Phenylenediamine-based bivalent glycocyclophanes: Synthesis and analysis of the influence of scaffold rigidity and ligand spacing on lectin binding in cell systems with different glycomic profiles

Article abstract of DOI:10.1039/b913010a

The conjugation of carbohydrates to synthetic scaffolds has the goal of preparing potent inhibitors of lectin binding. We herein report the synthesis of a panel of bivalent compounds (cyclophane and terephthalamide-derivatives) then used to establish the

Full text of DOI:10.1039/b913010a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Phenylenediamine-based bivalent glycocyclophanes: synthesis and analysis of
the influence of scaffold rigidity and ligand spacing on lectin binding in cell
systems with different glycomic profiles†
Sabine Andre´,a Trinidad Velasco-Torrijos,b Rosaria Leyden,b Sebastien Gouin,b Manuela Tosin,b
Paul V. Murphy*b,c and Hans-Joachim Gabiusa
Received 1st July 2009, Accepted 17th August 2009
First published as an Advance Article on the web 23rd September 2009
DOI: 10.1039/b913010a
The conjugation of carbohydrates to synthetic scaffolds has the goal of preparing potent inhibitors of
lectin binding. We herein report the synthesis of a panel of bivalent compounds (cyclophane and
terephthalamide-derivatives) then used to establish the influence of scaffold flexibility on respective
inhibitory potency in a medically relevant test system. Synthetic routes to two phenylenediamine-based
glycocyclophanes involving Ugi reactions of glucuronic acid derivatives and subsequent ring closing
metathesis are described, as are improvements for producing terephthalamide-based carbohydrate
carriers. Assays were performed with human tumour cells measuring quantitatively the influence of the
test compounds on fluorescent surface staining by labelled lectins. Biological evaluation using two
different lines of cancer cells as well as cells with known alterations in the glycomic profile (cells treated
with an inhibitor of glycan processing and a glycosylation mutant) reduced the risk of generating
premature generalizations regarding inhibitor potency. Bioactivity relative to free mannose was
invariably determined for the synthetic compounds. A clear trend for enhanced inhibitory properties for
macrocyclic compounds compared to non-macrocyclic derivatives was discerned for one type of
glycocyclophane. Herein we also document the impact of altering the spacing between the mannose
residues, altering cell surface ligand density and cell-type reactivity. The applied strategy for the cell
assays is proposed to be of general importance in the quest to identify medically relevant lectin
inhibitors.
medically relevant activities of cell surface glycans in regulating
Introduction
e. g. adhesion, migration, proliferation or tissue invasion, it is an
The ideal chemical properties of carbohydrates to generate an
attractive aim to develop strategies to efficiently interfere with
unsurpassed structural biodiversity of oligomers and to engage
the binding of certain effectors to cell surface glycans. In other
in intermolecular interactions underlie the concept of the sugar
words, the issue to define structure–activity relationships for lectin-
code.1 Indeed, glycans of cell surface glycoproteins are versatile
blocking compounds is of fundamental importance, especially at
biochemical signalling molecules by serving as direct contact site
the level of cell surfaces which represent the physiological platform
for receptors with carbohydrate specificity (trans-interactions) and
as potent molecular switches regulating the protein component’s
accessibility and functionality (cis-interactions).1,2 Even seemingly
for the interaction process.
Toward this end, we herein put the question to the experimental
test whether and how intramolecular flexibility of distinct scaffolds
carrying two docking sites for a model lectin affects its cell surface
binding. In detail, we compare two rather flexible bivalent com-
pounds (1a, 3a) with the corresponding rigid glycocyclophanes
minor structural changes in these complex carbohydrates, such as
a core substitution in N-glycans, trigger significant alterations
in glycan conformations and lectin affinity.3 Underlining their
salient role for cellular communication, several cases have been
(2a, 4a/4b) (Fig. 1), the central synthetic and cell biological part
elucidated in which orchestrated changes in glycan display and
of our report. In addition, other phenylenediamine-based bivalent
expression of the cognate lectin cooperate to turn sugar-encoded
compounds (5a, 10) and the monovalent compound (6a) were used
information swiftly into cellular responses such as apoptosis
as controls (Fig. 1). The impact of scaffold design on bioactivity is
induction or autoimmune suppression.4 Considering the multiple,
further evaluated by testing four bivalent terephthalamide-based
compounds (7c, 7d, 8, 9)5 and the bivalent mannoside 11, based
on a tetraethylene glycol scaffold (Fig. 1). The selection of the
bivalent ligand systems was based on choosing scaffolds with
varying degrees of rigidity with a view to exploring how changing
both rigidity and the spatial presentation of ligand would alter
inhibition of lectin binding. The spatial relationship between the
two mannose ligands in the two most constrained systems 2a and
4a is shown to be significantly different. The linkage patterns of
the mannose ligands and thus spatial relationship between these
aInstitute of Physiological Chemistry, Faculty of Veterinary Medicine,
Ludwig-Maximilians-University Munich, Veterina¨rstr. 13, 80539 Munich,
Germany
bSchool of Chemistry and Chemical Biology, Centre for Synthesis and
Chemical Biology, University College Dublin, Belfield, Dublin 4. E-mail:
paul.v.murphy@nuigalway.ie; Fax: +353 (0)91 525700
cSchool of Chemistry, National University of Ireland, Galway, University
Rd, Galway
† Electronic supplementary information (ESI) available: Experimental
details, NMR spectra. See DOI: 10.1039/b913010a
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4715–4725 | 4715
©

View Article Online
Fig. 1 Structures of the test panel.
residues is also different when comparing the terephthalamides
7–9, of which the initial synthetic and structural aspects were
reported previously.5 Since glycan display on cells is physiologi-
cally subject to dynamic changes, the design of suitable assays is
a key factor to assess inhibitory capacity of any compound. It
should account for shifts in the glycomic profile to reach valid
conclusions. Consequently, our respective efforts were guided by
the intention to introduce systems with deliberately tunable affinity
and density of cell surface ligands with a view to evaluating
glycocompound potency. Naturally, these two characteristics can
have a bearing on the competition between the synthetic agent
and cell surface glycans for binding to the lectin. Thus, a more
general inhibitory trend for a synthetic ligand warrants being
investigated and this can be achieved by adequate tailoring of the
test system. Toward this aim, treatment of cells with an inhibitor of
glycan processing as well as using a cell type expressing a suitable
glycosylation mutant2g are both viable options to alter glycomic
profile by altering N-glycan structure on the cell surface. Hereby,
the cell surface density of mannose residues and the presentation
of truncated complex-type N-glycans, respectively, are increased.
These changes will enhance binding of the test lectin concanavalin
A (ConA), facilitating inhibitory assays using cells with different
glycomic profiles. In detail, the binding of the labelled lectin to
native cells is monitored automatically on the single-cell level by
fluorescent staining, and an inhibitor, blocking lectin-binding, will
decrease staining intensity.
Experimentally, the lectin’s binding parameters to cells (mean
fluorescence intensity of stained cells and percentage of positive
cells) and reactivity to inhibitors were determined for two different
human tumour cell lines to trace cell-type-dependent alterations.
If the sugar derivatives maintain their bioactivity, then the
presence of a glycocompound will reduce fluorescent cell surface
staining by the labelled lectin, and these activities are routinely
compared to that of free mannose as the reference value. The
two approaches to cell surface ligand remodelling then enabled
to derive a relationship between the observed inhibitory activity
observed for a synthetic compound to changes in the profile
of N-glycans. More specifically, we shifted the glycomic profile
of one tumour cell system to high-mannose-type N-glycans by
applying the mannosidase I inhibitor 1-deoxymannojirimycin and
additionally tested the glycosylation mutant Lec8 of the Chinese
hamster ovary system. The latter presents high-affinity ligands
for ConA constituted by the N-acetylglucosamine (GlcNAc)-
terminated trimannoside core.6
Results and discussion
Synthesis
The various rigid and flexible compounds 1–5 for the presentation
of the mannose ligand were envisaged to be accessible via Ugi
reactions of glucuronic acid derivatives. As the first step, the
syntheses of required building blocks (14, 18, 19) were carried out.
4716 | Org. Biomol. Chem., 2009, 7, 4715–4725
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
In detail, the glycosidation of trichloroacetimidate 127 promoted
by boron trifluoride etherate in the presence of allyl alcohol
generated the allyl glycoside with b-configuration. Saponification
and subsequent treatment of the product with acetic anhydride
gave a 6,3-lactone derivative as an intermediate. This lactone was
converted to 1317–19 by reaction with allyl alcohol in presence of
molecular sieves and sodium acetate. Glycoside bond-forming
reaction of 13 with the Schmidt-Michel donor 15 8 gave the
expected a-glycoside product in good yield, and subsequent
selective hydrolysis of the allyl ester using Pd (0) catalysis9 resulted
in 14 (Scheme 1).
Scheme 2 Synthesis of 18.
Scheme 1 Synthesis of 14.
The a-glycoside 16, prepared as described previously,10 was
treated with sodium methoxide in MeOH, then acetylation
produced a 6,3-lactone derivative analogous to 12, except for its
a-configuration. Subsequent reaction of this lactone intermediate
with allyl alcohol in the presence of molecular sieves and sodium
acetate produced glycosyl acceptor 17. Glycosidation using the
trichloroacetimidate donor 15 and subsequent Pd (0)-catalysed
hydrolysis of the allyl ester as before gave 18 (Scheme 2). The
azide derivative 19 was prepared by literature procedure.5d
With these carboxylic acid derivatives in hand we were able to
generate divalent structures using Ugi reactions. This approach
to the bivalent tertiary anilides was superior to other routes
investigated. The less successful routes included amide coupling
reactions of 1,4-phenylenediamine with glucuronic acid derivatives
and efforts to subsequently alkylate the amides. In contrast, the
Ugi reactions of 14, 18 and 19 with formaldehyde, phenylene-
1,4-diamine and methyl isocyanoacetate proceeded efficiently to
give 20a–22b (Scheme 3). The N-butyl derivative 21b (structure
shown in Scheme 4) was obtained from the corresponding Ugi
reaction of 18 with n-butylisocyanide. Removal of ester-protecting
groups from 20a–22b proved difficult due to competing cleavage
of the tertiary anilides. The hydrolysis product 6a, for example,
was generated from attempts to convert 20a into 1a using a
variety of conditions, including Zemple´n deacetylation. The use
of triethylamine in the presence of water and MeOH or KCN–
Scheme 3 Synthesis of flexible glycoclusters 1a, 3a and 5a.
