Synthesis of glycopeptide dendrimers, dimerization and affinity for Concanavalin A

Article abstract of DOI:10.1016/j.bmc.2011.03.047

We described herein the synthesis of second generation glycopeptide dendrimers G2a-g presenting variable amino acids placed internally into the multivalent scaffold. The effect of such structural modulation on recognition processes by Concanavalin A (Con

Full text of DOI:10.1016/j.bmc.2011.03.047

Bioorganic & Medicinal Chemistry 19 (2011) 2879–2887
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis of glycopeptide dendrimers, dimerization and affinity
for Concanavalin A
Ronan Euzen a, , Jean-Louis Reymond b
⇑
a Technicians Training Laboratory, Institute of Chemical Sciences and Engineering (ISIC-GE), Ecole Polytechnique Fédérale de Lausanne, Lausanne CH-1015, Switzerland
b Department of Chemistry and Biochemistry, University of Berne, Freiestrasse 3, CH-3012 Berne, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
We described herein the synthesis of second generation glycopeptide dendrimers G2a–g presenting var-
iable amino acids placed internally into the multivalent scaffold. The effect of such structural modulation
on recognition processes by Concanavalin A (Con A), was then estimated by enhanced-sensitivity
Enzyme-Linked Lectin Assay (ELLA). In a complementary study, glycopeptide dendrons of different valen-
Received 1 December 2010
Revised 8 March 2011
Accepted 18 March 2011
Available online 8 April 2011
cies and including a L-cysteine residue before the dendritic core (G0SH, G1SH and G2SH), were also syn-
thesized and homodimerized. Then, the disulfide-containing glycopeptide dendrimers generated by this
convergent approach (G02S2, G12S2 and G22S2) were used as Con A inhibitors and assayed by ELLA.
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Glycodendrimer
Peptide
Disulfide
Concanavalin A
Enzyme-linked lectin assay
1. Introduction
example, subunits derivatized from L-cysteine (Cys) could lead to
disulfide-containing dendrimers, displaying hydrolytic activi-
ties.1b,8 However, to the best of our knowledge, few examples took
advantage of such a convergent strategy for designing multivalent
lectin inhibitors.7a
Dendrimers are monodisperse tree-like molecules built in a
generation-wise manner. These hyperbranched systems have
found many applications in homogeneous catalysis, organic syn-
thesis or in biology.1 Glycodendrimers, such as glycoclusters and
glycopolymers, have thus appeared as very useful molecules to
probe carbohydrate–lectin interactions.2,3 Construction of well-
defined multivalent scaffolds and their use as molecular probes is
then still a challenge.4
In this context, some peptide and glycopeptide dendrimers pre-
sented an interesting propensity to interact with lectins.2,5 We
showed for example that screening of combinatorial libraries could
lead to potent bacterial or plant lectin inhibitors.2b–g We also re-
cently completed these studies by underlining that nature of ami-
no acids located close to external carbohydrate moieties of such
glycopeptide dendrimers, could tune affinity for Concanavalin A
(Con A).2a Nevertheless, influence of amino acids incorporated
more internally into multivalent scaffolds was not fully explored.
We showed also, in agreement with previously reported data,
that valency of these lectin inhibitors was a key feature for recog-
nition processes.6 In many dendrimers studies, valency could be in-
creased by dimerization of thiol-functionnalized dendrons.7 For
Therefore, and considering our interest in (glyco-)peptide den-
drimers, we evaluated herein in a first part, if amino acids placed
relatively internally into mannosylated peptide dendrimers could
tune affinity for Con A.1b,2,7c,8,9 This protein was selected as a model
receptor since this mannose-binding plant lectin and its deriva-
tives were commercially available and widely used to probe carbo-
hydrate–protein interactions.6,10 After experiments, we found by
enhanced-sensitivity Enzyme-Linked Lectin Assay (ELLA) that
modulating internal amino acids composition of presented glyco-
peptide dendrimers did not improve significantly their inhibition
properties. On the basis of these results, mono-, di- and tetravalent
glycodendrons including Cys far from carbohydrate moieties were
synthesized and then homodimerized to give, respectively, the cor-
responding di-, tetra- and octavalent glycodendrimers. The inhibi-
tion potency of these disulfides was also measured by ELLA. As
expected, binding properties increased with valency of glycopep-
tide dendrimers. Their abilities to inhibit Con A activity were com-
parable to these of glycopeptide dendrimers obtained previously
within a divergent approach. Consequently, these results sug-
gested that inhibition properties of presented glycopeptide dendri-
mers depended mainly on their valency, on the amino acids located
close to carbohydrate units and little on the manner to construct
the rest of multivalent scaffolds.
⇑
Corresponding author. Tel.: +41 21 693 53 93; fax: +41 21 693 35 50.
E-mail address: ronan.euzen@epfl.ch (R. Euzen).
0968-0896/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2011.03.047

2880
R. Euzen, J.-L. Reymond / Bioorg. Med. Chem. 19 (2011) 2879–2887
Scheme 1. General synthesis of second generation glycopeptide dendrimers G2a–g. Reagents and conditions: (a) protected amino acid [Fmoc–Arg(Pbf)–OH, Fmoc–Ser(tBu)–
OH, Fmoc–Glu(OtBu)–OH or Fmoc–Lys(Fmoc)–OH], BOP, DIPEA, NMP, rt, 1–12 h; (b) Ac2O, CH2Cl2, rt, 10 min.; (c) piperidine, DMF, rt, 2ꢀ 10 min; (d) 1, HCTU, DIPEA, NMP, rt,
12 h; (e) NH3, H2O, MeOH, rt, 24 h; (f) TFA, TIS, H2O, rt, 4 h; (g) NH3, H2O, MeOH, rt, 24 h.
cleaved as previously described and all N-termini were acylated
once again with N,N0-bis-Fmoc protected Lys in order to introduce
the last branching units. Protected L-tyrosine (Tyr) moieties were
afterwards coupled to the supported tetravalent peptides before
being involved in the last piperidine-mediated deprotection step. Fi-
nally, all N-termini were glycosylated using 2-(6-chloro-1H-benzo-
triazole-1-yl)-1,1,3,3-tetramethyluronium hexafluoro-phosphate
(HCTU) as a coupling reagent and the known thiomannosidic build-
ing block 1.12 This latter could be easily obtained in an anomerically
pure form from D-mannose (Man), in few steps and with a good over-
all yield. After on-bead carbohydrates deacetylation reaction, acido-
lytic cleavage from resin, in-solution removal of the remaining
acetyl groups and purification by reversed phase preparative HPLC
(RP-HPLC), the target glycopeptide dendrimers G2a–c could be iso-
lated with 5–7% yields.13 The second generation glycopeptide den-
drimers G2e–g, including no variable amino acid between
branching units but one copy of Arg, Ser or Glu before the dendritic
core (X2), were also synthesized by SPPS. After coupling, deprotec-
tion, cleavage reactions and purification, G2e–g were obtained with
15%, 22% and 18% yields, respectively.
Figure 1. Example of the inhibition curve obtained for the second generation
glycopeptide dendrimer G2d by enhanced-sensitivity ELLA.
2.2. Affinity of second generation glycopeptide dendrimers
G2a–c and G2e–g for Con A
G2a–c were involved in an enhanced-sensitivity ELLA.2a As re-
ported, this competitive test initially relies on yeast mannan
adsorption on a 96-wells microtitration plate. Inhibitor is then
introduced as serial dilutions followed by biotinylated Con A. After
washings, the remaining lectin can be covered by a preformed Neu-
travidin™-biotinylated horseradish peroxidase complex. Finally,
addition of a chromogenic enzyme substrate leads quickly to reli-
able absorptions. This methodology, inspired by original works of
Duk et al. gives sigmoidal inhibition response curves from which
IC50 values can be measured (Fig. 1).14 These latters are then
compared to inhibition properties obtained, in the same conditions
(Table 1), with monovalent carbohydrate derivatives such as Man,
2. Results and discussion
2.1. Synthesis of second generation glycopeptide dendrimers
G2a–c and G2e–g
Glycopeptide dendrimers G2a–c incorporating variable amino
acids (X1) between branching points were initially synthesized
(Scheme 1). L-Arginine (Arg), L-serine (Ser) or L-glutamic acid (Glu)
were indeed selected as variable amino acids given their comple-
mentary charges at neutral pH. These multivalent glycoconjugates
were synthesized by standard Fmoc solid phase peptide synthesis
(SPPS).11 N,N0-Bis-Fmoc protected
L-lysine (Lys) was then coupled
p-nitrophenyl
side (Me -Man). This latter glycoconjugate is indeed used as a
monovalent reference in our ELLA. Within this approach, G2a–c
then displayed IC50’s of 92, 64 and 60 M corresponding to relative
potencies per carbohydrate moiety of 3.6, 5.2 and 5.5, respectively.
