Oligomannoside mimetics by glycosylation of 'octopus glycosides' and their investigation as inhibitors of type 1 fimbriae-mediated adhesion of Escherichia coli

Article abstract of DOI:10.1039/b610741a

The glycocalyx of eukaryotic cells is composed of glycoconjugates, which carry highly complex oligosaccharide portions. To elucidate the biological role and function of the glycocalyx in cell-cell communication and cellular adhesion processes, glycomimeti

Full text of DOI:10.1039/b610741a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Oligomannoside mimetics by glycosylation of ‘octopus glycosides’ and their
investigation as inhibitors of type 1 fimbriae-mediated adhesion of Escherichia
coli
Michael Dubber, Oliver Sperling and Thisbe K. Lindhorst*
Received 26th July 2006, Accepted 5th September 2006
First published as an Advance Article on the web 25th September 2006
DOI: 10.1039/b610741a
The glycocalyx of eukaryotic cells is composed of glycoconjugates, which carry highly complex
oligosaccharide portions. To elucidate the biological role and function of the glycocalyx in cell–cell
communication and cellular adhesion processes, glycomimetics have become targets of glycosciences,
which resemble the composition and structural complexity of the glycocalyx constituents. Here, we
report about the synthesis of a class of oligosaccharide mimetics of a high-mannose type, which were
obtained by mannosylation of spacered mono- and oligosaccharide cores. These carbohydrate-centered
cluster mannosides have been targeted as inhibitors of mannose-specific bacterial adhesion, which is
mediated by so-called type 1 fimbriae. Their inhibitory potencies were measured by ELISA and
compared to methyl mannoside as well as to a series of mannobiosides, and finally to the
polysaccharide mannan. The obtained results suggest a new interpretation of the mechanisms of
bacterial adhesion according to a macromolecular rather than a multivalency effect.
Introduction
which only glycosylation leads to products which closely resemble
the composition of structures found in nature. This approach is
outlined in Fig. 1. As octopus-type core molecules, uniformly O-
Eukaryotic cells are covered by a nano-dimensioned carbohydrate
layer, which is called the glycocalyx. The glycocalyx consists of
a large number of structurally highly diverse glycoconjugates,
which are partly embedded in the cell membrane with their non-
carbohydrate moiety or stick to the cell surface by other non-
covalent interactions.
Research in the field of glycobiology has revealed that the
glycocalyx is essential for cellular communication however, the
principles and mechanisms underlying its biological function are
not well understood and are difficult to investigate.
Particularly, this is due to the enormous complexity of the gly-
cocalyx, which can be regarded as a molecular super-system. One
useful approach in coping with the problems of experimental car-
bohydrate biology is the synthesis and investigation of molecules,
which mimic the highly complex and hyperbranched character of
the glycoconjugates found in the glycocalyx. Many examples of
(
x-hydroxy-spacer)-modified glycosides are employed, which are
turned into carbohydrate-centered cluster glycosides by classic
glycosylation. Degradation of such oligosaccharide mimetics
should lead to compounds found in the natural environment as
the only degradation products, namely the sugar components and
ethylene or propylene glycol.
The degree of difficulty of this synthetic pathway is mainly
defined by the glycosylation step, which is complicated by steric
hindrance occurring in the formed products. This demanding
oligo-glycosylation reaction can be regarded as a multi-step
synthesis as the reactivity of partially glycosylated intermediates
decreases with increasing degree of glycosylation, the last glyco-
sylation step becoming the most difficult one.
1
2
We have recently tackled this problem in the preparation of
anomerically functionalized carbohydrate-centered glycoclusters
and have achieved the exhaustive mannosylation of four hydroxyl
3
such glycomimetics have been published and reviewed. Among
the different architectures which have been employed for the
conjugation of glycoligands, the use of carbohydrates as core
molecules has led to the synthesis of the so-called carbohydrate-
6
groups using the mannosyl donor 3 (Scheme 1). Then it has
become our goal to establish this chemistry for the exhaustive
glycosylation of octopus glycosides based on monosaccharide as
4
centered glycoclusters.
These often large molecules have been built up using monosac-
charide cores, which were uniformly spacer-modified at their
hydroxyl groups to prevent steric hindrance upon ligation with
peripheral carbohydrates. For such spacered core molecules the
name ‘octopus glycosides’ has been coined alluding to their multi-
5
arm like shape.
Many different chemistries have been used for ligation of
4
glycoligands to the oligofunctional carbohydrate cores, among
Otto Diels Institute of Organic Chemistry, Christiana Albertina University
of Kiel, Otto-Hahn-Platz 4, 24098, Kiel, Germany. E-mail: tklind@oc.uni-
kiel.de; Fax: (+49)431 880 7410
Scheme 1
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 3901–3912 | 3901

View Article Online
Fig. 1 This cartoon exemplifies the structure of O-(x-hydroxy-spacer)-modified glycosides (I), which can be used as oligofunctional scaffold molecules
of an “octopus-type”. They can be turned into carbohydrate-centered cluster glycosides (II) by classic glycosylation.
well as oligosaccharide cores. This would allow us to access a class
of high-molecular weight oligosaccharide mimetics, which in the
case of hydrolysis, deliver propylene oxide as the only byproduct
in addition to the sugar constituents.
Results and discussion
Synthesis of the mannosyl donor
For mannosylation of the diverse spacer-modified carbohydrate
cores, the trichloroacetimidate method was selected. To reduce the
The first carbohydrate core we used in these experiments was
D-glucose, and for further studies we selected two disaccharides
and one trisaccharide (cf. Scheme 2a). For glycosylation of the
carbohydrate-derived core molecules we chose a-mannosylation
for two reasons: (i) high mannose-type glycoclusters are of consid-
9
probability of glycosyl orthoester formation during the mannosy-
lation step benzoyl protective groups were used instead of acetyl
10
groups. As we anticipated that we would need a considerable
excess of glycosyl donor to achieve high yields, we first optimized
the synthesis of the mannosyl trichloroacetimidate 3 (Scheme 1).
The synthesis of the reducing sugar 2 from the pentabenzoate 1
7
erable biological interest and (ii) a-mannosylation is afflicted with
8
the problem of glycosyl orthoester formation which we wished to
solve.
Scheme 2
3
902 | Org. Biomol. Chem., 2006, 4, 3901–3912
This journal is © The Royal Society of Chemistry 2006

View Article Online
could be significantly improved. While anomeric deacetylation is
easily achieved in good yield by different procedures, regioselec-
It turned out that orthoester formation during the attempted
synthesis of 7 could be avoided when a large excess of the
imidate 3 and elevated TMSOTf concentrations were used in
the glycosylation step. Exhaustive mannosylation of all hydroxyl
groups was accomplished in very concentrated solution, using
11
tive 1-O-debenzoylation is less common, especially in the manno
12
series. We found that ethanolic dimethylamine can be used to
effect anomeric debenzoylation of 1 in pyridine solution, yielding
13
2
in nearly 80% after 1.5 h. In addition, the synthesis of the
10 mg acceptor in 1 ml CH
2
2
Cl . This optimized mannosylation
imidate 3 was optimized by a prolonged reaction time of 12 h,
protocol delivered the target cluster mannoside 7 in quantitative
yield. The H NMR spectrum of this cluster is complex however,
1
which increased the yield in this step from the reported 75% (3 h
14
reaction time) to 97%.
it can be fully assigned. The chiral core glucose influences its
periphery, resulting in different signal sets for the protons of each
individual mannosyl residue.
Oligomannosylation of a spacered glucose core
16
Subsequent deacylation of 7 under Zempl e´ n conditions finally
furnished the unprotected target glycocluster 8 after 2 days, in high
overall yield.
The mannosyl trichloroacetimidate 3 was then employed for
the glycosylation of all five hydroxyl groups of the octopus
5
glucoside 6 (Scheme 3). As previously reported, this uniformly
When this synthetic pathway had been elaborated with glucose
as the core, we set out to apply this successful approach to
di- and oligosaccharides. Three sugars were selected as starting
material, the disaccharide allyl melibioside (mel series) and the
non-reducing glycosides trehalose (tre series) and raffinose (raf
series) (Scheme 2a). Whereas melibiose and trehalose are eight-
functional scaffolds, suited to the assembly of eight mannosyl
residues, even eleven sugar rings can be attached to a raffinose-
derived scaffold molecule.
O-(3-hydroxypropyl)-modified pentanol can be readily prepared
starting from the fully O-allylated glucose derivative 5, which is
derived from allyl a-D-glucoside (4). A hydroboration–oxidation
sequence leads to 6 in high yield. The published procedure for the
5
synthesis of 6 was modified in that the reaction mixture was kept
at reflux temperature for 1 h during hydroboration.
Firstly, we anticipated that the reaction conditions established
for the synthesis of the glucose-centered derivatives 5–8 (Scheme 3)
would also be suited to perform the analogous chemistry starting
from allyl melibioside (4-mel), trehalose (4-tre) and raffinose
(
4-raf), respectively. However, it turned out that some reaction
conditions had to be optimized for every individual sugar which
was used as the carbohydrate core.
While for the allylation of 4, phase transfer catalysis with
TBABr, 33% NaOH and allyl chloride was successful (Scheme 3),
perallylation of the higher polar trehalose 4-tre under the same
conditions gave only poor results. Perallylation of 4-tre to form
5
-tre was achieved with NaH and allyl bromide in DMF. Hydro-
17
boration of perallylated trehalose 5-tre to form 6-tre, as well as
of the melibiose- and the raffinose-centered analogs 5-mel and 5-
raf, proceeded as with the glucose derivative 5 however, instead
of using MgSO as a drying reagent, Sephadex LH-20 had to be
4
employed with dry methanol as the eluent.
For the critical mannosylation step, which leads to the desired
carbohydrate-centered oligomannoside mimetics (Scheme 2b), the
solubility of the carbohydrate-centered oligo-alcohols was an
important parameter. Because 6-tre, 6-mel, and 6-raf were of
higher polarity than the glucose-centered pentaol 6, which is
even soluble in CH
more polar solvent than CH
trehalose-centered octaol 6-tre with 3 as the donor was possible
in acetonitrile. In contrast to glycosylation of 7, mannosylation of
6-tre in acetonitrile had to be carried out in very dilute solution
with less that 0.1 mg acceptor per 1 mL acetonitrile.
2
Cl
2
, glycosylation had to be performed in a
Cl . Complete mannosylation of the
2
2
7
Scheme 3
1
The first attempt of mannosylation of the core molecule 6 using
the glycosyl donor 3 to achieve 7 was hampered by the formation
of glycosyl orthoester byproducts. This finding was unexpected
because it has been reported that the benzoyl-protected mannosyl
trichloroacetimidate 3 does not to lead to orthoester formation
in other difficult glycosylation reactions such as in case of the
mannosylation of a tether-functionalized carbohydrate-centered
Our search for optimized mannosylation conditions was
doomed to navigate between three undesired complications (i)
incomplete mannosylation, thus leading to structural defects in the
respective product, (ii) orthoester formation and (iii) cleavage of an
interglycosidic bond. The reaction conditions had to be carefully
fine-tuned for each case in order to obtain perfect products.
Optimization of the various glycosylation parameters was possible
by close monitoring the glycosylation reaction and the subsequent
6
15
tetraol or in the mannosylation of sterically hindered polyols.
Unfortunately, formation of glycosyl orthoester byproducts dur-
ing the mannosylation of spacered carbohydrate cores remained a
severe problem throughout all the work reported here.
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 3901–3912 | 3903

View Article Online
deprotection step by MALDI-TOF MS. The optimized reaction
conditions reported here (cf. Experimental section), resulted from
zoylated glucose-centered cluster glycoside 7 could be deprotected
16
under standard Zempl e´ n conditions, a stepwise deprotection
protocol was necessary for the oligosaccharide-centered cluster
mannosides. In this procedure the benzoylated sugar was first
dissolved in THF and treated with methanolic sodium methox-
ide solution. After 1–4 h the partially deprotected glycocluster
precipitated from the THF solution, which was concentrated
and subsequently subjected to a standard Zempl e´ n procedure in
methanol, leading to fast and complete debenzoylation in nearly
quantitative yields. The acidic ion exchange resin Amberlite IR
120, which was used after deprotection for neutralization also
effected cleavage of orthoesters, when these had been formed
during the glycosylation step.
