Novel furano analogues of duocarmycin C1 and C2: design, synthesis, and biological evaluation of seco-iso-cyclopropylfurano[2,3-e]indoline (seco-iso-CFI) and seco-cyclopropyltetrahydrofurano[2,3-f]quinoline (seco-CFQ) analogues.

Article abstract of DOI:10.1016/S0968-0896(02)00157-8

The design, synthesis and biological evaluation of novel seco-iso-cyclopropylfurano[2,3-e]indoline (seco-iso-CFI) and the seco-cyclopropyltetrahydrofurano[2,3-f]quinoline (seco-CFQ) analogues of the duocarmycins are described. These novel analogues (4-7) were designed on the premise that the lone pair of electrons on the furano-oxygen atom could enter into conjugation with the isocyclopropylfurano[e]indolone (iso-CFI) alkylating moiety, formed from the loss of HCl in compounds 4-7. The seco-iso-CFI DNA alkylating pharmacophore was synthesized through a well precedented approach of 5-exo-trig aryl radical cyclization with a vinyl chloride. In our studies, in addition to the formation of the seco-iso-CFI product, an equal amount of an unexpected seco-CFQ product was also generated during the radical cyclization reaction. Like CC-1065 and adozelesin, using Taq DNA polymerase stop and thermal cleavage assays, the seco-iso-CFI compounds (4 and 6) and the seco-CFQ compounds (5 and 7) were shown to preferentially alkylate the adenine-N3 position within the minor groove of long stretches of A residues. A MM2 energy optimized molecular model of a 1:1 complex of compound 6 with DNA reveals that the iso-CFI compound fits snugly within the minor groove. Using a MTT based experiment, the cytotoxicity of compounds 4-7 were determined against the growth of murine leukemia (L1210), mastocytoma (P815) and melanoma (B16) cell lines. The concentrations of compounds required to inhibit the growth of these tumor cells by 50% is in the range of 10(-8)M. These compounds were also tested against a panel of human cancer cells by the National Cancer Institute, demonstrating that the compounds exhibited a high level of activity against selected solid tumors. At a concentration of 0.0084 microM (based on the IC(50) of compound 17 (seco-CBI-TMI) against the growth L1210 cells), while compounds 4 and 17 were toxic against murine bone marrow cells as judged by a colony forming study of freshly isolated murine progenitor hematopoeitic cells, compound 5, a seco-CFQ compound, was significantly less toxic. Flow cytometric analysis of P815 cells that had been incubated for 24h with compounds 4 and 5 at their cytotoxic IC(50) concentrations indicated the induction of apoptosis in a large percentage of cells, thereby suggesting that this might be the mechanism by which the iso-CFI compounds kill cells.

Full text of DOI:10.1016/S0968-0896(02)00157-8

Bioorganic & Medicinal Chemistry 10 (2002) 2941–2952
Novel Furano Analogues of Duocarmycin C1 and C2:
Design, Synthesis, and Biological Evaluation of
seco-iso-Cyclopropylfurano[2,3-e]indoline (seco-iso-CFI)
and seco-Cyclopropyltetrahydrofurano[2,3-f ]quinoline
(seco-CFQ) Analogues
Tiffany T. Howard,a Brian M. Lingerfelt,a Bethany L. Purnell,a Adrienne E. Scott,a
Carly A. Price,a Heather M. Townes,a LuAnne McNulty,a Heather L. Handl,a
Kaitlin Summerville,a Stephen J. Hudson,b J. Phillip Bowen,c
Konstantinos Kiakos,d John A. Hartleyd and Moses Leea,*
aDepartment of Chemistry, Furman University, Greenville, SC 29613, USA
bDepartment of Biology, Furman University, Greenville, SC 29613, USA
cDepartment of Chemistry, University of Georgia, Athens, GA 30602, USA
dDepartment of Oncology, Royal Free University College Medical School, London, W1W 7BS, UK
Received 15 January 2002; accepted 16 April 2002
This paper is dedicated to Professor Gordon L. Lange on the occasion of his retirement.
Abstract—The design, synthesis and biological evaluation of novel seco-iso-cyclopropylfurano[2,3-e]indoline (seco-iso-CFI) and the
seco-cyclopropyltetrahydrofurano[2,3-f]quinoline (seco-CFQ) analogues of the duocarmycins are described. These novel analogues
(4–7) were designed on the premise that the lone pair of electrons on the furano-oxygen atom could enter into conjugation with the
isocyclopropylfurano[e]indolone (iso-CFI) alkylating moiety, formed from the loss of HCl in compounds 4–7. The seco-iso-CFI
DNA alkylating pharmacophore was synthesized through a well precedented approach of 5-exo-trig aryl radical cyclization with a
vinyl chloride. In our studies, in addition to the formation of the seco-iso-CFI product, an equal amount of an unexpected seco-
CFQ product was also generated during the radical cyclization reaction. Like CC-1065 and adozelesin, using Taq DNA polymerase
stop and thermal cleavage assays, the seco-iso-CFI compounds (4 and 6) and the seco-CFQ compounds (5 and 7) were shown to
preferentially alkylate the adenine-N3 position within the minor groove of long stretches of A residues. A MM2 energy optimized
molecular model of a 1:1 complex of compound 6 with DNA reveals that the iso-CFI compound fits snugly within the minor
groove. Using a MTT based experiment, the cytotoxicity of compounds 4–7 were determined against the growth of murine leuke-
mia (L1210), mastocytoma (P815) and melanoma (B16) cell lines. The concentrations of compounds required to inhibit the growth
of these tumor cells by 50% is in the range of 10ꢀ8 M. These compounds were also tested against a panel of human cancer cells by
the National Cancer Institute, demonstrating that the compounds exhibited a high level of activity against selected solid tumors. At
a concentration of 0.0084 mM (based on the IC50 of compound 17 (seco-CBI-TMI) against the growth L1210 cells), while com-
pounds 4 and 17 were toxic against murine bone marrow cells as judged by a colony forming study of freshly isolated murine
progenitor hematopoeitic cells, compound 5, a seco-CFQ compound, was significantly less toxic. Flow cytometric analysis of P815
cells that had been incubated for 24 h with compounds 4 and 5 at their cytotoxic IC50 concentrations indicated the induction of
apoptosis in a large percentage of cells, thereby suggesting that this might be the mechanism by which the iso-CFI compounds kill
cells. # 2002 Elsevier Science Ltd. All rights reserved.
*Corresponding author. Tel.: +1-864-294-3368; fax: +1-864-294-3559; e-mail: moses.lee@furman.edu
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00157-8

2942
T. T. Howard et al. / Bioorg. Med. Chem. 10 (2002) 2941–2952
Introduction
The duocarmycins, such as duocarmycin A (1), C1 (2)
and C2 (3) as depicted in Figure 1, and CC-1065 are
highly potent anticancer agents, and the development of
analogues of these compounds into potential drugs for
cancer treatment has received much attention.1 Several
analogues of these natural products have entered clin-
ical trials: adozelesin,2 carzelesin,3 KW2189,4 and bize-
lesin.5 While these compounds demonstrated significant
anticancer activity, their clinical efficacy was sig-
nificantly limited by their bone marrow toxicity, and
currently only bizelesin is still under clinical evaluation
(phase II).5c Analogues of the duocarmycins and
CC-1065 derive their cytotoxic potency by covalently
reacting with the adenine-N3 position within the minor
groove of AT-rich DNA sequences, including a pre-
ferred site 50-AGTTA-30 for CC-1065,1 and their
mechanism of cytotoxic action has been suggested to be
a result of drug induced apoptosis of cancer cells.6
Figure 1. Structures of duocarmycin A, C1, C2, CC-1065 and iso-CFI-
TMI.
In attempts to design novel analogues of the duo-
carmycins and CC-1065 with reduced toxicity to bone
marrow cells, a wide range of analogues of the cyclo-
propylpyrrolo[e]indolone (CPI) subunit were synthe-
sized and tested.1 Examples of such analogues,
cyclopropylbenzo[e]indolone (CBI),1,7 cyclopropylpyr-
azolo[e]indolone (CPzI),8 cyclopropylfurano[e]indolone
(CFI),9 and cyclopropylindolone (CI),10 are given in
Figure 2. From studies of these analogues several trends
of structure–activity relationships became apparent.1
First, the cytotoxicity of the analogues is directly related
to the chemical stability of the molecules in aqueous
acidic solutions.1dꢀh The stability is believed to arise
from vinylogous resonance conjugation of the lone pair
of electrons on the nitrogen atom of the carboxamido
moiety and the cyclopropylindolone unit.1d,11 It has also
been demonstrated that the electron withdrawing cap-
ability of the carboxamido unit will further stabilize the
cyclopropylindolone unit against solvolysis.12 Second,
inclusion of an extended aromatic unit to the CI unit, as
in the case of CBI, significantly enhances chemical sta-
bility and cytotoxicity.1,7,11 Third, attachment of an
electron withdrawing group to the extended aromatic
moiety also improves the chemical stability of the
pharmacophore.1d,13 The extensive SAR studies have
further revealed that the seco-prodrugs are equally
reactive with DNA and as cytotoxic as the cyclopropane
containing drugs.1,10 The seco-prodrugs readily lose
HCl in cells or in the cell culture media to generate the
active cyclopropane containing drugs to elicit the bio-
logical activity. It has also been shown that both the six-
and five-membered ring containing seco-prodrugs, duo-
carmcyin C1 and C2, respectively, undergo a loss of
HCl to generate Duocarmycin A.14
Figure 2. Examples of known DNA alkylating subunits of the duo-
carmycins and CC-1065.
addition to the carboxamido moiety, the furano-oxygen
should provide a second set of vinylogous conjugation
to the pharmacophore and stabilize the cyclopropy-
lindolone funtionality toward nucleophiles. Another
attractive feature of the iso-CFI pharmacophore is that
furans are known to be stable towards biological
degradation.9a In this paper, the synthesis, DNA bind-
ing properties, cytotoxicity, and bone marrow toxicity
of two classes of seco-prodrugs of iso-CFI compounds,
4–7 as depicted in Figure 3, are described. The seco-iso-
CFI compounds 4 and 6 are analogous to duocarmycin
C2, and the seco-CFQ agents 5 and 7 are congeners of
duocarmycin C1. While extensive studies have been
done on seco-chloromethylindoline precursors of the
cyclopropane-containing drugs,1 only limited work has
been done on seco-chloroquinoline analogues that are
related to duocarmycin C1.15 This paucity of work
might be partly due to the lack of available synthetic
methodologies. In this paper, formation of the seco-
CFQ products represents a direct and efficient synthesis
of chloroquinoline precursors of the cyclopropaneindo-
lone compounds.
In the present study, novel seco-prodrugs of iso-CFI
analogues of duocarmycins were synthesized and inves-
tigated. The ultimate alkylating agent iso-CFI, as
depicted in the iso-CFI-TMI molecule given in Figure 1,
was designed on the basis that the lone pair of electrons
of the furan-oxygen atom would enter into conjugation
(+R effect) with the cyclopropylindolone moiety. In

