An asymmetric C8/C8′-tripyrrole-linked sequence-selective pyrrolo[2,1-c][1,4]benzodiazepine (PBD) dimer DNA interstrand cross-linking agent spanning 11 DNA base pairs

Article abstract of DOI:10.1016/j.bmcl.2008.01.096

A novel sequence-selective extended PBD dimer 4 has been synthesized that binds with high affinity to an interstrand cross-linking site spanning 11 DNA base pairs. Despite its molecular weight (984.07) and length, the molecule has significant DNA interstrand cross-linking potency (～100-fold greater than the clinically used agent melphalan) and sub-micromolar cytotoxicity in a number of tumour cell lines, suggesting that it readily penetrates cellular and nuclear membranes to reach its DNA target.

Full text of DOI:10.1016/j.bmcl.2008.01.096

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 2073–2077
An asymmetric C8/C80-tripyrrole-linked sequence-selective
pyrrolo[2,1-c][1,4]benzodiazepine (PBD) dimer DNA
interstrand cross-linking agent spanning 11 DNA base pairs
Arnaud C. Tiberghien,a David A. Evans,a Konstantinos Kiakos,b
Christopher R. H. Martin,a John A. Hartley,a,b
a,c
David E. Thurstona,c, and Philip W. Howard
*
aSpirogen Ltd, 29/39 Brunswick Square, London WC1N 1AX, UK
bCancer Research UK Drug–DNA Interactions Research Group, UCL Cancer Institute, Paul O’Gorman Building,
Huntley Street, London WC1E 6BT, UK
cCR-UK Gene Targeting Drug Design Research Group, School of Pharmacy, University of London, 29/39 Brunswick Square,
London WC1N 1AX, UK
Received 3 December 2007; revised 23 January 2008; accepted 24 January 2008
Available online 30 January 2008
Abstract—A novel sequence-selective extended PBD dimer 4 has been synthesized that binds with high affinity to an interstrand
cross-linking site spanning 11 DNA base pairs. Despite its molecular weight (984.07) and length, the molecule has significant
DNA interstrand cross-linking potency (ꢀ100-fold greater than the clinically used agent melphalan) and sub-micromolar cytotox-
icity in a number of tumour cell lines, suggesting that it readily penetrates cellular and nuclear membranes to reach its DNA target.
Ó 2008 Elsevier Ltd. All rights reserved.
There is growing interest in small molecules that recog-
nize DNA sequence,1 and examples are known that
interact with DNA non-covalently (e.g., the hairpin
polyamides2) or covalently through monoalkylation
(e.g., the pyrrolobenzodiazepines3) or cross-linking
(e.g., bizelesin4) mechanisms. SJG-136 (1, Fig. 1) is a po-
tent, sequence-selective interstrand DNA cross-linking
agent containing pyrrolobenzodiazepine units that is
presently in Phase I evaluation in the clinic.5–7 The pro-
pensity of SJG-136 to preferentially target 50-Pu-
GATC-Py DNA sequences is thought to contribute to
its antitumour activity.8 In 2001, Bando and co-workers
reported the structure of an efficient CPI-based cross-
linking agent (2), the polypyrrole-imidazole core of which
allows sequence-specific cross-linking.9 More recently,
syntheses of the first examples of heterocyclic-linked
PBD dimers have been independently reported; Kumar
and Lown have described molecules of type 3, although
DNA cross-linking and sequence-selectivity data were
not reported,10 and similar extended PBD dimers have
been reported in the patent literature by our group.11
In order to systematically study the effect of length and
composition of the polyheterocyclic linker on DNA se-
quence-selectivity, binding affinity and cross-linking
efficiency of C8/C80-linked PBD dimers of type 3, we
have synthesized different families of dimers containing
a variety of heterocyclic linkers and with different regio-
chemical arrangements of constituent components.
These heterocyclic units are known from the work of
Dervan and co-workers to recognize sequences in the
minor groove.2
We report here the synthesis and evaluation of the ex-
tended asymmetrically-linked tripyrrole-containing
PBD dimer 4 (AT-235) which spans 11 DNA base pairs
with a sequence-selectivity that can be rationalized
based on both its covalent and non-covalent interactions
with DNA base pairs in the minor groove.
