Pyrimidine nucleotides with 4-alkyloxyimino and terminal tetraphosphate δ-ester modifications as selective agonists of the P2Y4 receptor

Article abstract of DOI:10.1021/jm101591j

P2Y₂ and P2Y₄ receptors are G protein-coupled receptors, activated by UTP and dinucleoside tetraphosphates, which are difficult to distinguish pharmacologically for lack of potent and selective ligands. We structurally varied phosphate and uracil moieties in analogues of pyrimidine nucleoside 5′-triphosphates and 5′-tetraphosphate esters. P2Y₄ receptor potency in phospholipase C stimulation in transfected 1321N1 human astrocytoma cells was enhanced in N⁴-alkyloxycytidine derivatives. OH groups on a terminal δ-glucose phosphoester of uridine 5′-tetraphosphate were inverted or substituted with H or F to probe H-bonding effects. N⁴-(Phenylpropoxy)-CTP 16 (MRS4062), Up ₄-[1]3′-deoxy-3′-fluoroglucose 34 (MRS2927), and N ⁴-(phenylethoxy)-CTP 15 exhibit ≤10-fold selectivity for human P2Y₄ over P2Y₂ and P2Y₆ receptors (EC ₅₀ values 23, 62, and 73 nM, respectively). δ-3-Chlorophenyl phosphoester 21 of Up₄ activated P2Y₂ but not P2Y ₄ receptor. Selected nucleotides tested for chemical and enzymatic stability were much more stable than UTP. Agonist docking at CXCR4-based P2Y₂ and P2Y₄ receptor models indicated greater steric tolerance of N⁴-phenylpropoxy group at P2Y₄. Thus, distal structural changes modulate potency, selectivity, and stability of extended uridine tetraphosphate derivatives, and we report the first P2Y₄ receptor-selective agonists.

Full text of DOI:10.1021/jm101591j

ARTICLE
pubs.acs.org/jmc
Pyrimidine Nucleotides with 4-Alkyloxyimino and Terminal
Tetraphosphate δ-Ester Modifications as Selective Agonists
of the P2Y4 Receptor
Hiroshi Maruoka,† M. P. Suresh Jayasekara,† Matthew O. Barrett,‡ Derek A. Franklin,‡
Sonia de Castro,† Nathaniel Kim,† Stefano Costanzi,§ T. Kendall Harden,‡ and Kenneth A. Jacobson*,†
†Molecular Recognition Section, Laboratory of Bioorganic Chemistry, NIDDK, National Institutes of Health, Bethesda,
Maryland 20892-0810, United States
‡Department of Pharmacology, University of North Carolina, School of Medicine, Chapel Hill, North Carolina 27599-7365,
United States
§Laboratory of Biological Modeling, NIDDK, National Institutes of Health, Bethesda, Maryland 20892, United States
S Supporting Information
b
ABSTRACT: P2Y2 and P2Y4 receptors are G protein-coupled receptors,
activated by UTP and dinucleoside tetraphosphates, which are difficult
to distinguish pharmacologically for lack of potent and selective ligands.
We structurally varied phosphate and uracil moieties in analogues of
pyrimidine nucleoside 50-triphosphates and 50-tetraphosphate esters.
P2Y4 receptor potency in phospholipase C stimulation in transfected
1321N1 human astrocytoma cells was enhanced in N4-alkyloxycytidine
derivatives. OH groups on a terminal δ-glucose phosphoester of uridine
50-tetraphosphate were inverted or substituted with H or F to probe
H-bonding effects. N4-(Phenylpropoxy)-CTP 16 (MRS4062), Up4-
[1]30-deoxy-30-fluoroglucose 34 (MRS2927), and N4-(phenylethoxy)-
CTP 15 exhibit g10-fold selectivity for human P2Y4 over P2Y2 and
P2Y6 receptors (EC50 values 23, 62, and 73 nM, respectively). δ-3-
Chlorophenyl phosphoester 21 of Up4 activated P2Y2 but not P2Y4 receptor. Selected nucleotides tested for chemical and
enzymatic stability were much more stable than UTP. Agonist docking at CXCR4-based P2Y2 and P2Y4 receptor models indicated
greater steric tolerance of N4-phenylpropoxy group at P2Y4. Thus, distal structural changes modulate potency, selectivity, and
stability of extended uridine tetraphosphate derivatives, and we report the first P2Y4 receptor-selective agonists.
’ INTRODUCTION
receptors.13 In contrast, the 20-azido analogue 3 is one of the few
analogues of 1 that displays any selectivity (only 5-fold in
comparison to the P2Y2 receptor) for the P2Y4 receptor.8 There
are no selective antagonists of the P2Y4 receptor.
The eight subtypes of P2Y receptors are nucleotide-activated G
protein-coupled receptors (GPCRs), and four of these subtypes,
the P2Y2, P2Y4, P2Y6, and P2Y14 receptors, respond to uracil
nucleotides.1,2 The P2Y2 and P2Y4 receptors are activated by
uridine 50-triphosphate (UTP, 1, Chart 1) and its analogues and
are difficult to distinguish pharmacologically, and a prominent
need exists for definitive ligand tools.3,4 The P2Y2 receptor also is
activated by analogues of adenine 50-triphosphate.
Dinucleotide P2Y2 receptor agonists, such as 2 (Up4U,
INS365), which are more resistant to enzymatic hydrolysis than
1, are being developed clinically for the treatment of eye and
pulmonary diseases.5ꢀ7 Compounds 1 and 2 are not selective for
the P2Y2 receptor. However, several studies have explored the
structureꢀactivity relationship (SAR) of agonists8ꢀ13 and
antagonists14 at the P2Y2 receptor, and potent and selective
P2Y2 receptor agonists, such as 4 (MRS2698)10 and C-linked
nucleotides9 have been identified. The nucleotide agonists 5ꢀ7
also are P2Y2 receptor-selective relative to both P2Y4 and P2Y6
In the present study, we further investigated SARs at the P2Y2
and P2Y4 receptors through synthesis of nucleotides with addi-
tional substitutions of the uracil and phosphate moieties and
combinations thereof. These novel derivatives incorporated
groups such as 4-methoxyimino and its homologues on the
pyrimidine ring, found previously to enhance the potency of
analogues of uridine 50-diphosphate (UDP) at P2Y receptors.15
We also probed the effects of substituting one uridine moiety of 2
with various sugar and aliphatic and aromatic alcohol moieties.
Two lead δ-phosphoesters of 50-tetraphosphates that were
recently reported as full agonists of the P2Y2 receptor,13 that
is, 7 (MRS2768), which was P2Y2 receptor-specific, and 8
Received: December 15, 2010
Published: April 29, 2011
r
2011 American Chemical Society
4018
dx.doi.org/10.1021/jm101591j J. Med. Chem. 2011, 54, 4018–4033
|

Journal of Medicinal Chemistry
ARTICLE
Chart 1. Structures and Potencies of Prototypical Agonist Ligands for Studying P2Y2 and P2Y4 Receptors
(MRS2732), with 7- and 26-fold selectivity in comparison to
P2Y4 and P2Y6 receptors, were modified in the present study.
Although P2Y2 receptor selectivity was not improved by these
changes, several analogues exhibited important increases in
potency and selectivity at the P2Y4 receptor. This new chemical
synthesis was combined with molecular docking studies based on
a model derived from the recently solved structure of the CXCR4
chemokine receptor16 to interrogate the molecular basis for
selectivity of interaction of several of these molecules for the
P2Y4 receptor over the P2Y2 receptor.
phosphorus oxychloride, and the reaction mixture was treated
immediately with bis(tri-n-butylammonium) pyrophosphate to
produce the 50-triphosphate. The favored method for
preparation of the β,γ-dichloromethylene derivative 10 was by
condensation of dichloromethylenediphosphonic acid with ur-
idine 50-monophosphate (UMP) using N,N0-diisopropylcarbo-
diimide (DIC). An alternate method involving reaction
with UMP-morpholide required 2 weeks of reaction to reach
completion.
The synthesis of 50-tetraphosphate derivatives that were
modified at the terminal phosphates was performed using a
carbodiimide coupling reagent in the presence of magnesium
chloride (Schemes 5 and 6). The addition of MgCl2 dramatically
increased the reactivity during the coupling reaction of mono-
phosphate and triphosphate,17 but it impeded the coupling of
monophosphate and diphosphate (data not shown). Phenyl
monophosphates 57ꢀ62 were prepared by phosphorylation
with a dibenzyl-protected phosphate group followed by depro-
tection of the benzyl groups using trimethylsilyl bromide (TMS-
Br, Scheme 2).15 Hexose sugar monophosphates 68ꢀ72 were
phosphorylated at the 10-position by full acetylation and treat-
ment with phosphoric acid. The synthesis of glucuronic acid-1-
monophosphate 74 was performed by oxidation of commercially
available glucose-1-monophosphate using 2,2,6,6-tetramethylpi-
peridine-1-oxyl (TEMPO) as a catalyst.18
’ RESULTS
Chemical Synthesis. Novel nucleotide derivatives (Table 1)
were prepared by the synthetic routes shown in Schemes 1ꢀ6.
The potencies of nine known reference compounds (1, 6ꢀ9,
17ꢀ19, and 27) are also listed in Table 1.11ꢀ13 The types of
nucleotide modifications made included analogues of cytidine 50-
triphosphate (CTP) containing a N4-alkoxy group (Scheme 1),
50-tetraphosphates derivatives of 17 (Up4) containing a substi-
tuted δ-phenyl phosphoester group at the terminal phosphate
(Schemes 2 and 5), derivatives of 17 containing a sugar group at
the terminal phosphate (Schemes 3 and 6), and analogues of 1
containing a β,γ-dichloromethylene-bridge in the triphosphate
group (Scheme 4).
The synthesis of N4-alkoxy cytidines 39ꢀ42 from cytidine was
performed using corresponding alkoxy amines. The resulting
N4-alkoxy cytidines were phosphorylated by standard methods6 to
give the desired N4-alkoxy-CTP analogues 11ꢀ16 (Scheme 1).
In each case, the unprotected nucleoside was first treated with
Pharmacological Activity. The activation of phospholipase C
(PLC) by a range of concentrations of each newly synthesized
nucleotide derivative (10ꢀ37) was studied in [3H]inositol-
labeled 1321N1 human astrocytoma cells stably expressing
the human P2Y2, P2Y4, or P2Y6 receptor by following the
4019
dx.doi.org/10.1021/jm101591j |J. Med. Chem. 2011, 54, 4018–4033

Journal of Medicinal Chemistry
ARTICLE
Table 1. Relative Potencies of 1 and Its Analogues for Activation of the Human P2Y2, P2Y4, and P2Y6 Receptors (Unless Noted:
R = H; X, Y = O; and Z = O)
4020
dx.doi.org/10.1021/jm101591j |J. Med. Chem. 2011, 54, 4018–4033

Journal of Medicinal Chemistry
ARTICLE
Table 1. Continued
a Agonist potencies reflect stimulation of PLC in 1321N1 human astrocytoma cells stably expressing the human P2Y2, P2Y4, or P2Y6 receptor. Potencies
are presented in the form of EC50 values, which represent the concentration of agonist at which 50% of the maximal effect is achieved. These values were
determined using a four-parameter logistic equation and the GraphPad software package (GraphPad, San Diego, CA). The results are presented as
means ( standard errors and are the average of 3ꢀ6 different experiments with each molecule. b Agonist potencies for compounds 2, 6, 8, 9, 18, and 19
were from published reports.8,13 Compounds 1, 7, 10, 17, and 27 were reported in Ko et al.13 but reassayed in this study. c Not determined. d e50% effect
at 10 μM. Values of >10 μM were determined by extrapolation. NE-no effect at 10 μM.
Scheme 1. Synthesis of Alkoxyimino Derivativesa
a Reagents and conditions: (i) RONH2, pyridine, 100 °C, overnight. (ii) (a) Proton Sponge, POCl3, PO(OMe)3, 0 °C, 3 h; (b) Bu3N,
tributylammonium pyrophosphate, 0 °C, 10 min; (c) triethylammonium bicarbonate(aq), room temperature, 30 min.
methodology (see the Experimental Section) that we described
previously in detail.13,15,19
Table 1 illustrates analogues of 1 and 17 that were designed for
possible interaction with the P2Y2 and P2Y4 receptors and their
4021
dx.doi.org/10.1021/jm101591j |J. Med. Chem. 2011, 54, 4018–4033

