Structure-based design of selective high-affinity telomeric quadruplex-binding ligands

Article abstract of DOI:10.1039/c0cc02917c

A library of triazole-based telomeric quadruplex-selective ligands has been developed that mimic an established family of tri-substituted acridine-based ligands, using crystal structure data as a starting-point for computer-based design. Binding affinities, estimated by electrospray mass spectrometry, are in accord with the design concept.

Full text of DOI:10.1039/c0cc02917c

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Structure-based design of selective high-affinity telomeric
quadruplex-binding ligandsw
a
a
a
b
´
Caterina Maria Lombardo, Iria Sanchez Martınez, Shozeb Haider, Valerie Gabelica,
´
´
Edwin De Pauw,b John E. Mosesc and Stephen Neidle*a
Received 30th July 2010, Accepted 6th October 2010
DOI: 10.1039/c0cc02917c
A
library of triazole-based telomeric quadruplex-selective
ligands has been developed that mimic an established family of
tri-substituted acridine-based ligands, using crystal structure
data as a starting-point for computer-based design. Binding
affinities, estimated by electrospray mass spectrometry, are in
accord with the design concept.
Oligonucleotides and nucleic acids containing G-tracts can be
organised as G-quadruplexes.1 These are polymorphic tertiary
structures, characterised by a hydrophobic core of G-quartets
and negatively-charged loops. Quadruplexes show exceptional
stability over other conformations in the presence of Na+ or
K+ ions. Putative quadruplex sequences have been identified
in G-rich genomic sequences,2 with over-representation in
telomeres,3 as well as in other genomic regions for example
in promoter sequences of a number of proto-oncogenes,4 such
as c-myc5a and c-kit,5b in 50 untranslated regions6a and in
introns.6b A number of these putative quadruplexes are appealing
targets for cancer therapeutics. For instance, inducing
the single-stranded telomeric DNA overhang to fold into
G-quadruplexes has been shown to inhibit telomerase activity7a
and cancer cell growth.7b Such precise targeting of human
telomeres is significant since in >80% of cancers telomerase is
up-regulated and contributes to the malignant phenotype by
maintaining cancer cell immortalization.7c
Fig. 1 (Left) Representation of a 45-mer telomeric DNA sequence,
with two G-quadruplexes and a trisubstituted ligand (compound 15)
in a low-energy docked position at the interface between the two
quadruplexes. The ligand is shown as a solvent-accessible surface
colored by charge. (Right) The BRACO-19 molecule.
and shows antitumour activity against xenograft models.11
The BRACO-19 molecule was designed using qualitative
molecular modeling, with the crystal structure of the native
parallel human telomeric quadruplex as a template. It was
rationalized that each of the three substituents emanating
from the acridine core of BRACO-19 would be able to interact
with a quadruplex groove.10 This feature would, it was suggested,
provide binding selectivity over duplex DNA, which has just
two grooves. A more recent crystal structure of a BRACO-19
complex with a bimolecular quadruplex has confirmed the
essential correctness of this hypothesis and has also provided a
more detailed view of the interactions involved.12 The struc-
ture has a parallel-stranded quadruplex arrangement, with the
biological unit being two 50 to 30 stacked quadruplexes. Each
bimolecular quadruplex in this structure contains three planar
stacked G-quartets with a BRACO-19 molecule stacking
directly onto the 30 end G-quartet face.
A considerable number of small organic molecules have
been found to stabilise quadruplex DNA structures.8 Many,
though not all, are based on polycyclic heteroaromatic cores,
with the acridine nucleus being especially well explored.9
However, the selectivity of many of these molecules for
G-quadruplexes over duplex DNA is frequently less than what
would be therapeutically acceptable, and their polycyclic
features can make their druggability a challenge.
