Synthesis of a novel class of cyclodextrin-based nanotubes

Article abstract of DOI:10.1021/ol201065u

The synthesis and characterization of a novel class of structurally well-defined nanotubes from β-cyclodextrin are described. These new hosts were formed using disulfide linkages that substitute all the primary hydroxyl groups of a β-cyclodextrin. A deep and rigid hydrophobic channel with a size of more than 1.5 nm is found in the molecules. Because of their unique geometry and the potential biodegradability of the disulfide bond, this class of molecules could find broad applications in biology and other areas of research.

Full text of DOI:10.1021/ol201065u

ORGANIC
LETTERS
2011
Vol. 13, No. 14
3572–3575
Synthesis of a Novel Class of
Cyclodextrin-Based Nanotubes
Aixia Wang, Wenling Li,† Ping Zhang, and Chang-Chun Ling*
Alberta Ingenuity Centre for Carbohydrate Science, Department of Chemistry,
University of Calgary, 2500 University Drive NW, Calgary, Alberta T2N 1N4, Canada
ccling@ucalgary.ca
Received April 22, 2011
ABSTRACT
The synthesis and characterization of a novel class of structurally well-defined nanotubes from β-cyclodextrin are described. These new hosts
were formed using disulfide linkages that substitute all the primary hydroxyl groups of a β-cyclodextrin. A deep and rigid hydrophobic channel
with a size of more than 1.5 nm is found in the molecules. Because of their unique geometry and the potential biodegradability of the disulfide
bond, this class of molecules could find broad applications in biology and other areas of research.
Cyclodextrins (CDs)1 attract attention from different
research fields because of their unique hydrophobic cav-
ities which can host a range of organic compounds via
noncovalent interaction. The building blocks of all natural
CDs consist of the same R-glucopyranosyl unit which
forms the wall of all cavities; therefore all CD cavities have
potential to interact with guest molecules with high affinity
and specificity by taking advantage of the cooperative
effect.4,5 Following Tabushi and Harada’s pioneer work,6
the syntheses of CD-based dimers,7 trimers,8 tetramers,9
(4) Liu, Y.; Chen, Y. Acc. Chem. Res. 2006, 39, 681.
(5) (a) Breslow, R.; Halfon, S.; Zhang, B. Tetrahedron 1995, 51, 377
and references therein. (b) Breslow, R.; Chung, S. J. Am. Chem. Soc.
1990, 112, 9659.
(6) (a) Tabushi, I.; Kuroda, Y.; Shimokawa, K. J. Am. Chem. Soc.
1614, 1979, 101. (b) Harada, A.; Furue, M.; Nozakura, S.-I. Polym. J.
1980, 12, 29.
(7) (a) Breslow, R.; Greenspoon, N.; Guo, T.; Zarzycki, R. J. Am.
Chem. Soc. 1989, 111, 8296. (b) Yuan, D.-Q.; Immel, S.; Koga, K.;
Yamaguchi, M.; Fujita, K. Chem.;Eur. J. 2003, 9, 3501. (c) Brady, B.;
Darcy, R. Carbohydr. Res. 1998, 309, 237. (d) de Jong, M. R.; Engbersen,
J. F. J.; Huskens, J.; Reinhoudt, D. N. Chem.;Eur. J. 2000, 6, 4034. (e)
Liu, Y.; Chen, Y.; Liu, S.-X.; Guan, X.-D.; Wada, T.; Inoue, Y. Org.
Lett. 2001, 3, 1657. (f) Dong, D.; Baigl, D.; Cui, Y.; Sinay, P.; Sollogoub,
M.; Zhang, Y. Tetrahedron 2007, 63, 2973. (g) Aime, S.; Gianolio, E.;
Palmisano, G.; Robaldo, B.; Barge, A.; Boffa, L.; Cravotto, G. Org.
Biomol. Chem. 2006, 4, 1124. (h) Fujita, K.; Ejima, S.; Imoto, T. Chem.