Journal of Medicinal Chemistry
ARTICLE
Scheme 2. Synthesis of Aryl Phosphatesa
Scheme 5. Synthesis of Up4-δ-phenyl Ester Derivativesa
a Reagents and conditions: (i) CBr4, iPr2NEt, dibenzyl phosphite, MeCN,
room temperature, 2 h. (ii) TMS-Br, CH2Cl2, room temperature, 1.5 h.
a Reagents and conditions: (i) (a) DIC, DMF, room temperature, 3 h;
(b) aryl-1-phosphate, MgCl2, DMF, room temperature, overnight.
Scheme 3. Synthesis of Sugar-1-phosphatesa
Scheme 6. Synthesis of Up4-δ-sugar Derivativesa
a Reagents and conditions: (i) Ac2O, NaOAc, 5 h, 110 °C. (ii) H3PO4,
50 °C, 3 h. (iii) TEMPO, H2O, NaOH, NaOCl, 0 °C.
a Reagent and conditions: (i) (a) DIC or DCC, DMF, room tempera-
ture, 3 h; (b) sugar-1-phosphate, MgCl2, DMF, room temperature, 2 h to
overnight. Unless noted, Y = O.
Scheme 4. Synthesis of β,γ-Dichloromethylene-UTP 10a
28- and 32-fold in comparison to the P2Y2 and P2Y6 receptors,
respectively. N4-Phenylethoxy substitution of CTP resulted in a
molecule 15 that exhibited somewhat lower potency (EC50 = 73
nM) and selectivity (16- and 17-fold in comparison to the P2Y2
and P2Y6 receptors, respectively) than 16 for the P2Y4 receptor.
Introduction of either a sugar moiety or a substituted phenyl ring
at the terminal δ-phosphate was examined in the series of
50-tetraphosphates. We previously reported that introduction of a
glucose or phenyl ester group at the δ-phosphate resulted in
increased selectivity for the P2Y2 receptor against the P2Y4 and
P2Y6 receptors.13 A 3-Cl derivative 21 showed higher potency than
the phenyl derivative 7, but its P2Y2 receptor selectivity in compar-
ison to the P2Y6 receptor was decreased. Similarly, all substitutions
of the δ-phenyl ester examined (20ꢀ25) were either less potent or
less selective for the P2Y2 receptor than the unsubstituted 7.
Consistent with previous findings, introduction of a β,γ-di-
chloromethylene substitution in 24 abolished activity at the
three P2Y receptors tested. On the basis of the P2Y2 receptor
selectivity of Up4-δ-[1]glucose 8,13 substitutions of the δ-glucose
moiety were examined (26ꢀ37). The types of glucose modifica-
tions included were inversion of hydroxyl groups (27ꢀ29),
replacement of a hydroxyl group with F (33ꢀ37), and other
omission or other replacement of hydroxyl functionality (30ꢀ32).
These modifications resulted in greatly varied potency and
selectivity for the P2Y receptors. Inversion of the 30-hydroxyl
group in the allose derivative 28 increased P2Y4 receptor potency
11-fold, but inversion of the 20-hydroxyl group in the mannose
derivative 29 decreased potency at P2Y2, P2Y4, and P2Y6
a Reagents and conditions: From 75, (i) (a) DIC, DMF, room tem-
perature, 3 h; (b) PO(OH)2CCl2PO(OH)2, MgCl2, DMF, room
temperature, overnight. From 76, (ii) PO(OH)2CCl2PO(OH)2,
DMF, room temperature, 2 weeks.
biological potencies. Introduction of β,γ-phosphonomethylene
bridges, which enhance the stability of nucleotides against the
action of ectonucleotidases, resulted in compounds 6 and 10 that
were inactive at the P2Y4 receptor.13
The possibility of a hydrophobic pocket beyond the 4-position
of the pyrimidine ring in the P2Y2 or P2Y4 receptor structure was
probed in analogues of 1 with N4-alkoxyimino moieties 11ꢀ16.
One of these molecules, N4-methoxy-CTP 11,15 was more
potent than the native ligand 1 at both subtypes. Within a related
series of homologous arylalkoxy derivatives, the N4-phenylpro-
poxy substitution of CTP in 16 was found to optimally increase
activity at the P2Y4 receptor with an EC50 value of 23 nM
(Figures 1 and 2). This compound displayed selectivities of
4022
dx.doi.org/10.1021/jm101591j |J. Med. Chem. 2011, 54, 4018–4033

Journal of Medicinal Chemistry
ARTICLE
Molecular Modeling. In light of the SAR data generated
through this work, in particular the different selectivity profile of
compounds 11 and 16, we revised our experimentally supported
modeling hypothesis for the binding model of pyrimidine-based
nucleotides to the P2Y2 and P2Y4 receptors, through the
incorporation of further experimental data. These new models
were not used to design the compounds but to offer a putative
structural interpretation of the results. Note that to facilitate the
comparison among receptors, throughout this section, amino
acid residues are designated both with their residue number in
the P2Y2 and P2Y4 sequence and the universal GPCR residue
identifier defined by the Ballesteros and Weinstein.20,28,29
In the absence of experimentally elucidated structures of any of
the members of the P2Y family, in the past decade, we have
constructed models of the P2Y2 and P2Y4 receptors through
homology modeling based on rhodopsin, which, at the time, was
the only crystallographically solved GPCR, followed by molec-
ular docking.8,10,20 We incorporated into the models several
hypotheses derived from experiments. These included the pre-
sence of two disulfide bridges and a salt bridge that putatively
contributes to the shape of the extracellular domains of the
receptors. Data from mutagenesis studies revealed that the
residues involved in the formation of these extracellular bridges
are fundamental to the function of P2Y receptors.22ꢀ24 The
binding mode of the ligand was also driven by mutagenesis and
SAR data. Specifically, our published models were constructed
using as anchors for the binding of the phosphate moieties of the
nucleotides three cationic residues located in transmembrane
R-helix 3 (TM3), TM6, and TM7 that are essential for ligand
recognition according to mutagenesis data.24ꢀ26 The Northern
(N) conformation adopted by the ribose moiety of 1 and the
orientation of the nucleobase toward the opening of the inter-
helical cavity are also supported by SAR data.11,15,21,30
Figure 1. Activity of native agonist 1 and alkyloxyimino derivatives
11, 15ꢀ16 at the P2Y4 receptor as indicated by activation of PLC in
1321N1 human astrocytoma cells stably expressing the human P2Y4
receptor.
Our rhodopsin-based models do not explain the different
selectivity profiles of compounds 11 and 16, because in those
models the area that surrounds the N4 substituents of the two
compounds is completely conserved between the P2Y2 and the
P2Y4 receptors (Figure 3). Here, we propose revised models of
the P2Y2 and P2Y4 receptors that incorporate the experimentally
supported features of the previous models and are consistent with
SAR data gathered in this work. The new models are based on the
recently solved X-ray crystal structure of the CXCR4 chemokine
receptor,16 which, among all of those currently available, is the
closest to the P2Y receptors.20 Among the crystallographically
solved GPCRs, this receptor shares the highest sequence similar-
ity with P2Y receptors, showing 24.6 and 25.7% of sequence
identity with the P2Y2 and the P2Y4 receptor—the sequence
identity was calculated according to the alignment shown in
Figure 4 and refers to the entire portion of the CXCR4 receptor
that was solved crystallographically, with the exclusion of the
C-terminal domain. Moreover, the CXCR4 contains a disulfide
bridge, putatively present also in the P2Y2 and P2Y4 receptors,
connecting the N-terminus (through a Cys located two positions
upstream when compared to the P2Ys) with the boundary
between extracellular loop 3 (EL3) and TM7 (through a Cys
located at position 7.25, as in the P2Y receptors)—see Figure 4.
Before the solution of the CXCR4 structure, none of the
structurally characterized GPCRs featured a similar disulfide
bridge. Like the previous rhodopsin-based models, our current
models incorporate all of the experimentally derived structural
features mentioned above, including the aforementioned pair of
extracellular disulfide bridges and the salt bridge connecting EL2
Figure 2. Activity of compound 16 [N4-(phenylpropoxy)-CTP] at
P2Y2, P2Y4, and P2Y6 receptors as indicated by activation of PLC in
1321N1 human astrocytoma cells stably expressing the human P2Y2,
P2Y4, or P2Y6 receptor. The effect of 1 corresponds to 100%.
receptors. Fluoro substitution at the 20-position of the terminal
δ-glucose moiety in 33 greatly reduced activity at all three receptor
subtypes, suggesting that the H-bond-donating property of the
20-hydroxyl group is necessary for this biological activity. The
hydroxymethylene group appeared to be important for activity
at the P2Y2 receptor, because the xylose derivative 30 exhibited
reduced activity (9-fold less potent than 8) at this subtype.
The 30- and 40-fluoro derivatives 34 and 37 maintained activity
at the P2Y2 receptor. However, fluoro substitution at the
30-position of the terminal glucose moiety in 34 greatly increased
potency (EC50 = 62 nM) at the P2Y4 receptor. Thus, P2Y4
receptor selectivities of 11.5- and 15.3-fold were observed for this
analogue in comparison to the P2Y2 and P2Y6 receptors,
respectively. Analogues 35 and 36 that included combination
of 30-deoxy-30-fluoroglucose with the N4-benzyloxyimino or
N4-methoxyimino modification in the nucleobase were less
active and selective than unmodified 34. Thus, simultaneous
introduction of these two P2Y4 receptor-directing modifications
did not produce further enhancement of potency or selectivity.
4023
dx.doi.org/10.1021/jm101591j |J. Med. Chem. 2011, 54, 4018–4033

Journal of Medicinal Chemistry
ARTICLE
Figure 3. Rhodopsin-based molecular models of the P2Y2 and P2Y4 receptors.8 The second EL (EL2) is schematically represented as a green tube. The area
that in these models would surround the N4 substituents of the imino derivatives described in this article, is completely conserved between the P2Y2 and the
P2Y4 receptor. This is in contrast with the fact that large N4 substituents are better tolerated by the P2Y4 than the P2Y2 receptor. Two nonconserved EL2
residues that could explain the selectivity profile of the N4 substituent compounds according to the CXCR4 models (see Figure 5) are not positioned in a way
that would allow interactions with such moieties in the rhodopsin-based models. Receptor residues are represented with gray carbons, while ligands are
represented with yellow carbons. See the figure of Jacobson et al. in ref 8 for further details on the rhodopsin-based models.
Figure 4. Sequence alignment of CXCR4, P2Y2, andP2Y4 receptors used for the construction of the homology models. The sequence of the P2Y1 receptor, on
which most of the mutagenesis studies were performed, is also shown. Green dots indicate the two nonconserved EL2 residues putatively responsible for the
selectivity of compound 16 for the P2Y4 receptor. Blue dots indicate the conserved cationic residues putatively responsible for coordination of the negatively
charged phosphate groups of the ligands. An additional cationic residue, Lys2897.36 of the P2Y4 receptor, is proposed to interact with the R-phosphate. The
cysteine residues involved in the formation of disulfide bridges are encased in a dark yellow box and connected by a dark yellow line. The glutamate and arginine
residues involved in the formation of the salt bridge that putatively connects EL2 to the extracellular end of TM6, conserved also in the P2Y1 receptor, are
encased in a red and a blue box, respectively, and are connected by a red and blue line. The secondary structure of thetemplate is indicated by a color-coded bar,
with R-helices in red, β-strands in yellow, and turns in blue. For each TM, black dots indicate the X.50 position, as defined in the GPCR indexing system.
Transmembrane helices (TMs), extracellular loops (ELs), and intracellular loops (ILs) are indicated by labels. The figure was generated with MOE.40
with the extracellular end of TM6. These features, contribute to
shaping of the opening of the interhelical cavity lined by residues
located in the extracellular N-terminus, the three ELs, and all of
the TMs (Figure 5).
4024
dx.doi.org/10.1021/jm101591j |J. Med. Chem. 2011, 54, 4018–4033