We report here the structure-based design, synthesis and
preliminary assessment of a novel series of non-polycyclic
trisubstituted ligands, whose affinity and selectivity for telomeric
G-quadruplex DNA has been evaluated using electrospray
mass spectrometry (ESI-MS).13 The goal has been to design
molecules (i) with potentially enhanced selectivity based on the
quadruplex concept of selective groove binding, and (ii) that
do not have a polycyclic heteroaromatic core, so potentially
enhancing drug-like features. We have used the BRACO-19-
quadruplex complex crystal structure12 as a starting point
for the structure-based design of non-polycyclic mimetics of
BRACO-19. The single-stranded overhang of human telomeric
DNA is 100–200 nucleotides in length, and in principle several
quadruplex structures can be formed along its length. We have
previously modeled such a higher-order arrangement using a
simple linker between 50 to 30 ends in this crystal structure to
form a continuous sequence, which does not involve any
The 3,6,9-trisubstituted acridine ligand BRACO-19 (Fig. 1)
has high affinity for human telomeric quadruplex DNAs and is
a potent inhibitor of the telomerase enzyme.10 It has selective
cytotoxic activity against a range of human tumour cell lines
a CRUK Biomolecular Structure Group, The School of Pharmacy,
University of London, London WC1N 1AX, UK.
E-mail: stephen.neidle@pharmacy.ac.uk; Tel: +44 (0)20 7753 5969
b Physical Chemistry and Mass Spectrometry Laboratory,
Department of Chemistry, University of Liege, B-4000 Liege,
Belgium
c The School of Chemistry, University of Nottingham,
University Park, Nottingham, NG7 2RD, UK
w Electronic supplementary information (ESI) available: Details of
synthetic chemistry, molecular modelling, electrospray mass spectro-
scopy and results of ESI-MS competition experiments. See DOI:
10.1039/c0cc02917c
c
9116 Chem. Commun., 2010, 46, 9116–9118
This journal is The Royal Society of Chemistry 2010

View Article Online
perturbation of the ligand binding site.14 The resulting 45-mer
bis-quadruplex ligand sandwich complex modeled structure has
been used here as the template for initial qualitative modelling
followed by in silico docking and binding energy calculations
for plausible ligands.15 We started with the hypothesis that such
ligands would require three substituents as in BRACO-19 itself,
and have explored their length and size. Ligand positions were
explored computationally at the interface between the two
quadruplexes, stacked on the 30 and 50 terminal G-quartet
surfaces. Ligands were derived from a previously devised16
series of bis-triazole quadruplex-binding ligands which contain
two alkylamine side-chain arms linked to a benzene core
through 1,4-triazoles. These had been prepared using a click
chemistry approach,17 which was also used here to add an
additional triazole ‘side arm’ to the benzene core.
Scheme 1 Synthesis of ligands 5–15: (i) 4-chlorobutyryl chloride
(4 eq.), 4 1C to rt, overnight; (ii) pyrrolidine (5 eq.), 4 1C to rt,
overnight; (iii) chloroacetyl chloride/3-chloropropionyl chloride, TEA
(2 eq.), THF, 41C to rt, 2 h; (iv) piperidine/pyrrolidine/diethylamine
(3 eq.), THF, 4 1C to rt, overnight; (v) anhydrous THF, H2, Pd/C
(10% m/m), rt, overnight; (vi) tBuONO (2.5 eq.), HCl (5.5 eq.), THF,
0 1C, 1.5 h; (vii) NaN3 (3 eq.), H2O, 0 1C to rt, overnight; (viii) 1,3,5-
triethynylbenzene, compound (4a–k) (4 eq.), CuSO4 (0.05 eq.), sodium
ascorbate (0.2 eq.), bathophenanthrolinedisulfonic acid disodium salt
hydrate (0.1 eq.), H2O–tBuOH, 110 1C, microwave, 15 min.
Initial qualitative modeling with the 45-mer suggested that
addition of the third alkylamine side-chain (Fig. 1) could
effectively mimic the key structural features of the BRACO-19
structure. This was followed by molecular dynamics simulations15
to ascertain whether the resulting complex was conformation-
ally stable. The resulting structure (Fig. 1) indicates that the
arrangement is structurally sound, with the central and three
attached phenyl rings constituting the core of the ligand, all
being involved in p–p stacking with G-quartet guanine bases
in the binding site. The three alkylamino arms are each
positioned in a groove structure, analogous to disposition of
the side-chains in the BRACO-19 complex crystal structure.
