Isotope effects on the enzymatic and nonenzymatic reactions of chorismate

Article abstract of DOI:10.1021/ja052929v

The important biosynthetic intermediate chorismate reacts thermally by two competitive pathways, one leading to 4-hydroxybenzoate via elimination of the enolpyruvyl side chain, and the other to prephenate by a facile Claisen rearrangement. Measurements with isotopically labeled chorismate derivatives indicate that both are concerted sigmatropic processes, controlled by the orientation of the enolpyruvyl group. In the elimination reaction of [4- ²H]chorismate, roughly 60% of the label was found in pyruvate after 3 h at 60 °C. Moreover, a 1.846 ± 0.057 ²H isotope effect for the transferred hydrogen atom and a 1.0374 ± 0.0005 ¹⁸O isotope effect for the ether oxygen show that the transition state for this process is highly asymmetric, with hydrogen atom transfer from C4 to C9 significantly less advanced than C-O bond cleavage. In the competing Claisen rearrangement, a very large ¹⁸O isotope effect at the bond-breaking position (1.0482 ± 0.0005) and a smaller ¹³C isotope effect at the bond-making position (1.0118 ± 0.0004) were determined. Isotope effects of similar magnitude characterized the transformations catalyzed by evolutionary unrelated chorismate mutases from Escherichia coli and Bacillus subtilis. The enzymatic reactions, like their solution counterpart, are thus concerted [3,3]-sigmatropic processes in which C-C bond formation lags behind C-O bond cleavage. However, as substantially larger ¹⁸O and smaller ¹³C isotope effects were observed for a mutant enzyme in which chemistry is fully rate determining, the ionic active site may favor a somewhat more polarized transition state than that seen in solution.

Full text of DOI:10.1021/ja052929v

Published on Web 08/27/2005
Isotope Effects on the Enzymatic and Nonenzymatic
Reactions of Chorismate
S. Kirk Wright,† Michael S. DeClue,‡ Ajay Mandal,‡ Lac Lee,† Olaf Wiest,§
W. Wallace Cleland,*,† and Donald Hilvert*,‡
Contribution from the Institute for Enzyme Research and Department of Biochemistry,
UniVersity of Wisconsin, 1710 UniVersity AVenue, Madison, Wisconsin 53726, Laboratory of
Organic Chemistry, Swiss Federal Institute of Technology, ETH Ho¨nggerberg, CH-8093 Zu¨rich,
Switzerland, and Department of Chemistry and Biochemistry, UniVersity of Notre Dame,
Notre Dame, Indiana 46556-5670
Received May 4, 2005; E-mail: hilvert@org.chem.ethz.ch
Abstract: The important biosynthetic intermediate chorismate reacts thermally by two competitive pathways,
one leading to 4-hydroxybenzoate via elimination of the enolpyruvyl side chain, and the other to prephenate
by a facile Claisen rearrangement. Measurements with isotopically labeled chorismate derivatives indicate
that both are concerted sigmatropic processes, controlled by the orientation of the enolpyruvyl group. In
the elimination reaction of [4-2H]chorismate, roughly 60% of the label was found in pyruvate after 3 h at 60
°C. Moreover, a 1.846 ( 0.057 2H isotope effect for the transferred hydrogen atom and a 1.0374 ( 0.0005
18O isotope effect for the ether oxygen show that the transition state for this process is highly asymmetric,
with hydrogen atom transfer from C4 to C9 significantly less advanced than C-O bond cleavage. In the
competing Claisen rearrangement, a very large 18O isotope effect at the bond-breaking position (1.0482 (
0.0005) and a smaller 13C isotope effect at the bond-making position (1.0118 ( 0.0004) were determined.
Isotope effects of similar magnitude characterized the transformations catalyzed by evolutionarily unrelated
chorismate mutases from Escherichia coli and Bacillus subtilis. The enzymatic reactions, like their solution
counterpart, are thus concerted [3,3]-sigmatropic processes in which C-C bond formation lags behind
C-O bond cleavage. However, as substantially larger 18O and smaller 13C isotope effects were observed
for a mutant enzyme in which chemistry is fully rate determining, the ionic active site may favor a somewhat
more polarized transition state than that seen in solution.
Scheme 1
The unimolecular rearrangement of chorismate (1) to prephen-
ate (2) is a key step in the biosynthesis of the aromatic amino
acids tyrosine and phenylalanine (Scheme 1). Formally, this
reaction can be described as a pericyclic Claisen rearrangement,
a 3,3-sigmatropic process that is rarely exploited in cellular
metabolism. Secondary tritium isotope effects have suggested
that the uncatalyzed reaction is concerted but asynchronous, with
C-O bond cleavage preceding C-C bond formation,1 while
helix bundle chorismate mutase domain of the Escherichia coli
stereochemical studies have established a chairlike geometry
chorismate mutase-prephenate dehydratase (EcCM).6 The AroH
for the transition state.2 These conclusions are supported by
class, exemplified by the monofunctional chorismate mutase
from Bacillus subtilis (BsCM), has a trimeric pseudo R/â-barrel
RHF/6-31* calculations.3
Chorismate mutases accelerate the chorismate to prephenate
rearrangement more than a million fold.4 Crystallographic
studies have shown that these enzymes fall into two structurally
distinct classes.5 The overwhelming majority of chorismate
mutases belong to the AroQ family, whose prototype is the all-
topology with three active sites at the subunit interfaces.7,8
Despite their different scaffolds, the enzymes have similarly
functionalized active sites, activities, and inhibition profiles.5
The origins of the catalytic rate enhancement achieved by
chorismate mutases have been elusive. The enzymatic reactions,
like their uncatalyzed counterpart, proceed via a chairlike
transition state.2 However, efforts to better define the structure
† University of Wisconsin.
‡ Swiss Federal Institute of Technology.
§ University of Notre Dame.
(1) Addadi, L.; Jaffe, E. K.; Knowles, J. R. Biochemistry 1983, 22, 4494-
(5) Lee, A. Y.; Stewart, J. D.; Clardy, J.; Ganem, B. Chem. Biol. 1995, 2,
195-203.
4501.
(2) Sogo, S. G.; Widlanski, T. S.; Hoare, J. H.; Grimshaw, C. E.; Berchtold,
G. A.; Knowles, J. R. J. Am. Chem. Soc. 1984, 106, 2701-2703.
(3) (a) Wiest, O.; Houk, K. N. J. Am. Chem. Soc. 1995, 117, 11628-11639.
(b) Wiest, O.; Houk, K. N. J. Org. Chem. 1994, 59, 7582-7584.
(4) Andrews, P. R.; Smith, G. D.; Young, I. G. Biochemistry 1973, 12, 3492-
3498.
(6) Lee, A. Y.; Karplus, P. A.; Ganem, B.; Clardy, J. J. Am. Chem. Soc. 1995,
117, 3627-3628.
(7) Chook, Y. M.; Gray, J. V.; Ke, H. M.; Lipscomb, W. N. J. Mol. Biol.
1994, 240, 476-500.
(8) Chook, Y. M.; Ke, H. M.; Lipscomb, W. N. Proc. Natl. Acad. Sci. U.S.A.
1993, 90, 8600-8603.
