Investigation of platelet aggregation inhibitory activity by phenyl amides and esters of piperidinecarboxylic acids

Article abstract of DOI:10.1016/S0968-0896(02)00599-0

A series of anilides and phenyl esters of piperidine-3-carboxylic acid (nipecotic acid) were synthesized and tested for the ability to inhibit aggregation of human platelet rich-plasma triggered by adenosine 5′-diphosphate (ADP) and adrenaline. As a rule,

Full text of DOI:10.1016/S0968-0896(02)00599-0

Bioorganic & Medicinal Chemistry 11 (2003) 1439–1450
Investigation of Platelet Aggregation Inhibitory Activity by
Phenyl Amides and Esters of Piperidinecarboxylic Acids
Modesto de Candia,a Luciana Summo,a,y Antonio Carrieri,a Cosimo Altomare,a,*
Adele Nardecchia,b Saverio Cellamarec and Angelo Carottia
a
`
Dipartimento Farmaco-chimico, Universita degli Studi, Via Orabona 4, 70125, Bari, Italy
bDipartimento di Medicina e Oncologia, Ospedale Policlinico, P.zza G. Cesare 1, 70125, Bari, Italy
c
`
Dipartimento di Scienze del Farmaco, Universita ‘G. D’Annunzio’, Via dei Vestini, 66100, Chieti, Italy
Received 2 August 2002; accepted 11 November 2002
Abstract—A series ofanilides and phenyl esters ofpiperidine-3-carboxylic acid (nipecotic acid) were synthesized and tested for the
ability to inhibit aggregation ofhuman platelet rich-plasma triggered by adenosine 5 0-diphosphate (ADP) and adrenaline. As a rule,
amides were about two times more active than the corresponding esters, and derivatives bearing substituents at the para position of
the phenyl ring were significantly more active than the meta-substituted ones. Among the tested compounds, 4-hexyloxyanilide of
nipecotic acid (18a) was found to be the most active one, its IC50 value being close to that ofthe most active bis-3-carbamoyl-
piperidines reported in literature (ca. 40 mM) and aspirin (ca. 60 mM) in ADP- and adrenaline-induced aggregation, respectively.
Compared with the isomeric 4-hexyloxyanilides ofpiperidine-2-carboxylic (pipecolinic) and piperidine-4-carboxylic (isonipecotic)
acids, compound 18a showed higher activity, and a Hansch-type quantitative structure–activity relationship (QSAR) study high-
lighted lipophilicity and increase in electron density ofthe phenyl ring as the properties which mainly increase the antiplatelet
activity (r2=0.74, q2=0.64). The interaction ofnipecotoyl anilides with phosphatidylinositol, a major component ofthe inner layer
ofthe platelet membranes, was investigated by means offlexible docking calculation methods to give an account ofa key event
underlying their biological action.
# 2003 Elsevier Science Ltd. All rights reserved.
Introduction
Although it is a weak platelet activator, ADP plays a
relevant role in platelet function. It can trigger platelet
activation, which is mediated by three purinergic recep-
tors (P2Y, P2T, P2X) showing distinct specificity, and is
responsible for the secondary wave of platelet aggrega-
tion, followed by ADP release from dense granules.7ꢀ9
Thrombosis and thromboembolism are major causes of
morbidity and mortality.1 Platelet aggregation is a cen-
tral step in the formation of intravascular thrombi, and
deposition ofplatelets is also involved in the formation
1,2
ofatherosclerotic plaques.
When vascular endothe-
lium is damaged, the subendothelial matrix exposes
adhesive proteins, such as collagen, and the contact with
them activates the platelets. As a consequence the pla-
telets change their discoidal shape, form a network in
the site ofinjury and release from their dense granules a
number ofphysiological proaggregatory agonists, such
as adenosine-50-diphosphate (ADP), platelet activating
factor (PAF), serotonin, thromboxane A2 (TxA2) and
cathecolamines.3ꢀ6 These agonists can activate the sur-
rounding platelets leading to thrombus formation.
Antiplatelet agents, such as aspirin, dipyridamole, thie-
nopyridines, and platelet glycoprotein IIb/IIIa antago-
nists have amply demonstrated their utility in
preventing and treating coronary artery thrombosis.10,11
In particular, ticlopidine and clopidogrel, active in
S-configuration (Fig. 1), both belonging to the thieno-
pyridine class, have been increasingly used to prevent
ischemic events. Moreover, their association with
aspirin are effective in suppressing thrombotic compli-
cations after coronary stenting and in other conditions
in which patients are at high risk ofatherothrombotic
events.12,13 Despite their benefits, administration of
thienopyridines can in certain cases be associated with
several undesired side effects (e.g., neutropenia, agranu-
locytosis, gastrointestinal toxicity),12 whereas a problem
*Corresponding author. Tel.: +39-080-544-2781; fax: +39-080-544-
2230; e-mail: altomare@farmchim.uniba.it
yPresent address: Berlin Chemie AG, Menarini Group, Glienicker weg
125, 12489 Berlin, Germany.
0968-0896/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00599-0

1440
M. de Candia et al. / Bioorg. Med. Chem. 11 (2003) 1439–1450
Figure 1. Structures ofantithrombotic thienopyridines (ticlopidine and clopidogrel), literature carbamoylpiperidines ( 1), and general stucture ofthe
piperidine-3-carboxylic acid derivatives examined in this paper (2–22).
in clinical use ofall the currently available antiplatelet
agents is the interference of physiological platelet func-
tion in haemostasis.14 Therefore, the development of
antiplatelet agents, showing limited side effects, is still a
relevant goal ofthe pharmaceutical research.
which in contrast should allow one to challenge the
hypothesis ofthe need ofa double amonium head and
to modulate the lipophilicity ofthe phenyl substituents.
As a part ofour ongoing research aimed at developing
new potential antithrombotic drugs,30,31 we screened a
series ofderivatives structurally simpler than com-
pounds 1 to gain insights into the minimal structural
features necessary for antiplatelet activity of nipecota-
mides. Thus, we synthesized and tested for their effects
on the human platelet aggregation (ADP and adrenaline
were chosen among the several known triggers) a series
ofnipecotic acid anilide congeners, and a number of
phenyl esters as well, bearing in meta and para positions
ofthe phenyl group substituents selected according to
the Hansch QSAR analysis principles. A comparison of
the most active nipecotic acid derivatives with their iso-
mers obtained from pipecolinic (piperidine-2-car-
boxylic) and isonipecotic (piperidine-4-carboxylic) acids
was also undertaken. Herein, we report on structure–
activity relationships ofthese compounds.
Among the numerous compounds reported in literature,
amides ofpiperidine-3-carboxylic acid (nipecotamides),
namely a,a0-bis[3-(N,N-dialkylcarbamoyl)piperidino]-p-
xylenes 1 (Fig. 1), found to inhibit in vitro human pla-
telet aggregation induced by ADP, collagen, thrombin,
adrenaline, and the stable TxA2-mimetic U46619, and
to be effective in in vivo thrombosis models,15ꢀ17
attracted our attention.
Bis-3-carbamoylpiperidine derivatives bearing n-hexy-
loxy (1a) or amino (1b) substituents at C-20 ofthe
phenyl group in the xylene bridge have been found to be
among the most active, whereas from structure–activity
studies15,17 it had been inferred that double cationic
head, amide carbonyl, appropriate stereochemistry of
the C-3, and a log P value ofabout 4.5 have to be con-
sidered relevant molecular determinants ofthe anti-
platelet activity. A hypothesis on the mechanism of
action ofnipecotamides has been suggested based on
several studies carried out by Gollamudi and co-
workers.15,17ꢀ29 Due to their lipophilicity, and surface
activity, nipecotamides can penetrate the platelet mem-
branes and interact with anionic phospholipids (mainly
phosphatidylinositol, PI, and phosphatidylserine, PS),
rendering them resistant to hydrolysis catalyzed by
phospholipase-C to the second messengers inositol
1,4,5-triphosphate (IP3) and s,n-1,2-diacylglycerol
(DAG). Nipecotamides inhibit the phosphoinositide
Chemistry
The phenyl piperidine-3-carboxamides (Table 1, 2a–
20a) and anilides 23a and 24a (Table 2) were synthesized
using known methods according to the pathway shown
in Scheme 1.
1-(Tert-butyloxycarbonyl)-piperidine carboxylic acids,
prepared by a reported method,32 were condensed with
para or meta substituted anilines, using dicyclohexyl-
carbodiimmide (DCC) and 1-hydroxybenzotriazole
hydrate (BtOH) as coupling reagents in dry dichloro-
methane. N-[4-(hexyloxy)phenyl]-N-methylamine, neces-
sary to synthesize derivative 19a, was obtained through
a one-pot reductive methylation of p-hexyloxy aniline in
the presence ofaqueous ofrmaldheyde (10% Pd/C as
catalyst). The nitroanilides 6a and 7a were prepared in
better yields by reacting the N-BOC-nipecotoyl chloride
with meta- and para-nitroaniline, respectively. N-BOC-
p-nitrophenyl amide (7a) was reduced by catalytic
turnover as well. As a consequence, the levels ofIP 3 and
2+
ofcytosolic Ca
concentrations, necessary for myosin
phosphorylation and subsequent platelet activation, are
reduced.
To the best ofour knowledge, literature is lacking of
information on the effects on platelet aggregation of
simple aryl amides ofpiperidine-3-carboxylic acid,

