A structure-based design approach to the development of novel, reversible AChE inhibitors

Article abstract of DOI:10.1021/jm010826r

Chimeras of tacrine and m-(N,N,N-Trimethylammonio)trifluoroacetophenone (1) were designed as novel, reversible inhibitors of acetylcholinesterase. On the basis of the X-ray structure of the apoenzyme, a molecular modeling study determined the favored attachment positions on the 4-aminoquinoline ring (position 3 and the 4-amino nitrogen) and the favored lengths of a polymethylene link between the two moieties (respectively 5-6 and 4-5 sp³ atoms). Seven compounds matching these criteria were synthesized, and their inhibitory potencies were determined to be in the low nanomolar range. Activity data for close analogues lacking some of the postulated key features showed that our predictions were correct. In addition, a subsequent crystal structure of acetylcholinesterase complexed with the most active compound 27 was in good agreement with our model. The design strategy is therefore validated and can now be developed further.

Full text of DOI:10.1021/jm010826r

J . Med. Chem. 2001, 44, 3203-3215
3203
Articles
A Str u ctu r e-Ba sed Design Ap p r oa ch to th e Develop m en t of Novel, Rever sible
ACh E In h ibitor s
Caroline Doucet-Personeni,† Philip D. Bentley,† Rodney J . Fletcher,† Adrian Kinkaid,† Gitay Kryger,‡
Bernard Pirard,† Anne Taylor,† Robin Taylor,§ J ohn Taylor,† Russell Viner,*,† Israel Silman,| J oel L. Sussman,‡
Harry M. Greenblatt,‡ and Terence Lewis†
Syngenta, J ealott’s Hill Research Station, Bracknell, Berkshire RG42 6EY, U.K., Departments of Structural Biology and
Neurobiology, Weizmann Institute of Science, Rehovot 76100, Israel, and Cambridge Crystallographic Data Centre,
12 Union Road, Cambridge CB2 1EZ, U.K.
Received J anuary 19, 2001
Chimeras of tacrine and m-(N,N,N-Trimethylammonio)trifluoroacetophenone (1) were designed
as novel, reversible inhibitors of acetylcholinesterase. On the basis of the X-ray structure of
the apoenzyme, a molecular modeling study determined the favored attachment positions on
the 4-aminoquinoline ring (position 3 and the 4-amino nitrogen) and the favored lengths of a
polymethylene link between the two moieties (respectively 5-6 and 4-5 sp3 atoms). Seven
compounds matching these criteria were synthesized, and their inhibitory potencies were
determined to be in the low nanomolar range. Activity data for close analogues lacking some
of the postulated key features showed that our predictions were correct. In addition, a
subsequent crystal structure of acetylcholinesterase complexed with the most active compound
27 was in good agreement with our model. The design strategy is therefore validated and can
now be developed further.
In tr od u ction
While the essential biological role of acetylcholin-
esterase (AChE) in the mediation of vertebrate and
invertebrate nervous transmission has been known for
more than 80 years, this remarkable enzyme continues
to be the focus of much interest in pharmaceutical and
agrochemical circles. Medicinally, it has been targeted
in treatments for Alzheimer’s disease, myasthenia
gravis, and glaucoma, and in the recovery of victims of
nerve agent exposure.1-4 Agrochemically, AChE is the
most widely exploited insecticide target: all the com-
mercial organophosphate and carbamate agents are
believed to exert their effects through irreversible
inhibition of the enzyme.5 More recently, researchers
in these areas have focused on two key objectives,
F igu r e 1. Examples of commercially important AChE inhibi-
tors.
discovering (a) selective inhibitors of the mammalian
enzyme for the effective treatment of Alzheimer’s dis-
ease, thereby avoiding the undesirable side effects of
the established drug, tacrine6 (Figure 1), and (b) selective
inhibitors of insect enzymes which would avoid the
problems of both acute and chronic toxic effects of
organophosphate (e.g., parathion) and carbamate (e.g.,
carbofuran) insecticides on human and other vertebrate
life.7
tous family of enzymes, the serine hydrolases (e.g.,
esterases, ligases, amidases, proteases), which share the
same catalytic mechanism. Any new commercial product
would require selectivity of this nature to be both safe
and successful.
When these goals were first tackled in the 1970s and
1980s, little was known about the structure of AChE.
Attempts at structure-based inhibitor design at this
stage were highly speculative, and successful inhibitor
discovery was driven more by “random” screening and
exhaustive analogue investigation than by an under-
standing of inhibitor-protein interactions. Despite these
challenges, the approach afforded the first truly selec-
tive, commercially viable AChE inhibitor in the form of
The emphasis in both of these areas is on obtaining
selectivity for AChE over other members of the ubiqui-
* To whom correspondence should be addressed: E-mail:
Russell.Viner@syngenta.com. Phone: +44 1344 414180. Fax: +44 1344
455629.
† Syngenta.
‡ Department of Structural Biology, Weizmann Institute of Science.
§ Cambridge Crystallographic Data Centre.
| Department of Neurobiology, Weizmann Institute of Science.
10.1021/jm010826r CCC: $20.00 © 2001 American Chemical Society
Published on Web 08/28/2001

3204 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20
Doucet-Personeni et al.
F igu r e 3. Binding mode of trifluoromethyl ketones at the
active site of serine hydrolase enzymes, compared with ace-
tylcholine binding to AChE.
F igu r e 2. Tacrine-Torpedo AChE complex X-ray structure
showing key interactions and the proximity of active site serine
(Ser200). The tacrine carbon atoms are magenta; all other
atoms are colored by atom type. The hydrogen bonds with the
corresponding acceptor-donor distances are in green.
the anti-Alzheimer drug Aricept (E2020) in 1995 (Figure
1).8 The elucidation of the crystal structure of Torpedo
californica AChE (TcAChE) in 19919 and subsequent
enzyme-inhibitor complexes10-18 has provided research-
ers with an invaluable tool for studying the structure
and function of this enzyme.
The unusual location of the catalytic site at the base
of a deep, narrow gorge lined with aromatic residues
has prompted extensive investigation across a diverse
range of topics, including the mechanism of substrate
binding and turnover,19,20 the role of cation-π and π-π
interactions,21 the possible contribution of electrostatics
in catalysis and noncholinergic functions,22 and of course
the attempted rationalization of known inhibitor bind-
ing and novel inhibitor design.23-29
Recently post-hoc rationalization of the binding mode
of E2020 has been possible by analysis of the crystal
structure of the inhibitor-TcAChE complex,14,30 which
has highlighted the importance of aromatic interactions
between protein and inhibitor which were largely
overlooked in the original predictions.
F igu r e 4. Model of the “Quinn inhibitor”-TcAChE complex
X-ray structure showing key inhibitor-protein interactions.
The inhibitor carbon atoms are magenta; all other atoms are
colored by atom type.
Tacrine forms only one direct H-bond with the protein,
with the overall binding mode apparently dominated by
aromatic and hydrophobic contacts. This mode of bind-
ing has since been shown to be important for a variety
of inhibitor and substrate types. It is also worth noting
that π-π interactions are not usually favorable except
when a charged ring system (such as the acridinium
system) is involved. The inhibitor occupies only a part
of the active site, while the immediate vicinity of the
catalytic serine 200 is occupied by water.
Of particular interest are inhibitors which act as
transition-state mimics, such as trifluoromethyl ke-
tones. A number of examples of such compounds acting
as inhibitors of serine hydrolases, including lipases,32
proteases,33-35 juvenile hormone esterase,36 and AChE36,37
are known. The substrate-like inhibitor 1 of AChE
(Figure 3) studied by Quinn’s group is of this class and
is one of the most potent reversible inhibitors of AChE
reported in the literature. Sussman, Silman, et al.
confirmed the nature of the binding mode of compound
1 complexed with Torpedo AChE by X-ray analysis
(Figure 4).10 The active site serine covalently bonds to
the activated carbonyl of the trifluoroacetyl group, while
the quaternary ammonium group interacts favorably
with the adjacent aromatic residues Trp84 and Phe330
among others.
Ba ck gr ou n d
We have been studying the structure and function of
AChE to design novel, reversible active site inhibitors,
with a view to developing selective insecticides which
would function efficiently without the potentially harm-
ful side effects of more traditional agrochemicals. The
current work was facilitated by the availability of
structural information of the known anti-Alzheimer
drug tacrine bound to the active site of Torpedo AChE
(Figure 2).31 An analysis of the complex highlighted a
number of key features of the binding mode.
The inhibitor binds in its protonated form and par-
ticipates in four different kinds of intermolecular inter-
actions:
(1) H-bonding to both the protein and “structural”
water molecules,
(2) aromatic (π-π stacking and edge-to-face) interac-
tions, mainly with W84 and F330,
(3) cationic nitrogen-aromatic interactions (N+-π),
mainly with W84 and F330, and
Molecu la r Design
(4) hydrophobic contacts involving non-H-bonding
groups of inhibitor and protein.
The overall philosophy guiding our design was to try
to find a single molecule that could exhibit the trifluo-

