Transition state structure and rate determination for the acylation stage of acetylcholinesterase catalyzed hydrolysis of (acetylthio)choline

Article abstract of DOI:10.1021/ja9933590

Rate-limiting steps and transition state structure for the acylation stage of acetylcholinesterase-catalyzed hydrolysis of (acetylthio)choline have been characterized by measuring substrate and solvent isotope effects and viscosity effects on the bimolecular rate constant k(E) (=k(cat)/K(m)). Substrate and solvent isotope effects have been measured for wild-type enzymes from Torpedo californica, human and mouse, and for various active site mutants of these enzymes. Sizable solvent isotope effects, (D₂O)k(E) ~ 2, are observed when substrate β-deuterium isotope effects are most inverse, (βD)k(E) = 0.95; conversely, reactions that have (D₂O)k(E) ~ 1 have substrate isotope effects of (βD)k(E) = 1.00. Proton inventories of k(E) provide a quantitative measure of the contributions by the successive steps, diffusional encounter of substrate with the active site and consequent chemical catalysis, to rate limitation of the acylation stage of catalysis. For reactions that have the largest solvent isotope effects and most inverse substrate isotope effects, proton inventories are linear or nearly so, consistent with prominent rate limitation by a chemical step whose transition state is stabilized by a single proton bridge. Reactions that have smaller solvent isotope effects and less inverse substrate isotope effects have nonlinear and upward bulging proton inventories, consistent with partial rate limitations by both diffusional encounter and chemical catalysis. Curve fitting of such proton inventories provides a measure of the commitment to catalysis that is in agreement with the effect of solvent viscosity on k(E) and with the results of a double isotope effect measurement, wherein (βD)k(E) is measured in both H₂O and D₂O. The results of these various experiments not only provide a model for the structure of the acylation transition state but also establish the validity of solvent isotope effects as a tool for quantitative characterization of rate limitation for acetylcholinesterase catalysis.

Full text of DOI:10.1021/ja9933590

J. Am. Chem. Soc. 2000, 122, 2981-2987
2981
Transition State Structure and Rate Determination for the Acylation
Stage of Acetylcholinesterase Catalyzed Hydrolysis of
(Acetylthio)choline
Siobhan Malany,†,# Monali Sawai,† R. Steven Sikorski,† Javier Seravalli,†,
Daniel M. Quinn,*,† Zoran Radic´,‡ Palmer Taylor,‡ Chanoch Kronman,§
Baruch Velan,§ and Avigdor Shafferman§
Contribution from the Department of Chemistry, The UniVersity of Iowa, Iowa City, Iowa 52242,
Department of Pharmacology, UniVersity of California-San Diego, La Jolla, California 92093, and
Israel Institute for Biological Research, Ness-Ziona 70450, Israel
ReceiVed September 15, 1999
Abstract: Rate-limiting steps and transition state structure for the acylation stage of acetylcholinesterase-
catalyzed hydrolysis of (acetylthio)choline have been characterized by measuring substrate and solvent isotope
effects and viscosity effects on the bimolecular rate constant kE ()kcat/Km). Substrate and solvent isotope effects
have been measured for wild-type enzymes from Torpedo californica, human and mouse, and for various
active site mutants of these enzymes. Sizable solvent isotope effects, D OkE ∼ 2, are observed when substrate
2
â-deuterium isotope effects are most inverse, âDkE ) 0.95; conversely, reactions that have D OkE ∼ 1 have
2
substrate isotope effects of âDkE ) 1.00. Proton inventories of kE provide a quantitative measure of the
contributions by the successive steps, diffusional encounter of substrate with the active site and consequent
chemical catalysis, to rate limitation of the acylation stage of catalysis. For reactions that have the largest
solvent isotope effects and most inverse substrate isotope effects, proton inventories are linear or nearly so,
consistent with prominent rate limitation by a chemical step whose transition state is stabilized by a single
proton bridge. Reactions that have smaller solvent isotope effects and less inverse substrate isotope effects
have nonlinear and upward bulging proton inventories, consistent with partial rate limitations by both diffusional
encounter and chemical catalysis. Curve fitting of such proton inventories provides a measure of the commitment
to catalysis that is in agreement with the effect of solvent viscosity on kE and with the results of a double
isotope effect measurement, wherein âDkE is measured in both H2O and D2O. The results of these various
experiments not only provide a model for the structure of the acylation transition state but also establish the
validity of solvent isotope effects as a tool for quantitative characterization of rate limitation for acetyl-
cholinesterase catalysis.
Introduction
and Q are the successive choline and acetate products. The
bimolecular rate constant kE (≡kcat/Km) for this mechanism is
given by eq 1:
The physiological reaction catalyzed by acetylcholinesterase
(AChE1) is the hydrolysis of the neurotransmitter acetylcholine
(ACh). This reaction proceeds, after reversible substrate binding,
by successive acylation and deacylation steps, as outlined in
the following kinetic mechanism:
k1k3
kE )
(1)
k2 + k3
k1
k3
k5
The rate constant kE is obtained by dividing the pseudo-first-
order rate constant Vmax/Km by [E]T, and therefore kinetic isotope
and other effects on Vmax/Km are effects on kE. As eq 1 shows,
kE is a function of microscopic rate constants that convert the
free E + free A reactant state to the successive transition states
of the substrate binding step (rate constant k1) and the chemical
step that produces the acylenzyme intermediate (rate constant
k3). As for many enzymes whose catalytic power is highly
evolved, kE for AChE approximates2 the diffusion-controlled
limit of Eigen and Hammes.3 Because of this, kE, which is
frequently referred to as the observed acylation rate constant,
E + A y
\
kz EA
98 F + P H O8 E + Q
2
2
E, A, EA, and F are the free enzyme, free substrate, Michaelis
complex, and the acylenzyme intermediate, respectively, and P
* To whom correspondence should be addressed: 319-335-1335 (phone);
319-335-1270 (fax); daniel-quinn@uiowa.edu.
† The University of Iowa.
