Synthesis and evaluation of 1,2,8,8a-tetrahydrocyclopropa[c]pyrrolo[3,2-e]indol-4(5H)-one, the parent alkylation subunit of CC-1065 and the duocarmycins: Impact of the alkylation subunit substituents and its implications for DNA alkylation catalysis

Article abstract of DOI:10.1021/jo000297j

The synthesis of 1,2,8,8a-tetrahydrocyclopropa[c]pyrrolo[3,2-e]indol-4(5H)-one (CPI), the parent CC-1065 and duocarmycin SA alkylation subunit, is detailed. The parent CPI alkylation subunit lacks the C7 methyl substituent of the CC-1065 alkylation subunit and the C6 methoxycarbonyl group of duocarmycin SA, and their examination permitted the establishment of the impact of these natural product substituents. The studies revealed a CPI stability comparable to the CC-1065 alkylation subunit but which was 6x more reactive than the (+)-duocarmycin SA alkylation subunit, and it displayed the inherent reaction regioselectivity (4:1) of the natural products. The single-crystal X-ray structure of (+)-N-BOC-CPI depicts a near identical stereoelectronic alignment of the cyclopropane accounting for the identical reaction regioselectivity and a slightly diminished vinylogous amide conjugation relative to (+)-N-BOC-DSA suggesting that the stability distinctions stem in part from this difference in the vinylogous amide as well as alterations in the electronic nature of the fused pyrrole. Establishment of the DNA binding properties revealed that the CPI-based agents retain the identical DNA alkylation selectivities of the natural products. More importantly, the C6 methoxycarbonyl group of duocarmycin SA was found to increase the rate (12-13x) and efficiency (10x) of DNA alkylation despite its intrinsic lower reactivity while the CC-1065 C7 methyl group was found to slow the DNA alkylation rate (4x) and lower the alkylation efficiency (ca. 4x). The greater DNA alkylation rate and efficiency for duocarmycin SA and related analogues containing the C6 methoxycarbonyl is proposed to be derived from the extended length that the rigid C6 methoxycarbonyl provides and the resulting increase in the DNA binding-induced conformational change which serves to deconjugate the vinylogous amide and activate the alkylation subunit for nucleophilic attack. The diminished properties resulting from the CC-1065 C7 methyl group may be attributed to the steric impediment this substituent introduces to DNA minor groove binding and alkylation. Consistent with this behavior, the duocarmycin SA C6 methoxycarbonyl group increases biological potency while the CC-1065 C7 methyl group diminishes it.

Full text of DOI:10.1021/jo000297j

J . Org. Chem. 2000, 65, 4101-4111
4101
Syn th esis a n d Eva lu a tion of
1,2,8,8a -Tetr a h yd r ocyclop r op a [c]p yr r olo[3,2-e]in d ol-4(5H)-on e, th e
P a r en t Alk yla tion Su bu n it of CC-1065 a n d th e Du oca r m ycin s:
Im p a ct of th e Alk yla tion Su bu n it Su bstitu en ts a n d Its
Im p lica tion s for DNA Alk yla tion Ca ta lysis
Dale L. Boger,* Alejandro Santilla´n, J r., Mark Searcey, Steven R. Brunette,
Scott E. Wolkenberg, Michael P. Hedrick, and Qing J in
Department of Chemistry and The Skaggs Institute for Chemical Biology, The Scripps Research Institute,
10550 North Torrey Pines Road, La J olla, California 92037
Received March 3, 2000
The synthesis of 1,2,8,8a-tetrahydrocyclopropa[c]pyrrolo[3,2-e]indol-4(5H)-one (CPI), the parent CC-
1065 and duocarmycin SA alkylation subunit, is detailed. The parent CPI alkylation subunit lacks
the C7 methyl substituent of the CC-1065 alkylation subunit and the C6 methoxycarbonyl group
of duocarmycin SA, and their examination permitted the establishment of the impact of these
natural product substituents. The studies revealed a CPI stability comparable to the CC-1065
alkylation subunit but which was 6× more reactive than the (+)-duocarmycin SA alkylation subunit,
and it displayed the inherent reaction regioselectivity (4:1) of the natural products. The single-
crystal X-ray structure of (+)-N-BOC-CPI depicts a near identical stereoelectronic alignment of
the cyclopropane accounting for the identical reaction regioselectivity and a slightly diminished
vinylogous amide conjugation relative to (+)-N-BOC-DSA suggesting that the stability distinctions
stem in part from this difference in the vinylogous amide as well as alterations in the electronic
nature of the fused pyrrole. Establishment of the DNA binding properties revealed that the CPI-
based agents retain the identical DNA alkylation selectivities of the natural products. More
importantly, the C6 methoxycarbonyl group of duocarmycin SA was found to increase the rate
(12-13×) and efficiency (10×) of DNA alkylation despite its intrinsic lower reactivity while the
CC-1065 C7 methyl group was found to slow the DNA alkylation rate (4×) and lower the alkylation
efficiency (ca. 4×). The greater DNA alkylation rate and efficiency for duocarmycin SA and related
analogues containing the C6 methoxycarbonyl is proposed to be derived from the extended length
that the rigid C6 methoxycarbonyl provides and the resulting increase in the DNA binding-induced
conformational change which serves to deconjugate the vinylogous amide and activate the alkylation
subunit for nucleophilic attack. The diminished properties resulting from the CC-1065 C7 methyl
group may be attributed to the steric impediment this substituent introduces to DNA minor groove
binding and alkylation. Consistent with this behavior, the duocarmycin SA C6 methoxycarbonyl
group increases biological potency while the CC-1065 C7 methyl group diminishes it.
In tr od u ction . CC-1065 and the duocarmycins form
the parent members of a class of potent antitumor
antibiotics that selectively bind and alkylate DNA (Figure
1).1 The stereoelectronically controlled adenine N3 ad-
dition to the least substituted carbon of the activated
cyclopropane occurs within selected minor groove AT-rich
sites, and extensive effort has been devoted to determine
the origin of the DNA alkylation selectivity, to establish
the link between DNA alkylation and the ensuing
biological properties, and to define the fundamental
principles underlying the relationships between struc-
ture, chemical reactivity, and biological activity.2-5
Key to recent studies was the identification of a
conformational change induced upon binding in the
narrow AT-rich minor groove, which leads to disruption
of the vinylogous amide stabilization of the alkylation
subunit and activation of the cyclopropane to nucleophilic
attack.6 This subtle element of the DNA alkylation
mechanism was initially recognized through the synthe-
(1) Ichimura, M.; Ogawa, T.; Takahashi, K.; Kobayashi, E.; Kawa-
moto, I.; Yasuzawa, T.; Takahashi, I.; Nakano, H. J . Antibiot. 1990,
43, 1037. Ichimura, M.; Ogawa, T.; Katsumata, S.; Takahashi, K.;
Takahashi, I.; Nakano, H. J . Antibiot. 1991, 44, 1045. Ohba, K.;
Watabe, H.; Sasaki, T.; Takeuchi, Y.; Kodama, Y.; Nakazawa, T.;
Yamamoto, H.; Shomura, T.; Sezaki, M.; Kondo, S. J . Antibiot. 1988,
41, 1515. Takahashi, I.; Takahashi, K.; Ichimura, M.; Morimoto, M.;
Asano, K.; Kawamoto, I.; Tomita, F.; Nakano, H. J . Antibiot. 1988,
41, 1915. Yasuzawa, T.; Iida, T.; Muroi, K.; Ichimura, M.; Takahashi,
K.; Sano, H. Chem. Pharm. Bull. 1988, 36, 3728. Ichimura, M.; Muroi,
K.; Asano, K.; Kawamoto, I.; Tomita, F.; Morimoto, M.; Nakano, H. J .
Antibiot. 1988, 41, 1285. Ishii, S.; Nagasawa, M.; Kariya, Y.; Yama-
moto, H.; Inouye, S.; Kondo, S. J . Antibiot. 1989, 42, 1713. Hanka, L.
J .; Dietz, A.; Gerpheide, S. A.; Kuentzel, S. L.; Martin, D. G. J . Antibiot.
1978, 31, 1211. Chidester, C. G.; Krueger, W. C.; Mizsak, S. A.;
Duchamp, D. J .; Martin, D. G. J . Am. Chem. Soc. 1981, 103, 7629.
(2) Boger, D. L.; Machiya, K.; Hertzog, D. L.; Kitos, P. A.; Holmes,
D. J . Am. Chem. Soc. 1993, 115, 9025.
(3) For mechanistic aspects see: Boger, D. L.; J ohnson, D. S. Angew.
Chem., Int. Ed. Engl. 1996, 35, 1438.
(4) For synthetic aspects see: Boger, D. L.; Boyce, C. W.; Garbaccio,
R. M.; Goldberg, J . A. Chem. Rev. 1997, 97, 787.
(5) Warpehoski, M. A.; Hurley, L. H. Chem. Res. Toxicol. 1988, 1,
315.
