Antimalarial activity of novel pyrrolizidinyl derivatives of 4-aminoquinoline

Article abstract of DOI:10.1016/j.bmcl.2008.05.042

Two pyrrolizidinylalkyl derivatives of 4-amino-7-chloroquinoline (MG2 and MG3) were prepared and tested in vitro against CQ-sensitive and CQ-resistant strains of Plasmodium falciparum and in vivo in a Plasmodium berghei mouse model of infection. Both compounds exhibited excellent activity in all tests and low toxicity against mammalian cells. Preliminary studies of the acute toxicity and of the metabolism of the most active compound MG3 indicate a promising profile as a new antimalarial drug candidate.

Full text of DOI:10.1016/j.bmcl.2008.05.042

Bioorganic & Medicinal Chemistry Letters 18 (2008) 3737–3740
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Antimalarial activity of novel pyrrolizidinyl derivatives of 4-aminoquinoline
b
a
b
b
Anna Sparatore a, , Nicoletta Basilico , Manolo Casagrande , Silvia Parapini , Donatella Taramelli ,
*
Reto Brun c, Sergio Wittlin c, Fabio Sparatore d
a Istituto di Chimica Farmaceutica e Tossicologica ‘‘Pietro Pratesi”, University of Milan, Via Mangiagalli 25, 20133 Milan, Italy
b Dipartimento di Sanita’ pubblica—Microbiologia—Virologia, University of Milan, Via Pascal, 36, 20133 Milan, Italy
c Swiss Tropical Institute, Socinstrasse 57, CH-4002 Basel, Switzerland
d CTG Pharma, Viale Gran Sasso 17, 20131 Milan, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Two pyrrolizidinylalkyl derivatives of 4-amino-7-chloroquinoline (MG2 and MG3) were prepared and
tested in vitro against CQ-sensitive and CQ-resistant strains of Plasmodium falciparum and in vivo in a
Plasmodium berghei mouse model of infection. Both compounds exhibited excellent activity in all tests
and low toxicity against mammalian cells. Preliminary studies of the acute toxicity and of the metabolism
of the most active compound MG3 indicate a promising profile as a new antimalarial drug candidate.
Ó 2008 Elsevier Ltd. All rights reserved.
Received 11 April 2008
Revised 8 May 2008
Accepted 10 May 2008
Available online 16 May 2008
Keywords:
4-Aminoquinoline derivatives
Pyrrolizidinyl derivatives
Chloroquine
Antimalarial agents
Novel, effective, safe and inexpensive antimalarial agents are
urgently needed to treat malaria in tropical and subtropical re-
gions, where it still represents a serious health problem, affecting
400–500 million people annually.
N
N
H
H
(CH2)n
NH
(CH2)n
NH
Plasmodium falciparum (Pf), which is responsible for >1 Mio
fatalities, has developed resistance to common antimalarials such
as chloroquine (CQ) or antifolates. Nevertheless, several CQ analogs
were shown to retain significant activity against CQ-resistant
strains of Pf, suggesting that resistance could be compound-spe-
cific and not related to changes in the structure of the drug target.1
Indeed, we have recently demonstrated that analogs of CQ and
quinacrine, whose 4-diethylamino-1-methylbutyl chain was re-
placed by a quinolizidinyl- or a quinolizidinyl-alkyl moiety (Fig.
1) exhibited high antimalarial activity against CQ-S and CQ-R
strains of Pf in vitro.2 Some of these compounds also showed effi-
cacy against P. berghei and P. yoelii when given orally (po) or intra-
peritoneally (ip) in a murine standard 4-day test.3
OCH3
n = 1 - 3
n = 0 - 3
Cl
N
Cl
N
Figure 1. Structures of quinolizidinyl analogs of chloroquine and quinacrine.
