Synthesis, biophysical and biological evaluation of 3,6-bis-amidoacridines with extended 9-anilino substituents as potent G-quadruplex-binding telomerase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2004.05.090

Telomerase and telomere maintenance are emerging targets for the treatment of human cancers. We report here on the targeting of the telomere-telomerase complex with a series of small molecules based on an acridine platform. A series of 3,6-bisamidoacridines with extended 9-anilino sidechains were designed and synthesised as potential telomeric G-quadruplex DNA (G4) interacting compounds. G4-stabilisation was assessed using a high-throughput FRET (fluorescence resonance energy transfer) assay and telomerase inhibition quantified by a modified TRAP (telomerase repeat amplification protocol) method. Within the series, the compounds showed significant G4-stabilising ability (ΔT _m values of 25-36°C at 1μM concentration) and telomerase inhibition in the nanomolar region (^telEC₅₀ values of 80-318nM). Furthermore, a direct correlation between the FRET and TRAP assays was observed, supporting the use of the rapid screening FRET assay for early assessment of potential G4-stabilising telomerase inhibitors.

Full text of DOI:10.1016/j.bmcl.2004.05.090

Bioorganic & Medicinal Chemistry Letters 14 (2004) 4347–4351
Synthesis, biophysical and biological evaluation of
3,6-bis-amidoacridines with extended 9-anilino substituents as
potent G-quadruplex-binding telomerase inhibitors
*
ꢀ
ꢁ
Christoph M. Schultes, Berangere Guyen, Javier Cuesta and Stephen Neidle
Cancer Research UK Biomolecular Structure Group, University of London School of Pharmacy, 29/39 Brunswick Square,
London WC1N 1AX, UK
Received 26 March 2004; revised 30 April 2004; accepted 18 May 2004
Available online
Abstract—Telomerase and telomere maintenance are emerging targets for the treatment of human cancers. We report here on the
targeting of the telomere–telomerase complex with a series of small molecules based on an acridine platform. A series of 3,6-bis-
amidoacridines with extended 9-anilino sidechains were designed and synthesised as potential telomeric G-quadruplex DNA (G4)
interacting compounds. G4-stabilisation was assessed using a high-throughput FRET (fluorescence resonance energy transfer) assay
and telomerase inhibition quantified by a modified TRAP (telomerase repeat amplification protocol) method. Within the series, the
compounds showed significant G4-stabilising ability (DTm values of 25–36 °C at 1 lM concentration) and telomerase inhibition in
the nanomolar region (telEC50 values of 80–318 nM). Furthermore, a direct correlation between the FRET and TRAP assays was
observed, supporting the use of the rapid screening FRET assay for early assessment of potential G4-stabilising telomerase
inhibitors.
Ó 2004 Elsevier Ltd. All rights reserved.
1. Introduction
The rapid screening of the DNA-binding abilities of
various compound series by fluorescence techniques is a
swift, semi-quantitative way of selecting promising new
candidates for further in-depth biological testing.8–10
Demonstration of a correlation between the results ob-
tained from this type of biophysical screen and the
TRAP method10;11 lends support to the proposed
mechanism of action for a particular series of ligands,
and would also speed up the process of comparing and
evaluating telomerase inhibitors that function in this
way.
The telomerase enzyme complex is increasingly regarded
as playing a key role in telomere maintenance and
tumourigenesis, and is thus an important target for the
development of novel cancer therapeutic agents.1;2
Various strategies have been proposed to induce telo-
merase inhibition3 and a promising method has been to
focus on the disruption of the telomere–telomerase
interaction via small molecules capable of interacting
with the enzyme substrate, thereby preventing telomere
elongation by the telomerase complex. A number of
structural studies have shown4–6 that single-stranded
DNA present at the 30-ends of telomeres can fold into
higher-order DNA quadruplex structures. The design of
small molecules capable of binding selectively to, and
stabilising, these structures thus holds promise as a
means of influencing telomere maintenance.1;7
Previous studies from this laboratory have focused on
generating acridine-based molecules with varying pat-
terns of substitution around the aromatic core.12;13
Coupled to the acridine scaffold, the basic 3,6-bis-(3-
pyrrolidin-1-yl-propionamido) substitution pattern has
been shown to be an effective starting point for further
chemical studies12 and recent studies in this laboratory
have thus focused on refining the structure–activity
relationships (SAR) of functional groups at the 9-posi-
tion (Fig. 1). The nature of this substitution is expected
to play a critical role both in enhancing ligand–DNA
interactions and in increasing the selectivity of these
molecules for quadruplex over duplex DNA (dsDNA),
Keywords: Telomerase; Telomere; Quadruplex DNA; FRET; TRAP;
Acridines.