9
10.1021/ja052929v CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 12957-12964
12957

A R T I C L E S
Wright et al.
sponding to the lysate of ca. 40 mg/mL) was added, and the pH of the
resulting mixture was adjusted to 8.1 using NaOH (1 N). Finally, sodium
dithionite (85 mg, 0.48 mmol) was added. The resulting solution was
sealed and gently stirred at 23 °C for 1 h. The reaction mixture was
then filtered through an Amicon 10K cutoff ultrafiltration membrane
at 4 °C. When nearly dry, the membrane was washed with Tris-HCl
(25 mL, 50 mM, pH 8.2). The combined filtrate was acidified to pH
1.5 and extracted with EtOAc (6 × 20 mL). The combined extracts
were dried over Na2SO4 and carefully concentrated in vacuo (while
the bath for the rotary evaporator was maintained at ca. 20 °C). The
residue was dissolved in a 1:1 mixture of H2O and MeOH and filtered
through a pad of charcoal. The filtrate was lyophilized to provide crude
labeled chorismic acid (80.0 mg, 40%). The residue was further purified
by C-18 reverse phase silica gel chromatography (10 mM, ammonium
acetate buffer). Rf (1 M, ammonium acetate buffer) 0.89. 1H NMR (300
MHz, D2O): δ 4.64 (d, J ) 2.0 Hz, 1H), 4.84 (m, 1H), 4.97 (dd, J )
10.0, 2.0 Hz, 1H), 5.26 (d, J ) 2.0 Hz, 1H), 6.02 (dd, J ) 10, 2.0 Hz,
1H), 6.38 (dt, J ) 8.0, 2.0 Hz, 1H), 6.65 (m, 1H). MS (ESI), m/z 451
(2M - H+), 225 (M - H+), 207 (M - H3O+).
of this species at the active site of an AroQ enzyme, the
bifunctional chorismate mutase-prephenate dehydrogenase from
E. coli, were thwarted by suppression of the secondary tritium
isotope effect, most likely due to a large forward commitment.1
More recently, heavy atom isotope effects were determined for
the BsCM-catalyzed reaction.9 The large 18O isotope effect at
O7 (ca. 5%), the bond-breaking position, and the small, normal
isotope effect at C1 (ca. 0.6%), the site of bond-making,
suggested that chemistry is significantly rate determining for
the AroH enzyme and that the enzymatic reaction also proceeds
through a concerted but asymmetric transition state.
In the present study, we have extended our work on BsCM
by determining the heavy atom isotope effect at C9 in the vinyl
side chain of chorismate, providing additional information about
the bond-making process. We also report heavy atom isotope
effects at C1 and O7 for the EcCM-catalyzed reaction and for
the uncatalyzed rearrangement of chorismate. Our results
indicate that both enzyme-catalyzed reactions, as well as the
spontaneous thermal reaction, proceed concertedly through
similar asynchronous pericyclic transition states. Furthermore,
analysis of the nonenzymatic conversion of chorismate to
4-hydroxybenzoate shows that this side reaction, which com-
petes with the Claisen rearrangement, also proceeds via a
pericyclic transition state.
[9-13C,10-13C]Chorismic Acid (1d). This compound was prepared
chemoenzymatically from ethyl [7-13C]shikimate10 (25 mg, 0.12 mmol)
and [3-13C]phosphoenolpyruvate11 (58.5 mg, 0.26 mmol), as described
1
for [10-12C]chorismic acid 1a (8.2 mg, 30%). H NMR (300 MHz,
D2O): δ 4.53 (m, 1H), 5.05 (dd, J ) 12.6, 1.6 Hz, 1H), 5.17 (m, 1H),
5.71 (m, 1H), 6.05 (m, 1H), 6.35 (m, 1H), 6.86 (m, 1H). MS (ESI),
m/z 227 (M - H+), 209 (M - H3O+). [10-12C]Chorismic acid and
[7-18O,10-13C]chorismic acid were mixed in ca. 99:1 ratio. To ensure
decomposition of all contaminating prephenate prior to determination
of the isotope effect, the mixture was treated with 300 µL of 5 N HCl.
After 1 h on ice, the chorismate acid solution was brought to pH to 6.5
by addition of KOH and sparged with moist CO2-free nitrogen gas.
3-((3aS,5R,6R,6aS)-6-Benzyloxy-2,2-dimethyl-tetrahydro-furo-
[2,3-d][1,3]dioxol-5-yl)-2-(diethoxy-phosphoryl)-propionic Acid Meth-
yl Ester (4). Triflic anhydride (1.14 mL, 6.8 mmol) was added dropwise
over 10 min to a stirred solution of 1,2-O-isopropylidene-3-O-benzyl-
D-arabinofuranose (3)16 (1.45 g, 5.2 mmol) in dry CH2Cl2 (15 mL) and
pyridine (920 µL, 11.4 mmol) at -30 °C under a flow of nitrogen.
After 15 min at -30 °C, the reaction was quenched by the addition of
H2O (5 mL). The resulting mixture was diluted with CH2Cl2 (20 mL),
washed with aqueous NaH2PO4 (1.0 M, 5 mL), dried over MgSO4,
and evaporated to afford 2.0 g (93%) of the triflate as a light pink oil.
This intermediate is highly reactive and should be used as quickly as
possible. The crude triflate (2.0 g, 4.8 mmol) was dissolved in dry
DMF (8.5 mL) and added to the ylid of methyl P,P-diethylphospho-
nacetate, prepared in advance by adding methyl P,P-diethylphospho-
nacetate (1.54 mL, 8.35 mmol) in dry DMF (8.5 mL) to a stirred
suspension of sodium hydride (300 mg, 7.5 mmol) in dry DMF (20
mL) at 0 °C. After addition of 15-crown-5 (3 drops), the mixture was
stirred overnight at room temperature to furnish a dark solution. The
reaction was subsequently cooled to 0 °C, quenched with aqueous NaH2-
PO4 (1.0 M, 25 mL), and then extracted with CHCl3 (3 × 40 mL). The
combined CHCl3 layers were washed with H2O (15 mL), dried over
MgSO4, and evaporated to yield a red oil. Purification by flash
chromatography with silica absorbent (EtOAc/hexanes, 1:1) gave 1.65
g (73%) of 4 as a mixture of diastereomers. 1H NMR (300 MHz,
CDCl3): δ 1.28-1.37 (m, 18H), 1.51 (s, 3H), 1.55 (s, 3H), 2.15-2.30
(m, 2H), 2.35-2.55 (m, 2H), 3.04-3.20 (m, 1H), 3.34-3.50 (m, 1H),
3.68-3.86 (m, 8H), 4.06-4.22 (m, 10 H), 4.54-4.66 (m, 6H), 5.84
(d, J ) 3.9 Hz, 1H), 5.89 (d, J ) 3.6 Hz, 1H), 7.28-7.40 (m, 10H).
HRMS-MALDI+ [M + Na]+: calcd (C22H33NaO9P) 495.1760; found
495.1755.
