Non-covalent polyvalent ligands by self-assembly of small glycodendrimers: A novel concept for the inhibition of polyvalent carbohydrate-protein interactions in vitro and in vivo

Article abstract of DOI:10.1002/chem.200500901

Polyvalent carbohydrate-protein interactions occur frequently in biology, particularly in recognition events on cellular membranes. Collectively, they can be much stronger than corresponding monovalent interactions, rendering it difficult to control them with individual small molecules. Artificial macromolecules have been used as polyvalent ligands to inhibit polyvalent processes; however, both reproducible synthesis and appropriate characterization of such complex entities is demanding. Herein, we present an alternative concept avoiding conventional macromolecules. Small glycodendrimers which fulfill single molecule entity criteria self-assemble to form non-covalent nanoparticles. These particles - not the individual molecules - function as polyvalent ligands, efficiently inhibiting polyvalent processes both in vitro and in vivo. The synthesis and characterization of these glycodendrimers is described in detail. Furthermore, we report on the characterization of the non-covalent nanoparticles formed and on their biological evaluation.

Full text of DOI:10.1002/chem.200500901

FULL PAPER
DOI: 10.1002/chem.200500901
Non-Covalent Polyvalent Ligands by Self-Assembly of Small
Glycodendrimers: A Novel Concept for the Inhibition of Polyvalent
Carbohydrate–Protein Interactions In Vitro and In Vivo
Gebhard Thoma,*[a] Markus B. Streiff,[a] Andreas G. Katopodis,[a] Rudolf O. Duthaler,[a]
Nicolas H. Voelcker,[b] Claus Ehrhardt,[a] and Christophe Masson[a]
Abstract: Polyvalent carbohydrate–pro-
tein interactions occur frequently in
biology, particularly in recognition
events on cellular membranes. Collec-
tively, they can be much stronger than
corresponding monovalent interactions,
rendering it difficult to control them
with individual small molecules. Artifi-
cial macromolecules have been used as
polyvalent ligands to inhibit polyvalent
processes; however, both reproducible
synthesis and appropriate characteriza-
tion of such complex entities is de-
manding. Herein, we present an alter-
native concept avoiding conventional
macromolecules. Small glycodendrim-
ers which fulfill single molecule entity
criteria self-assemble to form non-co-
valent nanoparticles. These particles—
not the individual molecules—function
as polyvalent ligands, efficiently inhib-
iting polyvalent processes both in vitro
and in vivo. The synthesis and charac-
terization of these glycodendrimers is
described in detail. Furthermore, we
report on the characterization of the
non-covalent nanoparticles formed and
on their biological evaluation.
Keywords: carbohydrate–protein
interactions · dendrimers · multi-
valency · polyvalency · self-assembly
Introduction
erties can be unique and substantially different from proper-
ties displayed by monovalent interactions of their constitu-
ents.[2] In particular, their affinity can be much greater than
in the case of a monovalent interaction between a carbohy-
drate and a protein (which is generally weak; KD = 10À3–
10À4 mÀ1).[3] As a consequence, the inhibition of disease-rele-
vant polyvalent biological processes with monovalent oligo-
saccharides has proven difficult. However, approaches utiliz-
ing non-natural polyvalent glycopolymers have led to highly
potent receptor blockers (Figure 1).[4]
The interaction of an immunoglobulin of class IgM with
multiple antigens presented on a cell surface is a typical
polyvalent recognition process (Figure 2c).[5] IgMs have ten
identical binding sites and consist of five monomeric subu-
nits, joined by disulfide bridges, each of them having a mo-
lecular weight of ꢀ180 kDa (Figure 2b). As an example,
the interaction of natural human or non-human primate anti
aGal IgMs and the aGal epitope (Gala1–3Galb1–4GlcNAc;
Figure 2a), which is expressed on the cells of mammals but
not of humans and Old World monkeys, causes hyperacute
rejection, thus hampering pig to primate xenotransplanta-
tion.[6] Removal or blockade is a potential prevention thera-
py. While blocking attempts with the monovalent trisacchar-
ide were unsuccessful, polymers providing multiple copies of
the aGal trisaccharide efficiently inhibited anti aGal IgMs
Polyvalent carbohydrate–protein interactions occur fre-
quently in recognition events on cellular membranes.[1] They
are characterized by the simultaneous contact of multiple
oligosaccharides (ligands)on one biological entity with mul-
tiple proteins (receptors)on another. Their collective prop-
[a] Dr. G. Thoma, Dr. M. B. Streiff, Dr. A. G. Katopodis,
Dr. R. O. Duthaler, Dr. C. Ehrhardt, Dr. C. Masson++
Novartis Institutes for BioMedical Research
Lichtstrasse 35, 4056 Basel (Switzerland)
Fax : (+41)61-324-6735
E-mail: gebhard.thoma@novartis.com
[b] Dr. N. H. Voelcker+
The Scripps Research Institute
Beckman Centre for Chemical Sciences
10550, N. Torrey Pines Rd., La Jolla, CA 92034 (USA)
[+] Present address: Flinders University of South Australia
School of Chemistry, Physics and Earth Sciences
GPO Box 2100 Bedford Park, South Australia 5042 (Australia)
[
++] Present address: GENFIT S.A.
885, avenue Eug›ne AvinØe
59120 Loos (France)
Supporting information for this article (1H NMR spectra of com-
pounds 11Gal, 19Gal, 24Gal, 29Gal, and 30Gal)is available on the
WWW under http://www.chemeurj.org/ or from the author.
Chem. Eur. J. 2006, 12, 99 – 117
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
99

Figure 2. Multivalent interaction between an IgM and the aGal epitope:
a)representation of the aGal trisaccharide; b)schematic representation
of a circulating, planar IgM directed against the aGal epitope; c)upon
polyvalent binding to multiple aGal trisaccharides presented on a cell
surface, the IgM adopts a staple form causing complement activation and
cell destruction.
Figure 1. Inhibition of a physiological polyvalent carbohydrate (ligand)–
protein (receptor)interaction with a)a monovalent mimic of the ligand
and b)an artificial polyvalent ligand. Polyvalent interactions can be
orders of magnitude stronger than monovalent interactions. Thus, in case
a)inhibition will only occur if the mimic of the physiological ligand is
orders of magnitude more potent than a single physiological ligand, or if
it is used at very high (millimolar)concentrations. However, in b)a phys-
iological polyvalent interaction competes with an artificial polyvalent in-
teraction. Appropriate design of the artificial polyvalent ligand can lead
to inhibition of the physiological process even if the “monovalent” phys-
iological ligand and its “monovalent” mimic show comparable potencies.
in vitro and in vivo. However, macromolecules with a mo-
lecular mass comparable to the mass of an IgM
[7,8]
(ꢀ1000 kDa)had to be used.
As both the reproducible preparation and the appropriate
characterization of non-homogeneous macromolecules are
very demanding, an alternative concept for the control of
polyvalent interactions with well-characterized small mole-
cules is desirable (Figure 3). Such an approach could be
based on the supramolecular chemistry[9] of compounds
comprising a self-assembling moiety (gray square)and a
ligand moiety (red ellipse). Ideally, intermolecular recogni-
tion of self-assembling moieties will lead to non-covalent
particles presenting multiple ligands. These aggregates—but
not the individual molecules (which are far too small)—can
function as polyvalent receptor blockers. Interactions of the
individual ligands with the polyvalent receptor contribute to
the stability of the receptor–ligand complex. The strength of
the inhibitory complex is best for a maximal number of both
the contacts between individual self-assembling moieties
and the individual ligand–receptor interactions. Therefore,
in the case of a dynamic, reversible self-assembly process,
the polyvalent receptor can be utilized as a template to opti-
mize its own polyvalent inhibitor (Figure 3).
Figure 3. Formation of a tailored polyvalent receptor–ligand complex by
dynamic self-assembly of small molecules comprising both a self-assem-
bling moiety (grey square) and a ligand moiety (red ellipse). Scenario a):
A preformed polyvalent ligand makes contact with the receptor. Further
optimization of the ligand-receptor complex occurs utilizing the poly-
valent receptor as a template. Scenario b): A relatively small, oligovalent
aggregate binds weakly to the polyvalent receptor. In a second step, the
highly potent polyvalent ligand “grows” on the surface of the receptor
leading to an optimized receptor–ligand complex.
In a preliminary communication, we recently published
our findings on novel glycodendrimers containing endgroups
modified with the aGal trisaccharide (ligand).[10,11] We
showed that, in aqueous solution, molecules with appropri-
100
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 99 – 117

Inhibition of Polyvalent Carbohydrate–Protein Interactions with Glycodendrimers
FULL PAPER
ate core structures (self-assembling moiety)self-assemble to
form non-covalent nanoparticles with molecular weights of
up to 7000 kDa. These particles — but not the individual
molecules — were found to efficiently inhibit the aGal–IgM
interaction, both in vitro and in vivo.[10] Here, we discuss the
synthesis and characterization of these and additional glyco-
dendrimers and present new evidence that self-assembly
strongly depends on the structure and size of the dendrimer
core. We show that particle formation is a dynamic process,
which is a prerequisite for the formation of an optimized re-
ceptor–ligand complex.
Syntheses and characterization of dendrimer core structures:
Dendrimer core structures were prepared by a convergent
“outside–in” approach[12] using a single, three-directional
building block (1)to allow for a minimum number of trans-
formations for dendrimer tier construction. The orthogonal-
ly protected, rigid wedge 1 consists of three aromatic rings
linked via amide bonds. Compound 1 displays a benzoate
ester function and two benzylic amines. It was prepared by
amide coupling of 2 and 3. Saponification gave carboxylate
4, whereas acid treatment furnished diamine 5 (Scheme 1).
Scheme 2.
endgroups)was obtained by reacting 6 with 0.5 equivalents
of 5 under the same reaction conditions. The fourth-, fifth-,
and sixth-generation core structures with 16 endgroups (8),
32 endgroups (9), and 64 endgroups (10), respectively, were
prepared following the same procedures (Scheme 3).
1
The integrity of the dendrimer cores was assessed by H
NMR spectroscopy using the integral values of indicative
protons (Figure 4). All of the compounds contain two types
of benzylic protons, giving rise to two distinct signals (see
structures in Figure 4). They are either adjacent to a carba-
mate (yellow)or adjacent to a benzamide (blue.) The ratio
of these two types of benzylic protons changes characteristi-
cally from generation to generation (see table in Figure 4).
The changes are considerable only up to the fourth genera-
tion core 8 (16 endgroups). For the highest generation cores
(9, 10), the ratios of the integrals of “yellow” and “blue”
protons are very similar and close to unity, impeding an ac-
curate distinction based on NMR. All measured and expect-
ed values were in agreement within the experimental error,
but for the reason mentioned above the data can only prove
the integrity of the cores up to the fourth generation. The
dendrimer core structures contain two types of trisubstituted
aromatic rings. These rings bear either a carboxylic acid or a
carboxamide moiety. The protons in the para positions give
rise to two characteristic signals (para to carboxylic acid:
red; para to carboxamide: green). Whereas all of the den-
drimers have only one “red” proton, the number of “green”
protons increases from generation to generation (see table
in Figure 4). The ratios can be determined very accurately
up to the fourth generation core 8. For the highest genera-
tion cores, 9 and 10, the ratios are very high (30:1, 60:1),
hampering a highly accurate determination. All measured
and expected values were in agreement within the experi-
Scheme 1. HBTU=O-benzotriazol-1-yl-N,N,N’,N’-tetramethyluronium
hexafluorophosphate; DMF=dimethylformamide.
The second-generation core 6 (four endgroups)was as-
sembled from 4 and 0.5 equivalents of 5 in DMF/iPr2NEt
using HBTU to promote the amide bond formation
(Scheme 2). To ensure complete consumption, it was impor-
tant to use a large excess of iPr2NEt. After completion of
the coupling, water and LiOH were added to the reaction
mixture to cleave the methyl ester. The product was isolated
by precipitation by adding either water or a less polar or-
ganic solvent. Further purification could be achieved by
adding a solution of the precipitate in DMF to either water
or an organic solvent. The third generation core 7 (eight
Chem. Eur. J. 2006, 12, 99 – 117
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
101

G. Thoma et al.
Scheme 3.
mental error. We conclude that all of the dendrimer core
structures are highly homogeneous. Up to the fourth genera-
tion core 8 (16 endgroups), their purity is >95%. This is re-
markable considering that not a single chromatographic pu-
rification was required. The convergent synthesis approach
allows the preparation of a subsequent dendrimer genera-
tion with nearly double the molecular weight by forming
just two amide bonds. Thus, side products stemming from in-
complete reaction and unconsumed starting materials differ
significantly from the product in terms of molecular weight
102
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 99 – 117

Inhibition of Polyvalent Carbohydrate–Protein Interactions with Glycodendrimers
FULL PAPER
Figure 4. 1H NMR spectra (d=4–9 ppm)of dendrimer cores 6–10 (600 MHz; DMSO/D2O, 75 8C). The color code for the assignment of selected charac-
teristic signals is illustrated for compounds 6 and 7. The table shows the expected and the measured ratios of the integrals of the selected signals. *signal
stems from CH2Cl2.
and size, allowing for purification by simple precipitation.
We did not attempt to prepare larger dendrimers because of
the envisaged difficulties in suitably characterizing these
molecules.
spectively. Compound 29 was prepared using building block
31, which was obtained from 32 and 3 (Scheme 7). Depro-
tection of 31 gave 33, which was coupled with 4 to furnish
29. Compound 30 was prepared from building block 34,
which, in turn, was obtained from benzoate 35 and 2
(Scheme 8). Compound 35 was available from 36, which was
protected to furnish 37. Subsequent N-methylation yielded
38. Deesterification and concomitant removal of the N-pro-
tecting groups gave access to 35. Selective deprotection of
34 gave 39 and 40, which were coupled to furnish 30.
Several analogues of 6 with more aliphatic core structures
were prepared. Compound 11 contains branched aliphatic
elements instead of the three trisubstituted aromatic rings of
6 (Scheme 4). Coupling of 2 with two equivalents of 12 gave
ketone 13, which was reacted with 14 to furnish alkene 15.
In some experiments, an isomer of 15 with the double bond
adjacent to a nitrogen atom was isolated. Subsequent hydro-
genation of 15 or its isomer gave 16. Building blocks 17 or
18 were obtained from 16 by selective deprotection. Den-
drimer 11 was prepared by condensation of 17 and 18.
Dendrimer 19 has cyclohexyl rings instead of the six dis-
ubstituted aromatic rings of 6. The core was assembled from
cyclohexyl carboxylic acid 20 and benzoate 3 (Scheme 5).
Building block 21 was selectively deprotected, and this was
followed by amide coupling of the resulting acid 22 and dia-
mine 23. Finally, the more flexible core structure 24 with
linear alkyl chains (C6)instead of the disubstituted aromatic
rings of 6 was prepared analogously to 19 using hexanoic
acid 25 instead of 20 (Scheme 6).
Functionalization of dendrimer core structures and charac-
terization of the glycodendrimers: The convergent synthesis
of the core structures requires just two transformations per
individual molecule to obtain the next generation dendrim-
er, which has quite distinct properties compared to the start-
ing materials and potential side products. On the contrary,
depending on the size of the core structure, modifications of
the endgroups can involve a large number of transforma-
tions per individual molecule. Thus, side products stemming
from incomplete reactions can have very similar properties
compared to the desired product, rendering purification
very difficult. To obtain highly homogeneous glycodendrim-
ers, all chemical reactions involved in the functionalization
need to proceed in near-quantitative yield. We relied on a
The second-generation core structures 29 and 30 bear me-
thylated benzamide and methylated anilide functions, re-
Chem. Eur. J. 2006, 12, 99 – 117
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
103

