Dinuclear ruthenium(II) complexes that induce and stabilise G-quadruplex DNA

Article abstract of DOI:10.1002/chem.201405991

A series of dinuclear ruthenium(II) complexes were synthesised, and the complexes were determined to be new highly selective compounds for binding to telomeric G-quadruplex DNA. The interactions of these complexes with telomeric G-quadruplex DNA were stud

Full text of DOI:10.1002/chem.201405991

DOI: 10.1002/chem.201405991
Full Paper
&
G-Quadruplexes
Dinuclear Ruthenium(II) Complexes That Induce and Stabilise G-
Quadruplex DNA
Li Xu,[a, b] Xiang Chen,[a] Jingheng Wu,[a] Jinquan Wang,[a] Liangnian Ji,[a] and Hui Chao*[a]
Abstract: A series of dinuclear ruthenium(II) complexes were
synthesised, and the complexes were determined to be new
highly selective compounds for binding to telomeric G-
quadruplex DNA. The interactions of these complexes with
telomeric G-quadruplex DNA were studied by using circular
dichroism (CD) spectroscopy, fluorescence resonance energy
transfer (FRET) melting assays, isothermal titration calorime-
try (ITC) and molecular modelling. The results showed that
the complexes 1, 2 and 4 induced and stabilised the forma-
tion of antiparallel G-quadruplexes of telomeric DNA in the
absence of salt or in the presence of 100 mm K+-containing
buffer. Furthermore, complexes 1 and 2 strongly bind to and
effectively stabilise the telomeric G-quadruplex structure and
have significant selectivity for G-quadruplex over duplex
DNA. In comparison, complex 3 had a much lesser effect on
the G-quadruplex, suggesting that possession of a suitably
sized plane for good p–p stacking with the G-quadruplets is
essential for the interaction of the dinuclear ruthenium(II)
complexes with the G-quadruplex. Moreover, telomerase
inhibition by the four complexes and their cellular effects
were studied, and complex 1 was determined to be the
most promising inhibitor of both telomerase and HeLa cell
proliferation.
Introduction
Ligand interaction with duplex DNA leads to acute toxicity and
drastic side effects in normal tissues.[14] In addition, G-quadru-
plexes are rich in sequence-dependent structural polymorph-
isms.[15] Different guanine-rich sequences are well known to
fold into different quadruplex conformations in numerous
ways. Even the same sequence can fold differently depending
on many factors related to the surrounding solution, such as
the counterions present (K+ and Na+).[16,17] Although telomere
sequences can exhibit a range of topologies, conformational
selectivity has been only a few considered in almost all of the
ligands or complexes recently reported to demonstrate good
G-quadruplex affinity and stabilisation ability.[18,19] Furthermore,
other authors have suggested that G-quadruplex stabilisers
should also be designed to selectively target antiparallel and
hybrid-type G-quadruplex folding topologies because the
parallel G-quadruplex structure of vertebrate telomeric repeat
sequences is not the preferred folding topology under
physiological conditions.[17,20] Thus, the improvement of
G-quadruplex selectivity is necessary because ligands that bind
to specific conformations of human telomeric DNA can lead to
the development of disease-specific therapeutic effects and to
their utilisation as personal drugs.
The DNA G-rich single strand can fold into higher-order
structures called G-quadruplexes under appropriate cationic
conditions.[1–3] Quadruplexes are present in many biologically
significant genomic regions, including telomeres,[4] immuno-
globulin switch regions,[5] mutational hot spots and regulatory
elements in oncogene promoters,[6–8] and can modulate several
biological key processes. For example, G-quadruplexes present
in telomeric DNA have been suggested to inhibit telomerase
activity (an enzyme that is over-expressed in 85–90% of cancer
cells) and to interfere with the biological function of telo-
mere.[9] The G-quadruplexes in many oncogene promoters
function as a transcriptional repressor element. Therefore,
small molecules that can promote G-quadruplex formation or
stabilisation have become an attractive approach towards anti-
cancer drug discovery[.[10–13] However, these G-quadruplex sta-
bilisers show relatively weak selectivity over duplex DNA.
[a] Dr. L. Xu,+ X. Chen,+ J. Wu, J. Wang, Prof. L. Ji, Prof. Dr. H. Chao
MOE Key Laboratory of Bioinorganic
and Synthetic Chemistry
School of Chemistry and Chemical Engineering
Sun Yat-Sen University
Guangzhou 510275 (P.R. China)
Ruthenium(II) complexes have been reported to be
G-quadruplex stabilisers[21–25] and probes.[26–29] Ruthenium(II)
complexes, especially dinuclear ruthenium(II) complexes, offer
substantial advantages over organic compounds,[30–35] other
metal complexes (Pt, Ni, etc.)[36–40] and even mononuclear
ruthenium(II) complexes,[21,25,28,29] including greater variations
in shape, charge and size. By connecting two metal centres
into a dinuclear complex, selectivity for G-quadruplex struc-
tures can be achieved and the drug tolerance of tumour cells
E-mail: ceschh@mail.sysu.edu.cn
[b] Dr. L. Xu+
School of Chemistry and Chemical Engineering
Guangdong Pharmaceutical University
Zhongshan, 528458 (China)
[+] These authors contributed equally to this work.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201405991.
Chem. Eur. J. 2015, 21, 1 – 14
1
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
to platinum anticancer drugs can be reduced. In addition, two
metals are preferred over one because metals incorporated as
a structural locus not only restrict the geometry of the planar
aromatic core but also serve as a binding element for the sub-
stituents of the core.[39] Several recent notable reports have
shown that dinuclear ruthenium(II) complexes can induce the
formation of and stabilise G-quadruplexes.[21,25] With all of the
important influencing factors in mind, we recently synthesised
four dinuclear ruthenium(II) complexes (Scheme 1). Surprising-
ly, some of them can efficiently induce the formation and sta-
bilisation of antiparallel G-quadruplex structures of the human
telomeric sequence (AG3(T2AG3)3, HTG22) and can even lead to
the conversion of a hybrid G-quadruplex to an antiparallel
G-quadruplex. These complexes can also selectively bind and
stabilise G-quadruplexes in the presence of excessive duplex
DNA. Importantly, the complexes can significantly inhibit
telomerase activity and HeLa cell proliferation.
All of these ruthenium(II) complexes were purified by column
chromatography and were characterised by elemental analysis,
1H NMR spectroscopy and ES-MS (see Figures S1–S4 in the
Supporting Information). Meanwhile, the four mononuclear
complexes were also synthesised (see Figures S5–S12 in the
Supporting Information). Because of the very poor solubility of
mononuclear ruthenium(II) complexes in aqueous solution, the
following experiments were tested only with the dinuclear
ruthenium(II) complexes.
Circular dichroism measurements
Circular dichroism (CD) spectroscopy is a conventional method
for determining the presence of a G-quadruplex structure and
its different folding structures and the effect of complex bind-
ing on the quadruplex structures.[41] The interaction of the
complexes with telomeric DNA was studied by using this
method. Human telomeric DNA can exist as a mixture of anti-
parallel and parallel G-quadruplex conformations. Previous
studies have shown that some G-quadruplex DNA binders can
interact with both conformations.[41–43] The CD spectrum of an
Results and discussion
Synthesis and characterisation
antiparallel-stranded G-quadruplex exhibits
a characteristic
Complexes 1–4 (Scheme 1) were prepared by the direct
reaction of the ebipcH2 (2,2’-(9-ethyl-9H-carbazole-3,6-diyl)bis-
(1H-imidazo[4,5-f][1,10]phenanthroline)), mbpibH2 (1,3-bis(1H-
imidazo[4,5-f][1,10]phenanthrolin-2-yl)benzene), TpipibH2 (N-(4-
(1H-imidazo[4,5-f][1,10]phenanthrolin-2-yl)phenyl)-4-(1H-imida-
zo[4,5-f][1,10]phenanthrolin-2-yl)-N-phenylaniline) or hbpibH2
(2,6-bis(1H-imidazo[4,5-f][1,10]phenanthrolin-2-yl)-4-methylphe-
nol) ligand with the appropriate molar ratio of the precursor
complex, cis-[Ru(bpy)2Cl2]·2H2O (bpy=bipyridine), in ethylene
glycol; the complexes were obtained in relatively high yields.
positive peak at l=295 nm, a smaller negative peak at l=
265 nm and a smaller positive peak at l=245 nm.[44,45] Howev-
er, the CD spectrum of a parallel-stranded G-quadruplex exhib-
its a positive peak at l=260 nm and a small negative peak at
l=240 nm. In the absence of salt, the CD spectrum of the
HTG22 oligonucleotide was partially dissociated to single-
stranded molecules with a small negative band centred at l=
240 nm, a major positive band at l=256 nm, a minor negative
band at l=270 nm and a positive band near l=295 nm (Fig-
ure 1a, black line). Upon the addition of an excess of complex
Scheme 1. Chemical structures of the ruthenium(II) complexes 1–4.
