Cyclodextrin-based artificial oxidases with high rate accelerations and selectivity

Article abstract of DOI:10.1016/j.tetlet.2014.02.100

Three cyclodextrin derivatives with one to four 2-O-formylmethyl groups attached to the secondary rim were prepared and investigated as catalysts for the oxidation of aminophenols in buffered dilute hydrogen peroxide. The derivatives were found to be Michaelis-Menten catalysts and to give rate accelerations of up to 20,000 for the oxidation of 2-aminophenol to 2-amino-phenoxazin-3-one, and 12,000 for the oxidation of 2-amino-p-cresol to 2-nitro-p-cresol. While a range of differently substituted substrates was oxidized the success of the reaction was highly dependent on the substituent pattern. The ability of one of the new artificial enzymes to oxidize selectively one aminophenol from a mixture of two was investigated giving substrate selectivities of up to 16:1.

Full text of DOI:10.1016/j.tetlet.2014.02.100

Tetrahedron Letters 55 (2014) 2304–2307
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Cyclodextrin-based artificial oxidases with high rate accelerations
and selectivity
⇑
You Zhou, Emil Lindbäck, Christian M. Pedersen, Mikael Bols
Department of Chemistry, University of Copenhagen, Universitetsparken 5, 2100 Kbh Ø, Denmark
a r t i c l e i n f o
a b s t r a c t
Article history:
Three cyclodextrin derivatives with one to four 2-O-formylmethyl groups attached to the secondary rim
were prepared and investigated as catalysts for the oxidation of aminophenols in buffered dilute hydro-
gen peroxide. The derivatives were found to be Michaelis–Menten catalysts and to give rate accelerations
of up to 20,000 for the oxidation of 2-aminophenol to 2-amino-phenoxazin-3-one, and 12,000 for the oxi-
dation of 2-amino-p-cresol to 2-nitro-p-cresol. While a range of differently substituted substrates was
oxidized the success of the reaction was highly dependent on the substituent pattern. The ability of
one of the new artificial enzymes to oxidize selectively one aminophenol from a mixture of two was
investigated giving substrate selectivities of up to 16:1.
Received 2 January 2014
Revised 5 February 2014
Accepted 25 February 2014
Available online 5 March 2014
Keywords:
Chemzymes
Hydrogen peroxide
Peroxidase
Ó 2014 Elsevier Ltd. All rights reserved.
Catalysis
Michaelis–Menten
Enzymes are potent, selective, and environmentally friendly cat-
alysts of chemical reactions, and as such appear to be ideally suited
to be the main agents in a more sustainable and efficient chemical
synthesis. However, the use of enzyme catalysis in organic chemis-
try is limited by a narrow scope as natural enzymes normally prefer
natural substrates.1 Enzyme mimics (or artificial enzymes) could
increase the scope of enzyme catalysis in synthesis, as the catalyst,
in principle, may be designed for specific substrates.2
COOH
OH
N
NH2
O
O
O
O
O
O2N
kcat/kuncat = 1068
kcat /kuncat = 272
1
R= 2-OH
R=
4-OH-3-COOH
H2O2, buffer
pH 7, 25 oC
N
NH2
R= 4-Me
N
R
Enzyme-catalyzed oxidation, which potentially could be highly
important, has found relatively little use in synthesis, perhaps
because impractical and expensive cofactors are typically
necessary. It is therefore of interest that cyclodextrins (CDs)
containing a ketone or aldehyde functional group act as artificial
oxidases catalyzing the oxidation of an amine or alcohol with
hydrogen peroxide as the only oxidant and cofactor.3,4 For
example, various aromatic amines were oxidized into azo, nitro,
or other condensation products in an aqueous solution of hydrogen
peroxide when the catalyst 1 was present (Fig. 1), with rate accel-
erations of 70–1068.3
kcat /kuncat = 70
Figure 1. Suggested mechanism for the oxidation of an aromatic amine catalyzed
by a cyclodextrin ketone.
