Artificial glycosyl phosphorylases

Article abstract of DOI:10.1002/chem.200500364

α- and β-Cyclodextrin 6^A,6^Ddiacids (1 and 2), β-cyclodextrin-6monoacid (14), β-cyclodextrin ⁶,6 ^D-diO-sulfate (16) and β-cyclodextrin-6heptasulfate (19) were synthesised. Acids 1, 2 and 14 were made from perbenz

Full text of DOI:10.1002/chem.200500364

Artificial Glycosyl Phosphorylases
[a]
Cyril Rousseau,[a] Fernando Ortega-Caballero,[a] Lars UlrikNordstrøm,
[b]
Brian Christensen,[b] Torben EllebækPetersen, and Mikael Bols*[a]
Abstract: a- and b-Cyclodextrin 6A,6D-
diacids (1 and 2), b-cyclodextrin-6-
monoacid (14), b-cyclodextrin 6A,6D-di-
O-sulfate (16) and b-cyclodextrin-6-
heptasulfate (19) were synthesised.
Acids 1, 2 and 14 were made from per-
benzylated a- or b-cyclodextrin, by di-
isobutylaluminum hydride (DIBAL)-
promoted debenzylation, oxidation and
deprotection. Addition of molecular
sieves was found to improve the deben-
zylation reaction. Sulfates 16 and 19
were made by sulfation of the appro-
priately partially protected derivatives
and deprotection. Catalysis of 4-nitro-
phenyl glycoside cleavage by these cy-
clodextrin derivatives was studied.
Compounds 1, 2 and 16 were found to
catalyse the reaction, with the catalysis
followingMichaelis–Menten kinetics
and dependingfirst order on the phos-
phate concentration. In a phosphate
buffer (0.5m, 598C, pH 8.0), KM varied
from 2–10 mm and the kcat/kuncat ratio
from 80–1000 dependingon the stereo-
chemistry of the substrate and the cata-
lyst, with 2 beingthe best catalyst and
with the sulfated 16 also displayingcat-
alytic ability. The monoacid 14 and the
heptasulfate 19 were not catalytic.
Keywords: artificial enzymes · cy-
clodextrins · glycosides · hydrolysis ·
kinetics
Introduction
and minor aliphatic groups.[12] Its only drawback is that car-
ryingout synthetic chemistry on CD derivatives is relatively
difficult,[13] makingthe preparation of many analouges a
slow process.
Enzymatic catalysis of chemical reactions occurs with im-
pressive selectivity and rate.[1] Thanks to enzymesꢀ abilities
to pick out specific molecules and to transform them selec-
tively, biosynthesis greatly surpasses even modern chemical
synthesis in its ability to make complex molecules. If chem-
ists could learn how to create catalysts with the characteris-
tics of enzymes—artificial enzymes (AEs)—chemical syn-
thesis would in all likelihood be revolutionised, so research
in artificial enzymes is an important frontier.[2–6]
Since enzymes achieve their astoundingrate enhance-
ments through proximity effects, artificial enzymes should
be attainable through endowment of a host molecule with
appropriate catalytic groups in order to mimic the active
site of an enzyme. A very promisinghost molecule for artifi-
cial enzymes is the cyclodextrin (CD) molecule,[7–11] which is
water-soluble and shows good binding constants to aromatic
For this work we wished to create a CD-based AE that
would be able to mimic the action of a glycosidase. It was
anticipated that a CD containingappropriate catalytic
groups at the secondary or primary rim should bind aromat-
ic groups and, provided that these groups were correctly
positioned, catalyse hydrolysis of an aryl glycoside
(Figure 1). Glycosidases were the first enzymes to have their
catalytic machinery unravelled; they each contain two car-
boxylate groups positioned 5.5 Š apart.[14] One carboxylate
group acts as a general acid catalyst, whilst the other func-
tions as a nucleophile. Presumably a CD containingtwo car-
boxylic acids at the rim with the appropriate separation and
locked in optimal conformation should catalyse aryl glyco-
side hydrolysis at optimal pH.
The application of CDs as glycosidase mimics has been re-
ported previously. Ohe et al. observed that a-CD (3) in-
creased the rate of hydrolysis of 4-nitrophenyl a-mannopyr-
anoside (25) by up to 7.6 times at pH 12, while the hydroly-
sis of the b-anomer was unaffected.[15] Conversion of other
nitrophenyl glycosides was increased by 3 by factors of up to
8.6, and experiments with partially methylated a-CD deriva-
tives indicated that the 2-OH group was necessary for the
rate accelerations. b-CD (4) did not affect the hydrolysis.
[a] Dr. C. Rousseau, Dr. F. Ortega-Caballero, L. U. Nordstrøm,
Prof. Dr. M. Bols
Department of Chemistry, University of Aarhus
Langelandsgade 140, 8000 Aarhus (Denmark)
Fax : (+45)861-961-99
E-mail: mb@chem.au.dk.
[b] B. Christensen, Dr. T. E. Petersen
Protein Chemistry Laboratory, Department of Molecular Biology
University of Aarhus, 8000 Aarhus C (Denmark)
5094
ꢁ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
DOI: 10.1002/chem.200500364
Chem. Eur. J. 2005, 11, 5094 – 5101

FULL PAPER
communication[17] we reported
that 1 and 2 acted as artificial
glycosidases of nitrophenyl gly-
cosides at neutral pH. The cat-
alysed conversion of 4-nitro-
phenyl glycosides was found to
follow Michaelis–Menten ki-
netics. At pH 7.4 and 598C,
kcat/kuncat varied from 12 to 35,
dependingon the stereochem-
istry of the substrate. We have
now found that this reaction is
phosphate-dependant and cat-
alyses glycoside cleavage with
kcat/kuncat ratios of up to 1000 at
high phosphate concentrations.
We also report the syntheses
of, and catalysis studies with,
Figure 1. Catalytic action of a cyclodextrin-containing artificial glycosidase.
the monoacid and sulfate ana-
logues of 1 and 2, which alto-
Since the 2-OH group of 3 is partially deprotonated at
pH 12, it appears likely that an alcoholate may be associated
with the observed acceleration. Nevertheless, it is also note-
worthy that 4 had little effect on the hydrolysis of 4-nitro-
phenyl glycosides even under these relatively extreme pH
conditions. Bennet and collaborators investigated the hy-
drolysis of 4-nitrophenyl 2-tetrahydropyranyl ether, a glyco-
side model, catalysed by 3 and 4.[16] They found that 3 accel-
erated the hydrolysis about fourfold, while 4 decreased the
rate of hydrolysis. A b-CD derivative with a single 2-O-car-
boxymethyl group (5) similarly decreased the hydrolysis
rate. They also observed that 4 catalysed the hydrolysis of a
2-deoxyglucopyranosyl pyridinium salt with kcat/kuncat = 7.5.
