Interaction of 1,4-Dihydropyridine and Pyridine Derivatives with Adenosine Receptors: Selectivity for A₃ Receptors

Abstract

1,4-Dihydropyridine and pyridine derivatives bound to three subtypes of adenosine receptors in the micromolar range. Affinity was determined in radioligand binding assays at rat brain A₁ and A_2A receptors using [³H]-(R)-PIA [[³H]-(R)-N⁶-(phenylisopropyl)adenosine] and [³H]CGS 21680 [[³H]-2-[[4-(2-carboxyethyl)phenyl]ethylamino]-5′-(N-ethylcarbamoyl)adenosine], respectively. Affinity was determined at cloned human and rat A₃ receptors using [¹²⁵I]AB-MECA [N⁶-(4-amino-3-iodobenzyl)-5′-(N-methylcarbamoyl)adenosine]. Structure–activity analysis at adenosine receptors indicated that sterically bulky groups at the 4-, 5-, and 6-positions are tolerated. (R,S)-Nicardipine, 12, displayed K_i values of 19.6 and 63.8 μM at rat A₁ and A_2A receptors, respectively, and 3.25 μM at human A₃ receptors. Similarly, (R)-niguldipine, 14, displayed K_i values of 41.3 and 1.90 μM at A₁ and A₃ receptors, respectively, and was inactive at A_2A receptors. A preference for the R- vs the S-enantiomer was observed for several dihydropyridines at adenosine receptors, in contrast with the selectivity at L-type Ca²⁺ channels. A 4-trans-β-styryl derivative, 24, with a K_i value of 0.670 μM at A₃ receptors, was 24-fold selective vs A₁ receptors (K_i) 16.1 μM) and 74-fold vs A_2A receptors (K_i) 49.3 μM). The affinity of 24 at L-type Ca²⁺ channels, measured in rat brain membranes using [³H]isradipine, indicated a K_i value of 0.694 μM, and the compound is thus nonselective between A₃ receptors and L-type Ca²⁺ channels. Inclusion of a 6-phenyl group enhanced A₃ receptor selectivity: Compound 28 (MRS1097; 3,5-diethyl 2-methyl-6-phenyl-4-(trans-2-phenylvinyl)-1,4(R,S)-dihydro-pyridine-3,5-dicarboxylate) was 55-fold selective vs A₁ receptors, 44-fold selective vs A_2A receptors, and over 1000-fold selective vs L-type Ca²⁺ channels. In addition, compound 28 attenuated the A₃ agonist-elicited inhibitory effect on adenylyl cyclase. Furthermore, whereas nicardipine, 12, displaced radioligand from the Na⁺-independent adenosine transporter with an apparent affinity of 5.36 ± 1.51 μM, compound 28 displaced less than 10% of total binding at a concentration of 100 μM. Pyridine derivatives, when bearing a 4-alkyl but not a 4-phenyl group, maintained affinity for adenosine receptors. These findings indicate that the dihydropyridines may provide leads for the development of novel, selective A₃ adenosine antagonists.

The 1,4-dihydropyridines have been developed extensively as potent blockers and activators of L-type calcium channels.¹ A number of these channel blockers, such as nifedipine (Figure 1, 9) and nicardipine (Figure 1, 12), are used therapeutically in the treatment of cardiovascular disorders, especially hypertension and coronary heart disease.²

Figure 1.

Structures of 1,4-dihydropyridines as potent calcium channel antagonists (nifedipine, 9, and nicardipine, 12) or as ligands at other receptor sites (SNAP 5089, an adrenoceptor antagonist, and 3-ethyl 5-[2-(phenylthio)ethyl] 2,4,6-trimethyl-1,4-dihydropyridine-3,5-dicarboxylate, 4, a PAF antagonist).

Dihydropyridines appear to be “privileged structures” in medicinal chemistry and pharmacology, i.e., they display affinity for many diverse binding sites.³ This adaptability of dihydropyridines has been utilized to optimize affinity in binding to α_1a-adrenergic receptors (e.g., the antagonist SNAP 5089, Figure 1),⁴ to platelet activating factor (PAF, 1-O-hexadecyl/octadecyl-2-O-acetyl-sn-glycero-3-phosphorylcholine) receptors,⁵ and at other receptor targets.⁶ Thus, by careful structural modification, it has been possible to select for affinity at sites other than Ca²⁺ channels. For example, a dihydropyridine derivative, 3-ethyl 5-[2-(phenylthio)-ethyl] 2,4,6-trimethyl-1,4-dihydropyridine-3,5-dicarboxylate (Figure 1, 4), has been found to inhibit PAF binding with an affinity of 69 nM, while the same derivative was inactive as a calcium channel blocker.⁵ Several dihydropyridines were also found to bind to A₁ adenosine receptors in rat brain.^7,8 In the present study we have used the 1,4-dihydropyridine nucleus as a template for probing structure–activity relationships (SAR) at several subtypes of central adenosine receptors (A₁, A_2A, and A₃), in an effort to design new selective antagonists. The classical7 (A₁ and A_2A) adenosine receptor antagonists are xanthines,⁹ but it would be desirable to expand the list of diverse chemical structures known to bind to adenosine receptors. There is already a large number of non-xanthine antagonists known for the A₁ receptor.^9–13 Although selective agents have been reported for both A₁ and A_2A subtypes of adenosine receptors,⁹ the development of ligands for the recently discovered A₃ receptor is lagging behind the other subtypes.¹⁴ Activation of the A₃ receptor results in hypotension and promotion of release of inflammatory mediators from mast cells.¹⁵ A₃ adenosine receptor antagonists are being sought as potential antiasthmatic,¹⁵ antiinflammatory,¹⁵ and cerebroprotective agents.¹⁶ There may also be an involvement of A₃ receptors in cancer¹⁷ and apoptosis.¹⁸ We have introduced the first selective agonists for the A₃ adenosine receptor subtype, including N⁶-(3-iodobenzyl)-5′-(N-methylcarbamoyl)adenosine (IB-MECA),^19,20 but studies of the SAR of xanthines at this subtype have thus far failed to identify principles of achieving selectivity.²¹ Furthermore, at A₃ receptors the affinity of xanthines is generally much weaker than at the A₁ or A_2A subtypes,^21,22 and considerable species variability in their affinity has been noted.^23–25 A screen of >100 cyclic compounds of diverse structure has provided several unexplored leads for A₃ selectivity, but potencies were generally low.¹¹ A recent study has identified additional natural product leads having moderate affinity, but the degree of selectivity was still minimal.¹³ Thus, the design of potent and selective A₃ antagonists has remained an unfulfilled challenge.¹⁴

Results

Synthesis.

The structures of the 1,4-dihydropyridines and related derivatives tested for affinity in radioligand binding assays at adenosine receptors are shown in Table 1. Some of the derivatives, i.e., both well-known Ca²⁺ channel ligands (nifedipine, 9; nitrendipine, 10; nicardipine, 12; nimodipine, 13; niguldipine, 14 and 15; Bay K 8644, 18 and 19) and a PAF antagonist (4),⁵ were obtained from commercial sources. Other compounds were synthesized using standard methodology (Scheme 1),^26–28 in fair yield, as reported in Table 2. The synthesis consisted of the Hantzsch condensation of a 3-amino-2-butenoate ester, 33a,b, an aldehyde, 34a–o, and a 3-ketopropionate ester derivative, 35a–e, that were dissolved in ethanol and heated to 100 °C. In order to obtain substitution at the 4-position, the aldehyde component, 34a–o, was varied. Substitution at the 6-position was achieved by varying the 3-ketopropionate ester component, 35a–e (Scheme 1).

Table 1.

