Structure-activity relationships for inhibition of human cholinesterases by alkyl amide phenothiazine derivatives

Article abstract of DOI:10.1016/j.bmc.2004.09.059

A number of phenothiazine N-alkyl amide derivatives were synthesized in order to examine the structure-activity relationships for inhibition of human acetylcholinesterase and butyrylcholinesterase. Several lines of evidence indicate that inhibition of butyrylcholinesterase (BuChE) is important in the treatment of certain dementias. Further testing of this concept requires inhibitors that are both BuChE-selective and robust. N-alkyl derivatives (2, 3, 4) of phenothiazine (1) have previously been found to inhibit only BuChE in a mechanism involving π-π interaction between the phenothiazine tricyclic ring system and aromatic residues in the active site gorge. To explore features of phenothiazines that affect the selectivity and potency of BuChE inhibition, a series of N-carbonyl derivatives (5-25) was synthesized and examined for the ability to inhibit cholinesterases. Some of the synthesized derivatives also inhibited AChE through a different mechanism involving carbonyl interaction within the active site gorge. Binding of these derivatives takes place within the gorge, since this inhibition disappears when the molecular volume of the derivative exceeds the estimated active site gorge volume of this enzyme. In contrast, BuChE, with a much larger active site gorge, exhibited inhibition that increased directly with the molecular volumes of the derivatives. This study describes two distinct mechanisms for binding phenothiazine amide derivatives to BuChE and AChE. Molecular volume was found to be an important parameter for BuChE-specific inhibition.

Full text of DOI:10.1016/j.bmc.2004.09.059

Bioorganic & Medicinal Chemistry 13 (2005) 211–222
Structure–activity relationships for inhibition of human
cholinesterases by alkyl amide phenothiazine derivatives
c
c
c
Sultan Darvesh,a,b,c, Robert S. McDonald, Andrea Penwell, Sarah Conrad,
Katherine V. Darvesh,c Diane Mataija,c Geraldine Gomez,c Angela Caines,c
Ryan Walshb and Earl Martinc
*
aDepartment of Medicine (Neurology), Dalhousie University, Halifax, Nova Scotia, Canada
bDepartment of Anatomy and Neurobiology, Dalhousie University, Halifax, Nova Scotia, Canada
cDepartment of Chemistry, Mount Saint Vincent University, Halifax, Nova Scotia, Canada
Received 28 May 2004; accepted 23 September 2004
Available online 22 October 2004
Abstract—Several lines of evidence indicate that inhibition of butyrylcholinesterase (BuChE) is important in the treatment of certain
dementias. Further testing of this concept requires inhibitors that are both BuChE-selective and robust.
N-alkyl derivatives (2, 3, 4) of phenothiazine (1) have previously been found to inhibit only BuChE in a mechanism involving p–p
interaction between the phenothiazine tricyclic ring system and aromatic residues in the active site gorge. To explore features of
phenothiazines that affect the selectivity and potency of BuChE inhibition, a series of N-carbonyl derivatives (5–25) was synthesized
and examined for the ability to inhibit cholinesterases.
Some of the synthesized derivatives also inhibited AChE through a different mechanism involving carbonyl interaction within the
active site gorge. Binding of these derivatives takes place within the gorge, since this inhibition disappears when the molecular vol-
ume of the derivative exceeds the estimated active site gorge volume of this enzyme. In contrast, BuChE, with a much larger active
site gorge, exhibited inhibition that increased directly with the molecular volumes of the derivatives. This study describes two distinct
mechanisms for binding phenothiazine amide derivatives to BuChE and AChE. Molecular volume was found to be an important
parameter for BuChE-specific inhibition.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
important in the treatment of AD.7–11 For example,
inhibition of BuChE leads to an increased level of brain
ACh.12 Furthermore, high levels of BuChE are found
associated with all the neuropathological lesions in
AD.13–15 In addition, BuChE is expressed in neurons
in areas of the brain involved in cognition and behav-
ior.16,17 Also, AChE knockout mice are viable, indicat-
ing that BuChE is able to compensate for the absence
of AChE.18 Finally, all of the drugs that have shown effi-
cacy in the treatment of AD inhibit both AChE and
BuChE, albeit to different degrees.19 For example, the
Ki value for inhibition of AChE by donepezil is
0.024lM while that for inhibition of BuChE is
2.21lM. On the other hand, the Ki value for inhibition
of AChE by tetrahydroaminoacridine is 0.136lM, while
that for inhibition of BuChE is 0.00528lM.19
In AlzheimerÕs disease (AD), loss of cholinergic neuro-
transmission in the brain contributes to the salient
cognitive and behavioral disturbances.1–3 Loss of
cholinergic cells is accompanied by reduced concentra-
tion of the neurotransmitter acetylcholine (ACh), and
of one of its hydrolyzing enzymes, acetylcholinesterase
(AChE EC 3.1.1.7).1,3 In contrast, the level of the related
enzyme, butyrylcholinesterase (BuChE EC 3.1.1.8), is
elevated.4 The use of cholinesterase inhibitors has been
shown to be effective for the treatment of AD.5,6 Clinical
efficacy of these drugs is thought to result from inhibi-
tion of AChE, thus prolonging the availability of ACh
for neurotransmission.5,6 However, several lines of evi-
dence suggest that inhibition of BuChE may also be
In order to further examine the importance of BuChE
inhibition in the treatment of AD, highly selective
BuChE inhibitors are being developed.8,20–22 A number
of phenothiazine derivatives have been shown to inhibit
*
Corresponding author. Address: Room 1308, Camp Hill VeteransÕ
Memorial Building 1, 5955 VeteransÕ Memorial Lane, Halifax, Nova
Scotia, Canada B3H 2E1; e-mail: sultan.darvesh@dal.ca
0968-0896/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2004.09.059

212
S. Darvesh et al. / Bioorg. Med. Chem. 13 (2005) 211–222
cholinesterases, especially BuChE (Fig. 1).23,24 Further-
more, many phenothiazine derivatives, such as chlor-
promazine (3) and ethopropazine (4) have been used
in clinical practice to treat several diseases of the cent-
ral nervous system, including schizophrenia,25 and
tremor.26–28 This implies that a phenothiazine derivative
that is a potent and highly selective BuChE inhibitor
would be expected to be tolerated as a therapeutic agent.
have been described elsewhere.29 In the present study,
all the compounds were fully characterized. All purified
compounds proved to be homogeneous by thin layer
chromatographic analysis, and 1H NMR revealed
all of them to be more than 98% pure. All compounds
1
were fully characterized by IR, and H and 13C NMR
spectroscopy, as well as low and high-resolution mass
spectrometry. Newly synthesized molecules were also
further characterized by elemental analysis.
In an attempt to develop phenothiazine derivatives that
are selective inhibitors of BuChE, we sought to examine
structure–activity relationships of inhibition of choli-
nesterases by phenothiazine derivatives. A series of N-
carbonyl derivatives of phenothiazine were, therefore,
synthesized. In this study, these N-10 alkyl amides were
examined with respect to how structural variations of
the derivatives affected potency and selectivity toward
cholinesterase inhibition. Inhibitory potency was evalu-
ated on the basis of inhibitor constants (Ki values), and
the trends observed were considered with respect to cal-
culated molecular parameters, namely, molecular vol-
ume and the angle of fold of the tricyclic ring system.
Replacement of the hydrogen atom on the nitrogen of
phenothiazine by a carbonyl led to AChE and BuChE
inhibition. The AChE inhibition is observed to be re-
stricted by the molecular volume of the alkyl amide phe-
nothiazine and this restriction can be related to the
smaller volume of the AChE active site gorge. On the
other hand, the potency of BuChE inhibition exhibited
a direct relationship with increasing molecular volumes
of the phenothiazine amide derivatives. This work ex-
plores the structural features of N-10 amide derivatives
of phenothiazine that influence the selectivity and po-
tency of such compounds toward BuChE in particular,
and toward human cholinesterases in general.
2.2. Biological evaluations
Each phenothiazine derivative was evaluated for its abil-
ity to inhibit BuChE and/or AChE using a modifica-
tion30 of EllmanÕs spectrophotometric method.31
Initially, each derivative was examined at the highest
concentration, depending on solubility limits (1–5mM
in 50% aqueous acetonitrile), to determine whether a
compound inhibited BuChE, AChE, or both. Subse-
quently, serial dilution (1:10) profiles were generated
for each compound to determine a range of concentra-
tions suitable for kinetic studies. Lineweaver–Burk plots
were then generated for each compound. A replot of
slopes of these lines versus inhibitor concentration gave
the inhibition constants (Ki; Table 1).
The enzyme kinetic studies with phenothiazine (1),
N-methyl phenothiazine (2), chlorpromazine (3), and
ethopropazine (4) (Fig. 1) showed that they inhibited
BuChE, without any effect on AChE, at concentra-
tions up to 0.167mM (Table 1). The presence of larger
N-alkyl amino groups, as in chlorpromazine (3) and
ethopropazine (4), substantially improved inhibition of
BuChE, over that of (1) or (2), as reflected by smaller
Ki values (Table 1).
Of the 13 noncyclic amide derivatives synthesized (5–
17), seven derivatives, (5–8, 11, 16, 17) inhibited both
BuChE and AChE (Table 1). Notably, the N-acetyl
derivative (5) had an inhibition potency toward BuChE
comparable to the parent, phenothiazine (1). However,
in contrast to the parent phenothiazine (1), derivative
5 inhibited AChE as well, with a Ki value similar to that
for inhibition of BuChE (Table 1). When the number
of carbon atoms attached to the carbonyl group was
increased, for example, acetyl (5) to propanoyl (6) to
butanoyl (7), BuChE inhibition was increased, while
inhibition toward AChE was diminished (Table 1).
Derivatives with five or more carbon atoms attached
to the N-position of phenothiazine showed marked
improvements in BuChE inhibition but did not inhibit
AChE, with the exception of derivative 11 (Table 1).
2. Results
2.1. Synthetic chemistry
The N-substituted phenothiazine amides were prepared
by refluxing phenothiazine with the appropriate acid
chloride (Fig. 2) and an equivalent amount of triethyl-
amine in dichloromethane solution. Purification of the
amides was performed by sequential chemically active
extraction, silica gel column chromatography, and crys-
tallization. Twenty phenothiazine derivatives were syn-
thesized. Some of the derivatives synthesized (Fig. 2)
Of the seven derivatives with cyclic side chains, two (18
and 19), inhibited both AChE and BuChE (Table 1).
Those analogs of this group with six or more carbon
atoms attached to the N-position of phenothiazine
inhibited BuChE only (Table 1).
The inhibition of BuChE by all the phenothiazine alkyl
amides tested was consistently mixed noncompetitive in
nature. In contrast, those derivatives that affected AChE
displayed competitive inhibition of this enzyme.
Figure 1. Structure of phenothiazine and several N-10-alkyl deriva-
tives.

