Very high rate enhancement of benzyl alcohol oxidation by an artificial enzyme

Article abstract of DOI:10.1002/anie.200600812

(Figure Presented) Promoting the atomic dialogue! At neutral pH and ambient temperature, a cyclodextrin-derived ketone causes hydrogen peroxide and benzyl alcohols to meet in its active site and react up to 60 000 times faster than they do outside the cyclodextrin.

Full text of DOI:10.1002/anie.200600812

Communications
DOI: 10.1002/anie.200600812
Artificial Enzymes
Very High Rate Enhancement of Benzyl Alcohol
Oxidation by an Artificial Enzyme**
Lavinia G. Marinescu and Mikael Bols*
Angewandte
Chemie
4
590
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 4590 –4593

Angewandte
Chemie
catalyzed reaction is observable even at cyclodextrin concen-
E
nzymes have fascinated scientists since their discovery,
trations that are 10000 times lower than the substrate
concentration and as low as 1 mm. GC-MS analysis revealed
that complete oxidation of the starting material can be
achieved. The reaction follows Michaelis–Menten kinetics
(Figure 1), can be inhibited by addition of cyclopentanol, and
and, over some decades, one aim in organic chemistry has
been to create molecules that mimic the active sites of
[
1–6]
enzymes and promote catalysis.
there are relatively few examples of enzyme models that
actually perform Michaelis–Menten catalysis under enzy-
matic conditions (i.e., water, pH 7, ambient temperature),
and very high rate accelerations under these conditions are
rare. On the other hand progress in synthetic chemistry, in
this context, in carbohydrate chemistry, now makes it
possible to prepare more sophisticated rigid catalysts.
Recently we reported the bridged ketocyclodextrins 1 and
Nevertheless even today
[
7]
[
8]
[
9]
[
10]
2
(Scheme 1)
and that they catalyze the oxidation of
Figure 1. Hanes plot of the 2-catalyzed ([2]=0.38 mm) oxidation of 2-
methoxybenzyl alcohol to 2-methoxybenzaldehyde in the presence of
7
2 mm H O , 258C.
2 2
is catalyzed neither by b-cyclodextrin nor by 1,3-dichloroa-
cetone (DCA, Table 1). 1,3-Diacetoxyacetone (DAA) cata-
lyzes the reaction slightly (Table 1). Under these reaction
conditions 1 and 2 are stable and preincubation with hydrogen
peroxide for a couple of hours does not decrease the catalytic
activity.
Scheme 1. Catalysts 1 and 2, which consist of a core of either a- or b-
cyclodextrin with dihydroxyacetone attached to the primary rim through
ester bonds.
The kinetic data (K and k ) for a range of benzyl alcohol
m
cat
substrates were obtained from vcat vs. [S] data (vcat: velocity of
the catalyzed reaction, S: substrate) in the usual manner using
nonlinear least-squares fitting and are shown in Table 1. For
pH 7 and 258C, the Km values range from (0.21 Æ 0.07) to
(5.0 Æ 1.6) mm which are typical values for the binding of
anilines to nitrobenzenes in the presence of hydrogen
peroxide (H O ) in a reaction with enzyme kinetics and
2
2
[
11]
with a ratio k /kuncat up to 1070.
The cup-shaped com-
cat
pounds 1 and 2 bind the aromatic amino group in their
hydrophobic cavities (K ꢀ 1–5 mm), while the ketone func-
small aromatic molecules by cyclodextrins. The k constants
m
cat
À6
À4 À1
tionality is believed to form a hydroperoxide adduct with
vary from about 10 to about 2.5 ” 10
s
and since the first-
H O2 that is responsible for the oxidation of the bound
order rate constant for the background oxidation (kuncat)
2
[
11]
À8
À9 À1
amine. We have now found that 1 and 2 also catalyze the
oxidation of benzylic alcohols to aldehydes, and that for this
reaction the rate acceleration for the catalyzed reaction is
very high.
under these condition is 10 –10
s
the ratio between the
reaction rates for the reactions inside and outside the cavity
(k /kuncat) varies from about 400 to about 29000 at 258C. The
cat
comparison with diacetoxyacetone is also significant: DAA
The cyclodextrins 1 and 2 catalyze the transformation of
benzylic alcohols into aldehydes (or ketones) in the presence
of hydrogen peroxide (72 mm) in aqueous solution, at room
temperature, and at pH 7.0 [Eq. (1)]. The reaction rate of the
catalysis gives a second-order rate constant k
of 4 ”
cat,2
À4 À1 À1
10
m s , which has to be compared to the k /K values
cat m
À2
À2 À1 À1
for 1 and 2 that are 1.35 ” 10
and 1.75 ” 10 m s ,
respectively; it follows that the supramolecular catalysts are
0 to 45 fold more efficient than DAA. As can be seen from
3
Table 1 secondary alcohols (Me or Ph substituents) are
oxidized as well. A range of substituents on the aromatic
ring is tolerated though with variations in the catalysis rate.
For three reactions, a study of pH and temperature effects
was performed (see Table 1, entries marked with [a]). The
reaction rate increases with temperature and pH value, but
above pH 9 the catalyst starts to decompose, presumably as a
result of the lability of the esters At pH 8.0 and 458C the rate
increase observed for the oxidation of 2-hydroxybenzyl
[
*] Dr. L. G. Marinescu, Prof. Dr. M. Bols
Department of Chemistry
University of Aarhus
4
alcohol is approximately 6.3 ” 10 . Though this value still is
8000 Aarhus (Denmark)
Fax: (+45)8619-6199
E-mail: mb@chem.au.dk
lower than that found for natural enzymes, where rate
[
12]
6
increases have been calculated
to lie between 10 and
1
8
2
[
**] We thank the Lundbeck foundation and the SNF(FNU) for support.
10 , it is fascinating to consider that since 2 has a 10 times
Angew. Chem. Int. Ed. 2006, 45, 4590 –4593ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4591

