Efficient preparation of glycoclusters from silsesquioxanes

Article abstract of DOI:10.1021/ol040043a

(Chemical Equation Presented) A new type of glycocluster based on polyhedral oligosilsesquioxanes (POSS) has been efficiently prepared from unprotected mannoside and lactoside employing a convergent approach of thiol-radical addition reaction. The versatility of this approach was demonstrated by functionalization of mannosides and lactosides of different-length spacers.

Full text of DOI:10.1021/ol040043a

ORGANIC
LETTERS
2004
Vol. 6, No. 20
3457-3460
Efficient Preparation of Glycoclusters
from Silsesquioxanes
Yongjun Gao,† Atsuko Eguchi,‡ Kazuaki Kakehi,‡ and Yuan C. Lee*,†
Department of Biology, Johns Hopkins UniVersity, 3400 North Charles Street,
Baltimore, Maryland 21218, USA, and Faculty of Pharmceutical Sciences,
Kinki UniVersity, Kowakae 3-4-1, Higashi-Osaka 577-8502, Japan
yclee@jhu.edu
Received July 7, 2004
ABSTRACT
A new type of glycocluster based on polyhedral oligosilsesquioxanes (POSS) has been efficiently prepared from unprotected mannoside and
lactoside employing a convergent approach of thiol-radical addition reaction. The versatility of this approach was demonstrated by functionalization
of mannosides and lactosides of different-length spacers.
Cell surface oligasaccharides are involved in numerous
biological processes such as cancer cell metastasis, inflam-
mation, and infections by bacteria and viruses.1 As the
oligasaccharide-receptor interactions are often multivalent,
the concept of cluster effect2 has attracted considerable
attention, and extensive efforts have been expended on
building glycoclusters with a variety of scaffolds such as
cyclodextrins (CDs),3 polymers,4 dendrimers,5 calix[4]-
allenes,6 crown ethers,7 and peptides.8
other polyhedral frameworks have been developed, but cube-
octameric clusters are most common. There are several
reasons for the current widespread interests in POSS, the
most important of which being the remarkable simplicity of
(3) (a) Laine´, V.; Coste-Sarguet, A.; Gadelle, A.; Defaye, J.; Perly, B.;
Djeda¨ıni-Pilard, F. J. Chem. Soc., Perkin Trans. 2 1995, 1479-1487. (b)
Ashton, P. R.; Ko¨eniger, R.; Stoddart, J. F.; Alker, D.; Harding, V. D. J.
Org. Chem. 1996, 903-908. (c) Kassab, R.; Fe´lix, C.; Parrot-lopez, H.;
Bonaly, R. Tetrahedron Lett. 1997, 38, 7555-7558.
(4) (a) Choi, S.-K.; Mammen, M.; Whitesides, G. M. Chem. Biol. 1996,
3, 97-104. (b) Nishimura, S.-I.; Lee, Y. C. In Polysaccharides: Structural
DiVersity and Functional Verstality; Dumitriu, S., Ed.; Marcel Dekker: New
York, 1998; pp 523-537.
(5) (a) Page´, D.; Zanini, D.; Roy, R. Bioorg. Med. Chem. 1996, 4, 1949-
1961. (b) Zanini, D.; Roy, R. J. Org. Chem. 1996, 61, 7348-7354. (c)
Zanini, D.; Roy, R. Bioconjugate Chem. 1997, 8, 187-192. (d) Jayaraman,
N.; Nepogodiev, S. A.; Stoddart, J. F. Chem. Eur. J. 1997, 3, 1193-1199.
(e) Ashton, P. R.; Hounsell, E. F.; Jayaraman, N.; Nilsen, T. M.; Spencer,
N.; Stoddart, J. F.; Young, M. J. Org. Chem. 1998, 63, 3429-3437.
(6) (a) Meunier, S. J.; Roy, R. Tetrahedron Lett. 1996, 37, 5469-5472.
(b) Fujimuto, T.; Shimizu, C.; Hayashida, O.; Aoyama, Y. J. Am. Chem.
Soc. 1998, 120, 601-602.
(7) (a) Dumont, B.; Joly, J. P.; Chapleur, Y.; Marsura, A. Bioorg. Med.
