Trisubstituted acridines as G-quadruplex telomere targeting agents. Effects of extensions of the 3,6- and 9-side chains on quadruplex binding, telomerase activity, and cell proliferation

Article abstract of DOI:10.1021/jm050555a

The synthesis is reported of a group of 3,6,9-trisubstituted acridine compounds as telomeric quadruplex-stabilizing ligands with systematic variations at the 3-, 6-, and 9-positions. A new microwave-assisted methodology has been developed for trisubstituted acridine synthesis. Structure-activity relationships are reported using surface plasmon resonance and a fluorescence melting assay to examine quadruplex binding, together with a telomerase inhibition assay. These reveal relationships between G-quadruplex stabilization and telomerase inhibition and optimal 3,6- and 9-substituent side-chain lengths for maximal activity. Qualitative molecular modeling using molecular dynamics simulations has been undertaken on four quadruplex-DNA complexes. Long-term exposure of MCF7 cancer cells to a subset of the most active compounds, at doses lower than the IC₅₀ values, showed that one compound produced a marked decrease in population growth, accompanied by senescence, which is consistent with telomere targeting by this agent.

Full text of DOI:10.1021/jm050555a

582
J. Med. Chem. 2006, 49, 582-599
Trisubstituted Acridines as G-quadruplex Telomere Targeting Agents. Effects of Extensions of
the 3,6- and 9-Side Chains on Quadruplex Binding, Telomerase Activity, and Cell Proliferation
Michael J. B. Moore,†,‡ Christoph M. Schultes,†,§,‡ Javier Cuesta,†,| Francisco Cuenca,† Mekala Gunaratnam,†
Farial A. Tanious, W. David Wilson, and Stephen Neidle*,†
Cancer Research UK Biomolecular Structure Group, The School of Pharmacy, UniVersity of London, 29-39 Brunswick Square,
London WC1N 1AX, UK, and Department of Chemistry, Georgia State UniVersity, Atlanta, Georgia 30303-3083
ReceiVed June 13, 2005
The synthesis is reported of a group of 3,6,9-trisubstituted acridine compounds as telomeric quadruplex-
stabilizing ligands with systematic variations at the 3-, 6-, and 9-positions. A new microwave-assisted
methodology has been developed for trisubstituted acridine synthesis. Structure-activity relationships are
reported using surface plasmon resonance and a fluorescence melting assay to examine quadruplex binding,
together with a telomerase inhibition assay. These reveal relationships between G-quadruplex stabilization
and telomerase inhibition and optimal 3,6- and 9-substituent side-chain lengths for maximal activity.
Qualitative molecular modeling using molecular dynamics simulations has been undertaken on four
quadruplex-DNA complexes. Long-term exposure of MCF7 cancer cells to a subset of the most active
compounds, at doses lower than the IC50 values, showed that one compound produced a marked decrease
in population growth, accompanied by senescence, which is consistent with telomere targeting by this agent.
Introduction
identified as one of the key stages in oncogenic transformation.15
Telomerase function and telomere capping have thus developed
into promising targets for anticancer chemotherapies.16
In addition to the targeting of the enzyme itself, a number of
ways of influencing the DNA substrate of the enzyme itself
have been studied, thus opening the way to compounds with a
dual role as initiators of telomere uncapping and telomerase
inhibitors. These two functions would be expected to result in
short-term and long-term effects respectively, by contrast with
other types of telomerase inhibition, which have focused on the
latter, with attendant problems of an extended time lag.
Telomere status is maintained and regulated in all cells
through a complex capping mechanism that prevents chromo-
somal ends from being recognized as damaged DNA and allows
faithful chromosome replication during the cell cycle.1,2 A host
of telomere-associated proteins (including TRF1, TRF2, and
POT1) ensure that the telomere, which typically consists of
several kilobasepair of double-stranded, G-rich DNA and a
100-200 bp single-stranded overhang on the 3′-strand, does
not elicit DNA-damage response pathways or lead to abnormal
chromosomal rearrangements.3-5 Exposure of the 3′-end due
to uncapping results in cellular senescence and apoptosis.6,7
The single-stranded telomeric overhang which is used by
telomerase as a primer during the elongation phase of its action
is capable of forming four-stranded DNA G-quadruplex struc-
tures.17,18 In vitro, the use of small molecule ligands to stabilize
such structures in DNA sequences mimicking the telomeric 3′-
overhang has been shown to result in the inhibition of telomerase
activity.19-23 Furthermore, cell culture experiments with a
number of these compounds have been shown to have pro-
nounced effects on cancer cell lines,24-26 suggesting that
stabilization of such structures is incompatible with the correct
functioning of telomerase and cancer cell telomere maintenance
in general, thereby causing destabilization of the telomere cap
and associated processes. The postulated mechanism of action
of these compounds is a denial of access to the telomere, both
for telomerase and other telomere-associated proteins, such as
the length-mediating single-strand-binding protein POT1.27-29
This can then lead to uncapping,25 causing chromosomal end-
to-end fusions,30 cell cycle arrest, and growth inhibition. Other
studies have also shown that the stabilization of quadruplexes
with small molecules hinders the ability of telomere-associated
helicases such as BLM and WRN to unwind these higher-order
structures,31-33 again impacting on the correct functioning of
the telomere cap. Recent in vivo findings in a mouse xenograft
model34 have demonstrated that exposure to the trisubstituted
acridine BRACO19 (1a) is associated with tumor growth
inhibition and killing of cancer cells, accompanied by atypical
mitoses in xenograft tissues, indicative of telomere dysfunction.
Analysis showed a loss of nuclear hTERT upon treatment with
The enzyme telomerase plays a crucial role in capping
assembly.2,8 It allows the extension of telomeres shortened over
successive rounds of replication due to the end-replication effect,
and in the absence of this process, telomeres are observed to
become critically short after a number of rounds of replication,
thus triggering growth arrest and replicative senescence or
apoptosis.9-11 It can thus be seen to work on two levels, the
more basic level being templating and extension of the telomere
3′-overhang while the higher level involves more general
participation in maintaining telomere homeostasis and other
putative roles in DNA-damage signaling pathways.
Telomeres are differentially regulated in normal somatic cells
compared to cancer cells, primarily through differing levels of
active telomerase enzyme. The correlation between the detection
of high levels of active telomerase in cancer cells and its absence
or down-regulation in surrounding noncancerous cells has been
well established,12-14 and the activation of telomerase has been
* Corresponding author. Tel 44 207 753 5969. Fax 44 207 753 5970.
E-mail: stephen.neidle@pharmacy.ac.uk.
† University of London.
‡ These authors contributed equally to this work.
§ Present address: Department of Molecular Oncology, Go¨ttinger Zen-
trum fu¨r Molekulare Biowissenschaften (GZMB), Justus von Liebig Weg
11, 37077 Go¨ttingen, Germany.
| Present Address: Max Planck Institut fu¨r Molekulare Physiologie, Otto
Hahn Strasse 11, 44227 Dortmund, Germany.
Georgia State University.
10.1021/jm050555a CCC: $33.50 © 2006 American Chemical Society
Published on Web 12/21/2005

G-Quadruplex Stabilization and Telomerase Inhibition
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2 583
heterogeneous reaction mixtures from which only a small
amount of product acridone (2% in the case of 8) and unreacted
starting material were obtained. Use of pyrrolidine as solvent
led to complete consumption of the starting material; however,
isolation of solid acridone materials from concentrated reaction
mixtures by partition between basic aqueous and organic phases
proved frustrating and irreproducible at times, as highlighted
in the synthesis of 8, where an oily residue was frequently
encountered. Employing cosolvents such as DMF and acetoni-
trile produced similar results, with hard to manipulate oily
residues being produced. Analysis of reaction products by LC/
MS indicated peaks [(M - C4H9N)H+, (M - C4H9N)H22+, (M
- (C4H9N)2)H+, (M - (C4H9N)2)H2+] consistent with species
arising from elimination of a pyrrolidine fragment (C4H9N).
Given our concerns regarding the questionable stability of
9-chloroacridines, it was important at this stage of the synthesis
to obtain pure acridone. Given the notorious insolubility of
acridones, a reputation upheld by 8-10, practical purification
methods are limited to washing or recrystallization with alcohol
or alcohol/DMF mixtures. These techniques, however, have not
proved effective in the purification of acridones 8-10.
Figure 1. Trisubstituted anilinoacridines.
compound and the colocalization of cytoplasmic hTERT with
ubiquitin, suggesting that the induced displacement of the
enzyme from the telomere targets it for enhanced degradation.
Inhibition of telomerase is thus an important effect associated
with telomere-targeting agents, but the full mechanism of action
is expected to occur through multiple pathways involving the
onset of senescence and apoptosis as well as the elicitation of
a DNA-damage response. Although early models of telomerase
inhibition suggested that a drawback of telomerase-targeted
therapies would be the extended time lag required before
telomeres reached critically shortened lengths,16,35-37a clear
evidence is emerging that G-quadruplex telomere-targeted agents
can influence events at the telomere on a much shorter time
scale, both in vitro and in vivo.34,37b
A large number of mostly planar, aromatic compounds have
been shown to be G-quadruplex-stabilizing ligands, including
those based on a central acridine scaffold.19-21 We have
previously reported on a number of such trisubstituted acridine
analogues with a variety of side-chain modifications and
stereoisomer variations, with the best compounds exhibiting
strong G-quadruplex binding, high selectivity for quadruplex
over duplex DNA, and accompanying telomerase inhibitory
activity in the nanomolar range.19,20,38,39 We present here data
from biophysical, biochemical, and cellular studies showing that
selective binding of 3,6,9-anilinoacridines (Figure 1) to qua-
druplex telomeric DNA correlates with telomere targeting and
telomerase inhibition and that there are optimal 3-, 6-, and
9-side-chain lengths for maximal activity. In addition to the use
of two biophysical methods (FRET and SPR assays) and the
biochemical TRAP assay, molecular dynamics simulations were
used to examine qualitative structure-activity relationships for
interaction with quadruplex DNA, with respect to variations at
the 3- and 6-positions.
The impurities associated with acridones 8-10 were found
to be transmitted along the synthesis pipeline. Analysis by LC/
MS of 9-chloroacridine intermediates 12-14 again indicated
the presence of elimination products and respective acridone
8-10, which had either remained unreacted or resulted from
hydrolysis of the 9-chloroacridine when isolated from the
reaction mixture. As feared, the 9-chloroacridines 12-14 proved
to be unstable, in contrast to 11, which can be stored, not
permitting purification at this point in the synthesis, so 12-14
were submitted immediately to the final couplings with excess
aniline. Target acridines were found to be thermally labile with
elimination of pyrrolidine to provide an olefinic product as
1
observed by H NMR analysis. Heating was restricted in the
final coupling reaction to apprehend this elimination but with
the effect of decreasing reaction efficiency. Crude mixtures were
obtained whose components included target acridine, aniline,
acridone, and elimination products from which pure target
acridines were extruded by repeated flash chromatography. Thus,
the series 1b-d (R ) N(CH3)2, n ) 3-5) and 1h-j (R )
3-(pyrrolidin-1-yl)propionamidoanilino, n ) 3-5) were prepared
in respective overall yields of 1-4% and 2-6% from 2, which
are lower than that observed for the synthesis of 3,6-bis-
propionamidoacridines 1e-g (n ) 2) (overall yield 15-16%).
Results
It was considered that problems associated with this synthetic
route may be bypassed by employing a route centered on 3,6-
diazido-9-chloroacridine (16), readily available from 2 by
diazotization and then treatment with thionyl chloride.42 In a
step that introduces the 9-substituent earlier in the synthesis,
compound 16 was coupled with the aniline 15b to give pure
9-substituted diazido acridine (17) in 83% yield [cf. an average
yield of 13% when introducing the 9-substituent later in the
synthesis (Scheme 1)]. Catalytic hydrogenation of 17 proceeded
smoothly (67%) to give 18. With the application of microwave
heating, the 3,6-bis(chloroalkylamido) derivatives 19-21 were
able to be prepared in parallel with yields ranging from 60 to
80%. Reaction with pyrrolidine at only room temperature
furnished pure target acridine in an average yield of 57%, after
only one round of flash chromatography.
Chemistry. Following methodology adapted by us to a
generic synthesis of 1 (n ) 2),20,40 acridines 1e-g (where n )
2) were prepared via the key 9-chloroacridine intermediate 11
(Scheme 1). The diaminoacridone 241 was acylated with
3-chloropropionyl chloride followed by amination with pyrro-
lidine and activation with phosphorus pentachloride and phos-
phoryl chloride to give 11 typically with an overall yield of
27% over the three steps. Heating a methanolic solution of 7
with the required aniline 15a-c, prepared in a three-step
sequence from available 4-nitroaniline, furnished pure 1e-g
typically in 55-61% yield from 11 (overall yield from 2 15-
16%) after purification by repeated flash chromatography.
The synthesis of 3,6-extended acridines 1b-d and 1h-j was
effected with modifications necessary to dissect differences in
reactivity and solubility of acridones (4-6) and stability of the
9-chloroacridines (11-14) relative to compounds 7 and 11,
respectively.
Comparing the two synthetic routes presented, we find that
acridines 1h, 1i, 1j are prepared in higher overall yield (19%,
24%, 9%, respectively) by employing the diazido route (Scheme
2) compared to the 9-chloroacridine route (Scheme 1) (3%, 6%,
2%), which is notable as the diazido route is a longer synthesis
Initial attempts at the synthesis of acridones 8-10 employed
conditions developed for the preparation of 7.19 However, these
conditions (heating with excess pyrrolidine in ethanol) gave

584 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2
Moore et al.
Scheme 1. 9-Chloroacridine Route to 3,6,9-Trisubstituted Acridinesa
a (i) Cl(CH2)n)2-5COCl, 80 °C; (ii) pyrrolidine, ∆; (iii) PCl5, POCl3, ∆; (iv) 15a-c or N,N-dimethylaminoaniline, MeOH, ∆.
by two steps. In addition, the diazido route is advantageous as
cleaner chemistries are employed.
In terms of the 3,6-extended analogues 1f and 1h-j, the 3,6-
bis-butanamido derivative 1h gave a relatively low FRET value
(∆Tm ) 22.63 °C), whereas the shortest 3,6-bis-propionamido
derivative 1f and longest 3,6-bishexanamido derivative 1j
showed the greatest stabilization. The 9-modified acridines 1e
and 1g compare well with the 3,6-modified acridines 1i and 1j
and have markedly higher ∆Tm values than 1h.
TRAP Assays of Telomerase Activity. Compounds 1a and
1b both show sub-micromolar inhibition of telomerase activity,
in agreement with an earlier study on 1a.19,20 The telEC50 value
for this compound reported here (113 nM) is within experimental
error of the two previously reported values (95 and 115 nM,
respectively). A large decrease in telomerase inhibitory activity
was found with increasing side-chain elongation (Table 1). This
closely parallels the trend in the FRET assay data for these four
compounds but is much more striking in the TRAP data, with
1c and 1d exhibiting telEC50 values of 1.9 and 6.9 µM,
respectively.
FRET Assays of Quadruplex Binding. The melting curves
obtained for compounds 1a-d were all of the same shape with
a major transition, yielding a [conc]∆T )20°C value and a small,
m
minor transition occasionally observed at lower temperatures
(Table 1). Only the major melting transitions are reported here,
as the smaller transitions were not large enough to yield
consistently accurate temperature measurements. For the main
melting step, the trend across the series was one of decreasing
quadruplex stabilization with increasing side-chain length.
Comparing FRET results for extended acridines 1a-d and
1e-j has been hampered by the inconsistent shape of some
melting curves, which may be quantified by the poor cor-
respondence between ∆Tm and [conc]∆T
values. In some
m)20°C
cases it is difficult to discern obvious trends. For 3,6-bis-
propionamido acridines modified at the 9-position, 1e-g, there
is a slight increase in ∆Tm with 9-substituent length that plateaus,
although such a trend does not hold when comparing the
Acridine 1f was found to be the most potent telomerase
inhibitor (telEC50 ) 67.4 nM) of the compounds presented in
[conc]∆T
parameter.
m)20°C