In the light of these promising results and pursuing the research
of drugs which are cheaper and easier to synthesize, we prepared
and studied two novel analogs of CQ (MG2 and MG3), where the
amino group of quinoline is linked to a pyrrolizidinylalkyl ((hexa-
hydro-1H-pyrrolizin-7a-yl)alkyl) moiety (Fig. 2).
In these new molecules a bulky, basic and lipophilic head is
again present, but such a synthetic bicyclic ring is linked in a posi-
The presence of a bulky, strongly basic and lipophilic bicyclic
moiety (as the quinolizidine ring) appears as an interesting struc-
tural feature able to overcome the resistance mechanisms by pre-
venting the metabolic oxidative dealkylation, which affects the
usual dialkylaminoalkyl chains of many CQ analogs. This metabolic
dealkylation significantly reduces the lipid solubility of the drug
and significantly increases cross-resistance up to and beyond that
seen with CQ.1
N
(CH2)n
NH
n=2 MG2
n=1 MG3
Cl
N
* Corresponding author. Tel.: +39 02 50319365; fax: +39 02 50319359.
Figure 2. Structures of new pyrrolizidinylalkyl derivatives of 4-amino-7-
E-mail address: anna.sparatore@unimi.it (A. Sparatore).
chloroquinoline.
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.05.042

3738
A. Sparatore et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3737–3740
N
NH2
CN
d
NH
N
c
N
N
a
MG2
Cl
Cl
Cl
O
N
b
NH2
CN
N
d
c
NH
N
MG3
Cl
N
Scheme 1. Reagents and conditions: (a) cyanoacetic acid, NH4OH, n-hexane, 15 °C, 40 h; (b) 2-amino-2-methyl-propanenitrile, NH3, MeOH, 20 °C, 24 h; (c) LiAlH4, dry Et2O,
rt, 5 h; (d) 4,7-dichloroquinoline, phenol, 180 °C, 4 h.
Table 1
In vitro antimalarial activities, b-hematin inhibitory activities (BHIA assay) and cellular cytotoxicity of MG2 and MG3
a
Compound
D-10 (CQ-S)
IC50 (nM)
W-2 (CQ-R)
IC50 (nM)
NF-54 (CQ-S)
IC50 (nM)
K1(CQ-R)
IC50 (nM)
BHIA Drug:haem
molar equivalent
WEHI 13
IC50 (nM)
HMEC-1
IC50 (nM)
K562 IC50
(nM)
a
a
b
b
a
a
MG2
MG3
CQc
5.5
30.6
16.5
54.7
28.9
293.5
6.33
10.9
4.7
19.6
10.6
48.5
1.66
1.06
1,69
61,453
24,717
>40,000
>60,000
>60,000
>40,000
17,587
18,069
31,828
a
Mean of 3–4 experiments, each performed in triplicate.
Mean of two experiments.
Diphosphate salt.
b
c
Table 2
In vivo single dose of MG2 and MG3 on P. berghei murine modela
tion to prevent the chirality present in the previous quinolizidinyl
analogs.
Compound
Dose (mg/Kg)
Route
Activity (%)
Mouse survival (days) Avg.
It is worth noting that a 7a-substituted pyrrolizidine ring is
present in several biologically active substances such as the antiar-
rhythmic pilsicainide,4 cerebral function activators,5 muscarinic
M1 and serotonine 5-HT4 receptor agonists, useful in improving
cognitive functions,5–7 and as gastro-intestinal prokinetic agents,8
respectively.