Development of Reversible AChE Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20 3205
position, and the extent to which the chain was folded.
Twelve candidate molecules were modeled using chain
lengths in this range and the four possible attachment
positions, and appending a hydrated trifluoroacetyl
group (C(OH)2CF3) to the end of the chain. A confor-
mational search was carried out for each candidate
molecule to identify reasonably low energy conforma-
tions where the distances in the candidate molecule best
matched the distances in the template. The conforma-
tions found were then energy minimized, but with the
nine distances constrained to stay near their template
values. The conformations were then overlaid on the
template and inspected in the active site. For each
candidate molecule, the best conformation was identi-
fied, attempting to take into account the goodness of fit
to the template, steric clashes with the protein, confor-
mational strain in the ligand, and solvent accessibility
of H-bonding groups in the active site. The same criteria
were used to rank the molecules against each other. The
favored molecules selected by this process were those
with chains linked to position 3 (chains of type -CH2-O-
(CH2)n, n ) 3 and 4) and to the 4-amino nitrogen (chains
of type -(CH2)m, m ) 4 and 5).
Synthetic targets were then designed on the basis of
these preferred attachment points and chain lengths.
Further rings and heteroatoms were introduced into the
chains to reduce flexibility (entropic benefit) and ease
the synthesis. These introductions were made so as to
preserve as much as possible the required conformations
of the chains, and to fit within the space constraints of
the active site.
The final list of novel putative inhibitor structures,
A-D, selected for synthesis is shown in Figure 6.
Precedents suggest that 8-substitution of tacrine tended
to increase binding so an equivalent halogen substituent
was usually retained in our target molecules.38,39
F igu r e 5. Distances and linkage positions used in the design
template.
roacetyl binding mode and the Trp84 ring stacking
binding mode simultaneously. Since both of these bind-
ing modes are observed for potent inhibitors of AChE,
it was hoped that combining these two effects would lead
to even higher affinity ligands with greater specificity
for AChE. The binding sites for the two modes appeared,
by inspection of the crystal structure of the apoenzyme,
to be sufficiently separate to allow both to be occupied
simultaneously, while at the same time being close
enough to make a bridging strategy attractive. The aim
was therefore to link a 4-aminoquinoline ring (as found
in the tacrine ring system) to a trifluoroacetyl group so
as to span the two sites. The nature of the linkage, and
its attachment position to the 4-aminoquinoline ring,
remained to be determined.
A manual docking of tacrine into the crystal structure
of the apoenzyme (PDB code 1ACE)9 indicated that
there was insufficient space to accommodate a chain at
positions 6, 7, and 8 of the 4-aminoquinoline ring
(Figure 5). However, positions 2, 3, and 5 and the
4-amino nitrogen all seemed to offer opportunities for
attaching a linking group.
A computational method was used to help choose the
best attachment position for the linker, as well as the
correct length. A template was defined by an amino-
quinoline ring stacked with Trp84 and a detached
trifluoroacetyl group docked into the oxyanion hole. This
template was characterized by measuring the nine
distances shown in Figure 5.
Preliminary modeling work using the template sug-
gested that the two groups could probably be linked by
a chain of 3-6 sp3 atoms, depending on the attachment
Ch em istr y
From a synthetic standpoint the tacrine-like tricyclic
targets A seemed most readily accessible. A successful
route to this class of compounds is outlined in Scheme
1. Alkylation of the known tacrine derivative 2 with
F igu r e 6. Structural types of inhibitors suggested by the computational approach, and some resulting synthetic targets.

3206 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20
Doucet-Personeni et al.
Sch em e 1. Synthesis of N-Alkylated Tacrine Analogues (Figure 6, Target A)a
a
Reagents: (a) (i) NaH, DMF, 50 °C, 1 h, (ii) 2-bromomethylthiophene, 60 °C, 2 h; (b) (i) LDA, THF, -78 °C, 1.5 h, (ii) N-methyl-N-
methoxytrifluoroacetamide, -78 to 0 °C, 0.5 h; (c) Raney Ni, THF, 4.5 h.
(bromomethyl)thiophene gave the N-substituted tacrine
3. This compound was trifluoroacetylated with N-
methyl-N-methoxytrifluoroacetamide to give 4, which
was hydrogenated to give the desired trifluoromethyl
ketone 5a together with the dechlorinated derivative 5b.
On exposure to atmospheric moisture these materials
were rapidly converted to their hydrates 6a and 6b,
which were then separated chromatographically. Previ-
ous studies have shown that although alkyl trifluoro-
methyl ketones readily hydrate to form geminal diols
they are in equilibrium, in solution, with sufficient
ketone to act as potent serine hydrolase inhibitors.40
The aminoquinoline targets B and C required a
different synthetic approach. By reacting the requisite
aminobenzonitrile 7a -c with a â-keto ester, using well-
known Lewis acid-mediated cyclization conditions,41 the
aminoquinolines 8a -c were efficiently formed (Scheme
2). They were then reduced to 9a -c and chlorinated to
give 10a -c. These chloromethyl intermediates were
reacted with phthalimide to give, after hydrazine cleav-
age, the amines 11a and 11b. Alkylation of these amines
proved unsuccessful, using a variety of standard condi-
tions. An alternative approach, as outlined in Scheme
2, relied upon the generation of a trifluoromethylated
lactone 12, a derivative of succinic anhydride, by reac-
tion with Rupert’s reagent.42 While purification of 12
proved troublesome, the crude material could be reacted
directly with amines 11a and 11b. Analysis of the
resulting products revealed them to be the lactones 13a
and 13b, rather than the desired open chain target
amides of type B.
gave only the substituted tetrahydrofuran 17. This,
presumably, resulted from preferential attack of the
thiolate of 14 on the ketone, followed by cyclodebromi-
nation.
We set out to make the final set of targets, the
substituted aminoquinolines D, from aminoquinoline
alcohol intermediates 9a -c. However, alkylation with
benzyl bromides proved ineffective; little reaction oc-
curred under standard basic or phase-transfer condi-
tions. However, benzyl ethers, such as 18, could be made
from the chloromethylaminoquinoline 10a by reaction
with excess benzyl alcohols at high temperature (Scheme
4). The key trifluoroacetylbenzyl alcohol 22 was ob-
tained from the 3-bromobenzyl alcohol (19), and this
similarly yielded the target compounds 23a -c with
10a -c.
The inaccessibility of the required benzyl thiols
dictated a different approach for the corresponding
sulfur-linked analogues, shown in Scheme 5. 3-Trifluo-
roacetylbenzyl bromide (26) was made in two straight-
forward stages from 3-bromotoluene (24). The thiol 14
was deprotonated with sodium hydride, and the result-
ant thiolate reacted, on heating with excess benzyl
bromides such as 26, to afford thioethers. Thus, the type
D target aminoquinoline 27 was made, together with
other analogues, such as related ketone 29. 27 was
subsequently reduced with sodium borohydride to afford
the trifluoromethyl alcohol 28.
In contrast to aliphatic trifluoromethyl ketone ex-
amples, the aryl trifluoromethyl ketones 23a -c and 27
were less susceptible to hydration and were isolated in
the ketone forms, from chloroform. Although they
hydrated substantially in water, a significant proportion
of the keto form was maintained. The unfluorinated
ketone 29 showed no tendency to hydrate.
The thiol 14, required for the S-linked target C, was
made by reaction of 10a with thiourea, followed by base
hydrolysis (Scheme 3). The 5-bromo-1,1,1-trifluoro-2-
pentanone (16) was obtained in two stages from butyro-
lactone via 15.42 Alkylation of 14 with 16 unexpectedly

Development of Reversible AChE Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20 3207
Sch em e 2. Synthesis of N-Alkylated Aminoquinolines (Figure 6, Target B)a
a
Reagents: (a) (i) SnCl4, PhMe, room temperature to reflux, 5 h, (ii) NaOH, 0-5 °C, 0.5 h; (b) LiAlH4, THF, 4 h; (c) SOCl2, reflux,
1 h; (d) (i) potasium phthalimide, dioxane, (ii) NH2NH2‚H2O, dioxane; (e) CF3TMS; (f) TBAF.
Sch em e 3. Synthesis of S-Alkylated Aminoquinolines (Figure 6, Target C)a
a
Reagents: (a) (i) thiourea, DMF, 100 °C, 2 h, (ii) NaOH, reflux, 2 h; (b) CF3TMS; (c) concentrated HBr; (d) (i) NaH, DMF, room
temperature to 50 °C, 0.5 h, (ii) 15 added at room temperature, heat 60 °C, 2 h.
Str u ctu r e-Activity Rela tion sh ip s
with the anticipated direct interaction of this group with
the enzyme. However, since none of the inhibitors of
type A was more active than the starting point 2, we
moved onto derivatives of aminoquinolines (targets B,
C, and D).
Some key intermediates and all final products of the
synthetic sequences described above were tested on
Torpedo AChE for inhibitory activity (Table 1).
The tacrine analogue 2 was chosen as a starting point
for the synthesis of target A because it contains both
the seemingly favored halogen substituent and the fused
seven-membered ring.38,39 The activities of compounds
6a and 6b, which conform to this target, compare well
with those of tacrine and 2. The IC50 of the more rigid
intermediate 4 is in the same range of activity, although
modeling had suggested that it would fit the active site
less well. The poorer fit to the active site may be offset
by the entropic benefit of greater rigidity. In comparison
with 3, the trifluoroacetyl group of 4 dramatically
enhances activity, by about 60-fold, which is consistent
Simple 4-aminoquinolines are relatively poor inhibi-
tors compared to tacrine, and this activity is not
significantly enhanced by the introduction of functional
side chains, such as those present in the intermediate
8a , 9a , or 11a (i.e., carboxyl ester, OH, NH2). These
compounds tend to have IC50 values in the micromolar
range, as do the more functional unexpected products
13a ,b and 17.
Unlike the simpler aminoquinolines, the target D
compounds 23a -c and 27 are very potent, all showing
IC50 values in the low nanomolar range. The nature of