G-Quadruplex Stabilization and Telomerase Inhibition
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2 585
Scheme 2. Diazido Route to 3,6,9-Trisubstituted Acridinesa
a (i) NaNO2, dilute H2SO4, NaN3, 0 °C to rt; (ii) SOCl2, rt; (iii) 15b, catalytic HCl, NMP; (iv) H2 20 psi, 10% Pd-C, MeOH/EtOAc; (v) Cl(CH2)n)3-5COCl,
TEA, toluene, µW; (vi) pyrrolidine, rt.
Table 1. Quadruplex FRET and TRAP Dataa
Table 2. SPR Dataa
∆Tm(1µM)
(°C)
[conc]∆T
error
(µM)
telEC50
(nM)
error
(nM)
ratio
(KG4DNA/
KdsDNA)
m)20°C
compd
(µM)
KdsDNA
KG4DNA
compd
(×106 M-1
)
(×106 M-1
)
1a
1b
1c
1d
1e
1f
1g
1h
1i
27.5
16.5
6.8
0.7
1.1
2.1
2.6
0.3
0.1
0.3
0.7
0.3
0.3
2.5
13.8
0.1
0.1
0.1
0.1
0.02
0.01
0.01
0.08
0.01
0.00
0.4
113
99
1930
6910
167
67
117
326
255
1
2
1a
1b
1c
1d
1e
1f
1g
1h
1i
1.0
0.1b
0.4
0.1b
6.8
1.0
6.0
5.0
0.7
0.3
1.1
nd
31.0
4.8
1.8
31.0
48.0
4.5
320
1400
16.1
1.5
7.9
16.1
23.3
1.8
nd
4.7
34.8
37.6
37.9
22.6
36.2
39.3
14.6
5.5
0.6
5.8
74.0
29.0
77.0
3.0
1.1
18.0
1.3
10.9
29.0
12.8
0.6
1.6
72.0
1.2
1j
22
23
146
5200
>20000
1j
22
23
1.7
nd
nd
nd
a ∆Tm(1µM), change in melting temperature at 1 µM drug concentration;
[conc]∆T )20°C, concentration of compound required to effect a 20 °C change
a Fitting errors for determination of K with a particular SPR data set are
(5%. Experimental errors in K based on reproducibility in repeat experi-
ments are (20% for the quadruplex studies. KdsDNA, binding constant of
compound to double-stranded DNA; KG4DNA, binding constant of compound
to quadruplex DNA; KG4DNA/KdsDNA, ratio of KdsDNA to KG4DNA; nd, not
determined. b Note that values of KdsDNA for 1b and 1d are upper limits.
Thus KG4DNA/KdsDNA values are likely to be greater than 48 and 5.8,
respectively.
m
in melting temperature; telEC50, concentration of compound required to
achieve 50% inhibition of telomerase inhibition; nd, not determined
this study. A propionamide-based linker in the 9-position is
optimal in this series, as the compounds with a longer 9-position
side chain [1e (acetamide) and 1g (butanamide)] have potencies
a factor of 2 worse than 1f. Analogous to the observations in
the FRET studies, 1h shows a significant decrease in inhibitory
potency, which nevertheless improves as the 3- and 6-side chains
become longer, although the effect is more pronounced in the
TRAP than in the FRET assay. Acridines 1e and 1g are roughly
equipotent with 1j (telEC50 values of 100-200 nM). Interest-
ingly, compound 1i, which displays similar G4 DNA stabiliza-
tion as 1f, is a less potent telomerase inhibitor. Telomerase
inhibition correlates well overall with the FRET data, providing
evidence for the utility of this technique in screening the potency
of potential telomerase inhibitors.
tion range examined, ligands bound significantly more tightly
to quadruplex DNA and had slower kinetics than binding to
duplex DNA, in particular with respect to dissociation (Figure
2a,b). As can be seen for compound 1a in Figure 2b, the duplex
curves reached a steady-state plateau soon after injection of the
compound. With the quadruplex, however, at the lowest
concentrations, it took significantly longer to reach a steady state
plateau (Figure 2a; note the different time scales in parts a and
b of Figure 2). The association reactions are bimolecular with
rates that are concentration-dependent, but the first-order
dissociation reactions can be directly compared. The kinetics
differences between the duplex and quadruplex reactions are
apparent in the plots. The steady-state RU (response units)
values were determined in the plateau regions of the sensorgrams
and are plotted versus the free compound concentration (from
the flow solution) in Figure 2c. Fitting of binding curves using
Surface Plasmon Resonance Studies of Quadruplex Af-
finity. SPR yields direct information on the binding of the
ligands to both the human quadruplex and duplex DNA and
thus enables the determination of the ratio of the two equilib-
rium-binding constants (KG4DNA/KdsDNA; Table 2), which is a
measure of G-quadruplex binding selectivity. In the concentra-

586 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2
Moore et al.
steric bulk of these ligands at the 3- and 6-positions is both
unfavorable for quadruplex binding and detrimental to duplex
affinity. Conversely, 1e and 1g exhibit the largest values for
KG4DNA but show lower ratios due to an increase in duplex
binding. The importance of duplex-quadruplex selectivity and
the favorable effect of trisubstitution was also evident when all
the 3,6,9-acridines studied were compared to the disubstituted
analogue 22, which demonstrated a roughly equal affinity for
both forms of DNA (Table 2).
Experiments on quadruplex binding kinetics were also
performed with 1a, although only sensorgrams at lower
concentrations that are dominated by the single strong binding
site were analyzed. As the kinetics for binding to duplex DNA
were too rapid to be accurately assessed by SPR methods, only
quadruplex binding was examined. Both the association and
dissociation rates were lower for the acridine-quadruplex
interaction than for the corresponding acridine-duplex complex,
but the larger affinity constants for quadruplex DNA were
mainly derived from the effect on the dissociation kinetics. The
binding constants calculated using the kinetics methods were
within experimental error of those determined by steady-state
measurements (data not shown).
Molecular Modeling. Molecular dynamics simulations were
performed on four G-quadruplex ligand complexes, with
compounds 1a-d (Figure 3a-d). An initial simulation for the
native quadruplex resulted in an average rmsd value of 1.59 Å
for the DNA residues 1-21 over the whole trajectory, which is
similar to the rmsd for the loop residues (average rmsd 1.52
Å). The rmsd of the G-quartets remained steady and low
throughout the course of the simulation, with an average value
of 0.58 Å, in accordance with previous reports on long-term
quadruplex simulations.43-45 The TTA loops, on the other hand,
showed much higher variation, again in accord with previous
simulation studies.45,46
Figure 2. (a, b) SPR sensorgrams for the interaction of compound 1a
with the human telomere G-quadruplex hairpin DNA (a, left) as well
as a duplex hairpin with a CGAATTCG stem sequence (b, right) are
shown. The concentrations of the compounds increase from 1 × 10-9
M (lower curve) to 8 × 10-7 M (highest curve) in both plots. The
experiments were conducted at 25 °C in HEPES buffer with 200 mM
K+. (c) Binding isotherms are shown for compound 1a with the -AATT-
duplex (blue squares) and G-quadruplex (black circles) and were
constructed with results from experiments such as those in a and b.
The RU values from the steady-state region of the sensorgrams were
plotted versus the concentration of the compounds. The lines in the
figures were obtained by nonlinear least-squares fits of the data to a
two-site-binding model.
The MD runs for the four ligand complexes showed that the
presence of the ligand has only a minor impact on the flexibility
of the quadruplex structure, apart from compound 1d. Its longer
side chains make favorable contacts with the loops, thereby
reducing the flexibility of the DNA backbone at those positions.
The behavior of the quadruplex-1a complex illustrates two
phenomena that have a significant impact on quadruplex
flexibility. The rotational barrier for the C-N bond between
the acridine and the 3- and 6-substituted amide appears to be
relatively low, and the amide can therefore “reverse” (with
respect to the starting conformation) into a conformation
whereby the carbonyl oxygen faces toward the center of the
uppermost quartet rather than away from it. This causes a shift
in the entire side chain and subsequently in the chromophore
position, resulting in a tighter “crescent” shape for the 3,6-
acridine, forcing the substituent in question to make new
contacts with the DNA. Although this did not result in major
changes in quadruplex structure, it indicates that the pyrroli-
dine-phosphate interactions made by the 3,6-side chains are
not very stable and can be influenced by intramolecular changes
in conformation, possibly due to nonoptimal side-chain length.
a two-binding-site model suggests that the acridines all had a
single strong binding site on the quadruplex with a weaker
secondary binding interaction associated with a 10-20-fold
lower affinity. This weaker interaction could not be accurately
evaluated in the experimental concentration range and only the
stronger binding constants are reported here. With the other
compounds, the kinetics of quadruplex dissociation are faster
and the RU values at each concentration are significantly lower
than with 1a. Sensorgrams for 1b are shown as examples in
the Supporting Information. The faster dissociation of 1b relative
to 1a can be seen, and the duplex shows very little binding of
1b. The duplex binding constants for 1b and 1d are quite low
and are upper limit estimates.
All K values determined by the steady-state method are given
in Table 2. KdsDNA values are in a smaller range (from 1 × 105
to 6.8 × 106 M-1) compared to KG4DNA values, which showed
greater variation, from 6 × 105 M-1 (for 1d) to 7.7 × 107 M-1
(for 1g). Elongation of the 3- and 6-side chains significantly
decreased quadruplex affinity, both in the group of compounds
containing an uncharged 9-side chain (1a-d) and in the set
containing the charged terminal pyrrolidine group at the
9-position (1e-j). The ratio of quadruplex-to-duplex binding
was still very favorable for 1b, due to its diminished duplex
affinity; however, this effect is even more noticeable for the
3,6-hexanamido derivative 1j, which due to its low dsDNA
affinity displayed the best quadruplex selectivity of all the
ligands studied here. This suggests that in general the added
Conformational changes in the loops also have an impact on
ligand position. As was seen for the native quadruplex, the loop
regions relaxed during the simulation and lost some of their
initial structure, accounting for their generally larger rmsd. As
the three TTA loops expanded during the simulations, the C5
methyl groups of the thymines tended to become more deeply
buried within the pocket, thereby making the groove on the
ligand-facing side of the loop shallower and displacing the
pyrrolidine group of the ligand. This caused disruption of the

G-Quadruplex Stabilization and Telomerase Inhibition
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2 587
Figure 3. Average (minimized) structures extracted from particular segments of the MD simulations for the ligand-DNA complexes containing
compounds 1a,b. Segments were chosen upon visual inspection of the trajectories. (a) 1a, red (0-375 ps), green (500-575 ps), blue (925-1000
ps); (b) 1b, red (0-200 ps), green (250-400 ps), blue (500-650 ps), magenta (700-900 ps); (c) 1c, red (0-150 ps), green (375-550 ps), blue
(575-700 ps), magenta (875-975 ps); and (d) 1d, red (200-300 ps), green (350-450 ps), blue (675-750 ps), magenta (800-950 ps).
loop contacts made by the 3- and 6-substituents, and although
the third loop demonstrated equivalent motion, the 9-substituent
was not buried deeply enough in the pocket to be affected.
Elongation of the side chains by one methylene with respect
to 1a (ie compound 1b) appeared to overcome both of these
problems to some extent, as the differing 3- and 6-side chains
make contacts that are neither affected by the loop thymines
nor prone to the amide reversal observed for 1a. Further
increases in the length of the side chains to give compounds 1c
and 1d caused greater variation in the ligand motion on the
quadruplex surface due to the greater flexibility of the 3- and
6-side chains (Figure 3c,d). It was noticeable for 1c that the
conformation of one of the 3,6-substituents differs with respect
to both 1a and 1b. The greater length caused it to straighten
out and probe the loop more thoroughly, whereas the shorter
groups of both the latter compounds caused interactions close
to the upper lip of the groove. Furthermore, for 1a-c the tight
binding of one of the 3,6-side chains was usually offset by looser
binding of the other substituent to the loop with which it was
in contact, accompanied by a correspondingly increased fluctua-
tion over the course of the simulation. With 1d, the longer 3,6-
side chains made tighter interactions with the DNA loops, and
the ligand position was considerably more stable over the course
of the simulation, as can be seen for averaged snapshots taken
during the experiment (Figure 3d).
cell death is the result of general, and irreversible, cell damage
can be determined by the calculation of an IC50 value. Those
for 1b were 8.75 µM in the MCF7 breast carcinoma cell line
and >25 µM in the IMR90 noncancer fibroblast line. In the
latter cell line, compound 1a showed a similar lack of toxicity
(IC50 > 25 µM), while results from six different tumor cell lines
have been previously determined for it (the average IC50 is 10.6
( 0.7 µM).24 Compound 1f exhibited a small decrease in cell
growth over the 96-h assay period, although 50% cytotoxicity
was not reached in the concentration range examined and the
IC50 values in the MCF7 and IMR90 cell lines were thus >25
µM. For all other compounds with 3,6- and 9-position modifica-
tions, IC50 values were .25 µM in the MCF7 and IMR90 cell
lines.
A series of long-term studies was undertaken in order to
assess the effects of selected compounds when administered at
concentrations below the IC50 values. Compound 1b produced
a large decrease in the rate of population growth in 4-week
exposure experiments, even at concentrations considerably
below its IC50 value in the MCF7 cell line (Figure 4a). At a 5
µM concentration, cell numbers were insufficient after the third
week for a continuation of the experiment, with a drop in the
cell number recorded after week 2. At 2 µM, a delayed effect
was observed, with the number of population doublings falling
significantly after 2 weeks, and even at 1 µM exposure, a
discernible difference was observed between the control and
treated cells. The rate of population growth remained relatively
low thereafter, although there were indications that the cells
may have started to recover after days 21-28.
Cell Biology Studies. Long-term cell viability studies were
undertaken on two compounds, representative of the most potent
telomerase inhibitors in each subgroup, 1a-d and 1e-j. Cellular
responses to exposure were examined in order to establish short-
term cytotoxicity profiles prior to these studies. The sulfor-
hodamine B (SRB) assay is a widely used method for the
assessment of cellular chemosensitivity in a high-throughput
format.47,48 The upper threshold level of cytotoxicity at which
Compound 1f produced only a minor decrease in the short-
term growth rate of MCF7 cells with respect to vehicle controls
(Figure 4c), despite being a significantly more potent telomerase
inhibitor, as judged by its telEC50 value. As in previous

