Unsymmetrical DNA cross-linking agents: Combination of the CBI and PBD pharmacophores

Article abstract of DOI:10.1021/jm020526p

A set of 10 compounds, each combining the seco-1,2,9,9a-tetrahydrocyclopropa[c]benz[e]indol-4-one (seco-CBI) and pyrrolo [2,1-c] [1,4] benzodiazepine (PBD) pharmacophores, was designed and prepared. These compounds were anticipated to cross-link between N3 of adenine and N2 of guanine in the minor groove of DNA. The compounds, which differ in the chain length separating the two alkylation subunits, and the configuration of the CBI portion, showed great variation in cellular toxicity (over 4 orders of magnitude in a cell line panel) with the most potent example exhibiting IC₅₀S in the pM range. Cytotoxicity correlated with the ability of the compounds to cross-link naked DNA. Cross-linking was also observed in living cells, at much lower concentrations than for a related symmetrical PBD dimer. A thermal cleavage assay was used to assess sequence selectivity, demonstrating that the CBI portion controlled the alkylation sites, while the PBD substituent increased the overall efficiency of alkylation. Several compounds were tested for in vivo activity using a tumor growth delay assay against WiDr human colon carcinoma xenografts, with one compound (the most cytotoxic and most efficient cross-linker) showing a statistically significant increase in survival time following a single iv dose.

Full text of DOI:10.1021/jm020526p

2132
J . Med. Chem. 2003, 46, 2132-2151
Un sym m etr ica l DNA Cr oss-Lin k in g Agen ts: Com bin a tion of th e CBI a n d P BD
P h a r m a cop h or es
Moana Tercel,* Stephen M. Stribbling, Hilary Sheppard, Bronwyn G. Siim, Kent Wu, Susan M. Pullen,
K. J ane Botting, William R. Wilson, and William A. Denny
Auckland Cancer Society Research Centre, Faculty of Medical and Health Sciences, University of Auckland,
Private Bag 92019, Auckland, New Zealand
Received November 20, 2002
A set of 10 compounds, each combining the seco-1,2,9,9a-tetrahydrocyclopropa[c]benz[e]indol-
4-one (seco-CBI) and pyrrolo[2,1-c][1,4]benzodiazepine (PBD) pharmacophores, was designed
and prepared. These compounds were anticipated to cross-link between N3 of adenine and N2
of guanine in the minor groove of DNA. The compounds, which differ in the chain length
separating the two alkylation subunits, and the configuration of the CBI portion, showed great
variation in cellular toxicity (over 4 orders of magnitude in a cell line panel) with the most
potent example exhibiting IC50s in the pM range. Cytotoxicity correlated with the ability of
the compounds to cross-link naked DNA. Cross-linking was also observed in living cells, at
much lower concentrations than for a related symmetrical PBD dimer. A thermal cleavage
assay was used to assess sequence selectivity, demonstrating that the CBI portion controlled
the alkylation sites, while the PBD substituent increased the overall efficiency of alkylation.
Several compounds were tested for in vivo activity using a tumor growth delay assay against
WiDr human colon carcinoma xenografts, with one compound (the most cytotoxic and most
efficient cross-linker) showing a statistically significant increase in survival time following a
single iv dose.
In tr od u ction
DNA cross-linking agents constitute an important
class of antitumor drugs.1 However, several of those
used in the clinic (e.g., nitrogen mustards, cisplatin,
mitomycin C) are compromised by serious side effects
associated with their poor selectivity for tumor cells.
Also, indiscriminate alkylation of multiple DNA sites,
or other cellular targets, makes it difficult to determine
their precise mechanism of action. Several attempts
have been made in recent years to construct new cross-
linking compounds by the dimerization of known and
more discriminating monoalkylating agents. Unusual
biological activities can be anticipated for those com-
pounds which cross-link at new sites (e.g., major or
minor groove, adenine or guanine bases), with different
selectivity or efficiency, or with various degrees of
distortion of the DNA structure. A further focus of
recent studies has been to increase the size of the dimer
recognition site. The hope is that enhanced sequence
selectivity might increase selectivity for tumor cells,2,3
with the ultimate goal of targeting predefined DNA
sites4 and thereby specifically inhibiting the expression
of those proteins critical for tumor cell proliferation.5
Two classes of minor groove monoalkylating agents
that have been used in these studies are the cyclopro-
paindolones6 and pyrrolo[2,1-c][1,4]benzodiazepines
(PBDs),7 each of which possess significant sequence
selectivity as monoalkylating agents. Cyclopropa-
adenines 6 or 7 base pairs (bps) apart in the minor
indolone dimers8 are exemplified by bizelesin 1. In this
groove, favoring AT-rich sequences such as 5′-TAATAA*-
compound two seco-CPI9 alkylating units derived from
terstrand cross-link between the N3 position of two
3′.11,12 Bizelesin is both more sequence selective and
the natural product CC-1065 are linked by a rigid bis-
cytotoxic than the monomeric agents from which it is
(indolyl)urea.10 Bizelesin very efficiently forms an in-
derived. Excellent preclinical activity was observed in
mice,13 and bizelesin advanced to clinical trial,14 al-
* To whom correspondence should be addressed. Phone:
64 9 3737 599. Fax: 64 9 3737 502. E-mail: m.tercel@auckland.ac.nz.
though significant myelotoxicity has been reported15 (in
10.1021/jm020526p CCC: $25.00 © 2003 American Chemical Society
Published on Web 04/25/2003

Unsymmetrical DNA Cross-Linking Agents
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2133
Sch em e 1. Target Compounds and Retrosynthesis
Analysis
have combined the CBI and PBD units to construct a
new class of unsymmetrical minor groove cross-linking
agents. These were anticipated to alkylate between N3
of adenine and N2 of guanine, a previously unknown
DNA lesion. During the course of this work a report
appeared describing the synthesis of a single CPI-PBD
dimer 5, designed and demonstrated to cross-link the 6
bp sequence 5′-CAATTA*-3′.27 Herein we provide full
details of our own work in this area.
Syn th esis
The target compounds for synthesis in the current
study are represented by the general structure high-
lighted in Scheme 1, i.e., minimal CBI and PBD phar-
macophores joined by the simplest possible linker, a
polymethylene chain. A set of 10 compounds was envis-
aged in which the linking chain contained between one
and five carbon atoms (n ) 1-5, to span a range which
should include matches to the bp register) and, for each
chain length, both enantiomers of the CBI subunit. The
latter was considered important since it is known that
the cyclopropaindolone enantiomeric form affects se-
quence selectivity (and toxicity)28,29 and, depending on
the binding preferences of the PBD component of the
dimers, could affect the balance of inter- versus intras-
trand cross-linking.
The most sensitive functional group in the target
molecules was anticipated to be the PBD imine, which
is very reactive to nucleophilic addition. The formation
of the imine was therefore planned as the final step,
from a suitably protected carbamate (P 1 in Scheme 1)
of the corresponding hemiaminal. This P 1 group must
be removed under conditions which are not strongly
basic, reductive, or nucleophilicsand the Aloc protecting
group, used by others in PBD syntheses25,30 was con-
sidered a suitable candidate. Retrosynthetic disconnec-
tion at the amide bond gives two fragments: the pair
of resolved Boc-protected CBI enantiomers 6R and 6S,
and a set of variable chain length PBD esters. The
former was prepared by the known route, in six steps
from 1,3-dihydroxynaphthalene20,31 and resolved by
semipreparative chiral HPLC, also as previously de-
scribed.29 For the latter, an ester protecting group P 2
was needed that was orthogonal to P 1 and also did not
require basic conditions for cleavage, as these would be
expected to racemize the crucial C11a position.32
The synthesis of the requisite PBD acids is illustrated
in Scheme 2. Vanillin was converted in three steps as
described to the nitrobenzoic acid 7,33 which was coupled
with commercially available (S)-2-pyrrolidinemethanol
to give the known amide 8.34 Hydrogenation over Pd/C
removed the benzyl protecting group and reduced the
nitro substituent to the aniline. The latter was selec-
tively reacted with allyl chloroformate at low temper-
ature to introduce the Aloc group in 10, while leaving
the phenol unaffected (reaction at room temperature
gave ca. 10% of the bis-Aloc product).
common with several cyclopropaindolone monomers16).
Bizelesin itself was a development of earlier flexibly
linked seco-CPI dimers 2 in which cross-linking ability
and cytotoxicity were found to be related to the number
of carbons in the linking polymethylene chain.17 Other
variants of cyclopropaindolone dimers have since been
reported,18 including extension to the seco-CBI (seco-
1,2,9,9a-tetrahydrocyclopropa[c]benz[e]indol-4-one, see
Scheme 1) alkylation subunit,19 which has the advan-
tages of being more synthetically accessible, more stable,
and more cytotoxic than it’s CPI analogue.20
An example of a recently developed PBD dimer is
DSB-120 3,21 formed by linking two molecules of the
PBD monomer DC-81 (see Scheme 5). DSB-120 was one
example selected from several dimers which differ in
the length of the connecting chain.22 Dimerization
increases the ability of these compounds to raise the
melting temperature of DNA, significantly enhances
their cytotoxicity, and extends the sequence selectivity
beyond the 5′-PuG*Pu-3′ motif favored by the mono-
mers. Molecular modeling and NMR studies of DSB-
120 demonstrated minor groove interstrand cross-
linking between the exocyclic amino groups of two
guanines 4 bps apart, with the molecule spanning a 6
bp sequence.23 Although DSB-120 showed minimal
antitumor activity in a murine model,24 an analogue
with C2 exo-methylene substitution (SJ G-136, 4)25 has
very recently been selected for clinical development on
the basis of activity in the National Cancer Institute’s
In Vitro and Hollow Fiber Assays.26
The phenol 10 was then alkylated with a variety of
4-bromobutanoic esters (i.e., n ) 3 chain length) in test
reactions to find an ester protecting group suitable for
the following steps. For example, reaction with the TCE
(trichloroethyl) ester 11c gave 12c, which was oxidized
with DMP (the Dess-Martin periodinane). The inter-
mediate aldehyde spontaneously cyclized to generate the
Given the success of this dimerization strategy to
provide interesting and biologically active molecules, we

2134 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Tercel et al.
Sch em e 2a
a
Conditions: (a) (COCl)2, DMF, then (S)-2-pyrrolidinemethanol, K2CO3; (b) H2, Pd/C; (c) allyl chloroformate, py, -78 °C; (d) K2CO3,
DMF, 20 °C; (e) DMP; (f) Zn powder, HCO2H; (g) CH2N2.
diazepine ring in 14c. Although this reaction was
accompanied by some overoxidation to the dione 13c,
requiring careful chromatographic separation, DMP
oxidation was found to be considerably higher yielding
than either Swern oxidation or the use of TPAP (tetra-
propylammonium perruthenate). Cleavage of the TCE
ester was accomplished using Zn/HCO2H,35 giving the
desired acid 15c in 89% yield. Not only was this higher
yielding than some alternatives (Zn/aq KH2PO436 or Zn/
aq NH4OAc37), it also gave the product with the highest
optical rotation, a value that was unchanged on re-
peated exposure to the reaction conditions. A portion of
15c was also converted to the methyl ester 16 and
examined by chiral HPLC: in comparison with a sample
obtained via base cleavage of the ester 14c, which
caused complete epimerization, 16 was clearly uncon-
taminated by racemization at C11a (see Supporting
Information).
Other possible ester protecting groups were less
successful: although both the TMSE (trimethylsilyl-
ethyl) ester and t-Bu ester analogues of 11c gave good
yields in the alkylation and oxidation steps, deprotection
of the ester resulted in decomposition (TMSE ester, HF/
py), racemization (TMSE ester, TBAF or CsF), or
considerably lower yields (t-Bu ester, HCl).
The TCE ester route was applied to the other chain
lengths (n ) 1, 2, 4, 5) and gave similar yields in all
steps, except for the propionate ester (n ) 2) for which
the phenol alkylation failed completely, due to reverse
Michael reaction. This necessitated an alternative syn-
thesis for the n ) 2 analogue, as shown in Scheme 3.
Vanillin was alkylated with 3-bromopropanol to give 17,
which underwent nitration using HNO3 to incorporate
the 5-NO2 group and also oxidize the side-chain alcohol
to the acid 19. This reaction is well precedented for
vanillic acid analogues and has been used in PBD
syntheses in the past.38 In this case the reaction was
not clean, and along with the desired product there was
also a significant amount of the nitrate 18, as well as
ipso substitution of the aldehyde to give the dinitro acid
20. The latter cocrystallized with 19, but was easily
separated following the later aldehyde oxidation. The
acid was protected as the TCE ester 21, the aldehyde
oxidized to the benzoic acid 23 (using sodium chlorite39),
and the latter coupled with (S)-2-pyrrolidinemethanol
to give 24. Reduction of the nitro group in the presence
of the TCE ester presented some problemsseventually,
brief hydrogenation over Pd/C was chosen ahead of some
reportedly more selective reagents (Ni2B,40 PtO2/H2) and
gave the desired product 25 in a workable 52% yield.
Steps as described above completed the synthesis of 15b
and furnished the complete set of five PBD acids.
The CBI and PBD units were linked together as
shown in Scheme 4. The Boc protecting group was
removed from the CBI with acid treatment, and the
resulting amines immediately reacted with the PBD
acids using EDCI. The final step, removal of the Aloc
protecting group using catalytic Pd(0), was not initially
successful and prompted some model reactions using an
equimolar mixture of the PBD ester 16 and the Boc-
protected CBI 6. It was found that weakly basic allyl
cation trapping reagents (e.g., dimedone41) were not able
to prevent significant allylation of the CBI phenol (nor
was p-methoxyphenol, tested as a more nucleophilic
phenol competitor), while stronger bases such as mor-
pholine or pyrrolidine caused competing spirocyclization
of the CBI. Eventually it was found that very brief
exposure to Pd(PPh3)4 and an excess of pyrrolidine (10
equiv, 1-5 min, 20 °C), followed by direct transfer of
the concentrated reaction mixture to the top of a

Unsymmetrical DNA Cross-Linking Agents
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2135
Sch em e 3a
aniline 28 was subjected to the same steps as above:
reaction with allyl chloroformate, DMP oxidation, and
cleavage of the Aloc protecting group to give the
intermediate 30, previously prepared by an alternative
route.42 The benzyl ether was removed, employing the
reported conditions42 to give 31 in moderate yield,
isolated for the first time as an analytically pure solid.
Interestingly, the measured optical rotation of 31 ([RD]
) +1239°) was considerably higher than the literature
values for any of the previous synthetic samples ([RD]
) +310°,42 +315°,43 +371°44).45 We note that the solvent
used in all cases was CHCl3, which commonly contains
1-2% EtOH as a stabilizer. Our measurements were
recorded using dried and fractionally distilled CHCl3,
and we found that addition of 2% EtOH to the CHCl3
solution immediately reduced the measured [RD] by
more than half.46
The symmetrical PBD dimer, DSB-120, was prepared
using similar chemistry (Scheme 6): the intermediate
10 was alkylated with 0.5 equiv of 1,3-dibromopropane,
the product was oxidized with DMP, and the Aloc
protecting groups were removed. The measured optical
rotation of the analytically pure solid ([RD] ) +1140°)
was again much higher than that previously reported
for this compound as an oil ([RD] ) +330°).47
The enantiomeric pair of CBIs in which the minor
groove binding component is truncated to a simple
acetyl substituent was also synthesized (32R and 32S,
Scheme 7), along with the previously described CBI
analogue 3329 bearing the trimethoxyindole substituent
common to the duocarmycin natural products.
a
Conditions: (a) c. HNO3, 0-20 °C; (b) (COCl)2, DMF, then
Cl3CCH2OH: (c) NaClO2, H2O2; (d) (COCl)2, DMF, then (S)-2-
pyrrolidinemethanol, K2CO3; (e) H2, Pd/C; (f) allyl chloroformate,
py, -78 °C; (g) DMP; (h) Zn powder, HCO2H.
Finally, two compounds were prepared that closely
mimic the structure of one of the cross-linking agents
(27eS, n ) 5, S-CBI enantiomer), except for minor
changes that render either end incapable of alkylating
DNA. For the ‘PBD inactive’ compound, the diazepine-
5,11-dione 35 was chosen. Retaining sp2 hybridization
at C11 is important to maintain the same overall shape
of the PBD, and diazepine-5,11-diones such as this have
been used in the past as nonalkylating structural
mimics of the PBDs.48 35 was synthesized as shown in
Scheme 8, starting from 13e, the byproduct from the
DMP oxidation step.49
chromatography column and immediate elution, pro-
vided the desired products in good yield (59-81%). This
method was successful too for the n ) 2 analogues where
again a reverse Michael reaction could have interfered.
In one case (27a S) the final product could not be
adequately purified, but the nine remaining compounds
1
were shown to be pure by H NMR and HPLC analysis
(>97%), and were carried forward to further testing.
Several reference compounds were also synthesized.
The PBD monomer DC-81 31 was prepared as shown
in Scheme 5. Reduction of 8 with SnCl2 selectively
reduced the nitro group as reported.34 The resulting
For the ‘CBI-inactive’ compound, the usual chlorom-
ethyl substituent was replaced with a simple methyl
Sch em e 4a
a
Conditions: (a) HCl, dioxane, then 15a -e, EDCI; (b) Pd(PPh3)4, pyrrolidine, CH2Cl2, 20 °C, e5 min.

2136 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Sch em e 5. Synthesis of DC-81a
Tercel et al.
a
Conditions: (a) SnCl2; (b) allyl chloroformate, py, -0 °C; (c)
DMP; (d) Pd(PPh3)4, pyrrolidine; (e) Pd/C, 1,4-cyclohexadiene.
F igu r e 1. Cytotoxicity of the CBI-PBD compounds 27a -eR
and 27b-eS in the AA8 cell line as a function of the linker
chain length (n). Filled symbols, R-CBI enantiomer; open
symbols, S-CBI enantiomer.
Sch em e 6. Synthesis of DSB-120a
3 and 5 being more potent than n ) 2 and 4. Presumably
this is due to a better match between the separation of
the alkylating functional groups and the base pair
register. A similar alternation in toxicity with polym-
ethylene chain length has been observed with sym-
metrical dimers in both CPI17 and PBD22 classes,
although the magnitude of the change was not so great.
For each chain length the S-CBI enantiomer is the more
potent, with the differential larger for the more toxic
chain lengths. These trends are more clearly seen in
Figure 1, where cytotoxicity in the AA8 cell line is
plotted on a log scale against chain length of the CBI-
PBD compounds.
a
Conditions: (a) Br(CH2)3Br, K2CO3, DMF; (b) DMP; (c) Pd-
(PPh3)4, pyrrolidine.
Sch em e 7. Synthesis of CBI Reference Compoundsa
There are no particularly large variations in sensitiv-
ity to a given compound across the cell lines in Table 1.
This is also true for 27cS, 27eR, 27eS (and 3) which
were tested in a more extensive panel of human tumor
cell lines: of colon (Colo205, HT29, WiDr), ovarian
(Skov3), or cervical (SiHa) origin (Table 2). Since the
Colo205 line is not adherent under culture conditions
an alternative assay employing continuous 5 day drug
exposure was used for this comparison.52 Under these
conditions of prolonged drug exposure Skov3 cells were
10-20 times more sensitive than in Table 1, but the
relative toxicity of the four compounds is similar in both
assays.
The similar sensitivity of the AA8 and UV4 cell lines
(Table 1) is worthy of note. UV4 is a sub-line of AA8
that is defective in nucleotide excision repair as a result
of a mutation in the ERCC1 gene53 and is hypersensitive
to many DNA alkylating agents. A hypersensitivity
factor [HF ) IC50(AA8)/IC50(UV4)] in the range 10-70
has been reported as typical of DNA cross-linking
agents,54 but for the CBI-PBD compounds the HF is
clearly much lower than this and mostly falls in the
range 1.5-3. It is possible that these compounds alky-
late with little distortion of the DNA structure, so that
the adducts are poorly recognized by the nucleotide
excision repair pathway.
a
Conditions: (a) HCl, dioxane; (b) AcCl, NaHCO3; (c) TMI acid,
EDCI.HCl.
group, as in structure 39. The synthesis of this utilized
an intermediate 3620 from the normal CBI synthesis
(Scheme 8). Alkylation with allyl bromide provided 37,50
which underwent the expected radical ring closure,
benzyl group deprotection, coupling with the PBD acid
15e, and deprotection, to give the desired product 39.
In Vitr o Cytotoxicity. Cellular toxicities were de-
termined as IC50 values after 4 h of drug exposure in
four cell lines: the Chinese hamster lines AA8 and UV4,
the murine mammary carcinoma EMT6, and the human
ovarian carcinoma Skov3.51 The results for the CBI-
PBD compounds, and for several reference compounds,
are collected in Table 1.
Immediately apparent is the enormous variation in
toxicity among the CBI-PBD compounds, with the ratio
between the least potent 27bR and most potent 27eS
being on the order of 35 000. There is also a clear
pattern in the toxicity data, with the chain lengths n )
It is also interesting to compare cytotoxicities with
those of the reference compounds in Table 1. DC-81 31,
the PBD monomer, has IC50s in the 1-2 µM range,
whereas the symmetrical dimer based on DC-81, DSB-