Affinities of Dihydropyridine and Pyridine Derivatives in Radioligand Binding Assays at A₁, A_2A, and A₃ Receptors^a-d

						Ki (mM) or % inhibitionc
compd	R2	R3	R4	R5	R6	rA1	rA2A	hA3
1	CH3	CO2CH3	CH3	CO2CH2CH3	CH3	32.6 ± 6.3	46.1 ± 6.8	32.3 ± 5.1
2	CH3	CO2CH3	CH3	CO2(CH2)2OCH3	CH3	49.2 ± 0.7	37 ± 18%(10−4)	62.3 ± 16.7
3	CH3	CO2CH3	CH3	CO2CH2Ph	CH3	6.45 ± 1.47	9.72 ± 0.63	2.78 ± 0.89
4	CH3	CO2CH2CH3	CH3	CO2(CH2)2SPh	CH3	6.50 ± 0.47	7.10 ± 2.46	5.56 ± 1.36
5	CH3	CO2CH3	CH2CH3	CO2CH2CH3	CH3	7.52 ± 2.79	9.56 ± 2.69	13.6 ± 2.0
6	CH3	CO2CH3	(CH2)2CH3	CO2CH2CH3	CH3	8.17 ± 1.58	11.5 ± 3.8	6.51 ± 0.74
7	CH3	CO2CH3	CH2CHCH3(CH2)2-CH=C(CH3)2(R, S)	CO2CH2CH3	CH3	9.10 ± 2.90	23.1 ± 8.6	7.90 ± 0.88
8	CH3	CO2CH3	Ph	CO2CH2CH3	CH3	11.0 ± 1.6	2.74 ± 0.85	12.0 ± 3.3
9(nifedipine)	CH3	CO2CH3	2-NO2Ph	CO2CH3	CH3	2.89 ± 0.23	18.2 ± 2.51	8.29 ± 2.41
10(nitrendipine)	CH3	CO2CH3	3-NO2Ph	CO2CH2CH3	CH3	8.96 ± 2.06	23.0 ± 3.7	8.30 ± 1.41
11	CH3	CO2CH2CH3	3-NO2Ph	CO2CH2CH3	CH3	3.34 ± 2.17	18.2 ± 7.9	2.51 ± 0.15
12(nicardipine)	CH3	CO2CH3	3-NO2Ph	CO2CH2CH2N(CH3)CH2Ph	CH3	19.6 ± 1.9	63.8 ± 4.2	3.25 ± 0.26
13(nimodipine)	CH3	CO2CH(CH3)2	3-NO2Ph	CO2CH2CH2OCH3	CH3	20.1 ± 1.7	44.3 ± 14.4	8.47 ± 2.75
14((R)-niguldipine)	CH3	CO2CH3	3-NO2Ph		CH3	41.3 ± 3.5	d (10−4)	1.90 ± 0.40
15((S)-niguldipine)	CH3	CO2CH3	3-NO2Ph		CH3	d (10−4)	d (10−4)	2.80 ± 0.35
16	CH3	CO2CH3	4-NO2Ph	CO2CH2CH3	CH3	37 ± 14% (10−4)	35.6 ± 1.9	5.90 ± 1.65
17	CH3	CO2CH3	2-CF3Ph	CO2CH2CH3	CH3	6.68 ± 2.37	20.7 ± 2.8	11.6 ± 1.7
18((R)-BayK8644)	CH3	CO2CH3	2-CF3Ph	NO2	CH3	0.785 ± 0.113	35.1 ± 10.1	2.77 ± 0.34
19((S)-BayK8644)	CH3	CO2CH3	2-CF3Ph	NO2	CH3	6.66 ± 1.89	86.3 ± 23.4	23.5 ± 0.6
20	CH3	CO2CH3	4-CH3OPh	CO2CH2CH3	CH3	2.75 ± 0.35	12.7 ± 3.8	4.10 ± 0.14
21	CH3	CO2CH3	3-CH3O-4-OHPh	CO2CH2CH3	CH3	51.0 ± 3.7	56.8 ± 1.9	32.1 ± 9.2
22	CH3	CO2CH3	3,4-OCH2OPh	CO2CH2CH3	CH3	3.66 ± 0.61	5.27 ± 1.97	4.58 ± 1.11
23	CH3	CO2CH3	PhCH2CH2	CO2CH2CH3	CH3	8.81 ± 0.92	6.71 ± 2.06	2.30 ± 0.70
24	CH3	CO2CH3	Ph-CH=CH-(trans)	CO2CH2CH3	CH3	16.1 ± 0.5	49.3 ± 12.5	0.670 ± 0.195
25	CH3	CO2CH3	Ph-C≡C-	CO2CH2CH3	CH3	5.39 ± 0.33	38.3 ± 7.9	0.940 ± 0.070
26	CH3	CO2CH2CH3	CH3	CO2CH2CH3	(CH2)3CH3	10.8 ± 3.52	38.0 ± 10.6	47.1 ± 10.8
27	CH3	CO2CH2CH3	CH3	CO2CH2CH3	Ph	25.9 ± 7.3	35.9 ± 15.3	7.24 ± 2.13
28(MRS1097)	CH3	CO2CH2CH3	Ph-CH=CH-(trans)	CO2CH2CH3	Ph	5.93 ± 0.27	4.77 ± 0.29	0.108 ± 0.012
29	CH3	CO2CH3	CH3	CO2CH2CH3	CH3	6.95 ± 2.66	8.96 ± 0.93	29.5 ± 2.1
30	CH3	CO2CH2CH3	CH3	CO2CH2CH3	Ph	7.41 ± 1.29	28.4 ± 9.1	4.47 ± 0.46
31	CH3	CO2CH3	2-NO2Ph	CO2CH3	CH3	d (10−4)	d (10−4)	d (10−4)
32	H	H	4-CH3OPh	H	H	44.5 ± 1.0	71 ± 29	d (10−4)

^aDisplacement of specific [³H]-(R)-PIA binding in rat brain membranes, expressed as K_i ± SEM in μM (n = 3–5).

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

^cDisplacement of specific [¹²⁵I]AB-MECA binding at human A₃ receptors expressed in HEK cells, in membranes, expressed as K_i ± SEM in μM (n = 2–3), or as a percentage of specific binding displaced at 10 μM.

^dDisplacement of ≤ 10% of specific binding at the indicated concentration. nd, not determined.

Scheme 1.

General Procedure for the Synthesis of 1,4-Dihydropyridine Derivatives

Table 2.

Characterization of Dihydropyridine and Pyridine Derivatives

no.	Tm (°C)	formula	MS	analysis	yield (%)	purification
1	125–126	C13H19NO4	253 (CI)	C,H,N	98.2	1; MeOH
2	88–89	C14H21NO5·0.25H2O	283 (CI)	C,H,N	53.0	2; TLC
3	oil	C18H21NO4·0.50H2O	315 (CI)	C,H,N	70.5	2; TLC
5	83–84	C14H21NO4	267 (CI)	C,H,N	50.5	1; MeOH
6	95–96	C15H23NO4	281 (CI)	H,Nb	73.3	1; MeOH
7	glass	C21H33NO4	363 (CI)	c	33.4	2; TLC
8	121–122	C18H21NO4	315 (CI)	C,H,N	78.6	1; MeOH
11	165–166	C19H22N2O6	374 (EI)	C,H,N	35.5	1; EtOAc
16	151–152	C18H20N2O6	360 (CI)	C,H,N	68.1	1; MeOH
17	114–115	C19H20F3NO4	383 (CI)	C,Hd	45.6	2; TLC
20	112–113	C19H23NO5·0.21EtOAc	345 (CI)	C,H,N	61.8	2; TLC
21	190	C19H23NO6	361 (CI)	C,H,N	79.6	1; MeOH
22	oil	C19H21NO6·0.32EtOAc	359 (CI)	C,H,N	63.0	2; TLC
23	87–88	C20H25NO4	343 (CI)	e	71.9	2; TLC
24	135–136	C20H23NO4	341 (CI)	C,H,N	87.9	1; MeOH
25	176–177	C20H21NO4	339 (EI)	C,H,N	43.3	1; EtOH
26	oil	C17H27NO4	309 (CI)	C,H,N	58.8	2; TLC
27	123–124	C19H23NO4	329 (CI)	C,H,N	53.0	1; pe 35–60
28	oil	C26H27NO4	417 (CI)	C,H,N	34.7	2; column
29	oil	C13H17NO4	251 (EI)	f	57.9	2; TLC
30	oil	C19H21NO4·0.20EtOH	327 (CI)	C,H,N	48.9	2; TLC

^aPurification was achieved by either (1) (re)crystallization from the solvent specified or (2) chromatography by the specified method, using EtOAc/pe, 20:80 (v/v), eluent. MeOH, methanol; EtOH, ethanol; EtOAc, ethyl acetate; pe, petroleum ether 35–60 °C fraction; TLC, preparative thin layer chromatography, silica 60, 1000 μm layer thickness; column, preparative column chromatography, silica 60, 220–440 mesh.