Communications
Table 1: Kinetic data for the oxidation of various benzylic alcohols to aldehydes catalyzed by 1 or 2 and
compared with those of other potential catalysts. Except otherwise noted the experiments were
performed in a 95 mm phosphate buffer at 258C and pH 7 with a H O concentration of 72 mm and with
smaller molecular weight than a typical
enzyme, the efficiency per mass unit is in
the same range.
2
2
a concentration of the catalyst of 0.38 mm.
From the series of p-substituted benzyl
alcohols (see Table 1) one can construct a
Hammett plot (Figure 2), which gives a
6
À1
Substrate
A or X
Catalyst
kcat [”10 s
]
Km [mm]
kcat/kuncat
1
2
26.9Æ3.2
26.3Æ2.0
244Æ6
2.0Æ0.7
1.5Æ0.4
–
1690Æ230
2
1650Æ150
correlation between kcat and s of r = 0.82,
[
a]
c]
[13]
2
2
989Æ28
and 1.9 as the reaction constant (1). This
[
no catalysis
no catalysis
2.2Æ0.2
1
1
value suggests that the transition state is
ionic and has some negative charge at the
reaction center. A comparison of 1-phenyl-
ethanol with its 1-deuterated analogue
(Table 1) reveals a small isotope effect (kH/
kD = 1.4). Based on this we suggest a
mechanism (Scheme 2) related to what has
been formulated for the tungsten-catalyzed
H O -oxidation of alcohols, which also gives
DCA
DAA
1
–
–
[
d]
1
H
24.4Æ1.5
0.72Æ0.30
5.0Æ1.6
0.77Æ0.63
5930Æ360
1
H
2
2
22.6Æ1.7
15.8Æ1.4
5490Æ420
2
H
not determined
1
2
18.6Æ0.3
18.3Æ0.4
1.1Æ0.1
973Æ22
960Æ34
0.84Æ0.1
b-CD
no catalysis
–
1
2
2
2
6.24Æ0.63
1.3Æ0.2
1.20Æ0.15
–
28600Æ9200
62900Æ5500
1
[14]
a relatively low isotope effect (k /k = 2).
H
D
[
a]
b]
2
2
62.1Æ0.8
This mechanism is also similar to what we
[
no catalysis
no catalysis
[
11]
have proposed for amine oxidation. The
DCA
–
1
[
15]
somewhat high 1 value
could be an
OMe
Cl
Br
2
2
2
2
1
2
3.41Æ0.07
0.52Æ0.08
0.61Æ0.17
0.71Æ0.19
1.2Æ0.4
1.83Æ0.51
1.09Æ0.56
2740Æ130
971Æ76
366Æ27
82Æ13
748Æ106
887Æ103
8.90Æ0.35
66.8Æ4.5
591Æ62
indication that the reaction is facilitated by
hydrogen bonding between the proton of
the benzylic alcohol and a Lewis base, such
as the keto hydroxy group, thus invoking the
cyclic mechanism outlined. The remarkable
supramolecular catalysis displayed by 1 and
[
a]
Br
5.62Æ0.44
4.55Æ0.47
OMe
OMe
OMe
OH
F
Cl
Br
Br
1
2
2
2
2
2
1
2
2
2
2
9.63Æ0.60
3.85Æ0.12
159Æ12
2.52Æ0.43
0.97Æ0.16
6.1Æ0.8
6760Æ570
2700Æ110
1760Æ130
not determined
402Æ41
2
under mild aqueous conditions is, as
[
a]
discussed above, in the best case as efficient
per mass-unit catalyst as that of some
natural enzymes. When this level of catal-
ysis can be achieved, practical applications
of such molecules are not far away.
1.28Æ0.08
9.10Æ0.49
21.5Æ0.4
33.7Æ0.9
24.2Æ0.1
3.93Æ0.98
1.57Æ0.03
16.0Æ0.8
0.09Æ0.08
0.66Æ0.23
0.37Æ0.05
1.15Æ0.12
0.93Æ0.02
0.78Æ0.93
0.50Æ0.06
0.21Æ0.07
707Æ155
1030Æ30
739Æ4
Me
OMe
Cl
945Æ236
2120Æ86
446Æ81
Experimental Section
Determining the rate of oxidation: Each reaction
was performed on 4–16 samples (2 mL each) of
the appropriate substrate at different concentra-
tions in 95 mm phosphate buffer containing
2
293Æ51
11.8Æ2.3
539Æ94
7
2 mm H O and either 1 or 2 (1 mg) or nothing
2 2
(as control). The reactions were followed at 258C
[
a] T=458C, pH 8. [b] Cyclopentanol (109 mm) was added as inhibitor. [c] H O was replaced with tert-
using UV absorption at an appropriate wave-
length and typically monitored for 5 h. Veloc-
2
2
À1 À1
À1
[16]
butyl hydroperoxide. [d] kcat,2 =0.4 mm
s
; kcat/kuncat =9.8 mm
.
ities were determined as the slope of the progress
curve of each reaction. The velocities of the
Figure 2. Hammett plot of the 2-catalyzed oxidation of p-substituted
benzyl alcohols at pH 7.0 in the presence of 72 mm H O , 258C. The
plot has a slope (1) of 1.93 and a correlation r of 0.82.
Scheme 2. Proposed mechanism of the catalytic process. First hydro-
gen peroxide is added to the ketone, and then bound substrate is
oxidized. The oxidation step may be assisted by a hydrogen bond.
2
2
2
4
592
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 4590 –4593

