High-avidity, low-affinity multivalent interactions and the block to polyspermy in Xenopus laevis

Article abstract of DOI:10.1021/ja020536f

The interaction of the lectin XL35 with the jelly coat protein (JCP) surrounding oocytes in Xenopus laevis is essential for the block to polyspermy. The molecular details of this event are poorly understood, and the present study has been undertaken with

Full text of DOI:10.1021/ja020536f

Published on Web 10/10/2002
High-Avidity, Low-Affinity Multivalent Interactions and the
Block to Polyspermy in Xenopus laevis
Esther Arranz-Plaza, Alex S. Tracy,† Aloysius Siriwardena, J. Michael Pierce,† and
Geert-Jan Boons*
Contribution from the Complex Carbohydrate Research Center, UniVersity of Georgia,
220 RiVerbend Road, Athens, Georgia 30602
Received April 15, 2002
Abstract: The interaction of the lectin XL35 with the jelly coat protein (JCP) surrounding oocytes in Xenopus
laevis is essential for the block to polyspermy. The molecular details of this event are poorly understood,
and the present study has been undertaken with a view to delineating the mechanism of formation of the
fertilization envelope. A range of JCP-derived oligosaccharides were synthesized, and all were installed
with an artificial aminopropyl arm. This arm allowed the preparation of monovalent derivatives by acetylation
of the amino group or the synthesis of polyvalent compounds by attachment to an activated polyacrylamide
polymer. A number of analytical techniques, including enzyme-linked lectin assays and surface plasmon
resonance, have been developed and utilized to study the interactions of the mono- and polyvalent
compounds with XL35. The results reveal that the lectin XL35 has remarkably broad specificity for galactose-
containing saccharides and the affinities are only slightly modulated by secondary features, such as anomeric
configuration of the terminal sugar or the identity and linkage pattern of branching sugars. Broad specificity
was also observed when the saccharides were presented in a polyvalent fashion. The glycopolymers
displayed 10-20-fold increases in valency-corrected affinities compared to the corresponding monovalent
counterparts. Although the synthetic polymers are not as potent as the JCP, the kinetics of their interactions
mirror closely those of the native ligand, and in each case extremely long-lived interactions were observed.
The results of this study indicate that, in X. laevis, the true biological function of multivalency is not to
create an extremely tightly binding complex between XL35 and its natural ligand but, instead, to create a
very stable protective layer that will not dissociate and is yet flexible enough to encapsulate the developing
embryo. It is postulated that, even if these partners are unable to attain true equilibrium on the time scale
of the biological event, their mode of interaction would, nevertheless, be expected to guarantee an
insurmountable physical block to polyspermy. This study has also highlighted that multivalent interactions
require a very long time to achieve equilibrium, and this feature may well be the origin of several of the
ambiguities reported in the literature when multivalent ligands have been evaluated.
Introduction
The South African clawed toad, Xenopus laeVis, is a useful
model animal in which to study phenomena associated with
Carbohydrates are key to many of the events leading to
fertilization,1-3 and examples include recognition of the egg by
the sperm,4 induction of the acrosome reaction, fusion of the
sperm and egg, and the formation of the fertilization layer. This
last process involves fusion of cortical granules with the plasma
membrane and the subsequent release of the granule contents
into the glycoprotein matrix enveloping the oocyte. Interaction
of the cortical granule content with the egg glycoprotein matrix
induces an irreversible change to the egg jelly coat, establishing
a block to polyspermy.
fertilization and early development. In this organism, the
interaction of the novel carbohydrate-binding protein XL355-7
with the O-glycans of the jelly coat protein (JCP) enrobing the
egg is considered key in the prevention of polyspermy. XL35
is the first of a newly identified class of carbohydrate-binding
proteins8 that is characterized by its requirement for calcium
ions to bind its ligand, despite the fact its members do not
contain the calcium-binding motif found in the well-studied
C-type lectins.9 Homologues of XL35 have been found in
mice,10 rats,6 and humans,8 and the latter have been shown to
* Corresponding author. E-mail: gjboons@ccrc.uga.edu.
† Department of Biochemistry & Molecular Biology, Life Sciences Bldg.,
University of Georgia, Athens, GA 30602.
(5) Roberson, M. M.; Barondes, S. H. J. Biol. Chem. 1982, 257, 7520-7524.
(6) Lee, J. K.; Buckhaults, P.; Wilkes, C.; Teilhet, M.; King, M. L.; Moremen,
K. W.; Pierce, M. Glycobiology 1997, 7, 367-372.
(1) Mengerink, K. J.; Vacquier, V. D. Glycobiology 2001, 11, 37R-43R.
(2) Takasaki, S.; Mori, E.; Mori, T. Biochim. Biophys. Acta, Gen. Subj. 1999,
1473, 206-215.
(7) Nishihara, T.; Wyrick, R. E.; Working, P. K.; Chen, Y. H.; Hedrick, J. L.
Biochemistry 1986, 25, 6013-6020.
(3) Dell, A.; Morris, H. R.; Easton, R. L.; Patankar, M.; Clark, G. F. Biochim.
Biophys. Acta, Gen. Subj. 1999, 1473, 196-205.
(8) Lee, J. K.; Schnee, J.; Pang, M.; Wolfert, M.; Baum, L. G.; Moremen, K.
W.; Pierce, M. Glycobiology 2001, 11, 65-73.
(9) Lis, H.; Sharon, N. Chem. ReV. 1998, 98, 637-674.
(4) Vo, L. H.; Hedrick, J. L. Biol. Reprod. 2000, 62, 766-774.
9
10.1021/ja020536f CCC: $22.00 © 2002 American Chemical Society
J. AM. CHEM. SOC. 2002, 124, 13035-13046
13035

A R T I C L E S
Arranz-Plaza et al.
display unusual patterns of cell-type expression and most likely
function in microbial surveillance.8
In the present study, the interaction of XL35 with the JCP
has been systematically investigated using a range of chemically
synthesized mono-, di-, and trisaccharide derivativesstruncated
analogues of JCP-derived oligosaccharides. These saccharides
were also appended to a high-molecular-weight polymer to
probe the importance of the cluster effect for binding of the
JCP. Independent analytical techniques including enzyme-linked
lectin assays (ELLA) and surface plasmon resonance (SPR) have
been exploited to probe the interactions of these compounds
with XL35. A comparison has been made of the binding
behavior of the various synthetic glycopolymers with that of
the corresponding monovalent ligands and also with that of the
natural polymeric ligand, the JCP. This study has allowed
important aspects of the mode of interaction of XL35 with the
JCP to be delineated and provides an explanation of how a
physical block to polyspermy can be mounted.
Although native XL35 is an oligomeric glycoprotein with a
molecular weight of approximately 450000, under reducing
SDS-PAGE it has an apparent molecular weight of about 45000
and has consequently been estimated to be an oligomer of
between nine and twelve carbohydrate-binding units.6,11 Other
questions relating to the structure of XL35 and to its putative
ligand, however, remain only partially resolved. Limited
information on the identity of the natural ligand has come from
inhibition studies with simple saccharides, and these have
underlined the importance of terminal D-galactopyranose moi-
eties present in the JCP for its recognition by XL35.7,12 The
difficulties encountered in establishing the identity of the
endogenous carbohydrate epitope that gives optimum affinity
for XL35 stem from the fact that the jelly coat which layers the
unfertilized oocyte turns out to be structurally complex. The
JCP is composed predominantly of high-molecular-weight (>2
× 106) mucins and is over 60% sugar by weight.13 Moreover,
NMR and mass spectroscopic studies have revealed the mucin
to be composed of a very heterogeneous set of O-glycans.14-17
Furthermore, several of the reported JCP-derived oligosaccha-
rides are novel, such as a dimeric H antigen and a blood group
A antigen terminated with a GalNAc residue, but which of these
myriad of structures is important in the binding of XL35 to the
JCP has, however, remained a matter of conjecture.
Results and Discussion
Design of Carbohydrate Epitopes and Synthesis of Mon-
ovalent Ligands. O-Glycans released upon mild base treatment
of the mucin samples derived from the oocytes of six individual
X. laeVis toads have been found by NMR spectroscopy to
comprise 23 different structural motifs, 11 of which are unique.16
A large number of these structures are seen to carry a terminal
R-Gal(1-4) moiety, and taken together with the results of the
earlier inhibition studies,5,7,28-31 these data establish a terminal
R-D-galactopyranoside unit as an important structural determi-
nant of the putative ligand of XL35. The early inhibition studies
also indicate that various features of nonterminal sugar moieties
might serve as recognition elements.29 The trisaccharide D-Galp-
[L-Fucp-(R-1-4)]â-D-Galp and the two disaccharides D-Galp-
(R-1-4)â-D-Galp and L-Fucp-(R-1-2)â-D-Galp appear as ter-
minal sequences in a number of the reported JCP-derived
O-glycans, and consequently, these oligosaccharide structures
as well as those of simpler R- and â-D-galactopyranosides were
selected for investigation. Thus, the monosaccharides 1 and 2,
disaccharides 3-5, and trisaccharide 6 were targeted for
chemical synthesis (Figure 1). All are “installed” with amino-
propyl arms via which their eventual attachment to a desired
scaffold for presentation in a multimeric form would be
facilitated. Compounds 1-3 were prepared by routine proce-
dures (see the Supporting Information). Disaccharides 4a and
5a and trisaccharide 6a were assembled in a convergent
manner32 from the readily available glycosyl acceptors 9, 11,
and 16, respectively, by coupling with the appropriate glycosyl
donor 12 or 14 (Scheme 1). Furthermore, acceptor 11 could be
obtained from acceptor 9 by a simple two-step reaction
sequence. Thus, regioselective monobenzylation of the C-3
hydroxyl of precursor 7 was accomplished via a stannylidene
acetal, formed in situ by treatment of this tetrol with dibutyltin
oxide in refluxing methanol.33 Reaction of the thus formed tin
Any attempt to elucidate the mechanism of binding of XL35
with its natural ligand is complicated not only by the hetero-
geneity of the JCP, but also by the fact that its interaction with
XL35 is probably “multivalent” in nature.11,12 Many receptors
of carbohydrates when multiply presented bind avidly to their
respective ligands but only when these are presented multi-
valently.18-21 This cluster of weakly binding monovalent ligands
when properly orientated can sometimes generate enhancements
in association constants, over and above those expected on the
basis of valency. The “cluster effect”, as it has been coined,22
is now routinely proposed as the mechanism by which intrinsi-
cally weak protein-carbohydrate interactions are strengthened
and are manifested in vivo. The implication is that inherently
weak interactions are unlikely to be of any biological relevance.
Studies with synthetic ligands and a variety of polyvalent lectins
have established that these partners can enter into a tight binding
relationship if the initially weakly binding monovalent carbo-
hydrate epitope is taken and presented multivalently and in the
appropriate orientation.22-27
(10) Komiya, T.; Tanigawa, Y.; Hirohashi, S. Biochem. Biophys. Res. Commun.
1998, 251, 759-762.
(11) Chamow, S. M.; Hedrick, J. L. FEBS Lett 1986, 206, 353-357.
(12) Barondes, S. H. Science 1984, 223, 1259-1264.
(13) Hedrick, J. L.; Nishihara, T. J. Electron Microsc. Tech. 1991, 17, 319-
335.
(14) Guerardel, Y.; Kol, O.; Maes, E.; Lefebvre, T.; Boilly, B.; Davril, M.;
Strecker, G. Biochem. J. 2000, 352, 449-463.
(15) Plancke, Y.; Wieruszeski, J. M.; Alonso, C.; Boilly, B.; Strecker, G. Eur.
J. Biochem. 1995, 231, 434-439.
(25) Mammen, M.; Choi, S. K.; Whitesides, G. M. Angew. Chem., Int. Ed. 1998,
37, 2755-2794.
(16) Strecker, G.; Wieruszeski, J. M.; Plancke, Y.; Boilly, B. Glycobiology 1995,
5, 137-146.
(26) Bovin, N. V.; Gabius, H. J. Chem. Soc. ReV. 1995, 24, 413-&.
(27) Roy, R. Trends Glycosci. Glycotechnol. 1996, 8, 79-99.
(28) Wyrick, R. E.; Nishihara, T.; Hedrick, J. L. Proc. Natl. Acad. Sci. U.S.A.
1974, 71, 2067-2071.
(17) Tseng, K.; et al. Anal. Biochem. 1997, 250, 18-28.
(18) Drickamer, K. Struct. Biol. 1995, 2, 437-439.
(19) Rini, J. M. Annu. ReV. Biophys. Biomol. Struct. 1995, 24, 551-577.
(20) Weis, W. I.; Drickamer, K. Annu. ReV. Biochem. 1996, 65, 441-473.
(21) Lundquist, J. J.; Toone, E. J. Chem. ReV. 2002, 102, 555-578.
(22) Lee, Y. C.; Lee, R. T. Acc. Chem. Res. 1995, 28, 321-327.
(23) Roy, R. Polym. News 1996, 21, 226-232.
(29) Nitta, K.; Takayanagi, G.; Kawauchi, H. Chem. Pharm. Bull. 1984, 32,
2325-2332.
(30) Quill, T. A.; Hedrick, J. L. DeV., Growth Differ. 1994, 36, 615-620.
(31) Quill, T. A.; Hedrick, J. L. Arch. Biochem. Biophys. 1996, 333, 326-332.
(32) Boons, G. J. Tetrahedron 1996, 52, 1095-1121.
(33) David, S.; Hanessian, S. Tetrahedron 1985, 41, 643-663.
(24) Kiessling, L. L.; Gestwicki, J. E.; Strong, L. E. Curr. Opin. Chem. Biol.
2000, 4, 696-703.
9
13036 J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002

