β-Carbolines as specific inhibitors of cyclin-dependent kinases

Article abstract of DOI:10.1016/S0960-894X(02)00094-X

Harmine (3), 7-fluoro-1-methyl β-carboline (35) and 1-(5-methyl-imidazol-4-yl) β-carboline (41) were potent and specific inhibitors of cyclin-dependent kinases. The degree of aromaticity of the tricyclic ring and the positioning of substituents are important for inhibitory activity. While most β-carbolines inhibited CDK2 and CDK5 to the same extent, selective inhibition against CDK2 was observed in 1-(2-chlorophenyl)- (12), 1-(2-fluorophenyl)- (15), and 1-(2-chloro-5-nitrophenyl)- (28) β-carbolines.

Full text of DOI:10.1016/S0960-894X(02)00094-X

Bioorganic & Medicinal Chemistry Letters 12 (2002) 1129–1132
ꢀ-Carbolines as Specific Inhibitors of Cyclin-Dependent Kinases
Yongcheng Song,a,b Jian Wang,c Su Fern Teng,b Djohan Kesuma,b Yu Deng,c
Jinao Duan,d Jerry H. Wang,c Robert Zhong Qib and Mui Mui Sima,*
aMedicinal and Combinatorial Chemistry Group, Institute of Molecular and Cell Biology, 30 Medical Drive, Singapore 117609, Singapore
bFunctional Proteomics Group, Institute of Molecular and Cell Biology, 30 Medical Drive, Singapore 117609, Singapore
cBiotechnology Research Institute and Department of Biochemistry, The Hong Kong University of Science and Technology,
Clear Water Bay, Kowloon, Hong Kong
dChina Pharmaceutical University, Nanjing, PR China
Received 31 October 2001; accepted 27 December 2001
Abstract—Harmine (3), 7-fluoro-1-methyl b-carboline (35) and 1-(5-methyl-imidazol-4-yl) b-carboline (41) were potent and specific
inhibitors of cyclin-dependent kinases. The degree of aromaticity of the tricyclic ring and the positioning of substituents are
important for inhibitory activity. While most b-carbolines inhibited CDK2 and CDK5 to the same extent, selective inhibition
against CDK2 was observed in 1-(2-chlorophenyl)- (12), 1-(2-fluorophenyl)- (15), and 1-(2-chloro-5-nitrophenyl)- (28) b-carbolines.
# 2002 Elsevier Science Ltd. All rights reserved.
Cyclin-dependent kinases (CDKs) play essential roles in
the regulation of cell division cycle.1 Much evidence has
shown that CDKs are deregulated in human cancers.
CDK5, a major protein kinase which causes hyperphos-
phorylation of a microtubule-binding protein tau, is
closely linked to Alzheimer disease.2 Several classes of
CDK inhibitors have been reported, including stauro-
sporine, olomoucine, flavopiridol, indirubins, paullones,
and a series of 2,6,9-substituted purines.3 X-ray crystal-
lographic studies showed that these compounds com-
pete with ATP-Mg2+ for the ATP-binding site of
CDKs.
reported to exhibit strong cytotoxicity against a number
of human tumor cell lines.7 However, the underlying
mechanism and the cellular target molecules responsible
for such activity were not identified. Although some
b-carbolines were reported to have DNA intercalating
activity8 and inhibitory activity of topoisomerase,9 the
weak activities indicated that they are unlikely to be
the targets for the potent inhibition of cell growth. To
the best of our knowledge, a few enzymes, including
cytochrome P45010 and monoamine oxidase A,11 have
been found to be potently inhibited by b-carbolines in
nanomolar scales, but none of them is directly involved
in the cell proliferation cycle. Chemical synthesis of
harmine analogues was subsequently carried out to
understand the structure–activity relationships (SARs)
and to further improve the activity of harmine.
In our continuing search for traditional Chinese medi-
cinal herbs-derived anticancer agents, harmine (3), a
b-carboline alkaloid, was first identified as a hit in the
screen towards CDK5 inhibition inhibition (data to be
published elsewhere). b-Carbolines are widely dis-
tributed in many plants and mammals and they exhibit
a wide spectrum of biological activities.4,5 Chemical and
pharmacological studies in China showed that harmine,
originally isolated from the plant Peganum harmala, was
effective in suppressing the growth and proliferation of
tested tumour cells.6 More recently, harmine was
Oxidation and decarboxylation of crude 1,2,3,4-tetra-
hydro-3-carboxyl-b-carboline (synthesized from Pictet
Spengler reaction) with 10% K2Cr2O7 (or SeO2 for
compounds 40–42) gave the desired b-carboline (Fig.
1).12 When tryptamine was employed as the starting
material, the corresponding tetrahydro-b-carboline
failed to undergo dehydrogenation reaction under simi-
lar condition, thus suggesting the importance of 3-car-
boxylic acid in facilitating the aromatization process. In
one exceptional case, tryptamine was refluxed with the
trifluoroacetaldehyde ethyl hemiacetal in 1,4-dioxane
without any catalyst for 5 h to afford the corresponding
*Corresponding author. Tel.: +65-6874-1443; fax: +65-6779-1117;
e-mail: mcbsimmm@imcb.nus.edu.sg
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00094-X

