A new class of symmetric bisbenzimidazole-based DNA minor groove-binding agents showing antitumor activity

Article abstract of DOI:10.1021/jm000297b

The synthesis and evaluation of the novel head-to-head bisbenzimidazole compound 2,2-bis-[4′-(3″-dimethylamino-1″-propyloxy)phenyl]-5,5-bi-1 H-benzimidazole is described. An X-ray crystallographic study of a complex with the DNA dodecanucleotide sequence d(CGCGAAT-TCGCG) shows the compound bound in the A/T minor groove region of a B-DNA duplex and that the head-to-head bisbenzimidazole motif hydrogen-bonds to the edges of all four consecutive A:T base pairs. The compound showed potent growth inhibition with a mean IC₅₀ across an ovarian carcinoma cell line panel of 0.31 μM, with no significant cross-resistance in two acquired cisplatin-resistant cell lines and a low level of cross-resistance in the P-glycoprotein overexpressing acquired doxorubicin-resistant cell line. Studies with the hollow fiber assay and in vivo tumor xenografts showed some evidence of antitumor activity.

Full text of DOI:10.1021/jm000297b

138
J . Med. Chem. 2001, 44, 138-144
Articles
A New Cla ss of Sym m etr ic Bisben zim id a zole-Ba sed DNA Min or Gr oove-Bin d in g
Agen ts Sh ow in g An titu m or Activity
J ohn Mann,†,‡ Anne Baron,†,‡ Yaw Opoku-Boahen,† Eric J ohansson,| Gary Parkinson,| Lloyd R. Kelland,§ and
Stephen Neidle*,|
Chemistry Department, Reading University, Whiteknights, Reading RG6 6AD, U.K., CRC Centre for Cancer Therapeutics,
Institute of Cancer Research, Sutton, Surrey SM2 5NG, U.K., and CRC Biomolecular Structure Unit,
Chester Beatty Laboratories, Institute of Cancer Research, 237 Fulham Road, London SW3 6J B, U.K.
Received J uly 17, 2000
The synthesis and evaluation of the novel head-to-head bisbenzimidazole compound 2,2-bis-
[4′-(3′′-dimethylamino-1′′-propyloxy)phenyl]-5,5-bi-1H-benzimidazole is described. An X-ray
crystallographic study of a complex with the DNA dodecanucleotide sequence d(CGCGAAT-
TCGCG) shows the compound bound in the A/T minor groove region of a B-DNA duplex and
that the head-to-head bisbenzimidazole motif hydrogen-bonds to the edges of all four consecutive
A:T base pairs. The compound showed potent growth inhibition with a mean IC50 across an
ovarian carcinoma cell line panel of 0.31 µM, with no significant cross-resistance in two acquired
cisplatin-resistant cell lines and a low level of cross-resistance in the P-glycoprotein overex-
pressing acquired doxorubicin-resistant cell line. Studies with the hollow fiber assay and in
vivo tumor xenografts showed some evidence of antitumor activity.
In tr od u ction
Studies on the noncovalent interactions of small mole-
cules with the minor groove of DNA continue to be a
fruitful area for the discovery of potential new thera-
peutic agents. Those compounds based on bis(phenyl-
amidinium)furans can have activity1-4 against Pneu-
mocystis carinii, a pathogen of major clinical importance
in the treatment of AIDS patients, as well as cytotoxic
activity against tumor cell lines.5 Others have demon-
strated experimental antitumor activity, notably the
synthetic compound6 PNU-166196 (an R-bromoacrylic
derivative of distamycin) and various asymmetric bis-
benzimidazole compounds.7 The parent of this latter
family is the cell stain Hoechst 33258, also known as
Pibenzimol (Figure 1), which has two benzimidazole
groups linked in a head-to-tail manner.7 This compound
was initially found to have activity against L1210
murine leukemia,8 which was followed by several phase
I trials in humans.9 The first trial was limited by
pancreatic toxicity, although in a second trial some
responses were seen in pancreatic cancer; a subsequent
phase II trial did not show any objective responses,10
and no further trials of this agent have been reported.
Hoechst 33258 is weakly mutagenic in Salmonella test
strains.11,12 It inhibits the binding of TATA-box binding
protein to DNA13 and is an effective inhibitor of mam-
malian DNA topoisomerase I.14
F igu r e 1. (a) Structure of Hoechst 33258. (b) Synthetic
scheme used in the synthesis of the lead compound 5.
Hoechst 33258 recognizes A/T sequences.15 The mode
of binding to DNA has been established from X-ray
crystallographic studies of complexes with several A/T-
containing oligonucleotides,16,17 as well as NMR studies
in solution.18,19 All of these studies concur in finding that
the two core benzimidazole groups of Hoechst 33258
together recognize three consecutive A:T base pairs, via
a pair of bifurcated hydrogen bonds, with van der Waals
* To whom correspondence should be addressed. Tel: 0044 020 7970
6043. Fax: 0044 020 7 352 8039. E-mail at s.neidle@icr.ac.uk.
† Reading University.
§ CRC Centre for Cancer Therapeutics.
| CRC Biomolecular Structure Unit.
‡ Present address: School of Chemistry, Queen’s University Belfast,
Belfast BT9 5AG, U.K.