Block to Polyspermy in Xenopus laevis
A R T I C L E S
Scheme 1 a
Figure 1. Truncated JCP-derived analogues targeted for synthesis in this
study.
intermediate with benzyl bromide in the presence of the activator
tetrabutylammonium iodide gave the triol 8, which, upon
treatment with benzene dimethyl acetal in the presence of a
catalytic amount of p-toluenesulfonic acid in acetonitrile, gave
the desired 4,6-cis-benzylidene acetal 9 in a yield of 78%.
Acetylation of the C-2 hydroxyl of acceptor 9 and treatment of
the resulting derivative 10 with NaCNBH3/HCl34 led to regi-
oselective opening of the benzylidene acetal function to give
the glycosyl acceptor 11 in a yield of 77%. IDCP-mediated
coupling35 of acceptor 9 with the highly reactive thiofucosyl
donor 1236 gave solely the R-disaccharide 13 in good yield.
Removal of the benzyl ether and benzylidene protecting groups
in 13 with concomitant reduction of the azido moiety was
accomplished smoothly by catalytic hydrogenation over Pd/C
in ethanol acidified with HCl, to give target disaccharide 4a.
Coupling of acceptor 11 with thioglycoside 14 occurred
efficiently in the presence of the thiophilic promoter NIS/
TMSOTf35 to give the (1-4)-linked galactoside 15 in a yield
of 83% as a separable mixture of anomers (R/â ) 9/1). Removal
of the acetyl function in disaccharide 15 under standard
conditions gave derivative 16, which upon catalytic hydrogena-
tion over Pd/C led to removal of the benzyl ether groups and
reduction of the azido function to give the target disaccharide
5a. Alternatively, derivative 16 could serve in the synthesis of
the target trisaccharide 6a. Thus, IDCP-mediated35 coupling of
donor 14 with acceptor 16 gave the R-trisaccharide 17 in a yield
of 75% as the only anomer. Removal of the benzyl ether groups
occurred simultaneously with reduction of the azido group under
standard conditions to afford the target trisaccharide 6a.
Chemoselective N-acetylation of target building blocks 1a-6a
with acetic anhydride in methanol gave the required monovalent
derivatives 1b-6b, respectively. Disaccharide 3a was also
biotinylated for use in subsequent SPR experiments (see later).
a Reagents and conditions: (i) Bu2SnO, MeOH, ∆, then BnBr, Bu4NI,
toluene; (ii) PhCH(OMe)2, CSA, CH3CN; (iii) Ac2O, pyridine; (iv)
NaCNBH3, HCl‚Et2O, THF; (v) IDCP, Et2O, CH2Cl2, MS, 4 Å; (vi) Pd-
C, H2, 5% HCl-EtOH; (vii) Ac2O, MeOH; (viii) NIS/TMSOTf, DCM,
MS, 4 Å; (ix) NaOMe, MeOH, then Dowex H+.
Synthesis of Glycopolymers. It was considered that a
relatively high-molecular-weight, water-soluble, linear polymer,
onto which the various monovalent aminopropyl-armed precur-
sors were appended, might constitute an adequate mucin mimic
with which to probe the binding of the JCP. Glycopolymers
derived from poly[N-(acryloyloxy)succinimide] (pNAS) and an
aminoalkyl-armed oligosaccharide are well-studied25,37-42 and
were chosen for the present study. The target glycopolymers
(37) Schnaar, R. L.; Weigel, P. H.; Roseman, S.; Lee, Y. C. Methods Enzymol.
1982, 83, 306-310.
(38) Wang, J. Q.; Chen, X.; Zhang, W.; Zacharek, S.; Chen, Y. S.; Wang, P. G.
J. Am. Chem. Soc. 1999, 121, 8174-8181.
(39) Mammen, M.; Dahmann, G.; Whitesides, G. M. J. Med. Chem. 1995, 38,
4179-4190.
(40) Lees, W. J.; Spaltenstein, A.; Kingery, W. J. E.; Whitesides, G. M. J. Med.
Chem. 1994, 37, 3419-3433.
(34) Garegg, P. J. Pure Appl. Chem. 1984, 56, 845.
(35) Veeneman, G. H.; Leeuwen, S. H. v.; Boom, J. H. v. Tetrahedron Lett.
1990, 31, 1331-1334.
(41) Choi, S.-K.; Mammen, M.; Whitesides, G. M. J. Am Chem. Soc. 1996,
119, 4103-4111.
(42) Siriwardena, A.; Jorgensen, M. L.; Wolfert, A. W.; Vandenplas, M. L.;
Moore; N., J.; Boons, G.-J. J Am. Chem. Soc. 2001, 123, 8145-8146.
(36) Garegg, P. J. AdV. Carbohydr. Chem. Biochem. 1997, 52, 179-205.
9
J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002 13037

A R T I C L E S
Arranz-Plaza et al.
Scheme 2 a
Figure 2. Saturation binding study to characterize ELLA assay. Increasing
amounts of JCP-BAP conjugate were added to microtiter wells containing
no protein (blank) or a fixed amount of passively immobilized XL35. The
result for each concentration of JCP-BAP conjugate is expressed as the
difference between the absorbance readings for the wells containing XL35
and that of the blank (total - nonspecific binding). The assay saturated
reproducibly (nonlinear regression analysis), with an apparent KD of 6 µM,
on the basis of the quantity of galactose present in the JCP mucin (valence-
corrected). The inset shows a Scatchard transformation of the binding
isotherm. (The transformed data were not used to calculate the apparent
KD).
Assay Development and Validation. An assay format was
developed wherein the JCP was conjugated directly to an
enzyme and XL35 coated onto the microtiter plate. The required
JCP-linked enzyme was obtained by conjugating bacterial
alkaline phosphatase (BAP) to the mucin using a heterobifunc-
tional cross-linker. The amine-reactive cross-linker S-acetylm-
ercaptosuccinic anhydride (SATA) was employed, as coupling
of components could be carried out at near-physiological pH,
safeguarding the activity of the enzyme. Coupling of the JCP
to the linker was followed by deprotection of the thiol function
(0.5 M hydroxylamine), and this intermediate was reacted with
BAP that had been previously functionalized with maleimide
to give the required JCP-BAP conjugate. Although BAP is
larger and has a lower catalytic activity than the more commonly
used HRP, the former enzyme is not glycosylated and was
expected to be less likely to give artifacts in the subsequent
binding assay or when the sugar content of the JCP conjugate
was determined. Saturation binding studies were performed in
this second format, again to ensure that the conjugate JCP-
BAP interacted in a specific fashion (in this case, with the
microtiter-plate-bound lectin). The assay saturated reproducibly
with a valence-adjusted KD of 6 µM for the JCP based on the
quantity of galactose present on the mucin (Figure 2). This result
was confirmed with an inhibition assay using a JCP derivative
as a homologous inhibitor. The latter inhibitor was obtained by
taking the JCP derivatized with a thiol linker and alkylating it
with iodoacetamide rather than conjugating it to BAP. Gratify-
ingly, the apparent KI for the JCP determined from the data of
this experiment (Figure 3) is very similar to the apparent KD
obtained from the original saturation binding experiments. To
verify the specificity of the binding reaction, several negative
controls were performed, including assays with monosaccha-
rides, known from previous agglutination studies not to interact
with XL35. For example, N-acetyl-D-glucosamine or D-mannose
proved not to inhibit even at concentrations as high as 50 mM.
These results support that coating XL35 onto the microtiter plate
did not alter its binding specificity (see the inset in Figure 3).
Taken together, these experiments established that the assay was
a Reagents and conditions: (i) aminopropyl saccharide 1a-6a, Et3N,
DMF, then NH4OH.
19-24 were prepared by condensation of pNAS 18 with each
of the aminopropyl glycosides 1a-6a (Scheme 2), respectively,
followed by quenching of the remaining succinimate ester
groups with aqueous ammonia. Purification was achieved by
exaustive dialysis against water using a membrane with a
∼14000 molecular weight cutoff. All glycopolymers were
prepared from a single batch of pNAS, which was itself obtained
by radical-mediated polymerization of N-(acryloyloxy)succin-
imide. By appropriately controlling the ratio of aminopropyl
glycosides to activated esters in pNAS 18, functionalized
polymers 19c and 20-24 with 17% loading were obtained from
the corresponding aminopropyl glycosides 1a-6a. Two ad-
ditional polymers of 4% and 8% loading (19a and 19b,
respectively) were also obtained from aminopropyl disaccharide
3. The proposed structures of the glycopolymers were supported
1
by characteristic features in their H NMR spectral data, but
the loading of sugar on each glycopolymer was accurately
determined by a phenol-sulfuric assay using the appropriate
molar ratio of monosaccharides. The pNAS-derived poly(acrylic
acid) sodium salt and pNAS-derived polyacrylamide “back-
bones” were synthesized by treatment of precursor polymer 18
with sodium hydroxide or excess aqueous ammonia, respec-
tively. A molecular weight of 156000 was determined for the
pNAS-derived polyacid by gel filtration using commercial poly-
(acrylic acid) and dextran standards. The pNAS-derived amide
and acid were subsequently employed to determine possible
biological effects of the polymer backbone.
9
13038 J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002