G. Thoma et al.
Scheme 5.
Scheme 4.
strategy which we had previously applied to the highly pre-
dictive
functionalization
of
polylysine.[4b]
The
butoxycarbonyl(Boc)-protected core structures 4, 6, 7, 8, 9,
10, 11, 19, 24, 29,and 30 were transformed to the free
amines using trifluoroacetic acid (TFA)/water in the pres-
ence of mercaptoethanol to trap tert-butyl cations,thus hin-
dering the formation of tert-butylamino groups. The benzylic
amines were isolated as TFA salts by precipitation in diethyl
ether. The endgroups were further activated by introduction
of a chloroacetamide group. Using a large excess of the ster-
ically hindered weak base 2,6-lutidine and chloroacetic an-
hydride in DMF led to complete conversion. The chloroace-
tamides were isolated by precipitation in diethyl ether. Gly-
codendrimers containing the aGal trisaccharide were ob-
tained by reacting the activated core structures with a small
excess of aGal-SH[7a] in DMF (Figure 5). Addition of 1,8-
diazabicyclo[5.4.0]undecene (DBU) led to complete func-
tionalization within a few minutes. The crude products were
precipitated,redissolved in water,and further purified by ul-
Scheme 6.
trafiltration. Glycodendrimer 7Lac with eight lactose resi-
dues was obtained using Lac-SH instead of aGal-SH.
All of the glycodendrimers were characterized by NMR
spectroscopy. The 1H NMR spectra of 6Gal, 7Gal, 8Gal,
9Gal,and 10Gal in DMSO/D2O (5:1) at 808C are shown in
Figure 6. The values of the integrals of characteristic signals
104
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA,Weinheim
Chem. Eur. J. 2006, 12,99 – 117

Inhibition of Polyvalent Carbohydrate–Protein Interactions with Glycodendrimers
FULL PAPER
Scheme 7.
Figure 5. Synthesis of glycodendrimers 4Gal, 6Gal, 7Gal, 8Gal, 9Gal,
10Gal, 11Gal, 19Gal, 24Gal, 29Gal, 30Gal,and 7Lac by functionaliza-
tion of endgroups of Boc-protected core structures 4, 6, 7, 8, 9, 10, 11, 19,
24, 29,and 30 (for simplicity only one endgroup is shown).
were used to quantify the degree of endgroup functionaliza-
tion. The signals of internal benzylic hydrogens of the den-
drimer core (dꢀ4.7 ppm; green) and the 1-hydrogens of the
terminal a-linked galactoses (dꢀ5.0 ppm; red) were always
separated at the low field of the aliphatic proton region.
Their ratios were characteristic for each dendrimer genera-
tion (Figure 7). Expected and measured values were in very
good agreement,indicating complete functionalization with
the aGal trisaccharide in all cases (see Figure 6). It is impor-
tant to note that for higher generation glycodendrimers the
changes of this ratio are very small as it approaches a value
of 2. Thus,in the cases of 9Gal and 10Gal,with 32 and 64
endgroups,respectively,the NMR data do not prove com-
plete homogeneity but indicate functionalization of all
endgroups. In the cases of 6Gal, 7Gal,and 8Gal with 4,8,
and 16 endgroups,respectively,the data indicate very high
purities (>95% for 6Gal and 7Gal; >90% for 8Gal).
Scheme 8.
Chem. Eur. J. 2006, 12,99 – 117
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA,Weinheim
www.chemeurj.org
105

G. Thoma et al.
(2200 kDa) and 8Gal (1200 kDa) were still very high (250
and 70 individual molecules/particle,respectively). The fifth
and sixth generation glycodendrimers 9Gal and 10Gal did
not aggregate considerably (7 and 3 individual molecules/
particle,respectively). Very interestingly,considering the
second to the sixth glycodendrimer generations,the molecu-
lar weight of the particles formed was seen to decrease with
increasing molecular weight of the individual molecules.
The particle weight was not found to depend significantly
on the concentration,as shown for dendrimer
6Gal
ꢁ1
(7100 kDa at 0.06 mgmLꢁ1,6000 kDa at 0.03 mgmL
,
3700 kDa at 0.01 mgmLꢁ1),but is significantly affected by
the temperature (Figure 9). It was seen to decrease with in-
creasing temperature (tenfold on going from 308C to 708C).
1
The H NMR spectra in D2O are also indicative of the for-
mation of particles. As an example,the spectrum of 6Gal
showed very broad signals in the aromatic region (7–8 ppm;
dendrimer core structure; Figure 9). With increasing temper-
ature,the signals sharpened. We believe that broad signals
are indicative of large particles,and that these fall apart at
elevated temperatures leading to sharp signals as typically
observed for individual small molecules. In DMSO/D2O
(5:1) mixtures,sharp signals were observed even at low tem-
peratures. It is likely that aggregation is inhibited by organic
solvents.
Comparable molecular weights of the particles and very
similar NMR spectra were observed at identical tempera-
tures when a hot solution was cooled or when a cold solu-
tion was heated. Thus,self-assembly is a dynamic process
which rapidly establishes thermodynamic equilibrium
(Figure 9).
Figure 6. 1H NMR spectra (600 MHz,DMSO/D 2O,5:1,80 8C) of 6Gal,
7Gal, 8Gal, 9Gal,and 10Gal. The protons giving rise to the signals
marked in red and green are shown for glycodendrimer 8Gal (Figure 7).
The particle weight proved to be strongly dependent on
the core structure,as revealed by studying the aggregation
properties of glycodendrimers with small core modifications
compared to 6Gal and 7Gal. Glycodendrimer 24Gal,con-
taining a flexible aliphatic spacer in place of the disubstitut-
ed aromatic rings of 6Gal,did not aggregate at all. Com-
pound 19Gal,with cyclohexyl rings instead of the disubsti-
tuted aromatic rings of 6Gal,which is more rigid than
24Gal but less aromatic than 6Gal,self-assembled but
formed aggregates less than a tenth of the size of those with
6Gal. Glycodendrimer 11Gal,with branched aliphatic re-
placements of the trisubstituted aromatic rings,formed very
small aggregates (133 kDa). The third generation com-
pounds 7Lac (8”Lac; 1900 kDa; 250 individual molecules/
particle) and 7Gal (8”aGal; 2200 kDa; 265 individual mol-
ecules/particle),with identical backbones but different car-
bohydrate endgroups,formed aggregates of very similar
size. Thus,the particle weight showed no strong dependence
on the size of the hydrophilic endgroups.
Satisfying MALDI mass spectra were obtained for all of
the glycodendrimers with the exception of 9Gal and 10Gal.
The spectrum of 7Gal with eight endgroups is shown in
Figure 8. The expected and measured molecular weights of
glycodendrimer 7Gal are in agreement within the experi-
mental error. Several lower molecular weight signals were
detected. According to NMR,the glycodendrimer 7Gal ex-
hibits very high purity. Therefore,we believe that the addi-
tional signals are caused by fragmentation. Fragmentation
was also observed for all of the other glycodendrimers,even
for the first generation compound 4Gal which had been pu-
rified by chromatography,and could not be avoided when
different matrices were used. Because of the aggregation of
our glycodendrimers,both HPLC and size-exclusion chro-
matography proved to be inappropriate for analyzing these
molecules.
Characterization of the non-covalent nanoparticles: The
self-assembly of our glycodendrimers to form non-covalent
nanoparticles was studied by multi-angle light scattering in
water (MALS; Table 1). Accordingly,the first generation
dendrimer 4Gal forms small aggregates (50 kDa),whereas
6Gal forms large particles of 7100 kDa (1500 individual
molecules/particle). The particle weights obtained for 7Gal
In aqueous solution,intermolecular hydrogen bonds be-
tween individual glycodendrimers seem not to be significant-
ly involved in self-assembly since particle weights were not
reduced by the H-bond disrupting reagent guanidinium hy-
drochloride (1.5m),as determined for 6Gal by light scatter-
ꢁ1
ing (7100 kDa at 0.06 mgmLꢁ1,6000 kDa at 0.03 mgmL
,
3700 kDa at 0.01 mgmLꢁ1). In addition,the 1H NMR signals
106
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA,Weinheim
Chem. Eur. J. 2006, 12,99 – 117

Inhibition of Polyvalent Carbohydrate–Protein Interactions with Glycodendrimers
FULL PAPER
Figure 7. Structure of glycodendrimer 8Gal. The integrals of the internal benzylic protons (green) and the H-1 protons of the terminal a-linked galactose
(red) have been used to quantify the degree of endgroup functionalization (Figure 6).
intramolecular hydrogen bonds formed in 6Gal,but could
also be explained by steric effects of the additional methyl
groups on the core conformation.
Rigid,highly aromatic core structures are required for
self-assembly. We believe that intramolecular p-stacking
leads to a pre-organized core conformation that allows
core–core contacts. Self-assembly to form large particles can
occur if a glycodendrimer core can be in contact with at
least two other core structures at the same time. The core of
4Gal seems to be too small for efficient core–core interac-
tions,whereas the second generation core of 6Gal is optimal
for self-assembly. The decreasing particle weight obtained
for higher generation glycodendrimers (7Gal–10Gal) could
be explained by the increasing size of the dendrimer core
and increasing number of large endgroups per individual
molecule inducing a more globular shape. As a consequence,
Figure 8. MALDI-MS of glycodendrimer 7Gal (matrix: 2,5-dihydroxycin-
namic acid); calcd. for the Na salt: C393H525N44NaO166S8 8801.21; found
8801 (broad signal).
the core is more efficiently shielded by carbohydrates,ren-
dering intermolecular core–core contacts more difficult. In-
creased crowding in the core with each generation number
also stabilizes interior binding through p–p interactions and
of 6Gal in D2O at room temperature remained very broad
in the presence of guanidinium deuteriumchloride (6m).
Compound 29 with two methylated benzamides self-assem-
bled to still relatively large particles (1100 kDa,260 individ-
ual molecules/particle),whereas compound 30 with six me-
thylated anilides formed only small aggregates (315 kDa,70
individual molecules/particle; Table 1). It is important to
note that the reduced tendency to form aggregates observed
for 29 and 30 is not necessarily caused by the breakdown of
hydrophobic contacts,which further reduce the conforma-
tional flexibility of the glycodendrimer.
This hypothesis was supported by molecular modeling.
Conformational analyses of the dendrimer core structures of
compounds 6Gal and 7Gal were carried out by applying a
combined MC/LMOD calculation in vacuum. All of the tri-
saccharide endgroups were removed to reduce the computa-
tion time. By visual inspection,low-energy conformations of
the core of 6Gal were selected that allow both re-attach-
Chem. Eur. J. 2006, 12,99 – 117
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA,Weinheim
www.chemeurj.org
107

G. Thoma et al.
Table 1. Properties of the Glycodendrimers.
ment of the four trisaccharides
without steric interference and
solvation of the carbohydrates.
Most of these conformations
were compact and disk-like,as
exemplified by the structure
IgM[b]
IC50 [mm]
Haemolysis[b]
IC50 [mm]
[a]
Compound
(endgroups)
MW
(ind. mol.) [kDa]
MWaggr
[kDa]
4Gal (2)
6Gal (4)
7Gal (8)
7Lac (8)
8Gal (16)
9Gal (32)
10Gal (64)
11Gal (4)
19Gal (4)
24Gal (4)
29Gal (4)
30Gal (4)
1.908
4.198
8.779
7.154
17.941
36.264
72.911
4.136
4.234
4.078
4.226
4.282
50[c]
7100[d]
2200[c]
1900[c]
1200[c]
270[c]
200[c]
133[d]
571[d]
7[c]
>10
0.025
0.010
>100
0.019
>1
>1
>1
>1
>1
>100
0.035
0.010
n.d.
with
the
lowest
energy
0.180
0.550
2.240
n.d.
(Figure 10). Parallel and edge-
to-face p–p interactions of the
phenyl rings lead to the forma-
tion of two hydrophobic surfa-
ces—the front and back side of
n.d.
>100
n.d.
1100[d]
315[d]
0.170
>1
n.d.
the disk—which remain accessi-
ble even with the trisaccharides
attached. Thus,intermolecular
core–core contacts are readily
possible and aggregation can
[a] The weight-average molar mass (MWaggr) for all of the glycodendrimers was determined by multi-angle
light scattering (MALS). Polydispersity values (dn/dc) were between 0.18 and 0.22 mLgꢁ1. [b] For active com-
pounds,the average of at least two experiments is quoted (variations within a factor of 2.5). Concentrations
refer to equivalent concentration of trisaccharide,not to concentration of oligovalent glycodendrimer. [c] Test
concentration: 0.3 mgmLꢁ1. [d] Test concentration: 0.06 mgmLꢁ1
.
occur. Although the calcula-
tions were performed in vacuo,
an aqueous environment with a
high dielectric constant would also favor the formation of
compact core structures. Low energy conformations of the
dendrimer core of compound 7Gal were also found to be
compact. However,due to the larger size of the molecule,
they are less disk-like and rather spherical. In addition,it is
more difficult to attach the eight trisaccharides without
shielding the hydrophobic core surface. Most conformations
showed only one hydrophobic surface,which could lead to
pair formation but not to aggregation to form large particles.
Only a few conformations are suited for aggregation,pro-
viding at least two accessible hydrophobic surfaces (DE to
the absolute minimum >6 kcal). Compared to 6Gal,the
more globular shape of the core of 7Gal leads to less effi-
cient intermolecular core–core contacts,thereby rationaliz-
ing the formation of smaller aggregates. These effects are
most probably even more pronounced for the higher genera-
tion dendrimers 8Gal–10Gal,resulting in the observed de-
creasing tendency to aggregate. Interestingly,compared to
6Gal,low energy conformations of the core of its partially
N-methylated derivative 30Gal were less compact and did
not provide accessible hydrophobic surfaces for aggregation.
This could explain the formation of only small aggregates.
The nanoparticles formed by our glycodendrimers can be
deposited on surfaces and then investigated by atom force
microscopy (AFM) and transmission electron microscopy.
Aqueous solutions of 6Gal–10Gal were applied to hydro-
philic mica surfaces and the solvent was evaporated. By
atom force microscopy,we observed disk-shaped particles.
Their diameters decreased with increasing mass of the indi-
vidual molecules (61 nm, 6Gal; 13 nm, 10Gal),showing the
same trend as the molecular weights in solution (Fig-
ure 11a). The size distribution for the individual generations
was found to be remarkably homogeneous (within a factor
of two). The particles exhibit a pronounced disk shape.
Their flatness (1–5 nm) might be caused by interactions with
the mica surface. The disk-like morphology might be slightly
disrupted by AFM tip convolution artifacts. Particles of
Figure 9. 1H NMR spectra (500 MHz; D2O) of glycodendrimers 6Gal and
24Gal. a) At 308C 6Gal shows very broad signals for the dendrimer core
structure (d=7–8 ppm),which become sharpened at elevated tempera-
tures. The weight-average molecular mass of the aggregates formed in
water (MWaggr) was determined by multi-angle light scattering (MALS).
Accordingly,big aggregates self-assemble at low temperatures but disso-
ciate at elevated temperatures. Small aggregates or individual molecules
gave rise to sharp NMR signals,whereas big aggregates led to very broad
signals. Very similar MWaggr and NMR spectra were obtained upon heat-
ing a cold aqueous solution of 6Gal or upon cooling a hot solution. Thus,
self-assembly of glycodendrimer 6Gal is a dynamic process that rapidly
reaches thermodynamic equilibrium. b) 24Gal,which does not form par-
ticles (MWaggr = 7 kDa),showed very sharp signals at 25 8C.
108
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA,Weinheim
Chem. Eur. J. 2006, 12,99 – 117