10.1021/jm000297b CCC: $20.00 © 2001 American Chemical Society
Published on Web 12/07/2000

Bisbenzimidazole DNA Minor Groove-Binding Agents
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 2 139
forces playing a major role in the overall binding.20
A
number of other head-to-tail linked benzimidazoles have
been studied as probes of minor groove recognition, and
all show the same pattern of base recognition,21-23
including compounds with three benzimidazole units.24-27
We hypothesized that the size of the binding site could
be extended by a symmetric head-to-head arrangement
for the two benzimidazole groups. Several such mol-
ecules were subsequently synthesized and shown by
DNA footprinting and X-ray crystallographic methods
to bind to A/T DNA sequences in this manner.28 It was
also suggested that replacement of the piperazine group
in Hoechst 33258 by pure DNA-binding groups would
enhance cytotoxicity and possibly lead to compounds
with antitumor activity yet lacking the toxicity of
Hoechst 33258. We report here a detailed biological
evaluation of the lead compound 2,2-bis[4′-(3′′-dimethyl-
amino-1′′-propyloxy)phenyl]-5,5-bi-1H-benzimidazole (5;
Figure 1b) in the series, together with synthetic and
structural studies.
Resu lts
Ch em istr y. The required bisbenzimidazoles were
synthesized via the route shown in Figure 1b. This
involved condensation between 3,3′,4,4′-tetraaminobi-
phenyl (prepared by Raney nickel reduction of 3,3′-
dinitro-4,4′-diaminobiphenyl (1)28) and the requisite
aromatic aldehyde by refluxing the reactants in ni-
trobenzene under argon at 150 °C for 8-12 h. Thus
4-methoxybenzaldehyde yielded bisbenzimidazole (2)
(29%); 4-hydroxybenzaldehyde provided compound 3
(23%); and 4-(3′-dimethylamino-1′-propyloxy)benzalde-
hyde (4) yielded compound 5 (35%). This last aldehyde
was itself synthesized via a Mitsunobu reaction between
4-hydroxybenzaldehyde and 3-dimethylaminopropan-1-
ol (Ph3P, DEAD, THF, 85%).
Cr yst a l St r u ct u r e Descr ip t ion . The crystallo-
graphic analysis confirms that the structure of the
complex of compound 5 with the sequence d(CGC-
GAATTCGCG) is isomorphous with other dodecamers
of the same sequence, with or without minor groove-
bound drug. The mean helical twist between successive
base pairs is 35.6°, and there are 10.1 base pairs/helical
turn. The mean rise per base pair is 3.34 Å. Figure 2a
shows a temperature factor representation of the com-
plex. The core central bisbenzimidazole region of the
drug is well-defined and relatively immobile. However,
the temperature factors for the drug side chains are
high. This is as expected for long aliphatic chains of this
nature, and indeed, the conformation shown is probably
only one of several possible ones which could have been
fitted to the rather diffuse electron density.
A close-up view of the central benzimidazole region
(Figure 2b) illustrates some of the factors that contrib-
ute to the binding of the drug. First, the molecule is
overall crescent-shaped, thus complementing the helic-
ity of the minor groove. The drug molecule further
adapts to the groove curvature by the twisting of its two
subunits, with a dihedral angle of 20° between them.
Second, the positions of the hydrogen-bond donors on
the drug complement that of the acceptor spacing of the
DNA. The two benzimidazole groups each hydrogen-
bond to two A:T base pairs, in accordance with the
initial modeling suggestion, and thus cover all four A:T
F igu r e 2. (a) View (drawn with the GRASP program40) of the
structure of the complex between d(CGCGAATTCGCG) and
compound 5, colored according to temperature factor. Note the
high temperature factors of the alkylamino termini of com-
pound 5, indicating significant mobility. (b) Surface represen-
tation of the minor groove region of the complex, showing the
binding of compound 5. Hydrogen bonds to base edges are
indicated by dashed lines.
base pairs in this sequence (Figure 2a). The overall
binding site size is 6 base pairs.
The minor groove widths were calculated with the
CURVES program29 and are shown in Figure 3, com-
pared to those of complexes with other related ligands
as well as to that of the native dodecamer. This shows
that the overall shape of the minor groove in these
structures is similar but that the actual widths vary.

140 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 2
Mann et al.
F igu r e 3. Plot of C4′-C4′ minor groove widths29 in the complex of d(CGCGAATTCGCG) with compound 5, compared with those
in several other complexes. The Nucleic Acid Database41 codes are given for the structures.
F igu r e 5. In vivo antitumor activity of compound 5 (4 mg/
kg, ip days 0, 4, 8) against CH1 human ovarian carcinoma
xenografts: O, control; 2, treated group. Data points are mean
F igu r e 4. Growth inhibition by compound 5 across a panel
of human ovarian carcinoma cell lines using 96-h continuous
drug exposure. Values are mean ( SD from 3 experiments.
( SD.
As can be seen, the minor groove width of the present
complex is very similar to that of the native dodecamer,
particularly in the narrow AATT region, where the
widths are around 3.3 Å. There is a tendency for a
greater groove width in other drug-dodecamer struc-
tures in the 5′-AATT region, which may be due to the
differences in base-drug hydrogen bonding.
compound 5 showed potent growth inhibition with a
mean GI50 across the 60 cell lines of 0.21 µM. Pearson
coefficient COMPARE analysis indicated that compound
5 does not share a response pattern with any other
agent in the databank (the highest GI50 coefficients
were 0.661 for MDR Rhod30 and 0.578 for echinomycin).
We are grateful to the NCI for communicating these
Eva lu a tion Aga in st a Hu m a n Ova r ia n Ca r ci-
n om a Cell Lin e P a n el. The growth inhibitory proper-
ties of compound 5 were initially determined using a
small panel of human ovarian carcinoma cell lines,
including sublines possessing acquired resistance to
cisplatin (cisR lines) and doxorubicin (doxR line). Re-
sults using 4-day continuous drug exposure are shown
in Figure 4. Compound 5 showed potent growth inhibi-
tion with a mean IC50 across the panel of 0.31 µM.
Moreover, there was no marked cross-resistance in two
acquired cisplatin-resistant cell lines and only low level
cross-resistance in the P-glycoprotein overexpressing
acquired doxorubicin-resistant cell line (RF of 5.8
compared to 90 for doxorubicin itself).
