Synthesis and molecular recognition studies of the HNK-1 trisaccharide and related oligosaccharides. the specificity of monoclonal anti-HNK-1 antibodies as assessed by surface plasmon resonance and STD NMR

Article abstract of DOI:10.1021/ja2083015

The human natural killer cell carbohydrate, HNK-1, plays function-conducive roles in peripheral nerve regeneration and synaptic plasticity. It is also the target of autoantibodies in polyneuropathies. It is thus important to synthesize structurally relate

Full text of DOI:10.1021/ja2083015

ARTICLE
pubs.acs.org/JACS
Synthesis and Molecular Recognition Studies of the HNK-1
Trisaccharide and Related Oligosaccharides. The Specificity of
Monoclonal Anti-HNK-1 Antibodies as Assessed by Surface Plasmon
Resonance and STD NMR
Yury E. Tsvetkov,† Monika Burg-Roderfeld,‡ Gabriele Loers,§ Ana Ardꢀa,^ Elena V. Sukhova,†
Elena A. Khatuntseva,† Alexey A. Grachev,† Alexander O. Chizhov,† Hans-Christian Siebert,‡
Melitta Schachner,§ Jesꢀus Jimꢀenez-Barbero,^ and Nikolay E. Nifantiev*,†
†Laboratory of Glycoconjugate Chemistry, N.D. Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences,
Leninsky prospect 47, 119991 Moscow, Russia
‡Institut f€ur Biochemie und Endokrinologie, Veterin€armedizinische Fakult€at, Justus-Liebig-Universit€at Giessen, Frankfurter Strasse 100,
35392 Giessen, Germany
§Zentrum f€ur Molekulare Neurobiologie, Universit€at Hamburg, Martinistrasse 52, 20246 Hamburg, Germany
^Chemical and Physical Biology, Center of Biological Research, CSIC, Ramiro de Maeztu 9, 28040 Madrid, Spain
S Supporting Information
b
ABSTRACT: The human natural killer cell carbohydrate, HNK-1,
plays function-conducive roles in peripheral nerve regeneration
and synaptic plasticity. It is also the target of autoantibodies in
polyneuropathies. It is thus important to synthesize structurally
related HNK-1 carbohydrates for optimizing its function-
conducive roles, and for diagnosis and neutralization of auto-
antibodies in the fatal GuillainꢀBarrꢀe syndrome. As a first step
toward these goals, we have synthesized several HNK-1 carbo-
hydrate derivatives to assess the specificity of monoclonal
HNK-1 antibodies from rodents: 2-aminoethyl glycosides of
selectively O-sulfated trisaccharide corresponding to the HNK-1 antigen, its nonsulfated analogue, and modified structures
containing 3-O-fucosyl or 6-O-sulfo substituents in the N-acetylglucosamine residues. These were converted, together with several
related oligosaccharides, into biotin-tagged probes to analyze the precise carbohydrate specificity of two anti-HNK-1 antibodies by
surface plasmon resonance that revealed a crucial role of the glucuronic acid in antibody binding. The contribution of the different
oligosaccharide moieties in the interaction was shown by saturation transfer difference (STD) NMR of the complex consisting of the
HNK-1 pentasaccharide and the HNK-1 412 antibody.
1. INTRODUCTION
Precise interaction of different cell types is essential to
maintain neuronal network functions. During these pro-
cesses, carbohydrate structures expressed on proteins repre-
sent recognition sites and cause structural diversity of carrier
molecules, resulting in tight regulation of cellꢀcell interac-
tion, recognition, and migration. One of the most character-
istic carbohydrate epitopes of the nervous system are the
Figure 1. Structure of HNK-1 antigenic trisaccharide 1 and its 6-O-
sulfated and 3-O-fucosylated derivatives 2 and 3.
HNK-1 (human natural killer-1) cell glycans. These oligo-
saccharides include a unique structure 1 (Figure 1) composed
of a sulfated glucuronic acid attached to the nonreducing
terminus of an N-acetyllactosamine residue.1,2 This structure
neuropathies, including the GuillainꢀBarrꢀe syndrome (for a
is involved in preferential motor reinnervation of the injured
review, see ref 9).
femoral nerve in mice3ꢀ7 and nonhuman primates.8 It binds
to the GABAB receptor, thereby affecting synaptic activity
and plasticity. The HNK-1 carbohydrate is the target of auto-
immune antibodies that lead to autoimmune-based peripheral
Received: September 2, 2011
Published: November 16, 2011
r
2011 American Chemical Society
426
dx.doi.org/10.1021/ja2083015 J. Am. Chem. Soc. 2012, 134, 426–435
|

Journal of the American Chemical Society
ARTICLE
Development and characterization of HNK-1-related struc-
tures have a tremendous impact on our knowledge on the
regulation of nervous system functions. This also applies for
modified HNK-1-like oligosaccharide structures as potential
modulators of regeneration and synaptic activity. Also, these
structures will help to define against which carbohydrate epitopes
antibodies are specifically directed in polyneuropathies and
thereby possibly allow performing diagnosis of distinct aspects
of the disease. Precise knowledge of the epitope variation in
different types of polyneuropathies will be decisive in generating
carbohydrates that will allow neutralization of autoantibodies for
therapy to be carried out.
Scheme 1. Attempts at the Synthesis of Trisaccharide 10 by
the [1 + 2] Schemea
Several syntheses of HNK-1 oligosaccharides have been de-
scribed including glycolipids with penta- and heptasaccharide
carbohydrate chains,10ꢀ13 trisaccharide octyl glycoside,14 di- and
pentasaccharides propyl15 and 2-aminoethyl16 glycosides. On the
other hand, no modified analogues of HNK-1 oligosaccharides,
natural or unnatural, have been synthesized thus far, except for our
synthesis of an analogue of 1 containing 3,6-di-O-sulfoglucose
instead of 3-O-sulfoglucuronic acid,17 and the very recent chemo-
bacterial synthesis of allyl 3-O-sulfoglucuronyl-30-lactoside.18
In this contribution, we report the synthesis of 2-aminoethyl
glycosides of the parent HNK-1 trisaccharide 1, its nonsulfated
analogue, and two structures modified in the N-acetylglucosa-
mine residue: 6-O-sulfated trisaccharide 2 and 3-O-α-L-fucosyl-
ated tetrasaccharide 3. The former disulfated trisaccharide was
shown to be a constituent of carbohydrate chains of P0
glycoprotein.19 The latter structure has not been found in nature
yet. Taking into account the important role of the 3-O-α-L-
fucosylation of lactosamine in various biological phenomena
such as adhesion,20 development,21 cellular differentiation,22
oncotransformation,23 and some others, the 3-O-α-L-fucosylated
tetrasaccharide could also be present in HNK-1 antigenic
glycoproteins. The oligosaccharides described here and some
related oligosaccharides synthesized previously16,17,24 (Table 1)
were transformed into biotin-tagged molecular probes and used
in molecular recognition studies to assess the specificity of
autoantibodies, which are involved in the development of acute
immune-mediated neuropathies and chronic immune-mediated
polyneuropathies, by surface plasmon resonance. The structural
requirements for the carbohydrate binding to some HNK-1
antibodies were qualitatively estimated by ELISA using synthetic
HNK-1 glycolipids,25,26 but no quantitative measurements of
these interactions have been made so far. To reveal the topology
of the carbohydrate binding, the saturation transfer difference
(STD) NMR studies of a complex of the HNK-1 pentasaccharide16
with the HNK-1 412 antibody were carried out as well.