Chem. Lett. 1994, 4, 1123-1126. (b) Ko¨nig, B.; Frickle, T.; Wamman, A.;
Krallman-Wenzel, U.; Lindhorst, T. K. Tetrahedron Lett. 1998, 39, 2307-
2310.
Polyhedral oligosilsesquioxanes (POSS) are an interesting
class of clusters derived from the hydrolytic condensation
of trifunctional organosilicon monomers. The chemistry of
POSS has been extensively discussed.9 A wide variety of
* Corresponding author.
† Johns Hopkins University.
‡ Kinki University.
(1) (a) Varki, A. Glycobiology 1993, 3, 97-130. (b) Gabius, H.-J. Eur.
J. Biochem. 1997, 243, 543-576. (c) Lis, H.; Sharon, N. Chem. ReV. 1998,
98, 637-674.
(2) (a) Lee, Y. C. Carbohydr. Res. 1978, 67, 509-514. (b) Lee, Y. C.;
Lee, R. T. Acc. Chem. Res. 1995, 28, 321-327. (c) Lee, R. T.; Lee, Y. C.
Glycoconjugate J. 2000, 17, 543-551.
10.1021/ol040043a CCC: $27.50
© 2004 American Chemical Society
Published on Web 09/09/2004

synthesis of these compounds. In addition, functionalization
of these cube molecules has expanded the range of available
cluster molecules for a wide range of applications.10 In
particular, POSS provides a versatile platform of well-defined
shape for construction of more sophisticated structures.11 The
use of POSS as cores for dendrimers is particularly attrac-
tive,12 because their polyhedral structures produce spherically
symmetric dendrimers with smaller generation numbers than
conventional cores.13
To the best of our knowledge, there is only one reported
example of synthesis of carbohydrate-functionalized silses-
quioxane that exhibits selective and reversible complexation
to carbohydrate-binding proteins.14 The strategy involved
attachment of glycodendrons to the eight amine groups of
POSS core [(H2NCH2CH2CH2)8Si8O12] via standard amide
bond formation with carbohydrate-derived lactones, but
achieved only 20-53% yields. Although this method is
useful for generating glycoclusters, some drawbacks are
apparent, especially its requirement to start with carbohydrate
lactones. As a consequence, the spacers could not be
synthesized independently, thus limiting the flexibility of this
procedure. In addition, the starting material octaamine [(H2-
NCH2CH2CH2)8Si8O12] is difficult to obtain and maintain.15
As part of our ongoing project involving synthesis and
molecular recognition of glycoclusters, we became interested
in using the rigid POSS structure as scaffold to construct
new forms of carbohydrate clusters, because multivalent
ligands with some rigidity can have enhanced affinity and
selectivity during a binding event.16 The unique geometry
(e.g., size, shape, symmetry) and certain rigidity of POSS
are likely to allow effective ligand presentation. To fulfill
such potentials, attachment of carbohydrate units to the POSS
core must be carried out to the maximum extent in a very
high yield to avoid formation of isomers of undersubstituted
regioisomers, which are very difficult to separate and purify.
After some preliminary consideration of possible reactions,
we chose to use the well-known photoaddition of thiols to
alkenes,17 since its usefulness has been extensively demon-
strated.18
Figure 1. Graphical representation of the key synthetic step
(involving the simultaneous reaction of eight carbohydrate units
with the POSS core) in the synthesis of POSS-based carbohydrate
cluster compounds.
Our strategy is to use radical-addition of ω-thioglycosides
to commercially available octavinyl-POSS (Figure 1).
This method takes advantage of the ease of preparation
of amino-terminated glycosides and the ease of introducing
thiol functions to the amino groups. The use of spacer arms
of varying lengths allows the distance between the carbo-
hydrate residue and the POSS core to be varied. The
conditions of highly efficient addition to thiol end-group via
free-radical route in an anti-Markovnikov fasion would not
incur Si-C or Si-O bond cleavage.19 This process also
eliminates the need to deprotect carbohydrate groups after
their attachment to POSS, which may cause difficulties in
some cases (such as cleavage of Si-C or Si-O bond).
Scheme 1 summarizes the synthetic routes of some
glycosides and their derivatives to be used for the coupling
reactions with γ-thiobutyrolactone.
