Development of novel telomerase inhibitors based on a bisindole unit

Article abstract of DOI:10.1016/S0960-894X(01)00002-6

Telomerase is the enzyme that elongates telomere repeat at the ends of a chromosome. As high telomerase activity is observed in most cancer cells, inhibitors of human telomerase have been expected as new chemotherapeutic agents for cancer. We describe here the discovery of novel inhibitors with IC₅₀ values in the submicromolar range. The structure of the novel inhibitors will be useful as a scaffold for construction of the library in the search for telomerase inhibitors.

Full text of DOI:10.1016/S0960-894X(01)00002-6

Bioorganic& Medicinal Chemistry Letters 11 (2001) 583±585
Development of Novel Telomerase Inhibitors Based on a
Bisindole Unit
Shigeki Sasaki,a,* Takeru Ehara,a Ikuhiro Sakata,a Yasuhiro Fujino,b
Naozumi Harada,b Junko Kimura,b Hideo Nakamurab and Minoru Maedaa
aGraduate School of Pharmaceutical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan
bMitsubishi-Tokyo Pharmaceuticals, Inc., 1000 Kamoshida-cho, Aoba-ku, Yokohama 227-8502, Japan
Received 17 November 2000; accepted 21 December 2000
AbstractÐTelomerase is the enzyme that elongates telomere repeat at the ends of a chromosome. As high telomerase activity is
observed in most cancer cells, inhibitors of human telomerase have been expected as new chemotherapeutic agents for cancer. We
describe here the discovery of novel inhibitors with IC50 values in the submicromolar range. The structure of the novel inhibitors
will be useful as a sca€old for construction of the library in the search for telomerase inhibitors. # 2001 Elsevier Science Ltd. All
rights reserved.
The DNA sequences at the ends of a chromosome
include tandem repeats of the sequence that is specia-
lized by a species such as 50-TTAGGG in humans. In
most human somaticeclls, telomeres shorten progres-
sively by 50±200 nucleotides with each cell division.1
Reduction in the telomere length is believed to lead
ultimately to senescence and growth arrest. In contrast,
around 85±90% of human tumor cells express telomer-
ase, the enzyme that maintains telomere length, and
acquires inde®nite replicative capacity leading to cel-
lular immortality. If the telomere lengths of cancer cells
were signi®cantly short, selective inhibitors of telomer-
ase would stop proliferation of such cells. In this sense,
telomerase is regarded as a speci®c target for develop-
ment of cancer chemotherapeutic agents, although stem
cells have signi®cant telomerase activity.2 As there have
been no clinical trials of inhibitors to date despite
intensive research, discovery of novel inhibitors will
contribute to development of new drugs in advanced
cancer chemotherapy.
been revealed, several strategies have been attempted for
identi®cation of inhibitors of telomerase such as with
high-throughput screening, hypothesis on the G-quartet
structure of telomere and modi®cation of known inhi-
bitors of DNA and RNA polymerase.4 Telomerase
contains RNA as a template for telomere and is a
member of reverse transcriptases; therefore, it seems
reasonable to take known inhibitors of transcriptases as
lead structures for new inhibitors. Previously, we have
identi®ed a novel inhibitor (1)5 of DnaA protein, an
initiation factor of DNA replication at oriC of E. coli,
and this compound was found to have inhibitory activ-
ity to human telomerase in a preliminary assay. This
®nding led us to develop novel inhibitors of telomerase.
In the present study, we wish to report development of a
unique class of telomerase inhibitors.
The 3-acetoxy-2,20-bi-1H-indole derived from indigo
was used as the starting material to obtain a variety of
derivatives (1±13). A preliminary study on the SAR
(Table 1) has shown the following: (1) the carboxylic
acid at the terminus is primarily important (1 vs 2, 11 vs
12); (2) the length of the spacer between the carboxylic
acid and the bisindole skeleton should not be shorter or
longer than around 14 methylene (1, 5±9); (3) 1N-alky-
lation of the indole skeleton is tolerable (11, 13).
Although the bisindole skeleton was proven to be a
useful lead structure for telomerase inhibitors, it turned
out that 1 is labile under acidic condition and easily
su€ers air oxidation. Then, we used a 3-carboxymethyl
indole skeleton (14) as an alternative structure of the 3-
substituted indole unit. As 1N-alkylation was shown to
Antisense oligodeoxynucleotides and related com-
pounds exhibit potent inhibition of telomerase in the
picomolar range.3 Development of a delivery method of
these oligomers will be needed for clinical applications.
Although inhibitors with small molecular weight should
be promising candidates in cancer chemotherapy, at
least the reported ones are relatively less potent and less
selective. As no detailed structure of telomerase has
*Corresponding author. Tel.: +81-92-642-6651; fax: +81-92-642-
6654; e-mail: sasaki@phar.kyushu-u.ac.jp
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00002-6

