Delineation and decomposition of energies involved in quaternary ammonium binding in the active site of acetylcholinesterase

Article abstract of DOI:10.1021/ja9933588

The quaternary ammonium binding locus in the active site of mammalian acetylcholinesterase is subtended by the side chains of Trp86, Tyr133, Glu202, and Tyr337. Linear free-energy relationships define the interactions involved in molecular recognition by mouse acetylcholinesterase of the quaternary ammonium moiety of ligands. For substrates CH₃C(=O)XCH₂CH₂Y [X = O, Y = CHMe₂, or CH₂CH₃; X = S, Y = H, NH⁺Me₂, or N⁺Me₃] and trifluoroacetophenone transition state analogue inhibitors m-YC₆H₄C(=O)CF₃ [Y = H, Me, Et, iPr, tBu, CF₃, NH₂, NO₂, NMe₂, or N⁺Me₃], log(k(cat)/K(m)) and pK(i) depend linearly on the molar refractivity, but not the hydrophobicity, of the substituents Y. These correlations indicate that, in the acylation stage of catalysis, interactions in the quaternary ammonium binding locus stabilize the tetrahedral intermediate (as modeled by transition state analogue affinity) by (5 x 10⁵)-fold (ΔΔG(TI) = -32.5 kJ mol^-1) and the transition state by (2 x 10⁴)-fold (ΔΔG(+) = -24.5 kJ mol^-1). To evaluate the contribution of cation-π interactions, Trp86 was convened into Tyr, Phe, and Ala by site-specific mutagenesis. For this set of enzymes, a linear free-energy relationship is observed between the pK(i) values for inhibitions by the respective neutral and cationic transition state analogue inhibitors, m-tert-butyltrifluoroacetophenone and m-(N,N,N- trimethylammonio)-trifluoroacetophenone, which indicates that the free energy released on interaction of the quaternary, ammonium moiety with Trp86 arises about equally from cation-π and charge-independent interactions.

Full text of DOI:10.1021/ja9933588

VOLUME 122, NUMBER 13
APRIL 5, 2000
©
Copyright 2000 by the
American Chemical Society
Delineation and Decomposition of Energies Involved in Quaternary
Ammonium Binding in the Active Site of Acetylcholinesterase
Daniel M. Quinn,* Shawn R. Feaster, Haridasan K. Nair,⊥ ,† Nathan A. Baker,⊥ ,|
,
⊥
⊥ ,‡
§
§
Zoran Radi c´ , and Palmer Taylor
Contribution from the Department of Chemistry, The UniVersity of Iowa, Iowa City, Iowa 52242, and
Department of Pharmacology, UniVersity of California-San Diego, La Jolla, California 92093
ReceiVed September 15, 1999
Abstract: The quaternary ammonium binding locus in the active site of mammalian acetylcholinesterase is
subtended by the side chains of Trp86, Tyr133, Glu202, and Tyr337. Linear free-energy relationships define
the interactions involved in molecular recognition by mouse acetylcholinesterase of the quaternary ammonium
moiety of ligands. For substrates CH3C(dO)XCH2CH2Y [X ) O, Y ) CHMe2, or CH2CH3; X ) S, Y ) H,
+
+
NH Me2, or N Me3 ] and trifluoroacetophenone transition state analogue inhibitors m-YC6H4C(dO)CF3 [Y
)
+
H, Me, Et, iPr, tBu, CF3, NH2, NO2, NMe2, or N Me3], log(kcat/Km) and pKi depend linearly on the molar
refractivity, but not the hydrophobicity, of the substituents Y. These correlations indicate that, in the acylation
stage of catalysis, interactions in the quaternary ammonium binding locus stabilize the tetrahedral intermediate
5
TI
-1
(
as modeled by transition state analogue affinity) by (5 × 10 )-fold (∆∆G ) -32.5 kJ mol ) and the transition
4 ‡ -1
state by (2 × 10 )-fold (∆∆G ) -24.5 kJ mol ). To evaluate the contribution of cation-π interactions,
Trp86 was converted into Tyr, Phe, and Ala by site-specific mutagenesis. For this set of enzymes, a linear
free-energy relationship is observed between the pKi values for inhibitions by the respective neutral and cationic
transition state analogue inhibitors, m-tert-butyltrifluoroacetophenone and m-(N,N,N-trimethylammonio)-
trifluoroacetophenone, which indicates that the free energy released on interaction of the quaternary ammonium
moiety with Trp86 arises about equally from cation-π and charge-independent interactions.
1
Introduction
acceleration. Acetylcholinesterase (AChE ) is an attractive target
for such endeavors, since the catalytic power of the enzyme is
highly developed. AChE catalysis is minimally described by
the following three-step mechanism, in which E, EA, and F are
the free enzyme, Michaelis complex, and acylenzyme interme-
diate, respectively, A is the substrate, and P and Q are the
respective choline and acetate products:
A fundamental challenge in the biophysical attribution of the
structural origins of enzyme catalytic power is to describe the
nature and free energetic consequences of molecular interactions
that stabilize the transition state, and that thereby effect rate
*
To whom correspondence should be addressed: 319-335-1335 (phone);
3
19-335-1270 (fax); daniel-quinn@uiowa.edu.