Compound 3 was obtained by glycosylation of com-
mercially available per-O-acetylated R-D-mannose 1 with
benzyloxycarbonyl (Cbz)-protected 5-amino-pentanol 220
21
using SnCl4 as promoter in 55% yield. Deacetylation of
compound 3 followed by hydrogenolysis of the Cbz-group
using Pd/C catalyst gave the compound 4 in 86% yield.
For the synthesis of compound 8 (Scheme 1), benzobro-
molactose 522 was used to glycosylate commercially available
Cbz-protected 2-aminoethanol 6 in the presence of AgOTf
to give compound 7 in 78% yield. Removal of the O-benzoyl
groups under Ze´mplen conditions followed by catalytic (Pd/
C) hydrogenolysis of the Cbz group gave compound 8 in
82% yield. Compound 11 was prepared similarly, but more
(8) (a) Sprengard, U.; Schudok, M.; Schmidt, W.; Kretzschmar, G.; Kunz,
H. Angew. Chem., Int. Ed. Engl. 1996, 35, 321-324. (b) Lee, R. T.; Lee,
Y. C. Bioconjugate Chem. 1997, 8, 762-765. (c) Kamiya, S.; Kobayashi,
K. Macromol. Chem. Phys. 1998, 199, 1589-1596. (d) Kamitakahara, H.;
Suzuki, T.; Nishigori, N.; Suzuki, Y.; Kanie, O.; Wong, C.-H. Angew.
Chem., Int. Ed. 1998, 37, 1524-28.
(9) Baney, R. H.; Itoh, M.; Sakakibara, A.; Suzuki, T.; Chem. ReV. 1995,
95, 1409-1430.
(10) Lichtenhan, J. D. Comments Inorg. Chem. 1995, 17, 115-130.
(11) (a) Gentle, T. E.; Bassindale, A. R. J. Inorg. Organomet. Pol. 1995,
5, 281-294. (b) Feher, F. J.; Soulivong, D.; Eklund, A. G.; Wyndham, K.
D. Chem. Commun. 1997, 1185-1186. (c) Zhang, C.; Laine, R. M. J.
Organomet. Chem. 1996, 521, 199-201.
(17) Griesbaum, K. Angew. Chem., Int. Ed. Engl. 1970, 9, 273-287.
(18) (a) Lee, R. T.; Lee, Y. C. Carbohydr. Res. 1974, 37, 193-201. (b)
Hanessian, S.; Benalil, A.; Laferrie`re, C. J. Org. Chem. 1995, 60, 4786-
4797. (c) Kieburg, C.; Dubber, M.; Lindhorst, T. K. Synlett 1997, 1447-
1449. (d) Leydet, A.; Moullet, C. Roque, J. P.; Witvrouw, M.; Pannecouque,
C.; Andrei, G.; Snoeck, R.; Neys, J.; Schols, D.; De Clercq, E. J. Med.
Chem. 1998, 41, 4927-4932. (e) Fulton, D. A.; Stoddart, J. F. Org. Lett.
2000, 2, 1113-1116.
(12) Bassidale, A. R.; Gentle, T. E. J. Mater. Chem. 1993, 3, 1319-
1325.
(13) Newkome, G. R.; Moorefield, C. N. Voegtle, F. In Dendritic
Molecules; Newkome, G. R., Moorefield, C. N., Voegtle, F., Eds.; VCH:
Weinheim: New York, 1996; p 261.
(14) Feher, F. J.; Wyndham, K. D.; Knauer, D. J. Chem. Commun. 1998,
2393-2394.
(15) (a) Feher, F. J.; Wyndham, K. D. Chem. Commun. 1998, 323-
324. (b) Gravel, M. C.; Zhang, C.; Dinderman, M.; Laine, R. M. Appl.
Organomet. Chem. 1999, 13, 329-336.
(19) Rozga-Wijas, K.; Chojnowski, J.; Zundel, T.; Boileau, S. Macro-
molecules 1996, 29, 2711-2720.
(20) Hirai, Y.; Nagatsu, M. Chem. Lett. 1994, 21-22.
(16) Vrasidas, I.; Andre, S.; Valentini, P.; Bock, C.; Lensch, M.; Kaltner,
H.; Liskamp, R. M. J.; Gabius, H. J.; Pieters, R. J. Org. Biomol. Chem.
2003, 1, 803-810.
(21) Ichikawa, Y.; Lee, Y. C. Carbohydr. Res. 1990, 198, 235-246.
