Rational design, synthesis, and DNA binding properties of novel sequence-selective peptidyl congeners of ametantrone

Article abstract of DOI:10.1002/cmdc.201000106

Natural and synthetic compounds characterized by an anthraquinone nucleus represent an important class of anti-neoplastic agents, the mechanism of action of which is related to intercalation into DNA. Ametantrone (AM) is a synthetic 9,10-anthracenedione bearing two (hydroxyethylamino)ethylamino residues at positions 1 and 4; along with other anthraquinones and anthracyclines, it shares a polycyclic intercalating moiety and charged side chains that stabilize DNA binding. All these drugs elicit adverse side effects, which represent a challenge for antitumor chemotherapy. In the present work the structure of AM was augmented with appropriate groups that target well-defined base pairs in the major groove. These should endow AM with DNA sequence selectivity. We describe the rationale for the synthesis and the evaluation of activity of a new series of compounds in which the planar anthraquinone is conjugated at positions 1 and 4 through the side chains of AM or other bioisosteric linkers to appropriate dipeptides. The designed novel AM derivatives were shown to selectively stabilize two oligonucleotide duplexes that both have a palindromic GC-rich hexanucleotide core, but their stabilizing effects on a random DNA sequence was negligible. In the case of the most effective compound, the 1,4-bis-[Gly-(l-Lys) ] derivative of AM, the experimental results confirm the predictions of earlier theoretical computations. In contrast, AM had equal stabilizing effects on all three sequences and showed no preferential binding. This novel peptide derivative can be classified as a strong binder regarding the sequences that it selectively targets, possibly opening the exploitation of less cytotoxic conjugates of AM to the targeted treatment of oncological and viral diseases.

Full text of DOI:10.1002/cmdc.201000106

MED
DOI: 10.1002/cmdc.201000106
Rational Design, Synthesis, and DNA Binding Properties of
Novel Sequence-Selective Peptidyl Congeners of
Ametantrone
Alessandra Gianoncelli,[a] Serena Basili,[a] Matteo Scalabrin,[a] Alice Sosic,[a] Stefano Moro,[a]
Giuseppe Zagotto,[a] Manlio Palumbo,[a] Nohad Gresh,[b] and Barbara Gatto*[a]
Natural and synthetic compounds characterized by an anthra-
quinone nucleus represent an important class of anti-neoplas-
tic agents, the mechanism of action of which is related to in-
tercalation into DNA. Ametantrone (AM) is a synthetic 9,10-an-
thracenedione bearing two (hydroxyethylamino)ethylamino
residues at positions 1 and 4; along with other anthraquinones
and anthracyclines, it shares a polycyclic intercalating moiety
and charged side chains that stabilize DNA binding. All these
drugs elicit adverse side effects, which represent a challenge
for antitumor chemotherapy. In the present work the structure
of AM was augmented with appropriate groups that target
well-defined base pairs in the major groove. These should
endow AM with DNA sequence selectivity. We describe the ra-
tionale for the synthesis and the evaluation of activity of a
new series of compounds in which the planar anthraquinone is
conjugated at positions 1 and 4 through the side chains of AM
or other bioisosteric linkers to appropriate dipeptides. The de-
signed novel AM derivatives were shown to selectively stabilize
two oligonucleotide duplexes that both have a palindromic
GC-rich hexanucleotide core, but their stabilizing effects on a
random DNA sequence was negligible. In the case of the most
effective compound, the 1,4-bis-[Gly-(l-Lys)] derivative of AM,
the experimental results confirm the predictions of earlier the-
oretical computations. In contrast, AM had equal stabilizing ef-
fects on all three sequences and showed no preferential bind-
ing. This novel peptide derivative can be classified as a strong
binder regarding the sequences that it selectively targets, pos-
sibly opening the exploitation of less cytotoxic conjugates of
AM to the targeted treatment of oncological and viral diseases.
Introduction
Anthraquinones have been used
as efficient compounds in treat-
ing several kinds of cancers. The
discovery of the antitumor activi-
ty of 1,4-bis[(aminoalkyl)ami-
no]anthracene-9,10-diones such
as ametantrone (AM) and mito-
xantrone (MX) (Figure 1) has led
to numerous studies on their tu-
moricidal mechanisms and on
those of their congeners.[1] They
are classic intercalating agents
that accumulate in the nucleus,
Figure 1. Structures of ametantrone (3), mitoxantrone, and tested peptidyl-anthraquinones.
where they poison topoisomera-
se II, and there is good evidence
that the impaired topoisomera-
[a] Dr. A. Gianoncelli, Dr. S. Basili, Dr. M. Scalabrin, Dr. A. Sosic, Prof. S. Moro,
Prof. G. Zagotto, Prof. M. Palumbo, Prof. B. Gatto
Dipartimento di Scienze Farmaceutiche, Universitꢀ di Padova
Via Marzolo 5, 35131 Padova (Italy)
se II (i.e., a stabilized DNA–topoisomerase II cleavable complex)
contributes to the cytotoxic effect.[2] In addition, it has also
been demonstrated that the oxidation products of MX, cata-
lyzed by horseradish peroxidase, form a covalent complex with
DNA in vitro,[3] and similar drug–DNA adducts were also detect-
ed when the drug was oxidized in vitro by the human enzyme
myeloperoxidase.[4]
Fax: 0039-049-8275366
E-mail: barbara.gatto@unipd.it
[b] Prof. N. Gresh
Laboratoire de Pharmacochimie Molꢁculaire et Cellulaire
U648 INSERM, UFR Biomꢁdicale
The anthraquinone ring system enables intercalation be-
tween base pairs of DNA in the B conformation.[5] In MX and
AM the nucleus is intercalated along a direction perpendicular
Universitꢁ Paris Descartes, Paris (France)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201000106.
1080
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1080 – 1091

Peptidyl-Ametantrone–DNA Complexes
to the long axis of the base pairs. This places the two arms, lo-
cated at positions 1 and 4, in the major groove of DNA and in
symmetrical antiparallel arrangements, each arm interacting
with a distinct strand.[6] Although AM is approximately 10-fold
less potent in vivo and 100-fold less potent in vitro than MX,[7]
it exhibits a lower genotoxicity.[8] The chemical and biological
activity exhibited by anthraquinone compounds are greatly af-
fected by the different substituents of the planar ring
system.[1,9] Side chains bear usually one or two positive charges
to establish an electrostatic interaction with the phosphate
backbone of the polynucleotide.
spectively by compounds 13 and 14. Of the Gly series, com-
pounds 8 and 9 were included as controls, whereas the alkyl
isostere was synthesized (7 in Scheme 3 below) but could not
be tested due to its limited water solubility.
We performed a DNA-unwinding assay to evaluate if the
new derivatives bind to DNA as intercalators, and whether in-
tercalation parallels the ability to inhibit human topoisomera-
se II. By using a fluorescence-quenching assay (FQA) coupled
to thermal denaturation of short double-stranded (ds) oligonu-
cleotides labeled with appropriate fluorescence/quenching
probes we aimed to demonstrate whether the novel deriva-
tives display the desired sequence-specific recognition. Our
earlier theoretical studies predicted that the peptide back-
bones of the two arms of the anthraquinone nucleus should
adopt an antiparallel b-sheet arrangement, enabling each arm
to interact with one distinct strand.[17,18] Their symmetrical dis-
position with respect to the chromophore is a reflection of the
palindromic nature of targeted sequences d(GGCGCC)2 and
d(CCCGGG)2 that were chosen for our studies.
Our group has explored the changes in the physicochemical
and biological properties of substituted anthraquinones in
which peptide chains are connected through an amide bond
to the planar ring.[10–16] A peptide sequence linked to a phar-
macophore group could, if properly selected and substituted,
confer drug specificity. In fact, it is well known that precise rec-
ognition of defined DNA sequences in biological systems is
mediated by enzymes and proteins that have appropriate
structural motives. Indeed, one of the major drawbacks of
DNA-directed anti-neoplastic agents is their inability to selec-
tively target cancer cells. Therefore the introduction, into pres-
ently available drugs, of specific recognition elements for mu-
tated oncogenes elements should represent a basic advance in
cancer chemotherapy.
Furthermore, an accurate structural and dynamical descrip-
tion of peptidyl anthraquinone complexes in aqueous solution,
which would help to understand the molecular basis of
ligand–DNA recognition can be obtained through molecular
dynamics (MD) simulations. Here we make use of the Amber
force field to investigate through MD simulation the dynamical
and stability properties of the novel peptidyl anthraquinone
complexes with the aim to verify earlier predictions and ration-
alize the observed structure–activity relationship.
The present study is part of our endeavor to design oligo-
peptide-conjugated intercalator molecules that target well-de-
fined DNA sequences in the major groove.[17–20] On the basis of
computational studies, anthraquinone derivatives were de-
signed, and their binding energies with d(GGCGCC)2 were com-
pared to those with the other concurrent palindromic sequen-
ces containing the CG step. The d(GGCGCC)2 sequence, a mu-
tational hot spot in DNA, is located in several oncogenes and
is also part of the primer binding site (PBS) of the long termi-
nal repeat (LTR) of the HIV-1 retrovirus that is necessary for the
initial steps of reverse transcriptase (RT)-mediated cDNA syn-
thesis.[19,21,22] Therefore, it also represents an important target
for the design of novel antiviral drugs.
Results and Discussion
Chemistry
To obtain the ametantrone isosteres 1 and 2 we treated leuco-
quinizarin (L) with two equivalents of amine to get the Schiff
bases, which were then isomerized and oxidized by air as
shown in Scheme 1.
To verify these modeling predictions experimental-
ly,[17,18] we synthesized and tested the series of com-
pounds depicted in Figure 1. The substituents are
linked to the anthraquinone by esterification of the
two arms of AM (3) with Gly or Gly-(l-Lys) (com-
pounds 9 and 15), in which 15 represents the peptid-
yl-AM predicted to exhibit sequence-selective proper-
ties. Extending the arms upon passing from Gly to
Scheme 1. Reagents and conditions: a) Na2S2O4, N2, H2O, 70–808C; b) NH2(CH2)2X-
Gly-Lys should enable the cationic Lys side chains to (CH2)2OH, N2, H2O, 1008C, overnight; c) O2, H2O, 258C, 6 h.
bind to electron-rich O6/N7 sites with G bases up-
stream of the intercalation site.[18] An important fea-
ture of these derivatives is the presence of a free ni-
trogen in the lateral chains (N-series), 17 was therefore synthe-
sized as a control. Earlier studies demonstrated that such a ni-
trogen plays an important role because no activity is observed
when it is replaced by a methylene group or another atom.[23]
Further appropriate controls for our designed peptidyl-deriva-
tive of AM bear different linkers between the planar nucleus
and the peptide chains (-CH2- and -O- series), exemplified re-
For the synthesis of ametantrone (3) we used a reported
procedure by starting from 1,4-difluoroanthraquinone (Sche-
me 2A)[1] because the former procedure with N-(2-hydroxy-
ethyl)ethylenediamine and leucoquinizarin gave the cycliza-
tion compound 1,2,3,4-tetrahydro-4-(2-hydroxyethyl)-6-[{2-[(2-
hydroxyethyl)amino]ethyl}amino]naphtho[2,3-f]quinoxaline-
7,12-dione 3’ (Scheme 2B).[24]
ChemMedChem 2010, 5, 1080 – 1091
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1081