Inhibition of Polyvalent Carbohydrate–Protein Interactions with Glycodendrimers
FULL PAPER
Figure 10. Minimum energy conformation of the core of 6Gal obtained by MC/LMOD calculations in vacuum. The core adopts a disk-like conformation.
The gray arrows indicate the hydrophobic front and back sides of the disk,which are suited for intermolecular core–core contacts allowing aggregati on
to large particles even in the presence of the trisaccharides.
6Gal were also deposited on a more hydrophobic polyfor-
maldehyde surface and investigated by transmission electron
microscopy. The particle diameters were determined to be
between 20 and 120 nm,the majority having a diameter of
50 nm (Figure 11b). This is in good agreement with the
AFM results on mica. We conclude that the particle size is
independent of the properties of the surface on which they
are deposited,indicating that the particles are preformed in
solution.
tency clearly correlates with the size of the aggregates,and
not with the size of the individual molecules. The inhibition
data suggest that optimal particle weight and size for IgM
inhibition are obtained with 7Gal. The particles formed by
6Gal might be slightly too big,but still showed high poten-
cy. The aggregates formed by 4Gal, 8Gal, 9Gal,and 10Gal
are apparently too small to accomplish polyvalent amplifica-
tion of IgM inhibition. Potency is strongly dependent on the
aGal trisaccharide because the aggregating control dendrim-
er 7Lac (identical backbone and similar aggregation proper-
ties as 7Gal,but Lac instead of aGal) did not inhibit bind-
ing. Thus,antibody inhibition by interaction of the dendrim-
er backbone can be ruled out. Among the second-genera-
tion dendrimers with modified core structures,only com-
pound 29Gal (which was the only one that formed
aggregates of considerable size) showed moderate inhibition
in the binding assay. All of the others (11Gal, 19Gal,
24Gal,and 30Gal) were found to be inactive. These experi-
ments provide additional evidence that big,non-covalent ag-
gregates and not individual molecules function as polyvalent
receptor blockers.
Biological evaluation: The potential of our compounds as
polyvalent IgM ligands was assessed by employing in vitro
assays measuring the inhibition of both the anti-aGal IgM
binding to the xenoantigen and the aGal-mediated lysis of
pig erythrocytes (Table 1). Concentrations refer to equiva-
lent concentrations of trisaccharide and not to concentra-
tions of glycodendrimers. In both assays,monomeric aGal
was inactive at 100 mm, which is in agreement with the find-
ing that individual carbohydrate-protein interactions are
often weak.[3] Divalent first generation dendrimer 4Gal,
which forms small aggregates (50 kDa),showed no effect in
either assay. The glycodendrimers 6Gal and 7Gal,which
form large nanoparticles,were highly potent in both assays
(0.025,0.035 mm and 0.010,0.010 mm,respectively). Com-
pound 8Gal also showed high potency in the binding assay
(0.019 mm),but was significantly less potent in the haemoly-
sis assay (0.18 mm). The potency dropped even more for the
larger dendrimers 9Gal and 10Gal,which do not aggregate.
Thus,especially in the more relevant haemolysis assay,po-
The most potent compound, 7Gal,was selected for in
vivo profiling in cynomolgus monkeys. Three animals were
dosed with 1 mgkgꢁ1 (i.v.). Assuming 100% bioavailability,
the initial blood concentration can be estimated to be
ꢀ10 mm. Within 5 minutes of the injection,the anti- aGal
IgMs detected by ELISA were reduced to 20% of the initial
value determined prior to the administration and remained
at low levels for >4 h. Interestingly,the initial IgM levels in
Chem. Eur. J. 2006, 12,99 – 117
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA,Weinheim
www.chemeurj.org
109

G. Thoma et al.
Conclusion
We have reported on novel,self-assembling small glycoden-
drimers fulfilling single molecule entity criteria,which form
non-covalent nanoparticles in water. The particle sizes are
remarkably homogeneous and can be varied over a broad
range by appropriate choice of the dendrimer generation.
The particles can be deposited on different surfaces by evap-
oration of the solvent. Their size distribution is rather homo-
geneous. Self-assembly of our glycodendrimers is a very
robust process,which occurs under in vitro assay conditions
and even more remarkably in vivo. We have shown that
self-assembly of these glycodendrimers is a dynamic equilib-
rium process. Thus,it is conceivable that non-covalent poly-
valent ligands are optimized with respect to size and shape
in the presence of natural polyvalent receptors. The control
of a broad variety of physiologically relevant polyvalent in-
teractions should be possible.
Experimental Section
General: All reactions were carried out in dry solvents under an argon
atmosphere. Reagents were purchased at the highest commercial quality
and were used without further purification. NMR spectra were recorded
on Bruker DRX-600,DRX-500,AMX-500,AMX-400 or Avance
DPX 400 instruments,as well as on Varian Unity 500 and Varian Unity
plus 600 instruments. Residual undeuterated solvents were used as an in-
ternal reference for the calibration. The following abbreviations are used
to denote the multiplicities: s = singlet,d = doublet,t = triplet,q
=
quartet,quint. = quintet,m = multiplet. ESI (electrospray ionization)
mass spectra were obtained on a Finnigan MAT 90 mass spectrometer.
MALDI (matrix-assisted laser desorption ionization) mass spectra were
obtained on a Voyager STR instrument using dihydroxybenzoic acid, a-
cyano-4-hydroxycinnamic acid,25, -dihydroxycinnamic acid,sinapinic acid
or ferulic acid as a matrix. Merck silica gel 60 (particle size 0.040–
0.063 mm) was used for flash column chromatography. Ultrafiltrations
were performed using Amicon 8010 stirred cells (vol.: 10 mL,diam.:
25 mm) and Amicon YM3 disc membranes (molecular weight cut-off:
3000).
Syntheses of dendrimer core structures
Preparation of 1: Ethyldiisopropylamine (30 mL) was added to a mixture
of 2 (10.50 g,41.84 mmol), 3 (5.00 g,20.92 mmol),and HBTU (15.86 g,
41.84 mmol) in dry DMF (100 mL). The solution was stirred for 72 h at
room temperature. Ethyl acetate (250 mL) was added and the resulting
solution was extracted with 1n HCl (2”250 mL),1
n NaOH (2”
250 mL),and brine (2”100 mL). The organic phase was treated with de-
colorizing charcoal (2.5 g),dried with magnesium sulfate,and concentrat-
ed to a volume of 40 mL. The solution was then added dropwise to a stir-
red mixture of diethyl ether (750 mL) and hexane (250 mL). The precipi-
tate formed was filtered off,washed with diethyl ether/hexane,and dried.
Compound 1 (10.35 g,78%) was isolated as a colorless solid. 1H NMR
(600 MHz,CD 3OD,room temperature): d = 1.46 (s,18H; 2” tert-butyl),
3.92 (s,3H; CO 2CH3),4.30 (s,4H; 2”Ar-C H2-NH-),7.41 (d, 3J(H,H) =
Figure 11. Visualization of particles formed by 6Gal–10Gal,a) on a mica
surface by atomic force microscopy; b) on polyformaldehyde by transmis-
sion electron microscopy.
cynomolgus monkeys were found to be 2.3-fold higher than
in pooled human sera. Most importantly,anti- aGal anti-
body-mediated haemolytic activity was completely eliminat-
ed.
8.0 Hz,4H; Ar-H),7.90 (d,
3J(H,H)
= 8.0 Hz,4H; Ar- H),8.14 (d,
3J(H,H) = 2.0 Hz,2H; Ar- H),8.40 ppm (m,1H; Ar- H); HR-MS: calcd
for C34H40N4O8 [M+Na]+ 655.2738; found 655.2739.
Preparation of 4: A mixture of 1 (5.00 g,7.91 mmol),2 n NaOH (30 mL),
and dioxane (30 mL) was stirred at room temperature for 16 h. Ethyl ace-
tate (150 mL) was added and the resulting mixture was washed with 1n
HCl (2”100 mL) and brine (2”100 mL). The organic phase was cooled
to 08C for 16 h. The precipitate that formed was filtered off,washed with
cold ethyl acetate,and dried. Compound 4 (4.42 g,90%) was isolated as
1
a beige solid. H NMR (500 MHz,CD OD,room temperature): d = 1.39
3
110
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA,Weinheim
Chem. Eur. J. 2006, 12,99 – 117

Inhibition of Polyvalent Carbohydrate–Protein Interactions with Glycodendrimers
FULL PAPER
(s,18H; 2” tert-butyl),4.24 (s,4H; 2” Ar-C H2-NH-),7.35 (d, 3J(H,H) =
additional 16 h at room temperature. The solution was added dropwise to
a mixture of acetone and 0.2n HCl (750 mL,1:3). The precipitate formed
was filtered off and washed with water,acetone,and CH 2Cl2. Compound
9 (1760 mg,95%) was isolated as a beige solid. 1H NMR (600 MHz,
[D6]DMSO,80 8C): d = 1.40 (s,288H; 32” tert-butyl),4.22 (m,64H;
32” Ar-CH2-NHBoc),4.58 (m,60H; 30” Ar-C H2-NH-CO-Ar),7.10
(brm,32H; 32” -NHBoc),7.39 (d, 3J(H,H) = 8.0 Hz,64H; Ar-H),7.50
(d, 3J(H,H) = 8.0 Hz,60H; Ar-H),7.92–8.07 (m,184H; Ar-H),8.11 (s,
2H; Ar-H),8.40 (m,30H; Ar-H),8.58 (m,1H; Ar-H),8.79 (m,30H;
30” Ar-CH2-NH-CO-Ar),10.19 ppm (m,62H; 62” Ar-N H-CO-Ar); no
MS could be obtained neither by ES nor by MALDI.
8.0 Hz,4H; Ar-H),7.84 (d,
3J(H,H)
= 8.0 Hz,4H; Ar-H),8.08 (d,
3J(H,H) = 2.0 Hz,2H; Ar-H),8.35 ppm (m,1H; Ar-H); HR-MS: calcd
for C33H38N4O8 [M+Na]+ 641.2582; found 641.2584.
Preparation of 5: A mixture of 1 (3.00 g,4.75 mmol),2 n HCl (12 mL),
and dioxane (25 mL) was stirred at 50–608C for 5 h. Further dioxane
(25 mL) was then added and the solution was cooled to 08C for 4 h. The
precipitate that formed was filtered off,washed with cold dioxane,and
dried. The dihydrochloride of compound 5 (1.98 g,83%) was isolated as
a beige solid. 1H NMR (500 MHz,D 2O,room temperature): d = 3.86 (s,
3H; CO2CH3),4.24 (s,4H; 2” Ar-C H2-NH-),7.57 (d, 3J(H,H) = 8.0 Hz,
4H; Ar-H),7.90 (m,6H; Ar- H),8.16 ppm (m,1H; Ar- H); HR-MS:
calcd. for C24H24N4O4 [M+Na]+ 455.1690; found 455.1691.
Preparation of 10: Ethyldiisopropylamine (2 mL) was added to a mixture
of 5 (12.9 mg,25.6 mmol), 9 (800 mg,51.2 mmol),and HBTU (19.4 mg,
51.2 mmol) in dry DMF (10 mL). The solution was stirred for 7 h at room
temperature. Water (1.7 mL,dropwise) and LiOH monohydrate (32 mg,
762 mmol) were added and the mixture was stirred for an additional 16 h
at room temperature. The solution was added dropwise to a mixture of
acetone and 0.2n HCl (450 mL,1:2). The precipitate formed was filtered
off and washed with water,acetone,and CH 2Cl2. Compound 10 (780 mg,
96%) was isolated as a beige solid. 1H NMR (600 MHz,[D 6]DMSO,
808C): d = 1.40 (s,576H; 64” tert-butyl),4.21 (m,128H; 64” Ar-C H2-
NHBoc),4.57 (m,124H; 62” Ar-C H2-NH-CO-Ar),7.10 (brm,64H; 64”
-NHBoc),7.39 (d, 3J(H,H) = 8.0 Hz,128H; Ar-H),7.50 (m,124H; Ar-
H),7.92–8.10 (m,378H; Ar-H),8.40 (m,62H; Ar-H),the expected
signal at d = 8.58 (m,1H; Ar-H) could not be detected in the noise,8.79
(m,60H; 62” Ar-CH 2-NH-CO-Ar),10.10–10.25 ppm (m,126H; 126”
Ar-NH-CO-Ar); no MS could be obtained neither by ES nor by
MALDI.
Preparation of 6: Ethyldiisopropylamine (15 mL) was added to a mixture
of
4
(3.68 g,5.96 mmol),
5 (1.50 g,2.98 mmol),and HBTU (2.26 g,
5.96 mmol) in dry DMF (50 mL). The solution was stirred for 7 h at
room temperature. DMF (50 mL),water (25 mL,dropwise),and LiOH
monohydrate (3.13 g,74 mmol) were added and the resulting mixture
was stirred for an additional 16 h at room temperature. The solution was
then added dropwise to a mixture of acetone and 1n HCl (500 mL,1:1).
The precipitate formed was filtered off and washed with water,acetone,
and CH2Cl2. Compound 6 (3.26 g,68%) was isolated as beige solid. 1H
NMR (600 MHz,[D 6]DMSO,80 8C): d = 1.41 (s,36H; 4” tert-butyl),
3
4.22 (d, J(H,H) = 5.5 Hz,8H; 4” Ar-C H2-NHBoc),4.58 (d, 3J(H,H) =
7.5 Hz,4H; 2” Ar-C H2-NH-CO-Ar),7.11 (brm,4H; 4” N HBoc),7.40
(d, 3J(H,H) = 8.0 Hz,8H; Ar-H),7.51 (d, 3J(H,H) = 8.0 Hz,4H; Ar-
H),7.97 (m,16H; Ar-H),8.12 (s,2H; Ar-H),8.41 (m,2H; Ar-H),8.57
(m,1H; Ar-H),8.79 (s,2H; 2” Ar-CH 2-NH-CO-Ar),10.17 (s,4H; 4”
Ar-NH-CO-Ar),10.22 ppm (s,2H; 2” Ar-N H-CO-Ar); HR-MS: calcd
for C89H94N12O18 [M+NH4]+ 1636.7147; found 1636.7147.
Preparation of 11: Ethyldiisopropylamine (15 mmol) was added to a mix-
ture of 17 (1.40 g,2.33 mmol), 18 (0.65 g,1.01 mmol),and HBTU (0.96 g,
2.5 mmol) in dry DMF (30 mL). The solution was stirred overnight at
room temperature. Water (5 mL,dropwise) and LiOH monohydrate
(3.17 g,75 mmol) were added and the mixture was stirred for an addi-
tional 16 h at room temperature. The solution was then added dropwise
to a mixture of acetone and water (500 mL,1:3). The precipitate formed
was filtered off and washed with 1n HCl,water,acetone,and CH 2Cl2.
Compound 11 (800 mg,69%) was isolated as a beige powder. 1H NMR
Preparation of 7: Ethyldiisopropylamine (10 mL) was added to a mixture
of
5
(0.47 g,0.93 mmol),
6 (3.00 g,1.85 mmol),and HBTU (0.70 g,
0.93 mmol) in dry DMF (35 mL). The solution was stirred for 6 h at
room temperature. DMF (15 mL),water (12 mL,dropwise),and LiOH
monohydrate (1.95 g,46.5 mmol) were added and the resulting mixture
was stirred for an additional 16 h at room temperature. The solution was
added dropwise to a mixture of acetone and 0.2n HCl (750 mL,1:3). The
precipitate formed was filtered off and washed with water,acetone,and
CH2Cl2. Compound 7 (3.10 g,92%) was isolated as a beige solid. 1H
NMR (600 MHz,[D 6]DMSO,80 8C): d = 1.41 (s,72H; 8” tert-butyl),
4.22 (m,16H; 8” Ar-C H2-NHBoc),4.58 (m,12H; 6” Ar-C H2-NH-CO-
Ar),7.29 (brm,8H; 8” -N HBoc),7.39 (d, 3J(H,H) = 8.0 Hz,16H; Ar-
H),7.51 (m,12H; Ar-H),7.97 (m,40H; Ar-H),8.11 (s,2H; Ar-H),8.41
(400 MHz,[D ]DMSO,120 8C): d = 1.41 (s,36H; 4” tert-butyl),2.34 (m,
6
6H; 3” -CH2-CH(CH2-NH-)2),2.41 (m,3H; 3” -CH 2-CH(CH2-NH-)2),
3.40 (m,12H; 3” -CH -CH(CH2-NH-)2),4.20 (d, 3J(H,H) = 6.0 Hz,8H;
2
2” Ar-CH2-NHBoc),4.36 (d, 3J(H,H) = 6.0 Hz,4H; 2” Ar-C H2-NH-),
6.76 (brs,4H; 4” Ar-CH 2-NHBoc),7.34 (m,12H; Ar-H),7.79 (m,12H;
Ar-H),8.05 ppm (m,6H; 3” -CH 2-CH(CH2-NH-)2); MALDI-MS: m/z:
1582 [M+Na]+; HR-MS: calcd. for C83H106N12O18 [M+2Na]2+ 802.3766;
found 802.3764.
(m,6H; Ar-H),8.58 (m,1H; Ar-H),8.79 (s,6H; 6” Ar-CH
2-NH-CO-
Ar),10.19 ppm (s,14H; 2” Ar-N H-CO-Ar); no MS could be obtained
neither by ES nor by MALDI.
Preparation of 13: Compound 12 (3.00 g,16.75 mmol) was mixed with
2
(8.42 g,33.5 mmol) in dry DMF (50 mL). HBTU (19.06 g,50.2 mmol)
Preparation of 8: Ethyldiisopropylamine (10 mL) was added to a mixture
of 5 (0.209 g,0.414 mmol), 7 (3.00 g,0.828 mmol),and HBTU (0.314 g,
0.828 mmol) in dry DMF (35 mL). The solution was stirred for 6 h at
room temperature. DMF (15 mL),water (10 mL,dropwise),and LiOH
monohydrate (1.95 g,46.5 mmol) were added and the mixture was stirred
for an additional 16 h at room temperature. The solution was then added
dropwise to a mixture of acetone and 0.2n HCl (750 mL,1:3). The pre-
cipitate formed was filtered off and washed with water,acetone,and
CH2Cl2. Compound 8 (2.85 g,90%) was isolated as a beige solid. 1H
NMR (600 MHz,[D 6]DMSO,80 8C): d = 1.40 (s,144H; 16” tert-butyl),
and ethyldiisopropylamine (29 mL,167 mmol) were added and the result-
ing solution was stirred overnight at room temperature. The product was
precipitated by slow addition of the crude mixture to well-stirred aque-
ous 1n HCl (500 mL). The solid was collected by filtration and washed
with water,1 n NaOH,1 n HCl,water,and diethyl ether. Compound 13
was isolated as a beige powder (7.30 g,79%). 1H NMR (400 MHz,
[D6]DMSO,room temperature): d = 1.39 (s,18H; 2” tert-butyl),4.18
(m,8H; 2” Ar-C H2-NH-,-NH-C H2-CO-CH2-NH-),7.32 (d, 3J(H,H) =
8.0 Hz,4H; Ar-H),7.46 (t, 3J(H,H) = 6.0 Hz,2H; 2” Ar-CH 2-NH-),
7.82 (d, 3J(H,H) = 8.0 Hz,4H; Ar-H),8.79 ppm (t, 3J(H,H) = 5.5 Hz,
2H; -NH-CH2-CO-CH2-NH-); HR-MS: calcd for C29H38N4O7 [M+Na]+
593.2372; found 592.2375.
3
4.22 (d, J(H,H) = 5.5 Hz,32H; 16” Ar-C H2-NHBoc),4.58 (d, 3J(H,H)
=
7.5 Hz,28H; 14” Ar-C H2-NH-CO-Ar),7.11 (brm,16H; 16”
-NHBoc),7.39 (d, 3J(H,H) = 8.0 Hz,32H; Ar-H),7.50 (m, 3J(H,H) =
8.0 Hz,28H; Ar-H),7.97 (m,88H; Ar-H),8.11 (m,2H; Ar-H),8.40 (m,
Preparation of 15: Potassium bis(trimethylsilyl)amide (53 mL of a 0.5m
solution in toluene) was added at 08C to a solution of phosphonoacetate
14 (6.9 mL,37.8 mmol) in THF (75 mL) and the mixture was stirred
under argon at room temperature for 15 min. At 08C,a solution of 13
(7.0 g,12.6 mmol) in THF (150 mL) was added and stirring was contin-
ued for 2.5 h. Saturated aqueous NH4Cl (150 mL) was added and the re-
sulting mixture was allowed to warm to room temperature. The mixture
was extracted with ethyl acetate and the organic phase was washed with
1n HCl (3”150 mL),1 n NaOH (3”150 mL),and brine (3”150 mL).
14H; Ar-H),8.58 (m,1H; Ar-H),8.79 (m,14H; 14” Ar-CH
2-NH-CO-
Ar),10.18 ppm (m,30H; 30” Ar-N H-CO-Ar); no MS could be obtained
neither by ES nor by MALDI.
Preparation of 9: Ethyldiisopropylamine (2 mL) was added to a mixture
of
5
(59.5 mg,0.118 mmol),
8 (1800 mg,0.236 mmol),and HBTU
(89.4 mg,0.236 mmol) in dry DMF (20 mL). The solution was stirred for
6 h at room temperature. Water (3.2 mL,dropwise) and LiOH monohy-
drate (149 mg,3.54 mmol) were added and the mixture was stirred for an
Chem. Eur. J. 2006, 12,99 – 117
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA,Weinheim
www.chemeurj.org
111