2
5 glycosylation experiments.
To achieve complete mannosylation of 6-tre, 6-mel, as well as
-raf, higher temperatures at the beginning of the glycosylation
6
reaction turned out to be beneficial. However, the interglycosidic
bonds in trehalose, melibiose, and raffinose, respectively, are of
different stability and are sensitive to elevated reaction tempera-
tures. When the raffinose-centered polyol 6-raf (Scheme 2b) and
the melibiose-centered polyol 6-mel were glycosylated for 2 h at
◦
8
0 C, the a(1 → 6)-interglycosidic bond in melibiose remained
unaffected, whereas the a(1 → 1)-interglycosidic linkage between
the melibiose and fructose subunits of raffinose was completely
cleaved. Finally we found out that for the mannosylation of the
melibiose-centered oligo-alcohol 6-mel the reaction mixture was
Then we wished to further extend the elaborated synthetic
protocol to spacered carbohydrate scaffolds with varied spacer
lengths. This was of interest as the spacer length of cluster
mannosides has been shown to be an important parameter when
◦
best kept at 75 C for 2 h and then stirred at room temperature,
whereas in the case of the trehalose analog, 6-tre stirring for 2 h
◦
◦
20
at 60 C gave the best results. The raffinose analog 6-raf was kept
ligand–receptor interactions are to be optimized. Indeed, the
at 60 C for only 10 minutes before the glycosylation mixture was
strategy reported herein for the synthesis of carbohydrate-centered
glycoclusters could be adapted to the hydroxyethyleneglycol-
modified glucoside 11 (Scheme 4). The latter was obtained
further stirred at room temperature.
The excess of mannosyl donor 3, which was used, as well
as the amount of Lewis acid employed, could be tuned to an
optimum, which was common for all three oligosaccharide series.
The reaction time was of minor importance for the result of the
glycosylation reaction.
5
from 5 by ozonisation and reductive work-up leading to 9,
followed by perallylation to deliver 10, which was finally submitted
to hydroboration–oxidation. Complete glycosylation of the 6-
hydroxy-3-oxa-hexyl-spacered octopus glycoside 11 employing
3 as the mannosyl donor led to 12 in yields around 70%.
Deprotection to the OH-free cluster mannoside 13 with extended
spacers was possible according to Zempl e´ n in quantitative yield.
At this point it can be concluded that the methodology reported
here allows the synthesis of multivalent glucose-, trehalose-,
melibiose-, and raffinose-centered cluster mannosides. Optimized
reaction conditions for each sugar series and each synthetic step
have been elaborated. The carbohydrate-centered glycoclusters
8, 8-tre, 8-mel, 8-raf, and 13 closely resemble the elemental
composition of the natural example structures and thus are prime
candidates as oligosaccharide mimetics of a high-mannose type.
Spectroscopic characterization of the carbohydrate-centered
cluster mannosides 7-mel, 7-tre and 7-raf, respectively, was
possible by means of 2D-NMR and mass spectrometry. As
the formation of b-mannosides during mannosylation reactions
using trichloroacetimidates in acetonitrile was reported in the
18
literature, the anomeric configuration of the prepared cluster
mannosides was confirmed by means of their JC-1,H-1 hetero
coupling constants, which are significantly lower for a-mannosides
19
than for the respective b-anomers.
Unfortunately, deprotection of the O-benzoylated glycoclusters
-tre, 7-mel and 7-raf (Scheme 2b) was not trivial. While the ben-
7
Scheme 4
3
904 | Org. Biomol. Chem., 2006, 4, 3901–3912
This journal is © The Royal Society of Chemistry 2006

View Article Online
Inhibition of type 1 fimbriae-mediated bacterial adhesion
bacterial adhesion to the mannan surface is inhibited by an inves-
tigated inhibitor. Results in triplicate were used for plotting of the
inhibition curves for each individual ELISA experiment (Fig. 2).
Typically, the IC50 values obtained from several independently
performed tests were in the range of ± 15% for example, IC50 values
determined for the standard MeMan varied between 1.2 and 6.9
millimolar. However, the relative inhibitory potencies calculated
from independent series of data were highly reproducible.
Many biological processes occurring at cell surfaces are dependent
21
on interactions with high-mannose glycoconjugates. Bacterial
adhesion is also often mediated through the interactions of
bacterial fimbriae with mannose-containing conjugates of the host
cell glycocalyx. Fimbriae are adhesive proteinogenous organells
22
on the bacterial surface, which carry lectin domains. Escherichia
coli possess so-called type 1 fimbriae, which have been attributed
a specificity for terminal a-mannosyl residues. The lectin portion
The IC50s determined for an individual cluster mannoside were
related to the IC50 value of the standard inhibitor MeMan,
which was measured on the same microtiter plate affording a
relative ranking of the tested compounds regarding their inhibitory
potency. Relative inhibitory potencies (RIP values) were thus
based on the inhibitory potency of MeMan in the same test
with RIP (MeMan) ≡ 1 (Table 1). Rather unnatural synthetic
mannosides such as methylumbelliferyl a-D-mannoside were not
23
of type 1 fimbriae has been identified as the protein FimH.
The oligomannoside mimetics prepared herein have been tar-
geted to investigate the inhibition of mannose-specific bacterial
adhesion to the glycocalyx of their host cells. They were tested for
their potency as inhibitors of type 1 fimbriae-mediated bacterial
adhesion and were expected to perform much better than methyl
a-D-mannoside (MeMan), a known inhibitor of type 1 fimbriae-
27
considered in this study (cf. the subsequent paper ).
24
mediated bacterial adhesion, owing to their complex multivalent
structure. We anticipated varying inhibitory potencies depending
on the number of a-mannosyl residues exposed on the carbohy-
drate core as well as depending on the spatial circumstances of
their orientation.
Unexpectedly, the inhibitory potencies of all carbohydrate-
centered cluster mannosides tested, 8, 8-tre, 8-mel, 8-raf, and
13, were almost the same when compared to MeMan with
RIP values around 200 (Fig. 3). Their inhibitory potency was
increased by two orders of magnitude compared to the simple
methylmannoside. To allow a better assessment of this finding, the
clusters were compared to a series of mannobiosides, allyl 2-O-a-D-
mannosyl-a-D-mannoside (1 → 2Dis), allyl 3-O-a-D-mannosyl-a-
D-mannoside (1 → 3Dis), allyl 4-O-a-D-mannosyl-a-D-mannoside
Table 1 summarizes the results which were obtained when the
unprotected oligomannoside mimetics 8, 8-tre, 8-mel, 8-raf, and
13 were tested for their potency as inhibitors of type 1 fimbriae-
mediated bacterial adhesion using an enzyme-linked immunosor-
bent assay (ELISA). The microtiter plates used were coated
with mannan from Saccharomyces cerevisiae constituting a highly
branched polysaccharide with an a-1,6-linked polymannoside
backbone from which a-D-mannosyl residues branch out from
(
1 → 4Dis), allyl 6-O-a-D-mannosyl-a-D-mannoside (1 → 6Dis),
and the branched mannotrioside allyl 3,6-di-O-(a-D-mannosyl)-
a-D-mannoside(1 → 3,6Tris), which were prepared according
28
to reported procedures. Whereas, the carbohydrate-centered
25
the 2- and the 3-position. Recombinant Escherichia coli bacteria
glycoclusters prepared herein form a group of inhibitors which
perform around two orders of magnitude better than MeMan
regardless of their valency or spatial arrangement, the tested
mannobiosides hardly exceed the inhibitory potency of MeMan
or are up to one order of magnitude more potent (1 → 3Dis and
26
(
E. coli HB101 pPKl4 ) were employed, carrying type 1 fimbriae
as the only fimbriae which are expressed on the bacterial surface.
IC50 values were determined for each cluster mannoside according
to standard methods and reflect the concentration at which 50% of
1
→ 3,6Tris), respectively (Table 1).
Table 1 Potencies of the prepared carbohydrate-centered cluster manno-
sides, mannobio- and triosides and mannan as inhibitors of mannose-
specific adhesion of E. coli, compared to MeMan, as determined by
ELISA. IC50 values are average values of at least three independent assays
and are listed together with their standard deviations (s.d.). So-called
relative inhibitory potencies (RIP) are relative to the IC50 value measured
for methyl a-D-mannoside (1); thus the inhibitory potency of 1 has been
defined as RIP ≡ 1. All RIP values are average values of at least three
independently determined RIPs and listed together with their standard
deviations
a
Mannose-containing molecules
IC50
s.d.
RIP
1
2.9
11
0.8
1.4
20
180
170
230
250
130
190
190
s.d.
—
MeMan
3.6
1.2
0.5
0.3
0.5
1.1
0.03
0.028
0.038
0.00058
0.0003
0.0032
0.0018
0.0038
b
1
1
1
1
1
8
8
8
8
1
8
→ 2Dis
→ 3Dis
→ 4Dis
→ 6Dis
→ 3,6Tris
1.1c
1.0
0.4
2.8
7.0
0.35
0.5
0.6
32
46
74
74
38
18
51
d
2.2
e
0.15
0.048
0.054
0.016
0.031
0.027
0.041
0.0086
-Tre
-Mel
-Raf
3
Fig. 2 Example for the sigmoidal fitting of the data obtained by ELISA,
from which IC50 values were obtained. On each ELISA plate, the standard
methyl a-D-mannoside (MeMan) was included.
-Mel(def)
Mannan
Our results with the known di- and trisaccharides are in rough
accordance with the literature and can be rationalized according
toligand–receptor interactions, inother words, onthe basis oftheir
a
e
b
28
c
28
d
28
[
mmolar]. Reported: 1.3.
Reported: 1.2.
Reported: 0.5.
29
28
Reported: 10.5.
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 3901–3912 | 3905

View Article Online
Fig. 3 Comparison of the relative inhibitory potencies of the prepared carbohydrate-centered cluster mannosides as well as mannan, based on MeMan
(
RIP ≡ 1, not shown). Standard deviations are indicated.
molecular interactions with the carbohydrate-recognition domain
CRD) of the fimbrial lectin FimH. However, the findings collected
This is a very interesting result, suggesting that mannan, as
well as the class of carbohydrate-centered cluster mannosides
prepared and investigated herein, other than the di- and trisac-
charides tested, share a structural feature which is responsible for
inhibition of the adhesion of E. coli in the employed ELISA. This
inhibitory effect can neither be interpreted according to standard
structure–activity relationships nor can it be understood on the
basis of solitary or multivalent interactions of ligands and their
(
for the various cluster mannosides cannot be rationalized this
way. Referring to reported multivalency effects in the inhibition
of type 1 fimbriae-mediated bacterial adhesion, we had expected
much better values (lower IC50s) for the novel hyperbranched
oligomannoside mimetics.
From molecular dynamics simulations, which we have per-
30
31
formed prior to our synthetic work, we have an idea of the
conformational behaviour of our target cluster mannosides 8, 8-
tre, 8-mel, and 8-raf. Modeling had revealed that the carbohydrate-
centered cluster mannosides enclose a spherical conformational
space. The average distance of the cluster center to an exposed
mannosyl residue is in rough accordance with the radius of the
receptors. Multivalency effects, which are important in biology
and especially in glycobiology have been reasoned according to
32
different principles, such as a chelate or a statistical effect; our
findings suggest the addition of a new mechanistic principle to
the knowledge reported so far. The observed inhibitory potencies
may owe to a feature which is typical for macromolecules and
the interactions they form, rather than for distinct molecular
epitopes. We will consider such a macromolecular effect in our
future experiments and we will attempt to elucidate the underlying
mechanistic principles.
