Initial development of a cytotoxic amino-seco-CBI warhead for delivery by prodrug systems

Article abstract of DOI:10.1016/j.bmc.2015.04.034

Abstract Cyclopropabenzaindoles (CBIs) are exquisitely potent cytotoxins which bind and alkylate in the minor groove of DNA. They are not selective for cancer cells, so prodrugs are required. CBIs can be formed at physiological pH by Winstein cyclisation of 1-chloromethyl-3-substituted-5-hydroxy-2,3-dihydrobenzo[e]indoles (5-OH-seco-CBIs). Corresponding 5-NH₂-seco-CBIs should also undergo Winstein cyclisation similarly. A key triply orthogonally protected intermediate on the route to 5-NH₂-seco-CBIs has been synthesised, via selective monotrifluoroacetylation of naphthalene-1,3-diamine, Boc protection, electrophilic iodination, selective allylation at the trifluoroacetamide and 5-exo radical ring-closure with TEMPO. This intermediate has potential for introduction of peptide prodrug masking units (deactivating the Winstein cyclisation and cytotoxicity), addition of diverse indole-amide side-chains (enhancing non-covalent binding prior to alkylation) and use of different leaving groups (replacing the usual chlorine, allowing tuning of the rate of Winstein cyclisation). This key intermediate was elaborated into a simple model 5-NH₂-seco-CBI with a dimethylaminoethoxyindole side-chain. Conversion to a bio-reactive entity and the bioactivity of this system were confirmed through DNA-melting studies (ΔT_m = 13°C) and cytotoxicity against LNCaP human prostate cancer cells (IC₅₀ = 18 nM).

Full text of DOI:10.1016/j.bmc.2015.04.034

Accepted Manuscript
Initial development of a cytotoxic amino-seco -CBI warhead for delivery by
prodrug systems
Elvis A. Twum, Amit Nathubhai, Pauline J. Wood, Matthew D. Lloyd, Andrew
S. Thompson, Michael D. Threadgill
PII:
S0968-0896(15)00332-6
DOI:
http://dx.doi.org/10.1016/j.bmc.2015.04.034
BMC 12246
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
2 February 2015
2 April 2015
10 April 2015
Please cite this article as: Twum, E.A., Nathubhai, A., Wood, P.J., Lloyd, M.D., Thompson, A.S., Threadgill, M.D.,
Initial development of a cytotoxic amino-seco -CBI warhead for delivery by prodrug systems, Bioorganic &
Medicinal Chemistry (2015), doi: http://dx.doi.org/10.1016/j.bmc.2015.04.034
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Graphical Abstract
To create your abstract, type over the instructions in the template box below.
Fonts or abstract dimensions should not be changed or altered.
Initial development of a cytotoxic amino-seco-
CBI warhead for delivery by prodrug systems
Leave this area blank for abstract info.
E. A. Twum, A. Nathubhai, P. J. Wood, M. D. Lloyd, A. S. Thompson, M. D. Threadgill
Dept. Pharmacy & Pharmacology, University of Bath, Bath BA2 7AY, UK
e
r
Me2N
Me2N
u
0.10
0.08
0.06
0.04
0.02
0.00
0.01 eq.
0.05 eq.
0.08 eq.
0.10 eq.
0.13 eq.
0.15 eq.
0.20 eq.
0.40 eq.
0.60 eq.
O
O
p
Me
Me
Cl
N
O
Me
Me
NH
pH 7.0
NH
NCOCF3
N
O
N
O
NHBoc
NH
NH
2
30
40
50
60
70
80
90 100
Temperature (°C)

Bioorganic & Medicinal Chemistry
journal homepage: www.els evier.com
Initial development of a cytotoxic amino-seco-CBI warhead for delivery by
prodrug systems
Elvis A. Twum, Amit Nathubhai, Pauline J. Wood, Matthew D. Lloyd, Andrew S. Thompson and Michael
D. Threadgill*
Medicinal Chemistry, Department of Pharmacy & Pharmacology, University of Bath, Claverton Down, Bath BA2 7AY, UK
Telephone +44 1225 386840; FAX +44 1225 386114; e-mail m.d.threadgill@bath.ac.uk
ARTICLE INFO
ABSTRACT
Article history:
Received
Cyclopropabenzaindoles (CBIs) are exquisitely potent cytotoxins which bind and alkylate in the
minor groove of DNA. They are not selective for cancer cells, so prodrugs are required. CBIs
can be formed at physiological pH by Winstein cyclisation of 1-chloromethyl-3-substituted-5-
hydroxy-2,3-dihydrobenzo[e]indoles (5-OH-seco-CBIs). Corresponding 5-NH2-seco-CBIs
should also undergo Winstein cyclisation similarly. A key triply orthogonally protected inter-
mediate on the route to 5-NH2-seco-CBIs has been synthesised, via selective monotrifluoroacet-
ylation of naphthalene-1,3-diamine, Boc protection, electrophilic iodination, selective allylation
at the trifluoroacetamide and 5-exo radical ring-closure with TEMPO. This intermediate has pot-
ential for introduction of peptide prodrug masking units (deactivating the Winstein cyclisation
and cytotoxicity), addition of diverse indole-amide side-chains (enhancing non-covalent binding
prior to alkylation) and use of different leaving groups (replacing the usual chlorine, allowing
tuning of the rate of Winstein cyclisation). This key intermediate was elaborated into a simple
model 5-NH2-seco-CBI with a dimethylaminoethoxyindole side-chain. Conversion to a bio-
reactive entity and the bioactivity of this system were confirmed through DNA-melting studies
(∆Tm = 13 deg. C) and cytotoxicity against LNCaP human prostate cancer cells (IC50 = 18 nM).
Received in revised form
Accepted
Available online
Keywords:
CBI
Prodrug
TEMPO
Amino-CBI
Cytotoxin
2009 Elsevier Ltd. All rights reserved.
1. Introduction
in tumour tissue. Many of these prodrugs release the cyclopropa-
benzaindole (CBI) alkylating warhead, rather than CPI; these
Duocarmycin SA 1 (Scheme 1) is a natural product of the
cyclopropapyrroloindole (CPI) class with exquisitely potent cyto-
toxicity, having IC50 = 10 pM in L1210 cells.1 It acts by binding
in the minor groove of DNA, with consequent alkylation by the
strained spirocyclopropane ring. In the structure of the related
natural product CPI 2 (CC1065), the side-chain, which binds
non-covalently in the minor groove of DNA prior to covalent
reaction, is extended to interact over a longer section of the
groove. This also provides high cytotoxic potency for this comp-
ound (IC50 = 24 pM in L1210 cells).2 The semisynthetic experim-
ental drug adozelesin 3 carries the same CPI war-head as does 2
but the side-chain is simplified; the binding of this molecule into
the minor groove as a π-stacked dimer has been studied in depth.3
The non-covalently-binding units can also be displayed on each
side of the CPI, as in yatakemycin 4,4,5 increasing potency yet
further (IC50 = 5 pM in L1210 cells). However, this extreme cyto-
toxicity is not selective for tumour cells; compounds such as 2
also cause a delayed lethality in mice and rabbits.6,7
alkylating units have similar potency but improved stability and
synthetic accessibility.8,9 Tietze has disclosed a prodrug 5, in
which the exocyclic OH of the seco compound is masked as a β-
D-galactoside.10,11 This prodrug is essentially non-toxic until the
β-D-galactoside is cleaved by a β-D-galactosidase in the context
of ADEPT; Winstein cyclisation then gives the active CBI. The
exocyclic oxygen is masked as a cyclic hydroxamate ester in 6;
bioreductive cleavage of the N–O bond exposes the phenol, trig-
gering the required Winstein cyclisation.12 Bioreduction of the
nitro group to amino in hypoxic tumour cells is also required for
activation of the nitro-prodrug 7.13 The amine then promotes
Winstein cyclisation, with a sulfonate as the leaving group; 7
shows ~1000-fold selectivity for killing hypoxic HT29 human
colon cancer cells vs. oxic HT29 cells in vitro. Oxidative activat-
ion is required for seco-CPI prodrugs such as 8, developed by the
Bradford group.14 Metabolic oxidation, by CYP1A1 or CYP2W1
(over-expressed in some tumours) gives 9, from which Winstein
cyclisation leads to the active cytotoxin 10 (Scheme 1).