B. Gatto et al.
MED
yl-anthraquinones are analyzed by their different
electrophoretic mobilities. DNA migration reflects the
extent of intercalation due to the complexed ligand
and also the persistence of the complex in an electric
field. AM (3), used as a positive control, was found to
have the strongest intercalation properties because
gel retardation was evident already at 10 mm
(Figure 2).
The novel peptidyl-anthraquinones in the nitrogen
series have weaker unwinding properties: 9 and 15
show an appreciable effect only at 10-fold higher
concentrations, whereas unwinding of plasmid DNA
is more pronounced for 17 and shows a good corre-
lation between DNA mobility and concentration. The
Scheme 2. A) Reagents and conditions: a) DMSO, room temperature, 96 h. B) Reagents
and conditions: a) Na2S2O4, N2, H2O, 70–808C; b) NH2(CH2)2NH(CH2)2OH, N2, H2O, 1008C,
overnight; c) O2, H2O, 258C, 6 h.
bioisostere control
8 seems not to appreciably
modify the topological state of plasmid pBR322 at
any concentration tested, and 14 demonstrates a
Then the alcoholic groups of both chains in 1–3 were esteri-
fied with Boc-Gly-OH, and the Gly was deprotected with 90%
trifluoroacetic acid (TFA) to give compounds 7–9 as shown in
Scheme 3.
weak unwinding at the highest concentration taken. The alkyl
isostere 13 shows DNA retardation at 10 mm, but at 100 mm
ethidium staining of the lane is reduced, and DNA is precipitat-
ed in the loading well (Figure 2).
Peptidyl conjugates of AM exhibit sequence--
selective DNA binding
Having established that the novel peptidyl-anthraqui-
nones derived from AM are intercalators that are able
to unwind DNA, it was essential to obtain informa-
tion on their DNA sequence-selective binding. There-
fore, we resorted to fluorescence-quenching assays
(FQA) to analyze the melting profiles of three differ-
ent double-stranded oligonucleotides upon peptidyl-
anthraquinone addition. All three oligos are 18 base
pairs long. The first sequence, used as a control, has
a random alternation of bases (denoted as random
oligo).[25] The two other ones have core palindromic
GC-rich hexanucleotides, namely, CCCGGG (Z1) or
GGCGCC (Z2). Their choice was made on the basis of
molecular modeling,[18] which predicted these as the
major targets for the Gly-Lys derivative 15.
Scheme 3. Reagents and conditions: a) DCC, DMAP, Boc-Gly-OH, THF, room temperature,
16 h, Ar(g); b) TFA 90% 1 h, Et2O.
FQA was run with FAM and DAB probes, as de-
scribed.[25] Briefly, forward strand oligonucleotides
were labeled at their 5’ ends with FAM, whereas the
The reaction of intermediates 7–9 with Boc-Lys(Boc)-OH and
the successive deprotection of all amino groups with 90% TFA
gave compounds 13–15, as shown in Scheme 4.
Upon treating Boc-Gly-OH with the ametantrone side chain,
both of the secondary aliphatic amines and both of the pri-
mary alcohols were acylated to give the tetrasubstituted com-
pound 17 (Scheme 5).
DAB quencher was labeled at the 3’ end of the complementary
sequence. The double-stranded oligonucleotide that resulted
when the two complementary strands were annealed is a
blunt-ended hybrid that places FAM and DAB so close together
that contact quenching occurs. Indeed, neither the peptidyl-
anthraquinones, nor the DAB-labeled oligonucleotides, nor the
hybridized blunt-end oligo showed an appreciable fluores-
cence when excited at 494 nm at room temperature, but the
free FAM-labeled oligonucleotides under the same conditions
showed an appreciable emission at 520 nm. As melting of the
hybridized oligo duplexes takes place upon temperature in-
crease, the FAM-labeled oligonucleotide is released from the
quenching effect and its fluorescent signal increases due to
the increasing distance between the two probes (not shown).
Peptidyl-anthraquinones are DNA intercalators
The electrophoresis mobility of different topological DNA
forms offers an easy way to analyze the intercalation potency
of the novel compounds. In this work the changes in topologi-
cal properties of pBR322 plasmid after incubation with peptid-
1082
www.chemmedchem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1080 – 1091

Peptidyl-Ametantrone–DNA Complexes
pounds 8 and 14 are poor bind-
ers in all cases, whereas 15 and,
to a lesser extent, 9 emerge as
sequence-selective binders.
Our experimental results con-
firm, as predicted, that the free
nitrogen in the linker of 15 and,
to a lesser extent, 9 is necessary
for long-range specificity when
the peptide is conjugated to the
ametantrone nucleus. Indeed, 17
and 9 further illustrate the con-
tribution of the free nitrogen: in
the former derivative, an addi-
tional Gly is conjugated through
an amide bond to the nitrogen
in the lateral chain: it is evident
that when the side chain nitro-
Scheme 4. Reagents and conditions: a) Boc-Lys(Boc)-OSu, DMF, CH2Cl2, Et3N, room temperature, 16 h, Ar(g); b) TFA
90% 1 h, Et2O.
Figure 2. DNA-unwinding assay. Negatively supercoiled plasmid at defined
conditions migrates with the highest mobility (covalently closed circular
DNA, CCC). Increasing quantities of intercalating agents provoke unwinding
of supercoiled forms of DNA, and hence variation in the hydrodynamic
volume of DNA. This effect is visualized by retardation in the electrophoretic
migration of supercoiled plasmid. P=0.25 mg supercoiled pBR322;
OC=0.1 mg plasmid in open circular form obtained by limited DNase action;
L=linearized pBR22. Indicated compound concentrations are expressed as
micromolar. Electrophoresis [1% agarose gel in TBE 1ꢁ (89 mm Tris, 89 mm
boric acid, 2 mm EDTA, pH 8.0)] was run overnight 15 mVcmꢀ1, and the gel
was stained with ethidium bromide.
Scheme 5. Reagents and conditions: a) DCC, DMAP, Boc-Gly-OH, THF, room
temperature, 16 h, Ar(g); b) TFA 90% 1 h, Et2O.
The Tm of each oligo pair was determined as the temperature
value at which the melting profile (fluorescence vs. tempera-
ture) presents a flex point. We then measured the Tm of the
three different double-stranded oligonucleotides in the pres-
ence of the AM control (3) and of each peptidyl-anthraquinone
tested at 1, 10, and 100 mm concentration, and calculated the
DTm values reported in Table 1.
gen is not free (compound 17) sequence-specific DNA binding
of 9 is lost.
Melting experiments, although consistent with DNA-unwind-
ing studies, indicate that the peptidyl-anthraquinones behave
in contrasting ways upon binding to the random and the two
sequences having GC-rich core palindromes. This is evidenced
in Figure 3, which reports the relative stabilization of each
oligo duplex at the tested peptidyl-anthraquinones/DNA ratios.
Indeed, 3, 13, and 17 act as nonselective binders because
they stabilize the three sequences by the same amount. Com-
Regarding the melting analysis, in the case of compound 13
with Z2 at 100 mm we never observed fluorescence signals at
any temperature, which is likely due to the precipitation of the
oligo in the capillary that was used for the melting analysis.
Therefore, the DTm determination for this particular sample
could not be performed. Precipitation of DNA by a high con-
centration of 13 was observed also in the unwinding experi-
ments, and further confirmed by CD spectroscopy (see the
ChemMedChem 2010, 5, 1080 – 1091
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1083