a-D-mannoside (pNP a-Man) or methyl a-D-manno-
to Rink amide resin using benzotriazole-1-yl-oxy-tris-(dimethyl-
amino)-phosphonium hexafluorophos-phate (BOP) as a condensa-
tion agent. After capping of potentially unreacted amines and
Fmoc groups removal by treatment with piperidine, protected vari-
able amino acids could be coupled. Terminal Fmoc groups were then
a
l

R. Euzen, J.-L. Reymond / Bioorg. Med. Chem. 19 (2011) 2879–2887
2881
Table 1
Inhibition properties of glycopeptide dendrimers G1, G2a–g, G3, G02S2, G12S2 and G22S2
Compound
na
IC50
(l
M)
r.p.b
r.p./nc
Mand
1
1
1
4
4
4
4
4
4
4
2
4
8
2
8
10456 ( 1953)
1320 ( 57)
644 ( 25)
92 ( 42)
64 ( 19)
60 ( 17)
0.13 ( 0.04)
1.0
0.13 ( 0.04)
1.0
Me
a
-Mand
-Mand
pNP
G2a
a
2.1 ( 0.2)
14.3 ( 8.6)
20.6 ( 9.0)
22.0 ( 9.3)
25.4 ( 6.5)
21.3 ( 11.2)
24.0 ( 9.1)
14.2 ( 6.1)
12.7 ( 4.5)
25.4 ( 5.0)
2.1 ( 0.2)
3.6 ( 2.1)
5.2 ( 2.3)
5.5 ( 2.3)
6.4 ( 1.6)
5.3 ( 2.8)
6.0 ( 2.3)
3.6 ( 1.5)
6.4 ( 2.3)
6.4 ( 1.6)
G2b
G2c
G2dd
G2e
52 ( 11)
62 ( 26)
55 ( 15)
93 ( 30)
104 ( 32)
52 ( 8)
2.0 ( 0.9)
150 ( 34)
2.9 ( 1.1)
G2f
G2g
G02S2
G12S2
G22S2
G1d
660 ( 326)
8.8 ( 2.5)
455 ( 203)
83 ( 41)
4.4 ( 1.3)
57 ( 25)
G3d
a
b
c
Number of mannosyl residues.
Relative potency compared to Me
Relative potency per carbohydrate moiety.
a-Man.
d
See Ref. 2a.
These values highlighted an improvement of affinity toward Con A
by cluster effect compared to that obtained with the monovalent
reference.15 In other words, the affinity of the lectin for inhibitor in-
creased with its valency. This result might arise from higher local
ligand concentration. However, these IC50’s remained close to that
found previously for the glycopeptide dendrimer G2d, presenting
only four external Tyr residues. Indeed, this latter compound was
greatly simplified by a subsequent reduction with DTT (1,4-
dithio- -threitol), thus allowing a partial purification of expected
D,L
thiols by preparative RP-HPLC.11 Then, air-promoted dimerization
reactions were conveniently performed. Completion of such an
oxidation process was easily monitored by analytical RP-HPLC.
After work-up and purification, the target disulfides G02S2, G12S2
and G22S2 were finally isolated with 16%, 6% and 4% overall yields,
respectively. Then, G02S2 and G12S2 were prepared in slightly low-
er yields than these associated to the synthesis of known counter-
associated within the same assay, to an IC50 value of 52 lM and
then to relative potency per carbohydrate of 6.4. Then, introduction
of variable amino acids between branching points influenced mod-
estly the inhibition properties of our glycopeptide dendrimers. The
inhibition properties of G2e–g were also estimated by enhanced-
sensitivity ELLA. This assay gave, respectively, IC50 values of 62,
parts G1 and G2d obtained within
a
divergent approach.2a
However, the octavalent derivatives G22S2 and the previously de-
scribed analog G3 were isolated with comparable yields.2a
Although this latter could only be isolated in a 4% yield by decreas-
ing of resin loading, this condition was not required for preparing
G22S2.
55 and 93 lM associated to relative potencies par carbohydrate of
5.3, 6.0 and 3.6. Thus, introduction of variable amino acids before
the first branching unit did not affect significantly inhibition prop-
erties of glycopeptide dendrimers in agreement with results found
with G2a–c. Consequently, although introduction close to carbohy-
drate moieties of amino acids charged at neutral pH seemed to dis-
rupt recognition processes, this behavior was not observed for
further variable positions.2a These results then contrasted with
those previously observed with external amino acids.
2.4. Affinity of disulfide-based glycopeptide dendrimers G02S2,
G12S2 and G22S2 for Con A
The ability of G02S2, G12S2 and G22S2 to inhibit Con A activity was
also estimated by enhanced-sensitivity ELLA. The obtained data
were additionally confirmed by standard ELLA.16 G02S2 then pre-
sented an IC50 value of 104 lM corresponding to a relative potency
2.3. Synthesis of disulfide-based glycopeptide dendrimers G02S2,
G12S2 and G22S2
per mannosyl residue of 6.4. In other words, this divalent disulfide
was an inhibitor as potent as the known first generation glycopep-
tide dendrimer G1, synthesized owing to a divergent approach.
Valency of multiantennary scaffolds is known to be an impor-
tant structural feature to tune affinity for Con A.6 This point was
also underlined in our previous study involving multivalent inhib-
itors built owing a divergent approach.2a Although convergent
strategies, and more particularly these relying on thiol-presenting
subunits dimerization, are well known in dendrimers chemistry,
disulfide-including glycopeptide dendrimers were rarely involved
as lectin inhibitors. Interestingly, in a recent study Lindhorst
et al. reported by employing a uropathogenic strain of Escherichia
coli that Cys-based mannoside glycoclusters were inhibitors of type
1 fimbriae bacterial adhesion.7a We then envisioned the synthesis
of di-, tetra- and octavalent disulfides G02S2, G12S2 and G22S2
(Scheme 2) from the corresponding glycodendrons G0SH, G1SH
and G2SH, respectively (see Supplementary data). These latters,
incorporating a Cys residue before the first branching Lys, could
be produced by classical SPPS. Once released from resin, analytical
RP-HPLC profiles unfortunately showed the formation of many
products. Complexity of such reaction mixtures was however
The dimer G12S2 was also assayed. With an IC50 of 52 lM, the inhib-
itory properties of this tetravalent compound were close to these
measured for its counterpart G2d. Indeed, a relative potency per car-
bohydrate of 6.4 was found for both glycopeptide dendrimers. Final-
ly, affinity of G22S2 for Con A was also estimated. The IC50 value of
2.0 lM obtained for this octavalent derivative could be associated
to a relative potency per carbohydrate of 83. Then, inhibition prop-
erties of G22S2 and G3 were both found in the micromolar range.
Considering the uncertainties of ELLA for high affinity ligands, rela-
tively similar inhibition behaviours were herein observed for both
glycopeptide dendrimers. Moreover, IC50 values obtained by en-
hanced-sensitivity ELLA for G02S2, G12S2 and G22S2 were compared
to these measured by standard ELLA. Indeed, in this latter method,
attachment of a 40 kDa peroxidase to the lectin would prevent
aggregation processes and thereby virtual increase of affinity.16 Gi-
ven that G02S2, G12S2 and G22S2 were associated, respectively, to
IC50 values of 102, 52 and 2.4
lM by standard ELLA, it was then likely
that presented disulfides were not involved in cross-linking of

2882
R. Euzen, J.-L. Reymond / Bioorg. Med. Chem. 19 (2011) 2879–2887
Scheme 2. Structure of glycopeptide dendrimers G1, G2d, G3, G02S2, G12S2 and G22S2.
biotinylated Con A within enhanced-sensitivity ELLA. Consequently,
affinity for Con A increased as expected with valency. Our results
were moreover consistent with pioneer works of Roy et al. in which
di- and tetravalent inhibitors were associated to IC50 values reaching
lined by ELLA that incorporation of internal amino acids did not
modify significantly the affinity for this plant lectin. This result then
contrastedwiththese previously obtained withvariable aminoacids
placed externally on the multivalent scaffold.2a Additionally, disul-
fide-containing glycopeptide dendrimers were built by homodimer-
ization of Cys-based glycodendrons. The affinity for Con A of these
dimers obtained by a convergent strategy, was also evaluated by
ELLA. Comparison of inhibition properties with results obtained
for known derivatives synthesized by a divergent approach showed
that affinity for Con A depended mainly on valency and few on the
way to assemble the multivalent scaffolds.
several dozens of
lM, and in which inhibition activity of an octamer
was in the
l
M range.6e It could be also deduced from presented re-
sults that ability of our glycopeptide dendrimers to inhibit Con A
activity was relatively independent on the nature of the employed
multivalent scaffold, as long as external amino acids were judi-
ciously chosen.
3. Conclusions
4. Experimental section
4.1. General methods
In conclusion, we contributed herein to study the relationship
existing between structure of multivalent glycoconjugates and
affinity for lectins. Different glycopeptide dendrimers, presenting
various amino acids inside or before the dendritic core, were initially
synthesized and used as ligands toward Con A. It was then under-
All standard chemicals were purchased either from Acros,
Aldrich or Fluka (Switzerland). Amino acids derivatives and

R. Euzen, J.-L. Reymond / Bioorg. Med. Chem. 19 (2011) 2879–2887
2883
coupling reagents were purchased from Seen Chemicals or
Novabiochem (Switzerland). Rink amide resin (NovaSynÒ TGR,
0.23 mmol gꢁ1) was purchased from Novabiochem (Switzerland).