MeOH did ultimately give the bivalent mannosides 1a (18%), 3a
(25%) and 5a (8%) in low yields (Scheme 3).
With the mannose derivatives 20a and 21a/21b in hand macro-
cyclisation was effected using ring-closing metathesis reactions,
generating conformationally constrained macrocyclic structures
(Scheme 4). The reaction of 20a with the Grubbs-I catalyst (91%),
subsequent catalytic hydrogenolysis (89%) and de-O-acetylation
using KCN in MeOH (11%) gave 2a (not shown). As in the
synthesis of acyclic derivatives 1a and 3a (Scheme 3), it was
not possible to prevent competing cleavage of the tertiary anilide
intermediate, which was the major reason for the low yields of
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4715–4725 | 4717
©

View Article Online
TMSOTf as promoter as acceptor gave 24. The acetonide groups
were then removed by heating in 80% acetic acid and the depro-
tected intermediate then acetylated to obtain a per-O-acetylated
disaccharide, where the glucose residue was in the pyranose form.
Next, treatment of this per-O-acetate with azidotrimethylsilane
in the presence of a Lewis acid gave a b-azide intermediate and
subsequent catalytic hydrogenation of the azide gave 25 as a
mixture of anomers. Coupling of 25 using freshly recrystallised
terephthaloyl chloride proceeded in 84% yield to produce the
acetylated dimeric precursor. Subsequent de-O-acetylation using
Zemple´n conditions proceeded smoothly to give 7d; 26 was also
isolated as a by-product from this reaction sequence. This sequence
was overall more efficient than that described for preparation of
7c previously.5c
The synthesis (not shown) of 11 was achieved by glycosidation
of 15 with tetraethylene glycol as the nucleophile and subsequent
deprotection. Having completed the synthetic procedures, we next
calculated spatial aspects of ligand presentation.
Spatial parameters of ligand presentation
The spatial way that mannose residues are presented on the
macrocyclic derivatives 2a and 4a was inferred computationally. It
was expected that the mobility of mannose residues in 2a and
4a/4b would be significantly more restricted than in both 1a
and 3a. Cyclophane analogues derived from phenylenediamine
which incorporate allyl glycosides have been noted to be rigid.10a
The macrocyclic scaffold of 2a contains two internal b-linked
glucuronic acid residues, whereas that in 4a contains two internal
a-linked glycosides. The effect of these scaffold configurational
differences on the spatial parameters of mannose presentation was
explored by molecular modelling. Low-energy conformations of
the core scaffold structures were first generated using conforma-
tional searching techniques and energy minimization in Macro-
model. The mannose residues were then incorporated so that
glycosidic bond torsion angles defining the linkage between the
mannose residues and the glucuronic acid residue were consistent
with low-energy structures for related disaccharides, which had
been calculated using the OPLS-AA force field in Macromodel.11
One of the mannose residues of 2a was then superimposed with
one of the mannose residues of 4a, and the overlapped model
structures are shown in Fig. 2. The modeling shows that the
spatial relationship between the two mannose residues in 2a is
clearly distinct from that in 4a. The distance (Man C-1 to Man
Scheme 4 Synthesis of rigid macrocyclic glycoclusters 4.
the desired macrocyclic glycocyclophane derivatives. A similar
sequence from 21a and 21b gave 4a and 4b (Scheme 4), respectively.
The yield for the deprotection of the ring-closed product obtained
from metathesis of 21b, which was achieved with triethylamine in
MeOH and water, was higher than that observed for deprotection
of the ring-closed product obtained from 21a, which involved use
of KCN–MeOH.
The synthesis of terephthalamide derivative 7d, adding to
prior experience,5c was accomplished from diacetone-D-glucose
23 (Scheme 5). Glycosidation of 23 using 15 as donor and with
˚
C-1) between the mannose residues in 2a is 15.4 A, whereas the
˚
equivalent distance in 4a is 13.0 A. Two other parameters were
selected to define ligand spatial presentation: (a) the dihedral
angle defined by the glycosidic bonds to the two mannose residues
(C1–O1–O1–C1); (b) the dihedral angle defined by C4–C1–C1–C4
of the mannose residues. For 2a these parameters were 83◦ and
-66◦, respectively, whereas for 4a these correspondingly were -2◦
and -141◦. To next prove bioactivity of the prepared compounds
and to trace structure–activity correlations of biological relevance
cell-binding assays were systematically performed.
Cell assays
Biological responses are triggered by binding of lectins to cell
surface glycans. Obviously, this process needs to be impaired for
Scheme 5 Synthesis of 7d.
4718 | Org. Biomol. Chem., 2009, 7, 4715–4725
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Fig. 2 Models of 2a and 4a superimposed show the different spatial
presentation of the mannose ligands.
Fig. 3 Semilogarithmic representation of the fluorescent surface staining
of cells of the human colon adenocarcinoma line SW480 by 1 mg ml-1
labelled ConA. The control value of cell positivity by the second-step
reagent in the absence of lectin is given as shaded area, the 100%
value (lectin staining in the absence of inhibitor) as black line. Numbers
characterizing staining (percentage of positive cells/mean fluorescence
intensity) in each panel are always given in the order of the given listing.
A/B: Inhibition of lectin binding by increasing concentrations of mannose
(A: 1 mM, 2 mM, 4 mM, 10 mM, 20 mM) and glucose (B: 1 mM, 2 mM,
4 mM, 10 mM, 20 mM). C: Staining parameters in the presence of 2 mM
mannose and 2 mM glucose. D: Staining parameters in the presence of
0.5 mM sugar in compounds 6a, 1a, 9 and 11 compared to those at 2 mM
free mannose in (C). E: Staining parameters in the presence of 0.5 mM
sugar in compounds 2a, 5a, 4b and 4a compared to those at 2 mM free
mannose in (C). F: Staining parameters in the presence of 0.1 mM sugar
in compounds 6a, 2a and 1a compared to those at 0.2 mM free mannose.
triggering a therapeutic effect. Labelling of a lectin without activity
impairment generates a reliable tool to monitor its association
with cell surfaces. The extent of carbohydrate-dependent cell
binding is experimentally determined by fluorescent staining in
FACScan analysis, obtaining the percentage of positive cells and
signal intensity on a logarithmic scale as read-out, as presented
in Fig. 3–7. The binding of the labelled lectin will shift the
curve representing the staining profile of each lectin-exposed cell
population from the position of the control (given as shaded area
in each Figure), and presence of an effective inhibitor will move
the curve back into the direction of the control. In other words, the
two mentioned staining parameters in the graphs will be lowered, if
a glycocompound successfully competes with the lectin’s docking
sites on the cell surface. The result of this competition is best
evaluated at a subsaturating lectin concentration in the linear
range of concentration dependence, in this case at 1 mg ml-1. It
is mandatory to run the assay with aliquots of the same cell batch.
As essential controls for the probe’s activity, we first ascertained
the respective effects of the known haptenic monosaccharides
mannose and glucose at this ConA concentration on binding
to colon adenocarcinoma cells. Both sugars affected extent of
lectin binding (Fig. 3A,B). The direct comparison at 2 mM
inhibitor concentration revealed superiority of mannose relative
to glucose (Fig. 3C). The measured relative potencies of the two
monosaccharide inhibitors reflected their known activities in other
assays.6a Galactose, used as osmolarity control, as expected, did
not interfere with lectin activity (not shown). These experimental
series validated the assay so that the panel of test compounds could
be systematically examined.
ferences in binding affinities for synthetic ligands were therefore
noted in comparative analyses exemplarily illustrated in Fig. 3D,
E. The presentation of mannose in the context of the derivatives
increased the lectin-blocking capacity more than fivefold based
on the sugar concentration in the assay, with the strongest effects
seen already with the monomer 6a (Fig. 3F). Macrocycle 2a was
rather effective and surpassed the level of inhibition observed for
the mannose-bearing cyclophanes 4a and 4b (Fig. 3E). The acyclic
and more flexible derivative 1a was also active (Fig. 3D). In direct
comparison at a normalized sugar concentration of 0.1 mM, the
establishment of a rigid scaffold for sugar presentation consistently
resulted in an increase in inhibition of ConA binding compared
to the less rigid scaffold (2a more potent than 1a) (Fig. 3F).
Among the terephthalamides compound 9 gave the highest level
of blocking lectin binding, surpassing that of the increasingly
flexible bisubstituted tetraethylene glycol 11 (Fig. 3D) and other
terephthalamides 7 and 8. In order to next infer dependence on
the type of lectin a respective experimental series were investigated
with another leguminous lectin (Fig. 4). When using the closely
related lectin Pisum sativum agglutinin as a probe, the relative
In principle, derivatization and attachment of mannose to a
variety of different scaffolds did not impair the bioactivity of the
mannose residue. However the nature of the scaffold employed
did have an influence. Routinely, aliquots of cell batches were
studied with the whole panel of compounds to avoid any influence
of differences among individual cell preparations. Gradual dif-
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4715–4725 | 4719
©

View Article Online
compounds remained similar, with bivalent compound 8 still
proving slightly more active than free mannose (Fig. 5B).
Second, we tested a cell system from a different tumour type,
using the human pancreatic carcinoma line Capan-1. Commonly,
cell surface glycosylation changes among different tumour types.