This lack of selectivity has led to research on the development
of prodrugs designed to release the potent cytotoxins selectively
Many of these prodrugs are activated by reduction or oxidat-
ion, with only 5 requiring enzyme-catalysed hydrolysis. As part

H2N
O
Non-covalent minor-
groove-binding unit
OH
N
OMe
OMe
OMe
DNA
NH
O
OMe
N
H
H
O
N
N
N
O
MeO2C
O
OH
N
H
O
OMe
Me
Me
N
N
H
H
N
N
1: Duocarmycin SA
O
2: CC1065
O
3: Adozelesin
N
H
N
H
OH
O
O
OMe
Me2N
MeSOC
N
H
N
H
N
H
N
N
H
N
O
O
O
MeO
OH
N
H
O
H
Cl
Cl
H
4: Yatakemycin
O
Me
N
H
N
OMe
H
N
N
O
O
OSO2Bn
N
O
O
O
O
N
OH
5
Me
6
O
7
HO
HO
OH
H2NOC
Cl
Cl
Cl
NO2
Cl
Cl
N
N
N
H
H
H
N
N
N
O
O
O
Oxidation
by CYP
Winstein
cyclisation
N
H
N
H
N
H
OH
O
8
9
10
Scheme 1. Structures of CPI extreme cytotoxins duocarmycin SA 1, CC1065 2, adozelesin 3 and yatakemycin 4 and of seco-CBI prodrugs 5-7; also metabolic
oxidation of seco-CPI prodrug 8 by CYP1A1 and CYP2W1, leading to insertion of the phenol in 9 and Winstein cyclisation to the active CPI cytotoxin 10.
of our studies towards prodrugs where seco-CBIs are released by
tumour-specific peptidases, we sought an appropriate seco-CBI
carrying an amine at the 5-position for linkage to the peptide
through an amide bond.
the present work. Here, the leaving group X is simply a chloride
(as in many seco-CPIs and seco-CBIs) and the side-chain exten-
sion is the water-solubilising dimethylaminoethoxy group used
by Tietze et al.10 in their galactosidase-activated prodrug 5.
2. Results & discussion
2.2. Chemical synthesis
2.1. Design of a protected seco-CBI
The key step in assembling the target seco-CBI-warhead 12 is
cyclisation of a suitable 1,2,4-trisubstituted naphthalene. Two
cyclisations were investigated: intramolecular 5-exo reaction of a
chiral 1-metallo-2-oxiranylmethylaminonaphthalene (to give
enantiopure material) and radical 5-exo cyclisation of a 1-iodo-2-
allylaminonaphthalene (giving a racemate).
Figure 1 shows our general design of target 5-amino-seco-
CBIs for later linkage to masking peptides. 5-Hydroxy-seco-CBIs
would be inappropriate, as the link to the peptides would have to
be aryl esters, which may be chemically labile in biofluids.
Figure 1 shows the general design of the target 5-amino-seco-
CBIs as structure 11. In this structure, X represents a suitable
leaving group (tuned to provide a balance of reactivity and stabil-
ity), R represents a suitable acid-labile protecting group (or H),
R’ represents substitution on the indole side-chain. This substitut-
ion may be the OH or OMe groups found in natural CPIs (con-
tributing to non-covalent binding in the minor groove), additional
indole or benzofuran units (also contributing to binding) or
water-solubilising entities. Structure 12 is the specific target for
Both routes required an orthogonally protected 1-iodonaphth-
alene-2,4-diamine. Scheme 2 shows the routes investigated there-
to. Firstly, attempted nitration of 4-aminonaphthalen-1-ol 13
towards 4-amino-2-nitronaphthalen-1-ol failed, giving only naph-
tho-1,4-quinone 14 by oxidation. Martius Yellow 15 is a conven-
ient starting material, with the correct pattern of functionalities
on the naphthalene ring, although the nitro groups need to be dis-
criminated. Selective reduction of the 4-nitro group with tin(II)

OH
O
iii
NMe2
RO2C
RO2C
ii
N
H
N
H
30: R = H
31: R = Et
32: R = Et
33: R = H
i
Scheme 3. Preparation of side-chain unit 33. Reagents & conditions: i, EtOH,
HCl, reflux; ii, Me2N(CH2)2Cl.HCl, K2CO3, H2O, CHCl3, reflux; iii, Cs2CO3,
MeOH, H2O, reflux.
Figure 1. General structure of target 5-amino-seco-CBIs 11 and structure of
synthetic target 12. R = H or protecting group; R’ = water-solubilising group
or extension providing additional non-covalent binding; X = leaving group.
chloride, in a modification of the literature method,15 afforded 16.
Attempted protection of the aniline under basic conditions gave
only the carbonate 17, arising from reaction at the more nucleo-
philic phenoxide anion. Under neutral conditions, the amine of
16 is more nucleophilic towards Boc2O, giving 18 in excellent
yield. The phenol was readily triflated but the attempted replace-
ment of OTf in 19 with iodine to give 20 failed; this had prev-
iously been successful in the conversion of 23 to 24.16 An attempt
to remove the triflyloxy group reductively (aiming for N-Boc 3-
nitronaphthalene-1-amine) served only to cleave the O−S bond to
regenerate 18. With that avenue closed, attempts were made to
reduce the nitro group of 18 to the corresponding amine, before
orthogonal protection and introduction of iodine. Hydrogenolysis
proceeded well in the presence of Pd but all attempts to isolate
the very electron-rich naphthalene product failed, giving the N-
Boc naphthoquinoneimine 21. Reduction, followed by trifluoro-
acetylation, gave orthogonally protected 22 in moderate yield but
no routes could be found to elaborate this towards 29.
Figure 2. Structures of non-interconverting atropisomers of 38. Chiral
centres are indicated by green arrows. The iodine is shown in CPK mode to
emphasise the steric bulk.
diamine 25. As reported previously,16 the phenolic OH of 15 was
triflated to give 23 and SNAr reaction with iodide ion furnished
24. From here, reduction of the nitro groups also caused loss of
the iodine, affording naphthalene-1,3-diamine 25. This electron-
rich naphthalene is unstable to autoxidation, so it was protected
as the diBoc derivative 26. The iodine was restored electrophilic-
ally with electrophilic iodine generated from N-iodosuccinimide
and toluenesulfonic acid, affording 27. However, all attempts to
remove one Boc group selectively from 27 failed, with many
acidic conditions also causing loss of the iodine, through a
Wheland intermediate which was observed by NMR.16 Changing
the point at which selectivity is achieved to the protection step,
rather than deprotection, treatment of 25 with trifluoroacetic an-
An alternative route was investigated, in which discrimination
between the nitrogens was achieved by selective protection of the
OH
OH
OH
OH
OTf
NO2
NO2
NO2
NO2
v
ii
iv
vi or vii
NH2
O
NO2
NH2
O
NHBoc
O
NHBoc
I
iii
ix
15
16
18
viii
19
13
i
vi
OBoc
NO2
NHCOCF3
NH2
NO2
NH2
OH
NBoc
R
NBoc
NHBoc
R
O
14
17
22
21
20
NO2
NO2
x
NH2
NHBoc
iv
v
NO2
NO2
NH2
NHBoc
23: R = OTf
24: R = I
26: R = H
27: R = I
I
15
25
xii,iv
vi
xi
NHCOCF3
xi
NHCOCF3
NHBoc
NHBoc
28
29
Scheme 2. Synthetic approaches to the key orthogonally protected 1-iodonaphthalene-2,4-diamine 29 Reagents & conditions: i, KNO3, CF3CO2H, -20°C; ii,
SnCl2, aq. HCl, EtOH, <35°C; iii, Boc2O, DMAP, CH2Cl2; iv, Boc2O, THF; v, Tf2O, pyridine, 0°C; vi, NaI, DMF, 80°C; vii, Pd(OAc)2, Ph3P, HCO2H, 65°C;
viii, H2, Pd/C, MeOH, then exposure to air; ix, K2CO3, Na2S2O4, H2O, CH2Cl2 then (F3CCO)2O, Pri2NEt, CH2Cl2; x, SnCl2, EtOAc; xi, N-iodosuccinimide,
TsOH, THF; xii, (F3CCO)2O, Pri2NEt, THF.

