Kinetics of the hydride reduction of an NAD+ analogue by isopropyl alcohol in aqueous and acetonitrile solutions: Solvent effects, deuterium isotope effects, and mechanism

Article abstract of DOI:10.1021/jo9007628

(Chemical Equation Presented) The rate constants of the hydride-transfer reactions from isopropyl alcohol (i-PrOH) to an NAD⁺ model, 9-phenylxanthyliumion (PhXn⁺), in acetonitrile (AN) and inwater containing AN (80% H₂O/20% AN) were determined over a temperature range from 36 to 67°C. The reactions follow second-order rate laws. In the latter solution, formation of the water adduct of PhXn⁺ was observed as a side-equilibrium (K). The observed inverse solvent kinetic isotope effect (k_H2O^obs/k_D2O^obs = 0.54), the larger than unity equilibrium isotope effect (K(H₂O)/K(D₂O) = 2.69), and the results of acid effect on the observed rate constants of the reactions are consistent with the "side-equilibriummechanism". Kinetic isotope effects at all three H/D positions of i-PrOH for the net hydride-transfer process were determined in both solutions at 60°C: KIE _α-D^H = 3.2(AN), 3.2(H₂O); KIE _β-D6^H = 1.05(AN), 1.16(H₂O); KIE _OD^H = 1.08(AN), 1.04(H₂O). These KIE values are consistent with the presence of the positively charged alcohol moiety in the transition state (TS) for cleavage of the α-C-H bond, the delocalization of the positive charge over the α-C-OH group, and the stepwise hydride and proton transfer processes. Comparison of the activation parameters for the reactions in the two solvent systems as well as those in the i-PrOH/AN (1:1 v/v) reported earlier suggests that the AN medium promotes the reaction by activating the ground-state alcohol reactant through weak interactions with the electron pairs on alcohol O, while water and parent alcohol media facilitate the reaction by H-bonding stabilization of the alcohol moiety of the TS. Results suggest that in the alcohol dehydrogenases without a Zn(II) cofactor in the active sites alcohols would be oxidized via hydride transfer to NAD⁺ coenzyme followed by the rapid deprotonation to the nearby basic species in the active site of the enzymes.

Full text of DOI:10.1021/jo9007628

pubs.acs.org/joc
þ
Kinetics of the Hydride Reduction of an NAD Analogue by Isopropyl
Alcohol in Aqueous and Acetonitrile Solutions: Solvent Effects, Deuterium
Isotope Effects, and Mechanism
Yun Lu,* Fengrui Qu, Yu Zhao, Ashia M. J. Small, Joshua Bradshaw, and Brian Moore
Department of Chemistry, Southern Illinois University Edwardsville, Edwardsville, Illinois 62026
yulu@siue.edu
Received April 12, 2009
þ
The rate constants of the hydride-transfer reactions from isopropyl alcohol (i-PrOH) to an NAD
þ
model, 9-phenylxanthylium ion (PhXn ), in acetonitrile (AN) and in water containing AN (80% H O/
2
2
order rate laws. In the latter solution, formation of the water adduct of PhXn was observed as a side-
0% AN) were determined over a temperature range from 36 to 67 °C. The reactions follow second-
þ
obs
obs
equilibrium (K). The observed inverse solvent kinetic isotope effect (k
/k
D O
2
=0.54), the larger
than unity equilibrium isotope effect (K(H O)/K(D O)=2.69), and the results of acid effect on the
H O
2
2
2
observed rate constants of the reactions are consistent with the “side-equilibrium mechanism”. Kinetic
isotope effects at all three H/D positions of i-PrOH for the net hydride-transfer process were deter-
H
R-D
H
β-D6 2
mined in both solutions at 60 °C: KIE
=3.2(AN), 3.2(H O); KIE
2
=1.05(AN), 1.16(H O);
H
KIE
= 1.08(AN), 1.04(H O). These KIE values are consistent with the presence of the positively
2
OD
charged alcohol moiety in the transition state (TS) for cleavage of the R-C-H bond, the delocalization
of the positive charge over the R-C-OH group, and the stepwise hydride and proton transfer processes.
Comparison of the activation parameters for the reactions in the two solvent systems as well as those in
the i-PrOH/AN (1:1 v/v) reported earlier suggests that the AN medium promotes the reaction by
activating the ground-state alcohol reactant through weak interactions with the electron pairs on
alcohol O, while water and parent alcohol media facilitate the reaction by H-bonding stabilization of
the alcohol moiety of the TS. Results suggest that in the alcohol dehydrogenases without a Zn(II)
þ
cofactorin the activesites alcohols would be oxidizedvia hydridetransfer to NAD coenzyme followed
by the rapid deprotonation to the nearby basic species in the active site of the enzymes.
1-4
Introduction
of the effects of the relevant factors on the kinetics.
Conventional kinetic investigations have, however, often
been greatly challenged since the kinetic data determined
usually do not completely correspond with the inherent
Alcohols in living organisms are metabolized to carbonyl
þ
compounds via oxidation by NAD coenzyme mediated
with alcohol dehydrogenases. The chemistry of the biologi-
hydride transfer processes due to the complexity of the
5-7
cal reactions involves hydride transfer from the R-C of the
alcohol to the 4-C of the NAD pyridinium ring. Much of
enzymatic reactions.
Although various approaches have
þ
been developed in order to kinetically unmask the chemistry
of the enzymatic reactions, it has been a great challenge to
the mechanistic understanding of the transformations has
come from determinations of the reaction kinetics and study
(3) Cook, P. F.; Cleland, W. W. Biochemistry 1981, 20, 1797–1805.
(
1) Klinman, J. P. In Enzyme Mechanisms from Isotope Effects; Cook, P.
F., Ed.; CRC Press, Boca Raton, 1991; pp 127-148.
2) Plapp, B. V. In Isotope Effects in Chemistry and Biology; Kohen, A.,
Limbach, H.-H., Eds.; CRC Press: Boca Raton, 2005; pp 811-835.
(4) Sekhar, V. C.; Plapp, B. V. Biochemistry 1990, 29, 4289–4295.
(5) Klinman, J. P. Biochemistry 1976, 15, 2018–2026.
(6) Cleland, W. W. Acc. Chem. Res. 1975, 8, 145–151.
