Cyclodextrins to limit substrate inhibition and alter substrate selectivity displayed by enzymes

Article abstract of DOI:10.1039/b008339i

The substrate inhibition exhibited by carboxypeptidase A in catalysing the hydrolysis of (S)-2-O-(N-benzoylglycyl)β-phenyllactate) is limited by addition of cyclodextrins. The cyclodextrins do not significantly change the maximum rate of reaction, but they increase the concentration of the substrate at which the maximum rate of reaction is observed, by more than an order of magnitude when 0.105 mol dm^-3 hydroxypropyl-β-cyclodextrin is used. Cyclodextrins also alter the substrate selectivity of α-chymotrypsin in catalysing the hydrolysis of (S)-N-acetylleucine methyl ester and (S)-N-acetylphenylalanine methyl ester, in favour of reaction of the former. Calculations show that these effects are due to complexation of the substrates by the cyclodextrins. Thus the results establish that cyclodextrins can be used to manipulate the concentrations Of enzyme substrates in free solution in a predictable manner.

Full text of DOI:10.1039/b008339i

View Article Online / Journal Homepage / Table of Contents for this issue
Cyclodextrins to limit substrate inhibition and alter substrate
selectivity displayed by enzymes
a
a
a
a
b
Christopher J. Easton,* Jason B. Harper, Sarah J. Head, Kitty Lee and Stephen F. Lincoln
1
a
Research School of Chemistry, Australian National University, Canberra ACT 0200,
Australia. E-mail: easton@rsc.anu.edu.au
Department of Chemistry, University of Adelaide, Adelaide SA 5005, Australia
b
Received (in Cambridge, UK) 17th October 2000, Accepted 22nd January 2001
First published as an Advance Article on the web 27th February 2001
The substrate inhibition exhibited by carboxypeptidase A in catalysing the hydrolysis of (S)-2-O-(N-benzoylglycyl)-
β-phenyllactate) is limited by addition of cyclodextrins. The cyclodextrins do not significantly change the maximum
rate of reaction, but they increase the concentration of the substrate at which the maximum rate of reaction is
Ϫ3
observed, by more than an order of magnitude when 0.105 mol dm hydroxypropyl-β-cyclodextrin is used. Cyclo-
dextrins also alter the substrate selectivity of α-chymotrypsin in catalysing the hydrolysis of (S)-N-acetylleucine
methyl ester and (S)-N-acetylphenylalanine methyl ester, in favour of reaction of the former. Calculations show
that these effects are due to complexation of the substrates by the cyclodextrins. Thus the results establish that
cyclodextrins can be used to manipulate the concentrations of enzyme substrates in free solution in a predictable
manner.
Introduction
Enzymes are highly efficient biological catalysts (Scheme 1).
Scheme 1
Scheme 2
However, some of the important features of enzyme catalysis
in vivo limit the utility of enzymes in vitro. In biological systems
enzymes catalyse reactions of substrates that are present only at
low concentrations, but for in vitro applications it is often desir-
able to have much higher substrate concentrations whereas the
substrates often have only low aqueous solubility. This has been
Results and discussion
Carboxypeptidase A catalyses the hydrolysis of (S)-2-O-(N-
benzoylglycyl)-β-phenyllactate (1) (Scheme 3). The rate of
8
1
addressed using enzymes in organic solvents and with surfact-
2
ants. Alternatively, cyclodextrins have been used to complex
enzyme substrates and increase the total amount taken up in
3,4
water, making them more accessible for reaction.
Enzyme catalysis in vivo is regulated by mechanisms which
also limit the utility of enzymes in vitro. Product inhibition is
5
one such form of regulation. Biochemically this prevents
wasteful metabolism of a substrate but in vitro it reduces the
rate and extent of conversion to product. Previously we have
shown that cyclodextrins can be used to reduce this inhibition,
3,6
through selective product complexation.
Phenylalanine
ammonia lyase shows product inhibition in catalysing the con-
7
version of (S)-phenylalanine to (E)-cinnamate, but the extent
of that inhibition is reduced and the efficiency of the catalysis
is increased by using cyclodextrins to selectively complex the
cinnamate in the reaction mixture as it forms. Another form of
in vivo regulation of enzyme catalysis is substrate inhibition,
where high substrate concentrations decrease enzyme activity.
