Combining 1,3-Ditriazolylbenzene and Quinoline to Discover a New G-Quadruplex-Interactive Small Molecule Active against Cancer Stem-Like Cells

Article abstract of DOI:10.1002/cmdc.201900243

Quadruplex nucleic acids are promising targets for cancer therapy. In this study we used a fragment-based approach to create new flexible G-quadruplex (G4) DNA-interactive small molecules with good calculated oral drug-like properties, based on quinoline

Full text of DOI:10.1002/cmdc.201900243

Accepted Article
Title: Combining 1,3-ditriazolyl-benzene and quinoline to discover
a new G-quadruplex interactive small molecule active against
cancer stem-like cells
Authors: Eduarda Mendes, Enrico Cadoni, Filipa Carneiro, Marta B.
Afonso, Hugo Brito, João Lavrado, Daniel J. V. A. dos Santos,
Jorge B. Vítor, Stephen Neidle, Cecília M. P. Rodrigues, and
Alexandra Paulo
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemMedChem 10.1002/cmdc.201900243
Link to VoR: http://dx.doi.org/10.1002/cmdc.201900243
A Journal of
www.chemmedchem.org

10.1002/cmdc.201900243
ChemMedChem
COMMUNICATION
Combining 1,3-ditriazolyl-benzene and quinoline to discover a new G-quadruplex
interactive small molecule active against cancer stem-like cells
Eduarda Mendes,a Enrico Cadoni,a Filipa Carneiro,a Marta B. Afonso,a Hugo Brito,a João
Lavrado,a Daniel J. V. A. dos Santos,a Jorge B. Vítor,a,b Stephen Neidle,c Cecília M. P.
Rodrigues,a Alexandra Paulo a*
[a]
Dr. E. Mendes, E. Cadoni, F. Carneiro, Dr. M. B. Afonso, H. Brito, Dr. J. Lavrado, Dr. D.J.V.A. dos Santos, Dr. J. B. Vítor, Prof. C.M.P. Rodrigues, Prof. A.
Paulo
Research Institute for Medicines (iMed.ULisboa), Faculty of Pharmacy, Universidade de Lisboa, Av. Prof. Gama Pinto, 1649 003 Lisbon, Portugal
E-mail: mapaulo@ff.ulisboa.pt
[b]
[c]
Dr. J. B. Vítor
Department of Biochemistry & Human Biology, Faculty of Pharmacy, Universidade de Lisboa, 1649-003 Lisbon, Portugal
Prof. S. Neidle
School of Pharmacy, UCL, London WC1N 1AX, UK
Supporting information for this article is given via a link at the end of the document.
Abstract: Quadruplex nucleic acids are promising targets for cancer
therapy. In this study we used a fragment-based approach to create
new flexible G-quadruplex (G4) DNA interactive small molecules
with good calculated oral drug-like properties, based on quinoline
and triazole heterocycles. G4 melting temperature and Polymerase
Chain Reaction-stop assays showed that two of these compounds
are selective G4 ligands, since they were able to induce and
stabilize G4s in a dose- and DNA sequence-dependent manner.
