Synthesis and P2Y Receptor Activity of Nucleoside 5′-Phosphonate Derivatives

Abstract

Ribose-based nucleoside 5′-diphosphates and triphosphates and related nucleotides were compared in their potency at the P2Y receptors with the corresponding nucleoside 5′-phosphonate derivatives. Phosphonate derivatives of UTP and ATP activated the P2Y₂ receptor but were inactive or weakly active at P2Y₄ receptors. Uridine 5′-(diphospho)phosphonate was approximately as potent at the P2Y₂ receptor as at the UDP-activated P2Y₆ receptor. These results suggest that removal of the 5′-oxygen atom from nucleotide agonist derivatives reduces but does not prevent interaction with the P2Y₂ receptor. Uridine 5′-(phospho)phosphonate as well as the 5′-methylphosphonate equivalent of UMP were inactive at the P2Y₄ receptor and exhibited maximal effects at the P2Y₂ receptor that were ≤50% of that of UTP suggesting novel action of these analogues.

The P2Y receptors are a family of eight G protein-coupled receptors (GPCRs) of class A that respond to diverse nucleotides.¹ The human P2Y₁, P2Y₁₂, and P2Y₁₃ receptors are activated preferentially by ADP, while the human P2Y₁₁ receptor is activated preferentially by ATP. The human P2Y₄, P2Y₆, and P2Y₁₄ subtypes respond exclusively to various uracil nucleotides, and the human P2Y₂ receptor responds to both UTP and ATP. Depending on the subtype, the effector coupling of these receptors is typically through G_q or G_i proteins, to stimulate phospholipase or to inhibit of adenylate cyclase, respectively.

P2Y receptors are targets for therapeutic approaches, some of which are currently being explored at an early stage.² Several clinical applications of P2Y receptor ligands are more advanced, such as P2Y₁₂ receptor antagonists as antithrombotic agents³ and P2Y₂ receptor agonists as drugs for cystic fibrosis and dry eye disease.⁴

To aid in the design of P2Y receptor ligands, we have reported rhodopsin-based molecular models of all known subtypes of P2Y receptors.^5,6 Various amino acid residues have been proposed to coordinate the bound nucleotide ligands in P2Y₁ receptors and other P2Y subtypes. The medicinal chemistry of P2Y receptors that are responsive to uracil nucleotides has recently been explored.^7–11 The conformational constraint or replacement of the hydroxyl moieties of the ribose moiety have been introduced as a means for increasing the selectivity at P2Y₁, P2Y₂, or P2Y₆ receptors.^12,13

In the present study, we have explored the removal of the 5′-oxygen or its replacement by carbon moieties, resulting in 5′-phosphonate derivatives. The activity of various phosphonate derivatives was compared with the native ribosides at various P2Y receptor subtypes. Potency was best preserved at the P2Y₂ receptor following this modification. However, the apparent efficacy of activation of the P2Y₂ receptor by several phosphonate derivatives was variable and significantly less than 100%, suggesting a potential change in the mode of interaction with the receptor. The introduction of the phosphonate linkage is also intended to increase stability of the alpha-phosphate toward hydrolysis by nucleotidases, which is frequently a limitation in the pharmacological use of nucleotide derivatives in in vitro and in vivo experiments.

Various adenine and uracil nucleotide phosphonate derivatives were synthesized and evaluated in functional assays of different P2Y receptors (Table 1). The derivatives in which the 5′-oxygen was omitted, i.e., 1, 2, and 4–6, were prepared by known synthetic routes.¹⁴ The synthetic route to the elongated 5′-phosphonate derivatives 3 and 7 is shown in Scheme 1 following earlier reported strategies.^15,19 5′, 6′-Vinyl phosphonate 9 was synthesized by oxidation of 2′, 3′-O-isopropylideneuridine (8) to the 5′-aldehyde intermediate, which was immediately reacted with freshly prepared [(diethoxyphosphinyl)methylidene] triphenylphosphorane.¹⁶ The E-configuration of the resulting alkene could be inferred from the large coupling constant (3J = 17.3 Hz). Catalytic hydrogenation of the resulting vinyl phosphonate ester 9 in the presence of palladium on carbon gave the corresponding saturated phosphonate ester 10. Simultaneous deprotection of the phosphonate diester and the acetonide was achieved upon treatment with TMSBr, yielding the corresponding 5′-phosphonate 7.

Table 1. In vitro.

pharmacological data for various nucleotide 5′-phosphonate derivatives analogues and the corresponding native ribosides (X = OCH₂) in the stimulation of PLC at recombinant human P2Y receptors expressed in 1321N1 astrocytoma cells.

No	Name	R	n	B	X	Receptor	EC50, μMa	Reference comp. (X = OCH2) EC50, μM7–11
1	adenosine 5′-(diphospho)-phosphonate	HO	2	A	CH2	P2Y2	1.4 ± 0.6	0.085 ± 0.012
						P2Y4	<50% at 10 μM	Antagonist12
2	uridine 5′-(diphospho)-phosphonate	HO	2	U	CH2	P2Y2	8.6 ± 3.1	0.049 ± 0.012
						P2Y4	NE	0.073 ± 0.02
						P2Y6	NE	>1010
3	uridine 5′-(diphospho)- vinylphosphonate	HO	2	U	trans -CH=CH	P2Y2	1.0 ± 0.4	0.049 ± 0.012
						P2Y4	<50% at 10 μM	0.09 ± 0.0121
						P2Y6	NE	NE
4	uridine 5′-phospho- phosphonate	HO	1	U	CH2	P2Y2	3.4 ± 1.1 (partial ag)	NE
						P2Y4	NE	NE
						P2Y6	1.9 ± 0.7	0.046 ± 0.020
5	di-(uridine- phosphonate)		1	U	CH2	P2Y2	<50% at 10 μM	NE8
						P2Y4	NE	NE8
						P2Y6	NE	1.26 ± 0.17b
6	uridine 5′-(glucose[1′]phos- pho)phosphonate		1	U	CH2	P2Y14	NE	0.301 ± 0.07
7	uridine 5 ′-methylene-phosphonate	HO	0	U	CH2	P2Y2	1.6 ± 0.4 (partial ag)	NE
						P2Y4	NE	NE
						P2Y6	NE	NE