Compound 5 has also been evaluated by the National
Cancer Institute (NCI) using 60 cell lines which are
derived from a diverse group of human cancers.30 In
contrast to the methods outlined above, the NCI used
a 48-h drug exposure, albeit followed by the same
SRB endpoint. In common with the above results,
data.
The effect of the time of drug exposure was investi-
gated using the SKOV-3 cell line. There was little
dependence on the time of drug exposure to growth
inhibition (IC50); values (in µM) were 2 h, 0.33; 6 h,
0.105; 24 h, 0.115; and 96 h, 0.1.
In Vivo An titu m or Activity. Initial in vivo antitu-
mor evaluation of compound 5 used the hollow fiber
assay with ip-implanted fibers containing CH1 cells and
ip-administered drug.31 Evidence of antitumor activity
was seen following a single dose of 4 mg/kg (T/C of 0.79).
Additional hollow fiber studies were also performed by
the NCI; a maximum effect of T/C of 0.56 was observed
against the OVCAR-3 human ovarian carcinoma cell
line (in ip-implanted fibers) using a dose of 4.5 mg/kg
given ip daily for 4 consecutive days (Table 1b; personal
communication, NCI).
Following the observation of some antitumor effect
in mice bearing CH1-containing hollow fibers, compound
5 was evaluated against the corresponding sc xenograft

Bisbenzimidazole DNA Minor Groove-Binding Agents
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 2 141
Ta ble 1.
The crystallographic study demonstrates that com-
pound 5 can bind effectively to the A/T-rich minor groove
of B-form DNA, in accord with our earlier observations
on the dihydroxy derivative 3.28 Both compounds are
effective inhibitors of transcription,28 with 5 being an
order of magnitude more potent. This parallels the
higher affinity for DNA shown by 5 compared to 3.28
Although we do not as yet have direct evidence that the
in vitro (and possibly the in vivo) properties of 5 are
due to its effect on transcription, we suggest that this
is a plausible mechanism. However we cannot rule out
at this stage alternative DNA-related modes of action
such as inhibition of DNA topoisomerases or helicases.
Many (though not all) minor groove-binding molecules
possessing significant cytotoxicity and antitumor prop-
erties also covalent bond in the minor groove (especially
to N3 of adenine or N2 of guanine). Compounds such
as the alkylating distamycin derivative tallimustine33
have antitumor activity whereas the parent distamycin
does not. However the discovery of in vivo activity for
the noncovalent R-bromoacryloyl derivative of distamy-
cin (PNU 151807)6 has demonstrated that covalent
binding to DNA is not always a requirement in order
for groove-binding molecules to have antitumor activity,
a conclusion reinforced by the present results on com-
pound 5. In the head-to-head bisbenzimidazole category,
a number of nonalkylating cytotoxic analogues have
been reported, some of which may target human tumor
helicases.34
a. Cytotoxicity of Bisbenzimidazole Compounds for
96-h Exposure in a Panel of Ovarian Cell Lines
IC50 (µM)
compd
NSC no. A2780 A2780cisR CH1 SKOV-3
2
3
5
D701062 13.5
D699126 4.3
D699127 0.235
5.1
2.7
3.8
16.5
16.0
0.375
1.05
0.24
26.5 >100
0.115
9.5
Hoechst 33258 322921
12.0
b. Activity in sc Human Tumor Xenograft (hollow fiber) Assays,
as Determined by the NCI31
compd
NSC no.
ip score
sc score
total score
2
3
5
D701062
D699126
D699127
8
10
12
0
0
4
8
10
16
using a maximum tolerated dose of 4 mg/kg (days 0, 4,
8 ip administration). Tumor growth curves for treated
versus control groups are shown in Figure 5. Compound
5 exhibited some evidence of in vivo antitumor activity
against this established sc-implanted xenograft; the
maximum T/C% was 0.56 on day 10 (this difference just
attained statistical significance P of 0.047, two-tailed
Students t-test). There was no evidence of drug-induced
toxicity in this study.
Discu ssion
Compound 5 possesses promising in vitro antitumor
growth inhibitory properties, with the concentration
required for 50% inhibition being around 200-300 nM.
Table 1a compares activity for 5 with those previously
reported by us28 for compounds 2 and 3, which shows
that compound 5 is consistently 10-20-fold more active
in the cell line panel. Furthermore, activity was ob-
served for 5 against two cell lines possessing resistance
to the DNA-damaging agent cisplatin and against a
P-glycoprotein overexpressing doxorubicin-resistant line.
Moreover, the two acquired cisplatin-resistant cell lines
are known to be resistant through distinct biochemical
mechanisms: A2780cisR, through multiple mechanisms
of reduced platinum transport, increased glutathione
and increased DNA repair, while CH1cisR increased
DNA repair or increased tolerance to platinum-DNA
adducts.32 Time of exposure experiments showed that
compound 5 was similarly potent whether drug expo-
sure was short (2 or 6 h) followed by wash-out or long
(24 or 96 h). This suggests that in vivo plasma concen-
trations required for in vitro growth inhibition (around
200 nM) only need to be attained for relatively short
periods (2-6 h) for antitumor efficacy to be obtained.
This effect is unlikely to be due to pharmacokinetic
factors, which would not usually be seen in an in vitro
setting but are more probably due to rapid saturation
of the drug and retained affinity for DNA. Furthermore,
data from the NCI COMPARE analysis suggest that
compound 5 does not share a pattern of growth inhibi-
tion with any other reported anticancer agents. Anti-
tumor activity was also observed in vivo using the
hollow fiber assay and, moreover, against an established
sc ovarian cancer xenograft. Using a day 0, 4, 8
schedule, activity was observed at day 10 post start of
treatment. The hollow fiber in vivo assay on all three
compounds (Table 1b) shows that compound 5 shows a
significant (though not outstanding) level of activity,
whereas the other two show only very modest responses.