˚
respective conformational sphere and this is increasing from 7.8 A
for 8 to 9.2 for 8-tre, 9.4 for 8-mel, and finally 9.7 for 8-raf. These
differences, however, were not reflected in the measured inhibitory
potencies and even a much more spaced cluster mannoside such as
1
3 received a similar ranking as all other cluster mannosides tested.
Moreover, a structurally defective analogue of 8-mel, termed as
-mel(def), hardly showed any difference when compared to the
On the other hand, a second part of our work on the mechanisms
of carbohydrate-based bacterial adhesion must focus on the
detailed inspection of the interactions between the CRD of FimH
and mannoside ligands, and this is discussed in the subsequent
8
perfect cluster 8-mel (Table 1).
27
While we had expected to measure different inhibitory potencies
depending on the valency of a specific cluster and its three-
dimensional characteristics, the findings summarized in Table 1
prompted us to consider an alternative mechanism for the involved
molecular interactions. Our considerations led us to test mannan
itself as an inhibitor of type 1 fimbriae-mediated bacterial adhesion
to the mannan-coated surface of the poylstyrene plate used in the
ELISA. The concentration of mannan was calculated on the basis
of its average molecular weight and this delivered a RIP value of
paper.
Conclusions
We have to conclude that understanding of fimbriae-mediated bac-
terial adhesion might require at least two different points of view.
One perspective deals with the interpretation of results obtained
from hemagglutination inhibition assays or ELISA. Inhibition of
bacterial adhesion as observed in this testing system cannot be
33
1
90. Thus, mannan is ranking in the range of all carbohydrate-
rationalized on the basis of the known crystal structure of FimH
centered cluster mannosides which we have tested before.
however, it can neither be interpreted in the sense of a classical
3
906 | Org. Biomol. Chem., 2006, 4, 3901–3912
This journal is © The Royal Society of Chemistry 2006

View Article Online
“
multivalency effect”. Rather than that, a macromolecular effect
diluted serially two-fold in PBSE. Bacterial suspension (50 ll
per well) was added and the plate was left at 37 C for 1 h to
◦
should be considered as rational for the inhibition of bacterial
adhesion to the gylcocalyx or a glycocalyx mimetic.
Nevertheless, in addition to more such supramolecular consid-
erations, the molecular details of interactions of the FimH CRD
and mannoside ligands have to be investigated, utilizing the known
crystal structure of FimH. This second aspect of our research is
allow sedimentation of the bacteria. Then each well was washed
four times with PBSE (150 ll) and 50 ll of the first antibody (anti-
fimA antibody, solution as optimized prior to the experiments) in
2% skimmed milk was added. The plates were incubated for 30 min
and then washed twice with PBSET and the second antibody
was added (50 ll). The plates were incubated for 30 min and
then washed three times with PBSET and once with PBSE and
substrate buffer. ABTS solution (50 ll) was added, incubated
27
highlighted in the successive contribution.
Experimental
◦
for 60 min at 37 C. For ELISA controls, bacterial adhesion to
General remarks
blocked, uncoated microtiter plates was checked, and the reaction
of the employed antibodies with yeast mannan was tested and
found to be negligible. The percentage inhibition was calculated
Optical rotations were determined with a Perkin Elmer 241
polarimeter (10 cm cells, Na-D-line: 589 nm). NMR spectra
were recorded at 400 or 500 MHz on Bruker AMX-400 and
−1
as [OD(nI) − OD(I) × 100 × [OD(nI)] (nI: no inhibitor, I: with
inhibitor).
Bruker DRX-500 instruments with Me
standard. All reactions were monitored by TLC on silica gel FG254
Merck) with detection by UV light and/or by charring with
0% ethanolic sulfuric acid. Flash column chromatography was
4
Si (d 0) as the internal
IC50 values are average values from at least three independent
assays and are listed together with their standard deviations.
Relative inhibitory potencies (RIPs) are based on the IC50 value of
methyl a-D-mannopyranoside (MeMan), with RIP (MeMan) ≡ 1.
(
1
performed on silica gel 60 (200–400 mesh, Macherey Nagel & Co).
Gel permeation chromatography was carried out on Sephadex G-
Tetra-O-2,3,4,6-benzoyl-mannopyranose (2). To a solution of
1
,2,3,4,6-penta-O-benzoyl-mannose (1, 5.00 g, 7.13 mmol) in
pyridine (50 mL), an ethanolic solution of dimethylamine (5.6 M,
5 mL) was added and the reaction mixture was stirred for 1.5 h
at rt. Then the reaction was quenched by the addition of toluene
100 mL), the organic phase was washed three times with brine,
1
5 (Pharmacia) if not otherwise stated. Elemental analyses were
determined by the Microanalytical Laboratory of the Department
of Organic Chemistry at the University of Hamburg. Optical
densities (ODs) were measured on an Asys DigiScan 400 ELISA
reader at 405 nm with the reference read to 492 nm. ELISA plates
were incubated at 37 C.
Methyl a-D-mannoside was purchased from Fluka, F-shaped
3
(
◦
dried over MgSO , filtered, concentrated and purified by column
4
chromatography (toluene–ethyl acetate, 5 : 1) to yield the product
(3.31 g, 78%) as a white solid. Anal. Calcd. for C34
28
H O10: C, 68.45;
9
6-well microtiter plates from Sarstedt. Mannan from Saccha-
H 4.73. Found: C, 68.60; H, 4.73%.
romyces cerevisiae was purchased from Sigma and was used
−
1
in 50 mM aq. Na
anti-fimA antibody was a kind gift from Prof. Dr J. Hacker
W u¨ rzburg) and peroxidase-conjugated goat anti-rabbit anti-
2
CO
3
(1 mg ml ; pH 9.6). The polyclonal
[
3-O-(2,3,4,6-Tetra-O-benzoyl-a-D-mannopyranosyloxy)propyl]
2,3,4,6-tetra-O-[3-(2,3,4,6-tetra-O-benzoyl-a-D-mannopyranosyl-
(
oxy)propyl]-a-D-glucopyranoside (7). To a solution of the pentaol
body (IgG, H + L) was purchased from Dianova. Skimmed
milk was from Ulzena, Tween 20 from Roth, ABTS [2,2-
azidobis-(3-ethylbenzothiazoline-6-sulfonic acid)] from Fluka and
thimerosal {2-(ethylmercurio)thiobenzoic acid, sodium salt} was
from Merck. A recombinant type 1 fimbriated E. coli strain, E. coli
5
6
2
0
(22 mg, 0.047 mmol) and the mannosyl donor 3 (1.59 g,
Cl (4 mL) TMS-OTf (5% in CH Cl
.2 mL,) was added under N and the solution was stirred at
(1 g) and CH Cl (50 mL) were
.14 mmol) in dry CH
2
2
2
2
,
2
rt overnight. Then NaHCO
3
2
2
added, the solution was filtered, concentrated and the residue
purified by flash chromatography (light petroleum ether–ethyl
acetate, 2 : 1 → 1 : 1) to yield the title cluster mannoside (157 mg,
2
6
34
HB101 (pPKl4), was used and cultured as described earlier.
Buffers. PBS (phosphate-buffered saline) was prepared by
2
D
0
quant.) as a white amorphous solid. [a] −21.9 (c 1.89, CH
2
2
Cl ).
dissolving 8 g NaCl, 0.2 g KCl, 1.44 g Na HPO ·2H O and 0.2 g
KH PO in 1000 ml of bidest. water (pH 7.2). PBSE was PBS
buffer + 100 mg l thimerosal, PBSET was PBSE buffer + 200 ll
2
4
2
1
H NMR (400 MHz, CDCl
3
): d = 8.11–7.91 (m, 30H, aryl-H),
2
4
−
1
7.87–7.77 (m, 10H, aryl-H), 7.56–7.48 (m, 10H, aryl-H), 7.42–
7.14 (m, 50H, aryl-H), 6.17, 6.15, 6.14, 6.14, and 6.14 (each
−
1
l
tween 20. Substrate buffer was 0.1 M sodium citrate dihydrate,
dd ≈ t, 5H, J3,4man
J4,5man 9.1 Hz, H-4man), 6.11, 5.98, 5.91, 5.91,
and 5.91 (each dd, 5H, J2,3man 3.0 Hz, H-3man), 5.77, 5.72, 5.72,
adjusted to pH 4.5 with citric acid. For preparation of the ABTS
solution, ABTS (1 mg per ml) was dissolved in substrate buffer
5.72, 5.70 (each dd, 5H, H-2man), 5.17, 5.14, 5.14, 5.12, 5.11 (d,
5H, J1,2man 1.5 Hz, H-1man), 5.05 (d, 1H, J1,2glc 3.6 Hz, H-1glc),
4.74–4.66 (m, 5H, H-6man), 4.56 (m, 1H, H-5man), 4.53–4.41 (m,
10H, 5 H-6man, 4 H-5man), 4.08–3.62 (m, 24H, H-5glc, 2 H-6glc, 10
and 0.1% H
2
2
O (25 ll per ml) was added.
ELISA
To determine the potencies of the various cluster mannosides
OCHHCH
2
CH
2
O, 10 OCH
2
CH
2
CHHO, H-3glc), 3.50 (dd ∼ t, 1H,
tested as inhibitors of type 1 fimbriae-mediated adhesion of E. coli,
J3,4glc 9.1 Hz, H-4glc), 3.40 (dd, 1H, J2,3glc 9.1 Hz, H-2glc), 2.15–2.04
34
1
1
an ELISA was used as published earlier. Polystyrene microtiter
(m, 10H, OCH
2
CHHCH
2
O) ppm. H- H COSY d = 3.70 (H-3)
3
1
3
plates were coated with mannan solution (100 ll per well) and
ppm. C NMR (100.67 MHz, CDCl
): d = 166.2–166.0, 165.5–
◦
dried overnight at 37 C. The plates were blocked once with
165.2 (20 COOR), 133.4–132.9, 130.0–128.2 (100 aryl-C), 97.8,
97.8, 97.7, 97.7, 97.5 (C-1man), 97.0 (C-1glc), 81.9 (C-3glc), 80.6 (C-
◦
5
% skimmed milk in PBSE for 30 min at 37 C. The wells
were washed with PBSE (150 ll) and then PBSE (50 ll) and
inhibitor solutions (50 ll) were added. Inhibitor solutions were
2
glc), 77.8 (C-4glc), 70.6, 70.6, 70.5, 70.5, 70.5 (C-2man), 70.5 (C-5glc),
70.4, 70.3, 70.3, 70.2, 70.2 (C-3man), 70.2 (CH CH OC-3glc), 69.4
2
2
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 3901–3912 | 3907

View Article Online
ꢀ
ꢀ
ꢀ
ꢀ
(
(
6
C-6glc), 69.1, 67.9, 67.9 (3 CH
C-5man), 67.0, 67.0, 66.9, 66.9, 66.8 (C-4man), 66.2, 65.7, 65.5, 65.3,
2
CH
2
O), 68.9, 68.9, 68.8, 68.8, 68.7
2,3,4,6,2 ,3 ,4 ,6 -Octa-O-[3-(2,3,4,6-tetra- O-benzoyl-a-D-man-
nopyranosyloxy)propyl]-D-trehalose (7-tre). The octaol 6-tre
(50 mg, 0.062 mmol) and the mannosyl donor 3 (5.0 g, 6.7 mmol)
were dissolved in dry acetonitrile (400 mL) under an argon
5.0, 64.9 (5 CH
2
CH
2
OC-1man, CH
2
CH
2
OC-1glc), 62.9, 62.8, 62.8,
CH CH O)
6
2.8, 62.8 (C-6man), 30.7, 30.4, 30.3, 29.8, 29.7 (OCH
2
2
2
◦
ppm. Anal. Calcd. for C191
6
H
172
O56: C, 68.21; H 5.15. Found: C,
atmosphere and the solution was heated to 65 C. Then TMS-OTf
(0.05 ml) was added and it was stirred for 1 h at 65 C. Then, it was
+
◦
8.12; H, 5.25%. MALDI-TOF MS: 3383.46 (M + Na) ion.