Molecular docking studies have suggested plausible quadruplex
binding to both the G-quartet and groove, with the quinoline module
playing the major role. Compounds were screened for cytotoxicity
acid structures as potential targets for selective anticancer
drugs.[1][3] Strikingly, it has been shown that G4-stabilizing
ligands might differentially target human cancer stem cells,[11][12]
a subpopulation of cancer cells implied in tumor formation,
metastases and recurrence due to their long-lasting properties
and resistance to chemotherapy.[13] Many G4-interactive small
molecules with anti-proliferative activity have been described to
date but only two of them have reached clinical trials,[14] possibly
in part because the rational design of many efficient and
selective G4-binders has led to molecules with poor or only
moderate pharmacokinetic properties. Differences between
nucleic acid sequences forming G4s can create diverse G4
topologies that may have major influence on G4-ligand
interactions. The design of flexible non-fused polycyclic G4-
interactive small molecules that may also target grooves and
loops of G4s, has emerged as one way to achieve selectivity
and also to lead to more drug-like compounds.[15][16][17] The 1,3-
di(1H-1,2,3-triazol-4-yl)benzene system was previously shown
by biophysical and molecular modeling studies to provide an
efficient central module, able to interact effectively with the 3’-
end G-quartet of the human telomeric parallel-stranded G4
structure.[18][19] Also, quinolines have been shown to be important
modules in the design of several potent G4 ligands.[15][20]
against
four
cancer
cell
lines,
where
4,4'-(4,4'-(1,3-
phenylene)bis(1H-1,2,3-triazole-4,1-diyl))bis(1-methylquinolin-1-ium)
(1d) showed the greater activity. Importantly, dose-response curves
show that 1d is cytotoxic in the human colon cancer HT-29 cell line
enriched in cancer stem-like cells,
a subpopulation of cells
implicated in chemo-resistance. Overall, this study identified a new
small molecule as a promising lead for the development of drugs
targeting G4 in cancer stem cells.
DNA binding compounds have been widely used in cancer
therapy, but their general toxicity due to lack of selectivity
demands new approaches. An emerging and promising new
approach to develop more selective anticancer drugs, has four-
stranded G-quadruplex (G4) DNA structures formed by certain
guanine rich-sequences, as targets.[1][2][3] The basic unit of a G4
is the G-quartet, a square-planar arrangement of four guanine
bases linked with hydrogen bonds. The G-quartets can be linked
by intervening loop sequences so that they can stack on top of
each other to form a G4, which is additionally stabilized in
solution by monovalent cations coordinated with the O6 atoms of
guanines.[1] There is now good evidence that G4s can be formed
in cells[4][5] where they have important regulatory roles such as in
telomere maintenance,[6] in DNA replication, in epigenetic
responses and in control of gene expression.[1][7][8] Moreover,
the enrichment of these structures in promoter regions of, in
particular, proto-oncogenes has been revealed by informatics
analysis of the human genome[9] and supported by antibody-
based G4 chromatin immunoprecipitation and high-throughput
sequencing.[10] The discovery that in many cases they negatively
control transcription, has led to the exploitation of these nucleic
In order to further explore triazole and quinoline rings in the
design of G4-interactive small molecules with potential
anticancer activity, we have designed compounds 1a-d with
predicted good oral drug-like properties (Table S1). Synthesis of
compounds 1a-c was readily achieved by microwave-assisted
Cu(I)-catalysed Huisgen cycloaddition reaction between 1,3-
diethynylbenzene and the appropriate azide as depicted in
Scheme
1 and described in the Supporting Information.
Interestingly, methylation of the 4-azidoquinoline nitrogen atom
precludes the cycloaddition reaction, even in the presence of
bathophenanthroline disulphonic acid which was previously
described as a ‘click catalyst’.[21] Instead, compound 1d was
obtained by direct methylation of 1c. The ability of compounds
1a-d to interact with different G4 DNA structures and a duplex
DNA sequence was evaluated by a Fluorescence Resonance
Energy Transfer (FRET)-melting assay. The DNA sequences
used, their respective folding topologies and melting
temperatures at the assay conditions, are presented in Table S2.
1
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900243
ChemMedChem
COMMUNICATION
A
B
Telo21
KRAS21
cMYC19
TLoop
1b
1d
Double strand DNA
(fold to G4)
Figure 1. G4 stabilizing efficiency and selectivity of compounds measured by
a FRET melting assay. (A) Melting temperature variations (ΔTm) of labeled
G4s present in promoters of k-RAS (KRAS21) and c-MYC (cMYC19), human
telomere (Telo21) and hairpin loop sequence (T-loop) at 0.2 M, stabilized by
compounds (5 M). ΔTm values are averages from two independent
experiments each in triplicate; std errors
< 0.25 °C. (B) FRET melting
competition assay data for 1b and 1d in complex with KRAS21 G4s,
challenged with increasing concentrations of non-labeled 26mer dsDNA
competitor.