Block to Polyspermy in Xenopus laevis
A R T I C L E S
galactopyranosides (data not shown). Disaccharide 4b (Fuc-
(â1,2)Gal) is a slightly poorer inhibitor than the corresponding
galactopyranoside derivative 1b, which suggests that the L-
fucoside moiety adds little to the binding affinity of the
galactopyranose epitope. This latter observation is consistent
with the failure to inhibit XL35 binding with high concentrations
of L-fucose in this assay (data not shown). Disaccharide
derivative 5b (GalR(1,4)Gal) is again no better an inhibitor than
the simpler R-galactoside derivative 1b, although trisaccharide
derivative 6b (GalR(1,4)[FucR(1,2]Gal), which contains a core
epitope found at the terminus of many of the mucin-derived
neutral sugars, is seen to be a measurably better inhibitor than
the other compounds tested. Collectively, our data suggest that
XL35 has a very broad specificity for galactose-containing
saccharides and the affinities are only slightly modulated by
secondary features, such as anomeric configuration of this
terminal sugar or the identity and linkage pattern of branching
sugars. Whereas earlier agglutination studies gave conflicting
results as to which D-galactopyranoside anomer was inhibitory,
the present assay establishes that both are recognized although
the R-anomer is preferred. It is interesting to note that the
inability of lectins to discriminate between anomers has only
been reported in a few cases. One such example pertinent to
this study relates to a family of oligosaccharide constructs
terminating in either R- or- â-D-galactose.43 Both anomers of
this family were found to be effective inhibitors of sperm
binding to mouse eggs.
Figure 3. Homologous inhibition experiments and negative controls. To
verify that the concentration of the JCP-BAP was appropriate for
subsequent inhibition studies, alkylated JCP was used as a homologous
inhibitor. Its apparent KI of 4.3 µM is in good agreement with the apparent
KD as determined previously from saturation binding experiments. Glycans
deemed to be irrelevant (from hemagglutination studies) were used as
negative controls and indeed show no effect on binding (insets).
Table 1. Inhibition Results of Monovalent Compounds Using
Inhibition ELLA (Immobilized XL35)
IC
IC
IC
50
(M × 10-3 a
)
50
50
compd
(M × 10-3 a
)
compd
(M × 10-3 a
)
compd
1b
2b
8 (6-10)
26 (21-31)
3b
4b
14 (8-17)
21 (16-28)
5b
6b
8 (6-11)
3 (2-4)
a Values in parentheses represent the 95% confidence intervals for each
IC50.
Inhibition by Multi- and Polyvalent Ligands. The inhibition
results for the glycopolymers 19-24 are presented in Table 2.
The data take into account the molecular weight of the parent
polymer backbone and are corrected for the total amount of
saccharide on each glycopolymer. As can be seen, the glyco-
polymers display a marked enhancement in activity; however,
on a valence-corrected basis the increase is only 10-20-fold.
Furthermore, the data of compounds 19a-c indicate that ligand
density has only a marginal effect on binding. The inhibition
assay also established that clustering of any of the monomeric
O-glycosides on a polymeric backbone has no bearing on the
specificity of the interaction with XL35, as the relative ranking
of the polymers parallels that of their corresponding monovalent
counterparts. Thus, XL35 retains its broad specificity even when
the ligands are presented in a polyvalent fashion. In contrast to
our results, other studies have shown that an oligosaccharide
clustering can increase or even switch the selectivities of
binding.44-48 Poly(acrylic acid) sodium salt and the polyamide
did not inhibit the binding, confirming that XL35 does not
interact with the polymer backbone.
Figure 4. Representative inhibition curves for monovalent compounds using
competition ELLA. XL35 was passively immobilized onto a microtiter plate,
and a fixed amount of JCP-BAP conjugate was competed away using
increasing concentrations of monovalent aminopropyl saccharides.
Binding curves for the JCP derivative (from a homogeneous
inhibition experiment), the glycopolymer 19c, and melibiose
alone are presented in Figure 5. The data demonstrate that, in
an equilibrium setting, the presentation of melibiose in a
multivalent fashion in the form of a synthetic polymer substan-
robust and that the newly synthesized JCP-BAP conjugate was
at the proper dilution to provide accurate IC50 values for other
inhibitors.
Inhibition by Monovalent Compounds. The newly devel-
oped ELLA inhibition assay was first used to evaluate the
affinities of the synthetic monovalent O-glycosides 1b-6b
(Figure 1). IC50 values are listed in Table 1, and representative
inhibition curves from which these values were obtained are
shown in Figure 4. Both monosaccharides 1b and 2b are
inhibitors, although the R-derivative is measurably preferred.
The same difference was also seen with R- and â-methyl
(43) Litscher, E. S.; Juntunen, K.; Seppo, A.; Penttila, L.; Niemela, R.; Renkonen,
O.; Wassarman, P. M. Biochemistry 1995, 34, 4662-4669.
(44) Adler, P.; Wood, S. J.; Lee, Y. C.; Lee, R. T.; Petri, W. A.; Schnaar, R. L.
J. Biol. Chem. 1995, 270, 5164-5171.
(45) Weatherman, R. V.; Chervenak, K. H. M.; Kiessling; L. L.; Toone, E. J.
Biochemistry 1996, 35, 3619-3624.
(46) Horan, N.; Yan, L.; Isobe, H.; Whitesides, G. M.; Kahne, D. Proc. Natl.
Acad. Sci. U.S.A. 1999, 96, 11782-11786.
(47) Liang, R.; Loebach, J.; Horan, N.; Ge, M.; Thompson, C.; Yan, L.; Kahne,
D. Biochemistry 1997, 97, 110554-10559.
(48) Roseman, D. S.; Baenziger, J. U. J. Biol. Chem. 2001, 276, 17052-17057.
9
J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002 13039

A R T I C L E S
Arranz-Plaza et al.
Table 2. Inhibition Data of Glycopolymers
compd IC50 (M × 10-5), IC50 (M × 10-3), compd IC50 (M × 10-5), IC50 (M × 10-3),
glycopolymer valence-adjusteda
glycopolymer valence-adjusteda
concentration
concentration
19 (4%)
19 (8%)
19 (17%)
20
1.6
1.6
0.7
1.8
0.4 (0.2-0.7) 21
0.9
1.9
2.4
0.5
0.4 (0.2-0.6)
1.2 (0.2-0.6)
1.7 (1.1-2.5)
0.4 (0.3-0.5)
0.7 (0.4-1.2) 22
0.6 (0.4-0.9) 23
2.5 (1.9-3.3) 24
a Values in parentheses represent the 95% confidence intervals for each
IC50.
Figure 6. SPR experiments designed to assess the effect of multivalency
on the kinetics of the interaction. The injection of glycopolymer 19c shows
a rapid but measurable association phase (from 200 to 300 s) that does not
come to equilibrium. After the initial 120 s of dissociation, the rate begins
to slow dramatically. The failure to reach equilibrium and the unusually
slow dissociation rate (varying with time) are both indicative of a multivalent
interaction.
such a calculation, the stability of a given complex would be
expected to be further enhanced, resulting in significant increases
in the estimated half-life of a given complex.
A series of SPR experiments were designed to study in detail
the avidity of the multivalent interaction by determining the
dissociation constants of monovalent melibiose and comparing
them with those of the melibiose glycopolymers. The kinetic
analyses were performed using a CM5 chip onto which was
coupled 5000 resonance units (RUs) of XL35. A series of
concentrations of polymer 19c were separately injected onto a
chip using the Kinject command with an association time of
120 s and a dissociation time of 1800 s. The particularly long
dissociation time was chosen in an attempt to maximize the
chances that bound polymer could dissociate. The data from a
representative sensorgram are shown in Figure 6 and represent
the signal from the XL35 surface minus the signal observed
for a surface to which no XL35 had been coupled. The
sensorgram shows a rapid but measurable association profile.
The initial dissociation phase was relatively rapid, with 15%
of the bound glycopolymer dissociating after 120 s, correspond-
ing to a rate of dissociation of 7.5%/min (Figure 6). The
population of glycopolymer capable of dissociating quickly
presumably represents the fraction tethered to the lectin by a
minimal number of epitopes and consequently able to dissociate.
After the initial 120 s, however, the rate of dissociation is seen
to drop sharply to 0.43%/min. After a 30 min dissociation phase
under flow, only 27% of the bound polymer was able to
dissociate. To verify that the multivalent interaction between
XL35 and the glycopolymer was glycan-specific, a negative
control experiment, utilizing a similar glycopolymer to which
N-acetylglucosamine, rather than melibiose, had been conju-
gated, was performed. This glycopolymer, when injected over
the XL35-derivatized surface, gave no detectable response,
demonstrating that the interactions observed for glycopolymer
19c with the XL35 coupled to the SPR chip were indeed
selective.
Figure 5. Comparison of monovalent and multivalent ligands using
competition ELLA. The results presented are valence-corrected and serve
to assess the magnitude of any cluster effect exhibited between XL35 and
any multivalent ligand. Glycopolymer 19c is approximately 16-fold more
potent an inhibitor as compared to monovalent melibiose, while the JCP is
approximately 60 times more potent than the glycopolymer.
tially alters its inhibition profile relative to that of monovalent
melibiose. The improvement in binding is, nevertheless, not as
dramatic as that observed for the native ligand. The JCP is
approximately 1000 times more potent than monovalent meli-
biose when the data are corrected for valence. The difference
in efficacy between the natural mucin and neoglycopolymers
may well be related to differences in their architectures: the
JCP is approximately 15 times larger in size than the melibiose
glycopolymer 19c and has a different backbone (protein vs
polyacrylamide backbone). Furthermore, the native ligand is
very heavily glycosylated, containing a quantity of carbohydrate
approximately 2-5 times the quantity of protein.
Assessing Multivalency and Ligand Binding Using Surface
Plasmon Resonance. SPR is an optical evanescent wave method
that is increasingly finding use in the study of interactions of
biological molecules.49 This approach has as an advantage that
it provides rate constants for both forward and reverse reactions.
If the interaction of a multivalent lectin and a polyvalent ligand
were to display a cluster effect, the observed SPR dissociation
profile would differ from that of a monovalent ligand, since
the apparent off-rate (kd(apparent)) is the microscopic kd raised
n
to the valency of the interaction (kd(apparent) ) kd ), where n
is the valency of the interaction. Qualitatively this may be
thought of as the decreased probability that n sites can dissociate
simultaneously, a condition that is required for the multivalent
ligand to completely dissociate from the complex. Estimates of
the half-lives for lectin-ligand complexes, when calculated in
this fashion, are seen to increase dramatically with valency. If
the effects of rebinding were also to be taken into account in
The observed stability of the multivalent interaction renders
its evaluation by an SPR analysis complex. Determination of
the KD of any multivalent interaction using SPR is not
straightforward since the data cannot be fit to any simple mono-
or bivalent model. An alternative approach is to take advantage
of an equilibrium SPR analysis to obtain the dissociation
(49) McDonnell, J. M. Curr. Opin. Chem. Biol. 2001, 5, 572-577.
9
13040 J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002