(6) Boger, D. L.; Garbaccio, R. M. Acc. Chem. Res. 1999, 32, 1043.
Boger, D. L.; Garbaccio, R. M. Bioorg. Med. Chem. 1997, 5, 263.
10.1021/jo000297j CCC: $19.00 © 2000 American Chemical Society
Published on Web 05/10/2000

4102 J . Org. Chem., Vol. 65, No. 13, 2000
Boger et al.
Sch em e 1
F igu r e 1.
methyl group from the alkylation subunit of CC-1065 (C7-
MeCPI)17 has not been previously investigated, and its
role has been the subject of much speculation. Because
this substituent protrudes into the floor of the minor
groove, it was not clear whether this removal of the C7
methyl group would be a productive modification leading
to an increase in DNA alkylation due to the removal of a
steric impediment to minor groove alkylation or detri-
mental leading to a decrease in reaction selectivity or
regioselectivity.
F igu r e 2.
sis of reversed analogues of duocarmycin SA,7 and
supporting evidence has been derived through solution
structural studies8,9 of the DNA adducts and the synthe-
sis of carefully designed analogues containing deep-
seated structural modifications.10-16
Herein, we describe the synthesis of 1,2,8,8a-tetrahy-
drocyclopropa[c]pyrrolo[3,2-e]indol-4(5H)-one (CPI),17 the
parent alkylation subunit of both duocarmycin SA and
CC-1065 which lacks the C6 methoxycarbonyl and C7
methyl group of the respective natural products (Figure
2). The removal of the C6 methyl ester from the duocar-
mycin SA alkylation subunit leads to a decrease in the
rigid length of the agent and should decrease the rate of
DNA alkylation derived from catalysis by the binding-
induced conformational change. The removal of the C7
The studies described herein demonstrate that the
duocarmycin SA C6 methoxycarbonyl conveys several
productive features to the molecule. In addition to
enhancing intrinsic chemical stability (6×) which results
in a comparable enhancement in biological potency, it
increases the rate of DNA alkylation approximately 12-
13-fold and the alkylation efficiency 10-fold without
altering the characteristic sequence selectivity. In con-
trast, the C7 methyl group of the CC-1065 alkylation
subunit does not appear to convey productive character-
istics to the natural product or its analogues. The agents
described herein lacking the C7 methyl group possessed
a comparable reactivity and reaction selectivity but
alkylated DNA with a faster rate and improved efficiency.
Thus, the C7 methyl group appears to be counterproduc-
tive, and the CPI analogues lacking this substituent
exhibited a biological potency approximately 2-10 times
greater than the corresponding CC-1065 analogue.
(7) Boger, D. L.; Bollinger, B.; Hertzog, D. L.; J ohnson, D. S.; Cai,
H.; Me´sini, P.; Garbaccio, R. M.; J in, Q.; Kitos, P. A. J . Am. Chem.
Soc. 1997, 119, 4987.
(8) Eis, P. S.; Smith, J . A.; Rydzewski, J . M.; Case, D. A.; Boger, D.
L.; Chazin, W. J . J . Mol. Biol. 1997, 272, 237.
(9) Schnell, J . R.; Ketchem, R. R.; Boger, D. L.; Chazin, W. J . J .
Am. Chem. Soc. 1999, 121, 5645.
(10) Boger, D. L.; Hertzog, D. L.; Bollinger, B.; J ohnson, D. S.; Cai,
H.; Goldberg, J .; Turnbull, P. J . Am. Chem. Soc. 1997, 119, 4977.
(11) Boger, D. L.; Garbaccio, R. M.; J in, Q. J . Am. Chem. Soc 1997,
62, 8875.
Syn th esis of N-BOC-CP I. Regioselective iodination
(NIS, TsOH) of indole 4,18 first prepared in synthetic
efforts on CC-1065 and available in two steps from di-
N-benzoyl-2-(benzyloxy)-p-quinonediimine, was followed
by N1-debenzoylation under basic conditions (2 M
NaOCH3, CH3OH, 10 min, 25 °C, 99%) to provide 6
(Scheme 1). Further removal of the N5-benzoyl group was
(12) Boger, D. L.; Santilla´n, A., J r; Searcey, M.; J in, Q. J . Am. Chem.
Soc. 1998, 120, 11554.
(13) Boger, D. L.; Santilla´n, A., J r.; Searcey, M.; J in, Q. J . Org.
Chem. 1999, 64, 5241.
(14) Boger, D. L.; Turnbull, P. J . Org. Chem. 1997, 62, 5849.
(15) Boger, D. L.; Turnbull, P. J . Org. Chem. 1998, 63, 8004.
(16) Boger, D. L.; Garbaccio, R. M. J . Org. Chem. 1999, 64, 5666.
(17) In preceding studies, CPI has been used as an abbreviation for
the CC-1065 alkylation subunit. Since it more accurately reflects the
abbreviation for the parent unsubstituted alkylation subunit described
herein, we have adopted CPI as its abbreviation and C7-MeCPI for
the CC-1065 alkylation subunit.
(18) Boger, D. L.; Zarrinmayeh, H. J . Org. Chem. 1990, 55, 1379.

Parent Alkylation Subunit of CC-1065
J . Org. Chem., Vol. 65, No. 13, 2000 4103
Ta ble 1. Resolu tion on Ch ir a lcel OD Colu m n
KW-2189 was found to undergo solvolysis and alkylate
DNA albeit less effectively than the free phenol,23 11
failed to provide evidence of spirocyclization and was
recovered unchanged upon treatment with NaH (THF),
DBU (CH3CN), Et3N-CH3CN-H2O (1:1:1), or pH 7
phosphate buffer. While this was only examined in
preliminary experiments which do not preclude such
behavior, it is suggestive that the spirocyclization of 11
and related agents is not facile and unlikely to be
significant.
agent
R1
R2
R3
R
10
11
12
13
BOC
H
H
Bn
Bn
H
BOC
BOC
BOC
-
1.30
1
1.53
1.35
-
-
accomplished by treatment of 6 with BOC2O and DMAP
at room temperature for 30 min (93%), followed by
hydrolysis of the benzoyl group (2 M NaOCH3, CH3OH,
10 min, 25 °C, 98%). Intermediate 8 was N-alkylated to
give the key substrate 9 for free radical cyclization (3
equiv of NaH, 3 equiv of 1,3-dichloropropene, DMF, 1 h,
25 °C, 100%).19 The cyclization occurred smoothly in
toluene (90 °C, 0.1 equiv of AIBN, 1 equiv of Bu3SnH, 1
h, 100%) to give 10, the fully protected precursor to
N-BOC-CPI. This free radical reaction could also be
conducted with catalytic amounts of tin and treatment
of 9 with (Bu3Sn)2O (0.1 equiv) and poly(methylhydrox-
ysiloxane) (PMHS) under conditions described by Fu and
co-workers20 cleanly provided 10 (55%).
Although not integral to the preparation of advanced
CPI analogues where both the N1 and N5 BOC protecting
groups are removed simultaneously, the preparation of
N-BOC-CPI itself required selective removal of the indole
N1-BOC group. This could be achieved under mild acidic
conditions (TFA, CH2Cl2, 1-2 h, 0 to 25 °C, 67%, typically
65-80%) providing 11. Debenzylation using transfer
hydrogenation21 (25% aq HCO2NH4, THF, 10% Pd/C, 2
h, 25 °C, 100%) followed by spirocyclization (DBU, CH3-
CN, 25 °C, 10 min, 100%) gave N-BOC-CPI in excellent
conversions. In fact, spirocyclization was so facile that
purification of 12 by SiO2 chromatography often resulted
in partial closure to 13.22 This same route was also
explored using the corresponding bromide derived from
treatment of 4 with NBS and found to give comparable
conversions, including that of the 5-exo-trig free radical
cyclization.19
Resolu tion . To assess the properties of both enanti-
omers of the CPI-based agents, direct chromatographic
resolution of 10-13 on a Chiralcel OD semipreparative
HPLC column (2 × 25 cm, 10% i-PrOH-hexane eluent, 7
mL/min) was investigated (Table 1). Compound 10, the
product of the free radical cyclization and immediate
precursor to the extended analogues, was effectively
resolved (R ) 1.30). Although compound 11 was incapable
of resolution under similar conditions, the intermediate
12 could also be resolved with excellent separation of the
enantiomers (R ) 1.53), as could N-BOC-CPI (13, R )
1.35). The absolute configuration was assigned on the
basis of the DNA alkylation selectivity of the analogues
prepared from the two enantiomers and was consistent
with expected [R]D rotation values, relative Chiralcel OD
retention times, and cytotoxic potency.