Keywords: DNA-binding; Sequence-selective; Interstrand; Cross-link-
ing; Anticancer agent; Pyrrolobenzodiazepine; PBD; PBD dimer.
The synthesis of 4 was achieved through sequential
amide coupling between the three main components 5,
6 (to give 7) and 8 as shown in Scheme 1.
*
Corresponding author. Tel.: +44 2077535932; fax: +44
2077535964; e-mail: david.thurston@pharmacy.ac.uk
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.01.096

2074
A. C. Tiberghien et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2073–2077
H
N
N
N
O
O
O
O
N
CH2
H
H
H
N
O
N
H3CO2C
N
N
N
O
N
O
O
O
N
1 (SJG-136)
H
O
2
2
O
X= CH or N
H
N
N
O
O
H
O
O
X
H
H
N
N
O
N
H
N
N
N
H
3
O
O
N
OCH3
n (n = 1 to 3)
O
3
O
N
O
H
NH
N
H
OMe
N
N
O
O
H
N
N
O
H
N
N
O
H
N
O
N
MeO
4 (AT-235)
O
Figure 1. Structures of the DNA interstrand cross-linking agents 1 (SJG-136), the CPI-based dimer 2, the previously reported PBD dimer 3 and the
new asymmetric PBD dimer 4 (AT-235).
BocHN
Alloc
N
H
N
OH
H
a,b
H2N
O
N
O
H
N
N
MeO
N
O
O
OH
N
O
5
6
H2N
Alloc
N
O
OMe
H
H
O
N
HO
N
O
H
N
MeO
N
N
Alloc
N
O
O
H
OH
H
N
O
N
8
O
N
MeO
O
c
7
Alloc
N
O
MeO
H
O
NH
H
N
N
OMe
N
O
O
H
d
N
4
N
Alloc
N
O
H
OH
H
N
O
N
9
O
N
MeO
O
Scheme 1. Reagents and condition: (a) EDCI, HOBt, DMF, 71%; (b) 4 N HCl in dioxane, quant; (c) oxalyl chloride, THF, DMF cat, then 7 and
DIPEA, 44%; (d) Pd(PPh3)4, pyrrolidine, CHCl3, 85%.

A. C. Tiberghien et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2073–2077
2075
The known Boc-tripyrrole acid 5 was prepared by an
improvement of a method of Boger and co-workers12
that required no complex work-up or purification proce-
dures (Scheme 2).
The novel chirally pure C8-aminopropyl-N10-alloc-pro-
tected PBD 6 was synthesized in 7 steps from the previ-
ously
reported
4-hydroxy-5-methoxy-2-nitro-
Figure 2. Cross-linking gel18 for extended PBD dimer 4 (AT-235)
using linearized pUC18 DNA. Controls: DS, double stranded DNA;
SS, single stranded DNA. The XL50 was determined to be 0.23 lM.
benzaldehyde starting material (14)13 (Scheme 3). The
aliphatic Boc amino side chain was introduced by Mits-
unobu phenol etherification and the aldehyde oxidized
with potassium permanganate to yield the nitro acid
15. Coupling to S-pyrrolidinemethanol gave the nitro
alcohol 16 which was hydrogenated followed by alloc
protection of the resulting amine to give alcohol 17.
Exposure to TEMPO/BAIB14 led to smooth oxidative
ring closure to give 18. The best conditions for the final
Boc deprotection were found to be TFA/DCM/water 47/
47/6 which afforded the capping unit 6 in an overall yield
of 17% over 7 steps. This intermediate required only one
chromatographic purification and was chirally pure.
Further protection of the C11-hydroxy functionality be-
fore the next coupling step was found to be unnecessary
due to the superior nucleophilicity of the amino func-
tionality. The C11-methyl ether (8) of the PBD acid cap-
ping unit described by Wells and co-workers15 was
employed in order to improve coupling efficiency and
avoid side-reactions. Finally, cascade removal of all pro-
tective groups of 9 under Deziel conditions16 yielded the
extended PBD dimer 417 in 85% yield.
Table 1. Average cytotoxicity values from the NCI 60-cell panel and
results from a K562 line, and DNA cross-linking data for 1 (SJG-136)
and extended dimer 4
b
a
Compound
NCI (lM)
TGIc
K562 GI50
(lM)
XL50
(lM)
b
d
GI50
LC50
0.562
22.9
1
4
0.007
0.014
0.087
0.49
0.008
0.037
0.060
0.23
a XL50: Dose providing 50% cross-linking of DS pUC18 plasmid
DNA.18
b GI50: Dose inhibiting 50% cell growth.
c TGI: Dose inhibiting 100% cell growth.
d LC50: Dose killing 50% of cells.
