The number and distances of positive charges of polyamine side chains in a series of perylene diimides significantly influence their ability to induce G-quadruplex structures and inhibit human telomerase

Article abstract of DOI:10.1016/j.bmc.2007.11.065

We have synthesized eight polyamine perylene diimides to conjugate the efficiency of perylene derivatives in stabilizing G-quadruplex structures and the polyamines' biological activity, due to specific interactions with different DNA domains. Our study was carried out by investigating the ability of these derivatives to induce inter- and intramolecular G-quadruplex structures by polyacrylamide gel electrophoresis (PAGE) and to inhibit telomerase in a modified TRAP assay. The two properties appear to be satisfactorily correlated and they show that the number and distances of positive charges in the side chains dramatically influence both these features. Although our previous studies on perylene derivatives with mono-positively charged side chains indicated that self assembly in aqueous solution leads to a major efficiency, the result observed with the spermine derivative suggests that a too strong aggregation is unfavourable, because it determines a lower solubility of the compounds.

Full text of DOI:10.1016/j.bmc.2007.11.065

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry 16 (2008) 2292–2304
The number and distances of positive charges of polyamine
side chains in a series of perylene diimides significantly
influence their ability to induce G-quadruplex structures
and inhibit human telomerase
a,
a,b
b
*
Marco Franceschin, Caterina Maria Lombardo, Emanuela Pascucci,
b
b
a,c
Danilo D’Ambrosio, Emanuela Micheli, Armandodoriano Bianco,
a
b,d
Giancarlo Ortaggi and Maria Savino
a
Dipartimento di Chimica, Universit a` di Roma ‘‘La Sapienza’’, Piazzale A. Moro 5, 00185 Roma, Italy
Dipartimento di Genetica e Biologia Molecolare, Universit a` di Roma ‘‘La Sapienza’’, Piazzale A. Moro 5, 00185 Roma, Italy
Istituto di Chimica Biomolecolare del CNR, Universit a` di Roma ‘‘La Sapienza’’, Piazzale A. Moro 5, 00185 Roma, Italy
Istituto di Biologia e Patologia Molecolare del CNR, Universit a` di Roma ‘‘La Sapienza’’, Piazzale A. Moro 5, 00185 Roma, Italy
b
c
d
Received 8 October 2007; revised 16 November 2007; accepted 23 November 2007
Available online 31 December 2007
Abstract—We have synthesized eight polyamine perylene diimides to conjugate the efficiency of perylene derivatives in stabilizing G-
quadruplex structures and the polyamines’ biological activity, due to specific interactions with different DNA domains. Our study
was carried out by investigating the ability of these derivatives to induce inter- and intramolecular G-quadruplex structures by poly-
acrylamide gel electrophoresis (PAGE) and to inhibit telomerase in a modified TRAP assay. The two properties appear to be sat-
isfactorily correlated and they show that the number and distances of positive charges in the side chains dramatically influence both
these features. Although our previous studies on perylene derivatives with mono-positively charged side chains indicated that self
assembly in aqueous solution leads to a major efficiency, the result observed with the spermine derivative suggests that a too strong
aggregation is unfavourable, because it determines a lower solubility of the compounds.
ꢀ
2007 Elsevier Ltd. All rights reserved.
1
. Introduction
loosing genetic information when cells replicate. Normal
somatic cells do not exhibit telomerase activity and their
DNA reaches a critic length (Hayflick limit ), after
4
Immortality and the related high proliferation of cancer
1
cells is often due to telomerase activation. Telomerase
is a specific reverse transcriptase with an endogenous
RNA template; it has a fundamental role in the mainte-
nance of the eukaryotic chromosomes’ terminal ends,
the telomeres, which are characterized by a repeated
which cellular senescence occurs. The majority of cancer
cells (>85%) show telomerase activity, where the enzyme
lengthens the single-stranded region of the telomere
through addition of discrete repeats of bases, thereby
overcoming the Hayflick limit and giving rise to the cel-
lular immortalization. Due to the selective activation of
telomerase in cancer cells compared to normal cells, the
enzyme has been proposed as target for non-cytotoxic
2
G-rich DNA, specific proteins and a single-stranded
3
terminal region, which shortens when a cellular replica-
tion occurs (‘end replication problem’). DNA polymer-
0
5,6
ase is unable to fully replicate the 3 -end of
chromosomes and so the non-coding telomeric region
of DNA acts to protect chromosomes from the risk of
anticancer agents. One of the strategies to inhibit the
enzyme is the modification of its substrate; since
0
telomeric DNA contains a single-stranded G-rich 3 -
overhang, it is possible to induce the formation of G-
quadruplex structures, in order to keep telomerase from
lengthening chromosome ends.
Keywords: Polyamine side chained perylene derivatives; G-quadruplex;
Telomerase inhibitors.
7
*
Corresponding author. Tel.: +39 0649913341; fax: +39 064462778;
e-mail: marco.franceschin@uniroma1.it
The G-quadruplex is an unusual DNA secondary struc-
ture, based on Hoogsteen G–G pairing, which gives rise
0
968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.11.065

