Aromatic Core Extension in the Series of N-Cyclic Bay-Substituted Perylene G-Quadruplex Ligands: Increased Telomere Damage, Antitumor Activity, and Strong Selectivity for Neoplastic over Healthy Cells

Article abstract of DOI:10.1002/cmdc.201200348

Based on previous work on both perylene and coronene derivatives as G-quadruplex binders, a novel chimeric compound was designed: N,N'-bis[2-(1-piperidino)-ethyl]-1-(1-piperidinyl)-6-[2-(1-piperidino)-ethyl]-benzo[ghi]perylene-3,4:9,10-tetracarboxylic diimide (EMICORON), having one piperidinyl group bound to the perylene bay area (positions 1, 12 and 6, 7 of the aromatic core), sufficient to guarantee good selectivity, and an extended aromatic core able to increase the stacking interactions with the terminal tetrad of the G-quadruplex. The obtained chimera molecule, EMICORON, rapidly triggers extensive DNA damage of telomeres, associated with the delocalization of telomeric protein protection of telomeres1 (POT1), and efficiently limits the growth of both telomerase-positive and -negative tumor cells. Notably, the biological effects of EMICORON are more potent than those of the previously described perylene derivative (PPL3C), and more interestingly, EMICORON appears to be detrimental to transformed and tumor cells, while normal fibroblasts expressing telomerase remain unaffected. These results identify a new promising G-quadruplex ligand, structurally and biologically similar on one side to coronene and on the other side to a bay-monosubstituted perylene, that warrants further studies.

Full text of DOI:10.1002/cmdc.201200348

MED
DOI: 10.1002/cmdc.201200348
Aromatic Core Extension in the Series of N-Cyclic Bay-
Substituted Perylene G-Quadruplex Ligands: Increased
Telomere Damage, Antitumor Activity, and Strong
Selectivity for Neoplastic over Healthy Cells
Marco Franceschin,[a] Angela Rizzo,[b] Valentina Casagrande,[a] Erica Salvati,[b]
Antonello Alvino,[a] Alessandro Altieri,[a] Alina Ciammaichella,[a] Sara Iachettini,[b]
Carlo Leonetti,[b] Giancarlo Ortaggi,[a] Manuela Porru,[b] Armandodoriano Bianco,*[a] and
Annamaria Biroccio*[b]
Based on previous work on both perylene and coronene deriv-
atives as G-quadruplex binders, a novel chimeric compound
was designed: N,N’-bis[2-(1-piperidino)-ethyl]-1-(1-piperidinyl)-
6-[2-(1-piperidino)-ethyl]-benzo[ghi]perylene-3,4:9,10-tetracar-
boxylic diimide (EMICORON), having one piperidinyl group
bound to the perylene bay area (positions 1, 12 and 6, 7 of the
aromatic core), sufficient to guarantee good selectivity, and an
extended aromatic core able to increase the stacking interac-
tions with the terminal tetrad of the G-quadruplex. The ob-
tained “chimera” molecule, EMICORON, rapidly triggers exten-
sive DNA damage of telomeres, associated with the delocaliza-
tion of telomeric protein protection of telomeres 1 (POT1), and
efficiently limits the growth of both telomerase-positive and
-negative tumor cells. Notably, the biological effects of EMI-
CORON are more potent than those of the previously de-
scribed perylene derivative (PPL3C), and more interestingly,
EMICORON appears to be detrimental to transformed and
tumor cells, while normal fibroblasts expressing telomerase
remain unaffected. These results identify a new promising G-
quadruplex ligand, structurally and biologically similar on one
side to coronene and on the other side to a bay-monosubsti-
tuted perylene, that warrants further studies.
Introduction
G-quadruplexes (G4) are non-canonical secondary structures
formed in DNA sequences containing consecutive runs of gua-
nines that are thought to play a role in key biological process-
es including telomere maintenance.[1] These findings have
prompted a search for small organic molecules as specific li-
gands for these structures, for their development as potential
anticancer agents.[2] The number of known G4 ligands has
grown rapidly over the past few years.[3] Features shared by
many of these ligands include an aromatic core that favors
stacking interactions with the G-tetrads, and, in most cases,
basic side chains that interact with the quadruplex grooves.[4]
Despite the promising results obtained in preclinical models,
only two drugs that target either a G4 (Quarfloxin) or fold into
a G4 structure (AS1411) are presently in phase II clinical trials.[5]
G4 ligands were initially designed to counteract telomerase
action at telomeres. Surprisingly, their antiproliferative effects
can occur in telomerase-negative cells and follow kinetics,
which cannot be merely explained by telomere shortening,
suggesting that these compounds affect other pathways that
are not necessarily related to telomere biology. Results from
different research groups,[6] including ours,[7] clearly demon-
strated that, in addition to their telomerase inhibitory proper-
ties, these drugs can exert an anticancer effect by telomeric
chromatin alteration, leading to the activation of damage foci
(i.e., gH2AX, p53 binding protein 1 (53BP1)) that co-localize
with the telomeric repeats. It emerged that G4-interacting
agents are more than simple telomerase inhibitors and that
their direct target is telomere instead of telomerase.
Our group recently reported an N-cyclic bay-monosubstitut-
ed perylene diimide 1 (PPL3C, Figure 1) as a selective G4
ligand able to induce a specific telomere damage and antipro-
liferative activity on both transformed and tumor cells.[8] In par-
ticular, we pointed out that the presence of bulky groups on
the perylene bay area (positions 1, 12 and 6, 7) gives a good
selectivity with respect to duplex DNA, which was missing in
previously reported unsubstituted perylene diimides. Further-
more, we found that only one piperidinyl group is necessary
to give such selectivity, whereas the second piperidinyl group
[a] Dr. M. Franceschin,+ Dr. V. Casagrande, Dr. A. Alvino, Dr. A. Altieri,
Dr. A. Ciammaichella, Prof. G. Ortaggi, Prof. A. Bianco
Dipartimento di Chimica
Sapienza Universitꢀ di Roma
P.le.A. Moro 5, 00185 Roma (Italy)
E-mail: Armandodoriano.Bianco@uniroma1.it
[b] Dr. A. Rizzo,+ Dr. E. Salvati, Dr. S. Iachettini, Dr. C. Leonetti, Dr. M. Porru,
Dr. A. Biroccio
Experimental Chemotherapy Laboratory
Regina Elena National Cancer Institute
Via delle Messi d’Oro 156, 00158 Rome (Italy)
E-mail: biroccio@ifo.it
[+] These authors contributed equally to the paper.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201200348.
2144
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2012, 7, 2144 – 2154

N-Cyclic Bay-Substituted Perylene G-Quadruplex Ligands
MED
formed cells, even if, consistently with the previously reported
high capability of 2 to bind G4 versus duplex DNA,[10d] DNA
damage is also localized at the telomeric regions (see Fig-
ure S1C and D in the Supporting Information). These results
suggest that 2 can also interact with additional targets outside
the telomere and indicates that it can have off-target effects.
