Synthesis and evaluation of 7-amino-2-(2(3)-furyl)-5-phenylethylamino- oxazolo[5,4-d]pyrimidines as potential A2A adenosine receptor antagonists for positron emission tomography (PET)

Article abstract of DOI:10.1016/j.ejmech.2005.07.018

The brain A_2A adenosine receptor (A_2AAR) participates with the dopamine D₂ receptor in the control of movement and also might influence behavior. Because PET is an important tool for studying the roles of receptors in dise

Full text of DOI:10.1016/j.ejmech.2005.07.018

European Journal of Medicinal Chemistry 41 (2006) 7–15
http://france.elsevier.com/direct/ejmech
Original article
Synthesis and evaluation of 7-amino-2-(2(3)-furyl)-5-
phenylethylamino-oxazolo[5,4-d]pyrimidines as potential A2A
adenosine receptor antagonists for positron emission tomography (PET)
Marcus H. Holschbach *, Dirk Bier, Stefan Stüsgen, Walter Wutz, Wiebke Sihver,
Heinz H. Coenen, Ray A. Olsson 1
Institut für Nuklearchemie, Forschungszentrum Jülich GmbH, D-52425 Jülich, Germany
Received 24 March 2005; received in revised form 11 July 2005; accepted 18 July 2005
Available online 09 November 2005
Abstract
The brain A2A adenosine receptor (A2AAR) participates with the dopamine D2 receptor in the control of movement and also might influence
behavior. Because PET is an important tool for studying the roles of receptors in disease, a ligand for imaging the brain A2AAR is desirable. This
report describes the synthesis and A2AAR antagonist activities of a panel of phenyl-substituted 7-amino-2-(2-furyl)-5-phenylethylamino-oxazolo
[5,4-d]pyrimidines, 11aa–af, and their 3-furyl congeners, 11ba–bd. In competitive binding studies all compounds displaced [3H]CGS21680 from
the A2AAR with Ki values of 14–33 nM with selectivity for the A2AAR over the A1AR of 5- to 94-fold. Autoradiography of brain sections
showed a high level of unspecific binding that obscured specific binding. Thus, these compounds are not promising PET ligands.
© 2005 Elsevier SAS. All rights reserved.
Keywords: A2A adenosine receptor; PET; Antagonist; Oxazolo[5,4-d]pyrimidines
1. Introduction
actions are important in the control of movement. A2AARs
with lower densities in the hippocampus and amygdala may
yet importantly affect behavior and may play a role in the
pathogenesis of conditions such as attention deficit hyperactiv-
ity syndrome, panic disorders and schizophrenia [5].
Central nervous system adenosine receptors play important
roles in neurotransmission. The adenosine A1 receptor (A1AR)
is widely distributed in the human brain, the highest densities
being in the cerebral cortex, hippocampus, thalamus and stria-
tum. This receptor is strongly neuromodulatory [1], and may
initiate the neuroprotective effects of ischemic preconditioning
[2]. In comparison autoradiographic studies show that the dis-
tribution of the A2AAR is more restricted, the highest densities
being in the neurons of the striatum, nucleus accumbens and
olfactory tubercle [3]. Many of these neurons coexpress the
A2AAR and the D2 dopaminergic receptor [4]; their reciprocal
Various modes of neuroimaging are revealing important
new information about the roles of various brain receptors in
health and disease. One, positron emission tomography (PET),
provides quantitative information about the distribution and
density of receptors and how diseases may alter those charac-
teristics [6]. Several A2AAR antagonists have been radiola-
beled with carbon-11, but rigorous proof that they were bind-
ing specifically to brain A2AARs is lacking [7]. The present
study continues the search for A2AAR ligands for brain ima-
ging that meet several nuclear medicine, pharmacological and
radiochemical requirements. The ligand should bind specifi-
cally to the A2AAR with high affinity and selectivity. It should
penetrate the blood-brain barrier to an extent that supports the
imaging of brain receptors. Metabolism particularly by brain
should be negligible. Since the A2AAR is a G protein-coupled
* Corresponding author. Tel.: +49 2461 61 6961; fax: +49 2461 61 2535.
E-mail addresses: m.holschbach@fz-juelich.de (M.H. Holschbach),
d.bier@fz-juelich.de (D. Bier), st.stuesgen@fz-juelich.de (S. Stüsgen),
w.wutz@fz-juelich.de (W. Wutz), w.sihver@fz-juelich.de (W. Sihver),
h.h.coenen@fz-juelich.de (H.H. Coenen), r.a.olsson@fz-juelich.de (R.A. Olsson).
1
Present address: Department of Internal Medicine, University of South Florida,
Tampa, FL 33612, USA.
0223-5234/$ - see front matter © 2005 Elsevier SAS. All rights reserved.
doi:10.1016/j.ejmech.2005.07.018

8
M.H. Holschbach et al. / European Journal of Medicinal Chemistry 41 (2006) 7–15
receptor, the ideal radioligand would be an antagonist, because
affinity for the receptor would not depend on the degree of
receptor-G protein coupling. Because the physical half-life of
carbon-11 is so brief (t1/2 = 20.4 min) a longer-lived isotope,
such as fluorine-18 (t1/2 = 109.8 min), that could be dispensed
from a central radiopharmacy rather than generated on-site,
would be advantageous. Finally, to achieve a high radiochemi-
cal yield and in the interest of safety, radiolabeling should oc-
cur late in the synthesis, preferably as the final step.
The known A2AAR antagonists are 6:5 bicyclic or 5:6:5 tri-
ciyclic heterocycles [8]. CGS15493, the first of the non-
xanthine antagonists [9], served as a template for more potent
and selective A2AAR antagonists, which differed according to
the heterocycle but retained the key exocyclic amino and furyl
substituents (or a phenyl group [10]). Although some of those
“second generation” antagonists had affinities in the low pico-
molar range, solubility and bioavailability remained problems
which the introduction of hydrophilic groups in the exocyclic
substituents did not improve [11]. Our previous work [12] on
6-substituted 4-amino-2-phenyl-1,2,4-triazolo[4,3-a]quinoxa-
lones [10] identified several compounds with good affinity
for the A2AAR, but these, too, were poorly water soluble. Ac-
cordingly, it seems reasonable that the poor solubility of such
molecules owes to the heterocyclic bases themselves, and that
the design of new antagonists should consider alternative
bases. A patent application [13] described the synthesis of
some furyl-substituted purines, oxazolopyrimidines and pteri-
dines. Although these compounds were said to antagonize ade-
nosine, the patent contained little data to support that claim.
Nonetheless, we decided to use one class of heterocycles em-
bodied in that patent, the oxazolo[5,4-d]pyrimidines, as a lead
for developing new A2AAR antagonists potentially suitable for
PET.
2. Chemistry
Scheme 1 shows the synthetic pathway to the target com-
pounds. The condensation [14] of thiourea, 1, with diethyl 2-
aminomalonate, 2, generated the starting material, 5-amino-
4,6-dioxo-2-methylmercaptopyrimidine, 3. Alkylation with
methyl iodide protected the mercapto group, permitting chemo-
selective acylation of the 5-amino group of 4 with either 2- or
3-furoyl chloride, forming amides 5a and 5b, respectively.