M. Franceschin et al. / Bioorg. Med. Chem. 16 (2008) 2292–2304
2293
to a hydrophobic planar ring of four guanines. The G-
tetrad is stabilized by the co-ordination of a monovalent
O
N
O
O
N
8
,9
cation in the hole ; multiple G-tetrads can stack to-
gether by p–p aromatic interactions to form a complex
quadruple helix, with four negatively charged grooves.
G-quadruplex can adopt different topologies depending
on the glycosidic bond conformation, on the monova-
R
R
O
R=
1
0
11,12
H
N
lent cation and on the oligonucleotide sequence.
These features give rise to many kinds of structures, dif-
ferent in number, orientation, polarity of the chains,
NH2
POL-1
POL-2
POL-3
POL-4
POL-5
POL-6
POL-7
POL-8
N
H
(11)
H
N
1
3
loops and grooves. Many studies showed that G-quad-
ruplex structures appear to be involved in many impor-
tant biological activities, such as DNA replication, gene
NHCOCH
3
N
H
(18)
(7)
H
1
4
N
NH2
expression and recombination, and their presence
1
N
H
5,16
in vivo has been demonstrated
logical systems.
in a number of bio-
H
NHCOCH3
N
N
H
(14)
(9)
G-quadruplex can be stabilized by the presence of small
organic molecules, which consequently act as telomerase
inhibitors. The types of molecules that are able to stabi-
lize G-quadruplexes are characterized by an aromatic
core, that favours stacking interactions with the G-tetr-
ads, and basic side chains (positively charged in physio-
logical conditions), which interact with the quadruple
helix grooves, according to the model of threading inter-
N
H
NH2
N
H
NHCOCH3
N
(16)
(3)
N
1
7,18
calation.
cules with these characteristics have been synthesized:
In order to inhibit telomerase, many mole-
H
1
9,20
19
anthraquinones
are examples of the first generation of these inhibitors.
and 3,6-disubstituted acridines
H
N
N
(5)
2
1
H
Subsequently, tri-substituted acridines and porphy-
2
2
rins have been synthesized to improve the efficiency
in inhibiting telomerase. Natural compounds, such as
Figure 1. Structures of the eight perylene diimides with different
polyamine side chains.
2
3
24
berberine and telomestatin, have also been shown
to inhibit human telomerase, the latter with high effi-
ciency. Perylene diimides present optimal features to
interact with the G-quadruplex and show a good ability
to induce different G-quadruplex structures and to inhi-
bit telomerase, depending on the side chains’ basicity
3
2
cancer targets. In particular, spermine shows a good
efficiency in interacting with G-quadruplex DNA and
triethylenetetramine, a synthetic polyamine, shows the
ability both to stabilize G-quadruplexes structures and
2
5
33
and length. Our research group has recently synthe-
2
to inhibit telomerase. As polycations, these com-
pounds are able to induce the same G-quadruplex struc-
tures as inorganic cations and their interaction with
DNA has been used for the synthesis of functionalized
6
sized new highly hydrosoluble perylene derivatives
2
and coronene derivatives with various side chains.
7
34
3
5
In this work, we present a series of perylene derivatives
with polyamine side chains (Fig. 1), with the aim to eval-
uate the role of the number and the distance of positive
charges in the side chains on G-quadruplex induction
and telomerase inhibition. Many polyamines are well
known, both as natural and synthetic products. Natural
polyamines (spermine, spermidine, putrescine) play a
key role in many genetic processes, such as DNA repli-
perylene derivatives as molecular fluorescent dyes.
Although these studies indicate that polyamines can
interact specifically with different DNA structures, a
study of the influence of the number and distances of
the positive charges in polyamines or polyamine deriva-
tives on their ability to induce G-quadruplexes and inhi-
bit telomerase is still lacking. The eight new polyamine
perylene derivatives hereby reported present different
numbers and distances of positive charges in their side
chains. Our results show that these two parameters are
important in determining the ability of these compounds
to self-assemble in water solution, to induce the
G-quadruplex and to inhibit telomerase.
2
8
cation and gene expression. They are also involved in
cellular growth, proliferation and differentiation. Acety-
lation of polyamines can play many biological roles,
2
9
above all in the regulation of polyamine activity. Nat-
ural polyamine metabolism is strongly regulated because
they can interact with DNA: they are positively charged
in physiological conditions and therefore able to form
specific electrostatic bonds with the DNA phosphate
2. Results and discussion
3
0
backbone. They can interact with A, B and Z DNA,
and have also been demonstrated to induce structural
In this work, the synthesis, the physico-chemical proper-
ties and the biological activity of eight perylene diimides
are reported. These new compounds are characterized
by side chains containing polyamines, which differ for
3
1
transitions among these conformations. Alterations
in their metabolism are involved in cancer development,
so they themselves have been identified as new anti-

2294
M. Franceschin et al. / Bioorg. Med. Chem. 16 (2008) 2292–2304
three features: the number of basic sites, the distance be-
tween them and the nature of the terminal group (pri-
mary, secondary, tertiary amine or amido-group,
Fig. 1).
the presence of a significant amount of the products.
This was because the polyamine side chains are suffi-
ciently polar to give rise to a satisfactory water solubility
of the synthesized compounds, even in these conditions.
In order to eliminate the polyamines that did not react,
the water solution was extracted with chloroform. Final-
ly, the hydrochlorides of compounds 3 and 5 were ob-
tained by adding a concentrated solution of HCl to
the aqueous layer, in order to guarantee the complete
saturation of the basic sites, and precipitating the
respective hydrochlorides with acetone.
2
.1. Chemical synthesis
Polyamines used as reactants are characterized by the
presence of one (2 and 4, Scheme 1) or two (6, 8 and
1
0, Schemes 2 and 3) primary amines as terminal
groups. Therefore, in order to avoid undesired side reac-
tions, two different synthetic schemes for the synthesis of
perylene diimides were carried out, depending on the
nature of the amine used: when only one free primary
amino group was present, Scheme 1 was followed, in
the case of two primary amino groups, Scheme 2 was
adopted, unless one of the two groups was previously
protected, according to Scheme 3.
2.1.2. Scheme 2. The perylene diimides 7, 9 and 11 were
obtained by treatment of the perylene anhydride 1 with
a large excess of polyamines 6, 8 and 10, respectively.
The large excess of reactants avoided the polymerization
which could occur. Compounds 6 and 8 are liquid at
room temperature, so that they assumed the roles of
both reactants and solvents, while, 10 being a solid, a
mixture of DMA and dioxane was used as solvents in
this case. The reaction was conducted under argon
atmosphere, heating under stirring at 100 ꢁC for 24 h,
2
.1.1. Scheme 1. The perylene diimides 3 and 5 were ob-
tained by treatment of the perylene anhydride 1 with an
almost stoichiometric amount of polyamines 2 and 4,
respectively (only 15% excess of reactant was used), in
DMA/dioxane under argon. After heating for 6 h under
reflux, the reaction was quenched by NaOH solution
and the mixture was cooled at 4 ꢁC overnight. The red
precipitate was filtered, washed repeatedly with water
and dried under vacuum to obtain the basic form of
the desired product. Differently from the previously de-
3
6
then at 170 ꢁC for 2 h. The purification of the products
was challenging, because of the physico-chemical prop-
erties of the synthesized compounds, quite similar to
those of the reacting polyamines. After quenching the
crude mixture as described in Scheme 1, solvents were
eliminated by filtration; the residue was then dissolved
in HCl solution and precipitated as a hydrochloride,
by adding acetone. The acetone solution showed a white
opalescence after the complete precipitation of the salts
2
5
scribed preparations with mono-amino side chains, the
remaining solution was strongly coloured, suggesting
O
O
O
O
DMA , dioxane
hreflux, under argon
O
O
6
O
O
N
O
R
R
N
H N
R
2
O
R=
(
1)
N
H
N
(2)
(4)
R=
N
H
N
(3)
(5)
H
N
N
H
H
N
N
H
Scheme 1.
O
O
O
O
O
N
O
O
O
1
. 100°C 24h
. 170°C 2h, under argon
2
N
O
R
NH2
H N
R
2
H N
R
NH2 (excess)
2
O
H
(
1)
R=
N
N
H
(6)
(8)
(10)*
H
N
R=
N
(7)
H
N
H
H
N
H
(9)
N
N
H
N
H
N
(11)
H
*
DMA and dioxane were added
Scheme 2.