The initial idea behind the present work was to mimic a
glycosidase by incorporating two carboxylates into a CD. It
was suggested that a derivative less conformationally flexi-
ble than 5 would be desirable, which suggested the idea of
investigating diacids 1 and 2. Inspection of models showed
that the distances between the carboxylates were 5.0 Š in 1
and 6.5 Š in 2, and thus relatively close to the distance
found in a glycosidase. At a pHꢀpKa a significant part of
the molecule should be in the monoprotonated form, simi-
larly to the active form of a glycosidase. In a preliminary
gether suggest that the catalysis is caused by electrostatic ef-
fects.
Results and Discussion
Synthesis of 1, 2 and 14: Compounds 1 and 2[18–20] were read-
ily obtained from the diols 8 and 9, made in turn by selec-
tive O-debenzylation of 6 or 7 by Pearce and Sinaþꢀs proce-
dure (Scheme 1).[21,22] This powerful method uses diisobuty-
laluminum hydride (DIBAL) in toluene to remove one or—
more importantly—two benzyl groups from the perbenzylat-
ed CD in a selective fashion. Pearce and Sinaþ obtained
yields of 82–83% of 8 or 9 through the use of DIBAL in tol-
uene (0.5m, 120–140 equiv) at 30–508C, and a 60% yield of
12 (Scheme 2) with DIBAL (0.1m, 35 equiv) at room tem-
perature. In some experimentation to minimize the expendi-
ture of DIBAL in this reaction we made the surprisingand
useful observation that 4 Š molecular sieves have a positive
effect (Table 1). Treatment of 6 with 0.1m DIBAL at room
temperature normally provided mainly monool (Table 1,
entry 1), but when 4 Š molecular sieves were added a good
yield of 8 was obtained (Table 1, entry 2). Other types of
Chem. Eur. J. 2005, 11, 5094 – 5101
www.chemeurj.org
ꢁ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
5095

M. Bols et al.
by treatment with Dess–Martin
reagent, and the dialdehyde
was further oxidised to the
diacid 10 (86%) by treatment
with
sodium
chlorite[23]
(Scheme 1). By similar meth-
ods 9 was converted to 11 in
84% yield. The two diacids 10
and 11 were hydrogenolysed to
1 and 2, respectively, in quanti-
tative yield.
The pKa values of these two
diacids were determined by ti-
tration. The individual pKa
values were too close to be de-
Scheme 1. Synthesis of diacids 1 and 2.
termined individually, but an
average pKa of 3.5 was deter-
mined for 1 and a pKa of 3.2
was determined for 2. Simula-
tion of the titration curves
with plausible values of pKa(1)
and pKa(2) showed that the
pKa values of 1 were definitely
within in the 2.5–4.5 range,
Scheme 2. Synthesis of monoacid 14.
whilst those of 2 were between
2.2 and 4.2.
The b-CD monoacid 14 was
made from 7 in a similar manner, with a DIBAL reaction
with 7 (Table 1, entry 10) beingemployed to obtain 12
(Scheme 2). Oxidation first with Dess–Martin reagent and
subsequently with NaClO2 converted 12 into acid 13 in 75%
yield. Hydrogenolysis of 13 gave the acid 14 in 94% yield.
Table 1. Isolated yields after chromatography of mono- and diol from
the reduction of 6 or 7 with DIBAL with different additives at room tem-
perature.[a]
Entry Substrate Additive
[gmL ꢁ1
Monool
[%]
Diol
[%]
DIBAL
conc. [m]
]
1
2
3
6
6
6
–
26
10
–
–
64
–
0.1
0.1
0.1
4 Š MS (0.08)
4 Š MS powder
(0.08)
Synthesis of sulfates 16 and 19: For comparison with the car-
boxylic acids it was also decided to synthesise O-sulfates.
These molecules should be negatively charged, similarly to a
carboxylate, but nonnucleophilic. Compound 9 was there-
fore converted into the disulfate 15 in 93% yield by treat-
ment with sulfur trioxide·pyridine complex[24] (Scheme 3).
Hydrogenolysis gave the CD disulfate 16, isolated as its
disodium salt, in 96% yield. This compound is essentially
neutral in aqueous solution.
4
5
6
6
3 Š MS (0.08)
sodium aluminate
(0.08)
19
–
23
–
0.1
0.1
6
6
sodium silicate
(0.08)
31
–
0.1
7
8
9
10
6
6
7
7
–
7
47
–
–
0.03
0.03
0.3
4 Š MS (0.08)
4 Š MS (0.08)
4 Š MS (0.08)
10
81
-
55
0.03
The desire to expand the negative charge at the primary
rim prompted us to synthesise 19, a b-CD with O-sulfates
on all six OH groups. This was made as outlined in
Scheme 3. The primary rim was silylated with tert-butyldi-
methylsilyl chloride (TBDMSCl), the secondary rim was
acetylated, and the silyl groups were removed with BF3 as
described previously, to provide 17 (Scheme 3).[25] Sulfation
with SO3/pyridine[25] gave 18 in 92% yield, and this was de-
protected with NaOMe to give the heptasulfate 19, isolated
as the heptasodium salt, in 88% yield.
[a] Reaction time was 21 h except at entry 8, which was 1.5 h. MS =
molecular sieves.
molecular sieves such as 3 Š (Table 1, entry 4) had a smaller
effect, as did sodium silicate and sodium aluminate (Table 1,
entries 5 and 6). These experiments suggest that the effect is
associated with the zeolite structure and not with the chemi-
cal constitution. With use of a lower DIBAL concentration
(0.03m), which otherwise had little effect (Table 1, entry 7),
the monool could be obtained quite effectively (Table 1,
entry 8). The conversion of 7 into 9 (Table 1, entry 9) or 12
(Table 1, entry 10) also worked better with the addition of
molecular sieves to either more or less concentrated solu-
tions of DIBAL. The diol 8 was oxidised to the dialdehyde
Catalysis experiments: At pH 6–8, compounds 1 and 2—and
also disulfate 16—catalysed the formation of 4-nitrophenol
from 4-nitrophenyl b-d-glucopyranoside (20; Scheme 4) at
catalyst concentrations of 1–3 mm in 50 mm phosphate
5096
ꢁ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 5094 – 5101

Artificial Glycosyl Phosphorylases
FULL PAPER
KM and kcat values could be
calculated. The kuncat constant
could be determined as the
slope of the plot of Vuncat
against substrate concentra-
tion.