# Current address: Department of Pharmacology, University of California-
San Diego, La Jolla, CA 92093.
Current address: Department of Biochemistry, University of Nebraska,
The Beadle Center N113, Lincoln, NE 68588.
‡ University of California-San Diego.
§ Israel Institute for Biological Research.
(2) (a) Bazelyansky, M.; Robey, E.; Kirsch, J. F. Biochemistry 1986,
25, 125-130. (b) Hasinoff, B. B. Biochim. Biophys. Acta 1982, 704, 52-
58. (c) Nolte, H.-J.; Rosenberry, T. L.; Neumann, E. Biochemistry 1980,
19, 3705-3711. (d) Quinn, D. M. Chem. ReV. 1987, 87, 955-979.
(3) Eigen, M.; Hammes, G. G. AdV. Enzymol. Relat. Subj. Biochem. 1963,
25, 1-38.
(1) Abbreviations: ACh, acetylcholine; AChE, acetylcholinesterase; Ala,
alanine; ATCh, (acetylthio)choline; DTNB, 5,5′-dithiobis(2-nitrobenzoic
acid); Glu, glutamate; Gln, glutamine; His, histidine; HuAChE, human
AChE.; MAChE, mouse AChE; Ser, serine; TcAChE, Torpedo californica
AChE; TX100, Triton X-100.
10.1021/ja9933590 CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/16/2000

2982 J. Am. Chem. Soc., Vol. 122, No. 13, 2000
Malany et al.
Table 1. Isotope and Viscosity Effects for Wild-Type and Mutant AChEs
D OkE
âDkE
Cc
D Ok3
k
3E
b
b
c
âD
d
relative kea
2
2
enzyme
T. californica wild-type
T. californica Glu199Gln
human wild-type
1.0
0.01
0.07
1.47 ( 0.06
2.08 ( 0.07
1.62 ( 0.04
1.00 ( 0.01
0.98 ( 0.01
0.97 ( 0.01e
0.95 ( 0.01e
0.96 ( 0.02
0.95 ( 0.02
1.000 ( 0.005
0.97 ( 0.01
3 ( 1
3.0 ( 0.6
0.85 ( 0.06
0.88 ( 0.04
0.87 ( 0.06
0.93 ( 0.03
1.2 ( 0.2f
1.8 ( 0.3
0.4 ( 0.2
human Glu202Gln
human Glu202Ala
mouse wild-type
mouse Glu202Gln
0.009
0.0004
0.05
1.84 ( 0.08
2.1 ( 0.1
1.26 ( 0.04
1.5 ( 0.1
3.1 ( 0.5
2.4 ( 0.4
g
∞
0.001
a Relative kE values were obtained by dividing measured kE values by the kE ) 1.0 × 109 M-1 s-1 for T. californica AChE. All kE values for these
comparisons were extrapolated to zero ionic strength, as described by Quinn et al.34 b Parameters were measured at 25.0 ( 0.2 °C in the following
buffers: TcAChE, 0.1 M sodium phosphate buffer, pH 7.16 and pD 7.78, µ ) 0.22; HuAChE, 0.1 M sodium phosphate buffer, pH 7.30 and pD
7.80, µ ) 0.35; MAChE, 0.04 M sodium phosphate buffer, pH 7.50 and pD 8.10, µ ) 0.10. c Except where indicated (see footnote f), commitment
values and intrinsic solvent isotope effects were calculated by fitting proton inventories to eq 4 of the text, as demonstrated in Figure 1. For the
wild-type human AChE reaction, the commitment in D2O was estimated in the text as CD O ) 1.2 ( 0.5. d Calculated according to eq 7, as explained
2
in the text. e The respective isotope effects were measured in buffered H2O and D2O. The Student t-test indicates that the isotope effects are
f
different at the 99.5% confidence level. Determined from dependence of kE on solvent viscosity; see Figure 2. g Curve-fitting of viscosity dependencies
as in Figure 2 gave very large values of C (>105), consistent with sole rate limitation by substrate binding, i.e. k1.
is not likely to be rate limited by just one of the successive
transition states.
These distinct experimental approaches provide complementary
information that allows postulation of a model for the geometry
of the acylation transition state. Implicit in the analyses described
herein is that mammalian and marine AChEs have transition
states that are structurally similar. This assumption is supported
by the observations that the active sites of T. californica and
mouse AChEs are similarly situated in the enzyme interiors at
the bottom of 20 Å deep active site gorges, and that the amino
acids that line the gorges are highly conserved.9 Solvent isotope
effect measurements in mixed H2O-D2O buffers (called proton
inventories10) provide a quantitative accounting of rate limitation
in the acylation stage of catalysis. The results of proton inventory
experiments are corroborated by substrate isotope and viscosity
effects. Correlations between these different experimental ap-
proaches establish the validity of solvent isotope effects as a
tool for characterizing the dynamics of AChE catalysis.
The surrogate substrate (acetylthio)choline (ATCh) provides
a convenient alternative to the physiological substrate ACh, since
hydrolysis of ATCh can be monitored by the sensitive and
continuous spectrophotometric Ellman assay.4 The values of kcat
and kcat/Km for Electrophorus electricus AChE-catalyzed hy-
drolyses of the two substrates are nearly identical,2d and kcat
values for the two substrates are rate-limited by acylation and
deacylation to comparable extents.5 Consequently, AChE ca-
talysis is impervious to the OfS elemental substitution in the
etherial position of the ester function of the substrate.
A reasonable strategy for increasing the rate limitation of kE
by the k3 step is to utilize mutant AChEs in which the effect of
mutation is to selectively destabilize the transition state of the
chemical step. Shafferman et al.6 suggest that reductions in kE
that accompany replacement of Glu202 of the active site of
human AChE with Gln, Ala, or Asp arise from slowing of the
acylation step k3, and ascribe this effect to disruption in the
mutant enzymes of an H-bond network that positions catalytic
and ligand binding residues. For the corresponding Glu199 to
Gln mutant of Torpedo californica AChE, Radic´ et al.7 also
suggests that the catalytic efficiency is affected. Consequently,
Glu202(199)8 mutants are utilized herein to increase rate
limitation by the chemical step in the acylation stage of AChE
catalysis.
