2-Phenyl-3-(quinolizidin-1-yl)-5-substituted indoles as platelet antiaggregating agents

Article abstract of DOI:10.1016/j.farmac.2003.11.009

A set of ten 2-phenyl-3-(quinolizidin-1-yl)-5-substituted indoles was prepared through the Fischer cyclization of lupinyl- and epi- lupinylphenylketone 4-substituted phenylhydrazones.Compounds were tested for antiaggregating activity on human platelets activated by adenosine diphosphate (ADP), collagen and adrenaline. At 2.5 × 10^-4 M concentration most compounds strongly inhibited the aggregation induced by all the agonists considered and many of them still displayed good activity at 0.625 × 10^-4 M concentration. The least active (1c) and one of the most active (1d) compounds were also tested for antiaggregating activity on rabbit platelets activated by ADP, PAF and sodium arachidonate. Both the compounds were active against ADP and PAF, but only 1d inhibited the arachidonate-induced aggregation (100% at 8 × 10^-6 M concentration) and increased the bleeding time in mice. The same compounds were subjected to a general pharmacological screening and found to display several activities; of particular interest was the dose dependent reduction of serum cholesterol and heparin precipitating betalipoproteins in hypercholesterolemic mice exerted by 1c, which was still significant at the oral dose of 10 mg/kg.

Full text of DOI:10.1016/j.farmac.2003.11.009

IL FARMACO 59 (2004) 101–109
www.elsevier.com/locate/farmac
2-Phenyl-3-(quinolizidin-1-yl)-5-substituted indoles as platelet
antiaggregating agents
Marcella Ercoli a, Lorenzo Mina a, Caterina Canu Boido a, Vito Boido a,*, Fabio Sparatore a,
Ugo Armani b, Antonietta Piana b
a Dipartimento di Scienze Farmaceutiche, Università di Genova, Viale Benedetto XV, 3, Genova 16132, Italy
b Centro di ricerca sulla trombosi e dell’aterosclerosi, DIMI, Università di Genova, Viale Benedetto XV, 6, Genova 16132, Italy
Received 2 July 2003; accepted 14 November 2003
Abstract
A set of ten 2-phenyl-3-(quinolizidin-1-yl)-5-substituted indoles was prepared through the Fischer cyclization of lupinyl- and epi-
lupinylphenylketone 4-substituted phenylhydrazones.Compounds were tested for antiaggregating activity on human platelets activated by
adenosine diphosphate (ADP), collagen and adrenaline. At 2.5 × 10–4 M concentration most compounds strongly inhibited the aggregation
induced by all the agonists considered and many of them still displayed good activity at 0.625 × 10–4 M concentration. The least active (1c) and
one of the most active (1d) compounds were also tested for antiaggregating activity on rabbit platelets activated by ADP, PAF and sodium
arachidonate. Both the compounds were active againstADP and PAF, but only 1d inhibited the arachidonate-induced aggregation (100% at 8 ×
10–6 M concentration) and increased the bleeding time in mice. The same compounds were subjected to a general pharmacological screening
and found to display several activities; of particular interest was the dose dependent reduction of serum cholesterol and heparin precipitating
betalipoproteins in hypercholesterolemic mice exerted by 1c, which was still significant at the oral dose of 10 mg/kg.
© 2003 Elsevier SAS. All rights reserved.
Keywords: Quinolizidine derivatives; Indoles; Platelet antiaggregating activity; Hypocholesterolemic activity
1. Introduction
significant platelet antiaggregating activity was often ob-
served in in vitro assays [4–12].
Among the tested compounds, a 100% inhibition of
150 µM sodium arachidonate-induced aggregation was ob-
served in four compounds [6,8,10,12] at concentrations in
the range of 2.46–7.7 µM, while three additional compounds
exhibited the same activity at 25 µM concentration [8–10].
The best compound was S-[(2-cinnamoylamino) phenyl]
thiolupinine [5] which, on the other hand, showed only mod-
est activity against adenosine diphosphate (ADP)- and PAF-
induced aggregation, even at much higher concentrations.An
even greater selective antagonist for arachidonate vs. ADP
and PAF activation was the 3-benzyl-1-lupinyl-6-
methoxyquinoxalin-2(1H)-one [6,13].
Thromboembolic diseases are a leading cause of mortality
and disability in developed countries. Present approaches for
their treatment and/or prevention include the use of antico-
agulant, thrombolytic and antiplatelet drugs.
An optimal antiplatelet drug should block platelet-
dependent thrombus formation in a reversible manner and
regardless of the nature of the activating agent, and should be
endowed with a proper balance of antithrombotic efficacy
and maintenance of hemostasis for long-term prophylactic
treatment [1–3].
Since the above requirements are not fulfiled by any avail-
able drug, the search for new antiaggregating agents is war-
ranted.
During our long-standing study on many quinolizidine
derivatives, synthesized for different pharmacological aims,
Generally higher concentrations (in the range 200–
300 µM) were needed to antagonize the ADP (1.5 µM)-
induced aggregation, which was observed in seven com-
pounds.
The best activity was seen for 10-homolupinanoyl-2-
methoxyphenothiazine [9], which inhibited the ADP-
induced aggregation for ≈50% at 73 µM concentration; the
* Corresponding author.
E-mail address: vito.boido@unige.it (V. Boido).