Block to Polyspermy in Xenopus laevis
A R T I C L E S
calculated for the apparent KI from the competition ELLA using
an exact correction to the Cheng-Prushoff50,51 approximation.
This finding is important because studies by Toone and others21
have shown that considerable discrepancies may be observed
when different methodologies are employed to determine
binding affinities.
An additional SPR experiment using a solution competition
format was designed to determine the KD for monovalent
melibiose. In this experiment, melibiose was immobilized on
the surface of a chip and the system calibrated using increasing
concentrations of XL35. Fixed concentrations of XL35 were
each mixed with increasing amounts of melibiose, and each of
these mixtures was then separately injected over the surface of
immobilized melibiose. As the concentration of melibiose in
the solution was increased, XL35 was competed away from the
melibiose surface, thereby decreasing the signal. Since the initial
concentrations of both XL35 and melibiose are known, the
signal obtained represents the quantity of XL35 that is free (and
therefore able to bind the surface) at a given concentration of
melibiose in solution. On the basis of the concentrations of free
and bound XL35 at each concentration of melibiose in solution,
the KD may be determined. The format of these experiments
takes advantage of XL35’s large size, since SPR responds
directly to the mass of material adsorbed onto the surface of
the chip. Additionally, prior experiments (data not shown)
demonstrated that attempts to directly measure the KD for
monovalent melibiose by injecting increasing concentrations of
melibiose over an XL35-conjugated chip yielded extremely low
signals due to the small mass of the disaccharide. Since these
experiments were designed to measure monovalent affinities
rather than multivalent interactions, a critical feature of the
experiment was that the ligand density of the melibiose on the
surface must be sufficiently low to prevent multivalent interac-
tions. This condition was achieved by biotinylating melibiose
derivative 3 and capturing the biotinylated disaccharide onto a
CM5 chip containing immobilized streptavidin. By contrast to
the strongly multivalent interaction between XL35 and the
glycopolymers where a minimal amount of the complex
dissociated even after 30 min of flow, under the low coupling
density conditions described here nearly 90% of the bound
complex dissociated after only 2.5 min of flow. The monovalent
dissociation profile demonstrated that the subsequent competi-
tion experiments measured monovalent affinities. On the basis
of these data, the KD for monovalent melibiose was determined
to be 1.57 × 10-3 M, which is an order of 10-fold less than the
corresponding IC50 established by competition ELLA. This
difference could be due to the fact that the SPR solution
competition assay developed in this study is designed to measure
binding at a single site, as opposed to evaluating the ability of
a monovalent inhibitor to compete multivalent ligands away
from a multivalent receptor.
Figure 7. Equilibrium SPR binding experiments using 5000 RUs of XL35
coupled to a CM5 chip. (A, top) Sensorgrams using 2-fold serial dilutions
of melibiose glycopolymer 19c (from 1.6 × 10-5 to 3.1 × 10-8 M). Injection
times were maximized to allow equilibrium to be reached (as the data
obtained would be amenable to a nonlinear regression analysis). (B, bottom)
A plot of the nonlinear regression analysis using the glycopolymer
concentrations and the associated equilibrium responses. This approach
allowed the KD of the interaction between XL35 and glycopolymer 19c to
be determined despite the complications associated with multivalent
interactions.
constant, although multivalency and the resulting stability of
the bound complex again need to be contended with. If one
considers the rate equation of a receptor-ligand interaction
d[RL]/dt ) ka[R][L] - kd[RL]
where [R] ) concentration of the receptor, [L] ) concentration
of the ligand, ka ) association constant, kd ) dissociation
constant, and d[RL]/dt represents the change in the extent of
binding with respect to time.
It follows that equilibrium is reached when d[RL]/dt ) 0,
but this is only true as t approaches infinity. After 5 half-lives,
however, the measured binding is within 3% of the true
equilibrium value, so this represents a practical basis for
achieving equilibrium to conduct equilibrium SPR experiments.
In the case of multivalent interactions, it is extremely difficult
to approach equilibrium, as this requires that, for every
receptor-ligand complex formed, another must dissociate,
which is statistically unlikely (see above). Additionally, since
a concentration of ligand 10 times that of the KD is required for
an equilibrium analysis to be properly evaluated (to saturate
the receptor), this type of experiment requires not only long
injections or contact times but also that large quantities of ligand
be injected. The sensorgrams shown in Figure 7 are derived
from experiments in which a balance has been struck between
the long injection times required to reach equilibrium and the
large quantities of material consumed under such conditions.
Even under these relatively drastic conditions, the system is seen
not to come completely to equilibrium. Nevertheless, using this
approach, a non-valence-corrected KD of 1.67 × 10-6 M can
be calculated from the sensorgram for the glycopolymer 19c,
The glycopolymers synthesized in this study are seen to
exhibit behavior akin to that of the natural ligand and, in
important respects, do serve as reasonable mimics for this mucin.
Although the SPR-derived equilibrium inhibition profiles dem-
onstrate that the synthetic polymers are not as potent as the
JCP, the kinetics of their interactions mirror closely those of
the native ligand. This is despite the fact that this mucin is 10
(50) Cheng, Y.-C.; Prusoff, W. H. Biochem. Pharmacol. 1973, 22, 3099-3108.
(51) Munson, P. J.; Rodbard, D. J. Recept. Res. 1988, 8, 533-546.
in excellent agreement with the value of 4.3 × 10-6
M
9
J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002 13041

A R T I C L E S
Arranz-Plaza et al.
Experimental Section
times larger than the polymers and shows a correspondingly
higher potential for interacting multivalently. It is likely that
XL35 induces cross-linking of the mucin, which would result
in formation of a three-dimensional matrix impenetrable to
sperm. In addition, this cross-linking would be expected to result
in a modest increase in valence-corrected affinity as is indeed
borne out by our results. It appears that XL35 has to sample
many different regions of the jelly coat matrix before forming
a thermodynamically favorable complex that has minimal chance
of dissociating from the jelly coat mucin. An alternative scenario
would be one in which XL35 underwent a tight and irreversible
multivalent binding interaction with a specific ligand on the
surface of the oocyte immediately upon secretion. Under such
circumstances, which are reminiscent of a classical cluster effect,
the lectin would not be able to diffuse through the jelly coat
enrobing the egg. As a consequence, the fertilization envelope
would be only partially formed and thus would not be expected
to provide an effective physical block to polyspermy.
General Information. Chemistry. Chemicals for synthesis were
purchased from Aldrich, Acros, Sigma, and Fluka and used without
further purification. Molecular sieves were activated at 350 °C in vacuo
for 3 h. All solvents were distilled from the appropriate drying agents.
All reactions were performed under anhydrous conditions according
to routine dry techniques unless otherwise indicated. Reactions were
monitored by thin-layer chromatography (TLC) on Kieselgel 60 F254
(Merck). Detection was by examination under UV light (254 nm) and
by charring with 10% sulfuric acid in methanol. Column chromatog-
1
raphy was performed on silica gel (Merck, 70-230 mesh). H NMR
and 13C NMR spectra were recorded on Varian Inova300, Inova500,
and Inova600 spectrometers equipped with Sun workstations. Spectra
were recorded in deuterated solvent and referenced appropriately.
Assignments were made, when required, with the aid of standard
gCOSY, gHSQC, TOCSY, and gHMBC experiments. Negative ion
matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF)
mass spectra were recorded using an HP-MALDI instrument using
either gentisic acid or dihydroxybenzoic acid as the matrix. High-
resolution mass spectroscopy was performed on a Voyager delayed
extraction STR using 2,5-dihydrobenzoic acid as an internal calibration
matrix.
The present study has also highlighted the requirement to
establish equilibrium if an assay is to report accurately on any
binding event. This condition is seen to be particularly difficult
to meet when the binding partners are multivalent. This may
well be the origin of several of the ambiguities reported in the
literature when multivalent ligands have been evaluated.52
3-Azidopropyl 3-O-Benzyl-â<rsf>-D-galactopyranoside (8). A
mixture of 3-azidopropyl â-D-galactopyranoside (1.58 g, 6.0 mmol)
and dibutyltin oxide (1.65 g, 6.7 mmol) in MeOH (45 mL) was heated
under reflux for 2 h, after which time the reactants had completely
dissolved. The methanol was then evaporated under reduced pressure.
The resulting syrup was dissolved in toluene (45 mL), and tetrabuty-
lammonium iodide (2.44 g, 6.7 mmol) and benzyl bromide (0.78 mL,
6.7 mmol) were added. This mixture was heated under reflux for 12 h
and then concentrated in vacuo. The resulting oil was purified by flash
column chromatography over silica gel (eluent ethyl acetate/hexanes,
Conclusion
The lectin XL35 shows a remarkable ability to bind a wide
variety of both monovalently and polyvalently presented D-
galactopyranosides. Taken together, the results strongly suggest
that broad lectin specificity and mucin heterogeneity are integral
to the interaction of XL35 with the JCP. That the observed broad
specificity is biologically relevant is supported by the observa-
tion that intraspecific O-glycan heterogeneity observed in
Xenopus does not have a crucial bearing on the viability of the
fertilized eggs.14 Multivalency is certainly important in the
binding observed for the synthetic glycopolymers synthesized
in this study, but the interactions manifest only modest increases
in their valency-corrected affinities, relative to their monomeric
counterparts. However, the net interaction and the kinetics of
binding displayed by the glycopolymers are seen to mirror
closely those of the JCP when examined by SPR. The results
suggest that in X. laeVis the true biological function of
multivalency is not to create an extremely tightly binding
complex between XL35 and its natural ligand. It serves, instead,
to create a very stable protective layer that will not dissociate
and is yet flexible enough to encapsulate the deVeloping embryo.
In the system under scrutiny, an extremely long-lived interaction
of polymeric ligand and its multimeric receptor is achieved,
essentially, through multiple and successive weak binding and
rebinding events and not through any one all-important and
static, tight binding interaction. Through this cascade of
reversible low-affinity interactions a dynamic interplay between
the JCP and XL35 is established, which must manifest itself
macroscopically as what is, in effect, a fluid hydrogel. These
partners might not be able to attain true equilibrium on the time
scale of the biological event, but their mode of interaction would,
nevertheless, be expected to guarantee an insurmountable
physical block to polyspermy.
2/1, v/v) to give triol 8 as a syrup (1.56 g, 74%). [R]23 +1.8 (c )
D
0.39, CH2Cl2). FAB-MS (m/z): 354.2 (M+ + 1). 1H NMR (300 MHz,
CDCl3): 7.40-7.25 (m, 5H, CHAr), 4.74 (s, 2H, CH2Ph), 4.25 (d, 1H,
J ) 7.7 Hz, H-1), 4.02-3.92 (m, 3H, SpOCH2 , H-6a, H-4), 3.86-
a
3.72 (m, 2H, H-2, H-6b), 3.70-3.62 (m, 1H, SpOCH2 ), 3.52-3.38
b
(m, 4H, CH2N3, H-3), 2.70-2.00 (br s, 3H, 3 OH), 1.87 (m, 2H, J )
6.3 Hz, Sp-CH2). 13C NMR (75 MHz, CDCl3): δ 137.7 (C-Ar), 128.8,
128.3, 128.1 (CH-Ar), 103.4 (C-1), 80.5, 74.6 (C-5, C-3), 72.4 (CH2-
Ph), 71.2 (C-2), 67.3 (C-4), 66.9 (Sp-OCH2), 62.7 (C-6), 48.6 (CH2N3),
29.5 (Sp-CH2). HR MS (m/z): calcd for C16H23N3O6Na 376.1485,
found 376.1460.
3-Azidopropyl 3-O-Benzyl-4,6-O-benzylidene-â-D-galactopyra-
noside (9). A solution of 8 (1.24 g, 3.5 mmol) in dry acetonitrile (20
mL) was treated with R,R-dimethoxytoluene (1.05 mL, 7 mmol) and
p-toluenesulfonic acid (33 mg). The mixture was stirred for 22 h at
room temperature and then neutralized with Et3N. The reaction mixture
was concentrated under reduced pressure and the residue purified by
silica gel column chromatography (eluent ethyl acetate/hexanes, 3/1,
v/v) to yield alcohol 9 as an amorphous solid (1.2 g, 77.5%). [R]24
D
+42.0 (c ) 4.1, CH2Cl2). FAB-MS (m/z): 442.2 (M+ + 1). H NMR
1
(300 MHz, CDCl3): δ 7.50-7.10 (m, 10H, ArCH), 5.36 (s, 1H,
OCHO), 4.72-4.60 (AB, 2H, J ) 12.4 Hz, CH2Ph), 4.28-4.14 (m,
2H, H-1, H-6a), 4.03 (d, 1H, J ) 3.1 Hz, H-4), 3.98-3.84 (m, 3H,
H-2, H-6b, Sp-OCH2 ), 3.60-3.50 (m, 1H, Sp-OCH2 ), 3.44-3.36
(dd, 1H, J ) 9.6 Hz, J ) 3.1 Hz, H-3), 3.33 (t, 2H, J ) 6.6 Hz, CH2N3),
3.28-3.22 (m, 1H, H-5), 1.90-1.70 (m, 2H, Sp-CH2). 13C NMR (75
MHz, CDCl3): δ 138.1, 137.8 (ArC), 129.0, 128.5, 128.2, 127.9, 126.4
(ArCH), 103.2 (C-1), 101.2 (OCHO), 79.4 (C-5), 73.2 (C-3), 71.7 (CH2-
Ph), 70.1 (C-2), 69.4 (C-6), 66.9 (C-4), 66.7 (Sp-OCH2), 48.7 (CH2N3),
29.4 (Sp-CH2). HR MS (m/z): calcd for C25H27N3O6Na 464.1798,
found 464.1794.
a
b
3-Azidopropyl 2-O-Acetyl-3-O-benzyl-4,6-O-benzylidene-â-D-ga-
lactopyranoside (10). A solution of 9 (1.2 g, 2.7 mmol) in pyridine
(20 mL) was treated with acetic anhydride (20 mL). After being stirred
(52) Mann, D. A.; Kanai, M.; Maly, D. J.; Kiessling, L. L. J. Am. Chem. Soc.
1998, 120, 10575-10582.
9
13042 J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002