588 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2
Moore et al.
Figure 4. (a). Plot of population doublings (PD) against time for cellular exposure to compound 1b (three separate concentrations; VC, vehicle
control). Long-term (4-week) cell growth experiments for 1b were performed using the breast adenocarcinoma cell line MCF7, with cells exposed
twice per week and counting carried out weekly. (b) Percentage of total cells counted staining for senescence-associated â-galactosidase (SA-â-
Gal) activity upon long-term (4-week) exposure to 1b at varying concentrations. * indicates that counting of stained cells following exposure to 1b
was difficult due to the shrunken morphology of the cells; see the text for further details. (c) Plot of population doublings (PD) against time for
cellular exposure to compound 1f (three separate concentrations; VC, vehicle control). Long-term (4-week) cell growth experiments for compound
1f were performed using the breast adenocarcinoma cell line MCF7, with cells exposed twice per week and counting carried out weekly. (d)
Percentage of total cells counted staining for senescence-associated â-galactosidase (SA-â-Gal) activity upon long-term (4-week) exposure to 1f at
varying concentrations. Cells were stained once per week following twice-weekly exposure.
experiments, no effect was observed in the first week of
exposure and the overall divergence from the controls was not
large. However, a decrease in the proliferation rate was noted
after the second week at 5 and 10 µM concentrations, although
the results suggest that after 3-4 weeks the growth rate of the
exposed cells was gradually returning to pre-exposure levels.
as a tendency for the fraction of stained cells to be higher in
regions of dense clustering. This may be due to paracrine effects
associated with the expression of senescence-inducing genes
in cells that have already undergone senescence themselves.51
Analogous four-week experiments carried out with 1f were
insufficient to establish any significant trends in senescence
staining over longer time periods. An increase in the number
of cells staining for senescence was initially observed after
which cell populations treated with 10 µM compound consis-
tently showed around 20% senescent cells (Figure 4d). The
corresponding values at 1 and 5 µM were lower, decreasing to
just above vehicle control levels after 28 days. This supports
the gradual return to control proliferation rates at these
concentrations after 3-4 weeks as observed in the growth rate
experiments. Morphologically, no shrinkage or distortion of
normal cell shape was observed even after 4 weeks of treatment
with 1f, and staining was consistent with what was observed
for compound 1b.
Senescence was estimated with the SA-â-Gal assay, in parallel
with the long-term tissue culture studies, using a fraction of
cells set aside after counting. The number of cells stained for
senescence thus corresponds directly to the data recorded in
the plots describing the population doubling behavior of the
cell populations. A clear increase in the percentage of cells
staining for SA-â-Gal was observed with respect to controls,
even at a 1 µM concentration for compound 1b (Figure 4b). At
higher concentrations, staining and counting of the SA-â-Gal
expressing fraction was affected by the morphology of the cells
following exposure. At 2 µM and above, shrunken, shriveled
cells were observed, making it difficult to distinguish the
characteristic blue coloration upon staining. Cells staining for
SA-â-Gal were observed to show localized staining in the
nucleus, the intensity of color ranging from faint at low
compound concentrations to increasingly intense as the amount
of compound was increased. Additionally, populations exposed
to compound displayed a proportion of stained cells exhibiting
an enlarged and flattened morphology that has been described
as a characteristic feature accompanying senescence,49,50 as well
Discussion
Biophysical and in Vitro Assays. The biophysical FRET
and SPR assays, together with the TRAP assay, all show that
the side-chain length in this series of trisubstituted acridines is
optimal for quadruplex binding and telomerase inhibition when
n ) 2 or 3 (i.e., compounds 1a and 1b, respectively).
Additionally, the effect of the third, 9-position substituent on

G-Quadruplex Stabilization and Telomerase Inhibition
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2 589
Figure 5. Biophysical correlations for compounds 1a-d and 1e-j. (a, b) FRET-SPR correlations; comparisons of [conc]∆T
and KG4DNA
)20°C
m
would be expected to show an inverse correlation, with strong quadruplex binders demonstrating low FRET values (red bars) and high SPR values
(black bars). (c, d) FRET-TRAP correlations; changes in [conc]∆T )20°C and telEC50 values should be directly proportional, with potent compounds
m
showing both low FRET values (red bars) and TRAP values (black bars). (e, f) TRAP-SPR correlations; as with the FRET-TRAP correlation, a
favorable KG4DNA/KdsDNA ratio (black bars) translates into increased telomerase inhibition in the TRAP assay (black bars). ∆Tm(1µM), change in
melting temperature at 1 µM drug concentration; [conc]∆T )20°C, concentration of compound required to effect a 20 °C change in melting temperature;
m
telEC50, concentration of compound required to achieve 50% inhibition of telomerase inhibition; KdsDNA, binding constant of compound to double-
stranded DNA; KG4DNA, binding constant of compound to quadruplex DNA; KG4DNA/KdsDNA, ratio of KdsDNA to KG4DNA
.
the acridine scaffold is highly additive, as can be seen for the
data obtained for the disubstituted acridine analogue 22.
We have compared the data derived from the FRET and SPR
experiments, since both provide measures of ligand-quadruplex
binding affinity. Although the FRET assay gives an indirect
estimate of quadruplex binding, since stabilization of the
quadruplex may not be directly related to binding affinity, a
indoloquinolines,52 benzoindoloquinolines,22 dibenzophenan-
throlines,53 and ethidium derivatives.54 The data for the present
series confirm that an improvement in quadruplex stabilization
tel
as determined by FRET corresponds directly to enhanced
-
EC50 values (Figure 5c,d), again providing support for high-
throughput FRET as a screening approach.
Addition of ligand to the DNA probe in FRET experiments
was also observed to cause the appearance of a minor melting
step prior to the main melting process from which the Tm values
are derived in a number of experiments. These compounds thus
increase the stabilization of the DNA structure as a whole, yet
also cause minor local destabilizing effects. This may be due
to interactions with the loop regions, which are likely to be less
stable than the central guanine-stacked core and may transmit
small effects to the fluorophores through small rearrangements
of the G-stacks due to changes in loop topology. The effects of
changes in loop structure on quadruplex stability have been
investigated using both biophysical55 and computational45,46
approaches, and they may have an important impact on
quadruplex targeting in the future as compounds designed to
interact more efficiently with the quadruplex loop sequences
plot of [conc]∆T
against KG4DNA for the ligand set
m)20°C
nevertheless shows that they have an inverse proportionality
relationship (Figure 5a,b), giving strong support to the use of
high-throughput screening of compounds using FRET as an
indicator of quadruplex affinity.39,52 Compound 1f is the
outstanding outlier from this pattern, with anomalously low
duplex and quadruplex binding indicated by the SPR studies,
which contrast with its high potency in both the FRET and
TRAP experiments. The second correlation that we observe is
between the FRET and TRAP data. The postulated mechanism
of action implies that increased quadruplex stabilization results
in more potent telomerase inhibition. This has been found
previously for structurally related 3,6,9-trisubstituted acridines,39
as well as for more structurally diverse molecules, such as

590 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2
Moore et al.
are investigated. These data also gives some insight into changes
in quadruplex structure upon melting and the steps involved in
this process. Computational studies have attempted to assess
the importance of various intermediates in the folding of
quadruplexes,56 and conversely, some of these processes may
be of importance in understanding quadruplex unfolding. One
of the principle factors governing quadruplex formation and
stabilization is the role of the monovalent cations that coordinate
within and between the G-stacks of the quadruplex core.56-58
Loss of cations from within the core causes significant
destabilization of the quadruplex structure. Compounds such
as 9-anilinoacridines, which, at physiological pH, contain
cationic chromophores capable of interacting with the uppermost
G-quartet through “pseudocation” effects,59 may therefore exert
their stabilizing effect partly by blocking, or diminishing, cation
loss from between the G-quartets.
chain structure may significantly alter the binding mode of the
compounds on the G-quartet surface.
The results presented here highlight a number of structural
issues related to ligand-DNA interactions that could usefully
be addressed in the further development of 3,6,9-acridines. The
choice of initial starting conformation for the ligand G-quartet
surface was based on structure-activity relationships (SARs)
from previous work19 and relied on three main assumptions
supporting the chosen orientations. First, the three acridine
substituents should attempt to contact the three loops of the
quadruplex, as it has been shown that modifications at all three
positions can cause significant changes in binding affinity and
potency.19,20 Second, a charge on the central nitrogen of the
acridine appeared favorable and may assert this influence
through positioning directly above the potassium axis that runs
through the center of the quadruplex structure, perpendicular
to the G-quartets. In this way, it may act as a “pseudo-potassium
ion” in the stabilization of the uppermost G-tetrad.59 Third, only
one exposed quartet of the quadruplex was considered as a
binding site on the basis of the crystal structure (the “3′-face”).
This face is exposed to water, whereas the second (5′) face lies
directly adjacent to the same face from a neighboring quadruplex
in the crystal packing arrangement and had been suggested to
be the more energetically favorable site for binding.19 It is
difficult to judge the extent to which the choice of starting
conformation played a role in the MD simulations performed.
Genetic algorithm (GA) conformational searches performed
during the setup of the ligand-DNA complexes indicated that
there was not a significant energetic difference between the
lowest energy conformers of the acridines. Changes in ligand
position are therefore likely to be dependent on overcoming
energetic barriers between local minima of similar energy on
the quadruplex surface, rather than energetic differences between
ligand conformations. Movement of the 3- and 6-side chains
can be seen to be dependent mainly on the position of the
charged terminal pyrrolidine group and the interaction this
moiety makes with the spatially relevant DNA backbone. The
side-chain length of compound 1a is less than optimal, as neither
side chain has the full potential to maximize its electrostatic
and van der Waals (VDW) energy interactions with the
corresponding loops. Compound 1b is more favorable, the extra
length in the side chains allowing more systematic sampling of
the loop cavity, and the added flexibility of the side chains
decreases the impact of variations in the loop structure on
binding site and orientation during the MD simulations.
However, with compounds 1c and 1d, the modeled structures
at first sight appear to contrast with the data obtained from the
biophysical experiment. In the model, the increased side-chain
lengths of these two compounds would be expected to lead to
a corresponding increase in quadruplex binding, yet this is not
the conclusion from visual inspection of the MD trajectories.
In a manner similar to the FRET/TRAP correlation, the
relationship between the SPR and TRAP data should be one
whereby increased ligand binding affinity for quadruplex DNA
corresponds to greater potency in the TRAP assay. However,
the TRAP assay contains a PCR step, the efficiency of which
is related to the duplex affinity of the compound tested, and it
was therefore decided to compare telomerase inhibition data
with the quadruplex-to-duplex binding ratio (KG4DNA/KdsDNA),
rather than directly with the quadruplex affinity data alone. In
this way, the impact of both the KG4DNA and KdsDNA properties
of the acridines on the inhibition assay are taken into account.
Plots of telEC50 values against the binding ratio confirmed this
inversely proportional relationship, and the use of the ratio gave
a better correlation than the use of KG4DNA alone (Figure 5e,f).
This is especially important in the comparison between com-
pounds 1a and 1b. Although the former shows a 10-fold greater
affinity for quadruplex DNA compared with the latter, the
diminished duplex affinity of 1b leads to both compounds
having comparable KG4DNA/KdsDNA ratios, which is reflected in
their similar telEC50 values. This was further underlined by the
results of the TAQ inhibition assay, which showed <25% PCR
inhibition even at 2 µM for all four trisubstituted acridines
examined (1a-d, data not shown).
The use of a two-binding-site model in the SPR experiments
and thus the identification of a weaker binding site to the
quadruplex DNA structure may be reflected in the corresponding
two-step FRET melting curves noted above. It has been
suggested that the two faces of the parallel quadruplex are not
energetically equivalent in terms of binding to molecules such
as acridines with extended side chains,19 and it is conceivable
that one face presents the major binding site for the compounds
examined in this study while the other represents the weaker
site. This highlights the varying extents to which the acridine
chromophore and the side chains are responsible for the
observed quadruplex affinity and ultimately potency against
telomerase in the TRAP assay. The final measured affinity is a
combination of these two features, with the acridine providing
a strong π-π stacking interaction with the G-quartet while the
side chains probe outlying DNA structure. This is confirmed
when trisubstituted acridines are compared with disubstituted
analogues, and the resulting improvement when going from two
to three side chains is due not simply to the decreased duplex
binding affinity of the 3,6,9-acridines but also to the increased
ligand-DNA interactions that can be attributed to the third
moiety. Comparisons between large sets of acridine com-
pounds19,39 show that correlations are evident among structurally
related subseries of compounds, but not among diverse com-
pounds in general. This suggests that large variations in side-
Although qualitative observations on ligand-quadruplex
interactions gave important insights into the behavior of the
complex, attempts at quantitative correlations between the
biophysical assay data and computationally derived ligand
binding free energies were unsuccessful. This reflects the
difficulties inherent in modeling DNA-ligand interactions using
explicit solvent systems. Factors influencing this include the
treatment of electrostatics and the relatively short time scale (1
ns) used for the MD simulations. Conformational sampling of
ligand positions on the quadruplex surface is compromised by
the short time scale nature of MD models, and the lack of an
experimental DNA-ligand structure for the human intramolecular
quadruplex complicates this issue. Additionally, it is likely that

G-Quadruplex Stabilization and Telomerase Inhibition
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2 591
the model neglects to give a complete representation of the
balance between solvent exposure and DNA binding of the side
chains. Increased side-chain flexibility may therefore lead not
so much to tighter quadruplex binding but to increased solvent
exposure, which is consistent with the experimental data.
SRB assay. Normal cell proliferation generally continued for a
1-week period before a major response was observed but then
rapidly decreased following compound exposure, with a con-
comitant increase in observed senescence. The shape of the
responses in terms of cell growth therefore suggests that the
early change in growth rate observed was the initial cellular
response to this type of damage: slowing population growth
and increasing the number of cells entering senescence.
Furthermore, the observed results are likely to be directly related
to compound structure, as similar data have been reported for
compound 1a.24,34 The shriveled morphology of the cells treated
with 1b may also be indicative of apoptotic cell death, and
although the phenotype could not be accurately described in
our hands, a shrunken cytoplasm and fragmented nuclei have
been described as hallmarks of this process.50 This reflects
results such as those reported for the pentacyclic acridinium
compound RHPS4, where onset of senescence has been
described at higher concentration levels and apoptosis at lower
ones,25 or TMPyP4, which showed the same initial lack of effect
but then elicited an apoptotic response in the following 2
weeks.64 Further experiments are needed to investigate the
possible induction of apoptosis under these experimental condi-
tions.
The SAR of the 9-substituent suggests that it is required not
just for its role in decreasing duplex intercalation but also as
an important third moiety for interaction, for example, with the
appropriate quadruplex loop. However, it is difficult to deter-
mine the extent to which quadruplex binding is dependent on
electronic effects exerted on the acridine chromophore by the
side chains. Changes in side-chain properties may thus have
an impact on the acridine that is difficult to predict or model
simply in terms of increases or decreases in binding due to VDW
or electrostatic ligand-DNA interactions. This may be espe-
cially the case with the 9-anilinoacridines, in which the presence
of the aniline group at the 9-position has been suggested to
play an important electronic role in the protonation state (and
hence physicochemical properties) of the acridine chromo-
phore.60-62
In short, the structure-activity relationships of all three
acridine substituents, taken together, are important for the overall
effectiveness of a particular compound, and the modeling is
consistent with the conclusion that it is difficult to break down
the molecule and attempt to make individual modifications
without altering the DNA-ligand contacts at other parts of the
molecule. Targeting all three loops simultaneously is crucial,
thereby increasing DNA-ligand contacts and facilitating stron-
ger binding as well as selectivity. The modeling also highlights
the key role of the quartet-acridine interaction, as was
previously observed in the crystal structure of a disubstituted
acridine with a dimeric quadruplex having diagonal loops.63
Both the guanines and the chromophore of the ligands remained
very stable over the course of the experiment, regardless of the
side chains attached, although these are important for improving
the selectivity and strength of binding, and variations in their
structure are significant, although further work is required to
accurately model the impact of side-chain modifications.
In terms of senescence, the profile of increased senescence
parallels the decrease in the number of population doublings
over the first 1-2 weeks, perhaps indicating the accelerated
senescence of the fraction of cells most affected by the
compound. Moreover, the stabilization of the number of cells
staining for senescence, as well as the population growth rate,
after 3-4 weeks shows that there is a subpopulation that is
particularly sensitive to exposure to the compound. This may
be due to heterogeneity in telomere length (or less likely in the
3′-overhang), implying a subpopulation of cells with telomeres
of shorter length than the population average. This is in accord
with observations of telomere length using Southern blots, which
show populations of cells with some having short telomeres.34
Shorter telomeres would subsequently accelerate the onset of
senescence,10 which may be accelerated by the presence of
the compounds, in agreement with the results of theoretical
studies on the effects of telomerase inhibitors on population
growth.65,66
Cellular Effects. The cellular changes expected following
exposure to telomere-targeting compounds are varied and
complex, reflecting the interactions of telomeric proteins, DNA,
and RNA involved in telomere maintenance. Targeting of
telomeric DNA in particular is further complicated by issues
of selectivity and specificity. However, the overall correlations
between the quadruplex-affinity (FRET, SPR) and biochemical
assay results (TRAP) suggest that these structure-activity
relationships are derived primarily from the strength of the
ligand-quadruplex DNA interaction, rather than more indirect
effects that might modulate activity in a cell-based assay, such
as interference with protein-quadruplex recognition or other
macromolecular interactions. Above all, the efficacy of such
compounds in the in vivo setting depends on their selectivity
for quadruplex over duplex DNA, as well as their underlying
drugability.
The observed halt to cell growth may be a consequence of
two separate but linked events, with senescence from telomere
length reduction not being solely responsible for the decrease
in population doublings It has been demonstrated that an
alteration of telomere state, rather than simply telomere length,
may be at least partially responsible for entry into replicative
senescence,67,68 and this altered state may rely heavily on the
condition of the 3′-overhang, as it can be shown that much of
this overhang is lost in senescent human cells.6,7 In this
hypothesis, a change in telomere state would signal initial entry
into senescence while shortening of the overall telomere length
results in diminished chromosome protection, triggering crisis.11
This is of particular significance with respect to the biological
data for compound 1b, which showed significantly less short-
term cytotoxicity (as measured in the SRB assay) in the human
fibroblast cell line IMR90 than in the cancer cell line MCF7.
This supports the postulate that telomeres and their associated
capping mechanism are differently regulated in normal and
cancer cells in a manner that is not simply dependent on the
presence or absence of telomerase and therefore they represent
a promising target for small-molecule intervention in their own
right. There is increasing interest in nontelomeric quadruplexes
as possible targets for selective drug action, and the promoter
The specific inhibition of telomerase would be expected to
result in an extended time lag associated with the gradual
shortening of telomeres prior to the onset of replicative
senescence. The MCF7 cell line has telomeres of average length
5-6 kb, and direct telomerase inhibition would thus be expected
to take considerably longer (a loss of 4 kb at a rate of 100 bp
per replication and 0.7 PD per day would occur after ap-
proximately 57 days). Compound 1b, however, produces a
pronounced antiproliferative effect within 1-2 weeks, at
concentrations well below the IC50 value as determined in the