Unsymmetrical DNA Cross-Linking Agents
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2137
Sch em e 8. Synthesis of Monoalkylating Analoguesa
a
Conditions: (a) Pd(PPh3)4, pyrrolidine; (b) EDCI, THF; (c) Zn powder, HCO2H; (d) 6S, HCl, doxane, then 34, EDCI, DMA; (e) allyl
bromide, NaH; (f) Bu3SnH, AIBN, PhH; (g) Pd/C, NH4HCl2; (h) HCl, dioxane, then 15e, EDCI, DMA.
Ta ble 1. In Vitro Cytotoxicity (4 h exposure), NCI Results, and Cross-Linking Ability of CBI-PBDs and Reference Compounds
compound
nd
IC50 (nM)a
EMT6
5.0
300
51
0.30
0.030
12
NCI screenb
GI50 (nM)f rangeg
enante
AA8
48
UV4
6.1
860
160
0.83
0.074
14
Skov3
C50c (µM)
27a R
27bR
27bS
27cR
27cS
27d R
27d S
27eR
1
2
2
3
3
4
4
5
5
R
R
S
R
S
R
S
R
S
16
960
200
1.9
0.14
35
91
3200
500
13
<10
130
79
200
35
50
1.21
7.90
5.00
0.025
0.017
1.05
0.56
0.13
0.022
1600
240
4.0
0.13
40
11
4.2
0.054
2000
340
14000
210
0.21
29
4500
3200
>10000
190
5.9
1.5
3.1
11
3.0
160
0.56
0.0078
18
<10
2400
>10000
27eS
0.023
0.032
1900
140
14000
360
0.29
6.2
2100
31 DC-81
3 DSB-120
32R CBI-Ac
32S CBI-Ac
33 CBI-TMI
35 ‘PBD-inactive’
39 ‘CBI-inactive’
1500
1100
86
1800
59
23
12000
280
0.029
R
S
S
S
0.090
0.13
3.2
840
13
2500
a
b
Drug concentration to reduce cell density to 50% of that of the controls. Values are the mean of 2-6 experiments. Results from the
National Cancer Institute’s In Vitro Cell Line Screening Project. c Concentration required for 50% cross-linking of linearized pcDNA3
plasmid DNA following 18 h exposure at 37 °C. Number of carbons in the linker chain. e Enantiomeric form at C1 of the seco-CBI.
d
f Mean drug concentration required for 50% inhibition of growth compared to controls across the 60 cell line panel. Ratio of LC50 values
g
between the least sensitive and most sensitive cell lines.
Ta ble 2. In Vitro Cytotoxicity (5 day exposure) of Selected
CBI-PBD Compounds and DSB-120 3 against Human Tumor
Cell Lines
A similar comparison can be made considering the
PBD as a substituent on a CBI parent. For example,
the toxicity of the S-CBI acetamide 32S is increased
13 000 times with an n ) 5 linked PBD substituent,
27eS, but only by 40-fold with a similarly spaced
nonalkylating diazepinedione analogue, 35. Compound
27eS, which has IC50s in the pM range, is in fact more
cytotoxic than 33, although it lacks the indole substitu-
ent of the latter which is thought to be responsible for
activation of CBI-type compounds in general on binding
in the minor groove of DNA.55
IC50 (nM)a
compound
Colo205
HT29
WiDr
Skov3
SiHa
27cS
27eR
0.037
0.18
0.021
0.25
0.014
0.28
0.021
0.20
0.012b
0.20
27eS
3 (DSB-120)
0.0035
7.0
0.0027
8.7
0.0039 0.0033 0.0028
10 6.3 7.3
a
Drug concentration to reduce [3H]thymidine uptake to 50%
of that of the controls. Values are the mean of 2-3 experiments
b
except where marked. Single determination.
Overall, the marked dependence of potency on linker
chain length, and on the presence of two alkylation-
competent functional groups, strongly suggests that
DNA cross-linking is a major contributor to the observed
toxicities of the CBI-PBD compounds.
In Vitr o Activity in th e NCI Scr een . All of the
CBI-PBD compounds were submitted for testing in the
National Cancer Institute’s In Vitro Cell Line Screening
Project.56 Each compound was tested against a panel
of about 60 different human cancer cell lines derived
from nine different tissue types, and the results were
expressed as the concentration required for 50% inhibi-
120 (3), is on average about 20 times as toxic. For DSB-
120, the linker chain length was identified as being
optimal for both toxicity and cross-linking ability.22
However, if DC-81 carries an S-CBI substituent, even
the poorest choice of linker (n ) 2) gives a compound
27bS equitoxic with DSB-120, while the optimal linker
(n ) 5), 27eS, increases toxicity by a further 3 orders
of magnitude. This is clearly due to alkylation of DNA,
rather than some favorable noncovalent interaction,
since 39, the ‘CBI inactive’ variant of 27eS, is no more
toxic than DC-81 itself.

2138 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Tercel et al.
F igu r e 2. Extent of cross-linking of linearized pcDNA3 plasmid DNA by CBI-PBD compounds 27a -eR and 27b-eS and DSB-
120 3 following 18 h exposure at 37 °C. Fitted curves are Logistic 3-parameter sigmoidal curves with maximum y value constrained
to 100%. Concentrations required for 50% cross-linking (C50) are presented in Table 1.
tion of growth compared to controls (GI50), no growth
compared to controls (TGI), or 50% reduction in cells
(LC50). The mean GI50 values (analogous to the IC50
values reported above) for each of the CBI-PBDs across
the cell line panel are reported in Table 1, and again
these cover a large range of concentrations, with gener-
ally the same ranking of compound toxicity as reported
above. With this more extensive range of cell lines some
differential sensitivity was observed and was consider-
ably greater for the more potent compounds (see range
of LC50 values in Table 1). Although the individual
patterns varied, some cell lines repeatedly showed up
as significantly more sensitive than the average, for
example (based on LC50 data) the NSCLC NCI-H522,
the colon cancer Colo205, and several melanomas such
as UACC-62 and SK-MEL-2. In contrast the CNS
cancers and leukemia panel were generally more resis-
tant than average. The latter is surprising given the
myelotoxicity encountered in clinical trials of cyclopro-
paindolone compounds.16
that do, for example cyanomorpholino-adriamycin, com-
bine alkylating ability with intercalation and minor
groove binding capability.
At the GI50 level of comparison the data set was not
complete since both 27cS and 27eS were too toxic to
achieve 50% cell growth at the concentration ranges
tested. The remaining compounds could be divided into
two groups: 27cR and 27eR which showed some
correlation (PCC 0.50-0.60) to a group of topoisomerase
II inhibitors (e.g., adriamycin, daunomycin), and 27a R,
27d R, and 27d S which showed stronger correlation
(PCC > 0.74) to a set of DNA cross-linking agents (e.g.,
chlorambucil, uracil nitrogen mustard).
Cr oss-Lin k in g of Na k ed DNA. The ability of the
CBI-PBD compounds to cross-link naked DNA was
determined using a slight modification of a previously
described assay.58 pcDNA3 plasmid DNA (5400 bp) was
linearized and 3′-end labeled using 32P-dATP. The DNA
was incubated with the compounds for 18 h at 37 °C
and then denatured by alkali treatment. The basis of
the assay is that if the DNA is cross-linked, the
denatured strands are held in close proximity and
readily renature, so that on agarose gel electrophoresis
they run as double-stranded DNA. Following phospho-
rimaging the amount of cross-linking can be calculated.
This assay has been used in the past for symmetrical
dimers of cyclopropaindolones59 and PBDs.22,25 In both
cases denaturation was induced by heating (90 °C for 2
min), but because of the known susceptibility of cyclo-
propaindolone-adenine adducts to thermal depurina-
tion,60 we chose to use alkali-induced denaturation.58
It is also known that alkali can reverse cyclopropain-
dolone-adenine adduct formation,61 but control experi-
ments showed that prolonging the alkali exposure before
electrophoresis did not alter the levels of cross-linking
observed for the CBI-PBD compounds.
A COMPARE analysis57 was also performed for seven
of the CBI-PBDs (all except the examples where n )
2, 27bR and 27bS) against the compounds in the NCI’s
Standard Agent Database. The COMPARE algorithm
ranks compounds for the similarity of their in vitro
growth patterns against the 60 cell line panel. The
results are calculated as Pearson correlation coefficients
(PCCs) where the ‘seed compound’ has a PCC ) 1.00.
High correlation values have been shown to indicate a
common mechanism of action between various com-
pounds, which may not be structurally related. For the
CBI-PBDs using LC50 data the highest PCCs were
observed with intercalating agents (or aromatic chro-
mophores) bearing minor groove binding side chains. In
particular the RNA synthesis inhibitors actinomycin D,
chromomycin A3, and mithramycin A, the quinoxaline
bisintercalator echinomycin, and the topoisomerase II
inhibitor adriamycin were ranked in the top seven
matches for all of the CBI-PBDs with average PCCs
greater than 0.73. (Full details are provided in the
Supporting Information.) Surprisingly, alkylating agents
hardly feature in the COMPARE matches, and those
Each of the compounds was tested at five or more
concentrations, and the assays repeated 2-4 times.
Each gel also included 3 as an internal standard, which
gave good reproducibility on repeat assay.62 Figure 2
shows how the amount of cross-linked DNA varied with
compound concentration. The data points were fitted to

Unsymmetrical DNA Cross-Linking Agents
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2139
F igu r e 3. Relationship between cross-linking ability and
cytotoxicity in the AA8 cell line for the CBI-PBD compounds
27a -eR and 27b-eS and DSB-120 3. Symbols as in Figure
2. Regression line calculated for all compounds except 3 (r2
0.88, slope ) 1.32).
)
a sigmoidal curve (constrained to a maximum y value
of 100%) which in each case gave good convergence to
the data, and allowed calculation of C50 (the concentra-
tion required for 50% cross-linking). These values are
collected in Table 1.
Inspection of Figure 2 reveals some interesting pat-
terns. Cross-linking ability is dependent on chain length
in the order n ) 2 < 1 < 4 < 5 e 3, and for each chain
length the S-CBI enantiomer (open symbols in Figure
2) is the more efficient cross-linker. Cross-linking ability
spans a wide range (C50 ) 7.90-0.017 µM), with three
of the CBI-PBDs (27cR, 27cS, and 27eS) and DSB-
120 363 ranking as extremely efficient cross-linkers. For
these compounds cross-linking was detectable at con-
centrations as low as 1 nM, corresponding to less than
one drug molecule for each segment of linear DNA in
solution under the assay conditions. Even at the C50
concentrations there are less than 20 molecules of these
compounds per DNA segment in the assay.
Figure 3 plots cross-linking ability against cytotoxicity
in AA8 cells and shows a good correlation between these
properties. Excluding the PBD dimer 3 the log-log
regression gave r2 ) 0.88 with a slope of 1.32, again
supportive of cross-linking as the major mechanism of
cytotoxicity. (Results for the other cell lines were similar
with r2 values from 0.89 to 0.92 and slopes of 1.31-1.39,
see Supporting Information.) Compound 3 is clearly an
outlier from this data set and is about 400-fold less toxic
than would be predicted from its cross-linking ability.
One possible explanation is that 3 may cross-link more
slowly than the CBI-PBDs, so that comparable C50
values measured after 18 h drug exposure do not reflect
the level of cross-links generated during the cytotoxicity
assay (4 h exposure). However, when the cross-linking
assay was repeated for 3 and 27cR (the CBI-PBD of
most similar cross-linking ability) at 2 h drug exposure,
again near identical C50s (0.13 and 0.11 µM, respec-
tively) and practically superimposable cross-link curves
were observed (Supporting Information).
F igu r e 4. Comet assay of Skov3 cell suspensions following
exposure to 27eS at various concentrations for 1 h. Cells were
washed, and then aliquots were irradiated on ice with a dose
of 10 Gy. Cells were analyzed for DNA breakage or cross-
linking, with the histograms showing the distribution of tail
moments from 100 comets per treatment condition. Calculated
cross-linking indices: (panel E, 0%); panel F, 6%; panel G, 70%;
panel H, 101%.
and to assess DNA cross-linking under more physiologi-
cally relevant cellular conditions, comet assays64 were
performed for the most potent CBI-PBD 27eS and for
DSB-120 3. A single cell suspension of Skov3 cells was
exposed to the compounds for 1 h at various concentra-
tions and then irradiated with a dose of 10 Gy in order
to induce single strand breaks. The cells were im-
mobilized in an agarose gel on a microscope slide and
lysed with alkali. Electrophoresis draws any DNA
fragments out of the nuclei to form cometlike tails.
These are stained with propidium, and the tail moments
(mean migration distance × percent DNA in the tail)
were determined using an image analysis system.
Typical results for 27eS are shown in Figure 4, from
which it is apparent that treatment with 27eS alone
(panels B-D) does not significantly alter the control tail
moment distribution (panel A). Irradiation causes sig-
nificant DNA strand breakage as expected (panel E),
and this fragmentation can effectively be reversed by
pretreatment with increasing concentrations of 27eS
(panels F-H). A significant shift was apparent even at
the lowest drug concentration tested (0.3 nM), a con-
centration at which clonogenic survival of the cells was
reduced to 10% of that of controls (data not shown).
Calculation of the cross-linking index (effectively, how
much a given drug treatment shifts the tail moment
DNA Cr oss-Lin k in g in Cells. To further investigate
this difference between the CBI-PBDs and DSB-120

2140 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Tercel et al.
F igu r e 5. Extent of cross-linking of cellular DNA (as mea-
sured by the cross-linking index, CI, derived from the comet
assay in Skov3 cells) by the CBI-PBD 27eS (open diamond)
and DSB-120 3 (filled circle) following 1 h exposure at 37 °C.
F igu r e 6. Thermal cleavage assay of DNA alkylated with
CBI-acetamides 32R and 32S and CBI-PBDs where n ) 5
(27eR and 27eS) and n ) 2 (27bR and 27bS), resolved using
denaturing polyacrylamide gel electrophoresis. The DNA was
incubated with the compounds for 6 h at 37 °C at the
concentrations indicated above the lanes (µM). The control lane
contained no alkylating agent. Dideoxy sequencing lanes are
labeled A, C, G, and T. The sequence is numbered relative to
the labeled strand 5′ end of the 512 bp PCR product which
was designated bp 1. Major alkylation sites are indicated by
the numbered arrows.
distribution from the radiation-only condition back to
the control condition) shows that for 27eS a CI50 (cross-
linking index of 50%) is achieved at a concentration of
approximately 1.5 nM (Figure 5).
A similar set of experiments was performed for DSB-
120 3 and generated a CI50 ) 3.0 µM (Figure 5). Thus,
although 3 and 27eS have essentially equal ability to
cross-link naked DNA (C50 ratio ) 1.3, Table 1), the
latter is at least 1000 times more efficient at cross-
linking cellular DNA. This ratio correlates much more
closely with the observed cytotoxicity differentials for
3 versus 27eS against Skov3 cells (4300 in Tables 1 and
1900 in Table 2).
Sequ en ce Selectivity. To gain some information on
the sequence selectivity of the alkylation events a
thermal cleavage assay was performed using a 512 bp
fragment of the lac UV5 promoter as the target DNA.
The 5′-32P-end labeled DNA was incubated with the
compounds of interest and then heated at 100 °C for 30
min. This assay is widely used with CBI-type com-
pounds since it induces depurination and strand cleav-
age at all CBI alkylation sites, and the fragments can
be readily identified following gel electrophoresis.60,65
Figure 6 shows results for a 90 bp section of the target
DNA for three pairs of compounds: the CBI-acetamides
32R and 32S, and the CBI-PBDs where n ) 2 (27bR
and S) and 5 (27eR and S). For each of these compounds
on all occasions (a total sequence of 240 bps was
analyzed by varying the electrophoresis conditions) the
only alkylation site observed was adenine, almost
exclusively with one or two AT bps on the 5′ side of the
alkylation site. This sequence matches the reported
binding preference of truncated CBI analogues65 (such
as 32R and 32S), but it was surprising to observe
exactly the same sites with approximately the same
relative affinities for the CBI-PBDs. It appears that
the CBI portion drives the overall sequence selectivity
of the dimers, and the presence of the flexibly linked
PBD neither limits the sites which are alkylated nor
directs the CBI to sites that were not previously
targeted. This was particularly unexpected for the pairs
of compounds which differ only in CBI enantiomeric
form (27bR and 27bS, 27eR and 27eS), since for these
to alkylate a given adenine on the labeled strand the
bulky PBD substituent must necessarily point in the
Ta ble 3. Maximum Tolerated Doses and Antitumor Activity of
Selected CBI-PBD Compounds and DSB-120 3
MTDa (µmol/kg)
MTD ratio time to end pointb
(iv/ip)
compound
ip
iv
a( SEM (days)
vehicle control
27cS
27eR
27eS
3 (DSB-120)
-
-
-
100
17.7
10
20.1 ( 2.9
26.0 ( 4.5
24.4 ( 2.1
40.8 ( 7.9
22.8 ( 2.5
0.00316 0.316
0.75
0.0178
1.0
13.3
0.178
4.21
4.21
a
Maximum tolerated dose for single injection to non-tumor-
b
bearing mice. Average time for WiDr tumor xenograft to reach
a diameter of 17 mm following single iv injection at day 0 (tumor
diameter 8 mm).
opposite direction along the minor groove and to there-
fore lie in a quite different sequence environment. The
PBD does however make the alkylation more efficient,
as similar intensity bands were observed for the CBI-
PBDs at much lower concentrations than for the CBI-
acetamides.
In Vivo Activity. Three CBI-PBDs were chosen for
in vivo study (27cS, 27eR, and 27eS), a subset that
includes the most efficient cross-linking agents and
representatives of two chain lengths (n ) 3, 5) and both
CBI configurations. DSB-120 3 was also included as a
comparison, although in vivo activity was not antici-
pated for this compound.24 Maximum tolerated doses
(MTDs) were determined for single intraperitoneal (ip)
or intravenous (iv) injections in nontumor-bearing C3H/
HeN mice (Table 3). Comparing routes of administra-
tion, iv injection was better tolerated for all compounds
(up to 100 times for 27cS) and caused acute toxicity at
doses above the MTD (time to death e17 days, see
Supporting Information), while higher doses by ip
injection were associated with later deaths (g33 days)
which appeared to result from local abdominal toxicity
(chemical peritonitis and subsequent organ adhesions).
The CBI-PBDs appeared intermediate in toxicity be-

Unsymmetrical DNA Cross-Linking Agents
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2141
Although this route would be adaptable to more complex
structures (e.g., CBIs bearing a minor groove binding
indole substituent, or PBDs with exo-unsaturation) the
emphasis in this study was to use the minimal alky-
lating agents and simplest possible linker to define the
structural requirements for efficient cross-linking.
Both linker chain length and CBI enantiomeric form
appear critical, with an enormous variation in cytotox-
icity, more than 35 000-fold in several cell lines, ob-
served among the set of compounds prepared. A good
correlation between toxicity and the ability to cross-link
naked DNA (Figure 3) supports the proposal that DNA
cross-linking is the major mechanism of cytotoxicity of
these compounds. This is emphasized by the relative
inactivity of the monoalkylating agents 35 and 39,
where even the latter is more than 2 orders of magni-
tude less toxic than its close structural analogue 27eS.
The most efficient CBI-PBD cross-linkers are excep-
tionally cytotoxic with IC50s in the pM range.
F igu r e 7. Survival curves for nude mice bearing WiDr
xenografts treated with single iv doses of CBI-PBD com-
pounds or DSB-120 3. Averages for groups of 5-9 mice.
Controlled tumor implies mean tumor diameter less than 17
mm.
Interestingly, the correlation between cross-linking
ability and cytotoxicity does not extend to the sym-
metrical PBD dimer 3, which appeared much less toxic
than anticipated (Figure 3). Compound 3 has previously
been characterized as a highly efficient cross-linker of
cellular DNA (using the technique of alkaline elution
in K562 cells),66 so it was remarkable to observe a CBI-
PBD compound cross-linking cellular DNA at 1000-fold
lower concentrations (Figure 5). The reasons for this
different behavior are not clearslarge differences in
cellular uptake between the similar structures 3 and
27eS would not be expected, and for a 1 h drug exposure
time differential repair of DNA adducts also seems
unlikely, especially as DSB-120 adducts are known to
be poorly repaired.66 One possible explanation lies in
the different reactivity of the CBI and PBD alkylating
agentsswhereas PBDs react (reversibly) with a range
of biological nucleophiles such as glutathione, cyclopro-
paindolones are highly selective for reaction with DNA,
so it is possible that the CBI portion of the mixed dimers
directs these compounds more productively to DNA as
the cellular target.
tween bizelesin 1 and DSB-120 3, which have reported
MTDs for single iv injections of 0.018 and 9.4 µmol/kg,
respectively.13,24 However, given the large IC50 differ-
ences compared to 3, the mice tolerated considerably
higher levels of the CBI-PBDs than anticipated, par-
ticularly by the iv route.
Antitumor activity was assessed using the human
WiDr colon carcinoma grown as a xenograft in CD-1
nude mice. In cytotoxicity experiments using cell cul-
tures (Table 2) this cell line displayed similar responses
to the cross-linking agents as Colo205, which was
highlighted in the NCI screening results as a particu-
larly sensitive line. Single ip or iv doses were adminis-
tered at the MTD to mice bearing established tumors
with an average diameter of 8 mm. This treatment was
well tolerated in the tumor-bearing mice as judged by
minimal early body weight changes (for the first 5 days
-0.3 to +0.8% compared to +0.9 to 1.3% for controls).
No significant antitumor effects were observed for any
of the compounds following ip injection, nor for 3, 27cS
or 27eR via the iv route. However 27eS administered
as a single iv dose gave a statistically significant (p <
0.05) growth delay of 21 days: the average time to end
point (mean tumor diameter of 17 mm) was 41 days
after 27eS treatment compared to 20 days for controls
(Figure 7 and Table 3). The group of 9 treated mice also
included one cure (no palpable tumor 150 days post-
treatment), which was not included in the determination
of growth delay. The individual tumor growth curves
are presented in the Supporting Information section.
The CBI portion also appears to drive the sequence
selectivity of the dimers, overriding the sequence prefer-
ence of the PBD and leading to alkylation at the same
sites as those targeted by truncated CBIs (32R and 32S)
(Figure 6). This is observed irrespective of the CBI
configuration. However, the thermal cleavage assay
used to generate Figure 6 only gives information on CBI
alkylation sites. It is possible that the vast majority of
these sites represent monoalkylation events, and only
a very small subset, those sites where the PBD finds
itself in a suitable environment for cross-linking, gener-
ate the lesions responsible for the observed toxicity. It
would be interesting to conduct experiments with short
oligonucleotides (as reported for 5)27 to confirm the
expected adenine-N3 to guanine-N2 cross-link and to
define the binding site size and sequence preference of
the CBI-PBDs. Another open question concerns the
reversibility of adduct formation, and in particular
whether a CBI-PBD monoalkylated via the CBI can
translocate to a site suitable for cross-linking, at which
it may then become trapped.
Con clu sion s
In this study we describe the preparation and inves-
tigation of a set of CBI-PBD compounds designed as
unsymmetrical DNA cross-linking agents. The conver-
gent and efficient synthesis, which introduces the vari-
able chain length linker at as late a stage as possible,
provides the target compounds in high purity and
reasonable overall yield and in no more than 12 linear
steps from commercially available starting materials.
Results from the NCI In Vitro Cell Line Assay and
COMPARE analysis were also interesting. Although

