Effect of p-substitution of aryl α-D-mannosides on inhibiting mannose-sensitive adhesion of Escherichia coli - Syntheses and testing

Article abstract of DOI:10.1002/(SICI)1099-0690(199808)1998:8<1669::AID-EJOC1669>3.0.CO;2-Q

A series of p-substituted aryl α-D-mannosides was synthesized and tested with regard to their inhibitory capacity in the hemagglutination of guinea pig erythrocytes by type 1 fimbriated Escherichia coli. Synthesis of the bivalent, phenyl-linked mannoside cluster 10 was complicated by orthoester formation during the glycosylation reaction. Surprisingly, the inhibitory potency of 10 was lower than that of p-nitrophenyl α-D-mannoside (1) and this is an important finding for the further design of clustered ligands.

Full text of DOI:10.1002/(SICI)1099-0690(199808)1998:8<1669::AID-EJOC1669>3.0.CO;2-Q

FULL PAPER
Effect of p-Substitution of Aryl ␣--Mannosides on Inhibiting
Mannose-Sensitive Adhesion of Escherichia coli – Syntheses and Testing
ˇ´
´
ˇ
Thisbe K. Lindhorst*a, Sven Köttera, Jirı Kubischb, Ulrike Krallmann-Wenzelc, Stefan Ehlersc, and Vladimır Krenb
Institut für Organische Chemie der Universität Hamburga,
Martin-Luther-King-Platz 6, D-20146 Hamburg, Germany
Institute of Microbiology, Academy of Sciences of the Czech Republic, Laboratory of Biotransformationb,
´
Videnska 1083, CZ-14220 Prague 4, Czech Republic
Forschungszentrum Borstel, Abteilung für Molekulare Infektionsbiologiec,
Parkallee 22, D-23845 Borstel, Germany
Received March 12, 1998
Keywords: Glycosides / Clusters / Noncovalent interactions / Ligands / Receptors
A
series of p-substituted aryl α-D-mannosides was complicated by orthoester formation during the glycosylation
synthesized and tested with regard to their inhibitory
capacity in the hemagglutination of guinea pig erythrocytes
by type 1 fimbriated Escherichia coli. Synthesis of the
bivalent, phenyl-linked mannoside cluster 10 was
reaction. Surprisingly, the inhibitory potency of 10 was lower
than that of p-nitrophenyl α- -mannoside (1) and this is an
important finding for the further design of clustered ligands.
D
Microbial adhesion to host cells, occurring prior to infec- aryl α--mannosides on the binding potencies of the re-
tion is often dependent on the recognition of cell surface spective mannosides towards type 1 fimbriae of E. coli HB
carbohydrates by microbial lectins[1]. In order to therapeuti- 101 (pPKL4). To enable structural comparison to 1, an-
cally manipulate these interactions, it is crucial to under- other p-R-benzyl α--mannosides became our target com-
stand the detailed specificities of the carbohydrate recog- pounds (Figure 1).
nition domains on microbial lectins.
Escherichia coli[2] is a commensal of the intestinal tract
of mammals and has been recognized as an important
Figure 1. Schematic presentation of α-mannoside binding to type
1 fimbriae: The mannose moiety is bound at the carbohydrate reco-
gnition domain (CRD); the hydrophobic part of the aromatic man-
nosides can improve its binding potencies by interacting with a hy-
pathogen involved in a variety of intestinal and extraintesti-
nal diseases. E. coli are equipped with long filamentous ap-
pendages on their surfaces, called fimbriae or pili. Fimbriae
carry lectin-like domains which enable them to bind to
specific sugar epitopes on the surface of potential host
cells[3]. Mannose-specific pili are called type 1 fimbriae and
belong to the important virulence factors of E. coli[4]. Test-
ing of their carbohydrate specificity has led to the following
key results[5]. (1) The agglutinin on type 1 fimbriae is
specific for α--mannosides, (2) several mannose-containing
trisaccharides show enhanced binding potency and (3) aro-
matic α-mannosides are especially powerful inhibitors of
hemagglutination by E. coli. These findings were integrated
in a model about the carbohydrate recognition domain on
type 1 fimbriae[6] in which the type 1 fimbrial lectin corre-
sponds to the approximate size of a trisaccharide and an
drophobic binding region, present inside the CRD or close to it; the
contribution of different p-substituents R was tested in inhibition
hemagglutination tests
Our previous studies on the specificity of type 1 fimbri-
additional hydrophobic binding region is present, inside or ated E. coli, using multivalent mannosyl clusters[8] have
close to the actual carbohydrate recognition domain. p- shown that divalent and trivalent clusters possessed
Nitrophenyl α--mannoside (1) became the prototype of strongly enhanced binding potencies towards type 1 fimbri-
potent aromatic mannoside ligands for type 1 fimbriae as ated E. coli. Thus, we were especially interested in the syn-
its potency was usually not surpassed by the various oligo- thesis and testing of 1,4-bis(α--mannopyranosyloxyme-
saccharides tested. Studies on the modification of the sub- thyl)benzene (10), representing a divalent mannoside clus-
stituents at the phenyl moiety revealed different binding po- ter. It could be expected to be an especially good inhibitor
tencies in several cases[7]. This prompted us to further inves- of E. coli adhesion, due to the combined effects of clustered
tigate the effect of different substituents in p-substituted mannose moieties and hydrophobicity of the aromatic
Eur. J. Org. Chem. 1998, 1669Ϫ1674
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1434Ϫ193X/98/0808Ϫ1669 $ 17.50ϩ.50/0
1669

T. K. Lindhorst
FULL PAPER
moiety. Syntheses and biological testing of the compounds
listed in Figure 1 are presented in this paper.
