Probing multivalent carbohydrate-lectin interactions by an enzyme-linked lectin assay employing covalently immobilized carbohydrates

Article abstract of DOI:10.1016/j.bmc.2007.08.063

We report here the synthesis of a series of mono- to trivalent N-acetylglucosamine (GlcNAc) derivatives as ligands for the plant lectin wheat germ agglutinin (WGA). Their WGA binding potencies were determined by an established enzyme-linked lectin assay (

Full text of DOI:10.1016/j.bmc.2007.08.063

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry 15 (2007) 7661–7676
Probing multivalent carbohydrate–lectin interactions
by an enzyme-linked lectin assay employing
covalently immobilized carbohydrates
Caroline Maierhofer, Katja Rohmer and Valentin Wittmann*
Universita¨t Konstanz, Fachbereich Chemie, Fach M 709, 78457 Konstanz, Germany
Received 5 June 2007; revised 27 August 2007; accepted 31 August 2007
Available online 6 September 2007
Abstract—We report here the synthesis of a series of mono- to trivalent N-acetylglucosamine (GlcNAc) derivatives as ligands for the
plant lectin wheat germ agglutinin (WGA). Their WGA binding potencies were determined by an established enzyme-linked lectin
assay (ELLA) employing microtiter plates with non-covalently immobilized porcine stomach mucin (PSM) as reference ligand and
an ELLA with a new GlcNAc derivative covalently immobilized via a thiourea linkage. Comparison of both assays revealed that the
type of presentation of GlcNAc residues on the microtiter plates either as part of a glycoprotein or as a covalently immobilized
monosaccharide derivative strongly influences the outcome of the assay. Although the apparent dissociation constants KEDLLA for
the interaction of peroxidase-labeled WGA with the microtiter plates are comparable for both surfaces, IC50 values obtained with
the PSM-free ELLA were substantially lower. Even more strikingly, this ELLA displayed a better differentiation between ligands of
different valency leading to significantly higher relative inhibitory potencies of multivalent ligands compared to monovalent. Addi-
tionally, problems associated with the use of PSM, such as maximum inhibition at considerably less than 100% and poor reproduc-
ibility of IC50 values could be overcome with this type of ELLA.
ꢀ 2007 Elsevier Ltd. All rights reserved.
1. Introduction
tency of a polyvalent inhibitor depends on the spatial
presentation of the carbohydrates and the associated
possibility of occupying several binding sites simulta-
neously26 and is often assessed in terms of a relative
inhibitory potency compared to the monovalent interac-
tion. If the 3D structure of a lectin is known, suitable
multivalent ligands may be obtained by rational de-
sign.27,28 However, if the three-dimensional orientation
of the binding sites of a lectin is not known, many poten-
tial ligands have to be synthesized and screened to deter-
mine the required orientation of the carbohydrate units.
The specific recognition of carbohydrate structures in
biological systems by carbohydrate-binding proteins
(lectins)1–4 is the basis of numerous intra- and intercellu-
lar events ranging from the control of protein folding to
cell–cell communication during development, inflamma-
tion, immune responses, and cancer metastasis.5–9 High
affinity lectin ligands are of great medicinal interest in
the diagnosis and inhibition of such carbohydrate-med-
iated processes.10–15 The generation of high affinity li-
gands, however, is not trivial because the interaction
of individual carbohydrate epitopes with lectins is in
many cases weak and undiscriminating.16 The binding
affinity can be dramatically increased by clustering of
lectin binding sites and carbohydrate recognition
units.17–23 The observation that the binding affinity in-
creases exponentially with the number of binding sites
has been termed the glycoside cluster effect.24,25 The po-
Enzyme-linked lectin assays (ELLA)29 provide a fast
method to evaluate lectin binding properties of soluble
natural or synthetic lectin ligands in solution.30–34
Microtiter plates are coated with a reference ligand (of-
ten a (neo)glycoprotein or polysaccharide) and IC50 val-
ues for inhibition of the binding of an enzyme-labeled
lectin to the immobilized reference ligand (the matrix)
by the soluble ligands to be tested are determined. The
advantage of ELLA compared to solid-phase binding
assays such as surface plasmon resonance (SPR)35 is
that both, lectin and ligand are in solution. However,
the use of glycoproteins from natural sources as refer-
ence ligands is limited because their varying complex
Keywords:
Carbohydrates;
Lectins;
Multivalency;
Assay;
Immobilization.
*
Corresponding author. Tel.: +49 7531 88 4572; fax: +49 7531 88
4573; e-mail: mail@valentin-wittmann.de
0968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.08.063

7662
C. Maierhofer et al. / Bioorg. Med. Chem. 15 (2007) 7661–7676
carbohydrate composition due to microheterogeneity36
leads to reduced reproducibility of IC50 values. Further-
more, incomplete inhibition of lectin binding to the plate
surface is often observed, that is, maximal inhibition oc-
curs at considerably less than 100%. To overcome the
problem of microheterogeneity, covalent immobilization
of defined synthetic carbohydrates and their application
in an enzyme-linked immunosorbent assay (ELISA)
employing PapG adhesin was reported.37
NH2
NH2
H2N
NH2
NH2
NH2
1
2
3
4
H2N
O
O
H2N
H2N
O
H2N
NH2
N
H
O
O
H2N
From screening of a 19440-membered split-and-mix li-
brary of conformationally restricted spatially diverse
glycoclusters, we were able to identify a number of tetra-
to hexavalent ligands of the plant lectin wheat germ
agglutinin (WGA) with affinities up to 620-fold com-
pared to the binding of the monovalent ligand N-acetyl-
glucosamine (GlcNAc).38 WGA is a 36-kDa lectin
composed of two identical glycine- and cysteine-rich
subunits.39 The WGA dimer contains a total of eight
binding sites for GlcNAc and oligomers thereof.40 The
results demonstrated that the binding affinity of the
identified ligands was not only determined by the num-
ber of GlcNAc residues but that ligand architecture is
important as well.
5
6
7
H2N
NH2
N
NH2
8
Figure 1. Amines used for the synthesis of WGA ligands.
Table 1. Coupling of activated carbonate 9 to amines 1–8a
OAc
O
AcO
AcO
O
O
O
+
amine 1–
8
Here we present a series of structurally diverse mono- to
trivalent GlcNAc derivatives which were synthesized to
explore the effects of scaffold structure on the binding
affinity toward WGA. IC50 values were determined by
a published ELLA38,41 and an ELLA employing a newly
developed GlcNAc derivative which was covalently
immobilized to microtiter plates via a thiourea linkage
leading to improved properties regarding sensitivity
and reproducibility. Comparison of both assays revealed
interesting insight into the influence of the type of carbo-
hydrate presentation on the microtiter plates on the re-
sults of the assay.
NHAc
O
9
NO2
EtN(i-Pr)2/CH2Cl2
product 10–17
Amine
Product
Yield (%)
1
2
3
4
5
6
7
8
10
11
12
13
14
15
16
17
85
94
97
96
83
92
94
99
2. Results and discussion
a For structures of products, see Scheme 1.
2.1. Synthesis of WGA ligands
O-deacetylation was conveniently carried out under salt
free conditions using N-ethyldimethylamine in
methanol.
In a previous study we identified several cyclic neoglyco-
peptides containing four to six GlcNAc residues as li-
gands for WGA.38 Interestingly,
a
pentavalent
compound showed a lower IC50 value than a hexavalent,
as determined by an ELLA. This indicated that the
binding affinity is not only determined by the number
of GlcNAc residues but that their spatial presentation
is important as well. In order to further investigate the
effects of the structure of multivalent GlcNAc deriva-
tives on the binding affinity toward WGA, a series of
compounds were prepared. Commercially available
amines 1–8 (Fig. 1) were reacted with activated carbon-
ate 9, which was already used for the preparation of
neoglycopeptides,42 leading to mono- to trivalent glyco-
conjugates 10–17 in high yields (Table 1). Subsequent O-
deacetylation resulted in WGA ligands 18–25 in quanti-
In addition, hydroxybutenyl glycoside 27, obtained by
O-deacetylation of compound 26,43 GlcNAc 28, and al-
lyl 2-acetamido-2-deoxy-b-D-glucopyranoside 29 were
used as monovalent WGA ligands (Scheme 2).
2.2. ELLA employing non-covalently immobilized porcine
stomach mucin as reference ligand
Binding potencies of compounds 18–25 and 27–29 were
determined by an ELLA employing non-covalently
immobilized porcine stomach mucin (PSM) as reference
ligand as described earlier.38,41 PSM-coated microtiter
plates were incubated with mixtures of horseradish per-
oxidase (HRP)-labeled WGA (1 lg mLꢀ1) and synthetic
WGA ligands in varying concentrations. It is worth
mentioning that precipitate formation during incubation
´
tative yields (Scheme 1). When reacted under Zemplen
conditions followed by neutralization using acidic ion
exchange resin, some of the products (especially the
amines 24 and 25) adhered to the resin. Therefore,

C. Maierhofer et al. / Bioorg. Med. Chem. 15 (2007) 7661–7676
7663
OR
O
OR
H
N
O
H
N
RO
RO
RO
RO
O
O
O
O
NHAc
NHAc
O
O
10 R = Ac
11 R = Ac
EtMe2N,
MeOH
NaOMe,
MeOH
18 R = H
19 R = H
O
OR
OR
N
O
O
O
O
H
N
OR
AcHN
RO
RO
H
O
O
OR
OR
NHAc
O
O
12 R = Ac
20 R = H
NaOMe,
MeOH/CHCl3
OR
O
H
N
H
AcHN
O
RO
RO
OR
OR
O
O
N
O
O
NHAc
O
13 R = Ac
21 R = H
EtMe2N,
MeOH
OR
O
OR
H
RO
RO
O
H
N
RO
RO
O
O
O
O
N
O
O
O
O
O
NHAc
NHAc
O
O
O
O
OAc
O
OR
O
H
RO
H
RO
RO
O
O
N
RO
N
O
NHAc
NHAc
O
14 R = Ac
15 R = Ac
23 R = H
EtMe2N,
MeOH
EtMe2N,
MeOH
22 R = H
OR
O
OR
H
N
H
N
AcHN
RO
RO
OR
O
O
O
O
O
OR
N
H
NHAc
O
O
16 R = Ac
EtMe2N,
MeOH
24 R = H
OR
H
AcHN
OR
OR
N
O
O
O
OR
O
H
RO
RO
O
O
N
O
O
N
NHAc
O
OR
17 R = Ac
EtMe2N,
MeOH
AcHN
OR
OR
HN
O
O
25 R = H
O
Scheme 1. Deacetylation of glycoconjugates 10–17.
OH
O
OH
O
OR
O
RO
RO
HO
HO
HO
HO
O
O
OH
OH
AcHN
NHAc
NHAc
26 R = Ac
27 R = H
Scheme 2. Additional monovalent WGA ligands 27–29.
EtMe2N,
MeOH
28
29

7664
C. Maierhofer et al. / Bioorg. Med. Chem. 15 (2007) 7661–7676
under the given experimental conditions was not ob-
served. After incubation, the plates were washed and
remaining labeled WGA bound to the reference ligand
was quantified by a HRP-catalyzed color reaction using
idate the assay, we determined the inhibitory potency of
allyl 2-acetamido-2-deoxy-a-D-glucopyranoside at
a
concentration of 15 mM to be 12% (data not shown).
This corresponds well to the value of 15% published
by Zanini and Roy.41 The monovalent ligands 18, 27,
and 29 are all slightly better inhibitors than GlcNAc
28 with a maximum b value of 4.2 for compound 18.
As expected, a further increase of the inhibitory potency
is observed for the multivalent ligands 20–25. The b val-
ues of divalent derivatives 22–24 are in a range of 13–25
and that of trivalent ligand 25 is 40. Remarkably, the
best inhibitors are the divalent compounds 21 and 20
(b = 50 and 67, respectively). On the first view, this
might indicate the influence of the aromatic spacer giv-
ing 20 and 21 a lower flexibility than 22–24. However,
a closer examination of the inhibition curves obtained
from these ligands containing aromatic spacers reveals
2,20-azinobis(3-ethylbenzothiazoline-6-sulfonic
acid)
diammonium salt (ABTS) as substrate. Dose–response
curves for inhibition of the binding of HRP-labeled
WGA to immobilized PSM are shown in Figure 2. From
these curves the concentrations that reduce the binding
of labeled WGA to PSM by 50% (IC50 values) were
determined as a measure of potency of the synthesized
inhibitors (Table 2, column A). Also shown in Table 2
are relative inhibitory potencies (b values) referenced
to GlcNAc 28 (b = 1).
As shown in Table 2, column A, the IC50 value of Glc-
NAc 28 was determined to be 1000 mM. In order to val-
a
propyl urethane 18
benzyl urethane 19
100
80
60
40
20
0
divalent 22
divalent 24
divalent m-arene 21
-6
-5
-4
-3
-2
-1
0
1
2
3
lg [inhibitor] / m
M
b
GlcNAc 28
allyl alcohol 27
allyl glycoside 29
divalent 23
100
80
60
40
20
0
trivalent 25
divalent p-arene 20
-6
-5
-4
-3
-2
-1
0
1
2
3
lg [inhibitor] / m
M
Figure 2. Dose–response curves for inhibition of the binding of HRP-labeled WGA to PSM-coated microtiter plates by (a) synthetic ligands 18, 19,
21, 22, and 24, (b) 20, 23, 25, and 27–29.