Journal of Medicinal Chemistry
ARTICLE
Figure 5. Contour of the interhelical cavity (in gray) of the P2Y2 and P2Y4 receptor deduced from the analysis of the new CXCR4 homology models.
According to these models, there is only one region, indicated by a dashed yellow circle, where the cavity is significantly smaller in the P2Y2 than in the
P2Y4 receptor. This difference is due to a nonconserved amino acid in EL2, which is a threonine (Thr182) in the P2Y4 receptor and a bulkier arginine
(Arg180) in the P2Y2 receptor. Of note, an arginine residue is present at the corresponding position of the CXCR4 receptor (Arg183). The orientation
of Arg180 in our model of the P2Y2 receptor follows the orientation of Arg183 in the crystal structure of the CXCR4 receptor. The disulfide bridge that
putatively links the N-terminus with EL3 and the salt bridge that putatively links TM6 with EL2 are shown. The disulfide bridge that links TM3 with EL2
is not clearly visible because it is obscured by the interhelical cavity. Compounds 11 and 16 are shown within the pockets of the P2Y2 and P2Y4 receptors,
respectively. The backbone of the receptor is schematically represented as a ribbon colored with a continuum spectrum going from red, at the
N-terminus, to purple, at the C-terminus. For convenience, TM7 is not shown.
We used the new CXCR4-based in silico structures of the P2Y2
and P2Y4 receptors to construct models of the putative com-
plexes of the two receptors with compounds 11 and 16. Im-
portantly, we incorporated in the binding mode and in the
conformation of the two ligands all of the experimentally sup-
ported features shown in our previous models.8,10,20 Moreover,
we incorporated novel experimental information that we gathered
through NMR spectroscopy and X-ray crystallography. Specifi-
cally, the tautomeric form of the imino group and its conforma-
tion were determined through NMR spectroscopy (chemical shift
of NH at 3-position of 39 and nuclear Overhauser effect between
MeO and NH). Additionally, an X-ray crystallographic structure
was obtained for N4-t-BuO-cytosine (Supporting Information),
which is in close agreement with a literature report for N4-MeO-
cytosine.27 The data clearly showed the presence of a hydrogen
atom at the 3-position and the absence of a hydrogen atom at the
N4-position. Bond lengths also indicate that the compound
adopts an exo double bond tautomeric form.
the binding of nucleotides appears shifted toward the extracell-
ular side in the new CXCR4-based models as compared to those
previously reported (Figure 6).8,10,20 This is due to the fact that in
the new models the side chain of Phe3.32 (113 for P2Y2, 115 for
P2Y4)—a residue common to all P2Y receptors and conserved as
an aromatic residue (Tyr1163.32) in the CXCR4 receptor—is
parallel to the plane of the membrane. Because of this orienta-
tion, which is consistent with the one shown by Tyr1163.32 in the
crystal structure of the CXCR4 receptor, Phe3.32 limits the depth
of the binding cavity and prevents the ligands from docking at a
position more distal to the extracellular side. Our previous
models, which were built prior to the availability of a template
containing an aromatic residue at position 3.32, featured the side
chain of Phe3.32 oriented perpendicularly to the plane of the
membrane and in π-stacking with the nucleobase of the ligands
and allowed for a deeper binding pocket.8,10,20 Because of the
shallower binding mode proposed here in comparison to our
previous rhodopsin-based models, the CXCR4-based models are
compatible with the influence of the substituent at position 4 of
the pyrimidine ring profile of compounds of the newly synthe-
sized alkoxyimino derivatives on their selectivity for the P2Y4
Interestingly, the new models show several distinctive features
that make them compatible with the selectivity profile of com-
pounds 11 and 16 for the P2Y2 and P2Y4 receptors. Specifically,
4025
dx.doi.org/10.1021/jm101591j |J. Med. Chem. 2011, 54, 4018–4033

Journal of Medicinal Chemistry
ARTICLE
Figure 6. Previously published rhodopsin-based model (A, in complex with 1) and new CXCR4-based model of the P2Y4 receptor (B, in complex with 16).
One residue of the CXCR4 receptor (Tyr1163.32, in red) was key to deduce the putative rotameric state of the corresponding residue in the P2Y2 and P2Y4
receptors (Phe1133.32 for P2Y2 and Phe1153.32 for P2Y4, in gray) in the new models. The CXCR4-based rotameric state of Phe3.32 caused a significant change of
the ligand binding mode with respect to the previously published one. The backbone of TM3 is schematically represented as a yellow ribbon.
versus the P2Y2 receptor (Figure 5). In particular, the phenyl
moiety of the N4 substituent of compound 16 protrudes toward
EL2, fitting into a cavity lined by two EL2 residues, namely,
Thr182 and Leu184 in the P2Y4 receptor. These two residues are
not conserved in the P2Y2 receptor, where their place is taken by
Arg180 and Thr182, respectively. Our models suggest that these
differences may account for the P2Y4 receptor selectivity of 16,
especially in light of the fact that the sterically bulky Arg180,
conserved in the CXCR4 receptor and oriented as in the CXCR4
crystal structure, seems to considerably reduce the volume of the
cavity of the P2Y2 receptor. As evident from Figure 5, this is the
only region of the interhelical cavity that shows a significantly
smaller volume in the P2Y2 than in the P2Y4 receptor.
Stability of Nucleotide Derivatives. Chemical and enzymatic
instability of nucleotide analogues in biological systems is a major
limitation in their use as pharmacological probes. Therefore,
selected nucleotide derivatives were evaluated for stability during
prolonged exposure to two different conditions. The derivatives
were incubated at 37 °C in aqueous medium at either low pH or
in the presence of membranes prepared from 1321N1 human
astrocytoma cells,15 which contain ectonucleotidases and are
representative of cells of the mammalian central nervous system.
At regular intervals, aliquots were taken for high-performance
liquid chromatography (HPLC) analysis (Figure 7).
All of the nucleotide analogues tested were more stable in
the presence of cell membranes than the native agonist 1, and the
most stable in the group was the Up4-δ-phenyl ester 7. The
4-chlorophenyl analogue 20 appeared to be more stable than the
corresponding 4-methoxy 22 and 4-nitro 24 analogues. Among
the nucleotide analogues tested, only Up4-[1]glucose 8 was less
stable at pH 1.5 than 1. In general, the Up4-δ-phenyl esters
displayed a relatively high degree of stability under both condi-
tions in comparison to 1. The stability at pH 1.5 was also
determined for the potent P2Y4 receptor agonists 15, 16, and
34. The 50-triphosphate derivatives 15 and 16 were similar in
stability to 1, and 34 was much more labile, similar to the closely
related δ-glucose phosphoester 8 (Supporting Information).
antagonists and the fact that many analogues of 1 and diuridine
tetraphosphates activate both subtypes. Two classical, general
antagonists for P2 receptors, suramin and Reactive Blue 2, have
been used as slightly selective antagonists of the P2Y2 and P2Y4
receptors, respectively.4,31,32 In the rat, inosine-50-triphosphate was
shown to be 13-fold selective as an agonist at the P2Y4 receptor in
comparison to the P2Y2 receptor.32 The work described here
identifies modifications that provide important new directions for
introduction of agonist selectivity for P2Y4 receptor ligands.
The P2Y4 receptor is often expressed in the same tissues or, in
some cases, the same cells as the widely distributed P2Y2 receptor.
This is the case for epithelial cells of the intestine. The study of
knockout mice has revealed that the stimulatory effect of 1 on
chloride secretion is mediated by the P2Y4 receptor,33,34 whereas
its inhibitory effect on sodium absorption involves the P2Y2
receptor.35 Protective effects of P2Y4 receptor activation were
observed during cardiac ischemia studied in murine HL-1
myocytes.36 P2Y4 mRNA isexpressed in microvascularendothelial
cells isolated from mouse heart but not lung.37 The effects of 1 on
migration, proliferation, and VCAM-1 expression in these cells
were greatly reduced in P2Y4 receptor-deficient mice.37 Macro-
phage adhesion induced by 1 also was reduced, and P2Y4 receptor-
null mice displayed defective angiogenesis in in vivo models.37 In
contrast, P2Y2 receptor-deficient mice exhibit reduced effects of 1
on leukocyte adhesion, eosinophil migration, and VCAM-1
expression.38 Thus, selective P2Y4 receptor agonists might prove
useful in the treatment of constipation, cardiac ischemia, or
inflammation and exhibit reduced proinflammatory effects in the
lung due to lack of interaction with the P2Y2 receptor.
Substitutions at the 4-position of the nucleobase were ex-
plored through the synthesis of imino derivatives of various sizes.
The exploration of this region, done through a classic medicinal
chemistry approach, was prompted by the observation, based on
published models of the P2Y2 and P2Y4 receptors,8,10,20 that
there might have been some space available for the expansion of
the ligands around that position. This process led to the synthesis
of the first P2Y4-selective compounds. In particular, while
N4-(methyl)-CTP (11) was unselective for the two subtypes,
N4-(phenylpropoxy)-CTP (16) and N4-(phenylethoxy)-CTP
(15) were selective for the P2Y4 receptor with EC50 values of
23 and 73 nM, respectively. The activity of these CTP analogues
contrasts with the weak interaction of CTP itself at the P2Y2 and
’ DISCUSSION
Pharmacological resolution of responses mediated by P2Y2
versus P2Y4 receptors has been problematic due to lack of selective
4026
dx.doi.org/10.1021/jm101591j |J. Med. Chem. 2011, 54, 4018–4033

Journal of Medicinal Chemistry
ARTICLE
Figure 7. (A) Time course of the hydrolysis of various nucleotides incubated at 37 °C and pH 1.5. The assays were started with 0.2 mg/mL of the
compound. The concentration of analogues in the medium was determined using HPLC analysis of aliquots. Data are expressed as a percent of the initial
peak. (B) Time course of the degradation of various nucleotides in the presence of membranes of 1321N1 astrocytoma cells. The cells were
homogenized, and the nuclear fraction was removed by centrifugation. The resultant pellets were suspended in Tris buffer (pH 7.4). The membrane was
diluted with Dulbecco's PBS (pH 7.4) and homogenized before use for the stability check. The final protein concentration was 14.9 μg/mL. The assays
were started with 0.2 mg/mL compound. Concentrations of molecules were analyzed using HPLC. Data are expressed as a percent of initial peak.
P2Y4 receptors.8 Because our rhodopsin-based models could not
explain the different selectivity profiles of the alkoxyimino
derivatives, we reconstructed molecular models of the P2Y2
and P2Y4 receptors based on the recently disclosed structure
of the CXCR4 receptor (Figure 5).16 The new models incorpo-
rate all of the experimentally supported features of the old ones
but show a binding mode in which the ligands are slightly shifted
toward the extracellular side (Figure 6), which makes them
compatible with the new data. In particular, the large substituent
of compound 16 protrudes toward EL2, fitting into a cavity
substantially larger in the P2Y4 receptor than in the P2Y2
receptor, due to the nonconservation of two specific residues.
This hypothesis is in good agreement with the selectivity for the
P2Y4 receptor shown by compound 16.
P2Y4 selective (EC50 value = 62 nM). However, neither
δ-phenyl esters nor methylenephosphonates were useful for
achieving selectivity for the P2Y4 receptor. Up4-δ-3-chloro-
phenyl ester 21 was inactive at P2Y4 receptors and displayed
an EC50 value at the P2Y2 receptor of 0.84 μM. The chemical
and enzymatic stability of selected derivatives in this series was
examined by HPLC to indicate greatly enhanced stability of
the δ-phenyl phosphoesters but not the δ-glucose phosphoe-
sters in comparison to 1. Thus, additional SAR studies will be
needed to enhance the stability of P2Y4 receptor selective
agonists. In particular, these studies suggest that the potency,
selectivity, and stability of extended uridine tetraphosphate
derivatives may be modulated by distal structural changes.
N4-Alkoxyimino analogues were analyzed in a molecular
model, which indicated greater steric tolerance of the
N4-phenylpropoxy group in the P2Y4 receptor.
Moreover, by carefully probing substitution of Up4-[1]glucose,
we identified Up4-[1]30-deoxy-30-fluoroglucose (34) as modestly
4027
dx.doi.org/10.1021/jm101591j |J. Med. Chem. 2011, 54, 4018–4033