M. Franceschin et al. / Bioorg. Med. Chem. 16 (2008) 2292–2304
2295
O
O
O
O
O
O
(
1)
EtOH 96%
4h r.t.
O
2
H N R NHCCH3
H N
R
NH2
2
2
O
OAc
OAc
H
H
OAc
OAc
DMA , dioxane
R=
N
R=
N
N
H
(6)
N
(13)
(15)
(17)
6hreflux
H
under argon
N
(8)
N
OAc
H
H
H
N
H
(
12)
N
N
(10)
N
H
H
O
O
N
O
O
O
N
O
R
NHCCH3
CH CNH
R
3
H
R=
N
N
(14)
(16)
(18)
H
N
H
H
N
N
H
Scheme 3.
of 7, 9 and 11, due to the partial precipitation of the
most soluble hydrochlorides of the residue reacting
amines. Their precipitation is slower than that of the
products, so the final excess of polyamines was elimi-
nated by washing the products with acetone, until the
loss of the opalescence.
2.2. Spectroscopic properties and self-aggregation of the
polyamine perylene derivatives
Absorption spectroscopy can be a good tool to evaluate
the self assembly of polyamine perylene derivatives in
water solution. In fact, the absorption spectra between
350 and 650 nm in MES/KCl buffer solution (pH 6.5)
2
.1.3. Scheme 3. This synthetic pathway was elaborated
give clear indications that all the synthesized compounds
can self-interact in solution. Their extinction coefficients
are characterized in all cases by a strong hypochromic
effect and the bands present a large broadening with
respect to the spectra in organic solvent, as shown in
Figure 2 and in Supplementary data. This indicates that
perylene moieties are stacked in water, as previously dis-
to operate in stoichiometric conditions also with poly-
amines 6, 8 and 10, to avoid the purification difficulties
due to the excess of reactants. The polyamines were
monoacetylated at one of the terminal amino groups,
1
obtaining the N Amonoacetylated compounds 13, 15
and 17, which can react with perylene anhydride 1 under
the same conditions of Scheme 1, to obtain perylene dii-
mides 14, 16 and 18. The acetyl group was chosen as a
2
5
cussed. Surprisingly, the hypochromic effect in water
solution is significantly higher in the case of POL-1
and POL-5 with respect to their monoacetylated deriva-
tives. This suggests that the assumption that the higher
the number of positive charges in the side chains the
greater the water solubility is surely too simple, while
it is necessary to take into account the complexity of
the interactions with water structure. However, as previ-
‘protecting group’ because it is simple to add and be-
cause of the many interesting biological roles of N -
1
8
29
and N -monoacetyl polyamines. The risk of this reac-
1
tion is to obtain the N Adiacetylated compound as the
main product, so that a strictly stoichiometric amount
of the mild acetylant a-D-glucose pentaacetate12 was
used. The reaction was conducted at room temperature
2
5
ously observed, a certain degree of stacking is impor-
tant in determining the ability of these compounds to
induce G-quadruplex structures (see the following sec-
tion), although a quantitative correlation between these
two features (amount of self-stacking and ability of
G-quadruplex induction) was not found so far and the
underlying mechanism requires surely further investiga-
tion. In particular, it has been recently shown by
3
7
for 24 h, under stirring, in 96% ethanol solution. Be-
cause of the presence of a small amount of diacetylated
product and of a discrete amount of the reactants, which
did not react completely in these mild conditions, the
monoacetylated products were purified by column chro-
matography silica
CH CH CH OH/NH = 5:2:2:1).
pounds 13, 15 and 17 obtained were used as reactants
in the reaction with 1, in the same conditions of Scheme
1
on
gel
3
(CHCl /MeOH/
3
The pure com-
7
3
2
2
3
2
5d
Palumbo and coworkers
that strong drug self-aggre-
gation is related to a minor telomerase inhibition and
weaker interactions with the G-quadruplex, suggesting
, obtaining compounds 14, 16 and 18, respectively.

2296
M. Franceschin et al. / Bioorg. Med. Chem. 16 (2008) 2292–2304
3
2
2
1
1
0
5
0
5
0
5
0
2.3. Inter- and intramolecular G-quadruplex DNA
induction
POL-1
POL-3
POL-5
The ability of the polyamine perylene derivatives to in-
duce G-quadruplex structures was investigated by poly-
25
acrylamide gel electrophoresis (PAGE) using the DNA
0
oligomer TSG4 (5 -GGGATTGGGATTGGGATT
0
GGGTT-3 ). TSG4 is able to form stable intramolecular
G-quadruplex and can act as a substrate for telomerase
3
8
elongation in a modified TRAP assay. This oligonu-
2
3
cleotide was used, in our previous works, on berberine,
27
2
5,26
perylene
and coronene derivatives. The oligonu-
cleotide was incubated in the presence of increasing con-
centrations of polyamine perylene derivatives and the
formation of quadruplex structures was investigated
by PAGE analysis as reported in Figure 3. Two different
bands were observed, that were identified as single
strand DNA (ss) and monomeric G-quadruplex (M).
The assignment of the bands was possible by compari-
son with the structures induced by 40 lM PIPER, that
was introduced in each PAGE as an internal standard.
The intramolecular G-quadruplex (M) corresponds to
the band showing the highest mobility; its high folded
structure favours the faster running in the gel grid with
respect to DNA single strand, which has the same
molecular weight. It is worth noting that this increase
of mobility can be less than that expected for naked
DNA, since it is reasonable to suppose that one or more
drug molecules could bind to the G-quadruplex struc-
ture: this explains the observed shift of this band.
3
50
400
450
500
λ (nm)
550
600
650
3
2
2
1
1
0
5
0
5
0
5
0
2
5
POL-2
POL-4
POL-6
The perylene derivatives, indicated as POL-1 and POL-
2, which are characterized by the spermine or monoacet-
ylated spermine as side chains, are unable to induce the
intramolecular G-quadruplex (Fig. 3). In the case of
POL-1, this can be attributed to its low water solubility,
which also gives easily rise to aggregation. All the other
perylene derivatives, except for POL-7, are able to in-
duce the intramolecular G-quadruplex: for these six
compounds we have evaluated the decrease of the free
single-strand DNA at different drug concentrations
3
50
400
450
500
550
600
650
λ (nm)
3
2
2
1
1
0
5
0
5
0
5
0
POL-7
POL-8
(
papers,
Table 1) to compare their efficiency. As in our previous
2
5–27
we preferred this method to the evaluation
of the monomeric G-quadruplex concentration, since this
second approach is affected by a larger error, due to the
binding of the drug. The trend appears to be more favour-
able in the case of POL-3 and POL-5, with respect to their
monoacetylated derivatives (POL-4 and POL-6, respec-
tively) and in the case of POL-8 with the decrease of dis-
tance between charges with respect to POL-7.
3
50
400
450
500
550
600
650
Since the more effective molecules appear also to be able
to induce small amounts of intermolecular G-quadru-
plex, we have carried out similar studies, using an oligo-
λ (nm)
Figure 2. UV–vis absorption spectra of perylene derivatives POL-1/8
in MES–KCl aqueous buffer (pH 6.5).
nucleotide able to form only intermolecular structures,
0
2
HTR (5 -AATCCGTCGAGCAGAGTTAGGGTTA
0
25
GGGTTAG-3 ). All the eight polyamine perylene
derivatives are able to induce intermolecular G-quadru-
plex structures (Fig. 4). The obtained results reported in
Figure 4 show that all the studied molecules are less
efficient in inducing inter- with respect to intramolecular
G-quadruplex, since, in the former case, also at the
highest drug concentration, a substantial amount of
that the higher selectivity for G-quadruplex arrange-
ments upon aggregation is due to a reduced binding effi-
ciency to duplex and single stranded DNA rather than a
greater affinity for G-quartets. We have recently derived
analogous conclusions in
2
derivatives.
a
series of coronene
7