Solvolysis: Rea ctivity. The relative reactivity of the
alkylation subunits as established by their rate of acid-
catalyzed solvolysis has been found to reflect a direct
relationship between stability and in vitro biological
potency.3 The solvolysis of N-BOC-CPI at pH 3.0 (50%
CH3OH-buffer, buffer ) 4:1:20 v/v/v 0.1 M citric acid,
0.2 M Na2HPO4, H2O) was followed spectrophotometri-
cally by UV with the disappearance of the long wave-
length absorption at 343 nm of the CPI chromophore and
with the appearance of a short wavelength absorption
at 238 nm attributable to the solvolysis product. The
reactivity of N-BOC-CPI (k ) 7.13 × 10-6 s-1, t1/2 ) 27
h) was similar to the BOC derivative of the CC-1065
alkylation subunit (t1/2 ) 37 h) but substantially greater
(6-7×) than the duocarmycin SA alkylation subunit
N-BOC-DSA (t1/2 ) 177 h), Figure 3. This indicates that
the C7 methyl group of the CC-1065 alkylation subunit
has little effect on the acid-catalyzed solvolysis reactivity,
whereas the electron-withdrawing nature of the duocar-
mycin SA C6 methyl ester has a stabilizing effect.
Solvolysis: Regioselectivity. The acid-catalyzed nu-
cleophilic addition of CH3OH to N-BOC-CPI was con-
ducted on a preparative scale to establish the regiose-
lectivity of addition. Treatment of 13 with catalytic
CF3SO3H (0.12 equiv, CH3OH, 0-25 °C, 3 h, 90%) led to
two products in a ratio of 4:1 (Scheme 2). The major
product 21 results from attack at the least-substituted
cyclopropane carbon, and the addition occurs with a
regioselectivity identical to that of the natural product
alkylation subunits.24,25 The minor ring expansion prod-
In conjunction with efforts to establish the properties
of the O-benzyl-protected CPI derivatives, we examined
the possibility that 11 and related agents may spirocy-
clize via participation of the indole N5-H (eq 1). Analogous
observations with KW-2189 have been made although in
this case the indole NH is much more acidic and the
leaving group (Br vs Cl) more reactive.23 Thus, while
(19) Patel, V. F.; Andis, S. L.; Enkema, J . K.; J ohnson, D. A.;
Kennedy, J . H.; Mohamadi, F.; Schultz, R. M.; Soose, D. J .; Spees, M.
M. J . Org. Chem. 1997, 62, 8868. Boger, D. L.; Boyce, C. W.; Garbaccio,
R. M.; Searcey, M. Tetrahedron Lett. 1998, 39, 2227.
(20) Lopez, R. M.; Hays, D. S.; Fu, G. C. J . Am. Chem. Soc. 1997,
119, 6949.
(21) Ram. S.; Ehrenkaufer, R. E. Synthesis 1988, 91.
(22) Efforts to further shorten this route by reacting 5 with BOC2O
(25 °C, 91%) followed by treatment of the resulting imide with 2 M
NaOCH3-CH3OH (25 °C, 4 h, 91%) provided 7-(benzyloxy)-5-[(tert-
butyloxycarbonyl)amino]-4-iodoindole cleanly. However, attempts to
alkylate N5 selectively with 1,3-dichloropropene provided mixtures of
N1 and N5 alkylation. Similarly, although this was not investigated in
detail, attempted alkylation of 5 with 1,3-dichloropropene failed to
provide the N-alkylated material cleanly.
(24) Warpehoski, M. A.; Harper, D. E. J . Am. Chem. Soc. 1994, 116,
7573. Warpehoski, M. A.; Harper, D. E. J . Am. Chem. Soc. 1995, 117,
2951.
(23) Asai, A.; Nagamura, S.; Saito, H. J . Am. Chem. Soc. 1994, 116,
4171.

4104 J . Org. Chem., Vol. 65, No. 13, 2000
Boger et al.
F igu r e 4.
X-r a y Cr ysta l Str u ctu r e of N-BOC-CP I a n d Rela -
tion sh ip s to Rea ctivity a n d Rea ction Regioselec-
tivity.26 The single-crystal X-ray structures of the alky-
lation subunits of CC-1065, the duocarmycins, and
synthetic analogues have provided insights into the
origins of their reaction regioselectivity and differences
in stability.3,6 The stability of the alkylation subunits is
a result of three structural features: the conjugative
stability provided by the fused aromatic ring, the non-
ideal alignment of the cyclopropane, and the cross-
conjugative stability provided by the vinylogous amide
with the latter being especially important. The difference
in reactivity of ring-expanded analogues of CBI (i.e., CBQ
and CNA)14,16 has been accurately correlated with an
increase in the bond length a and in the torsion angle ø1
diagnostic of the extent of vinylogous amide cross con-
jugation. This bond length and ø1 torsion angle for
N-BOC-CPI were 1.396 Å and 12.6°. These values are
close to those of N-BOC-DSA (1.393 Å, 6.9°) but diag-
nostic of less vinylogous amide conjugation, indicating
that the increased reactivity of CPI alkylation subunit
may stem in part from the decrease in cross conjugation
of the vinylogous amide (Figure 4). Further contributing
to the differences in reactivity is the electronic nature of
the pyrrole ring substitution arising from removal of the
electron-withdrawing methoxycarbonyl group of N-BOC-
DSA. The comparable stereoelectronic alignment of the
CPI and DSA cyclopropanes with the cyclohexadienone
π-system accounts for their near identical reaction regi-
oselectivity.
F igu r e 3.
Sch em e 2
uct 22 was found to be generated exclusively with
inversion of the reacting center stereochemistry, indicat-
ing that the attack of methanol occurs by clean SN2
addition (see Experimental Section). This is analogous
to our observations with N-BOC-DSA,10,25 N-BOC-DA,25
and related agents,14,16 but contrasts the purported
behavior of C7-MeCPI reported by Warpehoski24 and
disclosed without such conclusive evidence of the extent
of maintenance or loss of stereochemistry. Treatment of
N-BOC-CPI with HCl in THF (1.8 equiv, -78 °C) cleanly
gave only the product of addition to the less-substituted
cyclopropane, with no trace of formation of the ring-
expanded product. Identical observations have been
described for the natural alkylation subunits of CC-1065
and the duocarmycins,10,23-25 and results from the in-
creased preference for SN2 addition to the least-substi-
tuted carbon by large nucleophiles. Thus, the removal of
the CC-1065 C7 methyl group or the duocarmycin SA C6
methoxycarbonyl group has no impact on the intrinsic
reaction regioselectivity.
Syn th esis of Key Du oca r m ycin a n d CC-1065 An a -
logu es. Coupling of the CPI alkylation subunit to the
DNA binding subunits of duocarmycin, CC-1065, and
related analogues followed a general procedure previously
reported19 (Scheme 3). This approach was especially
attractive because of the successful resolution of the early
synthetic intermediate 10. Exhaustive BOC deprotection
(26) The atomic coordinates for this structure have been deposited
with the Cambridge Crystallographic Data Centre and may be obtained
upon request from the Director, Cambridge Crystallographic Data
Centre, 12 Union Road, Cambridge, CB2 1EZ, UK.
(25) Boger, D. L.; Goldberg, J .; McKie, J . A. Bioorg. Med. Chem. Lett.
1996, 6, 1955. Boger, D. L.; McKie, J . A.; Nishi, T.; Ogiku, T. J . Am.
Chem. Soc. 1997, 119, 311.

Parent Alkylation Subunit of CC-1065
J . Org. Chem., Vol. 65, No. 13, 2000 4105
Ta ble 2. Rela tive Alk yla tion Efficien cya
Sch em e 3
agent
R1
R2
relative alkylation efficiency
2
CO2CH3
CO2CH3
CO2CH3
H
H
H
5,6,7-OCH3
5-OCH3
H
5,6,7-OCH3
5-OCH3
H
1
1
0.05
0.1
0.1-0.05
0.01
38
39
40
a
At the w794 high affinity site 5′-AATTA.