&
&
Chem. Eur. J. 2015, 21, 1 – 14
www.chemeurj.org
2
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Figure 1. CD spectra of HTG22 (10 mm) induced after the addition of the ruthenium(II) complexes a) 1, b) 2, c) 3 and d) 4 at 258C in a Tris-HCl buffer (10 mm,
pH 7.4).
1 (5 mol equiv) to the HTG22, the CD spectrum changed sub-
stantially. The intensity of the band centred at l=295 nm de-
creased substantially, whereas the small negative band at l=
240 nm and the positive band at l=256 nm disappeared,
leading to the appearance of a positive band at l=245 nm
and a major negative band at l=265 nm (Figure 1a). These
changes suggest that complex 1 can induce single-stranded
guanine-rich HTG22 DNA to form an antiparallel G-quadruplex
structure. The CD spectra of the HTG22 oligonucleotide after
the addition of complexes 2 and 4 in the absence of salt were
similar to that after the addition of complex 1 (Figures 1b and
d). However, the band intensity in the CD spectrum after the
addition of complex 2 was weaker than in the spectra after the
addition of complexes 1 and 4. This result indicated that com-
plex 2 was less effective at inducing G-quadruplex formation
than with complexes 1 and 4. Furthermore, upon addition of
complex 3 to the HTG22 oligonucleotide, a new and still
unclear conformational change was observed in the CD
spectrum (Figure 1c).
oligonucleotide formed the hybrid-type G-quadruplex struc-
ture with a large positive band at l=290 nm, a shoulder at ap-
proximately l=270 nm, a small positive band at l=250 nm
and a negative band at l=235 nm (see Figure S13 in the Sup-
porting Information, black line).[46] When complex 1 was titrat-
ed into the described solution, the CD spectrum significantly
changed: the maximum band at l=292 nm increased in inten-
sity and the shoulder at l=270 nm gradually disappeared. In
addition, the small positive band at l=250 nm gradually dis-
appeared and led to the appearance of a positive band at l=
245 nm and
a
major negative band at l=260 nm
(see Figure S13A in the Supporting Information), indicating
that complex 1 can induce the formation of an antiparallel
G-quadruplex structure from the hybrid-type G-quadruplex
structure (see Figure S13A in the Supporting Information).
Similar changes in the CD spectrum were also induced by the
addition of complexes 2–4 (see Figures S13B–D in the
Supporting Information).
Second, we tested the effect of the complexes on the con-
formation of the HTG22 oligonucleotide formed in a 100 mm
NaCl buffer. As shown in Figure S14 in the Supporting Informa-
tion, in the presence of Na+ ions and the absence of the ruth-
enium(II) complexes, the human telomeric sequence adopted
an antiparallel conformation.[47] The addition of increasing
CD experiments were also measured in solutions containing
K+ and Na+ to determine whether the compounds could lead
to either structural conversion or stabilisation. First, CD
titration experiments were performed for the HTG22 oligo-
nucleotide in the presence of 100 mm K+ buffer. The HTG22
Chem. Eur. J. 2015, 21, 1 – 14
www.chemeurj.org
3
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
amounts of complexes 1, 3 and 4 to the HTG22 oligonucleo-
tide in a 100 mm Na+ buffer resulted in no significant changes
in the CD spectrum because the G-quadruplex was strongly
stabilised by Na+ (see Figures S14A, C and D in the Supporting
Information). However, the addition of complex 2 resulted in
substantial changes, including an increase in the intensity of
the positive band at l=295 nm and the disappearance of the
small positive band at l=240 nm, which led to the appear-
ance of a positive band at l=245 nm and a shoulder band at
l=270 nm (see Figure S14B in the Supporting Information).
These results suggest that complex 2 had the ability to induce
the anti-parallel structure of the HTG22 oligonucleotide to
form a hybrid-type G-quadruplex structure in the presence of
Na+ ions, whereas complexes 1, 3 and 4 did not. Overall, the
CD experiments clearly demonstrated that complexes 1 and 4
can induce the formation of antiparallel G-quadruplexes with
or without the addition of a stabilising salt.
Furthermore, the chiral interferences between the chiral
structures of the ruthenium(II)
Stabilising ability and selectivity studies with Fçrster
resonance energy transfer (FRET)
To investigate the stabilisation effect and the selectivity of the
Ru complexes for telomeric G-quadruplex DNA, a FRET melt-
ing-point assay was performed to detect the melting tempera-
ture. The human telomeric G-quadruplex DNA (F22T) contain-
ing fluorophores at both the 5’-end and the 3’-end was used
in this assay. The DTm values of the G-quadruplex DNA treated
with 1.0 mm solutions of the complexes are shown in Table S1
in the Supporting Information. As shown in Table S1 in the
Supporting Information, among all of the complexes, complex
2 had the most distinct ability to stabilise the G-quadruplex
DNA, with DTm values of 248C in a Na+ buffer and of 188C in
a K+ buffer. The DTm values for complex 1 were similar at ap-
proximately 208C in a Na+ buffer and of 128C in a K+ buffer,
whereas complexes 3 and 4 showed relatively lower DTm
values. The FRET melting data demonstrated that complexes
complexes interacting with a G-
quadruplex structure have also
been considered. We used
thermo slide-A-lyzer mini dialysis
units with 3500 MWCO to dialy-
sis the complex–DNA mixed so-
lution. Because the molecular
weight of the quadruplex–RuII
complex composite structures is
larger than 3500, only those
chiral ruthenium(II) complexes,
which did not bind to the DNA
can dialysis out of the dialysis
unit. If the complexes have chiral
selectivity towards the G-quad-
ruplex structure, this will gener-
ate a change of the CD signals,
which remained in the dialysis
unit. However, no significant CD
signal change was observed
after 24 h dialysis compared
with the samples without dialy-
sis. At the same time, the dialy-
sates outside the dialysis tubes
did not exhibit any CD signals.
These results demonstrated that
there is no obvious difference of
binding strength between the
different chiral structures of the
ruthenium(II) complexes and G-
quadruplex (see Figures S15–S17
in the Supporting Information).
Figure 2. FRET melting profiles of 0.2 mm F22T with the different ruthenium complexes a) 1, b) 2, c) 3 and d) 4 in
Na+ buffer. e) Plot of DTm versus the complex concentration.
&
&
Chem. Eur. J. 2015, 21, 1 – 14
www.chemeurj.org
4
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
1 and 2 could effectively stabilise G-quadruplex DNA with
much higher DTm values compared to the complexes reported
in previous studies. The concentration-dependent melting
curves for complexes 1–4 in Na+ and K+ buffers are shown in
Figures 2 and S7 in the Supporting Information, respectively. In
the presence of Na+, the Tm value of the antiparallel G-quadru-
plex is 508C. In the presence of 1 mm of complex 2, the Tm
value increased to 748C. This value further increased to 828C
when the concentration of complex 2 was increased to 3 mm.
With the high ionic strength of Na+ (Figure 2b), the observed
DTm value of 328C is uncommonly high compared to those re-
ported in previous studies.[33,36,40] In the presence of K+, we ob-
served a similar phenomenon, but with lower DTm values. For
example, the Tm value of the hybrid G-quadruplex in the pres-
ence of K+ is 608C; however, in the presence of 1 mm of com-
plex 2, the Tm value increases by 188C (to 788C). After the ad-
dition of 3 mm of complex 2, a remarkable melting temperature
increase of 248C was observed (see Figure S18B in the Sup-
porting Information). Although the DTm values observed for
complex 1 were smaller than those for complex 2 in both Na+
(Figure 2a) and K+ (see Figure S18A in the Supporting Infor-
mation) buffers, all of the results suggest that complexes
1 and 2 are suitable G-quadruplex stabilisers. FRET melting ex-
periments were also used to examine the binding selectivity of
the complexes on a G-quadruplex over duplex DNA. The DNA
competition FRET melting assay was performed to show the
DTm change for ruthenium(II) complex concentrations of
0.5 mm, 1.0 mm and 2.0 mm with 200 nm F22T through the addi-
tion of various concentrations of the double-stranded DNA
ds26. The results of the melting-point change of 1.0 mm ruthe-
nium(II) complexes induced by the addition of ds26 in Na+ or
K+ buffers are shown in Figure 3 and Figure S19 in the Sup-
porting Information, respectively. The DTm values for complex
1, 2 and 4 did not significantly decrease, particularly in the
Na+ buffer, despite the presence of ds26 at a concentration
Figure 4. ITC data for the binding of complexes a) 1, b) 2, c) 3 and d) 4 to
the human telomeric G-quadruplex. The data were obtained in a buffer that
consisted of 100 mm NaCl, 10 mm NaH2PO4/Na2HPO4 and 1 mm Na2EDTA
(EDTA=ethylenediaminetetraacetic acid) (pH 7.0) at 258C The binding
parameters of complexes 1–4 were obtained through analysis of the ITC
profiles.
