New polycyclic pyrimidine derivatives with antiplatelet in vitro activity: Synthesis and pharmacological screening

Article abstract of DOI:10.1016/S0968-0896(00)00272-8

The preparation and the pharmacological screening of novel anti-aggregatory/antiphlogistic polycyclic pyrimidine derivatives are described. The compounds were developed starting from bioactive 2-aminobenzopyranopyrimidine derivatives in order to assess the importance of the benzopyrano[4,3-d]pyrimidine structure and the role of an amino basic moiety in position 2. Antiplatelet activity was assessed in vitro against ADP and arachidonic acid-induced aggregation in guinea-pig plasma. Anti-inflammatory/analgesic/antipyretic activities were studied in rat paw oedema, mouse writhing test and E coli-induced rat fever. Ulcerogenic and gastroprotective effects were also investigated in vivo on rat gastric mucosa. Among the tested compounds, the 5-substituted benzopyranopyrimidine derivatives 3d and 4d proved to be the most active antiplatelet agents as potent as acetylsalicylic acid against arachidonic acid-stimulated aggregation. Furthermore the 2-methylthio derivative 4d was endowed with greater efficacy against ADP aggregation suggesting that additional non-TXA₂ dependent mechanisms are involved in its biological activity. Orally administered at 100mgkg^-1 in rats this latter compound displayed antiphlogistic acitivity comparable to indomethacin (10 mg kg ^-1) coupled with an unusual gastroprotective effect on ethanol-induced ulcers. In conclusion, these findings indicate that the 5-pyrrolidino-2-methylthiobenzopyrano[4,3-d]pyrimidine 4d fulfils the chemical requirements to exhibit antiplatelet activity associated with gastroprotective effect. Copyright

Full text of DOI:10.1016/S0968-0896(00)00272-8

Bioorganic & Medicinal Chemistry 9 (2001) 629±636
New Polycyclic Pyrimidine Derivatives with Antiplatelet In Vitro
Activity: Synthesis and Pharmacological Screening
Olga Bruno,a Silvia Schenone,a Angelo Ranise,a Francesco Bondavalli,a
Elisabetta Barocelli,b,* Vigilio Ballabeni,b Milena Chiavarini,b Simona Bertoni,b
Massimiliano Tognolinib and Mariannina Impicciatoreb
a
Á
Dipartimento di Scienze Farmaceutiche, Universita degli Studi, Viale Benedetto XV, 3-16132Genova, Italy
b
Á
Istituto di Farmacologia e Farmacognosia, Universita degli Studi, Campus Parco Area delle Scienze 27/A, 43100 Parma, Italy
Received 4 August 2000; accepted 12 October 2000
AbstractÐThe preparation and the pharmacological screening of novel anti-aggregatory/antiphlogistic polycyclic pyrimidine deri-
vatives are described. The compounds were developed starting from bioactive 2-aminobenzopyranopyrimidine derivatives in order
to assess the importance of the benzopyrano[4,3-d]pyrimidine structure and the role of an amino basic moiety in position 2. Anti-
platelet activity was assessed in vitro against ADP and arachidonic acid-induced aggregation in guinea-pig plasma. Anti-in¯am-
matory/analgesic/antipyretic activities were studied in rat paw oedema, mouse writhing test and E coli-induced rat fever.
Ulcerogenic and gastroprotective e€ects were also investigated in vivo on rat gastric mucosa. Among the tested compounds, the
5-substituted benzopyranopyrimidine derivatives 3d and 4d proved to be the most active antiplatelet agents as potent as acet-
ylsalicylic acid against arachidonic acid-stimulated aggregation. Furthermore the 2-methylthio derivative 4d was endowed with
greater ecacy against ADP aggregation suggesting that additional non-TXA2 dependent mechanisms are involved in its biological
activity. Orally administered at 100 mg kgÀ1 in rats this latter compound displayed antiphlogistic acitivity comparable to indo-
methacin (10 mg kg À1) coupled with an unusual gastroprotective e€ect on ethanol-induced ulcers. In conclusion, these ®ndings
indicate that the 5-pyrrolidino-2-methylthiobenzopyrano[4,3-d]pyrimidine 4d ful®ls the chemical requirements to exhibit antiplate-
let activity associated with gastroprotective e€ect. # 2001 Elsevier Science Ltd. All rights reserved.
Introduction
antiplatelet and other signi®cant activities, have been
published.6,7 In earlier experimental investigations8,9 it has
been demonstrated that several polycyclic fused molecules,
sharing the benzopyran moiety, could present a platelet
anti-aggregating activity in vitro superior or comparable
to that of acetylsalicylic acid (ASA). More recently we
have prepared and screened some 2-amino-5H-[1]ben-
zopyrano[4,3-d]pyrimidin-5-amines 1a±f10 (Fig. 1) with
the aim to examine carefully the relationships between
the presence of basic moieties and antiphlogistic/anti-
platelet activity. These compounds e€ective as analge-
sic/antipyretic agents exhibited a general inhibition of
ADP-induced platelet aggregation in vitro, with 1d and
1e resulting to be slightly more active than ASA.
Because compounds 1a±f proved to be devoid of gas-
trolesive e€ects and also capable to prevent gastric
mucosa from ethanol-induced ulcers in rats, we con-
sidered them interesting for a further investigation.
It is well known that actually thromboembolic diseases are
dramatic factors of morbidity and mortality. An e€ective
way to prevent thrombotic events has been demonstrated
to be the antiaggregant therapy with cyclo-oxygenase
inhibitor aspirin (ASA).1,2 However, a number of evi-
dence indicates that this chronic antiplatelet therapy is
frequently associated with gastro-intestinal adverse side
e€ects.3,4 Since multiple agonists can stimulate platelet
aggregation5 other antiplatelet agents acting with di€er-
ent mechanisms of action were developed in the recent
past. However, a demand still remains for innovative
antithrombotic drugs safer than those currently adopted.