Block to Polyspermy in Xenopus laevis
A R T I C L E S
at room temperature for 12 h, the reaction mixture was concentrated
under reduced pressure and the residue purified by flash silica gel
column chromatography (eluent ethyl acetate/hexanes, 1/1.5, v/v) to
139.2, 139.1, 138.9, 138.4, 138.3, 138.1 (ArC), 128.5, 128.4, 128.3,
128.2, 128.1, 128.0, 127.9, 127.8, 127.6, 127.5, 127.4, 127.3 (ArCH),
101.6 (C-1), 100.8 (C-1′), 79.3 (C-3), 79.2 (CH), 77.6 (CH2Ph), 77.2
(CH2), 76.8 (CH2), 76.6 (CH), 75.1 (CH2), 75.0 (CH), 74.1 (CH), 73.9
(CH2Ph), 73.7 (CH), 73.3 (CH2Ph), 73.2 (CH2Ph), 72.5 (CH2Ph), 71.8
(CH2Ph), 71.0 (C-2), 69.4 (C-5 or C-5′), 68.2 (C-6), 67.9 (C-6′), 65.8
(Sp-OCH2), 48.4 (CH2N3), 29.4 (Sp-CH2), 21.3 (CH3). HR MS (m/
z): calcd for C59H65N3O12Na 1030.4466, found 1030.4435.
give 10 as an oil (1.25 g, 95%). [R]23 +55.4 (c ) 0.57, CH2Cl2).
D
1
FAB-MS (m/z): 482.3 (M+ - 1). H NMR (300 MHz, CDCl3): δ
7.60-7.20 (m, 10 H, ArCH), 5.48 (s, 1H, OCHO), 5.40-5.31 (dd,
1H, J ) 9.9 Hz, J ) 8.0 Hz, H-2), 4.75-4.58 (AB, 2H, J ) 12.6 Hz,
CH2Ph), 4.39 (d, 1H, J ) 8.0 Hz, H-1), 4.32-4.24 (m, 1H, H-6a), 4.17
(d, 1H, J ) 3.3 Hz, H-4), 4.07-4.00 (m, 1H, H-6b), 4.00-3.90 (m,
3-Azidopropyl 3,6-Di-O-benzyl-4-O-(2,3,4,6-tetra-O-benzyl-r-D-
galactopyranosyl)-â-D-galactopyranoside (16). A solution of 15 (92
mg, 0.09 mmol) in MeOH (5 mL) was treated with a catalytic amount
of NaOMe. After being stirred for 24 h at room temperature, the reaction
mixture was neutralized with Dowex H+ resin, filtered, and concentrated
under reduced pressure. The residue was purified by silica gel column
chromatography (eluent ethyl acetate/hexanes, 1/3, v/v) to give 16 as
a syrup (84.6 mg, 96% yield). [R]25D +44.4 (c ) 0.44, CH2Cl2). FAB-
a
b
1H, Sp-OCH2 ), 3.63-3.52 (m, 2H, H-3, Sp-OCH2 ), 3.40-3.30 (m,
3H, H-5, CH2N3), 2.08 (s, 3H, CH3CO), 1.94-1.70 (m, 2H, Sp-CH2).
13C NMR (75 MHz, CDCl3): δ 169.4 (CdO), 138.1, 137.7 (ArC),
129.0, 128.9, 128.5, 128.4, 128.1, 127.8, 127.6, 126.4 (ArCH), 101.2,
101.1 (C-1, OCHO), 77.4 (C-5), 73.2 (C-3), 71.2 (CH2Ph), 70.2 (C-2),
69.2 (C-6), 66.7 (C-4), 65.6 (Sp-OCH2), 48.2 (CH2N3), 29.2 (Sp-
CH2), 21.2 (CH3). HR MS (m/z): calcd for C25H29N3O7Na 506.1903,
found 506.1841.
1
MS (m/z): 987.9 (M+ + Na). H NMR (300 MHz, CDCl3): δ 7.30-
7.10 (m, 30 H, ArCH), 4.95 (d, 1H, J ) 2.2 Hz, H-1′), 4.83 (d, 1H, J
) 11.3 Hz, aCH2Ph), 4.81 (d, 1H, J ) 11.8 Hz, aCH2Ph), 4.73 (d, 1H,
J ) 12.6 Hz, aCH2Ph), 4.67 (s, 2H, CH2Ph), 4.60 (d, 1H, J ) 11.8 Hz,
3-Azidopropyl 2-O-Acetyl-3,6-O-dibenzyl-â-D-galactopyranoside
(11). To a mixture of 10 (0.21 g, 0.44 mmol), NaBH3CN (0.35 g, 5.6
mmol), and powdered 3 Å molecular sieves in dry THF was added a
1 M HCl/ether solution dropwise until evolution of gas had ceased.
The reaction mixture was then stirred for 2 h at room temperature,
diluted with THF, and filtered through a pad of Celite and the filtrate
concentrated under reduced pressure. The residue was purified by silica
gel flash column chromatography (eluent ethyl acetate/hexanes, 1/2,
v/v) to afford 11 as an oil (0.16 g, 77%). [R]25D +14.5 (c ) 1.8, CH2-
Cl2). FAB-MS (m/z): 508.3 (M+ + Na). 1H NMR (300 MHz, CDCl3):
δ 7.40-7.20 (m, 10 H, ArCH), 5.10 (dd, 1H, J ) 9.6 Hz, J ) 8.0 Hz,
H-2), 4.62 (d, 1H, J ) 12.4 Hz, CH2Ph), 4.52 (s, 2H, CH2Ph), 4.45 (d,
1H, J ) 12.4 Hz, CH2Ph), 4.25 (d, 1H, J ) 8.0 Hz, H-1), 4.00 (d, 1H,
b
bCH2Ph), 4.45 (d, 1H, J ) 11.3 Hz, CH2Ph), 4.41 (d, 1H, J ) 12.6
Hz, bCH2Ph), 4.33 (dd, 1H, J ) 8.8 Hz, J ) 4.8 Hz, H-6a), 4.24-4.10
(m, 5H, bCH2Ph, bCH2Ph, CH2Ph, H-1), 4.04-3.82 (m, 6H, H-2′, H-3′,
H-4, H-5′, H-6′a, Sp-OCH2 ), 3.73 (dd, 1H, J ) 9.6 Hz, J ) 7.7 Hz,
a
H-2), 3.62-3.38 (m, 4H, H-4′, H-5, H-6′b, Sp-OCH2 ), 3.14 (t, 2H, J
b
) 6.6 Hz, CH2N3), 3.24-3.16 (m, 2H, H-3, H-6b), 1.95-1.85 (m, 2H,
Sp-CH2). 13C NMR (75 MHz, CDCl3): δ 139.0, 138.8, 138.7, 138.1
(ArC), 128.5, 128.4, 128.3, 128.2, 128.1, 128.0, 127.9, 127.7, 127.6,
127.5 (ArCH), 103.5 (C-1), 100.5 (C-1′), 80.7, 79.2, 76.7 (CH), 75.1
(CH2Ph), 75.0, 74.1 (CH), 73.9 (CH2), 73.7 (CH), 73.4, 73.2, 72.5,
71.9 (CH2), 70.8, 69.6 (CH), 68.2, 68.1, 66.7 (CH2), 48.6 (CH2N3),
29.4 (Sp-CH2). HR MS (m/z): calcd for C57H63N3O11Na 988.4360,
found 988.4426.
a
J ) 3.3 Hz, H-4), 3.90-3.80 (m, 1H, Sp-OCH2 ), 3.80-3.60 (m, 2H,
b
H-6), 3.52 (m, 1H, H-5), 3.50-3.44 (m, 1H, Sp-OCH2 ), 3.42 (dd,
1H, J ) 9.6 Hz, J ) 3.3 Hz, H-3), 3.28 (t, 2H, J ) 6.6 Hz, CH2N3),
1.98 (s, 3H, CH3CO), 1.90-1.80 (m, 2H, Sp-CH2). 13C NMR (75 MHz,
CDCl3): δ 169.6 (CdO), 138.1, 137.5 (ArC), 128.7, 128.6, 128.2, 127.9
(ArCH), 101.3 (C-1), 78.9 (C-3), 74.0 (CH2Ph), 73.7 (C-5), 71.8 (CH2-
Ph), 70.9 (C-2), 69.2 (C-6), 66.4 (C-4), 66.0 (Sp-OCH2), 48.3 (CH2N3),
29.3 (Sp-CH2), 21.2 (CH3). HR MS (m/z): calcd for C29H32N3O7Na
508.2060, found 508.1954.
3-Azidopropyl 3-O-Benzyl-4,6-O-benzylidene-2-O-(2,3,4-tri-O-
benzyl-r-L-fucopyranosyl)-â-D-galactopyranoside (13). Compounds
9 (0.25 g, 0.56 mmol) and 12 (0.32 g, 0.68 mmol), freshly activated 4
Å molecular sieves, and IDCP (0.53 g, 1.1 mmol) were stirred in ether/
CH2Cl2 (5/1, v/v, 18 mL) at room temperature for 3.5 h. The molecular
sieves were filtered off, and the filtrate was diluted with CH2Cl2 (20
mL) and washed with aqueous Na2S2O3 (20 mL) and then water (3 ×
20 mL). The organic layer was dried (MgSO4) and concentrated under
reduced pressure. The residue was purified by silica gel column
chromatography (eluent ethyl acetate/hexanes, 1/2, v/v) to yield
3-Azidopropyl 2-O-Acetyl-3,6-di-O-benzyl-4-O-(2,3,4,6-tetra-O-
benzyl-r/ â-D-galactopyranosyl)-â-D-galactopyranoside (15). A mix-
ture of 14 (0.118 g, 0.2 mmol), 11 (98 mg, 0.2 mmol), and freshly
activated 4 Å molecular sieves in dry ether/CH2Cl2 (4/1, v/v; 2.5 mL)
was stirred for 1 h at room temperature and then 30 min at -25 °C.
NIS (68 mg, 0.3 mmol) and TMSOTf (5.5 µL, 0.03 mmol) were then
added to the mixture. After 20 h of stirring at -25 °C, the molecular
sieves were filtered off, the filtrate was diluted with CH2Cl2 (20 mL),
and the organic phase was washed with aqueous Na2S2O3 (20 mL) and
water (3 × 20 mL), dried (MgSO4), and concentrated under reduced
pressure. The residue was purified by silica gel column chromatography
(eluent ethyl acetate/hexanes, 1/4, v/v) to obtain 15 as an R/â mixture
(9/1 by 1H NMR) of (0.167 g, 83%) which was separated by silica gel
column chromatography (eluent ethyl acetate/toluene, 0.5/10, v/v) to
disaccharide 13 as a white foam (0.38 g, 78% yield). [R]25 -28.9 (c
D
) 5.1, CH2Cl2). 1H NMR (300 MHz, CDCl3): δ 7.76-7.20 (m, 25 H,
ArCH), 5.73 (d, 1H, J ) 3.6 Hz, H-1′), 5.49 (s, 1H, OCHO), 5.03 (d,
a
a
1H, J ) 11.5 Hz, CH2Ph), 4.94 (d, 1H, J ) 11.8 Hz, CH2Ph), 4.83
b
a
(d, 1H, J ) 11.8 Hz, CH2Ph), 4.80 (d, 1H, J ) 11.8 Hz, CH2Ph),
4.75 (d, 1H, J ) 12.1 Hz, CH2Ph), 4.72 (d, 1H, J ) 11.5 Hz, CH2-
a
b
b
Ph), 4.64 (d, 1H, J ) 12.1 Hz, CH2Ph), 4.63 (d, 1H, J ) 11.8 Hz,
bCH2Ph), 4.54 (d, 1H, J ) 8.0 Hz, H-1), 4.45-4.34 (m, 2H, H-5′, H-6a),
4.26 (dd, 1H, J ) 8.0 Hz, J ) 9.3 Hz, H-2), 4.22 (d, 1H, J ) 3.6 Hz,
a
H-4), 4.12-3.98 (m, 4H, H-3′, H-2′, H-6b, Sp-OCH2 ), 3.88 (dd, 1H,
give the pure R anomer. [R]25 +57.9 (c ) 0.22, CH2Cl2), FAB-MS
D
J ) 9.3 Hz, J ) 3.6 Hz, H-3), 3.76 (m, 1H, H-4′), 3.69-3.59 (m, 1H,
(m/z): 1030.7 (M+ + Na). 1H NMR (300 MHz, CDCl3): δ 7.30-7.05
(m, 30 H, ArCH), 5.16 (dd, 1H, J ) 10.2 Hz, J ) 8.0 Hz, H-2), 4.94
(d, 1H, J ) 3.0 Hz, H-1′), 4.84 (d, 1H, J ) 12.1 Hz, aCH2Ph), 4.80 (d,
1H, J ) 11.3 Hz, aCH2Ph), 4.70 (s, 2H, CH2Ph), 4.68 (d, 1H, J ) 12.9
b
Sp-OCH2 ), 3.46-3.38 (m, 3H, H-5, Sp-CH2N3), 1.96-1.84 (m, 2H,
Sp-CH2), 1.22 (d, 3H, J ) 6.6 Hz, CH3). 13C NMR (75 MHz,
CDCl3): δ 139.1, 138.8, 138.6, 138.3, 137.8 (ArC), 128.9, 128.4, 128.3,
128.0, 128.2, 127.8, 127.6, 127.5, 127.4, 127.2, 126.3 (ArCH), 102.0
(C-1), 101.0 (OCHO), 97.3 (C-1′), 81.7 (C-3), 79.5 (C-3′), 78.2 (C-
4′), 76.1 (C-2′), 74.9 (CH2Ph), 73.1 (CH2Ph), 72.9 (C-4), 72.7 (CH2-
Ph), 72.0 (C-2), 70.8 (CH2Ph), 69.4 (C-6), 66.6 (C-5), 66.4 (C-5′), 66.2
(Sp-OCH2), 48.6 (CH2N3), 29.5 (Sp-CH2), 16.9 (CH3). HR MS (m/
z): calcd for C50H55N3O10Na 880.3785, found 880.3736.
a
b
Hz, CH2Ph), 4.60 (d, 1H, J ) 12.1 Hz, CH2Ph), 4.46 (d, 1H, J )
11.3 Hz, CH2Ph), 4.40 (dd, 1H, J ) 8.8 Hz, J ) 5.2 Hz, H-6a), 4.30
b
(d, 1H, J ) 12.9 Hz, bCH2Ph), 4.24 (d, 1H, J ) 8.0 Hz, H-1), 4.20 (d,
1H, J ) 12.0 Hz, aCH2Ph), 4.14 (d, 1H, J ) 12.0 Hz, bCH2Ph), 4.12-
4.00 (m, 6H, CH2Ph, H-2′, H-3′, H-4′, H-4), 3.98-3.78 (m, 2H, H-5′,
b
Sp-OCH2), 3.52-3.40 (m, 4H, H-5, H-6′, Sp-OCH2 ), 3.32 (dd, 1H,
J ) 10.4 Hz, J ) 2.7 Hz, H-3), 3.26 (t, 2H, J ) 6.9 Hz, CH2N3), 3.15
(dd, 1H, J ) 8.5 Hz, J ) 4.9 Hz, H-6b), 1.96 (s, 3H, CH3CO), 1.86-
1.66 (m, 2H, Sp-CH2). 13C NMR (75 MHz, CDCl3): δ 169.5 (CdO),
3-Azidopropyl 3,6-Di-O-benzyl-2-(2,3,4-tri-O-benzyl-r-L-fucopy-
ranosyl)-4-(2,3,4,6-tetra-O-benzyl-r-D-galactopyranosyl)-â-D-galac-
topyranoside (17). A mixture of 14 (27.6, 0.058 mmol), 16 (46.6 mg,
9
J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002 13043

