Synthesis of Readily Modifiable Cyclodextrin Analogues via Cyclodimerization of an Alkynyl-Azido Trisaccharide

Article abstract of DOI:10.1021/ja039374t

A convergent strategy for the synthesis of β-cyclodextrin analogues is reported, utilizing preferential cyclodimerization of an azido-alkyne trisaccharide via Cu(I)-catalyzed [3 + 2] dipolar cycloaddition of the alkyne and azide functional groups. The res

Full text of DOI:10.1021/ja039374t

Published on Web 01/27/2004
Synthesis of Readily Modifiable Cyclodextrin Analogues via
Cyclodimerization of an Alkynyl-Azido Trisaccharide
Kyle D. Bodine, David Y. Gin, and Mary S. Gin*
Department of Chemistry, UniVersity of Illinois at Urbana-Champaign, 600 South Mathews AVe,
Urbana, Illinois 61801
Received November 1, 2003; E-mail: mgin@scs.uiuc.edu
Scheme 1 a
Cyclodextrins (CDs) have found widespread use in supramo-
lecular chemistry due to their ability to bind hydrophobic molecules
within their cavities while dissolved in polar solvents such as water.
Functionalized cyclodextrins have found applications as molecular
reactors,1 enzyme mimics (catalysts),2 molecular machines,3 and
electrode surface modifiers.4 Often a key feature of these systems
is a cyclodextrin that has been selectively modified to display
necessary functional groups in specific positions on the macrocycle.
Indeed, differences in reactivity between the secondary hydroxyl
groups on one face of â-CD and the primary hydroxyl groups on
the other face, in conjunction with the geometric constraints of the
macrocycle, have allowed the selective substitution at one,5 two,5,6
three,5,7 or four8 positions about the macrocycle. Even with these
promising advances, the selective placement of different functional
groups about the cyclodextrin macrocycle remains challenging, thus
limiting the full potential of cyclodextrins as functional materials.
An alternative approach to selectively modified cyclodextrin
macrocycles is the stepwise synthesis of cyclodextrins and cyclo-
dextrin analogues from sugar residues that bear different functional
groups.9 Toward this goal, both stepwise oligosaccharide synthesis
followed by cycloglycosylation,10 as well as cyclooligomerization
of mono-,11 di-,9 and trisaccharides12 to give synthetic cyclooligo-
saccharides has been reported. Despite these remarkable synthetic
efforts, disadvantages include long synthetic sequences, low cycliza-
tion yields, anomeric mixtures, and for cyclooligomerizations, mix-
tures of macrocycles of different sizes that are difficult to separate.
Our interest in macrocycles with selectively positioned func-
tionality led us to consider the synthesis of cyclodextrin analogues
from already-functionalized sugar units. However, we set out to
streamline the synthesis by targeting a cyclodimerization reaction
of appropriately substituted trisaccharides. The ligation reaction that
we chose to investigate is the [3 + 2] Huisgen cyclization of an
azide with an alkyne to form a triazole.13 This reaction is chemo-
selective and high-yielding, and the precursors are otherwise stable
to many reaction conditions. Herein we report the highly convergent
preparation of macrocycle 1, a cyclodextrin analogue that exhibits
association characteristics with the hydrophobic fluorophore 8-anilino-
1-naphthalenesulfonate that are similar to those of â-CD.
a (a) NaH, HCCCH2Br, DMF; (b) CAN, 4:1 MeCN/H2O; (c) Ph2SO,
Tf2O, 3:1 PhMe/CH2Cl2; 3; Et3N; (d) Ph2SO, Tf2O, 3:1 PhMe/CH2Cl2;
TMSN3; Et3N.
Trisaccharide 2 was prepared by sequential dehydrative glycosyl-
ation of mannosides derived from the common intermediate 310b
(Scheme 1). The terminal alkyne was installed by C4-O-alkylation
of 3 (NaH, HCCCH2Br) to afford propargyl ether 4 in >99% yield.
Anomeric deprotection (CAN, H2O) afforded the hemiacetal 5
(68%, 9:1 R: â) which then served as an appropriate donor for
sulfoxide-mediated dehydrative glycosylation.14 Thus, hemiacetal
5 was activated with Ph2SO and Tf2O at -45 °C in the absence of
a triflic acid scavenger, allowing for glycosylation of the C-4
hydroxyl in 3 to give the disaccharide in 90% yield with complete
R-selectivity.15 Oxidative removal of the 4-methoxyphenyl acetal
as before afforded hemiacetal 6 in 76% yield. A second iteration
of glycosylation with C-4 nucleophile 3 and removal of the
anomeric protecting group gave the final hemiacetal 7 in 60% yield
for the two-step sequence. Glycosylation of trimethylsilyl azide
under the same dehydrative coupling conditions furnished the
trisaccharide 2 in 94% yield.
A critical feature of our synthetic strategy for the formation of
macrocycle 1 is the preferred cyclodimerization over unproductive
linear oligomerization. Thus, trisaccharide 2 was subjected to a
variety of [3 + 2] cycloaddition conditions. While cyclodimerization
was observed upon simply heating a solution of 2 in toluene, we
found these conditions to be sluggish and to lead to the formation
of multiple products. Decomposition could be suppressed by using
copper catalysts, a number of which have been shown to promote
the Huisgen cycloaddition of alkynes and azides.16 After surveying
the efficiency of reaction with a variety of Cu(I) sources (CuSO4,
sodium ascorbate; CuSO4, Cu0 powder; CuI‚P(OEt)3; CuI), additives
(DBU, tris(triazolyl)amine), and solvents (DMF/H2O, CH3CN,
toluene), we were pleased to discover that treatment of trisaccharide
2 with CuI and DBU in toluene at 50 °C afforded the desired
cyclodimer in 80% yield, with the corresponding cyclotrimer formed
in 15% yield (eq 1). Transfer hydrogenolysis to remove the 18
benzyl groups (NH4HCO2, Pd/C, 50 °C) effected quantitative
The trisaccharide used to test our strategy (2) consists of 1,4-
linked mannose units protected as 2,3,6-tribenzyl ethers, with an
anomeric azide and a 4-propargyl ether at the opposing termini.
9
1638
J. AM. CHEM. SOC. 2004, 126, 1638-1639
10.1021/ja039374t CCC: $27.50 © 2004 American Chemical Society

