Combining glycocluster synthesis with protein engineering: An approach to probe into the significance of linker length in a tandem-repeat-type lectin (galectin-4)

Article abstract of DOI:10.1016/j.carres.2013.12.024

Complementarity in lectin-glycan interactions in situ is assumed to involve spatial features in both the lectin and the glycan, giving a functional meaning to structural aspects of the lectin beyond its carbohydrate-binding site. In combining protein engineering with glycocluster synthesis, it is shown that the natural linker length of a tandem-repeat-type human lectin (galectin-4) determines binding properties in two binding assays (using surface-presented glycoprotein and cell surface assays). The types of glycocluster tested included bivalent lactosides based on tertiary amides of terephthalic, isophthalic, 2,6-naphthalic and oxalic acids as well as bivalent H(type 2) trisaccharides grafted on secondary/tertiary terephthalamides and two triazole-linker- containing cores. The presented data reveal a marked change in susceptibility to the test compounds when turning the tandem-repeat-type to a proto-type-like display. The testing of glycoclusters is suggested as a general strategy to help to delineate the significance of distinct structural features of lectins beyond their contact sites to the glycan.

Full text of DOI:10.1016/j.carres.2013.12.024

Carbohydrate Research 389 (2014) 25–38
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Combining glycocluster synthesis with protein engineering:
an approach to probe into the significance of linker length
in a tandem-repeat-type lectin (galectin-4)
Sabine André a, Guan-Nan Wang b, Hans-Joachim Gabius a, Paul V. Murphy b,
⇑
a Institute of Physiological Chemistry, Faculty of Veterinary Medicine, Ludwig-Maximilians-University Munich, Veterinärstr. 13, 80539 Munich, Germany
b School of Chemistry, National University of Ireland Galway, University Road, Galway, Ireland
a r t i c l e i n f o
a b s t r a c t
Article history:
Complementarity in lectin–glycan interactions in situ is assumed to involve spatial features in both the
lectin and the glycan, giving a functional meaning to structural aspects of the lectin beyond its carbohy-
drate-binding site. In combining protein engineering with glycocluster synthesis, it is shown that the nat-
ural linker length of a tandem-repeat-type human lectin (galectin-4) determines binding properties in
two binding assays (using surface-presented glycoprotein and cell surface assays). The types of glycoclus-
ter tested included bivalent lactosides based on tertiary amides of terephthalic, isophthalic, 2,6-naphtha-
lic and oxalic acids as well as bivalent H(type 2) trisaccharides grafted on secondary/tertiary
terephthalamides and two triazole-linker-containing cores. The presented data reveal a marked change
in susceptibility to the test compounds when turning the tandem-repeat-type to a proto-type-like dis-
play. The testing of glycoclusters is suggested as a general strategy to help to delineate the significance
of distinct structural features of lectins beyond their contact sites to the glycan.
Received 12 September 2013
Received in revised form 23 December 2013
Accepted 27 December 2013
Available online 11 January 2014
Keywords:
Agglutinin
Glycoprotein
Lectin
Modelling
Terephthalamides
Triazoles
Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction
terephthalamide, N,N0-diglucosylterephthalamide, glycophane
and triazole containing scaffolds.4 Terephthalamide was found to
Our understanding of the biological roles of complex carbohy-
drates is currently taking a quantum leap. Initially viewed to store
energy and confer rigidity to cell walls, the sugars’ unsurpassed
versatility for generating structural variability of oligomers is being
delineated as a platform for coding, with receptors (lectins) reading
and translating sugar-encoded messages into cellular effects.1 Fit-
tingly, cases of functional orchestration of glycan/lectin expression,
for example, in inflammation or by a tumour suppressor, are
unravelled, stimulating project lines to define the diagnostic/prog-
nostic potential of carbohydrate–protein interactions.2 Moving on
from work with plant lectins as models, the quest to explain the
inherent specificity and selectivity of the interplay of tissue lectins
with their (cognate) counterreceptors has inspired efforts to shape
synthetic binding partners, in terms of the structure of headgroups
and their spatial presentation.3 The asialoglycoprotein receptor on
hepatocytes has become a role model with its preference for triva-
lent binding partners, with its three contact sites separated by
about 15 Å, 22 Å and 23 Å.3a
be a scaffold suited for presenting carbohydrate ligands to plant
and human lectins, as exemplified by the bioactivity of compound
1.4a,b This glycocluster 1, which is a constrained compound, fea-
tures a distance of 7 to 8 Å between its two sugar headgroups
(intersugar or interheadgroup distance).4a In order to explore
structure–bioactivity relationships using such glycoclusters as
tools, we prepared compounds 2–7 (Chart 1) as a set of analogues
of 1, with lactose as the headgroup. Beside the terephthalamide
scaffold in 2, we generated acceptors for ligands based on tertiary
amides of oxalic acid (3), isopthalic acid (6) and 2,6-napthalic acid
(5). In these three cases, the interheadgroup distance is varied,
with possibly the headgroup orientation also being varied, while
maintaining the sugar and the acetylglycine residue bonded to
the nitrogen atoms. In compounds 2 and 4, a t-butyl residue was
incorporated instead of the acetate residue present in compounds
3, 5 and 6; it was anticipated that this modification would give an
indication as to whether the group attached to the anomeric nitro-
gen atom has an effect on the reactivity to a lectin. To assess the
impact of the headgroup, lactose was extended in four substances
to the trisaccharide fucosyl lactose, a positive regulator of reactiv-
ity for certain human lectins.4d,5 Compound 10 is a direct analogue
of fucosyl lactose containing 7, which contains secondary amides
rather than the tertiary amides found in 7 and the lactose deriva-
Our work in this area has thus far focused on bi- to tetravalent
compounds, where the headgroups were grafted to
⇑
Corresponding author. Tel.: +353 91492460.
E-mail address: paul.v.murphy@nuigalway.ie (P.V. Murphy).
http://dx.doi.org/10.1016/j.carres.2013.12.024
0008-6215/Ó 2014 Elsevier Ltd. All rights reserved.

26
S. André et al. / Carbohydrate Research 389 (2014) 25–38
OH
OH
OH
O
O
OH
O
OH
O
OH
O
R
HO
HO
HO
N
H
O
CO2Me
N
H
OH
O
OH
O
OH
O
O
HO
O
HO
O
HO
HO
HO
HO
O
O
N
O
N
O
N
HO
HO
HO
OH
N
O
HO
H
O
O
MeO2C
N
O
OH
OH
HO
OH
OH
HO
OH
N
O
O
N
O
O
HO
HO
H
O
4
O
1
2
3
R = CONHCH2CO2Me
R = CONHtBu
O
O
N
OH
OH
HO
OH
OH
HO
OH
R
HO
HO
OH
OH
OH
OH
O
O
OH
HO
OH
O
OH
O
O
HO
HO
N
CO2Me
N
H
CO2Me
O
OH
O
OH
O
H
O
O
N
H
O
CO2Me
OH
O
O
HO
HO
O
HO
HO
O
N
O
N
HO
N
O
HO
HO
HO
O
OH
OH
OH
O
OH
N
OH
O
HO
N
H
O
HO
N
H
MeO2C
O
O
OH
O
O
O
H
N
OH
MeO2C
N
N
HO
O
OH
OH
OH
O
OH
HO
HO
HO
O
HO
O
O
MeO2C
O
OH
HO
O
OH
OH
HO
OH
O
N
5
6
7
OH
OH
OH
O
OH
OH
O
OH
O
OH
O
OH
O
O
HO
HO
HO
OH
OH
OH
O
OH
O
N
H
O
HO
O
HO
N
O
HO
O
N
N
HO
O
H
O
O
O
O
O
N
N
N
OH
OH
O
OH
O
HO
HO
HO
O
HO
OH
OH
OH
O
O
OH
OH
OH
OH
OH
OH
O
N
OH
OH
OH
8
9
10
OH
N
N
O
HO
OH
N
O
N
O
O
OH
HO
O
OH
N
O
O
HO
OH
OH
O
OH
O
OH
HO
OH
O
OH
OH
OH
OH
OH
Chart 1. Structure of glycoclusters 1–10.
tives 2–6. Two additional bivalent fucosyl lactose-presenting sub-
stances, that is, 8 and 9, were also included, here with sugar
attached to triazole-containing linkers previously proven as being
compatible for lectin reactivity.4c Since bioactivity of compounds
of these classes has thus previously been documented for members
of the adhesion/growth-regulatory family of human/animal galec-
tins,4 we can herein proceed to test these bivalent clusters as sen-
sors to answer a distinct question on the structure–activity
relationship of a human galectin.
Among the galectins, one of their three groups is established by
the tandem-repeat design. Two different carbohydrate recognition
domains (CRDs) are covalently connected by a linker (Fig. 1).6 Hu-
man galectin-4 (Gal-4), a molecular transporter in glycoprotein
routing and delivery in enterocytes and neurons,7 has a single fixed
length of this connecting peptide (Fig. 1). In assay settings which
determine inhibition of lectin binding to glycoproteins, Gal-4 has
proven to be rather sensitive to the presence of glycoclusters,
secondary and tertiary terephthalamides among them.4a,8 This
capacity of glycoclusters to interfere with Gal-4 binding to a
ligand-exposing surface (on a plastic surface or cell surface)
enables addressing the issue of the relevance of linker length for
binding capacity to Gal-4. To do so, two variants, that is, Gal-4V
with length reduction and Gal-4P with complete linker removal
mimicking a dimer of the proto-type group, were designed and
produced (Fig. 1), then tested with the panel of compounds shown
Figure 1. Illustration of the design of the two Gal-4 variants by comparatively
presenting the sequence of the natural 42-amino-acid linker along with the
positions of deletions between the N- and C-terminal CRDs and the resulting
molecular masses (a) as well as the linker lengths (b) in Gal-4V and Gal-4P. The
products of recombinant expression show purity and the expected positions in gel
electrophoretic analysis (c).
(proto-type) galectin (CG-1A) with strong reactivity to N-glycans
with LacNAc termini was used as a control,9 and the sensor
in Chart
1 in solid-phase/in vitro assays. A homodimeric

