Facile syntheses of disubstituted bis(vinylquinolinium)benzene derivatives as G-quadruplex DNA binders

Article abstract of DOI:10.1016/j.tet.2013.04.045

Abstract A series of disubstituted bis(vinylquinolinium)benzene derivatives were designed, which were prepared through a facile three-component one-pot reaction in good yield. FRET results showed that 1,3-disubstituted benzene derivatives had much stronger stabilization effect on G-quadruplex DNA than that of 1,4-disubstituted benzene derivatives. The introduction of substituted amine side chain at quinolinium obviously increased the binding affinity of compounds to G-quadruplex DNA. It was also found that 1,3-disubstituted benzene derivatives and 1,4-disubstituted benzene derivatives had different effects on the conformation of G-quadruplex DNA by CD spectroscopy analysis. The differences for the interactions of these two classes of compounds with G-quadruplex were further studied and elaborated through molecular modeling experiments.

Full text of DOI:10.1016/j.tet.2013.04.045

Tetrahedron 69 (2013) 4922e4932
Contents lists available at SciVerse ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Facile syntheses of disubstituted bis(vinylquinolinium)benzene
derivatives as G-quadruplex DNA binders
Zhen-Quan Liu, Shi-Tian Zhuo, Jia-Heng Tan, Tian-Miao Ou, Ding Li, Lian-Quan Gu,
*
Zhi-Shu Huang
School of Pharmaceutical Sciences, Sun Yat-sen University, Guangzhou 510006, People’s Republic of China
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of disubstituted bis(vinylquinolinium)benzene derivatives were designed, which were prepared
through a facile three-component one-pot reaction in good yield. FRET results showed that 1,3-
disubstituted benzene derivatives had much stronger stabilization effect on G-quadruplex DNA than
that of 1,4-disubstituted benzene derivatives. The introduction of substituted amine side chain at qui-
nolinium obviously increased the binding affinity of compounds to G-quadruplex DNA. It was also found
that 1,3-disubstituted benzene derivatives and 1,4-disubstituted benzene derivatives had different effects
on the conformation of G-quadruplex DNA by CD spectroscopy analysis. The differences for the in-
teractions of these two classes of compounds with G-quadruplex were further studied and elaborated
through molecular modeling experiments.
Received 12 January 2013
Received in revised form 28 March 2013
Accepted 11 April 2013
Available online 16 April 2013
Keywords:
Bisquinolinium derivatives
One-pot reaction
SNAr reaction
Ó 2013 Elsevier Ltd. All rights reserved.
G-quadruplex
Binding agents
1. Introduction
derivatives, such as 360A (Fig. 1) with a amide linkage between
quinolinium moiety and pyridine ring, and the positive charge on
Guanine-rich sequences can self-associate into planar guanine
quartets (G-quartets) that stack on each other to form unusual
structures called G-quadruplexes.1 Studies revealed the prevalence
of G-quadruplex-forming sequences in promoter regions,2 such as
c-myc,3 c-kit,2d,4 and k-ras,5 as well as in the telomeric chromo-
somal terminals.6 The formation or stabilization of G-quadruplexes
in these regions may play important role in the regulation of gene
expression and the maintenance of telomere length. For example,
formation of quadruplex structures at the promoter regions can
regulate gene expression at the transcriptional level.7 Formation of
a quadruplex at the telomeric end can stop the function of the
telomerase enzyme.8 Small molecules that can selectively induce
G-quadruplex formation or stabilization are therefore promising
lead compounds for cancer treatment.7,9
quinolinium might be interacted with phosphate groups of the DNA
backbone.10 Czerwinska et al. have reported that 1,4-bis-(vinyl-
quinolinium) benzene (Fig. 1) with a vinyl linkage between qui-
nolinium moiety and benzene ring as a G-quadruplex binder, can be
controlled to recognize parallel and anti-parallel G-quadruplex
DNA by switching light.11 In the year of 2010, our group has also
reported a series of bisaryldiketene derivatives M1eM4 with a vinyl
linkage (Fig. 1) as effective c-myc G-quadruplex DNA binders with
high selectivity for G-quadruplex over duplex DNA.12 In addition,
€
Muller et al. have reported a series of pyridostatin analogues as
telomeric G-quadruplex DNA ligands and demonstrated that in-
troduction of amine side chain would increase the stabilizing ability
and selectivity of compounds to G-quadruplex.13
Based on the ligands mentioned above, we known that the in-
troduction of amine side chain would increase the affinities of li-
gands for G-quadruplex DNA and improve their selectivity for G-
quadruplex because the amine side chain may dock onto the groove
of G-quadruplex DNA and interact with the G-quadruplex DNA by
forming intermolecular hydrogen bond or electrostatic interaction.
On the other hand, the quinolium scaffold would also interact with
Up to now, several classes of G-quadruplex ligands have been
developed. Most of them share a common planar aromatic core,
interacting with the terminal G-tetrad via
pep stacking. One no-
table class of potent G-quadruplex ligands is bisquinolinium
G-quartet plane by pep stacking or with the phosphate backbone
Abbreviations: CD, circular dichroism; FRET, fluorescence resonance energy
transfer; PPA, polyphosphoric acid; SNAr, nucleophilic aromatic substitution
reaction.
via electrostatic interaction. Considering that those bisquinolinium
derivatives mentioned above were all bisquinolinium derivatives
with a central aromatic core and the bisaryldiketene derivatives
reported by our group were ones with a central alicyclic ketone. So
* Corresponding author. Tel./fax: þ86 20 39943056; e-mail address: ceshzs@
mail.sysu.edu.cn (Z.-S. Huang).
0040-4020/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tet.2013.04.045

Z.-Q. Liu et al. / Tetrahedron 69 (2013) 4922e4932
4923
Fig. 1. Structures of the previously reported bisquinolinium G-quadruplex ligands (360A, E,E-1,4-bis-(vinylquinolinium)benzene, M1eM4, and pyridostatin derivatives) and de-
signed target compounds (m-BQ and p-BQ derivatives).
we plan to modify the bisaryldiketene derivatives M1eM4 by
changing the central part with an aromatic benzene core. To the
best of our knowledge, bisquinolinium derivatives with amine side
chain at C-4 position of the quinolinium ring have not been syn-
thesized and characterized as G-quadruplex binders so far. Herein,
we designed a series of disubstituted bis(vinylquinolinium)ben-
zene derivatives, which were prepared through a facile three-
component one-pot reaction. As shown in Scheme 1, we success-
fully prepared the target compounds by combining a nucleophilic
aromatic substitution reaction (SNAr reaction) with a Knoevenagel
reaction together in one-pot. By using this method, we introduced
two quinolinium cores and two amine side chains to the target
molecules through one-step reaction. In order to explore the effect
of molecular shape on stabilization and conformation of G-quad-
ruplex DNA, we prepared not only a series of 1,3-disubstituted
benzene bisquinolinium derivatives (m-BQ derivatives) but also
a series of 1,4-disubstituted benzene bisquinolinium derivatives (p-
BQ derivatives). And their interactions with G-quadruplex DNA
were investigated by using FRET-melting, FID assay, CD spectros-
copy, and molecular modeling study.
2. Results and discussion
2.1. Chemistry
As shown in Scheme 1, in order to prepare m-BQ and p-BQ de-
rivatives, we must firstly obtain the intermediate 3. 2-Methyl-4-
hydroxylquinoline (1) was prepared by condensation reaction of
starting material aniline with ethyl acetoacetate catalyzed by poly-
phosphoric acid (PPA) at 130 ꢀC in 77% yield. Then compound 1 was
treated with phosphorylchloride at 120 ꢀC for 2 h to give 4-chloro-2-
methylquinoline (2) in 65% yield. Finally, 4-chloro-2-methyl-
quinoline was methylated using methyl iodide to give a mixture of
1,2-dimethyl-4-chloroquinolinium iodide (3) and 1,2-dimethyl-4-
iodoquinolinium chloride (4) with a total yield of 90%. It was found
that the ratio of two compounds was 2:1 according to 1H NMR and
MS analysis. Considering that this mixture has no effect on next step
reaction, the mixture was used directly without further purification.
For the synthesis of target compounds, a model reaction was
chosen to optimize reaction condition using 1,3-benzenedicar-
boxyaldehyde, intermediates 3 and 4, and pyrrolidine as starting
Scheme 1. Synthetic routes for bisquinolinium derivatives (m-BQ and p-BQ derivatives): (a) PPA, 130 ꢀC; (b) POCl3, 120 ꢀC; (c) CH3I, sulfolane, 80 ꢀC; (d) solvents, reflux.

4924
Z.-Q. Liu et al. / Tetrahedron 69 (2013) 4922e4932
materials. According to the method reported by Mullins et al.14 and
considering the reactivity of our substrate, we carried out the re-
action at 80 ꢀC. The solvent for the reaction was tested with xylene,
dioxane, THF, n-butanol, and ethanol. Our experimental results
demonstrated that compounds 3 and 4 were converted to an SNAr
product in the solvents of xylene, dioxane, and THF. Especially in
solvent THF, compounds 3 and 4 were reacted with pyrrolidine and
converted almost completely to an SNAr product. In ethanol and n-
butanol, the target compound could be obtained by combining an
SNAr reaction and a Knoevenagel reaction together in one-pot with
a good yield of 92% in ethanol and 50% in n-butanol, respectively.