11 DNA base pairs with a degree of selectivity within
this span, the probability of the perfectly matched site(s)
appearing within the pUC18 sequence is lower than that
for SJG-136 which spans 6 bp, and this could be one
possible explanation for the lower observed potency of
4 in this assay compared to 1.
PBD dimer 4 was shown to effectively interstrand cross-
link linear plasmid pUC18 DNA with an XL50 value of
0.23 lM (Fig. 2 and Table 1),18 a potency 4-fold less
than SJG-136 (0.06 lM) but approximately 100-fold
greater than the clinically used nitrogen mustard mel-
phalan (20.0 lM). As the molecule is designed to span
PBD dimer 4 was also shown to have significant cyto-
toxicity in both the NCI 60-cell line panel and in the
K562 leukaemia cell line (Table 1).
BocHN
O2N
O2N
H
H
N
a, b
N
c, d, e
N
N
OMe
O
O
H
N
N
OMe
O
O2N
N
N
BocHN
O
O
10
OH
12
CCl3
N
5
OBt
N
13
O
N
11
O
O
Scheme 2. Reagents and conditions: (a) 10% Pd/C, H2, DMF, quant; (b) 11, DMF, 50%; (c) 10% Pd/C, H2, DMF, quant; (d) 13, DMF, 94%; (e)
NaOH, MeOH, water, quant.
OH
HO
NO2
H
c
BocHN
O
NO2
OH
BocHN
O
NO2
a, b
d, e
N
MeO
MeO
MeO
O
O
O
14
15
16
Alloc
NH
Alloc
N
Alloc
N
OH
OH
H
OH
H
BocHN
O
f
BocHN
O
g
H2N
O
N
MeO
N
MeO
N
MeO
6
18
O
17
O
O
Scheme 3. Reagents and conditions: (a) PPh3, DEAD, (3-hydroxy-propyl)carbamic acid tert-butyl ester, THF; (b) KMnO4, water, acetone, 32% over
2 steps; (c) S-pyrrolidinemethanol, EDCI, HOBt, DMF, 93%; (d) 10% Pd/C, H2, EtOAc, quant; (e) allyl chloroformate, pyridine, DCM, 86%; (f)
TEMPO, BAIB, DCM, 72%; (g) TFA, DCM, water (47/47/6), quant.

2076
A. C. Tiberghien et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2073–2077
Table 2. Energy calculations using duplex DNA sequences A–F to
predict preferences for interstrand (A–C) versus intrastrand (D–F)
cross-linking and the optimal number of base pairs (6–8) separating
covalently modified guanines
The sub-micromolar GI50 and TGI cytotoxicity values
suggest that, despite its molecular weight (984.07) and
length (spanning 11 bp), 4 is still capable of traversing
cellular and nuclear membranes to reach the genome
of these cells.
Simulated binding sequence
(50 ! 30)
DDEbinda/kcal molꢁ1
DNA footprinting8 of 4 on a 543 bp fragment of BCL-2
DNA19 showed that it binds with high affinity (i.e.,
10 nM) to a 50-GCTTATAATGG-30 sequence (Fig. 3).
This may be rationalized through simulation studies
(Table 2) involving 4 bound to a range of potential bind-
ing sites which predict a preference for interstrand rather
than intrastrand cross-linking with the covalently mod-
ified guanines on opposite strands separated by 7 bp
(i.e., sequence B). The predicted binding energy at the
main footprint site (DDEbind = 42.6. kcal molꢁ1) was
higher than those values calculated for the hypothetical
cross-linking motifs shown in Table 2. However, as se-
quences A, B and C are not present in the footprinted
BCL-2 fragment, it is possible that higher-affinity bind-
ing sites may exist in longer DNA sequences. Figure 4
shows a molecular model in which 4 is bound to the
footprinted 50-GCTTATAATGG-30 sequence. The two
PBD moieties are bound to the 30-CGA (opposite 50-
GCT) and 50-TGG triplet sequences (covalent binding
A
B
C
D
E
F
CGCAGAAATTTCTGCG
CGCAGAAAATTTCTGCG
CGCAGAAAATTTTCTGCG
CGCAGAAATTTGTGCG
CGCAGAAAATTTGTGCG
CGCAGAAAATTTTGTGCG
13.6 0.6
0
7.4 0.6
10.1 0.6
17.6 0.6
23.5 0.6
The results predict that sequence B (50-AGAAAATTTCT-30) has the
lowest binding energy, suggesting that 4 is more likely to form an
interstrand rather than an intrastrand cross-link with a spacing of 7 bp
between the covalently-linked guanine residues.