592 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2
Moore et al.
(ii) Protonation of the 9-position extended side chains in
compounds 1e-j results in elevated ∆Tm values and generally
higher potency against telomerase.
(iii) Extending the 9-position substituent length does not have
a major effect on either binding or potency. It is possible to
counterbalance the adverse effects of extended 3,6- substituents
with an extended substituent at the 9-position.
(iv) Selectivity for quadruplex over duplex DNA is correlated
with both affinity and telomerase inhibition.
(v) Potency against telomerase and a high level of quadruplex
affinity are necessary features though not sufficient by them-
selves for cellular activity. Compounds need to have appropriate
cellular uptake and metabolic stability.
Of the compounds synthesized and evaluated in this study,
only two, 1a and 1b, can be said to meet these criteria. An
analogue of these has recently been chosen as a candidate for
clinical studies in human cancer.
Figure 6. Representative picture of MCF7 cells following long-term
(4-week) incubation with compound 1b (1 µM) after staining for
senescence-associated â-galactosidase (SA-â-Gal) activity (blue, ar-
rows). Approximate magnification factor ×200.
Materials and Methods
regions of oncogenes such as c-myc have received special
attention.69 Such sites are within double-stranded DNA, con-
trasting with the uniquely single-stranded 3′-overhang telomeric
DNA, and presumably are most accessible during transcription
or replication. It has been calculated68 that there are up to
350 000 potential quadruplex sites in the human genome
(although it is likely that many cannot form stable quadru-
plexes69). Thus, the issue of selectivity for a particular quadru-
plex may be important. There is little data available to date,
but one study has shown70 that compound 1a has a 5-fold
selectivity for the human telomeric quadruplex compared to the
c-myc one. This selectivity may be significantly higher in cells
when the accessibility of these targets is taken into account. In
a therapeutic setting, even a small effect that down-regulates
the expression of oncogenes is beneficial.
Biological results for compound 1b, which exhibits binding
and telomerase inhibitory activity equivalent to those of 1a,
while suggesting an improved SAR, underline the short time
scale of effects in the MCF7 breast cancer cell line and
demonstrate a link between cell growth inhibition and the onset
of senescence (Figure 6). Furthermore, these effects were
observed at concentrations considerably below the reported
short-term cytotoxicity (IC50) value of the compound, and these
studies showed that the compound is significantly more potent
in this cancer cell line than in a normal somatic cell fibroblast
line (IMR90). The inability of compound 1f to produce cell
growth arrest and senescence may be due to several factors;
metabolism at the 9-substituent amide bond or reduced cellular
uptake as a result of the additional charge on the pyrrolidine
group at the 9-position are both possibilities. The calculated
log P value for compounds 1a and 1b are respectively 1.4 and
1.9, whereas that for 1f is -0.36. We conclude that potent
telomerase inhibitory activity and quadruplex affinity are
necessary though by themselves insufficient features for cellular
activity. The extra charge at the 9-position in compound 1f may
well be one charge too many.
Fluorescence resonance energy transfer (FRET), PCR inhibition
(TAQ), and telomeric repeat amplification protocol (TRAP) assays
were carried out following previously published procedures,39,52
while surface plasmon resonance (SPR) experiments were also
conducted according to reported methods.19,20
FRET Assay. All oligonucleotides and their fluorescent conju-
gates (Eurogentec, Southampton, UK) were initially dissolved as a
stock 50 µM solution in purified water; further dilutions were carried
out in the relevant buffer. The labeled oligonucleotide F21T [5′-
FAM-d(GGG[TTAGGG]3)-TAMRA-3′; donor fluorophore FAM
is 6-carboxyfluorescein; acceptor fluorophore TAMRA is 6-car-
boxytetramethylrhodamine] used as the FRET probe was diluted
from stock solution to the correct concentration (400 nM) in a 50
mM potassium cacodylate buffer (pH 7.4) and then annealed by
heating to 85 °C for 10 min, followed by cooling to room
temperature in a heating block. Compounds were stored at -80
°C as 10 mM stock solutions in DMSO; final solutions were
prepared using 10 mM HCl in the initial 1:10 dilution, after which
50 mM potassium cacodylate buffer (pH 7.4) was used in all
subsequent steps. The 96-well plates (MJ Research, Waltham, MA)
were prepared by aliquoting 50 µL of the annealed DNA into each
well, followed by 50 µL of the compound solutions. Measurements
were made in triplicate on a DNA Engine Opticon (MJ Research)
with excitation at 450-495 nm and detection at 515-545 nm.
Fluorescence readings were taken at intervals of 0.5 °C over the
range 30-100 °C, with a constant temperature being maintained
for 30 s prior to each reading to ensure a stable value. Final analysis
of the data was carried out using the program Origin 7.0 (OriginLab
Corp., Northampton, MA).
TAQ Assay. The procedure used was based on the internal
control sequences that have been reported in the literature.73
A
master mix was prepared containing the TS forward primer (0.15
µg; 5′-AAT CCG TCG AGC AGA GTT-3′), NT reverse primer
(0.5 µM; 5′-ATC GCT TCT CGG CCT TTT-3′), TSNT template
(200 pM; 5′-AAT CCG TCG AGC AGA GTT AAA AGG CCG
AGA AGC GAT-3′), MgCl2 (1 mM), and equal amounts of dNTPs
(20 µM), as well as Taq polymerase (1.25 U; RedHot, ABgene,
Surrey, UK), in addition to the Taq reaction buffer (ABgene). All
primers were obtained from Invitrogen (Paisley, UK) and had been
HPLC-purified by the supplier. Compound solutions (prepared as
before) were then pipetted into the reaction tubes, and the
appropriate amount of master mix was added with mixing.
Amplification was carried out using a two-step protocol with an
initial denaturing step at 92 °C for 2 min, followed by 33 PCR
cycles (92 °C for 30 s, 52 °C for 30 s, 72 °C for 45 s), and the 36
bp product was visualized on a 10% v/v polyacrylamide gel using
SYBR Green I staining (Sigma).
The biophysical and biochemical data overall suggest a
number of structure-activity correlations for these trisubstituted
acridines:
(i) With the 9-position substituents not having a cationic
charge [i.e., compounds 1a-d, with R ) N(CH3)2], both
quadruplex binding and telomerase inhibition are optimal with
the shortest 3- and 6-side chains, i.e., with n ) 2; extending
the side chains by even one carbon atom beyond n ) 3 has a
highly deleterious effect on quadruplex binding and telomerase
potency.
TRAP Assay. Telomerase activity in the presence of the
compounds was assessed using a modified version of standard
published TRAP protocols.39,73-75 Cell extract from exponentially

G-Quadruplex Stabilization and Telomerase Inhibition
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2 593
growing A2780 human ovarian carcinoma cells was used as the
enzyme source and total protein upon extraction was quantitated
using the standard Bradford assay method.76 Briefly, the TRAP
assay was carried out in two steps with an initial primer-elongation
step and subsequent PCR amplification of the telomerase products
to enable detection. In part 1, a master reaction mix was prepared
containing the TS forward primer (0.1 µg; 5′-AAT CCG TCG AGC
AGA GTT-3′), TRAP buffer [20 mM Tris-HCl (pH 8.3), 68 mM
KCl, 1.5 mM MgCl2, 1 mM EGTA, 0.05% v/v Tween-20], BSA
(0.05 µg), and dNTPs (125 µM each). Compounds were prepared
to 10 times final concentration (or vehicle controls) as described
above, added to the PCR reaction tubes, and maintained at 0 °C.
Simultaneously, the required amount of protein was thawed to 0
°C on ice and diluted in TRAP lysis buffer (pH 7.4, 0.5% CHAPS,
10 mM Tris-HCl, 1 mM MgCl2, 1 mM EGTA, 5 mM â-mercap-
toethanol, 10% glycerol) to the required concentration (500 µg/
µL). The appropriate amount (2 µL) was then combined with the
master mix on ice and aliquoted into the PCR tubes to give a final
volume of 40 µL. The initial telomere elongation step was carried
out for 10 min at 30 °C, followed by a protein heat inactivation
step (4 min at 92 °C) and final maintenance of the mixture at 4 °C.
Following heat inactivation of telomerase, 10 µL of a PCR reaction
mix containing ACX primer (1 µM; 5′-GCG CGG [CTTACC]3
CTA ACC-3′) and 2 U Taq polymerase (RedHot, ABgene, Surrey,
UK) in TRAP reaction buffer was added to each tube to start the
PCR protocol for part 2, with thermal cycling being carried out in
three parts following an initial 2 min denaturing period at 92 °C
(33 cycles of 92 °C for 30 s, 55 °C for 30 s, 72 °C for 45 s). The
PCR-amplified reaction products were run out on a 10% v/v PAGE
gel and visualized using SYBR green I (Sigma). telEC50 values were
calculated by quantitating the TRAP product using a gel scanner
and GeneTools software (Syngene, Cambridge, UK). Measurements
were made with respect to a negative control run using the
equivalent TRAP-PCR conditions but with the protein extract
either omitted or heat inactivated by incubation at 92 °C (20 min).
SPR Experiments. The Biosensor-SPR experiments were per-
formed on BIAcore 2000 and 3000 optical biosensor systems using
four-channel streptavidin-coated sensor chips (Biacore, Inc.) as
previously described.19,20 Briefly, streptavidin-coated sensor chips
were preconditioned with three to five consecutive 1-mL injections
of 1 M NaCl in 50 mM NaOH and washing with Hepes buffer at
pH 7.4 [0.01 M Hepes, 0.20 M KCl, 3 mM EDTA, 0.005% (v/v)
surfactant P20]. Conditioning was continued until the change in
SPR signal was less than 1 RU/min. Immobilization of the DNA
on the surface was carried out by noncovalent capture of biotiny-
lated DNA sequences representative either of the human telomeric
quadruplex (5′-biotin-d[AG3(TTAG3)3]-3′) or a hairpin duplex DNA
sequence (5′-biotin-CGAATTCG-CTCT-CGAATTCG-3′). Manual
injection of 25 nM DNA in Hepes buffer (pH 7.4) at a flow rate of
2 µl/min was used in order to achieve long contact times with the
surface and to better control the amount of DNA bound to the
surface. Folding of the quadruplex DNA with respect to time under
similar conditions to those used on the sensor chips was verified
by a series of melting/cooling experiments using both CD and UV
methods. All the binding study procedures were automated using
cycles of sample injection and surface regeneration. To minimize
possible unwanted mass transport effects, all SPR kinetics experi-
ments were conducted at flow rates of 50-100 µL/min with low
surface densities of immobilized DNA. For the steady-state analysis,
flow rates were typically 10-30 µL/min. Global fitting of the
association and dissociation curves was performed using the
supplier’s software (BIA Evaluation Software, Biacore, Inc.) and
was done in a concentration range where the compounds bind
significantly only to the strongest binding site.
charges and atom types applied using the MMFF94 method.77,78
Stepwise minimizations using the MMFF94s force field79 were
subsequently carried out to a convergence of 0.02 kcal mol-1/Å
over 5000 steps using the Powell method80 with a cutoff of 11 Å.
This was followed by a conformational search on the ligand side
chains (all rotatable bonds) using a genetic algorithm (GA) as
implemented in SYBYL.81-83 The 100 lowest energy conformers
were generated using a search performed over 2 × 104 generations
with a population of 1000 and a selective pressure of 1.5,82 again
with a cutoff of 11 Å and a dielectric constant of 80.
Following output from the conformational searches, quadruplex
and ligand were brought into close proximity with the acridine
chromophore parallel to the G-quartet (approximately distance 3.4
Å) and the acridine nitrogen placed above the central potassium
channel. Ligand conformers with 3- or 6-side chains clashing with
the DNA were then excluded, yielding generally between three and
five structures for further work. Conformers chosen in this way
were merged with the quadruplex and the aggregate (DNA plus
ligand) and minimized in the same manner as has been described,
with the DNA remaining fixed as a static set (convergence of 0.02
kcal mol-1/Å, 5000 steps Powell minimization, cutoff 11 Å,
dielectric constant of 4). Once merged, the structures were
superimposed using the quadruplex as a template and the ligands
separated from the complex to yield a molecular database containing
the quadruplex and all the ligands docked in this manner.
Subsequent preparation of the DNA-ligand complexes (with
compounds 1a-d), further minimizations, and molecular dynamics
(MD) calculations were performed using the XLEAP, ANTE-
CHAMBER, and SANDER modules of the AMBER7 package.84,
85
All calculations were carried out in AMBER 7 using the Parm99
version of the Cornell et al. force field,86 the TIP3P potential for
waters,87 the parm99.dat parameter set for the nucleic acids,88 and
GAFF atom types85,89 for the ligand. Ligand charges were derived
using the MMFF94 method77 and defined in the ligand preparatory
files as part of the new residue input. Periodic boundary conditions
were applied, with the particle-mesh Ewald (PME) method90 used
to treat long-range electrostatic interactions. The solute was first
solvated in a water-box, the dimensions of which extended to a
distance at least 10 Å from any solute atom (total VDW box size
approximately 61 × 54 × 50 Å3). Potassium counterions were
subsequently added using XLEAP to attain overall system neutrality,
with standard potassium parameters as applied in the Cornell et al.
force field (VDW radius 2.658 Å).
Minimizations and MD runs were performed using SANDER,
with the SHAKE algorithm enabled for hydrogen atoms in dynamics
runs and a 1 fs time step with an 11 Å nonbonded cutoff. Overall
equilibration of the system was carried out in successive steps, with
initial equilibration of the system at constant volume followed by
further rounds at constant pressure once the desired system
temperature (300 K) had been reached. Following two primary
rounds of solvent minimization, solute restraints were gradually
reduced to 0.2 kcal mol-1/Å over six 10 ps MD steps (100, 50, 25,
10, 5, 0.2), followed by 100 ps of unrestrained MD. The final
geometry was then the starting point for the full 1 ns production
run, with energy information and averages printed out every 250
steps, coordinates captured at 0.5 ps intervals, and the nonbonded
list updated every 10 steps. Trajectories were examined visually
using the VMD software package.91 This allowed mapping of any
large conformational changes to distinct time periods in the
trajectory and the comparison with energy and rmsd values.
Sulforhodamine B (SRB) Growth Inhibition Aassay and
Long-Term Cell Culture Experiments. The SRB assay was
carried out in 96-well microtiter plates (Nunclon Surface, Nunc
A/S, Denmark) according to published procedures.30,47,48 Short-term
cytotoxicity measurements were carried out in the human breast
adenocarcinoma cell line MCF7 and in the human fibroblast cell
line IMR90, both of which were obtained from the European
Collection of Cell Cultures (Salisbury, UK).
Molecular Modeling. The crystal structure of the parallel 22mer
telomeric G-quadruplex in potassium buffer (PDB ID 1KF118) was
used as the starting point for the computational work. Preparation
and manipulation of the structure and initial building of the ligand-
DNA complex prior to the molecular dynamics (MD) simulations
was carried out using programs within the SYBYL 6.9 suite (Tripos,
Inc., St. Louis, MO). Ligands were constructed in SYBYL, with
Long-term experiments were carried out using the MCF7 cell
line. Counting and media changes were carried out according to

