Methyl analogues of the experimental Alzheimer drug phenserine: Synthesis and structure/activity relationships for acetyl- and butyrylcholinesterase inhibitory action

Article abstract of DOI:10.1021/jm010080x

With the goal of developing potential Alzheimer's pharmacotherapeutics, we have synthesized a series of novel analogues of the potent anticholinesterases phenserine (2) and physostigmine (1). These derivatives contain methyl (3, 4, 6), dimethyl (5, 7, 8, 10, 11) and trimethyl (14) substituents in each position of the phenyl group of the phenylcarbamoyl moieties, and with N-methyl and 6-methyl substituents (12, 13, 31, 33). We also quantified the inhibitory action of these compounds against human acetylcholinesterase (AChE) and butyrylcholinesterase (BChE). An analysis of the structure/anticholinesterase activity relationship of the described compounds, together with molecular modeling, confirmed the catalytic triad mechanism of the binding of this class of carabamate analogues within AChE and BChE and defined structural requirements for their differential inhibition.

Full text of DOI:10.1021/jm010080x

4062
J . Med. Chem. 2001, 44, 4062-4071
Meth yl An a logu es of th e Exp er im en ta l Alzh eim er Dr u g P h en ser in e: Syn th esis
a n d Str u ctu r e/Activity Rela tion sh ip s for Acetyl- a n d Bu tyr ylch olin ester a se
In h ibitor y Action
Qian-sheng Yu,† Harold W. Holloway,† J udith L. Flippen-Anderson,‡ Brian Hoffman,§ Arnold Brossi,# and
Nigel H. Greig*,†
Drug Design & Development Section, Laboratory of Neurosciences, Gerontology Research Center (4E02),
National Institute on Aging, National Institutes of Health, 5600 Nathan Shock Drive, Baltimore, Maryland 21224-6825,
Laboratory for the Structure of Matter, Naval Research Laboratory, Department of the Navy, Washington, D.C. 20375,
Medicinal Chemistry Section, National Institute on Drug Abuse, Baltimore, Maryland 21224-6825, and
School of Pharmacy, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599-7361
Received February 21, 2001
With the goal of developing potential Alzheimer’s pharmacotherapeutics, we have synthesized
a series of novel analogues of the potent anticholinesterases phenserine (2) and physostigmine
(1). These derivatives contain methyl (3, 4, 6), dimethyl (5, 7, 8, 10, 11) and trimethyl (14)
substituents in each position of the phenyl group of the phenylcarbamoyl moieties, and with
N-methyl and 6-methyl substituents (12, 13, 31, 33). We also quantified the inhibitory action
of these compounds against human acetylcholinesterase (AChE) and butyrylcholinesterase
(BChE). An analysis of the structure/anticholinesterase activity relationship of the described
compounds, together with molecular modeling, confirmed the catalytic triad mechanism of the
binding of this class of carabamate analogues within AChE and BChE and defined structural
requirements for their differential inhibition.
In tr od u ction
presently available that defines the relationship be-
tween the chemical structure of compounds and their
ChE inhibitory action.11,12 During recent years, detailed
X-ray diffraction studies of the two ChE enzymes,
acetyl- and butyrylcholinesterase (AChE: EC 3.1.1.7
and BChE: EC 3.1.1.8; respectively), as well as of
inhibitors, together with extensive biochemical research
has elucidated the three-dimensional fit of different
types of inhibitors within AChE and BChE.13-21 This
has provided us the opportunity to study the structure/
activity relationship of anti-ChEs in a new manner to
develop even more selective and potent inhibitors.
Phenserine (2), the unsubstituted phenylcarbamate
analogue of the natural product and classical anti-ChE,
physostigmine (1) (Figure 1),9,22-27 is a potent and
selective inhibitor of AChE (70-fold versus BChE ac-
tion). Phenserine (2) proved to be well tolerated in phase
I clinical trials for the treatment of Alzheimer’s disease
and is currently in phase II efficacy trials. We have
previously demonstrated9 that substitution into the
phenyl ring of its carbamate modifies the selectivity of
the resulting compounds, with 2′-substitution providing
increased AChE selectivity and 4′-substitution negating
this to provide moderate BChE selectivity. There are,
however, only a few reports in the literature in which a
systematic approach has been taken to define the
structure/activity relationship of each important sub-
stitution on a compound that bears the potential to
inhibit two structurally similar but separate critical
enzymes, such as AChE and BChE. We, herein, report
such a study.
Alzheimer’s disease (AD) is the most common neuro-
degenerative disorder that leads to dementia and ac-
counts for up to two-thirds of all such cases. It is, addi-
tionally, the fourth leading cause of death in the United
States of America and afflicts some 4 million adults.
Although several neurotransmitter systems are affected
in AD, in light of the early and dramatic cholinergic cell
loss in the AD brain,1-3 agents that augment the
cholinergic system have received the greatest attention
with regard to drug development and intervention.4,5
In this regard, three strategies have been investigated
to increase cholinergic neurotransmission. These have
involved the use of (i) precursors, such as choline, to
augment neurotransmitter synthesis at presynaptic
terminals,6 (ii) direct agonists to stimulate postsynaptic
muscarinic and/or nicotinic cholinergic receptors,7,8 and
(iii) anticholinesterases (anti-ChE) to inhibit the en-
zymes that metabolize naturally released acetylcholine
(ACh) and thereby amplify its postsynaptic action.4,5,9,10
The development of anti-ChEs has thus far proved to
be the most productive approach, with four drugs
currently approved for clinical use: specifically, tacrine
(Cognex, originally Pfizer/Warner-Lambert, MI, cur-
rently: Horizon Pharm, GE), donepezil (Aricept, Pfizer,
NY, Eisai, J apan), rivastigmine (Exelon, Novartis,
Switzerland), and galantamine (Reminyl, J anssen, NJ ).
In large part, the successful development of current
anti-ChEs has benefited from the extensive information
* To whom correspondence should be addressed. Phone: 410-558-
8278. Fax: 410-558-8323. E-mail: GreigN@vax.grc.nia.nih.gov.
† National Institute on Aging.
‡ Naval Research Laboratory.
Resu lts
§ National Institute on Drug Abuse. Current address: GeneFor-
matics Inc., San Diego, CA 92121.
Ch em istr y. The commercially unavailable isocyan-
ates 23 and 25 were prepared from anilines 19 and 20
#
University of North Carolina.
10.1021/jm010080x CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/20/2001

Methyl Analogues of Phenserine
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4063
Chloride (28) was reacted with eseroline (27) at 90
°C for 56 h to give the desired product, 12, in 25% yield
(Scheme 3). The Mannich reaction of eseroline (27) with
dimethylamine hydrochloride and formaldehyde gave
6-dimethylaminomethylene-eseroline (29) in a yield of
89%. Qualitative assessment of this product demon-
strated that no substitution in the 4′-position occurred.
Compound 29 was converted to 6-methyl eseroline (31)
quantitatively, by catalytic hydrogenation. Compound
30 then was reacted with methylisocyanate and phen-
ylisocyanate to give 6-methylphysostigmine (31) and
6-methylphenserine (13), respectively. Compound 29
also was reacted with phenylisocyanate to afford 6-di-
methylaminomethylene-phenserine (32). Compound 32
then was deaminated to give the desired 6-methylphen-
serine (13) by catalytic hydrogenation in a low yield
(Scheme 4). We have previously reported our procedures
for synthesizing known compounds 14-1823 and 33.29
Phenserine (2) and the fumarate of 2′,4′,6′-trimeth-
ylphenserine (14) were subjected to X-ray analysis to
define their configuration and conformation in their
crystalline state (Laboratory for the Structure of Matter,
Naval Research Laboratory, Washington, D.C.). Figure
2 illustrates their absolute configuration. Aligning the
tricyclic ring structures of both compounds to the same
plane demonstrates that substitution onto the phenyl
ring of the carbamate function causes its rotation in
relation to the rest of the molecule; torsion angles are
shown.
