Cluster mannosides as inhibitors of type 1 fimbriae-mediated adhesion of Escherichia coli: Pentaerythritol derivatives as scaffolds

Article abstract of DOI:10.1002/1099-0690(200006)2000:11<2027::AID-EJOC2027>3.0.CO;2-L

Pentaerythritol derivatives were used as core molecules for the synthesis of two cluster α-D-mannosides, which were designed as oligomannoside mimetics. The problem of glycosyl orthoester formation, which frequently occurs in oligo-mannosylations, was solved. The clusters were tested for their capacity to block binding of Escherichia coli to yeast mannan in vitro and were found to be more than 200 times more potent in inhibiting mannose-specific adhesion than methyl α-D-mannoside.

Full text of DOI:10.1002/1099-0690(200006)2000:11<2027::AID-EJOC2027>3.0.CO;2-L

FULL PAPER
Cluster Mannosides as Inhibitors of Type 1 Fimbriae-Mediated Adhesion of
Escherichia coli: Pentaerythritol Derivatives as Scaffolds
Thisbe K. Lindhorst,*[a] Michael Dubber,[a] Ulrike Krallmann-Wenzel,[b] and Stefan Ehlers[b]
Keywords: Cluster mannosides / Type 1 fimbriae / Antiadhesives / Oligosaccharides / Glycosylations
Pentaerythritol derivatives were used as core molecules for
capacity to block binding of Escherichia coli to yeast mannan
the synthesis of two cluster α- -mannosides, which were de- in vitro and were found to be more than 200 times more po-
D
signed as oligomannoside mimetics. The problem of glycosyl
orthoester formation, which frequently occurs in oligo-man-
nosylations, was solved. The clusters were tested for their
tent in inhibiting mannose-specific adhesion than methyl α-
-mannoside.
D
Introduction
target cells, bacteria use multivalent contacts between their
own lectins and parts of the host cell glycocalyx. Some bac-
terial lectin domains are assembled on long, proteinogenous
appendages on the bacterial surface, which are called pili or
fimbriae. Fimbriae are classified according to their sugar
specificity and mannose-specific fimbriae are called ‘‘type 1
fimbriae’’. While type 1-fimbriated E. coli are normal in-
habitants of the colon, they cause infections if they colonize
the urinary tract. Type 1 fimbriae of E. coli substantially
contribute to the adhesion and virulence of this micro-
organism.[6]
A most exciting feature of carbohydrates is their ability
to encode biological information in the form of three-di-
mensional arrangements of atoms and functional groups,
e.g. those represented by the complex oligosaccharide por-
tions present in glycoproteins and glycolipids.[1] This in-
formation is translated into a biological effect by the forma-
tion of noncovalent complexes between segments of oligos-
accharide structures and specialized proteins, the lectins
and selectins,[2] respectively. Such carbohydrate-protein
complexes are currently under investigation in order to elu-
cidate their structural and functional details.[3] One of the
tools optimally suited for the investigation and manipula-
tion of carbohydrate-protein interactions consists of oligos-
accharide mimetics that are structurally modeled on their
naturally occurring counterparts. The design of oligosacch-
aride mimetics may allow both an easier synthetic access to
the necessary ligands and the optimization of their affinities
to the respective receptors.[4]
A rational approach to the synthesis of oligosaccharide
mimetics involves subjecting the naturally occurring struc-
tures to a ‘‘molecular striptease’’ that substitutes the inner
regions of an oligosaccharide by a simple, noncarbohydrate
core molecule. The latter can then be functionalized with
the carbohydrate epitopes, which are essential for receptor
binding. In this way, oligoantennary oligosaccharides of
naturally occurring glycoconjugates may be mimicked by
so-called cluster glycosides.[5]
During our ongoing research program aimed at inhibit-
ing the mannose-specific adhesion of Escherichia coli bac-
teria to their host cells,[7] we synthesized a number of differ-
ent cluster mannosides to mimic structural ensembles pre-
sent in high-mannose type N-glycans.[8] Some tri- and tetra-
valent glycoclusters turned out to be especially potent
inhibitors of bacterial binding. We conclude from these re-
sults that the synthetic glycoclusters, completely or par-
tially, fit into the carbohydrate recognition domain (CRD)
of the bacterial lectin. This finding is in keeping with a
structural model for the bacterial CRD used for mannose-
specific adhesion, according to which the CRD consists of
three subsites and may accommodate mannose-containing
trisaccharides of a specific three-dimensional structure.[9]
These findings make it an attractive scientific goal to im-
prove the synthesis, yield and performance of cluster man-
nosides that can serve as oligomannoside mimetics and po-
tentially inhibit mannose-specific bacterial adhesion.
Such cluster glycosides are promising antiadhesives to
block bacterial infection, as bacterial adhesion is often a
prerequisite for infection. To adhere to the surface of their
In this context, we recently turned our attention to penta-
erythritol as a scaffold for clustering. Pentaerythritol has
been used successfully as a branched core molecule for the
preparation of glycomimetics as well as dendrimers.[10] It
was first used for sugar clustering with the goal of produ-
cing potentially megacaloric nutrients.[11] In this and other
cases,[12] as well as for the synthesis of sugar-based clus-
ters,[11] ester linkages were used for clustering. Peptide link-
ages were chosen to assemble tumor-associated saccharides
en route to haptens for anti-tumor vaccine development.[13]
[a]
Institute of Organic Chemistry, University of Hamburg
Martin-Luther-King-Platz 6, 20146 Hamburg, Germany
Fax: (internat.) ϩ 49-(0)40/42838-5592
E-mail: tklind@chemie.uni-hamburg.de
[b]
Division of Molecular Infection Biology,
Research Center Borstel
Parkallee 22, 23845 Borstel, Germany,
Fax: (internat.) ϩ 49-(0)4537/188-686
E-mail: sehlers@fz-borstel.de
Eur. J. Org. Chem. 2000, 2027Ϫ2034
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1434Ϫ193X/00/0611Ϫ2027 $ 17.50ϩ.50/0
2027

T. K. Lindhorst, M. Dubber, U. Krallmann-Wenzel, S. Ehlers
FULL PAPER
Scheme 1. Pentaerythritol-based cluster glycosides as depicted were
designed to serve as high mannose-type glycan mimetics, in which
C3 spacers, highlighted by black dots, are thought to substitute
monosaccharide units
Scheme 2. (i) BnBr, NaH, DMF, 86%; (ii) 9-BBN, THF, then
NaOH and H2O2, 91%; (iii) NaH, diglyme, 62%
In addition, galabioside clusters were synthesised on the ba-
sis of pentaerythritol derivatives and proved to be excellent
in vitro inhibitors of hemagglutination caused by Strepto-
coccus suis.[14] Furthermore, pentaerythritol has been used chain as the aglycon part of the carbohydrate portion prior
as the skeleton molecule for the synthesis of sialyl Lewis-X to clustering or (ii) realizing the required C3 spacer moieties
mimetics,[15] for the synthesis of mono-, di- and tridentate at the core molecule. In both synthetic routes, an allylation-
galactose clusters carrying photoaffinity labels to block and hydroboration sequence would allow the necessary hydro-
label the hexose transport system in erythrocytes[16] and to xypropyl spacers to be established.