(7) Dalziel, K. J. Biol. Chem. 1963, 238, 2850–2858.
(
DOI: 10.1021/jo9007628
r 2009 American Chemical Society
Published on Web 08/11/2009
J. Org. Chem. 2009, 74, 6503–6510 6503

Lu et al.
JOCArticle
correctly extract the kinetics of these hydride-transfer steps
for the derivation of the mechanistic information including
the structure of the important transition state (TS).
mechanistic information obtained. Studies of the effects of
isotopic substitution at the three H/D positions of i-PrOH, of
the acid concentration, of the alcohol concentration, and of
the temperature on the kinetics of the reactions were carried
out. The results were compared with those earlier obtained
for the same reaction in mixed solvents (i-PrOH/AN) as well
as a β-D KIE value for an alcohol dehydrogenase reaction
found in literature. Reaction mechanisms including the
structures of the TSs were proposed.
8
-10
Studies of the kinetics of relatively simple non-enzymatic
model reactions have been applied in order to provide useful
insights into the mechanism of the biological enzymatic reac-
tions. The corresponding investigations for the alcohol dehy-
11-15
drogenases have, however, not generally been successful
until very recently when we reported the thorough kinetic
studies of hydride-transfer reactions from isopropyl alcohol
þ
(
(
i-PrOH) to the two NAD analogues 9-phenylxanthylium ion
þ
Results
þ
PhXn ) and 10-methylacridinium ion (MA ), in the parent
þ
Hydration Equilibrium of PhXn and the Equilibrium Isotope
þ
alcohol medium containing various amounts of acetonitrile
The derived rate
16,17
AN) or a small amount of water (eq 1).
Effect (EIE). In 80% H O/20% AN, PhXn (counterion,
2
(
-
H
BF ) exists in equilibrium with its water adduct PhXnOH
4
constants for the hydride transfers (k ) together with the
observed deuterium kinetic isotope effects (KIE) at all three
H/D positions in i-PrOH were used to elucidate the hydride
transfer TS structure and the H-bonding stabilization effect on
þ
þ
(eq 2). The equilibrium concentration of PhXn ([PhXn ] )
increases with increasing [H ] (= [HBF ]). Since the ob-
eq
þ
4
obs
served rate constant (k ) of the hydride-transfer reaction
þ
17
between i-PrOH and PhXn in aqueous solution depends on
þ
the TS. These TS structural features bear some similarities to
those of the hydride-transfer processes of the alcohol dehy-
drogenase reactions.
[PhXn ] , it is necessary to determine the equilibrium con-
stant (K) in order to derive kinetic information on the net
eq
H
hydride-transfer step (k ) from the observed kinetic results in
obs
H
this reaction medium. The relationships between k , k , and
Kwill be discussed in the following sections (see eq 4). The value
of K in 80% H O/20% AN at 60 °C, defined as K(H O) =
2
þ
2
þ
eq
[
PhXnOH] [H ] /[PhXn ] , was determined to be equal to
eq
eq
0
.214 ( 0.035 M. Note that the K value might be comparable
þ
to the KR value for the PhXn that was determined in pure
acidic aqueous solutions at 25 °C. The K
M have been reported by Arnett and Feldman, respectively.
In order to understand the kinetic solvent isotope effect
þ
þ of 0.148 and 0.079
R
19
18
obs
obs
(
k
(H O)/k (D O)), K in 80% D O/20% AN at the same
2 2 2
temperature was also determined (K(D O) = 0.0797 ( 0.0134
2
M). The equilibrium isotope effect (EIE = K(H O)/K(D O))
was observed to be equal to 2.69. The larger than unity EIE
value indicates that under the same conditions [PhXn ] is
eq
2
2
Understanding the factors affecting the rates of the
non-enzymatic reactions and the stabilities of the corre-
sponding TSs is expected to illuminate the nature of the
enzymatic catalysis. In this paper, we will explore how
changes in environments/reaction media affect the kinetics
þ
higher in D O medium than in H O medium.
2
2
þ
of the hydride-transfer reaction from i-PrOH to PhXn and
the nature of its TS. We have reported the kinetic study of
the reaction in the parent alcohol medium containing AN
1
1:1 v/v); herein we will report the corresponding study in
7
(
aqueous solution (80% H O/20% AN (v/v)) and in aprotic
2
AN. Since i-PrOH concentrations used in this work were
low, complications due to the aggregation of the alcohols
that most likely exists in the i-PrOH/AN solvents are ex-
pected to be diminished or eliminated. The latter effect can
significantly affect the kinetics of the reactions and hence the
The observed EIE value is consistent with the known
deuterium fractionation factor (j) for deuterium between
þ
20,21
H
O and H
O sites.
In aqueous system, j is defined to
be the D/H ratio in the OL bond (L = H and D) divided by
2
3
þ
the D/H ratio in the OL bond, i.e., j = (D/H)OL
+
/(D/H)OL.
A jvalue of 0.69 has been determined for the hydronium ion in
20,21
the aqueous system.
This value means that, in D O solvent
2
(
(
8) Northrop, D. B. Biochemistry 1975, 14, 2644–2651.
9) Hermes, J. D.; Roeske, C. A.; O’Leary, M. H.; Cleland, W. W.
at room temperature, D has only 69% preference as compared
þ
Biochemistry 1982, 21, 5106–5114.
(
(
to H in H O solvent for placement in OL bond. Thus,
2
10) Hammes, G. H.; Schimmel, P. R. Enzymes 1970, 2, 67–114.
11) Shirra, A.; Suckling, C.-J. J. Chem. Soc., Perkin Trans. II 1977, 759–
þ
considering equilibrium 2 as PhXn þ 2L O = PhXnOL þ
2
þ
7
65.
L O and assuming H and D have the same opportunity for
3
(
(
12) Ohnishi, Y.; Kitami, M. Tetrahedron Lett. 1978, 4035–4036.
13) Kanomata, N; Suzuki, M.; Yoshida, M.; Nakata, T. Angew. Chem.,
placement in L O and PhXnOL, the EIE of the equilibrium
2
Int. Ed. 1998, 37, 1410–1411.
14) Kimura, E.; Shionoya, M.; Hoshino, A.; Ikeda, T.; Yamada, Y. J.