^b6 (C₁₅H₂₃NO₄) H, N; C: calcd, 64.03; found, 64.75. EI: calcd, 281.1627; found, 281.1630.

^c7 (C₂₁H₃₃NO₄). This compound decomposes upon prolonged standing. Biological data and mass spectrometry data were obtained with the freshly purified compound, but extensive drying and shipping proved detrimental for the elemental analysis. C, H, N: calcd, 69.39, 9.15, 3.85; found, 61.65, 7.27, 4.70. EI: calcd, 363.2410; found, 363.2398.

^d17 (C₁₉H₂₀F₃NO₄) C, H; N: calcd, 3.65; found, 4.17. EI: calcd, 383.1344; found, 383.1346.

^e23 (C₂₀H₂₅NO₄) EI: calcd, 343.1784; found, 343.1791.

^f29 (C₁₃H₁₇NO₄) EI: calcd, 251.1158; found, 251.1162.

For compound 26, the precursor butyl β-keto ester (35e) was prepared from the condensation of the enolate lithium salt of ethyl acetate and valeryl chloride.²⁸

Good yields of the 6-phenyl-1,4-dihydropyridines were obtained using a 72 h reaction time.

Oxidation of 1,4-dihydropyridines (1 and 27) to the corresponding pyridine derivatives (29 and 30, respectively) was carried out using tetrachloro-1,4-benzoquinone (chloranil, 37) in tetrahydrofuran (Scheme 2).²⁹

Scheme 2.

General Procedure for the Oxidation of 1,4-Dihydropyridine Derivatives (Compounds 29 and 30)

Most of the dihydropyridines examined, except for two pairs of pure enantiomers, 14 and 15, 18 and 19, and the nonchiral compounds 9 and 11, were racemic mixtures at position C-4.

Binding at Adenosine Receptors.

K_i values at A₁ and A_2A receptors were determined in radioligand binding assays in rat brain membranes vs [³H]-(R)-PIA [[³H]-(R)-N⁶-(phenylisopropyl)adenosine] or [³H]CGS 21680 [[³H]-2-[[4-[(2-carboxyethyl)phenyl]ethylamino]-5′-(N-ethylcarbamoyl)adenosine], respectively.^30,31 Affinity at cloned human A₃ receptors expressed in HEK-293 cells²³ was determined using [¹²⁵I]AB-MECA [N⁶-(4-amino-3-[¹²⁵I]iodobenzyl)-5′-(N-methylcarbamoyl)-adenosine].^13,32

SAR analysis at adenosine receptors indicated that sterically bulky groups are tolerated at the 4-, 5-, and 6-positions (Table 1). R₅ was varied from ethoxycarbonyl to larger ester groups in a homologous series (1 vs 2, 3, and 4), all having a methyl group at the 4-position, resulting in a general enhancement of affinity at all three adenosine receptor subtypes. Compound 4 is a potent PAF–acether antagonist⁵ and not a calcium channel antagonist; thus there is an uncoupling of the SAR in this series for interaction with both PAF and adenosine receptors on the one hand and L-type calcium channels on the other hand.

Another means of enhancing adenosine receptor affinity (relative to compound 1) was to enlarge the R₄ substituent. In the series of ethyl, n-propyl, and even larger alkyl substituents at the 4-position (5, 6, and 7) it was demonstrated that steric bulk at this position is tolerated in adenosine receptor binding and favored at the A₃ subtype and not detrimental at the A₁ subtype. The affinity at the A_2A subtype is generally decreased with increasing chain length of the alkyl substituent. Consecutively, 4-aryl substituents, typical of the potent Ca²⁺ channel blockers in clinical use, were also examined. In general, the affinity and receptor subtype selectivity of 4-aryl analogues was highly dependent on the substituents of the phenyl ring. For example, the unsubstituted 4-phenyl compound 8 was more potent (4-fold) at A_2A receptors than at either A₁ or A₃ receptors. Of the nitrophenyl derivatives, compound 10, bearing an o-nitro group, was 3-fold selective for A₁ vs A₃, and 6-fold for A₁ vs A_2A receptors, whereas the m-nitro derivative 10 was about equipotent at A₁ and A₃ receptors, but 3-fold less potent at A_2A receptors than at either A₁ or A₃ receptors. To complete the series, we also synthesized the p-nitro-substituted compound 16, which displayed a >17-fold selectivity for A₃ receptors vs A₁ receptors and 6-fold selectivity for A₃ receptors vs A_2A receptors. In addition, we prepared compounds with different substituents. Compound 17, bearing a o-trifluoromethyl group behaved similarly to the o-nitro-substituted compound 10 and was 2-fold selective for A₁ vs A₃ receptors. The p-methoxy derivative 20 was almost equipotent at A₁ (K_i = 2.75 μM), and A₃ receptors (K_i = 4.10 μM), but slightly less potent at A_2A receptors (K_i = 12.7 μM). Affinity at A₃ receptors did not appear to be a direct function of electron density of the phenyl substituents; i.e. compounds containing an electron-withdrawing p-nitro group (16) and an electron-donating p-methoxy group (20) were equipotent. Remarkably, the introduction of a hydroxy substituent on the para position of the 4-phenyl ring (in addition to a m-methoxy substituent) led to a general decrease in affinity (21; K_i = 51.0 μM at A₁, 56.8 μM at A_2A, and 32.1 μM at A₃ receptors). Thus, steric factors surrounding the phenyl ring may be more important than electronic factors. The piperonal derivative 22 had a generally increased affinity at A₁ and A₃ adenosine receptor subtypes (vs the unsubstituted phenyl compound 8), but did not display subtype selectivity.

The relatively high affinity of a 4-aryl group larger than phenyl (22), an arylalkyl group (23), and dehydro analogues thereof (24, 25) further indicated a bulk tolerance at this position. The 4-trans-β-styryl derivative, 24, was particularly potent (K_i) 0.670 μM) and selective (24- and 74- fold vs rat A₁ and A_2A, respectively) at human A₃ receptors. The phenylacetylenic analogue, 25, was nearly as potent at A₃ receptors, but not as selective as 24 for A₃ vs A₁ receptors (only 6-fold).

Substitution at the 3- and 5-positions of the dihydropyridine moiety also had a large effect on affinity at adenosine receptors. Changing an asymmetrical mixed methyl and ethyl ester of nitrendipine, 10, to the corresponding symmetrical diethyl ester, 11, resulted in a 3.3-fold enhancement of A₃ affinity. At A₁ and A_2A receptors a similar effect was observed, and subtype selectivity was therefore not affected. Thus, compound 11 displayed K_i values of 3.34 and 18.2 μM at rat A₁ and A_2A receptors, respectively, and 2.51 μM at human A₃ receptors. Nicardipine,³³ 12, differing from nitrendipine,³⁴ 10, in the presence of a sterically bulky ester group at the 5-position, displayed a moderate enhancement of affinity at human A₃ receptors of 2.6-fold, while at rat A₁ and A_2A receptors affinity was diminished 2–3 fold. Even larger ester substituents (i.e., 14, and 15) were well tolerated at human A₃ receptors, while affinity at A₁ and A_2A receptors was drastically diminished.

Affinities of enantiomeric pairs, i.e., of Bay K 8644 (18 and 19, R and S, a Ca²⁺ channel blocker and activator, respectively) and the Ca²⁺ channel blocker niguldipine (14 and 15, R and S, respectively), indicated a general preference for the R- over the S-enantiomer at all of the receptor subtypes. This is in contrast to the affinity at L-type calcium channels [(R)-(−)-niguldipine, K_i) 8.1 nM; (S)-(+)-niguldipine, K_i) 0.18 nM; or 45-fold stereoselectivity]³⁵ and at α_1a adrenoceptors [(R)-(−)-niguldipine, K_i) 4.7 nM; (S)-(+)-niguldipine, K_i) 0.16 nM; or 29-fold stereoselectivity],⁴ at which the (S)-(+)-enantiomer is preferred. Both niguldipine enantiomers tended particularly towards selectivity for the A₃ subtype (vs A₁ and A_2A receptors, but not vs Ca²⁺ channels). Thus, (R)-niguldipine, 14, displayed K_i values of 41.3 and >100 μM at rat A₁ and A_2A receptors, respectively, and 1.90 μM at human A₃ receptors.

