Synthesis and pharmacological assessment of derivatives of isoxazolo[4,5-d]pyrimidine

Article abstract of DOI:10.1016/j.bmc.2003.10.004

A series of new 5-alkyl and 5-arylisoxazolo[4,5-d]pyrimidinones (5a-g, 6-8) were prepared from 4-amino-3-oxo-isoxazolidine-5-carboxylic acid amide. Some of the aryl derivatives of isoxazolo[4,5-d]pyrimidine were tested pharmacologically in comparison with Diazepam. Compounds 5b-d and 7 demonstrated interesting anxiolytic activity.

Full text of DOI:10.1016/j.bmc.2003.10.004

Bioorganic & Medicinal Chemistry 12 (2004) 265–272
Synthesis and pharmacological assessment of derivatives
of isoxazolo[4,5-d]pyrimidine
.
Edwin Wagner,a,* Lilianna Becana and Elzbieta Nowakowskab
aDepartment of Drugs Technology, Wroclaw Medical University, Nankier Sq. 1,50-140 Wroclaw, Poland
bDepartment of Pharmacology, Poznan University of Medicine, Fredry Str. 10, 61-701 Poznan, Poland
Received 18 July 2003; accepted 1 October 2003
Abstract—A series of new 5-alkyl and 5-arylisoxazolo[4,5-d]pyrimidinones (5a–g, 6–8) were prepared from 4-amino-3-oxo-isoxa-
zolidine-5-carboxylic acid amide. Some of the aryl derivatives of isoxazolo[4,5-d]pyrimidine were tested pharmacologically in com-
parison with Diazepam. Compounds 5b–d and 7 demonstrated interesting anxiolytic activity.y
# 2003 Elsevier Ltd. All rights reserved.
1. Introduction
of new benzodiazepine receptor (BzR) ligands based on
reliable pharmacophore models has been proposed on
the basis of binding data displayed at multiple BzR
subtypes by various structural families of BzR ligands,
but it is also good to take into consideration the struc-
ture of the endogenous ligand of this receptor. Therefore,
in our study we have taken the structure of inosine
(hypoxantine) and that of THIP (4,5,6,7-tetra-
hydroisoxazolo[5,4-e]pyridin-3(2H)-one as the basis for
the synthesis of a new non-benzodiazepine series of
compounds with potential anxiolitic activity. Inosine
(hypoxantine) is an endogenous ligand for the brain
benzodiazepine receptor.13,14 The THIP mentioned above
is a highly selective ligand of the GABAA receptor.15
Benzodiazepine receptor (BzR) ligands are structurally
diverse compounds that bind to the GABAA/BzR com-
plex. They include b-carbolines1 imidazobenzodiaze-
pines,2
imidazopyridines,3
pyrazoloquinolines,4
pyridiindoles,5 triazolopyridazines such as CL218872,6
as well as some others.7À9 Molecular biological studies
have demonstrated that several receptor subunits
(a1–a6, b1–b3, g1–g3, d) combine to form the GABAA
receptor complex in the mammalian central nervous
system.10,11 One from two central benzodiazepine
receptors are probably formed by combinations of the
subunits a1b2g2, show many of the pharmacological
effects associated with the type I BzR, and a mixture of
the subunits a2–a3 and a5b2g2 characterize the biological
effects of the so-called type II BzR.12 Of the chemical
classes which have binding sites on this macromolecular
ionophore, the benzodiazepines are the most widely
studied and used. However, these compounds often
produce undesirable side effects, including sedation,
physical dependence, amnesia, and ethanol potentation.
When the benzodiazepines are used only as antianxiety
drugs, the sedative effect makes normal life most diffi-
cult for patients. For this reason, many laboratories
have been searching for anxiolytics that are devoid of or
exhibit markedly reduced side effects. A rational design
We have concentrated on the synthesis of derivatives of
isoxazolo[4,5-d]pyrimidine — a structure created from a
compilation of the structure of THIP and hypoxantine,
where the nitrogen atom in position 7 (hypoxantine) is
replaced by an oxygen atom. Inserting appropriately
substituted aryl groups in position 5 should also have
profitable influence on the activity of these derivatives
and to bring the structure closer to one more selective
for the classical type I BzR — CL-218872.6
Keywords: Isoxazolo[4,5-d]pyrimidinones; Anxiolytic compounds;
Benzodiazepine receptors; Anxiety.
* Corresponding author. E-mail: wagner@bf.uni.wroc.pl
E. Wagner, L. Becan, E. Nowakowska, Pol. P. 328352, 1998. E.
Wagner, L. Becan, E. Nowakowska, Pol. P. 328382.
y
0968-0896/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2003.10.004

266
E. Wagner et al. / Bioorg. Med. Chem. 12 (2004) 265–272
compound 6 which, after cyclizing, gave a new 5-(4-
chloro-phenyl)-3a,7a-dihydro-6H-isoxazolo[4,5-d]pyr-
imidine-3,7-dione (7). Compounds 5d and 7 were sub-
jected to oxidation with KMnO4 in acetone and the
result was identical for both compounds: 5-(4-chloro-
phenyl)-6H-isoxazolo[4,5-d]-pyrimidine-3,7-dione (8).
The oxidation also confirmed the structure of the new
compounds. Compounds 5b–e and 7 were subjected to
pharmacological study.
2.2. Influence of the compounds on locomotor activity
The aim of the study was to find out if compounds of dif-
fering chemical structure, that is, 5b–e and 7, show anxio-
lytic activity comparable with the effects of diazepam.