Treatment with POCl3 cyclized the amides to form a mixture
consisting of about 90–95% of the 7-chlorooxazolo-pyrimi-
dines 6a and 6b and 5–10% of 7-oxo-oxazolopyrimidines 7a
and 7b. This suggests that cyclization of 5a and 5b proceeded
quantitatively and the succeeding chlorination to an extent of
90–95%. The introduction of the 7-amino group consisted of
reacting 6a and 6b with NaN3 and then reducing the 7-azido
group of 8a and 8b with SnCl2 to obtain 9a and 9b. Perborate
oxidation [15] of 9a and 9b converted the methylmercapto
group to a sulfone, a better leaving group for displacement by
nucleophiles. The aminolysis of the sulfone with arylethyla-
mines required careful control of the reaction conditions,
namely, a 4:1 ratio of amine to 10a/b, 20 ml of acetonitrile/
mmol of sulfone and a reaction temperature of 110 °C. Devia-
tions from those parameters led to side reactions, primarily de-
gradative ones. The type of aryl substituent greatly influenced
the time required for the reaction to go to completion, which
ranged between 24 and 96 hours. The reaction sequence out-
Scheme 1. Synthesis of oxazolopyrimidines 11aa–af and 11ba–bd.

M.H. Holschbach et al. / European Journal of Medicinal Chemistry 41 (2006) 7–15
9
Scheme 2. Tritiation of 11aa; synthesis of [3H]11ab.
lined above resulted in the formation of 7-amino-2-(2-furyl)-5-
phenylethylamino-oxazolo[5,4-d]pyrimidines, 11aa–af (a-ser-
ies), and 7-amino-2-(3-furyl)-5-phenylethylamino-oxazolo[5,4-
d]pyrimidines, 11ba–bd (b-series) in acceptable overall yields.
overall yields of the literature synthesis were in the range of
1%. Second, a major aim of no-carrier-added (n.c.a.) radio-
syntheses is the introduction of the radiolabel at the latest step
possible to increase radiochemical yield and promote radiation
safety. In this example, the insertion of the arylethylamino
group, a favored site for radiolabeling, would occur in the sec-
ond step, which is early in the synthesis and, moreover, is one
of the two low-yield steps. Our synthetic strategy was based on
the formation of sulfones 10a and 10b, versatile intermediates
for nucleophilic displacement reactions. Thus, aminolysis of
10a and 10b with various differently substituted phenylethyla-
mines gave the target compounds 11aa–af and 11ba–bd with
yields ranging from 35% to 60%.
For the pharmacological evaluation of the synthesized oxa-
zolopyrimidines compound 11aa was tritiated with commer-
cially available CT3I under basic conditions. [3H]11ab was ob-
tained with a radiochemical yield of 32%, a radiochemical
purity exceeding 98% and a specific radioactivity of 3.1
GBq/μmol corresponding to that of the CT3I used.
For the pharmacological evaluation of these compounds de-
rivative 11ab was radiolabeled with tritium (Scheme 2) by re-
acting the potassium salt of 11aa, formed in situ using KOH in
DMSO, with CT3I.
3. Results and discussion
In the literature synthesis [13] the target compounds were
generated in four steps, beginning with the condensation of
methyl isothiouronium sulfate with 2-(2-furanecarbamoyl)-
malonic acid to give 4,6-dihydroxy-5-(2-furoylamino)-2-
methylmercaptopyrimidine, 5a, in one step. Displacement of
the methylmercapto group by a 2-arylethylamine followed by
simultaneous chlorination/cyclization with POCl3 generated a
5-(2-arylethylamino)-7-chloro-2-(2-furyl)oxazolo[5,4-d]pyrimi-
dine which on reaction with ammonia gave the target com-
pound. Even though this route entailed fewer steps (4 vs. 8
described above) it was not ideal for our purposes for two rea-
sons. First, owing to yields of <20% in the first two steps,
In equilibrium binding studies using [3H]11ab unspecific
binding was very high, accounting for about 95% of total bind-
ing, thus making accurate measurements of a KD value impos-
sible. Tables 1 and 2 summarize the results of competition as-
says and AR subtype selectivity of 11aa–af and 12ba–bd in
Table 1
A1 and A2AAR antagonist affinity, selectivity and hydrophobicity factors of compounds 11aa–af and 11ba–bd
Cpd.
11aa
11ab
11ac
11ad
11ae
11af
11ba
11bb
11bc
11bd
a
R
4-OH
A1ARa
A2AArb
A2A/A1
20
71
68
91
26
14
12
13
log k’w
3.47
3.50
2.97
3.81
4.08
3.27
3.53
4.20
3.53
3.89
360 (165–760)
1570 (120–2080)
1700 (440–5155)
1270 (465–3470)
440 (265–735)
410 (145–1170)
340 (37–3040)
435 (95–1965)
170 (22–1670)
250 (90–670)
18 (12–27)
22 (19–27)
25 (20–30
14 (8–24)
17 (10–28)
30 (11–78)
29 (22–37)
33 (26–43)
32 (25–41)
21 (15–28)
4-OMe
3-OMe, 4-OH
3,4-OMe
3-OMe
2-OMe
4-OH
4-Ome
3,4-Ome
3-Ome
5
12
Ki (mean and 95% CL) vs. [3H]CPDPX, 2.5 nM.
Ki (mean and 95% CL) vs. [3H]CGS21680, 5 nM.
b

10
M.H. Holschbach et al. / European Journal of Medicinal Chemistry 41 (2006) 7–15
Table 2
4. Experimental section
Elemental analyses
2-Furyl series
General methods. Melting points were measured on an Elec-
11aa, C17H15N5O3, Mol. Wt.: 337,33
calc.: C, 60,53; H, 4,48; N, 20,76;
11ab, C18H17N5O3, Mol. Wt.: 351,36
calc.: C, 61,53; H, 4,88; N, 19,93;
11ac, C18H17N5O4, Mol. Wt.: 367,36
calc.: C, 58,85; H, 4,66; N, 19,06;
11ad, C18H17N5O3, Mol. Wt.: 351,36
calc.: C, 61,53; H, 4,88; N, 19,93;
11ae, C18H17N5O3, Mol. Wt.: 351,36
calc.: C, 61,53; H, 4,88; N, 19,93;
11af, C19H19N5O4, Mol. Wt.: 381,39
calc.: C, 59,84; H, 5,02; N, 18,36;
3-Furyl series
11ba, C17H15N5O3, Mol. Wt.: 337,33
calc.: C, 60,53; H, 4,48; N, 20,76;
11bb, C18H17N5O3, Mol. Wt.: 351,36
calc.: C, 61,53; H, 4,88; N, 19,93;
11bc, C18H17N5O3, Mol. Wt.: 351,36
calc.: C, 61,53; H, 4,88; N, 19,93;
11bd, C19H19N5O4, Mol. Wt.: 381,39
calc.: C, 59,84; H, 5,02; N, 18,36;
trothermal™ apparatus and are uncorrected. The “Zentralabtei-
lung für chemische Analysen” at the Forschungszentrum Jue-
lich performed the elemental analyses, which were within
0.4% of the calculated composition. Thin layer chromatogra-
phy (TLC) employed precoated silica sheets (4 × 8 cm, Poly-
gram™, Macherey-Nagel, Düren) developed with ethyl acet-
ate/hexane: 90:10 (v/v). Mass spectra (MS), ESI (positive),
were obtained on a Finnigan Automass III mass spectrometer
Found: C, 60,48; H, 4,43; N, 20,84;
Found: C, 61,57; H, 4,92; N, 19,79;
Found: C, 58,77; H, 4,62; N, 19,17;
Found: C, 61,61; H, 4,84; N, 20,01;
Found: C, 61,50; H, 4,98; N, 19,99;
Found: C, 59,88; H, 4,96; N, 18,47;
1
(Thermo Quest, Dreieich). H and 13C NMR spectra were ob-
tained at 200.13 and 50.32 MHz, respectively, by means of a
Bruker DPX-200 spectrometer (Avance 200) in ≈ 5% solution
in DMSO-d6 at 25 °C. Chemical shifts are given in δ ppm
using the residual proton signal of DMSO-d6 at 2.52 ppm as
a reference. The multiplicity symbols s, d, t and m refer to
singlet, doublet, triplet and multiplet, respectively.