S. André et al. / Carbohydrate Research 389 (2014) 25–38
27
character of the compounds was further tested in a second system
of galectin/galectin variant. This was established by chimera-type
galectin-3 (constituted by the CRD and a stalk of the N-terminal
peptide with two sites for Ser phosphorylation and nine Pro/Gly-
rich collagen-like tandem repeats; CG-3) and its proteolytically
truncated form, that is, the CRD free of this collagenase-sensitive
N-terminal tail.10 The marked effects of protein engineering/trun-
cation on the reactivity of lectins to certain glycoclusters under-
score the structural significance of the natural linker length of
Gal-4 and the presence of the collagen-like tail in CG-3.
with N-(9-fluorenylmethoxycarbonyloxy)-succinimide (FmocOSu)
in pyridine and the free hydroxyl groups were acetylated by adding
acetic anhydride. The Fmoc was removed thereafter using morpho-
line in DMF to give the fucosyl lactosamine 18. This was subse-
quently reacted with formaldehyde, methyl isocyanoacetate and
terephthalic acid to yield a diamide. The protecting groups were
removed by Zemplén deacetylation and hydrogenolysis to give rise
to 7 (Scheme 3).
2.2. Structural analysis and molecular modelling
Compounds 1–7 and 10 are divalent glycosyl amides, structur-
ally related to terephthalamides described previously.11 In 1–7, the
amides are tertiary amides, whereas 10 harbours secondary
amides. In terms of amide configuration, the amides in 10 are trans
amides (or Z amides). This term trans is applied as in peptide chem-
istry. Compound 10 shows one set of signals in the NMR spectrum,
and the chemical shift of the anomeric proton appears at d 5.84 as a
doublet with a coupling constant of J = 9.3 Hz. This is consistent
with 10 adopting a trans–trans (ZZ) structure shown in Figure 2,
by comparison with related secondary amides.11 However, and as
shown previously for tertiary terephthalamides, the tertiary
amides in 1–7 adopt configurations where the cis (E) amide is pre-
ferred. The NMR spectra show two signal sets for all these com-
pounds. For example, for 2 (Fig. 3) the anomeric proton for the
cis (or E) amide appears at d 4.84 (J = 8.8 Hz), whereas that for
the trans amide was observed at d 5.76 (J = 9.3 Hz). By integration
of suitable signals in the 1H NMR spectra the ratio of cis:trans (E:Z)
and consequently the cis–cis (EE) to the cis–trans (EZ) isomeric ra-
tios are determined. Thus for compound 2 (Fig. 3) there are two
well separated signals for the galactose anomeric protons in iso-
meric structures and these are observed at d 4.50 (d, J = 7.8 Hz,
2H, H-10 Z) and d 4.44 (d, J = 7.8 Hz, 2H, H-10 E); integration of these
signals enabled establishment of the EE:EZ ratio in 2 to be 70:30.
These ratios are summarized for compounds 1–7 in Table 1. All
compounds 1–7 showed >2:1 preferences in solution for the cis–
cis isomer. There are essentially two conformations that can be
considered for the cis–cis isomers, and these are depicted in Fig-
ure 2. The carbohydrate headgroups could be stacked (stacked
cis–cis) or presented in an extended conformation (extended cis–
cis). The extended cis–cis structure is considered to be more rele-
vant for binding to lectins, because headgroups might be too close
for reactivity when stacked. It is also possible that lectins could
bind to their ligand in the EZ or cis–trans structure, which is pres-
ent to a lesser degree in solution. Dynamic transition to the EZ con-
former is possible in the equilibrium as a result of amide bond
rotation, and there is evidence that this exchange can take place
for compounds related to 1–7.11 While the trans–trans (ZZ) isomer
for compounds 1–7 (not shown) could also be adopted, it appears
to be present at concentrations too low to be detected by NMR.
Based on the NMR data for compounds 1–7 we conclude that
the major structural isomer is cis–cis or EE. Assuming that the ex-
tended cis–cis structure is more relevant than the stacked cis–cis
structure and cis–trans structures for presenting the headgroup
for binding to a lectin, we next carried out molecular dynamics
simulations using Macromodel 8.0 (Schrödinger Inc., LLC, New
York, NY, USA), to supplement our previous studies of divalent
compounds.4a The main purpose of this work is to estimate dis-
tances between the carbohydrate headgroups and respective ori-
entations. Models were built for the compounds 1–3 and 5–6,
and each compound was treated equally. Stochastic dynamics
was applied to the selected structure at 300 K with an equilibration
time of 1 ns and a time step of 1.5 fs using the OPLS-AA force field
in the gas phase. A simulation was carried out in Macromodel for
compound 2 using the GB/SA effective solvent model for water
(not explicit water) and no major differences were observed
2. Results and discussion
2.1. Synthesis of the glycoclusters
The preparation of the bivalent glycoclusters 1–6 with lactose
as headgroup started from the lactosamine 11 (Scheme 1) and in-
volved the Ugi reaction. The reaction of 11 with formaldehyde,
terephthalic acid and methyl isocyanoacetate gave 1 after deacet-
ylation, as described earlier.11 When 11 was reacted with oxalic
acid, 2,6-naphthalenedicarboxylic acid and isophthalic acid,
respectively, in the presence of formaldehyde and methyl iso-
cyanoacetate, and the acetyl groups removed, then compounds 3,
5 and 6 were obtained. When t-butyl isocyanide was reacted with
lactosamine 1, formaldehyde and terephthalic acid or isophthalic
acid in methanol, the bivalent compounds 2 and 4 resulted after
deacetylation.
The syntheses of bivalent glycoclusters presenting as headgroup
the trisaccharide fucosyl lactose were next carried out (Schemes 2
and 3). Firstly, the fucosylated azide 12 was prepared as previously
described.4d This azide was used in copper-catalysed azide alkyne
cycloaddition reactions12,13 with 14 and 15 to give, after deacetyla-
tion and subsequent HPLC-based purification, 8 and 9, respectively.
Although the azide 12 contained a small amount of the a-anomer
(<6%), it was possible to remove the cycloaddition products that re-
sulted from the presence of this anomer by careful chromatogra-
phy prior to the final deprotection step. Next, catalytic
hydrogenation of 12 resulted in the 20-O-fucosyl lactosamine 13
(Scheme 2), which when used in a coupling reaction with tereph-
thaloyl chloride in the presence of DIPEA and followed by deacet-
ylation was turned into 10. Since the glycosyl amine 13, formed by
the reduction of the azide, contained a small amount of its a-ano-
mer (<10%), a mixture of diamides was obtained from the coupling
reaction with the terephthaloyl chloride. Purification by HPLC was
necessary in order to obtain 10 as a single product.
Since the Ugi reactions of 11 were successful, we also were
interested to try to prepare the fucosyl lactose analogue of 1 from
13. However, the Ugi reaction did not succeed for peracetylated
compound 13. To solve this problem the Ugi reaction of amine
18 was carried out instead. Compound 16 was treated with satu-
rated solution of ammonium hydrogen carbonate to give 17. Next,
the amine group of 17 was temporarily protected by a treatment
AcO
OAc
O
1.Dicarboxylic acid,
OAc
O
Isocyanide, CH2O, MeOH
O
AcO
1-6
AcO
NH2
OAc
2. NaOMe, MeOH
OAc
11
C
N
Isocyanide =
or
C
N
CO2Me
Scheme 1. Synthesis of bivalent glycoclusters 1–6.

28
S. André et al. / Carbohydrate Research 389 (2014) 25–38
AcO
OAc
AcO
AcO
AcO
OAc
O
O
OAc
O
OAc
O
1, Terephthaloyl chloride
DIPEA, THF, rt, 87%
O
Pd/C, H2
O
O
O
AcO
AcO
OAc
12
1.
NH2
N3
10
AcO
AcO
EtOAc
95%
O
O
2. NaOMe, MeOH, 76%
after HPLC
OAc
OAc
OAc
13
AcO
AcO
O
O
1.
14
15
.
CuSO4 H2O
.
CuSO4 H2O
sodium ascorbate
MeOH-H2O, 92%
sodium ascorbate
MeOH-H2O, 88%
2. NaOMe, MeOH, 68%
after HPLC
2. NaOMe, MeOH, 68%
after HPLC
9
8
Scheme 2. Synthesis of 8–10.
OH
OH
OH
OH
O
O
HO
HO
OH
O
OH
O
O
HO
O
HO
NH4HCO3
O
O
NH2
HO
OH
CH3OH-H2O
HO
O
O
OBn
17
OBn
16
OBn
OBn
OBn
OBn
AcO
AcO
OAc
O
1. Terephthalic acid
CNCH2COOCH3
CH2O, MeOH
1. FmocOSu
Ac2O, Py.
OAc
O
O
O
AcO
OBn
18
NH2
7
2. Morpholine
DMF
AcO
O
2. NaOMe, MeOH
3. H2-Pd, MeOH
OBn
OBn
Scheme 3. Synthesis of 7.
between the gas phase simulation of 2 and estimating the GB/SA
effective water model for 2. During each of the subsequent simula-
tions 100–200 structures were sampled and an internal coordinate
system was used to define spatial parameters for these structures.
This system firstly involves measuring the distances for the sam-
pled structures between the anomeric carbon atoms of the glucose
unit in the disaccharide (in Å) to give the interheadgroup distance.
In addition, two dihedral angles were measured, a core dihedral
and a galactose dihedral, for the sampled structures, providing
parameters relevant to the interheadgroup orientation, and these
are defined in the top part of Figure 4. Scatter plots were generated
using the data obtained, and these are also shown in Figure 4.
These plots can be considered representative of the spatial
arrangements between the lactose headgroups accessible during
the simulations, which were run at identical conditions.
apparently not greatly influenced by replacing the terminal acetate
with the t-butyl group. Similarly, the distance between the two
disaccharides did not change noticeably, averaging between 7–
8 Å for both compounds during the simulations, as can also be ex-
pected for compound 7. For 6, which has the isophthalic acid, sim-
ilar interlactose distances to compounds 1 and 2 were observed,
but there was more flexibility in interheadgroup orientation, as
can also be expected for the structurally similar compound 4. With
regard to compound 5, which has naphthalene rather than ben-
zene, the distance between the lactose residues is increased, as
would be expected, and this varied between 9 and 10 Å during
the simulation. Molecular dynamics simulations for compound
10 was not carried out but, on the basis of studies with a related
lactose diamide,4a the distances between the residues would be ex-
pected to be 10 Å and similar to 5. The interheadgroup orientation
in 5 was similar to that found in compounds 1 and 2. Coming to the
ditriazoles, the distance between the two trisaccharides in 8 would
be expected to approximate to 13 Å, whereas compound 9 is more
flexible with distances of up to 16–17 Å being possible. At the other
As can be seen, by comparing data for compounds 1 and 2, the
measured galactose and core dihedrals did not vary significantly,
indicating that the relative orientations between the two potential
contact sites for galectins are similar for these compounds and

S. André et al. / Carbohydrate Research 389 (2014) 25–38
29
A: Divalent tertiary amides
OH
O
R
OH
O
R
OH
O
R
R2
HO
R2
HO
R2
HO
O
N
N
O
N
O
N
OH
OH
OH
C-2 axis
O
C-2 axis
R
OH
OH
R2
O
O
N
HO
R2
N
O
OH
O
HO
O
HO
R
OH
R
OH
R2
cis-trans
structur e
extended cis-cis structure
stacked cis-cis structure
HO
B: Divalent secondary amides
OH
R2
O
H
N
HO
O
OH
HO
N
O
O
R2
HO
H
trans-trans structure
HO
Figure 2. Amide structures.
Figure 3. Partial 1H NMR spectrum of compound 2 (D2O, 500 MHz, d 4.35–4.80 ppm).
side of the range of distances, for the oxalic acid derivative 3, its
lactose residues were separated by 3–4 Å and the interheadgroup
dihedrals were also significantly different.
biantennary complex-type N-glycans with backfolding (ꢀ5.9 Å),
without backfolding (8.1 Å) and with anti-parallel configuration
(22.1 Å), core substitutions regulating the position of the
a1,3/6-
Broadly spoken, the distances between sugar headgroups in this
range of compounds lie in the range of those of LacNAc termini in
arms, and 15–22 Å for the type 1 triantennary complex-type N-gly-
can.14 In this sense, the synthetic compounds share spatial

30
S. André et al. / Carbohydrate Research 389 (2014) 25–38
Table 1
branch-end LacNAc can become a binding partner for a CRD of hu-
Ratio of isomers of compounds 1–7
man galectins at saturation in solution, albeit with a gradient of
decreasing affinity,9b,16 the compounds of this panel may disclose
information relevant for the discussion on the significance of the
natural length of the linker. As in its function in glycoprotein rout-
ing, each Gal-4 is expected to react with headgroups of two syn-
thetic molecules. Space-filling models have been shown for
compounds 7–9 (Fig. 5).
OH
R1
OH
O
R1
N
R2
HO
O
R2
R3
N
O
HO
OH
OH
O
R3
cis (E)
trans (Z)
2.3. Analysis of inhibitory potency
In the first type of assay, a glycoprotein (asialofetuin with up to
nine LacNAc termini in the three triantennary complex-type N-gly-
cans) was adsorbed to the plastic surface of microtiter plate wells,
a matrix for carbohydrate-dependent binding of Gal-4. Binding
was saturable, dependent on the presence of the terminal galactose
moieties of the N-glycans and was blocked completely by an excess
of hapten (lactose), for all galectin proteins tested (not shown).
Titrations at an assay setting in the linear range of the OD-response
with increasing concentrations of the test compounds determined
Compound
cis:trans (E:Z) ratio
cis–cis:cis–trans ratio (EE:EZ, see Fig. 3)
1
2
3
4
5
6
7
85:15
85:15
85:15
85:15
92:8
70:30
70:30
70:30
70:30
84:16
84:16
76:24
92:8
88:12
the concentration to reduce extent of the signal by 50% (IC50
-
value), a relative measure of the capacity to interfere with
glycan–lectin interaction (for representative cases, please see
properties with a natural constellation. Since Gal-4 is strongly
reactive with N-glycans presenting LacNAc termini,15 and every
OH
O4
C4
HO
OH
O
O
OH
C1'
O
C1
O
HO
HO
HO
O
HO
HO
O
HO
OH
C4'
HO
Core dihedral = angle C4-C1-C1'-C4'
Galactose dihedral = angle O4-C4-C1'-O4'
O4'
Distance between lactose units = C1 to C1'
Figure 4. Spatial parameters for EE structures sampled during molecular dynamics simulations are represented using scatter plots. The definition for the dihedrals and
distance between lactose units is included in the upper part of this figure.

