Evidence for the existence of a specific gprotein-coupled receptor activated by guanosine

Article abstract of DOI:10.1002/cmdc.201100100

Guanosine, released extracellularly from neurons and glial cells, plays important roles in the central nervous system, including neuroprotection. The innovative DELFIA Eu-GTP binding assay was optimized for characterization of the putative guanosine receptor binding site at rat brain membranes by using a series of novel and known guanosine derivatives. These nucleosides were prepared by modifying the purine and sugar moieties of guanosine at the 6- and 5'-positions, respectively. Results of these experiments prove that guanosine, 6-thioguanosine, and their derivatives activate a Gprotein-coupled receptor that is different from the well-characterized adenosine receptors. Catching the elusive guanosine receptor: The innovative DELFIA Eu-GTP binding assay was applied to characterize the guanosine binding site by using novel and known guanosine derivatives. Some of the tested compounds, which proved to be full agonists with EC₅₀ values in the low nanomolar range, could be useful tools for further characterization of the putative guanosine receptor.

Full text of DOI:10.1002/cmdc.201100100

MED
DOI: 10.1002/cmdc.201100100
Evidence for the Existence of a Specific G Protein-Coupled
Receptor Activated by Guanosine
Rosaria Volpini,[a] Gabriella Marucci,[a] Michela Buccioni,[a] Diego Dal Ben,[a]
Catia Lambertucci,[a] Carmen Lammi,[a] Ram C. Mishra,[b] Ajiroghene Thomas,[a] and
Gloria Cristalli*[a]
Guanosine, released extracellularly from neurons and glial cells,
plays important roles in the central nervous system, including
neuroprotection. The innovative DELFIA Eu-GTP binding assay
was optimized for characterization of the putative guanosine
receptor binding site at rat brain membranes by using a series
of novel and known guanosine derivatives. These nucleosides
were prepared by modifying the purine and sugar moieties of
guanosine at the 6- and 5’-positions, respectively. Results of
these experiments prove that guanosine, 6-thioguanosine, and
their derivatives activate a G protein-coupled receptor that is
different from the well-characterized adenosine receptors.
Introduction
Guanosine (1, Figure 1) is a natural-
ly occurring nucleoside which is
also part of many structural and
regulatory components in cells.
However, it was recognized that
neurons and glial cells, like astro-
cytes and oligodendrocytes, are
able to release guanosine, which
plays an important role in the cen-
placing the binding of [3H]guanosine (at 50 nm) from rat brain
membranes: guanosine = 6-thioguanosine > 8-bromoguano-
sine > inosine > 7-methylguanosine = 3’-deoxyguanosine >
2’-deoxyguanosine = guanine = 6-thioguanine @ N2-methyl-
guanosine. Hence, guanosine itself and 6-thioguanosine were
equally effective in displacing [3H]guanosine in these binding
experiments (Ki =31 and 32 nm, respectively).
Limited modifications at other positions of guanosine result-
ed in compounds with lower affinity. Furthermore, the pres-
ence of the ribofuranosyl sugar moiety of guanosine or 6-thio-
guanosine (2, Figure 1) appears relevant but not essential; in
fact, the corresponding bases (guanine and 6-thioguanine), al-
though less active than the above-mentioned nucleosides (Ki =
219 and 251 mm, respectively), are still able to displace
[3H]guanosine from rat brain membranes. These binding stud-
ies confirmed that the [3H]guanosine-labeled sites are distinct
from the well-characterized ATP and adenosine binding sites.
In fact, neither adenine nucleotides, nor adenosine, nor selec-
tive adenosine receptor ligands decreased the binding of
[3H]guanosine.[8] However, after publication of these studies by
Traversa et al., there were no other reports related to the char-
acterization of guanosine binding sites, despite the authors’
suggestion that these sites, found in membranes from rat
tral nervous system. In fact, there
is increasing evidence that guano-
sine is able to mediate trophic ef-
fects[1] by inducing the synthesis
and release of trophic factors such
Figure 1. Structures of guano-
sine (1) and 6-thioguanosine
(2).
as nerve growth factor (NGF), fibroblast growth factor (FGF),
and adenosine.[2] Furthermore, guanosine shows neuroprotec-
tive effects, inducing glutamate uptake after ischemic
damage.[3,4] Under basal conditions, the guanosine concentra-
tion is ~0.47 mm, but during specific conditions such as hypo-
xia and hypoglycemia, its concentration is greatly increased.[5]
Some of guanosine’s effects appear to be exerted after its
uptake into cells. Many of these effects, however, particularly
trophic effects including promotion, division, and growth of as-
trocytes and other neuron-like cells, are still present after the
use of specific nucleoside uptake inhibitors such as nitroben-
zylthioguanosine, nitrobenzylthioinosine, and propentophyl-
line. This suggests the presence of a specific membrane bind-
ing site for guanosine.[6]
[a] Prof. R. Volpini,+ Prof. G. Marucci,+ Dr. M. Buccioni, Dr. D. Dal Ben,
Dr. C. Lambertucci, Dr. C. Lammi, A. Thomas, Prof. G. Cristalli
School of Pharmacy, Medicinal Chemistry Unit
University of Camerino
Some years ago, a binding site for guanosine was detected
on membrane preparations from rat brain.[7,8] Traversa and co-
workers carried out the initial tests on a series of compounds
belonging to the group of guanine- and adenine-based pu-
rines by using a radioligand receptor binding assay. Structure–
activity relationships drawn from their studies showed the fol-
lowing rank order of potency for the tested substances in dis-
via S. Agostino 1, 62032 Camerino (MC) (Italy)
Fax: (+39)0737402295
E-mail: gloria.cristalli@unicam.it
[b] Dr. R. C. Mishra
Biology Department, Georgia State University
100 Piedmont Avenue SE, Atlanta, GA 30303 (USA)
[+] These authors contributed equally to this work.