3208 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20
Doucet-Personeni et al.
Sch em e 4. Synthesis of 3-Benzyloxyaminoquinolines (Figure 6, Target D)a
a
Reagents: (a) TBDMSCl, imidazole, DMF; (b) (i) BuLi, Et2O, 0 °C, (ii) N-trifluoroacetylpiperidine, 0 °C to room temperature, 0.5 h;
(c) HCl/MeOH; (d) excess 22, 120 °C, 4 h; (e) excess benzyl alcohol, 10a , 120 °C, 4 h.
Sch em e 5. Synthesis of 3-Benzylthioaminoquinolines (Figure 6, Target D)a
a
Reagents: (a) (i) Mg, Et2O, reflux, (ii) trifluoroacetic acid, reflux; (b) NBS, CCl4, reflux, 3 h; (c) (i) 14, NaH, DMF, 60 °C, 1 h, (ii) 26,
60 °C, 2 h; (d) NaBH4, THF, room temperature, 1 h; (e) 14, NaH, DMF, 60 °C, 1 h, (ii) benzyl bromide, 60 °C, 2 h.
the heteroatom linking group makes no difference to the
activity (23a versus 27) and neither does the nature of
the atom at the 5-position (F, Cl, or H, for 23a -c).
Close analogues that lack the full trifluoromethyl
ketone moiety (18), or have been reduced to the second-
ary alcohol 28, or those that contain hydrogen atoms
instead of fluorines (29) are, respectively, 500-, 3700-,
and 5000-fold less active than their parent compound.
These observations are consistent with the type D target
compounds interacting, as we postulated, with the
active site serine 200 to form a transition-state ana-
logue, of the sort shown in Figure 3. Indeed, although
the secondary alcohol 28 could occupy both the oxyanion
hole and methyl binding pockets, it is incapable of
forming a covalent bond to serine 200, while the
unfluorinated ketone 29 is, presumably, insufficiently
reactive to act as a nucleophilic trap for the serine.
The study of compounds 18, 28, and 29 as well as the
intermediates 8a , 9a , and 11a thus showed the impor-
tance of the COCF3 group as a key moiety in the design
of our inhibitors.
However, the potent activity of 23a -c and 27 is not
only due to the trifluoromethyl ketone as compound 26,
which lacks the aminoquinoline moiety, is 100-fold less
active than 23a -c and 27.

Development of Reversible AChE Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20 3209
Ta ble 1. In Vitro IC50 Values for the Inhibition of Torpedo AChE
compd no.
IC50 (nM)
structure in
compd no.
IC50 (nM)
structure in
tacrine
2
3
4
6a
6b
8a
9a
11a
13a
8.2
2.4
740
12
36
14
Figure 1
13b
17
18
23a
23b
23c
26
27
28
29
7.2 × 103
18 × 103
2.0 × 103
3.8
5.1
3.0
Scheme 2
Scheme 3
Scheme 4
Scheme 4
Scheme 4
Scheme 4
Scheme 5
Scheme 5
Scheme 5
Scheme 5
Scheme 1
Scheme 1
Scheme 1
Scheme 1
Scheme 1
Scheme 2
Scheme 2
Scheme 2
Scheme 2
2.6 × 103
360
3.0
2.6 × 103
19 × 103
14 × 103
11 × 103
15 × 103
F igu r e 8. Simulated annealing omit map for the inhibitor
27 contoured at 4σ (purple density), calculated by CNS. The
protein residues are rendered as dark green stick models, and
the inhibitor is rendered as a light green ball stick model.
Carbon atoms are colored green, oxygen red, nitrogen dark
blue, and fluorine cyan. Gly118, Gly119, and Ala201, whose
nitrogen atoms make up the oxyanion hole, are not labeled
for clarity. Two key water molecules are shown as red spheres.
The contact between the amine nitrogen of the inhibitor and
a water molecule shown partialy behind the electron density
is 2.8 Å long. The orientation is similar to that shown in Figure
7. The picture was made with XtalView/Raster3D.51,52
F igu r e 7. Schematic representation of binding of 27 to
TcAChE. The secondary structure of the enzyme is rendered
as red cylinders for helices, cyan arrows for sheet structures,
and gray tubing for the random coil. The bound inhibitor is
shown as solid spheres (carbon in gray, nitrogen in blue,
oxygen in red, and fluorine in light blue). Trp84 is shown in
green tubing partially hidden by the inhibitor, and Trp279,
which marks the entrance to the gorge, is shown as green
tubing near the top/center of the picture.
We have therefore achieved the synergistic effect of
the aminoquinoline group and the COCF3 moiety that
was aimed for in this work.
aminopyridine ring (allowing a similar hydrogen-bond-
ing pattern) and also preserves the stacking interaction
of the bicycle with Trp84. The origin of the flip comes
from the template on which the design work was carried
out: the tacrine had been incorrectly modeled into the
active site in a flipped orientation (see the Computa-
tional Methods).
The conformation of the linker group was poorly
predicted, and this may in part be due to the unusual
conformation of this in the observed structure, which
has an eclipsed torsion in one of the C-S bonds. The
conformation of the linker group allows an intramo-
lecular stacking interaction to occur between the phenyl
ring and the 4-aminoquinoline, which was not predicted
in the modeled structure, and probably offsets the
penalty of the eclipsed bond.
Despite the differences between the predicted and
observed binding modes, the design strategy was suc-
cessful. Molecule 27 is a potent inhibitor with an IC50
value of 3 nM. Although the predicted binding was
wrong in some matters of detail, 27 does span the
trifluoroacetyl and tacrine binding sites that were
targeted at the outset.
Com p a r ison of th e An ticip a ted Bin d in g Mod e
w ith Th a t Exp er im en ta lly Obser ved
Ultimately, the extent to which structure-based de-
sign is successful depends on our ability to predict
protein-ligand interactions. To assess the method used
here, we thus solved the crystal structure of compound
27 complexed with T. californica AChE as described in
the Experimental Section. Figure 7 shows an overall
view of the complex while Figure 8 shows the simulated
annealing omit map for the inhibitor 27. As is detailed
below, there are both similarities and differences be-
tween the predicted and observed binding modes (Figure
9).
The position of the trifluoroacetyl group was predicted
well, which is perhaps not surprising because this is
constrained by a covalent bond to the protein.
The largest difference, however, occurs for the 4-ami-
noquinoline bicycle, which is flipped over in the pre-
dicted structure with respect to the observed structure.
Although at first sight this appears to be a major
change, the flip largely preserves the position of the

3210 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20
Doucet-Personeni et al.
over relative to tacrine. The two aminopyridine rings were in
much the same position, but the second ring of the amino-
quinoline was near the saturated ring of tacrine rather than
the unsaturated ring it more closely resembled.
The trifluoroacetyl group was located in the same position
as the acetyl group of acetylcholine as modeled into the
structure of the apoenzyme by Sussman et al. (1ACE). The
resulting template was characterized by the distances in
Figure 5.
The conformational searching was carried out with the
systematic search option of SYBYL (version 6.01, Tripos Inc.,
1699 South Hanley Rd., St. Louis, MO, 63144). This was used
to generate all conformations of the flexible linking chain of
each candidate molecule, such that the distance constraints
were met to within a predefined tolerance. Depending on the
molecule, tolerances of between (0.3 and (1.0 Å were used;
torsion increments were 5°, and the van der Waals reduction
factor was 0.85. An in-house program (CONPICKS) was used
to select up to 200 representative conformations from the
search output for each molecule. Each conformation was
energy minimized with the molecular mechanics program
AESOP (B. B. Masek, unpublished results), but with the nine
distances constrained to stay near their template values. The
strain energy of each conformation relative to the global
minimum was estimated, and conformations with strain
energies >10 kcal/mol were rejected. The surviving, unique
conformations were overlaid on the template and inspected
in the active site. Molecules were qualitatively assessed by
visual inspection, on the basis of their goodness of fit to the
template, steric clashes with the protein, conformational
strain, solvent accessibility of H-bonding groups, and number
of flexible torsions.
Ch em istr y. Melting points were determined with a Gal-
lenkamp apparatus and are uncorrected. 1H NMR and 19F
NMR spectra (270 MHz) were taken on a J EOL GSX270
spectrometer. Chemical shifts are given in δ values (ppm) with
tetramethylsilane (for 1H NMR) and CFCl3 (for 19F NMR) as
internal standard. Abbreviations used in the NMR analyses
are as follows: br s ) broad singlet, d ) doublet, m )
multiplet, q ) quartet, t ) triplet. The IR spectra were
recorded with a Perkin-Elmer 881 infrared spectrophotomer
using neat liquid films (for liquids) or Nujol Mulls (for solids).
The HRMS spectra were taken on a Micromass Autospecq
spectrometer, the electrospray mass spectra on a Micromass
Platform LC spectrometer in electrospray mode, and the
autoprobe mass spectra on a Micromass TRIO2000 spectrom-
eter. Where analyses are indicated only by symbols of the
elements, the analytical results obtained for those elements
(performed by Medac Ltd., Brunel University, U.K.) were
within (0.4% of theoretical values.
Gen er a l P r oced u r es. 11-Am in o-1-ch lor o-6H-cycloh ep -
ta [b]qu in olin e (2). ZnCl2 (anhydrous, 45.0 g, 0.33mol) was
added to a solution of 2-amino-6-chlorobenzonitrile (50.0 g,
0.33mol) in cycloheptanone (280 mL). The mixture was stirred
at 120 °C for 3 h. When a dense precipitate formed, the mixture
was cooled to room temperature and the solvent decanted off.
The residual solid was triturated with EtOAc (150 mL) and
filtered off. This was added to 2 M NaOH (400 mL) and stirred
for 1.5 h. CH2Cl2 (500 mL) was then added and stirring
continued for 0.5 h. Undissolved solid was filtered off. The
organic phase of the filtrate was washed with water, dried
(MgSO4), and evaporated to give the crude product. Recrys-
tallization from CH2Cl2/EtOAc gave the product as a white
solid (55.3 g, 68%): mp 195-196 °C; 1H NMR (CDCl3) δ 1.60-
1.95 (6H, m), 2.67-2.72 (2H, m), 3.05-3.12 (2H, m), 5.82 (1H,
br s), 7.30-7.42 (2H, m), 7.75 (1H, m). Anal. (C14 H15ClN2) C,
H, N.
1-Ch lor o-11-(2-Th iop h en ylm eth yl)a m in o-6H-cycloh ep -
ta [b]qu in olin e (3). NaH (60% in oil, 2.30 g, 0.0575mol) was
suspended in dry DMF (150 mL), and compound 2 (13.20 g,
0.054mol) in dry DMF (150 mL) was added at 20 °C with
stirring over 0.5 h. When effervescence had subsided, the
mixture was heated at 50 °C for 1 h and then cooled to room
temperature. A solution of 2-bromomethylthiophene (9.50 g,
F igu r e 9. Comparison of the crystallographically observed
binding of 27 (green carbon atoms) with the predicted binding
mode (magenta carbons). The wire-frame protein atoms relate
to the predicted structure. The orientation is similar to that
shown in Figure 7.
Con clu sion s
On the basis of the structures of enzyme-inhibitor
complexes, we have designed novel and potent reversible
inhibitors targeting AChE from T. californica which
occupy most of the functional region of the active site.
The modeled complex of one of these inhibitors has been
compared to the structure obtained by X-ray crystal-
lography, and despite some differences between the
predicted and observed complexes, the design strategy
can be argued to have been successful. Although it is
difficult to predict exact binding modes with the current
technologies, they provide us with approximate models
that are often good enough to allow the design of active
molecules.
The type D inhibitors, which present a good inhibitory
efficiency and a limited conformational flexibility, pro-
vided us with an ideal template, and starting point for
further synthesis. This has allowed us to improve the
potency of the inhibitors as well as to explore species
differences between AChEs, the design of insect-selec-
tive active site AChE inhibitors being of particular
interest to us. This work has been guided by the
crystallographic studies and will be reported elsewhere.
In a contemporaneous study to the work described in
this paper, researchers at the Universitat de Barcelona
have developed other potent AChE inhibitors using a
similar hybridization strategy, but instead combining
the features of tacrine with those of huperzine A.27,29,43-45
The success of both their work and ours indicates the
utility of hybridization strategies in structure-based
ligand design programs.
Exp er im en ta l Section
Com p u ta tion a l Meth od s. Details of the template defini-
tion are as follows. The aminoquinoline ring was positioned
as closely as possible to the tacrine location inferred from
preliminary crystallographic results, which had been made
available to the modeling group in advance of the full 3D
coordinates. The ring nitrogen was within H-bonding distance
of the His440 backbone carbonyl oxygen. When the fully
refined 3D coordinates of the tacrine complex became avail-
able, we realized that the docked aminoquinoline was flipped