B. Gatto et al.
MED
Molecular dynamics of DNA–14 and DNA–15 complexes
Table 1. DNA binding of peptidyl-anthraquinones expressed as a func-
tion of DTm.
The conformational features of the sequence specific peptidyl-
derivative 15 and of its non-sequence-specific isostere 14 in
complex with the core hexameric sequence of Z2 oligo (re-
spectively DNA–15 and DNA–14) during the MD trajectory
have been checked upon monitoring specific structural param-
eters as a function of time. The root-mean-square deviation
(RMSD) calculated for all atoms of both complexes are report-
ed in Figure 4. The deviation of the intercalative complex ge-
Compd
Oligo[a]
DTm [8C]
10 mm
1 mm
100 mm
Random[b]
Z1[c]
Z2[d]
1.26ꢁ0.80
1.08ꢁ0.07
1.24ꢁ0.19
5.54ꢁ0.96
5.85ꢁ0.51
6.11ꢁ1.22
14.60ꢁ1.37
16.66ꢁ0.65
15.45ꢁ0.13
3
Random
Z1
Z2
0.00
0.00
0.00
0.22ꢁ0.21
0.00
0.00
2.40ꢁ0.33
1.74ꢁ0.53
2.11ꢁ0.41
8
Random
Z1
Z2
0.00
0.00
0.00
0.18ꢁ0.16
2.98ꢁ0.04
2.48ꢁ0.33
2.80ꢁ0.86
10.74ꢁ0.60
9.84ꢁ0.29
9
Random
Z1
Z2
0.00
0.00
0.00
0.35ꢁ0.26
0.67ꢁ0.59
0.73ꢁ0.48
13.09ꢁ1.25
14.29ꢁ0.69
ND[e]
13
14
15
17
Random
Z1
Z2
0.00
0.00
0.00
0.00
0.00
0.00
2.31ꢁ1.01
4.19ꢁ0.31
3.52ꢁ0.40
Random
Z1
Z2
0.00
0.00
0.00
0.41ꢁ0.76
4.02ꢁ0.04
3.26ꢁ0.33
3.70ꢁ0.77
16.05ꢁ0.13
16.04ꢁ0.64
Random
Z1
Z2
0.24ꢁ0.17
0.00
0.00
2.21ꢁ0.21
2.69ꢁ0.11
2.38ꢁ0.33
8.23ꢁ0.25
9.10ꢁ1.24
8.78ꢁ0.28
[a] Annealed ds oligo (0.25 mm) in Tris 10 mm, EDTA 1 mm, NaCl 20 mm,
pH 7.5. [b] Random oligo Tm =48.07ꢁ0.718C. [c] Z1 sequence (core:
CCCGGG) Tm =45.95ꢁ0.148C. [d] Z2 sequence (core: GGCGCC) Tm =
48.34ꢁ0.408C. [e] Not determined.
Figure 4. Root-mean-square atom positional deviations of complexes DNA–
14 and DNA–15 (see labels) as a function of simulation time. The structures
were superimposed and the RMSDs were calculated for all atoms excluding
the first two and the last two base pairs.
ometry from the initial conformation increases with time al-
though over the whole trajectory the simulated decanucleo-
tide duplex structures remain closer to that of canonical B-
DNA. The RMSD from the average MD structure oscillates
around 1.5 ꢂ for the DNA–15 complex and around 2.2 ꢂ for
the DNA–14 complex. This indicates that both intercalative
complexes are stabilized in a well-defined configuration during
the simulation.
The higher RMSD average value measured during the trajec-
tory for the DNA–14 complex can be explained by the loss of
the electrostatic interactions that take place between the pro-
tonated secondary amine present in the linker of 15 and the
adjacent DNA in the corresponding DNA–15 complex (see
Figure 5). In 14, the secondary amine is replaced by an ether
moiety.
Figure 3. Sequence-specific recognition of AM-derived peptidyl-anthraqui-
nones. Differential stabilization (DT) of the designed oligonucleotides (Z1
and Z2 core sequences) with reference to a random duplex sequence by 10
and 100 mm ligand. Annealed ds oligos were 0.25 mm in 10 mm Tris, 1 mm
EDTA, 20 mm NaCl, pH 7.5. Data from FQA experiments reported in Table 1.
The essential role of such an electrostatic interaction can
also be evidenced by monitoring the interaction energy profile
of the complexes of 14 and 15 during MD. As observed in
Figure 6, the overall stability of the DNA–15 complex is higher
than that of the DNA–14 complex due to its more favorable
electrostatic energy contribution. For both complexes, on the
other hand, on each strand the Lys side chain binds to the O6/
N7 electron-rich atoms of two G bases upstream of the interca-
lation site, as shown in Figure 5. These electrostatic interac-
tions increase the stability of both complexes.
Supporting Information). This peculiar difference from its con-
geners 14 and 15 is clearly ascribable to the long alkyl linkers,
which can interact with each other in solution as well as in the
complex with DNA.
1084
www.chemmedchem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1080 – 1091

Peptidyl-Ametantrone–DNA Complexes
Table 2. Cell toxicities and logP values of peptidyl-anthraquinones.
Compd
EC50 [mm][a]
logP[b]
HL60
LoVo
HeLa
3
8
9
13
14
15
17
0.27ꢁ0.1
30.0ꢁ6.1
NA
27.7ꢁ3.7
NA
1.50ꢁ0.1
NA[c]
96.0ꢁ6.9
44.1ꢁ1.6
NA
0.90ꢁ0.11
66.5ꢁ14.3
NA
83.2ꢁ17.1
NA
0.36ꢁ0.07
1.30ꢁ0.08
ꢀ0.41ꢁ0.13
0.30ꢁ0.15
0.78ꢁ0.11
43.5ꢁ6.5
NA
NA
NA
NA
NA
ꢀ2.65ꢁ0.21
ꢀ1.08ꢁ0.05
[a] Determined at 72 h by MTT assay. [b] Determined experimentally at
pH 7.4. [c] Not active (EC50 ꢂ100 mm).
that the novel compounds are significantly less cytotoxic than
AM (3) in all of the tested lines, namely the leukemia cell line
HL60, the LoVo colon carcinoma cells, as well as the HeLa (cer-
vical carcinoma), with leukemia cells being the most sensitive.
Such a limited toxicity can be ascribed to an impaired permea-
tion in cells due to the increased charge of the peptidyl conju-
gates, in particular in 15, as we have observed for other highly
charged derivatives.[26] Therefore, we similarly attempted to
measure the uptake of the peptidyl-anthraquinones in the
HL60 cells to verify this hypothesis. However, because the pep-
tidyl-anthraquinones are not fluorescent, and their molar ab-
sorptivity at the absorbance maximum is low, the estimate of
concentration was not precise enough to safely determine the
amount of their uptake in HL60 cells. Hence, we experimentally
measured their permeation coefficient to determine the rela-
tive lipophilicity of these highly charged derivatives and corre-
late them with the biological data. The results, shown in
Table 2, indicate that all the peptidyl-derivatives of AM, namely
9, 15, and 17, exhibit negative permeation coefficients, with
the lysyl derivative 15 being the most hydrophilic of the
series; this offers an explanation for the low cytotoxicity.
However, 15 exhibits a residual cell cytotoxicity in the leuke-
mia cell line, whereas it is considered to be inactive (EC50 ꢂ
100 mm) in the other tested cell lines. Isosteres and controls
show different biological activity: in the series represented by
8 and 14 the less-charged derivative 8 has a limited cytotoxici-
ty consistent with its lipophilic character, whereas its more
charged Lys derivative 14, the isostere of 15, loses activity
while retaining a good permeation coefficient. The other iso-
stere, 13, shows high permeation that, coupled to a good al-
though nonselective DNA binding, allows it to retain cytotoxic
properties.
Figure 5. Structure of the binary complex of peptidyl-anthraquinones 14 (on
the right) and 15 (on the left) with the target DNA sequence. All complexes
are displayed from the side of the DNA longitudinal axes. DNA base pairs,
above and below the intercalation site, are represented by the correspond-
ing Connolly surface colored in light gray. Stabilizing interactions are indicat-
ed by green arrows.
Figure 6. Interaction energies as a function of simulation time of the DNA
duplex interacting with peptidyl-anthraquinones 14 and 15.
The b-sheet-like motif of the two peptidyl tails of both 14
and 15 is also conserved. In fact, the double charge–dipole in-
teractions mediated by the two lysines seem to guarantee this
stable conformation over the entire 2 ns course of simulation.
It is also gratifying to observe that such a motif, which had
been predicted earlier on the basis of simpler energy minimiza-
tions,[17,18] could be retained in the course of the 2 ns simula-
tion.
Hence, the fraction of 15 able to reach and bind to specific
DNA sequences might be able to elicit cell responses. Because
AM can affect topoisomerase II (topo II) activity, we analyzed
the activity of AM-derived peptidyl-anthraquinones toward the
catalytic activity of this enzyme. Inhibition of pBR322 relaxation
by topo II (Figure 7) performed with purified human enzyme
under these experimental conditions was positive for 3 at
10 mm, whereas 15 showed inhibition only at the highest con-
centration (100 mm). This result is consistent with those of the
unwinding experiments and rules out topo II as a primary
target for AM-derived peptidyl-anthraquinones.
Biological evaluation of peptidyl-anthraquinones
We have shown that a simple dipeptide can be conjugated to
the nucleus of a known intercalator to direct it to sequence
specific recognition of DNA. Because AM is a cytotoxic drug,
we investigated whether its cell toxicity could be affected by
peptide conjugation. The data reported in Table 2 show in fact
ChemMedChem 2010, 5, 1080 – 1091
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1085