cooled to rt, more mannosylimidate 3 (1.2 g, 1.6 mmol) was added
[3-O-(a-D-Mannopyranosyloxy)propyl] 2,3,4,6-tetra-O-[3-(a-D-
and the reaction mixture was stirred at rt overnight. The solution
mannopyranosyloxy)propyl]-a-D-glucopyranoside (8). To a solu-
tion of 7 (329 mg, 0.098 mmol) in dry MeOH (50 mL), NaOMe
was neutralized with solid NaHCO , it was filtered, neutralized
and concentrated. The residue was purified by three subsequent
purification steps. First GPC on Sephadex LH-20 was performed
3
(
10 mg Na in 20 mL MeOH) was added and the solution was
stirred for 2 d at rt. Then it was neutralized with Amberlite IR
20, filtered and purified on Sephadex LH-20 (with MeOH as the
eluent) to yield the unprotected title compound (102 mg, 82%)
(eluent CH
on silica gel (hexane–ethyl acetate, 45 : 55) and finally another
GPC on Sephadex LH-20 (eluent CH Cl –MeOH, 1 : 1) gave the
pure title cluster mannoside (318 mg, 0.059, 95%) as an amorphous
2
2
Cl –MeOH, 1 : 1), followed by flash chromatography
1
2
2
2
0
1
as a colorless amorphous solid. [a]D +13.5 (c 1.85, MeOH). H
NMR (400 MHz, D
-MeOH): d = 4.95 (d, 1H, J1,2glc 3.56 Hz,
H-1glc), 4.82, 4.81 (3×), 4.80 (d, 5H, H-1man), 3.97–3.51 (m, 54H,
2
0
1
4
white solid. [a]D −20.3 (c 0.64, CH
2
Cl
2
). H NMR (500 MHz, D -
6
DMSO, 403 K): d = 8.02–7.92, 7.87–7.82, 7.72–7.65, 7.64–7.56,
7.52–7.20 (each m, 32 H, 16 H, 16 H, 16 H 80 H, aryl-H), 6.03–
5.95 (m, 8H, 8 H-4man), 5.88, 5.85, 5.85, 5.84 (each dd, each 2H,
H-3glc, H-5glc, 2 H-6glc, 5 H-2man, 5 H-3man, 5 H-4man, 5 H-5man
0 H-6man, 20 OCHHCH
H-4glc), 3.30 (dd ≈ t, 1H, J 9.1 Hz, H-2glc), 2.00–1.83 (m, 10H,
OCH CHHCH
O) ppm. H- H COSY: d = 3.56 (H-3glc), 3.65 (H-
glc) ppm. H- C HMQC: d = 3.88 (H-6man), 3.85 (H-2man), 3.77
H-6 man), 3.74 (H-3man), 3.67 (H-4man), 3.56 (H-5man) ppm. C NMR
100.67 MHz, D
-MeOH): d = 102.8 (5 C-1man), 99.2 (C-1glc), 84.2
C-3glc), 83.0 (C-2glc), 80.5 (C-4glc), 75.8, 75.8 (3×), 75.7 (5 C-5man),
,
1
2
CH
2
O), 3.31 (dd ≈ t, 1H, J 9.1 Hz,
J
2,3man 3.0 Hz, J3,4man 9.9 Hz, 8 H-3man), 5.71, 5.68, 5.68, 5.65 (each
1
1
2
2
dd, each 2H, J1,2man 1.7 Hz, 8 H-2man), 5.24 (d, 2H, J1,2glc 3.6 Hz, 2
1
13
5
H-1glc), 5.23, 5.18, 5.16, 5.13 (each d, each 2H, 8 H-1man), 4.67–4.47
ꢀ
13
ꢀ
(
(
(
7
7
6
(m, 24H, 8 H-5man, 8 H-6man, 8 H-6 man), 4.05–3.64 (m, 40H, 2 H-3glc
,
ꢀ
4
2 H-5glc, 2 H-6glc, 2 H-6 glc, 8 glc-OCH
2
, 8 man-OCH ), 3.40 (dd
2
≈ t, 2H, J 9.1 Hz, 2 H-4glc), 3.39 (dd, 2H, J2,3glc 9.1 Hz, 2 H-2glc),
2.11–1.95 (m, 16H, 8 CH CH CH ) ppm. C NMR (125.77 MHz,
CDCl Ph), 134.5–129.5 (160 aryl-C),
1
3
3.9, 73.8 (4×) (5 C-3man), 73.3 (5×) (5 C-2man), 72.9 (C-5glc), 72.7,
2.0, 71.8, 70.4, 70.1 (C-6glc, CH CH O), 69.8, 69.8, 69.8, 69.7,
9.7 (5 C-4man), 67.1 (CH CH OC-1glc), 66.7, 66.6, 66.6, 66.5, 66.4
2
2
2
2
2
3
): d = 167.5–166.5 (32 CO
2
2
2
99.3, 99.0, 89.9, 89.9 (8 C-1man), 95.1 (2 C-1glc), 83.0 (2 C-3glc), 81.7
(2 C-2glc), 79.1 (2 C-4glc), 72.3 (2 C-5glc), 72.0, 71.9, 71.9, 71.8 (8
C-2man), 71.7, 71.6, 71.5, 71.5 (8 C-3man), 70.2, 70.1, 70.1, 70.0 (8 C-
(
(
CH
5 OCH
2
CH
2
OC-1man), 64.1 (5×) (5 C-6man), 32.9, 32.7, 32.5, 32.1, 31.8
1
13
2
CH
2
CH
2
O) ppm. H- C HMBC: d = 72.0 or 71.8 (C-6glc
)
+
ppm. MALDI-TOF MS: 1303.78 (M + Na) ion.
5
(
man), 70.7, 70.5, 69.6, 69.3 (8 (glc)OCH
8 C-4man), 67.5, 67.2, 66.6 (2×) (8 man-OCH
(2×) (8 C-6man), 32.0, 31.8, 31.7, 31.3 (8 OCH
* The signal for one CH moiety is superimposed by the signal
for C-3man. MALDI-TOF MS: m/z = 5454.33 [M + Na] (5431.71
calcd. for C308 91). Anal. Calcd. for C308 91 (5435.54): C,
68.08; H 5.16. Found: C, 67.77; H, 5.06%.
2
, 2 C-6glc*), 68.3 (3×), 68.1
), 64.2, 64.2, 64.1
CH CH O) ppm.
2
ꢀ
ꢀ
ꢀ
ꢀ
2
,3,4,6,2 ,3 ,4 ,6 -Octa-O-(3-hydroxypropyl)-D-trehalose (6-tre).
2
2
2
17
Perallylated trehalose 5-tre (370 mg, 0.56 mmol) was dissolved
in dry THF (20 mL), treated with 9-BBN (20 mL, 10 mmol)
and the reaction mixture was heated under reflux for 1 h. An
2
+
H
278
O
278
H O
◦
excess of hydride was hydrolyzed with water at 0 C and then
aqueous NaOH (3 M, 10 mL) was added, followed by dropwise
◦
ꢀ
ꢀ
ꢀ
ꢀ
addition of aqueous H
2
O
2
(30%, 10 mL) at 0 C. The reaction
2,3,4,6,2 ,3 ,4 ,6 -Octa-O-[3-(a-D-mannopyranosyloxy)propyl]-
D-trehalose (8-tre). To a solution of 7-tre (110 mg, 0.020 mmol)
in dry THF (50 mL), NaOMe (10 mg Na in 20 mL MeOH) was
added and it was stirred for 4 h at rt. Then, it was concentrated,
the residue dissolved in MeOH (50 mL) and NaOMe (10 mg Na
in 20 mL MeOH) was added. This reaction mixture was stirred
overnight at rt, then neutralized with Amberlite IR 120, filtered,
concentrated and purified on Sephadex LH-20 (with MeOH as the
eluent) to yield the title glycocluster (38 mg, 90%) as a colorless
mixture was stirred overnight at rt and saturated with K
phases were separated and the aqueous phase extracted with THF
40 mL) three times. The combined organic phases was concen-
trated, the residue dissolved in a minimum amount of methanol
and passed over a Sephadex LH-20 column (eluent MeOH). Then
2
CO . The
3
(
purification was accomplished on silica gel (CH
2
2
Cl –MeOH, 3 : 1)
to obtain the pure title compound (372 mg, 0.46 mmol, 83%) as a
2
0
1
colorless syrup. [a]D +70.1 (c 2.19, MeOH). H NMR (500 MHz,
2
0
1
D
4
-MeOH): d = 5.20 (d, 2H, J1,2 = 3.1 Hz, 2 H-1), 3.98–3.84, 3.79–
amorphous solid. [a] +72.9 (c 0.17, MeOH). H NMR (400 MHz,
D
3
2
1
.58 (m, 8H und 32H, 8 OCH
H-5, 2 H-6, 2 H-6 , 2 H-3), 3.35–3.28 (m, 4H, 2 H-2, 2 H-4), 1.92–
2
CH
2
CH
2
OH, 8 OCH
2
CH
2
CH
2
OH,
D
4
-MeOH): d = 5.22 (d, 2H, J1,2glc 3.5 Hz, 2 H-1glc), 4.82 (6×), 4.79
ꢀ
(2×) (each d ≈ s, 8H, 8 H-1man), 4.04–3.49 (m, 88H, 32 OCHHCH
2
,
,
1
1
ꢀ
.77 (m, 16H, 8 OCH
2
CH
2
CH
2
OH) ppm. H– H-COSY: d = 3.95
-MeOH): d =
4.9 (2 C-1), 83.8 (2 C-3), 82.7 (2 C-2), 80.5 (2 C-4), 73.3 (2 C-
), 72.6, 72.0, 71.9, 70.7, 70.5 (8 OCH CH CH OH, 2 C-6), 61.5,
1.3, 61.2, 61.2 (8 OCH CH CH OH), 35.7, 35.5, 35.4, 34.8 (8
2 H-3glc, 2 H-5glc, 2 H-6glc, 2-H-6 glc, 8 H-2man, 8 H-3man, 8 H-4man
1
3
ꢀ
(
H-5), 3.64 (H-3) ppm. C NMR (100.67 MHz, D
4
8 H-5man, 8 H-6man, 8 H-6 man), 3.37–3.29 (m, 4H + MeOH, 2 H-
1
3
9
5
6
2glc, 2 H-4glc), 2.00–1.77 (m, 16H, OCH
2
CHHCH
2
O) ppm.
C
2
2
2
NMR (100.67 MHz, D
4
-MeOH): d = 102.8 (2×), 102.7 (6×) (8
2
2
2
C-1man), 95.1 (2×) (2 C-1glc), 84.0 (2×) (2 C-3glc), 82.7 (2×) (2 C-
glc), 80.5 (2×) (2 C-4glc), 75.9 (2×), 75.8 (6×) (8 C-5man), 73.8 (8×)
(8 C-3man), 73.4 (4×), 73.3 (4×) (8 C-2man), 72.7 (2×), 71.8 (2×),
1
13
OCH
2
CH
2
CH
2
OH) ppm. H– C-HMBC: d = 72.0 or 71.9 (C-6)
2
+
ppm. MALDI-TOF MS: m/z = 829.38 [M + Na] (806.45 calcd.
for C36
H
70
O19). Anal. Calcd. for C36
H
70
O19 × 2 H
2
O (842.97): C,
71.7 (2×), 70.6 (2×), 70.5 (2×) (2 C-6glc, (glc)OCH ), 69.8 (6×),
2
5
1.29; H, 8.85. Found: C, 51.69; H, 8.80%.
69.7 (2×) (8 C-4man), 67.0 (2×), 66.7 (2×), 66.5 (2×), 66.4 (2×)
This journal is © The Royal Society of Chemistry 2006
3
908 | Org. Biomol. Chem., 2006, 4, 3901–3912

View Article Online
+
(
3
CH
2.6 (2×), 32.1 (2×) (8 OCH
MS: 2125.89 (M + Na) ion.
2
CH
2
OC-1man), 64.1 (8×) (8 C-6man), 33.0 (2×), 32.7 (2×),
(8 OCH
ion. Anal. Calcd. for C36
52.37; H, 8.66%.