S. Darvesh et al. / Bioorg. Med. Chem. 13 (2005) 211–222
213
Figure 2. Synthetic scheme for the preparation of the phenothiazine derivatives and the structures of compounds synthesized.
Table 1. Inhibition constants (Ki values) and molecular properties of phenothiazine and derivatives
3
Molecular volume (A )
aAChE (Ki lM)
Relative potency
BuChE:AChE
ÔButterflyÕ angle (ꢁ)
˚
Compounds (see Fig. 1)
BuChE (Ki lM)
Phenothiazine (1)b
31.8 10.4
36.9 2.7
3.22 1.82
0.166 0.05
35.8 2.5
25.6 5.4
16.2 3.5
13.2 2.7
10.2 1.5
6.46 1.11
6.33 0.18
6.37 0.44
6.43 1.22
3.59 0.47
8.31 1.07
10.9 3.9
29.6 6.3
23.1 0.8
4.8 0.56
4.21 0.4
1.26 0.22
0.86 0.16
2.22 0.40
1.68 0.23
None
None
212
223
8
3
180
160
162
161
159
160
157
161
158
157
154
160
157
155
160
165
157
147
165
162
159
163
160
159
N-Methyl phenothiazine (2)b
Chlorpromazine (3)b
Ethopropazine (4)b
None
None
340 15
363 17
5
38.9 3.5
46.7 10.3
30.4 5.9
36.2 4.9
None
1.1
1.8
1.9
2.7
244
271 17
296
9
6
7
9
8
299 23
315 12
318 17
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
None
29.5 4.5
None
None
4.7
308
7
327 11
342 17
323 17
358 13
None
None
79.8 10.9
24.1 2.4
44.8 4.1
55.6 24.7
None
7.3
0.8
275
281 13
296
8
1.9
11.6
9
309 23
315 10
343 11
362 18
334 22
365 27
None
None
None
None
a For all the compounds for which the Ki value is listed as ÔnoneÕ, there was no detectable inhibition at concentrations of the phenothiazine derivative
up to 1.67 · 10ꢀ4 M, or to the solubility limit of the compound.
b Reference compounds.

214
S. Darvesh et al. / Bioorg. Med. Chem. 13 (2005) 211–222
In summary, phenothiazine and derivatives with an
alkyl side chain attached directly to the nitrogen of the
ring inhibited BuChE only. Addition of a carbonyl car-
bon at this position of the ring led to inhibition of both
BuChE and AChE, as long as the substituent linked to
the carbonyl was small.
volume of each derivative were calculated to determine
whether either or both parameters were related to selec-
tivity and inhibition potency. When calculating molecu-
lar parameters such as the ÔbutterflyÕ angle and the
molecular volume, there are a number of possible meth-
ods available. One such method that is suited for a series
of mid-sized molecules such as ours, is molecular
mechanics. In this method, the energy of the compound
is expressed as a sum of contributions involving the geo-
metrical parameters of bond-stretching, bond-bending,
torsion, and nonbonded interactions. The specific func-
tion chosen, along with the choice of parameters used,
together describe the force field. The conformer having
the lowest energy was obtained using the MMFF94
force field (PC Spartan Pro, Wavefunction, Inc., 18401
Von Karman, Suite 370, Irvine, California, 92612).
The Ki values for the inhibition of BuChE ranged from
0.166 0.05lM to 36.9 2.7lM. For those derivatives
that also inhibited AChE, the Ki values ranged from
24.1 2.4lM to 79.8 10.9lM. The BuChE-specific
Ki values of some of the synthesized derivatives, such
as cyclopeptyl carbonyl (22) and cyclopeptyl acetyl
(24) (Table 1) compared favorably with drugs used to
treat AlzheimerÕs disease (donepezil; AChE = 0.024
0.0071lM and BuChE = 2.21 0.37lM), galantam-
ine; AChE = 0.520 0.031lM and BuChE = 2.09
0.78lM and tetrahydroaminoacridine; AChE =
0.136 0.027lM and BuChE = 0.00528 0.00037
lM).19
2.4. ‘Butterfly’ angles of the phenothiazine derivatives
Before calculating molecular parameters for all the
derivatives, the molecular mechanics calculation method
we chose was compared to other available methods,
2.3. Calculations of molecular parameters
namely,
Hartree-Fock/STO-3G
and
B3LYP/6-
Previous observations24 have suggested that the large
molecular volume of the phenothiazine derivative etho-
propazine (4) is a significant factor in determining its
BuChE-specificity. That is, the active site gorge of
BuChE, but not of AChE, can accommodate this mole-
cule. On the other hand, the nonplanar nature of the
phenothiazine ring system, which has been described
as having a ÔbutterflyÕ shape32 may also be important.
Variations in the angle between the two aromatic rings,
an angle which will be referred to as the ÔbutterflyÕ angle,
could also play a role in determining inhibitor specificity
and potency. The ÔbutterflyÕ angle and total molecular
31G(d)33–35 using two simpler compounds, namely, phe-
nothiazine (1) and acetyl phenothiazine (5) (Fig. 3),
employing the Gaussian 98 suite of programs.36 ÔButter-
flyÕ angles and computed energies of the derivatives,
whose geometries were optimized in the above manner,
are presented in Table 2.
The B3LYP/6-31G(d) results obtained for phenothia-
zine (1) in the present study (Table 1) were in accord
with the work at the B3LYP level carried out by Palafox
et al.37 They reported a calculated ÔbutterflyÕ angle of
152.2ꢁ for phenothiazine (1), in comparison with crystal
Figure 3. ÔButterflyÕ angles for phenothiazine (1) and acetyl phenothiazine (5).
Table 2. Comparison of calculation methods for ÔbutterflyÕ angles
Method
Energy (hartree)
Phenothiazine 1
Energy (hartree)
Acetylphenothiazine 5
ÔButterflyÕ angle (ꢁ)
Phenothiazine 1
ÔButterflyÕ angle (ꢁ)
Acetylphenothiazine 5
MM
HF/STO-3G
B3LYP/6-31G(d)
—
ꢀ901.0197
ꢀ915.6436
180
165
150
159
147
135
ꢀ1050.8336
ꢀ1068.2925