B. Gatto et al.
MED
The biological evaluation of the new compounds showed
that the presence of the lysine-containing peptide in the AM-
derived series, although necessary for sequence-specific DNA
recognition, confers a decisively high hydrophilic character, im-
pairing biological effects in the cell lines tested. However, the
residual cytotoxicity exhibited by the highly charged com-
pound 15 points to the possibility of exploiting appropriate
delivery systems to circumvent cell permeation. Furthermore,
in a new but promising line of research, the limited cytotoxic
effect exhibited by the peptidyl-AM 15 coupled to its se-
quence preferences for the d(GGCGCC)2 core, which is present
in the primer binding site (PBS) of HIV, offers a possibility for
the sequence-selective peptidyl-anthraquinones to be used as
drugs acting at early events of HIV replication.[29–31]
Figure 7. Human topoisomerase IIa relaxation assay. AM (3) and peptidyl-AM
derivative concentrations used were 0.1, 1, 10, and 100 mm as indicated, and
were incubated for 1 h at 378C with DNA and enzyme before stopping the
reaction. D=0.1 mg plasmid DNA; T=human topo IIa isoform (0.01 U). Elec-
trophoresis [1% agarose gel in 89 mm Tris, 89 mm boric acid, 2 mm EDTA,
pH 8.0] was run overnight 15 mVcmꢀ1 and stained after the run with ethid-
ium bromide.
Conclusions
The search for specificity is a critical challenge in pharmaceuti-
cal sciences. The ability to distinguish between healthy and
malignant or infected cells, to prevent recognition of multiple
targets, and avoid the onset of undesired effects, is a major in-
centive for the development of novel, effective drugs. Compu-
tational studies can be a very valuable asset toward the ration-
al design of selective DNA binding ligands. In the present
work, and on the basis of earlier computational studies, we in-
vestigated novel anthraquinone compounds derived from
ametantrone, a well-studied antitumor compound. One of the
novel derivatives was designed to target selectively the palin-
dromic GC-rich sequences (GGCGCC)2 and (CCCGGG)2.
Experimental Section
Chemistry
NMR spectra were recorded on a Bruker Avance AMX 300 spec-
1
trometer; H and 13C NMR spectra were run by using [D6]DMSO as
a solvent, and the solvent peak was used as an internal standard.
Chemical shifts (d) are expressed in parts per million (ppm) relative
to (CH3)4Si, and spin multiplicities are indicated as: s (singlet), brs
(broad singlet), d (doublet), dd (double doublet), t (triplet), and m
(multiplet), and the values are expressed in Hz. Analytical thin-layer
chromatography (TLC) was carried out on pre-coated silica gel
plates (Merck 60F254), and spots were visualized with UV light at
254 nm. Preparative column chromatography was performed by
using Merck silica gel (230–400 mesh). All starting materials not de-
scribed below, the solvents and the deuterated solvents were pur-
chased from commercial sources, mainly Aldrich and Fluka. All re-
agents and solvents were used as received from commercial sour-
ces without additional purification, unless stated otherwise. Analyt-
ical HPLC was carried out on Varian HPLC system by using a Agi-
lent Extend C18 column, 4.6 mmꢁ250 mm, with a particle size of
5 mm. The mobile phase consisted of A: 0.16m ammonium formate
buffer (70%), pH 2.7 (by formic acid), sodium hexanesulfonate was
added at a concentration of 0.025m as an ion-pair reagent; and B:
MeCN; the eluting gradient was: 90A:10B!10A:90B in 25 min and
then 10A:90B for an additional 5 min.[32] The eluate was monitored
at 260 nm, an absorption maximum of the anthraquinone nucleus.
HRMS data were obtained with a Mariner API-TOF instrument (Per-
septive Biosystems Inc., Framingham MA 01701, USA).
The change of topological forms of plasmid pBR322 upon
compound addition (unwinding assay) enabled us to examine
the intercalation properties of the new peptidyl-anthraqui-
nones relative to the parent compound AM, and melting pro-
files coupled to FQA gave us information on sequence selectiv-
ity and allowed us to evaluate whether the peptidyl derivatives
could selectively target the desired sequences.
Our new series of designed peptidyl-anthraquinones exhibit
a flexible and basic linker: 15, which bear a Gly-(l-Lys) dipep-
tide conjugated as a depsipeptide to the anthraquinone, is the
most sequence selective of the whole series, in complete
agreement with the theoretical predictions.[18] The two isostere
controls that were synthesized and tested were 13 (alkyl linker)
and 14 (ether linker), which conjugate the anthraquinone nu-
cleus to the same Gly-(l-Lys) dipeptide. Compound 14 is a
poor DNA binder in all cases, whereas the alkyl isostere 13
binds DNA without any sequence discrimination and causes
precipitation as evidenced in unwinding and in FQA experi-
ments, and confirmed in CD spectroscopy experiments.
1,4-bis[(5-Hydroxypentyl)amino]-9,10-anthracenedione
(1):
Na2S2O4 (1.0 g, 5.7 mmol) was added to a suspension of leucoquini-
zarin (0.5 g, 2.08 mmol) in 0.5% aq K2CO3 (50 mL) under a N2 at-
mosphere. The reaction mixture was heated at 1008C, and Na2S2O4
(0.5 g, 2.9 mmol) and the 5-aminopentan-1-ol were added (2.15 g,
20.80 mmol). The reaction mixture was held at reflux overnight,
cooled to room temperature, and air was introduced for 6 h under
vigorous stirring. H2O was added until complete precipitation of
the product, and the solid was filtered out and dried. The crude
product was then purified by column chromatography (EtOAc/
MeOH, 9:1) to obtain compound 1 as a blue powder (700 mg,
82%). HPLC: tR =17.7 min, purity >99%.[33] 1H NMR (300 MHz,
[D6]DMSO): d=10.92 (t, J=5.3 Hz, 2H), 8.25 (m, 2H), 7.79 (m, 2H),
7.50 (s, 2H), 4.40 (t, J=5.1 Hz, 2H), 3.49 (m, 8H), 1.70 (m, 4H),
1.50 ppm (m, 8H); 13C NMR (300 MHz, [D6]DMSO): d=23.5, 29.5,
32.6, 42.5, 61.0, 108.5, 126.1, 125.1, 132.6, 134.2, 146.5, 180.8 ppm;
MD analysis supports our binding data; interestingly, the
peptidyl tails of 15 adopt a b-sheet conformation. This motif
could be used as a scaffold to graft side chains at appropriate
locations to target specific bases upstream of the intercalation
site. DNA recognition in the major groove occurs in the majori-
ty of cases by proteins that adopt an a-helical conformation.
There exist however precedents for proteins that bind to DNA
in the major groove and adopt a b-sheet conformation, as in
the case of the dimeric Arc repressor.[27,28]
1086
www.chemmedchem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1080 – 1091