Journal of Medicinal Chemistry
ARTICLE
In conclusion, we have synthesized novel uracil nucleotide
derivatives that are directed toward selective activation of the
P2Y4 receptor. A series of homologous alkoxyimino derivatives
provided an entry to novel selectivity at this subtype. Several of
these compounds should be suitable for use as pharmacological
probes of the physiological roles of the P2Y4 receptor.
548.9035. HPLC RT 9.7 min (99%) in solvent system A, 1.1 min (98%)
in system B.
General Procedure for the Preparation of Nucleoside
Triphosphates (11ꢀ16). A solution of the nucleoside 39ꢀ4415
(0.073 mmol) and Proton Sponge (24 mg, 0.11 mmol) in trimethyl
phosphate (0.4 mL) was stirred for 10 min at 0 °C. Then, phosphorus
oxychloride (0.013 mL, 0.13 mmol) was added dropwise, and the
reaction mixture was stirred for 2 h at 0 °C. A solution of tributylam-
monium pyrophosphate (0.8 mL, 0.44 mmol) and tri-n-butylamine
(0.069 mL, 0.29 mmol) in DMF (1 mL) was added, and stirring was
continued at 0 °C for additional 10 min. A 0.2 M concentration of
triethylammonium bicarbonate solution (1.5 mL) was added, and the
clear solution was stirred at room temperature for 1 h. After solvents
were removed, the residue was purified by Sephadex-DEAE A-25 resin
and preparative HPLC.
’ EXPERIMENTAL SECTION
Chemical Synthesis. 1H NMR spectra were obtained with a
Varian Gemini 300 or Varian Mercury 400 spectrometers using D2O,
CDCl3, or DMSO-d6 as a solvent. The chemical shifts are expressed as
relative ppm from HOD (4.80 ppm). 31P NMR spectra were recorded at
room temperature by use of Varian XL 300 (121.42 MHz) or Varian
Mercury 400 (162.10 MHz) spectrometers; orthophosphoric acid
(85%) was used as an external standard. In several cases, the signal of
the terminal phosphate moiety was not visible due to high dilution.
The purity of final nucleotide derivatives was checked using a Hewlett
Packard 1100 HPLC equipped with a Zorbax SB-Aq 5 μm analytical
column (50 mm ꢁ 4.6 mm; Agilent Technologies Inc., Palo Alto, CA).
N4-Methoxycytidine-50-triphosphate Triethylammonium
Salt (11). Compound 11 (37.6 mg, 100%) was obtained as a white
1
solid using N4-methoxycytidine 39 (20 mg, 0.073 mmol). H NMR
(D2O): δ 7.26 (d, J = 8.2 Hz, 1H), 5.96 (d, J = 6.4 Hz, 1H), 5.82 (d, J =
8.2 Hz, 1H), 4.46ꢀ4.41 (m, 1H), 4.41ꢀ4.37 (m, 1H), 4.28ꢀ4.15 (m,
3H), 3.80 (s, 3H). 31P NMR (D2O): δ ꢀ11.39 (br), ꢀ22.90 (br).
HRMS-EI found, 511.9873 (M ꢀ Hþ)ꢀ. C10H17N3O15P3 requires
511.9857; purity >98% by HPLC (system B: 9.08 min).
Mobile phase: linear gradient solvent system:
5 mM TBAP
(tetrabutylammonium dihydrogenphosphate)ꢀCH3CN from 80:20 to
40:60 in 13 min; the flow rate was 0.5 mL/min. Peaks were detected by
UV absorption with a diode array detector at 254, 275, and 280 nm. All
derivatives tested for biological activity showed >97% purity by HPLC
analysis (detection at 254 nm).
N4-Ethoxycytidine-50-triphosphate Triethylammonium
Salt (12). Compound 12 (21.3 mg, 58%) was obtained as a white solid
using N4-ethoxycytidine 40 (20 mg, 0.070 mmol). 1H NMR (D2O): δ
7.26 (d, J = 7.8 Hz, 1H), 5.96 (d, J = 6.6 Hz, 1H), 5.82 (d, J = 7.8 Hz, 1H),
4.47ꢀ4.42 (m, 1H), 4.41ꢀ4.34 (m, 1H), 4.28ꢀ4.22 (m, 2H),
4.21ꢀ4.13 (m, 1H), 4.05 (dd, J = 14.2, 7.1 Hz, 2H). 31P NMR
(D2O): δ ꢀ11.29 (br), ꢀ22.10 (br). HRMS-EI found, 526.0029
(M ꢀ Hþ)ꢀ. C11H19N3O15P3 requires 526.0029; purity >98% by
HPLC (system B: 10.6 min).
High-resolution mass measurements were performed on Micromass/
Waters LCT Premier electrospray time of flight (TOF) mass spectro-
meter coupled with a Waters HPLC system, unless noted. Purification of
the nucleotide analogues for biological testing was carried out on
(diethylamino)ethyl (DEAE)-A25 Sephadex columns with a linear
gradient (0.01ꢀ0.5 M) of 0.5 M ammonium bicarbonate as the mobile
phase. Compound 10 was purified by Sephadex alone (and isolated in
the ammonium salt form), and compounds 11ꢀ16 were additionally
purified by HPLC with a Luna 5 μ RP-C18(2) semipreparative column
(250 mm ꢁ 10.0 mm; Phenomenex, Torrance, CA) and using the
following conditions: flow rate of 2 mL/min; 10 mM triethylammonium
acetate (TEAA)ꢀCH3CN from 100:0 to 95:5 (system A) [or up to 99:1
to 50:50 (system B)] in 30 min (and isolated in the triethylammonium
salt form). All other compounds were purified only by HPLC. All other
reagents were of analytical grade and were purchased from Sigma-
Aldrich (St. Louis, MO).
N4-t-Butyloxycytidine-50-triphosphate Triethylammonium
Salt (13). Compound 13 (7.6 mg, 22%) was obtained as a white solid
using N4-t-butoxycytidine 41 (20 mg, 0.063 mmol). 1H NMR (D2O): δ
7.24 (d, J = 8.2 Hz, 1H), 5.97 (d, J = 6.4 Hz, 1H), 5.87 (d, J = 8.2 Hz, 1H),
4.43ꢀ4.35 (m, 2H), 4.26ꢀ4.12 (m, 3H), 1.31 (s, 9H). 31P NMR (D2O):
δ ꢀ9.85 (br), ꢀ11.36 (br), ꢀ22.99 (br). HRMS-EI found, 554.0342
(M ꢀ Hþ)ꢀ. C13H23N3O15P3 requires 554.0341; purity >98% by HPLC
(system B: 9.59 min).
N4-Benzyloxycytidine-50-triphosphate Triethylammonium
Salt (14). Compound 14 (16.7 mg, 49%) was obtained as a white solid
using N4-benzyloxycytidine 42 (20 mg, 0.057 mmol). 1H NMR (D2O): δ
7.46ꢀ7.35 (m, 5H), 7.22 (d, J = 8.4Hz, 1H), 5.95 (d, J = 6.5Hz, 1H), 5.76
(d, J = 8.4Hz, 1H), 5.05 (s, 2H), 4.43ꢀ4.32 (m, 2H), 4.26ꢀ4.13 (m, 3H).
31P NMR (D2O): δ ꢀ10.18 (br), ꢀ11.56 (d, J = 14.6 Hz), ꢀ23.21
(t, J = 15.0 Hz). HRMS-EI found, 588.0186 (M ꢀ Hþ)ꢀ. C16H21N3O15P3
requires 588.0200; purity >98% by HPLC (system B: 10.14 min).
N4-Phenylethoxycytidine-50-triphosphate Triethylammo-
nium Salt (15). Compound 15 (1.29 mg, 5.5%) was obtained as a
white solid using N4-(2-phenylethoxy)-cytidine 43 (10 mg, 0.027 mmol).
1H NMR (D2O): δ 7.40ꢀ7.29 (m, 5H), 7.23 (d, J = 7.6 Hz, 1H), 5.94 (d,
J = 6.0 Hz, 1H), 5.81 (d, J = 8.4 Hz, 1H), 4.43ꢀ4.38 (m, 2H), 4.28ꢀ4.25
(m, 5H), 3.00 (dd, J = 6 Hz, J = 6.4 Hz). HRMS-EI found, 602.0342
(M ꢀ Hþ)ꢀ. C17H23N3O15P3 requires 602.0318; purity >98% by HPLC.
N4-Phenylpropoxycytidine-50-triphosphate Triethylam-
monium Salt (16). Compound 16 (0.83 mg, 11%) was obtained as
a white solid using N4-(3-phenylpropoxy)-cytidine 44 (5 mg, 0.013
mmol). 1H NMR (D2O): δ 7.38ꢀ7.31 (m, 5H), 7.23 (d, J = 8.4 Hz, 1H),
5.95 (d, J = 6.0 Hz, 1H), 5.78 (d, J = 8.4 Hz, 1H), 4.42ꢀ4.38 (m, 2H),
4.25ꢀ4.23 (m, 3H), 4.06 (dd, J = 5.5 Hz, J = 6.1 Hz, 2H), 2.75 (dd, J =
7.28 Hz, J = 7.45 Hz, 2H), 4.20 (m, 2H). 31P NMR (D2O): δ ꢀ5.51 (m),
ꢀ10.34 (m), ꢀ10.81 (m). HRMS-EI found, 616.0486 (M ꢀ Hþ)ꢀ.
C18H25N3O15P3 requires 616.0496; purity >98% by HPLC.
Uridine-50-β,γ-dichloromethylene Triphosphate Ammo-
nium Salt (10). Uridine monophosphate sodium salt (300 mg,
0.93 mmol) and the dichloromethylenediphosphonic acid disodium salt
(1 g, 3.46 mmol) were converted to the tributylammonium salts by
treatment with ion-exchange resin [DOWEX 50WX2-200 (H)] and tri-
n-butylamine. After the water was removed, the obtained tributylammo-
nium
salts were dried under high vacuum for 1 h. DIC (76.20 μL, 0.90 mmol)
was added to a solution of uridine monophosphate tributylammonium
salt (0.93 mmol) in N,N-dimethylformamide (DMF) (3 mL). After the
reaction mixture was stirred at room temperature for 3 h, a solution of
dichloromethylenediphosphonic acid tributylammonium salt (3.46
mmol) and MgCl2 (171.4 mg, 1.8 mmol) in DMF (2 mL) was added.
The reaction mixture was stirred at room temperature overnight. After
the solvent was removed, the MgCl2 was removed by treatment with ion-
exchange resin [DOWEX 50WX2-200 (H)] and ammonia bicarbonate,
and the product 10 was purified by ion-exchange column chromatog-
raphy with a Sephadex-DEAE A-25 resin. 1H NMR (D2O): δ 7.96 (d, J =
7.8 Hz, 1H), 5.96 (m, 2H), 4.43 (m, 1H), 4.37 (m, 1H), 4.36 (m, 2H).
31P NMR (D2O): δ ꢀ7.77 (d, J = 18.3 Hz), ꢀ10.54 (s), ꢀ10.79 (s).
HRMS-EI found, 548.9043 (M ꢀ H)þ. C10H14Cl2N2O14P3 requires
4028
dx.doi.org/10.1021/jm101591j |J. Med. Chem. 2011, 54, 4018–4033