1074
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1074 – 1080

A Guanosine-Activated GPCR
brain or primary astrocytes in culture, could belong to the
class of G protein-coupled receptors (GPCRs).[8]
Recently, a novel time-resolved fluorometric (TRF) method,
which exploits the unique fluorescence properties of lantha-
nide chelates and provides a powerful alternative to radioiso-
tope assays, was set up by our research group. It was validated
by testing the functional activity of ligands at adenosine A3 re-
ceptors that had already been studied in adenylyl cyclase ex-
periments.[9] This new assay allows the examination of GPCR
activation by monitoring the binding of a non-hydrolysable eu-
ropium-labeled GTP (Eu-GTP).[10] Hence, we decided to investi-
gate the putative guanosine binding site at rat brain mem-
branes with this innovative and non-radioactive assay in order
to clarify whether the putative guanosine receptor is a GPCR
that can be activated by guanosine as well as new guanosine
derivatives, synthesized on the basis of results obtained with
the above-mentioned binding study.
For this purpose, a series of guanine-based nucleosides, sub-
stituted at the 6-position, were prepared on the basis that gua-
nosine and 6-thioguanosine showed the same binding capaci-
ty. Hence, methyl, cyclopentyl, cyclohexyl, and phenyl groups
were introduced at the 6-position, linked to the oxygen or
sulfur atom (Figure 2). The synthesis of isosteres bearing a ni-
Scheme 1. Reagents and conditions: a) Ac2O, DMAP, Et3N, MeCN, RT, 30 min;
b) POCl3, N,N-DEA, TEA, MeCN, 1008C, 5 min; c) RXH, base, MeCN or DMF,
RT–1208C, 30 min–48 h; d) NH3/MeOH, RT, 16 h; e) RNH2, EtOH/H2O 1:1,
408C–reflux, 2–4 h.
Reaction of 4 with the suitable alcohol, previously stirred
with sodium hydroxide, gave the desired deprotected O-substi-
tuted compounds 5–7. Compound 5 was already described by
Maruyama et al.[12] 6-Thioguanosine (2) and 8-bromoguanosine
(18) were prepared by following reported procedures.[13,14] For
the synthesis of thio derivatives 8–11, compound 4 was react-
ed with the appropriate thiols in the presence of potassium
carbonate, and after evaporation, the crude mixture was treat-
ed with methanolic ammonia to give the desired 6-thio deriva-
tives. Synthesis of compound 8 was already described, but our
procedure afforded an improved yield.[15] Access to compound
11 was reported with a different procedure, starting from a nu-
cleoside in which the 2-amino group had been previously pro-
tected.[16] Reaction of compound 4 with the suitable amine af-
forded the desired nucleosides 12–14 (Scheme 1). These deriv-
atives had already been described, but were obtained with dif-
ferent procedures.[17–19] Reaction of compound 4 with hydroxyl-
amine and hydrazine to obtain derivatives 16 and 17 gave
mixtures that were difficult to purify. Hence, an alternative syn-
thetic route was set up by deprotecting 4 with methanolic am-
monia at room temperature to give compound 15,[12] the cou-
pling of which with hydroxylamine and hydrazine afforded the
desired nucleosides 16 and 17.[20]
Figure 2. General structures of the synthesized compounds.
trogen atom linked to the 6-position of the purine core were
carried out, and small hydrophilic groups such as hydroxy and
amino were also linked to the 6-nitrogen atom. The known 8-
bromoguanosine was re-synthesized and tested. Furthermore,
as in the case of adenosine receptors modification of the
ribose moiety improved ligand potency, 5’-carboxamido-, 5’-N-
ethylcarboxamido-, and 5’-N-cyclopentylcarboxamidoguano-
sine derivatives were prepared and tested.
Guanosine (1) was also used as starting material for the syn-
thesis of 5’-modified compounds 24–26 (Scheme 2). Com-
pound 1 was protected with acetone to give derivative 19,[21]
which was then oxidized by bis(acetoxy)iodobenzene (BAIB)
using 2,2,6,6-tetramethylpiperidin-1-yloxyl (TEMPO) as catalyst
to obtain carboxylate derivative 20.[22] The carboxylate moiety
of 20 was first converted into the acyl chloride by reaction
with thionyl chloride in the presence of a catalytic amount of
N,N-dimethylformamide; after removal of excess reagent, treat-
ment of the crude mixture with the suitable amines furnished
the corresponding amide derivatives 21–23. Deprotection of
the latter compounds under acidic conditions provided the de-
Results and Discussion
Chemistry
The 6-substituted-2-aminopurine riboside derivatives 5–14, 16,
17 were synthesized by starting from commercially available
guanosine (1) (Scheme 1). Compound 1 was protected by ace-
tylation with acetic anhydride using triethylamine and N,N-di-
methylaminopyridine as catalyst to give 3, which was reacted
with phosphorus oxychloride leading to the 6-chloro derivative
4.[11]
ChemMedChem 2011, 6, 1074 – 1080
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1075

G. Cristalli et al.
MED
Guanosine stability was assessed in the stimulation buffer
and the biological matrix by using classical HPLC experiments.
The results showed that no guanosine degradation was detect-
able until 30 min incubation time, confirming that this nucleo-
side is stable in our biological system and under the assay con-
ditions.