Experiments are underway to further investigate
these findings on compound 5, as well as to design more
active analogues.
Exp er im en ta l Section
2,2-Bis(4′-m eth oxyp h en yl)-5,5-bi-1H-ben zim id a zole (2).
4,4′-Diamino-3,3′-dinitrobiphenyl (2 g, 7.3 mmol) (1) in acetone
(70 mL) was reduced by hydrogenation using 40 psi of H2 and
a catalytic amount of Raney nickel for 2 h. After filtration the
solvent was removed under reduced pressure to obtain a brown
solid which was used in the next step without further purifica-
tion. The crude intermediate and p-anisaldehyde (1.8 mL, 14.6
mmol) in nitrobenzene (20 mL) were heated at 150 °C under
argon for 8 h. An excess of hexane (40 mL) was added to the
cooled mixture to afford a precipitate and this was removed
by filtration. The precipitate was then purified by column
chromatography on silica gel (gradient from 100% of ethyl
acetate to 20% acetone/ethyl acetate (v/v)). Concentration in
vacuo yielded a pale yellow solid, which was redissolved in
acetone (1 mL) and precipitated in an excess of hexane (40
mL). Filtration gave 945 mg (29%) of a pale yellow solid: mp
267-270 °C; 1H NMR (400 MHz, DMSO-d6) δ 12.82 (s, 2H,
H3), 8.15 (d, 4H, J ) 8.0, H2′, H6′), 7.50-7.90 (m, 6H, H4, H6,
H7), 7.13 (d, 4H, J ) 8.0, H3′, H5′), 3.85 (s, 6H, OCH3); 13C NMR
(DMSO-d6) δ 160.6, 151.9, 144.6, 143.2, 135.7, 135.2, 127.9,
122.7, 118.6, 116.5, 114.3, 55.3; MS (EI+) m/z 447 M + H (100),
223 (5); HRMS calcd for C28H23N4O2 447.1820, found 447.1827.
4′,4′-[5,5-Bi-1H -b en zim id a zole-2,2-d iyl]b isp h en ol (3).
4,4′-Diamino-3,3′-dinitrobiphenyl (500 mg, 1.82 mmol) in
acetone (20 mL) was reduced by hydrogenation using 40 psi
of H2 and a catalytic amount of Raney nickel for 2 h. After
filtration the solvent was removed under reduced pressure to
obtain a brown solid which was used in the next step without
further purification. The crude intermediate and p-hydroxy-
benzaldehyde (445 mg, 3.64 mmol) were heated in nitroben-
zene (5 mL) at 150 °C under argon for 12 h. An excess of
hexane (50 mL) was added to the cooled mixture to afford a
precipitate and this was removed by filtration. The precipitate
was then purified by column chromatography on silica gel
(gradient from 100% ethyl acetate to 10% methanol/ethyl

142 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 2
Mann et al.
acetate (v/v)). Concentration in vacuo afforded a pale yellow
solid. This solid was redissolved in acetone (1.5 mL) and an
excess of hexane (35 mL) was added to yield a precipitate.
Filtration afforded 175 mg (23%) of a pale yellow solid: mp
256-260 °C; 1H NMR (250 MHz, DMSO-d6) δ 12.71 (s, 2H,
H3), 9.97 (s, 2H, OH), 8.03 (d, 4H, J ) 8.6, H3′, H5′), 7.48-7.88
(m, 6H, H4, H6, H7), 6.93 (d, 4H, J ) 8.6, H2′, H6′); 13C NMR
(DMSO-d6) δ 159.1, 152.7, 144.6, 143.1, 135.7, 135.0, 128.1,
121.1, 118.4, 116.3, 115.5; MS (EI+) m/z 419 M + H (30), 107
(100), 66 (25); HRMS calcd for C26H19N4O2 419.1507, found
419.1510.
4-(3′,3′-Dim eth ylam in o-1′-pr opyloxy) ben zaldeh yde (4).
Diethyl azodicarboxylate (9.55 mL, 61.4 mmol) was added
dropwise at 0 °C under argon to a solution of p-hydroxyben-
zaldehyde (5 g, 40.94 mmol), 3-dimethylaminopropan-1-ol (7.27
mL, 61.4 mmol) and triphenylphosphine (16.1 g, 61.4 mmol)
in THF (100 mL). The mixture was warmed slowly to room
temperature and stirred for 3 h. The solvent was removed
under reduced pressure and the crude oil obtained purified
by column chromatography on silica gel (gradient from 100%
DCM to 20% methanol/DCM (v/v)) to give 7.17 g (85%) of a
pale yellow oil: 1H NMR (400 MHz, CDCl3) δ 9.88 (s, 1H, H1),
7.82 (d, 2H, J ) 8.8, H3, H7), 7.00 (d, 2H, J ) 8.8, H4, H6), 4.10
(t, 2H, J ) 6.4, H1′), 2.46 (t, 2H, J ) 7.1, H3′), 2.26, (s, 6H,
NCH3), 1.99 (tt, 2H, J ) 6.4, J ) 7.1, H2′); 13C NMR (CDCl3) δ
190.8, 164.3, 132.2, 129.8, 115.1, 66.1, 56.1, 45.5, 27.5; MS (EI+)
8 control untreated wells) and left to incubate under normal
growth conditions for 96 h, unless otherwise stated. The
medium was then removed and wells fixed with 10% trichlo-
roacetic acid and stained with 0.4% SRB in 1% acetic acid as
described previously.32 Basic amino acid content in each well
was then measured by solubilizing the bound SRB with 10
mM Tris and using a Titertek Multiscan MCC/340 MKII plate
reader set at 540 nm. Mean absorbance was then expressed
as a percentage of the control untreated well absorbance and
plotted vs drug concentration. Comparisons were made in
terms of IC50 values (the concentration that reduced the mean
absorbance to 50% of those in control wells) and resistance
factors (RF, the IC50 in a resistant subline versus its parent
line).