S. André et al. / Carbohydrate Research 389 (2014) 25–38
31
Figure 5. Space filling models of fucosyl lactose derivatives 7 (top left), 8 (top right) and 9 (bottom).
Fig. 6). As given in Table 2, Gal-4 binding was rather effectively
inhibited by the bivalent compounds, extending the previously col-
lected evidence with a tertiary terephthalamide (compound 1) and
The results obtained by testing these two variants under identi-
cal conditions clearly showed that: (i) reduction of linker length
decreased the lectin’s susceptibility to the test compounds; and
(ii) this truncation of the linker and its complete removal led to
similar results (Table 2). Most conspicuously, the two ditriazoles
8 and 9 dropped markedly in inhibitory capacity, a result definitely
warranting confirmation by a different assay type. However, a spe-
cial point precludes direct comparability for all data: the interpre-
tation of the data for fucosyl lactose-presenting compounds 7–10
should take into account that the reactivity of Gal-4V to this epi-
tope has been shown to decrease relative to that of the wild-type
protein.5c An impact of the truncation both on affinity and on li-
gand selection has furthermore been revealed for surface binding
to human neuroblastoma cells.17 To document how a dimeric
galectin will react with the glycoclusters we have added CG-1A.
The range of IC50-values of the two engineered Gal-4 proteins
was rather similar to that of this control protein (Table 2). Its level
of reactivity was in line with previous results for CG-1A and dilac-
tosyl diamides/di-, tri- and tetravalent fucosyl lactosides.4c,d
In order to exclude the possibility that characteristics inherent
to the solid-phase assay underlie the differences seen in Table 2
we next performed cell-binding assays. As a test principle, binding
parameters of the labelled galectin to the surface of cells in culture
can be affected by the presence of the test compounds and this ef-
fect was quantified in terms of the percentage of positive cells and
mean fluorescence intensity relative to mock-treated controls,
which were always processed in parallel. This assay reflects the
sensitivity of galectin binding to cell surface glycans in vitro, when
the compounds are tested comparatively. To ensure identical con-
ditions each series was performed with aliquots of the same cell
suspension. Matching Gal-4 specificity, two lines with an abun-
dance of LacNAc-presenting N-glycans were selected, that is, hu-
man pancreatic carcinoma cells expressing the tumour
suppressor p16INK4a and the Lec2 mutant of the Chinese hamster
ovary (CHO) cell panel.
N,N0-diglucosylterephthalamides.4a The addition of the
a1,2-linked
fucose moiety to the lactose headgroup expectedly led to an in-
crease in inhibitory potency, from 0.3 mM for lactose in the sec-
ondary terephthalamide4a to 0.05 mM (compound 10). The
introduction of the t-butyl group (for compounds 2 and 4) can
make a rather minor difference, and having oxalic acid in the back-
bone proved rather favourable, compared to isophthalic acid (Ta-
ble 2). The two ditriazoles with the trisaccharide were the most
potent inhibitors (Table 2). Overall, the data further substantiate
the sensitivity of this lectin to bi- and trivalent glycoclusters, here-
by establishing a solid basis to ask the question whether and how
linker-length reduction in the tandem-repeat-type lectin will affect
the reactivity towards the glycocluster. Two forms of engineered
Gal-4 were therefore tested, by maintaining linker parts phyloge-
netically conserved in mammals (Gal-4V) and by turning Gal-4
into a linkerless, proto-type-like form (Gal-4P) (Fig. 1).
In each case, titrations with lectin at 2 ꢁ 105 cells per assay
identified the linear range of the response, as in the solid-phase as-
says, to enable the work to be carried out with optimal sensitivity.
Instead of the OD-value, this type of analysis provides information
on number of positive cells and the mean fluorescence intensity.
Figure 6. Courses of titrations illustrating the extent of inhibition of binding of
labelled Gal-4V (5 lg/mL) to surface-presented asialofetuin in the solid-phase assay
upon increasing the concentration of lactose (A) as well as compound 9 (h) and
compound 10 (j) (B).

32
S. André et al. / Carbohydrate Research 389 (2014) 25–38
Table 2
IC50-values (in mM) of the test compounds for blocking binding of labelled galectins to the glycans of the surface-immobilized glycoprotein asialofetuina
Compound
Gal-4 (5
lg/mL)
Gal-4V (5
l
g/mL)
Gal-4P (2
lg/mL)
CG-1A (3 lg/mL)
1b
2
3
4
5
6
7
8
9
0.08
0.05
0.08
0.12
n.d.
n.d.
0.5
0.6
0.4
n.d.
0.4
n.d.
0.2
0.1
0.3
3
n.d.
0.4
0.6
0.5
n.d.
0.4
n.d.
0.3
0.3
0.4
6
n.d.
0.1
0.2
0.14
n.d.
0.12
n.d.
0.2
0.1
n.d.
0.008
0.01
0.05
0.3
0.4
0.3
10
Lac
1.6 (2.5b)
n.d.: not determined.
a
Given as concentration of sugar (not test compound).
From Ref. 4a.
b
Gal-4 bound to the human cells in a carbohydrate-dependent man-
ner (Fig. 7A). As exemplarily shown for compound 2 and also for
compound 5, lactose as a headgroup in glycoclusters was much
more active than lactose free in solution (Fig. 7B). The direct com-
parison between the compounds with terephthalic acid versus the
2,6-naphthalic acid cores showed that there is no major difference
between these scaffolds, also reflecting their rather similar inter-
headgroup orientation described above. A grading of activities
was seen for the trisaccharide-presenting clusters 7, 9 and 10
(Fig. 7C). The tertiary terephthalamide 7 proved very active, as also
seen for the ditriazole 9, and, fully in line with a rather low potency
of the lactose-presenting secondary terephthalamide in the solid-
phase assay,4a compound 10 was found to be the least active. Its
inhibitory capacity was in the range of that for more suited scaf-
folds 2 and 5 presenting lactose instead of the trisaccharide
(Fig. 7B and C).
When testing labelled Gal-4V, inhibition by free sugars corrob-
orated the comparatively reduced level of reactivity to the trisac-
charide mentioned above (Fig. 7D). The same applied to the
reactivity to the clusters 2 and 5 (Fig. 7E and F). Despite a 3-fold
higher sugar concentration the extent of inhibition was less for
Gal-4V than for Gal-4 binding (Fig. 7B and C versus Fig. 7E and
F), as seen in the solid-phase assays (Table 2). As similarly seen
for the second cell line (Lec2 CHO) and for labelled Gal-4P (not
shown), results in both assay types consistently indicated that
the linker has a bearing on the lectin’s reactivity to bivalent gly-
coclusters and to the trisaccharide (20-fucosyl lactose) headgroup.
Evidently, this peptide portion of Gal-4 does more than simply con-
necting the two CRDs, and there is corroborating evidence for this
assumption. Having measured the diffusion constant of the pro-
teins, the (counterintuitive) decrease from 1.02 0.01 to
0.92 0.01ꢁ10ꢂ6 cm2 sꢂ1 by truncation (reducing the molecular
mass), along with acquisition of lactose dependence of this param-
eter,18 argues in favour of an active role of the linker, above a cer-
tain length limit, in making Gal-4 structurally somehow compact.
In this spatial form, probably with some flexibility to adapt in
the presence of bi- to oligovalent ligands, the lectin is more sensi-
tive to the test compounds than its variants. Headgroup distance in
the tested range, which covers the distances in biantennary N-gly-
cans, altered sensitivity for Gal-4 to a limited extent, while invari-
ably being less active for Gal-4 after truncation.
of a protein part, here the tail, a natural proteolytic process, alters
the sensitivity. As shown in Figure 8 for experiments with the wild-
type CHO line, CG-3 was much more reactive with test compounds
than the truncated version, extending previous observations with,
for example, a bivalent glycocluster with triazole linkers that had a
more than 2-fold higher capacity to interfere with CG-3 binding.4d
The differences noted in the cytofluorimetric assays were in this
range (Fig. 8B and D).
3. Conclusions
Two factors combine to establish the reactivity of a glycan to a
lectin: the structure of the cognate epitope and its spatial presen-
tation.1 To attain the required high-level specificity and selectivity
in counterreceptor selection in vivo the way sites for contact are
positioned in lectins is assumed to matter markedly, too. If this is
true, glycoclusters should sense engineered changes in lectins,
serving as tools for structure–activity correlations. Our data using
two engineered Gal-4 variants and the test panel of bivalent gly-
coclusters revealed special properties for the protein with the nat-
ural linker length. Its reduction decreased the susceptibility to
glycoclusters in the two binding assays and the reactivity to the
H-type trisaccharide, encouraging further testing, e.g. in agglutina-
tion assays using glycodendrimersomes.20 Also considering the re-
sults with CG-3, this approach, teaming up protein engineering
with glycocluster synthesis and testing, is thus illustrated to be
useful on the way to eventually resolve the question on the phys-
iological significance of the length of sequence extension (Gal-3) or
linker (Gal-4) in endogenous lectins, for galectins and other lectin
families sharing these types of modular design.
4. Materials and methods
4.1. General experimental
Unless otherwise noted, all commercially available compounds
were used as provided without further purification. Solvents for
chromatography were technical grade. Petroleum ether 40–60 °C
was used for column chromatography and thin-layer chromatogra-
phy (TLC). NMR spectra were recorded (25 °C). The frequency was
500 MHz for 1H NMR and 125 MHz for 13C NMR. Data are reported
in the following order: chemical shift (d) in ppm; multiplicities are
indicated s (singlet), d (doublet), t (triplet), q (quartet), m (multi-
plet); coupling constants (J) are given in Hertz (Hz). Chemical shifts
are reported relative to internal Me4Si in CDCl3 (d 0.0) or HOD for
D2O (d 4.79, 25 °C) for 1H and Me4Si in CDCl3 (d 0.0) or CDCl3 (d
77.0) for 13C. 1H NMR signals were assigned with the aid of COSY,
Whether a different type of protein processing will also have a
bearing on sensitivity was tested in a second system. Galectin-3 is
special among galectins due to its N-terminal stalk with the two
sites for Ser phosphorylation and the collagen-like repeats, this
section connected as tail to the CRD. It is responsible for oligomer-
ization in the presence of polyvalent ligands and high-affinity
binding to clustered counterreceptors.19 In this case, too, removal