Biological evaluation
All the synthesized compounds were tested by using the opti-
mized DELFIA Eu-GTP assay, and in order to evaluate the spe-
cificity of the guanosine response, known adenosine receptor
agonists such as N6-cyclopentyladenosine (CPA, for the A1 sub-
type), 2-(aminoethylphenyl-4-propanoic acid)-5’-N-ethylcarbox-
amidoadenosine (CGS21680, for A2A), 5’-N-ethylcarboxamidoa-
denosine (NECA, for A2B), and 2-chloro-N6-(3-iodobenzyl)-5’-N-
methylcarboxamidoadenosine (Cl-IBMECA, for A3) were used as
control compounds (Table 1).[25] The results obtained showed
that the adenosine derivatives are unable to activate the re-
Scheme 2. Reagents and conditions: a) PTSA, DMP, (CH3)2CO, RT, 3 h; b) BAIB,
TEMPO, NaHCO3, MeCN/H2O 1:1, RT, 3 h; c) 1. SOCl2, DMF, 508C, 2 h, 2. RNH2,
À208C–RT, 90 min–16 h; d) HCl 2n, MeCN/H2O, 608C, 2 h.
sired derivatives 24–26. Synthesis of protected nucleoside 22
and compound 25 were already reported with a different pro-
cedure, which involved a greater number of steps.[23,24]
Table 1. Biological activity and response of the studied compounds at rat
brain membranes expressing the putative guanosine receptor using the
Eu-GTP binding assay.
Eu-GTP binding assay optimization
The knowledge acquired in using the innovative DELFIA Eu-
GTP binding assay at the adenosine A3 receptor[9] was exploit-
ed for its optimization at rat brain membrane preparations in
order to investigate the nature of the guanosine binding site
using known and novel ligands. This new assay allows the ex-
amination of GPCR activation by monitoring the binding of a
non-hydrolysable europium-labeled GTP: Eu-GTP. Upon ago-
nist–GPCR binding, Eu-GTP binds to the Ga subunit. After incu-
bation, the unbound Eu-GTP is washed away, and the signal
from GPCR-bound Eu-GTP is measured with a TRF reader. Spe-
cifically, Eu-GTP-labeled species fluoresce at l 615 nm after ex-
citation at l 340 nm. This fluorescence signal can be used to
construct specific dose–response curves of each agonist
tested, as an increase in agonist concentration produces an in-
crease in the fluorescence signal. For these assays, the re-
sponse elicited by guanosine was set at 100% of effect.
The first crucial point investigated in the optimization pro-
cess was determination of the suitable buffer for the specific
stimulation of the guanosine target at rat brain membranes. In
particular, using guanosine as reference compound, results
showed that at 1 mm MgCl2, 10 mm GDP, and 20 mm NaCl in
50 mm HEPES, the putative guanosine receptor stimulation at
rat brain membrane was achieved with high fluorescence
signal detected relative to the basal samples.
Compd
X
R
EC50 [nm][a]
Response [%][b]
1, Guanosine
5
6
7
2, 6-Thioguanosine
8
9
10
53
61
15
22
26
26
12
14
50
630
660
199
354
301
398
316
25
16
97
ND
ND
ND
ND
100
75
100
100
88
83
100
100
85
85
80
87
88
89
85
100
80
80
90
0
O
O
O
CH3
cC5H9
C6H5
S
S
S
S
NH
NH
NH
CH3
cC5H9
cC6H11
C6H5
H
CH3
cC5H9
11
12, 2-Aminoadenosine
13
14
15
16
17
18, 8-Bromoguanosine
24
25
26
CPA[c]
CGS21680[c]
NECA[c]
Cl-IBMECA[c]
NH
NH
OH
NH2
H
C2H5
cC5H9
0
0
0
After the optimization of the stimulation buffer, various gua-
nosine incubation times were investigated in order to obtain
the best response in fluorescence terms of guanosine itself.
The best incubation time was 5 min, and so the other com-
pounds were incubated at the same time and under the same
conditions.
[a] Concentration of agonist required to produce 50% of the maximum
effect; each concentration was tested five times in triplicate, and values
are given as the mean ÆSE. [b] Response is expressed as a percentage of
the maximal relative fluorescence and is referenced against guanosine,
set at 100%. [c] Compound tested at 100 mm; ND: not determined.
1076
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1074 – 1080

A Guanosine-Activated GPCR
ceptor at rat brain membranes, suggesting that the fluores-
cence response obtained after guanosine incubation is not
due to unspecific activation of adenosine receptors. Moreover,
these results confirm the undoubted putative guanosine re-
ceptor activation because the adenosine derivatives are unable
to stimulate this binding site. Moreover, both hypoxanthine
and guanine are unable to activate the receptor (data not
shown). These results underscore that the ribose moiety is im-
portant for receptor activation because in the absence of the
sugar, any fluorescence signal variation relative to the basal
level was detected.
spect to the adenosine receptor, all these nucleosides showed
partial agonist behavior, with responses ranging from 80 to
90%. Concentration–response curves of selected representa-
tive compounds, in comparison with guanosine (1) and 6-thio-
guanosine (2), are shown in Figure 3.
Guanosine and 6-thioguanosine (2) gave responses (relative
fluorescence, RF) of 100 and 88%, with EC50 values of 50 and
26 nm, respectively. Hence, the isosteric substitution of oxygen
with a sulfur atom at the 6-position led to a compound that is
slightly more active than guanosine, but unable to give the full
response.
On these bases, a series of nucleosides substituted at the 6-
position with methyl, cyclopentyl, cyclohexyl, or phenyl groups
linked by an oxygen or sulfur atom were prepared and tested
in the optimized DELFIA Eu-GTP assay (Table 1). Results of
these experiments show that an O- or S-linked methyl group
(compounds 5 and 8, respectively) maintains the activity of the
corresponding parent compounds 1 and 2, while lacking the
full agonism with respect to compound 1 (compound 5:
EC50 =61 nm, RF=75%; compound 8: EC50 =26 nm, RF=83%).