Circular dichroism studies on the binding of compound 15 to a
human telomeric quadruplex sequence show that the parallel
form is induced (see the ESIw), giving confidence to the use of
the parallel crystal structure in the modelling and simulation
studies. Even so, docking studies were also performed on an
alternative polymorph of the human telomeric quadruplex,
one of the (3 + 1) hybrid anti-parallel structures,18 but all low-
energy arrangements involved stereochemically unacceptable
buckling of ligand, DNA or both. These in silico trials were
then abandoned.
mixture was necessary to avoid elimination of the amine for
side-chains with n Z 2. The resulting compounds (5–15), all
have an extended aromatic surface. Initial attempts to study
DNA binding using techniques to measure the elevation in
melting temperature as a result of ligand were unsuccessful
since it appears that the compounds are unstable when UV
irradiated at the elevated temperatures required in melting
experiments. Their quadruplex and duplex DNA binding
abilities were therefore assessed by electrospray mass
spectrometry.
Binding of each ligand to the 22-mer human telomeric
G-quadruplex sequence d[AGGG(TTAGGG)3] (tel22) was
examined at 1 : 1 and 2 : 1 drug : DNA ratios (Table 1, see
also Fig. S1 and accompanying text, ESIw). All the Kd values
reported refer to the 1 : 1 complex (Q + L)5ꢀ. The meta-
substituted compounds are consistently the strongest binders,
with compounds 5, 8, 12 and 15 having lower Kd values than
that of BRACO-19 itself.20 All of these have short side chains
(n = 1, 2) with either pyrrolidine or diethylamine basic groups.
These compounds have consistently higher affinities than the
corresponding para compounds, with those having pyrrolidine
and diethylamine end-groups being superior to the piperidine
derivatives 11 and 13. Compounds 5 (n = 1) and 15 (n = 2)
are more active than the related compound (7) (n = 3), which
suggests that shorter side chains result in superior quadruplex
binding.
The synthesis of a small library of the trisubstituted click
ligands used a convergent click chemistry approach.17 A series
of azide building blocks with focussed structural diversity was
initially synthesised, then clicked onto a central core, the
commercially available trialkyne 1,3,5-triethynylbenzene
(Scheme 1). The diversity of building blocks was designed in
order to establish structure–activity relationships. They all
have the common features of a phenyl ring (providing
an aromatic surface) and a series of basic side-chains. These
were built onto a starting 2-nitroaniline or 3-nitroaniline in a
maximum of four steps. Briefly, the required nitroaniline was
acylated, then one-pot substituted with diverse amines. The
nitro group was readily reduced with H2, then the azide was
synthesised from the resulting aniline with one-pot diazotisa-
tion and azide substitution.19 The 1,3,5-triethynylbenzene was
then extended with one-pot click reaction formation of three
triazole rings via Cu(I) catalysed Huisgen 1,4-dipolar cyclo-
addition. Catalytic Cu(I) was formed in situ and a second
catalyst, bathophenanthrolinedisulfonic acid disodium salt
hydrate, was necessary to achieve complete trisubstitution.
The reaction was complete after only 15 minutes of microwave
irradiation; an excess of the required amine in the reaction
ESI-MS has also been used to assess quadruplex:duplex
selectivity in a competition experiment between three sequences.21
The telomeric 22-mer (tel22), the parallel intermolecular
G-quadruplex sequence (dTG4T)4 and
a duplex DNA,
d(CGCGAATTCGCG)2, each at 5 mM, were injected with
the ligand at 10 mM. Almost all the compounds showed
selectivity for G-quadruplex sequences and did not bind to
the duplex (Fig. S5, ESIw).
Several of the compounds have higher affinity for the tetra-
molecular TG4T quadruplex than for the telomeric quadruplex
c
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 9116–9118 9117

View Article Online
Table 1 ESI-measured dissociation constants for compounds 5–15
with a human telomeric 22-mer quadruplexa. NR2 groups are defined
in Scheme 1
COST Action MP0802 (to S.H. and V.G.), and the University
of Liege (FSR starting grant D-08/10 to V.G.).
Compound
Substitution pattern
n
NR2
Kd/mM
Notes and references
5
6
7
8
9
10
11
12
13
14
meta-
para-
meta-
meta-
para-
para-
meta-
meta-
para-
para-
meta-
n/a
1
2
3
2
2
1
1
1
1
1
2
n/a
pyr
pyr
pyr
dieth
dieth
pyr
pip
dieth
pip
6.0 ꢂ 0.6
70 ꢂ 50
35 ꢂ 9
1 D. J. Patel, A. T. Phan and V. Kuryavyi, Nucleic Acids Res., 2007,
35, 7429; S. Burge, G. N. Parkinson, P. Hazel, A. K. Todd and
S. Neidle, Nucleic Acids Res., 2006, 34, 5402; S. Neidle, Curr. Opin.
Struct. Biol., 2009, 19, 239.