584
S. Sasaki et al. / Bioorg. Med. Chem. Lett. 11 (2001) 583±585
have no e€ect on inhibitory activity (Table 1, 11 and
13), we designed new compounds connecting two indole
units by 1N-alkylation (Scheme 1).
As the terminal carboxyl group should be in its anionic
form at physiological pH, it was interesting to us to
introduce another anionic group such as a phosphate
group. Accordingly, the ester group of 30 was reduced
to the corresponding hydroxyl compound (31). Phos-
phodiester derivatives (32±38) were derived from 31 by
the conventional method and puri®ed as its TEA salt
(Scheme 3).
3-Carboxymethylindole was coupled with methyl 12-
aminododecanoate to give 14. Meantime, indole (3-Me
or 3-H) was alkylated with TsO-(CH2)2-OTHP or TsO-
(CH2)3-OTHP, and the product was transformed into
the corresponding tosylate (15). Alkylation of 14 was
then performed with 15 to furnish the linked bisindole
derivatives 16. Intramolecular coupling between the
2,20-position of 16 was carried out in tri¯uoroacetic acid
(TFA) and the following DDQ oxidation produced the
corresponding pentacyclic compounds, which were then
hydrolyzed to a€ord the carboxylic derivative 17 or 18.
The compounds without the 2,20-linkage of the bisin-
dole (19±22) were also obtained from 16 by hydrolysis.
Mono-indole compounds having a variety of alkyl
groups at 1N (24±29) were also synthesized for com-
parison with bisindole compounds (Scheme 2).
Inhibitory activity of all the synthesized compounds was
tested by a quantitative stretch PCR assay6 with the use
of telomerase extracted from HCT116 (American Type
Culture Collection) and the results are summarized in
Tables 2 and 3.
It should be noted that bisindole derivatives having an
ethyl or propyl N,N0-linker (17 and 18) exhibited stron-
ger inhibitory activity than the lead compound 1 (Table
2). More surprising was the ®nding that the compounds
(the type B; 19±22) without 2,20-linkage showed almost
equal potency with 17 and 18. The compounds of the
type C having one indole unit also displayed telomerase
inhibition to a similar extent with the type B com-
pounds. Substitution of the indole with pyrrole or pyr-
rolidine (25 and 26) diminished inhibitory activity,
whereas potency was retained with the compounds
having a 1N-naphthyloxyethyl (24), phenoxypropyl (28)
or c-hexyloxypropyl (29) group. These results have sug-
gested that a hydrophobicgroup of suitable volume
may be an essential factor for inhibition potency but not
the nitrogen atom. It has been also clearly shown that a
3-carboxymethylindole skeleton (14) can be a useful
unit for telomerase inhibitors.
Table 1. Inhibition ratio of telomerasea
Compounds
Inhibition (%)
No
R1
R2
1
2
3
NH(CH2)11COOH
NH(CH2)11COOCH3
OH
H
H
H
68
8
9
4
5
6
OCH3
NH(CH2)2COOH
NH(CH2)5COOH
H
H
H
1
0
0
7
8
9
10
11
12
13
NH(CH2)5CONHCH2COOH
NH(CH2)11CONHCH2COOH
NH(CH2)4CONH(CH2)5COOH
NH(CH2)11CONH(CH2)11COOH
NH(CH2)11COOH
H
H
H
H
CH3
CH3
CH2Ph
13
36
56
0
90
0
NH(CH2)11COOCH3
NH(CH2)11COOH
73
aDetermined by the stretch PCR assay with use of 100 mM of the
ligand.
Scheme 2. (a) (1) KOtBu, DMF; (2) TsOHꢀ H2O, MeOH; (3) TsCl,
Et3N, CH2Cl2; (b) (1) nucleophile, KOtBu, DMF; (2) aqueous NaOH,
THF.
Scheme 1. (a) KOtBu, DMF; (b) (1) TFA, CH2Cl2; (2) DDQ, toulene;
(c) aqueous NaOH, MeOH, THF.
Scheme 3. (a) LiAlH4, THF; (b) (1) iPr2NP(Cl)OCH2CH2CN, iPr2-
NEt, CH2Cl2; (2) R2OH, 1H-tetrazole; (3) tBuOOH; (4) Et3N, CHCl3.