G. Thoma et al.
The solvent was removed and the residue was subjected to flash chroma-
tography on SiO2 (ethyl acetate/CH2Cl2,2:1 !4:1). Compound 15 was
isolated as a beige solid (3.1 g,40%). 1H NMR (400 MHz,[D 6]DMSO,
808C): d = 1.40 (s,18H; 2” tert-butyl),3.67 (s,3H; -COOCH 3),4.09 (d,
3J(H,H) = 5.5 Hz,2H; -CH =C(CH2-NH-)2),4.57 (d, 3J(H,H) = 5.5 Hz,
2H; -CH=C(CH2-NH-)2),4.19 (d, 3J(H,H) = 6.0 Hz,4H; 2” Ar-C H2-
NH-),5.83 (s,1H; CH 3OOC-CH=C[CH2-NH-]2),7.04 (brs,2H; 2” Ar-
CH2-NH-),7.33 (d, 3J(H,H) = 8.0 Hz,4H; Ar-H),7.81 (d, 3J(H,H) =
8.0 Hz,4H; Ar-H),8.42 (t, 3J(H,H) = 5.5 Hz,1H; -CH =C(CH2-NH-)2),
8.46 ppm (t, 3J(H,H) = 5.5 Hz,1H; CH 3OOC-CH=C(CH2-NH-)2); HR-
MS: calcd for C32H42N4O8 [M+Na]+ 633.2895; found 633.2896.
(brs,4H; 4” -CH 2-NHBoc),7.63 (m,4H; Ar-H),7.87 (m,2H; Ar-H),
7.97 (brs,2H; 2” -CH 2-NHCOAr),8.05 (m,2H; Ar-H),8.17 (m,1H;
Ar-H),9.61 (s,4H; 4” Ar-N-CO-),9.69 (s,2H; 2” Ar-NH-CO-),
12.43 ppm (brs,1H; -COOH); MALDI-MS: m/z: 1678 [M+Na]+; HR-
MS: calcd for C89H130N12O18 [M+2Na]2+ 850.4705; found 850.4708.
Preparation of 21:
A mixture of 3 (2.5 g,10.45 mmol), 20 (5.38 g,
20.90 mmol),HBTU (7.93 g,20.90 mmol),and ethyldiisopropylamine
(15.5 mL,88.87 mmol) in DMF (50 mL) was stirred for 16 h at 20 8C. The
product was precipitated by slow addition of the crude mixture to well-
stirred 1n HCl (300 mL). The solid was collected by filtration and
washed with water,1 n NaOH,1 n HCl,and further water. It was then
dissolved in a small amount of ethyl acetate and precipitated by addition
to a mixture of cyclohexane (100 mL) and diethyl ether (300 mL). The
precipitate was filtered off,washed,and dried. Compound 21 was isolated
1
Isomer of 15 with the double bond adjacent to a nitrogen atom: H NMR
(400 MHz,[D 6]DMSO,80 8C): d = 1.40 (s,18H; 2” tert-butyl),3.40 (s,
2H; CH3OOC-CH2-C(=CH-NH-)(-CH2-NH-)),3.60 (s,3H; -COOCH 3),
3
as a beige solid (3.64 g,55%). 1H NMR (400 MHz,[D ]DMSO,80 8C): d
3.98 (d, J(H,H) = 5.5 Hz,2H; CH 3OOC-CH2-C(=CH-NH-)(-CH2-NH-)),
6
4.19 (m,4H; 2” Ar-C H2-NH-),7.00 (d, 3J(H,H)
=
10.0 Hz,1H;
CH3OOC-CH2-C(=CH-NH-)(-CH2-NH-)),7.07 (brs,2H; 2” Ar-CH
NH-),7.34 (m,4H; Ar-H),7.81 (m,4H; Ar-H),8.31 (t,
3J(H,H)
5.5 Hz,1H; CH 3OOC-CH2-C(=CH-NH-)(-CH2-NH-)),9.39 ppm (d, 3J-
= 0.94 (m,4H),1.37–1.47 (m,6H),1.41 (s,18H; 2”
tert-butyl),1.78 (m,
-
=
4H),1.87 (m,4H),2.28 (m,2H),2.83 (m,4H),3.85 (s,3H; -COOCH
3),
2
6.41 (brs,2H; 2” -CH -NH-Boc),7.90 (m,2H; Ar-H),8.19 (m,1H; Ar-
2
H),9.71 ppm (s,2H; 2” Ar-NH-CO-); HR-MS: calcd. for C 34H52N4O8
[M+Na]+ 667.3677; found 667.3676.
(H,H)
= 10.0 Hz,1H; CH 3OOC-CH2-C(=CH-NH-)(-CH2-NH-)); ES-
MS: m/z: 633 [M+Na]+.
Preparation of 22: A mixture of 21 (1.60 g,2.48 mmol),2 n NaOH
(10 mL),and dioxane (10 mL) was stirred at 20 8C for 16 h. The mixture
was added dropwise to well-stirred 1n HCl (200 mL),and the resulting
precipitate was filtered off,washed with diethyl ether,and dried. Com-
Preparation of 16: A mixture of 15 (3.00 g,4.91 mmol),ammonium for-
mate (15.5 g,245.7 mmol),and Pd (10% on activated charcoal,160 mg)
in methanol (100 mL) was stirred at 208C for 16 h. Additional ammoni-
um formate (15.5 g,245.7 mmol) and Pd (10% on activated charcoal,
160 mg) were added and stirring was continued for 24 h. The catalyst was
then filtered off,the solvent was removed,and the resulting solid was
washed with water. Compound 16 was isolated as a beige powder (2.6 g,
86%). 1H NMR (400 MHz,[D 6]DMSO,80 8C): d = 1.40 (s,18H; 2”
tert-butyl),2.39 (m,3H; -C H2-CH(CH2-NH-)2),3.36 (m,4H; -CH 2-CH-
(CH2-NH-)2),3.58 (s,3H; -COOCH 3),4.19 (d, 3J(H,H) = 6.0 Hz,4H;
2” Ar-CH2-NHBoc),7.05 (brs,2H; 2” Ar-CH 2-NHBoc),7.33 (d,
3J(H,H) = 8.0 Hz,4H; Ar-H),7.79 (d, 3J(H,H) = 8.0 Hz,4H; Ar-H),
8.22 ppm (m,2H; -CO-CH 2-CH(CH2-NH-)2); HR-MS: calcd for
C32H44N4O8 [M+Na]+ 635.3051; found 635.3052.
pound 22 was isolated as
a
beige solid (1.22 g,78%). 1H NMR
(400 MHz,[D ]DMSO,80 8C): d = 0.94 (m,4H),1.39–1.47 (m,6H),1.41
6
(s,18H; 2” tert-butyl),1.78 (m,4H),1.87 (m,4H),2.29 (m,2H),2.83
(m,4H),6.41 (brs,2H; 2” -CH 2-NH-Boc),7.86 (m,2H; Ar-H),8.16 (m,
1H; Ar-H),9.68 (s,2H; 2” Ar-NH-CO-),12.45 ppm (brs,1H; -COOH);
HR-MS: calcd for C33H50N4O8 [M+Na]+ 653.3521; found 653.3522.
Preparation of 23: A mixture of 21 (0.85 g,1.31 mmol),TFA (20 mL),
water (1.6 mL),and triethylsilane (0.8 mL) was stirred at 20 8C for 3 h.
The mixture was added dropwise to diethyl ether (200 mL),and the pre-
cipitate formed was filtered off and dried. The TFA salt of compound 23
was isolated as
[D6]DMSO/D2O,80 8C): d = 1.02 (m,4H),1.44 (m,4H),1.58 (m,2H),
a
beige solid (0.85 g,91%). 1H NMR (400 MHz,
Preparation of 17: A mixture of 16 (2.00 g,3.26 mmol),2 n NaOH
(15 mL),and dioxane (15 mL) was stirred at 20 8C for 16 h. The mixture
was added dropwise to well-stirred 1n HCl (200 mL),and the resulting
precipitate was filtered off,washed with diethyl ether,and dried. Com-
1.86 (m,8H),2.31 (m,2H),2.70 (m,4H),3.83 (s,3H; -COOCH
(d, 3J(H,H) = 2.0 Hz,2H; Ar-H),8.15 ppm (t, 3J(H,H) = 2.0 Hz,1H;
Ar-H); HR-MS: calcd for C24H36N4O4 [M+Na]+ 467.2629; found
467.2630.
3),7.85
pound 17 was isolated as
a
beige solid (1.62 g,83%). 1H NMR
(400 MHz,[D 6]DMSO,80 8C): d = 1.41 (s,18H; 2” tert-butyl),2.32 (m,
3H; -CH2-CH[CH2-NH-]2),3.37 (m,4H; -CH 2-CH(CH2-NH-)2),4.19 (d,
Preparation of 24: Ethyldiisopropylamine (1.5 mL) was added to a mix-
ture of 27 (1.16 g,2.00 mmol), 28 (0.62 g,1.00 mmol),and HBTU (0.76 g,
2.00 mmol) in dry DMF (5 mL). The solution was stirred for 16 h at
208C. Ethyl acetate (100 mL) was added and the mixture was extracted
with 1n HCl (2”100 mL),1 n NaOH (2”100 mL),and brine (100 mL).
The solution was dried with MgSO4,the solvent was removed,the resi-
due was redissolved in ethanol (5 mL),and this solution was added drop-
wise to diethyl ether (200 mL). The precipitate formed was filtered off
3J(H,H) = 6.0 Hz,4H; 2” Ar-C H2-NHBoc),7.05 (brs,2H; 2” Ar-CH
-
2
NHBoc),7.33 (d, 3J(H,H) = 8.0 Hz,4H; Ar-H),7.80 (d,
3J(H,H)
=
8.0 Hz,4H; Ar-H),8.23 ppm (m,2H,-CH
2-CH(CH2-NH-)2); HR-MS:
calcd for C31H42N4O8 [M+Na]+ 621.2895; found 621.2895.
Preparation of 18: A mixture of 16 (0.75 g,1.22 mmol),TFA (20 mL),
water (1.5 mL),and triethylsilane (0.75 mL) was stirred at 20 8C for 3 h.
The mixture was added dropwise to diethyl ether (200 mL),and the pre-
cipitate formed was filtered off and dried. The TFA salt of compound 18
and washed with diethyl ether. The methyl ester of 24 (1.30 g,86%) was
1
isolated as a beige solid. H NMR (400 MHz,CD OD,20 8C,selected sig-
3
was isolated as
a
beige solid (0.76 g,97%). 1H NMR (400 MHz,
nals): d = 1.41 (s,36H; 4” tert-butyl),2.38 (m,12H; 6” -CH 2-CH2-CO-),
3.03 (t, 3J(H,H) = 7.0 Hz,8H; 4” BocHN-C H2-CH2-),3.38 (t, 3J(H,H)
= 7.0 Hz,4H; 2” Ar-(CO)HN-C H2-CH2-),3.87 (s,3H; COOCH 3),7.68
(d, 3J(H,H) = 2.0 Hz,4H; Ar-H),7.94 (d, 3J(H,H) = 2.0 Hz,2H; Ar-
[D6]DMSO,80 8C): d = 2.42 (m,3H; -C H2-CH(CH2-NH-)2),3.38 (m,
4H; -CH2-CH(CH2-NH-)2),3.57 (s,3H; -COOCH 3),4.11 (m,4H; 2”
Ar-CH2-NH2),7.54 (d, 3J(H,H) = 8.0 Hz,4H; Ar-H),7.89 (d, 3J(H,H) =
8.0 Hz,4H; Ar-H),8.19 (brs,4H; 2” Ar-CH 2-NH2),8.34 ppm (m,2H;
-CO-CH2-CH(CH2-NH-)2); HR-MS: calcd for C22H28N4O4 [M+Na]+
435.2003; found 425.2002.
H),7.98 (t, 3J(H,H)
=
7.0 Hz,2H; Ar-H),8.07 ppm (t,
3J(H,H)
=
7.0 Hz,1H; Ar-H); ES-MS: m/z: 1535 [M+Na]+; HR-MS: calcd. for
C78H120N12O18 [M+2Na]2+ 779.4314; found 779.4310. Water (5 mL) was
added dropwise to
a solution of the methyl ester of 24 (1.00 g,
Preparation of 19: Ethyldiisopropylamine (3.9 mL) was added to a mix-
ture of 22 (0.75 g,1.12 mmol), 23 (1.41 g,2.23 mmol),and HBTU (0.85 g,
2.23 mmol) in dry DMF (22 mL). The solution was stirred for 16 h at
208C. Water (5 mL,dropwise) and LiOH monohydrate (2.33 g,55 mmol)
were added and the mixture was stirred for an additional 16 h at 208C.
The solution was then added dropwise to a mixture of acetone and water
(500 mL,1:3). The precipitate formed was filtered off and washed with
1n HCl,water,acetone,and CH 2Cl2. Compound 19 was isolated as a
beige solid (1.48 g,80%). 1H NMR (400 MHz,[D 6]DMSO,80 8C): d =
0.66 mmol) in DMF (12.5 mL). LiOH (0.28 g,6.60 mmol) was then added
and the mixture was stirred at room temperature for 16 h. Ethyl acetate
(50 mL) and ethanol (5 mL) were added and the resulting mixture was
washed with 1n HCl (2”50 mL),1 n NaOH (2”50 mL),and brine
(50 mL). The solvent was removed and the residue was dissolved in a 5:1
mixture of ethyl acetate and ethanol (10 mL). This solution was added
dropwise to diethyl ether (150 mL). The precipitate formed was filtered
off,washed with diethyl ether and hexanes,and dried. Compound 24 was
0.97 (m,12H),1.29–1.50 (m,16H),1.41 (s,36H; 4”
tert-butyl),1.57 (m,
isolated as
CD3OD,20 8C,selected signals): d = 1.41 (s,36H; 4” tert-butyl),2.38
a
colorless powder (0.94 g,95%). 1H NMR (400 MHz,
2H),1.70–1.95 (m,24H),2.29 (m,6H),2.83 (m,8H),3.15 (m,4H),6.40
112
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA,Weinheim
Chem. Eur. J. 2006, 12,99 – 117

