Synthesis of Carbohydrate-Centered Oligosaccharide Mimetics Equipped with a Functionalized Tether

Article abstract of DOI:10.1021/jo000432s

Synthetic glycoclusters have gained substantial attention as mimetics of multivalent glycoconjugates. For their proposed glycobiological applications, it is advantageous to incorporate a functionalized tether into the clusters, which allows coupling to so

Full text of DOI:10.1021/jo000432s

J . Org. Chem. 2000, 65, 5275-5281
5275
Syn th esis of Ca r boh yd r a te-Cen ter ed Oligosa cch a r id e Mim etics
Equ ip p ed w ith a F u n ction a lized Teth er
Michael Dubber and Thisbe K. Lindhorst*
Institute of Organic Chemistry, Christian-Albrechts-University,
Otto-Hahn-Platz 4, D-24098 Kiel, Germany
tklind@oc.uni-kiel.de
Received March 22, 2000
Synthetic glycoclusters have gained substantial attention as mimetics of multivalent glycoconjugates.
For their proposed glycobiological applications, it is advantageous to incorporate a functionalized
tether into the clusters, which allows coupling to solid supports and other molecules such as reporter
groups or even bioactive molecules. We herein report the use of carbohydrates as oligofunctional
scaffolds for the synthesis of tethered cluster mannosides. Glycocluster 11 was prepared following
two different pathways, starting either from glucose or the nonreducing disaccharide trehalose.
The oligo alcohols 5 and 14 served as acceptors in the subsequent oligo-mannosylation reaction, in
which three main problems were overcome: (i) incomplete glycosylation, (ii) cleavage of the core-
glycoside, and (iii) ortho ester formation. Optimum conditions for the glycosylation were identified
utilizing an advanced MALDI-TOF protocol.
In tr od u ction
It was recently shown that carbohydrate derivatives
can serve as multivalent scaffolds for the synthesis of
oligosaccharide mimetics.9-12 In this paper, an important
extension of this concept is described, in which carbohy-
drate-centered glycoclusters carrying a functionalized
tether are synthesized. Incorporation of a spacer arm to
allow further conjugation of the oligosaccharide mimetic
has just recently been recognized as a potential advan-
tage of the n-pentenyl glycoside methodology.13
In this paper two approaches for the synthesis of
tethered glycoclusters were followed: (i) the monosac-
charide pathway using D-glucose as the core molecule,
and (ii) the trehalose pathway, in which the nonreducing
disaccharide was appropriately modified to serve as
scaffold for oligoglycosylation, followed by cleavage of the
interglycosidic bond. A mannosyl donor was used for the
glycosylation reactions, as cluster mannosides are under
special investigation in our laboratory as inhibitors of
mannose-specific adhesion of Escherichia coli.7,14,15
One of the main objectives in glycobiology is the
investigation of carbohydrate-protein interactions, which
play essential roles in cell-cell communication.1,2 A major
benefit of a more detailed understanding of these molec-
ular recognition processes would be effective intervention
in the associated pathological phenomena, along with the
development of carbohydrate-based drugs such as car-
bohydrate-based antiadhesives to prevent microbial in-
fections.3 To simplify the synthesis of the necessary
saccharides as well as to improve their biological func-
tions, a large number of neoglycoconjugates4 and glyco-
mimetics5 have been synthesized including multivalent
glycoconjugates of various architectures.6 The latter have
been designed as to mimic oligosaccharides or oligo-
antennary glycoconjugates.7
For all potential glycobiological applications of (mul-
tivalent) glycomimetics, it is advantageous to have the
respective molecules equipped with a handle to allow the
attachment of reporter groups or labels such as various
dyes or biotin, thus facilitating the necessary glycobio-
logical assays. Furthermore, a functionalized tether
might be used for drug targeting or for the attachment
to solid phases,8 improving the perspectives of synthesis,
purification, and molecular diversity as well as isolation
of lectins.
Resu lts a n d Discu ssion
Mon osa cch a r id e P a th w a y. To transform D-glucose
into an oligo-hydroxylated core molecule equipped with
a functionalized tether at the anomeric center, the
most simple and shortest glycosidation method available
could be utilized as the first step. Starting with glucose
* To whom correspondence should be addressed. Tel: Int 49 431
880 2023. Fax: Int 49 431 880 7410.
(1) Varki, A. Glycobiology 1993, 3, 97.
(9) Wong, C.-H.; Ye, X.-S.; Zhang, Z J . Am. Chem. Soc. 1998, 120,
7137.
(2) Dwek, R. A. Chem. Rev. 1996, 96, 683.
(3) Kahane, I.; Ofek, I. Toward Anti-Adhesion Therapy for Microbial
Diseases. Adv. Exp. Med. Biol. 1996, 408.
(10) Wunberg, T.; Kallus, C.; Opatz, T.; Henke, S.; Schmidt, W.;
Kunz, H. Angew. Chem. 1998, 110, 2620; Angew. Chem., Int. Ed. 1998,
37, 2503.
(4) Bovin, N. V.; Gabius, H.-J . Chem. Soc. Rev. 1995, 413.
(5) Simanek, E. E.; McGarvey, G. J .; J ablonski, J . A.; Wong, C.-H.
Chem. Rev. 1998, 98, 833.
(6) J ayaraman, N.; Nepogodiev, S. A.; Stoddart, J . F. Chem. Eur. J .
1997, 3, 1193.
(7) Lindhorst, Th. K.; Dubber, M.; Krallmann-Wenzel, U.; Ehlers,
S. Eur. J . Org. Chem. 2000, 2027.
(8) Wijsman, E. R.; Filippov, D.; Valentijn, A. R. P. M.; van der
Marel, G. A.; van Boom, J . H. Recl. Trav. Chim. Pays-Bas 1996, 115,
397.
(11) Kitov, P. I.; Sadowska, J . M.; Mulvey, G.; Armstrong, G. D.;
Ling, H.; Pannu, N. S.; Read, R. J .; Bundle, D. R. Nature 2000, 403,
669.