Materials and Methods
General Methods and Materials. All reactions were carried out in
flame-dried glassware. THF and Et2O were dried by distillation from
sodium benzophenone ketyl. CH2Cl2 was dried by distillation from
CaH2. All other chemicals and solvents were reagent grade and used
as received. Chromatographic purifications were performed with
Kieselgel 60, 230-400 mesh silica gel (Fluka). Eluent mixtures are
reported as v:v percentages of the minor constituent in the major
constituent. Methyl [7-12C]shikimate,10 ethyl [7-13C]shikimate,10 and
potassium [3-13C]phosphoenolpyruvate11 were prepared by published
methods. Unlabeled chorismic acid was isolated from Escherichia coli
strain KA12 and purified by reverse phase chromatography.12 Mixtures
of [10-12C]chorismic acid and [7-18O,10-13C]chorismic acid (ca. 99:1)
and [10-12C]chorismic acid and [1,10-13C]chorismic acid (ca. 99:1) were
prepared chemoenzymatically as previously described.10 Wild-type
BsCM,13 C75S BsCM,13 and EcCM14 were produced and purified as
reported.
[10-12C]Chorismic Acid (1a). NaOH (1 M, 0.88 mL, 0.88 mmol)
was added to a solution of methyl [7-12C]shikimate10 (165 mg, 0.88
mmol) in a 1:1 mixture of THF and H2O (12 mL). The mixture was
stirred at 23 °C for 8 h. After the removal of THF by distillation in
vacuo, the aqueous mixture was lyophilized. The monosodium salt of
ATP (0.68 g, 1.2 mmol) was dissolved in Tris-HCl (24.0 mL, 50 mM)
while maintaining the pH of the solution between 6.5 and 9.0. To this
solution were added FMN (50 mg, 0.09 mmol), the unlabeled
monopotassium salt of phosphoenolpyruvate (0.57 g, 2.43 mmol), a
solution of KCl (1.0 M aqueous, 2.0 mL), a solution of MgSO4 (1.0 M
aqueous, 0.2 mL), and a solution of the sodium shikimate in Tris-HCl
(4.4 mL, 50 mM, pH 8.2). KA12/pKAD50 extract15 (30 mL, corre-
(9) Gustin, D. J.; Mattei, P.; Kast, P.; Wiest, O.; Lee, L.; Cleland, W. W.;
Hilvert, D. J. Am. Chem. Soc. 1999, 121, 1756-1757.
(10) Gustin, D. J.; Hilvert, D. J. Org. Chem. 1999, 64, 4935-4938.
(11) Hirschbein, B. L.; Mazenod, F. P.; Whitesides, G. M. J. Org. Chem. 1982,
47, 3765-3766.
(12) Grisostomi, C.; Kast, P.; Pulido, R.; Huynh, J.; Hilvert, D. Bioorg. Chem.
1997, 25, 297-305.
(13) Mattei, P.; Kast, P.; Hilvert, D. Eur. J. Biochem. 1999, 261, 25-32.
(14) MacBeath, G.; Kast, P.; Hilvert, D. Biochemistry 1998, 37, 10062-10073.
(15) Kast, P.; AsifUllah, M.; Hilvert, D. Tetrahedron Lett. 1996, 37, 2691-
2694. Dell, K. A.; Frost, J. W. J. Am. Chem. Soc. 1993, 115, 11581-
11589.
2-(Diethoxy-phosphoryl)-3-((3aS,5R,6R,6aS)-6-hydroxy-2,2-di-
methyl-tetrahydro-furo[2,3-d][1,3]dioxol-5-yl)-propionic Acid Meth-
yl Ester (5). Compound 4 (1.65 g, 3.5 mmol) was dissolved in 40 mL
of methanol and stirred with H2O (2.0 mL), ammonium formate (1.0
(16) Moore, B. S.; Cho, H. J.; Casati, R.; Kennedy, E.; Reynolds, K. A.; Mocek,
U.; Beale, J. M.; Floss, H. G. J. Am. Chem. Soc. 1993, 115, 5254-5266.
9
12958 J. AM. CHEM. SOC. VOL. 127, NO. 37, 2005

Isotope Effects on the Reactions of Chorismate
A R T I C L E S
g, 15.8 mmol), and 10% Pd/C (1.13 g) at 50 °C for 2 h. After cooling
to room temperature and filtering off the Pd/C, the solution was diluted
with H2O (10 mL) and concentrated to ca. 10 mL under reduced
pressure. Extraction with CHCl3 (3 × 40 mL), followed by drying over
MgSO4 and evaporation of the solvent, afforded 1.26 g (94%) of 5 as
a mixture of diastereomers as a clear oil. This material was used in the
next step without further purification. 1H NMR (300 MHz, CDCl3): δ
1.27-1.36 (m, 18H), 1.50 (s, 3H), 1.54 (s, 3H), 2.16-2.51 (m, 4H),
3.02-3.50 (m, 2H), 3.73 (s, 3H), 3.76 (s, 3H), 3.96-4.18 (m, 12H),
4.51 (s, 1H), 4.53 (s, 1H), 5.86 (d, J ) 3.9 Hz, 1H), 5.90 (d, J ) 3.6
Hz, 1H). HRMS-MALDI+ [M + Na]+: calcd (C15H27NaO9P) 405.1290;
found 405.1289.
130.9, 134.2, 166.9. HRMS-ESI+ [M + Na]+: calcd (C11H15DNaO5)
252.0957; found 252.0951.
Methyl [4-2H]Shikimate (9). Compound 8 (130 mg, 0.58 mmol)
was mixed with Dowex-50WX8-200 (220 mg, washed with methanol
three times and dried under vacuum for 30 min) in methanol (14 mL)
containing 0.7 mL of H2O. The mixture was stirred overnight at room
temperature under an atmosphere of nitrogen. The solution was then
filtered, and the resin was washed with methanol (3 × 5 mL). The
combined methanol layers were evaporated to afford a light yellow
oil. Purification by flash chromatography with silica absorbent (MeOH/
CH2Cl2, 1:9) yielded 81 mg (74%) of 9 as a white solid. mp 110-113
°C. 1H NMR (300 MHz, CD3OD): δ 2.10 (ddt, J ) 17.4, 3.9, 1.5 Hz,
1H), 2.60 (ddt, J ) 18.3, 4.5, 2.1 Hz, 1H), 3.64 (s, 3H), 3.89 (t, J )
5.1 Hz, 1H), 4.27 (t, J ) 1.8 Hz, 1H), 4.50 (br s, 1H), 6.69 (pent, J )
1.8 Hz, 1H). 13C NMR (75 MHz, CD3OD): δ 31.5, 51.0, 67.2, 68.4,
72.2 (t, J ) 22.5 Hz), 130.2, 139.2, 168.8. HRMS-ESI+ [M + Na]+:
calcd (C8H11DNaO5) 212.0644; found 212.0643.
