Design and synthesis of two cytotoxic analogs of the novel pyrrolo[1′,2′:1,2][1,4]diazepin [7,6-b]indol-5(6h)-one nucleus

Article abstract of DOI:10.1246/cl.2003.512

The design and synthesis of the two cytotoxic derivatives 15 and 16 of the novel pyrrolo[1′,2′:1,2][1,4]diazepin[7,6-b]indol-5(6H)-one nucleus is described. Readily available methyl 2-indolecarboxylates 5 and 6 are nitrosated with NaNO₂ in AcOH to give the analogs 7 and 8, which are then oxidized with KMnO₄ in aq. NaOH to provide the 3-NO₂ acids 9 and 10. These, in turn, are subjected to amidation with (2S)-pyrrolidine-2-carboxaldehyde diethyl thioacetal in the presence of EDCI and HOBt and then to a 7-exo-trig cyclization reaction to give the target molecules 15 and 16. The new analogs were evaluated in the human leukemic K ₅₆₂ cell line and were shown to have micromolar potency.

Full text of DOI:10.1246/cl.2003.512

512
Chemistry Letters Vol.32, No.6 (2003)
Design and Synthesis of Two Cytotoxic Analogs of the Novel Pyrrolo[10,20:1,2][1,4]diazepin
[7,6-b]indol-5(6H)-one Nucleus
Andrew Tsotinis,ꢀ Margarita Vlachou, Konstantinos Kiakos,y John A. Hartley,yy and David E. Thurstony
University of Athens, School of Pharmacy, Department of Pharmaceutical Chemistry,
Panepistimiopolis-Zografou, GR-15771, Athens, Greece
yCR-UK Gene Targeted Drug Design Research Group, The School of Pharmacy, University of London,
29/39 Brunswick Square, London WC1N 1AX, U.K.
yyCR-UK Drug-DNA Interactions Research Group, Department of Oncology, Royal Free & University College Medical School,
UCL, 91 Riding House Street, London WIP 8BT, U.K.
(Received February 28, 2003; CL-030170)
The design and synthesis of the two cytotoxic derivatives
15 and 16 of the novel pyrrolo[10,20:1,2][1,4]diazepin[7,6-b]in-
dol-5(6H)-one nucleus is described. Readily available methyl 2-
indolecarboxylates 5 and 6 are nitrosated with NaNO2 in AcOH
to give the analogs 7 and 8, which are then oxidized with
KMnO4 in aq. NaOH to provide the 3-NO2 acids 9 and 10.
These, in turn, are subjected to amidation with (2S)-pyrroli-
dine-2-carboxaldehyde diethyl thioacetal in the presence of
EDCI and HOBt and then to a 7-exo-trig cyclization reaction
to give the target molecules 15 and 16. The new analogs were
evaluated in the human leukemic K562 cell line and were shown
to have micromolar potency.
with the appropriate 3-dimensional shape to fit into the DNA
minor groove. Similarly, the aromatic A-ring substituents can
enhance binding of the PBD through the formation of hydrogen
bonds to DNA bases.4;5
It is known that the presence of a hydroxyl group at the C-9
position of the A-ring can lead to cardiotoxicity.6 In an attempt
to circumvent this, various analogs have been synthesized
which bear nitrogen-containing aromatic heterocycles (e.g. 3)
in place of the usual aromatic A-ring. We report here the design
and synthesis of the first indolic PBD analogs 15 and 16
(Scheme 1). The indole nucleus was chosen because it is known
to increase minor groove binding affinity without significantly
affecting sequence selectivity.7 Furthermore, the methoxyl
groups in the indole moiety were considered to be potential
sources of hydrogen bonds. In the structures of the new tetra-
cyclic analogs 15 and 16 the B and C-rings of the parent PBDs
have been retained.
The natural product anthramycin 1 and its naturally occur-
ring (e.g. DC-81, 2) and synthetic pyrrolo[2,1-c][1,4]benzodi-
azepine (PBD) congeners (e.g. 3 and 4)1;2 have in recent years
generated significant interest as the basis for novel types of anti-
tumour agents (Figure 1).3 The PBDs interact within the minor
groove of DNA by forming a covalent bond between their elec-
trophilic C11-position and the exocyclic C2-NH2 moiety of a
guanine residue.4 Most of the known PBDs, including anthra-
mycin, have a tricyclic skeleton which consists of an aromatic
A-ring, a 1,4-diazepin-5-one B-ring bearing a N10-C11 imine
functionality (or the equivalent) and a pyrrolidine C-ring.
Through extensive Structure-Activity Relationship (SAR) stu-
dies, each of these rings is known to play a distinct role in
the interaction with DNA. For example, the (S) configuration
of the C11a position at the B-C junction provides the molecules
The strategy for the synthesis of the target molecules 15 and
16 is shown in Scheme 1. Methyl 5-methoxy-2-indolecarboxy-
late (5) was prepared from commercially available 5-meth-
oxyindole-2-carboxylic acid by reaction with thionyl chloride
NO
CH3O
R
CH3O
R
a
b
CO2CH3
CO2CH3
N
H
N
H
5: R = H
6: R = OCH3
7: R = H (62%)
8: R = OCH3 (67%)
CH(SCH2CH3)2
H
NO2
NO
2
CH3O
CH3O
R
N
d
OH
c
CO2H
OCH3
H
H
N
10
9
N
H
N
N
H
H3C
HO
R
8
O
10
11
11
H
1
A
B
5
9: R = H (68%)
11: R = H (61%)
11a
C
11a
7
10: R = OCH3 (75%)
12: R = OCH3 (68%)
N
N
CH3O
6
4
2
CONH2
3
O
O
CH(SCH2CH3)2
H
H
NH2
N
1: Anthramycin
2: DC-81
CH3O
CH3O
R
N
e
N
N
H
N
R
OCH3
O
O
H
OCH3
H
N
N
10
N
13: R = H (79%)
14: R = OCH3 (82%)
15: R = H (60%)
16: R = OCH3 (65%)
10
11
11
H
N
H
11a
11a
Scheme 1. Reagents and conditions: (a) NaNO2, AcOH, 2 h;
(b) KMnO4, aq. NaOH (2M), 30 min; (c) (2S)-pyrrolidine-2-
carboxaldehyde diethyl thioacetal, EDCI, HOBt, Et3N,
N
N
CH3O
N
O
O
4
3
CH2Cl2-DMF(2:1), 18 h; (d) 10% Pd-C, H (50 atm), MeOH,
2
Figure 1.
3.5 h; (e) HgCl2, CaCO3, CH3CN-H2O (4:1), 8 h.
Copyright ꢀ 2003 The Chemical Society of Japan