All solvents used were analytical grade. Biotin labelled Concanava-
lin A (type IV, from Canavalia ensiformis) and yeast mannan (from
Saccharomyces cerevisiae) were purchased from Sigma (Switzerland).
Neutravidin™ biotin binding protein and ImmunoPureÒ bio-
tinylated horseradish peroxydase were purchased from Pierce Bio-
technology (Switzerland). C96 Maxisorp Nunc-Immuno plates and
nontreated mixing plates (EIA/RIA 96 well plates) were, respec-
tively, purchased from Milian (Switzerland) and Corning Incorpo-
rated (USA). Peptide syntheses were performed manually in a
glass reactor. Analysis by reversed phase high performance liquid
chromatography (RP-HPLC) was performed with a Waters (996
Photodiode array detector) chromatography system using a chro-
molith performance RP-18e column (4.6 ꢀ 100 mM, flow rate
3 mL minꢁ1). Preparative RP-HPLC was performed with HPLC-
grade CH3CN and MilliQ deionized H2O in a Waters prepak car-
tridge 500 g (RP-C18 20 nm, 300 Å pore size) installed on a waters
Prep LC4000 system from Millipore (flow rate 100 mL minꢁ1). The
gradients of eluent and the percentages referring to mixtures of
solvents were given as a ratio of volumes. Abbreviation ‘eluent A’
and ‘eluent B’ were used to refer, respectively, to H2O containing
0.1% TFA (trifluoroacetic acid), and CH3CN/H2O (3:2) containing
0.1% TFA. Compounds were detected by UV absorption at
214 nm. The structure of all compounds was ascertained by 1D
[1H, 13C, DEPT-135 (Distorsionless Enhancement by Polarization
Transfer)] and/or 2D NMR experiments [COSY (1H–1H Correlation
Spectroscopy), HMQC (1H–13C Heteronuclear Multiple Quantum
Correlation) and HMBC (1H–13C Heteronuclear Multiple Bond Cor-
relation) experiments]. The NMR spectra were recorded on a
DRX400 spectrometer (Bruker) at 400 MHz for 1H and 100 MHz
for 13C, or on DRX500 spectrometer (Bruker) at 500 MHz for 1H
and 125 MHz for 13C (Analytical Research and Services, University
of Berne). Chemical shifts were given in d units, measured from the
solvent signal. MS spectra were performed by ESI (Electrospray
Ionization) either in positive or negative mode, and were recorded
on an Applied Biosystems/Sciex Qtrap spectrometer (Analytical Re-
search and Services, University of Berne).
tored with the previously mentioned tests. Finally, the whole
procedure was repeated as many times as necessary to achieve
the synthesis of the desired oligopeptide.
4.2.2. Mannosylation reactions
Once realized the assembly of the peptide backbone and the
cleavage of the last Fmoc residue, the resin was acylated using
5 equiv per amine of the mannosidic building block 1, 5 equiv of
HCTU
[2-(6-chloro-1H-benzotriazole-1-yl)-1,1,3,3-tetramethyl-
uronium hexafluorophosphate] and 10 equiv of DIEA in NMP
(5 mL). The reaction was carried out overnight in dark and the resin
was then washed with NMP, MeOH and CH2Cl2 (3 ꢀ 5 mL of each
solvent). The completion of the glycosidic capping was checked
with the previously mentioned TNBS test. The reaction was sys-
tematically restarted and, after washings, the potentially free
amines were acetylated as described above.
4.2.3. On-bead O-deacetylation
The on-bead deacetylation of carbohydrate moieties was real-
ized by suspending beads in a mixture of MeOH, H2O and 25%
aqueous solution of NH4OH (4 mL, 8:1:1). After stirring for 24 h
and filtration, the beads were washed with NMP, MeOH and CH2Cl2
(3 ꢀ 5 mL of each solvent).
4.2.4. One-pot side chains deprotection and cleavage from resin
The side-chains deprotection and cleavage from resin were car-
ried out concomitantly by stirring the beads for 4 h in a mixture of
TFA, TIS (triisopropylsilane) and H2O (3 mL, 94:5:1). For cysteine-
containing dendrimers, 1% of EDT (1,2-ethandithiol) was added
to this cleavage mixture. After filtration and washing of the solid
support with TFA (2 ꢀ 1 mL), the peptide was precipitated by addi-
tion of MTBE (methyl tert-butyl ether, 45 mL) and collected by cen-
trifugation (10 min, 4 °C, 4400 rpm). The crude peptide was then
resuspended in clean MTBE (50 mL), recentrifugation and finally
dried under a gentle stream of N2.
4.2.5. In-solution O-deacetylation
The in-solution removal of residual acetyl groups was achieved
by stirring for 24 h the crude glycopeptide in a mixture of MeOH,
H2O and 25% aqueous solution of NH4OH (4 mL, 8:1:1). Then, the
solution was concentrated under reduced pressure, dissolved in
eluent A (10 mL) and freeze-dried. The optimal purification condi-
tions were then estimated by analytical RP-HPLC, and the target
glycopeptide was finally purified by preparative RP-HPLC.
4.2. Chemical synthesis
4.2.1. Synthesis of the peptide backbones
The Rink amide resin was washed and swollen successively
with CH2Cl2 (2 ꢀ 5 mL) and NMP (N-methylpyrrolidone,
1 ꢀ 5 mL). The coupling reaction was then performed by stirring
4.2.6. Second generation glycopeptide dendrimer G2a
the resin (ca. 200 mg, 46
lmol) with 3 equivalents (per amine) of
The glycopeptide dendrimer G2a was synthesized from Rink
N-Fmoc protected amino acid, 3 equiv of BOP [benzotriazole-1-
yl-oxy-tris-(dimethylamino)-phosphonium hexafluorophos-phate]
and 6 equiv of DIEA (N,N-diisopropylethylamine) in NMP (5 mL) for
1–12 h. Subsequently, the solid support was washed successively
with NMP, MeOH and CH2Cl2 (3 ꢀ 5 mL of each solvent) and the
completion of the condensation reaction was controlled by using
the TNBS (2,4,6-trinitrobenzensulfonic acid) test or selectively by
using the acetaldehyde–chloranil test when traces of free amines
derived from proline had to be detected.11 The condensation proce-
dure was repeated if the reaction was found to be uncompleted
and was also systematically renewed in the case of arginine cou-
pling. Once completed the reaction, the potentially free amines
were acetylated by stirring the resin for 10 min with an equivolu-
mic mixture of Ac2O and CH2Cl2 (5 mL), followed by washings as
described above. Then, the Fmoc groups were then removed by
stirring the resin with a solution (5 mL, 1:4) of piperidine in DMF
(N,N-dimethylformamide) for 10 min. After washings with NMP,
MeOH, CH2Cl2 (3 ꢀ 5 mL of each solvent), this deprotection proce-
dure was repeated and the completion of the reaction was moni-
amide resin (212 mg, 49 lmol). After purification by RP-HPLC, the
target TFA salt was isolated as a white amorphous (9 mg, 7%): Pre-
parative RP-HPLC: tr = 23.6 min (A/B = 100:0?40:60, 40 min); Ana-
lytical RP-HPLC: tr = 8.11 min (A/B = 100:0?40:60, 10 min); 1H
NMR (500 MHz, CD3CN/D2O (3:2) + 0.1% [D]-TFA): d 7.72–7.69 (m,
8H, H-30Tyr, H-5T0 yr), 7.41–7.39 (m, 8H, H-20Tyr, H-60Tyr), 5.87 (br s, 4H,
H-1Man), 5.14–5.10, 5.08–5.05 (2m, 4H, CH
CH Arg, CH Lys), 4.54, 4.52 (2br s, 4H, H-2Man), 4.50–4.40 (m, 8H,
H-5Man, H-6aMan), 4.35 [dd, 2J(H-6aMan, H-6bMan) = 12.2 Hz, 3J(H-
aTyr), 4.93–4.67 (m, 5H,
a
a
5
Man, H-6bMan) = 5.8 Hz, 4H, H-6bMan], 4.28–4.23 (m, 8H, H-3Man
,
,
H-4Man), 3.77–3.57, 3.50–3.35 (2m, 26H, CH2 Lys, CH2dArg, CH2aLys
e
SCH2), 3.23–3.12 (m, 8H, SCH2CH2), 2.48–2.37, 2.34–2.27, 2.25–
2.14, 2.13–2.07, 2.01–1.89, 1.84–1.69 ppm (6m, 26H, CH2bArg
CH2bLys, CH2 Arg, CH2
Lys, CH2dLys); 13C NMR (125 MHz, CD3CN/
,
c
c
D2O (3:2) + 0.1% [D]-TFA): d 175.7 [C(@O)NH2], 173.6, 173.1, 173.0
[C(@O)NH], 156.5 (C-10Arg), 155.0 (C-1T0 yr), 130.3 (C-30Tyr, C-50Tyr),
127.9, 127.7 (C-4T0 yr), 115.2, 115.2 (C-20Tyr, C-60Tyr), 84.9, 84.6 (C-
1
4
Man), 73.1 (C-5Man), 71.7, 71.6 (C-2Man), 71.2 (C-3Man), 66.9 (C-
Man), 60.8 (C-6Man), 55.4, 55.3 (CH Tyr), 53.6, 53.2, 52.9, 52.7
a

2884
R. Euzen, J.-L. Reymond / Bioorg. Med. Chem. 19 (2011) 2879–2887
(CH
(CH2bTyr), 35.9, 35.4 (SCH2CH2), 30.7, 30.2, 30.2 (CH2bArg), 30.4
(CH2bLys), 27.9 (CH2dLys), 26.5, 26.3 (SCH2), 24.5, 24.4, 24.4 (CH2 Arg),
a
Arg, CH
a
Lys), 40.8 (CH2dArg), 38.8, 38.7 (CH2
e
Lys), 36.5, 35.9, 35.9
(19 mg, 15%) by RP-HPLC purification: Preparative RP-HPLC:
tr = 28.4 min (A/B = 100:0?50:50, 50 min); Analytical RP-HPLC:
tr = 9.13 min (A/B = 100:0?50:50, 10 min); 1H NMR (400 MHz,
CD3CN/D2O (3:2) + 0.1% [D]-TFA): d 7.72–7.70 (m, 8H, H-30Tyr, H-
50Tyr), 7.41–7.38 (m, 8H, H-2T0 yr, H-60Tyr), 5.87 (br s, 4H, H-1Man),
c
2+
22.3 ppm (CH2cLys); MS (ESI+) calcd for C102H157N19O37S4
[M+2H]2+: 1184.0. Found: 1184.6.