Typically this is the result of a second substrate molecule bind-
Scheme 3
5
ing to the enzyme (Scheme 2). Again this prevents excessive
formation of the products 2 and 3 is a function of the concen-
tration of the substrate 1, as shown in Fig. 1a. This profile is
characteristic of substrate inhibition of enzyme catalysis. At
low concentrations of the substrate 1 (below 10 mol dm ) the
reaction efficiency is determined primarily by the extent of
binding of the substrate 1 to the enzyme, to form the binary
complex EؒS (Scheme 2). At higher concentrations, additional
substrate metabolism in a biological system but limits the
in vitro utility of the process because low concentrations must
be used. We now report the use of cyclodextrins to reduce sub-
Ϫ4
Ϫ3
8
strate inhibition, as displayed by carboxypeptidase A. We also
discuss the application of cyclodextrins to alter the substrate
selectivity shown by α-chymotrypsin.
5
84
J. Chem. Soc., Perkin Trans. 1, 2001, 584–587
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b008339i

View Article Online
substrate concentrations used. This correlates well with the
experimentally determined effects of the cyclodextrins (Fig. 1)
and confirms that they are adequately described by Scheme 2.
The principle described above, of using cyclodextrins to
manipulate the concentration of an enzyme substrate free in
solution, was also applied to alter the substrate selectivity dis-
played by α-chymotrypsin. Some enzymes catalyse reactions
of mixtures of substrates (Scheme 4). It was envisaged that
Fig. 1 Rate of hydrolysis of the lactate 1 catalysed by carboxy-
Ϫ3
Ϫ3
Scheme 4
peptidase A (20 units dm ) in 0.05 mol dm Tris buffer (I = 0.5 mol
Ϫ3
dm , KCl), at pH 7.5 and 298 K, containing either a) no cyclodextrin
Ϫ3
Ϫ3
(
᭹), b) β-cyclodextrin (9.96 × 10 mol dm ) (᭛) or c) hydroxy-
through selective complexation, cyclodextrins could be used to
change the ratio of the substrates free in solution, thereby
affecting the relative rates at which the reactions of the
substrates were catalysed.
α-Chymotrypsin catalyses the hydrolysis of N-acetylamino
acid methyl esters (Scheme 5). When a ca. 10 : 1 mixture of
Ϫ3
propyl-β-cyclodextrin (0.105 mol dm ) (᭜).
binding of the substrate 1 occurs to give the catalytically
inactive ternary complex EؒS , so the rate of reaction decreases.
2
When β-cyclodextrin and hydroxypropyl-β-cyclodextrin are
added to the reaction mixtures, bell-shaped logarithm of sub-
strate 1 concentration versus reactivity profiles are observed
(
Figs. 1b and 1c), similar to that seen in the absence of a cyclo-
dextrin. The maximum rate of reaction does not change signifi-
cantly. However, the concentration of the substrate 1 giving rise
to the maximum rate of catalysis is larger, by a factor of ca. 3.5
Ϫ3
for the reaction in the presence of β-cyclodextrin (9.96 × 10
Ϫ3
mol dm ) and ca. 10 when hydroxypropyl-β-cyclodextrin
0.105 mol dm ) is used. In practical terms, the latter result
Ϫ3
(
means that the scale of the reaction can be decreased by at least
an order of magnitude without adversely affecting the rate.
The effects of the cyclodextrins can be attributed to their
complexation of the substrate 1 and the effects of this complex-
ation on the interactions of the substrate 1 with the enzyme, in
accord with Scheme 2. At low concentrations of the substrate 1,
its complexation by the cyclodextrins decreases the extent of
formation of the binary complex EؒS, thereby reducing the rate
of reaction. At higher concentrations, cyclodextrin complex-
ation of the substrate 1 decreases the extent of formation of the
Scheme 5
Ϫ2
Ϫ3
(
S)-N-acetylleucine methyl ester (4a) (1.01 × 10 mol dm )
Ϫ3
and (S)-N-acetylphenylalanine methyl ester (4b) (1.00 × 10
Ϫ3
mol dm ) was treated with α-chymotrypsin, the ratio of form-
ation of the acids 5a and 5b was 1 : 3.9 when 25% of the phenyl-
alanine derivative 4b had been processed. This ratio decreased
as the extent of reaction increased, consistent with the changing
ratio of the residual substrates 4a and 4b. When 40% of the
phenylalanine derivative 4b had been processed, the ratio of the
product acids 5a and 5b was 1 : 2.7. On repeating the reaction
in the presence of hydroxypropyl-β-cyclodextrin (0.097 mol
ternary complex EؒS in favour of the binary complex EؒS, thus
2
increasing the rate of reaction.