For several years we are interested in synthetize new
molecules as potential therapeutic agents in cardiovascular
diseases: a number of 3,5-diphenyl-1H-pyrazole and of
(3,5-diphenyl-1H-pyrazol-1-yl)pyrimidine derivatives, with
So we planned to perform di€erent modi®cations of
compounds 1a±f in order to ascertain the importance of
the benzopyrano[4,3-d]pyrimidine structure and the role
*Corresponding author. Tel.: +39-521-905090; fax: +39-521-905091;
e-mail: barocell@unipr.it
0968-0896/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(00)00272-8

630
O. Bruno et al. / Bioorg. Med. Chem. 9 (2001) 629±636
Figure 1.
Scheme 1.
Figure 2.
of the basic moiety in position 2 of this heterocyclic
system. According to this purpose we synthesized com-
pounds 2 and 3a,b,c (Fig. 2) maintaining an amino-
substituted pyrimidine ring but fused with more simple
or isoster moieties, while in the compound 3d a phenyl
grouphas been introduced in position 5. Furthermore,
we prepared novel benzopyranopyrimidine derivatives
4a±f (Fig. 2) in which the basic groupin position 2 has
been substituted with a methylthio function.
Scheme 2. (a) NaOH 1M, Et3N, 80±90ꢀC; (b) EtOH absolute,
C2H5ONa, re¯ux; (c) TiCl4, toluene/anisole; primary/secondary amines.
Thus, in the present work we evaluated the in vitro
antiplatelet activity of these new derivatives in guinea-
pig plasma aggregated by ADP and arachidonic acid
(AA) in comparison with ASA, as well as their anti-
in¯ammatory, analgesic and antipyretic activities. Addi-
tionally, for the most active compounds, we tested e€ects
on gastric mucosa integrity in order to discuss the ®nd-
ings in comparison with the pharmacological data proven
for the corresponding 2-amino derivatives (1a±f).
Chemistry
Compounds 211 and 3a±d were prepared by condensation
of guanidine with the dimethylaminomethylene deriva-
tive of suitable ketones 5 or 6a±d obtained by well known
reactions12À15 (Scheme 1). The preparation of dimethyl-
aminomethylene ¯avanone 6d16 has been modi®ed in
order to increase the yield.
Scheme 3.
In this case the reaction proceeds through a Michael
addition at C-2 of the chromone moiety with con-
comitant opening of the pyrone ring; the intermediate 7
thus formed undergoes recyclization with subsequent
water elimination according two di€erent ways: route (a),
which is energetically favoured, gives the predominant 2-
amino-5-hydroxybenzopyrano[4,3-d]pyrimidine 12, while
route (b) leads to the by-product 2-amino-5-(2-
hydroxybenzoyl)pyrimidine 13. Ghosh and co-workers
claimed that all the nucleophiles behave similarly with
3-formylchromone and their observations have also
been supported by the subsequent study of Pene and
Hubert-Habart.19 Particularly, they demonstrated that
in the presence of an alkali or a metal alkoxide (required
to get the free bases) the yield of the by-product increa-
ses or, on the contrary, decreases when the formyl
Compounds 4a±f have been prepared following the
previous synthetic method for compounds 1a±f with the
proper modi®cations (Scheme 2).
3-Formylchromone 8, which was obtained from 2-hy-
droxyacetophenone 7 by a Vilsmeier reaction,17 has
been condensed with S-methylisothiourea sulfate to give
the intermediate 5-hydroxy-2-methylthio-5H-[1]benzo-
pyrano[4,3-d]pyrimidine 9, probably according to the
reaction pathway suggested by Ghosh in the case of the
condensation of 8 with guanidine carbonate18 (Scheme 3).

O. Bruno et al. / Bioorg. Med. Chem. 9 (2001) 629±636
631
groupof 8 is opportunely masked by oxime, dioxolane
or diacetyl groups.
antiplatelet agents against ADP and AA induced
aggregation of guinea pig PRP. In particular, the pyr-
rolidino derivative 4d exhibited the highest activity
being equipotent to ASA (IC50=0.061 mM) in inhibit-
ing platelet aggregation elicited by AA (IC50=
0.085 mM) and surpassing the reference drug when
ADP-induced aggregation was considered. Indeed,
while in this last model ASA produced only 40% inhi-
bition, compound 4d completely prevented the aggrega-
tion (IC50=0.18 mM). The isopropylamino substituted
derivative 4a as well as the cycloalkylamino 4c displayed
comparable antiplatelet activity against the two aggre-
gation inducers employed. Finally, the remaining 4b, 4e
and 4f were able to prevent AA e€ect only at millimolar
concentration and resulted ine€ective against ADP
evoked platelet activation even when tested at high
concentrations (1 mM) (Table 1).
The reaction of 3-formylchromone with S-methyl-
isothiourea sulfate has already been carried out by
Petersen20 who obtained only the 5-hydroxy-2-methyl-
thio-5H-[1]benzopyrano[4,3-d]pyrimidine 9 in the pre-
sence of 1M NaOH solution. We have now
reinvestigated this reaction and we have veri®ed that
really compound 9 is the predominant product, in the
Petersen method, but also a little amount of compound 10
has been formed. On the other hand, if the reaction is
carried out in ethanol solution, in the presence of equi-
molar sodium ethoxide, the yield of 10 is greater.
Then, the hemiacetalic hydroxy-groupin position 5 of the
benzopyrano[4,3-d]pyrimidine system has been replaced
with the proper amines in the presence of TiCl4, according
to a method already described (ref 10 and references
cited therein).
In vivo experiments
In the carrageenan paw oedema assay an anti-
in¯ammatory activity, comparable to the reference drug
indomethacin (10 mg kgÀ1 os), was detected for benzo-
thiopyranopyrimidine derivative 3b, pyrrolidino 4d and
piperidino 4e benzopyranopyrimidine derivatives (100
mg kgÀ1 os), which signi®cantly inhibited (about 50%)
rat paw swelling versus the control group. As regards
the analgesic activity, no compounds exhibited a protec-
tive activity in writhing test. Among the molecules under
study, 2-aminophenyl substituted derivative 3d and
cyclopropylamino derivative 4c prevented the febrile
response caused by Escherichia coli (Table 2).