Following these considerations, we designed a “chimera”
molecule, in which one half is similar to coronene and for the
other half to a bay-monosubstituted perylene (EMICORON (3),
Figure 1).
Results
Synthesis
Starting compound 4 (PIPER-Br) was prepared as previously de-
scribed.[11] A small amount (no more than 10% at this stage) of
the 1,6-isomer was always present, which was determined by
1
analyzing the H NMR spectra, as discussed in the cited litera-
ture. In Scheme 1, only the isomer 4 and the relative subse-
quent products are shown. The first step was the selective sub-
stitution of only one bromine atom with piperidine
(Scheme 1). Conditions were optimized to increase the ratio
between the desired mono derivative (6) and the disubstituted
product (5). Initially we experimented with the conditions em-
ployed for the synthesis of three-side-chained DAPER deriva-
tives[12] treating 4 with piperidine at room temperature. We ob-
tained 6 almost in equal amount with respect to the disubsti-
tuted product 5. After experimenting with several reaction
conditions (see Table S2 in the Supporting Information), we ob-
tained better results at higher temperatures, as described in
the Experimental Section. We also tried to add hydroquinone
to the reaction mixture, which is a molecule able to capture
radicals. In fact, in our experience, when hydroquinone is
added to different reactions to avoid the formation of dehalo-
genated products, it seems to be able to delay the displace-
ment of the second bromine atom; an observation that could
suggest the involvement of a radicalic mechanism (data not
shown). To optimize the reaction conditions we tried also to
use an excess of piperidine and the reaction was stirred at
1008C for a shorter time. We also compared the reaction yields
in the absence and presence of hydroquinone, identifying the
latter conditions as optimal. The products obtained by this re-
action are not easily and completely separable by column
chromatography, so the partially purified mixture was used in
the following synthetic step. The relative ratio values of com-
pounds 5/6 were calculated from the 1H NMR spectra.
Figure 1. Chemical structures of N-cyclic bay-substituted perylene and coro-
nene derivatives. “Chimera” molecule 3 has one half similar to coronene and
the other half similar to bay-monosubstituted perylene.
hinders the interaction with the terminal quadruplex G-
tetrad.[8] For these reasons, a bay-monosubstituted perylene
diimide such as 1 is characterized by decreased steric hin-
drance, which (compared with other derivatives with two
bulky substituents on the bay area) enables improved overlap
on the terminal G-tetrad to occur. From a structural point of
view, we noticed that the second bay area of monosubstituted
perylene diimides was free to be derivatized. Therefore, it may
be interesting to enlarge the planar aromatic area on this side
of the molecule, to improve interaction with the terminal G-
tetrad, provided that the selectivity with respect to duplex
DNA should be granted by the presence of the piperidinyl
group on the other side.
In fact, we previously reported a series of hydrosoluble coro-
nene derivatives, which are characterized by a large hydropho-
bic aromatic core, as G4 ligands and telomerase inhibitors.[9] By
using ESI-MS, we showed that coronene derivatives have
a much higher capability of binding to G4 DNA than two side-
chain perylene derivatives,[10] in agreement with the idea that
a larger aromatic core should give better interaction with the
G-tetrads.[4] Specific biological and molecular assays have been
performed on the reference compound of this series 2 (CORON
(2), Figure 1); we found that this G4 ligand is highly active, in
terms of inhibition of cell proliferation on both transformed-
and tumor cells, but this effect also occurs in normal telomer-
ized fibroblasts (see Table S1 and Figure S1A and B in the Sup-
porting Information). In agreement with these results, com-
pound 2 induces DNA damage, both in normal and trans-
To proceed towards the extension of the aromatic core,
a suitable functionalized alkyne (7) was prepared, having 1-pi-
peridine at the end of the linear chain.[11] This alkyne was suc-
cessively used in a Sonogashira cross-coupling[13] to synthesize
the intermediate compound 8. This reaction is catalyzed by
Pd0 complexes in the presence of CuI and a suitable base, lead-
ing to a new CÀC bond. In contrast to the previously reported
synthesis of coronene derivatives, in this case the Sonogashira
reaction conditions led to the asymmetrical compound 8 due
to the presence of only one bromine atom on compound 6. In
ChemMedChem 2012, 7, 2144 – 2154
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
2145

A. Biroccio et al.
MED
quadruplex structure formed by
the human telomeric sequence
and its selectivity with respect to
duplex DNA.[10] To this aim, an
oligonucleotide
representing
four human telomeric repeats
(21-TT, 5’-GGGTTAGGGTTAGGGT-
TAGGGTT-3’) and a model oligo-
nucleotide for duplex DNA
(DK66,
5’-CGTAAATTTACG-3’)
were used. The association con-
stants K1 and K2, relating to the
equilibrium for the formation of
drug–DNA complexes with 1:1
and 2:1 stoichiometry respec-
tively, as well as the correspond-
ing percentage of bound DNA,
are reported in Table 1, with the
relative standard deviations.
The collected data demon-
strates that 3 is a stronger G4
ligand than the previously re-
ported coronene derivative 2,[10d]
showing a K1 value substantially
equal to 2 but a K2 value that is
two orders of magnitude higher.
In fact, calculating the total
amount of ligand bound in the
1:1 ratio experiments, the values
obtained were higher than 99%,
showing full saturation. Com-
pound 3 displayed a low affinity
towards duplex oligonucleotide
by binding twice as much quad-
ruplex DNA as duplex DNA,
showing enhanced selectivity
compared to 2.[10d] On the other
hand, this selectivity is lower
than that showed by previously
reported 1, even though com-
pound 3 has a much higher G4-
binding ability. So we decided to
Scheme 1. Reagents and conditions: a) Hydroquinone, piperidine, anhydrous dioxane, 1008C, argon atmosphere,
40 min; b) THF, Et3N, CuI, [Pd(PPh3)4], argon atmosphere, 808C, 20 h: c) DBU, toluene, argon atmosphere, heated
to reflux, 20 h.
the Sonogashira coupling, the cyclization described in the next
step also partially occurs. Separation was not useful at this
stage, so the mixtures were used in the following
step. To complete the aromatic core, a base-catalyzed
perform competition experiments by using genomic DNA from
calf thymus (CT).[8] Due to the high affinity demonstrated to-
Table 1. Values of Log K1 and Log K2 and bound DNA (%).[a]
cyclization was performed with 1,8-diazabicyclo-
[5.4.0]undec-7-ene (DBU). In this way the desired
compound 3 was obtained (Scheme 1), which was
converted into the respective water-soluble hydro-
chloride by precipitation with diethyl ether from an
acidic (HCl) methanol solution.
21-TT
DK66
Log K1
Log K2
Bound
Log K1
Log K2
Bound
DNA [%]
DNA [%]
PPL3C[8] (1)
CORON[10d] (2)
EMICORON (3)
5.8Æ0.2
6.6Æ0.3
6.5Æ0.2
5.0Æ0.1
4.2Æ0.1
6.3Æ0.2
56Æ7
78Æ5
73Æ2
3.9Æ0.2
6.0Æ0.3
5.4Æ0.1
4.2Æ0.1
5.2Æ0.2
4.9Æ0.1
<5
59Æ5
38Æ1
[a] K1 and K2 values (reported on a logarithmic scale) and percentage of bound DNA
calculated at 1:1 drug/DNA ratio, as described in the Experimental section, for the in-
dicated oligonucleotides. Values represent the mean Æ standard deviation (SD) of at
least three independent experiments.