C. Maierhofer et al. / Bioorg. Med. Chem. 15 (2007) 7661–7676
7665
Table 2. Absolute and relative IC50 values of synthetic ligands 18–25 and 27–29 for inhibition of the binding of HRP-labeled WGA to (A) non-
covalently immobilized porcine stomach mucin (PSM) from dose–response curves shown in Figure 2 and (B) covalently immobilized GlcNAc 34
from dose–response curves shown in Figure 4
Compounda
(A) Reference ligand: non-
covalently immobilized PSM
(B) Reference ligand: covalently
immobilized GlcNAc
IC50 (mM)
bb
IC50 (mM)
bb
OH
O
HO
HO
28
27
29
18
19
1000
1
23
1
OH
O
AcHN
OH
O
HO
HO
500
280
240
100
2
12
10
—
—
1.9
2.3
—
OH
NHAc
OH
O
HO
HO
3.6
O
NHAc
H
N
4.2
R
R
N
H
10
—
R
O
O
N
H
23
22
75
60
13
17
0.288
80
—
O
H
N
R
R
O
N
H
—
H
N
O
R
H
R
N
R
24
25
21
40
25
20
25
40
50
—
—
N
N
H
H
R
HN
0.055
420
—
R
R
N
R
R
N
H
N
H
N
H
N
H
—
R
NH
20
38
15
—
67
—
0.088
0.016
260
HN R
1440
a
OH
O
HO
O
NHAc
O
R =
HO
O
b b = relative inhibitory potency.

7666
C. Maierhofer et al. / Bioorg. Med. Chem. 15 (2007) 7661–7676
that they are exceptionally steep. One possible explana-
tion for this behavior is increasing aggregation of the li-
gands at higher concentrations, leading to higher
valency and, therefore, better inhibition. In this regard,
it is conspicuous that the inhibition curve of monovalent
derivative 19 containing benzyl groups also displays this
behavior.
immobilization of GlcNAc leading to constructs with
different spacing between the carbohydrate and the sur-
face (Scheme 3). Commercially available amino modi-
fied microtiter plates 30 were treated with activated
carbonate 9 in the presence of Hunig’s base. O-Deacet-
´
¨
ylation under Zemplen conditions gave GlcNAc-modi-
fied microtiter plates 31 containing a short spacer.
Alternatively, microtiter plates 30 were reacted with
1,4-phenylene diisothiocyanate according to a modified
procedure originally described by Guo et al. for the fab-
rication of DNA microarrays50 to generate amine-reac-
tive isothiocyanate surface 32. Addition of the amino
spacer modified GlcNAc derivative 33 gave GlcNAc
surface 34 with a longer spacer.
As shown in Figure 2, the inhibitor concentrations
needed for the maximal inhibition of the binding of
HRP-labeled WGA to immobilized PSM could not be
reached in most cases due to insufficient solubility of
the ligands. In addition, we experienced that the repro-
ducibility of the assay is limited when the experimental
setup is changed, for example, differences in IC50 values
up to one order of magnitude were observed with differ-
ent sources of PSM and microtiter plates. The order of
inhibitory potencies of different ligands, however, is gen-
erally robust to these variables. The disadvantages asso-
ciated with PSM prompted us to develop a new ELLA
employing a surface covalently modified with a defined
GlcNAc derivative, thus avoiding the use of PSM.
For the synthesis of 33, activated carbonate 9 was re-
acted with mono tritylated diamine 35 to give urethane
36 in quantitative yield (Scheme 4). Detritylation using
trifluoroacetic acid (TFA) and triisopropyl silane (TIS)
and subsequent O-deacetylation of 37 yielded 33.
To determine the performance of the surfaces 31 and 34,
microtiter plates prepared under various conditions
were blocked with bovine serum albumin (BSA) and
incubated with a series of differently concentrated
WGA-HRP solutions. After washing, HRP substrate
ABTS was added and the absorption was measured at
405 nm (Fig. 3). This allowed quantification of HRP-la-
beled lectin bound to the plates. To generate surface 31,
the concentration of activated carbonate 9 was varied
between 0.06 and 100 mM. Incubation with WGA-
HRP occurred at concentrations between 0.2 and
25 lg mLꢀ1. As shown in Figure 3a, the amount of
bound WGA-HRP is very low (A405 generally below
2.3. Enzyme-linked lectin assay employing covalently
immobilized GlcNAc as reference ligand
2.3.1. Modification of amino functionalized microtiter
plates. Covalent immobilization of carbohydrates has
recently gained immense interest for the preparation of
carbohydrate microarrays and several methods for the
site-specific, covalent attachment of chemically modified
carbohydrates to appropriately derivatized surfaces
have been published. The field has been extensively re-
viewed.44–49 We compared two methods for chemical
OH
O
HO
HO
NHMe
O
O
O
1. 9, EtN(i-Pr)2, DMSO
NHAc
2. NaOMe, MeOH
NMe
30
31
SCN
EtN(i-Pr)2
DMSO
NCS
OH
SCN
O
H
N
H
N
H
N
HO
HO
S
O
O
O
S
NHAc
N
H
N
3
O
S
N
H
N
33
pH 10.1
32
34
OH
O
H
HO
HO
O
O
N
O
O
NH2
O
NHAc
O
33
Scheme 3. Preparation of covalently modified microtiter plates 31 and 34.

C. Maierhofer et al. / Bioorg. Med. Chem. 15 (2007) 7661–7676
7667
EtN(i-Pr)2, CH2Cl2
H2N
O
O
NHTr
9
+
O
35
quant.
OAc
O
H
N
AcO
AcO
O
O
O
O
NHTr
O
NHAc
O
36
OAc
TFA, TIS
O
H
N
AcO
O
NHAc
O
O
O
NH2
AcO
O
CH2Cl2
quant.
O
37
OH
NaOMe, PrNH2
O
H
N
HO
O
O
O
O
O
NH2
HO
O
MeOH
quant.
NHAc
33
Scheme 4. Synthesis of spacer-modified GlcNAc 33.
0.5) and does not correlate with the concentration of
activated carbonate 9 used for surface preparation.
From these results we conclude that the observed bind-
ing of WGA-HRP is not a matter of a specific interac-
tion with the immobilized GlcNAc residues. This may
originate from the short distance between the plate sur-
face and the GlcNAc residues reducing the accessibility
of the GlcNAc residues to the WGA binding sites.51
and trivalent ligands 23 and 25 on the other hand are
lowered by a much higher factor, resulting in b values
of 80 and 420 versus 13 and 40, respectively. The inhibi-
tion curve of dimer 20 with the aromatic spacer is not
any more as steep as in Figure 2b. The IC50 value of
20 is still lower than that of divalent 23 but is now higher
than the IC50 of trivalent 25. If aggregation caused the
steep increase of the curves shown in Figure 2b, it is
obviously not any more effective at the low concentra-
tion needed for inhibition of WGA binding to the cova-
lently modified plates. Additionally, due to the lower
IC50 values observed with this variant of the ELLA, lim-
ited solubility did not any more prevent reaching maxi-
mal inhibition with all tested ligands.
A completely different behavior was observed with long-
spacer surface 34. A clear relationship between the con-
centration of 33 used for derivatization of 32 and the
amount of WGA-HRP bound to the surface is observed
(Fig. 3b). At concentrations of amine 33 P1 mM, the
reactive surface of the plate is saturated and binding
curves consistent with simple ligand binding are ob-
tained with increasing WGA-HRP concentration. Addi-
tionally, the total amount of WGA-HRP bound to
surface 34 is much higher compared to 31. Therefore,
all following experiments were carried out with microti-
ter plates 34 obtained by treatment of 32 with 1 mM
solutions of GlcNAc derivative 33.
It is well known that the performance of multivalent car-
bohydrate ligands is strongly dependent on the assay
used to evaluate them because different phenomena
may be measured. This does not only apply to absolute
IC50 values but also to observed multivalency effects,
that is, relative IC50 values. Toone et al., for example,
compared a range of competitive and non-competitive
binding assays, including ELLA, hemagglutination inhi-
bition, and isothermal titration microcalorimetry32 and
Pieters et al. studied multivalent galectin ligands by a so-
lid-phase inhibition assay and fluorescence titrations in
solution.52 There are also few reports that show that
the use of ELLA with different matrices can lead to dif-
ferent relative inhibitory potencies.33,53 This phenome-
non, however, is not well understood.
2.3.2. Determination of inhibitory potencies. Inhibition
experiments with ligands 27–29, 20, 23, and 25 were car-
ried out as described above for PSM-coated microtiter
plates at WGA-HRP concentrations of 1 lg mLꢀ1
.
Dose–response curves for inhibition of the binding of
HRP-labeled WGA to covalently immobilized GlcNAc
34 are shown in Figure 4. IC50 values obtained from
these curves are given in Table 2, column B.
A likely explanation for different absolute IC50 values
found with different matrices could be that the lectins
to be investigated have different binding affinities to
the immobilized reference ligands and, therefore, vary-
ing concentrations of inhibitors are necessary for inhibi-
tion of this interaction. To investigate whether this
explanation could count for the lower IC50 values ob-
tained with our PSM-free ELLA, we compared the affin-
ities of WGA-HRP binding to PSM-coated microtiter
plates on one hand and covalently modified plates 34
As can be seen immediately, IC50 values obtained with
covalently modified microtiter plates were substantially
lower than those obtained with PSM-coated plates.
The IC50 value of GlcNAc 28 is 23 mM instead of
1000 mM. Similarly, inhibitory potencies of the other
monovalent compounds investigated are lowered by a
comparable factor of roughly 30–40, leading not only
to the same order of potencies but also to similar relative
inhibitory potencies (b values). The IC50 values of di-