2-(Diethoxy-phosphoryl)-3-((3aS,5R,6S,6aS,[6-2H])-6-hydroxy-
2,2-dimethyl-tetrahydro-furo[2,3-d][1,3]dioxol-5-yl)-propionic Acid
Methyl Ester (6). Crude 5 (1.26 g, 3.3 mmol) was stirred with
phosphorus pentoxide (765 mg, 2.7 mmol) in dry DMF (12.6 mL) and
dry DMSO (1.2 mL) under nitrogen at 70 °C for 3.5 h. After cooling
to room temperature and dilution with H2O (40 mL), the mixture was
extracted with CHCl3 (3 × 50 mL). The combined CHCl3 layers were
washed with H2O (15 mL), dried over MgSO4, and evaporated to give
1.1 g (88%) of the crude ketone as a light yellow oil. The latter was
dissolved in methanol (27.0 mL) and water (3.0 mL) and stirred for 5
min at 0 °C. Sodium borodeuteride (590 mg, 14.1 mmol) was then
slowly added to the solution over a 5 min period. After being stirred
at 0 °C for 1 h, the mixture was diluted with H2O (20 mL) and extracted
with CHCl3 (3 × 50 mL). The combined CHCl3 layers were dried over
MgSO4 and evaporated to yield 940 mg (85%) of 6 as a mixture of
diastereomers as a clear oil. This material was used in the next step
[4-2H]Chorismate (1e). This compound was prepared chemoenzy-
matically from compound 9 (40 mg, 0.21 mmol) as described for [10-
12C]chorismic acid (1a).10 It was obtained as an off-white solid (32
mg, 67%). 1H NMR (300 MHz, D2O, pD 6.0): δ 4.45 (d, J ) 2.7 Hz,
1H), 4.78 (d, J ) 2.1 Hz, 1H), 5.08 (d, J ) 2.7 Hz, 1H), 5.83 (d, J )
10.8 Hz, 1H), 6.20 (dd, J ) 9.9, 1.8 Hz, 1H), 6.43 (d, J ) 2.1 Hz,
2
1H). H NMR (61.4 MHz, CH3OH): δ 4.61 (br s, 12H).
Kinetic Isotope Effect Measurements. All manipulations were
performed on ice to suppress the nonenzymatic decomposition of
chorismate. Natural abundance chorismate or mixtures of specifically
labeled substrates were dissolved in 50 mM potassium phosphate buffer,
pH 7.5. The chorismate concentration, determined spectroscopically
at 274 nm (ꢀ 2630 M-1 cm-1), was between 1 and 2 mM for complete
conversion reactions and between 4 and 6 mM for partial conversion
reactions. The substrate solution (4 mL) was placed in a sidearm
decarboxylation flask fitted with a stopcock adapter. The solutions were
sparged with moist CO2-free nitrogen gas. For the enzymatic reactions,
the sample was warmed to 22 °C and an appropriate amount of
chorismate mutase (0.25 Units) was introduced through the sidearm.
Aliquots of the reaction mixture were removed, and the progress of
the reaction was monitored spectroscopically at 274 nm (∆ꢀ 2630 M-1
cm-1). At the desired endpoint (ca. 50% or 100% conversion), the
reaction was quenched by addition of 300 µL of 5 N HCl. After 45
min at room temperature, volatile products were collected under reduced
pressure, using two liquid nitrogen/pentane traps (-115 °C) to remove
contaminants and H2O and a liquid nitrogen trap (-196 °C) to trap
CO2. The 13CO2/12CO2 ratio of the sample (RP) was then determined
on a Finnigan delta E isotope ratio mass spectrometer as previously
described.9 The remaining reaction solution from partial conversion
experiments was subsequently treated with KOH to bring the pH to
6.5 and sparged with moist CO2-free nitrogen gas. Residual substrate
was then completely converted to product by addition of wild-type
BsCM. After 12 h at 22 °C, 5 N HCl was added and the decarboxylation
reaction was allowed to proceed for 45 min. The CO2 was distilled at
reduced pressure as described above, and the 13CO2/12CO2 ratio of the
residual substrate (RS) was determined by mass spectrometry. All
reactions were performed in triplicate.
1
without further purification. H NMR (300 MHz, CDCl3): δ 1.27-
1.39 (m, 18H), 1.59 (s, 3H), 1.62 (s, 3H), 2.22-2.51 (m, 4H), 3.14-
3.29 (m, 1H), 3.32-3.48 (m, 1H), 3.74 (d, J ) 0.6 Hz, 3H), 3.76 (d,
J ) 0.6 Hz, 3H), 3.92-4.08 (m, 2H), 4.10-4.22 (m, 8H), 4.60 (s,
1H), 4.61 (s, 1H), 5.69 (d, J ) 4.2, 1H), 5.72 (d, J ) 4.2 Hz, 1H).
HRMS-MALDI+ [M + Na]+: calcd (C15H26DNaO9P) 406.1352; found
406.1352.
2-(Diethoxy-phosphoryl)-3-((3aS,4R,6aS,[3a-2H])-6-hydroxy-2,2-
dimethyl-tetrahydro-furo[3,4-d][1,3]dioxol-4-yl)-propionic Acid Meth-
yl Ester (7). Compound 6 (940 mg, 2.45 mmol) was mixed with
Dowex-50WX8-200 (2.4 g, pre-washed with acetone three times and
dried under vacuum for 30 min) in dry acetone (25 mL) and stirred at
room temperature under nitrogen overnight. The solution was then
filtered, and the resin was washed with acetone (3 × 35 mL). The
combined acetone layers were evaporated to yield 823 mg (88%) of 7
as a mixture of diastereomers as a light yellow oil. This material was
used in the next step without further purification. 1H NMR (300 MHz,
CDCl3): δ 1.24-1.37 (m, 18H), 1.44 (s, 3H), 1.45 (s, 3H), 2.12-2.50
(m, 4H), 3.11-3.40 (m, 2H), 3.75 (s, 6H), 4.06-4.24 (m, 10H), 4.58
(s, 2H), 5.32 (s, 2H). HRMS-MALDI+ [M + Na]+: calcd (C15H26-
DNaO9P) 406.1352; found 406.1358.
Methyl 3,4-[4-2H](O-Isopropylidene)shikimate (8). A solution of
7 (820 mg, 2.15 mmol) in dry methanol (60 mL) was added dropwise
to a precooled solution of sodium methoxide, which was prepared by
adding dry methanol (60 mL) to sodium (160 mg, 7.0 mmol, washed
one time with hexanes), over 30 min under a flow of nitrogen while
maintaining the temperature at 0 °C. The mixture was warmed to room
temperature and stirred overnight. The reaction was then quenched by
addition of a saturated aqueous solution of NaHCO3 (30 mL) and
extracted with Et2O (3 × 60 mL). The combined Et2O layers were
washed with H2O (20 mL), dried over MgSO4, and evaporated to
provide a light yellow oil. Purification by flash chromatography with
silica absorbent (EtOAc/hexanes, 3:7) yielded 132 mg (27%) of 8 as a
For the nonenzymatic reactions, 4 mL chorismate samples (2-4 mM)
were incubated at 60 °C in the same buffer as that used in the enzyme-
catalyzed reactions. The progress of the reaction was monitored by 1H
NMR and UV-vis spectroscopy. After 30 min, corresponding to
conversion of ca. 50% of the starting material, the reaction was
quenched with 5 N HCl, and the 13CO2/12CO2 ratios, RP and RS, were
determined as described above. The R0 value for the samples,
corresponding to the 13CO2/12CO2 ratio obtained upon 100% conversion
of chorismate, was obtained by adding wild-type BsCM as described
above.