Where the effect of time of drug exposure was investigated,
drug was removed by aspiration at the times indicated (2, 6,
24, and 96 h) and replaced with fresh growth medium and
incubated for the remainder of the 96-h period.
2. Hollow F iber Assa y. The hollow fiber assay was used
for the initial in vivo evaluation of the compound.36 The CH1
ovarian cell line was prepared by trypsinization and seeded
at 1 × 107 single cells/mL into 2-cm polyvinylidene fluoride
(PVDF) fibers (Spectrum, Laguna Hills, CA) and heat-sealed.
Three days later, following incubation of fibers in growth
medium, fibers were implanted into the peritoneal cavity (ip)
of nude mice using a trochar. One day later (day 4 after seeding
into fibers), mice were treated with either a single maximum
tolerated dose of compound 5 (4 mg/kg, ip administration) or
vehicle (10% DMSO/90% arachis oil). Six days later, fibers
were removed and the numbers of viable cells assessed in
treated versus control fibers using the colorimetric MTT assay
and measuring absorbance at 540 nm.36 Results are expressed
as a % treated/control absorbance.
3. Hu m a n Tu m or Xen ogr a ft Stu d ies. Tumor fragments
from the CH1 human ovarian xenograft (corresponding to the
CH1 cell line) were implanted sc into the flanks of adult female
nude mice. Animals bearing established tumors (largest
diameter of 6-8 mm) were then randomized (day 0) into
treatment groups of 6 mice to receive either compound X (at
the predetermined maximum tolerated dose of 4 mg/kg) or
vehicle (10% DMSO/90% arachis oil). Animals were dosed by
ip injection on days 0, 4, and 8 following randomization.
Tumor size was determined twice weekly by caliper mea-
surements, and tumor volumes calculated (volume ) [a × b2
× π]/6, where a and b are orthogonal tumor diameters). Tumor
volumes were then expressed as a percentage of the volume
at the start of treatment (relative tumor volume). The effect
of the drug was assessed in terms of the optimum T/C (the
ratio of the mean relative tumor volume of treated to that of
control groups) calculated on each day of measurement.
All procedures involving animals were performed within the
guidelines set by the Animal Ethics Committee of the Institute
of Cancer Research and the United Kingdom Coordinating
Committee on the welfare of animals in experimental neopla-
sia.
m/z 208 M + H (46), 86 (48), 58 (100); HRMS calcd for C12H18
NO2 208.1337, found 208.1337.
-
2,2-Bis[4′-(3′-d im eth yla m in o-1′′-p r op yloxy)p h en yl]-5,5-
bi-1H-ben zim id a zole (5). 4,4′-Diamino-3,3′-dinitrobiphenyl
(1.5 g, 5.47 mmol) in acetone (100 mL) was reduced by
hydrogenation using 40 psi of H2 and a catalytic amount of
Raney nickel for 3 h. After filtration the solvent was removed
under reduced pressure to obtain a brown solid which was used
in the next step without further purification. The crude
intermediate and 4-(3′-dimethylamino-1′-propyloxy)benzalde-
hyde (2.27 g, 10.93 mmol) in nitrobenzene (15 mL) were heated
at 150 °C under argon for 12 h. The mixture was cooled to
room temperature and methanol (100 mL) was added. The
solution was then washed with hexane (3 × 100 mL) to remove
the excess of nitrobenzene and the methanol extract concen-
trated in vacuo to give a brown solid. The solid was then
purified by column chromatography on silica gel (hexane and
a gradient from 50% ethyl acetate/methanol to 4% Et3N/
methanol (v/v)). Concentration in vacuo afforded a solid, which
was then redissolved in acetone (40 mL), and an excess of
hexane (400 mL). Filtration from the solution of the resulting
precipitate yielded 1.126 g (35%) of a yellow solid; mp 218-
220 °C; 1H NMR (400 MHz, DMSO-d6) δ 12.80 (s, 2H, H3), 8.12
(d, 4H, J ) 8.8, H2′, H6′), 7.49-7.88 (m, 6H, H4, H6, H7), 7.10
(d, 4H, J ) 8.8, H3′, H5′), 4.07 (t, 4H, J ) 6.2, H1′′), 2.35 (t, 4H,
J ) 6.9, H3′′), 2.14 (s, 12H, NCH3), 1.85-1.88 (m, 4H, H2′′); 13
C
NMR (DMSO-d6) δ 160.9, 152.0, 143.3, 136.1, 135.8, 128.9,
123.4, 122.2, 119.8, 119.6, 115.7, 66.9, 56.5, 46.1, 27.7; MS (EI+)
m/z 588 M+ (15), 515 (10), 58 (100); HRMS calcd for C36H40N6O2
588.3210, found 588.3203.