7668
C. Maierhofer et al. / Bioorg. Med. Chem. 15 (2007) 7661–7676
2.5
2.0
1.5
1.0
0.5
0.0
a
[9] = 100 mM
[9] = 10 mM
[9] = 1 mM
[9] = 0,25 mM
[9] = 0,0625 mM
5
0
10
15
20
25
30
[WGA-HRP] / µg mL—1
2.5
2.0
1.5
1.0
0.5
0.0
b
[33] = 10 mM
[33] = 1 mM
[33] = 0.1 mM
[33] = 10 µM
[33] = 1 µM
[33] = 0.1 µM
0
2
4
6
8
10
[WGA-HRP] / µg mL—1
Figure 3. Binding of WGA-HRP to covalently modified microtiter plates 31 containing a short linker (a) and 34 containing a long linker (b) prepared
with varying concentrations of activated carbonate 9 and amine 33, respectively. Assays were carried out under identical conditions.
on the other. It was expected that the lower IC50 values
would result from significantly lower binding affinities of
the lectin to covalently modified plates 34. The apparent
dissociation constant KEDLLA for the interaction of WGA-
HRP with surface 34, which corresponds to the concen-
tration of WGA-HRP necessary to occupy 50% of the
surface binding sites, was readily available from the
binding curves shown in Figure 3b. Data for the PSM-
coated surface were obtained from an analogous exper-
iment. As shown in the semi-logarithmic plot in Figure
ing curves reveals two differences. Whereas the binding
curve for covalently modified plates 34 perfectly fits to
a Langmuir isotherm (Hill coefficient = 1), the binding
curve for PSM-coated plates is much steeper with a Hill
coefficient of 2. At the same time, PSM-coated plates
were estimated to bind roughly twice the amount of
WGA-HRP per unit surface area compared with surface
34. These results underline the importance of the topol-
ogy of carbohydrate presentation on surfaces, such as
cell membranes or carbohydrate microarrays, for the
affinity and specificity of carbohydrate–lectin interac-
tions. At this point, it is worth mentioning that the
IC50 value of GlcNAc determined by the new assay
(23 mM) is in much closer proximity to the KD value
of GlcNAc which has been determined by several groups
5, KDELLA of covalently modified plates 34 (0.26 lg mLꢀ1
)
is even slightly lower than KEDLLA of PSM-coated plates
(0.35 lg mLꢀ1) which corresponds to a higher binding
affinity and does not explain the observed lower IC50
values with surface 34. A closer examination of the bind-

C. Maierhofer et al. / Bioorg. Med. Chem. 15 (2007) 7661–7676
7669
GlcNAc 28
allyl alcohol 27
allyl glycoside 29
divalent 23
100
80
60
40
20
0
trivalent 25
divalent p-arene 20
neoglycopeptide 38
-6
-5
-4
-3
-2
-1
0
1
2
3
lg [inhibitor] / m
M
Figure 4. Dose–response curves for inhibition of the binding of HRP-labeled WGA to covalently modified microtiter plates 34 by synthetic ligands
20, 23, 25, 27–29, and 38.
2.5
2.0
1.5
1.0
KEDLLA
= 0.26
μ
g ml–1
0.5
0.0
= 0.35
μ
g ml–1
KEDLLA
-3
-2
-1
0
1
2
lg [WGA-HRP] / μ
g mL—1
Figure 5. Semi-logarithmic plot of the binding of WGA-HRP to (•) PSM-coated microtiter plates and (o) covalently modified microtiter plates 34
produced with a concentration of GlcNAc derivative 33 of 1 mM. Absolute absorption values are not comparable for both experiments due to
different assay conditions. PSM-coated plates were estimated to bind roughly twice the amount of WGA-HRP per unit surface area compared with
microtiter plates 34.
employing different methods54–58 to be in the range of
approximately 1–5 mM.
is a highly glycosylated glycoprotein existing in many
glycoforms. It is known that WGA also binds to N-acet-
ylneuraminic acid39 and it may be that this interaction is
not inhibited by GlcNAc derivatives. Therefore, we sub-
jected 38 to the ELLA with covalently immobilized Glc-
NAc. As can be seen in Figure 4, the maximal inhibition
of WGA-HRP binding to this surface presenting a
homogeneous carbohydrate epitope is now close to
100%. The IC50 value of 38 is 16 lM corresponding to
a b of 1440 (Table 2). The previously determined IC50
employing PSM-coated microtiter plates was 0.38 mM
(b = 218).38,59
As mentioned in the introduction, a disadvantage of
ELLA is that in many cases maximal inhibition occurs
at considerably less than 100%. When we, for example,
determined the IC50 values of cyclic neoglycopeptides,
such as 38, for inhibition of WGA-HRP binding to
PSM-coated microtiter plates, a maximum inhibition
of approximately 70% was observed.38 The reasons for
this behavior are obscure but might be rooted in the
presence of several different binding sites on PSM which

7670
C. Maierhofer et al. / Bioorg. Med. Chem. 15 (2007) 7661–7676
O
NHAc
O
HO
HO
O
O
NH
O
OH
OH
O
O
OH
O
HN
HN
H
N
AcHN
OH
OH
HN
H
N
O
HO
HO
O
O
O
O
O
O
O
N
H
O
O
NHAc
O
NH
O
N
NH2
H
NH
H
N
O
O
HN
OH
O
HN
Boc
HO
HO
NH
O
O
NHAc
O
38
3. Conclusions
Preparative flash column chromatography (FC) was car-
ried out on Merck silica gel 60 (40–63 lm). H NMR
1
In conclusion, synthetic mono- to tetravalent WGA li-
gands have been examined for their ability to inhibit
binding of WGA-HRP to carbohydrate-coated microti-
ter plates. IC50 values have been determined by an
ELLA with non-covalently immobilized PSM as refer-
ence ligand and an ELLA employing covalently immo-
bilized spacer-separated GlcNAc as reference ligand.
In general, the inhibitory potency of the ligands in-
creased with the number of GlcNAc residues without
a clear preference for a certain spacer. Comparison of
the results of both ELLA, however, indicated that the
type of presentation of GlcNAc residues on the microti-
ter plates either as part of a glycoprotein or as a cova-
lently immobilized monosaccharide derivative strongly
influences the outcome of the assay. Our investigations
revealed two remarkable aspects. Interestingly, although
the apparent dissociation constants KEDLLA for the inter-
action of WGA-HRP with the microtiter plates are com-
parable for both surfaces, IC50 values obtained with the
PSM-free ELLA were substantially lower. Furthermore,
this ELLA displays a better differentiation between li-
gands of different valency leading to significantly higher
relative inhibitory potencies (b values) of multivalent li-
gands. Covalent immobilization of carbohydrate deriva-
tive 33 allows for generation of a chemically defined
carbohydrate presentation. Problems associated with
the use of PSM, such as maximum inhibition at consid-
erably less than 100%, poor reproducibility of IC50 val-
ues, and insufficient solubility of the inhibitors at the
concentrations required for maximum inhibition, could
be overcome with this ELLA.
and 13C NMR spectra were recorded on Bruker AC
250, Bruker AMX 400, and Bruker Avance DRX 600
instruments. 1H chemical shifts are reported in ppm rel-
ative to the signal from residual solvent protons (CDCl3:
7.26; D2O: 4.67 ppm). 13C chemical shifts are referenced
to the solvent signal (CDCl3: 77.0 ppm) or in case of
D2O to internal standard [D3]MeCN (1.32 ppm).
Assignments of proton and carbon signals were carried
out with the aid of DQF-COSY, HMQC, and ROESY
experiments. MALDI mass spectra were obtained on a
Bruker Biflex III instrument with a-cyano-4-hydroxy-
cinnamic acid (CHCA) as the matrix (positive mode).
High resolution nano-ESI mass spectra were recorded
on an Bruker Apex II FT-ICR mass spectrometer
equipped with a 7 T actively shielded cryomagnet and
an Apollo ESI source. Samples were dissolved in
water/MeOH 1:1 (v/v) to a final concentration of
5 pmol lLꢀ1. If required, a solution of 40 mg mLꢀ1
NaI in MeOH was added to the sample solutions in a
ratio of 1:200 (v/v). Elemental analyses were performed
by the microanalytical facility at the Universita¨t Konstanz.
4.2. Synthesis of the WGA ligands
4.2.1. General procedure 1 (GP 1): preparation of the
acetylated WGA ligands. [(Z)-4-(p-Nitrophenoxycarbon-
yloxy)-but-2-en-1-yl]-2-acetamido-3,4,6-tri-O-acetyl-2-
deoxy-b-D-glucopyranoside 942 (1.1 equiv per amino
group) is dissolved in dry CH2Cl2 (3–11 mL per mmol
9). Addition of the amine and EtN(i-Pr)2 (1 equiv base
per amino group) affords a yellow colored solution.
The reaction mixture is stirred at room temperature un-
til the reaction is complete. After evaporation of the sol-
vent, the residue is purified by FC.
4. Experimental
4.1. General methods
4.2.2. General procedure 2 (GP 2): deprotection of the
acetylated WGA ligands. The acetylated WGA ligand is
dissolved in MeOH/EtNMe2 5:1 (20–30 mL per mmol
acetylated compound) and stirred at room temperature
until the deacetylation is complete. Evaporation of the
solvent gives the product in sufficient purity for further
reactions.
Solvents were purified by distillation and dried by nor-
mal procedures. Thin-layer chromatography (TLC)
was performed on Merck silica gel 60 F254 aluminum
sheets. The sheets were visualized by UV light and
immersion in the appropriate stain followed by heating.

