Design, synthesis, evaluation and QSAR analysis of N1-substituted norcymserine derivatives as selective butyrylcholinesterase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2010.01.057

We synthesized a series of N¹-substituted norcymserine derivatives 7a-p and evaluated their anti-cholinesterase activities. In vitro evaluation showed that the pyridinylethyl derivatives 7m-o and the piperidinylethyl derivative 7p improved the anti-butyrylcholinesterase activity by approximately threefold compared to N¹-phenethylnorcymserine (PEC, 2). A quantitative structure-activity relationship (QSAR) study indicated that logS might be a key feature of the improved compounds.

Full text of DOI:10.1016/j.bmcl.2010.01.057

Bioorganic & Medicinal Chemistry Letters 20 (2010) 1718–1720
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis, evaluation and QSAR analysis of N1-substituted
norcymserine derivatives as selective butyrylcholinesterase inhibitors
Jun Takahashi a, Ichiro Hijikuro e, Takeshi Kihara a, Modachur G. Murugesh e, Shinichiro Fuse d,
Ryo Kunimoto c, Yoshinori Tsumura b, Akinori Akaike b, Tetsuhiro Niidome a, Yasushi Okuno c,
Takashi Takahashi d, Hachiro Sugimoto a,
*
a Department of Neuroscience for Drug Discovery, Graduate School of Pharmaceutical Sciences, Kyoto University, 46-29 Yoshida-Shimoadachi-cho, Sakyo-ku, Kyoto 606-8501, Japan
b Department of Pharmacology, Graduate School of Pharmaceutical Sciences, Kyoto University, 46-29 Yoshida-Shimoadachi-cho, Sakyo-ku, Kyoto 606-8501, Japan
c Department of Systems Bioscience for Drug Discovery, Graduate School of Pharmaceutical Sciences, Kyoto University, 46-29 Yoshida-Shimoadachi-cho, Sakyo-ku,
Kyoto 606-8501, Japan
d Department of Applied Chemistry, Graduate School of Science and Engineering, Tokyo Institute of Technology, 2-12-1 Ookayama, Meguro, Tokyo 152-8552, Japan
e Department of Development, ChemGenesis Incorporated, 4-10-1 Nihonbashi-Honcho, Chuo-ku, Tokyo 103-0023, Japan
a r t i c l e i n f o
a b s t r a c t
We synthesized a series of N1-substituted norcymserine derivatives 7a–p and evaluated their anti-cho-
linesterase activities. In vitro evaluation showed that the pyridinylethyl derivatives 7m–o and the pipe-
ridinylethyl derivative 7p improved the anti-butyrylcholinesterase activity by approximately threefold
compared to N1-phenethylnorcymserine (PEC, 2). A quantitative structure-activity relationship (QSAR)
study indicated that logS might be a key feature of the improved compounds.
Article history:
Received 26 November 2009
Revised 7 January 2010
Accepted 8 January 2010
Available online 20 January 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
Acetylcolinesterase
Butyrylcholinesterase
Alzheimer’s disease
Carbamate
Alzheimer’s disease (AD) is one of the primary neurodegenera-
tive diseases that develops in the elderly, and its prevalence is ex-
pected to increase substantially in the future.1 While AD is
definitively characterized by its pathological hallmarks, such as se-
nile plaques and neurofibrillary tangles, in the postmortem brain,
AD is diagnosed by the progressive cognitive impairment observed
in living patients. Since the decline of cholinergic neurotransmis-
sion is responsible for cognitive dysfunction, considerable effort
has been devoted to developing effective cholinesterase inhibitors
(ChEIs).
Acetylcholinesterase (AChE) is the main cholinesterase (ChE) in
the brain, and hence most works aimed at developing ChEIs have
targeted AChE. To date, some AChE inhibitors have been clinically
approved, and use of these inhibitors has been successful in the
treatment of AD. However, in the AD brain, cholinergic neurons
are lost along with the progression of the disease and AChE is also
reduced.2
creased in the AD brain.2 Interestingly, it has been reported that
selective BuChE inhibitors (BuChEIs) increase the brain ACh,4 and
BuChE knockout mice and silent mutants in humans showed no
physiological disadvantage,5,6 inspiring the hypothesis that BuChE
may be a promising target for developing anti-AD drugs with no
adverse effect.