S. André et al. / Carbohydrate Research 389 (2014) 25–38
33
Figure 8. Semilogarithmic representation of fluorescent surface staining of parental
CHO cells by labelled CG-3 (5 g/mL; A and B) and proteolytically truncated CG-3
(2 g/mL; C and D); for further details, please see legend of Figure 7. Inhibition of
l
l
CG-3-dependent staining by 10 mM lactose, 0.5 mM fucosyl lactoside and 0.5 mM
lactose (A) and by 0.5 mM sugar presented by compounds 8, 9 and 7 (B). Inhibition
of staining by truncated CG-3 by 10 mM lactose, 0.5 mM fucosyl lactoside and
2 mM lactose (C) and by 0.5 mM sugar presented by compounds 8, 9 and 7 (D).
Purification System. Anhydrous DMF, pyridine and EtOH were used
as purchased from commercial suppliers.
4.2. Synthetic procedures
Figure 7. Semilogarithmic representation of fluorescent surface staining of human
pancreatic carcinoma cells (Capan-1), reconstituted for the expression of the
4.2.1. Typical procedure for the Zemplén deacetylation
The acetylated compound (0.02 mmol) was dissolved in metha-
nol (5 mL), a catalytic amount of NaOMe (0.1 mL of a 0.2 M solu-
tion in MeOH) was added and the resulting mixture was stirred
for 1 h at room temperature. Amberlite IR-120 (plus) was added,
the mixture was neutralized and the resin was then removed by fil-
tration and washed with water. Finally, the solvents were removed
under diminished pressure and the residue subjected to subse-
quent chromatographic purification.
tumour suppressor p16INK4a, by labelled Gal-4 (10
lg/mL; A–C) and its Gal-4V
variant (5
lg/mL; D–F). The control value (background) for the signal obtained by
processing cells with the fluorescent indicator in the absence of the lectin is drawn
as grey-shaded area of the respective scan data, the 100%-value (lectin-dependent
staining in the absence of a test compound) as thick black line. The measured
results on staining (percentage of positive cells/mean fluorescence intensity) are
given in each panel in the order of listing the lactose concentration/type of
compound (from top to bottom), all concentrations in sugar (not bivalent scaffold).
In detail, the top number defines the control, the next pairs the test cases in the
order of listing in the text and finally, at the bottom, the 100%-value. Inhibition of
Gal-4-dependent staining by 1 mM/0.5 mM lactose and 50
by 40 M sugar presented by compounds 2/5 (B) and by compounds 7, 9 and 10 (C)
relative to the 100% control. Inhibition of Gal-4V-dependent staining by 1 mM/
0.5 mM lactose and 120 M fucosyl lactoside (D), by 120 M sugar presented by
M compounds 7, 9 and 10 (F) relative to the 100%
lM fucosyl lactoside (A),
l
4.2.2. N,N0-Di(b-
D-galactopyranosyl-(1?4)-b-D-glucopyranosyl)-
N,N0-di[(t-butylcarbamoyl)-methyl]terephthalamide 2
l
l
compounds 5/2 (E) and by 120
l
Terephthalic acid (13.6 mg, 0.08 mmol), lactosyl amine
1
control.
(100 mg, 0.16 mmol) and formaldehyde (16 lL of a 37% solution,
0.19 mmol) were suspended in MeOH (5 mL) and the mixture
was stirred at room temperature for 1 h. t-Butyl isocyanide
(19 lL, 0.16 mmol) was then added and the mixture was stirred
13C NMR signals using DEPT, gHSQCAD and/or gHMBCAD. Low-
and high-resolution mass spectra were in positive and/or negative
mode, as indicated in each case. TLC was performed on aluminium
sheets precoated with silica gel and spots visualized by UV and
charring with H2SO4–EtOH (1:20), or cerium molybdate. Flash
chromatography was carried out with silica gel 60 (0.040–
0.630 mm) and using a stepwise solvent polarity gradient corre-
lated with TLC mobility. CH2Cl2, MeOH, toluene and THF reaction
solvents were used as obtained from a Pure Solv™ Solvent
at room temperature overnight. The reaction was heated to 45 °C
for 10 h and then solvent was then removed under reduced pres-
sure. Chromatography of the residue (CH2Cl2–CH3OH, gradient elu-
tion, 70:1 to 60:1 to 50:1) gave the protected intermediate as a
white amorphous solid (98 mg, 77%). Zemplén deacetylation and
subsequent purification using a BioGel P-2 gel column (eluent/
H2O) and subsequent C18 reverse-phase column (gradient elution,
H2O to H2O–CH3OH, 95:5) gave 2 as an interconverting mixture of

34
S. André et al. / Carbohydrate Research 389 (2014) 25–38
EE and EZ isomers (4:1); ½a D20
ꢃ
+28.0 (c 0.1, D2O); 1H NMR (500 MHz,
4.2.5. [N,N0-Di(b-
D-galactopyranosyl-(1?4)-b-D-glucopyranosyl)-
D2O) data for EE isomer d 7.70 (s, 4H), 4.84 (d, J = 8.8 Hz, 2H, H-1),
4.44 (d, J = 7.8 Hz, 2H, H-10), 4.20 (ABq, J = 16.3 Hz, 4H), 3.99 (d,
J = 10.9 Hz, 2H), 3.92 (d, J = 3.3 Hz, 2H), 3.86–3.63 (m, 14H), 3.60–
3.48 (m, 6H), 1.38 (s, 18H); selected 1H NMR data for EZ isomer d
7.68 (d, J = 7.8 Hz, 2H), 7.62 (d, J = 7.8 Hz, 2H), 5.76 (d, J = 9.3 Hz,
2H, H-1), 4.50 (d, J = 7.8 Hz, 2H, H-10), 1.18 (s, 18H); 13C NMR
(125 MHz, D2O) d 174.1 (C), 169.9 (C), 135.9 (C), 127.5, 102.8,
87.2, 77.5, 76.7, 75.3, 74.0, 72.4, 70.8, 69.7, 68.5 (each CH), 61.0
(CH2), 59.9 (CH2), 51.5 (C), 45.5 (CH2), 27.6 (CH3); HRMS-ESI: calcd
for C44H70N4O24Na: 1061.4278; Found: 1061.4293.
N,N0-di(1-methoxycarbonyl)-methylamino-2-oxoethyl]-
naphthalene-2,6-dicarboxamide 5
Naphthalene-2,6-dicarboxylic acid (18 mg, 0.08 mmol), lactosyl
amine 1(100 mg, 0.16 mmol) and formaldehyde (16 lL of a 37%
solution, 0.19 mmol) were suspended in MeOH (5 mL) and the
mixture was stirred at room temperature for 1 h. Methyl iso-
cyanoacetate (15 lL, 0.24 mmol) was then added and the mixture
was stirred at room temperature overnight. The reaction was then
heated to 45 °C for 12 h. The solvent was removed thereafter under
reduced pressure. Chromatography of the residue (CH2Cl2–CH3OH,
gradient elution, 70:1 to 60:1 to 50:1) led to the protected interme-
diate as a white amorphous solid (89 mg, 66%). Zemplén deacety-
lation and subsequent purification using a BioGel P-2 column
(eluent: H2O) and then C18 reverse-phase column (gradient elu-
tion, H2O to H2O–CH3OH, 95:5) gave 5 as an interconverting mix-
4.2.3. N,N0-Di(b-
D-galactopyranosyl-(1?4)-b-D-glucopyranosyl)-
N,N0-di[(1-methoxycarbonyl)-methylamino-2-
oxoethyl]oxalamide 3
Oxalic acid (7 mg, 0.08 mmol), lactosyl amine 11 (100 mg,
0.16 mmol) and formaldehyde (16
0.19 mmol) were suspended in MeOH (5 mL) and the mixture
was stirred at room temperature for 1 h. Methyl isocyanoacetate
ture of EE and EZ isomers (4:1); ½a D20
ꢃ
+36 (c 0.2, D2O); 1H NMR
lL
of
a
37% solution,
(500 MHz, D2O) data for EE isomer d 8.24 (s, 2H), 8.18 (d,
J = 8.5 Hz, 2H), 7.74 (d, J = 8.6 Hz, 2H), 4.89 (d, J = 8.9 Hz, 2H, H-
1), 4.54–4.36 (m, 6H), 4.17 (s, 4H), 4.01 (d, J = 11.9 Hz, 2H), 3.91
(d, J = 3.0 Hz, 2H), 3.88–3.58 (m, 20H), 3.53–3.38 (m, 6H); selected
1H NMR data for EZ isomer d 5.83 (d, J = 8.6 Hz, 2H); 13C NMR
(125 MHz, D2O) d 175.0 (C), 171.93(C), 171.87 (C), 132.9 (C),
132.4 (C), 129.7, 127.0, 124.7, 102.7, 87.4, 77.5, 76.9, 75.3, 74.1,
72.4, 70.8, 69.7, 68.4 (each CH), 60.9 (CH2), 60.0 (CH2), 52.8
(CH3), 44.7(CH2), 41.3 (CH2); HRMS-ESI: calcd for C46H64N4O28Na:
1143.3605; Found: 1143.3601.
(15 lL, 0.16 mmol) was then added and the mixture was stirred
at room temperature overnight. The reaction was then heated
to 45 °C for 12 h, solvent was removed thereafter under reduced
pressure. Chromatography of the residue (CH2Cl2–CH3OH, gradi-
ent elution, 80:1 to 60:1 to 50:1) gave the protected intermediate
as a white amorphous solid (84 mg, 67%). Zemplén deacetylation
and purification of the product using a BioGel P-2 gel column
(eluant/H2O) followed by C18 reverse-phase column (elution,
H2O) generated 3 as an interconverting mixture of EE and EZ iso-
4.2.6. N,N0-Di(b-
D-galactopyranosyl-(1?4)-b-D-glucopyranosyl)-
mers (3.1:1); ½a D20
ꢃ
+30.8 (c 0.5, D2O); 1H NMR (500 MHz, D2O)
N,N0-di[(1-methoxycarbonyl)-methylamino-2-
oxoethyl]isophthalamide 6
data for EE isomer d 4.98 (d, J = 8.7 Hz, 2H, H-1), 4.47 (d,
J = 7.7 Hz, 2H, H-10), 4.30 (ABq, J = 16.7 Hz, 4H), 4.11–4.07 (m,
4H), 3.98–3.90 (m, 4H), 3.89–3.62 (m, 24H), 3.56 (dd, J = 9.9,
7.7 Hz, 2H, H-20); selected 1H NMR data for EZ isomer d 5.65 (d,
J = 9.4 Hz, 2H); 13C NMR (125 MHz, D2O) d 171.8 (C), 170.5 (C),
165.1 (C), 102.8, 86.4, 77.3, 76.8, 75.3, 74.1, 72.5, 70.9, 69.4,
68.5 (each CH), 61.0 (CH2), 59.8 (CH2), 52.8 (CH3), 43.7 (CH2),
41.2 (CH2); HRMS-ESI: calcd for C36H58N4O28Na: 1017.3135;
Found: 1017.3108.
Isophthalic acid (20 mg, 0.12 mmol), lactosyl amine 121
(152 mg, 0.24 mmol) and formaldehyde (23 lL of a 37% solution,
0.29 mmol) were suspended in MeOH (5 mL) and the mixture
was stirred at room temperature for 1 h. Methyl isocyanoacetate
(23 lL, 0.24 mmol) was then added and the mixture was stirred
at room temperature overnight. The reaction was then heated to
45 °C for 12 h, solvent removed thereafter under reduced pressure.
Chromatography of the residue (CH2Cl2–CH3OH, gradient elution,
70:1 to 60:1 to 50:1) led to the protected intermediate as a white
amorphous solid (129 mg, 65%). Zemplén deacetylation of this
intermediate and purification by BioGel P-2 gel column (eluent/
H2O) and then C18 reverse-phase chromatography (gradient elu-
tion, H2O to H2O–CH3OH, 95:5) generated 2 as an interconverting
4.2.4. N,N0-Di(b-
D-galactopyranosyl-(1?4)-b-D-glucopyranosyl)-
N,N0-di[(t-butylcarbamoyl)-methyl]isophthalamide 4
Isophthalic acid (13.6 mg, 0.08 mmol), lactosyl amine
(100 mg, 0.16 mmol) and formaldehyde (16 lL of a 37% solution,
0.19 mmol) were suspended in MeOH (5 mL) and the mixture
was stirred at room temperature for 1 h. t-Butyl isocyanide
1
mixture of EE and EZ isomers (3.75:1); ½a D20
ꢃ
+27.3 (c 0.6, D2O); 1H
NMR (500 MHz, D2O) data for EE isomer d 7.79 (s, 1H), 7.78 (s,
2H), 7.75–7.71 (m, 1H), 4.83 (d, J = 8.8 Hz, 2H, H-1), 4.43 (d,
J = 7.6 Hz, 2H, H-10), 4.38 (d, J = 11.9 Hz, 4H), 4.14 (d, J = 3.5 Hz,
4H), 3.97 (d, J = 12.7 Hz, 2H), 3.93 (d, J = 3.3 Hz, 2H), 3.86–3.63
(m, 20H), 3.57–3.49 (m, 6H); selected 1H NMR data for EZ isomer
d 5.77 (d, J = 9.5 Hz, 2H), 4.42 (d, J = 7.2 Hz, 2H); 13C NMR
(125 MHz, D2O) d 174.0 (C), 171.8 (C), 134.0 (C), 130.0, 129.7,
102.8, 87.2, 77.5, 76.7, 75.3, 74.1, 72.4, 70.8, 69.7, 68.5 (each CH),
61.0 (CH2), 59.9 (CH2), 52.8(CH3), 44.8 (CH2), 41.3 (CH2); HRMS-
ESI: calcd for C42H62N4O28Na: 1093.3448; Found: 1093.3466.
(19 lL, 0.16 mmol) was then added and the mixture was stirred
at room temperature overnight. The reaction was then heated
to 45 °C for 12 h, followed by removing solvent under reduced
pressure. Chromatography of the residue (CH2Cl2–CH3OH, gradi-
ent elution, 80:1 to 60:1 to 50:1) gave the protected intermediate
as a white amorphous solid (101 mg, 79%). Zemplén deacetylation
and subsequent purification using a BioGel P-2 gel column (elu-
ent/H2O) and C18 reverse-phase column (gradient elution, H2O
to H2O–CH3OH, 95:5) produced 4 as an interconverting mixture
of EE and EZ isomers (4:1); ½a D20
ꢃ
+22.3 (c 0.27, D2O); 1H NMR
(500 MHz, D2O) data for EE isomer d 7.80–7.68 (m, 4H), 4.79 (d,
2H, H-1, determined by 2D-NMR), 4.43 (d, J = 7.8 Hz, 2H, H-10),
4.20 (ABq, J = 16.4 Hz, 4H), 4.04–3.94 (m, 2H), 3.92 (d, J = 3.2 Hz,
2H), 3.91–3.59 (m, 14H), 3.59–3.44 (m, 6H), 1.38 (s, 18H); se-
lected 1H NMR data for EZ isomer d 5.75 (d, J = 9.3 Hz, 2H), 4.32
(d, J = 7.3 Hz, 2H), 1.18 (s, 18H); 13C NMR (125 MHz, D2O) d
174.1 (C), 169.9 (C), 134.1 (C), 130.0, 129.5, 102.8, 87.3, 77.5,
76.7, 75.3, 74.1, 72.4, 70.8, 69.7, 68.5 (each CH), 61.0 (CH2),
59.9 (CH2), 51.5 (C), 45.7 (CH2), 27.6 (CH3); HRMS-ESI: calcd for
4.2.7. O-(2,3,4-Tri-O-benzyl-
tri-O-acetyl-b- -galactopyran osyl)-(1?4)-1,2,3,6-tetra-O-
acetyl-b- -glucopyranosyl amine 18
Compound 16 (450 mg, 0.59 mmol) was dissolved in CH3OH–
H2O (30 mL, 2:1) and treated with an excess of ammonium hydro-
gen carbonate for 7 days at 30 °C. The solution was concentrated to
half its original volume and diluted with water. This procedure was
repeated twice, followed by removing water. Toluene was then
evaporated from the residue (25 mL ꢁ 3 times) and the residue
containing 17 was suspended in pyridine (25 mL), Fmoc-OSu
a-L-fucopyranosyl)-(1?2)-O-(3,4,6-
D
D
C44H70N4O24Na: 1061.4278; Found: 1061.4257.