These results indicated that protic solvent favored the three-
component one-pot reaction, and aprotic solvent only favored the
SNAr reaction. This is consistent with the results reported by Yama-
shita et al.,15 which have shown that alcohol can accelerate Knoe-
venagel reaction. Our results indicated that the polarity of solvent
had a significant effect on the reaction, and ethanol was found to be
the best solvent for this reaction.
respectively, at a higher temperature of 120 ꢀC. When cyclic amines
were used, a lower temperature of 80 ꢀC could be used. It was
shown that reactive activity of cyclic aliphatic amines was much
stronger than that of acyclic amines, which is consistent with the
nucleophilicity of these amines.
Inorderto explore theeffectofmolecular shape ontheinteraction
of compound with G-quadruplex, a series of p-BQ derivatives were
also prepared by using 1,4-benzenedicarboxyaldehydeto replace 1,3-
benzenedicarboxyaldehyde. As shown in Table 1, the result was
similar to that of the V-shaped bisquinolinium derivatives, which
gave a higher yield (from 68 to 92%, entries 1e4) for the cyclic ali-
phatic amines and a lower yield for acyclic aliphatic amines (51%
yield in entry 5 and 62% yield in entry 6).
In addition, two bisquinolinium derivatives, m-BQ-4O and p-
BQ-4O, without amine side chain were synthesized, which were
used as the referenced compounds for the test. According to the
coupling constant of 16 Hz for the double bond protons, the con-
figuration of exocyclic double bond of the derivatives was identified
as E-form.
By using the above reaction, a series of m-BQ derivatives were
prepared with starting materials 1,3-benzenedicarboxyaldehyde,
1,2-dimethyl-4-chloroquinolinium iodide, and various amines (as
shown in Table 1) including cyclic aliphatic and acyclic aliphatic
amines. As shown in Table 1, entries 1 to 4, cyclic aliphatic amines
were generated with good yield ranging from 74 to 92%. Our results
2.2. Stabilizing ability and selectivity studies with FRET
The stabilizing ability of these bisquinolinium derivatives on
G-quadruplex DNA was evaluated with FRET-melting experi-
Table 1
Synthesis of V-shaped (m-BQs) and linear-shaped (p-BQs) disubstituted bisquinolinium derivatives
Entry
1
HNR1R2
Conditionsa
80 ꢀC, 3 h
Product
Yieldb (%)
m-BQ-Pyrro
p-BQ-Pyrro
92
92
2
3
80 ꢀC, 3 h
80 ꢀC, 3 h
m-BQ-OHEtP
p-BQ-OHEtP
78
68
m-BQ-Mor
p-BQ-Mor
91
91
4
80 ꢀC, 3 h
m-BQ-MP
p-BQ-MP
74
75
5c
120 ꢀC, overnight
m-BQ-DE-EA
p-BQ-DE-EA
52
51
6c
120 ꢀC, overnight
80 ꢀC, 3 h
m-BQ-DM-EA
p-BQ-DM-EA
m-BQ-4O
55
62
95
93
7d
d
p-BQ-4O
a
Isophthalaldehyde or terephthalaldehyde (1.0 mmol, 1.0 equiv), 1,2-dimethyl-4-chloroquinolinium iodide (2.0 mmol,2.0 equiv), amines (2.5 equiv) and the reaction was
performed at the specified temperature and for the specified length of time. All data are averaged by two experiments.
b
Yields were obtained by measuring the product after filtering off the reaction mixture and washing with ethanol.
c
Both reaction entries 6 and 7 were carried out at 120 ꢀC in a sealed tube.
d
m-BQ-4O and p-BQ-4O were synthesized by using isophthalaldehyde or terephthalaldehyde (1.0 mmol, 1.0 equiv), 1,2-dimethylquinolinium iodide (2.0 mmol, 2.0 equiv),
piperidine (0.5 equiv) as starting material.
also showed that acyclic aliphatic amines were generated with
a relatively low yield. As shown in entries 5 and 6 of Table 1, when
N1,N1-dimethylethanediamine and N1,N1-diethylethanediamine
were used as reagents, reactions led to yields of 55 and 52%,
ments.16 F21T (human telomeric G-quadruplex DNA), F22T (c-myc
promoter G-quadruplex DNA), F-c-kit1-T (c-kit1 promoter G-quad-
ruplex DNA), and F10T (hairpin duplex DNA) containing fluo-
rophores at both 50-end and 30-end, were used in this assay. The

Z.-Q. Liu et al. / Tetrahedron 69 (2013) 4922e4932
4925
melting temperatures of all G-quadruplexes were determined in the
presence of 1.0 M of bisquinolinium derivatives in TriseHCl buffer
concentration. The concentrations required to decrease the fluo-
rescence by 50%, reflecting binding to quadruplex (G4DC50) or du-
plex (dsDC50) structures, were reported in Table 3. As the results
shown in Table 3, most of synthesized BQ derivatives had a good G-
quadruplex DNA binding ability and a relatively weak duplex DNA
binding ability. p-BQ derivatives appear to be stronger binders for
DNA than that of m-BQ derivatives. And comparing with the ref-
erence compound M3, our BQ derivatives showed better c-myc
promoter G-quadruplex DNA binding ability than M3, which might
be related to that introduction of the amine side chain into the
quinolinium ring led to binding mode of ligands with G-quadruplex
DNA change to a certain extent.
m
containing 60 mM KCl. The enhanced melting temperatures (
were obtained as shown in Table 2.
DTm)
Table 2
Change in DNA melting temperatures
(DTm) determined by FRET-melting
experimenta
Ligands
F21Tb
Tm/ꢀC)
F22Tb
Tm/ꢀC)
F-c-kit1-Tb
Tm/ꢀC)
F10Tb
Tm/ꢀC)
(
D
(
D
(
D
(D
M3c
10c
9c
17c
0c
m-BQ-4O
6.1ꢂ0.7
19.0ꢂ0.8
20.3ꢂ0.1
16.8ꢂ1.0
8.42ꢂ0.2
9.0ꢂ0.1
9.0ꢂ0.3
2.4ꢂ0.4
4.8ꢂ0.2
2.8ꢂ0.3
9.3ꢂ0.1
3.5ꢂ0.3
1.8ꢂ0.6
2.5ꢂ0.2
6.0ꢂ0.4
6.2ꢂ0.5
16.3ꢂ0.1
19.3ꢂ0.2
20.2ꢂ1.0
7.7ꢂ1.0
ꢁ8.7ꢂ0.8
ꢁ5.5ꢂ0.5
3.3ꢂ0.1
3.3ꢂ0.1
4.8ꢂ0.1
4.7þ0.9
7.5ꢂ0.6
2.8ꢂ0.3
3.1ꢂ0.1
0.0ꢂ0.2
0.5ꢂ0.1
0.7ꢂ0.3
0.8ꢂ0.4
0.3ꢂ0.3
0.6ꢂ0.1
0.8ꢂ0.4
0.5ꢂ0.1
0.5ꢂ0.1
0.5ꢂ0.1
1.4ꢂ0.1
ꢁ0.3ꢂ0.3
0.5ꢂ0.1
0.6ꢂ0.2
m-BQ-Mor
m-BQ-Mp
m-BQ-OHEtP
m-BQ-Pyrro
m-BQ-DE-EA
m-BQ-DM-EA
p-BQ-4O
p-BQ-Mor
p-BQ-Mp
p-BQ-OHEtP
p-BQ-Pyrro
p-BQ-DE-EA
p-BQ-DM-EA
10.9ꢂ0.4
12.5ꢂ0.3
Table 3
ꢄ16.3ꢂ0.1
14.0ꢂ0.1
7.9ꢂ0.2
7.9ꢂ0.4
2.4ꢂ0.5
4.5ꢂ0.9
5.3ꢂ0.1
10.0ꢂ0.8
8.6ꢂ0.2
4.9ꢂ0.3
4.8ꢂ0.4
Binding affinities of ligands to quadruplexes and duplex DNA measured from FID
assay
a
a
Ligands
G4DC50
[m
M]
dsDC50
ds26
[mM]
HTG21
Pu27
M3b
m-BQ-4O
0.87
6.03
6.63
1.32
2.15
6.16
1.76
2.12
5.93
5.3
0.75
0.64
0.78
2.49
2.64
1.73
6.34
1.45
1.18
0.83
2.09
1.56
1.55
6.04
0.83
0.22
0.23
0.22
0.84
0.68
12.80
>25.00
41.43
6.00
10.89
10.29
18.8
m-BQ-Mor
m-BQ-Mp
m-BQ-OHEtP
m-BQ-Pyrro
m-BQ-DE-EA
m-BQ-DM-EA
p-BQ-4O
p-BQ-Mor
p-BQ-Mp
p-BQ-OHEtP
p-BQ-Pyrro
p-BQ-DE-EA
p-BQ-DM-EA
a
D
Tm¼Tm (DNAþligand)ꢁTm (DNA). In the absence of ligand, Tm values of annealed
F22T, F21T, F-c-kit1-T, and F10T are 78.0, 60.8, 56.7, and 61.9 ꢀC, respectively.
4.39
b
The primary structures of all oligomers used in this study were shown
>25.00
>25.00
6.67
1.33
5.41
in Supplementary data.
c
Reported in Ref. 12.
The FRET-melting data (Table 2 and Fig. S1) showed that the DTm
5.01
4.38
values for all compounds were in a wide range from 0.6 to 20.3 ꢀC.