M. Franceschin et al. / Bioorg. Med. Chem. 16 (2008) 2292–2304
2297
Figure 3. G-quadruplex structure formation induced by the eight polyamine side-chain perylene derivatives, studied by native PAGE. (A) Typical
autoradiographies obtained using TSG oligonucleotide (12 lM) in the presence of different drug concentrations: 1 lM (lane a), 2 lM (lane b), 5 lM
lane c and lane 1), 10 lM (lane d and lane 2), 20 lM (lane 3), 40 lM (lane 4) and with no drug (lane 0). As standard in lane P the structures induced
4
(
by PIPER at concentration 40 lM were reported. Major bands were identified as single-stranded DNA (SS), monomeric (M) and intermolecular (I)
G-quadruplex.
Table 1. Percentage of single strand DNA decrease as evaluated by
PAGE with TSG
concentrations
free system by means of a modified Telomerase Repeat
3
8
4
oligonucleotide (Fig. 3) at the indicated
Amplification Protocol (TRAP). Taking into account
that the influence of six of eight molecules in inducing
the intramolecular G-quadruplex appears significantly
larger than in inducing intermolecular G-quadruplex
structures, we have carried out the TRAP assay using
Compound
10 lM
20 lM
POL-3
POL-4
POL-5
POL-6
POL-7
POL-8
55%
56%
32%
70%
86%
47%
30%
55%
25%
26%
68%
17%
2
6,27
TSG4 as a telomerase substrate.
This oligonucleo-
tide is able to fold into the intramolecular G-quadruplex
at the KCl concentration used in the TRAP assay
(68 mM). Nevertheless, a similar structure may be effi-
Errors estimated on at least three independent experiments are about
5%.
ciently unfolded and extended by telomerase, unless a
suitable concentration of G-quadruplex stabilizing mol-
ecule is present. A typical experiment for each molecule
is reported in Figure 5a and the results obtained are re-
ported as an histogram in Figure 5b.
±
single strand DNA is present, and the behaviour of the
eight perylene derivatives appears practically equal in
the limit of experimental errors (Supplementary data).
The results obtained by performing the TRAP assay, in
the absence and in the presence of the polyamine pery-
lene derivatives in the 2–20 lM concentration range,
indicate a large difference in their ability to inhibit
telomerase. Namely POL-3 and POL-5 are effective
2
.4. Biological activity (TRAP assay)
The ability of the eight polyamine perylene derivatives
to inhibit human telomerase was investigated in a cell-

2298
M. Franceschin et al. / Bioorg. Med. Chem. 16 (2008) 2292–2304
Figure 4. G-quadruplex structure formation induced by the eight polyamine side-chain perylene derivatives, studied by native PAGE. (A) Typical
autoradiographies obtained using 2HTR oligonucleotide (12 lM) in the presence of different drug concentrations, as indicated in the legend of Figure
3. Major bands were identified as single-stranded DNA (SS), dimeric (D) and tetrameric (T) G-quadruplex.
telomerase inhibitors with respect to their acetylated
counterparts POL-4 and POL-6, while POL-8 is a better
inhibitor than POL-7. In the case of POL-1 (and analo-
gously POL-2 for comparison) the drug concentration
values were decreased to 2, 5 and 7 lM, for solubility
reasons. POL-2 has surprisingly shown an activity com-
parable to that of POL-3, despite it is acetylated at the
end of the side chain. POL-3 clearly emerges as the most
efficient molecule in inhibiting the enzyme, while POL-5
and POL-8 are both efficient, but at higher drug concen-
tration with respect to POL-3.
seem to be significant in determining the behaviour of
the different compounds are the distances between posi-
tive charges, the charges number and the length of the
side chain. As for the first aspect, the distance corre-
sponding to (CH ) appears to be the optimum size: in
2
2
fact, POL-3 shows the best activity, which decreases if
the distance between N atoms is increased to (CH ) ,
2
3
as in POL-5. This is also shown by the higher activity
of POL-8 with respect to POL-7. However, distance var-
3
9
iation should not be considered a critical parameter,
due to the high flexibility of polyamines. On the other
hand, POL-5 has a basic nitrogen atom less than
POL-3, reducing the number of positive charges in phys-
iological conditions. The importance of positive charge
number also emerges clearly from the strong decrease
of activity if the end of the side chains is acetylated
and thus not able to establish electrostatic interactions
with the DNA phosphates. This is shown by the behav-
iour of POL-4 and POL-6 with respect to POL-3 and
3. Conclusions
The reported results show a strong selectivity of poly-
amine perylene derivatives in inducing G-quadruplex
structures and in inhibiting telomerase in a cell-free sys-
tem. The structural features of the side chains which

M. Franceschin et al. / Bioorg. Med. Chem. 16 (2008) 2292–2304
2299
Figure 5. (A) Inhibition of human telomerase by the eight perylene derivatives by telomerase repeat amplification protocol assay (TRAP). Typical
autoradiographies obtained using TSG4 oligonucleotide in the presence of the indicated drug concentrations. In lane 0 cell extract was not added, in
lane E no drug was added. IS is a 130-bp ‘internal standard’ to control PCR amplification efficiency. (B) Percentage of telomerase inhibition for the
polyamine perylene derivatives at the indicated concentrations. The results of POL-1 are not reported because at concentrations higher than 2 lM
precipitation occurs. Errors estimated on at least three independent experiments are about ±5%.
POL-5. POL-2 shows a surprising behaviour, being al-
most as efficient as POL-3 in terms of activity in the
TRAP assay, despite its terminal acetylation. In princi-
ple, POL-1, carrying the same side chain as POL-2 with-
out the terminal acetyl group, should be even more
active than POL-2. The activity of POL-1 in the TRAP
assay looks very good indeed, since it shows about 30%
of inhibition at 2lM (Fig. 5b). Unfortunately, its poor
solubility and the great tendency to self-aggregate in
water (Fig. 2) cause precipitation of drug and DNA at
higher concentrations (Fig. 5a). It is interesting to note
that POL-1 and POL-2 are not good inducers of the
G-quadruplex, as derived by PAGE studies, but their
ability to inhibit telomerase could be related to their sta-
bilization of the G-quadruplex. In fact, in the modified
TRAP assay a preformed G-quadruplex structure by
TSG4 oligonucleotide is expected to be present.
In the case of G-quadruplex/spermine interactions, re-
4
0
cent studies carried out with NMR showed that this
polyamine interacts strongly with the G-quadruplex
and suggest that its binding could derive from two differ-
ent mechanisms: the spanning of the quadruplex groove
or the interaction with the fold of quadruplex loops.

