.9 Hz, H-1′), 8.36 (d, 1H, J = 4.1 Hz, N⁶-H), 8.45 (s, 1H, H-8), 8.56 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 14.37 (CH₃), 22.37 (CH₂), 25.21 (CH₂), 27.83 (N⁶-CH₃), 31.68 (CH₂), 62.14 (C-5′), 71.15 (C-3′), 74.36 (C-2′), 86.43 (C-4′), 87.88 (C-1′), 119.56 (C-5), 121.33 (C-5′′), 140.96 (C-8), 147.88 and 149.90 (C-2, C-4 and C-4′′), 156.07 (C-6). HRMS (ESI-MS) C₁₇H₂₅N₈O₄ [M + H]⁺: 405.1992 found, 405.1998 calcd. Anal. (C₁₇H₂₄N₈O₄) C, H, N.

2-[4-(2-Hydroxyethyl)-1,2,3-triazol-1-yl]-N⁶-methyl-9-(β-D-ribofuranosyl) adenine (4)

The reaction of 23 (70 mg, 0.217 mmol) with 3-butyn-1-ol (33 μL, 0.435 mmol) afforded compound 4 without precipitation. After solvent evaporation, the mixture was purified on a silica gel column (90:10 CH₂Cl₂/MeOH + 1% 7 N NH₃ in MeOH), yielding compound 11 as a white solid in 68% yield. ¹H NMR (300 MHz, DMSO-d₆): δ 2.85 (t, 2H, J = 6.9 Hz, CH₂), 3.03 (d, 3H, J = 4.7 Hz, N⁶-CH₃), 3.52–3.59 (m, 1H, H-5′A), 3.63–3.72 (m, 3H, H-5′B and CH₂), 3.94 (dd, 1H, J = 3.5 and 7.3 Hz, H-4′), 4.16 (dd, 1H, J = 4.8 and 8.1 Hz, H-3′), 4.59 (app q, 1H, J = 5.4 Hz, H-2′), 4.74 (t, 1H, J = 5.8 Hz, CH₂–OH), 4.97 (app t, 1H, J = 5.6 Hz, 5′-OH), 5.24 (d, 1H, J = 5.0 Hz, 3′-OH), 5.50 (d, 1H, J = 6.2 Hz, 2′-OH), 5.93 (d, 1H, J = 6.2 Hz, H-1′), 8.31 (d, 1H, J = 4.7 Hz, N⁶-H), 8.43 (s, 1H, H-8), 8.54 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 27.83 (N⁶-CH₃), 29.70 (CH₂), 60.85 (CH₂–OH), 62.13 (C-5′), 71.15 (C-3′), 74.35 (C-2′), 86.43 (C-4′), 87.81 (C-1′), 119.56 (C-5), 122.02 (C-5′′), 140.49 (C-8), 145.49, 149.88, and 149.65 (C-2, C-4, and C-4′′), 156.05 (C-6). HRMS (ESI-MS) C₁₅H₂₁N₈O₅ [M + H]⁺: 393.1630 found, 393.1634 calcd. Anal. (C₁₅H₂₀N₈O₅·H₂O) C, H, N.

2-(4-Dimethylaminomethyl-1,2,3-triazol-1-yl)-N⁶-methyl-9-(β-D-ribofuranosyl)adenine (5)

The reaction of 23 (70 mg, 0.217 mmol) with 1-dimethylamino-2-propyne (47 μL, 0.435 mmol) gave compound 5 without precipitation. The volatiles were removed under reduced pressure, and the residue was purified on a silica gel column (80:20 CH₂Cl₂/MeOH + 1% 7 N NH₃ in MeOH). Compound 5 was obtained as a white solid in 67% yield. ¹H NMR (300 MHz, DMSO-d₆): δ 2.20 (s, 6H, 2 × CH₃), 3.05 (d, 3H, J = 4.4 Hz, N⁶-CH₃), 3.54–3.62 (m, 3H, H-5′A and CH₂), 3.65–3.72 (m, 1H, H-5′B), 3.95 (dd, 1H, J = 4.1 and 7.6 Hz, H-4′), 4.18 (dd, 1H, J = 5.0 and 8.2 Hz, H-3′), 4.62 (app q, 1H, J = 5.4 Hz, H-2′), 4.98 (t, 1H, J = 5.6 Hz, 5′-OH), 5.23 (d, 1H, J = 5.0 Hz, 3′-OH), 5.49 (d, 1H, J = 6.2 Hz, 2′-OH), 5.95 (d, 1H, J = 6.2 Hz, H-1′), 8.38 (d, 1H, J = 4.4 Hz, N⁶-H), 8.45 (s, 1H, H-8), 8.64 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 27.85 (N⁶-CH₃), 45.18 (N(CH₃)₂), 53.85 (CH₂), 62.10 (C-5′), 71.12 (C-3′), 74.38 (C-2′), 86.41 (C-4′), 87.84 (C-1′), 119.64 (C-5), 123.328 (C-5′′), 140.97 (C-8), 144.41, 149.59, 149.79 (C-2, C-4, and C-4′′), 156.05 (C-6). HRMS (ESI-MS) C₁₆H₂₄N₉O₄ [M + H]⁺: 406.1944 found, 406.1951 calcd. Anal. (C₁₆H₂₃N₉O₄) C, H, N. N calcd, 31.09; found, 30.41.

N⁶-Methyl-2-(4-phenyl-1,2,3-triazol-1-yl)-9-(β-D-ribofuranosyl) adenine (6)

The reaction of 23 (70 mg, 0.217 mmol) with phenylacetylene (48 μL, 0.435 mmol) yielded compound 6 (54%) as a white solid. ¹H NMR (300 MHz, DMSO-d₆): δ 3.11 (d, 3H, J = 4.5 Hz, N⁶-CH₃), 3.55–3.63 (m, 1H, H-5′A), 3.67–3.74 (m, 1H, H-5′B), 3.97 (dd, J = 3.6 and 7.2 Hz, H-4′), 4.21 (dd, J = 4.8 and 8.1 Hz, H-3′), 4.66 (app q, J = 5.4 Hz, H-2′), 5.03 (app t, 1H, J = 5.6 Hz, 5′-OH), 5.29 (d, 1H, J = 5.0 Hz, 3′-OH), 5.54 (d, 1H, J = 6.2 Hz, 2′-OH), 5.99 (d, 1H, J = 5.9 Hz, H-1′), 7.4 (t, 1H, J = 7.3 Hz, Ph), 7.50 (t, 2H, J = 7.5 Hz, Ph), 8.06 (d, 2H, J = 7.3 Hz, Ph), 8.44 (d, 1H, J = 4.5 Hz, N⁶-H), 8.49 (s, 1H, H-8), 9.31 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 27.92 (N⁶-CH₃), 62.16 (C-5′), 71.17 (C-3′), 74,39 (C-2′), 86.48 (C-4′), 87.94 (C-1′), 119.85 (C-5), 120.68 (C-5′′), 126.27, 128.95, 129.61, and 130.85 (Ph), 141.082 (C-8), 147.02 and 149.77 (C-2, C-4, and C-4′′), 156.10 (C-6). HRMS (ESI-MS) C₁₉H₂₁N₈O₄ [M + H]⁺: 425.1689 found, 425.1685 calcd. Anal. (C₁₉H₂₀N₈O₄) C, H, N.