200 times that of F22T. These results indicate that the
ruthenium(II) complexes can specifically stabilise the G-
quadruplex structure with affecting duplex DNA.
Binding affinity studies with isothermal titration calorimetry
(ITC)
To assess the binding affinity and thermodynamic properties
of the interaction between the complexes and the
G-quadruplexes, we used ITC to determine the chemical ther-
modynamics, binding constants (Kb) and number of binding
sites (n). Representative ITC curves for the binding of com-
plexes 1–4 to the HTG22 G-quadruplex DNA at 298 K in a Na+-
containing solution are shown in Figure 4. The thermodynamic
parameters for the binding of the four complexes to quadru-
plex DNA are summarised in Table S2 in the Supporting Infor-
mation. The binding constants of complexes 1, 2 and 4 are ap-
proximately twice that of complex 3. These data demonstrate
the importance of the intercalating chromophores because
complexes 1–4 have identical ancillary ligands.
Figure 3. DTm change of F22T under different concentrations of double-
stranded DNA ds26 in Na+ buffer treated with 1 mm of the different Ru
complexes.
The human telomeric G-quadruplex binding stoichiometry
with quadruplex DNA was investigated through luminescence-
Chem. Eur. J. 2015, 21, 1 – 14
www.chemeurj.org
5
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
based Job plots (see Figure S20 in the Supporting Information).
Three major inflection points for three complexes were ob-
served at x=0.52 for complex 1, x=0.57 for complex 2 and
x=0.46 for complex 3. These data are consistent with a 1:1
[quadruplex]/[complex] binding mode. However, the intersec-
tion of the fit lines in the Job plot for human telomeric G-
quadruplex with complex 4 is approximately 0.69. This value
corresponds to a [complex]/[G-quadruplex] stoichiometric ratio
of 2:1. These results are consistent with the results obtained
from the ITC study.
Complexes that exhibit better inhibition of the amplification in
HTG22 generally possess a greater G-quadruplex structure-sta-
bilising ability. Meanwhile, the binding selectivity between the
HTG22 G-quadruplex and other representative G-quadruplexes
in gene promoters, such as c-myc, bcl-2 and c-kit2, has also
been tested by using the PCR-stop assay. Experimental results
demonstrate that none of these four complexes can
significantly inhibit the PCR amplification by interacting with
the c-myc, bcl-2 or c-kit2 gene promoters sequence at 14 mm
whereas the highest concentration that can inhibit the ampli-
fication in the HTG22 sequence is only 12 mm (see Figure S21
in the Supporting Information).
Polymerase chain reaction (PCR)-stop assay
To demonstrate that the ruthenium(II) complexes can induce
G-quadruplex formation from telomeric DNA, a PCR-stop assay
was performed. The PCR-stop assay is an effective method to
ascertain whether the complexes bind to the test oligomer,
that is, HTG22, and induce its transformation into a G-quadru-
plex structure.[48] In the lane in which G-quadruplex stabilisers
were absent, sequences of HTG22 and its corresponding com-
plementary sequence (HTG22rev) will, under the function of
Taq DNA polymerase, hybridise to a final double-stranded DNA
PCR product that can be observed as a clear band in both
lanes with 0 mm ruthenium(II) complexes. However, in the
presence of the four ruthenium(II) complexes, the template se-
quence HTG22 formed a G-quadruplex structure induced by
the ruthenium complexes that the Taq DNA polymerase
cannot elongate into a double-stranded structure; thus, the
PCR reaction is blocked. The greater the concentration of the
ruthenium(II) complex is, the smaller is the amount of final
product that can be detected.[19] As the ruthenium(II) complex
binds to all of the template sequence HTG22, the band of the
PCR product will completely disappear. Figure 5 illustrates the
disappearance of the final PCR product under increasing
concentrations of the complexes.
Telomere repeat amplification protocol (TRAP) assay
The complexes synthesised herein were designed to restrain
the activity of telomerase. The TRAP assay is a general and reli-
able method for evaluating the inhibitory properties of small
molecules against telomerase in vitro. In previous studies, com-
plexes with a high binding affinity and stabilising ability for
telomeric G-quadruplexes were suggested to exhibit inhibitory
effects on telomerase.[49] After the complex binds to the telo-
meric sequence and induces its structural transformation into
a G-quadruplex, the geometrical conformation of the telomeric
DNA is altered such that its telomerase binding site is lost. The
immediate consequence is telomerase inactivity. The intensity
of the DNA ladder will reflect the activity of telomerase. To
make the experiment more rigorous, we performed a positive
control by using the same process used for the other test sam-
ples but without the addition of any complexes. In addition,
a negative control was performed by using the same process
used for the positive control but without the addition of cell
extracts. These controls ensure that the DNA ladder is generat-
ed by telomerase. The internal control (IC) bands were gener-
ated by the sequence that cannot transform into a G-quadru-
plex; thus, their steady appearance will help to ensure that
other factors do not block the PCR during the process. As
shown in Figure 6, the number of bands clearly decreased with
increasing concentrations of the complexes, showing that the
in vitro inhibitory effects of the four complexes toward telo-
merase are concentration dependent. Complexes 1, 2 and 4
exhibit a remarkably high inhibitory effect on telomerase at
only nanomolar levels. Complex 1, in particular, causes the
DNA ladder to completely disappear at an encouraging con-
centration of approximately 200 nm. Although complex 3 ex-
hibits less inhibitory activity toward telomerase than the other
three complexes, its inhibition concentration of 2 mm still
makes it one of the best telomerase inhibitors. Compared with
previously reported G-quadruplex ligands,[24,50] all four
complexes synthesised herein exhibit an extraordinarily high
inhibition activity of telomerase in vitro, which indicates that
they are promising potential human telomerase inhibitors.
Figure 5. Effects of ruthenium(II) complexes 1–4 on the hybridisation of
HTG22 in the PCR-stop assay.
The inhibitory effects of all four of the complexes are con-
centration dependent. Complex 1 exhibited a better G-quadru-
plex stabilisation activity with complete inhibition of HTG22 at
a complex concentration of approximately 10 mm. The same
effect of complexes 2 and 4 was observed at approximately
12 mm, whereas the band was still slightly detected in the case
of complex 3, even at a complex concentration of 12 mm.
Short-term cell viability
All of the above results indicate that the four complexes can
induce the formation of telomere sequence DNA into anti-par-
&
&
Chem. Eur. J. 2015, 21, 1 – 14
www.chemeurj.org
6
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
liferation experiment. Figure 7
shows the cell viability (relative
to control) as a function of the
concentration of the four com-
plexes after incubation for 72 h.
All four ruthenium complexes
showed mild cytotoxicity com-
pared to cisplatin in the short-
term assay. The IC50 values of the
four complexes 1–4 are 33.1,
38.5, 54.2 and 40.8 mm, respec-
tively, whereas that for cisplatin
is (8.3Æ1.4) mm. In addition, the
four complexes at concentra-
tions less than 3.13 mm showed
no cytotoxicity toward HeLa
cells, indicating that concentra-
tions less than this value could
be used in the long-term prolif-
eration experiments.
Figure 6. Inhibition of telomerase activity by complexes 1–4. A positive control (+) was performed with telomer-
ase but without ligand. A negative control (À) was performed without either telomerase or ligand. The ladder
shows the product of telomerase elongation. The lower band is an internal control primer (IC).
Long-term proliferation experiments
As a consequence of telomerase inhibition, telomere sequence
elongation will be blocked. A large amount of evidence sug-
gests that a dysfunctional telomere could activate p53 to ini-
tiate cellular senescence or apoptosis.[52,53] Because telomere
shortening accompanies cell division, this effect will only be
observed after several cycles of cell division. Thus, long-term
experiments must be performed to eliminate the effects of
acute cytotoxicity or other nonspecific events. Therefore, sub-
cytotoxic concentrations (2.5 mm, far below the IC50 value) of
the complexes were evaluated in HeLa cells during long-term
exposure. Figure 8 shows the change in the CI, representing
cell growth. At the very beginning, cells that were and were
not treated with complexes could not be readily distinguished,
indicating that the complexes had no obvious short-term cyto-
toxic effects. After three days, all of the cells exposed to the
complexes were delayed in entering the log phase and the cell
number slowly decreased, indicating that the cell activity was
reduced. The growth condition of the cells exposed to com-
plexes 1 and 2 was inferior to that of the cells exposed to
complexes 3 and 4. All of these results indicate that the cells
were led to apoptosis in the long-term experiments by the
four complexes. Furthermore, the complexes showing stronger
telomerase inhibition ability in the TRAP assay also exhibited
more efficient cell inhibition in the long-term proliferation ex-
periments. Combined with the TRAP assay results, we believe
that the long-term cell death caused by the four complexes is
a result of the inhibition of telomerase, which leads to telo-
mere-length shortening. Previously reported efficient telomeric
G-quadruplex ligands and telomerase inhibitors exhibiting the
same experimental results[54] and phenomena support the find-
ings that these four complexes are significant telomerase in-
hibitors.