Each of the supplied compounds was investigated in a
locomotor activity test in doses of 0.1–10 mg/kg bw, 30
min after ip injection. The aim of the experiments was
to estimate the highest dose not causing sedation.
These modifications have been made for the purpose of
obtaining compounds with the expected anxiolytic
activity, but with possibly decreased side effects.
It was found that the following doses of the compounds
were free from sedative action: compounds 5b, 5c, 7 — 1
mg/kg ; compounds 5d, 5e — 0.1 mg/kg.
The classical benzodiazepine diazepam tested at the
dose level of 2.5 mg/kg showed no sedative activity.
2. Results and discussion
2.1. Synthesis
Table 1. The influence of the compounds 5b–e and 7 on locomotor
activity in rats, in comparison with diazepam
Of the four structural isomers of isoxazolopyrimidines,
only the isomer 4,5-d has not been thoroughly investi-
gated. There have been only three16À18 foreign papers
published concerning the synthesis of derivatives of this
isomer till now. We obtained the derivatives 5a–g, 7, 8
of this isomer from aminoamide 4 and the correspond-
ing aldehydes (Scheme 1). The aminoamid 4 was pre-
pared earlier by Stammer19 but in low total yield. In
contrast to previous work,19, we obtained compound 4
using another method. We used a more active dimethyl
ester 2 which was obtained from reaction between treo-
b-hydroxy-dl-aspartic acid dimethyl ester20 and PCl5.
Compound 2 (obtained in a very good yield) has the
identical mp and IR as the one obtained by Antolini21
from the appropriate aziridine. The intermediate 2 was
converted into the hydroxamic acid 3 (by reaction with
anhydrous hydroxylamine in dry methanol) and was
cyclized in the presence of lithium metoxide. Compound
3 is identical to the compound obtained by Stammer.19
It is a white amorphous solid which produces positive
ninhydrin and ferric chloride tests and a positive nitro-
prusside test, which is specific to the isoxazole ring.22
We converted compound 3 into aminoamide 4 with
perfect efficiency by applying the Mattingly and Miller
method of reduction of hydroxamic acids to amides by
means of titanium trichloride.23 Compound 4 was the
same as the compound obtained by Stammer,19 but was
prepared without the two low-productivity stages
(methanolysis and aminolysis).
Substances tested
n
Doses
(mg/kg)
Locomotor
activity ÆSEM
Control 0.5% CMC
Control H2O inj.
Compound 5b
Compound 5c
Compound 5d
Compound 5e
Compound 5e
Compound 7
8
8
8
8
8
8
8
8
8
0.4 mL
0.4 mL
92.0Æ10.3
90.2Æ9.5
1.0 mg/kg
1.9 mg/kg
0.1 mg/kg
0.1 mg/kg
1.0 mg/kg
1.0 mg/kg
2.5 mg/kg
88.0Æ8.6
101.0Æ10.1
92.9Æ9.4
103.0Æ12.3
114.0Æ14.1
92.0Æ8.2
DIAZEPAM
87.0Æ6.9
Diazepam and the tested compounds were administered 30 min before
the test. n, number of animals in the group.
Table 2. Investigation of anxiolytic activity of compounds 5b–e and 7
in comparison with placebo, in the two compartment exploratory test
in rats
Substance (mg/kg)
n
WSE Æ
BWT Æ
BSE Æ
SEM
SEM
SEM
Control 0.5% CMC
Control H2O inj.
8
8
8
8
8
8
8
8
8
1.2Æ0.8
1.1Æ0.8
2.9Æ0.1
2.1Æ0.1
6.0Æ0.5
6.7Æ0.5
Compound 5b 1.0 mg/kg
Compound 5c 1.0 mg/kg
Compound 5d 0.1 mg/kg
Compound 5e 0.1 mg/kg
Compound 5e 1.0 mg/kg
Compound 7 1.0 mg/kg
DIAZEPAM 2.5 mg/kg
18.5Æ2.9*
13.4Æ2.3*
4.9Æ1.1*
1.3Æ0.7
6.2Æ1.1* 11.2Æ1.8*
3.9Æ1.4
4.1Æ1.5
1.6Æ0.9
1.5Æ0.8
4.8Æ0.7*
9.6Æ2.8
7.4Æ2.9
4.1Æ2.4
3.9Æ1.9
9.7Æ0.3*
9.9Æ0.9*
1.5Æ0.4
9.9Æ1.1*
19.5Æ2.4* 12.8Æ3.1*
Compound 4 was subjected to reaction with the suitable
aldehydes. We obtained a number of new derivatives of
isoxazolo[4,5-d]pyrimidine 5a–g. The result of reaction
of compound 4 with p-chlorobenzoyl chloride was
WSE, white square entrance; BWT, black–white transition; BSE,
black square entrance. The experiments were carried out 30 min after
administration of the compound. *Significant differences against con-
trol group (P<0.05). n, number of rats in the group.

E. Wagner et al. / Bioorg. Med. Chem. 12 (2004) 265–272
267
As shown in Table 1, the compounds investigated did
not change the locomotor activity of the rats. The
proper experiments were carried out after a period of 3
days, during which the rats became accustomed to the
apparatus through daily exposure to it.
The control animals, receiving only H2O, CMC or H2O
with Tween injections, showed a marked preference for
the dark area (BSE=6.9Æ0.5 vs WSE<2.0 and
BWT=2.9Æ0.1).