* Corresponding author. Tel.: +44-207-753-5969; fax: +44-207-753-
5970; e-mail: stephen.neidle@ulsop.ac.uk
€
Present address: Max-Planck-Institut fur Molekulare Physiologie,
Otto-Hahn-Strasse 11, 44227 Dortmund, Germany.
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.05.090

4348
C. M. Schultes et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4347–4351
O
H
N
N
X
O
O
HN
HN
O
O
O
N
N
H
N
N
H
N
N
N
N
H
N
N
H
N
1
2
N
N
N
X =
O
1a
1b
1c
N
N
HN
HN
1d
1e
Figure 1. Structures of the trisubstituted acridines examined in this study (1a–e). The related compound 2 (BRACO-19)12;14 is included as a reference.
and a number of potent low-nanomolar inhibitors have
already been described.13;14
propionamido) acridine 1114 to yield 1a–e in 38–57%
yield (Scheme 3). The final compounds were subse-
quently purified by flash chromatography using various
mixtures of methanol (10–40%), dichloromethane (60–
99%) and triethylamine (1–10%).
2. Synthesis
Preparation of the 9-anilino side-chains was carried out
by reacting p-nitroaniline with adipic anhydride,15 to
afford the carbamoyl acid 9 in 86% yield. The latter was
then reacted with suitable primary or secondary amines
using the carbonic–carboxylic mixed anhydride method,
and subsequent reduction of the nitro group using Pd on
charcoal afforded the desired anilines 10a–e in a 60 to
83% yield for the three steps (Scheme 2). The synthesis
of the only noncommercially available amine 6 was
based on the Gabriel synthesis, as has been described16
(Scheme 1). Each of these anilines was then finally
coupled with the 9-chloro-3,6-bis-(3-pyrrolidin-1-yl-
3. Biophysical and biological assays
The compounds were examined in two assays, a FRET
(fluorescence resonance energy transfer) assay to assess
their G4-stabilising ability and a TRAP (telomerase re-
peat amplification protocol) method to quantify their
telomerase inhibitory effect (telEC50).
The FRET assay was conducted using the previously-
reported protocol,10 which enables the measurement of
relative G4-stabilisation values in a 96-well plate format.
O
O
N
a-c
d
OH
N
+
NH
N
Cl
H2N
O
O
3
4
5
6
Scheme 1. Reagents and conditions: (a) potassium phthalimide, DMF, 80 °C; (b) CBr4, PPH3, CH3CN, rt, (c) NHEt2, K2CO3, 1,4-dioxane, rt; (d)
H2NNH2, MeOH, rt.
O
O
OH
X
NH2
HN
HN
O
O
O
e
f-g
+
O
O
NO2
NO2
N
NH2
7
8
9
10a-e
O
N
N
N
X =
N
N
HN
HN
6
Scheme 2. Reagents and conditions: (a) 1,4-dioxane, reflux; (b) H2NR or HNR2, ethyl chloroformate, Et3N, CH2Cl2, 0 °C to rt; (c) 10% Pd/C,
ammonium formate, MeOH, rt.

C. M. Schultes et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4347–4351
4349
O
H
N
X
O
O
Cl
HN
O
O
O
h
N
N
H
N
N
H
N
N
N
H
N
N
H
N
11
1a-e
Scheme 3. Reagents and conditions: (a) 10a–e, MeOH, 60 °C.