Solvents and reagents, which were of the highest grade
available, were from Sigma-Aldrich, Deisenhofen, Germany,
or Lancaster Synthesis, Muelheim am Main, Germany. N,N-
Dimethylformamide (DMF) and dimethyl sulfoxide (DMSO)
were distilled under argon and stored in lightproof containers
over 4 Å molecular sieves.
Found: C, 60,54; H, 4,41; N, 20,87;
Found: C, 61,44; H, 4,96; N, 19,86;
Found: C, 61,51; H, 4,94; N, 20,04;
Found: C, 59,91; H, 4,94; N, 18,42;
pig brain membranes. All compounds antagonized binding of
[3H]CGS 21680 to the A2AAR with Ki values between 14 and
33 nM; whether the 2-substituent was a 2- or a 3-furyl group
was unimportant. The Ki values for antagonism of binding of
[3H]CPDPX to A1ARs were between 170 and 1700 nM, and,
because the Ki values for inhibition at the A2AAR varied over a
relatively narrow range, affinity for the A1AR importantly in-
fluenced selectivity. Selectivity for the A2AAR over the A1AR
was > 50-fold for compounds 11ab, 11ac and 11ad. Although
all contained either a 3- and/or a 4-methoxyphenyl group, these
results did not yield unambiguous structure-activity relations.
The selectivity of both 11ab (4-methoxy) and 11ad (3-meth-
oxy) were >50-fold, but that of 11af (3,4-dimethoxy) was only
14-fold. By contrast, the selectivity of the 3-furyl derivatives
11ba–bd were rather low, suggesting that the phenyl substitu-
ents are not important for selectivity.
4.1. Chemistry
4.1.1. 5-Amino-2-mercaptopyrimidine-4,6-diol (3) [14]
With exclusion of moisture thiourea (50 g, 657 mmol) and
diethyl aminomalonate hydrochloride (100 g, 473 mmol) were
added to a mechanically stirred solution of sodium ethoxide
formed from sodium (25 g, 1.087 mol) in absolute ethanol
(1.5 l). A thick orange paste formed immediately; this was
heated for 3 hours at reflux, cooled to room temperature and
the solid was collected by filtration through Whatman No. 2
paper. Dissolving the solid material in a minimum amount of
water (~ 3 l) and decantation removed traces of insoluble im-
purities. Acidification with conc. HCl precipitated an orange
crystalline solid that was filtered through Whatman No. 2 pa-
per and washed successively with water, ethanol and diethy-
lether. Drying gave 3 (66 g, 88%), m.p. > 300 °C. MS (m/z)
160.1 [M + H]+.
Tables 1 and 2 also include measurements of a hydrophobi-
city parameter (k’w). Log k’w varied over a rather narrow
range (2.9–4.2) and was unrelated to Ki at either receptor, evi-
dence that hydrophobicity was also not an important determi-
nant of affinity.
In vitro autoradiographic studies showed that neither ligands
selective for the A1AR nor for the A2AAR displaced [3H]11ab
bound to rat brain slices. Thus, the high degree of unspecific
binding obscured specific binding.
4.1.2. 5-Amino-2-methylmercaptopyrimidine-4,6-diol (4) [14]
Iodomethane (66 g, 29 ml, 465 mmol) was added slowly to
a solution of 3 (64 g, 407 mmol) in 400 ml 10% NaOH. The
solution was stirred for 1 hour at room temperature and then
acidified with glacial acetic acid. The precipitated product was
collected, washed with water and dried to give 56 g (73%) of
In summary, the oxazolopyrimidines examined had affi-
nities for the A2AAR in the low-nanomolar range, and some
were quite selective over the A1AR. Despite a high level of
unspecific binding, it was possible to demonstrate specific
binding to A2AARs in equilibrium binding assays of pig brain
membranes. However, unspecific binding obscured specific
binding in in vitro autoradiographic experiments. For that rea-
son these oxazolopyrimidines are not suitable candidates for
imaging brain A2AARs by PET.
1
solid material, m.p. > 300 °C. H NMR: 2.43 (s, 3H, SCH3),
6.15 (br s, 4H, OH, NH2). MS (m/z): 176.2 [M + H]+.
4.1.3. Furan-2-carboxylic acid-(4,6-dihydroxy-2-
methylsulfanyl-pyrimidin-5-yl)-amide (5a)
The addition of 2-furoyl chloride (14.2 g, 10.8 ml,
110 mmol) to an ice-cold solution of 4 (17.3 g, 100 mmol) in

M.H. Holschbach et al. / European Journal of Medicinal Chemistry 41 (2006) 7–15
11
1 N NaOH (250 ml) was followed by stirring for 1 hour at
room temperature. The mixture was carefully acidified with
conc. HCl and stirred for another 30 min. The solid product
was collected by filtration and washed with water, EtOH, and
diethylether. After drying the yield was 21 g (78%), m.p. 299–
300 °C, darkens at 285 °C. 1H NMR: 2.51 (s, 3H, SCH3), 6.64
(s, 1H, furyl-H-4), 7.23 (s, 1H, furyl-H-3), 7.86 (s, 1H, furyl-
H-5), 9.06 (s, 1H, NH), 12.11 (sbr, 2H, OH). 13C NMR: 13.83
(SCH3), 112.73 (furyl-C4), 114.90 (furyl-C3), 132.33 (C5),
141.27 (C2), 146.09 (furyl-C5), 148.62 (furyl-C2), 157.81
(C-OH), 160.76 (C-OH), 163.81 (C=O). MS (m/z): 268.1
[M + H]+.
SCH3), 5.92 (sbr, 2H, NH2), 6.65 (q, 1H, furyl-H-4), 7.24 (dd,
1H, furyl-H-3), 7.69 (dd, 1H, furyl-H-5), 13C NMR: 14.88 (S
CH3), 112.81 (furyl-C4), 114.97 (furyl-C3), 118.01 (C8),
142.24 (C9), 146.42 (furyl-C5), 151.24 (C2), 155.34 (furyl-
C2), 168.89 (C7 + C5). MS (m/z): 249.2 [M + H]+.