594 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2
Moore et al.
standard procedures as has been reported,24,30 with cells maintained
in Dulbecco’s modified Eagle’s medium (DMEM; Invitrogen,
Groningen, NL) supplemented with 10% v/v fetal calf serum (Life
Technologies, Paisley, UK), 0.5 µg/mL hydrocortisone (Acros
Chemicals, Loughborough, UK), 2 mM L-glutamine (Invitrogen),
and a 1× solution of nonessential amino acids (Invitrogen) under
a humidified 5% CO2, 95% air atmosphere. Cell culture followed
a regular weekly pattern based on the rate of growth and the cells
reaching confluence, with counting, reseeding, and the first treat-
ment on day 1 followed by reexposure at the same compound
concentration on day 4. This protocol was maintained for 4-8
weeks, with analysis and staining experiments being carried out
following the change of medium and transfer to a fresh flask on
day 1.
Staining for Senescence-Associated â-Galactosidase (SA-â-
Gal) Activity. Staining for SA-â-Gal activity was carried out
according to the instructions of the supplier (Cell Signaling
Technology, Inc., Beverly, MA). In brief, cells from long-term
exposure studies were retrieved at the end of each week and seeded
in 35 mm 6-well plates (Nunc A/S) at a density of 1 × 105 cells in
2 mL media and incubated overnight under standard conditions
together with the appropriate concentration of the compound under
investigation. After an approximately 24-h incubation period, the
growth medium was removed, and the cells were washed, fixed,
and stained using the supplied staining solution [400 mM citric
acid/sodium phosphate (pH 6.0), 1.5 M NaCl, 20 mM MgCl2, 5
mM potassium ferrocyanide, 5 mM potassium ferricyanide, 1 mg
of X-gal (5-bromo-4-chloro-3-indolyl-â-D-galactopyranoside)], fol-
lowed by incubation overnight at 37 °C (5% CO2). Cells were
examined by light microscope (mag. 200-800×) and counted the
next day for the characteristic senescence-associated development
of blue coloration.
Chemistry. Melting points (mp) were recorded on a Stuart
Scientific SMP1 melting point apparatus and are uncorrected. IR
spectra of solids were recorded using a Perkin-Elmer SPECTRUM
1000 FT-IR spectrometer. NMR spectra were recorded at 400 MHz
(1H) and 100 MHz (13C) on a Bruker spectrometer in either CDCl3
(Aldrich) or DMSO-d6 (GOSS) or CD3OD (GOSS) solutions using
TMS as an internal standard or in the case of 13C spectra the residual
solvent peak of CDCl3 and DMSO-d6. Mass spectrometry services
were provided by The School of Pharmacy (nominal and HRMS
ESI, MALDI) and the EPSRC National Mass Spectrometry Service
Centre (Chemistry Department, University of Wales, Swansea,
Singleton Park, Swansea SA2 8PP, HRMS ESI). TLC analysis was
carried out on silica gel (Merck 60F-254) with visualization at 254
and 366 nm. Treatment of an organic solution in the usual manner
refers to stepwise drying with magnesium sulfate, filtration, and
then evaporation of the filtrate in vacuo. Preparative flash chro-
matography was carried out with BDH silica gel (BDH 153325P).
Reagents and chemicals unless indicated were purchased from
Sigma-Aldrich. Solvents were purchased from BDH. N,N-Dim-
ethylaminoaniline was generated by neutralization of the com-
mercially available dihydrochloride salt. 6-Chlorohexanoyl chlo-
ride;88 acridones 2,41 3,19 and 7;19 9-chloroacridines 1119 and 16;42
and acridines 1a19 and 2238 were prepared as reported. Details for
the preparation of anilines 15a-c and acridine 23 by treatment of
9-chloroacridine with aniline 15b are presented in the Supporting
Information, together with IR data on all compounds reported here.
Analytical rpHPLC was performed with a Waters 600S Controller
HPLC system connected to a Thermosep Products Spectra Series
UV detector. HPLC method A1 had the following parameters:
column, Hamilton C18; flow, 0.6 mL/min; gradient, 0-5 min, 20%-
80% methanol in water containing 0.1% trifluoroacetic acid);
detection, 290 nm. HPLC method B1 had the following param-
eters: column Phemonex Prodigy C18; flow, 1 mL/min; gradient,
0-27 min, 0%-46% acetonitrile in 30% acetonitrile in water
containing 46 mM sodium octyl sulfonate with pH adjusted to 2.8
with phosphoric acid; detection 268 nm. Acridines 1b-j were found
to be >95% pure by applying HPLC method A1. The purity of 1b
and 1f, which were submitted for cell biology studies, was also
assessed with HPLC method B1. Details are presented in the
Supporting Information. Analytical data for compound 1a have been
previously presented.19
3,6-Bis(chloroalkylamido)acridones (4-6). A suspension of 2
and acid chloride was heated at the specified temperature for the
indicated time. The reaction suspension was allowed to cool to room
temperature and placed in an ice bath, and ice-cold anhydrous
diethyl ether was added. The solid produced was isolated by
filtration and stirred in a 1:1 ice-cold mixture of aqueous sodium
hydrogen carbonate and diethyl ether. This suspension was filtered,
and the solid obtained was washed with ice-cold water and diethyl
ether and dried in vacuo over phosphorus pentoxide to afford pure
title product.
3,6-Bis[(pyrrolidin-1-yl)alkylamido]acridones (8-10). A solu-
tion of 4, 5, or 6 (stated amount) in pyrrolidine (stated volume)
was heated under reflux for the specified time. The reaction solution
was allowed to cool to room temperature and then poured into a
stirred saturated aqueous sodium hydrogen carbonate solution
(stated volume). Stirring was continued until a solid formed which
was collected by suction filtration, washed with ice-cold diethyl
ether and water, and then dried in vacuo over phosphorus pentoxide
to afford title product.
3,6-Bis[(pyrrolidin-1-yl)alkylamido]-9-chloroacridines (12-
14). A suspension of 8, 9, or 10 and phosphorus pentachloride and
phosphorus oxychloride was heated under reflux with concomitant
dissolution for the stated time. The reaction solution was allowed
to cool to room temperature and then placed in an ice bath. Ice-
cold anhydrous diethyl ether (stated volume) was then added to
the cooled reaction solution. The resulting solid was isolated by
suction filtration, washed thoroughly with anhydrous ice-cold
diethyl ether, and then added portionwise to a stirred mixture of
ice-aqueous ammonia solution (0.880)-chloroform in an ice-bath.
Further ammonia solution was added to ensure the mixture remained
alkaline. The chloroform layer was separated and the aqueous layer
extracted with chloroform. The combined organic extracts were
washed with brine and treated in the usual manner to give typically
a dark brown residue that was triturated with ice-cold diethyl ether
to afford a crude product which was submitted directly to the
synthesis of trisubstituted acridines.
3,6-Bis(chloroalkylamido)-9-anilinoacridines (19-21). A sus-
pension of 18 and the appropriate acid chloride, triethylamine, and
toluene, if required, was microwaved at 150 °C for 5 min in fixed
hold time mode. The resulting reaction mixtures were filtered to
isolate the title compounds as solid products, which were then
triturated with ether, dried, and immediately submitted to the next
synthetic step.
3,6-Bis[(pyrrolidin-1-yl)alkylamido]-9-anilinoacridines (1).
Method A (1a-j). Either N,N-dimethylaminoaniline or 15a-c was
added portionwise to a solution of the appropriate 9-chloroacridine
11-14 in MeOH (stated volume) and the resulting solution heated
under reflux conditions for the stated time. The reaction solution
was evaporated and the resulting red brown solid partitioned
between chloroform and 20% aqueous ammonium hydroxide
solution. The organic layer was separated and washed with brine
and treated in the usual manner to give crude product as a red brown
solid which was purified by flash chromatography eluting with the
stated solvent mixture of dichloromethane and methanol with 10%
v/v triethylamine. Fractions containing product were evaporated
and partitioned again between chloroform and 20% dilute aqueous
ammonium hydroxide solution. The organic layer was separated
and the aqueous layer further extracted with chloroform. The
combined organic extracts were washed with brine and then treated
in the usual manner to give red brown solids, which were dried in
a pistol in vacuo over phosphorus pentoxide to afford the pure title
product.
Method B (1h-j). A solution of the 3,6-bis(chloroalkylamido)-
9-anilinoacridine (19-21) in anhydrous pyrrolidine (1 mL) was
stirred overnight at room temperature. LC/MS analysis at this point
indicated reaction completion. The reaction solution was evaporated
and the resulting crude product purified by flash chromatography
eluting with 3:6:1 methanol/dichloromethane/triethylamine as de-
scribed above for method A to give to pure title products.