© 2003 Elsevier SAS. All rights reserved.
doi:10.1016/j.farmac.2003.11.009

102
M. Ercoli et al. / IL FARMACO 59 (2004) 101–109
3 (α); 4 (β)
1 (α); 2 (β)
1
2
R=H (a); F (b); C1 (c); CH (d); OCH (e)
R=H (a); F (b); C1 (c); CH (d); OCH (e)
3
3
3
3
Scheme 1
Scheme 2
same compound was also active against PAF-induced aggre-
gation at ≈25 µM concentration.
ture of phenylhydrazone and dry zinc chloride (molar ratio
1:15) at 200 °C. In the case of 4-methoxyphenylhydrazone,
the last method gave a complex mixture of compounds, from
which the indole was not isolated. The desired indole was
finally obtained, though in poor yields, by refluxing the
absolute ethanol solution of the ketone, 4-methoxy-
phenylhydrazine hydrochloride and 85% phosphoric acid in
a molar ratio of 1:1:4.4.
The required lupinylphenylketone [1-phenyl-2-(quino-
lizidin-1a-yl)ethanone] (8) was prepared in two methods, as
outlined in Scheme 3, in all cases starting from l-lupinine,
which was extracted from cultivated Lupinus luteus L. seeds
or from wild Sardinian L. hispanicus Boiss et Reut. seeds.
The reaction of cyanolupinane with phenylmagnesium
bromide was the most profitable. Therefore, only the latter
procedure was applied to obtain the epi-lupinylphenylketone
[1-phenyl-2-(quinolizidin-1b-yl)ethanone] (12), starting
from epi-lupinine, prepared from lupinine by the action of
sodium ethoxide or sodium hydride in refluxing xylene
(scheme 4).
Nine compounds (at 20–30 µM concentrations) were ac-
tive (60–100% inhibition) against PAF (20 ng/ml)-induced
aggregation, one of which, 2-ethoxycarboamido-10-
homolupinanoylphenothiazine [9], continued to exhibit a
58% inhibition at 2.1 µM concentration.
Activities against ADP- and PAF- induced aggregation
were commonly exhibited, though at different concentra-
tions, by the same compounds while lacking activity against
arachidonate.
Therefore, the 9-(lupinylthio)xanthene [8] and the 10-
homolupinanoylphenothiazine [9] are somewhat outstanding
in exhibiting a good activity against the platelet aggregation
induced by the three tested agonists.
It is noteworthy that the introduction of different types of
substituents on position 2 of the phenothiazine nucleus abol-
ishes the activity against arachidonate-induced aggregation,
leaving the activity against ADP and PAF platelet activation
unchanged or improved.
Cyanolupinane and epi-cyanolupinane were already de-
scribed by Clemo and Rudinger [14], who obtained them by
refluxing the ethanol solution of KCN and the corresponding
bromoderivatives, the preparation of which is rather tedious.
We found that the reaction of corresponding chloroderiva-
tives with NaCN in DMSO was more profitable.
In the above situation, we thought it is interesting to
undertake a more systematic study of several kinds of quino-
lizidine derivatives as potential antiaggregating agents.
Therefore, as a first possibility, a set of 2-phenyl-3-
quinolizidin-1a/1b-yl-5-substituted indoles of structures 1
and 2 (Scheme 1) is described.
The cyclization of phenylhydrazones with structures 3 and
4 yielded the expected indoles only in moderate or defini-
tively poor yields. Nevertheless, the formation of byproducts
Related 2-unsubstituted or 2-methyl-3-quinolizidin-1a-
yl-5-substituted indoles were described by us several years
ago [5,7] and some of them were moderately active against
ADP-induced platelet aggregation.
Indoles 1 and 2 were tested for antiaggregating activity
against ADP-, collagen- and adrenaline- induced platelet
aggregation. Moreover, a couple of these (1c and 1d) were
subjected to a general pharmacological screening.
2. Chemistry
The required 2-phenyl-3-(quinolizidin-1a/1b-yl)-5-
substituted indoles (1 and 2) were prepared through a Fischer
indole cyclization of lupinyl/epi-lupinylphenylketone
4-substituted phenylhydrazones (Scheme 2).
Scheme 3
A preliminary attempt of cyclization by refluxing the
ethanol solution of phenylhydrazones in the presence of dry
hydrogen chloride, as carried out in the past for the analogous
derivatives of lupinylmethylketone [5], failed completely.
Therefore, the cyclization was effected by heating the mix-
Scheme 4

M. Ercoli et al. / IL FARMACO 59 (2004) 101–109
103
analogous to those obtained in the cyclization of lupinylm-
ethylketone phenylhydrazones [5] was never observed. In-
deed, in the latter case a peculiar aminoketone [1(4-
reduced pressure (0.01–0.05 torr); b.p. 160–170 °C (air bath
temperature). In the case of the epi-lupinyl derivative, the
distillation was repeated two or three times to completely
eliminate the unreacted nitrile.
aminophenyl)-1-quinolizidin-1a-yl)-propan-2-one]
was
isolated, as a result of a para-benzidine-like rearrangement
of the hydrazoene intermediate, which is postulated by the
Robinsons mechanism of Fischer reaction [15,16].
Structures of final indoles were supported by elemental
analyses and spectral (UV and NMR) data.
IR spectra of both ketones, in KBr pellets, exhibited a
sharp carbonyl band around 1685 cm–1.
8: oil; 85% yield; analysis (C, H, N) for C17H23NO;
oxime: m.p. 235–237 °C, analysis for C17H24N2O.