1
clear oil. H NMR (300 MHz, CDCl3): δ 1.40 (s, 3H), 1.45 (s, 3H),
2.03 (br s, 1H), 2.25 (ddt, J ) 16.5, 8.4, 1.8 Hz, 1H), 2.80 (ddt, J )
17.4, 4.8, 0.9 Hz, 1H), 3.77 (s, 3H), 3.90 (pent, J ) 3.9 Hz, 1H), 4.74
(d, J ) 3.3 Hz, 1H), 6.90-6.94 (m, 1H). 13C NMR (75 MHz, CD2-
Cl2): δ 25.9, 28.1, 29.8, 52.3, 69.0, 72.6, 77.9 (t, J ) 22.5 Hz), 109.9,
NMR Kinetics. The reaction of chorismate and [4-2H]chorismate
(20 mM) was monitored at 60 °C by 1H NMR spectroscopy in
deuterated phosphate buffer (200 mM, pD 7.5) containing dioxane (1.0
9
J. AM. CHEM. SOC. VOL. 127, NO. 37, 2005 12959

A R T I C L E S
Wright et al.
Scheme 2. Isotopically Labeled Chorismate Derivatives
Table 1. Kinetic Isotope Effects on the Rearrangement Catalyzed
by Wild-Type BsCM, C75S BsCM, and EcCMa
label
enzyme
18O-7
13C-1
13C-9
wt BsCM
1.045 ( 0.003b 1.0043 ( 0.0002b 1.0129 ( 0.0024
1.053 ( 0.002b 1.0057 ( 0.0002b 1.0110 ( 0.0022
C75S BsCM
EcCMc
1.046 ( 0.0002 1.015 ( 0.0002
1.0131
n.d.
Becke3LYP 6-31+G* d 1.049
1.0178
mM) as an internal standard. The concentrations of chorismate,
prephenate, phenylpyruvate, and 4-hydroxybenzoate were determined
periodically by integration of representative, well-resolved peaks and
comparison to the internal standard. The kinetic data were analyzed
with the program DynaFit.17 The decomposition of [4-2H]chorismate
a In 50 mM potassium phosphate buffer, pH 7.5, and 22 °C. KIEs are
corrected by division by the observed secondary kinetic isotope effect at
C10, which was measured to be 1.0009 ( 0.0001 for wild-type BsCM,
1.0016 ( 0.0002 for the C75S variant, and 1.0033 ( 0.0001 for EcCM.
b Reference 9. c n.d., not determined. d Calculated at 295.15 K.
2
in phosphate buffer (pH 6.0, 60 °C) was similarly monitored by H
NMR spectroscopy using dioxane-d8 as the internal standard. All
experiments were performed in duplicate.
ing only natural abundance 13C and the 1a/1b, 1a/1c, and 1a/
1d substrate mixtures were converted enzymatically to prephen-
ate in separate experiments. In each case, three complete
conversion reactions were performed with 4-6 mM chorismate
with wild-type BsCM, and three partial conversion reactions
(to ca. 50%) were carried out for the wild-type enzyme, the
C75S BsCM variant, and EcCM. At the desired endpoint, the
reaction was quenched with 5 M HCl and volatile products were
collected under reduced pressure, using two pentane/liquid N2
traps at -115 °C to remove H2O and other contaminants and a
liquid N2 trap at -196 °C to trap CO2. The 13CO2/12CO2 ratio
of the sample was then determined on a Finnigan delta E isotope
ratio mass spectrometer, and the heavy atom isotope effects were
calculated according to eq 1:
Computational Studies. All calculations were performed using the
G03 series of programs18 at the B3LYP/6-31+G* level of theory.
Transition structures were characterized by harmonic frequency calcula-
tions and had exactly one negative eigenvalue, corresponding to the
reaction coordinate, as confirmed by intrinsic reaction coordinate (IRC)
calculations. Isotope effects were calculated using QUIVER19 and the
harmonic frequencies, scaled by 0.961, at temperatures of 295.15 and
333.15 K and are corrected for tunneling with a one-dimensional
tunneling approximation as proposed by Bell.19,20 An extended, diequa-
torial conformation of chorismate served as the reference state for the
isotope effect calculations (see Supporting Information).
Results
Labeled Chorismates. Primary heavy atom isotope effects
on both the enzymatic and the nonenzymatic rearrangement of
chorismate to prephenate were determined by the remote label
method,21 exploiting the C10 carboxylate of chorismate as the
indicator position (Scheme 2). Prephenate decarboxylates
quantitatively under acidic conditions,22 releasing the carbon
at the indicator position as CO2. The isotopic composition of
the latter was determined by isotope ratio mass spectrometry,
enabling determination of isotope effects at the mechanistically
interesting key positions O7, C1, and C9.
A ca. 99:1 mixture of [10-12C]chorismate (1a) and [7-18O,-
10-13C]chorismate (1b) was used to determine a primary isotope
effect at the site of bond breaking, whereas ca. 99:1 mixtures
of [10-12C]chorismate (1a) and [1,10-13C]chorismate (1c) or [9,-
10-13C]chorismate (1d) were used to determine the isotope
effects at the sites of bond formation. The required chorismate
derivatives were prepared chemoenzymatically. Biosynthetic
transformation of mixtures of singly and doubly labeled
shikimates gave the 1a/1b and 1a/1c mixtures, as previously
described.10 In contrast, the 1a/1d mixture was prepared by
mixing independently synthesized 1a and 1d in the desired ratio.
The latter mixture was treated with acid prior to enzymatic
reaction to decompose contaminating prephenate originating
from either of the components, which could serve as a source
of CO2 and skew the isotopic ratios.
log(1 - f)
log[1 - f(RP/R0)]
KIE )
(1)
where R0 is the isotope ratio of the remote label in the starting
material (obtained by running the reaction to completion), RP
is the isotope ratio of the remote label in the product, and f is
the fractional conversion of substrate to product. The observed
values for the 1a/1b, 1a/1c, and 1a/1d mixtures were divided
by the secondary kinetic isotope effect at C10, measured with
unlabeled chorismate to give the desired 18O and 13C isotope
effects.
Data for the BsCM variants are summarized in Table 1,
together with theoretical isotope effects based on the transition
structure for the concerted reaction obtained at the Becke3LYP/
6-31+G* level of theory. As was previously reported,9 a very
large 18O effect on C-O cleavage and a smaller, but still
significant, 13C effect at C1 are observed for the BsCM-catalyzed
reactions. The 1.3% and 1.1% 13C effects at C9 for the BsCM
and C75S BsCM variants, respectively, confirm that C-C bond
formation lags considerably behind C-O bond cleavage in the
enzyme-catalyzed reaction. Data for the structurally unrelated
helical bundle mutase EcCM, which are also reported in Table
1, are qualitatively similar to those obtained with BsCM. The
4.6% 18O KIE at O7 and the smaller 1.5% 13C KIE at C1 closely
match the corresponding theoretical isotope effects.
Enzymatic Reactions. For determination of the isotope
effects on the enzymatic reaction, unlabeled chorismate contain-
Nonenzymatic Reactions. Chorismate decomposes with a
t1/2 of ∼30 min at 60 °C.4 Under these conditions, elimination
of the enolpyruvyl side chain to give 4-hydroxybenzoate
competes with the Claisen rearrangement to prephenate, and
prephenate is also further converted to phenylpyruvate, as
summarized in Scheme 3.4
(17) Kuzmic, P. Anal. Biochem. 1996, 237, 260-273.
(18) Frisch, M. J.; et al. Gaussian 03, revision C.01; Gaussian, Inc.: Wallingford,
CT, 2004.