G-Quadruplex Stabilization and Telomerase Inhibition
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2 595
3,6-Bis(4-chlorobutanamido)acridone (4). A brown suspen-
sion of 2 (5.019 g, 22 mmol) and 4-chlorobutyryl chloride (50 mL,
63 g, 450 mmol, 20 equiv) was heated at 80 °C overnight. The
now yellow suspension was treated as described in the general
methods to afford 4 (9.301 g, 96%) as a brown solid: mp 319 °C
(dec); 1H NMR (DMSO-d6) δ 2.076 (m, 4H, J ) 6.7 Hz), 2.571(t,
4H, J ) 7.3 Hz), 3.734 (t, 4H, J ) 6.5 Hz), 7.201 (dd, 2H, J )
1.9, 8.8 Hz), 8.089 (d, 2H, J ) 8.8 Hz), 8.127 (d, 2H, J ) 1.7 Hz),
10.370 (bs, 2H), 11.646 (bs, 1H); 13C NMR (DMSO-d6) δ 27.7,
33.5, 44.9, 48.5, 104.7, 113.2, 116.4, 126.8, 142.0, 143.1, 170.9,
175.0; HRMS (MALDI) m/z 434.1033 (MH+, C21H22N3O3 requires
434.1033).
3,6-Bis(5-chloropentanamido)acridone (5). A brown suspen-
sion of 2 (1.002 g, 4 mmol) and 5-chlorovaleroyl chloride (6 mL,
7.2 g, 46 mmol, 12 equiv) was heated at 70 °C for 4 h. Treatment
of the reaction mixture as described in the general methods gave 5
as a light brown solid (1.485 g, 72%): mp 337 °C (dec); 1H NMR
(DMSO-d6) δ 1.768 (m, 8H), 2.427 (t, 4H, J ) 6.8 Hz), 3.681 (t,
4H, J ) 6.1 Hz), 7.191 (dd, 2H, J ) 1.8, 8.8 Hz), 8.088 (d, 2H, J
) 8.6 Hz), 8.129 (d, 2H, J ) 1.4 Hz), 10.274 (bs, 2H), 11.614 (bs,
1H); 13C NMR (DMSO-d6) δ 22.3, 31.5, 35.6, 45.0, 104.6, 113.2,
116.3, 126.7, 142.0, 143.1, 171.6, 174.9; HRMS (MALDI) m/z
462.1346 (MH+, C23H26Cl2N3O3 requires 462.1346).
3,6-Bis(6-chlorohexanamido)acridone (6). A brown suspension
of 2 (2.223 g, 0.01 mol) and 6-chlorohexanoyl chloride (16.39 g,
0.1 mol, 10 equiv) was heated at 70 °C for 5 h. Treatment of the
reaction mixture as described in the general methods gave 6 as a
brown solid (2.454 g, 51%): mp >350 °C (dec); 1H NMR (DMSO-
d6) δ 1.448 (m, 4H), 1.654 (m, 4H), 1.761 (m, 4H), 2.397 (t, 4H,
J ) 7.3 Hz), 3.651 (t, 4H, J ) 6.5 Hz), 7.191 (d, 2H, J ) 8.8 Hz),
8.083 (d, 2H, J ) 8.8 Hz), 8.134 (s, 2H), 10.256 (bs, 2H), 11.615
(bs, 1H); 13C NMR (DMSO-d6) δ 24.2, 25.9, 31.7, 36.3, 45.2, 104.6,
113.2, 116.3, 126.7, 142.0, 143.1, 171.8, 174.9; HRMS (MALDI)
m/z 490.1659 (MH+, C25H30N3O3 requires 490.1625).
25.0, 26.7, 28.2, 36.5, 53.5, 55.6, 104.6, 113.2, 116.3, 126.7, 142.0,
143.2, 172.0, 174.9; HRMS (+ESI) m/z 560.3595 (MH+, C33H46N5O3
requires 560.3595).
3,6-Bis[4-(pyrrolidin-1-yl)butanamido]-9-chloroacridine (12).
A suspension of 5 (3.029 g, 6 mmol) and phosphorus pentachloride
(1.249 g, 6 mmol) in phosphorus oxychloride (20 mL) was refluxed
for 2.5 h. The resulting solution was treated as described in the
general methods to afford crude 12 as a brown solid (1.261 g, 40%).
3,6-Bis[5-(pyrrolidin-1-yl)pentanamido]-9-chloroacridine (13).
A brown suspension of 9 (983 mg, 1.85 mmol) and phosphorus
pentachloride (0.55 g, 2.6 mmol, 1.43 equiv) in phosphorus
oxychloride (20 mL) was heated under reflux for 3 h. The reaction
solution was treated as described in the general methods to afford
crude 13 as a brown solid (600 mg, 59%).
3,6-Bis[6-(pyrrolidin-1-yl)hexanamido]-9-chloroacridine (14).
A suspension of 10 (624 mg, 1.1 mmol) and phosphorus pentachlo-
ride (368 mg, 1.8 mmol, 1.6 equiv) in phosphorus oxychloride (15
mL) was heated under reflux for 3 h. The reaction solution was
treated as described in the general methods to afford crude 14 as a
brown solid (241 mg, 37%).
3,6-Diazido-9-{4′-[3′′-(pyrrolidin-1-yl)propanamido]anilino}-
acridine (17). Aniline 15b (62 mg, 0.2 mmol) was added to a
solution of the diazido acridine 16 (6 mg, 0.2 mmol) in NMP (3
mL) followed by catalytic HCl (5 drops) and the resulting solution
stirred at room temperature for 1 h, at which point TLC analysis
indicated reaction completion. Ethyl acetate (15 mL) was added
and the red solid produced isolated by filtration and washed with
ethyl acetate and then diethyl ether to yield 14 as a red solid (86
mg, 87%): mp 200-205 °C; 1H NMR (CD3OD) δ 2.028 (m, 2H),
2.140 (m, 2H), 2.932 (t, 2H, J ) 6.7 Hz), 3.132 (m, 2H), 3.537 (t,
2H, J ) 6.5 Hz), 3.684 (m, 2H), 6.574 (d, 2H, J ) 1.9 Hz), 6.608
(dd, 2H, J ) 1.9, 9.3 Hz), 7.111 (d, 2H, J ) 8.7 Hz), 7.605 (d, 2H,
J ) 8.7 Hz), 7.671 (d, 2H, J ) 9.3 Hz); HRMS (+ESI) m/z
493.2220 (MH+, C26H25N10O requires 493.2207).
3,6-Bis[4-(pyrrolidin-1-yl)butanamido]acridone (8). A solution
of 4 (1.735 g, 4 mmol) in pyrrolidine (7 mL, 5.964 g, 84 mmol, 21
equiv) was heated under reflux overnight. The reaction solution
was allowed to come to room temperature and cold saturated
aqueous sodium hydrogen carbonate (30 mL) added with stirring.
Stirring of the resulting oily mixture was continued until the point
a brown solid collected. This solid was taken up into 10%
hydrochloric acid, washed with dichloromethane, and then made
basic with 5 N aqueous sodium hydroxide. The light brown solid
formed was isolated by suction filtration and washed with cold ether
3,6-Diamino-9-{4′-[3′′-(pyrrolidin-1-yl)propanamido]anilino}-
acridine (18). A slurry of 10% Pd/C catalyst (3 mg) in ethyl acetate
(0.5 mL) was added to a solution of 17 (25 mg, 0.0508 mmol) in
methanol (5 mL) in a Parr hydrogenator flask and the reaction
suspension shaken at 20 psi. H2 for 1 h. At this point, LC/MS and
TLC analysis indicated reaction completion. The mixture was
filtered through Celite and the filtrate evaporated to give 18 as a
1
red solid (15 mg, 67%): mp 205-210 °C; H NMR (CD3OD) δ
2.067 (m, 2H), 2.179 (m, 2H), 2.970 (t, 2H, J ) 6.7 Hz), 3.171
(m, 2H), 3.576 (t, 2H, J ) 6.5 Hz), 3.723 (m, 2H), 6.613 (d, 2H,
J ) 1.9 Hz), 6.647 (dd, 2H, J ) 1.9, 9.4 Hz), 7.150 (d, J ) 8.7
Hz), 7.644 (d, 2H, J ) 8.7 Hz), 7.710 (d, J ) 9.3 Hz); 13C NMR
(DMSO-d6) δ 22.7, 32.0, 49.6, 52.9, 94.5, 105.5, 114.3, 119.9,
123.0, 127.2, 128.1, 128.8, 138.0, 142.9, 154.2, 167.6; HRMS
(+ESI) m/z 441.2417 (MH+, C26H29N6O requires 441.2397).
1
and water. Yield 8 (1.564 g, 78%): mp 311 °C (dec); H NMR
(DMSO-d6) δ 1.680-1.648 (m, 8H), 1.782 (m, 4H), 2.407-2.447
(m, 16H), 7.182 (dd, 2H, J ) 1.9, 8.8 Hz), 8.074 (d, 2H, J ) 8.8
Hz), 8.131 (d, 2H, J ) 1.8 Hz), 10.275 (bs, 2H), 11.626 (bs, 1H);
13C NMR (DMSO-d6) δ 23.0, 24.2, 34.6, 53.4, 55.0, 104.6, 113.2,
116.3, 125.7, 142.1, 143.2, 172.0, 175.0; HRMS (ESI+) m/z
504.2969 (MH+, C29H38O3N5 requires 504.2969).
3,6-Bis(4-chlorobutanamido)-9-{4′-[3′′-(pyrrolidin-1-yl)-
propanamido]anilino}acridine (19). A suspension of 18 (30 mg,
0.068 mmol), 3-chloropropionyl chloride (2 mL), and triethylamine
(37 µL, 27 mg, 0.27 mmol, 4 equiv) was microwaved and the
resulting reaction mixture treated as described in the general
methods to afford 19 as a red solid (35 mg, 79% yield): LC/MS
(+ESI) m/z 649.27 (MH+).
3,6-Bis(5-chloropentanamido)-9-{4′-[3′′-(pyrrolidin-1-yl)-
propanamido]anilino}acridine (20). A suspension of 18 (30 mg,
0.07 mmol), 5-chlorobutyryl chloride (2 mL), and triethylamine
(37 µL, 27 mg, 0.27 mmol, 4 equiv) was microwaved and the
resulting reaction mixture treated as described in the general
methods to afford 20 as a red solid (37 mg, 80%): LC/MS (+ESI)
m/z 677.30 (MH+).
3,6-Bis(6-chlorohexanamido)-9-{4′-[3′′-(pyrrolidin-1-yl)-
propanamido]anilino}acridine (21). A suspension of 18 (25 mg,
0.06 mmol), 6-chlorohexanoyl chloride (48 mg, 0.29 mmol, 5
equiv), and triethylamine (32 µL, 23 mg, 0.23 mmol, 4 equiv) was
microwaved and the resulting reaction mixture treated as described
in the general methods to afford 21 as a red solid (24 mg, 59%):
LC/MS (+ESI) m/z 705.25 (MH+).
3,6-Bis[5-(pyrrolidin-1-yl)pentanamido]acridone (9). The green
solution of 5 (466 mg, 1 mmol) in pyrrolidine (5 mL, 4.26 g, 60
mmol, 60 equiv) was heated at reflux overnight. Treatment of
reaction solution as described in the general methods gave 9 as a
1
light brown solid (514 mg, 90%): mp >350 °C (dec); H NMR
(DMSO-d6) δ 1.482 (m, 4H), 1.651 (m, 6H), 1.774 (m, 6H), 2.427
(m, 12H), 3.682 (t, 4H, J ) 5.6 Hz), 7.192 (d, 2H, J ) 8.4 Hz),
8.085 (d, 2H, J ) 8.6 Hz), 8.129 (s, 2H), 10.291 (bs, 2H), 11.642
(bs, 1H); 13C NMR (DMSO-d6) δ 23.0, 23.1, 28.0, 36.4, 53.5, 55.3,
104.6, 113.2, 116.3, 126.7, 142.1, 143.2, 172.0, 175.0; HRMS
(+ESI) m/z 532.3292 (MH+, C31H42N5O3 requires 532.3282).
3,6-Bis[6-(pyrrolidin-1-yl)hexanamido]acridone (10). A solu-
tion of 6 (3.831 g, 8 mmol) in pyrrolidine (14 mL, 11.928 g, 168
mmol, 20 equiv) was heated under reflux for 15 h. The reaction
solution was treated as described in the general methods to give
10 as a brown solid (3.295 g, 75%): mp >350 °C (dec); 1H NMR
(DMSO-d6) δ 1.297-1.367 (m, 4H), 1.419-1.490 (m, 4H), 1.634
(m, 12H), 2.370 (m, 12H), 3.309 (t, 4H, J ) 7.0 Hz), 7.186 (dd,
2H, J ) 1.3 Hz, 8.8 Hz), 8.080 (d, 2H, J ) 8.9 Hz), 8.132 (s, 2H),
10.239 (bs, 2H), 11.615 (bs, 1H); 13C NMR (DMSO-d6) δ 23.0,

596 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2
Moore et al.
3,6-[4-(Pyrrolidin-1-yl)butanamido]-9-[4′-(N,N-dimethylami-
no)anilino]acridine (1b). By Method A. A dark red brown solution
of N,N-dimethylaminoaniline (26 mg, 0.19 mmol, 1.1 equiv) and
12 (87 mg, 0.17 mmol) in anhydrous methanol (7 mL) was heated
under reflux for 4 d. Product isolation and purification by flash
chromatography eluting with a 3:6:1 mixture of methanol/dichlo-
romethane/triethylamine gave 1b as a red brown solid (13 mg,
13%): mp 112 °C (dec); 1H NMR (CDCl3, 3 mg/mL) δ 1.821 (m,
8H), 1.889 (m, 4H), 2.521 (m, 4H), 2.573 (m, 12H), 2.937 (s, 6H),
6.693 (d, 2H, J ) 8.6 Hz), 6.903 (d, 2H, J ) 7.8. Hz), 7.178 (bs,
2H), 7.825 (bs, 2H), 7.865 (bs, 2H), 10.229 (bs, 2H); HRMS (+ESI)
m/z 622.3866 (MH+, C37H48N7O2 requires 622.3864).
10.063 (s, 2H), 10.519 (s, 1H), 11.032 (s, 1H), 11.066 (s, 2H),
13.861 (s, 1H); HRMS (+ESI) m/z 691.4078 (C42H35N8O3 requires
691.4078).
3,6-Bis[3-(pyrrolidin-1-yl)propionamido]-9-{4′-[3′′-(pyrroli-
din-1-yl)butanamido]anilino}acridine (1g). By Method A. A
solution of 15c (83 mg, 0.3 mmol, 3 equiv) in MeOH (5 mL) was
added dropwise to a solution of 11 (50 mg, 0.1 mmol) in MeOH
(5 mL) and the resulting solution heated at 40 °C for 5 h. Product
isolation and then purification by recrystallization from MeOH/
diethyl ether afforded 1g as a red brown solid (39 mg, 55%): mp
168-170 °C; 1H NMR (CD3OD) δ 1.788-1.955 (m, 14H), 2.428
(t, 2H, J ) 7.2 Hz), 2.674-2.774 (m, 10H), 2.962 (t, 4H, J ) 7.1
Hz), 3.147 (t, 2H, J ) 6.9 Hz), 3.341-3.472 (m, 6H), 6.684
(d, 2H, J ) 8.8 Hz), 7.055(d, 2H, J ) 8.7 Hz), 7.137 (dd, 2H,
J ) 1.9, 9.4 Hz), 7.961 (d, 2H, J ) 9.4 Hz), 8.360 (d, 2H, J ) 1.9
Hz); HRMS (+ESI) m/z 705.4206 (MH+, C41H53N8O3 requires
705.4235).
3,6-Bis[4-(pyrrolidin-1-yl)butanamido]-9-{4′-[3′′-(pyrrolidin-
1-yl)propanamido]anilino}acridine (1h). By Method A. A solu-
tion of 9-chloroacridine 12 (534 mg, 1 mmol) and the aniline 15b
(725 mg, 3 mmol, 3 equiv) in methanol (20 mL) was heated under
reflux for 4 days. Product isolation and purification by flash
chromatography eluting with 3:6:1 methanol/dichloromethane/
triethylamine gave 1d as a red brown solid (82 mg, 11%).
3,6-Bis[5-(pyrrolidin-1-yl)pentanamido]-9-[4′-(N,N-dimeth-
ylamino)anilino]acridine (1c). By Method A. An ice-cooled
solution of N,N-dimethylaminoaniline dihydrochloride salt (170 mg,
0.8 mmol, 1.1 equiv) and triethylamine (125 µL, 91 mg, 0.9 mmol,
1.2 equiv) in anhydrous methanol (10 mL) was stirred for 15
minutes under nitrogen and then added slowly to an ice-cooled
solution of 13 (408 mg, 0.74 mmol) in methanol (5 mL). The
resulting solution was allowed to come to room temperature and
then heated at reflux for 2 days under nitrogen. Product isolation
and purification by flash chromatography eluting with 3:5:2
dichloromethane/methanol/triethylamine gave 1c as a red brown
solid (12 mg, 2%): mp 131-133 °C; 1H NMR (CDCl3, 1 mg/mL)
δ 1.596 (m, 4H), 1.751 (bs, 12H), 2.487 (bs, 16H), 2.918 (s, 6H),
6.867 (d, 2H, J ) 7.9 Hz), 6.662 (d, 2H, J ) 8.5 Hz), 7.405 (bm,
2H), 7.807 (bm, 2H), 7.869 (bm, 2H), 9.363 (bs, 2H); HRMS
(+ESI) m/ z 650.4180 (MH+, C39H52N7O2 requires 650.4177).
3,6-Bis[5-(pyrrolidin-1-yl)hexanamido]-9-[4′-(N,N-dimeth-
ylamino)anilino]acridine (1d). By Method A. An ice-cooled
solution of N,N-dimethylaminoaniline dihydrochloride salt (99 mg,
0.47 mmol, 1.1 equiv) and triethylamine (0.2 mL, 1.43 mmol, 3.3
equiv) in anhydrous methanol (10 mL) was stirred for 15 minutes
under nitrogen and then added slowly to a solution of 14 (248 mg,
0.43 mmol) in anhydrous methanol (10 mL). The resulting solution
was allowed to come to room temperature and then heated at reflux
for 2.5 d under nitrogen. Product isolation and purification by flash
chromatography eluting with 3:6:1 methanol/dichloromethane/
triethylamine gave 1d as a red solid (19 mg, 12%): mp 159-161
°C; 1H NMR (CDCl3, 2 mg/mL) δ 1.379 (m, 4H), 1.589 (m, 4H),
1.688 (m, 4H), 1.805 (bs, 8H), 2.509 (m, 8H), 2.624 (bs, 8H), 2.935
(s, 6H), 6.627 (d, 2H, J ) 8.2 Hz), 7.008 (d, 2H, J ) 9.5 Hz),
7.519 (m, 2H), 7.833 (dd, 2H, J ) 1.8, 7.5 Hz), 8.024 (bs, 2H),
9.977 (bs, 2H); HRMS (+ESI) m/z 678.4501 (MH+, C41H56N7O2
requires 678.4490).
3,6-Bis[3-(pyrrolidin-1-yl)propionamido]-9-{4′-[2′′-(pyrroli-
din-1-yl)acetamido]anilino}acridine (1e). By Method A. A
solution of 15a (132 mg, 0.6 mmol) and 11 (300 mg, 0.6 mmol) in
methanol (5 mL) was heated at 60 °C for 6 h. Product isolation
and purification by flash chromatography eluting with 3:6:1
methanol/dichloromethane/triethylamine gave 1e as a red brown
solid (412 mg, 61%): mp 207-211 °C; 1H NMR (1e·4TFA,
DMSO-d6) δ 1.832 (m, 2H), 1.886 (m, 6H), 2.028 (m, 6H), 2.814
(d, 2H, J ) 4.645 Hz), 2.974 (t, 4H, J ) 7.24 Hz), 3.086 (m, 6H),
3.493 (m, 6H), 3.572 (m, 6H), 7.336 (dd, 2H, J ) 1.6, 9.5 Hz),
7.382 (d, 2H, J ) 8.8 Hz), 7.719 (d, 2H, J ) 8.9 Hz), 8.094 (d,
2H, J ) 9.4 Hz), 8.508 (d, 2H, J ) 1.6 Hz), 10.003 (s, 2H), 10.187
(s, 1H), 10.860 (s, 1H), 11.039 (s, 1H), 11.064 (s, 2H), 13.899 (s,
1H); HRMS (+ESI) m/z 677.3949 (MH+, C42H35N8O3 requires
677.3922).
By Method B. A solution of 19 (10 mg, 0.015 mmol) in
anhydrous pyrrolidine (1 mL) was stirred overnight at room
temperature. Product isolation and purification afforded 1h as a
1
red brown solid (7 mg, 60%): mp 176-179 °C; H NMR (CD3-
OD) δ 1.977-2.100 (m, 16H, J ) 7.1 Hz), 2.530 (t, 2H, J ) 6.51),
2.607 (t, 4H, J ) 7.1 Hz), 2.78 (t, 2H, J ) 7.1 Hz), 3.026 (m, 4H),
3.088 (t, 4H, J ) 8.0 Hz), 3.180 (m, 8H), 7.171 (d, 2H, J ) 8.8
Hz), 7.266 (dd, 2H, J ) 1.9, 9.4 Hz), 7.646 (d, 2H, J ) 8.9 Hz),
7.991 (d, 2H, J ) 9.4 Hz), 8.405 (d, 2H, J ) 1.8 Hz); HRMS
(+ESI) m/z 719.4413 (MH+, C42H55N8O3 requires 719.4391).
3,6-Bis[5-(pyrrolidin-1-yl)pentanamido]-9-{4′-[3′′-(pyrrolidin-
1-yl)propanamido]anilino}acridine (1i). By Method A. A solution
of 15b (7 mg, 0.2 mmol) and 13 (100 mg, 0.2 mmol) in methanol
(10 mL) was heated under reflux for 4 days. Product isolation and
purification by flash chromatography eluting with 3:6:1 methanol/
dichloromethane/triethylamine afforded 1i as a red brown solid (21
mg, 16%).
By Method B. A solution of 20 (10 mg, 0.02 mmol) in anhydrous
pyrrolidine (1 mL) was stirred overnight at room temperature.
Product isolation and purification afforded 1e as a red brown solid
1
(8 mg, 74%): mp 187-189 °C; H NMR (CD3OD) δ 1.762 (m,
8H), 1.955 (m, 8H), 1.984 (t, 4H, J ) 7.1 Hz), 2.513 (t, 4H, J )
6.8 Hz), 2.668 (t, 2H, J ) 7.2 Hz), 2.789 (m, 4H), 2.909 (t, 4H, J
) 7.7 Hz), 3.001 (m, 8H), 3.217 (t, 2H, J ) 7.2 Hz), 7.037 (d, 2H,
J ) 8.8 Hz), 7.213 (dd, 2H, J ) 1.9, 9.3 Hz), 7.560 (d, 2H, J )
8.8 Hz), 7.950 (d, 2H, J ) 9.3 Hz), 8.312 (d, 2H, J ) 1.6 Hz);
HRMS (+ESI) m/z 747.4702 (MH+, C44H59N8O3 requires 747.4705).
3,6-Bis[4-(pyrrolidin-1-yl)hexanamido]-9-{4′-[3′′-(pyrrolidin-
1-yl)propanamido]anilino}acridine (1j). By Method A. A solu-
tion of 15b (128 mg, 0.6 mmol, 1.2 equiv) and 14 (284 mg, 0.5
mmol) in methanol was heated under reflux for 2.5 d. Product
isolation and purification as described in the general methods
afforded 1f as a red brown solid (44 mg, 12%).
By Method B. A solution of 21 (12 mg, 0.017 mmol) in
anhydrous pyrrolidine (1 mL) was stirred overnight at room
temperature. Product isolation and purification by flash chroma-
tography eluting with 3:6:1 methanol/dichloromethane/triethylamine
afforded 1f as a red brown solid (5 mg, 38%): mp 134-136 °C;
1H NMR (CD3OD) δ 1.471 (quintet, 4H, J ) 7.6 Hz), 1.695-
1.816 (m, 8H), 1.910 (m, 4H), 1.987 (m, 8H), 2.488 (t, 4H, J )
7.4), 2.691 (t, 2H, J ) 7.2 Hz), 2.833 (m, 4H), 2.969 (t, 4H, J )
8.1 Hz), 3.048 (t, 2H, J ) 7.0 Hz), 3.100 (m, 8H), 7.068 (d, 2H,
J ) 8.8 Hz), 7.230 (dd, 2H, J ) 2.0, 9.3 Hz), 7.582 (d, 2H, J )
8.8 Hz), 7.953 (d, 2H, J ) 9.3 Hz), 8.327 (d, 2H, J ) 1.8 Hz);
HRMS (+ESI) m/z 775.5029 (MH+, C46H63N8O3 requires 775.5018).
3,6-Bis[3-(pyrrolidin-1-yl)propionamido]-9-{4′-[3′′-(pyrroli-
din-1-yl)propanamido]anilino}acridine (1f). By Method A. A
solution of 15b (430 mg, 1.8 mmol, 3 equiv) and 11 (300 mg, 0.6
mmol) in methanol (50 mL) was heated under reflux for 15 h.
Product isolation and purification by flash chromatography eluting
with 3:6:1 methanol/dichloromethane/triethylamine gave 1f as a red
brown solid (290 mg, 60%): mp 147-149 °C; 1H NMR (1f·4TFA,
DMSO-d6) δ 1.888 (m, 6H), 2.035 (m, 6H), 2.874 (t, 2H, J ) 7.215
Hz), 2.969 (t, 4H, J ) 7.3 Hz), 3.086 (m, 6H), 3.485 (m, 6H),
3.571 (m, 6H), 7.337 (m, 4H), 7.724 (d, 2H, J ) 8.91 Hz), 8.087
(d, 2H, J ) 9.5 Hz), 8.492 (d, 2H, J ) 1.9 Hz), 9.932 (s, 1H),