12: m.p. 42–44 °C (dry ether–pentane), 45% yield; analy-
sis (C, H, N) for C17H23NO; oxime: m.p. 154–155 °C, analy-
sis (C, H, N) for C17H24N2O.
1H NMR spectra did not exhibit any peculiar feature, thus
only a few are described as examples (see Section 3).
3.1.2.2. Magnesium turnings (0.97 g, 40 mmol) were cov-
ered with 5 ml of dry ether and treated with a crystal of iodine
and a few drops of methyl iodide. After a few minutes, a
solution of freshly distilled chlorolupinane (7.5 g, 40 mmol)
in 25 ml of dry ether was added and the mixture was heated at
reflux, while stirring, for 4 h. After cooling, a solution of
4.12 g (40 mmol) of benzonitrile in 25 ml of dry ether was
added dropwise and then the mixture was refluxed for 12 h.
After cooling, a 10% solution of ammonium chloride was
added and the mixture was thoroughly extracted with ether.
The ether solution was then shaken with 1 N hydrochloric
acid (50 + 30 + 20 ml) and subsequently with water; the acid
solution was extracted with ether, basified and again ex-
tracted with ether. After removing the solvent, the oily resi-
due was distilled at 0.2 torr to obtain the ketone 8 (yield
75%).
3. Experimental
3.1. Chemistry
Melting points were determined by the capillary method
on an electrothermal apparatus and are uncorrected.
Elemental analyses were performed with a CE EA
1110 CHNS-O instrument and the results obtained for the
indicated elements were within 0.3% of the calculated val-
ues.
UV and IR spectra were recorded, respectively, on a
Varian Techtron model DMS 80 and on a Perkin Elmer
398 spectrophotometer; H NMR were taken on a Varian
1
Gemini 200 spectrometer, using CDCl3 (or d6-DMSO) as
solvent with TMS as internal standard.
3.1.3. Lupinyl- and epi-lupinylphenylketone 4-substituted
phenylhydrazones (3a–d, 4a–d)
3.1.1. Cyanolupinane (7) and epi-cyanolupinane (11)
Sodium cyanide (2 g, 36 mmol, 88% title) was added to a
solution of chlorolupinane [17] or epi-chlorolupinane [18]
(4.75 g, 25 mmol) in 12 ml of freshly distilled DMSO. The
solution was heated at 145 °C for 6 h.
The cooled suspension was diluted with water and after
having added 3 ml of 6-N NaOH, was thoroughly extracted
with ether. The ether solution was washed with a small
quantity of water, dried over sodium sulfate and evaporated
to dryness. The oily residue was distilled at 80–85 °C and
0.1 torr. In both cases about 3.9 g (87% yield) of nitrile was
obtained. Cyanolupinane was oily, while the epi-
cyanolupinane was solid and, after crystallization from pen-
tane, melted at 60–62 °C [19].
Equimolar amounts (5 mmol) of lupinylphenylketone or
epi-lupinylphenylketone and 4-substituted phenylhydrazine
were heated at 110 °C for 4 h under a slow stream of nitrogen.
The crude 4-substituted phenylhydrazones were used as such
for the indole cyclization.
3.1.4. 2-Phenyl-3-(quinolizidin-1␣/1b-yl)-5-substituted
indoles (1a–d, 2a–d)
The crude phenylhydrazone 3 or 4 (5 mmol) was thor-
oughly mixed with finely grounded anhydrous zinc chloride
in a dry-box. The mixture was gradually heated to 200 °C
under a slow stream of nitrogen and maintained at this tem-
perature for 10 min. After cooling, the mixture was taken up
with water, basified with a 2-N NaOH solution and extracted
with ether. The dried (Na2SO4) ether solution was evaporated
to dryness and the oily residue was worked up in different
manners depending on each compound.
Compounds 2a and 2b crystallized after the addition of a
few drops of anhydrous ether; crystals were collected and
washed with cold dry ether and the ether solution was chro-
matographed on neutral alumina (activity 1) to obtain a
further crop of crystalline compound.
3.1.2. 1-Phenyl-2-(quinolizidin-1␣/1b-yl)ethanones (8
and 12) (lupinyl- and epi-lupinylphenylketones)
3.1.2.1. A commercial 3 M ethereal solution of phenylmag-
nesium bromide (4 ml) was diluted with 20 ml of ether and
added dropwise to a solution of cyanolupinane or epi-
cyanolupinane (2 g, 11.2 mmol) in 20 ml of dry ether. The
mixture was heated at reflux for 9 or 18 h, respectively, and
then treated with 1 N hydrochloric acid (30 ml). The acid
solution was extracted with ether, basified and again ex-
tracted with ether.
In the case of compounds 1b, 1c, 2c and 2d, the oily
residue was distilled at 0.1 torr (120–125 °C, air bath tem-
perature) to remove the starting ketones; the residue was then
After removing the solvent, the residue was distilled under

104
M. Ercoli et al. / IL FARMACO 59 (2004) 101–109
Table 1
Characteristics of compounds 1a–e and 2a–e
Compound
Formula a
Solvent b
Yield
(%)
34
M.p.