(12) Kieburg, C.; Dubber, M.; Lindhorst, Th. K. Synlett 1997, 1447.
(13) Buskas, T.; So¨derberg, E.; Konradsson, P.; Fraser-Reid, B. J .
Org. Chem. 2000, 65, 958.
(14) Lindhorst, Th. K.; Kieburg, C.; Krallmann-Wenzel, U. Glyco-
conjugate J . 1998, 15, 605.
(15) Ko¨tter, S.; Krallmann-Wenzel, U.; Ehlers, S.; Lindhorst, Th.
K. J . Chem. Soc., Perkin Trans. 1 1998, 2193.
10.1021/jo000432s CCC: $19.00 © 2000 American Chemical Society
Published on Web 08/03/2000

5276 J . Org. Chem., Vol. 65, No. 17, 2000
Dubber and Lindhorst
Sch em e 1. Mon osa cch a r id e P a th w a ya
reported yields when the same procedure was used for
per-allylation of nonfunctionalized carbohydrates,18 and
this reflects the difficulties connected with modification
of the tethered derivatives described herein. The subse-
quent hydroboration step with 9-BBN-H was accompa-
nied with an unexpectedly large number of side reactions
which could, however, be reduced by modification of the
18
conditions previously described using a reaction time
of 1 h at room temperature. During the oxidation step
employing NaOH/H2O2, an intramolecular ring closure
with elimination of 9-BBN-Br was observed, as well as
several other side reactions. This observation is analo-
gous to work published by H. C. Brown19 and could
correspondingly be avoided by replacing aqueous NaOH
with aqueous NaOAc. Thus, the glucose-based tetraol
equipped with a bromo-functionalized tether (5) was
obtained, which is the acceptor molecule for the following
glycosylation reaction. Glycosylation of 5, however, is a
nontrivial task even in view of the advantages of rela-
tively well-advanced modern glycosidation methodology.
Multiple glycosylations of relatively small core molecules
are difficult, because the core becomes increasingly
sterically hindered during proceeding glycosylation and
the last glycosylation step consequently becomes the
slowest and most critical of all. This problem was
frequently monitored by MALDI-TOF measurements
during the glycosylation reaction. An additional problem
is that the separation of the target molecules 8 and 15
(vide supra) from a mixture of structurally imperfect,
however very similar products is hardly possible. There-
fore, formation of side products arising from incomplete
glycosylation of the core polyols had to be avoided by
using a large excess of the glycosyl donor. Glycosyl
imidates20 proved to be suited for this reaction. A problem
connected with the special case of mannosylations is the
formation of glycosyl ortho esters, which is especially
favored in the manno-series due to the relatively high
stability of the intermediate acyloxonium ion.21 While the
utilization of an ether-protected mannosyl trichloro-
acetimidate would exclude the possibility of ortho ester
formation, the ether-protected mannosyl donor was not
suitable for the preparation of 8. Finally, the benzoylated
mannosyl imidate 722 proved to be a suitable donor in
this reaction, being superior to acetylated as well as
benzylated analogues. However, a relatively high con-
centration of Lewis acid was necessary to avoid ortho
ester formation completely. This implied a restriction
concerning the reaction time, which must be limited to
exclude anomerization at the anomeric center of the
glucoside core. Whereas with the ω-azido-functionalized
glycoside 6 no time frame could be found in which ortho
ester rearrangement was complete while the anomeric
configuration remained unchanged, the bromide 5 al-
lowed the preparation of the ortho ester-free fully man-
nosylated glycocluster 8 when the reaction was performed
in concentrated solution in CH2Cl2. When the imidate
excess was reduced by 50%, NMR analysis of the products
revealed no difference, however a minor additional peak
occurring in the MALDI-TOF mass spectrum indicated
the formation of clusters missing one mannosyl residue.
a
Key: (a) BF3‚Et2O, 6-bromohexanol, CH2Cl2; (b) NaOMe,
MeOH; (c) allyl bromide, NaOH, tetrabutylammonium bromide,
rt; (d) 9-BBN-H, THF; (e) NaN3, DMF; (f) TMS-OTf, CH2Cl2; (g)
NaOMe, MeOH-THF; (h) NaN3, DMF; (i) H2, Pd-C.
pentaacetate (1) a BF3-etherate catalyzed reaction16 with
an equimolar amount of 6-bromohexanol led to formation
of glucoside 2. Subsequent deprotection under Zemple´n
conditions gave the unprotected glycoside 3 (Scheme 1).
The next step necessitated allylation of the primary and
secondary hydroxyl groups in 3, excluding inter- and
intramolecular side reactions arising from the participa-
tion of the bromide function of the aglycon moiety. This
was accomplished by modification of a literature-known
procedure17 using PTC conditions and by reducing the
reaction temperature from 40 to 20 °C. The yield of 42%
obtained in this step is significantly lower than the
(18) Dubber, M.; Lindhorst, Th. K. Carbohydr. Res. 1998, 310, 35.
(19) Brown, H. C.; Chen, J . C. J . Org. Chem. 1981, 46, 3978.
(20) Schmidt, R. R.; Kinzy, W. Adv. Carbohydr. Chem. Biochem.
1994, 50, 21.
(16) Takano, T.; Nakatsubo, F.; Murakami, K. Carbohydr. Res. 1990,
203, 341.
(17) Nouguier, R.; Mchich, M. J . Org. Chem. 1985, 50, 3296.
(21) Wulff, G.; Ro¨hle G. Angew. Chem. 1974, 86, 173; Angew. Chem.,
Int. Ed. Engl. 1974, 13, 157.
(22) Bien, F.; Ziegler, T. Tetrahedron: Asymmetry 1998, 9, 781.

Synthesis of Oligosaccharide Mimetics
J . Org. Chem., Vol. 65, No. 17, 2000 5277
Sch em e 2. Tr eh a lose P a th w a ya
a
Key: (a) allyl bromide, NaH, DMF; (b) 9-BBN-H, THF; (c) 7, TMS-OTf, acetonitrile; (d) aqueous TFA, THF, or acetonitrile; (e)
6-bromohexanol, BF3‚Et2O, CH2Cl2.