T. T. Howard et al. / Bioorg. Med. Chem. 10 (2002) 2941–2952
2943
afforded a carboxylic acid intermediate in 93% yield,
which upon treatment with the Shioiri–Yamada reagent
(DPPA)16 in t-butanol and triethylamine produced the
BOC protected amine 10 in quantitative yield. Treat-
ment of compound 10 with NBS and p-toluenesulfonic
acid at ꢀ78 ꢁC generated a bromide (55%), which was
alkylated with (E/Z)-1,3-dichloropropene in the pre-
sence of sodium hydride to give compound 11 in 90%
yield. A degassed solution of compound 11 in toluene
was heated at reflux in the presence of 2.7 equivalents of
tri-n-butlytin hydride and 0.6 equivalents of AIBN to
produce the 5-exo-trig aryl radical/vinyl chloride cycli-
zation reaction9b,17 and to give the desired seco-iso-CFI
alkylating subunit 12 in 40% yield after silica gel col-
umn purification. In addition to compound 12, a new
and slightly less polar product was also obtained in 48%
yield. The structure of compound 13 was assigned as the
protected seco-CFQ (seco-cyclopropanefuranoquino-
Figure 3. Target molecules of this study: seco-iso-CFI analogues (4 and
6) and seco-CFQ analogues (5 and 7). Structure of the reference com-
pound seco-CBI-TMI (17).20
1
line) on the basis of its 500 MHz H NMR and mass
spectral data. The proton signals of the chloroquinoline
subunit, 4.17 (t, 9.5, 2H), 3.5–3.65 (m, 3H), are con-
sistent with the pattern observed for duocarmycin C1.14
It is worthy to note that variation of the reaction tem-
perature (80–110 ꢁC) as well as the concentration (1–2
dilutions) did not significantly alter the chemical yields
and proportions of the isomers produced.
The formation of equal amounts of the seco-CFQ pro-
duct 13 in addition to the 5-exo-trig (seco-iso-CFI) pro-
duct 12 is note worthy because this type of radical
cyclization reactions on a diverse range of structures
have been widely reported, including the formation of
seco-CFI.9b However, in all cases, the five-membered
ring or 5-exo-trig product was obtained exclusively.9b,17
Under a similar reaction condition for the radical pro-
moted formation of compounds 12 and 13, and con-
sistent with literature precedence,17 the reaction of a
corresponding substituted naphthalene reactant gave
the seco-CBI structure exclusively. As indicated in Fig-
ure 4, we propose that the formation of the seco-CFQ is
most likely influenced by the furano-oxygen atom. The
lone pair of electrons provides stabilization on the
5-membered ring radical intermediate, formed via a
5-exo-trig ring closure reaction. Instead of immediately
reacting with tributyltin hydride, the radical inter-
mediate undergoes a ring expansion reaction to produce
a more stable intermediate, which has a six-membered
ring (seco-CFQ). The seco-CFQ radical is further stabi-
lized by conjugation of the radical at the benzylic posi-
tion with the benzofuran moiety. Subsequent reaction
of the seco-CFQ radical with tributyltin hydride affords
product 13.
Scheme 1. (a) Potassium t-butoxide, reflux; (b) acetic anhydride,
sodium acetate, reflux; (c) K2CO3, ethanol, reflux; (d) benzyl bromide,
NBu4I, DMF, K2CO3, rt; (e) (i) NaOH; (ii) HCl; (f) DPPA; t-butanol,
Et3N, heat; (g) NBS, pTSOH, methanol/THF, ꢀ78 ꢁC; (h) NaH,
DMF, 1,3-dichloro-2-propene; (i) AIBN, Bu3SnH, toluene, heat; (j) (i)
HCl, THF; (ii) TMI–CO2H, EDCI, DMF, rt; (k) ammonium formate,
10% Pd/C, aq THF, H2; (l) (i) H2, 10% Pd/C, THF; (ii) NaH, THF.
Results and Discussion
Synthesis
The synthetic approach utilized for the preparation of
the seco-CFI compounds is indicated in Scheme 1, and
it follows the general methodology reported by Boger
and coworkers.1dꢀh,7,13 A Stobbe condensation of
2-furaldehyde with diethyl succinate in the presence of
potassium t-butoxide afforded an isomeric mixture of
half-esters in 56% yield. The half-esters were converted
into the substituted benzofuran by Friedel–Crafts acy-
lation (AcONa, Ac2O, reflux), followed by removal of
the acetate group with ethanolic sodium carbonate to
provide phenol 8 in 54% yield. Protection of the
hydroxy group with benzyl bromide gave product 9 in
95% yield. Hydrolysis of the ester group in compound 9
To complete the synthesis of the target molecules 4 and
5, the BOC protecting group on compounds 12 and 13
were removed by treatment with HCl. The resulting
amines were directly coupled with 5,6,7-trimethoxy-
indole-2-carboxylic acid in the presence of EDCI to
produce amides 14 and 15 in 64 and 80% yields,
respectively. Removal of the benzyl moieties of com-
pounds 14 and 15 by catalytic hydrogenolysis provided
racemic mixtures of compounds 4 and 5 in almost
quantitative yields. The seco-iso-methyl-CFI (6) and

2944
T. T. Howard et al. / Bioorg. Med. Chem. 10 (2002) 2941–2952
Figure 4. Formation of the seco-CFQ and seco-iso-CFI structures via
a 5-exo-trig radical/vinyl chloride cyclization mechanism.
seco-methyl-CFQ (7) compounds were prepared in a
similar fashion as compounds 4 and 5, except the
synthesis began with 5-methylfurfural.
In order to test the relative solvolytic stability of the iso-
CFI pharmacophore, the cyclopropane-containing pro-
duct 16 was generated. Removal of the benzyl protect-
ing group in compound 12 was achieved by catalytically
hydrogenation in quantitative yield. Reaction of the
resulting phenol with sodium hydride in THF gave
compound 16 in 23% yield. A room temperature
0.1 mM solution of N-BOC protected 16 in a 1:1
methanol/water mixture at pH 3 was monitored by
UV–vis spectrophotometry. Following the dis-
appearance of the absorbance for the indolone moiety
at 320 nm with time, a half-life for compound 16 was
found to be 410 h. A similar study was undertaken on
N-BOC-CBI, and its half-life was found to be 235 h,
indicating that the iso-CFI molecule was somewhat
more stable under the conditions studied. The reported
half-life for N-BOC-CBI is 133 h.12a
Figure 5. Taq polymerase gel examining damage on the bottom strand
of a mixed AT/GC sequence (defined in the Experimental). 10 bp:
molecular weight marker; C and A: sequencing lanes specific for
cytosines and adenines; 0: control; adozelesin: 0.1 mM; compound 4: 1,
10, and 100 mM; compound 6: 1, 10, and 100 mM; compound 5: 10 and
100 mM; compound 7: 10 and 100 mM (arrows indicate bases alkylated
and their position on the plasmid sequence.
trimethoxyindole group in compounds 4–7. Even
though compounds 4–6 produced similar alkylation
efficiencies, compound 7 was somewhat less reactive for
reasons that are unclear at this stage.
The thermally induced DNA sequence selective strand
break assay is a commonly used method for determining
the covalent sequence selectivity of compounds pro-
vided that the alkylation occurs with purine N3 nucleo-
philes in the minor groove.1,19 For this study, the probe
sequence was obtained from PCR amplification of base
pairs 749–956 of plasmid pUC18 that was linearized
with Hind III. In addition to the above primer 1, a sec-
ond primer 2 (CTGTCGGGTTT) fragment was used as
a forward primer so that the final DNA fragment was
singly end labeled. Results from the thermal cleavage
assay for compounds 4, 5 and 17 (seco-CBI-TMI),20
shown in Figure 6, revealed that these agents could
covalently interact with the N3 position of adenine
residues in the minor groove. Consistent with results
from the Taq stop assay, all three of these agents
recognized and alkylated the sequence 30-AAAAA-50 at
position 865, with approximately the same efficiency.
All agents also recognized secondary sites, for example,
at A764 and A772 within AT-rich sequences.
DNA binding studies
The Taq DNA Polymerase stop assay developed by
Hartley and co-workers is very sensitive in determining
the sequence selective alkylation of DNA.18 This PCR
based assay works on the premise that Taq DNA poly-
merase activity will stop DNA synthesis at the site of
alkylation. Using a singly radiolabeled primer 50-32P-
CTCACTCAAAGGCGGTAATAC-30 and a Hind III
linearized pUC18 plasmid DNA, the region of the plas-
mid starting from position 749 was linearly amplified.
The autoradiograph shown in Figure 5 demonstrates
that, like adozelesin,2 compounds 4–7 could stop the
activity of Taq DNA polymerase at sites with con-
tiguous adenine residues, and all four compounds
showed similar covalent sequence selectivity at the
A(865) cluster. These results are not surprising since
both the seco-iso-CFI and seco-CFQ prodrugs would
undergo a loss of HCl to generate an identical iso-CFI-
TMI compound (see Fig. 1). It is worthy to note that
CC-1065 was more reactive with the A(865) cluster than
compounds 4–7. This observation could be a result of
the enhanced binding of CC-1065 to the minor groove
due to the larger PDE units, when compared to the
Molecular modeling studies
The MM2 energy minimized complex of iso-CFI-TMI
with duplex dGGCGGAGTTAGG in the B-form, and
alkylating at the adenine N3 position of the underlined
A residue, is shown in Figure 7. This oligonucleotide
was used by Hurley and coworkers in their NMR and
molecular modeling studies of a DNA complex with