Journal of Medicinal Chemistry
ARTICLE
General Procedure for the Preparation of Nucleoside
Tetraphosphate Phenyl Derivatives (20ꢀ25). The correspond-
ing aryl monophosphate15 (0.036 mmol) and 1 as a trisodium salt (10
mg, 0.018 mmol) were converted to the tributylammonium salts by
treatment with ion-exchange resin [DOWEX 50WX2-200 (H)] and tri-
n-butylamine. After the water was removed, the obtained tributylam-
monium salts were dried under high vacuum for 1 h. DIC (3.05 μL,
0.036 mmol) was added to a solution of uridine triphosphate tributy-
lammonium salt (0.018 mmol) in DMF (0.2 mL). After the reaction
mixture was stirred at room temperature for 1 h, a solution of the
corresponding aryl monophosphate tributylammonium salt (0.036
mmol) and MgCl2 (3.43 mg, 0.036 mmol) in DMF (0.1 mL) was
added. The reaction mixture was stirred at room temperature overnight.
After the solvent was removed, the MgCl2 was removed by treatment
with ion-exchange resin [DOWEX 50WX2-200 (H)] and ammonia
bicarbonate or tri-n-butylamine, and the residue was purified by a
semipreparative HPLC.
Uridine-50-(4-chlorophenyl)-tetraphosphate Triethylam-
monium Salt (20). Compound 20 (0.3 mg, 6%) was obtained as a
white solid using 1 (12 mg, 0.022 mmol). 1H NMR (D2O): δ 7.93 (d, J =
7.5 Hz, 1H), 7.38 (d, J = 8.4 Hz, 2H), 7.25 (d, J = 8.4 Hz, 2H), 5.97 (d, J =
5.4 Hz, 1H), 5.93 (d, J = 7.8 Hz, 1H), 4.42ꢀ4.30 (m, 2H), 4.27ꢀ4.21
(m, 3H). 31P NMR (D2O): δ ꢀ11.20 (br), ꢀ15.74 (br), ꢀ23.04 (br).
HRMS-EI found, 672.9218 (M ꢀ Hþ)ꢀ. C15H18N2O18P4Cl requires
672.9194; purity >98% by HPLC (system B: 11.9 min).
(m, 2H), 4.41ꢀ4.30 (m, 2H), 4.25ꢀ4.19 (m, 3H). 31P NMR (D2O): δ
ꢀ11.18 (d, J = 14.5 Hz), ꢀ16.89 (d, J = 14.5 Hz), ꢀ22.80ꢀ23.60 (m).
HRMS-EI found, 683.9434 (M ꢀ H)ꢀ. C15H18N3O20P4 requires
683.9438; purity >98% by HPLC (system A: 13.1 min).
Uridine-50-glucose-10-β,γ-dichloromethylene Tetraphosp-
hate Triethylammonium Salt (26). Compound 10 (3.4 mg,
0.006 mmol) and glucose-1-phosphate disodium salt hydrate (4 mg,
0.036 mmol) were converted to the tributylammonium salts by treatment
with ion-exchange resin [DOWEX 50WX2-200 (H)] and tri-n-butyla-
mine. After the water was removed, the obtained tributylammonium salts
were dried under high vacuum for 1 h. DIC (3.05 μL, 0.036 mmol) was
added to a solution of 10 (0.006 mmol) in DMF (0.2 mL). After the
reaction mixture was stirred at room temperature for 1 h, a solution of
glucose-1-phosphate tributylammonium salt (0.036 mmol) and MgCl2
(3.43 mg, 0.036 mmol) inDMF(0.1 mL) wasadded.Thereactionmixture
was stirred at room temperature overnight. After the solvent was removed,
the MgCl2 was removed by treatment with ion-exchange resin [DOWEX
50WX2-200 (H)] and ammonia bicarbonate or tri-n-butylamine, and the
residue was purified by a semipreparative HPLC to obtain 26 (0.32 mg,
1
7%) as a white solid. H NMR (D2O): δ 8.02 (d, J = 8.1 Hz, 1H),
6.04ꢀ5.97 (m, 2H), 5.71ꢀ5.67 (m, 1H), 4.53ꢀ4.47 (m, 1H), 4.44ꢀ4.38
(m, 1H), 4.37ꢀ4.25 (m, 3H), 4.00ꢀ3.73 (m, 4H), 3.57ꢀ3.41 (m, 2H).
31P NMR (D2O): δ 8.08 (d, J = 19.1 Hz), 1.90ꢀ0.02 (br), ꢀ10.48 (s),
ꢀ10.74 (s). HRMS-EI found, 790.9233 (M ꢀ Hþ)ꢀ. C16H26N2O22P4Cl
requires 790.9227; purity >98% by HPLC (system A: 10.3 min).
General Procedure for the Preparation of Nucleoside
Tetraphosphate Sugars (28ꢀ37). The appropriate sugar 63ꢀ67
was treated with a mixture of sodium acetate (50 mg) in acetic anhydride
(5 mL), and the mixture was stirred for 5 h at 110 °C while following the
course of the reaction by TLC. Excess Ac2O was removed using a rotary
evaporator, and the mixture was stirred with aqueous NaHCO3 for 15
min. The product was extracted into dichloromethane and washed with
brine. The corresponding crude acetyl derivative was purified by column
chromatography. The monophosphate derivatives (68ꢀ72) of these
acetyl sugars were prepared as the lithium salts using the MacDonald
procedure.39 The corresponding sugar monophosphate (0.036 mmol)
and uridine triphosphate trisodium salt (10 mg, 0.018 mmol) were
converted to the tributylammonium salts by treatment with ion-
exchange resin [DOWEX 50WX2-200 (H)] and tri-n-butylamine. After
the water was removed, the obtained tributylammonium salts were dried
under high vacuum for 1 h. DIC (3.05 μL, 0.036 mmol) was added to a
solution of 1 as a tributylammonium salt (0.018 mmol) in DMF
(0.2 mL). After the reaction mixture was stirred at room temperature
for 3 h, a solution of the corresponding sugar monophosphate tributy-
lammonium salt (0.036 mmol) and MgCl2 (3.43 mg, 0.036 mmol) in
DMF (0.1 mL) was added. The reaction mixture was stirred at room
temperature for 2 h to overnight. After the solvent was removed, the
MgCl2 was removed by treatment with ion-exchange resin [DOWEX
50WX2-200 (H)] and ammonia bicarbonate or tri-n-butylamine, and
the residue was purified by a semipreparative HPLC purification using
system C.
Uridine-50-(3-chlorophenyl)-tetraphosphate Triethylam-
monium Salt (21). Compound 21 (1.4 mg, 12%) was obtained as a
white solid using 1 (10 mg, 0.018 mmol). 1H NMR (D2O): δ 7.88 (d, J =
8.1 Hz, 1H), 7.36ꢀ7.12 (m, 4H), 5.91 (d, J = 5.1 Hz, 1H), 5.79 (d, J = 8.1
Hz, 1H), 4.39ꢀ4.28 (m, 2H), 4.23ꢀ4.17 (m, 3H). 31P NMR (D2O): δ
ꢀ11.20 (d, J = 18.3 Hz), ꢀ15.84 (d, J = 17.5 Hz), ꢀ23.14 (m). HRMS-
EI found, 672.9218 (M ꢀ Hþ)ꢀ. C15H18N2O18P4Cl requires 672.9194;
purity >98% by HPLC (system B: 12.2 min).
Uridine-50-(4-methoxyphenyl)-tetraphosphate Triethylam-
monium Salt (22). Compound 22 (2.7 mg, 22%) was obtained as a
white solid using 1 (10 mg, 0.018 mmol). 1H NMR (D2O): δ 7.88 (d, J =
8.4 Hz, 1H), 7.17 (d, J = 9.0 Hz, 2H), 6.92 (d, J = 9.0 Hz, 2H), 5.91 (d, J =
5.1 Hz, 1H), 5.88 (d, J = 8.4 Hz, 1H), 4.37ꢀ4.25 (m, 2H), 4.23ꢀ4.16 (m,
3H), 3.78 (s, 3H). 31P NMR (D2O):δꢀ11.19 (br, 1H), ꢀ15.04 (br, 1H),
ꢀ23.00 (br, 2H). HRMS-EI found, 668.9681 (M
ꢀ
Hþ)ꢀ.
C16H21N2O19P4 requires 668.9689; purity >98% by HPLC (system A:
12.1 min).
Uridine-50-(3-methoxyphenyl)-tetraphosphate Triethy-
lammonium Salt (23). Compound 23 (1.9 mg, 16%) was obtained
as a white solid using 1 (10 mg, 0.018 mmol). 1H NMR (D2O): δ 7.89
(d, J = 8.1 Hz, 1H), 7.26 (t, J = 8.1 Hz, 1H), 6.90ꢀ6.82 (m, 2H), 6.72 (d,
J = 8.1 Hz, 1H), 5.91 (d, J = 6.0 Hz, 1H), 5.88 (d, J = 8.7 Hz, 1H),
4.38ꢀ4.24 (m, 2H), 4.22ꢀ4.16 (m, 3H), 3.80 (s, 3H). 31P NMR (D2O):
δ ꢀ11.20 (d, J = 18.4 Hz, 1H), ꢀ15.69 (d, J = 17.5 Hz, 1H), ꢀ23.14 (m,
2H). HRMS-EI found, 668.9699 (M ꢀ Hþ)ꢀ. C16H21N2O19P4 requires
668.9689; purity >98% by HPLC (system A: 12.1 min).
Uridine-50-allose-10-tetraphosphate Triethylammonium Salt
(28). Compound 28 (1 mg, 8%) was obtained as a white solid using 1 (10
mg, 0.018 mmol) and D-allose-1-phosphate (69, 16.5 mg, 0.036 mmol). 1H
NMR (D2O): δ 7.95 (d, J = 8.1 Hz, 1H), 5.98 (m, 2H), 5.60 (m, 1H), 4.16
(m, 2H), 4.33 (m, 1H), 4.28ꢀ4.23 (m, 2H), 3.90ꢀ3.60 (m, 4H), 3.54 (m,
1H), 3.41 (m, 1H). 31P NMR (D2O): δꢀ11.29 (m), ꢀ12.40 (m), ꢀ22.86
(m). HRMS-EI found, 724.9831 (M ꢀ H)þ. C15H25N2O23P4 requires
724.9804. HPLC (purity 97%, system A: 11.5 min; purity 98%, system B:
0.8 min).
Uridine-50-mannose-10-tetraphosphate Triethylammo-
nium Salt (29). Compound 29 (3.18 mg, 24%) was obtained as a
white solid using 1 (10 mg, 0.018 mmol) and D-mannose-1-phosphate
(68, 16.5 mg, 0.036 mmol). 1H NMR (D2O): δ 7.99 (d, J = 8.4 Hz, 1H),
Uridine-50-(4-nitrophenyl)-tetraphosphate Triethylammo-
nium Salt (24). Compound 24 (2.2 mg, 23%) was obtained as a white
solidusing 1 (12mg, 0.022 mmol). 1H NMR (D2O):δ 8.23(d, J = 8.7Hz,
2H), 7.87 (d, J = 8.4 Hz, 1H), 7.41 (d, J = 8.7 Hz, 2H), 5.91 (d, J = 5.1 Hz,
1H), 5.87 (d, J = 8.4 Hz, 1H), 4.37ꢀ4.28 (m, 2H), 4.22ꢀ4.17 (m, 3H).
31P NMR (D2O): δ ꢀ11.14 (d, J = 17.5 Hz, 1H), ꢀ16.83 (d, J = 16.9 Hz,
1H), ꢀ23.06 (m, 2H). HRMS-EI found, 683.9434 (M ꢀ Hþ)ꢀ.
C15H18N3O20P4 requires 683.9434; purity >98% by HPLC (system A:
11.6 min).
Uridine-50-(3-nitrophenyl)-tetraphosphate Triethylammo-
nium Salt (25). Compound 25 (3.4 mg, 41%) was obtained as a white
solid using 1 (10 mg, 0.018 mmol). 1H NMR (D2O): δ 8.27 (d, J = 12.0
Hz, 2H), 7.92 (d, J = 8.1 Hz, 1H), 7.44 (d, J = 12.0 Hz, 2H), 5.96ꢀ5.88
4029
dx.doi.org/10.1021/jm101591j |J. Med. Chem. 2011, 54, 4018–4033