O
O
O
O
O
HO
X
O
OH
OH
HO
X= H or OH
NH
Ar
Both ketones5 and aldehydes6 appropriately attached to the pri-
mary rim of a- and/or b-cyclodextrin have been found to catalyze
Figure 2. Mechanism of oxidation of aromatic amines catalyzed by cyclodextrin 1.
oxidation reaction. The mechanism of oxidation is believed to oc-
cur by the addition of hydrogen peroxide to the ketone/aldehyde
and subsequent oxidation of the cyclodextrin-bound aminoben-
zene (Fig. 2) and/or hydroxylamine leading to a hydroxylamine
or nitroso-derivative, respectively. The nitroso-derivative then
condenses to produce azo-derivatives, or other products, or is fur-
ther oxidized to a nitro-compound.
In all the amine oxidase models investigated so far, the catalytic
group has been attached at the primary face of the cyclodextrin.
⇑
Corresponding author. Tel.: +45 35320160; fax: +45 35320214.
E-mail address: bols@chem.ku.dk (M. Bols).
http://dx.doi.org/10.1016/j.tetlet.2014.02.100
0040-4039/Ó 2014 Elsevier Ltd. All rights reserved.

Y. Zhou et al. / Tetrahedron Letters 55 (2014) 2304–2307
2305
Yet, a reasonable assumption is that the catalytic group is better
placed at the secondary rim since this opening is wider and there-
fore more likely to bind the substrate. We were thus interested in
investigating the effect of attaching an aldehyde to the secondary
rim and the results of this work are reported herein. We have
found that cyclodextrins with one to four aldehyde groups at-
tached to the secondary rim are the most powerful artificial amine
oxidase catalysts so far investigated.
Derivatives were chosen that have a formylmethyl moiety at-
tached to a secondary alcohol as these could readily be prepared
from the corresponding aldehyde. This led to the three b-cyclodex-
trin derivatives 2–4 (Fig. 3) with one, two, or four formylmethyl
groups attached, respectively. Compound 2 was prepared by the
method shown in Scheme 1 commencing with direct allylation of
b-cyclodextrin according to a literature procedure.7 This method
exploits the higher acidity of the 2-OH group, uses nucleophilic
catalysis, and proceeds in approximately 40% yield to give 5. This
compound was perbenzylated to afford 6 (89%), dihydroxylated
with OsO4/NMO to 7 (78%), and cleaved with NaIO4/silica to form
the aldehyde 8 (96%) that was deprotected to give 2 in quantitative
yield (Scheme 1).
previously been described,9 while 4 was made similarly from the
tetrol 9 as shown in Scheme 2. Allylation of 9 was carried out as
previously described10 to produce 10 (84%). Dihydroxylation of
10 with OsO4/NMO gave 11 in 66% yield. Periodate cleavage with
silica-supported sodium periodate11 converted 11 into 4 (79%).
With the compounds 2, 3, and 4 in hand, aromatic amine oxida-
tion catalysis was investigated in the manner previously used:3 the
aminophenol was dissolved (concentrations 1–18 mM) in 0.1 M
phosphate buffer (pH 7) with hydrogen peroxide (64 mM) and
the cyclodextrin (85 or 170 lM). For each aminophenol substrate
Michaelis–Menten catalysis was found with all three catalysts
and the kinetic parameters are shown in Table 1.