M. de Candia et al. / Bioorg. Med. Chem. 11 (2003) 1439–1450
1441
Table 1. Antiplatelet activity and lipophilicity ofX-substituted
phenyl esters and anilides ofnipecotic acid
hydrogenation to p-aminophenyl amide (8a), that in
turn was reacted with methanesulfonyl chloride to give
the N-BOC derivative ofcompound 14a, and trans-
formed into the dimethylamino congener 16a through a
one-pot reductive methylation with aqueous for-
maldheyde. Deprotection ofthe N-BOC intermediates
with HCl gas yielded the amide derivatives as hydro-
chloride salts. Physical and spectral data ofthe newly
synthesized compounds are reported in Table 3. The
phenyl esters ofpiperidinecarboxylic acid tested in this
study (2b, 4b, 7b, 13b, 17b, 18b, 21b–24b) were synthe-
sized using our previously reported procedures.33
Compd
Y
X
IC50 (mM) or% inhibition log Pb
at 100 mMa
ADP
Adrenaline
Control
2a
2b
3a
4a
4b
5a
6a
7a
15ꢁ1%
200
570
0%
162
15ꢁ1%
4ꢁ2%
0%
NH
O
NH
NH
O
NH
NH
NH
O
NH
NH
NH
NH
NH
NH
O
H
H
3-Br
4-Br
4-Br
4-Cl
3-NO2
4-NO2
4-NO2
4-NH2
4-CF3
4-CN
4-CH3
3-OCH3
4-OCH3
4-OCH3
1.53
1.78
2.65
2.65
2.78
2.50
1.82
1.82
1.82
0.30
2.86
1.56
2.03
1.60
1.60
1.87
0.34
1.44
1.69
3.35
3.61
4.25
4.52
4.31
5.20
3.04
3.64
39ꢁ13%**
245
10ꢁ1%
0%
28ꢁ6%**
10ꢁ1%
0%
12ꢁ7%
7b
8a
9a
Results and Discussion
372
0%
270
0%
The nipecotic acid derivatives were tested as antiplatelet
agents by measuring their inhibitory actions on the in
vitro aggregation ofhuman platelet-rich plasma (PRP)
triggered by ADP and adrenaline, both at 10 mM con-
centrations, by using a turbidimetric method.19,30,34,35
The aggregation tracing relative to the controls revealed
a typical biphasic aggregation with both agonists. In a
preliminary screening, carried out at a dose of100 mM,
a number ofthe test compounds showed platelet aggre-
gation inhibitory effects significantly different from
controls. In general, the most active compounds inhib-
ited both the primary (related to the receptor initial
events) and secondary (irreversible, due to the release of
agonists from the dense granules) phases of the in vitro
aggregation. Compounds showing inhibitory effects sig-
nificantly different from controls (p<0.001) were sub-
jected to a screening ofdose-dependent effects on the in
vitro aggregation ofhuman PRP and concentrations
inhibiting platelet aggregation by 50% (IC50) were
determined. The results are summarized in Table 1,
where the % inhibition at 100 mM is reported as the
mean (ꢁSEM) of3–4 individual determinations ofr
weak inhibitors and the average IC50 values from
2–3 individual determinations for medium to strong
inhibitors.
10a
11a
12a
13a
13b
14a
15a
16a
17a
17b
18a
18b
19ac
20a
21b
22b
1a
218
4ꢁ2%
5ꢁ1%
5ꢁ1%
23ꢁ3%*
0%
0%
269
NH 4-NHSO2CH3
NH
NH
NH
O
NH
O
4-COCH3
4-N(CH3)2
4-C(CH3)3
4-C(CH3)3
4-OnC6H13
4-OnC6H13
175
45.1
66.0
38ꢁ5%***
70.2
100
NCH3 4-OnC6H13
NH
O
120d
102
4-nC7H15
4-I
4-OCH2Ph
214
235
38.7
O
ASA
59.1
aPlatelet-rich plasma (PRP) was pre-incubated with the test compound
or with dmso (0.5% v/v, control) at 37 ꢂC for 5 min. The inducer,
ADP (10 mM) or adrenaline (10 mM), was then added. Activity is
expressed as IC50 or percentage ofthe inhibitory response ofr less
active compounds at 100 mM concentration (meansꢁSEM, nꢃ3);
*p<0.05, **p<0.01, ***p<0.001 significantly different from control.
bn-Octanol-water partition coefficient (log P) calculated by CLOG P
software vers. 4.0 (Biobyte, Claremont, CA, USA).
cMaleate salt.
dValue obtained by extrapolation; 100 mM was the maximum concen-
tration tested.
Table 2. Antiplatelet activity and lipophilicity of4-hexyloxyphenyl esters and amides ofpiperidine carboxylic acids
Compd
Position on
piperidine ring
Y
IC50 (mM) or % inhibition
at 100 mMa
log Pb
Half-lives in
10% human
serum at
37 ꢂC (min)c
ADP
Adrenaline
Control
23a
23b
18a
18b
15ꢁ1%
32ꢁ7%***
nd
15ꢁ1%
29ꢁ3%***
nd
2
2
3
3
4
4
NH
O
NH
O
NH
O
4.54
4.78
4.25
4.52
3.49
3.84
s
3.6
s
50
s
45.1
100
141
120
66.0
38.0ꢁ5%***
24a
24b
75.1
50
aSee footnote (a) in Table 1; nd, not determinable, because ofstrongly unreproducible results.
bSee footnote (b) in Table 1.
cData are means ofthree determinations (RSD less than 5%); s, stable, that is, no hydrolysis observed within 2 h ofincubation.