S. André et al. / Carbohydrate Research 389 (2014) 25–38
35
(211 mg, 0.65 mmol) was added and the mixture was stirred over-
night at ambient temperature. Acetic anhydride (12 mL) was then
added and the mixture was stirred for another 12 h. After concen-
tration and co-evaporation with toluene, the residue was purified
by chromatography (EtOAc- PE, gradient elution, 5:1 to 2:1) to give
a colourless oil. This oil was dissolved in DMF (5 mL) and morpho-
line (5 mL) was added. After 25 min the solution was diluted with
toluene (10 mL) and concentrated. The residue was purified by
chromatography (CH2Cl2–CH3OH, gradient elution, 80:1 to 70:1)
to lead to 18 as a colourless oil (237 mg, 40% for three steps), Rf
0.33 (CH2Cl2–CH3OH, 40:1). 1H NMR (500 MHz, CDCl3) d 7.35–
7.24 (m, 15H), 5.32 (d, J = 2.5 Hz, 1H), 5.22 (d, J = 3.4 Hz, 1H, H-
100), 5.13 (t, J = 9.5 Hz, 1H, H-3), 5.05 (dd, J = 10.0, 3.4 Hz, 1H),
4.97 (d, J = 11.6 Hz, 1H), 4.75 (t, J = 9.5 Hz, 1H, H-2), 4.72–4.63
(m, 5H), 4.49 (dd, J = 1.9, 11.5 Hz, 1H), 4.38 (d, J = 7.6 Hz, 1H, H-
10), 4.18–4.02 (m, 6H), 3.85–3.78 (m, 4H), 3.69 (d, J = 1.9 Hz, 1H),
3.55–3.51 (m, 1H), 2.08–2.07 (4s, 12H), 2.04 (s, 3H), 1.79 (s, 3H),
1.19 (d, J = 6.4 Hz, 3H); 13C NMR (125 MHz, CDCl3): 170.4, 170.34,
170.30, 170.0, 169.9, 169.8, 138.7, 138.6, 138.4 (each C), 128.4,
128.3, 128.2, 127.7, 127.6, 127.54, 127.50, 127.3, 100.9, 97.9,
84.9, 79.4, 77.7, 76.3 (each CH), 74.8 (CH2), 74.0 (CH), 73.8 (CH),
73.4 (CH2), 73.1 (CH), 73.0 (CH), 72.9 (CH2), 72.6 (CH), 72.1 (CH),
70.5(CH), 67.22 (CH), 67.20 (CH), 62.5 (CH2), 61.0 (CH2), 20.91,
20.88, 20.8, 20.7, 20.6, 16.5 (each CH3). HRMS-ESI: calcd for C51H64-
NO20Na: 1032.3841; Found: 1032.3837
CuSO4 (2.4 mg dissolved in 1 mL H2O, 14.6
l
mol) were subse-
quently added and the mixture was stirred overnight, after which
the solvent was removed and the residue was precipitated by CH2-
Cl2 (50 mL) and water (15 mL). The organic phase was washed by
water (15 mL ꢁ 2), dried by Na2SO4 and concentrated. The crude
residue was purified by flash chromatography (EtOAc-PE, gradient
elution, 2:1 to 2.5:1) to give a white foam (122 mg, 88%); The pro-
tecting groups were removed from the peracetylated intermediate
(41 mg, 0.021 mmol) by the Zemplén procedure to give 8 as a
white amorphous solid (17 mg, 68%) after preparative reverse-
phase HPLC (isocratic elution with water–CH3CN, 91:9, flow rate
10 mL/min) and lyophilization; ½a D20
ꢃ
ꢂ56.0 (c 0.1, D2O); 1H NMR
(500 MHz, D2O) d 8.60 (s, 2H), 7.87 (s, 4H), 5.81 (d, J = 9.1 Hz, 2H,
H-1), 5.35 (d, J = 1.9 Hz, 2H, H-100), 4.61 (d, J = 7.6 Hz, 2H, H-10),
4.27 (dd, J = 12.6, 6.0 Hz, 2H, H-500), 4.13 (t, J = 9.2 Hz, 2H, H-2),
4.04–4.00 (m, 4H), 3.95–3.70 (m, 24H), 1.30 (d, J = 6.3 Hz, 6H);
13C NMR (125 MHz, D2O) d 147.1 (C), 129.4 (C), 126.3, 121.3,
100.3, 99.4, 87.4, 78.1, 76.4, 75.2, 74.9, 74.4, 73.5, 72.1, 71.6,
69.6, 69.1, 68.2, 66.9 (each CH), 61.1 (CH2), 59.8 (CH2), 15.3
(CH3); HRMS-ESI: calcd for C46H68N6O28Na: 1175.3979; Found:
1175.3937.
4.2.10. 1,4-Di[(
(1?4)-b- -glucopyranosyl)-1,2,3-triazol-4-
ylmethyloxy]benzene 9
a-L-fucopyranosyl-(1?2)-b-D-galactopyranosyl-
D
4.2.8. N,N0-Di[
(1?4)-b-
-glucopyranosyl]-N,N0-di[(1-
methoxycarbonyl)methylamino-2-oxoethyl]terephthalamide 7
Terephthalic acid (5 mg, 32.2 mol), formaldehyde (37%, 6 L,
77.3 mol) and 18 (65 mg, 64.4 mol) were suspended in anhy-
drous methanol. After 1 h, the methyl isocyanoacetate (9 L,
96.6 mol) was added and the reaction was allowed to stirred at
a-L-fucopyranosyl-(1?2)-b-D-galactopyranosyl-
Compound 12 (90 mg, 0.10 mmol) was dissolved in CH3OH–
H2O (2:1, 6 mL), then p-bispropargyloxybenzene22 15 (9.4 mg,
0.05 mmol), sodium ascorbate (4 mg dissolved in 1 mL H2O,
D
20 lmol) and CuSO4 (1.6 mg dissolved in 1 mL H2O, 10 lmol) were
l
l
subsequently added and the mixture was stirred overnight, after
which the solvent was removed and the residue was participated
by CH2Cl2 (50 mL) and water (15 mL). The organic phase was
washed by water (15 mL ꢁ 2), dried by Na2SO4 and concentrated.
Flash silica gel chromatography (EtOAc-PE, gradient elution, 2:1
to 2.5:1) gave the dimeric intermediate as a white foam (91 mg,
l
l
l
l
rt for 4 h and 45 °C for 24 h. Then solvent was removed under re-
duced pressure. Chromatography of the residue (CH2Cl2–CH3OH,
gradient elution, 100:1 to 70:1 to 55:1) gave the protected inter-
mediate as a white amorphous solid (19 mg, 24%). Zemplén deacet-
ylation and subsequent C18 reverse-phase chromatography (H2O–
CH3OH, gradient elution, 1:1 to 1:2 to 1:3 to 1:4) were carried out
to produce the benzylated intermediate. Then the residue was dis-
solved in methanol, to which 10% Pd-C was added. The mixture
was stirred under an atmosphere of hydrogen for 24 h at ambient
temperature. When the reaction was completed, the mixture was
filtered over celite and concentrated. Reverse-phase chromatogra-
phy using a C-18 column (H2O–CH3OH, gradient elution, 1:0 to
98:2 to 97:3) gave 7 as an amorphous solid as an interconverting
92%), Rf 0.50 (PE-EtOAc, 1:4); ½a D20
ꢃ
ꢂ75.0 (c 1.0, CHCl3); 1H NMR
(500 MHz, CDCl3) d 7.80 (s, 2H), 6.92 (s, 4H), 5.84 (d, J = 9.3 Hz,
2H, H-1), 5.45 (t, J = 9.5 Hz, 2H), 5.40 (d, J = 3.8 Hz, 2H, H-100),
5.37–5.30 (m, 6H), 5.18–5.15 (m, 6H), 5.02–4.97 (m, 4H), 4.51 (d,
J = 12.2 Hz, 2H), 4.46 (d, J = 7.6 H, 2H, H-10), 4.41 (q, J = 6.5 Hz,
2H), 4.32 (dd, J = 12.2, 5.5 Hz, 2H), 4.17 (dd, J = 11.2, 6.5 Hz, 2H),
4.10 (dd, J = 11.2, 7.0 Hz, 2H), 4.03–3.97 (m, 4H), 3.92–3.84 (m,
4H), 2.17 (s, 6H), 2.13 (2s, 12H), 2.10 (s, 6H), 2.08 (s,6H), 2.00 (s,
6H), 1.99 (s, 6H), 1.97 (s, 6H), 1.87 (s, 6H), 1.24 (d, J = 6.5 H, 6H);
13C NMR (125 MHz, CDCl3): d 170.7, 170.6, 170.5, 170.3, 170.1,
169.9, 169.8, 169.7, 168.9, 152.8, 145.0 (each C), 121.2, 115.9,
100.2, 95.6, 85.8, 76.3, 73.8, 73.4, 71.8, 71.5, 71.0, 70.9, 70.2, 68.0,
67.3, 67.0, 65.0 (each CH), 62.6, 62.1, 60.8 (each CH2), 20.8, 20.70,
20.68, 20.66, 20.65, 20.63, 20.60, 20.2, 15.6 (each CH3); LRMS
(ESI) 1991.5 (M+Na+); HRMS-ESI: calcd for C84H108N6O48 Na:
1991.6092; Found: 1991.6145. This intermediate (37 mg,
0.019 mmol) was dissolved in methanol (5 mL) to which a catalytic
amount of NaOMe (0.1 mL of a 0.2 M solution in MeOH) was added
and the resulting solution was stirred for 1 h at room temperature.
Amberlite IR-120 (plus) was added to neutralize pH = 7, after
which the resin was removed by filtration and washed with water.
The solvent removed under diminished pressure to give 9 after
mixture of EE and EZ isomers (3:1); ½a D20
ꢃ
ꢂ28.0 (c 0.1, D2O); 1H
NMR (500 MHz, D2O) data for EE isomer d 7.73 (s, 4H), 5.27 (d,
J = 3.4 Hz, 2H, H-100), 4.74 (d, J = 8.9 Hz, 2H, H-1), 4.51 (d,
J = 7.8 Hz, 2H, H-10), 4.39 (ABq, J = 16.6 Hz, 4H), 4.14 (m, 6H), 3.98
(d, J = 11.6 Hz, 2H, H-6a), 3.93–3.68 (m, 28H), 3.64 (J = 8.5 Hz,
2H), 3.47 (t, J = 9.1 Hz, 2H), 3.30–3.28 (m, 2H), 1.08 (d, J = 6.6 Hz,
6H); selected 1H NMR data for EZ isomer d 7.70 (d, J = 7.5 Hz,
2H), 7.63 (d, J = 7.5 Hz, 2H), 5.74 (d, J = 7.7 Hz, 2H, H-1), 5.35 (bs,
2H, H-100), 4.57 (d, J = 7.5 Hz, 2H, H-10); 13C NMR (125 MHz, D2O)
d 174.0 (C), 171.8 (C), 171.7 (C), 136.0 (C), 127.5, 100.1, 99.5,
87.6, 77.5, 76.6, 75.2, 75.1, 74.2, 73.4, 71.6, 69.7, 69.5, 69.0, 68.2,
66.8 (each CH), 61.1(CH2), 60.1(CH2), 52.8 (CH3), 44.7 (CH2), 41.3
(CH2), 15.3 (CH3); HRMS-ESI: calcd for C54H82N4O36Na:
1385.4606; Found: 1385.4608.
lyophilization, as an amorphous solid (20 mg, 88%). ½a D20
ꢂ63.3 (c
ꢃ
0.12, D2O); 1H NMR (500 MHz, D2O) d 8.32 (s, 2H), 7.06 (s, 4H),
5.79 (d, J = 9.2 Hz, 2H, H-1), 5.35 (d, J = 3.2 Hz, 2H, H-100), 5.27 (s,
4H), 4.60 (d, J = 7.7 Hz, 2H, H-10), 4.26 (dd, J = 13.1, 6.5 Hz, 2H, H-
500), 4.08 (t, J = 9.3 Hz, 2H), 4.00–3.97 (m, 4H), 3.92–3.89 (m, 4H),
3.87–3.71 (m, 20H), 1.28 (d, J = 6.6 Hz, 6H); 13C NMR (125 MHz,
D2O) d 152.2 (C), 143.7 (C), 124.4, 116.9, 100.2, 99.3, 87.3, 78.1,
76.3, 75.2, 74.9, 74.4, 73.5, 72.0, 71.6, 69.6, 69.1, 68.1, 66.9 (each
4.2.9. 1,4-Di[
(1?4)-b- -glucopyranosyl-1,2,3-triazol-4-yl]benzene 8
Compound 12 (130 mg, 145.9 mol) was dissolved in CH3OH-
H2O (2:1, 12 mL), then 1,4-diethynyl benzene (9.6 mg, 73.0 mol),
sodium ascorbate (5.8 mg dissolved in 1 mL H2O, 29.2 mol) and
a-L-fucopyranosyl-(1?2)-b-D-galactopyranosyl-
D
l
l