Physostigmine (1), an alkaloid isolated from the Calabar bean,
the seed of Physostigma venenosum, is the most classical ChEI,
and still used in the treatment of glaucoma. Although many mech-
anistic studies have been carried out, clarification of three-dimen-
sional structures of AChE7 and BuChE8 have enabled us to
investigate the detailed enzyme-inhibitor interaction. For instance,
regarding the N1-position of the physostigmine derivatives, the
bulkiness of substituents at this position affects the enantioselec-
tivity at 3a, and the phenethyl group renders the compounds more
selective for BuChE than their corresponding N1-H and N1-Me ana-
logs.9–11 It has also been reported that N1-phenethylnorcymserine
(PEC, 2) exhibits the property of a potent and selective BuChEI.12
Although these findings were intriguing, these studies did not
investigate a large variety of substituents and, overall, they were
limited to relatively simple compounds. Hence, previous report
studying the quantitative structure–activity relationship (QSAR)
has pointed out that large polar and large hydrophobic substitu-
ents would have to be synthesized and tested.13 An investigation
Vertebrates have another acetylcholine (ACh) hydrolase known
as butyrylcholinesterase (BuChE). In contrast to AChE, BuChE is
mainly derived from glia in the brain.3 BuChE is therefore not re-
duced, and its contribution in degrading ACh is thought to be in-
* Corresponding author. Tel.: +81 75 753 9270; fax: +81 75 753 9269.
E-mail address: hsugimot@pharm.kyoto-u.ac.jp (H. Sugimoto).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.01.057

J. Takahashi et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1718–1720
1719
using more diverse derivatives may increase our understanding
and help improve the development of better compounds. In light
of this, we investigated the structure-activity relationship of the
N1 moiety of norcymserine derivatives 1, 2, 3 and 7a–p in this re-
port. Furthermore, for a detailed understanding, we carried out the
QSAR study. For all compounds, the carbamate moiety was fixed in
the shape of 4-isopropylphenylcarbamate, because this structural
arrangement increases the potency and the selectivity for BuChE.12
PEC (2) and cymserine (3) were prepared according to proce-
dures described previously.9,14 A series of N1-substituted norcym-
serine derivatives 7a–p were synthesized via intermediates 5a–p
as indicated in Scheme 1. Ammonium iodide (4), prepared from
physostigmine (1) using four steps,14 was treated with different
R–NH2 compounds (b, d–l and o, Scheme 1),9,11,15 followed by
demethylation with BBr39–12,16 to generate N1-substituted noreser-
olines (5b, 5d–l and 5o). Alternatively, the intermediate 5 was pre-
pared using only one step in a convenient one-pot procedure.17
Physostigmine (1) was reacted with KOH and MeI in DMSO at room
temperature for 1 h, before the addition of different R–NH2 com-
pounds (a, c, m, n and p, Scheme 1). After the reaction mixtures
were stirred at 120 °C for 3 h, N1-substituted noreserolines 5a,
5c, 5m, 5n and 5p were obtained at 10–21% yield, accompanied
by the corresponding methyl ethers 6a, 6c, 6m, 6n and 6p at
0–10% yield, respectively. Finally, the N1-substituted norcymserine
derivatives 7a–p were prepared from 5a–p using previously re-
ported methods.9–12,16
Anti-ChE activities were measured using a modified Ellman’s
colorimetric method.18 Briefly, we used the extracts from mice
brain and mice serum as sources of AChE and BuChE, respectively,
and all compounds were preincubated with enzymes for 1 h at
37 °C before adding acetylthiocholine or butyrylthiocholine in
combination with the coloring reagent 5,50-dithiobis-(2-nitroben-
zoic acid).
The IC50 values for AChE and BuChE are summarized in Table 1.
We first assessed whether the length of the alkyl chain had an
influence on anti-BuChE activity, and we found that the activity
did not differ between the N1-benzyl, -phenetyl, -phenylpropyl
and -phenylbutyl derivatives (2 and 7a–c). Since the length of
the alkyl chain did not influence the activity, we subsequently
investigated the effects of various substituents on the phenethyl
group. Neither electron-donating nor electron-withdrawing group
at the 4-position of the phenyl group influenced the activity (7d–f).
On the other hand, bulky substituent at the same position was
shown to reduce the activity (7g). Disubstituents at the 2 and 4-po-
sition of the phenyl group also lowered the potency of the com-
pounds (7h and 7i). The increased bulkiness of the alkyl chain of
the phenethyl group also decreased the anti-BuChE activity (7j–
l). Most physostigmine derivatives reported previously did not in-
clude a heteroatom in its N1-substituents. Therefore, we also exam-
ined the effects of the substitution of heterocycles for the phenyl
ring of the phenethyl group. Substitution of pyridine for the phenyl
group increased the anti-BuChE activity by approximately three-
fold regardless of the position of the nitrogen on the pyridine,
and the 1-piperidinylethyl group substitution exhibited a similar
activity (7m–p).
The QSAR study probing the N1-position of physostigmine
derivatives has ever been conducted using a small dataset, and
was used to evaluate for the anti-AChE activity.19 Therefore, in or-
der to rationalize the observed anti-BuChE activities, we carried
out the QSAR study.