T. T. Howard et al. / Bioorg. Med. Chem. 10 (2002) 2941–2952
2945
Figure 7. MM2 energy optimized (steepest descend) structure of a
complex of a covalent adduct between the chloromethyl moiety of
compound 4 and adenine–N3 of the underlined residue of this oligo-
nucleotide dGGCGGAGTTAAGG.
Table 1. In vitro cytotoxicity (IC50 in mM) of compounds 4–7 and 17
against murine cancer cells (the cells were incubated with the com-
pounds continuously for 3 days)
Compd/cells
L1210
P815
B16
4
5
6
7
17
0.053
0.064
0.030
0.032
0.0084
0.051
0.061
0.064
0.018
0.0086
0.040
0.062
0.092
0.064
0.076
Figure 6. Thermal cleavage gel showing purine–N3 lesions on the
upper strand of the mixed AT/GC region. 0: control; Compound 4: 3
and 10 mM; Compound 5: 3 and 10 mM; Compound 17 (seco-CBI-
TMI): 3 and 10 mM.
compound 17, and for B16 melanoma cells compounds
4, 5, and 7 are slightly more cytotoxic. Furthermore,
compounds 6 and 7, the methylated analogues of the
seco-iso-CFI and seco-CFQ compound, have nearly
indistinguishable cytotoxic properties with their parent
compounds 4 and 5, respectively.
CC.1c As depicted in Figure 7, the iso-CFI-TMI mole-
cule fits snugly and isohelically in the minor groove. The
angular plane between the iso-CFI and TMI units exhi-
bits a twist of 6ꢁ in order to accommodate the natural
curvature of the DNA.
The cytotoxic properties of compounds 4–7 were also
examined at the National Cancer Institute, against a
panel of 60 different human cancer cell lines.22 The via-
bility of the cells after 48 h of continuous exposure of
the compounds was determined using a sulforhodamine
B assay. The LC50 values (concentration for killing 50%
of the cells) for compounds 4 and 5 against the 60 dif-
ferent tumor cells are given in Figure 8. Bars extending
to the right indicate cells more sensitive than the average
to the particular compound, whereas bars to the left
indicate less sensitive cells. One of the interesting fea-
tures that are apparent from these studies is that agents
4–7 are not indiscriminately toxic to cells, and they
show unique patterns of cytotoxicity primarily against
cells derived from solid tumors such as breast, prostate,
renal, melanoma, CNS and colon cancers. Even though
these compounds demonstrated subtly different patterns
of cytotoxicity, they tend to be more active against
melanoma and renal cancer cells. Specific reasons for
the unique patterns of cytotoxicity for compounds 4–7
remain unknown, and experiments are underway to
address this issue.
Cytotoxicity studies
Although it is important to characterize the interaction
of potential anticancer agents with their DNA target,
the goal is to design compounds which are capable of
killing tumor cells. Using a MTT based growth inhibi-
tion assay, the IC50 values for compounds 4–7 against
murine L1210 leukemia, murine P815 mastocytoma,
and murine B16 melanoma cells were determined.21
Compound 17, which has a reported IC50 value of
30 pM against L1210 cells (3 days’ exposure),20 was used
as a reference compound for the cytotoxicity studies.
The results given in Table 1 demonstrate that the novel
seco-iso-CFI (4 and 6) and seco-CFQ (5 and 7) agents,
which contain an additional electron donating furan-
oxygen compared to a benzo moiety in CBI units, have
IC50 values in the 10ꢀ8 M range. This level of cyto-
toxicity is almost comparable to the average value for

2946
T. T. Howard et al. / Bioorg. Med. Chem. 10 (2002) 2941–2952
Figure 8. NCI mean graph (log LC50 values) for a 48-h in vitro assay on compounds 4 and 5.
Cell-cycle studies
damaging DNA and inducing the cells to undergo
apoptosis. For comparison, cisplatin at 100 mM, pro-
duced 17% of sub-G0 population of cells, which is con-
sistent with a literature precedent.6 Results from our
studies suggest that iso-CFI compounds are likely to
exert their cytotoxic activity through the induction of
apoptosis.
Evidence that duocarmycin analogues cause cellular
death through apoptosis was obtained through micro-
scopic examination of Molt-4 leukemia cells that had
been incubated with duocarmycin SA. Those cells
demonstrated morphological abnormalities and geno-
mic DNA degradation associated with apoptosis.6 Flow
cytometry is another way in which apoptotic cells can
be detected. It has been suggested that the sub-G0 peak
is indicative of the presence of apoptotic cells,23 that can
be determined by a propidium iodide staining method.
Following a 24-h incubation of P815 cells at the cyto-
toxic IC50 concentration of compound 5, the results
showed a substantial increase in the percentage of sub-
G0 stage (22% compared to <1% for untreated control
cells), indicating that compound 5 was capable of
Bone marrow toxicity studies
One of the most severe limitations in the development of
analogues of CC-1065 and the duocarmycins into useful
therapeutic agents is their severe non-selective toxicity
toward the bone marrow.2ꢀ5 For this reason, the toxi-
city of compounds 4 and 5 against hematopoietic pro-
genitor cells were assessed using a commercially
available colony forming assay (StemCell Technologies,