(°C)
UV
kmax (nm)
223; 298
223; 298
231; 304
223; 299
224; 304
224; 299
226; 302
230; 304
228; 302
226; 305
1a
1b
1c
1d
1e c
2a
2b
2c
2d d
2e
C
C
C
C
C
C
C
C
C
C
23H26N2 + HCl
a
b
c
a
d
b
b
b
b
b
>310
23H25FN2
18
166–167
162–163
>310
23H25ClN2
35
24H28N2 + HCl + 0.5 H2O
24H28N2O
35
14,5
40
169–171
249–250
260–262
239–240
175–176
190–192
23H26N2
23H25FN2
30
23H25ClN2
28,5
18
24H28N2
24H28N2O + 0.25 H2O
3
a All compounds were analyzed for C, H, N and results were within 0.3% of calculated values.
b a : Absolute ethanol; b: anhydrous ether; c: anhydrous ether/pentane; d: ethanol.
c 1H NMR (CDCl3): d1.16–2.3 (m, 13H, quinolizidine ring); 2.9–3.3 (m, 3H, a-protons on C3, C4 and C6 of quinolizidine ring); 3.65 (s, 3H, OCH3); 6.8–7.9
(m, 8H, arom.); 8.8 (bs, 1H, exchangeable, NH).
d H NMR (CDCl3): d 0.9–2.55 (m, 13H, quinolizidine ring + s, 3H at d 2.5, CH3); 2.8–3 (m, 3H, a-protons on C3, C4 and C6 of quinolizidine ring); 6.82–7.65
(m, 8H, arom.); 7.9 (bs, 1H, exchangeable, NH).
chromatographed on alumina, eluting with dry ether to ob-
tain crystalline compounds.
Finally, in the case of compounds 1a and 1d, the oily
residue was chromatographed directly on alumina, eluting
with ether; indoles were eluted as oils, which were converted
to crystalline hydrochlorides.
Platelet aggregation, performed in an Aggrecorder II
PA3220 aggregometer (Menarini, Firenze, Italy), was mea-
sured according to Born’s turbidimetric method [20] and
quantified by the maximal light transmission after 5 min.
A first sample of PRP was pre-incubated at 37 °C for
2 min before the addition of the agonist (2 and 5 µM ADP;
collagen at final concentrations of 4 and 2 µg/ml, 5 µM
adrenaline).
A second sample of PRP was incubated for 2 min with a
solution of the tested compound. Platelet aggregation was
then performed adding the agonist and comparing the maxi-
mal light transmission of the aggregation curves obtained
with and without the addition of the tested compound, the
percentage of inhibition of platelet aggregation was calcu-
lated.
Characteristics of compounds are reported in Table 1.
3.1.5. 5-Methoxy-2-phenyl-3-(quinolizidin-1␣/1b-yl)indoles
(1e and 2e)
4-Methoxyphenylhydrazine hydrochloride (0.69 g,
4 mmol) and 1.2 ml (17.5 mmol) of 85% phosphoric acid
were added to a solution of lupinylphenylketone or epi-
lupinylphenylketone (1.03 g, 4 mmol) in 12 ml of absolute
ethanol. The solution was heated at reflux for 12 h under a
stream of nitrogen and then evaporated to dryness under
reduced pressure. The residue was dissolved in water, bas-
ified with a 2-N NaOH solution and extracted with ether. The
solvent was evaporated and the residue was distilled at 120–
125 °C (0.1 torr) to remove the unreacted ketone (≈0.7 g).
The undistilled residue was chromatographed on neutral alu-
mina (activity 1) eluting with dry ether to obtain the crystal-
line indoles 1e and 2e.
All compounds were used as hydrochlorides.
3.3. General pharmacological screening
A general pharmacological screening was performed by
MDS-Panlabs, Bothell, WA, USA, on compounds 1c and 1d.
This screening consisted in the determination of the maxi-
mum tolerated dose in mice (MTD, p.o. and i.p.) with simul-
taneous behavioral examination (Irwin test) and in 34 pri-
mary in vivo tests (using a suitable MTD fraction, depending
on the test type) and in 20 in vitro tests, concerning CNS,
cardiovascular and gastrointestinal apparatuses, intermedi-
ary metabolism, allergy and inflammation.
For in vivo tests, compounds were generally administered
orally using a gastric tube in the form of aqueous solutions or
finely homogenized suspensions in Tween 80 (2%), when the
highest dose (300 mg/kg) was used. In a few cases com-
pounds were introduced i.p. as aqueous solutions (10 ml/kg).
Groups of three or five animals (rat or mice) were used.
For in vitro assays, the dissolution of a test compound in a
buffer or saline solution was speeded up by means of DMSO;
Characteristics of compounds are reported in Table 1.
3.2. Platelet aggregation assays
Human blood samples from normal volunteers were
drawn through a 19-gauge needle, carefully avoiding pro-
longed venous stasis. None of these subjects were treated
with any drug known to influence the platelet function.
Blood (36 ml) was collected in plastic tubes containing
3.8% trisodium citrate aqueous solution (4 ml). Platelet-rich
plasma (PRP) was obtained by centrifuging the blood at 100
g for 20 min. Platelet-poor plasma (PPP) was obtained by
centrifuging the remaining blood at 1100 g for 15 min.
Platelet count in PRP was maintained at 300,000/mmc.