The pH-dependence of the
reaction of 20 in the presence
of 2 in 50 mm phosphate buffer
is shown in Figure 2c. Since 2
is in dianionic form over the
entire pH range, the remarka-
ble change in kcat indicates de-
pendence on phosphate, since
the pH curve follows the ioni-
sation of secondary phosphate
(pKa = 7.2). Indeed, no catal-
ysis was observed when no
phosphate was present, such as
in a TRIS-sulfate buffer. When
the phosphate buffer concen-
tration was increased, kcat also
increased in a linear fashion
(Figure 2d). This means that
the data fit the rate law given
in Equation (1), where Pi is
phosphate and kcat = k’cat ”Pi.
In other words, bound sub-
strates undergo a second-order
Scheme 3. Synthesis of disulfate 16 and heptasulfate 19.
Scheme 4. Hydrolytic reaction and its substrates.
buffer. The stereoisomeric sub-
strates 23–24 (Scheme 4) were
also converted, whereas the 2-
nitrophenyl galactoside 25 was
not. b-CD (4), the monoacid
14 and the heptasulfates 18
and 19 were found not to be
catalytic under these condi-
tions.
The rate of catalysis (Vcat)
was obtained from the deter-
mination of absorption at
400 nm in parallel experiments
in the presence and in the ab-
sence of catalyst, giving a total
rate (Vtot) and an uncatalysed
rate (Vuncat), and Vcat = Vtotꢁ
Vuncat. With increasingsubstrate
concentration Vuncat increases
linearly while Vcat approaches
a maximum value, a clear sign
of Michealis–Menten-type ki-
netics (Figure 2a). A Haines
Figure 2. a) Vcat and Vuncat as a function of [S] (20) at pH 7.9, 598C in 50 mm phosphate buffer with 1.24 mm 2
as the catalyst. b) Haines plot for the conversion of 20 at pH 7.9, 598C in 50 mm phosphate buffer, with
1.24 mm 2 as the catalyst. c) pH-dependence of the formation of 22 from 20 catalysed by 2 at 598C. d) Phos-
phate-dependence of the formation of 22 from 20 catalysed by 1 at 358C, pH 8.0.
plot of [S]/Vcat against [S] gave
an excellent linear relationship
(Figure 2b), from which the
Chem. Eur. J. 2005, 11, 5094 – 5101
www.chemeurj.org
ꢁ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
5097

M. Bols et al.
reaction, which given the pH profile is likely to be substitu-
tion with monohydrogenphosphate.
Since 1 and 2 are in carboxylate form in the catalytic pH
range, the increased catalysis rate of bound substrate must
be associated with a proximity effect from the negatively
charged groups. Two possibilities can be imagined, either
nucleophilic attack or an electrostatic stabilisation of a posi-
tively charged transition state. This was the rationale for in-
vestigating the sulfate 16; nucleophilic catalysis is not possi-
ble in this compound and only electrostatic stabilisation can
play a role. The observation that 16 is a catalyst strongly
suggests that the catalysis is caused by electrostatic stabilisa-
tion of the transition state by the presence of the negative
charge in its vicinity. The kcat value of 16 is generally three
times lower than that of 2, which can be explained by the
negative charge in the sulfate being more remote from the
rim. The catalytic mechanism shown in Scheme5 is pro-
k1
k0cat
ƒƒ!
ƒ!
E þ S
ES þ P
E þ P
ð1Þ
ƒ
i
k
ꢁ1
k
ꢁ1þkcat
The KM value, which is defined as KM
=
, is essen-
k1
tially identical to the dissociation constant for the substrate–
catalyst complex, because kcat is likely to be much smaller
k
k1
ꢁ1
than kꢁ1 (i.e., KMffi ). The KM values, which typically range
from 3–10 mm, therefore reflect the efficiency of bindingof
the nitrophenyl group. This agrees well with the known dis-
sociation constants of the complexes formed between 20
and a-cyclodextrin (3) and between 20 and b-cyclodextrin
(4), which have been determined as 14 mm and 28 mm, re-
spectively.
The kinetic constants for the reaction of 20 at 500 mm
phosphate buffer at the optimum pH 8.0 are shown in
Table 2. Under these conditions the catalytic effect is consid-
erably more apparent than that observed in dilute phos-
phate,[17] with the catalytic efficiency over background reac-
tion (kcat/kuncat) beingup to a 989-fold increase in the best
case (Table 2).
The best substrate in these experiments is the b-glycoside
20, while the rate increases with a-glycoside substrates 23
and 24 are two to four times lower, so the configuration at
the anomeric position is of significance. The a-glycosides
also bind slightly more tightly (having lower KM values)
than 20, but this is not important since the KM value also re-
flects unproductive binding.
The stereochemistry of C4 in the substrate is, as one
would expect, of no significance as the conversions of 23
and 24 are essentially very similar.
Scheme 5. Proposed mechanism for the catalytic ability of the diacids.
The carboxylates stabilise positive charge in the transition state, facilitat-
ingsubstitution by secondary phosphate.
The best catalyst is 2, with 1 beingslightly inferior. The
disulfate 16 is somewhat weaker. Remarkably, the monoacid
14 has no catalytic effect, and in this respect is identical to
its precursor b-CD (4, Table 2). Interestingly, a-CD (3) has
a weakly catalytic effect, but unlike 1, 2 and 16 the catalysis
does not follow Michaelis–Menten-type kinetics but is a
simple substrate-dependant rate increase.
posed. The monoacid 14 displays no catalytic effect
(Table 2), presumably because there is insufficient negative
charge to have an appreciable effect. The heptasulfate 19
was investigated to increase the negative charge at the pri-
mary face. It was not catalytic, however, which is interpreted
as due to there beingtoo much crowdingat the primary
face to allow correctly aligned binding.