2 J. Huppert and S. Balasubramanian, Nucleic Acids Res., 2005, 33,
2908; A. Todd, M. Johnston and S. Neidle, Nucleic Acids Res.,
2005, 33, 2901.
3 M. J. McEachern, A. Krauskopf and E. H. Blackburn, Annu. Rev.
Genet., 2000, 34, 331.
4 J. L. Huppert and S. Balasubramanian, Nucleic Acids Res., 2006,
35, 406.
5 See for example: (a) A. Siddiqui-Jain, C. L. Grand, D. J. Bearss
and L. H. Hurley, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 11593;
Y. Qin, J. S. Fortin, D. Tye M. Gleason-Guzman, T. A. Brooks
and L. H. Hurley, Biochemistry, 2010, 49, 4208; (b) S. T. Hsu,
P. Varnai, A. Bugaut, A. P. Reszka, S. Neidle and
S. Balasubramanian, J. Am. Chem. Soc., 2009, 131, 13399;
M. Gunaratnam, S. Swank, S. M. Haider, K. Galesa,
A. P. Reszka, M. Beltran, F. Cuenca, J. A. Fletcher and
S. Neidle, J. Med. Chem., 2009, 52, 3774.
6 (a) For example: S. Kumari, A. Bugaut, J. L. Huppert and
S. Balasubramanian, Nat. Chem. Biol., 2007, 3, 218; M. Wieland
and J. S. Hartig, Chem. Biol., 2007, 14, 757; A. Bugaut,
R. Rodriguez, S. Kumari, S. T. Hsu and S. Balasubramanian,
Org. Biomol. Chem., 2010, 8, 2771; (b) J. Eddy and N. Maizels,
Nucleic Acids Res., 2007, 36, 1321.
7 (a) D. Sun, B. Thompson, B. E. Cathers, M. Salazar, S. M. Kerwin,
J. O. Trent, T. C. Jenkins, S. Neidle and L. H. Hurley, J. Med.
Chem., 1997, 40, 2113; (b) P. Phatak, J. C. Cookson, F. Dai,
V. Smith, R. B. Gartenhaus, M. F. Stevens and A. M. Burger, Br.
J. Cancer, 2007, 96, 1223; (c) J. W. Shay and W. E. Wright, Nat.
Rev. Drug Discovery, 2006, 5, 577.
8 A. De Cian, L. Lacroix, C. Douarre, N. Temime-Smaali,
C. Trentesaux, J. F. Riou and J. L. Mergny, Biochimie, 2008, 90,
131; D. Monchaud and M.-P. Teulade-Fichou, Org. Biomol.
Chem., 2008, 6, 627.
9 See for example: R. J. Harrison, S. M. Gowan, L. R. Kelland and
S. Neidle, Bioorg. Med. Chem. Lett., 1999, 9, 2463; M.-K. Cheng,
C. Modi, J. C. Cookson, I. Hutchinson, R. A. Heald, A. J. McCarroll,
S. Missailidis, F. Tanious, W. D. Wilson, J.-L. Mergny,
C. A. Laughton and M. F. G. Stevens, J. Med. Chem., 2008, 51, 963.
10 M. Read, R. J. Harrison, B. Romagnoli, F. A. Tanious,
S. H. Gowan, A. P. Reszka, W. D. Wilson, L. R. Kelland and
S. Neidle, Proc. Natl. Acad. Sci. U. S. A., 2001, 98, 4844.
11 A. M. Burger, F. Dai, C. M. Schultes, A. P. Reszka, M. J. Moore,
J. A. Double and S. Neidle, Cancer Res., 2005, 65, 1489.
12 N. H. Campbell, G. N. Parkinson, A. P. Reszka and S. Neidle,
J. Am. Chem. Soc., 2008, 130, 6722.
13 F. Rosu, E. De Pauw and V. Gabelica, Biochimie, 2008, 90, 1074;
V. Gabelica, Methods Mol. Biol., 2010, 613, 89.
14 S. M. Haider and S. Neidle, Biochem. Soc. Trans., 2009, 37, 583.
15 S. M. Haider and S. Neidle, Methods Mol. Biol., 2010, 608, 17. See
also the ESIw.