Journal of Medicinal Chemistry
ARTICLE
6.02 (m, 2H), 5.56 (m, 1H), 4.44 (m, 2H), 4.31ꢀ4.27 (m, 3H), 4.16 (m,
1H), 4.03ꢀ3.87 (m, 3H), 3.79 (m, 1H), 3.67 (m, 1H). 31P NMR (D2O):
δ ꢀ11.17 (d, J = 17.6 Hz), ꢀ12.63 (d, J = 17.5 Hz), ꢀ23.10 (m). HRMS-
EI found, 724.9796 (M ꢀ H)þ. C15H25N2O23P4 requires 724.9799.
HPLC (purity 99%, system A: 10.4 min; purity 99%, system B: 1.1 min).
Uridine-50-xylose-10-tetraphosphate Triethylammonium
Salt (30). Compound 30 (3.18 mg, 43%) was obtained as a white solid
using 1 (6.3 mg, 0.011 mmol) and D-xylose-1-phosphate (14 mg, 0.032
mmol). 1H NMR (D2O): δ 8.00 (d, J = 8.4 Hz, 1H), 6.07ꢀ5.98 (m, 2H),
5.64ꢀ5.58 (m, 1H), 4.50ꢀ4.40 (m, 2H), 4.36ꢀ4.23 (m, 3H),
3.83ꢀ3.76 (m, 1H), 3.69ꢀ3.60 (m, 1H), 3.58ꢀ3.49 (m, 1H),
3.47ꢀ3.37 (m, 1H). 31P NMR (D2O): δ ꢀ11.22 (d, J = 16.8 Hz),
ꢀ12.65 (d, J = 18.2 Hz), ꢀ22.6 to ꢀ23.4 (m). HRMS-EI found,
694.9693 (M ꢀ H)ꢀ. C14H23N2O22P4 requires 694.9656; purity
>98% by HPLC (system B: 12.3 min).
General Procedure for the Preparation of Nucleoside
Tetraphosphate Sugars (31 and 32). The corresponding sodium
salt of sugar monophosphate (0.036 mmol) and the trisodium salt of 1
(10 mg, 0.018 mmol) were converted to the tributylammonium salts by
treatment with ion-exchange resin [DOWEX 50WX2-200 (H)] and tri-
n-butylamine. After the water was removed, the obtained tributylam-
monium salts were dried under high vacuum 1 h. DIC (3.05 μL, 0.036
mmol) was added to a solution of 1 as a tributylammonium salt (0.018
mmol) in DMF (0.2 mL). After the reaction mixture was stirred at room
temperature for 3 h, a solution of the corresponding sugar monopho-
sphate tributylammonium salt (0.036 mmol) and MgCl2 (3.43 mg,
0.036 mmol) in DMF (0.1 mL) was added. The reaction mixture was
stirred at room temperature for 2 h to overnight. After the solvent was
removed, the MgCl2 was removed by treatment with ion-exchange resin
[DOWEX 50WX2-200 (H)] and ammonia bicarbonate or tri-n-butyla-
mine, and the residue was purified by a semipreparative HPLC purifica-
tion using system C.
D-3-deoxy-3-fluoroglucose-1-phosphate (71, 16.57 mg, 0.036 mmol).
1H NMR (D2O): δ 7.92 (d, J = 8.1 Hz, 1H), 5.98 (m, 2H), 5.61 (m, 1H),
4.35 (m, 2H), 4.27ꢀ4.21 (m, 4H), 3.99ꢀ3.72 (m, 3H). 31P NMR
(D2O): δ ꢀ11.08 (m), ꢀ12.87 (m), ꢀ22.38 (m). HRMS-EI found,
726.9763 (M ꢀ H)þ. C15H24FN2O23P4 requires 726.9761. HPLC
(purity 98%, system A: 10.4 min; purity 98%, system B: 1.0 min).
P1-(N4-Benzyloxycytidine-50-)P4-(30-deoxy-30-fluoroglucose-
10-)tetraphosphate Triethylammonium Salt (35). Compound
35 (0.550 mg, 2.6%) was obtained as a white solid using triethylammo-
nium salt of N4-benzyloxycytidine 50-triphosphate (14.45 mg, 0.018
mmol) and D-3-deoxy-3-fluoroglucose-1-phosphate (71, 14.76 mg,
1
0.036 mmol). H NMR (D2O): δ 7.44 (m, 5H), 7.23 (d, J = 8.8 Hz,
1H), 5.96 (d, J = 6.4 Hz, 1H), 5.77 (d, J = 8.4 Hz, 1H), 5.05 (s, 2H),
4.46ꢀ4.42 (m, 2H), 4.38ꢀ4.17 (m, 4H), 3.99ꢀ3.72 (m, 3H). 31P NMR
(D2O): δ ꢀ11.53 (m), ꢀ13.27 (m), ꢀ23.21 (m). HRMS-EI found,
832.0319 (M ꢀ H)þ. C22H31FN3O22P4 requires 726.9761.
P1-(N4-Methoxycytidine-50-)P4-(30-deoxy-30-fluoroglucose-
10-)tetraphosphate Triethylammonium Salt (36). Compound
36 (0.80 mg, 8.8%) was obtained as a white solid using triethylammo-
nium salt of N4-methoxycytidine 50-triphosphate (9 mg, 0.012 mmol)
and D-3-deoxy-3-fluoroglucose-1-phosphate (71, 10 mg, 0.025 mmol).
1H NMR (D2O): δ 7.26 (d, J = 7.6 Hz, 1H), 5.96 (d, J = 6.92 Hz, 1H),
5.83 (d, J = 8.2 Hz, 1H), 5.69 (m, 1H), 4.42ꢀ4.37 (m, 2H), 4.26ꢀ4.15
(m, 4H), 3.99ꢀ3.89 (m, 2H), 3.80 (m, 5H). HRMS-EI found, 756.0002
(M ꢀ H)þ. C16H27FN3O24P4 requires 756.0021; purity 97% by HPLC
(system B: 13.5 min).
P1-(Uridine-50-)P4-(40-deoxy-40-fluoroglucose-10-)tetraphos-
phate Triethylammonium Salt (37). Compound 37 (2.88 mg,
22%) was obtained as a white solid using 1 (10 mg, 0.018 mmol) and
D-4-deoxy-4-fluoroglucose-1-phosphate (72, 16.57 mg, 0.036 mmol).
1H NMR (D2O): δ 7.99 (d, J = 8.1 Hz, 1H), 6.02 (m, 2H), 5.66 (m, 1H),
4.45 (m, 2H), 4.28 (m, 3H), 4.17ꢀ4.06 (m, 2H), 3.99ꢀ3.77 (m, 2H),
3.60 (m, 1H). 31P NMR (D2O): δ ꢀ11.12 (d, J = 17.6 Hz), ꢀ12.70 (d,
J = 16.6 Hz), ꢀ23.17 (m). HRMS-EI found, 726.9753 (M ꢀ H)þ.
C15H24FN2O23P4 requires 726.9761. HPLC (purity 97%, system A:
11.0 min; purity 99%, system B: 1.1 min).
r-D-Glucuronic Acid 1-Phosphate Trisodium Salt (74).
TEMPO (10 mg, 64 μmol) was added to a solution of R-D-glucose-1-
phosphate disodium salt (73, 4.13 g, 13.6 mmol) in H2O (4 mL) at 0 °C.
To this solution, aqueous 1 M NaOH was added until pH 9 was reached,
and a NaOCl solution (1.2 mL, available chlorine: 10ꢀ13%) was added
slowly. The pH was maintained at 9 by adding 1 M NaOH(aq) several
times during the reaction. After 1.5 h, MeOH was added to the reaction
mixture, and the resulting precipitate was collected by filtration to give
74 (415 mg, 90%) as a white solid. 1H NMR (D2O): δ 5.46 (dd, 1H, J =
3.6, 7.5 Hz), 4.15 (d, 1H, J = 10.2 Hz), 3.79 (t, 1H, J = 9.6 Hz), 3.48
(d, 1H, J = 9.6 Hz), 3.47 (t, 1H, J = 9.6 Hz). HRMS-ESꢀ found, 294.9831
(M þ Na ꢀ 2H)ꢀ. C6H9O10PNa requires 294.9837.
P1-(Uridine-50-)P4-(20-deoxy-20-acetamidoglucose-10-)tet-
raphosphate Triethylammonium Salt (31). Compound 31 (5.21
mg, 38%) was obtained as a white solid using 1 (10 mg, 0.018 mmol) and
D-2-deoxy-2-acetamidoglucose-1-phosphate (12.5 mg, 0.036 mmol). 1H
NMR (D2O): δ 7.99 (d, J = 8.1 Hz, 1H), 6.02 (m, 2H), 5.53 (m, 1H),
4.44 (m, 2H), 4.29 (m, 3H), 4.02ꢀ3.80 (m, 4H), 3.53 (m, 1H), 2.11 (s,
3H), 1.93 (s, 1H). 31P NMR (D2O): δ ꢀ11.17 (d, J = 18.3 Hz), ꢀ12.77
(d, J = 19.0 Hz), ꢀ23.10 (m). HRMS-EI found, 766.0037 (M ꢀ H)þ.
C17H28N3O23P4 requires 766.0064. HPLC (purity 99%, system A: 10.4
min; purity 99%, system B: 0.8 min).
P1-(Uridine-50-)P4-(glucuronic acid-10-)tetraphosphate Trie-
thylammonium Salt (32). Compound 32 (2.46 mg, 18%) was
obtained as a white solid using 1 (10 mg, 0.018 mmol) and R-D-glucuronic
acid 1-phosphate (74, 12.2 mg, 0.036 mmol). 1H NMR (D2O): δ 7.95 (d,
J = 7.5 Hz, 1H), 5.99 (m, 2H), 5.65 (m, 1H), 4.39 (m, 2H), 4.27ꢀ4.23 (m,
3H), 4.10 (m, 1H), 3.80 (m, 1H), 3.51 (s, 2H). 31P NMR (D2O): δ
ꢀ11.18 (d, J = 17.6 Hz), ꢀ12.55 (d, J = 16.7 Hz), ꢀ22.71 (m). HRMS-EI
found, 738.9575 (M ꢀ H)þ. C175H23N2O24P4 requires 738.9591. HPLC
(purity 97%, system A: 11.4 min; purity 99%, system B: 1.0 min).
P1-(Uridine-50-)P4-(20-deoxy-20-fluoroglucose-10-)tetraphos-
phate Triethylammonium Salt (33). Compound 33 (3.28 mg,
25%) was obtained as a white solid using 1 (10 mg, 0.018 mmol) and
D-2-deoxy-2-fluoroglucose-1-phosphate (70, 16.57 mg, 0.036 mmol).
1H NMR (D2O): δ 7.99 (d, J = 8.1 Hz, 1H), 6.01 (m, 2H), 5.82 (m, 1H),
4.44 (m, 2H), 4.28 (m, 3H), 4.12ꢀ3.71 (m, 4H), 3.52 (m, 1H). 31P
NMR (D2O): δ ꢀ11.10 (d, J = 16.7 Hz), ꢀ13.10 (d, J = 17.5 Hz),
ꢀ23.07 (m). HRMS-EI found, 726.9753 (M ꢀ H)þ. C15H24FN2O23P4
requires 726.9761. HPLC (purity 99%, system A: 10.3 min; purity 98%,
system B: 1.1 min).
Assay of PLC Activity Stimulated by P2Y2, P2Y4, and P2Y6
Receptors. Stable cell lines expressing the human P2Y2, P2Y4, or P2Y6
receptor in 1321N1 human astrocytoma cells were generated as
described.19 Agonist-induced [3H]inositol phosphate production was
measured in 96-well plates that received 20000 cells/well 2 days prior
to assay. Sixteen hours before the assay, the inositol lipid pool of the cells
was radiolabeled by incubation in 100 μL of serum-free inositol-
free Dulbecco's modified Eagle's medium containing 1.0 μCi of
myo-[3H]inositol. No changes of medium were made subsequent to the
addition of [3H]inositol. On the day of the assay, cells were
challenged with 25 μL of a 5-fold concentrated solution of receptor
agonists in 200 mM HEPES (N-2-hydroxyethylpiperazine-N0-2-
ethanesulfonic acid), pH 7.3, in Hank's balanced salt solution,
containing 50 mM LiCl for 30 min at 37 °C. Incubations were
terminated by aspiration of the drug-containing medium and addi-
tion of 450 μL of ice-cold 50 mM formic acid. [3H]Inositol phosphate
P1-(Uridine-50-)P4-(30-deoxy-30-fluoroglucose-10-)tetraphos-
phate Triethylammonium Salt (34). Compound 34 (3.58 mg,
27%) was obtained as a white solid using 1 (10 mg, 0.018 mmol) and
4030
dx.doi.org/10.1021/jm101591j |J. Med. Chem. 2011, 54, 4018–4033