Diazepam administered in the dose of 2.5 mg/kg bw
showed a significant rise of all parameters: WSE,
BWT and BSE (Table 2), 30 min after administration
(the peak value). The pyrimidine compounds were
tested for anxiolytic activity in doses free from any
sedative effect (see Table 1). It was found that 1 mg/
kg of compound 5b, 1 mg/kg of compound 5c, 0.1
mg/kg of compound 5d, and 1 mg/kg of compound 7
showed anxiolytic activity.
2.3. Anxiolytic activity of the compounds in the ‘two-
compartment exploratory test’
An event was recorded when a rat entered a square with
all four paws (white square entrance WSE or black
square entrance BSE). If a rat crossed from the white
square area to the black one, or from the black to the
white, a black–white transmission (BWT) was recorded.
Figure 1. Investigation of anxiolytic activity of the tested compounds, in comparison with diazepam. The experiments were carried out 30 min after
compound administration. WSE, white square entrance, BSE, black square entrance, BWT, black–white transition. *Significant differences against
control group (p<0.05).
Table 3. The effects of Ro-15-1788 on anxiolytic action of diazepam and substances 5b and 7 in two compartment exploratory test
No
Group
n
WSE XÆSEM
BWT XÆSEM
BSE XÆSEM
1.
2.
Control
6
6
2.25Æ0.25
2.30Æ0.24
1.25ÆÀ0.25
1.00Æ0.00
7.75Æ0.75
7.25Æ1.10
Ro-15-1788 (10 mg/kg ip)
30 min before the test
Diazepam (2.5 mg/kg ip)
30 min before the test
Ro-15-1788+Diazepam
30 min before the test
Substance 5b (1 mg/kg ip)
30 min before the test
Substance 5b (1 mg/kg)
+Ro-15-1788 (10 mg/kg)
30 min before the test
Substance 7 (1 mg/kg ip)
30 min before the test
Substance 7 (1 mg/kg)
+Ro-15-1788 (10 mg/kg)
30 min before the test
3.
4.
5.
6.
6
6
6
6
24.40Æ1.36*
5.60Æ0.68*,**
18.17Æ0.60*
13.83Æ1.57*,{
8.60Æ0.60*
1.40Æ0.24*,y
5.40ÆÀ0.50*
3.33Æ0.61*,y
10.60Æ0.68*
12.00Æ1.45*
10.17Æ0.54
7.83Æ0.65y
7.
8.
6
6
8.83Æ0.60*
3.60Æ0.40*
3.50Æ0.56*
7.17Æ0.70
5.83Æ0.70
14.00Æ2.05*,{
Statistically significant against the control group (1); p<0.05. **Statistically significant against diazepam group (3); p<0.05. yStatistically significant
against group (5); p<0.05. {Statistically significant against group (7); p<0.05. Ro-15-1788 was administered in the intraperitoneal dose of 10 mg/kg
30 min before the anxiolytic test.

268
E. Wagner et al. / Bioorg. Med. Chem. 12 (2004) 265–272
After compounds 5b and 7, a rise in all parameters was
observed, while after compounds 5c and 5d only a rise
in WSE occurred (Table 2, Fig. 1).
2.4. Assay in association with Flumazenil
An event was recorded whenever a rat entered a square
with all four paws (white square entrance WSE or black
square entrance BSE). If a rat crossed from a white square
to a black one or from a black to a white, a BWT was
Compound 5e showed no anxiolytic activity at doses of
0.1 and 1.0 mg/kg bw.
Scheme 1.

E. Wagner et al. / Bioorg. Med. Chem. 12 (2004) 265–272
269
recorded. The control animals, which received only
H2O, CMC or H2O with Tween injections, showed a
marked preference for the dark area (BSE=7.75Æ0.75
vs WSE<2.25 and BWT 1.25Æ0.25).
and coupling constants in Hz. Mass spectra were recor-
ded on a Finnigan MAT 95. Two solvent systems,
butanol–acetic acid–water (3:1:1, BAW) and CH2Cl2–
metanol (7:1) were used. Elemental analyses were per-
formed on a Carbo Erba NA 1500 analyzer in our
faculty laboratories.
Diazepam administered in the doses of 2.5 mg/kg sig-
nificantly increased the value of all parameters, that is
WSE, BWT and BSE (Table 3) 30 min after its admin-
istration (peak value).
4.1.1. Dimethyl ꢀ-chloroaspartate hydrochloride (2). To
the stirred suspension of 213.5 g (1 mol) threo-b-
hydroxy-dl-aspartic acid dimethyl ester hydrochloride
in 1 L of absolute CH2Cl2 was added 228.5 g (1.1 mol)
PCl5 portion-wise over a period of 1 h. The reaction
mixture was stirred at room temperature for 3 h and left
Compounds 5b and 7 produced statistically significant
increases in the WSE and BWT values, which confirms
the anxiolytic potential of both compounds.
standing at room temperature for 12h. The excess PCl
5
Ro-15-1788 did not exhibit primary anxiolytic potential,
as the WSE, BWT and BSE parameters were not differ-
ent from there of the control group. Joint administration
of Ro-15-1788 and diazepam resulted in antagonizing
the anxiolytic effects of diazepam (decrease in WSE and
BWT values), and similar results could be observed by
joint administration of Ro-15-1788 and 5b. However, no
R-15-1788-related antagonism could be observed for
joint administration of flumazenil and compound 7.
was filtered and the solution was evaporated to ca. 400
mL. The remainder was cooled to 0 ꢀC, and the pre-
cipitate was filtered and washed successively with abso-
lute CH2Cl2 and dry ether. The solid was dried over
CaCl2 at room temperature in vacuo, giving 187.9 g
(81%) of crude product (2). A sample was recrystallized
three times from methanol–ether: mp 89–90 ꢀC (lit.11 mp
88–90 ꢀC).