M. Ercoli et al. / IL FARMACO 59 (2004) 101–109
105
Table 2
Human platelet antiaggregating activity of compounds 1a–e and 2a–e
Compound
Final concentration
(µM)
Inhibition of platelet aggregation induced by ADP, collagen and adrenalin in human plasma (%) a
ADP (5 µM)
ADP (2 µM)
Collagen
(4 µg/ml)
93
Collagen
(2 µg/ml)
93
Adrenalin
(5 µM)
87
1a
1b
1c
1d
1e
2a
2b
2c
2d
2e
250
125
62.5
250
125
62.5
250
125
62.5
250
125
62.5
250
125
62.5
250
125
62.5
250
125
62.5
250
125
62.5
250
125
62.5
250
125
62.5
1000
100
350
240
64
84
21
24
39
31
14
33
16
7
90
54
43
62
51
17
52
39
3
87
88
82
25
32
68
86
87
88
79
87
86
22
39
84
87
90
85
20
47
72
11
6
5
67
43
9
78
59
32
88
65
46
67
56
47
56
31
13
58
51
36
75
65
62
96
86
89
41
34
85
82
15
92
91
94
74
87
88
31
46
42
67
48
29
80
61
49
56
35
11
56
51
11
48
44
39
89
58
50
45
9
84
89
80
86
91
86
23
69
65
86
89
100
77
74
85
11
17
63
83
84
81
74
83
68
20
55
25
91
89
89
68
83
31
19
43
5
89
89
86
45
87
79
11
36
55
92
93
97
90
89
76
16
78
79
ASA
41
37
68
3
3
30
Adenosine
55
32
6
a Mean value of two or three determinations.
the final concentration of DMSO (not interfering with the
tests) was 0.1% for platelet aggregation and 0.5% for all the
others.
4. Results and discussion
Antiaggregating activity of compounds 1a–e and 2a–e
(Scheme 1) was tested on human platelet aggregation stimu-
lated by ADP at 5 and 2 µM concentration, collagen at 4 and
2 µg/ml and adrenaline at 5 µM concentration. Compounds
were assayed at 2.5, 1.25 and 0.62 × 10–4 M. Results are
collected in Table 2.
Most compounds, when tested at the concentration 2.5 ×
10–4 M, strongly inhibited the aggregation induced by all the
agonists considered, and many of them maintained good
activity at 1.25 × 10–4 M concentration.
The lowest inhibitory effects were observed against ADP,
while the highest were against adrenaline. All compounds
were definitely more active than acetylsalicylic acid (ASA) at
comparable concentrations, and many compounds in the
aforementioned concentrations were more active than ASA
The procedures for these assays have already been de-
scribed [7–10,21–23]: a pre-established level of response,
which is high enough to suggest significant activity, is indi-
cated for each assay. Doses (mg/kg) or concentrations
(mg/ml or µM) indicated in the methods were the highest
used routinely, depending on toxicity; when significant activ-
ity was detected, lower doses or concentrations were tested in
order to define the minimal effective ones and secondary
tests were performed to provide some insight for the possible
mechanisms of action.
In Tables 3–7 only the results of the more significant
assays are collected; the pertinent reference for the method
employed is indicated near each assay name.

106
M. Ercoli et al. / IL FARMACO 59 (2004) 101–109
Table 3
Inhibition of rabbit platelet aggregation induced by several agents and related in vivo assays
Test
Concentration (µm) Effect
Route; dose (mg/kg)
Inhibition by compound (%)
Reference drug
Aspirin
Concentration (µm)
Route; dose (mg/kg)
Inhibition (%)
100
1c
1d
Platelet aggregation induced by
a
Na
27
8
0
100
100
0
arachidonate
[8]
13.8
2.7
270
80
27
8
a
a
ADP [8]
57
9
100
20
Adenosine
RP48740 d
18.7
20.4
64
54
PAF [8]
97
48
nt
100
60
2.7
20
In vivo
b
c
Bleeding time
[10]
p.o. 100
nt
28
Aspirin
p.o. 100
p.o. 100
60
60
PAF
antagonism
(mice) [8]
p.o. 100
p.o. 30
40
nt
60
40
L-65273 e
a % inhibition of maximum non-reversible rabbit platelet aggregation induced by: Na arachidonate (50 µg/ml = 153 µM), ADP (0.8 µg/ml = 1.8 µM) or PAF
(10–20 ng/ml).
b % prolongation of bleeding time relative to a control group of mice.
c % inhibition of mortality observed 1 h after the i.v. administration of PAF–acether (200 µg/kg).
d 3-(3-Pyridyl)-1H,3H-pyrrolo[1,2-c]thiazole-7-carboxamide.
e trans-2,5-bis-(3,4,5-trimethoxyphenyl)tetrahydrofuran.
Table 4
Maximal tolerated dose in mice [8] for compounds 1c and 1d
Administration route
Dose (mg/kg)
Number of animals dead (and effects) after treatment with compound
1c
1d
Oral
i.p.