Commun. 1984, 1277. (i) Okabe, Y.; Yamamura, H.; Obe, K.; Ohta, K.;
Kawai, M.; Fujita, K. Chem. Commun. 1995, 581. (j) Breslow, R.;
Greenspoon, N.; Guo, T.; Zarycki, R. J. Am. Chem. Soc. 1989, 121,
8296. (k) Chiu, S.-H.; Myles, S. C.; Garrell, R. L.; Stoddart, J. F. J. Org.
Chem. 2000, 65, 2792.
˚
the same height (7.9 A). However, the cavityvolumes differ
according to the number of R-glucopyranosyl units in-
volved in forming the host molecule. For example, R-CD
constitutes six sugar units, generating a cavity with a fixed
3
˚
volume of 174 A . The other two homologues, the β- and
γ-CDs, which are formed with 7 and 8 sugar units, produce
3
larger but still fixed cavity volumes of 262 and 424 A ,
˚
respectively. There has been considerable interest in devel-
oping chemistry to create novel hosts with increased
volumes to augment CD’s inclusion capability. Among
the many strategies reported in the literature,2 oligomer-
ization of CD monomers3 using linkers of variable nature
is the most attractive because this creates linear or
branched CD hosts with multiple cavities that have the
(8) (a) Leung, D. K.; Atkins, J. H.; Breslow, R. Tetrahedron Lett.
2001, 42, 6255. (b) Okabe, Y.; Yamamura, M.; Obe, K.; Ohta, K.;
Kawai, M.; Fujita, K. Chem. Commun. 1995, 581. (c) Liu, Y.; Li, L.; Li,
X.-Y.; Zhang, H.-Y.; Wada, T.; Inoue, Y. J. Org. Chem. 2003, 68, 3646.
(d) Sasaki, K.; Nagasaka, M.; Kuroda, Y. Chem. Commun. 2001, 2630.
(9) (a) Jiang, T.; Li, M.; Lawrence, D. S. J. Org. Chem. 1995, 60, 7293.
(b) Breslow, R.; Zhang, X.; Xu, R.; Maletic, M.; Merger, R. J. Am.
† Present address: School of Chemical and Biological Engineering,
Lanzhou Jiaotong University, Lanzhou, Gansu 730070, P. R. of China.
(1) Szeitli, J. Cyclodextrin Technology; Kluwer Academic Publishers:
Dordrecht, 1988.
(2) Engeldinger, E.; Armspach, D.; Matt, D. Chem. Rev. 2003, 103,
4147.
ꢀ
ꢀ
(3) For a review, see: Sliwa, W.; Girek, T.; Koziol, J. J. Curr. Org.
Chem. 2004, 8, 1445.
Chem. Soc. 1996, 118, 11678. (c) Lecourt, T.; Bleriot, Y.; Auzely-Velty,
R.; Sollogoub, M. Chem. Commun. 2010, 46, 2238.
r
10.1021/ol201065u
Published on Web 06/13/2011
2011 American Chemical Society

and pentamers10 have been reported. The design and
synthesis of such oligomeric CD hosts can be challenging
depending on the chemical properties of linkers. Probably
one of the simplest approaches is to use disulfide as a linker
to bring together two CD molecules by taking advantage
of the thiolꢀdisulfide exchange which is a fundamental
biological process. This method has been explored pre-
viously to prepare different CD-based “head to head”,
“head to tail”, and “tail to tail” dimers (head: primary face,
tail: secondary face) using one disulfide linkage.7hꢀj Re-
cently, Kraus’s group succeeded in preparing two duplexes
of R-CD (Figure 1) using two and three disulfide linkages
respectively.11 They strategically placed the sulfur atoms at
the 6A,6D or 6A,6C,6E positions of R-CD and obtained
dimers 1 and 2 in high yields.
succeeded. Because of their unique molecular topology, as
well as of the presence of potentially biodegradable dis-
ulfide linkages,13 this class of host molecules could have
widespread applications in biology and areas involving
molecular recognition.