2142 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Tercel et al.
Gen er a l Meth od for th e P r ep a r a tion of TCE Ester s of
ω-Br om oa lk a n oic Acid s. P r ep a r a tion of 2,2,2-Tr ich lor o-
eth yl 4-Br om obu tyr a te (11c). (COCl)2 (5.1 mL, 59 mmol)
and DMF (3 drops) were added to a solution of 4-bromobutyric
acid (7.47 g, 45 mmol) in CH2Cl2 (70 mL), and the solution
was stirred at 20 °C for 16 h. The solvent was evaporated, the
residue was redissolved in CH2Cl2 (60 mL), and 2,2,2-trichlo-
roethanol (4.7 mL, 50 mmol) was added. The solution was
stirred at 20 °C for 4 h and then washed with aq NaHCO3.
The organic layer was dried (Na2SO4) and evaporated and the
residue purified by column chromatography (20:1 petroleum
ether:EtOAc) to give 11c as a colorless liquid (8.56 g, 64%);
1H NMR (CDCl3) δ 4.76 (s, 2 H), 3.50 (t, J ) 6.4 Hz, 2 H), 2.69
(t, J ) 7.2 Hz, 2 H), 2.25 (quint, J ) 6.8 Hz, 2 H), which was
used directly in the next step.
correlation of the CBI-PBD differential toxicity pat-
terns to those of intercalating agents seems initially
surprising, cross-correlation between alkylating agents
and topoisomerase poisons are not unknown.67 When
bizelesin itself is used as a seed compound for COM-
PARE analysis, it also generates a list from the Stan-
dard Agent Database that includes, for example, acti-
nomycin D, chromomycin A3, mithramycin A, adriamy-
cin, and daunomycin among the most closely correlated
hits.68
The marked differential toxicities for the most potent
CBI-PBDs in the 60 cell line panel (Table 1), coupled
with the apparent resistance of leukemia cell lines,
prompted an investigation of the in vivo antitumor
activity of the CBI-PBDs using a growth delay assay.
All three CBI-PBDs tested were better tolerated than
expected, given their much greater in vitro toxicity
compared to DSB-120 3, and one compound, 27eS,
provided a statistically significant increase in lifespan
for mice bearing WiDr human colon xenografts. Given
the demanding nature of this assay, this positive result
suggests that further testing is warranted against other
tumor models and using alternative dosing schedules.
The following compounds were prepared by the same
general method.
2,2,2-Tr ich lor oeth yl br om oa ceta te (11a )69 as a colorless
1
liquid (93%); H NMR (CDCl3) δ 4.82 (s, 2 H), 3.98 (s, 2 H).
2,2,2-Tr ich lor oet h yl 3-br om op r op ion a t e (11b ) as
a
1
colorless liquid (45%); H NMR (CDCl3) δ 4.80 (s, 2 H), 3.63
(t, J ) 6.7 Hz, 2 H), 3.09 (t, J ) 6.7 Hz, 2 H).
2,2,2-Tr ich lor oeth yl 5-br om op en ta n oa te (11d )70 as a
1
colorless liquid (82%); H NMR (CDCl3) δ 4.76 (s, 2 H), 3.43
(t, J ) 6.4 Hz, 2 H), 2.52 (t, J ) 7.2 Hz, 2 H), 1.99-1.83 (m, 4
H).
2,2,2-Tr ich lor oeth yl 6-br om oh exa n oa te (11e)71 as a
Exp er im en ta l Section
1
colorless liquid (79%); H NMR (CDCl3) δ 4.75 (s, 2 H), 3.41
Ch em istr y. Analyses were carried out in the Microchemical
Laboratory, University of Otago, NZ. Melting points were
determined using an Electrothermal Model 9200 digital melt-
ing point apparatus and are as read. NMR spectra were
obtained on a Bruker AM-400 spectrometer at 400 MHz (1H)
or 100 MHz (13C). Mass spectra were obtained on a Varian
VG 7070 mass spectrometer at nominal 5000 mass resolution.
CHCl3 used for optical rotation measurements was washed
with H2O (×3), dried (CaCl2), fractionally distilled from P2O5,
and stored in the dark.
5-Am in o-4-{[(2S)-2-(h yd r oxym eth yl)p yr r olid in yl]ca r -
bon yl}-2-m eth oxyp h en ol (9). A solution of {(2S)-1-[4-(ben-
zyloxy)-5-methoxy-2-nitrobenzoyl]pyrrolidinyl}methanol (8)34
(3.43 g, 8.87 mmol) in MeOH (100 mL) was hydrogenated over
Pd/C (5%, 0.37 g) at 50 psi for 55 min. The catalyst was filtered
off through Celite, the filtrate was evaporated, and the residue
was dissolved in CH2Cl2 and evaporated to give 9 as a tan foam
(2.36 g, 100%); 1H NMR (DMSO-d6) δ 9.09 (br s, 1 H), 6.69 (s,
1 H), 6.17 (s, 1H), 4.98 (s, 2 H), 4.75 (s, 1 H), 4.12-4.03 (m, 1
H), 3.64 (s, 3 H), 3.48-3.28 (m, 4 H), 1.98-1.78 (m, 3 H), 1.71-
1.61 (m, 1 H); 13C NMR δ 169.1, 149.1, 142.0, 138.4, 113.5,
110.5, 103.2, 61.5, 58.4, 56.6, 49.1, 27.1, 24.0, which was used
directly in the next step.
Allyl 5-H yd r oxy-2-{[(2S)-2-(h yd r oxym et h yl)p yr r oli-
d in yl]ca r bon yl}-4-m eth oxyp h en ylca r ba m a te (10). The
aniline 9 prepared in the preceding reaction (2.36 g, 8.87 mmol)
was dissolved in THF (100 mL), pyridine (0.93 mL, 11.3 mmol)
was added, and the solution was cooled to -78 °C. Allyl
chloroformate (0.94 mL, 8.87 mmol) was added dropwise, and
the mixture was stirred for 10 min at -78 °C. H2O and aq
HCl (2 N, 5 mL) were added and the reaction was allowed to
warm to room temperature. The THF was evaporated and the
aq residue was extracted with CH2Cl2 (×3). The extracts were
dried (Na2SO4) and evaporated, and the residue purified by
column chromatography (100:1 then 95:5 EtOAc:MeOH) to
give 10 as a white foam (2.65 g, 85%); [R]D -133° (c 0.314,
CHCl3); 1H NMR (CDCl3) δ 8.54 (br s, 1 H), 7.67 (s, 1 H), 6.79
(s, 1 H), 6.27 (br s, 1 H), 6.00-5.90 (m, 1 H), 5.34 (dq, J )
17.3, 1.4 Hz, 1 H), 5.23 (dq, J ) 10.5, 1.3 Hz, 1 H), 4.68-4.57
(m, 2 H), 4.45-4.35 (m, 1 H), 4.33 (br s, 1 H), 3.90-3.83 (m, 1
H), 3.85 (s, 3 H), 3.73-3.66 (m, 1 H), 3.61-3.54 (m, 1 H), 3.52-
3.44 (m, 1 H), 2.20-2.12 (m, 1 H), 1.95-1.84 (m, 1 H), 1.81-
1.63 (m, 2 H); 13C NMR δ 171.0, 153.6, 148.1, 141.9, 132.5,
131.8, 118.1, 116.1, 110.3, 108.3, 66.4, 65.8, 61.0, 56.4, 51.6,
28.3, 25.1; HRMS (EI) calcd for C17H22N2O6 350.1478, found
350.1476.
(t, J ) 6.7 Hz, 2 H), 2.49 (t, J ) 7.4 Hz, 2 H), 1.94-1.86 (m, 2
H), 1.77-1.69 (m, 2 H), 1.57-1.49 (m, 2 H).
Gen er a l Meth od for P h en ol Alk yla tion . P r ep a r a tion
of 2,2,2-Tr ich lor oeth yl 4-(5-{[(Allyloxy)ca r bon yl]a m in o}-
4-{[(2S)-2-(h ydr oxym eth yl)pyr r olidin yl]car bon yl}-2-m eth -
oxyp h en oxy)bu ta n oa te (12c). K2CO3 (0.58 g, 4.2 mmol) and
then 11c (TCE ester) (0.93 g, 3.1 mmol) were added to a
solution of 10 (phenol) (0.73 g, 2.08 mmol) in DMF (6 mL),
and the mixture was stirred at 20 °C for 8 h. H2O and aq HCl
(2 N, 6 mL) were added, and the mixture was extracted with
EtOAc (×3). The extracts were washed with aq NaCl (×3) and
then dried (Na2SO4) and evaporated. The residue was purified
by column chromatography (EtOAc) to give 12c as a white
foam (0.84 g, 71%); [R]D -69° (c 0.346, CHCl3); 1H NMR
(CDCl3) δ 8.74 (br s, 1 H), 7.78 (s, 1H), 7.27 (s, 1 H), 6.02-
5.91 (m, 1 H), 5.36 (dq, J ) 17.3, 1.4 Hz, 1 H), 5.25 (br d, J )
11.5 Hz, 1 H), 4.77 (s, 2 H), 4.65-4.62 (m, 2 H), 4.47-4.37 (m,
1 H), 4.25 (br s, 1 H), 4.16 (t, J ) 6.1 Hz, 2 H), 3.87-3.81 (m,
1 H), 3.83 (s, 3 H), 3.74-3.68 (m, 1 H), 3.63-3.56 (m, 1 H),
3.54-3.46 (m, 1 H), 2.71 (t, J ) 7.3 Hz, 2 H), 2.25 (quint, J )
6.7 Hz, 2 H), 2.21-2.14 (m, 1 H), 1.95-1.86 (m, 1 H), 1.81-
1.62 (m, 2 H); 13C NMR δ 171.4, 170.9, 153.6, 150.5, 144.0,
132.5, 132.0, 118.1, 115.5, 111.7, 105.7, 94.9, 74.0, 67.6, 66.7,
65.8, 61.2, 56.6, 51.6, 30.6, 28.4, 25.1, 24.3; HRMS (EI, 35Cl)
calcd for C23H29Cl3N2O8 566.0990, found 566.0978.
The following compounds were prepared by the same
general method.
2,2,2-Tr ich lor oeth yl (5-{[(a llyloxy)ca r bon yl]a m in o}-4-
{[(2S)-2-(h yd r oxym eth yl)p yr r olid in yl]ca r bon yl}-2-m eth -
oxyp h en oxy)a ceta te (12a ) as a white foam (80% after a
reaction time of 2 h, leaving the reaction for longer caused
decomposition and much lower yields); [R]D -71° (c 0.228,
1
CHCl3); H NMR (CDCl3) δ 8.67 (br s, 1 H), 7.77 (s, 1H), 6.87
(s, 1 H), 6.00-5.90 (m, 1 H), 5.35 (dq, J ) 17.2, 1.5 Hz, 1 H),
5.25 (dq, J ) 10.4, 1.3 Hz, 1 H), 4.91 (s, 2 H), 4.86 (d, J ) 0.8
Hz, 2 H), 4.67-4.57 (m, 2 H), 4.45-4.37 (m, 1 H), 3.87 (s, 3
H), 3.86-3.82 (m, 1 H), 3.74-3.67 (m, 1 H), 3.61-3.46 (m, 2
H), 2.21-2.13 (m, 1 H), 1.96-1.88 (m, 1 H), 1.81-1.63 (m, 2
H); 13C NMR δ 170.5, 166.8, 153.5, 148.9, 144.1, 132.4, 131.7,
118.2, 117.1, 111.9, 106.3, 94.3, 74.1, 66.4, 65.8, 65.2, 61.1, 56.7,
51.5, 28.3, 25.1; HRMS (FAB, 35Cl) calcd for C21H26Cl3N2O8
539.0755, found 539.0747.
2,2,2-Tr ich lor oeth yl 5-(5-{[(a llyloxy)ca r bon yl]a m in o}-
4-{[(2S)-2-(h ydr oxym eth yl)pyr r olidin yl]car bon yl}-2-m eth -
oxyp h en oxy)p en ta n oa te (12d ) as a colorless oil (88%); [R]D

Unsymmetrical DNA Cross-Linking Agents
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2143
-68° (c 0.326, CHCl3); 1H NMR (CDCl3) δ 8.77 (br s, 1 H), 7.79
(s, 1H), 6.83 (s, 1 H), 6.02-5.92 (m, 1 H), 5.36 (dq, J ) 17.3,
1.5 Hz, 1 H), 5.25 (dq, J ) 10.5, 1.2 Hz, 1 H), 4.75 (s, 2 H),
4.69-4.60 (m, 2 H), 4.47-4.38 (m, 1 H), 4.15-4.07 (m, 2 H),
3.88-3.81 (m, 1 H), 3.83 (s, 3 H), 3.75-3.68 (m, 1 H), 3.64-
3.57 (m, 1 H), 3.54-3.46 (m, 1 H), 2.58 (t, J ) 7.1 Hz, 2 H),
2.22-2.14 (m, 1 H), 1.98-1.85 (m, 5 H), 1.81-1.61 (m, 2 H);
13C NMR δ 171.7, 171.0, 153.6, 150.7, 144.0, 132.5, 132.1,
118.1, 115.2, 111.7, 105.5, 95.0, 73.9, 68.2, 66.8, 65.8, 61.2, 56.7,
51.7, 33.6, 28.4, 28.3, 25.1, 21.4; HRMS (EI, 35Cl) calcd for
1H NMR (CDCl3) δ 7.29 (s, 1H), 6.71 (s, 1 H), 5.88-5.74 (m, 1
H), 5.64-5.54 (m, 1 H), 5.21-5.10 (m, 2 H), 4.87-4.82 (m, 2
H), 4.84 (s, 2 H), 4.65 (dd, J ) 13.3, 5.3 Hz, 1 H), 4.51-4.43
(m, 1 H), 3.94 (s, 3 H), 3.76 (s, 1 H), 3.73-3.67 (m, 1 H), 3.60-
3.52 (m, 1 H), 3.49-3.44 (m, 1 H), 2.18-2.08 (m, 2 H), 2.06-
1.96 (m, 2 H); 13C NMR δ 166.9, 166.6, 156.0, 149.0, 148.5,
131.8, 128.1, 127.8, 118.3, 115.9, 111.5, 94.3, 86.0, 74.1, 66.9,
66.0, 59.7, 56.1, 46.4, 28.7, 23.0; HRMS (FAB, 35Cl) calcd for
C
21H24Cl3N2O8 537.0598, found 537.0593; and a llyl (11a S)-
7-m eth oxy-5,11-d ioxo-8-[2-oxo-2-(2,2,2-tr ich lor oeth oxy)-
e t h oxy]-2,3,11,11a -t e t r a h yd r o-1H -p yr r olo[2,1-c][1,4]-
ben zod ia zep in e-10(5H)-ca r boxyla te (13a ) as a colorless oil
C
24H31Cl3N2O8 580.1146, found 580.1130.
2,2,2-Tr ich lor oeth yl 6-(5-{[(a llyloxy)ca r bon yl]a m in o}-
1
(5%); [R]D +35° (c 0.277, CHCl3); H NMR (CDCl3) δ 7.38 (s,
4-{[(2S)-2-(h ydr oxym eth yl)pyr r olidin yl]car bon yl}-2-m eth -
oxyp h en oxy)h exa n oa te (12e) as a colorless oil (83%); [R]D
-66° (c 0.515, CHCl3); 1H NMR (CDCl3) δ 8.77 (br s, 1 H), 7.78
(s, 1H), 6.82 (s, 1 H), 6.02-5.91 (m, 1 H), 5.36 (dq, J ) 17.2,
1.5 Hz, 1 H), 5.25 (dq, J ) 10.4, 1.3 Hz, 1 H), 4.75 (s, 2 H),
4.69-4.59 (m, 2 H), 4.46-4.38 (m, 1 H), 4.12-4.04 (m, 2 H),
3.88-3.82 (m, 1 H), 3.83 (s, 3 H), 3.74-3.68 (m, 1 H), 3.64-
3.57 (m, 1 H), 3.54-3.46 (m, 1 H), 2.51 (t, J ) 7.4 Hz, 2 H),
2.22-2.14 (m, 1 H), 1.94-1.84 (m, 3 H), 1.83-1.62 (m, 4 H),
1.60-1.52 (m, 2 H); 13C NMR δ 171.8, 171.1, 153.6, 150.8,
143.9, 132.5, 132.1, 118.1, 115.1, 111.6, 105.5, 95.0, 73.9, 68.5,
66.7, 65.8, 61.2, 56.7, 51.7, 33.8, 28.6, 28.4, 25.5, 25.1, 24.5;
HRMS (EI, 35Cl) calcd for C25H33Cl3N2O8 594.1303, found
594.1297.
1H), 6.76 (s, 1 H), 5.94-5.81 (m, 1 H), 5.32 (dq, J ) 17.0, 1.3
Hz, 1 H), 5.27 (br d, J ) 10.4 Hz, 1 H), 4.93-4.77 (m, 4 H),
4.75-4.63 (m, 2 H), 4.13-4.09 (m, 1 H), 3.96 (s, 3 H), 3.82-
3.76 (m, 1 H), 3.59-3.51 (m, 1 H), 2.73-2.67 (m, 1 H), 2.13-
1.95 (m, 3 H); HRMS (FAB, 35Cl) calcd for C21H22Cl3N2O8
535.0442, found 535.0426.
Allyl (11a S)-11-h yd r oxy-7-m eth oxy-5-oxo-8-{[5-oxo-5-
(2,2,2-tr ich lor oeth oxy)p en tyl]oxy}-2,3,11,11a -tetr a h yd r o-
1H -p yr r olo[2,1-c][1,4]b en zod ia zep in e-10(5H )-ca r b oxy-
la te (14d ) as a white foam (60%); [R]D +95° (c 0.378, CHCl3);
1H NMR (CDCl3) δ 7.24 (s, 1H), 6.66 (s, 1 H), 5.86-5.74 (m, 1
H), 5.66-5.58 (m, 1 H), 5.20-5.10 (m, 2 H), 4.75 (s, 2 H), 4.72-
4.64 (m, 2 H), 4.49-4.41 (m, 1 H), 4.07-3.97 (m, 2 H), 3.91 (s,
3 H), 3.73-3.65 (m, 1 H), 3.60-3.45 (m, 2 H), 2.57 (t, J ) 7.0
Hz, 2 H), 2.16-2.09 (m, 2 H), 2.05-1.97 (m, 2 H), 1.95-1.85
(m, 4 H); 13C NMR δ 171.7, 166.9, 156.1, 150.1, 148.7, 131.8,
128.3, 125.9, 118.0, 113.8, 110.8, 95.0, 86.0, 73.9, 68.5, 66.7,
59.8, 56.1, 46.4, 33.5, 28.7, 28.3, 23.0, 21.4; HRMS (EI, 35Cl)
calcd for C24H29Cl3N2O8 578.0990, found 578.0983; and a llyl
(11a S)-7-m eth oxy-5,11-d ioxo-8-{[5-oxo-5-(2,2,2-tr ich lor o-
et h oxy)p en t yl]oxy}-2,3,11,11a -t et r a h yd r o-1H -p yr r olo-
[2,1-c][1,4]ben zod ia zep in e-10(5H)-ca r boxyla te (13d ) as a
colorless oil (7%); [R]D +34° (c 0.134, CHCl3); 1H NMR (CDCl3)
δ 7.32 (s, 1H), 6.71 (s, 1 H), 5.93-5.81 (m, 1 H), 5.30 (dq, J )
17.1, 1.5 Hz, 1 H), 5.25 (dq, J ) 10.4, 1.1 Hz, 1 H), 4.82-4.71
(m, 1 H), 4.80 (s, 2 H), 4.67-4.61 (m, 1 H), 4.14-4.11 (m, 1
H), 4.05-3.98 (m, 2 H), 3.93 (s, 3 H), 3.82-3.76 (m, 1 H), 3.59-
3.52 (m, 1 H), 2.74-2.68 (m, 1 H), 2.58 (t, J ) 7.0 Hz, 2 H),
2.14-1.84 (m, 7 H); 13C NMR (CCl3 carbon not seen) δ 171.6,
170.3, 165.1, 152.2, 150.3, 148.9, 130.8, 128.3, 124.3, 119.5,
111.4, 110.9, 73.9, 68.7, 68.1, 59.7, 56.2, 46.6, 33.4, 28.2, 26.4,
23.6, 21.4; HRMS (EI, 35Cl) calcd for C24H27Cl3N2O8 576.0833,
found 576.0825.
Gen er a l Meth od for DMP Oxid a tion . P r ep a r a tion of
Allyl (11aS)-11-Hydr oxy-7-m eth oxy-5-oxo-8-[4-oxo-4-(2,2,2-
tr ich lor oeth oxy)bu toxy]-2,3,11,11a-tetr ah ydr o-1H-pyr r olo-
[2,1-c][1,4]ben zodiazepin e-10(5H)-car boxylate (14c). DMP
(1.31 g, 3.1 mmol) was added to a solution of 12c (1.17 g, 2.06
mmol) in CH2Cl2 (30 mL), causing slight warming, and the
solution was stirred at room temperature for 30 min. Aqueous
Na2SO3 (10%, 35 mL) was added, followed by aq NaHCO3, and
the mixture was stirred for a further 15 min. The organic layer
was separated, the aq layer was extracted with CH2Cl2 (×2),
and the combined extracts were dried (Na2SO4) and evapo-
rated. The residue was purified by column chromatography
(EtOAc, repeating with the mixed fractions to separate the
slightly less polar over-oxidation product) to give 14c as a
white foam (770 mg, 66%); mp 124-126 °C (PhH-petroleum
1
ether); [R]D +110° (c 0.257, CHCl3); H NMR (CDCl3) (shows
the presence of ca. 10% PhH) δ 7.24 (s, 1H), 6.68 (s, 1 H), 5.86-
5.73 (m, 1 H), 5.67-5.59 (m, 1 H), 5.19-5.09 (m, 2 H), 4.76 (s,
2 H), 4.67 (dd, J ) 13.3, 5.4 Hz, 1 H), 4.50-4.42 (m, 1 H),
4.12-4.02 (m, 2 H), 3.91 (s, 3 H), 3.77 (br s, 1 H), 3.73-3.66
(m, 1 H), 3.60-3.52 (m, 1 H), 3.50-3.44 (m, 1 H), 2.71 (t, J )
7.2 Hz, 2 H), 2.22 (quint, J ) 6.7 Hz, 2 H), 2.16-2.09 (m, 2
H), 2.04-1.96 (m, 2 H); 13C NMR (one aromatic quaternary
carbon possibly coincident with PhH signal at δ 128.3) δ 171.4,
166.9, 156.0, 149.9, 148.7, 131.8, 126.1, 118.0, 114.0, 110.8,
94.9, 86.0, 74.0, 67.7, 66.7, 59.8, 56.1, 46.3, 30.4, 28.7, 24.2,
23.0. Anal. (C23H27Cl3N2O8‚1/8PhH) C, H, N.
Allyl (11a S)-11-h yd r oxy-7-m eth oxy-5-oxo-8-{[6-oxo-6-
(2,2,2-tr ich lor oeth oxy)h exyl]oxy}-2,3,11,11a -tetr a h yd r o-
1H -p yr r olo[2,1-c][1,4]b en zod ia zep in e-10(5H )-ca r b oxy-
la te (14e) as a white foam (49%); [R]D +99° (c 0.348, CHCl3);
1H NMR (CDCl3) δ 7.24 (s, 1H), 6.65 (s, 1 H), 5.86-5.74 (m, 1
H), 5.66-5.57 (m, 1 H), 5.20-5.10 (m, 2 H), 4.75 (s, 2 H), 4.68
(dd, J ) 13.4, 5.4 Hz, 1 H), 4.48-4.42 (m, 1 H), 4.04-3.96 (m,
2 H), 3.91 (s, 3 H), 3.73-3.63 (m, 2 H), 3.60-3.44 (m, 2 H),
2.51 (t, J ) 7.4 Hz, 2 H), 2.15-2.09 (m, 2 H), 2.04-1.97 (m, 2
H), 1.92-1.84 (m, 2 H), 1.82-1.74 (m, 2 H), 1.59-1.51 (m, 2
H); 13C NMR δ 171.8, 167.0, 156.1, 150.1, 148.6, 131.8, 128.3,
125.7, 118.0, 113.7, 110.7, 95.0, 86.0, 73.9, 68.7, 66.7, 59.8, 56.1,
46.3, 33.8, 28.7, 28.6, 25.4, 24.4, 23.0; HRMS (EI, 35Cl) calcd
for C25H31Cl3N2O8 592.1146, found 592.1139; and a llyl (11a S)-
7-m eth oxy-5,11-d ioxo-8-{[6-oxo-6-(2,2,2-tr ich lor oeth oxy)-
h exyl]oxy}-2,3,11,11a -tetr a h yd r o-1H-p yr r olo[2,1-c][1,4]-
ben zod ia zep in e-10(5H)-ca r boxyla te (13e) as a colorless oil
Also isolated from the same column was a llyl (11a S)-7-
m eth oxy-5,11-d ioxo-8-[4-oxo-4-(2,2,2-tr ich lor oeth oxy)bu -
t o x y ]-2,3,11,11a -t e t r a h y d r o -1H -p y r r o lo [2,1-c][1,4]-
ben zod ia zep in e-10(5H)-ca r boxyla te (13c) as a colorless oil
(208 mg, 18%); [R]D +38° (c 0.170, CHCl3); 1H NMR (CDCl3) δ
7.33 (s, 1H), 6.73 (s, 1 H), 5.93-5.82 (m, 1 H), 5.30 (dq, J )
17.4, 1.5 Hz, 1 H), 5.25 (dq, J ) 10.4, 1.1 Hz, 1 H), 4.76 (s, 2
H), 4.75-4.71 (m, 1 H), 4.67-4.62 (m, 1 H), 4.13-4.04 (m, 3
H), 3.93 (s, 3 H), 3.82-3.74 (m, 1 H), 3.60-3.51 (m, 1 H), 2.71
(t, J ) 7.2 Hz, 2 H), 2.72-2.67 (m, 1 H), 2.23 (quint, J ) 6.6
Hz, 2 H), 2.12-1.95 (m, 3 H); 13C NMR δ 171.3, 170.2, 165.0,
152.2, 150.1, 148.8, 130.8, 128.3, 124.5, 119.5, 111.6, 110.9,
94.9, 74.0, 68.1, 67.9, 59.6, 56.1, 46.6, 30.3, 26.4, 24.1, 23.6;
HRMS (EI, 35Cl) calcd for C23H25Cl3N2O8 562.0677, found
562.0654.
The following compounds were prepared by the same
general method.
Allyl (11a S)-11-h yd r oxy-7-m eth oxy-5-oxo-8-[2-oxo-2-
(2,2,2-t r ich lor oe t h oxy)e t h oxy]-2,3,11,11a -t e t r a h yd r o-
1H -p yr r olo[2,1-c][1,4]b e n zod ia ze p in e -10(5H )-ca r b ox-
yla te (14a ) as a white foam (73%); [R]D +106° (c 0.279, CHCl3);
1
(26%); [R]D +31° (c 0.440, CHCl3); H NMR (CDCl3) δ 7.32 (s,
1H), 6.71 (s, 1 H), 5.93-5.82 (m, 1 H), 5.29 (dq, J ) 17.2, 1.3
Hz, 1 H), 5.24 (dq, J ) 10.5, 1.1 Hz, 1 H), 4.78-4.72 (m, 1 H),
4.75 (s, 2 H), 4.67-4.61 (m, 1 H), 4.16-4.11 (m, 1 H), 4.03-
3.96 (m, 2 H), 3.93 (s, 3 H), 3.83-3.74 (m, 1 H), 3.60-3.51 (m,
1 H), 2.74-2.67 (m, 1 H), 2.51 (t, J ) 7.4 Hz, 2 H), 2.12-1.96
(m, 3 H), 1.93-1.85 (m, 2 H), 1.82-1.74 (m, 2 H), 1.60-1.52
(m, 2 H); 13C NMR δ 171.7, 170.3, 165.1, 152.2, 150.4, 148.8,
130.8, 128.3, 124.1, 119.4, 111.2, 110.8, 95.0, 73.8, 68.9, 68.0,
59.7, 56.2, 46.6, 33.7, 28.5, 26.4, 25.4, 24.4, 23.6; HRMS (EI,
35Cl) calcd for C25H29Cl3N2O8 590.0990, found 590.0977.

