Synthesis and anti-HIV properties of new hydroxyquinoline-polyamine conjugates on cells infected by HIV-1 LAV and HIV-1 BaL viral strains

Article abstract of DOI:10.1016/j.bmcl.2006.08.114

To find new derivatives that block different virus strains entry in cells bearing specific surface receptors represent an interesting challenge for medicinal chemists. Here, we report the synthesis and the anti-HIV properties of a new series of analogues based on the introduction of quinoline moiety on various polyamine backbones, including polyazamacrocycles. Three compounds 7, 8, and 10 of this series were found active on PBMCs cells infected by HIV-1 LAV or by HIV-1 BaL, in contrast the well-known reference compound 1a (AMD 3100) was found only active on HIV-1 LAV strain.

Full text of DOI:10.1016/j.bmcl.2006.08.114

Bioorganic & Medicinal Chemistry Letters 16 (2006) 5988–5992
Synthesis and anti-HIV properties of new hydroxyquinoline–
polyamine conjugates on cells infected by HIV-1 LAV
and HIV-1 BaL viral strains
Vincent Moret,a Nathalie Dereudre-Bosquet,b Pascal Clayette,b Younes Laras,a
Nicolas Pietrancosta,a Amandine Rolland,a Clement Weck,a Sylvain Marca and
Jean-Louis Krausa,*
aLaboratoire de Chimie Biomole´culaire, UMR-CNRS 6216, IBDML, Universite´ de la Me´diterrane´e,
Parc Scientifique de Luminy, case 901, 13288 Marseille cedex 9, France
bSPI-BIO, CEA, Pharmacologie des re´trovirus, 18 route du Panorama, BP6, 9226 Fontenay aux Roses, France
Received 27 July 2006; revised 29 August 2006; accepted 30 August 2006
Available online 26 September 2006
Abstract—To find new derivatives that block different virus strains entry in cells bearing specific surface receptors represent an inter-
esting challenge for medicinal chemists. Here, we report the synthesis and the anti-HIV properties of a new series of analogues based
on the introduction of quinoline moiety on various polyamine backbones, including polyazamacrocycles. Three compounds 7, 8,
and 10 of this series were found active on PBMCs cells infected by HIV-1 LAV or by HIV-1 BaL, in contrast the well-known ref-
erence compound 1a (AMD 3100) was found only active on HIV-1 LAV strain.
Ó 2006 Elsevier Ltd. All rights reserved.
Viruses that are broadly resistant to currently available
anti-retroviral medications represent a growing chal-
lenge for HIV therapy. Unlike many existing HIV drugs
that target the virus after it has infected a healthy cell, it
had been shown that compounds like AMD3100 1a, its
analogue 1b (Fig. 1) and other related analogues block
the virus from entering a healthy cell preventing the rep-
lication process. These compounds block cell surface G-
protein coupled receptors, such as CCR5 or CXCR4.1–3
AMD 3100 has reached phase II clinical trials and is
presently considered as a prototype of this class of com-
pounds. A recent review on these analogues has been
published by Hatse et al.4 Nevertheless, the existence
of two azamacrocycles is not a prerequisite for anti-
HIV activity.
CXCR4 and CCR5 receptors can prevent the relevant
HIV strains from entering and infecting the target cells.
An infected individual cell may harbor different levels of
both CXCR4 and CCR5. Research on HIV disease sug-
gests that approximately one-half of patients are infect-
ed with HIV strains that use CXCR4 as an entry
receptor as well as strains that use CCR5.
AMD3100 1a is extremely specific in its affinity for the
CXCR4 receptor, and this property depends, at least
in part, on an electrostatic interaction between the basic
(positively charged) nitrogens of the cyclam moieties
and the acid (negatively charged) carboxylates of the
aspartic acid residues located at positions 171, 182,
193, and 262 of the CXCR4 receptor. In particular,
In order to enter and infect cells, HIV must bind to
either CXCR4 or CCR5 chemokine receptors. Different
HIV strains prefer one receptor to the other. Blocking
NH HN
NH HN
NH
NH HN
HN
N
HN
NH
N
N
N
Keywords: Polyazamacrocycles; 8-Hydroxyquinoline; HIV-1 LAV;
HIV-1 Bal; CXCR4 receptor; CCR5 receptor.