synthesize α--mannosyl clusters as photoaffinity ligands
When we followed the first route, the spacer portion was
introduced as the aglycon moiety in mannoside 3, in order
for Concanavalin A.[17]
For the inhibition of mannose-specific bacterial adhe- to perform the necessary glycosylation step at an early stage
sion, we designed a pentaerythritol-based cluster-manno- of the synthesis. For the synthesis of 3, allyl α--mannoside
side as depicted in Scheme 1, in which pentaerythritol itself, (1) was prepared by Fischer glycosylation, benzylated to 2
as well as the included C3 spacers, are used as structural under standard conditions, and 2 was eventually subjected
substitutes of monosaccharide moieties. This principle was to a hydroboration/oxidation reaction to obtain the primary
introduced by Lehmann and co-workers with the synthesis alcohol 3. Hydroboration of the allyl group in 2 proceeded
of so-called spacer-modified oligosaccharides to investigate strictly regioselectively when the sterically demanding hy-
the sugar binding site of α-amylase.[18] We now report on droborane 9-BBN (9-borabicyclo[3.3.1]nonane) was used as
the synthesis of the target molecule 19 as well as its trivalent the reagent. The structure of 3 could be unequivocally
analogue 21, and on their antiadhesive properties in inhibit- proved by 1H-NMR spectroscopy after acetylation of a
ing mannose-specific binding of type 1-fimbriated E. coli.
small sample. Mannoside 3 was then used as the alcohol
component in a Williamson ether synthesis with pentaer-
ythritol tetrabromide (4). This reaction, however, led to a
mixture of mono-, di-, tri-, and tetradentate products, even
when an eight-fold excess of alcohol and forced reaction
conditions were used (Scheme 2). In the best cases, the
Results and Discussion
Synthesis
A retrosynthetic analysis reveals two possibilities for the tetra-cluster 5 could be isolated in 62% yield, although
synthesis of cluster mannoside 19; (i) incorporating a C3 yields that were lower by approximately 10% were also of-
2028
Eur. J. Org. Chem. 2000, 2027Ϫ2034

Cluster Mannosides
FULL PAPER
1
Table 1. H-NMR data (400 MHz, CDCl3) of core molecules 7Ϫ9
and 11Ϫ13
Positions of recorded protons
1
2
3
4
COCH3
CH3
7
3.46
(s)
3.95
(dddd)
3.95
(mc)
3.45
(t)
5.88
(mc)
5.87
(mc)
1.86
(mc)
1.86
(mc)
1.79
(mc)
1.80
(mc)
5.25 (ddd)
5.13 (ddd)
5.20
Ϫ
Ϫ
Ϫ
11
8
3.30
(s)
1.00
(s)
Ϫ
(mc)
4.13
3.35
(s)
2.05
(s)
(t)
12
9
3.23
(s)
3.45
(t)
4.15
2.05
(s)
0.91
(s)
Ϫ
(t)
3.39
(s)
3.58
(t)
3.59
(t)
3.74
Ϫ
(t)
13
3.30
(s)
3.75
Ϫ
0.90
(s)
(t)
ively, were isolated in significant yields. Attempts to isomer-
ize 15 and 16 to the respective mannosides were not success-
ful. 1,2-O-glycosyl orthoesters can be recognized in 1H-
NMR spectra by a typical high-field shift of 2-H (δ ഠ 4.6)
and 5-H (δ ഠ 3.7) and, moreover, by the chemical shift
of the orthoacetyl CH3 group (δ ϭ 1.78), which is clearly
separated from the acetyl CH3 groups resonating at δ ഠ 2.
To circumvent orthoester formation, which was shown to
be the dominant problem of the mannosylation reaction,
benzoyl-protected mannosyl trichloroacetimidate 17[20] was
used as the glycosyl donor. No orthoester formation was
observed in this case and, upon mannosylation of 9 and 13,
the protected cluster mannosides 18 and 20, respectively,
were isolated in excellent yields after purification on silica
gel followed by gel permeation chromatography (Scheme 5).
Thus, benzoylated trichloroacetimidate 17, which is ob-
tained in three steps from mannose is an excellent mannosyl
donor in cluster mannoside synthesis and completely avoids
the problem of orthoester formation. Deprotection to af-
ford the target clusters 19 and 21, respectively, was an easy
Scheme 3. (i) Allyl bromide, aq. NaOH, TBABr, 59% for 7, 47%
for 11; (ii) 9-BBN, THF, then NaOH and H2O2, then Ac2O, pyrid-
ine, 87% for 8, 84% for 12; (iii) NaOMe, MeOH, quant.
ten obtained. This result is in accordance with the known
reluctance with which neopentyl positions, as present in 4,
react in substitution reactions.
Therefore, an alternative route to target cluster 19 was
started, in which first pentaerythritol was modified to serve
as spacer-equipped tetrafunctional core molecule in the
subsequent glycosylation step. Derivatisation of penta-
erythritol was accomplished by per-allylation of the poorly
soluble pentaerythritol by reversed phase transfer catalysis
according to a published procedure.[19] The resulting allyl
ether 7 was then converted into pure tetra-alcohol 9 by a
hydroboration/oxidation reaction of the double bonds using
9-BBN, acetylation of the product to form 8, purification
´
and quantitative reaction under Zemplen conditions (using
sodium methanolate in methanol).
Clusters 18Ϫ21 show symmetrical well-resolved NMR
spectra that exhibit the signal set for one ‘‘arm’’ of the clus-
ters. MS analyses of the clusters give single peaks at the
expected masses.
Antiadhesive Properties
´
at this stage and then Zemplen deacetylation (Scheme 3).
Bacteria use specific fimbrial lectins to adhere to the gly-
The analogous reaction sequence was carried out with cocalyx of their host cells. Since adhesion is often the first
2,2,2-trihydroxymethylethane (10) to yield 13 and to allow step in bacterial pathogenesis, an anti-adhesive intervention
the synthesis of the trivalent cluster analogue 21, which we may effectively prevent bacterial infection. Mimetics of the
intended to compare to 19 in the biological studies. NMR oligosaccharides that bind to the bacterial carbohydrate re-
data of core molecules 7Ϫ9 and 11Ϫ13 are listed in Table 1. cognition domains are likely candidates for inhibiting bac-
The subsequent mannosylation of 9 and 13 turned out to terial adhesion. The antiadhesive potencies of cluster man-
be more difficult than expected. With 2,3,4,6-tetra-O- nosides 19 and 21 were tested in an ELISA-microplate for-
acetyl-α-mannosyl bromide as the glycosyl donor, only gly- mat using a genetically engineered E. coli strain, HB101
cosyl orthoesters were obtained and by using acetylated (pPKl4),[21] which expresses only type 1 fimbriae on its sur-
mannosyl trichloroacetimidate 14 as the glycosyl donor, no face. Serially diluted aqueous solutions of clusters 19 and
improvement was achieved in this regard (Scheme 4). With 21 were included in the adhesion reaction and the cluster
TMS triflate as the catalyst, as well as with BF3 Et2O used concentration that inhibits binding by 50% was determined
as the Lewis acid, orthoester clusters 15 and 16, respect- (IC50). Both clusters worked as inhibitors of bacterial adhe-
Eur. J. Org. Chem. 2000, 2027Ϫ2034
2029

T. K. Lindhorst, M. Dubber, U. Krallmann-Wenzel, S. Ehlers
FULL PAPER
Scheme 4. (i) TMSOTf; for 15 in acetonitrile, 64%; for 16 in toluene, 54%
Scheme 5. (i) CH2Cl2, TMSOTf, 92% for 18, 95% for 20; (ii) NaOMe, MeOH, quant.