Cr ysta llogr a p h ic Stu d ies. 1. Cr ysta lliza tion a n d Da ta
Collection . The DNA sequence d(CGCGAATTCGCG), puri-
fied by HPLC, was purchased from Oswell Ltd. (Southampton,
U.K.). Crystals were grown from hanging drops at 284 K as
yellow cubes. The crystal that was used for data collection was
grown from a drop containing 1 µL of 4 mM single-stranded
DNA dissolved in 40 mM sodium cacodylate buffer, pH 6.8, 1
µL of 2 mM compound 5 (in the same buffer) and 2 µL of
Hampton Screening kit reagent number 16, consisting of 10%
2-methylpentane-2,4-diol (MPD), 12 mM spermine tetrahy-
drochloride, 80 mM NaCl and 20 mM MgCl2. This drop was
equilibrated against 700 µL of 35% MPD solution and the
crystal grew over a period of 4 weeks. The crystal was flash-
frozen in liquid nitrogen and the data were collected using a
Rigaku RAXIS-IV image plate detector using mirror-focused
Cu KR radiation from a Rigaku RU200 rotating-anode genera-
tor. The crystal-detector distance was 100 cm and the
oscillation range was 1.5°, leading to data collection at
maximum resolution of 2.0 Å. 40 frames were collected and
Biologica l Stu d ies. 1. Gr ow th In h ibition Assa ys. As-
sessment of growth inhibition was performed in a small panel
of human ovarian carcinoma cell lines using the sulfo-
rhodamine B (SRB) assay.32 Included in this panel were two
sublines possessing acquired resistance to cisplatin (A2780cisR
12-fold resistant and CH1cisR 6-fold resistant)32 and a subline
possessing acquired resistance to doxorubicin (90-fold), through
overexpression of P-glycoprotein.35 This panel is representative
of that used by us in several previous preclinical anticancer
drug evaluation studies.
The cell lines grew as monolayers in Dulbecco’s modified
Eagle’s medium containing 10% fetal bovine serum (Imperial
Laboratories, Andover, U.K.) supplemented with 2 mM
glutamine and 0.5 µg/mL hydrocortisone in 6% CO2/94% air.
The compound was dissolved in 10% dimethyl sulfoxide. Cells
were seeded into 96-well microtiter plates at 3000-5000/well
and allowed to attach overnight. Serial dilutions of drug in
growth medium were then added to quadruplicate wells (with

Bisbenzimidazole DNA Minor Groove-Binding Agents
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 2 143
(2) Trent, J . O.; Clark, G. R.; Kumar, A.; Wilson, W. D.; Boykin, D.
W.; Hall, J . E.; Tidwell, B. L.; Blagburn, B. L.; Neidle, S.
Targeting the Minor Groove of DNA: Crystal Structures of Two
Complexes between Furan Derivatives of Berenil and the DNA
Dodecamer d(CGCGAATTCGCG)2. J . Med. Chem. 1996, 39,
4554-4562.
(3) Boykin, D. W.; Kimar, A.; Xiao, G.; Wilson, W. D.; Bender, B.
C.; McCurdy, D. R.; Hall, J . E.; Tidwell, R. T. 2,5-Bis[4-(N-
alkylamidino)phenyl]furans as Anti-Pneumocystis carinii Agents.
J . Med. Chem. 1998, 41, 124-129.
(4) Francesconi, I.; Wilson, W. D.; Tanious, F. A.; Hall, J . E.; Bender,
B. C.; Tidwell, R. R.; McCurdy, D.; Boykin, D. W. 2,4-Diphenyl
Furan Diamidines as Novel Anti-Pneumocystis carinii Pneumo-
nia Agents. J . Med. Chem. 1999, 42, 2260-2265.
(5) Neidle, S.; Kelland, L. R.; Trent, J . O.; Simpson, I. J .; Boykin,
D. W.; Kumar, A.; Wilson, W. D. Cytotoxicity of Bis(phenylami-
dinium) Furan Alkyl Derivatives in Human Tumour Cell
Lines: Relation to DNA Minor Groove Binding. Bioorg. Med.
Chem. Lett. 1997, 7, 1403-1408.
(6) Marchini, S.; Ciro, M.; Gallinari, F.; Geroni, C.; Cozzi, P.;
Dincalci, M.; Broggini, M. Alpha-bromoacryloyl Derivative of
Distamycin A (PNU 151807): a New Noncovalent Minor Groove
DNA Binder with Antineoplastic Activity. Br. J . Cancer 1999,
80, 991-997.
(7) Reddy, B. S. P.; Sondhi, S. M.; Lown, J . W. Synthetic DNA Minor
Groove-Binding Drugs. Pharmacol. Ther. 1999, 84, 1-111.
(8) Lammler, G.; Herzog, H.; Saupe, E.; Schutze, H. R.; Chemo-
therapeutic Studies on Litomosodoides carini Infection of Mas-
tomys nataleusis 1. The Filaricidal Action of 2,6-bis-Benzimi-
dazoles. WHO Bull. 1971, 44, 751-756.
(9) Kraut, E. H.; Masspeis, L.; Bakerzak, S.; Grever, M.; Evaluation
of Pibenzimol (NSC322921) in Refractory Solid Malignancies.
Proc. Am. Soc. Clin. Oncol. 1988, 7, 62.
(10) Kraut, E.; Fleming, T.; Segal, M.; Neidhart, J .; Behrens, B. C.;
MacDonald, J . Phase II Study of Pibenzimol in Pancreatic
Cancer - a Southwest Oncology Group Study. Invest. New Drugs
1999, 9, 95-96.
(11) Ferguson, L. R.; Denny, W. A. Microbial Mutagenic Effects of
the DNA Minor Groove Binder Pibenzimol (Hoechst 33258)
and a Series of Mustard Analogues. Mutat. Res. 1995, 329,
19-27.
(12) Turner, P. R.; Denny, W. A. The Mutagenic Properties of DNA
Minor-groove Binding Ligands. Mutat. Res. 1996, 355, 141-169.
(13) Chiang, S.-Y.; Welch, J .; Rauscher, F. J . III.; Beerman, T. A.
Effects of Minor Groove Binding Drugs on the Interaction of
TATA Box Binding Protein and TFIIA With DNA. Biochemistry
1994, 33, 7033-7040.
(14) Chen, A. Y.; Chiang, Y.; Gatto, B.; Liu, L. F. DNA Minor Groove-
binding Ligands: a Different Class of Mammalian DNA Topo-
isomerase I Inhibitors. Proc. Natl. Acad. Sci. U.S.A. 1993, 90,
8131-8135.