ESI-MS study of binding affinity and selectivity
The new synthesized derivative 3 was studied by ESI-
MS to evaluate its ability to bind the intramolecular
2146
www.chemmedchem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2012, 7, 2144 – 2154

N-Cyclic Bay-Substituted Perylene G-Quadruplex Ligands
MED
wards the G4 structure, it was possible to perform these ex-
periments with a lower drug concentration. At half the concen-
tration of drug and DNA both the 1:1 and the 2:1 complexes
were clearly visible (see Figure S2 in the Supporting Informa-
tion). Percentages of bound DNA in the absence and in the
presence of calf thymus DNA are reported in Table 2.
The results demonstrated that 3 was able to induce DNA
damage, as revealed by the increased percentage of both
gH2AX- and 53BP1-positive cells and this effect was selective
for transformed cells (Figure 3a). Indeed, gH2AX and 53BPI
were not significantly (P>0.05) activated in normal telomer-
ized fibroblasts upon treatment with 3, even at the highest
drug concentration (Figure 3a). The selective DNA damage in-
duced by 3 in transformed cells prompted further studies to
investigate the telomere specific effects of this ligand. Decon-
volution microscopy analysis, reported in Figure 3b, revealed
that some of the damaged foci (both gH2AX and 53BP1) in-
duced by 3 co-localized with TRF1, a good marker for inter-
phase telomeres[15] forming the so-called TIFs (telomere-dys-
function induced foci).[16] Quantitative analysis revealed that
treatment with 3 significantly (P<0.01) increased the percent-
age of cells with more than four gH2AX/TRF1 co-localizations
(the percentage of TIFs-positive cells reached about 50% upon
treatment with 0.1 mm concentration), with a mean of about
seven TIFs per nucleus (Figure 3c).
Compound 3 showed a good selectivity in competition ex-
periments, similar to 1[8] and 2[10d] toward the same oligonu-
cleotide (which is the biologically relevant human telomeric se-
quence), but with the advantage of an increased binding affini-
ty. In fact, it is worth noting that the results shown have been
obtained at a 0.5 drug/DNA ratio, whereas in the other cases
1:1 experiments were performed due to the minor binding
affinity of the drug.
Table 2. Competition experiments on the 21-TT oligonucleotide.[a]
21-TT/CT DNA ratio
no CT
1:1
N[%] at 1:1
1:5
N[%] at 1:5
PPL3C[8] (1)
CORON[10d] (2)
EMICORON (3)[b]
56Æ7 47Æ3
78Æ5 63Æ5
38Æ1 33Æ1
0.84
0.81
0.87
39Æ1
42Æ5
22Æ1
0.70
0.54
0.58
[a] Values of percentage of bound 21-TT and normalized percentage
bound quadruplex (0<N%<1), as defined in the Experimental Section,
for samples containing a fixed amount of both drug and G4 DNA and dif-
ferent amounts of calf thymus DNA (CT), at the indicated quadruplex/
duplex ratios (in phosphate ions). Values represent the mean Æ SD of at
least three independent experiments. [b] Data obtained at a drug/DNA
ratio of 0.5:1.
Effect on the telomerase/telomere pathway
Due to the inability of telomerase to extend a quadruplex-
folded telomeric substrate, G4 ligands were firstly designed as
telomerase inhibitors. We therefore investigated the ability of
3 to inhibit human telomerase activity in a cell-free system by
means of a telomerase repeat amplification protocol (TRAP)
assay. Of note, although a major risk of handling variability
exists, to avoid false-positive results and/or overestimate the
inhibitory effect of the G4 ligand, the TRAP assay has been per-
formed by removing the inhibitor after telomerase exten-
sion.[14] This is an important step during the PCR-based TRAP
assay because most G4 ligands can interfere with the PCR am-
plification of a sequence able to form a G4. As it is evident
from the Figure 2, compound 3 was able to inhibit telomerase
activity in a dose-dependent manner. However, this effect oc-
curred at a very high dose of the drug; the IC50 value is ap-
proximately 30 mm under the experimental conditions and
therefore 3 is a poor telomerase inhibitor.
Figure 2. Inhibition of telomerase activity. Telomerase activity was evaluated
by the telomerase repeat amplification protocol (TRAP) assay as reported in
the Experimental Section. Compound 3 was added at different concentra-
tions ranging from 1 to 200 mm. Before PCR, the samples were purified by
phenol/chloroform extraction. Products were resolved on 12% PAGE and vi-
sualized with SYBR Green staining. A representative TRAP assay is showed.
The quantitative analysis was undertaken as reported in the Experimental
Section, and data represent the mean of three independent experiments
with similar results. The intensities of TRAP products were normalized to the
intensities of the corresponding bands in the untreated sample. Telomerase
inhibition (%) was plotted against compound concentration to determine
the IC50 value (IC50 =27.0Æ2.0 mm).
Consequently, the ability of 3 to cause telomere uncapping
has been investigated (Figure 3). To this aim, a two-step analy-
sis was performed to establish, firstly, if the compound was
able to induce DNA damage and, secondly, if this DNA
damage was localized to the telomeres. In particular, BJ fibro-
blasts (human foreskin) expressing human telomerase reverse
transcriptase (BJ-hTERT) or hTERT and SV40 early region (BJ-
HELT), were exposed to different drug concentrations for 12 h.
To determine the cause of telomere uncapping, we investi-
gated the localization of TRF1, telomeric repeat binding factor
2 (TRF2), and protection of telomeres 1 (POT1), three telomeric
proteins inducing telomere dysfunction and evoking a DNA
damage signal when their levels are reduced at telomeres. A
ChIP assay showed that 3 delocalized POT1 from telomeres,
whereas TRF1 and TRF2 remained associated to the telomeres
upon treatment with the ligand (Figure 3d).
ChemMedChem 2012, 7, 2144 – 2154
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
2147

A. Biroccio et al.
MED
Figure 3. Activation of DNA damage response in transformed cells is associated with the displacement of POT1 from telomeres. a) Transformed (BJ-EHLT) and
normal telomerized (BJ-hTERT) human fibroblasts were treated with different concentrations of 3 for 24 h, fixed, and processed for IF by using antibodies
against gH2AX and 53PB1. Quantitative analysis, showing the percentage of gH2AX-(gray bars) and 53PB1-positive cells (black bars), is reported. Three inde-
pendent experiments were performed, and error bars indicate the standard deviation (SD); [**]=(p<0.01). b) Untreated and 3-treated BJ-EHLT cells were co-
immunostained with antibodies against TRF1 and gH2AX or 53BP1 and processed for IF. Representative images of IF were acquired with a Leica Deconvolu-
tion microscope (magnification 100ꢁ). Enlarged views are reported on the right of the merged images from 3-treated samples. c) Percentage of TIFs-positive
cells and average number of TIFs per nucleus in untreated treated cells. Cells with four or more gH2AX (gray bars) or 53BP1/TRF1 foci (black bars) were
scored as TIFs-positive. The mean of three independent experiments with similar results is reported. Error bars indicate the SD. D) Chromatin extracts from
the indicated samples were subjected to ChIP analysis using antibodies against TRF1, TRF2, and POT1. A b-actin antibody was used as the negative control.