Results and Discussion
Isotope Effect Comparisons for Wild-Type and Mutant
Enzymes. Table 1 lists the results of isotope effect and viscosity
experiments on various AChEs. Solvent isotope effects are
expressed as the ratio of kE values measured in equivalently
D2O
buffered solutions of H2O and D2O,11 i.e. D OkE ) kEH O/kE
;
2
2
â-deuterium isotope effects are ratios of kE values for hydrolyses
of (acetyl-1H3-thio)choline and (acetyl-2H3-thio)choline; i.e. âDkE
) kEH3/kED3. For AChEs from each of the species in Table 1,
there is a striking correlation between solvent and substrate
isotope effects. The larger the solvent isotope effect, the more
inverse (i.e. less than unity) is the â-deuterium isotope effect.
In this article, we describe the results of solvent and substrate
isotope effects and viscosity effects on kE for hydrolysis of
ATCh catalyzed by wild-type and mutant AChEs from Torpedo
californica (TcAChE), human (HuAChE), and mouse (MAChE).
(4) Ellman, G. L.; Courtney, K. D.; Andres, J. V.; Featherstone, R. M.
Biochem. Pharmacol. 1961, 7, 88-95.
(9) (a) The crystal structure of Torpedo californica AChE was first
described by Sussman and colleagues: Sussman, J. L.; Harel, M.; Frolow,
F.; Oefner, C.; Goldman, A.; Toker, L.; Silman, I. Science 1991, 253, 872-
879. (b) The crystal structure of the complex of mouse AChE and the snake
venom toxin fasciculin is found in the following: Bourne, Y.; Taylor, P.;
Marchot, P. Cell 1995, 83, 503-512. (c) Gentry and Doctor have compared
the primary sequences of various esterases of the R/â hydrolase fold
supergene family, which includes AChE: Gentry, M. K.; Doctor, B. P. In
Enzymes of the Cholinesterase Family; Quinn, D. M., Balasubramanian,
A. S., Doctor, B. P., Taylor, P., Eds.; Plenum Press: New York and London,
1995; pp 493-505.
(5) Froede, H. C.; Wilson, I. B. J. Biol. Chem. 1984, 259, 11010-11013.
(6) (a) Shafferman, A.; Ordentlich, A.; Barak, D.; Kronman, C.; Ariel,
N.; Velan, B. In Structure and Function of Cholinesterases and Related
Proteins; Doctor, B. P., Taylor, P., Quinn, D. M., Rotundo, R. L., Gentry,
M. K., Eds.; Plenum Press: New York and London, 1999; pp 203-209.
(b) Shafferman, A.; Ordentlich, A.; Barak, D.; Stein, D.; Ariel, N.; Velan,
B. Biochem. J. 1996, 318, 833-840.
(7) Radic´, Z.; Gibney, G.; Kawamoto, S.; MacPhee-Quigley, K.; Bon-
giorno, C.; Taylor, P. Biochemistry 1992, 31, 9760-9767.
(8) The sequence positions of amino acid residues are denoted by two
numbers. The first number in normal type is the sequence position in
mammalian AChE, and the number in italics and parentheses is the sequence
position in Torpedo californica AChE. This dual numbering scheme follows
the convention proposed in the following reference: Massoulie´, J.; Sussman,
J. L.; Doctor, B. P.; Soreq, H.; Velan, B.; Cygler, M.; Rotundo, R.;
Shafferman, A.; Silman, I.; Taylor, P. In Multidiscliplinary Approaches to
Cholinesterase Functions; Shafferman, A., Velan, B., Eds.; Plenum Press:
New York and London, 1992; pp 285-288.
(10) (a) Quinn, D. M.; Sutton, L. D. In Enzyme Mechanism from Isotope
Effects; Cook, P. F., Ed.; CRC Press: Boca Raton, 1991; pp 73-126.
Additional references to solvent isotope effect reviews can be found in this
article. (b) Kiick has discussed the relationship between the curvature of
proton inventories and commitment factors that are in turn estimated from
substrate isotope effects: Kiick, D. M. J. Am. Chem. Soc. 1991, 113, 8499-
8504.
(11) Equivalent H2O and D2O buffers are those that contain the same
concentrations of all solutes.

AChE Acylation Transition State
J. Am. Chem. Soc., Vol. 122, No. 13, 2000 2983
the carbonyl carbon of the substrate, with concomitant proton
transfer from Ser203(200)8 to His447(440)8 in the active site
triad.14 This single proton bridge gives rise to a linear
dependence of k3 on n:15
k3,n ) k3H O(1 - n + nφ3 )
(3)
T
2
Substitution of eqs 2 and 3 into eq 1 gives the following equation
for proton inventories of kE:
T
(1 + C)(1 - n + nφ3 )
H2O
kE,n ) kE
(4)
1 + 1.2nC(1 - n + nφ3 )
T
H2O
In this equation kE,n and kE
are the values of kE in H2O-
D2O mixtures of atom fraction of deuterium n and 0, respec-
tively; φ3T is the fractionation factor of the bridging proton that
stabilizes the transition state of the k3 step, and C ()k3/k2) is
Figure 1. Proton inventories for kE of human AChE-catalyzed
hydrolysis of ATCh. Reactions were monitored as described in the
Enzyme Kinetics and Data Analysis subsection in the Experimental
Section. Filled-circles are for wild-type AChE and arise from following
first-order time courses in 0.1 M sodium phosphate buffer, pH 7.34 in
H2O and pD 7.86 in D2O, that contained sufficient NaCl so that µ )
0.35; reactions also contained 190 pM AChE, 2 mM TX100, 0.5 mM
DTNB, and [ATCh]0 ) 25 µM ()Km/12). Open circles are for the
Glu202(199)Ala mutant and arise from initial rate measurements in
0.01 M sodium phosphate buffer, pH 7.26 in H2O and pD 7.86 in D2O,
that contained sufficient NaCl so that µ ) 0.1; reactions also contained
2.0 nM AChE, 2 mM TX100, 0.5 mM DTNB, and 31 µM ATCh ()Km/
50). The solid lines are least-squares fits to eq 4; the corresponding
the commitment to catalysis.16 The intrinsic isotope effect on
T
the k3 step is D Ok3 ) 1/φ3 . Figure 1 shows that the Glu202Ala
2
mutant of HuAChE, which according to Table 1 gives the most
inverse â-deuterium and the largest solvent isotope effects,
yields a nearly linear proton inventory. Equation 4 predicts a
linear dependence when C ) 0, i.e. when the k3 step is solely
rate limiting.17 The average parameters of three such proton
inventories are C ) 0.4 ( 0.2 and D Ok3 ) 2.4 ( 0.4. The
2
average commitment is consistent with 70 ( 10% rate limitation
by the k3 step,17 and the magnitude of the intrinsic solvent
isotope effect is reasonable for a step whose transition state is
stabilized by O to N proton transfer from Ser203(200) to
His447(440).10a On the other hand, wild-type HuAChE gives a
proton inventory that is nonlinear and upward bulging, as also
shown in Figure 1. Least-squares analysis of this proton
T
parameters for wild-type AChE are φ3 ) 0.33 ( 0.06 and C ) 3 (
T
1, and for Glu202(199)Ala AChE the parameters are φ3 ) 0.39 (
0.06 and C ) 0.4 ( 0.3.