T. T. Howard et al. / Bioorg. Med. Chem. 10 (2002) 2941–2952
2947
Inc., Vancouver, Canada). In our studies, bone marrow
cells were taken from the femurs of healthy mice that
had not been treated with drug. The cells were then
incubated in the MethocultTM media (StemCell, Inc.),
with continuous exposure to the drugs, where equal
concentrations (0.0084 mM, IC50 for compound 17 for
L1210 cells) were used for each of the compounds. On
day 12, the colonies were counted. The formation of
colonies indicates that the stem cells retain their activity
and thus stimulate the formation of progenitor cells.
The absence of colonies either indicates that exposure to
the drug solutions completely kills the bone marrow
cells or makes these cells non-viable. The results from
this study indicate that while compound 17 and 4 are
toxic (no colonies were observed in duplicated experi-
ments), the seco-CFQ analogue 5 displayed significantly
less toxicity (45ꢂ10 colonies compared to 60ꢂ3 colo-
nies for untreated control). Overall, the favorable activ-
ity against tumor cell lines and low toxicity to murine
bone marrow cells of the seco-CFQ pharmacophore to
murine bone marrow cells makes it a worthy candidate
for further drug development studies.
to reflux overnight, then the solvent was removed in
vacuo. The brown residue was then treated with water
(200 mL) and acidified to pH 6 with 6 M HCl. The
solution was then extracted with ethyl acetate (4 ꢃ
100 mL), and the combined organic layers were dried
and concentrated under reduced pressure to produce a
brown oil. This crude material was purified on a silica
gel column. A gradient eluent system was used begin-
ning with 10% EtOAc/hexane and ending with 30%
EtOAc/hexane to produce phenol 8 as a yellow solid
(1.7 g, 8.23 mmol, 54%). Mp 100–104 ꢁC. TLC (2.5%
MeOH/CHCl3) Rf 0.44. IR (neat) 3323, 3139, 1652,
1600, 1430, 1380, 1323, 1241. 1H NMR (CDCl3) 7.78 (t,
1.0, 1H), 7.61 (d, 2.5, 1H), 7.33 (d, 1.0, 1H), 6.83 (dd,
1.0, 2.5, 1H), 5.44 (s br, 1H), 4.33 (q, 7.0, 2H), 1.35 (t,
7.0, 3H). EI–MS (m/z, rel. intensity) 206 (M+, 72), 191
(10), 178 (30), 161 (100). Accurate mass for C11H10O4:
calcd 206.0579, obsd 206.0574.
Ethyl 4-benzyloxybenzofuran-6-carboxylate (9). Com-
pound 8 (3.38 g, 16.4 mmol) was combined with K2CO3
(2.9 g, 23.3 mmol) and Bu4NI (20 mg), and to the mix-
ture was added dry DMF (35 mL). Benzyl bromide
(2.3 mL, 20.8 mmol) was added to this suspension. The
reaction mixture was stirred overnight under positive N2
pressure. The DMF was removed under vacuum
(0.1 mmHg, 60 ꢁC) using a kugelrohr apparatus and the
resulting residue was dissolved in CHCl3. The crude
product was purified using a silica gel column and
CHCl3 as the eluent to yield the product 9 as an orange
solid (5.04 g, 17.0 mmol, 95%). Mp 52–58 ꢁC. TLC
(CHCl3) Rf 0.59. IR (Neat) 3035, 1708, 1600, 1445,
Experimental
Ethyl 4-hydroxybenzofuran-6-carboxylate, (8). Potas-
sium t-butoxide (1 M in t-butanol) (18.0 mL, 0.018 mol)
was added to a solution of 2-furaldehyde (11.0 mL,
0.13 mol) and diethyl succinate (32.0 mL, 0.19 mol). The
resulting dark brown mixture was kept under a nitrogen
atmosphere and heated to reflux in an oil bath (150 ꢁC).
After 45 min, another aliquot of diethyl succinate
(32.0 mL) and potassium t-butoxide (18.0 mL) were
added. The solution was allowed to reflux overnight.
After cooling, the dark solution was chilled on an ice
bath and carefully acidified (6 M HCl) to pH 1. The
mixture was then immediately extracted with ether (3 ꢃ
150 mL). The combined ether layers were extracted with
saturated Na2CO3 (5 ꢃ 75 mL). The Na2CO3 layers
were combined and then slowly acidified to a pH of 1
(6 M HCl) and extracted with ether (4 ꢃ 100 mL). The
ether layers were combined, washed with brine
(100 mL), and then dried over sodium sulfate. The sol-
vent was removed in vacuo to yield a dark brown oil
(15.19 g, 73 mmol, 56%), which was used directly in the
next step.
1
1328, 1308, 1236. H NMR (CDCl3) 7.82 (t, 1.0, 1H),
7.59 (d, 2.5, 1H), 7.43 (d, 7.5, 2H), 7.40 (d, 1.0, 1H), 7.34
(t, 7.5, 2H), 7.28 (t, 7.5, 1H), 6.87 (dd, 1.0, 2.5, 1H), 5.18
(s, 2H), 4.33 (q, 7.0, 2H), 1.35 (t, 7.0, 3H). EI–MS (m/z
rel. intensity) 296 (M+, 15), 91 (100). Accurate mass for
C18H16O4: calcd 296.1049, obsd 296.1042
4-Benzyloxy-6-(N-(tert-butyloxycarbonyl)amino)benzo-
furan (10). Compound 9 (5.04 g, 17.0 mmol) was dis-
solved in EtOH (200 mL) and 10% NaOH (200 mL) was
added. The yellow solution was heated to reflux for 3 h.
The ethanol was removed in vacuo and the remaining
aqueous suspension was stirred in an ice bath and acid-
ified to pH 1 using 3 M HCl. The resulting yellow pre-
cipitate was filtered and washed with water. The
5-benzyloxybenzofuran-7-carboxylic acid was then
dried in a vacuum oven (70 ꢁC at 0.1 mmHg). The fil-
trate was extracted with CHCl3 (3 ꢃ 75 mL), and the
organic layers were combined and dried with sodium
sulfate. The solvent was removed in vacuo to produce
more of the carboxylic acid as a yellow solid. The total
yield was 4.2 g, 15.7 mmol, 93%. Mp 192–196 ꢁC. TLC
(CHCl3) Rf 0.05. IR (neat) 2985, 1677, 1595, 1426, 1405,
The viscous brown oil was dissolved in acetic anhydride
(100 mL) and anhydrous sodium acetate (5.3 g,
64.6 mmol) was added. The solution was kept under a
positive pressure of nitrogen and heated in an oil bath at
150 ꢁC overnight. The solvent was removed by vacuum
flash distillation at about 80–100 ꢁC. The resulting resi-
due was partitioned with Na2CO3 (200 mL) and ethyl
acetate (200 mL), and the aqueous layer was further
extracted with ethyl acetate (2 ꢃ 150 mL). The com-
bined organic layers were dried with anhydrous sodium
sulfate and concentrated to yield a dark brown oil
(8.0 g) which was immediately dissolved in ethanol
(200 mL). To the solution was added K2CO3 (12 g,
86.8 mmol). The resulting reaction mixture was heated
1
1374, 1313, 1256. H NMR (CDCl3) 7.90 (s, 1H), 7.64
(d, 2.5, 1H), 7.45 (s, 1H), 7.44 (d, 7.0, 2H), 7.35 (t, 7.0,
2H), 7.30 (t, 7.5, 1H), 6.90 (dd, 0.5, 2.0, 1H), 5.20 (s,
1H). EI–MS (m/z rel. intensity) 268 (M+, 30), 91 (100).
The above carboxylic acid (2.0 g, 7.46 mmol) was sus-
pended in dry t-butanol (over molecular sieves 3A), and
diphenylphosphorylazide (DPPA) (2.3 mL, 10.6 mmol)