The bioactivity profiles of the test compounds disclosed a notable
activity for terephthalamide derivative 9 and tetraethylene glycol
11 (Fig. 6). However, macrocyclic ring-closure led only to a slight
improvement of the activity of compound 2a compared to open
form 1a (Fig. 6A, B); this contrasted to that observed for the colon
adenocarcinoma cells, especially after increasing high-mannose-
type surface glycosylation. The open form 1a displayed a value
close to that observed for the mannose reference at 22%/10.5,
corroborating the concept for a dependence of relative compound
activity on cell-surface characteristics.
Fig. 4 Staining of cells of the human colon adenocarcinoma line SW480
by 0.2 mg ml-1 labelled Pisum sativum agglutinin (for further details,
please see legend to Fig. 3). A: Inhibition of lectin binding by increasing
concentrations of mannose (0.1 mM, 0.5 mM, 2 mM, 8 mM). B: Inhibition
of lectin binding by 0.5 mM sugar in compound 9, as free mannose, in
compound 4a and in compound 8.
potencies did indeed not remain constant. The monomer 6a was
less active than free mannose (not shown). Macrocycle 4a was
a comparatively stronger inhibitor than 2a and also its acyclic
analogue 3a, but only barely displaying the inhibitory efficiency
of free mannose (Fig. 4B). Terephthalamide-based scaffolds also
proved to be bioactive, with 9 being again more potent than 7 and
8, demonstrating the significant impact on linkage types (Fig. 4B).
Taking these data into account it thus does not appear to be valid
to extrapolate data within a lectin family. To next reveal whether
the inhibitory potency on ConA will depend on the glycan profile,
we performed three series of experiments.
First, we treated the colon cancer cells with mannosidase I
inhibitor 1-deoxymannojirimycin to shift glycan processing from
generating complex-type to high-mannose-type glycosylation and,
indeed, obtained the expected increase in ConA binding (Fig. 5A).
Thus, we could address the issue on the relationship of inhibitory
potency to a variation in glycan display and subsequently ex-
amined each compound’s activity. Overall, the respective mea-
surements quantitatively yielded a rather similar distribution of
activity when compared to assays carried out with the untreated
cells. Of note, macrocycle 2a showed increased activity with a
more than 2.5-fold difference relative to its open form 1a (please
note: plots given in Fig. 5B were recorded at 0.2 mM mannose
concentration for compound 2a vs. 0.5 mM for compound 1a).
This result intimates again the enhanced potential of a suitable
macrocyclic compound to block binding at increased cell surface
ligand density. Activity grading among terephthalamide-based
Fig. 6 Staining of cells of the human pancreatic carcinoma line Capan-1
by 1 mg ml-1 labelled ConA (for further details, please see legend to Fig. 3).
A: Inhibition of lectin binding by 0.5 mM sugar in compounds 2a and
9 (reference value for free mannose: 22%/10.5). B: Inhibition of lectin
binding by 0.5 mM sugar in compounds 6a, 11 and 1a compared to staining
in the absence of any inhibitor.
Besides altering the cell-type and density of ligands on a
cell surface, the affinity of a lectin for a cell type can also be
manipulated. A method to increasing ConA affinity for a cell
involves truncation of complex-type N-glycans so that terminal
GlcNAc moieties are found in the branches.6b,c We thus finally
tested a glycosylation mutant of Chinese hamster ovary (CHO)
cells with impaired galactosylation (Lec8). Whereas ligand density
had previously appeared to alter reactivity to rigid macrocycle 2a,
the presentation of high-affinity ligands on CHO cells led to no
significant shifts relative to that observed in assays with colon
cancer cells. As noted before, the monomeric derivative 6a was
active, and the inhibitory grading including the 2a/1a comparison
remained rather constant, as exemplarily illustrated in Fig. 7.
Terephthalamide 9 was again active. Thus for two different cell
types, increasing the density of mannose residues and presenting
Fig. 5 Staining of cells of the human colon adenocarcinoma line SW480
after treatment with 1-deoxymannojirimicin by 1 mg ml-1 labelled ConA
(for further details, please see the legend to Fig. 3). A: Staining parameters
in the absence (dashed) and in the presence of 1-deoxymannojirimycin
(black) as well as its inhibition by the presence of 0.5 mM, 2 mM and
10 mM mannose, respectively. B: Inhibition of lectin binding by 0.2 mM
sugar in compound 2a, 0.5 mM sugar in compound 1a, 0.2 mM sugar in
compound 6a and 0.5 mM sugar in compound 8 compared to the following
reference data when using free sugar: 62%/20 at 0.5 mM mannose and
42%/11.9 at 2 mM mannose.
Fig. 7 Staining of cells of the Chinese hamster ovary glycosylation mutant
line Lec8 by labelled ConA (for further details, please see legend to Fig. 3).
A: Staining with increasing concentrations of labelled ConA (0.2 mg ml-1,
0.5 mg ml-1, 1 mg ml-1 and 2 mg ml-1). B: Inhibition of lectin binding at a
concentration of 1 mg ml-1 by 0.5 mM sugar in compounds 2a, 9 and 1a
compared to 1 mM free mannose and the control in the absence of sugar.
4720 | Org. Biomol. Chem., 2009, 7, 4715–4725
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
high-affinity ligands by N-glycan truncation were tested and the
results revealed a consistent effect of both scaffold rigidity and
structure on inhibitory activity. Particular rigid structures based
on glycocyclophane scaffolds, as the cell based tests with the two
tested leguminous lectins reveal, may selectively improve blocking
of cell binding, as parallel assays with the compound pairs 1a/2a
and 3a/4a documented. Of practical value, the outlined routes
to tailor cell surface glycosylation are proposed to be of general
importance for testing glycan derivatives and glycoclusters.
were added and the mixture was stirred at room temp and under
N2 for 20 min. The mixture was then cooled on an ice bath and
0.04 M TMSOTf in dry dichloromethane (0.5 mL, 0.02 mmol) was
then added dropwise. Stirring was continued on ice for 30 min and
solid NaHCO3 (90 mg) was then added and the mixture stirred
for a further 5 min. Filtration was followed by removal of solvent
under diminished pressure. Chromatography (dichloromethane–
EtOAc, 5 : 1) of the residue gave the protected disaccharide as a
white solid (107 mg, 78%); [a]D = +13.51 (c 0.21, CHCl3); Rf 0.7
1
(dichloromethane–MeOH, 95 : 5); H-NMR (CDCl3, 300 MHz):
d = 6.01–5.83 (m, 2H, CH=), 5.43–5.09 (m, 9H, overlapping H-
4 GlcA, H-4 Man, CH2=, H-3 Man, H-2 Man and H-2 GlcA),
5.01 (s, 1H, H-1 Man), 4.74–4.62 (m, 2H, CO2CH2), 4.54 (d, J
7.5, 1H, H-1 GlcA), 4.42–4.06 (m, 5H, overlapping of OCH2,
H-6a Man, H-6b Man and H-5 Man), 4.01 (d, J 9.6, 1H, H-5
GlcA), 3.96–3.90 (t, J 9.0, 1H, H-3 GlcA), 2.19 (s, 3H, OCOCH3),
2.16 (br s, 9H, OCOCH3), 2.07, 2.01 (each s, 3H, OCOCH3); 13C-
NMR (CDCl3, 75 MHz): d = 170.6, 170.1, 169.7, 169.6, 169.5,
169.2, 168.8 (each CO), 133.2, 131.2 (each CH), 119.4, 117.7 (each
CH2), 99.5 (overlapping C-1 GlcA, C-1 Man), 80.5, 72.6, 72.1,
70.2 (each CH), 69.9 (OCH2), 69.6, 69.5, 68.8 (each CH), 66.7
(CO2CH2), 65.2 (CH), 61.8 (CH2), 20.9, 20.8, 20.7, 20.7, 20.6,
20.6 (each CH3); HRMS (ES): calcd 711.2112 [M+Na]+, found
711.2087; IR (film from dichloromethane): 3369, 3197, 2981, 1751,
1608, 1373, 1224 cm-1. To the protected disaccharide (107 mg,
0.155 mmol) in dry acetonitrile (1 mL) at 0 ◦C Pd(Ph3P)4 (18 mg,
0.015 mmol) followed by pyrrolidine (12 mL, 0.16 mmol) were
Conclusion
Our report addresses the question of potency in cell assays of
bivalent mannosides presented on a variety of scaffolds. We
specifically compared compounds based on both flexible and
rigid scaffolds and also compared differing spatial arrangement
of ligands presented on two subtly distinctive rigid macrocyclic
scaffolds. Toward this end, we reported the facile synthesis of two
types of phenylenediamine-based glycocyclophanes and improve-
ments to preparing derivatives based on the terephthalamide-
based scaffold. Bioactivity was tested in cell assays, revealing
consistent lectin-binding properties of the derivatives. Obviously,
macrocycles12 can serve as an effective ligand-presenting scaffold,
as also documented for tetraaza[6.1.6.1]paracyclophane or also
cyclic peptides as macrocyclic frameworks.13,14 Also, terephthala-
mide derivatives were shown to be active. Herein, we move beyond
common haemagglutination, precipitation or solid-phase data to
let the glycocompounds compete with cell surface glycans for
lectin binding. The presented results therefore extend the available
data on bivalent compounds and ConA binding, delineating
an impact of intra-mannosyl distance and linkage type, e.g.
obtained with diamine cores.15 The reduction of the inherent
level of flexibility appeared to have a favorable impact especially
for cells displaying a high ligand density, and the structure
of the macrocycle can modulate inhibitory potency in inter-
lectin comparison (ConA/PSA). In addition to conformational
restraint the three-dimensional presentation of the ligands, highly
dependent on the scaffold employed, could factor into differences
observed in the assay system.
Of general importance beyond the nature of this model, we
have set up a strategy for cell assays, which minimizes the risk
for invalid generalizations and reaches high sensitivity to probe
influence of the nature of glycomic profiles. It consists of two
different lines and manipulations of the glycomic profile by
recruiting an inhibitor for glycan processing and a glycosylation
mutant. Because interference with cell binding is a medically
relevant parameter and tissue lectins can target distinct ligands
with context-dependent glycosylation, e.g. as recognition signals in
adhesion, induction of apoptosis or glycoprotein routing,2,4,16 this
approach is proposed to find applications in studies on endogenous
lectins.