Finally, we observed that some of the compounds 4a±f,
transformed in hydrochlorides, can be converted into
the corresponding 5-ethoxy-2-methylthio derivative 14
(Fig. 3) during the recrystallization process in absolute
ethanol at 70±80 ꢀC.
Preliminary studies suggested that this is a general
behaviour for all the prepared compounds, but only as
hydrochlorides. In fact no transformation has been evi-
denced during the recrystallization of the basic form of
compounds 4, or by re¯uxing them in absolute ethanol
and sodium ethoxide. So, to understand if this is a
common method to obtain a new series of 5-alkoxy
derivatives, we are planning to study the in¯uence of the
temperature and the hydrochloric acid concentration on
this reaction in the presence of di€erent alcohols.
As expected, the reference drug indomethacin (10
mg kgÀ1 os) caused gastric lesions while amongst the
aforementioned e€ective compounds the pyrimidine deri-
vatives 3b, 4d and 4e (100 mg kgÀ1 os) did not produce
Table 1. In vitro antiplatelet activity, expressed as ecacy (% max
e€ect) and potency (IC50 value), of acetylsalicylic acid (ASA) and
pyrimidine derivatives under study on guinea-pig PRPa aggregation
induced by ADP and arachidonic acid (AA)
Results
In vitro antiplatelet activity
ADP-induced
aggregationb
AA-induced
aggregationc
As regards the polycyclic pyrimidine derivatives 2, 3a±d
a complete inhibition of ADP and AA evoked aggrega-
tion was detected only for compound 3b (IC50= 0.52
and 0.35 mM versus ADP and AA, respectively). In
addition, only a partial inhibition of ADP induced aggre-
gation was observed for compound 3c (IC50=0.79 mM)
while the phenyl substituted analogue 3d caused an ASA-
like antiplatelet e€ect against AA-aggregation
(IC50=0.12 mM). As concerns the 2-methylthio sub-
stituted benzopyrano[4,3-d]pyrimidine derivatives three
compounds (4a, 4c, 4d) out of six proved to be e€ective
Compounds
n
Maximal
inhibition
(%)
IC50
(mM)
Maximal
inhibition
(%)
IC50
(mM)
d
2
6
Ð
Ð
0.52
0.79
0.82
56
60
100
32
0.91
0.82
0.35
Ð
3a
3b
3c
3d
6
6
6
6
40
100
78
52
100
0.12
4a
4b
4c
4d
4e
4f
6
6
6
6
6
6
100
d
0.58
Ð
100
100
100
100
100
100
0.26
0.58
0.54
0.08
0.65
0.75
100
100
0.54
0.18
Ð
d
d
Ð
ASA
10
43
Ð
100
0.06
aPRP, platelet-rich plasma.
b3 mM.
c50 mM.
Figure 3.
dIne€ective upto 1 mM.

632
O. Bruno et al. / Bioorg. Med. Chem. 9 (2001) 629±636
Table 2. Anti-in¯ammatory, analgesic and antipyretic activity
(expressed as percentage of inhibition) of indomethacin and pyr-
imidine derivatives under studya
Compounds
Dose
(mg kgÀ1 po)
n
Rat
paw
oedema
Mouse
writhing
test
Rat
E. coli
fever
2
100
100
100
100
100
10
10
10
10
10
11
0
44**
0
27
0
0
0
0
0
0
0
0
88*
3a
3b
3c
3d
10
4a
4b
4c
4d
4e
4f
100
100
100
100
100
100
10
10
10
10
10
10
0
0
0
58**
44**
12
0
0
17
29
0
23
33
100**
7
6
33
10
Indomethacin
10
28
45**
81**
93**
aSigni®cance as compared to controls: * p<0.05; ** p<0.01
any mucosal erosions (Fig. 4a). Within this groupof
molecules, only derivatives 3b and 4d, at the same dosage,
signi®cantly protected gastric mucosa from ethanol-
induced damage (Fig. 4b).
Discussion
Figure 4. E€ects on gastric mucosa following the oral administration of
selected pyrimidine derivatives 100mgkgÀ1 in conscious rats: (a) gastric
ulcerogenic activity of the compounds under study and indomethacin
(IND) (10 mgkgÀ1 os); (b) gastroprotective e€ects of the compounds and
vehicle alone on ethanol-induced gastric ulcers. The total length of gas-
tric lesions and the total area of ethanol-induced ulcers are considered
respectively. Values represent the mean of 6±8 animals each and the ver-
tical bars indicate SEM. ** p<0.01; * p<0.05.
The di€erent polycyclic pyrimidine derivatives 2, 3a±d
generally displayed minor antiplatelet and analgesic/anti-
pyretic activities with respect to the 2-aminobenzopyr-
anopyrimidine derivatives previously studied 1a±f. The
sulfurated derivative 3b coupled a moderate antiplatelet
activity against ADP and AA-induced aggregation with a
certain antiphlogistic action. It also was devoid of gastro-
lesive activity and signi®cantly prevented the necrotizing
activity of ethanol on rat gastric mucosa. Thus, these
experimental data indicate that amongst the di€erent
manipulations performed at the polycyclic skeleton in
compounds 2 and 3a±d only the presence of S atom in
the heterocyclic moiety 3b did not negatively in¯uence
the in vivo and in vitro biological activity.
ethanol-induced gastric ulcers. Since the occurrence of
microvascular injury is involved in this ulcer model, it is
likely that the anti-aggregating activity of such com-
pounds can contribute to prevent the necrotic erosions
induced by ethanol.23
Within the remaining compounds under study, iso-
propylamino 4a, cyclopropylamino 4c and pyrrolidino
4d derivatives revealed an attractive pharmacological
pro®le. Unlike the conventional cyclooxygenase inhi-
bitor ASA,21,22 they potently prevented both AA and
ADP stimulated aggregation of guinea-pig platelet in
vitro. In detail, 4d was as potent as ASA against AA
and exhibited the highest activity in this experimental
model being about 4 times more potent than 4a and 4c.