36
S. André et al. / Carbohydrate Research 389 (2014) 25–38
CH), 62.0, 61.1, 59.7 (each CH2), 15.3 (CH3); HRMS-ESI: calcd for
48H72N6O30 Na: 1235.4191; Found: 1235.4204.
4.58 (d, J = 7.7 Hz, 2H, H-10), 4.28 (dd, J = 13.0, 6.4 Hz, 2H), 3.99
(d, J = 11.5 Hz, 2H), 3.91–3.63 (m 28H), 1.29 (d, J = 6.5 Hz, 6H);
13C NMR (125 MHz, D2O) d 171.0 (C), 136.4 (C), 127.9, 100.2,
99.3, 79.8, 77.0, 76.2, 75.2, 75.1, 73.5, 71.6, 71.5, 69.6, 69.1, 68.1,
66.9 (each CH), 61.1(CH2), 59.9 (CH2), 15.2 (CH3); HRMS-ESI: calcd
for C44H68N2O30 Na: 1127.3755; Found: 1127.3728.
C
4.2.11. O-(2,3,4-Tri-O-acetyl-
tri-O-acetyl-b- -galactopyranosyl)-(1?4)-1,2,3,6-tetra-O-acetyl-
-glucopyranosyl amine 13
The azide 12 (1.2 g, 1.35 mmol) was dissolved in EtOAc to which
a-L-fucopyranosyl)-(1?2)-O-(3,4,6-
D
D
was added 10% Pd-C (0.1 g). The reaction was left to stir overnight
under H2. Then the reaction was diluted with EtOAc and filtered
through Celite. Removal of the solvent gave 13 as a white foam
4.3. Protein engineering, production and labelling
Starting with full-length cDNA for human Gal-4 cloned from
mRNA preparation of human leukaemic KG-1 cells, the BamHI
restriction site (position 769 to 774) was silenced without affecting
coding using the sense primer 50-CTGAATGGCTCGTGGGG
CTCAGAGGAGAAGAAGATCACC-30 (nucleotide exchanges under-
lined; melting temperature >78 °C) and the antisense primer 50-
GGTGATCTTCTTCTCCTCTGAGCCCCACGAGCCATTCAG-30 (nucleo-
tide exchanges underlined; melting temperature >78 °C) in a
modified QuikChangeÒ site-directed mutagenesis procedure
(Agilent Technologies, Böblingen, Germany). The extension
(1.16 g, >95% yield, mixture of anomers, b:a = 10:1) which was
used in the next step without further purification; Rf 0.24,
(EtOAc–CH2Cl2, 9:1); 1H NMR (500 MHz, CDCl3) data for the b-ano-
mer: d 5.38 (d, J = 3.9 Hz, 1H), 5.29 (dd, J = 10.3, 2.8 Hz, 2H), 5.18–
5.13 (m, 2H), 5.00–4.96 (m, 2H), 4.78 (t, J = 9.4 Hz, 1H), 4.48 (dd,
J = 11.9, 1.8 Hz, 1H), 4.45–4.40 (m, 2H), 4.24 (dd, J = 12.0, 6.1 Hz,
1H), 4.17-4.11 (m, 3H), 4.10–4.06 (m, 1H), 3.86–3.80 (m, 2H),
3.78 (t, J = 9.5 Hz, 1H), 3.64–3.61 (m, 1H), 2.16 (s, 3H), 2.15 (s,
3H), 2.12 (s, 3H), 2.08 (s, 3H), 2.07 (s, 3H), 2.06 (s, 3H), 2.00 (s,
3H), 1.98 (, 3H), 1.97 (s, 3H), 1.22 (d, J = 6.5 H, 3H). Selected 1H
reaction was performed in two steps. Briefly, two 50 lL reaction
NMR data for the
a
anomer: d 5.35 (d, J = 3.0 Hz, 1H), 5.23 (t,
mixtures were prepared in separate tubes containing 20 pmol
either of the sense or the antisense primer, 200 ng template plas-
mid and 1 U PfuTurboÒ DNA polymerase (Agilent Technologies).
After an initial preheating step at 95 °C for 30 s, three cycles (dena-
turation at 95 °C for 30 s, annealing at 55 °C for 1 min, extension at
68 °C for 8 min) were run. To complete the primer-directed
J = 9.4 Hz, 1H), 4.73 (t, J = 9.4 Hz, 1H), 4.28 (dd, J = 12.0, 5.5 Hz,
1H); 13C NMR (125 MHz, CDCl) for the b-anomer: d 170.74,
170.70, 170.6, 170.34, 170.30, 170.1, 170.03, 170.01, 169.67 (each
C), 100.2, 95.5, 85.0, 74.7, 74.0, 73.5, 72.4, 72.0, 71.4, 71.1, 70.7,
68.1, 67.5, 67.1, 64.8 (each CH), 62.7 (CH2), 60.9 (CH2), 20.92,
20.86, 20.8, 20.67, 20.66, 20.65, 20.63, 20.60, 20.59, 15.53 (each
CH3); LRMS (ESI) 866.2 [M+H]+; HRMS-ESI: calcd for
sequence extension 25 lL of each tube were transferred to one
tube and 1 U PfuTurboÒ DNA polymerase was added. Subse-
quently, thermal cycling which consisted of preheating at 94 °C
for 30 s and 18 cycles (denaturation at 94 °C for 30 s, annealing
at 55 °C for 1 min, extension at 68 °C for 5 min) was carried out.
After incubation in the presence of DpnI (10 U) at 37 °C for 2 h to
C36H53N1O23: 866.2921; Found: 866.2930.
4.2.12. N,N0-Di(
(1?4)-b- -glucopyranosyl) terephthalamide 10
a-L-fucopyranosyl-(1?2)-b-D-galactopyranosyl-
D
The lactosyl amine 13 (112 mg, 0.13 mmol) and DIPEA (0.16 mL,
0.19 mmol) in dry THF (10 mL) were added dropwise at room tem-
perature into freshly recrystallized terephthaloyl chloride (26 mg,
0.065 mmol) in dry THF. The reaction was stirred overnight and
then the solvent was removed. Chromatography of the residue
(EtOAc-PE, gradient elution, 1:1 to 2:1) gave the intermediate as
digest the methylated parental DNA template, 5 lL of the reaction
mixture were used to transform XL-1-Blue electrocompetent cells.
Plasmids were isolated from kanamycin-resistant colonies grown
on LB agar plates and treated with BamHI to ascertain the sequence
conversion. Then, the cDNAs of the N-domain extended by a nucle-
otide sequence encoding for a tetrapeptide stretch at its C-termi-
nus and of the C-domain with a 13mer-peptide stretch of the
linker were first amplified separately. Creating a new joining hinge
of the shortened cDNA using two BamHI restriction sites has the
advantage that the wild-type codons, at these positions encoding
Gly and Ser, were preserved. In detail, the following primer pairs
were used: the sense primer 50-CATATGGCCTATGTCCCCGCACCG-30
with an internal NdeI restriction site (underlined) and the anti-
a white amorphous solid (105 mg, 87% including trace of
a-ano-
mer); Rf 0.75 (PE-EtOAc, 1:4); ½a D20
ꢃ
ꢂ77.2 (c 1.1, CHCl3); 1H NMR
(500 MHz, CDCl3) d 7.82 (s, 4H), 7.03 (d, J = 9.0 Hz, 2H), 5.40–5.34
(m, 6H), 5.34 (d, J = 2.6 Hz, 2H), 5.30 (d, J = 3.6 Hz, 2H), 5.19 (dd,
J = 11.0, 3.2 Hz, 2H), 5.01–4.98 (m, 6H), 4.51–4.41 (m, 6H), 4.32–
4.29 (m, 2H), 4.18 (dd, J = 11.1, 6.6 Hz, 2H), 4.08 (dd, J = 11.1,
6.8 Hz, 2H), 3.89–3.84 (m, 8H), 2.17 (s, 6H), 2.14 (s, 6H), 2.13 (s,
6H), 2.11 (s, 6H), 2.10 (s, 6H), 2.05 (s, 6H), 2.00 (s, 6H), 1.984 (s,
6H), 1.976 (s, 6H), 1.25 (d, J = 6.7 Hz, 6H); 13C NMR (125 MHz,
CDCl3) d 171.7, 170.7, 170.6, 170.5, 170.3, 170.2, 170.0, 169.7,
169.6, 165.9, 136.2 (each C), 127.7, 99.9, 95.7, 79.0, 74.9, 74.1,
73.4, 71.7, 71.4, 71.1, 71.0, 70.8, 68.1, 67.4, 67.0, 65.0 (each CH),
62.3 (CH2), 61.0 (CH2), 20.9, 20.78, 20.76, 20.69, 20.67, 20.66,
20.64, 20.60, 15.6 (each CH3); LRMS (ESI) 1883.4 [M+Na]+;
´
´
sense primer 5-GGATCCTCCGATGAAGTTGATTGATTGAAGTTG-3
with an internal BamHI restriction site (underlined) for the
N-domain with the nucleotide sequence for the tetrapeptide
stretch of the linker and the sense primer 50-GGATCCCTGCCCACC
ATGGAAGGA-30 with an internal BamHI restriction site (under-
lined) and the antisense primer 50-GTCGACTTAGATCTGGACAT
AGG-30 with an internal SalI restriction site (underlined) in the
cDNA encoding for the 13mer-peptide stretch of the linker
followed by the sequence of the C-domain. The cDNAs were then
propagated in the pET-Blue-1 AccepTor vector (Novagen, Bad
Soden, Germany), digestion with the restriction enzymes and gel
extraction led to two vector-released cDNAs, which were ligated
into the pET24a expression vector (Novagen) yielding a 897 bp
insert encoding the Gal-4V sequence.cDNA for Gal-4P was
prepared by first amplifying the cDNAs of N- and C-domains
separately and reverting the artificial ClaI restriction site at the
new joining hinge to the wild-type codons at these positions. In
detail, we used the following primer pairs: the sense primer
50-CATATGGCCTATGTCCCCGCACCG-30 with an internal NdeI
HRMS-ESI: calcd for
1883.5609. This intermediate (40 mg, 21.5
minor -anomer) was dissolved in methanol (5 mL) to which a cat-
C
80H104N2O48 Na: 1883.5656; Found:
lmol, including the
a
alytic amount of NaOMe (0.1 mL of a 0.2 M solution in MeOH) was
added and the resulting solution was stirred for 1 h at room tem-
perature. Amberlite IR-120 (plus) was added to neutralize pH = 7,
after which the resin was removed by filtration and washed with
water. The solvent removed on a rotary evaporator to generate
10 as a yellow solid (23 mg, 97%). The compound was further puri-
fied by reverse-phase semi-preparative HPLC (isocratic elution
with water–CH3CN, 97:3, flow rate 10 mL/min, tR = 22 min) and
the product was obtained after lyophilization as a white solid;
a 2D0
ꢃ
ꢂ41.5 (c 0.325, D2O); 1H NMR (500 MHz, D2O) d 7.96 (s,
restriction site (underlined) and the antisense primer 5-ATCGAT
´
½
4H), 5.35 (d, J = 2.7 Hz, 2H, H-10’), 5.23 (d, J = 9.1 Hz, 2H, H-1),
GATTGATTGAAGTTGCAGATCCCC-3 with an internal ClaI restriction
´