◦
added under N2 and the mixture was stirred at 0 C for 25 min
and then filtered through celite. The solvent was removed under
diminished pressure and the yellow residue was dissolved in EtOAc
and washed with water. The aq phase was acidified to pH = 4 with
Amberlite-H+ resin and then filtered. The filtrate was extracted
with EtOAc (4 ¥ 10 mL) and the combined organic extracts were
dried (MgSO4), filtered and the solvent was removed to give 14 as
a pale yellow foam (65 mg, 65%); [a]D = +1.69 (c 0.53, CHCl3);
1H-NMR (CDCl3, 300 MHz): d = 5.90–5.78 (m, 1H, CH=), 5.42
(br s, 1H, CO2H), 5.36–5.11 (m, 6H, overlapping H-4 GlcA, H-4
Man, CH2=, H-3 Man, H-2 Man), 5.09–5.03 (t, J 7.5, 1H, H-
2 GlcA), 4.98 (d, J 1.5, 1H, H-1 Man), 4.56 (d, J 6.9, 1H, H-1
GlcA), 4.41–4.35 (dd, J 4.8, 13.2, 1H, H-6a Man), 4.26–4.21 (dd,
J 3.6, -12.6, 1H, H-6b Man), 4.16–4.01 (m and d overlapping,
4H, J 9, H-5 GlcA, OCH2, H-5 Man), 3.96–3.90 (t, J 8.7, 1H,
H-3 GlcA), 2.15, 2.14 (each s, 3H, OCOCH3), 2.11 (br s, 6H,
OCOCH3), 2.03, 1.97 (each s, 3H, OCOCH3); 13C-NMR (CDCl3,
75 MHz): d = 170.9, 170.1, 169.8, 169.3, 168.8 (each CO), 133.2
(CH), 117.8 (CH2), 99.3 (C-1 GlcA and C-1 Man), 78.8 (CH),
72.0 (2 CH), 70.1 (OCH2), 69.9 (CH), 69.5 (2 CH), 68.8, 65.3 (each
CH), 61.9 (C-6 Man), 20.9, 20.8, 20.8, 20.7, 20.6, (each OCOCH3);
HRMS (ES): calcd 647.1822 [M - H]-, found 647.1823; IR (film
from dichloromethane): 3623, 3423, 2943, 1749, 1645, 1431, 1373,
1225 cm-1.
1,4-Di-[(N-(1-methoxycarbonyl)methylamino-2-oxoethyl)-2,3,4,6-
tetra-O-acetyl-a-D-mannopyranosyl-(1→3)-2,4-di-O-acetyl-1-O-
allyl-b-D-glucopyranuronamido]benzene 20a
Experimental
Allyl 2,3,4,6-tetra-O-acetyl-a-D-mannopyranosyl-(1→3)-2,4-di-O-
acetyl-1-O-allyl-b-D-glucopyranosiduronate 14
Formaldehyde (14 mL from a 37% solution in water, 0.165 mmol)
was added to acid 14 (107 mg, 0.165 mmol) and 1,4-
phenylenediamine (8.2 mg, 0.076 mmol) in MeOH (1.5 mL)
Ester 13 20 (74 mg, 0.21 mmol) and 15 (112 mg, 0.22 mmol) were
dissolved in dry dichloromethane (5 mL) and molecular sieves
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4715–4725 | 4721
©

View Article Online
at room temp and the mixture was stirred for 1 h and then
methyl isocyanoacetate (15 mL, 0.17 mmol) was added and stirring
continued for a further 24 h. The solvent was evaporated to give
a dark residual solid. Chromatography of the residue (EtOAc,
then dichloromethane–MeOH 98 : 2) gave 20a as an off-white
solid (76 mg, 62%); [a]D = +13.3 (c 0.24, CHCl3); Rf 0.28
1,4-Di-[(N-(1-methoxycarbonyl)methylamino-2-oxoethyl)-a-D-
mannopyranosyl-(1→3)-1-O-allyl-a-D-glucopyranuronamido]-
benzene 3a
Deprotection of 21a (7 mg, 4.3 mmol) with KCN–MeOH as de-
scribed for 20a gave 3a as a white solid after lyophilization (1.2 mg,
1
25%); H-NMR (D2O, 300 MHz): d = 7.61 (s, 4H, aromatic H),
1
(dichloromethane–MeOH, 95 : 5); H-NMR (CDCl3, 300 MHz):
5.78–5.70 (m, 2H, CH=), 5.26–5.17 (m, 6H, overlapping of CH2=,
H-1 Man), 4.96–4.95 (d, J 3.5, 2H, H-1 GlcA), 4.79–4.76 (m,
NCH2CONH), 4.55–4.53 (d, J 10, 2H, NCH2CONH), 4.31–4.29
(d, J 10, 2H, H-5 GlcA), 4.06–4.05 (m, 2H, H-2 Man), 3.97–3.92
(m, 4H, overlapping H-2 GlcA, H-5 Man), 3.88–3.74 (m, 10H,
overlapping of H-3 Man, NHCH2CO2, H-6a Man, H-6b Man),
3.73–3.64 (m, 2H, OCH2), 3.59–3.54 (m, 2H, OCH2) overlapping
with 3.56–3.52 (t, J 9, 2H, H-4 GlcA), 3.47–3.42 (m, 2H, H-
3 GlcA), 3.31–3.27 (t, J 9, 2H, H-4 Man); LRMS (ES): 546.3
[M - 2H]2-.
d = 7.56 (s, 4H, aromatic H), 6.89–6.85 (t, J 5.4, 2H, NH), 5.82–
5.71 (m, 2H, CH=), 5.47–5.42 (t, J 9.3, 2H, H-4 GlcA), 5.33–
5.27 (t, J 10.2, 2H, H-4 Man), 5.23–5.08 (m, 8H, overlapping
of CH2=, H-3 Man, H-1 Man), 4.89–4.84 (m, 4H, overlapping
of H-2 Man, H-2 GlcA), 4.53–4.48 (d, J 15.6, 2H, CH2CONH),
4.28–4.14 (m, 8H, overlapping H-1 GlcA, CH2CONH, H-6a Man,
OCH2), 4.07–3.86 (m, 10H, overlapping H-5 GlcA, NHCH2CO2,
H-5 Man, H-6b Man and OCH2), 3.77–3.74 (m, 2H, H-3 GlcA),
3.74 (s, 6H, CO2CH3), 2.15, 2.12, 2.09, 2.04, 2.02, 1.96 (each s, 3H,
OCOCH3); 13C-NMR (CDCl3, 125 MHz): d = 171.3, 170.8, 170.2
(2 s), 169.8, 169.7, 169.4, 168.2, 166.6 (each CO), 142.2 (aromatic
C), 133.4 (CH), 129.4 (aromatic CH), 117.7 (CH2), 100.0, 99.7,
80.5, 77.4, 71.9, 71.0 (each CH), 70.1 (CH2), 69.8, 69.7, 69.0, 65.6
(each CH), 62.0 (C-6 Man), 54.6 (NCH2CONH), 52.4 (CO2CH3),
41.3 (NHCH2CO2CH3), 21.2, 21.1, 21.0 (2 s), 20.9, 20.8 (each
CH3); HRMS (ES): calcd 1649.5029 [M+Na]+, found 1649.4958;
IR (film from dichloromethane): 3631–3374, 2966, 1751, 1670,
1510, 1438, 1373, 1223, 1040 cm-1.
1,4-Di-[(N-(1-methoxycarbonyl)methylamino-2-oxoethyl)-a-D-
mannopyranosyl-(1→3)-1-azido-1-deoxy-b-D-
glucopyranuronamido]benzene 5a
The protected dimer 22 (24 mg, 0.015 mmol) was suspended in
water–MeOH-NEt3: (1 : 1 : 8, 5 mL) and the mixture stirred for
24 h at room temp and for a further 24 h at 50 ◦C and the solvent
was then removed. Semi-preparative HPLC of the residue (16 mg,
gradient elution, 0.1% aq TFA-CH3CN, 97 : 3–50 : 50) gave 5a as
a white solid (1.3 mg, 8%); 1H-NMR: (500 MHz, D2O; EE:EZ,21
10 : 1): peaks for the EE isomer: d 7.59 (s, 4H, Ar-H), 5.16 (d, J
1.2, 2H, H-1 Man), 4.57 (s, 4H, NCH2CONHCH2), 4.51 (d, J 8.8,
2H, H-1 GlcA), 4.13 (d, J 9.7, 2H, H-5 GlcA), 4.03 (dd, J 3.2, 1.7,
2H, H-2 Man), 3.95 (t, J 9.4, 2H, H-4 GlcA), 3.91 (ddd, J 10.0,
4.8, 2.4, 2H, H-5 Man), 3.85 (s, 4H, NCH2CONHCH2), 3.85–3.73
(m, 6H), 3.80 (dd, J = 12.3, 2.3, 2H), 3.75 (dd, J = 12.3, 4.9, 2H),
3.70 (t, J = 9.9, 2H), 3.57 (t, J = 9.1, 2H, H-3 GlcA), 3.33 (t, J =
9.0, 2H, H-2 GlcA); LRMS (ES): 1063.0 [M - H]-, 531.0 [M -
2H]2-; HRMS (ES): calcd 531.1449 [M - 2H]2-, found 531.1468.
Selected 1H-NMR data for the EZ isomer: d 7.54 (d, J = 8.64 Hz,
2H), 7.45 (d, J = 8.7 Hz, 2H), 5.28 (s, 1H), 4.95 (d, J = 8.4 Hz,
1H), 3.44 (t, J = 8.8 Hz, 1H).