NH HN
1a
1b
*
Corresponding author. Tel.: +33 4 91 82 91 41; fax: +33 4 91 82 94
16; e-mail: kraus@luminy.univ-mrs.fr
Figure 1. AMD 3100 structure 1a and a related analogue 1b.
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.08.114

V. Moret et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5988–5992
5989
the aspartate residues 171 and 262, located at the junc-
tion of the transmembraneous segments with the extra-
cellular loops of the CXCR4, have proven to be
crucial in the binding of 1a with its receptors.5 Analogue
1a affinity for CXCR4 can be enhanced by its complex-
ation with Cu2+, Zn2+ or Ni2+ metal ions6,7 and this en-
hanced affinity is explained through a better interaction
with the aspartate in position 262 (For additional details
on the anti-HIV activities of polyazamacrocycle metal
complexes, see references Hatse et al.4 and Princen
et al.8). These observations led us to investigate the pos-
sibility to design new anti-HIV drugs incorporating in
their structure specific ligands with transition metal che-
lating properties, which could block virus entry in
healthy cells mediated either by CXCR4 or CCR5 recep-
tors. As far as, 8-hydroxy-quinolines were known to
form tetradentate complexes with transition metals such
as copper and zinc with good binding affinity values9
(logKs values being around 10) they were selected as
suitable ligands for the design of new chemokine recep-
tor antagonists. As far as clioquinol 2 (5-chloro-7-iodo-
8-hydroxy-quinoline), a drug in clinical trials for Alzhei-
mer’s disease treatment, being known to form zinc and
copper complexes.10 We attempt to introduce 8-hy-
droxy-quinoline moieties on different polyamine back-
bones, to give rise to mono- or di-quinoline analogues
of general structures represented in Figure 2. Since
length and chemical structure of the linker joining the
two polyazamacrocycles in AMD3100 were determina-
tive for the anti-HIV activity, it was of interest to use
different diamines, polyamines or related conjugates as
clioquinol linkers. Finally, we report the synthesis of
new polyamine-conjugate and their anti-HIV properties.
Polyamine can be either: (a) a monocyclam or bicyclam
AMD 3100 or (b) the diamines: 1,3-diamino propane,
1,4-diaminomethylbenzene, and piperazine.
The new analogues were synthesized according to meth-
odologies summarized in Schemes 1 and 2.
The whole new analogues were obtained starting from a
common intermediate, chloromethyl quinoline 3 which
was synthesized from chloromethylation of commercial-
ly available quinoline, according to a method described
by Zheng.11
Several diamino linkers 1,3-diamino propane, 1,4-diami-
no xylene, piperazine, 1,4,8-tri-terbutoxy carbonyl
1,4,8,11-tetraazacyclotetradecane, 1,8,12-tri-terbutoxy
carbonyl 1,4,8,12-tetraazacyclopentadecane, and penta
Boc-protected bis-polyazamacrocycles 10d (The synthet-
ic scheme for 10d starting from cyclam is shown on
Scheme 2) were condensed on intermediate 3, in acetoni-
trile as solvent in the presence of K2CO3 as base. Final
compounds (4, 5, and 6), were obtained, respectively,
in 55%, 60%, and 50% yields, while the Boc-protected
analogues 7a and 8a were obtained in 45% and 55%
yields. After Boc-deprotection in acidic media, the final
corresponding analogues (7 and 8) were isolated in
quantitative yields (Scheme 1).
The synthesis of bis-polyazamacrocycle–quinoline con-
jugate 10 required a specific synthetic route summarized
in Scheme 2. Final compound 10 was obtained through
the synthesis of the key intermediate 1,11-diBoc-poly-
azamacrocycle 10b. This later was synthesized starting
from cyclam (three equivalents) by addition of one
equivalent of di terbutyldicarbonate reagent. After a
Linker
R
X
N
N
N
OH
OH
OH
H
H
2 R=Cl, X=I (clioquinol)
N
N
4 Linker =
5 Linker =
3 R=CH2Cl, X=H (chloromethyl quinoline)
N
N
N
H
H
N
6 Linker =
NH HN
NH HN
N
HN
N
NH HN
(CH2)n
NH
N
N
N
NH
NH
N
N
N
N
N
OH
OH
OH
OH
7 n=2
8 n=3
10
9
Figure 2. General structures of the new synthesized polyamine–quinoline conjugates.