C O M M U N I C A T I O N S
conversion to the desired macrocycle 1, which was purified by
dialysis to remove excess salts.
ANS + 1 fluorescence relative to that of ANS + â-CD. Simple
solvent effects were ruled out by addition of mannose to ANS,
which displayed fluorescence identical to that of ANS alone
(superimposable with black curve). An association constant of 38
( 10 M-1 was determined for ANS + 1 by fluorescence titration,21
which is comparable to that of â-CD (71 ( 4 M-1).22 While this
suggests that ANS associates similarly within the cavities of 1 and
â-CD, further studies will be necessary to determine the generality
of such host-guest interactions for 1.23
In conclusion, we have demonstrated a convergent strategy for
the synthesis of â-cyclodextrin analogues, exemplified by the
preferential cyclodimerization of trisaccharide 2 via a [3 + 2]
dipolar cycloaddition of the alkyne and azide functional groups.
The resultant oligosaccharide macrocycle retains the binding
propensity of cyclodextrins, as demonstrated by the similar ANS
association constants measured for macrocycle 1 and â-cyclodextrin.
This new synthetic strategy opens up new avenues for modular
preparation of functionally diverse cyclodextrin analogues that are
otherwise inaccessible. Efforts toward such selectively modified
cyclodextrin analogues are currently underway.
Acknowledgment. This work was supported by the Roy J.
Carver Charitable Trust (02-198) NIH (GM58833), and the
University of Illinois.
Supporting Information Available: Experimental procedures for
all compounds and for fluorescence studies (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
This synthetic protocol offers a facile, highly convergent method
for the preparation of cyclodextrin-like macrocycles with differ-
entially functionalized starting materials; however, there are distinct
differences between macrocycle 1 and â-CD. First, where â-CD
comprises seven sugar residues, macrocycle 1 has six sugar residues
and two triazole rings. Second, cyclodextrins contain glucose
residues, whereas the sugars in macrocycle 1 are derived from
mannose. This difference in C2-stereochemistry of the pyranose
rings precludes hydrogen-bonding between secondary hydroxyl
groups on adjacent residues of 1, which should result in diminished
structural rigidity relative to the cyclodextrins.17 In light of these
differences, it is imperative to determine whether macrocycle 1
displays a propensity for forming inclusion complexes analogous
to those of â-CD in order to assess its potential utility in
supramolecular systems.
A number of small molecules are known to form inclusion com-
plexes with cyclodextrins.18 We chose to investigate the interaction
of 8-anilino-1-naphthalenesulfonate (ANS) with macrocycle 1 as
complexation with â-CD is readily detected by changes in the
fluorescence of the ANS dye (Figure 1). In aqueous solutions, the
fluorescence of ANS is largely quenched (black curve).19 Addition
of an excess of â-CD leads to an increase in the fluorescence
intensity and a slight blue shift as the bound ANS encounters a
more hydrophobic environment (red curve).20 Addition of a similar
excess of 1 to a solution of ANS resulted in similar changes in the
fluorescence intensity (blue curve), indicating that macrocycle 1 is
able to serve as host to inclusion of ANS in a manner similar to