4.1.7. 7-Amino-2-furan-2-yl-5-methylsulfonyl-oxazolo[5,4-d]
pyrimidine (10a)
Compound 9a (2.48 g, 10 mmol) was added in one portion
to a well-stirred suspension of NaBO3 × H2O (5 g, 55 mmol) in
glacial acetic acid (100 ml) preheated to 65 °C (oil bath tem-
perature). The yellow suspension was stirred for 3 h at that
temperature, diluted with water (300 ml) and the title com-
pound was collected by filtration (Whatman type 2 paper, no
vacuum initially). The sulfone was obtained as pale yellow so-
lid, which was oven dried at 120 °C for 1 hour and kept in a
desiccator. Yield 2.72 g (97%) off-white solid, m.p. > 250 °C,
darkens at ~ 180 °C. 1H NMR: 3.35 (s, 3H, SO2CH3), 6.84 (q,
1H, furyl-H-4), 7.47 (dd, 1H, furyl-H-3), 8.11 (m, 1H, furyl-H-
5), 8.53 (sbr, 2H, NH2). 13C NMR: 40.09 (SO2CH3), 113.88
(furyl-C4), 116.69 (furyl-C3), 118.01 (C8), 141.74 (C9),
148.51 (furyl-C5), 153.31 (C2), 157.51 (furyl-C2), 161.39 (
C5), 163.78 (C7). MS (m/z): 282.1 [M + 1]+.
4.1.4. 7-Chloro-2-furan-2-yl-5-methylsulfanyl-oxazolo[5,4-d]
pyrimidine (6a)
A suspension of 5a (13.35 g, 50 mmol) in POCl3 (100 ml)
was stirred for 3 hours at 120 °C. Excess POCl3 was distilled
off (oil bath 140 °C) in vacuo, the residue azeotroped twice
with toluene (25 ml), poured into ice/water (100 ml) and stirred
vigorously for 30 min. The solid product was dissolved in re-
fluxing ethanol (~ 350 ml) for 15–20 min and filtered to re-
move insoluble impurities. Upon cooling 6a, contaminated
with some 7a, precipitated. The solid was filtered off, air-dried
and used for the next step without further purification. Yield
19.6 g, 73%.
4.1.8. 3-Furoyl chloride [16]
4.1.5. 7-Azido-2-furan-2-yl-5-methylsulfanyl-oxazolo[5,4-d]
pyrimidine (8a)
A mixture of 3-furoic acid (33.6 g, 300 mmol) and thionyl
chloride (69.3 g, 550 mmol) in 1,2-DCE (350 ml) was refluxed
for 12 h. Careful rotary evaporation removed the solvent and
excess thionyl chloride; distillation of the residue under re-
duced pressure gave the acid chloride as a colorless liquid
which solidified on standing. Yield 32.5 g, 83%; bp50 62–
63 °C (Ref. [16] bp47 65 °C), m.p. 28–29 °C (Ref. [16] m.p.
29 °C).
Sodium azide (3 g, 46 mmol) was added to a solution of
crude 6a/7a (11.2 g, 42 mmol) in dry DMF (100 ml). After
stirring for a few minutes at room temperature solid NaCl be-
gan to precipitate and after 12 h the reaction mixture was cen-
trifuged and the supernatant was decanted into water (150 ml).
The precipitated product was collected by filtration (7a, carried
over from the previous step, remained in solution), dissolved in
a minimum volume of boiling isopropanol, filtered while hot
and cooled to give 8a, which was pure by TLC. Yield 10.97 g,
1
4.1.9. Furan-3-carboxylic acid-(4,6-dihydroxy-2-
methylsulfanyl-pyrimidin-5-yl)-amide (5b)
95%, m.p. 151–2 °C. H NMR: 2.66 (s, 3H, SCH3), 6.64 (q,
1H, furyl-H-4), 7.35 (dd, 1H, furyl-H-3), 7.71 (dd, 1H, furyl-
H-5), 13C NMR: 15.08 (SCH3), 113.08 (furyl-C4), 116.8
(furyl-C3), 118.92 (C8), 141.56 (C9), 147.28 (furyl-C5),
153.04 (C2), 153.37 (furyl-C2), 168.81 (C7 + C5). MS (m/z):
275.2 [M + H]+.
The title amide was prepared as described for 5a. The yield
was 72%, m.p. 191–2 °C, darkens at 175 °C. 1H NMR, δ: 2.53
(s, 3H, –SCH3), 6.95 (d, 1H, furyl H-4), 7.75 (t, 1H, furyl H-
5), 8.28 (s, 1H, furyl H-2), 8.96 (s, 1H, –NH), 11.98 (br s, 2H,
–OH). 13C NMR, δ: 13.81 (–SCH3), 110.16 (furyl C-4), 123.50
(furyl C-3), 132.41 (C-5), 144.84 (furyl C-5), 144.88 (C-2)
146.57 (furyl C-2), 160.63 (C-OH), 161.75 (C-OH), 163.83
(C=O). MS calc. for C10H9N3O4S, molecular weight 267.26:
(m/z): 268.1 [M + H]+.
4.1.6. 7-Amino-2-furan-2-yl-5-methylsulfanyl-oxazolo[5,4-d]
pyrimidine (9a)
The slow addition of 8a (5.49 g, 20 mmol) in acetone
(400 ml) to a solution of SnCl2 × 2 H2O (9 g, 40 mmol) in
MeOH (50 ml) caused immediate gas evolution. After the ad-
dition the mixture was heated to 50 °C and stirred for an addi-
tional 2 h, concentrated in vacuo to ~ 50 ml, made alkaline
with conc. NH4OH (100 ml) and extracted twice with THF
(100 ml). Drying of the extracts over anhydrous Na2SO4, fil-
tration and rotary evaporation gave crude 9a as a yellow solid
that was recrystallized from EtOH. Yield 4.2 g, 85%, m.p.
4.1.10. 7-Chloro-2-furan-3-yl-5-methylsulfanyl-oxazolo[5,4-d]
pyrimidine (6b)
This compound was prepared as described for 6a. Pyrimi-
dine 6b, contaminated with some hydroxy compound, 7b,
(~ 5%) was obtained in 73% yield and used for the next step
without further purification.
1
189–90 °C. Rfazide: 0.71, Rfamine: 0.61. H NMR: 2.62 (s, 3H,

12
M.H. Holschbach et al. / European Journal of Medicinal Chemistry 41 (2006) 7–15
4.1.11. 7-Azido-2-furan-3-yl-5-methylsulfanyl-oxazolo[5,4-d]
pyrimidine (8b)
eous layer was reextracted with ethyl acetate (30 ml). The
pooled organic phases were washed with water (50 ml), brine
(50 ml) and dried over anhydrous Na2SO4. Filtration and ro-
tary evaporation at 30–35 °C gave the products (almost pure
by TLC). Further purification consisted of taking up the pro-
ducts in acetone (~ 10 ml/100 mg) treatment with ultrasound,
separation of insoluble material by filtration and rotary eva-
poration of the solvent at 30–35 °C. Recrystallization from
suitable solvents gave analytical samples.