a Reagents and conditions: (a) NIS, TfOH, MS 4 Å, CH2Cl2, ꢀ30 ꢀC,
70%; (b) (H2N)2CS, sym-collidine, MeOH, reflux, 90%; (c) BF3 Et2O,
3
MS AW-300, CH2Cl2, 0 ꢀC.
was detected upon the glycosylation of 7 with thioglycoside
donor 8, while imidate 9 provided 10 in an unsatisfactory yield of
about 10%. Therefore, the [1 + 2] synthetic scheme was
abandoned, and we focused our attention on the alternative
[2 + 1] approach to prepare the target structures.
Glucosamine acceptor 5 is suitable for the preparation of the
monosulfated structure 1, whereas a glucosamine derivative with
orthogonal protecting groups in the positions 3 and 6 was
necessary as a glycosyl acceptor for the rational synthesis of
the structures 2 and 3. For this aim, the known benzylidene
acetal 1117 was converted into diol 12, and the diol was
further transformed into 6-tosylate 13 by selective tosylation.
(Scheme 2). The subsequent treatment of tosylate 13 with
p-methoxyphenol in DMF in the presence of NaH afforded
necessary glucosamine acceptor 14 with the orthogonal protect-
ing group pattern.
Two different galactose acceptors 1529 and 1630 were exam-
ined to synthesize a glucuronosyl-(1f3)-galactose donor block
(Scheme 3). The successful glycosylation of thioglycoside 15
would lead, without any intermediate transformations, to the
necessary disaccharide donor 17. However, the BF3 E2O-
3
promoted reaction of 15 with imidate 9 produced target 17 in
minor yield due to the predominant transfer of the SEt group to
the uronic acid donor and some other side processes. The glycosyla-
tion of allyl galactoside 16 with the same donor 9 required the
thorough optimization of reaction conditions with respect to
2. RESULTS AND DISCUSSION
promoter (TMSOTf, BF3 E2O), solvent (CH2Cl2, toluene), the
3
molecular sieve type, and temperature. Under optimal con-
ditions found (see Scheme 3), disaccharide 18 was obtained in
64% yield.
2.1. Synthesis of Oligosaccharide Ligands Related to
HNK-1 Antigens. Two approaches to the assembly of the key
trisaccharide sequence 1 are possible, (1) by chain elongation
from the reducing end ([1 + 2] scheme) and (2) from the
nonreducing end ([2 + 1] scheme). We explored first the [1 + 2]
As the benzylidene group in 18 may be unstable under de-
allylation conditions, it was replaced by acid-stable benzoates to
provide completely acylated disaccharide 19. Removal of the
anomeric allyl group produced hemiacetal 20, which was then
converted to imidate 21. Surprisingly, the reaction of 21 with
acceptor 5 afforded no trisaccharide 23; only gradual destruction
of the imidate was observed, while acceptor 5 was nearly
quantitatively recovered from the reaction mixture. Then imidate
21 was transformed in thioglycoside 22; its subsequent coupling
synthetic scheme (Scheme 1). The NIS TfOH-promoted gly-
3
cosylation of 2-azidoethyl glycoside 517 with thiogalactoside 427
afforded disaccharide 6 in good yield; removal of the chloroacetyl
group from 6 with thiourea provided disaccharide acceptor 7.
To introduce a glucuronic acid residue into 7, two glucuronosyl
donors 828 and 928 containing a selectively removable 3-O-acetyl
group were tested. However, no formation of trisaccharide 10
427
dx.doi.org/10.1021/ja2083015 |J. Am. Chem. Soc. 2012, 134, 426–435

Journal of the American Chemical Society
ARTICLE
Scheme 2. Synthesis of Glycosyl Acceptor 14a
C-3 of the uronic acid residue (δ 72.1 f 78.5 for 28; δ 73.5 f
78.6 for 29) and for C-6 of N-acetylglucosamine (δ 61.2 f 66.0)
as compared to those of the parent hydroxyl compounds.
Simultaneous reduction of the azide in the aglycon and removal
of the benzyl group(s) by catalytic hydrogenolysis of 28 and 29
failed. After the fast reduction of the azido group, the hydro-
genolysis of benzyl groups almost stopped, apparently due to
poisoning of the catalyst by the amine formed.16 Moreover, a
considerable loss of the sulfate groups was observed on pro-
longed hydrogenolysis. To overcome this difficulty, two-step
hydrogenolysis was used. At the first step, the amino group
arising on azide reduction was blocked with a trifluoroacetyl
group, and then hydrogenolysis was continued until complete
debenzylation. Removal of the single benzyl group from 29 was
thus achieved practically without competitive desulfation to give
31 in 77% yield. In the case of 28 having two benzyl groups,
desulfation was much more pronounced, and buffering of the
reaction solution with NaOAc was necessary to obtain debenzyl-
ated product 30 in reasonable yield. Following two-step saponi-
fication of 30 and 31 smoothly gave target free sulfates 32 and 33
corresponding to the structures 1 and 2.
a Reagents and conditions: (a) 80% aq AcOH, 40 ꢀC, 86%; (b) TsCl, Py,
0 ꢀC f rt, 75%; (c) p-MeOC6H4OH, NaH, DMF, 60 ꢀC, 91%.
Scheme 3. Synthesis of Trisaccharides 23 and 24a
To avoid the loss of the sulfate upon catalytic hydrogenolysis,
an alternative reaction sequence in which benzyl groups were
replaced by benzoates before sulfation was also explored. The
catalytic reduction of the azido group in 23 followed by N-
trifluoroacetylation produced 34. The latter was subjected to
conventional catalytic hydrogenolysis, and the formed diol 35
was benzoylated with the formation of 36. Benzoic anhydride
was used at the latter step to avoid competitive N-benzoylation.