The total DNA (input) represents 10 and 1% of genomic DNA. Southern blot analysis was performed by using telomeric (Telo) or ALU repeat-specific probes.
A representative experiment is shown on the left panel. The signals obtained were quantified by densitometry, and the percentage of precipitated DNA for
each untreated (gray bars) or compound 3-exposed (black bars) sample was calculated as a ratio of input signals and plotted in the graph (right panel). Four
independent experiments were performed, and error bars indicate the SDs.
2148
www.chemmedchem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2012, 7, 2144 – 2154

N-Cyclic Bay-Substituted Perylene G-Quadruplex Ligands
MED
ing possibility that telomere
damages induced by 3 may rap-
idly promote growth inhibition
selectively in malignant cells.
Transformed- and normal telo-
merized fibroblasts were ex-
posed to the lower dose of drug
able to trigger telomere damage
when cells were exposed for
12 h. The growth curves of un-
treated and drug-treated cells
were analyzed from 2 to 8 days
of culture (Figure 4a). A signifi-
cant time-dependent decrease
of cell proliferation has been ob-
served in transformed BJ-EHLT
cells treated with the ligand,
reaching the maximum effect at
day 8. Interestingly, at the same
drug dose, compound 3 slightly
reduces the cell proliferation of
normal telomerized fibroblasts
(being the growth inhibition
about 20% at days 6–8 of cul-
ture), suggesting that this agent
would preferentially kill cancer
cells.
On the basis of these results,
compound 3 was studied for an-
tiproliferative activity against
human cancer cell lines by the
US National Cancer Institute
(NCI)/National
Institutes
of
Health (NIH) developmental
therapeutics program.[18] In par-
ticular, this compound has been
evaluated in the full panel of
human tumor cell lines derived
from nine cancer cell types (leu-
kemia, melanoma and cancers of
non-small cell lung, colon, CNS,
prostate, ovarian, renal, and
breast). Figure 4b shows the GI50
(dose inhibiting the 50% of the
tumor growth) of each cell line
by subpanel group; the full data
are reported in the Supporting
Information (Table S3). Evidently,
compound 3 is highly active in
Figure 4. Antiproliferative effect of compound 3 is specific for transformed and tumor cells. a) Upper panels: in
&
&
vitro growth curves of BJ-EHLT and BJ-hTERT untreated ( ) and treated ( ) with 3 (0.1 mm). The figures show rep-
resentative experiments performed in quintuplicate with SDs. Lower panels: percentage of growth inhibition of
drug-treated versus untreated cells calculated at day 2, 4, 6, and 8 of cell culture. b) Cytotoxicity of 3 in the NCI in
vitro 60 human cancer cell lines drug screen program.
Effect on cellular proliferation and tumor survival
inhibiting tumor cell proliferation; the GI50 median value of
each cell line from the subpanel groups was shown to be less
than 1 mm (Figure 4b).
The effect of 3 on tumor cells has been also evaluated by
clonogenic assay both in telomerase-negative and -positive
tumor cells. In particular, the analysis was performed on U2OS
cells, which maintain telomeres through alternative mechanism
(ALT) and on three telomerase-positive cell lines of different
Telomeres emerge as cellular integrators of various stresses.[17]
Consequently, changes in their structure profoundly affect the
ability of cells to proliferate and to adapt to a new environ-
ment. Notably, even if the molecular mechanism(s) were not
yet described, pharmacological telomere damage induced by
G4 ligands triggers a selective anticancer effect on transformed
and tumor cells. Therefore, the above results raise the interest-
ChemMedChem 2012, 7, 2144 – 2154
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
2149

A. Biroccio et al.
MED
Figure 5. Cytotoxic activity of compound 3 in different tumor cell lines: a) Human U2OS osteosarcoma, M14 melanoma; b) HT29 colorectal adenocarcinoma,
CG5 breast cancer cell lines, and c) wild-type or p53-deficient HCT116 colon cancer cell lines were treated with the indicated doses (ranging from 0.1 to
1.5 mm) of compound 3 for 96 h and the colony-forming ability was evaluated as reported in the Experimental Section. Surviving fractions were calculated as
the ratio of absolute survival of the treated sample/absolute survival of the control sample. Three independent experiments were performed, and error bars
indicate the SDs.
histotype, M14 melanoma, HT29 colon and CG5 breast carcino-
Discussion and Conclusions
mas. The results, reported in Figure 5, show that 3 inhibited
cell survival in a dose-dependent manner in all the tumor lines
employed, regardless the presence of telomerase, and more in-
terestingly, the IC50 values being always below to 1 mm (Fig-
ure 5a). More interestingly, even if 3 was active both in tumor
cells expressing the wild type or mutant p53, the lack of this
oncosuppressor in the HCT116 colon-carcinoma cell line, can
make the cells more sensitive to treatment with 3, the IC50
value being reduced up to about 40% in p53-/- compared to
the wild-type cells.
Our previous results on N-cyclic bay-substituted perylene dii-
mides showed that one piperidinyl group is necessary to
ensure quadruplex-versus-duplex selectivity; the second group
cannot bring any further advantage, making the interaction
with the ending tetrad of the quadruplex structure more diffi-
cult.[8]
Herein we have designed the “chimera” molecule 3, having
one piperidinyl group bound to the perylene bay area and an
extended aromatic, describing the synthesis of this new asym-
metric compound. Initially, by using ESI-MS, we studied the
2150
www.chemmedchem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2012, 7, 2144 – 2154

N-Cyclic Bay-Substituted Perylene G-Quadruplex Ligands
MED
binding of this compound to the human telomeric G-rich se-
quence and its selectivity with respect to duplex DNA, also in
the case of genomic DNA. We found that 3 shows a good
quadruplex-versus-duplex selectivity, but with the advantage
of higher binding constants than the previously reported com-
pounds, both with respect to N-cyclic bay-monosubstituted
perylene diimide 1 and to coronene derivative 2.
From a molecular point of view, compound 3 inhibits telo-
merase activity, even if this effect occurs at very high drug con-
centration. However, it is difficult to compare the IC50 value of
this compound with that of other G4 ligands available in the
literature because, as already noted in several papers, the IC50
value derived from the TRAP assay depends on the assay con-
ditions and primer concentration.[6b,19,20] More importantly,
many studies have used the TRAP assay without removing the
inhibitor prior to the PCR step. As a consequence, the inhibito-
ry effect of many G4 ligands has been overestimated. However,
since the biological effects were also quickly observed in telo-
merase-negative ALT cells, we can conclude that the impair-
ment of telomere elongation by 3 is not the main mechanism
of drug action.
ferative activity induced by 3 is clearly evident in transformed
cells, whereas normal telomerized fibroblasts are only weakly
affected by the treatment. This result strongly suggests that
this agent displays a marked narrow window for selectivity,
which presents an interesting situation in the light of the
future clinical development of this new class of antitumoral
drugs. These findings are in good agreement with the quadru-
plex-versus-duplex selectivity derived by ESI-MS, as well as
with the stronger G4 binding ability of this compound with re-
spect to the previously reported compounds of these series.