4.1.2. 4-Amino-3-oxo-isoxazolidine-5-carboxylic acid hy-
droxyamide (3). To a solution of sodium methoxide,
prepared from 27.6 g (1.2 mol) sodium in 650 mL
anhydrous methanol, was added 83.4 g (1.2mol)
hydroxylamine hydrochloride and stirred at room tem-
perature for 90 min. The mixture was cooled to 0 ꢀC, the
sodium chloride was filtered and a solution of 116 g (0.5
mol) dimethyl b-chloroaspartate hydrochloride (2) in
350 mL absolute methanol was added to the filtrate.
This solution was cooled to À10 ꢀC and a solution of
lithium methoxide (prepared from 600 mL anhydrous
methanol and 12g LiH) was added. The mixture was
cooled to À5 ꢀC in a refrigerator over night, the reac-
tion mixture was acidified (pH 4.5–5) with con-
centrated HCl at 5 ꢀC. The suspension was evaporated
to ca. 300 mL in vacuo. The remainder was cooled to
5 ꢀC, and the precipitate was filtered and washed with
absolute ethanol and ether. The solid was dissolved in
a minimum amount of water and the solution stirred
into 3 L of absolute EtOH. The precipitate was filtered
and dried over CaCl2 at room temperature in vacuo,
giving 155.4 g (96.4%) of amorphous powder. The IR
and TLC were identical with 4-amino-5-carbohydrox-
amido-3-isoxazolidone.10
3. Conclusions
The results of the biological study confirm our assump-
tion that the derivatives of isoxazolo[4,5-d]pyrimidine,
which are structurally similar to hypoxantine can have
anxiolytic activity comparable to diazepam. All doses of
the compounds investigated (including diazepam) were
lower than those causing the sedative effect. The results
of our study, depicted in Figure 1, show that compounds
5b, 5c and 7, even at 2.5 times smaller doses, have
approximately the same anxiolytic activity as diazepam,
whereas compound 5d has weaker anxiolytic activity, but
at a 25 times smaller dose. Only compound 5e lacks this
activity. Derivatives of isoxazolo[4,5-d]-pyrimidine may
constitute a new chemical tool for further research work
on the efficiency of anxiolytic activity.
If Ro 15-1788 antagonizes the anxiolytic effects of clas-
sic benzodiazepines (diazepam), which would confirm
the well-established hypothesis relating anxiolytic effects
of benzodiazepines to their action on benzodiazepine
receptors in the brain, it may be postulated that the
mechanism of anxiolytic action of 5b may be similar to
that of classic benzodiazepines.
4.1.3. 4-Amino-3-oxo-isoxazolidine-5-carboxylic acid
amide (4). To a stirred solution of 16.1 g (0.1 mol) of (3)
in 60 mL of water was added drop-wise a 20% aqueous
solution of TiCl3.14 The mixture was stirred under N2 and
the pH maintained at 7 with 10% LiOH. After comple-
tion addition of the TiCl3, the blue color persisted and a
light yellow suspension of Ti(OH)4 formed. The solid
Ti(OH)4 was removed by centrifugation and the solution
was evaporated in vacuo to ca. 30 mL and diluted with
100 mL of absolute ethanol. After cooling in a refrig-
erator for the night, the precipitate was filtered, washed
with 95% EtOH, and dried, giving 12.15 g of crude
product (4). Recrystallization from NH4OH–AcOH/
EtOH gave an analytically pure compound (4): 11.3 g
(78%); mp 203–4 ꢀC (lit. ca. 200 ꢀC10).
4. Experimental
4.1. General
Melting points were determined with a Kofler hot-stage
apparatus and are uncorrected. The progress of the
reaction and purity of compounds were monitored by
TLC analytical silica gel plates (Merck F254). IR spectra
were recorded on a Specord M 80 spectrometer for KBr
discs. 1HNMR spectra were recorded with a Bruker
Avance DRX-300 instrument; chemical shifts are repor-
ted in ppm downfield from the internal tetramethysilane

270
E. Wagner et al. / Bioorg. Med. Chem. 12 (2004) 265–272
4.2. General procedure for the preparation of derivatives
of tetrahydro-isoxazolo-[4,5-d]pyrimidine-3,7-dione (5a–g)
OCH3), 4.07 (d, J=4 Hz, 1H, 3a-CH), 4.37 (d, J=4 Hz,
1H, 7a-CH), 5.69 (s, 1H, 5-CH), 7.26 (s, 1H, 6-NH),
7.36–7.39 (d, J=8.5 Hz, 2H, 30/50-ArH), 7.76–7.80 (d,
J=8.5 Hz, 2H, 20/60-ArH), 8.25 (s, 1H, 2-NH). Anal.
calcd (C12H13N3O4): C, 54.75; H, 4.98; N, 15.96.
Found: C, 54.61; H, 5.06; N, 16.12.