Thus, no further attempts were made to optimize the
amount of imidate necessary for the glycosylation reac-
tion, especially in light of the fact that 7 is an inexpen-
sive, easily obtained and relatively stable derivative of
mannose.23
This reaction sequence was started with the perallyl-
ation of trehalose using allyl bromide and NaH in DMF
to yield 13, followed by the hydroboration-oxidation
sequence leading to the octaol 14 (Scheme 2). No difficul-
ties were observed in this step which gave the product
in 83% isolated yield, in contrast to the case of the
analogous synthesis of the monosaccharide derivative 5,
which gave low yields.
Mannosylation of 14, as with 5, was complicated by
the aforementioned problems of incomplete glycosylation,
ortho ester-formation and undesired cleavage of the
interglycosidic bond accompanied by subsequent glyco-
sylation of the anomeric center. Fortunately, intermo-
lecular benzoyl group migration leading to benzoylation
of the spacer hydroxyl groups, as it has been reported
for polymer-supported oligosaccharide synthesis,24 was
never observed.
To find optimum conditions allowing the synthesis of
structurally perfect 15 from 14, means for monitoring the
reaction and the formation of the various side products
had to be developed. When the course of the glycosylation
reaction was periodically monitored by MALDI-TOF-MS,
incomplete glycosylation as well as cleavage and destruc-
tion of the core could be seen, but no ortho ester formation
could be observed. However, when the product of the
glycosylation reaction was subjected to silica gel filtration
followed by basic deprotection and neutralization with
After 8 had been obtained in pure form, deprotection
was achieved under modified Zemple´n conditions. As the
benzoyl-protected cluster is hardly soluble in methanol
and deprotection in a THF-MeOH mixture did not go to
completion, a stepwise protocol was introduced. Thus, the
majority of the benzoate groups was first removed with
sodium methoxide in THF-MeOH, and then this solvent
mixture was replaced by pure MeOH to achieve complete
deprotection. The unprotected cluster 9 was then sub-
jected to an SN2 substitution reaction using sodium azide,
followed by catalytic reduction of the intermediate 10 to
obtain cluster 11 carrying the amino-functionalized
tether.
Tr eh a lose P a th w a y. Considering the difficulties con-
nected with the monosaccharide pathway which was used
first for the preparation of 11, an alternative route was
sought in which the symmetrical, nonreducing disaccha-
ride trehalose (12) was used as the starting material,
which can be considered an anomerically protected
glucose. Cleavage of the interglycosidic bond at an inter-
mediate stage of the synthesis would lead to two equiva-
lents of the reducing glucose-centered glycocluster 16.
(24) Nukada, T.; Berces, A.; Whitfield, D. M. J . Org. Chem. 1999,
64, 9030.
(23) See the Supporting Information.

5278 J . Org. Chem., Vol. 65, No. 17, 2000
Dubber and Lindhorst
Sch em e 3
acidic ion-exchange resin, MALDI-TOF analysis of the
deprotected cluster glycoside revealed ortho ester por-
tions formed during the glycosylation step. To use
MALDI-TOF-MS as a reliable tool in this context, the
reproducibility of MALDI-TOF spectra had to be ensured.
Thus, sample preparation for MALDI-TOF-MS in which
dihydroxybenzoic acid is commonly used with carbohy-
drates,25 leading to relatively low reproducibility of the
obtained spectra,26 was improved by evaluation of several
other matrixes. Finally the premix technique with nor-
Harmane27 was modified, using acetone and sodium
iodide as cationizer to avoid [M + matrix]+ peaks. With
this technique in hand, optimum conditions for the
glycosylation could be established.
Because of the high polarity of the polyol-type scaffold
14, acetonitrile had to be used as the solvent in the
glycosylation reaction. The first glycosylation experi-
ments, as evaluated for the synthesis of 8, led to
incomplete glycosylation with approximately five instead
of eight mannosyl residues attached to the core. Therefore
a two-step glycosylation procedure was attempted, in
which the second glycosylation step was carried out in
CH2Cl2. This approach resulted in clusters carrying up
to seven mannosyl moieties. To further enhance the
glycosylation reaction, more concentrated solutions in
acetonitrile were used. This approach, which was suc-
cessful with 8, failed completely in this case, yielding
products in which only two of eight spacer arms of the
core were mannosylated. Interestingly, extreme dilution
of the reaction partners opened the way to complete
glycosylation of 14. While for the glycosylation of 5 in
CH2Cl2 concentrations of >10 mg acceptor polyol/mL
were optimal, the same result was obtained with 14 in
acetonitrile at concentrations of < 0.1 mg acceptor polyol/
mL.
The remaining ortho ester problem could not be solved
by rearrangement reaction using elevated concentrations
of Lewis acid as shown in the synthesis of 8, however, it
could be supressed by using elevated reaction tempera-
tures during the initial phase of the glycoslyation reac-
tion.
Thus the “octopus” glycocluster 15 was ultimately
obtained in pure form. As cleavage of the interglycosidic
bond was shown to be a constant problem during its
synthesis, it was anticipated that this interglycosidic
cleavage would be easy to utilize in order to obtain two
equivalents of the reducing glycocluster 16 after the
mannosylation step. However, this approach was not
successful as 15 was stable even under the conditions of
up to 20% of 1 M aqueous TFA used in organic cosolvents
such as THF or acetonitrile at temperatures up to 80 °C.
Fortunately, it was possible to split 15 with BF3‚Et2O
and a trace of water in CH2Cl2, however further degrada-
tion also occurred during this reaction. Adding an alcohol
component instead of water decreased the extent of
degradation significantly. This led to the appealing
possibility to synthesize 8 from 15 using 6-bromoethanol
in the presence of a large excess BF3‚Et2O, which unhap-
pily produced an anomeric mixture.
(25) Harvey, D. J .; Ku¨ster, B.; Naven, T. J . P. Glycoconjugate J .
1998, 15, 333.