Biologica l Eva lu a tion . Table 1 illustrates the bio-
logical activity of compounds 3-18 and 31-33 against
freshly prepared human AChE and BChE, in compari-
son to physostigmine (1) and phenserine (2), whose
measured values are in accord with our previous
reports.9,22-27
F igu r e 1.
with phosgene by a general method, shown in Scheme
1.28 In this method, it proved important to use the
hydrochlorides of the anilines as the starting materials
and to undertake the reaction in a sealed vessel. Con-
sequently, highly volatile phosgene could not escape
from the reaction mixture, even at high temperature.
The aniline hydrochlorides proved to be poorly soluble
in the reaction solvent, toluene. Hence, after the reac-
tion, unreacted material was removed by filtration, and
the remaining phosgene, together with the solvent,
toluene, was readily removed by evaporation to give the
isocyanates as brownish oils. This procedure minimized
the handling of these irritant and moisture sensitive
materials and, additionally, avoided chromatography or
distillation.
Eseroline (27) then was reacted with isocyanates
(21-26) in ether to afford phenylcarbamates (6-11)
(Figure 1) according to a known procedure23 (Scheme
2). Usually, after the completion of the reaction, evapo-
ration of solvent provides a residue that invariably
contains a small quantity of unreacted isocyanate. This
then reacts with the carbamate reaction-product to form
N-phenyl-N-phenylallophanyl eseroline.29 Thereby the
yield is decreased, and purification is required. The
purification is complex as the Rf values of the allophanyl
and eseroline carbamate are similar. The original reac-
tion procedure was therefore improved by adding H2O
to destroy any trace of isocyanate, and the ether solution
was washed with dilute NaOH to remove unreacted
eseroline to give a highly pure product without the use
of chromatography, a notable improvement on the
synthetic scheme.
(a ) P h en ylca r ba m a te m od ifica tion , m on osu bsti-
tu tion : In comparison to unsubstituted phenserine (2),
which has a selectivity for AChE inhibition of 70-fold,
methyl substitution in the 2′ position, 3, provided a
2-fold improvement in AChE potency and a marginal
decline in BChE activity to afford an AChE selectivity
of 195-fold. The 2′-ethyl substitution, 15, rendered the
compound less BChE active still, with an AChE selec-
tivity of 290-fold. In contrast, methyl substitution in the
3′ position, 6, maintained AChE potency, compared to
phenserine (2), and increased BChE potency to afford
a 12-fold AChE selective compound. The 4′-methyl sub-
stitution, 4, reduced AChE and increased BChE potency,
compared to 2, to afford a compound that was unselec-
tive. Larger substitution in the 4′ position, such as with
an isopropyl, 18, further reduced AChE and increased
BChE potency to render a 15-fold BChE selective agent.
In contrast, similar substitution in the 2′ position, 17,
provided a 43-fold AChE selective compound.
(b) P h en ylca r ba m a te m od ifica tion , d isu bstitu -
tion : Compounds with dimethyl substituents in the 2′,3′
(7), 2′,4′ (5), and 2′,5′ positions (10) possessed a similar
AChE inhibitory activity to that of the lead compounds
physostigmine (1) and unsubstituted phenserine (2). All
these compounds, similar to 2, were AChE selective.
However, this was reduced for 7 and 10 as a conse-
quence of improved BChE potency. Dimethyl substitu-
tion in the 3′,4′-position, 8, likewise maintained AChE
activity but provided significant potency against BChE,

4064 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24
Yu et al.
Sch em e 1
Sch em e 2
Sch em e 3
Sch em e 4a
rendering an unselective inhibitor. In contrast, 2′,6′- (11)
and 3′,5′- (9) dimethyl substitutions reduced inhibitory
action against both enzyme subtypes, with the former
primarily affecting AChE and the latter BChE activity.
This was heightened by diethyl substitution (16).
(c) P h en ylca r ba m a te m od ifica tion , tr isu bstitu -
tion : The 2′,4′,6′-trimethyl substitution (14) was not
tolerated and resulted in a lack of inhibitory action
against both enzyme subtypes.
In summary, substitution into the 2′ position provided
improved AChE and reduced BChE inhibitory action,
compared to phenserine (2). In contrast, 4′ substitution
provided the reverse: a reduced AChE and improved
BChE activity. Substitution into the 3′ position main-
tained AChE activity and improved BChE action, yield-
ing a less selective compound. In general, disubstitution
was well tolerated when involving the 2′ position, with
the exception of 2′,6′-dimethylphenserine (11), and
a
Reagents: (a) HCl (pH 3), EtOH, 80 °C, 5 h; (b) Pd(HO)2/C,
MeOH, H2, 1 h; (c) PhNCO or CH3NCO, Et2O, Na; (d) PhNCO,
Et2O, Na; (e) Pd(HO)2/C, MeOH, H2.
resulted in improved BChE activity, compared to mono-
substitution, with a corresponding reduction in AChE
selectivity. Disubstitution was well tolerated when
involving the 4′ position for AChE inhibition, with
improved BChE action being derived from the 3′,4′
position. Disubstitutions involving the 3′ position were
well tolerated and provided compounds with activity

Methyl Analogues of Phenserine
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4065
Å in 14). The central five-membered ring is coplanar
with its adjoining phenyl ring ((0.05 Å in 2 and ( 0.02
Å in 14). The fused ring system is folded at the bond in
common to the two five-membered rings (C3a-C8a). In
2 there is only one weak intermolecular interaction
between N1 and N11 at 3.16 Å. The presence of the
fumarate in 14 provides stronger intermolecular hydro-
gen bonding with both N1 and N11, interacting with
the fumarate at 2.75 and 2.82 Å, respectively. There is
also a strong hydrogen bond between neighboring fu-
marate ions (O‚‚‚O ) 2.56 Å).
Discu ssion
Figure 3 shows a diagram of mammalian AChE into
which phenserine has been placed in its assumed
binding site. The binding domain for phenserine is
inside a 20 Å deep gorge that intrudes into the surface
of the enzyme.16,18,20 This is the very same domain that
binds and rapidly metabolizes ACh. Within the gorge
of AChE or BChE, three primary binding domains exist.
These include (i) an acyl pocket that defines the active
center involved in the catalysis of ACh and is centered
around an active serine residue, Ser203, in human
enzyme; (ii) an active center choline subsite involved
in the attraction and binding of the quaternary am-
monium of the choline moiety of ACh; and (iii) a
peripheral anionic site that is uninvolved in ACh
hydrolysis but is the binding site of the ChEIs, tacrine,
and donepezil and lies at the mouth of the gorge.
Binding to this latter domain blocks access of ACh to
the former binding sites and produces a conformational
change such that the binding domains cannot align
sufficiently for enzyme activity.20,21
Within the acyl binding domain of AChE, attack of
the carbonyl function of phenserine (2) and analogues
or ACh occurs through Ser203, via a charge relay system
within a catalytic triad of amino acids that involves the
imidazole ring of His447 and carboxylic group of
16,18,20
Glu334
.
The phenserine-AChE intermediate, like
physostigmine (1), likely exists in a tetrahedral confor-
mation that then collapses to a carbamylated drug-
enzyme complex. This complex is far more stable than
the acetylenzyme observed as a result of nucleophilic
attack on the carbonyl group of ACh, which rapidly
hydrolyzes to regenerate active enzyme.19 In contrast,
the carbamylated enzyme only slowly hydrolyzes at rate
that is dependent on the structure of the carbamate
moiety.19,30 For phenserine (2), this complex likely is
further stabilized by both hydrophobic and π electron
interactions, due to the π-π stacking of the phenyl
group of the phenylcarbamate between the flanking
phenyl moieties of Phe295 and Phe297 (Figure 3). These
amino acids are lacking within BChE, wherein they are
replaced by valine and leucine, respectively.18,20,21 These
latter residues that delineate the limit of the acyl
domain within BChE are smaller than phenyalanine
and thus create a pocket within BChE that allows the
additional steric bulk of 4′-substituted phenylcarbamate
analogues, 4 and 18. These same compounds are not
well tolerated within AChE and hence 4′ substitution
attentuates the selectivity of the analogues to favor
BChE inhibitory activity. This is in accord with site-
directed mutagenesis studies in both human31,32 and
Torpedo33 AChE, which illustrated that mutation of
F igu r e 2.
intermediate between 2′ and 4′ substitution. Trimethyl
substitution in the 2′,4′,6′ position resulted in a lack of
activity.