2
CH
2
CH
2
OH) ppm. MALDI-TOF MS: 828.59 (M + Na)
2
CH CH O) ppm. MALDI-TOF
2
2
H
70
O19·H
2
O: C, 52.41; H 8.80. Found: C,
+
Allyl 2,3,4-tri-O-allyl-6-(2,3,4,6-tetra-O-allyl-a-D-galactopy-
ranosyloxy)-b-D-glucopyranoside (5-mel). To a suspension of the
allyl melibioside (4-mel) (1.2 g, 3.1 mmol) in dry DMF (100 mL),
NaH (2.5 g, 65 mmol) was added at rt. 1 h later, allyl bromide
[3-(2,3,4,6-Tetra- O -benzoyl-a-D-mannopyranosyloxy)propyl]
2,3,4-tri-O-[3-(2,3,4,6-tetra-O-benzoyl-a-D-mannopyranosyloxy)-
propyl]-6-{2,3,4,6-tetra-O-[3-(2,3,4,6-tetra-O-benzoyl-a-D-mannopy-
ranosyloxy)propyl]-a-D-galactopyranosyloxy}-b-D-glucopyrano-
side (7-mel). A solution of the octaol 6-mel (0.034 g, 0.042 mmol)
and the mannosyl donor 3 (5.0 g, 6.7 mmol) in dry acetonitrile
(
3.1 mL, 37 mmol) was added and the reaction mixture was stirred
overnight at rt. Then the reaction was quenched with ice water at
◦
◦
0
C and diluted with toluene (100 mL). The organic phase was
(400 mL) was heated to 75 C, TMS-OTf (0.05 mL) was added
separated, consecutively washed with aqueous NaCl solution (2 ×)
and the reaction mixture was stirred at this temperature for
2 h. Then, additional donor 3 (1.2 g, 1.6 mmol) was added and
the mixture was stirred overnight at rt. Then it was neutralized
and water (6 ×), dried over MgSO
4
, filtered, concentrated and
the residue was purified by flash chromatography (toluene–ethyl
acetate, 7 : 1) to yield the title compound (1.4 g, 69%) as a colorless
with NaHCO
purified on Sephadex LH-20 (CH
3
(10 g), filtered, concentrated and the residue was
Cl –MeOH, 1 : 1), followed by
2
0
1
syrup. [a]D +48.0 (c 3.42, CH
d = 6.01–5.84 (dddd ≈ m, 8H, 8 OCH
H, 8 OCH CHCHH), 5.20–5.11 (m, 8H, 8 OCH
2
Cl
2
). H NMR (400 MHz, CDCl
CHCH ), 5.33–5.20 (m,
CHCHH), 5.05
3
):
2
2
2
2
flash chromatography (hexane–ethyl acetate, 9 : 11) to yield the
title cluster mannoside (145 mg, 64%) as a white amorphous solid.
8
2
2
2
D
0
1
(
d, 1H, J1,2gal 3.6 Hz, H-1gal), 4.41–3.94 (m, 18H, H-1glc, H-5gal, 16
[a] −36.0 (c 0.30, CH
2
Cl
2
). H NMR (500 MHz, D -DMSO, 353
6
ꢀ
OCHHCHCH
2
), 3.85–3.74 (m, 4H, H-6glc, H-6 glc, H-2gal, H-4gal),
K): d = 7.98–7.87, 7.84–7.79, 7.66–7.55, 7.48–7.18 (each m, 32 H,
16 H, 32H, 80H, aryl-H), 6.01–5.93 (m, 8 H, 8 H-4man), 5.83–5.75
(m, 8 H, 8 H-3man), 5.66–5.60 (m, 8H, 8 H-2man), 5.18–5.10 (m,
9H, H-1gal, 8 H-1man), 4.64–4.55, 4.53–4.43 (each m, 8H, 16H, 8
H-5man, 8 H-6man, 8 H-6 man), 4.37 (d, 1H, J1,2glc 7.6 Hz, H-1glc),
4.00–3.32 (m, 44H, H-3glc, H-4glc, H-5glc, H-6glc, H-6 glc, H-2gal
H-3gal, H-4gal, H-5gal, H-6gal, H-6 gal, 16 OCHHCH
OCH
ppm. C NMR (125.77 MHz, CDCl
aryl-C), 133.3–128.2 (160 aryl-C), 103.5 (C-1glc), 97.8–97.5 (C-1gal
8 C-1man), 84.9 (C-3glc), 82.6 (C-2glc), 78.5, 78.2, 77.1, 75.9 (C-2gal
3
.71 (dd, 1H, J2,3gal 10.2 Hz, J3,4gal 2.5 Hz, H-3gal), 3.59 (dd, 1H,
J
6,5gal 7.6 Hz, J6,6ꢀgal 9.2 Hz, H-6gal), 3.52 (dd, 1H, J6ꢀ,5gal 6.1 Hz,
ꢀ
H-6 gal), 3.40 (m, 1H, H-5glc), 3.38–3.30 (m, 2H, H-3glc, H-4glc),
1
3
ꢀ
3
(
1
1
.19 (dd ≈ t, 1H, J1,2glc and J2,3glc 8.4 Hz, H-2glc) ppm. C NMR
ꢀ
100.67 MHz, CDCl
35.3, 135.0, 134.6 (8 OCH
17.1, 117.0, 116.8, 116.6 (8 OCH
3
): d = 135.9, 135.7, 135.7, 135.6, 135.5,
,
ꢀ
2
CHCH
2
), 117.4, 117.2, 117.2, 117.1,
), 102.7 (C-1glc), 98.2
2
CH
CHHCH
2
OH, 16
O)
2
CHCH
2
2
CH
2
CHHO, H-2glc), 2.05–1.85 (m, 16H, OCH
2
2
1
3
(
(
C-1gal), 84.6 (C-3glc), 82.1 (C-2glc), 78.2 (C-3gal), 78.1 (C-4glc), 76.7
C-2gal), 75.3 (C-4gal), 75.0 (C-5glc), 74.7, 74.4, 74.1, 74.0, 72.6,
3
): d = 166.0–165.1 (32
,
,
7
(
2.2, 72.1, 70.2 (OCH
C-6glc) ppm. Anal. Calcd. for C36
C, 65.30; H, 8.27%.
2
CHCH
2
), 69.3 (C-5gal), 69.0 (C-6gal), 66.9
H
54
O11: C, 65.24; H 8.21. Found:
C-3gal, C-4gal, C-4glc), 75.0 (C-5glc), 70.6–70.4 (8 C-2man), 70.3–70.1
(8 C-3man), 69.1 (C-5gal,), 68.9–68.6 (8 C-5man), 66.9–66.8 (8 C-4man),
7
0 (2×), 69.6, 69.5, 69.4, 68.2, 67.8 (2×), 66.2–65.7 (C-6glc, C-6gal, 8
3
-Hydroxypropyl 2,3,4-tri-O-(3-hydroxypropyl)-6-[2,3,4,6-tetra-
man-OCH
11 OCH CH
ion.
2 2 2
, 4 glc-OCH
, 4 gal-OCH
), 62.7 (8 C-6man), 30.6–29.7
O-(3-hydroxypropyl)-a-D-galactopyranosyloxy]-b-D-glucopyrano-
side (6-mel). To a solution of 5-mel (240 mg, 0.36 mmol) in dry
THF (20 mL), 9-BBN (0.5 M solution in THF, 12 mL) was added
+
(
2
2
CH O) ppm. MALDI-TOF MS: 5454.1 (M + Na)
2
◦
under an N
2
atmosphere and the solution was stirred at 60 C for
[3-(a-D-Mannopyranosyloxy)propyl] 2,3,4-tri-O-[3-(a-D-manno-
pyranosyloxy)propyl]-6-{2,3,4,6-tetra-O-[3-(a-D-mannopyranosy-
loxy)propyl]-a-D-galactopyranosyloxy}-b-D-glucopyranoside (8-
mel). To a solution of the protected cluster 7-mel (81 mg, 0.015
mmol) in dry THF (50 mL), NaOMe (10 mg Na in 20 mL MeOH)
was added and the solution was stirred for 4 h at rt. Then it
was concentrated and the residue dissolved in MeOH (50 mL).
Then again NaOMe (10 mg Na in 20 mL MeOH) was added
and the reaction mixture was stirred overnight at rt, neutralized
with Amberlite IR 120, filtered and the residue was purified on
Sephadex LH-20 (eluent MeOH) to yield the unprotected title
1
h. Then the excess of 9-BBN was destroyed by dropwise addition
of water at 0 C. The hydroboration mixture was oxidized by the
addition of aqueous NaOH (3 M, 12 mL) and aqueous H
30%, 12 mL) at 0 C, followed by stirring at rt overnight. The
◦
2
O
2
◦
(
aqueous phase was saturated with K
2
CO and the THF phase was
3
separated. The aqueous phase was extracted twice with THF (40
mL). The combined organic phases were concentrated and purified
on Sephadex LH-20 (eluent MeOH) and by flash chromatography
CH
6%) as a colorless oil. [a]D +53.1 (c 0.52, MeOH). H NMR
500 MHz, D
(
5
(
1
2
Cl
2
–MeOH, 3 : 1) to afford the title compound (162 mg,
2
0
1
2
0
4
-MeOH): d = 5.14 (d ≈ s, 1H, H-1gal), 4.31 (d,
compound (32 mg, quant.) as a colorless amorphous solid. [a]
+66.0 (c 0.59, MeOH). H NMR (300 MHz, D
D
1
H, J1,2glc 7.6 Hz, H-1glc), 4.01 (ddd ≈ t, 1H, J5,6gal and J5,6ꢀgal
-MeOH): d = 5.14
4
6
.6 Hz, H-5gal), 3.99–3.60 (m, 38H, 16 OCHHCH
2
CH
2
OH, 16
(d ≈ s, 1H, H-1gal), 4.85–4.78 (m, 8H, 8 H-1man), 4.31 (d, 1H, J1,2glc
OCH
2
CH
2
CHHOH, H-2gal, H-3gal, H-4gal, H-6gal), 3.58 (dd, 1H,
7.3 Hz, H-1glc), 4.06–3.51 (m, 86H, H-5gal, 16 OCHHCH
16 OCH CH
2 2
ꢀ
CH
O,
ꢀ
J
6,6ꢀgal 9.2 Hz, H-6 gal), 3.42–3.26 (m, 3H + MeOH, H-3glc, H-4glc
H-5glc), 3.05 (dd ≈ t, 1H, J2,3glc 8.1 Hz, H-2glc), 1.92–1.77 (m, 16H,
OCH CHHCH OH) ppm. C NMR (100.67 MHz, D -MeOH): d
105.7 (C-1glc), 99.7 (C-1gal), 87.2 (C-3glc), 84.9 (C-2glc), 80.9, 80.5,
9.3, 78.1 (C-2gal, C-3gal, C-4gal, C-4glc), 77.0 (C-5glc), 72.7, 72.7, 72.1,
2.0, 72.0 (2×), 70.5, 70.1, 70.0, 68.8, 68.3 (8 OCH CH CH OH,
,
2
2
CHHO, H-2gal, H-3gal, H-4gal, H-6gal, H-6 gal, 8
ꢀ
H-4man, 8 H-3man, 8 H-2man, 8 H-5man, 8 H-6man, 8 H-6 man), 3.46–3.25
1
3
2
2
4
(m, 3H + MeOH, H-3glc, H-4glc, H-5glc), 3.03 (dd ≈ t, 1H, J2,3glc
1
3
=
8.9 Hz, H-2glc), 2.03–1.80 (m, 16H, OCH
2
CHHCH O) ppm. C
2
7
7
NMR (125.77 MHz, D
4
-MeOH): d = 104.9 (C-1glc), 102.1–101.9
2
2
2
(8×) (8 C-1man), 98.9 (C-1gal),. 86.3 (C-3glc), 84.2 (C-2glc), 80.0,
79.3, 78.8, 77.5 (C-2gal, C-3gal, C-4gal, C-4glc), 76.4 (C-5glc), 74.9
(8×) (8 C-5man), 72.9 (8×) (8 C-3man), 72.5 (8×) (8 C-2man), 71.0
C-6glc, C-6gal), 71.6 (C-5gal), 61.6, 61.6, 61.5, 61.5, 61.4, 61.3, 61.2,
1.1 (8 OCH CH CH OH), 35.7, 35.5, 35.3, 35.2, 35.2, 35.1, 35.0
6
2
2
2
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 3901–3912 | 3909

View Article Online
(
(
(
(
C-5gal), 71.9, 71.7, 71.3, 71.2, 70.9, 69.6, 69.1, 67.8, 67.0, 64.6
C-6glc, C-6gal, 8 glc-OCH
OCH
2
CH
2
CHHOH), 3.26 (dd, 1H, J2,3glc 9.1 Hz, H-2glc), 1.92–
1
1
2
),68.9 (8×) (8 C-4man), 66.0–65.7 (8×)
1.77 (m, 22H, 22 OCH
2
CHHCH OH) ppm. H- H COSY : d
2
13
1
8 man-OCH
8 OCH CH
2
), 63.2 (8×) (8 C-6man), 32.0, 31.8 (5×), 31.4, 31.2
= 3.57 (H-3glc), 3.68 (H-2gal). H- H HSQC : d = 3.51 (H-4glc),
+
ꢀ
2
2
CH
2
O) ppm. MALDI-TOF MS: 2126.1 (M + Na)
3.99 (H-5glc), 3.90 (H-6glc), 3.77 (H-6 glc), 3.69 (H-1frc), 3.49 (H-
ꢀ
ꢀ
ion.