This enforces the idea that ESI-MS can be a quick and efficient
method to screen G4 ligands and to predict their activity.
The interestingly activity on transformed fibroblasts has
been also observed on cancer cells as reported by the data
from the NCI Drug Screen Program. Indeed, compound 3 is
highly active on all the human tumor cell lines tested; the GI50
values being always below the 1 mm concentration, regardless
the tumor hystotype, thus providing a compelling rationale to
target telomere pathway for broad-spectrum cancer therapy.
Again, these results have been corroborated by using the clo-
nogenic assay, which is considered the most valid in vitro test
to measure the surviving fraction of cells after a drug treat-
ment, and are consistent with a cancer stem cell targeting
mechanism. Finally, the efficacy of 3 on all the transformed
and tumor cell lines tested, suggests that the oncosuppressor
p53, one of the main apoptosis inducers that is frequently mu-
tated on cancers, is not necessary for the response of tumor
cells to the G4 ligand. However, the use of a genetically-de-
fined tumor model of HCT116 colon carcinoma cell line, pos-
sessing the wild-type or mutant p53 in the same genetic back-
ground, revealed that the lack of this oncosuppressor can
make the cells more sensitive to the treatment with 3. Taken
together, these results indicate that p53 can be involved in the
repair of G4-induced telomere damage rather than in tumor
cell death.
Considering that (from a biological point of view) compound
3 cannot be considered a telomerase inhibitor, we still decided
to proceed in the analysis of its molecular and biological ef-
fects. This decision was based on data demonstrating that the
presence of telomerase in tumor cells (as a mechanism of telo-
mere maintenance), is not mandatory for the antitumoral activ-
ity of G4-interactive compounds.[6–8,20] Thus, we looked beyond
telomerase activity and function and we focused on the
capped status to evaluate telomere functionality following
treatment with the telomere-targeting agent. Our results clear-
ly demonstrated that 3 rapidly disrupts the telomere architec-
ture of cells resulting in a potent DNA damage response char-
acterized by the formation of several telomeric foci containing
phosphorylated H2AX and 53BP1, which are two of the main
hallmarks of DNA double-strand break. This is typical of the te-
lomere deprotection occurring during cellular senescence or
upon the loss of telomeric proteins.[15,21] Consistently, com-
pound 3 specifically delocalizes POT1 from telomeres, whereas
both TRF1 and TRF2 remain associated to the TTAGGG repeats.
Of note, compound 3 is more potent than the previously de-
scribed 1 in inducing telomere damage and (in agreement
with these results) it is able to limit the growth of transformed
cells at a 5-times-lower drug concentration. The ability of G4 li-
gands to uncap telomeres and to possess antitumoral activity
has been already described for other agents,[6,7,22] reinforcing
the notion that these agents can act as inhibitors of telomere-
related process and therefore the rationale for the develop-
ment of this class of inhibitors as antitumoral agents must be
found elsewhere other than in higher telomerase expression in
cancer cells.
In conclusion, in this paper we have identified the chimera
molecule 3 as a new promising G4 ligand having a strong anti-
tumoral effect, typical of the hydrosoluble coronene deriva-
tives, and an intriguing selectivity toward tumor cells, charac-
teristic of the perylene derivatives, providing a compelling ar-
gument to suggest that telomere is a well-validated target at
the preclinical level.
Experimental Section
Chemistry
General: All commercial reagents and solvents were purchased
from Fluka and Sigma–Aldrich, and used without further purifica-
tion. TLC glass plates (silica gel 60 F254) and silica gel 60 (0.040–
0.063 mm) were purchased from Merck. 1H and 13C NMR spectra
were performed with Varian Gemini 200 and Varian Mercury 300 in-
struments. ESI-MS spectra were recorded on a Micromass Q-TOF
MICRO spectrometer. Starting compounds 4 and 7 were prepared
as previously described[11] (for details, see the Supporting Informa-
tion). All compounds used in the biophysical and biological evalua-
tions were >95% pure as determined by HPLC (instrument: HPLC-
Since telomere is not a selective target for transformed and
tumor cells, the telomere damage and antiproliferative activity
of 3 has been evaluated on normal and transformed cells, with
the aim to verify if this new chimera molecule can maintain
the telomere specificity of the previous reported 1.[8] The re-
sults clearly demonstrated that telomere damage and antiproli-
ChemMedChem 2012, 7, 2144 – 2154
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
2151

A. Biroccio et al.
MED
Waters 2487; column: SUPELCO LC-Diol HPLC Column, 5 mm parti-
cle size, LꢁI.D. 25 cmꢁ4.6 mm).
(broad, 2H, Car-Npiperidine-CH2), 2.84 (broad, 18H, Npiperidine-CH),
1.69 ppm (broad, 24H, CHpiperidine); 13C NMR (300 MHz, CDCl3): d=
163.8 (C=O), 163.7 (C=O), 163.6 (C=O), 163.5 (C=O), 152.8 (Car),
138.7 (Car), 133.1 (Car), 132.4 (Car), 129.0 (Car), 128.9 (Car), 128.4 (Car),
127.2 (Car), 126.7 (Car), 126.1 (Car), 125.6 (Car), 125.0 (Car), 124.9 (Car),
123.3 (Car), 123.0 (Car), 122.9 (Car), 121.9 (Car), 121.7 (Car), 121.6 (Car),
121.2 (Car), 120.2 (Car), 60.3, 56.3, 54.8, 54.7, 54.6, 53.4, 37.9, 37.7,
30.8, 26.0, 25.9, 24.4, 24.3, 23.8 ppm; MS (ESI): m/z: calcd for
C52H59N6O4: 831.4598 [M+H+]; found: 831.4580.