Conclusion
Table 2. Kinetic constants for the conversion of substrates in 500 mm phosphate buffer at pH 8.0. The catalyst
concentration was from 0.1–2.0 mm.
In summary, the cyclodextrin
diacids have been found to cat-
Catalyst
S
Temperature
kcat [”107 sꢁ1
]
KM [mm]
kcat/kuncat
1
1
1
1
20
20
23
24
20
20
23
24
20
20
23
24
20
35
59
59
59
59
59
59
59
59
59
59
59
59
10.4ꢂ0.3
73.7ꢂ10.9
183ꢂ5
9.41ꢂ0.80
11.9ꢂ3.7
706ꢂ26
247ꢂ38
301ꢂ8
alyse
phosphate-dependant
7.46ꢂ0.72
4.89ꢂ1.33
7.90ꢂ1.92
7.38ꢂ0.94
4.77ꢂ1.03
3.87ꢂ1.57
glycoside cleavage with kcat
/
86.0ꢂ4.2
188ꢂ14
9.23ꢂ0.44
106ꢂ4
292ꢂ15
989ꢂ77
42ꢂ2
kuncat ratios of up to 1000. They
form complexes with nitro-
phenyl glycosides with dissoci-
ation constants of 0.5–15 mm,
thereby activatingthe nitro-
phenoxy groups towards being
substituted by phosphate. The
correspondingdisulfate dis-
plays similar catalytic behav-
iour. It is therefore concluded
2
2[a]
2
2
3
16
16
16
4, 14, 19
267ꢂ10
407ꢂ25
1.40ꢂ0.27[b]
309ꢂ4
141ꢂ8
0.437ꢂ0.077[b]
97.6ꢂ1.3
–
[c]
2.89ꢂ0.14
0.96ꢂ2.96
1.91ꢂ3.07
30.7ꢂ3.6
79.9ꢂ9.6
64.5ꢂ8.9
23.7ꢂ2.9
[c]
[c]
[c]
–
–
–
[a] Carried out in 50 mm phosphate buffer. [b] Catalysis did not follow Michaelis–Menten kinetics, but the
simple rate law V = (kcat +kuncat)[S] in the presence of 2.6 mm 3. [c] No catalysis observed.
5098
ꢁ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 5094 – 5101

Artificial Glycosyl Phosphorylases
FULL PAPER
CHCl3); 1H NMR (400 MHz, CDCl3): d = 7.40–6.68 (m, 80H; CHPh),
that the catalysis is caused by electrostatic stabilisation of a
positively charged transition state.
ꢁ
5.64 (d, J = 3.6 Hz, 2H; H-1), 5.48 (d, Jgem = 10.4 Hz, 2H; H CHPh),
ꢁ
ꢁ
5.06 (d, Jgem = 10.4 Hz, 2H; H CHPh), 4.90 (d, Jgem = 10.4 Hz, 2H; H
CHPh), 4.83–4.50 (m, 21H), 4.48–4.33 (m, 8H), 4.30–3.86 (m, 28H), 3.57
(d, J = 11.6 Hz, 2H), 3.52 (dd, J = 3.4 Hz, J = 9.8 Hz, 2H), 3.43 (dd, J
= 2.8 Hz, J = 8.8 Hz, 2H), 3.23–3.03 (m, 5H) ppm; 13C NMR (100 MHz,
CDCl3): d = 171.31 (C=O), 139.46, 139.18, 139.06, 138.41, 138.32, 138.28,
137.72, 136.29 (Cipso), 128.63, 128.44, 128.43, 128.37, 128.34, 128.31,
128.16, 128.14, 128.08, 128.06, 127.95, 127.88, 127.71, 127.64, 127.28,
127.19, 127.07, 126.96, 126.28 (CHPh), 99.66, 98.22, 96.17 (C-1), 84.17,
81.16, 81.12, 80.49, 80.42, 79.25, 79.20, 77.91, 76.77, 76.34, 75.95, 74.16,
73.66, 73.37, 72.83, 72.56, 71.38, 70.63, 70.03, 68.73 (CH2, CH) ppm;
HRMS: calcd for C148H152O32Na: 2464.0164; found 2464.0820.
Experimental Section
General: Solvents were distilled under anhydrous conditions. All re-
agents were used as purchased without further purification. Evaporation
was carried out on a rotary evaporator with the temperature kept below
408C. Glassware used for water-free reactions was dried (min. 2 h) at
1308C before use. Columns were packed with silica gel 60 (230–400
mesh) as the stationary phase. TLC plates (Merck 60, F254) were visual-
ised by sprayingwith cerium sulfate (1%) and molybdic acid (1.5%) in
10% H2SO4 and heatinguntil coloured spots appeared. 1H NMR, 13C
NMR and COSY were carried out on a Varian Mercury 400 instrument.
Mass spectra (MALDI MS) were obtained on a Voyager DE PRO mass
spectrometer (Applied Biosystems) with use of an a-cyanohydroxycin-
namic acid (a-CHCA) matrix. Spectra were calibrated with angiotensin I
(m/z 1296.69), adrenocorticotropic hormone (ACTH) (clip 1–17; m/z
2093.09), ACTH (clip 18–39; m/z 2465.20), and ACTH (clip 7–38; m/z
3657.93).
a-Cyclodextrin-6A,6D-dicarboxylic acid (1): Compound 10 (2.34 g,
0.96 mmol) was dissolved in an AcOEt/MeOH mixture (1:1, 150 mL).
Pd/C (250 mg) and TFA (cat.) were then added and the mixture was stir-
red overnight under hydrogen. Filtration through Celite and evaporation
of the solvent gave compound 1 (0.95 g, 100%) as a white foam: [a]D
=
+103.6 (c = 1.0, H2O); 1H NMR (400 MHz, D2O): d = 4.99 (d, J =
3.6 Hz, 2H; H-1), 4.96 (d, J = 3.6 Hz, 2H; H-1), 4.90 (d, J = 3.6 Hz,
2H; H-1), 4.02 (d, J = 9.2 Hz, 4H), 3.92–3.66 (m, 18H), 3.62–3.38 (m,
10H) ppm; 13C NMR (100 MHz, D2O): d
= 172.36 (C=O), 101.97,
101.80, 100.34 (C-1), 81.75, 81.08, 81.03, 73.29, 72.93, 72.72, 72.03, 71.82,
71.75, 71.62, 71.33, 71.23, 60.36, 59.95 (CH2, CH) ppm; HRMS: calcd for
C36H56O32Na: 1023.2652; found 1023.2448.