S. Darvesh et al. / Bioorg. Med. Chem. 13 (2005) 211–222
215
structure results of 158.5ꢁ for the orthorhombic form38
and 153.3ꢁ for the monoclinic form.39 As indicated in
Table 2, while molecular mechanics provided ÔbutterflyÕ
angles that were closer to planar than other methods, it
did reproduce the trend of ÔbutterflyÕ angle decrease
upon N-substitution. For this reason, in all subsequent
calculations on the phenothiazine derivatives, geometry
optimizations based on the best conformer were carried
out at the molecular mechanics level. ÔButterflyÕ angles
for phenothiazine (1) and the analogs considered in this
study varied from approximately 180ꢁ for phenothiazine
(1) itself to approximately 160ꢁ for most compounds, to
a low value of 145ꢁ for the cyclopropyl carbonyl deriv-
ative (18) (Table 1). The nature of the molecular
mechanics method is such that the same calculated Ôbut-
terflyÕ angle will be obtained every time calculations on a
given molecule are carried out. In this sense, precision is
much higher than the significant figures of the reported
value would suggest (Table 1). However, as is noted
above, there are variations of up to 17% between calcul-
ated results and crystallographic data. No consistent
pattern of ÔbutterflyÕ angle trends emerged that would
explain the observed variations in inhibitor potency
and selectivity for these derivatives.
Figure 4. Relationship between molecular volume and potency of
inhibition of butyrylcholinesterase by N-carbonyl phenothiazine
derivatives. ꢀlogKI (pKI) is plotted against calculated molecular
volumes. The linear regression equation was applied to determine the
line of best fit and follows the equation pKI = 0.0118 · volume +
1.4602 (R2 = 0.68).
2.5. Molecular volumes of the phenothiazine derivatives
acetylcholine. In BuChE, however, these residues are re-
placed by two smaller amino acid residues, L286 and
V288, allowing for accommodation of bigger substituent
groups, such as those in butyrylcholine, succinylcholine,
and benzoylcholine, to be held in place for catalysis. A
further important domain is found at the lip of the ac-
tive site gorge and is referred to as the peripheral anionic
site (Fig. 5). The amino acid side chain cluster of this do-
main in AChE consists of Y72, D74, Y124, W286, and
Y341.43 This domain is involved in substrate inhibition
in AChE.42 In BuChE, the amino acid residues at the
equivalent positions are N68, D70, Q119, A277, and
Y332 and are involved in substrate activation in this en-
zyme.50 These substitutions at the periphery of the gorge
in BuChE provide a larger opening for substrate and
inhibitor molecules to enter the active site gorge.
Molecular volume was expected to play a key role in the
inhibitory properties of each molecule as has been sug-
gested for ethopropazine (4).23 One way to represent
the molecular volume is to select a surface of fixed elec-
tron density (0.001e/bohr3, in this case) and to compute
the volume within that isodensity surface.40 Ab initio
methods were required in order to obtain an electron
density surface. Accordingly, single-point Hartree-
Fock/STO-3G calculations were carried out starting
from the atomic coordinates obtained at the molecular
mechanics level. Molecular volumes computed in this
manner, based on an average of five separate calcula-
tions, are summarized in Table 1 and range from
3
3
˚
˚
212A for phenothiazine (1) itself to 365A for the
cyclohexylacetyl derivative (24). This parameter was
found to be the major determinant for BuChE-specific-
ity and inhibitor potency for the N-carbonyl amides
studied here (Fig. 4).
Another domain located within the active site gorge is
the cation-p domain (Fig. 5). This is made up of two
amino acid residues, W86 and Y337, in AChE, with
Y337 being replaced by A328 in BuChE. These residues,
especially the conserved W86 (W82 in BuChE), have
been shown to be important in stabilizing the quater-
nary ammonium moiety of choline esters.41,42 As shown
previously,23 ethopropazine is able to bind to BuChE by
means of p–p interaction between the heterocyclic ring
system and amino acid residues Y332 and F329 in the
active site gorge. The amino acid residue Y337 in AChE
has been shown to interfere with this p–p binding.23
Replacement of this amino acid residue by A328 in
BuChE alleviates the interference with p–p interactions
for phenothiazine and its derivatives. Further stabiliza-
tion of the interaction of BuChE and ethopropazine
(4) appears to involve an electrostatic attraction between
the anionic side chain of D70 in the enzyme and the cat-
ionic side chain nitrogen of the inhibitor.23 These com-
bined ligands provide a robust inhibitor constant for
ethopropazine (4) (Table 1). It is clear that the amide
3. Discussion
Molecular modeling, site-directed mutagenesis, and X-
ray crystallographic studies have revealed a number of
distinct domains associated with the active sites of cho-
linesterases.23,41–46 Within each of these domains, there
are amino acid side chain clusters that control substrate
specificity and inhibitor binding.23,44
In both cholinesterases, a serine–histidine–glutamate
˚
catalytic triad is located near the bottom of a 20A deep
gorge (Fig. 5).41,46–49 Another domain is the Ôacyl pock-
etÕ that holds the alkyl substituent attached to the carb-
onyl of the substrate. In AChE, this pocket has two
bulky phenylalanine residues, F295 and F297, and thus
is only able to accommodate the small methyl moiety of

216
S. Darvesh et al. / Bioorg. Med. Chem. 13 (2005) 211–222
Figure 5. Active-site gorge of cholinesterases. The figure was generated using the Pymol program and crystal structures available in the protein
database.45–48
derivatives do not have the side chain nitrogen as in etho-
propazine (4) to provide additional binding at D70.
This gives some of the amide derivatives inhibitor prop-
erties less robust than ethopropazine (4). On the other
hand, in spite of the cationic side chain nitrogen being
absent, compound (22) which has a molecular volume
comparable to ethopropazine (4) (Table 1) has a Ki
value that is not very different from that of ethoprop-
azine (4) (Table 1).
(electron density in the tricyclic ring system, presence
of a polar carbonyl group, molecular volume) are
changing synchronously. Within the amide series only,
it is clear that BuChE inhibition in general becomes
more potent as the size of the alkyl group or cycloalkyl
ring increases. That the calculated molecular volume is a
good predictor of potency of BuChE inhibition is borne
out by the graph of pKi versus molecular volume (Fig.
4). The reasonably linear correlation between these
parameters illustrates that an increase in the molecular
volume by about 50% (5–22) causes a decrease in Ki
value by about 40-fold. It should be noted that the
volumes of all our derivatives are well below the esti-
The inability of many phenothiazine derivatives to inhi-
bit AChE is also related to the small size of the AChE
3
active site gorge volume (302.31A ) compared to that
˚
3
3
23
22
˚
˚
of BuChE (501.91A ). This undoubtedly plays a key
role in determining specificity toward BuChE inhibition
by certain phenothiazine derivatives such as ethopropa-
zine (4) whose van der Waals volume, calculated to be
mated active site gorge volume for BuChE (502A ).
Two observations provide compelling evidence that phe-
nothiazine derivative enter the active site gorge, namely,
a linear correlation between increasing volume and
inhibitor potency, and the earlier observation of p–p
interaction between the phenothiazine moiety and aro-
matic amino acid residues within the active site gorge.23
3
˚
317.6A , is larger than the volume of the AChE active
site gorge.23 It should be noted that our calculated
volume of ethopropazine (4) differs somewhat from the
3
3
volume reported by Saxena et al.23 (363A vs 317.6A ,
respectively). This discrepancy is due to the different
methods used in obtaining the volumes, but does not
alter the principle that the relative volumes of inhibitor
and active site gorge are important.
˚
˚
All of our amide derivatives show mixed noncompetitive
inhibition with BuChE, implying that both substrate
and inhibitor are able to bind to BuChE simultaneously.
There are several noteworthy differences in the specific-
ity and potency of inhibition of AChE by the phenothi-
azine amides, which point toward a fundamental
difference in the inhibition of this enzyme.
In the present study, molecular volume is also observed
to be a key factor in determining potency of BuChE
inhibition and the selectivity toward BuChE and AChE.
3.1. Butyrylcholinesterase inhibition
3.2. Acetylcholinesterase inhibition
All of the phenothiazine amides reported here inhibit
BuChE, as do the parent (1) and the N-alkyl substituted
phenothiazines (2, 3, 4) (Table 1). Direct comparison be-
tween the 10-alkyl and 10-acyl derivatives is problem-
atic, however, since several molecular parameters
Perhaps the most significant aspect of the present study,
in terms of understanding different modes of interaction
with inhibitors, is the observation of AChE inhibition by
the phenothiazine amides with small alkyl or cycloalkyl
groups attached to the N-carbonyl. To our knowledge,