2300
M. Franceschin et al. / Bioorg. Med. Chem. 16 (2008) 2292–2304
Table 2. Medium number of H-bonds per structure as derived by the
25 lowest energy complexes between the indicated ligands and the
monomeric G-quadruplex (Fig. 6) obtained by simulated annealing
of the studied compounds to interact with the G-
quadruplex structure. In fact, performing simulated
annealing experiments of the analyzed compounds
2
7
4
1
Compound
POL-3
POL-5
POL-7
POL-8
on a human intramolecular G-quadruplex as previ-
2
7
ously described, we estimated the medium number
of H-bonds per structure as derived by the 25 lowest
energy complexes (Table 2). We excluded from this
analysis the acetylated compounds, because of their
poor activity in the PAGE study, and the spermine
containing derivative POL-1, because of its poor sol-
ubility. The values obtained for the analyzed com-
pounds show a good correlation with their efficiency
in inhibiting telomerase. In particular, POL-3, which
is the best efficient compound of this series, shows
a very strong tendency to make hydrogen bonding
with the G-quadruplex structure. In Figure 6, the
ability of POL-3 to make a multiple pattern of H-
bonds is clearly shown, differently from the other
compounds.
Hbonds/structure
5.0
2.9
1.8
2.2
However, it is reasonable to assume that both these
interactions are disfavoured by the conjugation of the
polyamines with the perylene moiety. We suggest that
this conjugation leads to the formation of micellar
aggregates, which are too large in the case of POL-1
to allow the induction of the G-quadruplex, causing also
the poor water solubility of this compound, although
further investigations are necessary to safely assess this
topic.
Moreover, hydrogen bonding is surely another very
important feature in determining the different ability
27
41
Figure 6. Models obtained by molecular modelling for the complexes between a monomeric G-quadruplex (blue) and the perylene ligands POL-3
(A–C) and POL-7 (B–D). Semi-transparent surfaces are in blue for the DNA (A,B), while ligands are atom-type coloured. H-bonds are evidenced as
dotted yellow lines (C,D), with red coloured oxygen atoms on DNA phosphates when involved in hydrogen bonding.

M. Franceschin et al. / Bioorg. Med. Chem. 16 (2008) 2292–2304
2301
From this research, POL-3 combines a good water solu-
bility, an efficient ability to induce G-quadruplex struc-
tures and an activity of telomerase inhibition under
(0.97 g) and the hydrochloride (260 mg) of compound
5 were obtained. The overall yield was 73%.
1
5
lM, improving all these aspects with respect to the
previously synthesized perylene diimides, as in the case
H NMR (300 MHz, CF COOD): d 9.24 (4H, d,
3
2
5
J = 8 Hz, aromatic H), 9.14 (4H, d, J = 8 Hz, aromatic
H), 5.18 (4H, m, NimidicACH ), 4.3–4.1 (12H, m, N
2
6
of three and four side-chained perylene derivatives.
2
aminic
For these reasons, POL-3 clearly emerges to be consid-
ered as the best polyamine perylene diimide, which
surely deserves further testing in biological systems.
These studies have been currently carried out in our
laboratories.
ACH ), 3.96 (2H, m, J = 6 Hz, N
ACH (CH ) ),
2
aminic
3 2
1
3
1.78 (12H, d, J = 6 Hz, ACH ) ppm.
C NMR
3
(CF COOD): d 162.7 (C@O), 132.9 (ar), 129.2 (ar),
3
125.5 (ar), 122.5 (ar), 120.5 (ar), 117.7 (ar), 50.3, 45.1,
41.7, 38.2, 33.4, 13.4. MS (ESI) m/z: 647.3345
+
(M+H) ] (calcd for C H N O : 647.3346).
[
3
8
43
6
4
4. Experimental
4.1.2. Synthesis of compounds 7, 9 and 11: General
procedure (Scheme 2). Perylene anhydride (1) was trea-
ted with a large excess of the appropriate polyamine
4
.1. Chemistry
(
6, 8 or 10): when it is liquid (6 and 8), the amine could
All the commercial reagents and solvents were pur-
chased from Aldrich. TLC plates (silica gel 60 F254
be used as a reactant and a solvent, when solid (10),
DMA and dioxane were added. The mixture was heated
at 100 ꢁC for 24 h, then at 170 ꢁC for 2 h. The crude
product was treated with a mixture of diethyl ether/1-
propanol = 4:1, so that a colloidal precipitate was ob-
tained, which was filtered under vacuum. It was dis-
solved in HCl solution, then hydrochlorides of the
products were precipitated by adding acetone and they
were purified by washing many times with acetone, to
remove the hydrochloride of the reactant polyamine.
)
and silica gel 60 (0.063–0.200 mm) for column
chromatography were purchased from Merck. NMR
spectra were obtained with Varian Gemini 200 and
Varian Mercury 300 instruments. High resolution
ESI-MS spectra were recorded on Micromass
Q-TOF MICRO spectrometer. Elemental analyses
3
6
(
(
C, H, N) were carried out on EA1110 CHNS-O
CE instruments).
0
4
(
.1.1. Synthesis of compounds 3 and 5: General procedure
Scheme 1). Perylene anhydride (1) was treated with the
4.1.2.1. Synthesis of N,N -bis(8-amino-3,6-diazaoctyl)-
3,4,9,10-perylenetetracarboxylic diimide (7). Compound
1 (98 mg) was treated with 2.0 ml of triethylenetetramine
(6), according to the general procedure described, so
that 96 mg of the hydrochloride of compound 7 were
obtained, with a yield of 44%.
appropriate polyamine (2 or 4), in a refluxing mixture of
N,N-dimethylacetamide (DMA, 15 ml) and 1,4-dioxane
2
5
(
15 ml) for 6 h. The basic forms of the products (3
and 5, respectively) were precipitated by adding a di-
luted solution of aqueous NaOH, separated by filtration
and dried under vacuum. The solution was extracted
with chloroform and the aqueous layer was acidified
with concentrated HCl. Hydrochlorides of the products
were precipitated with acetone.
1
H NMR (300 MHz, CF COOD): d 9.19 (4H, d,
3
J = 8 Hz, aromatic H), 9.09 (4H, d, J = 8 Hz, aromatic
H), 5.12 (4H, m, N
ACH ), 4.3–4.0 (20H, br, N
imidic
2
aminic
1
3
ACH ) ppm. C NMR (CF COOD): d 162.8 (C@O),
2
3
132.7 (ar), 129.4 (ar), 125.6 (ar), 122.7 (ar), 120.5 (ar),
117.8 (ar), 45.1, 41.9, 41.5, 41.3, 33.9, 33.3. MS (ESI)
0
4
.1.1.1. Synthesis of N,N -bis(7-dimethylamino-4-aza-
+
m/z: 649.3273 [(M+H) ] (calcd for C H N O :
649.3251).
heptyl)-3,4,9,10-perylenetetracarboxylic diimide (3).
Compound 1 (1.02 g) was treated with 1 ml of N,N-dim-
ethyldipropylenetriamine (2), according to the general
procedure described, so that both the basic form
462 mg) and the hydrochloride (1.38 g) of compound
were obtained. The overall yield was 92%.
3
6
41
8
4
0
4.1.2.2. Synthesis of N,N -bis(7-amino-4-azaheptyl)-
3,4,9,10-perylenetetracarboxylic diimide (9). Compound
1 (107 mg) was treated with 2.0 ml of bis(3-aminopro-
pyl)amine (8), according to the general procedure de-
scribed, so that 102 mg of the hydrochloride of
compound 9were obtained, with a yield of 48%.
(
3
1
H NMR (300 MHz, CF COOD): d 9.20 (4H, d,
3
J = 8 Hz, aromatic H), 9.15 (4H, d, J = 8 Hz, aromatic
H), 4.88 (4H, m, NimidicACH ), 3.81 (12H, m, N
2
aminic
1
ACH ), 3.44 (12H, s, NACH ), 3.0–2.8 (8H, br, CH A
H NMR (300 MHz, CF COOD): d 9.00 (8H, m, aro-
2
3
2
3
1
3
CH ANaminic) ppm. C NMR (CF COOD): d 162.3
matic H), 4.71 (4H, m, NimidicACH ), 3.68 (12H, m,
2
3
2
(
1
1
C@O), 132.4 (ar), 129.2 (ar), 125.4 (ar), 122.5 (ar),
NaminicACH ), 2.9–2.5 (8H, br, CH ACH AN
)
aminic
2
2
2
1
3
20.5 (ar), 117.8 (ar), 51.4, 42.8, 41.6, 39.4, 33.7, 20.6,
8.0. MS (ESI) m/z: 675.3663 [(M+H) ] (calcd for
ppm. C NMR (CF COOD): d 161.7 (C@O), 131.7
3
+
(ar), 128.5 (ar), 124.8 (ar), 121.8 (ar), 119.7 (ar), 117.1
(ar), 41.9, 41.0, 33.3, 32.8, 19.9, 19.2. MS (ESI) m/z:
C H N O : 675.3659).
4
0
47
6
4
+
19.3054 [(M+H) ] (calcd for C H N O : 619.3033).
6
3
6
39
6
4
0
.1.1.2. Synthesis of N,N -bis(5-isopropylamino-3-aza-
4
0
4.1.2.3. Synthesis of N,N -bis(12-amino-4,9-diazadode-
cyl)-3,4,9,10-perylenetetracarboxylic diimide (11). Com-
pound 1 (72 mg) was treated with 0.82 g of spermine
(10), according to the general procedure described, in
pentyl)-3,4,9,10-perylenetetracarboxylic diimide1 (5).
Compound 1 (0.99 g) was treated with 1.1 ml of N Aiso-
propyldiethylenetriamine (4), according to the general
procedure described, so that both the basic form