On the other hand, introduction of a cyclopentyl or a phenyl
group afforded nucleosides that are slightly more potent than
guanosine and 6-thioguanosine while generally maintaining
the full agonism (compounds 6: EC50 =15 nm, RF=100%, 7:
EC50 =22 nm, RF=100%, 9: EC50 =12 nm, RF=100%, and 11:
EC50 =50 nm, RF=85%). Because the most potent compound
in the series proved to be the 6-thiocyclopentyl derivative 9,
the larger cyclohexyl homologue 10 was prepared and tested
as well. Results confirm that the presence of a cycloalkyl group
linked to a sulfur atom is favorable for activity and potency at
the putative guanosine receptor (10: EC50 =14 nm, RF=100%).
The synthesis of isosteres bearing a nitrogen atom linked to
the 6-position of the purine core led to compounds 12–14.
Small hydrophilic groups such as hydroxy and amino were also
linked to the 6-nitrogen atom yielding compounds 16 and 17.
Results of the DELFIA Eu-GTP assay show that all these modifi-
cations led to a decrease in response (80–89%) and EC50
values ranging from 199 nm (cyclopentyl derivative 14) to
660 nm (methyl derivative 13).
The known 8-bromoguanosine (18) was re-synthesized and
tested as well, proving to be a full agonist but with less poten-
cy than the parent guanosine (18: EC50 =316 nm, RF=100%).
Furthermore, because modification of the ribose group im-
proved ligand potency at some adenosine receptors, 5’-carbox-
amido-, 5’-N-ethylcarboxamido-, and 5’-N-cyclopentylcarboxa-
midoguanosine derivates 24–26 were prepared and tested.
The results of the Eu-GTP assay show that this sugar modifica-
tion is well tolerated in terms of activity. In particular, the 5’-N-
ethylcarboxamido derivative 25, with an EC50 value of 16 nm, is
one of the most potent compounds in this series. However, in
contrast to the corresponding adenosine derivatives with re-
Figure 3. Concentration–response curves in Eu-GTP functional assays at rat
brain membranes expressing the putative guanosine receptor of selected
representative compounds. Each point represents the mean of five experi-
ments performed in triplicate with a maximum SEM lower than Æ10
(RF=relative fluorescence).
Conclusions
A series of novel and known guanosine derivatives were pre-
pared and tested at rat brain membrane preparations by using
the innovative DELFIA Eu-GTP binding assay. The results prove
that guanosine and 6-thioguanosine and their derivatives acti-
vate a putative GPCR that is different from the well-character-
ized purinergic adenosine receptors. In fact, known adenosine
receptor agonists are unable to stimulate this binding site.
Among the guanosine derivatives tested, modifications at the
6- or 5’-positions may increase activity, although certain of
these variations are detrimental for efficacy. Notably, the novel
cycloalkyl guanosine and 6-thioguanosine derivatives 6, 9, and
10 are endowed with EC50 values in the low nanomolar range
and are able to induce a response at 100%. Hence, these
could be useful tools for further characterization of the puta-
tive guanosine receptor.
Experimental Section
Chemistry
Melting points were determined with a Bꢁchi apparatus and are
1
uncorrected. H NMR spectra were obtained with a Varian Mercury
400 MHz spectrometer; d values are in ppm, J values are in Hz. All
ChemMedChem 2011, 6, 1074 – 1080
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1077

G. Cristalli et al.
MED
exchangeable protons were confirmed by the addition of D2O.
Mass spectra were recorded on an HP 1100-MSD series instrument.
All measurements were performed with electrospray ionization
(ESI-MS) on a single quadrupole analyzer. Thin-layer chromatogra-
phy (TLC) was carried out on pre-coated TLC plates with silica gel
60 F254 (Fluka). For column chromatography, silica gel 60 (Merck)
and silica RP-18 (Aldrich) were used. Elemental analyses were de-
termined on a Fisons Instruments Model EA 1108 CHNS-O analyzer
and are within 0.4% of theoretical values. Purity of the compounds
is ꢀ98% according to elemental analysis data.
reacted with cyclopentanethiol. Flash chromatography eluting with
CHCl3/MeOH (98:2) afforded 9 as a light-yellow solid. Yield 81%;
1
mp: 838C (dec); H NMR (400 MHz, [D6]DMSO): d=1.58 (m, 4H, H-
cyclopentyl), 1.70 (m, 2H, H-cyclopentyl), 2.20 (m, 2H, H-cyclopen-
tyl), 3.52 (m, 1H, HCH-5’), 3.62 (m, 1H, HCH-5’), 3.87 (m, 1H, H-4’),
4.08 (m, 1H, H-3’), 4.24 (m, 1H, SCH), 4.45 (m, 1H, H-2’), 5.06 (t, J=
5.6 Hz, 1H, OH), 5.15 (d, J=4.8 Hz, 1H, OH), 5.42 (d, J=6.0 Hz, 1H,
OH), 5.76 (d, J=5.6 Hz, 1H, H-1’), 6.48 (brs, 2H, NH2), 8.15 ppm (s,
1H, H-8); 13C NMR (75 MHz, [D6]DMSO): d=24.9, 33.7, 41.2, 61.7,
70.7, 73.9, 85.6, 86.8, 122.3, 139.1, 151.3, 159.8, 160.9 ppm; ESI-MS:
positive mode m/z: 368.0 ([M+H]+), 390.0 ([M+Na]+), 757.0
([2M+Na]+), negative mode m/z: 401.5 ([M+Cl]À), 732.5
([2MÀH]À); Anal (C15H21N5O4S) C, H, N.
General procedure for the synthesis of 6-alkoxy derivatives 6
and 7. Compound 4 (300 mg, 0.7 mmol) was added to a suspen-
sion of the appropriate alcohol (7.0 mmol), and the solution was
stirred with NaOH (280 mg, 7.0 mmol) for 1 h at RT; the reaction
mixture was left under stirring for 48 h at RT. After removal of the
volatiles, the residue was purified as described for each com-
pound.