Figure 7. Effect of the ruthenium(II) complexes 1–4 on HeLa cell viability.
HeLa cells were treated with various concentrations of the ruthenium(II)
complexes (3.13–100 mm) for 72 h.
allel structures with high selectivity, leading to high telomerase
inhibition in the cell-free system. These results thus encour-
aged us to discover the function of these four complexes in
the cell. Inhibition of telomerase is well known to shorten the
telomere. Thus, cell apoptosis will not immediately occur; the
cells will enter programmed cell death until the telomere is
reduced to a certain point after several cell divisions.[51] On the
basis of these considerations, complexes that can become
potential telomerase inhibitors must not kill cells immediately.
Therefore, short-term cell viability assays will provide a specific
concentration that will not influence cell growth with an acute
toxicity, which is necessary for the subsequent long-term pro-
Chem. Eur. J. 2015, 21, 1 – 14
www.chemeurj.org
7
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
binding mode can easily be
found in the binding of K+-as-
sembled HTG DNA for the base
pairs that are exposed to the sol-
vent. Meanwhile, for an ionic so-
lution containing Na+ ions, the
ruthenium(II) complex is stacked
on the G-quadruplex and its
phenanthroline nitrogen atoms
overlap the negative electrostat-
ic potential provided by the po-
larised
carbonyl
centre
(Figure 9). It clearly shows the
importance of p–p interactions,
where the central aromatic rings
in the ruthenium(II) complex
stack onto the surface of the
guanines. To further investigate
these possible binding modes
and to achieve more reliable in-
formation, molecular dynamics
(MD) simulations and MM/GBSA
calculations were carried out. As
summarised in Table 1, similar
binding free energies (DGbind
)
values were obtained for exter-
nal stacking. Relatively, due to
the larger contact surfaces,
stronger ligand/DNA interactions
(lower DGbind values) were ob-
served for the intercalation bind-
ing mode, especially in the cases
of complex 1. Results are in ex-
cellent agreement with those re-
lated to the stabilising ability of
the FRET experiments and the
thermodynamic parameters (see
Table S2 in the Supporting Infor-
Figure 8. Long-term growth curves of HeLa cells treated with complexes 1 (bright green), 2 (blue), 3 (dark green)
and 4 (cyan). The red curve represents control cells not exposed to the complexes.
mation) obtained by using ITC, which indicated that complex
1 exhibits remarkable stabilising abilities, whereas complex 3
shows the worst. Furthermore, the last MD snapshot presented
in Figure 11 clearly shows that complexes 1, 2 and 4 stabilises
Molecular docking and MM/GBSA calculations
Molecular docking experiments were performed to elucidate
different interactions between the ligands and HTG DNA.[55]
The three forms of the complexes, that is, the (D,D), (D,L) and
(L,L) form were introduced in the docking analysis. The re-
sults showed that these compounds share very similar binding
sites (see Figures S22 and S23 in the Supporting Information).
For complex 4, which binds with the DNA as a ratio of 2:1, we
presented a binding model including nonspecific binding to-
gether with intercalation presented in Figure 9 (by using the
(L,L) form, the same as follows) because there is no evident
supporting that covalent bonding is involved between com-
plex 4 and the G-quadruplex. When the nonspecific binding
was formed, the complexes inserted into one of the DNA
grooves with the planar aromatic ring groups spanning the nu-
cleotides (Figure 9). The intercalation between the ligands and
the HTG DNA in an ionic solution containing K+ ions is shown
in Figure 10. As shown in Figure 10, The external-stacking
Table 1. Calculated binding free energies and their components for the
binding of the dinuclear ruthenium(II) complexes to Na+- (PDB 143D)/K+
(PDB 1KF1)-assembled human telomeric G-quadruplex DNA (in [kcal
molÀ1]).
PDB
Binding mode Complex DGele DEvdW
DGsolv-np DGbind
1
2
3
4
1
2
3
4
31.30 À107.35 À8.93 À84.98
30.12 À102.87 À8.97
16.06
À69.96 À6.89
À81.72
À60.79
À84.95
À50.93
À56.95
À50.06
À48.03
143D intercalation
35.21 À111.08 À9.08
7.10
9.05
5.78
5.99
À52.60 À5.43
À60.01 À6.01
À50.59 À5.26
À48.90 À5.12
external
1KF1
stacking
&
&
Chem. Eur. J. 2015, 21, 1 – 14
www.chemeurj.org
8
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Conclusion
In summary, four ruthenium(II) complexes have been
synthesised and characterised. The binding behaviours of
the ruthenium(II) complexes towards G-quadruplex DNA were
examined by CD spectroscopy and FRET melting assays. These
experiments showed that complexes 1, 2 and 4 can induce
high stabilisation of the human telomeric G-quadruplex and
possess an extraordinarily high selectivity for the G-quadruplex
versus duplex DNA. To examine the abilities of the complexes
to bind strongly to telomeric quadruplex DNA and to inhibit
cancer proliferation, we also performed TRAP assay and in
vitro cytotoxicity studies of the four complexes. The results in-
dicated that complex 1 is a potent inhibitor of telomerase and
HeLa cell proliferation. The results also indicated that a dinu-
clear complex has a higher solubility in aqueous solution and
that its geometry structure leads to a better selectivity to G-
quadruplex over duplex DNA than the traditional mononuclear
complex. In addition, we also achieve some certain results of
the structure–function relationship between the main ligands
of the metal complex and the G-quadruplex DNA. These results
will be valuable for designing more potent and selective telo-
meric G-quadruplex-interactive compounds.
Figure 9. External stacking binding mode of complex 4 and the Na+-
assembled HTG DNA (PDB ID: 143D).
Experimental Section
Materials: The human telomeric
sequence d[(TTAGGG)n] (HTG22)
DNA,
DNA
F22T
(5’-FAM-
d(AG3[TTAGGG]3)-TAMRA-3’)
(FAM06-carboxy
fluorescein,
TAMRA=6-carboxytetramethyl-
rhodamine), DNA ds26 (5’-CAATCG-
GATCGAATTCGAT-CCGATTG-3’), the
oligomers
HTG22
(5’-AGGGT-
TAGGGTTAGGGTTAGGG) and the
corresponding complementary se-
quences
ATACGCTTCTCGTCCCTAACCC), the
oligomers c-myc (5’-
TGGGGAGGGTGGGGAGGGTGGG-
GAAGG) and the corresponding
complementary sequences Rev c-
myc (5’-GATCTTCTTCGTCCTTCCC-
CA), the oligomers bcl-2 (5’-
CGGGCGCGGGAGGAAGGGGGCGG-
HTG22rev
(5’-
Figure 10. External stacking binding mode of complex 1 and the K+-assembled HTG DNA (PDB ID: 1 KF1).
a) Platform and b) side elevation, red stands for negative-charged regions, blue stands for positive-charged
regions.
the DNA strands by spanning over the G-quadruplex. Even
GAGC) and the corresponding complementary sequences Rev bcl-2
(5’-ATCGATCGCTTCT-CGTGCTCCCGCCC), the oligomers c-kit2 (5’-
CGGGCGGGCGCGAGGG-AGGGG) and the corresponding comple-
mentary sequences Rev c-kit2 (5’-TATATATATACCCCTCCCT) were ob-
tained from Sangon Biological Engineering Technology & Services
Co., Ltd. (Shanghai, China). Acrylamide (molecular biology grade)
and ethidium bromide were purchased from Sigma (USA). HeLa
(cervical) cell lines were purchased from American Type Culture
Collection. Other materials were commercially available and of re-
agent grade. Doubly distilled or DEPC-treated water (DEPC=dieth-
ylpyrocarbonate) were used to prepare the buffers. A solution of
HTG22 in the buffer gave a ratio of UV absorbance at l=260 and
280 nm of approximately (1.8–1.9):1, indicating that the DNA was
sufficiently free of protein.[53] The DNA concentration per nucleo-
without the central ions, the conformations of the planar G-
quadruplets did not change obviously in the stabilisation of
these two ligands. However, the DNA chain bound to the less
effective complex 3 was not stabilised, indicating that the G-
quadruplets did not maintain their original planar conforma-
tion and the bases scattered after 4.2 ns of MD equilibration.
According to the component analysis of the DGbind values
(Table 1), the primary contributors to the binding free energies
were van der Waals energies. Therefore, the lower stabilising
ability of complex 3 can be explained by its non-planar confor-
mation, which was disadvantageous toward the stacking of
complex 3 on the planar G-quadruplets.
Chem. Eur. J. 2015, 21, 1 – 14
www.chemeurj.org
9
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
2-hydroxy-5-methylisophthalaldehydein was substituted
for 4,4’-(phenylazanediyl)-dibenzaldehyde. Yield: 0.49 g,
60%; FAB-MS: m/z: 545 [M+1]; elemental analysis calcd
(%) for C33H20N8O: C 72.78, H 3.70, N 20.58; found: C
72.75, H 3.72, N 20.62.