1
frc), 4.02 (H-4frc), 3.89 (H-5frc), 3.71 (H-6frc), 3.71 (H-6 frc), 3.70
1
3
(
H-3gal), 3.86 (H-4gal), 3.97 (H-5gal) ppm. C NMR (100.67 MHz,
Undeca-O-allyl-D-raffinose (5-raf). To a suspension of raffi-
nose (4-raf) (3.0 g, 5.0 mmol) in dry DMF (100 mL), NaH (3.5 g,
8
were added and the reaction mixture was stirred overnight at
rt. Then the reaction was quenched with ice water at 0 C and
D
4
-MeOH): d = 106.7 (C-2frc), 99.9 (C-1gal), 92.0 (C-1glc), 86.5
(
(
(
7
6
C-3frc), 84.4 (C-4frc), 84.1 (C-3glc), 82.9 (C-2glc), 81.9 (C-5frc), 80.8
C-3gal), 79.9 (C-4glc), 79.3 (C-2gal), 78.0 (C-4gal), 74.4 (C-1frc), 74.1
C-6frc), 73.8 (C-5glc), 71.6 (C-5gal), 72.7, 72.5, 71.8 (2×), 70.8, 70.6,
0 mmol) and 1 h later, allyl bromide (6.1 mL, 72.1 mmol)
◦
0.5 (2×), 70.1, 70.0, and 69.9 (2× (C-6gal, 11 OCH
2
CH
2
CH OH),
2
toluene (100 mL) was added. The organic phase was separated,
7.8 (C-6glc), 61.6, 61.6, 61.6, 61.5, 61.4, 61.3 61.2 (3× 61.2, 61.0
consecutively washed with aqueous NaCl (2 ×) and water (6 ×),
(11 OCH
2
CH CH
2
2
OH), 35.7, 35.5, 35.3 (3×), 35.2, 35.2 (2×), and
dried over MgSO , filtered, concentrated and the residue was
4
3
4.9(3× (11 OCH
2
CH CH OH) ppm. MALDI-TOF MS: 1165.87
2
2
purified by flash chromatography (toluene–ethyl acetate, 7 : 1)
to yield the title compound (1.4 g, 69%) as a colorless syrup. [a]
+
(
M + Na) ion. Anal. Calcd. for C51
H
98
O27·2H
2
O: C, 51.94; H 8.72.
2
0
D
Found: C, 51.99; H, 8.64%.
1
+
81.4 (c 0.91, CH
.82 (dddd ≈ m, 11H, 11 OCH
H-1glc), 5.34–5.20 (ddd, 11H, 11 OCH
2
Cl
2
). H NMR (400 MHz, CDCl
CHCH ), 5.51 (d, 1H, J1,2glc 3.6 Hz,
CHCHH), 5.20–5.07 (ddd,
CHCHH), 5.06 (d, 1H, J1,2gal 3.6 Hz, H-1gal), 4.40–
.29 (m, 3H, OCHHCHCH ), 4.26–3.89 (m, 24H, H-3frc, H-4frc
), 3.84 (dd, 1H, H-6glc),
3
): d = 6.02–
5
2
2
Undeca-O-[3-(2,3,4,6-tetra-O-benzoyl-a-D-mannopyranosyloxy)-
propyl]-D-raffinose (7-raf). To a solution of 6-raf (39 mg, 0.034
mmol) and the mannosyl donor 3 (5.0 g, 6.7 mmol) in dry
2
1
4
1H, 11 OCH
2
2
,
acetonitrile (400 mL), TMS-OTf (0.05 mL) was added under N
2
◦
◦
H-5frc, H-5gal, H-5glc, 19 OCHHCHCH
3
1
2
at 60 C and the solution was stirred at 60 C for 10 min. Then
the heating was removed and stirring was continued for 2 h at rt.
Additional donor 3 (2.0 g, 2.7 mmol) was added and the reaction
.83 (dd, 1H, H-4gal), 3.80 (dd, 1H, J2,3gal 10.2 Hz, H-2gal), 3.70 (dd,
ꢀ
H, J3,4gal 3.1 Hz, H-3gal), 3.68–3.57 (m, 6H, H-1frc, H-6frc, H-6 frc
,
H-3glc, H-6glc, H-6gal), 3.54 (dd ≈ t, 1H, J3,4glc and J3,5glc 9.2 Hz,
mixture was stirred overnight at rt. Then NaHCO
added, it was filtered and purified by subsequent procedures, first
on Sephadex LH-20 (CH Cl –MeOH, 1 : 1) and then by flash
3
(10 g) was
ꢀ
H-4glc), 3.49 (dd, 1H, J6ꢀ,5gal 5.6 Hz, J6,6ꢀgal 9.2 Hz, H-6 gal), 3.43
ꢀ
(
d, 1H, J1,1ꢀfrc 11.2 Hz, H-1 frc), 3.26 (dd, 1H, J2,3glc 9.2 Hz, H-2glc
)
)
2
2
1
13
ppm. H- C HMQC: d = 4.24 (H-3frc), 4.01 (H-4frc), 3.97 (H-5frc
chromatography (hexane–ethyl acetate, 4 : 6) to yield the title
1
13
1
1
20
ppm. H- C HMBC: d = 3.92 (H-5gal) ppm. H- H COSY: d =
copound (231 mg, 91%) as a white amorphous solid. [a]D −30.0
1
3
1
3
1
.94 (H-5glc) ppm. C NMR (100.67 MHz, CDCl
35.9, 135.8, 135.8, 135.7, 135.5, 135.2, 135.2, 135.1, 135.0, 135.0
CHCH
), 117.4, 117.3, 117.3 (2×), 117.3, 117.2, 116.8,
16.7, 116.6 (2×), 116.4 (11 OCH CHCH ), 104.8 (C-1frc), 98.6
3
): d = 136.1,
(c 0.17, CH
2
Cl
2
). H NMR (500 MHz, CDCl
3
): d = 8.12–7.70 (m,
88H, aryl-H), 7.55–7.00 (m, 132H, aryl-H), 6.25–6.05 (m, 11 H,
11 H-4man), 5.98–5.86 (m, 11 H, 11 H-3man), 5.78–5.60 (m, 12 H, 11
H-2man, H-1glc), 5.23–4.92 (m, 12H, H-1gal, 11 H-1man), 4.80–4.56,
4.53–4.30 (each m, 11H, 22H, 11 H-5man, 11 H-6man, 11 H-6 man),
4.16–3.26 (m, 64H, H-3frc, H-3glc, H-4glc, H-5glc, H-6glc, H-6 glc, H-
(
11 OCH
2
2
1
2
2
ꢀ
(
C-1gal), 90.3 (C-1glc), 83.8 (C-3frc), 82.4 (C-4frc), 81.7 (C-3glc), 79.9,
9.8 (C-2glc, C-5frc), 78.2 (C-3gal), 77.6 (C-4glc), 76.8 (C-2gal), 75.2
ꢀ
7
ꢀ
ꢀ
(
7
(
C-4gal), 74.4 (2×), 74.1, 72.9, 72.7, 72.6, 72.1 (2×), 71.9, 71.9,
1
frc, H-1 frc, H-4frc, H-5frc, H-6frc, H-6 frc, H-2gal, H-3gal, H-4gal, H-5gal
,
ꢀ
1.8, 71.6 (C-6frc, OCH
2
CHCH
2
), 71.5 (C-5glc), 69.5 (C-5gal), 69.1
H-6gal, H-6 gal, 22 OCHHCH
2
CH
2
OH, 22 OCH
CHHCH O) ppm. C NMR
): d = 165.9–165.1 (44 aryl-C), 133.2–128.2
(220 aryl-C), 104.5 (C-2frc), 97.9–97.5 (C-1gal, 11 C-1man), 90.3
2
CH
2
CHHOH,
+
13
C-6gal), 66.4 (C-6glc) ppm. MALDI-TOF MS: 967.94 (M + Na)
H-2glc), 2.24–1.90 (m, 22H, OCH
(125.77 MHz, CDCl
2
2
ion. Anal. Calcd. for C51
H, 8.11%.
H
76
O16: C, 64.81; H, 8.10. Found: C, 64.65;
3
(
(
C-1glc), 83.8, 83.3, 82.2, 80.8, 79.6, 78.5, 77.2, 75.7, and 75.5
C-3frc, C-4frc, C-3glc, C-2glc, C-5frc, C-3gal, C-4glc, C-2gal, C-4gal),
Undeca-O-(3-hydroxypropyl)-D-raffinose (6-raf). To a solution
of 5-raf (535 mg, 0.57 mmol) in dry THF (20 mL), 9-BBN (0.5 M
7
3.4–71.5 (C-1frc, C-6frc, C-5glc), 70.5–70.1 (11 C-2man, 11 C-3man
C-5gal), 68.7 (11 C-5man), 70.0, 69.2, 69.1, 68.4–67.3 (11 (gal-,
glc-, frc-OCH , C-6gal), 66.7 (11 C-4man), 66.0–65.4 (C-6glc, 11
,
solution in THF, 26 mL) was added under an N atmosphere and
2
the solution was stirred at reflux temperature for 1 h. Then the
2
excess of 9-BBN was destroyed by dropwise addition of water at
man-OCH
2
), 62.6 (11 C-6man), 30.6–29.7 (11 OCH
2
CH
2
CH O)
2
◦
0
C. The hydroboration mixture was oxidized by the addition of
+
ppm. MALDI-TOF MS: 7525.7 (M + Na) ion.
aqueous NaOH (3 M, 13 mL) and aqueous H
2
O (30%, 13 mL) at
2
◦
0
C. Then it was stirred at rt overnight and the aqueous phase was
CO . The phases were separated and the aqueous
Undeca-O-[3-(a-D-mannopyranosyloxy)propyl]-D-raffinose (8-raf).