⊥
The University of Iowa.
k1
k3
H O
2
‡
Current address: Walter Reed Army Institute of Research, Division
E + A y
\
z EA
9
8 F + P
8 E + Q
k2
k5
of Biochemistry, Washington, DC 20307.
†
Current address: AlliedSignal Inc., Buffalo, NY 1421.
Current address: Department of Chemistry, University of California-
|
k k
1
3
San Diego, La Jolla, CA 92093.
k /K )
(1)
cat
m
§
k2 + k3
University of California-San Diego.
1
0.1021/ja9933588 CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/16/2000

2
976 J. Am. Chem. Soc., Vol. 122, No. 13, 2000
Quinn et al.
Table 1. Acylation Rate Constants and Inhibition Constants for
Scheme 1. Substrates and Inhibitors of Mouse
Acetylcholinesterase
Wild-Type Mouse Acetylcholinesterase
substrates
inhibitorsa
(M- s-1
1
)
Y
K
i
X
Y
k
cat/K
m
5
O
O
S
S
S
S
iPr
Et
H
1.2 ( 0.2 × 10
H
Me
Et
iPr
tBu
CF
46 ( 2 nM
8700 ( 400
2.85 ( 0.05 nM
0.51 ( 0.03 nM
15.6 ( 0.7 pM
2200 ( 300
+
6
NH Me
2
3.1 ( 0.5 × 10
+
c
c
7
b
N Me
3
6 ( 3 × 10
3.7 ( 0.6 pM
+
7
N Me
3
4 ( 1 × 10
3
0.78 ( 0.03 nM
69.8 ( 0.6 nM
3.30 ( 0.07 nM
14.4 ( 0.8 pM
NH
NO
2
2
NMe
2
+
c
Equation 1 shows that the bimolecular rate constant kcat/Km takes
one from the free E and free A reactant state through the
transition state of the first irreversible step (i.e. k3) of the
mechanism. Hence, kcat/Km can be viewed as an acylation rate
constant. Values of the acylation rate constant kcat/Km that are
N Me
3
3
3.7 ( 0.4 fM
4.7 ( 0.4 fM
+
c
N Me
a
K
i
values for trifluoroketone inhibitors refer to the free ketone, and
consequently have been corrected for equilibrium hydration, as
described by Nair et al.5 b Taken from Radi c´ et al.
26c
c
Replicate values
+
3
were determined for the substrate and inhibitor that have Y ) N Me .
9
-1 -1
>
10 M
s
have been reported for (acetylthio)choline (ATCh)
2
turnover. Moreover, AChE accelerates acetylcholine turnover
by at least 10 -fold, which corresponds to -74 kJ mol of
13
3
-1
6
by Dougherty and co-workers, that aromatic functions interact
with organic cations via stabilizing cation-π interactions. We
4
7
free energy of stabilization of the transition state. The purpose
of this report is to describe a multifarious array of structure-
activity relationships that characterize the nature of the physical
forces that mouse AChE utilizes in molecular recognition of
the quaternary ammonium moiety of substrates and trifluoro-
acetophenone inhibitors. The crux of this characterization is the
determination of free energies of stabilization that provide a
substantial fraction of the catalytic power of AChE. Moreover,
unique structure-activity relationships are described in which
the enzyme structure is varied, and which initiate efforts to
attribute components of the above-mentioned free energies of
stabilization to individual active site amino acid constituents
of the quaternary ammonium binding locus.
describe below mutagenesis experiments that indicate that both
cation-π and dispersion interactions contribute to molecular
recognition in the quaternary ammonium binding locus of mouse
AChE.
As Figure 1A shows, linear free-energy correlations indicate
that both catalysis and inhibition depend on the molar refractivity
of the substituents Y. The plot for inhibitors was constructed
by first fitting data to the following multiple linear free-energy
relationship:
pK ) Fσ + m‚MR + C
(2)
i
m
The constants F and m are the sensitivities of inhibitor potency
pKi to the electronic substituent constant (σm) and molar
refractivity (MR) of the meta substituent. The value of F, 1.6
Results and Discussion
(
0.5, is similar to that determined from a correlation of logKHyd
Linear Free-Energy Relationships for Catalysis by and
Inhibition of Wild-Type AChE. Scheme 1 shows the families
of substituted trifluoroacetophenone inhibitors and acetate ester
substrates that were used to gauge molecular recognition in the
quaternary ammonium binding locus of mouse AChE. Table 1
lists enzyme kinetic parameters for the substrates and inhibition
constants for the inhibitors for wild-type mouse AChE-catalyzed
reactions. Previous work indicated that the free energies of
interaction in the quaternary ammonium binding loci of T.
californica and E. electrophorus AChEs depend linearly on the
molar refractivity of the substituent, which was interpreted as
indicating that dispersion interactions contribute to molecular
versus σm (not shown), where KHyd is the equilibrium constant
for hydration of the various trifluoroacetophenone inhibitors.