Inhibition of Polyvalent Carbohydrate–Protein Interactions with Glycodendrimers
FULL PAPER
(m,12H; 6” -CH 2-CH2-CO-),3.03 (t, 3J(H,H)
BocHN-CH2-CH2-),3.38 (t, 3J(H,H)
7.0 Hz,4H; 2” Ar-(CO)HN-
=
7.0 Hz,8H; 4”
Preparation of 30: Ethyldiisopropylamine (1.6 mL) was added to a mix-
ture of 40 (0.32 g,0.46 mmol), 39 (0.60 g,0.93 mmol),and HBTU (0.37 g,
0.98 mmol) in dry DMF (12 mL). The solution was stirred for 16 h at
208C. Water (5 mL,dropwise) and LiOH monohydrate (0.98 g,23 mmol)
were added and the mixture was stirred for an additional 16 h at 208C.
The solution was then added dropwise to a mixture of acetone and 0.1n
HCl (500 mL,1:3). The precipitate formed was filtered off and washed
with water and diethyl ether. The solvent was removed and the residue
was purified by flash chromatography (SiO2,CH 2Cl2/MeOH,9:1). Com-
=
CH2-CH2-),7.68 (d, 3J(H,H) = 2.0 Hz,4H; Ar-H),7.95 (d, 3J(H,H) =
2.0 Hz,2H; Ar-H),8.00 (t, 3J(H,H) = 7.0 Hz,2H; Ar-H),8.10 ppm (t,
3J(H,H) = 7.0 Hz,1H; Ar-H); ES-MS: m/z: 1498 [MꢁH]ꢁ; HR-MS:
calcd. for C78H120N12O18 [M+2Na]2+ 772.4236; found 772.4235.
Preparation of 26: Ethyldiisopropylamine (8 mL) was added to a mixture
of 3 (2.33 g,9.75 mmol), 25 (4.50 g,19.50 mmol),and HBTU (7.39 g,
19.50 mmol) in dry DMF (32 mL). The solution was stirred for 72 h at
208C. Ethyl acetate (150 mL) was then added and the solution was ex-
tracted with 1n HCl (2”150 mL),1 n NaOH (2”150 mL),and brine (2”
100 mL). The organic phase was treated with decolorizing charcoal
(2.5 g),dried with magnesium sulfate,and the solvent was removed. The
residue was subjected to flash chromatography on SiO2 (ethyl acetate/
hexanes 2:1!4:1). Compound 26 was isolated as a colorless foam (4.70 g,
pound 30 was isolated as
a
beige solid (0.45 g,57%). 1H NMR
(400 MHz,[D 6]DMSO,80 8C): d = 1.36 (s,36H; 4” tert-butyl),3.20 (s,
12H; 4” Ar-NCH3-CO-Ar),3.21 (s,6H; 2” Ar-NC H3-CO-Ar),4.09 (d,
3J(H,H) = 6.0 Hz,8H; 4” Ar-C H2-NHBoc),4.35 (d, 3J(H,H) = 6.0 Hz,
4H; 2” Ar-CH2-NHCOAr),6.96 (brm,4H; 4” Ar-CH 2-NHBoc),7.12–
7.22 (m,26H; Ar-H),7.26 (m,1H; Ar-H),7.45 (m,6H; Ar-H),8.64 ppm
(t, 3J(H,H) = 6.0 Hz,2H; 2” Ar-CH 2-NHCOAr); MALDI-MS: m/z:
1726 [M+Na]+; HR-MS: calcd for C95H106N12O18 [M+2Na]2+ 874.3766;
found 874.3764.
81%). 1H NMR (400 MHz,CD 3OD,20 8C): d
=
1.31 (m,4H; 2”
BocHN-CH2-CH2-CH2-CH2-CH2-CO-),1.31 (s,18H; 2” t-butyl),1.40
(quint., 3J(H,H) = 7.5 Hz,4H; BocHN-CH 2-CH2-CH2-CH2-CH2-CO-),
1.61 (quint., 3J(H,H)
7.5 Hz,4H; BocHN-CH 2-CH2-CH2-CH2-CH2-
=
Preparation of 31:
A
mixture of
3
(1.0 g,4.18 mmol),
32 (2.2 g,
CO-),2.29 (t, 3J(H,H) = 7.5 Hz,4H; BocHN-CH 2-CH2-CH2-CH2-CH2-
C(O)-),2.93 (t, 3J(H,H) = 7.5 Hz,4H; BocHN-C H2-CH2-CH2-CH2-CH2-
C(O)-),3.80 (s,3H; CO 2CH3),7.88 (m,2H; Ar-H),8.03 ppm (m,1H;
Ar-H); HR-MS: calcd for C30H48N4O8 [M+Na]+ 615.3364; found
615.3365.
8.36 mmol),HBTU (3.17 g,8.36 mmol),and ethyldiisopropylamine
(6.2 mL,35.5 mmol) in DMF (25 mL) was stirred for 16 h at 20 8C. The
product was precipitated by dropwise addition of the crude mixture to
well-stirred 1n HCl (200 mL). The solid was collected by filtration and
washed with water. It was dissolved in ethyl acetate (75 mL) and washed
successively with 1n HCl (3”20 mL),1 n NaOH (3”20 mL),and brine
(3”20 mL),and dried with Na 2SO4. The solvent was removed and the
residue was purified by flash chromatography (SiO2,cyclohexane/EtOAc,
1:1) to give 31 as a beige solid (850 mg,31%). 1H NMR (400 MHz,
[D6]DMSO,80 8C): d = 1.44 (s,18H; 2” tert-butyl),2.83 (s,6H; 2”
-CH2-NCH3-CO-),3.90 (s,3H; -COOCH 3),4.47 (s,4H; 2” -C H2-NCH3-
CO-),7.37 (d, 3J(H,H) = 8.0 Hz,4H; Ar-H),7.99 (d, 3J(H,H) = 8.0 Hz,
4H; Ar-H),8.15 (d, 3J(H,H) = 2.0 Hz,2H; Ar-H),8.61 (t, 3J(H,H) =
2.0 Hz,1H; Ar-H),10.25 ppm (s,2H; 2” Ar-CO-NH-); HR-MS: calcd
for C36H44N4O8 [M+Na]+ 683.3051; found 683.3051.
Preparation of 27: Compound 26 (2.50 g,4.22 mmol) was dissolved in a
mixture of dioxane (20 mL) and water (25 mL). LiOH (1.76 g,
42.0 mmol) was added and the mixture was stirred for 16 h at 208C.
Ethyl acetate (50 mL) was then added and the mixture was extracted
with 1n HCl (2”100 mL),water (2”100 mL),and brine (100 mL). The
organic phase was dried with MgSO4 and the solvent was removed. Com-
pound 27 was isolated as a colorless foam (2.43 g,quant.). 1H NMR
(400 MHz,CD OD,20 8C): d = 1.31 (m,4H; 2” BocHN-CH -CH2-CH2-
3
2
CH2-CH2-CO-),1.31 (s,18H; 2”
t-butyl),1.40 (quint., 3J(H,H)
=
7.5 Hz,4H; BocHN-CH 2-CH2-CH2-CH2-CH2-CO-),1.62 (quint., 3J(H,H)
= 7.5 Hz,4H; BocHN-CH -CH2-CH2-CH2-CH2-CO-),2.29 (t, 3J(H,H) =
Preparation of 32: NaH (60%,3.18 g,79.59 mmol) was added portion-
wise to 2 (5 g,19.90 mmol) in THF (100 mL) and the resulting suspension
was stirred for 30 min at 208C. Methyl iodide (9.9 mL,159.2 mmol) was
added and stirring was continued for 40 h. The mixture was then
quenched with iced water and the pH was adjusted to 4 using 2n HCl.
The mixture was extracted with ethyl acetate (3”100 mL),and the com-
bined organic phases were washed with brine (3”100 mL) and dried with
Na2SO4. Filtration through SiO2 (EtOAc) gave 32 (4.6 g,87%) as a beige
2
7.5 Hz,4H; BocHN-CH 2-CH2-CH2-CH2-CH2-CO-),2.94 (t, 3J(H,H)
=
7.5 Hz,4H; BocHN-C H2-CH2-CH2-CH2-CH2-CO-),7.87 (m,2H; Ar-H),
8.06 ppm (m,1H; Ar-H); HR-MS: calcd for
601.3208; found 601.3208.
C
29H46N4O8 [M+Na]+
Preparation of 28: Compound 26 (1.00 g,1.69 mmol) was dissolved in a
mixture of TFA (25 mL),water (0.7 mL),and triethylsilane (0.7 mL). The
solution was stirred at 208C for 2 h. The solvents were removed and the
residue was co-evaporated with diethyl ether (3”50 mL). Compound 28
(1.05 g,quant.) was obtained as a colorless foam. 1H NMR (400 MHz,
CD3OD,20 8C): d = 1.37 (m,4H; 2” H 2N-CH2-CH2-CH2-CH2-CH2-CO-),
solid. 1H NMR (400 MHz,[D 6]DMSO,80 8C): d
= 1.42 (s,9H; tert-
butyl),2.81 (s,3H; -CH 2-NCH3-CO-),4.45 (s,2H; -C H2-NCH3-CO-),
7.32 (d, 3J(H,H) = 8.0 Hz,2H; Ar-H),7.92 (d, 3J(H,H) = 8.0 Hz,2H;
Ar-H),12.47 ppm (brs,1H; -COOH); HR-MS: calcd for C 8H10N2O2
[M+Na]+ 189.0635; found 189.0634.
1.63 (m,8H; H 2N-CH2-CH2-CH2-CH2-CH2-CO-),2.33 (t, 3J(H,H)
7.5 Hz,4H; H 2N-CH2-CH2-CH2-CH2-CH2-C(O)-),2.84 (t, 3J(H,H)
=
=
7.5 Hz,4H; H 2N-CH2-CH2-CH2-CH2-CH2-C(O)-),3.80 (s,3H; CO 2CH3),
7.85 (m,2H; Ar-H),8.06 ppm (m,1H; Ar-H); HR-MS: calcd for
C20H32N4O4 [M+Na]+ 415.2316; found 415.2316.
Preparation of 33: Compound 31 (0.53 g,0.80 mmol) was dissolved in a
mixture of TFA (15 mL),water (1.2 mL),and triethylsilane (0.6 mL). The
solution was stirred at 208C for 2 h. The solvents were removed and the
residue was co-evaporated with diethyl ether (3”50 mL). Compound 33
(1.05 g,89%) was obtained as a colorless solid. 1H NMR (400 MHz,
[D6]DMSO/D2O,20 8C): d = 2.60 (s,6H; 2” Ar-CH 2-NHCH3),3.87 (s,
3H; -COOCH3),4.20 (s,4H; 2” Ar-C H2-NHCH3),7.61 (d, 3J(H,H) =
8.0 Hz,4H; Ar-H),8.03 (d, 3J(H,H) = 8.0 Hz,4H; Ar-H),8.14 (m,2H;
Preparation of 29: Ethyldiisopropylamine (2.3 mL) was added to a mix-
ture of 33 (0.45 g,0.65 mmol), 4 (0.81 g,1.30 mmol),and HBTU (0.52 g,
1.37 mmol) in dry DMF (13 mL). The solution was stirred for 16 h at
208C. Water (5 mL,dropwise) and LiOH monohydrate (1.40 g,33 mmol)
were added and the mixture was stirred for an additional 16 h at 208C.
The solution was then added dropwise to a mixture of acetone and water
(500 mL,1:3). The precipitate formed was filtered off and washed with
1n HCl,water,acetone,and CH 2Cl2. Compound 29 was isolated as a
beige solid (1.01 g,94%). 1H NMR (400 MHz,[D 6]DMSO,80 8C): d =
1.41 (s,36H; 4” tert-butyl),2.98 (s,6H; 2” -CH 2-NCH3-CO-),4.22 (d,
3J(H,H) = 6.0 Hz,8H; 4” Ar-C H2-NHBoc),4.76 (s,4H; 2” Ar-C H2-
Ar-H),8.66 ppm (m,1H; Ar-H); HR-MS: calcd for
C
26H28N4O4
[M+Na]+ 483.2003; found 483.2004.
Preparation of 34: At 08C,DMF (2 drops) and oxalyl chloride (2.56 mL,
29.85 mmol) were successively added to suspension of (5 g,
a
2
19.90 mmol) in dry CH2Cl2 (150 mL) and the mixture was stirred at 08C
for 4 h. The solvent was then removed,the residue was redissolved in
CH2Cl2 (75 mL),and this solution was added at 0 8C to a solution of 35
(2.39 g,8.95 mmol) and ethyldiisopropylamine (8.7 mL,49.7 mmol) in
CH2Cl2 (75 mL). The ice bath was removed and the mixture was stirred
for 4 h. Water (100 mL) was then added,and the organic phase was sepa-
rated,washed with 1 n HCl (3”100 mL),1 n NaOH (3”100 mL),and
NCH3-CO-),7.09 (brs,4H; 4” Ar-CH 2-NHBoc),7.39 (d, 3J(H,H)
8.0 Hz,8H; Ar-H),7.47 (d, 3J(H,H) = 8.0 Hz,4H; Ar-H),7.63 (m,4H;
=
Ar-H),7.94 (d, 3J(H,H) = 8.0 Hz,8H; Ar-H),8.03 (d, 3J(H,H) = 8.0 Hz,
4H; Ar-H),8.12 (m,2H; Ar-H),8.38 (m,2H; Ar-H),8.57 (m,1H; Ar-
H),10.15 (s,4H; 4” Ar-CO-N H-),10.24 ppm (s,2H; 4” Ar-CO-NH-);
MALDI-MS: m/z: 1670 [M+Na]+; HR-MS: calcd for C91H98N12O18
[M+2Na]2+ 843.3453; found 846.3454.
brine (3”100 mL),and dried with Na SO4. The solvent was removed and
2
the residue was purified by flash chromatography (SiO2,EtOAc/cyclo-
Chem. Eur. J. 2006, 12,99 – 117
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA,Weinheim
www.chemeurj.org
113