(22) Lichtenthaler, F. W.; Kaji, E.; Weprek, S. J. Org. Chem. 1985, 50
(19), 3505-3515.
3458
Org. Lett., Vol. 6, No. 20, 2004

Scheme 1. Synthesis of Aminoalkyl Glycosides
Scheme 3. Synthesis of Glycoclusters
simply by (i) glycosylation of Fmoc-protected 6-aminohex-
anol 923 with benzobromolactose 522 using AgOTf catalyst
and then (ii) removal of the O-benzoyl and Fmoc groups
simultaneously under Ze´mplen conditions. Thus, compound
11 was obtained in 70% overall yield.
The thiol compounds 12, 13, 14 were prepared by reacting
the primary amino groups in 4, 8, and 11 with γ-thiobuty-
rolactone (6 equiv) at 50 °C in 0.25 M aqueous NaHCO3 (3
equiv) in 50% ethanol solution in the presence of DTT (2.5
equiv) for 8-10 h (Scheme 2).24 The thiol compounds were
Scheme 2. Synthesis of Thiols
The efficient addition of thiol-teminated glycoside residues
to the POSS core was achieved (Scheme 3) when a mixture
of octavinyl POSS and 12-14 (3.9-4.4 equiv for each vinyl
group) in H2O/THF (1:1) was irradiated with UV light (254
nm) in the presence of a catalytic amount of AIBN. After
1
24-48 h, the disappearance of the H resonance at 5.95-
6.18 ppm which corresponds to the vinyl group and
emergence of new signals corresponding to the incorporated
SiCH2 residues (1.06-1.09 ppm) indicated the intended
reaction was complete. Purification of the crude product from
the photochemical reaction by Sephadex G-15 gel filtration
(H2O as eluent) afforded pure glycoclusters 15, 16, and 17
in 70%, 66%, and 73% yield, respectively. It should be noted
that most of the unreacted carbohydrate thiols could be
recovered in pure form after the gel filtration. Only a trace
amount of disulfide (<5%) was obtained. To prevent the
oxidation of the carbohydrate thiols to disulfides during
storage, it is better to keep them under inert atmosphere in
refrigerator. The identities of these compounds were estab-
lished unequivocally from close inspection of their 1H NMR
and 13C NMR spectra.25 The structures were further con-
firmed by MALDI-TOF mass spectra.
purified by silica gel chromatography, and the products were
isolated in excellent yields ranging from 82 to 90%.
(23) Lin, P. Kong Thoo; Kuksa, V. A.; Maguire, N. M. Synthesis 1998,
6, 859-866.
(24) Norberg, T.; Blixt, O. J. Carbohydr. Chem. 1997, 16, 143-154.
Org. Lett., Vol. 6, No. 20, 2004
3459

Although the 1H NMR data at 25 °C showed the broaden-
ing of both of the signals for the POSS cores and carbohy-
drate portions, the ratios of the integrals for the signals of
saccharide residue protons and for the signals belonging to
the POSS cores are in accordance with the structures of the
We examined the binding of 16 or 17 by RCA120 (a
â-galactose specific lectin). The study was performed by
measuring the inhibitory effect of 16 or 17 on the binding
of asialo-oligosaccharides from human R1-acid glycoprotein
(AGP) by RCA120 using capillary affinity electrophoresis.26
Complete inhibition was shown by 10 µM of 16 or 17. A
typical example using 16 as an inhibitor is shown in the
Supporting Information. In contrast, the inhibition is not
shown even at 20 mM of lactose (data not shown). This
indicated that 16 and 17 showed 200 times or higher
inhibitory effect than lactose, presumably due to the cluster
effect.2 As predicted by binding specificity of RCA120, 15,
which contains only mannose residues did not show any
inhibition at 10 µM. The detailed binding characteristics will
be reported elsewhere.
1
expected products. The H NMR spectra of 15, 16, and 17
show signals at 1.06-1.09 ppm revealing the presence of
the SiCH2 groups on the POSS cores. In addition, the signals
at 2.54-2.55 ppm, corresponding to the terminal methylene
(CH2SH) of 12, 13, and 14 were replaced by two new signals
at 2.59-2.60 ppm and 2.69-2.70 ppm as a result of the
radical addition reactions.