Peptidyl-Ametantrone–DNA Complexes
HRMS (ES+): m/z: calcd for C24H30N2O4 [M+H]+: 411.5793, found:
411.5145.
16.2 min, purity 98%. 1H NMR (300 MHz, [D6]DMSO): d=11.25 (t,
J=5.2 Hz, 2H), 8.25 (s,2H), 7.80 (dd, J=2.5 Hz, J=5.9 Hz, 2H), 7.55
(dd, J=2.5 Hz, J=5.9 Hz, 2H), 6.50 (s, 2H), 4.20 (m, 4H), 3.95 (s,
4H), 3.20 (m, 4H), 2.80 (m, 8H), 1.40 ppm (s, 18H); 13C NMR
(300 MHz, [D6]DMSO): d=28.5, 42.3, 48.7, 49.5, 64.1, 79.5, 108.9,
125.5, 131.2, 132.7, 134.2, 138.4, 156.3, 169.6, 180.9 ppm; HRMS
(ES+): m/z: calcd for C36H50N6O10 [M+H]+: 727.8292, found:
727.8990.
1,4-bis{[2-(2-Hydroxyethoxy)ethyl]amino}-9,10-anthracenedione
(2): was obtained by using the same procedure as described
before by starting from leucoquinizarin (0.5 g, 2.08 mmol) and 2-(2-
aminoethoxy)ethanol (2.19 g, 20.80 mmol). The crude product was
purified by column chromatography (EtOAc/MeOH, 9:1), to yield
compound 2 as a blue powder (800 mg, 93%). HPLC: tR =12.9 min,
purity 99%. 1H NMR (300 MHz, [D6]DMSO): d=10.88 (t, J=5.2 Hz,
2H), 8.24 (dd, J=2.5 Hz, J=5.9 Hz, 2H), 7.79 (dd, J=2.5 Hz, J=
5.9 Hz, 2H), 7.50 (s, 2H), 4.62 (t, J=5.0 Hz, 2H), 3.71 (t, J=5.0 Hz,
4H), 3.61 (m, 4H), 3.53 ppm (m, 8H); 13C NMR (300 MHz,
[D6]DMSO): d=42.5, 60.6, 69.6, 72.7, 108.8, 125.0, 126.1, 132.7,
134.2, 146.3, 181.1 ppm; HRMS (ES+): m/z: calcd for C22H26N2O6
[M+H]+: 415.7658, found: 415.7209.
1,4-bis{[5-(Glycyl)hydroxypentyl]amino}-9,10-anthracenedione
bistrifluoroacetate (7): A solution of 4 (40 mg, 0.055 mmol) in
90% TFA (50 mL) was stirred for 1 h at room temperature and then
cold Et2O (5 mL) was added. The reaction mixture was kept at 08C
for 30 min and filtered; the precipitate was washed twice with Et2O
(5 mL). Compound 7 was obtained as a blue powder (39 mg, 94%).
1
HPLC: tR =15.7 min, purity 99%. H NMR (300 MHz, [D6]DMSO): d=
10.80 (t, J=5.2 Hz, 2H), 8.25 (dd, J=2.5 Hz, J=5.9 Hz, 2H), 7.78
(dd, J=2.5 Hz, J=5.9 Hz, 2H), 7.50 (s, 2H), 4.10 (s, 4H), 3.64 (m,
4H), 3.44 (m, 4H), 1.80 (m, 4H), 1.55 ppm (m, 8H); 13C NMR
(300 MHz, [D6]DMSO): d=23.3, 28.7, 29.9, 39.8, 45.1, 65.0, 108.5,
117.4 (TFA), 125.2, 126.5, 132.7, 134.2, 146.7, 161.4 (TFA), 167.4,
180.9 ppm; HRMS (ES+): m/z: calcd for C28H36N4O6 [M+2H]+:
263.3247, found: 263.3163.
1,4-bis[{2-[(2-Hydroxyethyl)amino]ethyl}amino]-9,10-anthracene-
dione (3): was obtained by following literature preparation, from
1,4-difluoroanthraquinone (200 mg, 0.819 mmol) in DMSO (4 mL)
and N-(2-Hydroxyethyl)ethylenediamine (256 mg, 2.457 mmol).
Compound 3 was prepared as a blue powder (120 mg, 36%).
HPLC: tR =9.4 min, purity >99%.[34] 1H NMR (300 MHz, [D6]DMSO):
d=10.69 (t, J=6.3 Hz, 2H), 8.24 (dd, J=3.2 Hz, J=5.8 Hz, 2H), 7.83
(dd, J=3.2 Hz, J=5.8 Hz, 2H), 7.68 (s, 2H), 3.59 (m, 4H), 3.30 (m,
4H), 2.80 ppm (m, 8H); 13C NMR (300 MHz, [D6]DMSO): d=48.5,
49.6, 52.5, 62.2, 109.7, 118.7, 129.8, 132.4, 133.8, 138.3, 182.2 ppm;
HRMS (ES+): m/z: calcd for C22H28N4O4 [M+H]+: 413.9581, found:
413.9256.
1,4-bis[{2-[2-(Glycyl)hydroxyethoxy]ethyl}amino]-9,10-anthrace-
nedione bistrifluoroacetate (8): With the same procedure de-
scribed for compound 7, from 5 (50 mg, 0.069 mmol) in 90% TFA
(50 mL), compound 8 was prepared as a blue powder (50 mg,
97%). HPLC: tR =10.9 min, purity 98%. 1H NMR (300 MHz,
[D6]DMSO): d=10.82 (t, J=5.2 Hz, 2H), 8.24 (dd, J=2.5 Hz, J=
5.9 Hz, 2H), 7.79 (dd, J=2.5 Hz, J=5.9 Hz, 2H), 7.50 (s, 2H), 4.25
(m, 4H), 3.67 (m, 4H), 3.62 (s,4H), 3.58 (m, 4H), 3.23 ppm (m, 4H);
13C NMR (300 MHz, [D6]DMSO): d=40.9, 44.8, 64.1, 69.0, 70.2, 109.5,
117.4, 118.2, 129.5, 132.4, 133.7, 138.7, 161.4, 169.6, 182.2 ppm;
HRMS (ES+): m/z: calcd for C26H32N4O8 [M+2H]+: 265.1183, found:
265.1156.
1,4-bis{[5-(Boc-Glycyl)hydroxypentyl]amino}-9,10-anthracene-
dione
0.122 mmol),
(4):
N,N-Dicyclohexylcarbodiimide
N,N-dimethylaminopyridine
(DCC;
(DMAP;
25.2 mg,
14.9 mg,
0.122 mmol) and Boc-Gly-OH (214 mg, 1.22 mmol) were added to a
solution of 1 (50 mg, 0.122 mmol) in THF (10 mL) under an argon
atmosphere. The reaction mixture was stirred at room temperature
overnight. Then the solvent was evaporated under vacuum and
the residue was purified by column chromatography (toluene/
EtOAc, 8:2), to yield compound 4 as a blue powder (52 mg, 59%).
HPLC: tR =28.6 min, purity >99%. 1H NMR (300 MHz, [D6]DMSO):
d=10.88 (t, J=5.4 Hz, 2H), 8.25 (dd, J=2.5 Hz, J=5.9 Hz, 2H), 7.78
(dd, J=2.5 Hz, J=5.9 Hz, 2H), 7.50 (s, 2H), 7.20 (t, J=6.3 Hz, 2H),
4.08 (t, J=6.3 Hz, 4H), 3.67 (m, 4H), 3.45 (m, 4H), 1.68 (m, 8H),
1.50 (m, 4H), 1.36 ppm (s, 18H); 13C NMR (300 MHz, [D6][D6]DMSO):
d=23.2, 28.2, 28.5, 29.2, 42.3, 64.5, 78.6, 108.6, 125.1, 126.1, 132.6,
134.2, 146.4, 156.2, 170.8, 180.9 ppm; HRMS (ES+): m/z: calcd for
C38H52N4O10 [M+H]+: 725.3534, found: 725.3557.
1,4-bis{[2-{[2-(2-Glycyloxyethyl)]amino}ethyl]amino}-9,10-anthra-
cenedione bistrifluoroacetate (9): By using the same procedure
described for the preparation of compound 7, and by starting
from 6 (20 mg, 0.028 mmol), compound 9 was prepared as a blue
powder (18 mg, 85%). HPLC: tR =9.3 min, purity >99%. 1H NMR
(300 MHz, [D6]DMSO): d=11.25 (t, J=5.3 Hz, 2H), 8.25 (s, 2H), 7.82
(dd, J=2.6 Hz, J=5.8 Hz, 2H), 7.57 (dd, J=2.6 Hz, J=5.8 Hz, 2H),
6.48 (s, 2H), 4.16 (m, 4H), 3.97 (s, 4H), 3.22 (m,4H), 2.84 ppm (m,
8H); 13C NMR (300 MHz, [D6]DMSO): d=40.4, 48.5, 49.7, 64.1, 108.9,
117.4, 125.5, 131.2, 132.7, 134.2, 138.4, 169.6, 171.4, 180.9 ppm,
HRMS (ES+): m/z: calcd for C26H34N6O6 [M+2H]+: 264.3654, found:
264.3156.
1,4-bis[{2-[2-(Boc-Glycyl)hydroxyethoxy]ethyl}amino]-9,10-an-
thracenedione (5): By using the previous procedure from com-
pound
2
(50 mg, 0.122 mmol) and Boc-Gly-OH (214 mg,
1,4-bis{[5-(Boc-Lys(Boc)-Glycyl)hydroxypentyl]amino}-9,10-an-
1.22 mmol), compound 5 was prepared as a blue powder (73 mg,
82%). HPLC: tR =20.6 min, purity >99%. 1H NMR (300 MHz,
[D6]DMSO): d=10.87 (t, J=5.2 Hz, 2H), 8.23 (dd, J=2.5 Hz, J=
5.9 Hz, 2H), 7.81 (dd, J=2.5 Hz, J=5.9 Hz, 2H), 7.50 (s, 2H), 7.22 (t,
J=6.1 Hz, 2H), 4.25 (m, 4H), 3.92 (t, J=6.1 Hz, 4H), 3.65 (m, 4H),
3.60 (m, 4H), 3.23 (m, 4H), 1.40 ppm (s, 18H); 13C NMR (300 MHz,
[D6]DMSO): d=28.5, 42.3, 44.0, 64.4, 69.1, 70.3, 79.5, 110.7, 118.7,
130.1, 132.2, 133.7, 138.2, 156.3, 169.6, 182.2 ppm; HRMS (ES+):
m/z: calcd for C36H48N4O12 [M+H]+: 729.8067, found: 729.3319.
thracenedione (10): A solution of Boc-Lys-(Boc)-OSu (80 mg,
0.180 mmol) in CH2Cl2 (14 mL), which had been adjusted to pH 9
with Et3N was added to a solution of 7 (31 mg, 0.042 mmol) in
DMF (2 mL). The reaction mixture was stirred for 2 h. Then H2O
was added, the product was extracted in CH2Cl2, and the organic
phase was evaporated under vacuum. The crude product was puri-
fied by column chromatography, (EtOAc/MeOH, 9:1), to obtain
compound 10 as a blue powder (25 mg, 50%). HPLC: tR =29.6 min,
purity 99%. 1H NMR (300 MHz, [D6]DMSO): d=10.82 (t, J=5.2 Hz,
2H), 8.25 (dd, J=2.5 Hz, J=5.9 Hz, 2H), 7.80 (dd, J=2.5 Hz, J=
5.9 Hz, 2H), 7.50 (s, 2H), 4.53 (m, 2H), 4.16 (s, 4H), 4.08 (m, 4H),
3.06 (m, 4H), 2.96 (m, 4H), 1.79 (m, 4H), 1.57 (m, 8H), 1.52 (m, 4H),
1.40 (s, 36H), 1.29 ppm (m, 8H); 13C NMR (300 MHz, [D6]DMSO): d=
23.3, 28.5, 28.7, 29.4, 29.9, 31.5, 40.8, 42.1, 45.1, 55.4, 65.0, 79.7,
1,4-bis{[2-{[2-(2-Boc-Glycyloxyethyl)]amino}ethyl]amino}-9,10-an-
thracenedione (6): By using the same procedure described previ-
ously for the preparation of compound 4, and starting from 3
(70 mg, 0.170 mmol) and Boc-Gly-OH (214 mg, 1.22 mmol), com-
pound 6 was obtained as a blue powder (22 mg, 18%). HPLC: tR =
ChemMedChem 2010, 5, 1080 – 1091
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1087