O
O
I
I
Me
NH
NO2
NBoc
O
O
II
OH
O
S
O O
NHBoc
OH
OBn
34
i
36
37
40
i
OH
H
O
H
I
I
Me
NBoc
NCOCF3
NHCOCF3
III
N
CF3
i or iv
X
O
NHBoc
v
NHBoc
NHBoc
OBn
35
29
38
39
vi
Me
Me
Me
Me
N
N
O
O
Me
vii
Me
Me
I
Me
I
NCOCF3
NH
N
CF3
N
CF3
viii
O
O
BocN
NHBoc
NHBoc
NHBoc
O
41
42
44
ix
43
O
O
NMe2
NMe2
Me
NMe2
Me
Me
N
Cl
HO
O
Me
NH
NH
NH
N
N
N
xiii
xii
O
O
O
xiv
O
N
NHBoc
NHBoc
x
47
45
O
NMe2
NMe2
50 NMe2
R1O
Cl
NH
NH
N
N
46: R1 = R2 = H
O
O
47: R1 = H; R2 = Boc
48: R1 = Ac; R2 = H
49: R1 = Ac; R2 = Boc
51: R = Boc
52: R = H
xi
xv
NHR2
NHR
Scheme 4. The ring-closure 34→35 developed by Tietze et al.17 and assembly of target 5-amino-seco-CBIs 47 and 48, via radical ring-closure. Reagents &
conditions: i, MeLi, CuCN, Et2O, THF, N2, -78°C; ii, 4-nitrobenzenesulfonyl chloride, Et3N, toluene; iii, 37, K2CO3, acetone, 50°C; iv, ButLi, ZnCl2, THF,
-78°C; v, H2C=CHCH2Br, K2CO3, acetone, 50°C; vi, H2C=CHCH2Br, KOBut, THF; vii, TEMPO, Bu3SnH, benzene, 60°C; viii, NaOH, H2O, THF; ix, 33, DIC,
DMF, 40°C; x, Zn, THF / AcOH / H2O (1 : 3 : 1), 70°C; xi, Boc2O, THF, reflux; xii, Zn, THF / AcOH / H2O (3 : 1 : 1), 70°C; xiii, MsCl (3 eq.), Et3N, DMF,
0°C, then LiCl; xiv, MsCl (1.4 eq.), Et3N, pyridine, 0°C then LiCl; xv, HCl, 1,4-dioxane.
hydride under very mild conditions led to reaction at the less-
sterically-encumbered 3-NH2; subsequent protection of the 1-
NH2 with Boc led to the orthogonally protected diamine 28 in
modest yield. As for 26→27, 28 was iodinated electrophilically
to provide the key intermediate 29.
a higher-order lithium cuprate (replacing the iodine); attack of the
aryl-copper on the oxirane then proceeds with chelation-controll-
led regioselectivity and stereoselectivity to form 35 (Scheme 4).
The 4-nitrobenzenesulfonate (nosylate) ester 37 of R-glycidol 36
was formed in the usual way; this leaving group was selected as
this and similar oxiranylmethyl nosylates react with nucleophiles
strictly in an SN2 manner,17 retaining the configuration at the
oxirane. The F3CCON‒H is the more acidic in 29 and selective
alkylation of the corresponding anion with 37 afforded 38 in
good yield. Interestingly, 38 was formed as a mixture of chromat-
ographically separable diastereomeric atropisomers 38A and
38B. The bulky iodine twists the N-oxiranylmethyltrifluoroacet-
amide out of the plane of the naphthalene by rotation about the
naphthalene‒N bond, generating a chiral centre. Interconversion
of the atropisomers by rotation about this C‒N bond is so sev-
erely impeded that it does not occur at ambient temperature.
Scheme 3 shows the preparation of the non-covalently binding
and water-solubilising indole side-chain. The carboxylic acid of
30 was protected as the ethyl ester 31 before the corresponding
phenoxide was alkylated with N-(2-chloroethyl)dimethylamine to
afford 32. Hydrolysis of the ester with caesium carbonate in
aqueous methanol gave the required carboxylic acid 33.
With these building blocks in hand, methods were investigated
to build the target seco-CBIs. Initial attempts sought to exploit
the ring-closure developed by Tietze et al.,17 in which the protec-
ted iodo-(oxiranylmethylamino)naphthalene 34 is metallated with

Owing to the presence of the conventional R-chiral centre in the
oxirane, the atropisomers are diastereomeric. 1H, 13C and 19F
NMR analysis of atropisomer 38A also showed a 3:2 mixture of
rotamers; this probably reflects restricted rotation of the tertiary
amide carbonyl‒N bond in the trifluoroacetamide. No similar
rotamers were evident in the corresponding spectra of 38B, sug-
gesting that one rotamer about the amide bond predominates
greatly in this diastereoisomer. Figure 2 shows the proposed
structures of these atropisomers, using structures minimised in
Sybyl and rendered with Accelrys DS.
Unfortunately, the metalation / cyclisation reaction developed
by Tietze et al.17 failed to form 39 from a mixture of diastereoiso-
mers of 38 The only isolable product (in high yield) was 40, der-
ived from loss of the F3CCO protection (Scheme 4). We specul-
ate that the failure to cyclise was due to the presence of the
BocN‒H, which may have quenched the cuprate before it could
Scheme 5. Proposed Winstein cyclisation of 52 in aq. buffer (pH 7.0).
stoichiometric MsCl in pyridine obviated this side-reaction to
generate the required protected seco-CBI 51. Finally the Boc pro-
tection was acidolysed to form 52 as the stable dihydrochloride,
with the cation being unable to trigger Winstein cyclisation.
-
engage in metallation. Nucleophilic CN may then have cleaved
the trifluoroacetamide. The NMR spectra of 40 showed the pres-
ence of the diastereomeric atropisomers, which were not separ-
able in this case. Furuyama et al. reported that 4-iodophenol
(carrying an acidic proton) is readily metallated with But4ZnLi2
and the arylzinc reacts with allyl bromide to give exclusively 4-
allylphenol.18 This process also failed to cyclise 38 to 40, giving
only unreacted 38. The door was closed on this potential route to
homochiral CBIs carrying NHBoc.
2.3. Biochemical and cell-biological evaluation
The CBIs act by binding and alkylating in the minor groove of
DNA. When they do so, they cause an increase in the temperature
at which the strands of double-stranded DNA separate, the so-
called melting temperature (Tm). Indeed, the CPI CC-1065 2
causes an increase in Tm of up to 31 deg. C.2 Thus, to confirm that
the seco-CBI 52 did form a DNA-reactive entity (presumably the
CBI 53 formed by Winstein cyclisation (Scheme 5)) in aqueous
buffer at pH 7.0 and that the expected 53 did bind and react with
double-stranded DNA, a DNA-melting study was carried out.
Firstly, to ensure that the putative electrophile 53 was fully
formed in the buffer, the dihydrochloride salt of 52 was incub-
ated with double-stranded calf-thymus DNA in phosphate buffer
at pH 7.0 for 1 h and for 24 h and the increase in melting temper-
ature (∆Tm) was determined for both incubations. The identical
values demonstrated that the naphthylamine was present as the
free base at pH 7.0 and that formation of an electrophilic product
(putatively 53) was complete within 1 h. Thus the full DNA-
melting study was conducted with an incubation time of 1 h (Fig-
ure 3). Panel A shows the primary changes in UV absorption at
256 nm with temperature, using nine different molar ratios of
52/53 to DNA (calculated per base-pair), whereas panel B shows
the first derivative of these primary curves, allowing the Tm to be
determined as the temperature at the apex of the derivative curve.
In panel C, ∆Tm is plotted against the molar ratio of 52/53 per
DNA base-pair. The data confirm that generation of the electro-
phile (putatively 53) takes place rapidly at pH 7.0 and that it does
bind to and alkylate double-stranded DNA, giving a maximum
∆Tm of 13 deg. C. This maximum is achieved with ca. 0.2 drug
molecules per base-pair. Of course, 52 and 53 are racemates. It is
known that only one enantiomer of CBIs binds strongly to DNA,
the other having a shape which does not match the curve of the
minor groove. Indeed, (-)-duocarmycin SA 1 alkylates DNA at
least ten-times more efficiently than does its (+)-enantiomer.21
Thus it may be argued that the appropriate ratio to consider is the
ca. 0.1 ratio of the correct enantiomer per base-pair.
Therefore, a route was developed in which cyclisation of an
achiral starting material would give a seco-CBI in racemic form.