Most of the compounds were stabilizer of G-quadruplex DNA. The
FRET results indicated that m-BQ derivatives had obviously higher
stabilizing ability to G-quadruplex than the corresponding p-BQ
derivatives, which suggested that the molecular shape made an
important contribution for the compounds to interact with G-
quadruplex. The data also showed different class of amines has
varying effect in stabilizing G-quadruplex. The compounds with
rigid cyclic amine displayed stronger stabilization ability for G-
quadruplex than those with flexible acylic amine.
a
DC50 is the ligand concentration required to displace 50% of thiazole orange
(TO). Experiments were performed at least in duplicate, and estimated error values
are within ꢂ5%.
b
Reported in Ref. 12.
2.4. Binding property studies with CD
2.4.1. Activity of compounds for inducing the formation of G-quad-
ruplex in the presence of potassium ion. In order to investigate the
activity of synthesized compounds for inducing the formation of G-
quadruplex structures, the G-rich DNAs and compounds were co-
annealed in the presence of 150 mM KCl, and CD experiments
were carried out. CD spectroscopy is a conventional method for
determining the conformation of G-quadruplex structures and the
effect of ligand binding on quadruplex structure.20 As shown in
Fig. 2A and B, in the presence of 150 mM Kþ, the CD spectrum of
HTG21 (human telomeric G-quadruplex DNA oligomer,
GGGTTAGGGTTAGGGTTAGGG) in the absence of any ligand showed
a major positive band at 290 nm, a shoulder at around 270 nm,
a small positive band at 250 nm, and a minor negative band near
234 nm, indicating that the formation of a mixture of anti-parallel
and parallel G-quadruplexes, possibly including hybrid-types as
well, which is consistent to those reported previously.17,21 Upon the
addition of different compounds to the above solution, the CD
spectra changed greatly. As shown in Fig. 2A and B, most of the
compounds induced the G-quadruplex DNA to mainly form a par-
allel structure from forms of parallel/antiparallel mixture and
hybrid-type, with a positive band at 295 nm disappearance, a pos-
itive band appearance at 260 nm, and a negative band at 245 nm
emergence. Only three compounds (m-BO-4O, p-BQ-4O, and p-
BQ-Mor) induced HTG21 G-quadruplex DNA to form a similar
conformation as those without these compounds. These results
suggested that most of the compounds favored to induce and sta-
bilize the HTG21 G-quadruplex DNA in form of parallel
Comparing with the reference compound M3 (a 1,3 ‘V’- shaped
bisquinolinium derivative) that we reported previously,12 m-BQ
derivatives m-BQ-Mor, m-BQ-Mp, and m-BQ-OHEtP showed
stronger stabilization ability to G-quadruplex DNA than that of M3,
and the p-BQ derivatives showed weaker stabilization ability than
that of M3.
Moreover, these derivatives had weak effect on the thermal
stability of the duplex DNA F10T (Table
2 and Fig. S1,
ꢁ0.3ꢂ0.3 ꢀCꢃ
D
Tmꢃ1.4ꢂ0.1 ꢀC), which implied that the derivatives
had good selectivity on G-quadruplex DNA over duplex DNA.17 In-
terestingly, we also found that m-BQ-DE-EA and m-BQ-DM-EA
could destabilize c-kit1 G-quadruplex DNA structures with
DTm
values of ꢁ8.7ꢂ0.8 ꢀC and ꢁ5.5ꢂ0.5 ꢀC, respectively.
2.3. Binding ability studies with FID
The binding affinities of ligands for the quadruplex and duplex
structures were evaluated by using fluorescence intercalator dis-
placement (FID) assays.18 The assays were performed in the pres-
ence of Kþ ions with the human telomeric sequences (HTG21), c-
myc promoter sequence Pu27, and duplex DNA ds26.
FID assay is based on the decrease in fluorescence that follows
when a ligand displaces the DNA intercalator thiazole orange (TO)
from a DNA structure.19 FID curves (Fig. S2) were obtained by
plotting the percentage displacement of TO against ligand

4926
Z.-Q. Liu et al. / Tetrahedron 69 (2013) 4922e4932
Fig. 2. CD spectra of HTG21 under the condition of 10 mM TriseHCl (7.4) in 150 mM KCl. (A, B) Effect of the m-BQ or p-BQ derivatives on HTG21 DNA in the presence of 150 mM KCl.
(C, D) CD titration spectra of compounds m-BQ-OHEtP and p-BQ-OHEtP on HTG21 DNA in the presence of 60 mM KCl.
conformation when DNA and compounds were co-annealed in the
presence of 150 mM KCl.
increase of the positive band at 290 nm and the negative band at
260 nm as the concentration of compounds increased. And it was
found that differently induced CD signals appeared, respectively,
with a negative band around 359 nm for m-BQ-OHEtP, but a posi-
tive band around 365 nm and a negative band around 430 nm for p-
BQ-OHEtP. These results indicated that the compounds could not
change the conformation of telomeric G-quadruplex DNA formed in
KCl solution, but some differences for the interaction of these two
compounds with G-quadruplex DNA might be existed. The same
experiments were also performed for c-kit1 and Pu27 G-quad-
ruplex DNA. As shown in Fig. S4, both m-BQ-OHEtP and m-BQ-
OHEtP could not change the conformation of these two G-quad-
ruplexes formed in KCl solution.
Secondly, we also tested the effect of compounds on confor-
mation of HTG21 G-quadruplex formed in 60 mM NaCl buffer. As
shown in Fig. 3A and B, in the presence of Naþ ions and absence of
the compounds, the human telomeric sequence adopted an anti-
parallel conformation.24 After addition of m-BQ-OHEtP to the so-
lution, the spectrum was almost unchanged (Fig. 3A). But the ad-
dition of p-BQ-OHEtP resulted in a great change, with a significant
decrease in the positive band at 295 nm, the disappearance of
negative band and the emergence of positive band at 260 nm, and
appearance of two inducing CD signals around 375 and 440 nm
(Fig. 3B). The results suggested that compound p-BQ-OHEtP could
induce anti-parallel structure of HTG21 G-quadruplex DNA to form
parallel structure in the presence of Naþ ions, but compound m-
BQ-OHEtP didn’t have this activity.
The same experiments were carried out for oncogene promoter
G-quadruplex DNA oligomer, c-kit1 (AGGGAGGGCGCTGGGAG-
GAGGG) and c-myc (Pu27, TGGGGAGGGTGGGGAGGGTGGG-
GAAGG), under the same condition mentioned above. The results
shown in Fig. S3 demonstrated that most of compounds could in-
duce and stabilize c-kit1 and Pu27 G-quadruplex DNAs in form of
parallel structures. It was also found that on addition of m-BQ-DE-
EA or m-BQ-DM-EA to the solution of c-kit1 G-quadruplex DNA, the
CD signal intensity obviously decreased at 260 nm (Fig. S3A). As it
was reported, the decrease in the CD signal suggested ligand-
induced disruption of the stacking between the bases of the G-
quadruplex tetrads consistent with an apparent unfolding effect.
This may be due to the ability of the two compounds to recognize
multiple binding sites with the G-quadruplex.22 On the other hand,
the structural differences of c-kit1 G-quadruplex DNA with telo-
meric and c-myc G-quadruplex DNA might be one of the most
important factors to produce the results (Table S2).23 In addition, it
was also speculated that relatively flexible amino terminal in the
side chain of two compounds could be more easily interact with the
grooves or loops of G-quadruplex DNA, which was not conducive to
the stability of G-quadruplex structure.
2.4.2. Effect of compounds on the conformation of G-quadruplex in
the presence of monovalent ion. In order to further explore whether
these synthetic compounds can change the conformation of G-
quadruplex structures, the G-quadruplex DNA was formed by
annealing in the presence of 60 mM KCl or NaCl, CD titration ex-
periments were carried out in the presence of monovalent ions.
According to FRET results, compounds m-BQ-OHEtP and p-BQ-
OHEtP were selected for the following studies.
Firstly, CD titration experiments were carried out for HTG21 G-
quadruplex DNA in the presence of 60 mM KCl. As shown in Fig. 2C
and D, upon addition of m-BQ-OHEtP or p-BQ-OHEtP to the so-
lution, the spectra for both didn’t have obvious overall changes. A
major anti-parallel structure (the positive band at 290 nm and the
negative band at 260 nm) was retained and accompanied by the
2.4.3. Effect of compounds on the conformation of G-quadruplex in
the absence of salt. In order to explore whether these synthetic
compounds can induce the formation of G-quadruplex structures
or not in the absence of salt, the CD experiments were carried out in
the absence of salt. And m-BQ-OHEtP and p-BQ-OHEtP were also
selected for the following studies. On one hand, the experiment
was performed for HTG21 DNA. As shown in Fig. 3C, as the con-
centration of m-BQ-OHEtP increased, the positive band at 255 nm
began to decrease and the positive band at 295 increased a little.
However, addition of p-BQ-OHEtP to the solution led to a decrease

Z.-Q. Liu et al. / Tetrahedron 69 (2013) 4922e4932
4927
Fig. 3. CD titration spectra of ligand m-BQ-OHEtP (A, C) or p-BQ-OHEtP (B, D) in solutions of preannealed quadruplex DNA (A, B, D, 2.5
mM) and (C, 5.0
mM) in (TriseHCl) buffer
(10 mM, pH 7.4) in the presence or absence of monovalent cations. A and B, telomeric DNA (HTG21) in the presence of 60 mM NaCl. C and D, HTG21 in the absence of monovalent
cation.
in the band at 295 nm, and a significant enhancement of the pos-
itive band at 260 nm (Fig. 3D). The results suggested that m-BQ-
OHEtP had a propensity to induce formation of anti-parallel type of
G-quadruplex structure and p-BQ-OHEtP seems more likely in-
duced the formation of parallel type structure in the absence of salt.