The protected cluster mannoside 7-raf (231 mg, 0.031 mmol) was
suspended in dry THF (50 mL), NaOMe (10 mg Na in 20 mL
MeOH) was added and the reaction mixture was stirred at rt for
4 h. Then, it was concentrated, the residue was dissolved in MeOH
(50 mL) and NaOMe (10 mg Na in 20 mL MeOH) was added. This
reaction mixture was stirred overnight at rt, then it was neutralized
with Amberlite IR 120, filtered and the residue was purified on
Sephadex LH-20 (eluent MeOH) to yield the unprotected title
saturated with K
2
3
phase was extracted twice with THF (40 mL). The combined
organic phases was concentrated and the residue was purified
by two subsequent procedures, first by GPC on Sephadex LH-
2
0 (eluent MeOH), followed by flash chromatography (CH
2
2
Cl –
MeOH, 3 : 1) to afford the pure title compound (538 mg, 85%) as
2
0
1
a colorless oil. [a]D +66.1 (c 0.36, MeOH). H NMR (500 MHz,
D
4
-MeOH): d = 5.63 (d, 1H, J1,2glc 3.6 Hz, H-1glc), 5.18 (d, 1H, H-
2
0
1gal), 4.15 (d, 1H, H-3frc), 4.05–3.47 (m, 61H, H-3glc, H-4glc, H-5glc
,
,
compound (64 mg, 71%) as a colorless amorphous solid. [a]
+71.5 (c 0.13, MeOH). H NMR (500 MHz, CDCl
1H, J1.2glc 3.5 Hz, H-1glc), 5.17 (s, 1H, H-1gal), 4.86–4.79 (m, 11H, 11
D
ꢀ
ꢀ
ꢀ
1
H-6glc, H-6 glc, H-1frc, H-1 frc, H-4frc, H-5frc, H-6frc, H-6 frc, H-2gal
H-3gal, H-4gal, H-5gal, H-6gal, H-6 gal, 22 OCHHCH
3
): d = 5.62 (d,
ꢀ
2
CH
2
OH, 22
3
910 | Org. Biomol. Chem., 2006, 4, 3901–3912
This journal is © The Royal Society of Chemistry 2006

View Article Online
H-1man), 4.15 (d, 1H, H-3frc), 4.07–3.47, 3.40–3.22 (m, 125H and
72.9, 72.8, 72.7, 72.4, 72.2, 72.0, 69.3 (C-6, 5 OCH
70.3 (5 OCH CH CH OH), 61.3 (5 OCH CH CH OH), 34.9 (5
OCH CH CH
OH) ppm. MALDI-TOF MS: m/z = 713.6 [M +
Na] (690.4 calcd. for C31
2
CH
2
O),
ꢀ
ꢀ
3
H + MeOH, H-3glc, H-5glc, H-6glc, H-6 glc, H-1frc, H-1 frc, H-4frc
,
2
2
2
2
2
2
ꢀ
ꢀ
H-5frc, H-6frc, H-6 frc, H-2gal, H-3gal, H-4gal, H-5gal, H-6gal, H-6 gal, 22
2
2
2
+
OCHHCH
2
CH
2
O, 22 OCH
2
CH
2
CHHO, 11 H-4man, 11 H-3man, 11
62
H O16).
ꢀ
H-2man, 11 H-5man, 11 H-6man, 11 H-6 man, H-2glc, H-4glc), 2.02–1.83
[6-O-(2,3,4,6-Tetra-O-benzoyl-a-D-mannopyranosyloxy)-3-oxa-
1
3
(
m, 22H, 22 OCH
2
CHHCH OH) ppm. C NMR (100.67 MHz,
2
hexyl] 2,3,4,6-tetra-O-[6-(2,3,4,6-tetra-O-benzoyl-a-D-mannopy-
ranosyloxy)-3-oxa-hexyl]-a-D-glucopyranoside (12). A mixture
of the pentaol 11 (45 mg, 0.065 mmol) and the mannosyl donor
D
4
-MeOH): d = 101.9 (11×) (11 C-1man), 99.3 (C-1gal), 91.3 (C-1glc),
8
8
5.6 (C-3frc), 83.5 (2×) (C-4frc, C-3glc), 82.1 (C-2glc), 81.0 (C-5frc),
0.1 (C-3gal), 79.1 (C-4glc), 78.8 (C-2gal), 77.4 (C-4gal), 74.9 (11×) (11
3
(2.30 g, 3.1 mmol) was dried under high vacuum and then
dissolved in dry CH Cl (4 mL) under an argon atmosphere. A
solution of TMS-OTf (5% in dry CH Cl , 0.2 mL) was added
and the reaction mixture was stirred overnight at rt. Then it
was neutralized with NaHCO (1 g), filtered, concentrated and
the residue was purified by flash chromatography on silica gel
C-5man), 73.3 (2×) (C-1frc, C-6frc), 73.0 (12×) (C-5glc, 11 C-3man), 77.5
2
2
(
11×) (11 C-2man), 71.7 (2×), 71.1 (2×), 69.7 (4×), and 69.1 (2×)
CH , C-6gal, 11 gal-, glc-, frc-OCH
), 70.9 (C-5gal), 68.9 (11×) (11
CH O(C-1man)), 63.2
11×) (11 C-6man), 32.0–31.2 (11×) (11 OCH CH CH O) ppm.
MALDI-TOF MS: 2948.6 (M + Na) ion.
2
2
(
2
2
C-4man), 66.0 (6×), 65.7 (6×) (C-6glc, 11 CH
2
2
3
(
2
2
2
+
(
cyclohexane–ethyl acetate, 3 : 2 → 1 : 1) to yield the title cluster
20
D
[
2-(Allyloxy)ethyl] 2,3,4,6-tetra-O-[2-(allyloxy)ethyl]-a-D-glu-
mannoside as an amorphous colorless solid (164 mg, 71%). [a]
5
1
copyranoside (10). The pentaol 9 (165 mg, 0.41 mmol) was
suspended in dry DMF (10 mL) and NaH (60% suspension in
paraffin oil, 200 mg, 5.0 mmol) and after 0.5 h, allyl bromide
−35.4 (c = 0.11, CH
2
Cl
2
). H NMR (400 MHz, CDCl
3
): d =
8.12–8.08, 8.06–8.02, 7.97–7.93, 7.85–7.81, 7.59–7.53, 7.44–7.48,
7.43–7.30, 7.26–7.20 (each m, 10H, 10H, 10H, 10H, 10H, 5H,
35H, 10H, aryl-H), 6.17–6.10 (m, 5H, 5 H-4man), 5.91 (dd, 5H,
(
0.35 mL, 4.1 mmol), were added. The reaction mixture was
◦
stirred overnight at rt, was then cooled to 0 C and water (100 mL)
and toluene (100 mL) were added. The phases were separated and
the organic phase was washed twice with satd. aqueous sodium
J2,3man 3.3 Hz, J3,4man 10.3 Hz, 5 H-3man), 5.70 (dd, 5H, J1,2man
1.6 Hz, 5 H-2man), 5.10–5.08 (m, 5H, 5 H-1man), 4.99 (d, 1H, J1,2glc
3.5 Hz, H-1glc), 4.72–4.68 (m, 5H, 5 H-6man), 4.51–4.46 (m, 5H,
ꢀ
chloride solution and twice with water. It was dried over MgSO
4
,
5 H-6 man), 4.45–4.40 (m, 5H, 5 H-5man), 4.08–4.00, 3.96–3.89,
filtered and concentrated. Purification on silica gel (toluene–ethyl
acetate, 1 : 1) delivered the title compound as a colorless syrup
3.86–3.75, 3.74–3.58 (each m, 2H, 6H, 6H, 30H, 44H, H-3glc
,
ꢀ
H-5glc, H-6glc, H-6 glc, 5 OCH
2
CH
2
OCH
2
CH
2
CH
2
O), 3.39–3.43
CH CH O)
): d = 166.1–165.4 (20
Ph), 133.4–128.3 (100 aryl-C), 97.7 (5 C-1man), 97.2 (C-1glc),
82.1 (C-3glc), 80.9 (C-2glc), 78.1 (C-4glc), 72.4, 72.2, 70.8–70.6, 70.1,
69.9, 69.7, 67.7, 66.8 (5 OCH CH O, C-6glc) 70.6 (5 C-2man), 70.3
(C-5glc), 70.2 (5 C-3man), 68.8 (5 C-5man), 67.0 (5 C-4man), 65.7–65.6
2
0
1
(
134 mg, 54%). [a]D +71.3 (c 0.38, CH
2
Cl
): d = 5.97–5.81 (m, 5H, 5 OCH CHCH
H, 5 OCH CHCHH), 5.19–5.12 (m, 5H, 5 OCH
2
). H NMR (300 MHz,
), 5.31–5.21 (m,
CHCHH),
(m, 2H, H-4glc, H-2glc), 2.03–1.94 (m, 10H, 5 OCH
2
2
2
1
3
CDCl
5
4
3
2
2
ppm. C NMR (125.76 MHz, CDCl
CO
3
2
2
2
.99 (d, 1H, J1,2 3.6 Hz, H-1), 4.10–3.86, 3.83–3.49 (each m, 13H
ꢀ
and 21H, H-3, H-5, H-6, H-6 , 5 OCH
2
CH
2
O, 5 OCH
2
CH
2
O, 5
2
2
OCH
2
CHCH
2,3 9.6 Hz, H-2) ppm. C NMR (75.47 MHz, CDCl
34.9, 134.8 (2×), 134.7 (5 OCH CHCH ), 116.9, 116.8 (2×),
16.7, 116.6 (5 OCH CHCH ), 97.1 (C-1), 82.0 (C-3), 80.9 (C-2),
7.9 (C-4), 70.1 (C-5), 72.3, 72.2, 72.1, 72.1 (2×), 72.0, 72.0, 70.8,
0.6, 69.8, 69.7, 69.6 (2×), 69.4, 69.1, 66.8 (C-6, 5 OCH CH O, 5
CH O, 5 OCH CHCH ) ppm.
2
), 3.40 (dd ≈ t, 1H, J 9.5 Hz, H-4), 3.38 (dd, 1H,
1
3
J
1
1
7
7
3
): d = 134.9,
(5 (C-1man)OCH
2
), 62.9 (5 C-6man), 29.8–29.7 (5 OCH
2
CH
2
CH O)
2
+
2
2
ppm. MALDI-TOF MS: m/z = 3604.6 [M + Na] (3581.2 calcd.
2
2
for C201
192
H O61).
[
6-O-(a-D-Mannopyranosyloxy)-3-oxa-hexyl] 2,3,4,6-tetra-O-
6-O-(a-D-mannopyranosyloxy)-3-oxa-hexyl]-a-D-glucopyranoside
13). The protected cluster mannoside 12 (137 mg, 0.038 mmol)
2
2
[
(
OCH
2
2
2
2
(
6-Hydroxy-3-oxa-hexyl) 2,3,4,6-tetra-O-(6-hydroxy-3-oxa-
was dissolved in THF (20 mL) and treated with NaOMe (0.02 M
in MeOH, 10 mL). The reaction mixture was stirred at rt for
0.5 h, then it was neutralized by the addition of ion exchange resin
Amberlite IR 120, it was filtered, concentrated and then purified
on Sephadex LH-20 with methanol as the eluent. The unprotected
title compound was obtained as a colorless amorphous solid
hexyl)-a-D-glucopyranoside (11). The perallylated octopus
glucoside 10 (136 mg, 0.23 mmol) was dissolved in dry THF
15 mL), treated with 9-BBN (5 mL, 2.5 mmol) and the reaction
mixture was heated under reflux for 1 h. The excess hydride was
destroyed by the addition of ice water and then aqueous NaOH
3 M, 2.5 mL) and aqueous H
(
2
0
1
(
2
O
2
(30%, 2.5 mL) were added
(57 mg, quant.). [a]D +72.8 (c 0.53, MeOH). H NMR (500 MHz,
◦
dropwise at 0 C. The reaction mixture was stirred overnight at
D
4
-MeOH): d = 5.02 (d, 1H, J1,2glc 3.7 Hz, H-1glc), 4.81–4.79 (m,
rt, then the solution was saturated with solid K
were separated and the aqueous phase was washed with THF (2 ×
0 mL). The combined organic phases was dried over MgSO
and the solvent was removed after filtration. Purification on silica
gel (CH Cl
as a colorless syrup (72 mg, 45%). [a] +53.7 (c 0.66, MeOH).