N⁶-Methyl-2-[4-pyridin-2-yl-1,2,3-triazol-1-yl]-9-(β-D-ribofuranosyl) adenine (7)

The reaction of 23 (70 mg, 0.217 mmol) with 2-ethynylpyridine (44 μL, 0.435 mmol) afforded compound 7 as a white solid in 55% yield. ¹H NMR (300 MHz, DMSO-d₆): δ 3.09 (d, 3H, J = 4.1 Hz, N⁶-CH₃), 3.57–3.66 (m, 1H, H-5′A), 3.67–3.76 (m, 1H, H-5′B), 3.97 (dd, 1H, J = 3.9 and 7.5 Hz, H-4′), 4.19 (dd, 1H, J = 4.8 and 8.1 Hz, H-3′), 4.64 (app q, 1H, J = 6.0 Hz, H-2′), 4.99 (t, 1H, J = 5.6 Hz, 5′-OH), 5.26 (d, 1H, J = 5.0 Hz, 3′-OH), 5.54 (d, 1H, J = 6.2 Hz, 2′-OH), 5.99 (d, 1H, J = 5.9 Hz, H-1′), 7.42 (m, 1H, pyridin-2-yl), 7.96 (m, 1H, pyridin-2-yl), 8.16 (d, 1H, J = 7.3 Hz, pyridin-2-yl), 8.42 (d, 1H, J = 4.1 Hz, 1H, N⁶-H), 8.49 (s, 1H, H-8), 8.68 (d, 1H, J = 4.1 Hz, pyridin-2-yl), 9.16 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 27.92 (N⁶-CH₃), 62.13 (C-5′), 71.14 (C-3′), 74.49 (C-2′), 86.47 (C-4′), 88.16 (C-1′), 119.92 (C-5), 120.53 (C-5′′), 121.07 (pyridin-2-yl), 122.25 (pyridin-2-yl), 137.79 (pyridin-2-yl), 141.28 (C-8), 148.47, 150.45 and 150.56 (C-2, C-4, C-4′′, and pyridin-2-yl), 156.28 (C-6). HRMS (ESI-MS) C₁₈H₁₉N₉O₄Na [M + Na]⁺: 448.1458 found, 448.1457 calcd. Anal. (C₁₈H₁₉N₉O₄) C, H, N.

N⁶-Methyl-2-[4-(4-propoxyphenyl)-1,2,3-triazol-1-yl]-9-(β-D-ribofuranosyl) adenine (8)

The reaction of 23 (70 mg, 0.217 mmol) with 1-eth-1-ynyl-4-propoxybenzene (57 μL, 0.435 mmol) afforded compound 8 as a white solid in 36% yield. ¹H NMR (300 MHz, DMSO-d₆): δ 0.99 (t, 3H, J = 7.3 Hz, CH₃), 1.71–1.78 (m, 2H, CH₂), 3.10 (d, 3H, J = 4.1 Hz, N⁶-CH₃), 3.54–3.62 (m, 1H, H-5′A), 3.64–3.72 (m, 1H, H-5′B), 3.98 (m, 3H, H-4′ and CH₂), 4.18 (dd, 1H, J = 4.8 and 8.1 Hz, H-3′), 4.64 (app q, 1H, J = 5.4 Hz, H-2′), 4.99 (t, 1H, J = 5, 7 Hz, 5′-OH), 5.25 (d, 1H, J = 4.1 Hz, 3′-OH), 5.50 (d, 1H, J = 5.9 Hz, 2′-OH), 5.97 (d, 1H, J = 6.2 Hz, H-1′), 7.03 (d, 2H, J = 8.8 Hz, Ph), 7.94 (d, 2H, J = 8.8 Hz, Ph), 8.36 (d, 1H, J = 4.1 Hz, 1H, N⁶-H), 8.45 (s, 1H, H-8), 9.15 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 11.08 (CH₃), 22.73 (CH₂), 27.90 (N⁶-CH₃), 62.16 (C-5′), 69.71 (OCH₂), 71.18 (C-3′), 74.38 (C-2′), 86.46 (C-4′), 87.91 (C-1′), 115.50 (Ph), 119.60 (C-5), 123.28 (C-5′′), 127.66 (Ph), 140.01 (C-8), 147.01, 149.82, 150.45 (C-4, C-2, and C-4′′), 156.10 (C-6). HRMS (ESI-MS) C₂₂H₂₇N₈O₅ [M + H]⁺: 483.2109 found, 483.2104 calcd. Anal. (C₂₂H₂₆N₈O₅) C, H, N.

N⁶-Methyl-2-(4-benzyl-1,2,3-triazol-1-yl)-9-(β-D-ribofuranosyl) adenine (9)

The reaction of 23 (70 mg, 0.217 mmol) with 3-phenyl-1-propyne (54 μL, 0.435 mmol) gave compound 9 as a white solid in 43% yield. ¹H NMR (300 MHz, DMSO-d₆): δ 3.03 (d, 3H, J = 3.8 Hz, N⁶-CH₃), 3.53–3.62 (m, 1H, H-5′A), 3.64–3.71 (m, 1H, H-5′B), 3.95 (dd, 1H, J = 3.7 and 7.2 Hz, H-4′), 4.11 (s, 2H, CH₂), 4.17 (dd, 1H, J = 4.8 and 8.1 Hz, H-3′), 4.61 (app q, 1H, J = 5.7 Hz, H-2′), 4.97 (app t, 1H, J = 5.3 Hz, 5′-OH), 5.22 (d, 1H, J = 4.7 Hz, 3′-OH), 5.47 (d, 1H, J = 5.9 Hz, 2′-OH), 5.94 (d, 1H, J = 5.9 Hz, H-1′), 7.23 (m, 1H, Ph), 7.32 (d, 4H, J = 4.4 Hz, Ph), 8.34 (d, 1H, J = 3.8 Hz, N⁶-H), 8.45 (s, 1H, H-8), 8.59 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 25.32 (N⁶-CH₃), 32.51 (CH₂), 62.11 (C-5′), 71.13 (C-3′), 74.43 (C-2′), 86.41 (C-4′), 87.96 (C-1′), 121.42 (C-5), 121.97 (C-5′′), 126.91, 129.12, 129.21, and 140.02 (Ph), 140.88 (C-8), 147.42, 149.58, and 149.81 (C-4, C-2, and C-4′′), 156.09 (C-6). HRMS (ESI-MS) C₂₀H₂₃N₈O₄ [M + H]⁺: 439.1846 found, 439.1842 calcd. Anal. (C₂₀H₂₂N₈O₄) C, H, N.

2-(4-Cyclopentylmethyl-1,2,3-triazol-1-yl)-N⁶-methyl-9-(β-D-ribofuranosyl) adenine (10)

The reaction of 23 (70 mg, 0.217 mmol) with 3-cyclopentyl-1-propyne (57 μL, 0.435 mmol) yielded compound 10 (32%) as a white solid. ¹H NMR (300 MHz, DMSO-d₆): δ 1.23–1.28 (m, 2H, cyclopentyl), 1.48–1.62 (m, 4H, cyclopentyl), 1.71–1.75 (m, 2H, cyclopentyl), 2.19–2.25 (m, 1H, cyclopentyl), 2.72 (d, 2H, J = 7.3 Hz, CH₂-cyclopentyl), 3.05 (d, 3H, J = 3.9 Hz, N⁶-CH₃), 3.53–3.61 (m, 1H, H-5′A), 3.63–3.72 (m, 1H, H-5′B), 3.96 (dd, 1H, J = 3.6 and 7.2 Hz, H-4′), 4.19 (dd, 1H, J = 4.8 and 8.1 Hz, H-3′), 4.62 (app q, 1H, J = 5.9 Hz, H-2′), 4.98 (t, 1H, J = 5.3 Hz, 5′-OH), 5.23 (d, 1H, J = 4.7 Hz, 3′-OH), 5.49 (d, 1H, J = 6.5 Hz, 2′-OH), 5.95 (d, 1H, J = 5.9 Hz, H-1′), 8.34 (d, 1H, J = 3.9 Hz, N⁶-H), 8.45 (s, 1H, H-8), 8.55 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 25.37 (cyclopentyl), 27.83 (N⁶-CH₃), 31.56 (CH₂), 32.55 (cyclopentyl), 62.15 (C-5′), 71.15 (C-3′), 74.34 (C-2′), 86.43 (C-4′), 87.87 (C-1′), 119.60 (C-5), 121.58 (C-5′′), 140.92 (C-8), 147.33, 149.89 (C-4, C-2, and C4′′), 156.07 (C-6). HRMS (ESI-MS) C₁₉H₂₇N₈O₄ [M + H]⁺: 431.2153 found, 431.2155 calcd. Anal. (C₁₉H₂₆N₈O₄) C, H, N.