1130
Y. Song et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1129–1132
1-trifluoromethyl - 1,2,3,4 - tetrahydro-b-carboline. The
optimized dehydrogenation was achieved by refluxing
the intermediate with activated MnO2 in benzene for 1 h
to give compound 44.
Firstly, while all compounds contain the core structure
of a tricyclic ring, full aromaticity of the tricyclic ring is
required for good CDK inhibitory activities. Dihydro-
b-carbolines 4 (10%) and 6 (13%) invariably showed
much lower activity than their fully aromatic counter-
parts, b-carbolines 3 (62%) and 5 (69%) respectively.
The tetrahydro b-carboline 7 (12%) was also inactive.
Secondly, substitution at the 6-position reduces the
potency of b-carboline. CDK inhibition was completely
abolished when the methoxy group at the 7-position (3)
was moved to the 6-position (47, 2%). 6-Benzyloxyl
b-carboline (48, 2%) was equally inactive. The same
trend was also observed in 7-methyl- (31, 23%), 6-
methyl- (33, 15%) and 5-methyl-b-carbolines (34, 30%),
as well as in 7-fluoro- (35, 65%), 6-fluoro- (45, 46%)
and 5-fluoro-b-carbolines (46, 55%).
Figure 1. General preparation of b-carbolines.
An amino group was also introduced at the 4-position
of the b-carboline ring.13 1-Propyl-1,2,3,4-tetrahydro-b-
carboline prepared from tryptamine was N-acylated to
give 49a. It was subjected to a DDQ mediated oxidation
in aqueous THF to give the corresponding 4-oxo-sub-
stituted product, which upon further reaction with
hydrazine at 120 ꢀC for 7 h, yielded the fully aromatized
1-propyl-4-amino-b-carboline (49) (Fig. 2). 7-Sub-
stituted b-carboline was prepared by either reacting
harmol with a variety of alkyl halide or aliphatic iso-
Thirdly, among the 1-methyl b-carbolines, compounds
that showed at least comparable inhibition as harmine 3
were those having substituents at the 7-position: harmol
(5, 69%), fluoro- (35, 65%) and N-propyl-aminocarbo-
nyloxy-b-carbolines (58, 61%). Harman 2 and all other
7-substituted compounds (31, 50–57, 59–62) showed
lower inhibitory activities than harmine 3. In addition,
1-methyl group was the optimized substituent among
the 7-fluoro b-carbolines (35–40). On the contrary, there
was no obvious steric or electronic requirement for C-1
substituent (1–2, 8–30, 41–44). The best compound was
the 1-(5-methylimidazol-4-yl) b-carboline (41, 89%).
Although synthesis was further carried out to make a
b-carboline containing 7-fluoro and 1-(5-methylimida-
zol-4-yl) substituents, the resulted compound (40) exhib-
ited lower activity than the parent compounds 26 and 41.
1
cyanates (Fig. 3). H–13C HMQC and HMBC NMR
experiments were also carried out to ensure that the
products were O-alkylated and not N-alkylated.
The structures of b-carbolines (1–62) and their inhibi-
tory activities against CDK5 and CDK2 were summar-
ized in Table 1. The inhibition assay was carried out in
duplicate and evaluated at 50 mMof each test com-
2+ 14ꢁ18
In general, the
pound and 15 mMof ATP-Mg
.
inhibition profiles of b-carbolines towards CDK5 and
CDK2 are similar. This is not surprising, as the proteins
have been known to exhibit similar substrate specificity.
For easy understanding, the SAR discussion shall focus
primarily on the CDK2 inhibition result.
Although most b-carbolines showed similar inhibitory
activities towards CDK5 and CDK2, 1-(2-chloro-
phenyl)- (12), 1-(2-fluorophenyl)- (15), 1-(2-chloro-5-
nitrophenyl)-b-carbolines (28) were found to display
potent and selective inhibition towards CDK2 (>60%).
Finally, the 4-amino b-carboline (49) was also inactive.
Of all active compounds (>60%), compounds 3, 28, 35,
and 41 were selected for IC50 determination (Table 2). All
except compound 28 inhibited CDC2/cyclin B, CDK2/
cyclin A and CDK5/p25nck5a in micromolar range. Con-
sistent with the data in Table 1, compound 28 showed
selective inhibition towards CDK2. These compounds
had negligible inhibition towards other tested kinases,
including cAMP-dependent protein kinase, protein kinase
C, mitogen-activated protein kinase, and the Src-related
protein tyrsine kinases Lck, Lyn and Fyn (>250 mM).
In conclusion, we have identified harmine 3 and com-
pound 41 as potent and specific CDK inhibitors (data to
be published elsewhere). SAR analysis demonstrated
that a complete aromatized tricyclic ring and the posi-
tioning of substituents are important for inhibitory
activity. As CDC2 and CDK2 are known to control the
cell division cycle and CDK5 is involved in the patho-
logical development of Alzheimer’s disease, develop-
ment of specific CDK inhibitors is believed to have
potential pharmaceutical applications.
Figure 2. Synthesis of 1-propyl-4-amino-b-carboline.
Figure 3. Synthesis of 7-substituted b-carbolines.