5.17–5.13, 5.09–5.05 (2m, 4H, CHaTyr), 4.88–4.72 (m, 4H, CHaArg,
4.2.7. Second generation glycopeptide dendrimer G2b
CHaLys), 4.56–4.52 (m, 4H, H-2Man), 4.51–4.40 (m, 8H, H-5Man, H-
The synthesis of the dendrimer G2b was performed on Rink
amide resin (226 mg, 52 mol). After work-up, this molecule was
6aMan), 4.35 [dd, 2J(H-6aMan, H-6bMan) = 12.0 Hz, 3J(H-5Man, H-
l
6bMan) = 5.8 Hz, 4H, H-6bMan], 4.29–4.23 (m, 8H, H-3Man, H-4Man),
isolated as a colourless solid by RP-HPLC (6 mg, 5%): Preparative
RP-HPLC: tr = 22.7 min (A/B = 100:0?60:40, 40 min); Analytical
RP-HPLC: tr = 7.97 min (A/B = 100:0?60:40, 10 min); 1H NMR
(500 MHz, CD3CN/D2O (3:2) + 0.1% [D]-TFA): d 7.70–7.67 (m, 8H,
H-30Tyr, H-5T0 yr), 7.39–7.37 (m, 8H, H-20Tyr, H-60Tyr), 5.85 (s, 4H, H-
3.77–3.57, 3.51–3.31 (2m, 24H, CH2bTyr, CH2dArg, CH2
3.23–3.11 (m, 8H, SCH2CH2), 2.47–2.37, 2.37–2.24, 2.24–2.13,
2.13–2.03, 2.00–1.89, 1.83–1.67 ppm (6m, 22H, CH2bArg, CH2bLys
CH2cArg CH2cLys
CH2dLys); 13C NMR (100 MHz, CD3CN/D2O
(3:2) + 0.1% [D]-TFA): d =175.7 [C(@O)NH2], 173.3, 173.2, 172.9,
172.8, 172.7, 172.5, 172.4, 172.20 [C(@O)NH], 155.0, 155.0, 154.9
(C-10Arg, C-1T0 yr), 130.3 (C-3T0 yr, C-5Tyr), 127.9, 127.7 (C-40Tyr), 115.2,
115.2 (C-20Tyr, C-6T0 yr), 84.9, 84.7, 84.6 (C-1Man), 73.1 (C-5Man), 71.7
(C-2Man), 71.2 (C-3Man), 67.0 (C-4Man), 60.8 (C-6Man), 55.2, 55.1,
eLys, SCH2),
,
,
,
1
Man), 5.15–5.11, 5.07–5.04 (2m, 4H, CH
a
Tyr), 5.00–4.98, 4.92–
Lys), 4.52–4.51 (m,
4.89 (2m, 2H, CH Ser), 4.87–4.71 (m, 3H, CH
a
a
4H, H-2Man), 4.48–4.36 (m, 12H, H-5Man, H-6aMan, CH2bSer), 4.33
[dd, 2J(H-6aMan, H-6bMan) = 12.0 Hz, 3J(H-5Man, H-6bMan) = 5.9 Hz,
4H, H-6bMan], 4.26–4.21 (m, 8H, H-3Man, H-4Man), 3.81–3.55, 3.51–
55.0 (CHaTyr), 53.5, 53.2 (CHaArg, CHaLys), 38.9, 38.7, 38.7 (CH2eLys),
3.31 (2m, 22H, CH2
2.45–2.37, 2.35–2.26, 2.24–2.15, 2.12–2.05, 2.00–1.88, 1.82–
1.68 ppm (6m, 18H, CH2bLys CH2cLys
CH2dLys); 13C NMR
(125 MHz, CD3CN/D2O (3:2) + 0.1% [D]-TFA): d 175.8 [C(@O)NH2],
175.0, 174.2, 173.5, 173.0, 173.0 [C(@O)NH], 154.9, 154.9 (C-10Tyr),
130.3 (C-40Tyr), 127.9, 127.7 (C-3T0 yr, C-50Tyr), 115.2, 115.1 (C-2T0 yr, C-
60Tyr), 84.8, 84.7 (C-1Man), 73.1 (C-5Man), 71.6 (C-2Man), 71.2 (C-
e
Lys, CH2bTyr, SCH2), 3.20–3.09 (m, 8H, SCH2CH2),
36.6, 36.1, 36.0 (CH2bTyr), 35.5, 35.4 (SCH2CH2), 30.8, 30.7, 30.7,
30.6, 30.6, 30.6 (CH2bArg, CH2bLys), 28.0, 27.9 (CH2dLys), 26.6, 26.5,
,
,
26.4 (SCH2), 22.5, 22.3, 22.3 ppm (CH2
c
Lys); MS (ESI–) calcd for
C96H149N15O40S4 [M+2H2O+2OH]2ꢁ : 1139.9. Found: 1140.8.17
2ꢁ
4.2.10. Second generation glycopeptide dendrimer G2f
The glycodendrimer G2f was obtained from Rink amide
3Man), 66.9 (C-4Man), 66.3 (ch2bSer), 60.8, 60.8 (C-6Man), 55.2, 55.1
resin (218 mg, 50 lmol) and isolated as a colourless solid (24
(CH Ser, CH Tyr), 53.6, 53.5 (CH Lys), 36.5, 36.4, 36.0 (CH2 Lys),
35.5, 35.4, 35.4 (CH2bTyr), 35.5, 35.4, 35.4 (SCH2CH2), 30.4, 30.4
a
a
a
e
mg, 22%) by RP-HPLC: Preparative RP-HPLC: tr = 9.02 min (A/
B = 100:0?50:50, 50 min); Analytical RP-HPLC: tr = 9.02 min (A/
B = 100:0?50:50, 10 min); 1H NMR (500 MHz, CD3CN/D2O
(3:2) + 0.1% [D]-TFA): d 7.69-7.63 (m, 8H, H-30Tyr, H-5T0 yr), 7.39–
7.32 (m, 8H, H-20Tyr, H-6T0 yr), 5.84–5.80 (m, 4H, H-1Man), 5.14–5.10,
(CH2bLys), 27.9, 27.9, 27.8 (CH2dLys), 26.5, 26.5, 26.4 (SCH2),
2+
22.3 ppm (CH2cLys); MS (ESI+) calcd for C96H143N13O39S4
[M+2H]2+: 1114.9. Found: 1115.4.