MG2
MG3
CQb
MG2
MG3
CQb
100
100
100
30
30
30
po
po
po
sc
sc
sc
98
12.3
13.7
14.7
8.7
10.3
9.7
99.5
99.2
99.1
99.7
99.2
Controls
7.0
The target compounds were prepared by reacting in the pres-
ence of phenol the 4,7-dichloroquinoline with the suitable (hexa-
hydro-1H-pyrrolizin-7a-yl)alkylamine.9 The amino compounds
were obtained by LiAlH4 reduction of the corresponding ni-
triles,10,11 which in turn were formed by cyclization of 1,7-di-
chloro-4-heptanone with NH3 and cyanoacetic acid or 2-amino-
2-methylpropanenitrile6,12,13 (Scheme 1).
a
Groups of three P. berghei-infected NMRI mice were treated 1-day post-infec-
tion with compounds formulated in 70/30 Tween 80/ethanol and diluted 10Â with
water before administration. Antimalarial efficacy was measured by percent
reduction in parasitemia on day 3 post-infection and animal survival time was
compared to untreated control groups (n = 5).
b
Diphosphate salt.
Compounds MG2 and MG3 were tested in vitro against CQ-S (D-
10 and NF-54) and CQ-R (W-2 and K1) strains of Pf. The antimalar-
ial activity (IC50) was quantified as inhibition of parasite growth,
measured with the production of parasite lactate dehydrogenase
(for D-10 and W-2 strains)14 or the incorporation of [3H]-hypoxan-
thine (for NF-54 and K1 strains).15 Moreover, to assess the mecha-
nism of action, the BHIA (b-hematin inhibitory activity) assay was
performed as previously described.16 Cytotoxicity on murine cells
WEHI, clone 13, or human cells HMEC-1 and K562, was assayed
using the MTT test.17 Finally, the ability to protect NMRI mice from
a lethal CQ-susceptible P. berghei GFP-transfected ANKA strain
infection18 was studied, administering the compounds either orally
or subcutaneously (sc) as a single dose or repeated doses for 4 con-
secutive days.19,20
The results of these assays are presented in Tables 1–3.
The two compounds MG2 and MG3 were active against both
CQ-S strains D-10 and NF-54 in a comparable range as CQ (Table
1). Importantly, these compounds exhibited high activity against
both CQ-R strains W-2 and K1, being 2.5- to 10-fold more active
than the reference drug CQ.
Table 3
In vivo 4-day administrations of MG2 and MG3 on P. berghei murine modela
Compound Dose (mg/kg):
Route Activity
(%)
Mouse survival (days)
Avg.
4Â
MG2
MG3
CQc
30
10
30
10
30
10
10
3
po
po
po
po
po
po
sc
>99.9
>99.9
>99.9
>99.9
>99.9
99.9
30.0 (3/3 mice curedb)
14.7
25 (2/3 mice cured)
25.7 (2/3 mice cured)
30 (3/3 mice cured)
22.3
15.3
9.3
MG2
MG3
CQc
>99.9
94
sc
10
3
sc
sc
>99.9
96
29.3 (2/3 mice cured)
7.3
10
3
sc
sc
>99.9
99.8
15.7
9.3
Controls
7.0
a
Groups of three P. berghei-infected NMRI mice were treated 1 day post-infection
with compounds formulated in 70/30 Tween 80/ethanol and diluted 10Â with
water before administration. Antimalarial efficacy was measured by percent
reduction in parasitemia on day 4 post-infection and animal survival time was
compared to untreated control groups (n = 5).
b
Compound MG3, though less active than MG2 on CQ-S strains,
was definitely better than the homolog on CQ-R strains, and was
Mice cured: no parasites present at day 30. Parasitemia was checked by microscopy.
Diphosphate salt.
c

A. Sparatore et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3737–3740
3739
Table 4
Toxicity and body weight variation after ip administration of MG3 in mice
Dose (mg/kg)
No. of mice
No. of surviving mice
Mean body weight (g)
% Variation
Initial
After 5 days
After 13 days
% Variation
150
100
50
Control
4
4
4
4
2a
4b
4c
4
29.5
29.4
29.9
32.7
33.30
32.45
32.38
34.36
+12.9
+10.1
+8.3
36.25
36.03
36.12
38.40
+22.9
+22.2
+20.8
+17.5
+5.1
a
Two mice died within 5–10 min; the survivors exhibited shivering and torpor but they recovered in the following 2 h.
