Highly efficient G-quadruplex recognition by bisquinolinium compounds

Article abstract of DOI:10.1021/ja067352b

Syntheses and telomeric G-quadruplex-DNA binding properties of novel bisquinolinium compounds are reported. This series exhibits remarkable efficiency both in terms of stabilization and selectivity, thus combining the performances of the most potent quadruplex binders reported so far. These bisquinolinium compounds then represent an ideal tradeoff between rapid synthetic access and efficient target recognition. The study also highlights important structural parameters that lead to the design of highly selective G-quadruplex binders. Copyright

Full text of DOI:10.1021/ja067352b

Published on Web 01/30/2007
Highly Efficient G-Quadruplex Recognition by Bisquinolinium Compounds
Anne De Cian,‡ Elsa DeLemos,† Jean-Louis Mergny,‡ Marie-Paule Teulade-Fichou,*,† and
David Monchaud†
Laboratoire de Chimie des Interactions Mole´culaires, Colle`ge de France, CNRS UPR285, 75005 Paris, France, and
Laboratoire de Biophysique, Muse´um National d’Histoire Naturelle, USM503, INSERM U565 CNRS UMR 5153,
75005 Paris, France
Received October 13, 2006; E-mail: mp.teulade-fichou@college-de-france.fr
Trying to control and regulate the expression of genes is
emerging as a very appealing anticancer strategy. Indeed, chemo-
therapy aiming at designing molecules able to interact with unusual
structures of nucleic acids is currently subjected to a close
examination.1 In that sense, quadruplex-DNA is a particularly
attractive high-order structure since it appears to be putatively
present in pivotal genomic regions such as telomeres,2 promoters
of oncogenes, and most growth control genes.3 Convincing reports
on the efficiency of quadruplex interacting molecules as therapeuti-
cally active agents are beginning to appear in the literature.4
Thus, intensive investigations are currently oriented toward the
design and development of new G-quadruplex ligands. Of particular
interest are molecules with high quadruplex affinity that exhibit
Figure 1. (A) Alternative conformations of pyridodicarboxamide unit (H-
bonds appear as dotted lines). (B) Structure of 1a/b and 2a/b (with triflate
the ability to discriminate quadruplex-DNA from nucleus pre-
dominant duplex-DNA. Up to now, the leading G-quadruplex binder
has been the natural product telomestatin.5 Nevertheless, its total
synthesis was achieved only very recently, and the complexity of
the process renders its convenient exploitation difficult.6 Recently,
some of us reported on bisquinolinium compounds that exhibit excep-
tional affinity and selectivity for quadruplex-forming oligonucleo-
tides.4c,7 The anti-proliferative activity of these compounds has been
demonstrated, as well as their preferential binding to telomeric
regions of human chromosomes,7b thus providing new insights on
quadruplex existence in vivo. These exciting results prompted us
to develop the bisquinolinium family of compounds which was
furthermore facilitated by a convenient and rapid synthetic access.
In the initial series, the two quinolinium moieties are connected
through a 2,6-pyridodicarboxamide unit.4c,7 This motif, well-known
to adopt an internally organized H-bonded syn-syn conformation
(Figure 1A),8 was shown to be critical for quadruplex binding.
Indeed, inversion of the amide connectivity leads to a loss of affinity
(Supporting Information).4c This suggests that the central pyridodi-
carboxamide unit locks the ligand in a crescent-shaped conformation
highly favorable for G-quartet overlap. On this basis, we reasoned
that expanding the aromaticity of the central core without disrupting
either the H-bonds network or modifying the cationic side-arm
nature could result in improved recognition properties.
To this end, 6,6′-disubstituted-2,2′-bipyridine and 2,9-disubsti-
tuted-1,10-phenanthroline units that are also susceptible to be
conformationally locked via H-bonding were considered as good
candidates for replacing the pyridine central core. Ligands 1a/b
and 2a/b (Figure 1B) were synthesized via straightforward three-
and four-step procedures respectively, from inexpensive com-
mercially available material (Supporting Information).