1,4-Di-[(N-(1-methoxycarbonyl)methylamino-2-oxoethyl)-a-D-
mannopyranosyl-(1→3)-1-O-allyl-b-D-glucopyranuronamido]-
benzene 1a
To a flask containing potassium cyanide (5.4 mg, 0.083 mmol),
which had been pre-dried under high vacuum, a solution of 20
(23 mg, 0.014 mmol) in dry MeOH (3 mL) was added under N2 and
at room temp and stirring was continued for 20 h. The solvent was
removed under diminished pressure and semi-preparative HPLC
of the residue (isocratic elution, H2O–CH3CN, 96 : 4) gave 1a as
1
a white solid after lyophilization (2.8 mg, 18%); H-NMR (D2O,
300 MHz): d = 7.61 (s, 4H, aromatic H), 5.98–5.92 (m, 2H, CH=),
5.39–5.29 (m, 4H, CH2=), 5.18 (s, 2H, H-1 Man), 4.93–4.86 (d, J
20.1, 2H, NCH2CONH), 4.72–4.65 (d, J 20.1, 2H, NCH2CONH),
4.58–4.55 (d, J 8.1, 2H, H-1 GlcA), 4.31–4.20 (m, 2H, OCH2),
4.22–4.20 (d, J 8.4, 2H, H-5 GlcA), 4.11–4.09 (m, 2H, OCH2),
4.05 (m, 2H, H-2 Man), 4.01–3.91 (m, 4H, overlapping of H-2
GlcA, H-5 Man), 3.87 (s, 4H, NHCH2CO2), 3.85–3.76 (m, 6H,
overlapping of H-3 Man, H-6a Man, H-6b Man), 3.74–3.67 (t, J
9.6, 2H, H-4 GlcA), 3.60–3.52 (m, 2H, H-3 GlcA), 3.39–3.33 (t,
J 8.1, 2H, H-4 Man); HRMS (ES): [M - 2H]2-calcd 546.1697,
found 546.1805, LRMS (ES): 1093.5 [M - H]-. The hydrolysis by-
product 6a was also obtained by HPLC (2.6 mg, 25%); 1H-NMR
(D2O, 300 MHz): d = 7.27–7.24 (d, J 8.7, 2H, aromatic H), 6.79–
6.76 (d, J 8.7, 2H, aromatic H), 6.06–5.87 (m, 1H, CH=), 5.38–
5.29 (m, 2H, CH2=), 5.18 (s, 1H, H-1 Man), 4.72 (s, overlapping
with HOD signal, NCH2CONH), 4.49 (s, 2H, NHCH2CONH),
4.32–4.26 (m, 1H, OCH2), 4.18–4.15 (d, J 8.1, 1H, H-1 GlcA),
4.10–4.05 (m, 2H, overlapping OCH2, H-2 Man), 3.97–3.77 (m,
6H, overlapping of H-5 GlcA, H-2 GlcA, H-5 Man, H-3 Man,
H-6a Man, H-6b Man), 3.85 (s, 4H, NHCH2CO2), 3.74–3.67 (t, J
9.6, 1H, H-4 GlcA), 3.50–3.56 (m, 1H, H-3 GlcA), 3.38–3.33 (t, J
8.1, 1H, H-4 Man); HRMS (ES): calcd 715.2310, found 715.2329,
[M - H].-
Glycocyclophane 2a
A degassed solution of the diamide 20a (47 mg, 0. 029 mmol)
in dry dichloromethane (17 mL, 1.7 mM) and under N2 was
treated with Grubbs-I catalyst (5.2 mg, ~10%) for 60 h. The
solvent was evaporated under diminished pressure and gave a
black residue. Chromatography (dichloromethane–MeOH 97 : 3)
gave the macrocyclic product as an inseparable 2 : 3 mixture of cis
and trans isomers (42 mg, 91%); Rf 0.15(dichloromethane–MeOH,
1
95 : 5); H-NMR (CDCl3, 300 MHz): d = 7.61 (s, 4 H, aromatic
H), 7.56 (s, 4 H, aromatic H), 6.85–6.82 (m, 2 H, NH), 5.74–5.70
(t, 2H, J 5.7, 2H, alkene CH) and 5.64 (br s, 2H, alkene CH], 5.46–
5.29 (m, 4H, overlapping of H-4 GlcA and H-4 Man), 5.18–5.16
(dd J 2.7, 6, 2H, H-3Man), 5.13 (bs, 4H, overlapping of H-1 Man
and H-2 Man), 4.89–4.81 (br m, 2H, H-2 GlcA), 4.78–4.72 (m, 2H,
OCH2), 4.32 (br s, 4H, overlapping of H-6 Man), 4.22–3.87 (m,
18H, overlapping of H-1 GlcA, NCH2CO, OCH2, NHCH2CO,
H-5 GlcA, H-5 Man, H-3 GlcA), 3.75 (s, 6H, CO2CH3), 2.20,
2.18, 2.11, 2.08, 2.05, 2.01 (each s, 3H, CO2CH3), 2.19, 2.17,
4722 | Org. Biomol. Chem., 2009, 7, 4715–4725
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
2.12, 2.09, 2.03, 1.96 (each s, 3H, CO2CH3); 13C-NMR (CDCl3,
125 MHz): d = 170.5, 170.4, 170.0, 169.6 (2 s), 169.5, 169.1, 167.8,
166.5, 166.4 (each CO), 141.9 (aromatic C), 131.0 (CH=), 129.0,
128.6 (each aromatic CH), 100.0, 99.7, 99.4, 92.2 (C-1 GlcA and
Man), 80.4, 72.4, 71.7, 71.5, 71.2, 69.9, 69.5, 69.4, 69.0, 68.9, 68.7,
68.6, 67.7, 66.2, 65.3, 65.2 (C-2 to C-5 GlcA and Man), 63.9,
62.6 (each OCH2), 61.6, 61.5 (each C-6 Man), 55.2 (NCH2CO),
52.3 (CO2CH3), 41.1 (NHCH2CO2), 20.9 (2 s), 20.8 (2 s), 20.7 (2
s), 20.6 (each COCH3); HRMS (ES): [M+Na]+ calcd 1621.4716
found 1621.4768; IR (film from dichloromethane): 3622–3401,
2963, 1750, 1660, 1373, 1225, 1041 cm-1. To this mixture (40 mg,
0.014 mmol) in EtOAc (2 mL), 10% Pd–C (6 mg) was added and
the reaction mixture was stirred at room temp under H2. The
mixture was then filtered through celite and the solvent removed
under diminished pressure. Chromatography (dichloromethane–
MeOH, 97 : 3) gave the protected glycophane (37 mg, 92%); Rf 0.17
(dichloromethane–MeOH, 95 : 5); [a]D = +38.7 (c 0.23, CHCl3);
1H-NMR (CDCl3, 300 MHz): d = 7.63 (br s, 4 H, aromatic H), 6.82
(br s, 2 H, NH), 5.40–5.32 (overlapping signals of H-4 GlcA and
H-4 Man) 5.22–5.15 (4H, overlapping signals of H-3 Man and H-1
Man), 4.95 (s, 2H, H-2 Man), 4.31 (br s, 4H, NCH2CO), 4.16–4.13
(dd, J 3.7, 5.5, 2H, H-2 GlcA), 4.09–3.99 (br m, 16H, overlapping
signals of H-6a Man, H-6b Man, NHCH2CO, H-1 GlcA, H-5 Man,
H-5 GlcA, H-3 GlcA), 3.84–3.75 (m, 2H, OCH2), 3.77 (s, 6H,
CO2CH3), 3.62–3.59 (m, 2H, OCH2), 2.20, 2.14, 2.10, 2.06, 2.04,
1.98 (each s, 3H, CO2CH3), 1.62 (m. 4H, CH2); 13C-NMR (CDCl3,
75 MHz): d = 170.6, 170.1 (2 s), 169.9, 169.7, 169.0, 167.8, 166.4
(each CO), 142.6 (broad signal, aromatic C), 130.2, 129.8 (broad
signals, each aromatic CH), 100.8, 99.9 (C-1 GlcA and Man), 80.9,
77.7, 69.6, 69.57, 69.5, 69.4, 69.0, 65.2 (2 s) (C-2-C-5 GlcA and
Man, OCH2), 61.5 (C-6 Man), 52.3 (NCH2CO), 52.3 (CO2CH3),
41.1 (NHCH2CO2), 20.9, 20.8 (2 s), 20.7 (2 s), 20.6 (each COCH3);
IR (film from dichloromethane): 3613–3405, 2959, 1748, 1655,
1509, 1437, 1373, 1259, 1223, 1039 cm-1. Deprotection of this
intermediate (32 mg, 0.02 mmol) with KCN and MeOH–water
(1 : 1) as described above for 1a and subsequent semi-preparative
HPLC (isocratic elution, water–CH3CN, 99 : 1) and lyophilisation
gave 2a (2.3 mg, 11%); 1H-NMR (D2O, 500 MHz): d = 7.67 (s, 4H,
aromatic H), 5.26 (s, 2H, H-1 Man), 4.80 (s, 2H, NCH2CONH),
4.65 (s, 2H, NCH2CONH), 4.44–4.43 (d, J 8, 2H, H-1 GlcA),
4.12–4.05 (m, 6H, overlapping of H-2 GlcA, OCH2, H-2 Man),
4.02–3.99 (m, 4H, overlapping H-5 GlcA, OCH2), 3.96 (s, 4H,
NHCH2CO2) 3.94–3.91 (dd, J 3, J 10, 2H, H-6 Man) overlapping
with 3.90–3.92 (m, 2H, H-5 Man), 3.88 (d, J 2, 2H, H-3 Man),
3.85–3.81 (dd, J 5, 12.5, 2H, H-6 Man), 3.79–3.75 (t, J 10, 2H,
H-4 GlcA), 3.64–3.60 (t, J 8.5, 2H, H-3 GlcA), 3.36–3.34 (t, J 8,
2H, H-4 Man), 1.90–1.83 (m, 4H, CH2 CH2); HRMS (ES): calcd
533.1619 [M - 2H]2-, found 533.1617.