To the suspension of compound 4 (4.35 g, 0.03 mol) in
100 mL of dry methanol was added 0.033 mol of the
corresponding aldehyde. The reaction mixture was stir-
red under reflux for 4 h. The mixture was concentrated
to ca. 30 mL. After cooling, the remainder was placed in
a refrigerator for 6 h, and the precipitate was filtered
and washed with methanol. Solids was purified by
recrystallization.
4.2.6. 5-(4-Nitro-phenyl)-tetrahydro-isoxaxolo[4,5-d]pyri-
midine-3,7-dione (5f). Ivory crystals obtained from 4 and
p-nitrobenzaldehyde and recrystallized from ethanol;
yield 72%; mp 206–207 ꢀC; 1H NMR (DMSO-d6) d 3.60
(br s, 1H, 4-NH), 4.19 (d, J=4 Hz, 1H, 3a-CH), 4.45 (d,
J=4 Hz, 1H, 7a-CH), 5.90 (s, 1H, 5-CH), 7.71 (s, 1H, 6-
NH), 8.07 (d, J=9 Hz, 20/60-ArH), 8.27–8.30 (m, 3H, 30/
50-ArH and 2-NH). Anal. calcd (C11H10N4O5): C, 47.49;
H, 3.62; N, 20.14. Found: C, 47.65; H, 3.51; N, 19.97.
4.2.1. 5-Ethyl-tetrahydro-isoxazolo[4,5-d]pyrimidine-3,7-
dione (5a). From 4 and propionyl aldehyde and recrys-
tallized from ethanol; yield 58% of white crystals; mp
ꢀ
1
189–190 C; H NMR (300 MHz, DMSO-d6) d 1.25 (t,
3H, CH3), 3.15 (m, 2H, CH2), 3.16 (br s, 1H, 4-NH),
3.70 (d, J=4 Hz, 1H, 3a-CH), 4.07 (d, J=4 Hz, 1H, 7a-
CH), 5.57 (s, 1H, 5-CH), 7.28 (s, 1H, 6-NH), 8.32 (s,
1H, 2-NH). EJ MS m/z 185 (M+) Anal. calcd
(C7H11N3O3): C, 45.40; H, 5.99; N, 22.69. Found: C,
45.55; H, 6.05; N, 22.75.
4.2.7. 5-(2-Hydroxy-phenyl)-tetrahydro-isoxazolo[4,5-
d]pyrimidine-3,7-dione (5g). Colorless crystals obtained
from 4 and salicylaldehyde and recrystallized from
methanol; yield 61%; mp 165–166 ꢀC; 1H NMR
(DMSO-d6) d 4.03 (br s, 1H, 4-NH), 4.24 (d, J=4 Hz,
1H, 3a-CH), 4.53 (d, J=4 Hz, 1H, 7a-CH), 5.96 (s, 1H,
5-CH), 7.29–7.40 (m, 3H, 40/50/60-ArH), 7.47 (s, 1H,
6-NH), 7.58–7.61 (d, 1H, J=6 Hz, 60-ArH), 7.94 (br s,
1H, OH), 8.39 (s, 1H, 2-NH). EJ MS m/z 249 (M+).
Anal. calcd(C11H11N3O4): C, 53.01;H, 4.45; N, 16.86.
Found: C, 53.18; H, 4.48: N, 16.72.
4.2.2. 5-Phenyl-tetrahydro-isoxazolo[4,5-d]pyrimidine-
3,7-dione (5b). Colorless needles obtained from 4 and
benzaldehyde and recrystallized from methanol; yield
79%; mp 168–169 ꢀC; 1H NMR (DMSO-d6) d 3.18 (br s,
1H, 4-NH), 4.08 (d, J+4 Hz, 1H, 3a-CH), 4.38 (d, J=4
Hz, 1H, 7a-CH), 5.63 (s, 1H, 5-CH), 7.34 (s, 1H, 6-NH),
7.44–7.51 (m, 3H, 30/40/50-PhH), 7.81–7.84 (m, 2H, 20/
60-PhH), 8.22 (s, 1H, 2-NH). EJ MS m/z 233 (M+).
Anal. calcd (C11H11N3O3): C, 56.65; H, 4.75; N, 18.02.
Found: C, 56.85; H, 4.90; N, 17.90.
4.2.8. 4-(4-Chloro-benzoylamino)-3-oxo-isoxazolidine-5-
carboxylic acid amide (6). To a suspension of 5.8 g
(0.04 mol) compound 4 in 80 mL anhydrous DMF and
4.5 g (0.045 mol) anhydrous TEA was added 7.87 g
(0.045 mol) p-chlorobenzoyl chloride at À5 to 0 ꢀC por-
tion-wise over a period of 30 min. The mixture was
stirred at 0 ꢀC for 2h and then at room temperature for
4.2.3. 5-(4-Fluoro-phenyl)-tetrahydro-isoxazolo[4,5-d]pyr-
imidine-3,7-dione (5c). Colorless needles obtained from
4 and p-fluorobenzaldehyde and recrystallized from
methanol; yield 83%; mp 165–166 ꢀC; 1HNMR
(DMSO-d6) d 3.75 (br s, 1H, 4-NH), 4.08 (d, J=4 Hz,
1H, 3a-CH), 4.39 (d, J=4 Hz, 1H, 7a-CH), 5.72(s, 1H,
5-CH), 7.41 (s, 1H, 6-NH), 7.52–7.54 (d, J=7 Hz, 20/
60-ArH), 7.87–7.92(d, J=7 Hz, 30/50-ArH), 8.21 (s, 1H,
2-NH). EJ MS m/z 251 (M+) Anal. calcd
(C11H10N3O3F): C, 52.59; H, 4.02; N, 16.73. Found: C,
52.73; H, 4.14; N, 16.51.