(2R,3R,4S,5R)-2-[2-Amino-6-(cyclohexylthio)-9H-purin-9-yl]-5-(hy-
droxymethyl)tetrahydrofuran-3,4-diol (10). Compound 4 was re-
acted with cyclohexanethiol. Flash chromatography eluting with
CHCl3/MeOH (97:3) afforded 10 as a white solid. Yield 91%; mp:
1
738C (dec); H NMR (400 MHz, [D6]DMSO): d=1.23 (m, 2H, H-cyclo-
(2R,3R,4S,5R)-2-[2-Amino-6-(cyclopentyloxy)-9H-purin-9-yl]-5-(hy-
droxymethyl)tetrahydrofuran-3,4-diol (6). Compound 4 was re-
acted with cyclopentyl alcohol. Preparative TLC eluting with CHCl3/
MeOH (97:3) afforded 6 as a white solid. Yield 18%; mp: 998C
hexyl), 1.44 (m, 3H, H-cyclohexyl), 1.58 (m, 1H, H-cyclohexyl), 1.71
(m, 2H, H-cyclohexyl), 2.00 (m, 2H, H-cyclohexyl), 3.41 (m, 1H,
HCH-5’), 3.60 (m, 1H, HCH-5’), 3.87 (m, 1H, H-4’), 4.09 (m, 2H, H-3’
and H-cyclohexyl), 4.44 (m, 1H, H-2’), 5.05 (m, 1H, OH), 5.13 (d, J=
4.0 Hz, 1H, OH), 5.41 (d, J=6.0 Hz, 1H, OH), 5.76 (d, J=5.2 Hz, 1H,
H-1’), 6.47 (brs, 2H, NH2), 8.15 ppm (s, 1H, H-8); 13C NMR (75 MHz,
[D6]DMSO): d=25.2, 25.5, 32.9, 33.0, 61.3, 70.4, 73.5, 85.3, 86.4,
120.4, 138.7, 151.0, 159.5, 159.7 ppm; Anal (C16H23N5O4S) C, H, N.
1
(dec); H NMR (400 MHz, [D6]DMSO): d=1.58 (m, 2H, H-cyclopen-
tyl), 1.73 (m, 4H, H-cyclopentyl), 1.96 (m, 2H, H-cyclopentyl), 3.50
(m, 1H, HCH-5’), 3.59 (m, 1H, HCH-5’), 3.86 (m, 1H, H-4’), 4.07 (m,
1H, H-3’), 4.44 (m, 1H, H-2’), 5.11 (m, 2H, 2ꢂOH), 5.39 (d, J=
6.0 Hz, 1H, OH), 5.57 (m, 1H, CH-cyclopentyl), 5.75 (d, J=6.0 Hz,
1H, H-1’), 6.36 (brs, 2H, NH2), 8.06 ppm (s, 1H, H-8); 13C NMR
(75 MHz, [D6]DMSO): d=23.4, 33.4, 61.3, 70.3, 73.7, 79.2, 85.2, 86.3,
116.5, 135.5, 151.3, 154.7, 156.6 ppm; Anal (C15H21N5O5) C, H, N.
(2R,3R,4S,5R)-2-[2-Amino-6-(phenylthio)-9H-purin-9-yl]-5-(hy-
droxymethyl)tetrahydrofuran-3,4-diol (11). Compound 4 was re-
acted with thiophenol. Flash chromatography eluting with CHCl3/
MeOH (98:2) afforded 11 as a white solid.[16] Yield 48%; mp: 1028C
(dec); Anal (C16H17N5O4S) C, H, N.
(2R,3R,4S,5R)-2-(2-Amino-6-phenoxy-9H-purin-9-yl)-5-(hydroxy-
methyl)tetrahydrofuran-3,4-diol (7). Compound 4 was reacted
with phenol. Flash chromatography eluting with CHCl3/MeOH
(98:2) afforded 7 as a white solid. Yield 13%; mp: 115–1178C;
1H NMR (400 MHz, [D6]DMSO): d=3.53 (m, 1H, HCH-5’), 3.62 (m,
1H, HCH-5’), 3.89 (m, 1H, H-4’), 4.10 (m, 1H, H-3’), 4.47 (m, 1H, H-
2’), 5.07 (m, 1H, OH), 5.18 (m, 1H, OH), 5.46 (m, 1H, OH), 5.80 (d,
J=6.0 Hz, 1H, H-1’), 6.44 (brs, 2H, NH2), 7.23 (m, 3H, H-Ph), 7.42
(m, 2H, H-Ph), 8.21 ppm (s, 1H, H-8); 13C NMR (75 MHz, [D6]DMSO):
d=61.7, 70.8, 73.9, 85.7, 86.9, 114.5, 122.1, 125.5, 130.0, 139.3,
152.9, 155.6, 160.1 ppm; Anal (C16H17N5O5) C, H, N.
(2R,3R,4S,5R)-2-(2,6-Diamino-9H-purin-9-yl)-5-(hydroxymethyl)te-
trahydrofuran-3,4-diol (12). NH3 (4 mL) was condensed into a
steel bomb at À808C, and compound 4 (200 mg, 0.47 mmol) was
added. The mixture was left at 708C for 16 h, then evaporated to
dryness and purified by flash chromatography eluting with CHCl3/
MeOH (90:10) to afford, after crystallization from MeOH, 12 as a
white solid.[17] Yield 19%; mp: 217–2198C; Anal (C10H14N6O4) C, H,
N.
(2R,3R,4S,5R)-2-[2-Amino-6-(methylamino)-9H-purin-9-yl]-5-(hy-
droxymethyl)tetrahydrofuran-3,4-diol (13). NH2CH3 (2 mL) was
poured into a cooled flask containing 4 (200 mg, 0.47 mmol). The
flask was sealed and let to stir at RT for 16 h. The mixture was
dried under vacuum and purified by flash chromatography eluting
with CHCl3/MeOH (95:5) to afford 13 as a white solid.[18] Yield 94%;
mp: 125–1278C; Anal (C11H16N6O4) C, H, N.