2144 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Tercel et al.
Gen er a l Meth od for TCE Ester Clea va ge.35 P r ep a r a -
tion of 4-({(11a S)-10-[(Allyloxy)ca r bon yl]-11-h yd r oxy-7-
m eth oxy-5-oxo-2,3,5,10,11,11a -h exa h yd r o-1H-p yr r olo[2,1-
c][1,4]ben zod ia zep in -8-yl}oxy)bu ta n oic Acid (15c). Zn
powder (0.68 g, 10.4 mmol) was added to a solution of (14c)
(ester) (1.18 g, 2.09 mmol) in HCO2H (88%, 30 mL), causing
slight warming, and the mixture was stirred at room temper-
ature for 1 h. The mixture was diluted with EtOAc, filtered
through Celite, and evaporated. The residue was dissolved in
EtOAc and extracted with aq NaHCO3 (×3). The aq extracts
were acidified with c.HCl and then extracted with EtOAc (×4).
The organic extracts were dried (Na2SO4) and evaporated, and
the resulting oil was crystallized from EtOAc-PhH to give 15c
as a white solid (0.81 g, 89%); mp 163-165 °C; [R]D +123° (c
0.237, CHCl3); 1H NMR (CDCl3) δ ca. 7.5 (v br s, 1 H), 7.25 (s,
1H), 6.77 (s, 1 H), 5.82-5.72 (m, 1 H), 5.62 (d, J ) 9.8 Hz, 1
H), 5.17-5.07 (m, 2 H), 4.64 (dd, J ) 13.3, 5.4 Hz, 1 H), 4.47-
4.39 (m, 1 H), 4.13-4.01 (m, 2 H), 3.90 (s, 3 H), 3.73-3.66 (m,
1 H), 3.59-3.44 (m, 2 H), 2.57 (t, J ) 6.8 Hz, 2 H), 2.19-2.07
(m, 4 H), 2.03-1.96 (m, 2 H); 13C NMR δ 177.3, 167.1, 156.2,
150.0, 148.7, 131.7, 125.8, 118.1, 114.1, 110.8, 86.0, 67.8, 66.9,
60.1, 56.1, 46.4, 30.0, 28.7, 23.8, 23.0. Anal. (C21H26N2O8‚PhH)
C, H, N.
72%). A sample crystallized from PhH-petroleum ether, mp
1
79-80 °C; H NMR (CDCl3) δ 9.85 (s, 1 H), 7.44 (dd, J ) 8.1,
1.9 Hz, 1 H), 7.41 (d, J ) 1.8 Hz, 1 H), 6.99 (d, J ) 8.1 Hz, 1
H), 4.28 (t, J ) 6.0 Hz, 2 H), 3.92 (s, 3 H), 3.89 (t, J ) 5.6 Hz,
2 H), 2.40 (br s, 1 H), 2.17-2.10 (m, 2 H); 13C NMR δ 190.9,
153.7, 149.8, 130.2, 126.7, 111.4, 109.1, 67.8, 60.8, 56.0, 31.6.
Anal. (C11H14O4) C, H.
3-(4-For m yl-2-m eth oxy-5-n itr oph en oxy)pr opion ic Acid
(19). The alcohol 17 (3.57 g, 17.0 mmol) was added in portions
over several minutes with stirring to c.HNO3 (50 mL) at 0 °C.
The internal temperature rose over 15 min to 18 °C, then fell
again after 10 min to 5 °C. The ice bath was removed and the
mixture allowed to warm to room temperature, with no further
exotherm. After 20 min the mixture was poured into ice-cold
H2O (400 mL) and the solid was filtered off. The solid was
dissolved in EtOAc, and the solution was washed with H2O
until the washings were only slightly acidic and then extracted
with aq NaHCO3 (×3). The aq extracts were acidified (c.HCl)
and then extracted with EtOAc (×3). The EtOAc extracts were
dried (Na2SO4) and evaporated to give 19 as a pale yellow solid
1
(2.19 g, 48%); H NMR (DMSO-d6) δ 12.47 (br s, 1 H), 10.20
(s, 1 H), 7.73 (s, 1 H), 7.37 (s, 1 H), 4.38 (t, J ) 5.9 Hz, 1 H),
3.95 (s, 3 H), 2.78 (t, J ) 5.9 Hz, 1 H); 13C NMR δ 188.5, 171.7,
152.5, 150.9, 143.5, 124.7, 110.1, 108.2, 65.3, 56.4, 33.6; HRMS
The following compounds were prepared by the same
general method.
1
(EI) calcd for C11H11NO7 269.0536, found 269.0534. H NMR
analysis also showed the presence of an impurity (ca. 10%)
which was not removed by attempted crystallization. This
impurity is tentatively identified as 3-(2-methoxy-4,5-dinitro-
phenoxy)propionic acid 20.
The EtOAc layer from the NaHCO3 extraction was also dried
(Na2SO4) and evaporated and the residue purified by column
chromatography (1:2 EtOAc:petroleum ether) to give 3-(4-
for m yl-2-m eth oxy-5-n itr op h en oxy)p r op yl n itr a te (18) as
a yellow solid (0.49 g, 10%); 1H NMR (CDCl3) δ 10.46 (s, 1 H),
7.61 (s, 1 H), 7.42 (s, 1 H), 4.71 (t, J ) 6.1 Hz, 1 H), 4.26 (t, J
) 6.0 Hz, 2 H), 4.00 (s, 3 H), 2.37-2.31 (m, 2 H); 13C NMR δ
187.7, 153.6, 151.2, 143.6, 126.0, 110.1, 108.3, 69.3, 65.5, 56.6,
26.7; HRMS (EI) calcd for C11H12N2O8 300.0594, found 300.0589.
({(11aS)-10-[(Allyloxy)car bon yl]-11-h ydr oxy-7-m eth oxy-
5-oxo-2,3,5,10,11,11a -h exa h yd r o-1H -p yr r olo[2,1-c][1,4]-
ben zod ia zep in -8-yl}oxy)a cetic a cid (15a ) as a white solid
(59%); [R]D +136° (c 0.196, CHCl3); 1H NMR (CDCl3) δ 7.28 (s,
1H), 6.66 (s, 1 H), 5.84-5.72 (m, 1 H), 5.60 (d, J ) 9.8 Hz, 1
H), 5.19-5.07 (m, 2 H), 4.77-4.64 (m, 2 H), 4.60 (dd, J ) 13.3,
5.4 Hz, 1 H), 4.49-4.42 (m, 1 H), 3.92 (s, 3 H), 3.74-3.66 (m,
1 H), 3.58-3.41 (m, 2 H), 2.10-1.95 (m, 4 H); 13C NMR (one
aromatic quaternary C not seen) δ 170.9, 166.9, 156.2, 148.8,
131.6, 128.0, 127.2, 118.3, 115.3, 111.2, 85.9, 67.1, 66.0, 60.1,
56.2, 46.5, 28.5, 23.0; HRMS (FAB) calcd for C19H23N2O8
407.1454, found 407.1445.
5-({(11a S)-10-[(Allyloxy)ca r bon yl]-11-h yd r oxy-7-m eth -
oxy-5-oxo-2,3,5,10,11,11a-h exah ydr o-1H-pyr r olo[2,1-c][1,4]-
ben zod ia zep in -8-yl}oxy)p en ta n oic a cid (15d ) as a cream
solid (67%); mp 96-98 °C (EtOAc-Et2O); [R]D +123° (c 0.349,
2,2,2-Tr ich lor oet h yl 3-(4-F or m yl-2-m et h oxy-5-n it r o-
p h en oxy)p r op ion a te (21). (COCl)2 (2.7 mL, 30.7 mmol) and
DMF (2 drops) were added to a suspension of 19 (6.89 g, 25.6
mmol) in CH2Cl2 (120 mL), and the suspension was stirred at
20 °C for 18 h. The solution was evaporated, the residue was
redissolved in CH2Cl2 (80 mL), and the solution was cooled to
0 °C. 2,2,2-Trichloroethanol (2.7 mL, 28 mmol) and pyridine
(2.7 mL, 33 mmol) were added, the ice bath was removed, and
the mixture was stirred for 90 min. Aqueous HCl (2 N, 40 mL)
was added, and the mixture was extracted with CH2Cl2 (×3).
The extracts were dried (Na2SO4) and evaporated, and the
residue was purified by column chromatography (1:4 then 1:2
EtOAc:petroleum ether) to give 21 as a yellow oil (6.94 g, 68%);
1H NMR (CDCl3) δ 10.46 (s, 1 H), 7.66 (s, 1 H), 7.41 (s, 1 H),
4.82 (s, 2 H), 4.49 (t, J ) 6.1 Hz, 2 H), 3.98 (s, 3 H), 3.09 (t, J
1
CHCl3); H NMR (CDCl3) δ 7.25 (s, 1H), 6.70 (s, 1 H), 5.86-
5.74 (m, 1 H), 5.63 (d, J ) 9.8 Hz, 1 H), 5.18-5.07 (m, 2 H),
4.67 (dd, J ) 13.4, 5.5 Hz, 1 H), 4.50-4.43 (m, 1 H), 4.09-
3.97 (m, 2 H), 3.91 (s, 3 H), 3.73-3.67 (m, 1 H), 3.60-3.47 (m,
2 H), 2.44 (t, J ) 7.1 Hz, 2 H), 2.15-2.07 (m, 2 H), 2.03-1.96
(m, 2 H), 1.94-1.77 (m, 4 H); 13C NMR δ 177.5, 167.1, 156.2,
150.1, 148.7, 131.8, 128.3, 125.7, 118.1, 114.0, 110.8, 86.0, 68.6,
66.8, 60.0, 56.1, 46.4, 33.4, 28.7, 23.0, 21.3. Anal. (C22H28N2O8‚
1/2H2O) C, H, N.
6-({(11a S)-10-[(Allyloxy)ca r bon yl]-11-h yd r oxy-7-m eth -
oxy-5-oxo-2,3,5,10,11,11a-h exah ydr o-1H-pyr r olo[2,1-c][1,4]-
ben zod ia zep in -8-yl}oxy)h exa n oic a cid (15e) as a white
) 6.1 Hz, 2 H); 13C NMR δ 187.7, 168.8, 153.5, 151.2, 143.5,
1
foam (78%); [R]D +116° (c 0.203, CHCl3); H NMR (CDCl3) δ
35
126.0, 110.1, 108.5, 94.6, 74.2, 64.7, 56.6, 33.9; HRMS (EI,
-
7.24 (s, 1 H), 6.68 (s, 1 H), 5.83-5.71 (m, 1 H), 5.62 (d, J ) 9.8
Hz, 1 H), 5.16-5.06 (m, 2 H), 4.65 (dd, J ) 13.4, 5.2 Hz, 1 H),
4.49-4.40 (m, 1 H), 4.02-3.95 (m, 2 H), 3.90 (s, 3 H), 3.72-
3.65 (m, 1 H), 3.58-3.45 (m, 2 H), 2.33 (t, J ) 7.3 Hz, 2 H),
2.15-2.07 (m, 2 H), 2.03-1.95 (m, 2 H), 1.87-1.78 (m, 2 H),
1.72-1.63 (m, 2 H), 1.52-1.42 (m, 2 H); 13C NMR δ 179.0,
167.1, 156.3, 150.1, 148.7, 131.7, 128.3, 125.7, 118.0, 114.0,
110.7, 86.0, 68.8, 66.8, 60.2, 56.1, 46.4, 34.4, 28.7, 28.2, 25.2,
24.5, 23.0; HRMS (EI) calcd for C23H30N2O8 462.2002, found
462.1998.
Cl) calcd for C13H12Cl3NO7 398.9679, found 398.9680. 1H NMR
analysis also showed the presence of an impurity (ca. 10%),
identified as 2,2,2-trichloroethyl 3-(2-methoxy-4,5-dinitrophe-
noxy)propionate (22) (see below).
5-Met h oxy-2-n it r o-4-[3-oxo-3-(2,2,2-t r ich lor oet h oxy)-
p r op oxy]ben zoic Acid (23). 21 (6.43 g, 16.0 mmol) was
dissolved in CH3CN (80 mL), and a solution of NaH2PO4‚2H2O
(0.67 g, 4.3 mmol) in H2O (12 mL) was added, followed by aq
H2O2 (35%, 1.64 mL, 16.8 mmol). A solution of NaClO2 (80%,
2.54 g, 22.4 mmol) in H2O (25 mL) was then added dropwise
over 20 min, causing slight warming.39 The mixture was
stirred at room temperature for 90 min, then Na2SO3 (0.42 g,
3.2 mmol) was added and the mixture stirred for a further 5
min to decompose excess H2O2. The mixture was diluted with
aq NaCl and extracted with EtOAc (×3), and the extracts were
washed with aq NaCl (×2) and extracted with aq NaHCO3
(×3). The alkaline extracts were acidified (c.HCl) and extracted
with EtOAc (×2), and the organic extracts were dried (Na2-
SO4) and evaporated to give 23 as a yellow foam (5.04 g, 75%);
4-(3-Hyd r oxyp r op oxy)-3-m eth oxyben za ld eh yd e (17). A
mixture of vanillin (5.08 g, 33.4 mmol), 3-bromopropanol (3.6
mL, 40.1 mmol), and K2CO3 (6.00 g, 43.4 mmol) in DMF (40
mL) was stirred at 90 °C for 30 min and then cooled to room
temperature. H2O was added, and the mixture was extracted
with EtOAc. The organic extracts were evaporated (to remove
extracted DMF), and the residue partitioned between EtOAc
and H2O. The organic layer was dried (Na2SO4) and evaporated
and the residue purified by column chromatography (1:1
EtOAc:petroleum ether) to give 17 as a colorless oil (5.07 g,