.
12h. The precipitate of TEA HCl was filtered and
washed with anhydrous DMF. The filtrate was evapo-
rated in vacuo to dryness. The residue was crushed with
30 mL of water and mixed with a solution of NaHCO3
(5%). The precipitate was filtered and washed water and
dried, giving 7.1 g of crude product (6). The product
was recrystallized from methanol giving a pure (TLC)
compound (6), 6.24 g (55% yield); mp 215–216 ꢀC; H
1
NMR (DMSO-d6) d 4.40 (d, J=4 Hz, 1H, 3a-CH), 4.76
(d, J=4 Hz, 1H, 7a-CH), 7.19 (br s, 1H, NH), 7.38 (br
s, 1H, NH), 7.43 (s, 1H, NH), 7.54 (d, J=9 Hz, 2H, 20/
60-ArH), 7.87(d, J=9 Hz, 2H, 30/50-ArH), 8.10 (s, 1H,
2-NH). EJ MS m/z 283 (M+). Anal. calcd
(C11H10N3O4Cl): C, 46.57; H, 3.55; N, 14.81. Found: C,
46.71; H, 3.61; N, 14.92.
4.2.4. 5 - (4 - Chloro - phenyl) - tetrahydro - isoxazolo[4,5 -
d]pyrimidine-3,7-dione (5d). Colorless needles obtained
from 4 and p-chlorobenzaldehyde and recrystallized
from ethanol; yield 92%; mp 176–177 ꢀC; 1H NMR
(DMSO-d6) d 3.81 (br s, 1H, 4-NH), 4.12(d, J=4 Hz,
1H, 3a-CH), 4.45 (d, J=4 Hz, 1H, 7a-CH), 5.80 (s, 1H,
5-CH), 7.40 (s, 1H, 6-NH), 7.51–7.53 (d, J=8 Hz, 2H,
20/60-ArH), 7.81–8.09 (d, J=8 Hz, 2H, 30/50-ArH), 8.30
(s, 1H, 2-NH). EJ MS m/z 267 (M+) Anal. calcd
(C11H10N3O3Cl): C, 49.35; H, 3.76; N, 15.70. Found: C,
49.56; H, 3.88; N, 15.63.
4.2.9. 5 - (4 - Chloro - phenyl) - 3a,7a - dihydro - 6H - isoxa-
zolo[4,5-d]pyrimidine-3,7-dione (7). To a suspension of
4.2g (0.014 mol) 6 in 100 mL anhydrous toluene was
added 20 mL of thionyl chloride, and refluxed for 8 h.
After cooling, the solvent and excess thionyl chloride
were distilled off under reduced pressure to dryness. The
residue was mixed with 50 mL of water and neutralized
with NaHCO3. The precipitate was filtered, washed with
water and dried. Two recrystallizations of the product
from methanol gave a pure compound (7), 2.2 g (59%
4.2.5. 5-(4-Methoxy-phenyl)-tetrahydro-isoxazolo[4,5-
d]pyrimidine-3,7-dione(5e). Colorless needles obtained
from 4 and p-methoxybenzaldehyde and recrystallized
from ethanol; yield 88%; mp. 179–180 ꢀC; 1H NMR
(DMSO-d6) d 3.76 (br s, 1H, 4-NH), 3.86 (s, 3H,
yield); mp 203–204 ꢀC; H NMR (300 MHz, DMSO-d6)
1

E. Wagner et al. / Bioorg. Med. Chem. 12 (2004) 265–272
271
d 4.47 (d, J=4 Hz, 3a-CH), 4.89 (d, J=4 Hz, 1H,
7a-CH), 7.56 (d, J=9 Hz, 2H, 20/60-ArH), 7.92(d, J=9
Hz, 2H, 30/50-ArH), 9.19 (s, 1H, 2-NH), 11.45 (br s, 1H,
NH). EJ MS m/z 265 (M+). Anal. calcd
(C11H8N3O3Cl): C, 49.73; H, 3.03; N, 15.81. Found: C,
49.91; H, 3.05; N, 15.92.
Procedure: The apparatus employed to test ‘approach-
avoidance behaviour’ was a conventional open field
(100ꢁ100 cm), the floor consisting of a sheet of white
plastic painted with a black grid dividing the field into 25
(5ꢁ5) equal squares. This surface was divided into two
compartments, one consisting of a square area (40ꢁ40
cm) (four squares) in one corner of the open field with
all the surfaces blackened and a roof fitted 35 cm from
the floor, to prevent light entering from above. The
second compartment was the remaining part of the open
field (21 squares), uniformly lit by a fluorescent lamp. No
partitions were provided between the two compartments,
the transition lines consisting of the internal half peri-
meter of the black area in the corner. At the beginning of
the test, all rats were placed gently in the same peripheral,
lighted white square near to the black compartment. The
number of transitions between the two-compartments
(BWT), square entries in the black compartment (BSE),
and square entries in the white compartment (WSE) were
recorded for 5 min by an observer sitting quietly 2m
away, unaware of the treatment.