This similarity is consistent with the observation that, in the
X-ray structure of the complex of T. californica AChE and the
inhibitor m-(N,N,N-trimethylammonio)trifluoroacetophenone
(TMTFA), the γ-O of Ser200 of the active site is within covalent
8
bond distance of the carbonyl carbon of the inhibitor. The linear
correlation for inhibition in Figure 1A is plotted as pKi - Fσm
versus molar refractivity, a presentation that emphasizes the
dependence of inhibitor potency on interaction in the quaternary
ammonium binding locus only, and which facilitates comparison
with the correlation of log(kcat/Km) and molar refractivity for
substituted substrates. The linear free-energy correlations of
Figure 1A suggest that mouse AChE, like the marine AChEs,5
also utilizes dispersion interactions to stabilize ligand complexes.
5
recognition of the quaternary ammonium function. However,
a powerful case has been developed in recent years, particularly
(
1) Abbreviations used: AChE, acetylcholinesterase; Ala, alanine; ATCh,
(6) (a) Dougherty, D. A.; Stauffer, D. A. Science 1990, 250, 1558-
1560. (b) Kearney, P. C.; Mizoue, R. S.; Kumpf, R. A.; Forman, J. E.;
McCurdy, A.; Dougherty, D. A. J. Am. Chem. Soc. 1993, 115, 9907-919.
(c) Mecozzi, S.; West, A. P.; Dougherty, D. A. J. Am. Chem. Soc. 1996,
118, 2307-2308. (d) Dougherty, D. A. Science 1996, 271, 163-168. (e)
Ma, J. C.; Dougherty, D. A. Chem. ReV. 1997, 97, 1303-1324.
(7) Shafferman and co-workers have championed the view that cation-π
interactions with Trp86 are important for molecular recognition by AChE
of quaternary ammonium ligands: (a) Ordentlich, A.; Barak, D.; Kronman,
C.; Flashner, Y.; Leitner, M.; Segall, Y.; Ariel, N.; Cohen, S.; Velan, B.;
Shafferman, A. J. Biol. Chem. 1993, 268, 17083-17095. (b) Ordentlich,
A.; Barak, D.; Kronman, C.; Ariel, N.; Segall, Y.; Velan, B.; Shafferman,
A. J. Biol. Chem. 1995, 270, 2083-2091.
(
acetylthio)choline; DTNB, 5,5′-dithiobis(2-nitrobenzoic acid); Glu, glutamic
acid; MEPQ, 7-(methylethoxyphosphinyloxy)-N-methyl quinolinium iodide;
MR, molar refractivity; Phe, phenylalanine; Ser, serine; TBTFA, m-tert-
butyltrifluoroacetophenone; TMTFA, m-(N,N,N-trimethylammonio)trifluo-
roacetophenone; Trp, tryptophan; Tyr, tyrosine.
(
2) Nolte, H.-J.; Rosenberry, T. L.; Neumann, E. Biochemistry 1980,
1
9, 3705-3711.
(
3) Quinn, D. M. Chem. ReV. 1987, 87, 955-979.
(4) The free energy of stabilization of the transition state was calculated
‡
-1
-1
as ∆∆G ) -RT ln C, where R ) 8.314 J K mol , T ) 298 K, and C
is the catalytic acceleration. All free energy calculations in this paper utilize
T ) 298 K and 1 M standard states for all relevant species.
(
5) Nair, H. K.; Seravalli, J.; Arbuckle, T.; Quinn, D. M. Biochemistry
(8) Harel, M.; Quinn, D. M.; Nair, H. K.; Silman, I.; Sussman, J. L. J.
Am. Chem. Soc. 1996, 118, 2340-2346.
1
994, 33, 8566-8576.

Cation-π Interactions in AChE
J. Am. Chem. Soc., Vol. 122, No. 13, 2000 2977
fragments from aqueous solution to neat organic phases.10 This
analysis suggests that the dependence of ligand affinity on molar
refractivity (or surface area) for interaction in the quaternary
ammonium binding locus of AChE is a result of the interplay
of various factors, which include the classical entropy-driven
hydrophobic effect as well as short-range van der Waals
(dispersion) interactions. The factor contributed by the charge
of the ligand is delineated later.
Figure 1B shows that, as expected, there is a linear relation-
ship between log(kcat/Km) and pKi. The slope of this Thompson-
1
1
Bartlett plot, m ) 0.64 ( 0.08, indicates that amino acids of
the mouse AChE active site that stabilize the tetrahedral
trifluoroketone complex also stabilize the acylation transition
state, albeit to a lesser extent. Moreover, these results suggest
that the tetrahedral intermediate in the acylation stage of
catalysis, to which the trifluoroketone complex bears closest
structural resemblance, is stabilized by the quaternary am-
monium binding locus more than is the acylation transition state.