Boger and McKie carried out a 5-exo radical cyclisation of 2-(N-
allyl-BocN)-4-benzyloxy-1-iodonaphthalene with TEMPO and
Bu3SnH to afford racemic 5-BnO-3-Boc-1-(2,2,6,6-tetramethyl-
piperidin-1-yloxy)methyl)-2,3-dihydrobenzo[e]indol-5-yl)carb-
amate.19 Similar cyclisations have been disclosed by Stevenson et
al.13 Alkylation of 29 with allyl bromide in the presence of excess
K2CO3 (Scheme 4) introduced two allyl groups (forming 41) but
selectivity for the more acidic N‒H was achieved using one equi-
valent of KOBut to form the monoanion, leading to 42. As for 38,
42 formed two separable but non-interconvertible conformers, in
a 4:1 ratio. Clearly, one of the centres of atropisomerism is the
naphthalene‒N(allyl)COCF3 bond which is restricted by the adj-
acent iodine but one may only speculate that the other source of
atropisomerism may be the naphthalene‒NHBoc bond, the
rotation of which might be impeded by the peri 5-H. Cyclisation
of the atropisomeric mixture to 43 was achieved in 60% yield
using TEMPO and Bu3SnH in boiling benzene. Tricycle 43 cont-
ains the three groups protected with completely mutually ortho-
gonal masking groups. Selective basic hydrolysis of the trifluoro-
acetyl protection revealed the secondary amine in 44, to which
the 5-(2-dimethylaminoethoxy)indole-2-carbonyl side-chain was
affixed in a carbodiimide coupling, affording 45. The O-(2,2,6,6-
tetramethylpiperidin-1-yl) protection was removed reductively,
after much optimisation. Reaction of 45 with zinc powder in THF
/ AcOH / H2O (1 : 3 : 1) gave a mixture of 46-49, in which the N-
Boc had been removed in 46 and 48 and the exposed alcohol had
been esterified in 48 and 49. Zinc acetate is a known Lewis acid
catalyst for acetylation of primary alcohols.20 The amine 48 was
reprotected to give 49. Reduction of 45 under less acidic condit-
ions (THF / AcOH / H2O (3 : 1 : 1)) allowed uneventful cleavage
of the N‒O bond to provide 47 in good yield. It only remained
for the OH to be converted into the chloride and for the N-Boc to
be removed. Reaction of 47 with mesyl chloride in DMF in the
presence of Et3N aimed to form the mesylate, which would be
displaced by chloride ion from added excess LiCl. These condit-
ions introduced the required chlorine but reaction of the MsCl
with solvent DMF generated a Vilsmeier reagent, which cleaved
the N‒Boc and formed the dimethylformamidine in 50. Use of
The cytotoxicity of 52 (presumably forming 53) towards
LNCaP human prostate carcinoma cells was measured using the
MTS assay. This cell line produces Prostate-Specific Antigen
(PSA) and its proliferation is sensitive to androgens.22 An IC50 of
18 3 nM was observed for the antiproliferative effect on these
cells. To demonstrate that masking the naphthylamine with a car-
bonyl (in the Boc-protected precursor 51) does diminish the cyto-
toxic potency, a similar assay was conducted for 51. This
compound indeed proved to be markedly less active, with IC50
=
218 47 nM. In each case, only one of the enantiomers is likely
to be potent.

group in 51. The target amino-seco-CBI 52 was generated by
simple rapid acidolysis of the Boc protection. The mutual ortho-
gonality of the protection in 43 means that this intermediate is
potentially highly versatile, with opportunities to modify the
order in which the subsequent steps are carried out, leading to
access to diverse amino-seco-CBIs and to the potential for attach-
ing the deactivating prodrug linker to the 5-NH2 prior to
“arming” the seco-CBI warhead by installing the leaving group.
1.2
1.0
A
0.01 eq.
0.05 eq.
0.08 eq.
0.10 eq.
0.13 eq.
0.15 eq.
0.20 eq.
0.40 eq.
0.60 eq.
0.8
0.6
0.4
0.2
0.0
Confirmation that the amino-seco-CBI system is a useful war-
head for release / delivery by peptide prodrugs was provided by
allowing 52 to undergo Winstein cyclisation in aqueous buffer at
pH 7.0 to provide the CBI 53. This was demonstrated to be com-
plete in < 1 h by the equivalence of DNA-binding at 1 h and 24 h
time-points. DNA-melting studies showed the generated 53 to
bind to and react with double-stranded DNA in the stoichiometry
of 0.2 molecules per base-pair (0.1 molecules if only the active
enantiomer is considered). The delivered CBI is also a highly
potent cytotoxin towards LNCaP human tumour cells, with IC50
in the nanomolar range. Further studies optimising the potency of
this warhead towards the picomolar range of 1-4 by modifying
the non-covalently-binding side-chain will be published later.
30
40
50
60
70
80
90 100
Temperature (°C)
0.10
0.08
0.06
0.04
0.02
0.00
0.01 eq.
0.05 eq.
0.08 eq.
0.10 eq.
0.13 eq.
0.15 eq.
0.20 eq.
0.40 eq.
0.60 eq.
B
Thus we have identified 43 as an important core intermediate
for development of potential warheads of extreme potency for
selective delivery through peptide prodrug systems.
30
40
50
60
70
80
90 100
Temperature (°C)
4. Experimental section
4.1. General
14.0
12.0
10.0
8.0
C
1H and 13C NMR spectra were recorded at 400.04 or 500.13
MHz for 1H NMR and 100.59 or 125.76 MHz for 13C NMR using
CDCl3, containing SiMe4, unless otherwise noted. 13C NMR
spectra were assigned using HSQC and HMBC. MS data were
obtained using electrospray ionisation using a microTOF instru-
ment (Bruker Daltonics, Bremen, Germany), calibrated using
sodium formate. Mps were obtained using a hot-stage micro-
scope (Reichert-Jung). Experiments were conducted at ambient
temperature, unless otherwise noted. Solutions in organic
solvents were dried with MgSO4. Synthetic methods for 14, 16-
19, 21, 22, 31-33, 37, 38, 40, 41 are reported in the Supplement-
ary Information; syntheses of 23-29 were reported previously.16
6.0
4.0
2.0
0.0
0.0
0.2
0.4
0.6
Molar equivalents of 52 / 53
Figure 3. DNA-melting studies on 52/53. Equivalents / molar equivalents
refer to the number of equivalents of 52/53 per base-pair. Panel A: Graphs of
absorbance (256 nm) vs. temperature for different molar ratios of 52/53.
Panel B: First derivative graphs from which the melting temperature (Tm)
values were measured. Panel C: Graph of Tm vs. molar equivalents of 52/53.
4.2. 1,1-Dimethylethyl N-(1-iodo-2-(N-(prop-2-enyl)-2,2,2-tri-
fluoroacetamido)naphthalen-4-yl)carbamate (42A & 42B)
Compound 29 (1.22 g, 2.5 mmol) was stirred with KOBut
(318 mg, 2.8 mmol) in THF (10 mL) under N2 for 30 min, foll-
owed by addition of 3-bromopropene (973 mg, 8.04 mmol). The
mixture was stirred for 2 h and heated to 50°C for 16 h. Sat. aq.
NaHCO3 was added. Extraction (EtOAc), drying, evaporation
3. Conclusions
In this paper, the challenge of assembling a seco-CBI carry-
ing a protected amine at the 5-position has been overcome. The
key steps were preparation of the orthogonally protected 1-iodo-
naphthalene-2,4-diamine 29, allowing selective allylation at the
2-trifluoroacetamido group, and radical 5-exo ring-closure of the
N-allyl compound 42 to the seco-CBI core in 43. This cyclisation
uses an achiral educt and achiral reagents, leading to racemic 43.
The biological activities of CBIs (binding in the minor groove of
double-stranded DNA, cytotoxicity) reside in one enantiomer
only; the development of a stereoselective radical ring-closure
will be the subject of a later publication. This route has the bene-
fit that 43 contains three functionalities with completely ortho-
gonal protection; the OH is protected by the O-tetramethylpiper-
idine (removable by reduction), the secondary amine site for
introduction of the non-covalently bonding side-chain is protec-
ted as the base-labile trifluoroacetamide and the primary aniline
(later to trigger Winstein cyclisation) is masked by the acid-labile
Boc group. These benefits were exploited by firstly exposing the
pyrrolidine nitrogen and attaching the indole side-chain in 45.
The OH was revealed and converted into the required leaving
→ petroleum ether /
EtOAc 19:1) gave 42A (200 mg, 15%) as a yellow solid: mp
and chromatography (petroleum ether
1
146-147°C; H NMR (NOESY) δ 1.56 (9 H, s, But), 3.95 (1 H,
dd, J = 14.2, 7.4 Hz, propenyl 1-H), 4.84 (1 H, dd, J = 14.2, 6.1
Hz, propenyl 1-H), 5.12 (1 H, dd, J = 17.0, 1.3 Hz, propenyl 3-
H), 5.21 (1 H, dd, J = 10.1, 1.1 Hz, propenyl 3-H), 5.94 (1 H, m,
propenyl 2-H), 7.27 (1 H, s, NH), 7.51 (1 H, dd, J = 8.2, 6.9, 1.2
Hz, 6-H), 7.60 (1 H, dd, J = 8.3, 6.9, 1.3 Hz, 7-H), 7.67 (1 H, d, J
= 8.3 Hz, 5-H), 8.17 (1 H, d, J = 8.1 Hz, 8-H), 8.27 (1 H, s, 3-H);
13C NMR δ 27.30 (CMe3), 53.30 (propenyl 1-C), 80.76 (CMe3),
91.92 (1-C), 115.16 (q, J = 288.6 Hz, CF3), 119.84 (3-C), 120.03
(propenyl 3-C), 121.93 (5-C), 125.63 (6-C), 126.43 (4a-C),
127.97 (7-C), 129.49 (propenyl 2-C), 131.56 (8-C), 134.31 (4-C),
135.26 (8a-C), 136.78 (2-C), 151.32 (Boc C=O), 156.35 (q, J =
36.3 Hz, CF3C=O); 19F NMR δ -68.20 (CF3); MS m/z 559.0149
(M + K) (C20H20F3IN2KO3 requires 559.0107); 543.0371 (M +

Na) (C20H20F3IN2NaO3 requires 543.0368); 538.0828 (M + NH4+)
(C20H24F3IN3O3 requires 538.0809). Further elution gave 42B
(750 mg, 57%) as a yellow oil: 1H NMR (NOESY) δ 1.55 (9 H,
s, But), 3.82 (1 H, dd, J = 14.5, 7.6 Hz, propenyl 1-H), 4.96 (1 H,
dd, J = 14.5, 5.6 Hz, propenyl 1-H), 5.18 (1 H, dq, J = 18.0, 1.3
Hz, propenyl 3-H) 5.22 (1 H, dq, J = 11.0, 1.0 Hz, propenyl 3-H),
5.95 (1 H, m, propenyl 2-H), 7.01 (1 H, s, NH), 7.61-7.67 (2 H,
m, 6,7-H), 7.85 (1 H, m, 5-H), 7.93 (1 H, s, 3-H), 8.30 (1 H, m,
8-H); 13C NMR δ 28.27 (CMe3), 54.00 (propenyl 1-C), 81.56
(CMe3), 115.89 (q, J = 288.7 Hz, CF3), 118.90 (3-C), 120.59
(propenyl 3-C), 120.66 (5-C), 127.92 (6-C or 7-C), 128.67 (7-C
or 6-C), 130.30 (propenyl 3-C), 134.38 (8-C), 134.70 (8a-C),
135.39 (4-C), 140.14 (2-C), 152.45 (Boc C=O), 156.53 (q, J =
36.5 Hz, F3CC=O); 19F NMR δ -68.55 (s, CF3); MS m/z 559.0154
(M + K) (C20H20F3IN2KO3 requires 559.0107); 543.0437 (M +
Na)+ (C20H20F3IN2NaO3 requires 543.0368); 521.0516 (M + H)+
(C20H21F3IN2O3 requires 521.0549); 486.9814 [(M - But) + Na)]+
(C16H12F3IN2NaO3 requires 486.9742).