S. Darvesh et al. / Bioorg. Med. Chem. 13 (2005) 211–222
217
this represents the first report of human AChE inhibition
by phenothiazine derivatives. However, this observation
is not due solely to the small molecular volume of these
derivatives since neither the parent phenothiazine (1)
nor its N-methyl derivative (2), both of which are smaller
than the N-acetyl derivative (Table 1), inhibit AChE.
Clearly, AChE inhibition must be a result of the N-car-
bonyl moiety, either a direct result of the carbonyl group
acting as a hydrogen bond acceptor or the indirect influ-
ence of the electron-withdrawing carbonyl on the elec-
tron density in the tricyclic ring system (vide infra). In
contrast to the BuChE inhibition results, and with the
exception of the 10-chloroacetyl derivative (16), there
appears to be no systematic trend in Ki (AChE) with
structure of the alkyl or cycloalkyl group since all Ki
values fall within the range 40 16lM. The 10-chloro-
acetyl derivative (16), has a Ki value twice that of
the average above. Examination of the spectral data
for this compound (see Experimental) clearly reveals that
the electron-withdrawing nature of the chlorine has a
much larger effect on the proximate carbonyl group than
it does on the more remote phenothiazine ring. The car-
bonyl thus becomes a much poorer hydrogen bond
acceptor and this leads to weaker AChE inhibition. Both
kinetic analysis and spectral studies point to the 10-
carbonyl group as the significant new factor in these
amide derivatives that permit AChE inhibition, possibly
through hydrogen bonding at a side chain group such
as that of Y124, and this, in conjunction with the molec-
ular volume limit for inhibition (vide infra) as well as
the competitive nature of that inhibition, indicates
that these derivatives enter the active site gorge for
binding.
do so competitively, which implies mutually exclusive
interactions. This indicates that the active site gorge of
AChE is blocked to the substrate when the enzyme–
inhibitor complex is formed, and vice versa.
It is clear from this study that phenothiazine amide
derivatives do enter the active site gorge of human cho-
linesterases and that binding to BuChE involves a dis-
tinct mechanism from that involved for inhibition of
AChE. In the former case the phenothiazine tricyclic
ring system provides the major ligand while for AChE
the amide carbonyl provides the point of interaction.
Molecular volume is an important factor in both cases
and provides a limiting parameter for exclusion of
AChE inhibition. Thus, molecular volume of the pheno-
thiazine amide derivatives provides a simple parameter
for obtaining BuChE-specific inhibition and increasing
the potency of inhibition toward this enzyme.
Phenothiazine derivatives are a class of heterocyclic syn-
thetic compounds, many of which, such as chlorproma-
zine (3) and ethopropazine (4) have been primarily used
for treatment of neurological disorders.25–28 Interest-
ingly, ethopropazine (4) has also been associated with
improved cognitive function.51 It is known that pheno-
thiazine derivatives are well tolerated as pharmaceutical
agents. Thus, derivatives of phenothiazine that are
inhibitors of cholinesterases may have a potential for
use in the treatment of dementias. This may be especially
true of many compounds that are both specific and
potent inhibitors of BuChE. Furthermore, these com-
pounds have the potential to clarify further the biologi-
cal functions of BuChE.
Careful examination of the AChE inhibition results in
Table 1 reveals that inhibition disappears dramatically
when the molecular volume of the derivative increases
above a certain critical volume. A fascinating illustra-
tion of this fact occurs with the series of derivatives with
five-carbon side chains. The straight-chain pentanoyl
4. Experimental section
4.1. Materials
Butyrylcholinesterase (BuChE, EC 3.1.1.8) from human
plasma was obtained from Roche. Recombinant human
acetylcholinesterase (AChE, EC 3.1.1.7), acetylthiocho-
line, butyrylthiocholine, chloropromazine, 5,50-dithio-
bis(2-nitrobenzoic acid)(DTNB), ethopropazine, pheno-
thiazine were purchased from Sigma (St. Louis, MO).
Acid chlorides were purchased from Aldrich Chemical
Co. Inc. (Milwaukee, Wi) and used without further puri-
fication. For those unavailable commercially, acid chlo-
rides were prepared from the corresponding acids using
oxalyl chloride according to the procedure below.
3
˚
derivative (9) has a molecular volume of 315A while
the branched-chain derivatives, 3-methylbutanoyl (10)
and dimethylpropanoyl (11), have molecular volumes
3
of 318 and 308A , respectively. Of these three, only
˚
the dimethylpropanoyl derivative (11) was found to in-
hibit AChE. Thus, there appears to be a critical volume
3
limit of about 310A above which no AChE inhibition is
˚
observed, presumably because the derivatives will not fit
into the active site gorge. It is intriguing that this limit is
3
remarkably close to the AChE gorge volume (302A ),
˚
estimated from X-ray crystallographic data.23
4.2. Chemistry
One other molecular parameter, the ÔbutterflyÕ angle,
was investigated regarding its role in determining the
presence of AChE inhibition. The fact that both the
N-methyl (2) and the N-acetyl derivatives have the same
ÔbutterflyÕ angle (ꢁ160ꢁ) while only the latter derivative
inhibits AChE appears to discount the significance of
the ÔbutterflyÕ angle in determining the presence of
AChE inhibition.
4.2.1. General. Melting points were determined on a
Mel-Temp II apparatus and are uncorrected. Thin layer
chromatography was carried out using silica gel sheets
with fluorescent indicator (0.20mm thickness; Mache-
rey–Nagel) and dichloromethane or a dichlorometh-
ane/ethyl acetate mixture as developing solvent. Plates
were visualized using a short wavelength UV lamp.
Infrared spectra were recorded as Nujol mulls between
sodium chloride plates on a Nicolet Model 205 FT-IR
spectrometer. Peak positions were obtained in the ÔPeak
In contrast to the noncompetitive inhibition of BuChE
by these phenothiazine amides, those that inhibit AChE

218
S. Darvesh et al. / Bioorg. Med. Chem. 13 (2005) 211–222
PickÕ mode, and were reproducible within 1–2cmꢀ1
.
139.02, 169.46. EIMS (low res): 241 (M+), 200, 199
(base), 198, 167, 166, 154, 127, 69, 43. EIMS (high res):
M+ (obs) 241.0560; calcd for C14H11NOS = 241.0561.
Nuclear magnetic resonance spectra were recorded at
the Atlantic Region Magnetic Resonance Centre at
Dalhousie University on a Bruker AC-250F spectro-
meter, operating at 250.1MHz for proton and
62.9MHz for carbon. Chemical shifts are reported in
ppm relative to TMS, in CDCl3 solution. Mass spectra
were recorded at Dalhousie University on a CEC 21-
110B spectrometer using electron ionization at 70V
and an appropriate source temperature with samples
being introduced by means of a heatable port probe.
Accurate mass measurements were also made on this
machine operated at a mass resolution of 8000 by com-
puter controlled peak matching to appropriate PFK ref-
erence ions. Mass measurements were routinely within
3ppm of the calculated value. Elemental microanalysis
of all new compounds was performed by Guelph Chem-
ical Laboratories Ltd, Guelph, ON, Canada.
4.4.2. Propanoyl phenothiazine (6). Using 2equiv of acid
chloride, after 18h of reflux, workup, followed by col-
umn chromatography, and recrystallization afforded a
49.6% yield of colorless crystals. Mp 85–87ꢁC (lit. mp
87–88ꢁC53). IR (Nujol): 1677, 1254, 1237, 1181, 1126,
965, 733, 725cmꢀ1 1H NMR (CDCl3): 1.10 (t, J =
.
7.5Hz, 3H), 2.46 (q, J = 7.5Hz, 2H), 7.19 (d of t,
J = 7.6, 1.5Hz, 2H), 7.30 (d of t, J = 7.6, 1.5Hz, 2H),
7.41 (d of d, J = 7.6, 1.5Hz, 2H), 7.49 (d of d, J = 7.6,
1.5Hz, 2H). 13C NMR (CDCl3): 9.57, 27.96, 126.79,
126.98, 127.34, 127.99, 133.28, 138.91, 173.02. EIMS
(low res): 255 (M+), 200, 199 (base), 198, 167, 166,
154, 127, 69, 57, 29. EIMS (high res): M+ (obs)
255.0705; calcd for C15H13NOS = 255.0718.
4.3. Synthesis
4.4.3. Butanoyl phenothiazine (7). Using 2.5equiv of acid
chloride, after a 3h reflux period, workup, followed by
column chromatography, and recrystallization afforded
a 49.4% yield of colorless crystals. Mp 87.5–90.5ꢁC
(lit. mp 94ꢁC54). IR (Nujol): 1678, 1298, 1283, 1249,
4.3.1. General method for N-substituted phenothiazine
amides. A solution containing phenothiazine (5.1mmol),
acid chloride (12.5–25mmol), and triethylamine
(5.0mmol) in dichloromethane (50mL) was refluxed
with stirring until TLC analysis revealed that all pheno-
thiazine was consumed. Reaction periods ranged from
1h to 19 days. The cooled reaction mixture was then
washed successively with 5% aqueous sodium bicarbo-
nate (3 · 40mL), 5% hydrochloric acid (3 · 40mL) fol-
lowed by water (40mL). The solution was dried
(MgSO4), filtered, and solvent removed under vacuum.
The crude solid product was then routinely purified by
column chromatography using silica gel 60 (63–200l)
(Caledon Laboratories Ltd) as adsorbant and dichloro-
methane or a dichloromethane/ethyl acetate mixture as
eluent. Fractions containing only the desired product
were combined, the solvent evaporated, and the solid
recrystallized from petroleum ether/dichloromethane
(2:1).
1180, 768, 755cmꢀ1
.
1H NMR (CDCl3): 0.86 (t,
J = 7.3Hz, 3H), 1.62 (sextet, J = 7.3Hz, 2H), 2.43 (t,
J = 7.3Hz, 2H), 7.20 (d of t, J = 7.6, 1.5Hz, 2H), 7.30
(d of t, J = 7.6, 1.5Hz, 2H), 7.43 (d of d, J = 7.6,
1.5Hz, 2H), 7.50 (broad d, J = 7.6Hz, 2H). 13C NMR
(CDCl3): 13.98, 19.03, 36.53, 127.02, 127.22, 127.61,
128.25, 133.57, 139.18, 172.37. EIMS (low res): 269
(M+), 200, 199, 198 (base), 154, 71, 69, 50, 42. EIMS
(high
res):
C16H15NOS = 269.0874.
M+
(obs) = 269.0871;
calcd
for
4.4.4. Methylpropanoyl phenothiazine (8). Using
2.5equiv of acid chloride, after a 3h reflux period,
workup, followed by column chromatography, and
recrystallization produced a 42.9% yield of colorless
crystals. Mp 147–149ꢁC (lit. mp 143–144ꢁC29). IR (Nu-
jol): 1674, 1458, 1388, 1272, 1251, 769, 756, 734cmꢀ1
.
4.3.2. General method for acid chloride preparation. To a
solution of carboxylic acids (12.5–25mmol) in benzene
(40mL) was added dropwise a 2–3-fold excess of oxalyl
chloride in benzene (10mL). The reaction mixture was
stirred for 1 day at room temperature at which time
the solvent and excess oxalyl chloride were removed
on the rotary evaporator. The residual acid chloride
was used directly in the phenothiazine amide synthesis
(see above).
1H NMR (CDCl3): 1.08 (d, J = 6.7Hz, 6H), 3.05 (septet,
J = 6.7Hz, 1H), 7.20 (d of t, J = 7.6, 1.5Hz, 2H), 7.30 (d
of t, J = 7.6, 1.5Hz, 2H), 7.43 (d of d, J = 7.6, 1.5Hz,
2H), 7.50 (d of d, J = 7.6, 1.5Hz, 2H). 13C NMR
(CDCl3): 19.64, 30.95, 126.73, 126.97, 127.21, 128.02,
133.55, 139.00, 176.59. EIMS (low res): 269 (M+), 200,
199, 198, 154, 127, 71, 69, 45, 43 (base), 41, 39, 27. EIMS
(high res): M+ (obs) 269.0874; calcd for
C16H15NOS = 269.0874.
4.4. Analytical data
4.4.5. Pentanoyl phenothiazine (9). Using 3equiv of acid
chloride, after 4 days of reflux, workup followed by col-
umn chromatography, and recrystallization afforded a
19.2% yield of colorless crystals. Mp 87.5–91.5ꢁC (lit.
mp 93ꢁC55). IR (Nujol): 1673, 1279, 1254, 1178, 1105,
4.4.1. Acetyl phenothiazine (5). Using 3equiv of acid
chloride, after a 1 day reflux period, workup, followed
by column chromatography, and recrystallization
afforded a 57.6% yield of colorless crystals. Mp
202–203ꢁC (lit. mp 204ꢁC52). IR (Nujol): 1671, 1318,
769, 755, 732cmꢀ1
.
1H NMR (CDCl3): 0.81
(t, J = 7.2Hz, 3H), 1.26 (m, 2H), 1.57 (m, 2H), 2.46 (t,
J = 7.5Hz, 2H), 7.22 (d of t, J = 7.6, 1.5Hz, 2H), 7.32
(d of t, J = 7.6, 1.5Hz, 2H), 7.44 (d of d, J = 7.6,
1.5Hz, 2H), 7.50 (br d, J ꢁ 7.6Hz, 2H). 13C NMR
(CDCl3): 13.75, 22.18, 27.43, 33.97, 126.73, 126.93,
127.33, 127.96, 133.32, 138.96, 172.29. EIMS (low res):
1259, 1237, 1029, 1012, 765, 729cmꢀ1
.
1H NMR
(CDCl3) 2.20 (s, 3H), 7.22 (d of t, J = 7.6, 1.5Hz, 2H),
7.32 (d of t, J = 7.6, 1.5Hz, 2H), 7.43 (d of d, J = 7.6,
1.5Hz, 2H), 7.50 (broad d, J ꢁ 7.6Hz, 2H). 13C NMR
(CDCl3): 23.13, 126.91, 127.08, 127.28, 128.04, 133.10,