G-Quadruplex Stabilization and Telomerase Inhibition
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2 597
(21) Gowan, S. M.; Heald, R.; Stevens, M. F.; Kelland, L. R. Potent
inhibition of telomerase by small-molecule pentacyclic acridines
capable of interacting with G-quadruplexes. Mol. Pharm. 2001, 60,
981-988.
(22) Alberti, P.; Schmitt, P.; Nguyen, C. H.; Rivalle, C.; Hoarau, M.;
Grierson, D. S.; Mergny, J. L. Benzoindoloquinolines interact with
DNA tetraplexes and inhibit telomerase. Bioorg. Med. Chem. Lett.
2002, 12, 1071-1074.
(23) Kim, M. Y.; Vankayalapati, H.; Shin-Ya, K.; Wierzba, K.; Hurley,
L. H. Telomestatin, a potent telomerase inhibitor that interacts quite
specifically with the human telomeric intramolecular G-quadruplex.
J. Am. Chem. Soc. 2002, 124, 2098-2099.
(24) Gowan, S. M.; Harrison, R. J.; Patterson, L.; Valenti, M.; Read, M.
A.; Neidle, S.; Kelland, L. R. A G-quadruplex-interactive potent
small-molecule inhibitor of telomerase exhibiting in vitro and in vivo
antitumor activity. Mol. Pharm. 2002, 61, 1154-1162.
Acknowledgment. This work has been supported by Cancer
Research UK (Program and Project Grants to S. N. and a
Research Studentship to C.M.S.), by grants from the Association
for International Cancer Research and Antisoma Ltd (to S.N.),
and by NIH Grant No. GM61587 (to W.D.W.). We are grateful
to various colleagues for advice and discussions, especially Prof.
Lloyd Kelland (Antisoma).
Supporting Information Available: Experimental procedures
and spectral data for compounds 15a-c and 23, rpHPLC purity
data, and SPR sensorgrams for compound 1b. This material is
available free of charge via the Internet at http://pubs.acs.org.
(25) (a) Leonetti, C.; Amodei, S.; D’Angelo, C.; Rizzo, A.; Benassi, B.;
Antonelli, A.; Elli, R.; Stevens, M. F. G.; D’Incalci, M.; Zupi, G.;
Biroccio, A. Biological Activity of the G-quadruplex ligand RHPS4
(3,11-difluoro-6,8,13-trimethyl-8H-quino[4,3,2-kl] acridinium metho-
sulfate) is associated with telomere capping alteration. Mol. Pharm.
2004, 66, 1138-1146. (b) Granotier, C.; Pennarun, G.; Riou, L.;
Hoffschir, F.; Gauthier, L. R.; de Chian, A.; Gomez, D.; Mandine,
E.; Riou, J.-F.; Mergny, J.-L.; Mailliet, P.; Dutrillaux, B.; Boussin,
F. D. Preferential binding of a G-quadruplex ligand to human
chromosome ends. Nucleic Acids Res. 2005, 33, 4182-4190. (c)
Pennarun, G.; Granotier, C.; Gauthier, L. R.; Gomez, D.; Hoffschir,
F.; Mandine, E.; Riou, J.-F.; Mergny, J.-L.; Mailliet, P.; Boussin, F.
Apoptosis related to telomere instability and cell cycle alterations in
human giolma cells treated by new highly selective G-quadruplex
ligands. Oncogene 2005, 24, 2917-2928.
(26) Gomez, D.; Paterski, R.; Lemarteleur, T.; Shin-ya, K.; Mergny, J.-
L.; Riou, J.-F. Interaction of telomestatin with the telomeric single-
strand overhang. J. Biol. Chem. 2004, 279, 41487-41494.
(27) Lei, M.; Podell, E. R.; Cech, T. R. Structure of human POT1 bound
to telomeric single-stranded DNA provides a model for chromosome
end-protection. Nature Struct. Mol. Biol. 2004, 11, 1223-1229.
(28) Colgin, L. M.; Baran, K.; Baumann, P.; Cech, T. R.; Reddel, R. R.
Human POT1 facilitates telomere elongation by telomerase. Curr.
Biol. 2003, 13, 942-946.
(29) (a) Baumann, P.; Cech, T. R. Pot1, the putative telomere end-binding
protein in fission yeast and humans. Science 2001, 292, 1171-1175.
(b) Zaug, A. J.; Podell, E. R.; Cech, T. R. Human POT1 disrupts
telomeric G-quadruplexes allowing telomerase extension in vitro.
Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 10864-10869. (c) Paeschek,
K.; Simonsson, T.; Postberg, J.; Rhodes, D.; Lipps, H. J. Telomere
end-binding proteins control the formation of G-quadruplex DNA
structures in vivo. Nature Struct. Mol. Biol. 2005, in press.
(30) Incles, C. M.; Schultes, C. M.; Kempski, H.; Koehler, H.; Kelland,
L. R.; Neidle, S. A G-quadruplex telomere targeting agent produces
p16-associated senescence and chromosomal fusions in human
prostate cancer cells. Mol. Cancer Ther. 2004, 3, 1201-1206.
(31) Sun, H.; Karow, J. K.; Hickson, I. D.; Maizels, N. The Bloom’s
syndrome helicase unwinds G4 DNA. J. Biol. Chem. 1998, 273,
27587-27592.
(32) Opresko, P. L.; Von Kobbe, C.; Laine, J. P.; Harrigan, J.; Hickson,
I. D.; Bohr, V. A. Telomere-binding protein TRF2 binds to and
stimulates the Werner and Bloom syndrome helicases. J. Biol. Chem.
2002, 277, 41110-41119.
(33) Li, J. L.; Harrison, R. J.; Reszka, A. P.; Brosh, R. M., Jr.; Bohr, V.
A.; Neidle, S.; Hickson, I. D. Inhibition of the Bloom’s and Werner’s
syndrome helicases by G-quadruplex interacting ligands. Biochemistry
2001, 40, 15194-15202.
(34) Burger, A. M.; Dai, F.; Schultes, C. M.; Reszka, A. P.; Moore, M.
J.; Double, J. A.; Neidle, S. The G-quadruplex-interactive molecule
BRACO-19 inhibits tumor growth, consistent with telomere targeting
and interference with telomerase function. Cancer Res. 2005, 65,
1489-1496.
(35) Chen, Z.; Corey, D. R. Telomerase inhibitors: A new option for
chemotherapy. AdV. Cancer Res. 2003, 87, 31-58.
(36) Shay, J. W.; Wright, W. E. Telomerase: A target for cancer
therapeutics. Cancer Cell 2002, 2, 257-265.
(37) (a) Mergny, J. L.; Riou, J. F.; Mailliet, P.; Teulade-Fichou, M. P.;
Gilson, E. Natural and pharmacological regulation of telomerase.
Nucleic Acids Res. 2002, 30, 839-865. (b) Kelland, L. R. Overcom-
ing the immortality of tumour cells by telomere and telomerase based
cancer therapeuticssCurrent status and future prospects. Eur. J.
Cancer 2005, 41, 971-979.
(38) Harrison, R. J.; Gowan, S. M.; Kelland, L. R.; Neidle, S. Human
telomerase inhibition by substituted acridine derivatives. Bioorg. Med.
Chem. Lett. 1999, 9, 2463-2468.
References
(1) Blackburn, E. H. Telomere states and cell fates. Nature 2000, 408,
53-56.
(2) Blackburn, E. H. Switching and signaling at the telomere. Cell 2001,
106, 661-673.
(3) Karlseder, J.; Smogorzewska, A.; de Lange, T. Senescence induced
by altered telomere state, not telomere loss. Science 2002, 295, 2446-
2449.
(4) Liu, D.; O’Connor, M. S.; Qin, J.; Songyang, Z. TelosomesA
mammalian telomere associated complex formed by multiple telo-
meric proteins. J. Biol. Chem. 2004, 279, 51338-51342.
(5) (a) Smogorzewska, A.; de Lange, T. Regulation of telomerase by
telomeric proteins. Annu. ReV. Biochem. 2004, 73, 177-208. (b) de
Lange, T. Shelterin: The protein complex that shapes and safeguards
human telomeres. Genes DeV. 2005, 19, 2100-2110.
(6) Li, G. Z.; Eller, M. S.; Firoozabadi, R.; Gilchrest, B. A. Evidence
that exposure of the telomere 3′ overhang sequence induces senes-
cence. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 527-531.
(7) Stewart, S. A.; Ben Porath, I.; Carey, V. J.; O’Connor, B. F.; Hahn,
W. C.; Weinberg, R. A. Erosion of the telomeric single-strand
overhang at replicative senescence. Nat. Genet. 2003, 33, 492-496.
(8) Steinert, S.; Shay, J. W.; Wright, W. E. Transient expression of human
telomerase extends the life span of normal human fibroblasts.
Biochem. Biophys. Res. Commun. 2000, 273, 1095-1098.
(9) Counter, C. M.; Avilion, A. A.; LeFeuvre, C. E.; Stewart, N. G.;
Greider, C. W.; Harley, C. B.; Bacchetti, S. Telomere shortening
associated with chromosome instability is arrested in immortal cells
which express telomerase activity. EMBO J. 1992, 11, 1921-1929.
(10) Hemann, M. T.; Strong, M. A.; Hao, L. Y.; Greider, C. W. The
shortest telomere, not average telomere length, is critical for cell
viability and chromosome stability. Cell 2001, 107, 67-77.
(11) Masutomi, K.; Yu, E. Y.; Khurts, S.; Ben-Porath, I.; Currier, J. L.;
Metz, G. B.; Brooks, M. W.; Kaneko, S.; Murakami, S.; DeCaprio,
J. A.; Weinberg, R. A.; Stewart, S. A.; Hahn, W. C. Telomerase
maintains telomere structure in normal human cells. Cell 2003, 114,
241-253.
(12) Counter, C. M.; Hirte, H. W.; Bacchetti, S.; Harley, C. B. Telomerase
activity in human ovarian carcinoma. Proc. Natl. Acad. Sci. U.S.A.
1994, 91, 2900-2904.
(13) Kim, N. W.; Piatyszek, M. A.; Prowse, K. R.; Harley, C. B.; West,
M. D.; Ho, P. L.; Corviello, G. M.; Wright, W. E.; Weinrich, S. L.;
Shay, J. W. Specific association of human telomerase activity with
immortal cells and cancer. Science 1994, 266, 2011-2015.
(14) Shay, J. W.; Bacchetti, S. A survey of telomerase activity in human
cancer. Eur. J Cancer 1997, 33, 787-791.
(15) Hahn, W. C.; Counter, C. M.; Lundberg, A. S.; Beijersbergen, R.
L.; Brooks, M. W.; Weinberg, R. A. Creation of human tumour cells
with defined genetic elements. Nature 1999, 400, 464-468.
(16) Neidle, S.; Parkinson, G. Telomere maintenance as a target for
anticancer drug discovery. Nat. ReV. Drug DiscoVery 2002, 1, 383-
393.
(17) Neidle, S.; Parkinson, G. N. The structure of telomeric DNA. Curr.
Opin. Struct. Biol. 2003, 13, 275-283.
(18) Parkinson, G. N.; Lee, M. P.; Neidle, S. Crystal structure of parallel
quadruplexes from human telomeric DNA. Nature 2002, 417, 876-
880.
(19) Harrison, R. J.; Cuesta, J.; Chessari, G.; Read, M. A.; Basra, S. K.;
Reszka, A. P.; Morrell, J.; Gowan, S. M.; Incles, C. M.; Tanious, F.
A.; Wilson, W. D.; Kelland, L. R.; Neidle, S. Trisubstituted acridine
derivatives as potent and selective telomerase inhibitors. J. Med.
Chem. 2003, 46, 4463-4476.
(20) Read, M.; Harrison, R. J.; Romagnoli, B.; Tanious, F. A.; Gowan,
S. M.; Reszka, A. P.; Wilson, W. D.; Kelland, L. R.; Neidle, S.
Structure-based design of selective and potent G quadruplex-mediated
telomerase inhibitors. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 4844-
4849.