Complete saponification of the acetylated methyl ester 13
gave the free acid 9. Diagnostic H-NMR chemical shifts
1
p-Nitrobenzyl α--mannopyranoside (5) was obtained by of the m-aryl protons in the p-substituted aryl mannosides
mannosylation of p-nitrobenzyl alcohol under Lewis acid 5Ϫ10 are listed in Table 1.
catalysis with the acetylated glycosyl trichloroacetimidate
Mannosylation of the diol 4-(hydroxymethyl)benzyl al-
11 as mannosyl donor and subsequent deacetylation of the cohol to obtain the bisglycoside 10, unexpectedly did not
resulting mannoside 12 in 86% overall yield (Scheme 1).
occur under conditions, which were effective for the syn-
Scheme 1
Chemoselective reduction of the nitro group in 5 to ob-
tain 6 had to be performed without cleavage of the benzyl
aglycon. This was achieved by Raney nickel catalyzed
1
Table 1. H-NMR chemical shifts of the m-aryl protons in the p-
substituted aryl mannosides 5Ϫ10
p-Substituents
1H-NMR chemical shifts of the
hydrogenation at short reaction times in yields around 70%.
´
m-aryl protons
Complete acetylation of 6, followed by Zemplen deprotec-
tion of the hydroxy groups led to the p-acetamidobenzyl
mannoside (7) in 94% overall yield. In analogy to the syn-
thesis of mannoside 12, glycosylation of methyl 4-(hydroxy-
methyl)benzoate with the acetylated mannosyl trichloroace-
timidate 11 gave access to the methoxycarbonyl-substituted
benzyl mannoside 13, which was carefully deacetylated by
methanolic NaOCH3 solution to give the target mannoside
NO2 (5)
NH2 (6)
8.25
7.15
7.31
8.04
7.83
7.28
NHAc (7)
CO2CH3 (8)
CO2H (9)
α-mannosyloxymethyl (10)
8 while leaving the p-methoxycarbonyl substituent un- thesis of 12 and 13. Glycosylation using the Koenigs-Knorr
changed (Scheme 2).
donor 2,3,4,6-tetra-O-acetyl-α--mannopyranosyl bromide
Scheme 2
1670
Eur. J. Org. Chem. 1998, 1669Ϫ1674

p-Substitution of Aryl α-D-Mannosides on Inhibiting Mannose-Sensitive Adhesion of Escherichia coli
FULL PAPER
(14) and different silver salts as promotor led to complex In this assay, the inhibition titre denotes the lowest concen-
mixtures of which only the 1,2-orthoacetate 15 could be tration of the inhibitor at which agglutination of guinea pig
isolated in its pure form. Typically, the signal of the anom- erythrocytes by E. coli is no longer visible by macroscopic
eric proton of the glycosyl 1,2-orthoester 15 is shifted inspection. As expected, the known high affinity inhibitor
downfield (δ ϭ 5.43) compared to the corresponding glyco- of type 1 fimbriae-mediated hemagglutination, p-nitrophe-
side, whereas the signals of 2-H and 5-H are shiftet upfield nyl α--mannoside (1) had an IT of 42 µ, 80-fold lower
(δ ϭ 4.49 and 3.61, respectively). Monosubstitution was re- when compared to methyl α--mannoside. The ITs of all
1
flected by the unsymmetry of the H-NMR spectrum re- other tested phenyl and benzyl mannosides are in a similar
garding the signals for the acceptor diol and by the inte- range and are shown in Figure 2.
gration ratios of the sugar ring protons and those of the
benzyl group. To further characterize the orthoester, it was
deprotected with triethylamine to give 16, which displayed
the conserved orthoester methyl group at δ ϭ 1.76. Unfor-
tunately, all our attempts (using acidic conditions, TMSOTf
catalysis or mercury salts) to isomerize the orthoacetate 15
to the corresponding glycoside failed. With benzyl-pro-
tected glycosyl donors the formation of glycosyl orthoesters
naturally is excluded. However, this approach was not suit-
able for the synthesis of the “benzyl-coupled” bisglycoside
10 as the final debenzylating deprotection step would also
cleave the glycosidic bonds. Finally, the divalent mannoside
17 was obtained in yields around 50% after employing the
acetylated donor 11 in acetonitrile at 0°C or at lower tem-
perature with the repeated addition of TMSOTf and long
reaction times.. Deacetylation of 17 was easily performed
to yield the unprotected target compound 10, which could
be unequivocally characterized by NMR spectroscopy.
Figure 2. Logarithmic presentation of the micromolar inhibition
titres (IT) of all tested compounds, as determined by inhibition
hemagglutination tests (MeMan: methyl α--mannopyranoside)
Scheme 3
The p-substituted phenyl mannosides 1Ϫ3 and benzyl
No substantial differences were observed. All tested com-
mannosides 4Ϫ10 were tested as inhibitors of mannose- pounds showed inhibitory concentrations 30- to 120-fold
specific adhesion in an inhibition hemagglutination test[9]
.
lower than methyl α--mannoside. The relative inhibition
Eur. J. Org. Chem. 1998, 1669Ϫ1674
1671

T. K. Lindhorst
FULL PAPER
determined with a Leitz heating stage microscope and are uncor-
titres (RIT) compared to 1 as the standard are shown in
Figure 3.