sion at concentrations in the low micromolar range, with pared to methyl α--mannoside, which was used as mono-
the tetravalent cluster 19 surpassing the inhibitory potency valent reference compound (Table 2).
of methyl mannoside by 257 and its trivalent analogue 21
by 200. On a valency-corrected basis, 19 and 21 performed
almost equally well. These results demonstrate that cluster
Conclusion
19 and also 21 can serve as oligomannoside mimetics, dis-
A number of conclusions can be drawn concerning the
playing a significantly elevated inhibitory potency com- synthesis and biological effects with type 1 fimbriae CRDs
2030
Eur. J. Org. Chem. 2000, 2027Ϫ2034

Cluster Mannosides
FULL PAPER
(ODs) were measured with a Dynatech 5000 ELISA reader at
405 nm with the reference read at 490 nm.
Table 2. Antiadhesive properties of the examined cluster mannos-
ides obtained in ELISA inhibition assays, in comparison to methyl
α--mannopyranoside (MeMan); inhibition values obtained in four
independently performed assays were averaged
Materials: Methyl α--mannoside was purchased from Merck, pen-
taerythritol tetrabromide from Aldrich, and TMSOTf (Fluka) was
used without further purification. F-shaped polystyrene plates from
Sarsted (with a lid) were used for inhibition tests. A recombinant
type 1 fimbriated E. coli strain, E. coli HB101 (pPKl4),[21] was used
and cultured as described in the literature.[8a] Mannan from Sacch-
aromyces cerevisiae was obtained from Sigma and was used in
50 m Na2CO3 solution (1 mg/mL, pH ϭ 9.6). A polyclonal anti
fimA antibody was used as the first ELISA antibody. The second
antibody was a goat-anti-rabbit peroxidase conjugated antibody
(IgG, HϩL) and was purchased from Dianova. Skimmed milk was
obtained from Glücksklee, Tween 20 from Roth, ABTS [2,2Ј-azi-
dobis(3-ethylbenzthiazoline-6-sulfonic acid)] from Sigma and thi-
merosal {2-[(ethylmercurio)thio]benzoic acid, sodium salt} from
Merck.
MeMan 19 (tetravalent) 21 (trivalent)
[a]
IC50
3000
12.5
12.5
[µM] (assay 1)
[a]
IC50 [µM] (assay 2) 3500
15
12.5
24
[a]
IC50 [µM] (assay 3) 4000
11
[a]
IC50 [µM] (assay 4) 2500
12
16
average IC50 [µM]
3250
645.5
1
12.63
5.4
257
64
16.25
1.7
standard deviation
[b]
RIC50
200
67
valency-corrected
1
[c]
RIC50
[a]
[b]
IC50: 50% of binding inhibited. Ϫ RIC50: Relative IC50 based
[c]
on MeMan (ϵ1). Ϫ
Relative IC50 based on moles mannoside
contained in one mole cluster.
Buffers: PBS (phosphate buffer saline) was prepared by dissolving
8 g of NaCl, 0.2 g of KCl, 1.44 g of Na2HPO4 ϫ 2 H2O and 0.2 g
of KH2PO4 in 1000 mL of distilled deionised water (pH ϭ 7.2).
PBSE (PBS buffer used for ELISA) was used as a 150 m solution
[pH ϭ 7.2, NaCl (136.90 m), KCl (2.68 m), Na2HPO4 (8.0 m),
KH2PO4 (2.4 m), 0.01% thimerosal]. PBSET was PBSE buffer ϩ
0.05% Tween 20. Substrate buffer was 0.1  sodium citrate dihy-
drate, adjusted to pH ϭ 4.5 with citric acid. The ABTS solution
of cluster mannosides such as 19 and 21. (i) The C3 spacer-
equipped branched oligofunctional alcohols 9 and 13 used
as glycosyl donor scaffolds can be readily prepared by an
allylation-hydroboration sequence starting from the
sterically more hindered alcohols 6 and 10, respectively. (ii)
These oligoalcohols can be quantitatively α-mannosylated
using benzoylated mannosyl trichloroactetimidate 17 as the was prepared by dissolving ABTS (1 mg) in substrate buffer (1 mL)
by sonication followed by the addition of 0.1% H2O2 (25 µL per
mL).
glycosyl donor without orthoester formation. (iii) Cluster
mannosides 19 and 21 are potent inhibitors of mannose-
specific adhesion of E. coli in the ELISA.
The observed inhibition of bacterial adhesion is most
likely explained by binding of cluster glycosides 19 and 21
ELISA: The ELISA protocol used to determine the inhibitory po-
tencies of the cluster mannosides towards type 1 fimbriated E. coli
was carried out according to a published procedure.[8c] ELISA
to single carbohydrate recognition domains (CRDs), which plates were incubated at 37 °C. In preliminary experiments, the
optimal concentrations for mannan, bacteria and antibody solu-
tions were determined. Microtiter plates were coated with mannan
solution (100 µL per well) and dried overnight at 37 °C. The plates
were washed once with PBS (150 µL per well) and then blocked
with 5% skimmed milk in PBSE for 30 min at 37 °C. The wells
were washed with PBSE (150 µL) and then PBSE (50 µL) and
inhibitor solution (50 µL) were added. Inhibitor solutions were di-
luted serially twofold in PBSE. Bacterial suspension (50 µL per
well) was added and the plate was left at 37 °C for 1 h to allow
are distributed along type 1 fimbriae, rather than by multi-
valent binding, which would reflect interaction of the sugar
clusters with more than one CRD.[22]
Experimental Section
General Methods: Flash chromatography was performed using
Merck silica gel 60 (0.040Ϫ0.063 mm, 230Ϫ400 mesh). Ϫ TLC was sedimentation of the bacteria. Each well was then washed four
performed on Kieselgel 60 F254 plates from Merck. Detection was
carried out under UV light or by spraying with 20% ethanolic sulf-
times with PBSE (150 µL) and then 50 µL of the first antibody (anti
fimA antibody, solution as optimized prior to the experiments) in
uric acid followed by heating. Core molecules were visualized with 2% skimmed milk was added. The plates were incubated for 30 min
anisaldehyde reagent[23] and subsequent heating. Ϫ Size exclusion
and then washed twice with PBSET and the second antibody was
chromatography was performed on Sephadex LH-20 from Pharma-
added (50 µL). The plates were incubated for 30 min and sub-
cia. Ϫ NMR spectra were measured with a Bruker AMX 400 sequently washed three times with PBSET, PBSE and substrate
1
(400 MHz for H and 100.67 MHz for 13C NMR) or a DRX 500 buffer. ABTS solution (50 µL) was added, the mixture incubated
(500 MHz for 1H and 125.84 MHz for 13C NMR) spectrometer.