The overall change in kcat/Km as one proceeds from Y ) H to
+
4
Me3N is (2 × 10 )-fold, while the overall change in Ki across
7
the series of inhibitors is (1.1 × 10 )-fold. For inhibitors, the
multiple linear free energy correlation of Figure 1A indicates
7
5
that, of the (1.1 × 10 )-fold potency range, a factor of 5 × 10
arises from interactions in the quaternary ammonium binding
locus. The corresponding substituent effects on the stabilities
of the acylation transition state and transition state analogue
‡
-1
TI
complex are ∆∆G ) -24.5 kJ mol and ∆∆G ) -40.2 kJ
-
1
mol , respectively; interactions in the quaternary ammonium
binding locus contribute a factor of -32.5 kJ mol- to ∆∆G .
1
TI
Linear Free-Energy Relationships for Wild-Type and
Mutant Acetylcholinesterases. The quaternary ammonium
binding locus whose function is gauged by the above-described
experiments consists of the amino acid side chains of Trp86,
Figure 1. Linear free-energy correlations for substrates and inhibitors
of mouse AChE. (A) Circles and diamonds are for inhibition by meta-
substituted trifluoroacetophenones and for substrate hydrolysis, respec-
8
,12
Tyr133, Glu202, and Tyr337
and three water molecules, as
8
shown in Figure 2. The closest interactions between AChE and
the quaternary ammonium function involve Trp86 and Glu202.
Free energy relationships for native mouse AChE, here denoted
Trp86, and the mutants Trp86Tyr, Trp86Phe, and Trp86Ala
provide a method for delineating the contributions of W86 of
this binding locus to the transition state analogue binding free
energies alluded to in the preceding paragraph. Michaelis-
Menten kinetic parameters and transition state analogue inhibi-
tion constants are given in Table 2 for the three Trp86 mutants
of mouse AChE. Mouse Trp86Ala AChE has a kcat/Km that is
1500-fold less than that of the native enzyme. Therefore,
tively. Inhibition data were fit to eq 2 of the text to give m ) 0.29 (
2
0
.03, F ) 1.6 ( 0.5, C ) 6.5 ( 0.4, and r ) 0.945. A rearranged
form of eq 2 that emphasizes the linear free-energy relationship between
inhibitor potency and molar refractivity of inhibitor substituents is
plotted, as described in the text. Substrate hydrolysis data were fit to
the following equation: log(kcat/K ) ) m‚MR + C. The parameters of
m
2
the fit are m ) 0.21 ( 0.03, C ) 2.8 ( 0.5, and r ) 0.886. (B)
Correlation between catalytic rate constant and inhibitor potency. The
solid line is a fit to the following equation: log(kcat/K
m
) ) m‚pK + C.
i
The parameters of the fit are m ) 0.64 ( 0.08, C ) -1.4 ( 0.8, and
2
r ) 0.934.
-
1
interaction with Trp86 contributes 18 kJ mol to stabilization
of the acylation transition state. This free energy change is about
On the other hand, correlations with the hydrophobic substituent
constant are linear for alkyl substituents but fail for nitrogen-
containing substituents (plot not shown), as observed previously
75% of that determined above for native mouse AChE-catalyzed
hydrolysis of substituted substrates, which provides an indication
of the dominant contribution of Trp86 to molecular recognition
in the quaternary ammonium binding locus.
5
for marine AChEs. However, one cannot exclude the hydro-
phobic effect altogether from contributing to molecular recogni-
tion in the quaternary ammonium binding locus, since a multiple
linear free-energy relationship among pKi, σm, and the substitu-
ent surface area performs nearly as well as the fit to eq 2
(10) (a) Reynolds et al. estimated that the free energy change per solute
surface area for transfer of hydrocarbons from aqueous medium to a
-
1
-2
hydrocarbon solvent is in the range -20 to -25 cal mol
Å
(-84 to
): Reynolds, J. A.; Gilbert, D. B.; Tanford, C. Proc.
Natl. Acad. Sci. U.S.A. 1974, 71, 2925-2927. (b) Herman reported a value
-
1
-2
-
105 J mol
Å
9
described above. However, the sensitivity of pKi to substituent
-
1
-2
-1
-2
-1
-2
surface area is -270 ( 30 J mol Å , a value that is well
of -30 cal mol
Å
(-126 J mol
Å ) for transfer of alkylbenzenes
from aqueous medium to neat aromatic hydrocarbon phases, determined
above estimates for the transfer of hydrophobic molecular
from solubility measurements: Herman, R. B. J. Phys. Chem. 1972, 76,
-
1
-2
2
754-2759. (c) A value for the hydrophobic effect of -47 cal mol
Å
-
1
-2
(9) Inhibition data were fit to the following equation, in which A is the
(-197 J mol
Å ) arises when solute-solvent size differences are taken
into account: Sharp, K. A.; Nicholls, A.; Fine, R. F.; Honig, B. Science
1991, 252, 106-109.
surface area of the meta substituent: pKi ) Fσm + a‚A + C. The parameter
-
2
values determined by the fit were F ) 2.2 ( 0.6, a ) 0.048 ( 0.006 Å
,
and C ) 4.5 ( 0.7. The correlation coefficient for this fit was r ) 0.962,
only marginally less than r ) 0.965 for the fit to eq 2. The free energy per
substituent surface area for transfer from aqueous solution to the active
(11) (a) Thompson, R. C. Biochemistry 1973, 12, 47-51. (b) Bartlett,
P. A.; Marlowe, C. Biochemistry 1983, 22, 4618-4624.
(12) Quinn, D. M. In ComprehensiVe Toxicology; Vol. 3, Biotransfor-
mation; Guengerich, F. P., Ed.; Elsevier: New York, 1997; pp 243-264.