Y. Song et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1129–1132
Table 1. Inhibition of CDK5/p25nck5a and CDK2 by b-carbolines
1131
Compda
R0
R2
R3
R4
R5
Inhibition of CDK5
(%) at 50 mM
Inhibition of CDK2
(%) at 50 mM
1 (Norharmane)
2 (Harmane)
3 (Harmine)
4
5 (Harmol)
6c (Harmalol)
7
H
Me
Me
Me
Me
Me
–COOH
Pr
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
MeO–
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
35
65
68
7
72
15
9
25
27
30
5
36
17
44
37
5
26
4
2
5
4
37
41
22
16
32
1
19
3
45
37
47
62
10
69
13
12
43
44
50
13
69
32
48
64
6
20
10
5
6
8
31
29
43
19
31
5
MeO–
MeO–
HO–
HO–
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
b (Harmaline)
8
9
iPr
tBu
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
Ph
2-Chlorophenyl
3-Chlorophenyl
4-Chlorophenyl
2-Fluorophenyl
3-Fluorophenyl
4-Fluorophenyl
2-Bromophenyl
3-Methoxyphenyl
4-Methoxyphenyl
2-Cyanophenyl
3-Cyanophenyl
4-Cyanophenyl
2-Nitrophenyl
3-Nitrophenyl
4-Nitrophenyl
2,4-Dimethylphenyl
2-Chloro-5-nitrophenyl
3-Chloro-4-nitrophenyl
2,4-Difluorophenyl
H
H
H
70
4
56
18
36
15
32
M
2-Fluorophenyl
e
H
H
e
e
H
H
H
M
e
H
H
H
Me
H
Me
H
19
13
Me
M
M
Et
Pr
Ph
M
H
e
H
H
H
H
F
F
F
F
F
69
65
H
H
H
H
H
H
H
H
H
e
e
H
H
NH2
H
H
e
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
55
48
40
28
55
94
55
2
54
60
57
55
51
89
30
11
39
2-Fluorophenyl
5-Methylimidazol-4-yl
5-Methylimidazol-4-yl
Imidazol-4-yl
CH2COOCH3
F
H
H
H
H
F
H
H
H
H
EtO–
CF3
M
M
Me
Me
Pr
Me
Me
M
Me
Me
Me
Me
Me
Me
Me
H
45
H
H
H
F
H
H
46
64
46
55
H
H
H
H
H
MeO–
BzO–
H
H
H
0
0
9
41
8
40
13
24
11
59
2
2
2
22
43
4
50
15
23
13
57
1
HOCH2CH2O–
HnBuO
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
NH2COCH2O
CH3COOCH2CH2O–
PhCOOCH2CH2O–
BnOOCCH2O–
(EtO)2CHCH2O–
PrNCOO–
69
64
61
53
c-HexNCOO–
60
M
e
H
H
H
2
10
(continued on next page)