(26) Wang, J .; Sporns, P.; Low, N. H. J . Agric. Food Chem. 1999,
47, 1547.
(27) Nonami, H.; Tanaka, K.; Fukuyama, Y.; Erra-Balsells, R. Rapid
Commun. Mass Spectrom. 1998, 12, 285.
La belin g of 11. The amino function of the tethered
cluster mannoside 11 can serve as a handle for a large

Synthesis of Oligosaccharide Mimetics
J . Org. Chem., Vol. 65, No. 17, 2000 5279
CDCl3) 171.1, 170.7, 169.8, 169.7, 101.2, 73.3, 72.2, 71.8, 70.3,
68.9, 62.4, 34.2, 33.0, 29.6, 28.2, 25.4, 21.2, 21.1, 21.0, 21.0
ppm. Anal. Calcd. for C20H31BrO10: C, 46.98; H, 6.11; Br, 15.63.
Found: C, 46.91;.H, 6.11; Br, 15.48.
number of further modifications (Scheme 3) which are
advantageous for many glycobiological applications.28
Thus the amino group allows attachment of labels such
as biotin, (fluorescent) dyes29,30 or coupling to solid
supports in order to form matrixes for affinity chroma-
tography.
6-Br om oh exyl â-D-Glu cop yr a n osid e (3). To a solution of
2 (1.3 g, 2.6 mmol) in dry MeOH (20 mL), a solution of NaOMe
in MeOH (0.1 M, 2 mL) was added. The reaction mixture was
stirred overnight, neutralized with Amberlite IR 120, filtered
and concentrated to yield 3 (0.91 g, quant) as a colorless
syrup: [R]D -15.4 (c 0.95, MeOH); 1H NMR (500 MHz, MeOH-
d4) 4.29 (d, 1H, J ) 7.9 Hz), 3.95 (m, 1H), 3.90 (dd, 1H, J )
4.7, 12.0 Hz), 3.70 (dd, 1H, J ) 1.9, 12.0 Hz), 3.60 (m, 1H),
3.49 (t, 2H, J ) 6.9 Hz), 3.39 (t, 1H + MeOH, J ) 8.5 Hz),
3.32 (t, 1H, J ) 8.5 Hz), 3.29 (m, 1H), 3.21 (t, 1H, J ) 8.5 Hz),
1.90 (m, 2H), 1.68 (m, 2H), 1.63-1.41 (m, 4H) ppm; 13C NMR
(100.6 MHz, MeOH-d4) 100.6, 79.3, 79.1, 76.3, 72.9, 71.9, 64.0,
35.5, 35.2, 31.8, 30.2, 27.4 ppm. Anal. Calcd. for C12H23BrO6:
C, 41.99; H, 6.75; Br, 23.28. Found: C, 41.79; H, 6.84; Br,
23.25.
6-Br om oh exyl 2,3,4,6-Tet r a -O-a llyl-â-D-glu cop yr a n o-
sid e (4). To a solution of 3 (1.27 g, 3.70 mmol) in water (5
mL), aqueous sodium hydroxide (40 g NaOH in 60 mL H2O)
and tetrabutylammonium bromide (2 g) were added. While the
reaction mixture was vigorously stirred at 20 °C, allyl bromide
(3 mL, 36 mmol) was added over 1 h. After an additional 2.5
h of vigorous stirring, the reaction was quenched by addition
of toluene (100 mL). The phases were separated and the
organic phase was washed with water (6×), dried (MgSO4),
filtered, concentrated and purified by flash chromatography
(toluene-EtOAc 9:1) to yield 4 (0.78 g, 42%) as a colorless
syrup: [R]D -4.6 (c 1.74, CH2Cl2); 1H NMR (500 MHz, CDCl3)
6.06-5.90 (m, 4H), 5.35-5.26 (m, 4H), 5.24-5.17 (m, 1H),
4.42-4.26 (m, 5H) 4.24-4.03 (m, 1H), 3.95 (m, 1H), 3.75 (dd,
1H, J ) 1.5, 11.2 Hz), 3.65 (dd, 1H, dd, J ) 2.5, 11.2 Hz), 3.52
(m, 1H), 3.45 (t, 2H, J ) 7.1 Hz), 3.43-3.34 (m, 3H), 3.23 (t,
1H, J ) 8.1 Hz), 1.90 (m, 2H), 1.66 (m, 2H), 1.54-1.39 (m,
4H) ppm; 13C NMR (100.6 MHz, CDCl3) 136.0, 135.8, 135.5,
135.4, 117.6, 117.6, 117.4, 117.3, 104.1, 84.8, 82.3, 78.3, 75.5,
75.1, 74.5, 74.3, 73.1, 70.4, 69.7, 34.4, 33.4, 30.1, 28.6, 25.9
ppm. Anal. Calcd for C24H39BrO6: C, 57.25; H, 7.81; Br, 15.87.
Found: C, 57.46; H, 7.85; Br, 15.90.