General procedure for de-O-benzylation of perbenzylated cyclodextrins
Synthesis of monool: Molecular sieves (4 Š, 1 g) were added at room
temperature under nitrogen to a solution of per-O-benzyl-a-cyclodextrin
(100 mg, 0.04 mmol) in anhydrous toluene (12 mL), and the system was
stirred for 1 h. After that, DIBAL (1.17 mL, 1.17 mmol, 1.0m in toluene)
was added dropwise. The reaction mixture was stirred at room tempera-
ture for 1 h. The mixture was cooled to 08C, water (10 mL) was carefully
added dropwise, and the mixture was stirred vigorously at room tempera-
ture for 15 min. It was diluted with EtOAc (10 mL) and filtered, with
washingwith EtOAc (3”10 mL). The oragnic layer was washed with
brine (2”10 mL), dried (MgSO4) and filtered, and the organic solvent
was removed in vacuo. The residue was purified by chromatography
(eluent gradient, EtOAc/pentane 1:5!1:2), to afford the monool (62 mg,
63%), the diol (19 mg, 20%) and starting material (16 mg, 16%). The as-
signment is in agreement with the assignment in Ref. [20].
Nonadeca-O-benzyl-b-cyclodextrin-6A,6D-dicarboxylic acid (11): The
Dess–Martin periodinane reagent (1.95 g, 4.60 mmol) was added to a so-
lution of diol 9 (2.6 g, 0.92 mmol) in CH2Cl2 (100 mL), and the reaction
mixture was stirred at room temperature for 4 h and then quenched by
addition of Et2O (100 mL) and saturated aqueous NaHCO3 containing
3 gof Na 2S2O3 (100 mL). After havingbeen stirred for an additional 2 h
the solution was diluted with Et2O (150 mL) and washed successively
with saturated aqueous NaHCO3 (50 mL) and water (50 mL). The organ-
ic phase was dried and concentrated.
NaClO2 (1.39 g, 15.37 mmol) and NaH2PO4 (0.8 g) in water (22 mL) were
added to a solution of the residue in tBuOH (53 mL), THF (22 mL) and
2-methylbut-2-ene (22 mL). The reaction mixture was stirred overnight
and then quenched with aqueous HCl (1m, 100 mL) and extracted with
EtOAc (4”20 mL). The organic extracts were dried and concentrated.
The remainingoil was purified by column chromatography on silica gel
(eluent, EtOAc/pentane 3:10, containing1% HCOOH), to provide com-
Synthesis of diol: Molecular sieves (4 Š, 50 g) were added at room tem-
perature under nitrogen to a solution of per-O-benzyl-b-cyclodextrin (5 g,
1.65 mmol) in anhydrous toluene (270 mL), and the system was stirred
for 1 h. After that, DIBAL (66 mL, 66.4 mmol, 1.0m in toluene) was
added dropwise. The reaction mixture was stirred for 21 h at room tem-
perature and was then cooled to 08C, water (30 mL) was carefully added
dropwise, and the mixture was stirred vigorously at room temperature
for 15 min. The mixture was diluted with EtOAc (50 mL) and filtered,
with washingwith EtOAc (3”100 mL). The organic layer was washed
with brine (2”75 mL), dried (MgSO4) and filtered, and the organic sol-
vent was removed in vacuo. The residue was purified by chromatography
(eluent gradient, EtOAc/pentane 1:5!1:2), to afford diol (3.79 g, 81%)
as a white foam. The assignment is in agreement with the assignment in
Ref. [20].
Hexadeca-O-benzyl-a-cyclodextrin-6A,6D-dicarboxylic acid (10): The
Dess–Martin periodinane reagent (0.97 g, 2.30 mmol) was added to a so-
lution of diol 8 (1.1 g, 0.46 mmol) in CH2Cl2 (50 mL), and the reaction
mixture was stirred at room temperature for 4 h and then quenched by
addition of Et2O (50 mL) and saturated aqueous NaHCO3 containing3 g
of Na2S2O3 (75 mL). After havingbeen stirred for an additional 2 h the
solution was diluted with Et2O (100 mL) and washed successively with
saturated aqueous NaHCO3 (50 mL) and water (50 mL). The organic
phase was dried and concentrated.
1
pound 11 (2.2 g, 84%) as a white foam: [a]D +30.7 (c = 1.0, CHCl3); H
NMR (400 MHz, CDCl3): d = 7.26–6.88 (m, 95H; CHPh), 5.78 (d, J =
4.0 Hz, 1H; H-1), 5.68 (d, J = 3.2 Hz, 1H; H-1), 5.43 (d, Jgem = 10.4 Hz,
ꢁ
ꢁ
1H; H CHPh), 5.37 (d, Jgem = 10.0 Hz, 1H; H CHPh), 5.16 (d, Jgem
=
ꢁ
ꢁ
11.2 Hz, 1H; H CHPh), 5.12 (d, Jgem = 10.8 Hz, 1H; H CHPh), 5.02 (d,
ꢁ
Jgem = 11.2 Hz, 2H; H CHPh), 4.92–4.80 (m, 6H), 4.78–4.46 (m, 20H),
4.43–4.26 (m, 14H), 4.22–3.78 (m, 22H), 3.73 (d, J = 11.6 Hz, 1H), 3.67
(d, J = 11.2 Hz, 1H), 3.60–3.47 (m, 4H), 3.43–3.15 (m, 8H) ppm; 13C
NMR (100 MHz, CDCl3): d
= 171.96, 171.89 (C=O), 139.66, 139.34,
139.28, 139.21, 139.15, 138.86, 138.65, 138.56, 138.47, 138.31, 138.28,
138.20, 138.13, 137.79, 137.73, 136.90, 136.67 (Cipso), 128.69, 128.47,
128.40, 128.37, 128.33, 128.28, 128.22, 128.18, 128.08, 128.05, 127.98,
127.93, 127.86, 127.76, 127.69, 127.65, 127.60, 127.57, 127.37, 127.25,
127.10, 127.02, 127.00, 126.95, 126.81, 126.53, 126.47 (CHPh), 100.79,
100.18, 98.84, 97.90, 97.24, 96.17, 95.76 (C-1), 83.23, 82.37, 81.07, 80.44,
80.24, 79.74, 79.22, 78.16, 77.73, 76.53, 76.25, 75.61, 74.55, 74.15, 73.37,
73.13, 72.93, 72.73, 72.66, 72.59, 71.29, 70.74, 70.57, 70.44, 69.38, 68.95,
60.46 (CH2, CH) ppm; HRMS: calcd for C175H180O37K: 2912.1840; found
2912.9507.