B. Gatto et al.
MED
108.9, 125.5, 131.2, 132.7, 134.2, 138.4, 156.0, 169.6, 171.4,
180.9 ppm; HRMS (ES+): m/z: calcd for C60H92N8O16 [M+H]+:
1181.6572, found: 1181.6617.
1,4-bis{[2-{[2-(2-Lys-Glycyloxyethyl)]amino}ethyl]amino}-9,10-an-
thracenedione tetrafluoroacetate (15): By the same procedure de-
scribed before, from compound 12 (20 mg, 0.017 mmol) product
15 as a blue powder was obtained (17 mg, 81%). HPLC: tR =
1,4-bis[{2-[2-(Boc-Lys(Boc)-Glycyl)hydroxyethoxy]ethyl}amino]-
9,10-anthracenedione (11): By using the same procedure de-
scribed for compound 10, and by starting from compound 8
(50 mg, 0.066 mmol) and Boc-Lys(Boc)-OSu (88 mg, 0.198 mmol),
compound 11 was obtained as a blue powder (36 mg, 46%).
HPLC: tR =26.8 min, purity >99%. 1H NMR (300 MHz, [D6]DMSO):
d=10.82 (t, J=5.2 Hz, 2H), 8.25 (dd, J=2.5 Hz, J=5.9 Hz, 2H), 7.80
(dd, J=2.5 Hz, J=5.9 Hz, 2H), 7.50 (s, 2H), 4.53 (m, 2H), 4.25 (m,
4H), 4.16 (s, 4H), 3.65 (m, 4H), 3.60 (m, 4H), 3.23 (m, 4H), 2.96 (m,
4H), 1.79 (m, 4H), 1.55 (m, 4H), 1.40 (s, 36H), 1.29 ppm (m, 4H);
13C NMR (300 MHz, [D6]DMSO): d=20.7, 28.5, 28.7, 29.4, 31.5, 40.8,
41.9, 44.0, 55.4, 65.0, 69.1, 70.2, 79.7, 108.9, 125.5, 131.2, 132.7,
134.2, 138.4, 156.0, 169.6, 171.4, 180.9 ppm; HRMS (ES+): m/z:
calcd for C58H88N8O18 [M+H]+: 1185.3995, found: 1185.3756.
1
11.7 min, purity >99%. H NMR (300 MHz, [D6]DMSO): d=10.80 (t,
J=5.3 Hz, 2H), 8.62 (s, 2H), 8.20 (dd, J=2.5 Hz, J=5.9 Hz, 2H), 7.78
(dd, J=2.5 Hz, J=5.9 Hz, 2H), 6.75 (s, 2H), 4.20 (m, 8H), 3.92 (m,
4H), 3.80 (m, 2H), 3.60 (m, 8H), 2.67 (m, 4H), 1.74 (m, 4H), 1.50 (m,
4H), 1.36 ppm (s, 4H); 13C NMR (300 MHz, [D6]DMSO): d=21.5,
29.3, 31.5, 40.7, 42.1, 48.5, 49.7, 55.2, 64.3, 108.9, 117.2, 125.5,
131.2, 132.7, 134.2, 138.4, 161.4, 169.6, 171.4, 180.9 ppm; HRMS
(ES+): m/z: calcd for C38H58N10O8 [M+2H]+: 392.2292, found:
392.2385.
1,4-bis{[2-{[2-(2-Boc-Glycyloxyethyl)]Boc-Gly-amino}ethyl]amino}-
9,10-anthracenedione (16): DCC (30 mg, 0.146 mmol), DMAP
(18 mg, 0.146 mmol), and Boc-Gly-OH (128 mg, 0.73 mmol) were
added to a solution of 3 (30 mg, 0.073 mmol) in THF (10 mL),
under an argon atmosphere. The reaction mixture was stirred at
room temperature overnight, the solvent was evaporated under
vacuum, and the residue was purified by column chromatography
(EtOAc/MeOH, 95:5), to obtain compound 16 as a blue powder
(25 mg, 33%). HPLC: tR =16.8 min, purity >99%. 1H NMR (300 MHz,
[D6]DMSO): d=10.95 (t, J=5.2 Hz, 2H), 7.85 (dd, J=2.5 Hz, J=
5.9 Hz, 2H), 7.65 (dd, J=2.5 Hz, J=5.9 Hz, 2H), 6.70 (s, 2H), 4.30
(m, 4H), 3.92 (s, 4H), 3.85 (s, 4H), 3.48 (m, 8H), 3.32 (s, 4H),
1.38 ppm (s, 36H); 13C NMR (300 MHz, [D6]DMSO): d=28.6, 42.5,
46.4, 46.8, 47.5, 61.5, 79.5, 108.9, 125.5, 131.2, 132.7, 134.2, 138.4,
156.3, 164.3, 169.8, 181.1 ppm; HRMS (ES+): m/z: calcd for
C50H72N8O16 [M+H]+: 1041.5066, found: 1041.4940.
1,4-bis{[2-{[2-(2-Boc-Lys(Boc)-Glycyloxyethyl)]amino}ethyl]ami-
no}-9,10-anthracenedione (12): By using the same procedure as
described above for compound 10, and by starting from com-
pound 9 and Boc-Lys-(Boc)-Osu (17 mg, 0.038 mmol), compound
12 was synthesized as a blue powder (21 mg, 93%). HPLC: tR =
22.1 min, purity >99%.1H NMR (300 MHz, [D6]DMSO): d=10.82 (t,
J=5.2 Hz, 2H), 8.60 (s, 2H), 8.25 (dd, J=2.5 Hz, J=5.9 Hz, 2H), 7.80
(dd, J=2.5 Hz, J=5.9 Hz, 2H), 6.80 (s, 2H), 4.20 (m, 8H), 3.90 (m,
4H), 3.75 (m, 2H), 3.60 (m, 8H), 2.66 (m, 4H), 1.70 (m, 4H), 1.52 (m,
4H), 1.40 (s, 36H), 1.35 ppm (m, 4H); 13C NMR (300 MHz,
[D6]DMSO): d=21.5, 28.5, 29.3, 31.5, 40.7, 42.1, 48.5, 49.7, 55.2,
64.3, 79.5, 108.9, 125.5, 131.2, 132.7, 134.2, 138.4, 156.0, 169.6,
171.4, 180.9 ppm; HRMS (ES+): m/z: calcd for C58H90N10O16 [M+H]+:
1183.4389, found: 1183.4295.
1,4-bis{[2-{[2-(2-Glycyloxyethyl)]Gly-amino}ethyl]amino}-9,10-an-
thracenedione tetrafluoroacetate (17): A solution of 16 (25 mg,
0.024 mmol) in 90% TFA (50 mL) was stirred for 1 h at room tem-
perature, and then cold Et2O was added. The reaction mixture was
cooled and held at 08C for 30 min, and the precipitate was filtered
off and dried to furnish compound 17 as a blue powder (14 mg,
91%). HPLC: tR =9.2 min, purity 98%. 1H NMR (300 MHz,
[D6]DMSO): d=11.25 (t, J=5.3 Hz, 2H), 7.90 (dd, J=2.6 Hz, J=
5.8 Hz, 2H), 7.84 (dd, J=2.6 Hz, J=5.8 Hz, 2H), 6.58 (s, 2H), 4.36
(m, 4H), 3.62 (s, 4H), 3.54 (s, 4H), 3.48 (m, 4H), 3.43 (m, 4H),
3.25 ppm (m, 4H); 13C NMR (300 MHz, [D6]DMSO): d=40.4, 41.5,
46.3, 46.7, 47.8, 61.4, 108.9, 117.4, 125.5, 131.2, 132.7, 134.2, 138.4,
164.5, 169.8, 171.4, 181.0 ppm; HRMS (ES+): m/z: calcd for
C30H40N8O8 [M+2H]+: 321.2985, found: 321.2529.
1,4-bis{[5-(Lys-Glycyl)hydroxypentyl]amino}-9,10-anthracene-
dione tetrafluoroacetate (13):
A
solution of 10 (20 mg,
0.017 mmol) in 90% TFA (50 mL) was stirred for 1 h at room tem-
perature, then cold Et2O was added. The reaction mixture was
cooled to, and kept at 08C for 30 min, and the precipitate was fil-
tered out and washed with cold Et2O and finally dried. Compound
13 was obtained as a blue powder (17 mg, 81%). HPLC: tR =
18.7 min, purity 98%. 1H NMR (300 MHz, [D6]DMSO): d=10.80 (t,
J=5.3 Hz, 2H), 8.25 (dd, J=2.5 Hz, J=5.9 Hz, 2H), 7.78 (dd, J=
2.5 Hz, J=5.9 Hz, 2H), 7.50 (s, 2H), 4.16 (s, 4H), 4.08 (m,4H), 3.56 (s,
2H), 3.06 (m, 4H), 2.65 (m, 4H), 1.79 (m, 4H), 1.55 (m, 8H), 1.50 (m,
4H), 1.28 ppm (m, 8H); 13C NMR (300 MHz, [D6]DMSO): d=20.7,
23.3, 28.7, 29.9, 32.1, 34.3, 40.8, 42.3, 45.1, 54.2, 65.0, 108.5, 117.4,
125.2, 126.5, 132.7, 134.2, 138.2, 161.4, 169.4, 171.4, 180.9 ppm;
HRMS (ES+): m/z: calcd for C40H60N8O8 [M+2H]+: 391.2340, found:
391.2468.
Compound solutions
Stock anthraquinones solutions were prepared in anhyd DMSO. All
compounds were characterized spectrophotometrically and fluori-
metrically. They presented similar absorbance profiles relative to
the parental drug, ametantrone (3). Concentrations of solutions
were therefore obtained by using the experimentally calculated
molar absorption coefficient of ametantrone, 5400mꢀ1 cmꢀ1 at
626 nm, 258C in deionized purified (Milli-Q) H2O.
1,4-bis[{2-[2-(LysGlycyl)hydroxyethoxy]ethyl}amino]-9,10-anthra-
cenedione tetrafluoroacetate (14): By the same procedure de-
scribed before, from compound 11 (30 mg, 0.025 mmol) com-
pound 14 was obtained as a blue powder (30 mg, 97%). HPLC:
tR =13.7 min, purity 98%. 1H NMR (300 MHz, [D6]DMSO): d=10.80
(t, J=5.3 Hz, 2H), 8.25 (dd, J=2.5 Hz, J=5.9 Hz, 2H), 7.78 (dd, J=
2.5 Hz, J=5.9 Hz, 2H), 7.50 (s, 2H), 4.25 (m, 4H), 4.16 (s, 4H), 3.65
(m,4H), 3.60 (m, 4H), 3.56 (m, 2H), 3.26 (m, 4H), 2.65 (m, 4H), 1.79
(m, 4H), 1.55 (m, 4H), 1.29 ppm (m, 4H); 13C NMR (300 MHz,
[D6]DMSO): d=20.7, 32.1, 34.3, 40.8, 42.1, 44.0, 54.1, 64.4, 69.1,
70.3, 108.5, 117.4, 125.2, 126.5, 132.7, 134.2, 138.2, 161.4, 169.4,
171.4, 180.9 ppm; HRMS (ES+): m/z: calcd for C38H56N8O10 [M+H]+:
785.4114, found: 785.4154.
Unwinding assays
This assay was chosen to evaluate anthraquinones intercalating ac-
tivity and possible unwinding of the double helix. Briefly, 0.25 mg
supercoiled pBR322 plasmid (MBI Fermentas), was incubated in in-
creasing quantity of the test compound into a final volume of
20 mL in Tris 10 mm, EDTA 1 mm at pH 8 (TE buffer). After 1 h incu-
bation time, all samples were loaded into a 1% agarose gel and
1088
www.chemmedchem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1080 – 1091