On the other hand, the CD titration experiment was performed
for c-kit1 and Pu27 DNA. As shown in Fig. S5A and S5B, c-kit1 se-
quence adopted a parallel structure in the absence of monovalent
cations. Upon addition of m-BQ-OHEtP and p-BQ-OHEtP, signifi-
cant changes in ellipticity were observed. And the positive band at
257 nm increased significantly with a 5 nm shift to 263 nm and the
negative band at 237 nm also increased with a shift of 2 nm to
239 nm. For Pu27G-quadruplex DNA, as shown in Fig. S5C and S5D,
two compounds could increase signal strength. The results in-
dicated that both compounds could induce and stabilize the c-kit1
and Pu27 sequences to form parallel structure of quadruplex in the
absence of salt.
2.5. Molecular modeling studies
In order to better understand how these bisquinolinium de-
rivatives interact with telomeric G-quadruplex DNA, the molecular
modeling study for the binding of compounds m-BQ-OHEtP and p-
BQ-OHEtP with the HTG21 G-quadruplex DNA was carried out
using AUTODOCK 4.0 program. The parallel propeller-type X-ray G-
quadruplex DNA structure (PDB ID 1KF1, RCSB Protein Data Bank)
was used as template for the modeling studies. For simplification,
the ratio of ligand and G-quadruplex DNA was assigned to 1:1.
As shown in Fig. 4, m-BQ-OHEtP and p-BQ-OHEtP interacted
with G-quadruplex DNA with a similar binding mode, the benzene
ring and two methylated quinoline ring of molecule stacked onto
the terminal quartet, with the nitrogen-containing side chain
docked into the grooves. But there are some differences in binding
mode for these two compounds, for m-BQ-OHEtP, the aromatic
moiety of compound interacted with the G-quartet mostly across
Fig. 4. Modeling studies for the complexes of the ligands (m-BQ-OHEtP and p-BQ-OHEtP) and G-quadruplex. Pictures were generated by using Discovery Studio 2.5 and were put
together by Photoshop CS 2.

4928
Z.-Q. Liu et al. / Tetrahedron 69 (2013) 4922e4932
the diagonal line of square G-quartet by end-stacking, while for p-
BQ-OHEtP, the aromatic moiety of the compound interacted with
the G-quartet across the two lateral of the square G-quartet, which
is responsible for the dimensions of the G-quadruplex and the two
compounds.
bind to the G-quadruplex than p-BQ-OHEtP. Derivative p-BQ-
OHEtP with large dimension could not stabilize with anti-parallel
structure for the hindrance of lateral and diagonal loops of this
G-quadruplex conformation and preferentially converted such
structure to parallel form. In addition, the dimensions of the
compounds were also too large for the dimensions of duplex DNA.
This could explain the selectivity of these two compounds to G-
quadruplex over duplex DNA.
Moreover, inducing CD signal will appear when an achiral ligand
binds tightly to a chiral host, such as DNA. Different induced CD
signal in the spectra suggested the different binding mode of DNA
binders.20b,26 As shown in all the CD titration results, different in-
ducing CD signals between 300 and 450 nm were observed for m-
BQ-OHEtP and p-BQ-OHEtP. These inducing CD signals further
suggested that both m-BQ-OHEtP and p-BQ-OHEtP interacted
with HTG21 G-quadruplex DNA in a different manner. Inspired by
these promising results, our research group is currently pursuing
structural optimization of this class of ligands by CADD methods in
parallel with organic synthesis.
The docking result showed that compound p-BQ-OHEtP had
binding free energy of
binding free energy of
m-BQ-OHEtP had stronger stabilizing ability for telomeric G-
quadruplex DNA than p-BQ-OHEtP, which is consistent with the
results from FRET-melting curve study. As the results shown in
D
D
G¼ꢁ8.50 kcal molꢁ1 and m-BQ-OHEtP had
G¼ꢁ9.03 kcal molꢁ1. It was indicated that
Table S1, the calculation of DG of m-BQ-OHEtP and p-BQ-OHEtP for
telomeric G-quadruplex DNA were ꢁ17.39 and 11.50 kJ molꢁ1
,
respectively.25
3. Conclusion
A series of 1,3- and 1,4-disubstituted benzene bisquinolinium
derivatives were designed and synthesized by a facile three-
component one-pot synthesis method, which combining a Knoe-
venagel reaction and an SNAr reaction to one-pot. By using this
method, we can introduce two quinolinium core and two nitrogen-
containing side chains to the target molecules in one step. FRET-
melting results and showed their significant selectivity for G-
quadruplex over duplex DNA, and their strong binding with G-
quadruplex DNA.
4. Experimental section
4.1. Synthesis and characterization
1H and 13C NMR spectra were recorded using TMS as the internal
standard in DMSO-d6 or CDCl3 with a Bruker BioSpin GmbH spec-
trometer at 400 and 100 MHz, respectively; Chemical shifts are re-
From the CD results, we found an obvious difference of these
two classes of compounds (1,3-disubstituted benzene and 1,4-
disubstituted benzene derivatives) in the effect on the G-quad-
ruplex DNA structures. Both of 1,3-disubstituted compound m-BQ-
OHEtP and 1,4-disubstituted compound p-BQ-OHEtP could
induce the G-quadruplex DNA to mainly form a parallel structure in
co-annealed condition in presence of Kþ. However, only p-BQ-
OHEtP could convert anti-parallel G-quadruplex to its parallel form
in presence of Naþ. After analyzing the possible conformational
forms causing by the rotation of exocyclic single bond of two classes
of compounds, we found that m-BQ-OHEtP had a variety of
ported in parts per million (ppm) relative to residual CHCl3 (
d
¼7.26,
1H;
d
¼77.0, 13C) and DMSO (
d
¼2.50,1H;
d
¼39.52,13C) in the corre-
sponding deuterated solvents. Mass spectra (MS) were recorded on
a Shimadzu LCMS-2010A instrument with an ESI or ACPI mass se-
lective detector, and high-resolution mass spectra (HRMS) on Shi-
madzu LCMS-IT-TOF. Melting points (mp) were determined using an
SRS-OptiMelt automated melting point instrument without cor-
rection. Flash column chromatography was performed with silica
gel (200ꢁ300 mesh) purchased from Qingdao Haiyang Chemical Co.
Ltd. The purities of synthesized compounds were confirmed to be
higher than 95% by analytical HPLC performed with a dual pump
Shimadzu LC-20AB system equipped with an Ultimate XB-C18 col-
umn and eluted with methanol/water (70%) containing 0.1% TFA at
ꢀ
structural forms with different dimensions (from 15.4 to 19.7 A)
ꢀ
while p-BQ-OHEtP had only one dimension (20.0 A), as shown in
Fig. 5. So m-BQ-OHEtP has more possible combination modes to
Fig. 5. Dimensions and comparison with m-BQ-OHEtP and p-BQ-OHEtP. Molecular lengths evaluated after molecular mechanics calculations (MM2, with Chem3D Ultra 8.0,
CambridgeSoft Corp., MA, USA).

Z.-Q. Liu et al. / Tetrahedron 69 (2013) 4922e4932
4929
a flow rate of 0.3 mL minꢁ1. Aniline, ethyl acetoacetate, poly-
phosphoric acid, POCl3 ethanol, n-butanol, 1,4-dioxane, THF, and all
the amines were commercially available.
4.5.1. Synthesis of 2,20-((1E,10E)-1,3-phenylenebis(ethene-2,1-diyl))
bis(1-methyl-4-(4-methylpiperazin-1-yl)-quinolin-1-ium)iodide (m-
BQ-MP). Following the general procedure for bisquinolinium de-
rivatives: product was obtained as yellow-green solid. Mp
254.3e255.5 ꢀC.1H NMR (400 MHz, DMSO-d6)
d: 8.39 (s,1H), 8.33 (d,
4.2. Synthesis of 2-methylquinolin-4-ol (1)27
2H, J¼8.0 Hz), 8.19 (d, 2H, J¼8.0 Hz), 8.07 (t, 2H, J¼8.0 Hz), 8.00 (d,
2H, J¼4.0 Hz), 7.92 (d, 2H, J¼16.0 Hz), 7.87 (d, 2H, J¼16.0 Hz), 7.77 (t,
2H, J¼8.0 Hz), 7.67 (t, 1H, J¼8.0 Hz), 7.55 (s, 2H), 4.28 (s, 6H), 3.87 (s,
To an equimolar solution of aniline (14.5 g, 155 mmol) and ethyl
acetoacetate (20.2 g, 155 mmol) was added polyphosphoric acid
(125.0 g). The reaction mixture was stirred at 130 ꢀC for 2 h. Re-
action completion was monitored by TLC. The reaction mixture was
poured into ice water (500 mL) slowly with vigorous stirring. The
precipitated solid was filtered and dried in vacuum oven for 4 h to
get the crude product as white solid. The crude product was taken
as such for the next step without further purification. Compound 1
was obtained as white solid with a yield of 77%. Mp 258.6e258.9 ꢀC.
8H), 2.63 (s, 8H), 2.31(s, 6H).13C NMR (100 MHz, DMSO-d6)
d: 159.50,
158.31, 153.47, 141.80, 140.84, 135.78, 133.92, 130.03,128.26, 126.72,
126.24, 121.17, 119.46, 119.18, 105.20, 54.27, 51.67, 45.41, 38.51. ESI-
HRMS: calcd for C40H46N6 ½Mꢅ2þ: 305.1886; found:305.1909.