S. Darvesh et al. / Bioorg. Med. Chem. 13 (2005) 211–222
219
283 (M+), 200, 199 (base), 198, 167, 166, 154, 85. EIMS
(high res): M+ (obs) 283.1037; calcd for
C17H17NOS = 283.1031. Anal. (C17H17NOS): calcd, C
72.05, H 6.05, N 4.94, S 11.32; found, C 72.49, H
6.37, N 4.98, S 11.24.
1307, 1282, 1246, 1179, 1148, 1127, 1029, 768, 753,
1
653cmꢀ1. H NMR (CDCl3): 0.88 (br m, 6H), 1.3–1.8
(br m, 4H), 2.7–2.8 (m, 1H), 7.20 (d of t, J = 7.6,
1.5Hz, 2H), 7.31 (d of t, J = 7.6, 1.5Hz, 2H), 7.43 (d
of d, J = 7.6, 1.5Hz, 2H), 7.50 (br d, J = 7.6Hz, 2H).
13C NMR (CDCl3): 12.15, 26.03, 44.90, 126.88, 127.05,
127.84, 128.21, 133.95, 139.18, 175.69. EIMS (low res):
297 (M+), 200, 199, 198 (base), 171, 166, 154, 127, 99,
71, 69, 55. EIMS (high res): M+ (obs) 297.1196; calcd
for C18H19NOS = 297.1187.
4.4.6. 3-Methylbutanoyl phenothiazine (10). Using
2.5equiv of acid chloride, after a 2.5h reflux period,
workup, followed by column chromatography, and
recrystallization gave a 67.8% yield of slightly off-white
crystals. Mp 157.5–160ꢁC. IR (Nujol): 1676, 1310,
1250, 767, 755cmꢀ1
.
1H NMR (CDCl3): 0.87 (d,
4.4.10. 3,3-Dimethylbutanoyl phenothiazine (14). Using
5equiv of acid chloride, after a 1h reflux period, work-
up, followed by column chromatography, and recrystal-
lization afforded a 45.6% yield of pale green crystals. Mp
125–128ꢁC. IR (Nujol): 1670, 1252, 1233, 1194, 1125,
J = 6.7Hz, 6H), 2.10 (monet, J = 6.7Hz, 1H), 2.38 (d,
J = 6.7Hz, 2H), 7.21 (d of t, J = 7.6, and 1.5Hz, 2H),
7.32 (d of t, J = 7.6, 1.5Hz, 2H), 7.44 (d of d, J = 7.6,
1.5Hz, 2H), 7.51 (broad d, J = 7.6Hz, 2H). 13C NMR
(CDCl3): 22.53, 25.91, 43.07, 126.79, 126.98, 127.47,
128.04, 133.47, 139.04, 171.63. EIMS (low res): 283
(M+), 238, 200, 199 (base), 198, 167, 166, 154, 127, 85,
84, 87, 69, 57. EIMS (high res): M+ (obs) 283.0999; calcd
for C17H17NOS = 283.1031. Anal. (C17H17NOS): calcd,
C 72.05, H 6.05, N 4.94, S 11.32; found, C 71.71, H
6.25, N 4.62, S 11.15.
1
1029, 761, 751, 728cmꢀ1. H NMR (CDCl3): 0.95 (s,
9H), 2.45 (s, 2H), 7.21 (d of t, J = 7.6, 1.5Hz, 2H),
7.32 (d of t, J = 7.6, 1.5Hz, 2H), 7.40 (d of d, J = 7.6,
1.5Hz, 2H), 7.50 (br d, J = 7.6Hz, 2H). 13C NMR
(CDCl3): 29.72, 31.56, 45.41, 126.65, 126.90, 127.56,
128.06, 133.60, 139.13, 170.76. EIMS (low res): 297
(M+), 200, 199 (base), 198, 167, 166, 154, 99, 71, 57.
EIMS (high res): M+ (obs) 297.1186; calcd for
C18H19NOS = 297.1187. Anal. (C18H19NOS): calcd, C
72.69, H 6.44, N 4.71, S 10.78; found, C 72.31, H
6.88, N 4.46, S 10.36.
4.4.7. Dimethylpropanoyl phenothiazine (11). Using
5equiv of acid chloride, after 19 days of reflux, a small
quantity of unreacted phenothiazine remained. Workup
followed by column chromatography, and recrystalliza-
tion afforded 18.1% yield of colorless crystals. Mp 119–
120ꢁC (lit. mp 150–152ꢁC29). IR (Nujol): 1670, 1297,
1285, 1257, 1223, 1152, 1126, 1029, 771, 754, 729,
4.4.11. Heptanoyl phenothiazine (15). Using 2equiv of
acid chloride, after a 5 day reflux period, workup, fol-
lowed by flash chromatography, and recrystallization
afforded a 47.8% yield of colorless crystals. Mp 110.5–
112ꢁC. IR (Nujol): 1682, 1311, 1303, 1256, 1172, 763,
1
667cmꢀ1. H NMR (CDCl3): 1.13 (S, 9H), 7.20 (d of
t, J = 7.6, 1.5Hz, 2H), 7.30 (d of t, J = 7.6, 1.5Hz,
2H), 7.45 (d of d, J = 7.6, 1.5Hz, 2H), 7.60 (d of d,
J = 7.6, 1.5Hz, 2H). 13C NMR (CDCl3): 29.16, 41.64,
126.56, 126.77, 127.76, 128.06, 134.83, 140.78, 178.10.
EIMS (low res): 283 (M+, base), 200, 199, 198, 167,
166, 85, 57, 40, 29. EIMS (high res): M+ (obs)
283.1025; calcd for C17H17NOS = 283.1031.
1
724cmꢀ1. H NMR (CDCl3): 0.82 (t, J = 6.9Hz, 3H),
1.1–1.3 (m, 6H), 1.58 (m, 2H), 2.45 (t, J = 7.5Hz, 2H),
7.20 (d of t, J = 7.6, 1.4Hz, 2H), 7.31 (d of t, J = 7.6,
1.4Hz, 2H), 7.43 (d of d, J = 7.6, 1.4Hz, 2H), 7.50 (br
d, J = 7.6Hz, 2H). 13C NMR (CDCl3): 14.13, 22.54,
25.40, 28.82, 31.57, 34.36, 126.81, 127.01, 127.42,
128.04, 133.41, 139.04, 172.40. EIMS (low res): 311
(M+), 200, 199 (base), 198, 167, 166, 154, 42. EIMS
(high res): M+ (obs) 311.1332; calcd for
C19H21NOS = 311.1344. Anal. (C19H21NOS): calcd, C
73.27, H 6.80, N 4.50, S 10.30; found, C 72.99, H
7.19, N 4.17, S 9.96.
4.4.8. Hexanoyl phenothiazine (12). Using 2.1equiv of
acid chloride, after a 3h reflux period, workup, followed
by column chromatography, and recrystallization af-
forded a 33.4% yield of colorless crystals. Mp 124–
126ꢁC. IR (Nujol): 1682, 1322, 1260, 1170, 1125, 764,
1
758, 724cmꢀ1. H NMR (CDCl3): 0.78–0.85 (m, 3H),
1.17–1.27 (m, 4H), 1.53–1.65 (m, 2H), 2.45 (t,
J = 7.5Hz, 2H), 7.20 (d of t, J = 7.6, 1.5Hz, 2H), 7.31
(d of t, J = 7.6, 1.5Hz, 2H), 7.42 (d of d, J = 7.6,
1.5Hz, 2H), 7.50 (br d, J = 7.6Hz, 2H). 13C NMR
(CDCl3): 14.18, 22.64, 25.31, 31.49, 34.53, 127.02,
127.22, 127.61, 128.25, 133.60, 139.21, 172.60. EIMS
(low res): 297 (M+), 200, 199 (base), 198, 167, 166,
154, 42. EIMS (high res): M+ (obs) 297.1197; calcd for
C18H19NOS = 297.11187. Anal. (C18H19NOS): calcd, C
72.69, H 6.44, N 4.71, S 10.78; found, C 72.26, H
6.81, N 4.40, S 10.53.
4.4.12. Chloroacetyl phenothiazine (16). Using 2.5equiv
of acid chloride, after a 2h reflux period, workup, col-
umn chromatography, and recrystallization gave a
49.5% yield of off-white crystals. Mp 115–117ꢁC (lit.
mp 113.5–114.5ꢁC56). IR (Nujol): 1693, 1672, 1348,
1
1336, 1249, 1170, 1126, 1031, 761, 650cmꢀ1. H NMR
(CDCl3): 4.19 (s, 2H), 7.27 (d of t, J = 7.6, 1.5Hz,
2H), 7.36 (d of t, J = 7.6, 1.5Hz, 2H), 7.47 (d of d,
J = 7.6, 1.5Hz, 2H), 7.59 (d of d, J = 7.6, 1.5Hz, 2H).
13C NMR (CDCl3): 41.97, 126.59, 127.37, 127.45,
128.16, 133.17, 137.88, 165.51. EIMS (low res): 277
M+ (37Cl), 275 M+ (35Cl), 200, 199, 198 (base), 171,
166, 165, 154, 127, 84, 77, 69, 49. EIMS (high res): M+
(obs) 275.0178; calcd for C14H10NOS35Cl = 275.0172.
4.4.9. 2-Ethylbutanoyl phenothiazine (13). Using 2equiv
of acid chloride, after a 1 day reflux period, workup, fol-
lowed by column chromatography, and recrystallization
afforded a 50.7% yield of colorless crystals. Mp 79.5–
81ꢁC (lit. mp 85–86ꢁC29). IR (Nujol): 1665, 1328,
4.4.13. Methoxyacetyl phenothiazine (17). Using
2.2equiv of acid chloride, after a 21h reflux period,