Furthermore, since the three compounds displayed only
2±3-fold lower potency against ADP induced aggregation,
it can be hypothesized that an additional, non TXA2-
dependent, mechanism can account for their anti-aggre-
gating activity.
By comparing the present data concerning compounds
4a±f with the ®ndings obtained for the 1a±f10 various
considerations can be drawn. The presence of the
2-methylthio instead of the 2-amino groupmarkedly
impaired the in vivo biological activities, since within
this new series only the cyclopropylamino 4c, pyrroli-
dino 4d, and piperidino 4e derivatives partly maintained
antiphlogistic/antipyretic properties while the corre-
sponding aminoderivatives exhibited all these activities.
As concerns the other isopropylamino 4a, terbutyla-
mino 4b and morpholino 4f derivatives a lack of any
signi®cant anti-in¯ammatory e€ect was reported. Fur-
thermore, this replacement of the amino basic moiety in
position 2 with a methylthio group seems to increase the
in vitro antiplatelet activity. In particular, the 2-methyl-
thio derivatives 4a, 4c, 4d bearing respectively iso-
When orally administered in rats, 4c and 4d, 4e acted as
antipyretic and antiphlogistic agents, respectively. The
compounds 4d and 4e, unlike indomethacin, did not
cause any gastric mucosal damage and compound 4d
displayed a signi®cant gastroprotective e€ect against
propylamino,
cyclopropylamino
and
pyrrolidino
substituents in position 5 of the heterocycle were more
e€ective than the corresponding 2-amino analogues in
inhibiting ADP-induced aggregation. They also prevented

O. Bruno et al. / Bioorg. Med. Chem. 9 (2001) 629±636
633
the platelet response to AA with a good potency.
Therefore, it seems reasonable to conclude that, within
this series, the nature of the substituent introduced in
position 5 could be crucial to determine the potential of
the biological activities in vitro. We can infer that a les-
ser steric hindrance of isopropylamino, cyclopropyla-
mino and pyrrolidino groups, in respect to the
terbutylamino, piperidino or morpholino moieties,
enhances antiplatelet activity. Furthermore, the pyrroli-
dino moiety of the most active compound 4d, has unlike
¯exibility compared to that of cyclopropylamino or iso-
propylamino group. On the contrary, the basic strength
of the 5-substituents does not seem important to raise
the antiplatelet activity, because the pyrrolidino 4d and
piperidino 4e derivatives have di€erent level of activity
in spite of their similar basicity. Moreover, it is note-
worthy that among the new compounds only the
pyrrolidino derivative 4d prevented ethanol-induced
gastric ulceration sharing the gastroprotective e€ect
displayed by almost all the 2-amino derivatives. It fol-
lows therefrom that within the benzopyranopyrimidine
series the occurrence of pyrrolidino moiety is associated
with remarkable pharmacological properties apart from
the presence of amino or methylthio 2-substitution.
[4,3-d]pyrimidine (3d). Guanidine hydrochloride (1.9 g,
20 mmol) was added to a solution of sodium ethoxide
obtained by dissolving sodium (0.46 g, 20 mmol) in dry
ethanol (20 mL). After stirring at room temperature for
15 min. the solid obtained (NaCl) was removed by ®l-
tration and washed with dry ethanol (5 mL). The di-
methylaminomethylene derivative of the suitable ketone
(10 mmol) dissolved in dry ethanol (30 mL) was slowly
added and the reaction mixture re¯uxed for 3±5 h. After
concentration under reduced pressure the residue was
poured into water (10 mL) and the suspension stirred for
1 h. The solid precipitated (white for 2, 3a, c, d, yellow for
3b) was ®ltered, washed with water (10 mL) and ethyl
ether (3 mL) and recrystallized by a proper solvent.
2. Yield: white crystals, 1.13 g, 75%. Mp=211±212 ꢀC
(from ethanol) (lit: 212±213 ꢀC11). Anal. calcd for
C8H11N3: C, 64.40; H, 7.43; N, 28.16. Found: C, 64.18;
H, 7.42; N, 28.26.
3a. Yield: 76%. Mp=187 ꢀC (from ethanol:ethyl ace-
tate, 1:2). 1H NMR, d (CDCl3): 2.60±3.05 (m, 4H,
2H5+ 2H6), 5.41 (br s, 2H, NH2, disappear swith D2O),
7.10±7.65 (m, 3H, H7+H8+H9), 8.10±8.40 (m, 2H,
H4+H10). IR, cmÀ1 (CHCl3): 3520±3410. Anal. calcd
for C12H11N3: C, 73.07; H, 5.62; N, 21.3. Found: C,
72.97; H, 5.63; N, 21.56.
In conclusion, from the present study it emerged that the
potent antiplatelet and gastric sparing compound 4d may
be useful for the development of new drugs endowed with
combined anti ADP/TXA2 activity. The bene®cial clinical
e€ect of such antiaggregating agents can be speculated on
the basis of the advantageous co-administration of
antithrombotic drugs acting with di€erent mechanisms in
the treatment of vaso-occlusive disorders.24,25
3b. Yield: 74%. Mp=167±168 ꢀC (from ethanol:ethyl
acetate, 1:1). 1H NMR, d (CDCl3): 3.84 (s, 2H, 2H5), 5.55
(br s, 2H, NH2, disappears with D2O), 7.10±7.50 (m, 3H,
H7+H8+H9), 8.10±8.45 (m, 2H, H4+H10). IR, cmÀ1
(CHCl3): 3530±3420. Anal. calcd for C11H9N3S: C, 61.37;
H, 4.21; N, 19.52. Found: C, 61.17; H, 4.17; N, 19.52.