300
100
30
0/3
0/3
No effect
0/3
Slight decrease limb tone and grip
0/3
No effect
nt
Slight tremor and muscolar relax
0/3
Slight decrease limb tone and grip
Table 5
Analgesic activity in mice
Test
Route; dose
Effect
Effect produced by
compound 1c
Reference drug
Route; dose (mg/kg)
Effect
(mg/kg1)
p.o. 100
p.o. 30
p.o. 100
p.o. 30
p.o. 10
i. p. 30
a
Phenylquinone
writhing [8]
Formalin
69
36
90
59
9
Aspirin
p.o. 100
p.o. 30
p.o. 300
p.o. 30
p.o. 10
i.p. 10
68
65
52
66
62
65
Ibuprofren
Aspirin
b
c
analgesia [21]
Ibuprofen
Morphine
Tifluadom
Tail flick [8]
10
a Inhibition of number of writhes.
b % reduction of induced paw licking time recorded between 20 and 30 min after formalin injection.
c % prolongation of the time to elicit pain response.
at 10–3 M concentration. Many compounds displayed good
activity even at 0.625 × 10–4 M concentration against ADP-,
collagen- and adrenaline-induced aggregations; in particular
at this concentration compounds 1d, 1e, 2a and 2c inhibited
ADP-induced aggregation for 40–50%. It is therefore evident
that there is a concentration-dependent inhibitory effect,
though the curve steepness is very variable for both the tested
compounds and the agonist considered.
quinolizidine ring to the indole nucleus does not produce a
homogeneous effect and, in any case, the differences ob-
served in the antiaggregating activity are not very high.
A somewhat more marked influence is exerted by the
nature of substituents on position 5, with methoxy, methyl
and unsubstituted compounds being better than fluoro- and
chloroderivatives.
Indeed compounds 1c and 1d, which resulted, respec-
tively, the least active and one of the most active as inhibitors
of human platelets aggregation are characterized by the pres-
It is worth noting that, with regard to the structure–activity
relationships, the different (a or b ) spatial connection of the

M. Ercoli et al. / IL FARMACO 59 (2004) 101–109
107
Table 6
In vivo assays for antihyperlipemic [21] and diuretic [8] activities of compounds 1c and 1d
Test
Dose mg/kg p.o.
Effect
Response to compound
Reference drug
Dose mg/kg p.o.
Response
1c
1d
nt
a
Total
cholesterol (in
mice)
300
100
30
10
–30
–27
–21
–16
–7
Bezafibrate
300
100
–39
–25
0
nt
nt
D-thyroxine
10
–30
3
nt
b
Heparin
precipitating
HPL (in mice)
300
100
30
10
–35
–29
–23
–21
–3
nt
0
Bezafibrate
D-thyroxine
100
10
–30
–35
nt
nt
3
nt
c
Ratio HPL/total 300
0.93
0.97
0.97
0.94
–20
nt
nt
cholesterol
100
30
10
30
10
30
10
30
0.96
nt
Bezafibrate
100
0.93
nt
D-thyroxine
Amiloride
10
20
0.93
+60
d
e
e
Urine volume
(in rats)
+50
+20
+200
+20
+20
Saluresis (in
rats)
+30
nt
Furosemide
5
+200
Kaluresis (in
rats)
–30
a Average percent reduction in serum cholesterol concentration ≥ 15% relative to hypercholestrolemic control mice indicates significant activity (established
control animals have a serum total cholesterol content of 350 28 mg/dl).
b Reduction of serum HPL (heparin precipitating betalipoproteins) concentration ≥ 20% indicates significant activity.
c Reduction of ratio below 0.92 indicates potential anti-atherogenic activity.
d % variation of Na+ or K+ excretion expressed as µeq/100 g body weight; a variation ≥ 50% indicates significant activity.
e Variation of Na+ or K+ excretion expressed as µeq/100 g body weight; a variation ≥ 50% indicates significant variety.
ence of a chlorine and a methyl group, respectively. They
were also tested for antiaggregating activity on rabbit plate-
lets against ADP, PAF and sodium arachidonate activation.
The activity against ADP-induced aggregation was con-
firmed by both the compounds, which, at lower concentra-
tions, also displayed high activity against the PAF-induced
aggregation, comparable or superior to that of the standard
resulting, respectively, comparable or superior to aspirin. In
the latter test compound 1c, at 30 mg/kg p.o., was compa-
rable to morphine at 10 mg/kg p.o. However, compound 1c
was inactive against the carragenan-induced rat paw edema,
suggesting that the observed antinociceptive activity may be
of central origin.
The relevant compound exerted a strong reduction in se-
rum cholesterol and heparin precipitating betalipoproteins in
hypercholesterolemic mice; the reduction was still signifi-
cant after an oral dose of only 10 mg/kg. Thus, compound 1c
is an interesting hit for the preparation of antihypercholester-
olemic agents not structurally related to existing drugs.
The simultaneous presence in the same molecule of hypo-
cholesterolemic and platelet antiaggregating activities repre-
sents an outstanding property, which could result useful in
the prevention and treatment of thrombotic disorders.
This double-faced activity is shared, as far as we know,
only by fibrates and has recently been described by some of
us [22] for 5-chloro-2-{4-[3-(dimethylammino)propoxy]
phenyl}benzotriazole. The latter compound, as well as
clofibric acid, while active against ADP-induced platelet
aggregation, was inactive against arachidonic acid-induced
aggregation, similarly to compound 1c.
agent
RP-48740
(3-(3-pyridyl)-1H,3H-pyrrolo[1,2-c]
thiazole-7-carboxamide). However, a clear-cut difference
was observed in the activity of compounds 1c and 1d against
the arachidonate-induced aggregation; while the first com-
pound was completely inactive at a 27 µM concentration, the
second one still gave a 100% inhibition at 8 µM concentra-
tion. Correspondingly, compound 1d produced a moderate
increase of bleeding time in mice.
Both compounds, but in particular 1d, protected mice
from mortality induced by i.v. injection of PAF, resulting
comparable to compound L-652731 [trans-2,5-bis(3,4,5-
trimethoxyphenyl)tetrahydrofuran] used as a reference drug.