2-(4-Cyclohexylmethyl-1,2,3-triazol-1-yl)-N⁶-methyl-9-(β-D-ribofuranosyl) adenine (11)

The reaction of 23 (70 mg, 0.217 mmol) with cyclohexyl-1-propyne (63 μL, 0.435 mmol) gave compound 11 in 82% yield. ¹H NMR (300 MHz, DMSO-d₆): δ 0.86–1.28 (br m, 6H, cyclohexyl), 1.54–1.72 (br m, 5H, cylcohexyl), 2.58 (d, 2H, J = 6.9 Hz, CH₂), 3.03 (d, 3H, J = 3.9 Hz, N⁶-CH₃), 3.51–3.59 (m, 1H, H-5′A), 3.63–3.70 (m, 1H, H-5′B), 3.94 (dd, 1H, J = 4.2 and 7.5 Hz, H-4′), 4.16 (dd, 1H, J = 4.8 and 8.1 Hz, H-3′), 4.59 (app q, 1H, J = 6.3 Hz, H-2′), 4.97 (t, 1H, J = 5.5 Hz, 5′-OH), 5.22 (d, 1H, J = 4.8 Hz, 3′-OH), 5.47 (d, 1H, J = 6.3 Hz, 2′-OH), 5.93 (d, 1H, J = 6.0 Hz, H-1′), 8.30 (d, 1H, J = 3.9 Hz, N⁶-H), 8.43 (s, 1H, H-8), 8.51 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 26.31 (cyclohexyl), 26.66 (cyclohexyl), 27.84 (N⁶-CH₃), 33.16 (cyclohexyl), 38.22 (CH₂), 62.14 (C-5′), 71.15 (C-3′), 74.35 (C-2′), 86.44 (C-4′), 87.86 (C-1′), 119.60 (C-5), 121.93 (C-5′′), 140.92 (C-8), 146.45, 149.92 (C-4, C-2, and C-4′′), 153.51 (C-6). HRMS (ESI-MS) C₂₀H₂₉N₈O₄ [M + H]⁺: 445.2305 found, 445.2311 calcd. Anal. (C₂₀H₂₈N₈O₄) C, H, N.

General Procedure for the Synthesis of 4′′-Substituted 2-(1,2,3-Triazol-4-yl)adenosine Derivatives 12–14

Compound 25 (100 mg, 0.32 mmol), sodium ascorbate (13 mg, 0.06 mmol mmol), and CuSO₄·5H₂O (3 mg, 0. 013 mmol) were suspended in 30 mL of H₂O/^tBuOH (3:1). The appropriate azide (2 equiv) was subsequently added, and the mixture was stirred overnight at room temperature. The 2-triazol-4-yl compounds (generally) precipitated from this reaction medium and were isolated by filtration with water.

2-(1-Benzyl-1,2,3-triazol-4-yl)-N⁶-methyl-9-(β-D-ribofuranosyl) adenine (12)

The reaction of 25 (100 mg, 0.32 mmol) with 85 mg (0.64 mmol) of benzylazide gave compound 12 in 78% yield (110 mg). ¹H NMR (300 MHz, DMSO-d₆): δ 3.03 (br s, 3H, N⁶-CH₃), 3.52–3.58 (m, 1H, H-5′A), 3.60–3.66 (m, 1H, H-5′B), 3.92 (app d, H-4′, J = 2.9 Hz, H-4′), 4.15 (dd, 1H, J = 4.7 and 7.6 Hz, H-3′), 4.60 (app q, 1H, J = 5.9 Hz, H-2′), 5.08 (t, 1H, J = 5.4 Hz, 5′-OH), 5.19 (d, 1H, J = 3.5 Hz, 3′-OH), 5.44 (d, 1H, J = 5.6 Hz, 2′-OH), 5.68 (s, 2H, CH₂), 5.97 (d, 6.2 Hz, H-1′), 7.38 (br s, 5H, Ph), 7.82 (br s, 1H, N⁶-H), 8.36 (s, 1H, H-8), 8.66 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 27.54 (N⁶-CH₃), 53.62 (CH₂), 62.29 (C-5′), 71.30 (C-3′), 74.30 (C-2′), 86.41 (C-4′), 87.77 (C-1′), 119.52 (C-5), 126.41 (C-5′′), 128.63, 128.86, 129.48 and 136.71 (Ph), 140.27 (C-8), 148.22, 153.56, 153.72 (C-2, C-4, and C-4′), 155.69 (C-6). HRMS (ESI-MS) C₂₀H₂₃N₈O₄ [M + H]⁺: 439.1834 found, 439.1842 calcd. Anal. (C₂₀H₂₂N₈O₄) C, H, N.

2-[1-(3-Methoxybenzyl)-1,2,3-triazol-4-yl)-N⁶-methyl-9-(β-D-ribofuranosyl) adenine (13)

The reaction of 25 (100 mg, 0.32 mmol) with 104 mg (0.64 mmol) of 3-methoxybenzylazide gave compound 13 in 80% yield (120 mg). ¹H NMR (300 MHz, DMSO-d₆): δ 3.03 (br s, 3H, N⁶-CH₃), 3.52–3.60 (m, 1H, H-5′A), 3.64–3.72 (m, 1H, H-5′B), 3.95 (app d, H-4′, J = 2.9 Hz, H-4′), 4.15 (dd, 1H, J = 4.7 and 7.6 Hz, H-3′), 4.60 (app q, 1H, J = 6.4 Hz, H-2′), 5.07 (t, 1H, J = 5.1 Hz, 5′-OH), 5.21 (d, 1H, J = 4.2 Hz, 3′-OH), 5.46 (d, 1H, J = 6.3 Hz, 2′-OH), 5.62 (s, 2H, CH₂), 5.95 (d, 1H, J = 6.6 Hz, H-1′), 6.91 (m, 3H, Ph), 7.29 (t, 1H, J = 7.9 Hz, Ph), 7.79 (br s, 1H, N⁶-H), 8.34 (s, 1H, H-8), 8.63 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 27.59 (N⁶-CH₃), 53.45 (CH₂), 55.82 (OCH₃), 62.30 (C-5′), 71.33 (C-3′), 74.25 (C-2′), 86.42 (C-4′), 87.70 (C-1′), 114.44 and 114.20 (Ph), 119.51 (C-5), 120.72 (Ph), 130.65 (C-5′′), 138.21 (C-8), 148.22, 153.56, 153.64 (C-2, C-4, and C-4′), 155.69 (C-6), 160.14 (Ph). HRMS (ESI-MS) C₂₁H₂₅N₈O₅ [M + H]⁺: 469.1938 found, 469.1947 calcd. Anal. (C₂₁H₂₄N₈O₅) C, H, N.

2-[1-(3-Chlorobenzyl)-1,2,3-triazol-4-yl)-N⁶-methyl-9-(β-D-ribofuranosyl) adenine (14)

The reaction of 25 (100 mg, 0.32 mmol) with 107 mg (0.64 mmol) of 3-chlorobenzylazide gave compound 14 in 73% yield (110 mg). ¹H NMR (300 MHz, DMSO-d₆): δ 3.04 (br s, 3H, N⁶-CH₃), 3.53–3.60 (m, 1H, H-5′A), 3.64–3.71 (m, 1H, H-5′B), 3.96 (app d, H-4′, J = 2.9 Hz, H-4′), 4.17 (dd, 1H, J = 4.7 and 7.6 Hz, H-3′), 4.64 (app q, 1H, J = 5.8 Hz, H-2′), 5.09 (t, 1H, J = 5.3 Hz, 5′-OH), 5.20 (d, 1H, J = 4.4 Hz, 3′-OH), 5.45 (d, 1H, J = 6.2 Hz, 2′-OH), 5.97 (d, 1H, J = 6.2 Hz, H-1′), 7.31–7.36 (m, 1H, Ph), 4.41–7.47 (m, 3H, Ph), 7.83 (br s, 1H, N⁶-H), 8.37 (s, 1H, H-8), 8.72 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 27.54 (N⁶-CH₃), 52.80 (CH₂), 62.31 (C-5′), 71.33 (C-3′), 74.27 (C-2′), 86.43 (C-4′), 87.69 (C-1′), 119.56 (C-5), 126.63 (C-5′′), 127.53, 128.53, 128.84, 131.42, 133.98, and 139.15 (Ph), 140.36 (C-8), 148.22, 150.03, and 153.43 (C-2, C-4, and C-4′), 155.78 (C-6). HRMS (ESI-MS) C₂₀H₂₂N₈O₄Cl [M + H]⁺: 473.1452 found, 473.1452 calcd. Anal. (C₂₀H₂₁N₈O₄Cl) C, H, N.