Am. Chem. Soc. 1992, 114, 10134–10137.
(
(18) Arnett, E. M.; Flowers, R. A.; Lundwig, R. T.; Meekhof, A. E.;
Ealek, S. A. J. Phys. Org. Chem. 1997, 10, 499–513.
(
15) Walz, R.; Vahrenkamp, H. Inorg. Chim. Acta 2001, 314, 58–62.
16) Lu, Y.; Endicott, D.; Kuester, W. Tetrahedron Lett. 2007, 48, 6356–
(19) Feldman, M. R.; Thame, N. G. J. Org. Chem. 1979, 44, 1863–1865.
(20) Alvarez, R.; Schowen, R. T. In Isotopes in Organic Chemistry;
Elsevier: New York, 1987; Vol 7, Chapter 1.
(
358.
6
(17) Lu, Y.; Qu, F.; Moore, B.; Endicott, D.; Kuester, W. J. Org. Chem.
008, 73, 4763–4770.
(21) Arrowsmith, C. H.; Guo, H.-X.; Kresge, A. J. J. Am. Chem. Soc.
1994, 116, 8890–8894.
2
6
504 J. Org. Chem. Vol. 74, No. 17, 2009

Lu et al.
JOCArticle
3
can be estimated to be 1/(0.69) = 3.04, which is not far from
the value of 2.69 determined from this work. Note that our
value was determined under reaction conditions (in 80% L O/
2
2
0
0% AN, 60 °C) different from those used to obtain the
.69 value (in pure L O, 25 °C), which probably contributes
2
to the difference between the two EIE values (2.69 vs 3.04).
1
Kinetics. When the hydride donor i-PrOH present in low
concentrations (e0.327 M in this work) was added into the
aqueous equilibrium system of PhXn and PhXnOH during
7
þ
kinetic experiments, the equilibrium formation of the alcohol
1
7
adduct (PhXnOPr-i) would not be expected to compete with
the existing hydration equilibrium 2. Moreover, even if the
PhXnOPr-i were formed, the acetal-like structure would be
readily hydrolyzed to PhXnOH that was further converted
back to its cationic precursor in acidic solution. While this is
most likely the case for the reaction in aqueous solution, the
equilibrium formation of the alcohol adduct in pure AN
solution might be expected since i-PrOH is the only nucleo-
phile present in the reaction solution. However, a UV-vis
spectral experiment showed that under the kinetic conditions
where [i-PrOH] e 0.327M, the formation of the PhXnOPr-i
was not observed. This differs markedly from our earlier
observations of the significant side-equilibrium formation of
FIGURE 1. Kinetic scans determined by analysis of the reaction
þ
aliquots, taken from the reaction mixtures of [PhXn ]
0
= 0.001 M
and [i-PrOH] = 0.0816 M in AN at 60 °C. The reaction aliquots
were diluted by 25 times with AN. Reaction times for spectra 1-5
are 30, 150, 359, 600, and 1295 min, respectively.
1
7
the PhXnOPr-i in the same hydride-transfer reaction in the
parent i-PrOH containing AN (1:1 v/v). Therefore, for the
hydride-transfer reaction in aqueous solution, the reaction
most likely follows mechanism 3 that involves the formation
of the hydration product (PhXnOH) in a rapid side-equilib-
rium with the carbocation. While in AN, the equilibrium
concentration of PhXnOPr-i is insignificant, which results in
-
5
(
maximum concentration of 4 ꢀ 10 M) in the diluted reac-
þ
tion aliquots to PhXn . Under these conditions, the rate of
formation of PhXnH is equal to the rate of decrease of
[PhXn ] . The rate law of the reaction in aqueous solution
derived from mechanism 3 is shown in eq 4.
þ
T
þ
obs
H
rate ¼ d½PhXnHꢁ=dt ¼ -d½PhXn ꢁ =dt
T
the simplification that k is equal to k .
H
þ
kH
s
PhXnH
¼ k ½PhXn ꢁ ½i-PrOHꢁ
-
H2O
'
eq
PhXnOHþHþ
PhXn þi-PrOH
þ
&
þH2O
H
þ
þ
þ
¼
¼
k ½PhXn ꢁ ½i-PrOHꢁ½H ꢁ=ð½H ꢁþKÞ
T
þacetoneþHþ
Determinations of the pseudo-first-order rate constants
pfo
) of the hydride-transfer reactions were carried out at
0 °C in 80% H O/20% AN with [PhXn] = 0.15 mM and
i-PrOH] e 0.327M and in AN with [PhXn] = 1 mM and
[i-PrOH] e 0.327M. In the latter solvent, reaction aliquots
taken at different times were diluted with AN, and the resultant
solutions were analyzed by UV-vis spectrophotometry and
kinetic scans were recorded (Figure 1). The k were computed
ð3Þ
obs
þ
k
½PhXn ꢁ ½i-PrOHꢁ
pfo
ð4Þ
T
þ
(
6
[
k
Effect of [H ] on k . The kinetics of the hydride-transfer
reaction were not measurable in neutral aqueous solution
2
0
þ
þ
0
0
due to low [PhXn ] . Sufficient acid was thus added to
eq
ensure that PhXn was present in high enough concentra-
tions for a measurable reaction to occur (see eq 5). Care had
to be taken so that [H ] was not so high that the protonation
of i-PrOH became significant, making the quantification of
0
þ
pfo
from the absorbance versus time data for the consumption of
þ
i-PrOH difficult. To this point, a pK value (-3.2) of the
a
-1
-1
þ
PhXn at 373 nm (ε = 29,800 M cm (AN)) and rate =
isopropyloxonium ion (i-PrOH ) in water estimated by
þ
2
þ
þ
22
Bartlett gave us some guides for selecting the [H ] range
d[PhXnH]/dt = -d[PhXn ]/dt. Since PhXn partitions be-
tween the hydride-transfer product and its water adduct (eq 3)
in aqueous solution, the rate definition becomes the following:
obs
for study. According to rate law 4, k should increase with
þ
þ
increasing [H ] and reach saturation at [H ] . K when
þ
þ
þ
T
rate = d[PhXnH]/dt = -d{[PhXn ] þ [PhXnOH] }/dt.