In brief, changes in the FRET signal upon melting of a
DNA probe (telomeric sequence d[GGG(TTAGGG)3]
dual-labelled with 6-FAM and TAMRA at the 50- and
30-ends, respectively) were observed under conditions
approximating those used in the crystallisation of the
published structure5 (50 mM cacodylate buffer, pH 7.4).
All five compounds 1a–e showed high levels of G4-sta-
bilisation, with changes in the melting temperature of
the native DNA sequence (DTm values) upon ligand
addition in the range 25–36 °C (at 1 lM ligand concen-
tration; Table 1). Curves fitted to the data using the
Origin package of software (OriginPro, OriginLab
Corp.) showed dose–response behaviour when plotting
DTm against concentration (Fig. 2), and enabled the
calculation of DTm values over a range of concentrations
tested. This provides a useful alternative for measuring
the stabilising ability (e.g., ½concnꢀDT ¼20 °C: concentra-
tion required to achieve a DTm value omf 20 °C; Fig. 2 and
Table 1), which is irrespective of the curve shape and
allows the comparative testing of compounds with large
differences in potency over a wide concentration range.
and telEC50 values below 100 nM (98 and 80 nM,
respectively). The flexibility of the 9-substituted ali-
phatic amine chains are likely to be an important factor
in G4 binding, as this would allow the chains to probe
the quadruplex surface in order to optimise charge–
charge contacts with the phosphate backbone. However,
increased G4–DNA binding in a cellular environment is
likely to be offset by the exposed nature of the terminal
amide on the 9-position, which might be subject to
intracellular amide bond cleavage.
Compounds 1a–c performed less well in the assays,
perhaps due to the greater rigidity and bulkiness of the
45
1a
1b
1c
1d
1e
2
[ligand]=1 µM
40
35
30
25
20
15
10
5
∆Tm=20oC
The modified TRAP assay was performed as previously
described,10;11 using 1 lg total protein extract from the
A2780 ovarian carcinoma cell line. Compounds 1a–e
were examined in this way to a concentration of 1 lM
(Fig. 3), and concomitant controls (PCR inhibition as-
say results, not shown) were carried out to verify that
there was no interference with the correct functioning of
Taq polymerase at these concentrations. Only 1d and 1e
showed significant PCR inhibition (40–50%) at 1 lM,
although this concentration is an order of magnitude
higher than the calculated telEC50 values (see below and
Table 1).
0
0.01
0.1
1
10
µ
M); Log scale
Concentration (
Figure 2. Dose–response curves fitted for the G4-stabilisation (FRET)
experiments carried out on compounds 1a–e and 2. The baseline
(DTm ¼ 0) was set at the melting temperature of the native structure
(53.2 °C). Dotted lines are intended to serve as visual references for the
The two compounds with diethyl amine side chains (1d
and 1e) showed the best results in both assays, with G4
stabilisation of over 30 °C at a 1 M ligand concentration
concn
and DTm (at 1 lM) values reported in Table 1 (hori-
ðDTm¼20 °CÞ
zontal and vertical lines, respectively).
Table 1. G4-stabilisation (FRET) and telomerase inhibition (TRAP) results for the acridine derivatives 1a–e and the reference compound 2
Compound
FRET
Concn
TRAP
Error (nM)
DTm (°C)a
(lM)b
telEC50 (nM)
ðDTm¼20 °CÞ
1a
1b
1c
1d
1e
2c
25.4
30.2
29.2
36.2
32.9
27.5
0.728
0.544
0.599
0.465
0.493
0.720
318
267
165
98
37
1
32
24
2
80
113
1
a DTm values reported at 1 lM compound concentration.
b Concn
is equivalent to the turning point of the depicted dose–response curves (Fig. 2).
ðDTm¼20 °CÞ
c The previously published12;14 acridine 2 (BRACO-19) is included as a reference (see text).