At the 6-position, n-butyl substitution (26) was tolerated, and phenyl substitution (27) enhanced A₃ adenosine receptor binding (4.5-fold vs 1), and lead to slight subtype selectivity. Combination of 6-phenyl and 4-trans-β-styryl substituents greatly enhanced both affinity and selectivity. Consequently, compound 28 was 55-fold selective for A₃ vs A₁ receptors, and 44-fold selective for A₃ vs A_2A receptors. Two other blockers of L-type Ca²⁺ channels, verapamil and diltiazem, which are structurally unrelated to the dihydropyridines, failed to bind appreciably to adenosine receptors.

Pyridine derivatives maintained affinity for adenosine receptors, particularly when bearing a 4-alkyl group, e.g., the oxidized compound 29, displayed a K_i value of 29.5 μM at human A₃ receptors, whereas the related dihydropyridine 1 displayed a K_i value of 32.3 μM at the same receptor subtype. Similarly, a comparison of the 6-phenyldihydropyridine 27 and the corresponding pyridine 30 shows only minor differences in affinity. In contrast, the nifedipine metabolite 31 no longer bound to adenosine receptors. The distantly related pyridine derivative 32 demonstrated that affinity of 4-arylpyridine analogues for adenosine receptors may be rescued to some extent by careful manipulation of the other substituents.

Binding at L-Type Calcium Channels.

Affinity at L-type Ca²⁺ channels, measured in rat brain membranes using [³H]isradipine³⁶ (Table 3), indicated a K_i value of 0.694 ± 0.165 μM for the styryl derivative 24 (cf. 0.670 μM at human A₃ receptors, and therefore not selective). In the same assay, (R)-niguldipine, 14, had a K_i value of 59.7 ± 0.1 nM (cf. 8.1 nM at guinea pig skeletal muscle³⁵ and 1.90 μM at human A₃ receptors, and therefore at least 32-fold selective for Ca²⁺ channels). Compound 30 at a concentration of 10 μM (K_i= 4.47 μM at A₃ receptors) or 28 (K_i = 0.108 μM at human A₃ receptors, and therefore at least 1000-fold A₃ selective vs Ca²⁺ channels) at a concentration of 100 μM displaced <10% of specific binding. Slightly more potent in Ca²⁺ channel binding were 27 (17% displacement at 100 μM; K_i = 7.24 μM at human A₃ receptors, and thus >14-fold A₃ selective) and 25 (23% displacement at 100 μM; K_i = 0.94 μM at human A₃ receptors, and thus >100-fold A₃ selective). The 5-benzyl ester 3 (K_i = 1.17 ± 0.08 μM at Ca²⁺ channels and 2.78 ± 0.89 μM at A₃ receptors) and the 4-(2-phenylethyl) analogue 23 (K_i = 0.910 ± 0.207 μM at Ca²⁺ channels and 2.30 ± 0.70 μM at A₃ receptors) were about 2.5-fold selective for Ca²⁺ channels compared to human A₃ receptors.

Table 3.

Inhibition of Specific Binding of [³H]Isradipine Binding at L-Type Calcium Channels by Various Dihydropyridine and Pyridine Derivatives in Rat Brain Membranes and the Selectivity Ratio vs Affinity at Cloned Human A₃ Receptors

compd	Ki (rCa2+)a	ratio of Ki (rCa2+)/Ki (hA3)
3	1.17 ± 0.08	0.42
14	0.0597 ± 0.0001	0.031
23	0.910 ± 0.207	0.40
24	0.694 ± 0.165	1.04
25	23 ± 3% (10−4)	>106
27	17 ± 5% (10−4)	>14
28	<10% (10−4)	>1000
30	<10% (10−5)	>2.2

^aExpressed in μM as K_i ± SEM or percent displacement of specific binding at the indicated concentration (M) for three determinations, each performed in duplicate. Rat brain membranes (ca. 100 μg of protein/tube) were incubated for 1 h at 25 °C with 0.1 nM [³H]isradipine and varying concentrations of the dihydropyridine or pyridine derivative in a 50 mM Tris buffer, pH 7.4 in a total volume of 0.5 mL. Nonspecific binding was determined in the presence of 10 μM nitrendipine. K_i values were calculated using the Cheng-Prusoff equation⁴³ assuming a K_d value of 0.13 nM for [³H]isradipine.

Functional Assays at A₃-Adenosine Receptors.

We examined the antagonist properties of two dihydropyridine derivatives in a functional assay utilizing cloned rat A₃ adenosine receptors expressed in CHO cells. Forskolin-stimulated adenylyl cyclase was inhibited by IB-MECA [N⁶-(3-iodobenzyl)-5′-(N-methylcarbamoyl)adenosine; 10⁻⁹–10⁻⁴ M) in CHO cells transfected with rat A₃ adenosine receptors (maximum degree of inhibition was 40–50%).²² The concentration of IBMECA used to test prospective antagonists was between 10 nM and 1.0 μM, i.e., respectively 0.1- and 10-fold the IC₅₀ of rat A₃ receptor mediated inhibition of adenylyl cyclase. Both nicardipine, 12, and (R)-Bay K 8644, 18, at a concentration of 10 μM showed a significant (p < 0.02) effect on the agonist-mediated inhibition of adenylyl cyclase (Table 4). The inhibition of adenylyl cyclase induced by 10⁻⁶ M IB-MECA (40.2 ± 3.3%) was attenuated to 27.2 ± 8.5% by 10 μM nicardipine, 12, and the inhibition of adenylyl cyclase activity caused by 10⁻⁷ M IB-MECA was nearly completely reversed. A similar, but not as extensive, effect could be observed for (R)-Bay K 8644, 18, even though the affinity of these compounds at human A₃ adenosine receptors was virtually identical. Compound 28 at a concentration of 50 μM attenuated the IB-MECA induced inhibition of adenylyl cyclase activity at all concentrations tested (Figure 2). This suggests that compound 28 is not only an effective displacer of the A₃-selective radioligand [¹²⁵I]AB-MECA but also an effective functional antagonist at this receptor subtype.

Table 4.

Effects of Dihydropyridine Derivatives at a Concentration of 10 μM on the Inhibition of Adenylyl Cyclase Elicited by the A₃ Agonist IB-MECA

compd	% inhibitiona
	IB-MECA (10−7 M)	n	IB-MECA (10−6M)	n
control	31.4 ± 4.0	5	40.2 ± 3.3	4
12	7.94 ± 3.15b	5	27.2 ± 8.5	4
18	19.4 ± 0.6b	2	27.8 ± 7.8	2

^aAssayed in membranes from CHO cells stably expressing the cloned rat A₃ receptor using a previously reported method in the presence of 1 μM forskolin.^15,28

^bStatistically significant (p < 0.02) vs the control value.

Figure 2.

Attenuation of A₃ agonist elicited inhibition of forskolin stimulated adenylyl cyclase activity via cloned rat A₃ adenosine receptors. Adenylyl cyclase activity was determined in membranes from CHO cells stably expressing the rat A₃ receptor, as reported previously.²⁸ Adenylyl cyclase activity was stimulated with 1 μM forskolin, and this stimulation was inhibited by the A₃ selective agonist IB-MECA (shaded bars). Addition of 50 μM MRS1097, 28, attenuated the IB-MECA induced inhibition of adenylyl cyclase activity (closed bars). The asterisk (*) denotes a statistically significant effect (p < 0.02) vs the control experiment.

Interaction of 1,4-Dihydropyridines with the NBI-Sensitive Nucleoside Transporter Protein.

It was shown previously that the nucleoside uptake inhibitor NBI (S-(4-nitrobenzyl)-6-thioinosine) binds to 1,4-dihydropyridine ([³H]nimodipine) binding sites on human red blood cell membranes.³⁷ We therefore investigated the possibility that 1,4-dihydropyridines may bind to the NBI-sensitive nucleoside transporter protein. Both the commercial Ca²⁺ channel blocker nicardipine, 12, and the A₃ adenosine receptor selective ligand 24 displaced [³H]NBI from its binding site in rat forebrain membranes, with K_i values of 5.36 ± 1.51 and 1.20 ± 0.11 μM, respectively. In contrast, the A₃-selective antagonist 28 displaced less than 10% of specific binding at a concentration of 100 μM. Compound 28 is therefore >900-fold selective for A₃ receptors compared to the NBI-sensitive nucleoside transporter.