(2R,3R,4S,5R)-2-[2-Amino-6-(methylthio)-9H-purin-9-yl]-5-(hy-
droxymethyl)tetrahydrofuran-3,4-diol (8). Sodium methanethio-
late (198 mg, 2.82 mmol) was added to a solution of 4 (200 mg,
0.47 mmol) in dry DMF (7 mL), and the mixture was stirred at RT
for 1 h. MeOH (7 mL) was then added, and the mixture was stirred
for another 2 h. After removing the volatiles, the residue was sepa-
rated by flash column chromatography, eluting with CHCl3/MeOH
(98:2) to give 8 as a white powder.[15] Yield 38%; mp: 125–1268C;
ESI-MS: positive mode m/z: 314.0 ([M+H]+), 336.0 ([M+Na]+),
648.9 ([2M+Na]+), negative mode m/z: 347.5 ([M+Cl]À), 624.2
([2M-H]À); Anal (C11H15N5O4S) C, H, N.
(2R,3R,4S,5R)-2-[2-Amino-6-(cyclopentylamino)-9H-purin-9-yl]-5-
(hydroxymethyl)tetrahydrofuran-3,4-diol (14). Cyclopentylamine
(139 mL, 1.41 mmol) and Et3N (197 mL, 1.41 mmol) were added to a
suspension of 4 (200 mg, 0.47 mmol) in EtOH (7 mL). After heating
the mixture at 608C for 18 h it was cooled to RT, and methanolic
NH3 (4 mL) was added to effect complete deprotection of the
sugar moiety. The solvents were evaporated under vacuum, and
the residue was separated by flash column chromatography elut-
ing with CHCl3/MeOH (97:3) to obtain 14 as a white solid.[19] Yield
57%; mp: 214–2158C; ESI-MS: positive mode m/z: 351.2 ([M+H]+),
723.3 ([2M+Na]+), negative mode m/z: 385.1 ([M+Cl]À), 699.3
([2MÀH]À); Anal (C15H22N6O4) C, H, N.
General procedure for the synthesis of 6-thioalkyl derivatives 9–
11. K2CO3 (276 mg, 2 mmol) and the suitable thiol (2 mmol) were
added to a solution of 4 (200 mg, 0.47 mmol) in dry DMF (7 mL),
and the mixture was left in a steel bomb at 1208C for 16 h. After
removal of the volatiles, methanolic NH3 (4 mL) was added, and
the mixture was stirred at RT for 4 h. After removal of the volatiles,
the residue was purified as described for each compound.
(2R,3R,4S,5R)-2-[2-Amino-6-(cyclopentylthio)-9H-purin-9-yl]-5-
(hydroxymethyl)tetrahydrofuran-3,4-diol (9). Compound 4 was
General procedure for the synthesis of compounds 21–23. SOCl2
(2.0 mL) and a catalytic amount of dry DMF (0.05 mL) were added
1078
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1074 – 1080

A Guanosine-Activated GPCR
to compound 20 (200 mg, 0.56 mmol), and the mixture was
heated at 508C for 90 min. After removal of the volatiles, the resi-
due was in turn reacted with the appropriate amine (2 mL) at the
temperature and duration reported for each compound. The mix-
ture was evaporated, and the residue was separated by chroma-
tography to obtain compounds 21–23 as powders.
11.31 ppm (brs, 1H, N1-H); 13C NMR (75 MHz, [D6]DMSO): d=23.8,
32.4, 32.5, 50.6, 73.4, 74.4, 83.8, 86.5, 116.9, 135.8, 152.0, 154.2,
157.1, 169.3 ppm; ESI-MS: positive mode m/z: 365.1 ([M+H]+),
387.1 ([M+Na]+), 750.9 ([2M+Na]+), negative mode m/z: 363.1
([MÀH]À), 399.0 ([M+Cl]À), 727.2 ([2MÀH]À); Anal (C15H20N6O5) C, H,
N.
(3aS,4S,6R,6aR)-6-[2-Amino-6-oxo-1H-purin-9(6H)-yl]-2,2-
dimethyltetrahydrofuro[3,4-d][1,3]dioxole-4-carboxamide (21).
The crude mixture was reacted with methanolic NH3 at RT for 16 h.
Flash chromatography eluting with CHCl3/MeOH (95:5) gave 21 as
a white powder. Yield 72%; mp: 2078C (dec); 1H NMR (400 MHz,
[D6]DMSO): d=1.31 (s, 3H, CH3), 1.49 (s, 3H, CH3), 4.42 (d, J=
2.4 Hz, 1H, H-4’), 5.22 (m, 1H, H-3’), 5.30 (dd, J=2.0 and 6.4 Hz,
1H, H-2’), 6.04 (d, J=2.0 Hz, 1H, H-1’), 6.47 (brs, 2H, NH2), 7.08
(brs, 1H, CONHA), 7.21 (brs, 1H, CONHB), 7.81 (s, 1H, H-8), 10.15
(brs, 1H, N1-H); Anal (C13H16N6O5) C, H, N.
Biological assays
Rat brain membranes
Animal testing was carried out according to the European Com-
munities Council Directive of 24 November, 1986 (86/609/EEC). Ex-
periments were performed on rat brain membranes of male Wistar
rats (200–250 g). Brains without cerebellum were quickly removed
and homogenized in five volumes of ice-cold buffer (50 mm Tris-
HCl pH 7.4, 0.2 mm EGTA) with protease inhibitor cocktails by
using a Politron Ultra-Turrax T8 instrument for 30 s at setting
level 5. The homogenate was centrifuged at 40000 g for 10 min at
48C in a Beckman vacuum ultracentrifuge with a 70 Ti rotor three
times. After each centrifugation the supernatant was discarded
and the pellet was resuspended in a minimum volume of ice-cold
buffer. During the quantification step the membranes were kept
on ice. Protein concentration was measured by the Bradford assay.