Further transformations of monoacetate 36 included selective
acidic deacetylation (f 37), sulfation (f 38), and final sapo-
nification to 32. Nonsulfated trisaccharide 39, necessary for the
evaluation of carbohydrate specificity of anti-HNK-1 antibodies,
was obtained by the saponification of diol 35.
Trisaccharide 24 was used as the starting compound for the
synthesis of the tetrasaccharide structure 3 (Scheme 5). The
reduction of the spacer azido group and debenzylation as
described above provided monohydroxyl trisaccharide acceptor
40, which was coupled with thiofucoside 41, obtained by
conventional benzoylation of the corresponding 3,4-diol,34 to
yield tetrasaccharide 42. The α-configuration of the fucoside
bond was confirmed by the corresponding coupling constant
a Reagents and conditions: (a) BF3 E2O, MS AW-300, toluene, 0 ꢀC f
3
4 ꢀC, 64%; (b) PPTS, 90% aq CH3CN, 80 ꢀC; (c) BzCl, pyridine, 92%,
two steps; (d) PdCl2, AcONa, 95% aq AcOH, 64%; (e) CCl3CN, DBU,
CH2Cl2, 75%; (f) EtSH, TMSOTf, MS 4 Å, CH2Cl2, 98%; (g) NIS,
TfOH, MS AW-300, CH2Cl2, ꢀ20 ꢀC f ꢀ10 ꢀC, 69% for 23, 60% for
24; (h) NIS, TfOH, MS 4 Å, CH2Cl2, ꢀ20 ꢀC f ꢀ10 ꢀC, 70%.
1
value J1,2 (3.4 Hz) in the H NMR spectrum of 42. As the
removal of benzyl and p-methoxyphenyl groups from sulfated
tetrasaccharide might be accompanied by desulfation (vide
supra), these groups were replaced by benzoates as described
above for 23. The catalytic hydrogenolysis of 42 followed by the
treatment with CAN afforded diol 43, which was benzoylated to
give 44. The acidic deacetylation of 44 provided derivative 45. It
is noteworthy that the hydroxyl group in 45 displayed much
lower reactivity toward sulfation than did the similar hydroxyl
groups in trisaccharides 26 and 27. Thus, the formation of only a
minor amount of sulfate 46 was detected under the conditions
used for the preparation of 28 and 29 (10-fold excess of the
with acceptors 5 and 14 in the presence of NIS, TfOH, and MS
AW-300 gave desired trisaccharides 23 and 24 in good yields. It is
noteworthy that the replacement in the reaction with 5 of non-
basic MS AW-300 with basic MS 4 Å resulted in the exclusive
formation of the glycosylation product at the acetamido group,
namely, imidate 25. The formation of similar imidates upon
glycosylation of N-acetylglucosamine acceptors has been reported
earlier.31,32
The only O-acetyl group in 23 was selectively removed by
mild acidic methanolysis;33 the splitting of the p-methoxyphenyl
group by ceric ammonium nitrate (CAN) preceded acidic
deacetylation in the case of 24 (Scheme 4). Monohydroxyl (26)
and and dihydroxyl (27) derivatives obtained thereby were
SO3 pyridine complex, rt, 2 h); complete conversion of 45 to 46
3
needed a larger excess of the sulfating reagent (50 equiv) and
prolonged reaction time (42 h). The saponification of 46 as
described above produced target sulfate 47.
2-Aminoethyl 3-O-sulfoglucuronide 51 was also prepared
(Scheme 6) to reveal a minimum carbohydrate element recog-
treated with the pyridine SO3 complex to provide the respective
3
sulfates 28 and 29. The location of the sulfate groups in 28 and 29
was confirmed by characteristic downfield shifts of the signals for
nizable by anti-HNK-1 antibodies. The BF3 Et2O-catalyzed
3
coupling of imidate 9 with 2-trifluoroacetamidoethanol afforded
428
dx.doi.org/10.1021/ja2083015 |J. Am. Chem. Soc. 2012, 134, 426–435

Journal of the American Chemical Society
ARTICLE
Scheme 4. Synthesis of Sulfated Trisaccharides 32 and 33a
a Reagents and conditions: (a) 6% anhydrous HCl in MeOH, 86% for 26, 85% for 37, 71% for 27, two steps; (b) CAN, aq CH3CN, 0 ꢀC; (c) complex
SO3 pyridine, DMF, 2 h, 86% for 28, 81% for 29, 96% for 38; (d) H2, PdO/C, MeOH, AcOH, 51% for 35, three steps; (e) CF3CO2Et, Et3N, MeOH;
3
(f) H2, PdO/C, MeOH, NaOAc, 45% for 30, three steps; (g) H2, PdO/C, MeOH, 77% for 31, three steps; (h) LiOH, aq THF, ꢀ10 ꢀC; (i) NaOH, aq
MeOH, 72% for 32, 83% for 33, 82% for 39; (j) Bz2O, DMAP, pyridine, 40 ꢀC, 96%.
Scheme 5. Synthesis of Sulfated Tetrasaccharide 47a
a Reagents and conditions: (a) H2, PdO/C, MeOH, AcOH; (b) CF3CO2Et, Et3N, MeOH; (c) H2, PdO/C, MeOH, 57%, three steps; (d) NIS, TfOH,
MS 4 Å, CH2Cl2, ꢀ20 ꢀC f ꢀ10 ꢀC, 52%; (e) CAN, aq CH3CN, 0 ꢀC, 64%, two steps; (f) BzCl, pyridine, ꢀ10 ꢀC, 95%; (g) 6% anhydrous HCl in
MeOH, 4 ꢀC, 75%; (h) complex SO3 pyridine, DMF pyridine (3:1), 42 h, 82%; (i) LiOH, aq THF, ꢀ10 ꢀC; (j) NaOH, aq MeOH, 92%.
3
3
glucuronide 48, which was converted to 51 by deacetylation,
sulfation, and saponification as described above for trisaccharide
36 and tetrasaccharide 44.
the necessary accuracy of measurements.35,41 Several methods
for immobilization of carbohydrate ligands to the surface have
been reported,37,38,41,42 but the application of biotin-tagged
oligosaccharides39,40,43,44 seems to be the most convenient be-
cause of the market appearance of streptavidin SPR sensor chips.