rected. Ϫ NMR: NMR spectra were recorded with a Bruker AM
1
400 (400 MHz for H NMR and 100.6 MHz for 13C NMR) or a
DRX 500 (500 MHz for 1H NMR and 125.84 MHz for 13C NMR);
chemical shifts are in ppm; NMR spectra were calibrated relative
to internal TMS (δ ϭ 0.00 for 1H and 13C NMR) or solvent peaks
{CDCl3: δ ϭ 7.26 for 1H NMR and 77.00 for 13C NMR;
[D4]MeOH: δ ϭ 3.35 for 1H NMR and 49.30 for 13C NMR;
[D6]DMSO: δ ϭ 2.49 for 1H NMR and 39.70 for 13C NMR; D2O:
δ ϭ 4.65 for 1H NMR and for 13C NMR [D3]acetonitile (δ ϭ 1.30)
or [D6]acetone (δ ϭ 30.60) was added}. Ϫ MS: Mass spectra were
measured with a VG Analytical 70-250S (FAB MS). Ϫ Bacteria: A
recombinant type 1 fimbriated E. coli strain, E. coli HB 101
(pPKl4)[11], was used and cultured as described elsewhere[8]. Ϫ In-
hibition hemagglutination tests with guinea pig erythrocytes were
performed as described[8][9]. Ϫ Carbohydrates: p-Nitrophenyl α--
mannoside (1) was from SENN chemicals and benzyl α--manno-
side (4) from Sigma. 2 was derived from 1 by catalytic hyrogena-
tion. The mannosyl donor 11 was prepared from -mannose in
three steps as described in the literature[12]. TMSOTf was from
Fluka and used without further purification.
Figure 3. Relative inhibition titres (RIT) of the tested compounds,
based on p-nitrophenyl α--mannoside (1) as standard (ϵ 1)
Considering the semi-quantitative character of the inhi-
bition hemagglutination test[10], four derivatives could be
identified to display relative inhibitory potencies different
from that of 1 (Figure 3). The p-methoxycarbonyl-substi-
tuted derivative 8 was 1.6-fold more effective as inhibitor
than 1 whereas p-nitrobenzyl mannoside 5, p-acetamido-
benzyl mannoside 7 and the acid 9 were found to be ap-
proximately half as potent as 1. The aromatic bismannoside
10 showed an inhibitory binding similar to that of 1. This
p-Nitrobenzyl 2,3,4,6-Tetra-O-acetyl-Ͱ--mannopyranoside (12):
A solution of 245 mg (1.60·10Ϫ3 mol) of 4-nitrobenzyl alcohol and
1.02 g (2.08·10Ϫ3 mol, 1.3 equiv.) of 11 in 20 ml of dry toluene was
twice concentrated with 20 ml of dry toluene and then dissolved in
˚
50 ml of dry acetonitrile. Molecular sieves (100 mg, 4 A) and 100
µl of a TMSOTf solution (0.02  in dichloromethane) were added
and the reaction mixture was stirred at 0°C for 45 min and at room
temp. for 1 h. Additional 200 µl of the TMSOTf solution were
was surprising as the combined effects of clustering and added and it was stirred at room temp. overnight. To quench the
reaction, 250 µl of Et3N was added and the mixture was concen-
trated to dryness. Purification by flash chromatography (light
petroleum ether/ethyl acetate, 2:1) afforded 0.7 g of 12 (90%), white
hydrophobicity were expected to add to better overall bind-
ing. This finding indicated that only one of the two man-
nose moieties was bound to the CRD of the fimbrial lectin.
From the obtained results no definite conclusions can be
drawn about the influence of the p-substituents in aromatic
mannosides on their inhibitory properties in the examined
adhesion system. Only smaller effects were detected, which
can not clearly be correlated to the chemical nature of the
substituents R. The conformation of the bismannoside 10
20
1
amorphous solid, [α]D ϭ ϩ44.5 (c ϭ 1.07, CHCl3). Ϫ H NMR
(400 MHz, CDCl3): δ ϭ 8.24 (d, J ϭ 8.7 Hz, 2 H, m-aryl-H), 7.53
(d, 2 H, o-aryl-H), 5.39 (dd, J2,3 ϭ 3.1 Hz, J3,4 ϭ 9.7 Hz, 1 H, 3-
H), 5.33 (dd, J1,2 ϭ 1.5 Hz, 1 H, 2-H), 5.33 (dd ഠ t, 1 H, 4-H),
4.92 (d, 1 H, 1-H), 4.84 (d, J ϭ 13.2 Hz, 1 H, benzyl-CHH), 4.67
(d, 1 H, benzyl-CHH), 4.29 (dd, J5,6 ϭ 5.6 Hz, J6,6Ј ϭ 12.7 Hz, 1
H, 6-H), 4.11 (dd, J5,6Ј ϭ 2.6 Hz, 1 H, 6Ј-H), 4.00 (ddd, 1 H, 5-
obviously did not allow complete binding to the CRD of H), 2.16, 2.11, 2.05, 2.01 [each s, each 3 H, 4 C(O)CH3]. Ϫ 13C
NMR (100.67 MHz, CDCl3): δ ϭ 170.52, 169.99, 169.90, 169.63
(each s, 4 CϭO), 147.71 (s, p-aryl-Cq), 143.54 (s, i-aryl-Cq), 128.12,
123.83 (each s, aryl-CH), 97.10 (s, C-1), 69.32 (s, C-4), 68.99 (s, C-
5), 68.87 (s, C-3), 68.35 (s, benzyl-CH2), 65.98 (s, C-2), 62.37 (s, C-
6), 20.80, 20.70, 20.64, 20.63 [each s, 4 C(O)CH3]. Ϫ C21H25NO12
(483.42): calcd. C 52.2, H 5.2, N 2.9; found C 52.3, H 5.1, N 2.8.
the fimbrial lectin. This implies that the type 1 fimbrial
carbohydrate binding site can not accomodate a structure
“sugar-aryl-sugar” as represented in 10. This is an import-
ant finding for further design of multivalent mannoside li-
gands for type 1 fimbriae.