This protective activity is rarely observed [24–26] and there-
fore, it deserves further investigation to define eventual thera-
peutic potentialities.
Compounds 1c and 1d, when subjected to the general
pharmacological screening resulted well tolerated up to the
maximum tested dose of 300 mg/kg p.o. and 100 mg i.p.
The 5-chloroderivative 1c exhibited analgesic activity in
phenylquinone writhing test and against formalin algesia,
On the other hand, the 5-methylderivative 1d did not have
analgesic and antihypocholesterolemic activities, but it was
endowed with saluretic activity (associated with increased
urine volume output in the hydrated rat) and with tracheal

108
M. Ercoli et al. / IL FARMACO 59 (2004) 101–109
Table 7
Antiarrhythmic activity in mice [8] and some in vitro assays related to cardiocirculatory function
Test
Route; dose (mg/kg)
Concentration (µg/ml)
Effect
Effects produced by compounds 1c and 1d and reference drugs
1c
1d
Lidocaine
Quinidine
Others
In vivo
a
Chloroform arrhythmia
i.p. 100
i.p. 30
2/3
0/3
nt
3/3
2/3
0/3
55
0
3/3
3/3h
i.p. 10
b
b
Aconitine arrhythmia
i. p. 100
i.p. 30
92
0
108h
113h
Aconitine ventricular
tachycardia
i.p. 100
115
87
i.p. 30
0
0
95h
46
89h
37i
In vitro
c
Left atria electrical driven 100
80
50
20
–19
0
nt
(guinea pig) Inotropic
effect (guinea pig left
atria)
30
10
10
10
3
51
0
d
d
–50
–13
–7
94
78
56
18
81
71
56
15
0
–20i
Chronotropic effect
(guinea pig right atria)
Ca2+ antagonism (guinea
pig ileum)
–18 (diltiazem)
nt
e
3
74
19
nt
65j (diltiazem)
1
0.3
0.1
3
nt
Ca2+ antagonism (guinea
pig left atria)
17
nt
60k (diltiazem)
f
1
0.3
0.1
100
30
10
3
nt
nt
g
g
Methacoline antag.
(guinea pig ileum) CCK
antagonist (guinea pig
ileum)
100
68
41
0
75l (atropine)
nt
nt
52
27
70l (lorglumide)
1
nt
a Number of protected over treated animals.
b % Time increase for onset of arrhythmia and ventricular tachycardia relative to control group.
c % Reduction in ability to follow electrically-induced contractile frequency.
d % Variation in contractile force or rate.
e % Inhibition of Ca2+-induced contraction.
f % Reduction of a Ca2+-induced increase in contractile force.
g % Reduction of contractile response.
h Tested at 50 mg/kg.
i Tested at 25 µg/ml.
j Tested at 0.01 µg/ml.
k : Tested at 2.5 µg/ml.
l Tested at 0.1 µg/ml.
relaxant activity in vitro. The latter appears unrelated to
b2-adrenergic agonism, since it was not inhibited by propra-
nolol, but may relate to possible inhibition of cyclooxyge-
nase, suggested by the aforementioned inhibition of arachi-
donic acid-induced platelet aggregation.
Both compounds 1c and 1d exhibited good antiarrhythmic
activity in mice, which underwent deep chloroform anesthe-
sia and aconitine infusion, as well as in electrically driven left
atria. This activity may relate to calcium antagonism, which
was seen in vitro in guinea pig ileum, but not in rat portal
vein; calcium antagonism was observed also in guinea pig
atria for compound 1d, but not for 1c. Moreover, in the case
of compound 1d, the antiarrhythmic activity may be also
related to the moderate (but long lasting) local anesthetic
activity observed in the guinea pig corneal surface stimula-
tion test. This activity was completely absent in compound
1c.
Finally, compound 1c antagonized the methacholine in-
duced, but not the CCK-induced, contractile response in
guinea pig ileum; the opposite was observed in compound
1d.
The results of the general pharmacological screening,
though limited to only two compounds, clearly indicate that
the nature of the substituent introduced on position 5 of
2-phenyl-3-(quinolizidin-1′-yl)indole may not only give rise
to quantitative differences in each activity (as already ob-
served for the human platelet antiaggregating activity), but
even to largely different biological properties.

M. Ercoli et al. / IL FARMACO 59 (2004) 101–109
109
[9] A. Sparatore, F. Sparatore, Preparation and pharmacological activities
of 10-homolupinanoyl-2-R-phenothiazines, Farmaco 49 (1994) 5–17.
[10] A. Sparatore, F. Sparatore, Preparation and pharmacological activities
of homolupinanoyl anilides, Farmaco 50 (1995) 153–166.
[11] G. Iusco, V. Boido, F. Sparatore, Synthesis and preliminary pharma-
cological investigation of N-lupinyl-2-methoxybenzamides, Farmaco
51 (1996) 159–174.
[12] F. Novelli, B. Tasso, F. Sparatore, Synthesis and pharmacological
evaluation of some thiolupinine derivatives, Farmaco 54 (1999) 354–
359.
Therefore, the 2-phenyl-3-(quinolizidin-1′-yl)indoles de-
serve further studies. The preparation of new 5-substituted
derivatives and corresponding position isomers will be pur-
sued, while the mechanisms of some of the observed activi-
ties will be investigated with particular concern to the pos-
sible connection of the antihypercholesterolemic activity
with the sigma receptor affinity and/or peroxysome
proliferator-activated receptor (PPAR) agonism.