6-Br om oh exyl 2,3,4,6-Tetr a -O-(3-h yd r oxyp r op yl)-â-D-
glu cop yr a n osid e (5). To a solution of 4 (0.78 mg, 1.55 mmol)
in dry THF (30 mL), 0.5 M 9-BBN-H solution in THF (28 mL,
14 mmol) was added and the reaction mixture was stirred at
room temperature for 1.5 h. The excess reagent was destroyed
by adding ice water. Oxidation was subsequently achieved by
the simultaneous dropwise addition of 3M aqueous NaOAc (14
mL) and 30% hydrogen peroxide (14 mL) followed by stirring
overnight at room temperature. The mixture was saturated
with K2CO3 and the phases were separated. The aqueous
phase was washed with THF (2x) and the combined organic
phases were concentrated and purified by flash chromatogra-
phy (EtOAc-MeOH 9:1) to yield 5 (0.48 g, 54%) as a colorless
syrup: [R]D -10.2 (c 0.96, CH2Cl2); 1H NMR (500 MHz, CDCl3)
4.35 (d, 1H, J ) 7.6 Hz), 4.14 (m, 1H), 4.08-3.97 (m, 4H),
3.95-3.75 (m, 14H), 3.60 (m, 1H), 3.53 (t, 2H, J ) 7.1 Hz),
3.43-3.34 (m, 3H), 3.21 (t, 1H, J ) 8.1 Hz), 2.88 (s, 4H), 2.03-
1.88 (m, 10H), 1.75 (m, 2H), 1.58 (m, 2H), 1.51 (m, 2H) ppm;
13C NMR (125.8 MHz, CDCl3) 103.7, 84.6, 82.3, 78.9, 74.9, 71.8,
71.7, 71.0, 70.6, 70.2, 70.1, 61.5, 61.4, 61.0, 60.6, 34.2, 33.3,
33.2, 33.0, 32.9, 32.4, 29.8, 28.2, 25.6 ppm. Anal. Calcd for
The most popular methodologies used for linking
molecules via their amino groups are peptide coupling
utilizing succinimidyl-activated esters, or thiourea-bridg-
ing using isothiocyanates as reaction partners. Thus,
glycocluster 11 could be turned into the biotin-labeled
derivative 17 using a tethered succinimidyl ester of
biotin. The fluorescein-labeled cluster 18 proved to be too
instable for definitive characterization, while for the
amide-linked analogue 19 sufficient 13C NMR spectral
data were obtained. Interestingly the NMR spectra of the
biotinylated cluster 17 changed when the NMR sample
in D2O was stored over several days. In the 13C NMR
spectrum a peak at 55.8 ppm for the central carbon of
the biotin bicycle was decreased while a corresponding
resonance started to show at 62.4 ppm. Correspondingly,
1
the multiplet for the attached proton showing in the H
NMR shifted from 3.91 to 4.31 ppm. To date this
phenomenon remains unexplained.
Su m m a r y
In conclusion, we have introduced two synthetic path-
ways leading to selectively functionalized cluster glyco-
sides such as 11, in which the amino-functionalized
aglycone can serve as a tether to allow various useful
modifications. This was demonstrated by the synthesis
of biotinylated cluster 17. The key problems connected
with the synthesis and analysis of oligomannosylation
and oligomannosylated clusters were solved so that this
type of molecule is now available for glycobiological
applications such as for the investigation of carbohydrate-
protein interactions and affinity chromatogaphy.
Exp er im en ta l Section
MALDI-TOF Tech n iqu es. For MALDI-TOF measure-
ments, acetone-soluble samples were prepared using the
premix technique with nor-Harmane27 and NaI in acetone as
matrix. Highly polar compounds were prepared using the
layer technique with dihydroxybenzoic acid (DHB) in water-
acetonitrile (2:1) with 0.1% TFA as matrix.25
6-Br om oh exyl 2,3,4,6-Tetr a -O-a cetyl-â-D-glu cop yr a n o-
sid e (2). To a solution of 1 (8 g, 21 mmol) and 1-bromohexanol
(3 mL, 22 mmol) in dry CH2Cl2 (20 mL) boron trifluoride-
etherate (10 mL, 81 mmol) was added. The reaction mixture
was stirred for 1 h in the dark, CH2Cl2 (100 mL) was added
and the solution was neutralized with saturated aqueous
NaHCO3 solution before being washed twice with water. The
organic phase was dried (MgSO4), filtered, concentrated,
purified by column chromatography (light petroleum-EtOAc
3:2) and recrystallized from diethyl ether to yield 2 (3.8 g, 36%)
as a white solid: mp 103 °C; [R]D -21 (c 1.2, CH2Cl2); 1H NMR
(400 MHz, CDCl3) 5.39 (t, 1H, J ) 9.2 Hz), 5.27 (t, 1H, J ) 9.2
Hz), 5.17 (dd, 1H, J ) 9.2, 8.1 Hz), 4.68 (d, 1H, J ) 8.1 Hz),
4.45 (dd, 1H, J ) 4.6, 12.2 Hz), 4.32 (dd, 1H, J ) 2.0, 12.2
Hz), 4.06 (m, 1H), 3.87 (dddd∼m, 1H), 3.67 (m, 1H), 3.58 (t,
1H), 2.27, 2.23, 2.21, 2.19 (4 s, 12H), 2.03 (m, 2H), 1.77 (m,
2H), 1.62 (m, 2H), 1.62 (m, 2H) ppm; 13C NMR (100.6 MHz,
C
24H47BrO10: C, 50.09; H, 8.23; Br, 13.88. Found: C, 49.95;
H, 8.39; Br, 13.75.
6-Br om oh exyl 2,3,4,6-Tetr a -O-[3-(2,3,4,6-tetr a -O-ben -
zoyl-r-D-m a n n op yr a n osyloxy)p r op yl]-â-D-glu cop yr a n o-
sid e (8). A mixture of 5 (0.055 g, 0.096 mmol) and 7 (2.8 g,
3.8 mmol) in dry CH2Cl2 was concentrated and dried under
high vacuum. The resulting solid was dissolved in dry CH2Cl2
(3 mL). TMS-OTf solution (0.1 mL, 0.5 mL TMS-Tf 0.5 mL in
10 mL CH2Cl2) was added and the reaction mixture was stirred
for 6 h at room temperature before being neutralized with
NaHCO3 (10 g), filtered, concentrated and purified by flash
chromatography (light petroleum-EtOAc 3:2 f 2:3). Subse-
quent gel filtration on Sephadex LH-20 (CH2Cl2-MeOH 1:1)
(28) McReynolds, K. D.; Hadd, M. J .; Gervay-Hague J . Bioconjugate
Chem. 1999, 10, 1021.
(29) Weatherman, R. V.; Kiessling, L. L. J . Org. Chem. 1996, 61,
534.
(30) Cao, S.; Tropper, F. D.; Roy, R. Tetrahedron 1995, 51, 6679.