For the QSAR study dataset, the structures of all compounds
excluding physostigmine, and their corresponding pIC50 values
for BuChE were used. The chemical structures were drawn using
CHEMDRAW Std 7.0 (Chembridge Soft. com) and the database was ex-
ported as an sdf file and imported into the MOE software (version
2008.10, Chemical Computing Group). Notably, the calculations
were performed using the appropriately ionized form. The QSAR
model generation was done using AutoQSAR packaged in MOE,
and the QSAR models were constructed based on the partial least
square method using 327 descriptors built in MOE. The equation
we obtained was shown below:
pIC50 ¼ 0:452ð0:044ÞlogS ꢀ 0:051ð0:009ÞweinerPol
þ 0:010ð0:002ÞPEOE VSA-0
þ 0:327ð0:049Þopr violation þ 10:556ð0:413Þ
n ¼ 18; RMSE ¼ 0:080; R2 ¼ 0:980;
cross-validated RMSE ¼ 0:187;
Q2 ¼ 0:909; S ¼ 0:094; F4;13 ¼ 160:2
In the QSAR equation, logS is the log of the aqueous solubility,
weinerPol is the Wiener polarity number,20 PEOE_VSA-0 is the sum
of the van der Waals surface area of atoms where the atomic par-
tial charges are ranging from ꢀ0,05 to 0, opr_violation is the num-
ber of violations of Oprea’s lead-like test,21 n is the number of
compounds, RMSE is the root mean square error, R2 is the goodness
1) R-NH2, CH3CN
H
N+
4 steps 15)
HO
N
O
I-
CH3O
110 , 5 h
N
N
R
R
OH
O
2) BBr3, CH2Cl2, rt
N
N
N
Physostigmine (1)
4
5a-p
KOH, CH3I, DMSO, rt, 1h
then, R-NH2, 120 , 3h
CH3O
R-NH2:
N
benzylamine (a), phenylpropylamine (b), phenylbutylamine (c), 4-methylphenethylamine (d),
4-fluorophenethylamine (e), 4-chlorophenethylamine (f), 4-isopropylphenethylamine (g),
2,4-dimethylphenethylamine (h), 2,4-dichlorophenethylamine (i), 2-phenylpropylamine (j),
2,2-diphenylethylamine (k), 1,2-diphenylethylamine (l), 2-(2-pyridinyl)ethylamine (m),
2-(3-pyridinyl)ethylamine (n), 2-(4-pyridinyl)ethylamine (o), 2-(1-piperidinyl)ethylamine (p).
N
6a, c, m, n, p
H
N
4-isopropylphenyl isocyanate
Na, Et2O, rt, 5 min
O
5a-p
N
R
O
N
7a-p
Scheme 1.

1720
J. Takahashi et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1718–1720
Table 1
descriptor was negatively correlated with the anti-BuChE activity.
Anti-ChE activity and enzyme selectivity
Since this value is defined as the number of carbon atoms pairs that
are separated by three carbon–carbon bonds, its value is generally
larger in the branched alkane than in the linear alkane. Namely,
this descriptor indicates that branched or spatially bulky substitu-
ents might be unfavorable.
In summary, by synthesizing various norcymserine derivatives
substituted on the N1 moiety, and by performing the QSAR study,
we found that pyridinylethyl and piperidinylethyl group substitu-
tion increased the anti-BuChE activity, and that the ionizable nitro-
gen of the substituent might contribute to this improvement.
Compounds
R
AChEa
(nM)
BuChEa Selectivityb
(nM)
1
Methyl (methylcarbamate, 40
physostigmine)
Benzyl
Phenethyl
Phenylpropyl
280
0.14
7a
2
>100,000 510
>100,000 540
>100,000 510
>100,000 550
>100,000 730
>100,000 500
>100,000 500
>196
>185
>196
>182
>137
>200
>200
35
7b
7c
7d
7e
7f
Phenylbutyl
4-Methylphenethyl
4-Fluorophenethyl
4-Chlorophenethyl
Methyl
3
8400
240
Acknowledgment
7g
7h
7i
7j
7k
7l
7m
7n
7o
7p
4-Isopropylphenethyl
2,4-Dimethylphenethyl
2,4-Dichlorophenethyl
2-Phenylpropyl
2,2-Diphenylethyl
1,2-Diphenylethyl
2-(2-Pyridinyl)ethyl
2-(3-Pyridinyl)ethyl
2-(4-Pyridinyl)ethyl
2-(1-Piperidinyl)ethyl
>100,000 3300
>100,000 1100
>100,000 2600
>100,000 720
>100,000 10,000
>100,000 15,000
>30
>91
>38
>139
>10
>7
26
>556
135
21
This work was financially supported by Eisai Co., Ltd.