The challenge to obtain such molecules is that the
dimerization/thiolꢀdisulfide exchange process could be
very complex because of numerous possibilities to form
disulfide bonds between two thiol groups located within
the same CD molecule as well as across two CD molecules.
In addition, the monomer can oligo- and polymerize to
form a complex mixture. Indeed, recently, Chechik’s group
has studied the polymerization of a β-CD derived hep-
tathiol (5) and they obtained large water-soluble nanocap-
sules cross-linked by disulfide linkages.14 According to the
molecular model and Kraus’s work,11a the formation of
disulfide linkages within the same CD molecules usually
results in considerable distortion of the CD macrocycle,
thus this would lead to the formation of CD intermediates
with a higher energy owing to the presence of excessive
amounts of torsional strains. If the total number of thiol
groups within the same CD is an odd number, under basic
conditions, this would result in an unpaired thiolate at any
time, which should serve as a driving force to promote
thiolꢀdisulfide equilibrium, and ultimately, the intramo-
lecular disulfide bonds would break in favor of the less
strained intermolecular disulfide linkages. We believe that,
after the formation of the first intermolecular disulfide
linkage between two CDs, if a second disulfide bond could
form between the pair, this would align the two CD
molecules in the “head to head” orientation; this could
promote the formation of other disulfide bonds between
the two molecules; ultimately, the thiolꢀdisulfide equili-
brated process would favor the formation of the desired
dimer, which should possess the least amount of torsional
strains.
From our previous research, we had access to periodi-
nated β-CD15 (3, Scheme 1). We thought that compound 3
could be a good substrate to use because it can be con-
verted to a molecule containing 7 thiol groups. Since the
molecule had an odd number of thiols, it should be an ideal
substrate to explore the scope of the thiolꢀdisulfide
equilibrium reaction. As shown in Scheme 1, we first
replaced all iodides of compound 3 with potassium
thioacetate in DMF. The intermediate thioacetate
was not isolated but subjected to an peracetylation
to afford the fully acetylated intermediate 4, which
was obtained in 76% yield by column chromatography
on silica gel. To test the dimerization reaction, the perace-
Figure 1. Structures of previously synthesized CD duplexes 1
and 2 containing 2ꢀ3 disulfide bonds and CPK models12 of
targeted duplexes fully substituted with disulfides.
One of the obvious advantages of using more than one
linker to bridge CD molecules is that the resulting hosts
such as 1 and 2 possess higher rigidity compared to those
containing only one bridge. However, for incompletely
substituted hosts like 1 and 2, there are still some polar
hydroxyl groups remaining, which make the middle zone
of the molecules still hydrophilic. One of the most exquisite
synthetic targets would be to replace all hydroxyl groups
with disulfide linkages (Figure 1); this would allow us to
obtaina novelclass ofmoleculeswhichare fully enclosed in
the middle, thus effectively connecting the two hydropho-
bic cavities with an additional hydrophobic zone created
by the fully encircled disulfide bonds. Molecular modeling
showed that such molecules would possess a genuine tube
which has a remarkable depth of more than 1.5 nm. To
date, the synthesis of such a class of molecules has not yet
(10) Rawal, G. K.; Zhang, P.; Ling, C.-C. Org. Lett. 2010, 12, 3096.
(11) (a) Kumprecht, L.; Budesinsky, M.; Vondrasek, J.; Vymetal, J.;
Cisarova, I.; Brynda, J.; Herzig, V.; Koutnik, P.; Zavada, J.; Cerny, J.;
Kraus, T. J. Org. Chem. 2009, 74, 1082. (b) Krejci, L.; Budesinski, M.;
Cisarova, I.; Kraus, T. Chem. Commun. 2009, 3557. (c) Kumprecht, L.;
Budesinsky, M.; Bour, P.; Kraus, T. New J. Chem. 2010, 34, 2254.