1442
M. de Candia et al. / Bioorg. Med. Chem. 11 (2003) 1439–1450
Among the compounds tested so far, 4-hexyloxyphenyl
derivatives ofnipecotic acid ( 18a and 18b) exhibited the
highest inhibition ofplatelet aggregation. Their dose–
activity relationships are shown in Figure 2.
structure–activity relationships (QSARs); the IC50 value
of 19a, the only one obtained by extrapolation from the
aggregation inhibition data, was also excluded from the
regression analysis. An inspection ofthe plot ofpIC
50
values against log Ps (Fig. 3) suggested a trend of
correlation between the inhibition ofADP-stimulating
platelet aggregation and lipophilicity.
para-Hexyloxyanilide ofnipecotic acid ( 18a) attained
50% inhibition activity ofADP-induced aggregation at
a concentration (45 mM) close to the IC50 value for the
bis-3-carbamoylpiperidine 1a (39 mM) determined by
Gollamudi et al.17 Importantly, it showed a degree of
inhibitory activity similar to that ofaspirin, taken as the
reference standard, in adrenaline-induced PRP aggre-
gation.
A detailed analysis ofthe subsets ofpiperidine-3-car-
boxylic acid derivatives revealed that the single para-
meter log P accounts for about 92% (r2=0.918) ofthe
variance ofthe activities ofthe phenyl esters; the corre-
lation with log P alone was, however, poorer within the
anilide subset (r2=0.465), even after omitting from the
regression the strongest outliers 2a and 13a, the two
compounds having the lowest lipophilicity. In an effort
to understand how, besides lipophilicity, other physico-
chemical properties can affect the antiplatelet activity,
albeit the limited range ofbiological values (about 1 log
unit), we carried out a Hansch-type multiple linear
regression (MLR) with cross-validation on a matrix in
which each compound was described by log P, calcu-
lated molar refractivity (CMR)36 and molar volume
(MV) as parameters ofbulkiness, Hammett sigma ( s),
resonance and inductive sigma constants (sR and sI) as
the electronic parameters. MV and sigma constants
were calculated with the ACDLab program.37 Data
analysis was carried out by using the Q-PARVUS
package ver. 3.0.38 With the set ofpiperidinecarboxylic
acid derivatives having IC50 values (compounds 2a and
13a omitted from regression), and retaining only
Structure–activity relationships
Aggregation data of Table 1 showed that, with the
exception of4- tert-butyl derivatives (17a and 17b), the
amides were about two times more active than the cor-
responding esters, and derivatives bearing substituents
at the para position ofthe phenyl ring were significantly
more active than the meta-substituted ones. Indeed, the
meta-substituted congeners, irrespective oftheir physi-
cochemical properties, did not show any detectable
activity at 100 mM, whereas the substituents at the
position para ofthe phenyl ring enhance the activity
mainly as a function of their lipophilicity, at least for
compounds having a log P value higher than ca. 2 as
calculated by the CLOG P software.36 Only the finite
antiplatelet activity data (IC50) were used in a Hansch-
type regression analysis aimed at deriving quantitative
Scheme 1. (a) BtOH, DCC, dry CH2Cl2, 35–60%; (b) HCl gas, CHCl3, 0–5 ꢂC, 35–50%; (c) SOCl2, TEA, CH2Cl2, reflux; (d) TEA, CH2Cl2, 46%;
(e) H2, 10% Pd/C, EtOH, 98%; (f) methanesulfonyl chloride, dry pyridine, dry dioxane, reflux, 67%; (g) H2, 40% H2CO, 10% Pd/C, EtOH, 91%.

Table 3. Physicochemical and spectroscopic data ofthe newly synthesized compounds
Compd
Mp (ꢂC)
Molecular formula
IR (cmꢀ1 a
)
1H NMR data (d, ppm)b
3a
4a
5a
6a
7a
8a
9a
10a
11a
12a
13a
14a
15a
16a
17a
18a
234–236
251–254
228–230
189–192
266–267
260–261
266–270
266–269
214–217
202–204
204–207
282–284
272–274
233–236
225–228
177–180
C12H15N2OBrꢄHCl
C12H15N2OBrꢄHCl
C12H15N2OClꢄHCl
C12H15N3O3ꢄHCl
C12H15N3O3ꢄHCl
C12H17N3Oꢄ2HClꢄH2O
C13H15N2OF3ꢄHCl
C13H15N3OꢄHCl
3290, 1663, 780
3445, 1684, 820
3457, 1627, 832
10.76 (s, 1H), 9.33 (s, 2H), 8.07 (d, J=3, 1H), 7.63 (m, 1H), 7.33 (m, 2H).
10.46 (s,1H), 8.27 (s, 2H), 7.58 (d, J=8.8, 2H), 7.47 (d, J=8.8, 2H).
10.33 (s, 1H), 9.27 (s, 2H), 7.76 (d, J=9, 2H), 7.41 (d, J=9, 2H).
3413, 1686, 1548, 1353, 739
3210, 1695, 1558, 1328, 857
3420, 1659, 1516, 823
3248, 1692, 840
11.20 (s, 1H), 9.43 (s, 2H), 8.79 (t, J=3, 1H), 8.07(m, J=9, 2H), 7.66 (t, J=3, 1H).
11.23 (s, 1H), 9.30 (s, 2H), 8.30 (d, J=9, 2H), 7.99 (d, J=9, 2H).
10.44 (s, 1H), 9.84 (s, 5H), 7.65 (d, J=8.5, 2H), 7.22 (d, J=8.5, 2H).
10.87 (s, 1H), 9.20 (s, 2H), 7.94 (d, J=9, 2H), 7.71 (d, J=9 Hz, 2H).
10.65 (s, 1H), 8.82 (s, 2H), 7.77 (d, J=8.7, 2H), 7.65 (d, J=8.7, 2H).
10.43 (s, 1H), 9.39 (s, 2H), 7.58 (d, J=9, 2H), 7.11 (d, J=9, 2H), 2.50 (s, 3H).
11.42 (s, 1H), 9.35 (s, 2H), 7.40 (d, J=3, 1H), 6.65 (m, 3H), 3.70 (s, 3H).
3251, 2219, 1702, 837
3267, 1686, 822
C13H18N2OꢄHCl
C13H18N2O2ꢄHCl
C13H18N2O2ꢄHCl
C13H19N3O3SꢄHCl
C14H18N2O2ꢄHCl
C14H21N3Oꢄ2HClꢄH2O
C16H24N2OꢄHClꢄH2O
C18H28N2O2ꢄHCl
3254, 1659, 1215, 1039, 781
3449, 1684, 1237, 1028, 834
3422, 1687, 1342, 1161, 840
3285, 1699, 1663, 840
3423, 1679, 840
10.26 (s, 1H), 9.40 (s, 2H), 7.60 (d, J=9, 2H), 6.88 (d, J=9, 2H), 3.70 (s, 3H).
11.20 (s, 1H), 10.83 (s, 1H), 9.80 (s, 2H), 7.84 (d, J=9, 2H), 7.47 (d, J=9, 2H), 3.25 (s, 3H).
10.83 (s, 1H), 9.36 (s, 2H), 7.95 (d, J=9, 2H), 7.81 (d, J=9, 2H), 2.50 (s, 3H).
10.01 (s, 1H), 9.22 (s, 3H), 7.31 (d, J=9, 2H), 7.12 (d, J=9, 2H), 3.05 (s, 6H).
10.40 (s, 1H), 9.40 (s, 2H), 7.64 (d, J=9, 2H), 7.41 (d, J=9, 2H), 1.23 (s, 9H).
10.15 (s, 1H), 9.06 (s, 2H), 7.48 (d, J=9, 2H), 6.83 (d, J=9, 2H), 3.88 (t, J=6, 2H),
1.80–1.54 (m, 2H), 1.45–1.20 (m, 6H), 0.85 (t, J=6 Hz, 3H).
3420, 1690, 840
3276, 2777, 1653, 1237, 829
19a
20a
23a
23b
24a
24b
131–135
166–170
192–193
135–136
181–183
212–214
C19H30N2O2ꢄC4H4O4
C19H30N2OꢄHCl
2922, 1654, 1514, 1249, 1125, 836
3351, 1673, 841
(CDCl3) 9.99 (s, 1H), 8.76 (s, 2H), 7.09 (d, J=9, 2H), 6.91 (d, J=9, 2H), 6.21 (s, 2H), 3.94
(t, J=7, 2H), 3.38–3.04 (m, 10H), 2.98–2.90 (m, 1H), 2.00–1.20 (m, 12H).
10.21 (s, 1H), 9.11 (s, 1H), 8.90 (s, 1H), 7.50 (d, J=8.5, 2H), 7.11 (d, J=8.5, 2H), 2.87
(t, J=7, 2H), 1.85–1.50 (m, 2H), 1.40–1.03 (m, 8H), 0.87–0.83 (m, 3H).
C18H28N2O2ꢄHClꢄH2O
C18H27NO3ꢄHClꢄH2O
C18H28N2O2ꢄHClꢄ2H2O
C18H27NO3ꢄHClꢄH2O
3435, 1654, 1240, 1027, 831
3449, 1767, 1199, 827
10.55 (s, 1H), 9.15 (s, 1H), 8.90 (s, 1H), 7.50 (d, J=9, 2H), 6.89 (d, J=9, 2H), 3.90 (t, J=6.5, 2H), 3.31–3.21
(m, 1H), 3.00–2.92 (m, 1H), 2.22–2.18 (m, 1H), 1.83–1.27 (m, 14H), 0.88–0.83 (m, 3H)
9.58 (s, 2H), 7.10 (d, J=9, 2H), 6.97 (d, J=9, 2H), 4.29–4.27 (m, 1H), 3.94 (t, J=6, 2H), 3.38–3.26 (m, 1H),
2.97–2.93 (m, 1H), 2.24–2.20 (m, 1H), 1.80–1.61 (m, 7H), 1.39–1.28 (m, 6H), 0.50 (t, J=7, 3H).
10.55 (s, 1H), 9.20 (s, 1H), 8.86 (s, 1H), 7.47 (d, J=9, 2H), 6.84 (d, J=9, 2H), 3.88 (t, J=6, 2H), 3.37–3.23
(m, 1H), 3.00–2.82 (m, 1H), 2.30–2.18 (m, 1H), 1.94–1.61 (m, 14H), 0.85 (t, J=7, 3H).
9.58 (s, 2H), 7.10 (d, J=9, 2H), 6.97 (d, J=9, 2H), 4.30–4.28 (m, 1H), 3.94 (t, J=6, 2H), 3.38–3.26 (m, 1H),
2.96–2.92 (m, 1H), 2.24–2.20 (m, 1H), 1.80–1.61 (m, 7H), 1.38–1.28 (m, 6H), 0.88–0.85 (m, 3H).
3458, 1653, 1238, 831
3298, 1754, 1250, 1170, 841
aOnly the most significant absorption bands have been reported.
bAll spectra have been recorded in DMSO-d6, ifnot differently indicated. Chemical shifts are expressed in d values; coupling constants are expressed in Hz; exchange with D2O was used to identify NH
protons. Only value ofNH, aromatic and signals ofsignificant protons are reported in the case ofnipecotic acid derivatives; the multiplets ofthe pip eridine protons were found within the following d
ranges: 3.60–2.50 (m, 5H) and 2.80–1.30 (m, 4H) ppm.