Affinity at Other Binding Sites.

Since compound 28 was the most potent and (A₃ adenosine receptor) selective compound of the series, its affinity at non-adenosine receptor binding sites was examined in a battery of radioligand binding assays (NovaScreen, Division of Oceanix Biosciences, Hanover, MD).³⁸ At a concentration of 10⁻⁵ M, there was no significant (0 ± 30%) displacement of radioligand from adrenergic (α₁, α₂, β), cholinergic (nicotinic and muscarinic M₁, M₂, M₃), dopaminergic (D₁ and D₂), histaminergic (H₁ and H₂), serotoninergic (5-HT₁, 5-HT₂, and 5-HT₃), C5a complement, central benzodiazepine (RO 151788), GABA_A (muscimol), GABA_B (baclofen), NMDA (kainate, quisqualate, and phencyclidine), glycine (strychnine sensitive and insensitive), σ (MK-801), angiotensin (AT-II), vasopressin V₁, neuropeptide Y, cholecystokinin (CCK-B), neurotensin, somatostatin, ANF1, and EGF receptors. There was also no significant displacement of binding of radioligand from second messenger sites (forskolin, phorbol ester, and inositol trisphosphate) and ion channels (N-type calcium channels, chloride channels, and low conductance potassium channels). Significant displacement of radioligand was observed for the CCK-A (40%) and VIP (45%) binding sites, but the (estimated) IC₅₀ for these compounds was still higher than 10 μM.

The Effect of Species Differences on Ligand Selectivity.

Ordinarily, a comparison of A₁, A_2A, and A₃ affinities within one species would be preferred. We chose the human A₃ receptor for our study instead of the rat A₃ receptor for the following reasons: Whereas the affinity of many ligands varies slightly between rat and human A₁ and A_2A adenosine receptors, the affinity of most known adenosine receptor antagonists is only marginal at rat A₃ receptors.^11,12,21,25 The human A₃ receptor, being more susceptible to displacement of a selective radioligand by competing compounds, therefore allowed for a better comparison between those compounds.²⁵ The comparison was justified by the similarity of affinity of dihydropyridines at rat and human non-A₃ adenosine receptors. For example, the calcium channel blocker nifedipine (9) displayed a K_i value of 5.6 ± 0.3 μM at human A_2A adenosine receptors (unpublished data) and a K_i value of 18 ± 3 μM at rat A_2A receptors (this manuscript), but the affinities of 1,4-dihydropyridines at rat A₃ receptors differed considerably from their affinity at human A₃ receptors (Table 5). Least affected by the species difference was (R)-Bay K 8644 (18) with K_i values of 2.77 ± 0.34 and 6.20 ± 2.12 μM at human and rat A₃ receptors, respectively. Thus, the species selectivity ratio for (R)-Bay K 8644 (18) is only 2.2-fold, but this value increases to 38.6-fold for our most potent and selective compound (28), which displayed a K_i value of 4.16 ± 1.55 μM at rat A₃ receptors. Consequently, the receptor subtype selectivity of the test compounds for A₃ vs either A₁ or A_2A adenosine receptors was disproportionately affected (Table 5).

Table 5.

Affinity of and Selectivity Ratio for Selected Dihydropyridine Derivatives at Rat A₃ vs Human A₃, Rat A₁, and Rat A_2A Adenosine Receptors, Respectively

compd	Ki (μM)a	selectivity ratio
		rA3/hA3	rA1/rA3	rA2A/rA3
12	11.0 ± 7.3	3.38	1.78	5.80
18	6.20 ± 2.12	2.24	0.13	5.66
24	12.1 ± 2.6	18.1	1.33	4.07
28	4.16 ± 1.55	38.6	1.43	1.15

^aAssayed in membranes from CHO cells stably expressing the cloned rat A₃ receptor using [¹²⁵I]-N⁶-(4-amino-3-iodobenzyl)-5′-(N-methylcarbamoyl)adenosine, as described previously.^13,28

Discussion

In the present study, we have demonstrated that structural modification and careful examination of the SAR of 1,4-dihydropyridines resulted in the development of one of the first A₃ adenosine receptor-selective non-xanthine ligands. Compound 28 [3,5-diethyl 2-methyl-6-phenyl-4-(trans-2-phenylvinyl)-1,4(R,S)-dihydropyridine-3,5-dicarboxylate; MRS1097] displayed a good (submicromolar) affinity in a radioligand binding assay (K_i = 0.108 ± 0.012 μM). In addition, we have shown that dihydropyridines can be effective in attenuating the IB-MECA elicited inhibition of adenylyl cyclase in CHO cells expressing the cloned rat A₃ adenosine receptor. Compound 28 was 55-fold selective vs A₁ receptors, 44-fold selective vs A_2A receptors, and over 1000-fold selective vs L-type Ca²⁺ channels.

It has been suggested that 1,4-dihydropyridines may bind to the nucleoside transporter protein of human red blood cells and human brain.^37,43 We therefore measured the ability of some 1,4-dihydropyridines to displace [³H](nitrobenzyl)thioinosine ([³H]NBI) from the Na⁺-independent adenosine transporter in rat forebrain. Whereas the commercial compound nicardipine (12) and the novel compound MRS1045 (24) displaced [³H]NBI with affinities similar to their affinity at human A₃ receptors (5.36 ± 1.51 and 1.20 ± 0.11 μM, respectively), the most selective compound of this series (MRS1097, 28) displaced less than 10% of total [³H]NBI binding at a concentration of 10⁻⁴ M.

In this study we have shown that the affinity of dihydropyridines can be optimized selectively for adenosine receptors vs L-type calcium channels and for A₃ vs A₁ and A_2A receptors. Hu et al. previously found no effect of nifedipine binding on A₁ receptor mediated adenylyl cyclase inhibition,⁷ but current work with the more recently cloned A₃ receptor²² demonstrates that IB-MECA-induced inhibition of adenylyl cyclase was attenuated by several compounds of the 1,4-dihydropyridine class (Table 4). Preliminary tests show that both nicardipine, 12, and (R)-Bay K 8644, 18 (both Ca²⁺ channel blockers), functionally antagonize the inhibitory effects of an A₃ selective agonist on adenylyl cyclase. Furthermore, we have shown that MRS1097, 28, significantly (p < 0.02) attenuates the effects of IB-MECA on adenylyl cyclase activity (Figure 2). Further pharmacological characterization will be required to determine if this is due to a competitive interaction with adenosine A₃ receptors or at some other stage of the functional cascade. Thus, we have discovered dihydropyridines as a promising lead in the search for A₃ adenosine receptor antagonists.

L-type calcium channel blockers such as nimodipine, 13, have been examined as cerebroprotective agents in experimental models of neurotoxicity in cell culture and in vivo.³⁹ Adenosine agonists and antagonists have also been evaluated in models of brain ischemia. It has also been postulated that an adenosine A₃ receptor antagonist would have cerebroprotective properties, based on the in vivo effects of the A₃ receptor agonist IB-MECA administered to gerbils in a chronic regimen.¹⁶ Thus, a compound that would display a dual action in vivo might be even more useful in treating stroke than either an L-type calcium channel blocker or A₃ receptor antagonist alone. Therefore, some of the compounds found in the present study to have affinity at both sites may be clinically useful as combined therapeutic agents.

Experimental Section

Materials.

(R)-(+)- and (S)-(−)-Bay K 8644, (R)-(−)- and (S)-(+)-niguldipine, nicardipine, nifedipine, nimodipine, and oxidized nifedipine were purchased from Research Biochemicals International (Natick, MA). Ethyl acetoacetate (35a), acetaldehyde (34a), propionaldehyde (34b), butyraldehyde (34c), benzaldehyde (34d), anisaldehyde (34e), vanillin (34f), methyl 3-aminocrotonate (33a), and trans-cinnamaldehyde (34g) were obtained from Fluka (Ronkonoma, NY). o-Nitrobenzaldehyde (34h), m-nitrobenzaldehyde (34i), p-nitrobenzaldehyde (34j), ethyl 3-aminocrotonate (33b), phenylpropargylaldehyde (34k), α,α,α -trifloro-o-tolualdehyde (34l), benzyl acetoacetate (35b), 2-methoxyethyl acetoacetate (35c), piperonal (34m), tetrachloro-1,4-benzoquinone (37), valeryl chloride, and ethyl benzoylacetate (35d) were from Aldrich (St. Louis, MO). Hydrocinnamaldehyde (34n) was from Eastman (Rochester, NY), and (±)-citronellal (34o) was from ICN (Plainview, NY). Compound 4 was from Tocris-Cookson (St. Louis, MO), and (S)-(4-nitrobenzyl)-6-thioguanosine was from Sigma (St. Louis, MO). [³H]PN200,110 (isradipine) was from DuPont NEN (Boston, MA). Compound 32 was prepared according to the literature.⁴⁰ All other materials were obtained from commercial sources.