9-(5-Ethylcarbamoyl-β-D-ribofuranosyl)-N⁶-methyl-2-(1,2,3-triazol-1-yl)adenine (15)

In a pressure tube was added 31 (110 mg, 0.30 mmol), trimethylsilylacetylene (259 μL, 1.81 mmol), and 4 mL of DMF. The mixture was stirred at 105 °C for 15 h. Solvent evaporation yielded a yellowish solid that was dissolved 6 mL of a 1.0 solution of tetrabutylammonium fluoride in THF and stirred for 5 h. After solvent evaporation, the residue was dissolved in ethyl acetate. Water was added, and the triazole precipitated in the water layer. After overnight cooling and filtration, the precipitate was further purified on a silica gel column (CH₂Cl₂/MeOH, 93:7) and yielded compound 15 as a white solid (49 mg, 42%). ¹H NMR (300 MHz, DMSO-d₆): δ 0.90 (t, 3H, CH₃), 3.05–3.21 (m, 2H, N-CH₂), 3.05 (d, 3H, J = 4.2 Hz, N⁶-CH₃), 4.26 (m, 1H, H-3′), 4.33 (d, 1H, J = 2.1 Hz, H-4′), 5.61 (d, 1H, J = 6.2 Hz, 3′-OH), 5.71 (d, 1H, J = 4.7 Hz, 2′-OH), 6.04 (d, 1H, J = 7.2 Hz, H-1′), 7.91 (d, 1H, J = 1.2 Hz, H-4′′), 8.07 (t, 1H, J = 5.3 Hz, NHCO), 8.41 (d, 1H, J = 4.2 Hz, N⁶-H), 8.54 (s, 1H, H-8), 8.82 (br s, 1H, H-5′′). ¹³CMR (300 MHz, DMSO-d₆): δ 15.16 (CH₃), 27.85 (N⁶-CH₃), 34.07 (CH₂), 73.64 (C-2′ and C-3′), 84.98 (C-4′), 88.02 (C-1′), 119.91 (C-5), 124.70 (C-5′′), 134.29 (C-4′′), 141.59 (C-8), 149.69 and 149.90 (C-2 and C-4), 156.14 (C-6), 169.73 (C═O). HRMS (ESI-MS) C₁₅H₂₀N₉O₄ [M + H]⁺: 390.1676 found, 390.1683 calcd. Anal. (C₁₅H₁₉N₉O₄) C, H, N.

General Procedure for the Synthesis of 4′′-Substituted 2-(1,2,3-Triazol-1-yl)adenosine Derivatives 16–18

To a mixture of 31 (100 mg, 0.28 mmol), CuI (5 mg, 0.03 mmol), and triethylamine (40 μL, 0.28 mmol) in water/acetonitrile (1:1), the appropriate alkyne (2 equiv) was added. The mixture was stirred for 5 days at room temperature. The reaction was monitored by ¹H NMR. The product was precipitated with water and cooled overnight. After filtration, the yellowish solid was purified on a silica gel column (CH₂Cl₂/MeOH, 90:10) to obtain the 1,2,3-triazol-1-yladenosine derivative as a white solid.

2-(4-Butyl-1,2,3-triazol-1-yl)-9-(5-ethylcarbamoyl-β-D-ribofuranosyl)-N⁶-methyladenine (16)

The reaction of compound 31 (100 mg, 0.28 mmol) with 1-hexyne (64μL, 0.56 mmol) gave 65 mg (53%) of compound 16 as a white solid. ¹H NMR (300 MHz, DMSO-d₆) δ 0.88–0.95 (m, 6H, 2 × CH₃), 1.31–1.43 (m, 2H, CH₂), 1.61–1.71 (m, 2H, CH₂), 2.72 (t, 2H, J = 7.8 Hz, C4′′-CH₂), 3.05–3.22 (m, 2H, N-CH₂), 3.05 (d, 3H, J = 4.1 Hz, N⁶-CH₃), 4.25 (m, 1H, H-3′), 4.33 (d, 1H, J = 2.1 Hz, H-4′), 4.73 (m, 1H, H-2′), 5.59 (d, 1H, J = 6.3 Hz, 3′-OH), 5.69 (d, 1H, J = 4.5 Hz, 2′-OH), 6.04 (d, 1H, J = 6.9 Hz, H-1′), 8.09 (t, 1H, J = 5.4 Hz), 8.37 (d, 1H, J = 4.1 Hz, N⁶-H), 8.53 (s, 1H, H-8), 8.56 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): 14.36 (CH₃), 15.17 (CH₃), 22.35 (CH₂), 25.21 (CH₂), 27.84 (N⁶-CH₃), 31.70 (CH₂), 34.07 (CH₂), 73.60 and 73.64 (C-2′ and C-3′), 84.99 (C-4′′), 87.91 (C-1′), 119.79 (C-5), 121.45 (C-5′′), 141.49 (C-8), 147.97, 149.72, and 149.93 (C-2, C-4, and C-4′′), 156.11 (C-6), 169.73 (C═O). HRMS (ESI-MS) C₁₉H₂₈N₉O₄ [M + H]⁺: 446.2256 found, 446.2264 calcd. Anal. (C₁₉H₂₇N₉O₄) C, H, N.

9-(5-Ethylcarbamoyl-β-D-ribofuranosyl)-N⁶-methyl-2-(4-pyridin-2-yl-1,2,3-triazol-1-yl)adenine (17)

The reaction of compound 31 (50 mg, 0.14 mmol) with 2-ethynylpyridine (56 μL, 0.56 mmol) gave 35 mg (54%) of compound 17 as a white solid. ¹H NMR (300 MHz, DMSO-d₆): δ 0.93 (t, 3H, J = 7.2 Hz, CH₃), 3.09–3.21 (m, 2H, N-CH₂), 3.10 (d, 3H, J = 4.2 Hz, N⁶-CH₃), 4.25 (m, 1H, H-3′), 4.35 (d, 1H, J = 2.1 Hz, H-4′), 4.74 (m, 1H, H-2′), 5.68 (d, 1H, J = 6.6 Hz, 2′-OH), 5.75 (d, 1H, J = 4.5 Hz, 3′-OH), 6.08 (d, 1H, J = 6.6 Hz, H-1′), 7.41 (m, 1H, pyridin-2-yl), 7.96 (m, 1H, pyridin-2-yl), 8.10 (t, 1H, J = 6.0 Hz, NHCO), 8.16 (d, 1H, J = 8.1 Hz, pyridin-2-yl), 8.48 (d, 1H, J = 4.1 Hz, 1H, N⁶-H), 8.60 (s, 1H, H-8), 8.67 (d, H, J = 5.1 Hz, pyridin-2-yl), 9.17 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 15.18 (CH₃), 27.94 (N⁶-CH₃), 34.11 (N-CH₂), 73.69 and 73.78 (C-2′ and C-3′), 84.95 (C-4′), 87.98 (C-1′), 120.04 (C-5), 120.65 (C-5′′), 122.20 and 124.11 (pyridin-2-yl), 138.027 (pyridin-2-yl), 141.64 (C-8), 147.95, 149,64, 149.67, and 149.98 (C-2, C-4, C-4′′, and pyridin-2-yl), 150.49 (pyridin-2-yl), 156.13 (C-6), 169.76 (C═O). HRMS (ESI-MS) C₂₀H₂₃N₁₀O₄ [M + H]⁺: 467.1899 found, 467.1903 calcd. Anal. (C₂₀H₂₂N₁₀O₄) C, H, N. N calcd, 30.03; found, 29.55.