(d ) C6 p osition m od ifica tion : Substitution in the
C6 position was not tolerated for either phenserine or
physostigmine (13, 31, 32). AChE and BChE inhibitory
activities were substantially reduced.
(e) N17 p osition m od ifica tion : Substitution in the
N17 position was not tolerated for either phenserine or
physostigmine. N-Methylphenserine (12) and N-meth-
ylphysostigmine (33) were dramatically less active than
2 and 1, respectively.
X-r a y Cr ysta llogr a p h y. Sin gle-cr ysta l X-r a y
a n a lysis of 2 a n d 14: The results of the X-ray studies
are illustrated in Figure 2. Compound 14 crystallized
as a fumarate salt. Except for the rotation of the phenyl
ring of the carbamate function, both compounds exhibit
the same overall conformation. The terminal five-
membered rings have an envelope conformation with
C2 being the out of plane atom (+0.61 Å in 2 and -0.65

4066 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24
Yu et al.
Ta ble 1. 50% Inhibitory Concentration (IC50, nM) + SEMa of Compounds versus Human Erythrocyte AChE and Plasma BChE
IC50 (nM)
no.
compounds
AChE
BChE
16 ( 3
selectivity
1
2
3
15
17
6
4
18
7
physostigmine
phenserine
2′-methylphenserine
2′-ethylphenserine
2′-isopropylphenserine
3′-methylphenserine
28 ( 2
2-fold BChE
70-fold AChE
195-fold AChE
290-fold AChE
43-fold AChE
12-fold AChE
2-fold AChE
15-fold BChE
13-fold AChE
130-fold AChE
19-fold AChE
3-fold BChE
none
2-fold AChE
10-fold AChE
3-fold AChE
2-fold AChE
15-fold AChE
none
22 ( 1
1560 ( 270
1950 ( 240
2915 ( 535
650 ( 45
165 ( 41
250 ( 8
10 ( 2
10 ( 1
15 ( 1
28 ( 4
4′-methylphenserine
140 ( 4
760 ( 20
23 ( 6
4′-isopropylphenserine
2′,3′-dimethylphenserine
2′,4′-dimethylphenserine
2′,5′-dimethylphenserine
2′,6′-dimethylphenserine
2′,6′-diethylphenserine
3′,4′-dimethylphenserine
3′,5′-dimethylphenserine
2′,4′,6′-trimethylphenserine
N-methylphysostigmine
N-methylphenserine
50 ( 1
170 ( 32
1820 ( 560
490 ( 79
290 ( 50
1100 ( 50
66 ( 7
5
14 ( 1
26 ( 1
785 ( 140
1500 ( 50
31 ( 7
78 ( 17
1300 ( 75
210 ( 40
690 ( 50
>10000
260 ( 130
2500 ( 100
10
11
16
8
9
798 ( 147
3290 ( 885
420 ( 120
>10000
8530 ( 235
5020 ( 180
3890 ( 1500
14
33
12
31
13
32
6-methylphysostigmine
6-methylphenserine
6-dimethylaminoethylenphysostigmine
19-fold AChE
none
a
The IC50 data of compounds 1-5 and 14-18 were cited from reference 23. The IC50 data of compound 33 is from reference 43.
steric energy E1 (Table 2); which is a combination of
bend, stretch-bend, torsion, van der Waals, and dipole/
dipole interactions. According to our biological evalua-
tion (Table 1), 2′-methylphenserine (3) possesses the
highest anti-AChE activity. We therefore chose its
conformation as the most favored for AChE interaction
and binding of the phenserine series. Our enzyme
kinetics studies support this choice, confirming the
greater potency of 3 over phenserine (2) and physostig-
mine (1) with respective inhibition constants, Ki, of
0.0047 µM (3), 0.048 µM (2), and 0.19 µM (1).35,36
Figure 4 shows the computed energy minimized
conformation of compounds 2, 3, 12-14, and 18, with
their critical dihedral angles, C17-C12-N11-C10 and
C10-O9-C5-C4, which for 2′-methylphenserine (3) are
3.4° and 159.8°, respectively. E1 values, as determined
by MM2,34 are likewise shown. The predicted conforma-
tion of 3 has the two phenyl groups on either side of its
carbonyl in less of an orthogonal relationship than
compounds 2, 12-14, and 18. To obtain a 2′-meth-
F igu r e 3.
ylphenserine-like conformation, the resulting torsion
angles for 2, 12-14, and 18 would change the energy
to E2, with ∆E (E2 - E1) being the energy difference
between the two conformations and representing the
conformational barrier.34 Values for ∆E for 2-18 and
32 versus AChE inhibitory activity are shown in Table
2. The differences in measured versus computed values
for 2 and 14 between Figures 2 and 4 likely are a
consequence of (i) the compounds assuming a different
low energy conformation in the noncrystalline versus
crystalline state and (ii) the assessment of 2′,4′,6′-
trimethylphenserine as a fumarate salt to obtain a
crystalline form for X-ray analysis.
Phe295 and Phe297 to their corresponding amino acids
for BChE provided the wide substrate specificity as-
sociated with BChE.
As shown by X-ray crystallography (Figure 2), the
hydrogen bonding between the N11-H of 14 and the
carbonyl group of fumaric acid, together with substitu-
tion onto the phenylcarbamate, causes rotation of the
phenyl moiety of the carbamate in relation to the ‘A’
ring of the remaining tricyclic structure, when compared
to 2. The torsion angles are indicated by dihedral angles
C17-C12-N11-C10 and C10-O9-C5-C4.
Computer aided molecular modeling was undertaken,
34
For compounds 12 and 13, unlike 3, the C17-C12-
N11-C10 and C10-O9-C5-C4 computed torsion
angles are dramatically different (by 19.9° and 0.98°,
and by -3.33° and -14.1°, respectively) from 2′-meth-
ylphenserine (3) (Figure 4). In addition, their ∆E values
(0.58 and 0.68 kcal/mol, respectively) are substantial
compared to compounds 2 and 3 (Figure 4, Table 2). For
example, they are equivalent to the free energy contri-
using MM2 force field, as available in CS Chem 3D.
When comparing target compounds, all parameters
were kept constant throughout. Initially, the three-
dimensional structure of each compound was built, then
conformational analysis was performed on each to
provide the energy minimized conformation of all sub-
stituted phenserines, 2-18 and 32, in their noncrystal-
line state, together with their corresponding minimum

Methyl Analogues of Phenserine
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4067
Ta ble 2. Relationship between IC50 for AChE and Conformational Energy Barrier ∆E (E2 - E1) for Substituted Phenserines
IC50 AChE
(nM)
∆E
(kcal/mol)
E1
E2
no.
compound
(kcal/mol)
(kcal/mol)
2
phenserine
22 ( 2
0.01
15.20
15.21
Methyl Substituted Phenserine
3
5
7
10
6
8
9
4
13
12
11
14
2′-methylphenserine
10 ( 2
0
15.16
14.82
16.34
14.86
14.89
15.75
14.49
14.83
16.12
29.13
20.29
19.83
15.16
14.84
16.35
14.87
14.89
15.75
14.50
14.84
16.70
29.81
20.43
19.88
2′,4′-dimethylphenserine
2′,3′-dimethylphenserine
2′,5′-dimethylphenserine
3′-methylphenserine
3′,4′-dimethylphrnserine
3′,5′-dimethylphenserine
4′-methylphenserine
6-methylphenserine
N-methylphenserine
2′,6′-dimethylphenserine
2′,4′,6′-trimethylphenserine
14 ( 1
0.02
0.01
0.01
0
23 ( 6
26 ( 1
28 ( 4
31 ( 7
0
78 ( 17
140 ( 4
210 ( 130
690 ( 50
785 ( 140
1300 ( 75
0.01
0.01
0.58
0.68
0.14
0.05
Other Substituted Phenserine
15
17
18
16
32
2′-ethylphenserine
10 ( 1
0.01
0.07
0
0.53
0.66
16.44
18.03
17.38
24.22
24.16
16.45
18.10
17.38
24.75
24.82
2′-isopropylphenserine
4′-isopropylphenserine
2′,6′-diethylphenserine
6-dimethylaminomethylenphenserine
15 ( 1
760 ( 20
1500 ( 50
2500 + 100
butions associated with the force of a weak van der
Waals or a hydrophobic methylene-methylene interac-
tion37 and are sufficient to potentially alter the binding
equilibrium constant, Ki, by some 2-fold (∆E° ) -RT
ln Ki).38 However, interactions between the enzyme
binding domains and an inhibitor may compensate for
or, vice versa, may increase the conformational barrier.