The azide was prepared as described for 8a. Pure 8b was
obtained in a yield of 91% after recrystallization from 2-pro-
1
panol. TLC (Rf 0.71), m.p. 142–3 °C (2-propanol). H NMR,
δ: 2.88 (s, 3H, –SCH3), 7.08 (q, 1H, furyl H-4), 7.98 (dd, 1H,
furyl H-5), 8.78 (dd, 1H, furyl H-2).13C NMR, δ: 15.08
(–SCH3), 113.08 (furyl C-4), 116.8 (furyl C-3), 118.92 (C-
8), 141.56 (C-9), 147.28 (furyl C-5), 153.04 (C-2), 153.37
(furyl C-2), 168.81 (C-7 and C-5). MS calc. for
C10H6N602S, molecular weight 274.26: (m/z): 275.1 [M + H]
4.1.14.2. Liquid amines, method B. The hot reaction mixture
was poured into ice/water (60 ml), the solid that separated
was taken up in ethyl acetate (100 ml) and worked up in the
same way as the products of solid amines.
+
.
4.1.12. 7-Amino-2-furan-3-yl-5-methylsulfanyl-oxazolo[5,4-
d]pyrimidine (9b)
4.1.15. 2-Furan-2-yl-N5-[2-(4-hydroxyphenyl)-ethyl]-oxazolo
[4,5-d]pyrimidine-5,7-diamine (11aa)
Compound 9b was prepared as described for 9a. The title
compound was purified by recrystallization from EtOH fol-
lowed by trituration of the crystals with ethyl acetate/hexane
90:10 (v/v) to remove traces of impurities. Yield 95%, m.p.
211–3 °C. TLC: Rfazide: 0.71, Rfamine: 0.61. 1H NMR, δ: 2.49
(s, 3H, –SCH3); 6.98 (q, 1H, furyl H-4); 7.77 (br s, 2H, –N
H2); 7.94 (t, 1H, furyl H-5); 8.55 (q, 1H, furyl H-2). 13C
NMR, δ: 14.57 (–SCH3), 109.18 (furyl C-4), 113.43 (C-8),
115.32 (furyl C-3) 144.88 (C-9), 145.60 (furyl C-2), 146.37
(furyl C-5), 153.09 (C-2), 165.16 (C-5), 167.32 (C-7). MS
calc. for C10H8N4O2S, molecular weight 248.26: (m/z):
249.2 [M + H]+.
Sulfone 10a reacted with 2-(4-hydroxyphenyl)-ethylamine
(tyramine, 550 mg) in 48 h. Rf sulfone: 0.54, Rf product: 0.48.
Method A gave a solid yellow residue for recrystallization
from ethanol (reflux, rt, freezer), yield 205 mg (61%), m.p.
1
211–2 °C (ethanol). H NMR, δ: 2.51 (t, 2H, -CH2Ph); 3.42
(m, 2H, CH2NH-); 6.69 (d, 2H, phenyl H); 6.75 (q, 1H, furyl
H-4); 6.94 (br s, –NH–); 7.06 (d, H, 2 phenyl H); 7.17 (d, 2H,
furyl-H-3) 7.36 (br s, 2H, –NH2), 7.96 (m, 1H, furyl H-5),
9.19 (br s, 1H, PhOH)). 13C NMR, δ: 35.19, 113.23,
113.32, 115.95, 130.40, 130.65, 142.82, 146.70, 156.43,
160.56, 166.64. MS calc for C17H15N5O3, molecular weight
337.33: (m/z) 338.2 [M + H]+.
4.1.13. 7-Amino-2-furan-3-yl-5-methylsulfonyl-oxazolo[5,4-
d]pyrimidine (10b)
Compound 10b was prepared as described for 10a in a
yield of 91% as an off-white solid, m.p. > 250 °C, darkens
4.1.16. 2-Furan-2-yl-N5-[2-(4-methoxyphenyl)-ethyl]-oxazolo
[4,5-d]pyrimidine-5,7-diamine (11ab)
1
at ~ 180 °C. H NMR, δ: 3.36 (s, 3H, SO2CH3), 7.02 (q,
Sulfone 10a reacted with 2-(4-methoxyphenyl)-ethylamine
(605 mg) in 24 h. Rf sulfone: 0.54, Rf product: 0.63. Method B
gave a reddish solid that was dissolved in a minimum amount
of hot methoxyethanol, cooled and diluted by volume with
water. The product precipitated as an orange solid, yield
1H, furyl-H-4), 7.97 (t, 1H, furyl-H-5), 8.45 (sbr, 2H, NH2),
8.63 (q, 1H, furyl-H-2). 13C NMR (DMSO-d6), δ: 40.08
(SO2CH3), 109.28 (furyl-C4’), 114.93 (C8), 117.95 (furyl-
C3’), 146.62 (furyl-C2’ and furyl-C5’) 156.59 (C9), 157.30 (
C2), 161.27 (C5), 163.86 (C7). MS calc. For C10H8N4O4S,
molecular weight 280.26: (m/z): 281.1 [M + H]+.
1
197 mg (56%), m.p. 182–3 °C (50% methoxyethanol). H
NMR, δ: 2.75 (t, 2H, PhCH2–); 3.41 (qbr, 2H, –CH2NH–);
3.72 (s, 3H, –OCH3); 6.74 (q, 1H, furyl-H-4), 6.84 (dbr, 3H,
2ArH + NH), 7.19 (dbr, 3H, 2ArH + furyl-H-3); 7.27 (br s,
2H, –NH2); 7.96 (s, 1H, furyl H-5). 13C NMR, δ: 35.19 (Ph
CH2), 43.96 (–NHCH2–), 113.22 (furyl C-4), 113.32 (furyl C-
3), 115.95 (PhCH–); 130.39 (PhCH–), 130.65 (C-8), 142.82
(C-9), 146.67 (furyl-C5), 156.43 (C2), 160.56 (furyl-C2),
161.39 (C5), 163.78 (C7). MS calc for C18H17N5O3, molecu-
lar weight 351.36: (m/z): 352.1 [M + H]+.
4.1.14. General procedure for the reaction of 10a with
amines
A suspension of 10a (280 mg, 1 mmol) in acetonitrile
(20 ml) and the appropriate amine (4 mmol) was refluxed
gently (oil bath 105–110 °C) for the specified time. During
the reaction with solid amines the suspension changed color
from off-white to dark yellow or light brown; liquid amines
gave clear solutions that turned orange–brown. Workup
(Method A or B, vide infra) depended on the type of amine.
Ethyl acetate/hexane 50/50 (v/v) was the solvent for TLC of
the products.
4.1.17. 2-Furan-2-yl-N5-[2-(4-hydroxy-3-methoxyphenyl)-
ethyl]-oxazolo[4,5-d]pyrimidine-5,7-diamine (11ac)
Dissolving the hydrochloride salt of 2-(4-hydroxy-3-meth-
oxyphenyl)-ethylamine (896 mg, 4.4 mmol) in 32% NH3
(30 ml) liberated the free base which was obtained in ~ 90%
yield by extraction with THF (2 × 30 ml), rotary evaporation
and azeotropic drying with CH3CN (30 ml). The reaction of
4.1.14.1. Solid amines, method A. The hot reaction mixture
was diluted with water (20 ml) and poured onto crushed ice
(~ 100 g). The solid that separated was taken up in ethyl acet-
ate (100 ml), extracted with 0.5 N HCl (50 ml) and the aqu-
the amine with 10a required 168 h. Rf sulfone: 0.54, Rf product
:

M.H. Holschbach et al. / European Journal of Medicinal Chemistry 41 (2006) 7–15
13
0.44. The yellowish solid obtained by method B was tritu-
4.1.21. General procedure for the reaction of 10b with
amines
rated with hot acetonitrile, yield 132 mg, 36%, m.p.