[PhXn ] = [PhXn ] . Under the kinetic saturation condi-
eq
obs
tions, k should be equated to k .
eq
eq
H
Therefore, simply following the decay of the UV absorbance
of [PhXn ] is inappropriate to evaluate the rate constant for
þ
pfo
obs
eq
The k (= k [i-PrOH]) was determined for the reaction
with [i-PrOH] = 0.327 M in the presence of HBF with
concentrations ranging from 0.030 to 1.90 M that were not
the formation of PhXnH. Also since both PhXnH and
PhXnOH absorb over the entire wavelength range, it is not
0
4
pfo
possible to determine the k under these circumstances. Thus,
þ
expected to favor the equilibrium formation of i-PrOH2
þ
pfo
(Figure 2). The results show that k increases with increas-
it is necessary to take into account the expression [PhXn ] =
T
þ
þ
þ
[
PhXn ] þ [PhXnOH] in order to evaluate the rate of
ing [H ], tends to saturate when [H ] is increased to about
0.5-1.0 M, but unexpectedly decreases afterward. This
kinetic behavior seems inconsistent with the rate law 4 or
eq
eq
formation of PhXnH. This was accomplished by diluting
aliquots of the reaction mixtures with AN/water (3:1, v/v)
containing 3 M HClO to provide samples for the spectro-
4
þ
photometric determination of [PhXn ] . The HClO concen-
T
4
(22) Barlett, P. D.; McCollum, J. D. J. Am. Chem. Soc. 1956, 78, 1441–
1450.
tration is high enough to convert all of the PhXnOH
J. Org. Chem. Vol. 74, No. 17, 2009 6505

Lu et al.
JOCArticle
pfo
FIGURE 2. Observed pseudo-first-order rate constants (k ) of the
þ
hydride-transfer reaction as a function of [H ] in 80% H O/20% AN
2
þ
þ
at 60 °C ([PhXn ]
0
= 0.00015 M, [i-PrOH]
0
= 0.327 M). [H ] are
0
The corresponding k
.03, 0.06, 0.12, 0.24, 0.48, 0.96, 1.24, 1.45, and 1.90 M, respectively.
pfo
are 0.000536, 0.00141, 0.00230, 0.00332,
1
-
0
.00510, 0.00524, 0.00470, 0.00409, and 0.00292 min
.
pfo
derived eq 5, which relates k to [H ].
þ
pfo
H
þ
þ
k
¼ k ½i-PrOHꢁ½H ꢁ=ð½H ꢁþKÞ
ð5Þ
One may assume that the decomposition of AN by hydro-
lysis in acid to acetamide and further to acetic acid might be
þ
responsible for the decay of the reaction rate at higher [H ]
through changing the solvent composition. It has been
shown, however, there was a very slow decomposition of
AN in 50% water/50% AN containing high [HClO ] (6 M) at
4
2
3
room temperature; AN decayed by about 4% in 2 weeks.
Thus, AN decomposes very slowly under our reaction con-
ditions, and this would not significantly affect the kinetic
determinations of our reactions.
pfo
FIGURE 3. Observed pseudo-first-order rate constants (k ) of
the hydride-transfer reaction as a function of [H ] (A) and the
þ
We consider the possibility that the observed dramatic rate
þ
pfo
corresponding double reciprocal plot (1/k ≈ 1/[H ]) (B), in 80%
þ
þ
reduction at high [H ] may result from changes in ionic
strength as the [HBF ] increases. The latter brings about an
H
2
O/20% AN containing NaClO
4
at 60 °C ([PhXn ]
0
= 0.00015M,
þ
þ
4
[i-PrOH] = 0.327 M, and [H ] þ [NaClO ] = 1.24 M). [H ] used
0
4
increase in the polarity of the reaction medium, thereby
affecting the rates of the reactions. This increase in polarity is
expected to lower the rate of the reaction by destabilizing the
hydride-transfer TS in which the overall 1þ charge on the
are 0.03, 0.06, 0.12, 0.24, 0.48, 0.96, and 1.24 M, respectively. The
corresponding k are 0.00233, 0.00325, 0.00395, 0.00468, 0.00470,
pfo
-
1
.00464, and 0.00470 min .
0
and 0.0339 M under the constant ionic strength conditions.
Note that this latter K value is very different from the one
þ
reactant (PhXn ) is dispersed between the two reactant moi-
þ
eties. To test this inference, we determined the effect of [H ] on
þ
determined under low [H ] conditions in which about 50% of
pfo
the observed value of k of the reaction under the conditions
the cation was hydrated in the equilibrium solution (see the first
paragraph of the Results section, K=0.214 M). This suggests
that the polar solvent with high salt concentration (or high
ionic strength) facilitates the presence of the cation in the
hydration equilibrium 2. Therefore, our observed bell-shaped
of constant ionic strength accomplished by adding an inert
electrolyte (NaClO ) to the acidic solutions, i.e., varying the
4
þ
[H ] but keeping ([HBF ] þ [NaClO ]) constant (1.24 M). As
predicted, the kinetic saturation effect with depressed k at
H ] ranging from 0.48 to 1.24 M was observed (Figure 3A), in
accord with the “side-equilibrium” mechanism (eq 3). Results
4
4
pfo
þ
[
pfo
þ
relationship between k and [H ] in Figure 2 resulted from
þ
the ionic strength changes of the kinetic solutions as [H ]
þ
pfo
were then fit to a double reciprocal relationship (1/k
þ
≈
changed, i.e., the increase in ionic strength as increasing [H ]
2
/[H ]) obtained by transforming eq 5 (Figure 3B, R =
1
lowers the rate of the hydride-transfer step and shifts the
hydration equilibrium to the cation side.
Inverse Kinetic Solvent Isotope Effect in Aqueous Solution.
0.990). From the obtained intercept and the slope of the latter
linear plot, k and K were evaluated to be 0.0153 M min
H
-1
-1
pfo
A comparison of the k for the reactions in H O and D O
2
2
(
23) Sze, Y.-K.; Irish, D. E. Can. J. Chem. 1975, 53, 427–435.
systems with the same alcohol concentrations (0.327 M) at low
6
506 J. Org. Chem. Vol. 74, No. 17, 2009

Lu et al.
