Substrate control through per-O-methylation of cyclodextrin acids

Article abstract of DOI:10.1039/c0cc02071k

Per-O-methylated cyclodextrins containing a single 2-O-(2-acetate), 2-O-(3-propanoate) or a 6-carboxylate were investigated for glycosidase activity on p-nitrophenyl glycosides. The former two compounds displayed enzyme catalysis giving rate accelerations of 500-1000, while the latter compound gave marginal catalysis. These results show that per-O-methylated cyclodextrins direct substrate binding from the secondary face leading to better catalysis. The Royal Society of Chemistry.

Full text of DOI:10.1039/c0cc02071k

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Substrate control through per-O-methylation of cyclodextrin acidsw
Thomas H. Fenger and Mikael Bols*
Received 24th June 2010, Accepted 2nd September 2010
DOI: 10.1039/c0cc02071k
Per-O-methylated cyclodextrins containing
a single 2-O-
(2-acetate), 2-O-(3-propanoate) or a 6-carboxylate were investi-
gated for glycosidase activity on p-nitrophenyl glycosides. The
former two compounds displayed enzyme catalysis giving rate
accelerations of 500–1000, while the latter compound gave
marginal catalysis. These results show that per-O-methylated
cyclodextrins direct substrate binding from the secondary face
leading to better catalysis.
Enzymes have fascinated chemists since the time of Emil
Fischer. This is not surprising since they are crucial for essen-
tially every biological event, and their power and selectivity are
of a magnitude beyond normal chemical standards.1 While a
century long major research effort in protein chemistry, bio-
chemistry, molecular biology and crystallography has unraveled
much about existing enzymes, actually building new enzymes
is still very difficult.2 Thanks to a monumental effort by
Breslow3 and others,4 cyclodextrin derivatives have proven
to be the most successful artificial enzymes, but the number of
compounds that actually display Michaelis–Menten kinetics is
still quite small.5 One of the problems with cyclodextrins is
perceived to be a tendency to indiscriminate binding allowing
guests to bind from both faces of the cavity and opposite
orientations.3b
Fig. 1 Cyclodextrin derivatives 1–3 having carboxylic acids at the
primary rim.
Table 1 Kinetic parameters for hydrolysis of nitrophenyl glycosides
in 0.5 M phosphate buffer, pH 8.0, 59 1C
Catalyst
Substratea
kcat (Â107 sÀ1
188
17
1.5
—
—
—
269
217
289
55
173
273
131
31
)
Km/mM
kcat/kuncat
17
26
3
3
3
3
4
4
4
4
5
5
5
PNP b-glc
PNP b-glc
PNP b-glc
PNP a-glc
PNP a-gal
ONP b-gal
PNP b-glc
PNP a-glc
PNP a-gal
ONP b-gal
PNP b-glc
PNP a-glc
PNP a-gal
ONP b-gal
7.90
—
—
—
—
—
4.68
4.88
3.02
0.59
5.53
6.37
1.97
—
989
122
14
—
—
Diacid 1 (Fig. 1) is an example of a cyclodextrin based
artificial enzyme.7 It catalyses hydrolysis8 of nitrophenyl
glycosides at 59 1C, pH 8 in phosphate buffer with a rate
increase (kcat/kuncat) of bound substrate (relative to unbound)
of 989 (Table 1). The catalysis is presumably a result of
electrostatic effects: when the substrate binds to the primary
face of 1 (Fig. 2, top left) the anomeric center comes close to
the negatively charged carboxylate, which promotes the
formation of a positively charged transition state. It is likely
that the substrate can bind in several orientations only some of
which allow catalysis. Thus a complex between 1 and substrate
with the sugar at the secondary face (Fig. 2, top right) will be
unproductive and, as it undoubtedly also is formed, leads
to a decrease in overall catalysis. If binding orientations
could be limited it would obviously pave the way for better
artificial enzymes. In this communication we report that per-
O-methylation can be used to obtain such substrate control.
Per-O-methylation of cyclodextrins is potentially attractive
since it gives these molecules a wider range of solvent
solubilities, while retaining or even increasing water solubility.
—
975
538
763
56
517
646
326
19
5
a
PNP = p-nitrophenyl, ONP = o-nitrophenyl, glc = D-glucopyranoside,
gal = D-galactopyranoside.
More important in this context it is known that per-O-methylated
cyclodextrins have a more narrow primary face opening.9 Also it
has been shown by crystallography that several guests, among
them p-nitrophenol, bind to cyclodextrin and per-O-methylated
cyclodextrin in opposite orientations.10
The per-O-methylated analogues of 1, 26 and 311 (Fig. 1)
were quite disappointing: both 2 and 3 displayed poor or
no enzymatic activity (Table 1). However in contrast the
O-methylated derivatives 411 and 511 that have a single carboxylic
acid attached to the secondary face (Fig. 3) were found to be
much better. Under similar conditions 4, which has an
acetic acid residue attached to 2-OH, catalysed hydrolysis
of p-nitrophenyl b-glucoside displaying Michaelis–Menten
Department of Chemistry, University of Copenhagen,
Universitetsparken 5, 2100 Kbh Ø, Denmark.
E-mail: bols@kemi.ku.dk
w Electronic supplementary information (ESI) available: Experimental
procedures, plot of kcat vs. phosphate, kinetic plots and data with
errors included and Roesy spectra. See DOI: 10.1039/c0cc02071k
c
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 7769–7771 7769