9-(5-Ethylcarbamoyl-β-D-ribofuranosyl)-N⁶-methyl-2-(4-benzyl-1,2,3-triazol-1-yl)adenine (18)

The reaction of compound 31 (70 mg, 0.19 mmol) with 3-phenyl-1-propyne (49μL, 0.56 mmol) gave 35 mg (38%) of compound 18 as a white solid. ¹H NMR (300 MHz, DMSO-d₆): δ 0.87 (t, 3H, J = 7.5 Hz, CH₃), 3.03–3.20 (m, 2H, N-CH₂), 3.03 (d, 3H, 4.5 Hz, N⁶-CH₃), 4.11 (s, 2H, CH–Ph), 4.24 (m, 1H, H-3′), 4.32 (d, 1H, J = 2.1 Hz, H4′), 4.72 (m, 1H, H-2′), 5.59 (d, 1H, J = 6.6 Hz, 3′-OH), 5.69 (d, 1H, J = 4.5 Hz, 2′-OH), 6.03 (d, 1H, J = 6.9 Hz, H1′), 7.22 (m, 1H, Ph), 7.32 (d, 4H, J = 4.2 Hz, Ph), 8.06 (t, 1H, J = 5.7 Hz, NHCO), 8.38 (d, 1H, J = 4.1 Hz, N⁶-H), 8.54 (s, 1H, H-8), 8.60 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 15.16 (CH₃), 27.84 (N⁶-CH₃), 31.72 (CH₂), 34.07 (N-CH₂), 73.64 (C-2′ and C-3′), 84.98 (C-4′), 87.87 (C-1′), 119.91 (C-5), 122.11 (C-5′′), 126.93, 129.13, 129.22, and 140.07 (Ph), 140.48 (C-8), 147.05, 149.87 (C-4, C-2, and C-4′′), 156.12 (C-6), 169.71 (C═O). HRMS (ESI-MS) C₂₂H₂₆N₉O₄ [M + H]⁺: 480.2098 found, 480.2107 calcd. Anal. (C₂₂H₂₅N₉O₄) C, H, N.

2-(4-Cyclopentylmethyl-1,2,3-triazol-1-yl)-9-(5-ethylcarbamoyl-β-D-ribofuranosyl)-N⁶-methyladenine (19)

Compound 31 (60 mg, 0.17 mmol), sodium ascorbate (13 mg, 0.066 mmol) and CuSO₄·5H₂O (3.5 mg, 0.013 mmol) were suspended in 4 mL of ^tBuOH/H₂O (1:1). 3-Cyclopentyl-1-propyne (58 μL, 0.44 mmol) was subsequently added, and the mixture was stirred for 2 days at room temperature. The 2-triazol-1-yl compound precipitated from the reaction medium. Water was added, and the mixture was cooled overnight. The precipitate was filtered off and washed with water and hexane to obtain 19 as a white solid in 33% yield. ¹H NMR (300 MHz, DMSO-d₆): δ 0.89 (t, 3H, CH₃), 1.19–1.28 (m, 2H, cyclopentyl), 1.45–1.75 (m, 6H, cyclopentyl), 2.15–2.25 (m, 1H, CH, cyclopentyl), 2.72 (d, 2H, J = 7.2 Hz, CH₂), 3.06 (d, 3H, J = 4.5 Hz, N⁶-CH₃), 3.09–3.22 (m, 2H, N-CH₂), 4.24 (dt, 1H, J = 1.5 and 4.8 Hz, H-3′), 4.33 (d, 1H, J = 2.1 Hz, H-4′), 4.71–7.76 (app q, 1H, J = 6.6 Hz, H-2′), 5.61 (d, 1H, J = 6.3 Hz, 2′-OH), 5.71 (d, 1H, J = 4.8 Hz, 3′-OH), 6.03 (d, 1H, J = 6.9 Hz, H-1′), 8.10 (t, 1H, J = 5.7 Hz, NHCO), 8.40 (d, 1H, J = 5.1 Hz, N⁶-H), 8.54 (s, 1H, H-8), 8.56 (s, 1H, H-5′′). ¹³C NMR (300 MHz, DMSO-d₆): δ 15.17 (CH₃), 25.36 (cyclopentyl), 27.85 (N⁶-CH₃), 31.53 (cyclopentyl), 31.51 (CH₂), 34.07 (CH₂), 73.63 and 73.56 (C-2′ and C-3′), 84.99 (C-4′), 87.93 (C-1′), 119.77 (C-5), 121.69 (C-5′′), 141.51 (C-8), 147.48 and 149.92 (C-2, C-4, and C-4′′), 156.10 (C-6), 169.74 (C═O). HRMS (ESI-MS) C₂₁H₃₀N₉O₄ [M + H]⁺: 472.2415 found, 472.2420 calcd. Anal. (C₂₁H₂₉N₉O₄) C, H, N.

N⁶-(5-Chloro-2-methoxybenzyl)-2-(4-cyclopentylmethyl-1,2,3-triazol-1-yl)-9-(β-D-ribofuranosyl)adenine (20)

Compound 33 (100 mg, 0.22 mmol), sodium ascorbate (17 mg, 0.086 mmol), and CuSO₄·5H₂O (3.5 mg, 0.017 mmol) were suspended in 4 mL of H₂O/^tBuOH (1:1). 3-Cyclopentyl-1-propyne (29 μL, 0.44 mmol) was subsequently added, and the mixture was stirred for 2 days at room temperature. The 2-triazol-1-yl compound precipitated from the reaction medium. Water was added, and the mixture was cooled overnight. The precipitate was filtered off and washed with water and hexane to obtain 20 as a white solid in 59% yield. ¹H NMR (300 MHz, DMSO-d₆): δ 1.23–1.28 (m, 2H, cyclopentyl), 1.48–1.61 (m, 4H, cyclopentyl), 1.66–1.73 (m, 2H, cyclopentyl), 2.15–2.25 (m, 1H, cyclopentyl), 2.72 (d, 2H, J = 7.2 Hz, CH₂-cyclopentyl), 3.56–3.61 (m, 1H, H-5′A), 3.66–3.71 (m, 1H, H-5′B), 3.86 (s, 3H, OCH₃), 3.97 (m, 1H, H-4′), 4.20 (m, 1H, H-3′), 4.64 (m, 1H, H-2′), 4.73 (br s, 2H, N⁶-CH₂), 4.96 (t, 1H, J = 6.0 Hz, 5′-OH), 5.22 (d, 1H, J = 4.8 Hz, 3′-OH), 5.48 (d, 1H, J = 5.7 Hz, 2′-OH), 5.95 (d, 1H, J = 6.3 Hz, H-1′), 7.04 (d, 1H, J = 9.0 Hz, Ph), 7.25–7.29 (m, 2H, Ph), 8.40 (s, 1H, H-8), 8.81 (s, 1H, H-5′′), 8.81 (br s, 1H, N⁶-H). ¹³C NMR (300 MHz, DMSO-d₆): δ 25.36 (cyclopentyl), 31.54 (CH₂), 32.54 (cyclopentyl), 38.86 (CH₂), 56.57 (OCH₃), 62.13 (C-5′), 71.15 (C-3′), 74.37 (C-2′), 86.46 (C-4′), 88.07 (C-1′), 113.15 (Ph), 119.58 (C-5), 121.37 (C-5′′), 124.66, 128.23, and 129.93 (Ph), 141.37 (C-8), 147.40, 149.66, and 150.12 (C-2, C-4, and C-4′′), 155.51 (Ph), 156.33 (C-6). HRMS (ESI-MS) C₂₆H₃₂N₈O₅Cl [M + H]⁺: 571.2184 found, 571.2184 calcd. Anal. (C₂₆H₃₁N₈O₅Cl) C, H, N.