Synthesis.

Proton nuclear magnetic resonance spectroscopy was performed on a Varian GEMINI-300 spectrometer, and spectra were taken in DMSO-d₆ or CHCl₃-d. Chemical-ionization (CI) mass spectrometry was performed with a Finnigan 4600 mass spectrometer, and electron-impact (EI) mass spectrometry with a VG7070F mass spectrometer at 6 kV. Elemental analysis was performed by Atlantic Microlab Inc. (Norcross, GA). All melting points were determined with a Unimelt capillary melting point apparatus (Arthur H. Thomas Co., Philadelphia, PA) and were uncorrected.

General Procedure for Preparation of 1,4-Dihydropyridine-3,5-dicarboxylate Esters (36; 1–3, 5–8, 11, 16, 17, 20–28).

Equimolar amounts (0.5 mmol) of the appropriate 3-amino-2-butenoate ester (33a,b), aldehyde (34a–o), and 3-ketopropionate ester (35a–e) derivative were dissolved in 5 mL of absolute ethanol. The solution was sealed in a glass tube and heated to 100 °C for at least 24 h, and at most 72 h. The solvent was then evaporated, and products were purified either by crystallization, column chromatography (silica 60; 220–440 mesh; Fluka, CH; 20% ethyl acetate-80% petroleum ether 35–60) or preparative TLC (silica 60; 1000 μm; Analtech, DE; 20% ethyl acetate-80% petroleum ether 35–60). From the moment the reactants were sealed into the glass tube, all procedures were performed under nitrogen and low-light conditions to prevent oxidation of the products. The products were shown to be homogeneous by TLC.

3-Methyl 5-ethyl 2,4,6-trimethyl-1,4(R,S)-dihydropyridine-3,5-dicarboxylate (1):

¹H NMR (CHCl₃-d) δ 0.98 (d, 3H, 4-CH₃, J = 6.7 Hz), 1.31 (t, 3H, 5-methyl, J = 7.2 Hz), 2.28 (s, 6H, 2- and 6-CH₃), 3.73 (s, 3H, 3-methyl), 3.83 (q, 1H, H-4, J = 6.4 Hz); 4.13–4.26 (m, 2H, 5-methylene), 5.51 (wide, 1H, H-1).

3-Methyl 5-(2-methoxyethyl) 2,4,6-trimethyl-1,4(R,S)-dihydropyridine-3,5-dicarboxylate (2):

¹H NMR (CHCl_3-d) δ 0.99 (d, 3H, 4-CH₃, J = 6.7 Hz), 2.28 (s, 6H, 2- and 6-CH₃), 3.41 (s, 3H, 5-methoxy), 3.66 (t, 2H, 5-(2-methylene), J = 4.9 Hz), 3.73 (s, 3H, 3-methyl), 3.85 (q, 1H, H-4, J = 6.5 Hz), 4.22–4.38 (m, 2H, 5-(1-methylene)), 5.54 (wide, 1H, H-1).

3-Methyl 5-benzyl 2,4,6-trimethyl-1,4(R,S)-dihydropyridine-3,5-dicarboxylate (3):

¹H NMR (CHCl₃-d) δ 0.98 (d, 3H, 4-CH₃, J = 5.0 Hz), 2.28 (s, 6H, 2- and 6-CH₃), 3.72 (s, 3H, 3-methyl), 3.89 (q, 1H, H-4, J = 6.5 Hz), 5.21 (q, 2H, 5-methylene, J = 14.8 Hz), 5.54 (wide, 1H, H-1), 7.30–7.41 (m, 5H, 5-phenyl).

3-Methyl 5-ethyl 2,6-dimethyl-4-ethyl-1,4(R,S)-dihydropyridine-3,5-dicarboxylate (5):

¹H NMR (DMSO-d₆) δ 0.63 (t, 3H, 4-methyl, J = 7.4 Hz), 1.18 (t, 3H, 5-methyl, J = 7.1 Hz), 2.19 (s, 6H, 2- and 6-CH₃), 3.58 (s, 3H, 3-methyl), 3.73 (t, 1H, H-4, J = 5.4 Hz), 4.00–4.13 (m, 4H, 4- and 5-methylene).

3-Methyl 5-ethyl 2,6-dimethyl-4-propyl-1,4(R,S)-dihydropyridine-3,5-dicarboxylate (6):

¹H NMR (DMSO-d₆) δ 0.77 (t, 3H, 4-methyl, J = 6.7 Hz), 1.12 (wide, 4H, 4a- and 4b-methylene), 1.19 (t, 3H, 5-methyl, J = 7.2 Hz), 2.19 (s, 6H, 2- and 6-CH₃), 3.59 (s, 3H, 3-methyl), 3.76 (t, 1H, H-4, J = 5.2 Hz), 4.07 (q, 2H, 5-methylene, J = 6.6 Hz).

3-Methyl 5-ethyl 2,6-dimethyl-4-(2(R,S),6-dimethylhexen-5-yl)-1,4(R,S)-dihydropyridine-3,5-dicarboxylate (7):

¹H NMR (CHCl₃-d) δ 0.91 (double d, 3H, 4-(2-CH₃), J₁ = 5.8 Hz, J₂ = 24.0 Hz), 1.30 (m, 7H), 1.63 (d, 6H, 4-(6- and 7-CH₃), J = 25.4 Hz), 2.28 (d, 6H, 2- and 6- CH₃, J = 7.8 Hz), 3.71 (s, 3H, 3-methyl), 3.97 (t, 1H, H-4, J = 7.2 Hz), 4.11–4.24 (m, 2H, 5-methylene), 5.07 (t, 1H, 4-(2-methyne), J = 6.5 Hz), 5.71 (wide, 1H, H-1).

3-Methyl 5-ethyl 2,6-dimethyl-4-phenyl-1,4(R,S)-dihydropyridine-3,5-dicarboxylate (8):

¹H NMR (DMSO-d₆) δ 1.11 (t, 3H, 5-methyl, J = 6.4 Hz), 2.23 (s, 6H, 2- and 6-CH₃), 3.51 (s, 3H, 3-methyl), 3.97 (m, 2H, 5-methylene), 4.84 (s, 1H, H-4), 7.08–7.20 (m, 5H, 4-phenyl).

3,5-Diethyl 2,6-dimethyl-4-(3-nitrophenyl)-1,4(R,S)-dihydropyridine-3,5-dicarboxylate (11):

¹H NMR (CHCl₃-d) δ 1.23 (t, 6H, 3- and 5-methyl, J = 6.9 Hz), 2.38 (s, 6H, 2- and 6-CH₃), 4.02–4.18 (m, 4H, 3- and 5-methylene), 5.10 (s, 1H, H-4), 5.68 (wide, 1H, H-1), 7.38 (t, 1H, H-5′, J = 8.0 Hz), 7.65 (d, 1H, H-6′, J = 7.8 Hz), 8.02 (d, 1H, H-4′, J = 7.4 Hz), 8.14 (s, 1H, H-2′).

3-Methyl 5-ethyl 2,6-dimethyl-4-(4-nitrophenyl)-1,4-(R,S)-dihydropyridine-3,5-dicarboxylate (16):

¹H NMR (DMSO-d₆) δ 1.11 (t, 3H, 5-methyl, J = 7.2 Hz), 2.26 (s, 6H, 2- and 6-CH₃), 3.52 (s, 3H, 3-methyl), 4.00 (m, 2H, 5-methylene), 4.96 (s, 1H, H-1), 7.39 (d, 2H, H-2′ and H-6′, J = 8.7 Hz), 8.09 (d, 2H, H-3′ and H-5′, J = 8.7 Hz).