N,N’-Bis[2-(1-piperidino)-ethyl]-1-(1-piperidinyl)-7-bromopery-
lene-3,4:9,10-tetracarboxylic diimide (6): Compound 4 (50 mg,
64.93 mmol; actually a mixture of the two possible isomers; for de-
tails, see the Supporting Information) and hydroquinone (25 mg,
227.27 mmol) were stirred in piperidine (2 mL) and anhydrous diox-
ane (2 mL) at 1008C under argon for 40 min. After cooling, water
was added (20 mL), and the crude product was extracted with
CHCl3 (3ꢁ50 mL). The organic layer was extracted with water until
the aqueous layer was neutral. After drying over Na2SO4, filtering,
and concentration in vacuo, the crude product was purified by
column chromatography on a silica gel (CHCl3/MeOH 98:2). The
complete separation of 5 and 6 was not possible; the fractions ob-
tained by chromatography showing a suitable aromatic pattern in
the 1H NMR spectra (as reported below) were collected, and the
mixture was used in the subsequent step. 1H NMR (200 MHz,
CDCl3): d=9.27 (d, J=8 Hz, 1H, aromatic H), 9.22 (d, J=8 Hz, 1H,
aromatic H), 8.66 (s, 1H, aromatic H), 8.36 (s, 1H, aromatic H), 8.52
(d, J=8 Hz, 1H, aromatic H), 8.40 (d, J=8 Hz, 1H, aromatic H), 4.59
(m, 4H, Nimidic-CH2), 3.37 (m, 2H, Car-Npiperidine-CH2), 2.95 (m, 2H,
Car-Npiperidine-CH2), 3.11 (broad, 12H, Npiperidine-CH2), 1.93 (broad, 8H,
Npiperidine-CH2-CH2), 1.73 (m, 4H, Car-Npiperidine-CH2-CH2), 1.62 (broad,
4H, Npiperidine-CH2-CH2-CH2), 1.48 ppm (broad, 2H, Car-Npiperidine-CH2-
CH2-CH2).
Electrospray ionization mass spectrometry (ESI-MS)
Complexes formed between ligands and quadruplex/duplex DNA
were determined by using ESI-MS. Single-stranded oligonucleo-
tides were purchased from Eurofins MWG Operon (Ebersberg, Ger-
many) with the following sequences: 5’-GGGTTAGGGTTAGGGT-
TAGGGTT-3’ (21-TT) and 5’-CGTAAATTTACG-3’ (DK66). ESI-MS spec-
tra were recorded on a Micromass Q-TOF MICRO spectrometer
(now Waters) in the negative ionization mode. The rate of sample
infusion into the mass spectrometer was 5 mLminÀ1 and the capil-
lary voltage was set to À2.6 kV. The source temperature was ad-
justed to 708C, the cone voltage to 30 V, and the collision energy
to 5 V. Data were analyzed by using the MassLynx software devel-
oped by Waters. Samples were prepared by mixing appropriate
volumes of ammonium acetate buffer (150 mm), annealed oligonu-
cleotide stock solution (50 mm), stock solutions of 3 (100 mm), and
methanol. The final concentration of DNA in each sample was
5 mm (in duplex or quadruplex unit) and the final volume of the
sample was 50 mL. After the binding equilibrium in ammonium
acetate was established, methanol (as 15% w/v) was added to the
mixture just before injection to obtain a stable electrospray signal.
As a reference, samples containing only 5 mm DNA with no drug
were prepared. Samples for competition experiments were pre-
pared following the procedure described above, adding an appro-
priate volume of CT DNA solution. Final concentrations of quadru-
plex DNA and drug solutions were always 5 mm and CT was added
at two different duplex/quadruplex ratios (1 and 5), calculated on
the basis of phosphate group concentrations. To minimize random
errors, each experiment has been repeated at least three times
under the same experimental conditions. Data were processed and
averaged with the SIGMA-PLOT software.
N,N’-Bis[2-(1-piperidino)-ethyl]-1-(1-piperidinyl)-7-[3-(1-piperidi-
no)-butinyl]-perylene-3,4:9,10-tetracarboxylic diimide (8): PP3CBr
6 (2.5 g, 3.23 mmol; mixture of mono and disubstituted deriva-
tives: see above), was dissolved in anhydrous THF (40 mL) and
Et3N (40 mL); then CuI (61 mg, 0.32 mmol) and [Pd(PPh3)4] (367 mg,
0.32 mmol) were added. After bubbling argon through the solu-
tion, the reaction mixture was heated at 808C with stirring and 1-
(3-butynyl)-piperidine 7 (829 mg) was added dropwise. The mixture
was then stirred at 808C overnight in an argon atmosphere. After
cooling, dilute HCl (10 mL) was added and, after neutralization
with 2m aq NaOH, the product was extracted with CH2Cl2 (3ꢁ
50 mL). The organic layer was washed with water until the aque-
ous layer was neutral. After treatment with anhydrous Na2SO4 and
filtration, the solvents were evaporated under vacuum. Since par-
tial cyclization can occur at this stage, complete separation of the
different products and a full characterization were not possible so
the crude product (3 g) was used in the following cyclization step
without further purification.
For drug–DNA complexes with 1:1 and 2:1 stoichiometry, which
have been shown to be the main species present in solution in all
the experiments, the formation of such complexes can be repre-
sented by two distinct equilibrium, which are described by the fol-
lowing two equations: K1 =[1:1]/([DNA] [drug]) and K2 =[2:1]/([1:1]
[drug]), in which [DNA], [drug], [1:1] and [2:1] represent respective-
ly the concentrations of the different species in solution: DNA
(duplex or quadruplex depending on the oligonucleotide used),
the ligand, the 1:1 and 2:1 drug–DNA complexes at equilibrium.
The association constants K1 and K2 can be calculated directly from
the relative intensities of the corresponding peaks found in the
mass spectra, with the assumption that the response factors of the
oligonucleotides alone and of the drug–DNA complexes are the
same, so that the relative intensities in the spectrum are supposed
to be proportional to the relative concentrations in the injected so-
lution.[10] The percentage of bound DNA was calculated according
to an equation developed by Brodbelt and co-workers,[10] which
represents the percentage of DNA bound ligand: Bound DNA
(%)=100([1:1]+[2:1])/([DNA]+[1:1]+[2:1]). To elaborate the data
obtained in the competition experiments, this percentage has
been normalized with respect to the same percentage obtained in
the presence and in the absence of calf thymus DNA, according to
N,N’-Bis[2-(1-piperidino)-ethyl]-1-(1-piperidinyl)-6-[2-(1-piperidi-
no)-ethyl]-benzo[ghi]perylene-3,4:9,10-tetracarboxylic
diimide
(3, EMICORON): Toluene (100 mL) and 1,8-diazabicyclo[5.4.0]undec-
7-ene (DBU) (1.66 mL) were added to the intermediate compound
8 (2.7 g, 3.25 mmol) and the reaction mixture was heated to reflux
with stirring under argon for 20 h. After cooling, CH2Cl2 (100 mL)
was added, and the organic layer was extracted with water until
the aqueous layer was neutral. The crude product was purified by
column chromatography on silica gel (CHCl3/MeOH 100:0, 98:2,
95:5, 90:10, 80:20, and 70:30) to give 1.5 g (55% yield) of the de-
sired compound. The product was then crystallized by dissolving
in a mixture of MeOH and 37% aq HCl solution (95:5) and precipi-
tating the respective hydrochloride with diethyl ether. From this
crystallization, 350 mg of hydrochloride salt was obtained from
1
420 mg of basic compound (71% yield). H NMR (300 MHz, CDCl3):
d=10.30 (d, J=8.7 Hz, 1H, aromatic H), 9.11 (s, 1H, aromatic H),
8.91 (s, 1H, aromatic H), 8.60 (d, J=8.7 Hz, 1H, aromatic H), 8.55 (s,
1H, aromatic H), 8.26 (s, 1H, aromatic H), 4.51 (m, 4H, Nimidic-CH2),
3.70 (broad, 2H, Car-CH2), 3.34 (broad, 2H, Car-Npiperidine-CH2), 2.94
2152
www.chemmedchem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2012, 7, 2144 – 2154

N-Cyclic Bay-Substituted Perylene G-Quadruplex Ligands
MED
the formula: N (%)=quadruplex bound in presence of CT (%)/
quadruplex bound in absence of CT (%).
line were seeded into 60 mm plates and, after 10–12 days, colonies
were stained with 2% methylene blue in 95% ethanol and count-
ed (>50 cells equaled one colony).