The reduced AChE inhibitory activities associated with
compounds 12 and 13, compared to 3 (Table 2), suggest
that a methyl group substituted in either the N11 or
C6 positions seriously impedes the free rotation of the
phenyl group of the phenylcarbamate and the ‘A’ ring
(Figure 1) around the carbonyl group. In contrast, when
substituted onto the phenylcarbamate (4-10), this does
not occur and ∆E values are comparatively low. Interac-
tion with the active center choline binding domain of
AChE and BChE likely occurs via the basic N1 position
of physostigmine (1), phenserine (2), and analogues.19,39
This choline binding site exists as a region of 14 con-
served aromatic amino acids centered around Trp86 in
humans.20,31 With required binding within the two
relatively fixed enzyme domains, rotation of the phen-
ylcarbamate in relation to the tricyclic moiety, bearing
the N1 group in the ‘C’ ring (Figure 1), can therefore
favor or disfavor the alignment of the carbonyl function
for attack by Ser203. Poor alignment, as a consequence
of a different conformation and impeded rotation, likely
explains the reduced activity of 12 and 13 compared to
1-3.
In contrast, this dynamic analysis cannot explain the
reduced AChE inhibitory activity of 18 compared to 3,
as its conformational barrier, ∆E, was zero, and its
conformation was similar to 3 (Figure 4 and Table 2);
indicating that 4′-substitutions elongate the molecule
without impeding free rotation around the carbonyl
moiety. It is this elongation within the size-restricted
gorge of AChE that perturbs the AChE versus BChE
inhibitory action of 18. Similarly, the reduced anti-
AChE action of 2′,4′,6′-trimethylphenserine (14) and
2′,6′-dimethylphenserine (11) also cannot be explained
by impeded rotation. Their low ∆E value suggests that
both compounds, similar to 2-10, should be capable of
aligning to the binding domains of the enzyme. Rather,
the reduced biological activity of 14 and 11 likely results
from their methyl substitutions obstructing the ap-
proach of Ser203 to the carbonyl moiety. For enzyme-
drug binding, the oxygen anion of Ser203 requires free
access to the positive carbon of the carbonyl group of
the inhibitor.19,21 The direction of approach of this
nucleophilic attack has been the subject of intensive
study. In general, and specifically for those cases that
proceed by the described tetrahedral mechanism, there
is no single definable and preferred transition state, but
rather a “cone” of trajectories exists.40 All approaches
within this cone then lead to reaction at comparable
rates. It is only when an approach comes from outside
the cone that the reaction rate falls, resulting in reduced
biological action.
It is additionally possible that the N-methyl of 12 and
C6 substitutions of 13 and 32 are likewise located within
the ‘cone’ area of the carbonyl group to hinder the
approach of Ser203, and, for these as well as for 16, a
combination of factors may reduce their final biological
activity.
In summary, for optimal interaction, binding, and
inhibition of AChE and BChE by phenserine (2) and
analogues, the active center choline subsite and acyl
pocket of the enzymes require holding the compounds’
tricyclic π-system and carbamate structures, respec-
tively, in an appropriate conformation to allow the
attack of Ser203 to the carbonyl group of the inhibitor to
form a tetrahedral intermediate and drug-enzyme
complex.19 We predict that amino acid differences that
discriminate AChE from BChE20,21 can then be utilized
to differentially stabilize either drug-AChE or drug-
BChE complexes by alkyl substitution into specific
positions on the phenylcarbamate (i.e., 2′-methylphen-
serine (3) and 4′-isopropylphenserine (18), respectively)
or the tricyclic moiety.27 Particular substitutions can,
however, sterically impede drug-enzyme interactions
and/or disfavor the required alignment between the
drug and enzyme for optimal binding and thus reduce
their inhibitory activity. In addition, hydrophobic as well
as π electron interactions, between the phenylcarbamate
of phenserine (2) and analogues and either the phenyl
moieties of Phe295 and Phe297 in AChE or flanking

4068 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24
Yu et al.
carried out using flash column chromatography (silica gel 60,
230-400 mesh) with the described solvents. All reactions
involving nonaqueous solutions were performed under an inert
atmosphere.
3,4-Meth ylp h en ylisocya n a te (23). 3,4-Dimethylaniline
(1.58 g, 10 mmol) was added to 10 mL of toluene in a sealed
tube with a small magnetic stirring bar. After addition of
phosgene (in toluene, 20%) (11 mL, approximately 21 mmol),
the sealed tube was closed tightly and inserted into a small
oil bath, maintained at 100 °C, that was set on top of a stir
plate. Following 6 h of reaction with vigorous stirring, the
suspended hydrochloride of aniline almost disappeared and
the reaction was stopped. Cooling to room temperature pro-
vided a reaction mixture that was filtered to give a clear,
transparent filtrate that was evaporated by vacuum to obtain
compound 23 (1.03 g, 70%) as a slight brownish oil: 1H NMR
(CDCI3) δ 7.75-6.82 (m, 3 H, Ph-H), 2.28 (s, 6H, 3,4-dimethyl);
EI-MS, m/z: 147 (M+).
2,5-Dim eth ylp h en ylisocya n a te (25). According to the
procedure for the synthesis of 23, compound 25 was obtained
as slightly brown oil: 1H NMR (CDCI3) δ 7.10-6.05 (m, 3 H,
Ph-H), 2.38 (s, 3H, 2-CH3), 2.28 (s, 3H, 5-CH3); EI-MS, m/z:
147 (M+).
(-)-(3a S )-1,3a ,8-Tr im e t h yl-1,2,3,3a ,8,8a -h e xa h yd r o-
p yr r olo[2,3-b]in d ol-5-yl N-(3′-Meth ylp h en yl)ca r ba m a te
(6). Eseroline (27) (108 mg, 0.5 mmol) was dissolved in
anhydrous ether (5 mL), and a piece of Na metal (approxi-
mately 1 mg) was added. The mixture was stirred at room
temperature for 1 min, then 3′-methylphenylisocyanate (21)
(66 mg, 0.5 mmol) was added. The reaction mixture was con-
tinuously stirred at room temperature and monitored by TLC.
When compound 27 had virtually disappeared, 1 mL of H2O
was added to destroy any remaining trace of unreacted isocy-
anate. The reaction mixture was diluted into 20 mL of Et2O
and washed with NaOH aqueous solution (approximately 5%
concentration) followed by brine. The ether solution was dried
over MgSO4, with stirring, and then filtered to obtain a clear
ether solution of the product. Evaporation of ether provided
compound 5 (158 mg, 90%) as a foam: [R] 2D0 -66.9° (c ) 0.2,
CHCl3); 1H NMR (CDCI3) δ 7.25-6.70 (m, 6H, Ar-H), 6.32 (d,
J ) 8.5 Hz, 1H, C7-H), 4.05 (s, 1H, C8a-H), 2.88 (s, 3H, CH3-
N8), 2.80-2.50 (m, 2H, C2-H2), 2.48 (s, 3H, CH3-N1), 2.28 (s,
3H, 3′-CH3), 1.90 (m, 2H, C3-H2), 1.40 (s, 3H, C3a-CH3); CI-
MS (NH3), m/z: 352 (MH+). The base of compound 5 was
dissolved in MeOH and then was added to a methanol solu-
tion containing an equivalent of L-(+)-tartaric acid. This was
concentrated under vacuum until it became syrup-like and
then was diluted with ether. A precipitated gum was sepa-
rated by decanting off the ether solution, and, thereafter, a
further batch of fresh ether was added. After scraping, the
gum was crystallized from the ether solution to give the
tartrate of compound 5: mp 96-98 °C. Anal. (C21H25N3O2‚
C4H6O6) C, H, N.