C. Maierhofer et al. / Bioorg. Med. Chem. 15 (2007) 7661–7676
7671
4.2.3.
4-(2-Acetamido-3,4,6-tri-O-acetyl-2-deoxy-b-D-
CDCl3 + 5% CD3OD, 20 ꢁC): d = 7.23 (br s, 4H; Har),
5.67–5.66 (m, 4H; Hvinyl), 5.20 (‘t’, J = 9.9 Hz, 2H; H-3),
5.01 (‘t’, J = 9.6 Hz, 2H; H-4), 4.67 (d, J = 8.2 Hz, 2H;
H-1), 4.64–4.63 (m, 2H; Hallyl), 4.58–4.55 (m, 2H; Hallyl),
4.33–4.27 (m, 8H; CH2N, Hallyl), 4.19 (dd, J = 12.2,
4.4 Hz, 2H; H-6a), 4.11–4.01 (m, 2H; H-6b), 3.86–3.83
(m, 2H; H-2), 3.68 (br s, 2H; H-5), 2.11 (s, 6H;
C(O)CH3), 2.06 (s, 6H; C(O)CH3), 1.99 (s, 6H;
C(O)CH3), 1.86 ppm (s, 6H; C(O)CH3); 13C NMR
(150.9 MHz, CDCl3 + 5% CD3OD, 20 ꢁC): d = 171.1,
171,1 171.1, 169.6 (C(O)CH3), 156.7 (OC(O)NH), 137.7
(Car,q), 128.6 (Colefin.), 128.4 (Colefin.), 127.6 (Car), 99.4
(C-1), 73.7 (C-3), 71.5 (C-5), 68.6 (C-4), 64.5 (Callyl),
62.1 (C-6), 60.7 (Callyl), 54.1 (C-2), 44.4 (CH2N), 22.7,
20.6, 20.6, 20.5 ppm (C(O)CH3); elemental analysis calcd
(%) for C46H62N4O22 (1023.00): C, 54.01; H, 6.11; N, 5.48;
found: C, 53.60; H, 5.98; N, 5.45.
glucopyranosyloxy)-(Z)-but-2-enyloxycarbonyl-propyl-
amine (10). According to GP 1, 9 (320 mg, 0.55 mmol),
CH2Cl2 (5 mL), propylamine (41 lL, 0.5 mmol), and
EtN(i-Pr)2 (87 lL, 0.5 mmol) were reacted overnight.
FC (CH2Cl2/MeOH 30:1) gave 10 (213 mg, 85%) as a
1
white solid. Rf = 0.32 (CH2Cl2/MeOH 15:1); H NMR
(600.1 MHz, CDCl3, 20 ꢁC): d = 5.84 (d, J = 8.2 Hz,
1H; NHAc), 5.73–5.66 (m, 2H; Hvinyl), 5.27 (‘t’,
J = 9.8 Hz, 1H; H-3), 5.06 (‘t’, J = 9.6 Hz, 1H; H-4),
4.86 (br s, 1H; NH), 4.74 (d, J = 8.3 Hz, 1H; H-1),
4.63–4.61 (m, 2H; Hallyl), 4.36 (dd, J = 12.9, 5.1 Hz,
1H; Hallyl), 4.30 (dd, J = 12.9, 6.8 Hz, 1H; Hallyl), 4.24
(dd, J = 12.2, 4.6 Hz, 1H; H-6a), 4.14 (dd, J = 12.2,
1.5 Hz, 1H; H-6b), 3.91–3.88 (m, 1H; H-2), 3.72–3.70
(m, 1H; H-5), 3.14–3.13 (m, 2H; CH2CH2CH3), 2.08
(s, 3H; C(O)CH3), 2.02 (s, 3H; C(O)CH3), 2.01 (s, 3H;
C(O)CH3), 1.93 (s, 3H; C(O)CH3), 1.54–1.50 (m, 2H;
CH2CH2CH3),
0.92 ppm
(t,
J = 7.4 Hz,
3H;
4.2.6. N,N0-Bis-(4-(2-acetamido-3,4,6-tri-O-acetyl-2-deoxy-
b-D-glucopyranosyloxy)-(Z)-but-2-enyloxy-carbonyl)-m-xyly-
lenediamine (13). According to GP 1, 9 (256 mg,
0.44 mmol), CH2Cl2 (5 mL), m-xylylenediamine (26 lL,
0.2 mmol), and EtN(i-Pr)2 (70 lL, 0.4 mmol) were re-
acted for 4.5 h. FC (CH2Cl2/MeOH 15:1) gave 13
(197 mg, 96%) as a white solid. Rf = 0.16 (CH2Cl2/MeOH
15:1); H NMR (600.1 MHz, CDCl3, 20 ꢁC): d = 7.31–
7.18 (m, 4H; Har), 6.10 (d, J = 8.5 Hz, 2H; NHAc),
5.74–5.69 (m, 4H; Hvinyl), 5.49 (br s, 2H; NH), 5.24 (‘t’,
J = 9.7 Hz, 2H; H-3), 5.04 (‘t’, J = 9.4 Hz, 2H; H-4),
CH2CH2CH3); 13C NMR (150.9 MHz, CDCl3, 20 ꢁC):
d = 170.8, 170.8, 170.3, 169.4 (C(O)CH3), 156.4
(OC(O)NH), 128.8 (Colefin.), 128.6 (Colefin.), 99.5 (C-1),
72.5 (C-3), 71.8 (C-5), 68.6 (C-4), 64.6 (Callyl), 62.1 (C-
6), 60.4 (Callyl), 54.6 (C-2), 42.8 (CH2CH2CH3), 23.3
(C(O)CH3), 23.2 (CH2CH2CH3), 20.7, 20.7, 20.6
(C(O)CH3), 11.2 ppm (CH2CH2CH3); elemental analy-
sis calcd (%) for C22H34N2O11 (502.51): C, 52.58; H,
6.82; N, 5.57; found: C, 52.62; H, 6.75; N, 5.55.
1
4.2.4. 4-(2-Acetamido-3,4,6-tri-O-acetyl-2-deoxy-b-D-glu-
copyranosyloxy)-(Z)-but-2-enyloxycarbonyl-benzyl-amine
(11). According to GP 1, 9 (291 mg, 0.5 mmol), CH2Cl2
(1.5 mL), benzylamine (60 lL, 0.55 mmol), and EtN(i-
Pr)2 (87 lL, 0.5 mmol) were reacted overnight. FC
(EtOAc) gave 11 (258 mg, 94%) as a slight brown solid.
Rf = 0.27 (EtOAc); 1H NMR (600.1 MHz, CDCl3,
20 ꢁC): d = 7.35–7.28 (m, 5H; Har), 5.82 (d, J = 8.7 Hz,
1H; NHAc), 5.74–5.66 (m, 2H; Hvinyl), 5.27 (‘t’,
J = 10.0 Hz, 1H; H-3), 5.25 (br s, 1H; NH), 5.06 (‘t’,
J = 9.6 Hz, 1H; H-4), 4.73 (d, J = 8.3 Hz, 1H; H-1), 4.70–
4.61 (m, 2H; Hallyl), 4.37–4.29 (m, 4H; Hallyl, CH2N),
4.23 (dd, J = 12.2, 4.7 Hz, 1H; H-6a), 4.13 (dd, J = 12.2,
1.55 Hz, 1H; H-6b), 3.88 (‘q’, J = 9.2 Hz, 1H; H-2), 3.70–
3.69 (m, 1H; H-5), 2.07 (s, 3H; C(O)CH3), 2.02 (s, 3H;
C(O)CH3), 2.01 (s, 3H; C(O)CH3), 1.90 ppm (s, 3H;
C(O)CH3); 13C NMR (150.9 MHz, CDCl3, 20 ꢁC):
d = 170.8, 170.6, 170.3, 169.4 (C(O)CH3), 156.4
(OC(O)NH), 138.3 (Car,q), 128.8 (Colefin.), 128.7 (Colefin.),
128.6, 127.5, 127.4 (Car), 99.4 (C-1), 72.5 (C-3), 71.7 (C-
5), 68.5 (C-4), 64.5 (Callyl), 62.1 (C-6), 60.7 (Callyl), 54.6
(C-2), 45.0 (CH2N), 23.3, 20.7, 20.7, 20.6 ppm
(C(O)CH3); elemental analysis calcd (%) for
C26H34N2O11 (550.55): C, 56.72; H, 6.22; N, 5.09; found:
C, 56.41; H, 6.08; N, 5.11.
4.73–4.59 (m, 6H; Hallyl, H-1), 4.42–4.33 (m, 8H; Hallyl,
CH2N), 4.22 (dd, J = 12.1, 4.0 Hz, 2H; H-6a), 4.14–4.12
(m, 2H; H-6b), 3.87 (‘q’, J = 9.3 Hz, 2H; H-2), 3.68–3.67
(m, 2H; H-5), 2.07 (s, 6H; C(O)CH3), 2.02 (s, 6H;
C(O)CH3), 2.01 (s, 6H; C(O)CH3), 1.90 ppm (s, 6H;
C(O)CH3); 13C NMR (150.9 MHz, CDCl3, 20 ꢁC):
d = 170.9, 170.8, 170. 6, 169.4 (C(O)CH3), 156.5
(OC(O)NH), 138.0 (Car,q), 129.0 (Colefin.), 128.4, 126.5,
126.0 (Car), 99.4 (C-1), 72.5 (C-3), 71.6 (C-5), 68.6 (C-4),
64.5 (Callyl), 62.1 (C-6), 60.7 (Callyl), 54.5 (C-2), 44.8
(CH2N), 23.2, 20.8, 20.7, 20.7 ppm (C(O)CH3); elemental
analysis calcd (%) for C46H62N4O22 (1023.00): C, 54.01;
H, 6.11; N, 5.48; found: C, 53.87; H, 6.29; N, 5.48.
4.2.7. 1,4-Bis-(3-(4-(2-acetamido-3,4,6-tri-O-acetyl-2-deo-
xy-b-D-glucopyranosyloxy)-(Z)-but-2-enyloxy-carbonylami-
no)-propoxy)-butane (14). Corresponding to GP 1, 9
(256 mg, 0.44 mmol), CH2Cl2 (5 mL), 1,4-bis-(3-amino-
propoxy)-butane (43 lL, 0.2 mmol), and EtN(i-Pr)2
(70 lL, 0.4 mmol) were reacted for 3 h. FC (CH2Cl2/
MeOH 15:1) gave 14 (182 mg, 83%) as a white solid.
1
Rf = 0.18 (CH2Cl2/MeOH 15:1); H NMR (600.1 MHz,
CDCl3, 20 ꢁC): d = 6.03 (d, J = 6.9 Hz, 2H; NHAc), 5.68
(br s, 4H; Hvinyl), 5.33 (br s, 2H; NH), 5.27 (‘t’,
J = 9.9 Hz, 2H; H-3), 5.06 (‘t’, J = 9.5 Hz, 2H; H-4),
4.74 (d, J = 8.3 Hz, 2H; H-1), 4.62–4.60 (m, 4H; Hallyl),
4.37 (dd, J = 12.6, 4.0 Hz, 2H; Hallyl), 4.30–4.24 (m, 4H;
H-6a, Hallyl), 4.14 (br d, J = 12.0 Hz, 2H; H-6b), 3.89
(‘q’, J = 9.1Hz, 2H, H-2), 3.72–3.71 (m, 2H; H-5), 3.49
(br s, 4H; NCH2CH2CH2O), 3.43 (br s, 4H; CH2), 3.28–
3.27 (m, 4H; NCH2CH2CH2O), 2.08 (s, 6H; C(O)CH3),
2.02 (s, 6H; C(O)CH3), 2.01 (s, 6H; C(O)CH3), 1.93 (s,
6H; C(O)CH3), 1.77 (br s, 4H; NCH2CH2CH2O),
1.62 ppm (br s, 4H; CH2); 13C NMR (150.9 MHz,
4.2.5.
N,N0-Bis-(4-(2-acetamido-3,4,6-tri-O-acetyl-2-
deoxy-b-D-glucopyranosyloxy)-(Z)-but-2-enyloxycarbon-
yl)-p-xylylenediamine (12). According to GP 1, 9 (256 mg,
0.44 mmol), CH2Cl2 (4 mL), p-xylylenediamine (28 mg,
0.2 mmol) in CH2Cl2 (1 mL), and EtN(i-Pr)2 (70 lL,
0.4 mmol) were reacted for 4.5 h. FC (CH2Cl2/MeOH
15:1) gave 12 (199 mg, 97%) as a white solid. Rf = 0.18
(CH2Cl2/MeOH 15:1); 1H NMR (600.1 MHz,