b-Cyclodextrin-6A,6D-dicarboxylic acid (2):[18–20] Compound
9 (2.2 g,
NaClO2 (0.78 g, 8.62 mmol) and NaH2PO4 (0.36 g) in water (11 mL) were
added to a solution of the residue in tBuOH (27 mL), THF (11 mL) and
2-methylbut-2-ene (11 mL). The reaction mixture was stirred overnight
and then quenched with aqueous HCl (1m, 75 mL) and extracted with
EtOAc (4”20 mL). The organic extracts were dried and concentrated,
and the remainingoil was purified by column chromatography on silica
gel (eluent, EtOAc/pentane 1:5, containing 1% HCOOH), to provide
compound 10 (0.92 g, 86%) as a white foam: [a]D =+21.5 (c = 1.1,
0.76 mmol) was dissolved in an AcOEt/MeOH mixture (1:1, 150 mL).
Pd/C (300 mg) and TFA (cat.) were then added and the mixture was stir-
red overnight under hydrogen. Filtration through Celite and evaporation
of the solvent gave compound 1 (0.9 g, 100%) as a white solid: [a]D
=
+112.3 (c = 1.0, H2O); 1H NMR (400 MHz, D2O): d = 5.05 (d, J =
3.6 Hz, 1H; H-1), 5.03 (d, J = 3.6 Hz, 1H; H-1), 5.00–4.93 (m, 5H; H-1),
4.18 (brd,
J = 8.0 Hz, 2H), 3.90–3.59 (m, 26H), 3.57–3.42 (m,
Chem. Eur. J. 2005, 11, 5094 – 5101
www.chemeurj.org
ꢁ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
5099

M. Bols et al.
10H) ppm; 13C NMR (100 MHz, D2O): d
=
171.83 (C=O), 101.93,
16 (205 mg, 96%) as a white foam: [a]D
=
+93.8 (c = 1.0, H2O); 1H
101.86, 101.68, 101.58 (C-1), 81.87, 81.81, 81.39, 81.20, 81.01, 80.48, 73.24,
73.14, 72.95, 72.90, 72.25, 72.17, 71.79, 71.73, 71.62, 60.17, 59.51 (CH2,
CH) ppm; HRMS: calcd for C42H66O37Na: 1185.3181; found 1185.3190.
NMR (400 MHz, D2O): d = 5.02–4.91 (m, 7H; H-1), 4.30–4.14 (m, 4H),
4.00–3.92 (m, 2H), 3.89–3.68 (m, 23H), 3.59–3.42 (m, 15H) ppm; 13C
NMR (100 MHz, D2O): d = 102.08, 101.97 (C-1), 81.35, 81.17, 80.97,
80.91, 80.86, 73.28, 73.15, 72.29, 72.11, 71.94, 71.85, 70.14, 67.22, 60.29,
60.14 ppm; HRMS: calcd for C42H68O41S2Na2: 1338.247; found 1315.375
[MꢁNa], 1213.417 [Mꢁ(SO3Na2+H)].
Eicosa-O-benzyl-b-cyclodextrin-6-carboxylic acid (13): The Dess–Martin
periodinane (255 mg, 0.60 mmol) was added to a solution of alcohol 12
(0.57 g, 0.20 mmol) in CH2Cl2 (20 mL), and the reaction mixture was stir-
red at room temperature for 1 h and then quenched by addition of Et2O
Tetradeca-O-acetyl-b-cyclodextrin-6A,6B,6C,6D,6E,6F,6G-heptasulfate
sodium salt (18): Sulfur trioxide pyridine complex (1.62 g, 10.18 mmol)
was added to a solution of 17 (500 mg, 0.29 mmol) in dry DMF (15 mL).
After the system had been stirred overnight at 408C the solvent was re-
moved in vacuo. The remainingoil was purified by column chromatogra-
phy on silica gel (eluent CH2Cl2/MeOH 10:3, containing2% Et 3N) and
column chromatography on IR-120 (sodium form) to give heptasulfate 18
(20 mL) and saturated aqueous NaHCO3 containing0.6 gNa
2S2O3
(20 mL). After havingbeen stirred for an additional 30 min, the solution
was diluted with Et2O (50 mL) and washed successively with saturated
aqueous NaHCO3 (20 mL) and water (20 mL). The organic phase was
dried (MgSO4) and concentrated.
The residue was dissolved in tBuOH (14 mL), THF (6 mL) and 2-methyl-
but-2-ene (5 mL), and a solution of NaClO2 (0.36 mg, 4.0 mmol) and
NaH2PO4 (0.4 g) in water (3 mL) was added. The reaction mixture was
stirred overnight, and was then quenched with aqueous HCl (1m, 10 mL)
and extracted with EtOAc (3”10 mL). The organic extracts were dried
(MgSO4) and concentrated. The remainingoil was purified by column
chromatography on silica gel (eluent gradient, EtOAc/pentane 1:5, then
EtOAc/pentane 2:5, containing1% HCOOH), to provide monoacid 13
1
(650 mg, 92%) as a white foam: [a]D = +48.0 (c = 1.0, H2O); H NMR
(400 MHz, D2O): d = 5.29 (t, J2,3 = J3,4 = 8.8 Hz, 7H; H-3), 5.20 (d, J1,
= 3.6 Hz, 7H; H-1), 4.81 (dd, 7H; H-2), 4.41 (brd, Jgem = 14.4 Hz, 7H;
H-6a), 4.26 (brd, 7H; H-6b), 4.11 (brd, 7H; H-5), 4.00 (t, J4,5 = 8.8 Hz,
7H; H-4), 2.04 and 2.03 (s, 42H; CH3CO) ppm; 13C NMR (100 MHz,
D2O): d = 173.24, 173.16 (C=O), 96.63 (C-1), 75.04, 71.32, 71.07, 70.26
(C-2, C-3, C-4, C-5), 66.82 (C-6), 20.78, 20.62 (CH3CO) ppm; HRMS:
calcd for C70H91O70S7Na7: 2436.0889; found 2255.2012 [Mꢁ
(SO3Na)2+2H+Na].