13C NMR spectra of 15 display only one anomeric carbon
signal at 99.34 ppm. 13C NMR spectra of 16 and 17 show
two anomeric carbon signals (102.55 ppm, 101.85 ppm for
16, 102.56 ppm, 101.82 ppm for 17). High-field chemical
shifts at 12.13 ppm for 15, 12.51 ppm for 16, and 12.15
ppm for 17 further confirm the presence of SiCH2 groups
on the POSS cores. Glycoclusters 15, 16, and 17 show only
one signal at 173.98 ppm for 15, 175.87 ppm for 16, and
173.99 ppm for 17 corresponding to their amide residues,
respectively.
In conclusion, we have described an efficient synthetic
route for a variety of POSS-based glycoclusters by photo-
addition of thiols to vinyl groups. This strategy should be
easily extended to attachment of other glycosides, including
those of more complicated oligosaccharides to POSS cores
with possible variations in spacers. This method is also
amenable to generate combinatorial libraries by changing the
nature of the sugar and spacing arms. Preliminary lectin
binding study indicated that these novel glycoclusters showed
strong inhibition of the binding of asialo-oligosaccharide
mixture derived from human R1-acid glycoprotein by
RCA120.
The results of MALDI-TOF mass spectra also supported
homogeneity of the products. The molecular masses of
glycoclusters 15, 16, and 17 were detected as peaks of the
desired molecular ions in agreement with their [M + K]+
adducts, respectively.
(25) Selected data for compound 15: 1H NMR (400 MHz, D2O, rt) δ
4.85 (s, 8H, H-1), 3.20 (bs, 16H), 2.70 (bs, 16H), 2.36 (bs, 16H), 1.90 (bs,
16H), 1.63 (bs, 16H), 1.55 (bs, 16H), 1.41 (bs, 16H), 1.09 (bs, 16H, SiCH2);
13C NMR (100 MHz, D2O, rt) δ 173.98 (8 CONH), 99.34 (C-1), 72.32,
70.35, 69.82, 67.04, 66.25, 60.48, 38.91, 34.50, 28.15, 25.07, 22.78, 12.13;
MALDI-TOF MS 3608.30 [M + K]+. Selected data for compound 16: 1H
NMR (400 MHz, D2O, rt) δ 4.49 (d, 8H, J ) 8.0 Hz, H-1′), 4.45 (d, 8 H,
J ) 7.6 Hz, H-1), 3.35 (bt, 8 H), 2.70 (bs, 16H), 2.60 (bs, 16H), 2.38 (bs,
16H), 1.90 (bs, 16 H), 1.06 (bs, 16H, SiCH2); 13C NMR (100 MHz, D2O,
rt) δ 175.30 (8 CONH), 102.55, 101.85 (8C-1 and 8C-1′), 78.00, 74.93,
74.36, 73.89, 72.38, 72.12, 70.52, 68.11, 60.60, 59.71, 38.89, 34.34, 24.76,
12.51; MALDI-TOF MS 4568.50 [M + K]+. Selected data for compound
17: 1H NMR (400 MHz, D2O, rt) δ 4.47 (bs, 8H, H-1′), 4.45 (bs, 8H,
H-1), 3.32 (bt, 8H), 3.19 (bs, 16H), 2.69 (bs, 16H), 2.59 (bs, 16H), 2.36
(bs, 16H), 1.89 (bs, 16H), 1.64 (bs, 16H), 1.53 (bs, 16H), 1.38 (bs, 32H),
1.08 (bs, 16H, SiCH2); 13C NMR (100 MHz, D2O, rt) δ 173.99 (CONH),
102.56, 101.82 (C-1 and C-1′), 78.04, 74.93, 74.33, 74.10, 72.45, 72.16,
70.53, 69.95, 68.12, 60.60, 59.77, 38.96, 34.53, 28.57, 24.63, 12.15;
MALDI-TOF MS: 5016.80 [M + K].
Acknowledgment. This work is supported by the Na-
tional Institutes of Health Research Grant No. R01 DK
009700.
Supporting Information Available: Detailed synthetic
procedures, biological experimental procedures, NMR, and
FAB or MALDI-TOF mass spectra data of all new com-
pounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
OL040043A
(26) Nakajima, K.; Oda, Y.; Kinoshita, M.; Kakehi, K. J. Proteome Res.
2003, 2, 81-88.
3460
Org. Lett., Vol. 6, No. 20, 2004