Hydroxypropyl-β-cyclodextrin and other modified cyclo-
dextrins are often used because they have higher affinities for9
particular guests and therefore display more marked effects.
Ϫ3
A
A
Ϫ3
dm ), 6 -amino-6 -deoxy-β-cyclodextrin (0.12 mol dm ) and
1
However, that was not the case in this work. Using H NMR
Ϫ3
α-cyclodextrin (0.056 mol dm ), the respective ratios of the
spectroscopy and the changes in chemical shift of signals of
product acids 5a and 5b were 1 : 1.0, 1 : 1.0 and 1 : 0.6, when
approximately 25% of the phenylalanine derivative 4b had been
consumed, and 1 : 0.9, 1 : 0.9 and 1 : 0.4, after approximately
0% reaction of the ester 4b.
Each of the cyclodextrins changes the ratio of formation of
10
the substrate 1 on complexation with the cyclodextrins, the
association constants with β-cyclodextrin and hydroxypropyl-3
β-cyclodextrin were calculated to be 250 ± 60 and 120 ± 20 dm
4
Ϫ1
mol , respectively. Instead, in this study the greater effect of
the hydroxypropylcyclodextrin relative to β-cyclodextrin on the
enzyme catalysis reflects the use of a larger amount of the
former. The solubility of β-cyclodextrin in water is limited to
the products 5a and 5b, this effect being greatest with α-cyclo-
dextrin. To determine if these effects are adequately accounted
for by the impact of cyclodextrin complexation as represented
in Scheme 4, it was necessary to know the association constants
for the cyclodextrin complexes. The value for the complex of
the leucine derivative 4a with α-cyclodextrin was determined to
Ϫ3
less than 0.02 mol dm but more of the hydroxypropylcyclo-
11
dextrin could be used because it is much more soluble.
The association constants of the complexes of the cyclo-
dextrins with the lactate 1 were calculated using eqn. (1). Based
3
Ϫ1
1
be 30 ± 29 dm mol , using H NMR spectroscopy. However,
9
despite considerable effort and using a variety of methods, it
was not feasible to determine the corresponding value in the
case of the phenylalanine derivative 4b. Using H NMR and
[
complex]
K =
(1)
1
[
guest][cyclodextrin]
ultraviolet spectroscopy, no significant changes were observed
on adding α-cyclodextrin to solutions of the amino acid deriv-
ative 4b. With circular dichroism and fluorescence spectroscopy,
the latter being used to study displacement of 6-(p-toluidino)-
naphthalene-2-sulfonate from the cyclodextrin cavity, practical
on this formula, in the experiments discussed above β-cyclo-
dextrin and the hydroxypropylcyclodextrin would be expected
to reduce the concentration of the substrate 1 free in solution
by factors of 3.5 and 13.5, respectively, across the range of
J. Chem. Soc., Perkin Trans. 1, 2001, 584–587
585

View Article Online
limitations resulted in very large experimental errors. As a
result, reliable association constants could not be measured
in these ways. Using H NMR spectroscopy, the association
Carboxypeptidase A catalysed hydrolysis of (S)-2-O-
(N-benzoylglycyl)-ꢀ-phenyllactate (1), in the presence and
absence of cyclodextrins
1
constants of the complexes of the amino acid derivatives 4a
Solutions of (S)-2-O-(N-benzoylglycyl)-β-phenyllactate (1)
and 4b with hydroxypropyl-β-cyclodextrin were calculated to be
Ϫ5
Ϫ2
Ϫ3
(
from 2.5 × 10 to 2.5 × 10 mol dm ) and carboxypeptidase
Ϫ3 Ϫ3
3
Ϫ1
3
0 ± 20 and 84 ± 15 dm mol , respectively. The large errors
A (ca. 20 units dm ) in 0.05 mol dm Tris buffer (I = 0.5 mol
dm , KCl), at pH 7.5 and 298 K, containing either no cyclo-
dextrin, β-cyclodextrin (9.96 × 10 mol dm ) or hydroxy-
propyl-β-cyclodextrin (0.105 mol dm ), were monitored for
changes in absorbance at 250–290 nm. Reaction rates were
calculated from the spectra using the UV-2101 PC Kinetics
software package. The results are illustrated in Fig. 1.