Pu27
5’TGGGGAGGGTGGGGAGGGTGGGGAAGG 3’
3‘ACCCCTTCCTGCTCTTCGCTAGCTA 5’
Pu27mut
5’TGGGGAGGGTGGAAAGGGTGGGGAAGG 3’
3‘ACCCCTTCCTGCTCTTCGCTAGCTA 5’
Scheme 1. Synthesis of compounds 1a-d. Conditions: a) Ethanolamine, TEA,
reflux, 24h; b) Methanesulphonyl chloride, pyridine, 0°C, 8h; c) NaN3, DMF,
110 °C, 2-3h; d) MeI (10 eq.), ACN, r.t.; e) 1,3-diethynylbenzene, CuSO4.5H2O,
Sodium L-ascorbate, tBuOH:H2O (1:1), MW 110 ˚C, 30 min.
A
B
The results shown in Figure 1A reveal that 1a-d are superior
stabilizers of G4 structures compared to duplex DNA (Tloop),
but no significant differences were observed between the
melting temperatures of DNA:ligand complexes with different
G4-sequences and folds (parallel or hybrid). Surprisingly, 1c
showed no capacity to stabilize G4s structures whereas
compounds 1b and 1d showed very similar ability to interact and
stabilize G4s DNA, possibly because their binding mode to G4 is
governed by a common chemical feature, the quinolinium cation.
Figure 1B shows the results of a FRET melting competition
assay, which indicates that 1d is more selective for G4
structures than 1b. In the presence of 50-fold higher molar
concentration of double strand DNA, the melting temperature of
the G4:1d complex decreases by only 10%.
50 bp
43 bp
product
50 bp
43 bp
product
Figure 2. PCR-Stop effect of compounds on two DNA sequences.
Polyacrylamide gels stained with ethidium bromide showing the effect of
compounds on polymerization of DNA constructs (A) containing the c-MYC
promoter wild type sequence (Pu27) or (B) with a mutated c-MYC promoter
sequence (Pu27mut). Legend: G-runs are shown in red, mutated bases in blue
and the hybridization region in grey; C – control of PCR product in the
absence of G4 ligand (0 M); M – PCR molecular weight marker.
To further assess the capacity of compounds to induce and
stabilize DNA G4 structures we performed a Polymerase Chain
Reaction (PCR)-stop assay using two different G4-forming
sequences. The 27 nucleotides sequence present in the wild-
type promoter region of the c-MYC oncogene (Pu27, Figure 2A)
is able to form three different G4s, the more stable loop-isomer
involving G-runs 2-3-4-5 and the less stable loop isomers 1-2-3-
4 and 1-2-4-5;[22][23] whereas substitution of two guanines in G-
run 3 (Pu27mut, Figure 2B) abrogates the formation of two of
the G4s. Figure 2 illustrates that compounds 1b and 1d can
inhibit DNA hybridization at G-run 5 and consequent formation of
the PCR product in a dose- and sequence-dependent manner,
clearly suggesting selective activity by induction of G4 formation.
The binding mode to G4s was investigated by molecular
docking using a well-studied G4, the human telomere G4
structure (PDB: 143D) and a previously validated protocol.[24]
The score results are presented in Table S5 and have the same
trend as the FRET results, i.e. the methylated compounds 1b
and 1d form more stable complexes and compound 1c is the
poorest performer. The top-ranked docking poses have a
quinoline ring sitting on top of the G-quartet formed by
nucleotides G1, G9, G21, G13, and below T11 with the rest of
the molecule occupying the groove, starting at G1 and ending at
G3 (Figures 3 and S1B). Although this stacking, to be effective,
tends to put the quinoline rings all in the same plane, they
occupy nevertheless different relative positions which impact on
the conformation of the molecule and change the possibilities for
suitable interactions. This is clearly seen in the number of direct
interactions found by the MOE software, with a 0.5 kcal mol-1
2
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900243
ChemMedChem
COMMUNICATION
cutoff interaction (Figure S2). While molecule 1a interacts with
G8 through a hydrogen bond with the triazole nitrogen atom at
position 2, 1b makes π-π interactions both with G1 and T11 and
arene-H with G1 and G2, while T11 also functions as a side-
chain acceptor. On the other hand, while molecule 1c only
makes π-π interactions with G1 and T11, 1d makes these same
interactions together with an extra arene-H interaction with G2.