It is clear from the data that the cyclodextrins 2–4 promote an
extraordinary high level of catalysis with some of the substrates
(Table 1). For the oxidation of 2-aminophenol, we obtained a kcat
value of up to 0.05 sÀ1, and therefore a turnover time (i.e. the aver-
age time the substrate is bound before being converted) down to
20 s (entry 3). For 2-amino-p-cresol we also obtained similarly
high kcat values (entries 15 and 16). When the kcat is compared to
the kuncat from the uncatalyzed reaction that was determined
simultaneously, we obtained kcat/kuncat values of up to 20,000 (en-
try 3). This is the highest rate acceleration ever obtained in this
reaction with an artificial enzyme; the best previous value was
Compounds 3 and 4 were prepared from the derivatives
obtained from 2A,3B-didesmethylation or 2A,3B,2D,3E-tetradesme-
thylation of permethylated b-cyclodextrin, as has been described
by Sollogoub.8 Compound 3 was prepared from the diol product
by allylation, dihydroxylation, and periodate cleavage as has
ca. 5000 and was obtained with a permethylated
a-cyclodextrin
with two aldehyde groups on the primary face (at C6 in A and D).6
When comparing the three cyclodextrin derivatives we see that
3 and 4 are more active than 2 in that the kcat and kcat/kuncat values
are higher. This is consistent with earlier observations that the
more carbonyl groups in the molecule the higher the catalysis rate,
and is presumably due to the higher probability that the substrate
reacts with the catalytic group when more are present. On the
other hand, 3 and 4 have a similar level of catalytic efficiency either
with similar kcat values or one or the other having a somewhat
higher value. So there is no real advantage in having four formylm-
ethyl groups rather than two, and presumably this reveals a limit
as to how many groups can interact effectively with the bound
substrate—or in other words: crowding in 4 prevents all the groups
from contributing effectively.
RO
OR3
OR
O
O
R3O
O
O
O
RO
OR
RO
O
OR
OR
RO
O
O
RO
RO
OR
O
RO
OR
O
OR
O
OR
RO
O
O
OR
O
We also see that per-O-methylation is no disadvantage in these
catalysts as the known permethylated analogue of 2, 2a,6 has
essentially the same activity as unmethylated 2 (Table 1). This is
in accordance with previous findings with C6 aldehydes that
worked equally well with or without per-O-methylation.6
The catalysts display some interesting substrate selectivities
(Table 1). While 2-aminophenol (entries 1–4) and 2-amino-p-cre-
sol (entries 13–16) are very similar substrates toward the catalysts,
OR1
R2O
2 R=H, R1=CH2CHO, R2-3=H
1
R=Me, R =CH2CHO, R2-3=Me
R=Me, R1-2=CH2CHO, R3=Me
R=Me, R1-3=CH2CHO
2a
3
4
Figure 3. Structures of cyclodextrins 2–4 investigated in this work.
(MeO)17
(MeO)17
CH2CHCH2Br
(BnO)20
β
β
KOtBu
84%
CH2CHCH2Br
LiH/LiI
OsO4
BnBr
β
β
β-CD
2AO
2AO
O3E
H
O3E
O3B
H
2DO
H
O3B
2DO
NaH
89%
NMO
78%
H
2O
DMSO
40%
2O
9
10
5
6
(MeO)17
(MeO)17
(BnO)20
(BnO)20
H2
NaIO4
79%
OsO4
NaIO4
96%
β
β
Pd/C
β
β
β
NMO
66%
2AO
2AO
HO
HO
O3E
O3E
100%
O3B 2DO
HO
O3B 2DO
2O
2O
2O
HO
HO
OH
OH
HO
HO
O
8
2
7
O
O
O
11
4
O
O
HO
Scheme 1. Synthesis of aldehyde 2.
Scheme 2. Synthesis of tetraaldehyde 4.

