1514 J. Am. Chem. Soc., Vol. 119, No. 7, 1997
Sun et al.
dehydrogenase reaction, followed at 360 nm with a Shimadzu 160U
UV/vis spectrophotometer interfaced to an IBM-compatible computer.
Hydration of Fluoropyruvate. The proton NMR spectrum of
fluoropyruvate in D2O at 30 °C in 0.1 M citrate buffer (pD 6.5) shows
doublets at δ ) 4.5 ppm (CH2 of the hydrate form) and 5.5 ppm (CH2
of the keto form), both peaks being split by a hydrogen-fluorine nuclear
spin-spin interaction with a splitting of J ) 46 Hz. From the relative
height of the two doublets, the percentage of hydration of fluoropyruvate
was calculated to be about 90%. Similar results were obtained by
Kokesh16 (85% hydration) and by Hurley and co-workers17 (95%
hydration). The predominance of the hydrate in aqueous solution is
confirmed by the natural-abundance 13C NMR: doublet at δ ) 86 ppm
(CH2F, J ) 159 Hz); doublet at δ ) 93 ppm (C(OH)2, J ) ca. 15 Hz);
singlet at δ ) 175 ppm (CO2-). No keto-carbonyl signal was detected
above the noise level, in agreement with Hurley and co-workers.17
If the keto form of fluoropyruvate is the substrate for SCPDC, at
sufficiently high enzyme concentrations the dehydration reaction should
become rate limiting. This does not occur under the conditions used
here. When the SCPDC concentration was increased by 1.20-fold, the
observed rate was increased by a factor of 1.12 ( 0.03 ([fluoropyruvate]
) 1 mM), 1.23 ( 0.06 ([fluoropyruvate] ) 10 mM), and 1.23 ( 0.08
([fluoropyruvate] ) 40 mM), with the rate becoming independent of
[fluoropyruvate] at ca. 5 mM. The observed rates are therefore not
the non-enzymic dehydration rates.
Competitive Measurements of 13C Isotope Effects. Procedures
for the competitive measurements followed those described by O’Leary.18
A solution containing 30 mM fluoropyruvate, 1 mM MgCl2, and 10
µM ThDP in 0.1 M potassium phosphate buffer, pH 6.5, was divided
into two portions. Each portion was placed in a closed vessel, and
then purged with N2 gas for 3 h to remove residual CO2. After the
solutions were incubated in a water bath at 25 °C for 30 min, pyruvate
decarboxylase was injected into both solutions to initiate the reactions.
The so called “low conversion” solution (20 mL) was allowed to
decarboxylate to the extent of approximate 10%, and then the reaction
was terminated by addition of concentrated phosphoric acid to lower
the pH below 2. The “high conversion” solution (2 mL) was allowed
to decarboxylate completely before addition of acid. The carbon
dioxide formed from the reactions was isolated on a vacuum line using
dry ice-isopropyl alcohol and liquid nitrogen traps. The vacuum line
maintained a vacuum of 400-800 µTorr throughout the distillation.
The extent of the reaction was checked by assay of the remaining
fluoropyruvate with lactate dehydrogenase. Isotopic ratios for isolated
CO2 were measured with a Finnigan Delta S isotope ratio mass
spectrometer, with appropriate correction for 17O. Isotope ratios and
isotope effects were calculated as described previously.19
Kinetic Technique for Solvent Isotope Effects. Citrate buffers
(0.1 M) were prepared with use of either pure H2O or 99.9% D2O as
solvent. Solvent isotope effects were calculated from the kinetic
parameters measured in 99.9% D2O buffers and H2O buffers. In proton
inventory studies, a mixture of appropriate volumes of D2O buffer and
H2O buffer gave the required atom fraction of deuterium n. All
reactions in H2O/D2O mixtures used the same substrate solution
(prepared in H2O), and were started by addition of the same enzyme
solutions (prepared in H2O), so that the errors from enzyme and
substrate solutions would tend to cancel. The value of n was corrected
for protium introduced in buffer preparation and by enzyme and
substrate solutions. Calibrations of fluoride electrode were done in
H2O, D2O, and 50% D2O (see Data Reduction section below).