JOCArticle
TABLE 2. Kinetic Isotope Effect of the Hydride-Transfer Step
pfo
TABLE 1. Pseudo-First-Order Rate Constants (k ) as a Function of
i-PrOH] in 80% H O/20% AN and in AN
a,b
[
2
H
H
H
reaction solvent
KIER-D
KIEβ-D6
KIEOD
o
in 80% water/20% AN at 52 C
a,b
o
in AN at 60 C
b,c
c
d
e
2
0% AN/80% H
AN
i-PrOH/AN (1:1 v/v)
2
O
3.2 ( 0.3 (2) 1.16 ( 0.05 (4) 1.04 ( 0.02 (4)
3.2 ( 0.1 (4) 1.05 ( 0.03 (6) 1.08 ( 0.01 (4)
f
pfo
3
pfo
3
[
i-PrOH]
M)
k
ꢀ10
[i-PrOH]
(M)
k
ꢀ10
f,g
-
1
-1
4.4
1.07
1.11
(
(min
)
(min )
a
At 60 °C, unless otherwise indicated; see texts for definitions of the
KIEs. Numbers in parentheses are times of determinations. [PhXn ]
0
0
0
0
.654
.327
.164
.0820
þ
5.49
2.75
1.37
0.673
þ
0.327
0.164
0.0820
0.0410
5.67
2.70
1.34
0.535
b
c
þ
0
=
þ
d
þ
0
.15 mM, [H ] = 0.06 M, unless otherwise indicated. At 54 °C, [H ] =
e
.96 M. Measurement error of the K was not considered; see text.
0
f
þ
g
PhXn ] = 1 mM. Reference 17.
[
0
a
b
] = 0.96 M. Based on three
[
PhXn ]
0
= 0.15 mM, [H ] = [HBF
4
c
þ
determinations. [PhXn ] = 1 mM.
0
TABLE 3. Second-Order Rate Constants and the Corresponding Acti-
vation Parameters for the Net Hydride-Transfer Reaction in Aqueous and
AN Solutions
þ
a
[
KIED2O (= k (H O)/k (D O)), which was evaluated to
H ] (0.06 M) gives the observed kinetic solvent isotope effect,
pfo
pfo
pfo
2
2
in 80% H
2
O/20% AN (v/v)
in AN
be equal to 0.53, i.e. an inverse solvent KIE was observed. Note
pfo
obs
that KIE
is also equivalent to KIE
D O
obs
by the second-order-rate constant (k ).
, which is defined
obs
3b
c
H
k
3b
H
3b
D O
temp (°C)
k
ꢀ 10
K (M)
ꢀ 10
temp (°C)
k
ꢀ 10
2
2
3
4
54
60
6
3
1.92
3.73
8.65
15.9
25.4
0.0895
0.0810
0.0746
0.0425
0.0450
2.10
4.04
9.32
16.6
26.6
36
42
54
60
67
3.48
5.47
9.88
16.5
24.8
The effect of acid concentration on the rate of the reaction
in AN was also investigated. No apparent effect was ob-
served, indicating that the formation of the alcohol adduct
PhXnOPr-i) is insignificant in AN.
pfo
Effect of [i-PrOH] on k . The effects of [i-PrOH] on the
6
7
H
(
H
Ea = 71.6 ( 1.6 kJ/mol
H
Ea = 58.3 ( 2.4 kJ/mol
qH
qH
4
S
a
= 68.9 ( 1.6 kJ/mol
4H = 55.6 ( 2.4 kJ/mol
qH
rates of the reactions in pure AN and in aqueous solution
þ
qH
4
= -107.3 ( 4.4 J/mol K
Containing [H ] = 0.96 M. M min . Equilibrium constant of
4S = -146.1 ( 7.9 J/mol K
c
with [H ] = 0.96 M were determined (see Table 1). The
results show that the reactions are first-order in the hydride
þ
b
-1
-1
þ
the equilibrium 2 in the aqueous solution containing [H ] = 0.96 M.
donor (i-PrOH) in both solvents. This is consistent with the
obs
proposed mechanism 3. Second-order-rate constants (k
)
obtained from the reactions in mixed solvent (i-PrOH/AN=
pfo
17
can therefore be calculated by dividing the k by [i-PrOH].
Kinetic Isotope Effects at the Three H/D positions in
i-PrOH. The effects of deuteration at the R- and β-C in
i-PrOH on the kinetics of the reactions were determined by
1/1) are also listed in Table 2 for comparison. Note that,
similarly, the larger uncertainty may be contained in the
H
KIEOD estimated in this mixed solvent system, as it was
separated from the observed KIE of the overall hydride-transfer
reaction by excluding the EIE of the side-equilibrium formation
pfo
comparing the k determined in the presence of i-PrOH-2-d
and (CD ) CHOH with that for the reaction of the normal
H
of the PhXnOPr-i adduct, i.e., KIEOD = KIEOD
obs
17
3
2
EIE.
3
pfo
i-PrOH under the same alcohol concentration conditions.
Since deuterium substitution at either position has no effect
Kinetic Temperature Effects. The k for the reactions in
both solvents with certain alcohol concentrations was also
determined at various temperatures and was converted to the
pfo
on the hydration equilibrium, KIE is equal to the KIE of
obs
obs
H
the net hydride-transfer process in both cases and are
H
second-order-rate constants (k ). In AN, k is equal to k
obs
(Table 3). In aqueous solution, k were determined under
H
referred to as KIE
H
and KIEβ-D6 , respectively. Note
R-D
þ
that KIEβ-D6 is a cumulative effect of six β-C-D bonds.
These results are listed in Table 2. The KIE
the condition of [H ] = 0.96 M (Table 3). According to eq 5,
obs
a conversion from k
H
H
to k thus needs the K values
in both
R-D
solvent systems are primary (3.2 (AN, 60 °C) and 3.2
determined under the specific acid conditions, i.e., specific
ionic strength conditions. Table 3 thus also includes the
H
(H O, 54 °C)), while the KIE
at 60 °C are secondary
2
β-D6
H
K and derived k for the reactions in the aqueous solution
(
1.05 (AN)) and 1.16 (H O)).