Chemical shifts are in ppm, relative to internal TMS (δ ϭ 0.00 for
1H and 13C NMR) or solvent peaks, which were calibrated as fol-
for 30 min at 37 °C and the reaction was stopped by adding 2%
oxalic acid (50 µL). IC50 values reflect the inhibitor concentration
that causes 50% inhibition of bacterial binding to yeast mannan.
lows: CDCl3 (δ ϭ 7.26 for 1H and δ ϭ 77.00 for 13C NMR), The percentage inhibition was calculated as [OD(nI)
Ϫ
[D4]methanol (δ ϭ 3.35 for 1H and δ ϭ 49.30 for 13C NMR). OD(I)] ϫ 100 ϫ [OD(nI)]Ϫ1. Duplicate results were used for the
Where necessary, two-dimensional 1H-1H or 1H-13C COSY experi-
ments were performed for complete signal assignments. Ϫ Optical
rotation values were obtained using a 341 or 243 PerkinϪElmer
polarimeter (Na-D line, 589 nm, cell length 10 cm). Ϫ Mass spectra
were measured with a VG Analytical 70Ϫ250S (FAB MS) or
Bruker Biflex III (MALDI-TOF MS) instrument. The matrix used
for MALDI was dihydroxybenzoic acid (DHB) in water/acetonitr-
construction of the inhibition curves for each individual ELISA
experiment. Typically, the intra-assay variation of an individual
ELISA is very small, whereas the IC50 values obtained from several
independently performed ELISAs differ significantly. However,
when relative inhibition potencies were used to determine the inhib-
itory potencies of the cluster mannosides towards type 1 fimbriated
E. coli the results are highly reproducible. For ELISA controls, the
ile (2:1) with 0.1% trifluoroacetic acid was used. Optical densities bacterial adhesion to blocked, uncoated microtiter plates was
Eur. J. Org. Chem. 2000, 2027Ϫ2034 2031

T. K. Lindhorst, M. Dubber, U. Krallmann-Wenzel, S. Ehlers
FULL PAPER
checked, and the reaction of antibodies with carbohydrate derivat-
was raised to 40 °C and then allyl bromide (10 mL, 118 mmol) was
ives as well as the reaction of the employed antibodies with yeast added dropwise over 1 h. The reaction mixture was stirred for 12 h
mannan was tested and found to be negligible. The low background
obtained from these control experiments was subtracted when cal-
culating the IC50 values.
at room temp. so vigorously that it appeared to be homogeneous.
Toluene (100 mL) was added, the organic phase was washed with
water, concentrated in vacuo and purified by flash chromatography
(toluene/ethyl acetate, 20:1) to yield the per-allylated product
(3.84 g, 13 mmol, 59%) as a colorless syrup; for NMR data see
Table 1.
3-Hydroxypropyl 2,3,4,6-Tetra-O-benzyl-α-D-mannopyranoside (3):
To a solution of benzylated allyl mannoside 2 (4.0 g, 6.89 mmol) in
dry dioxane (40 mL) was added 9-BBN (30 mL of a 0.5  solution
in THF) dropwise at 0 °C under argon. The reaction mixture was
stirred at room temp. for 20 h, cooled to 0 °C and 3  aq. NaOH
(80 mL) and 30% H2O2 (8 mL) were added. The reaction mixture
was stirred at room temp. for 2 d, the phases were separated and the
aqueous phase was extracted three times with ethyl acetate (100 mL
each). The organic phases were combined, washed with satd. NaCl
solution, and dried with MgSO4. The solution was filtered and the
filtrate concentrated in vacuo to give an almost colorless syrup,
which was purified by flash chromatography (toluene/ethyl acetate,
1:1) to yield the title alcohol (3.72 g, 6.2 mmol, 91%) as a colorless
syrup. Ϫ To facilitate the structural characterization by NMR, a
small amount of 3 was acetylated at room temp. in acetic anhyd-
ride/pyridine, coevaporated with toluene after complete conversion
and filtered through a short silica gel column to yield 3-acetoxypro-
pyl 2,3,4,6-tetra-O-benzyl-α--mannopyranoside. Ϫ 3: 1H NMR
(400 MHz, CDCl3): δ ϭ 1.79 (m, 2 H, CϪCH2ϪC), 3.50 (m, 1 H,
man-OϪCHH), 3.63Ϫ3.79 (m, 6 H, CHHϪOH, man-OϪCHH, 2-
H, 5-H, 6-H, 6-HЈ), 3.93 (t, 1 H, 4-H), 3.85 (m, 2 H, 3-H,
CHHϪOH), 4.47Ϫ4.77 (m, 7 H, 3 CH2-benzyl, CHH-benzyl), 4.85
(d, 1 H, J ϭ 2.0 Hz, 1-H), 4.86 (d, 1 H, CHH-benzyl), 7.10Ϫ7.40
Tetra-O-(3-acetoxypropyl)pentaerythritol (8): Allylated penta-
erythritol 7 (240 mg, 0.8 mmol) was dissolved in dry THF (50 mL),
9-BBN (14 mL, 7 mmol) was added and the reaction mixture was
stirred under reflux for 1 h. The excess hydride was destroyed by
adding a small amount of water. The mixture was cooled to 0 °C,
3  aq. NaOH (7 mL) was added, followed by dropwise addition
of 30% H2O2 (7 mL). The reaction mixture was stirred overnight
at room temp., the aqueous phase was carefully saturated with
K2CO3 and the organic phase was separated. The K2CO3-saturated
aqueous phase was extracted twice with THF (40 mL each) and
the combined organic phases were dried with MgSO4. After filtra-
tion, the filtrate was concentrated in vacuo, the residual syrup dis-
solved in pyridine (6 mL) and acetic anhydride (6 mL, 63 mmol)
and stirred at room temp. for 2 h. The mixture was coevaporated
with toluene on the rotary evaporator and purified on silica gel by
flash chromatography (petroleum ether/ethyl acetate, 3:2) to yield
the title compound (378 mg, 0.7 mmol, 87%) as a colorless syrup;
for NMR data see Table 1.
Tetra-O-(3-hydroxypropyl)pentaerythritol (9): The acetylated deriv-
ative 8 (430 mg, 0.68 mmol) was dissolved in dry MeOH (20 mL),
a little sodium was added to afford a catalytic amount of sodium
methoxide and the reaction mixture was stirred overnight at room
temp. The mixture was neutralized with ion exchange resin (Dowex
W50 Hϩ), filtered and the filtrate was concentrated in vacuo to
afford tetraol 9 (290 mg, 0.69 mmol, quant.) as a colorless syrup;
for NMR data see Table 1.