result from the small changes in chemical shift of signals of the
guests 4a and 4b on complexation by the cyclodextrin. Taking
the association constants at face value, it follows from eqn. (1)
that under the conditions used in the experiment discussed
above the cyclodextrin will complex ca. 75% of the leucine
derivative 4a and ca. 90% of the phenylalanine derivative 4b,
changing the ratio of the substrates 4a and 4b free in solution
by a factor of ca. 2.5. Selective complexation of the phenyl-
alanine derivative 4b by the cyclodextrin is consistent with
biasing the enzyme catalysis in favour of reaction of the leucine
derivative 4a, according to Scheme 4.
Ϫ3
Ϫ3
Ϫ3
Ϫ3
13
ꢁ-Chymotrypsin catalysed competitive hydrolysis of (S)-N-
acetylleucine methyl ester (4a) and (S)-N-acetylphenylalanine
methyl ester (4b), in the presence and absence of cyclodextrins
Ϫ2
In summary, the results discussed above provide further
examples of the use of cyclodextrins to increase the efficiency
Solutions of (S)-N-acetylleucine methyl ester (4a) (1.01 × 10
Ϫ3
mol dm ), (S)-N-acetylphenylalanine methyl ester (4b)
3
Ϫ3
Ϫ3
Ϫ3
of enzymes in vitro and they define parameters that should be
(1.0 × 10 mol dm ) and α-chymotrypsin (ca. 600 units dm )
Ϫ3
considered in order to predict the effect of a cyclodextrin on an
enzyme-catalysed process. They show that cyclodextrins can be
used to manipulate the concentrations of enzyme substrates
free in solution, to reduce substrate inhibition and alter sub-
strate selectivity in manners which may be rationalised accord-
ing to Schemes 2 and 4, respectively. In each system, it is
reasonable to expect that a substrate complexed by a cyclo-
dextrin will readily dissociate to maintain equilibrium with the
in 0.02 mol dm Tris buffer, at pH 7.9 and 310 K, containing
either no cyclodextrin, hydroxypropyl-β-cyclodextrin (0.097
Ϫ3
A
A
Ϫ3
mol dm ), 6 -amino-6 -deoxy-β-cyclodextrin (0.12 mol dm )
Ϫ3
or α-cyclodextrin (0.056 mol dm ), were analysed at 15 min
intervals using HPLC. The retention times of the esters 4a and
4b, and the acids 5a and 5b were 7.4, 9.8, 3.5 and 4.6 min,
respectively.
9
Association constants of cyclodextrin complexes
free substrate as it is consumed through reaction. Studies have
shown that dissociation rate constants of cyclodextrin com-
plexes are typically much greater than 1 s . Therefore the
cyclodextrins may be regarded as substrate reservoirs.
Solutions of (S)-2-O-(N-benzoylglycyl)-β-phenyllactate (1)
Ϫ1
Ϫ3
Ϫ3
Ϫ3
(
(
1.0 × 10 mol dm ) in 0.05 mol dm pH 7.5 Tris buffer
I = 0.5 mol dm , KCl), comprising 10% deuterium oxide and
Ϫ3
containing either β-cyclodextrin (0–10 mole equivalents) or
hydroxypropyl-β-cyclodextrin (0–100 mole equivalents), were
analysed at 298 K using H NMR spectroscopy. Changes in the
Experimental
1
difference between the chemical shifts of the signals due to the
ortho- and meta-protons of the benzoyl group of the lactate 1
as a function of cyclodextrin concentration were analysed by
non-linear regression to give association constants for 1 : 1
complexes of the lactate 1 with β-cyclodextrin and hydroxy-
Ultraviolet spectra were recorded on a Shimadzu UV-2101 PC
spectrophotometer coupled to a Shimadzu CPS temperature
controller. High performance liquid chromatography (HPLC)
TM
TM
was carried out using Waters 510 HPLC pumps, a Waters
TM
7
17plus Autosampler,TMa Waters
Tunable Absorbance
3
Ϫ1
propyl-β-cyclodextrin of 250 ± 60 and 120 ± 20 dm mol ,
Detector and a Waters 410 Differential Refractometer, on a
TM

respectively. The association constants of 30 ± 29, 30 ± 20 and
Waters Symmetry column, eluting with acetonitrile–water
3
Ϫ1
1
84 ± 15 dm mol , for the 1 : 1 complexes of the ester 4a with
α-cyclodextrin, and the esters 4a and 4b with hydroxypropyl-β-
cyclodextrin, respectively, were measured in a similar manner,
(
1 : 4, v/v) containing 0.01% TFA. H NMR spectra were
recorded at 500 MHz on a Varian Inova 500S spectrometer.