5.06–5.03 (2m, 4H, CH
(m, 3H, CH Lys), 4.51–4.50 (m, 4H, H-2Man), 4.49–4.36 (m, 10H,
H-5Man H-6aMan H-
CH2bSer), 4.32 [dd, 2J(H-6aMan
6bMan) = 12.2 Hz, 3J(H-5Man
H-6bMan) = 5.8 Hz, 4H, H-6bMan],
4.25–4.19 (m, 8H, H-3Man, H-4Man), 3.75–3.54 (m, 10H, CH2eLys
aTyr), 4.97–4.95 (m, 1H, CHaSer), 4.86–4.70
4.2.8. Second generation glycopeptide dendrimer G2c
a
The dendrimer G2c was obtained from Rink amide resin
,
,
,
(203 mg, 47
lmol) and was isolated by RP-HPLC as a white
,
foam (6 mg, 6%): Preparative RP-HPLC: tr = 22.8 min (A/B =
100:0?40:60, 40 min); Analytical RP-HPLC: tr = 8.18 min (A/
B = 100:0?40:60, 10 min); 1H NMR (500 MHz, CD3CN/D2O (3:2) +
0.1% [D]-TFA): d 7.72–7.69 (m, 8H, H-30Tyr, H-5T0 yr), 7.40–7.38 (m,
8H, H-20Tyr, H-60Tyr), 5.86 (br s, 4H, H-1Man), 5.15–5.12, 5.09–5.06
,
CH2bTyr), 3.47–3.30 (m, 12H, CH2bTyr, SCH2), 3.19–3.08 (m, 8H,
SCH2CH2), 2.45–2.37, 2.34–2.23, 2.19–2.09, 2.09–2.01, 1.97–1.85,
1.79–1.65 ppm (6m, 18H, CH2bLys, CH2
(125 MHz, CD3CN/D2O (3:2) + 0.1% [D]-TFA):
173.2, 173.1, 172.9, 172.9, 172.7, 172.5, 172.4, 172.1 [C(@O)NH2,
C(@O)NH], 154.9, 154.9 (C-10Tyr), 130.3 (C-30Tyr C-50Tyr), 127.9,
127.7 (C-40Tyr), 115.2, 115.2 (C-2T0 yr, C-60Tyr), 84.8, 84.7, 84.6 (C-
Man), 73.1 (C-5Man), 71.6 (C-2Man), 71.2 (C-3Man), 66.9 (C-4Man),
61.2 (CH2bSer),60.8 (C-6Man), 55.1, 55.1, 55.0 (CH Ser, CH Tyr),
53.7, 53.5, 53.4 (CH Lys), 38.6, 38.6 (CH2 Lys), 36.5, 36.0, 36.0
(CH2bTyr), 35.5, 35.4 (SCH2CH2), 30.7, 30.5, 30.3 (CH2bLys), 27.9
c
Lys, CH2dLys); 13C NMR
d
173.4, 173.4,
(2m, 4H, CH
4.67 (m, 3H, CH
4H, H-5Man), 4.44–4.40 (m, 4H, H-6aMan), 4.35 [dd, 2J(H-6aMan
H-6bMan) = 12.2 Hz, 3J(H-5Man, H-6bMan) = 6.0 Hz, 4H, H-6bMan],
4.28–4.23 (m, 8H, H-3Man, H-4Man), 3.80–3.57 (m, 10H, CH2eLys
CH2bTyr), 3.50–3.32 (m, 12H, CH2bTyr, SCH2), 3.23–3.11 (m, 8H,
SCH2CH2), 3.05–2.96 (m, 4H, CH2 Glu), 2.74–2.68, 2.60–2.49 (2m,
4H, CH2bGlu), 2.44–2.37, 2.35–2.25, 2.22–2.15, 2.13–2.06, 2.01–
1.90, 1.82–1.68 ppm (6m, 18H, CH2bLys, CH2
Lys, CH2dLys); 13C
aTyr), 4.95–4.91, 4.87–4.82 (2m, 2H, CHaGlu), 4.87–
aLys), 4.53–4.52 (m, 4H, H-2Man), 4.50–4.45 (m,
,
,
1
,
a
a
a
e
c
(CH2dLys), 26.5, 26.4, 26.3 (SCH2), 22.4, 22.3 ppm (CH2
(ESI–) calcd for C93H134N12O37S4
1069.8.
c
Lys); MS
2ꢁ
c
[M–2H]2ꢁ : 1069.4. Found:
NMR (125 MHz, CD3CN/D2O (3:2) + 0.1% [D]-TFA): d 175.9, 175.8,
175.7 [C(@O)NH2, CO2H], 173.4, 173.0, 172.9, 172.9, 172.8, 172.3,
172.2, 172.0 [C(@O)NH], 155.0, 155.0, 154.9 (C-10Tyr), 130.3 (C-
30Tyr, C-50Tyr), 127.9, 127.7, 127.7 (C-40Tyr), 115.2, 115.2 (C-2T0 yr, C-
60Tyr), 84.9, 84.6, 84.6 (C-1Man), 73.1, 73.1 (C-5Man), 71.6 (C-2Man),
4.2.11. Second generation glycopeptide dendrimer G2g
The glycopeptide dendrimer G2g was synthesized from Rink
amide resin (216 mg, 50 lmol) and isolated by RP-HPLC as a colour-
71.2 (C-3Man), 66.9, 66.9 (C-4Man), 60.8, 60.8 (C-6Man), 55.3, 55.2 (CH
less foam (19 mg, 18%): Preparative RP-HPLC: tr = 28.0 min (A/
B = 100:0?50:50, 50 min); Analytical RP-HPLC: tr = 9.16 min (A/
B = 100:0?50:50, 10 min); 1H NMR (400 MHz, CD3CN/D2O
(3:2) + 0.1% [D]-TFA): d 7.69–7.66 (m, 8H, H-30Tyr, H-5T0 yr), 7.38–
7.34 (m, 8H, H-20Tyr, H-6T0 yr), 5.83 (br s, 4H, H-1Man), 5.16–5.08,
a
-
Tyr), 53.7, 53.6, 53.3 (CH
(CH2 Lys), 36.6, 36.6, 36.0 (CH2bTyr), 35.5, 35.4, 35.3 (SCH2CH2),
30.7, 30.5, 30.4 (CH2bLys), 30.0, 29.9 (CH2 Glu), 27.9 (CH2dLys),
aLys), 52.8, 52.8 (CHaGlu), 38.8, 38.7
e
c
26.5, 26.3 (SCH2), 26.3, 26.0 (CH2bGlu), 22.4, 22.3, 22.3 ppm
2ꢁ
(CH2c
Lys); MS (ESI–) calcd for C100H143N13O41S4
[M–2H]2ꢁ
:
5.05–5.02 (2m, 4H, CH
9.6 Hz, 3J(CH
Glu,CH2bbGlu) = 4.7 Hz, 1H, CH
(m, 3H, CH Lys), 4.50–4.49 (m, 4H, H-2Man), 4.47–4.36 (m, 8H, H-
Man, H-6aMan), 4.31 [dd, 2J(H-6aMan, H-6bMan) = 12.2 Hz, 3J(H-5Man
H-6bMan) = 5.8 Hz, 4H, H-6bMan], 4.25–4.19 (m, 8H, H-3Man, H-
Man), 3.75–3.53 (m, 10H, CH2 Lys, CH2bTyr), 3.47–3.28 (m, 12H,
a
Tyr), 4.91–4.87 [dd, 3J(CH
Glu], 4.83–4.69
aGlu,CH2baGlu) =
1154.9. Found: 1155.6.
a
a
a
4.2.9. Second generation glycopeptide dendrimer G2e
The glycopeptide dendrimer G2e was obtained from Rink amide
resin (232 mg, 53 mol) and was isolated as a colourless TFA salt
5
,
l
4
e

R. Euzen, J.-L. Reymond / Bioorg. Med. Chem. 19 (2011) 2879–2887
2885
CH2bTyr, SCH2), 3.19–3.09 (m, 8H, SCH2CH2), 3.03–2.99 (m, 2H,
CH2 Glu), 2.74–2.66, 2.55–2.47 (2m, 2H, CH2bGlu), 2.44–2.33, 2.33–
2.21, 2.21–2.09, 2.09–1.98, 1.98–1.83, 1.79–1.63 ppm (6m, 18H,
CH2bLys CH2cLys
CH2dLys); 13C NMR (100 MHz, CD3CN/D2O
10 min); 1H NMR (500 MHz, CD3CN/D2O (3:2) + 0.1% [D]-TFA): d
7.76–7.71 (m, 4H, H-30Tyr H-5T0 yr), 7.44–7.40 (m, 4H, H-20Tyr
H-60Tyr), 5.89–5.85 (m, 2H, H-1Man), 5.25–5.15 (m, 4H, CH
Cys,
CH Tyr), 4.57–4.36 (m, 8H, H-2Man, H-5Man, H-6aMan, H-6bMan),
c
,
,
a
,
,
a
(3:2) + 0.1% [D]-TFA): d 175.8 (CO2H), 174.9 [C(@O)NH2],173.4,
173.2, 173.2, 173.1, 172.9, 172.9, 172.8, 172.5, 172.3, 172.1
[C(@O)NH], 154.9, 152.9 (C-10Tyr), 130.3 (C-30Tyr, C-50Tyr), 127.9, 127.7
(C-40Tyr), 115.1 (C-2T0 yr, C-6T0 yr), 84.8, 84.7, 84.6 (C-1Man), 73.1 (C-
4.29–4.24 (m, 4H, H-3Man, H-4Man), 3.84–3.78 (m, 2H, CH2bCys),
3.70–3.64 (m, 2H, CH2bTyr), 3.61–3.47 (m, 4H, CH2bCys, CH2bTyr),
3.46–3.36 (m, 4H, SCH2), 3.20–3.16 ppm (m, 4H, SCH2CH2); 13C
NMR (125 MHz, CD3CN/D2O (3:2) + 0.1% [D]-TFA): d 173.1,173.1,
172.3, 172.3 [C(@O)NH2, C(@O)NH], 154.7, 154.7 (C-10Tyr), 130.1,
5
Man), 71.7 (C-2Man), 71.2 (C-3Man), 66.9 (C-4Man), 60.8 (C-6Man),
55.1, 55.0 (CH Tyr), 53.7, 53.4 (CH Lys), 52.3 (CH Glu), 38.7, 38.6
(CH2 Lys), 36.5, 36.0, 35.9 (CH2bTyr), 35.5, 35.4, 35.3 (SCH2CH2),
30.7, 30.5, 30.3 (CH2bLys), 29.9 (CH2 Glu), 27.9, 27.9 (CH2dLys), 26.5,
a
a
a
130.1 (C-30Tyr, C-5T0 yr), 127.5, 127.5 (C-40Tyr), 115.0, 115.0 (C-20Tyr
,
e
C-60Tyr), 84.6, 84.6 (C-1Man), 72.9, 72.8 (C-5Man), 71.4, 71.4 (C-2Man),
71.0, 71.0 (C-3Man), 66.7, 66.7 (C-4Man), 60.6, 60.6 (C-6Man), 55.0,
c
26.4 (SCH2), 26.3 (CH2bGlu), 22.4, 22.3 ppm (CH2
c
Lys); MS (ESI–)
54.9 (CHaTyr), 51.9, 51.9 (CHaCys), 38.7, 38.7 (CH2bCys), 35.8,
calcd for C95H136N12O38S42– [M–2H]2–: 1090.4. Found: 1090.8.