2302
M. Franceschin et al. / Bioorg. Med. Chem. 16 (2008) 2292–2304
2
1
ml of anhydrous dioxane and 2 ml of DMA, so that
11 mg of the hydrochloride of compound 11 was ob-
both the basic form (124 mg) and the hydrochloride
(161 mg) of compound 14 were obtained. The overall
yield was 43%.
tained, with a yield of 62%.
1
1
H NMR (300 MHz, CF COOD): d 9.13 (8H, m, aro-
H NMR (300 MHz, CF COOD): d 9.17 (8H, m, aro-
3
3
matic H), 4.97 (4H, m, NimidicACH ), 3.86 (20H, m,
matic H), 5.15 (4H, m, NimidicACH ), 4.9–4.0 (20H,
2
2
NaminicACH ), 2.9–2.5 (16H, br, CH ACH ANaminic)
br, NaminicACH2, NamidicACH2), 2.67 (6H, s,
2
2
2
1
3
13
ppm. C NMR (CF COOD): d 161.4 (C@O), 131.5
CH AC@O) ppm. C NMR (CF COOD): d 174.4
3
3
3
(
(
ar), 128.5 (ar), 124.6 (ar), 121.7 (ar), 119.7 (ar), 117.2
ar), 43.8, 41.7, 41.2, 35.8, 33.4, 20.1, 19.3, 18.5. Elemen-
(C@O), 162.5 (C@O), 132.5 (ar), 129.3 (ar), 125.5 (ar),
122.5 (ar), 120.4 (ar), 117.7 (ar), 46.1, 41.7, 41.0, 33.2,
tal analysis: C H N O (6HCl): calcd C 53.9%, N
4
33.0, 16.1 (CH AC@O). MS (ESI) m/z: 367.1751
4
4
56
8
3
+
+
1
1.4%, H 6.3%; found C 52.8%, N 10.8%, H 6.2%.
[(M+2H) ] (calcd for C H N O : 367.1770).
20 23 4 3
0
4
.1.3. Synthesis of reactants 13, 15 and 17: General
4.1.4.2. Synthesis of N,N -bis(7-acetamido-4-azahep-
tyl)-3,4,9,10-perylenetetracarboxylic diimide (16). Com-
pound 1 (200 mg) was treated with 190 mg of 15,
according to the general procedure described, so that
both the basic form (189 mg) and the hydrochloride
(90 mg) of compound 16 was obtained. The overall yield
was 38%.
procedure. a-D-Glucose pentaacetate (12) was dissolved
in 40 ml of 96% ethanol, under stirring in conditions
of mild heating, and then were gradually added to the
opportune polyamine (6, 8 or 10). The crude product
is a light yellow dense liquid, which was purified by col-
umn chromatography on silica gel (CHCl /MeOH/
3
CH CH CH OH/NH = 5:2:2:1), obtaining the mono-
3
acetylated products (13, 15 and 17).
3
7
2
2
3
1
H NMR (300 MHz, CF COOD): d 9.05 (4H, d,
3
J = 8 Hz, aromatic H), 8.99 (4H, d, J = 8 Hz, aromatic
H), 4.68 (4H, m, NimidicACH ), 3.83 (4H, m, N
1
4
.1.3.1. Synthesis of the reactant N Aacetyltriethylen-
etetramine (13). Compound 12 (0.78 g) was mixed with
.5 ml of 6, according to the general procedure de-
2
amidic
ACH ), 3.53 (8H, m, N
ACH ), 2.7–2.3 (14H, br,
2
aminic
2
13
1
CH ACH AN
(CF COOD): d 173.4 (C@O), 162.2 (C@O), 132.3 (ar),
, CH AC@O) ppm.
C NMR
2
2
aminic
3
scribed, so that 0.82 g of reactant 13 were obtained, with
a yield of 44%.
3
129.0 (ar), 125.3 (ar), 122.3 (ar), 120.2 (ar), 117.6 (ar),
4
MS (ESI) m/z: 703.3244 [(M+H) ] (calcd for
2.2, 41.7, 33.2, 33.1, 21.0, 20.4, 15.5 (CH AC@O).
3
1
+
H NMR (200 MHz, CDCl ) : d 3.35 (2H, m, N
3
amidic
ACH ), 2.9–2.6 (10H, br, N
ACH ), 1.95 (3H, s,
C H N O : 703.3244).
40 43
2
aminic
2
6
6
CH AC@O) ppm.
3
0
.1.4.3. Synthesis of N,N -bis(12-acetamido-4,9-
4
1
4.1.3.2. Synthesis of the reactant N Aacetyldipropy-
diazadodecyl)-3,4,9,10-perylenetetracarboxylic diimide
(18). Compound 1 (100 mg) were treated with 136 mg
of 17, according to the general procedure described, so
that 121 mg of the hydrochloride of compound 18 were
obtained, with a yield of 48%.
lenetriamine (15). Compound 12 (0.78 g) was mixed with
.4 ml of 8, according to the general procedure de-
1
scribed. The crude product was purified as described be-
fore, so that 0.51 g of reactant 15 were obtained, with a
yield of 28%.
1
H NMR (300 MHz, CF COOD): d 9.13 (8H, m, aro-
3
1
H NMR (200 MHz, CDCl ) : d 3.45 (2H, m, N
matic H), 4.76 (4H, m, NimidicACH ), 3.85 (4H, m, N
3
amidic
2
amidic
ACH ), 2.9–2.6 (6H, br, N
ACH ), 1.90 (3H, s,
ACH ), 3.56 (16H, m, N
ACH ), 2.8–2.2 (22H, br,
2
aminic
2
2
aminic
2
13
CH AC@O), 1.8–1.5 (4H, br, CH ACH ANaminic) ppm.
CH ACH AN
, CH AC@O) ppm.
C NMR
3
2
2
2
2
aminic
3
(
CF COOD): d 173.8 (C@O), 162.2 (C@O), 132.3 (ar),
3
1
4.1.3.3. Synthesis of the reactant N Aacetylspermine
17). Compound 12 (0.76 g) was mixed with 2.01 g of 10,
129.1 (ar), 125.5 (ar), 122.4 (ar), 120.3 (ar), 117.7 (ar),
44.0, 42.3, 41.7, 33.4, 33.1, 21.0, 20.4, 18.9, 15.5
(
++
according to the general procedure described. The crude
product was purified as described before, so that 0.80 g
of reactant 17 were obtained, with a yield of 33%.
(CH AC@O). MS (ESI) m/z: 423.2380 [(M+2H)
]
3
(calcd for C H N O : 423.2396).
2
4
31
4
3
4
.2. UV–vis absorption spectroscopy
1
H NMR (200 MHz, CDCl ): d 3.35 (2H, m, N
3
amidic
ACH ), 2.9–2.6 (10H, br, N
ACH ), 1.95 (3H, s,
UV–vis absorption spectra were obtained using a JAS-
CO V-530 spectrophotometer. They were registered be-
tween 350 and 650 nm in quartz cuvettes. Four
hundred micromolar drug stocks, prepared by dissolv-
ing the hydrochlorides in DMSO, were diluted to ob-
serve spectra in aqueous solution (10 mM MES–
50 mM KCl buffer, pH 6.5).
2
aminic
2
CH AC@O), 1.8–1.5 (8H, br, CH ACH ANaminic) ppm.
3
2
2
4
.1.4. Synthesis of compounds 14, 16 and 18: General
procedure (Scheme 3). Perylene anhydride 1 was treated
with the opportune monoacetyl-polyamine (13, 15 or
1
7), following the same procedure of Scheme 1.
0
4
.1.4.1. Synthesis of N,N -bis(8-acetamido-3,6-dia-
4.3. Polyacrylamide gel electrophoresis (PAGE)
zaoctyl)-3,4,9,10-perylenetetracarboxylic diimide (14).
Compound 1 (312 mg) was treated with 331 mg of 13,
according to the general procedure described, so that
Previously 32P radio-labelled 2HTR and TSG4 oligonu-
cleotides were heated at 95 ꢁC for 10 min and quickly