Chemistry Letters Vol.32, No.6 (2003)
513
in methanol. Its bis-methoxy counterpart 6 was made from 3,4-
dimethoxybenzaldehyde by the two-step method reported by
Bunker et al.8 As direct nitration of the carboxylates 5 and 6
did not lead to the corresponding 3-NO2 indoles, 5 and 6 had
to be first nitrosated with NaNO2 in AcOH to give the analogs
7 and 8 followed by oxidation with KMnO4 in aq. NaOH to pro-
vide the desired 3-NO2 acids 9 and 10. Implementation of clas-
sical methods for the formation of the amides 11 and 12, such as
initial reaction of 9 and 10 with SOCl2 or (COCl)2 followed by
rings (e.g. the IC50 value for DC-81 2 in the K562 cell line after
1 h exposure is 3.0 mM), this suggests that they may be exerting
their effect either through nonselective interaction with DNA or
through an as yet unidentified non-DNA-interactive mechanism.
The difference in activity between analogues 15 and 16 suggest
that a full SAR study should be feasible and the synthesis of
further analogues is underway in order to study stability and
biological activity in this series.
coupling
to
(2S)-pyrrolidine-2-carboxaldehyde
diethyl
Ms. Vlachou and Dr. Tsotinis would like to thank the Uni-
versity of Athens (ELKE Account KA: 70/4/5879) for financial
support. Professor David Thurston would like to thank Cancer
Research UK for financial support (C180/A1060).
thioacetal9 or direct amidation of 9 and 10 with the thioacetal
in the presence of 1,10-carbonyldiimidazole (CDI), were not
successful. Their synthesis was eventually effected by a modi-
fied literature procedure using 1-[3-(dimethylamino)propyl]-3-
ethylcarbodiimide hydrochloride (EDCI) and 1-hydroxybenzo-
triazole hydrate (HOBt) in a mixture of CH2Cl2-DMF.10 The
C-3 NO2 groups of 11 and 12 were catalytically hydrogenated
to the corresponding amines 13 and 14, which then underwent a
7-exo-trig cyclization reaction to the target molecules 15 and 16
using the method described by Thurston and co-workers.9 The
overall yield from methyl 5-methoxy-2-indolecarboxylate (5) to
the new derivative 15 was 12%, while that of its congener 1611
from the ester 6 was 18%.
References and Notes
1
D. E. Thurston, in ‘‘Topics in Molecular and Structural
Biology: Molecular Aspects of Anticancer Drug-DNA Inter-
actions,’’ ed. by S. Neidle and M. J. Waring, The Macmillan
Press Ltd, London (1993), pp 54–88.
2
3
A. Kamal, M. V. Rao, N. Laxman, G. Ramesh, and G. S. K.
Reddy, Curr. Med. Chem.: Anti-Cancer Agents, 2, 215
(2002).
S. J. Gregson, P. W. Howard, J. A. Hartley, N. A. Brooks, L.
J. Adams, T. C. Jenkins, L. R. Kelland, and D. E. Thurston,
J. Med. Chem., 44, 737 (2001).