1
196–7 °C. H NMR, δ: 2.73 (t, 2H, ArCH2–); 3.45 (q, 2H,
Gentle reflux (oil bath 105–110 °C) of a mixture of 10b
(280 mg, 1 mmol) in acetonitrile (20 ml) and the appropriate
amine (4 mmol) for the specified time gave the target com-
pound. During the course of the reaction the suspension
cleared and changed color from light to dark yellow. Refrig-
eration usually precipitated crude products, which were sepa-
rated and crystallized from an appropriate solvent. If cooling
did not precipitate the product the reaction mixture was
poured onto crushed ice/water (~ 100 g). The solid that sepa-
rated was taken up in ethyl acetate (100 ml), extracted with
0.5 N HCl (50 ml) and the aqueous layer was back-extracted
with ethyl acetate (30 ml). The pooled organic phases were
washed with water (50 ml), brine (50 ml) and dried over an-
hydrous Na2SO4. Rotary evaporation at 30–35 °C gave the
products that, although pure by TLC (silica, eluent acetone/
hexane 50:50 (v/v), required further purification, which con-
sisted of taking up the products in acetone (~ 10 ml/100 mg)
treatment with ultrasound, separation of insoluble impurities
by filtration and rotary evaporation of the solvent at 30–35 °
C. Recrystallization from suitable solvents gave analytical
samples.
–NHCH2–); 3.77 (s, 3H, OCH3); 6.75 (m, 5H, –NH–, furyl
H-4 and 3 phenyl H); 7.18 (dd, 1H, furyl H-3); 7.25 (br s, 2H,
–NH2); 796, (q, 1H, furyl H-5); 8.71, (s, 1H, PhOH). 13C
NMR, δ: 35.65, 40.10, 43.83, 56.41, 113.07, 113.30,
113.71, 113.90, 116.21, 116.71, 121.67, 131.45, 141.77,
142.90, 145.58, 146.59, 148.25, 157.16, 157.52, 161.08,
161.41, 163.80, 166.65. MS calc for C18H17N504, molecular
weight 367.36: (m/z) 368.2 [M + H]+.
4.1.18. 2-Furan-2-yl-N5-[2-(3-methoxyphenyl)-ethyl]-oxazolo
[4,5-d]pyrimidine-5,7-diamine (11ad)
2-(3-Methoxyphenyl)-ethylamine (605 mg), reacted with
10a in 72 h. Rf sulfone: 0.54, Rf product: 0.59. Method B gave
an orange–red solid that was triturated with hot acetonitrile,
1
yield 214 mg, 61%, m.p. 166-7 °C. H NMR, δ: 2.83 (t, 2H,
PhCH2–); 3.55 (q, 2H, –NHCH2–); 3.76 (s, 3H, OCH3); 6.80
(m, 5H, NH2, furyl H-4, 2 phenyl H); 7.22 (m, 4H, –NH–,
furyl H-3, 2 phenyl H); 7.96, (s, 1H, furyl H-5). 13C NMR, δ:,
55.76, 112.27, 113.08, 113.29, 115.20, 121.82, 130.39,
142.31, 142.90, 144.88, 146.59, 157.17, 160.14, 161.05,
166.64. MS calc for C18H17N5O3, molecular weight 351.36:
(m/z): 352.1 [M + H]+.
4.1.22. 2-Furan-3-yl-N5-[2-(4-hydroxyphenyl)-ethyl]-oxazolo
[4,5-d]pyrimidine-5,7-diamine (11ba)
The reaction of 10b with 2-(4-hydroxyphenyl)-ethylamine,
4.1.19. 2-Furan-2-yl-N5-[2-(2-methoxyphenyl)-ethyl]-oxazolo
[4,5-d]pyrimidine-5,7-diamine (11ae)
(tyramine, 550 mg) required 96 h. Rf sulfone: 0.56, Rf product
:
0.36. After ~ 4 h at 105 °C a solid started to precipitate from
the initially homogeneous reaction mixture. The hot mixture
was filtered and the filtrate set aside; crude product contami-
nated with a red impurity precipitated slowly. The precipitate
was dissolved in acetonitrile (~ 10 ml) and cooled to room
temperature. After some hours the clear yellow solution con-
taining the product was decanted from a solid red impurity.
Refrigeration caused the slow crystallization of colorless pro-
duct. Yield 205 mg (61%), m.p. 143–4 °C (CH3CN). 1H
NMR, δ: 2.51 (t, 2H, ArCH2), 3.42 (qbr, 2H, CH2NH), 6.69
(d, 2H, 2ArH), 6.75 (q, 1H, furyl-H-4), 6.94 (sbr, NH), 7.06
(d, H, 2ArH), 7.17 (d, 2H, furyl-H-3) 7.36 (sbr, 2H, NH2),
7.96 (m, 1H, furyl-H-5), 9.19 (sbr, 1H, OH)). 13C NMR δ:
35.31, 42.91, 109.14, 113.99, 115.64, 130.41, 133.01,
144.27, 146.11, 150.67, 157.14, 158.36, 161.02, 166.84. MS
calc for C18H17N504, molecular weight 337.33: (m/z) 338.1
[M + H]+.
2-(2-Methoxyphenyl)-ethylamine (605 mg), reacted with
10a in 28 h. Rf sulfone: 0.54, Rf product: 0.68. The orange solid
obtained by method B was triturated with hot acetonitrile,
1
yield 179 mg, 51%, m.p. 164–5 °C. H NMR, δ: 2.85 (t,
2H, PhCH2–); 3.47 (m, 2H, –NHCH2–); 3.81 (s, 3H OCH3);
676 (q, 1H, furyl H-4); 6.90 (m, 2H, furyl H-3 and phenyl H);
7.05 (br s, 1H, –NH–); 7.21 (m, 3H, phenyl H); 7.49 (br s,
2H, –NH2); 7.97 (d, 1H, furyl H-5). 13C NMR, δ: 30.42,
42.15, 56.16, 108.74, 111.49, 113.32, 121.11, 128.32,
128.37, 130.93, 142.76, 146.73, 148.14, 158.13, 166.63. MS
calc for C18H17N5O3, molecular weight 351.36: (m/z): 352.2
[M + H]+.