For ester hydrolysis inverse â-deuterium effects arise from
partial sp2 to sp3 rehybridization of the carbonyl carbon as the
reactant state is converted into the transition state.12 Conse-
quently, the correlation of increasingly inverse â-deuterium
effects with increasingly normal (i.e. greater than unity) solvent
isotope effects suggests that mutant AChEs that react more
slowly, as manifested by relative kE values, do so with greater
rate limitation by the chemical step k3.
Solvent Isotope Effects and Acylation Reaction Dynamics.
A quantitative accounting of acylation reaction dynamics can
be had by conducting proton inventories of kE, as shown in
Figure 1. In these experiments kE is measured as a function of
the atom fraction of deuterium in mixed isotopic buffers. The
respective substrate binding and release rate constants k1 and
k2 are diffusion controlled, and therefore should be slower in
D2O than in H2O by a factor of 1.2, the relative viscosities of
the two solvents.10a The following pair of equations suggests
that the dependence of these rate constants on the atom fraction
of deuterium n in mixed isotopic solvents arises from a medium
effect:10a,13
inventory gives D Ok3 ) 3.0 ( 0.6 and C ) 3 ( 1, and therefore
2
the k3 step is now only 25 ( 8% rate limiting. These results
are consistent with the less inverse â-deuterium isotope effect
and smaller solvent isotope effect for the wild-type enzyme than
for the Glu202Ala mutant. Moreover, the intrinisic solvent
isotope effect D Ok3 for the wild-type enzyme is similar to that
2
for the Glu202Ala mutant, which supports the internal con-
sistency of the model described herein for AChE acylation
reaction dynamics.
Viscosity Effects and Acylation Reaction Dynamics. The
effect of viscosity of the reaction medium on kE, plotted in
Figure 2, supports the quantitative model for acylation reaction
dynamics presented above. As the viscosity ηrel increases, the
respective rate constants k1 and k2 for substrate binding and
release decrease:18 k1η ) k1/ηrel and k2η ) k2/ηrel. Incorporation
(14) Quinn, D. M.; Feaster, S. R. In ComprehensiVe Biological Catalysis;
Sinnott, M., Ed.; Academic Press: London, 1998; Vol. 1, pp 455-482.
(15) The following references provide precedents for linear proton
inventories for the chemical steps of AChE catalysis: (a) Pryor, A. N.;
Selwood, T.; Leu, L.-S.; Andracki, M. A.; Lee, B. H.; Rao, M.; Rosenberry,
T.; Doctor, B. P.; Silman, I.; Quinn, D. M. J. Am. Chem. Soc. 1992, 114,
3896-3900. (b) Kovach, I. M.; Larson, M.; Schowen, R. L. J. Am. Chem.
Soc. 1986, 108, 3054-3056. (c) Acheson, S. A.; Dedopoulou, D.; Quinn,
D. M. J. Am. Chem. Soc. 1987, 109, 239-245.
H2O
H2O
k1n ) 1.2-nk1
k2n ) 1.2-nk2
(2)
The k3 step is that for nucleophilic attack of Ser203(200)8 on
(12) (a) Hogg, J. L. In Transition States of Biochemical Processes;
Gandour, R. D., Schowen, R. L., Eds.; Plenum Press: New York, 1978;
pp 201-224. (b) Kovach, I. M.; Belz, M.; Larson, M.; Rousy, S.; Schowen,
R. L. J. Am. Chem. Soc. 1985, 107, 7360-7365.
(16) (a) Northrop, D. B. Biochemistry 1975, 14, 2644-2651. (b)
Northrop, D. B. In Isotope Effects on Enzyme-Catalyzed Reactions; Cleland,
W. W., O’Leary, M. H., Northrop, D. B., Eds.; University Park: Baltimore,
1977; pp 122-152.
(13) Mata-Segreda has reported that the viscosity of mixed isotopic waters
is a quadratic function of n: Mata-Segreda, J. F. ReV. Latinoam. Quim.
1979, 151-153. Use of a quadratic dependence for k1 and k2 in eq 4 instead
(17) The fraction of rate limitation by the k3 step is f3 ) 1/(1 + C); cf.
refs 2d and 10a.
(18) For a discussion of viscosity effects on enzyme reactions, including
the effect of the greater viscosity of D2O than of H2O, see: Karsten, W.
E.; Lai, C.-H.; Cook, P. F. J. Am. Chem. Soc. 1995, 117, 5914-5918.
of a medium effect has no significant effect on the least-squares estimates
T
of C and φ3
.

2984 J. Am. Chem. Soc., Vol. 122, No. 13, 2000
Malany et al.