-
1
-2
site of AChE is -2.3025RT‚a ) -270 ( 30 J mol
Å .

2978 J. Am. Chem. Soc., Vol. 122, No. 13, 2000
Quinn et al.
Figure 2. Crossed stereoview of the interaction of TMTFA in the
quaternary ammonium binding locus of Torpedo californica AChE.8
The residues that make close contacts with the quaternary ammonium
function of TMTFA include Trp86(84), Phe337(330), Glu202(199), and
three water molecules. The sequence numbers that follow the amino
acid abbreviation are those of mammalian and Torpedo californica
AChEs, respectively, the latter in italics and parentheses.29 The
mammalian AChE numbering scheme is used in the figure. In
mammalian AChE, Tyr replaces Phe at position 337.
Table 2. Acylation Rate Constants and Inhibition Constants For
Mutant Mouse Acetylcholinesterases
(M- s-1)a
1
K
(Y ) tBu)b
K
(Y ) N Me
3
)b
+
enzyme
W86c
k
cat/K
m
i
i
7
6 ( 3 × 10
3.7 ( 0.6 pM
3.7 ( 0.4 fM
4.7 ( 0.4 fM
0.11 ( 0.01 pM
0.37 ( 0.04 pM
6.7 ( 0.3 pM
7
4
( 1 × 10
6
6
4
W86F
W86Y
W86A
3.4 ( 0.4 × 10
2.0 ( 0.2 × 10
3.2 ( 0.8 × 10
6 ( 1 pM
33 ( 8 pM
80 ( 30 pM
a
The bimolecular rate constant kcat/K
m
is that for turnover of
(
acetylthio)choline, determined under the experimental conditions
b
outlined in the Methods subsection of the Experimental Section. Data
for the uncharged inhibitor, m-tert-butyltrifluoroacetophenone, are taken
2
6c
from Radi c´ et al.
i
K values for trifluoroketone inhibitors refer to the
free ketone, and consequently have been corrected for equilibrium
5
c
hydration, as described by Nair et al. Data for wild-type (W86) AChE
are reproduced from Table 1.
The slope of the linear free energy relationship in Figure 3A
between log(kcat/Km) for turnover of ATCh and pKi for inhibition
by TMTFA is m ) 1.0 ( 0.1. Hence, the sensitivities of catalysis
and transition state analogue binding to variation of the AChE
substituent at position 86 are the same. This comparison
indicates that, inasmuch as the tetrahedral trifluoroketone
complex resembles the tetrahedral intermediate in the acylation
Figure 3. Linear free-energy correlations for wild-type and mutant
mouse AChEs. Except where noted, the data used to construct these
correlations are contained in Tables 1 and 2. (A) The bimolecular rate
constant for AChE-catalyzed hydrolysis of ATCh is correlated with
inhibitory potency of TMTFA. The points from left to right are those
for Trp86Ala, Trp86Tyr, Trp86Phe, and wild-type AChE (rightmost
two points). The solid line is a fit to the equation in the legend of
8
stage of catalysis, the influence of Trp86 on the stabilities of
the tetrahedral intermediate and the acylation transition state is
the same. Figure 3B shows that, as for variation of substrate
and inhibitor substituents in Figure 1A, there is a linear free-
energy relationship between log(kcat/Km) and molar refractivity
of the varied amino acid at position 86. This correlation contains
data not only from the current study but also from kcat/Km values
reported by Shafferman et al.13 for wild-type human AChE and
the corresponding Trp86Phe, Trp86Ala, and Trp86Glu mutants.
Inclusion of the human AChE data underscores the general
applicability of the linear free-energy correlation to probing
molecular recognition in the quaternary ammonium binding
locus.
Figure 1B, and the parameters of the fit are m ) 1.0 ( 0.1, C ) -6
2
(
1, and r ) 0.969. (B) The bimolecular rate constant for AChE-
catalyzed hydrolysis of ATCh is correlated with the molar refractivity
of the amino acid side chain at position 86. Circles are from this study,
and from left to right are for Trp86Ala, Trp86Tyr, Trp86Phe, and wild-
type mouse AChEs. Squares were calculated from data of Shafferman
et al.13 for native and mutant human AChEs, and from left to right are
for Trp86Ala, Trp86Glu, Trp86Phe, and wild-type enzymes. The solid
line is a fit to the equation in the legend of Figure 1A, and the
2
parameters are m ) 0.100 ( 0.008, C ) 3.9 ( 0.2, and r ) 0.960.