2þ
4.5.2. Synthesis of 2,20-((1E,10E)-1,4-phenylenebis(ethene-2,1-diyl))
bis(1-methyl-4-(4-methylpiperazin-1-yl)-quinolin-1-ium)iodide (p-
BQ-MP). Following the general procedure for bisquinolinium de-
rivatives: product was obtained as yellow-green solid. Mp
1H NMR (400 MHz, DMSO-d6)
7.61e7.56 (m, 1H), 7.560e7.47 (m, 1H), 7.27e7.23 (m, 1H), 2.33(s,
3H). 13C NMR (100 MHz, DMSO-d6)
: 176.67, 149.58, 140.08, 131.35,
d
: 11.58 (s, 1H), 8.02 (d, 1H, J¼8.0 Hz),
265.6e267.2 ꢀC. 1H NMR (400 MHz, DMSO-d6)
d: 8.33 (d, 2H,
d
J¼8.0 Hz), 8.19 (d, 2H, J¼8.0 Hz), 8.08 (d, 2H, J¼4.0 Hz), 8.04 (s, 4H),
7.92 (d, 2H, J¼16.0 Hz), 7.86 (d, 2H, J¼16.0 Hz), 7.77 (t, 2H, J¼8.0 Hz),
7.55 (s, 2H), 4.26 (s, 6H), 3.85 (s, 8H), 2.63 (s, 8H), 2.31 (s, 6H).13C NMR
124.73, 124.46, 122.61, 117.69, 108.33, 19.40; MS (ESIþAPCI) m/z:
160.1 [MþH]þ1
.
(100 MHz, DMSO-d6) d: 159.46,153.46,141.66,140.88,136.83,133.88,
4.3. Synthesis of 4-chloro-2-methylquinoline (2)27
128.97, 126.72, 126.25, 121.35, 119.49, 119.19, 105.20, 54.25, 51.60,
45.41, 38.26. ESI-HRMS: calcd for C40H46N6 ½Mꢅ2þ: 305.1886;
2þ
A mixture of 1 (25 g, 88.9 mmol) and freshly distilled POCl3
(125 mL) was heated at 120 ꢀC for 2 h. The reactionwas monitored by
TLC. After completion of the reaction, excess of POCl3 was distilled off.
The residue thus obtained was stirred with ice water for 15 min, then
adjust the pH value to 7. And the compound was collected by filtra-
tion and washed with water. Finally purified by column chromatog-
raphy with CH2Cl2 as eluent. Compound 2 was obtained as pale
yellow solid with a yield of 65%. Mp 42.6e43.5 ꢀC.1H NMR (400 MHz,
found: 305.1916.
4.5.3. Synthesis of 2,20-((1E,10E)-1,3-phenylenebis(ethene-2,1-diyl))
bis(1-methylquinolin-1-ium)iodide (m-BQ-4O). Following the gen-
eral procedure for bisquinolinium derivatives, the product was ob-
tained as a brown solid. Mp 264.1e265.5 ꢀC. 1H NMR (400 MHz,
DMSO-d6)
d
: 9.18 (d, 2H, J¼8.0 Hz), 8.64 (d, 4H, J¼8.0 Hz), 8.53 (s,
1H), 8.42 (d, 2H, J¼8.0 Hz), 8.28e8.23 (m, 4H), 8.16 (d, 2H, J¼8.0 Hz),
8.10 (d, 2H, J¼16.0 Hz), 8.02 (t, 2H, J¼6.0 Hz), 7.76 (t, 1H, J¼8.0 Hz),
DMSO-d6)
d
: 8.14 (d, 1H, J¼8.0 Hz), 8.02 (d, 1H, J¼8.0 Hz), 7.72e7.68
(m, 1H), 7.56e7.52 (m, 1H), 7.34 (s, 1H), 2.69 (s, 3H). 13C NMR
4.65 (s, 6H). 13C NMR (100 MHz, DMSO-d6)
d: 155.90, 145.53, 144.60,
(100 MHz, DMSO-d6) d: 153.53, 143.27, 137.30, 125.10, 123.58, 121.38,
139.21, 135.62, 135.13, 131.41, 130.11, 129.87, 129.34, 129.27, 128.03,
119.41, 118.60, 116.63, 19.78; MS (ESIþAPCI) m/z: 179.1 [MþH]þ1
.
2þ
2þ
121.37, 120.68, 119.44, 40.46. ESI-HRMS: calcd for C30H26N2 ½Mꢅ
:
207.1043; found:207.1052.
4.4. Synthesis of 1,2-dimethyl-4-chloroquinolin-1-ium io-
dode (3) and 1,2-dimethyl-4-iodo-quinolin-1-ium chloride
(4)28
4.5.4. Synthesis of 2,20-((1E,10E)-1,3-phenylenebis(ethene-2,1-diyl))
bis(4-(4-(2-hydroxyethyl)piperazin-1-yl)-1-methylquinolin-1-ium)
iodide (m-BQ-OHEtP). Following the general procedure for bis-
quinolinium derivatives, the product was obtained as a yellow-
A suspension of compound 2 (8.0 g, 90 mmol), iodomethane
11.2 mL (13.0 g, 180 mmol), and sulfolane (25 g) was heated in
a sealed flask overnight at 80 ꢀC. A yellow precipitate was formed.
The reaction mixture was allowed to cool to room temperature and
then placed in an ice bath, precipitated further with a mixture of ice
cold anhydrous diethyl ether (250 mL), filtered, and washed thor-
oughly with anhydrous ethanol to afford 3 as a yellow solid with
green solid. Mp 215.8e216.7 ꢀC. 1H NMR (400 MHz, DMSO-d6)
d:
8.56 (s, 1H), 8.38 (d, 2H, J¼12.0 Hz), 8.23 (d, 2H, J¼8.0 Hz), 8.10 (t,
2H, J¼8.0 Hz), 8.03 (m, 4H), 7.95 (d, 2H, J¼16.0 Hz), 7.80 (t, 2H,
J¼8.0 Hz), 7.68 (t, 1H, J¼8.0 Hz), 7.63 (s, 2H), 4.34 (s, 6H), 3.97 (s,
8H), 3.72 (s, 4H), 2.34 (s, 12H). 13C NMR (100 MHz, DMSO-d6)
d:
159.27, 153.82, 142.28, 140.72, 135.75, 134.06, 130.33, 129.53, 128.21,
a yield of 90%. Compound 3: 1H NMR (400 MHz, DMSO-d6)
d: 8.67
126.65, 126.56, 121.06, 119.60, 119.31, 105.78, 58.74, 56.85, 56.64,
51.79, 50.08, 38.83. ESI-HRMS: calcd for C42H50N6O2 ½Mꢅ2þ
:
2þ
(d, 1H, J¼8.0 Hz), 8.56 (d, 1H, J¼12.0 Hz), 8.54 (s, 1H), 8.33 (t, 1H,
J¼8.0 Hz), 8.12 (t, 1H, J¼8.0 Hz), 4.44 (s, 3H), 3.08 (s, 3H); MS
(ESIþAPCI) m/z: 192.1. Compound 4: 1H NMR (400 MHz, DMSO-d6)
335.1992; found:335.2009.
4.5.5. Synthesis of 2,20-((1E,10E)-1,4-phenylenebis(ethene-2,1-diyl))
bis(1-methylquinolin-1-ium)iodide (p-BQ-4O). Following the gen-
eral procedure for bisquinolinium derivatives, the product was
obtained as a brown solid. Mp 296.5e297.8 ꢀC. 1H NMR (400 MHz,
d
: 8.90 (s, 1H), 8.56 (d, 1H, J¼12.0 Hz), 8.26 (t, 1H, J¼8.0 Hz), 8.05 (t,
1H, J¼8.0 Hz), 4.36 (s, 3H), 2.99 (s, 3H); MS (ESIþAPCI) m/z: 284.0.
4.5. General method for preparation of bisquinolinium
derivatives
DMSO-d6)
d
: 9.15 (d, 2H, J¼8.0 Hz), 8.62 (d, 4H, J¼8.0 Hz), 8.40 (d,
2H, J¼8.0 Hz), 8.24 (t, 4H, J¼12.0 Hz), 8.17 (s, 4H), 8.11 (d, 2H,
2þ
2þ
A suspension of 3 and 4 (0.64 g, 2.0 mmol), dialdehydes (ter-
ephthalaldehyde or isophthalaldehyde) (0.134 g, 1.0 mmol), and
ethanol or n-butanol (15 mL) was heated in flask or sealed tube 3 h
at 80 or 120 ꢀC. A precipitate was formed during the process of
reaction. The reaction mixture was allowed to cool to room tem-
perature and then placed the flask or sealed tube was placed in ice
bath, filtered, and washed thoroughly with anhydrous ethanol to
afford bisquinolinium derivatives as a solid.
J¼16.0 Hz), 8.00 (t, 2H, J¼8.0 Hz), 4.62 (s, 6H). C30H26N2 ½Mꢅ
:
207.1043; found:207.1041.
4.5.6. Synthesis of 2,20-((1E,10E)-1,4-phenylenebis(ethene-2,1-diyl))
bis(1-methyl-4-morpholinoquinolin-1-ium) iodide (p-BQ-Mor).