H3
D3
k1H3k3
k1D3k3
H3
D3
kE
)
kE
)
(6)
k2H3 + k3
k2D3 + k3
H3
D3
Taking the ratio of these equations provides the following:
âD
k
+
âDk1C
1 + C
3E
âDkE )
(7)
âD
H3
In this equation
k
) /
âDk1âDk3/âDk2, wherein âDk1 ) k1
3E
k1D3, for example. Therefore, âDk3E is the intrinsic â-deuterium
substrate isotope effect for conversion of the E + A reactant
state to the transition state of the k3 step. Similarly, âDk1 is the
intrinsic isotope effect for diffusional encounter of AChE and
ATCh. A reliable estimate of âDk1 is the following:19
mass of (acetyl-2H3-thio)choline
âDk1 )
) 1.009 (8)
Figure 2. Viscosity effects on wild-type mouse and human AChE-
catalyzed hydrolyses of ATCh. Values of kE were measured by first-
order kinetics in 0.05 M sodium phosphate buffer, pH 7.44, that
contained various concentrations of glycerol, 0.3 mM DTNB, and 2
mM TX100. Solid circles are for reactions that contained 37 pM human
AChE; kE values were calculated by dividing Vmax/Km values, determined
by fitting initial velocities to eq 11, by the enzyme concentration. Open
circles are for reactions that contained 100 pM mouse AChE and
[ATCh]0 ) 10 µM (<Km/10); kE values were determined by fitting
first-order time courses to eq 12. The solid lines are fits to eq 5. For
human AChE the corresponding least-squares parameters are C ) 1.2
( 0.2 and k1 ) 5.7 ( 0.3 × 107 M-1 s-1. For mouse AChE relative
viscosity increases are matched by relative decreases in kE, and therefore
eq 5 reduces to kEη ) k1/ηrel; from the least-squares fit k1 ) (2.40 (
mass of (acetyl-1H -thio)choline
x
3
By using this value for âDk1 and values of the commitment to
catalysis, C, that were determined in proton inventories, one
can solve eq 7 for âDk3E. This procedure gives âDk3E ) 0.85 (
0.06 for the wild-type human AChE-catalyzed hydrolysis of
ATCh, as tabulated in Table 1. One can also calculate the
intrinsic â-deuterium substrate isotope effect from the corre-
sponding isotope effect that was measured in D2O, i.e. âDkE )
0.95 ( 0.01. The commitment in D2O is calculated from the
commitment in H2O and the intrinsic solvent isotope effects on
the k2 and k3 steps: CD O
)
D Ok2CH O D O
/ k3 ) 1.2(3 ( 1)/(3.0
2
2
2
2
âD
0.07) × 107 M-1 s-1
.
( 0.6) ) 1.2 ( 0.5. Calculation according to eq 7 gives
k
3E
) 0.88 ( 0.04, in agreement with the value calculated for the
reaction in H2O. Therefore, the smaller observed â-D effect in
H2O than in D2O arises from more prominent rate limitation
by the k3 step in the latter than in the former solvent, and
indicates that the k3 step is more sensitive to solvent and
substrate isotopic substitution than the k1 and k2 steps.
Intrinsic â-deuterium substrate isotope effects were also
calculated for reactions of the Glu202Gln and Glu202Ala
mutants of human AChE. The weighted average of the four
intrinsic substrate isotope effects in Table 1 is âDk3E ) 0.90 (
0.03. This average effect can be used to estimate the fractional
change (f) in hybridization (sp2 f sp3) of ATCh and the bond
order between the γO of Ser203(200) and the carbonyl carbon
of ATCh in the transition state:12b
of these dependences into eq 1 gives:
k1C
kEη
)
(5)
1 + ηrelC
Analysis of the viscosity dependence of Figure 2 according to
eq 5 gives C ) 1.2 ( 0.2, which is consistent with 45 ( 8%
rate limitation by the k3 step. This result is in reasonable
quantitative agreement with that derived from the proton
inventory of Figure 1. Similar experiments for wild-type
MAChE indicate that kE is solely rate limited by the k1 step,
which is consistent with the diminutive solvent and â-deuterium
isotope effects in Table 1. In fact, the solvent isotope effect on
kE for wild-type MAChE is just that expected to arise from the
relative viscosities of H2O and D2O,10a as expected for a
diffusion-controlled reaction.
Acylation Transition State Structure from â-Deuterium
Isotope Effects. The effect of solvent isotope on the â-deuterium
effect also supports the quantitative assessment of acylation
reaction dynamics that derives from proton inventory experi-
ments. Consider the âDkE values of 0.97 ( 0.01 and 0.95 (
0.01 for wild-type HuAChE reactions in H2O and D2O,
respectively (cf. Table 1). The solvent isotope effect on the
â-deuterium substrate isotope effect is consistent with the idea
that the chemical k3 step is subject to more sizable isotope effects
than is the k1 step, and that the k3 step is more rate-limiting in
D2O than in H2O. The analysis that follows shows how this
information can be used to determine the intrinsic â-deuterium
isotope effect, from which a model of the transition state of the
k3 step can be constructed.
âD
k
) (âDKeq)f
(9)
K is that for equilibrium addition to the
eq
3E
âD
The isotope effect
carbonyl carbon, has a value of 0.87 for three deuteriums,12b
and provides a measure of the effect expected for full sp2 f
sp3 rehybridization. The value calculated from this equation for
the fractional change in hybridization is f ) 0.76 ( 0.22. The
salient feature of this estimate is that the chemical transition
state for the acylation stage of AChE-catalyzed hydrolysis of
ATCh bears considerable structural resemblance to the putative
tetrahedral intermediate. To provide a more quantitative view
of the transition state, a gas-phase model of the tetrahedral
intermediate for methoxide addition to carbonyl-protonated
(acetylthio)choline was generated by ab initio quantum me-
(19) The modest normal â-deuterium isotope effect calculated according
to eq 8 arises from the mass dependence of the frequency of collisions of
the respective isotopic (acetylthio)cholines with the enzyme: Leffler, J.