7672
C. Maierhofer et al. / Bioorg. Med. Chem. 15 (2007) 7661–7676
CDCl3): d = 170.8 170.4, 169.4 (C(O)CH3), 156.4
(OC(O)NH), 128.8 (Colefin.), 128.5 (Colefin.), 99.6 (C-1),
72.6 (C-3), 71.8 (C-5), 70.8 (CH2), 69.2 (NCH2CH2CH2O),
68.6 (C-4), 64.6 (Callyl), 62.1 (C-6), 60.4 (Callyl), 54.6 (C-2),
39.4 (NCH2CH2CH2O), 29.5 (NCH2CH2CH2O), 26.4
(CH2), 23.2, 20.7, 20.7, 20.6 ppm (C(O)CH3); elemental
analysis calcd (%) for C48H74N4O24 (1091.11): C, 52.84;
H, 6.84; N, 5.13; found: C, 52.53; H, 6.81; N, 5.06.
4.2.10. Tris-(2-(4-(2-acetamido-3,4,6-tri-O-acetyl-2-deoxy-
b-D-glucopyranosyloxy)-(Z)-but-2-enyloxycarbonyl)-ami-
noethyl)-amine (17). Corresponding to GP 1, compound
9 (384 mg, 0.66 mmol), CH2Cl2 (5 mL), tris-(2-amino-
ethyl)-amine (30 lL, 0.2 mmol), and EtN(i-Pr)2 (105
lL, 0.6 mmol) were reacted for 3.5 h. FC (CH2Cl2/
MeOH 9:1) gave 17 (292 mg, 99%) as a white solid.
1
Rf = 0.31 (CH2Cl2/MeOH 9:1); H NMR (600.1 MHz,
CDCl3, 20 ꢁC): d = 6.67 (br s, 3H; NHAc), 5.90 (br s,
3H; NH), 5.68 (br s, 6H; Hvinyl), 5.25 (‘t’, J = 9.6 Hz,
3H; H-3), 5.05 (‘t’, J = 9.0 Hz, 3H; H-4), 4.72 (d,
J = 8.1 Hz, 3H; H-1),4.63–4.55 (m, 6H; Hallyl), 4.36–
4.23 (m, 9H; Hallyl, H-6a), 4.14–4.12 (m, 3H; H-6b),
3.96 (‘q’, J = 9.0 Hz, 3H; H-2), 3.75–3.73 (m, 3H; H-
5), 3.21 (br s, 6H; CH2CH2), 2.55 (br s, 6H; CH2CH2),
2.07 (s, 9H; C(O)CH3), 2.00 (s, 9H; C(O)CH3), 2.00 (s,
9H; C(O)CH3), 1.92 ppm (s, 9H; C(O)CH3); 13C NMR
(150.9 MHz, CDCl3, 20 ꢁC): d = 170.9, 170.8, 170.6,
169.4 (C(O)CH3), 156.9 (OC(O)NH), 129.1, 128.3
(Colefin.), 99.8 (C-1), 72.7 (C-3), 71.6 (C-5), 68.7 (C-4),
64.6 (Callyl), 62.1 (C-6), 60.4 (Callyl), 54.2 (C-2), 53.7
(CH2CH2), 38.8 (CH2CH2), 23.2, 20.7, 20.7, 20.6 ppm
(C(O)CH3); elemental analysis calcd (%) for
C63H93N7O33 (1476.44): C, 51.25; H, 6.35; N, 6.64;
found: C, 50.85; H, 6.44; N, 6.67.
4.2.8. N,N0-Bis-(4-(2-acetamido-3,4,6-tri-O-acetyl-2-deoxy-
b-D-glucopyranosyloxy)-(Z)-but-2-enyloxycarbonyl)-4,7,10-
trioxa-1,13-tridecanediamine (15). According to GP 1, 9
(256 mg, 0.44 mmol), CH2Cl2 (5 mL), 4,7,10-trioxa-1,13-
tridecanediamine (44 lL, 0.2 mmol), and EtN(i-Pr)2
(70 lL, 0.4 mmol) were reacted for 3 h. FC (CH2Cl2/MeOH
9:1) gave 15 (204 mg, 92%) as a white solid. Rf = 0.34
(CH2Cl2/MeOH 9:1); 1H NMR (600.1 MHz, CDCl3,
20 ꢁC): d = 6.28 (d, J = 8.6 Hz, 2H; NHAc), 5.67 (m, 4H;
Hvinyl), 5.54 (br s, 2H; NH), 5.26 (‘t’, J = 9.9 Hz, 2H; H-3),
5.06 (‘t’, J = 9.6 Hz, 2H; H-4), 4.73 (d, J = 8.4 Hz, 2H; H-
1), 4.63–4.55 (m, 4H; Hallyl), 4.36 (dd, J = 12.8, 4.0 Hz,
2H; Hallyl), 4.29–4.23 (m, 4H; Hallyl, H-6a), 4.13 (dd,
J = 12.2, 1.4 Hz, 2H; H-6b), 3.91 (‘q’, J = 9.3 Hz, 2H; H-
2), 3.72–3.71 (m, 2H; H-5), 3.63–3.62 (m, 4H; CH2CH2),
3.59–3.58 (m, 4H; CH2CH2), 3.54 (t, J = 5.4 Hz, 4H;
NCH2CH2CH2O),
3.27
(‘q’,
J = 5.8 Hz,
4H;
NCH2CH2CH2O), 2.07 (s, 6H; C(O)CH3), 2.01 (s, 6H;
C(O)CH3), 2.01 (s, 6H; C(O)CH3), 1.93 (s, 6H; C(O)CH3),
1.76 ppm (‘t’, J = 5.9 Hz, 4H; NCH2CH2CH2O); 13C
NMR (150.9 MHz, CDCl3, 20 ꢁC): d = 170.8, 170.6,
169.4,163.2 (C(O)CH3), 156.4 (OC(O)NH), 128.7 (Colefin.),
128.5 (Colefin.), 99.6 (C-1), 72.5 (C-3), 71.7 (C-5), 70.4
(CH2CH2), 70.1 (CH2CH2), 69.6 (NCH2CH2CH2O), 68.6
(C-4), 64.6 (Callyl), 62.1 (C-6), 60.4 (Callyl), 54.5 (C-2), 39.2
(NCH2CH2CH2O), 29.3 (NCH2CH2CH2O), 23.2, 20.7,
20.7, 20.6 ppm (C(O)CH3); elemental analysis calcd (%)
for C48H74N4O25 (1107.11): C, 52.07; H, 6.74; N, 5.06;
found: C, 51.53; H, 6.69; N, 4.94.
4.2.11. 4-(2-Acetamido-2-deoxy-b-D-glucopyranosyloxy)-
(Z)-but-2-enyloxycarbonyl-propylamine (18). Corre-
sponding to GP 2, 10 (190 mg, 378 lmol) was dissolved
in MeOH/EtMe2N 5:1 and reacted for 48 h. Compound
18 was obtained in quantitative yield as a white solid.
Rf = 0.54 (MeCN/H2O 4:1); 1H NMR (600.1 MHz,
D2O, 20 ꢁC): d = 5.73–5.68 (m, 2H; Hvinyl), 4.55 (br s,
2H; Hallyl), 4.47 (d, J = 8.4 Hz, 1H; H-1), 4.33–4.25
(m, 2H; Hallyl), 3.84 (br d, J = 12.3 Hz, 1H; H-6a),
3.68 (br d, J = 10.0 Hz, 1H; H-6b), 3.62 (‘t’,
J = 9.3 Hz, 1H; H-2), 3.46 (‘t’, J = 8.2Hz, 1H; H-3),
3.39–3.37 (m, 2H; H-4, H-5), 3.00 (br s, 2H;
NCH2CH2CH3), 1.96 (s, 3H; C(O)CH3), 1.43–1.39 (m,
2H; NCH2CH2CH3), 0.80 ppm (t, J = 7.2 Hz, 3H;
NCH2CH2CH3); 13C NMR (150.9 MHz, D2O, 20 ꢁC):
d = 175.4 (C(O)CH3), 158.0 (OC(O)NH), 129.8 (Colefin.),
129.4 (Colefin.), 100.8 (C-1), 76.8 (C-5 or C-4), 74.8 (C-3),
70.7 (C-4 or C-3), 65.5 (Callyl) 61.7 (Callyl), 61.6 (C-6),
56.4 (C-2), 43.1 (NCH2CH2CH3), 23.2 (C(O)CH3),
23.0 (NCH2CH2CH3), 11.3 ppm (NCH2CH2CH3);
HRMS (nano ESI, pos. mode, H2O/MeOH) for
C16H28N2O8: [M+H]+ calcd 377.1918; found: 377.1914,
[M+Na]+ calcd 399.1738; found: 399.1737.
4.2.9. 1,7-Bis-(4-(2-acetamido-3,4,6-tri-O-acetyl-2-deoxy-
b-D-glucopyranosyloxy)-(Z)-but-2-enyloxycarbonyl)-dieth-
ylenetriamine (16). According to GP 1, 9 (384 mg,
0.66 mmol), CH2Cl2 (2 mL), diethylenetriamine (32 lL,
0.3 mmol), and EtN(i-Pr)2 (105 lL, 0.6 mmol) were re-
acted overnight. FC (CH2Cl2/MeOH 9:1) gave 16 (280
mg, 94%) as a white solid. Rf = 0.15 (CH2Cl2/MeOH
1
6:1); H NMR (600.1 MHz, CDCl3, 20 ꢁC): d = 6.35 (br
s, 2H; NHAc), 5.67 (br s, 4H; Hvinyl), 5.58 (br s, 2H;
NH), 5.25 (‘t’, J = 9.9 Hz, 2H; H-3), 5.03 (‘t’,
J = 9.5 Hz, 2H; H-4), 4.73 (d, J = 8.2Hz, 2H; H-1),
4.62–4.54 (m, 4H; Hallyl), 4.22 (dd, J = 12.2, 4.3 Hz, 2H;
H-6a), 4.11–4.07 (m, 2H; H-6b), 3.89 (m, 2H; H-2),
3.73–3.72 (m, 2H; H-5), 3.26–3.25 (m, 4H; CH2CH2),
2.75 (br s, 4H; CH2CH2), 2.06 (s, 6H; C(O)CH3), 2.00
(s, 6H; C(O)CH3), 1.99 (s, 6H; C(O)CH3), 1.91 ppm (s,
6H; C(O)CH3); 13C NMR (150.9 MHz, CDCl3, 20 ꢁC):
d = 171.1, 170.7, 170.5, 169.4 (C(O)CH3), 156.6
(OC(O)NH), 128.9, 128.4 (Colefin.), 99.5 (C-1), 72.5 (C-
3), 71.6 (C-5), 68.6 (C-4), 64.5 (Callyl), 62.1 (C-6), 60.3
(Callyl), 54.4 (C-2), 48.4 (CH2CH2), 40.5 (CH2CH2),
23.2, 21.0, 20.7, 20.6 ppm (C(O)CH3); elemental analysis
calcd (%) for C42H63N5O22 (989.97): C, 50.96; H, 6.41; N,
7.07; found: C, 51.11; H, 6.46; N, 6.67.
4.2.12. 4-(2-Acetamido-2-deoxy-b-D-glucopyranosyloxy)-
(Z)-but-2-enyloxycarbonyl-benzylamine (19). 11 (214 mg,
389 lmol) was suspended in dry MeOH (3 mL). After
addition of a 5.4 M methanolic NaOMe solution
(10 lL, 54 lmol) and stirring for 2.5 h, the reaction mix-
ture was neutralized using strong acidic ion exchanger
(DOWEX 50 WX 8, H+ form). Filtration and crystalli-
zation gave 19 (148 mg, 90%) as a slight yellow solid.
Rf = 0.67 (MeCN/H2O 4:1); 1H NMR (600.1 MHz,
D2O, 20 ꢁC): d = 7.35-7.34 (m, 2H; Har), 7.28–7.27 (m,
3H; Har), 5.74–5.68 (m, 2H; Hvinyl), 4.57 (br s, 2H;
Hallyl), 4.46 (d, J = 6.7 Hz, 1H; H-1), 4.30–4.24 (m,
4H; Hallyl, CH2N), 3.83 (br d, J = 12.2 Hz, 1H; H-6a),