The ligand’s ability to stabilize a quadruplex structure was
evaluated by FRET experiments using the quadruplex-forming
engineered oligonucleotide F21T (FAM-G3[T2AG3]3-Tamra), which
as counterions). (C) Dose-response curves for FRET results (∆T1/2) in
function of ligand concentration for 1a (green), 1b (orange), 2a (blue), and
2b (red). The baseline (∆T1/2 ) 0 °C) was set at the melting temperature
of the structure without ligand (48 °C).
mimics the human telomeric repeat (Supporting Information).9 As
shown in Figure 1C, compounds 2a/b (blue/red curves) appear as
remarkably strong quadruplex stabilizers, while more modest effects
are obtained with 1a/b (green/orange curves). Indeed, ∆T1/2 values
at 1 µM dose of ligand (vertical gray line, Figure 1C) are
significantly higher with 2a/b (∆T1/2(1µM) ) 29.7 and 28.5 °C,
respectively) than with 1a/b (∆T1/2(1µM) ) 15.2 and 9.6 °C,
respectively). Accordingly, the concentration required to achieve
a ∆T1/2 value of 20 °C (horizontal gray line, Figure 1C) is also
significantly lower with 2a/b ([conc](∆T )20°C) ) 0.48 and 0.61
1/2
µM, respectively) than with 1a/b ([conc](∆T )20°C) ) 1.19 and 1.69
1/2
µM, respectively).10 Altogether, these data reflect a very high level
of quadruplex stabilization for ligands 2 and represent a significant
improvement as compared to the pyridine series. Interestingly, these
results make this new series fully competitive with the high-affinity
G-quadruplex binders such as telomestatin, extended acridines, and
organometallic complexes all exhibiting ∆T1/2(1µM) > 20 °C.10 Most
important, the differences between the two series highlight that
structural rigidity is a key parameter for quadruplex recognition,
the free rotation around the biaryl axis of ligands 1 being responsible
for the lower performance of this series.
To gain further insights into the intrinsic qualities of these
ligands, competitive FRET experiments were performed in the
presence of various amounts of 26 bp duplex-DNA (ds26, from 0
to 10 µM, Supporting Information).11 Remarkably, the thermal
stabilization induced by 2a/b is only poorly affected (∼10% loss)
by the presence of 10 molar equiv of ds26 (Figure 2A,B). This
indicates that ligands 2 exhibit an exquisite quadruplex versus
duplex selectivity, thus behaving similarly to telomestatin in the
same conditions (Supporting Information).12 Again, bipyridine
derivatives 1 appear less competent since they are more sensitive
† Colle`ge de France.
‡ Muse´um National d’Histoire Naturelle.
9
1856
J. AM. CHEM. SOC. 2007, 129, 1856-1857
10.1021/ja067352b CCC: $37.00 © 2007 American Chemical Society

C O M M U N I C A T I O N S
preliminary data were obtained which suggest that inhibition
measured by TRAP does not actually reflect telomerase inhibition
but may result from an inhibition of the PCR amplification on
quadruplex-prone motifs even though the internal PCR control
(ITAS) is not affected (Supporting Information). Detailed explana-
tions will be reported elsewhere. Nevertheless, the obtained results,
among the best reported to date, represent a ∼20-fold improvement
as compared to the parent pyridine series4c,7 and may reflect
somehow the very good affinity of those ligands for quadruplex-
prone motifs, confirming FRET melting results.
In conclusion, the present paper describes the quadruplex-binding
properties of new members of the bisquinolinium family. Their easy
synthetic access combined with exceptional quadruplex affinities
and selectivities place these ligands among the most potent ones
reported so far. Further in vitro evaluations are currently underway.
Acknowledgment. This work was supported by ARC (#3365)
and E.U. FP6 “MolCancerMed” (LSHC-CT-2004-502943) grants.
The authors gratefully thank Dr. P. Mailliet for helpful discussions.
Figure 2. (A) FRET experiments carried out with 2a and F21T without
ligand (black) and with 1 µM 2a in absence (blue) or presence of competitive
duplex (ds26, 3 equiv (red) or 10 equiv (gray)). (B) FRET results (∆T1/2
,
Supporting Information Available: Synthesis and characterization
of 1a-2b; experimental procedures and additional FRET and CD data.