Development of Reversible AChE Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20 3211
0.054mol) in dry DMF (50 mL) was added, and the mixture
was heated at 60 °C for 2 h, then cooled to room temperature,
and poured into water. The mixture was adjusted to pH 12
with 2 M NaOH and extracted three times with CH2Cl2. The
organic extracts were combined, washed with water, dried
(MgSO4), and evaporated to give crude 3. This was purified
by chromatography on silica using CHCl3/EtOAc (80:20), giving
the product as a pale brown solid (9.8 g, 53%): mp 105-106
°C; 1H NMR (CDCl3) δ 1.70-2.00 (6H, m), 3.00-3.10 (2H, m),
3.10-3.20 (2H, m), 4.45 (2H, d), 5.95 (1H, br t), 6.95-7.02 (2H,
m), 7.20-7.27 (1H, m), 7.37-7.47 (2H, m), 7.85-7.92 (1H, m).
Anal. (C19H19ClN2S) C, H, N.
stirred at 0-5 °C for 0.5 h, and the resulting mixture was
extracted with EtOAc (3 × 250 mL). The extracts were
combined, washed with water, dried (MgSO4), and evaporated
to give a yellow crystalline solid (54.5 g, 78%): mp 70-71 °C;
1H NMR (CDCl3) δ 7.66 (2H, br s), 7.62 (1H, m), 7.55 (1H, m),
7.00 (1H, m), 2.77 (3H, s), 4.42 (2H, q), 1.44 (3H, t). Anal.
(C13H13FN2O2) C, H, N.
4-Am in o-3-ca r b oxyet h yl-5-ch lor o-2-m et h ylq u in olin e
(8b). 8b was obtained from 7b and purified by chromatography
on silica using CHCl3/MeOH (98:2) (75%). An analytical sample
1
was recrystallized from hexane: mp 70 °C; H NMR (CDCl3)
δ 8.25 (2H, br s), 7.30-7.77 (3H, m), 4.40 (2H, q), 2.74 (3H, s),
1.42 (3H, t); M+ (probe) 264 (1 Cl). Anal. C13H13ClN2O2) C, H,
N.
1-Ch lor o-11-(2-Tr iflu or oa cet yl-5-t h iop h en ylm et h yl)-
a m in ocycloh ep ta [b]qu in olin e (4). Diisopropylamine (0.59
g, 0.0058 mol) was dissolved in dry THF (20 mL) and
n-butyllithium (3.65 mL of a 1.6 M solution in hexane, 0.0058
mol) added dropwise, under nitrogen with stirring at -65 °C.
The solution was then stirred at this temperature for 10 min.
A solution of compound 3 (1.00 g, 0.0052 mol) in dry THF (15
mL) was added dropwise at -65 °C and stirred at this
temperature for 1.5 h. A solution of N-methyl-N-methoxytri-
fluoroacetamide in THF (5 mL) was added and the reaction
allowed to warm to 0 °C and maintained at this temperature
for 0.5 h. It was then allowed to warm to room temperature,
poured into cold water, made alkaline with 2 M NaOH, and
extracted three times with CH2Cl2. The organic extracts were
combined and washed with water, dried (MgSO4), and evapo-
rated to give the crude product which was purified by chro-
matography on silica using CHCl3/EtOAc (85:15) to give 4 as
4-Am in o-3-car boxyeth yl-2-m eth ylqu in olin e (8c). 8c was
obtained from 7c (72%). An analytical sample was recrystal-
1
lized from hexane: mp 137 °C; H NMR (CDCl3) δ 7.32-7.88
(4H, m), 7.05 (2H, br s), 4.41 (2H, q), 2.80 (3H, s), 1.45 (3H, t).
Anal. (C13H14N2O2) C, H, N.
4-Am in o-5-flu or o-3-h yd r oxym e t h yl-2-m e t h ylq u in o-
lin e (9a ). Lithium aluminum hydride (9.9 g, 0.26 mol) was
mixed with THF (1.2 L), and a solution of compound 8a (54.5
g, 0.22 mol) in THF (0.4L) was added to the stirred mixture,
over 1 h at 18-25 °C under a nitrogen atmosphere. After the
mixture was stirred for a further 3 h, water (75 mL) was added
dropwise with external ice/water cooling. The mixture was
then acidified with aqueous hydrochloric acid (2 M, 0.3 L), after
which it was made basic with saturated sodium hydroxide
solution (100 mL). The organic layer was separated, dried
(MgSO4), and evaporated. The residue was recrystallized from
ethanol (0.4 L) to give 4-amino-5-fluoro-3-hydroxymethyl-2-
methylquinoline as pale yellow crystals (31 g, 68%): mp 218-
1
a pale yellow solid (0.43 g, 34%): mp 105-106 °C; H NMR
(CDCl3) δ 1.75-1.91 (6H, m), 3.04-3.08 (2H, m), 3.16-3.20
(2H, m), 4.48 (2H, d, J ) 6.5), 5.97 (1H, br), 7.15 (1H, J )
4.2), 7.44-7.46 (2H, m), 7.87-7.93 (2H, m); 19F NMR (CDCl3)
1
220 °C; H NMR (DMSO) δ 6.69-7.45 (3H, m), 6.25 (2H, br
-72.6 ppm; MH+ (electrospray) 438 (1 Cl); IR (solid) 1684 cm-1
;
s), 4.95 (1H, t), 4.50 (2H, d), 2.45 (3H, s). Anal. (C11H11FN2O)
C, H, N.
Anal. (C21H18ClF3N2OS·0.25H2O) H, N, C calcd 57.47, found
56.88.
4-Am in o-5-ch lor o-3-h yd r oxym eth yl-2-m eth ylqu in olin e
(9b). 9b was obtained from compound 8b (65%): mp 240-
1-Ch lor o-11-(6,6-Dih yd r oxy-7,7,7-tr iflu or oh ep ta n -1-yl)-
am in ocycloh epta[b]qu in olin e (6a) an d 11-(6,6-Dih ydr oxy-
7,7,7-t r iflu or oh ep t a n -1-yl)a m in ocycloh ep t a -[b]q u in o-
lin e (6b). Compound 4 (0.900 g, 0.00205mol) was dissolved
in THF at 20 °C and Raney nickel added (7.0 mL of 50%
suspension, excess). At 1.5 h intervals more Raney nickel was
added (15 mL in three portions). After 4.5 h total the mixture
was filtered through Celite and the precipitate washed with a
mixture of THF/water (3:1, 50 mL). The filtrate was evaporated
under reduced pressure, and the residue was purified by
chromatography on silica gel using methanol/chloroform (8:
92), giving a mixture of 6a and 6b. These were separated by
chromatography on silica gel using dichloromethane/hexane/
triethylamine (80:20:2) to give 0.17 g (15%) of 6a and 0.21 g
(26%) of 6b as yellow oils. Data for compound 6a : 1H NMR
(CDCl3) δ 1.36-1.97 (14H, m), 2.65-2.80 (2H, m), 2.90-3.0
(2H, m), 3.05-3.15 (4H, m), 5.30 (1H, br), 7.35-7.45 (2H, m),
7.80-7.90 (1H, m); 19F NMR (CD3OD) -83.0 ppm; MH+
(electrospray) 432 (1 Cl). Anal. (C21H26ClF3N2O2) H, N, C calcd
58.54, found 57.18. Data for compound 6b: 1H NMR (CDCl3)
δ 1.35-1.90 (14H, m), 2.65-2.75 (2H, m), 2.85-2.95 (2H, m),
3.15-3.36 (4H, m), 3.50 (1H, br), 7.37-7.47 (1H, m), 7.52-
7.60 (1H, m), 7.80-7.90 (1H, m), 7.90-8.00 (1H, m); 19F NMR
1
241 °C; H NMR (DMSO) δ 7.30-7.70 (3H, m), 6.72 (2H, br
s), 5.10 (1H, t), 4.67 (2H, d), 2.57 (3H, s); M+ (probe) 222 (1
Cl). Anal. (C11H11ClN2O) C, H, N.
4-Am in o-3-h yd r oxym eth yl-2-m eth ylqu in olin e (9c). 9c
was obtained from compound 8c (84%): 1H NMR (DMSO) δ
7.27-8.07 (4H, m), 6.35 (2H, br s), 4.52 (2H, s), 2.40 (3H, s).
4-Am in o-3-ch lor om et h yl-5-flu or o-2-m et h ylq u in olin e
Hyd r och lor id e (10a ). Compound 9a (40.0 g, 0.19 mol) was
added in portions to stirred thionyl chloride (180 mL) over 5
min (exothermic reaction). The resulting mixture was stirred
under reflux for 1 h, after which it was evaporated to leave a
residue. The residue was washed with ethanol-free chloroform
(75 mL) and the white solid filtered off (49 g, quantitative):
1H NMR (DMSO) δ 2.70 (3H, s), 5.00 (2H, s), 7.35-7.45 (1H,
m), 7.80-7.90 (2H, m), 8.50 (2H, br s), 14.90 (1H, s). Anal.
(C11H11Cl2FN2) H, N; C calcd 50.60, found 50.08.
4-Am in o-5-ch lor o-3-ch lor om et h yl-2-m et h ylq u in olin e
Hyd r och lor id e (10b). 10b was obtained from compound 9b
(94%): 1H NMR (DMSO) δ 2.72 (3H, s), 5.00 (2H, s), 7.62-
8.10 (3H, m), 8.70 (2H, br s), -13.17 (1H, s).
4-Am in o-3-ch lor o-2-m et h ylq u in olin e H yd r och lor id e
(10c). 10c was obtained from compound 9c (88%).
4-Am in o-5-flu or o-2-m e t h yl-3-(a m in om e t h yl)q u in o-
lin e (11a ). Compound 10a (1.20 g, 0.0046 mol) was added to
a stirred suspension of potassium phthalimide (2.12 g, 0.0115
mol) in DMF (35 mL) at 100 °C, stirred for 2 h at 100 °C, then
cooled, and evaporated. The residue was triturated with water,
and the solid was filtered off, washed with 2 M NaOH and
then with cold water, and air-dried. This was suspended in
refluxing EtOH (30 mL), and hydrazine hydrate (0.60 mL,
excess) was added. It was stirred under reflux for 1 h,
evaporated, and concentrated. HCl (3 mL) and water (1.2 mL)
were then added, and the mixture was stirred under reflux
for 1 h, cooled, diluted with water (50 mL), and then made
alkaline with 2 M NaOH. It was extracted several times with
CH2Cl2, and the extracts were dried (MgSO4) and evaporated
to give the product as a pale yellow solid (0.28 g, 30%): mp
(CD3OD) -83.1 ppm; MH+ (electrospray) 397. Anal. (C21H27
F3N2O2) H, N, C calcd 63.62, found 62.10.
-
4-Amino-3-carboxyethyl-5-fluoro-2-methylquinoline (8a).
Ethyl acetoacetate (36.8 g, 0.28 mol) was added to a solution
of 7a (38.5 g, 0.28 mol) in toluene (800 mL). Stannic chloride
(147.5 g, 0.57 mol) was then added in a slow stream to the
stirred mixture while the temperature was kept within the
range 20-40 °C by external cooling using an ice/water bath.
The resultant mixture was cautiously heated to initiate an
exothermic reaction which was thereafter moderated by
external cooling with an ice bath. The mixture was then stirred
under reflux for 5 h and cooled to room temperature and the
supernatant liquid decanted from an oily residue. The residue
was triturated with EtOAc (750 mL) and a yellow solid filtered
off. This was added to sodium hydroxide solution (2 M) and