(15) Harshman, K. D.; Dervan, P. B. Molecular Recognition of B-DNA
by Hoechst 33258. Nucleic Acids Res. 1985, 13, 4825-4835.
(16) Spink, N.; Brown, D. G.; Skelly, J . V.; Neidle, S. Sequence
Dependent Effects in drug-DNA Interaction: the Crystal
Structure of Hoechst 33258 Bound to the d(CGCAAATTTGCG)2
Duplex. Nucleic Acids Res. 1994, 22, 1607-1612.
(17) Vega, M. C.; Coll, M.; Aleman, C. Intrinsic Conformational
Preferences of the Hoechst Dye Family and their Influence on
DNA Binding. Eur. J . Biochem. 1996, 239, 376-383.
(18) Parkinson, J . A.; Barber, J .; Douglas, K. T.; Rosamond, J .;
Sharples, D. Minor-groove Recognition of the Self-complemen-
tary Duplex d(CGCGAATTCGCG)2 by Hoechst 33258: a High-
field NMR Study. Biochemistry 1990, 29, 10181-10190.
(19) Embrey, K. J .; Searle, M. S.; Craik, D. J . Interaction of Hoechst
33258 with the Minor Groove of the A + T-rich DNA duplex
d(GGTAATTACC)2 Studied in Solution by NMR Spectroscopy.
Eur. J . Biochem. 1993, 211, 437-447.
(20) Czarny, A.; Boykin, D. W.; Wood, A. A.; Nunn, C. M.; Neidle, S.;
Zhao, N.; Wilson, W. D. Analysis of van der Waals and
Electrostatic Contributions in the Interactions of Minor-Groove
Binding Benzimidazoles with DNA. J . Am. Chem. Soc. 1995, 117,
4716-4717.
(21) Wood, A. A.; Nunn, C. M.; Czarny, A.; Boykin, D. W.; Neidle,S.
Variability in DNA Minor Groove Width Recognised by Ligand
Binding. The Crystal Structure of a bis-Benzimidazole Com-
pound Bound to the DNA Duplex d(CGCGAATTCGCG)2. Nucleic
Acids Res. 1995, 23, 3678-3684.
(22) Clark, G. R.; Squire, C. J .; Gray, E. J .; Leupin, W.; Neidle, S.
Designer DNA-Binding Drugs: The Crystal Structure of a meta-
Hydroxy Analogue of Hoechst 33258 bound to d(CGCGAAT-
TCGCG)2. Nucleic Acids Res. 1996, 24, 4882-4889.
Ta ble 2. Crystallographic Data
wavelength (Å)
unit cell dimensions
1.542
a ) 25.15 Å, R ) 90°
b ) 39.94 Å, â ) 90°
c ) 65.65 Å, ø ) 90°
space group
no. of reflections
P212121
4862
4664
4198
2465
3856
486
70
15
1
2.0
no. of unique reflections
no. of reflections used in refinement
no. of parameters refined
restraints
no. of atoms in DNA
no. of full-occupancy waters
no. of half-occupancy waters
no. of magnesium ions
max resolution of obsd reflections (Å)
max resolution of refinement (Å)
R factor (%)
2.0
24.4
30.0
Rfree (%)
the crystal did not appear to decay during collection. Data
reduction and scaling were carried out using the DENZO and
SCALEPACK programs.37 Data collection statistics are given
in Table 2.
2. Str u ctu r e Solu tion a n d Refin em en t. The structure
is isomorphous with other previously determined drug-
dodecanucleotide complex structures incorporating the same
sequence, and the structure of the native sequence (structure
BDL001 in the Nucleic Acid Database) was chosen as a
starting point for the analysis. The program XPLOR, version
3.1,38 was used to carry out positional refinement, initially for
individual groups and subsequently for individual atoms. The
resolution was progressively increased until the final indi-
vidual positional refinement included all data to 2.0 Å (2405
reflections), at which point further refinement was carried on
with the SHELX-97 program.39
A Fo - Fc map was generated and electron density corre-
sponding to the drug could clearly be seen, though the density
for the side chain regions was not very well-defined. The three-
dimensional structure for compound 5 was built with the
INSIGHTII package. Refinement was continued after intro-
duction of the drug as well as a dictionary of the drug
geometry. A hydrated magnesium ion was located in close
proximity to T17, as observed previously in other minor groove
structures. A total of 79 solvent molecules were located, of
which 65 had full occupancy, according to standard criteria.
The side chains of the drug molecule had high temperature
factors throughout the refinement, which contributed to the
final R and Rfree values being somewhat greater than normal:
24.4% and 30.0%, respectively. DNA helicoidal parameters
were generated with the program CURVES.29 Coordinates and
structure factors have been deposited with the Nucleic Acid
Database41 as entry no. DD0036.
Ack n ow led gm en t. We are grateful for support from
the Cancer Research Campaign (L.R.K.: Grant No. SP/
DC/STU2330/0201; S.N.: Grant No. SP1384). Y. O.-B
and A. B. thank the Government of Ghana and the
Royal Society of Chemistry Adrien Albert Bequest,
respectively, for studentships. We thank P. Rogers, F.
Boxall, M. Valenti, and L. Brunton for technical as-
sistance with SRB, hollow fiber, and xenograft assays.
We thank the Developmental Therapeutics Program of
the National Cancer Institute, Bethesda, MD, for the
hollow fiber data.
Refer en ces
(23) Clark, G. R.; Boykin, D. W.; Czarny, A.; Neidle, S. Structure of
a bis-amidinium derivative of Hoechst 33258 complexed to the
dodecanucleotide d(CGCGAATTCGCG)2: the role of hydrogen
bonding in minor groove drug-DNA recognition. Nucleic Acids
Res. 1997, 25, 1510-1515.