1132
Y. Song et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1129–1132
Table 1 (continued)
Compda
R0
M
R2
R3
R4
R5
Inhibition of CDK5
(%) at 50 mM
Inhibition of CDK2
(%) at 50 mM
61
e
H
H
H
H
H
22
27
8
62
M
e
H
2.5
aCompounds 1–7 were obtained from Aldrich.
b3,4-dihydro-b-carbolines.
c1,2,3,4-tetrahydro-b-carboline.
8. Taira, Z.; Kanzawa, S.; Dohara, C.; Ishida, S.; Matsumoto,
M.; Sakiya, Y. Jpn. J. Toxicol. Environ. Health 1997, 43, 83.
9. Funayama, Y.; Nishio, K.; Wakabayashi, K.; Nagao, M.;
Shimoi, K.; Ohira, T.; Hasegawa, S.; Saijo, N. Mutat. Res.
1996, 349, 183.
Table 2. IC50 (mM) of harmine (3), b-carbolines 28, 35 and 41
Compd
CDC2/cyclin B
CDK2/cyclin A
CDK5/p25nck5a
3
28
35
41
18
45
42
25
—
—
35
22
30
6
—
—
21
>1200
10. Stawowy, P.; Bonnet, R.; Rommelspacher, H. Biochem.
Pharmacol. 1999, 57, 511.
24
5
Olomoucine
Roscovitine
7
0.8
11. (a) Kim, H.; Sablin, S. O.; Ramsay, R. R. Arch. Biochem.
Biophys. 1997, 337, 137. (b) de Arriba, A. F.; Lizcano, J. M.;
Balsa, M. D.; Unzeta, M. J. Pharm. Pharmacol. 1994, 46, 809.
12. Preparation of b-carboline: To a solution of (substituted)
tryptophan (1 mmol) in deionised water (5 mL) and H2SO4 (1
mmol) was added the aldehyde (1.5 mmol). The reaction mixture
was heated at 65ꢀC for 12 h before adding glacial AcOH (1.5
mL) and 10% K2Cr2O7 (4 mL). It was refluxed for 3 min and
then cooled to rt. Saturated Na2SO3 was added to reduce the
excess oxidant and solid Na2CO3 was added to neutralize the
solution. The product was extracted into CH2Cl2 (3ꢂ20 mL) and
dried over Na2SO4. After removal of the solvent, the residue was
subjected to preparative TLC to yield the pure b-carboline.
13. Trudell, M. L.; Fukada, N.; Cook, J. M. J. Org. Chem.
1987, 52, 4293.
14. Protein kinase assays: Radioactive isotopes, [g-32P]ATP
(3000 Ci/mmol, 1 mCi/mL) and [3H]Thymidine (1 mCi/mL),
were purchased from Amersham. Other chemicals were
obtained from Sigma Chemicals. The CDK5/p25nck5a enzyme
was prepared from recombinant GST-Cdk5 and GST-p25nck5a
by reconstitution, thrombin cleavage of the GST moiety, and
then isolation of CDK5/p25nck5a. CDK2/cyclin A was pre-
pared by reconstitution of bacterially expressed CDK2 and
cyclin A and then activation of the complex by phosphoryla-
tion of CDK2 with CAK.15ꢁ17 Cdc2/cyclin B was purchased
from New England BioLabs. The kinase reaction of CDK5/
p25nck5a, CDC2/cyclin B, and CDK2/cyclin A was carried out
as previously described.18
Acknowledgements
This work is supported by Agency for Science, Tech-
nology and Research of Singapore, and Biotechnology
Research Institute, Hong Kong University of Science
and Technology. Dr. R. Z. Qi is the Principle Investi-
gator of the Functional Proteomics Group, Institute of
Molecular and Cell Biology and an Adjunct Staff
Member of the Department of Anatomy, Faculty of
Medicine, National University of Singapore. First and
second authors contributed equally to this work.
References and Notes
1. (a) Morgan, D. O. Annu. Rev. Cell Dev. Biol. 1997, 13, 261.
(b) Stein, G. S.; Baserga, R.; Giordano, A.; Denhardt, D. T.
The Molecular Basis of Cell Cycle and Growth Control; Wiley-
Liss Inc.: New York, 1999.
2. Lew, J.; Wang, J. H. Trends. Biochem. Sci. 1995, 20, 33.
3. (a) See reviews and the references cited therein: Gray, N.;
Detivaud, L.; Doerig, C.; Meijer, L. Curr. Med. Chem. 1999, 6,
859. (b) Senderowicz, A. M.; Sausville, E. A. J. Natl. Cancer
Inst. 2000, 92, 376. (c) Sielecki, T. M.; Boylan, J. F.; Benfield,
P. A.; Trainor, G. L. J. Med. Chem. 2000, 43, 1.
4. Torreilles, J.; Guerin, M. C.; Previero, A. Biochimie 1985,
67, 929.
15. Fesquet, D.; Labbe, J. C.; Derancourt, J.; Capony, J. P.;
Galas, S.; Girard, F.; Lorca, T.; Shuttleworth, J.; Doree, M.;
Cavadore, J. C. EMBO J. 1993, 12, 3111.
16. Poon, R. Y.; Yamashita, K.; Adamczewski, J. P.; Hunt,
T.; Shuttleworth, J. EMBO J. 1993, 12, 3123.
17. Solomon, M. J.; Harper, J. W.; Shuttleworth, J. EMBO J.
1993, 12, 3133.
18. Qi, Z.; Huang, Q.-Q.; Lee, K.-Y.; Lew, J.; Wang, J. H. J.
Biol. Chem. 1995, 270, 10847.
5. de Meester, C. Mutat. Res. 1995, 339, 139.
6. Zhan, T.; Li, C. J. Beijing Med. Univ. 1990, 22, 382.
7. Ishida, J.; Wang, H.-K.; Bastow, K. F.; Hu, C.-Q.; Lee,
K.-H. Bioorg. Med. Chem. Lett. 1999, 9, 3319.