3212 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20
111-112 °C; 1H NMR (CDCl3) δ 6.90-7.70 (3H, m), 6.50 (2H,
Doucet-Personeni et al.
of compound 14 (1.00 g, 0.0045 mol) in DMF (45 mL),
portionwise at 25 °C. This was stirred at 50 °C for 0.5 h and
then cooled to 25 °C, 16 (0.98 g, 0.0045 mol) was added and
stirred at 25 °C for 1 h and then at 60 °C for 2 h, and the
mixture was evaporated. The residue was mixed with water
(20 mL) and extracted with Et2O (3 × 20 mL). Ether extracts
were washed with brine (10 mL), dried (MgSO4), and evapo-
rated. The residue was purified by chromatography on silica
using EtOAc/hexane/NEt3 (9:1:0.05), giving 17 as a white solid
br), 4.11 (2H, m), 2.65 (3H, s), 1.40 (2H, br). Anal. (C11H12
FN3) C, H, N.
-
4-Am in o-5-ch lor o-2-m e t h yl-3-(a m in om e t h yl)q u in o-
lin e (11b). 11b was obtained from 10b and purified by
chromatography using CH2Cl2/MeOH/NEt3 (4:1:0.1) (65%): 1H
NMR (CDCl3) δ 7.30-7.75 (3H, m), 7.12 (2H, br-s), 4.08 (2H,
s), 2.61 (3H, s), 1.40 (2H, br s); M+ (probe) 221 (1 Cl).
5-Tr iflu or m e t h yl-5-t r im e t h ylsilyloxy-γ-b u t yr ola c-
ton e (12). Trimethylsilyltrifluoromethane (4.81 g, 0.034 mol)
was added to a stirred suspension of succinic anhydride (3.38
g, 0.034 mol) in dry THF (20 mL) under dry nitrogen.
Tetrabutylammonium fluoride (1.0 M in THF, 0.50 mL,
0.00050 mol) was then added, stirred at 60 °C for 7 h, and
evaporated. The residue was distilled (bp 120-130 °C, 15
mmHg), giving a clear colorless oil (2.47 g, 30%). GC/MS
analysis showed two peaks of equal intensity both having
molecular weight 242. The product was used in the preparation
of 13a and 13b.
1
(0.51 g, 31%): mp 134-135 °C; H NMR (CDCl3) δ 7.65 (1H,
d, J ) 7.4), 7.47 (1H, m), 7.00 (1H, m), 5.75 (2H, br), 4.25 (2H,
t, J ) 7.0), 4.10 (2H, ABq, J ) 52.7, 10.5), 2.68 (3H, s), 2.00-
2.50 (6H, m), 1.70 (3H, m); 19F NMR (CDCl3) -77.6 ppm; M+
(probe) 360. Anal. (C16H16F4N2OS) C, H, N.
4-Am in o-3-ben zyloxy-5-flu or o-2-m eth ylqu in olin e (18).
18 was obtained from 10a and benzyl alcohol and purified by
recrystallization from hexane (62%): mp 114-115 °C; 1HNMR
(CDCl3) δ 6.90-7.70 (8H, m), 5.95 (2H, br s), 4.77 (2H, s), 4.55
(2H, s), 2.57 (3H, s); M+ 296 (probe). Anal. (C18H17FN2O) C,
H, N.
1-Br om o-3-(t er t -b u t yld im e t h ylsilyloxym e t h yl)b e n -
zen e (20). tert-Butyldimethylsilyl chloride (44.8 g, 0.30 mol)
in DMF (200 mL) was added to a stirred solution of imidazole
(47.0 g, 0.69 mol) and 19 (50.0 g, 0.27 mol) in DMF (75 mL) at
20 °C (ice/water bath cooling), stirred at 50 °C for 20 h, and
cooled to room temperature, and water was (1.25 L) added.
The mixture was extracted with hexane (3 × 375 mL), and
the hexane extracts were washed with water (5 × 250 mL),
dried (MgSO4), and evaporated to leave a clear colorless oil
(74.5 g).
1-Tr iflu or oa cet yl-3-(ter t-b u t yld im et h ylsilyloxym et h -
yl)ben zen e (21). A 68.4 g (0.23 mol) sample of compound 20
was dissolved in dry Et2O (700 mL), and n-BuLi (100 mL, 2.5
M in hexanes, 0.25 mol) was added over 15 min, under dry
N2, at 0-5 °C. The clear yellow solution was allowed to warm
to 10 °C over 30 min and then recooled to 0 °C. A solution of
N-trifluoroacetylpiperidine (41.2 g, 0.23 mol) in Et2O (50 mL)
was then added over 30 min at 0-5 °C. The cooling bath was
removed, and the solution was stirred for 1 h and then recooled
to 0 °C. A saturated NH4Cl solution (250 mL) was then added.
The aqueous portion was extracted with Et2O (75 mL). The
combined ether extracts were washed with NH4Cl solution
(4 × 25 mL) and then water (4 × 25 mL), dried (MgSO4) and
evaporated, leaving a clear yellow oil (70.0 g).
3-Tr iflu or oa cetylben zyl Alcoh ol (22). A 70.0 g (0.22 mol)
sample of compound 21 was dissolved in MeOH (600 mL), and
concentrated HCl (60 mL) was added at 10-20 °C (ice/water
cooling bath). After the mixture was stirred for 30 min, it was
concentrated to 450 mL on a water pump at 25 °C. The residue
was diluted with Et2O (2 L), neutralized with a slurry of
NaHCO3/water, then washed with water, dried (MgSO4), and
evaporated. The residue was purified by chromatography on
silica using EtOAc/hexane (3:2), giving 22 as a pale yellow
viscous oil (31.7 g, 66% from 3-bromobenzyl alcohol): IR (Nujol
mull) 3334, 1719 cm-1; 1H NMR (CDCl3) δ 7.40-8.00 (4H, m),
4.80 (2H, s).
4-Am in o-5-flu or o-2-m et h yl-3-[N-(4-t r iflu or om et h yl-γ-
bu t yr ola ct on -4-yl)]a m in om et h ylq u in olin e (13a ). Com-
pound 12 (0.39 g, 0.0016mol) was added to a stirred suspension
of compound 11a (0.33 g, 0.0016 mol) in dry MeCN (15 mL),
and the mixture was stirred under reflux for 8 h and then
cooled to room temperature. The crystalline precipitate was
filtered off, dissolved in warm MeOH (5 mL), and filtered. The
filtrate was evaporated to give the product as a pale yellow
1
solid (0.15 g, 25%): mp 185-186 °C dec; H NMR (DMSO) δ
7.92 (1H, br), 7.37-7.47 (2H, m), 6.96-7.06 (1H, m), 4.55 (2H,
ABq), 2.60 (3H, s), 2.00-2.50 (4H, m); 19F NMR (CDCl3) -81.5
ppm; MH+ (electrospray) 358. Anal. (C16H15N3F4O2) C, H, N.
4-Am in o-5-ch lor o-2-m eth yl-3-[N-(4-tr iflu or om eth yl-γ-
bu tyr ola cton -4-yl)]a m in om eth ylqu in olin e (13b). 13b was
obtained from compounds 11b and 12 and purified by chro-
matography on silica using EtOAc/MeOH (8:2) (50%): mp
179-181 °C dec; 1H NMR (CDCl3) δ 7.20-7.60 (3H, br), 7.20-
7.00 (3H, m), 4.65 (2H, ABq), 2.65 (3H, s), 2.20-2.80 (4H, m);
19F NMR (CD3OD) -83.6 ppm; M+ (probe) 373 (1 Cl). Anal.
(C16H15F3ClN3O2) H, N, C calcd 51.42, found 50.91.
4-Am in o-5-flu or o-2-m eth yl-3-(2-m er ca p tom eth yl)qu in -
olin e (14). A solution of thiourea (2.63 g, 0.034 mol) in DMF
(15 mL) was added to a stirred suspension of compound 16
(9.0 g, 0.034 mol) in DMF (45 mL), and the mixture was stirred
at 100 °C for 2 h. The mixture was then evaporated, and the
residue was mixed with 2 M NaOH (65 mL) and stirred at
reflux under a nitrogen atmosphere for 2 h. The mixture was
cooled to room temperature and filtered. The filtrate was cooled
in ice, acidified with concentrated HCl, and then neutralized
with NaHCO3 solution. The yellow solid was filtered off and
washed with cold water (10 mL) (4.7 g, 61%): 1H NMR (D2O)
δ 7.60 (1H, m), 7.27 (1H, m), 7.15 (H, m), 3.65 (2H, s), 2.55
(3H, s).
2-Tr iflu or om et h yl-2-t r im et h ylsilyloxy-t et r a h yd r ofu -
r a n (15). γ-Butyrolactone (10.5 g, 0.12 mol) and CF3TMS (19.3
g, 0.13 mol) were dissolved in dry THF (50 mL), and 1.1 M
TBAF in THF (1 mL) was added dropwise, with stirring, at 0
°C and then stirred for 2 h at 20 °C. The solvent was removed
at atmospheric pressure and the residue distilled on the water
pump. The product was collected at 50-55 °C,12 mmHg, as a
clear colorless liquid (14 g, 50%).
1-Br om o-5,5,5-tr iflu or op en ta n -4-on e (16). 15 (14.0 g,
0.061 mol) and 48% HBr (35 mL) were heated at 115 °C in a
sealed glass tube for 90 h, cooled, mixed with water (200 mL),
and extracted with CH2Cl2 (3 × 30 mL). The extracts were
washed successively with water, saturated sodium bisulfite
solution, and then water. The product was dried (MgSO4), and
the solvent was removed at atmospheric pressure. The residue
was distilled on the water pump. The product was collected
at 40-50 °C, 12 mmHg, as a clear colorless liquid (8 g, 60%):
1H NMR (CDCl3) δ 3.40 (2H, t), 2.92 (2H, t), 2.15 (2H, m); M+
219.
4-Am in o-5-flu or o-2-m eth yl-3-(3-tr iflu or oa cetylben zyl-
oxym eth yl)qu in olin e (23a ). A mixture of compound 22 (31.0
g, excess) and compound 10a (7.80 g, 0.030 mol) was stirred
at 125-130 °C for 4 h, and HCl was evolved. After the mixture
was cooled to room temperature, water (250 mL) was added
and the mixture extracted with Et2O (5 × 500 mL). The clear
aqueous portion was cooled in ice and made alkaline to pH
10-12 with 2 M NaOH. This was then extracted with CH2Cl2
(5 × 100 mL), and the extracts were dried (MgSO4) and
evaporated. The residue was purified by chromatography on
silica using EtOAc/NEt3 (10:0.1), giving 23a as a white powder
1
(5.5 g, 47%): mp 136-137 °C; H NMR (CDCl3) δ 6.95-8.07
(7H, m), 5.97 (2H, br), 4.86 (2H, s), 4.62 (2H, s), 2.60 (3H, s);
19F NMR (CDCl3) -71.8 ppm (95%, free ketone) -84.8 ppm
(5%, hydrated ketone); M+ calcd for C20H16F4N2O2 392.114791,
found 392.116037. Anal. (C20H16F4N2O2‚0.5H2O) C, H, N.
4-Am in o-5-flu or o-2-m eth yl-3-[S-(2-tr iflu or om eth yl-tet-
r a h yd r ofu r a n -2-yl)th iom eth yl]qu in olin e (17). NaH (80%
in oil, 0.135 g, 0.0045 mol) was added to a stirred suspension
4-Am in o-5-ch lor o-2-m eth yl-3-(3-tr iflu or oa cetylben zyl-
oxym eth yl)qu in olin e (23b). 23b was obtained from com-