Peptidyl-Ametantrone–DNA Complexes
run overnight at 15 Vcmꢀ1 in TBE (89 mm Tris, 89 mm boric acid,
2 mm EDTA, pH 8.0). To obtain the open circular (OC) form, the
pBR322 plasmid was incubated in ice for 5 min with 0.001 UmLꢀ1
RQ1 DNase-RNase-free (Promega) in 40 mm Tris-HCl pH 8.0, 10 mm
MgSO4 and 1 mm CaCl2. Nicking was stopped by adding 20 mm
EGTA, pH 8.0 (Promega). Linearization of plasmid was performed
by digestion with EcoRI (Fermentas) as indicated by the supplier.
Agarose (supplied by BioRad) gels were poured at a 1% concentra-
tion and run in TBE buffer 1ꢁ (89 mm Tris, 89 mm boric acid,
20 mm EDTA). Ethidium bromide (Sigma) stocks at 10 mgmLꢀ1
were diluted to 50 mgmLꢀ1 in deionized purified (Milli-Q) H2O and
added to visualize the DNA after electrophoresis.
(0.01 U) was added to the + topo II samples. After 1 h incubation
at 378C, reactions were stopped by adding 0.2m NaCl, 0.025% bro-
mophenol blue, 5% glycerol (4 mL), loaded into a 1% agarose gel,
and run overnight at 15 Vcmꢀ1 in TBE (89 mm Tris, 89 mm boric
acid, 2 mm EDTA, pH 8.0). Gels were stained with ethidium bro-
mide after running (post-staining). Pictures were taken by using
the imaging system Geliance 600 (PerkinElmer).
Cytotoxicity
The human HL60 cell line was maintained in RPMI 1640 medium
supplemented with 10% heat-inactivated fetal bovine serum (FBS),
penicillin (100 UmLꢀ1), and streptomycin (100 mgmLꢀ1). LoVo cells
were maintained in F-12 HAM medium supplemented with 10%
heat-inactivated FBS, 1% glutamine, penicillin (100 U), and strepto-
mycin (100 mgmLꢀ1). The HeLa cell line was maintained in DMEM
supplemented with 10% heat-inactivated FBS, penicillin
(100 UmLꢀ1), and streptomycin (100 mgmLꢀ1). The cytotoxicity of
the compounds was measured by the colorimetric 3-(4,5-dime-
thylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Brief-
ly, cells (5ꢁ103 wellꢀ1) were cultured in triplicate in a 96-well plate
in the absence or presence of various concentrations of the test
compounds. After the indicated time of incubation at 378C, MTT
was added to each well, and the mixture was incubated for 4 h.
The medium was removed before addition of DMSO for another
30 min. The plates were then read at 540 nm with a Microplate
Reader (BioRad).
Melting assays
The determination of the melting temperatures of double-stranded
oligonucleotides of known sequences was determined by fluores-
cence quenching (FQA) essentially as described.[25] Briefly, three
pairs of different oligonucleotides were used: each pair was la-
beled at the 5’ end of the forward strand with the FRET probe FAM
(6’-carboxyfluorescein), and with DAB (dabcyl, N-4’-carboxy-4-(di-
methylamino)azobenzene) at the 3’ end of the reverse strand. FAM
is a fluorescent molecule characterized by a high quantum yield,
with maximum emission wavelength at 517 nm when excited at
494 nm (e494 =83000 cmꢀ1 mꢀ1), efficiently quenched by DAB.[35]
The random sequence was obtained annealing FAM–GTG AGA TAC
CGA CAG AGG (FAM-random) with CCT CTG TCG TGA TCT CAC–
DAB (DAB-random); the second and the third pairs were specific
sequences whose core was designed according to theoretical stud-
ies,[17,18] namely Z1 by annealing FAM–5’-ACT ATT CCC GGG TAA
TGA-3’ (FAM-Z1) with 5’-TCA TTA CCC GGG TAA AGT-3’–DAB (DAB-
Z1). FAM–5’-ACT ATT GGC GCC TAA TGA-3’ (FAM-Z2) and 5’-TCA
TTA GGC GCC AAT AGT-3’–DAB (DAB-Z2) oligonucleotide were an-
nealed to give the Z2 oligo. FQA was carried out in capillary
glasses in a Roche Light Cycler 1.5, and the software allowed us to
collect and analyze all the data. The variation of Tm (DTm) was the
parameter chosen to describe different blunt-ended hybrid oligo-
nucleotide’s stability over the temperature increase. It was calculat-
ed by monitoring the variation of fluorescence with the tempera-
ture. The inflexion point of the melting profile of the double-
stranded oligonucleotide that was derived by the software, gave
the Tm for each melting profile. Oligonucleotides solutions were
prepared by dissolving the lyophilized nucleic acids in a proper TE
volume to obtain 1 mm concentrated stock solution. Their concen-
tration was determined by UV measurement at 260 nm. Melting re-
actions took place into glass capillaries containing the incubation
mixture (20 mL): anthraquinones (2 mL) at a final concentration of 1,
10, and 100 mm; HB buffer (4 mL; HB 20: 10 mm Tris, 1 mm EDTA,
20 mm NaCl, pH 7.5). Before the assay took place, forward and re-
verse strand oligo solutions were mixed at equimolar concentra-
tions, denatured at 958C for 5 min, and allowed to anneal at room
temperature for 3 h. Thermal protocol of melting was a ramp 308–
958C modified by 18Cminꢀ1 by the light cycler apparatus. The fluo-
rescence in each capillary was read at 520 nm by the same appara-
tus.
LogP
Octanol–water partition coefficient (logP) values were determined
for each compound by the shake flask method. Peptidyl-anthraqui-
nones were dissolved at known concentrations in a solution of
buffer at pH 7.4 (10 mm Tris, 20 mm NaCl, 1 mm Mg(ClO4)2) saturat-
ed with octanol, mixed with an equal amount of buffer-saturated
octanol, shaken extensively and allowed to equilibrate over 16 h at
room temperature. The UV absorbance of the aqueous solution
was measured before and after being shaken. LogP was calculated
as the log of the ratio of the concentration in the octanol phase to
the concentration in the aqueous phase at pH 7.4.
Computational methodologies
All modeling studies were carried out on a 20 CPU (Intel Core2
Quad CPU 2.40 GHz) Linux cluster.
Structure building: Peptidyl-anthraquinone structures were built
using the “Builder” module of Molecular Operation Environment
(MOE, version 2008.10).[36] The software package MOPAC (ver-
sion 7),[37] implemented in MOE suite, was used for the structure
optimization process by using the HF/PM3 semiempirical method.
MOPAC software has been also used to calculate atomic charges
by fitting to electrostatic potential maps (EPS method).[37]
Duplex intercalation site preparation: the present study involved
the use of consensus dinucleotide intercalation geometry d(CpG)
initially obtained by using NAMOT2 software (Nucleic Acid MOdel-
ing Tool, Los Alamos National Laboratory, Los Alamos NM, USA).[38]
The d(CpG) intercalation site was contained in the center of a hex-
anucleotide duplex of sequence d(5’-GGCGCC-3’). Hexamers in the
B-form were built by using the “DNA Builder” module of Molecular
Operation Environment (MOE, version 2008.10).[36] The hexanucleo-
tide was minimized by using the Amber95 all-atom force field,[39]
Topoisomerase II-mediated DNA relaxation
For each sample, supercoiled pBR322 plasmid (100 ng; Fermentas)
was incubated with increasing concentrations (up to 100 mm) of
the test compounds and with the positive control at a final
volume of 20 mL in 40 mm Tris-HCl, pH 7.5, 80 mm KCl, 5 mm DTT,
15 ngmLꢀ1 BSA, 1 mm ATP and 10 mm MgCl2. Topoisomerase IIa
ChemMedChem 2010, 5, 1080 – 1091
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1089