The KIE at the OH/OD positions in i-PrOH, referred to
2
as a function of temperature. Note that the K=0.0425 M
þ
pfo
as KIEOD, was calculated from the k
of i-PrOH and i-PrOD with same concentrations. In AN,
KIEOD (= KIEOD ) is equal to KIE
In aqueous solution, k
be determined in D O solvent in order to avoid H/D exchange.
of the reactions
determined under [H ] = 0.96 M and 60 °C conditions
(Table 3) may be comparable to the slightly smaller K =
0.0339 M derived from the aforementioned double recipro-
pfo
obs
H
(1.08 in Table 2).
of the reaction of i-PrOD must
OD
pfo
pfo
þ
cal correlation between 1/k
obtained under the condition of larger “salt” concentra-
and 1/[H ], which was
2
obs
pfo
þ
Therefore, KIE
pfo
k
in aqueous solution is given by k (H O)/
2
OD
tions, i.e., ([H ] þ [NaClO ]) = 1.24 M. Activation para-
4
pfo
obs
(D O) = KIE
2
= KIE
. According to eq 4,
þ
D O
2
H
D O
2
meters including activation energies, activation enthalpies,
and activation entropies for the hydride-transfer steps were
evaluated using both the Arrhenius and the Eyring equa-
tions and are found in Table 3 as well. The activation
parameters obtained in i-PrOH/AN in our previous work
obs
þ
obs
KIEOD =KIEOD {([H ] þ K(D O))/([H ] þ K(H O))} =
2
2
obs
D O
2
KIE
, so the KIE
encompasses the components of
D O
2
H
. At
OD
both the equilibrium isotope effect (EIE) and the KIE
obs
þ
[
H ] = 0.06M, KIE
is equal to 0.53 (see previous sec-
D O
2
H
H
a
qH
tions), while the KIEOD under the same conditions was
evaluated to be equal to 1.04 (Table 2), characteristic of a
secondary KIE. Since the estimation of the KIEOD in aqueous
are E =71.1 ( 12.6 kJ/mol; 4H = 68.3 ( 12.6 J/mol;
qH
24
4
S
= -106.9 ( 38.0 J/mol K. Within experimental
3
obs
solution relies on the values of both K and KIE
, the KIE
(24) For a direct comparison, the activation entropy of the reaction
in i-PrOH/AN was calculated using the second-order rate constants
D O
2
value should contain the measurement errors of the two
quantities. The uncertainty of this KIE value in aqueous
solution would thus be higher than that in AN. Relevant KIEs
-
1 -1
(M
s
) converted from the pseudo-first-order rate constants reported in
our previous work by dividing the alcohol concentration (6.54 M); see Table
5 in ref 17.
J. Org. Chem. Vol. 74, No. 17, 2009 6507

Lu et al.
JOCArticle
error, these values are close to those obtained in aqueous
solution in this work.
SCHEME 1. An Unlikely Alternative Mechanism in Aqueous
Solution
-H2O
s
PhXnOHþHþ
h
PhXnOH2þþi-PrOH
PhXnHþacetoneþHþ
Discussion and Conclusions
and possibly cyclic TS structure that would be favored by the
solvent that does not provide such H-bonding stabilization
effect.
On the other hand, a comparison of the activation energies
of the reaction in aqueous solution (71.6 kJ/mol, Table 3)
and in AN (58.3 kJ/mol) reveals that the reaction is about
Solvent Effects on the Kinetics and Mechanism of the
Hydride-Transfer Reactions. The hydride-transfer reaction,
in aqueous solution or in mixed solvents (i-PrOH/AN),
involves the formation of the water adduct (PhXnOH) or
1
7
þ
the alcohol adduct (PhXnOPr-i) of the PhXn cation via
rapid equilibrium. While in pure AN solution, the formation
of PhXnOPr-i is insignificant. These conclusions follow from
the observation that acid does not affect the rate of the
hydride-transfer reaction in AN but does increase the rates
of the reactions in aqueous solution (Figures 2 and 3) and
1
3 kJ/mol less favored in the former than in the latter, despite
the above suggested H-bonding stabilization effect on the TS
of the reaction in the former solvent. This can be explained in
terms of the H-bonding to the reactant alcohol O in the
aqueous solution so that the ground-state reactant alcohol is
less reactive toward donating hydride ion than in AN.
1
7
in mixed solvents of i-PrOH/AN (see our previous work ).
The resulting hydride-transfer product, protonated acetone,
exists transiently and undergoes rapid proton transfer to the
basic species in the reaction solutions, forcing the reactions
to go to completion.
H
qH
The activation parameters (E = 71.1 kJ/mol; 4H
=
a
qH
6
8.3 J/mol; 4S = -106.9 J/mol K) that we obtained ear-
3
lier in i-PrOH/AN mixed solvents are close to the ones
observed in the aqueous solution within experimental error.
We suggest that the parent alcohol in i-PrOH/AN would
play a similar role as water does in deactivating the ground-
state alcohol and stabilizing the TS.
We have reported an inverse kinetic solvent isotope
effect (0.54) for this reaction in i-PrOH(D)/AN (1:1 v/v)
and a larger than unity equilibrium isotope effect (EIE =
K(i-PrOH)/K(i-PrOD)) (2.67) for the corresponding PhXn-
1
7
Thus, AN medium would promote the reaction by activat-
ing the alcohol reactant through weak interaction with the
electron pairs on alcohol O, whereas water and i-PrOH
would facilitate the reaction by stronger H-bonding stabili-
zation of the alcohol moiety of the TS. Interestingly, the
results derived from this study of the non-enzymatic reac-
tions are consistent with the traditionally accepted theories
for enzymatic rate enhancement, i.e., the “ground-state
destabilization theory” and the “TS stabilization theory”.
Furthermore, this H-bonding interaction in the TS is remi-
niscent to the substrate-enzyme interaction modes in alcohol
dehyhrogenase reactions. For example, in a human lactate
dehydrogenase reaction, H-bonding is known to be present
between the hydroxyl H of the lactate substrate and the
imidazole N of the His-193 during the reaction in the active
OPr-i formation equilibrium. These results were consistent
with the hydride-transfer mechanism involving the side-
equilibrium formation of PhXnOPr-i. The EIE value shows
þ
that [PhXn ] is higher in i-PrOD than in i-PrOH under
eq
the same reaction conditions, which is a consequence of the
fact that the reaction is faster in i-PrOD than in i-PrOH.