7.1 Hz, 8-H or 7-H), 7.51 (1 H, s, 4-H), 7.69 (2 H, m, 6,9-H2);
13C NMR δ 17.07 (piperidine 4-C), 20.07 (piperidine-Me), 20.26
(piperidine-Me), 28.33 (CMe3), 32.93 (piperidine-Me), 33.34
(piperidine-Me), 39.45 (piperidine 3-C or 5-C), 39.59 (piperidine
5-C or 3-C), 41.05 (1-C), 51.29 (2-C), 59.63 (piperidine 2-C or 6-
C), 59.72 (piperidine 6-C or 2-C), 76.68-77.32 (NOCH2), 80.54
(CMe3), 105.52 (4-C), 116.77 (5a,9b-C2), 120.98 (6-C or 9-C),
121.52 (7-C or 8-C), 123.23 (9-C or 6-C), 126.29 (8-C or 7-C),
131.41 (9a-C), 133.52 (5-C), 149.40 (3a-C), 153.34 (C=O); MS
m/z 476.2924 (M + Na)+ (C27H39N3NaO3 requires 476.2810),
454.3107 (M + H)+ (C27H40N3O3 requires 454.3070).
4.5. 1,1-Dimethylethyl N-(3-(5-(2-(dimethylaminoethoxy)-
indole-2-carbonyl)-1-((2,2,6,6-tetramethylpiperidin-1-yloxy)-
methyl)-2,3-dihydrobenzo[e]indol-5-yl)carbamate (45)
N,N’-Diisopropylcarbodiimide (106 mg, 0.81 mmol) was
stirred with 33 (101 mg, 0.41 mmol) and HOBt (125 mg, 0.81
mmol) in dry DMF (10 mL) at 0°C under N2 for 1 h. Compound
44 (203 mg, 0.37 mmol) in dry DMF (10 mL) was added. The
mixture was allowed to warm slowly to 20°C during 16 h and
was heated at 40°C for 2 h. Sat. aq. NaHCO3 was added and the
mixture was extracted (EtOAc). The extract was washed (water,
brine). Drying, evaporation and chromatography (EtOAc →
EtOAc / MeOH / Et3N 900:100:1) gave 45 (163 mg, 64%) as a
yellow oil: IR νmax 3447, 1727, 1693 cm-1; 1H NMR (NOESY) δ
0.88 (3 H, s, piperidine-Me), 1.00 (3 H, s, piperidine-Me), 1.04 (3
H, s, piperidine-Me), 1.17 (3 H, s, piperidine-Me), 1.24-1.50 (6
H, m, piperidine 3,4,5-H6), 1.54 (9 H, s, But), 2.45 (6 H, s,
NMe2), 2.88 (2 H, t, J = 5.5 Hz, CH2NMe2), 3.87 (1 H, t, J = 8.8
Hz, NOCH), 4.05 (1 H, m, 1-H), 4.13 (1 H, dd, J = 9.0, 4.8 Hz,
NOCH), 4.17 (2 H, t, J = 5.6 Hz, OCH2), 4.60 (1 H, t, J = 9.1 Hz,
2-H), 4.81 (1 H, dd, J = 10.2, 1.2 Hz, 2-H), 6.90 (1 H, s, Boc
NH), 7.00 (1 H, s, indole 7-H), 7.02 (1 H, dd, J = 8.9, 2.4 Hz,
indole 6-H), 7.13 (1 H, d, J = 2.1 Hz, indole 3-H), 7.36 (1 H, d, J
= 8.9 Hz, indole 7-H), 7.42 (1 H, t, J = 7.5 Hz, 7-H), 7.50 (1 H, t,
J = 7.2 Hz, 8-H), 7.84 (1 H, d, J = 8.2 Hz, 9-H), 7.89 (1 H, d, J =
8.5 Hz, 6-H), 8.90 (1 H, s, 4-H), 9.52 (1 H, s, indole NH); 13C
NMR δ 16.94 piperidine 4-C), 20.01 (piperidine-Me), 20.16
(piperidine-Me), 28.30 (CMe3), 33.02 (piperidine-Me), 33.07
(piperidine-Me), 39.48 (piperidine 3-C or 5-C), 39.53 (piperidine
5-C or 3-C), 40.20 (1-C), 45.51 (NMe2), 54.66 (2-C), 58.06
(OCH2CH2NMe2), 59.77 (piperidine 2-C or 6-C), 59.82
(piperidine 6-C or 2-C), 66.14 (OCH2), 77.64 (NOCH2), 80.69
(CMe3), 103.71 (indole 3-C), 105.58 (indole 4-C), 111.95 (4-C),
112.61 (indole 7-C), 116.79 (indole 6-C), 122.13 (6-C), 122.62
(9b-C), 123.91 (9-C), 124.47 (7-C), 126.60 (8-C), 128.31 (indole
3a-C), 130.07 (9a-C), 130.99 (5a-C), 131.26 (indole 7a-C),
133.71 (5-C), 141.39 (3a-C), 153.50 (indole 5-C + indole C=O),
160.23 (Boc C=O); MS m/z 706.3970 (M + Na)+ (C40H53N5NaO5
requires 706.3944).
4.3. 1,1-Dimethylethyl N-(1-(2,2,6,6-tetramethylpiperidin-1-
yloxy)methyl)-3-(trifluoroacetyl-2,3-dihydrobenzo[e]indol-5-
yl)carbamate (43)
TEMPO (35.7 mg, 0.23 mmol) and Bu3SnH (21 mg, 73 µmol)
were stirred with 42 (38.1 mg, 73 µmol) in benzene (2.0 mL) at
60°C for 135 min. Additional Bu3SnH (21 mg, 73 µmol) was
added at the 30 min, 60 min and 90 min time-points. Evaporation
and chromatography (petroleum ether / EtOAc 49:1 → 19:1)
gave 43 (24.1 mg, 60%) as a yellow solid: mp 160-161°C; IR
1
νmax 3364, 3129, 3266, 1700 cm-1; H NMR (NOESY) δ 0.90 (3
H, s, piperidine-Me), 1.00 (3 H, s, piperidine-Me), 1.07 (3 H, s,
piperidine-Me), 1.20 (3 H, s, piperidine-Me), 1.25, (1 H, m, pip-
eridine 4-H), 1.29 (1 H, m, piperidine 3-H or 5-H), 1.35 (1 H, m,
piperidine 3-H or 5-H), 1.42 (1 H, m, piperidine 3-H or 5-H),
1.54 (1 H, m, piperidine 4-H), 1.54 (9 H, s, But), 1.66 (1 H, m,
piperidine 3-H or 5-H), 3.83 (1 H, t, J = 8.8 Hz, NOCH), 3.99 (1
H, m, 1-H), 4.08 (1 H, dd, J = 8.9, 4.3 Hz, NOCH), 4.30 (1 H, t, J
= 8.7 Hz, 2-H), 4.61 (1 H, d, J = 10.9 Hz, 2-H), 6.10 (1 H, s,
NH), 7.47 (1 H, t, J = 7.8 Hz, 7-H), 7.53 (1 H, t, J = 7.6 Hz, 8-H),
7.81 (1 H, d, J = 8.3 Hz, 9-H), 7.90 (1 H, d, J = 8.4 Hz, 6-H),
8.79 (1 H, s, 4-H); 13C NMR δ 16.97 (piperidine 4-C), 19.97
(piperidine-Me), 20.05 (piperidine-Me), 28.30 (CMe3), 32.92
(piperidine-Me), 32.96 (piperidine-Me), 39.52 (piperidine 3-C or
5-C), 39.60 (piperidine 5-C or 3-C), 39.86 (1-C), , 52.42 (2-C),
59.88 (piperidine 2,6-C2), 77.26 (NOCH2), 81.05 (CMe3), 111.17
(4-C), 116.20 (q, J = 288.1 Hz, CF3), 122.15 (6-C), 123.13 (9b-
C), 124.01 (9-C), 125.36 (7-C), 125.52 (5a-C), 127.06 (8-C),
129.77 (9a-C), 134.19 (5-C), 139.56 (3a-C), 153.34 (Boc C=O),
154.28 (q, J = 37.5 Hz, CF3C=O); 19F NMR δ -72.21 (s, CF3);
MS m/z 572.2804 (M + Na)+ (C29H38F3N3NaO4 requires
572.2712), 550.2937 (M
550.2893), 516.2095 (M
(C25H30F3N3NaO requires 516.2086).