A R T I C L E S
Arranz-Plaza et al.
0.048 mmol), freshly activated 4 Å molecular sieves, and IDCP (45
mg, 0.096 mmol) was stirred in dry ether/CH2Cl2 (4/1, v/v, 3 mL) at
room temperature for 3 h. The molecular sieves were filtered off, the
filtrate was diluted with CH2Cl2 (10 mL), and the organic phase was
washed with aqueous Na2S2O3 (10 mL) and water (3 × 10 mL). The
organic layer was dried (MgSO4) and concentrated under reduced
pressure. The residue was purified by silica gel column chromatography
(eluent ethyl acetate/hexanes, 1/3, v/v) to give 17 (50 mg, 75% yield).
[R]25D -17.4 (c ) 0.91, CH2Cl2). MALDI-TOF MS (m/z): 1403.7 (M+
+ Na). 1H NMR (300 MHz, CDCl3): δ 7.30-6.90 (m, 45 H, ArCH),
5.61 (d, 1H, J ) 3.6 Hz, H-1′′), 4.94-4.52 (m, 11H, J1′,2′ ) 4.1 Hz, 5
CH2Ph, H-1′), 4.44-4.24 (m, 7H, J1,2 ) 8.0 Hz, H-6a, H-5′′, H-1, 2
CH2Ph), 4.09 (AX, 2H, J ) 11.8 Hz, CH2Ph), 4.06-3.75 (m, 11H,
v) to yield target trisaccharide 6a as its trisaccharide salt (18 mg, 70%
1
yield). [R]26 -7.7 (c ) 0.38, H2O). H NMR (600 MHz, D2O): δ
D
5.10 (d, 1H, J ) 3.9 Hz, H-1′′), 4.81 (d, 1H, J ) 3.9 Hz, H-1′), 4.45
(d, 1H, J ) 7.8 Hz, H-1), 4.24 (t, 1H, J ) 6.3 Hz, H-5′), 4.16 (q, 1H,
J ) 6.6 Hz, H-5′′), 4.73-4.14 (m, 15H, H-6, H-6′, H-4, H-4′, H-4′′,
H-2′, H-2′′, H-3, H-3′, H-3′′, H-5, Sp-OCH2), 3.68 (dd, 1H, J ) 9.9
Hz, J ) 7.7 Hz, H-2), 3.18 (t, 2H, J ) 7.1 Hz, CH2N), 2.14-2.02 (m,
2H, Sp-CH2), 1.28 (d, 3H, J ) 6.6 Hz, CH3). 13C NMR (75 MHz,
CD3OD): δ 103.9 (C-1), 102.6, 102.0 (C-1′, C-1′′), 79.9, 79.2, 75.1,
73.7, 72.7, 71.8, 71.5, 71.2, 70.9, 70.8 (CH), 69.0 (CH2), 68.4 (CH),
62.8 (CH2), 61.1 (CH2), 39.7 (CH2NH2), 33.3 (CH2), 17.0 (CH3). HR
MS (m/z): calcd for C21H39NO15Na 568.2217, found 568.2182.
3-(N-Acetylaminopropyl) 2-O-(r-L-Fucopyranosyl)-â-D-galacto-
pyranoside (4b). A solution of disaccharide 4a (10 mg, 0.026 mmol)
in MeOH (1 mL) was treated with acetic anhydride (0.3 mL) and Et3N
(3.7 µL) and the mixture stirred for 4 h at room temperature. The
reaction was concentrated in vacuo and the residue purified by
iatrobeads column chromatography (eluent chloroform/methanol/water,
12/6/0.5, v/v/v) to yield 4b as a syrup (9.6 mg, 83% yield). [R]26D -87.1
H-2′, H-3′, H-2′′, H-4, H-5′, H-6′a, H-2, H-3, H-4, CH2Ph, Sp-OCH2 ),
a
3.62 (br s, 1H, H-4′), 3.54-3.34 (m, 5H, H-3′′, H-4′′, H-6′b, H-5, Sp-
b
OCH2 ), 3.22 (t, 2H, J ) 6.9 Hz, CH2N3), 3.12 (dd, 1H, J ) 8.0 Hz,
J ) 4.9 Hz, H-6b), 1.70-1.80 (m, 2H, Sp-CH2), 1.10 (d, 3H, J ) 6.3
Hz, CH3). 13C NMR (75 MHz, CDCl3): δ 139.2, 139.0, 138.9, 138.5,
138.4, 138.3, 138.2 (ArC), 128.5, 128.4, 128.3, 128.2, 128.1, 128.0,
127.9, 127.8, 127.6, 127.5, 127.4, 127.3, 126.6 (ArCH), 102.3, 100.1,
97.4 (C-1, C-1′, C-1′′), 83.4, 79.7, 79.1, 78.4, 76.3 (CH), 75.1 (2 CH2-
Ph), 75.0, 73.9 (CH), 73.7 (CH2), 73.6 (CH), 73.4, 73.1, 72.4 (CH2),
72.3 (CH), 72.1 (CH2), 69.5 (CH), 68.5, 68.1 (CH2), 66.5 (CH), 66.1
(CH2), 48.7 (CH2N3), 29.7 (Sp-CH2), 17.0 (CH3). HR MS (m/z): calcd
for C84H91N3O15, 1404.6348, found 1404.6231.
1
(c ) 0.33, H2O). H NMR (600 MHz, CD3OD): δ 5.16 (br s, 1H,
H-1′), 4.34 (d, 1H, J ) 7.3 Hz, H-1), 4.24 (q, 1H, J ) 6.3 Hz, H-5′),
a
3.96-3.90 (m, 1H, Sp-OCH2 ), 3.83 (d, 1H, J ) 2.9 Hz, H-4), 3.79-
3.70 (m, 3H, H-6, H-2′, H-3′), 3.69 (dd, 1H, J ) 9.8 Hz, J ) 2.9 Hz,
b
H-3), 3.67-3.60 (m, 3H, H-4′, H-2, Sp-OCH2 ), 3.51 (t, 1H, J ) 5.9
Hz, H-5), 3.30-3.22 (m, 2H, CH2N), 1.94 (s, 3H, CH3), 1.84-1.74
(m, 2H, Sp-CH2), 1.20 (d, 3H, J ) 6.3 Hz, CH3). 13C NMR (75 MHz,
CD3OD): δ 104.9 (C-1), 103.2 (C-1′), 81.0 (C-2), 77.8 (C-5), 76.8,
74.9, 73.0 (C-2′, C-3, C-4′), 71.6 (C-4), 69.5 (Sp-OCH2), 69.3 (C-5′),
63.8 (C-6), 39.1 (CH2N), 31.8 (CH2), 23.8 (CH3), 18.0 (CH3). HR MS
(m/z): calcd for C17H31NO11Na 448.1795, found 448.1847.
3-Aminopropyl 2-O-(r-L-Fucopyranosyl)-â-D-galactopyranoside
(4a). Pd/C (160 mg) was added to a suspension of disaccharide 13 (80
mg, 0.09 mmol) in a mixture of EtOH/HCl (19/0.08 mL) and the
mixture stirred under an atmosphere of H2 for 4 h. The catalyst was
then filtered off and the filtrate neutralized with Dowex OH- resin
and concentrated under reduced pressure. The residue was purified using
an iatrobeads column (eluent methanol/water/acetic acid, 9/3/0.9, v/v/
v) to yield 4a as its acetate salt (26 mg, 64%). [R]26D -78.3 (c ) 0.34,
H2O). 1H NMR (300 MHz, CD3OD): δ 5.08 (d, 1H, J ) 2.7 Hz, H-1′),
4.38 (d, 1H, J ) 7.4 Hz, H-1), 4.18 (q, 1H, J ) 6.6 Hz, H-5′), 4.05-
3-(N-Acetylaminopropyl) 4-O-(r-D-Galactopyranosyl)â-D-galac-
topyranoside (5b). A solution of 5a (8.08 mg, 0.017 mmol) in MeOH
(1 mL) was treated with acetic anhydride (1.7 µL) and Et3N (2.4 µL)
and the mixture stirred for 4.5 h at room temperature. The reaction
mixture was then concentrated in vacuo and the residue purified by
iatrobeads column chromatography (eluent chloroform/methanol/water,
a
3.95 (m, 1H, Sp-OCH2 ), 3.87-3.50 (m, 10H, H-4, H-6, H-2, H-2′,
b
6/3/0.5, v/v/v) to yield amide 5b as a syrup (7.2 mg, 90% yield). [R]26
+72.8 (c ) 0.23, H2O). H NMR (300 MHz, D2O): δ 4.88 (d, 1H, J
H-3, H-3′, H-4′, H-5, Sp-OCH2 ), 3.16-2.98 (m, 2H, CH2N), 2.00-
D
1.90 (m, 2H, Sp-CH2), 1.22 (d, 3H, J ) 6.6 Hz, CH3). 13C NMR (75
MHz, CD3OD): δ 103.8 (C-1), 102.6 (C-1′), 81.2 (C-2), 76.8, 75.1,
73.7, 71.9, 70.9, 70.2 (C-5, C-2′, C-3, C-3′, C-4′, C-4), 68.7 (OCH2),
68.5 (C-5′), 62.6 (C-6), 39.5 (CH2N), 29.5 (CH2), 17.1 (CH3). HR MS
(m/z): calcd for C15H29NO10Na 406.1689, found 406.1669.
1
) 3.9 Hz, H-1′), 4.37 (d, 1H, J ) 7.7 Hz, H-1), 4.28 (t, 1H, J ) 6.6
Hz, H-5), 3.98-3.60 (m, 12H, H-6, H-6′, H-4, H-4′, H-5′, H-3, H-3′,
H-2′, Sp-OCH2), 3.46 (dd, 1H, J ) 7.7 Hz, J ) 10.0 Hz, H-2), 3.24-
3.14 (m, 2H, CH2N), 1.90 (s, 3H, CH3), 1.81-1.67 (m, 2H, Sp-CH2).
13C NMR (75 MHz, D2O): δ 170.6 (CO), 103.1 (C-10, 100.4 (C-1′),
77.4, 75.3, 72.6, 71.2, 71.1, 69.4, 69.2, 69.0 (C-2, C-2′, C-3, C-3′, C-4,
C-4′, C-5, C-5′), 68.0 (CH2), 60.8, 60.4 (CH2), 36.7 (CH2N), 28.7 (CH2),
22.2 (CH3CO). HR MS (m/z): calcd for C17H31NO12Na 464.1744, found
464.1829.
3-Aminopropyl 4-O-(r-D-Galactopyranosyl)-â-D-galactopyrano-
side (5a). Pd/C (360 mg) was added to a suspension of 16 (180 mg,
0.18 mmol) in EtOH/HCl (38/0.12 mL) and the mixture stirred under
an atmosphere of H2 for 3 h. The catalyst was then filtered off and the
filtrate neutralized with Dowex OH- resin and concentrated under
reduced pressure. The residue was purified using an iatrobeads column
(eluent methanol/water/acetic acid, 9/3/0.1, v/v/v) to yield 5a as the
acetate salt (57 mg, 66% yield). [R]26D +30.2 (c ) 1.5, H2O). 1H NMR
(300 MHz, CD3OD): δ 4.97 (d, 1H, J ) 3.6 Hz, H-1′), 4.34 (d, 1H,
J ) 7.1 Hz, H-1), 4.23 (t, 1H, J ) 5.9 Hz, H-5′), 4.30-3.98 (m, 2H,
3-(N-Acetylaminopropyl) 2-O-(r-L-Fucopyranosyl)-4-O-(r-D-ga-
lactopyranosyl)-â-D-galactopyranoside (6b). A solution of 6a (9.6
mg, 0.017 mmol) in MeOH (1 mL) was treated with acetic anhydride
(1.7 µL) and Et3N (2.4 µL) and the mixture stirred for 24 h at room
temperature. The reaction was then concentrated in vacuo and the
residue purified by iatrobeads column chromatography (eluent chloroform/
a
H-4 or H-4′, Sp-OCH2 ), 4.91 (d, 1H, J ) 2.5 Hz, H-4 or H-4′), 3.88-
b
3.44 (m, 10H, H-6, H-6′, H-2, H-2′, H-3, H-3′, H-5, Sp-OCH2 ), 3.20-
methanol/water, 6/4/0.5, v/v/v) to yield 6b as a syrup (8.4 mg, 81%).
3.02 (m, 2H, CH2N), 2.02-1.90 (m, 2H, Sp-CH2). 13C NMR (75 MHz,
D2O): δ 103.1 (C-1), 100.4 (C-1′), 77.5, 75.4, 72.6, 71.1, 71.0, 69.4,
69.2, 68.9 (C-2, C-2′, C-3, C-3′, C-4, C-4′, C-5, C-5′), 67.9 (CH2),
60.8, 60.4 (CH2), 37.8 (CH2N), 27.1 (CH2), 23.6 (CH3). HR MS (m/z):
calcd for C15H29NO11Na 422.1638, found 422.1601.
1
[R]26 -18.2 (c ) 0.27, H2O). H NMR (500 MHz, D2O): δ 5.20 (d,
D
1H, J ) 3.9 Hz, H-1′′), 4.91 (d, 1H, J ) 3.9 Hz, H-1′), 4.52 (d, 1H, J
) 7.8 Hz, H-1), 4.34 (t, 1H, J ) 6.3 Hz, H-5′), 4.27 (q, 1H, J ) 7.0
Hz, H-5′′), 4.00 (d, 1H, J ) 3.0 Hz, H-4′), 3.97 (d, 1H, J ) 2.6 Hz,
H-4), 3.94-3.65 (m, 9H, H-3′, Sp-OCH2, H-3, H-6, H-2′, H-4′′, H-2′′,
H-6′), 3.57 (dd, 1H, J ) 7.8 Hz, J ) 10.0 Hz, H-2), 3.22 (m, 2H,
CH2N), 1.94 (s, 3H, CH3), 1.80 (m, 2H, Sp-CH2), 1.16 (d, 3H, J )
7.0 Hz, CH3). 13C NMR (125 MHz, D2O): δ 104.4 (C-1), 103.2 (C-
1′), 102.3 (C-1′′), 80.6 (C-4), 79.9 (C-2), 77.6, 75.9, 74.5, 73.5 (C-5′),
72.1, 71.7, 71.6 (C-4′), 71.4, 71.0, 70.3 (Sp-OCH2), 69.5 (C-5′′), 63.2,
62.8 (C-6, C-6′), 39.5 (CH2N), 31.5 (Sp-CH2), 24.8 (CH3), 18.4 (CH3).
HR MS (m/z): calcd for C23H41NO16Na 610.2323, found 610.2433.
3-Aminopropyl 2-O-(r-L-Fucopyranosyl)-4-O-(r-D-galactopyra-
nosyl)-â-D-galactopyranoside (6a). Pd/C (120 mg) was added to a
solution of 17 (60 mg, 0.04 mmol) in EtOH/HCl (19/0.05 mL) and the
mixture stirred under an atmosphere of H2 for 5 h. The catalyst was
then filtered off and the filtrate neutralized with Dowex OH- resin
and concentrated under reduced pressure. The residue was purified using
an iatrobeads column (eluent methanol/water/acetic acid, 9/1/0.1, v/v/
9
13044 J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002