We have prepared a series of biotin-tagged HNK-1-related
oligosaccharides in which the carbohydrate and biotin moieties
are connected via a hydrophilic and flexible hexa(ethylene glycol)
spacer. The spacer allows one to avoid undesirable hydro-
phobic interactions and ensures an optimal orientation of a carbo-
hydrate ligand during the recognition process. The synthesis of a
2.2. Synthesis of Biotin-Tagged Oligosaccharide Ligands.
Surface plasmon resonance has become a powerful tool for
evaluation and quantification of the carbohydrateꢀprotein inter-
action (e.g., refs 35ꢀ40). Carbohydrate molecules are usually
immobilized to the surface, and a solution of protein is allowed
to run over the carbohydrates. The opposite case when low-
molecular-mass carbohydrates (normally 200ꢀ2000 D) run over
immobilized protein is less common, because it may not provide
429
dx.doi.org/10.1021/ja2083015 |J. Am. Chem. Soc. 2012, 134, 426–435

Journal of the American Chemical Society
ARTICLE
Table 1. List of the Parent Saccharide 2-Aminoethyl Glycosides and Biotinylated Ligands Therefrom (see Scheme 7)
parent 2-aminoethyl
glycoside
biotinylated ligand
R
of type 61
3-O-SO3ꢀ-β-D-GlcAp-(1f3)-β-D-Galp-(1f4)-β-D-GlcNAcp-(1f
3-O-SO3ꢀ-β-D-GlcAp-(1f3)-β-D-Galp-(1f4)-(6-O-SO3ꢀ-β-D-GlcNAcp)-(1f
3-O-SO3ꢀ-β-D-GlcAp-(1f3)-β-D-Galp-(1f
3-O-SO3ꢀ-β-D-GlcAp-(1f3)-β-D-Galp-(1f4)-[(α-L-Fucp-(1f3)]-β-D-GlcNAcp-(1f
3-O-SO3ꢀ-β-D-GlcAp-(1f3)-β-D-Galp-(1f4)-β-D-GlcNAcp-(1f3)-β-D-Galp-(1f4)-β-D-Glcp-(1f
β-D-Galp-(1f4)-β-D-GlcNAcp-(1f3)-β-D-Galp-(1f4)-β-D-Glcp-(1f
β-D-GlcAp-(1f3)-β-D-Galp-(1f4)-β-D-GlcNAcp-(1f
32
33
62
47
63
64
39
51
65
66
67
68
69
70
71
72
73
74
3-O-SO3ꢀ-β-D-GlcAp-(1f
3,6-di-O-(SO3ꢀ)2-β-D-Glcp-(1f3)-β-D-Galp-(1f4)-β-D-GlcNAcp-(1f
Scheme 6. Synthesis of 3-O-Sulfoglucuronide 51a
Scheme 7. Synthesis of Biotin-Tagged Oligosaccharidesa
a Reagents and conditions: (a) BF3 Et2O, CH2Cl2, ꢀ15 ꢀC, 62%;
3
(b) 6% anhydrous HCl in MeOH, 84%; (c) complex SO3 pyridine,
3
pyridine, 91%; (d) LiOH, aq THF, ꢀ10 ꢀC; (e) NaOH, aq MeOH, 84%.
reagent for the introduction of the biotin tag is depicted in the
Scheme 7.
A reaction sequence similar to that described for tetra(ethylene
glycol)45 was used to functionalize hexa(ethylene glycol) 52. The
base-catalyzed addition of 52 to tert-butyl acrylate furnished
monoester 53. Then the hydroxyl group was mesylated (f 54)
followed by substitution with azide (f 55). The catalytic reduc-
tion of the azido group provided amino ester 56, which was
acylated with biotin active ester 57 to afford 58. The removalofthe
tert-butyl group from 58 and the subsequent treatment of acid 59
formed with pentafluorophenyl trifluoroacetate resulted in the
formation of pentafluorophenyl ester 60.
a Reagents and conditions: (a) CH2dCHCO2Bu-t, NaH, THF, 78%;
(b) MsCl, Et3N, CH2Cl2; (c) NaN3, DMF, 75%, two steps; (d) H2,
Pd(OH)2/C, MeOH, 96%; (e) 57, DMF, pyridine, 76%; (f) TFA, CH2Cl2,
90%; (g) CF3CO2C6F5, pyridine, CH2Cl2, 92%; (h) Et3N, DMF.
Further acylation of oligosaccharide 2-aminoethyl glycosides
with active ester 60 produced necessary biotin-tagged oligosac-
charides of general formula 61. Trisaccharides 32 and 33, their
disaccharide fragment 62,16 tetrasaccharide 47, pentasaccharide
fragment of HNK-1 glycolipids 63,16 its tetrasaccharide fragment
6424 devoid of the glucuronic acid, nonsulfated trisaccharide
39, 3-O-sulfoglucuronide 51, and trisaccharide 6517 containing
3,6-di-O-sulfoglucose instead of 3-O-sulfoglucuronic acid were
subjected to this transformation to provide ligands 66-74 for SPR
measurements (Table 1).
antibodies was studied: with HNK-1 antibody (mouse mono-
clonal IgMk) and HNK-1 412 antibody first known as L2
antibody (rat monoclonal IgG2ak). It is known that both
antibodies recognize specifically sulfated HNK-1 carbohy-
drate structures, while the HNK-1 412 antibody is also able
to recognize the nonsulfated epitope.25,26 We have shown that
monosaccharide 73, disaccharide 68, trisaccharides 66 and 67,
fucosylated tetrasaccharide 69, and pentasaccharide 70 bind
to both antibodies in a concentration-dependent manner,
whereas HNK-1 412 antibody interacts also with nonsulfated
trisaccharide 72. Typical sensorgrams of oligosaccharideꢀ
antibody interactions exemplified by trisaccharide 66 are given
in Figure 2 (for a complete set of sensorgrams of interactions
studied, please see Supporting Information [SI].). None of the
antibodies recognized tetrasaccharide 71 devoid of the uronic
acid and trisaccharide 74 in which 3-sulfated glucuronic acid is
replaced by 3,6-disulfated glucose.
1
The ligands were characterized by H NMR spectra, which
were essentially the superposition of the spectra of the corre-
sponding saccharide and acid 59, and HRMS data. Then
molecular recognition studies to determine the precise carbohy-
drate specificity of two different HNK-1 antigen-recognizing
antibodies by using of the synthesized oligosaccharide derivatives
and both SPR and NMR methods were performed.