The new analogs 15 and 16 were evaluated in the human
leukemic K562 cell line and were shown to have micromolar po-
tency (Table 1). The three values for each compound appearing
in Table 1 represent evaluations on consecutive weeks using the
same stock solution. The decreasing activity suggests that the
new analogs are not as stable when stored in solution as parent
PBD molecules such as 1–4. A DNA footprinting experiment
involving incubation of 15 and 16 with plasmid DNA at concen-
trations of up to 100 mM for 5 h and using anthramycin 1 as a
control indicated that 15 and 16 showed no evidence of selec-
tive interaction with DNA. Given that the molecules are signif-
icantly less cytotoxic than the equiv. PBDs with benzenoid A-
4
5
X.-L. Yang and A. H.-J. Wang, Pharmacol. Ther., 83, 181
(1999).
D. E. Thurston, D. S. Bose, P. W. Howard, T. C. Jenkins, A.
Leoni, P. G. Baraldi, A. Guiotto, B. Cacciari, L. R. Kelland,
M.-P. Foloppe, and S. Rault, J. Med. Chem., 42, 1951
(1999).
6
7
P. G. Baraldi, B. Cacciari, A. Guitto, R. Romagnolin, A. N.
Zaid, and G. Spalluto, Il Farmaco, 54, 15 (1999).
Q. Zhou, W. Duan, D. Simmons, Y. Shayo, M. A.
Raymond, R. T. Dorr, and L. H. Hurley, J. Am. Chem.
Soc., 123, 4865 (2001).
8
9
A. M. Bunker, J. J. Edmunds, K. A. Berryman, D. M.
Walker, M. A. Flynn, K. M. Welch, and A. M. Doherty,
Bioorg. Med. Chem. Lett., 6, 1061 (1996).
Table 1. Cytotoxicity of compounds 15 and 16 in the K562 cell
linea
IC50 (ꢁM)b
D. R. Langley and D. E. Thurston, J. Org. Chem., 52, 91
(1987).
10 G. Kokotos, R. Verger, and A. Chiou, Chem.—Eur. J., 6,
15
16
18c
47
30c
38
4211 (2000).
11 Selected data for 16: H NMR (400 MHz, CDCl3): ꢀ 2.03–
1
50
58
2.14 (m, 2H, 2 Â H1), 2.32–2.45 (m, 2H, 2 Â H2), 3.66–
3.84 (m, 3H, 2 Â H3 + H12a), 3.90 (s, 3H, OCH3), 3.92
(s, 3H, OCH3), 6.86 (s, 1H, H7(10)), 7.22 (s, 1H, H10(7)),
7.42 (d, 1H, J ¼ 3:9 Hz, H12), 10.40 (bs, 1H, NH); 13C
NMR (50 MHz, CDCl3): ꢀ 24.4, 30.1, 46.1, 55.7, 56.0,
94.3, 99.8, 117.5, 121.6, 129.9, 130.6, 146.2, 150.0, 155.4,
aK562 is a human leukemia cell line in which IC50 values
were measured using a microculture tetrazolium assay (3-
(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bro-
ꢁ
mide, MTT) following a 1 h exposure to drug at 37 C.
bDose of drug required to inhibit cell growth by 50% com-
pared to drug-free control.
20
160.1; [a]D + 440.6ꢁ (c 0.014, CHCl3). Anal. Calcd. for
cThe three values for each compound represent evaluations
on consecutive weeks using the same stock solutions stored
C16H17N3O3: C, 64.20; H, 5.72; N, 14.04%. Found: C,
64.11; H, 5.75; N, 13.96%.
ꢁ
at ꢂ20 C in DMSO.
Published on the web (Advance View) May 13, 2003; DOI 10.1246/cl.2003.512