Journal of Medicinal Chemistry
ARTICLE
accumulation was quantified using scintillation proximity assay
methodology as previously described in detail.19
PolakꢀRibiere conjugate gradient algorithm and were performed with
a threshold on the potential energy gradient of 0.05 kcal/(mol Å). The
conformational searches were based on 100 steps of extended Monte
Carlo multiple minimum (MCMM) sampling protocol. Specifically,
the sampling regarded all torsion angles of residues and ligands as well as
ligand rotation and ligand translation.
The interhelical cavities of the P2Y2 and P2Y4 receptor were analyzed
with the SiteMap program, implemented in the Schrodinger package,
cropping site maps at 2 Å from the nearest site point.44
Data Analysis. Agonist potencies (EC50 values) were determined
from concentrationꢀresponse curves by nonlinear regression analysis
using the GraphPad software package Prism (GraphPad, San Diego,
CA). All experiments examining the activity of newly synthesized
molecules also included full concentration effect curves for the cognate
agonist of the target receptor: 1 for the P2Y2 and P2Y4 receptors and
UDP for the P2Y6 receptor. Each concentration of drug was tested in
triplicate assays, and concentration effect curves for each test drug were
repeated in at least three separate experiments with freshly diluted
molecule. The results are presented as means ( SEMs from multiple
experiments or in the case of concentration effect curves from a single
experiment carried out with triplicate assays that were representative of
results from multiple experiments.
Molecular Modeling. Molecular models of the P2Y2 and P2Y4
receptors were constructed with the molecular operating environment
(MOE)40 on the basis of the crystal structure of the CXCR4 chemokine
receptor.16 An initial sequence alignment was obtained with the
Blosum62 substitution matrix, with penalties for gap insertions and
extensions of 7 and 1, respectively. The proper alignment of the
conserved motifs that characterize each of the TM helices was checked
and, when necessary, manually adjusted. All of the gaps in the alignment
of the TM helices were eliminated; gaps in the loops were consolidated
into a single gap per loop and positioned where insertions or deletions
seemed compatible with the structure of the template. The final
alignment used for the construction of the model is provided in Figure 4.
Ten models were built and scored on the basis of electrostatic solva-
tion energy (GB/VI).41 Intermediate and final models were subjected
to energy minimizations with the Amber99 force field, according to
the protocol implemented in MOE, with the refinement level set to
medium.40
As mentioned, a disulfide bridge connects the N-terminus and the
extracellular end of TM7 in the CXCR4 receptor and, putatively, in the
P2Y receptors, according to mutagenesis data.22ꢀ24 However, as evident
from Figure 4, in the P2Y receptors, this cysteine residue is shifted two
residues downstream with respect to the CXCR4 receptor. Thus, before
the construction of the homology model, the N-terminal region of the
template was modified ad hoc. In particular, Cys28 and Arg30 of CXCR4
were mutated to alanine and cysteine, respectively, and a disulfide bridge
was built between Cys28 and Cys274. The mutated CXCR4 receptor
was then subjected to energy minimization with MOE,37 with the
Amber99 force field, until reaching a cutoff parameter on the potential
energy gradient of 0.05 kcal/(mol Å). During the minimization, the two
cysteines and all of the N-terminal residues were granted flexibility.
After the construction of the homology models, 1 was docked into the
P2Y2 and P2Y4 receptors superimposing our previously published
complexes to the new models. For each receptor, the ligand was then
subjected to two subsequent energy minimizations, the first holding the
receptor frozen, the second one granting flexibility to all residues within
a 5 Å radius from the ligand. The energy minimizations were conducted
in MOE,40 with the MMFFx force field, until reaching a cutoff parameter
on the potential energy gradient of 0.05 kcal/(mol Å). Compounds 11
and 16 were then sketched from 1 within the P2Y2 and P2Y4 receptors,
respectively, and subsequently subjected to the same two rounds of
minimization following the same protocol outlined above for 1. The
obtained P2Y2-11 and P2Y4-16 complexes were subsequently trans-
ferred to Maestro42 and subjected to conformational searches with the
MacroModel engine,43 as implemented in the Schrodinger package.
Flexibility was granted to the ligands and the residues within a 5 Å radius
from them. All of the atoms within an additional shell of a 3 Å radius were
also taken into account as a frozen environment. The calculations
were conducted with the MMFFs force field, using water as implicit
solvent (GB/SA model).44 The minimizations were based on the
Stability of Nucleotide Derivatives. Compound 1 was purchased
from Sigma (St. Louis, MO). Compound 2 was synthesized by the
procedure previously reported.13 HCl buffer (pH 1.5) solution was
prepared as follows. Concentrated HCl (7 mL, 12 N) was added to
2.0 g of sodium chloride and diluted with water to a volume of 1000 mL.
Membranes were prepared as follows. P2Y1 receptor-expressing astro-
cytoma cells were grown to 80% confluence and then harvested. The cells
were homogenized, and the nuclear fraction was removed by centrifuga-
tion at 100g for 5 min. The pellet was resuspended in 50 mM tris-
(hydroxymethyl)aminomethane (Tris) hydrochloride buffer (pH 7.4).
The suspension was homogenized with a polytron homogenizer
(Brinkmann) for 10 s and was ultracentrifuged at 20000g for 20 min at
4 °C. The resultant pellets were resuspended in Tris buffer (pH 7.4), and
the membrane was stored at ꢀ80 °C until the experiments. The protein
concentration of the membrane preparation was measured using the
Bradford assay.45 The membrane preparation was diluted with 4 mL of
Dulbecco's phosphate-buffered saline (pH 7.4) and homogenized before
use for the stability check. The finalprotein concentration was 14.9 μg/mL.
Ten microliters of a 2 mg/mL aqueous solution of each nucleotide
derivative was mixed with either 90 μL of HCl buffer (pH 1.2) or the
membrane suspension and incubated at 37 °C. At regular intervals, a 6 μL
aliquot of the above mixture (either membranes or low pH) was removed.
The injection sample was prediluted with 54 μL of 5 mM TBAP to allow
complete equilibration with the mobile phase to avoid the early HPLC
elution at ∼1 min. Then, 50 μL of the mixture was injected to the HPLC.
The areas under the curve (AUCs) of compounds were measured
using a Hewlett Packard 1100 HPLC equipped with a Zorbax Eclipse
5 mm XDB-C18 analytical column (250 mm ꢁ 4.6 mm; Agilent
Technologies Inc., Palo Alto, CA). Mobile phase: linear gradient solvent
system: 5 mM TBAP-CH3CN from 80:20 to 40:60 in 6.5 min; the flow
rate was 1 mL/min (system C) or system B. Peaks were detected by UV
absorption with a diode array detector at 254, 275, and 280 nm, and
AUCs were calculated based on the peak at 254 nm.
’ ASSOCIATED CONTENT
S
Supporting Information. NMR and mass spectra and
b
HPLC traces for selected derivatives, X-ray structure of
N4-t-BuO-cytosine, and NOESY experiment. This material is
available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*Tel: 301-496-9024. Fax: 301-480-8422. E-mail: kajacobs@helix.
nih.gov.
’ ACKNOWLEDGMENT
X-ray determination was performed by Dr. Yoshiaki Oyama,
Asubio Pharmaceuticals (Osaka, Japan). Mass spectral measure-
ments were performed by Dr. Noel Whittaker (NIDDK). We
thank Dr. Francesca Deflorian (NIDDK) and Dr. Andrei
A. Ivanov (Emory University) for helpful discussions. This
research was supported in part by NIGMS research grant
4031
dx.doi.org/10.1021/jm101591j |J. Med. Chem. 2011, 54, 4018–4033