Development of Reversible AChE Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20 3213
Ta ble 2. Data Collection and Processing Statistics
oscillation angle (deg)
0.25
251
409141
302.5
34411
0
total number of frames
total number of reflections
resolution range (Å)
number of unique reflections
redundancy
1
6.0
2
13.1
3
23.9
4
30.5
>5
24.1
% reflections
1.3
average redundancy (weighted)
overall completeness (%)
completeness in highest resolution shell (%)
3.5
98.4
91.8
Ta ble 3. Refinement Results
resolution range (Å)
number of protein atoms
number of nonprotein atoms
solvent (water, PEG, MES)
carbohydrate atoms
inhibitor
30.0-2.5
Rwork (%)
18.3
21.4
0.021
2.0
4244
Rfree (5% of reflections) (%)
rms bond deviations (Å)
rms angle deviations (deg)
171, 13, 12
42
28
pounds 10b and 22 and purified by chromatography on silica
using CH2Cl2/MeOH/NEt3 (96:2:2) (64%): mp 106-107 °C; 1H
NMR (DMSO) δ 7.27-7.65 (7H, m), 6.72 (2H, br), 4.65 (2H,
s), 4.55 (2H, s), 2.45 (3H, s); 19F NMR (DMSO) -71.0 ppm (11%
free ketone) -83.2 ppm (89% hydrated ketone); MH+ 409 (1
Cl, electrospray). Anal. (C20H16ClF3N2O2‚H2O) C, H, N.
4-Am in o-2-m eth yl-3-(3-tr iflu or oa cetylben zyloxym eth -
yl)qu in olin e (23c). 23c was obtained from compounds 10c
and 22 and purified by chromatography on silica using EtOAc/
NEt3 (99:1) (35%): mp 100-102 °C; 1H NMR (CDCl3) δ 7.40-
8.10 (8H, m), 5.40 (2H, br), 4.90 (2H, s), 4.61 (2H, s), 2.67 (3H,
s); 19F NMR (CDCl3) -71.9 ppm (99%, free ketone) -78.9 ppm
(1%, hydrated ketone); M+ calcd for C20H17F3N2O2 374.124213,
found 374.124035. Anal. (C20H17F3N2O2‚0.5H2O) H, N; C calcd
62.65, found 62.11.
C20H16F4N2OS 408.091948, found 408.092201. Anal. (C20H16F4N2-
OS·0.5H2O) C, H, N.
4-Am in o-5-flu or o-2-m eth yl-3-[3-(1,1-d ih yr oxy-2,2,2-tr i-
flu or oeth yl)-ben zylth iom eth yl]qu in olin e (28). A solution
of NaBH4 (0.15 g, 0.000395 mol) in EtOH (3 mL) was added
to a stirred solution of compound 27 in EtOH (2.8 mL). The
mixture was stirred at room temperature for 18 h and then
evaporated. The residue was mixed with water (0.25 mL) and
CH2Cl2 (5 mL). The CH2Cl2 extract was dried (MgSO4) and
evaporated, giving 28 as a white solid (0.083 g, 82%): mp 188-
189°; 1H NMR (DMSO) δ 6.77-7.47 (7H, m), 6.30 (2H, br),
5.10 (1H, m), 3.85 (2H, s), 3.75 (2H, s), 2.30 (3H, s); 19F NMR
(DMSO) -77.01 ppm, -77.04 ppm (1:1, CF3CH). Anal.
(C20H18F4N2OS) C, H, N.
4-Am in o-5-flu or o-2-m eth yl-3-(3-a cetylben zylth iom eth -
yl)qu in olin e (29). 29 was obtained from compound 14 and
3-acetylbenzyl bromide and purified by chromatography on
silica using CH2Cl2/MeOH (97:3) (73%). An analytical sample
was prepared by recrystallization from EtOAc: mp 157-158
3-Tr iflu or oa cetyltolu en e (25). To a suspension of Mg (9.5
g, 0.39 mol) and I2 (1 crystal) in dry Et2O (50 mL) was added
3-bromotoluene (44.5 g, 0.26 mol) with stirring, under N2, over
0.5 h. The mixture was refluxed for 1 h, diluted with Et2O
(250 mL), and a solution of CF3CO2H (10.0 g, 0.39 mol) in Et2O
(100 mL) was added at 10-20 °C with ice/water cooling over
1 h. It was stirred under reflux for 2 h, cooled in ice, poured
onto a mixture of ice (500 g), and concentrated HCl (100 mL).
The ether layer was separated, washed with saturated NaH-
CO3 solution (2 · 50 mL), dried (MgSO4), and evaporated. The
residue was distilled on a water pump, giving a clear colorless
liquid (12.8 g, 77%): bp 65-68 °C, 12 mmHg; 1H NMR (CDCl3)
1
°C; H NMR (CDCl3) δ 6.90-7.95 (7H, m), 5.55 (2H, br), 3.82
(2H, s), 3.75 (2H, s), 2.61 (3H, s), 2.55 (3H, s); M+ (probe) 354.
Anal. (C20H19FN2OS) C, H, N.
Bioch em istr y. Ma ter ia ls. T. californica AChE was puri-
fied as described previously.18,46 Acetylthiocholine iodide (ATCh)
and 5,5′-dithiobis(2-nitrobenzoic acid) (DTNB) were purchased
from Sigma.
En zym a tic Assa ys. AChE activity was evaluated spectro-
photometrically at 405 nm using the method of Ellman47 in
0.1 M phosphate buffer, pH 7.8. The enzyme (0.8 µg/L) was
incubated for 10 min at room temperature with 10 concentra-
tions of a compound using EtOH as cosolvent (maximum 2.5%).
The hydrolysis of the substrate was followed for 18 min after
addition of ATCh (500 µM) and DTNB (150 µM) and was
corrected for spontaneous hydrolysis of the substrate. The
activities were then compared to the activity of a control and
expressed as a percent of inhibition. Determinations of IC50
were made by least-squares fitting of the Hill equation to the
percentage inhibition data. The presented values are the
average of at least four determinations. Standard errors are
less than 30% of the mean IC50 values.
X-r a y Cr ysta llogr a p h y. Compound 27 was dissolved in
methanol to give a 100 mM stock solution. Soaking solutions
(10 mM) were prepared by adding the stock solution to a
solution containing 0.5 M MES, pH 5.6, and 36% PEG 200
(v/v). A crystal of TcAChE was transferred to an 8 mL drop of
the soaking solution. After 5 days, the crystal was transferred
to oil, mounted in a nylon loop, and flash-cooled in a cryo-
stream (120 K) mounted on a Rigaku FRC rotating anode.
Data collection parameters are summarized in Table 2. Data
were processed using DENZO and SCALEPACK.48 Initial
refinement was begun using SHELX,49 and finally using
CNS.50 Final refinement parameters are summarized in Table
3. The coordinates have been deposited in the Protein Data
Bank (PDB) with reference code 1HBJ .
δ 7.80-7.37 (4H, m), 2.40 (3H, s); IR 1719 cm-1
.
3-Tr iflu or oa cetylben zyl br om id e (26). To a solution of
compound 25 (12.2 g, 0.064 mol) in CCl4 (110 mL) were added
NBS (12.7 g, 0.071 mol) and benzoyl peroxide. The mixture
was stirred under reflux under N2 for 5 h, cooled, and filtered.
The filtrate was evaporated, and the residue was dissolved in
Et2O (75 mL), washed with saturated NaHCO3 (25 mL), dried
(MgSO4), and evaporated. The residue was then distilled on
an oil pump, collecting the product as a pale brown oil (9.4 g,
55%): bp 48-60 °C, 0.2 mmHg; 1H NMR (CDCl3) δ 7.55-8.10
(4H, m), 4.52 (2H, s).
4-Am in o-5-flu or o-2-m eth yl-3-(3-tr iflu or oa cetylben zyl-
th iom eth yl)qu in olin e (27). Compound 14 (2.00 g, 0.0090
mol) in DMF (50 mL) was stirred at room temperature, and
NaH (0.27 g, 80% in oil, 0.0090 mol) was added portionwise.
The mixture was stirred at room temperature for 10 min and
then at 50 °C for 0.5 h. The mixture was cooled to room
temperature, and 26 (2.40 g, 0.0090 mol) was added. The
mixture was stirred at 60 °C for 2 h and then evaporated. The
residue was mixed with water (50 mL) and extracted with CH2-
Cl2 (2 × 50 mL). The extracts were combined, dried (MgSO4),
and evaporated to leave a residue which was purified by
chromatography on silica using EtOAc/NEt3 (10:0.05), giving
1
27 as a yellow solid (1.00 g, 27%): mp 131-132 °C; H NMR
(CDCl3) δ 6.90-8.10 (7H, m), 5.57 (2H, br), 3.80 (2H, s), 3.85
(2H, s), 2.54 (3H, s); 19F NMR (CDCl3) -72.7 ppm (95%, free
ketone) -85.7 ppm (5%, hydrated ketone); M+ calcd for