2948
T. T. Howard et al. / Bioorg. Med. Chem. 10 (2002) 2941–2952
and dry triethylamine (1.5 mL,10.8 mmol) were added.
The resulting suspension was heated to reflux overnight
under a N2 atmosphere, at which time a clear light-
brown solution formed. The solution was concentrated
in vacuo. The resulting dark brown oil was purified
using a silica gel column with CHCl3 as the eluent to
provide product 10 as a yellow solid (2.52 g, 7.46 mmol,
100%). Mp 30–35 ꢁC. TLC (CHCl3) Rf 0.73. IR (neat)
3317, 3050, 1728, 1710, 1608, 1550, 1497, 1453, 1413,
viscous oils. cis isomer: (200 mg, 0.28 mmol, 45%). TLC
(10% EtOAc/hexane) Rf 0.38. IR (neat) 3033, 1701,
1630, 1594, 1528, 1488, 1457, 1404, 1368, 1328, 1302,
1
1253, 770. H NMR (CDCl3) 7.56 (s, 1H), 7.37 (d, 8.0,
1H), 7.32 (t, 7.5, 2H), 7.27 (t, 7.5, 1H), 6.89 (s, 1H), 6.51
(s, 1H), 5.93 (d, 7.0, 1H), 5.87 (q, 7.0, 1H), 5.13 (d, 11.0,
1H), 5.08 (d, 11.0, 1H), 4.41 (dd, 5.5, 15.5, 1H), 4.26
(dd, 6.0, 15.5, 1H), 1.26 (s, 9H). EI–MS (m/z rel. inten-
sity) 491 (M+, 15), 356 (100). Accurate mass for
C23H23NO749Br35Cl: calcd 491.0499, obsd 491.0491.
trans isomer: (200 mg, 0.28 mmol, 35%). TLC (10%
EtOAc/hexane) Rf 0.35. IR (Neat) 3122, 3069, 3025,
1
1368, 1311, 1239. H NMR (CDCl3) 7.40 (d, 8.0, 2H),
7.32 (t, 8.0, 2H), 7.27 (t, 8.0, 1H), 7.20 (s, 1H), 7.17 (d,
1.0, 1H), 6.76 (s, 1H), 6.74 (dd, 1.0, 2.5, 1H), 6.44 (s br,
1H), 5.11 (s, 2H), 1.46 (s, 9H). EI–MS (m/z rel. inten-
sity) 339 (M+, 20), 283 (42), 91 (100). Accurate mass for
C20H21NO4: calcd 339.1471, obsd 339.1464.
1
1701, 1590, 1484, 1453, 1404, 1364, 1320, 1258, 770. H
NMR (CDCl3) 7.57 (s, 1H), 7.38 (d, 6.5, 2H), 7.33 (t,
7.5, 2H), 7.27 (t, 7.0, 1H), 6.91 (s, 1H), 6.45 (s, 1H), 5.94
(m, 1H), 5.87 (d, 13.5), 5.13 (d, 11.0, 1H), 5.08 (d, 11.0,
1H), 4.30 (d, 6.5, 15, 1H), 4.27 (dd, 7.0, 15, 1H), 1.25 (s,
9H). EI–MS (m/z rel. intensity) 491 (M+, 12), 356 (100).
4-Benzyloxy-7-bromo-6-[N-(3-chloro-2-propenyl)-N-(tert-
butyloxycarbonyl)amino]benzofuran (11). Compound 10
(500 mg, 1.47 mmol) was dissolved in a 1:1 solution of
dry THF/MeOH (30 mL) and stirred under N2 at
ꢀ78 ꢁC. A solution of p-toluenesulfonic acid (20 mg,
0.105 mmol, dissolved in 1 mL of dry THF) and N-bromo-
succinimide (287 mg, 1.61 mmol, dissolved in 5 mL of
dry THF) were slowly added to the solution. The solu-
tion was allowed to stir at ꢀ78 ꢁC for about 2 h, then
5% NaHCO3 (8 mL) was added at ꢀ78 ꢁC. The reaction
mixture was partitioned with EtOAc (50 mL) and water
(50 mL). The aqueous layer was further extracted with
EtOAc (2 ꢃ 30 mL). The combined organic layers were
washed with brine, dried with sodium sulfate, and con-
centrated in vacuo on a rotary evaporator. The resulting
viscous yellow oil was purified on a silica gel column
that was pretreated with 0.5% triethylamine/hexane and
followed by a hexane wash. The product was eluted
with a gradient EtOAc/hexane solvent, beginning with
1.0% EtOAc and gradual increase of 0.5% EtOAc every
100 mL. The bromide intermediate was isolated as a
viscous clear oil which solidified upon refrigeration
(340 mg, 813 mmol, 55%). TLC (10% EtOAc/hexane)
Rf 0.47. Mp 86–90 ꢁC. IR (neat) 3415, 3362, 3069, 3033,
1736, 1652, 1599, 1581, 1534, 1481, 1454, 1410, 1374,
5-Benzyloxy-3-(tert-butoxycarbonyl)-1-chloromethyl-1,2-
dihydro-3H-furano[2,3-e]indole (12) and 6-benzyloxy-4-
(tert-butoxycarbonyl)-2-chloro-1,2,3,4-tetrahydro-4H-
furano[2,3-f]quinoline (13). To a mixture of cis and
trans 11 (1.37 g, 2.59 mmol) was added 2,20-azobisiso-
butyronitrile (AIBN) (250 mg, 1.52 mmol). The reagents
were then dissolved in freshly distilled toluene (140 mL),
and nBu3SnH (1.87 mL, 6.95 mmol) was added. The
solution was degassed with nitrogen and heated in an oil
bath (90 ꢁC) for 2 h. The toluene was removed in vacuo
to produce a light-yellow oil which was dissolved in
CH2Cl2 and loaded on a hexane/silica gel column. The
products were eluted with 1% EtOAc/hexane with the
composition of EtOAc gradually increased to 5%.
Fractions containing the products were concentrated in
vacuo. The expected five-membered ring product 12
(seco-iso-CFI) was obtained as a colorless clear oil
(430 mg, 1.03 mmol, 40%). TLC (10% EtOAc/hexane)
Rf 0.52. IR (neat) 3070, 3024, 1701, 1652, 1634, 1617,
1559, 1541, 1493, 1368, 1342, 1257, 788. 1H NMR
(CDCl3) 7.57 (s br, 1H), 7.41 (d br, 8.0, 2H), 7.35 (d,
2.5, 1H), 7.33 (t, 8.0, 2H), 7.28 (t, 8.0, 1H), 6.77 (d, 2.5,
1H), 5.13 (s, 2H), 4.15 (dd, 9.5, 10.0, 1H), 4.02 (dd, 4.0,
11.0, 1H), 3.98 (m, 1H), 3.92 (dt, 4.0, 10.0, 1H), 3.54 (t,
10.5, 1H), 1.51 (s, 9H). EI–MS (m/z rel. intensity) 413
(M+, 14). Accurate mass for C23H24NO345Cl: calcd
413.1394, obsd 413.1381.
1
1330, 1268, 1237, 802. H NMR (CDCl3) 7.77 (s, 1H),
7.45 (d, 2.0, 1H), 7.42 (d, 7.5, 1H), 7.33 (t, 7.5, 2H), 7.28
(t, 7.5, 1H), 6.96 (s br, 1H), 6.82 (d, 1.5, 1H), 5.13 (s,
2H), 1.97 (s, 9H). FAB-MS (m/z rel. intensity) 419
(M++ 2, 5), 417 (M+, 5). Accurate mass for
C20H20NO749Br: calcd 417.0576, obsd 417.0574.
In addition to the seco-iso-CFI product 12, an unex-
pected 6-membered ring substance 13 (seco-CFQ) was
also isolated as a colorless clear oil (513 mg, 1.24 mmol,
48%). TLC (10% EtOAc/hexane) Rf 0.51. IR (neat)
3122, 3069, 1701, 1639, 1612, 1493, 1457, 1417, 1368,
To a chilled (ice bath) and freshly washed sample (with
hexanes) of NaH (134 mg of 60% suspension in mineral
oil; 80.4 mg, 3.35 mmol) was added a solution of the
above bromo compound (350 mg, 0.84 mmol) in DMF
(9 mL). After 45 min, the flask was removed from the ice
bath and 1,3-dichloropropene (0.23 mL, 2.51 mmol) was
added to the solution, which was allowed to stir over-
night under N2. The DMF was removed under vacuum
(0.1 mmHg, 60 ꢁC) using a kugelrohr apparatus to pro-
duce a brown oil which was dissolved in CH2Cl2 and
purified on an silica gel column. A gradient solvent sys-
tem was used to elute the product, beginning with 1%
EtOAc/hexane and ending with 5% EtOAc/hexane. The
fractions containing either the cis or trans isomer were
combined and concentrated in vacuo to produce clear
1
1342, 1258, 1222, 756. H NMR (CDCl3) 7.57 (s br,
1H), 7.41 (d br, 7.5, 2H), 7.36 (d, 2.0, 1H), 7.33 (t, 7.5,
2H), 7.27 (t, 7.5, 1H), 6.76 (d, 2.0, 1H), 5.13 (s, 2H), 4.17
(t, 9.5, 2H), 3.5.65 (m, 3H), 1.49 (s, 9H). EI–MS (m/z
rel. intensity) 413 (M+, 4), 379 (M++ H-Cl, 20).
Accurate mass for C23H24NO345Cl: calcd 413.1394, obsd
413.1384.
Note: The radical cyclization of pure cis or trans isomer
of compound 11 gave similar mixtures of products 12
and 13.