d = 7.49 (s, 4H, aromatic H), 6.59–6.57 (t, J 5.0, 2H, NH), 5.63–
5.59 (t, J 9.5, 2H, H-4 GlcA), 5.30–5.26 (t, J 10, 2H, H-4 Man),
5.19–5.17 (dd, J 3, 10, 2H, H-3 Man), 5.10 (s, 2H, H-2 Man),
5.02 (s, 2H, H-1 Man), 5.01–5.00 (d, J 3.5, 2H, H-1 GlcA), 4.89–
4.86 (dd, J 3.0, 9.5, 2H, H-2 GlcA), 4.67–4.65 (d, J 10, 2H, H-5
GlcA), 4.50–4.47 (d, J -15.5, 2H, NCH2CONH), 4.22–4.19 (d,
J -15.5, 2H, NCH2CONH), 4.18–4.10 (m, 8H, overlapping of
H-5 Man, H-6a Man, H6b Man, and H-3 GlcA), 4.03–4.01 (t,
J 9, 4H, NHCH2CO), 3.77 (s, 6H, CO2CH3), 3.42–3.39 (br t, J
-7.5, 2H, OCH2), 3.31–3.29 (m 2H, OCH2), 2.26, 2.14, 2.08, 2.05,
2.03, 1.98 (each s, 3H, OCOCH3), 1.52–1.50 (m 2H, OCH2CH2),
1.00–0.98 (m 2H, OCH2CH2). 13C-NMR (CDCl3, 125 MHz): d =
170.6, 170.0, 169.9, 170.1, 169.8, 169.7, 169.5, 169.1, 167.5, 164.5
(each CO), 141.8 (aromatic C), 129.0 (aromatic CH), 98.2 (C-1
Man), 95.7 (C-1 GlcA), 76.3, 72.1, 71.2, 69.8, 69.4, 68.7 (each
CH), 66.5 (OCH2), 65.7, 65.3 (each CH), 62.2 (C-6 Man), 55.1
(NCH2CO), 52.4 (CO2CH3), 41.2 (NHCH2CO2), 26.4 (t, CH2),
20.9 (2 s), 20.8, 20.7, 20.6 (each OCOCH3); HRMS (ES): calcd
1623.4873 [M+Na]+, found 1623.4906. Deprotection using KCN–
MeOH as described for 1a gave the title compound 4a after semi-
preparative HPLC (isocratic elution, 0.1% aq TFA-CH3CN) and
lyophilization (1.8 mg, 15%); 1H-NMR (D2O, 500 MHz): d = 7.69
(s, 4H, aromatic H), 5.33 (s, 2H, H-1 Man), 4.96 (d, J 3.0, 2H, H-1
GlcA), 4.50–4.46 (m, NCH2CONH), 4.22 (m, 2H, NHCH2CO2),
4.19–4.10 (4H, overlapping OCH2, H-5 GlcA), 4.04–4.01 (m, 2H,
H-2 Man), 3.98–3.91 (m, 6H, overlapping H-5 Man, H-6 Man, H-
2 GlcA), 3.87–3.80 (m, 6H, overlapping OCH2, H-3 Man), 3.78–
3.74 (m, 4H, H-6b Man, H-4 GlcA), 3.50–3.46 (m, 4H, overlapping
H-3 GlcA, H-4 Man), 1.61 (br m, 2H, CH2 CH2), 1.02 (br m,
2H, CH2CH2); HRMS (ES): calcd 533.1619 [M - 2H]2-, found
533.1627.
Glycocyclophane 4b
Ring closing metathesis of 21b (97 mg, 0. 06 mmol) as described for
20a gave a mixture of cis and trans isomers of the macrocyclic prod-
uct as an off-white solid (67 mg 71%); Rf 0.3 (dichloromethane–
EtOAc 1 : 4); 13C-NMR (CDCl3, 75 MHz): d = 170.6, 169.9, 169.8,
169.6, 169.5, 169.4, 167.2 (each CO), 141.9 (aromatic C), 128.4
(aromatic CH), 125.4 (alkene CH), 98.1, 96.8, 94.1 (C-1 GlcA and
Man), 75.5, 71.9, 71.6, 69.7, 69.2, 68.6, 66.1, 65.8, (C-2-C-5 GlcA
and Man), 65.0, (OCH2), 62.1 (C-6 Man), 55.5 (NCH2CO), 39.4
(NHCH2), 31.4 (CH2CH2), 20.9, 20.8, 20.7, 20.6 (each COCH3),
20.0 (CH2CH3), 13.7 (CH2CH3). Catalytic hydrogenation of this
mixture (45 mg, 0.028 mmol) using 10% Pd–C gave the reduced
protected glycocyclophane intermediate as a white solid (44 mg,
95%); Rf 0.32 (dichloromethane–EtOAc, 1 : 4); [a]D = -18.2 (c
1
0.11, CHCl3); H-NMR (CDCl3, 600 MHz): d = 7.47 (s, 4H,
aromatic H), 6.11–6.09 (t, J 6, 2H, NH), 5.63–5.60 (t, J 9.6,
2H, H-4 GlcA), 5.30–5.27 (t, J 10.2, 2H, H-4 Man), 5.19–5.17
(dd, J 3.6, 10.2, 2H, H-3 Man), 5.10–5.09 (t, J 2.4, 2H, H-2
Man), 5.02 (d, J 1.8, 2H, H-1 Man), 4.99–4.98 (d, J 3.6, 2H,
H-1 GlcA), 4.89–4.87 (dd, J 3.6, 10.2, 2H, H-2 GlcA), 4.67–
4.65 (d, J 9.6, 2H, H-5 GlcA), 4.30–4.27 (d, J-15, 2H, NCH2CO),
4.19–4.17 (m, 4H, overlapping NCH2CO, H-6a Man), 4,14–4.09
(m, 6H, overlapping H-6 Man, H-5 Man, H-3 GlcA), 3.39–3.38
(t, J 7.8, 2H, OCH2), 3.28–3.19 (m, 6H, overlapping OCH2,
NCH2), 2.26, 2.14, 2.08, 2.05, 2.03, 1.98 (each s, 3H, OCOCH3),
1.51–1.46 (m, 6H, overlapping OCH2CH2, CH2CH2), 1.37–1.33
Glycocyclophane 4a
Ring closing metathesis of the diamide 21a (45 mg, 0. 027 mmol)
as described for 20a gave a mixture of cis and trans isomers of the
macrocyclic product, which was an off-white solid (32 mg, 74%);
Rf 0.42 (dichloromethane–MeOH, 90 : 10); HRMS (ES): calcd
1621.4716 [M+Na]+ found 1621.4757. Catalytic hydrogenation
of this mixture using 10% Pd–C gave the protected glycophane
as a white solid (24 mg, 88%); Rf 0.2 (dichloromethane–MeOH,
95 : 5); [a]D = -17.1 (c 0.18, CHCl3); 1H-NMR (CDCl3, 500 MHz):
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4715–4725 | 4723
©

View Article Online
(q, J 7.8, 4H, CH2CH3), 1.01–0.97 (m, 2H, OCH2CH2), 0.95–
0.91 (m, 6H, CH2CH3); 13C-NMR (CDCl3, 150 MHz): d = 170.6,
170.0, 169.8, 169.7, 169.5, 168.9, 167.4, 167.3 (each CO), 141.9
(aromatic C), 128.8 (aromatic CH), 98.2 (C-1 Man), 95.6 (C-1
GlcA), 76.3, 72.2, 71.2, 69.8, 69.4, 68.8 (each CH), 66.4 (OCH2),
65.8, 65.2 (each CH), 62.2 (C-6 Man), 55.6 (NCH2CO), 39.5
(NHCH2), 31.4 (CH2), 26.4 (OCH2CH2), 20.9, 20.8 (2 s), 20.7, 20.6
(each OCOCH3), 20.0 (CH2CH3), 13.7 (CH2CH3); HRMS (ES):
calcd 1591.5804 [M+Na]+, found 1591.5857. A solution of this
protected glycophane (44 mg, 0.028 mmol) in a mixture of MeOH–
water–NEt3 (8 : 1 : 1, 2 mL) was stirred 50 ◦C for 30 min and then
allowed to attain room temperature and stirring was continued for
another 40 h. The solvent was evaporated under reduced pressure
and subsequent semi-preparative HPLC (gradient elution: H2O–
CH3CN, 99 : 1 to 96 : 4) and lyophilisation gave 4b as a white solid
after lyophilization (8.7 mg, 30%); 1H-NMR (D2O, 500 MHz): d =
7.62 (s, 4H, aromatic H), 5.27 (s, 2H, H-1 Man), 4.90–4.88 (d, J
3. 2H, H-1 GlcA), 4.73–4.64 (m, 2H, NCH2CONH), overlapping
4.64–4.61 (d, J 9.6, 2H, H-5 GlcA), 4.24–4.18 (d, J 15.9, 2H,
NCH2CONH), 4.10–4.04 (m, 4H, overlapping H-2 Man, OCH2),
3.99–3.60 (m, 14H, overlapping OCH2, H-2 GlcA, H-5 Man,
H-3 Man, H-6a Man, H-6b Man, H-4 GlcA) 3.44–3.36 (m, 4H,
overlapping H-3 GlcA, H-4 Man), 3.17–3.12 (m, 4H, NHCH2),
1.56–1.50 (m, 2H, CH2 CH2), 1.40–1.34 (m, 4H, CH2CH2), 1.25–
1.18 (m, 4H, CH2CH3), 0.98–0.90 (m, 2H, CH2 CH2), 0.87–0.83
(t, J 7.5, 6H, CH2CH3); HRMS (ES): [M - H]- calcd 1063.4472,
found 1063.4431: calcd 1087.4434 [M+Na]+, found 1087.4475.
gave a white powder (37 mg, 92%) which was as a mixture of 7d
and 26, 85 : 15). Semi-preparative HPLC (isocratic elution, H2O–
CH3CN, 99 : 1) gave 7d; 1H NMR (600 MHz, D2O): d 7.93 (s, 4H,
aromatic H), 5.29 (d, J 1.2, 2H, H-1 Man), 5.23 (d, J 9.3, 2H, H-1
Glc), 4.08 (dd, J 3.2, 1.8, 2H, H-2 Man), 3.99 (ddd, J 9.9, 5.0,
2.3, 2H, H-5 Man), 3.93–3.83 (m, 6H, overlapping H-6b Glc H-3
Man, H-6b Man), 3.82–3.76 (m, 6H, overlapping H-6a Man, H-3
Glc, H-6a Glc), 3.74 (t, J 9.9, 2H, H-4 Man), 3.66–3.62 (m, 6H,
overlapping H-2 Glc, H-4 (Glc), H-5 Glc); 13C-NMR (150 MHz,
D2O): d 171.2 (NHCO), 136.8 (C), 128.1 (aromatic CH), 101.3
(C-1 Man) 83.0 (C-1 Glc), 77.7, 73.0, 70.8, 70.7, 70.5, 70.0, 66.8
(each CH), 61.0, 60.6 (each CH2); HRMS (ES): calcd 835.2596
[M+Na]+, found 835.2608.