Carboxypeptidase A (type I from bovine pancreas), α-chymo-
trypsin (type II from bovine pancreas) and the racemate of the
sodium salt of 2-O-(N-benzoylglycyl)-β-phenyllactate (1)† were
purchased from Sigma Chemical Co. (S)-N-Acetylleucine
methyl ester (4a) and (S)-N-acetylphenylalanine methyl ester
Ϫ3
except that the experiments were conducted in 0.02 mol dm
pH 7.9 Tris buffer at 310 K, and complexation was monitored
using changes in the difference between the chemical shifts of
the signals due to acetyl protons and side chain protons of the
esters 4a and 4b as a function of cyclodextrin concentration.
(
4b) were prepared using standard methods. α-Cyclodextrin
and β-cyclodextrin were obtained as generous gifts from Nihon
Shokuhin Kako Co. Hydroxypropyl-β-cyclodextrin was pur-
chased from Cyclolab Ltd., Hungary. The H NMR spectrum
Acknowledgements
J. B. H. gratefully acknowledges receipt of the Shell Australia
Postgraduate Scholarship.
1
of this material showed an average of 3.5 hydroxypropyl groups
per cyclodextrin moiety, giving the material an effective molec-
A
A
ular weight of 1347. 6 -Amino-6 -deoxy-β-cyclodextrin was
12
prepared as reported previously. All cyclodextrins were dried
to constant weight in vacuo over phosphorus pentaoxide prior
to use.
References
1
For examples see: (a) A. Freeman and M. D. Lilly, Appl. Microbiol.
Biotechnol., 1987, 25, 495; (b) C.-S. Chen and C. J. Sih, Angew.
Chem., Int. Ed. Engl., 1989, 28, 695; (c) C. H. Wong, Science,
1
989, 244, 1145; (d ) A. M. Klibanov, Acc. Chem. Res., 1990, 23, 114;
†
(
Racemic material was used in all experiments, in the knowledge that
R)-2-O-(N-benzoylglycyl)-β-phenyllactate is not hydrolysed by carb-
oxypeptidase A, and does not affect the interaction of the (S)-
(e) M. N. Gupta, Eur. J. Biochem., 1992, 203, 25; ( f ) A. L. Gutman
and M. Shapira, Adv. Biochem. Eng./Biotechnol., 1995, 52, 87;
(g) Y. L. Khmelnitsky and J. O. Rich, Curr. Opin. Chem. Biol., 1999,
3, 47.
1
3,14
enantiomer 1 with the enzyme.
The cyclodextrins were generally
used in large excess, so their association complexes formed with (R)-2-
O-(N-benzoylglycyl)-β-phenyllactate do not substantially affect their
complexation of the (S)-enantiomer 1. The H NMR studies showed
that complexation of 2-O-(N-benzoylglycyl)-β-phenyllactate by the
cyclodextrins is not stereoselective. Quantities reported in the Experi-
mental section refer to the (S)-enantiomer 1 only.
2 For examples see: (a) H. W. Cook and W. E. M. Lands, Can. J.
Biochem., 1975, 53, 1220; (b) E. G. Ceen, J. P. R. Herrmann and
P. Dunnill, Appl. Microbiol. Biotechnol., 1987, 25, 491; (c) R. Bru,
E. Bloechliger and P. L. Luisi, Arch. Biochem. Biophys., 1993, 307,
295; (d) J. M. López-Nicolás, R. Bru, A. Sánchez-Ferrer and
F. García-Carmona, Anal. Biochem., 1994, 221, 410.
1
5
86
J. Chem. Soc., Perkin Trans. 1, 2001, 584–587

View Article Online
3
4
For a review see: J. B. Harper, C. J. Easton and S. F. Lincoln,
Curr. Org. Chem., 2000, 4, 429.