+
H)+ (C29H39F3N3O4 requires
2-methylpropene
Na)+
–
+
4.6. 1,1-Dimethylethyl N-(3-(5-(2-dimethylaminoethoxy)-
indole-2-carbonyl)-1-hydroxymethyl-2,3-dihydrobenzo[e]-
indol-5-yl)carbamate (47)
4.4. 1,1-Dimethylethyl N-(1-((2,2,6,6-tetramethylpiperidin-1-
yloxy)methyl)-2,3-dihydrobenzo[e]indol-5-yl)carbamate (44)
Water (1.0 mL), Zn dust (934 mg, 14.4 mmol) and AcOH (1.0
mL) were added sequentially to 45 (123 mg, 0.18 mmol) in THF
(3.0 mL). The mixture was heated at 70°C in a sealed pressure
tube for 16 h. The cooled mixture was filtered (Celite®) and the
solvents were evaporated. Sat aq. NaHCO3 was added and the
suspension was extracted (EtOAc). The extract was washed
(water, brine). Drying, evaporation and chromatography (EtOAc
→ EtOAc / MeOH / Et3N 900:100:1) gave 47 as an off-white
solid (69.3 mg, 71%): mp 219-220°C; IR νmax 3567, 3423, 3293,
1684 cm-1; 1H NMR ((CD3)2SO) (NOESY) δ 1.56 (9 H, s, But),
2.36 (6 H, s, NMe2), 2.79 (2 H, t, J = 5.7 Hz, CH2NMe2), 3.23 (1
Compound 43 (70.4 mg, 0.13 mmol) in THF (4.0 mL) was
stirred with aq. NaOH (5.0 M, 4.0 mL) for 105 min. The mixture
was diluted with water and extracted with EtOAc. Washing
(brine), drying and evaporation gave 44 (60.7 mg, quant.) as a
yellow oil: IR νmax 3376, 1715 cm-1; 1H NMR (NOESY) δ 1.06 (3
H, s, piperidine-Me), 1.14 (6 H, s, 2 × piperidine-Me), 1.16 (3 H,
s, piperidine-Me), 1.16-1.42 (6 H, m, piperidine 3,4,5-H6), 1.54
(9 H, s, But), 3.70-3.75 (2 H, m, 2-H2), 3.80-3.90 (2 H, m, 1-H +
NOCH), 4.00 (1 H, dd, J = 7.6, 4.2 Hz, NOCH), 6.86 (1 H, s,
Boc NH), 7.20 (1 H, t, J = 7.5 Hz, 7-H or 8-H), 7.38 (1 H, t, J =

H, m, CHOH), 3.85 (1 H, m, CHOH), 4.02 (1 H, m, 1-H), 4.15 (2
H, t, J = 5.8 Hz, OCH2CH2), 4.74 (2 H, m, 2-H), 5.11 (1 H, br,
OH), 6.98 (1 H, dd, J = 8.9, 2.4 Hz, indole 6-H), 7.18 (1 H, d, J =
1.7 Hz, indole 3-H), 7.24 (1 H, d, J = 2.3 Hz, indole 4-H), 7.46 (1
H, d, J = 8.8 Hz, indole 7-H), 7.48 (1 H, t, J = 8.2 Hz, 7-H), 7.8
(1 H, t, J = 7.8 Hz, 8-H), 7.96 (1 H, d, J = 8.3 Hz, 9-H), 8.06 (1
H, d, J = 8.4 Hz, 6-H), 8.58 (1 H, s, 4-H), 9.29 (1 H, s, Boc NH),
11.66 (1 H, s, indole NH); 13C NMR ((CD3)2SO) δ 28.16 (CMe3),
42.81 (1-C), 45.35 (NMe2), 54 (2-C), 57.65 (CH2NMe2), 62.81
(CH2OH), 65.95 (OCH2CH2), 78.95 (CMe3), 103.31 (indole 4-C),
105.35 (indole 3-C), 113.13 (indole 7-C), 113.61 (4-C), 115.78
(indole 6-C), 123.38 (9b-C), 123.65 (9-C), 123.88 (6-C), 124.01
(7-C), 125.78 (5a-C), 126.62 (indole 2-C), 129.72 (9a-C), 131.08
(indole 3a-C), 131.66 (indole 7a-C), 134.20 (5-C), 141.01 (3a-C),
152.90 (indole 5-C), 154.07 (Boc C=O), 160.26 (indole C=O);
0.16 mmol). The mixture was stirred at 0°C for 1 h. LiCl (134
mg, 3.2 mmol) was added and the mixture was stirred at 20°C for
3 d. Water (5 mL) was added and the solvents were evaporated
→ EtOAc / MeOH / Et
3N
(< 30°C). Chromatography (EtOAc
950:50:1) gave 50 (21 mg, 76%) as a yellow solid: mp 229-
230°C; 1H NMR (NOESY) δ 2.42 (6 H, s, CH2NMe2), 2.84 (2 H,
t, J = 5.6 Hz, CH2NMe2), 3.04 (3 H, br, formamidine-Me), 3.10
(3 H, br, formamidine-Me), 3.45 (1 H, t, J = 10.9 Hz, CHCl),
3.98 (1 H, dd, J = 11.4, 3.1 Hz, CHCl), 4.14 (1 H, m, 1-H), 4.16
(2 H, t, J = 5.6 Hz, OCH2), 4.63 (1 H, t, J = 8.6 Hz, 2-H), 4.78 (1
H, dd, J = 9.3, 1.5 Hz, 2-H), 7.00 (1 H, d, J = 8.8 Hz, indole 6-
H), 7.02 (1 H, d, J = 2.3 Hz, indole 3-H), 7.14 (1 H, d, J = 2.2
Hz, indole 4-H), 7.35 (1 H, d, J = 8.9 Hz, indole 7-H), 7.38 (1 H,
t, J = 7.3 Hz, 7-H), 7.50 (1 H, dd, J = 8.0, 1.1 Hz, 8-H), 7.67-7.69
(2 H, m, 9-H + formamidine-H), 7.70 (1 H, s, 4-H), 8.47 (1 H, d,
J = 8.3 Hz, 6-H), 9.59 (1 H, s, indole NH); 13C NMR δ 35 (br,
formamidine-Me), 40 (br, formamidine-Me), 43.30 (1-C), 45.63
(CH2NMe2), 46.08 (CH2Cl), 55.05 (2-C), 58.21 (OCH2CH2N),
66.30 (OCH2), 103.70 (indole 4-C), 104.68 (4-C), 105.97 (indole
3-C), 112.63 (indole 7-C), 117.10 (indole 6-C), 118.26 (9b-C),
121.97 (9-C), 123.64 (7-C), 125.76 (6-C), 127.26 (8-C), 127.81
(5a-C), 128.25 (indole 3a-C), 129.75 (9a-C), 130.81 (indole 7-C),
131.36 (indole 7a-C), 142.03 (3a-C), 150.91 (5-C), 153.03 (form-
amidine CH), 153.66 (indole 5-C), 160.49 (indole C=O); MS m/z
MS m/z 567.2608 (M
567.258340), 545.2829 (M + H) (C31H37N4O5 requires 545.2764).