220
S. Darvesh et al. / Bioorg. Med. Chem. 13 (2005) 211–222
workup, column chromatography, and recrystallization
gave rise to a 56.2% yield of slightly off-white crystals.
Mp 135–136.5ꢁC. IR (Nujol): 1685, 1677, 1315, 1301,
calcd, C 73.19, H 5.80, N 4.74, S 10.85; found, C
72.80, H 5.86, N 4.33, S 10.72.
1285, 1257, 1124, 938, 769, 753, 724, 654cmꢀ1 1H
.
4.4.17. Cyclohexyl carbonyl phenothiazine (21). Using
3equiv of acid chloride, after a 1h reflux period, work-
up, followed by column chromatography, and recrystal-
lization afforded a 29.4% yield of off-white crystals. Mp
157–158.5ꢁC (lit. mp 148–150ꢁC29). IR (Nujol): 1677,
1316, 1272, 1252, 1231, 1165, 1122, 1029, 969, 767,
NMR (CDCl3): 3.40 (s, 3H), 4.15 (s, 2H), 7.24 (d of t,
J = 7.6, 1.5Hz, 2H), 7.33 (d of t, J = 7.6, 1.5Hz, 2H),
7.45 (d of d, J = 7.6, 1.5Hz, 2H), 7.53 (br d,
J = 7.6Hz, 2H). 13C NMR (CDCl3): 59.52, 71.19,
126.95, 127.43 (two peaks), 128.28, 133.27, 138.22,
168.32. EIMS (low res): 271 (M+), 199, 198
(base), 154, 127, 69, 45. EIMS (high res): M+ (obs)
271.0668; calcd for C15H13NO2S = 271.0667. Anal.
(C15H13NO2S): calcd, C 66.40, H 4.83, N 5.16, S
11.82; found, C 65.91, H 4.73, N 4.72, S 11.66.
1
753, 658cmꢀ1. H NMR (CDCl3): 1.0–1.8 (overlapping
multitriplets, 10H), 2.72 (t of t, J = 11.3, 3.3Hz, 1H),
7.21 (d of t, J = 7.6, 1.5Hz, 2H), 7.31 (d of t, J = 7.6,
1.5Hz, 2H), 7.43 (d of t, J = 7.6, 1.5Hz, 2H), 7.50 (d
of t, J = 7.6, 1.5Hz, 2H). 13C NMR (CDCl3): 25.45,
25.53, 29.44, 40.95, 126.60, 126.81, 127.08, 127.94,
133.42, 138.92, 175.34. EIMS (low res): 309 (M+), 200,
199 (base), 198, 171, 167, 166, 154, 127, 111, 83, 69,
55. EIMS (high res): M+ (obs) 309.1202; calcd for
C19H19NOS = 309.1187.
4.4.14. Cyclopropyl carbonyl phenothiazine (18). Using
2equiv of acid chloride, after a 3.5day reflux period,
workup, followed by column chromatography, and
recrystallization afforded a 72.2% yield of off-white crys-
tals. Mp 117–119ꢁC (lit. mp 112–114ꢁC29). IR (Nujol):
1672, 1395, 1256, 1173, 762, 752cmꢀ1
.
1H NMR
4.4.18. Cycloheptyl carbonyl phenothiazine (22). Using
3equiv of acid chloride, after a 5 day reflux period,
workup, followed by column chromatography, and
recrystallization afforded a 44.8% yield of pale green
crystals. Mp 100–102.5ꢁC. IR (Nujol): 1674, 1308,
(CDCl3): 0.77–0.89 (m, 2H), 1.16–1.25 (m, 2H), 1.78–
1.91 (m, 1H), 7.21 (d of t, J = 7.6, 1.5Hz, 2H), 7.32 (d
of t, J = 7.6, 1.5Hz, 2H), 7.44 (d of d, J = 7.6, 1.5Hz,
2H), 7.62 (d of d, J = 7.6, 1.5Hz, 2H). 13C NMR
(CDCl3): 9.77, 12.93, 126.59, 126.84, 127.26, 127.92,
132.90, 138.88, 172.31. EIMS (low res): 267 (M+), 200,
199 (base), 198, 154, 69, 40. EIMS (high res): M+
(obs) 267.0710; calcd for C16H13NOS = 267.0718.
1242, 1152, 769, 759, 658cmꢀ1 1H NMR (CDCl3):
.
1.2–2.0 (overlapping multiplets, 12H), 2.91 (m, 1H),
7.20 (d of t, J = 7.6, 1.5Hz, 2H), 7.30 (d of t, J = 7.6,
1.5Hz, 2H), 7.42 (d of d, J = 7.6, 1.5Hz, 2H), 7.49 (br
d, J = 7.6Hz, 2H). 13C NMR (CDCl3): 26.55, 28.35,
31.51, 41.66, 126.73, 126.95, 127.39, 128.05, 133.65,
139.08, 176.73. EIMS (low res): 323 (M+), 199
(base), 198, 171, 166, 154, 127, 97, 69, 55, 41.
EIMS (high res): M+ (obs) 323.1339; calcd for
C20H21NOS = 323.1344. Anal. (C20H21NOS): calcd, C
74.27, H 6.54, N 4.33, S 9.91; found, C 73.82, H 6.53,
N 4.37, S 9.75.
4.4.15. Cyclobutyl carbonyl phenothiazine (19). Using
2.2equiv of acid chloride, after a 1 day reflux period,
workup, followed by column chromatography, and
recrystallization afforded a 68.5% yield of off-white crys-
tals. Mp 145–147.5ꢁC. IR (Nujol): 1670, 1275, 1254,
1170, 765cmꢀ1 1H NMR (CDCl3): 1.7–1.9 (m, 4H),
.
2.2–2.4 (m, 2H), 3.48 (br pentet, J = 8Hz, 1H), 7.19 (d
of t, J = 7.6, 1.5Hz, 2H), 7.30 (d of t, J = 7.6, 1.5Hz,
2H), 7.40 (d of d, J = 7.6, 1.5Hz, 2H), 7.48 (br d,
J = 7.6Hz, 2H). 13C NMR (CDCl3): 17.55, 25.57,
38.17, 126.72, 126.97 (two peaks), 127.79, 133.18,
138.76, 174.24. EIMS (low res): 281 (M+), 200, 199
(base), 198, 197, 167, 154, 127, 83, 69, 54, 38. EIMS
(high res): M+ (obs) 281.0871; calcd for
C17H15NOS = 281.0874. Anal. (C17H15NOS): calcd, C
72.57, H 5.37, N 4.98, S 11.40; found, C 72.10, H
5.49, N 4.75, S 11.12.
4.4.19. Cyclopentylacetyl phenothiazine (23). Using
3equiv of acid chloride, after a 2h reflux period, work-
up, followed by column chromatography, and recrystal-
lization afforded a 50.8% yield of colorless crystals. Mp
152–154ꢁC. IR (Nujol): 1668, 1334, 1315, 1259, 1116,
1029, 765, 730cmꢀ1
.
1H NMR (CDCl3): 0.9–1.2
(m, 2H), 1.4–1.6 (m, 4H), 1.7–1.9 (m, 2H), 2.22 (m,
1H), 2.50 (d, J = 7.0Hz, 2H), 7.21 (d of t, J = 7.6,
1.5Hz, 2H), 7.32 (d of t, J = 7.6, 1.5Hz, 2H), 7.44 (d
of d, J = 7.6, 1.5Hz, 2H), 7.50 (br d, J = 7.6Hz, 2H).
13C NMR (CDCl3): 24.98, 32.51, 36.86, 40.28, 126.80,
127.01, 127.48, 128.05, 133.46, 139.05, 171.94. EIMS
(low res): 309 (M+), 200, 199 (base), 198, 171, 166,
154, 127, 83, 69, 67, 55, 41. EIMS (high res): M+
(obs) = 309.1193; calcd for C19H19NOS = 309.1187.
Anal. (C19H19NOS): calcd, C 73.75, H 6.19, N 4.53, S
10.36; found, C 73.27, H 6.22, N 4.05, S 10.28.
4.4.16. Cyclopentyl carbonyl phenothiazine (20). Using
2.6equiv of acid chloride, after a 1 day reflux period,
workup, followed by column chromatography, and
recrystallization afforded a 65.5% yield of very pale
green crystals. Mp 141–142.5ꢁC. IR (Nujol): 1670,
1583, 1309, 1250, 764, 751cmꢀ1 1H NMR (CDCl3):
.
1.4–1.9 (m, 8H), 3.13 (pentet, J = 7.7Hz, 1H), 7.22 (d
of t, J = 7.6, 1.5Hz, 2H), 7.30 (d of t, J = 7.6, 1.5Hz,
2H), 7.44 (d of d, J = 7.6, 1.5Hz, 2H), 7.53 (d of d,
J = 7.6, 1.5Hz, 2H). 13C NMR (CDCl3): 26.35, 30.96,
41.73, 126.67, 126.93, 127.30, 127.94, 133.59, 139.12,
176.07. EIMS (low res): 295 (M+), 200, 199 (base),
198, 166, 154, 69. EIMS (high res): M+ (obs) 295.1021;
calcd for C18H17NOS = 295.1031. Anal. (C18H17NOS):
4.4.20. Cyclohexylacetyl phenothiazine (24). Using
3equiv of acid chloride, after a 1h reflux period, work-
up, followed by column chromatography, and recrystal-
lization afforded a 64.5% yield of colorless crystals. Mp
173–174ꢁC. IR (Nujol): 1688, 1326, 1259, 1100, 770,
729cmꢀ1
.
1H NMR (CDCl3): 0.7–2.0 (overlapping