Synthesis of [(bpy)2Ru(TpipibH2)Ru(bpy)2](ClO4)4 (3): A
mixture of cis-[Ru(bpy)2Cl2]·2H2O (0.31 g, 0.58 mmol) and
TpipibH2 (0.20 g, 0.29 mmol) in ethylene glycol (40 cm3)
was heated to reflux under argon for 12 h to give a clear
red solution. After the solution was cooled, a red precipi-
tate was obtained by dropwise addition of a saturated
aqueous solution of NaClO4. The crude product was puri-
fied by column chromatography on neutral alumina with
acetonitrile as the eluent. The red band was collected
and the solvent was removed under reduced pressure to
1
give a red powder. Yield: 0.39 g, 70%; H NMR (500 MHz,
[D6]DMSO): d=10.07 (d, J=8 Hz, 2H), 9.03 (d, J=8 Hz,
2H), 8.89 (d, J=8 Hz, 2H), 8.85 (d, J=8 Hz, 4H), 8.82 (d,
J=8 Hz, 4H), 8.40 (d, J=7 Hz, 2H), 8.30 (d, J=8 Hz, 2H),
8.25 (t, J=8 Hz, 2H), 8.21 (t, J=5.5 Hz, 4H), 8.15 (t, J=
5 Hz, 4H), 8.11 (t, J=3 Hz, 2H), 7.92 (d, J=8 Hz, 2H),
7.86 (d, J=5 Hz, 4H), 7.84–7.79 (m, 4H), 7.65–7.56 (m,
7H), 7.45–7.35 (m, 6H), 7.27–7.18 ppm (m, 4H); ES-MS
Figure 11. The last snapshots of 4.2 ns of MD equilibration for the Na+-assembled HTG
DNA and the dinuclear ruthenium(II) complexes a) 1, b) 2, c) 3 and d) 4.
(CH3CN):
m/z:
803.4
535.8
[MÀ3ClO4ÀH]2+
,
753.6
502.3
[MÀ4ClO4À2H]2+
,
[MÀ3ClO4]3+
,
[MÀ4ClO4ÀH]3+, 376.8 [MÀ4ClO4]4+; elemental analysis
calcd (%) for C84H59N17Cl4O16Ru2: C 52.92, H 3.12, N 12.49; found: C
53.12, H 3.04, N 12.38.
tide was determined by absorption spectroscopy by using the
molar absorptivity (2.285ꢁ105 mol(quadruplex)À1 m3 cmÀ1) at l=
260 nm.[15] The buffers were prepared as follows: 1) 10 mm Tris-HCl,
pH 7.4; 2) 100 mm NaCl, 10 mm NaH2PO4/Na2HPO4 and 1 mm
Na2EDTA (pH 7.0) and 3) 100 mm KCl, 10 mm KH2PO4/K2HPO4 and
1 mmK2EDTA (pH 7.0).
Synthesis of [(bpy)2Ru(hbpibH2)Ru(bpy)2](ClO4)4 (4): This complex
was synthesised by a method identical to that described for the
preparation of [(bpy)2Ru(TpipibH2)Ru(bpy)2](ClO4)4, except that
hbpibH2 (0.16 g, 0.29 mmol) was used instead of TpipibH2. Yield:
1
0.33 g, 65%; H NMR (500 MHz, [D6]DMSO): d=9.78 (brs, 1H), 9.18
Physical measurements: Microanalysis (C, H and N) was performed
by using a Vario EL elemental analyser. Fast atomic bombardment
mass spectra (FAB-MS) were detected on a VG ZAB-HS spectrome-
ter in a 3-nitrobenzyl alcohol matrix. Electrospray mass spectra (ES-
(d, J=8 Hz, 4H), 8.87 (d, J=8 Hz, 4H), 8.83 (d, J=8 Hz, 4H), 8.26 (s,
2H), 8.22 (dt, J1 =J2 =8 Hz, 4H), 8.11 (dt, J1 =J2 =8 Hz, 4H), 8.00 (d,
J=5 Hz, 4H), 7.91–7.84 (m, 8H), 7.63 (d, J=5.5 Hz, 4H), 7.60 (dt,
J1 =J2 =5 Hz, 4H), 7.36 (dt, J1 =J2 =7 Hz, 4H), 2.50 ppm (s, 3H); ES-
1
MS) were recorded on an LCQ system (Finnigan MAT, USA). H NMR
MS (CH3CN) m/z: 734.1 [MÀ3ClO4ÀH]2+, 685.1 [MÀ4ClO4À2H]2+
,
spectra were recorded at room temperature on a Varian INOVA-500
spectrometer by using (CD3)2SO as the solvent and SiMe4 as an in-
ternal standard. All chemical shifts are reported relative to TMS. CD
spectra were recorded at room temperature on a JASCO J-810
spectropolarimeter in a cylindrical quartz cell with a path length of
0.2 cm.
457.6 [MÀ4ClO4ÀH]3+
,
342.4 [MÀ4ClO4]4+
; elemental analysis
calcd (%) for C73H52N16Cl4O17Ru2: C 49.56, H 2.96, N 12.67; found: C
49.43, H 3.01, N 12.56.
Circular dichroism measurements: CD experiments to investigate
the interaction of the complexes with the HTG22 G-quadruplex
were recorded on a JASCO J-810 spectropolarimeter at room tem-
perature by using a quartz cell with a path length of 0.2 cm. CD
spectra were collected from 200 to 400 nm with a scanning speed
of 500 nmminÀ1. The bandwidth was 8.53 nm, the response time
was 0.5 s and each sample was scanned three times to obtain an
average measurement. The solutions of HTG22 (10 mm) with
various concentrations of the RuII complexes (0–5 mol equiv) were
prepared in different buffers and stored at 48C overnight before
measurement.
Synthesis of ligands and the RuII complexes: The compounds
[(bpy)2Ru(ebipcH2)Ru(bpy)2](ClO4)4
(1),[56]
[(bpy)2Ru(mbpibH2)-
Ru(bpy)2](ClO4)4 (2),[57] 1,10-phenanthroline-5,6-dione,[58] cis-[Ru(b-
py)2Cl2]·2H2O[59] and 4,4’-(phenylazanediyl)-dibenzaldehyde[60] were
synthesised according to literature methods. All other reagents
were obtained commercially and used as received.
Synthesis of the TpipibH2 ligand: A mixture of 4,4’-(phenylazane-
diyl)dibenzaldehyde (0.45 g, 1.5 mmol), 1,10-phenanthroline-5,6-
dione (0.63 g, 3 mmol), ammonium acetate (4.62 g, 60 mmol) and
glacial acetic acid (50 cm3) was heated to reflux with stirring for
2 h. The cooled solution was then diluted with water and
neutralised with concentrated aqueous ammonia. The precipitate
was collected and purified by column chromatography on silica
gel (60–100 mesh) with ethanol as the eluent to give the com-
pound as a yellow powder. Yield: 0.59 g, 58%; FAB-MS: m/z: 683
[M+1]; elemental analysis calcd (%) for C44H27N9: C 77.52, H 3.99, N
18.49; found: C 77.02, H 4.08, N 18.35.
FRET assay: A FRET melting point assay was used to investigate
the ability of the RuII complexes to stabilise the G-quadruplex. The
fluorescent-labelled oligonucleotide F22T was prepared as
a 100 mm stock solution in buffers B and C and annealed at 908C
for 5 min. After being heated, the annealed sample was slowly
cooled to room temperature. Fluorescence melting curves were
measured on a Bio-Rad IQ 5 real-time PCR detection system by
using a total reaction volume of 25 mL, 500 nm of labelled oligonu-
cleotide and different concentrations of the ruthenium(II) com-
plexes. Fluorescence readings with excitation at l=470 nm and
detection at l=530 nm were taken at intervals of 18C in the tem-
Synthesis of the hbpibH2 ligand: This ligand was synthesised by
a method identical to that described for TpipibH2, except that
&
&
Chem. Eur. J. 2015, 21, 1 – 14
www.chemeurj.org
10
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
perature range of 25 to 958C. Each temperature was maintained
for 30 s before the reading was taken to ensure that the sample
had reached equilibrium. A DNA competition FRET melting assay
was performed to explore the selectivity of the RuII complexes be-
tween duplex DNA and G-quadruplex DNA. All of the conditions of
the reaction system were similar to those used in the FRET melting
point assay, except that different concentrations of duplex DNA
ds26 were added.[61,62]
incubated in a thermocycler under the following cycling condi-
tions: 948C for 2 min, followed by thirty cycles at 948C for 30 s,
588C for 30 s and 728C for 30 s. The PCR products were then sepa-
rated by electrophoresis on 12% non-denaturing polyacrylamide
gels in 1ꢁTBE and stained with ethidium bromide.