B. Gatto et al.
MED
[1] A. P. Krapcho, Z. Getahun, K. L. Avery, Jr., K. J. Vargas, M. P. Hacker, S.
Spinelli, G. Pezzoni, C. Manzotti, J. Med. Chem. 1991, 34, 2373–2380.
[2] N. A. Cornbleet, R. C. Stuart-Harris, I. E. Smith, E. E. Coleman, R. D.
Rubens, M. McDonald, H. T. Mouridsen, H. Rainer, A. T. van Oosterom,
J. F. Smyth, Eur. J. Cancer Clin. Oncol. 1984, 20, 1141–1146.
[3] K. Reszka, J. A. Hartley, P. Kolodziejczyk, J. W. Lown, Biochem. Pharmacol.
1989, 38, 4253–4260.
implemented by MOE modeling package, until the RMS value of
Truncated Newton method (TN) was <0.1 kcalmolꢀ1 ꢂꢀ1. To model
the effects of solvent more directly, a set of electrostatic interaction
corrections was used. The MOE suite implemented a modified ver-
sion of GB/SA contact function described by Still and co-authors.[40]
These terms model the electrostatic contribution to the free
energy of solvation in a continuum solvent model.
[4] C. Panousis, A. J. Kettle, D. R. Phillips, Anti-Cancer Drug Des. 1995, 10,
593–605.
Molecular docking protocol: Peptidyl-anthraquinones were
docked into both intercalation sites by using the flexible MOE-
Dock methodology. The purpose of MOE-Dock is to search for fa-
vorable binding configurations between a small, flexible ligand
and a rigid macromolecular target. Searching is conducted within a
user-specified 3D docking box, by using the “tabꢃ search” proto-
col[41] and MMFF94 force field.[42] Charges for peptidyl-anthraqui-
nones were imported from the MOPAC output files. MOE-Dock per-
forms a user-specified number of independent docking runs (55 in
the presented case) and writes the resulting conformations and
their energies to a molecular database file. The resulting DNA–pep-
tidyl-anthraquinone complexes were subjected to MMFF94 all-
atom energy minimization until the RMS of conjugate gradient
was <0.1 kcalmolꢀ1 ꢂꢀ1. GB/SA approximation has been used to
model the electrostatic contribution to the free energy of solvation
in a continuum solvent model.
[5] J. Kapuscinski, Z. Darzynkiewicz, Biochem. Pharmacol. 1985, 34, 4203–
4213.
[6] K. X. Chen, N. Gresh, B. Pullmann, Nucleic Acids Res. 1986, 14, 9103–
9115.
[7] F. Traganos, Pharmacol. Ther. 1983, 22, 199–214.
[8] W. W. Au, M. A. Butler, T. S. Matney, T. L. Loo, Cancer Res. 1981, 41, 376–
379.
[9] R. K. Zee-Cheng, C. C. Cheng, Drugs Future 1983, 8, 229–248.
[10] G. Palꢃ, M. Palumbo, G. A. Antonello, S. Marciani-Magno, Mol. Pharma-
col. 1986, 29, 211–217.
[11] M. Palumbo, G. Palꢃ, O. Gia, E. Ferrazzi, G. A. Meloni, Anti-Cancer Drug
Des. 1987, 1, 337–346.
[12] G. Zagotto, E. Uriarte, C. Antonello, M. T. Conconi, S. Marciani-Magno,
Bioorg. Med. Chem. Lett. 1992, 2, 659–662.
[13] B. Gatto, G. Zagotto, C. Sissi, C. Cera, E. Uriarte, G. Palꢃ, G. Capranico, M.
Palumbo, J. Med. Chem. 1996, 39, 3114–3122.
[14] B. Gatto, G. Zagotto, C. Sissi, M. Palumbo, Int. J. Biol. Macromol. 1997,
21, 319–326.
Molecular dynamics calculation: Peptidyl-anthraquinones were
immersed in a rectangular box filled with TIP3P[43] by imposing a
minimum solute-wall box distance of 10 ꢂ. The system was neu-
tralized with Na+ cations by using the AMBER leap module.[44] Each
system was modeled by using the AMBER95 all-atom force field[39]
with periodic boundary conditions. A cutoff radius of 9 ꢂ was used
for nonbonded interactions by updating the neighbor pair list
every 10 steps. The electrostatic interactions were calculated with
the Particle mesh Ewald method.[45] The SHAKE algorithm[46] was
used to constrain all bond lengths involving hydrogen. Optimiza-
tion and relaxation of solvent and ions were initially performed
while keeping the solute atoms constrained to their initial position
with decreasing force constants of 500, 25, 15, and 5 kcalmolꢀ1 ꢂ2.
Thereafter the system was minimized without any constraints and
warmed up. A 2 ns simulation was carried out. In the production
phase, the simulation was run with the isothermal-isobaric ensem-
ble (300 K and one atmosphere pressure), and pressure and tem-
perature were maintained by the Berendsen coupling algorithm[47]
with a 2 fs time step. Pressure and temperature coupling constants
were 0.5 ps. During the production run, from 0.5 to 2 ns, the atom
coordinates were saved every 0.1 ps for analysis. The MOE “Confor-
mational Geometry” module was used to check some structural
properties (root-mean-square deviation (RMSD), hydrogen bond).
Interaction energies are calculated for the entire trajectory by
using the MMFF94 force field.
[15] G. Zagotto, S. Moro, E. Uriarte, E. Ferrazzi, G. Palꢃ, M. Palumbo, Anti-
Cancer Drug Des. 1997, 12, 99–112.
[16] G. Zagotto, R. Supino, E. Favini, S. Moro, M. Palumbo, Farmaco 2000,
55, 1–5.
[17] N. Gresh, P. H. Kahn, J. Biomol. Struct. Dyn. 1991, 8, 827–846.
[18] N. Gresh, P. H. Kahn, J. Biomol. Struct. Dyn. 1990, 7, 1141–1160.
[19] S. Mohammadi, M. Perrꢄe-Fauvet, N. Gresh, K. Hillairet, E. Taillandier, Bio-
chemistry 1998, 37, 6165–6178.
[20] S. Far, A. Kossanyi, C. Verchꢅre-Bꢄaur, N. Gresh, E. Taillandier, M. Perrꢄe-
Fauvet, Eur. J. Org. Chem. 2004, 2004, 1781–1797.
[21] S. Wain-Hobson, P. Sonigo, O. Danos, S. Cole, M. Alizon, Cell 1985, 40,
9–17.
[22] L. Ratner, R. C. Gallo, F. Wong-Staal, Nature 1985, 313, 636–637.
[23] F. E. Durr in Anthracycline and Anthracenedione-Based Anticancer Agents,
Vol. 6 (Ed.: W. Lown), Elsevier, New York, 1988, pp. 401–405.
[24] R. K. Zee-Cheng, C. C. Cheng, J. Med. Chem. 1978, 21, 291–294.
[25] B. Gatto, G. Giaretta, M. Bertuzzi, E. Bianchini, G. Zagotto, M. Palumbo
in New Perspectives In Medicinal Chemistry, Vol. 1 (Ed.: S. Alcaro), Re-
search Signpost, Kerala, India, 2009, pp. 199–211.
[26] B. Gatto, O. Tabarrini, S. Massari, G. Giaretta, S. Sabatini, C. Del Vecchio,
C. Parolin, A. Fravolini, M. Palumbo, V. Cecchetti, ChemMedChem 2009,
4, 935–938.
[27] M. G. Zagorski, J. U. Bowie, A. K. Vershon, R. T. Sauer, D. J. Patel, Bio-
chemistry 1989, 28, 9813–9825.
[28] J. N. Breg, R. Boelens, A. V. George, R. Kaptein, Biochemistry 1989, 28,
9826–9833.
[29] J. R. Thomas, P. J. Hergenrother, Chem. Rev. 2008, 108, 1171–1224.
[30] B. D. Gooch, P. A. Beal, J. Am. Chem. Soc. 2004, 126, 10603–10610.
[31] J. A. Thomas, R. J. Gorelick, Virus Res. 2008, 134, 39–63.
[32] J. Johnson, A. Ahmad, S. Khan, Y.-F. Wang, A. Abu-Qare, J. Ayoub, A.
Zhang, I. Ahmad, J. Chromatogr. B 2004, 799, 149–155.
[33] V. Gibson, R. J. Anderson, T. C. Jenkins, D. Cairns, Pharmacy Pharmacol.
Commun. 1999, 5, 669–671.
[34] C. C. Cheng, R. K. Zee-Cheng, Prog. Med. Chem. 1983, 20, 83–118.
[35] S. Tyagi, D. Bratu, F. Kramer, Nat. Biotechnol. 1998, 16, 49–53.
[36] Molecular Operating Environment (MOE) v. 2008.10: Chemical Comput-
ing Group Inc., 010 Sherbrooke Street West, Suite 910, Montreal, QC,
H3A 2R7 (Canada).
Acknowledgements
This investigation was supported by the MIUR (PRIN, Rome, Italy),
AIRC (Associazione Italiana per la Ricerca sul Cancro), and the
University of Padova. S.M. is also very grateful to Chemical Com-
puting Group for the scientific and technical partnership.
Keywords: ametantrone conjugates · DNA recognition · drug
design · molecular docking · molecular dynamics
[37] MOPAC 7, J. J. P. Stewart, Fujitsu Ltd.: Tokyo (Japan), 1993.
[38] E. S. I. Carter, C. S. Tung, Comput. Appl. Biosci. 1996, 12, 25–30.
[39] W. D. Cornell, P. Cieplak, I. Bayly, I. R. Gould, K. M. Merz, Jr., D. M. Fergu-
son, D. C. Spellmeyer, T. Fox, J. W. Caldwell, P. A. Kollman, J. Am. Chem.
Soc. 1995, 117, 5179–5197.
1090
www.chemmedchem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1080 – 1091

Peptidyl-Ametantrone–DNA Complexes
[40] D. Qiu, P. S. Shenkin, F. P. Hollinger, W. C. Still, J. Phys. Chem. A 1997,
101, 3005–3014.
[41] C. A. Baxter, C. W. Murray, D. E. Clark, D. R. Westhead, M. D. Eldridge, Pro-
teins Struct. Funct. Genet. 1998, 33, 367–382.
[42] T. Halgren, J. Comput. Chem. 1996, 17, 490–519.
[43] W. L. Jorgensen, J. Chandrasekhar, J. D. Madura, R. W. Impey, M. L. Klein,
J. Chem. Phys. 1983, 79, 926–935.
[45] T. Darden, D. York, L. Pedersen, J. Chem. Phys. 1993, 98, 10089–10092.
[46] J. P. Ryckaert, G. Ciccotti, H. J. C. Berendsen, J. Comp. Phys. 1977, 23,
327–341.
[47] H. J. C. Berendsen, J. P. M. Postma, W. F. Van Gunsteren, A. Dinola, J. R.
Haak, J. Chem. Phys. 1984, 81, 3684–3690.
Received: March 15, 2010
Revised: March 22, 2010
[44] D. A. Pearlman, D. A. Case, J. W. Caldwell, W. S. Ross, T. E. Cheatham, S.
Debolt, D. M. Ferguson, G. Seibel, P. A. Kollman, Comp. Phys. Commun.
1995, 91, 1–41.
Published online on May 10, 2010
ChemMedChem 2010, 5, 1080 – 1091
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1091