(m, 20 H, aryl-H). Ϫ 3-Acetoxypropyl 2,3,4,6-Tetra-O-benzyl-α-D-
mannopyranoside: 1H NMR (400 MHz, CDCl3): δ ϭ 1.85 (mc, 2 H,
CϪCH2ϪC), 2.00 (s, 3 H, OAc), 3.42 (m, 1 H, OϪCHHϪCH2),
3.70Ϫ3.80 (m, 5 H, OCHHϪCH, 2-H, 5-H, 6-H, 6-HЈ), 3.89 (dd,
J ϭ 3.1, 10.1 Hz, 1 H, 3-H), 3.98 (t, J ϭ 10.1 Hz, 1 H, 4-H), 4.10
(m, 2 H, CH2OAc), 4.48Ϫ4.57 (m, 2 H, CH2-benzyl), 4.62Ϫ4.79
(m, 5 H, 2 ϫ CH2-benzyl, CHH-benzyl), 4.84 (d, J ϭ 1.5 Hz, 1 H,
1-H,), 4.87 (d, 1 H, CHH-benzyl), 7.10Ϫ7.40 (m, 20 H, aryl-H).
2,2,2-Tris(allyloxymethyl)ethane (11): 2,2,2-Tris(hydroxymethyl)-
ethane (10, 2 g, 16.6 mmol) was dissolved in 33% aqueous NaOH
solution (50 mL) at 40 °C. Allyl bromide (7.5 g, 62 mmol) and
tetrabutylammonium bromide (2 g, 6.2 mmol) were added and the
reaction mixture was stirred vigorously for 7 h at 40 °C. Water was
added, the aqueous phase was extracted three times with CH2Cl2
(100 mL each) and the combined organic phases were dried
(MgSO4). After filtration, the solvent was removed in vacuo. Flash-
chromatographic purification (toluene/ethyl acetate, 20:1) afforded
the title compound (1.87 g, 7.8 mmol, 47%) as a colorless syrup;
for NMR data see Table 1.
Benzyl-Protected Tetravalent Cluster Mannoside 5: Mannoside 3
(500 mg, 0.835 mmol) was dissolved in diglyme (diethylene glycol
dimethyl ether, 10 mL) and stirred with sodium hydride (100 mg)
for 1 h at room temp. under exclusion of moisture (N2). A solution
of pentaerythritol tetrabromide (4, 40 mg, 0.103 mmol) in diglyme
(5 mL) was added dropwise at reflux temperature and the reaction
mixture was stirred under reflux for 30 h. MeOH (10 mL) was ad-
ded at room temp., the mixture was filtered through a Celite bed
and the filtrate was concentrated in vacuo. Flash-chromatographic
purification of the residue (toluene/ethyl acetate, 10:1) afforded the
title cluster (157 mg, 0.063 mmol, 62%) as colorless syrup. Ϫ 1H
NMR (400 MHz, CDCl3): δ ϭ 1.84 (m, 8 H, CH2ϪOϪCH2Ϫ
CH2ϪCH2ϪO-man), 3.43 (m, 12 H, 4 ϫ CHHϪO-man, 4 ϫ
CϪCH2ϪO), 3.73 (mc, 20 H, 4 ϫ 2-H, 4 ϫ 3-H, 4 ϫ 5-H, 4 ϫ 6-
HЈ, 4 ϫ CHHϪO-man), 3.89 (dd, 4 H, 6-H), 3.98 (t, 4 H, 4-H),
4.05 (m, 8 H, CH2ϪOϪCH2ϪCH2ϪCH2ϪO-man), 4.44Ϫ4.85 (m,
32 H, benzylϪCH2), 4.83 (dd, 4 H, 1-H), 7.12Ϫ7.40 (m, 80 H, aryl-
2,2,2-Tris(acetoxypropyloxymethyl)ethane (12): The allylated core
molecule 11 (0.67 g, 2.79 mmol) was dissolved in dry dioxane
(20 mL) under N2 and the solution was cooled to 0 °C. 9-BBN
(51 mL of a 50% solution in THF, 25.5 mmol) was added dropwise
and the reaction mixture was stirred at room temp. for 24 h. The
mixture was cooled to 0 °C and 3  aq. NaOH (50 mL) was added
followed by the dropwise addition of 30% H2O2 (8 mL). The reac-
tion mixture was stirred overnight at room temp., the aqueous
phase was carefully saturated with K2CO3, and the organic layer
was separated. The K2CO3-saturated aqueous phase was extracted
twice with THF (40 mL each) and the combined organic phases
were dried with MgSO4. The solution was filtered and the filtrate
concentrated in vacuo. The residue was dissolved in pyridine
H).
Ϫ
13C NMR (100.67 MHz, CDCl3):
δ
ϭ
35.1
(OϪCH2ϪCH2ϪCH2ϪO), 44.1 (Cq, core-C), 68.0, 69.2
(OϪCH2ϪCH2ϪCH2ϪO), 69.3 (CqϪCH2ϪO), 72.0, 72.4, 73.2,
75.0 (benzylϪCH2), 64.3 (C-6), 71.7, 74.7, 74.8, 80.1 (C-2, C-3, C-
4, C-5), 97.8 (C-1), 127.3Ϫ138.3 (aryl-C). Ϫ (FAB-MS); m/z: 2460.3
[M ϩ Hϩ]; 2459.025 calcd. for C153H172O28.
Tetra-O-allyl-pentaerythritol (7): Pentaerythritol (6, 3 g, 22 mmol) (25 mL), acetic anhydride (20 mL) was added and the solution was
was dissolved in 33% aqueous NaOH solution (50 mL). Tetrabu-
tylammonium bromide (2 g, 6.2 mmol) was added, the temperature
stirred at room temp. for 2 h. The mixture was coevaporated with
toluene in the rotary evaporator and purified on silica gel by flash
2032
Eur. J. Org. Chem. 2000, 2027Ϫ2034

Cluster Mannosides
FULL PAPER
chromatography (toluene/ethyl acetate, 4:1) to yield the title com-
pound (680 mg, 1.62 mmol, 58%) as a colorless syrup; for NMR
data see Table 1.
76.3 (C-2, C-3, C-4, C-5), 77.3 (H3CϪCϪCH2), 97.2 (C-1), 124.8
(orthoacetyl-C), 169.0, 169.8, 170.2 (CϭO).
Ϫ
FAB-MS;
C53H78O30 [Mϩ]: calcd.1194.45; found 1193.98.
2,2,2-Tris(hydroxypropyloxymethyl)ethane (13): The acetylated de-
rivative 12 (600 mg, 1.42 mmol) was dissolved in dry MeOH
(30 mL), a catalytic amount of sodium was added and the reaction
mixture was stirred overnight at room temp. The mixture was neut-
ralized with ion exchange resin (Dowex W50 Hϩ), filtered and the
filtrate was concentrated in vacuo to afford triol 13 (420 mg,
quant.) as colorless syrup; for NMR data see Table 1.
Benzoyl-Protected Tetravalent Cluster Mannoside 18: Tetraol 9
(14 mg, 0.038 mmol) and the benzoylated mannosyl trichloroaceti-
midate 17 (0.91 g, 1.2 mmol) were dissolved in dry CH2Cl2 (8 mL)
under N2. TMSOTf (5% solution in CH2Cl2, 0.5 mL) was added
and the reaction mixture was stirred overnight at room temp.