H NMR (300 MHz, D
H-1), 4.09–3.58 (m, 44H, H-3, H-5, H-6, H-6 , 5 OCH
2
CO
3
, the phases
5H, 5 H-1man), 4.07–3.98, 3.94–3.53 (each m, 2H, 72H, H-3glc
H-5glc, 2 H-6glc, 5 H-2man, 5 H-3man, 5 H-4man, 5 H-5man, 10 H-6man
,
,
4
4
5 OCH
3.35 (dd ≈ t, 1H + MeOH, J 9.4 Hz, H-4glc), 1.95–1.87 (m, 10H,
5 OCH CH CH O) ppm. C NMR (125.76 MHz, D -MeOH): d
= 102.7 (5 C-1man), 99.4 (C-1glc), 84.5 (C-3glc), 83.2 (C-2glc), 80.4
(C-4glc), 75.7 (5 C-5man), 74.6, 74.4 ((C-3glc)OCH , (C-4glc)OCH ),
73.8 (5 C-3man), 73.4 (5 C-2man), 73.1, 72.9 (2x), 72.8, 72.7, 72.4,
72.2, 72.1 (C-6glc, 2 (glc)OCH , 5 (glc)OCH CH ), 72.8 (C-5glc),
70.2 (man-OCH CH CH O), 69.8 (5 C-4man), 69.4 (C-1glc-OCH ),
66.6 (5 man-OCH O)
2
CH
2
OCH
2
CH
2
CH O), 3.42 (dd, 1H, J2,3 9.7 Hz, H-2glc),
2
1
3
2
2
–MeOH, 6 : 1 → 5 : 1) yielded the title compound
2
2
2
4
2
D
0
1
4
-MeOH): d = 5.01 (d, 1H, J1,2 3.6 Hz,
2
2
ꢀ
2
CH O,
2
5
1
OCH
2
CH
2
CH
2
OH), 3.40 (d, 1H, J2,3 9.6 Hz, H-2), 3.35 (m,
CH CH OH)
-MeOH): d = 99.4 (C-1), 84.5
C-3), 83.2 (C-2), 80.4 (C-4), 72.7 (C-5), 74.6, 74.4, 73.1, 73.0,
2
2
2
H + MeOH, H-4), 1.88–1.76 (m, 10H, 5 OCH
2
2
2
2
2
2
2
1
3
ppm. C NMR (75.47 MHz, D
4
2
), 64.1 (5 C-6man), 32.1 (5 OCH
2
CH
2
CH
2
+
(
ppm. MALDI-TOF MS: m/z = 1523.8 [M + Na] (1500.7 calcd.
Org. Biomol. Chem., 2006, 4, 3901–3912 | 3911
This journal is © The Royal Society of Chemistry 2006

View Article Online
for C61
Found: C, 49.09; H, 7.69%.
H
112
O
41). Anal. Calcd. for C61
H
112
O
41: C, 48.79; H 7.52.
14 F. Bien and T. Ziegler, Tetrahedron: Asymmetry, 1998, 9, 781–790.
1
5 T. K. Lindhorst, M. Dubber, U. Krallmann-Wenzel and S. Ehlers,
Eur. J. Org. Chem., 2000, 2027–2034.
1
6 G. Zempl e´ n and E. Pascu, Ber. Dtsch. Chem. Ges., 1929, 62, 1613–1614.
Acknowledgements
17 M. Dubber and T. K. Lindhorst, Org. Lett., 2001, 3, 4019–4022.
1
8 R. R. Schmidt, M. Behrendt and A. Toepfer, Synlett, 1990, 694–696.
19 I. Tvaroska and F. R. Taravel, Adv. Carbohydr. Chem. Biochem., 1995,
1, 15–61.
This work was supported by the DFG (Deutsche Forschungsge-
meinschaft) in the frame of SFB 470 and by the Fonds of the
German Chemical Industry (FCI).
5
2
2
0 (a) G. Kretzschmar, U. Sprengard, H. Kunz, E. Bartnik, W. Schmidt,
A. Toepfer, B. H o¨ rsch, M. Krausse and D. Seiffge, Tetrahedron, 1995,
5
1, 13015–13030; (b) A. Kichler and F. Schubert, Glycoconjugate J.,
References
1995, 12, 275–281.
1 (a) N. Sharon and H. Lis, Glycobiology, 2004, 14, 53R–62R; (b) K. J.
Doores, D. P. Gamblin and B. G. Davis, Chem.–Eur. J., 2006, 12, 656–
665; (c) V. Bogoevska, A. Horst, B. Klampe, L. Lucka, C. Wagener and
P. Nollau, Glycobiology, 2006, 16, 197–209.
1
2
3
R. A. Dwek, Biochem. Soc. Trans., 1995, 23, 1–25.
A. Varki, Glycobiology, 1993, 3, 97–130.
(a) R. Roy, Top. Curr. Chem., 1997, 187, 241–274; (b) P. Sears and C.-H.
Wong, Angew. Chem., Int. Ed., 1999, 38, 1875–1917; (c) L. L. Kiessling,
J. E. Gestwicki and L. E. Strong, Curr. Opin. Chem. Biol., 2000, 4,
22 (a) P. Klemm and K. A. Krogfelt, in Fimbriae: Adhesion, Genetics,
Biogenesis and Vaccines, ed. P. Klemm, CRC Press, Boca Raton, 1994,
pp. 9–26; (b) J. Berglund and S. D. Knight, Adv. Exp. Med. Biol., 2003,
535, 33–52; (c) N. Sharon and H. Lis, Glycobiology, 2004, 14, 53–62.
23 M. Vetsch, C. Puorger, T. Spirig, U. Grauschopf, E.-U. Weber-Ban and
R. Glockshuber, Nature, 2004, 431, 330–332.
6
2
2
2
96–703; (d) C. R. Bertozzi and L. L. Kiessling, Science, 2001, 291,
357–2364; (e) W. B. Turnbull and J. F. Stoddart, Rev. Mol. Biotechnol.,
002, 90, 231–255; (f) T. K. Lindhorst, Top. Curr. Chem., 2002, 218,
01–235; (g) B. Davis, Chem. Rev., 2002, 102, 579–601.
4
(a) C. Kieburg, M. Dubber and T. K. Lindhorst, Synlett, 1997, 1447–
24 N. Sharon, FEBS Lett., 1987, 217, 145–157.
1
2
449; (b) M. Dubber and T. K. Lindhorst, J. Carbohydr. Chem., 2001,
0, 755–760.
25 G. H. Jones and C. E. Ballou, J. Biol. Chem., 1969, 244, 1043–1051.
26 P. Klemm, B. J. Jørgensen, I. van Die, H. de Ree and H. Bergmans,
Mol. Gen. Genet., 1985, 199, 410–414.
5
6
7
M. Dubber and T. K. Lindhorst, Carbohydr. Res., 1998, 310, 35–41.
M. Dubber and T. K. Lindhorst, J. Org. Chem., 2000, 65, 5275–5281.
(a) R. Nuck, C. Paul, B. Wieland, C. Heidrich, C. C. Geilen and W.
Reutter, Eur. J. Biochem., 1993, 216, 215–221; (b) A. P. Sherblom and
R. M. Smagula, Methods Mol. Biol., 1993, 14, 143–149; (c) C. Kuo, N.
Takahashi, A. F. Swanson, Y. Ozeki and S. Hakomori, J. Clin. Invest.,
27 O. Sperling, A. Fuchs and T. K. Lindhorst, Org. Biomol. Chem., 2006,
DOI: 10.1039/b610745a.
28 O. Sperling, Diploma Thesis, Hamburg and Kiel, 2001.
29 N. Firon, I. Ofek and N. Sharon, Carbohydr. Res., 1983, 120, 235–249.
30 C.-W. von der Lieth, M. Frank and T. K. Lindhorst, Rev. Mol.
Biotechnol., 2002, 90, 311–337.
1
996, 98, 2813–2818; (d) I. Botos and A. Wlodawer, Prog. Biophys.
Mol. Biol., 2005, 88, 233–282.
31 H.-J. Gabius, H.-C. Siebert, S. Andr e´ , J. Jim e´ nez-Barbero and H.
R u¨ diger, ChemBioChem, 2004, 5, 740–764.
8
9
G. Wulff and G. R o¨ hle, Angew. Chem., Int. Ed. Engl., 1974, 13, 157–171.
R. R. Schmidt and W. Kinzy, Adv. Carbohydr. Chem. Biochem., 1994,
32 (a) M. Mammen, S.-K. Choi and G. M. Whitesides, Angew. Chem., Int.
Ed., 1998, 37, 2755–2794; (b) R. T. Lee and Y. C. Lee, Glycoconjugate J.,
2000, 17, 543–551; (c) T. K. Lindhorst, Top. Curr. Chem., 2002, 218,
201–235; (d) T. Christensen, D. M. Gooden, J. E. Kung, J. I. Nasas and
E. J. Toone, J. Am. Chem. Soc., 2003, 125, 7357–7366; (e) P. I. Kitov
and D. R. Bundle, J. Am. Chem. Soc., 2003, 125, 16271–16284; (f) M.
Ambrosi, N. R. Cameron and B. G. Davis, Org. Biomol. Chem., 2005,
3, 1593–1608; (g) L. L. Kiessling, J. E. Gestwicki and L. E. Strong,
Angew. Chem., Int. Ed., 2006, 45, 2348–2368.
33 (a) D. Choudhury, A. Thompson, V. Stojanoff, S. Langerman, J.
Pinkner, S. J. Hultgren and S. Knight, Science, 1999, 285, 1061–1066;
(b) C. S. Hung, J. Bouckaert, D. Hung, J. Pinkner, C. Widberg, A.
Defusco, C. G. Auguste, R. Strouse, S. Langermann, G. Waksman and
S. J. Hultgren, Mol. Microbiol., 2002, 44, 903–918; (c) J. Bouckaert, J.
Berglund, M. Schembri, E. D. Genst, L. Cools, M. Wuhrer, C.-S. Hung,
J. Pinkner, R. Sl a¨ tteg a˚ rd, A. Zavialov, D. Choudhury, S. Langermann,
S. J. Hultgren, L. Wyns, P. Klemm, S. Oscarson, S. D. Knight and H. D.
Greve, Mol. Microbiol., 2005, 55, 441–455.
5
0, 21–123.
0 N. R o¨ ckendorf, O. Sperling and T. K. Lindhorst, Aust. J. Chem., 2002,
5, 87–93.
1 (a) R. M. Rowell and M. S. Feather, Carbohydr. Res., 1967, 4, 486–
1
1
5
4
1
91; (b) G. Excoffier, D. Gagnaire and J. O. Utille, Carbohydr. Res.,
975, 39, 368–373; (c) A. Nudelman, J. Herzig, H. E. Gottlieb and J.
Stirling, Carbohydr. Res., 1987, 162, 145–152; (d) J. Zhang and P. Kov a´ cˇ ,
J. Carbohydr. Chem., 1999, 18, 461–469; (e) J. Fiandor, M. T. Garc ´ı a-
L o´ pez, G. G. de las Heras and P. P. M e´ ndez-Castrill o´ n, Synthesis, 1985,
1
121–1123; (f) K. Bhaumik, P. D. Salgaonkar and K. G. Akamanchi,
Aust. J. Chem., 2003, 56, 909–911.
1
1
2 A. V. Nikolaev, I. A. Ivanova, V. N. Shibaev and N. K. Kochetkov,
Carbohydr. Res., 1990, 204, 65–78.
3 Dimethylamine in ethanol can be used for anomeric deacetylation
and debenzoylation of glucose and galactose derivatives as well. In
these cases THF should be used as the solvent instead of pyridine.
The reactivities in this anomeric deprotection reaction were found to
decrease as follows: Ac
Bz Gal.
5
Glc = Ac
5
Man < Ac
5
Gal ꢁ Bz
5
Glc < Bz
5
Man
34 T. K. Lindhorst, S. K o¨ tter, U. Krallmann-Wenzel and S. Ehlers,
J. Chem. Soc., Perkin Trans. 1, 2001, 823–831.
ꢁ
5
3
912 | Org. Biomol. Chem., 2006, 4, 3901–3912
This journal is © The Royal Society of Chemistry 2006