3c. Yield: 90%. Mp=209 ꢀC (from ethanol) (lit.:
Experimental
209 ꢀC26). H NMR, d (CDCl3): 5.16 (s, 2H, 2H5), 6.74
1
Materials and general methods
(br s, 2H, NH2, disappears with D2O), 6.90±7.60 (m,
3H, H7+H8+H9), 8.00±8.35 (m, 2H, H4+H10). IR,
cmÀ1 (KBr): 3330±3170.
All chemicals were obtained from Sigma-Aldrich S.r.L.
(Milan, Italy). Melting points are uncorrected and were
measured with a Buchi 530 instrument (Buchi Labor-
atoriums-Technik AG, Flawil, Schweiz). IR spectra
were recorded with a Perkin-Elmer 398 spectro-
photometer (Perkin-Elmer, Milan, Italy). 1H NMR were
recorded on a Hitachi Perkin-Elmer R-600 (60 MHz)
instrument (Perkin-Elmer, Milan, Italy); chemical shifts
are reported as d (ppm) relative to tetramethylsilane
(TMS) as internal stardard; the signals were character-
ized as s (singlet), d (doublet), t (triplet), m (multiplet),
br s (broad signal); J in Hz. Reactions were followed by
TLC on Kieselgel 60F254 (DC-Alufolien, E. Merck,
Darmstadt, Germany).
3d. Yield: 80%. Mp=164±165 ꢀC (from ethanol). 1H
NMR, d (CDCl3): 5.37 (br s, 2H, NH2, disappears with
D2O), 6.23 (s, 1H, H5), 6.85±7.40 and 7.60±7.85 (2m,
3H, H7+H8+H9), 7.47 (s, 5H, Ph), 8.10±8.30 (m, 2H,
H4+H10). IR, cmÀ1 (CHCl3):3540-3425. Anal. calcd for
C17H13N3O: C, 74.17; H, 4.76; N, 15.26. Found: C,
74.08; H, 4.74; N, 15.28.
2-(Dimethylaminomethylene)cyclohexanone 5,12 2-(di-
methylaminomethylene) -3,4 -dihydro-1(2H)- naphtale-
none 6a,13 3-(dimethylaminomethylene)-4-thiochroma-
none 6b,14 3-(dimethylaminomethylene)-4-chromanone
6c,15 were prepared following the literature references.
Analyses for C, H, N (Æ0.3% of the theoretical value),
were determined with the Elemental Analyzer EA 1110
(Fison-Instruments, Milan, Italy).
3-(Dimethylaminomethylene)-2-phenyl-4-chromanone
6d16 was prepared by the following modi®ed method:
Flavanone (3.5 g, 15mmol) dissolved in DMFDMA (di-
methylformamide dimethylacetal) (7mL) was heated at
90±100ꢀC for 45min. The excess DMFDMA was distilled
under reduced pressure, ethyl ether (15 mL) was added to
red oil residue and the pink-white solid crystallized was
®ltered. Yield: pink-white crystals. 1.75 g (42%);
General procedure for the preparation of 2-amino-5,6,
7,8-tetrahydroquinazoline (2), 2-amino-5,6-dihydro-[e]
quinazoline (3a), 2-amino-5H-[1]benzothiopyrano[4,3-d]
pyrimidine (3b), 2-amino-5H-[1]benzopyrano[4,3-d]pyri-
midine (3c) and 2-amino-5-phenyl-5H-[1]benzopyrano

634
O. Bruno et al. / Bioorg. Med. Chem. 9 (2001) 629±636
Mp=127±129ꢀC. Anal. calcd for C18H17 NO2: C, 77.40;
H, 6.13; N, 5.01. Found: C, 77.41; H, 6.16; N, 4.97.
(2 mL). The freshly distilled relevant amine (5 mL) in
dry toluene (5 mL) was then added and the coloured
resulting mixture (orange with primary and green with
secondary amines) was treated with 9 (2.15 g, 10 mmol)
and a further excess of the same amine (3 mL) in dry
toluene (10 mL).
Preparation of 5-hydroxy-2-methylthio-5H-[1]benzopyr-
ano[4,3-d]pyrimidine (9) and (2-hydroxyphenyl)-(2-methyl-
thio-pyrimidin-5-yl)-methanone (10)
Method A. S-Methylisothiourea sulfate (1.39 g, 5 mmol)
was added to a suspension of chromone-3-carbaldehyde 8
(1.74 g, 10 mmol) in 1 M NaOH (10 mL), followed by H2O
(20 mL) and triethylamine (0,5 mL). The reaction mixture
was heated under stirring for 3 h at 70±80ꢀC. After cooling
the 5-hydroxy-2-methylthio-5H-[1] benzopyrano[4,3-d]
pyrimidine 9 precipitated as a rose-white solid which was
®ltered and recrystallized by ethanol (yield: 2.13 g, 87%).
After re¯uxing for 6 h, the orange-yellow suspension
was cooled and then concentrated ammonia solution
(3 mL), isopropanol (2 mL) and Kieselgur (2 g) were
added in succession by stirring.
The resulting slurry was ®ltered and washed with tolu-
ene. The organic phase was washed once with little water,
dried (MgSO4), concentrated under reduced pressure
and chromatographed on Florisil (100±200 mesh).
Finally the solvent was evaporated under reduced pres-
sure and the yellow oil so obtained was crystallized from
absolute ethanol. Compound 4d is an oil which do not
crystallize, so the puri®cation was achieved by distillation
in vacuo.
The basic mother-waters were acidi®ed to pH 5±6 with
1 M HCl and extracted twice with CHCl3. The organic
phases were dried (MgSO4) and concentrated under
reduced pressure to give a yellow oil which crystallized,
with ether:petroleum ether (1:1) (5 mL), as a yellow
solid corresponding to (2-hydroxyphenyl)-(2-methyl-
thio-pyrimidin-5-yl)-methanone 10 (yield: 0.25 g, 10%).