The results also show that the reduced partial charge on the
methylated nitrogen atom, as in 1b and 1d, plays an important
role by allowing a tighter interaction with the quadruplex. It is
interesting to note that the central ditriazolyl-benzene moiety is
not responsible for defining the relative docking position of the
molecule as in previously ditriazolyl-benzene derivatives,[19] but
instead this role is played by the quinoline rings at both ends of
the molecules.
be used in controlling tumor progression, metastatic spread and
disease recurrence.[13]
In summary, we have used a fragment-based approach to
design a family of new flexible G4 ligands with good calculated
oral drug-like properties (Table S1).[26] Studies of interaction with
different DNA sequences indicate that compounds 1b and 1d
are able to induce and selectively stabilize G4 DNA structures in
a dose-dependent manner and preliminary molecular docking
studies suggest that these molecules target both the G-quartet
and the groove, with the quinoline rings having a major role in
the interaction with G4s. Importantly, compound 1d showed
short-term anti-proliferative activity, particularly against colon
cancer cells, and good activity against colon cancer stem-like
cells.
A
B
HT-29 enriched-CSC
DMSO
1b
1d
HT-29 adherent culture
log [compound], µM
Cell Line
Figure 4. Evaluation of the cytotoxic effect of compounds 1b and 1d. (A)
Human cancer cell lines HCT116, HT29, HepG2 and murine fibrosarcoma cell
line L929 were cultured in adherent conditions and incubated with compounds
at 50 M for 72 h. (B) Dose-response curves for compound 1d incubated with
HT29 cancer cells cultured in adherent conditions for 72 h and with HT-29-
derived tumor spheres (HT-29 enriched-CSC) for
7
days. Results are
independent
expressed as mean percentage SEM from at least
±
3
experiments and normalized to vehicle control (DMSO). *p < 0.05 from vehicle
control.
The fact that 1d binds to more than one G4 type suggests
that it can target multiple G4s. This would confer therapeutic
advantage in complex human cancers, where multiple cancer
genes and pathways need to be down-regulated in order to
achieve high levels of anticancer potency.[27] The relative
importance of G4 nucleic acids in the control of cancer stem
cells proliferation compared with cancer non-stem cells has
been poorly studied. A G4 has been implicated as a positive
regulator of octamer-binding protein 4 (Oct4) expression, a
Figure 3. Top-ranked docking poses for compounds 1b (carbon atoms in
green) and 1d (carbon atoms in yellow). Interactions are represented by
dotted lines and only bases closer than 4.5 Å to the molecules are represented.
transcription factor associated with pluripotent
properties
of embryonic and cancer stem cells.[28] Furthermore, a few
studies have suggested that cancer stem cells may be more
susceptible to loss of telomerase function, either by
transcriptional inhibition of TERT and c-MYC,[29] or by induction
of G4 formation at telomeres.[11][12] Overall, compound 1d is a
G4-stabilizing ligand based on quinoline and triazole
heterocycles, which shows activity against colon cancer stem-
like cells and good predicted oral drug-like properties. As such,
this compound could be a promising lead for the development of
novel drugs targeting long-lasting and chemo-resistant cancer
stem cells.