C. Maierhofer et al. / Bioorg. Med. Chem. 15 (2007) 7661–7676
7673
3.67–3.60 (m, 2H, H-6b, H-2), 3.46 (‘t’, J = 8.6 Hz, 1H;
H-3), 3.39–3.36 (m, 2H; H-4, H-5), 1.95 ppm (s, 3H;
C(O)CH3); 13C NMR (150.9 MHz, D2O, 20 ꢁC):
d = 175.3 (C(O)CH3), 159.2 (OC(O)NH), 138.0 (Car,q),
129.6 (Colefin.), 129.5 (Colefin.), 128.3, 127.9 (Car), 100.7
(C-1), 76.8 (C-5), 74.8 (C-3), 70.7 (C-4), 65.4 (Callyl),
61.9 (Callyl), 61.5 (C-6), 56.4 (C-2), 44.9 (CH2N),
23.0 ppm (C(O)CH3); HRMS (nano ESI, pos. mode,
H2O/MeOH) for C20H28N2O8: [M+Na]+ calcd
447.1738; found: 447.1743.
(600.1 MHz, D2O, 20 ꢁC): d = 5.73–5.69 (m, 4H;
Hvinyl-H), 4.55 (br s, 4H; Hallyl), 4.48 (d, J = 8.2 Hz,
2H; H-1), 4.31–4.28 (m, 4H; Hallyl), 3.85 (br d,
J = 12.1 Hz, 2H; H-6a), 3.68 (br d, J = 11.5 Hz, 2H;
H-6b), 3.62 (‘t’, 2H; H-2), 3.47–3.45 (m, 10H; CH2,
CH2, H-3), 3.37 (br s, 4H; H-5, H-4), 3.12 (br s, 4H;
CH2), 1.97 (s, 6H; C(O)CH3), 1.69 (br s, 4H; CH2),
1.55 ppm (br s, 4H; CH2); 13C NMR (150.9 MHz,
D2O, 20 ꢁC): d = 175.3 (C(O)CH3), 159.1 (OC(O)NH),
129.8 (Colefin.), 129.5 (Colefin.), 100.8 (C-1), 76.8 (C-4 or
C-5), 74.8 (C-3), 71.2 (CH2), 70.8 (C-5 or C-4), 68.7
(CH2), 65.5 (Callyl), 61.7 (Callyl), 61.6 (C-6), 56.4 (C-2),
38.5 (CH2), 29.7 (CH2), 26.2 (CH2), 23.0 ppm
(C(O)CH3); HRMS (nano ESI, pos. mode, H2O/MeOH)
for C36H62N4O18: [M+H]+ calcd 839.4132; found:
839.4124, [M+Na]+ calcd 861.3951; found: 861.3957.
4.2.13. N,N0-Bis-(4-(2-acetamido-2-deoxy-b-d-glucopyrano-
syloxy)-(Z)-but-2-enyloxycarbonyl)-p-xylylenediamine (20).
Compound 12 (190 mg, 186 lmol) was dissolved in
MeOH/CHCl3 1:1 (5 mL) and a 5.4 M methanolic
NaOMe solution (5 lL, 27 lmol) was added upon which
a white precipitate formed. After stirring for 2.5 h, water
was added until the precipitate disappeared. The reaction
mixture was neutralized using strong acidic ion exchanger
(DOWEX 50 WX 8, H+ form). Filtration and evaporation
gave 20 (142 mg, 99%) as a white solid. Rf = 0.35 (MeCN/
4.2.16. N,N0-Bis-(4-(2-acetamido-2-deoxy-b-D-glucopyr-
anosyloxy)-(Z)-but-2-enyloxycarbonyl)-4,7,10-trioxa-1,13-
tridecanediamine (23). According to GP 2, 15 (144 mg,
130 lmol) was dissolved in MeOH/EtMe2N 5:1 and re-
acted for 5 days. Compound 23 was obtained in quanti-
tative yield as a white solid. Rf = 0.32 (MeCN/H2O 4:1);
1H NMR (600.1 MHz, D2O, 20 ꢁC): d = 5.95–5.92 (m,
4H; Hvinyl), 4.78 (br s, 4H; Hallyl), 4.70 (d, J = 8.4 Hz,
2H; H-1), 4.54–4.50 (m, 4H; Hallyl), 4.07 (br d,
J = 11.9 Hz, 2H; H-6a), 3.90 (br d, J = 12.0 Hz, 2H; H-
6b), 3.86–3.82 (m, 10H; CH2, CH2, H-2), 3.73 (br s,
4H; CH2), 3.69 (br s, 2H; H-3), 3.60 (br s, 4H; H-4, H-
5), 3.35 (br s, 4H; CH2), 2.19 (s, 6H; C(O)CH3),
1.93 ppm (‘t’, J = 5.8 Hz, 4H; CH2); 13C NMR
(150.9 MHz, D2O, 20 ꢁC): d = 175.3 (C(O)CH3), 159.1
(OC(O)NH), 129.8 (Colefin.), 129.5 (Colefin.), 100.8 (C-1),
76.8 (C-5 or C-4), 74.8 (C-3), 70.7 (C-4 or C-5), 70.5
(CH2), 70.2 (CH2), 69.3 (CH2), 65.5 (Callyl), 61.7 (Callyl),
61.6 (C-6), 56.4 (C-2), 38.5 (CH2), 29.6 (CH2), 23.0 ppm
(C(O)CH3); HRMS (nano ESI, pos. mode, H2O/MeOH)
for C36H62N4O19: [M+H]+ calcd 855.4081; found:
855.4080, [M+Na]+ calcd 877.3900; found: 877.3907.
1
H2O 4:1); H NMR (600.1 MHz, D2O, 20 ꢁC): d = 7.21
(br s, 4H; Har), 5.71–5.66 (m, 4H; Hvinyl), 4.55 (br s, 4H;
Hallyl), 4.45 (d, J = 6.7 Hz, 2H; H-1), 4.29–4.25 (m, 4H;
Hallyl), 4.20 (br s, 4H; CH2N), 3.81 (br d, J = 12.2 Hz,
2H; H-6a), 3.65–3.58 (m, 4H; H-6b, H-2), 3.44 (‘t’,
J = 9.0 Hz, 1H; H-3), 3.37–s3.34 (m, 4H; H-4, H-5),
1.93 ppm (s, 6H; C(O)CH3); 13C NMR (150.9 MHz,
D2O, 20 ꢁC): d = 175.4 (C(O)CH3), 159.2 (OC(O)NH),
138.6 (Car,q), 129.7 (Colefin.), 129.5 (Colefin.), 128.2 (Car),
100.7 (C-1), 76.8 (C-5), 74.8 (C-3), 70.7 (C-4), 65.4 (Callyl),
61.9 (Callyl), 61.5 (C-6), 56.4 (C-2), 44.6 (CH2N), 23.0 ppm
(C(O)CH3); HRMS (nano ESI, pos. mode, H2O/MeOH)
for C34H50N4O16: [M+H]+ calcd 771.3295; found:
771.3281, [M+Na]+ calcd 793.3114; found: 793.3115.
4.2.14. N,N0-Bis-(4-(2-acetamido-2-deoxy-b-D-glucopyrano-
syloxy)-(Z)-but-2-enyloxy-carbonyl)-m-xylylenediamine (21).
According to GP 2, 13 (182 mg, 178 lmol) was dissolved in
MeOH/EtMe2N 5:1 and reacted for 9 days. 21 was ob-
tained in quantitative yield as a white solid. Rf = 0.51
4.2.17. 1,7-Bis-(4-(2-acetamido-2-deoxy-b-D-glucopyrano-
syloxy)-(Z)-but-2-enyloxycarbonyl)-diethylenetriamine (24).
Corresponding to GP 2, 16 (270 mg, 274 lmol) was dis-
solved in MeOH/EtMe2N 5:1 and reacted for 4 days.
Compound 24 was obtained in quantitative yield as a
slight yellow solid. Rf = 0.08 (MeCN/H2O 2:1); 1H
NMR (600.1 MHz, D2O, 20 ꢁC): d = 5.72–5.68 (m, 4H;
Hvinyl), 4.56 (br s, 4H; Hallyl), 4.47 (d, J = 8.4 Hz, 2H; H-
1), 4.30–4.26 (m, 4H; Hallyl), 3.84 (br d, J = 12.1 Hz, 2H;
H-6a), 3.67 (dd, J = 11.1, 1.7 Hz, 2H; H-6b), 3.61 (‘t’, J =
= 8.1 Hz, 2H; H-2), 3.46 (br s, 2H; H-3), 3.37 (br s, 4H;
H-4, H-5), 3.18 (br s, 4H; CH2), 2.65 (br s, 4H;
CH2),1.96 ppm (s, 6H; C(O)CH3); 13C NMR (150.9 MHz,
D2O, 20 ꢁC): d = 175.4 (C(O)CH3), 159.3 (OC(O)NH),
129.8 (Colefin.), 129.5 (Colefin.), 100.8 (C-1), 76.8 (C-4), 74.8
(C-3), 70.8 (C-5), 65.5 (Callyl), 61.9 (Callyl), 61.6 (C-6), 56.4
(C-2), 48.3 (CH2), 40.3 (CH2), 23.0 ppm(C(O)CH3);HRMS
(nano ESI, pos. mode, H2O/MeOH) for C30H51N5O16:
[M+H]+ calcd 738.3404; found: 738.3389.
1
(MeCN/H2O 4:1); H NMR (600.1 MHz, D2O, 20 ꢁC):
d = 7.32–7.29 (m, 2H; Har), 7.17–7.15 (m, 2H; Har), 5.73–
5.67 (m, 4H; Hvinyl), 4.56 (br s, 4H; Hallyl), 4.46 (d,
J = 7.6 Hz, 2H; H-1), 4.29–4.24 (m, 4H; Hallyl), 4.22 (br s,
4H; CH2N), 3.81 (br d, J = 12.0 Hz, 2H; H-6a), 3.66–3.59
(m, 4H; H-6b, H-2), 3.45 (‘t’, J = 9.2 Hz, 2H; H-3), 3.38–
3.35 (m, 4H; H-4, H-5), 1.94 ppm (s, 6H; C(O)CH3); 13C
NMR (150.9 MHz, D2O, 20 ꢁC): d = 175.4 (C(O)CH3),
159.2 (OC(O)NH), 140.0 (Car,q), 129.9 (Car), 129.7 (Colefin.),
129.5 (Colefin.), 126.9 (Car), 126.3 (Car), 100.8 (C-1), 76.8 (C-
5), 74.8 (C-3), 70.7 (C-4), 65.5 (Callyl), 61.9 (Callyl), 61.6 (C-
6), 56.4 (C-2), 44.8 (CH2N), 23.0 ppm (C(O)CH3); HRMS
(nano ESI, pos. mode, H2O/MeOH) for C34H50N4O16:
[M+H]+ calcd 771.3295, found: 771.3306, [M+Na]+ calcd
793.3114; found: 793.3125.
4.2.15. 1,4-Bis-(3-(4-(2-acetamido-2-deoxy-b-D-glucopyr-
anosyloxy)-(Z)-but-2-enyloxy-carbonylamino)-propoxy)-
butane (22). According to GP 2, 14 (150 mg, 137 lmol)
was dissolved in MeOH/EtMe2N 5:1 and reacted for 5
days. Compound 22 was obtained in quantitative yield
4.2.18.Tris-(2-(4-(2-acetamido-2-deoxy-b-D-glucopyranosyl-
oxy)-(Z)-but-2-enyloxycarbonyl)-aminoethyl)-amine (25).
Corresponding to GP 2, 17 (293 mg, 198 lmol) was dis-
1
as a white solid. Rf = 0.33 (MeCN/H2O 4:1); H NMR