Compounds 4d and 4e have been converted into their
corresponding hydrochlorides with a saturated hydro-
gen chloride ethereal solution. When these salts have
been heated for recrystallization in absolute ethanol at
70±80 ꢀC both were converted into the same 5-ethoxy-2-
methylthio-5H-[1]benzopyrano[4,3-d]pyrimidine 14: white
crystals, mp=96±97 ꢀC. 1H NMR, d (CDCl3):1.20 (t,
J=6.6, 3H, CH3), 2.66 (s, 3H, SCH3), 3.85 (q, J=6.6,
2H, CH2), 6.17 (s, 1H, H5), 7.00±7.70 (m, 3H,
H7+H8+H9), 8.20±8.55 (m, 2H, H10+H4). Anal. calcd
for C14H14N2O2S: C, 61.29; H, 5.14; N, 10.21. Found:
C, 61.12 ; H, 5.05; N, 10.42.
Method B. S-methylisothiourea sulfate (1.39 g, 5 mmol)
was added to a solution of sodium ethoxide obtained by
dissolving sodium (0.23 g, 10 mmol) in dry ethanol; after
stirring at room temperature for 10 min. the white solid
obtained (Na2SO4) was removed by ®ltration and chro-
mone-3-carbaldehyde 8 (1.74 g, 10 mmol) was added to
the ethanolic solution. The mixture was re¯uxed for 6 h,
then the solvent was evaporated under reduced pressure
to give a orange-red solid as a mixture of compounds 9
and 10 (TLC: Kieselgel, CHCl3:CH3OH 9.5:0.5) The
solid was dissolved in CHCl3 and the organic phase was
washed twice with 1 M NaOH (10 mL), dried (MgSO4)
and evaporated under reduced pressure to give a solid
corresponding to the sole 9 which was recrystallized by
ethanol (yield: 1.25 g, 51%). The basic solution was
acidi®ed (pH=5±6) with 1 M HCl and extracted twice
with CHCl3 (10 mL). The organic phase was dried
(MgSO4) and evaporated under reduced pressure to give
a yellow solid, recrystallized by ethanol, equivalent to
the compound 10 (yield: 0.85 g, 35%).
4a. Yield: 82%. Mp=93 ꢀC (from absolute ethanol). 1H
NMR, d (CDCl3): 0.90±1.35 (m, 6H, 2CH3), 2.13 (br s,
1H, NH, disappears with D2O), 2.61 (s, 3H, S-CH3),
3.10±3.70 (m, 1H, CH-isop.), 6.03 (br s, 1H, H5, became s
after exchange with D2O), 6.90±7.70 and 8.10±8.50 (2m,
4H, Ar), 8.45 (s, 1H, H4). Anal. calcd for C15H17N3OS:
C, 62.69; H, 5.96; N, 14.62. Found: C, 62.80; H, 5.88; N,
14.67.
4b. Yield: 81%. Mp=117±118 ꢀC (from absolute etha-
9. Mp=196±197 ꢀC (lit: 196±198 ꢀC20). 1H NMR, d
(DMSO-d6): 2.67 (s, 3H, SCH3), 6.61 (s, 1H, H5), 7.05±
7.80 (m, 4H, H7+H8+H9+OH, 1H disappears with
D2O), 8.20±8.50 (m, 1H, H10), 8.74 (s, 1H, H4). IR,
cmÀ1 (CHCl3): 3000±3400 (OH).
nol). H NMR, d (CDCl3): 1.26 (s, 9H, 3CH3), 1.80±
1
2.16 (br s, 1H, NH, disappears with D2O), 2.63 (s, 3H,
S-CH3), 6.20 (br s, 1H, H5 became s after exchange with
D2O), 6.85±7.56 and 8.10±8.50 (2m, 4H, Ar), 8.41 (s,
1H, H4). Anal. calcd for C16H19N3OS: C, 63.76; H,
6.35; N, 13.94. Found: C, 63.52; H, 6.30; N, 13.83.
ꢀ
1
10. Mp=112±114 C. H NMR, d (CDCl3):2.67 (s, 3H,
SCH3), 6.80±7.30 and 7.45±7.85 (2m, 4H, Ar), 8.87 (s,
2H, pyrim.), 11,70 (s, 1H, OH, disappears with D2O).
IR, cmÀ1 (CHCl3): 3460±2760 (OH); 1617 (CˆO). Anal.
calcd for C12H10N2O2S: C, 58.58; H, 4.09; N, 11.37.
Found: C, 58.43; H, 3.95; N, 11.22.
4c. Yield: 69%. Mp=104±105ꢀC (from absolute etha-
1
nol). H NMR, d (CDCl3): 0.30±0.70 (m, 4H, 2CH2),
2.40±3.10 (m, 2H, CH-Cyclop.+NH, disappears with
D2O), 2.61 (s, 3H, S-CH3), 5.96 (s, 1H, H5), 6.90±7.60 and
8.15±8.48 (2m, 4H, Ar), 8.38 (s, 1H, H4). Anal. calcd for
C15H14N3OS: C, 63.36; H, 4.96; N, 14.78. Found: C,
63.07; H, 5.24; N, 14.62.
General procedure for N-substituted 2-methylthio-5H-[1]
benzopyrano[4,3-d]pyrimidin-5-amines (4a±f)
4d. Yield: 69%. Bp=190 ꢀC/0.1 mmHg. 1H NMR, d
(CDCl3): 1.60±1.95 (m, 4H, 2CH2), 2.63 (s, 3H, S-CH3),
2.60±3.20 (m, 4H, 2CH2N), 6.20 (s 1H, H5), 6.83±7.60
TiCl4 (1.20 mL, 11 mmol) was added by stirring to an
ice-cooled solution of dry toluene (40 mL) and anisole

O. Bruno et al. / Bioorg. Med. Chem. 9 (2001) 629±636
635
and 8.10±8.45 (2m, 4H, Ar), 8.46 (s, 1H, H4). Anal.
calcd for C16H17N3OS: C, 64.19; H, 5.72; N, 14.04.
Found: C, 64.41 ; H, 5.68; N, 14.28.