3214 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20
Doucet-Personeni et al.
(23) Pang, Y. P.; Kozikowski, A. P. Prediction of the binding site of
1-benzyl-4-[(5,6-dimethoxy-1-indanon-2-yl)methyl]piperidine in
acetylcholinesterase by docking studies with the SYSDOC
program. J . Comput.-Aided Mol. Des. 1994, 8, 683-693.
(24) Pang, Y. P.; Kozikowski, A. P. Prediction of the binding sites of
huperzine A in acetylcholinesterase by docking studies. J .
Comput.-Aided Mol. Des. 1994, 8, 669-681.
Ack n ow led gm en t. This work was supported by the
European Union’s 4th Framework Program in Biotech-
nology.
Refer en ces
(25) Cho, S. J .; Garsia, M. L.; Bier, J .; Tropsha, A. Structure-based
alignment and comparative molecular field analysis of acetyl-
cholinesterase inhibitors. J . Med. Chem. 1996, 39, 5064-5071.
(26) Kawakami, Y.; Inoue, A.; Kawai, T.; Wakita, M.; Sugimoto, H.;
et al. The rationale for E2020 as a potent acetylcholinesterase
inhibitor. Bioorg. Med. Chem. 1996, 4, 1429-1446.
(27) Camps, P.; Cusack, B.; Mallender, W. D.; El Achab, R. E.; Morral,
J .; et al. Huprine X is a novel high-affinity inhibitor of acetyl-
cholinesterase that is of interest for treatment of Alzheimer’s
disease. Mol. Pharmacol. 2000, 57, 409-417.
(28) Han, Y. F.; Li, C. P.; Chow, E.; Wang, H.; Pang, Y. P.; et al.
Dual-site binding of bivalent 4-aminopyridine- and 4-amino-
quinoline-based AChE inhibitors: contribution of the hydropho-
bic alkylene tether to monomer and dimer affinities. Bioorg. Med.
Chem. 1999, 7, 2569-2575.
(29) Camps, P.; El Achab, R.; Gorbig, D. M.; Morral, J .; Munoz-
Torrero, D.; et al. Synthesis, in vitro pharmacology, and molec-
ular modeling of very potent tacrine-huperzine A hybrids as
acetylcholinesterase inhibitors of potential interest for the
treatment of Alzheimer’s disease. J . Med. Chem. 1999, 42, 3227-
3242.
(30) Kryger, G.; Silman, I.; Sussman, J . L. Three-dimensional
structure of a complex of E2020 with acetylcholinesterase from
Torpedo californica. J . Physiol. Paris 1998, 92, 191-194.
(31) Harel, M.; Silman, I.; Sussman, J . L. Private commmunications,
later published as ref 11, 1997.
(32) Street, I. P.; Lin, H. K.; Laliberte, F.; Ghomashchi, F.; Wang,
Z.; et al. Slow- and tight-binding inhibitors of the 85-kDa human
phospholipase A2. Biochemistry 1993, 32, 5935-5940.
(33) Imperiali, B.; Abeles, R. H. Inhibition of serine proteases by
peptidyl fluoromethyl ketones. Biochemistry 1986, 25, 3760-
3767.
(34) Veale, C. A.; Bernstein, P. R.; Bohnert, C. M.; Brown, F. J .;
Bryant, C.; et al. Orally active trifluoromethyl ketone inhibitors
of human leukocyte elastase. J . Med. Chem. 1997, 40, 3173-
3181.
(35) Neises, B.; Broersma, R. J .; Tarnus, C.; Piriou, F.; Remy, J . M.;
et al. Synthesis and comparison of tripeptidylfluoroalkane
thrombin inhibitors. Bioorg. Med. Chem. 1995, 3, 1049-1061.
(36) Szekacs, A.; Halarnkar, P. P.; Olmstead, M. M.; Prag, K. A.;
Hammock, B. D. Heterocyclic derivatives of 3-substituted-1,1,1-
trifluoro-2-propanones as inhibitors of esterolytic enzymes.
Chem. Res. Toxicol. 1990, 3, 325-332.
(37) Nair, H. K.; Lee, K.; Quinn, D. M. m-(N,N,N-Trimethylammo-
nio)trifluoroacetophenone: A femtomolar inhibitor of acetylcho-
linesterase. J . Am. Chem. Soc. 1993, 115, 9939-9941.
(38) Gregor, V. E.; Emmerling, M. R.; Lee, C.; Moore, C. J . The
synthesis and in vitro acetylcholinesterase and butyrylcholin-
esterase inhibitory activity of Tacrine (COGNEX®) derivatives.
Bioorg. Med. Chem. Lett. 1992, 2, 861-864.
(39) Desai, M. C.; Thadeio, P. F.; Lipinski, C. A.; Liston, D. R.;
Spencer, R. W.; et al. Physical parameters for brain uptake:
optimising LOG P, LOG D and pKa of THA. Bioorg. Med. Chem.
Lett. 1991, 1, 411-414.
(40) Gelb, M. H.; Svaren, J . P.; Abeles, R. H. Fluoro ketone inhibitors
of hydrolytic enzymes. Biochemistry 1985, 24, 1813-1817.
(41) Godard, A.; Queguiner, G. o-Aminoformylquinolines, New Het-
erocyclic Synthons. J . Heterocycl. Chem. 1980, 17, 465-473.
(42) Krishnamurti, R.; Bellew, D. R.; Prakash, G. K. S. Preparation
of Trifluoromethyl and Other Perfluoroalkyl Compounds with
(Perfluoroalkyl)trimethylsilanes. J . Org. Chem. 1991, 56, 984-
989.
(43) Camps, P.; El Achab, R.; Morral, J .; Munoz-Torrero, D.; Badia,
A.; et al. New Tacrine-Huperzine A Hybrids (Huprines): Highly
Potent Tight-Binding Acetylcholinesterase Inhibitors of Interest
for the Treatment of Alzheimer’s Disease. J . Med. Chem. 2000,
43, 4657-4666.
(1) Enz, A.; Amstutz, R.; Boddeke, H.; Gmelin, G.; Malanowski, J .
Brain selective inhibition of acetylcholinesterase: a novel ap-
proach to therapy for Alzheimer’s disease. Prog. Brain Res. 1993,
98, 431-438.
(2) Millard, C. B.; Broomfield, C. A. Anticholinesterases: medical
applications of neurochemical principles. J . Neurochem. 1995,
64, 1909-1918.
(3) Weinstock, M. Possible role of the cholinergic system and disease
models. J . Neural Transm., Suppl. 1997, 49, 93-102.
(4) Lindner, A.; Schalke, B.; Toyka, K. V. Outcome in juvenile-onset
myasthenia gravis: a retrospective study with long-term follow-
up of 79 patients. J . Neurol. 1997, 244, 515-520.
(5) Aldridge, W. N. Some Properties of Specific Cholinesterase with
Particular Reference to the Mechanism of Inhibition by Diethyl
p-Nitrophenyl Thiophosphate (E605) and Analogues. Biochem.
J . 1950, 46, 451-460.