NaHCO3 (5 g) was added and the suspension was quantitatively
transferred (rinse with CH2Cl2) to a silica gel column and purified
by flash chromatography (light petroleum ether/ethyl acetate, 2:1
Ǟ 1:2). This was followed by a second purification step on Se-
phadex LH-20 (elution with CH2Cl2/MeOH, 1:1) to yield the title
cluster (95 mg, 0.035 mmol, 92%) as a white amorphous solid. Ϫ
[α]2D1 ϭ Ϫ47.2 (c ϭ 0.63 in CH2Cl2) Ϫ 1H NMR (500 MHz,
CDCl3): δ ϭ 1.98 (mc, 8 H, 4 ϫ OCH2CH2CH2O), 3.48 [br. s, 8
H, (ROCH2)4C], 3.56 [mc, 8 H, (RCH2OCH2)4C], 3.69 (mc, 4 H, 4
ϫ manOCHH), 3.91 (mc, 4 H, 4 ϫ manOCHH), 4.44 (dddd ഠ m,
4 H, 4 ϫ 5-H), 4.47 (dd, J ϭ 4.1, 11.2 Hz, 4 H, 4 ϫ 6-HЈ), 4.69
(dd, J ϭ 1.5, 11.2 Hz, 4 H, 4 ϫ 6-H), 5.09 (d, J ϭ 1.5 Hz, 4 H, 4
ϫ 1-H), 5.70 (dd, J ϭ 1.5, 3.6 Hz, 4 H, 4 ϫ 2-H), 5.92 (dd, J ϭ
3.6, 10.2 Hz, 4 H, 4 ϫ 3-H), 6.13 (dd ഠ t, J ϭ 10.2 Hz, 4 H, 4 ϫ
4-H), 7.19Ϫ7.47, (m, 40 H ϩ CHCl3, 40 ϫ aryl-H), 7.55 (m, 8 H,
8 ϫ aryl-H), 7.82, 7.93, 8.03, 8.09 (each 8 H, each mഠd, 32 ϫ o-
aryl-H). Ϫ 13C NMR (100.67 MHz, CDCl3): δ ϭ 30.2 (CH2,
OCH2CH2CH2O), 46.0 (Cq), 63.3 (CH2, C-6), 66.3 [CH2, (man)-
OCH2], 67.4 (CH, C-4), 68.4 [CH2, (RCH2OCH2)4C], 69.2 (CH, C-
5), 70.4 [CH2, (ROCH2)4C], 70.6 (CH, C-3), 71.0 (CH, C-2), 98.2
(CH, C-1), 128.7, 128.8, 128.9, 129.0, 129.4, 129.6, 129.8, 130.1,
130.2, 130.3, 130.4, 133.4, 133.5, 133.8 (CH, aryl-C), 165.7, 165.8,
165.9, 166.5 (C, 16 ϫ CϭO). Ϫ MALDI-TOF-MS; C153H140O44
[M ϩ Na]ϩ: calcd. 2680.87; found 2703.83 .
Tetravalent 1,2-O-(Orthoacetyl)-β-D-mannopyranose Cluster 15:
The branched tetraol 9 (0.099 g, 0.275 mmol) and the trichloroacet-
imidate 14 (0.89 g, 1.82 mmol) were combined and coevaporated
three times with dry toluene (15 mL each) to remove traces of mois-
ture. This mixture was then dissolved in dry acetonitrile under ni-
trogen and TMSOTf (0.5 mL of a 0.02  solution in CH2Cl2) was
added at 0 °C. The reaction temperature was allowed to rise to
room temp. and after 2 h of stirring at ambient temperature more
TMSOTf (0.5 mL of a 0.02  solution in CH2Cl2) was added. The
reaction mixture was stirred at room temp. overnight, triethylamine
(100 µL) was added and the mixture was concentrated in vacuo.
Repeated flash chromatography (n-hexane/acetone, 4:5) gave the
title compound (0.298 g, 0.176 mmol, 64%) as colorless syrup. Ϫ
1H NMR (400 MHz, CDCl3): δ ϭ 1.73 (s, 12 H, 4 ϫ orthoacetyl-
CH3), 1.78 (mc, 8 H, 4 ϫ OϪCH2ϪCH2ϪCH2ϪO), 2.04, 2.06, 2.10
[3 ϫ s, each 12 H, each 4 ϫ C(O)CH3], 3.31 (m, 8 H, 4 ϫ
CϪCH2ϪO), 3.41 (mc, 8 H, 4 ϫ CCH2OCH2), 3.55 (mc, 8 H, 4 ϫ
manOCH2), 3.68 (ddd, 4 H, 4 ϫ 5-H), 4.14 (dd, J ϭ 2.5, 12.0 Hz,
4 H, 4 ϫ 6-H), 4.22 (dd, J ϭ 5.0, 12.0 Hz, 4 H, 4 ϫ 6-HЈ), 4.58
(dd ഠ t, 4 H, 4 ϫ 2-H), 5.16 (dd, J ϭ 4.0, 9.5 Hz, 4 H, 4 ϫ 3-H),
5.28 (dd ഠ t, J ϭ 9.5 Hz, 4 H, 4 ϫ 4-H), 5.47 (d, J ϭ 2.5 Hz, 4 H,
4 ϫ 1-H). Ϫ 13C NMR (100.62 MHz, CDCl3): δ ϭ 20.7, 20.7, 20.7
[3 ϫ C(O)CH3], 24.7 (orthoacetyl-CH3), 29.8, (OϪCH2ϪCH2Ϫ Unprotected Tetravalent Cluster Mannoside 19: The protected clus-
CH2ϪO), 45.4 (Cq, core-C), 59.8* (OϪCH2ϪCH2ϪCH2ϪO), 62.4 ter 18 (94 mg, 0.035 mmol) was dissolved in dry MeOH (100 mL).
(C-6), 65.7 (C-4), 68.0* (OϪCH2ϪCH2ϪCH2ϪO), 69.6 Sodium methoxide (1  in MeOH, 1 mL) was added and the reac-
(CϪCH2ϪO), 70.6 (C-3), 71.4 (C-5), 76.5 (C-2), 97.4 (C-1), 124.3
[orthoacetyl-C(OR)3], 169.4, 170.3, 170.6 [3 ϫ C(O)CH3]; assign-
ments of signals denoted by * may be interchanged. Ϫ FAB-MS;
C73H108O44 [Mϩ]: calcd. 1689.63 ; found 1689.45.
tion mixture was stirred at room temp. for 14 d. The mixture was
neutralized with ion exchange resin (Amberlite IR 120 Hϩ), filtered
and the filtrate concentrated in vacuo. The residue was purified by
size exclusion chromatography on Sephadex LH-20 (elution with
MeOH) to yield the unprotected title cluster (38 mg, 0.037 mmol,
quant.) as a colorless amorphous solid. Ϫ [α]1D9 ϭ ϩ60.2 (c ϭ 1.04,
Trivalent 1,2-O-(Orthoacetyl)-β-D-mannopyranose Cluster 16: Triol
13 (50 mg, 0.25 mmol) and trichloroacetimidate 14 (1.0 g,
2.0 mmol) were combined and coevaporated three times with dry
toluene to remove traces of moisture. The mixture was kept under
nitrogen and dry toluene (20 mL) was added to dissolve the starting
materials. The solution was cooled to 0 °C and TMSOTf (0.5 mL
of a 0.02  solution in CH2Cl2) was added to start the reaction.