DNA in aqueous buffer, thermolysis (100 °C, 30 min) to
induce strand cleavage at the sites of alkylation, dena-
turing polyacrylamide gel electrophoresis (PAGE) adja-
cent to Sanger sequencing lanes, and autoradiography
led to identification of the DNA cleavage and alkylation
sites. The full details of this procedure have been
disclosed elsewhere.27 A representative comparison of
DNA alkylation by (+)-CPI-TMI (38), (+)-39, and (+)-
CPI-indole (40) alongside (+)-duocarmycin SA within
w794 is illustrated in a Supporting Information figure
(Figure S1). All three of the (+)-CPI-derivatives alkylate
DNA with selectivities identical to (+)-duocarmycin
SA,28,29 but with increasingly decreased efficiency (Table
2). Both (+)-38 and (+)-39 alkylate with an efficiency that
is 10-50× less than the natural product. This behavior
is additionally magnified with the further incremental
loss of DNA alkylation efficiency seen with (+)-CPI-indole
(ca. 100×), in which both the C6 methyl ester and the
C5′ methoxy group of (+)-duocarmycin SA have been
removed. This decrease in alkylation rate and efficiency
is analogous to that observed while examining the role
of the duocarmycin SA C5′ methoxy group (Table 2).7,30
The behavior of CPI-indole (40) illustrates that both the
C6 methoxycarbonyl group and the C5′ methoxy group
independently contribute significantly to the DNA alky-
lation efficiency consistent with a potential role in
activation derived from a DNA binding-induced confor-
mational change. In this regard, it is notable that a
similar observation within a series of CBI-based ana-
logues (i.e., 17-19) revealed that the increase in DNA
alkylation efficiency and rate was not dependent upon
the electronic character of the analogous C7 substituent,
but rather related simply to its presence (19 ) 18 > 17,
CN ) OCH3 > H).31
(3.6 M HCl/EtOAc, 25 °C, 30 min) of 10 followed by
immediate coupling of the amine hydrochloride (3 equiv
of EDCI, 5 equiv of NaHCO3, DMF, 53-97%) with the
appropriate carboxylic acid (1.2 equiv of 23-27) provided
the benzyl protected seco-CPI derivatives 28-32 in good
yields. Debenzylation (25% aq HCO2NH4, 10% Pd/C,
THF, 94-100%) followed by base-catalyzed ring closure
(3 equiv of NaH, DMF, 43-100%) provided the desired
products 38-42. The advantages of proceeding through
the benzyl-protected seco-CPI agents are severalfold.
Purification of the coupled products was facilitated by
the relative nonpolar properties and favorable solubility
characteristics of the intermediate. Typically purification
of the coupling products derived from 12 were contami-
nated with the corresponding cyclopropane agents which
formed upon SiO2 purification. Thus, purifications of 28-
32 followed by their deprotection provided 33-37 suf-
ficiently clean as to not require further purification. In
addition, comparable attempts to couple the phenol
derived from deprotection of 12 proved less successful and
sometimes problematic. Finally and more importantly,
the benzyl-protected precursors 28-32 are relatively
nontoxic compared to 33-37, constituting easily purified
and readily stored intermediates.
The CPI unnatural enantiomer DNA alkylation is
considerably slower and was observed only after a more
prolonged incubation (72 versus 24 h, 25 °C, Supporting
DNA Alk yla tion Selectivity a n d Efficien cy. The
DNA alkylation properties of the agents were examined
within w794 duplex DNA for which comparison results
are available for related agents. The alkylation site
identification and the assessment of the relative selectiv-
ity among the available sites were obtained by thermally
induced strand cleavage of the singly 5′-end-labeled
duplex DNA after exposure to the agents. Following
treatment of the end-labeled duplex DNA with a range
of agent concentrations, the unbound agent was removed
by ethanol precipitation of the DNA. Redissolution of the
(27) Boger, D. L.; Munk, S. A.; Zarrinmayeh, H.; Ishizaki, T.;
Haught, J .; Bina, M. Tetrahedron 1991, 47, 2661.
(28) Boger, D. L.; J ohnson, D. S.; Yun, W. J . Am. Chem. Soc. 1994,
116, 1635. Boger, D. L.; J ohnson, D. S.; Yun, W.; Tarby, C. M. Bioorg.
Med. Chem. 1994, 2, 115.
(29) Boger, D. L.; Ishizaki, T.; Zarrinmayeh, H.; Munk, S. A.; Kitos,
P. A.; Suntornwat, O. J . Am. Chem. Soc. 1990, 112, 8961. Boger, D.
L.; Coleman, R. S.; Invergo, B. J .; Sakya, S. M.; Ishizaki, T.; Munk, S.
A.; Zarrinmayeh, H.; Kitos, P. A.; Thompson, S. C. J . Am. Chem. Soc.
1990, 112, 4263.
(30) Boger, D. L.; Bollinger, B.; J ohnson, D. S. Bioorg. Med. Chem.
Lett. 1996, 6, 2207.
(31) Boger, D. L.; Han, N.; Tarby, C.; Boyce, C. W.; Cai, H.; J in, Q.;
Kitos, P. A. J . Org. Chem. 1996, 61, 4894.

4106 J . Org. Chem., Vol. 65, No. 13, 2000
Boger et al.
Information Figure S1). Even with the longer reaction
times, the alkylation by the unnatural enantiomers is less
efficient, requiring higher agent concentrations to detect.
The DNA alkylation selectivity observed with ent-(-)-
CPI-TMI and ent-(-)-duocarmycin SA was indistinguish-
able but, consistent with the observations made with the
natural enantiomers, the efficiency of alkylation by (-)-
CPI-TMI was less (50-100×) than that of (-)-duocar-
mycin SA which possesses the more extended rigid
structure. Thus, the incorporation of the C6 methoxy-
carbonyl group increased the unnatural enantiomer DNA
alkylation even more significantly than the natural
enantiomer. The alkyation sites for the unnatural enan-
tiomers proved consistent with adenine N3 alkylation in
the minor groove with the agent binding in the 5′ to 3′
direction across a 3.5 base pair AT-rich sequence sur-
rounding the alkylation site.32 This is analogous to the
natural enantiomer alkylation selectivity which extends
in the reverse 3′ to 5′ direction in the minor groove and,
because of the diastereomeric nature of the adducts, is
offset by one base pair relative to the natural enanti-
omers.28
Rela tive Ra tes of Alk yla tion for (+)-CP I-TMI (38),
(+)-CP I-in d ole (40), a n d (+)-Du oca r m ycin SA (2).
The relative rates of DNA alkylation for (+)-CPI-TMI
(38), (+)-CPI-indole (40), and (+)-duocarmycin SA (2) at
the w794 high affinity site were measured (25 °C, 24 h,
10-5 and 10-4 M), and the results are shown in Figure 5
and summarized in Table 3. In previous studies, the
comparison of (+)-duocarmycin SA (2), (+)-DSA-5-meth-
oxyindole, and (+)-DSA-indole established that the re-
moval of the C5′ methoxy substituent leads to a substan-
tial decrease in the relative alkylation rate.7,30 The loss
of this group, which extends the rigid length of the
binding subunit, leads to a decrease in the inherent twist
in the helical conformation of the DNA bound agent. The
degree of twist in the N2 amide induced upon binding
contributes to the disruption of the vinylogous amide
conjugation and the increase in the inherent reactivity
of the agent, thus contributing to the catalysis of the DNA
alkylation reaction. This effect on the rate of DNA
alkylation is clearly demonstrated with the CPI-based
agents. A decrease in the extended rigid length of
duocarmycin SA through removal of the left-hand subunit
C6 methyl ester leads to a 12-13× decrease in the
relative rate of DNA alkylation. Thus, the C6 methoxy-
carbonyl group of duocarmycin SA contributes signifi-
cantly to the rate of DNA alkylation. This is especially
significant since 38 is intrinsically 6× more reactive than
2 yet alkylates DNA 12-13× slower. As reported previ-
ously, removal of the right-hand side C5′ methoxy group
led to a 20× reduction in the rate of DNA alkylation.7,30
Removing the groups on both the left and right-hand
subunits leads to a 250× decrease in the relative alkyl-
ation rate, demonstrating the predicted cumulative effect
(predict 240-260×) and highlighting the magnitude of
these structural variations on the activation of the DNA
bound agent.
F igu r e 5. Plots of percent integrated optical density (% IOD)
versus time established through autoradiography of 5′-32P end-
labeled DNA and used to monitor the relative rate of w794
DNA alkylation of the 5′-AATTA high affinity site for (+)-
duocarmycin SA (2), (+)-38 and (+)-40. Alkylation by (+)-2
and (+)-38 were compared at a concentration of 10-5 M and
(+)-38 and (+)-40 were compared at 10-4 M, 25 °C.
Ta ble 3. Rela tive Ra tes of DNA Alk yla tion a
agent
krel
Shortened Agents
(+)-2, duocarmycin SA
1
(+)-38, CPI-TMI
0.08
0.05
0.004
(+)-DSA-indole
(+)-40, CPI-indole
Extended Agents
(+)-41, CPI-indole2
(+)-43, DSA-indole2
(+)-44, CBI-indole2
(+)-45, C7-MeCPI-indole2
0.18
1
0.65
0.05
a
At the w794 high affinity site 5′-AATTA.
the 10-fold decrease in alkylation efficiency of (+)-CPI-
indole2 (10-5 M) analogous to (+)-C7-MeCPI-indole2
33
compared with (+)-DSA-indole2 which alkylates the same
single high affinity site in w794 DNA at 10-6 M concen-
tration. The relative rates of alkylation for (+)-CPI-
indole2 (41), (+)-CBI-indole2 (44), and (+)-DSA-indole2
(43) at the w794 high affinity site were measured (25 °C,
24 h, 10-5 M agent concentration), and the results are
shown in Figure 7 and summarized in Table 3. In
previous studies, the relative rates of alkylation for (+)-
43-45 were measured, and while the DSA and CBI
Rela tive Ra tes a n d Efficien cies of Alk yla tion for
(+)-CP I-in d ole2 (41), 45, a n d (+)-DSA-in d ole2 (43).