Unsymmetrical DNA Cross-Linking Agents
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2145
1H NMR (CDCl3) δ 7.45 (s, 1 H), 7.22 (s, 1 H), 4.82 (s, 2 H),
4.44 (t, J ) 6.2 Hz, 2 H), 3.96 (s, 3 H), 3.07 (t, J ) 6.1 Hz, 2
H); 13C NMR δ 169.5, 168.9, 152.5, 149.9, 141.9, 120.2, 111.7,
108.7, 94.6, 74.2, 64.7, 56.6, 33.9; HRMS (EI, 35Cl) calcd for
1H -p yr r olo[2,1-c][1,4]b e n zod ia ze p in e -10(5H )-ca r b ox-
yla te. The allyl carbamate was oxidized with DMP using the
general method described above to give the titile compound
1
as a white foam (78%); [R]D +110° (c 0.265, CHCl3); H NMR
C
13H12Cl3NO8 414.9629, found 414.9628.
(CDCl3) δ 7.24 (s, 1 H), 6.73 (s, 1 H), 5.87-5.74 (m, 1 H), 5.66-
5.58 (m, 1 H), 5.20-5.11 (m, 2 H), 4.80 (s, 2 H), 4.67 (dd, J )
13.3, 5.5 Hz, 1 H), 4.50-4.42 (m, 1 H), 4.37-4.28 (m, 2 H),
3.89 (s, 3 H), 3.78 (br s, 1 H), 3.73-3.66 (m, 1 H), 3.60-3.44
(m, 2 H), 3.02 (t, J ) 6.5 Hz, 2 H), 2.17-2.09 (m, 2 H), 2.04-
1.96 (m, 2 H); 13C NMR δ 169.2, 166.8, 156.0, 149.5, 148.9,
131.8, 128.2, 126.6, 118.2, 114.7, 111.1, 94.7, 86.0, 74.1, 66.8,
64.4, 59.8, 56.1, 46.4, 34.0, 28.7, 23.0; HRMS (FAB, 35Cl) calcd
for C22H26Cl3N2O8 551.0755, found 551.0754.
The EtOAc layer from the NaHCO3 extraction was also dried
(Na2SO4) and evaporated, and the residue triturated with
MeOH to give 2,2,2-tr ich lor oeth yl 3-(2-m eth oxy-4,5-d in i-
tr op h en oxy)p r op ion a te (22) as a pale yellow solid (0.85 g,
13%); 1H NMR (CDCl3) δ 7.40 (s, 1 H), 7.31 (s, 1 H), 4.82 (s, 2
H), 4.46 (t, J ) 6.0 Hz, 2 H), 3.97 (s, 3 H), 3.08 (t, J ) 6.1 Hz,
2 H); 13C NMR δ 168.6, 152.3, 150.6, 137.3, 136.3, 108.5, 107.3,
94.6, 74.2, 65.1, 56.9, 33.8; HRMS (EI, 35Cl) calcd for C12H11
Cl3N2O8 415.9581, found 415.9585.
-
3-({(11a S)-10-[(Allyloxy)ca r bon yl]-11-h yd r oxy-7-m eth -
oxy-5-oxo-2,3,5,10,11,11a-h exah ydr o-1H-pyr r olo[2,1-c][1,4]-
ben zod ia zep in -8-yl}oxy)p r op ion ic Acid (15b). The TCE
ester was cleaved using Zn/HCO2H according to the general
method described above to give 15b as a white solid (69%);
mp 168-170 °C (EtOAc); [R]D +144° (c 0.295, CHCl3); 1H NMR
(CDCl3) δ 7.26 (s, 1 H), 6.75 (s, 1 H), 5.84-5.73 (m, 1 H), 5.63
(d, J ) 9.7 Hz, 1 H), 5.18-5.10 (m, 2 H), 4.65 (dd, J ) 13.3,
5.3 Hz, 1 H), 4.49-4.41 (m, 1 H), 4.34-4.25 (m, 2 H), 3.89 (s,
3 H), 3.73-3.66 (m, 1 H), 3.59-3.45 (m, 2 H), 2.89 (t, J ) 6.6
Hz, 2 H), 2.15-2.07 (m, 2 H), 2.04-1.96 (m, 2 H); 13C NMR δ
174.5, 167.0, 156.2, 149.5, 149.0, 131.6, 128.2, 126.5, 118.3,
115.0, 111.0, 86.0, 66.9, 64.6, 60.0, 56.1, 46.4, 33.9, 28.7, 23.0.
Anal. (C20H24N2O8) C, H, N.
2,2,2-Tr ich lor oeth yl 3-(4-{[(2S)-2-(Hyd r oxym eth yl)p yr -
r olid in yl]ca r b on yl}-2-m et h oxy-5-n it r op h en oxy)p r op i-
on a te (24). (COCl)2 (0.25 mL, 2.9 mmol) and DMF (1 drop)
were added to a solution of 23 (1.00 g, 2.40 mmol) in CH3CN
(20 mL,) and the was solution stirred at 20 °C for 16 h. This
solution was then added dropwise over 15 min to a mixture of
(S)-(+)-2-pyrrolidinemethanol (242 mg, 2.4 mmol) and K2CO3
(0.80 g, 5.8 mmol) in CH3CN (15 mL) with stirring at -40 °C.
The mixture was allowed to warm to 0 °C over 40 min, then
H2O and aq HCl (2 N, 7 mL) were added. The CH3CN was
evaporated, and the aq residue was extracted with EtOAc (×3).
The extracts were dried (Na2SO4) and evaporated, and the
residue was purified by column chromatography (1:50 MeOH:
EtOAc) to give 24 as a pale yellow foam (1.10 g, 92%); [R]D
Gen er a l Meth od for Cou p lin g CBI Am in es a n d P BD
Acid s. P r ep a r a tion of Allyl (11a S)-8-{2-[(1R)-1-(Ch lor o-
m eth yl)-5-h yd r oxy-1,2-d ih yd r o-3H-ben zo[e]in d ol-3-yl]-2-
oxoet h oxy}-11-h yd r oxy-7-m et h oxy-5-oxo-2,3,11,11a -t et -
r a h yd r o-1H -p yr r olo[2,1-c][1,4]b en zod ia zep in e-10(5H )-
car boxylate (26aR). A solution of tert-butyl (1R)-1-(chlorometh-
yl)-5-hydroxy-1,2-dihydro-3H-benzo[e]indole-3-carboxylate
(6R)29,31 (161 mg, 0.48 mmol) in dioxane (10 mL) was saturated
with HCl gas and allowed to stand until TLC analysis
indicated complete consumption of the starting material (ca.
45 min). The solvent was evaporated, and 15a (PBD acid n )
1) (196 mg, 0.48 mmol), EDCI‚HCl [1-(3-dimethylaminopro-
pyl)-3-ethylcarbodiimide hydrochloride] (184 mg, 0.96 mmol),
and DMA (5 mL) were added. The mixture was stirred at 20
°C for 16 h, and then aq NaHCO3 was added. The mixture
was extracted with EtOAc (×3), and the extracts were washed
with aq NaCl (×2), then dried (Na2SO4) and evaporated. The
residue was purified by column chromatography (EtOAc) to
give 26a R as a pale green solid (230 mg, 77%); [R]D +90° (c
1
-67° (c 0.323, CHCl3); H NMR (CDCl3) δ 7.76 (s, 1 H), 6.80
(s, 1 H), 4.82 (s, 2 H), 4.44 (t, J ) 6.1 Hz, 2 H), 4.42-4.37 (m,
1 H), 3.95 (s, 3 H), 3.94-3.87 (m, 1 H), 3.82-3.77 (m, 1 H),
3.17 (t, J ) 6.7 Hz, 2 H), 3.06 (t, J ) 6.1 Hz, 2 H), 2.23-2.14
(m, 1 H), 1.95-1.65 (m, 3 H); HRMS (FAB, 35Cl) calcd for
C
18H22Cl3N2O8 499.0442, found 499.0440.
2,2,2-Tr ich lor oeth yl 3-(5-{[(Allyloxy)ca r bon yl]a m in o}-
4-{[(2S)-2-(h ydr oxym eth yl)pyr r olidin yl]car bon yl}-2-m eth -
oxyp h en oxy)p r op ion a te. 24 (4.77 g, 9.55 mmol) was dis-
solved in MeOH (70 mL) and hydrogenated over Pd/C (5%,
0.19 g) at 50 psi for 30 min. TLC analysis (1:20 MeOH:EtOAc)
showed mostly starting material, so more Pd/C (0.19 g) was
added and hydrogenation continued at 50 psi for a further 30
min. This process was repeated until TLC analysis showed
mostly one product, slightly more polar than the starting
material. The product was resolved from small amounts of
over-reduction products that were slightly more polar again.
Reduction to this point required a total of 160 min hydrogena-
tion and 0.53 g of Pd/C. The catalyst was filtered off through
Celite, the filtrate was evaporated, and the residue was
purified by column chromatography (1:20 then 1:10 MeOH:
EtOAc) to give crude 2,2,2-trichloroethyl 3-(5-amino-4-{[(2S)-
2-(hydroxymethyl)pyrrolidinyl]carbonyl}-2-methoxyphenoxy)-
propionate 25 (2.31 g, 52%) that was used directly in the next
step.
Allyl chloroformate (0.52 mL, 4.9 mmol) was added dropwise
to a solution of 25 (2.31 g, 4.92 mmol) and pyridine (0.52 mL,
6.4 mmol) in THF (120 mL) at -78 °C. After 10 min H2O and
aq HCl (2 N, 10 mL) were added, and the mixture was allowed
to warm to room temperature. The THF was evaporated, and
the aq residue was extracted with CH2Cl2 (×3). The extracts
were dried (Na2SO4) and evaporated, and the residue was
purified by column chromatography (EtOAc) to give the title
compound as a white foam (1.53 g, 56%); [R]D -73° (c 0.295,
CHCl3); 1H NMR (CDCl3) δ 8.71 (br s, 1 H), 7.83 (s, 1 H), 6.83
(s, 1 H), 6.02-5.91 (m, 1 H), 5.36 (dq, J ) 17.2, 1.4 Hz, 1 H),
5.25 (dq, J ) 10.4, 1.1 Hz, 1 H), 4.80 (s, 2 H), 4.69-4.60 (m, 2
H), 4.46-4.37 (m, 3 H), 3.89-3.81 (m, 1 H), 3.80 (s, 3 H), 3.75-
3.68 (m, 1 H), 3.62-3.55 (m, 1 H), 3.53-3.45 (m, 1 H), 3.02 (t,
J ) 6.2 Hz, 2 H), 2.22-2.14 (m, 1 H), 1.95-1.86 (m, 1 H), 1.81-
1.60 (m, 2 H); 13C NMR δ 170.8, 169.2, 153.6, 150.1, 144.1,
132.5, 132.0, 118.2, 116.1, 112.0, 106.2, 94.7, 74.1, 66.7, 65.8,
64.0, 61.2, 56.7, 51.6, 34.1, 28.4; HRMS (FAB, 35Cl) calcd for
1
0.162, CHCl3); H NMR (CDCl3) δ 9.28 (s, 1 H), 8.13 (d, J )
8.4 Hz, 1 H), 8.10 (s, 1 H), 7.52 (d, J ) 8.1 Hz, 1 H), 7.47 (dt,
J ) 7.5, 1.1 Hz, 1 H), 7.35 (dt, J ) 7.5, 1.2 Hz, 1 H), 7.28 (s,
1 H), 6.96 (s, 1 H), 5.68-5.55 (m, 1 H), 5.05 (d, J ) 14.7 Hz,
1 H), 5.00-4.90 (m, 2 H), 4.88 (d, J ) 14.7 Hz, 1 H), 4.75 (br
s, 1 H), 4.44-4.32 (m, 3 H), 4.20 (t, J ) 9.8 Hz, 1 H), 3.95 (s,
3 H), 3.88-3.80 (m, 2 H), 3.66-3.60 (m, 1 H), 3.51-3.42 (m, 1
H), 3.39-3.30 (m, 2 H), 1.99-1.80 (m, 4 H); 13C NMR δ 167.0,
166.8, 155.8, 154.8, 149.1, 148.8, 140.7, 131.8, 129.7, 128.2,
127.8, 127.6, 123.7 (two coincident CBI tertiary carbons), 122.9,
122.1, 117.8, 115.9, 115.0, 111.2, 100.3, 85.9, 69.3, 66.8, 60.2,
56.2, 52.4, 46.4, 46.1, 42.5, 28.5, 23.0; HRMS (FAB, 35Cl) calcd
for C32H32ClN3O8 621.1878, found 621.1864.
The following compounds were prepared by the same
general method.
Allyl (11a S)-8-{3-[(1R)-1-(ch lor om eth yl)-5-h yd r oxy-1,2-
d ih yd r o-3H -b e n zo[e]in d ol-3-yl]-3-oxop r op oxy}-11-h y-
dr oxy-7-m eth oxy-5-oxo-2,3,11,11a-tetr ah ydr o-1H-pyr r olo-
[2,1-c][1,4]ben zod ia zep in e-10(5H)-ca r boxyla te (26bR) as
a cream solid (60%); [R]D +90° (c 0.187, CHCl3); 1H NMR
(CDCl3) δ 8.84 (s, 1 H), 8.23 (d, J ) 8.3 Hz, 1 H), 8.02 (s, 1 H),
7.52-7.38 (m, 3 H), 7.23 (s, 1 H), 6.92 (s, 1 H), 5.78-5.64 (m,
2 H), 5.10-5.01 (m, 2 H), 4.81 (br s, 1 H), 4.62-4.55 (m, 2 H),
4.45-4.38 (m, 1 H), 4.36-4.30 (m, 1 H), 4.19-4.14 (m, 2 H),
3.85 (s, 3 H), 3.84-3.78 (m, 1 H), 3.73-3.65 (m, 1 H), 3.59-
3.38 (m, 3 H), 3.31 (t, J ) 10.3 Hz, 1 H), 3.23-3.14 (m, 1 H),
3.04-2.95 (m, 1 H), 2.10-1.90 (m, 4 H); 13C NMR δ 169.2,
C
22H28Cl3N2O8 553.0911, found 553.0903.
Allyl (11a S)-11-Hyd r oxy-7-m eth oxy-5-oxo-8-[3-oxo-3-
(2,2,2-t r ich lor oet h oxy)p r op oxy]-2,3,11,11a -t et r a h yd r o-

2146 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Tercel et al.
167.0, 155.9, 154.3, 149.6, 148.3, 140.8, 131.8, 129.8, 128.6,
127.5, 126.5, 123.7, 123.6, 122.4, 122.2, 117.8, 115.0, 114.3,
110.7, 100.4, 86.1, 66.7, 64.5, 60.4, 56.1, 53.4, 46.5, 46.4, 41.9,
35.6, 28.7, 23.1; HRMS (FAB, 35Cl) calcd for C33H34ClN3O8
635.2034, found 635.2031.
Allyl (11a S)-8-{2-[(1S)-1-(ch lor om eth yl)-5-h yd r oxy-1,2-
dih ydr o-3H-ben zo[e]in dol-3-yl]-2-oxoeth oxy}-11-h ydr oxy-
7-m et h oxy-5-oxo-2,3,11,11a -t et r a h yd r o-1H -p yr r olo[2,1-
c][1,4]ben zod ia zep in e-10(5H)-ca r boxyla te (26a S) as a pale
green solid (71%); [R]D -88° (c 0.122, CHCl3).
Allyl (11a S)-8-{3-[(1S)-1-(ch lor om eth yl)-5-h yd r oxy-1,2-
d ih yd r o-3H -b e n zo[e]in d ol-3-yl]-3-oxop r op oxy}-11-h y-
dr oxy-7-m eth oxy-5-oxo-2,3,11,11a-tetr ah ydr o-1H-pyr r olo-
[2,1-c][1,4]ben zod ia zep in e-10(5H)-ca r boxyla te (26bS) as
a cream solid (42%): [R]D +23° (c 0.105, CHCl3); 1H NMR
(CDCl3) δ 8.73 (s, 1 H), 8.25-8.19 (m, 2 H), 7.65 (d, J ) 8.2
Hz, 1 H), 7.52 (t, J ) 7.4 Hz, 1 H), 7.40 (t, J ) 7.7 Hz, 1 H),
7.23 (s, 1 H), 6.92 (s, 1 H), 5.77-5.64 (m, 2 H), 5.09-5.02 (m,
2 H), 4.60-4.53 (m, 2 H), 4.47-4.30 (m, 4 H), 4.09-4.02 (m, 1
H), 3.99-3.93 (m, 1 H), 3.88 (s, 3 H), 3.70-3.63 (m, 1 H), 3.51-
3.32 (m, 3 H), 3.27-3.17 (m, 1 H), 3.14-3.04 (m, 1 H), 2.02-
1.80 (m, 4 H).
Allyl (11a S)-8-{4-[(1R)-1-(ch lor om eth yl)-5-h yd r oxy-1,2-
dih ydr o-3H-ben zo[e]in dol-3-yl]-4-oxobu toxy}-11-h ydr oxy-
7-m et h oxy-5-oxo-2,3,11,11a -t et r a h yd r o-1H -p yr r olo[2,1-
c][1,4]ben zod ia zep in e-10(5H)-ca r b oxyla t e (26cR) as
a
white solid (46%); [R]D +152° (c 0.155, CHCl3); 1H NMR
(CDCl3) δ 9.61 (s, 1 H), 8.23 (d, J ) 8.2 Hz, 1 H), 8.19 (s, 1 H),
7.59 (d, J ) 8.3 Hz, 1 H), 7.49 (dt, J ) 7.6, 1.0 Hz, 1 H), 7.35
(dt, J ) 7.7, 0.8 Hz, 1 H), 7.24 (s, 1 H), 6.92 (s, 1 H), 5.70-
5.58 (m, 2 H), 5.04-4.95 (m, 1 H), 4.88 (br s, 1 H), 4.54-4.46
(m, 1 H), 4.37-4.29 (m, 1 H), 4.19-4.06 (m, 3 H), 4.01-3.93
(m, 1 H), 3.91-3.83 (m, 2 H), 3.89 (s, 3 H), 3.72-3.65 (m, 1
H), 3.57-3.43 (m, 2 H), 3.29 (t, J ) 11.2 Hz, 1 H), 2.74-2.65
(m, 1 H), 2.60-2.51 (m, 1 H), 2.39-2.23 (m, 2 H), 2.12-1.95
(m, 4 H); 13C NMR δ 171.7, 167.1, 156.3, 154.8, 149.9, 148.6,
141.1, 131.6, 129.9, 128.4, 127.6, 125.9, 123.8, 123.4, 122.6,
122.0, 117.9, 114.5, 114.2, 110.6, 100.6, 86.0, 67.7, 66.8, 60.2,
56.1, 53.3, 46.4 (two coincident carbons: CH2Cl and PBD C-3),
Allyl (11a S)-8-{4-[(1S)-1-(ch lor om eth yl)-5-h yd r oxy-1,2-
dih ydr o-3H-ben zo[e]in dol-3-yl]-4-oxobu toxy}-11-h ydr oxy-
7-m et h oxy-5-oxo-2,3,11,11a -t et r a h yd r o-1H -p yr r olo[2,1-
c][1,4]ben zodiazepin e-10(5H)-car boxylate (26cS) as a white
1
solid (45-55%); [R]D +57° (c 0.176, CHCl3); H NMR (CDCl3)
42.1, 31.7, 28.6, 23.7, 23.0; HRMS (FAB, 35Cl) calcd for C34H36
ClN3O8 649.2191, found 649.2193.
-
δ 9.44 (s, 1 H), 8.29-8.24 (m, 2 H), 7.64 (d, J ) 8.3 Hz, 1 H),
7.51 (dt, J ) 7.6, 0.9 Hz, 1 H), 7.37 (dt, J ) 7.7, 0.7 Hz, 1 H),
7.23 (s, 1 H), 6.85 (s, 1 H), 5.67-5.57 (m, 2 H), 5.05-4.95 (m,
2 H), 4.54-4.47 (m, 1 H), 4.45 (br s, 1 H), 4.33-4.10 (m, 4 H),
4.05-3.98 (m, 1 H), 3.93 (dd, J ) 11.3, 2.8 Hz, 1 H), 3.88 (s, 3
H), 3.72-3.65 (m, 1 H), 3.58-3.49 (m, 1 H), 3.47-3.41 (m, 1
H), 3.37 (t, J ) 10.8 Hz, 1 H), 2.89-2.74 (m, 2 H), 2.46-2.29
(m, 2 H), 2.12-1.93 (m, 4 H).
Allyl (11a S)-8-({5-[(1R)-1-(ch lor om et h yl)-5-h yd r oxy-
1,2-d ih yd r o-3H-ben zo[e]in d ol-3-yl]-5-oxop en tyl}oxy)-11-
h yd r oxy-7-m eth oxy-5-oxo-2,3,11,11a -tetr a h yd r o-1H-p yr -
r olo[2,1-c][1,4]ben zod ia zep in e-10(5H)-ca r boxyla te (26dR)
1
as a gray solid (41%); [R]D +138° (c 0.189, CHCl3); H NMR
(CDCl3) δ 9.27 (s, 1 H), 8.25 (d, J ) 8.3 Hz, 1 H), 8.19 (s, 1 H),
7.63 (d, J ) 8.3 Hz, 1 H), 7.50 (dt, J ) 7.5, 0.8 Hz, 1 H), 7.35
(dt, J ) 7.4, 0.7 Hz, 1 H), 7.20 (s, 1 H), 6.86 (s, 1 H), 5.79-
5.65 (m, 1 H), 5.64 (br d, J ) 9.5 Hz, 1 H), 5.12-5.02 (m, 2 H),
4.64-4.57 (m, 1 H), 4.46-4.39 (m, 1 H), 4.28-4.06 (m, 5 H),
4.02-3.91 (m, 2 H), 3.87 (s, 3 H), 3.71-3.64 (m, 1 H), 3.57-
3.41 (m, 2 H), 3.40 (t, J ) 10.6 Hz, 1 H), 2.76-2.56 (m, 2 H),
2.13-2.06 (m, 2 H), 2.05-1.92 (m, 6 H); 13C NMR δ 172.2,
167.1, 156.3, 154.8, 150.0, 148.7, 141.3, 131.7, 130.0, 128.3,
127.7, 125.8, 123.8, 123.5, 122.6, 122.0, 118.0, 114.7, 114.2,
110.6, 100.3, 86.1, 69.0, 66.9, 60.0, 56.1, 53.3, 46.4, 46.3, 42.3,
Allyl (11a S)-8-({5-[(1S)-1-(ch lor om eth yl)-5-h yd r oxy-1,2-
d ih yd r o-3H-ben zo[e]in d ol-3-yl]-5-oxop en tyl}oxy)-11-h y-
dr oxy-7-m eth oxy-5-oxo-2,3,11,11a-tetr ah ydr o-1H-pyr r olo-
[2,1-c][1,4]ben zod ia zep in e-10(5H)-ca r boxyla te (26d S) as
a gray solid (37%); [R]D +84° (c 0.161, CHCl3); 1H NMR (CDCl3)
δ 8.97 (s, 1 H), 8.27 (d, J ) 8.3 Hz, 1 H), 8.18 (s, 1 H), 7.64 (d,
J ) 8.4 Hz, 1 H), 7.51 (t, J ) 7.6 Hz, 1 H), 7.37 (t, J ) 7.7 Hz,
1 H), 7.20 (s, 1 H), 6.98 (s, 1 H), 5.83-5.72 (m, 1 H), 5.65 (d,
J ) 9.6 Hz, 1 H), 5.16-5.05 (m, 2 H), 4.66 (dd, J ) 13.3, 4.8
Hz, 1 H), 4.49-4.42 (m, 1 H), 4.30-4.12 (m, 4 H), 4.07-3.99
(m, 1 H), 3.95 (dd, J ) 11.1, 2.6 Hz, 1 H), 3.86 (s, 3 H), 3.72-
3.64 (m, 1 H), 3.57-3.49 (m, 1 H), 3.46-3.40 (m, 1 H), 3.38 (t,
J ) 10.8 Hz, 1 H), 2.74-2.55 (m, 2 H), 2.10-1.93 (m, 8 H).
35.4, 28.7, 27.9, 23.1, 20.8; HRMS (FAB, 35Cl) calcd for C35H39
ClN3O8 664.2426, found 664.2431.
-
Allyl (11a S)-8-({6-[(1R)-1-(ch lor om et h yl)-5-h yd r oxy-
1,2-d ih yd r o-3H -b en zo[e]in d ol-3-yl]-6-oxoh exyl}oxy)-11-
h yd r oxy-7-m eth oxy-5-oxo-2,3,11,11a -tetr a h yd r o-1H-p yr -
r olo[2,1-c][1,4]ben zod ia zep in e-10(5H)-ca r boxyla te (26eR)
as a pale gray-green solid (49%): [R]D +157° (c 0.161, CHCl3);
1H NMR (CDCl3) δ 9.43 (s, 1 H), 8.28 (d, J ) 8.3 Hz, 1 H),
8.25 (s, 1 H), 7.63 (d, J ) 8.3 Hz, 1 H), 7.50 (dt, J ) 7.5, 1.2
Hz, 1 H), 7.36 (dt, J ) 7.6, 1.0 Hz, 1 H), 7.24 (s, 1 H), 6.79 (s,
1 H), 5.78-5.65 (m, 2 H), 5.12-5.01 (m, 2 H), 4.63 (dd, J )
13.3, 5.2 Hz, 1 H), 4.47-4.41 (m, 1 H), 4.32 (br s, 1 H), 4.28
(dd, J ) 10.9, 1.7 Hz, 1 H), 4.22-4.16 (m, 1 H), 4.12-3.98 (m,
3 H), 3.95 (dd, J ) 11.2, 3.0 Hz, 1 H), 3.88 (s, 3 H), 3.73-3.66
(m, 1 H), 3.58-3.43 (m, 3 H), 3.41 (t, J ) 10.7 Hz, 1 H), 2.69-
2.51 (m, 2 H), 2.14-2.08 (m, 2 H), 2.03-1.95 (m, 2 H), 1.93-
1.82 (m, 4 H), 1.66-1.56 (m, 2 H); 13C NMR δ 172.5, 167.2,
156.2, 155.0, 149.9, 149.0, 141.2, 131.6, 130.0, 128.3, 127.6,
126.0, 123.9, 123.4, 122.7, 122.0, 117.9, 114.9, 114.6, 110.7,
100.6, 86.0, 69.1, 66.8, 60.4, 56.1, 53.4, 46.4 (two coincident
carbons: CH2Cl and PBD C-3), 42.2, 35.7, 28.7, 28.2, 25.3, 24.2,
23.1; HRMS (FAB, 35Cl) calcd for C36H41ClN3O8 678.2582,
found 678.2561.
The same general method, starting from tert-butyl (1S)-1-
(chloromethyl)-5-hydroxy-1,2-dihydro-3H-benzo[e]indole-3-car-
boxylate (6S)29,31 was used to prepare compounds 26a -eS,
which differ from those described above only in the configu-
ration of the chiral center at C1 of the seco-CBI. These
compounds were not fully characterized but were used directly
in the next step. In general their 1H NMR were very similar
to those of the corresponding analogues 26a -eR, except for
the case where n ) 1 (26a S) where very broad signals,
presumably a consequence of amide conformers, were ob-
served.
Allyl (11a S)-8-({6-[(1S)-1-(ch lor om eth yl)-5-h yd r oxy-1,2-
d ih yd r o-3H -b en zo[e]in d ol-3-yl]-6-oxoh exyl}oxy)-11-h y-
dr oxy-7-m eth oxy-5-oxo-2,3,11,11a-tetr ah ydr o-1H-pyr r olo-
[2,1-c][1,4]ben zod ia zep in e-10(5H)-ca r boxyla te (26eS) as
1
a pale gray-green solid (45%); [R]D +92° (c 0.176, CHCl3); H
NMR (CDCl3) δ 9.11 (s, 1 H), 8.29 (d, J ) 8.4 Hz, 1 H), 8.18 (s,
1 H), 7.65 (d, J ) 8.3 Hz, 1 H), 7.51 (dt, J ) 7.5, 1.1 Hz, 1 H),
7.37 (dt, J ) 7.7, 0.7 Hz, 1 H), 7.25 (s, 1 H), 6.81 (s, 1 H),
5.79-5.68 (m, 2 H), 5.14-5.02 (m, 2 H), 4.65 (dd, J ) 13.5,
5.3 Hz, 1 H), 4.49-4.42 (m, 1 H), 4.41 (br s, 1 H), 4.32-4.14
(m, 3 H), 4.08-3.99 (m, 2 H), 3.96 (dd, J ) 11.3, 3.0 Hz, 1 H),
3.90 (s, 3 H), 3.74-3.67 (m, 1 H), 3.59-3.44 (m, 2 H), 3.40 (t,
J ) 10.9 Hz, 1 H), 2.70-2.52 (m, 2 H), 2.14-2.08 (m, 2 H),
2.03-1.94 (m, 2 H), 1.92-1.82 (m, 4 H), 1.65-1.50 (m, 2 H).
Gen er a l Meth od for Rem ova l of th e Aloc P r otectin g
Gr ou p. P r epar ation of (11aS)-8-{2-[(1R)-1-(Ch lor om eth yl)-
5-h yd r oxy-1,2-d ih yd r o-3H -b e n zo[e]in d ol-3-yl]-2-oxo-
e t h oxy}-7-m e t h oxy-1,2,3,11a -t e t r a h yd r o-5H -p yr r olo-
[2,1-c][1,4]ben zod ia zep in -5-on e (27a R). Pyrrolidine (0.30
mL, 3.6 mmol) then Pd(PPh3)4 (21 mg, 5%) was added to a
solution of 26a R (222 mg, 0.36 mmol) in CH2Cl2 (4 mL) and
the solution stirred at 20 °C. TLC analysis (1:15 MeOH:EtOAc)
indicated that the reaction was complete within 30 s. After 5
min the solution was concentrated to ca. 1 mL at 20 °C and
directly purified by column chromatography (1:20 MeOH:
EtOAc). The product-containing fractions were dissolved in and
evaporated from CH2Cl2 (×3, to convert any hemiaminal to
the imine form44), and the residue was triturated with EtOAc
(4 mL) to give 27a R as a cream solid (126 mg, 68%); mp 190-
1
196 °C (dec); [R]D +557° (c 0.176, CHCl3); H NMR (CDCl3) δ
9.62 (s, 1 H), 8.09 (d, J ) 8.2 Hz, 1 H), 8.04 (s, 1 H), 7.60 (s, 1