The reaction mixture was stirred overnight at room temp. Triethyla-
mine (100 µL) was added and the mixture was concentrated in va-
cuo. Flash-chromatographic purification (toluene/acetone, 4:1)
gave the tris(orthoester) 16 (161 mg, 0.13 mmol, 54%) as a white
1
CHCl3). Ϫ H NMR (400 MHz, [D4]MeOH): δ ϭ 1.88 (mc, 8 H,
4 ϫ OCH2CH2CH2O), 3.42 [dd, 8 H, (ROCH2)4C], 3.50Ϫ3.59 [m,
16 H, 4 ϫ 5-H, 4 ϫ manOCHH, (RCH2OCH2)4C], 3.67 (dd ഠ t,
J ϭ 9.2 Hz, 4 H, 4 ϫ 4-H), 3.74 (dd, J ϭ 3.6, 9.2 Hz, 4 H, 4 ϫ 3-
H), 3.77 (dd, J ϭ 5.1, 12.2 Hz, 4 H, 4 ϫ 6-HЈ), 3.81Ϫ3.90 (m, 12
H, 4 ϫ 6-H, 4 ϫ 2-H, 4 ϫ manOCHH), 4.79 (d, J ϭ 1.2 Hz, 1 H,
4 ϫ 1-H). Ϫ 13C NMR (100.67 MHz, [D]4MeOH): δ ϭ 32.0 (CH2,
OCH2CH2CH2O), 47.8 (Cq), 64.0 (CH2, C-6), 66.7 (CH2, man-
OCH2CH2CH2OCH2), 69.7 (CH, C-4), 70.4 (CH2, OCH2CH2CH2-
OCH2), 72.0 [CH2, (ROCH2)4C], 73.4 (CH, C-2), 73.8 (CH, C-3),
75.3 (CH, C-5), 102.8 (CH, C-1). Ϫ MALDI-TOF-MS; C41H76O28
[M ϩ Na]ϩ: calcd. 1016.45; found 1139.36 .
1
foam. Ϫ H NMR (400 MHz, CDCl3): δ ϭ 0.87 (s, 3 H, CϪCH3),
1.73 (s,
9
H,
3
ϫ
orthoacetyl-CH3), 1.78 (mc,
6 H,
OCH2ϪCH2ϪCH2O), 2.04, 2.07, 2.10 (3 ϫ s, each 9 H, each 3 ϫ
COCH3), 3.20 (s, 6 H, 3 ϫ CCH2O), 3.40 (mc, 6 H, 3 ϫ Benzoyl-Protected Trivalent Cluster Mannoside (20): Triol 13
CCH2OCH2), 3.56 (mc, 6 H, 3 ϫ manOCH2), 3.70 (ddd, J ϭ 2.5, (58 mg, 0.20 mmol) and the benzoylated mannosyl trichloroacetim-
5.1, 9.7 Hz, 3 H, 3 ϫ 5-H), 4.17 (dd, J ϭ 2.5, 12.2 Hz, 3 H, 3 ϫ idate 17 (4.45 g, 6.0 mmol) were dissolved in dry CH2Cl2 (6 mL)
6-HЈ), 4.21 (dd, J ϭ 5.1, 12.2 Hz, 3 H, 3 ϫ 6-H), 4.60 (dd, J ϭ 2.5, under N2. TMSOTf (5% solution in CH2Cl2, 0.1 mL) was added
4.1 Hz, 3 H, 3 ϫ 2-H), 5.18 (dd, J ϭ 4.1, 9.7 Hz, 3 H, 3 ϫ 3-H),
5.30 (dd ഠ t, J ϭ 9.7 Hz, 3 H, 3 ϫ 4-H), 5.48 (d, J ϭ 2.5 Hz, 3 H, NaHCO3 (5 g) was added and the suspension was quantitatively
3 ϫ 1-H). Ϫ 13C NMR (100.62 MHz, CDCl3): δ ϭ 20.5, 20.6, 20.6
transferred (rinse with CH2Cl2) to a silica gel column and purified
and the reaction mixture was stirred overnight at room temp.
(3 ϫ COCH3), 24.5 (OCCH3), 29.6 (CϪCH2ϪC), 40.8 (Cq, core- by flash chromatography (light petroleum ether/ethyl acetate, 2:1
C), 62.3 (C-6), 59.6, 67.8 (OCH2ϪCH2ϪCH2ϪO), 65.5, 70.4, 71.2, Ǟ 1:1). This was followed by a second purification step on Se-
Eur. J. Org. Chem. 2000, 2027Ϫ2034
2033

T. K. Lindhorst, M. Dubber, U. Krallmann-Wenzel, S. Ehlers
FULL PAPER
phadex LH-20 (elution with CH2Cl2/MeOH, 1:1) to yield the pro-
tected title cluster (389 mg, 0.19 mmol, 95%) as a white amorphous
SFB 470 is gratefully acknowledged. We are thankful to Robert
Meinecke and Christian Seeberger for their assistance and to Dr.
solid. Ϫ [α]2D4 ϭ Ϫ41.9 (c ϭ 0.97 in CH2Cl2). Ϫ 1H NMR Markus Scheuring and Dr. Lore Brade for their friendly advice.
(400 MHz, CDCl3): δ ϭ 0.97 (s, 3 H, CH3), 1.96 (mc, 6 H, 3 ϫ
OCH2CH2CH2O), 3.33 [s, 6 H, (ROCH2)3CCH3], 3.55 (t, 6 H, 3 ϫ
[1]
[1a] A. Varki, Glycobiology 1993, 3, 97Ϫ130. Ϫ [1b] R. A. Dwek,
Chem. Rev. 1996, 96, 683Ϫ720.
manOCH2CH2CH2OCH2), 3.68 (mc, 3 H, 3 ϫ manOCHH), 3.92
(mc, 3 H, 3 ϫ manOCHH), 4.44 (ddd ഠ m, 3 H, 3 ϫ 5-H), 4.48
(dd, J ϭ 4.1, 12.2 Hz, 3 H, 3 ϫ 6-HЈ), 4.70 (dd, J ϭ 2.0, 12.2 Hz,
3 H, 3 ϫ 6-H), 5.09 (d, J ϭ 1.5 Hz, 3 H, 3 ϫ 1-H), 5.70 (dd, J ϭ
1.5, 3.6 Hz, 3 H, 3 ϫ 2-H), 5.92 (dd, J ϭ 3.6, 10.2 Hz, 3 H, 3 ϫ
3-H), 6.13 (dd ഠ t, J ϭ 10.2 Hz, 3 H, 3 ϫ 4-H), 7.22Ϫ7.60 (m, 36
H, aryl-H), 7.83, 7.94, 8.04, 8.10 (each mഠd, each 6 H, o-aryl-H).
[2]
[2b]
[2a] H. Lis, N. Sharon, Chem. Rev. 1998, 98, 637Ϫ674. Ϫ
E.
E. Simanek, G. J. McGarvey, J. A. Jablonowski, C.-H. Wong,
Chem. Rev. 1998, 98, 833Ϫ862.
W. I. Weis, K. Drickamer, Annu. Rev. Biochem. 1996, 65,
[3]
[4]
441Ϫ473.