Compounds 1b and 1d were first screened for their cell
growth inhibitory activity in a panel of cancer cell lines. The
results (Figure 4A) show a consistent potency trend of 1d > 1b,
which is particularly clear for human colon cancer HT-29 and
HCT116 cell lines. A dose-response curve confirmed the activity
of 1d for HT-29 cancer cells cultured in adherent conditions
(Figure 4B). It has been proposed that cancer stem cells play a
key role in colon cancer initiation, growth, metastasis and
therapy resistance.[13] To assess the inhibitory potency of 1d in
colon cancer stem-like cells, a dose-response curve was
calculated using human colon cancer HT-29-derived tumor
spheres.[25] Our results showed that 1d also targets HT-29 cells
enriched in cancer stem-like cells, with a IC50 value in the low
micromolar range (IC50 10.63 μM; 95% CI 7.39 – 20.08) (Figure
4B). Conventional chemotherapeutic regimens are effective in
reducing tumor mass, but often fail to eliminate stem-like cancer
cells, leading to eventual tumor relapse. Thus, the discovery of
drug-like small molecules active against cancer stem cells could
Experimental Section
Detailed experimental procedures, including compounds characterization
data and additional docking studies results can be found in the
supporting information file.
3
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900243
ChemMedChem
COMMUNICATION
[23] R. I. Mathad, E. Hatzakis, J. Dai, D. Yang, Nucleic Acids Res. 2011, 39,
9023–9033.
Acknowledgements
[24] J. Lavrado, S. A. Ohnmacht, I. Correia, S. Pisco, M. Gunaratnam, R.
Moreira, S. Neidle, J. V. A. Santos, A. Paulo, ChemMedChem 2015, 10,
836–849.
This work has received funding from from European Structural &
Investment Funds through the COMPETE Program – Programa
Operacional de Lisboa, under the Program grant LISBOA-01-
0145-FEDER-016405 and from National Funds through FCT,
under the Program grant SAICTPAC/0019/2015. J. Vitor’s lab
has been financed by New England Biolabs, Inc. (USA).
[25] D. M. Pereira, S. E. Gomes, P. M. Borralho, C. M. P. Rodrigues, Cell
Death Discov. 2019, 5, 1–13.
[26] D. F. Veber, S. R. Johnson, H. Cheng, B. R. Smith, K. W. Ward, K. D.
Kopple, J. Med. Chem. 2002, 45, 2615–2623.
[27] C. Marchetti, K. G. Zyner, S. A. Ohnmacht, M. Robson, S. M. Haider, J.
P. Morton, G. Marsico, T. Vo, S. Laughlin-Toth, A. A. Ahmed, et al., J.
Med. Chem. 2018, 61, 2500–2517.
Keywords: DNA • quadruplex • heterocycle • cancer stem cell •
drug design
[28] D. Renčiuk, J. Ryneš, I. Kejnovská, S. Foldynová-Trantírková, M.
Andäng, L. Trantírek, M. Vorlíčková, Biochim. Biophys. Acta - Gene
Regul. Mech. 2017, 1860, 175–183.
References:
[29] J. Chappell, S. Dalton, M. F. Roussel, G. W. Robinson, P. Gallant, Cold
Spring
Harb.
Perspect.
Med.
2013,
3,
DOI
[1]
[2]
A. Paulo, C. C. Castillo, S. Neidle, in Compr. Med. Chem. III (Eds.: S.
Chakckalamannil, D.P. Rotella, S.S. Ward), Oxford: Elsevier, 2017, pp.
308–340.
10.1101/cshperspect.a014381.
R. Hänsel-Hertsch, M. Di Antonio, S. Balasubramanian, Nat. Rev. Mol.
Cell Biol. 2017, DOI 10.1038/nrm.2017.3.
[3]
[4]
A. P. Francisco, A. Paulo, Curr. Med. Chem. 2017, 24, 4873–4904.
G. Biffi, D. Tannahill, J. McCafferty, S. Balasubramanian, Nat. Chem.
2013, 5, 182–186.
[5]
A. Henderson, Y. Wu, Y. C. Huang, E. A. Chavez, J. Platt, F. B.
Johnson, R. M. Brosh, D. Sen, P. M. Lansdorp, Nucleic Acids Res.