2306
Y. Zhou et al. / Tetrahedron Letters 55 (2014) 2304–2307
Table 1
Kinetic data for the oxidation of aminophenols by cyclodextrin aldehydes
Entry
Substrate
Catalyst
Main product
kcat (10À5 sÀ1
)
Km (mM)
kcat/kuncat
1
2
2
2a
894.5 73.8
1223 52
38.4 0.8
23.7 2.0
1846.1 152.4
2919 242
3
4
3
4
5031.6 2515
4290.9 571.3
31.8 5.2
15.7 1.2
19463 9728
5940 715
2-aminophenol
4-aminophenol
5
6
7
8
2
2a
3
4
354.8 120.9
2438 464
1739.3 231.1
1722.7 517.7
31.3 3.1
93.8 19
32.6 1.5
15.9 7.0
827.5 282.0
3401 740
2980.7 396.1
2226.5 669.2
9
10
2
2a
327.1.1 102.7
848 86
7.1 2.5
23.2 3.0
188.4 59.1
307 37
11
12
3
4
429.1 30.0
2125.2 527.3
1.7 0.3
5.0 2.6
247.1 17.3
1223.9 303.7
6-amino-m-cresol
2-amino-p-cresol
13
14
15
16
2
2a
3
4
235.9 87.5
33.4 3.76
7.4 0.2
157.9 1.0
52.0 2.9
656.3 243.6
860 99
11841 1467.8
10269 3094
220
3
5228.2 648.1
4928.9 1485.3
17
18
2
3
120.1 12.3
175.8 44.4
1.6 0.8
5.4 1.2
89.3 9.2
130.7 33
19
4
749.5 493
27 3.2
413 271
4-amino-o-cresol
4-amino-m-cresol
20
21
22
23
2
2a
3
4
2.39 0.2
2.36 3.5
26.5 7.2
0.2 0.6
2.2 0.8
17.4 1.4
10.6 6.1
14.7 1.4
11
2
162.8 43.9
120.2 69
26.6 15.2
24
25
2
2a
92.7 43.3
6.3 0.2
14.6 4.4
0.2 0.1
100.7 47
18
1
26
27
3
4
91.6 11.7
165.3 34.9
1.3 0.9
11.1 1.98
99.4 12.7
179.5 37.9
2-amino-m-cresol
28
29
2
3
114.5 27.4
668 187
6.1 3.0
66.5 3.2
48.3 11.5
281.8 79
30
4
245.78 98.7
15.5 3.7
110.5 44
3-aminophenol
The artificial enzyme concentration [E] was 170 lM.
A
B
28–30) is a surprisingly poor substrate in comparison. We suspect
that the former may be related to some mechanistic advantages of
having phenol as a hydrogen bond donor close to the reaction site.
This is supported by the fact that 2-aminophenol and 4-aminophe-
nol display essentially the same binding abilities to the catalyst. 3-
Aminophenol may have problems geometrically binding in a way
that enables catalysis. On the other hand, we observed that 2–4
are the first cyclodextrins that actually catalyze the oxidation of
3-aminophenol.
A methyl group on 4-aminophenol was found to decrease the
efficiency as a substrate: 4-amino-o-cresol (entries 17–19) and 4-
amino-m-cresol (entries 20–23) are much poorer substrates than
the unsubstituted analogue (entries 5–8). Again this is in accord
with a model for catalytic binding where the amino group has to
be close to the secondary rim. This is impossible in these substrates
without a steric clash either from the methyl group or the hydroxyl
group (Fig. 4B).
H3C
H3C
OH
Steric
clash
H2N
H2N
2O
2O
OH
O
O
H
H
O
O
OH
OH
Figure 4. (A) Proposed catalytic binding of methylated 2-aminophenols. (B) The
methylated 4-aminophenols poorly bind the CD cavity with the amino group at the
secondary rim.
that is, a 4-methyl group is well tolerated, 6-amino-m-cresol is a
worse substrate. This clearly shows that a 5-methyl group is poorly
tolerated. A 3-methyl group is even worse (entries 24–27). This
selectivity means that the catalytic binding is optimal when the
substrate has the methyl group at the primary face, and simulta-
neously the hydroxyl group in the center of the secondary face
opening and the amino group close to the secondary face rim
and the catalytic aldehyde (Fig. 4A).
We also performed three experiments to see how the substrate
selectivities observed above would be displayed when two sub-
strates had to compete for the enzyme active site (Table 2). These
experiments were performed by mixing two amino phenol sub-
strates at close to equimolar concentrations (7 mM) in buffer as
With respect to the position of the amino group in the amino-
phenol, 2-(ortho) is clearly the better substrate (entries 1–4), but
4- is also tolerated (entries 5–8), while 3-aminophenol (entries
above, with H2O2 (64 mM) and cyclodextrin derivative
3
(0.17 M). The ratio between the two substrates was then measured

Y. Zhou et al. / Tetrahedron Letters 55 (2014) 2304–2307
2307
Table 2
pared to 2-amino-p-cresol) and causes preferable conversion of
the otherwise less reactive substrate (2-amino-p-cresol).