T. T. Howard et al. / Bioorg. Med. Chem. 10 (2002) 2941–2952
2949
5-Benzyloxy-1-(chloromethyl)- 1,2-dihydro-3-[(5,6,7-tri-
methoxy-1H-indol-2-yl)carbonyl]-3H-furano[2,3-e]indo-
line (14). Compound 12 (178 mg, 0.430 mmol) was
dissolved in dry THF (3.2 mL) and chilled in an ice
bath. Concentrated HCl (1.3 mL) was slowly added to
the solution. After 15 min the reaction flask was
removed from the ice bath and allowed to stir at room
temperature under N2 for 4 h. The reaction was quen-
ched with 2.5 M Na2CO3 (25 mL) and saturated
NaHCO3 (10 mL), followed by an extraction with
EtOAc (2 ꢃ 75 mL). The organic layer was dried with
sodium sulfate and concentrated in vacuo to produce
the amine as a brown oil. To the amine was added
EDCI (247 mg, 1.29 mmol) and 5,6,7-trimethoxyindole-
2-carboxylix acid (130 mg, 0.51 mmol). The reaction
mixture was suspended in DMF (8.5 mL) and stirred
under N2 for 3 days. The DMF was removed in vacuo
and the resulting residue was dissolved in CH2Cl2 and
loaded on a hexane/silica gel column. The product was
eluted with 15% EtOAc/hexane with the percentage of
EtOAc slowly increased to 35%. Fractions containing
product were concentrated in vacuo to produce com-
pound 14 as an off white solid (150 mg, 64%). Mp 170–
174 ꢁC. TLC (30% EtOAc/hexane) Rf 0.45. IR (neat)
3433, 3050, 1639, 1621, 1590, 1528, 1460, 1413, 1382,
(0.135 g, 0.296 mmol, 100%). Mp 250 ꢁC (dec). TLC
(2.5% MeOH/CHCl3) Rf 0.27; (30% EtOAc/hexanes)
Rf 0.13. IR (neat) 3433, 3090, 1630, 1595, 1523, 1426,
1
1377, 1386, 1220, 739. H NMR (CDCl3) 9.42 (s br,
1H), 7.99 (s, 1H), 7.92 (s br, 1H), 7.52 (d, 2.0, 1 h), 7.08
(d, 2.0, 1H), 6.88 (s, 1H), 6.80 (d, 2.0, 1H), 4.76 (t, 10.0,
1H), 4.59 (dd, 4.5, 10.0, 1H), 4.18 (dd, 3.5, 11.0, 1H),
4.13 (s, 3H), 4.11 (d, 5.0, 1H), 3.96 (s, 3H), 3.92 (s, 3H),
3.67 (t, 10.0, 1H). Uv (ethanol) 226 (e 8.0 ꢃ 107).
EI–MS (m/z, rel. intensity) 456 (M+, 20), 422 (M++
H–Cl, 40). Accurate mass for C23H21N2O365Cl: calcd
456.1088, obsd 456.1091.
2-Chloro-6-hydroxy-4-[(5,6,7-trimethoxy-1H-indol-2-yl)-
carbonyl]-1,2,3,4-tetrahydro-4H-furano[2,3-f ]quinoline
(5). The synthesis of compound 5 followed a procedure
similar to that used in the synthesis of compound 4,
except compound 15 (169 mg, 0.309 mmol) was used as
a starting material. Compound 5 was isolated as a
creamy white powder (0.131 g, 0.287 mmol, 93%). Mp
260–264 ꢁC. TLC (30% EtOAc/hexanes) Rf 0.15. (Neat)
3433, 3123, 3008, 1621, 1586, 1523, 1431, 1382, 1311,
1258, 1222, 797. 1H NMR (CDCl3) 9.38 (s, 1H), 7.95 (s,
1H), 7.46 (d, 2.0, 1H), 6.91 (s, 1H), 6.89 (d, 2.5, 1H),
6.80 (d, 2.0, 1H), 6.79 (s, 1H), 4.72 (t, 9.5, 1H), 4.11 (dd,
6.0, 9.5, 1H), 4.07 (s, 3H), 4.05 (m, 1H), 3.93 (m, 1H),
3.89 (s, 3H), 3.85 (s, 3H), 3.82 (m, 1H). Uv (ethanol) 222
(e 7.4 ꢃ 107). EI–MS (m/z rel. intensity) 456 (M+, 4),
422 (M++ H–Cl, 64). Accurate mass for
C23H21N2O365Cl: calcd 456.1088, obsd 456.1088.
1
1226, 694. H NMR (CDCl3) 9.31 (s br, 1H), 8.01 (s,
1H), 7.443 (d, 7.0, 2H), 7.445 (d, 2.0, 1H), 7.35 (t, 7.0,
2H), 7.28 (t, 7.0, 1H), 6.92 (d, 2.0, 1H), 6.85 (d, 2.0, 1H),
6.82 (s, 1H), 5.19 (s, 2H), 4.71 (t, 10.0, 1H), 4.53 (dd,
4.5, 10.0, 1H), 4.10 (m, 1H), 4.02 (s, 3H), 4.01 (d, 4.0,
1H), 3.88 (s, 3H), 3.85 (s, 3H), 3.61 (t, 9.8, 1H). Uv
(ethanol) 222 (e 7.8 ꢃ 107). EI–MS (m/z rel. intensity)
546 (M+, 44), 512 (M++ H–Cl, 58). Accurate mass for
C30H27N2O365Cl: calcd 546.1558, obsd 546.1560.
1-(Chloromethyl)-5-hydroxy-1,2-dihydro-3-[(5,6,7-tri-
methoxy-1H-indol-2-yl)carbonyl]-3H-5-methylfurano[2,3-
e]indoline (6). Compound 6 was prepared following a
similar scheme used in the synthesis of compound 4,
except 5-methylfurfural was used as a starting material.
Compound 6 was isolated as a white powder (0.136 g,
0.289 mmol, 99% in the last hydrogenation step). TLC
(5% MeOH/CHCl3) Rf 0.50. Mp 263–267 ꢁC. IR (neat)
3433, 3100, 1963, 1927, 1634, 1586, 1524, 1457, 1391,
6-Benzyloxy-2-chloro-4-[(5,6,7-trimethoxy-1H-indol-2-yl)-
carbonyl]-1,2,3,4-tetrahydro-4H-furano[2,3-f ]quinoline
(15). The synthesis of compound 15 followed a proce-
dure similar to that used in the synthesis of compound
14, except compound 13 (168 mg, 0.41 mmol) was used.
The desired product 15 was isolated as an off white solid
(178 mg, 80%). Mp 186–190 ꢁC. TLC (30% EtOAc/
hexane) Rf 0.53. IR (neat) 3459, 3080, 1621, 1586, 1524,
1
1311, 1262, 1220, 800. H NMR (CDCl3) 9.45 (s, 1H),
9.42 (s, 1H), 7.82 (s, 1H), 7.64 (s, 1H), 6.98 (d, 2.0, 1H),
6.89 (s, 1H), 6.46 (d, 2.0, 1H), 4.75 (t, 10.0, 1H), 4.59
(dd, 4.5, 10.0, 1H), 4.19 (dd, 3.5, 11.0, 1H), 4.11 (s, 3H),
3.96 (s, 3H), 3.93 (s, 3H), 3.64 (t, 10.0, 1H), 2.46 (s, 3H).
FAB-MS (NBA) (m/z rel. intensity) 471 (M+H, 4).
Accurate mass (FAB-NBA) for C24H24N2O365Cl : calcd
471.1323, obsd 471.1307.
1
1413, 1377, 1311, 1222, 739. H NMR (CDCl3) 9.32 (s
br, 1H), 7.99 (s, 1H), 7.449 (d, 6.5, 2 h), 7.448 (d, 2.0,
1H), 7.35 (t, 6.5, 2H), 7.28 (t, 6.5, 1H), 6.86 (d, 2.0, 1H),
6.84 (d, 2.5, 1H), 6.79 (s, 1H), 5.19 (s, 2H), 4.72 (t, 10.0,
1H), 4.10 (dd, 6.0, 9.5, 1H), 4.01 (s, 3H), 4.00 (m, 1H),
3.88 (s, 3H), 3.92 (m, 1H), 3.85 (m, 1H), 3.84 (s, 3H).
UV (ethanol) 222 (e 6.3 ꢃ 107). EI–MS (m/z rel. inten-
sity) 546 (M+, 12) 512 (M++H–Cl, 60). Accurate mass
for C30H27N2O365Cl: calcd 546.1558, obsd 546.1562.
2-Chloro-6-hydroxy-4-[(5,6,7-trimethoxy-1H-indol-2-yl)-
carbonyl]-1,2,3,4-tetrahydro-4H-5-methylfurano[2,3-f ]-
quinoline (7). Compound 7 was prepared following a
similar scheme used in the synthesis of compound 5,
except 5-methylfurfural was used as a starting material.
Compound 7 was isolated as a light-yellow powder
(167 mg, 0.356 mmol, 72% in the final hydrogenation
step). TLC (25% EtOAc/hexane) Rf 0.49. Mp 258–
260 ꢁC. IR (neat) 3424, 1683, 1621, 1586, 1523, 1493,
1-(Chloromethyl)-5-hydroxy-1,2-dihydro-3-[(5,6,7-tri-
methoxy-1H-indol-2-yl)carbonyl]-3H-furano[2,3-e]indo-
line (4). To a mixture of 10% Pd/C (52 mg) and
compound 14 (152 mg, 0.278 mmol) was added in 25%
aqueous NH4HCO2 (1.27 mL) and ice-cold THF
(7.5 mL). The suspension was sealed with a rubber sep-
tum and flushed with H2 then stirred for 5 h at ambient
temperature. The reaction mixture was filtered over
Celite. The resulting clear solution was concentrated in
vacuo to produce compound 4 as a white powder
1
1382, 1306, 1217. H NMR (CDCl3) 9.41 (s, 1H), 7.97
(s, 1H), 6.99 (s br, 1H), 6.94 (s, 1H), 6.86 (s, 1H), 6.45 (s,
1H), 4.76 (t, 10.0, 1H), 4.14 (dd, 6.0, 9.5, 1H), 4.13 (s,
3H), 3.96 (s, 3H), 3.92 (s, 3H), 3.87 (m, 3H), 2.47 (s,
3H). EI–MS m/z (rel. intensity) 470 (M+,5), 436 (55),