S. Darvesh et al. / Bioorg. Med. Chem. 13 (2005) 211–222
221
multiplets, 11H), 2.36 (d, J = 7.0Hz, 2H), 7.21 (d of t,
References and notes
J = 7.6, 1.5Hz, 2H), 7.32 (d of t, J = 7.6, 1.5Hz, 2H),
7.43 (d of d, J = 7.6, 1.5Hz, 2H), 7.51 (broad d,
J = 7.6Hz, 2H). 13C NMR (CDCl3): 26.08, 26.19,
33.03, 35.26, 41.58, 126.91, 127.41, 127.95, 133.37,
139.00, 171.57. EIMS (low res): 323 (M+), 200, 199
(base), 198, 167, 166, 154, 97, 83, 54. EIMS (high res):
M+ (obs) 323.1334; calcd for C20H21NOS = 323.1344.
Anal. (C20H21NOS): calcd, C 74.27, H 6.54, N 4.33, S
9.91; found, C 73.98, H 6.62, N 3.90, S 9.74.
1. Whitehouse, P. J.; Price, D. L.; Clark, A. W.; Coyle, J. T.;
DeLong, M. R. Ann. Neurol. 1981, 10, 122.
2. Bartus, R. T.; Dean, R. L. I.; Beer, B.; Lippa, A. S. Science
1982, 217, 408.
3. Coyle, J. T.; Price, D. L.; DeLong, M. R. Science 1983,
219, 1184.
4. Perry, E. K.; Perry, R. H.; Blessed, G.; Tomlinson, E.
Neuropathol. Appl. Neurobiol. 1978, 4, 273.
5. Doody, R. S. J. Clin. Psychiatry 2003, 64, 11.
6. Rockwood, K.; Darvesh, S.. In Drug Advances; Gray, J.,
Ed.; Remedica Publishing: London, UK, 2003; pp 159–
177.
4.5. Biological evaluation
7. Ballard, C. G. Eur. Neurol. 2002, 47, 64.
8. Greig, N. H.; Lahiri, D. K.; Sambamurti, K. Int.
Psychogeriatr. 2002, 14, 77.
9. Giacobini, E. Butyrylcholinesterase: Its function and
Inhibitors; Martin Dunitz: London and New York, 2003.
10. OÕBrien, K. K.; Saxby, B. K.; Ballard, C. G.; Grace, J.;
Harrington, F.; Ford, G. A.; OÕBrien, J. T.; Swan, A. G.;
Fairbairn, A. F.; Wesnes, K.; del Ser, T.; Edwardson, J.
A.; Morris, C. M.; McKeith, I. G. Pharmacogenetics 2003,
13, 231.
11. Darvesh, S.; Hopkins, D. A.; Geula, C. Nat. Rev.
Neurosci. 2003, 4, 131.
12. Giacobini, E. In Cholinesterases and Cholinesterase Inhib-
itors; Giacobini, E., Ed.; Martin Dunitz: London and New
York, 2000; pp 181–226.
13. Mesulam, M.-M.; Geula, C. Ann. Neurol. 1994, 36, 722.
14. Guillozet, A. L.; Smiley, J. F.; Mash, D. C.; Mesulam,
M.-M. Ann. Neurol. 1997, 42, 909.
15. Lehmann, D. J.; Nagy, Z.; Litchfield, S.; Borja, M. C.;
Smith, A. D. Hum. Genet. 2000, 106, 447.
16. Darvesh, S.; Grantham, D. L.; Hopkins, D. A. J. Comp.
Neurol. 1998, 393, 374.
17. Darvesh, S.; Hopkins, D. A. J. Comp. Neurol. 2003, 463,
25.
4.5.1. Enzyme kinetic studies. The esterase activity of
AChE and BuChE was determined by a modification30
of the method described by Ellman et al.31 Briefly,
2.7mL of buffered DTNB solution (pH8.0), 0.1mL of
AChE (0.03units) or BuChE (0.05units) in 0.005%
aqueous gelatin and 0.1mL of 50% aqueous acetonitrile
or one of the phenothiazine derivative dissolved in this
solvent, in a quartz cuvette of 1cm path length. Serial
dilutions of each compound in 50% acetonitrile were
tested for the ability to inhibit either AChE or BuChE.
The mixture was zeroed at 412nm, and the reaction
was initiated by the addition of acetylthiocholine or
butyrylthiocholine in an aqueous solution at a final con-
centration of 1.6 · 10ꢀ4 M. The reactions were per-
formed at 23ꢁC. The rate of change of absorbance
(DA/min), reflecting the rate of hydrolysis of acetylthio-
choline or butyrylthiocholine, was recorded every 5s for
1min, using a Milton-Roy 1201 UV–visible spectropho-
tometer (Milton-Roy, Ivyland, PA) set at k = 412nm.
These experiments were generally done at least in tripli-
cate and the values averaged. Lineweaver–Burk plots
were generated by using a fixed amount of cholinest-
18. Mesulam, M.-M.; Guillozet, A.; Shaw, P.; Levey, A.;
Duysen, E. G.; Lockridge, O. Neuroscience 2002, 110, 627.
19. Darvesh, S.; Walsh, R.; Kumar, R.; Caines, A.; Roberts,
S.; Magee, D.; Rockwood, K.; Martin, E. Alzheimer Dis.
Assoc. Disord. 2003, 17, 117.
erase and varying amounts of substrate (3 · 10ꢀ5
–
1.6 · 10ꢀ4 M) in the presence or absence of the inhibi-
tors. The re-plot of the slopes of the above double recip-
rocal plots against inhibitor concentration gave the
inhibitor constant (Ki) as the intercept on the x-axis.
20. Yu, Q.; Holloway, H. W.; Utsuki, T.; Brossi, A.; Greig, N.
H. J. Med. Chem. 1999, 42, 1855.
21. Greig, N. H.; Sambamurti, K.; Yu, Q.-S.; Tracy, A. P.;
Holloway, H. W.; Haberman, F.; Brossi, A.; Ingram, D.;
Lahiri, D. In Butyrylcholinesterase: Its Function and
Inhibitors; Giacobini, E., Ed.; Martin Dunitz: London
and New York, 2003, pp 69–90.
22. Savini, L.; Gaeta, A.; Fattorusso, C.; Catalanotti, B.;
Campiani, G.; Chiasserini, L.; Pellerano, C.; Novellino,
E.; McKissic, D.; Saxena, A. J. Med. Chem. 2003, 46, 1.
23. Saxena, A.; Redman, A. M.; Jiang, X.; Lockridge, O.;
Doctor, B. P. Biochemistry 1997, 36, 14642.
24. Debord, J.; Merle, L.; Bollinger, J. C.; Dantoine, T.
J. Enzyme Inhib. Med. Chem. 2002, 17, 197.
25. Baldessarini, R. J.; Tarazi, F. I. In Goodman and GilmanÕs
the Pharmacological Basis of Therapeutics; Hardman, J.
G., Limbird, L. E., Gilman, A. G., Eds., 10th ed.;
McGraw-Hill Companies Inc.: New York, 2001; pp 485–
520.
26. Chouinard, G.; Annable, L.; Ross-Chouinard, A.; Krop-
sky, M. L. J. Clin. Psychiatry 1979, 40, 147.
27. Gardos, G.; Samu, I.; Kallos, M.; Cole, J. O. Psycho-
pharmacology 1980, 71, 29.
28. Fahn, S. Can. J. Neural Sci. 1987, 14, 528.
29. Fischer, H.; Buergin, W. Patent number CH 653 675 AD.
1986.
4.6. Calculation of molecular parameters
Molecular mechanics calculations were carried out using
the MMFF94 force field (PC Spartan Pro, Wavefunc-
tion, Inc., 18401 Von Karman, Suite 370, Irvine, Cali-
fornia, 92612). Hartree-Fock and density functional
theory calculations were carried out using the Gaussian
98 suite of programs.36
Acknowledgements
Canadian Institutes of Health Research, Dalhousie Uni-
versity Internal Medicine Research Fund, Queen Eliza-
beth II Health Science Center Research Fund, Heart
and Stroke Foundation of New Brunswick and Nova
Scotia, the Natural and Sciences and Engineering Re-
search Council of Canada, the Committee on Research
and Publications of Mount Saint Vincent University.
Andrew F. McDonald and Holly A. Cook provided
excellent technical support.