M. Franceschin et al. / Bioorg. Med. Chem. 16 (2008) 2292–2304
2303
cooled in ice, at a concentration of 12 lM. Then they
were incubated for 2 h at 30 ꢁC in MES–KCl buffer
3. Meyne, J.; Ratliff, R. L.; Moyzis, R. K. Proc. Natl. Acad.
Sci. U.S.A. 1989, 86, 7049.
4
5
6
7
. Shay, J. W.; Wright, W. E. Nat. Rev. Mol. Cell. Biol. 2000,
, 72.
. White, L. K.; Wright, W. E.; Shay, J. W. Trends
Biotechnol. 2001, 19, 114.
. Incles, C. M.; Schultes, C. M.; Neidle, S. Curr. Opin.
Investig. Drugs 2003, 4, 675.
. Neidle, S.; Parkinson, G. N. Nat. Rev. Drug Discov. 2002,
1, 383.
(
10 mM MES, pH 6,5, 50 mM KCl for 2HTR and
mM KCl for TSG4) in the presence of different drug
1
5
concentrations and with no drug. Complexes and struc-
tures formed after incubation were studied by native
PAGE (15% polyacrylamide gel, TBE 0.5·, KCl
2
0 mM, run overnight at room temperature). In all elec-
trophoresis runs, the G-quadruplex induction by PIPER
(
40 lM) was also reported as a useful standard to assign
8. Williamson, J. R. Curr. Opin. Struct. Biol. 1993, 3, 335.
9. Neidle, S.; Parkinson, G. N. Curr. Opin. Struct. Biol. 2003,
the electrophoresis bands to different DNA conforma-
tions. Major bands were so identified as single-stranded
DNA (ss), dimeric (D), tetrameric (T) and monomeric
1
3, 275.
0. Wang, Y.; Patel, D. J. Structure 1993, 1, 263.
11. Haider, S. M.; Parkinson, G. N.; Neidle, S. J. Mol. Biol.
002, 320, 189.
2. Crnugelj, M.; Hud, N. V.; Plavec, J. J. Mol. Biol. 2002,
20, 911.
3. Rhodes, D.; Giraldo, R. Curr. Opin. Struct. Biol. 1995, 5,
11.
1
2
5
(
M) G-quadruplex structures.
2
1
1
4.4. Telomerase repeat amplification protocol (TRAP)
assay
3
3
The reaction mixture for assaying inhibition of human
telomerase (50 ll) contains 50 lM dNTPs, 0.4 lM
14. Ghosal, G.; Muniyappa, K. Biochem. Biophys. Res.
Commun. 2006, 343, 1.
15. Schaffitzel, C.; Berger, I.; Postberg, J.; Hanes, J.; Lipps, H.
J.; Pluckthun, A. Proc. Natl. Acad. Sci. U.S.A. 2001, 98,
TSG4 primer and 1 ll of cell extract (prepared from
7
0 cultured HeLa cells) in TRAP buffer (20 mM Tris–
1
HCl, pH 8.3, 1.5 mM MgCl , 68 mM KCl, 0.05% Tween
8
6. Paeschke, K.; Simonsson, T.; Postberg, J.; Rhodes, D.;
Lipps, H. J. Nat. Struct. Mol. Biol. 2005, 12, 847.
7. Haider, S. M.; Parkinson, G. N.; Neidle, S. J. Mol. Biol.
572.
2
1
1
1
2
0 and 1 mM EGTA). In each sample, polyamine pery-
lene derivatives were added at different concentrations
and incubated for 2 h at 30 ꢁC, before addition of the
cell extract. After 30 min of incubation at 30 ꢁC, the
samples were purified by phenol/chloroform extraction.
2
003, 326, 117.
8. Clarck, G. R.; Pytel, P. D.; Squire, C. J.; Neidle, S. J. Am.
Chem. Soc. 2003, 125, 4066.
3
2
P radiolabelled TSG4 (0.14 lM), 0,4 lM CXext pri-
mer and 2U Taq DNA polymerase (Eppendorf) were
19. Sun, D.; Thompson, B.; Cathers, B. E.; Salazar, M.;
Kerwin, S. M.; Trent, J. O.; Jenkins, T. C.; Neidle, S.;
Hurley, L. H. J. Med. Chem. 1997, 40, 2113.
0
added and 30 PCR cycles were performed (94 ꢁC 10 ,
9
were loaded on a nondenaturing 12% polyacrylamide
gel. Samples with no drug and with no cell extract were
used as references. A 130-bp ‘internal standard’ (IS) was
00
00
0
00
2
0. Huang, H. S.; Chou, C. L.; Guo, C. L.; Yuan, C. L.; Lu,
Y. C.; Shieh, F. Y.; Lin, J. J. Bioorg. Med. Chem. 2005, 13,
2 ꢁC 30 , 54 ꢁC 30 , 72 ꢁC 5 30 ). Finally, the samples
1
435.
2
1. Read, M. A.; Harrison, R. J.; Romagnoli, B.; Tanious, F.
A.; Gowan, S. H.; Reszka, A. P.; Wilson, W. D.; Kelland,
L. R.; Neidle, S. Proc. Natl. Acad. Sci. U.S.A. 2001, 98,
3
8
used to evaluate PCR amplification efficiency.
4
844.
2
2
2. Seenisamy, J.; Bashyam, S.; Gocale, V.; Vankayalapati,
H.; Sun, D.; Siddiqui-Jain, A.; Streiner, N.; Shin-Ya, K.;
White, E.; Whilson, W. D.; Hurley, L. H. J. Am. Chem.
Soc. 2005, 127, 2944.
3. Franceschin, M.; Rossetti, L.; D’Ambrosio, A.; Schirripa,
S.; Bianco, A.; Ortaggi, G.; Savino, M.; Schultes, C.;
Neidle, S. Bioorg. Med. Chem. Lett. 2006, 16, 1707.
4. Kim, M. Y.; Vankayalapati, H.; Shin-ya, K.; Wierzba, K.;
Hurley, L. H. J. Am. Chem. Soc. 2001, 124, 2098.
5. (a) Fedoroff, O. Y.; Salazar, M.; Han, H.; Chemeris, V. V.;
Kerwin, S. M.; Hurley, L. H. Biochemistry 1998, 37,
Acknowledgments
Thanks are due to Antonello Alvino for the MS charac-
terization of the synthesized compounds, to F. Piccioni
for the NMR spectra, to S. Bacchetti for the protocol
of telomerase extraction and to Julie E. Reed for help
in the final revision of the manuscript. The research
has been financially supported by MIUR FIRB 2001,
MIUR COFIN 2005 and by Istituto Pasteur Fondazi-
one Cenci Bolognetti.
2
2
1
2367; (b) Rossetti, L.; Franceschin, M.; Bianco, A.;
Ortaggi, G.; Savino, M. Bioorg. Med. Chem. Lett 2002, 12,
527; (c) Rossetti, L.; Franceschin, M.; Schirripa, S.;
2
Supplementary data
Bianco, A.; Ortaggi, G.; Savino, M. Bioorg. Med. Chem.
Lett 2005, 15, 413; (d) Sissi, C.; Lucatello, L.; Paul
Krapcho, A.; Maloney, D. J.; Boxer, M. B.; Camarasa, M.
V.; Pezzoni, G.; Menta, E.; Palumbo, M. Bioorg. Med.
Chem. 2007, 15, 555.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmc.
2
007.11.065.
2
6. (a) Alvino, A.; Franceschin, M.; Cefaro, C.; Borioni, S.;
Ortaggi, G.; Bianco, A. Tetrahedron 2007, 63, 7858; (b)
Franceschin, M.; Pascucci, E.; Alvino, A.; D’Ambrosio,
D.; Bianco, A.; Ortaggi, G.; Savino, M. Bioorg. Med.
Chem. Lett. 2007, 17, 2515.
27. Franceschin, M.; Alvino, A.; Casagrande, V.; Mauriello,
C.; Pascucci, E.; Savino, M.; Ortaggi, G.; Bianco, A.
Bioorg. Med. Chem. 2007, 15, 1848.
References and notes
. Shay, J. W.; Wright, W. E. Cancer cell 2002, 2, 257.
. Cong, Y. S.; Wright, W. E.; Shay, J. W. Microbiol. Mol.
Bio. Rev. 2002, 66, 407.
1
2