2.3. SPR Investigation of the Interaction of the Oligosacchar-
ides with Monoclonal Antibodies. The interaction of synthetic
oligosaccharides with two different HNK-1 carbohydrate-specific
Two kinetic models were employed to calculate equilibrium
constants KD: (a) the Langmuir model (1:1 bimolecular interaction),
430
dx.doi.org/10.1021/ja2083015 |J. Am. Chem. Soc. 2012, 134, 426–435

Journal of the American Chemical Society
ARTICLE
Figure 2. Sensorgram of the binding of trisaccharide 66 to HNK-1 412 antibody (a) and HNK-1 antibody (b).
Figure 3. Fitting of the data of the interaction of trisaccharide 66 with HNK-1 412 (a, b) and HNK-1 (c, d) antibodies. Fitted data obtained using the
Langmuir model (a, c) and the two-state model (b, d) are represented by smooth curves.
and (b) the two-state model, which assumes that the ligandꢀ
analyte complex undergoes a conformational change. As it is
obvious from Figure 3, the two-state model provides better fit
of experimental and calculated curves especially in the case of
the HNK-1 412 antibody (a, b); the difference in the accuracy of
two kinetic models is less pronounced for the HNK-1 antibody
(c, d). These data show that the binding of carbohydrates to
HNK-1-related antibodies may be a more complicated process
than a simple 1:1 interaction. The bivalent analyte kinetic model
was tested as well, taking into account that HNK-1 412 antibody
has two binding sites and HNK-1 antibody, which is a hexamer,
has 12 binding sites. However, application of this model to both
antibodies showed even worse fitting than the Langmuir model
(data not shown). Equilibrium constants KD calculated using the
two-state model are given in the Table 2 (for complete kinetic
data, including association and dissociation constants, and stan-
dard errors, see Tables 1 and 2 in SI).
The KD(1) values of interactions with HNK-1 412 antibody
are approximately 2 orders of magnitude higher than those
with HNK-1 antibody (except nonsulfated structure 72), thus
indicating the higher affinity of the latter toward the carbo-
hydrate antigens studied. The minimum carbohydrate fragment
431
dx.doi.org/10.1021/ja2083015 |J. Am. Chem. Soc. 2012, 134, 426–435

Journal of the American Chemical Society
ARTICLE
recognizable by both antibodies was proven to be sulfated
glucuronic acid 73. The presence of the glucuronic acid is
the prerequisite for the recognition of an oligosaccharide by
both antibodies, whereas structures 71 and 74 devoid of this
monosaccharide showed no binding. The case of 74 is note-
worthy. Although both trisaccharides 66 and 74 are terminated
with the gluco-configured monosaccharides bearing two
negatively charged groups, only 66 is recognized by the anti-
bodies. Apꢀparently, the different spatial arrangement of the
weaker than the corresponding sulfated counterpart 66. In
other respects, the carbohydrate specificities of the antibodies
studied are similar, although the HNK-1 412 antibody seems
to be more sensitive to the structure of the carbohydrate
backbone.
Examination of binding abilities of HNK-1 antibodies to
linear structures revealed that both antibodies displayed the
weakest binding to sulfated uronic acid 73. Elongation of the
carbohydrate chain to di- (68), tri- (66), and pentasaccharide
(70) decreased KD(1) values in the case of HNK-1 412 by
almost 3 orders of magnitude. The contribution of the oligo-
saccharide chain to the binding was confirmed by STD NMR
investigation of a complex of pentasaccharide 63 with the
HNK-1 412 antibody (vide infra). The HNK-1 antibody
demonstrated the same trend within the series of monosacchar-
ide 73, disaccharide 68, and trisaccharide 66, although the
decrease in the KD(1) values was not as substantial as in the
case of the HNK-1 412 antibody. Unlike the HNK-1 412
antibody, weakening of the binding of pentasaccharide 70 as
compared to trisaccharide 66 was observed in the case of the
HNK-1 antibody. 6-O-Sulfatation (67) or 3-O-fucosylation (69)
in the N-acetylglucosamine residue of trisaccharide 66 caused
similar increase in the KD(1) values for both antibodies. The
KD(2) values characterizing a conformational change within an
antigenꢀantibody complex showed only slight variations and
ranged from 0.04 to 0.37.
ꢀ
CH2OSO3 and CO2 groups at C-5 of the terminal mono-
saccharides in 74 and 66 and different properties of the sulfate
and carboxylate groups, such as basicity and ability to bind
metal cations, are responsible for the dramatic difference in the
recognition of 66 and 74. Unlike the HNK-1 antibody that
recognizes only sulfated structures, the HNK-1 412 antibody
is able also to bind nonsulfated trisaccharide 72 although much
Table 2. Calculated (Two-State Kinetic Model) Equilibrium
Binding Constants of Oligosaccharides to HNK-1 412 and
HNK-1 Antibodies
HNK-1 412 antibody
HNK-1 antibody
KD(1) ꢁ 10ꢀ8
KD(2)
KD(1) ꢁ 10ꢀ10
KD(2)
oligosaccharide
(M)
(M)
(M)
(M)
66
67
68
69
70
72
73
0.371
1.45
0.37
0.10
0.12
0.10
0.37
0.05
0.07
0.490
2.04
8.92
0.15
0.12
0.06
0.04
0.10
2.4. STD NMR Investigation of a Complex of Pentasac-
charide 63 with the HNK-1 412 Antibody. Saturation transfer
difference (STD) NMR methods46 were then employed to
access residue-specific binding information. STD has widely
been applied to investigate ligandꢀreceptor biomolecular
interactions,47 including antigenꢀantibody recognition events.48
We exploited this methodology to explore, at atomic resolution,
10.9
7.25
34.2
1.61
no binding
12.8
0.153
22.2
91.5
0.06
Figure 4. 1H NMR (600 MHz) STD spectrum (saturation time 2 s with on-resonance irradiation at the aromatic region) of a 20:1 mixture of 63/HNK-
1 412 antibody at 310 K. The bottom trace shows the off-resonance spectrum, while the upper trace shows the STD signals. The region between 3.6 and
4.0 ppm is enlarged. Different responses from the different residues are observed, which were further analyzed with STD-TOCSY.
432
dx.doi.org/10.1021/ja2083015 |J. Am. Chem. Soc. 2012, 134, 426–435

Journal of the American Chemical Society
ARTICLE
Figure 5. NMR (500 MHz) regular 2D-TOCSY spectrum (left), and STD-2D-TOCSY spectrum (right), with saturation time 2 s, mixing time 60 ms,
and on-resonance irradiation at the aromatic region of a 20:1 mixture of 63/HNK-1 412 antibody at 310 K.