Molecular modelling
All calculations were carried out using the GB/SA solvation
model22 for water and the OPLS-AA force field23 in Macromodel
8.5. Low energy structures were obtained for both glycophane core
macrocycle structures of 2a and 4a (not containing the mannose
residue by a conformational searching structures sampled) with
using the SUMM method. For the purpose of this conformational
search N-methyl groups were incorporated instead of the
N-(1-methoxycarbonyl)methylamino-2-oxoethyl)
substituents.
The structures corresponding to the global minima were
subsequently used for the model building. The favoured angles
for the dihedrals U and W for the glycosidic linkage between the
mannose and glucose/glucuronic acid residues were calculated by
systematic exploration of U and W space using the dihedral drive
method and also by SUMM based conformational searching
techniques, of model disaccharides24 with a(1→3) linkages.
The glycosidic torsion U is defined as H1–C1–O3–C3 and W is
defined as C1–O3–C3–H3. and the torsion angles observed for
the lowest energy conformers were in good agreement with those
previously calculated for related disaccharides.25 See also the
supporting information section of reference 5c. These parameters
were then used to incorporate the mannose residues on to the
macrocyclic scaffold using the build function in Macromodel.
The N-(1-methoxycarbonyl)methylamino-2-oxoethyl substituent
was subsequently incorporated instead of the N-methyl group to
generate, after a final energy minimisation and superimposition,
the structures shown in Fig. 2.
N,N¢-Di-(a-D-mannopyranosyl-(1→3)-b-D-glucopyranosyl)-
terephthalamide 7b.
A solution of amine 25 (75.8 mg,
0.1193 mmol) and DIPEA (26.4 mL, 0.1551 mmol) in THF
(anhydrous, 2 mL) was added dropwise at room temp to fresh
terephthaloyl chloride (12 mg, 0.059 mmol) in THF (anhydrous,
2 mL). After 1 h the reaction was complete by TLC after which
the solvent was removed. Chromatography (EtOAc–cyclohexane
gradient elution, 1 : 1 to 1.5 : 1 and EtOAc–CH2Cl2, 9 : 1) gave
the protected terephthalamide dimer as a white amorphous solid
(70 mg, 84%); Rf 0.5, (EtOAc: CH2Cl2, 9 : 1); [a]D +8.0 (c 0.15,
CHCl3); 1H NMR (600 MHz, CDCl3, mixture of anomers, bb : ab,
83 : 17) peaks for the major bb anomer: d 7.83 (s, 4H, aromatic H)
7.06 (d, J 9.0, 2H, NH), 5.33 (t, J 9.3, 2H, H-1 Glc), 5.30 (t, J 10.2,
2H, H-4 Man), 5.21 (t, J 9.7, 2H, H-4 Glc), 5.16 (dd, J 10.1, 3.3,
2H, H-3 Man), 5.11 (dd, J = 3.2, 2.2, 2H, H-2 Man), 5.04–5.00
(m, 4H, overlapping H-1 Man, H-2 Glc), 4.30 (dd, J = 12.6, 4.4,
2H, H-6b Glc), 4.19 (dd, J = 12.4, 4.6, 2H, H-6a Man), 4.14–4.04
(m, 6H, overlapping H-6b Man, H-6a Glc, H-5 Man), 4.04 (t, J =
9.4, 2H, H-3 Glc), 3.80 (ddd, J 10.1, 4.3, 2.3, 2H, H-5 Glc), 2.17,
2.16, 2.12, 2.09, 2.09, 2.04, 1.98 (each s, 6H, COCH3); selected
NMR data for the ab anomer: d 7.92 (d, J 8.37, 2H, aromatic-
H), 7.86 (d, J 8.5, 2H, aromatic H), 7.11 (d, J 9.11, 1H, NH),
6.07–6.01 (m, 1H), 5.85–5.78 (m, 1H), 4.97 (d, J 1.6, 1H), 4.94–
4.91 (m, 1H), 4.89 (t, J 9.4, 1H); 13C-NMR (150 MHz, CDCl3):
d 172.0, 171.0, 170.9, 170.3, 170.0, 169.9 (2 s), (each COCH3),
166.5 (NHCO), 136.5 (C), 128.0 (CH), 99.2 (CH) 80.7 (CH), 79.2
(C-1 Glc), 74.1, 72.7, 70.1, 69.7, 69.0, 68.8, 65.7 (each CH), 62.3
(CH2), 61.9 (CH2), 21.2, 21.1, 21.0 (2 s), 20.9 (each COCH3);
HRMS (ES): anal. calcd 1423.4075 [M+Na]+, found 1423.4089;
IR (KBr): 3620–3361, 1759, 1678, 1544, 1375, 1227, 1042 cm-1.
Zemple´n deacetylation of this intermediate (70 mg, 0.05 mmol)
Cell assays
Cells of the human colon adenocarcinoma line SW480, also
treated with 150 mM 1-deoxymannojirimycin (Calbiochem, Darm-
stadt, Germany) for 24 h, the pancreatic carcinoma line Capan-1
and the CHO Lec8 mutant (kindly provided by P. Stanley, Albert
Einstein College of Medicine, Bronx, NY, USA) were cultured
and processed to monitor lectin-dependent cell staining using
biotinylated ConA as probe by quantitative fluorescence detection
in a FACScan instrument (Becton-Dickinson, Heidelberg, Ger-
many) with streptavidin/R-phycoerythrin (1 : 40; Sigma, Munich,
Germany) as indicator as described.26 Comparative analysis was
routinely done with aliquots of the same cell batch, and inter-assay
variability in percent did not exceed 12%.
4724 | Org. Biomol. Chem., 2009, 7, 4715–4725
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Angew. Chem., Int. Ed., 2004, 43, 2518–2521; (c) C. O’ Brien, M.
Pola´kova´, M. N. Pitt, M. Tosin and P. V. Murphy, Chem.–Eur. J., 2007,
13, 902–909.
Acknowledgements
Generous funding was provided by the European Commission
through Marie Curie IntraEuropean Fellowships ((MEIF-CT-
2003-500748,514958) to SGG and TVT as well as a Marie Curie
Research Training Network (contract no. CT-2005-19561), the
Programme for Research in Third-Level Institutions (PRTLI),
administered by the HEA (of Ireland) for funding, IRCSET and
Enterprise Ireland (MT), the research initiative LMUexcellent
and the Verein zur Fo¨rderung des biologisch-technologischen
Fortschritts in der Medizin e. V. (Heidelberg, Germany). We
gratefully acknowledge Dr B. Friday, Dr S. Namirha and Dr B. Tab
for inspiring discussion.
11 F. Mohamadi, N. G. J. Richards, W. C. Guida, R. Liskamp, M. Lipton,
C. Caufield, G. Chang, T. Hendrickson and W. C. Still, J. Comput.
Chem., 1990, 11, 440–467.
12 For selected publications on synthesis and applications of glycophanes
see: (a) R. R. Bukownik and C. S. Wilcox, J. Org. Chem., 1988, 53, 463–
467; (b) J. Jime´nez-Barbero, E. Junquera, M. Martin-Pastor, S. Sharma,
C. Vicent and S. J. Penades, J. Am. Chem. Soc., 1995, 117, 11198–11204;
(c) P. B. Savage, D. William and N. K. Dalley, J. Inclusion Phenom. Mol.
Recognit. Chem., 1997, 29, 335–346; (d) J. C. Morales and S. Penades,
Angew. Chem., Int. Ed., 1998, 37, 654–657; (e) T. Belghiti, J.-P. Joly, C.
Didierjean, S. Dahaoui and Y. Chapleur, Tetrahedron Lett., 2002, 43,
1441–1443.
13 O. Hayashida and I. Hamachi, J. Org. Chem., 2004, 69, 3509–3516.
14 For application of macrocyclic peptides for multivalent ligand display
see: (a) V. Wittmann and S. Seeberger, Angew. Chem., Int. Ed., 2000,
39, 4348–4352; (b) T. Ohta, N. Miura, N. Funitani, F. Nakajima, K.
Niikura, R. Sadamoto, C.-T. Guo, T. Suzuki, Y. Suzuki, M. Yasuo;, K.
Monde and S.-I. Nishimura, Angew. Chem., Int. Ed., 2003, 42, 5186–
5189.
References
1 H.-J. Gabius, ed. The Sugar Code. Fundamentals of Glycosciences,
Wiley-VCH, Weinheim, 2009.
15 (a) D. Page´, D. Zanini and R. Roy, Bioorg. Med. Chem., 1996, 4, 1949–
1961; (b) D. Page´ and R. Roy, Glycoconjugate J., 1997, 14, 345–356;
(c) R. Roy, D. Page´, S. F. Perez and V. V. Bencomo, Glycoconjugate J.,
1998, 15, 251–263.
2 (a) N. Sharon, and H. Lis, in Glycosciences: Status and Perspectives,
ed. H.-J. Gabius, and S. Gabius, Chapman & Hall, London, 1997, pp.