(12) β-CD coordinates were obtained from the protein databank
(http://www.pdb.org, 3M4E). All 6-OH’s were replaced with thiols, and
the resulting structure was minimized with an MM2 force field. The
heptathiol was manually dimerized by forming 7 disulfide linkages, and
the obtained dimer was minimized again. The obtained model has the
following average dihedral angles: C4ꢀC5ꢀC6ꢀS: ꢀ165°; C5ꢀC6ꢀ
SꢀS0: 165°; C6ꢀSꢀS0ꢀC60: 115°; SꢀS0ꢀC60ꢀC50: 165°; S0ꢀC60ꢀC50ꢀ
C40: ꢀ165°.
ꢀ
tylated heptathiol 4 was first subjected to a Zemplen
transesterification condition in methanol using sodium
methoxide as the base under an argon atmosphere; the
deacetylation finished within the first 30 min, and the
solution became clear. The reaction was continued in the
presence of a 1 M NaOH solution for 3 days under open
(13) Gilbert, H. F. Adv. Enzymol. 1990, 63, 69.
(14) Jones, L. C.; Lackowski, W. M.; Vasilyeva, Y.; Wilson, K.;
Chechik, K. Chem. Commun. 2009, 1377–1379.
(15) Gadelle, A.; Defaye, J. Angew. Chem., Int. Ed. Engl. 1991, 30, 78.
Org. Lett., Vol. 13, No. 14, 2011
3573

linked glucopyranoside. The two geminal protons H-6a and
H-6b were found at 3.53, 3.28 ppm. A 1D 13C NMR
spectrum also revealed that the compound had only 6 types
Scheme 1. Synthesis of Polyhydroxylated Dimer 6 from 3
13
of carbons. Figure 2c shows the 1Hꢀ C 2D HSQC NMR
spectra recorded at 360 K, which revealed unambiguiously
that the C-6 resonates at 39.96 ppm, confirming the
attachment of a sulfur atom to the C-6 carbon. The
simplicity of the NMR spectra of the isolated com-
pound could also suggest the presence of a monomer,
5. However, further characterization by high resolu-
tion ESI-QTOF mass spectrometry (Figure 3) con-
firmed the absence of monomer 5, but the presence of
dimer 6, as a peak at m/z 2496.3448 was observed which
correlates with the ammonium adduct of dimer 6
containing 7 disulfide linakges [(C84H126O56S14
þ
NH4)þ, calculated: 2496.3440]. This is further con-
firmed by the presence of a doubly charged ammonium
adduct peak (Figure 3). The isolated yield for dimer 6
was ∼25% which was not optimized.
air with vigorous agitation. During this time, we observed
the formation of a white precipitate, which was col-
lected at the end of the reaction by centrifugation. The
solid was repeatedly washed with water and methanol
alternatively.
The 1D 1H NMR spectrum of the precipitate in DMSO-d6
revealed that broad signals were observed, but the simplicity
of the 1H NMR spectra suggested that the solid contained a
simple or symmetrical compound; for example, a broad peak
at 4.89 ppm was observed at 299 K (Figure 2a), which could
be assigned to H-1’s. When we heated the NMR solution to
360 K, we observed a dramatic change in resolution
(Figure 2b); all signals became highly resolved. The broad
resonance of anomeric protons appeared as a sharp doublet
at 4.88 ppm (J = 3.7 Hz), and the remaining H-2, H-3, H-4,
and H-5 also had the expected coupling patterns for an R-
Figure 3. High resolution ESI-QTOF mass spectra of dimer 6.
It is interesting to note that one of the polymerization
procedures published by Chechik’s group was also carried
out using NaOH as a base, but they carried out the
experiment in an extremely dilute concentration (50 mg
in 500 mL of a 0.1 N NaOH solution). At basic pH, they
did not observe the formation of a precipitate, but after
neutralization with HCl, they indeed observed a slight
precipitation which was not characterized.14
Compound 6 has an interesting structural feature be-
cause it has 28 hydroxyl groups separated into two groups
of14, and theyare located atthe veryendofeachside of the
long hydrophobic channel. This provides a clear separa-
tion of polar and apolar groups in the molecule, which
should provide it with unusual chemical and physical
properties. We observed that dimer 6 had poor solubility
in water, methanol, and chloroform, which could be well
explained by its structural features.