View Article Online
Fig. 4 Dependance of kcat/kuncat for 4 on hydrolysis of PNP a-glc of
phosphate concentration at T = 59 1C in phosphate buffer.
phosphate (Fig. 4). Therefore these secondary face substituted
cyclodextrin acids appear more potent than the primary face
substituted cyclodextrin acids.
Fig. 2 Binding modes of phenyl glycoside substrates to cyclodextrin
acids 1 (top) and 4 (bottom). The glycoside can bind from the primary
face (left) or the secondary face (right).
Overall the data show that while per-O-methylation decreases
catalysis for cyclodextrin acids with the carboxylate at the primary
face, per-O-methylated cyclodextrins with a carboxylate at the
secondary face works well. This behavior can be explained
by a change in binding equilibria as shown in Fig. 2. If
unmethylated cyclodextrin derivatives, such as 1, bind nitro-
phenyl glycosides both from the primary and the secondary
face (Fig. 2, top), while per-O-methylated derivatives predo-
minantly bind at the secondary face (Fig. 2, bottom) catalysis
as the one observed would be expected. To verify this binding
behaviour NMR spectra of the complexes of 4-nitrophenyl
b-D-glucoside (PNP glc) with either b-cyclodextrin (CD) or
per-O-methyl-b-cyclodextrin (MCD) were analyzed. In the
complex between CD and PNP glc NOEs were observed
between the aromatic protons (in PNP glc) and H-5 and H-3
(in CD), while in the complex with MCD the o-nitro protons
(in PNP glc) show NOE to both H-3 and H-5 (in MCD), while
the m-nitro protons only have NOE to H-3 (see ESIw). This
clearly shows a change in binding conformations towards
more predominant binding from the secondary face when
MCD is host.
Fig. 3 Cyclodextrin derivatives 4–5 having carboxylic acids at the
secondary rim.
kinetics and giving a kcat/kuncat of 975 (Table 1). The corres-
ponding a-glucoside and galactoside substrates were also
accepted with kcat/kuncat of 538 and 763, respectively. Only the
o-nitrophenyl b-galactoside was a poor substrate probably
because the change in aromatic substitution leads to unproductive
or poor binding.
In summary this work shows that per-O-methylation of a
cyclodextrin artificial enzyme can direct its substrate binding
towards the secondary face. In this particular case the more
discriminative binding appears to cause more efficient catalysis.
Catalysis is presumably caused by electrostatic interactions
(for suggested transition state see ESIw). Limited substrate
control is undoubtedly one of the drawbacks when cyclo-
dextrins are used in artificial enzymes and per-O-methylation
may be one of the solutions to this problem.
The analogue 5, which had a 3-propionate at a 2-OH
(Fig. 3), behaved very similar to 4, the only difference being
that the rate accelerations for most substrates were slightly
smaller (Table 1).
FNU and KU are thanked for financial support, and
Christian Thorzen for assistance with NOE.
On first sight the catalysis displayed by 4 and 5 appears very
similar to that of unmethylated primary face substituted
cyclodextrin acids such as 1. Thus the catalysis by 4 shows
similarly a linear dependence in phosphate (Fig. 4), which
indicates that phosphate is involved in the reaction as a
nucleophilic catalyst.12 However, only one acid group is
required for the catalysis here while for the primary face
substituted acids two carboxylate groups are necessary. Also
4 (unlike 1) does not need phosphate assistance for catalysis as
a component of the catalysis will occur without the presence of
Notes and references
1 R. Wolfenden and M. J. Snider, Acc. Chem. Res., 2001, 34,
938–945.
2 A. Kirby and F. Hollfelder, From Enzyme Models to
Model Enzymes, RSC, Cambridge, 2009.
3 (a) R. Breslow, Artificial Enzymes, Wiley-VCH, Weinheim, 2005;
(b) R. Breslow and S. Dong, Chem. Rev., 1998, 98, 1997–2011.
4 (a) Y.-H. Zhou, M. Zhao, Z.-W. Mao and L.-N. Ji, Chem.–Eur. J.,
2008, 14, 7193–7201; (b) M. Fukudome, K. Shimosaki, K. Koga,
c
7770 Chem. Commun., 2010, 46, 7769–7771
This journal is The Royal Society of Chemistry 2010