Financial support by the Deutscher Akademischer Austausch-
dienst (DAAD), the Deutsche Forschungs-Gemeinschaft (DFG)
(SFB 470) and by the Fonds der Chemischen Industrie is gratefully
acknowledged. Special thanks are due to Prof. Dr. Joachim Thiem
for his kind support of our work.
p-Nitrobenzyl Ͱ--Mannopyranoside (5): To a solution of 424 mg
(0.88·10Ϫ3 mol) of 12 in 50 ml of dry MeOH a catalytic amount
of sodium was added and it was stirred at room temp. until the
deprotection reaction was complete (TLC ethyl acetate/MeOH/
water, 7:2:1). The solution was neutralized with ion exchange resin
(DOWEX-Hϩ), filtered and the filtrate concentrated. Purification
by flash chromatography (ethyl acetate/MeOH, 5:1) gave 263 mg
Experimental Section
General: TLC: Thin layer chromatography was carried out on
Alugram Sil G/UV254 plates (Machery & Nagel) or on Kieselgel
60 F254 plates (Merck). Detection was performed under UV light
20
of 5 (95%), white amorphous solid, [α]D ϭ ϩ71.04 (c ϭ 0.46,
MeOH). Ϫ 1H NMR (400 MHz, [D4]MeOH): δ ϭ 8.25 (d, J ϭ 8.7
Hz, 2 H, m-aryl-H), 7.63 (d, 2 H, o-aryl-H), 4.93Ϫ4.88 (d, 2 H, 1-
or by spraying with 20% ethanolic sulphuric acid or ninhydrin solu- H, benzyl-CHH), 4.69 (d, J ϭ 13.2, 1 H, benzyl-CHH), 3.90 (dd,
tion (2% in water/iPrOH, 1:1) with subsequent heating. Ϫ Flash J1,2 ϭ 1.5 Hz, J2,3 ϭ 3.6 Hz, 1 H, 2-H), 3.86 (dd, J5,6 ϭ 2.0 Hz,
chromatography: Merck silica gel 60 (0.040Ϫ0.063 mm, 230Ϫ400 J6,6Ј ϭ 11.7 Hz, 1 H, 6-H), 3.76 (dd, J3,4 ϭ 9.2 Hz, 1 H, 3-H), 3.73
mesh) was used. Ϫ Optical rotations: Optical rotation values were
(dd, J5,6Ј ϭ 6.1 Hz, 1 H, 6Ј-H), 3.65 (dd ഠ t, J4,5 ϭ 9.7 Hz, 1
obtained with a Perkin-Elmer polarimeter 341 or 243 (Na-D line, H, 4-H), 3.60 (ddd ഠ m, 1 H, 5-H). Ϫ 13C NMR (100.67 MHz,
589 nm, cell length 10 cm). Ϫ Melting points: Melting points were
[D4]methanol): δ ϭ 149.14, 147.21 (each s, aryl-Cq), 129.76, 124.91
Eur. J. Org. Chem. 1998, 1669Ϫ1674
1672

p-Substitution of Aryl α-D-Mannosides on Inhibiting Mannose-Sensitive Adhesion of Escherichia coli
FULL PAPER
(each s, aryl-CH), 101.66 (s, C-1), 75.59 (s, C-5), 73.03 (s, C-3), 8.04 (d,
72.46 (s, C-2), 69.09 (s, 4-H or benzyl-CH2), 69.03 (s, benzyl-CH2 o-aryl-H), 5.39 (dd, J2,3 ϭ 3.5 Hz, J3,4 ϭ 10.1 Hz, 1 H, 3-H), 5.31
or 4-H), 63.38 (s, C-6). Ϫ C13H17NO8 (315.27): calcd. C 49.5, H (dd ഠ t, J4,5 ϭ 10.1 Hz, 1 H, 4-H), 5.31 (dd, J1,2 ϭ 1.6 Hz, 1 H,
J ϭ 8.2 Hz, 2 H, m-aryl-H), 7.42 (d, 2 H,
5.4, N 4.4; found C 49.4, H 5.5, N 4.3.
2-H), 4.90 (d, 1 H, 1-H), 4.77 (d, J ϭ 12.6 Hz, 1 H, benzyl-CHH),
4.63 (d, 1 H, benzyl-CHH), 4.28 (dd, J5,6 ϭ 3.6 Hz, J6,6Ј ϭ 12.3
Hz, 1 H, 6-H), 4.07 (dd, J5,6Ј ϭ 2.5 Hz, 1 H, 6Ј-H), 4.00 (ddd, 1
H, 5-H), 3.93 (s, 3 H, CO2CH3), 2.15, 2.11, 2.04, 2.00 [each s, each
3 H, 4 C(O)CH3]. Ϫ 13C NMR (100.67 MHz, CDCl3): δ ϭ 170.54,
169.95, 169.85, 169.66 (each s, 4 CϭO), 166.68 (s, CO2CH3), 146.04
(s, p-aryl-Cq), 141.30 (s, i-aryl-Cq), 129.85, 127.59 (each s, aryl-CH),
96.98 (s, C-1), 69.44 (s, C-2), 69.09 (benzyl-CH2), 69.01 (s, C-5),
68.82 (s, C-3), 66.06 (s, C-4), 62.36 (s, C-6), 52.11 (s, CO2CH3),
20.78, 20.76, 20.62, 20.60 [each s, 4 C(O)CH3]. Ϫ C23H28O12
(496.46): calcd. C 55.6, H 5.7; found C 55.5, H 5.6.