Telomere repeat amplification protocol (TRAP) assay: All of this
evidence affirms that the four RuII complexes have the extraordina-
ry ability to induce and stabilise G-quadruplex formation from telo-
mere DNA. Therefore, further investigation, such as the telomerase
inhibition ability, is critical. The TRAP assay (telomeric repeat am-
plification protocol assay) is a well-developed method used to con-
firm that the RuII complexes have the ability to inhibit telomerase
in vitro.[24] All of the experiments were performed according to pre-
viously published methods that were slightly modified to suit our
requirements. NP-40 lysis buffer containing 10 mm Tris-HCl
(pH 8.0), 1 mm MgCl2, 1 mm EDTA, 1% (vol/vol) NP-40, 0.25 mm
sodium deoxycholate, 10% (vol/vol) glycerol, 150 mm NaCl;,5 mm
b-mercaptoethanol and 0.1 mm AEBSF were used to extract whole-
cell protein from HeLa cells. In the initial step, the TRAP buffer
(20 mm Tris-HCl (pH 8.3), 1.5 mm MgCl2, 63 mm KCl, 0.05%
Tween 20 and 1.0 mm EGTA) containing 125 mm dNTPs, 200 ng of
TS primer (5’-AATCCGTCGAGCAGAGTT), 100 ng of NT primer (5’-
ATCGCTTCTCGGCCTTTT), 100 ng of ACX primer (5’-GCGCGG-
CTTACCCTTACCCTTACCCTAACC), 1 mL of 0.01ꢁ10À18 molmLÀ1 TSNT
oligonucleotide (5’-AATCCGTCGAGCAGAGTTAAAAGGCCGAGAAG-
CGAT), 2 U Taq polymerase and 200 cells of protein extract (in 1 mL
of NP-40 lysis buffer) were mixed in an RNase-free tube. Secondly,
the RuII complexes were dissolved in DEPC-treated water and pre-
pared as fresh solutions, which were added to the tubes contain-
ing the previously described mixture in an appropriate concentra-
tion. After all of the components were homogeneously mixed, the
samples were incubated at 308C for 30 min to allow the extension
of the substrate by telomerase. To amplify the extension products
by telomerase, the following procedure was performed in a thermal
cycler: 958C for 5 min to inactivate the telomerase, followed by
twenty-four cycles at 958C for 30 s, 528C for 30 s and 728C for
30 s. Afterward, 10 mL of 6ꢁloading dye were added to each TRAP
reaction mixture, and 35 mL of the mixture with the loading dye
were loaded and resolved on an 8% non-denaturing acrylamide
gel in 1ꢁTBE (100 min, 25 VcmÀ1) that was stained with ethidium
bromide.[65]
Isothermal titration calorimetry (ITC): Calorimetric experiments
were performed by using a high-sensitivity isothermal titration cal-
orimeter (VP-ITC, MicroCal, Inc., Northampton, MA, USA). All of the
solutions were thoroughly degassed before use by being stirred
under vacuum for 0.5 h in buffer B at 258C. The human telomeric
G-quadruplex DNA solutions for use in the ITC experiments were
prepared by dilution of the stock DNA solution with the dialysis
buffer. The ruthenium complexes were dissolved in buffer B. The
sample cell was loaded with 1.43 mL of G-quadruplex DNA solu-
tion, and the reference cell was loaded with doubly distilled water.
For a typical titration, 10 mL of the complex solution were injected
into a sample cell of the DNA solution at 300 s intervals with a stir-
ring speed of 370 rpm. The heat output per injection was obtained
through integration and was corrected by subtraction of the dilu-
tion heat, which was determined in parallel experiments by using
an injection of the same concentrations of complex into the buffer
solution. The corrected binding isotherms were fitted by using the
Origin 7.0 software to obtain the Kb value, the number of binding
sites (n), the enthalpy change (DH) and the entropy change (DS).
Continuous variation analysis: Continuous variation analysis was
performed according to a previously reported literature proce-
dure.[63] Stock solutions of 100 mm complex were prepared. The
human telomeric G-quadruplex DNA solution was prepared to
match the concentration of the stock solutions in buffer B. The
concentrations of both the complex and the DNA were varied,
whereas the sum of the reactant concentrations was kept constant
at 10 mm. In the sample solutions, the mole fraction, c, of the com-
plex was varied from 0 to 1.0 in ratio steps of 0.1. The fluorescence
intensities of these mixtures were measured at 258C. The fluores-
cence intensity was plotted as a function of the mole fraction of
the complex to generate a Job plot. Linear regression analysis of
the data was performed by using the Origin 7.0 software.
PCR-stop assay: The PCR-stop assay was used to investigate the
effect of amplification inhibition induced by the RuII complexes. In
the absence of a G-quadruplex stabiliser, the human telomere se-
quence (HTG22) can amplify under the use of Taq DNA polymerase
and hybridise a final double-stranded DNA PCR product with its
corresponding complementary sequence, HTG22rev. After the RuII
complexes bind to HTG22 and stabilise it by transforming it into
a more stable formation, such as a G-quadruplex structure, the
PCR reaction will be restrained by this transformation. The increase
in the concentration of G-quadruplex stabilisers will decrease the
yield of the final PCR product. The PCR-stop assay was performed
by using a modified version of a protocol reported in a previous
study.[64] Four oligomers were used in the current study: HTG22 (5’-
AGGGTTAGGGTTAGGGTTAGGG), c-myc (5’-TGGGGAGGGTGGG-
GAGGG-TGGGGAAGG), bcl-2 (5’- CGGGCGCGGGAGGAAGGG-
GGCGGGAGC), c-kit2 (5’- CGGGCGGGCGCGAGGGAGGGG) and the
corresponding complementary sequence HTG22rev (5’-ATA-
CGCTTCTCGTCCCTAACCC), Rev c-myc (5’-GATCTTCTTCGTCCTTCC-
CCA), Rev bcl-2 (5’- ATCGATCGCTTCTCGTGCT-CCCGCCC) and Rev c-
kit2 (5’- TATATATATACCCCTCCCT). The reactions were performed in
a 1ꢁPCR buffer containing 600 nmol of HTG22 and 1000 nm
HTG22rev, 0.16 mm dNTPs and 2.5 U Taq polymerase with different
concentrations of the RuII complexes. The reaction mixtures were
Short-term cell viability: Short-term cytotoxicity of the complexes
was determined by a standard MTT assay. Because living cells trans-
form MTT to a purple formazan dye, the yield of formazan trans-
formed by the cells exposed to the proper concentration of com-
plexes was used to measure cytotoxicity.[66,67] First, cells were
grown in an RPMI 1640 medium containing FBS (10%), penicillin
(100 mgmLÀ1) and streptomycin (100 mgmLÀ1). The cells were incu-
bated at 378C in a humidified incubator with 5% CO2 to the expo-
nential growth phase. Then, the cells were diluted, seeded in 96-
well microassay culture plates (1ꢁ104 cells per well) and incubated
under the same conditions previously used. After being incubated
for 24 h, the cells were treated with various concentrations of the
complexes. The tested compounds were subsequently dissolved in
sterile water and diluted with RPMI 1640 to the required concen-
trations. The medium and drug-free control samples were prepared
simultaneously. After the cells were incubated for another 48 h,
a stock MTT dye solution (20 mL, 5 mgmLÀ1) was added to each
well. After further incubation (4 h), all of the solutions in the wells
were discarded and 150 mL of DMSO were added to dissolve the
MTT formazan. The plate was analysed on a microplate spectro-
photometer at a wavelength of l=570 nm. The absorption data
were plotted on a logarithmic graph of the percent viability read-
Chem. Eur. J. 2015, 21, 1 – 14
www.chemeurj.org
11
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
ing of the control versus the concentration. IC50 is the concentra-
tion at which 50% of the cells were viable relative to the control.
Each experiment was repeated a minimum of three times to
obtain mean values. The HeLa (cervical) cell line was the subject of
this study.
for the four complexes. After being neutrally charged by their own
ions in solution, the combined systems were solvated into an octa-
hedral box of TIP3P water, resulting in a 12 ꢂ separation of the
complexes. The normal state of 300 K and 1 atm were set in the
MD simulations. The 8 ꢂ distance of the non-bonded cut-off value
was adopted to avoid interactions from atoms in an adjacent
periodic box or from long-range electrostatic forces. The MD
calculations proceeded with the following steps: 200 steps of the
steepest descent minimisation, 20 ps MD simulation of hydrogen
atoms with constant pressure, 60 ps MD equilibration with con-
stant pressure and 4.2 ns of MD equilibrations. Fifty snapshots
were extracted from the last 0.2 ns of the trajectories (at 4 ps inter-
vals) in each binding system. Their binding free energy values,
DGbind, were obtained by performing calculations of molecular me-
chanics/generalised Born surface area (MM/GBSA).[71] The DGbind
value is given by summing the electrostatic energy values (DGele),
the non-electrostatic energy values (DGnon-ele) and the entropy
component (TDS). In a solvent system, DGele can be split into the
electrostatic energy in the gas phase and the electrostatic contri-
bution to the solvation free energy, whereas DGnon-ele consists of
the van der Waals (DEvdW) energy and the nonpolar contributions
to the solvation free energy. The entropy contributions were not
considered in this study because of the structural similarity of the
ruthenium complexes. All of the calculations of MD and MM/GBSA,
as well as their preparations, were performed by using Amber10.[47]
Moreover, the central cations were not included in the MD model-
ling because no Na+ ions were available in the crystal structure of
its assembling parallel quadruplexes.