7674
C. Maierhofer et al. / Bioorg. Med. Chem. 15 (2007) 7661–7676
solved in MeOH/EtMe2N 5:1 and reacted for 7 days. Com-
pound 25 was obtained as a white solid in quantitative yield.
20 ꢁC): d = 170.8, 170.7, 170.5, 169.5 (C(O)CH3), 156.4
(OC(O)NH), 129.0, 128.7, 128.6, 128.1 127.7, 126.2
(Car, Cvinyl), 99.5 (C-1), 72.7 (C-3), 71.6 (C-5), 70.9,
70.5, 70.4, 70.2, 70.0, 69.9, 69.4 (CH2), 69.0 (C-4), 64.5
(Callyl), 62.2 (C-6), 60.3 (Callyl), 54.4 (C-2), 41.1, 39.03
(CH2), 30.4 (CH2CH2CH2), 29.4 (CH2CH2CH2), 23.1,
21.0, 20.9, 20.6 ppm (C(O)CH3); MALDI-MS (pos.
mode, CHCA) for C48H63N3O14: [M+Na]+ calcd
928.4; found: 928.5, [M+K]+ calcd 944.4; found: 944.4.
1
Rf = 0.35 (MeCN/H2O 2:1); H NMR (600.1 MHz, D2O,
20 ꢁC): d = 5.71–5.70 (m, 6H; Hvinyl), 4.57 (br s, 6H; Hallyl),
4.48 (d, J = 8.4 Hz, 3H; H-1), 4.34–4.26 (m, 6H; Hallyl), 3.86
(br d, J = 12.0 Hz, 3H; H-6a), 3.69 (br d, J = 11.8 Hz, 3H;
H-6b), 3.63 (‘t’, J = 9.3 Hz, 3H; H-2), 3.48 (‘t’, J = 8.4 Hz,
3H; H-3), 3.40–3.38 (m, 6H; H-4, H-5), 3.15 (br s, 6H;
CH2), 2.60 (br s, 6H; CH2), 1.98 ppm (s, 9H; C(O)CH3);
13C NMR (150.9 MHz, D2O, 20 ꢁC): d = 175.2
(C(O)CH3), 159.0 (OC(O)NH), 129.6 (Colefin.), 100.8 (C-1),
76.8 (C-5 or C-4), 74.8 (C-3), 70.7 (C-4 or C-5), 65.5 (Callyl),
61.8 (Callyl), 61.6 (C-6), 56.4 (C-2), 53.5 (CH2), 38.9 (CH2),
23.1 ppm (C(O)CH3); HRMS (nano ESI, pos. mode,
H2O/MeOH) for C45H75N7O24: [M+H]+ calcd 1098.4936;
found: 1098.4938, [M+Na]+ calcd 1120.4736; found:
1120.4803.
4.2.21. 4-(2-Acetamido-3,4,6-tri-O-acetyl-2-deoxy-b-d-glu-
copyranosyloxy)-(Z)-but-2-enyloxycarbonyl-(3-(2-(2-(3-ami-
no-propoxy)-ethoxy)-ethoxy)propyl)-amine (37). Compound
36 (358 mg, 0.395 mmol) was dissolved in dry CH2Cl2
(6 mL). Trifluoroacetic acid (487 lL, 6.32 mmol) and tri-
iso-propylsilane (243 lL, 1.185 mmol) were added. After
stirring for 1.5 h, the solvent was removed. FC (CHCl3/
MeOH 6:1) gave 37 in quantitative yield as colorless oil.
Rf = 0.06 (CH2Cl2/MeOH 9:1); 1H NMR (600.1 MHz,
CDCl3, 20 ꢁC): d = 6.89 (br s, 1H; NHAc), 5.80 (br s, 1H;
NH), 5.60 (br s, 2H, Hvinyl), 5.19 (‘t’, J = 9.9 Hz, 1H; H-3),
4.96 (‘t’, J = 9.0 Hz, 1H; H-4), 4.68 (d, J = 8.4 Hz, 1H; H-
1), 4.58–4.50 (m, 2H; Hallyl), 4.32–4.16 (m, 3H; Hallyl, H-
6a), 4.04 (‘d’, J = 12.6 Hz, 1H; H-6b), 3.93 (‘q’, J = 9.2Hz,
1H; H-2), 3.76 (br s, 1H; H-5), 3.55–3.46 (m, 12H; OCH2),
3.25–3.18 (m, 2H; CH2N), 2.80 (br s, 2H; CH2N), 1.99 (s,
3H; C(O)CH3), 1.93 (s, 6H; C(O)CH3), 1.85 (s, 3H;
C(O)CH3), 1.69 ppm (br s, 4H; CH2CH2CH2); 13C NMR
(150.9 MHz, CDCl3, 20 ꢁC): d = 170.7, 170.6, 170.3, 169.4
(C(O)CH3), 156.6 (OC(O)NH), 128.7 (Colefin.), 128.6
(Colefin.), 99.6 (C-1), 72.7 (C-3), 71.6 (C-5), 70.4, 70.3, 70.0,
69.9, 69.6, 69.3 (OCH2), 68.8 (C-4), 64.6 (Callyl), 62.2 (C-6),
60.3 (Callyl), 54.3 (C-2), 39.6 (CH2N), 38.9 (CH2N), 31.6
(CH2CH2CH2) 29.4 (CH2CH2CH2), 23.2, 23.1, 21.0,
20.9 ppm (C(O)CH3); HRMS (nano ESI, pos. mode, H2O/
MeOH) for C29H49N3O14: [M+H]+ calcd 664.3287; found:
664.3261.
4.2.19. 4-Hydroxy-(Z)-but-2-enyl2-acetamido-2-deoxy-b-
D-glucopyranoside (27). According to GP 2, 4-hydroxy-
(Z)-but-2-enyl 2-acetamido-3,4,6-tri-O-acetyl-2-deoxy-
b-D-glucopyranoside 2643 (280 mg, 0.671 mmol) was dis-
solved in MeOH/EtMe2N 5:1 and reacted for 24 h.
Compound 27 was obtained as a white solid in quantita-
tive yield. Rf = 0.38 (MeCN/H2O 4:1); 1H NMR
(400.1 MHz, D2O, 20 ꢁC): d = 5.78–5.71 (m, 1H; Hvinyl),
5.64–5.53 (m, 1H; Hvinyl), 4.44 (d, J = 8.5 Hz, 1H; H-1),
4.28–4.18 (m, 2H; Hallyl), 4.09–4.07 (m, 2H; Hallyl), 3.83
(dd, 1H, J = 12.3, 1.2 Hz; H-6b), 3.73–3.63 (m, 1H; H-
6a), 3.59 (dd, 1H, J = 10.2, 8.6 Hz; H-2), 3.46–3.43 (m,
1H; H-5), 3.42–3.38 (m, 2H; H-3, H-4), 1.95 ppm (s,
3H; C(O)CH3); 13C NMR (100.6 MHz, D2O, 20 ꢁC):
d = 174.4 (C(O)CH3), 132.7 (Colefin.), 126.5 (Colefin.),
99.6 (C-1), 75.7 (C-4 or C-3), 69.7 (C-3 or C-4), 73.7
(C-5), 64.3 (Callyl), 60.5 (C-6), 56.9 (Callyl), 55.3 (C-2),
21.9 ppm (C(O)CH3); elemental analysis calcd (%) for
C12H21NO7 (291.30): C, 49.48; H, 7.27; N, 4.81; found:
C, 49.22; H, 7.44; N, 4.64.
4.2.22. 4-(2-Acetamido-2-deoxy-b-d-glucopyranosyloxy)-
(Z)-but-2-enyloxycarbonyl-(3-(2-(2-(3-amino-propoxy)-
ethoxy)-ethoxy)propyl)-amine (33). Compound 37
(259 mg, 0.39 mmol) was dissolved in dry MeOH
(8 mL). After addition of propylamine (192 lL,
2.34 mmol) and 0.5 M methanolic NaOMe solution
(450 lL, 225 lmol) and stirring for 2.5 h, the solvent
was evaporated. Compound 33 was obtained in quantita-
tive yield as colorless oil that was directly used for cova-
lent immobilization. 1H NMR (600.1 MHz, D2O, 20 ꢁC)
d = 5.70–5.58 (m, 2H; Hvinyl), 4.49–4.48 (m, 2H; Hallyl),
4.41 (d, J = 8.4 Hz, 1H; H-1), 4.25–4.21 (m, 2H; Hallyl),
3.78 (‘d’, J = 12.0 Hz, 1H; H-6a), 3.62 (dd, J = 12.0,
4.8 Hz, 1H; H-6b), 3.57–3.40 (m, 14H; H-2, H-5,
OCH2), 3.31–3.30 (m, 2H; H-4, H-5), 3.10–3.05 (m, 2H;
CH2N), 2.64 (‘t’, J = 6.9 Hz, 2H; CH2N), 1.90 (s, 3H;
C(O)CH3), 1.65–1.62 ppm (m, 4H; CH2CH2 CH2); 13C
NMR (150.9 MHz, D2O, 20 ꢁC) d = 175.3 (C(O)CH3),
159.2 (OC(O)NH), 129.9 (Colefin.), 129.5 (Colefin.), 100.9
(C-1), 76.9 (C-4 or C-5), 74.9 (C-3), 70.8 (C-5 or C-4),
70.6, 70.3, 69.8, 69.4 (CH2), 65.5 (Callyl), 61.8 (Callyl),
61.7 (C-6), 56.5 (C-2), 38.7, 38.5, 37.4 (CH2), 31.3, 29.7
(CH2CH2CH2), 23.1 ppm (C(O)CH3); HRMS (nano
ESI, pos. mode, H2O/MeOH) for C23H43N3O11:
[M+H]+ calcd 538.2970; found: 538.2965.
4.2.20. 4-(2-Acetamido-3,4,6-tri-O-acetyl-2-deoxy-b-D-
glucopyranosyloxy)-(Z)-but-2-enyloxycarbonyl-(3-(2-(2-
(3-(trityl-amino)-propoxy)-ethoxy)-ethoxy)propyl)-amine
(36). Compound 9 (1.36 g, 2.335 mmol) was dissolved in
dry CH2Cl2 (20 mL). Addition of 3-(2-(2-(3-(trityl-ami-
no)-propoxy)-ethoxy)-ethoxy)-propylamine
(900 mg,
1.946 mmol) dissolved in CH2Cl2 (2 mL) and EtN(i-
Pr)2 (340 lL, 1.946 mmol) afforded a yellow colored
solution. After stirring overnight at room temperature,
removal of the solvent and FC (EtOAc) gave 36 as a
white foam in quantitative yield. Rf = 0.18 (EtOAc);
1H NMR (600.1 MHz, CDCl3, 20 ꢁC): d = 7.38–7.36
(m, 6H; Har), 7.15–7.12 (m, 6H; Har), 7.06–7.03 (m,
3H; Har), 5.65 (br s, 1H; NH), 5.59 (br s, 2H, Hvinyl),
5.26 (‘t’, J = 9.9 Hz, 1H; H-3), 4.98 (‘t’, J = 9.6 Hz,
1H; H-4), 4.75 (d, J = 7.8 Hz, 1H; H-1), 4.54–4.47 (m,
2H; Hallyl), 4.29–4.18 (m, 3H; Hallyl, H-6a), 4.05 (‘d’,
1H, J = 10.8 Hz; H-6b), 3.84 (‘q’, J = 9.2 Hz, 1H; H-
2), 3.72–3.70 (m, 1H; H-5), 3.50–3.48 (m, 2H; CH2),
3.45–3.43 (m, 6H; CH2), 3.39–3.38 (m, 4H; CH2), 3.15
(‘d’, J = 6.0 Hz, 2H; CH2), 2.14 (br s, 2H; CH2), 1.95
(s, 3H; C(O)CH3), 1.90 (s, 3H; C(O)CH3), 1.89 (s, 3H;
C(O)CH3), 1.83 (s, 3H; C(O)CH3), 1.66–1.63 ppm (m,
4H; CH2CH2CH2); 13C NMR (150.9 MHz, CDCl3,