Figure 4 demonstrates that more than hydrophobic and
dispersion interactions are involved in molecular recognition
in the quaternary ammonium binding locus. For wild-type AChE
and the three Trp86 mutants, a linear free-energy relationship
is observed between pKi values for trifluoroacetophenone
transition state analogue inhibitors that have isosteric uncharged
The slope of this correlation is m ) 0.42 ( 0.07. Since the
uncharged and charged substituents have similar molar refrac-
14
tivities (19.62 and 21.20, respectively ), a slope of 0.93 is
expected if dispersion interactions alone are important for
quaternary ammonium recognition by Trp86. That m ) 0.42
suggests that the free energy released on interaction of the
quaternary ammonium moiety with Trp86 is about evenly
divided between contributions that are dependent on substituent
charge on one hand and those that are independent of charge
(m-tert-butyl) and charged (m-trimethylammonio) substituents.
(13) Shafferman, A.; Ordentlich, A.; Barak, D.; Kronman, N.; Ariel, M.;
Leitner, M.; Segall, Y.; Bromberg, A.; Reuveny, S.; Marcus, D.; Bino, T.;
Lazar, A.; Cohen, S.; Velan, B. In Enzymes of the Cholinesterase Family;
Quinn, D. M., Balasubramanian, A. S., Doctor, B. P., Taylor, P., Eds.;
Plenum: New York and London, 1995; pp 189-196.
(14) Hansch, C.; Leo, A. Substituent Constants for Correlation Analysis
in Chemistry and Biology; John Wiley: New York, 1979.

Cation-π Interactions in AChE
J. Am. Chem. Soc., Vol. 122, No. 13, 2000 2979
other investigations. Pressman et al. studied reactions of
antibodies homologous to quaternary ammonium haptenic
groups. They found that a hapten that contained the p-
azophenyltrimethylammonium moiety had a 15.5-fold higher
affinity than did the isosteric hapten that contained the p-azo-
tert-butylbenzene moiety. The corresponding Coulomb interac-
-
1
tion energy is -6.3 kJ mol , a value that is in reasonable
agreement with the estimates generated in this paper.
+
Ab initio calculations of cation-π interactions between Na
15
and aromatics indicate that the energy of complexation is about
evenly split between charge-dependent and charge-independent
contributions, just as observed experimentally herein. This
situation also obtains for calculated energies for interaction of
1
6
benzene with ammonium ions. The free energy difference of
-1
-
18.3 kJ mol reported herein for binding of TMTFA to wild-
type AChE versus the Trp86Ala mutant is in good agreement
with calculated free energies of complexation of tetramethyl-
-1 16,17
ammonium ion and benzene in the gas phase, -17 kJ mol ,
-
1 18
in aqueous solution, -13.8 kJ mol , and with the corre-
Figure 4. Linear free-energy correlation between inhibitory potencies
of uncharged and cationic transition state analogue inhibitors of wild-
type and mutant mouse AChEs. pK values for the m-tert-butyl and
i
sponding experimental free energy of association in the gas
-1 19
phase, -14.6 kJ mol . From the results described herein, both
log(kcat/Km) and pKi increase in the sequence Tyr < Phe , Trp
for the various aromatic residues at position 86. This trend agrees
with the trend in calculated binding energies for complexation
m-trimethylammonio inhibitors (TBTFA and TMTFA, respectively)
were calculated from data in Tables 1 and 2 and in Radi c´ et al.2 The
clusters of points are, from left to right, for Trp86Ala, Trp86Tyr,
Trp86Phe, and wild-type enzymes. The solid line is a fit to the following
6c
+
+
of Na and Me4N with the series of aromatics phenol, benzene,
2
0,21
and indole.
equation: pK
i
(TBTFA) ) m‚pK (TMTFA) + C. The parameters of the
i
2
22
fit are m ) 0.42 ( 0.07, C ) 5.4 ( 0.9, and r ) 0.854.
Zhong et al. used the in vivo nonsense-suppression method
to introduce unnatural amino acids into specific positions in
the nicotinic acetylcholine receptor. Large effects of unnatural
amino acid mutagenesis of Trp149 of the R subunit of the
receptor were noted, whereas mutagenesis at alternate Trp
residues had diminutive effects. Of particular note was the linear
free-energy correlation between the EC50 values for receptor
interaction with acetylcholine and the cation-π binding energies,
calculated by ab initio quantum mechanics, for the side chains
of Trp149 and various fluorinated analogues thereof. The range
of EC50 values was about 2 orders of magnitude, which
but dependent on the size of the substituent on the other. It is
reasonable to conclude that the 18 kJ mol-1 change in free
energy on interaction with Trp86 is about evenly split between
cation-π and dispersion/hydrophobic interactions.
Comparison with Other Studies. The contribution of charge-
dependent interactions to the total interaction free energy for
the quaternary ammonium binding site can be estimated in the
following way. As mentioned earlier, the affinity range for meta-
7
substituted trifluoroacetophenones is 1.1 × 10 , of which a factor
5
TI
-1
-1
of 5.5 × 10 (corresponding ∆∆G ) -32.5 kJ mol ) arises
from interaction with the quaternary ammonium binding locus
of AChE. From the Ki values of Table 1 one calculates that the
corresponds to a ∆∆G of ∼ -11 kJ mol , again in substantial
agreement with estimates generated herein of the interaction
energy of the quaternary ammonium moiety with Trp86 of
AChE.