5990
V. Moret et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5988–5992
N
N
OH
H
H
N
N
N
1
HO
N
a
5
N
N
N
OH
OH
4
OH
Cl
b
d
e
Boc
Boc
Boc
N
N
N
(CH2)n
N
H
c
OH
N
N
H
3
N
HO
g,h
N
N
N
6
OH
OH
NH HN
7a (n=2), 8a (n=3)
7, 8
f
N
N
N
N
9
OH
OH
Scheme 1. Reagents and conditions: (a) HCHO, HCl, 0 °C, overnight; (b) diamino propane–K2CO3, CH3CN, rt, overnight; (c) a,a0-diamino xylene,
CH3CN, rt, 3 days; (d) piperazine, K2CO3, CH3CN, rt, 3 days; (e) 1,4,8-tri-terbutoxy carbonyl-1,4,8,11-tetraazacyclotetradecane or 1,8,12-tri-
terbutoxy carbonyl-1,4,8,12-tetraazacyclopentadecane, DIEA, CH2Cl2, overnight, rt; (f) HCl, ether, rt; (g) analogue 10b (Scheme 2), DIEA, CH2Cl2,
overnight, rt; (h) HCl, ether, rt, overnight.
Br
Br
Boc
Boc
Boc
Boc
Boc
Boc
Boc
N
N
N
N
N
N
N
a
c
HN
Br
10a
10c
NH HN
NH HN
Boc
N
N
10b
d
NH HN
b
10b
NH HN
NH HN
Boc
Boc
Boc
Boc
Boc
N
N
N
N
N
N
N
3
NH
N
N
N
e, f
HN
N
10
10d
OH
Scheme 2. Reagents and conditions: (a), (b) Boc2O, CH2Cl2, rt, overnight; (c) K2CO3, DMF, 3 days, rt; (d) 10b, K2CO3, CH3CN; (e) 3, K2CO3,
CH3CN, rt; (f) HCl, ether, rt.
tedious column chromatography, 1,11-diBoc-protected
10b was isolated in 20% overall yield and its structure
was unambiguously assigned through NMR studies
consistently with NMR published data.12 The other pos-
sible expected analogue (1,8-diBoc-polyazamacrocycle)
which differs from analogue 10b (1,11-diBoc-polyaza-
macrocycle), only by the position of the Boc protecting
groups on the polyazamacrocyclic core, was isolated in
very low yield (less than 5%) (not shown in Scheme 2).
Condensation of one equivalent of the triprotected tetra-
azamacrocycle 10a already described in the literature13
with one equivalent of a,a0-dibromoxylene led to inter-
mediate 10c, which after condensation on the diBoc-
protected polyaza macrocycle 10b gave intermediate

V. Moret et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5988–5992
Table 1. Anti-HIV activities of the new conjugates on BaL and LAV HIV strains
5991
a
Compound
HIV-1 LAV EC50(lM)
HIV-1 BaL EC50(lM)
CC50 (lM)
1a
1b
0.47
0.05
Inactive
Inactive
Inactive
Inactive
Inactive
Inactive
Inactive
Inactive
Inactive
4.35
>10
>10
2
2
Inactive
5.10
3
2,3,2,3-Polyazamacrocycle
>10
>10
>10
>10
>10
>10
>10
>10
8
Inactive
Inactive
Inactive
Inactive
Inactive
4.10
2,3,3,3-Polyazamacrocycle
4
5
6
7
8
1.50
4.10
1.25
toxic
9
10
0.70
3.50
>10
EC50: concentration in lM required to inhibit HIV RT activity by 50%.
a CC50: concentration in lM to cause 50% death of uninfected cells. The highest tested concentration being 10 lM.
10d (overall yield 40%). Condensation of this later on
chloromethyl quinoline 3 led after Boc-standard depro-
tection to the desired analogue 10 (overall yield 60%).
The structures of the whole final analogues were clearly
identified by spectroscopic methods: 1H NMR, 13C
NMR, and MS (data available on request).
respective spacers, were totally denied of any anti-HIV
activity. Moreover monopolyazamacrocycles such as cy-
clam (2,3,2,3) or its larger analogue (2,3,3,3) were also
ineffective on both HIV virus strains. Clioquinol 2 it self
was found inactive.