1444
M. de Candia et al. / Bioorg. Med. Chem. 11 (2003) 1439–1450
QSARs with the cross-validated (leave-one-out proce-
dure, loo) parameter q2>0.5, the following two-variable
equation was obtained, which explains about 75% of
the variance in the biological data:
the Gollamudi data set (i.e., amides and esters instead of
amides only, mono- instead ofbis-amides, aryl instead
ofalkyl amide moieties), did not incorporate a quad-
ratic log P term, even though we could not exclude that
the correlation between pIC50 and log P is a non-linear
one for our molecular data set too. In fact, once a log P
value ofca. 4.2 has been reached with the 4-hexyloxy
nipecotanilide 18a, further increments of lipophilicity
due, for example, to methylation of the amide NH
(19a), more lipophilic substituent (20a), or replace-
ment ofamide with ester group ( 18b), resulted in a
2-fold decrease of activity against ADP-induced PRP
aggregation.
pIC50 ¼0:17ðÆ0:05Þlog P À 0:46ðÆ0:21ÞꢀR þ 3:05ðÆ0:16Þ
n ¼ 14; r2 ¼ 0:7390; q2 ¼ 0:6374; s ¼ 0:1518; F ¼ 15:57
ð1Þ
where n represents the number ofdata points, r2 the
squared correlation coefficient, q2 the loo cross-vali-
dation coefficient (as a figure ofthe ‘internal’ predictive
ability ofthe QSAR model), s the standard deviation of
the regression equation, F the statistical significance of
fit; 95% confidence intervals ofthe regression coeffi-
cients are given in parentheses.
The significant contribution ofthe resonance sigma
constant in eq 1 could be explained in the light ofargu-
ments brought up in the literature.17 It had been hypo-
thesized that a key interaction causing stabilization of
platelet membrane would be the binding ofnipecota-
mides 1 to membrane phosphoinositides, and computa-
tional evidence had been reported that highlighted H
bond formation between the 3-OH group of the inositol
portion ofphosphoinositides and the amide oxygen of
nipecotamides as a HB acceptor. The strength ofthis
HB should increase with increasing electronegativity of
the carbonyl oxygen. Such an argument was used to
explain the better activities of N,N-diethylamides
(CONEt2) over ethyl esters (COOEt) and methylketones
(COMe).17,40 Alkyloxy substituents, such as the hexyl-
oxy group in our set, due to their mesomeric electron-
donor properties, are those that render phenyl amides
more similar to alkyl amides from the electronic prop-
erty point ofview, their electronic parameters ofreso-
nance being superimposable (sR are ꢀ0.15 and ꢀ0.14
for alkyloxyphenyl and alkyl groups, respectively).
Within the series examined in this study, the importance
ofthe electron-donor mesomeric effects is demonstrated
by a number ofpairwise comparisons. Thus, of r exam-
ple, 4-methyl anilide (11a), despite its lower lipophili-
city, is 1.4-fold more active than the 4-trifluoromethyl
congener (9a).
In order to assess the relative contribution ofeach
independent variable, we computed regression on the
39
matrix ofautoscaled data,
standardized coefficients:
obtaining the following
À
Á
0
pIC50 ¼ 0:57ðlog PÞ0À0:41ðꢀRÞ0
ð2Þ
Eq 2 demonstrated that hydrophobic interactions, as
accounted for by log P term, play a dominant role in
modulating the inhibitory activity ofADP-induced PRP
aggregation ofphenyl esters and amides ofnipecotic acid.
Besides log P, a significant importance can be ascribed to
the electronic properties ofthe substituents in the para
position ofthe phenyl ring; the electron-donating groups,
especially due to mesomeric effects (a statistically poorer
equation was indeed obtained with s instead of sR),
increase the antiplatelet activity ofnipecotoyl deriva-
tives. Replacing log P with bulkiness parameters MR
and MV resulted in equations with lower predictive
abilities (q2=0.45 and 0.35 for the equations with MR
and MV, respectively).
It is worthy to point out that eq 1 holds fairly well only
for derivatives bearing substituents in the para position
ofthe phenyl ring and having log P values higher than a
minimum (ca. 2). In fact, not only nipecotamides 2a
(X=H) and 13a (X=4-OCH3) behaved as outliers, but
based on eq 1 the IC50 value ofa compound like
16a, bearing the strong electron-donor group (sR of
4-N(CH3)2 is ꢀ0.88), should be around 200 mM,
whereas instead it is a much weaker inhibitor, most
likely because ofits lower-than-minimum lipophilicity.
In summary, our data showed that, while the presence
ofa double amonium head should not be an essential
requisite, optimal lipophilicity and electron-donor
properties are major properties eliciting antiplatelet
activity from piperidine-3-carboxamides. Based on a
structural analogy criterion, it is conceivable that nipe-
cotanilides may act through a mechanism similar to that
ofthe bis-nipecotamides 1. Therefore, using molecular
modeling methods we examined the interaction ofthe
most active derivative 18a with phosphatidylinositol
(PI), a major lipid component ofthe platelet membrane
inner layer, taken as the main platelet target site.41ꢀ43
Molecular models ofthe ligate PI and ligand 18a (R and
S configurations) were built using the SYBYL frag-
mental library,44 and the aliphatic chains in both
structures fixed in extended conformations. Flexible
A dependence ofthe platelet aggregation inhibitory activ-
ity upon the hydrophobicity ofnipecotoyl derivatives was
somewhat expected based on the QSAR ofbis-nipeco-
tamides 1.15 A statistically significant parabolic rela-
tionship between activity and log P had been obtained
by regression analysis carried out on a series ofcom-
pounds 1 bearing different alkyl groups on the amide
nitrogen, and the optimum log P value calculated
around 4.5. Eq 1, which is based on a set ofpiper-
idinecarboxylic acid derivatives different from those of
docking calculations were performed using the DYN-
45
DOCK algorithm ofQXP sotfware package,
using
as starting geometries the minimum-energy con-
formations as found by molecular dynamics
(AMBER/MM2 force field, 1000 cycles, 300 K, time