[13] A. Sparatore, C. Boido Canu, F. Sparatore, Unpublished results.
[14] G.R. Clemo, J. Rudinger, Syntheses from lupinine, J. Chem. Soc.
(1951) 2714–2727.
Acknowledgements
[15] F. Sparatore, V. Boido, G. Pirisino, Formation of apparent para-
benzidine rearrangement products during indole syntheses, Tetrah.
Lett. 27 (1974) 2371–2372.
[16] F. Sparatore, V. Boido, Indole derivatives of pharmacological interest
and pathway of the cyclization reaction of some arylhydrazones,
Farmaco 43 (1988) 1097–1114.
Financial support from the Italian MIUR is gratefully
acknowledged.
References
[17] F. Sparatore, V. Boido, P. Preziosi, E. Miele, G. De Natale, Sintesi e
proprietà farmacodinamiche di alcuni derivati lupinanici (Synthesis
and pharmacodynamic properties of some lupinanic derivatives), Far-
maco, Ed. Sci. 24 (1969) 587–621.
[18] V. Boido, A. Sparatore, Chinolizidinilderivati di sistemi triciclici
(Quinolizidinylderivatives of tricyclic systems), Farmaco, Ed. Sci. 37
(1982) 63–73.
[1] S.A. Mousa, Antiplatelet therapies: from aspirin to GPIIb/IIIa-
receptor antagonists and beyond, Drug Discov. Today 4 (1999) 552–
561.
[2] P. Gresele, G. Agnelli, Novel approaches to the treatment of thrombo-
sis, Trend. Pharmacol. Sci. 23 (2002) 25–32.
[19] V. Boido, A. Boido, C. Boido Canu, F. Sparatore, Chinolizidinil-
alchilammine ad azione antiipertensiva (Quinolizidinyl-alkylamines
endowed with antihypertensive activity), Farmaco, Ed. Sci. 34 (1979)
673–687.
[20] G.V.R. Born, Aggregation of blood platelets by adenosine diphos-
phate and its reversal, Nature 194 (1962) 927–929.
[21] F. Novelli, F. Sparatore, Preparation and pharmacological activities of
spiro[3,4-dihydro-6/7-R-1,2,4-benzotriazine-3,4′-(1′substituted) pip-
eridines ], Farmaco 51 (1996) 541–550.
[3] D.L. Bhatt, E.J. Topol, Scientific and therapeutic advances in anti-
platelet therapy, Nature Rev. Drug Discov. 2 (2003) 15–28.
[4] V. Boido, F. Sparatore, Derivati di amminoalcoli e di diammine
naturali d’interesse farmacologico. Nota V. Nuovi derivati della
lupinina e dell’amminolupinano. Prime notizie sulla loro attività far-
macologica (Derivatives of natural aminoalcohols and diamines of
pharmacological interest. V. New derivatives of lupinine and amino-
lupinane. Preliminary report on their pharmacological activity), Ann.
Chim. 59 (1969) 526–538.
[22] A. Sparatore, F. Sparatore, Synthesis and preliminary pharmacologi-
[5] V. Boido, C. Boido Canu, Derivati di amminoalcoli e di diammine
naturali d’interesse farmacologico. Nota VI. 3-Chinolizidinil-e
3-chinolizidinilmetilindoli (Derivatives of natural aminoalcohols and
diamines of pharmacological interest. VI. 3-Quinolizidinyl- and
3-quinolizidinylmethylindoles), Ann. Chim. 63 (1973) 593–600.
[6] A. Sparatore, C. Boido Canu, F. Sparatore, P. Manca, M. Satta,
1-Lupinilchinossalin-2(1H)-oni 3,6-disostituiti d’interesse farmaco-
logico [3,6-Disubstituted 1-lupinylquinoxalin-2(1H)-ones of pharma-
cological interest], Farmaco, Ed. Sci. 40 (1985) 303–314.
[7] C. Boido Canu, V. Boido, F. Sparatore,A. Sparatore, Sintesi ed attività
farmacologica di 3-chinolizidin-1′-il-5R-indoli (Synthesis and phar-
macological activity of 3-quinolizidin-1′-yl-5R-indoles), Farmaco,
Ed. Sci. 43 (1988) 801–817.
cal
investigation
of
some
2-[4-(dialkylaminoalkoxy)-
phenyl]benzotriazole and their N-oxides, Farmaco 53 (1998) 102–
112.
[23] C. Canu Boido, F. Sparatore, Synthesis and preliminary pharmaco-
logical evaluation of some cytisine derivatives, Farmaco 54 (1999)
438–451.
[24] D.A. Handley, Preclinical and clinical pharmacology of platelet-
activating factor receptor antagonists, Med. Res. Rev. 10 (1990) 351–
370.
[25] D. Hosford, P. Braquet, Antagonists of platelet-activating factor:
chemistry, pharmacology and clinical applications, Progr. Med.
Chem. 27 (1990) 325–380.
[8] F. Novelli, F. Sparatore, Thiolupinine and some derivatives of phar-
macological interest, Farmaco 48 (1993) 1021–1049.
[26] M. Singh, M.K. Saraf, Platelet-activating factor: a new target site for
the development of nootropic agents, Drug. Fut. 26 (2001) 883–889.