E.; Grunwald, E. Rates and Equilibria in Organic Reactions; John Wiley:
New York and London, 1963; pp 57-59.
The relationship between the observed â-D effect on kE and
the intrinsic â-D effect on the k3 step is derived by writing
versions of eq 1 for each of the isotopic substrates:

AChE Acylation Transition State
J. Am. Chem. Soc., Vol. 122, No. 13, 2000 2985
agreement with theoretical models in the literature for the
acylation transition state in AChE catalysis. Fuxreiter and
Warshel22 used the effective valence bond/free energy perturba-
tion method to simulate the free energetics of formation of the
tetrahedral intermediate in the acylation stage of AChE-catalyzed
hydrolysis of acetylcholine and in the corresponding reaction
in water. Their calculations suggest that on the enzyme the
transition state lies 16 and 3.6 kcal mol-1 above the ester reactant
and the tetrahedral intermediate, respectively. From these
energetics one can estimate the fractional progress in the
transition state as f ) 0.68 by utilizing intersecting parabolas.23
The agreement with our experimentally determined value, f )
0.76 ( 0.22, is gratifying, in that both measures indicate that
the transition state more closely resembles the tetrahedral
intermediate than the reactant ester.
Figure 3. Crossed stereoview of a model for the transition state for
The quantum mechanical calculations of Wlodek et al.24 for
the formation of the tetrahedral intermediate in the acylation
stage of AChE catalysis suggest a smaller fractional progress
in the transition state than supported by the experiments
described herein. However, their gas-phase model for acylation
energetics understandably predicts that the tetrahedral intermedi-
ate is more stable than the reactant state. The efficacy of their
calculations resides in the quantitative accounting they provide
for the effect of the Glu202(199) to Gln mutation on the activity
of AChE.24
Conclusion. Much has been made in recent years of the fact
that amino acid side chains and peptide NH functions of
enzymes have a wide range, φ ) 0.3 to 1.8, of proton
fractionation factors.25 This observation has been taken as
impugning the ability of solvent isotope effects to produce useful
information on transition state structure and reaction dynamics
in enzyme-catalyzed reactions. The work detailed herein largely
allays this concern for AChE catalysis. Three independent
measures, viz. solvent and substrate isotope effects and viscosity
effects, give reasonably the same picture of fractional rate
limitation in the acylation stage of catalysis. The agreement of
these three methods inspires confidence not only in the model
for acylation reaction dynamics described herein but also in the
integrity of solvent isotope effect measurements to provide
quantitative information thereon.
methoxide addition to carbonyl-protonated (acetylthio)choline.
Table 2. Selected Structural Features of Protonated Ester,
Tetrahedral Intermediate, and Transition State Models
protonated
ester
tetrahedral
intermediate
transition
state
parametera
C1-O1
1.261
1.386
1.371
1.524
1.851
106.94
114.17
112.04
107.42
110.17
105.91
1.379
C1-O2, Å
1.45b
C1-C2, Å
C1-S, Å
1.491
1.707
117.55
1.519
1.839
C2-C1-S, deg
C2-C1-O2, deg
C2-C1-O1, deg
O1-C1-O2, deg
O1-C1-S, deg
O2-C1-S, deg
107.94
112.96
112.94
106.63
111.10
105.03
116.30
126.15
a Parameters with two-atom designations (e.g. C1-O1) are bond
distances, and those with three-atom designations (e.g. C2-C1-S) are
bond angles. See Figure 3 for specification of atom labels. b Distance
constrained as calculated according to eq 10 from the corresponding
distance in the tetrahedral intermediate. Dihedral angle constraints are
outlined in the Computational Modeling of the Transition State
subsection in the Experimental Section.
chanical calculations.20 This model possesses a nucleophilic
oxygen to carbonyl carbon bond distance of 1.373 Å. Using
this value as an estimate of the single bond distance, one can
calculate the corresponding distance in the transition state from
the following Pauling bond-order/bond-distance relationship:21
Experimental Section
D(f) ) D(1) - 0.6 log(f)
(10)
Materials. Reagents for the synthesis of isotopic (acetylthio)cholines
were purchased from the following sources: 2-aminoethanethiol
hydrochloride, Na2S‚9H2O, (2H3CCO)2O, dimethyl sulfate, and iodo-
methane, Aldrich Chemical Co.; (1H3CCO)2O, EM Science. (Acetyl-
thio)choline chloride, DTNB, NaH2PO4‚H2O, Na2HPO4‚7H2O, and
This calculation gives D(f) ) 1.45 ( 0.05 Å. The transition
state model that is displayed in Figure 3 was then generated by
constraining the nucleophilic methoxy oxygen-carbonyl carbon
distance at 1.45 Å. Table 2 compares various bond lengths and
angles for the tetrahedral intermediate and transition state
models, as well as for the starting carbonyl-protonated ester.
These data confirm the general impression that the transition
state bears appreciable structural similarity to the tetrahedral
intermediate, particularly with respect to bond angles in which
the carbonyl carbon is the central atom.
(22) Fuxreiter, M.; Warshel, A. J. Am. Chem. Soc. 1998, 120, 183-
194.
(23) Estimation of the fractional progress in the transition state from the
reaction energetics is achieved by calculating the intersection point of two
parabolas, each that have positive curvature. The minimum of one parabola
is set at f ) 0, where f is the fractional reaction progress, and y ) 0, where
y is the free energy in kcal mol-1, i.e. y ) af2. The minimum of the second
parabola is set at f ) 1 and y ) 12.4, which corresponds to the structure
and energetics of the tetrahedral intermediate; this gives y ) a(f - 1)2 +
12.4. Since the free energy of the transition state is 16 kcal mol-1 above
that of the reactant ester, y is set at 16 and the simultaneous equations are
solved for a ) 34.8 kcal mol-1 and f ) 0.68.
Comparison with Other Theoretical Models. The experi-
mental transition state structure described herein is in good
(20) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Kieth, T. A.; Petersson,
G. A.; Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zar-
krzewski, V. G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B.