b-Cyclodextrin-6A,6B,6C,6D,6E,6F,6G-heptasulfate sodium salt (19): An
aqueous solution of sodium hydroxide (4m, 5 mL) was added to a solu-
tion of 18 (600 mg, 0.24 mmol) in MeOH (10 mL). After stirring over-
night at room temperature, the reaction mixture was diluted with water
(10 mL) and Amberlite IR-120 (H+ form) until pH neutral, and the mix-
ture was filtered and concentrated in vacuo. The remainingoil was puri-
fied by column chromatography on IR-120 (sodium form) to give hepta-
(450 mg, 75%) as a white foam: [a]D
NMR (400 MHz, CDCl3): d = 7.28–6.90 (m, 100H; CHPh), 5.61 (d, J =
3.6 Hz, 1H; H-1), 5.55 (d, 3.6 Hz, 1H; H-1), 5.30–3.20 (m,
86H) ppm; 13C NMR (100 MHz, CDCl3): d
171.64 (C=O), 139.70,
=
+34.1 (c = 1.0, CHCl3); 1H
J
=
=
139.56, 139.44, 139.22, 139.05, 138.78, 138.65, 138.48, 138.41, 138.35,
138.26, 137.93, 136.89 (Cipso), 128.63, 128.59, 128.40, 128.30, 128.26,
128.15, 128.06, 127.89, 127.73, 127.63, 127.51, 127.46, 127.41, 127.22,
127.05, 126.90, 126.73 (CHPh), 99.60, 99.52, 98.89, 97.82, 97.57, 97.39 96.02
(C-1), 81.08, 80.92, 79.05, 73.56, 73.43, 73.26, 73.14, 72.93, 72.77, 72.61,
71.61, 69.10 (CH2, CH) ppm; HRMS: calcd for C182H188O36Na: 2972.2777;
found 2973.4368.
sulfate 19 (400 mg, 88%) as a white foam: [a]D
= +88.3 (c = 1.0,
H2O); 1H NMR (400 MHz, D2O): d = 5.09 (d, J1,2 = 3.6 Hz, 7H; H-1),
4.28 (brd, Jgem = 11.2 Hz, 7H; H-6a), 4.20 (brd, 7H; H-6b), 4.03 (brd,
7H; H-5), 3.90 (t, J2,3 = J3,4 = 9.5 Hz, 7H; H-3), 3.61 (t, J4,5 = 9.5 Hz,
7H; H-4), 4.81 (dd, 7H; H-2) ppm; 13C NMR (100 MHz, D2O): d =
101.05 (C-1), 79.40, 73.02, 72.06, 69.78 (C-2, C-3, C-4, C-5), 66.85 (C-
6) ppm. HRMS: calcd for C42H63O56S7: 1687.012; found 1631.206
[MꢁSO3+H+Na].
b-Cyclodextrin-6-monocarboxylic acid (14): Compound 13 (0.36 g,
0.12 mmol) was dissolved in an EtOAc/MeOH mixture (1:1, 60 mL). Pd/
C (35 mg) and TFA (cat,) were then added and the mixture was stirred
overnight under hydrogen. Filtration through Celite and evaporation of
the solvent gave compound 14 (0.13 g, 94%) as a white solid: [a]D =+
47.1 (c = 1.0, CHCl3); 1H NMR (400 MHz, H2O): d = 5.00–4.85 (m,
7H; H-1), 4.80–4.50 (m, 25H), 3.90–3.4 (m, 36H) ppm; 13C NMR
(100 MHz, H2O): d = 163.0 (C=O), 102.05, 102.01, 101.93, 101.74 (C-1),
82.24, 81.21, 80.20, 81.08, 73.28, 73.17, 72.51, 72.28, 72,17, 71.95, 71.89,
71.82, 71.68, 60.31 (CH2, CH) ppm; HRMS: calcd for C42H68O36Na:
1171.3388; found 1171.3064.
Nonadeca-O-benzyl-b-cyclodextrin-6A,6D-disulfate sodium salt (15):
Sulfur trioxide pyridine complex (279 mg, 1.75 mmol) was added to a so-
lution of diol 9 (500 mg, 0.17 mmol) in dry DMF (10 mL). After the
system had been stirred for 3 h at 408C the solvent was removed in
vacuo. The remainingoil was purified by column chromatorgaphy on
silica gel (eluent, CH2Cl2/MeOH 10:1) to give disulfate 15 (500 mg, 93%)
as a white foam: [a]D =+29.8 (c = 1.0, CHCl3); 1H NMR (400 MHz,
(CD3)2SO, 708C): d = 7.42–6.83 (m, 95H; CHPh), 5.57 (d, J = 2.8 Hz,
2H; H-1), 5.34 (d, J = 2.8 Hz, 1H; H-1), 5.23 (d, J = 2.8 Hz, 3H; H-1),
5.17 (d, J = 2.8 Hz, 1H; H-1), 5.09 (t, J = 12.0 Hz, 2H), 4.96–4.77 (m,
4H), 4.74–4.24 (m, 33H), 4.19–3.66 (m, 30H), 3.62–3.50 (m, 4H), 3.48–
3.22 (m, 8H) ppm; 13C NMR (100 MHz, (CD3)2SO): d = 139.85, 139.79,
139.15, 139.68, 139.61, 139.52, 139.13, 139.07, 139.00, 138.95, 138.85
(Cipso), 128.85, 128.71, 128.63, 128.56, 128.49, 128.44, 128.41, 128.07,
128.02, 127.97, 127.94, 127.85, 127.75, 127.64 (CHPh), 98.33, 97.85, 97.53,
96.63 (C-1), 81.51, 81.14, 80.68, 79.58, 78.85, 78.37, 77.51, 76.56, 75.70,
75.67, 75.48, 75.24, 74.97, 74.59, 73.14, 73.05, 72.96, 72.84, 72.72, 72.63,
72.41, 72.27, 71.96, 71.58, 69.61, 69.44, 65.75 (CH2, CH) ppm; HRMS:
calcd for C175H182O41S2Na3: 3072.1291; found 3072.5337
Procedure for determining the rate of hydrolysis: Each assay was per-
formed on 2 mL samples prepared from aqueous solutions (1 mL) of the
appropriate nitrophenyl glycoside at different concentrations mixed with
phosphate buffer (0.1m, 1 mL) containing 1, 2, 14, 16 or 19 (5 or 15 mg)
or nothing(control). For pH 9.0 a 0.1 m borate buffer was used. The reac-
tions were followed at 598C by UV absorption at 400 nm and were typi-