F igu r e 4. Conformation with minimum steric energy (E1) and
minimized energy in 2′-methylphenserine-like conformation
(E2).
aromatic amino acids in BChE, likely stabilize the
drug-enzyme complex in a manner that cannot occur
for physostigmine (1). Hydrolysis to regenerate active
enzyme then would require a higher free energy and
account for the long in vivo duration of enzyme inhibi-
tion of phenserine (2) and analogues (half-life 8 h)
compared to physostigmine (1) (half-life 30 min).9 Cur-
rently, X-ray crystallographic studies are being under-
taken on phenserine (2) and analogues within AChE
from Torpedo californica (Dr. Doriano Lamba, Istituto
di Strutturistica Chimica, Trieste, Italy) to further
elucidate the mechanism of action of the compounds and
the validity of our computer modeling.
Compounds 7-11 were synthesized from their correspond-
ing isocyanates (22-26) according to the above procedure for
synthesis of compound 6.
(-)-(3a S )-1,3a ,8-Tr im e t h yl-1,2,3,3a ,8,8a -h e xa h yd r o-
p yr r olo[2,3-b]in d ol-5-yl N-(2′,3′-Dim eth ylp h en yl)ca r ba m -
a te (7): [R] 2D0 62.5° (c ) 0.3, CHCl3); 1H NMR (CDCI3) δ
7.20-6.50 (m, 5 H, Ar-H), 6.28 (d, J ) 8.5 Hz, 1H, C7-H), 4.10
(s, 1H, C8a-H), 2.88 (s, 3H, CH3-N8), 2.75-2.50 (m, 2H, C2-
H2), 2.45 (s, 3H, CH3-N1), 2.25 (s, 3H, 2′-CH3), 2.15 (s, 3H,
3′-CH3), 1.90 (m, 2H, C3-H2), 1.35 (s, 3H, C3a-CH3); CI-MS
(NH3), m/z: 366 (MH+). Tartrate of compound 7: mp 108-
110 °C. Anal. (C22H27N3O2‚C4H6O6) C, H, N.
(-)-(3a S )-1,3a ,8-Tr im e t h yl-1,2,3,3a ,8,8a -h e xa h yd r o-
p yr r olo[2,3-b]in d ol-5-yl N-(3′,4′-Dim eth ylp h en yl)ca r ba m -
a te (8): [R] 2D0 62.0° (c ) 0.1, CHCl3); 1H NMR (CDCI3) δ
7.15-6.25 (m, 6 H, Ar-H), 4.15 (s, 1H, C8a-H), 2.88 (s, 3H,
CH3-N8), 2.75-2.50 (m, 2H, C2-H2), 2.48 (s, 3H, CH3-N1),
2.22 (s, 3H, 3′-CH3), 2.20 (s, 3H, 4′-CH3), 1.99 (m, 2H, C3-H2),
1.40 (s, 3H, C3a-CH3); CI-MS (NH3), m/z: 366 (MH+). Tartrate
of compound 8: mp 96-98 °C. Anal. (C22H27N3O2‚C4H6O6) C,
H, N.
Exp er im en ta l Section
Ch em istr y. Melting points (uncorrected) were measured
1
with a Fisher-J ohns apparatus; H NMR were recorded on a
Bruker (Bellevica, MA) AC-300 spectrometer; MS (m/z) were
recorded on a Hewlett-Packard 5890 GC-MS (EI) and on a
Finnigan-1015D mass spectrometer. Optical rotations were
measured by J ASCO, model DIP-370 (J apan, Spectroscopic
Co., Ltd.); elemental analyses were performed by Atlantic
Microlab, Inc. Unless otherwise indicated, all separations were

Methyl Analogues of Phenserine
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4069
(-)-(3a S )-1,3a ,8-Tr im e t h yl-1,2,3,3a ,8,8a -h e xa h yd r o-
p yr r olo[2,3-b]in d ol-5-yl N-(3′,5′-Dim eth ylp h en yl)ca r ba m -
a te (9): [R] 2D0 -67.8° (c ) 0.2, CHCl3); 1H NMR (CDCI3) δ
7.02-6.65 (m, 5 H, Ar-H), 6.30 (d, J ) 8.5 Hz, 1H, C7-H), 4.10
(s, 1H, C8a-H), 2.89 (s, 3H, CH3-N8), 2.75-2.55 (m, 2H, C2-
H2), 2.52 (s, 3H, CH3-N1), 2.28 (s, 6H, 3′-CH3 and 5′-CH3),
1.85-2.00 (m, 2H, C3-H2), 1.40 (s, 3H, C3a-CH3); CI-MS (NH3),
m/z: 366 (MH+). Tartrate of compound 9: mp 96-98 °C. Anal.
(C22H27N3O2‚C4H6O6) C, H, N.
of solvent gave a residue that was chromatographed on silica
gel (CH2Cl2/MeOH ) 10:1) to give compound 30 (163 mg,
90.0%) as a gum: [R] D20 -82.8° (c ) 0.3, CHCl3); 1H NMR
(CDCI3) δ 6.60 (s, 1H, C4-H), 6.34 (s, 1H, C7-H), 4.80 (s, 1H,
C8a-H), 2.95 (s, 3H, CH3-N8), 2.70-2.50 (m, 2H, C2-H2), 2.65
(s, 3H, CH3-N1), 2.15 (s, 3H, CH3-Ar), 2.05-1.95 (m, 2H, C3-
H2), 1.45 (s, 3H, C3a-CH3); CI-MS (NH3), m/z: 232 (MH+).
According to the procedure of making tartrate of compound
12, tartrate of compound 30 was obtained as solid: Anal.
(C14H20N2O‚1.5C4H6O6‚0.5H2O) C, H, N.
(-)-(3a S)-1,3a ,6,8-Tetr a m eth yl-1,2,3,3a ,8,8a -h exa h yd r o-
p yr r olo[2,3-b]in d ol-5-yl N-P h en ylca r ba m a te (13). (i) Com-
pound 30 (55 mg, 0.2 mmol) was dissolved in anhydrous ether
(2 mL), and a piece of Na metal (approximately 1 mg) was
added. The mixture was stirred under nitrogen at room
temperature for 5 min, then phenylisocyanate (23.6 mg, 0.2
mmol) was added. The reaction mixture was continuously
stirred at room temperature and monitored by TLC. When
compound 30 had almost disappeared, 0.5 mL of H2O was
added to destroy any trace of remaining unreacted phenyliso-
cyanate. The reaction mixture was diluted into 10 mL of ether
and washed with diluted NaOH aqueous solution (approxi-
mately 5% concentration), followed by brine. The ether solution
was dried over MgSO4, with stirring, and then filtered to
obtain a clear ether solution of the product. Evaporation of
solvent gave compound 13 (54 mg, 77.6%) as a foam: [R] D20
-80.0° (c ) 0.2, CHCl3); 1H NMR (CDCI3) δ 7.45-6.80 (m 5H,
Ar-H), 6.72 (s,1H, C4-H), 6.20 (s, 1H, C7-H), 4.20 (s, 1H, C8a-
H), 2.88 (s, 3H, CH3-N8), 2.75-2.55 (m, 2H, C2-H2), 2.50 (s,
3H, CH3-N1), 2.23 (s, 3H, 6-CH3), 2.00-1.85 (m, 2H, C3-H2),
1.45 (s, 3H, C3a-CH3); CI-MS (NH3), m/z: 351(MH+). In accord
with the procedure for preparing a tartrate of compound 12,
the tartrate of compound 13 was obtained as crystals: mp
127-130 °C. Anal. (C21H25N3O2‚C4H6O6.) C, H, N.
(ii) Compound 32 (45 mg, 0.12 mmol) was dissolved in
MeOH (2 mL), and Pd(OH)2/C (2 mg) was added. The reaction
mixture then was stirred under hydrogen at atmospheric
pressure (room temperature, 1 h). Following removal of the
catalyst by filtration, evaporation of the solvent gave a residue
that was chromatographed on silica gel (CH2Cl2/MeOH ) 10:
1) to give compound 13 (2 mg, 5%)
(-)-(3a S)-1,3a ,6,8-Tetr a m eth yl-1,2,3,3a ,8,8a -h exa h yd r o-
p yr r olo[2,3-b]in d ol-5-yl N-Meth ylca r ba m a te (31). In ac-
cord with the described procedure for making compound 11,
reaction of compound 30 with methylisocyanate gave com-
pound 31 (in 83.0% yield) as a foam: [R] 2D0 -63.4° (c ) 0.2,
CHCl3); 1H NMR (CDCI3) δ 6.65 (s, 1H, C4-H), 6.15 (s, 1H,
C7-H), 4.15 (s, 1H, C8a-H), 2.88 (s, 3H, CH3-N8), 2.70 (d, J )
5 Hz, 3H,CH3-NH-), 2.70-2.50 (m, 2H, C2-H2), 2.48 (s, 3H,
CH3-N1), 2.08 (s, 3H, C6-CH3), 2.00-1.80 (m, 2H, C3-H2), 1.40
(s, 3H, C3a-CH3); CI-MS (NH3), m/z: 290 (MH+). Similar to
the procedure for preparing the tartrate of compound 12, the
tartrate of compound 31 was obtained as a solid: Anal.