View Article Online
D.-Q. Yuan and K. Fujita, Tetrahedron Lett., 2007, 48, 7493–7497;
(c) P. G. McCracken, C. G. Ferguson, D. Vizitiu,
C. S. Walkinshaw, Y. Wang and G. R. J. Thatcher, J. Chem.
Soc., Perkin Trans. 2, 1999, 911–912; (d) H. Tsutsumi, H. Ikeda,
H. Mihara and A. Ueno, Bioorg. Med. Chem. Lett., 2004, 14,
723–726; (e) N. Masurier, F. Estour, M.-T. Froment, B. Lefevre,
J.-C. Debouzy, B. Brasme, P. Masson and O. Lafont, Eur. J. Med.
Chem., 2005, 40, 615–623; (f) A. Fragoso, R. Cao and M. Banos,
Tetrahedron Lett., 2004, 45, 4069–4071.
5 L. Marinescu and M. Bols, Trends Glycosci. Glycotechnol., 2010,
21, 309–323.
6 J. Bjerre, PhD thesis, University of Copenhagen, 2010.
7 C. Rousseau, F. O. Caballero, L. U. Nordstrøm, B. Christensen,
T. E. Pedersen and M. Bols, Chem.–Eur. J., 2005, 11, 5094–5101.
8 The reaction is dependent on phosphate which presumably as a
phosphorolysis external nucleophilic catalyst, but the final product
is the glycose.
9 S. Immel and F. W. Lichtenthaler, Starch, 1996, 48, 225–232.
10 (a) K. Harata, Bull. Chem. Soc. Jpn., 1977, 50, 1416–1424;
(b) K. Harata, K. Uekama, M. Otagiri and F. Hirayama, Bull. Chem.
Soc. Jpn., 1982, 55, 3904–3910; (c) K. Harata, K. Uekama, M. Otagiri,
F. Hirayama and Y. Sugiyama, Bull. Chem. Soc. Jpn., 1981, 54,
1954–1959; (d) K. Harata, K. Uekama, M. Otagiri, F. Hirayama and
Y. Sugiyama, Bull. Chem. Soc. Jpn., 1982, 55, 3386–3389.
11 Prepared from oxidation of the corresponding aldehyde, see ESIw.
12 We have not been able to show glucose-1-phosphate (Glc-1P) as a
product in these reactions only glucose, but Glc-1P is likely to
hydrolyse quite readily under the conditions.
c
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 7769–7771 7771