Long-term proliferation experiments: The traditional method to
detect long-term cell proliferation is to seed the cells onto a 96-
well plate and then culture, digest, dilute and reseed the cells for
more than ten days. After this process, the cells remaining in the
wells are stained with a SA-b-Gal staining kit for detection of the
expression of SA-b-Gal to determine the cell senescence and apop-
tosis.[68] Because this approach represents an end-point analytical
method, we could not monitor the cell growth conditions while
the experiment was in progress.[54] To overcome this shortcoming,
we used the xCELLigence system in our experiment. The xCELLi-
gence system offers a dynamic, real-time, label-free and non-inva-
sive analysis of a variety of cellular events. It can monitor cellular
events in real time without the incorporation of labels or any other
substances that may inhibit cell proliferation over a long period.
The system measures the electrical impedance switch as a dimen-
sionless parameter, termed the cell index (CI), across interdigitated
micro-electrodes on the bottom of tissue culture E-plates. Under
the same physiological conditions, the CI values are greater when
more cells are attached to the electrodes. Thus, the CI value can
represent the cell status, including the quantitative number of cells
present in a well. Herein, we designed a new method to exploit
the advantages of the xCELLigence system to their full potential.[21]
Long-term proliferation experiments were performed according to
the following procedure: HeLa cells (5.0ꢁ103 cells) were seeded in
a 16-well E-plate. A subcytotoxic concentration (2.5 mm) of a com-
plex was added to the experimental wells, whereas an equivalent
volume of 0.1% DMSO was added to the control wells, Each well
contained three parallel controls. Every three days, the cells in the
drug-exposed and control wells were collected, diluted tenfold and
reseeded into the same wells. This process enhances the different
statuses between the control cells and the drug-exposed cells, in-
cluding the cell number and shape. This experiment was continued
for approximately ten days.
Acknowledgements
This work was supported by the 973 program (2014CB845604),
the National Science Foundation of China (Nos. 21172273,
21171177 and 91122010), the Program for Changjiang Scholars
and Innovative Research Team in University of China (No.
IRT1298) and the Research Fund for the Doctoral Program of
Higher Education (20110171110013). L.X. thanks the support of
the National Science Foundation of China (21301034) and the
National Science Foundation of Guangdong Province
(S2013040014083).
Molecular modelling: Geometry optimisations for all of the ligands
were performed by the density functional B3LYP method with
a mixed basis set (SDD was used for Ru and 6-31G* was used for
the other atoms) by using the Gaussian 03 program.[69] Crystal
structures of parallel quadruplexes from the human telomeric DNA
sequence d(AG3[T2AG3]3) with K+ and Na+ were extracted from the
protein data bank (PDB ID 1KF1 and 143D, respectively). The
Amber FF99 force field was applied to the proteins after the sol-
vent/ions were removed and hydrogen atoms were added. Be-
cause the space between two G-quadruplets in PDB 143D was too
narrow, one of the phosphate backbones at the 5’-AG step was
broken to enlarge the space to hold these ligands, and then recon-
nected following by 1000 steps steepest descent minimisation.[70]
The Surflex-Dock (SFXC) module in SYBYL 7.3 (Tripos, Inc.) was
used to perform docking processes. In the docking analysis, pre-
dock and post-dock minimisations were not performed and the
ring flexibility was considered. Other parameters included 100 ad-
ditional starting conformations per molecule and an expanded
search grid of 5.00 ꢂ. To further elucidate the interactions between
the dinuclear ruthenium(II) complexes and the G-quadruplexes, the
best-scored docking conformations were introduced as the starting
structures in molecular dynamics (MD). Force field Amber FF99SB
was used for the DNA, whereas the GAFF force field was applied
for the ligands. The restrained electrostatic potential (RESP) charg-
es were calculated by the RESP module in AMBER[57] and were set
Keywords: conformation analysis · DNA · G-quadruplexes ·
ruthenium · telomeres
[1] G. N. Parkinson, M. P. H. Lee, S. Neidle, Nature 2002, 417, 876–880.
[2] J. T. Davis, Angew. Chem. Int. Ed. 2004, 43, 668–698; Angew. Chem.
2004, 116, 684–716.
[3] J. L. Huppert, Chem. Soc. Rev. 2008, 37, 1375–1384.
[4] E. Henderson, C. C. Hardin, S. K. Walk, I. Tinoco, Jr., E. H. Blackburn, Cell
1987, 51, 899–908.
[5] D. Sen, W. Gilbert, Nature 1988, 334, 364–366.
[6] T. S. Dexheimer, D. Sun, L. H. Hurley, J. Am. Chem. Soc. 2006, 128, 5404–
5415.
[7] S. Rankin, A. P. Reszka, J. Huppert, M. Zloh, G. N. Parkinson, A. K. Todd, S.
Ladame, S. Balasubramanian, S. Neidle, J. Am. Chem. Soc. 2005, 127,
10584–10589.
[8] T. Simonsson, P. Pecinka, M. Kubista, Nucleic Acids Res. 1998, 26, 1167–
1172.
[9] A. De Cian, L. Lacroix, C. Douarre, N. Temime-Smaali, C. Trentesaux, J. F.
Riou, J. L. Mergny, Biochimie 2008, 90, 131–155.
[10] P. Alberti, L. Lacroix, L. Guittat, C. Hꢃlꢄne, J.-L. Mergny, Mini-Rev. Med.
Chem. 2003, 3, 23–36.
&
&
Chem. Eur. J. 2015, 21, 1 – 14
www.chemeurj.org
12
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
[11] S. Balasubramanian, L. H. Hurley, S. Neidle, Nat. Rev. Drug Discovery
2011, 10, 261–275.
Merz, B. Wang, S. Hayik, A. Roitberg, G. Seabra, I. Kolossvꢅry, K. F. Wong,
F. Paesani, J. Vanicek, X. Wu, S. R. Brozell, T. Steinbrecher, H. Gohlke, L.
Yang, C. Tan, J. Mongan, V. Hornak, G. Cui, D. H. Mathews, M. G. Seetin,
C. Sagui, V. Babin, P. A. Kollman, University of California, San Francisco,
2008.
[12] S. Neidle, Curr. Opin. Struct. Biol. 2009, 19, 239–250.
[13] D. Yang, K. Okamoto, Future Med. Chem. 2010, 2, 619–646.
[14] J. H. Tan, L. Q. Gu„ J. Y. Wu, Mini-Rev. Med. Chem. 2008, 8, 1163–1178.
[15] G. N. Parkinson, F. Cuenca, S. Neidle, J. Mol. Biol. 2008, 381, 1145–1156.
[16] K. N. Luu, A. T. Phan, V. Kuryavyi, L. Lacroix, D. J. Patel, J. Am. Chem. Soc.
2006, 128, 9963–9970.
[48] X. H. Zheng, Y. F. Zhong, C. P. Tan, L. N. Ji, Z. W. Mao, Dalton Trans. 2012,
41, 11807–11812.
[49] G. W. Collie, R. Promontorio, S. M. Hampel, M. Micco, S. Neidle, G. N. Par-
kinson, J. Am. Chem. Soc. 2012, 134, 2723–2731.
[50] C. Y. Chen, Q. Wang, J. Q. Liu, Y. H. Hao, Z. Tan, J. Am. Chem. Soc. 2011,
133, 15036–15044.
[51] V. Hanusova, V. Kralova, L. Schroterova, L. Trilecova, A. Pakostova, L. Ska-
lova, Xenobiotica 2010, 40, 681–690.
[17] T. Wilson, P. J. Costa, V. Felix, M. P. Williamson, J. A. Thomas, J. Med.
Chem. 2013, 56, 8674–8683.
[18] T. Li, E. Wang, S. Dong, Anal. Chem. 2010, 82, 7576–7580.
[19] F. Hamon, E. Largy, A. Guedin-Beaurepaire, M. Rouchon-Dagois, A.
Sidibe, D. Monchaud, J. Mergny, J. Riou, C. Nguyen, M. Teulade-Fichou,
Angew. Chem. Int. Ed. 2011, 50, 8745–8749; Angew. Chem. 2011, 123,
8904–8908.
[52] Y. Chen, K. G. Qu, C. Q. Zhao, L. Wu, J. S. Ren, J. S. Wang, X. G. Qu, Nat.
Commun. 2012, 3, 1074.
[20] R. Hansel, F. Lohr, S. Foldynova-Trantirkova, E. Bamberg, L. Trantirek, V.
Dotsch, Nucleic Acids Res. 2011, 39, 5768–5775.
[21] C. Rajput, R. Rutkaite, L. Swanson, I. Haq, J. A. Thomas, Chem. Eur. J.