598 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2
Moore et al.
(39) Schultes, C. M.; Guyen, B.; Cuesta, J.; Neidle, S. Synthesis,
biophysical and biological evaluation of 3,6-bis-amidoacridines with
extended 9-anilino substituents as potent G-quadruplex-binding
telomerase inhibitors. Bioorg. Med. Chem. Lett. 2004, 14, 4347-
4351.
(40) Harrison, R. J.; Reszka, A. P.; Haider, S. M.; Romagnoli, B.; Morrell,
J.; Read, M. A.; Gowan, S. M.; Incles, C. M.; Kelland, L. R.; Neidle,
S. Evaluation of disubstituted acridone derivatives as telomerase
inhibitors: The importance of G-quadruplex binding. Bioorg. Med.
Chem. Lett. 2004, 14, 5845-5849.
(41) Matsumura, K. The synthesis of certain acridine compounds. J. Am.
Chem. Soc. 1929, 51, 816-820.
(42) Gamage, S. A.; Tepsiri, N.; Wilairat, P.; Wojcik, S. J.; Figgitt, D.
P.; Ralph, R. K.; Denny, W. A. Synthesis and in vitro evaluation of
9-anilino-3,6-diaminoacridines active against a multidrug-resistant
strain of the malaria parasite Plasmodium falciparum. J. Med. Chem.
1994, 37, 1486-1494.
(43) Spackova, N.; Berger, I.; Sponer, J. Nanosecond molecular dynamics
simulations of parallel and antiparallel guanine quadruplex DNA
molecules. J. Am. Chem. Soc. 1999, 121, 5519-5534.
(44) Spackova, N.; Berger, I.; Sponer, J. Structural dynamics and cation
interactions of DNA quadruplex molecules containing mixed guanine/
cytosine quartets revealed by large-scale MD simulations. J. Am.
Chem. Soc. 2001, 123, 3295-3307.
(45) Fadrna, E.; Spackova, N.; Stefl, R.; Koca, J.; Cheatham, T. E.; Sponer,
J. Molecular dynamics simulations of guanine quadruplex loops:
Advances and force field limitations. Biophys. J. 2004, 87, 227-
242.
(46) Hazel, P.; Huppert, J.; Balasubramanian, S.; Neidle, S. Loop-length-
dependent folding of G-quadruplexes. J. Am. Chem. Soc. 2004, 126,
16405-16415.
(47) Skehan, P.; Storeng, R.; Scudiero, D.; Monks, A.; McMahon, J.;
Vistica, D.; Warren, J. T.; Bokesch, H.; Kenney, S.; Boyd, M. R.
New colorimetric cytotoxicity assay for anti-cancer-drug screening.
J. Natl. Cancer Inst. 1990, 82, 1107-1112.
(48) Papazisis, K. T.; Geromichalos, G. D.; Dimitriadis, K. A.; Kortsaris,
A. H. Optimization of the sulforhodamine B colorimetric assay. J.
Immunol. Methods 1997, 208, 151-158.
(49) Hwang, E. S. Replicative senescence and senescence-like state
induced in cancer-derived cells. Mech. Aging DeV. 2002, 123, 1681-
1694.
(50) Roninson, I. B. Tumor senescence as a determinant of drug response
in vivo. Drug Resistance Updates 2002, 5, 204-208.
(51) Roninson, I. B.; Broude, E. V.; Chang, B. D. If not apoptosis, then
what? Treatment-induced senescence and mitotic catastrophe in tumor
cells. Drug Resistance Updates 2001, 4, 303-313.
(52) Guyen, B.; Schultes, C. M.; Hazel, P.; Mann, J.; Neidle, S. Synthesis
and evaluation of analogues of 10H-indolo[3,2-b]quinoline as G-
quadruplex stabilising ligands and potential inhibitors of the enzyme
telomerase. Org. Biomol. Chem. 2004, 2, 981-988.
(53) Mergny, J. L.; Lacroix, L.; Teulade-Fichou, M. P.; Hounsou, C.;
Guittat, L.; Hoarau, M.; Arimondo, P. B.; Vigneron, J. P.; Lehn, J.
M.; Riou, J. F.; Garestier, T.; Helene, C. Telomerase inhibitors based
on quadruplex ligands selected by a fluorescence assay. Proc. Natl.
Acad. Sci. U.S.A. 2001, 98, 3062-3067.
(54) Koeppel, F.; Riou, J. F.; Laoui, A.; Mailliet, P.; Arimondo, P. B.;
Labit, D.; Petitgenet, O.; Helene, C.; Mergny, J. L. Ethidium
derivatives bind to G-quartets, inhibit telomerase and act as fluores-
cent probes for quadruplexes. Nucleic Acids Res. 2001, 29, 1087-
1096.
(55) Risitano, A.; Fox, K. R. Influence of loop size on the stability of
intramolecular DNA quadruplexes. Nucleic Acids Res. 2004, 32,
2598-2606.
(60) Denny, W. A.; Cain, B. F.; Atwell, G. J.; Hansch, C.; Panthananickal,
A.; Leo, A. Potential antitumor agents. 36. Quantitative relationships
between experimental antitumor activity, toxicity, and structure for
the general class of 9-anilinoacridine antitumor agents. J. Med. Chem.
1982, 25, 276-315.
(61) Denny, W. A.; Atwell, G. J.; Baguley, B. C. Potential antitumor
agents. 39. Anilino ring geometry of amsacrine and derivatives:
Relationship to DNA binding and antitumor activity. J. Med. Chem.
1983, 26, 1625-1630.
(62) Gao, H.; Denny, W. A.; Garg, R.; Hansch, C. Quantitative structure-
activity relationships (QSAR) for 9-anilinoacridines: A com-
parative analysis. Chemico-Biol. Interactions 1998, 116, 157-
180.
(63) Haider, S. M.; Parkinson, G. N.; Neidle, S. Structure of a G-
quadruplex-ligand complex. J. Mol. Biol. 2003, 326. 117-25.
(64) Shammas, M. A.; Shmookler Reis, R. J.; Akiyama, M.; Koley, H.;
Chauhan, D.; Hideshima, T.; Goyal, R. K.; Hurley, L. H.; Anderson,
K. C.; Munshi, N. C. Telomerase inhibition and cell growth arrest
by G-quadruplex interactive agent in multiple myeloma. Mol. Cancer
Ther. 2003, 2, 825-833.
(65) Sidorov, I. A.; Hirsch, K. S.; Harley, C. B.; Dimitrov, D. S. Cancer
cell dynamics in the presence of telomerase inhibitors: Analysis of
in vitro data. J. Theor. Biol. 2002, 219, 225-233.
(66) Sidorov, I. A.; Hirsch, K. S.; Harley, C. B.; Dimitrov, D. S. Cancer
treatment by telomerase inhibitors: Predictions by a kinetic model.
Mathematical Biosciences 2003, 181, 209-221.
(67) Stewart, S. A.; Hahn, W. C.; O’Connor, B. F.; Banner, E. N.;
Lundberg, A. S.; Modha, P.; Mizuno, H.; Brooks, M. W.; Fleming,
M.; Zimonjic, D. B.; Popescu, N. C.; Weinberg, R. A. Telomerase
contributes to tumorigenesis by a telomere length-independent
mechanism. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 12606-12611.
(68) Zou, Y.; Sfeir, A.; Gryaznov, S. M.; Shay, J. W.; Wright, W. E.
Does a sentinel or a subset of short telomeres determine replicative
senescence? Mol. Biol. Cell 2004, 15, 3709-3718.
(69) Siddiqui-Jain, A.; Grand, C. L.; Bearss, D. J.; Hurley, L. H. Direct
evidence for a G-quadruplex in a promoter region and its targeting
with a small molecule to repress c-MYC transcription. Proc. Natl.
Acad. Sci. U.S.A. 2002, 99, 11593-11598.
(70) (a) Huppert, J.; Balasubramanian, S. Prevalence of quadruplexes in
the human genome. Nucleic Acids Res. 2005, 33, 2908-2916. (b)
Todd, A. K.; Johnston, M.; Neidle, S. Highly prevalent putative
quadruplex sequence motifs in human DNA. Nucleic Acids Res. 2005,
33, 2901-2907.
(71) Rankin, S.; Reszka, A. P.; Huppert, J.; Zloh, M.; Parkinson, G. N.;
Todd, A. K.; Ladame, S. L.; Balasubramanian, S.; Neidle, S. Putative
DNA quadruplex formation within the human c-kit oncogene. J. Am.
Chem. Soc. 2005, 127, 10584-10589.
(72) Lemarteleur, T.; Gomez, D.; Paterski, R.; Mandine, E.; Mailliet, P.;
Riou, J.-F. Stabilization of the c-myc gene promoter quadruplex by
specific ligands' inhibitors of telomerase. Biochem. Biophys. Res.
Commun. 2004, 323, 802-888.
(73) Kim, N. W.; Wu, F. Advances in quantification and characterization
of telomerase activity by the telomeric repeat amplification protocol
(TRAP). Nucleic Acids Res. 1997, 25, 2595-2597.
(74) Burger, A. M. Standard TRAP Assay. In Telomeres and Telom-
erase: Methods and Protocols; Double, J. A., Thompson, M. J., Eds.;
Humana Press Inc.: Totowa, NJ, 2002; pp 109-124.
(75) Gomez, D.; Mergny, J. L.; Riou, J. F. Detection of telomerase
inhibitors based on G-quadruplex ligands by a modified telomeric
repeat amplification protocol assay. Cancer Res. 2002, 62, 3365-
3368.
(76) Bradford, M. M. A rapid and sensitive method for the quantitation
of microgram quantities of protein utilizing the principle of protein-
dye binding. Anal. Biochem. 1976, 72, 248-254.
(77) Halgren, T. A. Merck Molecular Force Field. I. Basis, form, scope,
parametrization and performance of MMFF94. J. Comput. Chem.
1996, 17, 490-519.
(56) Stefl, R.; Cheatham, T. E.; Spackova, N.; Fadrna, E.; Berger, I.; Koca,
J.; Sponer, J. Formation pathways of a guanine-quadruplex DNA
revealed by molecular dynamics and thermodynamic analysis of the
substates. Biophys. J. 2003, 85, 1787-1804.
(57) Williamson, J. R.; Raghuraman, M. K.; Cech, T. R. Monovalent
cation-induced structure of telomeric DNA: The G-quartet model.
Cell 1989, 59, 871-880.
(58) Schultze, P.; Hud, N. V.; Smith, F. W.; Feigon, J. The effect of
sodium, potassium and ammonium ions on the conformation of the
dimeric quadruplex formed by the Oxytricha nova telomere repeat
oligonucleotide d(G(4)T(4)G(4)). Nucleic Acids Res. 1999, 27, 3018-
3028.
(59) Gavathiotis, E.; Heald, R. A.; Stevens, M. F.; Searle, M. S.
Recognition and Stabilization of quadruplex DNA by a potent new
telomerase inhibitor: NMR studies of the 2:1 complex of a
pentacyclic methylacridinium cation with d(TTAGGGT)(4). Angew.
Chem. Int. Ed. 2001, 40, 4749-4751.
(78) Halgren, T. A. Merck Molecular Force Field. VII. Characterization
of MMFF94, MMFF94s, and other widely available force fields for
conformational energies and for intermolecular-interaction energies
and geometries. J. Comput. Chem. 1999, 20, 730-748.
(79) Halgren, T. A. Merck Molecular Force Field. VI. MMFF94s option
for energy minimization studies. J. Comput. Chem. 1999, 20, 720-
729.
(80) Powell, M. J. D. Restart procedures for the conjugate gradient method.
Math. Programming 1977, 12, 241-254.
(81) Judson, R. S.; Jaeger, E. P.; Treasurywala, A. M.; Peterson, M. L.
Conformational searching methods for small molecules 2. Genetic
algorithm approach. J. Comput. Chem. 1993, 14, 1407-1414.
(82) McGarrah, D. B.; Judson, R. S. Analysis of the genetic algorithm
method of molecular conformation determination. J. Comput. Chem.
1993, 14, 1385-1395.

G-Quadruplex Stabilization and Telomerase Inhibition
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2 599
(83) Clark, M.; Cramer, R. D., III.; van Opdenbosch, N. Validation of
the general purpose Tripos 5.2 force field. J. Comput. Chem. 1989,
10, 982-1012.
(87) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J.; Klein, M. L.
Comparison of simple potential functions for simulating liquid water.
J. Chem. Phys. 1983, 79, 926-935.
(88) Wang, J.; Cieplak, P.; Kollman, P. A. How well does a restrained
electrostatic potential (RESP) model perform in calculating confor-
mational energies of organic and biological molecules? J. Comput.
Chem. 2000, 21, 1049-1074.
(89) Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D.
A. Development and testing of a general amber force field. J. Comput.
Chem. 2004, 25, 1157-1174.
(90) Darden, T.; York, D.; Pedersen, L. Particle mesh Ewald: An Nlog-
(N) method for Ewald sums in large systems. J. Chem. Phys. 1993,
98, 10089-10092.
(91) Humphrey, W.; Dalke, A.; Schulten, K. VMD: Visual molecular
dynamics. J. Mol. Graphics 1996, 14, 33-38.
(92) Tang, K. C.; Mariuza, R.; Coward, J. K. Synthesis and evaluation of
some stable multisubstrate adducts as specific inhibitors of spermidine
synthase. J. Med. Chem. 1981, 24, 1277-1284.
(84) Pearlman, D. A.; Case, D. A.; Caldwell, J. W.; Ross, W. S.;
Cheatham, T. E.; DeBolt, S.; Ferguson, D.; Seibel, G. L.; Kollman,
P. A. AMBER, a package of computer programs for applying
molecular mechanics, normal-mode analysis, molecular dynamics and
free energy calculations to simulate the structural and energetic
properties of molecules. Comput. Phys. Commun. 1995, 91, 1-41.
(85) Case, D. A.; Pearlman, D. A.; Caldwell, J. W.; Cheatham, T. E.;
Wang, J.; Ross, W. S.; Simmerling, C. L.; Darden, T. A.; Merz, K.
M.; Stanton, R. V.; Cheng, A. L.; Vincent, J. J.; Crowley, M.; Tsui,
V.; Gohlke, H.; Radmer, R. J.; Duan, Y.; Pitera, J.; Massova, I.;
Seibel, G. L.; Singh, U. C.; Weiner, P. K.; Kollman, P. A. AMBER
7. 2002; University of California, San Francisco.
(86) Cornell, W. D.; Cieplak, C. I.; Bayly, I. R.; Gould, I. R.; Merz, K.
M.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.;
Kollman, P. A. A second generation force field for the simulation of
proteins, nucleic acids and organic molecules. J. Am. Chem. Soc.
1995, 117, 5179-5197.
JM050555A