^aAgonist potencies were calculated using a four-parameter logistic equation and the GraphPad software package (GraphPad, San Diego, CA). EC₅₀ values (mean ± standard error) represent the concentration at which 50% of the maximal effect was achieved. A maximal effect (100% apparent efficacy) equivalent to that of the cognate agonist for the tested receptor was observed unless where indicated for two compounds (1 and 3) that exhibited <50% of maximal effect at the highest 10 μM concentration tested at the P2Y₄ receptor and for two compounds (4 and 7) that produced apparently saturable activation that was much less than that observed with UTP at the P2Y₂ receptor. The data for 4 and 7 are denoted as consistent with partial agonist (partial ag.) activity. N = 4 – 6 conc.-effect curves for each analogue.

^bUp₂U. NE - no effect at 10 μM.

Scheme 1.

Reagents and conditions: (a) (i) IBX, CH₃CN; (ii) (EtO)₂P(O)C=PPh₃, DMSO (44%); (b) H₂, Pd/C, MeOH (84%); (c) TMSBr, CH₂Cl₂ (62%); (d) Bu₃N, CDI, DMF; (Bu₃NH)₂H₂P₂O₇; TEAB 1M (22%); (e) TMSBr, CH₂Cl₂ (65%).

To prepare the (diphospho)phosphonate derivative 3, vinyl phosphonate 11, obtained by hydrolysis of both the phosphonate esters and the acetonide upon treatment with TMSBr,¹⁵ was activated with CDI in DMF, followed by the addition of bis(tri-n-butylammonium)pyrophosphate.

For biological assays, the human P2Y₁, P2Y₂, P2Y₄, and P2Y₆ receptors, which are coupled to activation of phospholipase C (PLC) through G_q, were stably expressed in 1321N1 human astrocytoma cells, which lack a native response to nucleotides.^17,18 Activity at the human P2Y₁₄ receptor was followed by quantification of phosphoinositide hydrolysis in COS-7 cells coexpressing a phospholipase C-activating chimeric G protein, Gαq/i, which is activated by G_i-coupled receptors.²⁰ Concentration effect curves were generated for each of the phosphonate derivatives at the relevant receptor(s), and the EC₅₀ values were compared to the potency of the corresponding nucleotide in the native riboside series (Table 1). Representative data for the uridine 5′-(diphospho)phosphonate 2 and the vinylicdiphosphophosphonate 3 at the P2Y₂ receptor are illustrated in Fig. 1.

Figure 1.

Activity of agonists 2 and 3 at P2Y₂ receptors as indicated by activation of PLC in stably transfected astrocytoma cells. The effect of UTP corresponds to 100%.

Full agonist activity was observed with 1, 2, and 3 at the P2Y₂ receptor, although the potencies were reduced by 16–176-fold relative to ATP or UTP. Although UTP is also a potent agonist at the human P2Y₄ receptor, neither uridine 5′-(diphospho)phosphonate 2, which is the simple phosphonate equivalent of UTP, nor the vinyl phosphonate derivative 3 exhibited activity at this receptor. Compounds 2 and 3 were inactive at the P2Y₆ receptor. Uridine 5′-phosphophosphonate 4, which is the simple phosphonate equivalent of UDP, was a full agonist at the P2Y₆ receptor exhibiting a potency approximately 40-fold less than UDP. In contrast, uridine 5′-(glucose-[1′]phospho)-phosphonate 6 exhibited no activity at the UDP-glucose activated P2Y₁₄ receptor.

A surprising result from our tests of activity was that uridine 5′-phosphophosphonate 4 was essentially as potent for activation of the P2Y₂ receptor as for activation of the P2Y₆ receptor and had no activity at the P2Y₄ receptor. However, the maximal effect observed with 4 for activation of the P2Y₂ receptor was only ~50% of that observed with UTP (Fig. 2A). Although UMP has no effect, uridine 5′-methylene-phosphonate 7 also was a relatively potent (EC₅₀ = 1.6 ± 0.4 μM) agonist at the P2Y₂ receptor. Similar to 4, the maximal effect observed at the P2Y₂ receptor with 7 was <50% of that observed with UTP (Fig. 2B). Analogue 7 had no effect on the P2Y₄ and P2Y₆ receptors.

Figure 2.

Partial agonist activity of 4 (A) and 7 (B) at P2Y₂ receptors as indicated by activation of PLC in stably transfected astrocytoma cells. The effect of UTP corresponds to 100%.

Preliminary experiments examining the capacity of high concentrations of compounds 4 and 7 to antagonize activation of the P2Y₂ receptor by UTP failed to provide convincing evidence that these molecules interact with the orthosteric binding site of the receptor. Therefore, these novel analogues potentially activate the P2Y₂ receptor through an allosteric mechanism.