Unsymmetrical DNA Cross-Linking Agents
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2147
H), 7.55 (d, J ) 4.3 Hz, 1 H), 7.44 (t, J ) 7.4 Hz, 1 H), 7.38-
7.31 (m, 2 H), 6.84 (s, 1 H), 5.25 (d, J ) 15.0 Hz, 1 H), 4.99 (d,
J ) 15.0 Hz, 1 H), 4.26-4.20 (m, 2 H), 4.03 (s, 3 H), 3.85-
3.74 (m, 2 H), 3.60-3.47 (m, 3 H), 3.30 (t, J ) 10.5 Hz, 1 H),
2.19-2.13 (m, 2 H), 2.01-1.94 (m, 2 H); 13C NMR δ 166.5,
164.5, 162.8, 154.9, 149.8, 148.1, 140.8, 140.4, 129.6, 127.2,
123.7, 123.3, 122.7, 122.1, 121.5, 114.3, 112.0, 111.2, 100.3,
1,2,3,11a-tetr ah ydr o-5H-pyr r olo[2,1-c][1,4]ben zodiazepin -
5-on e (27a S) as a cream solid (61%); mp 225-230 °C (dec);
1H NMR (DMSO-d6) δ 10.41 (s, 1 H), 8.10 (d, J ) 8.3 Hz, 1 H),
7.90 (s, 1 H), 7.80 (d, J ) 8.4 Hz, 1 H), 7.75 (d, J ) 4.5 Hz, 1
H), 7.52 (t, J ) 7.5 Hz, 1 H), 7.34 (t, J ) 7.6 Hz, 1 H), 6.91 (s,
1 H), 5.10 (q, J ) 15.8 Hz, 1 H), 4.45-4.37 (m, 1 H), 4.28-
4.18 (m, 2 H), 4.00 (dd, J ) 10.9, 2.8 Hz, 1 H), 3.87 (s, 3 H),
3.84-3.57 (m, 3 H), 3.43-3.35 (m, 1 H), 2.32-2.17 (m, 2 H),
1.98-1.89 (m, 2 H); 13C NMR δ 165.3, 164.2, 163.2, 154.3,
149.6, 146.8, 141.5, 140.3, 129.8, 127.3, 123.1, 122.9, 122.6,
121.9, 120.3, 113.8, 111.5, 110.9, 99.4, 66.8, 55.6, 53.3, 51.1,
68.1, 56.2, 53.7, 52.0, 46.7, 46.3, 42.2, 29.4, 24.1. Anal. (C28H26
-
ClN3O5‚H2O) C, H; N: calcd, 7.81; found 7.38.
The following compounds were prepared by the same
general method.
47.5, 46.3, 41.2, 28.7, 23.6; HRMS (FAB, 35Cl) calcd for C28H26
-
(11a S)-8-{3-[(1R)-1-(Ch lor om eth yl)-5-h yd r oxy-1,2-d ih y-
d r o-3H -b en zo[e]in d ol-3-yl]-3-oxop r op oxy}-7-m et h oxy-
1,2,3,11a-tetr ah ydr o-5H-pyr r olo[2,1-c][1,4]ben zodiazepin -
5-on e (27bR) as a cream solid (74%); mp 185-190 °C (dec);
ClN3O5 519.1561, found 519.1559. 1H NMR analysis also
showed the presence of an impurity (ca. 10%) which could not
be removed by trituration or attempted crystallization from a
number of solvents.
1
[R]D +591° (c 0.158, CHCl3); H NMR (CDCl3) δ 9.88 (s, 1 H),
(11a S)-8-{3-[(1S)-1-(Ch lor om eth yl)-5-h yd r oxy-1,2-d ih y-
d r o-3H -b en zo[e]in d ol-3-yl]-3-oxop r op oxy}-7-m et h oxy-
1,2,3,11a-tetr ah ydr o-5H-pyr r olo[2,1-c][1,4]ben zodiazepin -
5-on e (27bS) as a cream solid (76%); mp 185-195 °C (dec);
8.37 (d, J ) 8.3 Hz, 1 H), 8.24 (s, 1 H), 7.70 (d, J ) 4.3 Hz, 1
H), 7.55 (d, J ) 8.2 Hz, 1 H), 7.50 (t, J ) 7.1 Hz, 1 H), 7.49 (s,
1 H), 7.36 (dt, J ) 7.5, 0.9 Hz, 1 H), 7.12 (s, 1 H), 4.61-4.46
(m, 1 H), 4.15-4.11 (m, 1 H), 3.94-3.73 (m, 4 H), 3.88 (s, 3
H), 3.60-3.48 (m, 2 H), 3.31 (t, J ) 10.8 Hz, 1 H), 3.08-3.01
(m, 2 H), 2.30-2.24 (m, 2 H), 2.09-1.98 (m, 2 H); 13C NMR δ
169.6, 164.6, 162.3, 154.9, 150.4, 147.4, 140.8, 129.8, 127.5,
124.3, 123.5, 122.6, 122.0, 120.4, 114.7, 111.6, 110.9, 100.6,
64.6, 56.1, 53.7, 53.5, 46.7, 46.5, 42.0, 35.8, 29.6, 24.2. Anal.
(C29H28ClN3O5‚H2O) C, H, N.
1
[R]D +372° (c 0.137, CHCl3); H NMR (CDCl3) δ 9.49 (s, 1 H),
8.30 (d, J ) 8.3 Hz, 1 H), 8.25 (s, 1 H), 7.68 (d, J ) 4.4 Hz, 1
H), 7.55 (d, J ) 8.2 Hz, 1 H), 7.51 (s, 1 H), 7.48 (dt, J ) 8.2,
1.0 Hz, 1 H), 7.35 (dt, J ) 7.6, 0.9 Hz, 1 H), 7.08 (s, 1 H), 4.56-
4.40 (m, 2 H), 4.26 (br d, J ) 11.0 Hz, 1 H), 4.17-4.10 (m, 1
H), 3.90 (s, 3 H), 3.89-3.76 (m, 3 H), 3.61-3.48 (m, 2 H), 3.31
(t, J ) 11.3 Hz, 1 H), 3.16-3.08 (m, 1 H), 3.02-3.13 (m, 1 H),
2.29-2.18 (m, 2 H), 2.08-1.97 (m, 2 H). Anal. (C29H28ClN3O5‚
11/2H2O) C, H, N.
(11a S)-8-{4-[(1R)-1-(Ch lor om eth yl)-5-h yd r oxy-1,2-d ih y-
d r o-3H -b e n zo[e]in d ol-3-yl]-4-oxob u t oxy}-7-m e t h oxy-
1,2,3,11a-tetr ah ydr o-5H-pyr r olo[2,1-c][1,4]ben zodiazepin -5-
on e (27cR) as a cream solid (81%); mp 220-221 °C (dec); [R]D
+626° (c 0.159, CHCl3); 1H NMR (CDCl3) δ 9.64 (s, 1 H), 8.30
(s, 1 H), 8.25 (d, J ) 8.4 Hz, 1 H), 7.62 (d, J ) 8.3 Hz, 1 H),
7.58 (d, J ) 4.3 Hz, 1 H), 7.50 (t, J ) 7.3 Hz, 1 H), 7.49 (s, 1
H), 7.35 (t, J ) 7.6 Hz, 1 H), 6.85 (s, 1 H), 4.33-4.17 (m, 4 H),
4.03-3.88 (m, 5 H), 3.86 (s, 3 H), 3.83-3.76 (m, 1 H), 3.59-
3.53 (m, 2 H), 3.38 (t, J ) 10.7 Hz, 1 H), 2.90-2.75 (m, 2 H),
2.45-2.37 (m, 2 H), 2.24-2.17 (m, 2 H), 2.05-1.95 (m, 2H);
13C NMR δ 172.0, 164.6, 162.4, 155.0, 150.5, 147.8, 141.2,
140.6, 129.9, 127.5, 123.9, 123.4, 122.7, 122.0, 120.3, 114.7,
111.6, 110.8, 100.7, 67.5, 56.1, 53.6, 53.5, 46.7, 46.4, 42.2, 32.5,
29.6, 24.2, 24.1; HRMS (FAB, 35Cl) calcd for C30H31ClN3O5
548.1952, found 548.1915. Anal. (C30H30ClN3O5‚H2O) C, H, N.
(11a S)-8-{4-[(1S)-1-(Ch lor om eth yl)-5-h yd r oxy-1,2-d ih y-
d r o-3H -b e n zo[e]in d ol-3-yl]-4-oxob u t oxy}-7-m e t h oxy-
1,2,3,11a-tetr ah ydr o-5H-pyr r olo[2,1-c][1,4]ben zodiazepin -
5-on e (27cS) as a cream solid (71%); mp 189-195 °C (dec);
1
[R]D +475° (c 0.153, CHCl3); H NMR (CDCl3) δ 9.16 (s, 1 H),
8.26 (d, J ) 8.3 Hz, 1 H), 8.21 (s, 1 H), 7.64 (d, J ) 8.4 Hz, 1
H), 7.61 (d, J ) 4.4 Hz, 1 H), 7.50 (t, J ) 7.8 Hz, 1 H), 7.48 (s,
1 H), 7.37 (t, J ) 7.7 Hz, 1 H), 6.87 (s, 1 H), 4.35-4.21 (m, 2
H), 4.05-3.98 (m, 1 H), 3.95-3.89 (m, 1 H), 3.85 (s, 3 H), 3.84-
3.77 (m, 1 H), 3.64-3.52 (m, 2 H), 3.38 (t, J ) 10.8 Hz, 1 H),
2.93-2.85 (m, 1 H), 2.82-2.73 (m, 1 H), 2.47-2.33 (m, 2 H),
2.28-2.21 (m, 2 H), 2.06-1.99 (m, 2H). Anal. (C30H30ClN3O5‚
1/2H2O) H, N; C: calcd, 64.69; found 64.24.
(11a S)-8-({5-[(1S)-1-(Ch lor om et h yl)-5-h yd r oxy-1,2-d i-
h yd r o-3H-b en zo[e]in d ol-3-yl]-5-oxop en t yl}oxy)-7-m et h -
oxy-1,2,3,11a -t e t r a h yd r o-5H -p yr r olo[2,1-c][1,4]b e n zo-
d ia zep in -5-on e (27d S) as a cream solid (80%); mp 178-186
°C (dec); [R]D +463° (c 0.133, CHCl3); 1H NMR (CDCl3): apart
from phenol signal (δ 9.86 instead of 9.72) identical (( 0.03
ppm) to that reported for 27Rd , 13C NMR identical (( 0.1 ppm)
to that reported for 27Rd . Anal. (C31H32ClN3O5‚11/2H2O) C, H,
N.
(11a S)-8-({6-[(1S)-1-(Ch lor om et h yl)-5-h yd r oxy-1,2-d i-
h ydr o-3H-ben zo[e]in dol-3-yl]-6-oxoh exyl}oxy)-7-m eth oxy-
1,2,3,11a-tetr ah ydr o-5H-pyr r olo[2,1-c][1,4]ben zodiazepin -
5-on e (27eS) as a cream solid (64%); mp 118-125 °C (dec);
[R]D +505° (c 0.111, CHCl3); 1H NMR (CDCl3): apart from
phenol signal (δ 9.83 instead of 9.59) identical (( 0.03 ppm)
to that reported for 27Re, 13C NMR (CDCl3) δ 172.6, 164.7,
162.3, 155.2, 150.9, 147.9, 141.3, 140.6, 130.0, 127.5, 123.9,
123.4, 122.8, 122.0, 120.1, 114.6, 111.6, 110.7, 100.7, 68.7, 56.1,
53.7, 53.5, 46.7, 46.3, 42.3, 36.1, 29.6, 28.7, 25.6, 24.4, 22.2.
Anal. (C32H34ClN3O5‚11/2H2O) C, H, N.
2,2,2-Tr ich lor oeth yl 6-{[(11a S)-7-Meth oxy-5,11-d ioxo-
2,3,5,10,11,11a -h exa h yd r o-1H -p yr r olo[2,1-c][1,4]b en zo-
d ia zep in -8-yl]oxy}h exa n oa te. A sample of 13e which had
been left to stand at 20 °C for 16 months appeared to be mostly
hydrolyzed to the corresponding acid (2S)-1-(2-{[(allyloxy)-
carbonyl]amino}-5-methoxy-4-{[6-oxo-6-(2,2,2-trichloroethoxy)-
hexyl]oxy}benzoyl)-2-pyrrolidinecarboxylic acid; 1H NMR
(CDCl3) δ 8.63 (br s, 1 H), 7.78 (br s, 1 H), 6.84 (s, 1 H), 5.99-
5.89 (m, 1 H), 5.34 (dq, J ) 17.1, 1.4 Hz, 1 H), 5.22 (dq, J )
10.3, 1.4 Hz, 1 H), 4.75 (s, 2 H), 4.64-4.60 (m, 2 H), 4.08 (t, J
) 6.7 Hz, 2 H), 3.82 (s, 3 H), 3.70-3.50 (m, 3 H), 2.51 (t, J )
7.5 Hz, 2 H), 2.40-2.20 (m, 2 H), 2.07-1.98 (m, 2 H), 1.94-
(11a S)-8-({5-[(1R)-1-(Ch lor om et h yl)-5-h yd r oxy-1,2-d i-
h yd r o-3H-ben zo[e]in d ol-3-yl]-5-oxop en t yl}oxy)-7-m et h -
oxy-1,2,3,11a -t e t r a h yd r o-5H -p yr r olo[2,1-c][1,4]b e n zo-
d ia zep in -5-on e (27d R) as a cream solid (75%); mp 170-178
°C (dec); [R]D +611° (c 0.164, CHCl3); 1H NMR (CDCl3) δ 9.72
(s, 1 H), 8.29 (s, 1 H), 8.26 (d, J ) 8.3 Hz, 1 H), 7.65 (d, J )
4.6 Hz, 1 H), 7.63 (d, J ca. 9.1 Hz, 1 H), 7.50 (s, 1 H), 7.49 (t,
J ca. 7.1 Hz, 1 H), 7.35 (t, J ) 7.6 Hz, 1 H), 6.84 (s, 1 H), 4.30
(dd, J ) 10.8, 1.5 Hz, 1 H), 4.26-4.10 (m, 3 H), 4.05-3.98 (m,
1 H), 3.95 (dd, J ) 11.2, 2.7 Hz, 1 H), 3.89 (s, 3 H), 3.85-3.78
(m, 1 H), 3.70-3.65 (m, 1 H), 3.62-3.53 (m, 1 H), 3.41 (t, J )
10.8 Hz, 1 H), 2.79-2.62 (m, 2 H), 2.33-2.26 (m, 2 H), 2.10-
1.99 (m, 6 H); 13C NMR δ 172.4, 164.7, 162.4, 155.1, 150.8,
147.9, 141.3, 140.6, 129.9, 127.5, 123.9, 123.4, 122.7, 122.0,
120.2, 114.6, 111.6, 110.6, 100.6, 68.7, 60.4, 56.1, 53.7, 53.5,
46.7, 46.3, 42.3, 35.8, 29.6, 28.4, 24.2, 21.5. Anal. (C31H32
-
ClN3O5‚11/2H2O) H, N; C: calcd, 63.21; found 63.64.
(11a S)-8-({6-[(1R)-1-(Ch lor om et h yl)-5-h yd r oxy-1,2-d i-
h ydr o-3H-ben zo[e]in dol-3-yl]-6-oxoh exyl}oxy)-7-m eth oxy-
1,2,3,11a-tetr ah ydr o-5H-pyr r olo[2,1-c][1,4]ben zodiazepin -
5-on e (27eR) as a cream solid (59%); mp 181-185 °C (dec);
1
[R]D +598° (c 0.134, CHCl3); H NMR (CDCl3) δ 9.59 (s, 1 H),
8.30 (s, 1 H), 8.28 (d, J ) 8.7 Hz, 1 H), 7.65 (d, J ) 8.9 Hz, 1
H), 7.64 (d, J ) 5.0 Hz, 1 H), 7.50 (s, 1 H), 7.49 (t, J ca. 8.1
Hz, 1 H), 7.35 (t, J ) 7.6 Hz, 1 H), 6.82 (s, 1 H), 4.31 (br d, J
) 10.7 Hz, 1 H), 4.26-4.20 (m, 1 H), 4.19-4.00 (m, 3 H), 3.95
(dd, J ) 11.3, 2.8 Hz, 1 H), 3.89 (s, 3 H), 3.86-3.78 (m, 1 H),
3.71-3.66 (m, 1 H), 3.62-3.54 (m, 1 H), 3.41 (t, J ) 10.8 Hz,
1 H), 2.72-2.54 (m, 2 H), 2.33-2.26 (m, 2 H), 2.09-1.88 (m, 6
H), 1.70-1.60 (m, 2 H). Anal. (C32H34ClN3O5‚H2O) C, H, N.
(11a S)-8-{2-[(1S)-1-(Ch lor om eth yl)-5-h yd r oxy-1,2-d ih y-
d r o-3H -b e n zo[e]in d ol-3-yl]-2-oxoe t h oxy}-7-m e t h oxy-