Biological evaluation
Statistical analysis: The experiments have been repeated from
three to five times and the results obtained are presented as mean
standard deviations (ÆSD). Significant changes were assessed by
using the Student’s t-test for unpaired data, and P values <0.05
were considered significant.
Cells and culture conditions: BJ fibroblasts expressing hTERT (BJ-
hTERT) or hTERT and SV40 early region (BJ-HELT), M14 melanoma,
HT29 colon and CG5 breast carcinomas were obtained as previous-
ly reported.[7a] Human U2OS osteosarcoma and HCT116 colorectal
carcinoma cell lines were obtained from ATCC; HCT116 p53-/- cells
were a generous gifts from Bert Vogelstein at Johns Hopkins Medi-
cal Institutions, Baltimore, MD, USA. All the lines were grown in
Dulbecco’s Modified Eagle Medium (DMEM, Invitrogen Carlsbad,
CA, USA) supplemented with 10% fetal calf serum, 2 mm l-gluta-
mine and antibiotics.
Acknowledgements
This work was supported by the Italian Ministry of Education,
University and Research (MIUR) Program for Research of National
Interest (PRIN), the Sapienza Universitꢀ di Roma, Italian Associa-
tion for Cancer Research (AIRC) (grant no. 8633 and 11567) and
the Italian Ministry of Health. S.I. is recipient of a fellowship from
the Italian Foundation for Cancer Research (FIRC).
Immunofluorescence: Immunofluorescence was performed as previ-
ously reported.[7a] Cells were fixed in 2% formaldehyde and per-
meabilized in 0.25% Triton X100 in PBS for 5 min at room tempera-
ture. For immunolabeling experiments, cells were incubated with
primary antibody, then washed in PBS and incubated with the sec-
ondary antibodies. The following primary antibodies were used:
pAb and mAb anti-TRF1 (Abcam Ltd.; Cambridge UK); mAb anti-
gH2AX (Upstate, Lake Placid, NY), and pAb anti-53BP1 (Novus Bio-
logicals Inc., Littleton, CO). The following secondary antibodies
were used: TRITC-conjugated goat anti-rabbit and FITC-conjugated
goat anti-mouse (Jackson Lab.). Fluorescence signals were record-
ed by using a Leica DMIRE2 microscope equipped with a Leica
DFC 350FX camera and elaborated by Leica FW4000 deconvolution
software (Leica, Solms, Germany).
Keywords: antitumor agents · biological activity · cancer ·
DNA damage · G-quadruplexes
[1] a) S. Burge, G. N. Parkinson, P. Hazel, A. K. Todd, S. Neidle, Nucleic Acids
Res. 2006, 34, 5402–5415; b) H. J. Lipps, D. Rhodes, Trends Cell Biol.
2009, 19, 414–422.
[2] a) S. Neidle, G. N. Parkinson, Biochimie 2008, 90, 1184–1196; b) D. Yang,
K. Okamoto, Future Med. Chem. 2010, 2, 619–646; c) S. Neidle, G. N. Par-
kinson, Nat. Rev. Drug Discovery 2002, 1, 383–393.
Chromatin Immunoprecipitation assay (ChIP): BJ-HELT fibroblasts
were treated for 72 h with 0.5 mm of compound 3. The ChIP assay
was performed as previously described.[7a] The following antibodies
were used: pAb anti-TRF1 (Santa Cruz Biotechnology, Santa Cruz,
Ca); mAb anti-TRF2 (Imgenex, San Diego, CA); pAb anti-POT1-
(Abcam); mAb anti-b-actin (Sigma, Chemicals, Milano, Italy) was
used as a negative control in the ChIP assay.
[3] S. Neidle, FEBS J. 2010, 277, 1118–1125.
[4] M. Franceschin, Eur. J. Org. Chem. 2009, 2225–2238.
[5] a) D. Drygin, A. Siddiqui-Jain, S. O’Brien, M. Schwaebe, A. Lin, J. Bliesath,
C. B. Ho, C. Proffitt, K. Trent, J. P. Whitten, J. K. Lim, D. Von Hoff, K. An-
deres, W. G. Rice, Cancer Res. 2009, 69, 7653–7661; b) P. J. Bates, D. A.
Laber, D. M. Miller, S. D. Thomas, J. O. Trent, Exp. Mol. Pathol. 2009, 86,
151–164.
[6] a) N. Temime-Smaali, L. Guittat, A. Sidibe, K. Shin-ya, C. Trentesaux, J. F.
Riou, PLoS One 2009, 4, e6919; b) L. R. Gauthier, C. Granotier, F. Hoff-
schir, O. Etienne, A. Ayouaz, C. Desmaze, P. Mailliet, D. S. Biard, F. D.
Boussin, Cell. Mol. Life Sci. 2012, 69, 629–640; c) J. F. Riou, Curr. Med.
Chem. Anticancer Agents 2004, 4, 439–443.
[7] a) E. Salvati, C. Leonetti, A. Rizzo, M. Scarsella, M. Mottolese, R. Galati, I.
Sperduti, M. F. Stevens, M. D’Incalci, M. Blasco, G. Chiorino, S. Bauwens,
B. Horard, E. Gilson, A. Stoppacciaro, G. Zupi, A. Biroccio, J. Clin. Invest.
2007, 117, 3236–3247; b) A. Rizzo, E. Salvati, M. Porru, C. D’Angelo, M. F.
Stevens, M. D’Incalci, C. Leonetti, E. Gilson, G. Zupi, A. Biroccio, Nucleic
Acids Res. 2009, 37, 5353–5364; c) E. Salvati, M. Scarsella, M. Porru, A.
Rizzo, S. Iachettini, L. Tentori, G. Graziani, M. D’Incalci, M. F. Stevens, A.
Orlandi, D. Passeri, E. Gilson, G. Zupi, C. Leonetti, A. Biroccio, Oncogene
2010, 29, 6280–6293.
[8] V. Casagrande, E. Salvati, A. Alvino, A. Bianco, A. Ciammaichella, C.
D’Angelo, L. Ginnari-Satriani, A. M. Serrilli, S. Iachettini, C. Leonetti, S.
Neidle, G. Ortaggi, M. Porru, A. Rizzo, M. Franceschin, A. Biroccio, J. Med.
Chem. 2011, 54, 1140–1156.
[9] M. Franceschin, A. Alvino, V. Casagrande, C. Mauriello, E. Pascucci, M.