Angewandte
Chemie
À1
À1
uncatalyzed reactions were obtained directly from the control
samples, those of the catalyzed reactions were calculated by
subtracting the uncatalyzed rate from the total rate of the appropriate
cyclodextrin-containing sample. The vcat values were used to construct
Hanes plots ([S]/v vs. [S]) to ensure that the reaction follows
Michaelis–Menten kinetics. In that case Km and vmax were determined
using least-squares nonlinear regression fitting to the vmax vs. [S]
[16] The following extinction coefficients e [mm cm ] (258C, pH 7)
and wavelengths l [nm] of the products were determined and
used: benzaldehyde (1.23, 285), acetophenone (0.32, 300),
benzophenone (4.03, 280), 2-hydroxybenzaldehyde (2.92, 325),
2-methoxybenzaldehyde (2.75, 323), 2-chlorobenzaldehyde
(1.24, 300), 2-bromobenzaldehyde (0.47, 292), 3-methoxyben-
zaldehyde (2.48, 313), 4-methoxybenzaldehyde (7.53, 300), 4-
hydroxybenzaldehyde (6.1, 329), 4-fluorobenzaldehyde (0.88,
292), 4-chlorobenzaldehyde (1.09, 286), 4-bromobenzaldehyde
(1.59, 295), 4-methylbenzaldehyde (3.37, 285), 2,6-dimethoxy-
benzaldehyde (3.63, 323), 2,6-dichlorobenzaldehyde (1.03, 309),
acetylferrocene (1.68, 345 nm).
curve. kcat was calculated as vmax/[cyclodextrin]. k
as the slope from a plot of vuncat vs. [S].
was determined
uncat
[
16]
GC-MS experiments: A pseudo-preparative experiment was
performed by mixing 0.5 mmol substrate with 0.02 mmol catalyst 1
or 2 in 1.5 mL H O and slowly adding 1.5 mL H O , 35%. The
2
2
2
mixture was incubated at 258C from 3 h to several days. After
extraction of the mixture with dichloromethane and evaporation the
sample was analyzed in a Hewlett-Packard 5890A gas chromatograph
equipped with a 5971A MSD mass-selective detector.
Received: March 2, 2006
Published online: June 21, 2006
Keywords: biomimetics · cyclodextrins ·
.
Michaelis–Menten kinetics · supramolecular chemistry
[
1] A. J. Kirby, Angew. Chem. 1996, 108, 770 – 790; Angew. Chem.
Int. Ed. Engl. 1996, 35, 706 – 724.
[
[
2] R. Breslow, S. D. Dong, Chem. Rev. 1998, 98, 1997 – 2012.
3] E. E. Karakhanov, A. L. Maksimov, E. A. Runova, Y. S. Karda-
sheva, M. V. Terenina, T. S. Buchneva, A. Y. Guchkova, Macro-