4.1.20. N5-[2-(3,4-dimethoxyphenyl)-ethyl]-2-furan-2-yl-
oxazolo[4,5-d]pyrimidine-5,7-diamine (11af)
Liquid 2-(3,4-dimethoxyphenyl)-ethylamine (725 mg), re-
acted with 10a in 72 h. Rf sulfone: 0.54, Rf product: 0.47. Work-
up B gave a yellow solid that was triturated with hot acetoni-
trile, yield 198 mg, 52%, m.p. 165–6 °C. 1H NMR, δ: 2.78 (t,
2H, ArCH2); 3.47 (m, 2H, –CH2NH–), 6.75 (q, 1H, furyl H-
4), 6.78 (d, 1H, phenyl H), 6.86 (m, 3H, 2 phenyl H and –
NH-), 7.06 (d, 2H, 2 phenyl H), 7.17 (dd, 2H, furyl H-3) 7.25
(br s, 2H, –NH2), 7.96 (m, 1H, furyl H-5). 13C NMR, δ:
35.59, 43.71, 56.26, 56.38, 108.77, 112.76, 113.09, 113.30,
113.44, 133.16, 142.89, 146.59, 14782, 148.01, 149.47,
157.14, 161.06, 166.64, MS calc. for C19H19N5O4, molecular
weight 381.39: (m/z) 382.2 [M + H]+.
4.1.23. 2-Furan-3-yl-N5-[2-(4-methoxyphenyl)-ethyl]-oxazolo
[4,5-d]pyrimidine-5,7-diamine (11bb)
The reaction of 10b with liquid 2-(4-methoxyphenyl)-ethy-
lamine (605 mg) required 96 h. Rf sulfone: 0.54, Rf product: 0.61.
The reddish solid was triturated with hot acetonitrile and
cooled in the refrigerator. Product was filtered off and air
dried, yield 197 mg (56%), off-white solid, m.p. 187–8 °C.
1H NMR, δ: 2.82 (t, 2H, -CH2Ph); 3.43 (m, 2H, –NHCH2–);
3.73 (s, 3H, –OCH3); 665 (m, 5H, 3 phenyl H, –NH– and
furyl H); 7.18 (m, 4H, –NH2 and 2 phenyl H); 7.90 (m, 1H,

14
M.H. Holschbach et al. / European Journal of Medicinal Chemistry 41 (2006) 7–15
furyl H); 8.43 (m, 1H, furyl H). 13C NMR, δ: 35.18, 43.87,
55.82, 109.06, 114.58, 115.87, 130.48, 132.61, 144.32,
146.06, 15.043, 156.91, 158.45, 160.92, 166.80. MS calc for
C18H17N5O3, molecular weight 351.36: (m/z): 352.2
[M + H]+.
(v/v) + 0.1% Et3N. Flow 3.5 ml/min. UV at 260 nm. k’11aa
= 1.45, k’11ab = 5.35.
Quality control: Column LiChrospher 100, RP-18, 5 μm,
250 × 4 mm. Eluent as above. Flow 1 ml/min. UV as above.
k’11aa = 0.96, k’11ab = 3.88. For continuous measurement of
radioactivity the outlet of the UV detector was connected to
an in-line scintillation analyzer (Radiomatic™ 515 TR Series,
Packard, Dreieich, Germany) and the recorded data were pro-
cessed by an integrated software system.
Measurement of the specific radioactivity of the tritiated
compounds used HPLC, which measured the mass of product
by integrating the UV absorption of the product peak and
referenced to a standard curve. The product peak was col-
lected and an aliquot counted in a liquid scintillation counter.
4.1.24. 2-Furan-3-yl-N5-[2-(3-methoxyphenyl)-ethyl]-oxazolo
[4,5-d]pyrimidine-5,7-diamine (11bc)
The reaction of 10b with 2-(3-methoxyphenyl)-ethylamine
(605 mg) required 96 h. Rf sulfone: 0.56, Rf product: 0.59. The
orange–red solid that crystallized from the reaction mixture
was recrystallized from acetonitrile, yield 155 mg, 44%, m.
1
p. 146–7 °C, off-white crystals. H NMR, δ: 2.87, (t, 2H,
CH2-Ph); 3.45 (m, 2H, –NHCH2–); 3.75 (s, 3H, –OCH3);
6.73 (m, 3H, 1 phenyl H, –NH– and furyl H); 7.17 (m, 5H,
–NH2 and 3 phenyl H); 7.88 (m, 1H, furyl H); 8.43 (M, 1H,
furyl H). 13C NMR, δ: 34.91, 42.14, 108.92, 111.72, 114.24,
115.72, 120.12, 129.96, 130.11, 144.31, 146.34, 149.17,
151.02, 156.93, 158.41, 160.99, 161.43, 166.82. MS calc for
C18H17N5O3, molecular weight 351.36: (m/z): 352.1 [M + H]
4.2. Radioligand binding assays
Corpora striata (for A2AAR assays) and frontal cortices (for
A1AR assays) were dissected from pig brains and the tissue
was homogenized for 1 min in 20 volumes of ice-cold 50 mM
Tris–HCl buffer, pH 7.4 containing 10 mM MgCl2, soybean
trypsin inhibitor (20 μg/ml), bacitracin (200 μg/ml), and ben-
zamidine HCl (160 μg/ml) by means of an Ultra Turrax at
20,000 rpm. The homogenate was centrifuged at 48,000 × g
for 10 min at 4 °C (Beckmann Optima L, SW41Ti rotor). The
pellet was suspended in 20 volumes of Tris–HCl, pH 7.4,
containing adenosine deaminase (2 U/ml) and trypsin inhibi-
tor (20 μg/ml), and then was incubated for 30 min at 37 °C.
After centrifugation at 48,000 × g for 10 min at 4 °C the pel-
let was diluted in 20 volumes of 50 mM Tris–HCl, pH 7.4,
containing 10 mM MgCl2. Aliquots of the homogenate (1 ml)
were stored at –80 °C.
The assays were performed in triplicate by incubating ali-
quots of the membrane fractions (68–87 μg protein per assay)
in Tris–HCl, pH 7.4, containing adenosine deaminase (2 U/
ml), [3H]CPFPX (2.5 nM) and cortical homogenates for the
A1AR or [3H]CGS21680 (5 nM) and striatal homogenates for
A2AAR, in a total assay volume of 200 μl. Incubation was
carried out at 20 °C for 90 min. Centrifugation at 48,000 ×
g for 6 min at 8 °C separated bound from free ligand. Super-
natants were discarded, the pellets washed with ice cold buf-
fer (1 ml) and dissolved by incubating in SolvableTM (500 μl,
Canberra-Packard) for 120 min at 50 °C. Aliquots of 450 μl
were placed in scintillation vials with scintillation cocktail
(10 ml, Ultima Gold XR, Canberra-Packard). Radioactivity
was measured in a Liquid Scintillation Analyzer. Protein es-
timation was performed with a commercial assay (Bio-Rad
DC Protein Assay) after solubilization in 15% NH4OH con-
taining 2% SDS (w/v); human serum albumin served as a
standard. A computer-assisted curve-fitting program (Graph-
Pad Prism, version 3.0) calculated IC50, Ki and KD.
+
.
4.1.25. N5-[2-(3,4-Dimethoxyphenyl)-ethyl]-2-furan-3-yl-
oxazolo[4,5-d]pyrimidine-5,7-diamine (11bd)
The reaction of 10b with 2-(3,4-dimethoxyphenyl)-ethyla-
mine (725 mg) required 72 h. Rf sulfone: 0.56, Rf product: 0.54.