References and notes
4200
160
>100,000 180
1. Ferri, C.; Prince, M.; Brayne, C.; Brodaty, H.; Fratiglioni, L.; Ganguli, M.; Hall, K.;
Hasegawa, K.; Hendrie, H.; Huang, Y.; Jorm, A.; Mathers, C.; Menezes, P.;
Rimmer, E.; Scazufca, M. Lancet 2005, 366, 2112.
23,000
3500
170
170
2. Giacobini, E. Int. J. Geriatr. Psychiatry 2003, 18, S1.
3. Wright, C.; Geula, C.; Mesulam, M. Ann. Neurol. 1993, 34, 373.
4. Greig, N.; Utsuki, T.; Ingram, D.; Wang, Y.; Pepeu, G.; Scali, C.; Yu, Q.; Mamczarz,
J.; Holloway, H.; Giordano, T.; Chen, D.; Furukawa, K.; Sambamurti, K.; Brossi,
A.; Lahiri, D. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 17213.
5. Li, B.; Duysen, E.; Carlson, M.; Lockridge, O. J. Pharmacol. Exp. Ther. 2008, 324,
1146.
6. Primo-Parmo, S.; Bartels, C.; Wiersema, B.; van der Spek, A.; Innis, J.; La Du, B.
Am. J. Hum. Genet. 1996, 58, 52.
7. Sussman, J.; Harel, M.; Frolow, F.; Oefner, C.; Goldman, A.; Toker, L.; Silman, I.
Science 1991, 253, 872.
a
In order to estimate the enzyme activity, changes in the absorbance at 2 min
intervals were measured at 415 nm using spectrophotometer. The enzyme
a
activity is expressed as a percent of the activity of the solvent, DMSO. Seven dif-
ferent concentrations of each compound were used and the IC50 values of each
compound were calculated by nonlinear regression of the sigmoidal dose-response
curve using GraphPad Prism version 4.03. Only the results with correlation coeffi-
cients of r2 P 0.95 were accepted. The results are represented as the mean of the
IC50 obtained from at least four independent measurements.
b
The selectivity was calculated as follows; selectivity = IC50 AChE/IC50 BuChE.
8. Nicolet, Y.; Lockridge, O.; Masson, P.; Fontecilla-Camps, J.; Nachon, F. J. Biol.
Chem. 2003, 278, 41141.
of fit, Q2 is the cross-validated R2, S is the standard error, and F is
the ratio of the variance of the calculated values to that of the ob-
served values. Among the descriptors, logS was the relatively most
important factor and was positively correlated to the anti-BuChE
activity. This descriptor might be greatly influenced by the result
that the four most active compounds had an ionizable nitrogen
in their N1-substituents. Interestingly, a report previously de-
scribed that less hydrophobic substituents are favorable,13 consis-
tent with our result, although the former report focused on the
anti-AChE activity. Since it is considered that a part of the tricyclic
moiety itself interacts with Trp82 in the choline binding site (or
anionic site), it is not known how the ionizable nitrogen of pyridine
or piperidine interacts with BuChE, but it appears to be beneficial
to have the polar atom in the N1 substituent. Moreover, weinerPol
9. Pei, X.; Greig, N.; Bi, S.; Brossi, A. Med. Chem. Res. 1995, 5, 455.
10. Yu, Q.; Pei, X.; Holloway, H.; Greig, N.; Brossi, A. J. Med. Chem. 1997, 40, 2895.
11. Yu, Q.; Greig, N.; Holloway, H.; Brossi, A. J. Med. Chem. 1998, 41, 2371.
12. Yu, Q.; Holloway, H.; Utsuki, T.; Brossi, A.; Greig, N. J. Med. Chem. 1999, 42,
1855.
13. Kaur, J.; Zhang, M. Curr. Med. Chem. 2000, 7, 273.
14. Brzostowska, M.; He, X.; Greig, N.; Rapoport, S.; Brossi, A. Med. Chem. Res. 1992,
2, 238.
15. Pei, X.; Greig, N.; Flippen-Anderson, J.; Bi, S.; Brossi, A. Helv. Chim. Acta 1994, 77,
1412.
16. Yu, Q.; Brossi, A. Heterocycles 1988, 27, 745.
17. Yu, Q.; Luo, W.; Holloway, H.; Utsuki, T.; Perry, T.; Lahiri, D.; Greig, N.; Brossi, A.
Heterocycles 2003, 61, 529.
18. Ellman, G.; Courtnry, K.; Andres, V. J.; Feather-Stone, R. Biochem. Pharmacol.
1961, 7, 88.
19. Recanatini, M.; Cavalli, A.; Hansch, C. Chem. Biol. Interact. 1997, 105, 199.
20. Wiener, H. J. Am. Chem. Soc. 1947, 69, 17.
21. Oprea, T. J. Comput. Aided Mol. Des. 2000, 14, 251.