2-Azido-N⁶-methyl-9-(β-D-ribofuranosyl)adenine (23)

Sodium ascorbate (19.4 mg, 0.098 mmol) and CuSO₄·5H₂O (12.2 mg, 0.049 mmol) were added to a mixture of 22 (200 mg, 0.491 mmol), sodium azide (38,3 mg, 0.589 mmol), L-proline (11,3 mg, 0.098 mmol), and sodium carbonate (10.4 mg, 0.098 mmol) in 10 mL of H₂O/^tBuOH (1:1). The mixture was stirred overnight at 65 °C and was monitored by ¹H NMR. Then 50 mL of dilute NH₄OH was added and the crude mixture extracted with ethyl acetate (3 × 60 mL). The organic layer was washed with brine (60 mL), dried over MgSO₄, and purified on a silica gel column (CH₂Cl₂/MeOH, 95:5) to afford compound 23 as a slightly yellow solid in 66% yield. ¹H NMR (300 MHz, DMSO-d₆): δ 2.91 (d, 3H, J = 4.4 Hz, N⁶-CH₃), 3.48–3.55 (m, 1H, H-5′A), 3.58–3.66 (m, 1H, H-5′B), 3.90 (dd, 1H, J = 3.8 and 7.3 Hz, H-4′), 4.10 (dd, 1H, J = 4.7 and 9.7 Hz, H3′), 4.53 (app q, 1H, J = 5.9 Hz, H-2′), 5.04 (dd, 1H, J = 5.2 and 6.2 Hz, 5′-OH), 5.19 (d, 1H, J = 5.0 Hz, 3′OH), 5.43 (d, 1H, J = 6.2 Hz, 2′-OH), 5.78 (d, 1H, J = 6.2 Hz, H-1′), 8.12 (d, J = 4.4 Hz, N⁶-H), 8.27 (s, 1H, H-8). Small peaks from 1/6 tetrazole tautomeric form: δ 3.15 (d, 3H, J = 5.0 Hz, N⁶-CH₃), 3.95 (d, H-4′), 4.15 (d, H-3′), 5.51 (d, 2′-OH), 5.94 (d, H-1′), 8.51 (s, H-8). ¹³C NMR (300 MHz, DMSO-d₆): δ 27.54 (N⁶-CH₃), 62.21 (C-5′), 71.17 (C-3′), 74.12 (C-2′), 86.35 (C-4′), 88.01 (C-1′), 118.07 (C-5), 139.99 (C-8), 156.06 and 156.20 (C-2 and C-6). Small peaks from 1/6 tetrazole tautomeric form: δ 31.89 (N⁶-CH₃), 61.79 (C-5′), 70.77 (C-3′), 74.37 (C-4′), 112.30 (C-12), 142.91 (C-11), 147.60 (C-6). HRMS (ESI-MS) C₁₁H₁₅N₈O₄ [M + H]⁺: 323.1208 found, 323.1216 calcd. Anal. (C₁₁H₁₄N₈O₄) C, H, N.

N⁶-Methyl-9-(β-D-ribofuranosyl)-2-[2-trimethylsilylethyn-1-yl]adenine (24)

Compound 22 (500 mg, 1.23 mmol), CuI (12 mg, 0.062 mmol), and (Ph₃P)₃PCl₂ were dissolved in 9 mL of DMF. Triethylamine (205 μL, 1.47 mmol) and trimethylsilylacetylene (210 mg, 1.47 mmol) were added, and the reaction mixture was stirred overnight. After solvent evaporation, the residue was dissolved in CH₂Cl₂ and filtered through a pad of Celite. Purification on a silica gel column (CH₂Cl₂/MeOH, 95:5) yielded 305 mg (66%) of compound 23. ¹H NMR (300 MHz, DMSO-d₆): δ 0.00 (9H, s, (CH₃)₃Si), 2.7 (3H, N⁶-CH₃), 3.26–3.34 (m, 1H, H-5′A), 3.37–3.44 (m, 1H, H-5′B), 3.68 (dd, 1H, J = 3.5 Hz, H-4′), 3.86 (dd, 1H, J = 3.4 and 8.2 Hz, H-3′), 4.23 (app q, 1H, J = 5.9 Hz, H-2′), 4.87 (t, 1H, J = 5.0, 5′-OH), 4.94 (d, 1H, J = 4.99 Hz, 3′-OH), 5.21 (d, 1H, J = 6.2 Hz, 2′-OH), 5.63 (d, 1H, J = 6.2 Hz, H-1′), 7.69 (br s, 1H, N⁶-H), 8.19 (s, 1H, H-8). HRMS (ESI-MS) C₁₆H₂₄N₅O₄Si: [M + H]⁺: 378.1586 found; 378.1597 calcd.

2-Ethynyl-N⁶-methyl-9-(β-D-ribofuranosyl)purine (25)

An amount of 300 mg (0.8 mmol) of compound 24 was dissolved in 7 N ammonia in methanol and stirred for 2 h at 0 °C. After solvent evaporation, the residue was purified by silica gel chromatography (CH₂Cl₂/MeOH, 95:5) to obtain 160 mg (65%) of derivative 25. ¹H NMR (300 MHz, DMSO-d₆): δ 2.91 (s, 3H, N⁶-CH₃), 3.48–3.56 (m, 1H, H-5′A), 3.61–3.68 (m, 1H, H-5′B), 3.92 (m, 1H, H-4′), 4.02 (s, 1H, C≡CH), 4.11 (dd, 1H, J = 5.0 and 8.2 Hz, H-3′), 4.53 (app q, 1H, J = 5.9 Hz, H-2′), 5.15 (m, 2H, 3′-OH and 5′-OH), 5.44 (d, 1H, J = 6.15 Hz, 2′-OH), 5.84 (d, 1H, J = 6.16 Hz, H-1′), 7.95 (br s, 1H, N⁶-H), 8.40 (s, 1H, H-8). HRMS (ESI-MS) C₁₃H₁₆N₅O₄ [ M + H]⁺: 306.1197 found; 306.1202 calcd.

1-Deoxy-1-(6-methylamino-2-iodo-9H-purin-9-yl)-2,3-O-isopropylidene-β-D-ribufuranuronic Acid (27)

To a stirred solution of 3.8 g (8.5 mmol) of 26 in 560 mL of H₂O were added 1.4 g of KOH and, dropwise, a solution of 4.03 g (25.5 mmol) of KMnO₄ in 110 mL of H₂O. The mixture was stirred in the dark for 20 h, cooled to 0 °C, and quenched with 30 mL of 7% H₂O₂. The mixture was filtered through Celite. The filtrate was concentrated in vacuo and then acidified to pH 4 with 3 N HCl. The resulting precipitate was filtered off and successively washed with water and ether to give 2.98 g (76%) of 27 as a white solid. ¹H NMR (300 MHz, DMSO-d₆): δ 1.36 (s, 3H, CH₃), 1.51 (s, 3H, CH₃), 2.89 (d, 3H, J = 3.3 Hz, N⁶-CH₃), 4.68 (d, 1H, J = 1.8 Hz, H-4′), 5.40 (d, 1H, J = 6.0 Hz, H-2′), 5.47 (dd, 1H, J = 6.0 and 1.8 Hz, H-3′), 6.28 (s, 1H, H-1′), 8.08 (d, 1H, J = 3.3 Hz, N⁶-H), 8.16 (s, 1H, H-8). HRMS (ESI-MS) C₁₄H₁₆N₅O₅I [M + H]⁺: 462.0273 found; 462.0276 calcd.