4.2.10. 5-(4-Chloro-phenyl)- 6H-isoxazolo[4,5-d]pyrimi-
dine-3,7-dione (8). To a solution of compound 5d or 7
(0.001 mol) in acetone (250 mL) was added 2 g potas-
sium permanganate to compound 5d and 1 g to
compound 7 and the mixture was refluxed in a water
bath. The pink color of potassium permanganate slowly
disappeared. Addition of potassium permanganate
(about 0.3 g each time) and refluxing continued until the
pink color persisted. The hot suspension was filtered
and the excess acetone was distilled off to dryness. The
residue was mixed with 20 mL of water and excess
potassium permanganate destroyed with aq sodium
sulphite. The white precipitate was filtered washed with
water and dried. Two recrystallizations of the products
from ethanol gave identical pure colorless compound 8
crystals, mp 215–216 ꢀC, from compound (5d) with a
yield of 55%, and from compound 7 with a yield of
Whenever an animal crossed the transition line with all
four paws, an event was recorded. On removal of the rat
the floor was thoroughly cleaned. Each rat was assessed
in the novelty test only once, always between 10.00 a.m.
and 2.00 p.m.
1
62%. H NMR (300 MHz, DMSO-d6) d 7.55 (d, J=9
Hz, 2H, 20/60-ArH), 7.96 (d, 9 Hz, 2H, 30/50-ArH), 9.20
(br s, 1H, 2-NH), 11.35 (br s, 1H, NH). EJ MS m/z 263
(M+). Anal. calcd (C11H6N3O3Cl): C, 50.11; H, 2.29; N,
15.94. Found: C, 50.21; H, 2.36; N, 16.06.
Locomotor activity was measured in the treated and
control groups using eight 20.5ꢁ28ꢁ21 cm wire-grid
cages each with two horizontal infrared photocell
beams along the long axis 3 cm above the floor. Pho-
tocell interruptions were recorded by electro-
mechanical counters in an adjacent room. After 30
min of habituation to the novel cage, rats were injec-
ted with the moclobemide and the photocell activity
was recorded at 10-min intervals for 1 h. This test
provided an index of basal locomotor activity of ani-
mals in a familiar environment, necessary to indicate
the presence of the central stimulant effects of the
compounds used in the novelty test.
4.3. General methods for the biological activity evaluation
4.3.1. Material. Animals: male Wistar rats, 180–200 g,
purchased from a breeder (licencsed by the Ministry of
Agriculture, Warsaw, Poland), were used in this study.
The animals were housed under standard laboratory
conditions with a 12h light/dark cycle, light on at 6
a.m., in a temperature-controlled room at 21Æ2 ꢀC,
humidity 60% with free access to granulated standard
food and tap water. The rats were kept four per cage
(30ꢁ30ꢁ20 cm). Each experimental and control group
consisted of eight animals.
4.3.3. Statistics. The statistical significance of the results
of the anxiolytic trials was analysed according to the
two-tailed Student t-test.26 Statistical analysis for loco-
motor activity test was done by Dunnett’s t-test.
The study was approved by the Governmental Com-
mittee for Scientific Research, grant No. 8.364/98.
Drugs: Sodium carboxymethyl cellulose pure B.P.C. —
Koch Light Laboratories Ltd London, UK (CMC).
4.4. The effect of Ro-15-1788 on the anxiolytic efficacy
of diazepam and compounds 5b and 7 in the two-com-
partment exploratory test
1. H2O injection
2. compound 5b, 5c, 5d, 5e, 7.
4.4.1. Materials and methods. Animals: male Wistar
rats, 180–200 g, purchased from a breeder (licencsed of
the Ministry of Agriculture, Warsaw, Poland, were used
in this study.
Compounds 5b, 5c and 5e were administered as water
solutions intraperitoneally (ip) 30 min before the test.
Compounds 5d and 7 in a CMC suspension were also
injected intraperitoneally (ip) 30 min before the tests
(peak value time).
The animals were housed under standard laboratory
conditions with a 12-h light/dark cycle, light on at 6
a.m., in a temperature controlled room at 21Æ2 ꢀC,
humidity 60%, with free access to granulated standard
food and tap water. The rats were kept four per cage
(30ꢁ30ꢁ20 cm). Each experimental and control group
consisted of six animals.
4.3.2. Method used. The anxiolytic effects were deter-
mined according to the ‘Two-compartment exploratory
test’ of Crawley,24 in the modification of Merlo-Pich
and Samanin.25

272
E. Wagner et al. / Bioorg. Med. Chem. 12 (2004) 265–272
The study was approved by the Governmental Com-
mittee for Scientific Research, grant No. 8.364/98.
domized: (I) the distribution of the experimental data
was normal; (II) the variances were homogenous. Sta-
tistical analysis for the anxiolytic test was done by the
nonparametric Kruskal–Wallis’s test for unpaired data
and by the Friedman test for paired data. Statistical
significance was then tested by the post-hoc Man test.
4.5. Compounds
1. 3,7-Dioxo-5-phenyl-2,3,4,5,6,7,3a,7a-octahy-
droisoxazol[4,5-d]pyrimidine (5b) was adminis-
tered at 1 mg/kg ip 30 min before the anxiolytic
test.
References and notes
2. 3,7-Dioxo-5-(p-chlorophenyl)-2,3,6,7,3a,7a-hexa-
hydroisoxazol[4,5-d]pyrimidine (7) was adminis-
tered at 1 mg/kg ip 30 min before the anxiolytic
test.
3. Diazepam was administered at 2.5 mg/kg 30 min
before the test.