3-Methyl 5-ethyl 2,6-dimethyl-4-(2-α,α,α-trifluoromethylphenyl)-1,4(R,S)-dihydropyridine-3,5-dicarboxylate (17):

¹H NMR (DMSO-d₆) δ 1.05 (m, 3H, 5-methyl), 2.21 (s, 6H, 2- and 6-CH₃), 3.44 (s, 3H, 3-methyl), 3.84–4.10 (m, 2H, 5-methylene), 5.38 (s, 1H, H-4), 7.31 (t, 1H, J = 7.3 Hz), 7.51 (m, 3H, 4-phenyl).

3-Methyl 5-ethyl 2,6-dimethyl-4-(4-methoxyphenyl)-1,4(R,S)-dihydropyridine-3,5-dicarboxylate (20):

¹H NMR (CHCl₃-d) δ 1.23 (t, 3H, 5-methyl, J = 7.8 Hz), 2.34 (s, 6H, 2- and 6-CH₃), 3.65 (s, 3H, 4′-OCH₃), 3.76 (s, 3H, 3-methyl), 4.07–4.14 (m, 2H, 5-methylene), 4.94 (s, 1H, H-4), 5.57 (wide, 1H, H-1), 6.76 (d, 2H, H-2′ and H-6′, J = 8.5 Hz), 7.20 (d, 2H, H-3′ and H-5′, J = 8.6 Hz).

3-Methyl 5-ethyl 2,6-dimethyl-4-(4-hydroxy-3-methoxyphenyl)-1,4(R,S)-dihydropyridine-3,5-dicarboxylate

(21): ¹H NMR (DMSO-d₆) δ 1.13 (t, 3H, 5-methyl, J = 7.4 Hz), 2.21 (s, 6H, 2- and 6-CH₃), 3.52 (s, 3H, 3′-OCH₃), 3.66 (s, 3H, 3-methyl), 3.96–4.01 (m, 2H, 5-methylene), 4.74 (s, 1H, H-4), 6.49 (t, 1H, H-6′, J = 3.9 Hz), 6.58 (d, 1H, H-5′, J = 7.9 Hz), 6.66 (s, 1H, H-2′).

3-Methyl 5-ethyl 2,6-dimethyl-4-[3,4-(methylenedioxy)-phenyl]-1,4(R,S)-dihydropyridine-3,5-dicarboxylate (22):

¹H NMR (CHCl₃-d) δ 1.24 (t, 3H, 5-methyl, J = 7.3 Hz), 2.33 (s, 6H, 2- and 6-CH₃), 3.66 (s, 3H, 3-methyl), 3.73 (s, 1H, H-4), 4.09–4.22 (m, 2H, 5-methylene), 4.92 (s, 2H, 3′,4′-methylenedioxy), 5.57 (wide, 1H, H-1), 5.89 (s, 1H, H-2′), 6.66 (d, 1H, H-5′, J = 8.1 Hz), 6.75 (d, 1H, H-6′, J = 7.9 Hz).

3-Methyl 5-ethyl 2,6-dimethyl-4-(2-phenyl)ethyl-1,4-(R,S)-dihydropyridine-3,5-dicarboxylate (23):

¹H NMR (CHCl₃-d) δ 1.25–1.55 (m, 3H, 5-methyl), 1.63–1.71 (m, 2H, 4-(α-methylene)), 2.31 (s, 6H, 2- and 6-CH₃), 2.51–2.57 (m, 2H, 4-(β-methylene)), 3.73 (s, 3H, 3-methyl), 4.05 (t, 1H, H-4, J = 5.6 Hz), 4.12–4.27 (m, 2H, 5-methylene), 5.58 (wide, 1H, H-1), 7.15 (d, 2H, H-2′ and H-6′, J = 6.4 Hz), 7.24 (t, 3H, H-3′, and H-4′ and H-5′, J = 6.8 Hz).

3-Methyl 5-ethyl 2,6-dimethyl-4-(trans-2-phenylvinyl)-1,4(R,S)-dihydropyridine-3,5-dicarboxylate (24):

¹H NMR (CHCl₃-d) δ 1.31 (t, 3H, 5-methyl, J = 7.1 Hz), 2.34 (s, 6H, 2- and 6-CH₃), 3.74 (s, 3H, 3-methyl), 4.14–4.28 (m, 2H, 5-methylene), 4.63 (d, 1H, H-4, J = 5.4 Hz), 5.60 (wide, 1H, H-1), 6.19 (t, 1H, 4-(H-1 vinylidene), J = 6.0 Hz), 7.18 (d, 1H, 4-(H-2 vinylidene), J = 6.6 Hz), 7.24–7.34 (m, 5H, 4-phenyl).

3-Methyl 5-ethyl 2,6-dimethyl-4-(phenylethynyl)-1,4-(R,S)-dihydropyridine-3,5-dicarboxylate (25):

¹H NMR (CHCl₃-d) δ 1.35 (t, 3H, 5-methyl, J = 7.1 Hz), 2.36 (s, 6H, 2- and 6-CH₃), 3.80 (s, 3H, 3-methyl), 4.23–4.31 (m, 2H, 5-methylene), 4.99 (s, 1H, H-4), 5.71 (wide, 1H, H-1), 7.24 (t, 3H, H-3′, and H-4′, and H-5′, J = 3.2 Hz), 7.36 (d, 2H, H-2′ and H-6′, J = 3.6 Hz).

3,5-Diethyl 2,4-dimethyl-6-butyl-1,4(R,S)-dihydropyridine-3,5-dicarboxylate (26):

¹H NMR (CHCl₃-d) δ 0.92–0.98 (m, overlap, 6H, 6-(4-CH₃) and 4-CH₃, J = 7.0 Hz), 1.27 (t, 6H, 3- and 5-methyl, J = 7.0 Hz), 1.51 (m, 4H, 6-(2- and 3-CH₂)), 2.30 (s, 3H, 2-CH₃), 2.52–2.76 (m, 2H, 6-(1-CH₂)), 3.85 (q, 1H, H-4, J = 7.0 Hz), 4.18 (m, 4H, 3- and 5-methylene), 5.53 (wide, 1H, H-1).

3,5-Diethyl 2,4-dimethyl-6-phenyl-1,4(R,S)-dihydropyridine-3,5-dicarboxylate (27):

¹H NMR (CHCl₃-d) δ 0.89 (t, 3H, 3-methyl, J = 7.4 Hz), 1.14 (d, 3H, 5-methyl, J = 7.4 Hz), 1.31 (t, 3H, J = 7.0 Hz), 2.3 (s, 3H, CH₃), 3.94 (m, 3H), 4.23 (m, 2H), 7.28–7.41 (m, 5H, C₆H₅).

3,5-Diethyl 2-methyl-4-(trans-2-phenylvinyl)-6-phenyl-1,4(R,S)-dihydropyridine-3,5-dicarboxylate (28):

¹H NMR (CHCl₃-d) δ 0.92 (t, 3H, 5-methyl, J = 7.0 Hz), 1.32 (t, 3H, 3-methyl, J = 7.0 Hz), 2.37 (s, 3H, 2-CH₃), 3.94 (q, 3H, 5-methylene, J = 6.7 Hz), 4.22, (q, 3H, 3-methylene, J = 6.7 Hz), 4.76 (d, 1H, H-4, J = 6.2 Hz), 5.78 (wide, 1H, H-1), 6.30–6.38 (m, 2H, CH=CH), 7.20–7.44 (m, 10H, 2 x C₆H₅).

General Procedure for Oxidation of 1,4-Dihydropyri-dine-3,5-dicarboxylate Esters (38; 29, 30).

Equimolar amounts (0.25 mmol) of the 1,4-dihydropyridine-3,5-dicarboxylate ester (36; 1, 27) and tetrachloro-1,4-benzoquinone (37) in tetrahydrofuran (2 mL) were mixed and refluxed for up to 4 h. The solvent was then evaporated, and products were purified by preparative TLC (silica 60; 1000 μm; Analtech, DE; 20% ethyl acetate–80% petroleum ether 35–60).

3-Methyl 5-ethyl 2,4,6-trimethylpyridine-3,5-dicarboxylate (29):

¹H NMR (DMSO-d₆) δ 1.27 (t, 3H, 5-methyl, J 6.9 Hz), 2.16 (s, 3H, 4-CH₃), 2.39 (s, 6H, 2- and 6-CH₃), 3.87 (s, 3H, 3-methyl), 4.36 (q, 2H, 5-methylene, J = 7.4 Hz).