Following the general procedure for bisquinolinium derivatives, the
product was obtained as a brown solid. Mp 296.5e297.7 ꢀC. 1H NMR
(400 Hz, DMSO-d6)
d: 8.35 (d, 2H, J¼12.0 Hz), 8.23 (d, 2H, J¼8.0 Hz),

4930
Z.-Q. Liu et al. / Tetrahedron 69 (2013) 4922e4932
2þ
2þ
8.08 (t, 2H, J¼6.0 Hz), 8.04 (s, 4H), 7.93(d, 2H, J¼16.0 Hz), 7.87(d, 2H,
J¼16.0 Hz), 7.77 (t, 2H, J¼8.0 Hz), 7.56 (s, 2H), 4.28 (s, 6H), 3.90 (d, 8H,
J¼4.0 Hz), 3.87 (d, 8H, J¼4.0 Hz). ESI-HRMS: calcd for
42.78, 38.17, 37.94. ESI-HRMS: calcd for
293.1886; found: 293.1899.
C
38H46N6 ½Mꢅ
:
C38H40N4O2 ½Mꢅ2þ: 292.1570; found: 292.1589.
4.5.12. Synthesis of 2,20-((1E,10E)-1,3-phenylenebis(ethene-2,1-diyl))
bis(4-((2-(dimethylamino)ethyl)amino)-1-methylquinolin-1-ium)io-
dide (m-BQ-DM-EA). Following the general procedure for bisqui-
nolinium derivatives, the product was obtained as a yellow-green
2þ
4.5.7. Synthesis of 2,20-((1E,10E)-1,4-phenylenebis(ethene-2,1-diyl))
bis(4-(4-(2-hydroxyethyl)piperazin-1-yl)-1-methylquinolin-1-ium)
iodide (p-BQ-OHEtP). Following the general procedure for bisqui-
nolinium derivatives, the product was obtained as a yellow-green
solid. Mp 271.2e272.4 ꢀC. 1H NMR (400 MHz, DMSO-d6)
d: 9.13 (s,
1H), 8.59 (d, 2H, J¼8.0 Hz), 8.30 (d, 2H, J¼8.0 Hz), 8.10 (t, 2H,
J¼8.0 Hz), 7.86 (d, 2H, J¼16.0 Hz), 7.85 (t, 2H, J¼8.0 Hz), 7.67 (t, 1H,
J¼8.0 Hz), 4.25 (s, 6H), 4.10 (s, 2H), 3.50 (s, 8H), 2.92 (s, 12H). 13C
solid. Mp 236.9e238.2 ꢀC. 1H NMR (400 MHz, DMSO-d6)
d: 8.32
(d, 2H, J¼8.0 Hz), 8.19 (d, 2H, J¼8.0 Hz), 8.07 (d, 2H, J¼8.0 Hz), 8.04
(s, 4H), 7.86 (d, 2H, J¼16.0 Hz), 7.76 (t, 2H, J¼8.0 Hz), 7.53 (s, 2H),
4.53 (t, 2H), 4.26 (s, 6H), 3.86 (s, 8H), 3.61e3.57 (m, 4H), 2.75 (s, 8H),
NMR (100 MHz, DMSO-d6) d: 154.09, 141.58, 139.34, 135.83, 134.14,
130.07, 129.49, 127.87, 126.47, 123.87, 121.46, 119.01, 117.53, 99.49,
97.98, 54.40, 42.73, 38.27, 38.18. ESI-HRMS: calcd for
2.54 (t, 4H, J¼4.0 Hz). 13C NMR (100 MHz, DMSO-d6)
d: 159.32,
C38H46N6 ½Mꢅ2þ: 293.1886; found: 293.1898.
2þ
153.40, 141.61, 140.86, 136.81, 133.89, 128.97, 126.73, 126.18, 121.33,
119.40, 119.17, 105.08, 98.25, 59.79, 58.54, 52.83, 51.77. ESI-HRMS:
calcd for C46H60N8 ½Mꢅ2þ: 335.1992; found:335.1988.
4.5.13. Synthesis of 2,20-((1E,10E)-1,4-phenylenebis(ethene-2,1-diyl))
bis(4-((2-(diethylamino)ethyl)amino)-1-methylquinolin-1-ium)io-
dide (p-BQ-DE-EA). Following the general procedure for bisqui-
nolinium derivatives, the product was obtained as a yellow-green
2þ
4.5.8. Synthesis of 2,20-((1E,10E)-1,3-phenylenebis(ethene-2,1-diyl))
bis(1-methyl-4-morpholinoquinolin-1-ium) iodide (m-BQ-Mor).
Following the general procedure for bisquinolinium derivatives, the
product was obtained as a brown solid. Mp 309.8e311.5 ꢀC. 1H NMR
solid. Mp 307.4.3e308.5 ꢀC. 1H NMR (400 MHz, DMSO-d6)
d: 9.52
(s, 2H), 9.04 (s, 2H), 8.54 (d, 2H, J¼8.0 Hz), 8.30 (d, 2H, J¼8.0 Hz),
8.06 (s, 4H), 7.93 (d, 2H, J¼16.0 Hz), 7.85 (d, 2H, J¼16.0 Hz), 7.28 (s,
2H), 4.23 (s, 6H), 4.09 (t, 4H, J¼12.0 Hz), 3.52 (s, 12H), 1.24 (s, 12H).
(400 MHz, DMSO-d6)
d
: 8.41 (s,1H), 8.35 (d, 2H, J¼8.0 Hz), 8.24 (d, 2H,
J¼8.0 Hz), 8.08 (t, 2H, J¼4.0 Hz), 8.01 (d, 2H, J¼8.0 Hz), 7.95 (d, 2H,
J¼16.0 Hz), 7.89 (d, 2H, J¼16.0 Hz), 7.78 (t, 2H, J¼8.0 Hz), 7.68 (t, 1H,
J¼8.0 Hz), 7.57 (s, 2H), 4.30 (s, 6H), 3.90 (s, 8H), 3.88 (s, 8H). 13C NMR
13C NMR (100 MHz, DMSO-d6)
d: 154.02, 141.31, 139.40, 136.73,
134.18, 128.91, 126.61, 123.53, 121.70, 120.19, 119.11, 117.46, 97.97,
48.59, 46.58, 38.35, 37.96, 8.45. ESI-HRMS: calcd for
(100 MHz, DMSO-d6) d: 159.56,153.68,143.58,142.00,140.83,135.80,
C42H54N6 ½Mꢅ2þ: 321.2199; found: 321.2204.
2þ
133.98, 130.06, 126.71, 126.38, 121.15, 119.43, 119.26, 113.22, 105.26,
65.82, 52.11, 34.63. ESI-HRMS: calcd for C38H40N4O2 ½Mꢅ2þ
:
2þ
292.1570; found:292.1565.
4.5.14. Synthesis of 2,20-((1E,10E)-1,3-phenylenebis(ethene-2,1-diyl))
bis(4-((2-(diethylamino)ethyl)amino)-1-methylquinolin-1-ium)io-
dide (m-BQ-DE-EA). Following the general procedure for bisqui-
nolinium derivatives, the product was obtained as a yellow-green
4.5.9. Synthesis of 2,20-((1E,10E)-1,3-phenylenebis(ethene-2,1-diyl))
bis(1-methyl-4-(pyrrolidin-1-yl)quinolin-1-ium)iodide (m-BQ-Pyr-
ro). Following the general procedure for bisquinolinium de-
solid. Mp 289.7e291.3 ꢀC. 1H NMR (400 Hz, DMSO-d6)
d: 9.07 (s,
rivatives, the product was obtained as
a
brown solid. Mp
: 8.56 (d, 2H
1H), 8.58 (d, 2H, J¼8.0 Hz), 8.31 (d, 2H, J¼8.0 Hz), 8.09 (t, 2H,
J¼8.0 Hz), 8.01 (t, 2H, J¼8.0 Hz), 7.94 (d, 2H, J¼16.0 Hz), 7.84 (t, 2H,
J¼8.0 Hz), 7.65 (t,1H, J¼8.0 Hz), 7.31 (s, 2H), 4.27 (s, 6H), 4.16 (s, 2H),
314.1e316.2 ꢀC. 1H NMR (400 MHz, DMSO-d6)
d
J¼8.0 Hz), 8.32 (d, 1H, J¼8.0 Hz), 8.20 (d, 2H, J¼8.0 Hz), 8.03 (t, 2H,
J¼8.0 Hz), 7.94 (d, 2H, J¼8.0 Hz), 7.80 (d, 4H, J¼4.0 Hz), 7.70 (t, 2H,
J¼8.0 Hz), 7.64 (t,1H, J¼8.0 Hz), 7.05 (s, 2H), 4.15 (s, 6H), 4.05 (s, 8H),
3.35 (s, 16H), 1.27 (s, 12H). 13C NMR (100 MHz, DMSO-d6)
d: 154.44,
153.98, 141.54, 139.34, 135.77, 134.17, 130.14, 129.44, 127.92, 126.59,
123.61, 121.42, 119.10, 117.46, 97.96, 48.80, 46.64, 38.35, 38.09, 8.67.
2.08 (s, 8H). 13C NMR (100 MHz, DMSO-d6)
d: 154.48, 151.78, 140.70,
ESI-HRMS: calcd for C42H54N6 ½Mꢅ2þ: 321.2199; found: 321.2208.
2þ
140.24, 135.87, 133.44, 129.54, 127.80, 127.25, 124.99, 121.62, 118.17,
112.01, 111.96, 100.69, 53.58, 37.97, 25.12. ESI-HRMS: calcd for
C38H40N4 ½Mꢅ2þ: 276.1621; found: 276.1634.