+
Na) (C31H36N4NaO5 requires
4.7. 1,1-Dimethylethyl N-(1-acetoxymethyl-3-(5-(2-dimethyl-
aminoethoxy)indole-2-carbonyl)-2,3-dihydrobenzo[e]indol-5-
yl)carbamate (49)
Zn dust (354 mg, 5.4 mmol) was added to 45 (123 mg, 0.18
mmol) in THF (2.0 mL), AcOH (6.0 mL) and H2O (2.0 mL) in a
pressure tube. The tube was closed and the mixture was heated
for 70°C for 2 h. Further activated Zn dust (354 mg, 5.4 mmol)
was added, the tube was resealed and the mixture was heated at
70°C for 21 h, then cooled to 20°C. The mixture was diluted with
THF and filtered (Celite®). Evaporation and chromatography
520.2375 (M
+
H)+ (C29H3237ClN5O2 requires 520.2293),
518.2341 (M + H)+ (C29H3235ClN5O2 requires 518.2323).
4.9. 1,1-Dimethylethyl N-(1-chloromethyl-3-(5-(2-dimethyl-
aminoethoxy)indole-2-carbonyl)-2,3-dihydrobenzo[e]indol-5-
yl)carbamate (51)
→ EtOAc / MeOH / Et
(EtOAc
3N 950:50:1) gave two fractions,
each comprising two compounds. The first fraction (14.0 mg)
consisted of 46 (MS m/z 587.2945 (M + H) (C33H39N4O6 requires
587.2870)) and 47 (MS m/z 487.2432 (M + H) (C28H31N4O4
requires 487.2345)), whilst the second fraction (41.3 mg) consis-
ted of 48 (MS m/ z 445.2328 (M + H) (C26H29N4O3 requires
445.2240)) and 49 (MS m/z 545.2815 (M + H) (C31H37N4O5
requires 545.2764)). The first fraction (46 + 47) was boiled under
reflux with Boc2O (63 mg, 0.29 mmol) in dry THF (10 mL)
under N2 for 16 h. Evaporation and chromatography (EtOAc →
EtOAc / MeOH / Et3N 900:100:1) gave 49 as a pale buff solid
(14.5 mg, 14%): mp 170-171°C; 1H NMR (CD3OD) (NOESY) δ
1.61 (9 H, s, But), 2.03 (3 H, s, COMe), 2.46 (6 H, s, NMe2), 2.88
(2 H, t, J = 5.4 Hz, CH2NMe2), 4.10 (1 H, dd, J = 10.8, 8.0 Hz,
CHOAc), 4.19 (2 H, t, J = 5.4 Hz, OCH2CH2NMe2), 4.19 (1 H,
m, 1-H), 4.59 (1 H, dd, J = 10.8, 3.6 Hz, CHOAc), 4.68 (2 H, d, J
= 4.4 Hz, 2-H), 7.04 (1 H, dd, J = 9.0, 2.4 Hz, indole 6-H), 7.11
(1 H, s, indole 3-H), 7.23 ( 1 H, d, J = 2.1 Hz, 4-H), 7.45 (1 H, d,
J = 8.9 Hz, indole 7-H), 7.51 (1 H, ddd , J = 8.2, 6.8, 1.0 Hz, 7-
H), 7.60 (1 H, ddd, J = 8.0, 6.8, 1.0 Hz, 8-H), 8.01 (1 H. d, J =
8.3 Hz, 9-H), 8.08 (1 H, d, J = 8.5 Hz, 6-H); 13C NMR (CD3OD)
δ 20.73 (COMe), 28.78 (CMe3), 41.13 (1-C), 45.82 (NMe2),
56.45 (2-C), 59.27 (CH2NMe2), 66.48 (CH2OAc), 67.12
(OCH2CH2NMe2), 81.30 (CMe3), 104.65 (indole 4-C), 107.35
(indole 3-C), 114.01 (indole 7-C), 115.05 (4-C), 117.67 (indole
6-C), 123.93 (9b-C), 124.64 (6-C, or 9-C), 124.71 (9-C or 6-C),
125.87 (7-C), 127.90 (5a-C), 128.21 (8-C), 129.37 (indole 3a-C),
131.54 (9a-C), 132.06 (indole 2-C), 133.70 (indole 7a-C), 135.96
(indole 5-C), 142.52 (indole 3a-C), 154.91 (indole 5-C), 156.70
(Boc C=O), 162.89 (indole C=O), 172.77 (OAc C=O); MS m/z
587.2945 (M + H) (C33H39N4O6 requires 587.2870).
MsCl (7.0 mg, 61 µmol) was stirred with 47 (24 mg, 44 µmol)
in dry pyridine (0.7 mL) under N2 at 0°C for 1 h. LiCl (92 mg,
2.2 mmol) was added and the mixture was stirred at 20°C for 7 d.
EtOAc was added, followed by sat. aq. NaHCO3. The mixture
was extracted with EtOAc. The extract was washed (water,
→
brine). Drying, evaporation and chromatography (EtOAc
EtOAc / MeOH / Et3N 980:20:1) gave 51 (16.3 mg, 66%) as a
yellow solid: mp >250°C; 1H NMR (COSY / NOESY) δ 1.55 (9
H, s, But), 2.37 (6 H, s, NMe2), 3.47 (1 H, t, J = 10.7 Hz, ClCH),
3.96 (1 H, dd, J = 11.2, 2.9 Hz, ClCH), 4.13 (2 H, t, J = 5.8 Hz,
OCH2), 4.14 (1 H, m, 1-H), 4.66 (1 H, t, J = 8.6 Hz, 2-H), 4.82 (1
H, dd, J = 9.0, 1.7 Hz, 2-H), 6.91 (1 H, s, Boc NH), 7.01 (1 H, d,
J = 1.4 Hz, indole 3-H), 7.04 (1 H, dd, J = 8.9, 2.4 Hz, indole 6-
H), 7.14 (1 H, d, J = 2.2 Hz, indole 4-H), 7.35 (1 H, d, J = 8.9
Hz, indole 7-H), 7.45 (1 H, ddd, J = 8.3, 7.0, 1.1 Hz, 7-H), 7.55
(1 H, ddd, J = 7.9, 6.8, 0.9 Hz, 8-H), 7.77 (1 H, d, J = 8.2 Hz, 9-
H), 7.90 (1 H, d, J = 8.4 Hz, 6-H), 8.92 (1 H, s, 4-H), 9.42 (1 H,
s, indole NH); 13C NMR δ 28.28 (CMe3), 43.40 (1-C), 45.84
(NMe2 + CH2Cl), 54.94 (2-C), 58.37 (CH2NMe2), 66.59 (OCH2),
80.92 (CMe3), 103.64 (indole 4-C), 105.90 (indole 3-C), 111.46
(4-C), 112.59 (indole 7-C), 117.32 (indole 6-C), 120.77 (9b-C),
122.37 (6-C), 123.04 (9-C), 124.71 (5a,7-C2), 127.23 (8-C),
128.24 (indole 3a-C), 129.68 (9a-C), 130.49 (indole 2-C), 131.32
(indole 7a-C), 134.75 (5-C), 141.66 (3a-C), 153.28 (Boc C=O),
153.86 (indole 5-C), 160.41 (indole C=O); MS m/z 565.2382 (M
+ H)+ (C31H3637ClN4O4 requires 565.2396), 563.2443 (M + H)+
(C31H3637ClN4O4 requires 563.2425), 427.2157 ((M – (Boc + Cl)
+ H)+ (C26H27N4O2 requires 427.2134).
4.10. 1-Chloromethyl-3-(5-(2-dimethylaminoethoxy)indole-2-
carbonyl)-2,3-dihydrobenzo[e]indole-5-amine dihydrochlor-
ide (52)
4.8. 1-Chloromethyl-3-(5-(2-dimethylaminoethoxy)indole-2-
carbonyl)-5-(N,N-dimethylformamidino)-2,3-dihydrobenzo-
[e]indole (50)
Compound 51 (18.6 mg, 33 µmol) was stirred with HCl in
1,4-dioxane (4.0 M, 2.0 mL) for 2 h. The solvent and excess HCl
were evaporated at 20°C. The residue was triturated with MeCN
Et3N (28 mg, 0.27 mmol) was added to 47 (30 mg, 54 µmol)
in dry DMF (0.4 mL) under N2 at 0°C, followed by MsCl (19 mg,

and Et2O. Drying gave 52 (15.3 mg, quant.) as a pale buff solid:
,
Muhammad, Tsz-Huen Pang and Mujeeb Zia (University of
Bath) for help with the DNA-melting studies.