4350
C. M. Schultes et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4347–4351
Figure 3. TRAP gels for compounds 1a–e. In all cases, lanes 1–6 correspond to telomerase-containing cell extract incubated with the following
compound concentrations: lane 1, 0 (positive control); lane 2, 0.1 lM; lane 3, 0.2 lM; lane 4, 0.5 lM; lane 5, 1.0 lM; lane 6, 0 (negative control; no
protein). The last lane (negative control) containing no protein extract is intended as verification that there are no PCR artefacts present.
400
350
300
250
200
150
100
50
40
35
30
terminal functional groups on the 9-position. Although
1c, with an extra positive charge on the piperazine ring,
might have been expected to bind best to G4–DNA, it
gave a slightly worse FRET result than the closely re-
lated 1b, the added steric presence of the N-methyl
group possibly offsetting any gain due to the additional
charge. However, in this case G4-stabilising ability and
telomerase inhibition were not correlated, with 1c being
a significantly more potent telomerase inhibitor, having
TRAP
FRET
a
telEC50 value of 165 nM compared to 267 nM for 1b.
Furthermore, when compared with the reference com-
pound 2 the FRET–TRAP correlation is not straight-
forward. This compound, which has a considerably
shorter and uncharged 9-anilino-substituent (Fig. 1),
shows a comparatively low DTm value at 1 lM, yet also
has high telomerase activity, with a very low telEC50
value (113 nM). This can be rationalised in terms of side-
chain bulk and charge. The presence of the 9-anilino
group in 2 is clearly advantageous for the mechanism of
telomerase inhibition while the lack of a charge on this
9-substituent (compared to 1c–e, for instance) would
explain the decreased ability of 2 to stabilise the G-
quadruplex (see Fig. 4 for a direct comparison of the
FRET and TRAP data obtained for compounds 1a–e
and 2). Both FRET and TRAP values are still within the
general range observed for other 3,6,9-acridines.
25
0
0
1a
1b
1c
1d
1e
2
Compound
Figure 4. Correlation between values obtained from the FRET and
TRAP experiments for compounds 1a–e and 2. Compounds increase in
telomerase inhibition as the telEC50 value (left hand y axis) decreases,
while G4-stabilising ability increases with increasing DTm value (right
hand y axis). The overall trend seen shows the compounds increasing
in their ability to stabilise G4–DNA and inhibit telomerase proces-
sivity going from 1a to 1e (left to right); the red box indicates the
reference 2, which shows high comparative telomerase activity but
relatively poor FRET stabilising ability (see text).
dominated by charge–charge contacts. It thus seems
possible to make broad conclusions about telomerase
inhibition based on preliminary FRET results, but de-
tailed predictions are not always going to be reliable.
Some of our future work will focus on extending the
FRET–TRAP correlation studies to a larger set of
compounds, in order to understand this further.
We take these results to indicate that although com-
paring the FRET and TRAP data for compounds within
a structurally related set can yield meaningful structure–
activity relationships, as has been previously shown for
a small set of nonacridine based molecules10 and is ob-
served here for compounds 1a–e, a lower correlation is
found when comparing compounds that have greater
chemical diversity but are still closely related structur-
ally. Thus whilst a broad, overall trend can be found for
the data with compounds that have major structural
variations (such as a different number of charges), there
may be significant deviations from this trend within yet
larger sets of molecules possessing for example the same
number of charges or types of terminal functional
groups, but differing in their other chemical features.
This perhaps reflects the inherent difficulty in selectively
targeting compounds to specific types of DNA structure
or sequence when the ligand–DNA interactions are
Although no significant PCR inhibition was observed at
1 lM for compounds 1a–c, the observed Taq interfer-
ence of 1d and 1e at the same concentration raises the
question of compound selectivity for G4–DNA over
dsDNA. Testing for Taq inhibition is not directly
indicative of the strength of duplex binding but this
assay nevertheless yields important information as to
whether interference of other cellular DNA maintenance
processes can be expected due to unintended interac-
tions of these compounds with dsDNA, and whether
telomeres and telomerase can be pinpointed as the sole

C. M. Schultes et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4347–4351
4351
targets in the event of a cellular reaction to longer-term
References and notes
treatment. Binding studies using surface plasmon reso-
nance (SPR) have however shown12;13 that there is a
significant difference in the affinity of related 3,6,9-tri-
substituted acridines for the two types of DNA, thus
leaving open an exploitable window for telomerase
inhibition. In the case of compounds 1d and 1e, this
window stretches over an order of magnitude in con-
centration, as the telEC50 values lie just under 100 nM
and 50% Taq inhibition is not observed until over 1 lM.