a DDEbind is the MM-PBSA22,23 binding energy relative to the lowest
energy binding sequence which was
B (thus appearing as
0 kcal molꢁ1). The error estimate is the standard error in the mean
over all included trajectory snapshots.
Figure 4. Representative snapshot from a molecular dynamics simu-
lation (Sander program in Amber820) showing 4 bound to a 50-
GCTTATAATGG-30 sequence with the two PBD units covalently
bound to guanines on opposite strands (positions underlined).
Figure 3. DNA footprinting gel of 4 bound to a 543 bp fragment of
BCL-2 DNA19 showing a well-defined footprint at a 50-AGCTTA
TAATGG-30 site at a concentration of 10 nM or above after 18 h
incubation.
base pair positions underlined) with the heterocyclic lin-
ker interacting with the central 50-TATAA base pairs as
predicted by the literature.2,15 According to this model,

A. C. Tiberghien et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2073–2077
2077
M. G.; Schweikart, K. M.; Tomaszewski, J. E.; Sausville,
E. A.; Gregson, S. J.; Howard, P. W.; Thurston, D. E.
Cancer Res. 2004, 64, 6693.
the dimer takes up the curvature of the DNA minor
groove with little distortion of the helix.
8. Martin, C.; Ellis, T.; McGurk, C. J.; Jenkins, T. C.;
Hartley, J. A.; Waring, M. J.; Thurston, D. E. Biochem-
istry 2005, 44, 4135.
9. Bando, T.; Iida, H.; Saito, I.; Sugiyama, H. J. Am. Chem.
Soc. 2001, 123, 5158.
10. Kumar, R.; Lown, J. W. Eur. J. Med. Chem. 2005, 40, 641.
11. Howard, P. W.; Gregson, S. J.; Tiberghien, A.C. WO
2005/085250 A1 Sep 15th, 2005.
12. Boger, D. L.; Fink, B. E.; Hedrick, M. P. J. Am. Chem.
Soc. 2000, 122, 6382.
In Figure 3 an additional weaker footprint was observed
at a 50-TCACTATCTCCCGGTTA-30 sequence at a
concentration of approximately 1 lM. This sequence
does not contain a predicted cross-linking site for 4,
but does contain potential sites of monoalkylation for
one PBD unit such as 30-GGG, a preferred monomeric
PBD binding sequence.21 This could explain the lower
preference (by 2 orders of magnitude) of 4 for this site.
13. Fukuyama, T.; Lin, S. C.; Li, L. P. J. Am. Chem. Soc.
1990, 112, 7050.
14. DeMico, A.; Margarita, R.; Parlanti, L.; Vescovi, A.;
Piancatelli, G. J. Org. Chem. 1997, 62, 6974.
15. Wells, G.; Martin, C. R. H.; Howard, P. W.; Sands, Z. A.;
Laughton, C. A.; Tiberghien, A.; Woo, C. K.; Masterson,
L. A.; Stephenson, M. J.; Hartley, J. A.; Jenkins, T. C.;
Shnyder, S. D.; Loadman, P. M.; Waring, M. J.; Thur-
ston, D. E. J. Med. Chem. 2006, 49, 5442.
Through a greater understanding of SAR for extended
PBD dimers of this type, we hope to provide the basis
of a novel approach to DNA sequence recognition
through an interstrand cross-linking mechanism. This
is being further explored by the synthesis and evaluation
of analogues of 4 with differently-structured linkers
which will be reported in due course.