G. Thoma et al.
hexane,6:4). Compound 34 was isolated as a colorless solid (3.1 g,52%).
cipitate formed was filtered off. The TFA salts of the products were iso-
lated as beige solids.
1H NMR (400 MHz,[D ]DMSO,80 8C): d = 1.36 (s,18H; 2” tert-butyl),
6
4N: Yield 78%; 1H NMR (600 MHz,D 2O,20 8C): d = 4.17 (s,4H; 2”
3.23 (s,6H; 2” Ar-NC H3-CO-Ar),3.77 (s,3H; -COOCH 3),4.09 (d,
3J(H,H) = 6.5 Hz,4H; 2” Ar-C H2-NHBoc),6.96 (brm,2H; 2” Ar-
CH2-NHBoc),7.14 (d, 3J(H,H) = 8.0 Hz,4H; Ar-H),7.19 (d, 3J(H,H) =
8.0 Hz,4H; Ar-H),7.34 (t, 3J(H,H) = 2.0 Hz,1H; Ar-H),7.44 ppm (d,
3J(H,H) = 2.0 Hz,2H; Ar-H); HR-MS: calcd for C 36H44N4O8 [M+Na]+
683.3051; found 683.3050.
Ar-CH2-NH2),7.48 (d, 3J(H,H) = 8.5 Hz,4H; Ar-H),7.76 (d, 3J(H,H) =
2.0 Hz,2H; Ar-H),7.79 (d,
3J(H,H)
=
8.5 Hz,4H; Ar-H),8.12 ppm
(brt, 3J(H,H)
[2M+H]+.
=
2.0 Hz,1H; Ar-H); ES-MS: m/z: 419 [M+H]+,837
6N: Yield 93%; 1H NMR (600 MHz,[D 6]DMSO/D2O,80 8C): d = 4.12
(s, 8H; 4” Ar-CH2-NH2),4.57 (s,4H; 2” Ar-C H2-NH-CO-Ar),7.48 (d,
3J(H,H) = 8.0 Hz,4H; Ar-H),7.56 (d, 3J(H,H) = 8.0 Hz,8H; Ar-H),
Preparation of 35: A mixture of 38 (9.6 g,25.23 mmol) and HCl (37%,
15 mL) in methanol (150 mL) was heated at 648C for 16 h. The crude so-
lution was concentrated and then diethyl ether (500 mL) was added. The
precipitate formed was filtered off,washed with diethyl ether,and dried.
Compound 35 was isolated as a colorless powder (6.2 g,92%). 1H NMR
(400 MHz,[D 6]DMSO/D2O,20 8C): d = 2.79 (s,6H; 2” Ar-NHC H3),
3.82 (s,3H; -COOCH 3),6.80 (m,1H; Ar-H),7.14 ppm (m,2H; Ar-H);
HR-MS: calcd for C10H14N2O2 [M+H]+ 195.1128; found 195.1128.
7.91 (m,8H; Ar-H),7.99 (d,
3J(H,H) = 8.0 Hz,8H; Ar-H),8.05 (d,
3J(H,H) = 2.0 Hz,2H; Ar-H),8.36 (m,2H; Ar-H),8.43 ppm (m,1H;
Ar-H); MALDI-MS (matrix: dihydroxybenzoic acid): m/z: 1220 [M+H]+
.
7N: Yield: 83%; 1H NMR (600 MHz,[D 6]DMSO/D2O,80 8C): d = 4.11
(s,16H; 8” Ar-C H2-NH2),4.56 (s,12H; 6” Ar-C H2-NH-CO-Ar),7.46
(m,12H; Ar-H),7.54 (d, 3J(H,H) = 8.0 Hz,16H; Ar-H),7.88 (m,24H;
Preparation of 37: Di-tert-butyl dicarbonate (30.1 g,138.0 mmol) was
added portionwise to a solution of 36 (10 g,65.72 mmol), tert-butyl alco-
hol (75 mL),and 1 n NaOH (75 mL) and the mixture was stirred at 208C
for 16 h. The solution was then washed with cyclohexane (3”100 mL).
The aqueous phase was acidified by dropwise addition of 1n HCl and ex-
tracted with ethyl acetate (3”100 mL). The combined organic phases
were washed with brine (3”100 mL) and dried with Na2SO4. The product
was purified by filtration through SiO2 (EtOAc). Compound 37 was iso-
lated as a beige solid (20.2 g,87%). 1H NMR (400 MHz,[D 6]DMSO,
208C): d = 1.47 (s,18H; 2” tert-butyl),7.68 (m,2H; Ar-H),7.86 (m,
1H; Ar-H),9.49 (s,2H; 2” Ar-NHBoc),12.82 ppm (s,1H; -COOH);
HR-MS: calcd for C17H24N2O6 [M+Na]+ 375.1527; found 375.1526.
Ar-H),7.95 (d, 3J(H,H)
=
8.0 Hz,16H; Ar-H),8.03 (d,
3J(H,H)
2.0 Hz,2H; Ar-H),8.28 (d,
3J(H,H)
=
2.0 Hz,2H; Ar-H),8.25 (t,
=
3J(H,H) = 2.0 Hz,4H; Ar-H),8.31 ppm (t, 3J(H,H) = 2.0 Hz,1H; Ar-
H); MALDI-MS (matrix: dihydroxybenzoic acid): m/z: 2822 [M+H]+.
8N: Yield 91%; 1H NMR (600 MHz,[D 6]DMSO/D2O,80 8C): d = 4.10
(s,32H; 16” Ar-C H2-NH2),4.55 (s,28H; 14” Ar-C H2-NH-CO-Ar),7.47
(m,28H; Ar-H),7.54 (m,32H; Ar-H),7.85–7.95 (m,88H; Ar-H),8.02
(m,2H; Ar-H),8.23 (m,6H; Ar-H),8.27 (m,8H; Ar-H),8.33 ppm (m,
1H; Ar-H); MALDI-MS (matrix: a-cyano-4-hydroxycinnamic acid): m/z:
6025 [M+H]+.
9N: Yield 92%; 1H NMR (600 MHz,[D 6]DMSO/D2O,80 8C): d = 4.10
(s,64H; 32” Ar-C H2-NH2),4.52 (s,60H; 30” Ar-C H2-NH-CO-Ar),
7.35–7.55 (m,124H; Ar-H),7.77–8.00 (m,186H; Ar-H),8.10–8.25 ppm
(m,31H; Ar-H); no MS could be obtained neither by ES nor by
MALDI.
Preparation of 38: NaH (60%,8.3 g,206.4 mmol) was added portionwise
to a solution of 37 (9.7 g,27.53 mmol) in THF (100 mL). The resulting
suspension was stirred for 1 h at 208C. Methyl iodide (25.8 mL,
412.9 mmol) was added and stirring was continued for 16 h. Iced water
(100 mL) was then carefully added to quench the reaction. The solution
was acidified with 2n HCl and extracted with ethyl acetate (3”150 mL).
The organic phase was washed with 1n HCl (3”100 mL) and brine (3”
100 mL). The solution was dried with Na2SO4 and the product was puri-
fied by filtration through SiO2 (EtOAc). Compound 38 was isolated as a
beige solid (9.7 g,93%). 1H NMR (400 MHz,[D 6]DMSO,20 8C): d =
1.41 (s,18H; 2” tert-butyl),3.21 (s,6H; 2” Ar-NCH 3Boc),7.47 (m,1H;
Ar-H),7.63 (m,2H; Ar-H),13.11 ppm (s,1H; -COOH); HR-MS: calcd
for C19H28N2O6 [M+Na]+ 425.1659; found 425.1660.
1
10N: Yield 97%; H NMR (600 MHz,[D 6]DMSO/D2O,80 8C): d = 4.11
(s,128H; 64” Ar-C H2-NH2),4.56 (s,124H; 62” Ar-C H2-NH-CO-Ar),
7.43–7.60 (m,252H; Ar-H),7.77–8.00 (m,378H; Ar-H),8.25–8.50 ppm
(m,63H; Ar-H); no MS could be obtained neither by ES nor by
MALDI.
11N: Yield 97%; 1H NMR (400 MHz,[D 6]DMSO,80 8C): d = 2.32 (m,
6H; 3” -CH2-CH(CH2-NH-)2),2.42 (m,3H; 3” -CH 2-CH(CH2-NH-)2),
3.37 (m,12H; 3” -CH 2-CH(CH2-NH-)2),4.11 (s,8H; 2” Ar-C H2-NH2),
4.35 (d, 3J(H,H) = 5.5 Hz,4H; 2” Ar-C H2-NH-),7.35 (m,4H; Ar-H),
7.54 (m,8H; Ar-H),7.79 (m,4H; Ar-H),7.90 (m,8H; Ar-H),8.25–
8.40 ppm (m,14H; 3” -CH 2-CH(CH2-NH-)2,4” Ar-CH 2-NH2);
MALDI-MS: m/z: 1159 [M+H]+.
Preparation of 39:
A
mixture of 34 (1.2 g,1.82 mmol),2
n NaOH
(10 mL),and dioxane (10 mL) was stirred at 20 8C for 16 h. The mixture
was then added dropwise to well-stirred 1n HCl (200 mL),and the re-
sulting precipitate was filtered off,washed with diethyl ether,and dried.
Compound 39 was isolated as a beige solid (1.05 g,89%). 1H NMR
(400 MHz,[D 6]DMSO,80 8C): d = 1.36 (s,18H; 2” tert-butyl),3.23 (s,
6H; 2” Ar-NCH3-CO-Ar),4.09 (d, 3J(H,H) = 6.0 Hz,4H; 2” Ar-C H2-
1
19N: Yield 96%; H NMR (400 MHz,[D 6]DMSO/D2O,20 8C): d = 0.99
(m,12H),1.40 (m,12H),1.54 (m,6H),1.84 (m,24H),2.30 (m,6H),
2.68 (m,8H),3.09 (m,4H),7.61 (d,
3J(H,H) = 2.0 Hz,4H; Ar-H),8.88
NHBoc),6.96 (brm,2H; 2” Ar-CH 2-NHBoc),7.14 (d, 3J(H,H)
8.0 Hz,4H; Ar-H),7.19 (d, 3J(H,H) = 8.0 Hz,4H; Ar-H),7.31 (m,1H;
Ar-H),7.42 (m,2H; Ar-H),12.67 ppm (s,1H; -COOH); HR-MS: calcd
for C35H42N4O8 [M+Na]+ 669.2895; found 669.2896.
=
3
(d, J(H,H) = 2.0 Hz,2H; Ar-H),7.13 (t, 3J(H,H) = 2.0 Hz,2H; Ar-H),
8.19 ppm (t, 3J(H,H)
= 2.0 Hz,1H; Ar-H); MALDI-MS: m/z: 1255
[M+H]+.
24N: Yield 93%; 1H NMR (400 MHz,CD 3OD,20 8C): d = 2.42 (m,
12H; 6” -CH2-CH2-CO-),2.96 (t, 3J(H,H) = 7.0 Hz,8H; 4” H 2N-CH2-
CH2-),3.40 (t, 3J(H,H) = 7.0 Hz,4H; 2” Ar-CONH-C H2-CH2-),7.70
(m,4H; Ar-H),7.95 (m,2H; Ar-H),8.02 (m,2H; Ar-H),8.10 ppm (m,
1H; Ar-H); ES-MS: m/z: 1100 [M+H]+.
Preparation of 40: A mixture of 34 (0.60 g,0.91 mmol),TFA (15 mL),
water (1.2 mL),and triethylsilane (0.6 mL) was stirred at 20 8C for 3 h.
The mixture was then added dropwise to diethyl ether (200 mL),and the
precipitate formed was filtered off and dried. The TFA salt of compound
40 was isolated as a beige solid (0.53 g,85%). 1H NMR (400 MHz,
[D6]DMSO/D2O,20 8C): d = 3.20 (s,6H; 2” Ar-NC H3-CO-Ar),3.73 (s,
1
29N: Yield 96%; H NMR (400 MHz,[D 6]DMSO/D2O,80 8C): d = 2.97
3H; -COOCH3),3.98 (s,4H; 2” Ar-C H2-NH2),7.28 (d, 3J(H,H)
8.0 Hz,4H; Ar-H),7.33 (d, 3J(H,H) = 8.0 Hz,4H; Ar-H),7.46 (m,1H;
=
(s,6H; 2” Ar-CH 2-NCH3-CO-Ar),4.13 (s,8H; 4” Ar-C H2-NH2),4.75
(s,4H; 2” Ar-C H2-NCH3-CO-Ar),7.46 (d, 3J(H,H) = 8.0 Hz,4H; Ar-
H),7.60 (m,12H; Ar-H),8.01 (m,12H; Ar-H),8.09 (d,
3J(H,H)
=
Ar-H),7.48 ppm (m,2H; Ar-H); HR-MS: calcd for
C
26H28N4O4
[M+Na]+ 483.2003; found 483.2004.
2.0 Hz,2H; Ar-H),8.40 (m,2H; Ar-H),8.55 ppm (m,1H; Ar-H);
MALDI-MS: m/z: 1247 [M+H]+.
Modifications of dendrimer endgroups: General procedure for the remov-
al of Boc groups: The Boc-protected dendrimer 4, 6, 7, 8, 9, 10, 11, 19,
24, 29 or 30 (100 mg) was dissolved in a mixture of TFA (5 mL),water
(0.4 mL),and mercaptoethanol (0.2 mL). The solution was stirred at
208C for 3 h,then added dropwise to diethyl ether (50 mL),and the pre-
1
30N: Yield 91%; H NMR (400 MHz,[D 6]DMSO/D2O,20 8C): d = 3.18
(s,18H; 6” Ar-NC H3-CO-),3.97 (s,8H; 4” Ar-C H2-NH2),4.29 (s,4H;
2” Ar-CH2-NHCOAr),7.11–7.45 ppm (m,33H; Ar-H); MALDI-MS:
m/z: 1303 [M+H]+.
114
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA,Weinheim
Chem. Eur. J. 2006, 12,99 – 117