Block to Polyspermy in Xenopus laevis
A R T I C L E S
General Procedure for the Preparation of Glycopolymers 19-
24. pNAS was synthesized by an AIBN-initiated radical polymerization
of N-(acryloyloxy)succinimide dissolved in dry benzene as previously
described,39 and the same batch of material was used as the precursor
for the synthesis of all sugar-bearing polymers. For example, the
polymer 23c containing a 17% loading of saccharide was prepared by
suspending the 3-aminopropyl saccharide 3 (2.2 mg), pNAS (7.3 mg),
and Et3N (5 µL) in dry DMF (1 mL) and stirring the mixture for 22 h
at room temperature under argon and for a further 6 h at 65 °C. The
reaction was allowed to cool to room temperature and quenched with
concentrated aqueous ammonia (0.1 mL). Exhaustive dialysis of the
reaction mixture against distilled water with a molecular weight cutoff
of ∼14000 followed by freeze-drying gave the saccharide-bearing
polymer 19c of 17% loading as a white, fluffy solid. Polymers 19a
and 19b (with 5% and 10% sugar content, respectively) were obtained
from glycoside 3 in the same manner except that the amounts of pNAS
were adjusted accordingly to give the desired final loading in each case.
The following are selected NMR spectroscopic data. (Polymer 19a)
1H NMR (600 MHz, D2O): δ 4.90 (br s, H-1′), 4.40 (br s, H-1).
captoethanol, pH 8.9 (three 5 min washes) until the egg jelly was
solubilized. The egg jelly was then decanted, dialyzed against TCS,
and lyophilized prior to conjugation with BAP.
Thiolation of the JCP. The JCP (∼10 mg) was dissolved in 4 M
guanidine hydrochloride, 0.1 M Tris, pH 8, and treated for 1 h at 37
°C using 10 mM DTT. Iodoacetamide was then added to the reduced
JCP at a final concentration of 50 mM and the resulting mixture left
for 1 h at 37 °C. The thus formed carboxymethylated JCP (RCM-JCP)
was centrifuged at 15000g for 1 h to remove particulates and then
desalted twice against PBS with 5 mM EDTA, pH 8.0, in preparation
for thiolation. SATA (3.1 mg) was dissolved in DMSO (100 µL), and
30 µL of this stock was added to RCM-JCP in PBS (3 mL) containing
5 mM EDTA, pH 8 (final concentration of SATA 1.3 mM). After 2 h
at 25 °C, excess SATA was removed by dialyzing the reaction mixture
three times against PBS with 5 mM EDTA. The JCP-SATA conjugate
(1 mL, 2.6 mg of protein) thus obtained was treated with 0.5 M
hydroxylamine in PBS with 25 mM EDTA, pH 7.5, for 2 h and then
desalted against PBS with 25 mM EDTA, pH 7.2, to prepare it for
conjugation to derivatized BAP.
Conjugation of the JCP to BAP. BAP (5 mg) was desalted three
times against PBS with 5 mM EDTA, pH 8. To this BAP solution was
added SMCC in DMSO (60 mM; 50 µL), and the resulting solution
was left for 1 h at 25 °C. The mixture was then desalted twice against
PBS with 5 mM EDTA, pH 7.2, and the resulting BAP-SMCC added
to the deacetylated JCP-SATA conjugate. The reaction was left for 2
h at 25 °C and overnight at 4 °C and then dialyzed against two buffer
changes of TBS with 1 mM MgCl2. The activity of the final conjugate
was confirmed using saturation binding assays.
Competition ELLA (Lectin on Plates). Purified XL35 (20 µg) in
50 mM sodium bicarbonate, pH 9.4, was coated onto Corning half-
area ELISA plates for 1 h at 25 °C, and the plates were blocked using
TCS with 1% BSA and 0.1% Tween 20 as described above. The
blocking buffer was discarded and replaced with fresh blocking buffer
(20 µL). Inhibitors dissolved in blocking buffer were added to the first
column and serially diluted across the microtiter plate. The JCP-BAP
conjugate (20 µL of a 1/2500 stock in blocking buffer) was added to
each well and allowed to incubate overnight at 4 °C. The plates were
washed three times using TCS and developed at 37 °C using
p-nitrophenol in 0.2 M Tris, pH 10.2 (120 µL), before being read at
405 nm.
Saturation Binding Assays. Assays were performed in a fashion
similar to that of the corresponding inhibition assays with the exception
that no inhibitor was added. Instead, saturating concentrations of JCP-
BAP were added to the first column of wells and serially diluted across
the microtiter plate. Following incubation of the primary conjugate,
the plates were washed three times and developed.
1
(Polymer 19b) H NMR (600 MHz, D2O): δ 4.96 (br s, H-1′), 4.46
1
(br s, H-1). (Polymer 19c) H NMR (600 MHz, D2O): δ 4.97 (br s,
H-1′), 4.45 (br s, H-1). 13C NMR (125 MHz, D2O): δ 98.3 (C-1′),
1
102.5 (C-1). (Polymer 20) H NMR (600 MHz, D2O): δ 4.40 (br s,
H-1), 4.00-3.85 (m), 3.80-3.45 (m). (Polymer 21) 1H NMR (600 MHz,
D2O): δ 4.89 (br s, H-1). (Polymer 22) 1H NMR (600 MHz, D2O): δ
5.22 (br s, H-1′), 4.5 (br s), 4.3 (br s, H-5′). 13C NMR (150 MHz,
D2O): δ 104.5 (C-1), 102.4 (C-1′), 102.9. (Polymer 23) 1H NMR (600
MHz, D2O): δ 4.9 (br s, H-1′), 4.45 (br s, H-5′), 4.37 (br s, H-1). 13
C
NMR (150 MHz, D2O): δ 105.2 (C-1), 102.9 (C-1′). (Polymer 24) 1H
NMR (600 MHz, D2O): δ 5.25 (br s, H-1′′), 4.95 (br s, H-1′), 4.52 (br
s, H-1), 4.52 (br s, H-1), 4.37 (br s, H-5′), 4.27, (br s, H-5′′).
pNAS-Derived Polyacid and Polyamide Controls. These com-
pounds were obtained as previously described, from the same batch of
pNAS as the glycopolymers, purified by exhaustive dialysis and
1
characterized by H NMR spectroscopy (data not shown).38-40
General Information. Biochemistry. SATA, 4-(N-maleimidom-
ethyl)cyclohexane-1-carboxylate (SMCC), N-Hydroxysuccinimide-
biotin (NHS-biotin), 1-ethyl-3-(dimethylaminopropyl)carbodiimide
hydrochloride (EDC), NHS, sulfo-NHS-LC-biotin, and Streptavidin-
HRP were purchased from Pierce and used without further purification.
All chromatography solvents were HPLC grade and were filtered using
a 0.45 µm filter prior to use. Commercially available saccharides were
obtained from Sigma, and poly(acrylic acid) and dextran standards were
purchased from Fluka. BAP was purchased from Worthington and
dialyzed into phosphate-buffered saline (PBS) to which 5 mM EDTA
had been added.
Purification of Lectin. XL35 was purified from oocytes as
previously described53 except that an anion-exchange step was added
prior to melibiose affinity chromatography. Briefly, mature female
xenopi were anesthetized using 3-aminobenzoic acid ethyl ester
methanesulfonate (MS-222, Sigma), and the oocytes were obtained
surgically and homogenized in TCS prechilled to 4 °C. Cold acetone
(9 volumes) was added to the initial homogenate, and the pellet was
captured by filtration and then washed three times with cold acetone
to remove residual lipids. The delipidated pellet was resuspended in
0.3 M galactose (in TCS) and centrifuged at 10000g for 20 min. The
clarified supernatant was dialyzed against 50 mM Tris, pH 8.5, prior
to being applied to a 2.6 × 10 cm DEAE-650M column (TosoHaas)
equilibrated in the same buffer. XL35 was eluted using 50 mM Tris,
1.2 M NaCl, pH 8.5, and the protein fractions were pooled and applied
to a 3 mL melibiose affinity column equilibrated in TCS. The melibiose
column was washed extensively with TCS, and purified XL35 eluted
with TBS + 50 mM EDTA.
SPR Analyses. SPR experiments were conducted on a BIAcore 3000
apparatus (BIACORE AB, Uppsala, Sweden). Lectin was coupled to a
CM5 chip following the manufacturer’s instructions. Purified XL35
(100 µg) was dialyzed against 50 mM sodium acetate, pH 5, for 2 h.
Immediately prior to coupling, the dialyzed XL35 was diluted 10-fold
in 10 mM sodium acetate, pH 4. EDC and NHS were weighed out and
suspended in water immediately before use at final concentrations of
400 and 100 mM, respectively. The coupling reaction was carried out
using the BIAcore control software, and 15000 RUs was selected as
the target level using the “Aim for Immobilization Level” program.
Glycopolymer 19c and a glycopolymer loaded with 17% N-acetylglu-
cosamine (negative control) were separately dissolved in running buffer
(10 mM Hepes, 10 mM CaCl2, 150 mM NaCl, with 0.005% Tween
20, pH 7.2) at a concentration of 0.625 mg/mL. The flow rate was set
to 15 µL/min, and each sample (50 µL) was injected using the “Kinject”
command.
Purification of the JCP. Total JCP was obtained as previously
described54 by treating oviposited eggs with TBS + 50 mM â-mer-
Kinetic and Equilibrium Analyses. Serial dilutions of the glyco-
polymer 19c were injected over a chip surface containing 5000 RUs
of XL35 prepared as described above. Kinetic analyses were conducted
at a flow rate of 15 µL/min to minimize mass transport effects although
(53) Hedrick, J. L.; Hardy., D. M. Methods Cell Biol. 1991, 36, 231.
(54) Yurewicz, E. C.; Hedrick, J. L. Biol. Reprod. 1973, 9, 72-73.
9
J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002 13045