B.; Nanayakkara, A.; Challocombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen,
W.; Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin,
R. L.; Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.;
Head-Gordon, M.; Gonzalez C.; Pople, J. A. 1995, GAUSSIAN 94 (Revision
D.4); Gaussian, Inc.: Pittsburgh, PA.
(24) Wlodek, S. T.; Antosiewicz, J.; Briggs, J. M. J. Am. Chem. Soc.
1997, 119, 8159-8165.
(25) Loh, S. N.; Markley, J. L. Biochemistry 1994, 33, 1029-1036. A
notable feature of the fractionation factors reported by Loh and Markley is
that their average value is ∼0.8, which will generate a diminutive solvent
isotope effect of ∼1.2. Moreover, unless there is a palpable shift from this
value in the transition state of an enzymic reaction (a supposition for which
there is no evidence), the wide variation of fractionation factors reported
by Loh and Markley is of no relevance to the interpretation of solvent
deuterium kinetic isotope effects.
(21) Pauling, L. The Nature of the Chemical Bond, 3rd ed.; Cornel
University Press: Ithaca, NY, 1960; pp 255-256.

2986 J. Am. Chem. Soc., Vol. 122, No. 13, 2000
Malany et al.
T
Triton X-100 were purchased from Sigma Chemical Co. H2O used in
buffer preparation was distilled and subsequently deionized by passage
through a Barnstead D8922 mixed-bed ion-exchange column (Sybron
Corp). D2O was used as purchased from Isotec Inc. Buffer salts were
commercially available reagent grade materials. A Corning model 125
pH meter, equipped with a glass combination electrode, was used to
measure buffer pH values. For D2O buffers, pD values were determined
by adding 0.4 to the pH meter reading.10a
Proton inventory parameters (i.e. C and φ3 ) and â-deuterium
secondary isotope effects reported herein are weighted means of means
from repetitive measurements that were conducted on various days.
Weighted means were calculated according to the following equation:
n
n
Xh )
WiXi/ Wi
(13)
∑ ∑
i)1
i)1
Wild-type and mutant recombinant AChEs from Torpedo califor-
nica,7 mouse,26 and homo sapiens27 were produced and characterized
as described previously. Stock enzyme solutions were diluted in reaction
buffers prior to use.
The weighting factors are Wi ) 1/σi2, where σi2 is the square of the
standard error of the mean for â-deuterium isotope effects or of the
standard errors of the least-squares estimates for proton inventory
parameters. The standard error of the mean of means is also calculated
as a weighted mean according to eq 13 from the standard errors of the
repetitive measurements.
Synthesis of Isotopic (Acetylthio)cholines. Acetyl-2H3-thiocholine
was synthesized as follows. 2-(N,N-Dimethylamino)ethanethiol was
produced by successive additions of 1 equiv of dimethyl sulfate to
2-aminoethanethiol hydrochloride, as described by Nair et al. for the
dimethylation of 3-bromoaniline.28 2-(N,N-Dimethylamino)ethanethiol
hydrochloride (7.03 g, 50 mmol) was dissolved in aqueous sodium
sulfide (40 mL, 0.6 M) and extracted with 3 × 25 mL of CHCl3. 2H3-
acetic anhydride (2.69 g, 24.9 mmol) was added to the combined CHCl3
extracts in a round-bottom flask, and the reaction mixture was stirred
at room temperature for 4 h. The reaction mixture was washed with 3
× 10 mL of Na2CO3 (1.0 M); the aqueous extracts were combined,
extracted with 10 mL of CHCl3, dried over MgSO4, filtered, and
concentrated by rotary evaporation. Acetyl-2H3-thio-2-(N,N-dimeth-
ylamino)ethane was purified by fractional distillation (90-95 °C, ∼20
mmHg) in a short-path distillation apparatus. The final methylation
was effected by slow addition with stirring of iodomethane (10.8 g, 70
mmol) to 1.97 g (13 mmol) of acetyl-2H3-thio-2-(N,N-dimethylamino)-
ethane in dry acetone in a round-bottom flask. The highly exothermic
reaction formed a precipitate almost immediately. The precipitate was
collected by filtration, washed with dry Et2O (40 mL), and dried in
vacuo at room temperature to give acetyl-2H3-thiocholine as a white
powder. High-resolution mass spectra, acquired on an AutoSpec
spectrometer operated in the electron spray ionization mode, gave a
molecular mass for 2H3CCOSCH2CH2N+(CH3)3 of 165.21 u (calculated
165.29 u), and indicated that the percent deuteration of the acetyl group
was >97%. Proton NMR spectra, acquired at 300 MHz on an MSL-
300 spectrometer, were consistent with the expected structure of the
compound.
Viscosity Experiments. The viscosities of aqueous buffers in which
AChE reactions were conducted were adjusted by adding glycerol, as
described by Bazelyansky et al.2a The viscosigen glycerol not only slows
k1 and k2, but also has an inhibitory solvent effect on AChE catalysis.
Consequently, as did Bazelyansky et al.,2a the rate constants plotted in
Figure 2 were adjusted by the degree to which hydrolysis of a slow
substrate, in the experiments detailed herein (butyrylthio)choline, was
inhibited by glycerol. This was done by dividing observed kE values
for ATCh hydrolysis by the fractional inhibition of BuTCh hydrolysis
at the various glycerol concentrations.