In summary, we have found that cyclodextrin derivatives with
an aldehyde attached to the secondary rim are the best artificial
enzymes reported so far for the oxidation of aminophenols.
The catalysts show some interesting substrate selectivities such
as preference for oxidation of 2-amino-p-cresol over other
amino-cresols, and are able to oxidize this phenol preferably in
the presence of another aminophenol. This is the first step toward
a synthetic procedure where an artificial enzyme selectively
picks out a substrate, among several analogues, and transforms
it accordingly.
Results of competition experiments of two aminophenols (7 mM each) reacted with
H2O2 (64 mM) and cyclodextrin 3 (0.17 mM) in phosphate buffer at pH 7 and 25 °C
Exp.
1
Substrates
kcat/kuncat
kauncat
Ratio at endb
1:1.8
2-Aminophenol
4-Amino-m-cresol
2-Aminophenol
2-Amino-m-cresol
4-Amino-o-cresol
2-Amino-p-cresol
19463
163
19463
99
131
11841
0.42
0.26
0.42
0.35
1.33
0.26
2
3
1:1.9
16:1
The ratio, initially 1:1, is shown at the end the experiment. (a) in 10À5 sÀ1, (b) after
2, 2 or 4 days of reaction time, respectively.
Acknowledgment
We thank the Danish National Research Council for support of
this project.
Supplementary data
Supplementary data (experimental procedures for the synthesis
of 2 and 4 and NMR spectra of the new compounds) associated
with this article can be found, in the online version, at http://
dx.doi.org/10.1016/j.tetlet.2014.02.100.
Figure 5. The ratio of 4-amino-m-cresol to 2-aminophenol as a function of time in a
competition experiment as measured by HPLC. In the experiment the two
aminophenols (7 mM each) together in phosphate buffer, pH 7 were treated at
25 °C with H2O2 (64 mM) and cyclodextrin 3 (0.17 mM).
References and notes
1. Faber, K. Biotransformations in Organic Chemistry, 3rd ed.; Springer: Berlin,
1997.
2. Kirby, A. J.; Hollfelder, F. From Enzyme Models to Model Enzymes; RSC
Publishing: Cambridge, 2009.
3. Marinescu, L.; Mølbach, M.; Rousseau, C.; Bols, M. J. Am. Chem. Soc. 2005, 127,
17578–17579.
by HPLC as a function of time. The change in the substrate ratio as a
function of time in one of these experiments (experiment 1) is
shown in Figure 5.
4. Marinescu, L.; Bols, M. Angew. Chem., Int. Ed. 2006, 45, 4590.
5. Fenger, T. H.; Marinescu, L.; Bols, M. Org. Biomol. Chem. 2009, 7, 933–943.
6. Fenger, T. H.; Bjerre, J.; Bols, M. ChemBioChem 2009, 10, 2494–2503.
7. Hanessian, S.; Benalil, A.; Laferriere, C. J. Org. Chem. 1995, 60, 4786.
8. Xiao, S.; Yang, M.; Sinaÿ, P.; Blériot, Y.; Sollogoub, M.; Zhang, Y. Eur. J. Org. Chem.
2010, 1510–1516.
9. Zhou, Y.; Pedersen, C. M.; Bols, M. Tetrahedron Lett. 2013, 54, 2458–2461.
10. Zhou, Y.; Lindbäck, E.; Marinescu, L. G.; Pedersen, C. M.; Bols, M. Eur. J. Org.
Chem. 2012, 4063–4070.
The ratio of unreacted substrates when the experiments were
terminated is shown in Table 2. In all the experiments, we found
that the better substrate in terms of kcat is preferentially converted,
though most clearly in experiment 3, where we see that the ratio
between 4-amino-o-cresol and 2-amino-p-cresol has changed from
1:1 to 16:1 after four days of reaction. This shows that the cyclo-
dextrin derivative overrides the intrinsic higher reactivity of 4-
amino-o-cresol (as seen from its 5 times higher kuncat value com-
11. Zhang, Y. L.; Shing, T. K. M. J. Org. Chem. 1997, 62, 2622–2624.