This material is available free of charge via the Internet at http://
pubs.acs.org.
°C) for 1a/b and 2a/b (1 µM) in absence (blue) or presence of competitive
ds26 (3 equiv (red) or 10 equiv (gray)). (C and D) CD titration of 22AG (3
µM in 10 mM lithium cacodylate, pH 7.2, 100 mM NaCl buffer) by
increasing amounts of 2a: (C) some CD spectra at 296 nm from the titration
experiment; the arrow indicates the increasing amounts of ligand (from black
to red curves: 0, 1.2, 2.4, 3.6, 4.7, 5.9, 7.1, 8.2, 9.4, and 10.9 µM); (D) CD
signal as a function of 2a molar equivalents (0.4 µM increments).
References
(1) Wheelhouse, R. T.; Jennings, S. A.; Phillips, V. A.; Pletsas, D.; Murphy,
P. M.; Garbett, N. C.; Chaires, J. B.; Jenkins, T. C. J. Med. Chem. 2006,
49, 5187.
(2) Neidle, S.; Parkinson, G. N. Curr. Opin. Struct. Biol. 2003, 13, 275.
(3) (a) Cogoi, S. C.; Xodo, L. E. Nucleic Acids Res. 2006, 34, 2536. (b)
Dexheimer, T. S.; Sun, D.; Hurley, L. H. J. Am. Chem. Soc. 2006, 128,
5404. (c) Dai, J.; Dexheimer, T. S.; Chen, D.; Carver, M.; Ambrus, A.;
Jones, R. A.; Yang, D. J. Am. Chem. Soc. 2006, 128, 1096. (d) Xu, Y.;
Sugiyama, H. Nucleic Acids Res. 2006, 34, 949. (e) De Armond, R.; Wood,
S.; Sun, D.; Hurley, L. H.; Ebbinghaus, S. W. Biochemistry 2005, 44,
16341. (f) Phan, A. T.; Kuryavyi, V.; Gaw, H. Y.; Patel, D. J. Nat. Chem.
Biol. 2005, 1, 167. (g) Sun, D.; Guo, K.; Rusche, J. J.; Hurley, L. H.
Nucleic Acids Res. 2005, 33, 6070. (h) Rankin, S.; Reszka, A. P.; Huppert,
J.; Zloh, M.; Parkinson, G. N.; Todd, A. K.; Ladame, S.; Balasubramanian,
S.; Neidle, S. J. Am. Chem. Soc. 2005, 127, 10584.
Figure 3. Selected dimensions of a G-quartet (left, determined from X-ray
structure) and of 2a (right, after molecular mechanics (MM2) calculations
(Chem3D Ultra 8.0, CambridgeSoft Corp., MA)).
(4) (a) Burger, A. M.; Dai, F.; Schultes, C. M.; Reszka, A. P.; Moore, M. J.
B.; Double, J. A.; Neidle, S. Cancer Res. 2005, 65, 1489. (b) Kim, M.
Y.; Vankayalapati, H.; Shin-Ya, K.; Wierzba, K.; Hurley, L. H. J. Am.
Chem. Soc. 2002, 124, 2098. (c) Pennarun, G.; Granotier, C.; Gauthier,
L. R.; Gomez, D.; Hoffschir, F.; Mandine, E.; Riou, J.-F.; Mergny, J.-L.;
Mailliet, P.; Boussin, F. Oncogene 2005, 24, 2917.
(5) (a) Shin-ya, K.; Wierzba, K.; Matsuo, K.-I.; Ohtani, T.; Yamada, Y.;
Furihata, K.; Hayakawa, Y.; Seto, H. J. Am. Chem. Soc. 2001, 123, 1262.
(b) Tahara, H.; Shin-ya, K.; Seimiya, H.; Yamada, H.; Tsuruo, T.; Ide, T.