Inhibition of Polyvalent Carbohydrate–Protein Interactions with Glycodendrimers
FULL PAPER
1
General procedure for chloroacetylation: The TFA salt of the dendrimer
4N, 6N, 7N, 8N, 9N, 10N, 11N, 19N, 24N, 29N or 30N (100 mg) was dis-
solved in a mixture of DMF (1 mL) and 2,6-lutidine (0.6 mL). A solution
of chloroacetic anhydride (3.0 equiv./amino group) in DMF (1 mL) was
added dropwise. The mixture was stirred at 208C for 3 h. Dropwise addi-
tion to 1n HCl (50 mL) gave a beige precipitate,which was filtered off
and washed with water,acetone,CH 2Cl2,and diethyl ether.
30Cl: Yield 97%; H NMR (400 MHz,[D ]DMSO/D2O,80 8C): d = 3.19
6
(s,12H; 4” Ar-NC H3-CO-Ar),3.20 (s,6H; 2” Ar-NC H3-CO-Ar),4.04
(s,8H; 4” -CH 2-Cl),4.26 (s,8H; 4” Ar-C H2-NH-CO-CH2Cl),4.32 (s,
4H; 2” Ar-CH2-NH-CO-Ar),7.11–7.19 (m,26H; Ar-H),7.29 (m,1H;
Ar-H),7.42 ppm (m,6H; Ar-H); MALDI-MS: m/z: 1631 [M+Na]+.
General procedure for the functionalization with oligosaccharides: The
chloroacetylated dendrimer 4Cl, 6Cl, 7Cl, 8Cl, 9Cl, 10Cl, 11Cl, 19Cl,
24Cl, 29Cl or 30Cl (25 mg) and the trisaccharide aGal-SH[7a] or disac-
charide Lac-SH[7a] (1.5 equiv/-CH2Cl) were dissolved in degassed DMF
(2.0 mL). At room temperature,DBU (5 equiv/-CH 2Cl) was added and
the solution was stirred under nitrogen for 1 h. It was then added drop-
wise to a mixture of ethanol and diethyl ether (40 mL,1:2). The precipi-
tate that formed was collected,washed with diethyl ether (3”40 mL),
dried,dissolved in water (to which 10 drops of 1 n HCl had been added),
and subjected to ultrafiltration. Ultrafiltrations were performed using
Amicon 8010 stirred cells (vol.: 10 mL,diam.: 25 mm) and Amicon YM3
disc membranes (molecular weight cut-off: 3000). Ultrafiltrations were
repeated eight times (from 10 mL down to 2 mL) with the volume being
made up with distilled water on each occasion. Lyophilization afforded
the products as colorless solids. No ultrafiltration was performed for
4Gal. Instead,the product was precipitated by dropwise addition of the
reaction mixture to a mixture of ethanol and diethyl ether (1:1). The pre-
cipitate formed was collected and purified by chromatography on an RP-
18 column; gradient H2O!H2O/MeOH,1:1.
4Cl: Yield 73%; 1H NMR (400 MHz,[D 6]DMSO,20 8C): d = 4.14 (s,
4H; 2” -CH2-Cl),4.41 (d, 3J(H,H) = 6.0 Hz,4H; 2” Ar-C H2-NH-CO-
CH2-Cl),7.43 (d, 3J(H,H)
=
8.0 Hz,4H; Ar-H),7.97 (d,
3J(H,H)
=
8.0 Hz,4H; Ar-H),8.12 (m,2H; Ar-H),8.62 (m,1H; Ar-H),8.83 (t,
3J(H,H) = 6.0 Hz,2H; NH),10.41 ppm (s,2H; NH); ES-MS:
m/z: 569
[M+H]+.
6Cl: Yield 93%; 1H NMR (600 MHz,[D 6]DMSO/D2O,80 8C): d = 4.08
(s,8H; 4” -CH 2-Cl),4.38 (s,8H; 4” Ar-C H2-NH-CO-CH2-Cl),4.56 (s,
4H; 2” Ar-CH2-NH-CO-Ar),7.40 (d, 3J(H,H) = 8.0 Hz,8H; Ar-H),
7.48 (d, 3J(H,H) = 8.0 Hz,4H; Ar-H),7.92 (m,16H; Ar-H),8.08 (m,
2H; Ar-H),8.32 (m,2H; Ar-H),8.42 ppm (m,1H; Ar-H).
7Cl: Yield 83%; 1H NMR (600 MHz,[D 6]DMSO/D2O,80 8C): d = 4.05
(s,16H; 8” -C H2-Cl),4.36 (s,16H; 8” Ar-C H2-NH-CO-CH2-Cl),4.54 (s,
12H; 6” Ar-CH2-NH-CO-Ar),7.38 (d, 3J(H,H) = 8.0 Hz,16H; Ar-H),
7.46 (m,12H; Ar-H),7.87 (m,40H; Ar-H),8.06 (m,2H; Ar-H),8.25 (m,
6H; Ar-H),8.33 ppm (m,1H; Ar-H).
8Cl: Yield 91%; 1H NMR (600 MHz,[D 6]DMSO/D2O,80 8C): d = 4.06
(s,32H; 16” -C H2-Cl),4.36 (s,32H; 16” Ar-C H2-NH-CO-CH2-Cl),4.54
(s,28H; 14” Ar-C H2-NH-CO-Ar),7.38 (m,32H; Ar-H),7.46 (m,28H;
Ar-H),7.89 (m,88H; Ar-H),8.06 (m,2H; Ar-H),8.28 (m,14H; Ar-H),
8.38 ppm (m,1H; Ar-H).
4Gal: Yield 31 mg (50%); 1H NMR (400 MHz,[D 6]DMSO/D2O,80 8C,
selected signals): d = 1.83 (s,6H; 2” -NHCOC H3),3.19 (s,4H; 4”
-CO-CH2-S-),4.31 (2” d, 3J(H,H) = 8.0 Hz,4H; 2” GlcNAc H-1,2”
bGal H-1),4.36 (s,4H; 2” Ar-C H2-NH-CO-CH2-S),4.89 (d, 3J(H,H) =
3.5 Hz,2H; 2” aGal H-1),7.40 (d, 3J(H,H) = 8.0 Hz,4H; Ar-H),7.89
(d, 3J(H,H) = 8.0 Hz,4H; Ar-H),8.05 (d, 3J(H,H) = 2.0 Hz,2H; 2”
Ar-H),8.33 ppm (t, 3J(H,H) = 2.0 Hz,1H; Ar-H); HR-MS: calcd for
C81H118N8O40S2 [M+2Na]2+ 976.3336; found 976.3331.
9Cl: Yield 92%; 1H NMR (600 MHz,[D 6]DMSO/D2O,80 8C): d = 4.07
(s,64H; 32” -CH -Cl),4.37 (s,64H; 32” Ar-C H2-NH-CO-CH2-Cl),4.55
2
(s,60H; 30” Ar-C H2-NH-CO-Ar),7.39 (m,64H; Ar-H),7.47 (m,60H;
Ar-H),7.92 (m,184H; Ar-H),8.07 (m,2H; Ar-H),8.33 ppm (m,30H;
Ar-H); the expected signal at dꢀ8.38 ppm (m,1H; Ar-H) could not be
6Gal: Yield 36 mg (70%); 1H NMR (600 MHz,[D 6]DMSO/D2O,80 8C,
selected signals): d = 1.81 (s,12H; 4” -NHCOC H3),3.17 (s,8H; 4”
-CO-CH2-S-),4.28–4.34 (m,16H; 4” Ar-C
H2-NH-CO-CH2-S-,4”
detected in the noise.
GlcNAc H-1,4” bGal H-1),4.55 (s,4H; 2” Ar-C H2-NH-CO-Ar),4.89
(d, 3J(H,H) = 3.5 Hz,4H; 4” aGal H-1),7.37 (d, 3J(H,H) = 8.0 Hz,
8H; Ar-H),7.45 (d, 3J(H,H) = 8.0 Hz,4H; Ar-H),8.19 ppm (m,3H;
Ar-H); HR-MS: calcd for C185H254N20O82S4 [Mꢁ3H]3ꢁ 1397.4995; found
1397.4998.
7Gal: Yield 243 mg (80%); 1H NMR (600 MHz,[D 6]DMSO/D2O,80 8C,
selected signals): d = 1.81 (s,24H; 8” -NHCOC H3),3.18 (s,16H; 8”
1
10Cl: Yield 97%; H NMR (600 MHz,[D ]DMSO/D2O,80 8C): d = 4.04
6
(s,128H; 64” -CH 2-Cl),4.35 (s,128H; 64” Ar-C H2-NH-CO-CH2-Cl),
4.55 (s,124H; 62” Ar-C H2-NH-CO-Ar),7.36 (m,128H; Ar-H),7.44 (m,
124H; Ar-H),7.88 (m,376H; Ar-H),8.05 (m,2H; Ar-H),8.26 ppm (m,
30H; Ar-H); the expected signal at dꢀ8.38 ppm (m,1H; Ar-H) could
not be observed in the noise.
1
11Cl: Yield 87%; H NMR (400 MHz,[D ]DMSO/D2O,80 8C): d = 2.29
6
-CO-CH2-S-),4.29–4.35 (m,32H; 8” Ar-C
H2-NH-CO-CH2-S-,8”
(m,6H; 3” -C H2-CH(CH2-NH-)2),2.37 (m,3H; 3” -CH 2-CH(CH2-NH-)2),
3.35 (m,12H; 3” -CH 2-CH(CH2-NH-)2),4.08 (s,8H; 4” -CO-C H2Cl),
4.32 (s,4H; 2” Ar-C H2-NH-),4.36 (s,8H; 4” Ar-C H2-NH-),7.34 (m,
GlcNAc H-1,8” bGal H-1),4.56 (s,12H; 6” Ar-C H2-NH-CO-Ar),4.86
(d, 3J(H,H) = 3.5 Hz,8H; 8” aGal H-1),7.39 (d, 3J(H,H) = 8.0 Hz,
16H; Ar-H),7.47 (m,12H; Ar-H),8.29 (m,1H; Ar-H),8.34 (m,4H; Ar-
H),8.38 ppm (m,2H; Ar-H); MALDI-MS: m/z: 8801 [M+Na]+.
12H; Ar-H),7.78 ppm (m,12H; Ar-H); MALDI-MS:
m/z: 1465
[M+H]+.
7Lac: Yield 8 mg (75%); 1H NMR (600 MHz,[D ]DMSO/D2O,80 8C,se-
6
19Cl: Yield 91%; 1H NMR (400 MHz,[D 6]DMSO,80 8C): d = 1.00 (m,
12H),1.39–1.47 (m,16H),1.59 (m,2H),1.79–1.90 (m,24H),2.30 (m,
6H),3.01 (m,8H),3.15 (m,4H),4.03 (s,8H; 4” -COC
4H; Ar-H),7.86 (m,2H; Ar-H),7.89 (brs,4H; 4” -CH 2-NHCOCH2Cl),
7.97 (brs,2H; 2” -CH 2-NHCOAr),8.05 (m,2H; Ar-H),8.17 (m,1H;
Ar-H),9.61 (s,4H; 4” Ar-NH-CO-),9.68 (s,2H; 2” Ar-NH-CO-),
12.43 ppm (brs,1H; Ar-COOH); MALDI-MS: m/z: 1583 [M+Na]+.
lected signals): d = 3.17 (s,16H; 8” -CO-CH 2-S-),4.11 (d, 3J(H,H) =
8.0 Hz,8H; 8” H-1 carbohydr.),4.19 (m,8H; 8” H-1 carbohydr.),4.34
(m,16H; 8” Ar-C H2-NH-CO-CH2-S-),4.56 (s,12H; 6” Ar-C H2-NH-
CO-Ar),7.38 (d, 3J(H,H) = 8.0 Hz,16H; Ar-H),7.47 (m,12H; Ar-H),
8.26–8.29 ppm (m,7H; Ar-H); MALDI-MS: m/z: 7176 [M+Na]+.
H2Cl),7.63 (m,
8Gal: Yield 57 mg (70%); 1H NMR (600 MHz,[D 6]DMSO/D2O,80 8C,
selected signals): d = 1.80 (s,48H; 16” -NHCOC H3),3.16 (s,32H; 16”
-CO-CH2-S-),4.28–4.34 (m,64H; 16” Ar-C H2-NH-CO-CH2-S-,16”
GlcNAc H-1,16” bGal H-1),4.54 (s,28H; 14” Ar-C H2-NH-CO-Ar),
4.87 (d, 3J(H,H) = 3.5 Hz,16H; 16” aGal H-1),7.37 (d, 3J(H,H) =
8.0 Hz,32H; Ar-H),7.46 (m,28H; Ar-H),8.27 (m,13H; Ar-H),
8.29 ppm (m,2H; Ar-H); MALDI-MS: m/z: 17989 (broad signal) within
experimental error for [M+K]+.
9Gal: Yield 13 mg (90%); 1H NMR (600 MHz,[D 6]DMSO/D2O,80 8C,
selected signals): d = 1.81 (s,96H; 32” -NHCOC H3),3.17 (s,64H; 32”
-CO-CH2-S-),4.28–4.35 (m,128H; 32” Ar-C H2-NH-CO-CH2-S-,32”
GlcNAc H-1,32” bGal H-1),4.55 (s,60H; 30” Ar-C H2-NH-CO-Ar),
4.86 (d, 3J(H,H) = 3.5 Hz,32H; 32” aGal H-1),7.38 (m,64H; Ar-H),
7.47 (m,60H; Ar-H),8.30–8.35 ppm (m,31H; Ar-H); no MS could be
obtained by MALDI.
24Cl: Yield 88%; 1H NMR (400 MHz,CD 3OD,80 8C,selected signals):
d = 2.39 (m,12H; 6” -CH 2-CH2-CO-),3.22 (t, 3J(H,H) = 7.0 Hz,8H;
4” Cl-CH2(CO)NH-CH2-CH2-),3.38 (t, 3J(H,H) = 7.0 Hz,4H; 2” Ar-
(CO)HN-CH2-CH2-),4.00 (s,8H; 4” Cl-C H2(CO)NH-),7.69 (m,4H;
Ar-H),7.94 (m,2H; Ar-H),7.97 (m,2H; Ar-H),8.10 ppm (m,1H; Ar-
H); ES-MS: m/z: 1427 [M+Na]+.
1
29Cl: Yield 77%; H NMR (400 MHz,[D ]DMSO/D2O,80 8C): d = 2.96
6
(s,6H; 2” Ar-CH -NCH3-CO-Ar),4.09 (s,8H; 4” -C H2Cl),4.39 (s,8H;
2
4” Ar-CH2-NH-CO-CH2Cl),4.74 (s,4H; 2” Ar-C H2-NCH3-CO-Ar),
7.41 (m,12H; Ar-H),7.60 (m,4H; Ar-H),7.92 (d,
8H; Ar-H),7.99 (d, 3J(H,H) = 8.0 Hz,4H; Ar-H),8.10 (m,2H; Ar-H),
3J(H,H) = 8.0 Hz,
8.31 (m,2H; Ar-H),8.50 ppm (m,1H; Ar-H); MALDI-MS:
m/z: 1575
[M+Na]+.
Chem. Eur. J. 2006, 12,99 – 117
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA,Weinheim
www.chemeurj.org
115