+
quaternary ammonium inhibitor TMTFA (Y ) N Me3) binds
Barak et al.23 recently reported an insightful study of the effect
of mutation of Trp86 on the rates of aging of the structurally
varied human AChE phosphonyl conjugates of Scheme 2. The
first-order aging rate constants decrease in the order Trp86 >
Phe86 > Ala86; this trend is observed for all degrees of
branching at the â-carbon. However, the range of aging rate
constants for Trp86 versus Ala86 AChE decreased systemati-
more tightly than the isosteric neutral inhibitor TBTFA (Y )
tert-butyl) by a factor of 880, which corresponds to an
-1
interaction free energy difference of -16.8 kJ mol . A portion
of this higher affinity is due to the Hammett substituent effect
on the stability of the hemiketal adduct between the inhibitor
and the γ-O of Ser203(200). This factor is estimated as
-
1
∆
∆GHammett ) -2.303RT‚F(σTMTFA - σTBTFA) ) -10 kJ mol ,
14
+
where σTMTFA and σTBTFA are the σm values for the m-Me3N
(
15) Electrostatics contribute 58% and 59% to the calculated interaction
and the m-tert-butyl substituents of TMTFA and TBTFA,
respectively. This calculation leaves a difference of -6.8 kJ
+
energies of Na with benzene and phenol, respectively. See ref 6c.
(16) Kim, K. S.; Lee, J. Y.; Lee, S. J.; Ha, T.-K.; Kim, D. H. J. Am.
Chem. Soc. 1994, 116, 7399-7400.
-
1
mol of interaction free energy that arises specifically from
the positive charge of the quaternary ammonium moiety.
The affinity of wild-type mouse AChE for the quaternary
ammonium inhibitor TMTFA is 1600-fold greater than that of
the Trp86Ala mutant, which corresponds to a change in
(
17) Lee, J. Y.; Lee, S. J.; Choi, H. S.; Cho, S. J.; Kim, K. S.; Ha, T.-K.
Chem. Phys. Lett. 1995, 232, 67-71.
18) Duffy, E. M.; Kowalczyk, P. J.; Jorgensen, W. L. J. Am. Chem.
Soc. 1993, 115, 9271-9275.
19) Meot-Ner, M.; Deakyne, C. A. J. Am. Chem. Soc. 1985, 107, 469-
74.
20) Mecozzi, S.; West, A. P.; Dougherty, D. A. Proc. Natl. Acad. Sci.
U.S.A. 1996, 93, 10566-10571.
21) Basch, H.; Stevens, W. J. J. Mol. Struct. (THEOCHEM) 1995, 338,
03-315.
22) Zhong, W.; Gallivan, J. P.; Zhang, Y.; Li, L.; Lester, H. A.;
(
(
4
-1
interaction free energy of -18.3 kJ mol . The linear free energy
(
-
1
correlation of Figure 4 suggests that -10.6 kJ mol of this
free energy can be attributed to the charge of the quaternary
ammonium inhibitor. Therefore, this estimate and that in the
preceding paragraph suggest that positive charge accounts for
(
3
(
Dougherty, D. A. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 12088-12093.
(23) Barak, D.; Ordentlich, A.; Segall, Y.; Velan, B.; Benschop, H. P.;
De Jong, L. P. A.; Shafferman, A. J. Am. Chem. Soc. 1997, 119, 3157-
-
1
∼
-7 to -11 kJ mol of the energy that is released on binding
of ligands to the quaternary ammonium binding locus of AChE.
It is instructive to compare the interaction energies estimated
in the preceding two paragraphs with similar estimates from
3
158. The aging of phosphyl conjugates of AChE is the hydrolytic cleavage
of the O-CR bond of Scheme 2, a reaction that is thought to involve a
carbocation intermediate and/or carbocationic transition state.

2980 J. Am. Chem. Soc., Vol. 122, No. 13, 2000
Quinn et al.
Scheme 2. Phosphonyl Conjugates of Acetylcholinesterase
Studied by Barak et al.23
following sources: NaH
2
PO
4
‚H
2
O and Na
2
HPO
4
, Sigma Chemical Co.;
NaCl, Fisher Scientific. Water for buffer preparation was distilled and
then deionized by passage through a Barnstead D8922 mixed-bed ion-
exchange column (Sybron Corp.). Recombinant mouse AChEs were
expressed and characterized as described by Radi c´ et al. 7-(Methyl-
ethoxyphosphinyloxy)-N-methyl quinolinium iodide (MEPQ) was
obtained gratis from Yacov Ashani of the Israel Institute for Biological
Research, Ness-Ziona.