16. Deziel, R. Tetrahedron Lett. 1987, 28, 4371.
17. Data for 4. 1H NMR (400 MHz, DMSO-d6): d 9.90 (s,
3H), 8.10 (m, 1H), 7.79 (d, 2H, J = 4.36 Hz), 7.35 (s, 2H),
7.25 (s, 1H), 7.20 (2s, 2H), 7.05 (s, 1H), 6.90 (s, 2H), 6.84
(s, 2H), 4.21–4.01 (m, 4H), 3.85–3.81 (m, 15H), 3.61 (m,
2H), 3.46–3.35 (m, 6H), 2.45 (m, 2H), 2.38–2.17 (m, 4H),
2.13–1.89 (m, 8H); 13C NMR (100 MHz, DMSO-d6): d
168.7, 164.2, 158.3, 150.2, 150.1, 146.8, 140.5, 122.7, 122.0,
121.9, 119.7, 118.4, 118.4, 117.8, 114.8, 111.2, 110.0, 104.6,
104.2, 103.9, 67.7, 66.4, 55.8, 55.5, 53.6, 53.3, 48.5, 46.3,
36.0, 35.8, 35.5, 30.2, 28.9, 28.9, 28.7, 24.7, 23.6, 22.3; IR
(cmꢁ1): m 3306, 2949, 1638, 1597, 1555, 1508, 1465, 1435,
1405, 1262, 1216, 1090, 1064. MS (ES+) m/z (relative
intensity): 984.4 ([M+H]+Å, 100); HRMS (TOF MS ES+):
Calcd for C51H57N11O10 ([M+H]+): 984.4363. Found:
984.4359.
Acknowledgments
Marissa Coffils and Dr. Robert Schultz are thanked for
providing the K562 and NCI 60-cell line data, respec-
tively, and Laura Calderon is acknowledged for initially
synthesizing the PBD intermediate 6.
References and notes
1. Thurston, D. E. Chemistry and Pharmacology of Antican-
cer Drugs; CRC Press (Taylor & Francis): Boca Raton,
Florida, USA, 2006.
2. Kielkopf, C. L.; White, S.; Szewczyk, J. W.; Turner, J. M.;
Baird, E. E.; Dervan, P. B.; Rees, D. C. Science 1998, 282,
111.
18. Hartley, J. A.; Berardini, M. D.; Souhami, R. L. Anal.
Biochem. 1991, 193, 131.
19. Accession No. NC_000018.8 (Chromosome 18; DNA
Bases 59136813–59136823).
3. Thurston, D. E. Br. J. Cancer 1999, 80, 65.
4. Schwartz, G. H.; Patnaik, A.; Hammond, L. A.; Rizzo, J.;
Berg, K.; Von Hoff, D. D.; Rowinsky, E. K. Ann. Oncol.
2003, 14, 775.
20. Amber
Molecular
amber.scripps.edu.
Dynamics
Package:
http://
21. Thurston, D. E.. In Molecular Aspects of Anticancer
Drug–DNA Interactions; Neidle, S., Waring, M. J., Eds.;
The Macmillan Press Ltd.: London, UK, 1993; Vol. 1, p
54.
5. Alley, M. C.; Hollingshead, M. G.; Pacula-Cox, C. M.;
Waud, W. R.; Hartley, J. A.; Howard, P. W.; Gregson, S.
J.; Thurston, D. E.; Sausville, E. A. Cancer Res. 2004, 64,
6700.
22. Kollman, P. A.; Massova, I.; Reyes, C.; Kuhn, B.; Huo, S.
H.; Chong, L.; Lee, M.; Lee, T.; Duan, Y.; Wang, W.;
Donini, O.; Cieplak, P.; Srinivasan, J.; Case, D. A.;
Cheatham, T. E. Acc. Chem. Res. 2000, 33, 889.
23. Srinivasan, J.; Cheatham, T. E.; Cieplak, P.; Kollman, P.
A.; Case, D. A. J. Am. Chem. Soc. 1998, 120, 9401.
6. Gregson, S. J.; Howard, P. W.; Hartley, J. A.; Brooks, N.
A.; Adams, L. J.; Jenkins, T. C.; Kelland, L. R.; Thurston,
D. E. J. Med. Chem. 2001, 44, 737.
7. Hartley, J. A.; Spanswick, V. J.; Brooks, N.; Clingen, P.
H.; McHugh, P. J.; Hochhauser, D.; Pedley, R. B.;
Kelland, L. R.; Alley, M. C.; Schultz, R.; Hollingshead,