S. André et al. / Carbohydrate Research 389 (2014) 25–38
37
M.; Gress, T. M.; Nishimura, S.-I.; Rosewicz, S.; Gabius, H.-J. FEBS J. 2007, 274,
3233–3256; (c) Sanchez-Ruderisch, H.; Fischer, C.; Detjen, K. M.; Welzel, M.;
Wimmel, A.; Manning, J. C.; André, S.; Gabius, H.-J. FEBS J. 2010, 277, 3552–
3563; (d) Amano, M.; Eriksson, H.; Manning, J. C.; Detjen, K. M.; André, S.;
Nishimura, S.-I.; Lehtiö, J.; Gabius, H.-J. FEBS J. 2012, 279, 4062–4080; (e)
Gebert, J.; Kloor, M.; Lee, J.; Lohr, M.; André, S.; Wagner, R.; Kopitz, J.; Gabius,
H.-J. Histochem. Cell Biol. 2012, 138, 339–350; (f) Dawson, H.; André, S.;
Karamitopoulou, E.; Zlobec, I.; Gabius, H.-J. Anticancer Res. 2013, 33, 3053–
3059; (g) Scott, D. W.; Patel, R. P. Glycobiology 2013, 23, 622–633.
site (underlined) for the N-domain as well as the sense primer
50-ATCGATGTGCCATATTTCGGGAGG-30 with an internal ClaI
restriction site (underlined) and the antisense primer 50-GTCGA
CTTAGATCTGGACATAGG-30 with an internal SalI restriction site
(underlined) for the C-domain. The cDNAs were then propagated
as described above, digestion with restriction enzymes (NdeI/ClaI,
ClaI/SalI) followed by elution of the 846 bp insert encoding Gal-4P
from the gel. The inserted ClaI restriction site between the N- and
C-domains encoding for Ile and Asp finally needed to be reversed to
the wild-type codons, at these positions encoding for Asn and Pro,
in a modified QuikChangeÒ site-directed mutagenesis procedure
(Agilent Technologies) using a sense primer 50-CTGCAACTTCAATC
AATCAACCCTGTGCCATATTTCGGGAGGCTG-30 (nucleotide exchanges
underlined) and the antisense primer 50-CAGGGTCCCGAAATAT
GGCACAGGGTTGATTGATTGAAGTTGCAG-30 (nucleotide exchanges
underlined). Plasmids were isolated from kanamycin-resistant colo-
nies grown on LB agar plates and sequencing ascertained absence of
any deviations. Proteins were produced in the BL21(DE3) pLysS
E. coli strain with TB medium (Roth, Karlsruhe, Germany) at 22 °C
with optimal yields of 9 mg/L for Gal-4V and 3 mg/L for Gal-4P
3. (a) Neoglycoconjugates. Preparation and Applications; Lee, Y. C., Lee, R. T., Eds.;
Academic Press: San Diego, 1994; (b) Roy, R. Trends Glycosci. Glycotechnol. 2003,
15, 291–310; (c) Chabre, Y. M.; Roy, R. The Chemist’s Way to Prepare
Multivalency. In The Sugar Code. Fundamentals of glycosciences; Gabius, H.-J.,
Ed.; Wiley-VCH: Weinheim, 2009; pp 53–70; (d) Oscarson, S. The Chemist’s
Way to Synthesize Glycosides. In The Sugar Code. Fundamentals of glycosciences;
Gabius, H.-J., Ed.; Wiley-VCH: Weinheim, 2009; pp 31–51; (e) Murphy, P. V.;
André, S.; Gabius, H.-J. Molecules 2013, 18, 4026–4053.
4. (a) Leyden, R.; Velasco-Torrijos, T.; André, S.; Gouin, S.; Gabius, H.-J.; Murphy, P.
V. J. Org. Chem. 2009, 74, 9010–9026; (b) André, S.; Velasco-Torrijos, T.; Leyden,
R.; Gouin, S.; Tosin, M.; Murphy, P. V.; Gabius, H.-J. Org. Biomol. Chem. 2009, 7,
4715–4725; (c) André, S.; Jarikote, D. V.; Yan, D.; Vincenz, L.; Wang, G.-N.;
Kaltner, H.; Murphy, P. V.; Gabius, H.-J. Bioorg. Med. Chem. Lett. 2012, 22, 313–
318; (d) Wang, G.-N.; André, S.; Gabius, H.-J.; Murphy, P. V. Org. Biomol. Chem.
2012, 10, 6893–6907.
5. (a) Wu, A. M.; Wu, J. H.; Liu, J.-H.; Singh, T.; André, S.; Kaltner, H.; Gabius, H.-J.
Biochimie 2004, 86, 317–326; (b) Krzeminski, M.; Singh, T.; André, S.; Lensch,
M.; Wu, A. M.; Bonvin, A. M. J. J.; Gabius, H.-J. Biochim. Biophys. Acta 1810, 2011,
150–161; (c) Vokhmyanina, O. A.; Rapoport, E. M.; André, S.; Severov, V. V.;
Ryzhov, I.; Pazynina, G. V.; Korchagina, E.; Gabius, H.-J.; Bovin, N. V.
Glycobiology 2012, 22, 1207–1217.
6. (a) Kasai, K.-i.; Hirabayashi, J. J. Biochem. 1996, 119, 1–8; (b) Cooper, D. N. W.
Biochim. Biophys. Acta 2002, 1572, 209–231; (c) Kaltner, H.; Gabius, H.-J. Histol.
Histopathol. 2012, 27, 397–416.
reached with a final IPTG concentration of 75 lM. Recombinant pro-
duction of the two CG-3 proteins followed the previously described
protocol.10a Quality controls by one- and two-dimensional electro-
phoresis, gel filtration and haemagglutination to ascertain homoge-
neity, quaternary structure and activity as well as biotinylation
under activity-preserving conditions followed by product analysis
to quantify label incorporation and activity assessment were
performed as described.23
7. (a) Delacour, D.; Gouyer, V.; Zanetta, J.-P.; Drobecq, H.; Leteurtre, E.; Grard, G.;
Moreau-Hannedouche, O.; Maes, E.; Pons, A.; André, S.; Le Bivic, A.; Gabius, H.-
J.; Manninen, A.; Simons, K.; Huet, G. J. Cell Biol. 2005, 169, 491–501; (b)
Stechly, L.; Morelle, W.; Dessein, A. F.; André, S.; Grard, G.; Trinel, D.; Dejonghe,
M. J.; Leteurtre, E.; Drobecq, H.; Trugnan, G.; Gabius, H.-J.; Huet, G. Traffic 2009,
10, 438–450; (c) Stancic, M.; Slijepcevic, D.; Nomden, A.; Vos, M. J.; de Jonge, J.
C.; Sikkema, A. H.; Gabius, H.-J.; Hoekstra, D.; Baron, W. Glia 2012, 60, 919–935;
(d) Velasco, S.; Díez-Revuelta, N.; Hernández-Iglesias, T.; Kaltner, H.; André, S.;
Gabius, H.-J.; Abad-Rodríguez, J. J. Neurochem. 2013, 125, 49–62.
4.4. Solid-phase/cell assays
The plastic surface of microtiter plate wells was coated with the
8. (a) André, S.; Specker, D.; Bovin, N. V.; Lensch, M.; Kaltner, H.; Gabius, H.-J.;
Wittmann, V. Bioconjug. Chem. 2009, 20, 1716–1728; (b) André, S.; Lahmann,
M.; Gabius, H.-J.; Oscarson, S. Mol. Pharm. 2010, 7, 2270–2279; (c) André, S.;
Grandjean, C.; Gautier, F. M.; Bernardi, S.; Sansone, F.; Gabius, H.-J.; Ungaro, R.
Chem. Commun. 2011, 6126–6128; (d) André, S.; Renaudet, O.; Bossu, I.; Dumy,
P.; Gabius, H.-J. J. Pept. Sci. 2011, 17, 427–437.
9. (a) Beyer, E. C.; Zweig, S. E.; Barondes, S. H. J. Biol. Chem. 1980, 255, 4236–4239;
(b) Gupta, D.; Kaltner, H.; Dong, X.; Gabius, H.-J.; Brewer, C. F. Glycobiology
1996, 6, 843–849; (c) Solís, D.; Romero, A.; Kaltner, H.; Gabius, H.-J.; Díaz-
Mauriño, T. J. Biol. Chem. 1996, 271, 12744–12748; (d) Varela, P. F.; Solís, D.;
Díaz-Mauriño, T.; Kaltner, H.; Gabius, H.-J.; Romero, A. J. Mol. Biol. 1999, 294,
537–549; (e) Wu, A. M.; Wu, J. H.; Tsai, M.-S.; Kaltner, H.; Gabius, H.-J. Biochem.
J. 2001, 358, 529–538; (f) André, S.; Kaltner, H.; Lensch, M.; Russwurm, R.;
Siebert, H.-C.; Fallsehr, C.; Tajkhorshid, E.; Heck, A. J. R.; von Knebel Doeberitz,
M.; Gabius, H.-J.; Kopitz, J. Int. J. Cancer 2005, 114, 46–57; (g) Göhler, A.;
Büchner, C.; André, S.; Doose, S. H.; Kaltner, H.; Gabius, H.-J. Biochimie 2012, 94,
2649–2655.
glycoprotein asialofetuin (0.5 lg per well in 50 lL phosphate-buf-
fered saline) overnight at 4 °C, and further processing followed the
protocol given in detail previously.4 Assays were routinely done in
triplicates with up to six independent series. Cell binding was mea-
sured for human pancreatic carcinoma (Capan-1) cells expressing
the tumour suppressor p16INK4a (kindly provided by K. M. Detjen,
Berlin, Germany)2b as well as with the wild-type (parental)/Lec2
mutant lines of the CHO system (kindly provided by P. Stanley,
Bronx, USA).24 Process steps and controls were performed as de-
scribed previously.2b,4,25 Aliquots of cell suspensions at the same
passage were routinely processed at least in duplicates, with at
least three independent series and standard deviations not exceed-
ing 13.6% after normalization of data based on internal controls
and background values.
10. (a) Kaltner, H.; Kübler, D.; López-Merino, L.; Lohr, M.; Manning, J. C.; Lensch,
M.; Seidler, J.; Lehmann, W. D.; André, S.; Solís, D.; Gabius, H.-J. Anat. Rec. 2011,
294, 427–444; (b) Yongye, A. B.; Calle, L.; Ardá, A.; Jiménez-Barbero, J.; André,
S.; Gabius, H.-J.; Martínez-Mayorga, K.; Cudic, M. Biochemistry 2012, 51, 7278–
7289.
Acknowledgements
11. (a) Bradley, H.; Fitzpatrick, G.; Glass, W. K.; Kunz, H.; Murphy, P. V. Org. Lett.
2001, 3, 2629–2632; (b) Murphy, P. V.; Bradley, H.; Tosin, M.; Pitt, N.;
Fitzpatrick, G. M.; Glass, W. K. J. Org. Chem. 2003, 68, 5693–5704.
12. (a) Campo, V. L.; Carvalho, I.; Da Silva, C. H. T. P.; Schenkman, S.; Hill, L.;
Nepogodiev, S. A.; Field, R. A. Chem. Sci. 2010, 1, 507–514; (b) Meldal, M.;
Tornoe, C. W. Chem. Rev. 2008, 108, 2952–3015; (c) Morvan, F.; Meyer, A.;
Jochum, A.; Sabin, C.; Chevolot, Y.; Imberty, A.; Praly, J. P.; Vasseur, J. J.;
Souteyrand, E.; Vidal, S. Bioconjugate Chem. 2007, 18, 1637–1643; (d)
Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew. Chem., Int.
Ed. 2002, 41, 2596–2599.
The generous financial support by the Science Foundation
Ireland (08/SRC/B1393), the EC research/ITN programs GlycoHIT
(contract no. 260600), COST (CM1102) and GLYCOPHARM
(contract no. 217297) and the Verein zur Förderung des
biologisch-technologischen Fortschritts in der Medizin e.V. is
gratefully acknowledged. We are thankful to Drs. B. Friday and
B. Tab for helpful discussions.
13. Tornøe, C. W.; Christensen, C.; Meldal, M. J. Org. Chem. 2002, 67, 3057–3064.
References
ˇ
14. (a) Lee, Y. C. J. Biochem. 1997, 121, 818–825; (b) André, S.; Kozár, T.; Schuberth,
R.; Unverzagt, C.; Kojima, S.; Gabius, H.-J. Biochemistry 2007, 46, 6984–6995;
ˇ
(c) André, S.; Kozár, T.; Kojima, S.; Unverzagt, C.; Gabius, H.-J. Biol. Chem. 2009,
390, 557–565.
1. (a) Sharon, N. Complex Carbohydrates. Their Chemistry, Biosynthesis, and
Functions; Addison-Wesley: London, 1975; (b) Rauvala, H. Trends Biochem.
Sci. 1983, 8, 323–325; (c) The Sugar Code. Fundamentals of glycosciences;
Gabius, H.-J., Ed.; Wiley-VCH: Weinheim, 2009; (d) Gabius, H.-J.; André, S.;
Jiménez-Barbero, J.; Romero, A.; Solís, D. Trends Biochem. Sci. 2011, 36, 298–
313.
2. (a) Sperandio, M. FEBS J. 2006, 273, 4377–4389; (b) André, S.; Sanchez-
Ruderisch, H.; Nakagawa, H.; Buchholz, M.; Kopitz, J.; Forberich, P.; Kemmner,
W.; Böck, C.; Deguchi, K.; Detjen, K. M.; Wiedenmann, B.; von Knebel Doeberitz,
15. (a) Ideo, H.; Seko, A.; Yamashita, K. J. Biol. Chem. 2005, 280, 4730–4737; (b)
Morelle, W.; Stechly, L.; André, S.; van Seuningen, I.; Porchet, N.; Gabius, H.-J.;
Michalski, J.-C.; Huet, G. Biol. Chem. 2009, 390, 529–544.
16. Dam, T. K.; Gabius, H.-J.; André, S.; Kaltner, H.; Lensch, M.; Brewer, C. F.
Biochemistry 2005, 44, 12564–12571.
17. Kopitz, J.; Ballikaya, S.; André, S.; Gabius, H.-J. Neurochem. Res. 2012, 37, 1267–
1276.