35.8 (CH2bTyr), 35.2, 35.2 (SCH2CH2), 26.2, 26.2 ppm (SCH2); MS
(ESIꢁ) calcd for C42H59N6O16S4 [MꢁH]ꢁ : 1063.3. Found: 1063.0.
ꢁ
4.2.12. Synthesis of cysteine-containing glycopeptides
After peptide assembly, glycosylation reaction, on-bead deacet-
ylation and cleavage from the solid support as described above, the
crude material was solubilised in 25 mL of a 50 mM aqueous solu-
4.2.18. Tetravalent disulfide G12S2
After dimerization of the thiol G1SH and purification by RP-
HPLC, the disulfide G12S2 was obtained as a colourless amorphous
solid (6 mg, 6%, overall): Preparative RP-HPLC: tr = 28.8 min (A/
B = 100:0?50:50, 50 min); Analytical RP-HPLC: tr = 8.62 min (A/
B = 100:0?50:50, 10 min); 1H NMR (500 MHz, CD3CN/D2O
(3:2) + 0.1% [D]-TFA): d 7.74–7.66 (m, 8H, H-30Tyr, H-5T0 yr), 7.43–
7.36 (m, 8H, H-20Tyr, H-6T0 yr), 5.88–5.85 (m, 4H, H-1Man), 5.25–5.20
tion of (NH4)2CO3. DTT (1,4-dithio-D,L-threitol) was added (ca.
71 mg, 460 mol) and the mixture was stirred under argon for
l
2 h. The solution was freeze-dried, the residue thus obtained dis-
solved in eluent A (20 mL) and subjected to an other freeze-drying
step. The composition of the reaction mixture was evaluated by
analytical RP-HPLC and partial purification of intermediates
G0SH, G1SH and G2SH was achieved by preparative RP-HPLC.
The fractions containing mercaptans were then pooled and
freeze-dried.
(m, 2H, CH
4.79 (m, 2H, CH
H-6bMan), 4.29–4.23 (m, 4H, H-3Man, H-4Man), 3.89–3.82 (m, 2H,
CH2bCys), 3.76–3.54 (m, 10H, CH2bCys, CH2bTyr, CH2 Lys), 3.52–
3.32 (m, 12H, CH2bTyr, SCH2), 3.22–3,11 (m, 8H, SCH2CH2), 2.40–
2.29, 2.28–2.15, 2.03–1.88, 1.84–1.68 ppm (4m, 12H, CH2bLys
CH2cLys
CH2dLys); 13C NMR [125 MHz, CD3CN/D2O, 3:2
aCys), 5.18–5.12, 5.12–5.05 (2m, 4H, CH
aTyr), 4.82–
aLys), 4.56–4.32 (m, 16H, H-2Man, H-5Man, H-6aMan
,
e
4.2.13. Monovalent glycopeptide G0SH
,
The intermediate G0SH was synthesized from Rink amide resin
,
(410 mg, 94
l
mol) and partially purified by RP-HPLC: Preparative
(v:v) + 0.1% TFA-d]: d 173.5, 172.9 [C(@O)NH2, C(@O)NH], 154.8,
154.8 (C-10Tyr), 130.2 (C-30Tyr, C-5T0 yr), 127.9, 127.6 (C-40Tyr), 115.1,
115.1 (C-20Tyr, C-6T0 yr), 84.8, 84.5 (C-1Man), 73.0 (C-5Man), 71.6 (C-
RP-HPLC: tr = 8.4 min (A/B = 90:10?40:60, 50 min); Analytical
RP-HPLC: tr = 3.34 min (A/B = 90:10?40:60, 10 min); MS (ESI+)
calcd for C21H32N3O9S2 [M+H]+: 534.2. Found: 534.2.
2
Man), 71.1 (C-3Man), 66.9 (C-4Man), 60.7 (C-6Man), 55.1 (CH
53.7 (CH Lys), 52.0 (CH Cys), 38.6, 38.6 (CH2bCys, CH2 Lys), 36.5
(CH2bTyr), 35.4, 35.3 (SCH2CH2), 30.3, 30.3 (CH2bLys), 27.8, 27.8
aTyr),
+
a
a
e
4.2.14. Divalent glycopeptide G1SH
The synthesis of the intermediate G1SH was performed from
(CH2dLys), 26.5, 26.3 (SCH2), 22.3, 22.2 ppm (CH2
c
Lys); MS (ESIꢁ)
calcd for C90H128N12O36S6 [Mꢁ2H]2ꢁ : 1072.3. Found: 1073.0.
2ꢁ
Rink amide resin (403 mg, 93
l
mol) and partially purified by RP-
HPLC: Preparative RP-HPLC: tr = 25.4 min (A/B = 100:0?40:60,
60 min); Analytical RP-HPLC: tr = 7.36 min (A/B = 100:0?40:60,
4.2.19. Octavalent disulfide G22S2
10 min); MS (ESI+) calcd for C45H67N6O18S3 [M+H]+: 1075.4.
The synthesis of G22S2 could then achieved from the mercap-
tan G2SH and the target disulfide was obtained, after RP-HPLC
purification, as a colourless foam (9 mg, 4%, overall): Preparative
RP-HPLC: tr = 22.6 min (A/B = 100:0?40:60, 60 min); Analytical
RP-HPLC: tr = 8.39 min (A/B = 100:0?40:60, 10 min); 1H NMR
(500 MHz, CD3CN/D2O (3:2) + 0.1% [D]-TFA): d 7.73–7.70 (m, 16H,
H-30Tyr, H-5T0 yr), 7.42–7.39 (m, 16H, H-20Tyr, H-60Tyr), 5.89–5.87 (br s,
+
Found: 1075.6.
4.2.15. Tetravalent glycopeptide G2SH
The intermediate thiol G2SH was synthesized on Rink amide re-
sin (411 mg, 94 lmol) and partially purified by RP-HPLC: Prepara-
tive RP-HPLC: tr = 22.0 min (A/B = 90:0?40:60, 50 min); Analytical
RP-HPLC: tr = 7.56 min (A/B = 90:0?40:60, 10 min); MS (ESI+)
8H, H-1Man), 5.36–5.33, 5.29–5.26 (2m, 2H, CH Cys), 5.20–5.16,
a
calcd for C93H138N12O36S5 [M+2H]2+: 1079.4. Found: 1080.2.
5.11–5.07 (m, 8H, CH Tyr), 4.90–4.70 (m, 6H, CHaLys), 4.56–4.54
a
2+
(m, 8H, H-2Man), 4.52–4.41 (m, 16H, H-5Man, H-6aMan), 4.37 [dd,
2J(H-6aMan, H-6bMan) = 12.4 Hz, 3J(H-5Man, H-6bMan) = 5.9 Hz, 8H,
H-6bMan], 4.29–4.26 (m, 16H, H-3Man, H-4Man), 3.89–3.84 (m, 2H,
CH2bCys), 3.77–3.58 (m, 22H, CH2bCys, CH2bTyr, CH2eLys), 3.51–3.35
(m, 24H, CH2bTyr, SCH2), 3.19–3.15 (m, 24H, SCH2CH2), 2.46–2.39,
4.2.16. Homodimerization of cysteine-containing glycopeptides
The intermediate thiol was solubilised in a mixture MeOH, H2O
and 25% aqueous solution of NH4OH (4 mL, 8:1:1) and the reaction
medium subjected to an air-bubbling for 48 h. After concentration
under reduced pressure, acidification by adding a small amount of
eluent A (ca. 10 mL) and freeze-drying, the optimal purification
conditions were estimated by analytical RP-HPLC. Finally, the de-
sired disulfide was isolated by preparative RP-HPLC.