1. Lippke, K. P.; Schunack, W. G.; Wennig, W.; Muller,
W. E. J. Med. Chem. 1983, 26, 499.
2. Liu, R.; Hu, R. J.; Zhang, P.; Skolnick, P.; Cook, J. M. J.
Med. Chem. 1996, 39, 1928.
3. Arbilla, S.; Depoortere, H.; George, P.; Langer, S. Z.
Naunyn-Schmiedebergs Arch. Pharmacol. 1985, 330, 248.
4. Yokoyama, N.; Ritter, B.; Neubert, A. D. J. Med. Chem.
1982, 25, 337.
4. Ro-15-1788 was administered ip 30 min before
the test.
5. Trudell, M. L.; Lifer, S. L.; Tan, Y. C.; Martin, M. J.;
Deng, L.; Skolnick, P.; Cook, J. M. J. Med. Chem. 1990,
33, 2412.
4.6. Anxiolytic effects
6. Gee, K. W.; Brinton, R. E.; Yamamura, H. J. Life Sci.
1983, 32, 1037.
Anxiolytic effects were determined based on the two-
compartment exploratory test of Crawley24 in the
modification of Merlo-Pich and Samanin.25
7. Tarzia, G.; Occelli, E.; Toja, E.; Barone, D.; Corsico, N.;
Gallico, L.; Luzzani, F. J. Med. Chem. 1988, 31, 1115.
8. Petke, J. D.; Im, H. K.; Im, W. B.; Blakeman, D. P.;
Pregenzer, J. F.; Jacobsen, E. J.; Hamilton, B. J.; Carter,
D. B. Mol. Pharmacol. 1992, 42, 294.
4.6.1. Procedure. The apparatus for testing the
approach-avoidance behavior was a conventional open
field (100ꢁ100 cm), the floor consisting of a sheet of
white plastic painted with a black grid dividing the field
into 25 (5ꢁ5) equal squares. This surface was divided
into two compartments, one consisting of a square area
(40ꢁ40 cm) (four squares) in one corner of the open
field with all the surfaces blackened and a roof fitted 35
cm from the floor, to prevent light entering from above.
The second compartment was the remaining part of the
open field (21 squares), uniformly lit by a fluorescent
lamp. No partitions were provided between the two
compartments, the transition lines consisting of the
internal half perimeter of the black area in the corner.
At the beginning of the test all rats were placed gently in
the same peripheral, illuminated white square, near to
the black compartment. The number of transitions
between the two compartments (BWT), square entries
in the black compartment (BSE) and square entries in
the white compartment (WSE) were recorded for 5 min
by an observer sitting quietly 2m away, unaware of the
treatment.
9. Selleri, S; Bruni, F.; Costagli, C.; Costanzo, A.; Guerrini,
G.; Ciciani, G.; Gratteri, P.; Bonaccini, C.; Malmberg, A.;
Besnard, F.; Renard, S.; Costa, B.; Martini, C. J. Med.
Chem. 2003, 46, 310.
10. Metha, A. K.; Ticku, M. K. Brain Res. Rev. 1999, 29, 196.
11. Sieghart, W. Pharm. Rev. 1995, 47, 181.
12. Chang, Y.; Wang, R.; Barot, S.; Weiss, D. S. J. Neurosci.
1996, 16, 5415.
13. Asano, T.; Spector, S. Proc. Natl. Acad. Sci. U.S.A. 1979,
76, 977.
14. Skolnick, P.; Syapin, P. J.; Paugh, B. A.; Moneada, V.;
Marangos, P. J.; Paul, S. M. Proc. Natl. Acad. Sci. U.S.A.
1979, 76, 1515.
15. Krogsgaard-Larsen, P.; Johnston, G. A.; Lodge, D.;
Curtis, D. P. Nature 1977, 268, 53.
16. Desimoni, G.; Gruenanger, P.; Vita Finzi, P. Ric. Sci.
1966, 36, 130.
17. Desimoni, G.; Gruenanger, P.; Vita Finzi, P. Tetrahedron
1967, 23, 675.
18. Laurent, S.; Barbieux-Flammang, M.; Van Haverbeke,
Y.; Flammang, R.; Wentrup, C. Bull. Soc. Chim. Belg.
1994, 103, 181.
19. Stammer, Ch.H.; Sato, M. J. Med. Chem. 1978, 21, 709.
20. Jones, C. W., III; Leyden, D. E.; Stammer, Ch.H. Can. J.
Chem. 1969, 47, 4363.
21. Antolini, L.; Bucciarelli, M.; Caselli, E.; Davoli, P.; Forni,
A.; Moretti, I.; Prati, F.; Torre, G. J. Org. Chem. 1997,
62, 8784.
22. Jones, L. R. Anal. Chem. 1956, 28, 39.
23. Mattingly, Ph.; Miller, M. J. J. Org. Chem. 1980, 45, 410.
24. Crawley, J. N. Pharmacol. Biochem. Behav. 1981, 15, 695.
25. Merlo-Pich, E.; Samanin, R. Pharmacol. Res. 1989, 21,
595.
An event was recorded whenever the animal crossed the
transition line with all four paws. On removal of the rat,
the floor was thoroughly cleaned. Each rat was assessed
in the novelty test only once daily, always between 10.00
a.m. and 2.00 p.m.
4.7. Statistics
The data are shown as mean valuesÆSEM. The alloca-
26. Ther, L., Grundlagen der exp. Arzneimittelforschung Wiss,
¨
Verlagsgesellschaft: Stuttgart.
tion of the animals to the experimental groups was ran-