5280 J . Org. Chem., Vol. 65, No. 17, 2000
Dubber and Lindhorst
yielded 8 (262 mg, 95%) as an amorphous white solid: [R]D
-45.9 (c 0.23, CH2Cl2); 1H NMR (500 MHz, CDCl3) 8.10-8.06,
8.04-7.97, 7.94-7.86, 7.83-7.78, 7.57-7.50 (m, each 8H),
7.48-7.16 (m, 40H + CDCl3), 6.18-6.09 (m, 4H), 5.98-5.89
(m, 4H), 5.73, 5.72, 5.70, 5.69 (each dd, 4H, J ) 1.6 and 3.1
Hz), 5.15, 5.13, 5.11, 5.06 (each d, 4H, J ) 1.6 Hz), 4.73-4.64
(m, 4H), 4.52-4.41 (m, 8H), 4.25 (d, 1H, J ) 7.6 Hz), 4.05-
3.65 (m, 19H), 3.49 (m, 1H), 3.37-3.31 (m, 3H), 3.28 (t, 2H, J
) 6.9 Hz), 3.12 (t, 1H, J ) 8.1 Hz), 2.12-1.96 (m, 8H), 1.80-
the reaction mixture was stirred overnight at room tempera-
ture. The reaction was quenched with ice water at 0 °C and
toluene (100 mL) was added. The organic phase was separated,
consecutively washed with aqueous sodium chloride solution
(2×) and water (6×), dried (MgSO4), filtered, concentrated and
purified by flash chromatography (toluene-EtOAc 7:1) to yield
13 (4.22 g, 70%) as a colorless syrup: [R]D +134.8 (c 1.27,
CH2Cl2); 1H NMR (400 MHz, CDCl3) 6.03-5.77 (m, 8H), 5.34-
5.20 (m, 8H), 5.20-5.07 (m, 10H), 4.36, 4.32, 4.25 (each m,
each 2H), 4.16-4.01 (m, 10H), 3.96 (m, 2H), 3.70 (t, 2H, J )
9.2 Hz), 3.62 (dd, 2H, J ) 10.7, 3.6 Hz), 3.54 (dd, 2H, J ) 10.7,
2.0 Hz), 3.45 (t, 2H, J ) 9.7 Hz), 3.37 (dd, 2H, J ) 9.7, 4.1 Hz)
ppm; 13C NMR (125.8 MHz, CDCl3) 135.9, 135.5, 135.2, 135.0,
117.5, 117.1, 117.1, 116.6, 94.6, 81.6, 79.6, 77.7, 74.6, 74.3, 72.9,
72.2, 70.8, 68.8 ppm. Anal. Calcd for C36H54O11: C, 65.24; H,
8.21. Found: C, 65.23; H, 8.18.
1.73 (m, 2H), 1.64-1.56 (m, 2H), 1.42-1.30 (m, 4H) ppm; 13
C
NMR (100.6 MHz, CDCl3) 166.1, 165.-165.2, 133.3-133.0,
129.9-128.2, 103.4, 97.7 (3×), 97.6, 84.9, 82.4, 78.1, 74.8, 70.5
(2×), 70.5 (2×), 70.3, 70.2, 70.2 (2×), 70.2, 69.8, 69.7, 69.5,
69.1, 68.8 (3×), 68.7, 68.0, 66.1, 66.0, 65.4 (2×), 66.9, 66.9 (2×),
66.9, 62.8 (4×), 33.9, 32.7, 30.6, 30.4 (2×), 29.7, 29.5, 27.9, 25.2
ppm. Anal. Calcd for C160H151BrO46: C, 66.50; H, 5.27; Br, 2.77.
Found: C, 66.48; H, 5.22; Br, 2.80.
2,3,4,6,2′,3′,4′,6′-Octa-O-(3-h ydr oxypr opyl)tr eh alose (14).
To a solution of 13 (0.37 g, 0.56 mmol) in dry THF (20 mL),
9-BBN-H (0.5 M solution in THF, 20 mL, 10 mmol) was added
and the reaction mixture was stirred for 1 h under reflux. The
excess of reagent was destroyed with ice water and then
oxidation was achieved by treatment with 3 M aqueous NaOH
(10 mL) and dropwise addition of 30% hydrogen peroxide (10
mL). The reaction mixture was stirred overnight and saturated
with K2CO3, and the phases were separated. The aqueous
phase was washed with THF (2×) and the combined organic
phases were concentrated and purified on Sephadex LH-20
(MeOH) followed by flash chromatography (CH2Cl2-MeOH
3:1) to yield 14 (372 mg, 83%) as a colorless syrup: [R]D +70.1
(c 2.19, MeOH); 1H NMR (400 MHz, MeOH-d4) 5.20 (d, 2H, J
) 3.1 Hz), 3.98-3.84 (m, 8H), 3.79-3.58 (m, 32H), 3.35-3.28
(m, 4H), 1.92-1.77 (m, 16H) ppm; 13C NMR (100.6 MHz,
MeOH-d4) 94.9, 83.8, 82.7, 80.5, 73.3, 72.6, 72.0, 71.9, 70.7,
70.5, 61.5, 61.3, 61.2, 61.2, 35.7, 35.5, 35.4, 34.8 ppm; MALDI-
TOF-MS m/z 829.4 [M + Na]+, calcd for C36H70O19 806.5. Anal.
Calcd for C36H70O19 2 H2O: C, 51.29; H, 8.85. Found: C, 51.69;
H, 8.80.
2,3,4,6,2′,3′,4′,6′-Octa -O-[3-(2,3,4,6-Tetr a -O-ben zoyl-r-D-
m a n n op yr a n osyloxy)p r op yl)]tr eh a lose (15). A solution of
14 (0.050 g, 0.062 mmol) and 7 (5 g, 6.7 mmol) in dry
acetonitrile (400 mL) was heated to 65 °C, TMS-OTf (0.5 mL)
was added and the reaction mixture was stirred at this
temperature for 1 h. Additional 7 (1.2 g, 1.6 mmol) was added
and the solution was stirred overnight at room temperature.