2148 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Tercel et al.
1.83 (m, 3 H), 1.78 (quint, J ) 7.6 Hz, 2 H), 1.60-1.50 (m, 2
H); MS (APCI -ve, 35Cl), 607 (100%, M - H+).
stirred under N2 at 20 °C for 90 min. The catalyst was filtered
off through Celite, and the filtrate was diluted with H2O and
extracted with EtOAc (×2). The extracts were dried (Na2SO4)
and evaporated, and the resulting oil was triturated with hot
petroleum ether to give 38 as a white powder (67 mg, 93%);
mp 175-177 °C; 1H NMR (CDCl3) δ 8.17 (d, J ) 8.1 Hz, 1 H),
7.79 (br s, 1 H), 7.67 (d, J ) 8.4 Hz, 1 H), 7.45 (t, J ) 7.5 Hz,
1 H), 7.31 (t, J ) 7.3 Hz, 1 H), 6.71 (br s, 1 H), 4.16 (dd, J )
11.0, 9.3 Hz, 1 H), 3.80-3.67 (m, 2 H), 1.59 (s, 9 H), 1.38 (d, J
) 6.8 Hz, 3 H); 13C NMR δ 153.1, 152.5, 139.6, 130.4, 126.9,
123.2, 122.6, 122.5, 121.4, 120.7, 99.3, 81.0, 57.0, 33.0, 28.5,
21.6. Anal. (C18H21NO3) C, H, N.
This acid (374 mg, 0.61 mmol) was dissolved in CH2Cl2 (10
mL) and pyrrolidine (0.26 mL, 3.1 mmol) and Pd(PPh3)4 (24
mg, 3%) were added. After stirring for 30 min at 20 °C, the
solvent was evaporated and the residue dissolved in and
evaporated from THF several times, until the residue was free
of the excess pyrrolidine. The residue was redissolved in THF
(10 mL), EDCI‚HCl (0.24 g, 1.2 mmol) was added, and the
mixture was stirred at 20 °C for 20 h. The THF was evaporated
and the residue partitioned between H2O and EtOAc. The
EtOAc layer and two further EtOAc extracts were dried (Na2-
SO4) and evaporated. The residue was purified by column
chromatography (EtOAc) to give the title compound as a white
solid (ca. 170 mg, ca. 55%), contaminated with a little coeluting
(11a S)-8-{[6-(5-Hyd r oxy-1-m eth yl-1,2-d ih yd r o-3H-ben -
zo[e]in d ol-3-yl)-6-oxoh exyl]oxy}-7-m eth oxy-1,2,3,11a -tet-
r a h yd r o-5H-p yr r olo[2,1-c][1,4]ben zod ia zep in -5-on e (39).
38 was deprotected with HCl and coupled with the PBD acid
15e using EDCI according to the general method described
above, to give allyl (11aS)-11-hydroxy-8-{[6-(5-hydroxy-1-
methyl-1,2-dihydro-3H-benzo[e]indol-3-yl)-6-oxohexyl]oxy}-7-
methoxy-5-oxo-2,3,11,11a-tetrahydro-1H-pyrrolo[2,1-c][1,4]-
benzodiazepine-10(5H)-carboxylate as a pale green oil (38%).
The Aloc group was directly deprotected using Pd(PPh3)4, as
described above, and the crude product was triturated with
EtOAc to give 39 as a pale tan powder (70%); mp 170-176
1
OPPh3; H NMR (CDCl3) δ 7.80 (s, 1 H), 7.44 (s, 1H), 6.43 (s,
1 H), 4.75 (s, 2 H), 4.08-4.06 (m, 1 H), 4.02 (t, J ) 6.5 Hz, 2
H), 3.90 (s, 3 H), 3.81-3.74 (m, 1 H), 3.65-3.57 (m, 1 H), 2.78-
2.69 (m, 1 H), 2.51 (t, J ) 7.4 Hz, 2 H), 2.09-1.98 (m, 3 H),
1.90 (quint, J ) 7.1 Hz, 2 H), 1.78 (quint, J ) 7.6 Hz, 2 H),
1.60-1.52 (m, 2 H); 13C NMR δ 171.8, 170.7, 165.2, 151.7,
146.7, 129.3, 119.4, 112.4, 104.9, 95.0, 73.9, 68.8, 56.8, 56.2,
47.3, 33.7, 28.5, 26.2, 25.4, 24.4, 23.5; HRMS (EI, 35Cl) calcd
for C21H25Cl3N2O6 506.0778, found 506.0772.
1
°C; H NMR (CDCl3) (several signals split 1:1 due to diaster-
6-{[(11aS)-7-Meth oxy-5,11-dioxo-2,3,5,10,11,11a-h exah y-
d r o -1H -p y r r o lo [2,1-c][1,4]b e n zo d ia ze p in -8-y l]o x y }-
h exa n oic Acid (34). The TCE ester was cleaved using Zn/
HCO2H according to the general method described above to
give 34 as a colorless oil (85%); [R]D +265° (c 0.145, CHCl3);
1H NMR (CDCl3) δ 9.45 (s, 1 H), 7.44 (s, 1 H), 6.50 (s, 1 H),
4.08-3.99 (m, 3 H), 3.89 (s, 3 H), 3.81-3.74 (m, 1 H), 3.64-
3.56 (m, 1 H), 2.77-2.68 (m, 1 H), 2.39 (t, J ) 7.3 Hz, 2 H),
2.07-1.96 (m, 3 H), 1.88 (quint, J ) 7.1 Hz, 2 H), 1.75-1.67
(m, 2 H), 1.57-1.49 (m, 2 H); 13C NMR δ 178.1, 171.8, 165.4,
151.8, 146.7, 129.5, 119.0, 112.3, 105.0, 68.8, 56.8, 56.2, 47.4,
33.8, 28.4, 26.1, 25.2, 24.3, 23.5; HRMS (EI) calcd for C19H24N2O6
376.1634, found 376.1628.
(11a S)-8-({6-[(1S)-1-(Ch lor om et h yl)-5-h yd r oxy-1,2-d i-
h ydr o-3H-ben zo[e]in dol-3-yl]-6-oxoh exyl}oxy)-7-m eth oxy-
2,3-d ih yd r o-1H -p yr r olo[2,1-c][1,4]b en zod ia zep in e-5,11-
(10H,11a H)-d ion e (35). The acid 34 was coupled with 6S
using the general method described above, and the product
triturated with EtOAc to give 35 as a pale gray powder (32%);
mp 166-170 °C (dec); [R]D +157° (c 0.151, THF); 1H NMR
(DMSO-d6) δ 10.33 (s, 1 H), 10.20 (s, 1 H), 8.08 (d, J ) 8.2 Hz,
1 H), 7.99 (s, 1 H), 7.77 (d, J ) 8.3 Hz, 1 H), 7.48 (t, J ) 7.5
Hz, 1 H), 7.31 (t, J ) 7.6 Hz, 1 H), 7.23 (s, 1 H), 6.71 (s, 1 H),
4.37-4.30 (m, 1 H), 4.19-4.10 (m, 2 H), 4.08-3.93 (m, 4 H),
3.81-3.74 (m, 1 H), 3.77 (s, 3 H), 3.57-3.51 (m, 1 H), 3.48-
3.41 (m, 1 H), 2.62-2.44 (m, 3 H), 1.98-1.65 (m, 7 H), 1.56-
1.48 (m, 2 H). Anal. (C32H34ClN3O6‚H2O) C, H, N.
omers) δ 9.05, 9.01 (2×s, 1 H), 8.25 (d, J ) 8.4 Hz, 1 H), 7.68
(d, J ) 8.3 Hz, 1 H), 7.64 (d, J ) 4.3 Hz, 1 H), 7.51 (s, 1 H),
7.45 (t, J ) 7.3 Hz, 1 H), 7.32 (t, J ) 7.3 Hz, 1 H), 6.84, 6.83
(2×s, 1 H), 4.33-4.27 (m, 1 H), 4.19-4.07 (m, 2 H), 3.91 (s, 3
H), 3.86-3.77 (m, 3 H), 3.71-3.66 (m, 1 H), 3.61-3.54 (m, 1
H), 2.65-2.48 (m, 2 H), 2.34-2.25 (m, 2 H), 2.09-1.86 (m, 6
H), 1.68-1.58 (m, 2 H), 1.41, 1.40 (2×d, J ) 6.7 Hz, 3 H). Anal.
(C32H35N3O5‚EtOAc) C, H, N.
In Vitr o Cytotoxicity. For the cell lines AA8, UV4, EMT6,
and Skov3 in Table 1, inhibition of proliferation of log-phase
monolayers was assessed in 96-well plates as described previ-
ously.51 The drug exposure time was 4 h followed by sulfor-
hodamine B staining 3 days later. The IC50 was determined
by interpolation as the drug concentration required to inhibit
cell density to 50% of that of the controls on the same plate.
For the cell lines Colo205, HT29, WiDr, Skov3, and SiHa
in Table 2, inhibition of [3H]thymidine uptake by cell suspen-
sions was assessed in 96-well plates as described previously.52
Cells were exposed to drugs continuously for 5 days, with
addition of [3H]thymidine (0.25 µCi/mL, 0.1 µM) and 5-fluoro-
2′-deoxyuridine (0.1 µM) 4 h prior to culture termination.
Pronase (2 mg/mL) in Na4EDTA (4 mM) (150 µL/well, 30-60
min treatment) was used to release any adherent cells from
the plate prior to aspiration and filtration. The IC50 was
determined by interpolation as the drug concentration required
to reduce [3H]thymidine uptake to 50% of that of the controls
on the same plate.
ter t-Bu tyl 5-(Ben zyloxy)-1-m eth yl-1,2-d ih yd r o-3H-ben -
zo[e]in d ole-3-ca r boxyla te. A solution of tert-butyl allyl[4-
(benzyloxy)-1-bromo-2-naphthyl]carbamate (37)50 (369 mg,
0.76 mmol), Bu3SnH (0.25 mL, 0.91 mmol), and AIBN (2,2′-
azobisisobutyronitrile) (10 mg, 8%) in dry PhH (25 mL) was
degassed by bubbling with N2, then stirred at reflux under N2
for 1 h. The solvent was evaporated and the residue diluted
with petroleum ether and extracted with CH3CN (×4). The
extracts were evaporated and the residue purified by column
chromatography (1:20 EtOAc:petroleum ether) to give the title
compound as a colorless oil (257 mg, 84%). A sample crystal-
lized from petroleum ether, mp 89-90 °C; 1H NMR (CDCl3) δ
8.28 (d, J ) 8.4 Hz, 1 H), 7.87 (br s, 1 H), 7.68 (d, J ) 8.3 Hz,
1 H), 7.55 (br d, J ) 7.0 Hz, 2 H), 7.50-7.28 (m, 5 H), 5.26 (s,
2 H), 4.18 (dd, J ) 11.0, 9.4 Hz, 1 H), 3.84-3.67 (m, 2 H), 1.60
(s, 9 H), 1.38 (d, J ) 6.9 Hz, 3 H); 13C NMR (two aromatic
quaternary carbons not seen) δ 155.0, 152.8, 137.1, 130.2,
128.5, 127.9, 127.5, 127.0, 123.4, 122.8, 122.4, 96.6, 80.6, 70.2,
56.9, 32.9, 28.5, 21.5. Anal. (C25H27NO3) C, H, N.
Cr oss-Lin k in g of Na k ed DNA. A mixture containing
pcDNA3 plasmid (Invitrogen, 5 µg, 5 µL), EcoRI restriction
endonuclease (5 µL, 50 units), Boehringer buffer H (5 µL), and
sterile H2O (17 µL) was incubated at 37 °C for 90 min to
linearize the DNA. The DNA was labeled at the 3′ end by
adding R-32P-dATP (2 µL, 20 µCi), Klenow polymerase (1 µL,
5 units), Boehringer buffer H (1 µL), and sterile H2O (8 µL)
and incubating the mixture at 37 °C for 30 min. The labeled
DNA was purified using a High Pure PCR purification kit
(Boehringer) and resuspended in Tris-EDTA (TE, pH 7.5)
buffer at 50 ng/µL (OD260 nm).
A mixture of the linearized labeled plasmid (1 µL, 50 ng)
and test drug (1 µL, various concentrations in DMSO, to give
the desired concentration in a final volume of 10 µL) in TE
buffer (8 µL) was incubated at 37 °C for 18 h and then diluted
with strand separation buffer (6 µL of DMSO, 2 µL of 0.5 M
NaOH, and 2 µL of H2O containing 0.25% bromophenol blue,
0.25% xylene cyanol, and 30% glycerol). The mixture was
allowed to stand at room temperature for 5-10 min and then
loaded directly onto a 0.8% agarose gel. Control experiments
showed that extending the alkali treatment time to 30 min
gave no change in the amount of cross-linked DNA. Electro-
ter t-Bu tyl 5-Hyd r oxy-1-m eth yl-1,2-d ih yd r o-3H-ben zo-
[e]in d ole-3-ca r boxyla te (38). A mixture of the benzyl ether
prepared above (94 mg, 0.24 mmol), aq NH4HCO2 (25%, 0.5
mL, 2.0 mmol), and Pd/C (10%, 25 mg) in THF (5 mL) was

Unsymmetrical DNA Cross-Linking Agents
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2149
phoresis was performed at 75 V for 4 h. The gels were dried,
and the bands were visualized and quantified by phosphorim-
agery, using Imagequant software. Results were expressed as
percentage double-stranded DNA of the total present for each
lane and averaged over 2-4 assays.
Control double-strand (ds) lanes were prepared in the
absence of alkali treatment, and control single-strand (ss) lanes
in the absence of drug. Control ds lanes were always free of
ss DNA, but occasionally (3 of 16 gels) the ss lanes contained
2-4% ds DNA, indicating incomplete denaturation. This was
not considered significant in calculating C50. Each gel con-
tained 3 as an internal standard, which gave good reproduc-
ibility on repeat assay.
work on the cross-linking of plasmid DNA, Dr. J ared
B. Milbank for assistance in the statistical analysis of
the cross-linking results, and Hui Hui Phua for as-
sistance with the comet assay. We acknowledge the
National Cancer Institute’s Developmental Therapeu-
tics Program for performing the in vitro 60 cell line
screen and COMPARE analysis. This work was carried
out with the support of the Auckland Division of the
Cancer Society of New Zealand and the Health Research
Council of New Zealand.
Su p p or tin g In for m a tion Ava ila ble: Synthesis of 3, 16,
30, 31, 32R, 32S; chiral HPLC of 16; optical rotation measure-
ments for 3; cross-linking of linearized plasmid DNA by 3 and
each CBI-PBD compound; relationship between cross-linking
ability and cytotoxicity in UV4, EMT6, and Skov3 cell lines;
cross-linking by 3 and 27cR after 2 and 18 h; COMPARE
analysis results; in vivo toxicity data; growth delay curves for
WiDr xenografts in nude mice treated with iv 27eS. This
material is available free of charge via the Internet at http://
pubs.acs.org.
The data points were fitted to a Logistic 3 parameter
sigmoidal curve using Sigmaplot software. The curve is defined
by the equation y ) a/[1 + (x/x0)b] where a is the maximum y
value (constrained to 100 in this case), b is a measure of the
slope, and x0 is the midpoint (equal to C50). In all cases the
data converged to this fitted curve and was tolerated. The
individual data points, fitted curves, and curve parameters are
included in the Supporting Information.
Cr oss-Lin k in g of Cellu la r DNA (Com et Assa y). This
was performed as previously described,72 except that Skov3
cell suspensions were used (105 cells/mL) under aerobic
conditions with a drug exposure time of 1 h. The cross-linking
index (CI) was calculated from the median tail moment (TM)
for each treatment as follows:
Refer en ces
(1) Rajski, S. R.; Williams, R. M. DNA Cross-Linking Agents as
Antitumor Drugs. Chem. Rev. 1998, 98, 8, 2723-2795.
(2) Woynarowski, J . M.; Napier, C.; Trevino, A. V.; Arnett, B.
Region-Specific DNA Damage by AT-Specific DNA-Reactive
Drugs is Predicted by Drug Binding Specificity. Biochemistry
2000, 39, 9917-9927.
(3) Woynarowski, J . M.; Trevino, A. V.; Rodriguez, K. A.; Hardies,
S. C.; Benham, C. J . AT-rich Islands in Genomic DNA as a Novel
Target for AT-Specific DNA-Reactive Antitumor Drugs. J . Biol.
Chem. 2001, 276, 40555-40566.
(4) Bando, T.; Iida, H.; Saito, I.; Sugiyama, H. Sequence-Specific
DNA Interstrand Cross-Linking by Imidazole-Pyrrole CPI Con-
jugate. J . Am. Chem. Soc. 2001, 123, 5158-5159.
(5) Thurston, D. E. Nucleic Acid Targeting: Therapeutic Strategies
for the 21st Century. Br. J . Cancer 1999, 80, 65-85.
(6) Boger, D. L.; J ohnson, D. S. CC-1065 and the Duocarmycins:
Understanding Their Biological Function Through Mechanistic
Studies. Angew. Chem., Int. Ed. Engl. 1996, 35, 1438-1474.
(7) Thurston, D. E.; Bose, D. S. Synthesis of DNA-Interactive
Pyrrolo[2,1-c][1,4]benzodiazepines. Chem. Rev. 1994, 94, 433-
465.
(8) Fukuda, Y.; Seto, S.; Furuta, H.; Ebisu, H.; Oomori, Y.; Terash-
ima, S. Cyclopropapyrroloindole (CPI) Bisalkylators. Drugs
Future 1998, 23, 1317-1330.
(9) Seco-CPI is the ring-opened form of the CPI (1,2,8,8a-tetrahydro-
7-methylcyclopropa[c]pyrrolo[3,2-e]indol-4-one) alkylation sub-
unit. In general the seco forms of cyclopropaindolones readily
ring close under culture conditions and have the same biological
properties as the cyclo parents.6
(10) Mitchell, M. A.; Kelly, R. C.; Wicnienski, N. A.; Hatzenbuhler,
N. T.; Williams, M. G.; Petzold, G. L.; Slightom, J . L.; Siemien-
iak, D. R. Synthesis and DNA Cross-Linking by a Rigid CPI
Dimer, J . Am. Chem. Soc. 1991, 113, 8994-8995.
(11) Cross-linking occurs between the adenine marked with an
asterisk and that opposite the underlined thymine.
(12) Sun, D.; Hurley, L. H. Analysis of the Monoalylation and Cross-
Linking Sequence Specificity of Bizelesin, a Bifunctional Alky-
lation Agent Related to (+)-CC-1065. J . Am. Chem. Soc. 1993,
115, 5925-5933.
(13) Carter, C. A.; Waud, W. R.; Li, L. H.; DeKoning, T. F.; McGovren,
J . P.; Plowman, J . Preclinical Antitumor Activity of Bizelesin
in Mice. Clin. Cancer Res. 1996, 2, 1143-1149.
TM(radiation only) - TM(drug + radiation)
CI )
TM(radiation only) - TM(control)
and the cross-linking index of 50% (CI50) was derived by
interpolation from Figure 5.
Th er m a l Clea va ge Assa y. This was performed as previ-
ously described,31c except that the 5′-32P-end labeled DNA was
a 512 bp fragment of the lac UV5 promoter. To generate this
fragment the TC1 primer (5′-ctggcacgacaggtttcc-3′) was end-
labeled using T4 PNK and PCR was performed using this,
unlabeled TC3 primer (5′-tgggctgcaggtcgacgctct-3′), and the
plasmid pCC1.73 Alkylation of the labeled DNA and thermal
cleavage followed the method described,31c with sequencing
lanes generated using the fmol DNA sequencing system
(Promega) according to manufacturer’s instructions, with
primer TC1 and plasmid pCC1.
Toxicity a n d An titu m or Activity in Mice. All compounds
were formulated in 100% DMSO and administered as single
doses at 0.001 mL/g of body weight. At this dose level no
toxicity was observed in vehicle only treated controls via ip or
iv routes. The toxicity in nontumor-bearing female C3H/HeN
mice was determined using dose increments of 1.33-fold on a
fixed scale with a 60 day observation time. The MTD was
defined as the highest dose causing no deaths in a group of 6
mice and e15% mean body weight loss on day 4.
For the growth delay assay, WiDr cells (107 cells/mouse in
100 µL) were injected subcutaneously into the flank of CD-1
nude female mice. Mice were randomized to treatment groups
(5-11 mice/group) when mean tumor diameter reached 8 mm
and received a single dose either ip or iv at the predetermined
MTD (day 0). Tumor size was measured periodically thereafter
using calipers and mice were culled once the mean tumor
diameter reached 17 mm. The statistical significance of drug
effects was evaluated by ANOVA followed by Dunnett’s test
to determine p values for individual groups.
(14) Pitot, H. C.; Reid, J . M.; Sloan, J . A.; Ames, M. M.; Adjei, A. A.;
Rubin, J .; Bagniewski, P. G.; Atherton, P.; Rayson, D.; Goldberg,
R. M.; Erlichman, C. A Phase I Study of Bizelesin (NSC 615291)
in Patients with Advanced Solid Tumors. Clin. Cancer Res. 2002,
8, 712-717.
Ack n ow led gm en t. We would like to thank Prof.
David E. Thurston for the gift of some 3-({(11aS)-10-
[(9H-fluoren-9-ylmethoxy)carbonyl]-11-hydroxy-7-meth-
oxy-5-oxo-2,3,5,10,11,11a-hexahydro-1H-pyrrolo[2,1-c]-
[1,4]benzodiazepin-8-yl}oxy)propanoic acid (the FMOC-
protected analogue of 15c). We would also like to thank
Dr Swarna A. Gamage for the synthesis of 6, Karin Tan
for the chiral HPLC resolution of this material, Drs.
Michael A. Gieseg and Charley de Bock for preliminary
(15) Volpe, D. A.; Tomaszewski, J . E.; Parchment, R. E.; Garg, A.;
Flora, K. P.; Murphy, M. J .; Grieshaber, C. K. Myelotoxic Effects
of the Bifunctional Alkylating Agent Bizelesin on Human,
Canine and Murine Myeloid Progenitor Cells. Cancer Chemother.
Pharmacol. 1996, 39, 143-149.
(16) (a) Markovic, S. N.; Suman, V. J .; Vukov, A. M.; Fitch, T. R.;
Hillman, D. W.; Adjei, A. A.; Alberts, S. R.; Kaur, J . S.; Braich,
T. A.; Leitch, J . M.; Creagan, E. T. Phase II Trial of KW2189 in
Patients With Advanced Malignant Melanoma. Am. J . Clin.
Oncol. 2002, 25, 308-312. (b) Pavlidis, N.; Aamdal, S.; Awada,
A.; Calvert, H.; Fumoleau, P.; Sorio, R.; Punt, C.; Verweij, J .;
van Oosterom, A.; Morant, R.; Wanders, J .; Hanauske, A-R.