p-Aminobenzyl Ͱ--Mannopyranoside (6): To a solution of 170
mg (0.54·10Ϫ3 mol) of 5 in 10 ml of dry MeOH 200 mg of T1
Raney nickel (freshly prepared[13] and stored in dry ethanol) was
added and the mixture was stirred under hydrogen for 30 min. The
course of the reaction was carefully controlled by TLC (ethyl ace-
tate/MeOH/water, 7:2:1) and the reduction reaction was stopped
when the starting material was almost consumed and before cleav-
age of the glycosidic bond (mannose formation) occurred. After
filtration, concentration and purification by flash chromtography
(ethyl acetate/MeOH, 8:2) 118 mg of 6 (77%) was obtained, colour-
less syrup, Rf (ethylacetate/MeOH/water, 7:2:1) ϭ 0.61 (5), 0.52 (6),
p-(Methoxycarbonyl)benzyl Ͱ--Mannopyranoside (8): A solu-
tion of 490 mg (0.987·10Ϫ3 mol) of 13 in 20 ml of dry MeOH was
0.26 (mannose), [α]D ϭ ϩ71.49 (c ϭ 0.35, H2O). Ϫ 1H NMR
20
(400 MHz, D2O): δ ϭ 7.15 (d, J ϭ 8.14 Hz, 2 H, m-aryl-H), 6.75 treated with 2 ml of an NaOMe solution (1  in MeOH) and stirred
(d, 2 H, o-aryl-H), 4.85 (d, J1,2 ϭ 1.5 Hz, 1 H, 1-H), 4.55 (d, J ϭ at room temp. for 20 min. Then the reaction mixture was neu-
11.2 Hz, 1 H, benzyl-CHH), 4.36 (d, 1 H, benzyl-CHH) 3.80 (dd,
tralized with ion exchange resin (Levatit SP 1080 Hϩ), the resin
J2,3 ϭ 3.6 Hz, 1 H, 2-H), 3.77 (dd, J5,6 ϭ 2.0 Hz, J6,6Ј ϭ 12.2 Hz, was filtered off and the filtrate was concentrated. The resulting
1 H, 6-H), 3.67 (dd, J3,4 ϭ 9.7 Hz, 1 H, 3-H), 3.64Ϫ3.53 (m, 3 H, crude material was purified by flash chromatography (ethyl acetate/
4-H, 5-H, 6Ј-H). Ϫ 13C NMR (100.67 MHz, D2O): δ ϭ 149.43 (s, MeOH, 10:1) to yield 230 mg of 8 (71%), white amorphous solid,
i-aryl-Cq), 146.78 (s, p-aryl-Cq), 130.65, 116.72 (each s, aryl-CH); [α]D ϭ ϩ80.0 (c ϭ 0.45, MeOH). Ϫ 1H NMR (400 MHz,
20
99.40 (s, C-1); 75.59 (s, C-5); 73.21, 70.97, 70.47, 67.12 (each s, C-
[D4]MeOH): δ ϭ 8.04 (d, J ϭ 8.14 Hz, 2 H, m-aryl-H), 7.52 (d,
2, C-3, C-4, C-5), 69.68 (benzyl-CH2), 61.24 (C-6). Ϫ No elemental J ϭ 8.65 Hz, 2 H, o-aryl-H), 4.90 (d, J1,2 ϭ 1.5 Hz, 1 H, 1-H), 4.87
analysis was obtained.
(d, J ϭ 12.71 Hz, 1 H, benzyl-CHH), 4.64 (d, 1 H, benzyl-CHH),
3.94 (s, 3 H, OCH3), 3.91 (dd, J2,3 ϭ 3.6 Hz, 1 H, 2-H), 3.88 (dd,
J5,6 ϭ 2.0 Hz, J6,6Ј ϭ 11.7 Hz, 1 H, 6-H), 3.79 (dd, J2,3 ϭ 3.6 Hz,
J3,4 ϭ 9.2 Hz, 1 H, 3-H), 3.75 (dd, J5,6Ј ϭ 5.6 Hz, 1 H, 6Ј-H), 3.67
(dd ഠ t, J4,5 ϭ 9.7 Hz, 1 H, 4-H), 3.62 (ddd, 1 H, 5-H). Ϫ 13C
NMR (100.67 MHz, CDCl3): δ ϭ 168.67 (s, CO2CH3), 145.03 (s,
aryl-Cq), 130.91 (s, m-aryl-C), 128.99 (s, o-aryl-C), 101.35 (s, C-1),
75.35 (s, C-5), 72.94 (s, C-3), 72.42 (s, C-2), 69.51 (s, benzyl-CH2),
68.95 (s, C-4), 63.26 (s, C-6), 52.89 (s, OCH3). Ϫ C15H20O8
(328.31): calcd. C 54.9, H 6.1; found C 55.0, H 5.9.