Subsequently, the cell membrane quantity (500 mgmLÀ1) was split
into sterile tubes and each tube was centrifuged at 13000 g for
35 min at 48C in a Rotina 38R instrument with a 1711 rotor (Het-
tich Zentrifugen, Tuttlingen, Germany). The supernatant was dis-
carded, and the membrane pellet was snap frozen with liquid ni-
trogen and stored at À808C.
(3aS,4S,6R,6aR)-6-[2-Amino-6-oxo-1H-purin-9(6H)-yl]-N-ethyl-2,2-
dimethyltetrahydrofuro[3,4-d][1,3]dioxole-4-carboxamide (22).
The crude mixture was reacted with EtNH2 at À208C for 90 min.
Flash chromatography eluting with CHCl3/MeOH (98:2) gave 22 as
a
pale-yellow powder.[23] Yield 79%; mp: 1988C (dec); Anal
(C15H20N6O5) C, H, N.
(3aS,4S,6R,6aR)-6-[2-Amino-6-oxo-1H-purin-9(6H)-yl]-N-cyclopen-
tyl-2,2-dimethyltetrahydrofuro[3,4-d][1,3]dioxole-4-carboxamide
(23). The crude mixture was reacted with cyclopentylamine at
À208C for 90 min. Flash chromatography eluting with CHCl3/MeOH
(99:1) gave 23 as a yellow powder. Yield 34%; mp: 2188C (dec);
1H NMR (400 MHz, [D6]DMSO): d=0.89 (m, 1H, H-cyclopentyl), 1.22
(m, 1H, H-cyclopentyl), 1.31 (s, 3H, CH3), 1.40 (m, 5H, H-cyclopen-
tyl), 1.48 (s, 3H, CH3), 3.80 (m, 1H, NHCH), 4.46 (d, J=2.0 Hz, 1H, H-
4’), 5.26 (d, J=6.4 Hz, 1H, H-3’), 5.43 (dd, J=2.0 and 6.0 Hz, 1H, H-
2’), 6.11 (d, J=2.0 Hz, 1H, H-1’), 6.44 (brs, 2H, NH2), 7.09 (d, J=
7.2 Hz, 1H, NHCH), 7.80 (s, 1H, H-8), 10.64 ppm (brs, 1H, N1-H);
Anal (C18H24N6O5) C, H, N.
General procedure for the synthesis of compounds 24–26. A so-
lution of 2n HCl (5 mL), MeCN (3 mL), and H2O (3 mL) was in turn
added to compounds 21–23 (50 mg, 0.15 mmol), and the mixture
was heated at 608C for 2 h. After removal of the volatiles, the resi-
due was lyophilized followed by crystallization from MeOH. Com-
pounds 24–26 were obtained as white solids.
DELFIA Eu-GTP assays
Compounds listed in Table 1 were evaluated for receptor activity at
the putative guanosine receptor in rat brain membranes. Stimula-
tion buffer (1 mm MgCl2, 10 mm GDP, and 20 mm NaCl) was pre-
pared fresh and kept on ice. Rat brain membranes were resuspend-
ed in 50 mm HEPES. Experiments were carried out in 96 well-
plates: wells contained 5 mL solutions of the compounds under
study at various concentrations, 50 mL stimulation buffer, and 50 mL
membrane preparation (10 mg). After incubation at RT for 5 min
under slow shaking, Eu-GTP (10 nm) was added to each well. After
55 min incubation, the plate was washed twice with 300 mL ice-
cold 1ꢂ Eu-GTP wash solution, and fluorescence signal was mea-
sured with the time-resolved fluorometer (GENIosPro, TECAN).
(2S,3S,4R,5R)-5-[2-Amino-6-oxo-1H-purin-9(6H)-yl]-3,4-dihydroxy-
tetrahydrofuran-2-carboxamide (24). Yield 61%; mp: 1988C (dec);
1H NMR (400 MHz, [D6]DMSO): d=4.08 (m, 1H, H-3’), 4.20 (d, J=
1.6 Hz, 1H, H-4’), 4.44 (m, 1H, H-2’), 5.52 (d, J=6.4 Hz, 1H, OH),
5.64 (d, J=4.4 Hz, 1H, OH), 5.74 (d, J=8.0 Hz, 1H, H1’), 6.53 (brs,
2H, NH2), 7.26 (brs, 1H, CONHA), 7.95 (s, 1H, H-8), 8.17 (brs, 1H,
CONHB), 10.15 ppm (brs, 1H, N1-H); 13C NMR (75 MHz, [D6]DMSO):
d=72.2, 73.2, 84.1, 87.0, 117.3, 136.9, 150.6, 153.8, 156.6,
171.8 ppm; Anal (C10H12N6O5) C, H, N.
(2S,3S,4R,5R)-5-[2-Amino-6-oxo-1H-purin-9(6H)-yl]-N-ethyl-3,4-di-
hydroxytetrahydrofuran-2-carboxamide (25).[24] Yield 55%; mp:
2218C (dec); Anal (C12H16N6O5) C, H, N.