M. de Candia et al. / Bioorg. Med. Chem. 11 (2003) 1439–1450
1445
Figure 2. Concentration–response curves in logarithmic scale ofcompounds 18a (top) and 18b (bottom) on the platelet aggregation induced by ADP
step: 3 fs). Figure 4 shows the lowest-energy docked
structures in the most populated cluster, and GRID
contours46 (calculated with water probe) highlight
hydrophobic (yellow) and polar (blue) interactions.
The calculated binding energy (ꢀ10.1 and ꢀ8.1 kcal/
mol for R-18a and S-18a, respectively) indicated that
nipecotamide 18a (especially in R-configuration),
similarly to bis-nipecotamide 1a, may form a stable
complex with PI.
In an additional QSAR study, we used the same inde-
pendent variables of eq 1 in the regression analysis of
another data set taken from literature,17 that is bis-nipe-
cotamides 1 with varying aromatic bridges connecting
the two nipecotamide moieties (e.g., substituted-p-xylene
connecting bridge, structure 1 in Fig. 1). The ‘best’ two-
parameter equation and the corresponding standardized
regression coefficients39 obtained by the stepwise
regression analysis are as follows:

1446
M. de Candia et al. / Bioorg. Med. Chem. 11 (2003) 1439–1450
Figure 3. Plot ofpIC 50 against log P ofpiperidin-3-carboxylic acid derivatives (symbols: ^ anilides, & esters).
and aromatic electron density have to be considered the
main forces involved in the molecular recognition of
nipecotamides at the platelet target site.
pIC50 ¼ 0:44ðÆ0:08Þlog P À 0:58ðÆ0:26Þꢀ þ 2:12ðÆ0:32Þ
n ¼ 15; r2 ¼ 0:8588; q2 ¼ 0:7969; s ¼ 0:2940; F ¼ 36:49
ð3Þ
We finally wanted to compare activities ofthe nipeco-
toyl p-hexyloxyphenyl derivatives 18a and 18b with the
isomeric pipecolinic (23a and 23b) and isonipecotic (24a
and 24b) acid derivatives, and the results are summar-
ized in Table 2.
À
Á
0
pIC50 ¼ 0:72ðlog PÞ0À0:30ðꢀÞ0
ð4Þ
Eqs 3 and 4, which explain about 86% ofthe variance in
the biological data, with a good predictive ability (q2 of
ca. 0.80), demonstrate that antiplatelet activity is pri-
marily correlated with the lipophilicity ofthe connecting
bridge, as graphically shown in Figure 5.
Isonipecotamide 24a was about 1/3 as active as its iso-
mer nipecotamide 18a in inhibiting the ADP-induced
(and not the adrenaline-induced) PRP aggregation.
Likely, once again lipophilicity may account for such a
different antiplatelet properties, since compound 24a is
significantly less lipophilic than 18a. In contrast, a dif-
ference in lipophilicity is highly unlikely to account
alone for the observed difference in activity between
nipecotamide 18a and pipecolinamide 23a (their lipo-
philicity differs in just 0.3 log P units).
Moreover, the significant contribution of s (slightly
poorer statistics were obtained using sR in this case),
that multiplies a negative coefficient, proved the
increased activity to be also related to an increase in
electron density ofthe aromatic bridge. The worst-fit-
ting compound was the 20-amino derivative 1b; it was
indeed omitted from the regression eq 3.
In Table 2, the data ofstability determined in buffered
aqueous solution containing human serum could
account for variations in activity of the phenyl esters.
Nipecotate 18b and isonipecotate 24b were found to be
stable enough during the time ofthe aggregation test
(the hydrolyzed fraction was less than 15%), whereas
pipecolinate 23b underwent a fast cleavage in human
serum solution at 37 ꢂC, its half-life being less than 4
min. The fact that pipecolinic acid ester 23b, similar to
a-amino acid esters or related short-chained aliphatic
amino acid esters,47 is less resistant than the isomers 18b
and 24b to the chemical (results not shown) and serum
esterase-catalyzed hydrolysis can be a consequence of
either the electron withdrawing effect ofthe protonated
amino group, which activates the ester linkage toward
OH- attack, and (predominantly) the intramolecular
Eq 3 derived in our present study is much better, from a
statistical viewpoint, than the QSAR reported by Golla-
mudi and coworkers in the original paper, which involved
MV, MV2 and sigma as the independent variables. We
recomputed that regression equation, using the cross-
validation procedure as a diagnosis ofits prediction ability,
and found a negative q2 (i.e., no model at all) for the whole
set of16 compounds and q2 of0.576 after omitting com-
pound 1a from the data set, as did the authors, which is
one ofthe most active derivatives within the series. In
contrast, eq 3 showed statistics ofsome interest even when
applied to the whole data set (n=16, r2=0.715, q2=0.548).
Eqs 3 and 4, quite similar in their physicochemical
meaning to eqs 1 and 2, showed that hydrophobicity

M. de Candia et al. / Bioorg. Med. Chem. 11 (2003) 1439–1450
1447
catalysis by the same neighboring protonated amino
group that promotes ester cleavage. Due to its poor
stability in plasma solution, pipecolinate 23b did not-
show reproducible effects in the PRP aggregation
assay.
Conclusions
A number ofsimple phenyl amides ofpiperidine-3-car-
boxylic acid (nipecotanilides), designed with the aim of
simplifying the a,a0-bis[3-(N,N-dialkylcarbamoyl)piper-
idino]-p-xylenes (1) reported by others as potential
antithrombotic agents,15,17 exhibited appreciable and
dose-dependent inhibition activities on platelet aggre-
gation induced by ADP and adrenaline. Lipophilicity
and presence ofelectron-donating groups in the para
position ofthe phenyl ring were detected as the proper-
ties mainly responsible for the increase in activity. A log
P value of4.2 appeared desirable ofr optimum PRP
aggregation inhibition within the examined property
space, and N-[4-(hexyloxy)phenyl]piperidine-3-carbox-
amide hydrochloride (18a) was found as active as
aspirin in inhibiting adrenaline-induced platelet aggre-
gation. Docking calculations provided computational
support to complex formation between the nipecotoyl
anilide 18a and phosphatidylinositol, a major compo-
nent ofthe inner layer ofthe platelet membranes. This
Figure 4. View ofthe complex R-18a/PI as resulted from flexible
docking calculations performed with DYNDOCK algorithm of the
QXP software.45 The lowest-energy docked structures ofthe ligand, R-
18a, and the ligate, PI, in the most populated cluster are represented
(calculated binding energy: ꢀ10.1 kcal/mol). The molecular interac-
tion fields are visualized as favorable H-bonding regions (blue-colored,
contour level: ꢀ3.0 kcal/mol) and hydrophobic interaction regions
(yellow-colored, contour level: +1.5 kcal/mol) derived from GRID
calculations46 with a water probe.
interaction may be considered as a key event triggering
24,25
the antiplatelet action ofnipecotamides.
Work is in
progress to evaluate the effects ofother modifications of
the chemical structure ofnipecotamides and to gain
more information into their mechanism of action.
Figure 5. Plot ofpIC 50 against Clog P of a,a0-bis[3-(N,N-dialkylcarbamoyl)piperidino]-p-xylene analogues (Gollamudi data set).17