(6) Davis, K. L.; Powchik, P. Tacrine. Lancet 1995, 345, 625-630.
(7) Ballantyne, B.; Marrs, T. C. Clinical and experimental toxicology
of organophosphates and carbamates; Butterworth-Heine-
mann: Oxford.
(8) Sugimoto, H.; Iimura, Y.; Yamanishi, Y.; Yamatsu, K. Synthesis
and structure-activity relationships of acetylcholinesterase
inhibitors: 1-benzyl-4-[(5,6-dimethoxy-1-oxoindan-2-yl)methyl]-
piperidine hydrochloride and related compounds. J . Med. Chem.
1995, 38, 4821-4829.
(9) Sussman, J . L.; Harel, M.; Frolow, F.; Oefner, C.; Goldman, A.;
et al. Atomic structure of acetylcholinesterase from Torpedo
californica: a prototypic acetylcholine-binding protein. Science
1991, 253, 872-879.
(10) Harel, M.; Quinn, D. M.; Nair, H. K.; Silman, I.; Sussman, J . L.
The X-Ray Structure of a Transition State Analog Complex
Reveals the Molecular Origins of the Catalytic Power and
Substrate Specificity of Acetylcholinesterase. J . Am. Chem. Soc.
1996, 118, 2340.
(11) Harel, M.; Schalk, I.; Ehret-Sabatier, L.; Bouet, F.; Goeldner,
M.; et al. Quaternary ligand binding to aromatic residues in the
active-site gorge of acetylcholinesterase. Proc. Natl. Acad. Sci.
U.S.A. 1993, 90, 9031-9035.
(12) Millard, C. B.; Kryger, G.; Ordentlich, A.; Greenblatt, H. M.;
Harel, M.; et al. Crystal structures of aged phosphonylated
acetylcholinesterase: nerve agent reaction products at the
atomic level. Biochemistry 1999, 38, 7032-7039.
(13) Greenblatt, H. M.; Kryger, G.; Lewis, T.; Silman, I.; Sussman,
J . L. Structure of acetylcholinesterase complexed with (-)-
galanthamine at 2.3 A resolution. FEBS Lett. 1999, 463, 321-
326.
(14) Kryger, G.; Silman, I.; Sussman, J . L. Structure of acetylcho-
linesterase complexed with E2020 (Aricept): implications for the
design of new anti-Alzheimer drugs. Struct. Fold. Des. 1999, 7,
297-307.
(15) Harel, M.; Kleywegt, G. J .; Ravelli, R. B.; Silman, I.; Sussman,
J . L. Crystal structure of an acetylcholinesterase-fasciculin
complex: interaction of a three-fingered toxin from snake venom
with its target. Structure 1995, 3, 1355-1366.
(16) Bourne, Y.; Taylor, P.; Marchot, P. Acetylcholinesterase inhibi-
tion by fasciculin: crystal structure of the complex. Cell 1995,
83, 503-512.
(17) Bourne, Y.; Taylor, P.; Bougis, P. E.; Marchot, P. Crystal
structure of mouse acetylcholinesterase. A peripheral site-
occluding loop in a tetrameric assembly. J . Biol. Chem. 1999,
274, 2963-2970.
(18) Raves, M. L.; Harel, M.; Pang, Y. P.; Silman, I.; Kozikowski, A.
P.; et al. Structure of acetylcholinesterase complexed with the
nootropic alkaloid, (-)-huperzine A. Nat. Struct. Biol. 1997, 4,
57-63.
(19) Gilson, M. K.; Straatsma, T. P.; McCammon, J . A.; Ripoll, D.
R.; Faerman, C. H.; et al. Open “back door” in a molecular
dynamics simulation of acetylcholinesterase. Science 1994, 263,
1276-1278.
(20) Axelsen, P. H.; Harel, M.; Silman, I.; Sussman, J . L. Structure
and dynamics of the active site gorge of acetylcholinesterase:
synergistic use of molecular dynamics simulation and X-ray
crystallography. Protein Sci. 1994, 3, 188-197.
(44) Badia, A.; Banos, J . E.; Camps, P.; Contreras, J .; Gorbig, D. M.;
et al. Synthesis and evaluation of tacrine-huperzine A hybrids
as acetylcholinesterase inhibitors of potential interest for the
treatment of Alzheimer’s disease. Bioorg. Med. Chem. 1998, 6,
427-440.
(45) Barril, X.; Orozco, M.; Luque, F. J . Predicting Relative Binding
Free Energies of Tacrine-Huperzine A Hybrids as Inhibitors of
Acetylcholinesterase. J . Med. Chem. 1999, 42, 5110-5119.
(46) Sussman, J . L.; Harel, M.; Frolow, F.; Varon, L.; Toker, L.; et
al. Purification and crystallization of a dimeric form of acetyl-
cholinesterase from Torpedo californica subsequent to solubili-
zation with phosphatidylinositol-specific phospholipase C. J . Mol.
Biol. 1988, 203, 821-823.
(21) Ordentlich, A.; Barak, D.; Kronman, C.; Ariel, N.; Segall, Y.; et
al. Contribution of aromatic moieties of tyrosine 133 and of the
anionic subsite tryptophan 86 to catalytic efficiency and allosteric
modulation of acetylcholinesterase. J . Biol. Chem. 1995, 270,
2082-2091.
(22) Felder, C. E.; Botti, S. A.; Lifson, S.; Silman, I.; Sussman, J . L.
External and internal electrostatic potentials of cholinesterase
models. J . Mol. Graph. Model 1997, 15, 318-327, 335-317.

Development of Reversible AChE Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20 3215
(47) Ellman, G. L.; Courtney, K. D.; Andres, B. J .; Featherstone, R.
M. A new and rapid colorimetric determination of acetylcho-
linesterase activity. Biochem. Pharmacol. 1961, 7, 88-95.
(48) Otwinowski, Z.; Minor, W. Processing of X-ray Diffraction Data
Collected in Oscillation Mode. Methods Enzymol. 1997, 276,
307-326.
for macromolecular structure determination. Acta Crystallogr.,
D 1998, 54, 905-921.
(51) McRee, D. E. XtalView/Xfit--A versatile program for manipulat-
ing atomic coordinates and electron density. J . Struct. Biol. 1999,
125, 156-165.
(52) Merritt, E. A.; Bacon, D. J . Raster3D: Photorealistic molecular
graphics. Methods Enzymol. 1997, 277, 505-524.
(49) Sheldrick, G. M.; Schneider, T. M. SHELXL: High-Resolution
Refinement. Methods Enzymol. 1997, 277, 319-343.
(50) Brunger, A. T.; Adams, P. D.; Clore, G. M.; DeLano, W. L.; Gros,
P.; et al. Crystallography & NMR system: A new software suite
J M010826R