A representative example of DNA alkylation by (+)-41
within w794 is illustrated in Figure 6 and demonstrates
(32) The exceptions to this constitute alkylation sites observed for
the contaminate natural enantiomer (<1%) which alkylates DNA more
rapidly and efficiently.
(33) Boger, D. L.; Ishizaki, T.; Sakya, S. M.; Munk, S. A.; Kitos, P.
A.; J in, Q.; Besterman, J . M. Bioorg. Med. Chem. Lett. 1991, 1, 115.

Parent Alkylation Subunit of CC-1065
J . Org. Chem., Vol. 65, No. 13, 2000 4107
F igu r e 6. Thermally induced strand cleavage of w794 DNA
(SV40 DNA segment, 144 bp, nucleotide no. 138-5238); DNA-
agent incubation for 24 h at 25 °C, removal of unbound agent
and 30 min of thermolysis (100 °C), followed by denaturing
8% PAGE and autoradiography; lanes 1 and 2, (+)-CC-1065
(1, 1 × 10-5 and 1 × 10-6 M); lanes 3 and 4, (+)-duocarmycin
SA (2, 1 × 10-5 and 1 × 10-6 M); lanes 5-8, Sanger G, C, A
and T sequencing reactions; lane 9, control DNA; lanes 10-
12, (+)-CPI-indole2 (41, 1 × 10-4 to 1 × 10-6 M); lanes 13-15,
(+)-CBI-indole2 (44, 1 × 10-4 to 1 × 10-6 M); lanes 16-18,
(+)-DSA-indole2 (43, 1 × 10-4 to 1 × 10-6 M).
F igu r e 7. Plots of percent integrated optical density (% IOD)
versus time established through autoradiography of 5′-32P end-
labeled DNA and used to monitor the relative rate of w794
alkylation of the 5′-AATTA high affinity site for (+)-DSA-
indole2 (43), (+)-CBI-indole2 (44), (+)-CPI-indole2 (41) and (+)-
C7-MeCPI-indole2 (45) at a concentration of 10-5 M compound,
25 °C. The relative rates of alkylation are as follows: (+)-DSA-
indole2 (21.5×) > (+)-CBI-indole2 (14.0×) > (+)-CPI-indole2
(4.0×) > (+)-C7-MeCPI-indole2 (1.0×).
agents were similar in reactivity with DSA-indole2 having
a rate 1.5-2.0× faster than that of CBI-indole2, the CC-
1065 alkylation subunit analogue reacted much more
slowly with a rate 14-15× less than that of CBI-indole2
despite its greater intrinsic reactivity.31 To define the
effect of the CC-1065 C7 methyl substituent on the DNA
alkylation rate, CPI-indole2 (41) was compared with the
preceding agents. Its relative rate of DNA alkylation was
established to be faster than that of C7-MeCPI-indole2
(45). The absence of the C7 methyl substituent, which
has been suggested to contribute either a steric impedi-
ment to the binding and positioning of the alkylating
agent in the minor groove or to contribute to the selectiv-
ity of alkylation by precluding unproductive alkylations,
was found to slow the relative alkylation rate, with (+)-
45 having a rate 4× slower than 41 and >20× slower
than DSA-indole2. This indicates that the duocarmycin
SA C6 methoxycarbonyl group enhances the rate of DNA
alkylation 6× despite the fact that the alkylation subunit
is intrinsically 6× less reactive and that the C7 methyl
group of CC-1065 slows the relative DNA alkylation rate
(4×) and lowers the efficiency of DNA alkylation.
based analogues were more potent than the correspond-
ing analogue containing the CC-1065 alkylation subunit
and were comparable or slightly less potent than the
analogues containing the duocarmycin SA alkylation
subunit. Instructive in this regard is the L1210 activity
(IC50) of the N-BOC derivatives N-BOC-CPI (13, 160 nM),
N-BOC-C7-MeCPI (14, 330 nM), and N-BOC-DSA (16, 6
nM) and that of (+)-C7-MeCPI-indole2 (45), (+)-CPI-
indole2 (41), and DSA-indole2 (43): IC50 ) 40, 3, and 3
pM, respectively. Like prior observations, the natural
enantiomers were significantly more potent than the
corresponding unnatural enantiomers (10-200×). How-
ever, the relative distinctions were typically larger than
those observed with the DSA-based analogues (5-50×)
and more similar to the differences observed with the CC-
1065 analogues (1-1000×). Analogous to prior observa-
tions, the phenol precursors 33-37 displayed cytotoxic
activity that was not distinguishable from the final
agents containing the cyclopropane, indicating spirocy-
clization readily occurs under the conditions of the assay.
The corresponding precursors in which the phenol is
protected as the benzyl ether were substantially less
potent (ca. 10-200×), but not inactive. Even these deriv-
Cytotoxic Activity. The compounds displayed cyto-
toxic activity consistent with their DNA alkylation
properties and relative stabilities (Table 4). The CPI-

4108 J . Org. Chem., Vol. 65, No. 13, 2000
Ta ble 4. In Vitr o Cytotoxic Activity
Boger et al.
compd
IC50, nM (L1210)a
compd
IC50, nM (L1210)a
natural enantiomers
13, (+)-N-BOC-CPI
38, (+)-CPI-TMI
39, (+)-CPI-(5-MeO)indole
40, (+)-CPI-indole
41, (+)-CPI-indole2
42, (+)-CPI-CDPI1
unnatural enantiomers
(-)-N-BOC-CPI
(-)-CPI-TMI
160
(210)
1000
1
1
10
1
(1000)
(1)
(10)
(70)
(1)
0.007
0.009
0.050
0.003
0.009
(0.008)
(0.008)
(0.070)
(0.002)
(0.009)
(-)-CPI-(5-MeO)indole
(-)-CPI-indole
(-)-CPI-indole2
(-)-CPI-CDPI1
2
(4)
1, (+)-CC-1065
0.020
0.006
330
6
(-)-CC-1065
0.020
0.100
2, (+)-duocarmycin SA
14, (+)-N-BOC-C7-MeCPI
16, (+)-N-BOC-DSA
(-)-duocarmycin SA
(-)-N-BOC-DSA
60
O-benzyl derivatives
28
29
30
31
32
0.1
0.1
nd
0.5
1.3
50
50
100
100
20
a
The value in parentheses is the IC50 for the corresponding seco derivatives 12, 33-37. The assays were conducted at least four times,
and the average IC50 is reported; see ref 2.
atives retain significant activity, with the natural enan-
tiomers exhibiting IC50’s of ca. 0.1-1 nM, and the un-
natural enantiomers were again less potent (20-100 nM).
these observations, the CC-1065 C7 methyl group dimin-
ishes the biological potency relative to the CPI-based
agents.
Con clu sion s. A concise synthesis of the parent CPI
alkylation subunit is described which lacks the C6
methoxycarbonyl group of duocarmycin SA and the C7
methyl group of CC-1065. The evaluation of its properties
revealed that the duocarmycin SA C6 methyl ester
substantially increases stability (6×) while the CC-1065
C7 methyl group has no or only a relatively small effect
on the alkylation subunit reactivity. Neither substituent
has any impact on the intrinsic reaction regioselectivity
of the activated cyclopropane. Most remarkable is the
effect of the substitution pattern on the reactivity of the
agents toward DNA alkylation. The duocarmycin SA C6
methoxycarbonyl group substantially increases both the
efficiency and rate of DNA alkylation despite its intrinsic
greater stability, without altering the inherent sequence
selectivity, while the CC-1065 C7 methyl group slows the
rate and lowers the efficiency of DNA alkylation. The
removal of the methoxycarbonyl group and the resulting
decrease in rigid length in going from duocarmycin SA
to CPI-TMI (38) leads to a 12-13-fold decrease in the
relative rate of DNA alkylation. A similar effect has been
previously described for DSA-indole, in which removal
of the C5′ methoxy substituent on the right-hand subunit
led to a 20-fold decrease in the rate of DNA alkylation.7,21
The even larger 250-fold decrease in alkylation rate seen
with CPI-indole (40) in which both substituents are
removed is the predicted product of the incremental
decreases resulting from removal of each substituent. We
have interpreted this behavior to lend further support
for the model of agent activation derived from a DNA
binding-induced twist in the linking amide disrupting the
stabilizing vinylogous amide cross-conjugation. The duo-
carmycin SA C6 methyl ester and C5′ methoxy group lie
deep in the minor groove, increasing the rigid length of
the agent and thus the degree of the binding-induced
twist. Their removal decreases the degree of activation
leading to the slower and less-efficient DNA alkylation.