1448
M. de Candia et al. / Bioorg. Med. Chem. 11 (2003) 1439–1450
Experimental
acid (1.0 g, 4.60 mmol) and triethylamine (1.2 mL, 8.80
mmol) in 40 mL ofdichloromethane, and the mixture
heated to 90 ꢂC for 1 h. The solution was cooled to 0–
5 ꢂC and 4-nitro aniline (0.6 g, 4.36 mmol) in 30 mL of
methylene chloride was added dropwise, then the reac-
tion mixture was allowed to stir at rt overnight. Evap-
oration ofthe solvent under reduced pressure yielded a
solid residue, which was purified by silica gel column
chromatography (mobile phase: chloroform/methanol,
98:2 v/v) to give 0.70 g ofthe N-BOC-protected product
(46% yield), which was subsequently deprotected to
give, after crystallization from ethanol–ethyl acetate, 0.4
g of 7a (32% yield).
Chemistry
Melting points were determined in open capillary tubes
by SMP3 Stuart Scientific melting point apparatus and
are uncorrected. Infrared absorption spectra were
recorded as KBr pellets using a Perkin-Elmer Spectrum
One FT IR instrument (Perkin-Elmer Ltd, Buck-
inghamshire, UK). 1H NMR spectra were recorded on a
Varian 300 MHz instrument. Chemical shifts are
expressed in d values and the coupling constants in Hz;
exchange with D2O was used to identify NH pro-
tons. The following abbreviations have been used: s,
singlet; d, doublet; t, triplet; m, multiplet(s). Ele-
mental analysis were performed by the Analytical
Laboratory Service ofthe Dipartimento Farmaco-chi-
mico ofUniversity ofBari on Euro EA3000 analyzer
(Eurovector, Milan, Italy) and agreed with theoretical
Compound 6a was synthesized using the same proce-
dure.
N-(4-Aminophenyl)piperidine-3-carboxamide hydrochlo-
ride (8a). A suspension of0.5 g of 7a (1.75 mmol) and
0.15 g of10% Pd/C in 30 mL ofethanol was maintained
for 8 h at rt under hydrogen atmosphere, until TLC
analysis revealed the complete disappearance ofthe
starting material. The reaction mixture was filtered on
Celite pad and the catalyst was washed with ethanol.
The filtrate was evaporated under reduced pressure to
give 0.45 g (98% yield) ofthe N-BOC-protected deriva-
tive 8a, which was then deprotected with HCl gas to
give 0.21 g of 8a (51% yield).
1
values to within ꢁ0.40%. IR, H NMR and elemental
analysis data were consistent with the reported
structures (Table 3). Chemicals were purchased from
Aldrich Chemical Company (Milwaukee, WI). Silica
gel 60 (Merck 70–230 mesh) was used for column
chromatography.
General procedure for the synthesis of N-phenylpiperi-
dine carboxamides. As an example, the synthesis of
N-phenylpiperidine-3-carboxamide hydrochloride (2a)
is reported in detail. 1-(t-Butyloxycarbonyl)piperidine-
3-carboxylic acid was prepared in 85% yield following
reported method;32 mp 148–150 ꢂC (lit. Mp 149–
151 ꢂC).
N-{4-[(Methylsulfonyl)amino]phenyl}piperidine-3-car-
boxamide hydrochloride (14a). Methanesulfonylchloride
(0.1 mL, 1.3 mmol) was added dropwise to a stirred
solution of 8a as N-BOC-derivative (0.3 g, 1.0 mmol) in
10 mL ofdry dioxane and 1 mL ofdry pyridine, upon
cooling to 0–5 ꢂC. The mixture was refluxed for 1 h, and
then poured into ice-cold 2 N HCl. The resulting solu-
tion was extracted with chloroform (3ꢄ10 mL). The
extracts were combined, dried (Na2SO4), and evapo-
rated under reduced pressure to yield an oil, which
solidified upon trituration with diethyl ether to give 0.25
g (67% yield) of N-BOC-protected product. Subsequent
removal ofthe BOC-protecting group with HCl gas
yielded 0.12 g of 14a (38% yield).
N-BOC-nipecotic acid (1.0 g, 4.36 mmol), hydrated N-
hydroxy-benzotriazole hydrated (0.6 g, 4.36 mmol) and
dicyclohexylcarbodiimmide (0.9 g, 4.36 mmol) were
suspended in 15 mL ofdry dichloromethane. After stir-
ring for 15 min, aniline (0.4 mL, 4.36 mmol) was added
in one portion and the mixture stirred at rt for 10 h.
Solid dicyclohexylurea (DCU) was filtered off, and,
after removal of the solvent, the residue was dissolved in
50 mL ofethyl acetate and washed with 1 N HCl (3 ꢄ20
mL), 5% NaHCO3 (3ꢄ20 mL), and finally with water
(3ꢄ20 mL). The organic phase was dried (Na2SO4) and
evaporated under reduced pressure to yield generally a
solid upon trituration with diethyl ether. Following
purification ofthe solid residue by silica gel column
chromatography (mobile phase: petroleum ether/ethyl
acetate, 70:30 v/v), the product (0.80 g, 60% yield) was
subjected to the subsequent reaction for the removal of
the N-BOC-protecting group.
N-[4-(Dimethylamino)phenyl]piperidine-3-carboxamide
hydrochloride (16a). A solution ofthe N-BOC deriva-
tive of 8a (0.3 g, 1.0 mmol) and 40% aqueous formal-
dehyde (0.14 mL, 1.9 mmol) in 20 mL ofethanol was
stirred at rt temperature under hydrogen atmosphere in
the presence of10% Pd/C (0.1 g) of r 6 h. The reaction
mixture was filtered on a Celite pad and the catalyst
washed with ethanol. The combined filtrate, after con-
centration under reduced pressure, gave an oil which
solidified upon trituration with diethyl ether. The solid
residue was purified by silica gel column chromato-
graphy (mobile phase: petroleum ether/ethyl acetate,
70:30 v/v) to yield 0.3 g ofthe N-BOC-protected pro-
duct, which gave 0.12 g of 16a (38% yield) after the
BOC-deprotection reaction with HCl gas.
A solution of1 mmol ofthe N-BOC-protected nipeco-
tanilide in 100 mL ofchloroof rm was cooled to 0–5 ꢂC
and saturated with HCl gas. The solvent was removed
under reduced pressure to yield a solid. Recristallyza-
tion from ethanol-ethyl acetate yielded the desired N-
phenylpiperidinecarboxamide as hydrochloride salt.
The total yields were in the range 30–40%.
N-(4-Nitrophenyl)piperidine-3-carboxamide hydrochloride
(7a). Thionyl chloride (0.70 mL, 8.8 mmol) was added
dropwise to a cold stirred mixture of N-BOC-nipecotic-
N-[4-(Hexyloxy)phenyl]-N-methylamine. A suspension
of4-hexyxloxy-aniline (1g, 5.2 mmol) in 20 mL ofetha-
nol containing 40% aqueous formaldehyde (0.4 mL, 5.2