Savino, G. Ortaggi, A. Bianco, Bioorg. Med. Chem. 2007, 15, 1848–1858.
[10] a) F. Rosu, E. De Pauw, V. Gabelica, Biochimie 2008, 90, 1074–1087;
b) C. L. Mazzitelli, J. S. Brodbelt, J. T. Kern, M. Rodriguez, S. M. Kerwin, J.
Am. Soc. Mass Spectrom. 2006, 17, 593–604; c) F. Rosu, V. Gabelica, C.
Houssier, P. Colson, E. D. Pauw, Rapid Commun. Mass Spectrom. 2002,
16, 1729–1736; d) V. Casagrande, A. Alvino, A. Bianco, G. Ortaggi, M.
Franceschin, J. Mass Spectrom. 2009, 44, 530–540.
TRAP assay: Telomerase enzyme activity was measured with the
PCR-based TRAP kit (Chemicon International, MA, USA), as previ-
ously reported.[23] Before the PCR step, to purify the elongated
products and remove the bound ligand, extraction with phenol/
chloroform/isoamyl-alcohol (50:49:1) was performed and DNA was
precipitated over night at À208C. Reaction products were ampli-
fied in the presence of a 36 bp ITAS and each set of TRAP assay in-
cluded a control reaction without extract (negative control). Sam-
ples were separated on 12% PAGE and visualized with SYBR Green
staining (Sigma–Aldrich). Gels were quantified using a gel scanner.
The intensity data were obtained by scanning and integrating the
total signal of each PCR product ladder in the denaturing gels.
Drug samples were corrected for background by subtracting the
fluorescence reading of negative controls. The data was collected
at a range of concentrations to obtain dose-response curves from
with the IC50 values (the concentration required for 50% enzyme
inhibition) were obtained.
Cytotoxic assays: Cell growth was assessed by seeding 5ꢁ104 cells
in 60 mm Petri plates (Nunc, MasciaBrunelli, Milano, Italy) and 24 h
after plating, freshly dissolved 3 (0.1 mm) was added to the culture
medium. Cell counts (Coulter Counter, Kontron Instruments,
Milano, Italy) and viability (trypan blue dye exclusion) were deter-
mined daily, from day 1 to day 8 of culture. To evaluate the cell
colony-forming ability, cells were seeded at a density of 2ꢁ105
cells per 60 mm dish and exposed for 96 h to different doses of 3
(0.1 to 1.5 mm). At the end of the treatment, aliquots of each cell
[11] M. Franceschin, A. Alvino, G. Ortaggi, A. Bianco, Tetrahedron Lett. 2004,
45, 9015–9020.
[12] A. Alvino, M. Franceschin, C. Cefaro, S. Borioni, G. Ortaggi, A. Bianco, Tet-
rahedron 2007, 63, 7858–7865.
ChemMedChem 2012, 7, 2144 – 2154
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
2153

A. Biroccio et al.
MED
[13] a) U. Rohr, P. Schlichting, A. Bohm, M. Gross, K. Meerholz, C. Brauchle, K.
Mullen, Angew. Chem. 1998, 110, 1463–1467; Angew. Chem. Int. Ed.
1998, 37, 1434–1437; b) U. Rohr, C. Kohl, K. Mullen, A. van de Craats, J.
Warman, J. Mater. Chem. 2001, 11, 1789–1799; c) K. Sonogashira, Y.
Tohda, N. Hagihara, Tetrahedron Lett. 1975, 16, 4467–4470.
[14] L. Rossetti, M. Franceschin, A. Bianco, G. Ortaggi, M. Savino, Bioorg.
Med. Chem. Lett. 2002, 12, 2527–2533.
[15] B. van Steensel, T. de Lange, Nature 1997, 385, 740–743.
[16] H. Takai, A. Smogorzewska, T. de Lange, Curr. Biol. 2003, 13, 1549–1556.
[17] E. H. Blackburn, Nature 2000, 408, 53–56.
[18] US National Cancer Institute (NCI)/National Institutes of Health (NIH)
Developmental Therapeutics Program: http://dtp.nci.nih.gov (Last ac-
cessed: September 26, 2012).
[19] a) B. Cathers, D. Sun, L. Hurley, Anti-Cancer Drug Des. 1999, 14, 367–
372; b) J. E. Reed, M. Gunaratnam, M. Beltran, A. P. Reszka, R. Vilar, S.
Neidle, Anal. Biochem. 2008, 380, 99–105; c) A. De Cian, G. Cristofari, P.
Reichenbach, E. De Lemos, D. Monchaud, M. P. Teulade-Fichou, K. Shin-
Ya, L. Lacroix, J. Lingner, J. L. Mergny, Proc. Natl. Acad. Sci. USA 2007,
104, 17347–17352.
H. Demir, S. Mazumder, S. Okabe, T. Yamori, M. S. Viapiano, K. Shin-Ya,
H. Seimiya, I. Nakano, Clin Cancer Res. 2012, 18, 1268–80.
[21] a) B. van Steensel, A. Smogorzewska, T. de Lange, Cell 1998, 92, 401–
413; b) K. Ancelin, C. Brun, E. Gilson, Bioessays 1998, 20, 879–883; c) F.
d’Adda di Fagagna, P. M. Reaper, L. Clay-Farrace, H. Fiegler, P. Carr, T.
von Zglinicki, G. Saretzki, N. P. Carter, S. P. Jackson, Nature 2003, 426,
194–198; d) A. Biroccio, A. Rizzo, R. Elli, C. E. Koering, A. Belleville, B. Be-
nassi, C. Leonetti, M. F. Stevens, M. D’Incalci, G. Zupi, E. Gilson, Eur. J.
Cancer 2006, 42, 1881–1888; e) D. Loayza, T. De Lange, Nature 2003,
423, 1013–1018; f) D. Hockemeyer, A. J. Sfeir, J. W. Shay, W. E. Wright, T.
de Lange, EMBO J. 2005, 24, 2667–2678.
[22] a) A. M. Burger, F. Dai, C. M. Schultes, A. P. Reszka, M. J. Moore, J. A.
Double, S. Neidle, Cancer Res. 2005, 65, 1489–1496; b) T. Tauchi, K.
Shin-ya, G. Sashida, M. Sumi, S. Okabe, J. H. Ohyashiki, K. Ohyashiki, On-
cogene 2006, 25, 5719–5792.
[23] A. Biroccio, S. Amodei, B. Benassi, M. Scarsella, A. Cianciulli, M. Motto-
lese, D. Del Bufalo, C. Leonetti, G. Zupi, Oncogene 2002, 21, 3011–3019.
[20] a) G. Pennarun, C. Granotier, L. R. Gauthier, D. Gomez, F. Hoffschir, E.
Mandine, J. F. Riou, J. L. Mergny, P. Mailliet, F. D. Boussin, Oncogene
2005, 24, 2917–2928; b) T. Miyazaki, Y. Pan, K. Joshi, D. Purohit, B. Hu,
Received: July 16, 2012
Published online on October 24, 2012
2154
www.chemmedchem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2012, 7, 2144 – 2154