4.6. FRET-melting experiments
2þ
4.5.10. Synthesis of 2,20-((1E,10E)-1,4-phenylenebis(ethene-2,1-diyl))
bis(1-methyl-4-(pyrrolidin-1-yl)quinolin-1-ium)iodide (p-BQ-Pyr-
ro). Following the general procedure for bisquinolinium de-
To evaluate the stabilization and selectivity of bisquinolinium
derivatives for G-quadruplex DNA, FRET-melting experiments were
carried out. The human telomeric G-quadruplex DNA [F21T, 50-
FAM-d(GGG[TTAGGG]3)-TAMRA-30] and two oncogene promoter
G-quadruplexes c-myc [F22T, FAM-d(50-TGAG3TG3TAG3TG3TA2)-
TAMRA-30], c-kit1[F-c-kit1-T, 50-FAM-d(AGGGAGGGCGCTGGG-
AGGAGGG)-TAMRA-30] containing fluorophores at both 50-end and
30-end were used in this assay. A sequence that forms hairpin du-
plex DNA structure F10T (50-FAM-dTATAGCTATAHEGTATAGC-
TATATAMRA-30) as a non-quadruplex control was also involved.
The melting temperature of all G-quadruplexes was determined in
rivatives, the product was obtained as
a
brown solid. Mp
: 8.56 (d, 2H,
307.2e308.5 ꢀC. 1H NMR (400 MHz, DMSO-d6)
d
J¼12.0 Hz), 8.20 (d, 2H, J¼8.0 Hz), 8.03 (t, 2H, J¼8.0 Hz), 7.98 (s, 4H),
7.79 (s, 4H, J¼8.0 Hz), 7.70 (t, 2H, J¼4.0 Hz), 7.05 (s, 2H), 4.13 (s, 6H),
4.05 (s, 8H), 2.08 (s, 8H). 13C NMR (100 MHz, DMSO-d6)
d: 154.44,
151.78, 145.35, 140.73, 140.06, 136.66, 133.43, 128.71, 127.23, 124.97,
121.69, 118.17, 100.66, 53.46, 37.84, 25.09. ESI-HRMS: calcd for
C38H40N4 ½Mꢅ2þ: 276.1621; found: 276.1638.
2þ
the presence of 1
buffer containing 60 mM KCl. The result of FRET-melting assay was
represented as enhanced melting temperature ( Tm) reported in
mM of bisqinolinium derivatives in TriseHCl
4.5.11. Synthesis of 2,20-((1E,10E)-1,4-phenylenebis(ethene-2,1-diyl))
bis(4-((2-(dimethylamino)ethyl)amino)-1-methylquinolin-1-ium)io-
dide (p-BQ-DM-EA). Following the general procedure for bisqui-
nolinium derivatives, the product was obtained as a yellow-green
D
Table 1. The melting of the G-quadruplex was monitored alone or in
the presence of various compounds. Final analysis of the data was
carried out using Excel and Origin 7.5 (OriginLab Corp.).
solid. Mp 307.4e308.2 ꢀC. 1H NMR (400 MHz, DMSO-d6)
d: 9.88
(s, 2H), 9.12 (s, 2H), 8.67 (d, 2H, J¼8.0 Hz), 8.30 (d, 2H, J¼8.0 Hz),
8.07 (s, 4H), 7.94 (d, 2H, J¼16.0 Hz), 7.87 (d, 2H, J¼16.0 Hz), 7.30 (s,
2H), 4.23 (s, 6H), 4.10 (s, 4H), 3.51 (s, 4H), 2.93 (s, 12H). 13C NMR
4.7. TO FIDs assay
(100 MHz, DMSO-d6)
d
: 154.38, 154.06, 141.36, 139.35, 136.74,
Thiazole orange (TO) was purchased from Aldrich and used
134.17, 128.90, 126.49, 123.71, 121.67, 119.03, 117.53, 98.04, 54.41,
without
further
purification.
Three
quadruplex-forming

Z.-Q. Liu et al. / Tetrahedron 69 (2013) 4922e4932
4931
oligonucleotides used were: HTG21, Pu27. And duplex DNA ds26
was also used. The oligonucleotides were annealed at 10 mm in Kþ
buffer, containing lithium cacodylate (10 mm) at pH 7.3 and KCl
(100 mm), by heating at 95 ꢀC for 5 min and cooling over ice to favor
the intramolecular folding by kinetic trapping. The duplex DNA
ds26 was prepared by heating the self-complementary strand at
95 ꢀC followed by a slow cooling to room temperature over 3 h.
Each experiment was performed in a cell with a path length of 1 cm
in Kþ buffer with a total volume of 1.25 mL. The FID assay was
designed as follows: 0.25 mm prefolded DNA is mixed with TO
partial atomic charges were assigned. The nonpolar hydrogens of
the ligands were merged, and rotatable bonds were assigned. The
resulting G-quadruplex structure was used as an input for the
AUTOGRID program. AUTOGRID gave a precalculated atomic af-
finity grid map for each atom type in the ligand, plus an electro-
statics map, and a separate desolvation map present in the
substrate molecule. The dimensions of the active site box, which
was placed at the center of the G-quadruplex, were set to
ꢀ
ꢀ
ꢀ
ꢀ
60 Aꢆ60 Aꢆ60 A with the grid points 0.375 A apart. Docking cal-
culations were carried out using the Lamarckian genetic algorithm
(LGA). Initially, we used a population of random individuals (pop-
ulation size: 150), a maximum number of 25,000,000 energy
evaluations, a maximum number of generations of 27,000, and
a mutation rate of 0.02. One hundred independent docking runs
were carried out for each ligand. The resulting positions were
(0.250 mM for HTG21 and Pu27; 0.50 mM for ds26). Each ligand-
addition step (11 steps of 4.0 mL, from 0.2 to 10 equiv) was fol-
lowed by a 3 min equilibration period after which the fluorescence
spectrum was recorded with the background subtracted. The per-
centage of displacement was calculated from the fluorescence area
(FA, 510e750 nm, Ex¼501 nm) by using percentage of dis-
placement¼100ꢁ[(FA/FA0)ꢆ100], with FA0 being the fluorescence
of TO bound to DNA without added ligand. The percentage of dis-
placement was plotted as a function of the concentration of added
ligand. Final analysis of the data was carried out by using Excel and
Origin 7.5 software.
ꢀ
clustered according to a root-mean-square criterion of 0.5 A.
Acknowledgements
We thank the Natural Science Foundation of China (21172272,
91213302), the Specialized Research Fund for the Doctoral Program
of Higher Education (20110171110051), and the Science Foundation
of Guangzhou (2009A1-E011-6) for financial support of this study.
4.8. CD measurements experiments
CD measurements were performed on a Chirascan circular di-
chroism spectrophotometer (Applied Photophysics) using a quartz
cuvettes of 2 mm optical path length and over a wavelength range
of 230e450 nm at 1 nm bandwidth, 1 nm step size, and 0.5 s time
per point. The oligomer HTG21 (GGGTTAGGGTTAGGGTTAGGG) at
Supplementary data
Supplementary data associated with this article can be found in
the online version, at http://dx.doi.org/10.1016/j.tet.2013.04.045.
a final concentration of 5 mM was resuspended in TriseHCl buffer
References and notes
(10 mM, pH 7.4) containing the specific cations and the compounds
to be tested. The samples were heated to 95 ꢀC for 5 min, then
gradually cooled to room temperature and incubated at 4 ꢀC
overnight. A buffer baseline was collected in the same cuvette and
subtracted from the sample spectra. The CD spectra were obtained
by taking the average of at least three scans at 25 ꢀC. Then, CD ti-
1. (a) Lipps, H. J.; Rhodes, D. Trends Cell Biol. 2009, 19, 414e422; (b) Huppert, J. L.
Chem. Soc. Rev. 2008, 37, 1375e1384.
2. (a) Cogoi, S.; Xodo, L. E. Nucleic Acids Res. 2006, 34, 2536e2549; (b) Todd, A. K.;
Johnston, M.; Neidle, S. Nucleic Acids Res. 2005, 33, 2901e2907; (c) Qin, Y.;
Rezler, E. M.; Gokhale, V.; Sun, D.; Hurley, L. H. Nucleic Acids Res. 2007, 35,
7698e7713; (d) Bejugam, M.; Sewitz, S.; Shirude, P. S.; Rodriguez, R.; Shahid, R.;
Balasubramanian, S. J. Am. Chem. Soc. 2007, 129, 12926e12927; (e) Huppert, J. L.;
Balasubramanian, S. Nucleic Acids Res. 2005, 33, 2908e2916; (f) Huppert, J. L.;
Balasubramanian, S. Nucleic Acids Res. 2006, 35, 406e413.
3. (a) Siddiqui-Jain, A.; Grand, C. L.; Bearss, D. J.; Hurley, L. H. Proc. Natl. Acad. Sci. U.
S.A. 2002, 99, 11593e11598; (b) Phan, A. T.; Kuryavyi, V.; Gaw, H. Y.; Patel, D. J.
Nat. Chem. Biol. 2005, 1, 167e173; (c) Yang, D.; Hurley, L. H. Nucleosides Nucle-
otides Nucleic Acids 2006, 25, 951e968; (d) Phan, A. T.; Modi, Y. S.; Patel, D. J. J.
Am. Chem. Soc. 2004, 126, 8710e8716; (e) Kumar, N.; Basundra, R.; Maiti, S.
Nucleic Acids Res. 2009, 37, 103321e103331; (f) Thakur, R. K.; Kumar, P.; Halder,
K.; Verma, A.; Kar, A.; Parent, J. L.; Basundra, R.; Kumar, A.; Chowdhury, S.