M. de Candia et al. / Bioorg. Med. Chem. 11 (2003) 1439–1450
1449
mmol) and 10% Pd/C (0.2 g) was stirred at rt under
hydrogen atmosphere for 16 h. The reaction mixture
was filtered on a Celite pad and the catalyst washed with
ethanol. The combined filtrate was evaporated under
reduced pressure to give an oil residue, which was pur-
ified by silica gel column chromatography (mobile
phase: petroleum ether/ethyl acetate, 70:30 v/v) to
afford 0.60 g ofthe product (56% yield).
accomplished by measuring peak areas in relation to
those ofexternal standards.
For comparison, hydrolysis reactions were studied also
in the absence ofhuman serum. Pseudo-first-order rate
constants for the hydrolysis were determined from the
slopes oflinear plots ofthe logarithm ofthe residual
piperidine carboxylic ester against time.
N-(4-hexyloxyphenyl)-N-methylpiperidine-3-carboxamide
maleate (19a). N-BOC-protected nipecotic-acid (1.00 g,
4.36 mmol), hydrated N-hydroxy-benzotriazole (0.6 g,
4.36 mmol) and dicyclohexylcarbodiimmide (0.9 g, 4.36
mmol) were suspended in 20 mL ofdry dichloro-
methane. After stirring for 15 min, N-methyl-4-hexyl-
oxyaniline (0.9 g, 4.36 mmol) in 5 mL ofdry
dichloromethane was added dropwise. After stirring
overnight at rt, the solid dicyclohexylurea was filtered
off, and the solvent removed by evaporation under
reduced pressure. The residue was dissolved in 50 mL of
ethyl acetate and washed with 1N HCl (3ꢄ20 mL), 5%
NaHCO3 solution (3ꢄ20 mL), and finally with water
(3ꢄ20 mL). The ethyl acetate solution was dried
over Na2SO4, and evaporated under reduced pres-
sure to yield an oil, which solidified upon tritura-
tion with diethyl ether. The solid residue was then
purified by silica gel column chromatography (mobile
phase: petroleum ether/ethyl acetate, 80:20 v/v) and the
obtained product BOC-deprotected with HCl gas. The
free base, obtained by treatment with NaHCO3 and
extraction with chloroform, was dissolved in ethyl ace-
tate and added dropwise with an equimolar solution of
maleic acid in isopropyl alcohol dropwise. Upon cool-
ing 19a was obtained as maleate salt (0.30 g, 16%
yield).
Measurements of inhibitory activity of platelet
aggregation
The effects on the in vitro aggregation ofhuman plate-
let-rich plasma (PRP) induced by ADP and adrenaline
were determined by the turbidimetric method ofBorn, 35
as described by Quintana et al.,19,30 using a four-chan-
nel aggregometer PACKS-4 (Helena Laboratories,
Beaumont, TX, USA).
All blood and blood products were handled in plastic
ware, except for siliconized glass cuvettes and silico-
nized stir bars. With the sole exception ofcompound
19a prepared as maleate salt, all the test compounds
were prepared as hydrochlorides and dissolved in redis-
tilled water containing dimethylsulfoxide (0.5% v/v final
concentration). Solutions ofthe PRP aggregation indu-
cers were prepared by dissolving ADP sodium salt or
adrenaline in Tyrode’s buffer (NaCl 137 mM, NaHCO3
11.9 mM, KCl 2.7 mM, NaH2PO4 0.36 mM, glucose
560 mM; pH 7.4 by dropwise addition of1 N HCl).
Human blood was obtained from healty volunteers who
had not ingested any known antiplatelet drugs within
three days prior to the blood drawing. Platelet-rich
plasma (PRP) was obtained after centrifugation of
venous blood (9 mL), mixed with 0.129 mol/L sodium
citrate (1:9 to blood) to prevent it from clotting, at 100g
for 10 min at 23 ꢂC. PRP platelet numbers were counted
Hydrolysis catalyzed by human serum
and found as within the range 150,000–320,000/mLꢀ1
.
The susceptibility ofthe ester derivatives 18b, 23b and
24b, and the corresponding amides toward hydrolysis
catalyzed by human serum esterases was measured in
phosphate buffer (40 mM, pH 7.4; ionic strength of0.5
was maintained by adding a calculated amount ofKCl)
containing 10% ofhuman serum at 37 ꢁ0.2 ꢂC. Each
reaction were initiated by adding 100 mL ofthe metha-
nolic stock solution ofcompound under examination to
1.9 mL ofpreheated serum solution (final concentration
about 1ꢄ10ꢀ3 M) and the mixture was maintained in
water bath at 37 ꢂC. At appropriate times, 100 mL sam-
ples were withdrawn and added to 500 mL ofcold ace-
tonitrile in order to deproteinize the serum. After
mixing and centrifugation (10 min at 4000 rpm), 5 mL of
the clear supernatant were filtered through 0.2 mm
membrane filter (Waters, PTFE 0.2 mm) and analyzed
by HPLC. HPLC analysis was carried out on a Waters
1525 apparatus (constant-flow binary pump equipped
with Waters 2487 variable wavelength detector). A
Symmetry C18 column, 150ꢄ4.6 mm i.d. (Waters,
Milan, Italy), was used as the stationary phase, and
suitable water/MeOH mixtures were used as the mobile
phases. A constant flow rate of1.0 mL/min was main-
tained and the column effluent was continuously mon-
itored at 254 nm. Quantification ofthe compounds was
PRP (250 mL) was pre-incubated with the test com-
pound solutions (5 mL), or with dimethylsulfoxide
(0.5% v/v, control) at 37 ꢂC for 5 min, during which the
suspension was stirred at 900 rpm. Then the inducer, 50
mL Tyrode’s solution containing ADP (10 mM) or adre-
naline (10 mM), was added to the stirred sample and the
change in trasmittance at 640 nM was recorded for 10
min. The control cuvette containing the vehicle treated
PRP followed the same sequence of events.
Inhibition ofaggregation was expressed as a percent of
the maximum response, and antiplatelet activity data
were reported as meansꢁSEM (n=3–6). A preliminary
antiplatelet screen was obtained by measuring the effect
ofeach compound at 100 mM concentration on ADP-
induced aggregation. For compounds showing inhibi-
tory activities significantly different from the respective
control value, dose-response relations were determined
in a concentration interval ranging from 12.5 to 400 mM
and IC50 (concentration effecting 50% inhibition of
aggregation) was obtained by regression ofaggregation
inhibition on log concentration oftest compound.
Aspirin was used a reference standard in the adrenaline-
induced PRP aggregation assay.