9-(5-Ethylcarbamoyl-β-D-ribofuranosyl)-2-iodo-N⁶-methyladenine (30)

p-Nitrophenol (402 mg, 2.89 mmol) and 1-[3-(dimethylamino) propyl]-3-ethylcarbodiimide hydrochloride (506 mg, 2.65 mmol) were added to a solution of 27 (1.11 g, 2.41 mmol) in 10 mL of dry DMF. The reaction mixture was stirred for 3 h at room temperature and cooled to 0 °C, and 1.6 mL (24.1 mmol) of ethylamine was added. The solution turned yellow immediately and was further stirred for 1 h at room temperature. After evaporation of the volatiles, the residue was partitioned between ethyl acetate (3 × 100 mL) and H₂O (100 mL). The organic layer was washed with brine (100 mL), dried over MgSO₄, and concentrated to dryness. The residue was dissolved in 80% aqueous TFA (20 mL) and stirred for 2 h at room temperature. The mixture was concentrated in vacuo, coevaporated several times with EtOH, and purified by silica gel chromatography (CH₂Cl₂/MeOH, 96:4). Compound 30 was obtained as a white solid in 78% yield (840 mg). ¹H NMR (300 MHz, DMSO-d₆): δ 1.05 (t, 3H, J = 7.2 Hz, CH₃), 2.91 (d, 3H, J = 4.4 Hz, N⁶-CH₃), 3.19–3.29 (m, 2H, N-CH₂), 4.16 (dt, 1H, J = 2.1 and 4.4 Hz, H-3′), 4.31 (d, 1H, J = 2.1 Hz, H-4′), 4.58 (app q, 1H, J = 5.6 Hz, H-2′), 5.59 (d, 1H, J = 6.5 Hz, 2′-OH), 5.71 (d, 1H, J = 4.7 Hz, 3′-OH), 5.92 (d, 1H, J = 7.1 Hz, H-1′), 8.12 (t, 1H, J = 5.5 Hz, NHCO), 8.19 (d, 1H, J = 4.4 Hz, N⁶-H), 8.38 (s, 1H, H-8). HRMS (ESI-MS) C₁₃H₁₈N₆O₄I [M + H]⁺: 449.0429 found, 449.0436 calcd.

2-Azido-9-(5-ethylcarbamoyl-β-D-ribofuranosyl)-N⁶-methyladenine (31)

Sodium ascorbate (69 mg, 0.34 mmol) and CuSO₄· 5H₂O (5.6 mg, 0.17 mmol) were added to a mixture of 30 (780 mg, 1.74 mmol), sodium azide (226 mg, 3.48 mmol), L-proline (40 mg, 0.35 mmol), and sodium carbonate (37 mg, 0.35 mmol) in 20 mL of H₂O/^tBuOH (1:1). The mixture was stirred for 2 days at 65 °C and monitored by ¹H NMR. An amount of 100 mL of dilute NH₄OH was added, and the crude mixture was extracted with ethyl acetate (5 × 150 mL) and washed with brine (150 mL). The organic layer was dried over MgSO₄ and purified on a silica gel column (CH₂Cl₂/MeOH, 96:4) to afford compound 31 as a white solid in 79% yield (500 mg). ¹H NMR (300 MHz, DMSO-d₆): δ 1.06 (t, 3H, J = 7.2 Hz, CH₃), 2.96 (d, 3H, J = 4.0 Hz, N⁶-CH₃), 3.16–3.29 (m, 2H, N-CH₂), 4.14 (dt, 1H, J = 1.8 and 4.8 Hz, H-3′), 4.29 (d, 1h, J = 1.8, H-4′), 4.53–4.59 (app q, 1H, J = 6.3 Hz, H-2′), 5.52 (d, 1H, J = 6.3 Hz, 2′-OH), 5.68 (d, 1H, J = 4.8 Hz, 3′-OH), 5.90 (d, 1H, J = 7.2 Hz, H-1′), 8.20 (d, 1H, J = 4.0 Hz, N⁶-H), 8.35 (s, 1H, H-8), 8.49 (t, 1H, J = 6.0 Hz, NHCO). Small peaks from 1/5 tetrazole tautomeric form: δ 1.08–1.12 (t, 3H, J = 7.2 Hz, CH₃), 4.34 (d, 1H, J = 2.1 Hz, H-4′), 4.67–4.73 (app q, 1H, J = 7.2 Hz, H-2′), 5.54 (d, 1H, J = 4.5 Hz, 2′-OH) 6.00 (d, 1H, J = 7.2 Hz, H-1′). ¹³C NMR (300 MHz, DMSO-d₆): δ 15.48 (CH₃), 27.52 (N⁶-CH₃), 33.94 (N-CH₂), 73.06 and 73.75 (C-2′ and C-3′), 85.15 (C-4′), 88.09 (C-1′), 118.36 (C-5), 140.69 (C-8), 156.032 and 156.143 (C-2 and C-4), 169.79 (C-6). HRMS (ESI-MS) C₁₃H₁₈N₉O₄ [M + H]⁺: 364.1473 found; 364.1481 calcd. Anal. (C₁₃H₁₇N₉O₄) C, H, N.

N⁶-(5-Chloro-2-methoxybenzyl)-2-iodo-9-(β-D-ribofuranosyl)-adenine (32)

Compound 21 (1 g, 1.86 mmol) was dissolved in EtOH (30 mL). 5-Chloro-2-methoxybenzylammonium chloride (580 mg, 2.79 mmol) and Et₃N (392 μL, 2.79 mmol) were added, and the solution was refluxed overnight. The mixture was concentrated to dryness, dissolved in 7 N NH₃ in methanol, and stirred at room temperature for 2 h to deprotect the 2′-hydroxyl group. The volatiles were removed under reduced pressure, and the residue was purified by silica gel column (CH₂Cl₂/MeOH, 97:3). The product, compound 32, was realized in 80% yield. ¹H NMR (300 MHz, DMSO-d₆): δ 3.51–3.58 (m, 1H, H-5′A), 3.63–3.68 (m, 1H, H-5′B), 3.85 (s, 3H, OCH₃), 3.94 (m, 1H, H-4′), 4.13 (m, 1H, H-3′), 4.52–4.59 (m, 3H, N⁶-CH₂ and H-2′), 5.03 (t, 1H, J = 5.6 Hz, 5′-OH), 5.21 (d, 1H, J = 5.0 Hz, 3′-OH), 5.48 (d, 1H, J = 5.8 Hz, 2′-OH), 5.83 (d, 1H, J = 6.2 Hz, H-1′), 7.03 (d, 1H, J = 8.8 Hz, Ph), 7.16 (d, 1H, J = 2.7 Hz, Ph), 7.29 (dd, 1H, J = 2.7 and 8.8 Hz, Ph), 8.35 (s, 1H, H-8), 8.62 (br s, 1H, N⁶-H). HRMS (ESI-MS) C₁₈H₂₀N₅O₅-ICl [M + H]⁺: 548.0204 found; 548.0199 calcd.