Furthermore, the generally low acute cytotoxicity of the
9-anilino-3,6-bis-amidoacridines supports this pre-
mise,14 and future work will attempt to address these
issues in more detail.
1. Neidle, S.; Parkinson, G. Nat. Rev. Drug Discov. 2002, 1,
383.
2. Rezler, E. M.; Bearss, D. J.; Hurley, L. H. Curr. Opin.
Pharmacol. 2002, 2, 415.
3. Mergny, J. L.; Riou, J. F.; Mailliet, P.; Teulade-Fichou,
M. P.; Gilson, E. Nucleic Acids Res. 2002, 30, 839.
4. Wang, Y.; Patel, D. J. Structure 1993, 1, 263; Phan, A. T.;
Patel, D. J. J. Am. Chem. Soc. 2003, 125, 15021; Ying, L.;
Green, J. J.; Li, H.; Klenerman, D.; Balasubramanian, S.
Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 14629.
5. Parkinson, G. N.; Lee, M. P.; Neidle, S. Nature 2002, 417,
876.
6. Neidle, S.; Parkinson, G. N. Curr. Opin. Struct. Biol. 2003,
13, 275.
7. Cuesta, J.; Read, M. A.; Neidle, S. Mini. Rev. Med. Chem.
2003, 3, 11.
8. Mergny, J. L.; Maurizot, J. C. Chembiochemistry 2001, 2,
124.
9. 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.;
4. Conclusion
Within this series of 3,6,9-trisubstituted acridines, a
direct correlation has been observed between G4-sta-
bilisation, as measured in the FRET assay, and
telomerase inhibition (Fig. 4). This supports the use of
the FRET assay as a screening tool for the rapid iden-
tification of novel G4-stabilising ligands so that prom-
ising candidates can be chosen for more detailed
biological, biophysical and structural studies in order to
further elucidate their interactions with G4–DNA.
ꢀ ꢁ
Helene, C. Proc. Natl. Acad. Sci. U.S.A. 2002, 98, 3062.
10. Guyen, B.; Schultes, C. M.; Hazel, P.; Mann, J.; Neidle, S.
Org. Biomol. Chem. 2004, 2, 981.
11. Burger, A. M. In Methods in Molecular Biology; Double,
J. A., Thompson, M. J., Eds.; Humana: Totowa, NJ, 2002;
Vol. 191, pp 109–124.
12. Read, M.; Harrison, R. J.; Romagnoli, B.; Tanious, F. A.;
Gowan, S. H.; Reszka, A. P.; Wilson, W. D.; Kelland, L.
R.; Neidle, S. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 4844.
13. 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. J. Med. Chem. 2003, 46, 4463.
14. Gowan, S. M.; Harrison, J. R.; Patterson, L.; Valenti, M.;
Read, M. A.; Neidle, S.; Kelland, L. R. Mol. Pharmacol.
2002, 61, 1154.
Acknowledgements
This work has been supported by project and pro-
gramme grants to S.N. from Cancer Research UK, by a
project grant from The International Association for
Cancer Research and by a Cancer Research UK Stu-
dentship to C.M.S.
15. Hill, J. W. J. Am. Chem. Soc. 1930, 52, 4110.
~ꢀ
16. Avino, A.; Garcia, R. G.; Albericio, F.; Mann, M.; Wilm,
M.; Neubauer, G.; Eritja, R. Bioorg. Med. Chem. 1996, 4,
1649.