Data Reduction. A calibration curve for the fluoride ion electrode
was obtained for each set of experiments by fitting the potential in
millivolts (mV) as a function of fluoride concentration [F-] at constant
ionic strength to the Nernstian line: mV ) A - B log [F-]. The slope
B and intercept A were obtained for n ) 0, 0.5, and 1, with B essentially
constant at 59 mV and A typically a few percent larger in D2O. Values
of A and B at n between 0 and 0.5, or n between 0.5 and 1, were
calculated assuming a linear relation of A or B vs n. The fluoride ion
systems of Figure 1a (intracatalytic control of active-site access
through sulhydryl addition-elimination at the regulatory site:
exemplified by pyruvate with SCPDC and wheat-germ PDC),
Figure 1b (occupancy of the regulatory site blocks the active
site open: exemplified by fluoropyruvate with SCPDC), and
Figure 1c (no occupancy of a regulatory site required, active
site permanently open: exemplified by fluoropyruvate with
wheat-germ PDC, pyruvate with Zymomonas PDC, and pyruvate
with the cysteine-221-serine mutant constructed by Baburina
and co-workers4). If this is so then interaction of SCPDC with
fluoropyruvate would carry it from its normal, intracatalytically
regulated conformation into a conformation with the active site
frozen open by the presence of fluoropyruvate in the regulatory
site. The enzyme could be driven into this form either directly
or by passage through a number of abnormal catalytic cycles
involving formation and hydrolysis of acetyl-ThDP in the active
site. Interaction of fluoropyruvate with wheat-germ PDC would
similarly carry that enzyme into a form with permanently open
active site not requiring a regulator molecule in the regulatory
site, again either directly or by way of a number of abnormal
catalytic cycles. If this model were correct, it could emerge
that the Zymomonas enzyme would exhibit mainly conforma-
tional differences from SCPDC. Such differences might involve
only the disposition of groups forming the access machinery of
the active site; these groups might consitute as small an array
as the side chains of one or a few residues, or as large an array
as the domains of a subunit or the subunits themselves.14
Experimental Section
Materials. Thiamin diphosphate hydrochloride (“cocarboxylase”),
sodium fluoropyruvate, potassium fluoride, and NADH were purchased
from Sigma, sodium fluoride standard solution (0.1000 M), phosphoric
acid, and anhydrous citric acid from Fisher, and sodium citrate dihydrate
and magnesium sulfate (anhydrous) from J. T. Baker. Deuterium oxide
was obtained from Aldrich (99.9% D).
Enzymes. Pyruvate decarboxylase (EC 4.1.1.1) from Saccharomyces
cereVisaie (specific activity 9 units/mg; specific activity15 of the pure
enzyme, 80 units/mg) was purchased from Sigma, suspended in a
solution of 5% glycerol, 3.2 M ammonium sulfate, 5 mM potassium
phosphate, 1 mM magnesium acetate, 0.5 mM EDTA, and 25 µM
ThDP, pH 6.5. Alcohol dehydrogenase (EC 1.1.1.1) from baker’s yeast
with a specific activity of 360 units/mg and lactate dehydrogenase from
rabbit muscle Type II were also obtained from Sigma. Alcohol
dehydrogenase (EC 1.1.1.1) from horse liver was purchased from Fluka
with a specific activity of 1.6 units/mg of protein.
Kinetic Method. The decarboxylation of fluoropyruvate was
followed by measurement of the fluoride ion product with use of a
fluoride ion electrode (Fisher) vs a calomel reference electrode with a
Beckman model 4500 pH meter (read to 0.1 mV) interfaced to an IBM
compatible computer (data acquisition at 10-30-s intervals for a period
of 5-10 min). Data reduction is described below. The reaction was
conducted at 30.0 °C in a glass vessel contained in a jacketed beaker
through which thermostated water was circulated (reaction solution
volume 25.5 mL). The fluoride electrode was calibrated at 30 °C by
a standard NaF solution in 0.1 M citrate buffer containing 5 mM MgSO4
and 5 mM ThDP over the range of fluoride ion concentrations produced
in the reaction before each experiment, and 50 µM potassium fluoride
was added to the reaction solution to provide a stable background
reading. The reaction was started by addition of a SCPDC solution to
the reaction vessel. The final concentrations for the reaction were as
follows: SCPDC ) 11.5 nM, ThDP ) 5 mM, MgSO4 ) 5 mM, KF
) 50 µM, fluoropyruvate ) 0.1-30 mM, 0.1 M citrate buffer (all at
pH 6.0).
Detection of aldehyde formation from decarboxylation of fluoropy-
ruvate was attempted by coupling to the yeast or horse liver alcohol
(16) Kokesh, F. C. J. Org. Chem. 1976, 41, 3593-3599.
(17) Hurley, T. J.; Carrell, H. L.; Gupta, R. K.; Schwartz, J.; Glusker, J.
P. Arch. Biochem. Biophys. 1979, 193, 478-486.
(18) O’Leary, M. H. Methods Enzymol. 1980, 64, 83-104.
(19) O’Leary, M. H.; Richards, D. T.; Hendrickson, D. W. J. Am. Chem.
Soc. 1970, 92, 4435-4440.
(14) This matter has been cogently discussed: Lobell, M.; Crout, D. H.
G. J. Am. Chem. Soc. 1996, 118, 1867-1873.
(15) Ullrich, J. Methods Enzymol. 1970, 18A, 109-115. Zehender, H.;
Trescher, D.; Ullrich, J. Eur. J. Biochem. 1987, 167, 149-154.