â-CD. It should be noted that macrocycle 1 displays weak
fluorescence, which may account for the increased intensity of the
References
(1) Takahashi, K. Chem. ReV. 1998, 98, 2013.
(2) (a) Breslow, R.; Dong, S. D. Chem. ReV. 1998, 98, 1997. (b) Rizzarelli,
E.; Vecchio, G. Coord. Chem. ReV. 1999, 188, 343.
(3) Harada, A. Acc. Chem. Res. 2001, 34, 456.
(4) (a) Ferancova, A.; Labuda, J. Fresenius’ J. Anal. Chem. 2001, 370, 1. (b)
Willner, I.; Willner, B. J. Mater. Chem. 1998, 8, 2543.
(5) Khan, A. R.; Forgo, P.; Stine, K. J.; D’Souza, V. T. Chem. ReV. 1998,
98, 1977.
(6) Teranishi, K. Tetrahedron 2003, 59, 2519.
(7) Atsumi, M.; Izumida, M.; Yuan, D.-Q.; Fujita, K. Tetrahedron Lett. 2000,
41, 8117.
(8) Poorters, L.; Armspach, D.; Matt, D. Eur. J. Org. Chem. 2003, 1377.
(9) Gattuso, G.; Nepogodiev, S. A.; Stoddart, J. F. Chem. ReV. 1998, 98,
1919.
(10) (a) Mori, M.; Ito, Y.; Uzawa, J.; Ogawa, T. Tetrahedron Lett. 1990, 31,
3191. (b) Mori, M.; Ito, Y.; Ogawa, T. Carbohydr. Res. 1989, 192, 131.
(c) Hoffmann, B.; Bernet, B.; Vasella, A. HelV. Chim. Acta 2002, 85,
265.
(11) Chong, P. Y.; Petillo, P. A. Org. Lett. 2000, 2, 1093.
(12) Sakairi, N.; Kuzuhara, H. J. Chem. Soc., Chem. Commun. 1993, 1874.
(13) (a) Huisgen, R.; Knorr, R.; Mobius, L.; Szeimies, G. Chem. Ber. 1965,
98, 4014. (b) Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem.,
Int. Ed. 2001, 40, 2004. (c) Peto, C.; Batta, G.; Gyorgydeak, Z.; Sztaricskai,
F. J. Carbohydr. Chem. 1996, 15, 465. (d) Katritsky, A. R.; Zhang, Y.;
Singh, S. K. Heterocycles 2003, 60, 1225.
(14) Garcia, B. A.; Gin, D. Y. J. Am. Chem. Soc. 2000, 122, 4269.
(15) Maintaining the reaction temperature below -20 °C allows acid-catalyzed
post-coupling anomerization to afford the desired R-anomer as the sole
product.
(16) (a) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew.
Chem., Int. Ed. 2002, 41, 2596. (b) Pe´rez-Balderas, F.; Ortega-Mun˜oz,
O.; Morales-Sanfrutos, J.; Herna´ndez-Mateo, F.; Calvo-Flores, F. G.;
Calvo-As´ın, J. A.; Isac-Garc´ıa, J.; Santoyo-Gonza´les, F. Org. Lett. 2003,
5, 1951. (c) Tornoe, C. W.; Christensen, C.; Meldal, M. J. Org. Chem.
2002, 67, 3057.
(17) Saenger, W.; Jacob, J.; Gessler, K.; Steiner, T.; Hoffmann, D.; Sanbe,
H.; Koizumi, K.; Smith, S. M. Takaha, T. Chem. ReV. 1998, 98, 1787.
(18) Rekharsky, M. V.; Inoue, Y. Chem. ReV. 1998, 98, 1875.
(19) Ebbesen, T. W.; Ghiron, C. A. J. Phys. Chem. 1989, 93, 7139.
(20) Nishijo, J.; Nagai, M.; Yasuda, M. Carbohydr. Res. 1993, 245, 43.
(21) Connors, K. A. Binding Constants: The Measurement of Molecular
Complex Stability; John Wiley & Sons: New York, 1987.
(22) Nishijo, J.; Nagai, M. J. Pharm. Sci. 1991, 80, 58.
(23) A 1H NMR spectrum of a 1:1 mixture of 1 and ANS in D2O displayed
complexation induced shifts (CIS) for both compounds. Those for ANS
were of different magnitude and occasionally different direction compared
to the CIS of ANS with â-CD. This highlights that the specific structure
of the complex of ANS + 1 is distinct from that of ANS + â-CD. Further
studies to explore the solution structure of 1 and its complexes are currently
underway.
Figure 1. Fluorescence emission spectra of ANS (0.1525 mM) in the
presence and absence of â-cyclodextrin (21.12 mM) and macrocycle 1
(21.11 mM). All solutions were 0.1 M aqueous sodium phosphate buffer
(pH 7.40), λex ) 365 nm.
JA039374T
9
J. AM. CHEM. SOC. VOL. 126, NO. 6, 2004 1639