The yellow solid was triturated with acetonitrile, yield
210 mg, 55%, m.p. 161–3 °C. 1H NMR δ: 2.78 (t, 2H,
–CH2Ph), 3.47 (m, 2H, –CH2NH–), 6.75 (q, 5H, –NH–, furyl
H and 3 phenyl H); 7.21 (br s, 2H –NH2); 7.90 (m, 1H, furyl
H); 8.43 (m, 1H, furyl H). 13C NMR, δ: 35.60, 43.66, 56.28,
56.39, 108.79, 109.06, 112.79, 113.48, 115.83, 121.37,
133.17, 144.36, 146.08, 148.04, 149.49, 150.54, 156.71,
160.69, 166.78. MS calc. For C19H19N5O4, molecular weight
381.39: (m/z) 382.2 [M + H]+.
4.1.26. Radiosynthesis of [3H]11ab
CT3I (85 Ci/mmol, 1 mCi/μl in toluene) was from Amer-
sham Bioscience. The labeling reaction was performed in a
conical 1 ml glass vial sealed with a Teflon coated rubber
septum and containing a magnetic stirring bar.
The precursor 11aa (1.7 mg, 5 μmol) was added to a stir-
red suspension of powdered KOH (1.1 mg, 20 μmol) in dry
DMSO (250 μl) and the mixture was stirred at ambient tem-
perature for 1 h. [3H]CH3I (10 mCi, 10 μl of the toluene stock
solution in 240 μl dry DMSO) was added, the mixture was
stirred at ambient temperature for 30 min and the reaction
stopped by the addition of eluent (250 μl). The product was
isolated by semipreparative HPLC, the eluent was evaporated
in vacuo at 40 °C and the residue dried by azeotroping with
MeCN (2 × 5 ml). For in vitro studies the tritiated compound
was taken up in DMSO. Radiochemical yield for [3H]11ab
was 32%, the radiochemical purity exceeded 98% and the
specific activity corresponded to that of the CT3I used.
Semipreparative HPLC: 2 Columns in series of Multospher
120, RP-18, 5 μm, 250 × 8 mm. Eluent: MeCN/H2O 50:50
4.3. Measurement of an index of lipophilicity
A method employing HPLC [17] served for estimates of
the lipophilicity of all compounds. This method is based on

M.H. Holschbach et al. / European Journal of Medicinal Chemistry 41 (2006) 7–15
15
measurements of the retention times of the test compounds on
Acknowledgements
a 4.6 × 250 mm column of 5 μm Kromasil RP18 eluted at a
rate of 1 ml/min with each of the following proportions (v/v)
of methanol/water: 90:10, 85:15, 80:20, 75:25, 70:30 and
monitoring of the effluent at 260 nm. Converting each datum
to a capacity factor k’ by the formula k’ = (t – to)/to, where t
and to are the retention time and void time, respectively, and
solving the linear regression of log k’ on methanol: water ratio
for methanol = 0 gave the lipophilicity parameter log k’w.
Professor R.A. Olsson is in part supported by the Ed C.
Wright Chair in Cardiovascular Research, University of South
Florida.
References
[1] T.V. Dunwiddie, Int. Rev. Neurobiol. 27 (1985) 63–139.
[2] C. Heurteaux, I. Lauritzen, C. Widmann, M. Lazdunski, Proc. Natl.
Acad. Sci. USA 92 (1995) 4666–4670.
4.4. Autoradiography
[3] E. Lindström, E. Ongini, B.B. Fredholm, Naunyn Schmiedebergs Arch.
Pharmacol. 354 (1996) 539–541.
[4] B. Johansson, V. Georgiev, B.B. Fredholm, Neuroscience 80 (1997)
1187–1207.
Frozen rat brains (–20 °C) were sectioned at a thickness of
20 μm in the horizontal plane (Leica AG Microsystems, Ger-
many), mounted on gelatine-coated microscope slides (Labor-
optik GmbH, Germany), dried at 4 °C and stored at –80 °C
until use. Tissue sections were preincubated for 15 min at
4 °C in 50 mM Tris–HCl, pH 7.4, containing 1 mM MgCl2
and 2 U/ml adenosine deaminase, and then for 90 min at 23 °
C in the same buffer containing either 0.9 nM [3H]CPDPX,
13 nM [3H]11ab or 2.2 nM [3H]ZM241385, alone or with
different concentrations of 11ab. The sections were washed
twice for 5 min in 50 mM Tris–HCl at 4 °C, rinsed in deio-
nized water, dried in a stream of cold air and placed on a
phosphor image plate (Fuji, Tokyo, Japan) and exposed for
4 days. Laser scanning of the plates employed a phosphor
imager BAS 5000 (Fuji) controlled with software provided
by the vendor (Raytest Isotopenmeßgeräte, Straubenhardt,
Germany, version 2.11a). The advanced image analyzer pro-
gram (AIDA 3.10, Raytest Isotopenmeßgeräte) processed data
from regions of interest (ROIs). Analysis of binding data ar-
rayed in Microsoft Excel 2000 employed a non-linear, curve-
fitting analysis program (GraphPad Prism 3.00, GraphPad
Software, Inc., San Diego, CA 92130).
[5] J.-L. Moreau, G. Huber, Brain Res. Rev. 31 (1999) 65–82.
[6] M.E. Phelps, Proc. Natl. Acad. Sci. USA 97 (2000) 9226–9233.
[7] M.H. Holschbach, R.A. Olsson, Curr. Pharm. Des. 8 (2002) 99–110.
[8] B. Cacciari, G. Pastorin, G. Spalluto, Curr. Top. Med. Chem. 3 (2003)
403–411.
[9] J.E. Francis, W.D. Cash, S. Psychoyos, G. Ghai, P. Wenk, R.C. Fried-
man, C. Atkins, V. Warren, P. Furness, J.L. Huyn, G.A. Stone, M. De-
sai, M. Williams, J. Med. Chem. 31 (1988) 1014–1020.
[10] V. Colotta, D. Catarzi, F. Varano, L. Cecchi, G. Filacchioni, C. Martini,
L. Trincavelli, A. Lucacchini, Arch. Pharm. Med. Chem. 332 (1999)
39–41.
[11] P.G. Baraldi, B. Cacciari, R. Romagnoli, G. Spalluto, A. Monopoli, E.
Ongini, K. Varani, P.A. Borea, J. Med. Chem. 45 (2002) 115–126.
[12] M.H. Holschbach, D. Bier, W. Wutz, W. Sihver, M. Schüler, R.A. Ols-
son, Eur. J. Med. Chem. 40 (2005) 421–437.
[13] M.H. Block, A. Harrison, R.B. Hargreaves, Eur. Pat. Appl., 544,445, 2
Jun 93, Chem. Abst. 119 (Imperial Chemical Industries, PLC) (1993)
160314j.
[14] M.R. Harmden, D.T. Hurst, Aust. J. Chem. 43 (1990) 55–62.
[15] A. McKillop, J.A. Tarbin, Tetrahedron 43 (1987) 1753–1758.
[16] H. Gilman, R.R. Burtner, J. Org. Chem. 55 (1987) 2903–2909.
[17] D.J. Minnick, D.A. Brent, J. Frenz, J. Chromatog 461 (1989) 177–191.