2006, 12, 4611–4619.
[22] D. Sun, Y. Liu, D. Liu, R. Zhang, X. Yang, J. Liu, Chem. Eur. J. 2012, 18,
4285–4295.
[23] X. Chen, J. H. Wu, Y. W. Lai, R. Zhao, H. Chao, L. N. Ji, Dalton Trans. 2013,
42, 4386–4397.
[24] D. D. Sun, R. Zhang, F. Yuan, D. Liu, Y. H. Zhou, J. Liu, Dalton Trans.
2012, 41, 1734–1741.
[25] S. Shi, J. Liu, T. M. Yao, X. T. Geng, L. F. Jiang, Q. Y. Yang, L. Cheng, L. N.
Ji, Inorg. Chem. 2008, 47, 2910–2912.
[53] Z. H. Siddik, Oncogene 2003, 22, 7265–7279.
[54] W. B. Wu, S. H. Chen, J. Q. Hou, J. H. Tan, T. M. Ou, S. L. Huang, D. Li,
L. Q. Gu, Z. S. Huang, Org. Biomol. Chem. 2011, 9, 2975–2986.
[55] E. S. Baker, J. T. Lee, J. L. Sessler, M. T. Bowers, J. Am. Chem. Soc. 2006,
128, 2641–2648.
[56] F. R. Liu, K. Z. Wang, G. Y. Bai, Y. A. Zhang, L. H. Gao, Inorg. Chem. 2004,
43, 1799–1806.
[57] H. Chao, Y. X. Yuan, F. Zhou, L. N. Ji, Transit. Met. Chem. 2006, 31, 465–
469.
[58] W. Paw, R. Eisenberg, Inorg. Chem. 1997, 36, 2287–2293.
[59] J. Bolger, A. Gourdon, E. Ishow, J. P. Launay, Inorg. Chem. 1996, 35,
2937–2944.
[26] S. Shi, J. Zhao, X. Gao, C. Y. Lv, L. Yang, J. Hao, H. L. Huang, J. L. Yao,
W. L. Sun, T. M. Yao, L. N. Ji, Dalton Trans. 2012, 41, 5789–5793.
[27] J. L. Yao, X. Gao, W. L. Sun, X. Z. Fan, S. Shi, T. M. Yao, Inorg. Chem. 2012,
51, 12591–12593.
[28] T. Wilson, M. P. Williamson, J. A. Thomas, Org. Biomol. Chem. 2010, 8,
2617–2621.
[60] L. J. Chen, G. D. Qian, Y. J. Cui, X. F. Jin, Z. Y. Wang, M. Q. Wang, J. Phys.
Chem. B 2006, 110, 19176–19182.
[61] A. D. Moorhouse, A. M. Santos, M. Gunaratnam, M. Moore, S. Neidle,
J. E. Moses, J. Am. Chem. Soc. 2006, 128, 15972–15973.
[62] A. De Cian, E. Delemos, J. L. Mergny, M. P. Teulade-Fichou, D. Monchaud,
J. Am. Chem. Soc. 2007, 129, 1856–1857.
[29] L. Xu, D. Zhang, J. Huang, M. G. Deng, M. Zhang, X. Zhou, Chem.
Commun. 2010, 46, 743–745.
[30] G. A. L. Bare, B. Liu„ J. C. Sherman, J. Am. Chem. Soc. 2013, 135, 11985–
11989.
[31] W. J. Chung, B. Heddi, M. Tera, K. Iida, K. Nagasawa, A. T. Phan, J. Am.
Chem. Soc. 2013, 135, 13495–13501.
[32] X. X. Wang, C. X. Zhou, J. W. Yan, J. Q. Hou, S. B. Chen, T. M. Ou, L. Q. Gu,
Z. S. Huang, J. H. Tan, ACS Med. Chem. Lett. 2013, 4, 909–914.
[33] M. Franceschin, Eur. J. Org. Chem. 2009, 2225–2238.
[34] N. Ranjan, E. Davis, L. Xue, D. P. Arya, Chem. Commun. 2013, 49, 5796–
5798.
[63] P. Job, Ann. Chim. 1928, 9, 113–203.
[64] T. Lemarteleur, D. Gomez, R. Paterski, E. Mandine, P. Maillet, J. F. Riou,
Biochem. Biophys. Res. Commun. 2004, 323, 802–808.
[65] J. Reed, M. Gunaratnam, M. Beltran, A. P. Reszka, R. Vilar, S. Neidle, Anal.
Biochem. 2008, 380, 99–105.
[66] S. Kemp, N. J. Wheate, D. P. Buck, M. Nikac, J. G. Collins, J. R. Aldrich-
Wright, J. Inorg. Biochem. 2007, 101, 1049–1058.
[67] J. Ai, W. Guo, B. Li, T. Li, D. Li, E. Wang, Talanta 2012, 88, 450–455.
[68] J. F. Riou, L. Guittat, P. Mailliet, A. Laoui, E. Renou, O. Petitgenet, F.
Megnin-Chanet, C. Helene, J. L. Mergny, Proc. Natl. Acad. Sci. USA 2002,
99, 2672–2677.
[35] T. M. Ou, Y. J. Lu, J. H. Tan, Z. S. Huang, K. Y. Wong, L. Q. Gu, ChemMed-
Chem 2008, 3, 690–713.
[69] Gaussian 03, Revision D.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven,
K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone,
B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsu-
ji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Na-
kajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P.
Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts,
R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Och-
terski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg,
V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K.
Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui,
A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko,
P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham,
C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wallingford
CT, 2004.
[36] D. L. Ma, C. M. Che, S. C. Yan, J. Am. Chem. Soc. 2009, 131, 1835–1846.
[37] S. Bianco, C. Musetti, A. Waldeck, S. Sparapani, J. D. Seitz, A. P. Krapcho,
M. Palumbo, C. Sissi, Dalton Trans. 2010, 39, 5833–5841.
[38] K. Suntharalingam, O. Mendoza, A. A. Duarte, D. J. Mann, R. Vilar, Metal-
lomics 2013, 5, 514–523.
[39] K. Suntharalingam, A. J. P. White, R. Vilar, Inorg. Chem. 2010, 49, 8371–
8380.
[40] S. N. Georgiades, N. H. Abd Karim, K. Suntharalingam, R. Vilar, Angew.
Chem. Int. Ed. 2010, 49, 4020–4034; Angew. Chem. 2010, 122, 4114–
4128.
[41] J. Zhou, G. Yuan, Chem. Eur. J. 2007, 13, 5018–5023.
[42] D. P. N. Goncalves, S. Ladame, S. Balasubramanian, J. K. M. Sanders, Org.
Biomol. Chem. 2006, 4, 3337–3342.
[43] R. Giraldo, M. Suzuki, L. Chapman, D. Rhodes, Proc. Natl. Acad. Sci. USA
1994, 91, 7658–7662.
[70] M. Read, R. J. Harrison, B. Romagnoli, F. A. Tanious, S. H. Gowan, A. P.
Reszka, W. D. Wilson, L. R. Kelland, S. Neidle, Proc. Natl. Acad. Sci. USA
2001, 98, 4844–4849.
´
´
[44] V. Dapic, V. Abdomerovic, R. Marrington, J. Peberdy, A. Rodger, J. O.
Trent, P. J. Bates, Nucleic Acids Res. 2003, 31, 2097–2107.
[45] E. M. Rezler, J. Seenisamy, S. Bashyam, M. Y. Kim, E. White, W. D. Wilson,
L. H. Hurley, J. Am. Chem. Soc. 2005, 127, 9439–9447.
[71] J. Srinivasan, T. E. Cheatham III, P. Cieplak, P. A. Kollman, D. A. Case, J.
Am. Chem. Soc. 1998, 120, 9401–9409.
[46] Y. Xue, Z. Y. Kan, Q. Wang, Y. Yao, J. Liu, Y. H. Hao, Z. Tan, J. Am. Chem.
Soc. 2007, 129, 11185–11191.
Received: November 6, 2014
[47] AMBER 10, D. A. Case, T. A. Darden, T. E. Cheatham III, C. L. Simmerling,
J. Wang, R. E. Duke, R. Luo, M. Crowley, R. C. Walker, W. Zhang, K. M.
Published online on && &&, 0000
Chem. Eur. J. 2015, 21, 1 – 14
www.chemeurj.org
13
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
&
G-Quadruplexes
L. Xu, X. Chen, J. Wu, J. Wang, L. Ji,
H. Chao*
&& – &&
Dinuclear Ruthenium(II) Complexes
That Induce and Stabilise G-
Quadruplex DNA
Being blocked by Ru: A series of
dinuclear ruthenium(II) complexes (see
figure) were found to promote high
stabilisation of the G-quadruplexes of
human telomeric DNA and exhibited
a high telomerase inhibiting activity.
&
&
Chem. Eur. J. 2015, 21, 1 – 14
www.chemeurj.org
14
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!