2-Azido-N⁶-(5-chloro-2-methoxybenzyl)-9-(β-D-ribofuranosyl) adenine (33)

Sodium ascorbate (14 mg, 0.073 mmol) and CuSO₄· 5H₂O (9 mg, 0.037 mmol) were added to a mixture of 32 (200 mg, 0.365 mmol), sodium azide (47 mg, 0.73 mmol), L-proline (8 mg, 0.073 mmol), and sodium carbonate (8 mg, 0.073 mmol) in 4 mL of H₂O/^tBuOH (1:1). The mixture was stirred for 2 days at 65 °C and monitored by ¹H NMR. An amount of 10 mL of dilute NH₄-OH was added, and the crude mixture was extracted with ethyl acetate (5 × 15 mL) and washed with brine (15 mL). The organic layer was dried over MgSO₄ and purified on a silica gel column (CH₂Cl₂/MeOH, 97:3) to afford compound 33 as a white solid in 82% yield. ¹H NMR (300 MHz, DMSO-d₆): δ 3.51–3.56 (m, 1H, H5′-A), 3.60–3.66 (m, 1H, H5′-B), 3.83 (s, 3H, OCH₃), 3.93 (m, 1H, H 4′), 4.13 (m, 1H, H-3′), 4.56–4.62 (m, 3H, N⁶-CH₂ and H-2′), 5.03 (t, 1H, J = 5.4 Hz, 5′-OH), 5.17 (d, 1H, J = 4.8 Hz, 3′-OH), 5.41 (d, 1H, J = 6.3 Hz, 2′-OH), 5.81 (d, 1H, J = 6.0 Hz, H-1′), 7.02 (d, 1H, J = 8.7 Hz, Ph), 7.15 (d, 1H, J = 3.0 Hz, Ph), 7.28 (dd, 1H, J = 2.7 and 8.7 Hz, Ph), 8.34 (s, 1H, H-8), 8.63 (br s, 1H, N⁶-H). ¹³C NMR (300 MHz, DMSO-d₆): δ 38.59 (CH₂), 56.49 (OCH₃), 62.19 (C-5′), 71.15 (C-3′), 74.13 (C-2′), 86.36 (C-4′), 88.13 (C-1′), 113.02 (Ph), 119.58 (C-5), 124.29 (Ph), 128.10 (Ph), 129.93 (Ph), 140.43 (C-8), 150.69 (C-4), 155.51 (Ph and C-2), 156.33 (C-6). HRMS (ESI-MS) C₁₈H₂₀N₈O₅Cl [M + H]⁺: 463.1248 found, 463.1245 calcd. Anal. (C₁₈H₁₉N₈O₅Cl) C, H, N.

Cell Culture and Membrane Preparation

CHO cells expressing recombinant human ARs or the rat A₃AR were cultured in DMEM (Dulbecco’s modified Eagle’s medium) and F12 (1:1) supplemented with 10% fetal bovine serum, 100 units/mL penicillin, 100 μg/mL streptomycin, and 2 μmol/mL glutamine. After harvest and homogenization, the cells were centrifuged at 500g for 10 min. The pellet was resuspended in 50 mM Tris-HCl buffer (pH 7.4) containing 10 mM MgCl₂ and 1 mM EDTA. The suspension was homogenized with an electric homogenizer for 10 s and was then recentrifuged at 20000g for 20 min at 4 °C. The resulting pellets were resuspended in buffer containing 3 units/mL of adenosine deaminase, and the suspension was stored at −80 °C prior to the binding experiments. The protein concentration was measured using the Bradford assay.³⁹

Radioligand Binding Studies

For the A₃AR binding experiments, the procedures were similar to those previously described.¹⁹ Briefly, each tube contained 100 μL of membrane suspension, 50 μL of [¹²⁵I]I-AB-MECA ([¹²⁵I]N⁶-(4-amino-3-iodobenzyl)adenosine-5′-N-methyluronamide, final concentration of 0.5 nM), and 50 μL of increasing concentrations of compounds in Tris-HCl buffer (50 mM, pH 7.4) containing 10 mM MgCl₂ and 1 mM EDTA. Nonspecific binding was determined using 10 μM NECA (adenosine-5′-N-ethyluronamide). The mixtures were incubated at 25 °C for 60 min. Binding reactions were terminated by filtration through Whatman GF/B filters under reduced pressure using a MT-24 cell harvester (Brandel, Gaithersburg, MD). Filters were washed three times with ice-cold buffer. Radioactivity was determined in a Beckman 5500B γ-counter. The binding of [³H]CCPA (2-chloro-N⁶-cyclopentyladenosine) to the recombinant hA₁AR and the binding of [³H]CGS21680 (2-[p-(2-carboxyethyl)phenylethylamino]-5′-N-ethylcarboxamidoadenosine) to the recombinant hA_2AAR was performed as previously described.^20,44

Cyclic AMP Accumulation Assay

Intracellular levels of 3′,5′-cyclic AMP were measured by the competitive protein binding method.³² CHO cells expressing recombinant human⁴⁰ ARs were harvested by trypsinization. After resuspension in the medium, cells were plated in 24-well plates in 0.5 mL of medium/well. After 24 h the medium was removed and cells were washed three times with 1 mL/well of DMEM containing 50 mM N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid, pH 7.4. Cells were then treated with agonists and/or test compounds in the presence of rolipram (10 μM) and adenosine deaminase (3 units/mL) and incubated at 37 °C. For the A₃AR, after 45 min forskolin (10 μM) was added to the medium, and incubation was continued for an additional 15 min. The reaction was terminated upon removal of the medium, and the cells were lysed with 200 μL/well of 0.1 M ice-cold HCl. The cell lysate was resuspended and stored at −20 °C. For determination of cyclic AMP production, protein kinase A (PKA) was incubated with [³H]cyclic AMP (2 nM) in K₂HPO₄/EDTA buffer (K₂HPO₄, 150 mM; EDTA, 10 mM), 20 μL of the cell lysate, and 30 μL of 0.1 M HCl. Bound radioactivity was separated by rapid filtration through Whatman GF/C filters under reduced pressure and washed once with cold buffer. Bound radioactivity was subsequently measured by scintillation spectrometry. Calculation of the relative maximal efficacy at the A₃AR was determined at a fixed concentration of the nucleoside analogue (10 μM) and expressed as a relative percent of the effect of 10 μM NECA determined in each experiment, which typically reached ×50% inhibition of the forskolin stimulated cyclase.

Molecular Modeling

All calculations were performed on a Silicon Graphics Octane2 workstation (600 MHz IP30 processor, MIPS R14000). Compound 10, 2-(4-cyclopentylmethyl-1,2,3-triazole)-N⁶-methyladenosine, was constructed with the use of the Sketch Molecule of SYBYL 7.1 (Tripos Inc., 1699 South Hanley Road, St. Louis, MO 63144). A grid search was performed in which flexible bonds were rotated by 0° and 180° for t1 (C₅–C₆–N⁶–C_Me) at the N⁶ position, t2 (4′O–4′C–5′C–5′OH) at the 5′-position, and t3 (N₃–C₂–N₁′–N₂′) and by 60°, 180°, and −60° for t4 (N₃′–C₄′–C_Me–C_Cyc) and t5 (C₄′–C_Me–C_Cyc–C_Cyc) at the C2 position. The low-energy conformers from the grid search were reoptimized, removing all torsional constraints. Merck molecular force field (MMFF)⁴¹ and charges were applied with the use of distance-dependent dielectric constants and the conjugate gradient method until the gradient reached 0.05 kcal·mol⁻¹·Å⁻¹. After the low-energy conformers from the result of the grid search were clustered, the representative ones from all groups were reoptimized by semiempirical molecular orbital calculations with the PM3 method in the MOPAC 6.0 package.⁴²

A human A₃AR model (PDB code 1OEA) constructed by homology to the X-ray structure of bovine rhodopsin with 2.8 Å resolution (PDB code 1F88)³⁸ was used for the docking study. All atom types were assigned by the Amber7_FF99 force field.⁴³ Amber charges for protein and MMFF charges for ligand were calculated. The starting geometry of the ligand conformation was chosen from the human A₃AR complex model with Cl-IB-MECA,¹⁹ which was already validated by point mutation. The ribose binding position was fixed, using an atom-by-atom fitting method for the carbon atoms of the ribose ring. To determine the binding region of the 2-(4-cyclopentylmethyl-1,2,3-triazole) moiety at the adenine 2 position, the flexible bond defining a χ₁ (O–C_1′–N₉–C₄) angle was searched while docked within the putative binding cavity through various low-energy conformers with diverse t1–t5 angles, rotating by −60°, −110°, and −160°, assuming an anti conformation. Several conformations without any steric bump were selected for further optimization. The initial structures of all complexes were optimized using the Amber force field with a fixed dielectric constant of 4.0 and a terminating gradient of 0.05 kcal·mol⁻¹·Å⁻¹.