(19) Saunders, M.; Laidig, K. E.; Wolfsberg, M. J. Am. Chem. Soc. 1989, 111,
8989-8994.
(20) Bell, R. P. The Tunnel Effect in Chemistry; Chapman and Hall: New York,
1980; pp 60-63.
(21) O’Leary, M. H. Methods Enzymol. 1980, 64, 83-104.
(22) Zamir, L. O.; Tiberio, R.; Jensen, R. A. Tetrahedron Lett. 1983, 24, 2815-
2818.
The thermal reaction can be directly monitored by 1H NMR
spectroscopy. A typical time course for the decomposition
9
12960 J. AM. CHEM. SOC. VOL. 127, NO. 37, 2005

Isotope Effects on the Reactions of Chorismate
A R T I C L E S
Figure 1. Kinetic analysis of the nonenzymatic decomposition of chorismate (A) and [4-2H]chorismate (B) in 200 mM phosphate buffer in D2O at 60 °C.
The reactions were monitored by 1H NMR spectroscopy. Relative concentrations of chorismate (red b), prephenate (blue 1), phenylpyruvate (green 9),
4-hydroxybenzoate (magenta 2), and pyruvate (black [) are plotted as a function of time. The data were analyzed with the program DynaFit.17
Scheme 3
Table 2. Kinetic Isotope Effects on the Thermal Reactions of
Chorismatea
label
product
18O-7
13C-1
2H-4
prephenate
1.0482 ( 0.0005 1.0118 ( 0.0004 1.07 ( 0.04
1.0119 0.87
Becke3LYP 6-31+G*b 1.0427
4-hydroxybenzoate 1.0374 ( 0.0005 1. 0113 ( 0.0004 1.85 ( 0.06
Becke3LYP 6-31+G*b 1.0465
1.0022 1.45
of unlabeled chorismate at 60 °C in 200 mM sodium phosphate
buffer in D2O (pD 7.5) is shown in Figure 1A. The rate constants
obtained are k1 ) (2.98 ( 0.01) × 10-4 s-1, k2 ) (2.59 ( 0.05)
× 10-5 s-1, and k3 ) (3.26 ( 0.09) × 10-5 s-1. Although a
k1/k3 ratio of ca. 4 was previously reported based on indirect
measurements of the reaction constituents,4 we find a value of
9. Under the conditions of our experiment, the pyruvate that is
produced in the elimination step also exchanges with sol-
vent with an apparent rate constant of k4 ) (6.8 ( 0.6) × 10-4
a In 100 mM potassium phosphate, pH 7.5, heated for 30 min at 60 °C.
KIEs are corrected by division by the observed secondary kinetic isotope
effect at C10, which was measured to be 1.0011 ( 0.002 for the formation
of both prephenate and 4-hydroxybenzoic acid. The buffer concentration
for the deuterium kinetic isotope effects was 200 mM. b Calculated at 333.15
K (see Supporting Information).
4-hydroxybenzoate estimated from the product distribution
ratios, and R ) [f(RP/R0)]/[1 - (1 - f)(Rs/R0)]. The values are
summarized in Table 2.23 The 3.74% 18O isotope effect indicates
extensive C-O bond cleavage in the transition state for the
elimination. The 1.13% 13C isotope effect at C1 reflects the
decrease in bond order associated with the increase in positive
charge at this site.
s-1
.
Heavy atom isotope effects on the nonenzymatic Claisen
rearrangement to give prephenate, 18k1 and 13k1, were determined
at 60 °C as described for the enzymatic reactions and calculated
from the following relationship:
To gain further insight into the competing elimination
reaction, [4-2H]chorismate (1e) was prepared chemoenzymati-
cally according to Scheme 4. Its decomposition was monitored
at 60 °C in 200 mM sodium phosphate buffer in D2O (pD 7.5)
[1 - (1 - f)(RS/R0)]
[f(RP/R0)]
log(1 - f)
log[(1 - f)(RS/R0)]
xk1 )
(2)
(
)(
)
where xk1 is the apparent isotope effect, expressed as a ratio of
the rate constants for the light and heavy (x) isotopes. A
derivation of eq 2 is provided in the Supporting Information.
The 18O isotope effect at O7 and the 13C effect at C1 are in
excellent agreement with the theoretical isotope effects (Table
2).
Because only small amounts of hydroxybenzoic acid are
formed in the thermal reaction, it was not possible to determine
the heavy atom isotope effects, 18k3 and 13k3, on the elimination
reaction directly. However, these KIEs were estimated according
to eq 3 (for a derivation, see Supporting Information):
1
by H NMR spectroscopy as shown in Figure 1B. The rate
constants obtained are k1 ) (2.78 ( 0.10) × 10-4 s-1, k2 )
(2.19 ( 0.04) × 10-5 s-1, and k3 ) (1.77 ( 0.03) × 10-5 s-1
,
affording deuterium isotope effects of 1.07 ( 0.04, 1.19 ( 0.03,
and 1.85 ( 0.06 on the k1, k2, and k3 steps, respectively. Based
on the value of Dk3, C-H bond cleavage is not far advanced in
the transition state for the elimination. The fate of the deuterium
atom at C4 in this pathway was determined by 2H NMR
spectroscopy at 60 °C and pH 6.0 (to minimize exchange with
solvent). In addition to [4-2H]phenylpyruvate, formation of
(23) The agreement between theory and experiment is poorer for the heavy atom
isotope effects on the elimination reaction than for the Claisen rearrange-
ment, probably due to the indirect nature of the experimental determinations.
The discrepancy in the case of the deuterium isotope effects can be attributed
to the fact that the one-dimensional tunneling correction is insufficient.
An explicit quantum mechanical treatment of the nuclear hydrogen
coordinate is likely to be required, but is currently not possible for systems
of this size.
xk1
xk )
(3)
(1 + k1/k3)/R - (k1/k3)
3
where xk1 is the isotope effect value obtained from eq 2, k1 and
k3 are the rate constants for the formation of prephenate and
9
J. AM. CHEM. SOC. VOL. 127, NO. 37, 2005 12961

A R T I C L E S
Wright et al.
Scheme 4. Synthesis of [4-2H]Chorismate (1e)a
For the chorismate-to-prephenate transformation, density
functional theory (DFT) calculations in vacuo predict a some-
what looser aromatic transition state than the canonical aliphatic
Claisen rearrangement, with bond lengths of 2.127 and 2.660
Å for the breaking C-O and forming C-C bonds, respectively
(Figure 3).9,28 In qualitative agreement with this model, second-
ary tritium isotope effects on the thermal reaction indicated
significant cleavage of the C-O bond (ca. 40%) but little or
no formation of the C-C bond.1 The heavy atom isotope effects
determined in the current study confirm and extend this picture.