Since it is known that in specific conditions, CD4+ cells
express both CXCR4 and CCR5 mRNA co-receptors,18
it could be possible that PBMCs used in this screening
express both co-receptors, in this hypothesis the struc-
ture of new analogues allows interactions with both
receptors, which results in both cases to a blockade of
virus entry. BaL strains use CCR5 receptors to penetrate
specific cells while HIV-1 LAV strains use more specifi-
cally CXCR4 receptors to enter in MT4 cells (classifica-
tion of different HIV strains for the cells expressing
different kind of receptors has been given by Princen
et al.19). At present, we have no evidence that introduc-
tion of quinoline moiety on the bis-polyazamacrocycle
backbone allows interactions with both CXCR4 and
CCR5. It could be hypothesized that the bis-polyazama-
cycle moiety interacts with CXCR4 receptors, while
quinoline moiety allows interactions with CCR5 co-re-
ceptors. Nevertheless, this hypothesis of inhibitors en-
dowed with a dual activity on both types of receptors
(CXCR4 and CCR5) could be a new concept of interest,
if it can be confirmed by additional studies. Moreover
such HIV drugs could find therapeutic application in
the case of AIDS-related dementia complex (ADC) very
often caused by HIV-1 BaL strain infections. Indeed it is
known that productive HIV infection in the CNS is
limited to macrophages and microglial cells, which con-
tribute to the production of a number of putative neuro-
toxins including quinolinic acid.20 Elevated levels of
quinolinic acid are observed in vivo in cerebrospinal flu-
id of patients with AIDS-related dementia complex
(ADC).21 In this feature, such new analogues after opti-
mization could be of potential value for ADC treatment.
The anti-HIV activities of the whole series of com-
pounds were assayed according to previously described
methods.14,15 PBMCs were infected with two virus
strains: HIV-1 LAV15 and HIV-1 BaL.16 It was of inter-
est to assay the new analogues against these two strains
which use two different chemokines (GPCRs, CXCR4,
and CCR5) to enter the cells. Results are expressed as
means of 50% effective concentrations (EC50) and cyto-
toxic concentrations (CC50) represent concentrations re-
quired to cause 50% death of uninfected PBMCs. The
results are presented in Table 1. Clioquinol 2, 5-chloro-
methyl-8-hydroxy-quinoline 3, and bis-polyazamacrocy-
cle 1a were used as benchmarking compounds for
comparative anti-HIV activity with that of the new ana-
logue antiviral activity.
As it can be observed in Table 1, compounds 7, 8, and 10
show anti-HIV activities on both strains HIV-1 LAV and
HIV-1 BaL with EC50 values ranging from 0.5 to 5 lM
(LAV strain) and from 1.2 to 5 lM (BaL strain). The
case of compound 9 should be dissociated from the other
active derivatives since its EC50 value is very close to its
CC50 value, it cannot be considered as an active com-
pound. In contrast, Biscyclam 1a or AMD 3100 or its
analogue 1b was found totally inactive on HIV-1 BaL
strain. This result was quite interesting since these bis-
polyazamacrocycles are reported specific CXCR4 antag-
onist prototypes. They do not interact with any other
CXCR or CCR receptors, they block X4 HIV-1 replica-
tion only through CXCR4 antagonization.17
It can be also observed that the structure of diamino
spacers on which the quinoline moiety is linked appears
to be determinative for the HIV-activities on both viral
strains. Compounds 4, 5, and 6, including 1,3-diamino-
propane, piperazine or 1,4-diaminomethylbenzene as
Nevertheless, other action mechanisms for these new
drugs cannot be ruled out since the new analogues can
also interact with other targets within the viral replica-
tion cycle.

5992
V. Moret et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5988–5992
7. Hunter, T. M.; McNae, I. W.; Liang, X.; Bella, J.;
Acknowledgments
Parsons, S.; Walkinshaw, M. D.; Sadler, P. J. Proc. Natl.
Acad. Sci. U.S.A. 2005, 102, 2288.
8. Princen, K.; Hatse, S.; Vermeire, K.; Aquaro, S.; De
Clercq, E.; Gerlach, L. O.; Rosenkilde, M.; Schwartz, T.
W.; Skerlj, R.; Bridger, G.; Schols, D. J. Virol. 2004, 78,
12996.