cally monitored for 18 h. Velocities were determined as the slope of the
progress curve of each reaction. Uncatalysed velocities were obtained di-
rectly from the control samples. Catalysed velocities were calculated by
subtractingthe uncatalysed velocity from the velocity of the appropriate
cyclodextrin-containingsample. The catalysed velocities were used to
construct Haines plots ([S]/V versus [S]) from which Km and Vmax were
determined. kcat was calculated as Vmax/[cyclodextrin], kuncat was deter-
mined as the slope from a plot of Vuncat versus [S]. The followingextinc-
tion coefficients were determined for 4-nitrophenolate and used in the
calculations:
e
=
17.04 mmꢁ1 cmꢁ1 (pH 8.0, 598C), 16.9 mmꢁ1 cmꢁ1
(pH 7.9, 598C), 15.5 mmꢁ1 cmꢁ1 (pH 7.9, 258C), 15.3 mmꢁ1 cmꢁ1 (pH 7.4,
598C), 10.4 mmꢁ1 cmꢁ1 (pH 6.8, 598C), 7.41 mmꢁ1 cmꢁ1 (pH 6.8, 258C),
2.07 mmꢁ1 cmꢁ1 (pH 5.9, 598C) and 0.96 mmꢁ1 cmꢁ1 (pH 5.9, 258C). The
inhibition experiments were made by addingcyclopentanol or aniline
(15 mL) to a catalysed sample. Determination of glucose was made by use
of the Sigma Glucose Assay Kit GAGO-20, based on glucose oxidase-
catalysed oxidation of glucose to gluconic acid and hydrogen peroxide,
and subsequent determination of the hydrogen peroxide by peroxidase-
catalysed oxidation of o-dianisidine and absorption measurement at
540 nm.
b-Cyclodextrin-6A,6D-disulfate sodium salt (16): Compound 15 (500 mg,
0.16 mmol) was dissolved in an EtOH/H2O mixture (1:1, 60 mL).
Pd(OH)2 (300 mg) and TFA (cat.) were then added and the mixture was
stirred overnight under hydrogen. After filtration through Celite and
evaporation of the solvent, the remainingoil was subjected to column
chromatography on Amberlite IR-120 (sodium form) to give compound
5100
ꢁ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 5094 – 5101

Artificial Glycosyl Phosphorylases
FULL PAPER
[14] D. L. Zechel, S. G. Withers, Acc. Chem. Res. 2000, 33, 11–18.
[15] T. Ohe, Y. Kajiwara, T. Kida, W. Zhang, Y. Nakatsuji, I. Ikeda,
Chem. Lett. 1999, 921–922.
[16] T. H. Doug, J. Z. Chou, X. Huang, A. J. Bennet, J. Chem. Soc.
Perkin Trans. 2 2001, 83–89.
Acknowledgements
We thank the Lundbeck foundation and the SNF for financial support, as
well as the Ramón Areces foundation (F.O.C.).
[17] C. Rousseau, N. Nielsen, M. Bols, Tetrahedron Lett. 2004, 45, 8709–
8711.
[18] C. Fraschini, M. R. Vignon, Carbohydr. Res. 2000, 328, 585–589.
[19] J. Yoon, S. Hong, K. A. Martin, A. W. Czarnik, J. Org. Chem. 1995,
60, 2792–2795.
[20] Y. Kuroda, O. Kobayashi, Y. Suzuki, H. Ogoshi, Tetrahedron Lett.
1989, 30, 7225–7228.
[21] A. J. Pearce, P. Sinaþ, Angew. Chem. 2000, 112, 3756–3758; Angew.
Chem. Int. Ed. 2000, 39, 3610–3612.
[22] T. Lecourt, A. Herault, A. J. Pearce, M. Sollogoub, P. Sinaþ, Chem.
Eur. J. 2004, 10, 2960–2971.
[23] M. H. Clausen, M. R. Jørgensen, J. Thorsen, R. Madsen, J. Chem.
Soc. Perkin Trans. 1 2001, 543–551.
[24] J.-C. Jacquinet, Carbohydr. Res. 2004, 339, 349–359.
[25] K. Takeo, H. Mitoh, K. Uemura, Carbohydr. Res. 1989, 187, 203–
221.
[1] R. Wolfenden, Acc. Chem. Res. 2001, 34, 938–945.
[2] R. Breslow, S. D. Dong, Chem. Rev. 1998, 98, 1997–2011.
[3] A. J. Kirby, Angew. Chem. 1994, 106, 573–576; Angew. Chem. Int.
Ed. Engl. 1994, 33, 551–553.
[4] Y. Murakami, J. I. Kikuchi, Y. Hisaeda, O. Hayashida, Chem. Rev.
1996, 96, 721–758.
[5] W. B. Motherwell, M. J. Bingham, Y. Six, Tetrahedron 2001, 57,
4663–4686.
[6] R. Breslow, Acc. Chem. Res. 1995, 28, 146–153.
[7] R. Breslow, C. Schmuck, J. Am. Chem. Soc. 1996, 118, 6601–6605.
[8] R. Breslow, B. Zhang, J. Am. Chem. Soc. 1992, 114, 5882–5883.
[9] T. Akiike, Y. Nagano, Y. Yamamoto, A. Nakamura, H. Ikeda, A.
Veno, F. Toda, Chem. Lett. 1994, 1089–1092.
[10] R. Breslow, B. Zhang, J. Am. Chem. Soc. 1994, 116, 7893–7894.
[11] N. M. Milovic, J. D. Badjic, N. M. Kostic, J. Am. Chem. Soc. 2004,
126, 696–697.
[12] M. V. Rekharsky, Y. Inoue, Chem. Rev. 1998, 98, 1875–1917.
[13] A. R. Khan, P. Forgo, K. J. Stine, V. T. D’Souza, Chem. Rev. 1998,
98, 1977–1996.
Received: March 31, 2005
Published online: June 27, 2005
Chem. Eur. J. 2005, 11, 5094 – 5101
www.chemeurj.org
ꢁ 2005 Wiley-VCH VerlagGmbH & Co. KGaA, Weinheim
5101