The very large 18O effect at O7, the site of bond breaking, and
the smaller, but still significant 13C effect at C1, the site of bond
making, are both in excellent accord with the kinetic isotope
effects calculated at the Becke3LYP/6-31+G* level of theory
for a pericyclic transition state (Table 1). This picture is rounded
2
out by the observation of a small secondary H effect for the
hydrogen at C4, consistent with hyperconjugative stabilization
of the developing positive charge on the ring. In this case, the
agreement between theory and experiment is not as good
(observed, 1.07 ( 0.04; predicted 0.87), presumably due to the
fact that the dihedral angle between the hydrogen at C4 and
the cyclohexadienyl ring is not predicted accurately; even
relatively small changes in this angle will influence the extent
of hyperconjugation, leading to significant uncertainty in the
calculated isotope effect at this position. Together, the data
provide strong experimental evidence that the thermal rear-
rangement is a concerted, albeit asynchronous, [3,3]-sigmatropic
process.29
a Conditions: (a) (CF3SO2)2O, py, CH2Cl2, -30 °C; (b) (EtO)2P(O)CH2-
CO2Me, NaH, DMF, 83% over two steps; (c) Pd/C, MeOH, HCO2NH4, 50
°C, 94%; (d) P2O5, DMSO, DMF, 70 °C; (e) NaBD4, MeOH, 0 °C, 85%
over two steps; (f) Dowex-50w, acetone, 88%; (g) MeOH, Na0, 0 °C, 27%;
(h) Dowex-50w, MeOH/H2O, 74%; (i) NaOH, THF, H2O; (j) KA12/
pKAD50, Tris-HCl, PEP, ATP, FMN, KCl, MgSO4, NaS2O4, 67% over
two steps.
[3-2H]pyruvate was directly observed (Figure 2). Based on the
ratio of integrated signals for pyruvate to 4-hydroxybenzoate
In principle, chorismate mutases, which accelerate this
transformation more than a million-fold, could lower the energy
barrier for rearrangement simply by stabilizing the pericyclic
transition state observed in solution. Alternatively, the enzymes
might perturb this structure or even favor alternative mecha-
nisms, such as a two-step dissociative pathway or other routes.30
The observation of a large 18O effect at O7, and a smaller 13C
effect at C1, for the reaction catalyzed by the AroH mutase
from B. subtilis had previously been used to argue that the
enzymatic reaction, like its solution counterpart, is a concerted
pericyclic process with some C-C bond formation accompany-
ing C-O bond cleavage.9 The 13C isotope effect of 1.0129 (
0.0024 at C9 of chorismate with wild-type BsCM, which was
determined in the current study, strengthens this conclusion. A
stepwise mechanism should be insensitive to isotopic substitu-
tion at C9, because the ion pair intermediate would be expected
to partition preferentially forward to give prephenate, rather than
1
observed by H NMR spectroscopy after lyophilization of the
sample, 62% of the pyruvate formed after 3 h contains deuterium
originating from C4 of the labeled chorismate.
Discussion
Aliphatic and aromatic Claisen rearrangements have been
extensively investigated by theory24 and experiment.25,26 High-
level calculations indicate that these reactions are concerted and
proceed via cyclically delocalized, chairlike transition states in
which carbon-oxygen bond cleavage precedes carbon-carbon
bond formation. Experimental 13C and 2H isotope effects have
been determined by multisite NMR methods for the archetypal
rearrangement of allyl vinyl ether and allyl phenyl ether, which
exhibit excellent quantitative agreement with calculated isotope
effects.27 These findings provide confidence in the ability of
theory to account for the multidimensional bonding changes that
occur along the reaction coordinate.
(27) Meyer, M. P.; DelMonte, A. J.; Singleton, D. A. J. Am. Chem. Soc. 1999,
121, 10865-10874.
(24) Dewar, M. J. S.; Healy, E. F. J. Am. Chem. Soc. 1984, 106, 7127-7136.
Vance, R. L.; Rondan, N. G.; Houk, K. N.; Jensen, F.; Borden, W. T.;
Komornicki, A.; Wimmer, E. J. Am. Chem. Soc. 1988, 110, 2314-2315.
Dewar, M. J. S.; Jie, C. J. Am. Chem. Soc. 1989, 111, 511-519. Wiest,
O.; Black, K. A.; Houk, K. N. J. Am. Chem. Soc. 1994, 116, 10336-
10337. Davidson, M. M.; Hillier, I. H.; Vincent, M. Chem. Phys. Lett. 1995,
246, 536-540. Yamabe, S.; Okumoto, S.; Hayashi, T. J. Org. Chem. 1996,
61, 6218-6226.
(28) This structure has slightly shorter C-O and C-C bond lengths than the
transition state reported previously (ref 9), but the effect on the calculated
heavy atom isotope effects is small. The relative looseness of the calculated
transition state probably reflects electrostatic repulsion of the negatively
charged carboxylate groups in the chair geometry, as well as the known
bias of the B3LYP method (ref 27).
(29) An alternative two-step mechanism involving an ion pair intermediate was
also considered. However, efforts to locate the transition structure for a
fully dissociative process computationally inevitably led to transition
structures for the concerted elimination pathway (Figure 4). Moreover, such
a mechanism would predict identical deuterium kinetic isotope effects on
both the rearrangement and the elimination processes, which are not
observed experimentally (Table 2). Finally, the >1% 13C isotope effect
seen at C9 in the enzymatic reactions (Table 1) also argues against a
stepwise mechanism, because the bond order to C9 is not expected to
decrease substantially in such a process. Although the C-C-O-C torsional
mode would be lost in the enolate intermediate, this will not give a full
1% isotope effect.
(25) Ziegler, F. E. Chem. ReV. 1988, 88, 1423-1452. Gajewski, J. J.; Conrad,
N. D. J. Am. Chem. Soc. 1979, 101, 2747-2748. McMichael, K. D.; Korver,
G. L. J. Am. Chem. Soc. 1979, 101, 2746-2747. Coates, R. M.; Rogers,
B. D.; Hobbs, S. J.; Peck, D. R.; Curran, D. P. J. Am. Chem. Soc. 1987,
109, 1160-1170. Kupczyk-Subotkowska, L.; Saunders, W. H., Jr.; Shine,
H. J. J. Am. Chem. Soc. 1988, 110, 7153-7159. Kupczyk-Subotkowska,
L.; Subotkowski, W.; Saunders, W. H., Jr.; Shine, H. J. J. Am. Chem. Soc.
1992, 114, 3441-3445. Kupczyk-Subotkowska, L.; Saunders, W. H., Jr.;
Shine, H. J.; Subotkowski, W. J. Am. Chem. Soc. 1993, 115, 5957-5961.
Gajewski, J. J.; Brichford, N. L. J. Am. Chem. Soc. 1994, 116, 3165-
3166. Gajewski, J. J. Acc. Chem. Res. 1997, 30, 219-225.
(26) Gajewski, J. J.; Jurayj, J.; Kimbrough, D. R.; Gande, M. E.; Ganem, B.;
Carpenter, B. K. J. Am. Chem. Soc. 1987, 109, 1170-1186.
(30) Guilford, W. J.; Copley, S. D.; Knowles, J. R. J. Am. Chem. Soc. 1987,
109, 5013-5019.
9
12962 J. AM. CHEM. SOC. VOL. 127, NO. 37, 2005

Isotope Effects on the Reactions of Chorismate
A R T I C L E S
2
Figure 2. Decomposition of [4-2H]chorismate (1e) in 100 mM phosphate buffer at pH 6.0 and 60 °C monitored by H NMR spectroscopy.
electronically relative to their gas-phase counterparts,3a,32 al-
though the computed C5-O7 and C1-C9 distances for the
enzyme-bound transition state are longer and shorter, respec-
tively, than found experimentally. These discrepancies probably
reflect a combination of experimental inaccuracy, chemical
differences between the wild-type and mutant enzymes, and
computational limitations with such large systems.