The reaction mixture was neutralized with NaHCO3 (10 g),
filtered, concentrated and purified on Sephadex LH-20 (CH2Cl2-
MeOH 1:1) followed by flash chromatography (hexanes-EtOAc
9:11) to yield 15 (318 mg, 95%) as a white amorphous solid:
[R]D -20.3 (c 0.64, CH2Cl2); 1H NMR (500 MHz, DMSO-d6, 403
K) 8.02-7.92 (m, 32 H), 7.87-7.82 (m, 16 H), 7.72-7.65 (m,
16 H), 7.64-7.56 (m, 16 H), 7.52-7.20 (m, 80 H), 6.03-5.95
(m, 8 H), 5.88, 5.85, 5.85, 5.84 (each dd, each 2H, J ) 2.8, 9.9
Hz), 5.71, 5.68, 5.68, 5.65 (each dd, each 2H, J ) 1.7, 3.0 Hz),
5.24 (d, 2 H, J ) 3.6 Hz), 5.23, 5.18, 5.16, 5.13 (each d, each
2H, J ) 1.7 Hz), 4.67-4.47 (m, 24H), 4.05-3.64 (m, 40H), 3.40
(t, 2 H, J ) 9.1 Hz), 3.39 (dd, 2 H), 2.11-1.95 (m, 16H) ppm;
13C NMR (125.8 MHz, CDCl3) 167.5-166.5, 134.5-134.3,
131.3-129.5, 99.3, 99.0, 89.9, 89.9, 95.1, 83.0, 81.7, 79.1, 72.3,
72.0, 71.9, 71.9, 71.8, 71.7, 71.6, 71.5, 71.5, 70.2, 70.1, 70.1,
70.0, 70.7, 70.5, 69.6, 69.3, 68.3 (3×), 68.1, 67.5, 67.2, 66.6 (2×),
64.2, 64.2, 64.1 (2×), 32.0, 31.8, 31.7, 31.3 ppm; MALDI-TOF-
MS m/z 5454.3 [M + Na]+, calcd for C308H278O91 5431.7. Anal.
Calcd for C308H278O91: C, 68.06; H, 5.16. Found: C, 67.77; H,
5.06.
6-Br om oh exyl 2,3,4,6-Tetr a -O-[3-(r-D-m a n n op yr a n osyl-
oxy)p r op yl]-â-D-glu cop yr a n osid e (9). To a solution of 8
(0.418 g, 0.145 mmol) in THF (20 mL) a solution of NaOMe in
MeOH (0.1 M, 10 mL) was added. The reaction mixture was
stirred for 1 h at room temperature and then concentrated.
The residue was dissolved in MeOH (50 mL), then a solution
of NaOMe in MeOH (0.1 M, 10 mL) was added and the reaction
mixture was stirred overnight at room temperature before
being neutralized with Amberlite IR 120, filtered, concentrated
and purified by gel permeation chromatography on Sephadex
LH-20 (MeOH) to yield 9 (167 mg, 94%) as a colorless syrup:
[R]D + 47.1 (c 1.9, MeOH); 1H NMR (500 MHz, MeOH-d4) 4.81,
4.80, 4.80, 4.79 (each d, 4H, J ) 1.6 Hz), 4.30 (d, 1H, J ) 7.6
Hz), 3.96-3.82 (m, 17H), 3.79-3.54 (m, 27H), 3.50 (t, 2H, J )
6.8 Hz), 3.34 (m, 1H + MeOH), 3.31-3.26 (m, 2H), 3.03 (t,
1H, J ) 8.3 Hz), 1.98-1.85 (m, 10H), 1.67 (m, 2H), 1.58-1.43
(m, 4H) ppm; 13C NMR (100.6 MHz, MeOH-d4) 105.9, 102.8
(4×), 87.2, 84.5, 80.5, 77.0, 75.8 (3×), 75.7, 73.8 (4×), 73.4 (4×),
72.8, 71.8 (4×), 70.5, 69.8 (2×), 69.7 (2×), 66.9, 66.8, 66.5 (2×),
64.1 (4×), 35.7, 35.1, 32.9, 32.8, 32.7, 32.1, 31.8, 30.1, 27.6 ppm;
MALDI-TOF-MS m/z 1245.2 [M + Na]+, calcd for C48H87BrO30
1222.4.
6-Azid oh exyl 2,3,4,6-Tetr a -O-[3-(r-D-m a n n op yr a n osyl-
oxy)p r op yl]-â-D-glu cop yr a n osid e (10). A solution of 9
(0.172 g, 0.141 mmol) in DMF (20 mL) was treated with
sodium azide (0.056 g, 0.860 mmol) and the reaction mixture
stirred at 60 °C for 1 h. The solution was concentrated and
purified on Sephadex LH-20 (MeOH) yielding 10 (163 mg, 97%)
as a colorless syrup: [R]D + 40.1 (c 1.7, MeOH); 1H NMR (500
MHz, MeOH-d4) 4.82-4.78 (m, 4H), 4.30 (d, 1H, J ) 7.6 Hz),
3.97-3.82 (m, 17H), 3.79-3.54 (m, 27H), 3.36-3.32 (m, 3H +
MeOH), 3.30-3.26 (m, 2H), 3.03 (t, 1H, J ) 8.0 Hz), 1.98-
1.85 (m, 8H), 1.71-1.61 (m, 4H), 1.53-1.42 (m, 4H) ppm; 13
C
NMR (100.6 MHz, MeOH-d4) 105,9, 102.8 (3×), 102.7, 87.2,
84.9, 80.5, 77.0, 75.8 (3×), 75.7, 73.8 (4×), 73.4 (4×), 72.8, 71.9,
71.8 (3×), 70.5, 69.8 (2×), 69.8, 69.7, 66.9, 66.8, 66.5 (2×), 64.1
(4×), 53.6, 32.9, 32.7, 32.1, 31.9, 31.1, 28.7, 28.0 ppm; MALDI-
TOF-MS m/z 1208.3 [M + Na]+, calcd for C48H87N3O30 1185.5.