Statistical analysis
(2S,3S,4R,5R)-5-[2-Amino-6-oxo-1H-purin-9(6H)-yl]-N-cyclopentyl-
3,4-dihydroxytetrahydrofuran-2-carboxamide (26). Yield 60%;
mp: 2298C (dec); H NMR (400 MHz, [D6]DMSO): d=1.37 (m, 2H, H-
cyclopentyl), 1.48 (m, 2H, H-cyclopentyl), 1.59 (m, 2H, H-cyclopen-
tyl), 1.80 (m, 2H, H-cyclopentyl), 4.02 (m, 1H, NHCH), 4.13 (t, J=
4.0 Hz, 1H, H-4’), 4.27 (d, J=2.4 Hz, 1H, H-3’), 4.48 (t, J=5.6 Hz, 1H,
H-2’), 5.50 (m, 2H, 2ꢂOH), 5.79 (d, J=4.4 Hz, 1H, H-1’), 6.58 (brs,
2H, NH2), 8.06 (d, J=7.6 Hz, 1H, NHCH), 8.29 (s, 1H, H-8),
Responses are expressed as a percentage of the maximal relative
fluorescence and are referenced against guanosine, taken as 100%.
Concentration–response curves were fitted by nonlinear regression
using Prism 4.0 (GraphPAD Software, San Diego, CA, USA). To quan-
tify agonist potency, EC50 values were calculated; EC50 is the con-
centration of agonist required to produce 50% of the maximum
effect. Each concentration was tested five times in triplicate, and
values represent the mean Æ standard error (SE).
1
ChemMedChem 2011, 6, 1074 – 1080
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1079

G. Cristalli et al.
MED
[11] G. Cristalli, R. Volpini, S. Vittori, E. Camaioni, A. Monopoli, A. Conti, S. Di-
onisotti, C. Zocchi, E. Ongini, J. Med. Chem. 1994, 37, 1720–1726.
[12] T. Maruyama, N. Moriwaka, Y. Demizu, M. Ohtsuka, Tetrahedron Lett.
2005, 46, 8225–8228.
[13] A. J. Jankowski, D. S. Wise, L. B. Townsend, Nucleosides Nucleotides 1989,
8, 339–348.
[14] N. Amann, H.-A. Wagenknecht, Synlett 2002, 5, 687–691.
[15] E. Sharon, S. A. Lꢄvesque, M. N. Munkonda, J. Sꢄvigny, D. Ecke, G. Reiser,
B. Fischer, ChemBioChem 2006, 7, 1361–1374.
[16] S. Huss, G. Gosselin, J. L. Imbach, J. Org. Chem. 1988, 53, 499–506.
[17] B. Zhang, H. Muller-Steffner, F. Schuber, B. V. Potter, Biochemistry 2007,
46, 4100–4109.
[18] M. Ohno, Z. G. Gao, P. Van Rompaey, S. Tchilibon, S. K. Kim, B. A. Harris,
A. S. Gross, H. T. Duong, S. Van Calenbergh, K. A. Jacobson, Bioorg. Med.
Chem. 2004, 12, 2995–3007.
[19] B. K. Trivedi, R. F. Bruns, J. Med. Chem. 1989, 32, 1667–1673.
[20] K. Too, D. M. Brown, E. Bongard, V. Yardley, L. Vivas, D. Loakes, Bioorg.
Med. Chem. 2007, 15, 5551–5562.
[21] E. Defrancq, A. Leterme, N. Pelloux, M.-F. Lhomme, J. Lhomme, Tetrahe-
dron 1991, 47, 5725–5736.
[22] J. B. Epp, T. S. Widlanski, J. Org. Chem. 1999, 64, 293–295.
[23] R. C. Foitzik, S. M. Devine, N. E. Hausler, P. J. Scammells, Tetrahedron
2009, 65, 8851–8857.
Acknowledgements
This work was supported by grants from the Italian Ministry for
University and Research (PRIN2008) and the Ministry for Health
(Progetto Ordinario Neurolesi: RF-CNM-2007-662855).
Keywords: Eu-GTP
· GPCR · guanosine · nucleosides ·
receptors
[1] M. P. Rathbone, P. J. Middlemiss, J. W. Gysbers, C. Andrew, M. A. Herman,
J. K. Reed, R. Ciccarelli, P. Di Iorio, F. Caciagli, Prog. Neurobiol. 1999, 59,
663–690.
[2] P. J. Middlemiss, J. W. Gysbers, M. P. Rathbone, Brain Res. 1995, 677,
152–156.
[3] M. Rathbone, L. Pilutti, F. Caciagli, S. Jiang, Nucleosides Nucleotides Nu-
cleic Acids 2008, 27, 666–672.
[4] A. P. Schmidt, D. O. Souza, Open Neurosci. J. 2010, 4, 102–113.
[5] R. Ciccarelli, P. Di Iorio, P. Giuliani, I. D’Alimonte, P. Ballerini, F. Caciagli,
M. P. Rathbone, Glia 1999, 25, 93–98.
[6] J. W. Gysbers, M. P. Rathbone, Neuroreport 1992, 3, 997–1000.
[7] U. Traversa, G. Bombi, P. Di Iorio, R. Ciccarelli, E. S. Werstiuk, M. P. Rath-
bone, Br. J. Pharmacol. 2002, 135, 969–976.
[24] K.-H. Jung, R. R. Schmidt, Liebigs Ann. Chem. 1988, 10, 1013–1014.
[25] G. Cristalli, R. Volpini, Curr. Top. Med. Chem. 2003, 3, 355–469.
[8] U. Traversa, G. Bombi, E. Camaioni, A. Macchiarulo, G. Costantino, C. Pal-
mieri, F. Caciagli, R. Pellicciari, Bioorg. Med. Chem. 2003, 11, 5417–5425.
[9] R. Volpini, M. Buccioni, D. Dal Ben, C. Lambertucci, C. Lammi, G. Marucci,
A. T. Ramadori, K.-N. Klotz, G. Cristalli, J. Med. Chem. 2009, 52, 7897–
7900.
Received: February 17, 2011
Revised: March 11, 2011
[10] H. Frang, V. M. Mukkala, R. Syystç, P. Ollikka, P. Hurskainen, M. Scheinin,
I. Hemmilꢃ, Assay Drug Dev. Technol. 2003, 1, 275–280.
Published online on April 15, 2011
1080
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1074 – 1080