In vivo experiments
Wistar female rats (250±300 g) and Swiss female mice
(25±35 g) were used (Morini, S.Polo-RE, Italy). The test
compounds were suspended in 0.5% methylcellulose
and were orally administered at 100 mg kgÀ1 in 0.5 ml/
100 g body weight to rats and at 0.15 ml/100 g body
weight to mice 1 h before the application of phlogogen,
algogen and pyretogen agent. Indomethacin 10 mg kgÀ1
os was used as reference drug in all the experimental tests
while control animals received an equivalent volume of
vehicle alone. Antiphlogistic, analgesic, antipyretic and
gastric ulcerogenic/gastroprotective activities were eval-
uated following the experimental procedures already
described.10 Brie¯y, antiphlogistic activity was studied in
rats by inducing paw oedema through subplantar injec-
tion of carrageenan, analgesic activity was evaluated in
mice by acetic acid writhing test and antipyretic activity
was determined in rats with E. coli-induced fever. The
pharmacological activities were expressed as the per-
centage of inhibition calculated from the di€erence in
the response between the treated and the control group
at the time the maximum noxious e€ect occurred.
4e. Yield: 89%. Mp=97±98 ꢀC (from absolute ethanol).
1H NMR, d (CDCl3): 1.30±1.75 (m, 6H, 3CH2), 2.35±
3.20 (m, 4H, 2CH2N), 2.62 (s, 3H, SCH3), 5.98 (s, 1H,
H5), 6.80±7.60 and 8.10±8.45 (2m, 4H, Ar), 8.42 (s, 1H,
H4). Anal. calcd for C17H19N3OS: C, 65.15; H, 6.11; N,
13.41. Found: C, 64.95; H, 6.13; N, 13.38.
4f. Yield: 87%. Mp=117 ꢀC (from absolute ethanol).
1H NMR, d (CDCl3): 2.62 (s, 3H, SCH3), 2.60±3.00 (m,
4H, 2CH2N), 3.5±3.8 (m, 4H, 2CH2O), 5.98 (s, 1H, H5),
6.80±7.60 and 8.10±8.45 (2m, 4H, Ar), 8.45 (s, 1H, H4).
Anal. calcd for C16H17N3O2S: C, 60.93; H, 5.43; N,
13.32. Found: C, 60.91; H, 5.44; N, 13.25.
Pharmacological methods
All the new polycyclic pyrimidine derivatives were
screened for their in vitro antiplatelet activity in guinea-
pig platelet-rich plasma inducing the aggregation by
ADP or AA. In vivo experiments were performed in
order to evaluate their antiphlogistic, analgesic and
antipyretic activities in rat paw oedema, mouse writhing
test and rat E. coli-induced pyrexia, respectively. Selec-
ted compounds were tested for their erosive e€ects on
gastric mucosa and their gastroprotective activity
against ethanol-induced gastric ulcers in rats.
Selected compounds were studied to estimate their
ulcerogenic/gastroprotective e€ects. In detail, 5 h after
oral administration of these molecules the acute gastric
ulcerogenic activity was determined by examining rat
stomachs by means of a stereomicroscope connected to
an image analyser system (Leitz, ASM 68K) and the
total length (mm) of the gastric lesions was considered
for each stomach. For the study of the gastroprotective
activity the molecules were administered 1 h before the
necrotizing agent ethanol (1 mL/rat 90% v/v). One hour
after the oral administration of ethanol, the stomachs
were microscopically analyzed and the surface of each
haemorrhagic lesion was measured.
In vitro antiplatelet activity
Male guinea-pig (Morini, S.Polo-RE, Italy) blood
obtained by cardiac puncture after CO2 euthanasia was
collected in plastic tubes containing sodium citrate
(3.8% w/v; 9 parts blood: 1 part sodium citrate). The
blood was centrifuged for 10 min at 200 g to obtain
platelet rich plasma (PRP) and from the remaining
blood, which was centrifuged for 10 min at 2000 g
platelet poor plasma (PPP) was produced. Platelet
aggregation was performed in an Aggrecorder PA 3220
aggregometer (A. Menarini, Firenze, Italy) following
Born's turbidimetric method.27 Aggregation was recor-
ded as percent change in light transmission: the baseline
was set by using PRP and full transmission (100%) was
set by using PPP. PRP (250 mL) was preincubated at
37 ꢀC for 5 min with solvent (dimethyl sulphoxide,
DMSO), the compounds under study or the reference
drug ASA (from 10À5 to 10À3 M) before the addition
of the platelet aggregatory agent. PRP aggregation was
induced by 3 mM ADP (25 mL) (Sigma) or by 50 mM
AA (Menarini) and using concentrations sucient to
achieve maximum aggregation. Tests were performed
within 3 h to avoid platelet inactivation. The e€ects of
test compounds and acetylsalycilic acid, used as refer-
ence drug, were determined as percent inhibition cal-
culated from the total aggregation in 5 min. Control
samples received the same volume addition of DMSO
at the ®nal concentration of 0.5%. This concentration
of solvent did not interfere with platelet assay.
Statistical analysis
The results were expressed as meanÆSEM and the
means were compared using Student's t-test, p value
<0.05 or <0.01 being considered as statistically sig-
ni®cant or highly signi®cant, respectively.
Acknowledgements
The authors wish to thank C. Rossi and F. Tuberoni for
spectral recording, G. Domenichini for skillful help in
the experimental work. Financial support from MURST
(Co®nanziamento Nazionale) is gratefully acknowl-
edged.
References and Notes
1. Final report on the aspirin component of the ongoing
Physician' Health Study Research Group. N. Engl. J. Med.,
1989, 321, 129±135.
2. ATC (antiplatelet trialist collaboration). Collaborative
overview of randomized trials of antiplatelet therapyÐI: pre-
vention of death, myocardial infarction, and stroke by pro-

636
O. Bruno et al. / Bioorg. Med. Chem. 9 (2001) 629±636
longed antiplatelet therapy in various categories of patients.
Br. Med. J. 1994, 308, 81
14. Bruno, O.; Schenone, S.; Ranise, A. Antiin¯ammatory
agents: new series of N-substituted amino acids with complex
pyrimidine structures endowed with antiphlogistic activity. Il
Farmaco 1999, 54, 95.