p-Acetamidobenzyl Ͱ--Mannopyranoside (7): A solution of 100
mg (0.35·10Ϫ3 mol) of 6 in 5 ml of pyridine was treated with 2 ml
of acetic anhydride and stirred at room temp. until the reaction was
complete. The reaction mixture was repeatedly coconcentrated with
toluene and the remaining syrup was dissolved in 10 ml of dry
MeOH and treated with 100 µl of an NaOMe solution (1  in
MeOH) until the deprotection reation was complete (TLC: ethyl
acetate/MeOH/water, 7:2:1). The solution was neutralized with ion
exchange resin (DOWEX-Hϩ), filtered and the filtrate concentrated
and purified by flash chromatography (ethyl acetate/MeOH, 8:2)
to yield 105 mg of 7 (92%), colourless syrup, [α]D20 ϭ ϩ89.44 (c ϭ
p-Carboxybenzyl Ͱ--Mannopyranoside (9): A solution of 559
mg (1.13·10Ϫ3 mol) of 13 in 25 ml of MeOH/water (3:1) and 5 ml
0.59, H2O). Ϫ 1H NMR (400 MHz, D2O) δ ϭ 7.31 (d, 2 H, m- of aqueous NaOH (10%) was strirred at room temp. for 3 h. Then
aryl-H), 6.75 (d, 2 H, o-aryl-H), 4.85 (d, J1,2 ϭ 1.5 Hz, 1 H, 1-H), the reaction mixture was neutralized with ion exchange resin (Leva-
4.62 (d, J ϭ 11.7 Hz, 1 H, benzyl-CHH), 4.43 (d, 1 H, benzyl-
CHH) 3.84 (dd, J2,3 ϭ 3.6 Hz, 1 H, 2-H), 3.76 (dd ഠ d, J6,6Ј
tit SP 1080 Hϩ), the resin was filtered off and the filtrate was con-
centrated. The resulting crude material was purified by size ex-
ϭ
12.2 Hz, 1 H, 6-H), 3.69 (dd, J3,4 ϭ 9.7 Hz, 1 H, 3-H), 3.67Ϫ3.57 clusion chromatography on Sephadex G-15 (2 ϫ 90 cm column,
(m, 3 H, 4-H, 5-H, 6Ј-H). Ϫ 13C NMR (100.67 MHz, D2O): δ ϭ distilled H2O as eluent) to yield 333 mg of 9 (94%), white lyophilis-
173.20 (s, NHC(O)CH3), 137.31 (s, i-aryl-Cq), 133.96 (s, p-aryl-Cq),
ate, [α]D ϭ ϩ52.1 (c ϭ 1.13, DMSO). Ϫ 1H NMR (400 MHz,
20
129.71, 122.16 (each s, aryl-CH); 99.69 (s, C-1); 73.30, 70.97, 70.44, [D6]DMSO): δ ϭ 7.83 (d, 2 H, J ϭ 7.63 Hz, m-aryl-H), 7.23 (d, 2
67.09 (each s, C-2, C -3, C -4, C-5), 69.31 (s, benzyl-CH2), 61.23 H, J ϭ 8.12 Hz, o-aryl-H), 4.68 (d, 1 H, J1,2 ϭ 1.5 Hz, 1-H), 4.65
(s, C-6). Ϫ C15H21NO7 (327.33): calcd. C 55.0, H 6.5, N 4.3; found (d, 1 H, J ϭ 12.21 Hz, benzyl-CHH), 4.40 (d, 1 H, benzyl-CHH),
C 55.0, H 6.5, N 4.1.
3.67 (dd, 1 H, J6,6Ј ϭ 12.2 Hz, 6Ј-H),3.64 (dd, 1 H, J2,3 ϭ 3.1 Hz,
2-H), 3.54Ϫ3.41 (m, 3 H, 3-H, 5-H, 6-H), 3.48 (dd ഠ t, 1 H, J4,5 ϭ
9.2 Hz, 4-H). Ϫ 13C NMR (100.67 MHz, [D6]DMSO): δ ϭ 169.81
(s, CO2H), 138.47 (s, aryl-Cq), 129.18 (s, m-aryl-C), 126.74 (s, o-
aryl-C), 99.20 (s, C-1), 74.46, 71.23, 67.29 (each s, C-3, C-4, C-5),
70.51 (s, C-2), 67.63 (s, benzyl-CH2), 61.50 (s, C-6). Ϫ No elemental
analysis was obtained.
p-(Methoxycarbonyl)benzyl 2,3,4,6-Tetra-O-acetyl-Ͱ--manno-
pyranoside (13): A solution of 3.54 g (7.2·10Ϫ3 mol) of 11 and 929
mg (5.52·10Ϫ3 mol) of methyl 4-(hydroxymethyl)benzoate was twice
concentrated with 20 ml of dry toluene. The remaining mixture was
dissolved in 50 ml of dry dichloromethane, 2.5 ml of a TMSOTf
solution (0.02  in dry dichloromethane) was added and the reac-
tion mixture was stirred at room temp. for 3 h. Then, 1 additional
ml of TMSOTf (0.02  in dry dichloromethane) was added and
stirring was continued until the reaction was complete (TLC: light
petroleum ether/ethyl acetate, 1:1). To stop the reaction 300 µl of
Et3N was added and the solution was concentrated to dryness and
3,4,6-Tri-O-acetyl-1,2-O-[p-(hydroxymethyl)benzyloxy-
ethylidene]-β--mannopyranose (15): A mixture of 330 mg (1.2·10Ϫ3
˚
mol) of silver carbonate, 1 g of molecular sieves (3 A) and 80 mg
(0.58·10Ϫ3 mol) of 4-(hydroxymethyl)benzyl alcohol in 20 ml of dry
acetonitrile (20 ml) was stirred at room temp. under nitrogen and
exclusion of light. Then a solution of 576 mg (1.40·10Ϫ3 mol) of 14
in dry acetonitrile (20 ml) was added dropwise and the reaction
purified by flash chromatography (light petroleum ether/ethyl acet-
20
ate, 1:1) to yield 2.44 g of 13 (89%), colourless syrup, [α]D
ϭ
ϩ42.6 (c ϭ 1.25, CHCl3). Ϫ 1H NMR (400 MHz, CDCl3): δ ϭ mixture was stirred at room temp. for 20 h. TLC (toluene/ethyl
Eur. J. Org. Chem. 1998, 1669Ϫ1674 1673

T. K. Lindhorst
FULL PAPER
acetate, 1:1) showed at least 4 product spots. The mixture was fil-
(s, C-4), 69.0 (s, benzyl-CH2), 68.9 (s, C-3), 68.5 (s, C-5), 65.9 (s,
tered through Celite, concentrated and purified by flash chromatog- C-2), 62.2 (s, C-6), 20.7, 20.6, 20.49, 20.47 [each s, C(O)CH3]. Ϫ
raphy (toluene/ethyl acetate, 4:1) to yield 120 mg of 15 (Rf ϭ 0.23 FAB-MS; m/z: 799.5 [M ϩ H]ϩ&csdilla; 821.5 [M ϩ Na]ϩ.