(C16H23N3O2‚C4H6O6.) C, H, N.
(-)-(3a S )-1,3a ,8-Tr im e t h yl-1,2,3,3a ,8,8a -h e xa h yd r o-
p yr r olo[2,3-b]in d ol-5-yl N-(2′,5′-Dim eth ylp h en yl)ca r ba m -
a te (10): [R] 2D0 -61.0° (c ) 1.0, CHCl3); H NMR (CDCI3) δ
1
7.10-6.30 (m, 6 H, Ar-H), 4.13 (s, 1H, C8a-H), 2.95 (s, 3H,
CH3-N8), 2.80-2.55 (m, 2H, C2-H2), 2.55 (s, 3H, CH3-N1), 2.32
(s, 6H, 2′-CH3), 2.27 (s, 3H, 5′-CH3), 2.00-1.90 (m, 2H, C3-
H2), 1.42 (s, 3H, C3a-CH3); CI-MS (NH3), m/z: 366 (MH+).
Tartrate of compound 10: mp 108-110 °C. Anal. (C22H27N3O2‚
C4H6O6) C, H, N.
(-)-(3a S )-1,3a ,8-Tr im e t h yl-1,2,3,3a ,8,8a -h e xa h yd r o-
p yr r olo[2,3-b]in d ol-5-yl N-(2′,6′-Dim eth ylp h en yl)ca r ba m -
a te (10): [R] 2D0 -60.8° (c ) 0.4, CHCl3); H NMR (CDCI3) δ
1
7.10-6.15 (m, 6 H, Ar-H), 4.10 (s, 1H, C8a-H), 2.83 (s, 3H,
CH3-N8), 2.80-2.55 (m, 2H, C2-H2), 2.48 (s, 3H, CH3-N1), 2.30
(s, 6H, 2′-CH3 and 6′-CH3), 2.00-1.80 (m, 2H, C3-H2), 1.42 (s,
3H, C3a-CH3); CI-MS (NH3), m/z: 366 (MH+). Tartrate of
compound 10: mp 128-130 °C. Anal. (C22H27N3O2‚C4H6O6) C,
H, N.
(-)-(3a S )-1,3a ,8-Tr im e t h yl-1,2,3,3a ,8,8a -h e xa h yd r o-
pyr r olo[2,3-b]in dol-5-yl N-(Meth yl)-ph en ylcar bam ate (12).
Eseroline 27 (218 mg, 1 mmol) was dissolved in pyridine (5
mL), and N-methyl-phenylcarbamoyl chloride (28) (678 mg, 4
mmol) was added. The reaction mixture was heated in an oil
bath at 90 °C and stirred for 56 h. Evaporation of pyridine by
vacuum gave a residue which was chromatographed on a silica
gel (CH2Cl2/MeOH ) 20:1) to give compound 12 (90 mg, 25.6%)
as a gum: [R] 2D0 -32.5° (c ) 0.3, CHCl3); H NMR (CDCI3) δ
1
7.40-7.10 (m, 5 H, Ar-H), 6.80-6.50 (m, 2H, C4-H and C6-
H), 6.22 (d, J ) 8.5 Hz, 1H, C7-H), 4.08 (s, 1H, C8a-H), 3.37
(s, 3H, CH3-N), 2.87 (s, 3H, CH3-N8), 2.75-2.45 (m, 2H, C2-
H2), 2.47 (s, 3H, CH3-N1), 1.85 (m, 2H, C3-H2), 1.35 (s, 3H,
C3a-CH3); CI-MS (NH3), m/z: 352 (MH+); EI-MS, m/z (relative
intensity): 218 (MH+ - PhNCH3CO-, 15), 134 (PhNCH3O,
100), 106 (PhNCH3, 50), 77 (Ph, 60). The base of compound
12 was dissolved in MeOH, and then was added to a methanol
solution containing an equivalent of L-(+)-tartaric acid. The
methanol solution was concentrated under vacuum until it
became syrup-like and then was diluted with ether. A pre-
cipitated gum was separated by decanting off the ether
solution, and, thereafter, a further batch of fresh ether was
added. After scraping, the gum was solidified from the ether
solution to provide the tartrate of compound 12: Anal.
(C21H25N3O2‚C4H6O6) C, H, N.
(-)-(3a S)-1,3a ,8-Tr im eth yl-6-(d im eth yla m in o)m eth yl-
en e-1,2,3,3a ,8,8a -h exa h yd r op yr r olo[2,3-b]in d ol-5-ol (29).
Eseroline 27 (334 mg, 1 mmol), hydrochloride of dimethy-
lamine (405 mg, 6 mmol), and formaldehyde (37.9%) aqueous
solution (1 mL, about 10 mmol) were added into 10 mL of
EtOH in a sealed tube. The reaction mixture was stirred for 5
h at 80 °C in an oil bath. After cooling, the reaction mixture
was concentrated to provide a residue which was directly
chromatographed on silica gel (CH2Cl2/MeOH ) 10:1) to give
compound 29 (245 mg, 89.0%) as a gum: [R] 2D0 -3.4° (c ) 0.1,
CHCl3); 1H NMR (CDCI3) δ 6.55 (s, 1H, C4-H), 6.12 (s, 1H,
C7-H), 4.12 (s, 1H, C8a-H), 3.50 (s, 2H, Ar-CH2-N), 2.88 (s,
3H, CH3-N8), 2.85-2.55 (m, 2H, C2-H2), 2.54 (s, 3H, CH3-N1),
2.28 (s, 6H, Me2N-Ar), 2.10-1.80 (m, 2H, C3-H2), 1.45 (s, 3H,
C3a-CH3); CI-MS (NH3), m/z: 275 (MH+). HR-MS m/z: Calcd
for C16H25N3O: 275.19995; Found: 275.1990.
(-)-(3a S)-1,3a ,8-Tr im eth yl-6-(d im eth yla m in o)m eth yl-
en e-1,2,3,3a ,8,8a -h exa h yd r op yr r olo[2,3-b]in d ol-5-yl N-
P h en ylca r ba m a te (32). Reaction of compound 29 with phen-
ylisocyanate, according to the procedure for preparing com-
pound 13, provided compound 32 (in 90.0% yield) as a foam:
[R] 2D0 -48.7° (c ) 0.2, CHCl3); H NMR (CDCI3) δ 7.40-6.90
1
(m, 5H, Ar-H), 6.72 (s, 1H, C4-H), 6.25 (s, 1H, C7-H), 4.05 (s,
1H, C8a-H), 3.30 (s, 2H, Ar-CH2-N), 2.90 (s, 3H, CH3-N8),
2.70-2.50 (m, 2H, C2-H2), 2.47 (s, 3H, CH3-N1), 2.22 (s, 6H,
Me2N-Ar), 1.95-1.80 (m, 2H, C3-H2), 1.40 (s, 3H, C3a-CH3);
EI-MS, m/z: 394 (M+). HR-MS m/z: Calcd for C23H30N4O2:
394.2371; Found: 394.2369.
Sin gle-Cr ysta l X-r a y An a lysis of 2 a n d 14. Data for
compound 2 were collected on a Bruker P4 serial automatic
diffractometer with a graphite monochromator in the incident
beam. The crystal is orthorhombic in space group P212121 with
a ) 7.689(1), b ) 14.459 (1), and c ) 16.636(3) Å. Data for
compound 14 were collected on a Bruker SMART 1K CCD
system mounted on a 6 Kw Cu rotating anode using Gobels
mirrors to focus the beam. This crystal was also orthorhombic
(-)-(3a S)-1,3a ,6,8-Tetr a m eth yl-1,2,3,3a ,8,8a -h exa h ydr o-
p yr r olo[2,3-b]in d ol-5-ol (30). Compound 29 (215 mg, 0.78
mmol) was dissolved in MeOH (5 mL), and Pd(OH)2/C (10 mg)
was added. The reaction mixture was stirred under hydrogen
at atmospheric pressure and room temperature for 1 h.