mp >250°C. A sample in EtOAc was washed (sat. aq. NaHCO3
brine). Drying and evaporation gave the free base as a yellow
1
solid: mp > 250°C; H NMR ((CD3)2SO) (NOESY) δ 2.90 (6 H,
References
s, NMe2), 3.55 (2 H , t, J = 4.2 Hz, CH2NMe2), 3.77 (1 H, dd, J =
11.0, 8.2 Hz, ClCH), 3.99 (1 H, dd, J = 11.1, 3.1 Hz, ClCH), 4.15
(1 H, m, 1-H), 4.33 (2 H, t, J = 4.9 Hz, OCH2), 4.52 (1 H, dd, J =
10.8, 1.5 Hz, 2-H), 4.75 (1 H, t, J = 10.7 Hz, 2-H), 7.00 (1 H, d, J
= 8.8 Hz, indole 6-H), 7.10 (1 H, d, J = 1.7 Hz, indole 3-H), 7.26
(1 H, d, J = 1.9 Hz, indole 4-H), 7.32 (1 H, t, J = 7.9 Hz, 7-H),
7.45 (1 H, d, J = 8.9 Hz, indole 7-H), 7.48 (1 H, t, J = 7.4 Hz, 8-
H), 7.75-7.80 (2 H, m, 4,9-H2), 8.08 (1 H, d, J = 8.5 Hz, 6-H),
11.66 (1 H, s, indole NH); 13C NMR ((CD3)2SO) δ 41.40 (1-C),
42.89 (NMe2), 47.51 (ClCH2), 55.04 (2-C), 55.67 (CH2NMe2),
62.70 (OCH2), 99.50 (4-C), 104.01 (indole 4-C), 105.00 (indole
3-C), 113.30 (indole 7-C), 115.54 (indole 6-C), 117.55 (9b-C),
120.67 (4a-C), 122.46 (7-C), 123.05 (9-C), 123.37 (6-C), 126.91
(8-C), 127.11 (indole 2-C), 130.01 (9a-C), 131.52 (indole 3a-C),
131.95 (indole 7a-C), 142.78 (5-C), 151.97 (indole 5-C), 157.90
(3a-C), 158.17 (C=O); MS m/z 465.1910 (M + H) (C26H2837Cl-
N4O2 requires 465.1871), 463.1874 (M + H) (C26H2835ClN4O2
requires 463.1901).
1. Ichimura, M.; Ogawa, T.; Takahashi, K.; Kobayashi, E.; Kawamoto, I.;
Yasuzawa, T.; Takahashi, I.; Nakano, H. J. Antibiot. 1990, 43, 1037.
2. Li, L. H.; Swenson, D. H.; Schpok, S. L. F.; Kuentzel, S. L.; Dayton, B.
D.; Krueger, W. C. Cancer Res. 1982, 42, 999.
3. Hopton, S. R.; Thompson, A. S. Biochemistry 2011, 50, 4143.
4. Tichenor, M. S.; Boger, D. L. Nat. Prod. Rep. 2008, 25, 220.
5. Igarashi, Y.; Futamata, K.; Fujita, T.; Sekine, A.; Senda, H.; Naoki, H.;
Furumai, T. J. Antibiot. 2003, 56, 107.
6. McGovren, J. P.; Clarke, G. L.; Pratt, E. A.; Dekoning, T. F.. J. Antibiot.
1984, 37, 63.
7. Mohamadi, F.; Spees, M. M.; Staten, G. S.; Marder, P.; Kipka, J. K.;
Johnson, D. A.; Boger, D. L.; Zarrinmayeh, H. J. Med. Chem. 1994, 37,
232.
8. Boger, D. L.; Ishizaki, T.; Sakya, S. M.; Munk, S. A.; Kitos, P. A.; Jin,
Q.; Besterman, J. M. Bioorg. Med. Chem. Lett. 1991, 1, 115.
9. Atwell, G. J.; Tercel, M.; Boyd, M.; Wilson, W. R.; Denny, W. A. J.
Org. Chem. 1998, 63, 9414.
10. Tietze, L. F.; Major, F.; Schuberth, I. Angew. Chem. Int. Ed. 2006, 45,
6574.
11. Tietze, L. F.; Major, F.; Schuberth, I.; Spiegl, D. A.; Krewer, B.;
Maksimenka, K.; Bringmann, G.; Magull, J. Chem. Eur. J. 2007, 13,
4396.
12. Wolfe, A. L.; Duncan, K. K.; Parelkar, N. K.; Brown, D.; Vielhauer, G.
A.; Boger, D. L. J. Med. Chem. 2013, 56, 4104-4115.
4.21. DNA-melting studies
13. Stevenson, R. J.; Denny, W. A.; Tercel, M.; Pruijn, F. B.; Ashoorzadeh,
A. J. Med. Chem. 2012, 55, 2780.
The buffer was prepared by dissolving NaH2PO4 (1.17 g, 9.75
mM), ethylenediaminetetraacetic acid (EDTA) disodium salt
dihydrate (372 mg, 1.0 mM) and NaCl (438 mg, 7.5 mM) in 18.2
MΩ MilliQ water (800 mL). The pH was adjusted to pH 7.0
using aq. NaOH (0.1 M) or aq. HCl (0.1 M) as required and
Milli-Q water was added to 1.00 L. The calf thymus DNA stock
solution was prepared by dissolving DNA in 9.75 mM phosphate
buffer to a concentration of 1.0 mg mL-1. The stock was diluted
with buffer until an absorbance of 0.6 at 256 nm was obtained.
This gave a concentration of ca. 32 µg mL-1, using the Beer-
Lambert Law with ε = 6600 M-1 cm-1. Compound 52 (1.04 mg,
1.94 µmol) was dissolved in DMSO (1.00 mL) to make a stock
solution. Different volumes of this solution were added to the
DNA solution (3.00 mL) in the appropriate quartz cuvettes and
the mixtures were incubated at room temperature for 1.0 h to
allow alkylation of the DNA. Five parallel cuvettes were prep-
ared: A: blank containing buffer and DMSO (40 µL); B,C: con-
trols containing DNA, buffer and DMSO (40 µL); D,E: tests con-
taining 52/53, DNA, buffer and DMSO. All cuvettes contained
the same total volume of DMSO (40 µL). The absorbance at 256
nm was monitored for each sample over the range of 40°C →
95°C with the temperature rising at a rate of 0.5 deg. C min-1,
using a complete PC-controlled spectrometer system (Perkin
Elmer Lambda BioDNA Melt system including PTP-6 temper-
ature-controlled programmer for simultaneous curves, UVTemp-
lab software and in-cuvette temperature sensor). Both direct and
first-derivative graphs were plotted and the Tm was taken as
temperature at the maximum in the first-derivative plot.
14. Sheldrake, H. M.; Travica, S.; Johansson, I.; Loadman, P. M.;
Sutherland, M.; Elsalem, L.; Illingworth, N.; Cresswell, A. J.; Reuillon,
T.; Shnyder, S. D.; Mkrtchian, S.; Searcey, M.; Ingelman-Sundberg, M.;
Patterson, L. H.; Pors, K. J. Med. Chem. 2013, 56, 6273.
15. Hodgson, H. H.; Smith, E. W. J. Chem. Soc. 1935, 671.
16. Twum, E. A.; Woodman, T. J.; Wang, W.; Threadgill, M. D. Org.
Biomol. Chem. 2013, 11, 6208.
17. Tietze, L. F.; Schuster, H. J.; Hampel, S. M.; Rühl, S.; Pfoh, R. Chem.
Eur. J. 2008, 14, 895.
18. Furuyama, T.; Yonehara, M.; Arimoto, S.; Kobayashi, M.; Matsumoto,
Y.; Uchiyama, M. Chem. Eur. J. 2008, 14, 10348.
19. Boger, D. L.; McKie, J. A. J. Org. Chem. 1995, 60¸ 1271.
20. Threadgill, M. D. PhD thesis, University of Cambridge, 1981.
21. Boger, D. L.; Johnson, D. S. Angew. Chem. Int. Ed. Engl. 1996, 35,
1438.
22. Reiter, R. J.; Tan, D. X.; Manchester, L. C.; Korkmaz, A.; Fuentes-
Broto, L.; Hardman, W. E.; Rosales-Corral, S. A.; Qi, W. Cancer Invest.
2013, 31, 365.
23. Paine, H. A.; Nathubhai, A.; Woon, E. C. Y.; Sunderland, P. T.; Wood,
P. J.; Mahon, M. F.; Lloyd, M. D.; Thompson, A. S.; Haikarainen, T.;
Narwal, M.; Lehtiö, L.; Threadgill, M. D. J. Med. Chem., submitted.
4.22. Cytotoxicity assays
The human prostate cancer cell line LNCaP (Sigma-Aldrich)
was cultured in Dulbecco’s Modified Eagle’s Medium (DMEM)
with high glucose (4.5 g L-1), foetal bovine serum (FBS, 20%
(v/v)), penicillin (100 U mL-1) and streptomycin sulfate (100 µg
mL-1) at 37°C in a humidified atmosphere containing 5% CO2
(v/v). The MTS assay was conducted as described previously.23
Acknowledgements
We thank Dr. Timothy J. Woodman (University of Bath) for
help with the NMR spectra and Prostate Cancer UK (Grant
110816) for generous funding. We also thank Malham Al-