in toluene/ethyl acetate, 1:1; 44%), colourless crystals, m.p.
1,4-Bis(Ͱ--mannopyranosyloxymethyl)benzene (10): A solution
of 42 mg (0.053·10Ϫ3 mol) of 17 in 10 ml of dry MeOH (10 ml)
was treated with 150 µl of an NaOMe solution (1  in MeOH) and
stirred at room temp. until the reaction was complete (TLC: ethyl
acetate/MeOH/water, 7:2:1). The mixture was neutralized with ion
exchange resin (Levatit SP 1080 Hϩ, Merck), filtered and concen-
trated in vacuo. The residue was purified by size exclusion chroma-
tography on Sephadex LH-20 (1.5 ϫ 90 cm column, MeOH as
eluent) to yield 22 mg of 10 (90%), colourless syrup, [α]D21 ϭ ϩ99.5
135Ϫ142°C, [α]D ϭ ϩ4.7 (c ϭ 0.55, dichloromethane). Ϫ 1H
22
NMR (400 MHz, CDCl3): δ ϭ 7.28 (s, 4 H, aryl-H), 5.43 (d, J1,2 ϭ
2.5 Hz, 1 H, 1-H), 5.23 (dd ഠ t, J3,4 ϭ 9.7 Hz, J4,5 ϭ 9.4 Hz, 1 H,
4-H), 5.05 (dd, J2,3 ϭ 3.6 Hz, J3,4 ϭ 9.7 Hz, 1 H, 3-H), 4.59 (d, 2
H, benzyl-CH2), 4.54Ϫ4.45 (m, 3 H, benzyl-CH2, 2-H), 4.17 (dd,
J5,6 ϭ 5.1 Hz, J6,6Ј ϭ 12.2 Hz, 1 H, 6-H), 4.08 (dd, J5,6Ј ϭ 2.5 Hz,
1 H, 6Ј-H), 3.61 (ddd, 1 H, 5-H), 2.02, 2.00, 1.98 [each s, each 3
H, 3 C(O)CH3], 1.75 (s, 3 H, orthoester-C(O)CH3).
1
(c ϭ 1.0, MeOH). Ϫ H NMR (400 MHz, [D4]MeOH): δ ϭ 7.28
(s, 4 H, aryl-H), 4.87 (d, J1,2 ϭ 1.5 Hz, 2 H, 2 1-H), 4.79 (d, 2 H,
1,2-O-[p-(Hydroxymethyl)benzyloxyethylidene]-β--manno-
pyranose (16): A mixture of 100 mg (0.21·10Ϫ3 mol) of 15 and 3 ml
of Et3N in 10 ml of MeOH and 0.7 ml of water was stirred for 20
h. Then the mixture was coconcentrated with toluene and purified
by flash chromatography (ethyl acetate/MeOH, 7:3) to yield 45 mg
benzyl-CH2), 4.57 (d, 2 H, benzyl-CH2), 3.89 (dd, J5,6 ϭ 2.0, J6, 6Ј
ϭ
12.2 Hz, 2 H, 2 6-H), 3.86 (dd, J2,3 ϭ 3.6 Hz, 2 H, 2 2-H), 3.76
(dd, J5,6Ј ϭ 5.1 Hz, 2 H, 2 6Ј-H), 3.77 (dd, J2,3 ϭ 3.6, J3,4 ϭ 8.1
Hz, 2 H, 2 3-H), 3.67 (dd ഠ t, J4,5 ϭ 8.1 Hz, 2 H, 2 4-H), 3.63 (m,
2 H, 2 5-H). Ϫ 13C NMR (100.6 MHz, [D4] MeOH): δ ϭ 138.9 (s,
aryl-Cq), 129.5 (s, aryl-CH), 101.0 (s, C-1), 75.2 (s, C-5), 72.9 (s, C-
3), 72.5 (s, C-2), 69.9 (s, benzyl-CH2), 68.9 (s, C-4), 63.2 (s, C-6).
Ϫ FAB-MS; m/z: 485.3 [M ϩ Na]ϩ.
22
of 16 (62%), colourless syrup, [α]D ϭ ϩ27.5 (c ϭ 0.38, dichloro-
1
methane). Ϫ H NMR (500 MHz, CDCl3): δ ϭ 7.28 (s, 4 H, aryl-
H), 5.53 (d, 1 H, J1,2 ϭ 2.5 Hz, 1-H), 4.68Ϫ4.60 (m, 4 H, benzyl-
CH2), 4.58 (dd, 1 H, J2,3 ϭ 3.8 Hz, 2-H), 3.90 (dd, 1 H, J5,6 ϭ 2.5,
J6,6Ј ϭ 12.0 Hz, 6-H), 3.76 (dd, 1 H, J3,4 ϭ 9.5 Hz, 3-H), 3.72 (dd,
1 H, J5,6Ј ϭ 6.3 Hz, 6Ј-H), 3.63 (dd ഠ t, 1 H, J4,5 ϭ 9.5 Hz, 4-H),
3.29 (ddd, 1 H, 5-H), 1.76 (s, 3 H, orthoester-CH3).