Thereafter, the catalyst was removed by filtration. Evaporation

4070 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24
Yu et al.
in space group P212121 with a ) 8.632(1), b ) 12.571(1), and
c ) 12.495(1) Å. The structures were solved by direct methods
and refined by full-matrix least-squares on F2 values using
programs in the SHELXTL-PLUS package.41 The parameters
refined included the coordinates and anisotropic thermal
parameters for all non-hydrogen atoms and coordinates only
for the hydrogen atoms on the fumarate hydroxyl. All other
hydrogen atoms were included using a riding model in which
the coordinate shifts of their covalently bonded atoms were
applied to the attached hydrogens with C-H ) 0.96 Å and
N-H ) 0.86 Å. H angles were idealized and Uiso(H) set at
fixed ratios of Uiso values of bonded atoms. Final R-factors
were 0.054 for 1822 observed reflections for 2 and 0.060 for
1468 observed reflections for 14. The coordinates for both
compounds have been deposited as Supporting Information
and are also available from the Cambridge Crystallographic
Database.42
Qu a n tita tion of An tich olin ester a se Activity. The action
of compounds 1-18 and 31-33 to inhibit the ability of freshly
prepared human AChE and BChE, derived from plasma and
whole red blood cells, respectively, to enzymatically degrade
the specific substrates acetyl-(â-methyl)thiocholine and s-
butyrylthiocholine (0.5 mmol/L) (Sigma Chemical Co., St. Lois,
MO) were quantified.23-27,43 Compounds were dissolved in
Tween 80/EtOH 3:1 (v:v; <150 µL total volume) and were
diluted in 0.1 M Na3PO4 buffer (pH 8.0) in half-log concentra-
tions to provide a final concentration range that spanned 0.3
nM to 30 mM. Tween 80/EtOH was diluted to in excess of 1 in
5000 and no inhibitory action on either AChE or BChE was
detected in separate prior experiments.
National Institute on Drug Abuse (NIDA), NIH, for use
of NMR and polarimeter equipment. J .L.F.-A. thanks
ONR and NIDA for financial support.
Su p p or tin g In for m a tion Ava ila ble: Crystallographic
information for compounds 2 and 14 and tables of the
parameters used in MM2 for the comformational analysis of
cabamates. This material is available free of charge via the
Internet at http://pubs.acs.org.
Refer en ces
(1) Perry, E. K.; Tomlinson, B. E.; Blessed, G.; Bergman, K.; Gibson,
P. H.; Perry, R. H. Correlation of cholinergic abnormalities with
senile plaques and mental test scores in senile dementia Br.
Med. J . 1978, 2, 1457-1459.
(2) Whitehouse, P. J .; Price, D. L.; Struble, R. G.; Clark, A. W.; Coyle,
J . T.; Delon, M. R. Alzheimer’s disease and senile dementia: loss
of neurons in the basal forebrain. Science 1982, 215, 1237-1239.
(3) Bartus, R. T.; Dean, R. L.; Beer, B.; Lippa, A. S. The cholinergic
hypothesis of geriatric memory dysfunction. Science 1982, 217,
408-417.
(4) Giacobini, E. Cholinesterase inhibitors do more than inhibit
cholinesterase. In Alzheimer’s Disease: Molecular Biology to
Therapy; Becker, R., Giacobini, E., Eds.; Birkha¨user: Boston,
1997; pp 188-204.
(5) Giacobini, E. Cholinesterase inhibitors: from the Calabar bean
to Alzheimer therapy. In Cholinesterases and Cholinesterase
Inhibitors; Giacobini, E., Ed.; Martin Dunitz: London, 2000; pp
181-226.
(6) Blusztajn, J . K.; Wurtman, R. J . Choline and cholinergic
neurons. Science 1983, 221, 614-621.
(7) Newhouse, P. A.; Sunderland, T.; Tariot, P. N.; Blumhardt, M.
A.; Weingartner, H.; Mellow, A. Intravenous nicotine in Alzhe-
imer’s disease: a pilot study. Psychopharmacology 1988, 95,
171-175.
(8) O’Neill J .; Halgren, E.; Markinovic, K.; Siembieda, D.; Refai, D.;
Fitten, L. J .; Perryman, K.; Fisher, A. Effects of muscarinic and
adrenergic agonism on auditory P300 in the macaque. Physiol.
Behav. 2000, 70, 163-170.
(9) Greig, N. H.; Pei, X. F.; Soncrant, T.; Ingram, D. K.; Brossi, A.
Phenserine and Ring C Hetero-Analogues: Drug Candidates for
Treatment of Alzheimer’s Disease. Med. Res. Rev. 1995, 15,
3-31.
(10) Becker, R. E.; Moriearty, P.; Unni, L.; Vicari, S. Cholinesterase
inhibitors as therapy in Alzheimer’s disease: benefit to risk
considerations in clinical application. In Alzheimer’s Disease:
Molecular Biology to Therapy; Becker, R., Giacobini, E., Ed.;
Birkha¨user: Boston, 1997; pp 257-268.
(11) Long, J . P. Structure-activity relationships of the reversible
anticholinesterase agents. In Cholinesterase and Anticholinest-
erase Agents; Koelle, G. B., Ed.; Handbach der Experimentellen
Pharmakologie; Springer-Verlag: Berlin, 1963; Vol. 15; pp 374-
427.
(12) Usdi, E. Reactions of cholinesterase with substrates, inhibitors
and reactivators. In Anticholinesterase Agents, Karczmar, A. G.,
Ed.; International Encyclopedia of Pharmacology and Thera-
peutics, Sec. 13.; Pergamon Press Ltd.: Oxford, 1970; Vol. 1, pp
47-354.
(13) Quinn, D. M. Acetylcholinesterase: enyme structure, reaction
dynamics and virtual transition states. Chem. Rev. 1987, 87,
955-979.
For the preparation of AChE, freshly collected blood was
centrifuged (10000g, 10 min, 4 °C), and plasma was removed
and diluted 1:125 with 0.1 M Na3PO4 buffer (pH 7.4). For
BChE preparation, whole red blood cells were washed five
times in isotonic saline, lysed in 9 volumes of 0.1 M Na3PO4
buffer (pH 7.4) containing 0.5% Triton-X (Sigma), and then
were diluted with an additional 19 volumes of buffer to a final
dilution of 1:200.
Analysis of anticholinesterase activity, utilizing a 25 µL
sample of each enzyme preparation, was undertaken at their
optimal working pH, 8.0, in 0.1 M Na3PO4 buffer (0.75 mL total
volume). Compounds were preincubated with enzymes (30 min,
room temperature) and then were incubated with their respec-
tive substrates and 5,5′-dithiobis-2-nitrobenzoic acid (25 min,
37 °C). Production of a yellow thionitrobenzoate anion was
measured by spectrophotometer at 412 nm λ. To correct for
nonspecific substrate hydrolysis, aliquots were co-incubated
under conditions of absolute enzyme inhibition (by the addition
of 1 × 10-5 M physostigmine (1)), and the associated alteration
in absorbance was subtracted from that observed through the
concentration range of each test compound. Each agent was
analyzed on four separate occasions and assayed alongside
phenserine (2) and physostigmine (1), as control and external
standards whose activity we have previously reported.23-27
The enzyme activity at each concentration of test compound
was expressed as a percent of activity in the absence of
compound, transformed into a logit format (logit ) %activity/
100 - %activity) and then was plotted as a function of its log
(14) Soreq, H.; Zakut, H. Amplification of butyrylcholinesterase and
acetylcholinesterase genes in normal and tumor tissues: puta-
tive relationship to organophosphorous poisoning. Pharm. Res.
1990, 7, 1-7.
(15) Taylor, P. The cholinesterases. J . Biol. Chem. 1991, 266, 4025-
4028.