mol. Symp. 2003, 204, 159 – 173.
[
[
4] S. Sasaki, K. Koga, Stud. Org. Chem. 1992, 45, 265 – 310.
5] W. B. Motherwell, M. J. Bingham, Y. Six, Tetrahedron 2001, 57,
4663 – 4686.
[
6] R. Breslow, Science 1982, 218, 532 – 537.
[
7] For some recent examples, see: a) M. J. Han, S. K. Yoo, J. Y.
Chang, T.-K. Ha, Angew. Chem. 2000, 112, 355 – 357; Angew.
Chem. Int. Ed. 2000, 39, 347 – 349; b) R. Breslow, X. Zhang, Y.
Huang, J. Am. Chem. Soc. 1997, 119, 4535 – 4536; c) M.
Kunishima, K. Yoshimura, H. Morigaki, R. Kawamata, K.
Terao, S. Tani, J. Am. Chem. Soc. 2001, 123, 10760 – 10761;
d) S. H. Yoo, B. J. Lee, H. Kim, J. Suh, J. Am. Chem. Soc. 2005,
127, 9593 – 9602; e) L. Jiang, Z. L. Liu, Z. Liang, Y. Gao, Bioorg.
Med. Chem. 2005, 13, 3673 – 3680; f) N. M. Milovic, J. D. Badjic,
N. M. Kostic, J. Am. Chem. Soc. 2004, 126, 696 – 697.
[
8] High rate acceleration has however been observed in the
transacylation of cyclodextrins at high pH, see: G. L. Trainer,
R. Breslow, J. Am. Chem. Soc. 1981, 103, 154 – 158.
[
9] A. J. Pearce, P. Sinaþ, Angew. Chem. 2000, 39, 3610 – 3611;
Angew. Chem. Int. Ed. 2000, 39, 3610 – 3611.
[
10] C. Rousseau, B. Christensen, M. Bols, Eur. J. Org. Chem. 2005,
734 – 2739.
2
[
11] L. Marinescu, M. Mølbach, C. Rousseau, M. Bols, J. Am. Chem.
Soc. 2005, 127, 17578 – 17579.
[
[
12] R. Wolfenden, M. J. Snider, Acc. Chem. Res. 2001, 34, 938 – 945.
2
13] A plot of kcat/K vs. s (not shown) gave a poorer correlation (r =
m
0
.55) and 1.2 as 1.
14] S. E. Jacobson, D. A. Muccigrosso, F. Mares, J. Org. Chem. 1979,
4, 921 – 924.
15] Ruthenium- and palladium-catalyzed oxidation of benzylic
alcohols gives reaction constants that are small and negative
[
[
4
(
1 ꢀ À0.5; see: a) A. Dijksman, A. Marino-Gonzalez, A. Mair-
ata I Payeras, I. W. C. E. Arends, R. A. Sheldon, J. Am. Chem.
Soc. 2001, 123, 6826 – 6833; b) J. A. Mueller, C. P. Goller, M. S.
Sigman, J. Am. Chem. Soc. 2004, 126, 9724 – 9734). As the
present reaction does not involve formation of a positively
charged metal center, its high 1 value is not surprising.
Angew. Chem. Int. Ed. 2006, 45, 4590 –4593ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4593