The authors thank L. Charlier for technical assistance.
´
Conseil Regional PACA (Provence Alpes Cote d’azur),
INSERM (Institut National de la Sante et de la Recher-
ˆ
´
´
che Medicale), and ANRS (Agence Nationale pour la
Recherche sur le SIDA) are greatly acknowledged for
9. Barnham, K. J.; Masters, C. L.; Bush, A. I. Nat. Rev. Drug
Discov. 2004, 3, 205.
´
financial support. Professor K. Dudley (Universite de
la Mediterranee, IBDML) is greatly acknowledged for
´
´
10. Di Vaira, M.; Bazzicalupi, C.; Orioli, P.; Messori, L.;
Bruni, B.; Zatta, P. Inorg. Chem. 2004, 43, 3795.
11. Zheng, H.; Weiner, L. M.; Bar-Am, O.; Epsztejn, S.;
Cabantchik, Z. I.; Warshawsky, A.; Youdim, M. B.;
Fridkin, M. Bioorg. Med. Chem. 2005, 13, 773.
12. Boitrel, B.; Andrioletti, B.; Lachkar, M.; Guilard, R.
Tetrahedron Lett. 1995, 36, 4995.
proofreading the manuscript.
Supplementary data
Supplementary data associated with this article can be
found in the online version at doi:10.1016/j.bmcl.
2006.08.114.
13. Dessolin, J.; Galea, P.; Vlieghe, P.; Chermann, J. C.;
Kraus, J. L. J. Med. Chem. 1999, 42, 229.
14. Barre-Sinoussi, F.; Chermann, J. C.; Rey, F.; Nugeyre, M.
T.; Chamaret, S.; Gruest, J.; Dauguet, C.; Axler-Blin, C.;
Vezinet-Brun, F.; Rouzioux, C.; Rozenbaum, W.; Mon-
tagnier, L. Science 1983, 220, 868.
15. Pauwels, R.; Balzarini, J.; Baba, M.; Snoeck, R.; Schols,
D.; Herdewijn, P.; Desmyter, J.; De Clercq, E. J. Virol.
Methods 1988, 20, 309.
References and notes
1. De Clercq, E. Antiviral. Res. 2005, 67, 56.
2. De Clercq, E. Int. J. Biochem. Cell Biol. 2004, 36, 1800.
3. Hatse, S.; Princen, K.; Gerlach, L. O.; Bridger, G.;
Henson, G.; De Clercq, E.; Schwartz, T. W.; Schols, D.
Mol. Pharmacol. 2001, 60, 164.
4. Hatse, S.; Princen, K.; De Clercq, E.; Rosenkilde, M. M.;
Schwartz, T. W.; Hernandez-Abad, P. E.; Skerlj, R. T.;
Bridger, G. J.; Schols, D. Biochem. Pharmacol. 2005, 70,
752.
16. Gartner, S.; Markovits, P.; Markovitz, D. M.; Kaplan, M.
H.; Gallo, R. C.; Popovic, M. Science 1986, 233, 215.
17. De Clercq, E. Nat. Rev. Drug Discov. 2003, 2, 581.
18. Mosoian, A.; Klotman, M.E. In 5th conference on retro-
viruses and opportunistic infections February 1–5, 1998.
19. Princen, K.; Schols, D. Cytokine Growth Factor Rev. 2005,
16, 659.
20. Nottet, H. S.; Flanagan, E. M.; Flanagan, C. R.; Gelbard,
H. A.; Gendelman, H. E.; Reinhard, J. F., Jr. J.
Neurovirol. 1996, 2, 111.
5. Gerlach, L. O.; Skerlj, R. T.; Bridger, G. J.; Schwartz, T.
W. J. Biol. Chem. 2001, 276, 14153.
6. Inouye, Y.; Kanamori, T.; Sugiyama, M.; Yoshida, T.;
Koike, T.; Shionoya, M.; Enomoto, K.; Suehiro, K.;
Kimura, E. Antiviral Chem. Chemother. 1995, 6, 337.
21. Sei, S.; Saito, K.; Stewart, S. K.; Crowley, J. S.; Brouwers,
P.; Kleiner, D. E.; Katz, D. A.; Pizzo, P. A.; Heyes, M. P.
J. Infect. Dis. 1995, 172, 638.