222
S. Darvesh et al. / Bioorg. Med. Chem. 13 (2005) 211–222
30. Darvesh, S.; Kumar, R.; Roberts, S.; Walsh, R.; Martin,
E. Cell. Mol. Neurobiol. 2001, 21, 285.
31. Ellman, G. L.; Courtney, K. D.; Andres, V. J.; Feather-
stone, R. M. Biochem. Pharmacol. 1961, 7, 88.
32. Ragg, E.; Fronza, G.; Mondelli, R. J. Chem. Soc., Perkin
Trans. 2 1983, 9, 1289.
33. Becke, A. D. J. Chem. Phys. 1993, 98, 1372.
42. Shafferman, A.; Velan, B.; Ordentlich, A.; Kronman, C.;
Grosfeld, H.; Leitner, M.; Flashner, Y.; Cohen, S.; Barak,
D.; Ariel, N. EMBO J. 1992, 11, 3561.
´
43. Vellom, D. C.; Radic, Z.; Li, Y.; Pickering, N. A.; Camp,
S.; Taylor, P. Biochemistry 1993, 32, 12.
´
44. Radic, Z.; Pickering, N. A.; Vellom, D. C.; Camp, S.;
Taylor, P. Biochemistry 1993, 32, 12074.
45. Gnatt, A.; Loewenstein, Y.; Yaron, A.; Schwarz, M.;
Soreq, H. J. Neurochem. 1994, 62, 749.
34. Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
35. Lee, C.; Yang, W.; Parr, G. Phys. Rev. B 1988, 37, 785.
36. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G.
E.; Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.;
Montgomery, Jr., J. A.; Stratmann, R. E.; Burant, J. C.;
Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin, K. N.;
Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi,
M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.;
Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.;
Cui, Q.; Morokuma, K.; Rega, N.; Salvador, P.; Dannen-
berg, J. J.; Malick, D. K.; Rabuck, A. D.; Raghavachari,
K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Baboul,
A. G.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.;
Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.;
Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara,
A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-
Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 98,
Revision A.11.2. Gaussian, Inc., Pittsburgh PA, 2001.
37. Palafox, M. A.; Gil, M.; Nunez, J. L.; Tardajos, G. Int. J.
Quantum Chem. 2002, 89, 147.
46. Nicolet, Y.; Lockridge, O.; Masson, P.; Fontecilla-Camps,
J. C.; Nachon, F. J. Biol. Chem. 2003, 278, 41141, PDB
ID: 1POP.
47. Berman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.;
Bhat, T. N.; Weissig, H.; Shindyalov, I. N.; Bourne, P. E.
The protein data bank. Nucleic Acids Res. 2000, 28, 235
(http://www.pdb.org/).
48. Kryger, G.; Harel, M.; Giles, K.; Toker, L.; Velan, B.;
Lazar, A.; Kronman, C.; Barak, D.; Ariel, N.; Shafferman,
A.; Silman, I.; Sussman, J. L. Acta Crystallogr., Sect. D
2000, 56, 1385, PDB ID: 1B41.
49. DeLano, W. L. The PyMOL Molecular Graphics System.
DeLano Scientific, San Carlos, CA, USA. 2002 (http://
www.pymol.org).
50. Masson, P.; Froment, M. T.; Bartels, C. F.; Lockridge, O.
Eur. J. Biochem. 1996, 235, 36.
51. Darvesh, S.; MacKnight, C.; Rockwood, K. Int. Psycho-
geriatr. 2001, 13, 461.
52. Roseboom, H.; Forch, A. D. J. Pharm. Sci. 1979, 68,
515.
53. Siska, M.; Szporny, L.; Clauder, O. Acta Pharm. Hung.
1961, 31, 91.
38. McDowell, J. J. H. Acta Crystallogr., Sect. B 1976, 32, 5.
39. Bell, J. D.; Blount, J. F.; Briscoe, O. V.; Freeman, H. C.
Chem. Commun. 1968, 1656.
40. Frisch, Æ.; Frisch, M. J. Gaussian 98 UserÕs Reference,
2nd ed.; Gaussian Inc.: Pittsburg, PA, 1999.
41. Sussman, J. L.; Harel, M.; Frolow, F.; Oefner, C.;
Goldman, A.; Toker, L.; Silman, I. Science 1991, 253, 872.
54. Cauquil, G.; Casadevall, A. Compt. Rend. 1953, 236, 1569.
55. Stoss, P.; Satzinger, G. Chem. Ber. 1978, 111, 1453.
56. Gilman, H.; Nelson, R. D. J. Amer. Chem. Soc. 1953, 75,
5422.