Blank experiments were performed for the pentasaccharide
63 in the absence of the HNK-1 412 antibody, as well as for
tetrasaccharide 64 devoid of the nonreducing sulfated GlcA unit
in the presence of the HNK-1 412 antibody. No STD effects were
observed in the first case, while very weak effects (maximum
0.4%) were observed for the tetrasaccharide 64, and only at a very
high ligand:antibody (200:1) molar ratio, indicating nonspecific
binding under these experimental conditions. No STD was
observed for 64 at a molar ratio of 20:1 employed for the study
of the interaction of pentasaccharide 63 with the antibody.
Moreover, DOSY experiments were also performed for both
63 and 64 oligosaccharides in the absence and in the presence
(5:1 molar ratio) of the HNK-1 412 antibody. No variation of the
diffusion coefficient of the sugar signals of 64 were observed
upon addition of the antibody, while large changes were evi-
denced for the pentasaccharide sample when the HNK-1 412
antibody was present.53 Thus, the NMR data nicely complement
the SPR results described above and provide a detailed structural
view of the interaction process. In fact, these experimental
approaches provide an alternative protocol to that previously
described54 using molecular modeling approaches to study the
recognition of carbohydrate HNK-1 antigens and their mimetics
by protein receptors.
Figure 6. Values of STD signals in the complex of pentasaccharide 63
with HNK-1 412 antibody.
the binding epitope of pentasaccharide 63 for the HNK-1 412
antibody. Various experimental conditions were employed to obtain
the best results.49 Finally, a sample containing HNK-1 412 antibody
(30 μM) in the presence of a 20 molar excess of 63 (600 μM) was
chosen. In the presence of the HNK-1 412 antibody, some of the
signals of 63 were broadened and, especially, H5 of the sulfated GlcA
moiety was significantly upfield shifted. The most intense STD signals
were observed at 310 K. The results (Figure 4) indicated the existence
of a clear binding epitope, since not all of the 1H NMR signals of the
pentasaccharide were displayed in the STD spectrum.50,51 Moreover,
stronger STD intensities were observed for the signals belonging to
the GlcA moiety than for the other residues.
At this stage, due to the heavy overlapping observed between
3.6 and 3.9 ppm, no fine details of the contributing moieties
could be obtained in an unambiguous manner. Therefore, a STD-
TOCSY experiment52 was performed to resolve the STD signals
in a second dimension (Figure 5). In this manner, a linear binding
epitope was clearly defined, starting from the sulfated GlcA unit,
which showed the major interaction with HNK-1 412 antibody,
followed by its vicinal Gal residue and the subsequent GlcNAc
moiety. The terminal lactose fragment provided very weak STD
effects, whereas the linker atoms gave no STD intensity at all,
indicating that they mainly remain outside of the binding site.
Thus, the major contributions to binding may be schematized as
displayed in Figure 6.
3. CONCLUSION
To gain insights into the specificity of HNK-1 carbohydrate
binding antibodies, three sulfated and one nonsulfated oligosac-
charides related to the HNK-1 antigen were synthesized and
converted, together with some other oligosaccharides of this
type, to biotin-tagged molecular probes. They were used to
determine the carbohydrate specificity of HNK-1 412 and HNK-
1 monoclonal antibodies by SPR. It has been shown that the
presence of 3-O-sulfated glucuronic acid is a decisive feature for
the binding of oligosaccharides to both antibodies. Prior to this
study, a little was known about the role of further monosaccha-
rides of the HNK-1 glycan chain in the binding. Here we have
433
dx.doi.org/10.1021/ja2083015 |J. Am. Chem. Soc. 2012, 134, 426–435

Journal of the American Chemical Society
ARTICLE
shown that the monosaccharides of the contiguous N-acetyllac-
tosamine moiety are also involved in the molecular recognition
process of the HNK-1 glycan by the antigen-binding sites of
the studied antibodies. In contrast, the lactose fragment at the
“reducing end” of the HNK-1 pentasaccharide only barely
interacts with the carbohydrate epitope binding site. This fact
was clearly evidenced from the STD NMR study of the HNK-1
pentasaccharideꢀHNK-1 412 antibody complex.
0.5, 1, and 2 s saturation times (by concatenation of 50-ms Gaussian
pulses separated by 1 ms), whereas the STD TOCSY experiment was
performed only with 2 s. In all cases, the on-resonance frequency was
adjusted at 6.8 ppm.
’ ASSOCIATED CONTENT
S
Supporting Information. Synthetic protocols used to
b
prepare oligosaccharides and their biotin-tagged derivatives.
Sensorgrams of interaction of biotin-tagged oligosaccharides
with mAbs and kinetic data. Copies of NMR spectra. Complete
references 49 and 54. This material is available free of charge via
the Internet at http://pubs.acs.org.
Our results illustrate that HNK-1 antigenic trisaccharide or
longer chains are preferentially detected by the examined HNK-1
antibodies. The use of synthesized HNK-1 related structures
alone or in combination with carbohydrate ligands of other types
will allow one in the future and after necessary validation studies
to screen autoantibodies in the series of patients with polyneuro-
pathies in order to identify different types of this disease and to
develop new therapeutic strategies using selected HNK-1 struc-
tures for neutralization of HNK-1 autoantibodies in the affected
patients. It should be also pointed out that HNK-1 antigenic
carbohydrate ligands are carried by several adhesion mole-
cules55,56 that determine their ability to bind to the neurite out-
growth promoting the extracellular matrix molecule laminin.57
Thus, the described results can also form the rationale for the
selection of optimal oligosaccharides for enhancing peripheral
nerve regeneration and GABAergic synaptic transmission, the
importance of which is implicated in down-tuning of the central
nervous system overactivity, as seen in epilepsy. Initial studies in
this direction have been started by us58 previously and will be
continued with the use of the glycoconjugate tools described here.
’ AUTHOR INFORMATION
Corresponding Author
nen@ioc.ac.ru
’ ACKNOWLEDGMENT
This work was supported in part by collaborative grants RUS
09/A03 and RUS 10/011 from German Bundesministerium f€ur
Bildung und Forschung. The group at Madrid thanks the
Ministry of Science and Innovation of Spain for funding
(CTQ2009-08536; J.J.-B.). The group in Moscow thanks the
Division of Chemistry and Materials Science of the RAS for the
funding within Program 9 (N.E.N.), and the grant program of the
President of the Russian Federation to Young Scientists (Grant
MK-5544.2010.3 to A.A.G.). M.S. is supported by the New Jersey
Commission for Spinal Cord Research.
4. EXPERIMENTAL SECTION
For general analytical procedures and synthetic protocols used to
prepare oligosaccharides and their biotin-tagged derivatives, please see
the SI.