P. Sears, C.-H. Wong, Chem. Commun. 1998, 1161Ϫ1170.
[5] [5a]
D. Zanini, R. Roy in Carbohydrate Mimics (Ed.: Y. Chapl-
eur), Wiley-VCH, Weinheim 1998, p. 385Ϫ415. Ϫ [5b] Y. C. Lee,
Ϫ
13C NMR (125.76 MHz, CDCl3): δ ϭ 17.9 (CH3, CH3), 30.1
[5c]
R. T. Lee, Methods Enzymol. 1987, 138, 424Ϫ429. Ϫ
Th.
(CH2, OCH2CH2CH2O), 41.4 (C, CCH3), 63.3 (CH2, C-6), 66.2
(CH2, manOCH2CH2CH2OCH2), 67.4 (CH, C-4), 68.3 (CH2, man-
OCH2CH2CH2OCH2), 69.2 (CH, C-5), 70.6 (CH, C-3), 71.0 (CH,
C-2), 74.1 [CH2, (ROCH2)3CCH3], 98.1 (CH, C-1), 128.7, 128.8,
128.9, 129.4, 129.5, 129.8, 130.1, 130.2, 133.3, 133.4, 133.5, 133.8
(CH, aryl-C), 165.9, 166.5 (C, 12 CϭO). Ϫ MALDI-TOF-MS;
C116H108O33 [M ϩ Na]ϩ: calcd. 2028.68; found 2051.70.
K. Lindhorst, Nachr. Chem. Techn. Lab. 1996, 44, 1073Ϫ1079.
H. Conell, W. Agace, P. Klemm, M. Schembri, S. Marlid, C.
[6]
[7]
˚
Svanborg, Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 9827Ϫ9832.
I. Ofek, R. J. Doyle (Eds.), Bacterial Adhesion to Cells and Tis-
sues, Chapman & Hall, New York, 1994.
[8] [8a]
Th. K. Lindhorst, C. Kieburg, U. Krallmann-Wenzel, Gly-
[8b]
coconjugate J. 1998, 15, 605Ϫ613. Ϫ
B. König, T. Fricke,
A. Waßmann, U. Krallmann-Wenzel, Th. K. Lindhorst, Tetra-
[8c]
hedron Lett. 1998, 39, 2307Ϫ2310. Ϫ
S. Kötter, U.
Krallmann-Wenzel, S. Ehlers, Th. K. Lindhorst, J. Chem. Soc.,
Perkin Trans. 1 1998, 2193Ϫ2200.
Unprotected Trivalent Cluster Mannoside (21): The benzoylated
cluster 20 (333 mg, 0.164 mmol) was dissolved in dry MeOH
(100 mL), sodium methoxide (1  in MeOH, 1 mL) was added and
the reaction mixture was stirred at room temp. for 48 h. The mix-
ture was neutralized with ion exchange resin (Amberlite IR 120
Hϩ), filtered and the filtrate concentrated in vacuo. The residue
was purified by size exclusion chromatography on Sephadex LH-20
(elution with MeOH) to yield the unprotected title cluster (128 mg,
0.164 mmol, quant.) as a white amorphous solid. Ϫ [α]2D3 ϭ ϩ36.1
(c ϭ 2.24 in MeOH). Ϫ 1H NMR (400 MHz, CDCl3): δ ϭ 0.96 (s,
3 H, CH3), 1.87 (mc, 6 H, 3 ϫ OCH2CH2CH2O), 3.31 [br. s, 6 H,
(ROCH2)3CCH3], 3.50Ϫ3.59 (m, 12 H, 3 ϫ 5-H, 3 ϫ manOCHH,
3 ϫ manOCH2CH2ϪCH2OCH2), 3.65Ϫ3.89 (m, 18 H, 3 ϫ 3-H, 3
ϫ 2-H, 3 ϫ 4-H, 3 ϫ 6-H, 3 ϫ 6-HЈ, 3 ϫ manOCHH), 4.80 (d ഠ
s, 3 H, 3 ϫ 1-H). Ϫ 13C NMR (125.76 MHz, CDCl3): δ ϭ 19.2
(CH3, CH3), 32.0 (CH2, OCH2CH2CH2O), 43.1 (Cq, CCH3), 63.9
(CH2, C-6), 66.7 (CH2, manOCH2CH2CH2OCH2), 69.7 (CH, C-
4), 70.3 (CH2, manOCH2CH2CH2OCH2), 73.3 (CH, C-2), 73.8
(CH, C-3), 75.6 (CH, C-5), 75.7 [CH2, (ROCH2)3CCH3], 102.7
[9]
N. Sharon, FEBS Lett. 1987, 217, 145Ϫ157.
[10] [10a]
G. R. Newkome, X. Lin, Macromolecules 1991, 24,
1443Ϫ1444. Ϫ [10b] G. R. Newkome, J. K. Young, G. R. Baker,
R. L. Potter, L. Audoly, D. Cooper, C. D. Weis, Macromolec-
ules 1993, 26, 2394Ϫ2396.
[11]
S. Hanessian, C. Hoornaert, Carbohydr. Res. 1985, 137,
C14ϪC16.
[12]
[13]
W. H. Binder, W. Schmid, Monatsh. Chem. 1995, 126, 923Ϫ931.
S. Hanessian, D. Qiu, H. Prabhanjan, G. V. Reddy, B. Lou,
Can. J. Chem. 1996, 74, 1738Ϫ1747.
H. C. Hansen, S. Haatja, J. Finne, G. Magnusson, J. Am.
Chem. Soc. 1997, 119, 6974Ϫ6979.
S. Hanessian, H. Prabhanjan, Synlett 1994, 868Ϫ870.
J. Lehmann, M. Scheuring, Carbohydr. Res. 1995, 276, 57Ϫ74.
J. Lehmann, U. P. Weitzel, Carbohydr. Res. 1996, 294, 65Ϫ94.
S. Jegge, J. Lehmann, Carbohydr. Res. 1984, 133, 247Ϫ254.
R. M. Nougier, M. Mchich, J. Org. Chem. 1985, 50,
3296Ϫ3298.
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
F. Bien, T. Ziegler, Tetrahedron: Asymmmetry 1998, 9,
781Ϫ790.
P. Klemm, B. J. Jørgensen, I. van Die, H. de Ree, H. Bergmans,
Mol. Gen. Genet. 1985, 199, 410Ϫ414.
(CH, C-1).
Ϫ MALDI-TOF-MS; C32H60O21 [M ϩ
Na]ϩ:
[22] [22a]
calcd.780.36; found 803.43 .
Y. C. Lee, R. T. Lee, Acc. Chem. Res. 1995, 28, 321Ϫ327.
M. Mammen, S.-K. Choi, G. M. Whitesides, Angew.
[22b]
Ϫ
Chem. 1998, 110, 2908Ϫ2953; Angew. Chem. Int. Ed. 1998,
Acknowledgments
Financial support for this work by the Deutsche Forschungs-
gemeinschaft (DFG) in the framework of the collaborative network
37, 2754Ϫ2794.
[23]
E. Stahl, U. Kaltenbach, J. Chromatogr. 1961, 5, 351Ϫ355.
Received October 26, 1999
[O99602]
2034
Eur. J. Org. Chem. 2000, 2027Ϫ2034