6-Am in oh exyl 2,3,4,6-Tetr a -O-[3-(r-D-m a n n op yr a n osyl-
oxy)p r op yl]-â-D-glu cop yr a n osid e (11). A solution of 10
(0.159 g, 0.134 mmol) in MeOH (20 mL) was stirred with 5%
Pd-C under 1 atm of hydrogen for 2 h. The mixture was
filtered over a thin bed of Sephadex LH-20, which was washed
with MeOH, and the filtrate was concentrated to give 11 (149
1
mg, 96%) as a colorless syrup: [R]D + 55.1 (c 1.1, MeOH); H
NMR (400 MHz, MeOH-d4) 4.82-4.78 (m, 4H), 4.30 (d, 1H, J
) 7.6 Hz), 3.97-3.82 (m, 17H), 3.79-3.53 (m, 27H), 3.40-3.33
(m, 1H + MeOH), 3.30-3.25 (m, 2H), 3.03 (t, 1H, J ) 7.8 Hz),
2.74 (t, 1H, J ) 7.1 Hz), 1.98-1.85 (m, 8H), 1.71-1.63 (m, 2H),
1.62-1.52 (m, 2H), 1.51-1.39 (m, 4H) ppm; 13C NMR (125.8
MHz, MeOH-d4) 105,8, 102.8 (4×), 87.2, 84.9, 80.5, 77.0, 75.8
(4×), 73.8 (4×), 73.4 (4×), 72.8, 71.9, 71.9, 71.8 (2×), 70.5, 69.8,
69.8, 69.8, 69.7, 66.9, 66.8, 66.6 (2×), 64.1 (4×), 43.3, 33.9, 32.9,
32.7, 32.7, 32.1, 31.9, 28.8, 28.2 ppm; MALDI-TOF-MS m/z
1182.3 [M + Na]+, calcd for C48H89NO30 1159.6.
6-d-Biotin am idoh exyl 2,3,4,6-Tetr a-O-[3-(r-D-m an n opy-
r a n osyloxy)p r op yl]-â-D-glu cop yr a n osid e (17). To a solu-
tion of 11 (0.026 g, 0.022 mmol) in aqueous NaHCO3 (0.02M,
8 mL), sulfosuccinimidyl biotin (0.010 g, 0.023 mmol) was
added. The solution was stirred for 5 h at room temperature,
then poured onto Biogel-P2 and eluted with water to yield 17
(31 mg, quant.) as a white lyophilisate: 1H NMR (500 MHz,
D2O) 4.80 (s, 4H), 4.55 (dd, 1H, J ) 4.6, 8.1 Hz), 4.37 (m, 2H),
3.92-3.68 (m, 28H), 3.65-3.51 (m, 16H), 3.42 (m, 1H), 3.37-
3.23 (m, 3H), 3.18-3.00 (m, 3H), 2.95 (dd, 1H, J ) 5.1, 13.2
Hz), 2.73 (d, 1H, J ) 12.7 Hz), 2.20 (t, 2H, J ) 7.1 Hz), 1.94-
2,3,4,6,2′,3′,4′,6′-Octa -O-a llyl-tr eh a lose (13). To a suspen-
sion of 12 (3.1 g, 9.1 mmol) in dry DMF (100 mL), NaH (3.8 g,
99 mmol) and allyl bromide (8 mL, 95 mmol) were added and

Synthesis of Oligosaccharide Mimetics
J . Org. Chem., Vol. 65, No. 17, 2000 5281
1.81 (m, 8H), 1.75-1.53 (m, 6H), 1.50-1.44 (m, 2H), 1.41-
1.27 (m, 6H) ppm; 13C NMR (100.6 MHz, D2O, DEPT) 102.8,
100.1 (4×), 84.2, 81.9, 78.2, 74.0, 73.1 (4×), 71.4, 71.1, 71.0
(4×), 70.7, 70.6, 70.5 (4×), 69.2, 68.5, 67.1 (4×), 64.8 (2×), 64.7-
(2×), 62.5, 61.3 (4×), 60.6, 56.5, 55.8, 40.1, 39.6, 35.9, 29.9,
29.8, 29.7, 29.1, 29.0, 28.7, 28.2, 28.1, 26.1, 25.6, 25.3 ppm;
MALDI-TOF-MS m/z 1408.7 [M + Na]+, calcd for C58H103N3O32S
1385.6.
6-Ca r b oxyflu or escein yla m in oh exyl 2,3,4,6-Tet r a -O-
[3-(r-D-m a n n op yr a n osyloxy)p r op yl]-â-D-glu cop yr a n o-
sid e (19). To a solution of 11 (0.055 g, 0.047 mmol) and
NaHCO3 (100 mg) in MeOH (8 mL), 5(6)-carboxy-fluorescein-
N-succinimidyl ester (25 mg) was added. The yellow solution
was stirred for 5 h at room temperature in the dark, then
poured onto Sephadex LH-20 and eluted with MeOH to yield
19 (21 mg, 29%) as a very instable light-sensitive, red solid:
13C NMR (100.6 MHz, D2O, DEPT) 131.7, 129.3, 128.6, 123.6,
102.8, 100.2 (4×), 84.2, 81.8, 78.1, 73.9, 73.1 (4×), 71.2, 71.0
(4×), 70.6, 70.5 (4×),70.3 (2×), 67.2 (2×), 67.1 (2×), 64.8-64.7
(4×), 61.3-61.2 (4×), 40.4, 29.9, 29.8 (2×), 29.0 (2×), 28.4, 25.9,
25.1 ppm; MALDI-TOF-MS m/z 1540.3 [M + Na]+, calcd for
C
69H99NO36 1517.6.
Ack n ow led gm en t. This work was financed by the
BMBF (Glycobiotechnology Grant) and the FCI. We are
grateful to Dr. Volker Sinnwell for measuring the
sophisticated NMR spectra, to Claudia Heuer for tech-
nical assistance, and to Angela Scheppokat B.Sc. (Hons)
for English corrections.
Su p p or tin g In for m a tion Ava ila ble: Additional experi-
mental data for 7, 14, 18, and 19. This material is available
free of charge via the Internet at http://pubs.acs.org.
J O000432S