In this work, an inverse kinetic solvent isotope effect (0.53)
was observed for the hydride-transfer reaction in water
and a larger than unity EIE (2.69) for the hydration equilib-
þ
rium of PhXn was observed. Likewise, the latter EIE
þ
value indicates a higher [PhXn ] in D O solvent than
eq
2
in H O, giving rise to an inverse solvent KIE. These results
2
are consistent with mechanism 3 in which hydration is a
side-equilibrium inhibiting the net hydride-transfer step. Our
results also exclude an unlikely alternative mecha-
nism in which the PhXnOH adduct is an intermediate that
2
5
site of the enzyme.
KIE and the TS Structure. Kinetic isotope effects on the
hydride-transfer step at the three H/D positions of i-PrOH
can provide information about the structure of the TS. As
expected, the KIE brought about by deuterium substitution
undergoes an acid-catalyzed S 2-like substitution of OH
N
with hydride acting as a nucleophile (Scheme 1). This mecha-
nism is highly unlikely but is included for the sake of
completeness.
at the R-C in all three reaction media are primary (Table 2,
We have suggested in our previous work that the TS of the
hydride-transfer step in i-PrOH/AN (1:1 v/v) is the one in
which a positive charge develops on the alcohol moiety
during the reaction and is stabilized by H-bonding between
H
KIE
), indicating that the hydride-transfer step is rate-
determining. The secondary KIEs, resulting from the six
R-D
H
β-C-H/C-D bonds (KIE
) and the O-H/O-D bonds
β-D6
H
δþ
(KIEOD ) (Table 2), are consistent with a partially positively
charged alcohol moiety in the TS. The former is a result
of the orbital overlap between the β-C-H σ orbital and the
the solvent O and the positively charged H of its R-CdOH
δþ
(
T R-C -OH) moiety. This H-bonding interaction can be
further supported by the comparison of the activation
entropies of the reaction in aqueous solution with that in
δþ
p orbital on the carbon of the R-C -OH in the TS, affecting
the vibration of the β-C-H bonds. The latter results from the
effect of the developing positive charge on the vibration of
pure AN (-107.3 (aqueous) vs -146.1 J/mol K (AN),
3
Table 3). The observation that the activation entropy is
3
δþ
the O-H bond in the R-CdOH moiety of the TS. Since the
9 J/mol K less negative for the reaction in aqueous solution
3
β-C-H and the O-H bonds are more easily weaken and
therefore more readily allow dispersion of the positive
than in AN suggests that the TS in the former solvent is much
less ordered than in the latter solvent. This is consistent with
the fact that water solvent loosens the associated TS struc-
ture by H-bonding interaction between water and the posi-
tively charged alcohol moiety in the TS. On the other hand,
the observed large activation entropy in AN suggests that the
hydride-transfer process would accompany a highly ordered
δþ
charge on the β-CH -C -OH moiety as compared to
3
the corresponding C-D/O-D bonds, normal secondary
(25) Gulotta, M.; Deng, H.; Deng, H.; Dyer, R. B.; Calender, R. H.
Biochemistry 2002, 41, 3353–3363.
6
508 J. Org. Chem. Vol. 74, No. 17, 2009

Lu et al.
JOCArticle
KIEs resulted. Note that the observed primary R-C-D KIE
and secondary O-D KIE indicate that the deprotonation
from the O-H bond during the overall alcohol oxidation to
the carbonyl compound is not concerted with but after the
(50% i-PrOH/50% AN) that we reported earlier and the β-D
KIE for a lactate dehydrogenase reaction found in literature.
Protic solvents, water and alcohol, stabilize the hydride-
transfer TS by H-bonding interaction between their hydro-
xyl O and the positively charged hydroxyl H of the alcohol
moiety in the TS but retards the reaction by tying up the lone
pairs of electrons on the OH of the ground-state reactant
alcohols by solvation/H-bonding. Both trends were reversed
in pure AN due to the lack of these H-bonding effects.
Positive charge developing in the TS is delocalized over the
alcohol R-C-OH group. Our results are important for the
understanding of the TS of the alcohol dehydrogenase
reactions, especially for those enzymes that do not contain
a Zn(II) cofactor in the active sites, such as lactate dehydro-
genases. In the alcohol dehydrogenases with a Zn(II) cofac-
tor, the hydride source is most likely a zinc alkoxide. Our
study suggests that in the former dehydrogenases, the alco-
hol group would be oxidized via hydride transfer from the
rate-limiting hydride-transfer step.
H
KIE
in i-PrOH/AN (4.4) is larger than those observed
R-D
in the aqueous and AN solutions (3.2). The difference is an
indication that the transferring H is more equally bound to
reactive C sites of both reactants in the former than in the
latter two solutions. Similar magnitudes of β-D KIEs in
i-PrOH/AN (1.07) and in AN (1.05) suggest that the electron
density of the R-C of the alcohol moiety of the TSs in both
solutions would be the same. The slightly larger one observed
in acidic aqueous solution (1.16) may be explained in terms
of the H-bonding stabilization of the lone pairs of electrons
on the hydroxyl O so that the developing positive charge in
the TS is more localized at the R-C. Unlike the β-C-H(D)
bond in the TS whose vibrations were not expected to be
affected by solvation, the alcohol hydroxyl O-H(D) vibra-
tions would be influenced through H(D)-bonding to the
solvent electronegative atom. The O-D KIE magnitude
would thus be affected by both the positive charge density
on O and the strength of the solvation effect on the O-H(D)
group. Since how solvation affects the O-H(D) vibra-
tions is not known, a quantitative comparison between the
O-D KIEs in three solvent systems may thus be meaningless
þ
R-C to the NAD coenzyme followed by the rapid deproto-
nation from the hydroxyl O to the nearby basic species in the
active sites of the enzymes, rather than via a concerted
hydride-proton transfer process.
Experimental Section
General Procedures. 9-Phenylxanthylium tetrafluoroborate
þ
-
(
PhXn BF
4
) was synthesized according to the published pro-
((CD CHOH) was prepared by the
with NaBH according to a procedure
(1.04 (aqueous), 1.08 (AN), 1.11 (i-PrOH/AN), Table 2).