Computational Modeling of the Transition State. A model of the
transition state was constructed by the following procedure: (1) a model
of the tetrahedral intermediate for CH3O- addition to the carbonyl-
protonated form of ATCh was constructed by using the Spartan29
program; (2) the dihedral angles for the C-S-CH2-CH2 and S-CH2-
CH2-N substructures of the model were constrained at 180°, in accord
with the preference of AChE for binding of the fully extended
conformation of the substrate;30 (3) the Spartan model was then
geometry optimized by using the Gaussian 94 program and the
6-31+G** basis set,20 albeit with these dihedral angle constraints; (4)
the bond length in the transition state of the methoxy oxygen-carbonyl
carbon bond was calculated according to eq 10; and (5) a model of the
transition state was generated in a subsequent Gaussian 94 geometry
optimization, albeit with the methoxy oxygen-carbonyl carbon bond
length and dihedral angle constraints outlined above. This general
procedure was also followed for the carbonyl-unprotonated form of
ATCh. In this case, however, the tetrahedral adduct was less stable
than the products, methyl acetate and thiocholine zwitterion. Because
the intrinsic â-deuterium secondary isotope effects suggest that the
Enzyme Kinetics and Data Analysis. Reaction time courses for
AChE-catalyzed hydrolysis of ATCh were monitored at 25.0 ( 0.2 °C
by the Ellman assay,4 which utilizes coupling of the thiocholine product
with DTNB to generate a product that absorbs at λ ) 412 nm (ꢀ )
13600 M-1 cm-1). Initial velocities were calculated by linear least-
squares analysis of time courses for <10% turnover of the initial
substrate concentration. The kinetic parameters Km and Vmax were
determined by fitting initial velocities to the Michaelis-Menten
equation:
(29) Deppmeier, B. J.; Driessen, A. J.; Hehre, W. J.; Johnson, J. A.;
Klunzinger, P. E.; Lou, L.; Yu, J.; Baker, J.; Carpenter, J. E.; Dixon, R.
W.; Fielder, S. S.; Johnson, H. C.; Kahn, S. D.; Leonard, J. M.; Pietro, W.
J.;, SPARTAN SGI Version 5.0.2; Wavefunction Inc.: Irvine, CA, 1997.
(30) Harel, M.; Quinn, D. M.; Nair, H. K.; Silman, I.; Sussman, J. L. J.
Am. Chem. Soc. 1996, 118, 2340-2346.
(31) Ordentlich, A.; Barak, D.; Kronman, C.; Ariel, N.; Segall, Y.; Velan,
B.; Shafferman, A. J. Biol. Chem. 1998, 273, 19509-19517.
V
max[A]
νi )
(11)
Km + [A]
(32) A preliminary account of the work described in this paper was
reported in the following reference: Malany, S.; Sikorski, R. S.; Seravalli,
J.; Medhekar, R.; Quinn, D. M.; Radic´, Z.; Taylor, P.; Velan, B.; Kronman,
C.; Shafferman, A. In Structure and Function of Cholinesterases and Related
Proteins; Doctor, B. P., Taylor, P., Quinn, D. M., Rotundo, R. L., Gentry,
M. K., Eds.; Plenum Press: New York and London, 1998; pp 197-202.
(33) Note Added in Proof: Since the submission of this paper, work
has begun on computational models of the tetrahedral intermediate and
transition state in which electrophilic interaction at the carbonyl oxygen is
provided by three water molecules. The tetrahedral intermediate is more
exploded than the corresponding intermediate described in Table 2. For
example, the C1-O2 and C1-S distances are 1.413 and 1.971 Å,
respectively. Moreover, bond angles about the carbonyl carbon that include
sulfur are closely similar to (e.g. O1-C1-S ) 110.51°) or are smaller than
(e.g. O2-C1-S ) 101.34°) those of the tetrahedral intermediate model of
Table 2. A transition state model was generated by constraining the C1-
O2 distance at 1.48 Å, calculated according to eq 10 with f ) 0.76 and
D(1) ) 1.413 Å; this procedure is the same as that used to generate the
transition state model described in Table 2 and Figure 3. As before, this
new transition state model bears appreciable structural similarity to its
cognate tetrahedral intermediate, particularly with respect to bond angles
in which the carbonyl carbon is the central atom.
The first-order rate constant k ) Vmax/Km ) kcat[E]T/Km was also
determined at [A]0 < Km/10 by fitting reaction time courses, followed
for >3 half-lives, to the following equation, in which A, A0, and A∞
are absorbances at times t, 0, and infinity, respectively:
A ) (A0 - A∞)e-kt + A∞
(12)
(26) (a) Vellom, D. C.; Radic´, Z.; Li, Y.; Pickering, N. A.; Camp, S.;
Taylor, P. Biochemistry 1993, 32, 12-17. (b) Radic´, Z.; Pickering, N. A.;
Vellom, D. C.; Camp, S.; Taylor, P. Biochemistry 1993, 32, 12074-12084.
(27) (a) Shafferman, A.; Velan, B.; Ordentlich, A.; Kronman, C.;
Grosfeld, H.; Leitner, M.; Flashner, Y.; Cohen, S.; Barak, D.; Ariel, N.
EMBO J. 1992, 11, 3561-3568. (b) Ordentlich, A.; Kronman, C.; Barak,
D.; Stein, D.; Ariel, N.; Marcus, D.; Velan, B.; Shafferman, A. FEBS Lett.
1993, 334, 215-220.
(28) Nair, H. K.; Lee, K.; Quinn, D. M. J. Am. Chem. Soc. 1993, 115,
9939-9941.

AChE Acylation Transition State
J. Am. Chem. Soc., Vol. 122, No. 13, 2000 2987
transition state closely resembles the tetrahedral intermediate, these
computational comparisons indicate that the tripartite oxyanion hole
of AChE30,31 provides the stabilization needed for the tetrahedral
intermediate to form.32,33
Acknowledgment. We acknowledge the technical assistance
of Rohit Medhekar. This work was supported by NIH grant
NS21334 to D.M.Q., NIH grant GM18360 to P.T., and by a
contract to A.S. from the US Army Research and Development
Command (DAMD17-96-C-6088). S.M. acknowledges the
support of a predoctoral fellowship from the University of Iowa
Center for Biocatalysis and Bioprocessing.
(34) Quinn, D. M.; Seravalli, J.; Nair, H. K.; Medhekar, R. A.; Husseini,
B.; Radic´, Z.; Vellom, D. C.; Pickering, N.; Taylor, P. In Enzymes of the
Cholinesterase Family; Quinn, D. M., Balasubramanian, A. S., Doctor, B.
P., Taylor, P., Eds.; Plenum Press: New York and London, 1995; pp 203-
208.
JA9933590