[1]
K.-A. Karlsson, Curr. Opin. Struct. Biol. 1995, 5, 622Ϫ635.
[2]
J. Hacker, Curr. Top. Microbiol. Immunol. 1990, 151, 1Ϫ27.
[3]
K. A. Krogfelt, H. Bergmany, P. Klemm, Infect. Immun. 1990,
1,4-Bis(2,3,4,6-tetra-O-acetyl-Ͱ- -mannopyranosyloxy-
methyl)benzene (17): A mixture of 212 mg (0.44·10Ϫ3 mol) of 11
and 20 mg (0.14·10Ϫ3 mol) of 4-(hydroxymethyl)benzyl alcohol was
coconcentrated with 10 ml of dry toluene three times and then
dissolved in 10 ml of dry acetonitrile under nitrogen. The reaction
was started by the dropwise addition of 100 µl of a TMSOTf solu-
tion (0.02  in dry dichloromethane) at 0°C over 10 min. The reac-
tion mixture was stirred at 0°C for 30 min followed by stirring at
room temp. for 30 min. Then additional 100 µl of a TMSOTf solu-
tion (0.02  in dry dichloromethane) was added and the mixture
was stirred for 3 d at room temp. The reaction was stopped by
adding 100 µl of Et3N, it was coconcentrated with toluene and the
residue was purified by flash chromatography (light petroleum
ether/ethyl acetate, 1:1) yielding 61 mg of 17 (53%), colourless
58, 1995Ϫ1998. Ϫ P. Klemm, K. A. Krogfelt, in Fimbriae, Ad-
hesion, Genetics, Biogenesis and Vaccines (Ed.: P. Klemm), CRC
Press, Boca Raton, 1994, p. 9Ϫ26.
˚
[4]
H. Connell, W. Agace, P. Klemm, M. Schembri, S. Marlid, C.
Svanborg, Proc. Natl. Acad. Sci. USA 1996, 93, 9827Ϫ9832. Ϫ
E. V. Sokurenko, V. Chesnokova, R. J. Doyle, D. L. Hasty, J.
Biol. Chem. 1997, 272, 17880Ϫ17886.
[5]
N. Firon, I. Ofek, N. Sharon, Carbohydr. Res. 1983, 120,
235Ϫ249. Ϫ N. Firon, I. Ofek, N. Sharon, Infect. Immun. 1984,
43, 1088Ϫ1090.
[6]
N. Sharon, FEBS Lett. 1987, 217, 145Ϫ157.
[7]
N. Firon, S. Ashkenazi, D. Mirelman, I. Ofek, N. Sharon, In-
fect. Immun. 1987, 55, 472Ϫ476.
[8]
T. K. Lindhorst, C. Kieburg, U. Krallmann-Wenzel, Glycocon-
jugate J., in press.
[9]
E. A. Kabat, M. M. Mayer, Experimental Immunochemistry, re-
viewed and enlarged ed., C. C. Thomas, Springfiled, USA,
22
1
syrup, [α]D ϭ ϩ68.6 (c ϭ 0.48, CHCl3). Ϫ H NMR (400 MHz,
CDCl3): δ ϭ 7.28 (s, 4 H, aryl-H), 5.38 (dd, J2,3 ϭ 3.1 Hz, J3,4
1975.
[10]
The relative inhibitory potencies obtained in independent
ϭ
experiments were highly reproducable, however the obtained
absolute inhibition titres differed up to approximately 10% in
independent inhibition agglutination assays.
9.7 Hz, 2 H, 2 3-H), 5.33Ϫ5.28 (m, 4 H, 2 2-H, 2 4-H), 4.89 (d,
J1,2 ϭ 1.5 Hz, 2 H, 2 1-H), 4.73 (d, 2 H, benzyl-CH2), 4.56 (d, 2
H, benzyl-CH2), 4.29 (dd, J5,6 ϭ 5.1 Hz, J6,6Ј ϭ 12.2 Hz, 2 H, 2 6-
H), 4.08 (dd, J5,6Ј ϭ 2.6 Hz, 2 H, 2 6Ј-H), 4.03 (ddd, J4,5 ϭ 10.2 Hz,
2 H, 2 5-H), 2.15, 2.12, 2.04, 1.99 [each s, each 6 H, 8 C(O)CH3]. Ϫ
13C NMR (62.9 MHz, CDCl3) δ ϭ 170.4, 169.8, 169.7, 169.5 (each
s, CϭO), 136.0 (s, aryl-Cq), 128.1 (s, aryl-CH), 96.5 (s, C-1), 69.3
[11]
P. Klemm, B. J. Jørgensen, I. van Die, H. de Ree, H. Bergmans,
Mol. Gen. Genet. 1985, 199, 410Ϫ414.
[12]
J. Kerekgyarto, J. P. Kamerling, J. B. Bouwstra, J. F. G. Vliegen-
´
thart, A. Liptak, Carbohydr. Res. 1989, 186, 51Ϫ62.
[13]
H. A. Bruson, T. W. Riener, J. Am. Chem. Soc. 1943, 65, 23Ϫ27.
[98101]
1674
Eur. J. Org. Chem. 1998, 1669Ϫ1674