2.39–2.26, 2.26–2.13, 2.13–2.00, 2.00–1.90, 1.82–1.68 ppm (6m,
36H, CH2bLys, CH2c
Lys, CH2dLys); 13C NMR (125 MHz, CD3CN/D2O
(3:2) + 0.1% [D]-TFA): d 173.4, 172.6, 172.0 [C(@O)NH2, C(@O)NH],
154.9 (C-10Tyr), 130.5 (C-3T0 yr, C-50Tyr), 127.4 (C-4T0 yr), 116.1 (C-2T0 yr, C-
60Tyr), 85.8 (C-1Man), 74.0 (C-5Man), 72.6 (C-2Man), 72.2 (C-3Man), 67.8
4.2.17. Divalent disulfide G02S2
(C-4Man), 61.8 (C-6Man), 55.9 (CHaTyr), 54.3 (CHaLys), 53.0 (CHaCys),
The thiol G0SH was then dimerized to give the disulfide G02S2
which was isolated by RP-HPLC as a colourless foam (8 mg, 16%,
overall): Preparative RP-HPLC: tr = 22.9 min (A/B = 100:0?50:50,
50 min); Analytical RP-HPLC: tr = 7.36 min (A/B = 100:0?50:50,
40.1 (CH2bCys), 39.7 (CH2 Lys), 37.6 (CH2bTyr), 36.4 (SCH2CH2), 31.5
e
(CH2bLys), 29.6 (CH2dLys), 27.3 (SCH2), 23.4 ppm (CH2
cLys); MS
(ESIꢁ) calcd for C186H266N24O72S10 [Mꢁ4H]4ꢁ : 1078.0. Found:
4ꢁ
1076.9.

2886
R. Euzen, J.-L. Reymond / Bioorg. Med. Chem. 19 (2011) 2879–2887
4.3. Enzyme-linked lectin assays (ELLA)
experimental data was the closest to the average value). The relative
potencies (r.p. and r.p./n) were then calculated as follows: r.p. =
IC50-Me -Man/IC50-inhibitor and r.p./n = IC50-Me -Man/(IC50-inhibitor ꢀ n).
4.3.1. Preparation of buffers used for the inhibition
measurements
a
a
The corresponding errors were then
D
r.p. = (IC
/IC50-inhibitor)
-Man
a
PBS (phosphate saline buffer, 10 mM, pH 7.3): To a solution of
NaCl (8.77 g, 150 mmol) in deionized water (500 mL) were added
successively NaH2PO4, H2O (0.26 g, 1.9 mmol) and Na2HPO4,
7H2O (2.17 g, 8.1 mmol). If necessary, the pH could be adjusted
to 7.3 by adding few drops of 1 M aqueous solutions of NaOH or
HCl. The volume was then completed to 1 L with deionized water.
PBST (phosphate saline-Tween 20 buffer, 6.4 mM, pH 7.2): NaCl
(9.00 g, 154 mmol) was dissolved in deionized water (500 mL).
NaH2PO4, H2O (0.32 g, 2.3 mmol), Na2HPO4, 7H2O (1.09 g,
ꢀ(
D
IC50-Me
-Man/IC50-Me
+D
IC50-Inhibitor5/0-IMCe 50-Inhibitor
)
and
a
a-Man
D
(r.p./n) = (IC
-Man/IC50-inhibitor) ꢀ (
a
D
IC
-Man/IC
+
50-Me
50-Me
a
50-Me
a
-Man
D
IC50-Inhibitor/IC50-Inhibitor)/n. If the glycopeptide dendrimers were
not completely soluble in pure PBS, DMSO (from 10% to 30%) could
be added without significant modification of the IC50 values (data
not shown).
4.3.3. Standard ELLA
Maxisorp microtitration plates were treated as described above.
Serial two-fold dilutions of the inhibitors were performed as previ-
ously mentioned and completed with a solution of HRP-conjugated
4.1 mmol) and Tween 20 (500 lL) were then added. The pH value
could also be corrected as mentioned above if necessary. Finally,
the volume was adjusted to 1 L by addition of deionized water. Tris
[tris(hydroxymethyl)aminomethane, 50 mM, pH 7.6]: Tris base
(6.05 g, 50 mmol) was dissolved in deionized water (800 mL).
The pH was then adjusted to 7.6 by adding successively concen-
trated and molar aqueous solution of HCl. The volume was finally
completed to 1 L with deionized water. Citrate–phosphate buffer
(200 mM, pH 4.0): Citric acid (7.70 g, 40 mmol) was dissolved in
PBS (100 mL). The pH was then adjusted to 4.0 by adding the min-
imal amount of aqueous solutions of NaOH (5.0 M and 1.0 M solu-
tions were successively used). Finally, the volume was completed
to 200 mL with PBS.
Con A in PBS (60 lL/well of a 5 l
g mL–1 solution). After preincuba-
tion for 1 h at 37 °C, the mixtures were transferred into Maxisorp
microtitration plates and incubated. These plates were washed
with three times with PBST (250
was added (50 L/well). After 20 min at rt, a molar solution of sul-
phuric acid was added (50 L/well) and the absorbances were
lL/well) and the ABTS solution
l
l
measured at 415 nm. The inhibition percentages, relative potencies
and uncertainties were then calculated as described previously.
Acknowledgments
Authors thank the University of Berne and the Swiss National
Science Foundation for financial support, and Dr. Tamis Darbre
(University of Berne, Switzerland) for her advices.
4.3.2. Enhanced-sensitivity ELLA
A solution of yeast mannan (10
diluting ten times a stock solution. CaCl2 and MnCl2 were then
added to reach the final concentrations of 1
mol mL–1 for each salt.
l
g mL–1) in PBS was prepared by
l
Supplementary data
Maxisorp microtitration plates were coated overnight with this di-
luted yeast mannan solution and the plates were washed three
Supplementary data (characterization data, including ELLAs,
NMR, MS spectra and HPLC profiles are provided for G2a–c, G2e–
g, G02S2, G12S2 and G22S2) associated with this article can be
found, in the online version, at doi:10.1016/j.bmc.2011.03.047.
times with PBST (250
1 h at 37 °C with PBSA [PBS containing 10 mg mL–1 of BSA (bovine
serum albumin), 100 L/well] and were washed as described above.
Separately, serial two-fold dilutions of inhibitors (60 L/well) were
lL/well). The wells were then blocked for
l
l
performed in nontreated mixing plates. Stock solutions of mono-
saccharides (1–2 M) and of glycopeptide derivatives (1–10 mM)
in PBS were then used. These wells were completed with a solution
References and notes
1. (a) Helms, B.; Frechet, J. M. J. Adv. Synth. Catal. 2006, 348, 1125; (b) Koefed, J.;
Reymond, J.-L. Curr. Opin. Chem. Biol. 2005, 9, 656; (c) Patri, A. K.; Kukowska-
Latallo, J. F.; Baker, J. R., Jr. Adv. Drug Delivery 2005, 57, 2203; (d) Lee, C. C.;
MacKay, J. A.; Frechet, J. M.; Szoka, F. C. Nat. Biotechnol. 2005, 23, 1517;
Newlome, G. R.; Moorfield, C. N.; Vögtle, F. Dendritic Molecules: Concepts,
Synthesis, Applications; VCH: Weinheim, 2001; (f) Smith, D. K.; Diederich, F. Top.
Curr. Chem. 2000, 210, 183; (g) Bosman, A. W.; Jansen, H. M.; Meijer, E. W.
Chem. Rev. 1999, 99, 1665; (h) Smith, D. K.; Diederich, F. Chem. Eur. J. 1998, 4,
1353; (i) Tomalia, D. A.; Dvornic, P. R. Nature 1994, 372, 617.
2. (a) Euzen, R.; Reymond, J.-L. Mol. BioSyst. 2011, 7, 411; (b) Johansson, E. M. V.;
Dubois, J.; Darbre, T.; Reymond, J.-L. Bioorg. Med. Chem. 2010, 18, 6589; (c)
Kolomiets, E.; Swiderska, M. A.; Kadam, R. U.; Johansson, E. M. V.; Jaeger, K.-E.;
Darbre, T.; Reymond, J.-L. Chem. Med. Chem. 2009, 4, 562; (d) Darbre, T.;
Reymond, J.-L. Curr. Top. Med. Chem. 2008, 8, 1286; (e) Johansson, E. M. V.;
Crusz, S. A.; Kolomiets, E.; Buts, L.; Kadam, R. U.; Cacciarini, M.; Bartels, K.-M.;
Diggle, S. P.; Cámara, M.; Wiliams, P.; Loris, R.; Nativi, C.; Rosenau, R.; Jaeger, K.-
E.; Darbre, T.; Reymond, J.-L. Chem. Biol. 2008, 15, 1249; (f) Johansson, E. M. V.;
Kolomiets, E.; Rosenau, F.; Jaeger, K.-E.; Darbre, T.; Reymond, J.-L. New J. Chem.