In contrast, the CC-1065 C7 methyl substituent was
found to slow the rate of DNA alkylation 4× and lower
the efficiency of DNA alkylation without affecting the
reaction selectivity or regioselectivity. Presumably this
arises from steric interactions with the minor groove floor
which hinder binding and alkylation. Consistent with
Exp er im en ta l Section
1-Ben zoyl-5-(b en zoyla m in o)-7-(b en zyloxy)-4-iod oin -
d ole (5). A solution of indole 417 (118 mg, 0.26 mmol) and
TsOH (26 mg, 0.13 mmol) in THF (1 mL) was cooled to 0 °C
and treated with NIS (71 mg, 0.312 mmol) in THF (1 mL).
The reaction was allowed to warm to 25 °C. After 16 h,
additional NIS (15 mg, 0.065 mmol) was added and the
reaction was stirred for 24 h. The reaction mixture was diluted
with 10% aqueous NaHCO3 and extracted with CHCl3. The
organic layers were combined, dried (MgSO4), and concen-
trated. Flash chromatography (SiO2, 2.5 × 15 cm, 30% EtOAc/
hexane) afforded 5 (67 mg, 45%) as a yellow solid: mp 148 °C;
1H NMR (CDCl3, 400 MHz) δ 8.15 (s, 1H, NH), 7.99 (d, J )
6.0 Hz, 2H), 7.64 (d, J ) 8.0 Hz, 2H), 7.58-7.47 (m, 4H), 7.30-
7.17 (m, 5H), 6.58 (d, J ) 2.8 Hz, 1H), 4.92 (s, 2H); 13C NMR
(CDCl3, 62.5 MHz) δ 167.9, 165.4, 147.2, 136.2, 135.8, 134.8,
134.7, 134.1, 132.9, 132.1, 129.8, 129.5, 128.9, 128.4, 128.2,
127.8, 127.7, 127.1, 122.4, 110.6, 102.1, 71.8, 70.5; FABHRMS
(NBA/CsI) 704.9677 (M+ + Cs, C29H21IN2O3 requires 704.9651).
Anal. Calcd for C29H21IN2O3: C, 60.85; H, 3.70; N, 4.89.
Found: C, 60.81; H, 3.73; N, 4.80.
5-(Ben zoyla m in o)-7-(ben zyloxy)-4-iod oin d ole (6). Com-
pound 5 (1.60 g, 2.80 mmol) in CH2Cl2 (28 mL) was treated
with NaOCH3 in CH3OH (2 M, 4.2 mL, 8.4 mmol), and the
solution was stirred at 25 °C for 10 min. H2O and EtOAc were
added, and the organic layer was separated, dried (MgSO4),
and concentrated. Flash chromatography (SiO2, 3 × 30 cm,
25% EtOAc/hexanes) afforded 6 (1.3 g, 99%) as a colorless
1
solid: mp 165-166 °C; H NMR (CDCl3, 400 MHz) δ 8.61 (s,
1H), 8.31 (s, 1H), 8.05 (s, 1H), 8.00 (d, J ) 6.8 Hz, 1H), 7.53
(m, 5H), 7.37 (m, 3H), 7.21 (dd, J ) 2.8, 1.4 Hz, 1H), 6.43 (dd,
J ) 2.7, 1.2 Hz, 1H), 5.28 (s, 2H); 13C NMR (CDCl3, 62.5 MHz)
δ 165.7, 145.8, 136.5, 135.1, 132.7, 132.1, 132.0, 129.1, 128.8,
128.5, 128.3, 127.3, 124.6, 123.4, 106.9, 99.7, 72.7, 70.7;
FABHRMS (NBA/CsI) m/z 600.9408 (M+ + Cs, C22H17IN2O2
requires 600.9389). Anal. Calcd for C22H17IN2O2: C, 56.43; H,
3.66; N, 5.98. Found: C, 56.61; H, 3.63; N, 5.90.
N5,1-Bis(ter t-b u t yloxyca r b on yl)-5-(b en zoyla m in o)-7-
(b en zyloxy)-4-iod oin d ole (7). Compound 6 (255 mg, 0.54
mmol) in CH2Cl2 (6 mL) was treated with di-tert-butyl dicar-
bonate (687 mg, 3.16 mmol) and DMAP (128 mg, 1.04 mmol).
After 30 min at 25 °C the solution was subjected directly to
flash chromatography (SiO2, 2 × 15 cm, 20% EtOAc/hexane)
to give 7 (390 mg, 93%) as an oil: (Rf 0.80, SiO2, 25% EtOAc/
hexane); 1H NMR (CDCl3, 400 MHz) δ 7.81 (d, J ) 6.9 Hz,
2H), 7.55 (d, J ) 3.6 Hz, 1H), 7.45 (m, 5H), 7.31 (m, 3H), 6.82
(s, 1H), 6.57 (d, J ) 3.6 Hz, 1H), 5.15 (s, 2H), 1.47 (s, 9H),
1.22 (s, 9H); 13C NMR (CDCl3, 62.5 MHz) δ 171.8, 171.4, 152.5,

Parent Alkylation Subunit of CC-1065
J . Org. Chem., Vol. 65, No. 13, 2000 4109
148.9, 148.2, 137.8, 137.1, 136.9, 136.5, 131.6, 129.5, 128.6,
128.5, 128.3, 128.2, 127.7, 111.2, 109.6, 84.2, 83.7, 71.5, 60.5,
28.0, 27.8; FABHRMS (NBA/CsI) m/z 801.0402 (M+ + Cs,
mL, 2.18 mmol) was added to a stirred solution of 12 (132 mg,
0.41 mmol) in anhydrous acetonitrile (13 mL). After 10 min,
the solvent was removed and the residue was subjected directly
to chromatography (SiO2, 2 × 15 cm, 0-5% CH3OH-CHCl3
gradient) to afford 13 (122 mg, 100%) as a tan solid: mp 198
°C (decomp); 1H NMR (CDCl3, 400 MHz) δ 10.80 (br s, 1H,
NH), 7.07 (dd, J ) 2.8, 2.7 Hz, 1H), 6.45 (br s, 1H), 6.00 (dd,
J ) 2.5, 2.4 Hz, 1H), 3.96 (m, 2H), 2.75 (m, 1H), 1.57 (dd, J )
7.7, 3.8 Hz, 1H), 1.51 (s, 9H), 1.33 (t, J ) 4.6 Hz, 1H); 13C
NMR (CDCl3, 62.5 MHz) δ 178.5, 160.0, 152.1, 131.1, 129.2,
125.0, 109.0, 102.1, 83.0, 53.7, 32.7, 28.5, 27.1, 23.5; UV (THF)
C
32H33IN2O6 requires 801.0438).
1-(ter t-Bu tyloxyca r bon yl)-5-[(ter t-bu tyloxyca r bon yl)-
a m in o]-7-(ben zyloxy)-4-iod oin d ole (8). Compound 7 (600
mg, 0.9 mmol) in CH2Cl2 (33 mL) was treated with NaOCH3
in CH3OH (2 M, 0.94 mL, 1.9 mmol). After 10 min at 25 °C,
H2O and EtOAc were added, and the organic layer was
separated. The aqueous layer was washed with EtOAc, and
the combined organic layers were dried (MgSO4) and concen-
trated. Chromatography (SiO2, 2 × 15 cm, 10-25% EtOAc/
hexanes gradient) afforded 8 (496 mg, 98%) as a colorless oil:
1H NMR (CDCl3, 400 MHz) δ 7.83 (br s, 1H), 7.53 (m, 3H),
7.37 (m, 3H), 6.84 (br s, 1H), 6.45 (d, J ) 3.4 Hz, 1H), 5.21 (s,
2H), 1.54 (s, 9H), 1.45 (s, 9H); 13C NMR (CDCl3, 62.5 MHz) δ
161.0, 152.9, 148.0, 136.7, 135.3, 129.0, 128.3, 128.2, 127.9,
127.8, 120.8, 110.4, 102.0, 98.0, 83.7, 71.0, 66.7, 28.3, 27.6;
FABHRMS (NBA/CsI) m/z 697.0176 (M+ + Cs, C25H29IN2O5
requires 697.0176). Anal. Calcd for C25H29IN2O5: C, 53.20; H,
5.18; N, 4.96. Found: C, 53.27; H, 5.26; N, 4.75.
λ
max (∈) 319 (15500), 278 (23600), 209 (15600) nm; FABHRMS
(NBA/NaI) m/z 287.1409 (M+ + H, C16H18N2O3 requires
287.1400).
(+)-(8a R,7bS)-13: [R]22 +161 (c 0.19, acetone).
D
(-)-(8a S, 7bR)-13: [R]22 -160 (c 0.20, acetone).
D
Resolu tion . A solution of 12 (27 mg, 2.18 mmol) in 10%
2-propanol-hexane was resolved on a semipreparative Diacel
Chiralcel OD column (10 µm, 2 × 25 cm) using 10% 2-pro-
panol-hexane as eluent (7 mL/min). The effluent was moni-
tored at 254 nm, and the enantiomers eluted with retention
times of 27.0 and 41.7 min (R ) 1.53). The fractions were
collected and concentrated to afford (-)-(1S)-12 (12 mg, 89%,
>99% ee) and ent-(+)-(1R)-12 (12 mg, 89%, >99% ee). Com-
pounds 10 and 13 were resolved using similar conditions and
with excellent separations (R ) 1.30 and 1.35, respectively).