1450
M. de Candia et al. / Bioorg. Med. Chem. 11 (2003) 1439–1450
Supplementary Table. Elemental analysis data ofthe
newly synthesized compounds
16. Lawrence, W. P.; Howell, R. D.; Gollamudi, R. J. Pharm.
Sci. 1994, 83, 222.
17. Guo, Z.; Zheng, X.; Thompson, W.; Dugdale, M.; Golla-
mudi, R. Bioorg. Med. Chem. 2000, 8, 1041.
18. Lasslo, A.; Quintana, R. P.; Johnson, R. W.; Naylor, J. L.;
Dugdale, M. Biochim. Biophys. Acta 1984, 772, 84.
19. Quintana, R. P.; Lasslo, A.; Dugdale, M.; Goodin, L. L.
Thromb. Res. 1981, 22, 665.
Elemental analysis (C, H, N,)
Compd Molecular formula
Calculated
Found
C
C
H
N
H
N
20. Lasslo, A.; Dillingham, E. O.; Gollamudi, R.; Carter-
Burks, G. Biochim. Biophys. Acta 1987, 923, 443.
21. Lasslo, A.; Dillingham, E. O.; Gollamudi, R.; Carter-
Burks, G. Res. Commun. Chem. Pathol. Phrmacol. 1987, 58,
129.
22. Quintana, R. P.; Lasslo, A.; Greer, T. Thromb. Res. 1981,
24, 379.
23. Bevers, E.M.; Rosing, J.; Zwaal, R.F.A. In Platelets in
Biology and Pathology—III; MacIntyre, D.E., Gordon, J.L.
Eds.; Elsevier: Amsterdam, 1987; p 135.
24. (a) Gollamudi, R.; Dillingham, E. O.; Bond, S. E.
Thromb. Haemost. 1993, 69, 1322. (b) Lasslo, A.; Quintana,
R. P.; Crisan, D.; Baier, R. E.; Mayer, A. E.; Fornalik, M. S.
Med. Prog. Technol. 1984, 10, 71.
25. Lasslo, A. Fed. Proc. 1984, 43, 1382.
26. Gollamudi, R.; Dillingham, E. O.; Bond, S. E. Blood.
Coag. Fibrinol. 1994, 5, 375.
27. Handa, R. K.; Dillingham, E. O.; Gollamudi, R. Life Sci.
1995, 57, 983.
28. Zheng, X.; Salgia, S. R.; Thompson, W. B.; Dillingham,
E. O.; Bond, S. E.; Feng, Z.; Prasad, K. M.; Gollamudi, R. J.
Med. Chem. 1995, 38, 180.
3a
4a
5a
6a
7a
8a
9a
10a
11a
12a
13a
14a
15a
16a
17a
18a
19a
20a
23a
23b
24a
24b
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
12H15N2OBrꢄHCl
12H15N2OBrꢄHCl
12H15N2OClꢄHCl
12H15N3O3ꢄHCl
12H15N3O3ꢄHCl
45.09 5.05 8.76 45.30 5.03 8.83
45.09 5.05 8.76 45.33 5.19 8.83
52.36 5.86 10.18 52.44 5.98 10.32
50.44 5.64 14.71 50.40 5.60 14.70
50.44 5.64 14.71 50.66 5.82 14.71
12H17N3Oꢄ2HClꢄH2O 46.46 6.82 13.55 46.56 6.81 13.59
13H15N2OF3ꢄHCl
13H15N3OꢄHCl
13H18N2OꢄHCl
13H18N2O2ꢄHCl
13H18N2O2ꢄHCl
13H19N3O3SꢄHCl
14H18N2O2ꢄHCl
50.58 5.22 9.07 50.65 5.46 9.25
58.76 6.07 15.81 59.01 6.25 15.88
61.29 7.52 11.00 61.30 7.70 11.21
57.67 7.07 10.35 57.68 7.01 10.38
57.67 7.07 10.35 57.83 7.34 10.40
46.77 6.04 12.58 46.90 6.21 12.70
59.47 6.77 9.91 60.09 7.02 10.02
14H21N3Oꢄ2HClꢄH2O 49.71 7.45 12.42 49.92 7.50 12.48
16H24N2OꢄHClꢄH2O
18H28N2O2ꢄHCl
61.04 8.64 8.90 61.23 8.64 8.82
63.42 8.57 8.22 63.78 8.59 8.44
63.57 7.89 6.45 63.67 8.33 6.43
67.33 9.22 8.27 67.47 9.35 8.45
19H30N2O2ꢄC4H4O4
19H30N2OꢄHCl
18H28N2O2ꢄHClꢄH2O 60.24 8.71 7.80 60.47 8.77 8.09
18H27NO3ꢄHClꢄH2O
18H28N2O2ꢄHClꢄ2H2O 57.36 8.82 7.43 57.60 8.86 7.56
18H27NO3ꢄHClꢄH2O 60.07 8.40 3.89 60.21 8.66 4.01
60.07 8.40 3.89 60.19 8.36 4.00
29. Dillingham, E. O.; Lasslo, A.; Carter-Burks, G.; Bond,
S. E.; Gollamudi, R. Biochim. Biophys. Acta 1989, 990, 128.
30. Altomare, C.; Summo, L.; Varlamov, A. V.; Voskres-
sensky, L. G.; Borisova, T. N.; Carotti, A. Bioorg. Med. Chem.
Lett. 2000, 10, 581.
31. Voskressensky, L.G.; de Candia, M.; Carotti, A.; Bor-
isova, T.N.; Kulikova, L.N.; Valamov, A.V.; Altomare, C. J.
Pharm. Pharmacol. in press.
32. Bonina, F. P.; Arenare, L.; Palagiano, F.; Saija, A.; Nava,
F.; Trombetta, D.; De Caprariis, P. J. Pharm. Sci. 1999, 88, 561.
33. Altomare, C.; Carotti, A.; Cellamare, S.; Ferappi, M.;
Cagiano, R.; Renna, G. Int. J. Pharm. 1988, 48, 91.
34. Hellberg, M.; Stubbins, J. F.; Glennon, R. A. Bioorg.
Med. Chem. 2000, 8, 1917.
Acknowledgements
Part ofthis study was financially supported by Italian
Ministry ofUniversities and Research (MIUR, Rome,
Italy). The expert assistance ofDr. M. Micelli (Labor-
atorio di coagulazione, Policlinico Bari) in antiplatelet
testing is gratefully acknowledged.
References and Notes
1. Jang, I.K.; Fuster, V. In Thrombolitic Therapy; Comerota,
A. J. Ed.; J.B. Lippincott Co.: PA, 1995; p 17.
2. Majerus, W.; Broza, G.J.; Miletich, J.P.; Tollesfsen, D.M.
In The Goodman Gilman’s Pharmacological Basis of Thera-
peutics; Goodman, L. S., Gilman, A. G., Limbird, L. E. Eds.;
McGraw-Hill: New York, 1996; p 1341.
3. Lam, J. Y. T.; Chesebro, J. H.; Steele, P. M.; Dewanjee, M. K.;
Badimon, L.; Fuster, V. J. Am. Coll. Cardiol. 1986, 8, 1380.
4. Edmunds, L. H. Ann. Thorac. Surg. 1987, 44, 430.
5. Solymoss, B. C.; Nadeau, P.; Millette, D.; Campeau, L.
Circulation 1988, 78 (Suppl I), 140.
6. Fuster, V.; Pumphrey, C. W.; McGoon, M. D.; Chesebro, J. H.;
Pluth, J. R.; McGoon, D. C. Circulation 1982, 66 (Suppl I), 157.
7. Gachet, C. Thromb. Haemost. 2001, 86, 222.
35. Born, G. V. R. Nature 1962, 194, 927.
36. CLOG P for Windows, vers. 4.0; BioByte Corp.: Clar-
emont, CA, USA.
37. ACD/Labs package, release 5.0, Advanced Chemistry
Development Inc.: Toronto, Ontario, Canada.
38. Forina, M.; Lanteri, S.; Armanino, C. Q-PARVUS, vers.
3.0, An Extendable Package of Programs for Data Explorative
Analysis, Classification and Regression Analysis; Dipartimento
di Chimica e Tecnologie Farmaceutiche ed Alimentari: Gen-
ova, Italy.
39. Mager, H.; Barth, A. Pharmazie 1979, 34, 557.
40. Quintana, R. P.; Lasslo, A.; Dugdale, M. Biophys. J. 1982,
37, 130.
41. Shattil, S. J.; Cooper, R. A. Biochemistry 1976, 15, 4832.
42. Chap, H.; Perret, B.; Manco, G.; Simon, M. F.; Douste-
Blazy, L. Agents Actions 1979, 9, 400.
8. Lages, B.; Weiss, H. J. Thromb. Haemost. 1980, 43, 147.
9. Kunapuli, S. P. Trends Pharmacol. Sci. 1998, 19, 391.
10. Van De Graaff, E.; Steinhubl, S. R. Ann. Inter. Med. 2000,
32, 561.
43. Nozawa, Y.; Nakashima, S.; Nogata, K. Biochim. Bio-
phys. Acta 1991, 1082, 219.
11. Calverley, D. C. Clin. Geriatr. Med. 2001, 17, 31.
12. Berger, P. B. Curr. Cardiol. Rep. 1999, 1, 192.
13. Hankey, G. J. Cerebrovasc. Dis. 2001, 11 (Suppl 2), 11.
14. Schror, K. Drugs 1995, 50, 7.
44. SYBYL software, vers. 6.7; Tripos Associates: Saint
Louis, MO, USA.
45. Mc Martin, C.; Bohacek, R. S. J. Computer Aided Mol.
Des. 1997, 11, 333.
15. Feng, Z.; Gollamudi, R.; Dillingham, E. O.; Bond, S. E.;
Lyman, B. A.; Purcell, W. P.; Hill, R. J.; Korfmacher, W. A.
J. Med. Chem. 1992, 35, 2952.
46. Goodford, P. J. J. Med. Chem. 1985, 28, 849.
47. Bundgaard, H.; Larsen, C.; Thorbek, P. Int. J. Pharm.
1984, 18, 67.