C. Maierhofer et al. / Bioorg. Med. Chem. 15 (2007) 7661–7676
7675
4.3. Screening of WGA ligands
4.3.5. Protocol for the ELLA employing covalently
modified microtiter plates (34). This assay was carried
out as described above for microtiter plates with non-
covalently immobilized porcine stomach mucin except
that 100 lL per well of each ABTS solution and
H2SO4 were used. All tests were performed in duplicate.
4.3.1. Protocol for non-covalent immobilization of porcine
stomach mucin. Microtiter plates (Becton–Dickinson,
Falcon ProBind, Cat. No. 353915) were incubated with
a solution of porcine stomach mucin (Serva, fraction 5,
pH 7.0) (4.12 lg mLꢀ1) in PBS buffer (10 mM sodium
phosphate, 150 mM NaCl, pH 7.3) overnight at room
temperature (100 lL per well). The plates were then
washed five times with PBST buffer (PBS buffer contain-
ing 0.05% Tween 20).
Acknowledgments
Financial support by the Deutsche Forschungsgemeins-
chaft (FOR 434 ‘‘Oligosaccharide and DNA Chips—
Analysis of Secondary Gene Products’’) is gratefully
acknowledged. We thank Sonja Seeberger for a sample
of neoglycopeptide 38 and Reinhold Weber for record-
ing the HRMS spectra.
4.3.2. Protocol for the ELLA employing microtiter plates
with non-covalently immobilized porcine stomach mucin.
The coated microtiter plates were blocked with 1% bo-
vine serum albumin (BSA) in PBS buffer for 90 min at
37 ꢁC (150 lL per well). The plates were then washed
five times with PBST buffer. Solutions of WGA-HRP
(final concentration of 1 lg mLꢀ1) and any tested inhib-
itor in serial dilutions in PBS buffer were pre-incubated
for 60 min at 37 ꢁC. Then the WGA-HRP/inhibitor
solutions were transferred to the blocked microtiter
plates (100 lL per well) and incubated at 37 ꢁC for
90 min. After having washed the plates five times with
PBST buffer, a solution of 2,20-azinobis(3-ethylbenzo-
thiazoline-6-sulfonic acid) diammonium salt (ABTS)
(12.5 mg per 50 mL) in citrate/phosphate buffer
(200 mM, pH 4.0) containing 0.015% of H2O2 was
added (50 lL per well). The plates were incubated for
30 min at room temperature in the dark. The reaction
then was stopped by addition of 1 M H2SO4 (50 lL
per well). After 5 min in darkness, the absorption was
measured at 405 nm using a plate reader (BMG Lab-
tech, FLUOstar OPTIMA). Percent inhibition was cal-
culated with the following equation:
References and notes
1. Sharon, N. J. Biol. Chem. 2007, 282, 2753–2764.
2. Gabius, H.-J., Ed.; Animal Lectins. Biochim. Biophys.
Acta 2002, 1572, 163–434.
3. Dam, T. K.; Brewer, C. F. Chem. Rev. 2002, 102, 387–429.
4. Lis, H.; Sharon, N. Chem. Rev. 1998, 98, 637–674.
5. Dube, D. H.; Bertozzi, C. R. Nat. Rev. Drug Discov. 2005,
4, 477–488.
´
´
6. Gabius, H.-J.; Siebert, H.-C.; Andre, S.; Jimenez-Barbero,
J.; Rudiger, H. ChemBioChem 2004, 5, 740–764.
¨
7. Shriver, Z.; Raguram, S.; Sasisekharan, R. Nat. Rev. Drug
Discov. 2004, 3, 863–873.
8. Essentials of Glycobiology; Varki, A., Cummings, R.,
Esko, J., Freeze, H., Hart, G., Marth, J., Eds.; Cold
Spring Harbor Laboratory Press: Cold Spring Harbor,
1999.
9. Dwek, R. A. Chem. Rev. 1996, 96, 683–720.
10. Werz, D. B.; Seeberger, P. H. Chem. Eur. J. 2005, 11,
3194–3206.
11. Nangia-Makker, P.; Conklin, J.; Hogan, V.; Raz, A.
Trends Mol. Med. 2002, 8, 187–192.
12. Rudiger, H.; Siebert, H.-C.; Solis, D.; Jimenez-Barbero, J.;
¨
Romero, A.; von der Lieth, C.-W.; Diaz-Maurino, T.;
Gabius, H.-J. Curr. Med. Chem. 2000, 7, 389–416.
13. Gardiner, J. M. Expert Opin. Invest. Drugs 1998, 7, 405–411.
14. Lingwood, C. A. Curr. Opin. Chem. Biol. 1998, 2, 695–700.
15. Lee, Y. C.; Lee, R. T. J. Biomed. Sci. 1996, 3, 221–237.
16. Collins, B. E.; Paulson, J. C. Curr. Opin. Chem. Biol. 2004,
8, 617–625.
17. Mammen, M.; Choi, S.-K.; Whitesides, G. M. Angew.
Chem. 1998, 110, 2908–2953; Angew. Chem., Int. Ed. 1998,
37, 2755–2794.
18. Wittmann, V. In Highlights in Bioorganic Chemistry:
Methods and Applications; Schmuck, C., Wennemers, H.,
Eds.; Wiley-VCH: Weinheim, 2004; pp 203–213.
19. Kiessling, L. L.; Pontrello, J. K.; Schuster, M. C.. In
Carbohydrate-Based Drug Discovery; Wong, C.-H., Ed.;
Wiley-VCH: Weinheim, 2003; Vol. 2, pp 575–608.
20. Roy, R. Trends Glycosci. Glycotechnol. 2003, 15, 291–310.
21. Lindhorst, T. K. Top. Curr. Chem. 2002, 218, 201–235.
22. Houseman, B. T.; Mrksich, M. Top. Curr. Chem. 2002,
218, 1–44.
% inhibition ¼ ð½Aðno inhibitorÞ ꢀ Aðwith inhibitorÞꢁ=Aðno inhibitorÞÞ
ꢂ 100
IC50 values were reported as the concentration of solu-
ble ligand required for inhibition of 50% of the binding
of WGA-HRP to coated microtiter plates. All tests were
performed in duplicate.
4.3.3. Preparation of microtiter plates 32 with amino-
reactive isothiocyanate surface. Microtiter plates (Nunc,
CovaLink NH modules, Cat. No. 478042) 30 were trea-
ted with a 0.5% solution of 1,4-phenylene diisothiocya-
nate (PDC) in DMSO50,60 containing 0.4% EtN(i-Pr)2
for 3 h at room temperature (100 lL per well). Then
the plates were emptied and rinsed twice with isopropa-
nol, twice with water, and twice with sodium carbonate/
bicarbonate buffer (100 mM, pH 10.1) to give modified
microtiter plates (32).
4.3.4. Protocol for covalent immobilization of compound
33 to the modified microtiter plates (32). The modified
microtiter plates 32 were incubated with a 1 mM solu-
tion of compound 33 in sodium carbonate/bicarbonate
buffer (100 mM, pH 10.1) overnight at room tempera-
ture (100 lL per well). The wells were then washed once
with isopropanol and twice with PBST buffer (pH 7.3) to
give microtiter plates 34.
23. Lee, R. T.; Lee, Y. C. Glycoconjugate J. 2000, 17, 543–551.
24. Lee, Y. C.; Lee, R. T. Acc. Chem. Res. 1995, 28, 321–327.
25. Lundquist, J. J.; Toone, E. J. Chem. Rev. 2002, 102, 555–
578.
26. Gestwicki, J. E.; Cairo, C. W.; Strong, L. E.; Oetjen, K. A.;
Kiessling, L. L. J. Am. Chem. Soc. 2002, 124, 14922–14933.

7676
C. Maierhofer et al. / Bioorg. Med. Chem. 15 (2007) 7661–7676
27. Kitov, P. I.; Sadowska, J. M.; Mulvery, G.; Armstrong, G.
D.; Ling, H.; Pannu, N. S.; Read, R. J.; Bundle, D. R.
Nature 2000, 403, 669–672.
28. Fan, E.; Zhang, Z.; Minke, W. E.; Hou, Z.; Verlinde, C. L.
M. J.; Hol, W. G. J. J. Am. Chem. Soc. 2000, 122, 2663–
2664.
43. Wittmann, V.; Lennartz, D. Eur. J. Org. Chem. 2002,
1363–1367.
44. de Paz, J. L.; Seeberger, P. H. QSAR Comb. Sci. 2006, 25,
1027–1032.
45. Shin, I.; Park, S.; Lee, M.-r. Chem. Eur. J. 2005, 11, 2894–
2901.
29. McCoy, J. P., Jr.; Varani, J.; Goldstein, I. J. Anal.
Biochem. 1983, 130, 437–444.
46. Feizi, T.; Chai, W. Nat. Rev. Mol. Cell Biol. 2004, 5, 582–
588.
30. Rendle, P. M.; Seger, A.; Rodrigues, J.; Oldham, N. J.;
Bott, R. R.; Jones, J. B.; Cowan, M. M.; Davis, B. G. J.
Am. Chem. Soc. 2004, 126, 4750–4751.
47. Wang, D. Proteomics 2003, 3, 2167–2175.
48. Feizi, T.; Fazio, F.; Chai, W.; Wong, C.-H. Curr. Opin.
Struct. Biol. 2003, 13, 637–645.
31. Koehn, M.; Benito, J. M.; Mellet, C. O.; Lindhorst, T. K.;
Fernandez, J. M. G. ChemBioChem 2004, 5, 771–777.
32. Corbell, J. B.; Lundquist, J. J.; Toone, E. J. Tetrahedron:
Asymmetry 2000, 11, 95–111.
49. Ortiz Mellet, C.; Garcia Fernandez, J. M. ChemBioChem
2002, 3, 819–822.
50. Hong, B. J.; Sunkara, V.; Park, J. W. Nucleic Acids Res.
2005, 33, e106.
´
33. Andre, S.; Pieters, R. J.; Vrasidas, I.; Kaltner, H.;
Kuwabara, I.; Liu, F.-T.; Liskamp, R. M. J.; Gabius,
51. Non-specific binding of WGA to microtiter plates
with chemically immobilized GlcNAc employing
bifunctional cross-linking reagent divinyl sulfone has
been reported earlier: Hatakeyama, T.; Murakami, K.;
Miyamoto, Y.; Yamasaki, N. Anal. Biochem. 1996,
237, 188–192.
H.-J. ChemBioChem 2001, 2, 822–830.
´
34. Roy, R.; Page, D.; Figueroa Perez, S.; Verez Bencomo, V.
Glycoconjugate J. 1998, 15, 251–263.
35. Haseley, S. R.; Kamerling, J. P.; Vliegenthart, J. F. G.
Top. Curr. Chem. 2002, 218, 93–114.
´
52. Vrasidas, I.; Andre, S.; Valentini, P.; Bo¨ck, C.; Lensch,
36. Wittmann, V.. In Glycoscience: Chemistry and Chemical
Biology; Fraser-Reid, B., Tatsuta, K., Thiem, J., Eds.;
Springer-Verlag: Heidelberg, 2001; Vol. III, pp 2253–2287.
37. Nilsson, U.; Striker, R. T.; Hultgren, S. J.; Magnusson, G.
Bioorg. Med. Chem. 1996, 4, 1809–1817, In the meantime,
microtiter plates with covalently immobilized carbohy-
drates are also commercially available from Lundonia
Biotech (Lund, Sweden). However, no information on the
type of linker is given.
38. Wittmann, V.; Seeberger, S. Angew. Chem. 2004, 116, 918–
921; Angew. Chem., Int. Ed. 2004, 43, 900–903.
39. Wright, C. S. J. Mol. Biol. 1980, 141, 267–291.
40. Wright, C. S.; Kellogg, G. E. Protein Sci. 1996, 5, 1466–
1476.
M.; Kaltner, H.; Liskamp, R. M. J.; Gabius, H.-J.; Pieters,
R. J. Org. Biomol. Chem. 2003, 1, 803–810.
53. Pieters, R. J. Trends Glycosci. Glycotechnol. 2004, 16,
243–254.
54. Kristiansen, A.; Nysæter, A.; Grasdalen, H.; Varum, K.
˚
˚
M. Carbohydr. Polym. 1999, 38, 23–32.
55. Bains, G.; Lee, R. T.; Lee, Y. C.; Freire, E. Biochemistry
1992, 31, 12624–12628.
56. Jordan, F.; Bassett, E.; Redwood, W. R. Biochem.
Biophys. Res. Commun. 1977, 75, 1015–1021.
57. Nagata, Y.; Burger, M. M. J. Biol. Chem. 1974, 249, 3116–
3122.
58. Privat, J.-P.; Delmotte, F.; Mialonier, G.; Bouchard, P.;
Monsigny, M. Eur. J. Biochem. 1974, 47, 5–14.
59. In this experiment the IC50 of GlcNAc was 83 mM.
60. Guo, Z.; Guilfoyle, R. A.; Thiel, A. J.; Wang, R.; Smith,
L. M. Nucleic Acids Res. 1994, 22, 5456–5465.
41. Zanini, D.; Roy, R. Bioconjugate Chem. 1997, 8, 187–192.
42. Wittmann, V.; Seeberger, S. Angew. Chem. 2000, 112,
4508–4512; Angew. Chem., Int. Ed. 2000, 39, 4348–4352.