2950
T. T. Howard et al. / Bioorg. Med. Chem. 10 (2002) 2941–2952
203 (100). Accurate mass (EI–MS) for C24H23N2O365Cl:
calcd 470.1245, obsd 470.1250.
for 72 h at 37 ꢁC in a 5% CO2 atmosphere. MTT [3-(4,5-
dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide]
was dissolved in PBS (5 mg/mL). After the indicated cell
incubation period, 10 mL stock MTT solution was
added to each well, and the plates were further incu-
bated for 4 h at 37 ꢁC in a 5% CO2 atmosphere. After
this final incubation, 100 mL acidic isopropanol solution
(16 mL 12.1 N HCl in 5 mL isopropanol) was added to
each well. The contents of the wells were mixed thor-
oughly by pipetting the cell suspension up and down
and the plates were allowed to sit at room temperature
for 15 min in order to allow for full development of the
purple color. The plates were then read on a Dynatech
Plate Reader, utilizing Dynex Revelation 3.2 software,
with a test wavelenth of 570 nm and a reference wave-
length of 630 nm. The dose inhibiting the growth by
50% (IC50) was extrapolated from curves generated
based on the averages of the absorbance data (4 points/
concentration).
Note: Complete detailed procedures and characteri-
zation for compounds 6 and 7 can be obtained from the
authors.
N-BOC-iso-CFI (16). To a flame dried 25 mL round
bottom flask was added NaH (11.5 mg, 0.477 mmol),
which was washed with dry hexane (2 mL). A solution
of the phenol generate from catalytic hydrogenation of
compound 12 (51 mg, 0.16 mmol) was dissolved in dry
THF (2 mL) and added to the reaction flask, which was
kept in an ice bath. The reaction mixture was stirred for
3 h, at which time the THF was removed in vacuo
resulting in a brown solid which was dissolved in water
(20 mL) and EtOAc (20 mL). The water layer was
extracted with EtOAc (3 ꢃ 20 mL) and the organic lay-
ers were collected and concentrated to produce a brown
oil, which was purified by TLC (1% EtOAc/hexane).
The product was isolated as a clear oily residue (14 mg,
0.037 mmol, 23%). IR (neat) 3020, 1662, 1645, 1483,
1458, 1437, 1391, 1366, 1262. UV–vis (MeOH/water,
pH 3.0) 325 (e 222), 272 (e 833), 242 (e 1477). 1H NMR
(CDCl3) 6.98 (d, 1.5, 1H), 6.55 (d, 1.5, H), 6.54 (s, 1H),
3.75 (m, 1H), 3.71 (t, 7.5, 1H), 3.47 (m, 1H), 1.54 (s,
2H), 1.42 (s, 9H). FABMS (NBA) m/z (rel. intensity)
288 (M+H, 10).
Hematopoietic progenitor assay
For direct incubation of the drug solutions with the
bone marrow, the concentration of drug solution was
chosen to be 0.0084 mM (the IC50 value for the seco-
CBI-TMI compound 17 against L1210 cells). Bone
marrow cells were harvested from the femurs of healthy
DBA mice by flushing the femur with IMEM media
(ꢄ1 mL). These cells were counted and diluted to a
concentration of 2.0 ꢃ 105 cells/mL in additional IMEM
media. The drug solutions were prepared by making a
2.52 ꢃ 10ꢀ2 M stock solution of the drug in DMSO. In
order to make duplicate cultures, 0.3 mL of cells in
IMEM were added to 3.0 mL of the MethocultTM
M3434 media (StemCell Technologies) in a 15 mL con-
ical tube. The tube was then vortexed to ensure com-
plete mixing of the cells and media. Then, 11 mL of the
drug stock solution was added to the tube and thor-
oughly mixed via vortexing. This cell suspension (1 mL)
was then transferred (via 16G needle) into the wells of a
six-well plate. The plates were incubated for 12 days at
37 ꢁC and 5% CO2. The colonies were counted using a
dissecting microscope with a blue filter film attached to
add contrast to the cells.
Cytotoxicity studies
Compounds 4–7 and 17 were dissolved in DMSO to
obtain 1.75 ꢃ 10ꢀ2 M stock solutions, which were dilu-
ted with DMSO to prepare standard solutions from
concentrations of 1.75 ꢃ 10ꢀ3 to 1.75 ꢃ 10ꢀ10 M. These
solutions were then further diluted with DMEM (24 mL
standard solution in 176 mL media). Addition of these
new drug solutions (5 mL) to wells containing 100 mL
media/cell suspension resulted in final drug concen-
trations ranging from 1.0 ꢃ 10ꢀ12 to 1.0 ꢃ 10ꢀ4 M.
The murine mastocytoma P815 and murine leukemia
L1210 cell lines were obtained from American Type
Tissue Culture Collection (ATCC). The murine mela-
noma B16-F0 cell line was obtained from the Cancer
Center of the Greenville Hospital System. The cell lines
were grown in Delbecco’s Modified Eagle Medium
(DMEM, Atlanta Biochemicals) supplemented with
10% fetal bovine serum, Hepes Buffer (2 mM, Media-
tech Cellgro, 25–060-Cl), l-glutamine (2 mM, Media-
tech Cellgro), and penicillin/streptomycin (50,000 units
penicillin, 50,000 mg streptomycin, Atlanta Biologicals).
Cells were maintained at 37 ꢁC in a 5% humidified CO2
atmosphere.
Flow cytometry studies
The P815 cell line was used for this experimental proce-
dure. The cells were incubated with drugs in a final
concentration equaling their IC50 value. The IC50 con-
centrations for cisplatin and compound 4 were 100 and
0.061 mM, respectively. Cisplatin was used as a positive
control. The drug stock solutions were made in DMSO
so that the addition of 5 mL of the solution into 5 mL of
media would result in the desired concentrations listed
above. Cells were counted and re-suspended at a con-
centration of 1.5 ꢃ 105 cells/mL, and 5 mL of this cell
suspension were pipetted into a small culture flask. The
drug solutions were added (5 mL), and the cells were
incubated at 37 ꢁC and 5% CO2 for 24 h. The cells were
then harvested via centrifugation (1000 rpm, 10 min,
4 ꢁC) and washed twice with sample buffer (0.5 g glucose
in 500 mL Ca2+/Mg2+ free PBS; Electron Microscopy
Cultured cells were counted using a hemocytometer and
were resuspended in fresh DMEM at a concentration of
8 ꢃ 104 cells/mL. The cell suspension (100 mL) was added
to 96-well flat-bottom cell culture plates. At this con-
centration, 8000 cells were seeded in each well. The drug
solutions diluted in DMEM (detailed above) were added to
each well (5 mL/well). Quadruplicate wells were prepared
for each drug concentration. The plates were incubated

T. T. Howard et al. / Bioorg. Med. Chem. 10 (2002) 2941–2952
2951
Sciences, Fort Washington, PA, USA). To fix the cells,
the cells were vortexed vigorously for ꢄ10 s, after which
the vortexing was continued at a slower rate during the
dropwise addition of 1 mL of ice-cold 70% ethanol. The
tubes were capped and stored at 4 ꢁC until the day of the
flow cytometry analysis.
the prominent sites of damage on the bottom strand was
PCR amplified, using the pUC1 primer and the syn-
thetic primer 50-TGGTATCTTTATAGTCCTGTCG-30,
50-end labeled and binding on the complementary
(upper) strand at positions 956–935. The 166 base pairs
singly end-labeled fragment generated, was purified by
agarose gel electrophoresis and isolated using a Bio101
kit according to the manufacturer’s instruction.
Approximately 1 h before the desired time of analysis,
the cells were stained with propidium iodide (PI). The
PI staining solution was prepared by the addition of
0.6 mL of the PI stock solution (made from 0.8 mg of PI
with 0.8 mL of water) to 12 mL of sample buffer. RNase
A (EC 3.2.27.5; Sigma) was added (1200 Kunitz units,
ꢄ17 mg). This solution was then thoroughly mixed
before it (1 mL) was added to the cells, which were pel-
leted by centrifugation at 3000 rpm for 5 min and at
4 ꢁC. The cells were kept on ice until they were analyzed
by flow cytometry. The analysis was completed on a
Becton Dickinson FACScan.
The dry DNA pellets from the drug–DNA incubations
were re-suspended in sodium citrate buffer (pH 7.2) and
heated to 90 ꢁC for 30 min to thermally cleave at sites of
adenine– or guanine–N3 lesions as described by Rey-
nolds.24 Samples were chilled, precipitated and dried.
Acrylamide gel electrophoresis
Samples were dissolved in formamide loading dye, heat-
denatured for 3 min at 95 ꢁC, cooled on ice and electro-
phoresed at 2000 V for ꢅ2 h on a 6% acrylamide dena-
turing gel (Sequagel, National Diagnostics). The gels
were dried under vacuum at 80 ꢁC and exposed to film
(X-OMAT, Kodak). Densitometry was performed on a
Bio-Rad GS-670 imaging densitometer.
Drug–DNA reactions
All drug–DNA reactions were performed in 25 mM
triethanolamine, 1 mM EDTA, pH 7.2, at 37 ꢁC for 5 h.
Following incubation, DNA was precipitated by addi-
tion of 1/10 volume of 3 M sodium acetate and 3
volumes of 95% ethanol and washed with 70% ethanol.
The resulting pellet was dried by lyophilization.
Acknowledgements
Support (to M.L.) from the Dreyfus Foundation,
Research Corporation, National Science Foundation-
REU, Enzacta Ltd. (UK), and the National Cancer
Institute are gratefully acknowledged. J.A.H. thanks the
Cancer Research Campaign for support. The authors
thank Ms. Lillia Holmes at the Greenville Oncology
Research Center, Greenville, SC, for performing the
flow cytometry experiments. The authors also thank Dr.
Robert Kelly of Pharmacia & Upjohn for a generous
gift of adozelesin.
Taq polymerase stop assay
The procedure employed was previously described by
Ponti et al.18 Prior to drug/DNA incubation, plasmid
pUC18 was linearised with the Hind III enzyme, which
cuts at only one site in the plasmid (position 399), pro-
viding a stop for the Taq polymerase downstream from
the primer. The synthetic primer 50-CTCACT-
CAAAGGCGGTAATAC-30 binds to the com-
plementary (bottom) strand at positions 749–769 and
was used to examine the alkylation patterns on the bot-
tom strand. The oligodeoxynucleotide primer was 50 end
labeled prior to amplification using T4 polynucleotide
kinase and [g-32P]-ATP (5000 Ci/mmol, Amersham,
UK). The labeled primer was purified by elution
through Bio-Rad spin columns. Linear amplification of
DNA was carried out in a total volume of 100 mL con-
taining 0.5 mg of template DNA, 5 pmol of labeled pri-
mer, 250 mM of each dNTP, 10 U Taq polymerase,
10 mM Tris–HCl (pH 9.0), 50 mM KCl and 0.1% Tri-
ton1 X-100, 2.5 mM MgCl2 and 0.01% Gelatin.
References and Notes
1. (a) Li, L. H.; Wallace, T. L.; DeKoning, T. F.; Warpehoski,
M. A.; Kelly, R. C.; Prairie, M. D.; Drueger, W. C. Invest.
New Drugs 1987, 5, 329. (b) Li, L. H.; Kelly, R. C.; Warpe-
hoski, M. A.; McGovern, J. P.; Gebhard, I.; DeKoning, T. F.
Invest. New Drugs 1991, 9, 137. (c) Warpehoski, M. In
Advances in DNA Sequence Specific Agents. Hurley, L. H.,
Ed.; JAI: Greenwich, 1992: Vol. 1, p. 217. (d) Boger, D. L.;
Garbaccio, R. M. Acc. Chem. Res. 1999, 32, 1043. (e) Boger,
D. L.; Johnson, D. S. Angew. Chem., Int. Ed. Engl. 1996, 35,
1438. (f) Boger, D. L.; Johnson, D. S. Proc. Natl. Acad. Sci.
U.S.A. 1995, 92, 3642. (g) Boger, D. L. Acc. Chem. Res. 1995,
28, 20. (h) Boger, D. L. Chemtracts: Org. Chem. 1991, 4, 329.
2. (a) Hightower, R. D.; Sevin, B. U.; Pevras, J. P.; Untch,
M.; Angioli, R.; Averette, H. Gynecol. Oncol. 1992, 42, 186.
(b) Cote, S.; Momparler, R. L. Anti-Cancer Drugs 1993, 4,
327. (c) Fleming, G. F.; Ratain, S. M.; O’Brien, R. L.;
Schilsky, P. C.; Ramos, R.; Richards, J. M.; Vogelzang, N. J.;
Kasunic, D. A.; Earhart, R. H. J. Natl. Cancer Inst. 1994, 86,
368. (d) Foster, B. J.; Larusso, P. M.; Poplin, E.; Zalupski, M.;
Valdivieso, M.; Wozniak, A.; Flaherty, L.; Kasunic, D. A.;
Earhart, R. H.; Baker, L. H. Invest. New Drugs 1996, 13, 321.
(e) Burris, H. A.; Dieras, V. C.; Tunca, M.; Earhart, R. H.;
After an initial denaturation at 94 ꢁC for 4 min, the
cycling conditions were 94 ꢁC for 1 min, 58 ꢁC for 1 min,
74 ꢁC for 1 min+1 s per cycle, for a total of 30 cycles.
After amplification the samples were ethanol pre-
cipitated and washed with 70% ethanol.
Sequence selectivity of purine–N3 alkylation
The sites of covalent binding for the compounds tested,
as determined by the polymerase stop assay, were con-
firmed as minor groove purine N3 lesions using the ther-
mal cleavage assay. The region of the plasmid containing