2
6
Methods. Molar refractivities and σ
m
values for inhibitor and
1
4
substrate substituents were taken from Hansch and Leo. Substituent
van der Waals surface areas were calculated by using the MOLCAD
feature of the molecular modeling program Sybyl 6.1 (Tripos Associates
Inc., St. Louis, MO). Reactions were run at 25.0 ( 0.2 °C in 0.05 M
sodium phosphate buffers, pH 7.00, that contained sufficient ATCh
and/or NaCl to give an ionic strength (I) of 0.13. AChE-catalyzed
reactions were monitored on HP8452 or Beckman DU40 UV-visible
spectrophotometers. Hydrolyses of the thioesters ATCh, ethyl thioace-
tate, and 2-(N,N-dimethylamino)ethyl thioacetate were followed at 412
nm by coupling the released thiol product with DTNB, as described
by Ellman et al. for ATCh.27 Hydrolyses of the oxyesters butyl acetate
and isoamyl acetate were followed at 400 nm by a coupled assay that
cally with decreasing â branching, from 1120-fold when N )
to 60-fold when N ) 0. Barak et al.2 interpret these results
3
3
as indicating that aromatic residues at position 86, and in
particular Trp, stabilize the carbocationic transition state of the
aging reaction by cation-π interactions. For all but the least
branched conjugate, log(ka) is a linear function of the molar
2
refractivity of the amino acid at position 86; r values are 0.99,
0
.98, and 0.97 for N ) 3, 2, and 1, respectively. The activation
energy differences for the wild-type (i.e. Trp86) and Ala86
-
1
enzymes are -17.4, -14.7, and -14.7 kJ mol , respectively,
in good agreement with our estimate of the energetics of
interaction of Trp86 with quaternary ammonium ligands.
5
utilized p-nitrophenol as the indicator, as previously described. Buffer
pH was measured on a Corning Model 125 pH meter that was equipped
with a glass combination electrode. Enzyme concentrations were
determined by active site titration with MEPQ, as described by Ashani
2
2
The studies reported herein, and those of Zhong et al. and
23
Barak et al., show that linear free-energy relationships (LFERs)
are observed either when ligand substituent or enzyme/receptor
structure is systematically varied. The use of site-directed
mutants to construct LFERs adds an important new tool to the
arsenal of approaches for structure-function investigations of
cholinesterases.
28
et al. Michaelis-Menten kinetic parameters were determined by fitting
initial velocity data to an equation that includes substrate inhibition at
high substrate concentrations, as described by Radi c´ et al.2 Slow, tight-
binding inhibitions of wild-type and mutant mouse AChEs by TMTFA
6a
5
24
and TBTFA were characterized as described by Nair et al.
Acknowledgment. This work was supported by NIH grants
NS21334 to D.M.Q. and GM18360 to P.T.
Experimental Section
JA9933588
Materials. Substrates were used as received from the following
sources: (acetylthio)choline iodide, Sigma Chemical Co.; butyl acetate
and isoamyl acetate, Aldrich Chemical Co.; ethyl thioacetate, MTM
Research Chemicals; 2-(N,N-dimethylamino)ethyl thioacetate, from a
(
25) (a) Nair, H. K.; Lee, K.; Quinn, D. M. J. Am. Chem. Soc. 1993,
1
1
15, 9939-9941. (b) Nair, H. K.; Quinn, D. M. Bioorg. Med. Chem. Lett.
993, 3, 2619-2622.
(26) (a) Radi c´ , Z.; Pickering, N.; Vellom, D. C.; Camp, S.; Taylor, P.
5
previous study. Trifluoroacetophenone was used as received from
Biochemistry 1993, 32, 12074-12084. (b) Vellom, D. C.; Radi c´ , Z.; Li,
Aldrich Chemical Co.; meta-substituted trifluoroacetophenone inhibitors
were synthesized and structurally characterized as described previ-
Y.; Pickering, N. A.; Camp. S.; Taylor, P. Biochemistry 1993, 32, 12-17.
(
c) Radi c´ , Z.; Quinn, D. M.; Vellom, D. C.; Camp, S.; Taylor, P. J. Biol.
5
,25
Chem. 1995, 270, 20391-20399.
27) Ellman, G. L.; Courtney, K. D.; Andres, V., Jr.; Featherstone, R.
M. Biochem. Pharmacol. 1961, 7, 88-95.
28) Levy, D.; Ashani, Y. Biochem. Pharmacol. 1986, 35, 1079-1085.
ously. DTNB and p-nitrophenol were used as received from Sigma
(
Chemical Co. Buffer components were used as received from the
(
(
24) The structure-function analysis of sets of mutant enzymes by the
(29) Massouli e´ , J.; Sussman, J. L.; Doctor, B. P.; Sorey, H.; Velan, B.;
Cygler, M.; Rotundo, R.; Shafferman, A.; Silman, I.; Taylor, P. In
Multidisciplinary Approaches to Cholinesterase Functions; Shafferman, A.,
Velan, B., Eds.; Plenum: New York and London, 1992; pp 285-288.
use of linear free-energy relationships (LFERs) was pioneered by Alan
Fersht: Fersht, A. R.; Leatherbarrow, R. J.; Wells, T. N. C. Biochemistry
1
987, 26, 6030-6043.