2007, 7, 1291; (g) Kolomiets, E.; Johansson, E. M. V.; Renaudet, O.; Darbre, T.;
Reymond, J.-L. Org. Lett. 2007, 9, 1465.
3. (a) Roy, R. Trends Glycosci. Glycotechnol. 2003, 15, 291; (b) Sadler, K.; Tam, J. P.
Rev. Mol. Biotechnol. 2002, 90, 195; (c) Turnbull, W. B.; Stoddart, J. F. Rev. Mol.
Biotechnol. 2002, 90, 231; (d) Dam, T. K.; Brewer, C. F. Chem. Rev. 2002, 102, 387;
(e) Lindhorst, T. K. Top. Curr. Chem. 2002, 218, 201; (f) Bertozzi, R.; Kiessling, L.
L. Science 2001, 291, 2357.
4. Gestwicki, J. E.; Cairo, C. W.; Strong, L. E.; Oetjen, K. A.; Kiessling, L. L. J. Am.
Chem. Soc. 2002, 124, 14922.
5. (a) Renaudet, O.; Dumy, P. Bioorg. Med. Chem. Lett. 2005, 15, 3619; (b)
Wittmann, V.; Seeberger, S. Angew. Chem., Int. Ed. 2004, 43, 900; (c) Scott, J. K.;
Loganathan, D.; Blaine Easley, R.; Gong, X.; Goldstein, I. J. Proc. Natl. Acad. Sci.
U.S.A. 1992, 89, 5398.
of biotin-conjugated Con A (60 lL/well of a 5 l
g mL–1 solution in
PBS) and the plates were incubated 1 h at 37 °C. These solutions
were transferred into the yeast mannan coated plates, incubated
1 h at 37 °C and washed as mentioned previously. The complex
Neutravidin™-biotinylated HRP was then preformed separately:
to Tris buffer (9.6 mL) were successively added
(1.2 mL) of Neutravidin™ (100
g mL–1 in Tris) and a solution
(1.2 mL) of biotin-conjugated HRP (25
g mL–1 in Tris). The mixture
was shaken for 30 min at rt and immediately transferred into the
yeast mannan-biotinylated Con A coated plates (60 L/well). After
L/well)
a solution
l
l
l
1 h at 37 °C, these plates were washed with Tris (2 ꢀ 250
l
and an ABTS solution (0.25 mg mL–1 in citric–phosphate buffer con-
taining 0.015% of a 30% aqueous solution of H2O2) was added
(50 lL/well). After 5 min at rt, the absorbances (Abs) were mea-
sured at 415 nm and the inhibition percentages were calculated
as follows: % Inhibition = 100 ꢀ (Absno inhibitor ꢁ Abswith inhibitor)/
Absno inhibitor. The concentration of the inhibitor leading to 50% of
the maximal activity was reported as IC50. Each test was done at
least as a triplicate and the relative potency of all glycoconjugates
was calculated from the IC50 of methyl
a-D-mannoside (Me a-
Man) measured in the same conditions and on the same plate.
The inhibitory properties of each neoglycoconjuagte were virtually
made independent of the experiment series by correcting the IC50
values obtained for the reference to a common experimental value
6. (a) Gouin, S. G.; García Fernández, J. M.; Vanquelef, E.; Dupradeau, F.-Y.;
Salomonsson, E.; Leffler, H.; Ortega-Muñoz, M.; Nilsson, U. J.; Kovensky, J.
Chem. Bio. Chem. 2010, 11, 1430; (b) Schwefel, D.; Maierhofer, C.; Beck, J. G.;
Seeberger, S.; Diederichs, K.; Möller, H. M.; Welte, W.; Wittmann, V. J. Am.
(the value of 1320
lM was chosen for the IC50 associated to methyl
a-D-mannoside among all the obtained measurements since this

R. Euzen, J.-L. Reymond / Bioorg. Med. Chem. 19 (2011) 2879–2887
2887
Chem. Soc. 2010, 132, 8704; (c) Roy, R.; Page, D.; Figueroa Perez, S.; Verez
Bencomo, V. Glycoconjugate J. 1998, 15, 251; (d) Page, D.; Roy, R. Glycoconjugate
J. 1997, 14, 345; (e) Page, D.; Zanini, D.; Roy, R. Bioorg. Med. Chem. 1996, 4, 1949.
7. (a) Schierholt, A.; Hartmann, M.; Schwekendieck, K.; Lindhorst, T. K. Eur. J. Org.
Chem. 2010, 3120; (b) Xu, H.; Regino, C. A. S.; Koyama, Y.; Hama, Y.; Gunn, A. J.;
Bernardo, M.; Kobayashi, H.; Choyke, P. L.; Brechbiel, M. W. Bioconjugate Chem.
2007, 18, 1474; (c) Darbre, T.; Reymond, J. L. Acc. Chem. Res. 2006, 39, 925; (d)
Tomalia, D. A.; Huang, B.; Swanson, D. R.; Brothers, H. M., II; Klimash, J. W.
Tetrahedron 2003, 59, 3799.
8. For selected publications, see: (a) Delort, E.; Szöcs, E.; Widmer, R.; Siegenthaler,
H.; Reymond, J.-L. Macromol. Biosci. 2007, 7, 1024; (b) Delort, E.; Nguyen-Trung,
N.-Q.; Darbre, T.; Reymond, J.-L. J. Org. Chem. 2006, 71, 4468; (c) Clouet, A.;
Darbre, T.; Reymond, J.-L. Biopolymers 2006, 84, 114.
9. For recent articles, see: (a) Biswas, R.; Maillard, N.; Kofoed, J.; Reymond, J.-L.
Chem. Commun. 2010, 8746; (b) Uhlich, N. A.; Natalello, A.; Kadam, R. U.;
Doglia, S. M.; Reymond, J.-L.; Darbre, T. Chem. Bio. Chem. 2010, 11, 358; (c)
Uhlich, N.; Sommer, P.; Bühr, C.; Schürch, S.; Reymond, J.-L.; Darbre, T. Chem.
Commun. 2009, 6237; (d) Javor, S.; Reymond, J.-L. Isr. J. Chem. 2009, 49, 129; (e)
Sommer, P.; Fluxa, V. S.; Darbre, T.; Reymond, J.-L. Chem. Bio. Chem. 2009, 10,
1527; (f) Maillard, N.; Darbre, T.; Reymond, J.-L. J. Comb. Chem. 2009, 11, 667;
(g) Javor, S.; Reymond, J.-L. J. Org. Chem. 2009, 74, 3665; (h) Maillard, N.; Clouet,
A.; Darbre, T.; Reymond, J.-L. Nat. Protoc. 2009, 4, 132; (i) Javor, S.; Natalello, A.;
Doglia, S. M.; Reymond, J.-L. J. Am. Chem. Soc. 2008, 130, 17248; (j) Sommer, P.;
Uhlich, N.; Reymond, J.-L.; Darbre, T. Chem. Bio. Chem. 2008, 9, 689.
10. (a) Gómez-García, M.; Benito, J. M.; Gutiérrez-Gallego, R.; Maestre, A.; Ortiz
Mellet, C.; García-Fernandez, J. M.; Jiménez Blanco, J. L. Org. Biomol. Chem. 2010,
8, 1849; (b) Mangold, S. L.; Cloninger, M. J. Org. Biomol. Chem. 2006, 4, 2458; (c)
Kanai, M.; Mortell, K. H.; Kiessling, L. L. J. Am. Chem. Soc. 1997, 119, 9931; (d)
Toone, E. J. Curr. Opin. Struct. Biol. 1994, 4, 719; (e) Williams, B.; Chervenak, M.;
Toone, E. J. J. Biol. Chem. 1992, 267, 22907.
11. Chan, W. C.; White, P. D. Fmoc Solid Phase Peptide Synthesis:
Approach; Oxford University Press: New York, 2004.
A Practical
12. Ponpipom, M. M.; Bugianesi, R. L.; Robbins, J. C.; Doebber, T. W.; Shen, T. Y. J.
Med. Chem. 1981, 24, 1388.
13. Lateral chains of amino acids were deprotected concomitantly during work-up.
14. Duk, M.; Lisowska, E.; Wu, J. H.; Wu, A. M. Anal. Biochem. 1994, 221, 266.
15. Lee, Y.; Lee, R. Acc. Chem. Res. 1995, 28, 321.
16. (a) Lundquist, J. J.; Toone, E. J. Chem. Rev. 2002, 102, 555; (b) Lundquist, J. J.;
Toone, E. J. Tetrahedron: Asymmetry 2000, 11, 95.
17. Surprisingly, G2e seemed to be easily prone to fragmentation reactions when
positive mass spectrum was recorded. Then, negative mode spectrum had to be
considered from a basic hydro-organic solution of G2e. As a consequence, a
base peak corresponding to the ionisation cluster [M+2H2O+2OH]2ꢁ was
observed.