2952
T. T. Howard et al. / Bioorg. Med. Chem. 10 (2002) 2941–2952
Earhart, J. R.; Rodriguez, G. L.; Shaffer, D. S.; Fields, S. M.;
Campbell, E.; Schaaf, L.; Kasunic, D.; Von Hoff, D. D. Anti-
Cancer Drugs 1997, 8, 588.
11. (a) Boger, D. L.; Goldberg, J.; McKie, J. A. Bioorg. Med.
Chem. Lett. 1996, 6, 1955. (b) Boger, D. L.; Mesini, P.; Tarby,
C. M. J. Am. Chem. Soc. 1994, 116, 6461.
3. (a) Li, L. H.; Dekoning, J. F.; Kelly, R. C.; Krueger, W. C.;
McGovern, J. P.; Padbury, G. E.; Petzold, G. L.; Wallace,
T. L.; Ouding, R. J.; Praire, M. D.; Gebhard, I. Cancer Res.
1992, 52, 4904. (b) van Telligen, O.; Punt, C. J. A.; Awada, A.;
Wagener, D. J. T.; Piccart, M. J.; Groot, Y.; Schaaf, L. J.;
Henrar, R. E. C.; Nooijen, W. J.; Beijnen, J. H. Cancer Che-
mother. Pharmacol. 1998, 41, 366.
4. (a) Nagamura, S.; Kobayashi, E.; Gomi, K.; Saito, H.
Bioorg. Med. Chem. Lett. 1996, 6, 214, and references therein.
(b) Alberts, S. R.; Erlichman, C.; Reid, J. M.; Sloan, J. A.;
Ames, M. M.; Richardson, R. L.; Goldberg, R. M. Clin.
Cancer Res. 1998, 4, 2111.
5. (a) Walker, D. L.; Reid, F. M.; Ames, M. M. Cancer Che-
mother. Pharmacol. 1994, 34, 317. (b) Volpe, D. A.; Tomas-
zewski, J. E.; Parchment, R. E.; Garg, A.; Flora, K. P.;
Murphy, M. J.; Grieshaber, C. K. Cancer. Chemother. Phar-
macol. 1996, 39, 143. (c) Pitot IV, H. C.; Erlichman, C.; Reid,
J. M.; Sloan, J. A.; Ames, M. M.; Bagniewski, P. G.; Ather-
ton-Skaaf, P.; Adjei, A. A.; Rubi, J.; Rayson, D.; Goldberg,
R. M. Proc. Am. Assoc. Cancer Res. 1999, 40, 91.
6. (a) Wrasidlo, W.; Johnson, D. S.; Boger, D. L. Bioorg.
Med. Chem. Lett. 1994, 4, 631. (b) Wang, Y.; Yuan, H.; Ye,
W.; Wright, S. C.; Wang, H.; Larrick, J. W. J. Med. Chem.
2000, 43, 1541.
12. (a) Boger, D. L.; McKie, J. A.; Han, N.; Tarby, C. M.;
Riggs, H. W.; Kitos, P. A. Bioorg. Med. Chem. Lett. 1996, 6,
659. (b) Castedo, L.; Delamano, J.; Enjo, J.; Fernandez, J.;
Gravalos, D. G.; Leis, R.; Lopez, C.; Marcos, C. F.; Rios, A.;
Tojo, G. J. Am. Chem. Soc. 2001, 123, 5102.
13. Boger, D. L.; Han, N.; Tarby, C. M.; Boyce, C. W.; Cai,
H.; Jin, Q.; Kitos, P. A. J. Org. Chem. 1996, 61, 4894.
14. (a) Gomi, K.; Kobayashi, E.; Miyoshi, K.; Ashizawa, T.;
Okamoto, A.; Ogawa, T.; Katsumata, A.; Mihara, A.; Okabe,
M.; Hirata, T. Jpn. J. Cancer Res. 1992, 83, 113. (b) Ichimura,
M.; Ogawa, T.; Takahashi, K.; Mihara, A.; Takahashi, I.;
Nakano, H. Oncol. Res. 1993, 5, 165. (c) Yasuzawa, T.;
Muroi, K.; Ichimura, M.; Takahashi, I.; Ogawa, T.; Takaha-
shi, K.; Sano, H.; Saitoh, Y. Chem. Pharm. Bull. 1995, 43, 378.
15. (a) Boger, D. L.; McKie, J. A.; Boyce, C. W. Synlett 1997,
515. (b) Muratake, H.; Hayakawa, A.; Natsume, M. Tetra-
hedron Lett. 1997, 38, 7577. (c) Muratake, H.; Itsuko, A.;
Natsume, M. Tetrahedron Lett. 1994, 35, 2573. (d) Muratake,
H.; Matsumura, N.; Natsume, M. Chem. Pharm. Bull. 1998,
46, 559. (e) Arisoff, P. A.; Johnson, P. D. J. Org. Chem. 1992,
57, 6234. (f) Arisoff, P. A.; Johnson, P. D.; Sun, D.; Hurley,
L. H. J. Med. Chem. 1993, 36, 1956.
16. Ninomiya, K.; Shioiri, T.; Yamada, S. Tetrahedron 1974,
30, 2151.
7. (a) Boger, D. L.; McKie, J. A. J. Org. Chem. 1995, 60,
1271. (b) Boger, D. L.; Yun, W.; Han, N. Bioorg. Med. Chem.
1995, 3, 1429.
8. Baraldi, P. G.; Cacciari, B.; Pinedo de las Infantas, M. J.;
Romagnoli, R.; Spatullo, G.; Cozzi, P.; Mongelli, N. Anti-
Cancer Drug Des. 1997, 12, 67.
17. Boger, D. L.; Boyce, C. W.; Garbaccio, R. M.; Searcey,
M. Tetrahedron Lett. 1998, 39, 2227.
18. Ponti, M.; Forrow, S. M.; Souhami, R. L.; D’Incalci, M.;
Hartley, J. A. Nucl. Acids Res. 1991, 19, 2929.
19. Wyatt, M. D.; Lee, M.; Garbiras, B. J.; Souhami, R. L.;
Hartley, J. A. Biochem. 1995, 34, 13034.
9. (a) Mohamadi, F.; Spees, M. M.; Staten, G. S.; Marder, P.;
Kipka, J. K.; Johnson, D. A.; Boger, D. L.; Zarrinmayeh, H.
J. Med. Chem. 1994, 37, 232. (b) Patel, V. F.; Andis, S. L.;
Enkema, J. K.; Johnson, D. A.; Kennedy, J. H.; Mohamadi,
F.; Schultz, R. M.; Soose, D. J.; Spees, M. M. J. Org. Chem.
1997, 62, 8868.
10. (a) Boger, D. L.; Munk, S. A.; Zarrinmayeh, H.; Ishizaki,
T.; Haught, J.; Bina, M. Tetrahedron 1991, 47, 2661. (b)
Boger, D. L.; Ishizaki, T.; Zarrinmayeh, H.; Munk, S. A.;
Kitos, P. A.; Suntorwart, O. J. Am. Chem. Soc. 1990, 112,
8961.
20. (a) Boger, D. L.; Yun, W. J. Am. Chem. Soc. 1994, 116,
7996. (b) Atwell, G. J.; Wilson, W. R.; Denny, W. A. Bioorg.
Med. Chem. Lett. 1997, 7, 1493.
21. Lee, M.; Rhodes, A. L.; Wyatt, M. D.; D’Incalci, M.;
Forrow, S.; Hartley, J. A. J. Med. Chem. 1993, 36, 863.
22. Boyd, M. R. Principles Practices Oncol. 1989, 10, 1.
23. Pecere, T.; Gazzola, M. V.; Mucignat, C.; Parolin, C.;
Vecchia, F. D.; Cavaggioni, A.; Basso, G.; Diaspro, A.; Sal-
vato, B.; Carli, M.; Palu, G. Cancer Res. 2000, 60, 2800.
24. Reynolds, V. L.; Molineaux, I. J.; Kaplan, D. J.; Swenson,
D. H.; Hurley, L. H. Biochemistry 1985, 24, 2929.