S. Sasaki et al. / Bioorg. Med. Chem. Lett. 11 (2001) 583±585
Table 2. Inhibitory activity of indole derivativesa
585
Compound
No.
Type
R
n
Inhibition ratio (%)b
IC50 (mM)
1
17
18
Ð
A
A
Ð
H
CH3
Ð
2
3
68
62
82
>100
62
50
19
20
21
22
B
B
B
B
H
CH3
H
2
2
3
3
77
7
46
100
41
>100
92
CH3
20
24
25
26
27
28
29
C
C
C
C
C
C
1-Naphthyloxy
N-Pyrrolyl
N-Pyrrolidyl
Phenoxy
Phenoxy
c-Hexyloxy
2
3
3
2
3
3
65
0
13
0
75
75
35
>100
>100
Ð
33
67
aDetermined by the stretch PCR assay.
bInhibition with use of 100 mM of the ligand.
Inhibitory activities of the compounds bearing a phos-
phodiester group, the type D and E (32±38), are sum-
marized in Table 3. Non-anionicocmpounds ( 32 and
33) did not show signi®cant e€ect, whereas anionic ones
turned out to be potent inhibitors. In particular, 34 and
36 exhibited the strongest inhibition among the com-
pounds tested in this study. The compounds with a
relatively large hydrophobicgroup; m-chlorophenoxy of
36, phenoxy of 37, c-hexyloxy of 38, are more potent
than 35 with a smaller 2-cyanoethyl unit of phospho-
diester, indicating that the hydrophobic nature of the
phosphodiester terminus is also an important factor for
the inhibitory activity. The important role of the 2-
chlorophenylphosphodiester group was also shown by
the fact that the compound 25 of the type C became a
potent inhibitor with IC50=4.2 mM when it was con-
verted to the corresponding 2-chlorophenylpho-
sphodiester structure. The compounds (34, 36±38)
showed less potency in inhibition of Klenow enzyme
with IC50 values of 79, 70, 61 and 101 mM, respectively,
indicating selective inhibition toward telomerase.
In conclusion, we have developed novel inhibitors for
human telomerase. The most potent inhibitors (34 and
36) are constituted of two indole units, a long alkyl
chain and a phosphodiester group. This novel structure
will be useful as a sca€old to construct a compound
library of telomerase inhibitors. Further study is now
ongoing along this line for the construction of a new
compound library.
Acknowledgements
The authors are grateful for Ms. Kaori Chuujou, Mit-
subishi-Tokyo Pharmaceuticals, Inc., for her technical
assistance.
Table 3. Telomerase inhibition by phosphate derivativesa
References
1. Kipling, D. The Telomere; Oxford University Press:
Oxford, 1995.
2. (a) Recent reviews: Senior, K. Lancet 2000, 355, 2226. (b)
Newbold, R. F. Anti-Cancer Drug Des. 1999, 14, 349. (c) Neidle,
S.; Kelland, L. R. Anti-Cancer Drug Des. 1999, 14, 341. (d)
Lic htsteiner, S. P.; Lebkowski, J. S.; Vasserot, A. P.Ann. N.Y.
Acad. Sci. 1999, 886, 1.
Compounds
Inhibition
(%)b
IC50
(mM)
3. See references cited in refs 2c and 2d.
No Type
R1
R2
4. Recent telomerase inhibitors are summarized, Mergny, J.-
L.; Mailliet, P.; Lavelle, F.; Riou, J.-F.; Laoui, A.; Helene, C.
Anti-Cancer Drug Des. 1999, 14, 327.
5. (a) Sasaki, S.; Mizushima, T.; Hashimoto, T.; Maeda, M.;
Sekimizu, K. Bioorg. Med. Chem. Lett. 1994, 4, 1771. (b)
Mizushima, T.; Sasaki, S.; Ohishi, H.; Kobayashi, M.;
Katayama, T.; Miki, T.; Maeda, M.; Sekimizu, K. J. Biol.
Chem. 1996, 271, 25178.
32
33
34
35
36
37
38
D
E
D
E
E
E
E
OCH2CH2CN
OCH2CH2CN
OHꢀ Et3N
OPhm-Cl
OPhm-Cl
OPhm-Cl
OCH2CH2CN
OPhm-Cl
Oph
20
6
100
41
100
100
100
>100
>100
3.4
>100
2.5
OHꢀ Et3N
OHꢀ Et3N
OHꢀ Et3N
6.7
8.7
OHꢀ Et3N
Oc-hex
aDetermined by the Stretch PCR assay.
bInhibition with use of 100 mM of the ligand.
6. Tatematsu, K.; Nakayama, J.; Danbara, M.; Shionoya, S.;
Sato, H.; Omine, M.; Ishikawa, F. Oncogene 1996, 13, 2265.