Aqu eou s Solvolysis of N-BOC-CP I (13). A sample of 13
(150 µg) was dissolved in CH3OH (1.5 mL), and the resulting
solution was mixed with aqueous buffer (pH 3.0, 1.5 mL, 4:1:
20 (v:v:v) 0.1 M citric acid, 0.2 M Na2HPO4, and deionized
water, respectively). The UV spectrum of the solution was
measured immediately after mixing and then every 2 h for 24
h and every 8 h for 2 weeks. The total reaction time reflects
that required to observe no further change in the spectrum.
The decrease in absorbance was monitored at 343 nm while
the increase was monitored at 238 nm. The solvolysis rate (r2
) 0.998) was calculated from the plot of ln[(Af - Ai)/(Af - A)]
vs time; k ) 7.13 × 10-6 s-1, t1/2 ) 27 h (Figure 8).
5-(Ben zyloxy)-1-(ch lor om eth yl)-3,6-bis(ter t-bu tyloxy-
ca r bon yl)-1,2-d ih yd r o-3H-p yr r olo[3,2-e]in d ole (10). A so-
lution of 9 (394 mg, 0.61 mmol) and AIBN (39 mg) in toluene
(40 mL) was degassed with N2 for 15 min. The solution was
warmed to 90 °C, and Bu3SnH (0.17 mL, 0.61 mmol) was added
in 4 portions over 1 h. The solution was cooled and concen-
trated, and the residue was dissolved in EtOAc. Flash chro-
matography (SiO2, 1 × 10 cm, 10% EtOAc/hexanes) afforded
10 (313 mg, 100%) as a colorless film: 1H NMR (CDCl3, 400
MHz) δ 7.74 (br s, 1H), 7.52 (m, 2H), 7.29 (m, 3H), 7.27 (m,
1H), 6.40 (d, J ) 1 Hz, 1H), 5.20 (br s, 2H), 4.12 (m, 2H), 3.86
(m, 2H), 3.48, (t, J ) 10 Hz, 1H), 1.57 (s, 9H), 1.47 (s, 9H); 13
C
NMR (CDCl3, 62.5 MHz) δ 161.0, 148.8, 140.9, 136.9, 130.1,
128.6, 128.2, 127.7, 121.5, 116.0, 103.6, 97.0, 83.3, 71.0, 53.1,
46.8, 45.6, 28.5, 28.3; FABHRMS (NBA/CsI) m/z 512.2094 (M+,
C
28H33ClN2O5 requires 512.2078).
(-)-(1S)-10: [R]22 -11 (c 1.3, EtOAc).
D
(+)-(1R)-10: [R]22 +11 (c 1.4, EtOAc).
Acid -Ca ta lyzed Ad d ition of CH3OH to N-BOC-CP I. A
solution of 13 (5.0 mg, 17 µmol, 1.0 equiv) in CH3OH (1.5 mL)
was treated with CF3SO3H (0.19 mL, 0.011 M, 2.1 µmol, 0.12
equiv) at 0 °C. The reaction was allowed to warm to 25 °C
and after 3 h, was quenched by the addition of NaHCO3 (5
mg), filtered through Celite, and concentrated. PTLC (SiO2,
20 × 20 cm, 25% EtOAc/hexane) afforded 21 (4.1 mg, 76%) as
a clear oil: 1H NMR (CDCl3, 400 MHz) δ 8.37 (d, J ) 1.9 Hz,
1H), 7.49 (s, 1H), 7.12 (dd, J ) 3.0, 2.7 Hz, 1H), 6.39 (d, J )
2.2 Hz, 1H), 4.03 (m, 2H), 3.88 (m, 1H), 3.81 (d, J ) 9.0 Hz,
1H), 3.71 (m, 1H), 3.33 (s, 3H), 1.53 (s, 9H; FABHRMS (NBA/
NaI) m/z 319.1666 (M+ + H, C17H22N2O4 requires 319.1658),
and 22 (1.0 mg, 18%) as a clear oil: 1H NMR (CDCl3, 400 MHz)
δ 8.43 (s, 1H), 7.49 (s, 1H), 7.11 (dd, J ) 3.0, 2.4 Hz, 1H), 6.40
(s, 1H), 3.64 (dd, J ) 12.0, 6.9 Hz, 1H), 3.48 (s, 3H), 3.34 (m,
2H), 3.22 (dd, J ) 6.9, 12.0 Hz, 1H), 2.81 (dd, J ) 6.0, 12.7
Hz, 1H), 1.51 (s, 9H); FABHRMS (NBA/NaI) m/z 319.1672 (M+
+ H, C17H22N2O4 requires 319.1658).
D
5-(Ben zyloxy)-3-(ter t-bu tyloxyca r bon yl)-1-(ch lor om e-
th yl)-1,2-d ih yd r o-3H-p yr r olo[3,2-e]in d ole (11). A solution
of 10 (100 mg, 0.2 mmol) in CH2Cl2 (18 mL) was stirred under
N2 and treated with CF3CO2H (0.115 mL, 1 mmol). After 30
min, a further portion of CF3CO2H (0.115 mL, 1 mmol) was
added. After 1 h, NaHCO3 (400 mg) was added. The solution
was diluted with CH2Cl2, washed with H2O, dried (MgSO4),
and concentrated. Flash chromatography (SiO2, 2 × 15 cm,
10% EtOAc/hexanes) afforded pure 11 (54 mg, 67%) as a
colorless film: 1H NMR (acetone-d6, 400 MHz) δ 8.38 (s, 1H),
8.00 (br s, 1H), 7.45 (m, 5H), 7.16 (dd, J ) 2.7, 2.8 Hz, 1H),
6.84 (br s, 1H), 5.20 (br s, 2H), 4.10 (m, 2H), 3.98 (m, 2H),
3.50 (t, J ) 10.4 Hz, 1H), 1.56 (s, 9H); 13C NMR (CDCl3, 62.5
MHz) δ 153.0, 141.9, 140.9, 130.2, 128.7, 127.4, 127.3, 126.6,
124.9, 124.1, 116.4, 102.1, 96.8, 77.8, 70.9, 49.3, 46.4, 41.6, 28.7;
FABHRMS (NBA/CsI) m/z 412.1529 (M+, C23H25ClN2O3 re-
quires 412.1554).
3-(ter t-Bu tyloxyca r bon yl)-1-(ch lor om eth yl)-5-h ydr oxy-
1,2-d ih yd r o-3H-p yr r olo[3,2-e]in d ole (12). Compound 11
(168 mg, 0.41 mmol), HCO2NH4 (25% aq, 2.2 mL), and 10%
Pd/C (84 mg) were stirred in THF (10 mL), and Ar was bubbled
through the solution for 15 min. The mixture was then
vigorously stirred under Ar for 2 h. The solution was diluted
with EtOAc, dried (MgSO4), filtered, and concentrated. Chro-
matography (SiO2, 2 × 15 cm, 25% EtOAc/hexane) afforded
12 (132 mg, 100%) as a colorless film: 1H NMR (acetone-d6,
400 MHz) δ 9.22 (br s, 1H), 7.35 (br s, 1H), 7.22 (t, J ) 2.8 Hz,
1H), 7.19 (br s, 1H), 6.39 (dd, J ) 2.1, 3.1 Hz, 1H), 4.09 (m,
1H), 3.97 (m, 2H), 3.89 (m, 1H), 3.68 (dd, J ) 8.3, 10.8 Hz,
HPLC resolution of the reactions of (()- and (+)-13 indicat-
ing the reaction regioselectivity (4:1) and the SN2 addition to
provide a single enantiomer of 22 from (+)-13 is shown in
Figure 9.
Ad d ition of HCl to N-BOC-CP I. A solution of 13 (5.0 mg,
17 µmol, 1.0 equiv) in THF (0.40 mL) was cooled to -78 °C
and treated with 3 M HCl-EtOAc (10.0 µL, 30 µmol, 1.8
equiv). The mixture was stirred for 2 min before the solvent
was removed in vacuo. Chromatography (SiO2, 0.7 × 7 cm,
50% EtOAc/hexane) afforded 12 (5.3 mg, 97%).
Gen er a l P r oced u r e for th e Syn th esis of 5-Ben zyloxy
seco-CP I Der iva tives. A sample of 10 (1 equiv) was treated
with 3.6 M HCl/EtOAc (0.5 mL) at 24 °C for 30 min. The
solvent was removed under a stream of N2, and the residue
was dried under vacuum. The salt was dissolved in DMF (0.1
M) and treated sequentially with NaHCO3 (5 equiv), the
carboxylic acid (1.2 equiv), 1-(3-(dimethylamino)propyl)-3-
ethylcarbodiimide hydrochloride (EDCI, 3 equiv), and the
suspension was stirred at 24 °C for 15-18 h under Ar. The
1H), 1.53 (s, 9H); IR (film) νmax 3376, 2977, 1660, 1577 cm-1
;
FABHRMS (NBA/NaI) m/z 323.1155 (M+ + H, C16H19ClN2O3
requires 323.1162).