(16) Sussman, J . L.; Harel, M.; Farlow, F.; Oefner, C.; Goldman, A.;
Toker, L.; Silman, I. Atomic structure of acetylcholinesterase
from Torpedo Californica: a prototypic acetylcholine-binding
protein. Science 1991, 253, 872-879.
concentration. Inhibitory activity was calculated as an IC50
,
defined as the concentration of compound (nM) required to
inhibit 50% of enzymatic activity, and determined from a
correlation between log concentration and logit activity.
Com p u ter Aid ed Molecu la r Mod elin g. Computer aided
molecular modeling was undertaken using MM2 force field34
and the default parameters. A PDB entry for AChE was used
(Figure 3) in which phenserine was docked manually; there-
after molecular mechanics were applied to optimize the
geometries between it and the active site.
(17) Taylor, P.; Radic, Z. The cholinesterases: from genes to proteins.
Annu. Rev. Pharmacol. Toxicol. 1994, 34, 281-320.
(18) Silman, I.; Harel, M.; Eichler, J .; Sussman, J . L.; Anselmet, A.;
Massoulie, J . Structure-activity relationship in the binding of
reversible inhibitors in the active-site gorge of acetylcholinest-
erase. In Alzheimer’s Disease: Therapeutic Strategies; Becker,
R., Giacobini, E., Ed.; Birkha¨user: Boston, 1994; pp 88-92.
(19) Brufani, M.; Filoccamo, L. Rational design of cholinesterase
inhibitors. In Cholinesterases and Cholinesterase Inhibitors;
Giacobini, E., Ed.; Martin Dunitz: London, 2000, pp 27-46.
(20) Silman, I.; Sussman, J . L. Structural studies of acetylcholinest-
erase. In Cholinesterases and Cholinesterase Inhibitors; Giaco-
bini, E., Ed.; Martin Dunitz: London, 2000; pp 9-26.
(21) Reiner, E.; Radic, Z. Mechanism of action of cholinesterase
inhibitors. In Cholinesterases and Cholinesterase Inhibitors;
Giacobini, E., Ed.; Martin Dunitz: London, 2000; pp 103-120.
Ack n ow led gm en t. The authors are grateful to Noel
F. Whittaker, National Institute of Diabetes and Diges-
tive and Kidney Diseases, NIH, for undertaking the MS
analysis of the compounds reported herein and are
indebted to Dr. Amy Newman, Medicinal Chemistry,

Methyl Analogues of Phenserine
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 24 4071
(22) Brossi, A.; Pei, X. F.; Greig, N. H. Invited Review. Phenserine,
a Novel Anticholinesterase Related to Physostigmine: Total
Synthesis and Biological Properties. Aust. J . Chem. 1996, 49,
171-181.
(23) Brzostowska, M.; He, X. S.; Greig, N. H.; Brossi, A. Selective
Inhibition of Acetyl- and Butyrylcholinesterases by Phenylcar-
bamates of (-)-Eseroline, (-)-(N1)-Noreseroline and Physovenol.
Med. Chem. Res. 1992, 2, 238-246.
(24) Yu, Q. S.; Pei, X. F.; Holloway, H. W.; Greig, N. H.; Brossi, A.
Total syntheses and anticholinesterase activities of (3aS)-N(8)-
norphysostigmine, (3aS)-N(8)-norphenserine, their antipodal
isomers, and other N(8)-substituted analogues. J . Med. Chem.
1997, 40, 2895-2901.
determinants of substrate specificity. J . Biol. Chem. 1993, 268,
17083-17095.
(32) Vellom, D. C.; Radic, Z.; Li, Y.; Pickering, N. A.; Camp, S.; Taylor,
P. Amino acid residues controlling acetylcholinesterase and
butyrylesterase specificity. Biochemistry 1993, 32, 12-17.
(33) Harel, M.; Sussman, J . L.; Krejci, E.; Bon, S.; Chanal, P.;
Massoulie, J .; Silman, I. Conversion of acetylcholinesterase to
butyrylcholinesterase: modeling and mutagenesis. Proc. Natl.
Acad. Sci. U.S.A. 1992, 89, 10827-10831.
(34) Molecular modeling was undertaken with CS Chem 3D 4.0
version, provided by Cambridge Soft Corporation. Energy cal-
culations were based on the compilation of molecular mechanic
methods (MM2 force field programme).
(25) Yu, Q. S.; Greig, N. H.; Holloway, H. W.; Brossi, A. Syntheses
and anticholinesterase activities of (3aS)-N(1), N(8)-bisnor-
phenserine(3aS)-N(1), N(8)-bisnorphysostigmine, their anti-
podal isomers, and other potential metabolites of phenserine.
J . Med. Chem. 1998, 41, 2371-2379.
(26) Pei, X. F.; Yu, Q. S.; Holloway, H. W.; Brossi, A.; Greig, N. H.
Syntheses and biological evaluation of ring-C opened analogues
of the cholinesterase inhibitors physostigmine, phenserine and
cymserine. Med. Chem. Res. 1999, 9, 50-60.
(27) Yu, Q. S.; Holloway, H. W.; Utsuki, T.; Brossi, A.; Greig, N. H.
The synthesis of novel selective inhibitors of butyrylcholinest-
erase based on phenserine for Alzheimer’s disease. J . Med.
Chem. 1999, 42, 1855-1861.
(28) For reviews, see: (a) Patai. The Chemistry of Cyanates and their
Thio Derivatives, Wiley: New York, 1977; Part 2. (b) Twitchett.
Chem. Soc. Rev. 1974, 3, 209-230.
(29) Yu, Q. S.; Atack, J . R.; Rapoport, S. T.; Brossi, A. Carbamate
Analogues of (-)-Physostigmine: In Vitro Inhibition of Acetyl-
and Butlcholinesterases. FEBS Lett. 1988, 234, 127.
(30) Yu, Q. S.; Greig, N. H.; Holloway, H. W.; Flippen-Anderson, F.;
Brossi, A. (-)-(3aS)-Eseroline carbamate (II), a potent cholinest-
erase inhibitor and close analogue of physostigmine: reanalysis.
Med. Chem. Res. 2000, 10, 186-199.
(31) Ordentlich, A.; Barak, D.; Kronan, C.; Flashner, Y.; Leitner, M.;
Segall, Y.; Ariel, N.; Cohen, S.; Velan, B.; Shafferman, A.
Dissection of the human acetylcholinesterase active center
(35) Kamal, M. A.; Greig, N. H.; Alhomida, A. S.; Al-J afari, A. A.
Human acetylcholinesterase inhibition kinetics of the experi-
mental Alzheimer therapeutic, tolserine. Biochem. Pharmacol.
2000, 60, 561-570.
(36) Al-J afari, A. A.; Kamal, M. A.; Greig, N. H.; Alhomida, A. S.;
Perry, E. R. Kinetics of human erythrocyte acetylcholinesterase
inhibition by a novel derivative of physostigmine: phenserine.
Biochem. Biophys. Res. Commun. 1998, 248, 180-185.
(37) Silverman, R. B. The Organic Chemistry of Drug Design and
Drug Action; Academic Press: San Diego, CA, 1992; pp 61-62.
(38) Silverman, R. B. The Organic Chemistry of Drug Design and
Drug Action; Academic Press: San Diego, CA, 1992; p 55.
(39) Stein, H. H.; Lewis, G. J . Noncometitive inhibition of acetylcho-
line by eserine. Biochem. Pharmacol. 1969, 31, 1679-1684.
(40) For discussion, see: (a) Menger. Tetrahedron 1983, 39, 1013-
1040. (b) Liotta et al. J . Am. Chem. Soc. 1984, 106, 4849.
(41) Bruker. SHELXTL-PLUS Version 5.03; Bruker Analytical X-ray
Systems, Madison, WI, 1995.
(42) Cambridge Crystallographic Database, 12 Union Road, Cam-
bridge CB2 1EZ, UK (deposit@ccdc.cam.ac.uk).
(43) Atack, J . R.; Yu, Q. S.; Soncrant, T. T.; Brossi, A. Comparative
inhibitory effect of various physostigmine analogues against
acetyl- and butyrylcholinesterases. J . Pharmacol. Exp. Ther.
1989, 249, 194-202.
J M010080X