(1) Boykin, D. W.; Kumar, A.; Spychala, J .; Zhou, M.; Lombardy,
R. J .; Wilson, W. D.; Dykstra, C. C.; J ones, S. K.; Hall, J . E.;
Tidwell, R. R.; Laughton, C.; Nunn, C. M.; Neidle, S. Dicationic
Diaryl Furans as anti-Pneumocystis carinii Agents. J . Med.
Chem. 1995, 38, 912-916.

144 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 2
Mann et al.
(24) Clark, G. R.; Gray, E. J .; Li, Y.-H.; Leupin, W.; Neidle, S.
Isohelicity and Phasing in Drug-DNA sequence recognition:
Crystal Structure of a Tris (Benzimidazole)-Oligonucleotide
Complex. Biochemistry 1996, 35, 13745-13752.
(25) Pilch, D. S.; Xu, Z.; Sun, Q.; LaVoie, E. J .; Liu, L. F.; Breslauer,
K. J . A Terbenzimidazole that Preferentially Binds and Confor-
mationally Alters Structurally Distinct DNA Duplex Domains:
a Potential Mechanism for Topoisomerase I Poisoning. Proc.
Natl. Acad. Sci. U.S.A. 1997, 94, 13565-13570.
(26) Kim, J . S.; Sun, Q.; Chaing, Y.; Liu, A.; Liu, L. F.; LaVoie, E. J .
Quantitative Structure-activity Relationships on 5-substituted
Terbenzimidazoles as Topoisomerase I Poisons and Antitumor
Agents. Bioorg. Med. Chem. 1998, 6, 163-172.
(27) Xu, Z.; Tsai-Kun, L.; Kim, J . S.; LaVoie, E. J .; Breslauer, K. L.;
Liu, L. F.; Pilch, D. S. DNA Minor Groove Binding-directed
Poisoning of Human DNA Topoisomerase I by Terbenzimida-
zoles. Biochemistry 1998, 37, 3558-3566.
(28) Neidle, S.; Mann, J .; Rayner, E. L.; Baron, A.; Opoku-Boachen,
Y.; Simpson, I. J .; Smith, N. J .; Fox, K. R.; Hartley, J . A.;
Kelland, L. R. Symmetric bis-Benzimidazoles, a New Class of
Sequence-Selective DNA-binding Molecules. J . Chem. Soc.,
Chem. Commun. 1999, 929-930.
(29) Lavery, R.; Sklenar, H. Defining the Structure of Irregular
Nucleic Acids - Conventions and Principles. J . Biomol. Struct.
Dyn. 1989, 6, 655-667.
(30) Monks, A.; Scudiero, D. A.; J ohnson, G. S.; Paull, K. D.; Sausville,
E. A. The NCI Anti-cancer Drug Screen: a Smart Screen to
Identify Effectors of Novel Targets. Anti-Cancer Drug Des. 1997,
12, 533-541.
(31) Hollingshead, M. G.; Alley, M. C.; Camalier, R. F.; Abbott, B.
J .; Mayo, J . G.; Malspeis, L.; Grever, M. R. In vivo Cultivation
of Tumor cells in Hollow Fibres. Life Sci. 1995, 57, 131-141.
(32) Kelland, L. R.; Barnard, C. F. J .; Mellish, K. J .; J ones, M.;
Goddard, P. M.; Valenti, M.; Bryant, A.; Murrer, B. A.; Harrap,
K. R. A Novel trans-Platinum Coordination Complex Possessing
in vitro and in vivo Antitumor Activity. Cancer Res. 1994, 54,
5618-5622.
(33) Broggini, M.; D′Incalci, M. Modulation of Transcription factor-
DNA Interactions by Anticancer Drugs. Anti-Cancer Drug Des.
1994, 9, 373-387.
(34) Soderlind, K.-J .; Gorodetsky, B.; Singh, A. K.; Bachur, N. R.;
Miller, G. G.; Lown, J . W. Bis-benzimidazole anticancer agents:
targeting human tumour helicases. Anti-Cancer Drug Des. 1999,
14, 19-36.
(35) Sharp, S. Y.; Rowlands, M.; J arman, M.; Kelland, L. R. Effects
of a New Antioestrogen Idoxifene on Cisplatin- and Doxorubicin-
sensitive and Resistant Human Ovarian Carcinoma Cell Lines.
Br. J Cancer 1994, 70, 409-414.
(36) Mosmann T, J . Rapid Colorimetric Assay for Cellular Growth
and Survival: Application to Proliferation and Cytotoxicity
Assays. Immunol. Methods 1983, 65, 55-63.
(37) Otwinowski, Z.; Minor, W. In Data Collection and Processing;
Sawyer, L., Isaacs, N. W., Bailey, S., Eds.; SERC Daresbury
Laboratory: Warrington, U.K., 1993.
(38) Bru¨nger, A. T.; Kuriyan, J .; Karplus, M. Crystallographic R
factor Refinement by Molecular Dynamics. Science 1987, 235,
458-460.
(39) Sheldrick, G. M. SHELX-97, a crystallographic refinement
program; University of Go¨ttingen: Germany, Go¨ttingen, 1997.
(40) Nicholls, A.; Sharp, K. A.; Honig, B. Protein Folding and
Association: Insights from the Interfacial and Thermodynamic
Properties of Hydrocarbons. Proteins 1991, 11, 281-296.
(41) Berman, H. M.; Olson, W. K.; Beveridge, D. L.; Westbrook, J .;
Gelbin, A.; Demeny, T.; Hsieh, S.-H.; Srinivasan, A. R.; Schneider,
B. The Nucleic Acid Database: a Comprehensive Relational
Database of Three-Dimensional Structures of Nucleic Acids.
Biophys. J . 1992, 63, 751-759.
J M000297B