3,5-Diethyl 2,4-dimethyl-6-phenyl-pyridine-3,5-dicarboxylate (30):

¹H NMR (CHCl₃-d) δ 1.00 (t, 3H, 3-methyl, J = 7.4 Hz), 1.43 (t, 3H, 5-methyl, J = 7.5 Hz), 2.37 (s, 3H,4-CH₃), 2.62 (s, 3H, 2-CH₃), 4.10 (q, 2H, 3-methylene, J = 7.4 Hz), 4.47 (q, 2H, 5-methylene, J = 7.4 Hz), 7.40–7.58 (m, 5H, C₆H₅).

Ethyl Valerylacetate (35e).

A solution of n-butyllithium (6.3 mL, 1.6 M in hexane) was slowly added to a solution of N-isopropylcyclohexylamine (1.41 g, 10 mmol) in dry THF (20 mL) at −5 °C. The mixture was stirred for 15 min and then cooled to −78 °C. Ethyl acetate (440 mg, 5 mmol) was added dropwise over a period of 5 min, followed by valeryl chloride (600 mg, 5 mmol). The reaction mixture was allowed to stir an additional 10 min, at which point it was quenched with 5 mL of 20% HCl in water. The organic layer was separated off, and the aqueous layer was extracted with 10 mL of diethyl ether twice. The combined organic phases were washed with a saturated sodium bicarbonate solution (10 mL × 2) and brine (5 mL × 2) and dried over anhydrous sodium sulfate. The solvent was evaporated and the residue purified by column chromatography (silica gel 60, eluted with ethyl acetate/petroleum ether, 1:9) to yield ethyl valerylacetate (772 mg, 90%): ¹H NMR (CHCl₃-d) δ 0.90 (t, 3H, CH₃, J = 7.0 Hz), 1.20–1.65 (m, 7H, CH₃ and 2 × CH₂), 2.30 (t, 2H, COCH₂, J = 7.0 Hz), 2.65 (s, 2H, COCH₂CO), 4.15 (q, 2H, COOCH₂, J = 7.0 Hz); MS (EI) 85 [CH₃(CH₂)₃CO⁺], 172 (base), 57 [CH₃-(CH₂)₃]⁺. The ethyl valerylacetate (35e) was then used in the procedure described above for the preparation of compound 26.

Pharmacology: Radioligand Binding Studies.

Binding of [³H]-(R)-N⁶-(phenylisopropyl)adenosine ([³H]-(R)-PIA) to A₁ receptors from rat cerebral cortex membranes and of [³H]-2-[[4-(2-carboxyethyl)phenyl]ethylamino]-5′-N-(ethylcarbamoyl)-adenosine ([³H]CGS 21680) to A_2A receptors from rat striatal membranes was performed as described previously.^30,31 Adenosine deaminase (3 units/mL) was present during the preparation of the brain membranes, in a preincubation of 30 min at 30 °C, and during the incubation with the radioligands.

Binding of [¹²⁵I]-N⁶-(4-amino-3-iodobenzyl)-5′-(N-methylcarbamoyl)adenosine ([¹²⁵I]AB-MECA) to membranes prepared from HEK-293 cells stably expressing the human A₃ receptor (Receptor Biology, Inc., Baltimore, MD) or to membranes prepared from CHO cells stably expressing the rat A₃ receptor was performed as described.^13,32 The assay medium consisted of buffer containing 50 mM Tris, 10 mM MgCl₂, and 1 mM EDTA at pH 8.0. The glass incubation tubes contained 100 μL of the membrane suspension (0.3 mg of protein/mL, stored at −80 °C in the same buffer), 50 μL of [¹²⁵I]AB-MECA (final concentration 0.3 nM), and 50 μL of a solution of the proposed antagonist. Nonspecific binding was determined in the presence of 100 μM N⁶-(phenylisopropyl)adenosine (R-PIA).

Binding of [³H]isradipine to rat cerebral cortex membranes was performed essentially as described in Moody et al.³⁶ Briefly, an aliquot of a membrane suspension corresponding to 100 μg of protein was incubated in a total volume of 0.5 mL of Tris buffer (50 mM, pH 7.4, 25 °C) for 1 h in the presence of 0.1 nM [³H]isradipine and varying concentrations of the dihydropyridine or pyridine derivatives. Nonspecific binding was determined in the presence of 10 μM nitrendipine.

All nonradioactive compounds were initially dissolved in DMSO and diluted with buffer to the final concentration, where the amount of DMSO never exceeded 2%.

Incubations were terminated by rapid filtration over What-man GF/B filters, using a Brandell cell harvester (Brandell, Gaithersburg, MD). The tubes were rinsed three times with 3 mL of buffer each.

At least five different concentrations of competitor, spanning 3 orders of magnitude adjusted appropriately for the IC₅₀ of each compound, were used. IC₅₀ values, calculated with the nonlinear regression method implemented in the InPlot program (Graph-PAD, San Diego, CA), were converted to apparent K_i values using the Cheng–Prusoff equation⁴¹ and K_d values of 1.0, 14, and 0.13 nM for [³H]-(R)-PIA, [³H]CGS 21680, and [³H]isradipine, respectively, and 0.59 nM for binding of [¹²⁵I]-AB-MECA at human A₃ receptors, respectively.

Inhibition of Adenylyl Cyclase Activity.

Adenylyl cyclase activity was determined in membranes from CHO cells stably expressing the rat A₃ receptor, prepared as reported previously.³² [α−³²P]ATP was added to a membrane suspension containing 1 μM forskolin to stimulate adenylyl cyclase. The reaction was terminated by addition of a stop buffer containing 20 000 cpm/mL [³H]cyclic AMP. Cyclic AMP was isolated by consecutive chromatography over columns of Dowex 50 ion exchange resin and alumina. Maximal stimulation of adenylyl cyclase activity was obtained in the presence of 1 μM forskolin (circa 6–8-fold), and maximal inhibition of forskolin-stimulated adenylyl cyclase activity (40–50% residual activity from the stimulated level) was obtained in the presence of 10 μM IB-MECA. Statistical significance was calculated in a t-test, with 4 degrees of freedom at a significance level of 98%.

Interaction of 1,4-Dihydropyridines with the NBI-Sensitive Nucleoside Transporter Protein.

Competition of 1,4-dihydropyridines for binding of [³H]NBI to the NBI-sensitive nucleoside transporter protein was carried out by the slightly modified procedure of Marangos et al.⁴² Rat forebrain membranes were incubated for 30 min at 23 °C with 0.7 nM [³H]NBI and varying concentrations of competitor in Tris buffer (50 mM, pH 7.4) in a total of 500 μL. Then 5 μM (S)-(4-nitrobenzyl)-6-thioguanosine was added to determine non-specific binding. Nonspecific binding did not exceed 15% of total binding. A K_d value of 0.15 nM was used for the calculation of the K_ivalues.⁴²

Abbreviations:

[¹²⁵I]AB-MECA

[¹²⁵I]-N⁶-(4-amino-3-iodobenzyl)-5′-(N-methylcarbamoyl)adenosine

Bay K 8644

3-methyl 2,6-dimethyl-5-nitro-4-(2-(α,α,α-trifluoromethyl)phenyl)-1,4-dihydropyridine-3-carboxylate

CGS 21680

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

CHA

N⁶-cyclohexyladenosine

CHO cells

Chinese hamster ovary cells

DMSO

dimethyl sulfoxide

HEK cells

human embryonic kidney cells

IB-MECA

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

K _i

equilibrium inhibition constant

NBI

S-(4-nitrobenzyl)-6-thioinosine

NECA

(N-ethylcarbamoyl)adenosine

PAF

1-O-hexadecyl/octadecyl-2-O-acetyl-sn-glycero-3-phosphorylcholine, platelet activating factor

(R)-PIA

(R)-N⁶-(phenylisopropyl)adenosine

SAR

structure–activity relationship

Tris

tris(hydroxymethyl)aminomethane

XAC

8-[4-[[[[(2-aminoethyl)amino]carbonyl]methyl]oxy]phenyl]-1,3-dipropylxanthine, xanthine amine congener