Journal of Medicinal Chemistry
ARTICLE
GM38213 and by the Intramural Research Program of NIDDK,
National Institutes of Health. H.M. thanks Asubio Pharma Co.,
Ltd. for financial support. S.d.C. thanks Ministerio de Educaciꢀon
y Ciencia (Spain) for financial support.
(11) Kim, H. S.; Ravi, R. G.; Marquez, V. E.; Maddileti, S.;
Wihlborg, A. -K.; Erlinge, D.; Malmsj€o, M.; Boyer, J. L.; Harden,
T. K.; Jacobson, K. A. Methanocarba modification of uracil and adenine
nucleotides: High potency of Northern ring conformation at P2Y1,
P2Y2, P2Y4 and P2Y11, but not P2Y6 receptors. J. Med. Chem. 2002, 45,
208–218.
(12) El-Tayeb, A.; Qi, A.; M€uller, C. E. Synthesis and Structure-
Activity Relationships of Uracil Nucleotide Derivatives and Analogues as
Agonists at Human P2Y2, P2Y4, and P2Y6 Receptors. J. Med. Chem.
2006, 49, 7076–7087.
(13) Ko, H.; Carter, R. L.; Cosyn, L.; Petrelli, R.; de Castro, S.;
Besada, P.; Zhou, Y.; Cappellacci, L.; Franchetti, P.; Grifantini, M.; Van
Calenbergh, S.; Harden, T. K.; Jacobson, K. A. Synthesis and potency of
novel uracil nucleotide analogues as P2Y2 and P2Y6 receptor agonists.
Bioorg. Med. Chem. 2008, 16, 6319–6332.
(14) Sauer, R.; El-Tayeb, A.; Kaulich, M.; M€uller, C. E. Synthesis of
uracil nucleotide analogs with a modified, acyclic ribose moiety as P2Y2
receptor antagonists. Bioorg. Med. Chem. 2009, 17, 5071–5079.
(15) Maruoka, H.; Barrett, M. O.; Ko, H.; Tosh, D. K.; Melman, A.;
Burianek, L. E.; Balasubramanian, R.; Berk, B.; Costanzi, S.; Harden,
T. K.; Jacobson, K. A. Pyrimidine Ribonucleotides with Enhanced
Selectivity as P2Y6 Receptor Agonists: Novel 4-Alkyloxyimino, (S)-
Methanocarba, and 50-Triphosphate γ-Ester Modifications. J. Med.
Chem. 2010, 53, 4488–4501.
(16) Wu, B.; Chien, E. Y.; Mol, C. D.; Fenalti, G.; Liu, W.; Katritch,
V.; Abagyan, R.; Brooun, A.; Wells, P.; Bi, F. C.; Hamel, D. J.; Kuhn, P.;
Handel, T. M.; Cherezov, V.; Stevens, R. C. Structures of the CXCR4
chemokine GPCR with small-molecule and cyclic peptide antagonists.
Science 2010, 330, 1066–1071.
’ ABBREVIATIONS USED
CTP, cytidine 50-triphosphate; DIC, N,N0-diisopropylcarbodii-
mide; DMF, N,N-dimethylformamide;EL, extracellular loop;
GPCR, G protein-coupled receptor; HEPES, N-2-hydroxyethylpi-
perazine-N0-2-ethanesulfonic acid; HPLC, high-performance
liquid chromatography;IL, intracellular loop; MOE, molecular
operating environment; SAR, structureꢀactivity relationship;
TBAP, tetrabutylammonium dihydrogenphosphate; TEAA, trie-
thylammonium acetate; TEMPO, 2,2,6,6-tetramethylpiperidine-
1-oxyl; TM, transmembrane R-helix; TMS, trimethylsilyl;UMP,
uridine 50-monophosphate; UDP, uridine 50-diphosphate; UTP,
uridine 50-triphosphate
’ REFERENCES
(1) Abbracchio, M. P.; Burnstock, G.; Boeynaems, J. M.; Barnard,
E. A.; Boyer, J. L.; Kennedy, C.; Fumagalli, M.; King, B. F.; Gachet, C.;
Jacobson, K. A.; Weisman, G. A. International Union of Pharmacology
LVIII: Update on the P2Y G protein-coupled nucleotide receptors:
From molecular mechanisms and pathophysiology to therapy. Pharma-
col. Rev. 2006, 58, 281–341.
(17) Kogo, S.; Yamada, K.; Iwai, Y.; Osawa, K.; Hayakawa, H.
Process for Producing Di(pyrimidine nucleoside 50-)polyphosphate.
WO2008/012949 A1, 2008.
(2) Brunschweiger, A.; M€uller, C. E. P2 Receptors activated by uracil
nucleotides—An update. Curr. Med. Chem. 2006, 13, 289–312.
(3) Nicholas, R. A.; Lazarowski, E. R.; Watt, W. C.; Li, Q.;
Harden, T. K. Uridine nucleotide selectivity of three phospholipase
C-activating P2 receptors: Identification of a UDP-selective, a UTP-
selective, and an ATP- and UTP-specific receptor. Mol. Pharmacol. 1996,
50, 224–229.
(4) Communi, D.; Motte, S.; Boeynaems, J. M.; Pirotton, S. Phar-
macological characterization of the human P2Y4 receptor. Eur. J.
Pharmacol. 1996, 317, 383–389.
(5) Shaver, S. R.; Rideout, J. L.; Pendergast, W.; Douglass, J. G.;
Brown, E. G.; Boyer, J. L.; Patel, R. I.; Redick, C. C.; Jones, A. C.; Picher,
M.; Yerxa, B. R. Structure activity relationships of dinucleotides: Potent
and selective agonists of P2Y receptors. Purinergic Signalling 2005, 1,
183–191.
(18) Heeres, A.; van Doren, H. A.; Gotlieb, K. F.; Bleeker, I. P.
Synthesis of R- and β- D-glucopyranuronate 1-phosphate and R-D-
glucopyranuronate 1-fluoride: Intermediates in the synthesis of D-glu-
curonic acid from starch. Carbohydr. Res. 1997, 299, 221–227.
(19) Bourdon, D. M.; Wing, M. R.; Edwards, E. B.; Sondek, J.;
Harden, T. K. Quantification of isozyme-specific activation of phospho-
lipase C-β 2 by Rac GTPases and phospholipase C-ε by Rho GTPases in
an intact cell assay system. Methods Enzymol. 2006, 406, 489–499.
(20) Costanzi, S.; Mamedova, L.; Gao, Z.; Jacobson, K. Architecture
of P2Y nucleotide receptors: structural comparison based on sequence
analysis, mutagenesis, and homology modeling. J. Med. Chem. 2004,
47, 5393–5404.
(21) de Castro, S.; Maruoka, H.; Hong, K.; Kilbey, S. M., 2nd;
Costanzi, S.; Hechler, B.; Brown, G. G., Jr; Gachet, C.; Harden, T. K.;
Jacobson, K. A. Functionalized congeners of P2Y1 receptor antagonists:
2-alkynyl (N)-methanocarba 20-deoxyadenosine 30,50-bisphosphate ana-
logues and conjugation to a polyamidoamine (PAMAM) dendrimer
carrier. Bioconjugate Chem. 2010, 21, 1190–205.
(22) Hoffmann, C.; Moro, S.; Nicholas, R. A.; Harden, T. K.;
Jacobson, K. A. The role of amino acids in extracellular loops of the
human P2Y1 receptor in surface expression and activation processes. J.
Biol. Chem. 1999, 274, 14639–14647.
(6) Yerxa, B. R.; Sabater, J. R.; Davis, C. W.; Stutts, M. J.; Lang-Furr,
M.; Picher, M.; Jones, A. C.; Cowlen, M.; Dougherty, R.; Boyer, J.;
Abraham, W. M.; Boucher, R. C. Pharmacology of INS37217
[P(1)-(uridine 50)-P(4)- (20-deoxycytidine 50)tetraphosphate, tetraso-
dium salt], a next-generation P2Y2 receptor agonist for the treatment of
cystic fibrosis. J. Pharmacol. Exp. Ther. 2002, 302, 871–880.
(7) Kellerman, D.; Rossi Mospan, A.; Engels, J.; Schaberg, A.;
Gorden, A.; Smiley, L. Denufosol: A review of studies with inhaled
P2Y2 agonists that led to Phase 3. Pulm. Pharmacol. Ther. 2008, 21,
600–607.
(23) Moro, S.; Hoffmann, C.; Jacobson, K. A. Role of the extra-
cellular loops of G protein-coupled receptors in ligand recognition: a
molecular modeling study of the human P2Y1 receptor. Biochemistry
1999, 38, 3498–3507.
(24) Hillmann, P.; Ko, G. Y.; Spinrath, A.; Raulf, A.; von Kugelgen, I.;
Wolff, S. C.; Nicholas, R. A.; Kostenis, E.; Holtje, H. D.; Muller, C. E.
Key determinants of nucleotide-activated G protein-coupled P2Y2 recep-
tor function revealed by chemical and pharmacological experiments,
mutagenesis and homology modeling. J. Med. Chem. 2009, 52, 2762–
2775.
(8) Jacobson, K.; Costanzi, S.; Ivanov, A.; Tchilibon, S.; Besada, P.;
Gao, Z.; Maddileti, S.; Harden, T. Structure activity and molecular
modeling analyses of ribose- and base-modified uridine 50-triphosphate
analogues at the human P2Y2 and P2Y4 receptors. Biochem. Pharmacol.
2006, 71, 540–549.
(9) Davenport, R. J.; Diaz, P.; Galvin, F. C.; Lloyd, S.; Mack, S. R.;
Owens, R.; Sabin, V.; Wynn, J. Novel nucleotide triphosphates as potent
P2Y2 agonists with enhanced stability over UTP. Bioorg. Med. Chem. Lett.
2007, 17, 558–561.
(10) Ivanov, A.; Ko, H.; Cosyn, L.; Maddileti, S.; Besada, P.; Fricks,
I.; Costanzi, S.; Harden, T.; Calenbergh, S.; Jacobson, K. Molecular
modeling of the human P2Y2 receptor and design of a selective agonist,
20-amino-20-deoxy-2-thiouridine 50-triphosphate. J. Med. Chem. 2007,
50, 1166–1176.
(25) Moro, S.; Guo, D.; Camaioni, E.; Boyer, J. L.; Harden, T. K.;
Jacobson, K. A. Human P2Y1 receptor: molecular modeling and site-
directed mutagenesis as tools to identify agonist and antagonist recogni-
tion sites. J. Med. Chem. 1998, 41, 1456–66.
4032
dx.doi.org/10.1021/jm101591j |J. Med. Chem. 2011, 54, 4018–4033

Journal of Medicinal Chemistry
ARTICLE
(26) Jiang, Q.; Guo, D.; Lee, B. X.; van Rhee, A. M.; Kim, Y. C.;
Nicholas, R. A.; Schachter, J. B.; Harden, T. K.; Jacobson, K. A. A
mutational analysis of residues essential for ligand recognition at the
human P2Y1 receptor. Mol. Pharmacol. 1997, 52, 499–507.
(27) Van Meervelt, L. Structure of 30,50-di-O-acetyl-N4-methoxycy-
tosine. Acta Crystallogr. 1991, C47, 2635–2637.
(28) Ballesteros, J. A.; Weinstein, H. Integrated methods for the
construction of three-dimensional models and computational probing of
structure-function relations in G protein-coupled receptors. In Methods
in Neurosciences; Stuart, C. S., Ed.; Academic Press: San Diego, 1995;
Vol. 25, pp 366ꢀ428.
(29) van Rhee, A. M.; Jacobson, K. A. Molecular architecture of G
protein-coupled receptors. Drug Dev. Res. 1996, 37, 1–38.
(30) Costanzi, S.; Tikhonova, I.; Ohno, M.; Roh, E.; Joshi, B.;
Colson, A.; Houston, D.; Maddileti, S.; Harden, T.; Jacobson, K. P2Y1
antagonists: Combining receptor-based modeling and QSAR for a
quantitative prediction of the biological activity based on consensus
scoring. J. Med. Chem. 2007, 50, 3229–3241.
(31) Yitzhaki, S.; Shneyvais, V.; Jacobson, K. A.; Shainberg, A.
Involvement of uracil nucleotides in protection of cardiomyocytes from
hypoxic stress. Biochem. Pharmacol. 2005, 69, 1215–1223.
(32) Wildman, S. S.; Unwin, R. J.; King, B. F. Extended pharmaco-
logical profiles of rat P2Y2 and rat P2Y4 receptors and their sensitivity to
extracellular Hþ and Zn2þ ions. Br. J. Pharmacol. 2003, 140, 1177–1186.
(33) Robaye, B.; Ghanem, E.; Wilkin, F.; Fokan, D.; Van Driessche,
W.; Schurmans, S.; Boeynaems, J. M.; Beauwens, R. Loss of nucleotide
regulation of epithelial chloride transport in the jejunum of P2Y4-null
mice. Mol. Pharmacol. 2003, 63, 777–783.
(34) Ghanem, E.; Robaye, B.; Leal, T.; Leipziger, J.; Van Driessche,
W.; Beauwens, R.; Boeynaems, J. M. The role of epithelial P2Y2 and
P2Y4 receptors in the regulation of intestinal chloride secretion. Br. J.
Pharmacol. 2005, 146, 364–369.
(35) Matos, J. E.; Sorensen, M. V.; Geyti, C. S.; Robaye, B.;
Boeynaems, J. M.; Leipziger, J. Distal colonic Naþ absorption inhibited
by luminal P2Y2 receptors. Pflugers Arch. 2007, 454, 977–987.
(36) Cosentino, S.; Barcella, S.; Lo, H.; Burbiel, J. C.; M€uller, C. E.;
Tremoli, E.; Banfi, C.; Abbracchio, M. P. Role of P2 receptors in a
cardiomyocyte ischemic hypoxia model. Purines2010, Tarragona, Spain.
May 30ꢀJune 2, 2010. Abstract P10ꢀ11.
(37) Horckmans, M.; Lantz, N.; Dol-Gleizes, F.; Savi, P.; Gachet, C.;
Boeynaems, J. M.; Robaye, B.; Communi, D. Involvement of the P2Y4
receptor in inflammation and angiogenesis. Purines 2008 Meeting June
29ꢀJuly 2, 2008, Copenhagen, Denmark. Purinergic Signalling 2008, 4
(Suppl. 1), 118.
(38) Vanderstocken, G.; Bondue, B.; Horckmans, M.; Di Pietranto-
nio, L.; Robaye, B.; Boeynaems, J. M.; Communi, D. P2Y2 receptor
regulates VCAM-1 membrane and soluble forms and eosinophil accu-
mulation during lung inflammation. J. Immunol. 2010, 185, 3702–3707.
(39) MacDonald, D. L. [11] Chemical synthesis of aldose 1-phos-
phates. Methods Enzymol. 1966, 8, 121–125.
(40) The Molecular Operating Environment (MOE); Chemical Com-
puting Group, Inc.: Montreal, Canada, 2009; Vol. 10; www.ccg.com.
(41) Labute, P. The generalized Born/volume integral implicit
solvent model: Estimation of the free energy of hydration using London
dispersion instead of atomic surface area. J. Comput. Chem. 2008, 29,
1693–1698.
(42) Maestro, 9.1; Schrodinger, LLC: New York, NY; www.
schrodinger.com.
(43) MacroModel, 9.8; Schrodinger, LLC: New York, NY; www.
schrodinger.com.
(44) SiteMap, 2.4; Schr€odinger, LLC: New York, NY; www.
schrodinger.com.
(45) Bradford, M. M. A rapid and sensitive method for the quantita-
tion of microgram quantities of protein utilizing the principle of protein
dye binding. Anal. Biochem. 1976, 76, 248–254.
4033
dx.doi.org/10.1021/jm101591j |J. Med. Chem. 2011, 54, 4018–4033