EcCM, the second enzyme examined in this study, is
representative of the AroQ family of chorismate mutases. It is
evolutionarily unrelated to BsCM, but exhibits similar steady-
state parameters14,33 and has a comparably configured active
site.6 In this case, the isotope effects closely match the predicted
values (Tables 1), indicating that an ionic active site environment
need not necessarily alter the structure of the Claisen transition
state. However, this conclusion rests on the uncertain assumption
that chemistry is fully rate determining in this system. Although
viscosogens have no effect on EcCM activity,13 it is conceivable
that a kinetically significant, but isotopically insensitive, con-
formational isomerization of bound chorismate precedes the
actual rearrangement. This step might correspond to conversion
of chorismate from its preferred pseudoequatorial conformation
to a pseudoaxial conformation in which the carboxyvinyloxy
group is aligned for reaction.34 Such a scenario has been
proposed to explain why isotope effects on the reaction catalyzed
by the related E. coli chorismate mutase-prephenate dehydro-
genase are suppressed.1
In the absence of enzyme, conversion of chorismate to
4-hydroxybenzoate and pyruvate competes with the Claisen
rearrangement leading to prephenate.4 As in the transformation
of 3-(methoxycarbonyl)-5-ethenoxycyclohexadiene-cis-6-d to
give methyl benzoate and CH2DCHO,26 this side reaction occurs
via a concerted pericyclic mechanism in which C9 removes a
proton (deuteron) from C4 rather than adding to C1 (Scheme
Figure 3. Becke3LYP/6-31+G* transition structure for the Claisen
rearrangement of chorismate.
revert to chorismate.29 The significant isotope effects observed
at both C1 and C9, the two ends of the reacting π system,
strengthen the conclusion that the C-C bond is partially formed
in the transition state.
Significant slowing of the BsCM-catalyzed reaction by
viscosogens such as glycerol points to the chemical step being
only partially rate determining in this system.13 As a conse-
quence, the intrinsic isotope effects may not be fully expressed.
Consistent with this possibility, the sluggish C75S variant, which
is insensitive to changes in microviscosity,13 exhibits a signifi-
cantly larger 18O isotope effect than BsCM itself or the
theoretical prediction.9 Moreover, the 13C isotope effects at C1
and C9 are somewhat smaller than predicted by theory (Table
1). Together, these data imply that the enzymatic reaction may
have a more dissociative transition state than the thermal process.
Such a species could be stabilized by ionic groups that line the
active site,7 including Arg90, which directly contributes a
hydrogen bond to the ether oxygen of chorismate in the
transition state.31 Hartree-Fock calculations support the ability
of such interactions to alter the reactant and transition state
(32) Worthington, S. E.; Roitberg, A. E.; Krauss, M. Int. J. Quantum Chem.
2003, 94, 287-292.
(33) Stewart, J.; Wilson, D. B.; Ganem, B. J. Am. Chem. Soc. 1990, 112, 4582-
4584.
(31) Kienho¨fer, A.; Kast, P.; Hilvert, D. J. Am. Chem. Soc. 2003, 125, 3206-
3207.
(34) Copley, S. D.; Knowles, J. R. J. Am. Chem. Soc. 1987, 109, 5008-5013.
9
J. AM. CHEM. SOC. VOL. 127, NO. 37, 2005 12963

A R T I C L E S
Scheme 5
Wright et al.
5). Substantial transfer of deuterium from C4 of chorismate to
pyruvate is observed, arguing against the involvement of an
external base. The amount of label lost to solvent (ca. 40% after
3 h reaction) is readily explained by exchange under the
experimental conditions. The 2H isotope effect on the elimination
(1.85 ( 0.06) is similar in magnitude to that observed for
3-(methoxycarbonyl)-5-ethenoxycyclohexadiene-cis-6-d. In ad-
dition, like the Claisen rearrangement, elimination of the
enolpyruvyl group is subject to a large 18O isotope effect at O7
(1.0374 ( 0.0005). These results show that the C-O bond is
substantially cleaved in the transition state but that transfer of
the hydrogen from C4 to C9 is not far advanced, in qualitative
agreement with theoretical calculations (Figure 4, Table 2).23
Both reactions, elimination and rearrangement, thus appear to
be concerted but highly asynchronous.
Figure 4. Becke3LYP/6-31+G* transition structure for the conversion of
chorismate to 4-hydroxybenzoate.
about the nature of the transition state of the chorismate mutase
reaction both in solution and at the active sites of evolutionarily
unrelated enzymes. In all cases, the reaction proceeds via a
pericyclic transition state in which C-C bond formation lags
considerably behind C-O bond cleavage. Excellent quantitative
agreement between theory and experiment is found for the
uncatalyzed reaction, whereas the ionic active site of the
enzymes may polarize the transition state somewhat as compared
to its solution counterpart. The competing reaction of chorismate
leading to 4-hydroxybenzoate, an intermediate in ubiquinone
biosynthesis, likewise appears to occur via an asynchronous
concerted mechanism. The latter finding raises the intriguing
possibility that additional protein catalysts promoting pericyclic
processes will be found.
Although the elimination is a minor side reaction thermally,
the enzyme chorismate pyruvate-lyase efficiently converts
chorismate to 4-hydroxybenzoate for the ubiquinone biosynthetic
pathway.35 Our results suggest that the latter enzyme and the
related isochorismate pyruvate-lyase, which converts isocho-
rismate to salicylate for the biosynthesis of many siderophores,36
might operate by otherwise rare pericyclic mechanisms.
Acknowledgment. This work was supported by NIH Grant
GM 18938 to W.W.C., NIH Grant GM 20571 to S.K.W., NSF
grant DMR0079647 to O.W., and the Schweizer Nationalfonds
to D.H. We are grateful to Prof. J. Frost for his gift of plasmid
pKAD50 and to Prof. D. Singleton for helpful discussions.
In summary, competitive experiments with isotopically
labeled chorismate derivatives have provided direct information
Supporting Information Available: Coordinates and com-
putational details for calculation of the theoretical isotope effects,
plus derivations of eqs 2 and 3. This material is available free
of charge via the Internet at http://pubs.acs.org.
(35) Siebert, M.; Severin, K.; Heide, L. Microbiology-UK 1994, 140, 897-
904. Holden, M. J.; Mayhew, M. P.; Gallagher, D. T.; Vilker, V. L. Biochim.
Biophys. Acta-Protein Struct. Mol. Enzymol. 2002, 1594, 160-167.
(36) Serino, L.; Reimmann, C.; Baur, H.; Beyeler, M.; Visca, P.; Haas, D. Mol.
Gen. Genet. 1995, 249, 217-228. Serino, L.; Reimmann, C.; Visca, P.;
Beyeler, M.; Della Chiesa, V.; Haas, D. J. Bacteriol. 1997, 179, 248-257.
Gaille, C.; Kast, P.; Haas, D. J. Biol. Chem. 2002, 277, 21768-21775.
JA052929V
9
12964 J. AM. CHEM. SOC. VOL. 127, NO. 37, 2005