133–162; (b) G. Reuter and H.-J. Gabius, Cell. Mol. Life Sci., 1999,
55, 368–422; (c) H.-J. Gabius, H.-C. Siebert, S. Andre´, J. Jime´nez-
Barbero and H. Ru¨diger, ChemBioChem, 2004, 5, 740–764; (d) H.-J.
Gabius, Crit. Rev. Immunol., 2006, 26, 43–79; (e) S. Andre´, T. Kozˇa´r, S.
Kojima, C. Unverzagt and H.-J. Gabius, Biol. Chem., 2009, 390, 557–
566; (f) K. Honke, and N. Taniguchi, in The Sugar Code. Fundamentals
of Glycosciences, ed. H.-J. Gabius, Wiley-VCH, Weinheim, pp. 385–
402; (g) C. Zuber, and J. Roth, in The Sugar Code. Fundamentals of
Glycosciences, ed. H.-J. Gabius, Wiley-VCH, Weinheim, pp. 87–110.
3 (a) S. Andre´, C. Unverzagt, S. Kojima, X. Dong, C. Fink, K. Kayser and
H.-J. Gabius, Bioconjugate Chem., 1997, 8, 845–855; (b) C. Unverzagt,
S. Andre´, J. Seifert, S. Kojima, C. Fink, G. Srikrishna, H. H. Freeze,
K. Kayser and H.-J. Gabius, J. Med. Chem., 2002, 45, 478–491; (c) S.
Andre´, C. Unverzagt, S. Kojima, M. Frank, J. Seifert, C. Fink, K.
Kayser, C.-W. von der Lieth and H.-J. Gabius, Eur. J. Biochem.,
2004, 271, 118–134; (d) S. Andre´, S. Kojima, I. Prahl, M. Lensch,
C. Unverzagt and H.-J. Gabius, FEBS J., 2005, 272, 1986–1998; (e) S.
Andre´, T. Kozˇa´r, R. Schuberth, C. Unverzagt, S. Kojima and H.-J.
Gabius, Biochemistry, 2007, 46, 6984–6995.
4 (a) J. Kopitz, C. von Reitzenstein, S. Andre´, H. Kaltner, J. Uhl, V.
Ehemann, M. Cantz and H.-J. Gabius, J. Biol. Chem., 2001, 276,
35917–35923; (b) S. Andre´, H. Sanchez-Ruderisch, H. Nakagawa, M.
Buchholz, J. Kopitz, P. Forberich, W. Kemmner, C. Bo¨ck, K. Deguchi,
K. M. Detjen, B. Wiedenmann, M. von Knebel Doeberitz, T. M. Gress,
S.-I. Nishimura, S. Rosewicz and H.-J. Gabius, FEBS J., 2007, 274,
3233–3256; (c) D. D. Marathe, E. V. Chandrasekaran, J. T. Y. Lau, K. L.
Matta and S. Neelamegham, FASEB J., 2008, 22, 4154–4167; (d) R.
Schwartz-Albiez, in The Sugar Code. Fundamentals of Glycosciences,
ed. H.-J. Gabius, Wiley-VCH, Weinheim, pp. 447–468; (e) J. Wang, Z.-
H. Lu, H.-J. Gabius, C. Rohowsky-Kochan, R. W. Ledeen and G. Wu,
J. Immunol., 2009, 182, 4036–4045.
5 (a) M. Avalos, R. Babiano, M. J. Carretero, P. Cintas, F. J. Higes, J. L.
Jime´nez and J. C. Palacios, Tetrahedron, 1998, 54, 615–628; (b) P. V.
Murphy, H. Bradley, M. Tosin, N. Pitt, G. M. Fitzpatrick and W. K.
Glass, J. Org. Chem., 2003, 68, 5692–5704; (c) M. Tosin, S. G. Gouin
and P. V. Murphy, Org. Lett., 2005, 7, 211–214; (d) M. Tosin and P. V.
Murphy, J. Org. Chem., 2005, 70, 4107–4117.
6 (a) I. J. Goldstein, and R. D. Poretz, in The Lectins. Properties,
Functions, and Applications in Biology and Medicine, ed. I. E. Liener,
N. Sharon, and I. J. Goldstein, Academic Press, San Diego, 1986,
pp. 33–247; (b) H. Debray, D. Decout, G. Strecker, G. Spik and J.
Montreuil, Eur. J. Biochem., 1981, 117, 41–55; (c) S. K. Patnaik and P.
Stanley, Methods Enzymol., 2006, 416, 159–182.
7 R. T. Brown, F. Scheinmann and A. V. Stachulski, J. Chem. Res. (S),
1997, 370–371.
8 M. Upreti, D. Ruhela and R. A. Vishwakarma, Tetrahedron, 2000, 56,
6577–6584.
9 H. Kunz and H. Waldmann, Angew. Chem., Int. Ed. Engl., 1984, 23,
71–72.
16 (a) S. Andre´, S. Kojima, N. Yamazaki, C. Fink, H. Kaltner, K. Kayser
and H.-J. Gabius, J. Cancer Res. Clin. Oncol., 1999, 125, 461–474;
(b) G. Rappl, H. Abken, J. M. Muche, W. Sterry, W. Tilgen, S. Andre´,
H. Kaltner, S. Ugurel, H.-J. Gabius and U. Reinhold, Leukemia, 2002,
16, 840–845; (c) A. M. Wu, J. H. Wu, J.-H. Liu, T. Singh, S. Andre´,
H. Kaltner and H.-J. Gabius, Biochimie, 2004, 86, 317–326; (d) E. I.
Park, Y. Mi, C. Unverzagt, H.-J. Gabius and J. U. Baenziger, Proc.
Natl. Acad. Sci. U. S. A., 2005, 102, 17125–17129; (e) A. M. Wu, T.
Singh, J. H. Wu, M. Lensch, S. Andre´ and H.-J. Gabius, Glycobiology,
2006, 16, 524–537; (f) A. Beer, S. Andre´, H. Kaltner, M. Lensch, S.
Franz, K. Sarter, C. Schulze, U. S. Gaipl, P. Kern, M. Herrmann and
H.-J. Gabius, Cytometry, Part A, 2008, 73A, 139–147; (g) H.-J. Gabius,
Biochem. Soc. Trans., 2008, 36, 1491–1498; (h) S. Saussez, S. Cludts,
A. Capouillez, G. Mortuaire, K. Smatana Jr., H. Kaltner, S. Andre´,
X. Leroy, H.-J. Gabius and C. Decaestecker, Int. J. Oncol., 2009, 34,
433–439; (i) L. Stechly, W. Morelle, A.-F. Dessein, S. Andre´, G. Grard,
D. Trinel, M.-J. Dejoghe, E. Leteurtre, H. Drobecq, G. Trugnan, H.-J.
Gabius and G. Huet, Traffic, 2009, 10, 438–450.
17 A. Trynda, J. Madoj, A. Konitz and A. Wisniewski, Carbohydr. Res.,
2000, 329, 249–252.
18 A. Y. Chernyak, K. V. Antonov and N. K. Kotchetkov, Bioorgan. Khim.,
1987, 13, 958–966.
19 A. Y. Chernyak, L. O. Kononov and K. V. Antonov, Bull. Acad. Sci.
USSR, Div. Chem. Sci. (Engl. Transl.), 1988, 37, 1475–1482.
20 T. Velasco-Torrijos and P. V. Murphy, Org. Lett., 2004, 6, 3961–3964.
21 (a) M. Tosin, C. O’ Brien, G. M. Fitzpatrick, H. Mu¨ller-Bunz, W. K.
Glass and P. V. Murphy, J. Org. Chem., 2005, 70, 4096–4106; (b) H.
Bradley, G. Fitzpatrick, W. K. Glass, H. Kunz and P. V. Murphy, Org.
Lett., 2001, 3, 2629–2632.
22 W. C. Still, A. Tempczyk, R. C. Hawley and T. Hendrickson, J. Am.
Chem. Soc., 1990, 112, 6127–6129.
23 W. L. Jorgensen, S. D. Maxwell and J. Tirado-Rives, J. Am. Chem. Soc.,
1996, 118, 11225–11236.
24 A. Imberty and S. Pe´rez, Chem. Rev., 2000, 100, 4567–4588.
25 P. E. Jansson, L. Kenne, K. Persson and G. Widmalm, J. Chem. Soc.,
Perkin Trans. 1, 1990, 591–598.
26 (a) S. Andre´, B. Liu, H.-J. Gabius and R. Roy, Org. Biomol. Chem.,
2003, 1, 3909–3916; (b) J. C. Manning, K. Seyrek, H. Kaltner, S.
Andre´, F. Sinowatz and H.-J. Gabius, Histol. Histopathol., 2004, 19,
1043–1060; (c) S. Andre´, Z. Pei, H.-C. Siebert, O. Ramstro¨m and H.-J.
Gabius, Bioorg. Med. Chem., 2006, 14, 6314–6326; (d) S. Andre´, C. E. P.
Maljaars, K. M. Halkes, H.-J. Gabius and J. P. Kamerling, Bioorg. Med.
Chem. Lett., 2007, 17, 793–798; (e) S. Andre´, F. Sansone, H. Kaltner, A.
Casnati, J. Kopitz, H.-J. Gabius and R. Ungaro, ChemBioChem, 2008,
9, 1649–1661; (f) H. Kaltner, D. Sol´ıs, J. Kopitz, M. Lensch, M. Lohr,
J. C. Manning, M. Mu¨rnseer, M. Schno¨lzer, S. Andre´, J. L. Sa´iz and
H.-J. Gabius, Biochem. J., 2008, 409, 591–599; (g) H. Kaltner, D. Sol´ıs,
S. Andre´, M. Lensch, J. C. Manning, M. Mu¨rnseer, J. L. Sa´iz and H.-J.
Gabius, Biochemistry, 2009, 48, 4403–4416.
10 (a) T. Velasco-Torrijos and P. V. Murphy, Tetrahedron: Asymmetry,
2005, 16, 261–272; (b) M. Pola´kova´, N. Pitt, M. Tosin and P. V. Murphy,
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 4715–4725 | 4725
©