Exhibited sluggishness, recovering before 2 h.
Asymptomatic.
b
c
not susceptible to the resistance mechanism exhibiting a resis-
tance factor (ratio IC50 on CQ-R/ IC50 on CQ-S strains) equal 1.
Both compounds displayed a cytotoxicity on murine and human
cells that was comparable to that of CQ.
Acknowledgments
The financial support from the University of Milan (FIRST 2004-
06) is acknowledged.
The inhibition of b-hematin formation in the BHIA assay sug-
gests that both compounds interfere with the haem detoxification
process of parasites, thus acting with a mechanism similar to that
of CQ.
The in vivo studies, in the murine P. berghei malaria model,
showed that MG2 and MG3 behaved similarly to CQ in the single
dose treatment.
We thank Dr. Mauro Galli for skilled technical assistance in the
synthetic work and Christian Scheurer, Josefina Santo Tomas and
Christopher Snyder of the STI for assistance with the determination
of the antiplasmodial activity. We also thank the AVIS Association,
Milano for providing fresh A+ human blood from healthy donors.
References and notes
Repeated treatment (once a day for 4 consecutive days) with
MG2 and MG3 inhibited parasitemia by >99.9% when given at
10 mg/kg po and sc. Particularly, MG3 at 10 mg/kg po and sc pro-
duced a mean survival of 25.7 and 29.3 days, respectively and 2/
3 mice were cured (parasite-free on day 30).
For CQ diphosphate, at the same dose, the mean survival times
were 22.3 days (po) and 15.7 (sc), with no mice cured regardless of
the route of administration.
1. O’Neill, P. M.; Ward, S. A.; Berry, N. G.; Jeyadevan, J. P.; Biagini, G. A.; Asadollaly,
E.; Park, B. K.; Bray, P. G. Curr. Top. Med. Chem. 2006, 6, 479.
2. Sparatore, A.; Basilico, N.; Parapini, S.; Romeo, S.; Novelli, F.; Sparatore, F.;
Taramelli, D. Bioorg. Med. Chem. 2005, 13, 5338.
3. Lucantoni, L.; Sparatore, A.; Basilico, N.; Parapini, S.; Yardley, V.; Stewart, L.;
Habluetzel, A.; Pasqualini, L.; Esposito, F.; Taramelli, D. In: Third Cost B22
Annual Congress ‘Drug discovery and development for parasitic diseases’,
Athens, 1–4 October 2006, Book of abstracts, p. 137.
4. Miyano, S.; Sumoto, K.; Satoh, F.; Shima, K.; Hayashimatsu, M.; Morita, M.;
Aisaka, K.; Noguchi, T. J. Med. Chem. 1985, 28, 714.
5. Oka, M.; Matsumoto, Y.; Hirooka, K.; Suzuki, T. Chem. Pharm. Bull. 2000, 48,
1121.
6. Suzuki, T.; Usui, T.; Oka, M.; Suzuki, T.; Kataoka, T. Chem. Pharm. Bull. 1998, 45,
1265.
7. Suzuki, T.; Oka, M.; Maeda, K.; Furusawa, K.; Uesaka, H.; Takaoka, T. Chem.
Pharm. Bull. 1999, 47, 28.
8. Takeda, M.; Tsukamoto, K.; Yamano, M.; Uesaka, H. Jpn. J. Pharmacol. 1999, 292.
9. A mixture of 2-(hexahydro-1H-pyrrolizin-7a-yl)ethanamine10 or (hexahydro-
1H-pyrrolizin-7a-yl)methanamine11 (1.94 mmol), 4,7-dichloroquinoline
(384 mg, 1.94 mmol) and phenol (1.21 g, 13.58 mmol) was heated for 4 h at
180 °C, stirring under nitrogen. After cooling, the mixture was basified with 2 N
NaOH and the product extracted thoroughly with ether. The organic extracts
were washed with 2 N NaOH, then with water and finally with 5% acetic acid.