Nucleic Acids Res. 2009, 37, 172e183; (g) Simonsson, T.; Pecinka, P.; Kubista, M.
Nucleic Acids Res. 1998, 26, 1167e1172.
4. (a) 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,
10584e10589; (b) Fernando, H.; Reszka, A. P.; Huppert, J.; Ladame, S.; Rankin,
S.; Venkitaraman, A. R.; Neidle, S.; Balasubramanian, S. Biochemistry 2006, 45,
7854e7860; (c) Waller, Z. A. E.; Sewitz, S. A.; Hsu, S. T.; Balasubramanian, S. J.
Am. Chem. Soc. 2009, 131, 12628e12633.
5. (a) Ou, T.-M.; Lu, Y.-J.; Huang, Z.-S.; Wang, X.-D.; Tan, J.-H.; Chen, Y.; Ma, D.-L.;
Wong, K.-Y.; Tang, J.-C.; Chan, A.-S.; Gu, L.-Q. J. Med. Chem. 2007, 50, 1465e1474;
(b) Cogoi, S.; Paramasivam, M.; Spolaore, B.; Xodo, L. E. Nucleic Acids Res. 2008,
36, 3765e3780.
6. (a) Collie, G. W.; Parkinson, G. N. Chem. Soc. Rev. 2011, 40, 5867e5892; (b) Xu, Y.
Chem. Soc. Rev. 2011, 40, 2719e2740.
7. Balasubramanian, S.; Hurley, L. H.; Neidle, S. Nat. Rev. Drug Discov. 2011, 10,
261e275.
8. Zahler, A. M.; Williamson, J. R.; Cech, T. R.; Prescott, D. M. Nature 1991, 350,
718e720.
9. Neidle, S. Curr. Opin. Struct. Biol. 2009, 19, 239e250.
10. (a) Granotier, C.; Pennarun, G.; Riou, L.; Hoffschir, 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, 4182e4190; (b) Rodriguez, R.; Mller, S.;
Yeoman, J. A.; Trentesaux, C.; Riou, J.-F.; Balasubramanian, S. J. Am. Chem. Soc.
2008, 130, 15758e15759.
tration was performed at a fixed HTG21 concentration (2.5
mM)
with various concentrations of the compounds in TriseHCl buffer in
the absence or in the presence of 60 mM KCl or NaCl. After each
addition of compound, the reaction mixture was stirred and
allowed to equilibrate for at least 5 min (until no elliptic changes
were observed) and a CD spectrum was collected. Final analysis of
the data was carried out using Origin 7.5 (OriginLab Corp.).
The CD experiments for DNA oligomers HTG21, Pu27
(TGGGGAGGGTGGGGAGGGTGGGGAAGG), and c-kit1 (AGG-
GAGGGCGCTGGGAGGAGGG) were also carried out at a final con-
centration of 5
containing the tested ligands (25
KCl or NaCl. In the presence of 150 mM KCl, The samples containing
m
M in 10 mM TriseHCl buffer, pH 7.4 at 25 ꢀC
mM) in the absence or presence of
HTG21(5 mM) and compounds (25 mM) in 10 mM TriseHCl buffer,
pH 7.4, were annealed by heating at 95 ꢀC for 5 min, and then
gradually cooled to room temperature and incubated at 4 ꢀC
overnight. The recording of CD spectra was carried out following
the CD experiment. For CD titration experiments, the samples
containing HTG21, Pu27 and c-kit1 DNA (2.5
mM) in 10 mM
TriseHCl buffer in the absence of salt or the presence of 60 mM KCl
or NaCl, pH 7.4, were annealed by heating at 95 ꢀC for 5 min, and
then gradually cooled to room temperature and incubated at 4 ꢀC
overnight. The recording of CD spectra was carried out following
the CD experiment.
4.9. Molecular modeling study
11. Czerwinska, I.; Juskowiak, B. Int. J. Biol. Macromol. 2012, 51, 576e582.
12. Peng, D.; Tan, J.-H.; Chen, S.-B.; Ou, T.-M.; Gu, L.-Q.; Huang, Z.-S. Bioorg. Med.
Chem. 2010, 18, 8235e8242.
Docking studies were carried out using the AUTODOCK 4.0
program.29 By using ADT,30 nonpolar hydrogens of telomeric G-
quadruplex were merged to their corresponding carbons, and
€
13. Muller, S.; Sanders, D. A.; Di Antonio, M.; Matsis, S.; Riou, J.-F.; Rodriguez, R.;
Balasubramaniana, S. Org. Biomol. Chem. 2012, 10, 6537e6546.

4932
Z.-Q. Liu et al. / Tetrahedron 69 (2013) 4922e4932
14. Mullins, S. T.; Annan, N. K.; Cook, P. R.; Lowe, G. Biochemistry 1992, 31, 842e849.
15. Yamashita, K.; Tanaka, T.; Hayashi, M. Tetrahedron 2005, 61, 7981e7985.
16. (a) 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.; Gar-
estier, T.; Helene, C. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 3062e3067; (b) De
Cian, A.; Guittat, L.; Kaiser, M.; Sacca, B.; Amrane, S.; Bourdoncle, A.; Al-
berti, P.; Teulade-Fichou, M. P.; Lacroix, L.; Mergny, J. L. Methods 2007, 42,
183e195.
17. (a) Shirude, P. S.; Gillies, E. R.; Ladame, S.; Godde, F.; Shin-ya, K.; Huc, I.; Bala-
subramanian, S. J. Am. Chem. Soc. 2007, 129, 11890e11891; (b) Mueller, S.;
Pantos, G. D.; Rodriguez, R.; Balasubramanian, S. Chem. Commun. 2009, 80e82.
18. Monchaud, D.; Allain, C.; Bertrand, H.; Smargiasso, N.; Rosu, F.; Gabelica, V.; De
Cian, A.; Mergny, J. L.; Teulade-Fichou, M. P. Biochimie 2008, 90, 1207e1223.
19. Monchaud, D.; Allain, C.; Teulade-Fichou, M. P. Bioorg. Med. Chem. Lett. 2006, 16,
4842e4845.
20. (a) Paramasivan, S.; RuJan, I.; Bolton, P. H. Methods 2007, 43, 324e331; (b)
White, E. W.; Tanious, F.; Ismail, M. A.; Reszka, A. P.; Neidle, S.; Boykin, D. W.;
Wilson, W. D. Biophys. Chem. 2007, 126, 140e153.
21. (a) Tan, J.-H.; Ou, T.-M.; Hou, J.-Q.; Lu, Y.-J.; Huang, S.-L.; Luo, H.-B.; Wu, J.-Y.;
Huang, Z.-S.; Wong, K.-Y.; Gu, L.-Q. J. Med. Chem. 2009, 52, 2825e2835; (b)
Moorhouse, A. D.; Santos, A. M.; Gunaratnam, M.; Moore, M.; Neidle, S.; Moses,
J. E. J. Am. Chem. Soc. 2006, 128, 15972e15973.
22. (a) Dash, J.; Waller, Z. A. E.; Pantos, G. D.; Balasubramanian, S. Chem.dEur. J.
2011, 17, 4571e4581; (b) Kaluzhny, D.; IIyinsky, N.; Shchekotikhin, A.; Sinke-
vich, Y.; Tsvetkov, P. O.; Tsvetkov, V.; Veselovsky, A.; Livshits, M.; Borisova, O.;
Shtil, A.; Shchyolkina, A. PloS One 2011, 6 e271511e6.
23. Ou, T.-M.; Lu, Y.-J.; Tan, J.-H.; Huang, Z.-S.; Wong, K.-Y.; Gu, L.-Q. ChemMedChem
2008, 3, 690e713.
24. (a) Wang, Y.; Patel, D. J. Structure 1993, 1, 263e282; (b) Rezler, E. M.; Seenisamy,
J.; Bashyam, S.; Kim, M. Y.; White, E.; Wilson, W. D.; Hurley, L. H. J. Am. Chem.
Soc. 2005, 127, 9439e9447.
25. (a) Rachwal, P. A.; Fox, K. R. Methods 2007, 43, 291e301; (b) Lane, A. N.; Chaires,
J. B.; Gray, R. D.; Trent, J. O. Nucleic Acids Res. 2008, 36, 5482e5515.
26. Dash, J.; Shirude, P. S.; Hsu, S.-T. D.; Balasubramanian, S. J. Am. Chem. Soc. 2008,
130, 15950e15956.
27. (a) Eswaran, S.; Adhikari, A. V.; Chowdhury, I. H.; Pal, N. K.; Thomasa, K. D. Eur. J.
Med. Chem. 2010, 45, 3374e3383; (b) Ong, C. W.; Liu, M.-C.; Lee, K.-D.; Chang,
K. W.; Yang, Y.-T.; Tung, H.-W.; Fox, K. R. Tetrahedron 2012, 68, 5453e5457.
28. Lu, Y.-J.; Ou, T.-M.; Tan, J. H.; Hou, J.-Q.; Shao, W.-Y.; Peng, D.; Sun, N.; Wang, X.-
D.; Wu, W.-B.; Bu, X.-Z.; Huang, Z.-S.; Ma, D.-L.; Wong, K.-Y.; Gu, L.-Q. J. Med.
Chem. 2008, 51, 6381e6392.
29. Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart, W. E.; Belew, R. K.;
Olson, A. J. J. Comput. Chem. 1998, 19, 1639e1662.
30. Sanner, M. F. J. Mol. Graph. Model. 1999, 17, 57e61.