15. Bruno, O.; Schenone, S.; Ranise, A. Synthesis and phar-
macological pro®le of novel N-substituted N-5H-1benzopyr-
ano4,3-dpyrimidin-2-yl-N-methylglicinamides. Farmaco 1996,
51, 137.
16. Litkei, G.; Patonay, T.; Kardos, J. Synthesis of Mannich
compounds from ¯avanones. Org. Prep. Proc. Int 1990, 22, 47.
17. Harnisch, H. Chromon-3-carbaldehyde. Liebigs Ann.
Chem. 1972, 765, 8.
18. Ghosh, C. K.; Khan, S. Heterocyclic system: Part VII.
Reaction of 4-oxo-4H-[1]benzopyran-3-carbaldehyde and -3-
carboxylic ester with guanidine. Ind. J. Chem. 1979, 18B, 128.
Â
19. Pene, C.; Hubert-Habart, M. Di€erent possibilities for the
formation of pyrimidines from 3-formylchromone. J. Hetero-
cyclic Chem. 1980, 17, 329.
20. Petersen, U.; Heitzer, H. Heterocyclensynthesen mit
4-oxo-4H-chromen-3-carbaldehyd. Liebigs Ann. Chem. 1976,
9, 1663.
21. Dohi, M.; Sakata, Y.; Seki, J. The antiplatelet actions of
FR 122047, a novel cyclooxygenase inhibitor. Eur. J. Phar-
macol. 1993, 243, 179.
22. Yokota, K.; Yamamoto, N.; Morimoto, Y. E€ect of
KBT-3022, a new diphenilthiazole derivative on platelet func-
tions. J. Pharm. Pharmacol. 1995, 47, 768.
3. AMIS (aspirin myocardial infarction study researh group) A
randomized controlled trial of aspirin in persons recovered from
myocardial infarction. J. Amer. Med. Ass., 1980, 243, 661.
4. Bousser, M. G.; Eschwege, E.; Haguenau, M. ``AICLA'' con-
trolled trial of aspirin and dipyridamole in the secondary preven-
tion of athero-thrombotic cerebral ischemia. Stroke 1983, 14, 5.
5. Hourani, S. M. O.; Cusack, N. J. Pharmacological recep-
tors on blood platelet. Pharmacol. Rev. 1991, 43, 243.
6. Bruno, O.; Bondavalli, F.; Ranise, A. N-Substituted 3-(4-
hydroxy-3,5-diphenyl-1H-pyrazol-1-yl)propanamides, propa-
namines and 2-(4-hydroxy-3,5-diphenyl-1H-pyrazol-1-yl)etha-
namines with platelet antiaggregating, hypotensive, anti-
arrhythmic and other activities. Il Farmaco 1991, 46, 477 and
references therein cited.
7. Bondavalli, F.; Bruno, F.; Ranise, A. 3,5-Diphenyl-1H-
pyrazole derivatives IX: 2-substituted 4-phenyl-5-(3,5-diphe-
nyl-1H-pyrazol-1-yl)pyrimidines with platelet antiaggregating
and other activities. Il Farmaco 1992, 47, 171.
8. Longobardi, M.; Bargagna, A.; Mariani, E. 2H-1-Benzo-
pyran derivatives with platelet antiaggregating and other
activities. Il Farmaco 1993, 48, 1121.
9. Mosti, L.; Menozzi, G.; Schenone, P. Ethyl or methyl
6-substituted 3-(benzoylamino)-2-oxo-2H-pyran-5-carboxylates
and 3-(benzoylamino)-7,8-dihydro-2H-1-benzopyran-2,5(6H)-
diones with local anesthetic, platelet antiaggregating and other
activities. Il Farmaco 1994, 49, 45.
10. Bruno, O.; Schenone, S.; Ranise, A. Synthesis and phar-
macological screening of novel non-acidic gastric-sparing
antipyretic/antiin¯ammatory agents with antiplatelet proper-
ties: 5-alkyl/cycloalkylamino substituted 2-amino-5H-[1]ben-
zopyrano[4,3-d]pyrimidines. Arzneim.-Forsch./Drug Res. 2000,
50, 140.
11. Bredereck, H.; E€enberger, F.; Botsch, H.; Rehn, H.
Umsetzungen von vinylogen Carbonsaurenamiden zu Hetero-
cyclen. Chem. Ber. 1965, 96, 1081 and references cited therein.
12. Shuda, P. F.; Ebner, C. B.; Morgan, T. M. The synthesis
of Mannich bases from ketones and esters via enaminones.
Tetrahedron Lett. 1986, 27, 2567.
13. Wagner, R. M.; Jutz, C. Reaction of vinylogous for-
mamidinium salts with nucleophiles. I. Variance of the pyr-
imidine synthesis. Chem. Ber. 1971, 104, 2975.
23. Ohno, T.; Majima, M.; Hamur, M. Involvement of plate-
let aggregation in ethanol-induced rat gastric mucosal injury.
Gastroenterology 1997, 112, A242.
24. Hall, P.; Nakamura, S.; Maillo, L. A randomized com-
parison of combined ticlopidine and aspirin therapy versus
aspirin therapy alone after successful intravascular ultra-
sound-guided stent implantation. Circulation 1996, 93, 215.
25. Goods, C. M.; Al-Shaibi, K. F.; Liu, M. W. Comparison
of aspirin alone versus aspirin plus ticlopidine after coronary
artery stenting. Am. J. Cardiol. 1996, 78, 1042.
26. Bellasoued, F.; Maitte, P. Reactivite de b-enaminones
Â
cyclique II. Cyclocondensation directe d'amidine, d'hydrazine
et d'hydroxylamines sur la pyrrolidinomethylene-3 chroma-
none-4. J. Het. Chem. 1986, 23, 1753.
27. Born, G. V. R. Aggregation by ADP and its reversal.
Nature 1962, 194, 927.