Chemistry, Imperial College Press, London, 1999; (b) S. F. Lincoln
and C. J. Easton, in Structural Diversity and Functional Versatility of
Polysaccharides, ed. S. Dumitriu, Marcel Dekker, New York, 1998,
pp. 473–521.
For examples see: (a) J. M. López-Nicolás, R. Bru, A. Sánchez-
Ferrer and F. García-Carmona, Biochem. J., 1995, 308, 151; (b) H. J.
Woerdenbag, N. Pras, H. W. Frijlink, C. F. Lerk and T. M. Malingré,
Phytochemistry, 1990, 29, 1551; (c) D. L. Alexander and J. F. Fisher,
Steroids, 1995, 60, 290; (d) D. Schlosser, S. Irrgang and H.-P.
Schmauder, Appl. Microbiol. Biotechnol., 1993, 39, 16; (e) T.
Zelinski and M.-R. Kula, Biocatal. Biotransform., 1997, 15, 57;
10 (a) S. E. Brown, J. H. Coates, S. F. Lincoln, D. R. Coghlan
and C. J. Easton, J. Chem. Soc., Faraday Trans., 1991, 87, 2699;
(b) S. E. Brown, J. H. Coates, P. A. Duckworth, S. F. Lincoln,
C. J. Easton and B. L. May, J. Chem. Soc., Faraday Trans., 1993,
89, 1035.
11 (a) K. Uekama and M. Otagiri, CRC Crit. Rev. Ther. Drug Carrier
Syst., 1987, 3, 1; (b) A. Yoshida, H. Arima, K. Uekama and J. Pitha,
Int. J. Pharm., 1988, 46, 217.
12 S. E. Brown, J. H. Coates, D. R. Coghlan, C. J. Easton, S. J. van Eyk,
W. Janowski, A. Lepore, S. F. Lincoln, Y. Luo, B. L. May, D. S.
Schiesser, P. Wang and M. L. Williams, Aust. J. Chem., 1993, 46,
953.
13 (a) R. C. Davies, J. F. Riordan, D. S. Auld and B. L. Vallee,
Biochemistry, 1968, 7, 1090; (b) J. R. Whitaker, F. Menger and
M. L. Bender, Biochemistry, 1966, 5, 386; (c) W. O. McClure,
H. Neurath and K. A. Walsh, Biochemistry, 1964, 3, 1897; (d) M. L.
Bender, J. R. Whitaker and F. Menger, Proc. Natl. Acad. Sci. U.S.A.,
1965, 53, 711.
(
f ) W. Van Uden, J. A. Bos, G. M. Boeke, H. J. Woerdenbag and
N. Pras, J. Nat. Prod., 1997, 60, 401.
(a) The Enzymes, ed. P. D. Boyer, Academic Press, New York, 1970;
5
6
(
b) I. H. Siegel, Enzyme Kinetics: Behaviour and Analysis of Rapid
Equilibrium and Steady-State Enzyme Systems, Wiley Interscience,
New York, 1975.
C. J. Easton, J. B. Harper and S. F. Lincoln, J. Chem. Soc., Perkin
Trans. 1, 1995, 2525.
E. A. Havir and K. R. Hanson, Biochemistry, 1968, 7, 1904.
See, for example: (a) W. O. McClure, H. Neurath and K. A. Walsh,
Biochemistry, 1964, 3, 1897; (b) J. R. Whitaker, F. Menger and
M. L. Bender, Biochemistry, 1966, 5, 386; (c) R. C. Davies, J. F.
Riordan, D. S. Auld and B. L. Vallee, Biochemistry, 1968, 7, 1090;
7
8
(
5
d) J. W. Bunting and S. S.-T. Chu, Biochim. Biophys. Acta, 1978,
24, 393.
14 (a) J. W. Bunting and J. Murphy, Biochem. Biophys. Res. Commun.,
1972, 48, 1316; (b) J. Murphy and J. W. Bunting, Can. J. Chem.,
1975, 53, 283; (c) J. W. Bunting and S. S.-T. Chu, Biochemistry, 1976,
15, 3237.
9
For reviews see: (a) C. J. Easton and S. F. Lincoln, Modified
Cyclodextrins. Scaffolds and Templates for Supramolecular
J. Chem. Soc., Perkin Trans. 1, 2001, 584–587
587