29
cedure. 2-Propanol-d
reduction of acetone-d
6
3 2
)
Comparison to the β-D Isotope Effects in the Non-enzymatic
6
0
4
3
and Enzymatic Reactions in Literature. While the O-D KIE
has never been determined for either non-enzymatic or enzy-
matic hydride-transfer reactions involving alcohols in litera-
ture, a β-D KIE value of 1.19 was reported by Cleland and
described in literature, and the D content was determined to
be 99.4% per C-D bond. Commercially available i-PrOD,
i-PrOH-2-d, and isopropyl alcohol were all purified by distilla-
tion over dry K CO . Acetonitrile was distilled twice.
2
3
Determination of the Equilibrium Constant (K). The equilib-
rium constant (K) of the equilibrium between PhXn and
co-workers for the oxidation of lactate-β-d mediated by
3
þ
2
6
lactate dehydrogenase. To the knowledge of the authors,
this is the only β-D KIE reported by other research groups
for such hydride-transfer reactions. This value (1.06/β-D) is
much larger than the corresponding ones observed in our
non-enzymatic reactions (1.006/β-D in AN, 1.008/β-D in
i-PrOH/AN, 1.02/β-D in aqueous solution, Table 2), suggest-
ing that more positive charge is developed at the alcohol
moiety of the TS in the enzymatic reaction than in non-
enzymatic reactions. In addition to this experimental deter-
mination of the β-DKIE, theoretical calculations werecarried
out to evaluate the hyperconjugation stabilization effect by
PhXnOH in aqueous solution (80% water/20% AN or with
8
0% D O) was determined spectrophotometrically (UV-vis)
2
þ
by recording the absorbance at 373 nm attributable to PhXn .
The equilibrium was established by adding a certain amount of
the PhXnOH stock solution in AN to the 80% water/20% AN
þ
solution of various concentrations of H (= [HBF ]) at 60 °C
4
-
5
þ
([PhXnOH] = 2 ꢀ 10 M). The [H ] range (about 0.04-0.16
M) was chosen so that about 30-60% of the PhXnOH were
converted to the cation. Determinations based on three acid
0
concentrations were performed, and the average K(H O)
2
and K(D O) together with their standard deviations were re-
2
ported. For the Ks determined under high acid concentration
([H ] = 0.96 M) and different temperature conditions, only the
particular acid concentration was used. In the latter case,
β-C-H(D) bonds on the adjacent hydroxy-carbocations
þ
þ
(
C -OH). For example, a calculated isotope effect value of
1
.32 per three deuteriums on protonation of acetaldehyde was
2
þ
[PhXn ]
0
= 0.00015 M, which was same as the initial concen-
7
þ
reported by Hess et al., whereas a much smaller value of 1.19
per six deuteriums on protonation of acetone was calculated
by Alston et al.
In conclusion, the rate constants of the hydride-transfer
reactions from i-PrOH to PhXn as well as the correspond-
ing deuterium KIEs at the three H/D positions of i-PrOH
tration of the PhXn in the kinetic solutions, was used, and the
absorbance at 477 nm, which is also attributable to the cation,
was recorded for the calculation of the [PhXn ]eq and the
2
8
þ
K values.
1
7
þ
General Kinetic Determination Procedure. An80μLor12μL
portion of a 0.1 M stock solution of PhXn in AN was added to
þ
8
1
mL AN or aqueous solution containing i-PrOH in a well sealed
0 mL reaction vial, which was thermostatted in a water bath at
and the activation parameters in pure AN and in aqueous
solution (80% H O/20% AN) were determined. These re-
sults were compared with those observed in mixed solvents
2
the desired temperature (<70 °C). About 0.3 mL aliquots of the
reaction were periodically taken into sample vials precooled in
ice. The samples were immediately placed in a freezer (∼ -20 °C)
(
26) Cook, P. F.; Oppenheimer, N. J.; Cleland, W. W. Biochemistry 1981,
0, 1817–1825.
27) Hess, R. A.; Hengge, A. C.; Cleland, W. W. J. Am. Chem. Soc. 1998,
20, 2703–2709.
28) Alston, W. C.; Haley, K.; Kanski, R.; Murray, C. J.; Pranata, J. J.
Am. Chem. Soc. 1996, 118, 6562–6569.
2
1
(
(29) Dauben, H. J. Jr; Honnen, L. R.; Harmon, K. M. J. Org. Chem. 1960,
25, 1442–1445.
(30) Thompson, M. S.; Meyer, T. J. J. Am. Chem. Soc. 1982, 104, 4106–
(
4115.
J. Org. Chem. Vol. 74, No. 17, 2009 6509

Lu et al.
JOCArticle
pfo
(k ) of the reaction. The linear plots usually had regression
until 6-8 reaction aliquots within 1-3 half-lives of the reaction
had been collected. The aliquots of the reactions were then
analyzed by dilution of 80 μL of them in 1.92 mL of AN/water
2
coefficients (R ) greaterthan 0.995. Each kinetic runwas repeated
at least 3 times in most cases. The KIEs were obtained
by parallel determinations and taking the ratios of the pseudo-
first-order rate constants observed for the reactions involving
normal and deuterated i-PrOH, respectively.
(3:1, v/v) containing 3 M HClO by UV-vis spectroscopy, an
4
acid concentration that is high enough to convert all of the
PhXnOH in equilibrium with the PhXn in the aqueous aliquots
to PhXn . For the reactions in AN, analyses using AN as a
þ
þ
diluting agent gave the same results within experimental error.
The corresponding UV spectra at different reaction times, i.e.,
the kinetic scans, were then obtained. Absorbance (Abs) de-
creasing with time (t) at 373 nm due to the total PhXn
absorption was recorded (also see Results). The obtained
Acknowledgment. This work was supported by the Petro-
leum Research Fund administrated by the American Che-
mical Society (ACS-PRF: GB-44844) and a SIUE summer
research fellowship. We thank Professor Donald Bethell
for providing useful comments on the manuscript. We
thank Daniel Kabotso for helping collect some of the kinetic
data in Figure 3.
þ
Abs-t data were fit to the first-order integrated rate equation,
-
ln(Abs) = k t þ constant, and the slope of the linear plot of
3
-
ln(Abs) vs t was taken as the pseudo-first-order rate constant
6
510 J. Org. Chem. Vol. 74, No. 17, 2009