Figure 2. Varibale temperature 1D 1H NMR spectra of dimer 6
recorded at (a) 299 K and (b) 360 K. (c) 1Hꢀ C HSQC NMR
13
To demonstrate the applicability of the methodology,
wealsodesignedanotherdimer11thathadall thehydroxyl
spectra of 6 in DMSO-d6 recorded at 360 K.
3574
Org. Lett., Vol. 13, No. 14, 2011

groups protected with a methyl group. We expected the
new target to have better solubility in organic solvents
because it contains only apolar groups.
Thus, starting from heptol 7,16 we carried out a permesyl-
ation in a mixture of pyridineꢀdichloromethane (Scheme 2)
to afford the permesylate 8 in 68% yield. All mesylates
were displaced with potassium thioacetate to give the
desired hepta-6-thioacetate 9 in 56% yield.
from the 1H NMR spectra. The 13C spectra also confirmed
the existence of 8 types of carbon signals. This is further
supported by high-resolution ESI-QTOF mass spectro-
metry which detected a peak at m/z 2893.7389, which
correlates well with the expected value for the sodium
adduct of C112H182O56S14: 2893.7381 [M þ Na]þ.
Additionally, we also prepared another target, 12
(Figure 4), which possesses half the amount of hydroxyl
groups of 6 and half the amount of methyl groups of 11.
This is realized by strategically placing a methyl group on
the O-2 position of each glucopyranosyl unit. Target 12
should have interesting properties because it has balanced
hydrophobic/hydrophilic groups. The analogous duplex
12 was synthesized in an analogous manner as 11 from per-
6-O-tert-butyldimethylsilyl-3-O-methyl-β-CD (13)16 (see
Supporting Information).
Scheme 2. Synthesis Route to Permethylated Dimer 11
Figure 4. Structure of synthesized dimer 12 and its precursor 13.
Preliminary studies showed that both compounds 11
and 12 were fully soluble in many common organic
solvents including chloroform. However, they still have
poor solubility in water.
In conclusion, we demonstrated that the thiolꢀdisulfide
exchange equilibrium can be applied to the synthesis of a
novel class of fully enclosed β-CD dimers containing 7
disulfide linkages. This class of new hosts could find wide-
spread utilities in many areas of research because of their
uniquely deep hydrophobic pocket and molecular geometry.
Compound 9 was subsequently subjected to an in situ
deacetylation in methanol using sodium methoxide as a
strong base, to provide the intermediate hepta-6-thiol 10
which was not isolated; after stirring in open air under
basic conditions for another three days, dimerization read-
ily occurred as we observed the formation of a new spot on
TLC with an Rf 0.27 (hexane/acetone 3:2) and another
major band below, which is probably due to oligo-/poly-
merization. The upper spot was isolated by column chro-
matography in 21% yield. Like the previously synthesized
dimer 6, the 1D 1H and 13C NMR spectra of compound 11
showed expected symmetry and simplicity as 7 types of
sugar protons and 2 types of methyl groups were observed
Acknowledgment. We thank the Alberta Ingenuity
(now part of Alberta Innovates ꢀ Technology Futures)
and the University of Calgary for financial support of the
current project. The financial support from the Canadian
Foundation for Innovation (Leadership Opportunity Fund)
and the Government of Alberta (Small Equipment Grant
Program) is also gratefully acknowledged.
Supporting Information Available. Experimental pro-
cedures and related analytical data for compounds 4ꢀ6,
8ꢀ11, and 12. This material is available free of charge via
the Internet at http://pubs.acs.org.
(16) Takeo, K.; Mitoh, H.; Uemura, K. Carbohydr. Res. 1989, 187,
203.
Org. Lett., Vol. 13, No. 14, 2011
3575