2150 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Tercel et al.
Carzelesin Pase II Study in Advanced Breast, Ovarian, Colorec-
tal, Gastric, Head and Neck Cancer, Non-Hodgkin’s Lymphoma
and Malignant Melanoma: a Study of the EORTC Early Clinical
Studies Group (ECSG). Cancer Chemother. Pharmacol. 2000, 46,
167-171. (c) Cristofanilli, M.; Bryan, W. J .; Miller, L. L.; Chang,
A. Y-C.; Gradishar, W. J .; Kufe, D. W.; Hortobagyi, G. N. Phase
II Study of Adozelesin in Untreated Metastatic Breast Cancer.
Anti-Cancer Drugs 1998, 9, 779-782.
Synthesis of CC-1065 and Duocarmycin Analogues via Intramo-
lecular Aryl Radical Cyclization of a Tethered Vinyl Chloride.
Tetrahedron Lett. 1998, 39, 2227-2230. (c) Gieseg, M. A.;
Matejovic J .; Denny, W. A. Comparison of the Patterns of DNA
Alkylation by Phenol and Amino seco-CBI-TMI Compounds: Use
of a PCR Method for the Facile Preparation of Single End-
Labeled Double-Stranded DNA. Anti-Cancer Drug Design 1999,
14, 77-84.
(17) Mitchell, M. A.; J ohnson, P. D.; Williams, M. G.; Aristoff, P. A.
Interstrand DNA Cross-Linking with Dimers of the Spirocyclo-
propyl Alkylating Moiety of CC-1065. J . Am. Chem. Soc. 1989,
111, 6428-6429.
(18) (a) Park, H-J .; Kelly, R. C.; Hurley, L. H. The Chemical Evolution
of DNA-DNA Interstrand Cross-Linkers That Recognize Mixed
AT and GC Sequences. J . Am. Chem. Soc. 1996, 118, 10041-
10051. (b) Fukuda, Y.; Seto, S.; Furuta, H.; Ebisu, H.; Oomori,
Y.; Terashima, S. Novel Seco Cyclopropa[c]pyrrolo[3,2-e]indole
(32) Petrusek, R. L.; Anderson, G. L.; Garner, T. F.; Fannin, Q. L.;
Kaplan, D. J .; Zimmer, S. G.; Hurley, L. H. Pyrrolo[1,4]-
benzodiazepine Antibiotics. Proposed Structures and Charac-
teristics of the In Vitro Deoxyribonucleic Acid Adducts of
Anthramycin, Tomaymycin, Sibiromycin, and Neothramycins A
and B. Biochemistry 1981, 20, 1111-1119.
(33) Althuis, T. H.; Hess, H-J . Synthesis and Identification of the
Major Metabolites of Prazosin Formed in Dog and Rat. J . Med.
Chem. 1977, 20, 146-149.
(34) (a) Sagnou, M. J .; Howard, P. H.; Gregson, S. J .; Eno-Amooquaye,
E.; Burke, P. J .; Thurston, D. E. Design and Synthesis of Novel
Pyrrolobenzodiazepine (PBD) Prodrugs for ADEPT and GDEPT.
Bioorg. Med. Chem. Lett. 2000, 10, 2083-2086. (b) Thurston,
D. E.; Howard, P. W. Preparation of Pyrrolobenzodiazepines
(PBDs) as Antitumor Antibiotics. PCT Int. Appl. WO 0012507,
2000.
(35) Wright, I. G.; Ashbrook, C. W.; Goodson, T.; Kaiser, G. V.; Van
Heyningen, E. M. Chemistry of Cephalosporin Antibiotics. 23.
2-Methyl- and 2-methylenecephalosporins. J . Med. Chem. 1971,
14, 420-426.
(36) J ust, G.; Grozinger, K. A Selective, Mild Cleavage of Trichloro-
ethyl Esters, Carbamates, and Carbonates to Carboxylic Acids,
Amines, and Phenols using Zinc/Tetrahydrofuran/pH4.2-7.2
Buffer. Synthesis 1976, 457-459.
(37) J ou, G.; Gonzalez, I.; Albericio F.; Lloyd-Williams, P.; Giralt, E.
Total Synthesis of Dehydrodidemnin B. Use of Uronium and
Phosphonium Salt Coupling Reagents in Peptide Synthesis in
Solution. J . Org. Chem. 1997, 62, 354-366.
(38) (a) Baraldi, P. G.; Balboni, G.; Cacciari, B.; Guiotto, A.; Man-
fredini, S.; Romagnoli, R.; Spalluto, G.; Thurston, D. E.; Howard,
P. H.; Bianchi, N.; Rutigliano, C.; Mischiati, C.; Gambari, R.
Synthesis, In Vitro Antiproliferative Activity, and DNA-Binding
Properties of Hybrid Molecules Containing Pyrrolo[2,1-c][1,4]-
benzodiazepine and Minor-Groove-Binding Oligopyrrole Carri-
ers. J . Med. Chem. 1999, 42, 5131-5141. (b) Thurston, D. E.;
Bose, D. S.; Howard, P. H.; J enkins, T. C.; Leoni, A.; Baraldi, P.
G.; Guiotto, A.; Cacciari, B.; Kelland, L. R.; Foloppe, M-P.;
Rault, S. Effect of A-Ring Modifications on the DNA-Binding
Behavior and Cytotoxicity of Pyrrolo[2,1-c][1,4]benzodiazepines.
J . Med. Chem. 1999, 42, 1951-1964.
(39) Dalcanale, E.; Montanari, F. Selective Oxidation of Aldehydes
to Carboxylic Acids with Sodium Chlorite-Hydrogen Peroxide.
J . Org. Chem. 1986, 51, 567-569.
(40) Seltzman, H. H.; Berrang, B. D. Nickel Boride Reduction of Aryl
Nitro Compounds. Tetrahedron Lett. 1993, 34, 3083-3086.
(41) Kunz, H.; Unverzagt, C. Protective Group Dependent Stability
of Intersaccharide Bond. Synthesis of Fucosylchitobiose-Glu-
copeptides. Angew. Chem. 1988, 100, 1763-1765.
(42) Thurston, D. E.; Murty, V. S.; Langley, D. R.; J ones, G. B.
O-Debenzylation of a Pyrrolo[2,1-c][1,4]benzodiazepine in the
Presence of a Carbinolamine Functionality: Synthesis of DC-
81. Synthesis 1990, 81-84.
(43) Molina, P.; Diaz, I.; Tarraga, A. Synthesis of Pyrrolo[2,1-c][1,4]-
benzodiazepines via an Intramolecular Aza-Wittig Reaction.
Synthesis of the Antibiotic DC-81. Tetrahedron 1995, 51, 5617-
5630.
(44) Bose, D. S.; J ones, G. B.; Thurston, D. E. New Approaches to
Pyrrolo[2,1-c][1,4]benzodiazepines: Synthesis, DNA-Binding and
Cytotoxicity of DC-81. Tetrahedron 1992, 48, 751-758.
(45) These [RD] values in turn are considerably higher than that
measured for the natural product itself, [RD] ) +135°, but this
has been attributed to likely epimerization during the isolation
steps.42
(46) The recorded value in 2% EtOH/CHCl3 was +567°, but this does
not necessarily represent the equilibrium situation. For further
discussion see Supporting Information.
(47) Thurston, D. E.; Bose, D. S.; Thompson, A. S.; Howard, P. H.;
Leoni, A.; Croker, S. J .; J enkins, T. C.; Neidle, S.; Hartley, J .
A.; Hurley, L. H. Synthesis of Sequence-Selective C8-Linked
Pyrrolo[2,1-c][1,4]benzodiazepine DNA Interstrand Cross-Link-
ing Agents. J . Org. Chem. 1996, 61, 8141-8147.
(48) J ones, G. B.; Davey, C. L.; J enkins, T. C.; Kamal, A.; Kneale, G.
G.; Neidle, S.; Webster, G. D.; Thurston, D. E. The Non-Covalent
Interaction of Pyrrolo[2,1-c][1,4]benzodiazepine-5,11-diones with
DNA. Anti-Cancer Drug Design 1990, 5, 249-264.
Bisalkylators Bearing a 3,3′-Arylenebisacryloyl Group as
a
Linker. J . Med. Chem. 2001, 44, 1396-1406. (c) Boger, D. L.;
Searcey, M.; Tse, W. C.; J in, Q. Bifunctional Alkylating Agents
Derived from Duocarmycin SA: Potent Antitumor Activity with
Altered Sequence Selectivity. Bioorg. Med. Chem. Lett. 2000, 10,
495-498.
(19) J ia, G.; Lown, J . W. Design, Synthesis and Cytotoxicity Evalu-
ation of 1-Chloromethyl-5-hydroxy-1,2-dihydro-3H-benz[e]indole
(seco-CBI) Dimers. Bioorg. Med. Chem. 2000, 8, 1607-1617.
(20) Boger, D. L.; Ishizaki, T.; Kitos, P. A.; Suntornwat, O. Synthesis
of N-(tert-Butyloxycarbonyl)-CBI, CBI, CBI-CDPI1, and CBI-
CDPI2: Enhanced Functional Analogues of CC-1065 Incorporat-
ing the 1,2,9,9a-Tetrahydrocyclopropa[c]benz[e]indol-4-one (CBI)
Left-Hand Subunit. J . Org. Chem. 1990, 55, 5823-5832.
(21) Bose, D. S.; Thompson, A. S.; Ching, J .; Hartley, J . A.; Berardini,
M. D.; J enkins, T. C.; Neidle, S.; Hurley, L. H.; Thurston, D. E.
Rational Design of a Highly Efficient Irreversible DNA Inter-
strand Cross-Linking Agent Based on the Pyrrolobenzodiazepine
Ring System. J . Am. Chem. Soc. 1992, 114, 4939-4941.
(22) Bose, D. S.; Thompson, A. S.; Smellie, M.; Berardini, M. D.;
Hartley, J . A.; J enkins, T. C.; Neidle, S.; Thurston, D. E. Effect
of Linker Length on DNA-Binding Affinity, Cross-Linking
Efficiency and Cytotoxicity of C8-Linked Pyrrolobenzodiazepine
Dimers. J . Chem. Soc., Chem. Commun. 1992, 1518-1520.
(23) J enkins, T. C.; Hurley, L. H.; Neidle, S.; Thurston, D. E.
Structure of a Covalent Minor Groove Adduct with
a Pyr-
rolobenzodiazepine Dimer: Evidence for Sequence-Specific In-
terstrand Cross-Linking. J . Med. Chem. 1994, 37, 4529-4537.
(24) Walton, M. I.; Goddard, P.; Kelland, L. R.; Thurston, D. E.;
Harrap, K. R. Preclinical Pharmacology and Antitumor Activity
of the Novel Sequence-Selective DNA Minor-Groove Cross-
Linking Agent DSB-120. Cancer Chemother. Pharmacol. 1996,
38, 431-438.
(25) Gregson, S. J .; Howard, P. W.; Hartley, J . A.; Brooks, N. A.;
Adams, L. J .; J enkins, T. C.; Kelland, L. R.; Thurston, D. E.
Design, Synthesis, and Evaluation of a Novel Pyrrolobenzodi-
azepine DNA-Interactive Agent with Highly Efficient Cross-
Linking Ability and Potent Cytotoxicity. J . Med. Chem. 2001,
44, 737-748.
(26) Hartley, J . A.; Spanswick, V. J .; Pedley, B.; Hollingshead, M.
G.; Alley, M. C.; Schultz, R.; Gregson, S. J .; Howard, P. H.;
Thurston, D. E. In Vitro Antitumor Activity and In Vivo DNA
Interstrand Cross-linking by the Novel Pyrrolobenzodiazepine
Dimer SJ G-136 (NSC 694501). Proceedings of the American
Association for Cancer Research; 93rd Annual Meeting, San
Francisco, April 6-10, 2002, Vol. 43, Abstract 2432 (http://
www.aacr.org).
(27) Zhou, Q.; Duan, W.; Simmons, D.; Shayo, Y.; Raymond, M. A.;
Dorr, R. T.; Hurley, L. H. Design and Synthesis of a Novel DNA-
DNA Interstrand Adenine-Guanine Cross-Linking Agent. J . Am.
Chem. Soc. 2001, 123, 4865-4866.
(28) Hurley, L. H.; Warpehoski, M. A.; Lee, C-S.; McGovren, J . P.;
Scahill, T. A.; Kelly, R. C.; Mitchell, M. A.; Wicnienski, N. A.;
Gebhard, I.; J ohnson, P. D.; Bradford, V. S. Sequence Specificity
of DNA Alkylation by the Unnatural Enantiomer of CC-1065
and Its Synthetic Ananlogues. J . Am. Chem. Soc. 1990, 112,
4633-4649.
(29) Boger, D. L.; Yun, W. CBI-TMI: Synthesis and Evaluation of a
Key Analogue of the Duocarmycins. Validation of a Direct
Relationship between Chemical Solvolytic Stability and Cyto-
toxic Potency and Confirmation of the Structural Features
Responsible for the Distinguishing Behavior of Enantiomeric
Pairs of Agents. J . Am. Chem. Soc. 1994, 116, 7996-8006.
(30) Fukuyama, T.; Liu, G.; Linton, S. D.; Lin, S-C.; Nishino, H. Total
Synthesis of (+)-Porothramycin B. Tetrahedron Lett. 1993, 34,
2577-2580.
(31) (a) Boger, D. L.; McKie, J . A. An Efficient Synthesis of 1,2,9,9a-
Tetrahydrocyclopropa[c]benz[e]indol-4-one (CBI): An Enhanced
and Simplified Analogue of the CC-1065 and Duocarmycin
Alkylation Subunits. J . Org. Chem. 1995, 60, 1271-1275. (b)
Boger, D. L.; Boyce, C. W.; Garbaccio, R. M.; Searcey, M.
(49) 13e was found to have almost completely hydrolyzed on storage,
so following Aloc deprotection, an extra EDCI-mediated dehy-
dration step was used to reform the N10/C11 amide bond.

Unsymmetrical DNA Cross-Linking Agents
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2151
(50) Boger, D. L.; Mesini, P. Design, Synthesis and Evaluation of CC-
1065 and Duocarmycin Analogues Incorporating the 2,3,10,10a-
Tetrahydro-1H-cyclopropa[d]benzo[f]quinol-5-one (CBQ) Alky-
lation Subunit: Identification and Structural Origin of Subtle
Stereoelectronic Features That Govern Reactivity and Regio-
selectivity. J . Am. Chem. Soc. 1994, 116, 11335-11348.
(51) Wilson, W. R.; Denny, W. A.; Pullen, S. M.; Thompson, K. M.;
Li, A. E.; Patterson, L. H.; Lee, H. H. Tertiary Amine N-Oxides
as Bioreductive Drugs: DACA N-Oxide, Nitracrine N-Oxide
and AQ4N. Br. J . Cancer 1996, 74 (Suppl. XXVII), S43-S47.
(52) Marshall, E. S.; Finlay, G. J .; Matthews, J . H. L.; Shaw, J . H.
F.; Nixon, J .; Baguley, B. C. Microculture-Based Chemosensi-
tivity Testing: A Feasibility Study Comparing Freshly Ex-
planted Human Melanoma Cells With Human Melanoma Cell
Lines. J . Natl. Cancer Inst. 1992, 84, 340-345.
(53) Hoeijmakers, J . H. J .; Bootsma, D. Molecular Genetics of
Eukaryotic DNA Excision Repair. Cancer Cells 1990, 2, 311-
320.
(54) Hoy, C. A.; Salazar, E. P.; Thompson, L. H. Rapid Detection of
DNA-Damaging Agents Using Repair-Deficient CHO Cells.
Mutat. Res. 1984, 130, 321-332.
(55) Boger, D. L.; Garbaccio, R. M. Shape-Dependent Catalysis:
Insights into the Source of Catalysis for the CC-1065 and
Duocarmycin DNA Alkylation Reaction. Acc. Chem. Res. 1999,
32, 1043-1052.
(56) (a) 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
Design 1997, 12, 533-541. (b) Monks, A.; Scudiero, D.; Skehan,
P.; Shoemaker, R.; Paull, K.; Vistica, D.; Hose, C.; Langley, J .;
Cronise, P.; Vaigro-Wolff, A.; Gray-Goodrich, M.; Campbell, H.;
Mayo, J .; Boyd, M. Feasibility of a High-Flux Anticancer Drug
Screen Using a Diverse Panel of Cultured Human Tumor Cell
Lines. J . Natl. Cancer Inst. 1991, 83, 757-766.
(57) Paull, K. D.; Shoemaker, R. H.; Hodes, L.; Monks, A.; Scudiero,
D. A.; Rubinstein, L.; Plowman, J .; Boyd, M. R. Display and
Analysis of Patterns of Differential Activity of Drugs Against
Human Tumor Cell Lines: Development of Mean Graph and
COMPARE Algorithm. J . Natl. Cancer Inst. 1989, 81, 1088-
1092.
(58) Hartley, J . A.; Berardini, M. D.; Souhami, R. L. An Agarose Gel
Method for the Determination of DNA Interstrand Cross-linking
Applicable to the Measurement of the Rate of Total and “Second-
Arm” Cross-link Reactions. Anal. Biochem. 1991, 193, 131-134.
(59) Lee, C-S.; Gibson, N. W. DNA Damage and Differential Cyto-
toxicity Produced in Human Carcinoma Cells by CC-1065
Analogues, U-73,975 and U-77,779. Cancer Res. 1991, 51, 6586-
6591.
(61) (a) Warpehoski, M. A.; Harper, D. E.; Mitchell, M. A.; Monroe,
T. J .Reversibility of the Covalent Reaction of CC-1065 and
Analogues with DNA. Biochemistry 1992, 31, 2502-2508. (b)
Boger, D. L.; Yun, W. Reversibility of the Duocarmycin A and
SA DNA Alkylation Reaction. J . Am. Chem. Soc. 1993, 115,
9872-9873.
(62) See Supporting Information for graphs of the individual data
points and fitted curves for each of the compounds.
(63) A C50 value for 3 has been reported (0.055 µM).22 Although
similar to the value found here the results are not directly
comparable since a different DNA sequence (pBR322 plasmid,
4400 bp) and exposure time (2 h) were used.
(64) Fairbairn, D. W.; Olive, P. L.; O’Neill, K. L. The Comet Assay:
a Comprehensive Review. Mutat. Res. 1995, 339, 37-59.
(65) Boger, D. L.; Munk, S. A.; Zarrinmayeh, H.; Ishizaki, T.; Haught,
J .; Bina, M. An Alternative and Convenient Strategy for
Generation of Substantial Quantities of Singly 5′-32P-End-
Labeled Double-Stranded DNA for Binding Studies: Develop-
ment of a Protocol for Examination of Functional Features of
(+)-CC-1065 and the Duocarmycins that Contribute to their
Sequence-Selective DNA Alkylation Properties. Tetrahedron
1991, 47, 2661-2682.
(66) Smellie, M.; Kelland, L. R.; Thurston, D. E.; Souhami, R. L.;
Hartley, J . A. Cellular Pharmacology of Novel C8-Linked An-
thramycin-Based Sequence-Selective DNA Minor Groove Cross-
Linking Agents. Br. J . Cancer 1994, 70, 48-53.
(67) Paull, K. D.; Hamel, E.; Malspeis, L. Prediction of Biochemical
Mechanism of Action from the In Vitro Antitumor Screen of the
National Cancer Institute. In Cancer Chemotherapeutic Agents,
Foye, W. O., Ed.; American Chemical Society: Washington, DC,
1995; pp 9-45.
(68) http://dtp.nci.nih.gov/docs/cancer/searches/cancer_open_compound-
s.html, bizelesin (NSC 615291) at log(high concentration) -5.00,
LC50 comparison.
(69) Seuring, B.; Seebach, D. Syntheses and Determinations of the
Absolute Configurations of Norpyrenophorin, Pyrenophorin, and
Vermiculine. J ustus Liebigs Ann. Chem. 1978, 12, 2044-2073.
(70) Trost, B. M.; Verhoeven, T. R. Cyclization Catalyzed by Pal-
ladium(0). Initial Studies and Macrolide Formation. J . Am.
Chem. Soc. 1980, 102, 4743-4763.
(71) Alexander, J .; Renyer, M. L.; Veerapanane, H. A Convenient
Method for the Conversion of Halides to Alcohols. Synth.
Commun. 1995, 25, 3875-3881.
(72) Tercel, M.; Lee, A. E.; Hogg, A.; Anderson, R. F.; Lee, H. H.;
Siim, B. G.; Denny, W. A.; Wilson, W. R. Hypoxia-Selective
Antitumor Agents. 16. Nitroarylmethyl Quaternary Salts as
Bioreductive Prodrugs of the Alkylating Agent Mechlorethamine.
J . Med. Chem. 2001, 44, 3511-3522.
(73) White, R. J .; Phillips, D. R. Transcriptional Analysis of Multisite
Drug-DNA Dissociation Kinetics: Delayed Termination of
Transcription by Actinomycin D. Biochemistry 1988, 27, 9122-
9132.
(60) Reynolds, V. L.; Molineux, I. J .; Kaplan, D. J .; Swenson, D. H.;
Hurley, L. H. Reaction of the Antitumor Antibiotic CC-1065 with
DNA. Location of the Site of Thermally Induced Strand Break-
age and Analysis of DNA Sequence Specificity. Biochemistry
1985, 24, 6228-6237.
J M020526P