2014, 42, 860–869.
[6]
[7]
M. Ghosh, M. Singh, Nucleic Acids Res. 2018, 46, 10246–10261.
N. Maizels, L. T. Gray, PLoS Genet. 2013, 9, DOI
10.1371/journal.pgen.1003468.
[8]
[9]
B. Herdy, C. Mayer, D. Varshney, G. Marsico, P. Murat, C. Taylor, C.
D’Santos, D. Tannahill, S. Balasubramanian, Nucleic Acids Res. 2018,
46, 11592–11604.
J. L. Huppert, S. Balasubramanian, Nucleic Acids Res. 2007, 35, 406–
413.
[10] R. Hänsel-hertsch, D. Beraldi, S. V Lensing, G. Marsico, K. Zyner, A.
Parry, M. Di Antonio, J. Pike, H. Kimura, M. Narita, et al., Nat. Genet.
2016, 48, 1267–1272.
[11] P. Phatak, J. C. Cookson, F. Dai, V. Smith, R. B. Gartenhaus, M. F. G.
Stevens, A. M. Burger, Br. J. Cancer 2007, 96, 1223.
[12] T. Nakamura, S. Okabe, H. Yoshida, K. Iida, Y. Ma, S. Sasaki, T.
Yamori, K. Shin-Ya, I. Nakano, K. Nagasawa, et al., Sci. Rep. 2017, 7,
1–11.
[13] A. Zeuner, M. Todaro, G. Stassi, R. De Maria, Cell Stem Cell 2014, 15,
692–705.
[14] H. Xu, M. Di Antonio, S. McKinney, V. Mathew, B. Ho, N. J. O’Neil, N.
Dos Santos, J. Silvester, V. Wei, J. Garcia, et al., Nat. Commun. 2017,
8, 14432.
[15] A. R. Duarte, E. Cadoni, A. S. Ressurreição, R. Moreira, A. Paulo,
ChemMedChem 2018, 13, 869–893.
[16] J. Amato, A. Pagano, D. Capasso, S. Di Gaetano, M. Giustiniano, E.
Novellino, A. Randazzo, B. Pagano, ChemMedChem 2018, 13, 406–
410.
[17] D. Musumeci, J. Amato, P. Zizza, C. Platella, S. Cosconati, C.
Cingolani, A. Biroccio, E. Novellino, A. Randazzo, B. Pagano, Biochim.
Biophys. Acta - Gen. Subj. 2017, 1861, 1341–1352.
[18] A. D. Moorhouse, S. Haider, M. Gunaratnam, D. Munnur, S. Neidle, J.
E. Moses, Mol. Biosyst. 2008, 4, 629–642.
[19] A. D. Moorhouse, A. M. Santos, M. Gunaratnam, M. Moore, S. Neidle, J.
E. Moses, J. Am. Chem. Soc. 2006, 128, 15972–15973.
[20] S. Müller, K. Laxmi-Reddy, P. V Jena, B. Baptiste, Z. Dong, F. Godde,
T. Ha, R. Rodriguez, S. Balasubramanian, I. Huc, Chembiochem 2014,
15, 2563–2570.
[21] C. M. Lombardo, I. S. Martínez, S. Haider, V. Gabelica, E. De Pauw, J.
E. Moses, S. Neidle, Chem. Commun. 2010, 46, 9116–9118.
[22] T. A. Phan, Y. S. Modi, D. J. Patel, J. Am. Chem. Soc. 2004, 126,
8710–8716.
4
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900243
ChemMedChem
COMMUNICATION
Entry for the Table of Contents
Targeting cancer stem-cells with G4 ligands. Conventional chemotherapeutic regimens are effective in reducing tumor mass, but
often fail to eliminate cancer stem cells (CSC) leading to tumor relapse. Combining heterocycles we discovered a new G-quadruplex
(G4)-interactive small molecule with predicted oral drug-like properties and good activity against colon CSC.
5
This article is protected by copyright. All rights reserved.