16 A. D. Moorhouse, A. M. Santos, M. Gunaratnam, M. Moore,
S. Neidle and J. E. Moses, J. Am. Chem. Soc., 2006, 128, 15972.
17 H. C. Kolb, M. G. Finn and K. B. Sharpless, Angew. Chem., Int.
Ed., 2001, 40, 2004.
18 A. T. Phan, V. Kuryavyi, K. N. Luu and D. J. Patel, Nucleic Acids
Res., 2007, 35, 6517; A. Ambrus, D. Chen, J. Dai, T. Bialis,
R. A. Jones and D. Yang, Nucleic Acids Res., 2006, 34, 2723.
19 J. Das, S. N. Patil, R. Awasthi, C. P. Narasimhulu and S. Trehan,
Synthesis, 2005, 1801.
20 G. Collie, A. P. Reszka, S. M. Haider, V. Gabelica,
G. N. Parkinson and S. Neidle, Chem. Commun., 2009, 7482.
21 A. Casagrande, A. Alvino, G. Bianco, G. Ortaggi and
M. Franceschin, J. Mass Spectrom., 2009, 44, 530.
22 T. P. Garner, H. E. L. Williams, K. I. Gluszyk, S. Roe,
N. J. Oldham, M. F. G. Stevens, J. E. Moses and M. S. Searle,
Org. Biomol. Chem., 2009, 7, 4194.
5.0 ꢂ 1.0
100 ꢂ 19
49 ꢂ 26
250 ꢂ 130
3.0 ꢂ 0.6
77 ꢂ 12
32 ꢂ 5
dieth
pyr
n/a
15
BRACO-19
4.9 ꢂ 1.3
7.9 ꢂ 1.4
a
Abbreviations: (pyr) pyrrolidino, (dieth) diethyl amino, (pip)
piperidino. BRACO-19 has been used as a reference with data taken
from ref. 20. Esds from two ligand concentrations (5 and 10 mM) and
three voltage settings each.
or for duplex DNA. By examining the ESI-MS spectra (see Fig. S1
and accompanying text, ESIw), we noted that, in addition to
binding to tel22 to form 1 : 1 complexes, all ligands with para
substituents were causing partial dimerization of the (dTG4T)4
quadruplex. However, compounds 5, 8, 12 and 15 have high
affinity for the telomeric sequence, even in the presence of
(dTG4T)4. The correlation with the Kd values is shown in
Fig. S3 (ESIw). These are also the compounds that have
the highest affinity for the telomeric quadruplex. Calculated
binding energies15 (DEbinding) for 5, 7 and 15 are ꢀ8.64, ꢀ7.99
and ꢀ9.63 kcal moleꢀ1, respectively, which concurs with the
ranking order in Table 1.
Ligand 11 did not bind significantly to either of the other
sequences. None of the ligands bound to the duplex sequence.
Mixtures of 5 mM DNA duplex + 40 mM ligand were run
in an attempt to determine the Kd for ligand binding to duplex,
in order to calculate the selectivity of the ligands for the
telomeric quadruplex; no complex formation was detected
with any of the ligands examined in these particular experi-
ments (see Fig. S4, ESIw). From the noise level where the
complex should be detected, we estimated the Kd that would
lead to a detectable complex (S/N = 3 for the complex), and
deduced a Kd value for duplex binding of >3000 mM. This gives
a quadruplex/duplex selectivity ratio > 1000 for ligand 12, for
example. Although Kd values for the disubstituted triazole
compounds are not available, a Ka value22 for the analogue of
compound 6, with pyrrolidino end-groups of 7.7 ꢁ 105 Mꢀ1
,
suggests that the trisubstitution has resulted in an enhancement
of affinity, at least for the para-substituted compounds.
In summary a group of ligands have been designed which
demonstrate high affinity for an intramolecular human telomeric
quadruplex that is comparable to, and in several instances,
exceeds that of the established acridine ligand BRACO-19.
The tris-triazole ligands have higher selectivity for quadruplex
over duplex DNA compared to BRACO-19, with >1000 fold
difference in Kd values. Three compounds 8, 12 and 15 have been
selected for biological studies, which are currently underway.
This work was supported by a CRUK Programme Grant
(to S.N.), a CRUK Research Studentship (to J.E.M. and S.N.
for C.M.L.), the FNRS (research associate position to V.G.),
c
9118 Chem. Commun., 2010, 46, 9116–9118
This journal is The Royal Society of Chemistry 2010