(-)-(1S)-12: [R]22 -12 (c 0.46, THF).
D
(+)-(1R)-12: [R]22 +12 (c 0.47, THF).
D
2-(ter t-Bu tyloxyca r bon yl)-1,2,8,8a -tetr a h yd r ocyclopr o-
pa[c]pyr r olo[3,2-e]in dol-4-on e (N-BOC-CP I, 13). DBU (0.264

4110 J . Org. Chem., Vol. 65, No. 13, 2000
Boger et al.
F igu r e 9. HPLC analysis (Chiracel OD, 10% i-PrOH in
hexanes) of the acid-catalyzed addition of methanol to N-BOC-
CPI. (a) (+)-N-BOC-CPI. (b) Reaction mixture following acid-
catalyzed methanol addition to (+)-N-BOC-CPI. (c) (()-N-BOC-
CPI. (d) Reaction mixture following acid-catalyzed methanol
addition to (()-N-BOC-CPI.
1-(Ch lor om eth yl)-1,2-d ih yd r o-3-[(5,6,7-tr im eth oxy-1H-
in d ol-2-yl)ca r bon yl]-3H-p yr r olo[3,2-e]in d ole (seco-CP I-
TMI, 33). A sample of 28 (7.4 mg, 0.014 mmol) in 25% aqueous
HCO2NH4 (0.05 mL) and THF (0.20 mL) afforded 33 (5.8 mg,
94%) as a tan solid: 1H NMR (acetone-d6, 400 MHz) δ 10.30
(br s, 1H), 10.16 (br s, 1H), 8.72 (s, 1H), 7.83 (br s, 1H), 7.35
(t, J ) 2.9 Hz, 1H), 7.06 (d, J ) 2.4 Hz, 1H), 6.99 (s, 1H), 6.55
(dd, J ) 2.0, 2.9 Hz, 1H), 4.75 (app t, J ) 10.4 Hz, 1H), 4.57
(dd, J ) 3.7, 10.4 Hz, 1H), 4.08-4.16 (m, 2H), 4.03 (s, 3H),
3.88 (s, 3H), 3.86 (s, 3H), 3.79 (dd, J ) 8.5, 10.5 Hz, 1H);
FABHRMS (DHB) m/z 455.1248 (M+, C23H22ClN3O5 requires
455.1248).
F igu r e 8. Top: Solvolysis study (UV spectra) of N-BOC-CPI
(13) in 50% CH3OH-aqueous buffer (pH 3, 4:1:20 (v/v/v) 0.1
M citric acid, 0.2 M NaH2PO4, and H2O, respectively. For
clarity, UV traces at only a few time points are shown: t ) 0,
2, 4, 8, 12, 18, 24, 30, and 36 h. Bottom: Data from solvolysis
at 343 nm.
mixture was diluted with H2O and extracted with EtOAc. The
organic layer was dried (Na2SO4), filtered, and concentrated
in vacuo. Chromatography (40% EtOAc/hexane or 2-4% CH3-
OH/CH2Cl2) afforded the desired product.34
(+)-(1S)-33: tan solid: [R]23 +24 (c 2.8 × 10-3, DMF).
5-(Ben zyloxy)-1-(ch lor om et h yl)-1,2-d ih yd r o-3-[(5,6,7-
tr im eth oxy-1H-in d ol-2-yl)ca r bon yl]-3H-p yr r olo[3,2-e]in -
d ole (28). A sample of 10 (25.0 mg, 0.049 mmol), EDCI (28.1
mg, 0.147 mmol), 5,6,7-trimethoxyindole-2-carboxylic acid (14.7
mg, 0.059 mmol), and NaHCO3 (20.5 mg, 0.244 mmol) in DMF
(0.49 mL) afforded 28 (14.8 mg, 56%) as a tan solid: 1H NMR
(acetone-d6, 400 MHz) δ 10.60 (br s, 1H), 10.20 (br s, 1H), 8.06
(br s, 1H), 7.58 (d, J ) 7.4 Hz, 2H), 7.33-7.44 (m, 4H), 7.08
(d, J ) 2.4 Hz, 1H), 6.99 (s, 1H), 6.60-6.61 (m, 1H), 5.27 (s,
2H), 4.78 (app t, J ) 10.4 Hz, 1H), 4.60 (dd, J ) 3.7, 10.4 Hz,
1H), 4.13-4.18 (m, 2H), 4.04 (s, 3H), 3.88 (s, 3H), 3.86 (s, 3H),
3.83-3.87 (m, 1H); MALDI-FTMS (DHB) m/z 545.1697 (M+,
D
(-)-(1R)-33: tan solid: [R]23 -27 (c 1.2 × 10-3, DMF).
D
Gen er a l P r oced u r e for th e Syn th esis of CP I An a -
logu es. A solution of agents 33-37 (1 equiv) in DMF (0.03
M) was treated with NaH (3 equiv, 60% in oil). The suspension
was stirred at 24 °C for 30 min, quenched with the addition
of phosphate buffer (pH 7), extracted with EtOAc, dried (Na2-
SO4), filtered, and concentrated in vacuo. Chromatography
(75% EtOAc/hexane or 4% CH3OH/CH2Cl2) afforded the de-
sired product.34
N2-[(5,6,7-Tr im eth oxyin d ol-2-yl)ca r bon yl]-1,2,8,8a -tet-
r ah ydr ocyclopr opa[c]pyr r olo[3,2-e]in dol-4-on e (CP I-TMI,
38). A sample of 33 (4.3 mg, 0.010 mmol) and NaH (1.1 mg,
0.028 mmol) in DMF (0.32 mL) afforded 38 (4.0 mg, 100%) as
a tan solid: 1H NMR (acetone-d6, 400 MHz) δ 10.82 (br s, 1H),
10.45 (br s, 1H), 7.14 (t, J ) 2.6 Hz, 1H), 7.08 (d, J ) 2.4 Hz,
1H), 6.94 (s, 1H), 6.74 (s, 1H), 6.08 (t, J ) 2.6 Hz, 1H), 4.53
(dd, J ) 5.6, 10.3 Hz, 1H), 4.43 (d, J ) 10.3 Hz, 1H), 4.00 (s,
3H), 3.86 (s, 6H), 2.92 (m, 1H), 1.72 (dd, J ) 4.3, 8.1 Hz, 1H),
1.57 (t, J ) 4.3 Hz, 1H); MALDI-FTMS (DHB) m/z 420.1565
(M+ + H, C23H21N3O5 requires 420.1559). This ring closure was
also effectively accomplished with DBU in CH3CN.
C
30H28ClN3O5 requires 545.1717).
(+)-(1S)-28: tan solid: [R]23 +12 (c 7.4 × 10-3, EtOAc).
D
(-)-(1R)-28: tan solid: [R]23 -14 (c 6.1 × 10-3, EtOAc).
D
Gen er a l P r oced u r e for th e Syn th esis of seco-CP I
Der iva tives. The agents 28-32 (1 equiv) in THF (0.07 M)
and aqueous HCO2NH4 (25%, 22 equiv) were treated with a
catalytic amount of 10% Pd/C, and the suspension was stirred
at 24 °C for 1 h under Ar. The suspension was filtered (Celite)
and concentrated in vacuo. Chromatography (75% EtOAc/
hexane or 4% CH3OH/CH2Cl2) afforded the desired product.34
(+)-(8a R,7bS)-38: tan solid: [R]23 +220 (c 1.7 × 10-3
,
D
DMF).
(-)-(8a S,7bR)-38: tan solid: [R]23 -220 (c 6.1 × 10-4
,
D
(34) Experimental details and full characterization (29-32, 34-37,
and 39-42) are provided in the Supporting Information.
DMF).

Parent Alkylation Subunit of CC-1065
J . Org. Chem., Vol. 65, No. 13, 2000 4111
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
tails for the DNA alkylation studies (efficiency, selectivity, and
rate), experimental details and full characterization for 29-
Ack n ow led gm en t. We gratefully acknowledge the
financial support of the National Institutes of Health
(CA41986) and The Skaggs Institute for Chemical
Biology. We thank the Achievement Rewards for College
Scientists (ARCS) for a predoctoral fellowship (S.E.W.),
the American Cancer Society for a postdoctoral fellow-
ship (S.R.B., 1999-2001), and Dr. Raj K. Chadha for
the X-ray structure determination.
1
32, 34-37, and 39-42, a gel figure (Figure S1), and H NMR
spectra of 5-13, 21, 22, and 28-42. This material is available
free of charge via the Internet at http://pubs.acs.org.
J O000297J