G. Thoma et al.
10Gal: Yield 13 mg (85%); 1H NMR (600 MHz,[D 6]DMSO/D2O,80 8C,
selected signals): d = 1.81 (s,192H; 64” -NHCOC H3),3.17 (s,128H;
64” -CO-CH2-S-),4.29–4.35 (m,256H; 64” Ar-C H2-NH-CO-CH2-S-,
64” GlcNAc H-1,64” bGal H-1),4.56 (s,124H; 62” Ar-C H2-NH-CO-
Ar),4.86 (d, 3J(H,H) = 3.5 Hz,64H; 64” aGal H-1),7.39 (d, 3J(H,H) =
8.0 Hz,128H; Ar-H),7.47 (m,124H; Ar-H),8.30–8.35 ppm (m,63H; Ar-
H); no MS could be obtained by MALDI.
11Gal: Yield 38 mg (68%); 1H NMR (400 MHz,[D 6]DMSO/D2O,80 8C,
selected signals): d = 1.82 (s,12H; 4” NHCOC H3),4.31–4.37 (m,20H;
6” Ar-CH2-NH-CO-,4” GlcNAc H-1,4” bGal H-1),4.87 (d, 3J(H,H)
Computational methods: All calculations were performed within Macro-
model 8.5 (Schrçdinger,L. L. C.) applying the AMBER* force field with
a distance-dependent dielectric constant of 4.0 *R. The starting structures
were submitted to 10000 steps of Monte Carlo (MC) minimizations. Two
sets of MC perturbations were applied: new starting structures were gen-
erated by randomly modifying torsion angles and by variations along
random linear combinations of the 20 lowest vibrational modes
(LMOD).[13] All rotatable bonds (those not adjacent to amide bonds)
were included. To achieve a good balance between searching within one
well of the potential energy surface and jumping from one well to anoth-
er,a 50% mixture of both perturbations was applied. Of the minimized
= 3.5 Hz,4H; 4” aGal H-1),7.33 (m,12H; Ar-H),7.78 ppm (d,
3J-
= 8.0 Hz,12H; Ar-H); HR-MS: calcd for C 179H266N20O82S4
ꢁ1
structures,only those within 200 kJmol
of the current global minimum
(H,H)
were accepted. Building and visualization of structures was performed in
the framework of the molecular graphics program WitnotP (A. Widmer,
unpublished results).
[Mꢁ4H]4ꢁ 1032.8963; found 1032.8968.
19Gal: Yield 41 mg (89%); 1H NMR (400 MHz,[D 6]DMSO/D2O,80 8C,
selected signals): d = 1.83 (s,12H; 4” -NHCOC H3),3.13 (s,8H; 4”
-CH2-S-),4.32 (d, 3J(H,H) = 7.0 Hz,4H; 4” GlcNAc H-1),4.36 (d,
3J(H,H) = 8.0 Hz,4H; 4” bGal H-1),4.87 (d, 3J(H,H) = 3.5 Hz,4H;
4” aGal H-1),7.60 (d, 3J(H,H) = 2.0 Hz,4H; Ar-H),7.84 (m,2H; Ar-
H),8.03 (m,2H; Ar-H),8.12 ppm (m,1H; Ar-H); HR-MS: calcd for
ELISAs: a) human serum: On plate A,wells were coated with 0.5 mg of
aGal-HSA conjugate in 100 mL of NaHCO3 (100 mm,pH 9.6) for 2 h at
378C. Blocking with 250 mL of PBS containing 0.5% Tweenꢁ20 for 2 h at
ambient temperature was followed by removal of the liquid. On plate B,
pooled human serum was diluted 1:20 in PBS containing 10% Sea-
blockTM and 0.2% Tweenꢁ20 and incubated for 30 min with serial dilu-
tions of the glycodendrimers. Thereafter,100 mL from each well on plate
B was transferred to plate A,incubated for 1 h at ambient temperature,
and washed three times with PBS containing 0.15% Tweenꢁ20. Then,the
mixtures were incubated with 100 mL of PBS containing anti-IgM secon-
dary antibody peroxidase conjugate (diluted 1/2000),10% Blocker-
BSATM,and 0.2% Tween ꢁ20. After washing six times with 300 mL of PBS
containing 0.1% Tween 20,color development was initiated by adding
100 mL of TMB substrate solution and stopped after 5 min by adding
50 mL of 1m H2SO4. The A450 value was measured.
C
185H290N20O82S4 [Mꢁ4H]4ꢁ 1056.9432; found 1056.9426.
1
24Gal: Yield 42 mg (47%); H NMR (400 MHz,D O,20 8C,selected sig-
2
nals): d = 1.90 (s,12H; 4” -NHCOC H3),3.05 (s,8H; 4” -CO-C H2-S-),
4.31 (d, 3J(H,H) = 8.0 Hz,4H; 4” GlcNAc H-1),4.41 (d, 3J(H,H) =
8.0 Hz,4H; 4” bGal H-1),5.03 (d, 3J(H,H) = 3.5 Hz,4H; 4” aGal H-
1),7.35–8.00 (m,6H; Ar-H),7.50 (m,1H; Ar-H),7.72 ppm (m,2H; Ar-
H); HR-MS: calcd for C173H278N20O82S4 [Mꢁ4H]4ꢁ 1017.9198; found
1017.9206.
29Gal: Yield 25 mg (66%); 1H NMR (400 MHz,[D 6]DMSO/D2O,80 8C,
selected signals): d = 1.82 (s,12H; 4” NHCOC H3),2.96 (s,6H; 2” Ar-
CH2-NCH3-CO-Ar),3.18 (s,8H; 4” -CO-C H2-S-),4.31–4.36 (m,16H;
b) Cynomolgus monkey serum: Wells were coated with 0.5 mg of aGal-
polymer[7] in 100 mL of PBS overnight at 48C. Blocking with 300 mL of
PBS containing 10% SeablockTM for 2 h at ambient temperature was fol-
lowed by removal of the liquid. The wells were incubated for 30 min with
100 mL of serially diluted serum sample in PBS containing 10% Sea-
blockTM and 0.2% Tweenꢁ20,and subsequently washed three times with
PBS containing 0.15% Tween 20. Incubation with secondary antibody
and color development were performed as described above.
4” Ar-CH2-NH-CO-CH2-S-,4” GlcNAc H-1,4”
4H; 2” Ar-CH2-NCH3-CO-Ar),4.87 (d, 3J(H,H)
bGal H-1),4.73 (s,
3.5 Hz,4H; 4”
=
aGal H-1),7.59 (d, 3J(H,H) = 2.0 Hz,4H; Ar-H),8.08 (m,2H; Ar-H),
8.31 (m,2H; Ar-H),8.48 ppm (m,1H; Ar-H); HR-MS: calcd for
C
187H258N20O82S4 [Mꢁ3H]3ꢁ 1406.8433; found 1406.8426.
30Gal: Yield 43 mg (92%); 1H NMR (400 MHz,[D 6]DMSO/D2O,80 8C,
selected signals): d = 1.82 (s,12H; 4” NHCOC H3),2.50 (s,12H; 4”
Ar-N(CH3)-CO-),2.54 (s,6H; 2” Ar-N(C H3)-CO-),3.12 (s,8H; -CO-
CH2-S-),4.23 (s,8H; 4” Ar-C H2-NH-CO-CH2-S-),4.31–4.37 (m,12H;
2” Ar-CH2-NH-CO-Ar,4” GlcNAc H-1,4” bGal H-1),4.57 (s,4H; 2”
Ar-CH2-NH-CO-Ar),4.87 (d, 3J(H,H) = 3.5 Hz,4H; 4” aGal H-1),
7.30 (m,1H; Ar-H),7.43 ppm (m,6H; Ar-H); HR-MS: calcd for
Haemolysis assays: a) human serum: Pig erythrocytes were obtained
from heparinized pig blood,washed three times with CFD buffer
(pH 7.4),and suspended in CFD at a concentration of 1”10 9 mLꢁ1. On
low-bind U-shaped plates,seven serial dilutions of inhibitor were pre-
pared in 50 mL of CFD. Human serum was serially diluted in 50 mL of
CFD (ten dilutions assayed for each inhibitor concentration). Then,
50 mL of pig erythrocyte solution was added to each well and the plate
was incubated for 60 min at 378C with mild shaking. Plates were centri-
fuged for 10 min at 2000” g to precipitate the unlysed erythrocytes,the
supernatant was transferred to a flat-bottomed plate,and released hemo-
globin (A420) was measured.
C
191H266N20O82S4 [M+3Na]3+ 1449.5273; found 1449.5268.
Multi-angle light scattering (MALS): All determinations of the weight-
average molar mass (MWAggr) and of dn/dc were performed at 208C,
unless otherwise stated. They were carried out by Wyatt Technology
Deutschland GmbH and by Wellensiek Laborgeraete using a DAWNꢁ
EOS and
a Wyatt Optilab 903 refractometer. Concentrations were
b) Cynomolgus monkey serum: As above,but with the addition of 50 mL
of baby rabbit complement (heterologous complement source at a final
dilution of 1:10) in CFD and 50 mL of pig erythrocyte solution to each
well.
0.3 mgmLꢁ1 for 4Gal, 7Gal, 7Lac, 8Gal, 9Gal, 10Gal,and 24Gal and
0.06 mgmLꢁ1 for 6Gal, 11Gal, 19Gal, 29Gal,and 30Gal. Aqueous solu-
tions were heated for 30 min at 808C and then allowed to cool to 208C.
The samples were passed through Spartan 30 or Schleicher & Schuell
0.45 mm filters.
Atom force microscopy (AFM): Aqueous solutions (10 mL,2 mgmLꢁ1 for
6Gal–9Gal,0.2 mgmLꢁ1 for 10Gal) were spotted on freshly cleaved
mica. The solvent was allowed to evaporate for at least 120 min. A Multi-
Mode AFM with a Nanoscope III controller (Digital Instruments) was
used to generate all of the images. The microscope was operated in tap-
ping mode in air with Olympus etched silicon probes (cantilever length
120 mm,tip height 10 mm).
Acknowledgement
G.T. thanks Prof. M. Reza Ghadiri for the opportunity of a sabbatical
stay in his laboratory at the Scripps Research Institute.
Transmission electron microscopy: An aqueous solution of 6Gal (1 mL,
2 mgmLꢁ1) was diluted with 1 mL of a 2% sodium silicotungstate solution
for negative contrasting. The mixture was spotted on a copper net with a
Formvar (polyformaldehyde) surface and the solvent was allowed to
evaporate. A transmission electron microscope (EM910/FA Leo) was
used to generate the image.
[1] a) A. Varki, Glycobiology 1993, 3,97–130; b) R. A. Dwek, Chem.
Rev. 1996, 96,622; c) Essentials in Glycobiology,(Eds.: A. Varki,R.
Cummings,J. Esko,H. Freeze,G. Hart,J. Marth),Cold Spring
Harbor Laboratory Press,Plainview,NY, 1999.
116
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA,Weinheim
Chem. Eur. J. 2006, 12,99 – 117

Inhibition of Polyvalent Carbohydrate–Protein Interactions with Glycodendrimers
FULL PAPER
[2] a) M. Mammen,S.-K. Choi,G. M. Whitesides, Angew. Chem. 1998,
110,2908–2953; Angew. Chem. Int. Ed. 1998, 37,2754–2794;
b) R. T. Lee,Y. C. Lee, Glycoconjugate J. 2000, 17,543–551; c) J. J.
Lundquist,E. J. Toone, Chem. Rev. 2002, 102,555–578; d) G. Erco-
lani, J. Am. Chem. Soc. 2003, 125,16097–16103; e) P. I. Kitov,D. R.
Bundle, J. Am. Chem. Soc. 2003, 125,16271–16284; f) L. L. Kies-
sling,J. K. Pontrello,M. C. Schuster, Carbohydrate-Based Drug Dis-
covery 2003, 2,575–608; g) L. L. Kiessling,T. Young,K. H. Mortell,
Glycosci. 2001, 2,1817–1861.
[3] Y. C. Lee,R. T. Lee, Acc. Chem. Res. 1995, 28,321–327.
[4] a) G. Thoma,J. T. Patton,J. L. Magnani,B. Ernst,R. Öhrlein,R. O.
Duthaler, J. Am. Chem. Soc. 1999, 121,5919–5929; b) G. Thoma,
R. O. Duthaler,J. L. Magnani,J. T. Patton, J. Am. Chem. Soc. 2001,
2002, 110,1869–1877; d) J.-Q. Wang,X. Chen,W. Zhang,S. Zachar-
ek,Y. Chen,P. G. Wang, J. Am. Chem. Soc. 1999, 121,8174–8181.
[8] An elegant,alternative approach is the use of a1–3-galactosyltrans-
ferase gene-knockout pigs as donors. a) K. Yamada,K. Yazawa,A.
Shimizu,T. Iwanaga,Y. Hisashi,M. Nuhn,P. OꢂMalley,S. Nobori,
P. A. Vagefi,C. Patience,J. Fishman,D. K. C. Cooper,R. J. Hawley,
J. Greenstein,H.-J. Schuurman,M. Awwad,M. Sykes,D. H. Sachs,
Nature Med. 2005, 11,32–34; b) K. Kuwaki,Y.-L. Tseng,F. J. M. F.
Dor,A. Shimizu,S. L. Houser,T. M. Sanderson,C. J. Lancos,D. D.
Prabharasuth,J. Cheng,K. Moran,Y. Hisashi,N. Mueller,K.
Yamada,J. L. Greenstein,R. J. Hawley,C. Patience,M. Awwad,
J. A. Fishman,S. C. Robson,H.-J. Schuurman,D. H. Sachs,D. K. C.
Cooper, Nature Med. 2005, 11,29–31; c) C. J. Phelps,C. Koike,T. D.
Vaught,J. Boone,K. D. Wells,S.-H. Chen,S. Ball,S. M. Specht,
I. A. Polejaeva,J. A. Monahan,P. M. Jobst,S. B. Sharma,A. E.
Lamborn,A. S. Garst,M. Moore,A. J. Demetris,W. A. Rudert,R.
123,10113–10114; c) A. Spaltenstein,G. M. Whitesides,
J. Am.
Chem. Soc. 1991, 113,686–687; d) S. K. Choi,M. Mammen,G. M.
Whitesides, J. Am. Chem. Soc. 1997, 119,4103–4111; e) N. V. Bovin,
Glycoconjugate J. 1998, 15,431–446; f) K. H. Mortell,R. V. Weath-
erman,L. L. Kiessling, J. Am. Chem. Soc. 1996, 118,2297–2298;
Bottino,S. Bertera,M. Trucco,T. E. Starzl,Y. Dai,D. L. Ayares,
Sci-
ence 2003, 299,411–414; d) L. Lai,D. Kolber-Simonds,K.-W. Park,
H.-T. Cheong,J. L. Greenstein,G.-S. Im,M. Samuel,A. Bonk,A.
Rieke,B. N. Day,C. N. Murphy,D. B. Carter,R. J. Hawley,R. S.
Prather, Science 2002, 295,1089–1092.
g) C. W. Cairo,J. E. Gestwicki,M. Kanai,L. L. Kiessling,
J. Am.
Chem. Soc. 2002, 124,1615–1619; h) E. Arranz-Plaza,A. S. Tracy,
A. Siriwardena,J. M. Pierce,G.-J. Boons, J. Am. Chem. Soc. 2002,
124,13035–13046; i) S. Thobhani,B. Ember,A. Siriwardena,G.-J.
Boons, J. Am. Chem. Soc. 2003, 125,7154–7155; j) R. Roy,F. D.
Tropper, J. Chem. Soc. Chem. Commun. 1988,1058–1060; k) R.
[9] a) J. M. Lehn, Supramolecular Chemistry,VCH,Weinheim,Germa-
ny,1995; b) D. N. Reinhoudt,M. Crego-Calama, Science 2002, 295,
2403–2407; c) G. M. Whitesides,B. Grzybowski, Science 2002, 295,
2418–2421.
Roy,W. K. Park,O. M. Srivastava,C. Foxall,
Bioorg. Med. Chem.
Lett. 1996, 6,1399–1402; l) M. N. Matrosovich,L. V. Mochalova,
N. E. Byramova,N. V. Bovin, FEBS Lett. 1990, 272,209–212;
m) A. S. Gambaryan,A. B. Tuzikov,A. A. Chinarev,L. R. Juneja,
N. V. Bovin,M. N. Matrosovich, Antiviral Res. 2002, 55,201–205.
[5] a) A. Feinstein,E. A. Munn, Nature 1969, 224,147–149; b) T.
Tobita,M. Oda,T. Azuma, Mol. Immunol. 2004, 40,803–811.
[6] a) Subcellular Biochemistry Vol. 32,(Eds.: U. Galili,J. L. Avila),
Kluwer Academic/Plenum Publishers,New York, 1999; b) M. S. San-
drin,I. F. C. McKenzie, Immunological Reviews 1994, 141,169;
c) D. K. C. Cooper, Lancet 2003, 362,557–559.
[10] Parts of this work have been reported in a preliminary communica-
tion: G. Thoma,A. G. Katopodis,N. Voelcker,R. O. Duthaler,M. B.
Streiff, Angew. Chem. 2002, 114,3327–3330; Angew. Chem. Int. Ed.
2002, 41,3195–3198.
[11] For work on other glycodendrimers,see the following papers and
references therein: a) R. Roy, Polym. News 1996, 21,226–232; b) R.
Roy,M.-G. Baek,K. Rittenhouse-Olson, J. Am. Chem. Soc. 2001,
123,1809–1816; c) N. Rockendorf,T. K. Lindhorst,
Top. Curr.
Chem. 2001, 217,201–238; d) T. K. Lindhorst, Top. Curr. Chem.
2002, 218,201–235; e) W. B. Turnbull,J. F. Stoddart, Rev. Mol. Bio-
technol. 2002, 90,231–255; f) M. J. Cloninger, Curr. Opin. Chem.
Biol. 2002, 6,742–748.
[7] a) R. O. Duthaler,A. G. Katopodis,W. Kinzy,R. Ohrlein,G.
Thoma,WO 9847915, 1998; b) R. Zhong,Y. Luo,H. Yang,B.
Garcia,A. Ghanekar,P. Luke,S. Chakrabarti,G. Lajoie,M. J. Phil-
lips,A. G. Katopodis,R. O. Duthaler,M. Cattral,W. Wall,A. Jevni-
[12] a) J. M. J. Frꢃchet, Science 1994, 263,1710–1715; b) G. R. Newkome,
C. N. Moorefield,F. Voegtle, Dendritic Molecules: Concepts, Synthe-
ses, Perspectives,VCH,Weinheim, 1996.
kar,M. Bailey,G. A. Levy,D. R. Grant,
Transplantation 2003, 75,
10–19; c) A. G. Katopodis,R. G. Warner,R. O. Duthaler,M. B.
Streiff,A. Bruelisauer,O. Kretz,B. Dorobek,E. Persohn,H.
Andres,A. Schweitzer,G. Thoma,W. Kinzy,V. F. J. Quesniaux,E.
[13] I. Kolossvary,W. C. Guida, J. Am. Chem. Soc. 1996, 118,5011–5119.
Received: July 27,2005
Cozzi,H. F. S. Davies,R. Manez,D. White,
J. Clin. Investigation
Published online: October 17,2005
Chem. Eur. J. 2006, 12,99 – 117
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA,Weinheim
www.chemeurj.org
117