The acetic solution was alkalized with concd NH3 and extracted with ether.
After evaporation of the solvent, MG2 or MG3 as white solid were obtained.
Compound MG2: Yield: 66%; mp (Büchi): 123.5–125.5 °C (Et2O). 1H NMR
(Varian Mercury 300VX, CDCl3): d = 9.10 (br s, 1H, collapses with D2O); 8.45 (d,
J = 5.5 Hz, 1H); 7.90 (d, J = 1.9 Hz, 1H); 7.65 (d, J = 9.0 Hz, 1H); 7.35 (dd, J = 1.9,
8.8 Hz, 1H); 6.25 (d, J = 5.5, 1H); 3.40 to 3.30 (m, 2H); 3.20 to 3.00 (m, 2H); 2.80
to 2.60 (m, 2H); 2.00 to 1.60 (m, 10H). Anal. (Carlo Erba-EA-1110 CHNS-O
instrument) Calcd for C18H22N3Cl: C, 68.45; H, 7.02; N, 13.30, found: C, 68.62;
H, 7.14; N, 13.32. Compound MG3: Yield: 59.4%; mp: 112–113 °C (Et2O). 1H
NMR (CDCl3): d = 8.50 (d, J = 5.5 Hz, 1H); 7.95 (d, J = 1.9 Hz, 1H); 7.70 (d,
J = 9.0 Hz, 1H); 7.35 (dd, J = 1.9, 8.8 Hz, 1H); 6.40 (d, J = 5.5 1H); 6.00 (br s, 1H,
collapses with D2O); 3.15 (m, 4H); 2.80 to 2.60 (m, 2H); 2.00 to 1.60 (m, 10H).
HRMS (APEX II ICR-FTMS Bruker Daltonics, ESI) m/z calcd for C17H21N3Cl
[M+H]+ 302.14185, found 302.14187.
The in vivo acute toxicity and the effect on body weight of MG3
was evaluated in CD1 mice treated with a single dose ip and ob-
served for 13 days (Table 4): the toxic dose (less than 100% sur-
vival) was >100 mg/kg for MG3.
Finally, preliminary studies of the in vitro metabolism21 of MG3
and of its capability to inhibit cytochrome P450 isoforms22 were
undertaken.
The in vitro clearance of MG3 was 1.50 0.58 ml/min/g (human
hepatic microsomes), while that of 7-ethoxycoumarine, a common
reference standard, was in the range 7–9.9 ml/min/g. Moreover,
after 30 min of incubation, only small amounts of two metabolites
(with a MH+at m/z 300 and 316; parent compound MH+at m/z 302)
were detected at the HPLC-MS, thus supporting the expected sta-
bility of the pyrrolizidine ring to phase 1 oxidative metabolism.
The interaction of MG3 (3 lM) with CYP isoforms (1A2, 2C9,
2C19, 2D6, and 3A4) was very low; particularly the inhibition of
the last two isoforms (that together are responsible for the metab-
olism of about 60% of the most clinically important drugs) was only
5.7% and 4.3%, whereas CQ-induced inhibition was 20.4% and 6.8%,
respectively. Therefore MG3 should not interfere extensively with
the metabolism of other drugs.
In conclusion, the pyrrolizidinylalkyl derivatives of 4-amino-7-
chloroquinoline MG2 and MG3 exhibited excellent activity in vitro
against CQ-S and CQ-R strains of Pf and in vivo against P. berghei.
Thus the bicyclic pyrrolizidine moiety resulted, as already seen
for the quinolizidine, as a structural feature able to overcome the
Pf resistance mechanism.
Moreover, the novel compounds exhibited low toxicity against
mammalian cells, whereas in preliminary in vivo studies the most
active MG3 was also well tolerated up to a dose of 100 mg/kg.
These data combined with the absence of chiral centers and the
low cost of the intermediates required for its synthesis, make MG3
an interesting compound deserving further study as a promising
antimalarial agent.