38
S. André et al. / Carbohydrate Research 389 (2014) 25–38
18. Göhler, A.; André, S.; Kaltner, H.; Sauer, M.; Gabius, H.-J.; Doose, S. Biophys. J.
2010, 98, 3044–3053.
23. (a) Gabius, H.-J.; Engelhardt, R.; Cramer, F.; Bätge, R.; Nagel, G. A. Cancer Res.
1985, 45, 253–257; (b) Gabius, H.-J.; Wosgien, B.; Hendrys, M.; Bardosi, A.
Histochemistry 1991, 95, 269–277; (c) Purkrábková, T.; Smetana, K., Jr.;
Dvoránková, B.; Holíková, Z.; Böck, C.; Lensch, M.; André, S.; Pytlík, R.; Liu,
F.-T.; Klíma, J.; Smetana, K.; Motlík, J.; Gabius, H.-J. Biol. Cell 2003, 95, 535–545;
(d) Habermann, F. A.; André, S.; Kaltner, H.; Kübler, D.; Sinowatz, F.; Gabius, H.-
J. Histochem. Cell Biol. 2011, 135, 539–552; (e) Martín-Santamaría, S.; André, S.;
Buzamet, E.; Caraballo, R.; Fernández-Cureses, G.; Morando, M.; Ribeiro, J. P.;
Ramírez-Gualito, K.; de Pascual-Teresa, B.; Cañada, F. J.; Menéndez, M.;
Ramström, O.; Jiménez-Barbero, J.; Solís, D.; Gabius, H.-J. Org. Biomol. Chem.
2011, 9, 5445–5455; (f) Giguère, D.; André, S.; Bonin, M.-A.; Bellefleur, M.-A.;
Provencal, A.; Cloutier, P.; Pucci, B.; Roy, R.; Gabius, H.-J. Bioorg. Med. Chem.
2011, 19, 3280–3287.
19. (a) Kopitz, J.; von Reitzenstein, C.; André, S.; Kaltner, H.; Uhl, J.; Ehemann, V.;
Cantz, M.; Gabius, H.-J. J. Biol. Chem. 2001, 276, 35917–35923; (b) André, S.; Liu,
B.; Gabius, H.-J.; Roy, R. Org. Biomol. Chem. 2003, 1, 3909–3916; (c) Ahmad, N.;
Gabius, H.-J.; André, S.; Kaltner, H.; Sabesan, S.; Roy, R.; Liu, B.; Macaluso, F.;
Brewer, C. F. J. Biol. Chem. 2004, 279, 10841–10847; (d) Kopitz, J.; Bergmann,
M.; Gabius, H.-J. IUBMB Life 2010, 62, 624–628.
20. Percec, V.; Leowanawat, P.; Sun, H.-J.; Kulikov, O.; Nusbaum, C. D.; Tran, T. M.;
Bertin, A.; Wilson, D. A.; Peterca, M.; Zhang, S.; Kamat, N. P.; Vargo, K.; Moock,
D.; Johnston, E. D.; Hammer, D. A.; Pochan, D. J.; Chen, Y.; Chabre, Y. M.; Shiao,
T. C.; Bergeron-Brlek, M.; André, S.; Roy, R.; Gabius, H.-J.; Heiney, P. A. J. Am.
Chem. Soc. 2013, 135, 9055–9077.
21. Aravind, S.; Park, W. K. C.; Brochu, S.; Roy, R. Tetrahedron Lett. 1994, 35, 7739–
7742.
22. Srinivasan, M.; Sankararaman, S.; Hopf, H.; Dix, I.; Jones, P. G. J. Org. Chem.
2001, 66, 4299–4303.
24. Patnaik, S. K.; Stanley, P. Methods Enzymol. 2006, 416, 159–182.
25. Kopitz, J.; Fík, Z.; André, S.; Smetana, K., Jr.; Gabius, H.-J. Mol. Pharm. 2013, 10,
2054–2061.