Oncogene 2006, 25, 1955. (c) Gomez, D.; O’Donohue, M. F.; Wenner,
T.; Douarre, C.; Macadre, J.; Koebel, P.; Giraud-Panis, M. J.; Kaplan,
H.; Kolkes, A.; Shin-ya, K.; Riou, J.-F. Cancer Res. 2006, 66, 6908.
(6) Doi, T.; Yoshida, M.; Shin-ya, K.; Takahashi, T. Org. Lett. 2006, 8, 4165.
(7) (a) Lemarteleur, T.; Gomez, D.; Paterski, R.; Mandine, E.; Mailliet, P.;
Riou, J.-F. Biochem. Biophys. Res. Commun. 2004, 323, 802. (b) Granotier,
C.; Pennarun, G.; Riou, L.; Hoffshir, F.; Gauthier, L. R.; De Cian, A.;
Gomez, D.; Mandine, E.; Riou, J.-F.; Mergny, J.-L.; Mailliet, P.;
Dutrillaux, B.; Boussin, F. D. Nucleic Acids Res. 2005, 33, 4182.
(8) Berl, V.; Huc, I.; Khoury, R. G.; Krische, M. J.; Lehn, J.-M. Nature 2000,
407, 720.
(9) Mergny, J.-L.; Maurizot, J.-C. ChemBioChem 2001, 2, 124.
(10) (a) Schultes, C. M.; Guyen, B.; Cuesta, J.; Neidle, S. Bioorg. Med. Chem.
Lett. 2004, 14, 4347. (b) Moore, M. J. B.; Schultes, C. M.; Cuesta, J.;
Cuenca, F.; Gunaratnam, M.; Tanious, F. A.; Wilson, W. D.; Neidle, S.
J. Med. Chem. 2006, 49, 582. (c) Reed, J. E.; Arnal, A. A.; Neidle, S.;
Vilar, R. J. Am. Chem. Soc. 2006, 128, 5992.
(11) Mergny, J.-L.; Lacroix, L.; Teulade-Fichou, M.-P.; Hounsou, C.; Guittat,
L.; Hoarau, M.; Arimondo, P. B.; Vigneron, J. P.; Lehn, J.-M.; Riou, J.-
F.; Garestier, T.; He´le`ne, C. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 3062.
(12) De Cian, A.; Guittat, L.; Shin-ya, K.; Riou, J.-F.; Mergny, J.-L. Nucleic
Acids Symp. Ser. 2005, 49, 235.
to the duplex competition (29 and 35% loss in stabilization for 1a/
b, respectively (Figure 2B).
To determine the stoichiometry of association of 2a and
G-quadruplexes, CD titrations were carried out using the 22AG
sequence (AG3[T2AG3]3, in Na+ buffer, Figure 2C and Supporting
Information).13 Analysis of the data indicates that the curve
inflection occurs at ∼2:1 ligand/quadruplex ratio (Figure 2D). This
2:1 stoichiometry is consistent with a binding mode based on the
stacking of the ligand onto the two external G-quartets of the
quadruplex (Supporting Information).
A close examination of the crystal structure of 22AG (1KF1,
RCSB Protein Data Bank)14 shows that a G-quartet can be
considered as a square aromatic surface whose dimensions are
closely related to that of 2a (Figure 3 and Supporting Information).
The strong stabilization properties of ligands 2 could thus originate
in this accurate geometrical complementarity. Subsequently, the
molecular size may be unfavorable for interaction with a classical
base pair in duplex DNA,1 resulting in the high preference for the
quadruplex. Consequently, the molecular organization of the central
core (internal H-bonds) and electronic/electrostatic properties (two
quinolinium side arms) make phenanthroline bisquinolinium deriva-
tives 2 perfectly fitted for the recognition of the quadruplex target.
Finally, the inhibitory properties of these ligands were evaluated
via a classical TRAP assay (Supporting Information). However,
(13) Allain, C.; Monchaud, D.; Teulade-Fichou, M.-P. J. Am. Chem. Soc. 2006,
128, 11890.
(14) Parkinson, G. N.; Lee, M. P. H.; Neidle, S. Nature 2002, 417, 876.
JA067352B
9
J. AM. CHEM. SOC. VOL. 129, NO. 7, 2007 1857