2304
M. Franceschin et al. / Bioorg. Med. Chem. 16 (2008) 2292–2304
2
2
8. J a¨ nne, J.; Alhonen, L.; Pietil a¨ , M.; Kein a¨ nen, T. A. Eur. J.
Biochem. 2004, 271, 877.
9. Mudumba, S.; Menezes, A.; Fries, D.; Blankenship, J.
Biochem. Pharmacol. 2002, 63, 2011.
0. Keniry, M. A. FEBS Lett. 2003, 542, 153.
36. Tuntiwechapikul, W.; Taka, T.; B e´ tencourt, M.; Mak-
onkawkeyoon, L.; Randall, L. T. Bioorg. Med. Chem.
Lett. 2006, 16, 4120.
37. Kuswik-Rabiega, G.; Bruenger, F. W.; Miller, S. C.
Synthetic Commun. 1992, 22, 1307.
38. Gomes, D.; Mergny, J. L.; Riou, J. F. Cancer Res. 2002,
62, 3365.
39. Liquori, A. M.; Costantino, L.; Crescenzi, V.; Elia, V.;
Giglio, E.; Puliti, R.; de Santis Savino, M.; Vitagliano, V.
J. Mol. Biol. 1967, 24, 113.
3
3
3
3
1. Thomas, T. J.; Messner, R. P. J. Mol. Biol. 1988, 201, 463.
2. Thomas, T.; Thomas, T. J. J. Cell. Mol. Med. 2003, 7, 113.
3. Yin, F.; Liu, J.; Peng, X. Bioorg. Med. Chem. Lett. 2003,
1
3, 3923.
4. Miyoshi, D.; Nakao, A.; Sugimoto, N. Biochemistry 2002,
1, 15017.
5. Krauß, S.; Lysetska, M.; W u¨ rthner, F. Lett. Org. Chem.
005, 2, 349.
3
3
4
40. Keniry, M. A. FEBS Lett. 2003, 542, 153.
41. Parkinson, G. N.; Lee, M. P.; Neidle, S. Nature 2002, 417,
876.
2