10. Miyano, S.; Yamashita, O.; Sumoto, K.; Shima, K.; Hayashimatsu, M.; Satoh, F. J.
Heterocycl. Chem. 1987, 24, 271.
11. Miyano, S.; Yamashita, O.; Sumoto, K.; Shima, K.; Hayashimatsu, M.; Satoh, F. J.
Heterocycl. Chem. 1987, 24, 47.
12. Oka, M.; Baba, K.; Nakamura, L.; Dong, L.; Hamajima, H.; Unno, R.; Matsumoto,
Y. J. Heterocycl. Chem. 2003, 40, 177.
13. Oka, M.; Baba, K.; Suzuki, T.; Masumoto, Y. Heterocycles 1997, 45, 2317.
14. Makler, M. T.; Ries, J. M.; Williams, J. A.; Bancroft, J. E.; Piper, R. C.; Gibbins, B.
L.; Hinrichs, D. J. Am. J. Trop. Med. Hyg. 1993, 48, 739.
15. Desjardins, R. E.; Canfield, C. J.; Haynes, J. D.; Chulay, J. D. Antimicrob. Agents
Chemother. 1979, 16, 710.
16. Parapini, S.; Basilico, N.; Pasini, E.; Egan, T. J.; Olliaro, P.; Taramelli, D. Exp.
Parasitol. 2000, 96, 249.
17. D’Alessandro, S.; Gelati, M.; Basilico, N.; Parati, E. A.; Haynes, R. K.; Taramelli, D.
Toxicology 2007, 241, 66.
18. Donation from A. P. Waters and C. J. Janse, Leiden University.

3740
A. Sparatore et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3737–3740
19. Ridley, R. G.; Matile, H.; Jaquet, C.; Dorn, A.; Hofheinz, W.; Leupin, W.;
Masciadri, R.; Theil, F. P.; Richter, W. F.; Girometta, M. A.; Guenzi, A.; Urwyler,
H.; Gocke, E.; Potthast, J. M.; Csato, M.; Thomas, A.; Peters, W. Antimicrob.
Agents Chemother. 1997, 41, 677.
20. Peters, W.. In Chemotherapy and Drug Resistance in Malaria; Academic Press:
London, 1987; Vol. 1. pp 97–174.
21. Test compound at the final concentration of 1 lM was dissolved in DMSO and
pre-incubated for 10 min at 37 °C in potassium phosphate buffer pH 7.4, with
human microsomes (Xenotech) at the final concentration of 0.5 mg/ml. After
the pre-incubation period, the reaction was started by adding the cofactor
mixture (NADP, G6P, G6P-DH, MgCl2 3 mM and UDPGA 1 mM); samples were
check the chemical stability of test compound. 7-Ethoxycoumarin was used as
reference standard.
22. The inhibition of the P450 isoforms was measured using specific substrates
that become fluorescent upon CYP metabolism. Compounds, dissolved in
DMSO, are tested at least in triplicate (n = 3 or 4) at the single concentration of
3 lM in a 96-well plate containing incubation/NADPH regenerating buffer.
Specific isoenzymes and substrates are added and incubated at 37 °C. Reaction
was terminated at different times, depending on the assays, and plates are read
on a Fluoroskan Ascent at the appropriate emission/excitation wavelengths.
Concentration–response curves (seven concentrations) performed in duplicate
for known inhibitors for each isoenzyme are tested in every assay as positive
control. For each compound the percentage inhibition vs. control without
inhibitor is calculated. An inhibition <5% is considered no effect. For standards
the IC50 (concentration at 50% inhibition) is then determined by using Grafit v.
6.0.
taken at time
0 and 30 min and added to acetonitrile to stop reaction,
centrifuged and supernatant analyzed by LC–MS/MS (PremiereXE, Waters, in
full scan ESI pos mode). Acquired data were processed by Metabolynx
software. A control sample without cofactor was always added in order to