4.1. Antibodies. Both monoclonal anti-HNK-1 antibodies (412
and HNK-1) were prepared as described.55 The purified IgG or IgM
fractions, respectively, were used in SPR studies.
’ REFERENCES
(1) Chou, D. K. H.; Ilyas, A. A.; Evans, J. E.; Costello, C.; Quarles,
R. H.; Jungalwala, F. B. J. Biol. Chem. 1986, 261, 11717–11725.
(2) Voshol, H.; van Zuylen, C. W. E. M.; Orberger, G.; Vliegenthart,
J. F. G.; Schachner, M. J. Biol. Chem. 1996, 271, 22957–22960.
(3) Martini, R.; Xin, Y.; Schmitz, B.; Schachner, M. Eur. J. Neurosci.
1992, 4, 628–639.
(4) Martini, R.; Schachner, M.; Brushart, T. M. J. Neurosci. 1994, 14,
7180–7191.
(5) Schachner, M.; Martini, R.; Hall, H.; Orberger, G. Prog. Brain Res.
1995, 105, 183–188.
(6) Eberhardt, K. A.; Irintchev, A.; Al-Majed, A. A.; Simova, O.; Brushart,
T. M.; Gordon, T.; Schachner, M. Exp. Neurol. 2006, 198, 500–510.
(7) Simova, O.; Irintchev, A.; Mehanna, A.; Liu, J.; Dihnꢀe, M.;
B€achle, D.; Sewald, N.; Loers, G.; Schachner, M. Ann. Neurol. 2006,
60, 430–437.
(8) Irintchev, A.; Lee, H.; Wu, M. M.; Zhu, H.; Feng, Y. P.; Liu, Y. S.;
Bernreuther, C.; Loers, G.; You, S. W.; Schachner, M. Mol. Ther. 2011in
press.
(9) Kleene, R.; Schachner, M. Nat. Rev. Neurosci. 2004, 5, 195–208.
(10) Nakano, T.; Ito, Y.; Ogawa, T. Carbohydr. Res. 1993, 243,
43–69.
(11) Isogai, Y.; Ishida, H.; Kiso, M.; Hasegawa, A. J. Carbohydr.
Chem. 1996, 15, 1001–1026.
(12) Isogai, Y.; Ishida, H.; Kiso, M.; Hasegawa, A. J. Carbohydr.
Chem. 1996, 15, 1119–1137.
(13) Chevalier, R.; Colsch, B.; Afonso, C.; Baumann, N.; Tabet,
J.-C.; Mallet, J.-M. Tetrahedron 2006, 62, 563–577.
(14) Ding, Y.; Fukuda, M.; Hindsgaul, O. Bioorg. Med. Chem. Lett.
1998, 8, 1903–1908.
(15) Kononov, L. O.; Kornilov, A. V.; Sherman, A. A.; Zyryanov,
E. V.; Zatonsky, G. V.; Shashkov, A. S.; Nifantiev, N. E. Russ. J. Bioorg.
Chem. 1998, 24, 537–550.
4.2. SPR Measurements. SPR experiments were performed using
a ProteOn XPR-36 instrument (Bio Rad). PBST (phosphate-buffered
saline, 10 mM phosphate, 150 mM NaCl, pH 7.40 containing 0.005%
Tween 20) was used as a running buffer. The biotinylated oligosaccha-
rides (150 μL of 0.5ꢀ100 nM solutions in PBST depending on the
saccharide structure) were immobilized at a flow rate of 30 μL/min on
neutravidin-coated NLC-chips. One horizontal ligand channel was
employed per oligosaccharide. One of the six horizontal channels was
loaded with biotin (150 μL of 10 μM solution in PBST) and used as a
reference channel. Interaction was measured by running the antibody
solutions vertically over the immobilized carbohydrate structures using
one analyte channel per antibody concentration (200 μL of antibody
solution in PBST, flow rate 100 μL/min, duration of the association step
120 s, duration of the dissociation step 400ꢀ600 s). Four concentrations
were measured for the calculation of kinetic constants. After each
analysis, the surface was washed with a regeneration buffer (10 mM
NaOH) and equilibrated with the running buffer. Data processing was
performed with the ProteOn Manager software.
STD NMR Experiments. These spectra were recorded at 310 K in a
phosphate buffer (D2O) at pH 7.3, uncorrected for isotope effects, on
Bruker AVANCE 500- and 600 MHz spectrometers equipped with a
triple-channel cryoprobe, as described.47 The STD (10k scans) and STD
TOCSY experiments (mixing time 60 ms with 256 increments of 80
scans each) were performed as described, using a 20:1 ligand to receptor
molar ratio with a 600 μM concentration of oligosaccharides in Shigemi
NMR tubes. One-dimensional STD experiments were performed with
434
dx.doi.org/10.1021/ja2083015 |J. Am. Chem. Soc. 2012, 134, 426–435

Journal of the American Chemical Society
ARTICLE
(16) Kornilov, A. V.; Sherman, A. A.; Kononov, L. O.; Shashkov,
A. S.; Nifantiev, N. E. Carbohydr. Res. 2000, 329, 717–730.
(17) Sukhova, E. V.; Dubrovskii, A. V.; Tsvetkov, Y. E.; Nifantiev,
N. E. Russ. Chem. Bull. 2007, 56, 1655–1670.
(18) Bastide, L.; Priem, B.; Fort, S. Carbohydr. Res. 2011, 346, 348–351.
(19) Gutiꢀerrez Gallego, R.; Jimꢀenez Blanco, J.; Thijssen-Zuylen, C.
W. E. M.; Gotfredsen, C. H.; Voshol, H.; Duus, J. Ø.; Schachner, M.;
Vliegenthart, J. F. G. J. Biol. Chem. 2001, 276, 30834–30844.
(20) Barthel, S. R.; Wiese, G. K.; Cho, J.; Opperman, M. J.; Hays,
D. L.; Siddiqui, J.; Pienta, K. J.; Furie, B.; Dimitroff, C. J. Proc. Natl. Acad.
Sci. U.S.A. 2009, 106, 19491–19496.
(50) Mayer, M.; Meyer, B. J. Am. Chem. Soc. 2001, 123, 6108–6111.
(51) Bernardi, A.; Arosio, D.; Potenza, D.; Sanchez-Medina, I.; Mari, S.;
Ca~nada, F. J.; Jimenez-Barbero, J. Chem.—Eur. J. 2004, 10, 4395–4405.
(52) Johnson, M. A.; Pinto, B. M. Carbohydr. Res. 2004, 339,
907–928.