A R T I C L E S
Arranz-Plaza et al.
mass transport limitations were not observed even at lower flow rates.
Equilibrium analyses were performed using the same surface but at a
flow rate of 4 µL/min.
the JCP obtained by compositional analysis,55 standards were prepared
by using the appropriate molar combination of monosaccharides
suspended in blocking buffer (TCS with 1% BSA and 0.1% Tween
20) as previously described.56 Briefly, a 5% phenol (w/w) (10 µL)
solution in deionized water was added to individual wells of a microtiter
plate followed by the addition of sample (10 µL) or standard.
Concentrated H2SO4 (150 µL) was added to each of the wells, and the
contents were immediately mixed (five times). Microtiter plates were
then placed on a shaker for 10 min prior to the absorbance being read
at 495 nm. The assays were performed in duplicate, and a linear
regression was used to obtain the sugar concentration. The same
procedure was used to determine the sugar content of the glycopoly-
mers.
SPR Analysis of Monovalent Ligands. Disaccharide 3 was bioti-
nylated, using sulfo-NHS-LC-biotin and the product purified by
preparative reversed-phase chromatography using a linear gradient of
water/methanol (0-100% in 50 min). The absorbance was monitored
at 206 nm, and fractions were initially screened using MALDI-TOF
prior to confirmation by NMR spectroscopy (data not shown). Strepta-
vidin (10000 RU) was coupled to a CM5 chip using the NHS and EDC
as described above, but the coupling was carried out by manually
injecting the NHS/EDC mixture for 40 min followed by a 40 min
injection of streptavidin (25 µg/mL in 10 mM sodium acetate, pH 4.0).
Any unreacted NHS esters on the chip were then blocked using an
injection of 1 M ethanolamine, pH 8. Biotinylated 3 was then injected
over the chip surface and was immobilized by the coupled streptavidin.
A calibration curve was generated for XL35 by diluting it serially (200
µg/mL starting concentration), and a concentration of 75 µg/mL was
used for solution competition experiments. Increasing concentrations
of melibiose (up to 10 mM final concentration) were used to compete
XL35 from the streptavidin-melibiose surface.
Acknowledgment. This work was supported by the NIH
Resource Center for Biomedical Complex Carbohydrates (Grant
P41-RR-5351). We thank Dr. Steven Levery for performing the
compositional analysis and Dr. Kumar Koli for performing HR
MS. A.S. thanks the CNRS (France) for partial financial support.
Supporting Information Available: Procedures for the prepa-
1
ration of compounds 1-3, H NMR spectra of compounds
Molecular Weight Determination of the Glycopolymers. The size
of the glycopolymers was determined using gel filtration on a Sephacryl
S-400 column (1.5 × 50 cm) equilibrated in 200 mM NaCl with 10
mM H3PO4 titrated to pH 6.0 using NaOH. The flow rate for all analyses
was 2.5 mL/min, and detection was performed at 206 nm. The system
was calibrated using a combination of commercially available polyacid
and dextran standards.
1-17, and Cheng-Prusoff correction for calculating KI (PDF).
This material is available free of charge via the Internet at
http://pubs.acs.org.
JA020536F
(55) Fan, J. Q.; Namiki, K.; Matsuoka, K.; Lee, Y. C. Anal. Biochem. 1994,
219, 375.
(56) Saha, S. K.; Brewer, C. F. Carbohydr. Res. 1994, 254, 157-167.
Phenol-Sulfuric Assay for Determination of the Sugar Content
of the JCP and the Glycopolymers. Using the molar composition of
9
13046 J. AM. CHEM. SOC. VOL. 124, NO. 44, 2002