(53) Groves, P.; Rasmussen, M.; Molero, M. D.; Samain, E.; Ca~nada,
F. J.; Driguez, H.; Jimꢀenez-Barbero, J. Glycobiology 2004, 14, 451–456.
(54) Bhunia, A.; et al. J. Am. Chem. Soc. 2010, 132, 96–105.
(55) Kruse, J.; Mailhammer, R.; Wernecke, H.; Faissner, A.; Sommer, I.;
Goridis, C.; Schachner, M. Nature 1984, 311, 153–155.
(56) Voshol, H.; van Zuylen, C. W.; Orberger, G.; Vliegenthart, J. F.;
(21) Gooi, H. C.; Feizi, T.; Kapadia, A.; Knowles, B. B.; Solter, D.;
Evans, M. J. Nature 1981, 292, 156–158.
(22) Andressen, C.; Arnhold, S.; Mai, J. K. Anat. Embryol. 1997, 197,
209–215.
(23) Wang, J.-W.; Ambros, R. A.; Weber, P. B.; Rosano, T. G. Cancer
Res. 1995, 55, 3654–3658.
Schachner, M. J. Biol. Chem. 1996, 271, 22957–22960.
(57) Hall, H.; Deutzmann, R.; Timpl, R.; Vaughan, L.; Schmitz, B.;
Schachner, M. Eur. J. Biochem. 1997, 246, 233–242.
(58) Saghatelyan, A. K.; Gorissen, S.; Meigel, I.; Mosbacher, J.;
Kaupmann, K.; Bettler, B.; Kornilov, A. V.; Nifantiev, N. E.; Schachner,
M.; Dityatev, A. Mol. Cell. Neurosci. 2003, 24, 271–282.
(24) Sherman, A. A.; Yudina, O. N.; Mironov, Y. V.; Sukhova, E. V.;
Shashkov, A. S.; Menshov, V. M.; Nifantiev, N. E. Carbohydr. Res. 2001,
336, 13–46.
(25) Schmitz, B.; Schachner, M.; Ito, Y.; Nakano, T.; Ogawa, T.
Glycocon. J. 1994, 11, 345–352.
(26) Tokuda, A.; Ariga, T.; Isogai, Y.; Komba, S.; Kiso, M.; Hasegawa,
A.; Tai, T.; Yu, R. K. J. Carbohydr. Chem. 1998, 17, 535–546.
(27) Higson, A. P.; Tsvetkov, Y. E.; Ferguson, M. A. J.; Nikolaev,
A. V. J. Chem. Soc., Perkin Trans. 1 1998, 2587–2596.
(28) Kornilov, A. V.; Sukhova, E. V.; Nifantiev, N. E. Carbohydr. Res.
2001, 336, 309–313.
(29) Garegg, P. J.; Kvarnstr€om, I.; Niklasson, A.; Niklasson, G.;
Svensson, S. C. T. J. Carbohydr. Chem. 1993, 12, 933–954.
(30) Kornilov, A. V.; Kononov, L. O.; Zatonskii, G. V.; Shashkov,
A. S.; Nifantiev, N. E. Russ. J. Bioorg. Chem. 1997, 23, 597–607.
(31) Liao, L.; Auzanneau, F.-I. Org. Lett. 2003, 5, 2607–2610.
(32) Liao, L.; Auzanneau, F.-I. J. Org. Chem. 2005, 70, 6265–6273.
(33) Byramova, N. E.; Ovchinnikov, M. V.; Backinowsky, L. V.;
Kochetkov, N. K. Carbohydr. Res. 1983, 124, C8–C11.
(34) Mukherjee, A.; Palcic, M. M.; Hindsgaul, O. Carbohydr. Res.
2000, 326, 1–21.
(35) Duverger, E.; Frison, N.; Roche, A.-C.; Monsigny, M. Biochimie
2003, 85, 167–179.
(36) Kobayashi, Y.; Nakamura, H.; Sekiguchi, T.; Takanami, R.;
Murata, T.; Usui, T.; Kawagushi, H. Anal. Biochem. 2005, 336, 87–93.
(37) Suda, Y.; Arano, A.; Fukui, Y.; Koshida, S.; Wakao, M.; Nashimura,
T.; Kusumoto, S.; Sobel, M. Bioconjugate Chem. 2006, 17, 1125–1135.
(38) Murthy, B. N.; Voelcker, N. H.; Jayaraman, N. Glycobiology
2006, 16, 822–832.
(39) Collot, M.; Sendid, B.; Fievez, A.; Savaux, C.; Standaert-Vitse,
A.; Tabouret, M.; Drucbert, A. S.; Danzꢀe, P. M.; Poulain, D.; Mallet, J.-M.
J. Med. Chem. 2008, 51, 6201–6210.
(40) Linman, M. J.; Taylor, J. D.; Yu, H.; Chen, X.; Cheng, Q. Anal.
Chem. 2008, 80, 4007–4013 and references sited therein.
(41) Mann, D. A.; Kanai, M.; Maly, D. J.; Kiessling, L. L. J. Am. Chem.
Soc. 1998, 120, 10575–10582.
(42) Liang, P.-H.; Wang, S.-K.; Wong, C.-H. J. Am. Chem. Soc. 2007,
129, 11177–11184.
(43) Zhu, J.; Wan, Q.; Danishefsky, S. J. Tetrahedron Lett. 2009,
50, 712–714.
(44) Karamanska, R.; Clarke, J.; Blixt, O.; MacRae, J. I.; Zhang, J. Q.;
Crocker, P. R.; Laurent, N.; Wright, A.; Flitsch, S. L.; Russel, D. A.; Field,
R. A. Glycoconjugate J. 2008, 25, 69–74.
(45) Herzner, H.; Kunz, H. Carbohydr. Res. 2007, 342, 541–557.
(46) Meyer, B.; Peters, T. Angew. Chem., Int. Ed. 2003, 42, 864–890.
(47) Canales, A.; Matesanz, R.; Gardner, N.; Andreu, J. M.; Paterson,
I.; Díaz, F.; Jimꢀenez-Barbero, J. Chem.—Eur. J. 2008, 14, 7557–69.
(48) Oberli, M. A.; Tamborrini, M.; Tsai, Y.-H.; Werz, D. B.;
Horlacher, T.; Adibekian, A.; Gauss, D.; Moller, H. M.; Pluschke, G.;
Seeberger, P. H. J. Am. Chem. Soc. 2010, 132, 10239–10241.
(49) Groves, P.; et al. Magn. Reson. Chem. 2007, 45, 745–748.
435
dx.doi.org/10.1021/ja2083015 |J. Am. Chem. Soc. 2012, 134, 426–435